Sentence similarity python

# sentence_transformer: Python module using Sentence BERT 
# check documentation in source link for further details
from sentence_transformers import SentenceTransformer
from sentence_transformers.util import pytorch_cos_sim # cosine silarity

# stsb-roberta-large is one of the possible pre-trained models
model = SentenceTransformer('stsb-roberta-large')

# list of sentences (type: str)
sentences = list(...) 

# calculate the embeddings
embeddings = bert_model.encode(sentences)

# example: cosine similarity among first and second sentences
cos_similarity = pytorch_cos_sim(embeddings[0], embeddings[1])

Source: sbert.net

Similar codes

2 votes318 views4 answers
ValueError: Input contains NaN, infinity or a value too large for dtype('float32').
Python
0 vote322 views1 answer
Bokeh xlabel rotate
Python
0 vote309 views1 answer
Deep copy a dataframe
Python
0 vote293 views1 answer
Np.linspace() in python
Python
1 vote308 views2 answers
Python conditional operator one line
Python
-1 vote294 views1 answer
Pytube3
Python
0 vote325 views1 answer
Exec eval python
Python
18 votes313 views5 answers
Continue statement in python
Python
3 votes384 views3 answers
Tensorflow to numpy
Python
0 vote308 views1 answer
\n appears in json dump
Python