How to represent sentences in a high-dimensional vector space, a.k.a., sentence embedding.
A Generic Sentence Embedding Library
In natural language processing, we need to vectorize or encode text data to let machine proocess it. In the past, we mostly use encoders such as one-hot, term frequency, and TF-IDF (normalized term frequency). There are many challenges with these techniques that you can read here. In the recent years, the deep learning advancements give us opportunity to encode sentences or words in more meaningful format. The word2vec library was one of the major advancement in this field. Or, the BERT language model provides us a powerful sentence encoders that can be used in many projects.
The sentence embedding or encoding is an important step of many NLP projects. Plus, we believe that a flexible sent2vec library is needed to build a prototype fast. That is why we have initiated this project. In the early releases, you will have access to the standard encoders. We will add more curated techniques in the later releases. Hope you can use this library in your exciting NLP projects.
The library requires the following libraries:
It can be installed using pip:
pip install sent2vec
This is how to initialize the library and provide the data.
from sent2vec.vectorizer import Vectorizer
sentences = [
"This is an awesome book to learn NLP.",
"DistilBERT is an amazing NLP library.",
"We can interchangeably use embedding, encoding, or vectorizing.",
vectorizer = Vectorizer(sentences)
If you want to use the pre-trained DistilBertModel, you must use the code below.
vectors = vectorizer.sent2vec_bert()
Now, you can compute distance between
vectors as well as distance between
vectors. You can see that the third sentence is different from other sentences.
And, that's pretty much it!
Release history Release notifications | RSS feed
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.