Metadata-Version: 2.1
Name: nk-sent2vec
Version: 1.4.2
Summary: Embeds text documents using sent2vec
Home-page: UNKNOWN
Author: New Knowledge
License: BSD-3-Clause
Platform: UNKNOWN
Description-Content-Type: text/markdown
Requires-Dist: Cython (==0.29.24)
Requires-Dist: numpy (>=1.15.4)
Requires-Dist: pytest
Requires-Dist: sent2vec (<0.3,>=0.2)

# Sentence Embedding
A python wrapper for embedding short texts or sentences using [sent2vec](https://github.com/epfml/sent2vec), which draws on FastText.

To embed a list of strings `documents`, use:

```
from nk_sent2vec import Sent2Vec 

vectorizer = Sent2Vec(path = '/root/models/torontobooks_unigrams.bin')

print(vectorizer.embed_sentences(sentences=documents))
```

## Testing
Tests can be run using `pytest -s tests`

Also see `makefile` for default commands


