Rank phrases and text based on query by leveraging hugging-face models.
Project description
spacy-rerank
Rank phrases and text based on query by leveraging hugging-face models. Currently, we are only leveraging tiny-bert model.
Installation and Implementation
- Install the package using the below
pip install spacyrerank
- Code for simple implementation
from spacyrerank.rerank import Reranker
query = "done effort is wasted"
texts = ["work done", "valuable man", "effort wasted", "Great work", "great work mate"]
reranker = Reranker(query, texts=texts)
reranker()
[{'rank': 2, 'text': 'effort wasted', 'similarity-score': 0.877},
{'rank': 0, 'text': 'work done', 'similarity-score': 0.86},
{'rank': 3, 'text': 'Great work', 'similarity-score': 0.773},
{'rank': 4, 'text': 'great work mate', 'similarity-score': 0.757},
{'rank': 1, 'text': 'valuable man', 'similarity-score': 0.719}]
CPU times: user 331 ms, sys: 76 ms, total: 407 ms
Wall time: 564 ms
Project details
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
spacyrerank-0.0.5.tar.gz
(3.6 kB
view hashes)
Built Distribution
Close
Hashes for spacyrerank-0.0.5-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 56f27ffacf33cc1a2ef937dff32c495e50bb405aaa22d5233b34d26dfc6dee8b |
|
MD5 | e13f8996f48e375fae73f391593a2e8b |
|
BLAKE2b-256 | 4b0b8325e1409c07685b3caeda421774917fe6ed587572aa5832ae6a46cf9ac8 |