BM25S retriever #30815
hackerdjaKODJO
started this conversation in
General
BM25S retriever
#30815
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
BM25S retriever
I wonder if Langchain currently supports BM25S, an ultrafast Python implementation of BM25 that leverages Scipy sparse matrices for efficient text retrieval. BM25S significantly outperforms other BM25 implementations in terms of speed and memory usage, and it offers a simple and flexible API, supports various BM25 variants, and integrates seamlessly with Hugging Face for model sharing and community use. If BM25S is not yet supported, I would like to request its inclusion as a feature in Langchain. The integration of BM25S would greatly enhance the performance and flexibility of text retrieval tasks within Langchain.
[Langchain BM25 retriever]
BM25 (Wikipedia) also known as the Okapi BM25, is a ranking function used in information retrieval systems to estimate the relevance of documents to a given search query.
[BM25S]
BM25S is an ultrafast Python implementation of BM25, leveraging Scipy sparse matrices for efficient text retrieval, significantly outperforming other BM25 implementations in terms of speed and memory usage.
It offers a simple and flexible API, supports various BM25 variants, and integrates seamlessly with Hugging Face for model sharing and community use.
Beta Was this translation helpful? Give feedback.
All reactions