Ohad Eytan b5d670498f partners/milvus: allow creating a vectorstore with sparse embeddings (#25284) ...

# Description
Milvus (and `pymilvus`) recently added the option to use [sparse
vectors](https://milvus.io/docs/sparse_vector.md#Sparse-Vector) with
appropriate search methods (e.g., `SPARSE_INVERTED_INDEX`) and
embeddings (e.g., `BM25`, `SPLADE`).

This PR allow creating a vector store using langchain's `Milvus` class,
setting the matching vector field type to `DataType.SPARSE_FLOAT_VECTOR`
and the default index type to `SPARSE_INVERTED_INDEX`.

It is only extending functionality, and backward compatible. 

## Note
I also interested in extending the Milvus class further to support multi
vector search (aka hybrid search). Will be happy to discuss that. See
[here](https://github.com/langchain-ai/langchain/discussions/19955),
[here](https://github.com/langchain-ai/langchain/pull/20375), and
[here](https://github.com/langchain-ai/langchain/discussions/22886)
similar needs.

---------

Co-authored-by: Erick Friis <erick@langchain.dev>

2024-08-30 02:30:23 +00:00

..

integration_tests

partners/milvus: allow creating a vectorstore with sparse embeddings (#25284)

2024-08-30 02:30:23 +00:00

unit_tests

milvus: New langchain_milvus package and new milvus features (#21077)

2024-05-28 08:24:20 -07:00

__init__.py

milvus: New langchain_milvus package and new milvus features (#21077)

2024-05-28 08:24:20 -07:00