FastEmbedSparse

class langchain_qdrant.fastembed_sparse.FastEmbedSparse(model_name: str = 'Qdrant/bm25', batch_size: int = 256, cache_dir: str | None = None, threads: int | None = None, providers: Sequence[Any] | None = None, parallel: int | None = None, **kwargs: Any)

An interface for sparse embedding models to use with Qdrant.

Sparse encoder implementation using FastEmbed - https://qdrant.github.io/fastembed/ For a list of available models, see https://qdrant.github.io/fastembed/examples/Supported_Models/

Parameters:

• model_name (str) – The name of the model to use. Defaults to “Qdrant/bm25”.

• batch_size (int) – Batch size for encoding. Defaults to 256.

• cache_dir (str, optional) – The path to the model cache directory. Can also be set using the FASTEMBED_CACHE_PATH env variable.

• threads (int, optional) – The number of threads onnxruntime session can use.

• providers (Sequence[Any], optional) – List of ONNX execution providers. parallel (int, optional): If >1, data-parallel encoding will be used, r Recommended for encoding of large datasets. If 0, use all available cores. If None, don’t use data-parallel processing, use default onnxruntime threading instead. Defaults to None.

• kwargs (Any) – Additional options to pass to fastembed.SparseTextEmbedding

• parallel (int | None)

Raises:

ValueError – If the model_name is not supported in SparseTextEmbedding.

Methods

__init__([model_name, batch_size, ...])	Sparse encoder implementation using FastEmbed - https://qdrant.github.io/fastembed/ For a list of available models, see https://qdrant.github.io/fastembed/examples/Supported_Models/
aembed_documents(texts)	Asynchronous Embed search docs.
aembed_query(text)	Asynchronous Embed query text.
embed_documents(texts)	Embed search docs.
embed_query(text)	Embed query text.

__init__(model_name: str = 'Qdrant/bm25', batch_size: int = 256, cache_dir: str | None = None, threads: int | None = None, providers: Sequence[Any] | None = None, parallel: int | None = None, **kwargs: Any) → None

Sparse encoder implementation using FastEmbed - https://qdrant.github.io/fastembed/ For a list of available models, see https://qdrant.github.io/fastembed/examples/Supported_Models/

Parameters:

• model_name (str) – The name of the model to use. Defaults to “Qdrant/bm25”.

• batch_size (int) – Batch size for encoding. Defaults to 256.

• cache_dir (str, optional) – The path to the model cache directory. Can also be set using the FASTEMBED_CACHE_PATH env variable.

• threads (int, optional) – The number of threads onnxruntime session can use.

• providers (Sequence[Any], optional) – List of ONNX execution providers. parallel (int, optional): If >1, data-parallel encoding will be used, r Recommended for encoding of large datasets. If 0, use all available cores. If None, don’t use data-parallel processing, use default onnxruntime threading instead. Defaults to None.

• kwargs (Any) – Additional options to pass to fastembed.SparseTextEmbedding

• parallel (int | None)

Raises:

ValueError – If the model_name is not supported in SparseTextEmbedding.

Return type:

None

async aembed_documents(texts: List[str]) → List[SparseVector]

Asynchronous Embed search docs.

Parameters:

texts (List[str])

Return type:

List[SparseVector]

async aembed_query(text: str) → SparseVector

Asynchronous Embed query text.

Parameters:

text (str)

Return type:

SparseVector

embed_documents(texts: List[str]) → List[SparseVector]

Embed search docs.

Parameters:

texts (List[str])

Return type:

List[SparseVector]

embed_query(text: str) → SparseVector

Embed query text.

Parameters:

text (str)

Return type:

SparseVector