QdrantSparseEmbeddingRetriever

Retrieve documents from a QdrantDocumentStore using sparse vector embeddings.

Key Features

• Sparse vector-based retrieval from a Qdrant vector database.
• Configurable number of results with top_k and minimum score threshold.
• Supports metadata filtering to narrow down the search space.
• Supports document grouping by a payload field.
• Optional score scaling and sparse embedding return.

Configuration

1. Drag the QdrantSparseEmbeddingRetriever component onto the canvas from the Component Library.
2. Click on the component to open the configuration panel.
3. On the General tab:
   • Configure the QdrantDocumentStore with your Qdrant instance details. Make sure your Qdrant index uses sparse embeddings (use_sparse_embeddings=True).
   • Set top_k to control the maximum number of documents to retrieve.
4. Go to the Advanced tab to configure filter_policy, score_threshold, scale_score, return_embedding, group_by, and group_size.

Connections

QdrantSparseEmbeddingRetriever receives sparse query embeddings from a sparse text embedder. It sends retrieved documents to downstream components such as PromptBuilder or a ranker.

Source Code

To check this component's source code, open retriever.py in the Haystack Core Integrations repository.

Usage Examples

Basic Configuration

  QdrantSparseEmbeddingRetriever:
    type: qdrant.src.haystack_integrations.components.retrievers.qdrant.retriever.QdrantSparseEmbeddingRetriever
    init_parameters: {}

components:
  QdrantSparseEmbeddingRetriever:
    type: qdrant.src.haystack_integrations.components.retrievers.qdrant.retriever.QdrantSparseEmbeddingRetriever
    init_parameters:

Parameters

Inputs

Parameter	Type	Description
query_sparse_embedding	SparseEmbedding	Sparse Embedding of the query.
filters	Optional[Union[Dict[str, Any], models.Filter]]	Filters applied to the retrieved Documents. The way runtime filters are applied depends on the filter_policy chosen at retriever initialization.
top_k	Optional[int]	The maximum number of documents to return. If using group_by parameters, maximum number of groups to return.
scale_score	Optional[bool]	Whether to scale the scores of the retrieved documents or not.
return_embedding	Optional[bool]	Whether to return the embedding of the retrieved Documents.
score_threshold	Optional[float]	A minimal score threshold for the result. Score of the returned result might be higher or smaller than the threshold depending on the Distance function used.
group_by	Optional[str]	Payload field to group by, must be a string or number field. If the field contains more than one value, all values will be used for grouping. One point can be in multiple groups.
group_size	Optional[int]	Maximum amount of points to return per group. Default is 3.

Outputs

Parameter	Type	Description
documents	List[Document]	The retrieved documents.

Init Parameters

These are the parameters you can configure in Pipeline Builder:

Parameter	Type	Default	Description
document_store	QdrantDocumentStore		An instance of QdrantDocumentStore.
filters	Optional[Union[Dict[str, Any], models.Filter]]	None	A dictionary with filters to narrow down the search space.
top_k	int	10	The maximum number of documents to retrieve. If using group_by parameters, maximum number of groups to return.
scale_score	bool	False	Whether to scale the scores of the retrieved documents or not.
return_embedding	bool	False	Whether to return the sparse embedding of the retrieved Documents.
filter_policy	Union[str, FilterPolicy]	FilterPolicy.REPLACE	Policy to determine how filters are applied. Defaults to "replace".
score_threshold	Optional[float]	None	A minimal score threshold for the result. Score of the returned result might be higher or smaller than the threshold depending on the Distance function used. For example, for cosine similarity only higher scores will be returned.
group_by	Optional[str]	None	Payload field to group by, must be a string or number field. If the field contains more than one value, all values will be used for grouping. One point can be in multiple groups.
group_size	Optional[int]	None	Maximum amount of points to return per group. Default is 3.

Run Method Parameters

These are the parameters you can configure for the component's run() method. This means you can pass these parameters at query time through the API, in Playground, or when running a job. For details, see Modify Pipeline Parameters at Query Time.

Parameter	Type	Default	Description
query_sparse_embedding	SparseEmbedding		Sparse Embedding of the query.
filters	Optional[Union[Dict[str, Any], models.Filter]]	None	Filters applied to the retrieved Documents. The way runtime filters are applied depends on the filter_policy chosen at retriever initialization. See init method docstring for more details.
top_k	Optional[int]	None	The maximum number of documents to return. If using group_by parameters, maximum number of groups to return.
scale_score	Optional[bool]	None	Whether to scale the scores of the retrieved documents or not.
return_embedding	Optional[bool]	None	Whether to return the embedding of the retrieved Documents.
score_threshold	Optional[float]	None	A minimal score threshold for the result. Score of the returned result might be higher or smaller than the threshold depending on the Distance function used. For example, for cosine similarity only higher scores will be returned.
group_by	Optional[str]	None	Payload field to group by, must be a string or number field. If the field contains more than one value, all values will be used for grouping. One point can be in multiple groups.
group_size	Optional[int]	None	Maximum amount of points to return per group. Default is 3.