TransformersSimilarityRanker

Rank documents based on their semantic similarity to the query.

Legacy Component

This component is considered legacy and will no longer receive updates. It may be deprecated in a future release, with removal following after a deprecation period. Consider using SentenceTransformersSimilarityRanker instead, which provides the same functionality along with additional features.

Key Features

• Uses a pre-trained cross-encoder model from Hugging Face to rank documents by semantic similarity to the query.
• Configurable number of results returned via top_k.
• Supports score scaling via Sigmoid activation for normalized similarity scores.
• Supports calibration of probabilities via calibration_factor.
• Supports filtering by score threshold.

Configuration

1. Drag the TransformersSimilarityRanker component onto the canvas from the Component Library.
2. Click on the component to open the configuration panel.
3. On the General tab:
   • Set the ranking model. Pass a local path or the Hugging Face model name of a cross-encoder model.
   • Set top_k to control how many documents to return.
4. Go to the Advanced tab to configure scale_score, calibration_factor, score_threshold, and batch_size.

Connections

TransformersSimilarityRanker accepts a query string and a list of documents as inputs. Connect it after a retriever or DocumentJoiner in a query pipeline.

It outputs a ranked list of documents sorted from most to least relevant to the query. Connect its documents output to ChatPromptBuilder, AnswerBuilder, or another downstream component.

Source Code

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

Usage Examples

Basic Configuration

  TransformersSimilarityRanker:
    type: components.rankers.transformers_similarity.TransformersSimilarityRanker
    init_parameters: {}

components:
  TransformersSimilarityRanker:
    type: components.rankers.transformers_similarity.TransformersSimilarityRanker
    init_parameters:

Parameters

Inputs

Parameter	Type	Default	Description
query	str		The input query to compare the documents to.
documents	List[Document]		A list of documents to be ranked.
top_k	Optional[int]	None	The maximum number of documents to return.
scale_score	Optional[bool]	None	If True, scales the raw logit predictions using a Sigmoid activation function. If False, disables scaling.
calibration_factor	Optional[float]	None	Use this factor to calibrate probabilities with sigmoid(logits * calibration_factor). Used only if scale_score is True.
score_threshold	Optional[float]	None	Return documents only with a score above this threshold.

Outputs

Parameter	Type	Default	Description
documents	List[Document]		A list of documents closest to the query, sorted from most similar to least similar.

Init Parameters

These are the parameters you can configure in Pipeline Builder:

Parameter	Type	Default	Description
model	Union[str, Path]	cross-encoder/ms-marco-MiniLM-L-6-v2	The ranking model. Pass a local path or the Hugging Face model name of a cross-encoder model.
device	Optional[ComponentDevice]	None	The device on which the model is loaded. If None, overrides the default device.
token	Optional[Secret]	Secret.from_env_var(['HF_API_TOKEN', 'HF_TOKEN'], strict=False)	The API token to download private models from Hugging Face.
top_k	int	10	The maximum number of documents to return per query.
query_prefix	str		A string to add at the beginning of the query text before ranking. Use it to prepend the text with an instruction, as required by reranking models like bge.
document_prefix	str		A string to add at the beginning of each document before ranking.
meta_fields_to_embed	Optional[List[str]]	None	List of metadata fields to embed with the document.
embedding_separator	str	\n	Separator to concatenate metadata fields to the document.
scale_score	bool	True	If True, scales the raw logit predictions using a Sigmoid activation function.
calibration_factor	Optional[float]	1.0	Use this factor to calibrate probabilities with sigmoid(logits * calibration_factor). Used only if scale_score is True.
score_threshold	Optional[float]	None	Return documents with a score above this threshold only.
model_kwargs	Optional[Dict[str, Any]]	None	Additional keyword arguments for AutoModelForSequenceClassification.from_pretrained when loading the model.
tokenizer_kwargs	Optional[Dict[str, Any]]	None	Additional keyword arguments for AutoTokenizer.from_pretrained when loading the tokenizer.
batch_size	int	16	The batch size to use for inference. The higher the batch size, the more memory is required.

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	str		The input query to compare the documents to.
documents	List[Document]		A list of documents to be ranked.
top_k	Optional[int]	None	The maximum number of documents to return.
scale_score	Optional[bool]	None	If True, scales the raw logit predictions using a Sigmoid activation function. If False, disables scaling.
calibration_factor	Optional[float]	None	Use this factor to calibrate probabilities with sigmoid(logits * calibration_factor). Used only if scale_score is True.
score_threshold	Optional[float]	None	Return documents only with a score above this threshold.