SpladeConfig

class lightning_ir.models.bi_encoders.splade.SpladeConfig(query_length: int | None = 32, doc_length: int | None = 512, similarity_function: 'cosine' | 'dot' = 'dot', sparsification: 'relu' | 'relu_log' | 'relu_2xlog' | None = 'relu_log', query_pooling_strategy: 'first' | 'mean' | 'max' | 'sum' = 'max', query_weighting: 'contextualized' | 'static' | None = 'contextualized', query_expansion: bool = True, doc_pooling_strategy: 'first' | 'mean' | 'max' | 'sum' = 'max', doc_weighting: 'contextualized' | 'static' | None = 'contextualized', doc_expansion: bool = True, **kwargs)[source]

Bases: SingleVectorBiEncoderConfig

Configuration class for a SPLADE model.

__init__(query_length: int | None = 32, doc_length: int | None = 512, similarity_function: 'cosine' | 'dot' = 'dot', sparsification: 'relu' | 'relu_log' | 'relu_2xlog' | None = 'relu_log', query_pooling_strategy: 'first' | 'mean' | 'max' | 'sum' = 'max', query_weighting: 'contextualized' | 'static' | None = 'contextualized', query_expansion: bool = True, doc_pooling_strategy: 'first' | 'mean' | 'max' | 'sum' = 'max', doc_weighting: 'contextualized' | 'static' | None = 'contextualized', doc_expansion: bool = True, **kwargs) → None[source]

A SPLADE model encodes queries and documents separately. Before computing the similarity score, the contextualized token embeddings are projected into a logit distribution over the vocabulary using a pre-trained masked language model (MLM) head. The logit distribution is then sparsified and aggregated to obtain a single embedding for the query and document.

Parameters:

  • query_length (int | None) – Maximum number of tokens per query. If None does not truncate. Defaults to 32.

  • doc_length (int | None) – Maximum number of tokens per document. If None does not truncate. Defaults to 512.

  • similarity_function (Literal["cosine", "dot"]) – Similarity function to compute scores between query and document embeddings. Defaults to “dot”.

  • sparsification (Literal['relu', 'relu_log', 'relu_2xlog'] | None) – Whether and which sparsification function to apply. Defaults to None.

  • query_weighting (Literal["contextualized", "static"] | None) – Whether to reweight query embeddings. Defaults to “contextualized”.

  • query_expansion (bool) – Whether to allow query expansion. Defaults to True.

  • query_pooling_strategy (Literal["first", "mean", "max", "sum"]) – Pooling strategy for query embeddings. Defaults to “max”.

  • doc_pooling_strategy (Literal["first", "mean", "max", "sum"]) – Pooling strategy for document embeddings. Defaults to “max”.

  • doc_weighting (Literal["contextualized", "static"] | None) – Whether to reweight document embeddings. Defaults to “contextualized”.

  • doc_expansion (bool) – Whether to allow document expansion. Defaults to True.

Methods

__init__([query_length, doc_length, ...])

A SPLADE model encodes queries and documents separately.

Attributes

embedding_dim

model_type

Model type for a SPLADE model.
model_type: str = 'splade'

Model type for a SPLADE model.