UniCoilConfig

class lightning_ir.models.bi_encoders.coil.UniCoilConfig(query_length: int | None = 32, doc_length: int | None = 512, similarity_function: 'cosine' | 'dot' = 'dot', projection: 'linear' | 'linear_no_bias' = 'linear', **kwargs)

Bases: SingleVectorBiEncoderConfig

Configuration class for UniCOIL models.

__init__(query_length: int | None = 32, doc_length: int | None = 512, similarity_function: 'cosine' | 'dot' = 'dot', projection: 'linear' | 'linear_no_bias' = 'linear', **kwargs) → None

A UniCOIL model encodes queries and documents separately, and computes a similarity score using the maximum similarity of token embeddings between query and document.

Parameters:

• query_length (int | None) – Maximum query length in number of tokens. Defaults to 32.

• doc_length (int | None) – Maximum document length in number of tokens. Defaults to 512.

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

• projection (Literal["linear", "linear_no_bias"], optional) – Whether and how to project the embeddings. Defaults to “linear”.

Methods

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

A UniCOIL model encodes queries and documents separately, and computes a similarity score using the maximum similarity of token embeddings between query and document.

Attributes

embedding_dim
model_type	Model type for UniCOIL models.

model_type: str = 'unicoil'

Model type for UniCOIL models.