Struct RetrievalConfig 

pub struct RetrievalConfig {
    pub hybrid_alpha: f64,
    pub candidate_k_keyword: i64,
    pub candidate_k_vector: i64,
    pub final_limit: i64,
    pub group_by: String,
    pub doc_agg: String,
    pub max_chunks_per_doc: usize,
}

Search and retrieval tuning parameters.

These settings control how keyword and semantic search results are merged in hybrid mode, and the overall result limits.

Hybrid Scoring

The hybrid_alpha weight determines the blend between keyword (BM25) and semantic (cosine similarity) scores:

hybrid_score = (1 - α) × keyword_score + α × semantic_score

• α = 0.0 → pure keyword search
• α = 1.0 → pure semantic search
• α = 0.6 (default) → 60% semantic, 40% keyword

See docs/HYBRID_SCORING.md for the full specification.

Fields

hybrid_alpha: f64

Weight for semantic vs. keyword scores in hybrid mode. Range: [0.0, 1.0]. Default: 0.6.

candidate_k_keyword: i64

Number of keyword candidates to fetch before merging. Default: 80.

candidate_k_vector: i64

Number of vector candidates to fetch before merging. Default: 80.

final_limit: i64

Maximum number of results to return after merging and ranking. Default: 12.

group_by: String

Grouping strategy for results. Default: "document".

doc_agg: String

Aggregation method for document-level scores. Default: "max".

max_chunks_per_doc: usize

Maximum chunks per document in results. Default: 3.

Trait Implementations

impl Clone for RetrievalConfig

fn clone(&self) -> RetrievalConfig

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for RetrievalConfig

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for RetrievalConfig

fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.