Struct RuVectorDB 

pub struct RuVectorDB {
    pub embeddings: TritMatrix,
    pub metadata: Vec<String>,
}

A high-performance ternary-quantized vector database.

Fields

embeddings: TritMatrix

The document embeddings matrix [num_docs x embedding_dim].

metadata: Vec<String>

Metadata associated with each document (e.g. original text or ID).

Implementations

impl RuVectorDB

pub fn from_f32(embeddings: &[Vec<f32>], metadata: Vec<String>) -> Result<Self>

Create a new database from f32 embeddings.

pub fn search(&self, query_f32: &[f32], top_k: usize) -> Vec<SearchResult>

Search for the top-k most similar documents using ternary Sparse GEMV.

The query vector is quantized to ternary before search.

Trait Implementations

impl<'de> Deserialize<'de> for RuVectorDB

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

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for RuVectorDB

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.