Struct VectorConfig 

pub struct VectorConfig {
    pub dimensions: usize,
    pub distance: Distance,
    pub m: Option<u32>,
    pub ef_construction: Option<u32>,
    pub ef_search: Option<u32>,
    pub lazy_embedding: bool,
    pub embedding_model: Option<String>,
    pub compression: Option<CompressionConfig>,
}

Configuration for a vector collection, using a builder-style API that mirrors VectorConfig in the Python SDK.

Fields

dimensions: usize

Number of dimensions in each embedding.

distance: Distance

Distance metric (default: Distance::Cosine).

m: Option<u32>

HNSW M parameter – connections per node (default: library default).

ef_construction: Option<u32>

HNSW ef_construction – build quality (default: library default).

ef_search: Option<u32>

HNSW ef_search – query quality (default: library default).

lazy_embedding: bool

Enable lazy embedding mode (store text, embed on demand).

embedding_model: Option<String>

Model name to use for lazy embedding.

compression: Option<CompressionConfig>

Optional vector compression settings.

Implementations

impl VectorConfig

pub fn new(dimensions: usize) -> Self

Create a minimal config with the given number of dimensions.

pub fn with_distance(self, distance: Distance) -> Self

Set the distance metric.

pub fn with_m(self, m: u32) -> Self

Set the HNSW M parameter.

pub fn with_ef_construction(self, ef: u32) -> Self

Set the ef_construction parameter.

pub fn with_ef_search(self, ef: u32) -> Self

Set the ef_search parameter.

pub fn with_lazy_embedding(self, model: impl Into<String>) -> Self

Enable lazy embedding using an embedding model name.

pub fn with_compression(self, config: CompressionConfig) -> Self

Set a custom compression configuration.

pub fn with_delta_compression(self) -> Self

Enable delta compression with default settings.

pub fn with_quantized_compression(self) -> Self

Enable quantised-delta compression with default settings.

pub fn to_json(&self) -> String

Serialise to JSON for passing to the native C API.

Trait Implementations

impl Clone for VectorConfig

fn clone(&self) -> VectorConfig

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for VectorConfig

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

Formats the value using the given formatter.