Struct ModelArtifact 

pub struct ModelArtifact {
    pub fingerprint: SchemaFingerprint,
    pub params: ModelParams,
    pub tag: Option<String>,
    pub trained_on: u64,
    pub em_iterations: usize,
}

Everything that must be persisted after a successful EM training run.

Serializes to roughly 2–10 KB per artifact (bincode).

Fields

fingerprint: SchemaFingerprint

Fingerprint of the schema and data distribution this model was trained on.

params: ModelParams

Learned Fellegi-Sunter m/u parameters and decision thresholds.

tag: Option<String>

Optional human-readable label, e.g. "brp_2024_q1".

trained_on: u64

Unix timestamp (seconds) when EM training completed.

em_iterations: usize

Number of EM iterations performed.

Implementations

impl ModelArtifact

pub fn to_bytes(&self) -> Result<Vec<u8>, ZerError>

Serialize this artifact to bytes using bincode.

pub fn from_bytes(bytes: &[u8]) -> Result<ModelArtifact, ZerError>

Deserialize an artifact from bincode bytes.

Trait Implementations

impl Clone for ModelArtifact

fn clone(&self) -> ModelArtifact

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for ModelArtifact

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

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for ModelArtifact

fn deserialize<__D>( __deserializer: __D, ) -> Result<ModelArtifact, <__D as Deserializer<'de>>::Error>

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for ModelArtifact

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

where __S: Serializer,

Serialize this value into the given Serde serializer.