Struct RunLatency 

pub struct RunLatency {
    pub generation_ms: f64,
    pub ttft_ms: Option<u64>,
}

Per-run latency a runner reports alongside token/cost accounting. Self-reported, not verified — benchmarking/observability only.

Fields

generation_ms: f64

Sum of model GENERATION (decode) wall-clock across turns, in ms, EXCLUDING tool execution. The honest denominator for decode tokens/sec (output_tokens / generation_ms) — total wall-clock would conflate decode with tool latency.

ttft_ms: Option<u64>

Time-to-first-token of the FIRST generation call, in ms, when the backend measures it (InferenceEngine::generate_tracked → time_to_first_token_ms; None on paths that don’t surface it).

Trait Implementations

impl Clone for RunLatency

fn clone(&self) -> Self

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for RunLatency

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

Formats the value using the given formatter.

impl Default for RunLatency

fn default() -> Self

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for RunLatency

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

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for RunLatency

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

where __S: Serializer,

Serialize this value into the given Serde serializer.