Struct BenchmarkResult 

pub struct BenchmarkResult {
    pub status: RunStatus,
    pub phase: String,
    pub progress: Option<(u64, u64)>,
    pub summary: Vec<Stat>,
    pub table: Option<ResultTable>,
    pub verdict: Option<Verdict>,
    pub metrics: BTreeMap<String, f64>,
    pub log: Vec<LogLine>,
    pub elapsed: Duration,
    pub hardware_state: Option<HardwareStateReport>,
    pub dataset_fingerprint: Option<String>,
}

2026-09-26: One frame of a benchmark run.

Fields

status: RunStatus

phase: String

2026-09-26: What is happening right now, e.g. "warmup".

progress: Option<(u64, u64)>

2026-09-26: (done, total) for the progress bar; None when the total is unknown.

summary: Vec<Stat>

table: Option<ResultTable>

verdict: Option<Verdict>

metrics: BTreeMap<String, f64>

2026-09-26: Raw headline numbers, keyed by stable metric name. The gate record copies them and scores them against BENCH.toml bounds, so the gate does not parse the formatted summary strings.

log: Vec<LogLine>

2026-09-26: Lines appended to the run log since the previous frame (not cumulative).

elapsed: Duration

hardware_state: Option<HardwareStateReport>

2026-09-26: What state the box was in before and after this run.

The executor (crate::executor) stamps it on terminal frames, the frame crate::RunRecord keeps and the gate record is built from. None means unmeasured, which is not the same as a healthy box.

dataset_fingerprint: Option<String>

2026-09-26: Content identity of the dataset the run scored against, when the benchmark sets one (the MLPerf agentic leg does). Carried into the gate record: the metrics map is f64-only, so without this a record could pin a draw’s size but not its content.

Implementations

impl BenchmarkResult

pub fn running(phase: impl Into<String>, elapsed: Duration) -> Self

pub fn completed(phase: impl Into<String>, elapsed: Duration) -> Self

pub fn failed( phase: impl Into<String>, reason: impl Into<String>, elapsed: Duration, ) -> Self

pub fn with_progress(self, done: u64, total: u64) -> Self

pub fn with_summary(self, summary: Vec<Stat>) -> Self

pub fn with_metrics(self, metrics: BTreeMap<String, f64>) -> Self

pub fn with_table(self, table: ResultTable) -> Self

pub fn with_verdict(self, verdict: Verdict) -> Self

pub fn with_log(self, log: Vec<LogLine>) -> Self

pub fn log_line(self, line: LogLine) -> Self

pub fn with_hardware_state(self, report: HardwareStateReport) -> Self

Trait Implementations

impl Clone for BenchmarkResult

fn clone(&self) -> BenchmarkResult

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for BenchmarkResult

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

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for BenchmarkResult

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

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for BenchmarkResult

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

where __S: Serializer,

Serialize this value into the given Serde serializer.