Struct TaskEvalResult 

pub struct TaskEvalResult {

    pub task: Task,
    pub dataset: DatasetId,
    pub backend: String,
    pub backend_display: Option<String>,
    pub seed: u64,
    pub success: bool,
    pub error: Option<String>,
    pub metrics: HashMap<String, f64>,
    pub num_examples: usize,
    pub duration_ms: Option<f64>,
    pub label_shift: Option<LabelShift>,
    pub robustness: Option<RobustnessResults>,
    pub stratified: Option<StratifiedMetrics>,
    pub confidence_intervals: Option<ConfidenceIntervals>,
    pub kb_version: Option<String>,
}

Results from evaluating a task-dataset-backend combination.

Fields

task: Task

Task being evaluated

dataset: DatasetId

Dataset used

backend: String

Backend name

backend_display: Option<String>

Backend display name (may include composition details, e.g. stacked(regex+heuristic)).

Best-effort: when absent, callers should fall back to backend.

seed: u64

Random seed used for sampling/examples.

success: bool

Whether evaluation succeeded

error: Option<String>

Error message if failed

metrics: HashMap<String, f64>

Metrics (task-specific, stored as JSON-serializable map)

num_examples: usize

Number of examples evaluated

duration_ms: Option<f64>

Time taken in milliseconds (if available)

label_shift: Option<LabelShift>

Label shift/familiarity metrics (if computed for zero-shot)

robustness: Option<RobustnessResults>

Robustness scores (if robustness testing was enabled)

stratified: Option<StratifiedMetrics>

Stratified metrics by various dimensions

confidence_intervals: Option<ConfidenceIntervals>

Confidence intervals for key metrics (if computed)

kb_version: Option<String>

KB version used (if available from dataset metadata)

Implementations

impl TaskEvalResult

pub fn is_skipped(&self) -> bool

Check if this is a “skipped” result (feature not available or incompatible) vs actual failure

pub fn primary_f1(&self) -> Option<f64>

Get primary F1 metric for ranking

Trait Implementations

impl Clone for TaskEvalResult

fn clone(&self) -> TaskEvalResult

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for TaskEvalResult

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

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for TaskEvalResult

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

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl From<&TaskEvalResult> for EvalHistoryEntry

fn from(result: &TaskEvalResult) -> Self

Converts to this type from the input type.

impl Serialize for TaskEvalResult

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

where __S: Serializer,

Serialize this value into the given Serde serializer.