Enum PruningStrategy 

pub enum PruningStrategy {
    None,
    Median {
        n_warmup_steps: usize,
    },
    Percentile {
        percentile: f64,
        n_warmup_steps: usize,
    },
    Hyperband,
}

Pruning strategy for early stopping of unpromising trials.

Variants

None

No pruning.

Median

Prune if metric is below median of completed trials at same step.

Fields

n_warmup_steps: usize

Steps a trial runs unconditionally before pruning checks begin — early metrics are too noisy to kill on.

Percentile

Prune if metric is below given percentile.

Fields

percentile: f64

Percentile (0–100) of completed trials’ values at the same step the trial must reach to survive. 50.0 is PruningStrategy::Median.

n_warmup_steps: usize

Steps a trial runs unconditionally before pruning checks begin.

Hyperband

Bracket-based pruning (used with Hyperband). Declared for forward compatibility: the StudyRunner currently builds no pruner for it, so it behaves like PruningStrategy::None.

Trait Implementations

impl Clone for PruningStrategy

fn clone(&self) -> PruningStrategy

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for PruningStrategy

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

Formats the value using the given formatter.

impl<'de> Deserialize<'de> for PruningStrategy

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

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for PruningStrategy

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

where __S: Serializer,

Serialize this value into the given Serde serializer.