pub struct SpeculativeDecodingTelemetry {Show 14 fields
pub execution_topology: String,
pub target_tokens: usize,
pub draft_tokens: usize,
pub accepted_tokens: usize,
pub accept_rate: f64,
pub rounds: usize,
pub accept_lens: Vec<usize>,
pub emitted_tokens: usize,
pub optimistic_draft_tokens: usize,
pub reused_optimistic_tokens: usize,
pub discarded_optimistic_tokens: usize,
pub adaptive_lookahead_disabled: bool,
pub optimistic_draft_seconds: f64,
pub verification_in_flight_seconds: f64,
}Expand description
Speculative-decoding observations for one request.
Fields§
§execution_topology: StringStable target/assistant execution-placement topology label.
target_tokens: usizeTarget tokens evaluated.
draft_tokens: usizeAssistant proposals.
accepted_tokens: usizeAccepted assistant proposals.
accept_rate: f64Proposal acceptance fraction.
rounds: usizeVerification rounds.
accept_lens: Vec<usize>Accepted proposal count per round.
emitted_tokens: usizeEmitted tokens, including terminal EOS where applicable.
optimistic_draft_tokens: usizeOptimistically drafted tokens.
reused_optimistic_tokens: usizeOptimistically reused tokens.
discarded_optimistic_tokens: usizeOptimistically discarded tokens.
adaptive_lookahead_disabled: boolWhether adaptive accounting disabled further lookahead.
optimistic_draft_seconds: f64Host time spent in optimistic drafting.
verification_in_flight_seconds: f64Target verification in-flight wall time.
Trait Implementations§
Source§impl Clone for SpeculativeDecodingTelemetry
impl Clone for SpeculativeDecodingTelemetry
Source§fn clone(&self) -> SpeculativeDecodingTelemetry
fn clone(&self) -> SpeculativeDecodingTelemetry
Returns a duplicate of the value. Read more
1.0.0 (const: unstable) · Source§fn clone_from(&mut self, source: &Self)
fn clone_from(&mut self, source: &Self)
Performs copy-assignment from
source. Read moreSource§impl Debug for SpeculativeDecodingTelemetry
impl Debug for SpeculativeDecodingTelemetry
Source§impl<'de> Deserialize<'de> for SpeculativeDecodingTelemetry
impl<'de> Deserialize<'de> for SpeculativeDecodingTelemetry
Source§fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>where
__D: Deserializer<'de>,
fn deserialize<__D>(__deserializer: __D) -> Result<Self, __D::Error>where
__D: Deserializer<'de>,
Deserialize this value from the given Serde deserializer. Read more
impl StructuralPartialEq for SpeculativeDecodingTelemetry
Auto Trait Implementations§
impl Freeze for SpeculativeDecodingTelemetry
impl RefUnwindSafe for SpeculativeDecodingTelemetry
impl Send for SpeculativeDecodingTelemetry
impl Sync for SpeculativeDecodingTelemetry
impl Unpin for SpeculativeDecodingTelemetry
impl UnsafeUnpin for SpeculativeDecodingTelemetry
impl UnwindSafe for SpeculativeDecodingTelemetry
Blanket Implementations§
Source§impl<T> BorrowMut<T> for Twhere
T: ?Sized,
impl<T> BorrowMut<T> for Twhere
T: ?Sized,
Source§fn borrow_mut(&mut self) -> &mut T
fn borrow_mut(&mut self) -> &mut T
Mutably borrows from an owned value. Read more