pub struct RequestOutput<'a> {Show 13 fields
pub rid: &'a str,
pub output_ids: &'a [i32],
pub finish_reason: Option<&'a str>,
pub prompt_tokens: i64,
pub completion_tokens: i64,
pub cached_tokens: i64,
pub output_token_logprobs: &'a [f64],
pub finish_details: Option<&'a FinishDetails>,
pub spec_verify_ct: Option<i64>,
pub accept_draft_tokens: Option<f64>,
pub input_token_logprobs: Option<(&'a [Option<f64>], &'a [i32])>,
pub input_top_logprobs: Option<PromptTopLogprobRows<'a>>,
pub output_top_logprobs: Option<TopLogprobRows<'a>>,
}Expand description
One request’s slice of a BatchTokenIDOutSlim.
Fields§
§rid: &'a strRequest id.
output_ids: &'a [i32]Token ids generated this step.
finish_reason: Option<&'a str>The finish type, or None while the request runs.
prompt_tokens: i64Prompt length.
completion_tokens: i64Tokens generated so far.
cached_tokens: i64Prompt tokens served from the prefix cache.
output_token_logprobs: &'a [f64]Logprobs of this step’s tokens; empty when not requested.
finish_details: Option<&'a FinishDetails>The finish reason in full, once the request ended (from a producer that sends it).
spec_verify_ct: Option<i64>Verify rounds so far, with speculative decoding.
accept_draft_tokens: Option<f64>Accepted tokens per verify round, once finished with speculative decoding.
input_token_logprobs: Option<(&'a [Option<f64>], &'a [i32])>Prompt logprobs and their tokens, in the frame that ships them.
input_top_logprobs: Option<PromptTopLogprobRows<'a>>The prompt positions’ top-k alternatives, in the frame that ships them.
output_top_logprobs: Option<TopLogprobRows<'a>>This step’s tokens’ top-k alternatives, when the request asked for them.
Trait Implementations§
Source§impl<'a> Clone for RequestOutput<'a>
impl<'a> Clone for RequestOutput<'a>
impl<'a> Copy for RequestOutput<'a>
Source§impl<'a> Debug for RequestOutput<'a>
impl<'a> Debug for RequestOutput<'a>
Source§impl<'a> PartialEq for RequestOutput<'a>
impl<'a> PartialEq for RequestOutput<'a>
impl<'a> StructuralPartialEq for RequestOutput<'a>
Auto Trait Implementations§
impl<'a> Freeze for RequestOutput<'a>
impl<'a> RefUnwindSafe for RequestOutput<'a>
impl<'a> Send for RequestOutput<'a>
impl<'a> Sync for RequestOutput<'a>
impl<'a> Unpin for RequestOutput<'a>
impl<'a> UnsafeUnpin for RequestOutput<'a>
impl<'a> UnwindSafe for RequestOutput<'a>
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