Struct RealtimeConfig 

pub struct RealtimeConfig {
    pub latency_target_ms: f32,
    pub throughput_target_tps: f32,
    pub min_batch_size: usize,
    pub max_batch_size: usize,
    pub kv_cache_pressure_threshold: f32,
    pub enable_speculative: bool,
    pub speculative: SpeculativeConfig,
    pub batch_strategy: BatchSizeStrategy,
    pub kv_policy: KvCachePressurePolicy,
    pub max_memory_bytes: usize,
    pub optimization_interval_ms: u64,
}

Configuration for the realtime optimizer

Fields

latency_target_ms: f32

Target latency for TTFT (milliseconds)

throughput_target_tps: f32

Target throughput (tokens per second)

min_batch_size: usize

Minimum batch size

max_batch_size: usize

Maximum batch size

kv_cache_pressure_threshold: f32

KV cache pressure threshold (0.0 - 1.0)

enable_speculative: bool

Enable speculative decoding

speculative: SpeculativeConfig

Speculative decoding configuration

batch_strategy: BatchSizeStrategy

Batch sizing strategy

kv_policy: KvCachePressurePolicy

KV cache pressure policy

max_memory_bytes: usize

Maximum memory budget (bytes)

optimization_interval_ms: u64

Optimization interval (how often to recompute decisions)

Trait Implementations

impl Clone for RealtimeConfig

fn clone(&self) -> RealtimeConfig

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for RealtimeConfig

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

Formats the value using the given formatter.

impl Default for RealtimeConfig

fn default() -> Self

Returns the “default value” for a type.

impl<'de> Deserialize<'de> for RealtimeConfig

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

where __D: Deserializer<'de>,

Deserialize this value from the given Serde deserializer.

impl Serialize for RealtimeConfig

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

where __S: Serializer,

Serialize this value into the given Serde serializer.