[Router] Add load-aware prefill admission and bounded policy proposals (#37843)
Signed-off-by: Vincent Gao <vincentbo@linux.alibaba.com> Co-authored-by: Kangyan Zhou <zky314343421@gmail.com> Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Co-authored-by: Shangming Cai <csmthu@gmail.com>
This commit is contained in:
co-authored by
Kangyan Zhou
Claude Opus 4.8
Shangming Cai
parent
d50e9a9756
commit
ecd97de1fc
@@ -25,6 +25,7 @@ use sgl_router::config::{
|
||||
PolicyKind, ProxyConfig, ServerConfig, StaticUrlsDiscoveryConfig,
|
||||
};
|
||||
use sgl_router::discovery::{ModelId, WorkerId, WorkerMode, WorkerSpec};
|
||||
use sgl_router::policies::engine_load::EngineLoadTable;
|
||||
use sgl_router::policies::factory::build_registry;
|
||||
use sgl_router::policies::kv_events::{BlockSizeOracle, HashTree};
|
||||
use sgl_router::proxy::Proxy;
|
||||
@@ -54,6 +55,7 @@ fn config() -> Config {
|
||||
circuit_breaker: None,
|
||||
cache_aware: Some(CacheAwareConfig::default()),
|
||||
sticky: None,
|
||||
affinity: None,
|
||||
fused: None,
|
||||
eligibility: None,
|
||||
},
|
||||
@@ -88,6 +90,7 @@ fn build_ctx(url: String) -> Arc<AppContext> {
|
||||
Arc::new(HashTree::new()),
|
||||
Arc::clone(&tokenizers),
|
||||
BlockSizeOracle::new(),
|
||||
EngineLoadTable::new(),
|
||||
)
|
||||
.unwrap(),
|
||||
);
|
||||
|
||||
Reference in New Issue
Block a user