feat(frontend): session id plumbing into requests (#48048)

Signed-off-by: Karen Chung <karenc@nvidia.com>
This commit is contained in:
Karen Chung
2026-08-03 22:18:58 +00:00
committed by GitHub
parent e578de311c
commit f57123aa2d
39 changed files with 438 additions and 4 deletions
+1
View File
@@ -59,6 +59,7 @@ pub fn lower_text_request(
cache_salt: request.cache_salt.clone(),
priority: request.priority,
data_parallel_rank: request.data_parallel_rank,
session_id: request.session_id.clone(),
reasoning_parser_kwargs: request.reasoning_parser_kwargs.clone(),
lora_request: request.lora_request.clone(),
arrival_time: request.arrival_time,
+4
View File
@@ -183,6 +183,9 @@ pub struct TextRequest {
/// Override data parallel rank.
#[serde(default)]
pub data_parallel_rank: Option<u32>,
/// Stable session identity shared by related requests.
#[serde(default)]
pub session_id: Option<String>,
/// Optional reasoning-parser kwargs forwarded to engine-side structured
/// output logic.
#[serde(default)]
@@ -212,6 +215,7 @@ impl TextRequest {
cache_salt: None,
add_special_tokens: false,
data_parallel_rank: None,
session_id: None,
reasoning_parser_kwargs: None,
lora_request: None,
arrival_time: None,