feat: nvext field to OpenAI APIs and add worker_id reporting (vLLM) (#4372)

Signed-off-by: Zhongxuan Wang <daniewang@nvidia.com>

lib/llm/tests/test_jail.rs

@@ -46,6 +46,7 @@ mod tests {
                 object: "chat.completion.chunk".to_string(),
                 usage: None,
                 service_tier: None,
+                nvext: None,
             };
 
             Annotated {
@@ -84,6 +85,7 @@ mod tests {
                 object: "chat.completion.chunk".to_string(),
                 usage: None,
                 service_tier: None,
+                nvext: None,
             };
 
             Annotated {
@@ -126,6 +128,7 @@ mod tests {
                 object: "chat.completion.chunk".to_string(),
                 usage: None,
                 service_tier: None,
+                nvext: None,
             };
 
             Annotated {
@@ -169,6 +172,7 @@ mod tests {
                 object: "chat.completion.chunk".to_string(),
                 usage: None,
                 service_tier: None,
+                nvext: None,
             };
 
             Annotated {
@@ -212,6 +216,7 @@ mod tests {
                 object: "chat.completion.chunk".to_string(),
                 usage: None,
                 service_tier: None,
+                nvext: None,
             };
 
             Annotated {
@@ -2050,6 +2055,7 @@ mod parallel_jail_tests {
             object: "chat.completion.chunk".to_string(),
             usage: None,
             service_tier: None,
+            nvext: None,
         };
 
         Annotated {

lib/llm/tests/test_reasoning_parser.rs

@@ -36,6 +36,7 @@ fn create_mock_response_chunk(
         object: "chat.completion.chunk".to_string(),
         usage: None,
         service_tier: None,
+        nvext: None,
     };
 
     Annotated {

lib/llm/tests/test_streaming_tool_parsers.rs

@@ -108,6 +108,7 @@ fn load_test_data(file_path: &str) -> TestData {
                 object: "chat.completion.chunk".to_string(),
                 usage: None,
                 service_tier: None,
+                nvext: None,
             };
 
             Annotated {

lib/llm/tests/test_streaming_usage.rs

@@ -108,6 +108,7 @@ fn build_backend_outputs_with_cached_tokens(cached_tokens: Option<u32>) -> Vec<B
             finish_reason: None,
             index: Some(0),
             completion_usage: None,
+            disaggregated_params: None,
         },
         BackendOutput {
             token_ids: vec![1917],
@@ -119,6 +120,7 @@ fn build_backend_outputs_with_cached_tokens(cached_tokens: Option<u32>) -> Vec<B
             finish_reason: None,
             index: Some(0),
             completion_usage: None,
+            disaggregated_params: None,
         },
         BackendOutput {
             token_ids: vec![0],
@@ -139,6 +141,7 @@ fn build_backend_outputs_with_cached_tokens(cached_tokens: Option<u32>) -> Vec<B
                 }),
                 completion_tokens_details: None,
             }),
+            disaggregated_params: None,
         },
     ]
 }