[Bugfix] Expose usage field in GenerateResponse for disaggregated serving (#42748)

Signed-off-by: AIvashov <ivashov.aleksey@proton.me>
Signed-off-by: NickLucche <nicolo.lucchesi@mistral.ai>
Co-authored-by: NickLucche <nicolo.lucchesi@mistral.ai>

vllm/entrypoints/scale_out/token_in_token_out/protocol.py

@@ -221,8 +221,10 @@ class GenerateResponse(BaseModel):
             "through out the inference process and return in response."
         ),
     )
+    model: str | None = None
+    created: int | None = None
     choices: list[GenerateResponseChoice]
-
+    usage: UsageInfo | None = Field(default=None)
     prompt_logprobs: list[dict[int, Logprob] | None] | None = None
 
     kv_transfer_params: dict[str, Any] | None = Field(

vllm/entrypoints/scale_out/token_in_token_out/serving.py

@@ -320,7 +320,7 @@ class ServingTokens(OpenAIServing):
         request_metadata.final_usage_info = usage
 
         response = GenerateResponse(
-            id=request_id,
+            request_id=request_id,
             created=created_time,
             model=model_name,
             choices=choices,