fix(request): make callbacks not require semantic_cache (#465)

Broke in 48312e7#diff-3e852cfcf107dfd78ee2b7c7e88e05428b7f2bffcfc5ca1dce0c070caf563f2aR139

Same as #399

Signed-off-by: Max Wittig <[email protected]>

Author: max-wittig

src/vllm_router/services/request_service/request.py

@@ -141,10 +141,10 @@ async def process_request(
         await store_in_semantic_cache(
             endpoint=endpoint, method=request.method, body=body, chunk=cache_chunk
         )
-        if background_tasks and hasattr(request.app.state, "callbacks"):
-            background_tasks.add_task(
-                request.app.state.callbacks.post_request, request, full_response
-            )
+    if background_tasks and hasattr(request.app.state, "callbacks"):
+        background_tasks.add_task(
+            request.app.state.callbacks.post_request, request, full_response
+        )
 
 
 async def route_general_request(