fix(vertex_ai): enable inputAudioTranscription in realtime session config

Add inputAudioTranscription to the Vertex AI realtime setup so the backend returns transcripts of user speech, allowing guardrails to inspect voice input.

Made-with: Cursor

litellm/llms/vertex_ai/realtime/transformation.py

@@ -124,6 +124,8 @@ class VertexAIRealtimeConfig(GeminiRealtimeConfig):
                     "silenceDurationMs": 800,
                 }
             },
+            # Return input transcript so guardrails can inspect user speech.
+            "inputAudioTranscription": {},
             # Return output transcript so clients can read what the model said.
             "outputAudioTranscription": {},
         }