Update src/ai_processor.py

src/ai_processor.py

@@ -161,6 +161,7 @@ def _vlm_infer_gpu(messages, model_id: str, max_new_tokens: int, token: Optional
         txt = out[0].get("generated_text", "")
     return (txt or "").strip() or "⚠️ Empty response"
 
+@_SPACES_GPU(enable_queue=True)
 def generate_medgemma_report(  # kept name so callers don't change
     patient_info: str,
     visual_results: Dict,