Update src/ai_processor.py

src/ai_processor.py

@@ -45,9 +45,9 @@ class AIProcessor:
             HfFolder.save_token(self.config.HF_TOKEN)
             logging.info("HuggingFace token set successfully")
 
-        # YOLO detection on CPU
+        # YOLO detection on CPU (force CPU to avoid CUDA init)
         try:
-            self.models_cache['det'] = YOLO(self.config.YOLO_MODEL_PATH)
+            self.models_cache['det'] = YOLO(self.config.YOLO_MODEL_PATH, device='cpu')
             logging.info("✅ YOLO detection model loaded on CPU")
         except Exception as e:
             logging.warning(f"YOLO model not available: {e}")
@@ -171,7 +171,10 @@ class AIProcessor:
         if not vs:
             return "Clinical guidelines unavailable"
         docs = vs.as_retriever(search_kwargs={'k':10}).invoke(query)
-        return '\n\n'.join(f"Source: {d.metadata.get('source','?')}, Page: {d.metadata.get('page','?')}\n{d.page_content}" for d in docs)
+        return '\n\n'.join(
+            f"Source: {d.metadata.get('source','?')}, Page: {d.metadata.get('page','?')}\n{d.page_content}"
+            for d in docs
+        )
 
     @spaces.GPU(enable_queue=True, duration=120)
     def generate_final_report(self, patient_info, visual_results, guideline_context, image_pil, max_new_tokens=None):
@@ -214,13 +217,21 @@ class AIProcessor:
             do_sample=False
         )
         report = out[0]['generated_text'][-1].get('content','')
-        return report or self._generate_fallback_report(patient_info, visual_results, guideline_context)
+        return report or self._generate_fallback_report(
+            patient_info, visual_results, guideline_context
+        )
 
     def _generate_fallback_report(self, patient_info, visual_results, guideline_context):
         """Produce text-only fallback."""
         dp = visual_results.get('detection_image_path','N/A')
         sp = visual_results.get('segmentation_image_path','N/A')
-        return f"# Report\n{patient_info}\nType: {visual_results['wound_type']}\nDetection Image: {dp}\nSegmentation Image: {sp}\nGuidelines: {guideline_context[:200]}..."
+        return (
+            f"# Report\n{patient_info}\n"
+            f"Type: {visual_results['wound_type']}\n"
+            f"Detection Image: {dp}\n"
+            f"Segmentation Image: {sp}\n"
+            f"Guidelines: {guideline_context[:200]}..."
+        )
 
     def save_and_commit_image(self, image_pil):
         """Save locally and optionally to HuggingFace."""
@@ -260,7 +271,6 @@ class AIProcessor:
             image = Image.open(image)
         return self.full_analysis_pipeline(image, questionnaire_data)
 
-        
     def _assess_risk_legacy(self, questionnaire_data):
         """Legacy risk assessment for backward compatibility"""
         risk_factors = []