FIX method

rag_system.py

@@ -192,18 +192,26 @@ Question: {query}
 
 Answer:"""
 
-        # Generate response
+        # Generate response using chat completion
         try:
-            response = self.llm_client.text_generation(
-                prompt,
-                max_new_tokens=max_tokens,
+            messages = [
+                {
+                    "role": "user",
+                    "content": prompt
+                }
+            ]
+
+            response = self.llm_client.chat_completion(
+                messages=messages,
+                max_tokens=max_tokens,
                 temperature=temperature,
-                do_sample=temperature > 0,
             )
 
-            # Clean up response
-            if isinstance(response, str):
-                answer = response.strip()
+            # Extract answer from response
+            if hasattr(response, 'choices') and len(response.choices) > 0:
+                answer = response.choices[0].message.content.strip()
+            elif isinstance(response, dict) and 'choices' in response:
+                answer = response['choices'][0]['message']['content'].strip()
             else:
                 answer = str(response).strip()