fix: Migrate to OpenAI-style messages format

UserWarning: The 'tuples' format for chatbot messages is deprecated

🔧 Changes:
- Change Chatbot type from 'tuples' to 'messages'
- Migrate from [[user, bot]] to [{role, content}] format
- Update history parsing to use OpenAI-style messages
- Change history[-3:] tuples to history[-6:] messages (last 3 turns)

✨ Benefits:
- Future-proof against Gradio deprecation
- Compatible with OpenAI message format
- Better interoperability with other tools
- Cleaner message structure

🔄 Format Migration:
- Old: [["user msg", "bot msg"]]
- New: [{"role": "assistant", "content": "bot msg"}]

🤖 Generated with [Claude Code](https://claude.com/claude-code)

Co-Authored-By: Claude <noreply@anthropic.com>

app.py

@@ -91,18 +91,18 @@ def generate_response_impl(message, history):
         current_model, current_tokenizer = load_model_once()
 
         if current_model is None or current_tokenizer is None:
-            return history + [[message, "❌ 모델을 로드할 수 없습니다."]]
+            return history + [{"role": "assistant", "content": "❌ 모델을 로드할 수 없습니다."}]
 
         # Get device
         device = next(current_model.parameters()).device
 
         # Build conversation context (last 3 turns)
         conversation = ""
-        for user_msg, bot_msg in history[-3:]:
-            if user_msg:
-                conversation += f"사용자: {user_msg}\n"
-            if bot_msg:
-                conversation += f"어시스턴트: {bot_msg}\n"
+        for msg in history[-6:]:  # Last 3 turns (6 messages: 3 user + 3 assistant)
+            if msg["role"] == "user":
+                conversation += f"사용자: {msg['content']}\n"
+            elif msg["role"] == "assistant":
+                conversation += f"어시스턴트: {msg['content']}\n"
 
         conversation += f"사용자: {message}\n어시스턴트:"
 
@@ -142,7 +142,7 @@ def generate_response_impl(message, history):
         if not response:
             response = "죄송합니다. 응답을 생성할 수 없었습니다."
 
-        return history + [[message, response]]
+        return history + [{"role": "assistant", "content": response}]
 
     except Exception as e:
         import traceback
@@ -151,7 +151,7 @@ def generate_response_impl(message, history):
         print(f"Error: {error_msg}")
         print(traceback.format_exc())
         print("=" * 50)
-        return history + [[message, f"❌ 오류: {error_msg[:200]}"]]
+        return history + [{"role": "assistant", "content": f"❌ 오류: {error_msg[:200]}"}]
 
 
 # Conditionally apply ZeroGPU decorator
@@ -207,7 +207,7 @@ with gr.Blocks(title="🤖 Llama-2-Ko Chatbot") as demo:
 
     gr.Markdown(header)
 
-    chatbot = gr.Chatbot(height=400, type="tuples", show_label=False)
+    chatbot = gr.Chatbot(height=400, type="messages", show_label=False)
 
     with gr.Row():
         msg = gr.Textbox(