Update app.py

app.py

@@ -49,7 +49,7 @@ class ChatHistory:
     def clear(self):
         self.messages = []
 
-# Load environment variables and setup (same as before)
+# Load environment variables and setup
 load_dotenv()
 
 HF_TOKEN = os.getenv("HF_TOKEN")
@@ -62,7 +62,6 @@ if not HF_TOKEN:
 
 embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
 
-# Qdrant setup (same as before)
 try:
     client = QdrantClient(
         url=os.getenv("QDRANT_URL"),
@@ -102,7 +101,7 @@ retriever = db.as_retriever(
 llm = ChatCerebras(
     model="llama-3.3-70b",
     api_key=C_apikey,
-    streaming=True  # Enable streaming
+    streaming=True
 )
 
 template = """
@@ -140,7 +139,7 @@ def create_rag_chain(chat_history: str):
 
 chat_history = ChatHistory()
 
-def process_stream(stream_queue: Queue, history: List[dict]) -> Generator[List[dict], None, None]:
+def process_stream(stream_queue: Queue, history: List[List[str]]) -> Generator[List[List[str]], None, None]:
     """Process the streaming response and update the chat interface"""
     current_response = ""
     
@@ -151,19 +150,21 @@ def process_stream(stream_queue: Queue, history: List[dict]) -> Generator[List[d
             
         current_response += chunk
         new_history = history.copy()
-        new_history[-1]["content"] = current_response
+        new_history[-1][1] = current_response  # Update the assistant's message
         yield new_history
 
 @spaces.GPU()
-def ask_question_gradio(question: str, history: List[dict]) -> Generator[tuple, None, None]:
+def ask_question_gradio(question: str, history: List[List[str]]) -> Generator[tuple, None, None]:
     try:
+        if history is None:
+            history = []
+            
         chat_history.add_message("user", question)
         formatted_history = chat_history.get_formatted_history()
         rag_chain = create_rag_chain(formatted_history)
         
-        # Update history with user message
-        history.append({"role": "user", "content": question})
-        history.append({"role": "assistant", "content": ""})
+        # Update history with user message and empty assistant message
+        history.append([question, ""])  # User message
         
         # Create a queue for streaming responses
         stream_queue = Queue()
@@ -184,7 +185,7 @@ def ask_question_gradio(question: str, history: List[dict]) -> Generator[tuple,
         # Yield updates to the chat interface
         response = ""
         for updated_history in process_stream(stream_queue, history):
-            response = updated_history[-1]["content"]
+            response = updated_history[-1][1]
             yield "", updated_history
         
         # Add final response to chat history
@@ -192,7 +193,9 @@ def ask_question_gradio(question: str, history: List[dict]) -> Generator[tuple,
         
     except Exception as e:
         logger.error(f"Error during question processing: {e}")
-        history.append({"role": "assistant", "content": "An error occurred. Please try again later."})
+        if not history:
+            history = []
+        history.append([question, "An error occurred. Please try again later."])
         yield "", history
 
 def clear_chat():