Update app.py

app.py

@@ -33,8 +33,12 @@ def load_pipeline():
 def chat_interface(inputs):
     question = inputs["input"]
     chat_history = inputs["history"]
-    # Assuming `chain` is your instance of ConversationalRetrievalChain
-    result = chain.run({"input": question, "history": chat_history})
+    chat_history_tuples = []
+    for message in chat_history:
+        chat_history_tuples.append((message[0], message[1]))
+    
+    #result = llm_chain({"input": query, "history": chat_history_tuples})
+    result = llm_chain.invoke({"input": question, "history": chat_history})
     return result["response"]
 
 llm = load_pipeline()
@@ -67,13 +71,6 @@ with gr.Blocks() as demo:
     clear = gr.Button("Clear")
     #llm_chain, llm = init_chain(model, tokenizer)
 
-    # Convert chat history to list of tuples
-    chat_history_tuples = []
-    for message in chat_history:
-        chat_history_tuples.append((message[0], message[1]))
-    
-    result = llm_chain({"input": query, "history": chat_history_tuples})
-    
     iface = gr.Interface(
         fn=chat_interface, 
         inputs=[