Print messages, history messing up is the guess

app.py

@@ -39,6 +39,9 @@ def chat_fxn_caller(message, history, temperature=0.6, top_p=0.95, max_tokens=12
     """
     messages = []
 
+    print("history")
+    print(history)
+
     for turn in history:
         # Each turn is [user_msg, bot_msg] or [user_msg] depending on history length
         if isinstance(turn, list):
@@ -52,6 +55,9 @@ def chat_fxn_caller(message, history, temperature=0.6, top_p=0.95, max_tokens=12
     # Append current user message
     messages.append({"role": "user", "content": str(message)})
 
+    print("messages")
+    print(messages)
+
     prompt = chatbot.tokenizer.apply_chat_template(messages, tokenize=False)
 
     # Generate response
@@ -67,9 +73,17 @@ def chat_fxn_caller(message, history, temperature=0.6, top_p=0.95, max_tokens=12
         pad_token_id=chatbot.tokenizer.eos_token_id
     )
 
+    print("response:")
+    print(response)
+    
+
     # Extract only new text
     generated = response[0]["generated_text"][len(prompt):].strip()
 
+    print("generated")
+    print(generated)
+    
+
     # Optional: truncate to avoid huge outputs on free GPU
     max_response_len = 512
     if len(generated) > max_response_len: