fix

app.py

@@ -66,7 +66,7 @@ def respond(
         elif msg['role'] == 'assistant':
             conversation += f"Assistant: {msg['content']}\n"
     conversation += "Assistant: "
-    inputs = tokenizer.encode(messages, return_tensors='pt')
+    inputs = tokenizer.encode(conversation, return_tensors='pt')
     # text_streamer = TextStreamer(tokenizer, skip_prompt = True)
     # for response in model.generate(input_ids = inputs, streamer = text_streamer, max_new_tokens = max_tokens, use_cache = True,
     #                         temperature = temperature, min_p = top_p):