Update app.py

app.py

@@ -9,11 +9,16 @@ model = AutoModelForCausalLM.from_pretrained(
     model_name, low_cpu_mem_usage=True, device_map="auto", torch_dtype="auto"
 )
 
-def predict(history):
+def predict(history, message):
     """
     history: list of [user, bot] message pairs from the Chatbot
+    message: new user input string
     """
-    # Convert history into the 'messages' format for chat template
+    # Add the latest user message to the conversation
+    history = history or []  # make sure it's a list
+    history.append((message, ""))
+
+    # Convert to messages format for Qwen
     messages = []
     for human, bot in history:
         if human:
@@ -21,24 +26,28 @@ def predict(history):
         if bot:
             messages.append({"role": "assistant", "content": bot})
 
+    # Apply chat template
     text = tokenizer.apply_chat_template(
         messages, tokenize=False, add_generation_prompt=True
     )
+
     model_inputs = tokenizer([text], return_tensors="pt").to(model.device)
 
+    # Generate response
     generated_ids = model.generate(**model_inputs, max_new_tokens=512)
     generated_ids = [
         output_ids[len(input_ids):]
         for input_ids, output_ids in zip(model_inputs.input_ids, generated_ids)
     ]
-
     reply = tokenizer.batch_decode(generated_ids, skip_special_tokens=True)[0]
-    history.append((messages[-1]["content"] if messages else "", reply))
-    return history
 
-with gr.Blocks() as server:
+    # Update last message with bot reply
+    history[-1] = (message, reply)
+    return history, ""  # return history + clear textbox
+
+with gr.Blocks() as demo:
     chatbot = gr.Chatbot()
     msg = gr.Textbox(placeholder="Type your message here...")
-    msg.submit(predict, [chatbot], chatbot)
+    msg.submit(predict, [chatbot, msg], [chatbot, msg])
 
-server.launch(server_name="0.0.0.0", server_port=7860, share=False)
+demo.launch(server_name="0.0.0.0", server_port=7860, share=False)