diff --git a/web_demo2.py b/web_demo2.py
index 6c66308..8b85b10 100644
--- a/web_demo2.py
+++ b/web_demo2.py
@@ -2,14 +2,12 @@ from transformers import AutoModel, AutoTokenizer
 import streamlit as st
 from streamlit_chat import message
 
-
 st.set_page_config(
     page_title="ChatGLM2-6b 演示",
-    page_icon=":robot:",
+    page_icon=":robot:"
     layout='wide'
 )
 
-
 @st.cache_resource
 def get_model():
     tokenizer = AutoTokenizer.from_pretrained("THUDM/chatglm2-6b", trust_remote_code=True)
@@ -20,19 +18,20 @@ def get_model():
     model = model.eval()
     return tokenizer, model
 
-
 MAX_TURNS = 20
 MAX_BOXES = MAX_TURNS * 2
 
+#在启动时加载模型
+get_model()
 
-def predict(input, max_length, top_p, temperature, history=None):
+def predict(input, history=None):
     tokenizer, model = get_model()
     if history is None:
         history = []
 
     with container:
         if len(history) > 0:
-            if len(history)>MAX_BOXES:
+            if len(history) > MAX_BOXES:
                 history = history[-MAX_TURNS:]
             for i, (query, response) in enumerate(history):
                 message(query, avatar_style="big-smile", key=str(i) + "_user")
@@ -42,12 +41,15 @@ def predict(input, max_length, top_p, temperature, history=None):
         st.write("AI正在回复:")
         with st.empty():
             for response, history in model.stream_chat(tokenizer, input, history, max_length=max_length, top_p=top_p,
-                                               temperature=temperature):
+                                                       temperature=temperature):
                 query, response = history[-1]
                 st.write(response)
 
     return history
 
+#清除对话历史
+def clean():
+    st.session_state["state"] = None
 
 container = st.container()
 
@@ -72,4 +74,7 @@ if 'state' not in st.session_state:
 if st.button("发送", key="predict"):
     with st.spinner("AI正在思考，请稍等........"):
         # text generation
+        clean_button = st.button("新对话", on_click=clean)
         st.session_state["state"] = predict(prompt_text, max_length, top_p, temperature, st.session_state["state"])
+
+    st.session_state["state"]