Update app.py

app.py

@@ -5,7 +5,7 @@ import gradio as gr
 # 加载指令模型
 model = AutoModelForCausalLM.from_pretrained(
     "MediaTek-Research/Breeze-7B-Instruct-v1_0",
-    device_map="auto",  # 保留这一行以使用设备映射
+    device_map="auto",
     torch_dtype=torch.bfloat16,
 )
 
@@ -15,14 +15,16 @@ tokenizer = AutoTokenizer.from_pretrained("MediaTek-Research/Breeze-7B-Instruct-
 # 定义SYS_PROMPT
 SYS_PROMPT = "You are a helpful AI assistant built by MediaTek Research. The user you are helping speaks Traditional Chinese and comes from Taiwan."
 
+# 定义对话历史记录变量
+chat_history = []
+
 def generate_response(user_input):
-    # 定义聊天内容
-    chat = [
-        {"role": "user", "content": user_input},
-    ]
+    global chat_history
+    # 将用户输入添加到对话历史记录
+    chat_history.append({"role": "user", "content": user_input})
     
     # 应用聊天模板
-    prompt = tokenizer.apply_chat_template(chat, tokenize=False)
+    prompt = tokenizer.apply_chat_template(chat_history, tokenize=False)
     full_prompt = f"<s>{SYS_PROMPT} [INST] {prompt} [/INST]"
 
     # 生成文本
@@ -38,10 +40,22 @@ def generate_response(user_input):
 
     # 解码输出
     generated_text = tokenizer.decode(outputs[0], skip_special_tokens=True)
+    
+    # 将生成的文本添加到对话历史记录
+    chat_history.append({"role": "assistant", "content": generated_text})
+    
     return generated_text
 
 # 创建Gradio界面
-iface = gr.Interface(fn=generate_response, inputs="text", outputs="text")
+iface = gr.Interface(
+    fn=generate_response,
+    inputs=gr.Textbox(lines=5, placeholder="請輸入你的問題..."),
+    outputs=gr.Textbox(lines=10),
+    title="醫療問答助手",
+    description="這是一個基於 MediaTek-Research/Breeze-7B-Instruct-v1_0 模型的醫療問答助手。",
+    theme="default",  # 可以选择不同的主题，例如 "huggingface"
+    live=True,
+)
 
 # 启动Gradio界面并共享链接
 iface.launch(share=True)