Update app.py

app.py

@@ -15,7 +15,7 @@ tokenizer = AutoTokenizer.from_pretrained(model_id, use_fast=True,)
 def respond(prompt, history):
     # 構建聊天模板
     messages = [
-        {"role": "system", "content": "用戶是繁體中文使用者. 包括think 回答限縮在1024token"},
+        {"role": "system", "content": "使用中文，直接回答用戶的問題，不要生成 <think> 區段。"},
         {"role": "user", "content": prompt}
     ]    
     text = tokenizer.apply_chat_template(
@@ -32,7 +32,10 @@ def respond(prompt, history):
     # 生成回應
     generated_ids = model.generate(
         **model_inputs,
-        max_new_tokens=2048
+        max_new_tokens=2048,
+        temperature=0.7,  # 降低隨機性
+        top_p=0.9,        # 限制生成的多樣性
+        do_sample=True    # 啟用採樣
     )
     print("Generated IDs:", generated_ids)
 
@@ -50,12 +53,12 @@ def respond(prompt, history):
 # 設定 Gradio 的聊天界面
 demo = gr.ChatInterface(
     fn=respond,
-    title="Qwen2.5-0.5B-Instruct-openvino-4bit",
-    description="Qwen2.5-0.5B-Instruct-openvino-4bit"
+    title="DeepSeek-R1-Distill-Qwen-1.5B-openvino",
+    description="DeepSeek-R1-Distill-Qwen-1.5B-openvino"
 )
 
 if __name__ == "__main__":
     print("Launching Gradio app...")
-    #demo.launch(server_name="0.0.0.0", server_port=7860)
-    demo.launch()
+    demo.launch(server_name="0.0.0.0", server_port=7860)
+    #demo.launch()