Update app.py

app.py

@@ -3,10 +3,12 @@ import time
 import openvino_genai as ov_genai
 import numpy as np
 import gradio as gr
+import re
 
 # 下載模型
 model_id = "OpenVINO/Qwen3-0.6B-int4-ov"
 model_path = "Qwen3-0.6B-int4-ov"
+match = re.search(r"<think>(.*?)</think>(.*)", text, re.DOTALL)
 
 hf_hub.snapshot_download(model_id, local_dir=model_path, local_dir_use_symlinks=False)
 
@@ -18,9 +20,15 @@ tokenizer.set_chat_template(tokenizer.chat_template)
 
 
 def generate_response(prompt):
-    output = pipe.generate([prompt], max_length=1024)
-    tokenpersec=f'Throughput: {output.perf_metrics.get_throughput().mean:.2f} tokens/s'
-    return tokenpersec,output
+    generated = pipe.generate([prompt], max_length=1024)
+    tokenpersec=f'{output.perf_metrics.get_throughput().mean:.2f}'
+    match = re.search(r"<think>(.*?)</think>(.*)", generated, re.DOTALL)
+    
+    if match:
+        thinking = match.group(1).strip()
+        content = match.group(2).strip()  
+    return tokenpersec, thinking, content
+    
 
 # 建立 Gradio 介面
 demo = gr.Interface(