update

app.py

@@ -11,13 +11,12 @@ load_dotenv()
 API_URL = os.getenv("API_URL")
 API_TOKEN = os.getenv("API_TOKEN")
 
-# 验证必要的环境变量
 if not API_URL or not API_TOKEN:
-    raise ValueError("make sure API_URL & API_TOKEN")
+    raise ValueError("invalid API_URL || API_TOKEN")
 
 print(f"[INFO] starting:")
 print(f"[INFO] API_URL: {API_URL[:6]}...{API_URL[-12:]}")
-print(f"[INFO] API_TOKEN: {API_TOKEN[:10]}...{API_TOKEN[-10:]}")  # 只显示token的前10位和后10位
+print(f"[INFO] API_TOKEN: {API_TOKEN[:10]}...{API_TOKEN[-10:]}")  
 
 """
 For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
@@ -56,10 +55,7 @@ def respond(
     }
 
     print(f"[INFO] process user msg...")
-    print(f"[INFO] sysMsg: {system_message}")
     print(f"[INFO] userMsg: {message}")
-    print(f"[INFO] modelParam: temperature={temperature}, top_p={top_p}")
-    print(f"[INFO] reqData: {data}")
 
     try:
         with requests.post(API_URL, headers=headers, json=data) as r:
@@ -68,8 +64,8 @@ def respond(
                 if 'choices' in json_response and len(json_response['choices']) > 0:
                     content = json_response['choices'][0].get('message', {}).get('content', '')
                     if content:
-                        # if '<think>' in content and '</think>' in content:
-                            # content = content.split('</think>')[-1].strip()
+                        if '<think>' in content and '</think>' in content:
+                            content = content.split('</think>')[-1].strip()
                         print(f"[INFO] response: {content}")
                         return content
             return "Service temporarily unavailable"
@@ -95,7 +91,7 @@ demo = gr.ChatInterface(
             label="Top-p (nucleus sampling)",
         ),
     ],
-    type="messages"  # 指定使用新的消息格式
+    type="messages"  
 )