Update app.py

app.py

@@ -50,20 +50,36 @@ def chat_stream(message: str, history: list, system_prompt: str) -> Iterator[str
     messages.append({"role": "user", "content": message})
     
     # 调用API进行流式对话
-    response = openai.ChatCompletion.create(
-        model="mini",
-        messages=messages,
-        stream=True
-    )
+    headers = {
+        "Authorization": f"Bearer {API_KEY}",
+        "Content-Type": "application/json"
+    }
+    data = {
+        "model": "gpt-3.5-turbo",
+        "messages": messages,
+        "stream": True
+    }
+    
+    response = requests.post(f"{API_BASE}/chat/completions", headers=headers, json=data, stream=True)
     
     collected_messages = []
-    for chunk in response:
-        if chunk and hasattr(chunk.choices[0].delta, "content"):
-            chunk_message = chunk.choices[0].delta.content
-            collected_messages.append(chunk_message)
-            partial_message = "".join(collected_messages)
-            formatted_message = format_think_tags(partial_message)
-            yield formatted_message
+    for line in response.iter_lines():
+        if line:
+            line = line.decode('utf-8')
+            if line.startswith("data: "):
+                json_str = line[6:]
+                if json_str.strip() == "[DONE]":
+                    break
+                try:
+                    chunk = json.loads(json_str)
+                    if chunk['choices'][0]['delta'].get('content'):
+                        chunk_message = chunk['choices'][0]['delta']['content']
+                        collected_messages.append(chunk_message)
+                        partial_message = "".join(collected_messages)
+                        formatted_message = format_think_tags(partial_message)
+                        yield formatted_message
+                except json.JSONDecodeError:
+                    continue
 
 # 自定义CSS样式
 custom_css = """
@@ -148,4 +164,4 @@ with gr.Blocks(css=custom_css) as demo:
 # 启动应用
 if __name__ == "__main__":
     demo.queue()
-    demo.launch(share=False, server_name="0.0.0.0", server_port=7860)
+    demo.launch()