Update app.py

app.py

@@ -11,14 +11,11 @@ def get_model_response(client, messages, max_tokens, temperature, top_p):
             max_tokens=max_tokens,
             temperature=temperature,
             top_p=top_p,
-            stream=True
+            stream=False  # 스트리밍을 비활성화하여 전체 응답을 한 번에 받습니다.
         )
-        for message in response:
-            token = message.choices[0].delta.content if hasattr(message.choices[0], 'delta') else message.choices[0].text
-            if token:
-                yield token
+        return response.choices[0].message.content
     except Exception as e:
-        yield f"모델 추론 실패: {str(e)}"
+        return f"모델 추론 실패: {str(e)}"
 
 def respond(message, history, system_message, max_tokens, temperature, top_p, selected_model):
     try:
@@ -28,15 +25,12 @@ def respond(message, history, system_message, max_tokens, temperature, top_p, se
         messages.extend([{"role": "user" if i % 2 == 0 else "assistant", "content": m} for h in history for i, m in enumerate(h) if m])
         messages.append({"role": "user", "content": message})
         
-        response = ""
-        for token in get_model_response(client, messages, max_tokens, temperature, top_p):
-            response += token
-            yield response
+        response = get_model_response(client, messages, max_tokens, temperature, top_p)
         
-        if not response:
-            yield "모델이 응답을 생성하지 못했습니다. 다른 입력이나 모델을 시도해보세요."
+        history.append((message, response))
+        return "", history
     except Exception as e:
-        yield f"오류 발생: {str(e)}"
+        return "", history + [(message, f"오류 발생: {str(e)}")]
 
 models = {
     "deepseek-ai/DeepSeek-Coder-V2-Instruct": "DeepSeek-Coder-V2-Instruct",
@@ -66,10 +60,10 @@ with gr.Blocks() as demo:
         top_p = gr.Slider(minimum=0.1, maximum=1.0, value=0.90, step=0.05, label="Top-p (핵 샘플링)")
         model = gr.Radio(list(models.keys()), value=list(models.keys())[0], label="언어 모델 선택", info="사용할 언어 모델을 선택하세요")
 
-    send.click(respond, inputs=[msg, chatbot, system_message, max_tokens, temperature, top_p, model], outputs=chatbot)
-    msg.submit(respond, inputs=[msg, chatbot, system_message, max_tokens, temperature, top_p, model], outputs=chatbot)
-    regenerate.click(respond, inputs=[msg, chatbot, system_message, max_tokens, temperature, top_p, model], outputs=chatbot)
-    clear.click(lambda: None, None, chatbot, queue=False)
+    send.click(respond, inputs=[msg, chatbot, system_message, max_tokens, temperature, top_p, model], outputs=[msg, chatbot])
+    msg.submit(respond, inputs=[msg, chatbot, system_message, max_tokens, temperature, top_p, model], outputs=[msg, chatbot])
+    regenerate.click(lambda h: respond(h[-1][0] if h else "", h[:-1], *h.values()), inputs=[chatbot, system_message, max_tokens, temperature, top_p, model], outputs=[msg, chatbot])
+    clear.click(lambda: (None, None), outputs=[msg, chatbot])
 
 if __name__ == "__main__":
     if not hf_token: