Update app.py

app.py

@@ -83,7 +83,7 @@ def format_prompt(message, history):
     prompt += f"[INST] {message} [/INST]"
     return prompt
 
-def generate(prompt, history=[], temperature=0.1, max_new_tokens=24000, top_p=0.95, repetition_penalty=1.0):
+def generate(prompt, history=[], temperature=0.1, max_new_tokens=25000, top_p=0.95, repetition_penalty=1.0):
     input_tokens = len(tokenizer.encode(prompt))
     available_tokens = 32768 - input_tokens
     max_new_tokens = min(max_new_tokens, available_tokens)
@@ -100,12 +100,20 @@ def generate(prompt, history=[], temperature=0.1, max_new_tokens=24000, top_p=0.
             if isinstance(response, dict) and 'generated_text' in response:
                 output += response['generated_text']
             else:
-                output += str(response)
+                output += str(response)  # 이 부분을 문자열로 처리
         return output, f"Used tokens: {input_tokens + max_new_tokens}"
     except Exception as e:
         return f"Error: {str(e)}", "Used tokens: 0"
 
+demo = gr.Interface(
+    fn=generate,
+    inputs=[gr.Textbox(label="질문을 입력하세요", placeholder="여기에 질문을 입력하세요...", lines=2), gr.JSON(label="History", value=[])],
+    outputs=[gr.Markdown(), gr.Label()],
+    title="AIQ 코드파일럿: OpenLLM v1.12",
+    description="AIQ Codepilot과 상호작용해 보세요."
+)
 
+demo.launch(show_api=False)
 
 
 mychatbot = gr.Chatbot(