fix

app.py

@@ -7,7 +7,6 @@ from server import setup_mixinputs, launch_vllm_server
 
 API_URL = "http://localhost:8000/v1/chat/completions"
 
-@spaces.GPU(duration=120)
 def chat_with_moi(message, history, temperature, top_p, beta):
     # Set the MIXINPUTS_BETA env var *per request*
     os.environ["MIXINPUTS_BETA"] = str(beta)

server.py

@@ -2,6 +2,7 @@ import subprocess
 import threading
 import os
 import time
+import spaces
 
 def setup_mixinputs():
     # Step 1: Run mixinputs setup