max 30s

app.py

@@ -90,7 +90,7 @@ def load_model():
     except Exception as e:
         return f"❌ Error loading model: {str(e)}"
 
-@gpu_decorator
+@spaces.GPU(duration=30)
 def chat_predict(message, history, max_length, temperature, top_p, repetition_penalty, system_prompt):
     """Generate text using the loaded model with streaming and history"""
     global model, tokenizer