Update app.py

app.py

@@ -50,7 +50,7 @@ tokenizer, llava_model, image_processor, context_len = load_pretrained_model(
     args.load_4bit, 
     device=args.device)
 
-@spaces.GPU
+
 def bot_streaming(message, history):
     print ("triggered")
     print(message)
@@ -70,6 +70,8 @@ def bot_streaming(message, history):
         return
     
     streamer = TextIteratorStreamer(tokenizer, skip_prompt=True, skip_special_tokens=True)
+
+    @spaces.GPU
     def generate():
         print('Running chat')
         output = chat_llava(