Update model_worker.py

model_worker.py

@@ -29,6 +29,8 @@ from transformers import TextIteratorStreamer
 from threading import Thread
 from icecream import ic
 
+import spaces # for use zero of huggingface
+
 
 GB = 1 << 30
 
@@ -70,7 +72,7 @@ class ModelWorker:
 
         self.is_multimodal = True
 
-
+    @spaces.GPU
     @torch.inference_mode()
     def generate_stream(self, params):
         tokenizer, model = self.tokenizer, self.model