update

app.py

@@ -1,4 +1,4 @@
-import spaces
+#import spaces
 
 import os
 
@@ -92,7 +92,7 @@ class Chat:
         state.append_message(state.roles[1], None)
         return state
 
-    @spaces.GPU(duration=120)
+#    @spaces.GPU(duration=120)
     @torch.inference_mode()
     def generate(self, tensor: list, modals: list, prompt: str, first_run: bool, state, temperature, top_p, max_output_tokens):
         # TODO: support multiple turns of conversation.
@@ -132,7 +132,7 @@ class Chat:
         return outputs, state
 
 
-@spaces.GPU(duration=120)
+#@spaces.GPU(duration=120)
 def generate(image, video, state, state_, textbox_in, temperature, top_p, max_output_tokens, dtype=torch.float16):
     if not textbox_in:
         if len(state_.messages) > 0: