Update handler.py

handler.py

@@ -12,7 +12,7 @@ class EndpointHandler():
         self.model = LlavaForConditionalGeneration.from_pretrained(
             model_id, 
             torch_dtype=torch.float16, 
-            low_cpu_mem_usage=True, 
+            low_cpu_mem_usage=True,
             load_in_4bit=True
         )
         self.processor = AutoProcessor.from_pretrained(model_id)