fix cuda inference code

README.md

@@ -25,6 +25,8 @@ please note int2 **may be slower** than int4 on CUDA due to kernel issue.
 ~~~python
 import transformers
 from transformers import AutoModelForCausalLM, AutoTokenizer
+from auto_round import AutoRoundConfig ##must import for auto-round format
+
 
 #  https://github.com/huggingface/transformers/pull/35493
 def set_initialized_submodules(model, state_dict_keys):