diff --git a/python/llm/src/ipex_llm/transformers/npu_models/qwen2_mp.py b/python/llm/src/ipex_llm/transformers/npu_models/qwen2_mp.py
index ed5fbd30cd37..7a61ad9d24b9 100644
--- a/python/llm/src/ipex_llm/transformers/npu_models/qwen2_mp.py
+++ b/python/llm/src/ipex_llm/transformers/npu_models/qwen2_mp.py
@@ -39,7 +39,6 @@
 from transformers.modeling_outputs import BaseModelOutputWithPast
 from ipex_llm.transformers.npu_models.mp_models_base import run_model
 from ipex_llm.transformers.npu_models.mp_models_base import LLMBaseNNFactory
-from transformers.models.qwen2.modeling_qwen2 import Qwen2MLP
 
 
 class LowBitQwenMultiDecoderlayer(LLMBaseNNFactory):