diff --git a/python/llm/example/NPU/HF-Transformers-AutoModels/Model/llama2/README.md b/python/llm/example/NPU/HF-Transformers-AutoModels/Model/llama2/README.md
index dc289c8cc2c..ff4a9c1c059 100644
--- a/python/llm/example/NPU/HF-Transformers-AutoModels/Model/llama2/README.md
+++ b/python/llm/example/NPU/HF-Transformers-AutoModels/Model/llama2/README.md
@@ -21,6 +21,8 @@ pip install --pre --upgrade ipex-llm[xpu] --extra-index-url https://pytorch-exte
 
 # below command will install intel_npu_acceleration_library
 pip install intel-npu-acceleration-library==1.3
+
+pip install transformers==4.40
 ```
 
 ### 2. Runtime Configurations
diff --git a/python/llm/src/ipex_llm/transformers/npu_models/llama.py b/python/llm/src/ipex_llm/transformers/npu_models/llama.py
index 8a830d9e1a3..e06d61e16e8 100644
--- a/python/llm/src/ipex_llm/transformers/npu_models/llama.py
+++ b/python/llm/src/ipex_llm/transformers/npu_models/llama.py
@@ -106,7 +106,7 @@ def llama_model_forward(
     from ipex_llm.transformers.kv import DynamicNormalCache
     if use_cache and not isinstance(past_key_values, DynamicNormalCache):
         past_key_values = DynamicNormalCache.from_legacy_cache(past_key_values)
-        past_seen_tokens = past_key_values.set_seq_length()
+        past_seen_tokens = past_key_values.get_seq_length()
 
     if cache_position is None:
         cache_position = torch.arange(past_seen_tokens, past_seen_tokens + inputs_embeds.shape[1],