[NPU] fix transpose_value = False for NPU optimize_model=True (#1…

…2525)
intel-analytics · Dec 11, 2024 · 41ef497 · 41ef497
1 parent 588bfa2
commit 41ef497
Showing 1 changed file with 1 addition and 0 deletions.
diff --git a/python/llm/src/ipex_llm/transformers/npu_models/mp_models_base.py b/python/llm/src/ipex_llm/transformers/npu_models/mp_models_base.py
@@ -225,6 +225,7 @@ def attention(self,
             head_dim=head_dim,
         )
         new_key_states = key_states
+        new_value_states = value_states
 
         if mode == "decode":
             key_states = self.concat(past_key, key_states, axis=-2)