diff --git a/vllm/model_executor/layers/fused_moe/moe_torch_iterative.py b/vllm/model_executor/layers/fused_moe/moe_torch_iterative.py
index 481d2f3329e20..bcff55f4fdf16 100644
--- a/vllm/model_executor/layers/fused_moe/moe_torch_iterative.py
+++ b/vllm/model_executor/layers/fused_moe/moe_torch_iterative.py
@@ -48,4 +48,4 @@ def fused_moe(
         else:
             final_hidden_states = final_hidden_states + current_hidden_states
 
-    return final_hidden_states.view(orig_shape)
+    return final_hidden_states.view(orig_shape)  # type: ignore