"refactor: refactor_repeated_interfaces" (#32486)

Signed-off-by: tom-zju <tanjianpingzju1990@gmail.com>
Co-authored-by: Wentao Ye <44945378+yewentao256@users.noreply.github.com>

vllm/model_executor/model_loader/bitsandbytes_loader.py

@@ -22,6 +22,7 @@ from vllm.distributed import (
     get_tensor_model_parallel_world_size,
 )
 from vllm.logger import init_logger
+from vllm.lora.utils import is_moe_model
 from vllm.model_executor.layers.fused_moe import FusedMoE
 from vllm.model_executor.layers.linear import (
     LinearBase,
@@ -52,11 +53,6 @@ from vllm.utils.torch_utils import set_default_torch_dtype
 logger = init_logger(__name__)
 
 
-def is_moe_model(model: torch.nn.Module) -> bool:
-    """Checks if the model contains FusedMoE layers."""
-    return bool(any(isinstance(module, FusedMoE) for module in model.modules()))
-
-
 class BitsAndBytesModelLoader(BaseModelLoader):
     """Model loader to load model weights with BitsAndBytes quantization."""