diff --git a/vllm/model_executor/layers/quantization/online/mxfp4.py b/vllm/model_executor/layers/quantization/online/mxfp4.py index 877819cfa1..3607a4bdf7 100644 --- a/vllm/model_executor/layers/quantization/online/mxfp4.py +++ b/vllm/model_executor/layers/quantization/online/mxfp4.py @@ -254,7 +254,6 @@ class Mxfp4OnlineMoEMethod(OnlineMoEMethodBase): mxfp4_backend=self.mxfp4_backend, experts_cls=self.experts_cls, routing_tables=layer._expert_routing_tables(), - layer=layer, ) def get_fused_moe_quant_config(