diff --git a/vllm_ascend/ops/linear.py b/vllm_ascend/ops/linear.py index 3b6f7100..688520a3 100644 --- a/vllm_ascend/ops/linear.py +++ b/vllm_ascend/ops/linear.py @@ -299,7 +299,7 @@ class AscendRowParallelLinear(RowParallelLinear): def forward( self, input_, - is_prefill: bool = True, + **kwargs, ) -> Union[torch.Tensor, tuple[torch.Tensor, Optional[Parameter]]]: if self.custom_op is not None: return self.custom_op.apply(input_)