diff --git a/onnxruntime/contrib_ops/cuda/quantization/matmul_nbits.cc b/onnxruntime/contrib_ops/cuda/quantization/matmul_nbits.cc index 540fef7e7126f..2c3db2a3d2b47 100644 --- a/onnxruntime/contrib_ops/cuda/quantization/matmul_nbits.cc +++ b/onnxruntime/contrib_ops/cuda/quantization/matmul_nbits.cc @@ -40,7 +40,7 @@ Status MatMulNBits::PrepackedGemm( zero_points_ptr, zero_points_size, Y->MutableData(), Y->Shape().Size()); } -#endif // !USE_ROCM +#endif // !USE_ROCM template Status MatMulNBits::ComputeInternal(OpKernelContext* ctx) const { diff --git a/onnxruntime/contrib_ops/cuda/quantization/matmul_nbits.h b/onnxruntime/contrib_ops/cuda/quantization/matmul_nbits.h index 61761c566eaad..a8008e9cdcfa7 100644 --- a/onnxruntime/contrib_ops/cuda/quantization/matmul_nbits.h +++ b/onnxruntime/contrib_ops/cuda/quantization/matmul_nbits.h @@ -42,7 +42,7 @@ class MatMulNBits final : public CudaKernel { [[maybe_unused]] Tensor* Y) const { return ORT_MAKE_STATUS(ONNXRUNTIME, NOT_IMPLEMENTED, "Prepacked gemm is not supported for MatMulNBits op."); } -#endif // !USE_ROCM +#endif // !USE_ROCM Status ComputeInternal(OpKernelContext* context) const override;