|
|
|
@ -77,15 +77,6 @@ void matmul(const framework::Tensor& matrix_a, bool trans_a,
|
|
|
|
|
framework::Tensor* matrix_out, T beta,
|
|
|
|
|
platform::DeviceContext* context);
|
|
|
|
|
|
|
|
|
|
// // matrix multiply with continuous memory
|
|
|
|
|
// template <typename Place, typename T>
|
|
|
|
|
// void matmul(const framework::Tensor& matrix_a, bool trans_a,
|
|
|
|
|
// const framework::Tensor& matrix_b, bool trans_b,
|
|
|
|
|
// framework::Tensor* matrix_out,
|
|
|
|
|
// platform::DeviceContext* context) {
|
|
|
|
|
// matmul(matrix_a, matrix_b, trans_a, trans_b, 1, matrix_out, 0, context);
|
|
|
|
|
// }
|
|
|
|
|
|
|
|
|
|
} // namespace math
|
|
|
|
|
} // namespace operators
|
|
|
|
|
} // namespace paddle
|
|
|
|
|