1#pragma once
2
3// @generated by torchgen/gen.py from NativeFunction.h
4
5#include <c10/core/Scalar.h>
6#include <c10/core/Storage.h>
7#include <c10/core/TensorOptions.h>
8#include <c10/util/Deprecated.h>
9#include <c10/util/Optional.h>
10#include <c10/core/QScheme.h>
11#include <ATen/core/Reduction.h>
12#include <ATen/core/Tensor.h>
13#include <tuple>
14#include <vector>
15
16
17namespace at {
18namespace native {
19TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor,int64_t,int64_t,int64_t,int64_t,at::Tensor> _scaled_dot_product_flash_attention_cuda(const at::Tensor & query, const at::Tensor & key, const at::Tensor & value, double dropout_p=0.0, bool is_causal=false, bool return_debug_mask=false);
20TORCH_API ::std::tuple<at::Tensor,at::Tensor,at::Tensor,at::Tensor,int64_t,int64_t,int64_t,int64_t,at::Tensor> _scaled_dot_product_flash_attention_nestedtensor_cuda(const at::Tensor & query, const at::Tensor & key, const at::Tensor & value, double dropout_p=0.0, bool is_causal=false, bool return_debug_mask=false);
21} // namespace native
22} // namespace at
23