flash_attn_kernel.h

// Copyright (c) 2023 PaddlePaddle Authors. All Rights Reserved.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
//
//     http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.

#pragma once

#include "paddle/phi/core/dense_tensor.h"
#include "paddle/phi/core/device_context.h"

namespace phi {

template <typename T, typename Context>
void FlashAttnUnpaddedKernel(
    const Context& ctx,
    const DenseTensor& q,
    const DenseTensor& k,
    const DenseTensor& v,
    const DenseTensor& cu_seqlens_q,
    const DenseTensor& cu_seqlens_k,
    const paddle::optional<DenseTensor>& fixed_seed_offset,
    int64_t max_seqlen_q,
    int64_t max_seqlen_k,
    float scale,
    float dropout,
    bool causal,
    bool return_softmax,
    bool is_test,
    const std::string& rng_name,
    DenseTensor* out,
    DenseTensor* softmax,
    DenseTensor* softmax_lse,
    DenseTensor* seed_offset);

template <typename T, typename Context>
void FlashAttnKernel(const Context& ctx,
                     const DenseTensor& q,
                     const DenseTensor& k,
                     const DenseTensor& v,
                     const paddle::optional<DenseTensor>& fixed_seed_offset,
                     float dropout,
                     bool causal,
                     bool return_softmax,
                     bool is_test,
                     const std::string& rng_name,
                     DenseTensor* out,
                     DenseTensor* softmax,
                     DenseTensor* softmax_lse,
                     DenseTensor* seed_offset);

}  // namespace phi