wehub-resource-sync
52 lines
1.4 KiB
C++
52 lines
1.4 KiB
C++
// 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/backends/gpu/gpu_context.h"
|
|
|
|
namespace phi {
|
|
namespace funcs {
|
|
|
|
// This functor involves a fusion calculation in Ernie or Bert.
|
|
// The fusion mode is as follows:
|
|
//
|
|
// | |
|
|
// matmul
|
|
// |
|
|
// eltwise_add
|
|
// |
|
|
// softmax /
|
|
// \ /
|
|
// matmul
|
|
// |
|
|
template <typename T>
|
|
class MultiheadGPUComputeFunctor {
|
|
public:
|
|
void operator()(const GPUContext &dev_ctx,
|
|
int batch,
|
|
int seq_len,
|
|
int head_num,
|
|
int head_size,
|
|
T *qkptr,
|
|
const T *bias_qk_ptr,
|
|
bool bias_is_mask,
|
|
T *tptr,
|
|
T alpha,
|
|
T beta);
|
|
};
|
|
|
|
} // namespace funcs
|
|
} // namespace phi
|