elementwise_div_op.cu 1.7 KB
Newer Older
1
/* Copyright (c) 2016 PaddlePaddle Authors. All Rights Reserved.
G
gongweibao 已提交
2

L
Luo Tao 已提交
3 4 5
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
G
gongweibao 已提交
6

L
Luo Tao 已提交
7
    http://www.apache.org/licenses/LICENSE-2.0
G
gongweibao 已提交
8

L
Luo Tao 已提交
9 10 11 12 13
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. */
G
gongweibao 已提交
14 15

#define EIGEN_USE_GPU
Y
Yi Wang 已提交
16
#include "paddle/fluid/operators/elementwise_div_op.h"
17
#include "paddle/fluid/platform/float16.h"
G
gongweibao 已提交
18 19

namespace ops = paddle::operators;
20
namespace plat = paddle::platform;
G
gongweibao 已提交
21

Q
QI JUN 已提交
22
REGISTER_OP_CUDA_KERNEL(
G
gongweibao 已提交
23
    elementwise_div,
Q
QI JUN 已提交
24 25 26
    ops::ElementwiseDivKernel<paddle::platform::CUDADeviceContext, float>,
    ops::ElementwiseDivKernel<paddle::platform::CUDADeviceContext, double>,
    ops::ElementwiseDivKernel<paddle::platform::CUDADeviceContext, int>,
27 28 29
    ops::ElementwiseDivKernel<paddle::platform::CUDADeviceContext, int64_t>,
    ops::ElementwiseDivKernel<paddle::platform::CUDADeviceContext,
                              plat::float16>);
Q
QI JUN 已提交
30
REGISTER_OP_CUDA_KERNEL(
G
gongweibao 已提交
31
    elementwise_div_grad,
Q
QI JUN 已提交
32 33 34
    ops::ElementwiseDivGradKernel<paddle::platform::CUDADeviceContext, float>,
    ops::ElementwiseDivGradKernel<paddle::platform::CUDADeviceContext, double>,
    ops::ElementwiseDivGradKernel<paddle::platform::CUDADeviceContext, int>,
35
    ops::ElementwiseDivGradKernel<paddle::platform::CUDADeviceContext, int64_t>,
Q
QI JUN 已提交
36
    ops::ElementwiseDivGradKernel<paddle::platform::CUDADeviceContext,
37
                                  plat::float16>);