elementwise_sub_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. */
W
Wu Yi 已提交
14
#include "paddle/fluid/operators/elementwise/elementwise_sub_op.h"
15
#include "paddle/fluid/platform/float16.h"
G
gongweibao 已提交
16 17 18

namespace ops = paddle::operators;

Q
QI JUN 已提交
19
REGISTER_OP_CUDA_KERNEL(
G
gongweibao 已提交
20
    elementwise_sub,
Q
QI JUN 已提交
21
    ops::ElementwiseSubKernel<paddle::platform::CUDADeviceContext, float>,
22 23
    ops::ElementwiseSubKernel<paddle::platform::CUDADeviceContext,
                              paddle::platform::float16>,
Q
QI JUN 已提交
24 25
    ops::ElementwiseSubKernel<paddle::platform::CUDADeviceContext, double>,
    ops::ElementwiseSubKernel<paddle::platform::CUDADeviceContext, int>,
26
    ops::ElementwiseSubKernel<paddle::platform::CUDADeviceContext, int64_t>);
Q
QI JUN 已提交
27
REGISTER_OP_CUDA_KERNEL(
G
gongweibao 已提交
28
    elementwise_sub_grad,
Q
QI JUN 已提交
29
    ops::ElementwiseSubGradKernel<paddle::platform::CUDADeviceContext, float>,
30 31
    ops::ElementwiseSubGradKernel<paddle::platform::CUDADeviceContext,
                                  paddle::platform::float16>,
Q
QI JUN 已提交
32 33 34 35
    ops::ElementwiseSubGradKernel<paddle::platform::CUDADeviceContext, double>,
    ops::ElementwiseSubGradKernel<paddle::platform::CUDADeviceContext, int>,
    ops::ElementwiseSubGradKernel<paddle::platform::CUDADeviceContext,
                                  int64_t>);