Skip to content
体验新版
项目
组织
正在加载...
登录
切换导航
打开侧边栏
wk1234123
Paddle
提交
58ac8f46
P
Paddle
项目概览
wk1234123
/
Paddle
与 Fork 源项目一致
Fork自
PaddlePaddle / Paddle
通知
4
Star
0
Fork
0
代码
文件
提交
分支
Tags
贡献者
分支图
Diff
Issue
0
列表
看板
标记
里程碑
合并请求
0
Wiki
0
Wiki
分析
仓库
DevOps
项目成员
Pages
P
Paddle
项目概览
项目概览
详情
发布
仓库
仓库
文件
提交
分支
标签
贡献者
分支图
比较
Issue
0
Issue
0
列表
看板
标记
里程碑
合并请求
0
合并请求
0
Pages
分析
分析
仓库分析
DevOps
Wiki
0
Wiki
成员
成员
收起侧边栏
关闭侧边栏
动态
分支图
创建新Issue
提交
Issue看板
体验新版 GitCode,发现更多精彩内容 >>
提交
58ac8f46
编写于
9月 26, 2017
作者:
Y
Yibing Liu
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
apply more general dims for multiplex_op
上级
089f8e2d
变更
3
隐藏空白更改
内联
并排
Showing
3 changed file
with
8 addition
and
8 deletion
+8
-8
paddle/operators/multiplex_op.cc
paddle/operators/multiplex_op.cc
+4
-4
paddle/operators/multiplex_op.cu
paddle/operators/multiplex_op.cu
+2
-2
paddle/operators/multiplex_op.h
paddle/operators/multiplex_op.h
+2
-2
未找到文件。
paddle/operators/multiplex_op.cc
浏览文件 @
58ac8f46
...
...
@@ -44,7 +44,8 @@ class MultiplexOp : public framework::OperatorWithKernel {
"one candidate input tensors."
);
auto
in_dim
=
ins
[
0
]
->
dims
();
PADDLE_ENFORCE
(
in_dim
.
size
()
==
2
,
"Candidate tensors must be matrix."
);
PADDLE_ENFORCE
(
in_dim
.
size
()
>=
2
,
"The rank of candidate tensors must be not less than 2."
);
for
(
size_t
i
=
1
;
i
<
num_ins
;
i
++
)
{
auto
dim
=
ins
[
i
]
->
dims
();
PADDLE_ENFORCE
(
in_dim
==
dim
,
...
...
@@ -65,8 +66,7 @@ class MultiplexOpMaker : public framework::OpProtoAndCheckerMaker {
AddOutput
(
"Out"
,
"The output tensor of multiplex operator."
);
AddComment
(
R"DOC(Multiplex operator
Multiplex multiple tensors according to the index provided by the first
input tensor.
Multiplex multiple tensors according to the index provided by the index tensor.
Ids: the index tensor.
X[0 : N - 1]: the candidate tensors for output (N >= 2).
...
...
@@ -75,7 +75,7 @@ the (Ids[i])-th tensor.
For i-th row of the output tensor:
y[i]
[j] = x_{k}[i][j], j = 0,1, ... , (x_{0}.width - 1)
y[i]
= x_{k}[i]
where y is the output tensor. `x_{k}` is the k-th input tensor
and `k = Ids[i]`.
...
...
paddle/operators/multiplex_op.cu
浏览文件 @
58ac8f46
...
...
@@ -30,7 +30,7 @@ class MultiplexGPUKernel : public framework::OpKernel {
out
->
mutable_data
<
T
>
(
ctx
.
GetPlace
());
auto
rows
=
ins
[
0
]
->
dims
()[
0
];
auto
cols
=
ins
[
0
]
->
dims
()[
1
]
;
auto
cols
=
ins
[
0
]
->
numel
()
/
rows
;
// copy index to cpu
Tensor
index_t_cpu
;
index_t_cpu
.
CopyFrom
<
int32_t
>
(
*
ids
,
platform
::
CPUPlace
());
...
...
@@ -67,7 +67,7 @@ class MultiplexGradGPUKernel : public framework::OpKernel {
}
auto
rows
=
ins
[
0
]
->
dims
()[
0
];
auto
cols
=
ins
[
0
]
->
dims
()[
1
]
;
auto
cols
=
ins
[
0
]
->
numel
()
/
rows
;
// copy index to cpu
Tensor
index_t_cpu
;
index_t_cpu
.
CopyFrom
<
int32_t
>
(
*
ids
,
platform
::
CPUPlace
());
...
...
paddle/operators/multiplex_op.h
浏览文件 @
58ac8f46
...
...
@@ -33,7 +33,7 @@ class MultiplexCPUKernel : public framework::OpKernel {
out
->
mutable_data
<
T
>
(
ctx
.
GetPlace
());
auto
rows
=
ins
[
0
]
->
dims
()[
0
];
auto
cols
=
ins
[
0
]
->
dims
()[
1
]
;
auto
cols
=
ins
[
0
]
->
numel
()
/
rows
;
auto
index
=
ids
->
data
<
int32_t
>
();
Place
place
=
boost
::
get
<
Place
>
(
ctx
.
GetPlace
());
for
(
auto
i
=
0
;
i
<
rows
;
i
++
)
{
...
...
@@ -65,7 +65,7 @@ class MultiplexGradCPUKernel : public framework::OpKernel {
}
auto
rows
=
ins
[
0
]
->
dims
()[
0
];
auto
cols
=
ins
[
0
]
->
dims
()[
1
]
;
auto
cols
=
ins
[
0
]
->
numel
()
/
rows
;
auto
*
index
=
ids
->
data
<
int32_t
>
();
Place
place
=
boost
::
get
<
Place
>
(
ctx
.
GetPlace
());
for
(
auto
i
=
0
;
i
<
rows
;
i
++
)
{
...
...
编辑
预览
Markdown
is supported
0%
请重试
或
添加新附件
.
添加附件
取消
You are about to add
0
people
to the discussion. Proceed with caution.
先完成此消息的编辑!
取消
想要评论请
注册
或
登录