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
机器未来
Paddle
提交
244efd6b
P
Paddle
项目概览
机器未来
/
Paddle
与 Fork 源项目一致
Fork自
PaddlePaddle / Paddle
通知
1
Star
1
Fork
0
代码
文件
提交
分支
Tags
贡献者
分支图
Diff
Issue
1
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
1
Issue
1
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
提交
244efd6b
编写于
9月 22, 2017
作者:
L
Luo Tao
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
refine doc of sequence_pool_op
上级
0449b9c8
变更
2
隐藏空白更改
内联
并排
Showing
2 changed file
with
23 addition
and
22 deletion
+23
-22
paddle/operators/sequence_pool_op.cc
paddle/operators/sequence_pool_op.cc
+21
-20
paddle/operators/sequence_pool_op.h
paddle/operators/sequence_pool_op.h
+2
-2
未找到文件。
paddle/operators/sequence_pool_op.cc
浏览文件 @
244efd6b
...
...
@@ -47,9 +47,11 @@ class SequencePoolOpMaker : public framework::OpProtoAndCheckerMaker {
SequencePoolOpMaker
(
framework
::
OpProto
*
proto
,
framework
::
OpAttrChecker
*
op_checker
)
:
OpProtoAndCheckerMaker
(
proto
,
op_checker
)
{
AddInput
(
"X"
,
"A LoDTensor, the variable-length input of SequencePoolOp"
);
AddOutput
(
"Out"
,
"A LoDTensor, the variable-length output of SequencePoolOp."
);
AddInput
(
"X"
,
"A float LoDTensor, the variable-length input of SequencePoolOp"
);
AddOutput
(
"Out"
,
"A float LoDTensor, the variable-length output of SequencePoolOp."
);
AddAttr
<
int
>
(
"strategy"
,
"(int, default AVERAGE) the pooling strategy of SequencePoolOp."
)
...
...
@@ -58,23 +60,22 @@ class SequencePoolOpMaker : public framework::OpProtoAndCheckerMaker {
AddComment
(
R"DOC(
SequencePoolOp pools features of all time-steps of each instance.
For a mini-batch of 3 variable lengths sentences, containing 2, 3, and 2 words:
X = [[1, 3], [2, 4, 6], [5, 1]],
and X->lod()[0] = [0, 2, 5, 7]
then, for different strategy, we get:
- AVERAGE: Out = [2, 4, 3], where 2=(1+3)/2, 4=(2+4+6)/3, 3=(5+1)/2
- SUM: Out = [4, 12, 6], where 4=1+3, 12=2+4+6, 6=5+1
- SQRT: Out = [2.82, 6.93, 4.24], where 2.82=(1+3)/sqrt(2), 6.93=(2+4+6)/sqrt(3),
4.24=(5+1)/sqrt(2)
- MAX: Out = [3, 6, 5], where 3=max(1,3), 6=max(2,4,6), 5=max(5,1)
- LAST: Out = [3, 6, 1], where 3=last(1,3), 6=last(2,4,6), 1=last(5,1)
- FIRST: Out = [1, 2, 5], where 1=first(1,3), 2=first(2,4,6), 5=first(5,1)
and X->lod() is nullptr.
For a mini-batch of 3 variable lengths sentences, containing 2, 3, and 2 time-steps:
Assume X is a [7,M,N] float LoDTensor, and X->lod()[0] = [0, 2, 5, 7].
Besides, for the sake of simplicity, we assume M=1 and N=1,
and the value of X = [[1, 3], [2, 4, 6], [5, 1]].
Thus, Out is a [3,1,1] float LoDTensor, but Out->lod() is nullptr.
And for different strategy, the value of Out is as follows:
- AVERAGE: [2, 4, 3], where 2=(1+3)/2, 4=(2+4+6)/3, 3=(5+1)/2
- SUM: [4, 12, 6], where 4=1+3, 12=2+4+6, 6=5+1
- SQRT: [2.82, 6.93, 4.24], where 2.82=(1+3)/sqrt(2),
6.93=(2+4+6)/sqrt(3), 4.24=(5+1)/sqrt(2)
- MAX: [3, 6, 5], where 3=max(1,3), 6=max(2,4,6), 5=max(5,1)
- LAST: [3, 6, 1], where 3=last(1,3), 6=last(2,4,6), 1=last(5,1)
- FIRST: [1, 2, 5], where 1=first(1,3), 2=first(2,4,6), 5=first(5,1)
)DOC"
);
}
};
...
...
paddle/operators/sequence_pool_op.h
浏览文件 @
244efd6b
...
...
@@ -67,7 +67,7 @@ class SequencePoolKernel : public framework::OpKernel {
out_e
.
device
(
place
)
=
in_e
.
sum
(
Eigen
::
array
<
int
,
1
>
({{
0
}}));
break
;
default:
LOG
(
FATAL
)
<<
"unsupported pooling strategy"
;
PADDLE_THROW
(
"unsupported pooling strategy"
)
;
}
}
}
...
...
@@ -105,7 +105,7 @@ class SequencePoolGradKernel : public framework::OpKernel {
in_g_e
.
device
(
place
)
=
(
out_g_e
).
broadcast
(
bcast
);
break
;
default:
LOG
(
FATAL
)
<<
"unsupported pooling strategy"
;
PADDLE_THROW
(
"unsupported pooling strategy"
)
;
}
}
}
...
...
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