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
BaiXuePrincess
Paddle
提交
b94584cf
P
Paddle
项目概览
BaiXuePrincess
/
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
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
提交
b94584cf
编写于
8月 01, 2017
作者:
D
dangqingqing
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
Rename recurrent_network_op recurrent_op.
上级
79e89ef4
变更
5
隐藏空白更改
内联
并排
Showing
5 changed file
with
13 addition
and
10 deletion
+13
-10
paddle/operators/CMakeLists.txt
paddle/operators/CMakeLists.txt
+2
-4
paddle/operators/recurrent_op.cc
paddle/operators/recurrent_op.cc
+9
-4
paddle/operators/recurrent_op.h
paddle/operators/recurrent_op.h
+0
-0
paddle/operators/recurrent_op_test.cc
paddle/operators/recurrent_op_test.cc
+1
-1
paddle/pybind/CMakeLists.txt
paddle/pybind/CMakeLists.txt
+1
-1
未找到文件。
paddle/operators/CMakeLists.txt
浏览文件 @
b94584cf
...
...
@@ -55,7 +55,5 @@ op_library(fc_op SRCS fc_op.cc DEPS mul_op rowwise_add_op sigmoid_op
op_library
(
sgd_op SRCS sgd_op.cc sgd_op.cu
)
op_library
(
recurrent_network_op SRCS recurrent_network_op.cc DEPS op_desc
tensor op_registry operator net
)
cc_test
(
recurrent_network_op_test SRCS recurrent_network_op_test.cc DEPS
recurrent_network_op gtest mul_op add_op
)
op_library
(
recurrent_op SRCS recurrent_op.cc DEPS op_desc tensor op_registry operator net
)
cc_test
(
recurrent_op_test SRCS recurrent_op_test.cc DEPS recurrent_op gtest mul_op add_op
)
paddle/operators/recurrent_
network_
op.cc
→
paddle/operators/recurrent_op.cc
浏览文件 @
b94584cf
...
...
@@ -12,7 +12,7 @@
See the License for the specific language governing permissions and
limitations under the License. */
#include "paddle/operators/recurrent_
network_
op.h"
#include "paddle/operators/recurrent_op.h"
#include <glog/logging.h>
#include <cstring>
...
...
@@ -108,8 +108,13 @@ void LinkMemories(std::vector<std::shared_ptr<Scope>>& scopes,
std
::
shared_ptr
<
Scope
>
scope
=
scopes
[
step_id
];
std
::
shared_ptr
<
Scope
>
linked_scope
=
scopes
[
step_id
+
offset
];
for
(
auto
&
attr
:
memories
)
{
PADDLE_ENFORCE
(
scope
->
HasVariable
(
attr
.
pre_var
),
"the pre-memory [%s] is not in scope."
,
attr
.
pre_var
);
PADDLE_ENFORCE
(
linked_scope
->
HasVariable
(
attr
.
var
),
"the memory [%s] is not in linked scope."
,
attr
.
var
);
auto
mem
=
scope
->
GetVariable
(
attr
.
pre_var
)
->
GetMutable
<
Tensor
>
();
// maybe share variable is better?
auto
linked_mem
=
linked_scope
->
GetVariable
(
attr
.
var
)
->
GetMutable
<
Tensor
>
();
if
(
infer_shape_mode
)
{
mem
->
Resize
(
linked_mem
->
dims
());
...
...
@@ -295,12 +300,12 @@ public:
const
auto
&
name
=
RecurrentOp
::
kArgName
;
// inputs and outputs stored in proto
AddInputs
(
name
.
inlinks
,
"the input that need to be segmented for each step."
);
"the input
s
that need to be segmented for each step."
);
AddInputs
(
name
.
boot_memories
,
"variables to initialize memories."
);
AddInput
(
name
.
step_net
,
"network shared by all steps."
);
AddOutputs
(
name
.
outlinks
,
"the output that need to concated for all steps."
);
"the output
s
that need to concated for all steps."
);
AddOutput
(
name
.
step_scopes
,
"step scopes"
);
// Attributes stored in AttributeMap
...
...
paddle/operators/recurrent_
network_
op.h
→
paddle/operators/recurrent_op.h
浏览文件 @
b94584cf
文件已移动
paddle/operators/recurrent_
network_
op_test.cc
→
paddle/operators/recurrent_op_test.cc
浏览文件 @
b94584cf
...
...
@@ -18,7 +18,7 @@
#include "paddle/framework/op_registry.h"
#include "paddle/framework/operator.h"
#include "paddle/framework/tensor.h"
#include "paddle/operators/recurrent_
network_
op.h"
#include "paddle/operators/recurrent_op.h"
namespace
paddle
{
namespace
operators
{
...
...
paddle/pybind/CMakeLists.txt
浏览文件 @
b94584cf
cc_library
(
paddle_pybind SHARED SRCS pybind.cc DEPS pybind python
add_op fc_op sgd_op cross_entropy_op recurrent_
network_
op
)
add_op fc_op sgd_op cross_entropy_op recurrent_op
)
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