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
wmsofts
Paddle
提交
d723022e
P
Paddle
项目概览
wmsofts
/
Paddle
与 Fork 源项目一致
Fork自
PaddlePaddle / Paddle
通知
1
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
d723022e
编写于
6月 22, 2018
作者:
Y
Yancey1989
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
fix compile error
上级
56a903d3
变更
2
隐藏空白更改
内联
并排
Showing
2 changed file
with
9 addition
and
2 deletion
+9
-2
paddle/fluid/operators/listen_and_serv_op.cc
paddle/fluid/operators/listen_and_serv_op.cc
+5
-1
paddle/fluid/operators/send_recv_op_test.cc
paddle/fluid/operators/send_recv_op_test.cc
+4
-1
未找到文件。
paddle/fluid/operators/listen_and_serv_op.cc
浏览文件 @
d723022e
...
...
@@ -106,7 +106,11 @@ void ListenAndServOp::RunSyncLoop(
PADDLE_ENFORCE_GE
(
num_blocks
,
2
,
"server program should have at least 2 blocks"
);
auto
optimize_prepared
=
executor
->
Prepare
(
*
program
,
optimize_block_id_list
);
std
::
vector
<
int
>
optimize_blocks_idx
;
for
(
auto
blk
:
optimize_blocks
)
{
optimize_blocks_idx
.
push_back
(
blk
->
ID
());
}
auto
optimize_prepared
=
executor
->
Prepare
(
*
program
,
optimize_blocks_idx
);
// Insert placeholder for block0 which holds current op itself.
optimize_prepared
.
insert
(
optimize_prepared
.
begin
(),
...
...
paddle/fluid/operators/send_recv_op_test.cc
浏览文件 @
d723022e
...
...
@@ -129,7 +129,10 @@ void StartServerNet(bool is_sparse, std::atomic<bool> *initialized) {
// sub program run in listen_and_serv_op, for simple test we use sum
f
::
ProgramDesc
program
;
const
auto
&
root_block
=
program
.
Block
(
0
);
std
::
vector
<
framework
::
BlockDesc
*>
optimize_blocks
;
auto
*
optimize_block
=
program
.
AppendBlock
(
root_block
);
optimize_blocks
.
push_back
(
optimize_block
);
auto
*
prefetch_block
=
program
.
AppendBlock
(
root_block
);
// X for server side tensors, RX for received tensors, must be of same shape.
AddOp
(
"sum"
,
{{
"X"
,
{
"x0"
,
"x1"
}}},
{{
"Out"
,
{
"Out"
}}},
{},
optimize_block
,
...
...
@@ -139,7 +142,7 @@ void StartServerNet(bool is_sparse, std::atomic<bool> *initialized) {
attrs
.
insert
({
"Fanin"
,
1
});
attrs
.
insert
({
"ParamList"
,
std
::
vector
<
std
::
string
>
({
"Out"
})});
attrs
.
insert
({
"GradList"
,
std
::
vector
<
std
::
string
>
({
"x1"
})});
attrs
.
insert
({
"
OptimizeBlock"
,
optimize_block
});
attrs
.
insert
({
"
optimize_blocks"
,
optimize_blocks
});
attrs
.
insert
({
"PrefetchBlock"
,
prefetch_block
});
attrs
.
insert
({
"grad_to_block_id"
,
std
::
vector
<
std
::
string
>
({
""
})});
attrs
.
insert
({
"sync_mode"
,
true
});
...
...
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