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
PaddlePaddle
PaddleRec
提交
b23282f3
P
PaddleRec
项目概览
PaddlePaddle
/
PaddleRec
通知
68
Star
12
Fork
5
代码
文件
提交
分支
Tags
贡献者
分支图
Diff
Issue
27
列表
看板
标记
里程碑
合并请求
10
Wiki
1
Wiki
分析
仓库
DevOps
项目成员
Pages
P
PaddleRec
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
27
Issue
27
列表
看板
标记
里程碑
合并请求
10
合并请求
10
Pages
分析
分析
仓库分析
DevOps
Wiki
1
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
b23282f3
编写于
9月 18, 2019
作者:
X
xiexionghang
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
fix shuffler bug
上级
7654080c
变更
3
显示空白变更内容
内联
并排
Showing
3 changed file
with
43 addition
and
12 deletion
+43
-12
paddle/fluid/train/custom_trainer/feed/process/learner_process.cc
...luid/train/custom_trainer/feed/process/learner_process.cc
+4
-0
paddle/fluid/train/custom_trainer/feed/shuffler/shuffler.cc
paddle/fluid/train/custom_trainer/feed/shuffler/shuffler.cc
+17
-7
paddle/fluid/train/custom_trainer/feed/trainer_context.h
paddle/fluid/train/custom_trainer/feed/trainer_context.h
+22
-5
未找到文件。
paddle/fluid/train/custom_trainer/feed/process/learner_process.cc
浏览文件 @
b23282f3
...
@@ -77,6 +77,7 @@ int LearnerProcess::update_cache_model(uint64_t epoch_id, ModelSaveWay way) {
...
@@ -77,6 +77,7 @@ int LearnerProcess::update_cache_model(uint64_t epoch_id, ModelSaveWay way) {
return
0
;
return
0
;
}
}
int
LearnerProcess
::
wait_save_model
(
uint64_t
epoch_id
,
ModelSaveWay
way
,
bool
is_force_dump
)
{
int
LearnerProcess
::
wait_save_model
(
uint64_t
epoch_id
,
ModelSaveWay
way
,
bool
is_force_dump
)
{
ContextStatusGurad
status_guard
(
_context_ptr
,
TrainerStatus
::
Saving
);
auto
fs
=
_context_ptr
->
file_system
;
auto
fs
=
_context_ptr
->
file_system
;
auto
*
ps_client
=
_context_ptr
->
pslib
->
ps_client
();
auto
*
ps_client
=
_context_ptr
->
pslib
->
ps_client
();
auto
*
environment
=
_context_ptr
->
environment
.
get
();
auto
*
environment
=
_context_ptr
->
environment
.
get
();
...
@@ -154,6 +155,7 @@ int LearnerProcess::load_model(uint64_t epoch_id) {
...
@@ -154,6 +155,7 @@ int LearnerProcess::load_model(uint64_t epoch_id) {
if
(
!
environment
->
is_master_node
(
EnvironmentRole
::
WORKER
))
{
if
(
!
environment
->
is_master_node
(
EnvironmentRole
::
WORKER
))
{
return
0
;
return
0
;
}
}
VLOG
(
2
)
<<
"Start Load Model"
;
auto
*
fs
=
_context_ptr
->
file_system
.
get
();
auto
*
fs
=
_context_ptr
->
file_system
.
get
();
std
::
set
<
uint32_t
>
loaded_table_set
;
std
::
set
<
uint32_t
>
loaded_table_set
;
auto
model_dir
=
_context_ptr
->
epoch_accessor
->
checkpoint_path
();
auto
model_dir
=
_context_ptr
->
epoch_accessor
->
checkpoint_path
();
...
@@ -177,6 +179,7 @@ int LearnerProcess::load_model(uint64_t epoch_id) {
...
@@ -177,6 +179,7 @@ int LearnerProcess::load_model(uint64_t epoch_id) {
loaded_table_set
.
insert
(
itr
.
first
);
loaded_table_set
.
insert
(
itr
.
first
);
}
}
}
}
VLOG
(
2
)
<<
"Finish Load Model"
;
return
0
;
return
0
;
}
}
...
@@ -223,6 +226,7 @@ int LearnerProcess::run() {
...
@@ -223,6 +226,7 @@ int LearnerProcess::run() {
//Step2. 运行训练网络
//Step2. 运行训练网络
{
{
ContextStatusGurad
status_guard
(
_context_ptr
,
TrainerStatus
::
Training
);
std
::
map
<
std
::
string
,
paddle
::
framework
::
Channel
<
DataItem
>>
backup_input_map
;
std
::
map
<
std
::
string
,
paddle
::
framework
::
Channel
<
DataItem
>>
backup_input_map
;
for
(
auto
&
executor
:
_executors
)
{
for
(
auto
&
executor
:
_executors
)
{
environment
->
barrier
(
EnvironmentRole
::
WORKER
);
environment
->
barrier
(
EnvironmentRole
::
WORKER
);
...
...
paddle/fluid/train/custom_trainer/feed/shuffler/shuffler.cc
浏览文件 @
b23282f3
...
@@ -39,8 +39,8 @@ public:
...
@@ -39,8 +39,8 @@ public:
virtual
int
initialize
(
YAML
::
Node
config
,
virtual
int
initialize
(
YAML
::
Node
config
,
std
::
shared_ptr
<
TrainerContext
>
context_ptr
)
{
std
::
shared_ptr
<
TrainerContext
>
context_ptr
)
{
Shuffler
::
initialize
(
config
,
context_ptr
);
Shuffler
::
initialize
(
config
,
context_ptr
);
_max_concurrent_num
=
config
[
"max_concurrent_num"
].
as
<
int
>
(
4
);
// 最大并发发送数
_max_concurrent_num
=
config
[
"max_concurrent_num"
].
as
<
int
>
(
6
);
// 最大并发发送数
_max_package_size
=
config
[
"max_package_size"
].
as
<
int
>
(
1024
);
// 最大包个数,一次发送package个数据
_max_package_size
=
config
[
"max_package_size"
].
as
<
int
>
(
256
);
// 最大包个数,一次发送package个数据
_shuffle_data_msg_type
=
config
[
"shuffle_data_msg_type"
].
as
<
int
>
(
3
);
// c2c msg type
_shuffle_data_msg_type
=
config
[
"shuffle_data_msg_type"
].
as
<
int
>
(
3
);
// c2c msg type
_finish_msg_type
=
config
[
"finish_msg_type"
].
as
<
int
>
(
4
);
// c2c msg type
_finish_msg_type
=
config
[
"finish_msg_type"
].
as
<
int
>
(
4
);
// c2c msg type
...
@@ -62,6 +62,8 @@ public:
...
@@ -62,6 +62,8 @@ public:
data_channel
.
swap
(
input_channel
);
data_channel
.
swap
(
input_channel
);
set_channel
(
data_channel
);
set_channel
(
data_channel
);
_item_send_count
=
0
;
_item_receive_count
=
0
;
auto
*
environment
=
_trainer_context
->
environment
.
get
();
auto
*
environment
=
_trainer_context
->
environment
.
get
();
auto
worker_num
=
environment
->
node_num
(
EnvironmentRole
::
WORKER
);
auto
worker_num
=
environment
->
node_num
(
EnvironmentRole
::
WORKER
);
std
::
vector
<
std
::
vector
<
std
::
future
<
int
>>>
waits
(
concurrent_num
);
std
::
vector
<
std
::
vector
<
std
::
future
<
int
>>>
waits
(
concurrent_num
);
...
@@ -86,8 +88,9 @@ public:
...
@@ -86,8 +88,9 @@ public:
status
=
1
;
status
=
1
;
break
;
break
;
}
}
_item_send_count
+=
read_size
;
for
(
int
i
=
0
;
i
<
worker_num
;
++
i
)
{
for
(
int
i
=
0
;
i
<
worker_num
;
++
i
)
{
send_buffer_worker
.
clear
();
send_buffer_worker
[
i
]
.
clear
();
}
}
for
(
int
i
=
0
;
i
<
read_size
;
++
i
)
{
for
(
int
i
=
0
;
i
<
read_size
;
++
i
)
{
auto
worker_idx
=
_shuffle_key_func
(
send_buffer
[
i
].
id
)
%
worker_num
;
auto
worker_idx
=
_shuffle_key_func
(
send_buffer
[
i
].
id
)
%
worker_num
;
...
@@ -119,19 +122,19 @@ public:
...
@@ -119,19 +122,19 @@ public:
}
}
}
}
}
}
VLOG
(
5
)
<<
"start send finish, worker_num: "
<<
worker_num
;
VLOG
(
2
)
<<
"start send finish, worker_num: "
<<
worker_num
;
waits
[
0
].
clear
();
waits
[
0
].
clear
();
for
(
int
i
=
0
;
i
<
worker_num
;
++
i
)
{
for
(
int
i
=
0
;
i
<
worker_num
;
++
i
)
{
waits
[
0
].
push_back
(
send_finish
(
i
));
waits
[
0
].
push_back
(
send_finish
(
i
));
}
}
VLOG
(
5
)
<<
"wait all finish"
;
VLOG
(
2
)
<<
"wait all finish"
;
for
(
int
i
=
0
;
i
<
worker_num
;
++
i
)
{
for
(
int
i
=
0
;
i
<
worker_num
;
++
i
)
{
if
(
waits
[
0
][
i
].
get
()
!=
0
)
{
if
(
waits
[
0
][
i
].
get
()
!=
0
)
{
LOG
(
WARNING
)
<<
"fail to send finish "
<<
i
;
LOG
(
WARNING
)
<<
"fail to send finish "
<<
i
;
status
=
-
1
;
status
=
-
1
;
}
}
}
}
VLOG
(
5
)
<<
"finish shuffler, status: "
<<
status
;
VLOG
(
2
)
<<
"finish shuffler_send_channel, total_send:"
<<
_item_send_count
;
return
status
<
0
?
status
:
0
;
return
status
<
0
?
status
:
0
;
}
}
...
@@ -174,6 +177,7 @@ private:
...
@@ -174,6 +177,7 @@ private:
// 同步所有worker,在所有写入完成后,c2c_msg返回前,重置channel
// 同步所有worker,在所有写入完成后,c2c_msg返回前,重置channel
if
(
wait_num
==
0
)
{
if
(
wait_num
==
0
)
{
reset_channel
();
reset_channel
();
VLOG
(
2
)
<<
"finish shuffle_receive_channel, receive_count: "
<<
_item_receive_count
;
_wait_num_mutex
.
unlock
();
_wait_num_mutex
.
unlock
();
}
else
{
}
else
{
std
::
lock_guard
<
bthread
::
Mutex
>
lock
(
_wait_num_mutex
);
std
::
lock_guard
<
bthread
::
Mutex
>
lock
(
_wait_num_mutex
);
...
@@ -182,7 +186,7 @@ private:
...
@@ -182,7 +186,7 @@ private:
}
}
int32_t
write_to_channel
(
std
::
vector
<
DataItem
>&&
items
)
{
int32_t
write_to_channel
(
std
::
vector
<
DataItem
>&&
items
)
{
size_t
items_size
=
items
.
size
();
size_t
items_size
=
items
.
size
();
VLOG
(
5
)
<<
"write_to_channel, items_size: "
<<
items_size
;
_item_receive_count
+=
items_size
;
return
_out_channel
->
Write
(
std
::
move
(
items
))
==
items_size
?
0
:
-
1
;
return
_out_channel
->
Write
(
std
::
move
(
items
))
==
items_size
?
0
:
-
1
;
}
}
...
@@ -207,6 +211,8 @@ private:
...
@@ -207,6 +211,8 @@ private:
}
}
std
::
future
<
int32_t
>
send_shuffle_data
(
int
to_client_id
,
std
::
vector
<
DataItem
>&
items
)
{
std
::
future
<
int32_t
>
send_shuffle_data
(
int
to_client_id
,
std
::
vector
<
DataItem
>&
items
)
{
// server端也开启了client, worker节点为偶数编号
to_client_id
=
2
*
to_client_id
;
VLOG
(
5
)
<<
"send_shuffle_data, to_client_id: "
<<
to_client_id
<<
", items_size: "
<<
items
.
size
();
VLOG
(
5
)
<<
"send_shuffle_data, to_client_id: "
<<
to_client_id
<<
", items_size: "
<<
items
.
size
();
paddle
::
framework
::
BinaryArchive
ar
;
paddle
::
framework
::
BinaryArchive
ar
;
ar
<<
items
;
ar
<<
items
;
...
@@ -215,6 +221,8 @@ private:
...
@@ -215,6 +221,8 @@ private:
}
}
std
::
future
<
int32_t
>
send_finish
(
int
to_client_id
)
{
std
::
future
<
int32_t
>
send_finish
(
int
to_client_id
)
{
// server端也开启了client, worker节点为偶数编号
to_client_id
=
2
*
to_client_id
;
VLOG
(
5
)
<<
"send_finish, to_client_id: "
<<
to_client_id
;
VLOG
(
5
)
<<
"send_finish, to_client_id: "
<<
to_client_id
;
static
const
std
::
string
empty_str
;
static
const
std
::
string
empty_str
;
return
_trainer_context
->
pslib
->
ps_client
()
->
send_client2client_msg
(
_finish_msg_type
,
to_client_id
,
empty_str
);
return
_trainer_context
->
pslib
->
ps_client
()
->
send_client2client_msg
(
_finish_msg_type
,
to_client_id
,
empty_str
);
...
@@ -230,6 +238,8 @@ private:
...
@@ -230,6 +238,8 @@ private:
bthread
::
Mutex
_wait_num_mutex
;
bthread
::
Mutex
_wait_num_mutex
;
std
::
atomic
<
int
>
_wait_num
;
std
::
atomic
<
int
>
_wait_num
;
std
::
atomic
<
uint32_t
>
_item_send_count
;
std
::
atomic
<
uint32_t
>
_item_receive_count
;
};
};
REGIST_CLASS
(
Shuffler
,
GlobalShuffler
);
REGIST_CLASS
(
Shuffler
,
GlobalShuffler
);
...
...
paddle/fluid/train/custom_trainer/feed/trainer_context.h
浏览文件 @
b23282f3
...
@@ -95,20 +95,23 @@ private:
...
@@ -95,20 +95,23 @@ private:
class
TrainerContext
{
class
TrainerContext
{
public:
public:
TrainerContext
()
{
trainer_status
.
resize
(
2
,
0
);
}
inline
paddle
::
ps
::
PSClient
*
ps_client
()
{
inline
paddle
::
ps
::
PSClient
*
ps_client
()
{
return
pslib
->
ps_client
();
return
pslib
->
ps_client
();
}
}
inline
bool
is_status
(
TrainerStatus
status
)
{
inline
bool
is_status
(
TrainerStatus
status
)
{
auto
bit
_idx
=
static_cast
<
uint32_t
>
(
status
);
auto
status
_idx
=
static_cast
<
uint32_t
>
(
status
);
return
((
trainer_status
>>
bit_idx
)
&
1
)
>
0
;
return
trainer_status
[
status_idx
]
>
0
;
}
}
// 非线程安全, 其实TrainerContext所有成员的线程安全性 取决于 成员本身的线程安全性
// 非线程安全, 其实TrainerContext所有成员的线程安全性 取决于 成员本身的线程安全性
inline
void
set_status
(
TrainerStatus
status
,
bool
on
)
{
inline
void
set_status
(
TrainerStatus
status
,
bool
on
)
{
auto
bit
_idx
=
static_cast
<
uint32_t
>
(
status
);
auto
status
_idx
=
static_cast
<
uint32_t
>
(
status
);
trainer_status
=
trainer_status
&
(
1L
<<
bit_idx
)
;
trainer_status
[
status_idx
]
=
on
?
1
:
0
;
}
}
uint32_t
trainer_status
;
// trainer当前,由于可同时处于多种状态,这里分bit存储状态
std
::
vector
<
uint32_t
>
trainer_status
;
YAML
::
Node
trainer_config
;
YAML
::
Node
trainer_config
;
paddle
::
platform
::
CPUPlace
cpu_place
;
paddle
::
platform
::
CPUPlace
cpu_place
;
...
@@ -122,6 +125,20 @@ public:
...
@@ -122,6 +125,20 @@ public:
std
::
shared_ptr
<
SignCacheDict
>
cache_dict
;
//大模型cache词典
std
::
shared_ptr
<
SignCacheDict
>
cache_dict
;
//大模型cache词典
};
};
class
ContextStatusGurad
{
public:
ContextStatusGurad
(
TrainerContext
*
context
,
TrainerStatus
status
)
:
_context
(
context
),
_status
(
status
)
{
_context
->
set_status
(
_status
,
true
);
}
virtual
~
ContextStatusGurad
()
{
_context
->
set_status
(
_status
,
false
);
}
private:
TrainerStatus
_status
;
TrainerContext
*
_context
=
nullptr
;
};
}
// namespace feed
}
// namespace feed
}
// namespace custom_trainer
}
// namespace custom_trainer
}
// namespace paddle
}
// namespace paddle
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