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
ed2129cc
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
ed2129cc
编写于
5月 21, 2018
作者:
T
tangwei12
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
revert distribute_transpiler.py
上级
01975ec1
变更
1
隐藏空白更改
内联
并排
Showing
1 changed file
with
1 addition
and
125 deletion
+1
-125
python/paddle/fluid/transpiler/distribute_transpiler.py
python/paddle/fluid/transpiler/distribute_transpiler.py
+1
-125
未找到文件。
python/paddle/fluid/transpiler/distribute_transpiler.py
浏览文件 @
ed2129cc
...
...
@@ -14,7 +14,6 @@
from
__future__
import
print_function
import
os
import
math
import
distributed_splitter
as
splitter
...
...
@@ -27,10 +26,6 @@ LOOKUP_TABLE_TYPE = "lookup_table"
LOOKUP_TABLE_GRAD_TYPE
=
"lookup_table_grad"
RPC_CLIENT_VAR_NAME
=
"RPC_CLIENT_VAR"
# for checkpoint
SUCCESS
=
"_SUCCESS"
SERIAL_VAR_NAME
=
"SERIAL_NUMBER"
class
VarBlock
:
def
__init__
(
self
,
varname
,
offset
,
size
):
...
...
@@ -161,8 +156,7 @@ class DistributeTranspiler:
pservers
=
"127.0.0.1:6174"
,
trainers
=
1
,
split_method
=
splitter
.
round_robin
,
sync_mode
=
True
,
checkpoint_dir
=
None
):
sync_mode
=
True
):
"""
Transpile the program to distributed data-parallelism programs.
The main_program will be transformed to use a remote parameter server
...
...
@@ -216,12 +210,6 @@ class DistributeTranspiler:
self
.
pserver_endpoints
=
pserver_endpoints
self
.
optimize_ops
,
params_grads
=
self
.
_get_optimize_pass
()
# is_chief (no.0 triner) for checkpoint
# the no.0 trainer will save all variables and its own reader offset to checkpoint
# other trianers will save its own reader offset to checkpoint
self
.
_is_chief
=
trainer_id
==
0
self
.
checkpoint_dir
=
checkpoint_dir
# process lookup_table_op
# 1. check all lookup_table_op is distributed
# 2. check all lookup_table_op share the same table.
...
...
@@ -327,24 +315,6 @@ class DistributeTranspiler:
"epmap"
:
eplist
,
"sync_mode"
:
self
.
sync_mode
})
if
self
.
checkpoint_dir
and
self
.
_is_chief
:
program
.
global_block
().
create_var
(
name
=
SERIAL_VAR_NAME
,
persistable
=
True
,
type
=
core
.
VarDesc
.
VarType
.
RAW
)
save_vars
=
[]
for
var
in
self
.
origin_program
.
list_vars
():
if
self
.
_is_persistable
(
var
):
save_vars
.
append
(
var
.
name
)
program
.
global_block
().
append_op
(
type
=
"checkpoint_save"
,
inputs
=
{
"X"
:
save_vars
},
attrs
=
{
"overwrite"
:
True
,
"dir"
:
self
.
checkpoint_dir
})
# step4: Concat the parameters splits together after recv.
for
varname
,
splited_var
in
param_var_mapping
.
iteritems
():
if
len
(
splited_var
)
<=
1
:
...
...
@@ -525,37 +495,6 @@ class DistributeTranspiler:
pserver_program
.
sync_with_cpp
()
return
pserver_program
def
get_train_startup_program
(
self
):
"""
Get train startup program.
If self.checkpoint_dir is None, rerurn default startup program.
IF self.checkpoint_dir is Exist, add checkpoint_load op and load Var.
"""
startup_prog
=
default_startup_program
()
if
not
self
.
checkpoint_dir
:
return
startup_prog
load_vars
=
[]
for
var
in
startup_prog
.
list_vars
():
if
self
.
_is_persistable
(
var
):
load_vars
.
append
(
var
.
name
)
serial_number
=
self
.
_get_lastest_checkpoint_dir
(
self
.
checkpoint_dir
)
startup_prog
.
global_block
().
create_var
(
name
=
SERIAL_VAR_NAME
,
persistable
=
True
,
type
=
core
.
VarDesc
.
VarType
.
RAW
)
startup_prog
.
global_block
().
append_op
(
type
=
"checkpoint_load"
,
inputs
=
{
"X"
:
load_vars
},
outputs
=
{
"Argv"
:
[]},
attrs
=
{
"dir"
:
self
.
checkpoint_dir
,
"Serial"
:
serial_number
})
return
startup_prog
def
get_startup_program
(
self
,
endpoint
,
pserver_program
):
"""
Get startup program for current parameter server.
...
...
@@ -581,7 +520,6 @@ class DistributeTranspiler:
created_var_map
[
var
.
name
]
=
tmpvar
# 2. rename op outputs
load_vars
=
[]
for
op
in
orig_s_prog
.
global_block
().
ops
:
new_inputs
=
dict
()
new_outputs
=
dict
()
...
...
@@ -609,70 +547,8 @@ class DistributeTranspiler:
inputs
=
new_inputs
,
outputs
=
new_outputs
,
attrs
=
op
.
attrs
)
for
var
in
new_outputs
.
values
():
load_vars
.
append
(
var
.
name
)
# add checkpoint op
if
not
self
.
checkpoint_dir
:
return
s_prog
serial_number
=
self
.
_get_lastest_checkpoint_dir
(
self
.
checkpoint_dir
)
s_prog
.
global_block
().
create_var
(
name
=
SERIAL_VAR_NAME
,
persistable
=
True
,
type
=
core
.
VarDesc
.
VarType
.
RAW
)
s_prog
.
global_block
().
append_op
(
type
=
"checkpoint_load"
,
inputs
=
{
"X"
:
load_vars
},
outputs
=
{
"Argv"
:
[]},
attrs
=
{
"dir"
:
self
.
checkpoint_dir
,
"Serial"
:
serial_number
})
return
s_prog
def
_is_persistable
(
self
,
var
):
"""only save LodTensor variable"""
if
var
.
desc
.
type
()
==
core
.
VarDesc
.
VarType
.
FEED_MINIBATCH
or
\
var
.
desc
.
type
()
==
core
.
VarDesc
.
VarType
.
FETCH_LIST
or
\
var
.
desc
.
type
()
==
core
.
VarDesc
.
VarType
.
RAW
:
return
False
return
var
.
persistable
def
_get_lastest_checkpoint_dir
(
self
,
checkpoint_dir
):
"""
get the biggest number in checkpoint_dir, which has _SUCCESS
"""
if
not
checkpoint_dir
.
strip
():
return
""
def
has_success
(
checkpoint_dir
,
cur_dir
):
"""
is _SUCCESS in this dir
"""
if
not
os
.
path
.
isdir
(
os
.
path
.
join
(
checkpoint_dir
,
cur_dir
)):
return
-
1
try
:
int
(
cur_dir
)
except
ValueError
:
return
-
1
success_path
=
os
.
path
.
join
(
checkpoint_dir
,
cur_dir
,
SUCCESS
)
if
os
.
path
.
isfile
(
success_path
):
return
int
(
cur_dir
)
if
not
os
.
path
.
isdir
(
checkpoint_dir
):
return
"-1"
current_dir
=
0
dirs
=
os
.
listdir
(
checkpoint_dir
)
for
cur_dir
in
dirs
:
success_num
=
has_success
(
checkpoint_dir
,
cur_dir
)
if
success_num
>
current_dir
:
current_dir
=
success_num
return
str
(
current_dir
)
# transpiler function for dis lookup_table
def
_replace_lookup_table_op_with_prefetch
(
self
,
program
,
rpc_client_var
,
eplist
):
...
...
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