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
PaddleRec
提交
21630068
P
PaddleRec
项目概览
BaiXuePrincess
/
PaddleRec
与 Fork 源项目一致
Fork自
PaddlePaddle / PaddleRec
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
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
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
21630068
编写于
4月 29, 2020
作者:
T
tangwei
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
mac/windows fix
上级
6ef9bab0
变更
1
隐藏空白更改
内联
并排
Showing
1 changed file
with
40 addition
and
9 deletion
+40
-9
fleetrec/core/trainers/cluster_trainer.py
fleetrec/core/trainers/cluster_trainer.py
+40
-9
未找到文件。
fleetrec/core/trainers/cluster_trainer.py
浏览文件 @
21630068
...
...
@@ -64,6 +64,7 @@ class ClusterTrainer(TranspileTrainer):
assert
strategy
is
not
None
self
.
strategy
=
strategy
return
strategy
def
init
(
self
,
context
):
...
...
@@ -96,20 +97,50 @@ class ClusterTrainer(TranspileTrainer):
def
dataset_train
(
self
,
context
):
self
.
_exe
.
run
(
fleet
.
startup_program
)
fleet
.
init_worker
()
dataset
=
self
.
_get_dataset
()
reader
=
self
.
_get_dataloader
()
epochs
=
envs
.
get_global_env
(
"train.epochs"
)
for
i
in
range
(
epochs
):
self
.
_exe
.
train_from_dataset
(
program
=
fluid
.
default_main_program
(),
dataset
=
dataset
,
fetch_list
=
self
.
fetch_vars
,
fetch_info
=
self
.
fetch_alias
,
print_period
=
self
.
fetch_period
)
self
.
save
(
i
,
"train"
,
is_fleet
=
True
)
context
[
'status'
]
=
'terminal_pass'
program
=
fluid
.
compiler
.
CompiledProgram
(
fleet
.
main_program
).
with_data_parallel
(
loss_name
=
self
.
model
.
get_cost_op
().
name
,
build_strategy
=
self
.
strategy
.
get_build_strategy
(),
exec_strategy
=
self
.
strategy
.
get_execute_strategy
())
metrics_varnames
=
[]
metrics_format
=
[]
metrics_format
.
append
(
"{}: {{}}"
.
format
(
"epoch"
))
metrics_format
.
append
(
"{}: {{}}"
.
format
(
"batch"
))
for
name
,
var
in
self
.
model
.
get_metrics
().
items
():
metrics_varnames
.
append
(
var
.
name
)
metrics_format
.
append
(
"{}: {{}}"
.
format
(
name
))
metrics_format
=
", "
.
join
(
metrics_format
)
for
epoch
in
range
(
epochs
):
reader
.
start
()
batch_id
=
0
try
:
while
True
:
metrics_rets
=
self
.
_exe
.
run
(
program
=
program
,
fetch_list
=
metrics_varnames
)
metrics
=
[
epoch
,
batch_id
]
metrics
.
extend
(
metrics_rets
)
if
batch_id
%
10
==
0
and
batch_id
!=
0
:
print
(
metrics_format
.
format
(
*
metrics
))
batch_id
+=
1
except
fluid
.
core
.
EOFException
:
reader
.
reset
()
fleet
.
stop_worker
()
context
[
'status'
]
=
'terminal_pass'
def
infer
(
self
,
context
):
context
[
'status'
]
=
'terminal_pass'
...
...
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