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
9bbba6c0
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
9bbba6c0
编写于
9月 16, 2020
作者:
Y
yinhaofeng
浏览文件
操作
浏览文件
下载
差异文件
hhhMerge branch 'master' of
https://github.com/PaddlePaddle/PaddleRec
into change-dataset
上级
838928dd
777be5a0
变更
2
隐藏空白更改
内联
并排
Showing
2 changed file
with
8 addition
and
12 deletion
+8
-12
core/trainers/framework/runner.py
core/trainers/framework/runner.py
+4
-6
models/rank/dnn/config.yaml
models/rank/dnn/config.yaml
+4
-6
未找到文件。
core/trainers/framework/runner.py
浏览文件 @
9bbba6c0
...
@@ -209,12 +209,10 @@ class RunnerBase(object):
...
@@ -209,12 +209,10 @@ class RunnerBase(object):
if
save_step_interval
>=
1
and
batch_id
%
save_step_interval
==
0
and
context
[
if
save_step_interval
>=
1
and
batch_id
%
save_step_interval
==
0
and
context
[
"is_infer"
]
==
False
:
"is_infer"
]
==
False
:
if
context
[
"fleet_mode"
]:
if
context
[
"fleet_mode"
].
upper
()
==
"PS"
:
if
context
[
"fleet_mode"
].
upper
()
==
"PS"
:
train_prog
=
context
[
"model"
][
model_dict
[
"name"
]][
train_prog
=
context
[
"model"
][
model_dict
[
"main_program"
]
"name"
]][
"main_program"
]
else
:
elif
not
context
[
"is_fleet"
]
or
context
[
"fleet_mode"
].
upper
()
==
"COLLECTIVE"
:
train_prog
=
context
[
"model"
][
model_dict
[
"name"
]][
train_prog
=
context
[
"model"
][
model_dict
[
"name"
]][
"default_main_program"
]
"default_main_program"
]
startup_prog
=
context
[
"model"
][
model_dict
[
"name"
]][
startup_prog
=
context
[
"model"
][
model_dict
[
"name"
]][
...
...
models/rank/dnn/config.yaml
浏览文件 @
9bbba6c0
...
@@ -114,15 +114,13 @@ runner:
...
@@ -114,15 +114,13 @@ runner:
print_interval
:
1
print_interval
:
1
phases
:
[
phase1
]
phases
:
[
phase1
]
-
name
:
local_ps
_train
-
name
:
single_multi_gpu
_train
class
:
local_cluster_
train
class
:
train
# num of epochs
# num of epochs
epochs
:
1
epochs
:
1
# device to run training or infer
# device to run training or infer
device
:
cpu
device
:
gpu
selected_gpus
:
"
0"
# 选择多卡执行训练
selected_gpus
:
"
0,1"
# 选择多卡执行训练
work_num
:
1
server_num
:
1
save_checkpoint_interval
:
1
# save model interval of epochs
save_checkpoint_interval
:
1
# save model interval of epochs
save_inference_interval
:
4
# save inference
save_inference_interval
:
4
# save inference
save_step_interval
:
1
save_step_interval
:
1
...
...
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