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
wzpei9
Paddle
提交
c72a4f4d
P
Paddle
项目概览
wzpei9
/
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
体验新版 GitCode,发现更多精彩内容 >>
提交
c72a4f4d
编写于
5月 14, 2018
作者:
Y
yuyang18
浏览文件
操作
浏览文件
下载
差异文件
Merge branch 'develop' of
https://github.com/PaddlePaddle/Paddle
into feature/exec_strategy
上级
08295f98
8231960f
变更
3
展开全部
隐藏空白更改
内联
并排
Showing
3 changed file
with
4264 addition
and
7 deletion
+4264
-7
cmake/external/mkldnn.cmake
cmake/external/mkldnn.cmake
+2
-0
patches/mkldnn.hpp
patches/mkldnn.hpp
+4252
-0
python/paddle/fluid/tests/unittests/test_network_with_dtype.py
...n/paddle/fluid/tests/unittests/test_network_with_dtype.py
+10
-7
未找到文件。
cmake/external/mkldnn.cmake
浏览文件 @
c72a4f4d
...
...
@@ -56,6 +56,8 @@ ExternalProject_Add(
GIT_TAG
"v0.14"
PREFIX
${
MKLDNN_SOURCES_DIR
}
UPDATE_COMMAND
""
# Patch MKLDNN to compile with gcc 4.8, the related issue is in intel/mkl-dnn#237.
PATCH_COMMAND
${
CMAKE_COMMAND
}
-E copy_if_different
${
CMAKE_CURRENT_SOURCE_DIR
}
/patches/mkldnn.hpp
${
MKLDNN_SOURCES_DIR
}
/src/extern_mkldnn/include/mkldnn.hpp
CMAKE_ARGS -DCMAKE_INSTALL_PREFIX=
${
MKLDNN_INSTALL_DIR
}
CMAKE_ARGS -DCMAKE_BUILD_TYPE=
${
CMAKE_BUILD_TYPE
}
CMAKE_ARGS -DMKLROOT=
${
MKLML_ROOT
}
...
...
patches/mkldnn.hpp
0 → 100644
浏览文件 @
c72a4f4d
此差异已折叠。
点击以展开。
python/paddle/fluid/tests/unittests/test_network_with_dtype.py
浏览文件 @
c72a4f4d
...
...
@@ -27,12 +27,15 @@ class TestNetWithDtype(unittest.TestCase):
def
set_network
(
self
):
self
.
dtype
=
"float64"
self
.
init_dtype
()
self
.
x
=
fluid
.
layers
.
data
(
name
=
'x'
,
shape
=
[
13
],
dtype
=
self
.
dtype
)
self
.
y
=
fluid
.
layers
.
data
(
name
=
'y'
,
shape
=
[
1
],
dtype
=
self
.
dtype
)
y_predict
=
fluid
.
layers
.
fc
(
input
=
self
.
x
,
size
=
1
,
act
=
None
)
main
=
fluid
.
Program
()
with
fluid
.
program_guard
(
main
):
self
.
x
=
fluid
.
layers
.
data
(
name
=
'x'
,
shape
=
[
13
],
dtype
=
self
.
dtype
)
self
.
y
=
fluid
.
layers
.
data
(
name
=
'y'
,
shape
=
[
1
],
dtype
=
self
.
dtype
)
y_predict
=
fluid
.
layers
.
fc
(
input
=
self
.
x
,
size
=
1
,
act
=
None
)
cost
=
fluid
.
layers
.
square_error_cost
(
input
=
y_predict
,
label
=
self
.
y
)
avg_cost
=
fluid
.
layers
.
mean
(
cost
)
cost
=
fluid
.
layers
.
square_error_cost
(
input
=
y_predict
,
label
=
self
.
y
)
avg_cost
=
fluid
.
layers
.
mean
(
cost
)
self
.
program
=
main
self
.
fetch_list
=
[
avg_cost
]
sgd_optimizer
=
fluid
.
optimizer
.
SGD
(
learning_rate
=
0.001
)
...
...
@@ -45,7 +48,7 @@ class TestNetWithDtype(unittest.TestCase):
exe
=
fluid
.
Executor
(
place
)
exe
.
run
(
fluid
.
default_startup_program
())
for
data
in
train_reader
():
exe
.
run
(
fluid
.
default_main_program
()
,
exe
.
run
(
self
.
program
,
feed
=
feeder
.
feed
(
data
),
fetch_list
=
self
.
fetch_list
)
# the main program is runable, the datatype is fully supported
...
...
@@ -68,7 +71,7 @@ class TestNetWithDtype(unittest.TestCase):
# TODO(dzhwinter): make sure the fp16 is runable
# class TestFloat16(
SimpleNet
):
# class TestFloat16(
TestNetWithDtype
):
# def init_dtype(self):
# self.dtype = "float16"
...
...
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