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
wmsofts
Paddle
提交
3678cae2
P
Paddle
项目概览
wmsofts
/
Paddle
与 Fork 源项目一致
Fork自
PaddlePaddle / Paddle
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
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
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 >>
未验证
提交
3678cae2
编写于
2月 27, 2023
作者:
张
张春乔
提交者:
GitHub
2月 27, 2023
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
support fp16 on AlphaDropout (#50917)
上级
5f60b597
变更
2
隐藏空白更改
内联
并排
Showing
2 changed file
with
26 addition
and
2 deletion
+26
-2
python/paddle/fluid/tests/unittests/test_dropout_op.py
python/paddle/fluid/tests/unittests/test_dropout_op.py
+24
-0
python/paddle/nn/functional/common.py
python/paddle/nn/functional/common.py
+2
-2
未找到文件。
python/paddle/fluid/tests/unittests/test_dropout_op.py
浏览文件 @
3678cae2
...
...
@@ -980,6 +980,30 @@ class TestAlphaDropoutCAPI(unittest.TestCase):
result
.
numpy
(),
result_np
,
rtol
=
1e-05
)
def
test_static_fp16_gpu
(
self
):
if
paddle
.
fluid
.
core
.
is_compiled_with_cuda
():
place
=
paddle
.
CUDAPlace
(
0
)
with
paddle
.
static
.
program_guard
(
paddle
.
static
.
Program
(),
paddle
.
static
.
Program
()
):
input
=
np
.
random
.
random
([
2
,
3
]).
astype
(
"float16"
)
x
=
paddle
.
static
.
data
(
name
=
"x"
,
shape
=
[
2
,
3
],
dtype
=
"float16"
)
m
=
paddle
.
nn
.
AlphaDropout
(
p
=
0.0
)
y
=
m
(
x
)
exe
=
paddle
.
static
.
Executor
(
place
)
res
=
exe
.
run
(
paddle
.
static
.
default_main_program
(),
feed
=
{
"x"
:
input
,
},
fetch_list
=
[
y
],
)
np
.
testing
.
assert_allclose
(
res
[
0
],
input
,
rtol
=
1e-05
)
class
TestDropoutWithDeterminateSeedGenerator
(
unittest
.
TestCase
):
def
setUp
(
self
):
...
...
python/paddle/nn/functional/common.py
浏览文件 @
3678cae2
...
...
@@ -1384,7 +1384,7 @@ def alpha_dropout(x, p=0.5, training=True, name=None):
Alpha Dropout fits well to SELU activate function by randomly setting activations to the negative saturation value.
Args:
x (Tensor): The input tensor. The data type is float32 or float64.
x (Tensor): The input tensor. The data type is float
16, float
32 or float64.
p (float | int): Probability of setting units to zero. Default 0.5.
training (bool): A flag indicating whether it is in train phrase or not. Default True.
name (str, optional): Name for the operation (optional, default is None). For more information, please refer to :ref:`api_guide_Name`.
...
...
@@ -1416,7 +1416,7 @@ def alpha_dropout(x, p=0.5, training=True, name=None):
if
not
in_dynamic_mode
():
check_variable_and_dtype
(
x
,
'x'
,
[
'float32'
,
'float64'
],
'alpha_dropout'
x
,
'x'
,
[
'float
16'
,
'float
32'
,
'float64'
],
'alpha_dropout'
)
if
training
:
...
...
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