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
Paddle
提交
ee21f2f6
P
Paddle
项目概览
PaddlePaddle
/
Paddle
12 个月 前同步成功
通知
2292
Star
20931
Fork
5422
代码
文件
提交
分支
Tags
贡献者
分支图
Diff
Issue
1423
列表
看板
标记
里程碑
合并请求
543
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
1,423
Issue
1,423
列表
看板
标记
里程碑
合并请求
543
合并请求
543
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
ee21f2f6
编写于
1月 23, 2018
作者:
W
wanghaoshuang
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
Change default value of drop_rate in img_conv_group to 0
上级
37a94370
变更
1
隐藏空白更改
内联
并排
Showing
1 changed file
with
15 addition
and
15 deletion
+15
-15
python/paddle/v2/fluid/nets.py
python/paddle/v2/fluid/nets.py
+15
-15
未找到文件。
python/paddle/v2/fluid/nets.py
浏览文件 @
ee21f2f6
...
...
@@ -52,7 +52,7 @@ def img_conv_group(input,
conv_act
=
None
,
param_attr
=
None
,
conv_with_batchnorm
=
False
,
conv_batchnorm_drop_rate
=
None
,
conv_batchnorm_drop_rate
=
0
,
pool_stride
=
1
,
pool_type
=
None
):
"""
...
...
@@ -120,21 +120,21 @@ def sequence_conv_pool(input,
def
glu
(
input
,
dim
=-
1
):
"""
The gated linear unit composed by split, sigmoid activation and elementwise
multiplication. Specifically, Split the input into two equal sized parts
:math:`a` and :math:`b` along the given dimension and then compute as
The gated linear unit composed by split, sigmoid activation and elementwise
multiplication. Specifically, Split the input into two equal sized parts
:math:`a` and :math:`b` along the given dimension and then compute as
following:
.. math::
{GLU}(a, b)= a \otimes \sigma(b)
Refer to `Language Modeling with Gated Convolutional Networks
Refer to `Language Modeling with Gated Convolutional Networks
<https://arxiv.org/pdf/1612.08083.pdf>`_.
Args:
input (Variable): The input variable which is a Tensor or LoDTensor.
dim (int): The dimension along which to split. If :math:`dim < 0`, the
dim (int): The dimension along which to split. If :math:`dim < 0`, the
dimension to split along is :math:`rank(input) + dim`.
Returns:
...
...
@@ -157,24 +157,24 @@ def dot_product_attention(querys, keys, values):
"""
The dot-product attention.
Attention mechanism can be seen as mapping a query and a set of key-value
pairs to an output. The output is computed as a weighted sum of the values,
where the weight assigned to each value is computed by a compatibility
Attention mechanism can be seen as mapping a query and a set of key-value
pairs to an output. The output is computed as a weighted sum of the values,
where the weight assigned to each value is computed by a compatibility
function (dot-product here) of the query with the corresponding key.
The dot-product attention can be implemented through (batch) matrix
The dot-product attention can be implemented through (batch) matrix
multipication as follows:
.. math::
Attention(Q, K, V)= softmax(QK^\mathrm{T})V
Refer to `Attention Is All You Need
Refer to `Attention Is All You Need
<https://arxiv.org/pdf/1706.03762.pdf>`_.
Note that batch data containing sequences with different lengths is not
Note that batch data containing sequences with different lengths is not
supported by this because of the (batch) matrix multipication.
Args:
query (Variable): The input variable which is a Tensor or LoDTensor.
key (Variable): The input variable which is a Tensor or LoDTensor.
...
...
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