Skip to content
项目
群组
代码片段
帮助
当前项目
正在载入...
登录 / 注册
切换导航面板
P
pytensor
项目
项目
详情
活动
周期分析
仓库
仓库
文件
提交
分支
标签
贡献者
图表
比较
统计图
议题
0
议题
0
列表
看板
标记
里程碑
合并请求
0
合并请求
0
CI / CD
CI / CD
流水线
作业
日程
统计图
Wiki
Wiki
代码片段
代码片段
成员
成员
折叠边栏
关闭边栏
活动
图像
聊天
创建新问题
作业
提交
问题看板
Open sidebar
testgroup
pytensor
Commits
872799cd
提交
872799cd
authored
2月 08, 2016
作者:
carriepl
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
Add section on navigating between Scan inputs and outputs
上级
d2611b94
隐藏空白字符变更
内嵌
并排
正在显示
1 个修改的文件
包含
44 行增加
和
0 行删除
+44
-0
scan.txt
doc/developer/scan.txt
+44
-0
没有找到文件。
doc/developer/scan.txt
浏览文件 @
872799cd
...
...
@@ -214,3 +214,47 @@ The Scan op also defines a few helper functions for this purpose, such as
`inner_nitsot_outs()` or `mitmot_out_taps()`, but they are often poorly
documented and easy to misuse. These should be used with great care.
Navigating between outer inputs/outputs and inner inputs/outputs
----------------------------------------------------------------
Navigation between these four sets of variables can be done in two ways,
depending on the type of navigation that is required.
If the goal is to navigate between variables that are associated with the same
states (ex : going from an outer sequence input to the corresponding inner
sequence input, going from an inner output associated with a recurrent state
to the inner input(s) associated with that same recurrent state, etc.), then
the `var_mappings` attribute of the Scan op can be used.
This attribute is a dictionnary with 12 {key/value} pairs. The keys are listed
below :
* "outer_inp_from_outer_out"
* "inner_inp_from_outer_out"
* "inner_out_from_outer_out"
* "inner_inp_from_outer_inp"
* "inner_out_from_outer_inp"
* "outer_out_from_outer_inp"
* "outer_inp_from_inner_inp"
* "inner_out_from_inner_inp"
* "outer_out_from_inner_inp"
* "outer_inp_from_inner_out"
* "inner_inp_from_inner_out"
* "outer_out_from_inner_out"
Every corresponding value is a dictionary detailing a mapping from one set of
variables to another. For each of those dictionaries the keys are indices of
variables in one set and the values are the indices of the corresponding
variables in another set. For mappings to outer variables, the values are
individual indices or -1 if there is not corresponding outer variable.
For mappings to inner variables, the values are list of indices because
multiple inner variables may be associated with the same state.
If the goal is to navigate between variables that are *connected* (meaning
that one of them is used to compute the other), the methods
`connection_pattern()` and `inner_connection_pattern()` can be used.
The method `connection_pattern()` returns a list of list detailing, for every
pair of outer input and outer output whether they are connected or not. The
method `inner_connection_pattern()` accomplishes the same goal but for every
possible pair of inner output and inner input.
编写
预览
Markdown
格式
0%
重试
或
添加新文件
添加附件
取消
您添加了
0
人
到此讨论。请谨慎行事。
请先完成此评论的编辑!
取消
请
注册
或者
登录
后发表评论