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
2b83b6ac
提交
2b83b6ac
authored
8月 04, 2015
作者:
Frédéric Bastien
浏览文件
操作
浏览文件
下载
差异文件
Merge pull request #3238 from f0k/fix-corrmm-docstring
Fix outdated description of border_mode in GpuCorrMM constructor
上级
07e9332a
fbf3dcdd
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
11 行增加
和
10 行删除
+11
-10
blas.py
theano/sandbox/cuda/blas.py
+11
-10
没有找到文件。
theano/sandbox/cuda/blas.py
浏览文件 @
2b83b6ac
...
@@ -726,7 +726,7 @@ class BaseGpuCorrMM(GpuOp):
...
@@ -726,7 +726,7 @@ class BaseGpuCorrMM(GpuOp):
'do not use pad for BaseGpuCorrMM; please set padding in '
'do not use pad for BaseGpuCorrMM; please set padding in '
'border_mode parameter, see the docstring for more details'
)
'border_mode parameter, see the docstring for more details'
)
if
border_mode
!=
"valid"
:
if
border_mode
!=
"valid"
:
raise
ValueError
(
"border_mode must be 'valid'"
)
raise
ValueError
(
"border_mode must be 'valid'
if pad is given
"
)
border_mode
=
pad
border_mode
=
pad
if
isinstance
(
border_mode
,
int
):
if
isinstance
(
border_mode
,
int
):
border_mode
=
(
border_mode
,
border_mode
)
border_mode
=
(
border_mode
,
border_mode
)
...
@@ -995,20 +995,21 @@ class BaseGpuCorrMM(GpuOp):
...
@@ -995,20 +995,21 @@ class BaseGpuCorrMM(GpuOp):
class
GpuCorrMM
(
BaseGpuCorrMM
):
class
GpuCorrMM
(
BaseGpuCorrMM
):
"""GPU correlation implementation using Matrix Multiplication.
"""GPU correlation implementation using Matrix Multiplication.
:param border_mode: currently supports "valid" only; "full" can be
:param border_mode: the width of a border of implicit zeros to pad the
simulated by setting `pad="full"` (at the cost of performance), or
input with. Must be a tuple with 2 elements giving the numbers of rows
by using `GpuCorrMM_gradInputs`
and columns to pad on each side, or a single integer to pad the same
on all sides, or a string shortcut setting the padding at runtime:
``'valid'`` for ``(0, 0)`` (valid convolution, no padding), ``'full'``
for ``(kernel_rows - 1, kernel_columns - 1)`` (full convolution),
``'half'`` for ``(kernel_rows // 2, kernel_columns // 2)`` (same
convolution for odd-sized kernels). Note that the two widths are each
applied twice, once per side (left and right, top and bottom).
:param subsample: the subsample operation applied to each output image.
:param subsample: the subsample operation applied to each output image.
Should be a tuple with 2 elements.
Should be a tuple with 2 elements.
`(sv, sh)` is equivalent to `GpuCorrMM(...)(...)[:,:,::sv, ::sh]`,
`(sv, sh)` is equivalent to `GpuCorrMM(...)(...)[:,:,::sv, ::sh]`,
but faster.
but faster.
Set to `(1, 1)` to disable subsampling.
Set to `(1, 1)` to disable subsampling.
:param pad: the width of a border of implicit zeros to pad the input
:param pad: deprecated alias for `border_mode`.
image with. Should be a tuple with 2 elements giving the numbers of
rows and columns to pad on each side, or "half" to set the padding
to `(kernel_rows // 2, kernel_columns // 2)`, or "full" to set the
padding to `(kernel_rows - 1, kernel_columns - 1)` at runtime.
Set to `(0, 0)` to disable padding.
:note: Currently, the Op requires the inputs, filters and outputs to be
:note: Currently, the Op requires the inputs, filters and outputs to be
C-contiguous. Use :func:`gpu_contiguous
C-contiguous. Use :func:`gpu_contiguous
...
...
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