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
2301_77200941
mindspore
提交
57b27c9f
M
mindspore
项目概览
2301_77200941
/
mindspore
与 Fork 源项目一致
Fork自
MindSpore / mindspore
通知
1
Star
1
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
M
mindspore
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 >>
提交
57b27c9f
编写于
8月 26, 2020
作者:
L
lizhenyu
浏览文件
操作
浏览文件
下载
电子邮件补丁
差异文件
code refine for BN docs
上级
37bae5bf
变更
2
隐藏空白更改
内联
并排
Showing
2 changed file
with
32 addition
and
15 deletion
+32
-15
mindspore/ccsrc/runtime/device/gpu/cuda_env_checker.cc
mindspore/ccsrc/runtime/device/gpu/cuda_env_checker.cc
+0
-1
mindspore/ops/operations/nn_ops.py
mindspore/ops/operations/nn_ops.py
+32
-14
未找到文件。
mindspore/ccsrc/runtime/device/gpu/cuda_env_checker.cc
浏览文件 @
57b27c9f
...
...
@@ -118,7 +118,6 @@ std::pair<std::string, bool> CudaEnvChecker::IsCudaRealPath(const std::string &p
valid_path
=
(
end
==
real_path
.
size
()
-
1
)
?
true
:
((
end
==
real_path
.
size
()
-
2
)
&&
(
real_path
.
back
()
==
'/'
));
return
{
real_path
.
substr
(
0
,
end
+
1
),
valid_path
};
}
}
// namespace gpu
}
// namespace device
}
// namespace mindspore
mindspore/ops/operations/nn_ops.py
浏览文件 @
57b27c9f
...
...
@@ -625,7 +625,21 @@ class FusedBatchNorm(Primitive):
class
FusedBatchNormEx
(
PrimitiveWithInfer
):
r
"""
FusedBatchNormEx is an extension of FusedBatchNorm
FusedBatchNormEx is an extension of FusedBatchNorm, FusedBatchNormEx has one more output(output reserve)
than FusedBatchNorm, reserve will be used in backpropagation phase. FusedBatchNorm is a BatchNorm that
moving mean and moving variance will be computed instead of being loaded.
Batch Normalization is widely used in convolutional networks. This operation applies
Batch Normalization over input to avoid internal covariate shift as described in the
paper `Batch Normalization: Accelerating Deep Network Training by Reducing Internal
Covariate Shift <https://arxiv.org/abs/1502.03167>`_. It rescales and recenters the
feature using a mini-batch of data and the learned parameters which can be described
in the following formula.
.. math::
y = \frac{x - mean}{\sqrt{variance + \epsilon}} * \gamma + \beta
where :math:`\gamma` is scale, :math:`\beta` is bias, :math:`\epsilon` is epsilon.
Args:
mode (int): Mode of batch normalization, value is 0 or 1. Default: 0.
...
...
@@ -635,21 +649,25 @@ class FusedBatchNormEx(PrimitiveWithInfer):
Momentum value should be [0, 1]. Default: 0.9.
Inputs:
- **input_x** (Tensor) - Tensor of shape :math:`(N, C)`.
- **scale** (Tensor) - Tensor of shape :math:`(C,)`.
- **bias** (Tensor) - Tensor of shape :math:`(C,)`.
- **mean** (Tensor) - Tensor of shape :math:`(C,)`.
- **variance** (Tensor) - Tensor of shape :math:`(C,)`.
- **input_x** (Tensor) - The input of FusedBatchNormEx, Tensor of shape :math:`(N, C)`,
data type: float16 or float32.
- **scale** (Tensor) - Parameter scale, same with gamma above-mentioned, Tensor of shape :math:`(C,)`,
data type: float32.
- **bias** (Tensor) - Parameter bias, same with beta above-mentioned, Tensor of shape :math:`(C,)`,
data type: float32.
- **mean** (Tensor) - mean value, Tensor of shape :math:`(C,)`, data type: float32.
- **variance** (Tensor) - variance value, Tensor of shape :math:`(C,)`, data type: float32.
Outputs:
Tuple of 6 Tensor, the normalized input and the updated parameters.
- **output_x** (Tensor) - The same type and shape as the `input_x`.
- **updated_scale** (Tensor) - Tensor of shape :math:`(C,)`.
- **updated_bias** (Tensor) - Tensor of shape :math:`(C,)`.
- **updated_moving_mean** (Tensor) - Tensor of shape :math:`(C,)`.
- **updated_moving_variance** (Tensor) - Tensor of shape :math:`(C,)`.
- **reserve** (Tensor) - Tensor of shape :math:`(C,)`.
Tuple of 6 Tensor, the normalized input, the updated parameters and reserve.
- **output_x** (Tensor) - The input of FusedBatchNormEx, same type and shape as the `input_x`.
- **updated_scale** (Tensor) - Updated parameter scale, Tensor of shape :math:`(C,)`, data type: float32.
- **updated_bias** (Tensor) - Updated parameter bias, Tensor of shape :math:`(C,)`, data type: float32.
- **updated_moving_mean** (Tensor) - Updated mean value, Tensor of shape :math:`(C,)`, data type: float32.
- **updated_moving_variance** (Tensor) - Updated variance value, Tensor of shape :math:`(C,)`,
data type: float32.
- **reserve** (Tensor) - reserve space, Tensor of shape :math:`(C,)`, data type: float32.
Examples:
>>> input_x = Tensor(np.ones([128, 64, 32, 64]), mindspore.float32)
...
...
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