Skip to content
Merged
Show file tree
Hide file tree
Changes from 6 commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 3 additions & 3 deletions docs/api/paddle/audio/features/LogMelSpectrogram_cn.rst
Original file line number Diff line number Diff line change
Expand Up @@ -14,19 +14,19 @@ LogMelSpectrogram
- **n_fft** (int,可选) - 离散傅里叶变换中频率窗大小,默认 512。
- **hop_length** (int,可选) - 帧移,默认 512。
- **win_length** (int,可选) - 短时 FFT 的窗长,默认为 None。
- **window** (str,可选) - 窗函数名,支持的窗函数类型'hamming','hann','gaussian','general_gaussian','exponential','triang','bohman','blackman','cosine','tukey','taylor','bartlett','kaiser','nuttall',默认``hann``。
- **window** (str,可选) - 窗函数名,支持的窗函数类型'hamming','hann','gaussian','general_gaussian','exponential','triang','bohman','blackman','cosine','tukey','taylor','bartlett','kaiser','nuttall',默认 ``hann``
- **power** (float,可选) - 幅度谱的指数,默认是 2.0。
- **center** (bool,可选) - 对输入信号填充,如果 True,那么 t 以 t*hop_length 为中心,如果为 False,则 t 以 t*hop_length 开始,默认是 True。
- **pad_mode** (str,可选) - 如果 center 是 True,选择填充的方式,默认值是'reflect'。
- **n_mels** (int,可选) - mel bins 的数目,默认是 64。
- **f_min** (float,可选) - 最小频率(hz),默认 50.0。
- **f_max** (float,可选) - 最大频率(hz),默认为 None。
- **htk** (bool,可选) - 在计算 fbank 矩阵时是否用在 HTK 公式缩放,默认是 False。
- **norm** (Union[str,float],可选) - 计算 fbank 矩阵时正则化的种类,默认是``slaney``,你也可以 norm=0.5,使用 p-norm 正则化。
- **norm** (Union[str,float],可选) - 计算 fbank 矩阵时正则化的种类,默认是 ``slaney`` ,你也可以 norm=0.5,使用 p-norm 正则化。
- **ref_value** (float,可选) - 参照值,如果小于 1.0,信号的 db 会被提升,相反 db 会下降,默认值为 1.0。
- **amin** (float,可选) - 输入的幅值的最小值,默认是 1e-10。
- **top_db** (float,可选) - log-mel 谱的最大值(db),默认是 None。
- **dtype** (str|paddle.dtype|np.dtype,可选) - 输入和窗的数据类型,默认是``float32``。
- **dtype** (str|paddle.dtype|np.dtype,可选) - 输入和窗的数据类型,默认是 ``float32``


返回
Expand Down
Original file line number Diff line number Diff line change
Expand Up @@ -16,13 +16,13 @@ compute_fbank_matrix
- **f_min** (float,可选) - 最小频率(hz),默认是 0.0。
- **f_max** (Optional[float],可选) - 最大频率(hz),默认是 None。
- **htk** (bool,可选) - 是否使用 htk 缩放,默认是 False。
- **norm** (Union[str, float],可选) - norm 的类型,默认是``slaney``。
- **dtype** (str|paddle.dtype|np.dtype,可选) - 返回矩阵的数据类型,默认``float32``。
- **norm** (Union[str, float],可选) - norm 的类型,默认是 ``slaney``。
- **dtype** (str|paddle.dtype|np.dtype,可选) - 返回矩阵的数据类型,默认 ``float32``

返回
:::::::::

``paddle.Tensor``,Tensor shape (n_mels, n_fft//2 + 1)。
``paddle.Tensor`` ,Tensor shape (n_mels, n_fft//2 + 1)。

代码示例
:::::::::
Expand Down
8 changes: 4 additions & 4 deletions docs/api/paddle/distributed/fleet/UtilBase_cn.rst
Original file line number Diff line number Diff line change
Expand Up @@ -16,11 +16,11 @@ all_reduce(input, mode="sum", comm_world="worker")

- **input** (list|tuple|numpy.array) – 归约操作的输入。
- **mode** (str) - 归约操作的模式,包含求和,取最大值和取最小值,默认为求和归约。
- **comm_world** (str) - 归约操作的通信集合,包含:server 集合(``server``),worker 集合(``worker``)及所有节点集合(``all``),默认为 worker 集合。
- **comm_world** (str) - 归约操作的通信集合,包含:server 集合( ``server`` ),worker 集合( ``worker`` )及所有节点集合( ``all`` ),默认为 worker 集合。

**返回**

Numpy.array|None:一个和``input``形状一致的 numpy 数组或 None。
Numpy.array|None:一个和 ``input`` 形状一致的 numpy 数组或 None。

**代码示例**

Expand All @@ -32,7 +32,7 @@ barrier(comm_world="worker")

**参数**

- **comm_world** (str) - 阻塞操作的通信集合,包含:server 集合(``server``),worker 集合(``worker``)及所有节点集合(``all``),默认为 worker 集合。
- **comm_world** (str) - 阻塞操作的通信集合,包含:server 集合( ``server`` ),worker 集合( ``worker`` )及所有节点集合( ``all`` ),默认为 worker 集合。

**代码示例**

Expand All @@ -45,7 +45,7 @@ all_gather(input, comm_world="worker")
**参数**

- **input** (int|float) - 聚合操作的输入。
- **comm_world** (str) - 聚合操作的通信集合,包含:server 集合(``server``),worker 集合(``worker``)及所有节点集合(``all``),默认为 worker 集合。
- **comm_world** (str) - 聚合操作的通信集合,包含:server 集合( ``server`` ),worker 集合( ``worker`` )及所有节点集合( ``all`` ),默认为 worker 集合。

**返回**

Expand Down
14 changes: 7 additions & 7 deletions docs/api/paddle/nn/functional/dropout_cn.rst
Original file line number Diff line number Diff line change
Expand Up @@ -50,7 +50,7 @@ axis 参数的默认值为 None。当 ``axis=None`` 时,dropout 的功能为
将输入 x 和生成的 mask 点积,就得到了随机丢弃部分元素之后的结果:
[[0 2 0]
[4 0 6]]
假定 dropout 的概率使用默认值,即 ``p=0.5``,若 mode 参数使用默认值,即 ``mode='upscale_in_train'`` ,
假定 dropout 的概率使用默认值,即 ``p=0.5`` ,若 mode 参数使用默认值,即 ``mode='upscale_in_train'`` ,
则在训练阶段,最终增大后的结果为:
[[0 4 0 ]
[8 0 12]]
Expand All @@ -70,8 +70,8 @@ axis 参数的默认值为 None。当 ``axis=None`` 时,dropout 的功能为

- axis 应设置为:``[0, 1, ... , ndim(x)-1]`` 的子集(ndim(x) 为输入 x 的维度),例如:

- 若 x 的维度为 2,参数 axis 可能的取值有 4 种:``None``, ``[0]``, ``[1]``, ``[0,1]``
- 若 x 的维度为 3,参数 axis 可能的取值有 8 种:``None``, ``[0]``, ``[1]``, ``[2]``, ``[0,1]``, ``[0,2]``, ``[1,2]``, ``[0,1,2]``
- 若 x 的维度为 2,参数 axis 可能的取值有 4 种: ``None`` , ``[0]`` , ``[1]`` , ``[0,1]``
- 若 x 的维度为 3,参数 axis 可能的取值有 8 种: ``None`` , ``[0]`` , ``[1]`` , ``[2]`` , ``[0,1]`` , ``[0,2]`` , ``[1,2]``, ``[0,1,2]``

- 下面以维度为 2 的输入 Tensor 展示 axis 参数的用法:

Expand All @@ -80,7 +80,7 @@ axis 参数的默认值为 None。当 ``axis=None`` 时,dropout 的功能为
假定 x 是形状为 2*3 的 2 维 Tensor:
[[1 2 3]
[4 5 6]]
(1) 若 ``axis=[0]``,则表示只在第 0 个维度做 dropout。这时生成 mask 的形状为 2*1。
(1) 若 ``axis=[0]`` ,则表示只在第 0 个维度做 dropout。这时生成 mask 的形状为 2*1。
例如,我们可能会得到这样的 mask:
[[1]
[0]]
Expand All @@ -92,7 +92,7 @@ axis 参数的默认值为 None。当 ``axis=None`` 时,dropout 的功能为
[0 0 0]]
之后依据其它参数的设置,得到最终的输出结果。

(2) 若 ``axis=[1]``,则表示只在第 1 个维度做 dropout。这时生成的 mask 形状为 1*3。
(2) 若 ``axis=[1]`` ,则表示只在第 1 个维度做 dropout。这时生成的 mask 形状为 1*3。
例如,我们可能会得到这样的 mask:
[[1 0 1]]
这个 1*3 的 mask 在和 x 做点积的时候,会首先广播成一个 2*3 的矩阵:
Expand All @@ -101,9 +101,9 @@ axis 参数的默认值为 None。当 ``axis=None`` 时,dropout 的功能为
点积所得结果为:
[[1 0 3]
[4 0 6]]
(3) 若 ``axis=[0, 1]``,则表示在第 0 维和第 1 维上做 dropout。此时与默认设置 ``axis=None`` 的作用一致。
(3) 若 ``axis=[0, 1]`` ,则表示在第 0 维和第 1 维上做 dropout。此时与默认设置 ``axis=None`` 的作用一致。

若输入 x 为 4 维 Tensor,形状为 `NCHW`,其中 N 是批尺寸,C 是通道数,H 是特征高度,W 是特征宽度,当设置 ``axis=[0,1]`` 时,则只会在通道 `N` 和 `C` 上做 dropout,通道 `H` 和 `W` 的元素是绑定在一起的,即:``paddle.nn.functional.dropout(x, p, axis=[0,1])``,此时对 4 维 Tensor 中的某个 2 维特征图(形状为 `HW`),或者全部置 0,或者全部保留,这便是 dropout2d 的实现。详情参考 :ref:`cn_api_paddle_nn_functional_dropout2d` 。
若输入 x 为 4 维 Tensor,形状为 `NCHW`,其中 N 是批尺寸,C 是通道数,H 是特征高度,W 是特征宽度,当设置 ``axis=[0,1]`` 时,则只会在通道 `N` 和 `C` 上做 dropout,通道 `H` 和 `W` 的元素是绑定在一起的,即:``paddle.nn.functional.dropout(x, p, axis=[0,1])`` ,此时对 4 维 Tensor 中的某个 2 维特征图(形状为 `HW`),或者全部置 0,或者全部保留,这便是 dropout2d 的实现。详情参考 :ref:`cn_api_paddle_nn_functional_dropout2d` 。

类似的,若输入 x 为 5 维 Tensor,形状为 `NCDHW`,其中 D 是特征深度,当设置 ``axis=[0,1]`` 时,便可实现 dropout3d。详情参考 :ref:`cn_api_paddle_nn_functional_dropout3d` 。

Expand Down