mindspore-Ecosystem
/
mindspore
21
0
Fork 27
Code Issues Pull Requests 6 Projects Releases Wiki Activity

modify the format and order of API files

This commit is contained in:
zhangyi 2022-08-19 16:27:41 +08:00
parent 4a00560a8a
commit cf3550e306
49 changed files with 173 additions and 168 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
docs/api/api_python/dataset/mindspore.dataset.Dataset.d.rst
View File

@ -53,7 +53,7 @@
- **create_data_info_queue** (bool, 可选) - 是否创建一个队列，用于存储每条数据的数据类型和shape。默认值：False，不创建。
.. note::
如果设备类型为Ascend，每次传输的数据大小限制为256MB。
如果设备类型为Ascend，数据的特征将被逐一传输。每次传输的数据大小限制为256MB。
返回：
Dataset，用于帮助发送数据到设备上的数据集对象。

2
docs/api/api_python/dataset/mindspore.dataset.Dataset.g.rst
View File

@ -115,7 +115,7 @@
- **create_data_info_queue** (bool, 可选) - 是否创建存储数据类型和shape的队列，默认值：False。
.. note::
该接口在将来会被删除或不可见，建议使用 `device_queue` 接口。
该接口在将来会被删除或不可见。建议使用 `device_queue` 接口。
如果设备为Ascend，则逐个传输数据。每次数据传输的限制为256M。
返回：

2
docs/api/api_python/dataset_audio/mindspore.dataset.audio.Angle.rst
View File

@ -5,7 +5,7 @@ mindspore.dataset.audio.Angle
计算复数序列的角度。
.. note:: 待处理音频维度需为(..., complex=2)，其中第0维代表实部，第1维代表虚部。
.. note:: 待处理音频维度需为(..., complex=2)。第0维代表实部，第1维代表虚部。
异常：
- **RuntimeError** - 当输入音频的shape不为<..., complex=2>。

2
docs/api/api_python/dataset_audio/mindspore.dataset.audio.ComplexNorm.rst
View File

@ -5,7 +5,7 @@ mindspore.dataset.audio.ComplexNorm
计算复数序列的范数。
.. note:: 待处理音频维度需为(..., complex=2)，其中第0维代表实部，第1维代表虚部。
.. note:: 待处理音频维度需为(..., complex=2)。第0维代表实部，第1维代表虚部。
参数：
- **power** (float, 可选) - 范数的幂，取值必须非负，默认值：1.0。

2
docs/api/api_python/dataset_audio/mindspore.dataset.audio.TimeStretch.rst
View File

@ -5,7 +5,7 @@ mindspore.dataset.audio.TimeStretch
以给定的比例拉伸音频短时傅里叶（Short Time Fourier Transform, STFT）频谱的时域，但不改变音频的音高。
.. note:: 待处理音频维度需为(..., freq, time, complex=2)，其中第0维代表实部，第1维代表虚部。
.. note:: 待处理音频维度需为(..., freq, time, complex=2)。第0维代表实部，第1维代表虚部。
参数：
- **hop_length** (int, 可选) - STFT窗之间每跳的长度，即连续帧之间的样本数，默认值：None，表示取 `n_freq - 1`。

2
docs/api/api_python/dataset_text/mindspore.dataset.text.BasicTokenizer.rst
View File

@ -1,5 +1,5 @@
mindspore.dataset.text.BasicTokenizer
=====================================
======================================
.. py:class:: mindspore.dataset.text.BasicTokenizer(lower_case=False, keep_whitespace=False, normalization_form=NormalizeForm.NONE, preserve_unused_token=True, with_offsets=False)

6
docs/api/api_python/mindspore/mindspore.Model.rst
View File

@ -43,7 +43,7 @@
.. warning:: 这是一个实验性接口，后续可能删除或修改。
.. note:: 如果预先调用该接口构建计算图，那么 `Model.train` 会直接执行计算图。预构建计算图目前仅支持GRAPH_MOD模式和Ascend处理器，仅支持数据下沉模式。
.. note:: 如果预先调用该接口构建计算图，那么 `Model.train` 会直接执行计算图。预构建计算图目前仅支持GRAPH_MOD模式和Ascend处理器。仅支持数据下沉模式。
参数：
- **train_dataset** (Dataset) - 一个训练集迭代器。如果定义了 `train_dataset` ，将会构建训练计算图。默认值：None。
@ -58,8 +58,8 @@
使用PyNative模式或CPU处理器时，模型评估流程将以非下沉模式执行。
.. note::
如果 `dataset_sink_mode` 配置为True，数据将被发送到处理器中。此时数据集与模型绑定，数据集仅能在当前模型中使用。如果处理器是Ascend，数据特征将被逐一传输，每次数据传输的上限是256M。
该接口会构建并执行计算图，如果使用前先执行了 `Model.build` ，那么它会直接执行计算图而不构建。
如果 `dataset_sink_mode` 配置为True，数据将被发送到处理器中。此时数据集与模型绑定，数据集仅能在当前模型中使用。如果处理器是Ascend，数据特征将被逐一传输。每次数据传输的上限是256M。
该接口会构建并执行计算图。如果使用前先执行了 `Model.build` ，那么它会直接执行计算图而不构建。
参数：
- **valid_dataset** (Dataset) - 评估模型的数据集。

2
docs/api/api_python/mindspore/mindspore.SparseTensor.rst
View File

@ -8,7 +8,7 @@ mindspore.SparseTensor
`SparseTensor` 只能在 `Cell` 的构造方法中使用。
.. note::
此接口从 1.7 版本开始弃用，并计划在将来移除，请使用 `COOTensor`。
此接口从 1.7 版本开始弃用，并计划在将来移除。请使用 `COOTensor`。
对于稠密张量，其 `SparseTensor(indices, values, shape)` 具有 `dense[indices[i]] = values[i]` 。

4
docs/api/api_python/nn/mindspore.nn.Cell.rst
View File

@ -460,7 +460,7 @@
递归设置该Cell中的所有算子的并行策略为数据并行。
.. note:: 仅在图模式、全自动并行(AUTO_PARALLEL)模式下生效。
.. note:: 仅在图模式，使用auto_parallel_context = ParallelMode.AUTO_PARALLEL生效。
.. py:method:: set_grad(requires_grad=True)
@ -527,7 +527,7 @@
其中的每一个元素指定对应的输入/输出的Tensor分布策略，可参考： `mindspore.ops.Primitive.shard` 的描述，也可以设置为None，会默认以数据并行执行。
其余算子的并行策略由输入输出指定的策略推导得到。
.. note:: 需设置为PyNative模式，并且全自动并行(AUTO_PARALLEL)，同时设置 `set_auto_parallel_context` 中的搜索模式(search mode)为"sharding_propagation"。
.. note:: 需设置为PyNative模式，并且ParallelMode.AUTO_PARALLEL，同时设置 `set_auto_parallel_context` 中的搜索模式(search mode)为"sharding_propagation"。
参数：
- **in_strategy** (tuple) - 指定各输入的切分策略，输入元组的每个元素可以为元组或None，元组即具体指定输入每一维的切分策略，None则会默认以数据并行执行。

2
docs/api/api_python/nn/mindspore.nn.OneHot.rst
View File

@ -8,7 +8,7 @@ mindspore.nn.OneHot
输入的 `indices` 表示的位置取值为on_value，其他所有位置取值为off_value。
.. note::
如果indices是n阶Tensor，那么返回的one-hot Tensor则为n+1阶Tensor，新增 `axis` 维度。
如果indices是n阶Tensor，那么返回的one-hot Tensor则为n+1阶Tensor。新增 `axis` 维度。
如果 `indices` 是Scalar，则输出shape将是长度为 `depth` 的向量。

2
docs/api/api_python/nn/mindspore.nn.optim_note_sparse.rst
View File

@ -1,2 +1,2 @@
如果前向网络使用了SparseGatherV2等算子，优化器会执行稀疏运算。通过设置 `target` 为CPU，可在主机（host）上进行稀疏运算。
如果前向网络使用了SparseGatherV2等算子，优化器会执行稀疏运算，通过设置 `target` 为CPU，可在主机（host）上进行稀疏运算。
稀疏特性在持续开发中。

2
docs/api/api_python/nn/mindspore.nn.optim_note_weight_decay.rst
View File

@ -2,4 +2,4 @@
参数分组情况下，可以分组调整权重衰减策略。
分组时，每组网络参数均可配置 `weight_decay` ，若未配置，则该组网络参数使用优化器中配置的 `weight_decay` 。
分组时，每组网络参数均可配置 `weight_decay` 。若未配置，则该组网络参数使用优化器中配置的 `weight_decay` 。

4
docs/api/api_python/nn_probability/mindspore.nn.probability.bijector.Bijector.rst
View File

@ -19,11 +19,11 @@ mindspore.nn.probability.bijector.Bijector
但所有参数都应具有相同的float类型，否则将引发TypeError。
具体来说，参数类型跟随输入值的数据类型，即当 `dtype` 为None时，Bijector的参数将被强制转换为与输入值相同的类型。
具体来说，参数类型跟随输入值的数据类型。即当 `dtype` 为None时，Bijector的参数将被强制转换为与输入值相同的类型。
当指定了 `dtype` 时，参数和输入值的 `dtype` 必须相同。
当参数类型或输入值类型与 `dtype` 不相同时，将引发TypeError。只能使用mindspore的float数据类型来指定Bijector的 `dtype` 。
当参数类型或输入值类型与 `dtype` 不相同时，将引发TypeError。只能使用mindspore.float_type数据类型来指定Bijector的 `dtype` 。
.. py:method:: cast_param_by_value(value, para)

1
docs/api/api_python/nn_probability/mindspore.nn.probability.distribution.Beta.rst
View File

@ -20,6 +20,7 @@ mindspore.nn.probability.distribution.Beta
.. note::
- `concentration1` 和 `concentration0` 中元素必须大于零。
- `dist_spec_args` 是 `concentration1` 和 `concentration0`。
- `dtype` 必须是float，因为 Beta 分布是连续的。
异常：

2
docs/api/api_python/nn_probability/mindspore.nn.probability.distribution.Geometric.rst
View File

@ -15,7 +15,7 @@ mindspore.nn.probability.distribution.Geometric
- **name** (str) - 分布的名称。默认值：'Geometric'。
.. note::
`probs` 必须是合适的概率（0<p<1）。
`probs` 必须是合适的概率（0<p<1）。`dist_spec_args` 是 `probs`。
异常：

2
docs/api/api_python/nn_probability/mindspore.nn.probability.distribution.Poisson.rst
View File

@ -18,7 +18,7 @@ mindspore.nn.probability.distribution.Poisson
- **name** (str) - 分布的名称。默认值：'Poisson'。
.. note::
`rate` 必须大于0。
`rate` 必须大于0。 `dist_spec_args` 是 `rate`。
异常：
- **ValueError** - `rate` 中元素小于0。

2
docs/api/api_python/ops/mindspore.ops.AllReduce.rst
View File

@ -6,7 +6,7 @@
使用指定方式对通信组内的所有设备的Tensor数据进行规约操作，所有设备都得到相同的结果
.. note::
AllReduce操作暂不支持"prod"。集合中的所有进程的Tensor必须具有相同的shape和格式。用户在使用之前需要设置环境变量，运行下面的例子。获取详情请点击官方网站 `MindSpore <https://www.mindspore.cn/docs/zh-CN/master/api_python/mindspore.ops.html#通信算子>`_ 。
AllReduce操作暂不支持"prod"。集合中的所有进程的Tensor必须具有相同的shape和格式。用户在使用之前需要设置环境变量，运行下面的例子，获取详情请点击官方网站 `MindSpore <https://www.mindspore.cn/docs/zh-CN/master/api_python/mindspore.ops.html#通信算子>`_ 。
参数：
- **op** (str) - 规约的具体操作，如"sum"、"max"、和"min"。默认值：ReduceOp.SUM。

3
docs/api/api_python/ops/mindspore.ops.Broadcast.rst
View File

@ -6,7 +6,8 @@
对输入数据整组广播。
.. note::
集合中的所有进程的Tensor的shape和数据格式相同。
集合中的所有进程的Tensor的shape和数据格式相同。在运行下面样例时，用户需要预设通信环境变量，请在 `MindSpore \
<https://www.mindspore.cn/docs/zh-CN/master/api_python/mindspore.ops.html#通信算子>`_ 官网上查看详情。
参数：
- **root_rank** (int) - 表示发送源的进程编号。除发送数据的进程外，存在于所有进程中。

2
docs/api/api_python/ops/mindspore.ops.Print.rst
View File

@ -9,7 +9,7 @@ mindspore.ops.Print
.. note::
在PyNative模式下，请使用Python print函数。在Graph模式下，bool、int和float将被转换为Tensor进行打印，str保持不变。
该方法用于代码调试，当同时print大量数据时，为了保证主进程不受影响，可能会丢失一些数据，这时推荐使用 `Summary` 功能，具体可查看
该方法用于代码调试。当同时print大量数据时，为了保证主进程不受影响，可能会丢失一些数据，这时推荐使用 `Summary` 功能。具体可查看
`Summary <https://www.mindspore.cn/mindinsight/docs/zh-CN/master/summary_record.html?highlight=summary#>`_。
输入：

2
docs/api/api_python/ops/mindspore.ops.Range.rst
View File

@ -4,6 +4,6 @@ mindspore.ops.Range
.. py:class:: mindspore.ops.Range(maxlen=1000000)
返回从 `start` 开始，步长为 `delta` ，且不超过 `limit` （不包括 `limit` ）的序列。
序列的长度不能超过`maxlen`。 `maxlen`的默认值为1000000。
序列的长度不能超过 `maxlen`。 `maxlen` 的默认值为1000000。
更多参考详见 :func:`mindspore.ops.range`。

6
docs/api/api_python/ops/mindspore.ops.UniformInt.rst
View File

@ -24,11 +24,11 @@ mindspore.ops.UniformInt
- **maxval** (Tensor) - 分布参数， :math:`b` 。
决定生成随机数的上限，数据类型为int32。需为标量。
输出：
Tensor。shape为输入 `shape` ，数据类型支持int32。
异常：
- **TypeError** - `seed` 或 `seed2` 不是int类型。
- **TypeError** - `shape` 不是Tuple。
- **TypeError** - `minval` 或 `maxval` 不是Tensor。
- **ValueError** - `shape` 不是常量值。
输出：
Tensor。shape为输入 `shape` ，数据类型支持int32。

52
docs/api/api_python/ops/mindspore.ops.func_conv2d.rst
View File

@ -17,33 +17,31 @@ mindspore.ops.conv2d
请参考论文 `Gradient Based Learning Applied to Document Recognition <http://vision.stanford.edu/cs598_spring07/papers/Lecun98.pdf>`_ 。更详细的介绍，参见：http://cs231n.github.io/convolutional-networks/。
**参数：**
- **x** (Tensor): shape为 :math:`(N, C_{in}, H_{in}, W_{in})` 的Tensor.
- **weight** (Tensor) - 设置卷积核的大小为 :math:`(\text{kernel_size[0]}, \text{kernel_size[1]})` ，则shape为 :math:`(C_{out}, C_{in}, \text{kernel_size[0]}, \text{kernel_size[1]})` 。
- **kernel_size** (Union[int, tuple[int]]) - 数据类型为int或一个包含2个int组成的元组。指定二维卷积核的高度和宽度。单个整数表示该值同时适用于内核的高度和宽度。包含2个整数的元组表示第一个值用于高度，另一个值用于内核的宽度。
- **mode** (int) - 指定不同的卷积模式。此值目前未被使用。默认值：1。
- **pad_mode** (str) - 指定填充模式。取值为"same"，"valid"，或"pad"。默认值："valid"。
参数：
- **x** (Tensor): shape为 :math:`(N, C_{in}, H_{in}, W_{in})` 的Tensor.
- **weight** (Tensor) - 设置卷积核的大小为 :math:`(\text{kernel_size[0]}, \text{kernel_size[1]})` ，则shape为 :math:`(C_{out}, C_{in}, \text{kernel_size[0]}, \text{kernel_size[1]})` 。
- **kernel_size** (Union[int, tuple[int]]) - 数据类型为int或一个包含2个int组成的元组。指定二维卷积核的高度和宽度。单个整数表示该值同时适用于内核的高度和宽度。包含2个整数的元组表示第一个值用于高度，另一个值用于内核的宽度。
- **mode** (int) - 指定不同的卷积模式。此值目前未被使用。默认值：1。
- **pad_mode** (str) - 指定填充模式。取值为"same"，"valid"，或"pad"。默认值："valid"。
- **same**: 输出的高度和宽度分别与输入整除 `stride` 后的值相同。填充将被均匀地添加到高和宽的两侧，剩余填充量将被添加到维度末端。若设置该模式，`pad_val` 的值必须为0。
- **valid**: 在不填充的前提下返回有效计算所得的输出。不满足计算的多余像素会被丢弃。如果设置此模式，则 `pad_val` 的值必须为0。
- **pad**: 对输入 `x` 进行填充。在输入的高度和宽度方向上填充 `pad_val` 大小的0。如果设置此模式， `pad_val` 必须大于或等于0。
- **pad_val** (Union(int, tuple[int])) - 输入 `x` 的高度和宽度方向上填充的数量。数据类型为int或包含4个int组成的tuple。如果 `pad_val` 是一个int，那么上、下、左、右的填充都等于 `pad_val` 。如果 `pad_val` 是一个有4个int组成的tuple，那么上、下、左、右的填充分别等于 `pad_val[0]` 、 `pad_val[1]` 、 `pad_val[2]` 和 `pad_val[3]` 。值应该要大于等于0，默认值：0。
- **stride** (Union(int, tuple[int])) - 卷积核移动的步长，数据类型为int或两个int组成的tuple。一个int表示在高度和宽度方向的移动步长均为该值。两个int组成的tuple分别表示在高度和宽度方向的移动步长。默认值：1。
- **dilation** (Union(int, tuple[int])) - 卷积核膨胀尺寸。数据类型为int或由2个int组成的tuple。若 :math:`k > 1` ，则卷积核间隔 `k` 个元素进行采样。垂直和水平方向上的 `k` ，其取值范围分别为[1, H]和[1, W]。默认值：1。
- **group** (int) - 将过滤器拆分为组。默认值：1。
- **data_format** (str) - 数据格式的可选值有"NHWC"，"NCHW"。默认值："NCHW"。
- **same**: 输出的高度和宽度分别与输入整除 `stride` 后的值相同。填充将被均匀地添加到高和宽的两侧，剩余填充量将被添加到维度末端。若设置该模式，`pad_val` 的值必须为0。
- **valid**: 在不填充的前提下返回有效计算所得的输出。不满足计算的多余像素会被丢弃。如果设置此模式，则 `pad_val` 的值必须为0。
- **pad**: 对输入 `x` 进行填充。在输入的高度和宽度方向上填充 `pad_val` 大小的0。如果设置此模式， `pad_val` 必须大于或等于0。
- **pad_val** (Union(int, tuple[int])) - 输入 `x` 的高度和宽度方向上填充的数量。数据类型为int或包含4个int组成的tuple。如果 `pad_val` 是一个int，那么上、下、左、右的填充都等于 `pad_val` 。如果 `pad_val` 是一个有4个int组成的tuple，那么上、下、左、右的填充分别等于 `pad_val[0]` 、 `pad_val[1]` 、 `pad_val[2]` 和 `pad_val[3]` 。值应该要大于等于0，默认值：0。
- **stride** (Union(int, tuple[int])) - 卷积核移动的步长，数据类型为int或两个int组成的tuple。一个int表示在高度和宽度方向的移动步长均为该值。两个int组成的tuple分别表示在高度和宽度方向的移动步长。默认值：1。
- **dilation** (Union(int, tuple[int])) - 卷积核膨胀尺寸。数据类型为int或由2个int组成的tuple。若 :math:`k > 1` ，则卷积核间隔 `k` 个元素进行采样。垂直和水平方向上的 `k` ，其取值范围分别为[1, H]和[1, W]。默认值：1。
- **group** (int) - 将过滤器拆分为组。默认值：1。
- **data_format** (str) - 数据格式的可选值有"NHWC"，"NCHW"。默认值："NCHW"。
**返回：**
返回：
Tensor，卷积后的值。shape为 :math:`(N, C_{out}, H_{out}, W_{out})` 。
Tensor，卷积后的值。shape为 :math:`(N, C_{out}, H_{out}, W_{out})` 。
**异常：**
- **TypeError** - `kernel_size` 、 `stride` 、 `pad_val` 或 `dilation` 既不是int也不是tuple。
- **TypeError** - `out_channel` 或 `group` 不是int。
- **ValueError** - `kernel_size` 、 `stride` 或 `diation` 小于1。
- **ValueError** - `pad_mode` 不是"same"、"valid"或"pad"。
- **ValueError** - `pad_val` 是一个长度不等于4的tuple。
- **ValueError** - `pad_mode` 不等于"pad"，`pad_val` 不等于(0, 0, 0, 0)。
- **ValueError** - `data_format` 既不是"NCW"，也不是"NHWC"。
异常：
- **TypeError** - `kernel_size` 、 `stride` 、 `pad_val` 或 `dilation` 既不是int也不是tuple。
- **TypeError** - `out_channel` 或 `group` 不是int。
- **ValueError** - `kernel_size` 、 `stride` 或 `diation` 小于1。
- **ValueError** - `pad_mode` 不是"same"、"valid"或"pad"。
- **ValueError** - `pad_val` 是一个长度不等于4的tuple。
- **ValueError** - `pad_mode` 不等于"pad"，`pad_val` 不等于(0, 0, 0, 0)。
- **ValueError** - `data_format` 既不是"NCW"，也不是"NHWC"。

1
docs/api/api_python/ops/mindspore.ops.func_scatter_mul.rst
View File

@ -21,6 +21,7 @@ mindspore.ops.scatter_mul
Tensor，更新后的 `input_x` ，shape和类型与 `input_x` 相同。
异常：
- **TypeError** - `use_locking` 不是bool。
- **TypeError** - `indices` 不是int32。
- **ValueError** - `updates` 的shape不等于 `indices.shape + x.shape[1:]` 。
- **RuntimeError** - 当 `input_x` 和 `updates` 类型不一致，需要进行类型转换时，如果 `updates` 不支持转成参数 `input_x` 需要的数据类型，就会报错。

18
mindspore/python/mindspore/nn/metrics/metric.py
View File

@ -30,6 +30,15 @@ def rearrange_inputs(func):
This decorator is currently applied on the `update` of :class:`mindspore.nn.Metric`.
Args:
func (Callable): A candidate function to be wrapped whose input will be rearranged.
Returns:
Callable, used to exchange metadata between functions.
Supported Platforms:
``Ascend`` ``GPU`` ``CPU``
Examples:
>>> from mindspore.nn import rearrange_inputs
>>> class RearrangeInputsExample:
@ -52,15 +61,6 @@ def rearrange_inputs(func):
>>> outs = rearrange_inputs_example.update(5, 9)
>>> print(outs)
(9, 5)
Args:
func (Callable): A candidate function to be wrapped whose input will be rearranged.
Returns:
Callable, used to exchange metadata between functions.
Supported Platforms:
``Ascend`` ``GPU`` ``CPU``
"""
@functools.wraps(func)
def wrapper(self, *inputs):

6
mindspore/python/mindspore/nn/optim/sgd.py
View File

@ -60,6 +60,12 @@ class SGD(Optimizer):
Here : where p, v and u denote the parameters, accum, and momentum respectively.
Note:
If parameters are not grouped, the `weight_decay` in optimizer will be applied on the network parameters without
'beta' or 'gamma' in their names. Users can group parameters to change the strategy of decaying weight. When
parameters are grouped, each group can set `weight_decay`. If not, the `weight_decay` in optimizer will be
applied.
Args:
params (Union[list[Parameter], list[dict]]): Must be list of `Parameter` or list of `dict`. When the
`params` is a list of `dict`, the string "params", "lr", "grad_centralization" and

6
mindspore/python/mindspore/nn/probability/bijector/bijector.py
View File

@ -38,9 +38,6 @@ class Bijector(Cell):
dtype (mindspore.dtype): The type of the distributions that the Bijector can operate on. Default: None.
param (dict): The parameters used to initialize the Bijector. Default: None.
Supported Platforms:
``Ascend`` ``GPU``
Note:
`dtype` of bijector represents the type of the distributions that the bijector could operate on.
When `dtype` is None, there is no enforcement on the type of input value except that the input value
@ -51,6 +48,9 @@ class Bijector(Cell):
When `dtype` is specified, it is forcing the parameters and input value to be the same dtype as `dtype`.
When the type of parameters or the type of the input value is not the same as `dtype`, a TypeError will be
raised. Only subtype of mindspore.float_type can be used to specify bijector's `dtype`.
Supported Platforms:
``Ascend`` ``GPU``
"""
def __init__(self,

6
mindspore/python/mindspore/nn/probability/distribution/beta.py
View File

@ -44,9 +44,9 @@ class Beta(Distribution):
name (str): The name of the distribution. Default: 'Beta'.
Note:
`concentration1` and `concentration0` must be greater than zero.
`dist_spec_args` are `concentration1` and `concentration0`.
`dtype` must be a float type because Beta distributions are continuous.
- `concentration1` and `concentration0` must be greater than zero.
- `dist_spec_args` are `concentration1` and `concentration0`.
- `dtype` must be a float type because Beta distributions are continuous.
Raises:
ValueError: When concentration1 <= 0 or concentration0 >=1.

6
mindspore/python/mindspore/nn/probability/distribution/categorical.py
View File

@ -39,15 +39,15 @@ class Categorical(Distribution):
dtype (mindspore.dtype): The type of the event samples. Default: mstype.int32.
name (str): The name of the distribution. Default: Categorical.
Supported Platforms:
``Ascend`` ``GPU``
Note:
`probs` must have rank at least 1, values are proper probabilities and sum to 1.
Raises:
ValueError: When the sum of all elements in `probs` is not 1.
Supported Platforms:
``Ascend`` ``GPU``
Examples:
>>> import mindspore
>>> import mindspore.nn as nn

6
mindspore/python/mindspore/nn/probability/distribution/cauchy.py
View File

@ -41,9 +41,6 @@ class Cauchy(Distribution):
dtype (mindspore.dtype): The type of the event samples. Default: mstype.float32.
name (str): The name of the distribution. Default: 'Cauchy'.
Supported Platforms:
``Ascend``
Note:
`scale` must be greater than zero.
`dist_spec_args` are `loc` and `scale`.
@ -54,6 +51,9 @@ class Cauchy(Distribution):
ValueError: When scale <= 0.
TypeError: When the input `dtype` is not a subclass of float.
Supported Platforms:
``Ascend``
Examples:
>>> import mindspore
>>> import mindspore.nn as nn

6
mindspore/python/mindspore/nn/probability/distribution/exponential.py
View File

@ -40,9 +40,6 @@ class Exponential(Distribution):
dtype (mindspore.dtype): The type of the event samples. Default: mstype.float32.
name (str): The name of the distribution. Default: 'Exponential'.
Supported Platforms:
``Ascend`` ``GPU``
Note:
`rate` must be strictly greater than 0.
`dist_spec_args` is `rate`.
@ -52,6 +49,9 @@ class Exponential(Distribution):
ValueError: When rate <= 0.
TypeError: When the input `dtype` is not a subclass of float.
Supported Platforms:
``Ascend`` ``GPU``
Examples:
>>> import mindspore
>>> import mindspore.nn as nn

6
mindspore/python/mindspore/nn/probability/distribution/gamma.py
View File

@ -45,9 +45,6 @@ class Gamma(Distribution):
dtype (mindspore.dtype): The type of the event samples. Default: mstype.float32.
name (str): The name of the distribution. Default: 'Gamma'.
Supported Platforms:
``Ascend``
Note:
`concentration` and `rate` must be greater than zero.
`dist_spec_args` are `concentration` and `rate`.
@ -57,6 +54,9 @@ class Gamma(Distribution):
ValueError: When concentration <= 0 or rate <= 0.
TypeError: When the input `dtype` is not a subclass of float.
Supported Platforms:
``Ascend``
Examples:
>>> import mindspore
>>> import mindspore.nn as nn

6
mindspore/python/mindspore/nn/probability/distribution/geometric.py
View File

@ -37,9 +37,6 @@ class Geometric(Distribution):
dtype (mindspore.dtype): The type of the event samples. Default: mstype.int32.
name (str): The name of the distribution. Default: 'Geometric'.
Supported Platforms:
``Ascend`` ``GPU``
Note:
`probs` must be a proper probability (0 < p < 1).
`dist_spec_args` is `probs`.
@ -47,6 +44,9 @@ class Geometric(Distribution):
Raises:
ValueError: When p <= 0 or p >= 1.
Supported Platforms:
``Ascend`` ``GPU``
Examples:
>>> import mindspore
>>> import mindspore.nn as nn

6
mindspore/python/mindspore/nn/probability/distribution/gumbel.py
View File

@ -43,9 +43,6 @@ class Gumbel(TransformedDistribution):
dtype (mindspore.dtype): type of the distribution. Default: mstype.float32.
name (str): the name of the distribution. Default: 'Gumbel'.
Supported Platforms:
``Ascend`` ``GPU``
Note:
`scale` must be greater than zero.
`dist_spec_args` are `loc` and `scale`.
@ -55,6 +52,9 @@ class Gumbel(TransformedDistribution):
ValueError: When scale <= 0.
TypeError: When the input `dtype` is not a subclass of float.
Supported Platforms:
``Ascend`` ``GPU``
Examples:
>>> import mindspore
>>> import numpy as np

6
mindspore/python/mindspore/nn/probability/distribution/log_normal.py
View File

@ -44,9 +44,6 @@ class LogNormal(msd.TransformedDistribution):
dtype (mindspore.dtype): type of the distribution. Default: mstype.float32.
name (str): the name of the distribution. Default: 'LogNormal'.
Supported Platforms:
``Ascend`` ``GPU``
Note:
`scale` must be greater than zero.
`dist_spec_args` are `loc` and `scale`.
@ -56,6 +53,9 @@ class LogNormal(msd.TransformedDistribution):
ValueError: When scale <= 0.
TypeError: When the input `dtype` is not a subclass of float.
Supported Platforms:
``Ascend`` ``GPU``
Examples:
>>> import numpy as np
>>> import mindspore

6
mindspore/python/mindspore/nn/probability/distribution/logistic.py
View File

@ -41,9 +41,6 @@ class Logistic(Distribution):
dtype (mindspore.dtype): The type of the event samples. Default: mstype.float32.
name (str): The name of the distribution. Default: 'Logistic'.
Supported Platforms:
``Ascend`` ``GPU``
Note:
`scale` must be greater than zero.
`dist_spec_args` are `loc` and `scale`.
@ -53,6 +50,9 @@ class Logistic(Distribution):
ValueError: When scale <= 0.
TypeError: When the input `dtype` is not a subclass of float.
Supported Platforms:
``Ascend`` ``GPU``
Examples:
>>> import mindspore
>>> import mindspore.nn as nn

6
mindspore/python/mindspore/nn/probability/distribution/normal.py
View File

@ -43,9 +43,6 @@ class Normal(Distribution):
dtype (mindspore.dtype): The type of the event samples. Default: mstype.float32.
name (str): The name of the distribution. Default: 'Normal'.
Supported Platforms:
``Ascend`` ``GPU``
Note:
`sd` must be greater than zero.
`dist_spec_args` are `mean` and `sd`.
@ -55,6 +52,9 @@ class Normal(Distribution):
ValueError: When sd <= 0.
TypeError: When the input `dtype` is not a subclass of float.
Supported Platforms:
``Ascend`` ``GPU``
Examples:
>>> import mindspore
>>> import mindspore.nn as nn

6
mindspore/python/mindspore/nn/probability/distribution/poisson.py
View File

@ -37,9 +37,6 @@ class Poisson(Distribution):
dtype (mindspore.dtype): The type of the event samples. Default: mstype.float32.
name (str): The name of the distribution. Default: 'Poisson'.
Supported Platforms:
``Ascend``
Note:
`rate` must be strictly greater than 0.
`dist_spec_args` is `rate`.
@ -47,6 +44,9 @@ class Poisson(Distribution):
Raises:
ValueError: When rate <= 0.
Supported Platforms:
``Ascend``
Examples:
>>> import mindspore
>>> import mindspore.nn as nn

14
mindspore/python/mindspore/nn/probability/distribution/transformed_distribution.py
View File

@ -40,13 +40,6 @@ class TransformedDistribution(Distribution):
will use this seed; elsewise, the underlying distribution's seed will be used.
name (str): The name of the transformed distribution. Default: 'transformed_distribution'.
Supported Platforms:
``Ascend`` ``GPU`` ``CPU``
Raises:
TypeError: When the input `bijector` is not a Bijector instance.
TypeError: When the input `distribution` is not a Distribution instance.
Note:
The arguments used to initialize the original distribution cannot be None.
For example, mynormal = msd.Normal(dtype=mindspore.float32) cannot be used to initialized a
@ -58,6 +51,13 @@ class TransformedDistribution(Distribution):
distribution. Derived class can overwrite `default_parameters` and `parameter_names` by calling
`reset_parameters` followed by `add_parameter`.
Raises:
TypeError: When the input `bijector` is not a Bijector instance.
TypeError: When the input `distribution` is not a Distribution instance.
Supported Platforms:
``Ascend`` ``GPU`` ``CPU``
Examples:
>>> import numpy as np
>>> import mindspore

5
mindspore/python/mindspore/nn/probability/distribution/uniform.py
View File

@ -41,9 +41,6 @@ class Uniform(Distribution):
dtype (mindspore.dtype): The type of the event samples. Default: mstype.float32.
name (str): The name of the distribution. Default: 'Uniform'.
Supported Platforms:
``Ascend`` ``GPU``
Note:
`low` must be strictly less than `high`.
`dist_spec_args` are `high` and `low`.
@ -53,6 +50,8 @@ class Uniform(Distribution):
ValueError: When high <= low.
TypeError: When the input `dtype` is not a subclass of float.
Supported Platforms:
``Ascend`` ``GPU``
Examples:
>>> import mindspore

2
mindspore/python/mindspore/ops/composite/math_ops.py
View File

@ -560,7 +560,7 @@ def batch_dot(x1, x2, axes=None):
Default: None.
Returns:
Tensor, batch dot product of `x1` and `x2`. For example: The Shape of output
Tensor, batch dot product of `x1` and `x2`. For example, the Shape of output
for input `x1` shapes (batch, d1, axes, d2) and `x2` shapes (batch, d3, axes, d4) is (batch, d1, d2, d3, d4),
where d1 and d2 means any number.

2
mindspore/python/mindspore/ops/function/array_func.py
View File

@ -1075,7 +1075,7 @@ def slice(input_x, begin, size):
size (Union[tuple, list]): The size of the slice. Only constant value is allowed.
Returns:
Tensor, the shape is : input `size`, the data type is the same as `input_x`.
Tensor, the shape is input `size`, the data type is the same as `input_x`.
Raises:
TypeError: If `begin` or `size` is neither tuple nor list.

2
mindspore/python/mindspore/ops/function/math_func.py
View File

@ -2239,7 +2239,7 @@ def matrix_solve(matrix, rhs, adjoint=False):
adjoint(bool): Indicating whether to solve with matrix or its (block-wise) adjoint. Default: False.
Returns:
x (Tensor): The dtype and shape is the same as 'rhs'.
x (Tensor), The dtype and shape is the same as 'rhs'.
Raises:
TypeError: If adjoint is not the type of bool.

42
mindspore/python/mindspore/ops/function/nn_func.py
View File

@ -381,15 +381,15 @@ def celu(x, alpha=1.0):
Returns:
Tensor, has the same data type and shape as the input.
Supported Platforms:
``Ascend`` ``CPU`` ``GPU``
Raises:
TypeError: If `alpha` is not a float.
ValueError: If `alpha` has the value of 0.
TypeError: If `x` is not a Tensor.
TypeError: If dtype of `x` is neither float16 nor float32.
Supported Platforms:
``Ascend`` ``CPU`` ``GPU``
Examples:
>>> x = Tensor(np.array([-2.0, -1.0, 1.0, 2.0]), mindspore.float32)
>>> output = ops.celu(x, alpha=1.0)
@ -563,14 +563,14 @@ def kl_div(logits, labels, reduction='mean'):
Tensor or Scalar, if `reduction` is 'none', then output is a tensor and has the same shape as `logits`.
Otherwise it is a scalar.
Supported Platforms:
``Ascend`` ``CPU`` ``GPU``
Raises:
TypeError: If `reduction` is not a str.
TypeError: If neither `logits` nor `labels` is a Tensor.
TypeError: If dtype of `logits` or `labels` is not float32.
Supported Platforms:
``Ascend`` ``CPU`` ``GPU``
Examples:
>>> class Net(nn.Cell):
... def __init__(self):
@ -616,14 +616,14 @@ def hardshrink(x, lambd=0.5):
Returns:
Tensor, has the same data type and shape as the input `x`.
Supported Platforms:
``Ascend`` ``GPU`` ``CPU``
Raises:
TypeError: If `lambd` is not a float.
TypeError: If `x` is not a tensor.
TypeError: If dtype of `x` is neither float16 nor float32.
Supported Platforms:
``Ascend`` ``GPU`` ``CPU``
Examples:
>>> x = Tensor(np.array([[ 0.5, 1, 2.0], [0.0533,0.0776,-2.1233]]), mindspore.float32)
>>> output = ops.hardshrink(x)
@ -772,9 +772,6 @@ def interpolate(x, roi=None, scales=None, sizes=None, coordinate_transformation_
Returns:
Resized tensor, with the same data type as input `x`.
Supported Platforms:
``Ascend`` ``CPU`` ``GPU``
Raises:
TypeError: If `x` is not a Tensor.
TypeError: If the data type of `x` is not supported.
@ -787,6 +784,9 @@ def interpolate(x, roi=None, scales=None, sizes=None, coordinate_transformation_
TypeError: If `mode` is not a string.
ValueError: If `mode` is not in the support list.
Supported Platforms:
``Ascend`` ``CPU`` ``GPU``
Examples:
>>> # case 1: linear mode
>>> x = Tensor([[[1, 2, 3], [4, 5, 6]]], mindspore.float32)
@ -972,12 +972,12 @@ def selu(input_x):
Returns:
Tensor, with the same type and shape as the `input_x`.
Supported Platforms:
``Ascend`` ``GPU`` ``CPU``
Raises:
TypeError: If dtype of `input_x` is neither float16 nor float32.
Supported Platforms:
``Ascend`` ``GPU`` ``CPU``
Examples:
>>> input_x = Tensor(np.array([[-1.0, 4.0, -8.0], [2.0, -5.0, 9.0]]), mindspore.float32)
>>> output = ops.selu(input_x)
@ -1046,9 +1046,6 @@ def deformable_conv2d(x, weight, offsets, kernel_size, strides, padding, bias=No
\text{dilations[3]} - 1 }{\text{stride[1]}} + 1} \right \rfloor \\
\end{array}
Supported Platforms:
``Ascend`` ``GPU`` ``CPU``
Raises:
TypeError: If `strides`, `padding`, `kernel_size` or `dilations` is not a tuple with integer elements.
TypeError: If `modulated` is not a bool.
@ -1066,6 +1063,9 @@ def deformable_conv2d(x, weight, offsets, kernel_size, strides, padding, bias=No
with "numpy.ones()".
- `kernel_size` should meet the requirement::math:`3 * kernel\_size[0] * kernel\_size[1] > 8`.
Supported Platforms:
``Ascend`` ``GPU`` ``CPU``
Examples:
>>> x = Tensor(np.ones((4, 3, 10, 10)), mstype.float32)
>>> kh, kw = 3, 3
@ -1725,12 +1725,12 @@ def mish(x):
Returns:
Tensor, with the same type and shape as the `x`.
Supported Platforms:
``Ascend`` ``CPU``
Raises:
TypeError: If dtype of `x` is neither float16 nor float32.
Supported Platforms:
``Ascend`` ``CPU``
Examples:
>>> input_x = Tensor(np.array([[-1.0, 4.0, -8.0], [2.0, -5.0, 9.0]]), mindspore.float32)
>>> output = ops.mish(input_x)

8
mindspore/python/mindspore/ops/function/sparse_func.py
View File

@ -320,7 +320,7 @@ def dense_to_sparse_coo(tensor):
tensor: A dense tensor, must be 2-D.
Returns:
COOTensor, a sparse representation of the original dense tensor, containing:
COOTensor, a sparse representation of the original dense tensor, containing the following parts.
- indices (Tensor): 2-D integer tensor, indicates the positions of `values` of the dense tensor.
- values (Tensor): 1-D tensor, indicates the non-zero values of the dense tensor.
@ -366,7 +366,7 @@ def dense_to_sparse_csr(tensor):
tensor: A dense tensor, must be 2-D.
Returns:
CSRTensor, a sparse representation of the original dense tensor, containing:
CSRTensor, a sparse representation of the original dense tensor, containing the following parts.
- indptr (Tensor): 1-D integer tensor, indicates the start and end point for `values` in each row.
- indices (Tensor): 1-D integer tensor, indicates the column positions of all non-zero values of the input.
@ -454,8 +454,8 @@ def sparse_concat(sp_input, concat_dim=0):
Outputs:
- **output** (COOtensor) - the result of concatenates the input SparseTensor along the
specified dimension. OutShape: OutShape[non concat_dim] is equal to InShape[non concat_dim] and
OutShape[concat_dim] is all input concat_dim axis shape accumulate.
specified dimension. OutShape: OutShape[non concat_dim] is equal to InShape[non concat_dim] and
OutShape[concat_dim] is all input concat_dim axis shape accumulate.
Raises:
ValueError: If only one sparse tensor input.

2
mindspore/python/mindspore/ops/operations/comm_ops.py
View File

@ -188,7 +188,7 @@ class AllGather(PrimitiveWithInfer):
Note:
The tensors must have the same shape and format in all processes of the collection. The user needs to preset
communication environment variables before running the following example, please check the details on the
communication environment variables before running the following example. Please check the details on the
official website of `MindSpore \
<https://www.mindspore.cn/docs/en/master/api_python/mindspore.ops.html#communication-operator>`_.

2
mindspore/python/mindspore/ops/operations/debug_ops.py
View File

@ -433,7 +433,7 @@ class Print(PrimitiveWithInfer):
str remains unchanged.
This function is used for debugging. When too much data is printed at the same time,
in order not to affect the main process, the framework may discard some data. At this time,
if you need to record the data completely, you can recommended to use the `Summary` function. Please check
if you need to record the data completely, you can recommended to use the `Summary` function, and can check
`Summary <https://www.mindspore.cn/mindinsight/docs/zh-CN/master/summary_record.html?highlight=summary#>`_.
Inputs:

31
mindspore/python/mindspore/ops/operations/math_ops.py
View File

@ -3074,13 +3074,12 @@ class MulNoNan(_MathBinaryOp):
Tensor, the shape is the same as the shape after broadcasting,
and the data type is the one with higher precision among the two inputs.
Raises:
TypeError: If neither `x` nor `y` is a Tensor.
Supported Platforms:
``Ascend`` ``GPU`` ``CPU``
Raises:
TypeError: If neither `x` nor `y` is a Tensor.
Examples:
>>> # case 1 : same data type and shape of two inputs, there are some 0 in y.
>>> x = Tensor(np.array([[-1.0, 6.0, np.inf], [np.nan, -7.0, 4.0]]), mindspore.float32)
@ -5375,7 +5374,7 @@ class MatrixInverse(Primitive):
result may be returned.
Note:
The parameter 'adjoint' is only supporting False right now. Because complex number is not supported at present.
The parameter 'adjoint' is only supporting False right now, because complex number is not supported at present.
Args:
adjoint (bool) : An optional bool. Default: False.
@ -6254,23 +6253,23 @@ class RaggedRange(Primitive):
Inputs:
- **starts** (Tensor) - The starts of each range, whose type is int32, int64, float32 or float64,
and shape is 0D or 1D.
and shape is 0D or 1D.
- **limits** (Tensor) - The limits of each range, whose type and shape should be same as input `starts`.
- **deltas** (Tensor) - The deltas of each range, whose type and shape should be same as input `starts`,
and each element in the tensor should not be equal to 0.
and each element in the tensor should not be equal to 0.
Outputs:
- **rt_nested_splits** (Tensor) - The nested splits of the return `RaggedTensor`,
and type of the tensor is `Tsplits`,
shape of the tensor is equal to shape of input `starts` plus 1.
and type of the tensor is `Tsplits`,
shape of the tensor is equal to shape of input `starts` plus 1.
- **rt_dense_values** (Tensor) - The dense values of the return `RaggedTensor`,
and type of the tensor should be same as input `starts`.
Let size of input `starts`, input `limits` and input `deltas` are i,
if type of the input `starts`, input `limits` and input `deltas`
are int32 or int64, shape of the output `rt_dense_values` is equal to
sum(abs(limits[i] - starts[i]) + abs(deltas[i]) - 1) / abs(deltas[i])),
if type of the input `starts`, input `limits` and input `deltas`
are float32 or float64, shape of the output `rt_dense_values` is equal to
sum(ceil(abs((limits[i] - starts[i]) / deltas[i]))).
and type of the tensor should be same as input `starts`.
Let size of input `starts`, input `limits` and input `deltas` are i,
if type of the input `starts`, input `limits` and input `deltas`
are int32 or int64, shape of the output `rt_dense_values` is equal to
sum(abs(limits[i] - starts[i]) + abs(deltas[i]) - 1) / abs(deltas[i])),
if type of the input `starts`, input `limits` and input `deltas`
are float32 or float64, shape of the output `rt_dense_values` is equal to
sum(ceil(abs((limits[i] - starts[i]) / deltas[i]))).
Raises:
TypeError: If any input is not Tensor.
TypeError: If the type of `starts` is not one of the following dtype: int32, int64, float32, float64.

22
mindspore/python/mindspore/ops/operations/nn_ops.py
View File

@ -695,12 +695,12 @@ class Mish(PrimitiveWithInfer):
Outputs:
Tensor, with the same type and shape as the `x`.
Supported Platforms:
``Ascend`` ``GPU`` ``CPU``
Raises:
TypeError: If dtype of `x` is neither float16 nor float32.
Supported Platforms:
``Ascend`` ``GPU`` ``CPU``
Examples:
>>> x = Tensor(np.array([[-1.0, 4.0, -8.0], [2.0, -5.0, 9.0]]), mindspore.float32)
>>> mish = ops.Mish()
@ -743,12 +743,12 @@ class SeLU(Primitive):
Outputs:
Tensor, with the same type and shape as the `input_x`.
Supported Platforms:
``Ascend`` ``GPU`` ``CPU``
Raises:
TypeError: If dtype of `input_x` is not int8, int32, float16, float32, float64.
Supported Platforms:
``Ascend`` ``GPU`` ``CPU``
Examples:
>>> from mindspore.ops.operations.nn_ops import SeLU
>>> input_x = Tensor(np.array([[-1.0, 4.0, -8.0], [2.0, -5.0, 9.0]]), mindspore.float32)
@ -7104,7 +7104,7 @@ class Dropout2D(PrimitiveWithInfer):
Dropout2D can improve the independence between channel feature maps.
Note:
The keep probability :math:`keep\_prob` is equal to 'ops.dropout2d' input '1-p'.
The keep probability :math:`keep\_prob` is equal to :math:`1 - p` in :func:`mindspore.ops.dropout2d`.
Refer to :func:`mindspore.ops.dropout2d` for more detail.
@ -7138,7 +7138,7 @@ class Dropout3D(PrimitiveWithInfer):
Note:
The keep probability :math:`keep\_prob` is equal to 'ops.dropout3d' input '1-p'.
The keep probability :math:`keep\_prob` is equal to :math:`1 - p` in :func:`mindspore.ops.dropout2d`.
Refer to :func:`mindspore.ops.dropout3d` for more detail.
@ -8571,9 +8571,6 @@ class Conv3DTranspose(Primitive):
Tensor of shape :math:`(N, C_{out}//group, D_{out}, H_{out}, W_{out})`,
where :math:`group` is the Args parameter.
Supported Platforms:
``Ascend`` ``GPU``
Raises:
TypeError: If `in_channel`, `out_channel` or `group` is not an int.
TypeError: If `kernel_size`, `stride`, `pad` , `dilation` or `output_padding` is neither an int not a tuple.
@ -8586,6 +8583,9 @@ class Conv3DTranspose(Primitive):
TypeError: If data type of dout and weight is not float16.
ValueError: If bias is not none. The rank of dout and weight is not 5.
Supported Platforms:
``Ascend`` ``GPU``
Examples:
>>> dout = Tensor(np.ones([32, 16, 10, 32, 32]), mindspore.float16)
>>> weight = Tensor(np.ones([16, 3, 4, 6, 2]), mindspore.float16)

6
mindspore/python/mindspore/ops/operations/random_ops.py
View File

@ -566,15 +566,15 @@ class UniformInt(Primitive):
- **maxval** (Tensor) - The distribution parameter, b.
It defines the maximum possibly generated value, with int32 data type. Only one number is supported.
Outputs:
Tensor. The shape is the same as the input 'shape', and the data type is int32.
Raises:
TypeError: If neither `seed` nor `seed2` is an int.
TypeError: If `shape` is not a tuple.
TypeError: If neither `minval` nor `maxval` is a Tensor.
ValueError: If `shape` is not a constant value.
Outputs:
Tensor. The shape is the same as the input 'shape', and the data type is int32.
Supported Platforms:
``Ascend`` ``GPU`` ``CPU``