Col_lin
/
mindspore2022
forked from huawei/mindspore2022
0
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

!32008 Fix format in dump API 

Merge pull request !32008 from maning202007/code_docs_r1.7
This commit is contained in:
i-robot 2022-03-29 03:15:53 +00:00 committed by Gitee
parent 2e80383c8e a8de0a73e2
commit b5dcfe7b2d
No known key found for this signature in database
GPG Key ID: 173E9B9CA92EEF8F
8 changed files with 35 additions and 33 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
docs/api/api_python/mindspore/mindspore.set_dump.rst
View File

@ -3,9 +3,9 @@ mindspore.set_dump
.. py:class:: mindspore.set_dump(target, enabled=True)
启用或者禁用target及其子节点的Dump数据功能。
启用或者禁用 `target` 及其子节点的Dump数据功能。
target为 `Cell <https://www.mindspore.cn/docs/api/zh-CN/master/api_python/nn/mindspore.nn.Cell.html#mindspore.nn.Cell>`_ 或 `Primitive <https://www.mindspore.cn/docs/api/zh-CN/master/api_python/ops/mindspore.ops.Primitive.html#mindspore.ops.Primitive>`_ 的实例。请注意，此API仅在开启异步Dump功能且Dump配置文件中的 `dump_mode` 字段为"2"时生效。有关详细信息，请参阅 `Dump功能文档 <https://mindspore.cn/docs/programming_guide/zh-CN/master/dump_in_graph_mode.html>`_ 。默认状态下，Cell和Primitive实例不使能Dump数据功能。
`target` 为 :class:`mindspore.nn.Cell` 或 :class:`mindspore.ops.Primitive` 的实例。请注意，此API仅在开启异步Dump功能且Dump配置文件中的 `dump_mode` 字段为"2"时生效。有关详细信息，请参阅 `Dump功能文档 <https://mindspore.cn/docs/programming_guide/zh-CN/master/dump_in_graph_mode.html>`_ 。默认状态下， :class:`mindspore.nn.Cell` 和 :class:`mindspore.ops.Primitive` 实例不使能Dump数据功能。
.. Warning::
此类中的所有API均为实验版本，将来可能更改或者删除。
@ -13,10 +13,10 @@ mindspore.set_dump
.. Note::
- 此API只在Ascend后端的图模式有效。
- 此API只支持训练开始前调用。如果在训练过程中调用这个API，可能不会有效果。
- 使用set_dump(Cell, True)后，Cell正向计算和反向计算（梯度运算产生的计算）中的算子会被Dump。
- 对于 `nn.SoftMaxCrossEntropyWithLogits 层 <https://www.mindspore.cn/docs/api/zh-CN/master/api_python/nn/mindspore.nn.SoftmaxCrossEntropyWithLogits.html#mindspore.nn.SoftmaxCrossEntropyWithLogits>`_ ，正向计算和反向计算使用同一组算子。因此，只能看到反向计算中的Dump数据。请注意，当使用sparse=True和reduce=“mean”初始化时，nn.SoftmaxCrossEntropyWithLogits层也将在内部使用这些算子。
- 使用 `set_dump(Cell, True)` 后，Cell正向计算和反向计算（梯度运算产生的计算）中的算子会被Dump。
- 对于 :class:`mindspore.nn.SoftMaxCrossEntropyWithLogits` 层，正向计算和反向计算使用同一组算子。因此，只能看到反向计算中的Dump数据。请注意，当使用 `sparse=True` 和 `reduce=“mean”` 初始化时， :class:`mindspore.nn.SoftmaxCrossEntropyWithLogits` 层也将在内部使用这些算子。
**参数：**
- **target** (Union[Cell, Primitive]) - 要设置Dump标志的Cell或Primitive的实例。
- **enabled** (bool) - True表示启用Dump，False表示禁用Dump，默认值: True。
- **enabled** (bool，可选) - True表示启用Dump，False表示禁用Dump，默认值：True。

1
docs/api/api_python/train/mindspore.train.callback.SummaryCollector.rst
View File

@ -6,6 +6,7 @@
SummaryCollector还可以允许通过 `summary算子 <https://www.mindspore.cn/mindinsight/docs/zh-CN/master/summary_record.html#summarysummarycollector>`_ 将数据收集到summary文件中。
.. note::
- 使用SummaryCollector时，需要将代码放置到 `if __name__ == "__main__"` 中运行。
- 不允许在回调列表中存在多个SummaryCollector实例。
- 并非所有信息都可以在训练阶段或评估阶段收集。
- SummaryCollector始终记录summary算子收集的数据。

1
docs/api/api_python/train/mindspore.train.callback.SummaryLandscape.rst
View File

@ -3,6 +3,7 @@
SummaryLandscape可以帮助您收集loss地形图的信息。通过计算loss，可以在PCA（Principal Component Analysis）方向或者随机方向创建地形图。
.. note::
- 使用SummaryLandscape时，需要将代码放置到 `if __name__ == "__main__"` 中运行。
- SummaryLandscape仅支持Linux系统。
**参数：**

3
docs/api/api_python/train/mindspore.train.summary.SummaryRecord.rst
View File

@ -7,6 +7,7 @@
它通过执行 `record` 方法将数据写入文件。除了通过 `summary算子 <https://www.mindspore.cn/mindinsight/docs/zh-CN/master/summary_record.html#summarysummarycollector>`_ 记录网络的数据外，SummaryRecord还支持通过 `自定义回调函数和自定义训练循环 <https://www.mindspore.cn/mindinsight/docs/zh-CN/master/summary_record.html#callback>`_ 记录数据。
.. note::
- 使用SummaryRecord时，需要将代码放置到 `if __name__ == "__main__"` 中运行。
- 确保在最后关闭SummaryRecord，否则进程不会退出。请参阅下面的示例部分，了解如何用两种方式正确关闭SummaryRecord。
- 每次训练只允许创建一个SummaryRecord实例，否则会导致数据写入异常。
- SummaryRecord仅支持Linux系统。
@ -93,7 +94,7 @@
- **step** (int) - 表示当前的step。
- **train_network** (Cell) - 表示用于保存计算图的训练网络。默认值：None，表示当原始网络的图为None时，不保存计算图。
- **plugin_filter** (Callable[[str], bool]) - 过滤器函数，用于过滤需要写入的标签项。默认值：None。
- **plugin_filter** (Callable[[str], bool], 可选) - 过滤器函数，用于过滤需要写入的标签项。默认值：None。
**返回：**

28
mindspore/python/mindspore/common/dump.py
View File

@ -21,14 +21,13 @@ from mindspore._c_expression import security
def set_dump(target, enabled=True):
"""
Enable or disable dump for the target and its contents.
Enable or disable dump for the `target` and its contents.
Target should be an instance of `Cell <https://www.mindspore.cn/docs/api/en/master/api_python/nn/
mindspore.nn.Cell.html#mindspore.nn.Cell>`_ or `Primitive <https://www.mindspore.cn/docs/api/en/master/
api_python/ops/mindspore.ops.Primitive.html#mindspore.ops.Primitive>`_ . Please note that this API takes
effect only when Asynchronous Dump is enabled and the dump_mode field in dump config file is 2. See the
`dump document <https://mindspore.cn/docs/programming_guide/en/master/dump_in_graph_mode.html>`_
for details. The default enabled status for a cell or primitive is False.
`target` should be an instance of :class:`mindspore.nn.Cell` or :class:`mindspore.ops.Primitive` .
Please note that this API takes effect only when Asynchronous Dump is enabled and the `dump_mode`
field in dump config file is "2". See the `dump document <https://mindspore.cn/docs/programming_guide/
en/master/dump_in_graph_mode.html>`_ for details.
The default enabled status for a :class:`mindspore.nn.Cell` or :class:`mindspore.ops.Primitive` is False.
.. warning::
This is an experimental prototype that is subject to change or deletion.
@ -37,23 +36,20 @@ def set_dump(target, enabled=True):
1. This API is only effective for GRAPH_MODE with Ascend backend.
2. This API only supports being called before training starts.
If you call this API during training, it may not be effective.
3. After using set_dump(cell, True), operators in forward and backward
3. After using `set_dump(Cell, True)` , operators in forward and backward
computation (computation generated by the grad operations) of the
cell will be dumped.
4. For `nn.SparseSoftmaxCrossEntropyWithLogits
<https://www.mindspore.cn/docs/api/en/master/api_python/nn/
mindspore.nn.SoftmaxCrossEntropyWithLogits.html#mindspore.nn
.SoftmaxCrossEntropyWithLogits>`_ layer, the forward
4. For :class:`mindspore.nn.SoftmaxCrossEntropyWithLogits` layer, the forward
computation and backward computation use the same set of
operators. So you can only see dump data from backward computation.
Please note that nn.SoftmaxCrossEntropyWithLogits layer will also use
the above operators internally when initialized with sparse=True and
reduction="mean".
Please note that :class:`mindspore.nn.SoftmaxCrossEntropyWithLogits` layer will also use
the above operators internally when initialized with `sparse=True` and
`reduction="mean"` .
Args:
target (Union[Cell, Primitive]): The Cell instance or Primitive instance
to which the dump flag is set.
enabled (bool): True means enable dump, False means disable dump.
enabled (bool, optional): True means enable dump, False means disable dump.
Default: True.
Supported Platforms:

3
mindspore/python/mindspore/train/callback/_landscape.py
View File

@ -169,7 +169,8 @@ class SummaryLandscape:
It can create landscape in PCA direction or random direction by calculating loss.
Note:
1. SummaryLandscape only supports Linux systems.
1. When using SummaryLandscape, you need to run the code in `if __name__ == "__main__"` .
2. SummaryLandscape only supports Linux systems.
Args:
summary_dir (str): The path of summary is used to save the model weight,

11
mindspore/python/mindspore/train/callback/_summary_collector.py
View File

@ -75,11 +75,12 @@ class SummaryCollector(Callback):
SummaryCollector also enables the summary operator to collect data to summary files.
Note:
1. Multiple SummaryCollector instances in callback list are not allowed.
2. Not all information is collected at the training phase or at the eval phase.
3. SummaryCollector always record the data collected by the summary operator.
4. SummaryCollector only supports Linux systems.
5. The Summary is not supported when compile source with `-s on` option.
1. When using SummaryCollector, you need to run the code in `if __name__ == "__main__"` .
2. Multiple SummaryCollector instances in callback list are not allowed.
3. Not all information is collected at the training phase or at the eval phase.
4. SummaryCollector always record the data collected by the summary operator.
5. SummaryCollector only supports Linux systems.
6. The Summary is not supported when compile source with `-s on` option.
Args:
summary_dir (str): The collected data will be persisted to this directory.

11
mindspore/python/mindspore/train/summary/summary_record.py
View File

@ -101,11 +101,12 @@ class SummaryRecord:
can be added by calling add_value.
Note:
1. Make sure to close the SummaryRecord at the end, otherwise the process will not exit.
1. When using SummaryRecord, you need to run the code in `if __name__ == "__main__"` .
2. Make sure to close the SummaryRecord at the end, otherwise the process will not exit.
Please see the Example section below to learn how to close properly in two ways.
2. Only one SummaryRecord instance is allowed at a time, otherwise it will cause data writing problems.
3. SummaryRecord only supports Linux systems.
4. The Summary is not supported when compile source with `-s on` option.
3. Only one SummaryRecord instance is allowed at a time, otherwise it will cause data writing problems.
4. SummaryRecord only supports Linux systems.
5. The Summary is not supported when compile source with `-s on` option.
Args:
log_dir (str): The log_dir is a directory location to save the summary.
@ -334,7 +335,7 @@ class SummaryRecord:
step (int): Represents training step number.
train_network (Cell): The spare network for saving graph.
Default: None, it means just do not save the graph summary when the original network graph is None.
plugin_filter (Optional[Callable[[str], bool]]): The filter function, \
plugin_filter (Callable[[str], bool], optional): The filter function, \
which is used to filter out which plugin should be written. Default: None.
Returns: