From: @wudenggang
Reviewed-by: @ouwenchang,@Hanshize,@kingxian
Signed-off-by: @kingxian

From: @jiang-shuqiang
Reviewed-by: 
Signed-off-by:

If device target is GPU, the device number may be set to 1.

Note: When using sys.exit method, it is found that when the main process is killed, the summary process enters a sleep state and cannot execute the exit operation

To prevent write data conflicts in multi-card scenarios, the file on each card is increased by device_id

Fix can not collect input data when batch size is 1 and total step
number is 1

Fixed spelling errors

From: @ouwenchang
Reviewed-by: 
Signed-off-by:

1. Support explainer raise an RuntimeError exception
2. fix the ut of SummaryRecord

Environment variables are used to specify a maximum number of OpenBLAS threads.

In ubuntu(GPU) environment, numpy will use too many threads for computing,
it may affect the start of the summary process.
Notice: At present, the performance of setting the thread to 2 has been tested to be more suitable.
If it is to be adjusted, it is recommended to test according to the scenario first

a

bugfix

undo unexecpted changes

modify summary format

bugfix

fix CI problem

enhance comment

fix comment typo

use mindspore.log for errors

fix pylint issues

The SummaryRecord.add_value() method is extended to record the data of
MindExplain.

- handle collection for multiple trains
- how many tensors to collect when sunk
- change loglevel for get_learning_rate
- update calculation of `max_file_size`
- fix how collect_tensor_freq counting

- fix writer pool not exit when max_file_size too small
- fix API docs to illustrate `collect_tensor_freq` and `max_file_size`

Add more detail note for SummaryCollector and SummaryRecord,
else if it is used not right, some proplem will be caused.

Merge pull request !2770 from LiHongzhang/capture_time

I added a SummaryCollector to help users automatically collect information
such as the network, loss, learning rate and so on, making it easier to collect this information.

It also can collect train lineage and eval lineage information which is
collected by TrainLineage Callback and EvalLineage Callback in
MindInsight.

I also add some UT for SummaryCollect to keep the code correct.

Merge pull request !2193 from wenkai/wk1_log_level_0617

1. When initing SummaryRecord, if check failed, self._close is not set,
   which leads to 'SummaryRecord' object has no attribute '_closed'
   AttributeError later on.

2. There may be too many processes for handling summary adapting.
   See issue #I1K6K7