jiangshuqiang
b21a98b062
fix function description for SummaryCollector and SummaryRecord
5 years ago
mindspore-ci-bot
13844d85cb
!12073 modify asyn ckpt
From: @changzherui
Reviewed-by: @kingxian,@zhoufeng54
Signed-off-by: @kingxian
5 years ago
changzherui
3b87921ad0
modify asyn ckpt
5 years ago
ougongchang
3240b2d8e1
Add device id to summary file name
To prevent write data conflicts in multi-card scenarios, the file on each card is increased by device_id
5 years ago
luopengting
7f1a5448dc
support recording dict as string for dataset graph
5 years ago
ougongchang
c6e4b0c85f
Add more log when collect graph and use summary operators
Fix can not collect input data when batch size is 1 and total step
number is 1
Fixed spelling errors
5 years ago
jiangshuqiang
7d79376bec
remove max_file_size limitation for export files
5 years ago
jiangshuqiang
ab5cc10250
add the tensor collection feature when record summary
5 years ago
mindspore-ci-bot
1f06cd63f3
!10436 Support to control whether to throw runtime exceptions in SummaryRecord
From: @ouwenchang
Reviewed-by:
Signed-off-by:
5 years ago
ougongchang
06be546b52
Support to control whether raise RuntimeError exception in SummaryRecord
1. Support explainer raise an RuntimeError exception
2. fix the ut of SummaryRecord
5 years ago
caozhou
450f149ff4
fix remove meta in pynative mode error
5 years ago
mindspore-ci-bot
8bd048cb0f
!10604 modify save checkpoint file
From: @changzherui
Reviewed-by: @zhunaipan,@kingxian
Signed-off-by: @kingxian
5 years ago
caozhou
08a13ab71e
delete meta
5 years ago
changzherui
0b7464eece
modify save ckpt
5 years ago
wqx
a84b779d2f
实现LearningRateScheduler
5 years ago
ougongchang
bc23af20d6
fix the docstring of SummaryCollector and SummaryRecord
5 years ago
changzherui
1c5182afb4
modify timemonitor and ckpt info
5 years ago
caozhou
e774e91620
modify print
5 years ago
changzherui
ffeacf13e4
add exoirt air test
5 years ago
ougongchang
e3596bbd42
SummaryCollector is not supported to record data on Ascend device in
dataset sink mode
5 years ago
mindspore-ci-bot
74886f790d
!8085 Loss saved does not reduce precision in SummaryCollector
Merge pull request !8085 from ougongchang/fix_round
5 years ago
ougongchang
87bf9fbb40
Loss saved does not reduce precision in SummaryCollector
5 years ago
ougongchang
57b8603bd3
In data sink mode, remind the user not to record input data
5 years ago
mindspore-ci-bot
b9180a9dab
!7910 ignore the value in dataset graph if its type is an object, log messag, log message about image only if the dataset_sink_mode is False
Merge pull request !7910 from luopengting/summary
5 years ago
luopengting
e410462a4b
ignore the value in dataset graph if its type is an object, log message about image only if the dataset_sink_mode is False
5 years ago
caozhou
4eab76d211
fix comment problem
5 years ago
luopengting
1be752f098
collect custom lineage data in optimizer auto
5 years ago
mindspore-ci-bot
2af6313f53
!7534 SummaryRecord support to record mindexplain data
Merge pull request !7534 from ougongchang/feature_mindexplain
5 years ago
ougongchang
6072b25a07
SummaryRecord support to record mindexplain data
The SummaryRecord.add_value() method is extended to record the data of
MindExplain.
5 years ago
caozhou
120f6a372a
add api example
5 years ago
caozhou
76ab430d4f
custom ckpt save and load
5 years ago
chenzomi
d471d32e87
[ME] change `check_integer` to format `check_positive_int` and `check_integeter`
5 years ago
chenzomi
d4e8e94981
[ME] delete check_bool and replace with Validate.check_bool
5 years ago
mindspore-ci-bot
7f390467e9
!6781 Change prefix for server ckpt callback
Merge pull request !6781 from ZPaC/master-change-prefix-for-server-ckpt
5 years ago
ZPaC
28c57f3f29
Change prefix for server ckpt callback
5 years ago
caozhou
5221041490
fix int in ckpt name changed to float bug
5 years ago
李鸿章
548b931f9d
flush summary when appropriate
5 years ago
ougongchang
e93365c664
Add a note for summary only supports linux systems
5 years ago
nhussain
3bac9d3713
switch input columns and operation
change ImagefolderDV2 name
change ds.transforms.vision to ds.vision
change batch api to match map api more closely
compose op changes
test_pylint
remove compose op from vision, move to transform module, refactor map and batch to use column_order
5 years ago
ZPaC
87bf2a7dcd
Add PS context.
5 years ago
Li Hongzhang
066950f69e
GPU dataset_sink_mode collect inputs
5 years ago
mindspore-ci-bot
4ec343961e
!5482 modify save_checkpoint
Merge pull request !5482 from liuyang/md_save_checkpoint
5 years ago
liuyang_655
4683de3443
modify save_checkpoint
5 years ago
Li Hongzhang
f95d3f21fb
fix assertion: Tensor(0) is falsy
5 years ago
mindspore-ci-bot
5b738794d2
!5389 Copy the default specified data when collect_specified data is None
Merge pull request !5389 from ougongchang/fix_summarycollector
5 years ago
Li Hongzhang
9050f2ad64
forkserver multiprocessing context
5 years ago
ougongchang
458c69a22c
Copy the default specified data when collect_specified data is None
5 years ago
ZPaC
830172201a
Fix multi server precision error.
5 years ago
wanyiming
3d354d76fd
mod_callback
5 years ago
Li Hongzhang
de43c11e2e
fix several issues
- 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
5 years ago