6e230e807a
op pad
5 years ago
e6c596a9d9
!3653 add epoch_num description master
...
Merge pull request !3653 from panfengfeng/add_epoch_num_description_master
5 years ago
34864fbc56
fix weight init and add data aug
5 years ago
17d3982d46
fix scatter error msg
5 years ago
6ea2aa4e73
!3672 fix serving input numbers
...
Merge pull request !3672 from hexia/fix_input_check
5 years ago
76733ce816
fix cpu multi graph mem error
5 years ago
389cb35740
!3661 Alarm modification
...
Merge pull request !3661 from shenwei41/sw_master
5 years ago
8f35d2ed29
!3664 Modify the order of init and open of TDT
...
Merge pull request !3664 from hanjun996/master
5 years ago
b73ea6a7aa
!3668 Modify collecting graph and dataset graph to step end stage
...
Merge pull request !3668 from ougongchang/fix_collect_dataset
5 years ago
567509affc
!3522 add tinybert scripts
...
Merge pull request !3522 from wanghua/c75_tinybert
5 years ago
abbd7b50db
optimize the vgg script
5 years ago
4307c1fa61
change the column order and add drop_reminder option to make this script compatible with BertCLS model
5 years ago
a2ffc9530e
stuff added
...
box copying fix
common function for IOU
new sort func
final update + unit test
remove one comment
fix lintig -1
lint fix 2
lint fix 3
last lint fix
value fix in ST nms with mask
addressing comments
pylint fix
pylint fix 1
test file fix
5 years ago
e94d91ba95
remove import probability from nn/__init__.py
5 years ago
269b477684
use np.testing.assert instead of asserting
...
fix line lengths
fix spacing
reverse a few changes
5 years ago
6f70146153
!3660 modify readme for maskrcnn
...
Merge pull request !3660 from meixiaowei/master
5 years ago
d66e6b33bf
!3665 support multy node training in deeplabv3
...
Merge pull request !3665 from zhouyaqiang0/master
5 years ago
afce1c3a40
!3341 GPU maxpool with argmax op
...
Merge pull request !3341 from tom_chen/maxpool_with_argmax
5 years ago
98e2ee90de
fix optimizer parallel problems
5 years ago
1dafb2c6f5
Modify collecting graph and dataset graph to step end stage
...
We collect graph and dataset graph in begin stage before,
If there compile graph fail in GPU, we also collect graph
and dataset graph to summary dir, it will confuse user.
So we collect graph and dataset graph in step end stage now,
If there compile graph fail, we will not collect graph and dataset
graph.
5 years ago
051c290d8b
Modify patches and alerts
5 years ago
00f7a936bf
add resnet50 support multi node training
5 years ago
b0004a1791
support multy node training and remove code
5 years ago
387dac5832
!3651 change num_samples definition
...
Merge pull request !3651 from jiangzhiwen/dataset/change_num_samples
5 years ago
20ccf83826
modify tdt
5 years ago
a3e7c4c754
!3625 Optimize tensor data
...
Merge pull request !3625 from hewei/optimize_tensor_data
5 years ago
fe514bd1cc
!3644 [MD] fix minddataset core dump when file list size ia greater than 1000.
...
Merge pull request !3644 from liyong126/fix_mindrecord_bug
5 years ago
3100824703
fix input
5 years ago
a337a02732
!3638 fix codex and support akg op profiling
...
Merge pull request !3638 from geekun/yjk_master
5 years ago
9aa84b3d14
add strided slice op
5 years ago
b39d524d44
set out format to nhwc4
5 years ago
8950952fe3
modify readme
5 years ago
af39ca8252
modify the wrong word
5 years ago
9cad0fec1d
gpu broadcast to
5 years ago
1b69923472
!3643 Throw exception if different communication ops which are divided to the same segement share the same input
...
Merge pull request !3643 from huanghui/communication-op-fusion
5 years ago
d4b52ac59f
!3489 use kernelruntime::mem_manager to reduce rtMalloc and rtFree time in trans data format
...
Merge pull request !3489 from lvchangquan/master
5 years ago
4644085e92
add epoch_num
5 years ago
7dd5e78fde
add tinybert scripts
5 years ago
1eda0ef071
change num_samples definition
5 years ago
fcdad59ce6
!3594 fix batchnorm issue under mix precision in pynative mode
...
Merge pull request !3594 from wangqiuliang/fix-batchnorm-under-mix-precision-in-pynative
5 years ago
2309e7369a
add_python_distribute_pretrain_script
...
Signed-off-by: GuoMengHao <guomenghao@huawei.com>
5 years ago
12a150bb5d
!3630 not reuse refnode input's memory
...
Merge pull request !3630 from laiyongqiang/refnode_input_fix
5 years ago
c57ad1528f
!3635 fix dataset & train gil lock of gpu process master
...
Merge pull request !3635 from panfengfeng/fix_dataset_train_gil_of_gpu_master
5 years ago
44e739ae31
!3627 fix: device occupied tdt hung
...
Merge pull request !3627 from guozhijian/fix_device_occupied_tdt_hung
5 years ago
17d71280b8
fix codex and support akg op profiling
5 years ago
9ccc6889eb
!3624 fix GeneratorDataset time out
...
Merge pull request !3624 from yanghaitao/yht_generator_timeout
5 years ago
4834a6b347
!3574 Rename AnfNode::user_data related functions to follow naming rule
...
Merge pull request !3574 from hewei/rename_user_data_func
5 years ago
ed70de8070
fix coredump when number of file list more than 1000.
5 years ago
311d8ea1f9
add exception when different communication op in one segment shared the same input
5 years ago
e4a7ca7f08
!3637 Lowering value checking threshold to support training with very small steps or
...
Merge pull request !3637 from thlinh/dev_Jul28_lower_checking_threshold
5 years ago
ling
mindspore-ci-bot
zhouyaqiang
fangzehua
kswang
ms_yan
dessyang
danish
Xun Deng
tony_liu2
Ziyan
ougongchang
shenwei41
gengdongjie
hanjun996
hexia
yangzhenzhang
wandongdong
meixiaowei
limingqi107
wilfChen
panfengfeng
wanghua
jiangzhiwen
GuoMengHao
geekun
liyong
huanghui