i-robot
|
de95400c52
|
!30587 check platform for resize bilinear
Merge pull request !30587 from yangzhenzhang/check-platform-for-resize-bilinear
|
4 years ago |
zhoufeng
|
f49b195c39
|
extract common as an independent shared library
Signed-off-by: zhoufeng <zhoufeng54@huawei.com>
|
4 years ago |
yangzhenzhang
|
43e6e16da3
|
check platform for resizebilinear
|
4 years ago |
wangjun
|
24d448239c
|
add pynative_parallel
|
4 years ago |
i-robot
|
2e8eac8341
|
!30367 auto_parallel_adasum_support_data_parallel
Merge pull request !30367 from yao_yf/auto_parallel_adasum_support_data_parallel
|
4 years ago |
i-robot
|
0f24b679ec
|
!29819 Add GlobalNorm Search
Merge pull request !29819 from huangxinjing/add_global_norm_search
|
4 years ago |
yao_yf
|
19236b1a70
|
auto parallel adasum support data parallel and hybrid parallel
|
4 years ago |
liuluobin
|
9d6a6b7959
|
Code Alarm clearance for parallel modules.
|
4 years ago |
huangxinjing
|
092ba035e3
|
Add global norm parallel support
|
4 years ago |
huoxinyou
|
eb04e1102b
|
optimize syncbatchnorm
remove format in syncbatchnorm
|
4 years ago |
i-robot
|
dd593d2b87
|
!30157 auto_parallel_adasum_cpp_part
Merge pull request !30157 from yao_yf/auto_parallel_adasum_cpp_part
|
4 years ago |
yao_yf
|
20f642a014
|
auto parallel support adasum cpp part
|
4 years ago |
liuluobin
|
b0b79ef8fb
|
Add backend check for RandomChoiceWithMask
|
4 years ago |
i-robot
|
ddafecfa81
|
!30108 Add format attribution for SyncBatchNorm operator
Merge pull request !30108 from Bert0108/sync_bn_format
|
4 years ago |
Bert0108
|
974cc9466a
|
add format for syncbatchnorm in semi parallel
|
4 years ago |
liuluobin
|
80fd640819
|
Implementation of ROIAlignInfo and CropAndResizeInfo
|
4 years ago |
yangzhenzhang
|
4061c11e18
|
support pad mode for conv2d transpose
|
4 years ago |
i-robot
|
a55b1b5e05
|
!29453 support pad mode for conv2d parallel operator
Merge pull request !29453 from yangzhenzhang/support-pad-mode-for-conv2d
|
4 years ago |
i-robot
|
b0b3d7ede7
|
!29421 Implementation of BoundingBoxEncodeInfo, IOUInfo and RandomChoickWithMaskInfo
Merge pull request !29421 from liuluobin/parallel_ops_master
|
4 years ago |
i-robot
|
55ba926a04
|
!29087 optimize sharding propagation&add ut on pangu
Merge pull request !29087 from bichaoyang/master
|
4 years ago |
yangzhenzhang
|
6df5e9aa5d
|
support pad mode for conv2d
|
4 years ago |
liuluobin
|
fb79c508f2
|
Implementation of BoundingBoxEncodeInfo, IOUInfo, RandomChoiceWithMaskInfo.
|
4 years ago |
b00518648
|
2ff3425c76
|
1.consider comm cost inside a op when select a strategy;2.deal with the ops that share same param;3.add ut to fix the perfermence at pangu_alpha
|
4 years ago |
yangzhenzhang
|
4d0b1a47ee
|
support config group for batchnorm
|
4 years ago |
yangzhenzhang
|
000c5b5394
|
support dilation for conv2d
|
4 years ago |
yangzhenzhang
|
5514189257
|
support group for conv2d
|
4 years ago |
i-robot
|
ced581ce7b
|
!29181 fix stride slice parallel mask
Merge pull request !29181 from fangzehua/fix_strided_mask_master
|
4 years ago |
i-robot
|
dcf3095302
|
!28837 fix a bug of onehot
Merge pull request !28837 from bichaoyang/master_1
|
4 years ago |
fangzehua
|
18fe447fb9
|
fix stridedslice mask in parallel
|
4 years ago |
yangzhenzhang
|
6dd7333c0b
|
fix bug for conv2d transpose
|
4 years ago |
i-robot
|
888364be7b
|
!28996 Fix bugs in edge_costmodel.
Merge pull request !28996 from liuluobin/master_codex
|
4 years ago |
liuluobin
|
9761d70bfd
|
Fix bugs in edge_costmodel.
|
4 years ago |
yangzhenzhang
|
a9d2e255f5
|
support single direction exchange for conv2d
|
4 years ago |
b00518648
|
4ac7e1c34a
|
fix a bug of onehot
|
4 years ago |
i-robot
|
7d65ce6b3b
|
!28642 split h and w dimension for conv2d
Merge pull request !28642 from yangzhenzhang/compute-top-bottom-overlap-for-conv2d
|
4 years ago |
yangzhenzhang
|
e5df74e9e4
|
compute top bottom overlap for conv2d
|
4 years ago |
liuluobin
|
7b597d2361
|
Clearing codecheck for parallel
|
4 years ago |
lichenever
|
78839a2413
|
parallel_support_make_tuple
|
4 years ago |
i-robot
|
ec485cb436
|
!28307 convert EnvSetItem/EnvGetItem to primitive EnvironSet/EnvironGet
Merge pull request !28307 from xychow/EnvGetItem-conversion
|
4 years ago |
zhousiyi
|
19e9571149
|
rename env_getitem/setitem to EnvironGet/EnvironGet
|
4 years ago |
i-robot
|
03b4829b86
|
!28431 [Auto parallel] MoE Changes
Merge pull request !28431 from Xiaoda/122-moe-changes
|
4 years ago |
Xiaoda Zhang
|
6d8320fa66
|
1) fix the exact division in moe;
2) changing CumSum from composition to a single Operator;
3) add InferMirrorOps for CumSumInfo.
|
4 years ago |
yao_yf
|
b5d56af0ad
|
recompute genmask
|
4 years ago |
Xiaoda Zhang
|
66c7474e5a
|
remove CastInfo from CNODE
|
4 years ago |
yangzhenzhang
|
8a68577756
|
use neighbor-exchange-v2 for conv2d
|
4 years ago |
yangzhenzhang
|
5f6477b022
|
add output strategy for gather op
|
4 years ago |
i-robot
|
bf0fce0ebe
|
!27432 sharding propatation for pangu
Merge pull request !27432 from bichaoyang/master
|
4 years ago |
i-robot
|
dea78627c6
|
!27480 Bugfix: cell level op suport double-recursive algo
Merge pull request !27480 from zhuyuxiao/I4LE27
|
4 years ago |
b00518648
|
90bd01904a
|
1.adding compare function for strategy_map;
2.add FAST_GELU to step_auto_parallel.cc
3.remove the AUTO_PARALLEL constraint in transformer
|
4 years ago |
yao_yf
|
30576c6a75
|
fix reshape bool type in auto parallel
|
4 years ago |