0105b2166e
!10370 Add BatchMatMul for CPU
...
From: @xukailun_1
Reviewed-by:
Signed-off-by:
4 years ago
d9830ddf63
add c.dot opt
4 years ago
9bb4785777
!10428 unpack of cpu
...
From: @lishixing3
Reviewed-by:
Signed-off-by:
4 years ago
79893d6fef
!10336 Add Pack op for CPU Kernel
...
From: @zyx5256
Reviewed-by:
Signed-off-by:
4 years ago
4257288bb7
!10066 Add CumSum Kerner for CPU
...
From: @shaoxiangdong
Reviewed-by:
Signed-off-by:
4 years ago
9366200dba
!9646 Add Expm1 Operator for CPU
...
From: @xukailun_1
Reviewed-by:
Signed-off-by:
4 years ago
ef13ff4bef
!10613 add ut for primitve c of conv2d
...
From: @lianliguang
Reviewed-by: @kisnwang,@zhoufeng54
Signed-off-by: @zhoufeng54
4 years ago
69f882c083
batchmatmul
4 years ago
aecd96cdb2
!10424 Compute weighted loss
...
From: @xuguoyang5566
Reviewed-by: @wuxuejian
Signed-off-by:
4 years ago
b67aaf6773
!9832 expose_allgather_fusion_to_users
...
From: @gong_zi_yan
Reviewed-by:
Signed-off-by:
4 years ago
728fac6c9f
Expand J for innermost graph first when the graph also contains J primitive
4 years ago
a5c16ba5c4
update tests/st/networks/models/resnet50/src/dataset.py.
4 years ago
bbf8ec82b9
expose allgather fusion interface to users
4 years ago
0c42c5b979
!10611 support len of a Tensor
...
From: @zhangbuxue
Reviewed-by: @zh_qh,@zhunaipan
Signed-off-by: @zhunaipan
4 years ago
59f7605987
adjust cost threshold for bert ci_smoke
4 years ago
1cbef74372
!10592 deepfm network use mindrecord dataset
...
From: @shuzigood
Reviewed-by: @linqingke,@wuxuejian
Signed-off-by: @linqingke
4 years ago
cb909c7242
add ut for conv2d
4 years ago
1a7c161e73
support len of a Tensor
4 years ago
dafa401247
!10423 add testcases of dump
...
From: @rainyhorse
Reviewed-by: @chujinjin,@kisnwang
Signed-off-by: @kisnwang
4 years ago
35ff7c58f2
add pack for cpu kernel; bugfix adagrad gpu
4 years ago
677127ef5e
add grad definition for AsinGrad and ACosGrad
4 years ago
435a524f06
deepfm network use mindrecord dataset
4 years ago
3c6101a8b1
!10559 update L1Regularizer operation
...
From: @dangjiaqi1
Reviewed-by: @liangchenghui,@wuxuejian
Signed-off-by: @liangchenghui
4 years ago
ad814dc2ad
!10452 add cpu operators: Div/Mod/Floor/Minimum/ArgMinWithValue
...
From: @shibeiji
Reviewed-by: @liangchenghui
Signed-off-by:
4 years ago
e960830d92
!10536 [cpu] Add full support for cast OP on cpu
...
From: @yanglf1121
Reviewed-by: @wuxuejian,@zhunaipan
Signed-off-by: @wuxuejian
4 years ago
24345595f3
added collective send and receive
4 years ago
352a362878
Add new GPU operators: Asinh, AsinhGrad, Acosh, AcoshGrad, Atan and AtanGrad
4 years ago
7fe25d6444
Add weights loss compute
4 years ago
d043eeb834
add CPU operators: Div/Mod/Minimum/Floor/ArgMinWithValue
4 years ago
86bbef5146
uadate L1Regularizer,upadate test_l1_regularizer_op.py
4 years ago
8987b3dbd1
Add full support for cpu cast op
4 years ago
accf7c4671
Add CumSum Kerner for CPU
4 years ago
216aa8535b
!10412 Delete the redundant inputs of GNMTv2 model
...
From: @gaojing22
Reviewed-by: @c_34,@oacjiewen
Signed-off-by: @c_34
4 years ago
1f3d9ddff8
delete schema_json
4 years ago
4df0baa6c7
!10392 getitem fix bug
...
From: @yepei6
Reviewed-by: @zhunaipan,@kingxian,@zhunaipan
Signed-off-by: @kingxian
4 years ago
40cd715a4c
!10448 Fix warning information
...
From: @lizhenglong1992
Reviewed-by: @liucunwei,@jonyguo,@heleiwang
Signed-off-by: @liucunwei
4 years ago
221d5b67d2
!10442 change code in modelzoo to import dataset APIs from mindspore.dataset rather than mindspore.dataset.engine
...
From: @tiancixiao
Reviewed-by: @heleiwang,@liucunwei
Signed-off-by: @liucunwei
4 years ago
425c3c6abc
Fix warning information.
4 years ago
31fed1a2f6
change code to import APIs from mindspore.dataset rather than mindspore.dataset.engine
4 years ago
07578ca31e
!10398 Dvpp debug
...
From: @lizhenglong1992
Reviewed-by:
Signed-off-by:
4 years ago
261bb736de
!10063 Modify profiler directory structure to fit the new Run Package
...
From: @gzhcv
Reviewed-by:
Signed-off-by:
4 years ago
3ba3ffedd4
!10203 Synchronize latest Ascend software suite 18 Dec 2020
...
From: @nicholas_yhr
Reviewed-by:
Signed-off-by:
4 years ago
3bc94b97d7
fix some liness
4 years ago
b8345d03b6
Synchronize latest Ascend software 18 Dec 2020, with profiler fixes
4 years ago
1343c9cca2
!10255 using do_copy option to choose Tensor.from_numpy to do Tensor convert
...
From: @ms_yan
Reviewed-by:
Signed-off-by:
4 years ago
6e690ab725
!9949 add testcase and minor fix for SparseApplyProximalAdagrad/Ftrl
...
From: @TFbunny
Reviewed-by: @liangchenghui,@linqingke
Signed-off-by: @liangchenghui
4 years ago
c9b438e97b
add dump testcase
4 years ago
d032b69c2d
Update Error message for dvpp operator.
4 years ago
f9dbebd958
!10410 add dtype supports[fp64/fp32/fp16/int8/int16/int32/int64/uint8] for relu/reluv2/relugradv2
...
From: @yuan_shen_zhou
Reviewed-by: @wilfchen,@liangchenghui,@linqingke
Signed-off-by: @liangchenghui
4 years ago
e1248a7246
!10389 Fix the Backward of the AllSwap Operation
...
From: @huangxinjing
Reviewed-by: @stsuteng,@stsuteng,@zhunaipan
Signed-off-by: @stsuteng,@stsuteng
4 years ago
0a31f563a7
minor fixes
4 years ago
2e2c01d6f0
add dtype supports for relu/reluv2/relugradv2
4 years ago
4f78ee0077
fix getitem bug
4 years ago
deb1e6e965
use from_numpy and add do_copy option
4 years ago
44abd94b42
!9408 [ME][Pynative]Fix mem leak problem in pynative mode
...
From: @chenfei52
Reviewed-by:
Signed-off-by:
4 years ago
70dc40551a
!10020 Improve fancy index from 1st dimension to arbitrarily dimension
...
From: @yepei6
Reviewed-by:
Signed-off-by:
4 years ago
20df37e428
!10287 Add ut for dataset to_number
...
From: @luoyang42
Reviewed-by: @heleiwang
Signed-off-by:
4 years ago
565ce81b29
Fix allsawp
4 years ago
1f20e552ed
handle mem leak in pynative mode
...
use enter and leave construct
4 years ago
c1d4a0f80c
!10361 Introduce a master disable flag to ProfilIngManager
...
From: @ziruiwu
Reviewed-by:
Signed-off-by:
4 years ago
a1355a827c
introduce a master switch to turn on and off profiler
4 years ago
04392f3120
recovery-already-run
4 years ago
61ed05f133
!10102 test high order grad with respect to parameter
...
From: @chen_tanjie
Reviewed-by: @wang_zi_dong,@zh_qh
Signed-off-by: @zh_qh
4 years ago
176ea4676c
!10261 change the int32 restrict to int restric
...
From: @yepei6
Reviewed-by: @kingxian,@kingxian
Signed-off-by: @kingxian,@kingxian
4 years ago
32dc10a735
!9864 Adjust the location of cleaning unuse memory in value node
...
From: @joylvliang
Reviewed-by:
Signed-off-by:
4 years ago
b8b4ce7442
!10323 Rewrite gpu implementation for op maxpool_with_argmax/maxpool_with_argmax_grad
...
From: @yuan_shen_zhou
Reviewed-by: @linqingke,@liangchenghui
Signed-off-by: @linqingke
4 years ago
5c9982729d
fancy index getitem
4 years ago
9b3226c2b6
!10289 Ignore the graphs with defer_inline flag when specializing on graph arguments
...
From: @ginfung
Reviewed-by: @zh_qh
Signed-off-by: @zh_qh
4 years ago
236bfb75e3
change the int32 restrict to int
4 years ago
1490947ff0
adjust-the-location-of-cleaning-unuseless-memory-in-value-node
4 years ago
e1aa34a030
Ignore the graphs with defer_inline flag when specializing on graph arguments
4 years ago
00611b1c41
!10175 Add tensor.ndim and rename tensor.size() to tensor.size
...
From: @yanglf1121
Reviewed-by: @kingxian,@zhunaipan
Signed-off-by: @kingxian
4 years ago
09b68f9006
rewrite gpu implementation of op maxpool_with_argmax and maxpool_with_argmax_grad
4 years ago
f1e1a2afd6
!10223 fixed codex and change bare pointer to shared pointer
...
From: @anancds
Reviewed-by: @limingqi107
Signed-off-by:
4 years ago
2212a5d728
!10305 Fix core dump when num_worker is negative value
...
From: @ziruiwu
Reviewed-by: @mikef,@nsyca
Signed-off-by: @nsyca
4 years ago
c6f3b1826c
!10009 Add infer shape function for tile in graph kernel
...
From: @looop5
Reviewed-by: @gaoxiong1,@ckey_dou
Signed-off-by: @ckey_dou
4 years ago
2b64b7f295
!10312 SequenceMask move API to composite
...
From: @peilin-wang
Reviewed-by: @liangchenghui,@wuxuejian
Signed-off-by: @liangchenghui
4 years ago
412665f707
wrap sequenceMask in composite
4 years ago
299509dbfc
changed cast to round to zero if casting from float to integral
4 years ago
0e2f7a9e9e
add validate param to base class DatasetNode
4 years ago
da6de519ae
Add ut for dataset to_number
4 years ago
f8e5bffe4d
!10286 fix Wrong TypeError in numpy_native
...
From: @wangrao124
Reviewed-by: @zhunaipan,@kingxian
Signed-off-by: @kingxian
4 years ago
441e283e21
!10197 Add input check for the embedding
...
From: @huangxinjing
Reviewed-by: @stsuteng,@zhunaipan
Signed-off-by: @stsuteng
4 years ago
8bbe723603
add Tile infer shape function
4 years ago
2fb3f0ef4f
fix Wrong TypeError in numpy_native
4 years ago
ee23af58d8
fixed codex
4 years ago
4477b97465
!10167 BatchNormGrad infer gpu kernel
...
From: @jonwe
Reviewed-by: @liangchenghui
Signed-off-by:
4 years ago
2f3230e761
!10240 add restriction optimizer shard
...
From: @gong_zi_yan
Reviewed-by: @stsuteng,@zhunaipan
Signed-off-by: @stsuteng
4 years ago
2f783a6e25
fix process hang when training is over.
4 years ago
c5c905fdf5
add restriction for opt shard
4 years ago
cff3141d99
!10218 add CPU ops: Sign, NotEqual, Split, AvgPool, AvgPoolGradCpu
...
From: @caojian05
Reviewed-by:
Signed-off-by:
4 years ago
5143f7b99c
!10198 Add CPU OP resize_bilinear and resize neighbor
...
From: @wanyiming
Reviewed-by: @wuxuejian,@oacjiewen
Signed-off-by: @wuxuejian
4 years ago
88155de042
!10159 momentum weight fusion
...
From: @wilfchen
Reviewed-by: @cristoval,@limingqi107
Signed-off-by: @cristoval
4 years ago
a3b6f57842
add CPU ops: Sign, NotEqual, Split, AvgPool, AvgPoolGradCpu
4 years ago
5ff6de83eb
!10173 Fix Strategy Check for UnsortedSegments
...
From: @huangxinjing
Reviewed-by: @zhunaipan,@stsuteng
Signed-off-by: @stsuteng
4 years ago
e6febd85bb
!10200 Fix dir of test case
...
From: @luoyang42
Reviewed-by: @liucunwei,@heleiwang
Signed-off-by: @liucunwei
4 years ago
d99189f681
!10207 Fix some minor doc, typo related to AutoNumWorker
...
From: @ziruiwu
Reviewed-by: @mikef
Signed-off-by:
4 years ago
335c3f7602
BatchNormGrad infer gpu kernel
4 years ago
4b7e9975a0
!10090 [MS][GPU][CI Alarm] Fixed CI Alarm for Pad GPU kernel + code clean up
...
From: @danishnxt
Reviewed-by: @robingrosman
Signed-off-by: @robingrosman
4 years ago
4ae7102fd3
address 2 tickets
4 years ago
1c43115eef
!9540 [MS][API] - Update BroadcastTo to be able to support dynamic -1
...
From: @danishnxt
Reviewed-by: @robingrosman,@robingrosman
Signed-off-by: @robingrosman
4 years ago
2283eac723
add_resize_ops
4 years ago
a83e058f42
!9858 [MS][GPU][API] UnsortedScatterND_int64_Support
...
From: @danishnxt
Reviewed-by: @tom__chen,@robingrosman,@robingrosman
Signed-off-by:
4 years ago
b40af48366
Fix dir of test case
4 years ago
a177de60ff
!10186 Dvpp debug
...
From: @lizhenglong1992
Reviewed-by: @pandoublefeng,@heleiwang,@liucunwei
Signed-off-by: @liucunwei
4 years ago
83a29a0a54
!10176 Add PyNative Hccl st
...
From: @jojobugfree
Reviewed-by:
Signed-off-by:
4 years ago
dd0da1542d
Add input check
4 years ago
09e10e18bb
momentum weightdecay fusion
4 years ago
e6f0654b61
Add PyNative Hccl st
4 years ago
918679daa3
add tensor.ndim and rename tensor.size() to tensor.size
4 years ago
04da5bae5d
Fix dvpp parameter validation bug and make information clear, besides, modify Dvpp ut.
4 years ago
a8446af1ab
Fix condition check
4 years ago
19a0b61fc9
fix 310 testcase compile
...
Signed-off-by: zhoufeng <zhoufeng54@huawei.com>
4 years ago
bad8c0e422
Modify profiling dir structure to fit the new Run Package
4 years ago
54e5025c1a
equal op gpu kernel
4 years ago
cb0ecf3de7
!10061 add optimizer parallel level0 testcase
...
From: @zhiyuan_mei
Reviewed-by:
Signed-off-by:
4 years ago
893d59afd7
!10097 Fix bugs for np.mean, np.asarray, np.concatenate, np.linspace
...
From: @yanglf1121
Reviewed-by:
Signed-off-by:
4 years ago
eb22cf7e5a
!10073 [MS][CompositeOp][API] TensorDot pr fix
...
From: @danishnxt
Reviewed-by: @tom__chen
Signed-off-by:
4 years ago
ac109b601a
pad_CI_Alarm fix
...
typo fix
test addition for half - pad value checking
4 years ago
904f61b7fd
!9857 [MS][GPU] SparseApplyFtrl_int64_Support
...
From: @danishnxt
Reviewed-by: @tom__chen,@robingrosman
Signed-off-by: @robingrosman
4 years ago
b0b768335e
PR 10072 fix + new sts
...
lintfix
doc fix + code cleaned
doc update 2
4 years ago
98ea8fa6ea
Added tests for chain samplers, fixed GetDatasetSize issue with chained samplers
4 years ago
eb1fd8bd1c
!10113 Add YoloV3 with dvpp preprocess test
...
From: @lizhenglong1992
Reviewed-by: @jonyguo,@liucunwei,@pandoublefeng
Signed-off-by: @liucunwei
4 years ago
1dd4be55dc
!10112 add CPU unsortedsegmentsum
...
From: @zhao_ting_v
Reviewed-by: @wuxuejian,@liangchenghui
Signed-off-by: @wuxuejian
4 years ago
fc11b7dd68
!10074 Fix a testcase error when shuffle is above cache
...
From: @lixiachen
Reviewed-by: @liucunwei,@pandoublefeng
Signed-off-by: @liucunwei,@liucunwei
4 years ago
691e3baab6
!10109 [MD] Fix Decode doc & RandomNode log
...
From: @luoyang42
Reviewed-by: @heleiwang,@liucunwei
Signed-off-by: @liucunwei
4 years ago
6035823c4b
!10044 fix get num classes of concat
...
From: @luoyang42
Reviewed-by: @liucunwei,@heleiwang,@liucunwei
Signed-off-by: @liucunwei,@liucunwei
4 years ago
13f25fa936
add optimizer_parallel level0 case
4 years ago
760cd1870a
corner cases check for np.asarray, np.mean, np.concatenate
4 years ago
bdbdc291f5
high grad
...
Signed-off-by: Daniel <chentanjie@huawei.com>
4 years ago
edf3083510
add CPU unsortedsegmentsum
4 years ago
595b7a782f
[MD] Fix Decode doc & RandomNode log
4 years ago
13204cb8e3
Add YoloV3 with Dvpp test.
4 years ago
734f50f36e
!10083 support non tensor inputs in model predict
...
From: @zhangbuxue
Reviewed-by: @zhunaipan,@zh_qh
Signed-off-by: @zh_qh
4 years ago
f90f77c16e
support non tensor inputs in model predict
4 years ago
da74482812
!9769 add testcases of type error
...
From: @echo_hua
Reviewed-by: @chujinjin,@zhunaipan
Signed-off-by: @zhunaipan
4 years ago
fe41af8129
!10035 remove the 'raise' in construct of Cell
...
From: @zhangbuxue
Reviewed-by:
Signed-off-by:
4 years ago
aeeef7607a
remove the 'raise' in construct of Cell
4 years ago
a7f0e13234
Remove special treatment for cache above repeat
4 years ago
b645f7959d
!9749 [MS][API] UnsortedSegmentMax_Min added check for bug fix - num_segments
...
From: @danishnxt
Reviewed-by:
Signed-off-by:
4 years ago
d1fcfaf0f2
First commit - scatter_nd_int64_update
...
lint fix - no lint required for previous code
fix int64 templates
removed extra files
4 years ago
bc84e9b4e7
range check - UortedSegmin/max num_seg tensors
...
added in checks in PythonAPI to trigger Value errors in case of non dyn
added required check for python API
ST update + op check update
revert segSum change
updated gatherV2 and SplitOp check
added shape -1 sanity check - splitOp
slight refactor in unseg-Min/Max
lintfix
lint fix
4 years ago
a3d4dded12
!9979 BUGFIX: Correct the calculation of the output 'd_x' of the operator LayernormGradGrad
...
From: @david-he91
Reviewed-by: @liangchenghui,@linqingke
Signed-off-by: @liangchenghui
4 years ago
dd5fce4ffb
!10014 apply dynamic shape for ComputeAccidentalHits
...
From: @yanzhenxiang2020
Reviewed-by:
Signed-off-by:
4 years ago
b4ecf65e71
add typeerror testcases
4 years ago
6cc9d2c087
BUGFIX: Correct the calculation of the output 'd_x' of the operator LayernormGradGrad
4 years ago
f75280f6c2
!10006 support int32 dtype of Mul cpu op
...
From: @wuxuejian
Reviewed-by: @kisnwang,@c_34
Signed-off-by: @c_34
4 years ago
e9eb1ebac8
fix shape of ComputeAccidentalHits
4 years ago
bccce3ac21
!10028 adjust st of resnet50 in ascend pynative to gpu pynative
...
From: @lianliguang
Reviewed-by: @kisnwang,@zhoufeng54
Signed-off-by: @zhoufeng54
4 years ago
df9f4f41e8
fix get num classes of concat
4 years ago
mindspore-ci-bot
xutianming
x00540480
yujianfeng
wangmin0104
Ziyan
LianLiguang
buxue
zhuyuxiao
wangshuide2020
wsq3
chendongsheng
hedongdong
xuguoyang
shibeiji
“dangjiaqi1”
yanglf1121
shaoxiangdong
gaojing
Zhenglong Li
Xiao Tianci
lishixing3
yanghaoran
zhengbin (G)
Zirui Wu
zhouyuanshen
Payne
ms_yan
huangxinjing
chenfei
lvliang
Peilin Wang
YangLuo
looop5
wangrao
liyong
caojian05
Jonathan Yan
wanyiming
wilfChen
caifubi
zhoufeng
gzhcv
danishnxt
Mahdi
meizhiyuan
Daniel
zhaoting
Lixia Chen
wanghua
yanzhenxiang2020