mindspore-Ecosystem
/
mindspore
21
0
Fork 27
Code Issues Pull Requests 6 Projects Releases Wiki Activity
48,700 Commits 23 Branches 30 Tags 4 GiB
Commit Graph

10548 Commits

Author SHA1 Message Date
i-robot bdc8b0efd4
!37660 unsorted segment ids less than 0 
Merge pull request !37660 from ling/pr
 2022-07-11 01:22:17 +00:00
fan-jibin 8e65d5208c add adaptive_maxpool3d rank dynamic 2022-07-10 20:15:16 +08:00
hedongdong 4de4eb1a43 Refactor Ops MaxUnpool2D, MaxUnpool3D, MultiMarginLoss, MultilabelMarginLoss, TripletMarginLoss, BartlettWindow 2022-07-10 16:58:09 +08:00
lvxinyu 0914ecf339 [feat][assistant][I4OFJL] Add new aicpu operator BartlettWindow 2022-07-10 12:27:00 +08:00
zhou ruikang 4c1ed474af [feat][assistant][I4OFKB] Add new aicpu operator TripletMarginLoss 2022-07-10 12:26:59 +08:00
zsalliance 9afcffa05e [feat][assistant][I40FKS, I4CRKS]Add MultilabelMarginLoss and MultilabelMarginLossGrad 2022-07-10 12:26:59 +08:00
桂宁馨 144457f51d [feat] [assistant] [I40FKL, I40FKN] Add MultiMarginLoss, MultiMarginLossGrad 2022-07-10 12:26:59 +08:00
fateld bedfb1ab5c [feat][assistant][I4OFJT, I4OFJX, I4OFJY, I4OFJZ] Add new aicpu operator MaxUnpool2D, MaxUnpool2DGrad, MaxUnpool3D and MaxUnpool3DGrad 2022-07-10 12:26:58 +08:00
Huang-Jingg fea9347a38 [feat][assistant][I4OFKG, I4OFKI] Add new aicpu operator Mvlgamma and MvlgammaGrad 2022-07-10 12:26:55 +08:00
i-robot c3174b48cb
!37503 Add CPU version of SparseApplyFtrl op 
Merge pull request !37503 from TinaMengtingZhang/sparse_apply_ftrl_op_dev
 2022-07-09 23:58:02 +00:00
i-robot 81c51ff22e
!37703 fix CropResize op on CPU 
Merge pull request !37703 from zhangbuxue/fix_crop_resize_master
 2022-07-09 10:32:08 +00:00
wangshengnan12@huawei.com 494826b77c renorm_support_complex_gpu 2022-07-09 17:57:18 +08:00
i-robot 3638e7fb88
!35858 add concat batch function 
Merge pull request !35858 from jiangshuqiang/add_concat_batch
 2022-07-09 09:20:09 +00:00
buxue c4c9f907d2 fix CropResize op on CPU 2022-07-09 16:56:34 +08:00
i-robot cd201a75d6
!37044 add bce_with_logits_loss vmap 
Merge pull request !37044 from zhuzhongrui/pub_master2
 2022-07-09 08:14:33 +00:00
i-robot 1057e8440c
!37656 fix bias add bug 
Merge pull request !37656 from luoyang/biasadd_high_grad-fix
 2022-07-09 06:59:22 +00:00
z00512249 eeef1c3ab9 add bce_with_logits_loss vmap and cpu performance opt 2022-07-09 14:55:12 +08:00
i-robot 39f2b6803f
!37596 remove dynamic tests of adam_apply_one_with_decay op 
Merge pull request !37596 from Yanzhi_YI/remove_tests
 2022-07-09 05:31:35 +00:00
i-robot f089b01fa3
!37505 add random_gamma support dynamic shape and functional api 
Merge pull request !37505 from zhangqi/0706
 2022-07-09 02:59:57 +00:00
caifubi 9e1f0b0d41 Bugfix for hccl broadcast 2022-07-09 10:25:33 +08:00
i-robot 2fd95519c8
!37658 fix CropResize op on CPU 
Merge pull request !37658 from zhangbuxue/fix_crop_size
 2022-07-09 01:55:03 +00:00
i-robot 8521c3e55b
!37609 Fix UniqueConsecutive 0 dimension 
Merge pull request !37609 from huangbingjian/unique_consecutive_0d
 2022-07-09 01:45:05 +00:00
TinaMengtingZhang 063ff19296 fix precision issue of GridSampler2D/3D forward and backward op 2022-07-08 17:01:45 -04:00
YangLuo aa1fc379ae fix bias add bug 2022-07-08 23:57:57 +08:00
i-robot 99bf262b1c
!37640 Fix an ownership bug in Tensor.asnumpy() 
Merge pull request !37640 from hewei/fix_master
 2022-07-08 13:05:45 +00:00
buxue 973616168e fix CropResize op on CPU 2022-07-08 20:09:18 +08:00
i-robot 9b50df63df
!35847 [FIX] [BUG] [MD] [FUNC] RandomEqualize 
Merge pull request !35847 from yangwm/randomequalize
 2022-07-08 09:46:31 +00:00
i-robot 467248e24f
!36476 [feat] [assistant] [ops] [I4ZZUE] New GPU operator implementation NextAfter 
Merge pull request !36476 from 路雄博/NextAfter
 2022-07-08 09:15:58 +00:00
i-robot 949e4dd443
!34893 [MS][GPU] UpsampleNearest3D grad Op + UpsampleTrilinear3D Op validation fix 
Merge pull request !34893 from danishfarid/UpsampleNearest3D_Grad
 2022-07-08 09:09:22 +00:00
ling b528dda646 unsorted segment ids less than 0 2022-07-08 17:05:30 +08:00
zhujingxuan 2b138c80e2 expand support types for inplace_add and inplace_sub 2022-07-08 16:52:04 +08:00
yanghaoran 7573f1dfba
回退 'Pull Request !37403 : soft shrink cpu optimize' 2022-07-08 08:18:39 +00:00
i-robot 31256de8e4
!37619 Take offline ge not urgency testcase 
Merge pull request !37619 from weiyang/rm_ge_str
 2022-07-08 08:02:23 +00:00
i-robot f02867e3b8
!37403 soft shrink cpu optimize 
Merge pull request !37403 from ling/op
 2022-07-08 06:15:30 +00:00
He Wei f753bb6b34 Fix an ownership bug in Tensor.asnumpy() 
Ownership should be handled correctly in Tensor.asnumpy(),
otherwise there are potential risks for 'use after free' problem.
 2022-07-08 11:52:54 +08:00
huangbingjian 8a2f835803 Fix UniqueConsecutive 0 dimension 2022-07-08 11:48:58 +08:00
i-robot d82e31e029
!37556 Fix several bugs in vision processing 
Merge pull request !37556 from xiaotianci/fix_bugs
 2022-07-08 03:43:19 +00:00
yanghaoran f9de9bff39
!37635 回退 'Pull Request !37407 : [MS][ASCEND] add smooth_l1_loss ascend op' 
Merge pull request !37635 from yanghaoran/revert-merge-37407-master
 2022-07-08 03:12:46 +00:00
i-robot 3928a3dbdc
!37508 optimize gpu dropout 
Merge pull request !37508 from kisnwang/optimize-gpu-dropout
 2022-07-08 03:12:23 +00:00
yanghaoran c249e9680c
回退 'Pull Request !37407 : [MS][ASCEND] add smooth_l1_loss ascend op' 2022-07-08 03:11:47 +00:00
panshaowu e908c6effc add dynamic shape supports and remove Tensor interface of StandardLaplace operator 2022-07-08 11:09:23 +08:00
i-robot 2133d0333e
!37407 [MS][ASCEND] add smooth_l1_loss ascend op 
Merge pull request !37407 from mengyuanli/smooth_l1_loss_ascend
 2022-07-08 02:59:39 +00:00
i-robot 3334a4efef
!37598 Add st for primitive cache. 
Merge pull request !37598 from liqiliang/accelerate_function_cache
 2022-07-08 02:52:19 +00:00
yangzhenzhang 3b7fc4db29 modify strategys to strategies 2022-07-08 10:48:05 +08:00
zhangqi ae6c8c807d add random_gamma support dynamic shape and functional api 2022-07-08 10:15:32 +08:00
i-robot 389dc9ac27
!37493 Support to use methods/attributes of Cell object in graph compiler. 
Merge pull request !37493 from 张清华/opt0
 2022-07-08 02:03:45 +00:00
zhengzuohe b4e7178269 add testcases for tensor index operations 2022-07-08 10:02:13 +08:00
i-robot a2c401e6c3
!37367 modify gather op 
Merge pull request !37367 from yangzhenzhang/modify-gather-op
 2022-07-08 02:02:02 +00:00
yangwei 36437a4503 remove ge string case 2022-07-08 09:54:04 +08:00
danishfarid ca24d970e7 UpsampleNearest3D_Grad GPU 
reimplement changes as per comments

UpTri3D_Fix
 2022-07-07 21:20:12 +00:00
TinaMengtingZhang 4ecbbb7b08 Add SparseApplyFtrl cpu op 2022-07-07 13:57:32 -04:00
fan-jibin fd36536036 fix masked_fill with float value 2022-07-07 23:36:38 +08:00
i-robot 599cf51f1d
!37452 sparse_concat调整cpu算子输入格式调整，以符合算子统一定义 
Merge pull request !37452 from baochong/master
 2022-07-07 14:23:33 +00:00
i-robot 96f6177253
!37405 skip test_autotune_pymultiproc_train_simple_model 
Merge pull request !37405 from luoyang/skip_autotune_st
 2022-07-07 13:22:01 +00:00
i-robot c5b1d3db41
!36726 [MS][LITE] StandardNormal dynamic support 
Merge pull request !36726 from luoyuan/StandardNormal-dynamic-support
 2022-07-07 13:18:58 +00:00
Xiao Tianci fb474f0f1e fix several bugs 2022-07-07 20:19:33 +08:00
liqiliang 795a1ad710 Add st for primitive cache. 2022-07-07 19:49:20 +08:00
yiyanzhi_akane d5f4d255ce remove dynamic tests of adam_apply_one_with_decay op 2022-07-07 19:28:06 +08:00
buxue 6ddd62ff65 support range reverse 2022-07-07 19:08:23 +08:00
mengyuanli bae519664f add smooth_l1_loss_v2 tbe op 2022-07-07 17:32:23 +08:00
ling e70a22039c soft shrink cpu optimize 2022-07-07 17:22:23 +08:00
Zhang Qinghua 8d9962d2f7 Support to use methods/attributes of Cell object in graph compiler. 2022-07-07 16:18:25 +08:00
i-robot 6b3ad624e9
!37458 [OP] add vmap for MaxpoolGradGradWithArgmax 
Merge pull request !37458 from yangruoqi713/argmax
 2022-07-07 07:26:26 +00:00
kswang a342686581 optimize gpu dropout 2022-07-07 15:10:50 +08:00
yangzhenzhang 25009f9051 modify gather op 2022-07-07 14:59:50 +08:00
i-robot 4df4fc949f
!37249 upgrade_ascend_20220704_master 
Merge pull request !37249 from andy_wangrui/upgrade_ascend_20220704_master
 2022-07-07 03:58:23 +00:00
i-robot 2581b2d54b
!37444 Add supported type for argmin with value 
Merge pull request !37444 from ZPaC/argmin-support
 2022-07-07 03:34:12 +00:00
i-robot f2c46820ee
!37322 use funcgraph with dynamic shape 
Merge pull request !37322 from xychow/fix-bprop-relu-with-different-input-shape
 2022-07-07 03:33:14 +00:00
i-robot dffd4295c9
!37247 BCEWithLogitsLoss suit for dynamic shape 
Merge pull request !37247 from hanhuifeng/bce_dyn
 2022-07-07 02:41:18 +00:00
i-robot 6f0d5f6a40
!36650 [feat] [assistant] [I4ZZSA] Add new GPU operator MulNoNan 
Merge pull request !36650 from 徐锦明/MulNoNan2
 2022-07-07 02:31:16 +00:00
i-robot 68b1cb0eb9
!37426 [Fallback] Support issubclass built-in functions in constant scenarios. 
Merge pull request !37426 from Margaret_wangrui/builtin_type_fallback
 2022-07-07 01:42:19 +00:00
i-robot 08c0bc583f
!37390 [MD][UT] Transforms Unification - Verify selected op implementation 
Merge pull request !37390 from cathwong/ckw_xtranuni_ut_validate
 2022-07-06 18:50:29 +00:00
baochong a3913ca1eb modify sparse_concat args 2022-07-06 19:49:19 +08:00
luoyang 47116f31a6 skip test_autotune_pymultiproc_train_simple_model 2022-07-06 18:34:56 +08:00
i-robot 0af86fa175
!37408 [Renorm] Add Complex support for renorm 
Merge pull request !37408 from Xiaoda/135-renorm-adaption
 2022-07-06 10:29:18 +00:00
luoyuan 4e0ea6566f [MS][LITE] StandardNormal ops dynamic support and vmap 2022-07-06 17:34:43 +08:00
ZPaC c7e46d508f Add supported type for argmin with value 2022-07-06 16:41:33 +08:00
i-robot ddec1fe9bc
!37087 unsorted segment min/max npu ds 
Merge pull request !37087 from ling/pr
 2022-07-06 08:02:22 +00:00
格物致知 cd15645693 [feat] [assistant] [I4ZZSA] Add new GPU operator MulNoNan 2022-07-06 15:52:39 +08:00
yangruoqi713 391435d56d [OP] add vmap for MaxpoolGradGradWithArgmax 2022-07-06 15:45:04 +08:00
wenyuan_echo 7a724aa0b0 [feat] [assistant] [ops] [I4ZZUE] New GPU operator implementation NextAfter 2022-07-06 15:42:21 +08:00
i-robot 54359e30ba
!37447 takedown test_dyn_ab to ensure gate stability 
Merge pull request !37447 from yanghaoran/takedown_testcases
 2022-07-06 07:30:29 +00:00
zhousiyi d119345240 use bprop funcgraph with dynamic shape if multiple version exists. 2022-07-06 07:26:48 +00:00
i-robot a6ff9e0d77
!37269 soft shrink grad test case 
Merge pull request !37269 from ling/bug
 2022-07-06 07:26:27 +00:00
i-robot 519edd4bb6
!37195 Drop aot custom op test on mac 
Merge pull request !37195 from zichun_ye/aot_test_fix
 2022-07-06 07:10:39 +00:00
i-robot 2980e73b2d
!37203 fix pipeline.cc log 
Merge pull request !37203 from huanghui/fix-log
 2022-07-06 06:56:27 +00:00
shenyu 2c9cebab70 fix bug function 2022-07-06 14:53:47 +08:00
yanghaoran 5e5db83a8e takedown test_dyn_ab to ensure gate stability 2022-07-06 14:46:39 +08:00
i-robot 266e25216d
!37241 [OP] add maxpool grad grad with argmax cpu and gpu backend 
Merge pull request !37241 from yangruoqi713/argmax
 2022-07-06 06:27:27 +00:00
Xiaoda Zhang a72ae2acff add renorm complex support 2022-07-06 13:44:15 +08:00
i-robot ebd0fe1d7a
!33307 [assistant][ops] Add SparseTensorDenseMatMul 
Merge pull request !33307 from 余彦松/SparseTensorDenseMatMulv3
 2022-07-06 04:00:42 +00:00
Margaret_wangrui 2897ac9af2 [Fallback] Support issubclass built-in functions in constant scenarios. 2022-07-06 11:41:10 +08:00
i-robot ff86996b45
!37309 fix implicitly converted error log for master 
Merge pull request !37309 from chujinjin/fix_error_log_for_master
 2022-07-06 01:48:43 +00:00
i-robot 7ffc074a39
!35835 Support arg min with value 
Merge pull request !35835 from ZPaC/argmin-support
 2022-07-06 01:29:52 +00:00
yangruoqi713 d4f97e8aaf [OP] add maxpool grad grad with argmax cpu and gpu backend 2022-07-06 09:26:20 +08:00
i-robot 2a953992f5
!37127 Support high grad of StridedSlice 
Merge pull request !37127 from luoyang/stridedslices_high_grad
 2022-07-06 01:12:28 +00:00
Cathy Wong 8e045ba83d [MD][UT] Transforms Unification - Verify selected op implementation 2022-07-05 20:53:07 -04:00
i-robot 512a45c3f2
!36864 Fix vmap problem when using tuple input 
Merge pull request !36864 from LiangZhibo/vmap
 2022-07-05 13:38:59 +00:00
余彦松 5a26611fe8 Add SparseTensorDenseMatMul 2022-07-05 21:07:19 +08:00
i-robot d384d6acc8
!37280 fix adaptive_maxpool3d testcase 
Merge pull request !37280 from 范吉斌/fix_adaptive_maxpool3d
 2022-07-05 12:57:42 +00:00
i-robot 2415cebdd7
!37205 [feature]support operator Tan in Taylor differentiation 
Merge pull request !37205 from chenzhuo/master
 2022-07-05 12:51:07 +00:00
huanghui c293362075 fix outermost network inputs check log 2022-07-05 20:12:19 +08:00
ling 1296dd3c98 unsorted segment min/max npu ds 2022-07-05 17:36:09 +08:00
hanhuifeng2020 19a7b5be46 BCEWithLogitsLoss suit for dynamic shape 2022-07-05 17:31:59 +08:00
i-robot 5dc8e850e2
!37209 Fix bug of list_getitem eliminate 
Merge pull request !37209 from chenfei_mindspore/master
 2022-07-05 09:27:46 +00:00
chujinjin 456092b3b0 fix implicityly converted log error 2022-07-05 16:49:47 +08:00
i-robot 7e7cbe55a4
!37117 [assistant][ops] Add SegmentMean, SegmentProd 
Merge pull request !37117 from 王乐/SegmentMean
 2022-07-05 08:32:35 +00:00
i-robot 6fb3345288
!32257 add linspace pass 
Merge pull request !32257 from jiangzhenguang/add_linspace_pass
 2022-07-05 07:48:48 +00:00
YangLuo 9b559be6db Support high grad of StridedSlice 2022-07-05 15:46:45 +08:00
liangzhibo 03df4e4bf5 Fix vmap problem 2022-07-05 15:30:10 +08:00
jiangshuqiang 90e58867ec add concat batch function 2022-07-05 15:18:02 +08:00
ZPaC bbd9a04b0f Support arg min with value 2022-07-05 15:11:11 +08:00
ling 5e181366cc soft shrink grad test case 2022-07-05 15:09:25 +08:00
i-robot a0125cc73b
!37215 Since the float64 tensor_add operation is not supported in the Ascend platform, modify the test case. 
Merge pull request !37215 from Margaret_wangrui/float_numpy
 2022-07-05 06:31:57 +00:00
chenzhuo 4be29b46b5 support operator Tan in Taylor differentiation 2022-07-05 11:47:58 +08:00
i-robot 2bb9ecdb41
!37103 Change vmap rule for IndexFill opeartor. 
Merge pull request !37103 from hezhenhao1/add_index_fill
 2022-07-05 03:40:54 +00:00
i-robot d33cf03653
!37216 Support None shape input for ArgMaxWithValue 
Merge pull request !37216 from YuJianfeng/arg_max_with_value_ds
 2022-07-05 03:40:24 +00:00
i-robot 3a73522ee5
!37101 for gpu operation ApplyAdagradDA vmap and testcase 
Merge pull request !37101 from 于振华/ApplyAdagradDAD_vmap_testcase_0702
 2022-07-05 03:29:02 +00:00
fan-jibin 30c0989b00 fix adaptive_maxpool3d testcase 2022-07-05 10:46:44 +08:00
andy_wangrui 7dfa94c483 add jjf 6c8dc54bce 2022-07-05 09:55:15 +08:00
chenfei 7b917b83be fix bug of tuple setitem getitem pass 2022-07-05 09:34:28 +08:00
Bronyale ae87c7dd6b Add SegmentMean and SegmentProd 2022-07-05 01:45:27 +08:00
i-robot d824f2ae8e
!37116 Add some exception informations. 
Merge pull request !37116 from liqiliang/matrixbandpart-gpu
 2022-07-04 15:00:02 +00:00
i-robot 961a8ebca0
!36926 Support high grad of BiasAdd 
Merge pull request !36926 from luoyang/biasadd_high_grad
 2022-07-04 13:46:42 +00:00
i-robot 2ddd5014ae
!36976 [OP] optimieze the performance of max pool grad grad 
Merge pull request !36976 from yangruoqi713/pool
 2022-07-04 12:43:29 +00:00
hezhenhao1 73a08a4fd7 Change vmap rule for IndexFill opeartor. 2022-07-04 20:00:49 +08:00
i-robot 8b71af74c9
!36561 Support ms_function for user defined class methods, not only for cell methods. 
Merge pull request !36561 from 张清华/opt0
 2022-07-04 11:08:45 +00:00
liqiliang d399dc7b44 Add some exception informations. 2022-07-04 17:51:26 +08:00
yujianfeng 0e23044a82 Support None shape input for ArgMaxWithValue 2022-07-04 17:43:27 +08:00
Margaret_wangrui b74b945923 Since the float64 tensor_add operation is not supported in the Ascend platform, modify the test case. 2022-07-04 17:37:11 +08:00
yangruoqi713 247a0d3d86 [OP] optimieze the performance of max pool grad grad 2022-07-04 16:10:14 +08:00
Zichun Ye 9b4a19b503 update mac test case 2022-07-04 16:01:22 +08:00
i-robot 528bb57ae9
!36420 [MS][LITE][parallel] modify default thread num 
Merge pull request !36420 from yefeng/342-modify_default_param
 2022-07-04 07:51:21 +00:00
YangLuo a9353e45b7 Support high grad of BiasAdd 2022-07-04 15:47:56 +08:00
Zhang Qinghua 8d48c81857 Optimize the error log description for CreateInstance. 
Support ms_function for user defined class methods, not only for cell methods.
 2022-07-04 14:53:08 +08:00
i-robot bade14590c
!37146 adjust inputs and outputs sequence 
Merge pull request !37146 from zhangbuxue/adjust_inputs_and_outputs_sequence
 2022-07-04 06:49:00 +00:00
i-robot 12b3284605
!36659 支持coleasce算子 
Merge pull request !36659 from 黄晓/coleace_0625
 2022-07-04 06:11:56 +00:00
jiangzhenguang c4b670c6c5 add linspace pass 2022-07-04 13:43:20 +08:00
i-robot 55da1a87d5
!35287 Add tensor&functional interface for TensorScatterMax ops. 
Merge pull request !35287 from wuweikang/tensor-scatter-max
 2022-07-04 03:11:11 +00:00
buxue c9b3b5a176 adjust inputs and outputs sequence 2022-07-04 11:03:10 +08:00
muxiyin 2da36d1a1c support coleasce 2022-07-04 09:52:21 +08:00
i-robot f704ba6c32
!36968 GraphKernel Add enable debug mode flag 
Merge pull request !36968 from ZengZitao/debug_test
 2022-07-04 01:50:33 +00:00
i-robot 1bbb3fe743
!37118 support dynamic shape for batchnorm_grad_grad op 
Merge pull request !37118 from zhangbuxue/support_batchnormgradgrad_op_dynamic_shape
 2022-07-04 01:17:10 +00:00
buxue c4255d3822 support dynamic shape for batchnorm_grad_grad op 2022-07-03 17:21:27 +08:00
i-robot 0de2e9da96
!37115 fix the result wrong for Range op on CPU 
Merge pull request !37115 from zhangbuxue/fix_Range_op_result_wrong_on_cpu
 2022-07-03 04:32:30 +00:00
buxue b3f4b25587 fix range op result wrong on CPU 2022-07-02 19:14:21 +08:00
i-robot 69e5e25567
!36833 Add dynamic shape support for Squeeze 
Merge pull request !36833 from zhujingxuan/squeeze
 2022-07-02 10:42:02 +00:00
wuweikang 1ca9c8984b Add tensor&functional interface for TensorScatterMax 2022-07-02 18:22:37 +08:00
zhujingxuan c9f0a04f70 Add dynamic shape support for Squeeze 2022-07-02 17:33:55 +08:00
buxue 981e9013b4 develop batchnormgradgrad on GPU 2022-07-02 14:45:10 +08:00