Nigel
/
mindspore
forked from mindspore-Ecosystem/mindspore
0
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
56,551 Commits 24 Branches 0 Tags 3.9 GiB
Commit Graph

303 Commits

Author SHA1 Message Date
changzherui 264dccfd0d clean code for python 0918 2022-09-22 19:58:36 +08:00
He Wei 352571d7be Add experimental API for MapParameter 2022-09-19 19:04:25 +08:00
i-robot 242e5c819e
!40599 [feature] change vjp 
Merge pull request !40599 from chenzhuo/master_vjp
 2022-09-15 01:41:13 +00:00
chenzhuo b742f8aa76 modify vjp 2022-09-14 15:08:16 +08:00
chenfei 8402dc3570 fix some ut cases mode set 2022-09-06 16:01:00 +08:00
chenzhuo 5f39ebd788 support grad and value_and_grad with has_aux 2022-08-12 16:43:03 +08:00
huangxinjing 7bd91c0351 Fix warnning clear 2022-08-05 22:34:13 +08:00
yangzhenzhang 5fa63f8e46 delete auto parallel flag 2022-07-30 15:19:56 +08:00
Zhang Qinghua 8d9962d2f7 Support to use methods/attributes of Cell object in graph compiler. 2022-07-07 16:18:25 +08:00
fengyihang 1ed27daee3 delete extra check 2022-06-30 18:29:37 +08:00
i-robot 6c6d8931c3
!36329 Supply Tanhshrink activation function and test cases 
Merge pull request !36329 from 冯一航/Supply_Tanhshrink
 2022-06-24 01:33:32 +00:00
fengyihang 98bf50ffcf Supply Tanhshrink activation function and test cases 2022-06-23 11:34:53 +08:00
i-robot 7943638d00
!35948 Add CPU and GPU operators of ApplyAdamWithAmsgrad with dynamic shape and vmap features support 
Merge pull request !35948 from OwenSec/aawag
 2022-06-23 01:49:18 +00:00
i-robot 5e1b73a836
!36204 Add RReLU activation function and test cases 
Merge pull request !36204 from shaojunsong/feature/RReLU2
 2022-06-22 10:33:36 +00:00
OwenSec 65467272e0 add cpu and gpu operators of applyadamwithamsgrad 2022-06-22 11:33:30 +08:00
shaojunsong 260c920b3d add RReLU, in-place is not supported 2022-06-22 10:58:39 +08:00
i-robot 2919e625e3
!35699 Fix MultiHeadAttention dtype Error 
Merge pull request !35699 from huangxinjing/fix_attention_dtype
 2022-06-20 11:19:32 +00:00
fujianzhao 862752fdce feat: add ops.applyAdadelta CPU backend, add nn.Adadelta, ops.applyAdadelta CPU backend support vmap. 2022-06-19 08:01:14 +08:00
huangxinjing 1918f058c4 Fix linear dtype 2022-06-17 10:29:46 +08:00
i-robot f058997021
!35071 Add feature ReflectionPad1d & ReflecitonPad2d 
Merge pull request !35071 from shaojunsong/feature/reflectionPad
 2022-06-04 13:00:45 +00:00
shaojunsong c0b9d74987 added reflectionpad1d and 2d 2022-06-04 17:43:49 +08:00
lvyufeng ee30ffc3bf add nll_loss and cross_entropy for ops.functional and nn 2022-06-02 11:13:34 +08:00
i-robot 4f1a8b1eef
!33877 operators: Softmin, SiLU 
Merge pull request !33877 from Henry Shi/branch_operators
 2022-06-01 01:53:19 +00:00
Henry Shi 63bd6f035b operators:Softmin & SiLU 2022-05-29 10:42:31 +08:00
i-robot f3148dbe6e
!34842 add nn.Threshold 
Merge pull request !34842 from 吕昱峰（Nate.River）/threshold
 2022-05-25 09:35:33 +00:00
lvyufeng ce1f15666d add nn.Threshold 2022-05-24 15:25:48 +08:00
i-robot dd85238b91
!34579 set uni folder for ge test 
Merge pull request !34579 from xiaoyao/master
 2022-05-23 07:20:37 +00:00
He Wei 2d7c4334ec Support fusion_size argument for flatten_weights 2022-05-20 17:11:48 +08:00
xiao_yao1994 fbdfdb1127 "change default backend in debug graph mode and set uni folder for ge 
test"
 2022-05-20 09:12:03 +08:00
i-robot 20f3a911f1
!34094 Remove the context option: 'enable_sparse'. 
Merge pull request !34094 from 张清华/opt
 2022-05-20 00:52:05 +00:00
Zhang Qinghua d7762f1c8d Add bprop_return_sparse flag for sparse bprop primitive and remove context option: 'enable_sparse'. 2022-05-19 17:31:57 +08:00
fujianzhao 7fcb75beaa [feat] [assistant] [ops] [I51VS2] Add new operator nn.AdaptiveMaxPool1d 2022-05-18 15:50:44 +08:00
fujianzhao cb56dd895d [feat] [assistant] [ops] [I51VS1] Add new operator nn.AdaptiveAvgPool1d 2022-05-16 16:19:24 +08:00
i-robot c07819796c
!34221 [feat] [assistant] [ops] [I51VSC] Add new operator nn.HuberLoss 
Merge pull request !34221 from fujianzhao/huberloss
 2022-05-13 02:57:19 +00:00
i-robot bfd377ca3c
!34125 add nn.ConstantPad1d nn.ConstantPad2d nn.ConstantPad3d and nn.ZeroPad2d 
Merge pull request !34125 from ZhidanLiu/master
 2022-05-13 01:53:23 +00:00
ZhidanLiu fdd6f7c759 add nn.ConstantPad1d nn.ConstantPad2d nn.ConstantPad3d and nn.ZeroPad2d 2022-05-12 10:34:52 +08:00
fujianzhao 4fd3433d23 [feat] [assistant] [ops] [I51VSC] Add new operator nn.HuberLoss 2022-05-11 15:39:02 +08:00
fujianzhao 5bb57674da [feat] [assistant] [ops] [I51VSB] Add new operator nn.Hardtanh 2022-05-11 14:23:11 +08:00
liqiliang 0fabbc4ccb Add a nn cell for softsign. 2022-05-10 17:09:33 +08:00
He Wei 8b0d4e7751 Refactor optimizers to support flatten parameters 2022-04-20 20:23:31 +08:00
He Wei 0d12199928 Enable flatten_gradients for Adam optimizer 2022-04-15 19:49:57 +08:00
He Wei 723543eb5d Modify Optimizer to support flatten parameters 
1. Provide method to create Parameter from Tensor without copy;
2. Adapt optimizer.py to support flattened parameters;
3. Add unit test cases for parameter and optimizer.
 2022-04-07 14:11:32 +08:00
xiao_yao1994 94df0869ad uni_backend: add ut ge test 2022-03-10 16:08:36 +08:00
Margaret_wangrui 462c38813d [ME] Add parameter name check. 2022-02-14 10:07:43 +08:00
liuyang_655 15a153a40d Modify error info 2022-01-06 01:28:53 -05:00
wanyiming ab65f4ac98 addut 2021-11-26 09:48:34 +08:00
chenzhuo d07f24462c add examples of grad, jvp and vjp 2021-11-16 14:15:26 +08:00
chenzhuo f36edec3ff add function jvp 2021-11-12 08:44:41 +08:00
Zhang Qinghua b43693be23 Fix fallback and cell kwargs issues. 2021-10-27 14:06:02 +08:00
Bingya Weng be20f00299 modify ut of probability 2021-10-14 19:33:34 +08:00
huanghui ba66c0d491 add security isolate for save_graphs 2021-09-14 10:24:12 +08:00
i-robot 76fac95591 !23069 Modify error content 
Merge pull request !23069 from huchunmei/master
 2021-09-11 03:50:52 +00:00
huchunmei c75773cf37 Modify error content 2021-09-10 15:21:33 +08:00
yanghaoran 0ced8775c4 upgrade Ascend software package 09 Sep 21 2021-09-09 20:30:22 +08:00
huangxinjing 6cea07f749 Add args check 2021-09-08 09:12:35 +08:00
yao_yf 82889ec56b fixed sparse attention 2021-08-31 16:11:56 +08:00
i-robot ff2e34b9f9 !22509 Add ut and st for Jvp, Vjp and Grad 
Merge pull request !22509 from LiangZhibo/fix
 2021-08-30 08:21:54 +00:00
Zhang Qinghua a137fa1d0b Optimize the Executors routines. 
- Fix the key generating.
- Distinguish the executors.
 2021-08-28 17:36:12 +08:00
l00591931 98fad9f08e Add st and ut for jvp vjp and grad 2021-08-28 17:34:23 +08:00
zhihenghu ce12c02343 Add Sparse Attention 
adjut the file structure and name

Deleted extra information

Do some formatting work

Add test case and fix some document

fix imports
 2021-08-28 11:50:58 +08:00
huangxinjing 62496d75f3 less the interface exposed 2021-08-25 21:22:40 +08:00
ms_yan 36a8886ca2 Revert "[feat] [assistant] [I3T96T] add new Dataset operator CMUARCTICDataset" 
This reverts commit b077aa1cab.

Revert "[feat] [assistant] [I3T96X] add new Dataset operator LibriSpeechDataset"

This reverts commit 4e6f7dc97d.

delete pass_registry_test.cc

comment  hiai_nlu_model_multi.pb related  line
 2021-08-23 01:46:38 +08:00
djc 4e6f7dc97d [feat] [assistant] [I3T96X] add new Dataset operator LibriSpeechDataset 2021-08-22 13:39:37 +08:00
huangxinjing d777742904 1. Move the class to mindspore.parallel, support activation sharding 2021-08-20 14:34:10 +08:00
huangxinjing 18044aff0f 1. Add docstring, elimitate attention mask, tuple append the deocoder return layer past 
split attention inputs from two into three

Add input check and modify test case

Fix check error and add check to transformer inputs

add input check to the transformers
 2021-08-16 21:08:48 +08:00
i-robot 2650aae9ba !21351 complex support 
Merge pull request !21351 from zhouyaqiang0/complex_support
 2021-08-12 09:39:15 +00:00
zhouyaqiang b81ba100b7 add complex number support 2021-08-12 10:36:43 +08:00
jiangzhenguang a3bb780ce4 optimize run memory 2021-08-10 20:45:25 +08:00
huangxinjing 615d1a179d Add transformer layer 2021-07-29 11:10:43 +08:00
huangbingjian ad8f73f32b set default context mode to GRAPH_MODE 2021-06-09 14:40:49 +08:00
huanghui 29f632613f optimize log 2021-06-04 12:32:44 +08:00
huanghui 5acf3c8ade [dfx]optimize exception info print 2021-05-31 14:57:29 +08:00
wangnan39@huawei.com 2761776699 add check for optimizer parameters 2021-05-17 22:23:03 +08:00
jiaqi cdb7330818 diceloss bug 2021-03-09 16:23:45 +08:00
Zhang Qinghua df866f7248 Add TopoSort Rhs First attribute for special CNode, such as Depend CNode with isolated nodes. 2021-03-06 12:39:36 +08:00
Zhang Qinghua 8b8c59f01e Optimize the compile performance in Parser, FG, Manager and Renormalize: 
---
Remove the routine of handling isolated nodes in Renormalize.
Add isolated nodes from Parser&Resolver.
Modify isolated nodes handling in FG&Manager.
Optimize the renormalize routines.
Other optimizations.
 2021-03-04 16:01:54 +08:00
Jiaqi 112b8488c5 add rmse loss 2021-02-25 17:03:07 +08:00
Jiaqi fd1a4726b7 add focal loss 2021-02-22 19:33:26 +08:00
Jiaqi 952e245000 add multclass diceloss 2021-02-19 15:43:30 +08:00
He Wei 7d9a783993 [auto-monad] Support side-effects by auto-monad 
The basic idea is: exploits data dependency to control the execution order
of side-effect operations, and keep the semantics of ANF unchanged.

The ControlDepend primitive is removed and there are two primitives added:

1. UpdateState:
```
  a = Assign(para, value)
```
became:
```
  a = Assign(para, value, u)
  u = UpdateState(u, a)
```

2. Load:
```
  x = Add(para, value)
```
became:
```
  p = Load(para, u)
  x = Add(p, value)
  u = UpdateState(u, p)
```
 2021-02-08 09:01:15 +08:00
wangshuide2020 30f99f2722 add raises description for BCELoss, ReLU, BatchNorm1d, etc. operators. 2021-02-02 20:52:00 +08:00
Jiaqi 804ef79c18 modify dice and diceloss 2021-01-29 15:10:05 +08:00
mindspore-ci-bot 775d71f9c1 !11408 Update PP example 
From: @zichun_ye
Reviewed-by: @wang_zi_dong,@sunnybeike
Signed-off-by:
 2021-01-20 09:17:23 +08:00
Zichun Ye ca316f4422 update doc example in probability 
fix typo in probability module

fix pylint in msp
 2021-01-19 11:51:43 -05:00
Jiaqi 87ecd44f67 develop dice loss 2021-01-16 15:54:06 +08:00
lilei b858097ae4 modify MetaTensor and Tensor 2021-01-04 20:58:34 +08:00
huangxinjing dd0da1542d Add input check 2020-12-18 19:54:49 +08:00
buxue aeeef7607a remove the 'raise' in construct of Cell 2020-12-17 10:54:42 +08:00
l00591931 130c61cfc0 change name from interpolate to resizeBilinear 2020-12-10 11:46:40 +08:00
wangnan39@huawei.com 0fe9e2e4cb support import dynamic_lr from nn 2020-12-09 17:47:52 +08:00
mindspore-ci-bot 231fccf66c !9600 Add Adagrad Optimizer 
From: @zyx5256
Reviewed-by: @liangchenghui,@kingxian
Signed-off-by: @liangchenghui
 2020-12-08 18:29:18 +08:00
zhuyuxiao 37bebc751b add adagrad optim 2020-12-08 09:38:15 +08:00
mindspore-ci-bot e6ec9be6c2 !9386 l1_regularizer开发 
From: @dangjiaqi1
Reviewed-by: 
Signed-off-by:
 2020-12-07 21:58:58 +08:00
“dangjiaqi1” 39948670dd new update basic.py which add class L1Regularizer(Cell). new add test_l1_regularizer.py. new add test_l1_regularizer.py 2020-12-07 20:08:35 +08:00
Xun Deng 4ef9e9e57f fix minor issues in distribution classes 2020-12-03 14:47:08 -05:00
peixu_ren 37d40bc495 Add Beta distribution 2020-11-27 17:07:54 -05:00
peixu_ren 93d4ef428f Add Gamma distribution 2020-11-25 09:40:46 -05:00
peixu_ren 01f5da0a14 Add Poisson distribution 2020-11-24 11:55:39 -05:00
mindspore-ci-bot 7f725b93a3 !8234 Remove expm1_generic and log1p_generic from PP utils 
From: @peixu_ren
Reviewed-by: 
Signed-off-by:
 2020-11-23 09:56:07 +08:00
mindspore-ci-bot e3371a01d5 !8751 Fix minor bugs in bijector and distribution 
From: @shallydeng
Reviewed-by: @zichun_ye,@sunnybeike
Signed-off-by:
 2020-11-23 08:50:57 +08:00