Nigel
/
mindspore
forked from mindspore-Ecosystem/mindspore
0
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
45,876 Commits 24 Branches 0 Tags 3.9 GiB
Commit Graph

930 Commits

Author SHA1 Message Date
i-robot 1b3e06faf1
!34645 masked select grad gpu implement 
Merge pull request !34645 from zhaodezan/master
 2022-05-23 05:53:59 +00:00
i-robot eb856f0d79
!34695 [kernel]fix review comments 
Merge pull request !34695 from 张学同/opt_kernel
 2022-05-23 03:34:31 +00:00
yangzhenzhang 3b187616e7 add vmap for masked select 2022-05-23 09:10:35 +08:00
zhangxuetong 7463142a05 add apply_proximal_adagrad_gpu_kernel 2022-05-23 09:08:30 +08:00
Signed-off-by: jakcmanftr c4bbbfa6c7 [assistant][ops]New operator implementation for Bessel 
Signed-off-by: jakcmanftr <18407825861@163.com>
Signed-off-by: zhangyukun <38148677+jameszhangyukun@users.noreply.github.com>

[assistant][ops]New operator implementation for Bessel

Signed-off-by: jakcmanftr <18407825861@163.com>
Signed-off-by: zhangyukun <38148677+jameszhangyukun@users.noreply.github.com>
 2022-05-22 21:35:38 +08:00
zhaodezan 6d9721fbf9 masked select grad gpu first commit 2022-05-22 16:53:34 +08:00
i-robot 68a9ec284d
!33726 [feat] [assistant] [ops] [I4ZZUR] New GPU operator implementation, include RightShift 
Merge pull request !33726 from zhixinaa/RightShift
 2022-05-21 09:23:36 +00:00
ling 16c573e1ea softshrink softshrinkgrad 2022-05-20 17:09:07 +08:00
zhixinaa d30d43b70d [feat] [assistant] [ops] [I4ZZUR] New GPU operator implementation, include RightShift 2022-05-20 15:33:37 +08:00
limingqi107 817e60a308 add tensor/functional interface for diag op 2022-05-20 15:05:11 +08:00
zhangxuetong c089794e62 fix review comments 2022-05-20 14:53:26 +08:00
i-robot 698ec02a9d
!34670 adapt vmap for scatter_min operator 
Merge pull request !34670 from hujiahui8/scatter_min
 2022-05-20 02:58:07 +00:00
i-robot 92f73da146
!34616 Tan gpu inference 
Merge pull request !34616 from ling/gpu
 2022-05-20 01:34:17 +00:00
i-robot 527539ef7c
!34409 unsorted segment prod gpu inference 
Merge pull request !34409 from ling/core
 2022-05-20 01:20:54 +00:00
i-robot e4148639c4
!34652 [kernel][gpu]add assert kernel 
Merge pull request !34652 from 张学同/assert
 2022-05-20 01:17:16 +00:00
hujiahui8 2ffa88726c adapt vmap for scatter_min operator 2022-05-19 23:47:44 +08:00
i-robot 4dab24be13
!34479 add gpu diag op 
Merge pull request !34479 from limingqi107/bug_fix3
 2022-05-19 12:16:51 +00:00
zhangxuetong 66f56e41eb add assert gpu kernel 2022-05-19 19:18:19 +08:00
ling bbbe133362 Tan gpu inference 2022-05-19 19:02:04 +08:00
ling 437deb9da2 unsorted segment prod gpu inference 2022-05-19 18:59:19 +08:00
liangzhibo 46d4c8dc37 Add test case for vmap wrong in_axis of range 2022-05-19 14:49:15 +08:00
i-robot 7c11e3a820
!34480 gpu support atanh kernels 
Merge pull request !34480 from chenweifeng/gpu-atanh-kernel
 2022-05-19 06:23:32 +00:00
limingqi107 ef55422c04 add gpu diag op 2022-05-19 11:56:13 +08:00
huanghui 636192a116 Support vmap for Svd 2022-05-19 09:29:13 +08:00
wilfChen 1af2656199 gpu support atanh kernel 2022-05-18 15:46:26 +08:00
yangzhenzhang c7a59aa805 add gpu impl of masked select 2022-05-18 09:14:07 +08:00
mengyuanli 49c5b52b9f 1.scatter elements support update input or not 
2.add scatter gpu kernel
 2022-05-16 20:45:03 +08:00
i-robot 26a1c55fd8
!34334 Support high dimension input for SliceGrad on CPU and GPU 
Merge pull request !34334 from xiaotianci/slice_grad
 2022-05-16 11:14:17 +00:00
i-robot 9118cf48ca
!33420 Support GPU Svd, and add tensor, functional, primitive 
Merge pull request !33420 from huanghui/svd
 2022-05-16 02:11:34 +00:00
i-robot 0614c34a9f
!34386 Add a gpu kernel, Padding. 
Merge pull request !34386 from liqiliang/padding-gpu
 2022-05-16 02:11:13 +00:00
Xiao Tianci 0006a63066 SliceGrad supports 8D input in CPU and GPU 2022-05-16 10:06:22 +08:00
i-robot 9e7b82c9cb
!34324 [MS][OPS] Ger FastGeLU FastGeLUGrad pr3 
Merge pull request !34324 from KXiong/master
 2022-05-16 01:14:39 +00:00
i-robot c25f2bb443
!34333 Add a gpu kernel, MatrixBandPart. 
Merge pull request !34333 from liqiliang/matrixbandpart-gpu
 2022-05-14 10:40:00 +00:00
liqiliang 5daf1dd2d4 Add a gpu kernel, Padding. 2022-05-14 17:46:36 +08:00
huanghui cc1f903ce3 1. add Svd Tensor, functional, primitive. 
2. support GPU Svd kernel.
 2022-05-14 15:34:19 +08:00
i-robot f080d84646
!34374 Update NonZero primitive to support dynamic shape. 
Merge pull request !34374 from hezhenhao1/add_nonzero
 2022-05-14 06:29:37 +00:00
hezhenhao1 7d9f53b9fb Update NonZero primitive to support dynamic shape. 2022-05-14 11:38:16 +08:00
liqiliang 4be416f40f Add a gpu kernel, MatrixBandPart. 2022-05-14 10:46:38 +08:00
i-robot 377ff71a78
!34166 Add GPU implementation of MatrixDiagV3 operator. 
Merge pull request !34166 from hezhenhao1/add_matrix_diag
 2022-05-14 02:30:59 +00:00
liangzhibo 1b41721978 Add select tensor, functional interface and vmap 2022-05-13 17:02:21 +08:00
xiongkun f80d30e415 [MS][ops]adjust ger cpu operator 
[MS][ops]adjust ger cpu operator

ajust FastGelu and FastGeluGrad

ajust FastGelu and FastGeluGrad

adjust fast_gelu

adjust fast_gelu

adjust fast_gelu

adjust fast_gelu

adjust fast_gelu

adjust bitwise doc

adjust bitwise doc

adjust bitwise doc

adjust bitwise doc

add fast_gelu initsizelist

add fast_gelu initsizelist

ajust fast_gelu_grad_interface

ajust fast_gelu_grad_interface

add fast_gelu doc

add fast_gelu doc

add fast_gelu doc

add fast_gelu doc

add fast_gelu doc

add fast_gelu doc

add fast_gelu doc

add fast_gelu doc

add fast_gelu doc

add FastGeLU FastGeLUGrad vmap

add FastGeLU FastGeLUGrad vmap

add FastGeLU FastGeLUGrad vmap

add FastGeLU FastGeLUGrad vmap

add FastGeLU FastGeLUGrad vmap

add ger vmap

add ger vmap

add ger vmap

add ger vmap

adjust code

adjust code

adjust code

adjust code

adjust code

adjust code

add ger doc

add ger doc

adjust doc

remove bitwise kernelmod

adjust doc

adjust doc

replace tensor

replace tensor

ajust doc

adjust test

adjust test

adjust doc
 2022-05-13 15:43:47 +08:00
yuan d04a96e0bb [feat] [assistant] [ops] [I4ZZV4,I4ZZV5] New GPU operator implementation, include tril,triu 2022-05-13 15:16:59 +08:00
hezhenhao1 d5ee2e260c Add GPU implementation of MatrixDiagV3 operator. 2022-05-13 14:57:03 +08:00
i-robot 0968b28871
!34131 scatter_min: add dynamic_shape case for ascend 
Merge pull request !34131 from hujiahui8/scatter_min
 2022-05-13 06:35:46 +00:00
yangruoqi713 01a06810bf [OP] add function and tensor interface for hard shrink op 2022-05-13 09:37:42 +08:00
i-robot 7c7aa8430a
!34187 Add range functional and vmap 
Merge pull request !34187 from LiangZhibo/range_ops
 2022-05-12 17:40:17 +00:00
i-robot 4d9e0618c7
!34216 [kernel]add assign sub 
Merge pull request !34216 from 张学同/assign_sub
 2022-05-12 09:19:23 +00:00
hujiahui8 71f98273ea scatter_min: add dynamic_shape case for ascend 2022-05-12 17:08:02 +08:00
liangzhibo a30897fcfa Add range functional, tensor and vmap 2022-05-12 15:17:59 +08:00
zhangxuetong 8ac51461f5 add assign sub kernel 2022-05-12 09:27:03 +08:00
liangzhibo 7fa0a77e08 Add round functional, tensor interface and vmap 2022-05-12 09:22:38 +08:00
zhangxuetong e727deddce add ceil kernel 2022-05-11 16:15:28 +08:00
hanhuifeng2020 0659d49d56 fix the bug of StridedSlice op and DynamicStitch op 2022-05-10 20:54:00 +08:00
i-robot 309d022ff1
!33850 Add IndexFill primitive and adapt GPU implementation to new interface. 
Merge pull request !33850 from hezhenhao1/add_index_fill
 2022-05-10 09:26:26 +00:00
i-robot ca4c85f2d2
!33924 [GPU][OP] add hard shrink gpu kernel and hard shrink grad gpu kernel 
Merge pull request !33924 from yangruoqi713/hshrink
 2022-05-10 03:14:14 +00:00
i-robot 0ad97e624a
!34016 [kernel]add celu 
Merge pull request !34016 from 张学同/celu
 2022-05-10 01:17:16 +00:00
hezhenhao1 15eb164049 Add IndexFill primitive and adapt GPU implementation to new interface. 2022-05-10 09:09:08 +08:00
yangruoqi713 7905a4ab92 [GPU][OP] add hard shrink gpu kernel and hard shrink grad gpu kernel 2022-05-09 15:01:48 +08:00
zhangxuetong 16dfb85f86 add celu kernel 2022-05-07 16:57:49 +08:00
hezhenhao1 35a70030cd Add GPU implementation of NonZero operator. 2022-05-07 15:45:59 +08:00
i-robot b8d85a03eb
!33830 ScatterMax support dynamic shape 
Merge pull request !33830 from polyhedral/max
 2022-05-06 02:46:41 +00:00
i-robot 497fb1d575
!33683 Add half registration for Cosh 
Merge pull request !33683 from zhujingxuan/master
 2022-05-05 08:09:02 +00:00
polyhedral b177a68bfb ScatterMax support dynamic shape 2022-05-05 15:24:28 +08:00
hujiahui8 9cc94628fa scatter_min: add c++ implementation to adapt to dynamic shape 2022-04-29 10:33:16 +08:00
zhujingxuan 9037c5496b Add half registration 2022-04-28 14:57:44 +08:00
i-robot 3e80716b0c
!33437 SactterMin operator supports GPU, and add function interface 
Merge pull request !33437 from hujiahui8/scatter_min
 2022-04-28 02:39:44 +00:00
i-robot b85647eead
!33626 index_add_gpu_kernel use new interface 
Merge pull request !33626 from looop5/index_add_gpu
 2022-04-28 02:26:24 +00:00
i-robot a312a1684b
!33602 add selu kernel for gpu backend. 
Merge pull request !33602 from zhuzhongrui/pub_master
 2022-04-27 18:08:57 +00:00
z00512249 228f63d5e2 fix matrix_determinant kernel for gpu backend st. 2022-04-27 17:55:56 +08:00
hujiahui8 b8ec35c3b7 SactterMin operator supports GPU, and add function interface 2022-04-27 16:42:21 +08:00
z00512249 65d0166abd add selu kernel for gpu backend. 2022-04-27 12:52:40 +08:00
looop5 7dc011e772 index_add_gpu_kernel use new interface 
fix compiling

fix dst_out

fix undefined symbol
 2022-04-27 11:45:28 +08:00
z00512249 653efcadf6 fix log_matrix_determinant accuracy for gpu backend. 2022-04-27 10:27:43 +08:00
i-robot 69d5bd0320
!33563 add log matrix_determinant and matrix_determinant kernel for gpu backend. 
Merge pull request !33563 from zhuzhongrui/pub_master
 2022-04-26 08:17:02 +00:00
i-robot c1707043f9
!33534 add dropout2d kernel for gpu backend. 
Merge pull request !33534 from zhuzhongrui/pub_master3
 2022-04-26 06:21:01 +00:00
z00512249 17e20ffde2 add log matrix_determinant and matrix_determinant kernel for gpu backend. 2022-04-26 11:51:18 +08:00
yanghaoran 73ecbea67b
回退 'Pull Request !33484 : add matrix_determinant kernel for gpu backend.' 2022-04-26 02:46:50 +00:00
i-robot 9812c1072e
!33457 SactterMax operator supports GPU, and add function interface 
Merge pull request !33457 from polyhedral/max
 2022-04-26 02:10:46 +00:00
i-robot 5cfdef1447
!33484 add matrix_determinant kernel for gpu backend. 
Merge pull request !33484 from zhuzhongrui/pub_master2
 2022-04-26 01:33:17 +00:00
z00512249 b7d31ad7c4 add dropout2d kernel for gpu backend. 2022-04-25 21:46:00 +08:00
polyhedral d4c415deae ScatterMax support GPU and add function interface 2022-04-25 18:46:59 +08:00
i-robot 329018a490
!33332 add MaskedFill for GPU 
Merge pull request !33332 from 范吉斌/masked_gpu
 2022-04-25 08:00:21 +00:00
z00512249 1883dbf573 add matrix_determinant kernel for gpu backend. 2022-04-25 14:40:01 +08:00
hezhenhao1 9f7db2afab Add CPU/GPU implementation of Cummax operator. 2022-04-25 12:32:29 +08:00
hezhenhao1 a5ae70d5fb Add GPU implementation of Cummin operator. 2022-04-24 21:50:50 +08:00
i-robot 662abf0f05
!33345 support AvgPool3DGrad for gpu 
Merge pull request !33345 from 范吉斌/avg3dgrad
 2022-04-24 09:03:19 +00:00
i-robot 0bd136473e
!33369 [MS][ops] add GPU operator FastGeLU FastGeLUGrad 
Merge pull request !33369 from KXiong/master
 2022-04-24 08:07:08 +00:00
fan-jibin 8040c330db add masked fill 2022-04-24 14:35:58 +08:00
fan-jibin 5c1768d246 add avgpool3dgrad for gpu 2022-04-23 18:37:43 +08:00
i-robot 19afa6fc9a
!33356 Add a st for the MishGrad kernel on cpu and gpu respectively. 
Merge pull request !33356 from liqiliang/mishgrad-cpu
 2022-04-22 09:18:24 +00:00
i-robot ae565ae5c9
!33325 add lpNorm gpu implement 
Merge pull request !33325 from zhuzhongrui/pub_master4
 2022-04-22 08:01:38 +00:00
xiongkun b15f56697b [MS][ops] add GPU operator FastGeLU GastGeLUGrad 
[MS][ops] add GPU operator FastGeLU GastGeLUGrad

[MS][ops] add GPU operator FastGeLU GastGeLUGrad

[MS][ops] add GPU operator FastGeLU GastGeLUGrad

[MS][ops] add GPU operator FastGeLU GastGeLUGrad

[MS][ops] add GPU operator FastGeLU GastGeLUGrad

[MS][ops] add GPU operator FastGeLU GastGeLUGrad

[MS][ops] add GPU operator FastGeLU GastGeLUGrad
 2022-04-22 11:30:33 +08:00
z00512249 8b3935ca63 add lpNorm gpu implement 2022-04-22 10:23:00 +08:00
ling 2128f94f38 TensorScatterDiv 2022-04-21 15:51:41 +08:00
liqiliang 78238f0165 Add a st for the MishGrad kernel on cpu and gpu respectively. 2022-04-21 15:07:58 +08:00
i-robot 17b6837e9d
!33161 The DynamicAttrIntValue suit for new Interface 
Merge pull request !33161 from hanhuifeng/depend_sample
 2022-04-21 02:03:13 +00:00
i-robot f4f8a31ece
!33156 Support DenseToCSR GPU Kernel 
Merge pull request !33156 from 杨林枫/dense_to_csr
 2022-04-21 01:43:32 +00:00
yanglf1121 946f60e60e add dense_to_csr_matrix 2022-04-20 17:30:31 +08:00
hanhuifeng2020 dc7af49d84 The DynamicAttrIntValue suit for new interface 2022-04-20 16:44:14 +08:00
i-robot 76948bf439
!33281 Add a gpu kernel, Invert. 
Merge pull request !33281 from liqiliang/invert-gpu
 2022-04-20 01:21:21 +00:00
zhujingxuan 0b4907082a Add ScatterNdMin GPU kernel 2022-04-19 16:49:46 +08:00
liqiliang 1a33cd84b3 Add a gpu kernel, Invert. 2022-04-19 15:16:09 +08:00
hezhenhao1 cae3b8e5ca Fix cuda implementation of IndexFill operator. 2022-04-18 20:10:44 +08:00
i-robot 88b9151867
!33084 Add GPU implementation of ScatterNdMax operator. 
Merge pull request !33084 from hezhenhao1/add_scatter_max
 2022-04-16 07:26:55 +00:00
hezhenhao1 09e5b2a2a4 Add GPU implementation of ScatterNdMax operator. 2022-04-16 11:44:23 +08:00
liqiliang b75a0a2021 Add a gpu kernel, InvGrad. 2022-04-16 10:37:07 +08:00
i-robot db3001e9fb
!33054 Add a gpu kernel, Inv. 
Merge pull request !33054 from liqiliang/inv-gpu
 2022-04-15 06:42:24 +00:00
i-robot c52ef8ed33
!32970 Add GPU implementation of ScatterNdMul operator. 
Merge pull request !32970 from hezhenhao1/add_scatter_mul
 2022-04-15 03:45:45 +00:00
liqiliang 9a91f38f93 Add a gpu kernel, Inv. 2022-04-14 21:22:23 +08:00
hezhenhao1 6c453ba65f Add GPU implementation of ScatterNdMul operator. 2022-04-14 10:36:08 +08:00
i-robot 2b1d714550
!32932 Add GPU implementation of IndexFill operator. 
Merge pull request !32932 from hezhenhao1/add_index_fill
 2022-04-13 12:25:13 +00:00
hezhenhao1 7ca2564730 Add GPU implementation of IndexFill operator. 2022-04-13 16:51:14 +08:00
zhujingxuan f56410af46 Add Cosh GPU Kernel 2022-04-13 10:34:08 +08:00
i-robot 9a8d775dd3
!32610 add tensor_scatter_arithmetic_binary_op st 
Merge pull request !32610 from zhuzhongrui/pub_master2
 2022-04-08 06:07:44 +00:00
z00512249 fa226a8003 add tensor_scatter_arithmetic_binary_op st 2022-04-07 09:58:04 +08:00
zhujingxuan 5303f20e89 Refactor testcases for scatter_nd. 2022-04-07 09:16:58 +08:00
hanhuifeng2020 790118a4e8 [Dynamic][Gpu]fix some dynamic operators and add test cases 2022-03-29 15:31:57 +08:00
hanhuifeng2020 2de960a637 [DynamicShape]fix some problems of InferShapeForNopNode 2022-03-15 19:47:58 +08:00
lianliguang 55e7f2d761 modify DynamicShape to TensorShape 2022-03-07 14:19:12 +08:00
jiangzhenguang 64da650786 add half pixel center to resize bilinear 2022-03-05 15:43:20 +08:00
hanhuifeng2020 b571e67795 [DynamicShape][GPU]add dynamic shape support some ops for new network 2022-02-28 22:09:20 +08:00
zong-shuai 5547c8620c debug 2022-02-25 17:41:25 +08:00
hanhuifeng2020 662c51c019 [DynamicShape][GPU]add dynamic shape support of Concat and its backward for DCN 2022-02-23 19:18:21 +08:00
zhaoting b2b74d8ab2 add MaxPoolGradGrad on GPU 2022-02-22 09:49:37 +08:00
VectorSL 8b45b00f7c update tensorarry stack for static usage 2022-02-14 09:14:35 +08:00
zong-shuai 7e0cc63eb1 einsum operator 2022-01-17 16:55:58 +08:00
王南 eab11f8677 functional select support x or y is a scalar 2022-01-05 10:12:44 +08:00
VectorSL c84ef13e89 del dynamic reshape test file 2021-12-24 15:45:40 +08:00
i-robot 7d274f1702 !27276 GPU add dynamic kernels 
Merge pull request !27276 from VectorSL/gpu-dynmic-kernel-for-merge
 2021-12-23 01:00:41 +00:00
i-robot deea2c9b6b !26942 unify GRU and LSTM operaters and change LSTMCell to right version 
Merge pull request !26942 from 吕昱峰（Nate.River）/rnn_op
 2021-12-14 16:12:06 +00:00
i-robot a4cf337666 !27331 Add unify ClipGradNorm op on GPU 
Merge pull request !27331 from JoyLvliang/add_unify_clip_grad_norm_op_on_gpu
 2021-12-14 03:32:17 +00:00
VectorSL 91766dbcd8 add bool for tesorarray 2021-12-13 15:07:51 +08:00
7347157+joylvliang@user.noreply.gitee.com 7a9fd2d7df add_unify_clip_grad_norm_op_on_gpu 2021-12-13 11:40:57 +08:00
VectorSL b17d0a08c9 add dynamic kernels 2021-12-09 19:42:29 +08:00
lvyufeng 7de53ec3fe unify GRU and LSTM operaters and change LSTMCell to right version 2021-12-08 16:51:51 +08:00
z00512249 a1b8c2f065 while inplace op for cpu backend 2021-12-06 17:41:40 +08:00
i-robot 9cb5120594 !26936 move testcases to scipy_st 
Merge pull request !26936 from zhujingxuan/move_tesecases
 2021-12-01 01:23:57 +00:00
huangbingjian 66309632c1 fix the negative axis problem of reduce_eliminate 2021-11-30 13:11:01 +08:00
zhujingxuan a18b9e703e move testcases 2021-11-30 10:52:02 +08:00
VectorSL 710289a72d add tensor array 2021-11-24 16:12:04 +08:00
i-robot 30d182ac18 !26626 fix reduce ops axis multiple bug in GPU 
Merge pull request !26626 from zhangbuxue/fix_reduce_ops_axis_multiple_bug_in_GPU
 2021-11-24 06:15:10 +00:00
buxue 89a688f3be fix reduce ops axis multiple bug in GPU 2021-11-23 16:39:41 +08:00
hezhenhao1 accc6368aa Add support float64 as input type for ReduceProd GPU op. 2021-11-23 11:44:40 +08:00
wenbean 13409f519f Unify GPU/CPU ops input/output(col/rolmajor), modify related testcases, add linalg function and testcases 2021-11-19 14:40:27 +08:00
wenbean 9b305b231d Add GPU eigenvalues/eigenvector for symmetric mtrix(real and complex) 
add cuda ops real to complex(imag is 0j)
cpu/gpu eigh ops alreas returns complex eigenvalues for complex input
retuns real eigenvalues for real input
 2021-11-18 11:23:36 +08:00
i-robot dc5278516c !26130 add cholesky, cho_factor primitive and backend gpu implements 
Merge pull request !26130 from zhuzhongrui/gmres
 2021-11-16 02:37:32 +00:00
z00512249 36032e7ee2 add cholesky, cho_factor primitive and backend gpu implements 2021-11-15 21:19:56 +08:00
zuochuanyong 1f289803e9 fix LessEqual bug 2021-11-14 14:07:15 +08:00
hezhenhao1 cb13783059 Add fp64 as input type fot GPU op of ReduceMin and ReduceMean. 2021-11-08 17:15:32 +08:00
i-robot b6833ec3f2 !25872 add solve_triangular 
Merge pull request !25872 from zhujingxuan/solve_triangular
 2021-11-05 06:15:35 +00:00