Commit Graph

51519 Commits

Author SHA1 Message Date
zhangxuetong fcb0a7057c backend decoupling 2022-08-31 16:11:42 +08:00
i-robot 1edd58625f
!40834 Fix getattr, hasattr problem.
Merge pull request !40834 from LiangZhibo/null
2022-08-29 02:38:30 +00:00
i-robot aaeae5d3ae
!40723 rename DeviceQueueOp=>DataQueueOp, TransferNode=>DataQueueNode
Merge pull request !40723 from zhoufeng/delete-macro
2022-08-29 02:24:58 +00:00
i-robot 4d52420798
!41025 support dyn rank in SpaceToBatchND
Merge pull request !41025 from zichun_ye/op_dev
2022-08-29 02:22:58 +00:00
i-robot 3a07dc6315
!41034 add adaptive_avg_pool3d to en rst
Merge pull request !41034 from 范吉斌/code_docs_adaptive_avg_pool3d
2022-08-29 02:13:18 +00:00
i-robot 0c24957987
!40977 fix nll_loss and cross_entropy
Merge pull request !40977 from 吕昱峰(Nate.River)/master
2022-08-29 02:08:59 +00:00
i-robot 63c2d7b9e3
!40777 use c++ split on Ascend
Merge pull request !40777 from looop5/split_ascend_commit
2022-08-29 01:58:47 +00:00
fan-jibin 1bfeebbd6f add adaptive_avg_pool3d to en rst 2022-08-29 09:40:07 +08:00
i-robot bfb0a7d54a
!40821 Purge old ps arch
Merge pull request !40821 from ZPaC/purge-old-ps
2022-08-29 01:32:38 +00:00
i-robot f49bef6ca1
!40669 add adaptive_avg_pool3d interface
Merge pull request !40669 from 范吉斌/adaptive_avg_pool3d
2022-08-29 01:32:15 +00:00
i-robot c21a7613ca
!40661 intopk gpu inherit from NativeGpuKernelMod
Merge pull request !40661 from 范吉斌/intopk_gpu
2022-08-29 01:30:30 +00:00
i-robot 33dd1fa1ea
!41012 modify the inconsistence on webpage 0827
Merge pull request !41012 from 宦晓玲/code_docs_08277
2022-08-29 01:28:48 +00:00
i-robot ccf3db786c
!40668 move ascend const input to attr pass to unify mindir
Merge pull request !40668 from laiyongqiang/new_pass
2022-08-29 01:25:58 +00:00
i-robot 0d6581c54f
!41007 kernel actor support the somas for intergration of dynamic and static memory
Merge pull request !41007 from limingqi107/new_actor_runtime
2022-08-29 01:24:43 +00:00
i-robot 7cffdeeb8c
!40496 [MS]{Lite][Task] delete no need lite lib for litert
Merge pull request !40496 from 刘力力/master_develop
2022-08-29 01:21:20 +00:00
i-robot 1483877469
!40852 adapt ms.log to windows
Merge pull request !40852 from chenminmin/master
2022-08-29 01:14:55 +00:00
zhangyi fb53daacf0 modify the inconsistence on webpage 0827 2022-08-29 08:57:22 +08:00
i-robot 0b71a52aea
!40893 add muxsend/receive prim
Merge pull request !40893 from VectorSL/add-prim-muxsend/recev
2022-08-28 11:23:24 +00:00
i-robot 96bd8af06a
!41004 gpu priority replay buffer support distributed training
Merge pull request !41004 from chenweifeng/prb-multi-process
2022-08-28 09:48:09 +00:00
limingqi107 a3a317050e kernel actor support the somas for intergration of dynamic and static memory 2022-08-28 11:32:26 +08:00
Zichun Ye 94f23e9e84 support dyn rank in spacetobatchnd
fix code check problem
2022-08-28 09:09:14 +08:00
i-robot c208e21fc5
!40317 [GraphKernel]add target info flags.
Merge pull request !40317 from DeshiChen/0812_cpu_info
2022-08-27 13:16:54 +00:00
i-robot ef91729227
!40998 Fix the mux send gpu kernel
Merge pull request !40998 from chengang/fix_mux_send
2022-08-27 12:59:52 +00:00
i-robot 2a0156ba21
!40980 Code self-review
Merge pull request !40980 from LiangZhibo/review
2022-08-27 12:42:29 +00:00
i-robot c474727607
!41005 optimize CTCLossV2 backward by thread parallelization and splitting kernels
Merge pull request !41005 from zhujingxuan/CTCLossGrad
2022-08-27 10:56:08 +00:00
taipingchangan 86efc91cb2 adapt log to windows 2022-08-27 18:43:22 +08:00
i-robot 555b628b2d
!41001 supplement qat weight quant param
Merge pull request !41001 from liyan2022/dev_qat_smoke
2022-08-27 10:19:06 +00:00
ZPaC 3bb04bb3f3 Purge old ps arch 2022-08-27 18:16:57 +08:00
i-robot 28226b4104
!40950 code clean
Merge pull request !40950 from yuchaojie/code-clean
2022-08-27 09:58:13 +00:00
i-robot 9b3f2f4050
!40952 cube vector split
Merge pull request !40952 from zhangyinxia/master
2022-08-27 09:55:08 +00:00
wilfChen 2c76db0e72 gpu priority replay buffer support distributed training 2022-08-27 17:54:31 +08:00
i-robot eba7decc38
!40911 support for windows msvc
Merge pull request !40911 from qiuzhongya/qiuzhongya_msvc4
2022-08-27 09:39:25 +00:00
i-robot 21d9238935
!40959 fix: add info comment for api
Merge pull request !40959 from guozhijian/fix_docs_usage
2022-08-27 09:37:42 +00:00
i-robot 3199b14175
!40903 fix code check
Merge pull request !40903 from laiyongqiang/code_check_19
2022-08-27 09:27:29 +00:00
LaiYongqiang e809917a50 move ascend const input to attr pass to unify mindir 2022-08-27 17:21:00 +08:00
i-robot 03f8889ef1
!40962 Layernorm Expander on Lite
Merge pull request !40962 from jiaoy1224/expand
2022-08-27 09:15:23 +00:00
i-robot 940bbf6597
!41010 takedown test_linalg.py::test_gmres_against_scipy to ensure gate stability
Merge pull request !41010 from yanghaoran/takedown_testcases
2022-08-27 09:06:12 +00:00
liu lili c589e7e0cb mindspore lite: delete no need lib for cpu litert 2022-08-27 17:03:52 +08:00
yanghaoran 1bea98c821 takedown test_linalg.py::test_gmres_against_scipy to ensure gate stability 2022-08-27 16:49:29 +08:00
looop5 51bf00831d use c++ split on Ascend 2022-08-27 16:48:31 +08:00
i-robot 78d6fb088d
!41006 skip repeatability ut temporary and wait for fix
Merge pull request !41006 from luoyang/skip
2022-08-27 08:43:43 +00:00
lvyufeng 29601c250a fix nll_loss and cross_entropy 2022-08-27 16:39:48 +08:00
YangLuo 87fd13c30d skip repeatability ut temporary and wait for fix 2022-08-27 16:34:46 +08:00
i-robot 0e25f1e2fc
!40523 Change specialize order of graph nodes
Merge pull request !40523 from chenfei_mindspore/change-specialize-order
2022-08-27 08:34:08 +00:00
i-robot 9772a743af
!40957 Disable async backend init by default
Merge pull request !40957 from hewei/fix_master1
2022-08-27 08:32:34 +00:00
i-robot e523715b7b
!40882 Add vmap for StridedSliceGrad
Merge pull request !40882 from OwenSec/master
2022-08-27 08:30:55 +00:00
i-robot e22c122959
!40760 [MS][lite] kronecker production decomposition warning fixed
Merge pull request !40760 from lz/pr_40736
2022-08-27 08:26:24 +00:00
VectorSL b6e66ae44e add prim for muxsend/recevive 2022-08-27 16:21:38 +08:00
zhujingxuan 2bc4ecb565 optimize CTCLossV2 GPU backward by thread parallelization and splitting kernels 2022-08-27 16:13:40 +08:00
zhangyinxia d749af230b cube and vector split 2022-08-27 15:52:55 +08:00