Commit Graph

3732 Commits

Author SHA1 Message Date
i-robot d66af6eb86
!26399 [assistant][ops] Add CropAndResizeGradBoxes
Merge pull request !26399 from wangyantao/CropAndResizeGradBoxes
2022-03-26 09:41:38 +00:00
lianliguang 6304c2ef3d support str.format 2022-03-26 14:53:06 +08:00
lianliguang 424c16d116 fix abs(tensor) when tesnor data is a 0-dim number 2022-03-26 14:20:03 +08:00
wangjun 789539cbaa modify interface name for shard 2022-03-26 09:10:14 +08:00
i-robot 3f18180606
!31742 [MD][Autotune] Change autotune config
Merge pull request !31742 from harshvardhangupta/at_config_suggestion
2022-03-25 21:00:12 +00:00
wangyantao 668cdb936f [feat] [assistant] [I471CP] add new operator CropAndResizeGradBoxes 2022-03-26 00:30:34 +08:00
ZCX 0cb72a21f3 [feat] [assistant] [I471DI] add new Ascend operator NonDeterministicInts 2022-03-25 22:29:47 +08:00
harshvardhangupta 0086f4ee05 Modify the Dataset AutoTune config #31742 2022-03-25 09:03:38 -04:00
i-robot 40015e003f
!31896 fix isolated node error
Merge pull request !31896 from lianliguang/fix-isolated-node-bug
2022-03-25 11:52:21 +00:00
lianliguang 3d7c56c4cc fix isolate node bug when using fallback interpret node 2022-03-25 16:41:16 +08:00
i-robot 1af8b365de
!31880 [MD] add OBSMindDataset
Merge pull request !31880 from liyong126/add_obs_mindrecord_dataset
2022-03-25 06:50:58 +00:00
i-robot 5f343bcd09
!31824 add waring suggestion log and map op error check
Merge pull request !31824 from guozhijian/fix_more_info
2022-03-25 03:24:07 +00:00
liyong 419109cf0a add OBSMindDataset 2022-03-25 11:03:53 +08:00
i-robot 3c83d3f600
!31668 [Auto-Par] [D-Rec] Modify the name of the functions and variables of Parameter shared User strategy treatment & Add new treatment case of MatMul/Gather
Merge pull request !31668 from 王浩然/D-Rec-deliver
2022-03-25 02:11:41 +00:00
jonyguo 317c48a933 add log warning and op error check 2022-03-25 09:26:29 +08:00
7347157+joylvliang@user.noreply.gitee.com 39e70e8519 fix_bug_of_tensor_copy_canot_work_in_int64_with_D 2022-03-24 20:47:42 +08:00
haoran.wang fdfbe2dedc Modify the name of the funtions and variables of Parameter shared User strategy treatment 2022-03-24 06:16:55 +00:00
liyong bff0fbc76f fix GIL acquire Misuse 2022-03-24 08:45:30 +08:00
i-robot 4110c3d6d6
!31646 [dataset][bugfix] 修复config 接口typeerror的问题
Merge pull request !31646 from xiefangqi/md_fix_interface_comments
2022-03-23 02:45:01 +00:00
i-robot f59de99ed6
!31654 [ME]Support raise in constant scenarios.
Merge pull request !31654 from Margaret_wangrui/raise
2022-03-23 02:28:37 +00:00
Margaret_wangrui 6d10adc9c5 [ME]Support raise in constant scenarios. 2022-03-22 19:33:47 +08:00
i-robot 200a77c3d1
!31393 fix csr/coo docs and checks
Merge pull request !31393 from wangrao124/pr_0316
2022-03-22 05:15:38 +00:00
xiefangqi f81c7967ec fix config interface comments 2022-03-22 10:04:12 +08:00
i-robot 4bb2fbdb5f
!30695 Support third-party modules
Merge pull request !30695 from huangbingjian/import_module
2022-03-21 08:36:39 +00:00
huangbingjian a62484c5ad Added recognition of third-party libraries 2022-03-21 14:10:23 +08:00
i-robot 91d9585df7
!31346 add vmap ut and st
Merge pull request !31346 from Erpim/vmap_v12
2022-03-21 03:49:08 +00:00
Erpim 0e7e1f8400 add vmap st and ut 2022-03-21 09:38:38 +08:00
liyong 00371d523f fix mindrecord issue 2022-03-19 12:04:27 +08:00
wangrao124 5e5415f617 fix csr/coo 2022-03-18 16:46:38 +08:00
i-robot 2e516c735b
!28640 [assistant] [ops] Add the FillV2 operator
Merge pull request !28640 from 爱科学的小松鼠/OpFill
2022-03-17 09:48:14 +00:00
hu-daiwang 3b093696e7 [feat] [assistant] [I48O6L] add the fillV2 operator 2022-03-17 11:19:59 +08:00
jonyguo 11079eabbb fix: code check 2022-03-17 08:23:12 +08:00
i-robot 1d6834ce4d
!31334 [MD][Autotune] Modify the generated autotune config to improve readability
Merge pull request !31334 from harshvardhangupta/modify_at_config
2022-03-16 18:18:57 +00:00
i-robot 88691d054e
!31321 Convert AbstractList to AbstractTuple recursively
Merge pull request !31321 from hewei/fix_opt_perf
2022-03-16 07:09:52 +00:00
i-robot c2212f88b4
!31164 Fix the global norm missing insert allreduce
Merge pull request !31164 from huangxinjing/fx_global_norm_error
2022-03-16 06:46:13 +00:00
i-robot 72ea797e4d
!30339 [BUG][MD][FUNC]RandomInvert
Merge pull request !30339 from yangwm/randominvert
2022-03-16 06:39:53 +00:00
i-robot 8e8391dadb
!30429 BUG][MD][FUNC]RandomAutoContrast
Merge pull request !30429 from yangwm/autocontrast
2022-03-16 06:38:54 +00:00
He Wei 51c147ab12 Convert AbstractList to AbstractTuple recursively 2022-03-16 09:59:27 +08:00
i-robot bf03f0e030
!31252 Implementation of element wise parallel ops
Merge pull request !31252 from liuluobin/element_wise_ops
2022-03-16 01:32:29 +00:00
i-robot 5f75c701d5
!31332 Fix bug in terminating a failed pipeline
Merge pull request !31332 from h.farahat/fix_reset
2022-03-15 20:17:05 +00:00
harshvardhangupta accc32da6c modify the at config file 2022-03-15 13:49:00 -04:00
i-robot 12d8906e2e
!30253 Support more ops for dump flag
Merge pull request !30253 from huanghui/enhance-dump-flag
2022-03-15 16:49:11 +00:00
hesham 3d16a4dee2 Fix bug in terminating a failed pipeline 2022-03-15 11:35:28 -04:00
liuluobin f13d342986 Implementation of element wise parallel ops 2022-03-15 19:24:17 +08:00
i-robot 313a247ae8
!31178 upgrade Ascend package 11 Mar 22
Merge pull request !31178 from shenwei41/upgrade_ascend_20220311
2022-03-15 11:08:32 +00:00
i-robot 3be552ca39
!31156 test centos ut
Merge pull request !31156 from luoyang/test_centos
2022-03-15 09:38:15 +00:00
shenwei41 f1fde6a262 upgrade Ascend package 11 Mar 22 2022-03-15 16:31:27 +08:00
huangxinjing 31f55b6525 1. The main gol: Fix mixing inserting the AllReduce when where is no mirror appeared
2. remove pattern match error as the origin pattern match will find no operator if there is only one parameter
2022-03-15 15:54:06 +08:00
i-robot 01078763df
!31227 don't handle reshape when using adafactor optimizer in auto parallel
Merge pull request !31227 from yangzhenzhang/adafactor-parallel-skip-handle-reshape
2022-03-15 02:47:31 +00:00
huanghui 4482acf586 Support API set_dump for more type of ops 2022-03-15 10:24:15 +08:00
i-robot d4fc47ef0c
!31211 Clean iterators before forking
Merge pull request !31211 from h.farahat/pytest_problem
2022-03-15 01:47:28 +00:00
yangzhenzhang 1f98ffb79c adafactor parallel skip handle reshape 2022-03-15 09:18:19 +08:00
i-robot 6be7c42052
!31191 [MD] more tests for failover reset
Merge pull request !31191 from Mohammad Motallebi/failover_more_tests
2022-03-14 21:50:37 +00:00
h.farahat 318e6b8803 fix segfault 2022-03-14 14:55:21 -04:00
mohammad 583eb32723 [MD] Add more UT and a ST for failover reset 2022-03-14 13:52:09 -04:00
luoyang 74d8969b91 ut failed with pillow>=9.0.0, need more validation 2022-03-14 23:58:09 +08:00
i-robot dcb5cd670c
!30953 Dynamic Weight Decay
Merge pull request !30953 from wanyiming/dynamic_wd
2022-03-14 06:03:24 +00:00
lilei 690c58ebcf modify virtualdataset bug for master 2022-03-13 10:49:15 +08:00
i-robot 8b7d7a1e1a
!31141 [MD][Offload] Offload RandomSharpness Op Pure Sharpness Support & UT
Merge pull request !31141 from alashkari/fix-random-sharpness
2022-03-11 21:06:13 +00:00
Rescue face5e51a0 Added sharpness without randomness support and sharpnes op ut.
fix pad issue
2022-03-11 17:30:49 +00:00
i-robot 603ff25afb
!31177 Add ut for fallback python buildin function with tensor input
Merge pull request !31177 from LiangZhibo/syntax
2022-03-11 08:38:53 +00:00
i-robot 3c0e34ada0
!30855 Support user-defined classes by ms_class decorators
Merge pull request !30855 from huangbingjian/ms_class_dev
2022-03-11 08:37:12 +00:00
liangzhibo a3f183bdad Add test case for fallback buildin function with constant tensor 2022-03-11 15:35:29 +08:00
i-robot c2a5cc1486
!31040 Produce parallel operators for Argmin/max, SquareSumAll and UnsortedSegmentProd
Merge pull request !31040 from Bert0108/reduce_operators_arg
2022-03-11 06:15:56 +00:00
shenyu 443da2280d [FIX][BUG][MD][FUNC] RandomInvert 2022-03-11 11:53:04 +08:00
shenyu c14a8084c7 [FIX][BUG][MD][FUNC] RandomAuoContrast 2022-03-11 10:29:49 +08:00
i-robot a064d0855b
!30728 add ci test for unify backend
Merge pull request !30728 from xiaoyao/master
2022-03-11 01:50:03 +00:00
i-robot eecf6db426
!31121 [MD] revert OBSMindDataset
Merge pull request !31121 from liyong126/revert_OBSMindDataset
2022-03-11 01:21:10 +00:00
i-robot 216e7c6a92
!31041 add check for conv2d
Merge pull request !31041 from yangzhenzhang/add-check-for-conv2d
2022-03-11 01:20:52 +00:00
i-robot 67d10ce3be
!31066 Fix a bug where ROIAlign and CropAndResize distributed op do not support GPU
Merge pull request !31066 from liuluobin/fix_roialign
2022-03-11 01:16:46 +00:00
huangbingjian 4f7c9cb27f Support user-defined classes through ms_class decorators. 2022-03-10 20:12:54 +08:00
i-robot 99910ef27f
!31088 convert C++ data to python when run fallback function
Merge pull request !31088 from lianliguang/add-fallback-builtin-function
2022-03-10 11:56:01 +00:00
yangzhenzhang c00d29f223 rebase 2022-03-10 19:17:06 +08:00
liyong 50fc6f5736 Revert "!30810 [MD] add OBSMindDataset "
This reverts commit 9fcab9184e, reversing
changes made to afb33fff62.
2022-03-10 19:10:11 +08:00
liuluobin 8f045d02e3 Fix a bug where ROIAlign and CropAndResize distributed op do not support GPU 2022-03-10 17:31:36 +08:00
lianliguang d9390afad1 convert MS C++ dtype to MS python 2022-03-10 16:09:34 +08:00
xiao_yao1994 94df0869ad uni_backend: add ut ge test 2022-03-10 16:08:36 +08:00
Bert0108 bfc5e4345c add distributed operators for argmax/min sqauresumall and unsortedsetmentprod 2022-03-10 14:44:37 +08:00
liyong 10ada68fa9 add OBSMindDataset 2022-03-10 11:18:59 +08:00
i-robot 059ac45ae8
!30920 Call Draw function directly
Merge pull request !30920 from huanghui/fix-draw
2022-03-10 03:15:56 +00:00
wanyiming a124ec4de7 add dynamic_decay 2022-03-10 11:02:27 +08:00
i-robot b32daf3cb4
!31000 add fallback example of function
Merge pull request !31000 from lianliguang/add-fallback-builtin-function
2022-03-10 01:57:37 +00:00
i-robot 698b99aaa3
!18945 [assistant][ops]New operator implementation, include OmniglotDataset
Merge pull request !18945 from Wangsong95/omniglot_dataset
2022-03-09 06:51:15 +00:00
lianliguang f3b1e3291c add jit fallback example 2022-03-09 11:25:03 +08:00
qkeys ce7597f500 [feat][assistant][I3J6VB] add new data operator OmniglotDataset 2022-03-08 21:36:22 +08:00
i-robot 0b207fce31
!29735 unique support multi_dim tensor
Merge pull request !29735 from jiangzhenguang/unique_multi_tensor
2022-03-08 08:30:57 +00:00
i-robot f73987dcf2
!30949 Change the Parameter type name from 'ParameterTensor' to 'Parameter'
Merge pull request !30949 from 张清华/opt
2022-03-08 06:17:31 +00:00
jiangzhenguang 28f89f36b4 unique support multi-dim tensor 2022-03-08 14:09:19 +08:00
Zhang Qinghua 973008ebcf Change the Parameter type name from 'ParameterTensor' to 'Parameter' 2022-03-08 11:23:56 +08:00
i-robot 82bc2b722d
!30417 add history and lambda callbacks
Merge pull request !30417 from liutongtong9/add_callbacks
2022-03-08 02:17:46 +00:00
i-robot 75e1939f50
!30899 Enable python buildin function type for graph fallback and add some ut for python buildin function for graph fallback
Merge pull request !30899 from LiangZhibo/syntax
2022-03-08 01:35:47 +00:00
i-robot 2655d64720
!30865 [MD] Modify Batch number of workers with Python multiprocessing
Merge pull request !30865 from cathwong/ckw_batch_workers_pymultip
2022-03-07 18:21:16 +00:00
Cathy Wong bc29dd9d72 [MD] Modify Batch number of workers with Python multiprocessing 2022-03-07 10:39:05 -05:00
i-robot 335ef1c270
!30459 Add ut validate function for parallel
Merge pull request !30459 from liuluobin/ut_master
2022-03-07 15:23:14 +00:00
i-robot 4a9a65d2a6
!30919 Produce parallel operators for ReduceAll and ReduceProd
Merge pull request !30919 from Bert0108/reduce_operators_merge
2022-03-07 11:30:37 +00:00
liuluobin b797a410cc Add validate function for parallel ut 2022-03-07 16:19:52 +08:00
huanghui ce0f500a3f convert draw_ to call Draw function derectly 2022-03-07 16:12:52 +08:00
Bert0108 dfc92f1791 add distributed parallel operators for reduceall and reduceprod 2022-03-07 15:40:11 +08:00
liangzhibo 1a81031d92 Add testcases for fallback python buildin function 2022-03-07 14:20:24 +08:00
lianliguang 55e7f2d761 modify DynamicShape to TensorShape 2022-03-07 14:19:12 +08:00
i-robot a2303a92bd
!30741 [dataset][dfx] 易用性-Python多进程统一整改专项
Merge pull request !30741 from xiefangqi/md_add_multiprocessing_optimization
2022-03-07 01:29:21 +00:00
i-robot c4cc0e6f26
!30742 auto parallel support not ony power of 2 device numbers
Merge pull request !30742 from yao_yf/auto_parallel_support_not_only_2_power
2022-03-07 01:09:01 +00:00
yao_yf b60e54e0d5 support not only power of 2 2022-03-06 18:08:46 +08:00
xiefangqi 2de6c8908b add a clean up subprocess to clean subprocess when main process exit 2022-03-05 17:24:11 +08:00
liutongtong 7afcdfd211 add history and lambda callbacks 2022-03-04 17:24:57 +08:00
i-robot b90cf43562
!30553 Support dataset reset() to recover after failure
Merge pull request !30553 from h.farahat/reset
2022-03-04 02:19:45 +00:00
i-robot f01d841113
!27943 [MD][Autotune] Save/Load Autotune config
Merge pull request !27943 from harshvardhangupta/save_load_at_config
2022-03-03 20:42:23 +00:00
i-robot 7b26a32e98
!22608 [assistant][ops]New operator implementation, include KITTIDataset
Merge pull request !22608 from Wangsong95/kitti_dataset
2022-03-03 07:05:40 +00:00
h.farahat a3dec34833 Dataset failover reset 2022-03-02 23:53:56 -05:00
harshvardhangupta bd61adbb72 Implement save/load of autotune dataset pipeline configuration #27943 2022-03-02 15:08:08 -05:00
i-robot 70e61e9014
!30781 [MD] test_pyfunc_multiprocess_autotune.py - enable AutoTune
Merge pull request !30781 from cathwong/ckw_ut_fixup_map_python_multiproc
2022-03-02 19:50:29 +00:00
i-robot edcc6b790d
!30714 [MD][Offload] Add TypeCast op to offload
Merge pull request !30714 from markuskunej/offload_typecast
2022-03-02 19:48:58 +00:00
Cathy Wong eb931c7f88 [MD] test_pyfunc_multiprocess_autotune.py - enable AutoTune 2022-03-02 13:25:35 -05:00
i-robot 4aa82cc21e
!30679 [MD] Add tests for Python Multiprocessing with AutoTune
Merge pull request !30679 from cathwong/ckw_ut_map_python_multiproc
2022-03-02 14:26:09 +00:00
zx 93617ce91e [feat][assistant][I3J6VO] add new data operator KITTI 2022-03-02 19:29:17 +08:00
Cathy Wong b151f90b8d [MD] Add tests for Python Multiprocessing with AutoTune 2022-03-01 20:20:59 -05:00
markuskunej 6de37045ff Added TypeCast op in dataset offload. 2022-03-01 21:55:27 +00:00
wangjun 46612fabfb add st for shard 2022-03-01 15:32:37 +08:00
i-robot ad9757ccf0
!30661 [Auto parallel] [MoE] Fix an error of configuring MoE parallel
Merge pull request !30661 from Xiaoda/124-moe-changes
2022-03-01 02:26:37 +00:00
i-robot 789c1d6bd3
!30614 [AutoParallel] Fix Overflow As the cast is inserted before Mirror
Merge pull request !30614 from huangxinjing/fix_cast_error
2022-02-28 09:04:46 +00:00
Xiaoda Zhang 81e5abe580 fix an error of configuring parallel 2022-02-28 17:03:53 +08:00
i-robot a92c54b206
!30496 [Fallback] Suppport scipy module.
Merge pull request !30496 from huangbingjian/support_scipy
2022-02-28 08:33:34 +00:00
i-robot a8686ae3d9
!18827 [assistant][ops]New operator implementation, include LFWDataset
Merge pull request !18827 from Wangsong95/lfw_dataset
2022-02-28 08:31:11 +00:00
zx 2132f62d98 [feat][assistant][I3J6VQ] add new data operator LFW 2022-02-28 13:43:05 +08:00
huangxinjing 896daee845 [AutoParallel]Fix insert error for the mirror 2022-02-26 17:54:42 +08:00
yangzhenzhang 43e6e16da3 check platform for resizebilinear 2022-02-26 09:11:32 +08:00
i-robot 5deccfe64b
!30260 [MD][AutoTune] Re-enable AT for non-sink models
Merge pull request !30260 from danishfarid/re-enable_AT
2022-02-25 22:26:14 +00:00
i-robot 0341d96dd6
!30469 add shard function to support part of the graph executed in auto_parallel under pynative mode
Merge pull request !30469 from wangjun/0223_pp
2022-02-25 06:52:24 +00:00
huangbingjian a69d13bc44 [Fallback] Suppport scipy module. 2022-02-25 10:13:43 +08:00
i-robot cfe0f76d2b
!30491 ut for allgather fusion
Merge pull request !30491 from jiahongQian/master
2022-02-25 01:30:24 +00:00
danishfarid 6c4697fc8b reenable AT for non-sink
tests re-enable

code check fix 1

remove self

lint fixing

lint fix 2

remove test as per req
2022-02-25 00:02:58 +00:00
wangjun 24d448239c add pynative_parallel 2022-02-24 21:07:51 +08:00
i-robot 981eae461a
!30118 自动优化器并行特性
Merge pull request !30118 from zhuyuxiao/I4S85V
2022-02-24 11:09:44 +00:00
jiahongQian 25f57505bf ut for allgather fusion 2022-02-24 17:40:32 +08:00
i-robot bbcfbce9e0
!29997 [Auto parallel] [MoE] Support data_parallel + expert_parallel
Merge pull request !29997 from Xiaoda/124-moe-changes
2022-02-24 09:23:47 +00:00
zhuyuxiao d0e0e305d3 good 2022-02-24 16:10:18 +08:00
i-robot f2130e7434
!30483 [AutoParallel]Pipeline Automatic detection Opt
Merge pull request !30483 from lichen/pipeline_opt_detection
2022-02-24 03:40:15 +00:00
yao_yf e21f878e14 adasum ut fix 2022-02-24 09:31:04 +08:00
Xiaoda Zhang b714451937 implementing expert_parallel+data_parallel in MoE:
1) extending _Linear's input as 4-dimension tensor: [outer_batch, expert_dim, -1, hidden], and _Liner's BatchMatMul becomes BatchMatMul(4_dim_tensor, 3_dim_tensor);
2) configuring the _Linear's BatchMatMul sharding strategy as [[dp, ep, 1, 1], [ep, 1, mp]];
3) introducing a new parameter 'expert_parallel' in TransformerOpParallelConfig, creating a new class MoEParallelConfig to include 'data_parallel', 'model_parallel' and 'expert_parallel';
4) changing parallel config for FeedForward, TransformerEncoderLayer, TransformerDecoderLayer.
2022-02-24 09:24:27 +08:00
wangshengnan12@huawei.com acbefd80ea pipeline_opt_detection 2022-02-24 09:23:17 +08:00
i-robot 2a00ffd3b1
!30259 [MD]Update set_autotune_enable API to add save filepath
Merge pull request !30259 from cathwong/ckw_at_save_api
2022-02-24 01:00:59 +00:00
Cathy Wong 46e223e569 [MD] Update set_autotune_enable API to add save filepath 2022-02-23 14:50:55 -05:00
i-robot 81260a2319
!30466 takedown test_auto_parallel_adasum.py to ensure stability, again
Merge pull request !30466 from yanghaoran/master
2022-02-23 09:15:58 +00:00
i-robot 14393503b7
!30431 allreduce allgather fusion
Merge pull request !30431 from jiahongQian/master
2022-02-23 08:52:40 +00:00
yanghaoran 71d6b7d506 takedown test_auto_parallel_adasum.py to ensure stability, again 2022-02-23 16:42:13 +08:00
i-robot 2e8eac8341
!30367 auto_parallel_adasum_support_data_parallel
Merge pull request !30367 from yao_yf/auto_parallel_adasum_support_data_parallel
2022-02-23 06:50:26 +00:00
jiahongQian 8a2151d8bb allgather reducescatter fusion 2022-02-23 10:47:30 +08:00
i-robot eeb731ae3e
!18738 [assistant][ops]New operator implementation, include LSUNDataset
Merge pull request !18738 from Wangsong95/lsun_dataset
2022-02-23 02:28:40 +00:00
i-robot 5bee7156b9
!30369 add_virtualdataset_ut
Merge pull request !30369 from lilei/add_virtualdataset_ut
2022-02-23 02:21:34 +00:00
i-robot 0f24b679ec
!29819 Add GlobalNorm Search
Merge pull request !29819 from huangxinjing/add_global_norm_search
2022-02-23 01:15:08 +00:00