Commit Graph

6329 Commits

Author SHA1 Message Date
i-robot 8c6d4a05fc !22397 Add JVP for forward mode auto diff.
Merge pull request !22397 from LiangZhibo/fwd2
2021-08-26 11:28:22 +00:00
i-robot 20f25c497f !22400 Create new graph for pynative mode.
Merge pull request !22400 from gaoyong10/for_commit_0825
2021-08-26 11:05:12 +00:00
haizhouye d96d9e64b0 [feat] [assistant] [I3ALCA] add new LowpassBiquad operator
[feat] [assistant] [I3ALCA] add new LowpassBiquad operator

bugfix

fix

fix
2021-08-26 18:49:21 +08:00
i-robot 7083ec9d37 !21925 add pad mode to maxpool3d
Merge pull request !21925 from jiangzhenguang/amend_maxpool3d
2021-08-26 09:08:37 +00:00
l00591931 c07c286fcd Add forward mode auto diff 2021-08-26 16:35:41 +08:00
i-robot 8d00a8d803 !22360 Fix Transformer Mirror Error
Merge pull request !22360 from huangxinjing/fix_transformer_mirror_error
2021-08-26 08:16:33 +00:00
zhangyihui 3e5cb3b506 fix bugs for device_id_to_rank_id 2021-08-26 16:15:18 +08:00
i-robot 32281f84e7 !19000 update LayerNormGrad split pass to V2
Merge pull request !19000 from yuchaojie/ir_fusion2
2021-08-26 08:03:33 +00:00
i-robot 28e63116f2 !21921 Re-enabling offline debugger ut test
Merge pull request !21921 from sabrinasun_59ee/uttest
2021-08-26 07:50:01 +00:00
i-robot 471c57d3d7 !20046 [assistant][ops] Add New MaskedFill
Merge pull request !20046 from 孟权令/MaskedFill
2021-08-26 07:43:18 +00:00
i-robot a646eb27ab !21616 modify ms function cache key
Merge pull request !21616 from chujinjin/modify_ms_function_cache_key
2021-08-26 07:36:50 +00:00
liangyongxiong 44ad1cf23b add ST for RDR::TriggerAll() 2021-08-26 15:01:11 +08:00
i-robot 8e39dd4ec7 !22173 ascend add nontask sink mode
Merge pull request !22173 from baihuawei/graph_mode_nonsink_part3-2
2021-08-26 06:20:36 +00:00
i-robot 23a5c64ce0 !22218 Add graph kernel userdefine op support
Merge pull request !22218 from zichun_ye/graph_kernel_userdefine
2021-08-26 04:04:40 +00:00
i-robot 785e5fe6fd !22359 add st for pynative synchronize
Merge pull request !22359 from chujinjin/add_st_for_pynative_synchronize
2021-08-26 03:35:11 +00:00
gaoyong10 9fb5ce3b87 Create new graph for pynative mode. 2021-08-26 11:16:15 +08:00
i-robot 0faf36d5c5 !22187 modify api comment
Merge pull request !22187 from changzherui/mod_dir_auth
2021-08-26 03:01:33 +00:00
i-robot 07eaa1969b !22344 fix compile error when weight is used in while condition
Merge pull request !22344 from xychow/fix-while-loop-with-weight-in-condition
2021-08-26 03:00:38 +00:00
i-robot c710b57efe !22336 add rolling cpu kernel
Merge pull request !22336 from zhujingxuan/master
2021-08-26 02:38:12 +00:00
Xiaoda Zhang b2703879c6 fix the scope setting error when cloning nodes 2021-08-26 10:25:38 +08:00
i-robot edcbb68d71 !22386 fix neighborexchange empty input case
Merge pull request !22386 from zhoufeng/fix-neighbor-empty-input-bak
2021-08-26 01:52:10 +00:00
i-robot 5aeec398bb !22369 add test cases for conv2d
Merge pull request !22369 from yangzhenzhang/add-test-cases
2021-08-26 01:46:10 +00:00
i-robot 457a01fd09 !20055 [assistant][ops] Add math operator IndexAdd
Merge pull request !20055 from 孟权令/IndexAdd
2021-08-26 01:15:39 +00:00
sabrinasun a9c6bf45a3 re-enabling offline debugger ut test 2021-08-25 19:05:19 -04:00
changzherui 46319f3198 modify comment 2021-08-26 00:44:44 +08:00
i-robot 76a37daa43 !22358 thor generalization code submit
Merge pull request !22358 from wangshuangling/master
2021-08-25 13:46:27 +00:00
zhoufeng e5a1582e4b fix neighborexchange empty input case
Signed-off-by: zhoufeng <zhoufeng54@huawei.com>
2021-08-25 21:41:22 +08:00
huangxinjing 62496d75f3 less the interface exposed 2021-08-25 21:22:40 +08:00
i-robot 956ae2bbfd !22223 [assistant][TimeMask and FreqMask]
Merge pull request !22223 from QingfengLi/TimeFreqMask
2021-08-25 10:43:24 +00:00
i-robot acee9b24bc !22284 Change Op name in hccl to Op name in step trace
Merge pull request !22284 from 张毅辉/op_name_of_hccl_to_op_name_of_step_trace
2021-08-25 09:32:55 +00:00
chujinjin 5937d03355 add st for pynative synchronize 2021-08-25 17:27:54 +08:00
yangzhenzhang 0b9b2a9458 add test cases 2021-08-25 17:22:20 +08:00
i-robot 9b28bd6308 !22342 [Auto parallel] Adjust the device matrix of OneHot operator
Merge pull request !22342 from Xiaoda/85-adapt-onehot
2021-08-25 09:02:44 +00:00
jiangzhenguang 8a78b2b7f6 add pad mode to maxpool3d 2021-08-25 16:24:02 +08:00
i-robot 60634c5d11 !21745 Unified format of single machine and multi machine performance data directory
Merge pull request !21745 from 张毅辉/Device_id_to_rank_id
2021-08-25 07:49:29 +00:00
sl_wang 222b6684e0 second order optimizer thor generalization 2021-08-25 15:38:01 +08:00
doit a5193df648 [feat] [assistant] [I3T92B] add new array operator MaskedFill 2021-08-25 15:29:37 +08:00
i-robot 36238032a2 !22245 Add pynative switch layer ut test case
Merge pull request !22245 from zjun/switch_layer_ut_test
2021-08-25 07:24:28 +00:00
baihuawei a9694a9230 ascend add nontask sink mode 2021-08-25 14:45:58 +08:00
chujinjin 551fa9ab8b modify ms function cache key 2021-08-25 14:40:45 +08:00
Zichun Ye a7d89f6686 add graph kernel userdefine op support
fix code check
2021-08-25 14:37:28 +08:00
Xiaoda Zhang ae2d3b4832 adjust the device matrix for onehot use in moe 2021-08-25 14:36:06 +08:00
hulx c70731d601 [feat][assistant][I3T924] add new math operator IndexAdd 2021-08-25 14:19:17 +08:00
zhujingxuan ad389316cb add rolling cpu kernel 2021-08-25 11:57:18 +08:00
zhangyihui dab750d1a5 Mapping op_name of hccl to op_name of step trace 2021-08-25 10:51:52 +08:00
chenx2ovo 2f1d974eeb [fix] [assistant] [I465ZG] revise error message and change input param datatype 2021-08-25 10:47:27 +08:00
i-robot 698eb406e2 !20129 [assistant][ops] Add math operator Erfinv
Merge pull request !20129 from 孟权令/Erfinv
2021-08-25 02:44:18 +00:00
i-robot 2edaba38bf !21389 Add Tensor Base and Stat info to offline debugger
Merge pull request !21389 from parastooashtari/tensor_info_levels
2021-08-25 02:08:34 +00:00
i-robot 0adacd1b68 !22226 [AutoParallel]Refacor pipeline code
Merge pull request !22226 from lichen/Refactor_part_of_pipeline
2021-08-25 01:54:18 +00:00
zhangyihui 3d19949eb4 device_id to rank_id 2021-08-25 09:52:23 +08:00
zhousiyi 44bdcb101c fix while loop compile error when weight is used in while condition 2021-08-25 01:33:34 +00:00
i-robot abc9d8e6fe !22068 DynamicRNN supports scenarios that hidden_size is not multiples of 16
Merge pull request !22068 from yuchaojie/op_select2
2021-08-25 01:32:26 +00:00
chenfei 72cab0262e code clean 2021-08-25 09:29:01 +08:00
i-robot 4e81ea7bb0 !22300 Add ut and st case for non tensor input in pynative
Merge pull request !22300 from JoyLvliang/add_st_ut_case_for_pynative_non_tensor_input
2021-08-25 01:23:34 +00:00
Parastoo Ashtari ff83dee704 Add tensor base and stat info to offline debugger 2021-08-24 18:19:59 -04:00
i-robot 2cd255b54c !22311 When load mindir for ms_function, RuntimeError is thrown!
Merge pull request !22311 from lanzhineng/infer_optv3
2021-08-24 14:10:21 +00:00
lichenever 5812076512 Refactor_part_of_pipeline 2021-08-24 21:37:41 +08:00
hulx ca6f3ef155 [feat][assistant][I3T925] add new math operator Erfinv 2021-08-24 21:25:04 +08:00
i-robot 661c83e17f !22274 MD Profiling UT: Use pytest setup and teardown
Merge pull request !22274 from cathwong/ckw_mon_py_analyze_ut5
2021-08-24 13:24:49 +00:00
i-robot 1e47ff7bc3 !22108 modify check strategy for conv2d
Merge pull request !22108 from yangzhenzhang/modify-check-strategy-for-conv2d
2021-08-24 12:01:23 +00:00
hedongdong 09bc1bddf1 [feat][assistant][I3T92H] add new array operator SplitV 2021-08-24 19:54:26 +08:00
i-robot ab758a981c !20044 [assistant][ops] Add New Lerp
Merge pull request !20044 from 孟权令/Lerp
2021-08-24 11:25:44 +00:00
lanzhineng 6daabf9a86 mindir:add @ms_function testcase 2021-08-24 17:03:25 +08:00
7347157+joylvliang@user.noreply.gitee.com a30800ecc3 add_st_ut_case_for_pynative_non_tensor_input 2021-08-24 16:44:05 +08:00
i-robot 822cedec0a !22153 Convert some ops bprops to mindir
Merge pull request !22153 from YuJianfeng/bprop_mindir
2021-08-24 08:07:46 +00:00
yangzhenzhang f1afaeac5a modify check strategy for conv2d 2021-08-24 16:06:16 +08:00
i-robot 5e56f1952c !22285 Change monad.py as internal usage only
Merge pull request !22285 from hewei/core_user_api
2021-08-24 07:55:00 +00:00
yuchaojie 1d1490df0b xupdate LayerNormGrad split pass to V2 2021-08-24 14:43:09 +08:00
i-robot e49904b417 !21982 [MS][GPU] SyncBatchNormFwd/Grad - GPU Kernel
Merge pull request !21982 from danishfarid/SyncBatchNorm_GPU
2021-08-24 06:16:58 +00:00
i-robot fd06532b59 !22243 add Ascend and CPU ST for enabling RDR
Merge pull request !22243 from yuximiao/yuximiao_rdr
2021-08-24 03:50:05 +00:00
zjun 779b655cfe Add switch_layter pynative ut test
Signed-off-by: zjun <zhangjun0@huawei.com>
2021-08-24 11:21:13 +08:00
yuchaojie 7b3943bc91 DynamicRNN supports scenarios that hidden_size is not multiples of 16 2021-08-24 11:20:44 +08:00
He Wei c9ecb27db8 Change monad.py as internal usage only 2021-08-24 10:45:40 +08:00
i-robot 6082835501 !22146 dataset: support return bytes in GeneratorDataset
Merge pull request !22146 from ms_yan/decode_bytes
2021-08-24 02:41:18 +00:00
Cathy Wong 3581e7962e MD Profiling UT: Use pytest setup and teardown 2021-08-23 21:09:48 -04:00
yujianfeng 712b9bd013 convert some ops bprops to mindir 2021-08-24 09:09:09 +08:00
danishfarid c34e52c3d6 first commit
updated files

lint fix

lint fix 2

file name changes

CI run issue fix
2021-08-23 17:48:48 -04:00
i-robot ef195c9de6 !21631 Deserialize part4
Merge pull request !21631 from zetongzhao/deserialize_3
2021-08-23 20:09:23 +00:00
i-robot db44b88e1e !22094 MD Profiling: For mismatch op info btwn files, skip bottleneck analysis
Merge pull request !22094 from cathwong/ckw_mon_py_analyze_fixes3
2021-08-23 18:28:51 +00:00
zetongzhao b17464e30c deserialize and tests 2021-08-23 14:11:05 -04:00
ms_yan 6ecfc34524 add np.frombuffer to support return bytes 2021-08-23 17:31:38 +08:00
i-robot af34229974 !22206 fix gather_v2
Merge pull request !22206 from liubuyu/bug_fix
2021-08-23 06:16:06 +00:00
i-robot a84f4feb1a !22214 Add env one card
Merge pull request !22214 from chenfei_mindspore/add_env_one_card
2021-08-23 03:50:51 +00:00
i-robot 866d204658 !22117 Fix resnet st probabilistic failure in daily version.
Merge pull request !22117 from linqingke/resnet
2021-08-23 03:49:04 +00:00
chenfei 303146dea6 add env one card 2021-08-23 10:59:15 +08:00
lby e551d1614d fix tbe op gather_v2 2021-08-23 10:20:56 +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 b077aa1cab [feat] [assistant] [I3T96T] add new Dataset operator CMUARCTICDataset 2021-08-22 16:26:45 +08:00
djc 4e6f7dc97d [feat] [assistant] [I3T96X] add new Dataset operator LibriSpeechDataset 2021-08-22 13:39:37 +08:00
liangyongxiong ee9d25bff9 add Ascend and CPU ST for enabling RDR 2021-08-22 11:23:19 +08:00
i-robot 9f08cdc4ab !21958 Fix the execution sequence problem of the load in maketuple
Merge pull request !21958 from Margaret_wangrui/load_in_maketuple
2021-08-21 15:18:54 +00:00
i-robot 0e7844daa9 !22093 revert merge request 21685
Merge pull request !22093 from guozhijian/revert-merge-21685-master
2021-08-21 09:25:32 +00:00
i-robot d3c21c003c !22157 gather_v2 case disable
Merge pull request !22157 from liubuyu/bug_fix
2021-08-21 08:55:51 +00:00
lby f09ca7f061 bug fix 2021-08-21 15:00:37 +08:00
i-robot 0f3de77e93 !21149 Sync MD code clean to master
Merge pull request !21149 from xiefangqi/md_sync_codeclean_master
2021-08-21 06:43:20 +00:00
i-robot 4ba337e0f5 !16667 [assistant][ComplexNorm]
Merge pull request !16667 from QingfengLi/ComplexNorm
2021-08-21 02:44:18 +00:00
doit bf23333b27 [feat] [assistant] [I3T927] add new math operator Lerp 2021-08-20 20:19:41 +08:00
i-robot f34426095c !22084 Move Transformer Code from Mindspore.nn.parallel to mindspore.parallel.nn
Merge pull request !22084 from huangxinjing/transformer_doc_check
2021-08-20 10:40:01 +00:00
huangxinjing d777742904 1. Move the class to mindspore.parallel, support activation sharding 2021-08-20 14:34:10 +08:00
linqingke dd30c66985 Fix resnet st probabilistic failure in daily version. 2021-08-20 11:55:04 +08:00
i-robot 09a1a7f1f2 !21924 Add subdirectory for glog and ir_dump
Merge pull request !21924 from huanghui/add-submodule-for-dfx-files
2021-08-20 02:25:52 +00:00
i-robot 954864bfba !22037 incoroporate getitem if indexed node is zeros_like
Merge pull request !22037 from xychow/always-incorporate-getitem-if-zeros-like
2021-08-20 02:14:35 +00:00
i-robot 36f0b3c353 !22062 upgrade ascend package 19 Aug 21
Merge pull request !22062 from yanghaoran/upgrade_ascend_0819
2021-08-20 02:03:07 +00:00
Margaret_wangrui d0d0cc2c71 Fix the execution sequence problem of the load in maketuple 2021-08-20 09:37:42 +08:00
i-robot 5d1bb097e2 !22052 Fix probability core dump in MindData coverage test
Merge pull request !22052 from luoyang/issues2
2021-08-20 01:18:06 +00:00
robert_luo_yibo d837cbee0f [feat][assistant][I3CEG8] add new data operator ComplexNorm 2021-08-20 09:07:21 +08:00
Cathy Wong 85931175bc MD Profiling: For mismatch op info btwn files, skip bottleneck analysis.
Add to summary output: per_pipeline_time and per_push_queue_time.
Enhance UT
2021-08-19 12:19:59 -04:00
guozhijian e1df9a2152 revert 'Pull Request !21685 : add function to invert_op' 2021-08-19 22:49:51 +08:00
yanghaitao1 8fc11cb676 adapt delete libms_profiler_fwk.a 2021-08-19 19:33:53 +08:00
i-robot d978db22ca !22007 update shift op
Merge pull request !22007 from zhujingxuan/update_shift
2021-08-19 10:40:53 +00:00
i-robot 4113aee7d6 !18668 [assistant][ops]New operator implementation, include Flowers102Dataset
Merge pull request !18668 from Wangsong95/flowers102_dataset
2021-08-19 08:41:44 +00:00
YangLuo 9352935ff6 Fix probability core dump in MindData coverage test 2021-08-19 14:34:57 +08:00
i-robot c93bb9936e !20011 [assistant][ops] Add nn operator Roll
Merge pull request !20011 from 孟权令/Roll
2021-08-19 06:28:28 +00:00
huanghui 1630dcb0c8 add subdirectory for log and ir_dump 2021-08-19 14:11:16 +08:00
i-robot 3ad641cb65 !21685 add function to invert_op
Merge pull request !21685 from Killer/fake_killer
2021-08-19 05:13:39 +00:00
i-robot 8ec16bbab9 !21880 Check ref of update parameters
Merge pull request !21880 from chenfei_mindspore/ascend-control-use-vm
2021-08-19 03:49:12 +00:00
zhousiyi dcd374961e incoroporate getitem if indexed node is zeros_like so AddN(MakeTuple(x, zeros_like)) can be optimized 2021-08-19 03:28:24 +00:00
i-robot c144e37bab !21996 Re-add F.depend to the script
Merge pull request !21996 from huangbingjian/depend_yolov3
2021-08-19 01:37:43 +00:00
i-robot 5e718c5676 !21973 GetRankId failed in PyNative mode
Merge pull request !21973 from caifubi/master-hccl-get-rank-id
2021-08-19 01:22:23 +00:00
xiefangqi 0e2efc2501 fix minddata code clean 2021-08-19 09:11:30 +08:00
chenfei 4428db97e3 check abstract ref of update parameters 2021-08-18 23:14:14 +08:00
fake_killer b0bc4bf5da [feat] [assistant] [I44TRS] extend invert_op's function 2021-08-18 21:46:15 +08:00
i-robot dde05c36b8 !21551 auto_parallel_context interface dataset_strategy union
Merge pull request !21551 from yao_yf/auto_parallel_context_dataset_strategy_union
2021-08-18 09:03:18 +00:00
caifubi dfe0e94466 Fix PyNative get_rank_id/get_rank_size 2021-08-18 16:39:18 +08:00
zhujingxuan 7a32af0d91 update shift API and st
(cherry picked from commit 347abe4fbaf9b64cb574a8e5762bf7bebe561f21)
2021-08-18 16:25:07 +08:00
i-robot 415275ae17 !21805 support adafactor model parallel
Merge pull request !21805 from yangzhenzhang/auto-parallel-support-adafactor-opt
2021-08-18 07:20:39 +00:00
huangbingjian 4fccefdfd8 Re-add F.depend to the script 2021-08-18 14:54:29 +08:00
i-robot 7c95c2dece !21964 use temporary dir as dump dir
Merge pull request !21964 from yelihua/dev
2021-08-18 02:50:41 +00:00
hedongdong d76ed25e96 [feat][assistant][I3T92I] add new nn operator Roll 2021-08-18 10:15:17 +08:00
i-robot 264cb4d2e8 !21920 MD Profiling Analyze: Search for device trace file, independent of device_target
Merge pull request !21920 from cathwong/ckw_mon_py_analyze_device_file
2021-08-18 00:57:50 +00:00
i-robot 1e4dace193 !20683 deserializer 2nd part
Merge pull request !20683 from zetongzhao/deserialize_2
2021-08-17 19:50:43 +00:00
i-robot 8c81b18c15 !21912 Add Multinomial kernel on CPU.
Merge pull request !21912 from chengang/multinomial_cpu_kernel
2021-08-17 12:15:29 +00:00
i-robot f0a7fa47c4 !21908 Add UniformInt & UniformReal kernels on CPU
Merge pull request !21908 from chengang/random_ops
2021-08-17 12:15:18 +00:00
yelihua a3cba3857e use temporary dir as dump dir 2021-08-17 20:11:33 +08:00
i-robot 0d839fa7c6 !21809 Improved Transformer Struture and Add Args Check
Merge pull request !21809 from huangxinjing/transformer_improved
2021-08-17 12:05:25 +00:00
i-robot a77a0b968d !21761 comm_recompute_interface.
Merge pull request !21761 from yao_yf/comm_recompute_interface
2021-08-17 09:14:29 +00:00
i-robot debf4d8ece !21785 Fix Programming Specifications of Crowd Contributing Code
Merge pull request !21785 from xiaotianci/fix_crowd_contributing
2021-08-17 07:53:50 +00:00
cristoval 0a3e584eec add multinomial cpu kernel 2021-08-17 14:54:38 +08:00
yangzhenzhang 7ca64d2235 auto parallel support adafactor opt 2021-08-17 10:36:42 +08:00
yao_yf 5277b229be add cell comm recompute interface 2021-08-17 09:23:21 +08:00
Cathy Wong a2cbd4b5fa MD Profiling Analyze: Search for device trace file
MinddataProfilingAnalyzer() - remove device_target input parm
2021-08-16 15:41:11 -04:00
zetongzhao ea1f2d0d7c deserialize and tests 2021-08-16 10:23:36 -04:00
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 16d5427743 !21863 update the method for get rank_id
Merge pull request !21863 from yelihua/new-dev
2021-08-16 12:13:01 +00:00
cristoval b7c6f7ec07 add uniform_int and uniform_real on cpu 2021-08-16 20:11:18 +08:00
Xiao Tianci aae886bcd7 fix crowd contributing code 2021-08-16 19:33:10 +08:00
i-robot 692b39f306 !21821 RL add buffer sample update
Merge pull request !21821 from VectorSL/buffer-sample-update
2021-08-16 09:16:28 +00:00
i-robot ecb2f35665 !21818 add logs when query running jobs
Merge pull request !21818 from laiyongqiang/fusion_dbg
2021-08-16 06:13:58 +00:00
VectorSL 680d319290 update buffer sample gpu 2021-08-16 14:05:51 +08:00
yelihua a6dc9a0a07 get rank id when set hccl env for single card train 2021-08-16 13:44:02 +08:00