p47068915
/
mindspore
forked from mindspore-Ecosystem/mindspore
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
34,741 Commits 23 Branches 0 Tags 3.9 GiB
Commit Graph

4562 Commits

Author SHA1 Message Date
hesham cf3ff77cfd AutoTune support non-sink mode 
Add time sts to iterator tracing
 2021-11-24 17:51:25 -05:00
i-robot eef5f3f8d4 !22925 [assistant][ops] add operator Cummin 
Merge pull request !22925 from yangwm/cummin
 2021-11-24 11:46:23 +00:00
i-robot d251800c62 !26705 Add the check of function return None. 
Merge pull request !26705 from Margaret_wangrui/return_none
 2021-11-24 11:09:27 +00:00
i-robot cfc6ea32ff !24714 replace rtmemcpyxx to acl memcpy 
Merge pull request !24714 from jjfeing/br_replace_rtmemcpyxx_with_acl_api
 2021-11-24 07:47:43 +00:00
Margaret_wangrui f9a384456a Add the check of function return None. 2021-11-24 15:30:35 +08:00
i-robot 2cc51099b1 !26718 [ME][Fallback] Add some fallback numpy test cases 
Merge pull request !26718 from Margaret_wangrui/fallback
 2021-11-24 06:43:13 +00:00
i-robot b472850a75 !26594 Replace std::unordered_map/set with robin-hood-hashing 
Merge pull request !26594 from hewei/use_robin_hood
 2021-11-24 06:30:36 +00:00
i-robot 519f14a909 !26006 slice recompute activation 
Merge pull request !26006 from yao_yf/add_transformer_slice_activation_config
 2021-11-24 06:14:28 +00:00
i-robot de8b6db07c !26585 add warnings for set_dump API, see I4AUIR for details 
Merge pull request !26585 from wenkai/wk1120setdumpwarning1
 2021-11-24 03:42:25 +00:00
l00591931 21df240f23 Enable mindir to load initialize weight from python 2021-11-24 11:33:43 +08:00
Margaret_wangrui 16ac3638d3 [ME][Fallback] Add some fallback numpy test cases 2021-11-24 11:20:06 +08:00
He Wei 41dcac9c49 Replace std::unordered_map/set with robin-hood-hashing 
Robin-hood-hashing (https://github.com/martinus/robin-hood-hashing)
is considered faster then std::unordered_map/set,
so we use it to improve mindspore performance.

1. robin_hood head file in `third_party/robin_hood/include`;
2. In `utils/hash_map.h` and `utils/hash_set.h`, we define:
 - mindspore::HashMap as an alias of robin_hood::unordered_map;
 - mindspore::HashSet as an alias of robin_hood::unordered_set;
3. Replace:
 - `#include <unordered_map>` --> `#include "utils/hash_map.h"`;
 - `#include <unordered_set>` --> `#include "utils/hash_set.h"`;
 - `std::unordered_map` --> `mindspore::HashMap`;
 - `std::unordered_set` --> `mindspore::HashSet`;
 - `map.insert(std::pair(key, value))` --> `map.emplace(key, value)`;
 - `[] (const std::pair<K, V> &p) {..} ` --> `[] (const auto &p) {..} `;
4. Fix issues found by switch to robin_hood:
 - AnfNodeConfig hash and equal;
 - Fix a bug in `Slice::operator==()`;
 - Fix a bug in `CNode::HasPrimalAttr()`;
 - Fix map.erase() usage bugs: `map.erase(iter++)` --> `iter = map.erase(iter)`;
 - Fix some iterator invalidated problem;
5. Some std::unordered_map/set can not replace by robin_hood:
 - As parameter of functions that exposed to python by pybind11;
 - Use bad hash that cause robin_hood::map over_flow, such as AbstractBasePtrListHasher;
6. Update cpp unit tests;
7. Add build option '-F' to enable robin_hood, default on.
 2021-11-24 10:47:40 +08:00
i-robot 9a5fd32bd2 !26280 fix mindir export's error when using _encrypt 
Merge pull request !26280 from lianliguang/master
 2021-11-24 01:48:32 +00:00
wzw 86c5ad20c8 parallel_ut_refactoring1 2021-11-24 09:41:52 +08:00
luon 5485ff0cb6 [feat] [assistant] [I40GGH] add new ascend operator Cummin 2021-11-23 21:57:23 +08:00
wenkai c493f19b6b add warnings for set_dump API, see I4AUIR for details 2021-11-23 21:55:36 +08:00
i-robot 875f35d6d8 !26541 Fix file name and field type changes generated by HCCL in profiler. 
Merge pull request !26541 from casgj/master_hccl
 2021-11-23 13:16:28 +00:00
casgj b15c09db6d Fix file name and field type changes generated by HCCL. 2021-11-23 12:48:59 +08:00
i-robot 76206a8894 !26162 Add automatic detection for the changes of python scripts when enable compilation cache 
Merge pull request !26162 from YuJianfeng/compile_cache
 2021-11-23 02:54:38 +00:00
i-robot 310bbcebe1 !26539 [GraphKernel] support lite_adapter for graph kernel. 
Merge pull request !26539 from chenlei_autodiff/lite_callback
 2021-11-23 02:44:13 +00:00
jjfeing 05485d991c replace api with acl api 2021-11-23 09:06:28 +08:00
yujianfeng f2bd7de442 Add automatic detection for the changes of python scripts when enable compile cache 2021-11-22 21:27:02 +00:00
i-robot 4f55a3f52f !26275 [API] Tensor python float trans to ms float32 
Merge pull request !26275 from kingxian/master
 2021-11-22 11:13:15 +00:00
chenlei_autodiff 4c0d5dcfe7 [GraphKernel] support lite_adapter for graph kernel. 2021-11-22 15:10:30 +08:00
lianliguang a6e5dcfb22 if parameter total larger than 1T split to another file 2021-11-22 11:17:33 +08:00
kingxian 545dc83fcb fix: `cell` does not support deep copy, tensor default dtype -> float32 2021-11-22 09:08:32 +08:00
i-robot 1b8c2ff0e9 !26414 fault_recover_by_mirror_group_fix_opt_shard 
Merge pull request !26414 from yao_yf/fault_recover_by_mirror_group_fix_opt_shard
 2021-11-20 03:06:45 +00:00
i-robot 4f34c32c6d !26235 AutoTune v1 
Merge pull request !26235 from danishfarid/autotune
 2021-11-19 23:40:40 +00:00
Rescue da986710f0 Updated concat-zip check and UT. 2021-11-19 17:38:16 +00:00
hesham 40797c40da AutoTune 2021-11-19 09:27:12 -05:00
i-robot 70363899e7 !24080 [assistant][ops] Add new audio operator ComputeDeltas 
Merge pull request !24080 from Rainfor/compute_detlas
 2021-11-19 12:29:49 +00:00
i-robot c6821bde0a !21648 [assistant][ops] Add new data loading operator YesNoDataset 
Merge pull request !21648 from 杨旭华/YesNoDataset
 2021-11-19 09:05:07 +00:00
vddong f465cd0242 [feat][assistant][I3T96L]add new dataset operator YesNo 2021-11-19 13:36:24 +08:00
i-robot f38df5c888 !26234 New dataset config option, auto_offload 
Merge pull request !26234 from markuskunej/offload_config
 2021-11-19 03:44:56 +00:00
i-robot 55463892e4 !26402 add mem offload strategy 
Merge pull request !26402 from kisnwang/add-mem-offload-strategy
 2021-11-19 03:42:52 +00:00
yao_yf 188d39da83 slice_activation_in_recompute 
slice recompute activation
 2021-11-19 10:51:54 +08:00
i-robot 7559d5b798 !26494 [Auto parallel] Adjusting sharding propagation 
Merge pull request !26494 from Xiaoda/102-adjusting-sharding-propagation
 2021-11-19 02:00:46 +00:00
yao_yf 01dc4bbdf9 fix fault recover in optimizer shard 2021-11-19 09:46:27 +08:00
i-robot 3c39afad11 !24547 [MS][crowdfunding]New operator implementation, AutoAugment 
Merge pull request !24547 from yangwm/autoaugment
 2021-11-19 01:00:51 +00:00
markuskunej 29cf442521 Added dataset.config.set_auto_offload(), which will automatically attempt to offload dataset map operations if set to True. 2021-11-18 21:13:20 +00:00
i-robot 117df75afc !26126 MD Profiling - Add Start Stop Python UT 
Merge pull request !26126 from cathwong/ckw_ut_prof_startstop
 2021-11-18 14:05:00 +00:00
i-robot ce00ee1ad1 !25367 use acl api to control profiling 
Merge pull request !25367 from yanghaitao/yht_condation_start_profiler
 2021-11-18 12:36:14 +00:00
zhuchuang 4c3931f5c9 [feat] [assistant] [I3CKEE] add new audio operator ComputeDeltas 2021-11-18 17:18:43 +08:00
i-robot 9d6248194e !26310 MindSpore support load custom aicpu kernels. 
Merge pull request !26310 from linqingke/aicpu
 2021-11-18 08:09:26 +00:00
Xiaoda Zhang df67e74eaf making sharding_propagation smooth, add a reshape justification: 
1) when propagate sharding strategy from one op to another, try to find the strategy with zero communication cost;
2) if there is no such strategy, find the strategy with minimum communication cost, and raise a warning;
 2021-11-18 15:55:50 +08:00
kswang 7c312bd38c add mem offload strategy 2021-11-18 15:54:16 +08:00
i-robot 9f52343a6a !26350 add check for resizenearestneighbor parallel op 
Merge pull request !26350 from yangzhenzhang/add-check-for-resize-op
 2021-11-18 05:45:57 +00:00
despicablemme 09897ce7ec [feat] [assistant] [I40GZP] add new data op AutoAugment 2021-11-18 11:38:18 +08:00
i-robot 707e7fbbf9 !26453 [Fallback] Add some fallback numpy test cases 
Merge pull request !26453 from Margaret_wangrui/fallback_tests
 2021-11-18 03:01:47 +00:00
Margaret_wangrui c6582b9390 [Fallback] Add some fallback numpy test cases 2021-11-18 10:18:13 +08:00
Cathy Wong b1094b9d81 MD Profiling - Add Start Stop Python UT 2021-11-17 13:49:43 -05:00
yanghaitao1 c94aa6b872 use profiler acl api instead 2021-11-17 09:33:31 -05:00
i-robot 4e88f2e5e8 !26345 add examples of function grad, jvp and vjp 
Merge pull request !26345 from chenzhuo/jvp
 2021-11-17 08:35:38 +00:00
linqingke bef2923acf MindSpore support load custom aicpu ops. 2021-11-17 14:58:50 +08:00
i-robot 36517de69c !24519 [assistant][ops] Add new audio operator SlidingWindowCmn. 
Merge pull request !24519 from Rainfor/SlidingWindowCmn
 2021-11-17 06:40:43 +00:00
i-robot 9f2e1edc00 !26279 neighborexchangev2 fix bug 
Merge pull request !26279 from TuDouNi/neighborexchangev2_bug
 2021-11-17 02:29:21 +00:00
wangkc123 706f9e2cbb [feat] [assistant] [I3CKEH] add new audio operator SlidingWindowCmn 2021-11-16 22:00:09 +08:00
yangzhenzhang ba99e4c505 add check for resize op 2021-11-16 15:39:52 +08:00
i-robot 58b69a05ee !20576 [assistant][ops] add new dataset loading operator LJSpeechDataset 
Merge pull request !20576 from 杨旭华/LJSpeechDataset
 2021-11-16 07:00:20 +00:00
chenzhuo d07f24462c add examples of grad, jvp and vjp 2021-11-16 14:15:26 +08:00
ttudu 33ac1de062 fix bug 2021-11-16 11:14:58 +08:00
i-robot 7cdef3da6e !26334 [MD] fix occasional fail in random rampler ut 
Merge pull request !26334 from liyong126/fix_random_sampler
 2021-11-16 02:49:59 +00:00
i-robot 6bdd38399a !25811 fault_recover_by_mirror_group 
Merge pull request !25811 from yao_yf/fault_recover_by_mirror_group
 2021-11-16 02:20:58 +00:00
liyong 9663884db3 fix occasional failed in random sampler 2021-11-16 09:53:38 +08:00
i-robot 5af8572cd7 !25488 add the set_dump API 
Merge pull request !25488 from wenkai/wk1027setdump1
 2021-11-16 01:47:32 +00:00
i-robot 0113b3eee1 !26236 Add MD Profiler save() 
Merge pull request !26236 from Mohammad Motallebi/add_profiler_save
 2021-11-16 01:04:53 +00:00
mohammad 5c8ab5f60c add MD Profiler Save() 2021-11-15 13:18:36 -05:00
wenkai d2eca80d40 add the set_dump API, see I4AUIR for details 2021-11-15 21:28:40 +08:00
i-robot 9dfe08b805 !25909 [Fallback] Add global Py param when processing global namespace info 
Merge pull request !25909 from huangbingjian/update_global_params
 2021-11-15 11:43:35 +00:00
i-robot 9522ee9686 !25703 [assistant][ops] Add math operator LpNorm 
Merge pull request !25703 from 孟权令/LpNorm
 2021-11-15 02:36:54 +00:00
vddong 6fe2d35ae2 [feat][assistant][I3T96D] add new dataset loading operator LJSpeechDataset 2021-11-14 14:52:15 +08:00
yao_yf 501b978d16 find data parallel common group in auto parallel 2021-11-13 10:23:45 +08:00
i-robot 04da5c2808 !26140 Add callback support to BatchOp Datasets to enable Add/Remove workers 
Merge pull request !26140 from h.farahat/cb_batch
 2021-11-12 17:23:19 +00:00
i-robot e154dca4b1 !22193 [assistant][ops] Add new loader operator DBpedia 
Merge pull request !22193 from 杨旭华/DBpedia
 2021-11-12 09:02:28 +00:00
i-robot 0f07408425 !26034 F.grad support sens_param and fix graph_mode bug 
Merge pull request !26034 from zhang_sss/grad
 2021-11-12 06:38:15 +00:00
i-robot e773f0621e !25918 add Function jvp and vjp 
Merge pull request !25918 from chenzhuo/jvp
 2021-11-12 06:37:46 +00:00
i-robot 5869e02fa1 !26135 Enable tuple input of optimizer as variable 
Merge pull request !26135 from LiangZhibo/broaden3
 2021-11-12 06:29:35 +00:00
i-robot 6c587dc2d3 !25391 Support to profiling parallel strategy 
Merge pull request !25391 from ougongchang/profiling_stategy
 2021-11-12 06:07:30 +00:00
Carry955 0d26c38693 [feat][assistant][I40GXT] add new loader DBpedia 2021-11-11 19:02:22 -08:00
i-robot 7a73bae5c3 !26036 add output strategy for matmul operator 
Merge pull request !26036 from yangzhenzhang/add-output-strategy-for-op-init
 2021-11-12 01:22:22 +00:00
i-robot e03544685b !26049 [Auto parallel] Support reshape in sharding propagation 
Merge pull request !26049 from Xiaoda/99-ad-reshape-adjusting-in-sharding-propagation
 2021-11-12 01:14:18 +00:00
chenzhuo f36edec3ff add function jvp 2021-11-12 08:44:41 +08:00
hesham 2662b6d5c3 Add callback to Batch op 2021-11-11 10:58:01 -05:00
zhang__sss 0dd68aa7eb grad 2021-11-11 23:18:35 +08:00
i-robot 1792d5f8a4 !26131 Offload Hardware Acceleration Computing 
Merge pull request !26131 from alashkari/hw_accel_move_copy
 2021-11-11 13:18:07 +00:00
ougongchang 9229f1c1ff profiler support to collect parallel strategy info 
If SetNodeOutputType functions forcibly splits into multiple functions, the readability decreases, so it blocks lizard scans
 2021-11-11 20:11:42 +08:00
i-robot b910870ecc !19986 [assistant][ops] Add new dataset loading operator AGNEWS 
Merge pull request !19986 from 杨旭华/AGNEWS
 2021-11-11 06:22:25 +00:00
Xiaoda Zhang a772767265 support reshape in sharding propagation: 
1) using 'swc index of strategy_cost_' as reshape's selected strategy;
2) when encountering reshape in BFS, select the 'swc index' with zero communication cost;
3) when encountering a reshape that is already visited, check whether there exists communication between reshape and current operator. It is OK if communication happens between two configured operators;
4) currently, two consecutive reshapes are not supported;
5) adjusting BFS structure in graph_costmodel.cc;
6) adjusting some code in step_auto_parallel.cc to avoid cyclomatic complexity.
 2021-11-11 11:35:43 +08:00
l00591931 9353bd10a4 Enable tuple as variable in optimizer 2021-11-11 11:31:05 +08:00
zetongzhao 2de8142c44 profiler test 2021-11-10 11:55:19 -05:00
Megalomania c34bbabed9 [feat] [assistant] [I3T96M] add new loader AGNEWS 2021-11-09 23:51:19 -08:00
yangzhenzhang 8431ba616c add output strategy for op init 2021-11-10 15:35:45 +08:00
rescue 22e03ab12e Adding offload hardware accelerator. 2021-11-10 07:03:43 +00:00
i-robot 8bf7e28fa6 !25410 add dump flag for fusion nodes 
Merge pull request !25410 from yuchaojie/ir_fusion3
 2021-11-10 02:25:08 +00:00
i-robot d4c8dde6d2 !25923 Move IsRealKernel and IsRealCNodeKernel from AnfAlgo to AnfUtils 
Merge pull request !25923 from DeshiChen/1101_anfalgo
 2021-11-10 01:56:07 +00:00
i-robot 2e174427c9 !25963 [MD][Autotune] Add WaitFor 
Merge pull request !25963 from harshvardhangupta/hesh_tree_mod
 2021-11-09 22:43:00 +00:00
i-robot b7244ac987 !25852 MD C++ UT: Remove old tests part 2 
Merge pull request !25852 from hetshah/hs_remove_old_tests
 2021-11-09 22:07:59 +00:00
hetshah 5c703015e4 remove old tests and move coverage 2021-11-09 21:27:37 +00:00
i-robot 61b53fee3a !25980 Add callback support to Mappable Datasets to enable Add/Remove workers 
Merge pull request !25980 from h.farahat/cb_mappable
 2021-11-09 20:05:01 +00:00
i-robot f1f7731fb2 !25964 Fix serdes multithreading bug 
Merge pull request !25964 from zetongzhao/serdes_multithreading
 2021-11-09 15:48:06 +00:00