yankf21
/
mindspore2022
forked from huawei/mindspore2022
0
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
1,217 Commits 18 Branches 22 Tags 3.6 GiB
Commit Graph

24 Commits

Author SHA1 Message Date
mindspore-ci-bot 5b3327d103 !746 reducescatter backforward operator 
Merge pull request !746 from lirongzhen1/bp_reducescatter
 2020-04-28 14:11:53 +08:00
lirongzhen1 0b4648881b add reducescatter bprop 2020-04-28 09:58:34 +08:00
Xiaoda Zhang e227415673 support-the-multiple-subgraphs-in-the-ANF 2020-04-27 09:34:41 +08:00
ch-l f806b72447 use DeviceMemory for memory control 2020-04-23 17:55:17 +08:00
zhoufeng c2b3360d69 update clang format rule 2020-04-21 21:21:19 +08:00
mindspore-ci-bot 46acf23825 !405 [AutoParallel] Adapte rec-prog generator to new parser 
Merge pull request !405 from Chong/str-gen
 2020-04-21 16:49:46 +08:00
mindspore-ci-bot 5b6b1ad727 !394 [AutoParallel] Simplify rec-prog parser mechanism 
Merge pull request !394 from Chong/parser
 2020-04-21 14:07:24 +08:00
ch-l c71234f383 improve rec-prog str generator 2020-04-21 01:14:49 +08:00
Ziyan 0d208e00bd Model ALLTOALL as a single operator in cost model; scale the ALLTOALL, 
ALLGATHER, and REDUCESCATTER with different factors; change the BETA and
GAMMA value in cost model.
 2020-04-20 09:11:03 +08:00
Chong b1f5e44cd4 improve parser 2020-04-18 05:22:41 +02:00
Xiaoda Zhang 79de8f4bdf Adjusting backward communication cost of some operators 2020-04-17 11:17:34 +08:00
yangzhenzhang 6d522f0a4f add parallel op for layernorm 2020-04-15 19:30:07 +08:00
Xiaoda Zhang ffb2cb03a4 Change 'NOT_FULLY_USE_DEVICES' to 'FULLY_USE_DEVICES' and make ALL-1 user-specified-strategy valid in auto-parallel 2020-04-15 10:14:46 +08:00
Xiaoda Zhang 0ac50a19f5 Model the memory cost in auto-parallel. It is calculated by the output of operators, plus the parameters. Additionally, modify the graph-operations in auto_parallel to include memory_cost. 2020-04-14 11:39:31 +08:00
c00425699 d62f560b50 add_bool_type_check_in_comm_op 2020-04-13 17:34:29 +08:00
c00425699 c8cdb6b331 support distributed GatherV2 operator 2020-04-12 16:00:47 +08:00
yangzhenzhang b34c0e7a17 add parallel op for dropoutdomask 2020-04-11 12:11:22 +08:00
c00425699 b413638f23 refactor OperatorCostPtr in OperatorInfo 2020-04-09 20:37:52 +08:00
Xiaoda Zhang a153fad874 This commit is to separate the computation cost and memory cost in auto_parallel. Some related memory correction is removed. 2020-04-08 11:52:19 +08:00
Su Teng 60b68a1470 sort include file in parallel dir 2020-04-02 15:14:47 +08:00
Xiaoda Zhang 3d35792877 change_star_elimination: make the non-identity triangle_eliminatin exact 2020-04-01 19:24:18 +08:00
Xiaoda Zhang c080ec7874 change star elimination: remove some redundant and checking works 2020-04-01 09:16:36 +08:00
lichenever f946aea10d fix grpah mode loop sink bug in auto parallel 2020-03-31 16:16:53 +08:00
zhunaipan 930a1fb0a8 initial version 
Signed-off-by: leonwanghui <leon.wanghui@huawei.com>
 2020-03-27 22:54:54 +08:00