Xiaoda Zhang
|
9f4b8a3cd1
|
changing the succive edges order in GetAliveSuccEdges() so that Triangle and Star Elimination can be merged into particular node; adding some check information
|
5 years ago |
hongxing
|
2f7c5f7a59
|
fix edge bug
|
5 years ago |
hongxing
|
9c8e750c9e
|
maximize strategy dynamically
|
5 years ago |
hongxing
|
d84ccfe87f
|
complete element-wise list
|
5 years ago |
hongxing
|
8f04adf1c3
|
feature : eliminate graph
|
5 years ago |
hongxing
|
dc290d7959
|
support squeeze and reduce op
|
5 years ago |
Xiaoda Zhang
|
a05aa21cc2
|
calculating PEAK memory cost in the inference phase
|
5 years ago |
mindspore-ci-bot
|
7dc31684b6
|
!1073 fix reshape tensor redistribution bug
Merge pull request !1073 from yao_yf/reshape_bug_fix
|
5 years ago |
yao_yf
|
b0921c15e9
|
xreshape tensor_redistrinution bug fix
|
5 years ago |
mindspore-ci-bot
|
95d4665db9
|
!1051 auto parallel reshape reconstruct
Merge pull request !1051 from yao_yf/auto_parallel_reshape_reconstruct
|
5 years ago |
mindspore-ci-bot
|
2dd97a0780
|
!1028 [AutoParallel] refined memory check & added odd num tensor shape support
Merge pull request !1028 from Chong/memctl
|
5 years ago |
yao_yf
|
716def7c0a
|
move InferStraByTensorInfo to tensor_info.h
|
5 years ago |
mindspore-ci-bot
|
dd2062bf8d
|
!1023 add_gatherv2_distributed_op
Merge pull request !1023 from lichen/add_gatherv2_distributed_op
|
5 years ago |
lichenever
|
19a24b86ac
|
add gatherv2 distributed op
|
5 years ago |
yao_yf
|
f0bf438a55
|
reshape strategy search
|
5 years ago |
sheng
|
59bb014144
|
refine mem ctl & odd num ctl & fuse str writeback
|
5 years ago |
Xiaoda Zhang
|
def8573275
|
implementing-searching-strategy-for-inference
|
5 years ago |
ch-l
|
caac6bce5c
|
adjustements w.r.t. distributed execution
|
5 years ago |
mindspore-ci-bot
|
69ab46e624
|
!727 [AutoParallel] complete cost for recursive programming
Merge pull request !727 from Chong/cost
|
5 years ago |
ch-l
|
309060b1c2
|
complet cost models
|
5 years ago |
mindspore-ci-bot
|
5b3327d103
|
!746 reducescatter backforward operator
Merge pull request !746 from lirongzhen1/bp_reducescatter
|
5 years ago |
lirongzhen1
|
0b4648881b
|
add reducescatter bprop
|
5 years ago |
Xiaoda Zhang
|
e227415673
|
support-the-multiple-subgraphs-in-the-ANF
|
5 years ago |
ch-l
|
f806b72447
|
use DeviceMemory for memory control
|
5 years ago |
zhoufeng
|
c2b3360d69
|
update clang format rule
|
5 years ago |
mindspore-ci-bot
|
46acf23825
|
!405 [AutoParallel] Adapte rec-prog generator to new parser
Merge pull request !405 from Chong/str-gen
|
5 years ago |
mindspore-ci-bot
|
5b6b1ad727
|
!394 [AutoParallel] Simplify rec-prog parser mechanism
Merge pull request !394 from Chong/parser
|
5 years ago |
ch-l
|
c71234f383
|
improve rec-prog str generator
|
5 years ago |
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.
|
6 years ago |
Chong
|
b1f5e44cd4
|
improve parser
|
5 years ago |
Xiaoda Zhang
|
79de8f4bdf
|
Adjusting backward communication cost of some operators
|
5 years ago |
yangzhenzhang
|
6d522f0a4f
|
add parallel op for layernorm
|
5 years ago |
Xiaoda Zhang
|
ffb2cb03a4
|
Change 'NOT_FULLY_USE_DEVICES' to 'FULLY_USE_DEVICES' and make ALL-1 user-specified-strategy valid in auto-parallel
|
5 years ago |
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.
|
6 years ago |
c00425699
|
d62f560b50
|
add_bool_type_check_in_comm_op
|
6 years ago |
c00425699
|
c8cdb6b331
|
support distributed GatherV2 operator
|
6 years ago |
yangzhenzhang
|
b34c0e7a17
|
add parallel op for dropoutdomask
|
6 years ago |
c00425699
|
b413638f23
|
refactor OperatorCostPtr in OperatorInfo
|
6 years ago |
Xiaoda Zhang
|
a153fad874
|
This commit is to separate the computation cost and memory cost in auto_parallel. Some related memory correction is removed.
|
6 years ago |
Su Teng
|
60b68a1470
|
sort include file in parallel dir
|
6 years ago |
Xiaoda Zhang
|
3d35792877
|
change_star_elimination: make the non-identity triangle_eliminatin exact
|
6 years ago |
Xiaoda Zhang
|
c080ec7874
|
change star elimination: remove some redundant and checking works
|
6 years ago |
lichenever
|
f946aea10d
|
fix grpah mode loop sink bug in auto parallel
|
6 years ago |
zhunaipan
|
930a1fb0a8
|
initial version
Signed-off-by: leonwanghui <leon.wanghui@huawei.com>
|
6 years ago |