7d7c9c4fee
!311 GPU add akg kernel select
...
Merge pull request !311 from VectorSL/select
5 years ago
d9dd6aa0b8
!277 fix: use correct permissions when opening mindrecord files
...
Merge pull request !277 from guozhijian/fix_read_with_exact_option
5 years ago
57aee805ce
!274 GPU multiple stream feature
...
Merge pull request !274 from ZPaC/gpu-backend-supports-multiple-streams
5 years ago
99f12f9105
gpu uses dynamic memory pool by default
5 years ago
7595c91730
!301 add custom op st to ci
...
Merge pull request !301 from dengwentao/ci_st
5 years ago
fb19655ea6
!214 modify GPU operator information registration
...
Merge pull request !214 from Maoweiyong/modify-gpu-op-register
5 years ago
2dd60b4ce3
pack via cpack
5 years ago
950d75e376
add get_class_indexing exception info
5 years ago
66377e4c92
!293 Register all tbe op info with new registration mode
...
Merge pull request !293 from zjun/Modify_all_tbe_op
5 years ago
3ea3d9e5a4
1.GPU supports multiple streams.
...
2.GPU commnication stream and compute stream overlap.
5 years ago
a30ee78001
!295 bugfix(side effect): fix cell object can not free normally.
...
Merge pull request !295 from gongchen/fix_free_cell_obj
5 years ago
e09364df48
!298 bugfix(side effect): fix adding wrong control depend between AllReduce and GetStatus
...
Merge pull request !298 from gongchen/fix_allreduce_control_depend
5 years ago
48f90eb7bc
add custom op st to ci
5 years ago
d248b05a98
gpu add kernel select
5 years ago
a007e4812b
modify gpu operator information registration
5 years ago
b571fabd77
!289 Add cnode mapping after graph match
...
Merge pull request !289 from YuJianfeng/find_op
5 years ago
791442288c
fix format typo in examples
5 years ago
86911b7cdc
!287 [MS][Docs]Fix "ValueError: The `outputs` is not tuple. ", when use model.eval.
...
Merge pull request !287 from vlne-v1/I1EFUD-the-output-is-not-tuple
5 years ago
02aca06451
Add all other tbe op info register
5 years ago
c06d2c6c2a
!299 modify annotation of DepthwiseConv2d
...
Merge pull request !299 from wangnan39/modify_annotation_of_depthwise_conv
5 years ago
e5c67b9088
Add cnode to equal map when opt matching
5 years ago
7a6fdaf132
!302 fix workspace memory reuse
...
Merge pull request !302 from kisnwang/fix-workspace-malloc
5 years ago
e364936bd3
!297 add alternate path for stdclient library
...
Merge pull request !297 from yanghaoran/master
5 years ago
39b9e831cb
!291 disable memory reuse for GetNext op
...
Merge pull request !291 from caifubi/dev-getnext-mem-reuse-off
5 years ago
ea5da25d10
fix: use exactly read option
5 years ago
e1c8f248e0
Fix the output is not tuple, when eval
5 years ago
e53376092f
fix workspace reuse bug
5 years ago
8ba68220dd
modify annotation of DepthwiseConv2d
5 years ago
61e959a9f3
bugfix(side effect): fix cell object cann't free normally.
5 years ago
e8ca9c12c2
add alternate path for stdclient library
5 years ago
5d4144de11
bugfix(side effect): fix adding wrong control depend between AllReduce and GetStatus.
5 years ago
9bda080bb5
!260 refactor padding strategy
...
Merge pull request !260 from lianliguang/refactor-padding-strategy
5 years ago
1ab430072e
!232 [Auto parallel] Model the memory_cost in cost model
...
Merge pull request !232 from Xiaoda/model-memory-cost-in-auto-parallel
5 years ago
8674e0ad96
!288 fix nopnode output bug
...
Merge pull request !288 from kisnwang/fix-nopnode-output-bug
5 years ago
94589ce611
!226 expend conv stride and dilation to 2d
...
Merge pull request !226 from wangnan39/expend_conv_stride_to_2d
5 years ago
2aad57c595
getnext disable memory reuse
5 years ago
a4cf9028ee
!275 fix bert example script bugs
...
Merge pull request !275 from yoonlee666/master
5 years ago
0ca4ceb73f
!281 bug fix in gpu aware quantizaiton ops
...
Merge pull request !281 from SanjayChan/origin/quant_liantiao
5 years ago
ae675c5cf8
fix nopnode output bug
5 years ago
652ab6c386
add test case for aware quantizaiton
5 years ago
2604acedcb
extend conv stride and dilation to 2d
5 years ago
5d225f934f
change the padding strategy & refactor insert transdata
5 years ago
46c590b787
Add new flag to watch for the need of Renormalize pass when parsing the python input program
...
Signed-off-by: Hoai Linh Tran h00472437 <hoai.linh.tran@huawei.com>
Add optimizer checking: For a group of passes in "optimizer", if flagged then it will check and collect the newly generated nodes without types (i.e. abstract() == nullptr).
Before calling Renormalize(), the optimizer will check if there is any node needed retyping. If not the Renormalize pass will not be called.
Add checking for non-null abstract but still needs renorm;
Add flags to Substitution to help watching Renormalize
Simpler pass result checker, change Bool to Enum type
5 years ago
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.
5 years ago
c9fba7f091
!283 adapt graphengine update
...
Merge pull request !283 from jjfeing/master
5 years ago
f1fa2a9941
!273 [MD] update subset random sampler in minddataset
...
Merge pull request !273 from liyong126/mindrecord_subset_sampler_python
5 years ago
0b75e289b1
Dump the perf data by executed sequence, not alphabetic.
5 years ago
1fc576c502
adapt graphengin update
5 years ago
511acd2961
!269 Modify the improper code in bug info.
...
Merge pull request !269 from ZhangQinghua/master
5 years ago
7ffb8bb19f
!250 Add nn.pad to support three modes
...
Merge pull request !250 from casgj/gaojing_new4
5 years ago
mindspore-ci-bot
limingqi107
zhoufeng
Yanjun Peng
ZPaC
dengwentao
VectorSL
maoweiyong
万万没想到
zjun
YuJianfeng
jonyguo
Wei Luning
kswang
wangnan39@huawei.com
gong chen
yanghaoran
jojobugfree
chenzomi
lianliguang
Hoai Linh Tran h00472437
Xiaoda Zhang
Zhang Qinghua
jjfeing