chenjianping
1286767d0e
support building on windows
6 years ago
buxue
62807da0c4
Develop operator Unfold,take the ge backend,dock with tbe's ExtractImagePatches operator
6 years ago
zjun
b55f1324a3
Fix kernel query log level
6 years ago
wilfChen
cc93646207
gpu concat kernel support 4 inputs
6 years ago
wilfChen
5b7790a2a7
Gpu Slice kernel performance improvement
6 years ago
mindspore-ci-bot
7d7c9c4fee
!311 GPU add akg kernel select
Merge pull request !311 from VectorSL/select
6 years ago
mindspore-ci-bot
57aee805ce
!274 GPU multiple stream feature
Merge pull request !274 from ZPaC/gpu-backend-supports-multiple-streams
6 years ago
ZPaC
3ea3d9e5a4
1.GPU supports multiple streams.
2.GPU commnication stream and compute stream overlap.
6 years ago
VectorSL
9e372073e2
gpu add assigin
6 years ago
VectorSL
d248b05a98
gpu add kernel select
6 years ago
mindspore-ci-bot
9bda080bb5
!260 refactor padding strategy
Merge pull request !260 from lianliguang/refactor-padding-strategy
6 years ago
mindspore-ci-bot
94589ce611
!226 expend conv stride and dilation to 2d
Merge pull request !226 from wangnan39/expend_conv_stride_to_2d
6 years ago
chenzomi
652ab6c386
add test case for aware quantizaiton
6 years ago
wangnan39@huawei.com
2604acedcb
extend conv stride and dilation to 2d
6 years ago
lianliguang
5d225f934f
change the padding strategy & refactor insert transdata
6 years ago
mindspore-ci-bot
e95945969b
!270 fix complite error while using .so with clang on macos
Merge pull request !270 from SanjayChan/mac_os
6 years ago
mindspore-ci-bot
878a076dbd
!200 Add quantization aware training related operators
Merge pull request !200 from chengang/quant
6 years ago
chenzomi
65a49a21e5
fix complite error while using .so with clang on macos
6 years ago
mindspore-ci-bot
c478be0ff0
!248 Fix aicpu getnext bug
Merge pull request !248 from zjun/fix_getnext_bug
6 years ago
panfengfeng
cc80c76687
add quantizaiton gpu op
6 years ago
buxue
0db45147a4
dock relu6 for open source process and fix pow bprop
6 years ago
zjun
7c19a4c2fd
fix getnext op bug
6 years ago
mindspore-ci-bot
0d23520016
!233 modify Float64 to float64, del double
Merge pull request !233 from jjfeing/master
6 years ago
jjfeing
4fb76e6d7e
modify Float64 to float64, del double
6 years ago
lizhenyu
8cd3308a4c
change package name akg to _akg
6 years ago
zjun
16296da5c7
add aicpu opinfo register
6 years ago
mindspore-ci-bot
cc75cb357c
!168 remove mindspore::make_unique and make_unique.h
Merge pull request !168 from Alexey_Shevlyakov/remove_mindspore_make_unique
6 years ago
buxue
7541d3b067
Develop op MaxPoolWithArgMax
6 years ago
Alexey Shevlyakov
84d780c1a4
remove make_unique.h
6 years ago
zjun
651d9aae40
add new mode for operator info register
6 years ago
zhaoting
0c81759ae6
add YOLOv3 infer scipt and change dataset to MindRecord
6 years ago
Wei Luning
73ba399364
remove ge depend in cpu
6 years ago
mindspore-ci-bot
8f76c28017
!56 remove cce dependency
Merge pull request !56 from jojobugfree/dev-rm-cce
6 years ago
jojobugfree
e6c15b82c8
remove cce dependency
6 years ago
wenchunjiang
5a00d8cb58
This fixes an issue about mindspore process cannot exit when calling python api op_select_format failed in select kernel steps.
Previously function op_select_format and check_supported raise an exception directly on the tbe_process python side, but we don't deal with the exception, and raise an exeception on c++ side to frontend ME, that will cause some conflict when recycle resource on ME and tbe_process python interpreter.
This changes adding try...catch in function op_select_format and check_supported on the python side, and return the Exception string to c++ side, so that we can raise an exception to frontend ME and ME will deal with resouce clearning and exit.
6 years ago
zhunaipan
930a1fb0a8
initial version
Signed-off-by: leonwanghui <leon.wanghui@huawei.com>
6 years ago