ZPaC
a4d2a3e6a1
Add collective operation impl.
4 years ago
ZPaC
2b7429c5d2
1.Purge not used API.
2.Adapt for collective_init.h
4 years ago
ZPaC
055a493903
Add implement of collective manager API
4 years ago
i-robot
20c6263123
!26520 Add collective API for python layer
Merge pull request !26520 from ZPaC/dir-of-distributed
4 years ago
ZPaC
e243c1e021
Add collective API for python layer
4 years ago
i-robot
3269c9b881
!26335 Support MindSpore on MacOS
Merge pull request !26335 from xulei/ms_mac_compile_br
4 years ago
xiefangqi
24ac2382ff
mac compile
4 years ago
ZPaC
6d236baaf2
Add MindSpore communciation framework network building and communication lib implementation.
4 years ago
ZPaC
87057fdc27
Dynamic load ompi and nccl
4 years ago
ZPaC
9e18bad126
Add nvidia collective lib implementation.
4 years ago
ZPaC
0b732431a1
Add MPI implementation.
4 years ago
ZPaC
e5789757d6
Add collective adapter
4 years ago
limingqi107
2a5d9f964b
unified runtime support the sub graph sink
4 years ago
i-robot
b271aa7a25
!24969 device address add the key of device
Merge pull request !24969 from limingqi107/new_actor_runtime
4 years ago
maning202007
8ee33007b1
add const to the data_ptr definition
Fix pclint for debugger
fix bugs
fix duplicated code issue
Fix the depth of method over 4 issue
fix comments
fix CI errors
remove redundancy arguments
4 years ago
limingqi107
be100476d6
device address add the key of device
4 years ago
fangzehua
53009276d1
add cpu dynamic memory
4 years ago
i-robot
c5585a9451
!24467 Add GraphKernelFlags into namespace mindspore::graphkernel
Merge pull request !24467 from DeshiChen/0918_namespace
4 years ago
dayschan
7502345c8f
Add GraphKernelFlags into namespace mindspore::graphkernel
4 years ago
LaiYongqiang
79599546ee
add ascend memory adapter for ascend memory management
4 years ago
i-robot
06b0beced7
!22731 GraphKernel supports CPU
Merge pull request !22731 from DeshiChen/0901_graphkernel_cpu
4 years ago
dayschan
32ecd8ee79
GraphKernel supports CPU
only Linux system is supported now.
change the default value of `ENABLE_AKG` to off, and controlled by option `-K`.
the `ENABLE_AKG` is auto enabled when `ENABLE_GPU` or `ENABLE_D` is on.
since now, we can use `ENABLE_AKG` to control the compilation of graphkernel
and akg codes.
fix usage description for option "-K", it should be "[-K on|off]".
LLVM is required by akg for cpu kernels, so AKG for cpu is default disabled now.
4 years ago
limingqi107
0a9cbd2220
imporve performance of CPU LSTMGrad
4 years ago
i-robot
5bd7e4def9
!24156 Move InsertCastCpu before BuildKernel
Merge pull request !24156 from DeshiChen/0926_insert_cast_cpu
4 years ago
hwjiaorui
eaa3dc57b7
add dynamic shape attr pass
4 years ago
dayschan
5ed932e5fd
Move InsertCastCpu before BuildKernel
It's necessary to immediately insert cast for the nodes that dtype was changed in SelectKernel,
instead of after BuildKernel. otherwise, some fusion passes cannot work as expected,
like Graph-Kernel optimization.
By now, the shape of LSTM's reserved output will be changed in InitKernel (called by BuildKernel),
and this output is only used by its gradient operator, so we can ignore the reserved output in this pass.
4 years ago
i-robot
811dd1778b
!23959 fix link error if compiled with -s on when import mindspore
Merge pull request !23959 from yanghaitao/yht_sync_shangyong
4 years ago
lizhenyu
0b6d7e4adb
fix code review warning
4 years ago
yanghaitao1
7f216aac41
fix link error if import mindspore
4 years ago
sabrinasun
1b8b453373
add remain isolation for dump
4 years ago
djc
b077aa1cab
[feat] [assistant] [I3T96T] add new Dataset operator CMUARCTICDataset
4 years ago
djc
4e6f7dc97d
[feat] [assistant] [I3T96X] add new Dataset operator LibriSpeechDataset
4 years ago
limingqi107
5766234426
code review of gpu backend
4 years ago
limingqi107
09ceb1cf5b
unified runtime pclint fixed
4 years ago
limingqi107
fe36a158af
unified runtime pclint fixed
4 years ago
lizhenyu
9404bfb9d9
fix pclint warnings
4 years ago
buxue
733f765063
add security compilation options
4 years ago
lizhenyu
209dc5f584
fix codedex warnings
4 years ago
i-robot
b27a0fc73b
!19856 [bugfix] Reinitialize for cpu kernels which do not support multi-thread
Merge pull request !19856 from zyli2020/mindrt_debug
4 years ago
lizhenyu
cde12f6d98
bugfix: reinitialize for cpu kernels which do not support multi-thread
4 years ago
limingqi107
ae46d0beb9
fix bug of CPU not support more then 2G copy
4 years ago
i-robot
83bcf936b6
!19403 [bugfix] CPU PyNative hook occur select op error
Merge pull request !19403 from zyli2020/fix_mindtest_bug
4 years ago
lizhenyu
823621d8f3
[bugfix] CPU PyNative hook occur select op error
4 years ago
lizhenyu
efcf910b80
Adjust CPU optimize graph pass
4 years ago
limingqi107
e761655a42
actor runtime support CPU dump
4 years ago
limingqi107
9d2bf9cf6b
actor runtime memory optimize of communication op
4 years ago
limingqi107
7f634d12f0
actor runtime support GraphKrenel
4 years ago
mindspore-ci-bot
099a73ef0e
!17523 fix CPU lstm bug in pynative mode
From: @huaweib
Reviewed-by: @zhoufeng54,@kisnwang
Signed-off-by: @kisnwang
4 years ago
limingqi107
21d168e980
optimize the performance of actor runtimie dynamic shape
4 years ago
baihuawei
e594463960
fix cpu lstm and precision issues
4 years ago