lizhenyu
|
a62e9fbda2
|
remove tuple_get_item combination with make_tuple pattern from graph
|
5 years ago |
lizhenyu
|
9d0b74b473
|
move add graph manager to gpu session
|
5 years ago |
leopz
|
a9e30a96db
|
separate py from debug and utils
|
5 years ago |
lizhenyu
|
a25b84055c
|
refine data copy in multi-graph
|
5 years ago |
mindspore-ci-bot
|
2fef359c4d
|
!1396 Cancel NoOp optimizer on gpu backend until memory reuse ready
Merge pull request !1396 from chenweifeng/NoOp
|
5 years ago |
wilfChen
|
56d751b5d2
|
Cancel NoOp optimizer in GPU until memory reuse ready
|
5 years ago |
xiefangqi
|
8b585c76e2
|
fix py_transform gpu hang problem
|
5 years ago |
mindspore-ci-bot
|
a915cc3bd9
|
!1225 gpu NoOp optimizer
Merge pull request !1225 from chenweifeng/NoOp
|
5 years ago |
wilfChen
|
23b4b4d106
|
Gpu NoOp optimizer
|
5 years ago |
lvliang
|
2e284cb849
|
pynative-fix-topk-run-failed
|
5 years ago |
chujinjin
|
13bf42ba6f
|
abstract input tensor
|
6 years ago |
YuJianfeng
|
39945d0f79
|
Add AllGather fusion pass
|
6 years ago |
mindspore-ci-bot
|
9b80c160a0
|
!276 pynative-add-op-supported
Merge pull request !276 from JoyLvliang/pynative-add-op-supported
|
6 years ago |
lvliang
|
ffe8b5d3ec
|
pynative-add-op-supported
|
6 years ago |
laiyongqiang
|
3e05f50f5f
|
getnext_memcpy_elimination
|
6 years ago |
ZPaC
|
3ea3d9e5a4
|
1.GPU supports multiple streams.
2.GPU commnication stream and compute stream overlap.
|
6 years ago |
kswang
|
fb343bd607
|
add mem manager
|
6 years ago |
mindspore-ci-bot
|
3289d7bb69
|
!109 Fix some typo errors in session and device module
Merge pull request !109 from leonwanghui/typo-fix
|
6 years ago |
Wei Luning
|
73ba399364
|
remove ge depend in cpu
|
6 years ago |
leonwanghui
|
fdb47860cc
|
Fix some typo errors in session and device module
Signed-off-by: leonwanghui <leon.wanghui@huawei.com>
|
6 years ago |
zhunaipan
|
930a1fb0a8
|
initial version
Signed-off-by: leonwanghui <leon.wanghui@huawei.com>
|
6 years ago |