zhoufeng
b7bb53ff73
build libmindspore_backend.so
...
Signed-off-by: zhoufeng <zhoufeng54@huawei.com>
2022-03-17 19:14:45 +08:00
zhoufeng
f49b195c39
extract common as an independent shared library
...
Signed-off-by: zhoufeng <zhoufeng54@huawei.com>
2022-02-26 10:55:51 +08:00
shenwei41
ff75314719
upgrade ascend 20220211
2022-02-25 21:19:45 +08:00
marui
d47e79b04c
Refactor ccsrc directories and CMakeLists files
2022-02-13 14:37:51 +08:00
huangbingjian
63a89925ff
remove ControlDepend and its use
2021-04-07 11:30:40 +08:00
liubuyu
518818fbef
reshape type for 3d nodes
2021-03-10 08:56:10 +08:00
He Wei
7d9a783993
[auto-monad] Support side-effects by auto-monad
...
The basic idea is: exploits data dependency to control the execution order
of side-effect operations, and keep the semantics of ANF unchanged.
The ControlDepend primitive is removed and there are two primitives added:
1. UpdateState:
```
a = Assign(para, value)
```
became:
```
a = Assign(para, value, u)
u = UpdateState(u, a)
```
2. Load:
```
x = Add(para, value)
```
became:
```
p = Load(para, u)
x = Add(p, value)
u = UpdateState(u, p)
```
2021-02-08 09:01:15 +08:00
Yi Huaijie
d7faa77b5e
support int64 shape
2020-11-05 21:15:49 +08:00
fary86
fcbb3e0edc
Refactor ms_context implementation
2020-08-30 22:51:40 +08:00
liubuyu
d81862a916
decoupling core and context
2020-08-05 08:15:14 +08:00
WilliamLian
0179724dcd
spilit unspported transdata to two transdata from special format -> defualt -> default -> special
2020-07-28 14:23:31 +08:00
chenfei
1f1a07e645
don't insert assign from condition to true branch of while
2020-07-26 14:38:42 +08:00
root
1b6f85dec8
split tuple parameter to parameters
...
add function trans tuple to maketuple
2020-07-23 16:46:55 +08:00
liubuyu
43c79eb853
mindspore path adjust
2020-07-14 18:07:28 +08:00
He Wei
43e0967024
Decouple ir::Tensor class from python
2020-06-24 10:28:59 +08:00
gong chen
a6dfa281ea
Init GraphKernel.
...
- It provides a unified style to express graph and kernel for user.
- It provides a unified IR to represent graph and kernel for developer.
- It breaks the boundary between graph and kernel.
- It provides more opportunities to do compile optimization.
2020-06-20 22:31:54 +08:00
yujianfeng
f15cb6b7c9
Add sort by index for each group of AllReduce
2020-06-18 09:18:01 +08:00
WilliamLian
b86016a26f
remove the useless transdata and cast of control depend node
2020-06-15 14:54:47 +08:00
chujinjin
7465abc798
optimize transdata for pynative
2020-06-05 14:10:55 +08:00
mindspore-ci-bot
04398cf88e
!1433 add tensor_minnie and separate py from ir
...
Merge pull request !1433 from leopz/tensor_base
2020-05-28 16:00:56 +08:00
leopz
4508134ceb
add tensor_minnie and separate py from ir
2020-05-28 14:26:26 +08:00
huanghui
1d65ae598a
extract const_to_attr_strided_slice_grad pass
2020-05-27 17:30:17 +08:00
yujianfeng
31d8db9fc6
Add broadcast fusion pass
2020-05-13 15:30:38 +08:00
lianliguang
00e4306518
convert all tuple output to maketuple
2020-04-17 17:31:42 +08:00
mindspore-ci-bot
9be1a01db6
!85 Optimize depend edge with make tuple input
...
Merge pull request !85 from YuJianfeng/master
2020-04-02 22:00:18 +08:00
Wei Luning
73ba399364
remove ge depend in cpu
2020-04-02 19:38:42 +08:00
YuJianfeng
7f53bb062d
Optimize depend edge with make tuple input
2020-04-02 11:33:34 +08:00
zhunaipan
930a1fb0a8
initial version
...
Signed-off-by: leonwanghui <leon.wanghui@huawei.com>
2020-03-27 22:54:54 +08:00