26e39a9692
!13670 fix centernet loss error
...
From: @caojian05
Reviewed-by: @oacjiewen,@c_34
Signed-off-by: @c_34
4 years ago
dc8ec9d87f
[auto-monad] Prepare to support recursive calls
...
1. Find and mark recursive calls and graphs;
2. Do not eliminate argument Assign for recursive graphs;
3. Disable tail call optimization for recursive graphs;
4. Add attribute to output parameter Assign to disable stack push.
4 years ago
0da0f72c6c
!13329 [MS][RDR] optimize saving FuncGraph to using same params with save_graphs
...
From: @louie5
Reviewed-by: @zh_qh
Signed-off-by:
4 years ago
cf5eaf8590
!13050 Don't insert UpdateState for HyperMap func graph call, move auto monad eliminator out from CSE, and eliminate auto monad nodes for output node.
...
From: @zh_qh
Reviewed-by:
Signed-off-by:
4 years ago
47085ceefe
opt warpctc performance
4 years ago
e4695d47d1
Fix bugs: 13BZDF, 138MR1. Support CPU BatchNormGrad
4 years ago
e3056ed9b2
show accurate code line when use uninitialized var in for
4 years ago
ec7f9f395a
fix cmake lint
...
Signed-off-by: zhoufeng <zhoufeng54@huawei.com>
4 years ago
e43005242c
concat output fix npu pass
4 years ago
6090bf0849
gelu optimize
4 years ago
9aa05b177f
fixed the wait usage of shell script
4 years ago
f0a85b9220
fix layernorm grad bug
4 years ago
c59349e096
avoid double free for tdt channel
4 years ago
d436b9ee98
!11754 [MS][LITE][Develop]fix context error for unfied API
...
From: @lx0095
Reviewed-by: @zhang_xue_tong,@zhanghaibo5
Signed-off-by: @zhang_xue_tong
4 years ago
21a2c7bcbe
fix centernet loss error.
4 years ago
ad7ce7881d
fix context error for unfied API
4 years ago
9f33381425
modify example
4 years ago
6b9de24797
!13659 replace tensorflow io API with python standard library API
...
From: @zhouneng2
Reviewed-by: @liangchenghui,@c_34
Signed-off-by: @liangchenghui
4 years ago
d965070755
merge path
4 years ago
2e9a105709
!13655 [MS_LITE] fix ios
...
From: @YeFeng_24
Reviewed-by: @hangangqiang,@zhang_xue_tong
Signed-off-by: @hangangqiang
4 years ago
72f64291ec
!13622 [lite]fix bug for tf
...
From: @xu_anyue
Reviewed-by: @hangangqiang,@jpc_chenjianping
Signed-off-by: @hangangqiang
4 years ago
38b48fb0e8
!13608 Add CPU LogSoftMax
...
From: @zhao_ting_v
Reviewed-by: @wuxuejian
Signed-off-by: @wuxuejian
4 years ago
bfffbe3ca9
!13649 Add float64 support for list of gpu ops for end of version 1.1
...
From: @peilin-wang
Reviewed-by: @liangchenghui,@wuxuejian
Signed-off-by: @liangchenghui
4 years ago
d2777500ad
reg-10
4 years ago
686d3ca18d
!13560 Add need_check_supported for MaxPoolWithArgmax and MaxPoolGradWithArgmax.
...
From: @liu_xiao_93
Reviewed-by:
Signed-off-by:
4 years ago
dac00c6db3
!13611 [MSLITE][Develop] fix LogicalAnd
...
From: @sunsuodong
Reviewed-by: @zhanghaibo5,@zhang_xue_tong
Signed-off-by: @zhang_xue_tong
4 years ago
b8b833b762
!13558 fix some bugs in resnet, ssd and naml
...
From: @zhao_ting_v
Reviewed-by:
Signed-off-by:
4 years ago
7df661fbaf
!13596 modify comm fusion and grad accu api
...
From: @gong_zi_yan
Reviewed-by: @guoqi1024,@stsuteng
Signed-off-by: @stsuteng
4 years ago
62fbbf9eb7
fix explainer doc bugs
4 years ago
542142af8a
!13565 [Pynative]Fix Pynative AMP Backprop Bugs
...
From: @chenyijie6
Reviewed-by: @wilfchen,@chujinjin
Signed-off-by: @chujinjin
4 years ago
d8bf9ab8da
fix fcns2
4 years ago
dafd2713ac
!13637 [MD] Revert profiling get device id rule
...
From: @xiefangqi
Reviewed-by: @heleiwang,@liucunwei
Signed-off-by: @liucunwei
4 years ago
e212d12257
!13347 add cpu mixprecision
...
From: @huaweib
Reviewed-by: @zhoufeng54
Signed-off-by:
4 years ago
5c6f0ed2f4
!13648 Add float64 support to cumsum
...
From: @peilin-wang
Reviewed-by: @liangchenghui,@wuxuejian
Signed-off-by: @liangchenghui
4 years ago
5ff9c08680
!13515 cpp api modify
...
From: @zhoufeng54
Reviewed-by:
Signed-off-by:
4 years ago
8232628001
!13597 refactor normfusion
...
From: @wangzhe128
Reviewed-by: @hangangqiang,@hangangqiang,@zhang_xue_tong
Signed-off-by: @hangangqiang
4 years ago
5504116718
!13647 BroadcastTo add general -1 dim behavior
...
From: @peilin-wang
Reviewed-by: @liangchenghui,@wuxuejian
Signed-off-by: @liangchenghui
4 years ago
07ef04ad4b
!13614 [MD] Fix Canny when the ksize is set to 7
...
From: @tiancixiao
Reviewed-by: @heleiwang,@liucunwei
Signed-off-by: @liucunwei
4 years ago
ae6a4eb99f
!13569 fix minddata issues
...
From: @luoyang42
Reviewed-by: @heleiwang,@liucunwei
Signed-off-by: @liucunwei
4 years ago
a432828f18
remove zero or negative check in rsqrt_fp16 op
4 years ago
f508d63546
!13633 add fix fcn8s
...
From: @wukesong
Reviewed-by: @liangchenghui,@c_34
Signed-off-by: @liangchenghui
4 years ago
8754aaeb74
add CPU LogSoftMax
4 years ago
57cb024e03
add float64 support for ops for end of version 1.1
...
fix typo
4 years ago
7bb35d8ce4
!13635 add float64 support to Addn gpu
...
From: @TFbunny
Reviewed-by: @liangchenghui,@wuxuejian
Signed-off-by: @liangchenghui
4 years ago
04e3dbaad0
!13559 broadcast_to op supported on cpu
...
From: @wangyanling10
Reviewed-by:
Signed-off-by:
4 years ago
8ab9d99714
!13507 dynamic shape adapter
...
From: @liubuyu
Reviewed-by: @zhoufeng54,@jjfeing
Signed-off-by: @jjfeing
4 years ago
c0ebc36c78
change cumsum op python to allow for float64 input
4 years ago
6cead43bdf
add general -1 dim behavior for BroadcastTo op
4 years ago
30cb84cb88
fix st bug
4 years ago
99ebe71b39
Fix Pynative AMP Backprop Bugs
4 years ago
mindspore-ci-bot
He Wei
gengdongjie
yoni
buxue
zhoufeng
zhaozhenlong
lzk
chendongsheng
zhengjun10
ms_yan
CaoJian
lixian
Jiaqi
zhujingxuan
yefeng
lixiaohui
wukesong
hangangqiang
zhaoting
Peilin Wang
caifubi
chenyijie6