i-robot
|
e2790c37bd
|
!27157 [MS][LITE][ToD]Add LSTM grad op for backpropagation.
Merge pull request !27157 from amirama/export_amirama
|
4 years ago |
amiram
|
81e3f2c0ca
|
add lstm grad
|
4 years ago |
i-robot
|
6a8f5f3cbc
|
!26988 [MS][LITE][CPU] avx512
Merge pull request !26988 from liuzhongkai/avx512
|
4 years ago |
sunsuodong
|
cfb5477c1d
|
fix high version gcc build
|
4 years ago |
lzk
|
6978726378
|
avx512
|
4 years ago |
i-robot
|
647c22fa64
|
!26386 optimize cpu fused adam
Merge pull request !26386 from kisnwang/optimize-fused-adam
|
4 years ago |
chenping38
|
3a65a28895
|
mindir 端侧验证成功
|
4 years ago |
i-robot
|
8cb8660051
|
!26970 [MS][LITE]support call parameter populate
Merge pull request !26970 from mengyuanli/populate_call_para
|
4 years ago |
i-robot
|
acbaff17f7
|
!26960 [MSLITE][DEVELOP] fix bug of some op attr check
Merge pull request !26960 from yangruoqi713/master_fuzz
|
4 years ago |
mengyuanli
|
0da2c8d145
|
add call parameter and populate call
|
4 years ago |
yangruoqi713
|
2e57f87be2
|
[MSLITE][DEVELOP] fix bug of some op attr check
|
4 years ago |
i-robot
|
5f0f595d1b
|
!26907 [MS][LITE][develop] add fp16 kernel
Merge pull request !26907 from sunsuodong/add_fp16_kernel_b
|
4 years ago |
i-robot
|
7ee6e3060d
|
!26637 [MS][LITE][develop] add fp16 kernel
Merge pull request !26637 from sunsuodong/add_fp16_kernel
|
4 years ago |
sunsuodong
|
533a6574b2
|
add fp16 kernel
|
4 years ago |
sunsuodong
|
90b1318fcc
|
add fp16 kernel
|
4 years ago |
zhaosida
|
7de4c114b3
|
fix avx512 compile option
|
4 years ago |
i-robot
|
7fb6df8384
|
!26545 cast between fp16 int8 optimize
Merge pull request !26545 from zhaozhenlong/lite/issue/cast-fp16-int8
|
4 years ago |
zhangxuetong
|
9ef0fa84bb
|
support parameter cache and distribution predict
|
4 years ago |
kswang
|
427610d3a9
|
add global norm for fused adam
|
4 years ago |
i-robot
|
821aba8907
|
!26245 sync commercial pclint clean
Merge pull request !26245 from zhaodezan/master
|
4 years ago |
sunsuodong
|
47785a8f12
|
fix_deconv_winograd_fp16
|
4 years ago |
i-robot
|
3269c9b881
|
!26335 Support MindSpore on MacOS
Merge pull request !26335 from xulei/ms_mac_compile_br
|
4 years ago |
zhaozhenlong
|
e885bce606
|
cast between fp16 int8
|
4 years ago |
sunsuodong
|
bf208e243e
|
add fp16 kernel
|
4 years ago |
wangyanling
|
ac30c5f317
|
optimize matmul broadcast
|
4 years ago |
xiefangqi
|
24ac2382ff
|
mac compile
|
4 years ago |
i-robot
|
c7805f0fc7
|
!26294 [MS][LITE]Fix fuzz
Merge pull request !26294 from gongdaguo/master_fuzz
|
4 years ago |
i-robot
|
dafa363bde
|
!26326 matmul int8 sdot
Merge pull request !26326 from zhaozhenlong/lite/issue/matmul-int8
|
4 years ago |
zhaozhenlong
|
34b46257b8
|
matmul int8 sdot
|
4 years ago |
z00512249
|
21699e6855
|
fix bugs of gmres for backend cpu
|
4 years ago |
gongdaguo
|
51d4299bf3
|
fix fuzz
|
4 years ago |
sunsuodong
|
a7f84e86d0
|
add fp16 kernel
|
4 years ago |
i-robot
|
50bbdfef17
|
!26319 [lite]fix codedex
Merge pull request !26319 from 徐安越/master
|
4 years ago |
xuanyue
|
6b6d2bd196
|
fix codedex
|
4 years ago |
gongdaguo
|
a2afb88f0d
|
add int8 clip
|
4 years ago |
i-robot
|
a14a777464
|
!26050 [MS][LITE]Add int8 clip
Merge pull request !26050 from gongdaguo/add_int8_clip
|
4 years ago |
xuanyue
|
3cecadc8d9
|
fix fuzz bug and element size overflow
|
4 years ago |
i-robot
|
c8e303f9a7
|
!26253 [MS][LITE][develop] fix vs build and run bug
Merge pull request !26253 from sunsuodong/fix_vs_build
|
4 years ago |
sunsuodong
|
83875cf9b8
|
fix vs build
|
4 years ago |
sunsuodong
|
3d1066b6a0
|
add fp16 resize kernel
|
4 years ago |
zhaodezan
|
88b73e340d
|
sync commercial pclint clean
|
4 years ago |
i-robot
|
9a0be9152c
|
!26174 [MS][LITE] add tf unpack parser and allow ConstantOfShape to output tensor without elements
Merge pull request !26174 from XianglongZeng/myms_new
|
4 years ago |
i-robot
|
ddf345bdb3
|
!26185 [MSLITE][DEVELOP] fix bug of memory leak
Merge pull request !26185 from yangruoqi713/master
|
4 years ago |
i-robot
|
1c71ff0392
|
!26181 [MSLITE] deconv range expanddims transpose fuzz bug
Merge pull request !26181 from ling/pr
|
4 years ago |
zengxianglong
|
ac089bde12
|
add tf unpack parser and allow ConstantOfShape to output tensor without elements
|
4 years ago |
ling
|
bdeeffa84a
|
[MSLITE] deconv range expanddims transpose fuzz bug
|
4 years ago |
yangruoqi713
|
fef3426172
|
[MSLITE][DEVELOP] fix bug of memory leak
|
4 years ago |
mengyuanli
|
af1deb1c10
|
fix bug of lite
|
4 years ago |
gongdaguo
|
136b9ba298
|
add int8 clip
|
4 years ago |
i-robot
|
b4c1d771c1
|
!25855 [ms][lite][cpu] deconv optimize
Merge pull request !25855 from liuzhongkai/deconv_114
|
4 years ago |