VectorSL
|
7d7b81348e
|
gpu fix conv bug
|
5 years ago |
mindspore-ci-bot
|
13d1738ff3
|
!4706 fix SmoothL1Loss gpu kernel
Merge pull request !4706 from Peilin/smoothL1Loss-fix
|
5 years ago |
mindspore-ci-bot
|
87c7cf46b6
|
!4475 Add 2 reserve output for BatchNormGrad
Merge pull request !4475 from JonathanY/batchnormgrad
|
5 years ago |
lizhenyu
|
839ec02542
|
Add FusedBatchEx support
|
5 years ago |
mindspore-ci-bot
|
a3dae6344b
|
!4758 GPU kernel support NHWC
Merge pull request !4758 from VectorSL/nhwc-support
|
5 years ago |
mindspore-ci-bot
|
b9bde4c826
|
!4559 codex warning
Merge pull request !4559 from chenweifeng/warning
|
5 years ago |
VectorSL
|
e939d61a2c
|
conv pooling pad support NHWC
|
5 years ago |
wilfChen
|
061bbf1f87
|
code warning
|
5 years ago |
lizhenyu
|
19d50fea3e
|
add FusedBatchNormGradEx gpu kernel
|
5 years ago |
Peilin Wang
|
0d5220d33c
|
modified documentation and gpu kernel for smoothL1Loss
fix pylint
changed doc and code for SmoothL1Loss to be same a dchip. fixed grad kernel
fix ci
|
5 years ago |
Jonathan Yan
|
72d2597cb7
|
Add 2 reserve output for BatchNormGrad v1
|
5 years ago |
lizhenyu
|
7ddddc41a9
|
add FusedBatchNoramEx gpu kernel
|
5 years ago |
lizhenyu
|
d667d6ee92
|
bugfix:SigmoidCrossEntropyWithLogitsGrad need multiply dout
|
5 years ago |
Jonathan Yan
|
19d7488195
|
ROIAlignGrad kernel output size is wrong
|
5 years ago |
mindspore-ci-bot
|
01962afd23
|
!4024 Support half data type in ROIAlign/ROIAlignGrad Kernel
Merge pull request !4024 from JonathanY/roihalf
|
5 years ago |
mindspore-ci-bot
|
eb84ae4593
|
!4048 Fix broadcast, scatternd, reduce ops.
Merge pull request !4048 from linqingke/new_ops
|
5 years ago |
hanhuifeng2020
|
ab6f7420b5
|
modify some bug and add test case for gpu dropout op
|
5 years ago |
linqingke
|
fb405ee6f4
|
broadcast, slice, scatter_nd ops optimizer.
|
5 years ago |
mindspore-ci-bot
|
8f17535045
|
!3831 CUDA - GPU MirrorPad New Op
Merge pull request !3831 from danishnxt/GPU_One
|
5 years ago |
danish
|
081249b53f
|
commit 1 - mirror pad
commit 2
lint fix
lint fix 2
updated backprop + st test
test_file_fix
test_file_fix_2
fixed header_guards
comments addressed
clangFormatFix
|
5 years ago |
Jonathan Yan
|
43094bf78e
|
suport half for roi align
|
5 years ago |
mamba_ni
|
4fce4c7c34
|
support img2col for resnet50_thor GPU
primitive for im2col
fix bug
clang code format
clang format fix
fix pylint
fix license
delete useless code
|
5 years ago |
mindspore-ci-bot
|
49ba473bcc
|
!3803 add gpu klDivLoss op
Merge pull request !3803 from baihuawei/loss
|
5 years ago |
mindspore-ci-bot
|
82b103a740
|
!3780 add gpu BinaryCrossEntropy
Merge pull request !3780 from baihuawei/losscuda
|
5 years ago |
baihuawei
|
9eca56635d
|
add KLDiv loss
|
5 years ago |
baihuawei
|
aa9ea1707c
|
add binary cross entropy
|
5 years ago |
mindspore-ci-bot
|
afce1c3a40
|
!3341 GPU maxpool with argmax op
Merge pull request !3341 from tom_chen/maxpool_with_argmax
|
5 years ago |
tom__chen
|
5c3be0114f
|
add maxpool_with_argmax/grad cuda kernel
|
5 years ago |
Jonathan Yan
|
ad40e00228
|
roi align grad v1
|
5 years ago |
mindspore-ci-bot
|
380db207e8
|
!3344 GPU Pad Op
Merge pull request !3344 from danishnxt/PR-GPU-Pad
|
5 years ago |
mindspore-ci-bot
|
183cf5cf5d
|
!3285 Add Encode,Decode,SGD,floordiv,ScatterNd,GatherNd ops.
Merge pull request !3285 from linqingke/gpu_ops
|
5 years ago |
danish
|
adf59d2ded
|
commit one - PadOp Support files
clang format fix
fix format - cpplint
|
5 years ago |
linqingke
|
f679568d86
|
gpu ops code and test case.
|
5 years ago |
mindspore-ci-bot
|
d15b4c5d61
|
!3201 RoI Align GPU kernel
Merge pull request !3201 from JonathanY/main
|
5 years ago |
Jonathan Yan
|
661b993475
|
roi align v1
|
5 years ago |
mindspore-ci-bot
|
bad04340d6
|
!3240 GPU update CAST and conv2d_pad
Merge pull request !3240 from VectorSL/update
|
5 years ago |
mindspore-ci-bot
|
1625a27ae5
|
!3244 Support 2-dimensions target of CTCLossV2
Merge pull request !3244 from yangyongjie/yangyongjie
|
5 years ago |
yangyongjie
|
67d1ba0fc3
|
support 2-dimension target of CTCLossV2
|
5 years ago |
VectorSL
|
90f15df037
|
add int64-->fp16 and update conv pad
|
5 years ago |
liubuyu
|
76dc80e7b7
|
Unified code style
|
5 years ago |
VectorSL
|
140174182d
|
gpu add fusion: replace momentum cast
|
5 years ago |
liubuyu
|
43c79eb853
|
mindspore path adjust
|
5 years ago |