Gitiles
Code Review
Sign In
review.mlplatform.org
/
ml
/
ethos-u
/
ethos-u-vela
/
refs/tags/3.5.0
/
ethosu
/
vela
/
graph_optimiser.py
tag
1add1eb8184deff93fb44f74267a582f496884be
tagger
Tim Hall <tim.hall@arm.com>
Tue Aug 30 10:11:44 2022 +0100
object
1bbd06bf538ff9d290e9703b6bac05b9617f2ab0
22.08 Release.
0957e3e
MLBEDSW-5102 Update removal of memory only operators
by Jonas Ohlsson
· 2 years, 10 months ago
8f1f9aa
MLBEDSW-4838 Added basic TOSA support.
by Patrik Gustavsson
· 3 years ago
d8339a7
MLBEDSW-4034: New Scheduler Size or Performance Optimisation
by Tim Hall
· 3 years, 1 month ago
c875aa6
[MLBEDSW-4530] Improve --verbose-graph output
by Fredrik Svedberg
· 3 years, 2 months ago
8244ce0
MLBEDSW-4539: MEAN axis check exception fix
by Dwight Lidman
· 3 years, 2 months ago
5fabfca
MLBEDSW-4350 Use padding instead of skirt for merged SplitSlice
by Henrik G Olsson
· 3 years, 2 months ago
dec6fbc
MLBEDSW-4501: Support MEAN single axis variation
by Dwight Lidman
· 3 years, 2 months ago
a93f026
MLBEDSW-4132 Fix off-by-one error for negative packing axis
by Henrik G Olsson
· 3 years, 3 months ago
e22f96b
MLBEDSW Vela: Fix check for format restrictions
by Patrik Gustavsson
· 3 years, 2 months ago
3645d00
MLBEDSW-4397 Fix Reshape ifm/ofm prod/cons by cpu op
by Patrik Gustavsson
· 3 years, 2 months ago
ee99bb1
MLBEDSW-4334 Non-linear format decision in graph opt.
by Patrik Gustavsson
· 3 years, 3 months ago
95b279f
MEAN implementation changed to Average Pool
by Dwight Lidman
· 3 years, 3 months ago
9bb1e2e
MLBEDSW-4163: OFM zero point outside valid range
by Dwight Lidman
· 3 years, 3 months ago
9b37918
MLBEDSW-4215: Add support for MEAN to match QuantizedMeanOrSum implementation
by Dwight Lidman
· 3 years, 3 months ago
c822d62
MLBEDSW-4223: Full support for PAD operator
by Louis Verhaard
· 3 years, 4 months ago
8ba0792
MLBEDSW-4070: Addresses errors with the LeakyReLU operator.
by erik.andersson@arm.com
· 3 years, 4 months ago
4f728c0
MLBEDSW-1499: Add MEAN operator
by Dwight Lidman
· 3 years, 6 months ago
1a92f78
MLBEDSW-4022: support PAD followed by pool operator
by Louis Verhaard
· 3 years, 5 months ago
2446e59
MLBEDSW-3774 Fix avoid cascading for spilling
by Patrik Gustavsson
· 3 years, 5 months ago
e3b1b91
MLBEDSW-3774 Remove SplitSliceRead
by Patrik Gustavsson
· 3 years, 5 months ago
138d47f
MLBEDSW-3774 Removed ConcatSliceWrite
by Patrik Gustavsson
· 3 years, 5 months ago
da2b003
MLBEDSW-3937 Fix check for NHCWB16 for FC
by Patrik Gustavsson
· 3 years, 5 months ago
73e843f
vela: Change Shape4D mutability usage
by Tim Hall
· 3 years, 5 months ago
133ba7e
MLBEDSW-3772 Removed fixup_act_reorder
by Patrik Gustavsson
· 3 years, 5 months ago
ebf4af6
MLBEDSW-3903: Bug fix PAD operator
by Louis Verhaard
· 3 years, 5 months ago
189f748
MLBEDSW-3224: Support HardSwish
by Diqing Zhong
· 3 years, 5 months ago
2c2522d
MLBEDSW-3772 Fix FC with changed inp shape
by Patrik Gustavsson
· 3 years, 5 months ago
7bada40
MLBEDSW-3772 Fix unfusing ConcatTFlite
by Patrik Gustavsson
· 3 years, 5 months ago
3a26920
MLBEDSW-3772 Reshape removal
by Patrik Gustavsson
· 3 years, 5 months ago
ece4e65
MLBEDSW-3418: Bug fixes 16-bit leaky relu
by Louis Verhaard
· 3 years, 6 months ago
e82be7c
[MLBEDSW-2787] Remove op.attrs["rescale"] in softmax.py
by Fredrik Svedberg
· 3 years, 5 months ago
224e99b
MLBEDSW-3654 Fix for split/concat ops
by Patrik Gustavsson
· 3 years, 5 months ago
c509d33
MLBEDSW-3654 Fix setting op ifm ofm for LeakyRelu
by Patrik Gustavsson
· 3 years, 6 months ago
69b8480
MLBEDSW-3495: Tensor resampling mode for transpose
by Louis Verhaard
· 3 years, 6 months ago
ae2d553
MLBEDSW-3499: Support for PAD operator
by Louis Verhaard
· 3 years, 7 months ago
016b827
MLBEDSW-1493: Optimise strided conv
by Diqing Zhong
· 3 years, 6 months ago
3d73717
MLBEDSW-3791 Fix converting axis to 4D axis
by Patrik Gustavsson
· 3 years, 6 months ago
cc6915c
MLBEDSW-3790 Fix for cpu ops has no op.ifm_shapes
by Patrik Gustavsson
· 3 years, 6 months ago
eeb8515
Revert "Revert "MLBEDSW-3645 4D class for op ifm/ofm shapes""
by patrik.gustavsson
· 3 years, 6 months ago
df0a590
Revert "MLBEDSW-3645 4D class for op ifm/ofm shapes"
by patrik.gustavsson
· 3 years, 6 months ago
bf31d64
MLBEDSW-3645 4D class for op ifm/ofm shapes
by Patrik Gustavsson
· 3 years, 6 months ago
2349d42
MLBEDSW-3654 Add/use op ifm/ofm shapes
by Patrik Gustavsson
· 3 years, 7 months ago
1689548
MLBEDSW-3694 Replace padding with enum
by Michael McGeagh
· 3 years, 6 months ago
7a6f843
MLBEDSW-2066 Improve Exception messages
by Michael McGeagh
· 3 years, 7 months ago
94457b1
Revert "Revert "MLMBED-3450: Do not convert batched fully connected to conv""
by Diqing Zhong
· 3 years, 7 months ago
f3e3ad7
MLBEDSW-2836 Change sets to tuples
by Michael McGeagh
· 3 years, 7 months ago
d9c2c42
[MLBEDSW-3690] Refactor Softmax
by Fredrik Svedberg
· 3 years, 7 months ago
c21d1c7
MLBEDSW-3583: Prevent DMA for converted FullyConnected
by Louis Verhaard
· 3 years, 7 months ago
43f8f64
Revert "MLMBED-3450: Do not convert batched fully connected to conv"
by Patrik Gustavsson
· 3 years, 7 months ago
2.0.0.rc2
907cb03
vela: Tanh and Sigmoid broken in fixup_act_reorder
by Tim Hall
· 3 years, 8 months ago
8956761
vela: Improve the scaling is equal check
by Tim Hall
· 3 years, 8 months ago
15a8e80
MLMBED-3450: Do not convert batched fully connected to conv
by Andreas Nevalainen
· 3 years, 7 months ago
e8a5a78
MLBEDSW-839: Code generation using external API
by Louis Verhaard
· 3 years, 8 months ago
2.0.0.rc1
dda21af
MLBEDSW-3463: StridedSlice fixup function causes infinite recursion
by Dwight Lidman
· 3 years, 8 months ago
e6ccd87
MLBEDSW-3019: Add profiling debug database
by Tim Hall
· 3 years, 8 months ago
73320a4
MLBEDSW-3377: fixup_stridedslice_output may silently change CPU ops
by Dwight Lidman
· 3 years, 8 months ago
c7c0b1b
MLBEDSW-3283: Bug fix: StridedSlice Op is placed on CPU
by Diqing Zhong
· 3 years, 8 months ago
9358296
vela: Improve the scaling is equal check
by Tim Hall
· 3 years, 10 months ago
c6ac194
MLBEDSW-3004: UnpackReshaped can't be serialised
by Dwight Lidman
· 3 years, 9 months ago
2e186c7
MLBEDSW-3154 Bug fix for LUT ops with IFM from SplitSliceRead
by Louis Verhaard
· 3 years, 9 months ago
aee5d75
MLBEDSW-3148: Refactor Operation
by Louis Verhaard
· 3 years, 9 months ago
3010d9b
MLBEDSW-3060 Adjust check if weights fit in sram
by Patrik Gustavsson
· 3 years, 9 months ago
8912c53
MLBEDSW-3153: Fix overflow in sigmoid LUT
by Louis Verhaard
· 3 years, 9 months ago
0f98b36
[MLBEDSW-2802] Fix 5D tensor crash
by Fredrik Svedberg
· 3 years, 9 months ago
7ad862a
MLBEDSW-3032 Fixed issue in removal of reshapes
by Patrik Gustavsson
· 3 years, 9 months ago
f03bad3
MLBEDSW-2031: LUT support tanh/sigmoid
by Louis Verhaard
· 3 years, 9 months ago
f3d737e
MLBEDSW-2811: Add rescaling to ReLus with different scaling
by Andreas Nevalainen
· 3 years, 9 months ago
6c97e9a
MLBEDSW-3018 Change to check for avoiding NHCWB16
by Patrik Gustavsson
· 3 years, 9 months ago
cb33704
MLBEDSW-1693 Convert batched FC to Conv
by Patrik Gustavsson
· 3 years, 9 months ago
c3862c2
MLBEDSW-3008: Replace Split ops with Identity
by Dwight Lidman
· 3 years, 10 months ago
8dbf8cf
MLBEDSW-2745 Support relus with differing scales
by Michael McGeagh
· 3 years, 10 months ago
fa4cb29
MLBEDSW-2994 Remove undesired reshape OPs
by Patrik Gustavsson
· 3 years, 10 months ago
11b0bdb
vela: Refactor op sets in graph_optimiser
by Michael McGeagh
· 3 years, 10 months ago
98a3499
MLBEDSW-2935: LUT fusing with preceding operator
by Louis Verhaard
· 3 years, 10 months ago
90033f3
MLBEDSW-2906: Only add padding fields to Npu ops
by Jacob Bohlin
· 3 years, 10 months ago
d7911c4
MLBEDSW-2688: LUT calculation with different in/out scale
by Louis Verhaard
· 3 years, 10 months ago
a41cd4d
Small fix for Softmax regression
by Jacob Bohlin
· 3 years, 10 months ago
58520b9
MLBEDSW-2688: use LeakyRelu for int16
by Louis Verhaard
· 3 years, 10 months ago
d47cc27
MLBEDSW-2847: Fix for TransposeConv crash and u8 output diff
by Jacob Bohlin
· 3 years, 10 months ago
87c1350
MLBEDSW-2654: Convert Resizebilinear to a number of 2x2 pools
by Charles Xu
· 3 years, 11 months ago
b9fc33c
MLBEDSW-2688: LeakyRelu rewrite to LUT or MUL/MAX
by Louis Verhaard
· 3 years, 11 months ago
67e0d8f
MLBEDSW-2663: Handle optional tensors
by Jacob Bohlin
· 3 years, 10 months ago
eebb1c2
MLBEDSW-2779 Avoid NHCWB16 in some SplitSliceRead cases
by Patrik Gustavsson
· 3 years, 10 months ago
29d568e
MLBEDSW-2779 Consider num dimensions, in check for NHCWB16
by Patrik Gustavsson
· 3 years, 10 months ago
458a208
MLBEDSW-2570 Avoid usage of NHCWB16 for some cases
by Patrik Gustavsson
· 3 years, 11 months ago
c5b549b
MLBEDSW-2637 Utilise new tensor and operator funcs
by Michael McGeagh
· 3 years, 11 months ago
36ffaf3
Skip the NOP resizebilinear op
by Charles Xu
· 3 years, 11 months ago
a0c3624
[MLBEDSW-2335] SoftMax int16
by Fredrik Svedberg
· 4 years, 1 month ago
9a03fdf
MLBEDSW-2569:Support 1x1 IFM ResizeBilinear
by Charles Xu
· 4 years ago
8d939c0
MLBEDSW-2660 Convert FC-like convs to FC
by Michael McGeagh
· 4 years ago
6a8d424
vela: Move common functionality
by Michael McGeagh
· 4 years ago
9b64ba0
MLBEDSW-2584: Support cascading of Transpose Convolution
by Jacob Bohlin
· 4 years ago
e843d33
MLBEDSW-2548: Fix for Double Buffer size estimate
by Jacob Bohlin
· 4 years ago
c30f495
Code clean-up using black and flake8
by Tim Hall
· 4 years ago
e0ef273
MLBEDSW-2379: Fix 1-element tensors that were marked as scalars
by Louis Verhaard
· 4 years, 1 month ago
cf7da10
MLBEDSW-1716: Transpose Convolution support
by Jacob Bohlin
· 4 years, 1 month ago
b2fb212
MLBEDSW-2420: Improved support for dilated convolution
by Louis Verhaard
· 4 years, 1 month ago
42fed94
MLBEDSW-2372: Failing assert for ResizeBilinear with upscale != 2x
by Dwight Lidman
· 4 years, 1 month ago
7db7896
MLBEDSW-2067: added custom exceptions
by Louis Verhaard
· 4 years, 1 month ago
7879222
Add elementwise vector scalars support
by Charles Xu
· 4 years, 1 month ago
Next »