Age | Commit message (Collapse) | Author |
|
Reworked the interface of GemmLowp in order to make easy the integration
in Android NN
- Added support for different output stage
- Added validation for both matrix multiplication and output stage
- Added bounded relu support in the output stage
- Added in32_t bias support
- Added optimized path for vector by matrix case
This rework is required for:
- Convolution quantized
- Fully connected quantized
Change-Id: I512283d406099cf8c614dd89d0a97ed411143afc
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/110625
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com>
|
|
Is there any reason why we might have OpenCL enabled but wouldn't want to make the opencl_timer instrument available ?
Change-Id: I946d643028fc394a607478415e4f84c4e8fa5bc4
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/110466
Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com>
Reviewed-by: Kevin Petit <kevin.petit@arm.com>
Reviewed-by: Ioan-Cristian Szabo <ioan-cristian.szabo@arm.com>
Reviewed-by: Michalis Spyrou <michalis.spyrou@arm.com>
Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
Reviewed-by: Giorgio Arena <giorgio.arena@arm.com>
|
|
Change-Id: Icdaf7c2666a7fbb9b3c32ba2d70b667a59a6eeb9
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/110468
Reviewed-by: Pablo Tello <pablo.tello@arm.com>
Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com>
|
|
Integrate generic gemmlowp assembly version for u8.
Change-Id: I17ed4494c25a132b2bac581febe1544e49b4f352
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/110114
Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com>
Reviewed-by: Pablo Tello <pablo.tello@arm.com>
|
|
Change-Id: I6f51ffe6c324d9da500716b52c97c344f2a2a164
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/110486
Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
|
|
Will help with scripting to split the tests to run across several boards
Change-Id: I32806c3cd03cd1b4af1865cd4fdb0422d609eff0
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/110535
Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com>
Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
|
|
Change-Id: Iccd686be18381e96bcf09b14c7017c6dda0f38d8
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/109824
Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com>
Reviewed-by: Pablo Tello <pablo.tello@arm.com>
|
|
Change-Id: If1dcca724e5e5f5ab363ffc16b0ef8c943e0b657
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/110105
Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com>
Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Change-Id: I4853a4d83a13b162d5759434a4e800c8f9e05cce
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/110108
Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
|
|
- Added --no-allow-shlib-undefined to library builds to make sure we
don't leave any unresolved symbol
- Added --allow-shlib-undefined to the graph examples build lines so that
it doesn't matter whether or not OpenCL is supported by the Graph
librarylibrary
- Don't link tests statically anymore on Android (This was probably
required in the past because we were duplicating symbols in the
core and runtime libraries, but it doesn't seem to be an issue
anymore)
- Updated Doxygen accordingly
Change-Id: I9905f388e0838bc8e3369de52838bf980b992ead
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95172
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
|
|
Change-Id: I8a470cc1351593ad8eeaf4ec92e04865e83d4f3c
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/96147
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Change-Id: I8ae6b5111b594f224ad9b600cdf078459a8cac36
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/93281
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Change-Id: If09afa444c6b3e91117d1b1a529faa0778457cd3
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/96099
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Change-Id: Id9c50abdbd28e7e95fd460240cec51af1d6f1036
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95580
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
|
|
- Added min and max arguments for QuantizeDownInt32ToUint8Scale in order
to apply bounded relu
- Added support for int32_t biases
- Extended tests
Change-Id: I015dae17faa7284766b5435ca33bcf593c1b2b69
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/96512
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
|
|
Adds validation support to:
- CLDirectConvolution
- CLNormalizationLayer
- CLSoftmaxLayer
Change-Id: I9bd1e925e6db057c799169405f82ed21d20b87ee
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95939
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
|
|
Change-Id: I9b49a505201bfd6e0c4a40ad64ce302cbc3f2cf3
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/109907
Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
|
|
CLDepthwiseConvolution3x3 kernel (#28)
Change-Id: I51bebe74e3814c1245812ad575fe7854d460674f
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/109864
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com>
|
|
Biases were incorrectly passed as uchar8 in asymmetric quantized
calculations while they should be in int32.
Change-Id: I461f5e4ef6eb44374c1094378a1bfe9ade5e247d
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/96244
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Change-Id: Ibf15544e8624977815ee355071a038c07e1f01c2
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/94874
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
|
|
Change-Id: I21a4191b032df17611e01d20d89cc21e8eaf68fc
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/109605
Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Adds a is_scaled parameter in the NormalizationLayerInfo that flags if
the alpha parameter should be scaled by the normalization size of not.
Unscaled parameter is used by [Krichevksy 2012] which is used in
AndroidNN and TensorFlow LRN layer.
Change-Id: Iad2aa5e688cf4dcd6cc77a6e28c0663764f34ccb
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/96102
Reviewed-by: Diego Lopez Recas <diego.lopezrecas@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
|
|
Added support for S8 input in NEGEMMLowp Matrix Multiply Kernel.
Added a new function to run assembly kernels such that A*B=C (no offsets involved)
Added new tests for the assembly gemmlowp kernels (no offsets)
Integrated the assembly kernel for the A57
Change-Id: Ib3e39c1f3f7f1baa0d39be69485f61cd18e3c9b3
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95864
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Which means if someone specifies --log-file or changes --log-format then it will still output to the console.
Setting back to the original behaviour: if --log-file or --log-format is specified then there is no dual output.
The user need to explicitely ask for it by using --pretty-printer
Change-Id: I1a7cd37fd78286a566de3ba1fa4bcb0522f24077
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/96095
Reviewed-by: Pablo Tello <pablo.tello@arm.com>
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
|
|
Changes in CL and reference in terms of border handling.
Change-Id: I5bed95b1f4c308629d7113455dc8a55d74500bcd
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95742
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
|
|
Change-Id: Ic98f64ffe30739437a1fe31ef98d83ee900741e3
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95512
Reviewed-by: Michalis Spyrou <michalis.spyrou@arm.com>
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Change-Id: Ic6ac4cd6df6970593a5e2e6310b6d61951c88898
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/93887
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Change-Id: I257a09860dd82e7bb7a767edf96dcaf31b512855
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95865
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
|
|
(NEON and CL) for Histogram
Change-Id: Iccf6b4483cb8394dab2f861a737583126f9bed81
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/91601
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
|
|
This messes up badly the opencl_timer if we measure the first run and are using a CLTuner at the same time.
Basically all the tuner combinations appear in the JSON file and this makes it really hard to post process
Change-Id: I59377eb752ca8ddc95b884ab04b1ef3f65389931
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95860
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
|
|
Change-Id: Iead7497cc03e1e7bde440d2965a7bf54cbfa88bf
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95579
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Joel Liang <joel.liang@arm.com>
Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
|
|
Change-Id: I5f9658fee9175140d602e1ba516a67d4c83c6d87
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95671
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Michalis Spyrou <michalis.spyrou@arm.com>
|
|
The new interface makes NEGEMMLowp able to work with ASYMM8 data types.
Implemented 2 new functions:
- NEGEMMLowpMatrixMultiplyCore
- NEGEMMLowpOutputStage
These functions should make the integration in android NN doable
For more information about GEMMLowp:
https://github.com/google/gemmlowp/blob/master/doc/low-precision.md
Change-Id: Ie2c775f45234f68ca53dba644b3a912b997fd890
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95504
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Pablo Tello <pablo.tello@arm.com>
|
|
Change-Id: I380a11f41ca2158de1dd0a6339ed9c884feb8f69
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95385
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Michalis Spyrou <michalis.spyrou@arm.com>
|
|
compiler)
Change-Id: Ibb779dd3a8d10786da6d8f70590e654e14654d7b
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95530
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Pablo Tello <pablo.tello@arm.com>
|
|
Change-Id: I62b843b544fe9048837fd64c22e970fc6a0aaf23
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/94881
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
|
|
Change-Id: Iefb7a1025507ae1f01ae3cd5e9abf7c2eb74f306
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/93863
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
|
|
Change-Id: Iad11ce70a8a0878a48e445a092035c49c926cece
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/94855
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Last commit:
commit b25c5f68042b0c81bf611d59a1bb8535e1c42497
Author: Xinghang Zhou <xinghang.zhou@arm.com>
Date: Wed Oct 25 18:48:10 2017 +0800
Synced validation's tolerances of GCSoftmax from cl side
Change-Id: Ibe72054205c1c8721845d679a31af7ed0a7c5cf6
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/93283
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
|
|
Quotes were missing around the JSON strings
Change-Id: I3257721f76215a5c900428c453a843090d12a596
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95314
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
|
|
Change-Id: Ia59c5ac7b6ab7a53fb73103539984e310b918fca
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/94616
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
|
|
Change-Id: I90e5e28d3cf3b484f01fe5aaa394ee4d159898df
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/94043
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
|
|
variable test_filter.
This is particularly useful when building the tests for running them on Fast models because it reduces considerably
the execution time on the MODELS mainly due to the fact that there are fewer test suites in the executable.
Usage example:
PATH=$PATH:/work/local_tools/gcc-linaro-6.2.1-2016.11-x86_64_aarch64-linux-gnu/bin/ scons arch=arm64-v8a neon=1 opencl=0 openvx=0 opencv=0 Werror=0 validation_tests=1 benchmark_tests=0 debug=0 Werror=0 asserts=1 examples=0 cppthreads=0 openmp=0 standalone=1 test_filter=GEMMLowp.* -j9
Will build only the NEON GEMMLowp tests as a standalone binary.
Change-Id: I5cfb242d32276408e42da68c17455ea375902156
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95052
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
|
|
the valid ones
Change-Id: I4f5acc9af09587f0ec46482f6aea737fa42f35fc
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95186
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
|
|
Change-Id: I1037054615593205f07e25fb9b16fecd13407c2c
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95142
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
|
|
Change-Id: Ib92f62e8982a7f07799083ea41086e652437ab7e
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95100
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Pablo Tello <pablo.tello@arm.com>
|
|
Everything used to be stored as double which led to some numbers appearing in scientific notation and some counters values getting corrupted.
Now measurements can be stored as either floating point or integer values.
Added support for raw_data in order to output more detailed information to the JSON files (Will make use of that in the OpenCL timer instrument)
Change-Id: Ie83776b347a764c8bf45b47d7d9d7bec02b04257
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95035
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
|
|
Change-Id: I33439a31e1bcb995410ac023e5eca92360646564
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/94927
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Change-Id: I6504fb1e2d3d2b454a043d7157112891cef272a3
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95045
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Change-Id: I7a4126f96aa7ef7ed768ebe5b4e2b1f84228f8e6
Reviewed-on: http://mpd-gerrit.cambridge.arm.com/95060
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Tested-by: Kaizen <jeremy.johnson+kaizengerrit@arm.com>
|