Age | Commit message (Collapse) | Author |
|
Resolves: COMPMID-4813
Signed-off-by: Sheri Zhang <sheri.zhang@arm.com>
Change-Id: I773e5b1be2951431e072012fe393993411f68b66
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/6269
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Gunes Bayir <gunes.bayir@arm.com>
Reviewed-by: Giorgio Arena <giorgio.arena@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
|
|
The issue is caused by the number of iterations passed to
LOOP_UNROLLING. When we use the manual LOOP_UNROLLING, the number of
iterations must be less than or equal to 128.
To overcome this problem, we create a utility function to check if
any of the critical iterations (kernel dimensions) are beyond that
limit. If so, the utility function, disable the manual loop unrolling.
Resolves COMPMID-4609
Change-Id: I7221c967609e462a5abd1cbb74e2a120f344fcb3
Signed-off-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/5913
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
|
|
- Remove dedicated kernels for NCHW. Now we only use NHWC with permute
- Remove specialized kernels for 3x3 NHWC
- Simplify CLDepthwiseConvolutionLayer.cpp to call just the native
implementation for both floating-point and quantized data types
- Develop two parametric opencl kernels for depthwise convolution layer NHWC
(floating-point and quantized)
- Add support to export the weights to cl_image
- Extend test for depthwise convolution on opencl
Resolves COMPMID-4417
Change-Id: Ibe533f79c2860f9cac8e921895d5a8f947753a5c
Signed-off-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/5893
Reviewed-by: Giorgio Arena <giorgio.arena@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
|
|
* Implement in-place graph node mutator for 1x1 depthwise convolution
* Add in-place to validation fixture except for
DepthwiseConvolutionLayerNativeValidationFixture as it would be a
duplicate test otherwise (DepthwiseConvolutionLayerNative test tests
the underlying kernel)
Resolves: COMPMID-4432
Change-Id: Id7f10f5ebdce7d49f550c0b62dbaaab7f5b59d29
Signed-off-by: SiCongLi <sicong.li@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/5874
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Michele Di Giorgio <michele.digiorgio@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
|
|
This reverts commit 561c176598cd14245e2e7918fdf136d1c888d1da.
Reason for revert: <validation>
Change-Id: I6f2d61c27520439bb538e9265736532104b24cf8
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/5127
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
|
|
- Remove dedicated kernels for NCHW. Now we only use NHWC with permute
- Remove specialized kernels for 3x3 NHWC
- Simplify CLDepthwiseConvolutionLayer.cpp to call just the native
implementation for both floating-point and quantized data types
- Develop two parametric opencl kernels for depthwise convolution layer NHWC
(floating-point and quantized)
- Add support to export the weights to cl_image
- Extend test for depthwise convolution on opencl
Resolves COMPMID-4417
Change-Id: I253dd5d959a70783c82e62b1771a5e9f91621cb0
Signed-off-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/5806
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Giorgio Arena <giorgio.arena@arm.com>
|
|
* Replace assembly kernels for depthwise convolution with more optimized
ones.
* Add int8 assembly kernels.
* Fix implicit padding on optimized kernels
Resolves: COMPMID-3867, COMPMID-4361
Change-Id: I0b0867e05f61be4f368f62190d55e14d0ab3ebf2
Signed-off-by: Michele Di Giorgio <michele.digiorgio@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/5622
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
|
|
Relates to COMPMID-4385
Change-Id: Ibc1d67f766b7c1a399dbeacf26a4b9d9f7323785
Signed-off-by: Michele Di Giorgio <michele.digiorgio@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/5549
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
|
|
Resolves: COMPMID-4185
Change-Id: Ib5f22356356a022d567bb18d44ea272b62d10ebf
Signed-off-by: Michalis Spyrou <michalis.spyrou@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/5424
Reviewed-by: Michele Di Giorgio <michele.digiorgio@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
|
|
Add artificial implicit padding testing for the following fixtures:
- Scale
- FullyConnected
- Pooling
- DepthwiseConvolution
- DirectConvolution
- Winograd
- FFT
- GEMM/GEMMLowp
Create utility function that loops through a list of tensor and adds random padding based on the global seed (only for NHWC layer layout).
Remove GEMMLowpAssemblyFixture since it wasn't used
Remove some AssetsLibrary headers since they weren't used
Resolve COMPMID-4161
Change-Id: Ib6f4f7f113ae69b993d7b2a9e04abbf3de8c99fe
Signed-off-by: Giorgio Arena <giorgio.arena@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/5327
Reviewed-by: Michele Di Giorgio <michele.digiorgio@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
|
|
Change data layouts after the configure in validation tests for:
- Scale
- Pooling
- FullyConnected
- DepthwiseConvolution
- DirectConvolution
- FFTConvolution
- WinogradConvolution
- GEMMConvolution (Indirect GEMM included)
Extending fixtures
Fixes for new mixed data layout tests
Resolves: COMPMID-4162
Change-Id: I2f2eb2075f7e24ab3872249d88cadb57b82c5dde
Signed-off-by: Manuel Bottini <manuel.bottini@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/5326
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
|
|
This is needed in order to validate OpenCL kernel run-time compilation, without necessarily running or validating the kernels' execution
- Add a run-time option for our validation suite to only configure one target function, without allocating, running or validating
- Avoid to map/unmap tensors in CLAccessor if no allocation/validation is required
- Create a new Fixture macro that accepts fixtures split into configure/allocate_and_run/reference, and do the last two only if required
- Adjust fixture and validation files for the first trial function(s) (DepthwiseConvolutionLayer)
Signed-off-by: Giorgio Arena <giorgio.arena@arm.com>
Change-Id: I56fa1ce5ef4ac0c86bcabda686cc277ef5ec69c8
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/5048
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Manuel Bottini <manuel.bottini@arm.com>
Reviewed-by: Sang-Hoon Park <sang-hoon.park@arm.com>
|
|
Resolves: COMPMID-4056
Signed-off-by: Giorgio Arena <giorgio.arena@arm.com>
Change-Id: I6623eb9c0e66e52af4e0e9fb386031f4a09125b7
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/4722
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Michele Di Giorgio <michele.digiorgio@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
|
|
- Change name of uniform_real_distribution_fp16 to uniform_real_distribution_16bit, and make it also accept bfloat16 data type
Resolves: COMPMID-4057
Signed-off-by: Giorgio Arena <giorgio.arena@arm.com>
Change-Id: Id2f1a84b9c9f09cb260a0785add4fc5954d5853a
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/4768
Reviewed-by: Michele Di Giorgio <michele.digiorgio@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
|
|
- Change uniform_real_distribution_fp16's arguments to float to avoid maybe-uninitialized warnings in half data type's copy constructor and converter
- Add increment method to Dimensions in order to avoid possible overflows when incrementing single dimensions
- Remove internal_only flag documentation
Resolves: COMPMID-4061
Change-Id: I7c50487193bd7b47b03d52fc595840c7df4e7104
Signed-off-by: Giorgio Arena <giorgio.arena@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/4760
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
|
|
Resolves: COMPMID-4050
Signed-off-by: Giorgio Arena <giorgio.arena@arm.com>
Change-Id: I182548bf4b944c499a7134ac005b137877e61baf
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/4700
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Michele Di Giorgio <michele.digiorgio@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
|
|
preferred presentation
Change-Id: Ib7dcfcbb24b408999dfae366b9da396485aacf78
Signed-off-by: Michele Di Giorgio <michele.digiorgio@arm.com>
Reviewed-on: https://review.mlplatform.org/c/ml/ComputeLibrary/+/3525
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
Tested-by: Georgios Pinitas <georgios.pinitas@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
|
|
Change-Id: I1f292f98bc3a213ba5b26ac88aa78160c809cb87
Signed-off-by: Michele Di Giorgio <michele.digiorgio@arm.com>
Reviewed-on: https://review.mlplatform.org/c/2540
Reviewed-by: Michalis Spyrou <michalis.spyrou@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
|
|
COMPMID-2470 Implement a new and generic depthwise convolution for NEON QASYMM8 NHWC
COMPMID-2477 Enable FP16 data type for the new generic convolution on NEON for NHWC
COMPMID-2625 Remove old implementation files for the generic NEDepthwiseConvolution
Change-Id: I8f6deda4fc69dd7e472fba3228b1ed5dad172f3e
Signed-off-by: Giorgio Arena <giorgio.arena@arm.com>
Reviewed-on: https://review.mlplatform.org/c/2094
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
|
|
Reference
This patch modifies the reference implementation and the fixtures of
depthwise convolution layer to support QSYMM8_PER_CHANNEL quantization.
Change-Id: I28adb5c110308b1024a213bec2d35a89180a46dc
Signed-off-by: Michele Di Giorgio <michele.digiorgio@arm.com>
Reviewed-on: https://review.mlplatform.org/c/2063
Reviewed-by: Giorgio Arena <giorgio.arena@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Giuseppe Rossini <giuseppe.rossini@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
|
|
(Fp32/FP16-NHWC)
Part 1
Change-Id: I5e1d27a7006199e9229e455a1df9bfc2ed4e8341
Signed-off-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
Reviewed-on: https://review.mlplatform.org/c/1898
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Giorgio Arena <giorgio.arena@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
|
|
Change-Id: I45cacf75b08bb9d867343037507e56f200ad6ac0
Signed-off-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
Reviewed-on: https://review.mlplatform.org/c/1637
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Giorgio Arena <giorgio.arena@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
|
|
Change-Id: I2b883785c0500d4bdb6ee4700382ee058be2cd36
Signed-off-by: Giorgio Arena <giorgio.arena@arm.com>
Reviewed-on: https://review.mlplatform.org/c/1538
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
|
|
Change-Id: I18c39a7708a68861764b548c8d2bea3100be3612
Signed-off-by: Georgios Pinitas <georgios.pinitas@arm.com>
Reviewed-on: https://review.mlplatform.org/c/1511
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Michele Di Giorgio <michele.digiorgio@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
|
|
There was a problem in depthwise_conv where when needing to run
permute it was calling auto_init to initialise the temporary tensor
with input's qinfo instead of output's qinfo. This patch fixes
this and adds some tests to exercise the cases where we have different
qinfos
Change-Id: I396fbb4d73f6aa5efe03dc2dd727b3e13154d6f5
Signed-off-by: Pablo Tello <pablo.tello@arm.com>
Reviewed-on: https://review.mlplatform.org/c/1270
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Giuseppe Rossini <giuseppe.rossini@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
|
|
for F32
Change-Id: I39dd23696b6d8573e172a59b9e327b6a69886f08
Signed-off-by: Manuel Bottini <manuel.bottini@arm.com>
Reviewed-on: https://review.mlplatform.org/c/973
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Usama Arif <usama.arif@arm.com>
Reviewed-by: Giuseppe Rossini <giuseppe.rossini@arm.com>
|
|
Change-Id: I3106aa34bd168985a56791613d95072756be6e9b
Signed-off-by: Usama Arif <usama.arif@arm.com>
Reviewed-on: https://review.mlplatform.org/c/958
Comments-Addressed: Arm Jenkins <bsgcomp@arm.com>
Reviewed-by: Pablo Marquez <pablo.tello@arm.com>
Tested-by: Arm Jenkins <bsgcomp@arm.com>
|
|
Removed Fixed point position arguments from test sources
Change-Id: I8343724723b71611fd501ed34de0866d3fb60e7e
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/136382
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Michele DiGiorgio <michele.digiorgio@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Change-Id: I00b95f560548da76718298b642c8166f92421097
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/129520
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Michele DiGiorgio <michele.digiorgio@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Change-Id: I03f32c62350e5ea43e77bb15fc5a832d83719e3b
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/126657
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Michele DiGiorgio <michele.digiorgio@arm.com>
Reviewed-by: Georgios Pinitas <georgios.pinitas@arm.com>
|
|
Change-Id: I89de432f3fbcba7abf9e1d4f8396a4334b4fa2c2
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/118324
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Gian Marco Iodice <gianmarco.iodice@arm.com>
|
|
(optimized case).
Change-Id: Icdfd6c02ed526daf4f59a4b76c7bbc1bc48fde74
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/125938
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Change-Id: I1a9ed6c3420ddf8978aeaad48d9915333b006b49
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/116374
Tested-by: Jenkins <bsgcomp@arm.com>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|
|
Change-Id: I051b7e56b60bf1a55cdf014539ef71346d3aee26
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/114737
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Tested-by: Jenkins <bsgcomp@arm.com>
|
|
Removed biases shape from the dataset which is redundant as it can be computed from the weights shape.
Change-Id: I078b6eeb9e0a42f697e7b311deb86d2a1795872b
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/112904
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>
|
|
Change-Id: I147644349547c4e3804a80b564a9ad95131ad2d0
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/111560
Reviewed-by: Michalis Spyrou <michalis.spyrou@arm.com>
Tested-by: BSG Visual Compute Jenkins server to access repositories on http://mpd-gerrit.cambridge.arm.com <bsgcomp@arm.com>
|
|
Change-Id: I4b1f3f0da9ff5342c7de7083736fe91871d14e5b
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/110351
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>
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
|