projects / platform / upstream / armnn.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 5191033) | patch
Update the CL pin to the latest master
authorFerran Balaguer <ferran.balaguer@arm.com>
Mon, 24 Jun 2019 11:43:38 +0000 (12:43 +0100)
committerFerran Balaguer Arm <ferran.balaguer@arm.com>
Wed, 26 Jun 2019 15:22:31 +0000 (15:22 +0000)
commitb2b5a26e7b7250532df189cd7cf7cf1d4b528a28
tree6cbb1d44f074455a58ab00c2f6edd5ab5c3e8ebetree | snapshot
parent5191033c76330509666e0993857b7286e2a325fccommit | diff
Update the CL pin to the latest master

* Update SoftMaxUint8Workload for CL and NEON to deal with
  quantization parameters as vectors.
* Change Sigmoid Activation function QAsymm8 tests to use
  scale=1.f/256.f and offset=0 as quatization output parameters.

!android-nn-driver:1417

Signed-off-by: Ferran Balaguer <ferran.balaguer@arm.com>
Change-Id: Ief91f10193fbbbad0c0124ece41f0bf4e0dcd992
scripts/get_compute_library.sh diff | blob | history
src/backends/backendsCommon/test/ActivationTestImpl.hpp diff | blob | history
src/backends/cl/workloads/ClSoftmaxUint8Workload.cpp diff | blob | history
src/backends/neon/workloads/NeonSoftmaxUint8Workload.cpp diff | blob | history
Domain: Machine Learning / ML Framework;