projects / platform / core / ml / nnfw.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 540e520) | patch
Support ReduceMax kernel for cl up to 4-dimensions (#3340)
author장지섭/동작제어Lab(SR)/Engineer/삼성전자 <jiseob.jang@samsung.com>
Mon, 5 Nov 2018 07:04:39 +0000 (16:04 +0900)
committer오형석/동작제어Lab(SR)/Staff Engineer/삼성전자 <hseok82.oh@samsung.com>
Mon, 5 Nov 2018 07:04:39 +0000 (16:04 +0900)
commit4af4ed2b4622ec7c2a842b1510c362be8594ef45
tree3e239f5f4ccee192deef6d957177f7d9297e4356tree | snapshot
parent540e52029b0dd90ced52333fe6dc664c9e5f4ad1commit | diff
Support ReduceMax kernel for cl up to 4-dimensions (#3340)

* Support ReduceMax kernel for cl up to 4-dimensions

This commit supports ReduceMax kernel for cl up to 4-dimensions.

Signed-off-by: jiseob.jang <jiseob.jang@samsung.com>
* Optimize ReduceMax Kernel for cl

This commit optimizes ReduceNMax kernel for cl.
  - Change calling kernel from at once kernel to call separated kernels multiple times.

Signed-off-by: jiseob.jang <jiseob.jang@samsung.com>
libs/ARMComputeEx/arm_compute/core/CL/kernels/CLReduceMaxKernel.h diff | blob | history
libs/ARMComputeEx/arm_compute/runtime/CL/functions/CLReduceMax.h diff | blob | history
libs/ARMComputeEx/src/core/CL/cl_kernels/reduce_max.cl diff | blob | history
libs/ARMComputeEx/src/core/CL/kernels/CLReduceMaxKernel.cpp diff | blob | history
libs/ARMComputeEx/src/runtime/CL/functions/CLReduceMax.cpp diff | blob | history
runtimes/pure_arm_compute/src/compilation.cc diff | blob | history
Domain: Machine Learning / ML Framework;