projects / platform / core / ml / nnfw.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: ae077ed) | patch
Introduce cpu quant8 convolution kernel (#4910)
author오형석/On-Device Lab(SR)/Staff Engineer/삼성전자 <hseok82.oh@samsung.com>
Mon, 1 Apr 2019 10:08:26 +0000 (19:08 +0900)
committerGitHub Enterprise <noreply-CODE@samsung.com>
Mon, 1 Apr 2019 10:08:26 +0000 (19:08 +0900)
commit2ade35e42320121d583e907d4737fe29a903f6c8
tree1cff9ffec2d42ecb967f12e753cfb92fd5711820tree | snapshot
parentae077ed17e17f5204def511792d81b9c2c603853commit | diff
Introduce cpu quant8 convolution kernel (#4910)

Introduce cpu quantized int8 convolution kernel from tflite and gemmlowp
Use kernel in neurun cpu backend

Signed-off-by: Hyeongseok Oh <hseok82.oh@samsung.com>
libs/cker/README.md diff | blob | history
libs/cker/include/cker/FixedPoint.h [new file with mode: 0644] blob
libs/cker/include/cker/Utils.h diff | blob | history
libs/cker/include/cker/operation/Conv.h diff | blob | history
runtimes/neurun/backend/cpu/kernel/ConvolutionLayer.cc diff | blob | history
runtimes/neurun/backend/cpu/kernel/OperationUtils.cc diff | blob | history
runtimes/neurun/backend/cpu/kernel/OperationUtils.h diff | blob | history
Domain: Machine Learning / ML Framework;