projects / platform / upstream / armnn.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: d0dc770) | patch
IVGCVSW-3609 Fix decoding and encoding of INT32 tensors
authorAron Virginas-Tar <Aron.Virginas-Tar@arm.com>
Fri, 2 Aug 2019 17:54:28 +0000 (18:54 +0100)
committerAron Virginas-Tar <Aron.Virginas-Tar@arm.com>
Fri, 2 Aug 2019 18:08:11 +0000 (19:08 +0100)
commit198ee400733aa633ddeb867e42cc241a684e9787
tree4e9e11aa8a270f7ae4a816cd8711a4ea3a9d130ftree | snapshot
parentd0dc770aeb9423081a1ef0245f7449dcf3fb2ee2commit | diff
IVGCVSW-3609 Fix decoding and encoding of INT32 tensors

* Added Int32Decoder and Int32Encoder to decode INT32 tensors
* Changed MakeDecoder to return ScaledInt32Decoder only
  if the scale is different from 0, i.e. for quantized bias
  tensors

Signed-off-by: Aron Virginas-Tar <Aron.Virginas-Tar@arm.com>
Change-Id: I278061d445d1c549c7ace11f51aa172ce7c691ae
src/backends/reference/workloads/BaseIterator.hpp diff | blob | history
src/backends/reference/workloads/Decoders.hpp diff | blob | history
src/backends/reference/workloads/Encoders.hpp diff | blob | history
Domain: Machine Learning / ML Framework;