diff options
| author | Sadik Armagan <sadik.armagan@arm.com> | 2020-09-22 14:35:19 +0100 |
|---|---|---|
| committer | Sadik Armagan <sadik.armagan@arm.com> | 2020-09-22 16:21:40 +0000 |
| commit | 283a8b4aeaebf27c7f14e0c9c4cbfaf06a577cf5 (patch) | |
| tree | 15867a066249115b309a0969ff73cf6de75957a4 /tests | |
| parent | a2364ed649290f74e0608d0208c4746664b0abbc (diff) | |
| download | armnn-283a8b4aeaebf27c7f14e0c9c4cbfaf06a577cf5.tar.gz | |
IVGCVSW-5318 'Create a Neon/CL Workload Unit Test fast_math option enabled'
* Unit test implemented to make sure it returns WINOGRAD
* Updated the enable-fast-math option in ExecuteNetwork to be consistent
Signed-off-by: Sadik Armagan <sadik.armagan@arm.com>
Change-Id: Id64f114ae47966def69a9eef0770a4251ee56a41
Diffstat (limited to 'tests')
| -rw-r--r-- | tests/ExecuteNetwork/ExecuteNetwork.cpp | 4 |
1 files changed, 2 insertions, 2 deletions
diff --git a/tests/ExecuteNetwork/ExecuteNetwork.cpp b/tests/ExecuteNetwork/ExecuteNetwork.cpp index 5924348763..c15e33fc13 100644 --- a/tests/ExecuteNetwork/ExecuteNetwork.cpp +++ b/tests/ExecuteNetwork/ExecuteNetwork.cpp @@ -138,7 +138,7 @@ int main(int argc, const char* argv[]) ("infer-output-shape", po::bool_switch()->default_value(false), "Infers output tensor shape from input tensor shape and validate where applicable (where supported by " "parser)") - ("enable_fast_math", po::bool_switch()->default_value(false), + ("enable-fast-math", po::bool_switch()->default_value(false), "Enable fast_math computation of Convolution2D operator where applicable (where supported by backend)"); } catch (const std::exception& e) @@ -189,7 +189,7 @@ int main(int argc, const char* argv[]) bool parseUnsupported = vm["parse-unsupported"].as<bool>(); bool timelineEnabled = vm["timeline-profiling"].as<bool>(); bool inferOutputShape = vm["infer-output-shape"].as<bool>(); - bool enableFastMath = vm["enable_fast_math"].as<bool>(); + bool enableFastMath = vm["enable-fast-math"].as<bool>(); if (enableBf16TurboMode && enableFp16TurboMode) { |