diff options
author | Finn Williams <Finn.Williams@arm.com> | 2021-08-09 13:00:08 +0100 |
---|---|---|
committer | Matthew Sloyan <matthew.sloyan@arm.com> | 2021-08-10 10:30:28 +0000 |
commit | d218d9804723e78da9bbd36e6211b3310426852b (patch) | |
tree | 38d4fc63d6a3ed376a094d8b4867e8c15f7d73ad /tests/InferenceModel.hpp | |
parent | b20d1d4888c270d4d57a0bdcc011ded89a2f5b38 (diff) | |
download | armnn-d218d9804723e78da9bbd36e6211b3310426852b.tar.gz |
IVGCVSW-6289 Separate tensor shape inference and validation calls
* Pass m_shapeInferenceMethod to OptimizerOptions in ExecuteNetwork
Signed-off-by: Finn Williams <Finn.Williams@arm.com>
Change-Id: I90280fb7629092d3b66e8a3968ca9e35a0df854a
Diffstat (limited to 'tests/InferenceModel.hpp')
-rw-r--r-- | tests/InferenceModel.hpp | 3 |
1 files changed, 3 insertions, 0 deletions
diff --git a/tests/InferenceModel.hpp b/tests/InferenceModel.hpp index 31075939ce..4d2b167522 100644 --- a/tests/InferenceModel.hpp +++ b/tests/InferenceModel.hpp @@ -440,6 +440,9 @@ public: options.m_ReduceFp32ToBf16 = params.m_EnableBf16TurboMode; options.m_Debug = params.m_PrintIntermediateLayers; + options.m_shapeInferenceMethod = params.m_InferOutputShape ? + armnn::ShapeInferenceMethod::InferAndValidate : armnn::ShapeInferenceMethod::ValidateOnly; + armnn::BackendOptions gpuAcc("GpuAcc", { { "FastMathEnabled", params.m_EnableFastMath }, |