diff options
author | Gian Marco Iodice <gianmarco.iodice@arm.com> | 2018-05-14 14:21:39 +0100 |
---|---|---|
committer | Anthony Barbier <anthony.barbier@arm.com> | 2018-11-02 16:52:35 +0000 |
commit | a8aef2916379402e241d9f2c5e0faf3f99c860f7 (patch) | |
tree | accf1f74bb836766260dbdb90aad7b6048c675d2 /examples/graph_inception_v4.cpp | |
parent | cb0010b02281245c66d5c996fa9ef8b22f036a2d (diff) | |
download | ComputeLibrary-a8aef2916379402e241d9f2c5e0faf3f99c860f7.tar.gz |
COMPMID-792 - Re-enabled Winograd on NEON in all graph examples.
Since now the input transform can be multi-threaded, I re-ebaled Winograd in all graph examples
Change-Id: I39ef78243bb47fdae135e18dcae2102af0675b3b
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/131048
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Tested-by: Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'examples/graph_inception_v4.cpp')
-rw-r--r-- | examples/graph_inception_v4.cpp | 8 |
1 files changed, 3 insertions, 5 deletions
diff --git a/examples/graph_inception_v4.cpp b/examples/graph_inception_v4.cpp index 827370ec5e..ed95baa99e 100644 --- a/examples/graph_inception_v4.cpp +++ b/examples/graph_inception_v4.cpp @@ -54,10 +54,9 @@ public: std::unique_ptr<IPreprocessor> preprocessor = arm_compute::support::cpp14::make_unique<TFPreproccessor>(); // Set target. 0 (NEON), 1 (OpenCL). By default it is NEON - const int target = argc > 1 ? std::strtol(argv[1], nullptr, 10) : 0; - Target target_hint = set_target_hint(target); - ConvolutionMethod convolution_hint = target_hint == Target::NEON ? ConvolutionMethod::GEMM : ConvolutionMethod::DEFAULT; - FastMathHint fast_math_hint = FastMathHint::DISABLED; + const int target = argc > 1 ? std::strtol(argv[1], nullptr, 10) : 0; + Target target_hint = set_target_hint(target); + FastMathHint fast_math_hint = FastMathHint::DISABLED; // Parse arguments if(argc < 2) @@ -114,7 +113,6 @@ public: get_weights_accessor(data_path, "/cnn_data/inceptionv4_model/Conv2d_1a_3x3_BatchNorm_beta.npy"), 0.001f) << ActivationLayer(ActivationLayerInfo(ActivationLayerInfo::ActivationFunction::RELU)) - << convolution_hint // Conv2d_2a_3x3 << ConvolutionLayer(3U, 3U, 32U, get_weights_accessor(data_path, "/cnn_data/inceptionv4_model/Conv2d_2a_3x3_weights.npy"), |