| // |
| // Copyright © 2020-2021,2023 Arm Ltd and Contributors. All rights reserved. |
| // SPDX-License-Identifier: MIT |
| // |
| |
| #pragma once |
| |
| #include <CommonTestUtils.hpp> |
| |
| #include <armnn/INetwork.hpp> |
| #include <armnn/TypesUtils.hpp> |
| |
| #include <ResolveType.hpp> |
| |
| #include <doctest/doctest.h> |
| |
| namespace |
| { |
| |
| armnn::INetworkPtr CreateRankNetwork(const armnn::TensorInfo& inputTensorInfo, |
| const armnn::TensorInfo& outputTensorInfo) |
| { |
| armnn::INetworkPtr network(armnn::INetwork::Create()); |
| |
| armnn::IConnectableLayer* inputLayer = network->AddInputLayer(0, "Input"); |
| armnn::IConnectableLayer* rankLayer = network->AddRankLayer("Rank"); |
| armnn::IConnectableLayer* outputLayer = network->AddOutputLayer(0, "Output"); |
| |
| Connect(inputLayer, rankLayer, inputTensorInfo, 0, 0); |
| Connect(rankLayer, outputLayer, outputTensorInfo, 0, 0); |
| |
| return network; |
| } |
| |
| template<armnn::DataType ArmnnType, typename T = armnn::ResolveType<ArmnnType>> |
| void RankEndToEnd(const std::vector<armnn::BackendId>& backends) |
| { |
| using namespace armnn; |
| |
| std::vector<float> floatInputData{ |
| 1, 2, 3, 4, 5, |
| 11, 12, 13, 14, 15, |
| 21, 22, 23, 24, 25 |
| }; |
| std::vector<T> inputData = armnnUtils::QuantizedVector<T>(floatInputData); |
| |
| std::vector<int32_t> expectedOutputData{ 4 }; |
| |
| TensorInfo inputInfo ({ 1, 1, 5, 3 }, ArmnnType, 0.0f, 0, true); |
| TensorShape outputShape (Dimensionality::Scalar); |
| TensorInfo outputInfo(outputShape, DataType::Signed32); |
| |
| armnn::INetworkPtr network = CreateRankNetwork(inputInfo, outputInfo); |
| |
| CHECK(network); |
| |
| std::map<int, std::vector<T>> inputTensorData = {{ 0, inputData }}; |
| std::map<int, std::vector<int32_t>> expectedOutputTensorData = {{ 0, expectedOutputData }}; |
| |
| EndToEndLayerTestImpl<ArmnnType, DataType::Signed32>(std::move(network), |
| inputTensorData, |
| expectedOutputTensorData, |
| backends); |
| } |
| |
| } // anonymous namespace |