2 // Copyright © 2017 Arm Ltd. All rights reserved.
3 // See LICENSE file in the project root for full license information.
6 #include "NeonSoftmaxFloat32Workload.hpp"
11 NeonSoftmaxFloat32Workload::NeonSoftmaxFloat32Workload(const SoftmaxQueueDescriptor& descriptor,
12 const WorkloadInfo& info, std::shared_ptr<arm_compute::MemoryManagerOnDemand>& memoryManager)
13 : FloatWorkload<SoftmaxQueueDescriptor>(descriptor, info)
14 , m_SoftmaxLayer(memoryManager)
16 m_Data.ValidateInputsOutputs("NeonSoftmaxFloat32Workload", 1, 1);
18 // The ArmCompute softmax layer uses 2D input/output tensors, so flatten the first three dimensions.
19 arm_compute::ITensor& input = boost::polymorphic_downcast<INeonTensorHandle*>(m_Data.m_Inputs[0])->GetTensor();
20 arm_compute::ITensor& output = boost::polymorphic_downcast<INeonTensorHandle*>(m_Data.m_Outputs[0])->GetTensor();
22 m_SoftmaxLayer.configure(&input, &output, m_Data.m_Parameters.m_Beta);
25 void NeonSoftmaxFloat32Workload::Execute() const
27 ARMNN_SCOPED_PROFILING_EVENT_NEON("NeonSoftmaxFloat32Workload_Execute");