30 auto convertFunc = [](uint8_t* dst,
const uint8_t* src,
size_t size)
32 auto input =
reinterpret_cast<const float*
>(src);
33 auto output =
reinterpret_cast<Half*
>(dst);
34 size_t numElements = size/2;
38 for (
const auto& pair : m_TensorHandlePairs)
virtual void Execute() const override
const QueueDescriptor m_Data
#define ARMNN_SCOPED_PROFILING_EVENT_NEON(name)
void ValidateInputsOutputs(const std::string &descName, unsigned int numExpectedIn, unsigned int numExpectedOut) const
Copyright (c) 2020 ARM Limited.
static void ConvertFloat32To16(const float *srcFloat32Buffer, size_t numElements, void *dstFloat16Buffer)
Converts a buffer of FP32 values to FP16, and stores in the given dstFloat16Buffer.
NeonConvertFp32ToFp16Workload(const ConvertFp32ToFp16QueueDescriptor &descriptor, const WorkloadInfo &info)
Contains information about inputs and outputs to a layer.
void CopyTensorContentsGeneric(const ITensorHandle *srcTensor, ITensorHandle *dstTensor, CopyFunc copy)
void GatherTensorHandlePairs(const DescriptorType &descriptor, std::vector< std::pair< SrcTensorHandleType *, DstTensorHandleType *>> &tensorHandlePairs)