blob: 2bcc4e19176c88fc2f166c401baa30c4345433aa (
plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
|
//
// Copyright © 2017 Arm Ltd. All rights reserved.
// SPDX-License-Identifier: MIT
//
#include "ConvertFp32ToFp16Layer.hpp"
#include "LayerCloneBase.hpp"
#include <armnn/TypesUtils.hpp>
#include <backendsCommon/WorkloadData.hpp>
#include <backendsCommon/WorkloadFactory.hpp>
namespace armnn
{
ConvertFp32ToFp16Layer::ConvertFp32ToFp16Layer(const char* name)
: Layer(1, 1, LayerType::ConvertFp32ToFp16, name)
{
}
std::unique_ptr<IWorkload> ConvertFp32ToFp16Layer::CreateWorkload(const Graph& graph,
const IWorkloadFactory& factory) const
{
ConvertFp32ToFp16QueueDescriptor descriptor;
return factory.CreateConvertFp32ToFp16(descriptor, PrepInfoAndDesc(descriptor, graph));
}
ConvertFp32ToFp16Layer* ConvertFp32ToFp16Layer::Clone(Graph& graph) const
{
return CloneBase<ConvertFp32ToFp16Layer>(graph, GetName());
}
void ConvertFp32ToFp16Layer::ValidateTensorShapesFromInputs()
{
VerifyLayerConnections(1, CHECK_LOCATION());
auto inferredShapes = InferOutputShapes({ GetInputSlot(0).GetConnection()->GetTensorInfo().GetShape() });
BOOST_ASSERT(inferredShapes.size() == 1);
ConditionalThrowIfNotEqual<LayerValidationException>(
"ConvertFp32ToFp16Layer: TensorShape set on OutputSlot[0] does not match the inferred shape.",
GetOutputSlot(0).GetTensorInfo().GetShape(),
inferredShapes[0]);
}
} // namespace armnn
|
