Gitiles
Code Review Sign In
review.mlplatform.org / ml / armnn / a0162e17c56538ee6d72ecce4c3e0836cbb34c56 / . / docs / 01_02_deserializer_serializer.dox
blob: 81b4ad0e9b6ee64f19ad00b5b56178e5699d4880 [file] [log] [blame]
Jan Eilers31a7c892021-01-29 14:18:13 +0000[diff] [blame]1/// Copyright (c) 2021 ARM Limited and Contributors. All rights reserved.
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]2///
3/// SPDX-License-Identifier: MIT
4///
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]5
6namespace armnn
7{
8/**
Jan Eilers1ed388f2021-01-28 14:45:09 +0000[diff] [blame]9@page serializer Serializer
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]10@tableofcontents
11
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]12The `armnnSerializer` is a library for serializing an Arm NN network to a stream.
13
Jan Eilers1ed388f2021-01-28 14:45:09 +0000[diff] [blame]14@section serializersupport Supported Layers
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]15
Jan Eilers1ed388f2021-01-28 14:45:09 +0000[diff] [blame]16This reference guide provides a list of layers which can be serialized by the Arm NN SDK.
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]17
Jan Eilers1ed388f2021-01-28 14:45:09 +0000[diff] [blame]18@subsection serializersupportflully Fully supported
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]19
20The Arm NN SDK Serializer currently supports the following layers:
21
22- Activation
23- Addition
24- ArgMinMax
25- BatchToSpaceNd
26- BatchNormalization
mathad01b392e982021-04-07 12:07:30 +0100[diff] [blame]27- Cast
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]28- Comparison
29- Concat
30- Constant
31- Convolution2d
32- DepthToSpace
33- DepthwiseConvolution2d
34- Dequantize
35- DetectionPostProcess
36- Division
37- ElementwiseUnary
Jan Eilers1ed388f2021-01-28 14:45:09 +0000[diff] [blame]38- Fill
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]39- Floor
40- FullyConnected
41- Gather
42- Input
43- InstanceNormalization
44- L2Normalization
James Ward590f3f72020-12-02 10:20:48 +0000[diff] [blame]45- Logical
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]46- LogSoftmax
47- Lstm
48- Maximum
49- Mean
50- Merge
51- Minimum
52- Multiplication
53- Normalization
54- Output
55- Pad
56- Permute
57- Pooling2d
58- Prelu
James Ward590f3f72020-12-02 10:20:48 +0000[diff] [blame]59- QLstm
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]60- Quantize
61- QuantizedLstm
James Ward590f3f72020-12-02 10:20:48 +0000[diff] [blame]62- Rank
Jan Eilers1121c6b2021-02-19 09:38:03 +0000[diff] [blame]63- Reduce
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]64- Reshape
65- Resize
Keith Davis3ae3f972021-05-21 16:33:48 +0100[diff] [blame]66- Shape
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]67- Slice
68- Softmax
69- SpaceToBatchNd
70- SpaceToDepth
71- Splitter
72- Stack
73- StandIn
74- StridedSlice
75- Subtraction
76- Switch
James Ward590f3f72020-12-02 10:20:48 +0000[diff] [blame]77- Transpose
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]78- TransposeConvolution2d
Narumol Prangnawarata0162e12021-07-23 14:47:49 +0100[diff] [blame^]79- UnidirectionalSequenceLstm
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]80
81More machine learning layers will be supported in future releases.
82
Jan Eilers1ed388f2021-01-28 14:45:09 +0000[diff] [blame]83@subsection serializersupportdeprecated Deprecated layers
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]84
85Some layers have been deprecated and replaced by others layers. In order to maintain backward compatibility, serializations of these deprecated layers will deserialize to the layers that have replaced them, as follows:
86
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]87- Abs will deserialize as ElementwiseUnary
Jan Eilers1ed388f2021-01-28 14:45:09 +0000[diff] [blame]88- Equal will deserialize as Comparison
89- Greater will deserialize as Comparison
90- Merger will deserialize as Concat
91- ResizeBilinear will deserialize as Resize
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]92- Rsqrt will deserialize as ElementwiseUnary
93<br/><br/><br/><br/>
94
Jan Eilers1ed388f2021-01-28 14:45:09 +0000[diff] [blame]95@page deserializer Deserializer
96@tableofcontents
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]97
98The `armnnDeserializer` is a library for loading neural networks defined by Arm NN FlatBuffers files
99into the Arm NN runtime.
100
Jan Eilers1ed388f2021-01-28 14:45:09 +0000[diff] [blame]101@section deserializersupport Supported Layers
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]102
Jan Eilers1ed388f2021-01-28 14:45:09 +0000[diff] [blame]103This reference guide provides a list of layers which can be deserialized by the Arm NN SDK.
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]104
Jan Eilers1ed388f2021-01-28 14:45:09 +0000[diff] [blame]105@subsection deserializersupportfully Fully supported
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]106
107The Arm NN SDK Deserialize parser currently supports the following layers:
108
109- Abs
110- Activation
111- Addition
112- ArgMinMax
113- BatchToSpaceNd
114- BatchNormalization
mathad01b392e982021-04-07 12:07:30 +0100[diff] [blame]115- Cast
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]116- Concat
117- Comparison
118- Constant
119- Convolution2d
120- DepthToSpace
121- DepthwiseConvolution2d
122- Dequantize
123- DetectionPostProcess
124- Division
James Ward590f3f72020-12-02 10:20:48 +0000[diff] [blame]125- ElementwiseUnary
126- Fill
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]127- Floor
128- FullyConnected
129- Gather
130- Input
131- InstanceNormalization
132- L2Normalization
James Ward590f3f72020-12-02 10:20:48 +0000[diff] [blame]133- Logical
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]134- LogSoftmax
135- Lstm
136- Maximum
137- Mean
138- Merge
139- Minimum
140- Multiplication
141- Normalization
142- Output
143- Pad
144- Permute
145- Pooling2d
146- Prelu
147- Quantize
Jan Eilers1ed388f2021-01-28 14:45:09 +0000[diff] [blame]148- QLstm
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]149- QuantizedLstm
James Ward590f3f72020-12-02 10:20:48 +0000[diff] [blame]150- Rank
Jan Eilers1121c6b2021-02-19 09:38:03 +0000[diff] [blame]151- Reduce
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]152- Reshape
James Ward590f3f72020-12-02 10:20:48 +0000[diff] [blame]153- Resize
Jan Eilers1ed388f2021-01-28 14:45:09 +0000[diff] [blame]154- ResizeBilinear
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]155- Slice
156- Softmax
157- SpaceToBatchNd
158- SpaceToDepth
159- Splitter
160- Stack
161- StandIn
162- StridedSlice
163- Subtraction
164- Switch
Jan Eilers1121c6b2021-02-19 09:38:03 +0000[diff] [blame]165- Transpose
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]166- TransposeConvolution2d
Narumol Prangnawarata0162e12021-07-23 14:47:49 +0100[diff] [blame^]167- UnidirectionalSequenceLstm
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]168
169More machine learning layers will be supported in future releases.
170
Jan Eilers1ed388f2021-01-28 14:45:09 +0000[diff] [blame]171@subsection deserializersupportdeprecated Deprecated layers
Ryan OSheaf3a43232020-02-12 16:15:27 +0000[diff] [blame]172
173Some layers have been deprecated and replaced by others layers. In order to maintain backward compatibility, serializations of these deprecated layers will deserialize to the layers that have replaced them, as follows:
174
175- Equal will deserialize as Comparison
176- Merger will deserialize as Concat
177- Greater will deserialize as Comparison
178- ResizeBilinear will deserialize as Resize
179
180**/
181}