xref: /aosp_15_r20/external/ComputeLibrary/tests/datasets/SmallConvolutionLayerDataset.h (revision c217d954acce2dbc11938adb493fc0abd69584f3) 
1 /*
2  * Copyright (c) 2017-2021 Arm Limited.
3  *
4  * SPDX-License-Identifier: MIT
5  *
6  * Permission is hereby granted, free of charge, to any person obtaining a copy
7  * of this software and associated documentation files (the "Software"), to
8  * deal in the Software without restriction, including without limitation the
9  * rights to use, copy, modify, merge, publish, distribute, sublicense, and/or
10  * sell copies of the Software, and to permit persons to whom the Software is
11  * furnished to do so, subject to the following conditions:
12  *
13  * The above copyright notice and this permission notice shall be included in all
14  * copies or substantial portions of the Software.
15  *
16  * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
17  * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
18  * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
19  * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
20  * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
21  * OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
22  * SOFTWARE.
23  */
24 #ifndef ARM_COMPUTE_TEST_SMALL_CONVOLUTION_LAYER_DATASET
25 #define ARM_COMPUTE_TEST_SMALL_CONVOLUTION_LAYER_DATASET
26 
27 #include "tests/datasets/ConvolutionLayerDataset.h"
28 
29 #include "utils/TypePrinter.h"
30 
31 #include "arm_compute/core/TensorShape.h"
32 #include "arm_compute/core/Types.h"
33 
34 namespace arm_compute
35 {
36 namespace test
37 {
38 namespace datasets
39 {
40 class SmallWinogradConvolutionLayer3x3Dataset final : public ConvolutionLayerDataset
41 {
42 public:
SmallWinogradConvolutionLayer3x3Dataset()43     SmallWinogradConvolutionLayer3x3Dataset()
44     {
45         // Channel size big enough to force multithreaded execution of the input transform
46         add_config(TensorShape(8U, 8U, 32U), TensorShape(3U, 3U, 32U, 1U), TensorShape(1U), TensorShape(6U, 6U, 1U), PadStrideInfo(1, 1, 0, 0));
47         // Batch size 1
48         add_config(TensorShape(8U, 8U, 2U), TensorShape(3U, 3U, 2U, 1U), TensorShape(1U), TensorShape(6U, 6U, 1U), PadStrideInfo(1, 1, 0, 0));
49         // Batch size 4
50         add_config(TensorShape(23U, 27U, 5U, 4U), TensorShape(3U, 3U, 5U, 21U), TensorShape(21U), TensorShape(21U, 25U, 21U, 4U), PadStrideInfo(1, 1, 0, 0));
51         add_config(TensorShape(8U, 8U, 2U), TensorShape(3U, 3U, 2U, 1U), TensorShape(1U), TensorShape(8U, 8U, 1U), PadStrideInfo(1, 1, 1, 1));
52         add_config(TensorShape(3U, 9U), TensorShape(3U, 3U), TensorShape(1), TensorShape(3U, 9U), PadStrideInfo(1, 1, 1, 1));
53     }
54 };
55 
56 class SmallWinogradConvolutionLayer3x1Dataset final : public ConvolutionLayerDataset
57 {
58 public:
SmallWinogradConvolutionLayer3x1Dataset()59     SmallWinogradConvolutionLayer3x1Dataset()
60     {
61         // Channel size big enough to force multithreaded execution of the input transform
62         add_config(TensorShape(8U, 8U, 32U), TensorShape(3U, 1U, 32U, 1U), TensorShape(1U), TensorShape(6U, 8U, 1U), PadStrideInfo(1, 1, 0, 0));
63         // Batch size 1
64         add_config(TensorShape(8U, 8U, 2U), TensorShape(3U, 1U, 2U, 1U), TensorShape(1U), TensorShape(6U, 8U, 1U), PadStrideInfo(1, 1, 0, 0));
65         // Batch size 4
66         add_config(TensorShape(23U, 27U, 5U, 4U), TensorShape(3U, 1U, 5U, 21U), TensorShape(21U), TensorShape(21U, 27U, 21U, 4U), PadStrideInfo(1, 1, 0, 0));
67         add_config(TensorShape(8U, 8U, 2U), TensorShape(3U, 1U, 2U, 1U), TensorShape(1U), TensorShape(8U, 8U, 1U), PadStrideInfo(1, 1, 1, 0));
68     }
69 };
70 
71 class SmallWinogradConvolutionLayer1x3Dataset final : public ConvolutionLayerDataset
72 {
73 public:
SmallWinogradConvolutionLayer1x3Dataset()74     SmallWinogradConvolutionLayer1x3Dataset()
75     {
76         // Channel size big enough to force multithreaded execution of the input transform
77         add_config(TensorShape(8U, 8U, 32U), TensorShape(1U, 3U, 32U, 1U), TensorShape(1U), TensorShape(8U, 6U, 1U), PadStrideInfo(1, 1, 0, 0));
78         // Batch size 1
79         add_config(TensorShape(8U, 8U, 2U), TensorShape(1U, 3U, 2U, 1U), TensorShape(1U), TensorShape(8U, 6U, 1U), PadStrideInfo(1, 1, 0, 0));
80         // Batch size 4
81         add_config(TensorShape(23U, 27U, 5U, 4U), TensorShape(1U, 3U, 5U, 21U), TensorShape(21U), TensorShape(23U, 25U, 21U, 4U), PadStrideInfo(1, 1, 0, 0));
82         add_config(TensorShape(8U, 8U, 2U), TensorShape(1U, 3U, 2U, 1U), TensorShape(1U), TensorShape(8U, 8U, 1U), PadStrideInfo(1, 1, 0, 1));
83     }
84 };
85 
86 class SmallWinogradConvolutionLayer5x5Dataset final : public ConvolutionLayerDataset
87 {
88 public:
SmallWinogradConvolutionLayer5x5Dataset()89     SmallWinogradConvolutionLayer5x5Dataset()
90     {
91         add_config(TensorShape(8U, 8U, 2U), TensorShape(5U, 5U, 2U, 1U), TensorShape(1U), TensorShape(4U, 4U, 1U), PadStrideInfo(1, 1, 0, 0));
92         add_config(TensorShape(8U, 8U, 2U), TensorShape(5U, 5U, 2U), TensorShape(1U), TensorShape(8U, 8U, 1U), PadStrideInfo(1, 1, 2, 2));
93     }
94 };
95 
96 class SmallWinogradConvolutionLayer5x1Dataset final : public ConvolutionLayerDataset
97 {
98 public:
SmallWinogradConvolutionLayer5x1Dataset()99     SmallWinogradConvolutionLayer5x1Dataset()
100     {
101         add_config(TensorShape(8U, 8U, 2U), TensorShape(5U, 1U, 2U, 1U), TensorShape(1U), TensorShape(4U, 8U, 1U), PadStrideInfo(1, 1, 0, 0));
102         add_config(TensorShape(8U, 8U, 2U), TensorShape(5U, 1U, 2U), TensorShape(1U), TensorShape(8U, 8U, 1U), PadStrideInfo(1, 1, 2, 0));
103     }
104 };
105 
106 class SmallWinogradConvolutionLayer1x5Dataset final : public ConvolutionLayerDataset
107 {
108 public:
SmallWinogradConvolutionLayer1x5Dataset()109     SmallWinogradConvolutionLayer1x5Dataset()
110     {
111         add_config(TensorShape(8U, 8U, 2U), TensorShape(1U, 5U, 2U, 1U), TensorShape(1U), TensorShape(8U, 4U, 1U), PadStrideInfo(1, 1, 0, 0));
112         add_config(TensorShape(8U, 8U, 2U), TensorShape(1U, 5U, 2U), TensorShape(1U), TensorShape(8U, 8U, 1U), PadStrideInfo(1, 1, 0, 2));
113     }
114 };
115 
116 class SmallWinogradConvolutionLayer7x1Dataset final : public ConvolutionLayerDataset
117 {
118 public:
SmallWinogradConvolutionLayer7x1Dataset()119     SmallWinogradConvolutionLayer7x1Dataset()
120     {
121         add_config(TensorShape(14U, 14U, 2U), TensorShape(7U, 1U, 2U, 1U), TensorShape(1U), TensorShape(8U, 14U, 1U), PadStrideInfo(1, 1, 0, 0));
122         add_config(TensorShape(14U, 14U, 2U), TensorShape(7U, 1U, 2U), TensorShape(1U), TensorShape(14U, 14U, 1U), PadStrideInfo(1, 1, 3, 0));
123     }
124 };
125 
126 class SmallWinogradConvolutionLayer1x7Dataset final : public ConvolutionLayerDataset
127 {
128 public:
SmallWinogradConvolutionLayer1x7Dataset()129     SmallWinogradConvolutionLayer1x7Dataset()
130     {
131         add_config(TensorShape(14U, 14U, 2U), TensorShape(1U, 7U, 2U, 1U), TensorShape(1U), TensorShape(14U, 8U, 1U), PadStrideInfo(1, 1, 0, 0));
132         add_config(TensorShape(14U, 14U, 2U), TensorShape(1U, 7U, 2U), TensorShape(1U), TensorShape(14U, 14U, 1U), PadStrideInfo(1, 1, 0, 3));
133     }
134 };
135 
136 class SmallFFTConvolutionLayerDataset final : public ConvolutionLayerDataset
137 {
138 public:
SmallFFTConvolutionLayerDataset()139     SmallFFTConvolutionLayerDataset()
140     {
141         add_config(TensorShape(8U, 7U, 3U), TensorShape(3U, 3U, 3U, 2U), TensorShape(2U), TensorShape(8U, 7U, 2U), PadStrideInfo(1, 1, 1, 1));
142         add_config(TensorShape(64U, 32U, 5U), TensorShape(5U, 5U, 5U, 10U), TensorShape(10U), TensorShape(64U, 32U, 10U), PadStrideInfo(1, 1, 2, 2));
143         add_config(TensorShape(192U, 128U, 8U), TensorShape(9U, 9U, 8U, 3U), TensorShape(3U), TensorShape(192U, 128U, 3U), PadStrideInfo(1, 1, 4, 4));
144     }
145 };
146 
147 class SmallConvolutionLayerDataset final : public ConvolutionLayerDataset
148 {
149 public:
SmallConvolutionLayerDataset()150     SmallConvolutionLayerDataset()
151     {
152         // 1D Kernel
153         add_config(TensorShape(1U, 5U, 2U), TensorShape(1U, 3U, 2U, 3U), TensorShape(3U), TensorShape(1U, 7U, 3U), PadStrideInfo(1, 1, 0, 0, 2, 2, DimensionRoundingType::FLOOR));
154 
155         // 1x1 Kernel with Stride (1, 1) and NHWC data layout in order to test skipping Im2Col
156         add_config(TensorShape(1U, 5U, 2U), TensorShape(1U, 1U, 2U, 3U), TensorShape(3U), TensorShape(1U, 5U, 3U), PadStrideInfo(1, 1, 0, 0));
157 
158         // Batch size 1
159         add_config(TensorShape(23U, 27U, 5U), TensorShape(3U, 3U, 5U, 2U), TensorShape(2U), TensorShape(11U, 25U, 2U), PadStrideInfo(2, 1, 0, 0));
160         add_config(TensorShape(33U, 27U, 7U), TensorShape(5U, 5U, 7U, 3U), TensorShape(3U), TensorShape(11U, 12U, 3U), PadStrideInfo(3, 2, 1, 0));
161         add_config(TensorShape(17U, 31U, 2U), TensorShape(5U, 5U, 2U, 4U), TensorShape(4U), TensorShape(15U, 15U, 4U), PadStrideInfo(1, 2, 1, 1));
162         add_config(TensorShape(3U, 3U, 1U), TensorShape(2U, 2U, 1U, 5U), TensorShape(5U), TensorShape(2U, 2U, 5U), PadStrideInfo(1, 1, 0, 0));
163 
164         // Batch size different than one
165         add_config(TensorShape(23U, 27U, 5U, 4U), TensorShape(3U, 3U, 5U, 21U), TensorShape(21U), TensorShape(11U, 25U, 21U, 4U), PadStrideInfo(2, 1, 0, 0));
166         add_config(TensorShape(33U, 27U, 7U, 4U), TensorShape(5U, 5U, 7U, 16U), TensorShape(16U), TensorShape(11U, 12U, 16U, 4U), PadStrideInfo(3, 2, 1, 0));
167         add_config(TensorShape(17U, 31U, 2U, 4U), TensorShape(5U, 5U, 2U, 19U), TensorShape(19U), TensorShape(15U, 15U, 19U, 4U), PadStrideInfo(1, 2, 1, 1));
168 
169         // FC convolution
170         add_config(TensorShape(1U, 1U, 1024U), TensorShape(1U, 1U, 1024U, 1001U), TensorShape(1001U), TensorShape(1U, 1U, 1001U), PadStrideInfo(1, 1, 0, 0));
171 
172         // Asymmetric padding
173         add_config(TensorShape(33U, 27U, 3U), TensorShape(5U, 7U, 3U, 4U), TensorShape(4U), TensorShape(11U, 12U, 4U), PadStrideInfo(3, 2, 1, 1, 2, 0, DimensionRoundingType::FLOOR));
174         add_config(TensorShape(33U, 27U, 3U), TensorShape(5U, 7U, 3U, 4U), TensorShape(4U), TensorShape(11U, 12U, 4U), PadStrideInfo(3, 2, 1, 1, 0, 2, DimensionRoundingType::FLOOR));
175         add_config(TensorShape(33U, 27U, 3U), TensorShape(5U, 7U, 3U, 4U), TensorShape(4U), TensorShape(11U, 12U, 4U), PadStrideInfo(3, 2, 2, 1, 2, 0, DimensionRoundingType::FLOOR));
176         add_config(TensorShape(33U, 27U, 3U), TensorShape(5U, 7U, 3U, 4U), TensorShape(4U), TensorShape(11U, 12U, 4U), PadStrideInfo(3, 2, 1, 3, 0, 2, DimensionRoundingType::FLOOR));
177         add_config(TensorShape(33U, 27U, 3U), TensorShape(5U, 7U, 3U, 4U), TensorShape(4U), TensorShape(10U, 11U, 4U), PadStrideInfo(3, 2, 1, 0, 1, 0, DimensionRoundingType::FLOOR));
178         add_config(TensorShape(33U, 27U, 3U), TensorShape(5U, 7U, 3U, 4U), TensorShape(4U), TensorShape(10U, 11U, 4U), PadStrideInfo(3, 2, 0, 1, 0, 1, DimensionRoundingType::FLOOR));
179 
180         add_config(TensorShape(5U, 4U, 3U, 2U), TensorShape(4U, 4U, 3U, 1U), TensorShape(1U), TensorShape(2U, 1U, 1U, 2U), PadStrideInfo(1, 1, 0, 0, 0, 0, DimensionRoundingType::FLOOR));
181     }
182 };
183 
184 class SmallConvolutionLayerPrePaddingDataset final : public ConvolutionLayerDataset
185 {
186 public:
SmallConvolutionLayerPrePaddingDataset()187     SmallConvolutionLayerPrePaddingDataset()
188     {
189         // output shape is calculated by accounting pre-padding layer as well -- all the data is in nchw
190         add_config(TensorShape(17U, 31U, 2U), TensorShape(5U, 5U, 2U, 19U), TensorShape(19U), TensorShape(17U, 16U, 19U), PadStrideInfo(1, 2, 1, 1));
191         add_config(TensorShape(33U, 27U, 7U), TensorShape(5U, 5U, 7U, 16U), TensorShape(16U), TensorShape(12U, 13U, 16U), PadStrideInfo(3, 2, 2, 0));
192     }
193 };
194 
195 class SmallConvolutionLayerReducedDataset final : public ConvolutionLayerDataset
196 {
197 public:
SmallConvolutionLayerReducedDataset()198     SmallConvolutionLayerReducedDataset()
199     {
200         // Batch size 1
201         add_config(TensorShape(23U, 27U, 5U), TensorShape(3U, 3U, 5U, 21U), TensorShape(21U), TensorShape(11U, 25U, 21U), PadStrideInfo(2, 1, 0, 0));
202         add_config(TensorShape(33U, 27U, 7U), TensorShape(5U, 5U, 7U, 16U), TensorShape(16U), TensorShape(11U, 12U, 16U), PadStrideInfo(3, 2, 1, 0));
203         add_config(TensorShape(17U, 31U, 2U), TensorShape(5U, 5U, 2U, 19U), TensorShape(19U), TensorShape(15U, 15U, 19U), PadStrideInfo(1, 2, 1, 1));
204 
205         // Asymmetric padding
206         add_config(TensorShape(33U, 27U, 7U, 5U), TensorShape(5U, 7U, 7U, 16U), TensorShape(16U), TensorShape(11U, 12U, 16U, 5U), PadStrideInfo(3, 2, 1, 1, 2, 0, DimensionRoundingType::FLOOR));
207         add_config(TensorShape(33U, 27U, 7U, 5U), TensorShape(5U, 7U, 7U, 16U), TensorShape(16U), TensorShape(11U, 12U, 16U, 5U), PadStrideInfo(3, 2, 1, 1, 0, 2, DimensionRoundingType::FLOOR));
208         add_config(TensorShape(33U, 27U, 7U, 5U), TensorShape(5U, 7U, 7U, 16U), TensorShape(16U), TensorShape(11U, 12U, 16U, 5U), PadStrideInfo(3, 2, 2, 1, 2, 0, DimensionRoundingType::FLOOR));
209     }
210 };
211 
212 class SmallGroupedConvolutionLayerDataset final : public ConvolutionLayerDataset
213 {
214 public:
SmallGroupedConvolutionLayerDataset()215     SmallGroupedConvolutionLayerDataset()
216     {
217         // Batch size 1
218         // Number of groups = 2
219         add_config(TensorShape(23U, 27U, 8U), TensorShape(1U, 1U, 4U, 24U), TensorShape(24U), TensorShape(12U, 27U, 24U), PadStrideInfo(2, 1, 0, 0));
220         add_config(TensorShape(33U, 27U, 12U), TensorShape(5U, 5U, 6U, 16U), TensorShape(16U), TensorShape(11U, 12U, 16U), PadStrideInfo(3, 2, 1, 0));
221         // Number of groups = 4
222         add_config(TensorShape(23U, 27U, 8U), TensorShape(1U, 1U, 2U, 24U), TensorShape(24U), TensorShape(12U, 27U, 24U), PadStrideInfo(2, 1, 0, 0));
223         add_config(TensorShape(33U, 27U, 12U), TensorShape(5U, 5U, 4U, 15U), TensorShape(15U), TensorShape(11U, 12U, 15U), PadStrideInfo(3, 2, 1, 0));
224 
225         // Batch size 4
226         // Number of groups = 2
227         add_config(TensorShape(23U, 27U, 8U, 4U), TensorShape(1U, 1U, 4U, 24U), TensorShape(24U), TensorShape(12U, 27U, 24U, 4U), PadStrideInfo(2, 1, 0, 0));
228         add_config(TensorShape(33U, 27U, 12U, 4U), TensorShape(5U, 5U, 6U, 16U), TensorShape(16U), TensorShape(11U, 12U, 16U, 4U), PadStrideInfo(3, 2, 1, 0));
229         // Number of groups = 4
230         add_config(TensorShape(23U, 27U, 8U, 4U), TensorShape(1U, 1U, 2U, 24U), TensorShape(24U), TensorShape(12U, 27U, 24U, 4U), PadStrideInfo(2, 1, 0, 0));
231         add_config(TensorShape(33U, 27U, 12U, 4U), TensorShape(5U, 5U, 4U, 15U), TensorShape(15U), TensorShape(11U, 12U, 15U, 4U), PadStrideInfo(3, 2, 1, 0));
232 
233         // Arbitrary batch size
234         add_config(TensorShape(23U, 27U, 8U, 5U), TensorShape(1U, 1U, 4U, 24U), TensorShape(24U), TensorShape(12U, 27U, 24U, 5U), PadStrideInfo(2, 1, 0, 0));
235         add_config(TensorShape(33U, 27U, 12U, 3U), TensorShape(5U, 5U, 6U, 16U), TensorShape(16U), TensorShape(11U, 12U, 16U, 3U), PadStrideInfo(3, 2, 1, 0));
236         // Number of groups = 4
237         add_config(TensorShape(23U, 27U, 8U, 2U), TensorShape(1U, 1U, 2U, 24U), TensorShape(24U), TensorShape(12U, 27U, 24U, 2U), PadStrideInfo(2, 1, 0, 0));
238         add_config(TensorShape(33U, 27U, 12U, 5U), TensorShape(5U, 5U, 4U, 15U), TensorShape(15U), TensorShape(11U, 12U, 15U, 5U), PadStrideInfo(3, 2, 1, 0));
239 
240         // Asymmetric padding
241         add_config(TensorShape(33U, 27U, 8U, 5U), TensorShape(5U, 7U, 2U, 16U), TensorShape(16U), TensorShape(11U, 12U, 16U, 5U), PadStrideInfo(3, 2, 1, 1, 2, 0, DimensionRoundingType::FLOOR));
242         add_config(TensorShape(33U, 27U, 8U, 5U), TensorShape(5U, 7U, 4U, 16U), TensorShape(16U), TensorShape(11U, 12U, 16U, 5U), PadStrideInfo(3, 2, 1, 1, 0, 2, DimensionRoundingType::FLOOR));
243         add_config(TensorShape(33U, 27U, 6U, 5U), TensorShape(5U, 7U, 3U, 16U), TensorShape(16U), TensorShape(11U, 12U, 16U, 5U), PadStrideInfo(3, 2, 2, 1, 2, 0, DimensionRoundingType::FLOOR));
244     }
245 };
246 } // namespace datasets
247 } // namespace test
248 } // namespace arm_compute
249 #endif /* ARM_COMPUTE_TEST_SMALL_CONVOLUTION_LAYER_DATASET */
250