Gitiles
Code Review Sign In
review.mlplatform.org / ml / ComputeLibrary / 5e549fa8ff058eb06cb74de43b9f89a08d0b4a9c / . / tests / validation / CL / Pooling3dLayer.cpp
blob: 84d630e6cff7f46031797441d0f54259c8f47c09 [file] [log] [blame]
ramelg0137515692022-02-26 22:06:20 +0000[diff] [blame]1/*
2 * Copyright (c) 2022 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
25#include "arm_compute/core/TensorShape.h"
26#include "tests/framework/datasets/Datasets.h"
27
28#include "arm_compute/core/Types.h"
29#include "arm_compute/runtime/CL/CLTensor.h"
30#include "arm_compute/runtime/CL/CLTensorAllocator.h"
31#include "arm_compute/runtime/CL/functions/CLPooling3dLayer.h"
32#include "tests/CL/CLAccessor.h"
33#include "tests/PaddingCalculator.h"
34#include "tests/datasets/Pooling3dLayerDataset.h"
35#include "tests/datasets/PoolingTypesDataset.h"
36#include "tests/datasets/ShapeDatasets.h"
37#include "tests/framework/Asserts.h"
38#include "tests/framework/Macros.h"
39#include "tests/framework/datasets/Datasets.h"
40#include "tests/validation/Validation.h"
41#include "tests/validation/fixtures/Pooling3dLayerFixture.h"
42
43namespace arm_compute
44{
45namespace test
46{
47namespace validation
48{
49namespace
50{
51/** Input data sets for floating-point data types */
52const auto Pooling3dLayerDatasetFP = combine(combine(combine(combine(datasets::PoolingTypes(), framework::dataset::make("PoolingSize", { Size3D(2, 3, 2) })),
53 framework::dataset::make("Stride", { Size3D(1, 1, 1), Size3D(2, 1, 1), Size3D(1, 2, 1), Size3D(2, 2, 1) })),
54 framework::dataset::make("Padding", { Padding3D(0, 1, 0), Padding3D(1, 1, 1) })),
55 framework::dataset::make("ExcludePadding", { true, false }));
56
57const auto Pooling3dLayerDatasetFPSmall = combine(combine(combine(combine(datasets::PoolingTypes(), framework::dataset::make("PoolingSize", { Size3D(2, 2, 2), Size3D(3, 3, 3) })),
58 framework::dataset::make("Stride", { Size3D(2, 2, 2), Size3D(2, 1, 1) })),
59 framework::dataset::make("Padding", { Padding3D(0, 0, 0), Padding3D(1, 1, 1), Padding3D(1, 0, 0) })),
60 framework::dataset::make("ExcludePadding", { true, false }));
61
Mohammed Suhail Munshi5e549fa2022-03-16 11:14:06 +0000[diff] [blame^]62const auto Pooling3DLayerDatasetQuantized = combine(combine(combine(combine(framework::dataset::make("PoolingType", { PoolingType::MAX, PoolingType::AVG }),
63 framework::dataset::make("PoolingSize", { Size3D(2, 3, 2) })),
64 framework::dataset::make("Stride", { Size3D(1, 1, 1), Size3D(2, 1, 1), Size3D(1, 2, 1), Size3D(1, 1, 2), Size3D(2, 2, 1)})),
65 framework::dataset::make("Padding", { Padding3D(0, 0, 0), Padding3D(1, 1, 1), Padding3D(1, 0, 0) })),
66 framework::dataset::make("ExcludePadding", { true }));
67
ramelg0137515692022-02-26 22:06:20 +0000[diff] [blame]68using ShapeDataset = framework::dataset::ContainerDataset<std::vector<TensorShape>>;
69
Mohammed Suhail Munshi5e549fa2022-03-16 11:14:06 +0000[diff] [blame^]70constexpr AbsoluteTolerance<float> tolerance_f32(0.001f); /**< Tolerance value for comparing reference's output against implementation's output for 32-bit floating-point type */
71constexpr AbsoluteTolerance<float> tolerance_f16(0.1f); /**< Tolerance value for comparing reference's output against implementation's output for 16-bit floating-point type */
72constexpr AbsoluteTolerance<int8_t> tolerance_qasymm8_signed(1); /**< Tolerance value for comparing reference's output against implementation's output for QASYMM8_SIGNED integer datatype*/
73constexpr AbsoluteTolerance<uint8_t> tolerance_qasymm8(1); /**< Tolerance value for comparing reference's output against implementation's output for 8-bit asymmetric type */
ramelg0137515692022-02-26 22:06:20 +0000[diff] [blame]74
75} // namespace
76
77TEST_SUITE(CL)
78TEST_SUITE(Pooling3dLayer)
79
80// *INDENT-OFF*
81// clang-format off
82DATA_TEST_CASE(Validate, framework::DatasetMode::ALL, zip(zip(zip(
83 framework::dataset::make("InputInfo", { TensorInfo(TensorShape(2U, 27U, 13U, 4U, 3U), 1, DataType::F32, DataLayout::NDHWC), // Mismatching data type
84 TensorInfo(TensorShape(2U, 27U, 13U, 4U, 2U), 1, DataType::F32, DataLayout::NDHWC), // Invalid pad/size combination
85 TensorInfo(TensorShape(2U, 27U, 13U, 4U, 2U), 1, DataType::F32, DataLayout::NDHWC), // Invalid pad/size combination
86 TensorInfo(TensorShape(2U, 27U, 13U, 4U, 3U), 1, DataType::F32, DataLayout::NDHWC), // Invalid output shape
87 TensorInfo(TensorShape(5U, 13U, 15U, 2U, 3U), 1, DataType::F32, DataLayout::NDHWC), // Global Pooling
88 TensorInfo(TensorShape(13U,13U, 5U, 1U, 2U), 1, DataType::F32, DataLayout::NDHWC), // Invalid output Global Pooling
89 TensorInfo(TensorShape(5U, 13U, 13U, 4U, 4U), 1, DataType::F32, DataLayout::NDHWC), // Invalid data type
90 TensorInfo(TensorShape(5U, 13U, 13U, 4U, 4U), 1, DataType::F32, DataLayout::NDHWC),
91 TensorInfo(TensorShape(5U, 13U, 13U, 5U, 4U), 1, DataType::F32, DataLayout::NDHWC),
92 TensorInfo(TensorShape(1U, 16U, 1U, 3U, 4U), 1, DataType::F32, DataLayout::NDHWC),
93 TensorInfo(TensorShape(5U, 13U, 13U, 4U, 3U), 1, DataType::F32, DataLayout::NDHWC),
94 TensorInfo(TensorShape(5U, 13U, 13U, 4U, 2U), 1, DataType::F32, DataLayout::NDHWC),
95 TensorInfo(TensorShape(5U, 13U, 13U, 4U, 3U), 1, DataType::F32, DataLayout::NDHWC),
96 TensorInfo(TensorShape(5U, 13U, 13U, 4U, 3U), 1, DataType::F32, DataLayout::NDHWC),
97 }),
98 framework::dataset::make("OutputInfo",{ TensorInfo(TensorShape(2U, 25U, 11U, 3U, 3U), 1, DataType::F16, DataLayout::NDHWC),
99 TensorInfo(TensorShape(2U, 30U, 11U, 3U, 2U), 1, DataType::F32, DataLayout::NDHWC),
100 TensorInfo(TensorShape(2U, 25U, 16U, 3U, 2U), 1, DataType::F32, DataLayout::NDHWC),
101 TensorInfo(TensorShape(2U, 27U, 13U, 3U, 3U), 1, DataType::F32, DataLayout::NDHWC),
102 TensorInfo(TensorShape(5U, 1U, 1U, 1U, 3U), 1, DataType::F32, DataLayout::NDHWC), // Global pooling applied
103 TensorInfo(TensorShape(5U, 2U, 2U, 2U, 2U), 1, DataType::F32, DataLayout::NDHWC), // Invalid output Global Pooling
104 TensorInfo(TensorShape(5U, 12U, 12U, 3U, 4U), 1, DataType::F32, DataLayout::NDHWC),
105 TensorInfo(TensorShape(5U, 12U, 12U, 3U, 4U), 1, DataType::QASYMM8, DataLayout::NDHWC), // Invalid data type
106 TensorInfo(TensorShape(5U, 1U, 1U, 1U, 4U), 1, DataType::F32, DataLayout::NDHWC),
107 TensorInfo(TensorShape(1U, 15U, 1U, 2U, 4U), 1, DataType::F32, DataLayout::NDHWC), // Output width larger than input
108 TensorInfo(TensorShape(5U, 6U, 6U, 2U, 3U), 1, DataType::F32, DataLayout::NDHWC),
109 TensorInfo(TensorShape(5U, 6U, 6U, 2U, 2U), 1, DataType::F32, DataLayout::NDHWC),
110 TensorInfo(TensorShape(5U, 6U, 6U, 2U, 3U), 1, DataType::F32, DataLayout::NDHWC),
111 TensorInfo(TensorShape(5U, 6U, 6U, 2U, 3U), 1, DataType::F32, DataLayout::NDHWC),
112 })),
113 framework::dataset::make("PoolInfo", { Pooling3dLayerInfo(PoolingType::AVG, 3, Size3D(1, 1, 1), Padding3D(0, 0, 0)),
114 Pooling3dLayerInfo(PoolingType::AVG, 2, Size3D(1, 1, 1), Padding3D(2, 0, 0)),
115 Pooling3dLayerInfo(PoolingType::AVG, 2, Size3D(1, 1, 1), Padding3D(0, 0, 0)),
116 Pooling3dLayerInfo(PoolingType::L2, 3, Size3D(1, 1, 1), Padding3D(0, 0, 0)),
117 Pooling3dLayerInfo(PoolingType::AVG),
118 Pooling3dLayerInfo(PoolingType::MAX),
119 Pooling3dLayerInfo(PoolingType::AVG, 2, Size3D(), Padding3D(), false),
120 Pooling3dLayerInfo(PoolingType::AVG, 2, Size3D(1U, 1U, 1U), Padding3D(), false),
121 Pooling3dLayerInfo(PoolingType::AVG),
122 Pooling3dLayerInfo(PoolingType::MAX, 2, Size3D(1, 1, 2), Padding3D(0, 0, 0), false),
123 Pooling3dLayerInfo(PoolingType::AVG, 2, Size3D(2U, 2U, 2U), Padding3D(), false),
124 Pooling3dLayerInfo(PoolingType::AVG, 1, Size3D(2U, 2U, 2U), Padding3D(2, 2, 2), true), // Pool size is smaller than the padding size with padding excluded
125 Pooling3dLayerInfo(PoolingType::AVG, 1, Size3D(2U, 2U, 2U), Padding3D(2, 2, 2), false), // Pool size is smaller than the padding size with padding included
126 Pooling3dLayerInfo(PoolingType::AVG, 3, Size3D(2U, 2U, 2U), Padding3D(2,1,2,2,1,2), false, false, DimensionRoundingType::CEIL), // CEIL with asymmetric Padding
127 })),
128 framework::dataset::make("Expected", { false, false, false, false, true, false, false, false, true , false, true, false, false, false})),
129 input_info, output_info, pool_info, expected)
130{
131 ARM_COMPUTE_EXPECT(bool(CLPooling3dLayer::validate(&input_info.clone()->set_is_resizable(false), &output_info.clone()->set_is_resizable(false), pool_info)) == expected, framework::LogLevel::ERRORS);
132}
133
134
135template <typename T>
136using CLPooling3dLayerFixture = Pooling3dLayerValidationFixture<CLTensor, CLAccessor, CLPooling3dLayer, T>;
137
138template <typename T>
139using CLSpecialPooling3dLayerFixture = SpecialPooling3dLayerValidationFixture<CLTensor, CLAccessor, CLPooling3dLayer, T>;
140
141template <typename T>
142using CLPooling3dLayerGlobalFixture = Pooling3dLayerGlobalValidationFixture<CLTensor, CLAccessor, CLPooling3dLayer, T>;
143
Mohammed Suhail Munshi5e549fa2022-03-16 11:14:06 +0000[diff] [blame^]144template <typename T>
145using CLPooling3dLayerQuantizedFixture = Pooling3dLayerValidationQuantizedFixture<CLTensor, CLAccessor, CLPooling3dLayer, T>;
146
ramelg0137515692022-02-26 22:06:20 +0000[diff] [blame]147// clang-format on
148// *INDENT-ON*
Mohammed Suhail Munshi5e549fa2022-03-16 11:14:06 +0000[diff] [blame^]149TEST_SUITE(QUANTIZED)
150
151TEST_SUITE(QASYMM8)
152// Small Dataset Quantized Dataset
153FIXTURE_DATA_TEST_CASE(RunSmall, CLPooling3dLayerQuantizedFixture<uint8_t>, framework::DatasetMode::PRECOMMIT, combine(combine(combine(datasets::Small5dShapes(),
154 combine(Pooling3DLayerDatasetQuantized,
155 framework::dataset::make("DataType", DataType::QASYMM8))),
156 framework::dataset::make("InputQuantInfo", { QuantizationInfo(1.f / 127.f, 10), QuantizationInfo(1.f / 127.f, 10) })),
157 framework::dataset::make("OutputQuantInfo", { QuantizationInfo(1.f / 127.f, 5), QuantizationInfo(1.f / 127.f, 10) })))
158{
159 // Validate output
160 validate(CLAccessor(_target), _reference, tolerance_qasymm8);
161}
162
163// Large Dataset Quantized Dataset
164FIXTURE_DATA_TEST_CASE(RunLarge, CLPooling3dLayerQuantizedFixture<uint8_t>, framework::DatasetMode::NIGHTLY, combine(combine(combine(datasets::Large5dShapes(),
165 combine(Pooling3DLayerDatasetQuantized,
166 framework::dataset::make("DataType", DataType::QASYMM8))),
167 framework::dataset::make("InputQuantInfo", { QuantizationInfo(1.f / 127.f, 10), QuantizationInfo(1.f / 127.f, 10) })),
168 framework::dataset::make("OutputQuantInfo", { QuantizationInfo(1.f / 127.f, 5), QuantizationInfo(1.f / 127.f, 10) })))
169{
170 // Validate output
171 validate(CLAccessor(_target), _reference, tolerance_qasymm8);
172}
173TEST_SUITE_END()
174
175TEST_SUITE(QASYMM8_SIGNED)
176
177// Large Dataset Quantized Dataset Signed
178FIXTURE_DATA_TEST_CASE(RunSmall, CLPooling3dLayerQuantizedFixture<int8_t>, framework::DatasetMode::PRECOMMIT, combine(combine(combine(datasets::Small5dShapes(),
179 combine(Pooling3DLayerDatasetQuantized,
180 framework::dataset::make("DataType", DataType::QASYMM8_SIGNED))),
181 framework::dataset::make("InputQuantInfo", { QuantizationInfo(1.f / 127.f, -10), QuantizationInfo(1.f / 127.f, -10) })),
182 framework::dataset::make("OutputQuantInfo", { QuantizationInfo(1.f / 127.f, -5), QuantizationInfo(1.f / 127.f, -10) })))
183{
184 // Validate output
185 validate(CLAccessor(_target), _reference, tolerance_qasymm8_signed);
186}
187
188// Large Dataset Quantized pooling test
189FIXTURE_DATA_TEST_CASE(RunLarge, CLPooling3dLayerQuantizedFixture<int8_t>, framework::DatasetMode::NIGHTLY, combine(combine(combine(datasets::Large5dShapes(),
190 combine(Pooling3DLayerDatasetQuantized,
191 framework::dataset::make("DataType", DataType::QASYMM8_SIGNED))),
192 framework::dataset::make("InputQuantInfo", { QuantizationInfo(1.f / 127.f, -10), QuantizationInfo(1.f / 127.f, -10) })),
193 framework::dataset::make("OutputQuantInfo", { QuantizationInfo(1.f / 127.f, -5), QuantizationInfo(1.f / 127.f, -10) })))
194{
195 // Validate output
196 validate(CLAccessor(_target), _reference, tolerance_qasymm8_signed);
197}
198
199TEST_SUITE_END()
200TEST_SUITE_END()
201
ramelg0137515692022-02-26 22:06:20 +0000[diff] [blame]202TEST_SUITE(Float)
203TEST_SUITE(FP32)
204
205FIXTURE_DATA_TEST_CASE(RunSpecial, CLSpecialPooling3dLayerFixture<float>, framework::DatasetMode::ALL, datasets::Pooling3dLayerDatasetSpecial() * framework::dataset::make("DataType", DataType::F32))
206{
207 // Validate output
208 validate(CLAccessor(_target), _reference, tolerance_f32);
209}
210
211FIXTURE_DATA_TEST_CASE(RunSmall, CLPooling3dLayerFixture<float>, framework::DatasetMode::PRECOMMIT, combine(datasets::Small5dShapes(), combine(Pooling3dLayerDatasetFPSmall,
212 framework::dataset::make("DataType", DataType::F32))))
213{
214 // Validate output
215 validate(CLAccessor(_target), _reference, tolerance_f32);
216}
217
218FIXTURE_DATA_TEST_CASE(RunLarge, CLPooling3dLayerFixture<float>, framework::DatasetMode::NIGHTLY, combine(datasets::Large5dShapes(), combine(Pooling3dLayerDatasetFP,
Mohammed Suhail Munshi5e549fa2022-03-16 11:14:06 +0000[diff] [blame^]219 framework::dataset::make("DataType", DataType::F32))))
ramelg0137515692022-02-26 22:06:20 +0000[diff] [blame]220{
221 // Validate output
222 validate(CLAccessor(_target), _reference, tolerance_f32);
223}
224
225TEST_SUITE(GlobalPooling)
226// *INDENT-OFF*
227// clang-format off
228FIXTURE_DATA_TEST_CASE(RunSmall, CLPooling3dLayerFixture<float>, framework::DatasetMode::ALL,
229 combine(combine(combine(combine(combine(combine(
230 framework::dataset::make("InputShape", { TensorShape(3U, 27U, 13U, 4U),
231 TensorShape(4U, 27U, 13U, 4U, 2U)
232 }),
233 framework::dataset::make("PoolingType", { PoolingType::AVG, PoolingType::L2, PoolingType::MAX })),
234 framework::dataset::make("PoolingSize", { Size3D(27, 13, 4) })),
235 framework::dataset::make("Strides", Size3D(1, 1, 1))),
236 framework::dataset::make("Paddings", Padding3D(0, 0, 0))),
237 framework::dataset::make("ExcludePadding", false)),
238 framework::dataset::make("DataType", DataType::F32)))
239{
240 // Validate output
241 validate(CLAccessor(_target), _reference, tolerance_f32);
242}
243
244FIXTURE_DATA_TEST_CASE(RunSmallGlobal, CLPooling3dLayerGlobalFixture<float>, framework::DatasetMode::ALL,
245 combine(combine(
246 framework::dataset::make("InputShape", { TensorShape(27U, 13U, 4U, 3U),
247 TensorShape(27U, 13U, 4U, 4U, 2U)
248 }),
249 framework::dataset::make("PoolingType", { PoolingType::AVG, PoolingType::L2, PoolingType::MAX })),
250 framework::dataset::make("DataType", DataType::F32)))
251{
252 // Validate output
253 validate(CLAccessor(_target), _reference, tolerance_f32);
254}
255FIXTURE_DATA_TEST_CASE(RunLarge, CLPooling3dLayerFixture<float>, framework::DatasetMode::NIGHTLY,
256 combine(combine(combine(combine(combine(combine(
257 framework::dataset::make("InputShape", { TensorShape(4U, 79U, 37U, 11U),
258 TensorShape(4U, 79U, 37U, 11U, 2U)
259 }),
260 framework::dataset::make("PoolingType", { PoolingType::AVG, PoolingType::L2, PoolingType::MAX })),
261 framework::dataset::make("PoolingSize", { Size3D(79, 37, 11) })),
262 framework::dataset::make("Strides", Size3D(1, 1, 1))),
263 framework::dataset::make("Paddings", Padding3D(0, 0, 0))),
264 framework::dataset::make("ExcludePadding", false)),
265 framework::dataset::make("DataType", DataType::F32)))
266{
267 // Validate output
268 validate(CLAccessor(_target), _reference, tolerance_f32);
269}
270// clang-format on
271// *INDENT-ON*
272TEST_SUITE_END() // GlobalPooling
273TEST_SUITE_END() // FP32
274
275TEST_SUITE(FP16)
276
277FIXTURE_DATA_TEST_CASE(RunSmall, CLPooling3dLayerFixture<half>, framework::DatasetMode::PRECOMMIT, combine(datasets::Small5x5Shapes(), combine(Pooling3dLayerDatasetFPSmall,
278 framework::dataset::make("DataType", DataType::F16))))
279{
280 // Validate output
281 validate(CLAccessor(_target), _reference, tolerance_f16);
282}
283
284FIXTURE_DATA_TEST_CASE(RunLarge, CLPooling3dLayerFixture<half>, framework::DatasetMode::NIGHTLY, combine(datasets::Large5dShapes(), combine(Pooling3dLayerDatasetFP,
Mohammed Suhail Munshi5e549fa2022-03-16 11:14:06 +0000[diff] [blame^]285 framework::dataset::make("DataType", DataType::F16))))
ramelg0137515692022-02-26 22:06:20 +0000[diff] [blame]286{
287 // Validate output
288 validate(CLAccessor(_target), _reference, tolerance_f16);
289}
290
291TEST_SUITE(GlobalPooling)
292// *INDENT-OFF*
293// clang-format off
294FIXTURE_DATA_TEST_CASE(RunSmall, CLPooling3dLayerFixture<half>, framework::DatasetMode::ALL,
295 combine(combine(combine(combine(combine(combine(
296 framework::dataset::make("InputShape", { TensorShape(3U, 27U, 13U, 4U),
297 TensorShape(4U, 27U, 13U, 4U, 2U)
298 }),
299 framework::dataset::make("PoolingType", { PoolingType::AVG, PoolingType::L2, PoolingType::MAX })),
300 framework::dataset::make("PoolingSize", { Size3D(27, 13, 4) })),
301 framework::dataset::make("Strides", Size3D(1, 1, 1))),
302 framework::dataset::make("Paddings", Padding3D(0, 0, 0))),
303 framework::dataset::make("ExcludePadding", false)),
304 framework::dataset::make("DataType", DataType::F16)))
305{
306 // Validate output
307 validate(CLAccessor(_target), _reference, tolerance_f16);
308}
309
310FIXTURE_DATA_TEST_CASE(RunSmallGlobal, CLPooling3dLayerGlobalFixture<half>, framework::DatasetMode::ALL,
311 combine(combine(
312 framework::dataset::make("InputShape", { TensorShape(27U, 13U, 4U, 3U),
313 TensorShape(27U, 13U, 4U, 4U, 2U)
314 }),
315 framework::dataset::make("PoolingType", { PoolingType::AVG, PoolingType::L2, PoolingType::MAX })),
316 framework::dataset::make("DataType", DataType::F16)))
317{
318 // Validate output
319 validate(CLAccessor(_target), _reference, tolerance_f16);
320}
321FIXTURE_DATA_TEST_CASE(RunLarge, CLPooling3dLayerFixture<half>, framework::DatasetMode::NIGHTLY,
322 combine(combine(combine(combine(combine(combine(
323 framework::dataset::make("InputShape", { TensorShape(4U, 79U, 37U, 11U),
324 TensorShape(4U, 79U, 37U, 11U, 2U)
325 }),
326 framework::dataset::make("PoolingType", { PoolingType::AVG, PoolingType::L2, PoolingType::MAX })),
327 framework::dataset::make("PoolingSize", { Size3D(79, 37, 11) })),
328 framework::dataset::make("Strides", Size3D(1, 1, 1))),
329 framework::dataset::make("Paddings", Padding3D(0, 0, 0))),
330 framework::dataset::make("ExcludePadding", false)),
331 framework::dataset::make("DataType", DataType::F16)))
332{
333 // Validate output
334 validate(CLAccessor(_target), _reference, tolerance_f16);
335}
336// clang-format on
337// *INDENT-ON*
338TEST_SUITE_END() // GlobalPooling
339TEST_SUITE_END() // FP16
340TEST_SUITE_END() // Float
341TEST_SUITE_END() // Pooling3dLayer
342TEST_SUITE_END() // CL
343} // namespace validation
344} // namespace test
345} // namespace arm_compute