Isabella Gottardi | bc4484a | 2018-02-02 11:27:32 +0000 | [diff] [blame] | 1 | /* |
| 2 | * Copyright (c) 2018 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 | */ |
Georgios Pinitas | d8734b5 | 2017-12-22 15:27:52 +0000 | [diff] [blame] | 24 | #include "arm_compute/graph2.h" |
Isabella Gottardi | bc4484a | 2018-02-02 11:27:32 +0000 | [diff] [blame] | 25 | #include "support/ToolchainSupport.h" |
| 26 | #include "utils/GraphUtils.h" |
| 27 | #include "utils/Utils.h" |
| 28 | |
| 29 | #include <cstdlib> |
| 30 | #include <tuple> |
| 31 | |
| 32 | using namespace arm_compute::utils; |
Georgios Pinitas | d8734b5 | 2017-12-22 15:27:52 +0000 | [diff] [blame] | 33 | using namespace arm_compute::graph2::frontend; |
Isabella Gottardi | bc4484a | 2018-02-02 11:27:32 +0000 | [diff] [blame] | 34 | using namespace arm_compute::graph_utils; |
Isabella Gottardi | bc4484a | 2018-02-02 11:27:32 +0000 | [diff] [blame] | 35 | |
| 36 | namespace |
| 37 | { |
| 38 | } // namespace |
| 39 | |
| 40 | /** Example demonstrating how to implement Squeezenet's v1.1 network using the Compute Library's graph API |
| 41 | * |
| 42 | * @param[in] argc Number of arguments |
| 43 | * @param[in] argv Arguments ( [optional] Target (0 = NEON, 1 = OpenCL), [optional] Path to the weights folder, [optional] image, [optional] labels ) |
| 44 | */ |
| 45 | class GraphSqueezenet_v1_1Example : public Example |
| 46 | { |
| 47 | public: |
| 48 | void do_setup(int argc, char **argv) override |
| 49 | { |
| 50 | std::string data_path; /* Path to the trainable data */ |
| 51 | std::string image; /* Image data */ |
| 52 | std::string label; /* Label data */ |
| 53 | |
Georgios Pinitas | 140fdc7 | 2018-02-16 11:42:38 +0000 | [diff] [blame] | 54 | // Create a preprocessor object |
| 55 | const std::array<float, 3> mean_rgb{ { 122.68f, 116.67f, 104.01f } }; |
| 56 | std::unique_ptr<IPreprocessor> preprocessor = arm_compute::support::cpp14::make_unique<CaffePreproccessor>(mean_rgb); |
Isabella Gottardi | bc4484a | 2018-02-02 11:27:32 +0000 | [diff] [blame] | 57 | |
Michele Di Giorgio | e3fba0a | 2018-02-14 14:18:01 +0000 | [diff] [blame] | 58 | // Set target. 0 (NEON), 1 (OpenCL), 2 (OpenCL with Tuner). By default it is NEON |
Georgios Pinitas | 9a8c672 | 2018-03-21 17:52:35 +0000 | [diff] [blame] | 59 | const int target = argc > 1 ? std::strtol(argv[1], nullptr, 10) : 0; |
| 60 | Target target_hint = set_target_hint2(target); |
Isabella Gottardi | bc4484a | 2018-02-02 11:27:32 +0000 | [diff] [blame] | 61 | |
Georgios Pinitas | 2870516 | 2018-03-21 20:10:53 +0000 | [diff] [blame] | 62 | ConvolutionMethod convolution_hint = (target_hint == Target::CL) ? ConvolutionMethod::WINOGRAD : ConvolutionMethod::GEMM; |
| 63 | |
Isabella Gottardi | bc4484a | 2018-02-02 11:27:32 +0000 | [diff] [blame] | 64 | // Parse arguments |
| 65 | if(argc < 2) |
| 66 | { |
| 67 | // Print help |
| 68 | std::cout << "Usage: " << argv[0] << " [target] [path_to_data] [image] [labels]\n\n"; |
| 69 | std::cout << "No data folder provided: using random values\n\n"; |
| 70 | } |
| 71 | else if(argc == 2) |
| 72 | { |
| 73 | std::cout << "Usage: " << argv[0] << " " << argv[1] << " [path_to_data] [image] [labels]\n\n"; |
| 74 | std::cout << "No data folder provided: using random values\n\n"; |
| 75 | } |
| 76 | else if(argc == 3) |
| 77 | { |
| 78 | data_path = argv[2]; |
| 79 | std::cout << "Usage: " << argv[0] << " " << argv[1] << " " << argv[2] << " [image] [labels]\n\n"; |
| 80 | std::cout << "No image provided: using random values\n\n"; |
| 81 | } |
| 82 | else if(argc == 4) |
| 83 | { |
| 84 | data_path = argv[2]; |
| 85 | image = argv[3]; |
| 86 | std::cout << "Usage: " << argv[0] << " " << argv[1] << " " << argv[2] << " " << argv[3] << " [labels]\n\n"; |
| 87 | std::cout << "No text file with labels provided: skipping output accessor\n\n"; |
| 88 | } |
| 89 | else |
| 90 | { |
| 91 | data_path = argv[2]; |
| 92 | image = argv[3]; |
| 93 | label = argv[4]; |
| 94 | } |
| 95 | |
Isabella Gottardi | bc4484a | 2018-02-02 11:27:32 +0000 | [diff] [blame] | 96 | graph << target_hint |
Georgios Pinitas | d8734b5 | 2017-12-22 15:27:52 +0000 | [diff] [blame] | 97 | << InputLayer(TensorDescriptor(TensorShape(227U, 227U, 3U, 1U), DataType::F32), |
| 98 | get_input_accessor(image, std::move(preprocessor))) |
Georgios Pinitas | 2870516 | 2018-03-21 20:10:53 +0000 | [diff] [blame] | 99 | << ConvolutionMethod::DIRECT |
Isabella Gottardi | bc4484a | 2018-02-02 11:27:32 +0000 | [diff] [blame] | 100 | << ConvolutionLayer( |
| 101 | 3U, 3U, 64U, |
| 102 | get_weights_accessor(data_path, "/cnn_data/squeezenet_v1_1_model/conv1_w.npy"), |
| 103 | get_weights_accessor(data_path, "/cnn_data/squeezenet_v1_1_model/conv1_b.npy"), |
| 104 | PadStrideInfo(2, 2, 0, 0)) |
| 105 | << ActivationLayer(ActivationLayerInfo(ActivationLayerInfo::ActivationFunction::RELU)) |
| 106 | << PoolingLayer(PoolingLayerInfo(PoolingType::MAX, 3, PadStrideInfo(2, 2, 0, 0, DimensionRoundingType::CEIL))) |
Georgios Pinitas | 2870516 | 2018-03-21 20:10:53 +0000 | [diff] [blame] | 107 | << convolution_hint |
Isabella Gottardi | bc4484a | 2018-02-02 11:27:32 +0000 | [diff] [blame] | 108 | << ConvolutionLayer( |
| 109 | 1U, 1U, 16U, |
| 110 | get_weights_accessor(data_path, "/cnn_data/squeezenet_v1_1_model/fire2_squeeze1x1_w.npy"), |
| 111 | get_weights_accessor(data_path, "/cnn_data/squeezenet_v1_1_model/fire2_squeeze1x1_b.npy"), |
| 112 | PadStrideInfo(1, 1, 0, 0)) |
| 113 | << ActivationLayer(ActivationLayerInfo(ActivationLayerInfo::ActivationFunction::RELU)) |
| 114 | << get_expand_fire_node(data_path, "fire2", 64U, 64U) |
| 115 | << ConvolutionLayer( |
| 116 | 1U, 1U, 16U, |
| 117 | get_weights_accessor(data_path, "/cnn_data/squeezenet_v1_1_model/fire3_squeeze1x1_w.npy"), |
| 118 | get_weights_accessor(data_path, "/cnn_data/squeezenet_v1_1_model/fire3_squeeze1x1_b.npy"), |
| 119 | PadStrideInfo(1, 1, 0, 0)) |
| 120 | << ActivationLayer(ActivationLayerInfo(ActivationLayerInfo::ActivationFunction::RELU)) |
| 121 | << get_expand_fire_node(data_path, "fire3", 64U, 64U) |
| 122 | << PoolingLayer(PoolingLayerInfo(PoolingType::MAX, 3, PadStrideInfo(2, 2, 0, 0, DimensionRoundingType::CEIL))) |
| 123 | << ConvolutionLayer( |
| 124 | 1U, 1U, 32U, |
| 125 | get_weights_accessor(data_path, "/cnn_data/squeezenet_v1_1_model/fire4_squeeze1x1_w.npy"), |
| 126 | get_weights_accessor(data_path, "/cnn_data/squeezenet_v1_1_model/fire4_squeeze1x1_b.npy"), |
| 127 | PadStrideInfo(1, 1, 0, 0)) |
| 128 | << ActivationLayer(ActivationLayerInfo(ActivationLayerInfo::ActivationFunction::RELU)) |
| 129 | << get_expand_fire_node(data_path, "fire4", 128U, 128U) |
| 130 | << ConvolutionLayer( |
| 131 | 1U, 1U, 32U, |
| 132 | get_weights_accessor(data_path, "/cnn_data/squeezenet_v1_1_model/fire5_squeeze1x1_w.npy"), |
| 133 | get_weights_accessor(data_path, "/cnn_data/squeezenet_v1_1_model/fire5_squeeze1x1_b.npy"), |
| 134 | PadStrideInfo(1, 1, 0, 0)) |
| 135 | << ActivationLayer(ActivationLayerInfo(ActivationLayerInfo::ActivationFunction::RELU)) |
| 136 | << get_expand_fire_node(data_path, "fire5", 128U, 128U) |
| 137 | << PoolingLayer(PoolingLayerInfo(PoolingType::MAX, 3, PadStrideInfo(2, 2, 0, 0, DimensionRoundingType::CEIL))) |
| 138 | << ConvolutionLayer( |
| 139 | 1U, 1U, 48U, |
| 140 | get_weights_accessor(data_path, "/cnn_data/squeezenet_v1_1_model/fire6_squeeze1x1_w.npy"), |
| 141 | get_weights_accessor(data_path, "/cnn_data/squeezenet_v1_1_model/fire6_squeeze1x1_b.npy"), |
| 142 | PadStrideInfo(1, 1, 0, 0)) |
| 143 | << ActivationLayer(ActivationLayerInfo(ActivationLayerInfo::ActivationFunction::RELU)) |
| 144 | << get_expand_fire_node(data_path, "fire6", 192U, 192U) |
| 145 | << ConvolutionLayer( |
| 146 | 1U, 1U, 48U, |
| 147 | get_weights_accessor(data_path, "/cnn_data/squeezenet_v1_1_model/fire7_squeeze1x1_w.npy"), |
| 148 | get_weights_accessor(data_path, "/cnn_data/squeezenet_v1_1_model/fire7_squeeze1x1_b.npy"), |
| 149 | PadStrideInfo(1, 1, 0, 0)) |
| 150 | << ActivationLayer(ActivationLayerInfo(ActivationLayerInfo::ActivationFunction::RELU)) |
| 151 | << get_expand_fire_node(data_path, "fire7", 192U, 192U) |
| 152 | << ConvolutionLayer( |
| 153 | 1U, 1U, 64U, |
| 154 | get_weights_accessor(data_path, "/cnn_data/squeezenet_v1_1_model/fire8_squeeze1x1_w.npy"), |
| 155 | get_weights_accessor(data_path, "/cnn_data/squeezenet_v1_1_model/fire8_squeeze1x1_b.npy"), |
| 156 | PadStrideInfo(1, 1, 0, 0)) |
| 157 | << ActivationLayer(ActivationLayerInfo(ActivationLayerInfo::ActivationFunction::RELU)) |
| 158 | << get_expand_fire_node(data_path, "fire8", 256U, 256U) |
| 159 | << ConvolutionLayer( |
| 160 | 1U, 1U, 64U, |
| 161 | get_weights_accessor(data_path, "/cnn_data/squeezenet_v1_1_model/fire9_squeeze1x1_w.npy"), |
| 162 | get_weights_accessor(data_path, "/cnn_data/squeezenet_v1_1_model/fire9_squeeze1x1_b.npy"), |
| 163 | PadStrideInfo(1, 1, 0, 0)) |
| 164 | << ActivationLayer(ActivationLayerInfo(ActivationLayerInfo::ActivationFunction::RELU)) |
| 165 | << get_expand_fire_node(data_path, "fire9", 256U, 256U) |
| 166 | << ConvolutionLayer( |
| 167 | 1U, 1U, 1000U, |
| 168 | get_weights_accessor(data_path, "/cnn_data/squeezenet_v1_1_model/conv10_w.npy"), |
| 169 | get_weights_accessor(data_path, "/cnn_data/squeezenet_v1_1_model/conv10_b.npy"), |
| 170 | PadStrideInfo(1, 1, 0, 0)) |
| 171 | << ActivationLayer(ActivationLayerInfo(ActivationLayerInfo::ActivationFunction::RELU)) |
| 172 | << PoolingLayer(PoolingLayerInfo(PoolingType::AVG)) |
| 173 | << FlattenLayer() |
| 174 | << SoftmaxLayer() |
Georgios Pinitas | d8734b5 | 2017-12-22 15:27:52 +0000 | [diff] [blame] | 175 | << OutputLayer(get_output_accessor(label, 5)); |
Gian Marco | c1b6e37 | 2018-02-21 18:03:26 +0000 | [diff] [blame] | 176 | |
Georgios Pinitas | d8734b5 | 2017-12-22 15:27:52 +0000 | [diff] [blame] | 177 | // Finalize graph |
Georgios Pinitas | 9a8c672 | 2018-03-21 17:52:35 +0000 | [diff] [blame] | 178 | GraphConfig config; |
| 179 | config.use_function_memory_manager = true; |
| 180 | config.use_tuner = (target == 2); |
| 181 | graph.finalize(target_hint, config); |
Isabella Gottardi | bc4484a | 2018-02-02 11:27:32 +0000 | [diff] [blame] | 182 | } |
| 183 | void do_run() override |
| 184 | { |
| 185 | // Run graph |
| 186 | graph.run(); |
| 187 | } |
| 188 | |
| 189 | private: |
Georgios Pinitas | d8734b5 | 2017-12-22 15:27:52 +0000 | [diff] [blame] | 190 | Stream graph{ 0, "SqueezeNetV1.1" }; |
Isabella Gottardi | bc4484a | 2018-02-02 11:27:32 +0000 | [diff] [blame] | 191 | |
| 192 | BranchLayer get_expand_fire_node(const std::string &data_path, std::string &¶m_path, unsigned int expand1_filt, unsigned int expand3_filt) |
| 193 | { |
| 194 | std::string total_path = "/cnn_data/squeezenet_v1_1_model/" + param_path + "_"; |
Georgios Pinitas | d8734b5 | 2017-12-22 15:27:52 +0000 | [diff] [blame] | 195 | SubStream i_a(graph); |
Isabella Gottardi | bc4484a | 2018-02-02 11:27:32 +0000 | [diff] [blame] | 196 | i_a << ConvolutionLayer( |
| 197 | 1U, 1U, expand1_filt, |
| 198 | get_weights_accessor(data_path, total_path + "expand1x1_w.npy"), |
| 199 | get_weights_accessor(data_path, total_path + "expand1x1_b.npy"), |
| 200 | PadStrideInfo(1, 1, 0, 0)) |
| 201 | << ActivationLayer(ActivationLayerInfo(ActivationLayerInfo::ActivationFunction::RELU)); |
| 202 | |
Georgios Pinitas | d8734b5 | 2017-12-22 15:27:52 +0000 | [diff] [blame] | 203 | SubStream i_b(graph); |
Isabella Gottardi | bc4484a | 2018-02-02 11:27:32 +0000 | [diff] [blame] | 204 | i_b << ConvolutionLayer( |
| 205 | 3U, 3U, expand3_filt, |
| 206 | get_weights_accessor(data_path, total_path + "expand3x3_w.npy"), |
| 207 | get_weights_accessor(data_path, total_path + "expand3x3_b.npy"), |
| 208 | PadStrideInfo(1, 1, 1, 1)) |
| 209 | << ActivationLayer(ActivationLayerInfo(ActivationLayerInfo::ActivationFunction::RELU)); |
| 210 | |
| 211 | return BranchLayer(BranchMergeMethod::DEPTH_CONCATENATE, std::move(i_a), std::move(i_b)); |
| 212 | } |
| 213 | }; |
| 214 | |
| 215 | /** Main program for Squeezenet v1.1 |
| 216 | * |
| 217 | * @param[in] argc Number of arguments |
| 218 | * @param[in] argv Arguments ( [optional] Target (0 = NEON, 1 = OpenCL), [optional] Path to the weights folder, [optional] image, [optional] labels ) |
| 219 | */ |
| 220 | int main(int argc, char **argv) |
| 221 | { |
| 222 | return arm_compute::utils::run_example<GraphSqueezenet_v1_1Example>(argc, argv); |
| 223 | } |