| surmeh01 | bceff2f | 2018-03-29 16:29:27 +0100 | [diff] [blame^] | 1 | // |
| 2 | // Copyright © 2017 Arm Ltd. All rights reserved. |
| 3 | // See LICENSE file in the project root for full license information. |
| 4 | // |
| 5 | |
| 6 | #include <boost/test/unit_test.hpp> |
| 7 | #include "armnnTfParser/ITfParser.hpp" |
| 8 | #include "ParserPrototxtFixture.hpp" |
| 9 | #include <string> |
| 10 | #include <iostream> |
| 11 | |
| 12 | BOOST_AUTO_TEST_SUITE(TensorflowParser) |
| 13 | |
| 14 | struct DepthwiseConvolution2dFixture : public ParserPrototxtFixture<armnnTfParser::ITfParser> |
| 15 | { |
| 16 | explicit DepthwiseConvolution2dFixture(const char* paddingType) |
| 17 | { |
| 18 | m_Prototext = "node { \n" |
| 19 | " name: \"graphInput\" \n" |
| 20 | " op: \"Placeholder\" \n" |
| 21 | " attr { \n" |
| 22 | " key: \"dtype\" \n" |
| 23 | " value { \n" |
| 24 | " type: DT_FLOAT \n" |
| 25 | " } \n" |
| 26 | " } \n" |
| 27 | " attr { \n" |
| 28 | " key: \"value\" \n" |
| 29 | " value { \n" |
| 30 | " tensor { \n" |
| 31 | " dtype: DT_FLOAT \n" |
| 32 | " tensor_shape { \n" |
| 33 | " dim { \n" |
| 34 | " size: 1 \n" |
| 35 | " } \n" |
| 36 | " dim { \n" |
| 37 | " size: 1 \n" |
| 38 | " } \n" |
| 39 | " dim { \n" |
| 40 | " size: 3 \n" |
| 41 | " } \n" |
| 42 | " dim { \n" |
| 43 | " size: 3 \n" |
| 44 | " } \n" |
| 45 | " } \n" |
| 46 | " tensor_content: \"\\000\\000\\200?\\000\\000\\000@\\000\\000@@\\000\\000\\200@" |
| 47 | "\\000\\000\\240@\\000\\000\\300@\\000\\000\\340@\\000\\000\\000A\\000\\000\\020A\" \n" |
| 48 | " } \n" |
| 49 | " } \n" |
| 50 | " } \n" |
| 51 | " } \n" |
| 52 | " node { \n" |
| 53 | " name: \"Const_1\" \n" |
| 54 | " op: \"Const\" \n" |
| 55 | " attr { \n" |
| 56 | " key: \"dtype\" \n" |
| 57 | " value { \n" |
| 58 | " type: DT_FLOAT \n" |
| 59 | " } \n" |
| 60 | " } \n" |
| 61 | " attr { \n" |
| 62 | " key: \"value\" \n" |
| 63 | " value { \n" |
| 64 | " tensor { \n" |
| 65 | " dtype: DT_FLOAT \n" |
| 66 | " tensor_shape { \n" |
| 67 | " dim { \n" |
| 68 | " size: 1 \n" |
| 69 | " } \n" |
| 70 | " dim { \n" |
| 71 | " size: 3 \n" |
| 72 | " } \n" |
| 73 | " dim { \n" |
| 74 | " size: 3 \n" |
| 75 | " } \n" |
| 76 | " dim { \n" |
| 77 | " size: 3 \n" |
| 78 | " } \n" |
| 79 | " } \n" |
| 80 | " tensor_content: \"\\000\\000\\000?\\000\\000\\200?\\000\\000\\000?" |
| 81 | "\\000\\000\\000?\\000\\000\\200?\\000\\000\\000?" |
| 82 | "\\000\\000\\000?\\000\\000\\200?\\000\\000\\000?" |
| 83 | "\\000\\000\\000?\\000\\000\\200?\\000\\000\\000?" |
| 84 | "\\000\\000\\000?\\000\\000\\200?\\000\\000\\000?" |
| 85 | "\\000\\000\\000?\\000\\000\\200?\\000\\000\\000?" |
| 86 | "\\000\\000\\000?\\000\\000\\200?\\000\\000\\000?" |
| 87 | "\\000\\000\\000?\\000\\000\\200?\\000\\000\\000?" |
| 88 | "\\000\\000\\000?\\000\\000\\200?\\000\\000\\000?\" \n" |
| 89 | " } \n" |
| 90 | " } \n" |
| 91 | " } \n" |
| 92 | "} \n" |
| 93 | "node { \n" |
| 94 | " name: \"potato\" \n" |
| 95 | " op: \"DepthwiseConv2dNative\" \n" |
| 96 | " input: \"graphInput\" \n" |
| 97 | " input: \"Const_1\" \n" |
| 98 | " attr { \n" |
| 99 | " key: \"T\" \n" |
| 100 | " value { \n" |
| 101 | " type: DT_FLOAT \n" |
| 102 | " } \n" |
| 103 | " } \n" |
| 104 | " attr { \n" |
| 105 | " key: \"data_format\" \n" |
| 106 | " value { \n" |
| 107 | " s: \"NHWC\" \n" |
| 108 | " } \n" |
| 109 | " } \n" |
| 110 | " attr { \n" |
| 111 | " key: \"padding\" \n" |
| 112 | " value { \n" |
| 113 | " s: \""; |
| 114 | m_Prototext.append(paddingType); |
| 115 | m_Prototext.append("\"\n" |
| 116 | " } \n" |
| 117 | " } \n" |
| 118 | " attr { \n" |
| 119 | " key: \"strides\" \n" |
| 120 | " value { \n" |
| 121 | " list { \n" |
| 122 | " i: 1 \n" |
| 123 | " i: 1 \n" |
| 124 | " i: 1 \n" |
| 125 | " i: 1 \n" |
| 126 | " } \n" |
| 127 | " } \n" |
| 128 | " } \n" |
| 129 | " attr { \n" |
| 130 | " key: \"use_cudnn_on_gpu\" \n" |
| 131 | " value { \n" |
| 132 | " b: false \n" |
| 133 | " } \n" |
| 134 | " } \n" |
| 135 | "} \n"); |
| 136 | |
| 137 | SetupSingleInputSingleOutput({ 1, 1, 3, 3 }, "graphInput", "potato"); |
| 138 | } |
| 139 | }; |
| 140 | |
| 141 | struct DepthwiseConvolution2dSameFixture : DepthwiseConvolution2dFixture |
| 142 | { |
| 143 | DepthwiseConvolution2dSameFixture() : DepthwiseConvolution2dFixture("SAME") { } |
| 144 | }; |
| 145 | |
| 146 | BOOST_FIXTURE_TEST_CASE(ParseDepthwiseConv2DSame, DepthwiseConvolution2dSameFixture) |
| 147 | { |
| 148 | RunTest<4>({ 1, 2, 3, 4, 5, 6, 7, 8, 9 }, |
| 149 | { 2.5f, 5.f, 2.5f, 3.5f, 7.f, 3.5f, 4.5f, 9.f, 4.5f, |
| 150 | 6.f, 12.f, 6.f, 7.5f, 15.f, 7.5f, 9.f, 18.f, 9.f, |
| 151 | 5.5f, 11.f, 5.5f, 6.5f, 13.f, 6.5f, 7.5f, 15.f, 7.5f}); |
| 152 | } |
| 153 | |
| 154 | struct DepthwiseConvolution2dValidFixture : DepthwiseConvolution2dFixture |
| 155 | { |
| 156 | DepthwiseConvolution2dValidFixture() : DepthwiseConvolution2dFixture("VALID") { } |
| 157 | }; |
| 158 | |
| 159 | BOOST_FIXTURE_TEST_CASE(ParseDepthwiseConv2DValid, DepthwiseConvolution2dValidFixture) |
| 160 | { |
| 161 | RunTest<4>({ 1, 2, 3, 4, 5, 6, 7, 8, 9 }, // input data |
| 162 | { 6.f, 12.f, 6.f, 7.5f, 15.f, 7.5f, 9.f, 18.f, 9.f }); // output expected data |
| 163 | } |
| 164 | |
| 165 | |
| 166 | BOOST_AUTO_TEST_SUITE_END() |