/
valemobil
/
ncnn
GigaCode
beta
ОбзорЦентр заботыВойти

ncnn

Форк
0
Ветки: 2Коммиты: 3311Теги: 43
ncnn
/tests
/
test_convolution1d.cpp 
221 строка · 8.4 Кб
1
// Tencent is pleased to support the open source community by making ncnn available.
2
//
3
// Copyright (C) 2021 THL A29 Limited, a Tencent company. All rights reserved.
4
//
5
// Licensed under the BSD 3-Clause License (the "License"); you may not use this file except
6
// in compliance with the License. You may obtain a copy of the License at
7
//
8
// https://opensource.org/licenses/BSD-3-Clause
9
//
10
// Unless required by applicable law or agreed to in writing, software distributed
11
// under the License is distributed on an "AS IS" BASIS, WITHOUT WARRANTIES OR
12
// CONDITIONS OF ANY KIND, either express or implied. See the License for the
13
// specific language governing permissions and limitations under the License.
14


15
#include "testutil.h"
16


17
static int test_convolution1d(int w, int h, int outh, int kernel, int dilation, int stride, int pad, int bias)
18
{
19
    ncnn::Mat a = RandomMat(w, h);
20


21
    ncnn::ParamDict pd;
22
    pd.set(0, outh);     // num_output
23
    pd.set(1, kernel);   // kernel_w
24
    pd.set(2, dilation); // dilation_w
25
    pd.set(3, stride);   // stride_w
26
    pd.set(4, pad);      // pad_w
27
    pd.set(5, bias);     // bias_term
28
    pd.set(6, outh * h * kernel);
29


30
    int activation_type = RAND() % 6; // 0 1 2 3 4 5
31
    ncnn::Mat activation_params(2);
32
    activation_params[0] = RandomFloat(-1, 0); // alpha
33
    activation_params[1] = RandomFloat(0, 1);  // beta
34
    pd.set(9, activation_type);
35
    pd.set(10, activation_params);
36


37
    std::vector<ncnn::Mat> weights(bias ? 2 : 1);
38
    weights[0] = RandomMat(outh * h * kernel);
39
    if (bias)
40
        weights[1] = RandomMat(outh);
41


42
    int ret = test_layer("Convolution1D", pd, weights, a);
43
    if (ret != 0)
44
    {
45
        fprintf(stderr, "test_convolution1d failed w=%d h=%d outh=%d kernel=%d dilation=%d stride=%d pad=%d bias=%d act=%d actparams=[%f,%f]\n", w, h, outh, kernel, dilation, stride, pad, bias, activation_type, activation_params[0], activation_params[1]);
46
    }
47


48
    return ret;
49
}
50


51
static int test_convolution1d_0()
52
{
53
    static const int kdsp[16][4] = {
54
        {1, 1, 1, 0},
55
        {1, 1, 2, 0},
56
        {2, 1, 1, 1},
57
        {2, 1, 2, -233},
58
        {3, 1, 1, 1},
59
        {3, 1, 2, 1},
60
        {3, 2, 1, 1},
61
        {4, 1, 1, 2},
62
        {4, 1, 2, -233},
63
        {4, 2, 1, -234},
64
        {5, 1, 1, -234},
65
        {5, 1, 2, 2},
66
        {5, 2, 2, 2},
67
        {7, 1, 1, 3},
68
        {7, 1, 2, 3},
69
        {7, 2, 1, -233},
70
    };
71


72
    for (int i = 0; i < 16; i++)
73
    {
74
        const int k = kdsp[i][0];
75
        const int d = kdsp[i][1];
76
        const int s = kdsp[i][2];
77
        const int p = kdsp[i][3];
78
        const int b0 = i % 2;
79
        const int b1 = 1 - b0;
80


81
        int ret = 0
82
                  || test_convolution1d(9, 1, 1, k, d, s, p, b0)
83
                  || test_convolution1d(9, 1, 3, k, d, s, p, b1)
84
                  || test_convolution1d(9, 1, 7, k, d, s, p, b0)
85
                  || test_convolution1d(9, 1, 15, k, d, s, p, b1)
86
                  || test_convolution1d(9, 1, 31, k, d, s, p, b0)
87
                  || test_convolution1d(9, 3, 1, k, d, s, p, b1)
88
                  || test_convolution1d(9, 3, 3, k, d, s, p, b0)
89
                  || test_convolution1d(9, 3, 7, k, d, s, p, b1)
90
                  || test_convolution1d(9, 3, 15, k, d, s, p, b0)
91
                  || test_convolution1d(9, 3, 31, k, d, s, p, b1)
92
                  || test_convolution1d(9, 7, 1, k, d, s, p, b0)
93
                  || test_convolution1d(9, 7, 3, k, d, s, p, b1)
94
                  || test_convolution1d(9, 7, 7, k, d, s, p, b0)
95
                  || test_convolution1d(9, 7, 15, k, d, s, p, b1)
96
                  || test_convolution1d(9, 7, 31, k, d, s, p, b0)
97
                  || test_convolution1d(9, 15, 1, k, d, s, p, b1)
98
                  || test_convolution1d(9, 15, 3, k, d, s, p, b0)
99
                  || test_convolution1d(9, 15, 7, k, d, s, p, b1)
100
                  || test_convolution1d(9, 15, 15, k, d, s, p, b0)
101
                  || test_convolution1d(9, 15, 31, k, d, s, p, b1)
102
                  || test_convolution1d(9, 31, 1, k, d, s, p, b0)
103
                  || test_convolution1d(9, 31, 3, k, d, s, p, b1)
104
                  || test_convolution1d(9, 31, 7, k, d, s, p, b0)
105
                  || test_convolution1d(9, 31, 15, k, d, s, p, b1)
106
                  || test_convolution1d(25, 28, 31, k, d, s, p, b0)
107
                  || test_convolution1d(25, 31, 28, k, d, s, p, b1)
108
                  || test_convolution1d(25, 28, 28, k, d, s, p, b0)
109
                  || test_convolution1d(25, 24, 28, k, d, s, p, b1)
110
                  || test_convolution1d(25, 24, 31, k, d, s, p, b0)
111
                  || test_convolution1d(25, 28, 24, k, d, s, p, b1)
112
                  || test_convolution1d(25, 31, 24, k, d, s, p, b0)
113
                  || test_convolution1d(25, 24, 24, k, d, s, p, b1)
114
                  || test_convolution1d(25, 28, 48, k, d, s, p, b0)
115
                  || test_convolution1d(25, 31, 48, k, d, s, p, b1)
116
                  || test_convolution1d(25, 24, 48, k, d, s, p, b0)
117
                  || test_convolution1d(25, 48, 28, k, d, s, p, b1)
118
                  || test_convolution1d(25, 48, 31, k, d, s, p, b0)
119
                  || test_convolution1d(25, 48, 24, k, d, s, p, b1)
120
                  || test_convolution1d(25, 31, 31, k, d, s, p, b0)
121
                  || test_convolution1d(25, 48, 48, k, d, s, p, b1);
122


123
        if (ret != 0)
124
            return -1;
125
    }
126


127
    return 0
128
           || test_convolution1d(7, 1, 4, 3, 1, 1, 1, 1)
129
           || test_convolution1d(14, 1, 4, 3, 1, 2, 1, 1)
130
           || test_convolution1d(15, 4, 4, 3, 1, 1, 1, 1)
131
           || test_convolution1d(15, 8, 8, 3, 1, 1, 1, 1)
132
           || test_convolution1d(11, 8, 16, 3, 1, 1, 1, 1)
133
           || test_convolution1d(13, 16, 24, 3, 1, 1, 1, 1)
134
           || test_convolution1d(8, 16, 24, 3, 1, 1, 1, 0)
135
           || test_convolution1d(4, 16, 24, 3, 1, 1, 1, 1)
136
           || test_convolution1d(4, 16, 24, 3, 1, 1, 1, 0)
137
           || test_convolution1d(6, 64, 64, 3, 1, 2, 0, 1);
138
}
139


140
static int test_convolution1d_dynamic(int w, int h, int outh, int kernel, int dilation, int stride, int pad, int bias)
141
{
142
    ncnn::Mat a = RandomMat(w, h);
143


144
    ncnn::ParamDict pd;
145
    pd.set(0, 0);
146
    pd.set(1, 0);
147
    pd.set(2, dilation);
148
    pd.set(3, stride);
149
    pd.set(4, pad);
150
    pd.set(5, bias);
151
    pd.set(6, 0);
152
    pd.set(19, 1); // dynamic weight
153


154
    int activation_type = RAND() % 7; // 0 1 2 3 4 5 6
155
    ncnn::Mat activation_params(2);
156
    activation_params[0] = (activation_type == 6) ? RandomFloat(0, 1) : RandomFloat(-1, 0); // alpha
157
    activation_params[1] = RandomFloat(0, 1);                                               // beta
158
    pd.set(9, activation_type);
159
    pd.set(10, activation_params);
160


161
    std::vector<ncnn::Mat> as(bias ? 3 : 2);
162
    as[0] = a;
163
    as[1] = RandomMat(kernel, h, outh);
164
    if (bias)
165
        as[2] = RandomMat(outh);
166


167
    std::vector<ncnn::Mat> weights(0);
168


169
    int ret = test_layer("Convolution1D", pd, weights, as);
170
    if (ret != 0)
171
    {
172
        fprintf(stderr, "test_convolution1d_dynamic failed w=%d h=%d outh=%d kernel=%d dilation=%d stride=%d pad=%d bias=%d act=%d actparams=[%f,%f]\n", w, h, outh, kernel, dilation, stride, pad, bias, activation_type, activation_params[0], activation_params[1]);
173
    }
174


175
    return ret;
176
}
177


178
static int test_convolution1d_1()
179
{
180
    static const int kdsp[7][4] = {
181
        {1, 1, 1, 0},
182
        {1, 1, 2, 0},
183
        {2, 1, 1, 1},
184
        {2, 1, 2, -233},
185
        {3, 1, 1, 1},
186
        {3, 1, 2, 1},
187
        {3, 2, 1, -234},
188
    };
189


190
    for (int i = 0; i < 7; i++)
191
    {
192
        const int k = kdsp[i][0];
193
        const int d = kdsp[i][1];
194
        const int s = kdsp[i][2];
195
        const int p = kdsp[i][3];
196


197
        int ret = 0
198
                  || test_convolution1d_dynamic(11, 1, 1, k, d, s, p, 1)
199
                  || test_convolution1d_dynamic(11, 4, 13, k, d, s, p, 0)
200
                  || test_convolution1d_dynamic(11, 13, 4, k, d, s, p, 1)
201
                  || test_convolution1d_dynamic(11, 12, 12, k, d, s, p, 0)
202
                  || test_convolution1d_dynamic(11, 8, 12, k, d, s, p, 1)
203
                  || test_convolution1d_dynamic(11, 8, 13, k, d, s, p, 0)
204
                  || test_convolution1d_dynamic(11, 13, 8, k, d, s, p, 1)
205
                  || test_convolution1d_dynamic(11, 12, 16, k, d, s, p, 0)
206
                  || test_convolution1d_dynamic(11, 15, 15, k, d, s, p, 0)
207
                  || test_convolution1d_dynamic(11, 16, 16, k, d, s, p, 0);
208


209
        if (ret != 0)
210
            return -1;
211
    }
212


213
    return 0;
214
}
215


216
int main()
217
{
218
    SRAND(7767517);
219


220
    return test_convolution1d_0() || test_convolution1d_1();
221
}
222

Использование cookies

Мы используем файлы cookie в соответствии с Политикой конфиденциальности и Политикой использования cookies.

Нажимая кнопку «Принимаю», Вы даете АО «СберТех» согласие на обработку Ваших персональных данных в целях совершенствования нашего веб-сайта и Сервиса GitVerse, а также повышения удобства их использования.

Запретить использование cookies Вы можете самостоятельно в настройках Вашего браузера.