• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1# Copyright 2020 Huawei Technologies Co., Ltd
2#
3# Licensed under the Apache License, Version 2.0 (the "License");
4# you may not use this file except in compliance with the License.
5# You may obtain a copy of the License at
6#
7# http://www.apache.org/licenses/LICENSE-2.0
8#
9# Unless required by applicable law or agreed to in writing, software
10# distributed under the License is distributed on an "AS IS" BASIS,
11# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
12# See the License for the specific language governing permissions and
13# limitations under the License.
14# ============================================================================
15""" test_initializer_fuzz """
16import pytest
17
18import mindspore.nn as nn
19from mindspore import Model
20
21
22class Net(nn.Cell):
23    """ Net definition """
24
25    def __init__(self, in_str):
26        a, b, c, d, e, f, g, h = in_str.strip().split()
27        a = int(a)
28        b = int(b)
29        c = int(b)
30        d = int(b)
31        e = int(b)
32        f = int(b)
33        g = int(b)
34        h = int(b)
35
36        super(Net, self).__init__()
37        self.conv = nn.Conv2d(a, b, c, pad_mode="valid")
38        self.bn = nn.BatchNorm2d(d)
39        self.relu = nn.ReLU()
40        self.flatten = nn.Flatten()
41        self.fc = nn.Dense(e * f * g, h)
42
43    def construct(self, x):
44        x = self.conv(x)
45        x = self.bn(x)
46        x = self.relu(x)
47        x = self.flatten(x)
48        out = self.fc(x)
49        return out
50
51
52class LeNet5(nn.Cell):
53    """ LeNet5 definition """
54
55    def __init__(self, in_str):
56        super(LeNet5, self).__init__()
57
58        a1, a2, a3, a4, a5, a6, a7, a8, a9, a10, a11, a12, a13, a14, a15 = in_str.strip().split()
59        a1 = int(a1)
60        a2 = int(a2)
61        a3 = int(a3)
62        a4 = int(a4)
63        a5 = int(a5)
64        a6 = int(a6)
65        a7 = int(a7)
66        a8 = int(a8)
67        a9 = int(a9)
68        a10 = int(a10)
69        a11 = int(a11)
70        a12 = int(a12)
71        a13 = int(a13)
72        a14 = int(a14)
73        a15 = int(a15)
74
75        self.conv1 = nn.Conv2d(a1, a2, a3, pad_mode="valid")
76        self.conv2 = nn.Conv2d(a4, a5, a6, pad_mode="valid")
77        self.fc1 = nn.Dense(a7 * a8 * a9, a10)
78        self.fc2 = nn.Dense(a11, a12)
79        self.fc3 = nn.Dense(a13, a14)
80        self.relu = nn.ReLU()
81        self.max_pool2d = nn.MaxPool2d(kernel_size=a15)
82        self.flatten = nn.Flatten()
83
84    def construct(self, x):
85        x = self.max_pool2d(self.relu(self.conv1(x)))
86        x = self.max_pool2d(self.relu(self.conv2(x)))
87        x = self.flatten(x)
88        x = self.relu(self.fc1(x))
89        x = self.relu(self.fc2(x))
90        x = self.fc3(x)
91        return x
92
93
94def test_shape_error():
95    """ for fuzz test"""
96    in_str = "3 6 5 6 -6 5 16 5 5 120 120 84 84 3 2"
97    with pytest.raises(ValueError):
98        net = LeNet5(in_str)  # neural network
99        Model(net)
100