def __init__()

in utils/gluon/utils/mobilenetv2.py [0:0]

• 24 lines of code
• 9 McCabe index (conditional complexity)

    def __init__(self, in_planes, mid_planes, out_planes, strides=1,
                 norm_kwargs=None, last_gamma=False, name_prefix=None,
                 **kwargs):
        super(_BottleneckV1, self).__init__(prefix=name_prefix)

        self.use_shortcut = strides == 1 and in_planes == out_planes

        with self.name_scope():
            num_group = sum((c if c > 0 else 0 for c in mid_planes))

            # extract information
            self.conv1 = nn.Conv2D(channels=mid_planes, in_channels=in_planes,
                                  kernel_size=1, use_bias=False, prefix='conv1')
            self.bn1 = nn.BatchNorm(in_channels=mid_planes, prefix='bn1',
                                  **({} if norm_kwargs is None else norm_kwargs))
            self.relu1 = _op_act('relu6')
            # capture spatial relations
            self.conv2 = nn.Conv2D(channels=mid_planes, in_channels=mid_planes,
                                  kernel_size=3, padding=1, groups=num_group,
                                  strides=strides, use_bias=False, prefix='conv2')
            self.bn2 = nn.BatchNorm(in_channels=mid_planes, prefix='bn2',
                                  **({} if norm_kwargs is None else norm_kwargs))
            self.relu2 = _op_act('relu6')
            # embeding back to information highway
            self.conv3 = nn.Conv2D(channels=out_planes, in_channels=mid_planes,
                                  kernel_size=1, use_bias=False, prefix='conv3')
            self.bn3 = nn.BatchNorm(in_channels=out_planes, prefix='bn3',
                                  gamma_initializer='zeros' if (last_gamma and \
                                  self.use_shortcut) else 'ones',
                                  **({} if norm_kwargs is None else norm_kwargs))