def __init__()

in ss_baselines/savi/models/smt_resnet.py [0:0]

• 40 lines of code
• 13 McCabe index (conditional complexity)

    def __init__(self, block, layers, num_input_channels=3, num_classes=64,
                 zero_init_residual=False, groups=16, width_per_group=16,
                 replace_stride_with_dilation=None, norm_layer=None):
        super(CustomResNet, self).__init__()
        if norm_layer is None:
            norm_layer = nn.GroupNorm
        self._norm_layer = norm_layer

        self.inplanes = 16
        self.dilation = 1
        if replace_stride_with_dilation is None:
            # each element in the tuple indicates if we should replace
            # the 2x2 stride with a dilated convolution instead
            replace_stride_with_dilation = [False, False, False]
        if len(replace_stride_with_dilation) != 3:
            raise ValueError("replace_stride_with_dilation should be None "
                             "or a 3-element tuple, got {}".format(replace_stride_with_dilation))
        self.groups = groups
        self.base_width = width_per_group
        self.conv1 = nn.Conv2d(num_input_channels, self.inplanes, kernel_size=7,
                               stride=1, padding=3, bias=False)
        self.bn1 = norm_layer(groups, self.inplanes)
        self.relu = nn.ReLU(inplace=True)
        self.layer1 = self._make_layer(block, groups, 16, layers[0])
        self.layer2 = self._make_layer(block, groups, 32, layers[1], stride=2,
                                       dilate=replace_stride_with_dilation[0])
        self.layer3 = self._make_layer(block, groups, 64, layers[2], stride=2,
                                       dilate=replace_stride_with_dilation[1])
        self.layer4 = self._make_layer(block, groups, 128, layers[3], stride=2,
                                       dilate=replace_stride_with_dilation[2])
        # Assumes that the input is a 64x64 image
        self.fc = nn.Linear(128 * block.expansion * 8 * 8, num_classes)

        for m in self.modules():
            if isinstance(m, nn.Conv2d):
                nn.init.kaiming_normal_(m.weight, mode='fan_out', nonlinearity='relu')
            elif isinstance(m, (nn.BatchNorm2d, nn.GroupNorm)):
                nn.init.constant_(m.weight, 1)
                nn.init.constant_(m.bias, 0)

        # Zero-initialize the last BN in each residual branch,
        # so that the residual branch starts with zeros, and each residual block behaves like an identity.
        # This improves the model by 0.2~0.3% according to https://arxiv.org/abs/1706.02677
        if zero_init_residual:
            for m in self.modules():
                if isinstance(m, Bottleneck):
                    nn.init.constant_(m.bn3.weight, 0)
                elif isinstance(m, BasicBlock):
                    nn.init.constant_(m.bn2.weight, 0)