src/controlnet_aux/mlsd/models/mbv2_mlsd_large.py [28:178]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
             b = F.interpolate(b, scale_factor=2.0, mode='bilinear', align_corners=True)
        return torch.cat((a, b), dim=1)


class BlockTypeB(nn.Module):
    def __init__(self, in_c, out_c):
        super(BlockTypeB, self).__init__()
        self.conv1 = nn.Sequential(
            nn.Conv2d(in_c, in_c,  kernel_size=3, padding=1),
            nn.BatchNorm2d(in_c),
            nn.ReLU()
        )
        self.conv2 = nn.Sequential(
            nn.Conv2d(in_c, out_c, kernel_size=3, padding=1),
            nn.BatchNorm2d(out_c),
            nn.ReLU()
        )

    def forward(self, x):
        x = self.conv1(x) + x
        x = self.conv2(x)
        return x

class BlockTypeC(nn.Module):
    def __init__(self, in_c, out_c):
        super(BlockTypeC, self).__init__()
        self.conv1 = nn.Sequential(
            nn.Conv2d(in_c, in_c,  kernel_size=3, padding=5, dilation=5),
            nn.BatchNorm2d(in_c),
            nn.ReLU()
        )
        self.conv2 = nn.Sequential(
            nn.Conv2d(in_c, in_c,  kernel_size=3, padding=1),
            nn.BatchNorm2d(in_c),
            nn.ReLU()
        )
        self.conv3 = nn.Conv2d(in_c, out_c, kernel_size=1)

    def forward(self, x):
        x = self.conv1(x)
        x = self.conv2(x)
        x = self.conv3(x)
        return x

def _make_divisible(v, divisor, min_value=None):
    """
    This function is taken from the original tf repo.
    It ensures that all layers have a channel number that is divisible by 8
    It can be seen here:
    https://github.com/tensorflow/models/blob/master/research/slim/nets/mobilenet/mobilenet.py
    :param v:
    :param divisor:
    :param min_value:
    :return:
    """
    if min_value is None:
        min_value = divisor
    new_v = max(min_value, int(v + divisor / 2) // divisor * divisor)
    # Make sure that round down does not go down by more than 10%.
    if new_v < 0.9 * v:
        new_v += divisor
    return new_v


class ConvBNReLU(nn.Sequential):
    def __init__(self, in_planes, out_planes, kernel_size=3, stride=1, groups=1):
        self.channel_pad = out_planes - in_planes
        self.stride = stride
        #padding = (kernel_size - 1) // 2

        # TFLite uses slightly different padding than PyTorch
        if stride == 2:
            padding = 0
        else:
            padding = (kernel_size - 1) // 2

        super(ConvBNReLU, self).__init__(
            nn.Conv2d(in_planes, out_planes, kernel_size, stride, padding, groups=groups, bias=False),
            nn.BatchNorm2d(out_planes),
            nn.ReLU6(inplace=True)
        )
        self.max_pool = nn.MaxPool2d(kernel_size=stride, stride=stride)


    def forward(self, x):
        # TFLite uses  different padding
        if self.stride == 2:
            x = F.pad(x, (0, 1, 0, 1), "constant", 0)
            #print(x.shape)

        for module in self:
            if not isinstance(module, nn.MaxPool2d):
                x = module(x)
        return x


class InvertedResidual(nn.Module):
    def __init__(self, inp, oup, stride, expand_ratio):
        super(InvertedResidual, self).__init__()
        self.stride = stride
        assert stride in [1, 2]

        hidden_dim = int(round(inp * expand_ratio))
        self.use_res_connect = self.stride == 1 and inp == oup

        layers = []
        if expand_ratio != 1:
            # pw
            layers.append(ConvBNReLU(inp, hidden_dim, kernel_size=1))
        layers.extend([
            # dw
            ConvBNReLU(hidden_dim, hidden_dim, stride=stride, groups=hidden_dim),
            # pw-linear
            nn.Conv2d(hidden_dim, oup, 1, 1, 0, bias=False),
            nn.BatchNorm2d(oup),
        ])
        self.conv = nn.Sequential(*layers)

    def forward(self, x):
        if self.use_res_connect:
            return x + self.conv(x)
        else:
            return self.conv(x)


class MobileNetV2(nn.Module):
    def __init__(self, pretrained=True):
        """
        MobileNet V2 main class
        Args:
            num_classes (int): Number of classes
            width_mult (float): Width multiplier - adjusts number of channels in each layer by this amount
            inverted_residual_setting: Network structure
            round_nearest (int): Round the number of channels in each layer to be a multiple of this number
            Set to 1 to turn off rounding
            block: Module specifying inverted residual building block for mobilenet
        """
        super(MobileNetV2, self).__init__()

        block = InvertedResidual
        input_channel = 32
        last_channel = 1280
        width_mult = 1.0
        round_nearest = 8

        inverted_residual_setting = [
            # t, c, n, s
            [1, 16, 1, 1],
            [6, 24, 2, 2],
            [6, 32, 3, 2],
            [6, 64, 4, 2],
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -



src/controlnet_aux/mlsd/models/mbv2_mlsd_tiny.py [27:177]:
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
        b = F.interpolate(b, scale_factor=2.0, mode='bilinear', align_corners=True)
        return torch.cat((a, b), dim=1)


class BlockTypeB(nn.Module):
    def __init__(self, in_c, out_c):
        super(BlockTypeB, self).__init__()
        self.conv1 = nn.Sequential(
            nn.Conv2d(in_c, in_c,  kernel_size=3, padding=1),
            nn.BatchNorm2d(in_c),
            nn.ReLU()
        )
        self.conv2 = nn.Sequential(
            nn.Conv2d(in_c, out_c, kernel_size=3, padding=1),
            nn.BatchNorm2d(out_c),
            nn.ReLU()
        )

    def forward(self, x):
        x = self.conv1(x) + x
        x = self.conv2(x)
        return x

class BlockTypeC(nn.Module):
    def __init__(self, in_c, out_c):
        super(BlockTypeC, self).__init__()
        self.conv1 = nn.Sequential(
            nn.Conv2d(in_c, in_c,  kernel_size=3, padding=5, dilation=5),
            nn.BatchNorm2d(in_c),
            nn.ReLU()
        )
        self.conv2 = nn.Sequential(
            nn.Conv2d(in_c, in_c,  kernel_size=3, padding=1),
            nn.BatchNorm2d(in_c),
            nn.ReLU()
        )
        self.conv3 = nn.Conv2d(in_c, out_c, kernel_size=1)

    def forward(self, x):
        x = self.conv1(x)
        x = self.conv2(x)
        x = self.conv3(x)
        return x

def _make_divisible(v, divisor, min_value=None):
    """
    This function is taken from the original tf repo.
    It ensures that all layers have a channel number that is divisible by 8
    It can be seen here:
    https://github.com/tensorflow/models/blob/master/research/slim/nets/mobilenet/mobilenet.py
    :param v:
    :param divisor:
    :param min_value:
    :return:
    """
    if min_value is None:
        min_value = divisor
    new_v = max(min_value, int(v + divisor / 2) // divisor * divisor)
    # Make sure that round down does not go down by more than 10%.
    if new_v < 0.9 * v:
        new_v += divisor
    return new_v


class ConvBNReLU(nn.Sequential):
    def __init__(self, in_planes, out_planes, kernel_size=3, stride=1, groups=1):
        self.channel_pad = out_planes - in_planes
        self.stride = stride
        #padding = (kernel_size - 1) // 2

        # TFLite uses slightly different padding than PyTorch
        if stride == 2:
            padding = 0
        else:
            padding = (kernel_size - 1) // 2

        super(ConvBNReLU, self).__init__(
            nn.Conv2d(in_planes, out_planes, kernel_size, stride, padding, groups=groups, bias=False),
            nn.BatchNorm2d(out_planes),
            nn.ReLU6(inplace=True)
        )
        self.max_pool = nn.MaxPool2d(kernel_size=stride, stride=stride)


    def forward(self, x):
        # TFLite uses  different padding
        if self.stride == 2:
            x = F.pad(x, (0, 1, 0, 1), "constant", 0)
            #print(x.shape)

        for module in self:
            if not isinstance(module, nn.MaxPool2d):
                x = module(x)
        return x


class InvertedResidual(nn.Module):
    def __init__(self, inp, oup, stride, expand_ratio):
        super(InvertedResidual, self).__init__()
        self.stride = stride
        assert stride in [1, 2]

        hidden_dim = int(round(inp * expand_ratio))
        self.use_res_connect = self.stride == 1 and inp == oup

        layers = []
        if expand_ratio != 1:
            # pw
            layers.append(ConvBNReLU(inp, hidden_dim, kernel_size=1))
        layers.extend([
            # dw
            ConvBNReLU(hidden_dim, hidden_dim, stride=stride, groups=hidden_dim),
            # pw-linear
            nn.Conv2d(hidden_dim, oup, 1, 1, 0, bias=False),
            nn.BatchNorm2d(oup),
        ])
        self.conv = nn.Sequential(*layers)

    def forward(self, x):
        if self.use_res_connect:
            return x + self.conv(x)
        else:
            return self.conv(x)


class MobileNetV2(nn.Module):
    def __init__(self, pretrained=True):
        """
        MobileNet V2 main class
        Args:
            num_classes (int): Number of classes
            width_mult (float): Width multiplier - adjusts number of channels in each layer by this amount
            inverted_residual_setting: Network structure
            round_nearest (int): Round the number of channels in each layer to be a multiple of this number
            Set to 1 to turn off rounding
            block: Module specifying inverted residual building block for mobilenet
        """
        super(MobileNetV2, self).__init__()

        block = InvertedResidual
        input_channel = 32
        last_channel = 1280
        width_mult = 1.0
        round_nearest = 8

        inverted_residual_setting = [
            # t, c, n, s
            [1, 16, 1, 1],
            [6, 24, 2, 2],
            [6, 32, 3, 2],
            [6, 64, 4, 2],
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -