convex_resnet.py

from typing import Any, Callable, List, Optional, Type, Union

import torch
import torch.nn as nn
from torch import Tensor

from torchvision.utils import _log_api_usage_once
from torchvision.models._utils import _ovewrite_named_param

from convex_modules import ConvexConv2d, ConvexLinear, NegExpPositivity

__all__ = [
    "ConvexResNet",
    "convex_resnet18",
    "convex_resnet34",
    "convex_resnet50",
    "convex_resnet101",
    "convex_resnet152",
    "convex_resnext50_32x4d",
    "convex_resnext101_32x8d",
    "convex_resnext101_64x4d",
    "convex_wide_resnet50_2",
    "convex_wide_resnet101_2",
]


def conv3x3(in_planes: int, out_planes: int, stride: int = 1, groups: int = 1, dilation: int = 1) -> nn.Conv2d:
    """3x3 convolution with padding"""
    return ConvexConv2d(
        in_planes,
        out_planes,
        kernel_size=3,
        stride=stride,
        padding=dilation,
        groups=groups,
        bias=False,
        dilation=dilation,
        positivity=NegExpPositivity(),
    )


def conv1x1(in_planes: int, out_planes: int, stride: int = 1) -> nn.Conv2d:
    """1x1 convolution"""
    return ConvexConv2d(in_planes, out_planes, kernel_size=1, stride=stride, bias=False, positivity=NegExpPositivity())


class BasicBlock(nn.Module):
    expansion: int = 1

    def __init__(
        self,
        inplanes: int,
        planes: int,
        stride: int = 1,
        downsample: Optional[nn.Module] = None,
        groups: int = 1,
        base_width: int = 64,
        dilation: int = 1,
        norm_layer: Optional[Callable[..., nn.Module]] = None,
    ) -> None:
        super().__init__()
        if norm_layer is None:
            norm_layer = nn.BatchNorm2d
        if groups != 1 or base_width != 64:
            raise ValueError("BasicBlock only supports groups=1 and base_width=64")
        if dilation > 1:
            raise NotImplementedError("Dilation > 1 not supported in BasicBlock")
        # Both self.conv1 and self.downsample layers downsample the input when stride != 1
        self.conv1 = conv3x3(inplanes, planes, stride)
        self.bn1 = norm_layer(planes)
        self.relu = nn.ReLU(inplace=True)
        self.conv2 = conv3x3(planes, planes)
        self.bn2 = norm_layer(planes)
        self.downsample = downsample
        self.stride = stride

    def forward(self, x: Tensor) -> Tensor:
        identity = x

        out = self.conv1(x)
        out = self.bn1(out)
        out = self.relu(out)

        out = self.conv2(out)
        out = self.bn2(out)

        if self.downsample is not None:
            identity = self.downsample(x)

        out += identity
        out = self.relu(out)

        return out


class Bottleneck(nn.Module):
    # Bottleneck in torchvision places the stride for downsampling at 3x3 convolution(self.conv2)
    # while original implementation places the stride at the first 1x1 convolution(self.conv1)
    # according to "Deep residual learning for image recognition"https://arxiv.org/abs/1512.03385.
    # This variant is also known as ResNet V1.5 and improves accuracy according to
    # https://ngc.nvidia.com/catalog/model-scripts/nvidia:resnet_50_v1_5_for_pytorch.

    expansion: int = 4

    def __init__(
        self,
        inplanes: int,
        planes: int,
        stride: int = 1,
        downsample: Optional[nn.Module] = None,
        groups: int = 1,
        base_width: int = 64,
        dilation: int = 1,
        norm_layer: Optional[Callable[..., nn.Module]] = None,
    ) -> None:
        super().__init__()
        if norm_layer is None:
            norm_layer = nn.BatchNorm2d
        width = int(planes * (base_width / 64.0)) * groups
        # Both self.conv2 and self.downsample layers downsample the input when stride != 1
        self.conv1 = conv1x1(inplanes, width)
        self.bn1 = norm_layer(width)
        self.conv2 = conv3x3(width, width, stride, groups, dilation)
        self.bn2 = norm_layer(width)
        self.conv3 = conv1x1(width, planes * self.expansion)
        self.bn3 = norm_layer(planes * self.expansion)
        self.relu = nn.ReLU(inplace=True)
        self.downsample = downsample
        self.stride = stride

    def forward(self, x: Tensor) -> Tensor:
        identity = x

        out = self.conv1(x)
        out = self.bn1(out)
        out = self.relu(out)

        out = self.conv2(out)
        out = self.bn2(out)
        out = self.relu(out)

        out = self.conv3(out)
        out = self.bn3(out)

        if self.downsample is not None:
            identity = self.downsample(x)

        out += identity
        out = self.relu(out)

        return out


class ConvexResNet(nn.Module):
    def __init__(
        self,
        block: Type[Union[BasicBlock, Bottleneck]],
        layers: List[int],
        num_classes: int = 1000,
        groups: int = 1,
        width_per_group: int = 64,
        replace_stride_with_dilation: Optional[List[bool]] = None,
        norm_layer: Optional[Callable[..., nn.Module]] = None,
    ) -> None:
        super().__init__()
        _log_api_usage_once(self)
        if norm_layer is None:
            norm_layer = nn.BatchNorm2d
        self._norm_layer = norm_layer

        self.inplanes = 64
        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 "
                f"or a 3-element tuple, got {replace_stride_with_dilation}"
            )
        self.groups = groups
        self.base_width = width_per_group
        self.conv1 = ConvexConv2d(3, self.inplanes, kernel_size=7, stride=2, padding=3, bias=False, positivity=NegExpPositivity())
        self.bn1 = norm_layer(self.inplanes)
        self.relu = nn.ReLU(inplace=True)
        self.maxpool = nn.MaxPool2d(kernel_size=3, stride=2, padding=1)
        self.layer1 = self._make_layer(block, 64, layers[0])
        self.layer2 = self._make_layer(block, 128, layers[1], stride=2, dilate=replace_stride_with_dilation[0])
        self.layer3 = self._make_layer(block, 256, layers[2], stride=2, dilate=replace_stride_with_dilation[1])
        self.layer4 = self._make_layer(block, 512, layers[3], stride=2, dilate=replace_stride_with_dilation[2])
        self.avgpool = nn.AdaptiveMaxPool2d((1, 1))
        self.fc = ConvexLinear(512 * block.expansion, num_classes, positivity=NegExpPositivity())

        for m in self.modules():
            if isinstance(m, ConvexConv2d):
                m.reset_parameters()
            elif 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)
                m.weight.requires_grad_(False)
                nn.init.constant_(m.bias, 0)

    def _make_layer(
        self,
        block: Type[Union[BasicBlock, Bottleneck]],
        planes: int,
        blocks: int,
        stride: int = 1,
        dilate: bool = False,
    ) -> nn.Sequential:
        norm_layer = self._norm_layer
        downsample = None
        previous_dilation = self.dilation
        if dilate:
            self.dilation *= stride
            stride = 1
        if stride != 1 or self.inplanes != planes * block.expansion:
            downsample = nn.Sequential(
                conv1x1(self.inplanes, planes * block.expansion, stride),
                norm_layer(planes * block.expansion),
            )

        layers = []
        layers.append(
            block(
                self.inplanes, planes, stride, downsample, self.groups, self.base_width, previous_dilation, norm_layer
            )
        )
        self.inplanes = planes * block.expansion
        for _ in range(1, blocks):
            layers.append(
                block(
                    self.inplanes,
                    planes,
                    groups=self.groups,
                    base_width=self.base_width,
                    dilation=self.dilation,
                    norm_layer=norm_layer,
                )
            )

        return nn.Sequential(*layers)

    def _forward_impl(self, x: Tensor) -> Tensor:
        # See note [TorchScript super()]
        x = self.conv1(x)
        x = self.bn1(x)
        x = self.relu(x)
        x = self.maxpool(x)

        x = self.layer1(x)
        x = self.layer2(x)
        x = self.layer3(x)
        x = self.layer4(x)

        x = self.avgpool(x)
        x = torch.flatten(x, 1)
        x = self.fc(x)

        return x

    def forward(self, x: Tensor) -> Tensor:
        return self._forward_impl(x)


def convex_resnet18(**kwargs: Any) -> ConvexResNet:
    """ResNet-18 from `Deep Residual Learning for Image Recognition <https://arxiv.org/pdf/1512.03385.pdf>`__.
    Args:
        **kwargs: parameters passed to the ``torchvision.models.resnet.ResNet``
            base class. Please refer to the `source code
            <https://github.com/pytorch/vision/blob/main/torchvision/models/resnet.py>`_
            for more details about this class.
    .. autoclass:: torchvision.models.ResNet18_Weights
        :members:
    """
    return ConvexResNet(BasicBlock, [2, 2, 2, 2], **kwargs)


def convex_resnet34(**kwargs: Any) -> ConvexResNet:
    """ResNet-34 from `Deep Residual Learning for Image Recognition <https://arxiv.org/pdf/1512.03385.pdf>`__.
    Args:
        **kwargs: parameters passed to the ``torchvision.models.resnet.ResNet``
            base class. Please refer to the `source code
            <https://github.com/pytorch/vision/blob/main/torchvision/models/resnet.py>`_
            for more details about this class.
    .. autoclass:: torchvision.models.ResNet34_Weights
        :members:
    """
    return ConvexResNet(BasicBlock, [3, 4, 6, 3], **kwargs)


def convex_resnet50(**kwargs: Any) -> ConvexResNet:
    """ResNet-50 from `Deep Residual Learning for Image Recognition <https://arxiv.org/pdf/1512.03385.pdf>`__.
    .. note::
       The bottleneck of TorchVision places the stride for downsampling to the second 3x3
       convolution while the original paper places it to the first 1x1 convolution.
       This variant improves the accuracy and is known as `ResNet V1.5
       <https://ngc.nvidia.com/catalog/model-scripts/nvidia:resnet_50_v1_5_for_pytorch>`_.
    Args:
        **kwargs: parameters passed to the ``torchvision.models.resnet.ResNet``
            base class. Please refer to the `source code
            <https://github.com/pytorch/vision/blob/main/torchvision/models/resnet.py>`_
            for more details about this class.
    .. autoclass:: torchvision.models.ResNet50_Weights
        :members:
    """
    return ConvexResNet(Bottleneck, [3, 4, 6, 3], **kwargs)


def convex_resnet101(**kwargs: Any) -> ConvexResNet:
    """ResNet-101 from `Deep Residual Learning for Image Recognition <https://arxiv.org/pdf/1512.03385.pdf>`__.
    .. note::
       The bottleneck of TorchVision places the stride for downsampling to the second 3x3
       convolution while the original paper places it to the first 1x1 convolution.
       This variant improves the accuracy and is known as `ResNet V1.5
       <https://ngc.nvidia.com/catalog/model-scripts/nvidia:resnet_50_v1_5_for_pytorch>`_.
    Args:
        **kwargs: parameters passed to the ``torchvision.models.resnet.ResNet``
            base class. Please refer to the `source code
            <https://github.com/pytorch/vision/blob/main/torchvision/models/resnet.py>`_
            for more details about this class.
    .. autoclass:: torchvision.models.ResNet101_Weights
        :members:
    """
    return ConvexResNet(Bottleneck, [3, 4, 23, 3], **kwargs)


def convex_resnet152(**kwargs: Any) -> ConvexResNet:
    """ResNet-152 from `Deep Residual Learning for Image Recognition <https://arxiv.org/pdf/1512.03385.pdf>`__.
    .. note::
       The bottleneck of TorchVision places the stride for downsampling to the second 3x3
       convolution while the original paper places it to the first 1x1 convolution.
       This variant improves the accuracy and is known as `ResNet V1.5
       <https://ngc.nvidia.com/catalog/model-scripts/nvidia:resnet_50_v1_5_for_pytorch>`_.
    Args:
        **kwargs: parameters passed to the ``torchvision.models.resnet.ResNet``
            base class. Please refer to the `source code
            <https://github.com/pytorch/vision/blob/main/torchvision/models/resnet.py>`_
            for more details about this class.
    .. autoclass:: torchvision.models.ResNet152_Weights
        :members:
    """
    return ConvexResNet(Bottleneck, [3, 8, 36, 3], **kwargs)


def convex_resnext50_32x4d(**kwargs: Any) -> ConvexResNet:
    """ResNeXt-50 32x4d model from
    `Aggregated Residual Transformation for Deep Neural Networks <https://arxiv.org/abs/1611.05431>`_.
    Args:
        **kwargs: parameters passed to the ``torchvision.models.resnet.ResNet``
            base class. Please refer to the `source code
            <https://github.com/pytorch/vision/blob/main/torchvision/models/resnet.py>`_
            for more details about this class.
    .. autoclass:: torchvision.models.ResNeXt50_32X4D_Weights
        :members:
    """
    _ovewrite_named_param(kwargs, "groups", 32)
    _ovewrite_named_param(kwargs, "width_per_group", 4)
    return ConvexResNet(Bottleneck, [3, 4, 6, 3], **kwargs)


def convex_resnext101_32x8d(**kwargs: Any) -> ConvexResNet:
    """ResNeXt-101 32x8d model from
    `Aggregated Residual Transformation for Deep Neural Networks <https://arxiv.org/abs/1611.05431>`_.
    Args:
        **kwargs: parameters passed to the ``torchvision.models.resnet.ResNet``
            base class. Please refer to the `source code
            <https://github.com/pytorch/vision/blob/main/torchvision/models/resnet.py>`_
            for more details about this class.
    .. autoclass:: torchvision.models.ResNeXt101_32X8D_Weights
        :members:
    """
    _ovewrite_named_param(kwargs, "groups", 32)
    _ovewrite_named_param(kwargs, "width_per_group", 8)
    return ConvexResNet(Bottleneck, [3, 4, 23, 3], **kwargs)


def convex_resnext101_64x4d(**kwargs: Any) -> ConvexResNet:
    """ResNeXt-101 64x4d model from
    `Aggregated Residual Transformation for Deep Neural Networks <https://arxiv.org/abs/1611.05431>`_.
    Args:
        **kwargs: parameters passed to the ``torchvision.models.resnet.ResNet``
            base class. Please refer to the `source code
            <https://github.com/pytorch/vision/blob/main/torchvision/models/resnet.py>`_
            for more details about this class.
    .. autoclass:: torchvision.models.ResNeXt101_64X4D_Weights
        :members:
    """
    _ovewrite_named_param(kwargs, "groups", 64)
    _ovewrite_named_param(kwargs, "width_per_group", 4)
    return ConvexResNet(Bottleneck, [3, 4, 23, 3], **kwargs)


def convex_wide_resnet50_2(**kwargs: Any) -> ConvexResNet:
    """Wide ResNet-50-2 model from
    `Wide Residual Networks <https://arxiv.org/abs/1605.07146>`_.
    The model is the same as ResNet except for the bottleneck number of channels
    which is twice larger in every block. The number of channels in outer 1x1
    convolutions is the same, e.g. last block in ResNet-50 has 2048-512-2048
    channels, and in Wide ResNet-50-2 has 2048-1024-2048.
    Args:
        **kwargs: parameters passed to the ``torchvision.models.resnet.ResNet``
            base class. Please refer to the `source code
            <https://github.com/pytorch/vision/blob/main/torchvision/models/resnet.py>`_
            for more details about this class.
    .. autoclass:: torchvision.models.Wide_ResNet50_2_Weights
        :members:
    """
    _ovewrite_named_param(kwargs, "width_per_group", 64 * 2)
    return ConvexResNet(Bottleneck, [3, 4, 6, 3], **kwargs)


def convex_wide_resnet101_2(**kwargs: Any) -> ConvexResNet:
    """Wide ResNet-101-2 model from
    `Wide Residual Networks <https://arxiv.org/abs/1605.07146>`_.
    The model is the same as ResNet except for the bottleneck number of channels
    which is twice larger in every block. The number of channels in outer 1x1
    convolutions is the same, e.g. last block in ResNet-101 has 2048-512-2048
    channels, and in Wide ResNet-101-2 has 2048-1024-2048.
    Args:
        **kwargs: parameters passed to the ``torchvision.models.resnet.ResNet``
            base class. Please refer to the `source code
            <https://github.com/pytorch/vision/blob/main/torchvision/models/resnet.py>`_
            for more details about this class.
    .. autoclass:: torchvision.models.Wide_ResNet101_2_Weights
        :members:
    """
    _ovewrite_named_param(kwargs, "width_per_group", 64 * 2)
    return ConvexResNet(Bottleneck, [3, 4, 23, 3], **kwargs)