shufflenetv1 pytorch代码

以下是ShuffleNetV1的PyTorch代码实现，包括ShuffleNetV1的网络结构和训练代码：

import torch.nn as nn
import math

__all__ = ['ShuffleNetV1', 'shufflenetv1']

def conv3x3(in_planes, out_planes, stride=1):
    """3x3 convolution with padding"""
    return nn.Conv2d(in_planes, out_planes, kernel_size=3, stride=stride,
                     padding=1, bias=False)

class BasicBlock(nn.Module):
    def __init__(self, inplanes, planes, stride=1, downsample=None):
        super(BasicBlock, self).__init__()
        self.conv1 = conv3x3(inplanes, planes, stride)
        self.bn1 = nn.BatchNorm2d(planes)
        self.relu = nn.ReLU(inplace=True)
        self.conv2 = conv3x3(planes, planes)
        self.bn2 = nn.BatchNorm2d(planes)
        self.downsample = downsample
        self.stride = stride

    def forward(self, x):
        residual = 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:
            residual = self.downsample(x)

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

        return out

class ShuffleNetV1(nn.Module):
    def __init__(self, num_classes=1000, groups=3, width_mult=1):
        super(ShuffleNetV1, self).__init__()
        self.groups = groups
        self.stage_repeats = [3, 7, 3]
        if groups == 1:
            self.stage_out_channels = [-1, 24, 144, 288, 576]
        elif groups == 2:
            self.stage_out_channels = [-1, 24, 200, 400, 800]
        elif groups == 3:
            self.stage_out_channels = [-1, 24, 240, 480, 960]
        elif groups == 4:
            self.stage_out_channels = [-1, 24, 272, 544, 1088]
        elif groups == 8:
            self.stage_out_channels = [-1, 24, 384, 768, 1536]
        else:
            raise ValueError("""{} groups is not supported for
                1x1 Grouped Convolutions""".format(num_groups))

        # building first layer
        input_channels = 3
        output_channels = self.stage_out_channels[1]
        output_channels = int(output_channels * width_mult)
        self.conv1 = nn.Conv2d(input_channels, output_channels, kernel_size=3, stride=2, padding=1, bias=False)
        self.bn1 = nn.BatchNorm2d(output_channels)
        self.relu = nn.ReLU(inplace=True)
        self.maxpool = nn.MaxPool2d(kernel_size=3, stride=2, padding=1)

        # building stages
        self.stage2 = self._make_stage(2, width_mult)
        self.stage3 = self._make_stage(3, width_mult)
        self.stage4 = self._make_stage(4, width_mult)

        # building last several layers
        self.conv_last = nn.Conv2d(self.stage_out_channels[-2], self.stage_out_channels[-1], kernel_size=1, stride=1, padding=0, bias=False)
        self.globalpool = nn.AvgPool2d(7)
        self.fc = nn.Linear(self.stage_out_channels[-1], num_classes)

    def _make_stage(self, stage, width_mult):
        modules = OrderedDict()
        stage_name = "ShuffleUnit_Stage{}".format(stage)
        # stage_repeats = self.stage_repeats[stage]
        unit1 = ShuffleUnit(self.stage_out_channels[stage-1], self.stage_out_channels[stage], 2, groups=self.groups, width_mult=width_mult)
        modules[stage_name+"_unit1"] = unit1
        for i in range(self.stage_repeats[stage-2]):
            name = stage_name + "_unit" + str(i+2)
            module = ShuffleUnit(self.stage_out_channels[stage], self.stage_out_channels[stage], 1, groups=self.groups, width_mult=width_mult)
            modules[name] = module

        return nn.Sequential(modules)

    def forward(self, x):
        x = self.conv1(x)
        x = self.bn1(x)
        x = self.relu(x)
        x = self.maxpool(x)

        x = self.stage2(x)
        x = self.stage3(x)
        x = self.stage4(x)

        x = self.conv_last(x)
        x = self.globalpool(x)
        x = x.view(-1, self.stage_out_channels[-1])
        x = self.fc(x)

        return x

class ShuffleUnit(nn.Module):
    def __init__(self, in_channels, out_channels, stride, groups=3, width_mult=1):
        super(ShuffleUnit, self).__init__()

        if stride != 1 or in_channels != out_channels:
            self.use_res_connect = False
            self.shortcut = nn.Sequential(
                nn.Conv2d(in_channels, in_channels, kernel_size=3, stride=stride, padding=1, groups=in_channels, bias=False),
                nn.BatchNorm2d(in_channels),
                nn.Conv2d(in_channels, out_channels, kernel_size=1, stride=1, padding=0, bias=False),
                nn.BatchNorm2d(out_channels),
                nn.ReLU(inplace=True),
            )
        else:
            self.use_res_connect = True
            self.shortcut = nn.Sequential()

        self.groups = groups
        mid_channels = int(out_channels / 4 * width_mult)
        self.conv1 = nn.Conv2d(in_channels, mid_channels, kernel_size=1, stride=1, padding=0, bias=False)
        self.bn1 = nn.BatchNorm2d(mid_channels)
        self.conv2 = nn.Conv2d(mid_channels, mid_channels, kernel_size=3, stride=stride, padding=1, groups=groups, bias=False)
        self.bn2 = nn.BatchNorm2d(mid_channels)
        self.conv3 = nn.Conv2d(mid_channels, out_channels, kernel_size=1, stride=1, padding=0, bias=False)
        self.bn3 = nn.BatchNorm2d(out_channels)
        self.relu = nn.ReLU(inplace=True)

    def _channel_shuffle(self, x, groups):
        batchsize, num_channels, height, width = x.data.size()
        channels_per_group = num_channels // groups

        x = x.view(batchsize, groups, channels_per_group, height, width)
        x = torch.transpose(x, 1, 2).contiguous()
        x = x.view(batchsize, -1, height, width)

        return x

    def forward(self, x):
        if self.use_res_connect:
            shortcut = x

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

        x = self.conv2(x)
        x = self.bn2(x)
        x = self.relu(x)
        x = self._channel_shuffle(x, self.groups)

        x = self.conv3(x)
        x = self.bn3(x)

        if self.use_res_connect:
            shortcut = self.shortcut(shortcut)

        x += shortcut
        x = self.relu(x)

        return x

def shufflenetv1(**kwargs):
    """
    Constructs a ShuffleNetV1 model
    """
    return ShuffleNetV1(**kwargs)

在使用ShuffleNetV1时，可以通过以下方式进行实例化：

import torch
from shufflenetv1 import shufflenetv1

model = shufflenetv1(groups=3, width_mult=1)
input = torch.randn(1, 3, 224, 224)
output = model(input)
print(output.shape)