ZXL-Vague-CGAN.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [],
   "source": [
    "from __future__ import print_function\n",
    "import torchvision\n",
    "import argparse\n",
    "import os\n",
    "import random\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.parallel\n",
    "import torch.backends.cudnn as cudnn\n",
    "from torch.autograd import Variable\n",
    "import torch.optim as optim\n",
    "import torch.utils.data\n",
    "import torchvision.datasets as dset\n",
    "import torchvision.transforms as transforms\n",
    "import torchvision.utils as vutils\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.animation as animation\n",
    "from IPython.display import HTML\n",
    "###############\n",
    "import torchvision.models as models\n",
    "import imageio\n",
    "from torch import autograd\n",
    "from torchvision.utils import make_grid\n",
    "import imageio\n",
    "from IPython.display import HTML\n",
    "import time\n",
    "import math\n",
    "manualseed = 47\n",
    "random.seed(manualseed)\n",
    "torch.manual_seed(manualseed)\n",
    "import torch.nn.functional as F\n",
    "###############\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "2 Define variables"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 72,
   "metadata": {},
   "outputs": [],
   "source": [
    "workers = 2 # 数据加载器能够使用的进程数量\n",
    "batch_size = 64  # 训练时的批大小\n",
    "\n",
    "# 训练图片的大小，所有的图片都将改变到该大小\n",
    "image_size = 32  # 转换器使用的大小.\n",
    "nc = 1  # 训练图片的通道数，彩色图片是3\n",
    "nz = 100  # 本征向量z的大小(生成器的输入大小)\n",
    "ngf = 32  # 生成器中特征图个数大小\n",
    "ndf = 64  # 判别器中特征图个数大小\n",
    "num_epochs = 3\n",
    "lr = 0.0002  # 优化器学习率\n",
    "beta1 = 0.5  # Adam优化器的Beta1超参\n",
    "ngpu = 1  # 可利用的GPU数量，使用0将运行在CPU模式。\n",
    "device = torch.device(\"cuda:0\" if (torch.cuda.is_available() and ngpu > 0) else \"cpu\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "3 创建数据集"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "################这里可不要，修改后放到下面了#################\n",
    "# 创建数据集\n",
    "dataset = torchvision.datasets.MNIST(root='C:\\\\Users\\\\lz172\\\\Desktop\\\\jupyter\\\\defense-fl-gan-attack-main\\MNIST_data', train=True, transform=transforms.Compose([transforms.Resize(image_size),\n",
    "                               transforms.CenterCrop(image_size),\n",
    "                               transforms.ToTensor(),\n",
    "                               transforms.Normalize((0.5,), (0.5,)),\n",
    "                           ]), download=False)\n",
    "# 创建数据加载器\n",
    "dataloader = torch.utils.data.DataLoader(dataset=dataset, batch_size=batch_size, shuffle=True)  #600*100*([[28*28],x])\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x14e17c60400>"
      ]
     },
     "execution_count": 74,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 576x576 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "real_batch = next(iter(dataloader))\n",
    "plt.figure(figsize=(8,8))\n",
    "plt.axis(\"off\")\n",
    "plt.title(\"Training Images\")\n",
    "plt.imshow(np.transpose(vutils.make_grid(real_batch[0].to(device)[:64], padding=2, normalize=True).cpu(),(1,2,0)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "def weights_init(m):\n",
    "    classname = m.__class__.__name__\n",
    "    if classname.find('Conv') != -1:\n",
    "        nn.init.normal_(m.weight.data, 0.0, 0.02)\n",
    "    elif classname.find('BatchNorm') != -1:\n",
    "        nn.init.normal_(m.weight.data, 1.0, 0.02)\n",
    "        nn.init.constant_(m.bias.data, 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Generator(nn.Module):\n",
    "    def __init__(self, ngpu):\n",
    "        self.ngpu = ngpu\n",
    "        super(Generator,self).__init__()\n",
    "        self.gen=nn.Sequential(\n",
    "            # 输入是 Z, 对Z进行卷积\n",
    "            nn.ConvTranspose2d(110, ngf*8, 4, 1, 0, bias=False),\n",
    "            nn.BatchNorm2d(ngf*8),\n",
    "            nn.ReLU(inplace=True),\n",
    "            # 输入特征图大小. (ngf*8) x 4 x 4\n",
    "            nn.ConvTranspose2d(ngf*8, ngf*4, 4, 2, 1, bias=False),\n",
    "            nn.BatchNorm2d(ngf*4),\n",
    "            nn.ReLU(inplace=True),\n",
    "            # 输入特征图大小. (ngf*4) x 8 x 8\n",
    "            nn.ConvTranspose2d(ngf*4, ngf*2, 4, 2, 1, bias=False),\n",
    "            nn.BatchNorm2d(ngf * 2),\n",
    "            nn.ReLU(inplace=True),\n",
    "            # 输入特征图大小. (ngf*2) x 16 x 16\n",
    "            nn.ConvTranspose2d(ngf*2, 1, 4, 2, 1, bias=False),\n",
    "            nn.Tanh()\n",
    "            # 输入特征图大小. (nc) x 32 x 32\n",
    "        )\n",
    "    def forward(self, x):\n",
    "        x=self.gen(x)\n",
    "        return x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 77,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Generator(\n",
      "  (gen): Sequential(\n",
      "    (0): ConvTranspose2d(110, 256, kernel_size=(4, 4), stride=(1, 1), bias=False)\n",
      "    (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "    (2): ReLU(inplace=True)\n",
      "    (3): ConvTranspose2d(256, 128, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "    (4): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "    (5): ReLU(inplace=True)\n",
      "    (6): ConvTranspose2d(128, 64, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "    (7): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "    (8): ReLU(inplace=True)\n",
      "    (9): ConvTranspose2d(64, 1, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "    (10): Tanh()\n",
      "  )\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "\n",
    "# 创建生成器\n",
    "netG = Generator(ngpu).to(device)\n",
    "\n",
    "# 如果期望使用多个GPU，设置一下。\n",
    "if (device.type == 'cuda') and (ngpu > 1):\n",
    "    netG = nn.DataParallel(netG, list(range(ngpu)))\n",
    "\n",
    "# 使用权重初始化函数 weights_init 去随机初始化所有权重\n",
    "#  mean=0, stdev=0.2.\n",
    "netG.apply(weights_init)\n",
    "\n",
    "# 输出该模型\n",
    "print(netG)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Discriminator(nn.Module):\n",
    "    def __init__(self, ngpu):\n",
    "        super(Discriminator, self).__init__()\n",
    "        self.ngpu = ngpu\n",
    "        self.main = nn.Sequential(\n",
    "            # state size. (ndf) x 32 x 32\n",
    "            nn.Conv2d(64, ngf, 4, 2, 1, bias=False),  #这里源码的输入1通道需要改成3通道\n",
    "            nn.LeakyReLU(0.2, inplace=True),\n",
    "            # 输入大小. (ndf*2) x 16 x 16\n",
    "            nn.Conv2d(ngf, ngf * 2, 4, 2, 1, bias=False),\n",
    "            nn.BatchNorm2d(ngf * 2),\n",
    "            nn.LeakyReLU(0.2, inplace=True),\n",
    "            # 输入大小. (ndf*2) x 8 x 8\n",
    "            nn.Conv2d(ngf * 2, ngf * 4, 4, 2, 1, bias=False),\n",
    "            nn.BatchNorm2d(ngf * 4),\n",
    "            nn.LeakyReLU(0.2, inplace=True),\n",
    "            # 输入大小. (ndf*4) x 4 x 4\n",
    "            nn.Conv2d(ngf * 4, 11, 3, 1, 0, bias=False),\n",
    "            nn.Sigmoid()\n",
    "        )\n",
    "\n",
    "    def forward(self, x):\n",
    "        return self.main(x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Discriminator(\n",
      "  (main): Sequential(\n",
      "    (0): Conv2d(64, 32, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "    (1): LeakyReLU(negative_slope=0.2, inplace=True)\n",
      "    (2): Conv2d(32, 64, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "    (3): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "    (4): LeakyReLU(negative_slope=0.2, inplace=True)\n",
      "    (5): Conv2d(64, 128, kernel_size=(4, 4), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "    (6): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "    (7): LeakyReLU(negative_slope=0.2, inplace=True)\n",
      "    (8): Conv2d(128, 11, kernel_size=(3, 3), stride=(1, 1), bias=False)\n",
      "    (9): Sigmoid()\n",
      "  )\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "\n",
    "# 创建判别器\n",
    "netD = Discriminator(ngpu).to(device)\n",
    "\n",
    "# 如果期望使用多GPU，设置一下\n",
    "if (device.type == 'cuda') and (ngpu > 1):\n",
    "    netD = nn.DataParallel(netD, list(range(ngpu)))\n",
    "\n",
    "# 使用权重初始化函数 weights_init 去随机初始化所有权重\n",
    "#  mean=0, stdev=0.2.\n",
    "netD.apply(weights_init)\n",
    "\n",
    "# 输出该模型\n",
    "print(netD)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "# 初始化 BCE损失函数\n",
    "criterion = nn.BCELoss()\n",
    "criterion_pic = nn.CrossEntropyLoss()\n",
    "# 创建一个批次的本征向量用于可视化生成器训练的过程。\n",
    "fixed_noise = torch.randn(64, nz + 10, 1, 1, device=device)\n",
    "\n",
    "# 建立一个在训练中使用的真实和假的标记\n",
    "real_label = 1\n",
    "fake_label = 0\n",
    "\n",
    "# 为G和D都设置Adam优化器 # 刚看到说G的lr是D的lr的2倍会好一些\n",
    "optimizerD = optim.Adam(netD.parameters(), lr=lr, betas=(beta1, 0.999))\n",
    "optimizerG = optim.Adam(netG.parameters(), lr=lr, betas=(beta1, 0.999))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Starting Training Loop...\n",
      "Congratulations! Training Finished!\n"
     ]
    }
   ],
   "source": [
    "# 训练循环\n",
    "\n",
    "#############resnet18##############\n",
    "model = models.resnet18(pretrained=True)\n",
    "for p in model.parameters():\n",
    "    p.requires_grad = False\n",
    "conv1 = model.conv1\n",
    "conv1.stride = 1\n",
    "conv1.padding = (0, 0)\n",
    "conv1 = conv1.to(device)\n",
    "\n",
    "\n",
    "\n",
    "# 保存跟踪进度的列表\n",
    "img_list = []\n",
    "G_losses = []\n",
    "D_losses = []\n",
    "iters = 0\n",
    "errD_real = float(0)\n",
    "print(\"Starting Training Loop...\")\n",
    "# 每个epoh\n",
    "for epoch in range(num_epochs):\n",
    "    # 数据加载器中的每个批次\n",
    "    for i, data in enumerate(dataloader, 0):\n",
    "        ############################\n",
    "        # (1) 更新 D 网络: 最大化 log(D(x)) + log(1 - D(G(z)))\n",
    "        ###########################\n",
    "        ## 使用所有真实样本批次训练\n",
    "        netD.zero_grad()\n",
    "        # 格式化批\n",
    "#         real_cpu = data[0].to(device)\n",
    "#         data = Variable(data)\n",
    "#         data = data.to(device)\n",
    "        \n",
    "        real_cpu = data[0].to(device)\n",
    "        ############reshape real_cpu###########\n",
    "        real_cpu = torch.cat((real_cpu,real_cpu,real_cpu),1)\n",
    "        #######################################\n",
    "        \n",
    "#         print(real_cpu.shape)\n",
    "        b_size = real_cpu.size(0)\n",
    "        label = torch.full((b_size,), real_label, device=device,dtype=torch.float)\n",
    "        \n",
    "        \n",
    "        #############提取original图片的特征###############\n",
    "        original_feature = conv1(real_cpu)\n",
    "        #################################################\n",
    "#         print(original_feature.shape)\n",
    "        \n",
    "        # 通过D向前传递真实批次的feature map\n",
    "        output = netD(original_feature)\n",
    "        \n",
    "        output = output.view(b_size,11)\n",
    "#         print(output.shape)\n",
    "        # output = output.view([b_size, 11]　　　　　\n",
    "        real_label_label = output[:, 0]  # 第一维就是预测是否是真实图片\n",
    "        real_label_pic = output[:, 1:]  # 后十维是onehot，判断是哪个数字\n",
    "        #   print(real_label_pic.shape)\n",
    "        # 对所有真实样本批次计算损失\n",
    "#         print(label.shape)\n",
    "        errD_real = criterion(real_label_label, label)\n",
    "        errD_real.backward(retain_graph=True)  # 连续多个backward，除了最后一个不叫retain_graph=True,其余都要加\n",
    "\n",
    "        errD_fake_pic = criterion_pic(real_label_pic, data[1].cuda(device))\n",
    "        errD_fake_pic.backward()\n",
    "\n",
    "        # 计算后向传递中D的梯度\n",
    "\n",
    "        D_x = real_label_label.mean().item()\n",
    "\n",
    "        ## 使用所有假样本批次训练\n",
    "        # 生成本征向量批次\n",
    "        #   noise = torch.randn(b_size, nz+10, 1, 1, device=device)  # 注释这行表示输入110维的noise，也就是DCGAN的输入\n",
    "        noise = torch.randn(b_size, nz, device=device)\n",
    "\n",
    "        labels_onehot = np.zeros((b_size, 10))\n",
    "        labels_onehot[np.arange(b_size), data[1].numpy()] = 1\n",
    "        noise = np.concatenate((noise.cpu().numpy(), labels_onehot), axis=1)\n",
    "        # 每张图片是0到9之间的数字\n",
    "        labels_onehot = Variable(torch.from_numpy(labels_onehot).float()).to(device)\n",
    "        noise = noise.reshape([-1, 110, 1, 1])\n",
    "        noise = Variable(torch.from_numpy(noise).float()).to(device)\n",
    "\n",
    "        # 使用生成器G生成假图片\n",
    "        fake = netG(noise)\n",
    "        label.fill_(fake_label)                               \n",
    "        \n",
    "        ############混肴视觉特征#####################\n",
    "        b,c,w,h = fake.size()\n",
    "        total_variance1 = 0\n",
    "        for i in range(3,w,3):\n",
    "            for j in range(3,h,3):\n",
    "                variance1=fake[:,:,i-3:i,j-3:j].var()\n",
    "                total_variance1=total_variance1+variance1\n",
    "        loss1=-total_variance1\n",
    "        ############################################\n",
    "        \n",
    "        #############提取fake图片的特征##############\n",
    "        fake = torch.cat((fake,fake,fake),1)\n",
    "        fake_feature = conv1(fake)\n",
    "        #############################################\n",
    "        \n",
    "        \n",
    "        # 使用判别器分类所有的假批次样本的feature map\n",
    "        output = netD(fake_feature.detach())\n",
    "        # 计算判别器D的损失对所有的假样本批次\n",
    "        #         output = output.view([b_size, 11])\n",
    "        output = output.view([b_size, 11])\n",
    "        real_label_label = output[:, 0]\n",
    "        real_label_pic = output[:, 1:]\n",
    "        errD_fake = criterion(real_label_label, label)\n",
    "        # 对这个批次计算梯度\n",
    "        errD_fake.backward(retain_graph=True)\n",
    "        \n",
    "        errD_fake_pic = criterion_pic(real_label_pic, data[1].cuda(device)) + 0.5*loss1\n",
    "        errD_fake_pic.backward(retain_graph=True)\n",
    "        D_G_z1 = real_label_label.mean().item()\n",
    "        # 把所有真样本和假样本批次的梯度加起来\n",
    "        errD = errD_real + errD_fake + errD_fake_pic \n",
    "        # 更新判别器D\n",
    "        optimizerD.step()\n",
    "\n",
    "        ############################\n",
    "        # (2) 更新 G 网络: 最大化 log(D(G(z)))\n",
    "        ###########################\n",
    "        netG.zero_grad()\n",
    "        label.fill_(real_label)  # 假样本的标签对于生成器成本是真的\n",
    "        # 因为我们之更新了D，通过D执行所有假样本批次的正向传递\n",
    "        output = netD(fake_feature).view(-1) #这里是将netD的输入fake换成了fake_feature\n",
    "        # 基于这个输出计算G的损失\n",
    "        output = output.view([b_size, 11])\n",
    "        real_label_label = output[:, 0]\n",
    "        real_label_pic = output[:, 1:]\n",
    "        errG = criterion(real_label_label, label)\n",
    "        # 为生成器计算梯度\n",
    "        errG.backward(retain_graph=True)\n",
    "        errD_fake_pic = criterion_pic(real_label_pic, data[1].cuda(device)) + 0.5*loss1\n",
    "        errD_fake_pic.backward()\n",
    "        D_G_z2 = output.mean().item()\n",
    "        # 更新生成器G\n",
    "        optimizerG.step()\n",
    "\n",
    "        # 输出训练状态\n",
    "        if i % 50 == 0:\n",
    "            print('[%d/%d][%d/%d]\\tLoss_D: %.4f\\tLoss_G: %.4f\\tD(x): %.4f\\tD(G(z)): %.4f / %.4f'\n",
    "                  % (epoch, num_epochs, i, len(dataloader),\n",
    "                     errD.item(), errG.item(), D_x, D_G_z1, D_G_z2))\n",
    "\n",
    "        # 为以后画损失图，保存损失\n",
    "        G_losses.append(errG.item())\n",
    "        D_losses.append(errD.item())\n",
    "\n",
    "        # 检查生成器generator做了什么，通过保存的fixed_noise通过G的输出\n",
    "        if (iters % 500 == 0) or ((epoch == num_epochs - 1) and (i == len(dataloader) - 1)):\n",
    "            with torch.no_grad():\n",
    "                fake = netG(fixed_noise).detach().cpu()\n",
    "            img_list.append(vutils.make_grid(fake, padding=2, normalize=True))\n",
    "\n",
    "        iters += 1\n",
    "print(\"Congratulations! Training Finished!\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 87,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#Draw Losses\n",
    "# 对照组\n",
    "plt.figure(figsize=(10,5))\n",
    "plt.subplot(1,2,1)\n",
    "plt.title(\"Generator Loss\")\n",
    "plt.plot(G_losses,label=\"G\")\n",
    "plt.legend()\n",
    "plt.subplot(1,2,2)\n",
    "plt.title(\"Discriminator Loss\")\n",
    "plt.plot(D_losses,label=\"D\")\n",
    "# plt.plot(D_penalties,label=\"D_penalty\")\n",
    "# plt.xlabel(\"iterations\")\n",
    "# plt.ylabel(\"Loss\")\n",
    "plt.legend()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'dataloader' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-2-4a37c170bcc6>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m()\u001b[0m\n\u001b[0;32m      1\u001b[0m \u001b[1;31m# 从数据加载器中获取一批真实图像\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m----> 2\u001b[1;33m \u001b[0mreal_batch\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mnext\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0miter\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mdataloader\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      3\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      4\u001b[0m \u001b[1;31m# 画出真实图像\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      5\u001b[0m \u001b[0mplt\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfigure\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mfigsize\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;36m15\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m15\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mNameError\u001b[0m: name 'dataloader' is not defined"
     ]
    }
   ],
   "source": [
    "# 从数据加载器中获取一批真实图像\n",
    "real_batch = next(iter(dataloader))\n",
    "\n",
    "# 画出真实图像\n",
    "plt.figure(figsize=(15,15))\n",
    "plt.subplot(1,2,1)\n",
    "plt.axis(\"off\")\n",
    "plt.title(\"Real Images\")\n",
    "plt.imshow(np.transpose(vutils.make_grid(real_batch[0].to(device)[:64], padding=5, normalize=True).cpu(),(1,2,0)))\n",
    "\n",
    "# 画出来自最后一次训练的假图像\n",
    "plt.subplot(1,2,2)\n",
    "plt.axis(\"off\")\n",
    "plt.title(\"Fake Images\")\n",
    "plt.imshow(np.transpose(img_list[-1],(1,2,0)))\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'torch' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-1-dd4274784762>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m()\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m \u001b[0mtorch\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msave\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mnetG\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mstate_dict\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;34m\"C:\\\\Users\\\\lz172\\\\Desktop\\\\jupyter\\\\defense-fl-gan-attack-main\\\\saved-models\\\\CGAN-Generator-MaxLogVar48-epoch60-Fashion_MNIST-6\\\\netG_params.pkl\"\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      2\u001b[0m \u001b[0mtorch\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msave\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mnetD\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mstate_dict\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;34m\"C:\\\\Users\\\\lz172\\\\Desktop\\\\jupyter\\\\defense-fl-gan-attack-main\\\\saved-models\\\\CGAN-Discriminator-MaxLogVar48-epoch60-Fashion_MNIST-6\\\\netD_params.pkl\"\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      3\u001b[0m \u001b[0mnoise\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mrandn\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mb_size\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mnz\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mdevice\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mdevice\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      4\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      5\u001b[0m \u001b[0mlabels_onehot\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mzeros\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mb_size\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;36m10\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mNameError\u001b[0m: name 'torch' is not defined"
     ]
    }
   ],
   "source": [
    "torch.save(netG.state_dict(), \"C:\\\\Users\\\\lz172\\\\Desktop\\\\jupyter\\\\defense-fl-gan-attack-main\\\\saved-models\\\\CGAN-Generator-MaxLogVar48-epoch60-Fashion_MNIST-6\\\\netG_params.pkl\")\n",
    "torch.save(netD.state_dict(), \"C:\\\\Users\\\\lz172\\\\Desktop\\\\jupyter\\\\defense-fl-gan-attack-main\\\\saved-models\\\\CGAN-Discriminator-MaxLogVar48-epoch60-Fashion_MNIST-6\\\\netD_params.pkl\")\n",
    "noise = torch.randn(b_size, nz, device=device)\n",
    "        \n",
    "labels_onehot = np.zeros((b_size,10))\n",
    "labels_onehot[np.arange(b_size),0]=1 # 这个9换成想生成的数字\n",
    "noise=np.concatenate((noise.cpu().numpy(), labels_onehot),axis=1)\n",
    "# 每张图片是0到9之间的数字\n",
    "labels_onehot = Variable(torch.from_numpy(labels_onehot).float()).to(device)\n",
    "noise=noise.reshape([-1, 110, 1, 1])\n",
    "noise=Variable(torch.from_numpy(noise).float()).to(device)\n",
    "\n",
    "# 使用生成器G生成假图片\n",
    "fake = netG(noise)\n",
    "plt.imshow(fake.cpu().detach().numpy()[5][0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.0"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}