PyTorch图像分类器.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "gpuType": "T4",
      "authorship_tag": "ABX9TyNmH2WfUN7BAUny4hR0DfUG",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/DENGCPU/CSS-StudyDemos/blob/main/PyTorch%E5%9B%BE%E5%83%8F%E5%88%86%E7%B1%BB%E5%99%A8.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "id": "yYiPfLkGN7F2"
      },
      "outputs": [],
      "source": [
        "# 使用CIFAR10数据集，，它包含十个类别：‘airplane’, ‘automobile’, ‘bird’, ‘cat’, ‘deer’, ‘dog’, ‘frog’, ‘horse’, ‘ship’, ‘truck’。\n",
        "# CIFAR-10 中的图像尺寸为33232，\n",
        "# 也就是RGB的3层颜色通道，每层通道内的尺寸为32*32"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import torch\n",
        "import torchvision\n",
        "import torchvision.transforms as transforms\n",
        "\n",
        "transform = transforms.Compose(\n",
        "    [transforms.ToTensor(),\n",
        "     transforms.Normalize((0.5,0.5,0.5),(0.5,0.5,0.5))]\n",
        ")\n",
        "trainset = torchvision.datasets.CIFAR10(root='./data',train=True,\n",
        "                                       download=True,transform=transform)\n",
        "trainloader = torch.utils.data.DataLoader(trainset,batch_size=4,\n",
        "                                          shuffle=True,num_workers=2)\n",
        "testset  = torchvision.datasets.CIFAR10(root='./data',train=False,\n",
        "                                        download=True,transform =transform)\n",
        "testloader = torch.utils.data.DataLoader(testset,batch_size=4,\n",
        "                                         shuffle=False,num_workers=2)\n",
        "classes = ('plane', 'car', 'bird', 'cat',\n",
        "           'deer', 'dog', 'frog', 'horse', 'ship', 'truck')\n"
      ],
      "metadata": {
        "id": "Z65PmMswOQcr",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "3cbe438e-56f0-4f51-d89e-2a7d71e83645"
      },
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz to ./data/cifar-10-python.tar.gz\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "100%|██████████| 170498071/170498071 [00:02<00:00, 66163765.81it/s]\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Extracting ./data/cifar-10-python.tar.gz to ./data\n",
            "Files already downloaded and verified\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# 展示一些训练图片\n",
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "\n",
        "def imshow(img):\n",
        "  #反归一化\n",
        "  img = img / 2 + 0.5\n",
        "  npimg = img.numpy()\n",
        "  #转置\n",
        "  # 在PyTorch和TensorFlow等深度学习框架中，图像的维度顺序通常是(通道数, 高度, 宽度)\n",
        "  # 然而，对于matplotlib.imshow()函数，图像的维度顺序应该是(高度, 宽度, 通道数)\n",
        "  plt.imshow(np.transpose(npimg,(1,2,0)))\n",
        "  plt.show()\n",
        "\n",
        "# get some random training images\n",
        "dataiter = iter(trainloader)\n",
        "images, labels = next(dataiter)\n",
        "\n",
        "# show images\n",
        "imshow(torchvision.utils.make_grid(images))\n",
        "# print labels\n",
        "print(' '.join('%5s' % classes[labels[j]] for j in range(4)))\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 210
        },
        "id": "LjfTjdUCQgt-",
        "outputId": "4c013b92-002a-499b-bbc6-681864064684"
      },
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "  cat  ship plane  bird\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# 定义一个卷积神经网络\n",
        "\n",
        "import torch.nn as nn\n",
        "import torch.nn.functional as F\n",
        "\n",
        "\n",
        "class Net(nn.Module):\n",
        "    def __init__(self):\n",
        "        super(Net, self).__init__()\n",
        "        self.conv1 = nn.Conv2d(3, 6, 5)\n",
        "        self.pool = nn.MaxPool2d(2, 2)\n",
        "        self.conv2 = nn.Conv2d(6, 16, 5)\n",
        "        self.fc1 = nn.Linear(16 * 5 * 5, 120)\n",
        "        self.fc2 = nn.Linear(120, 84)\n",
        "        self.fc3 = nn.Linear(84, 10)\n",
        "\n",
        "    def forward(self, x):\n",
        "        x = self.pool(F.relu(self.conv1(x)))\n",
        "        x = self.pool(F.relu(self.conv2(x)))\n",
        "        x = x.view(-1, 16 * 5 * 5)\n",
        "        x = F.relu(self.fc1(x))\n",
        "        x = F.relu(self.fc2(x))\n",
        "        x = self.fc3(x)\n",
        "        return x\n",
        "\n",
        "\n",
        "net = Net()\n"
      ],
      "metadata": {
        "id": "E7Zyem13S1AS"
      },
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# 定义一个损失函数Cross-Entropy和优化器SGD\n",
        "import torch.optim as optim\n",
        "\n",
        "criterion = nn.CrossEntropyLoss()\n",
        "optimizer = optim.SGD(net.parameters(),lr=0.001,momentum=0.9)\n",
        "\n",
        "for epoch in range(2):\n",
        "  running_loss = 0.0\n",
        "  for i ,data in enumerate(trainloader,0):\n",
        "    inputs, labels = data\n",
        "    optimizer.zero_grad()\n",
        "    #forward+ backward + optimize\n",
        "    outputs = net(inputs)\n",
        "    loss = criterion(outputs,labels)\n",
        "    optimizer.step()\n",
        "\n",
        "    running_loss += loss.item()\n",
        "    if i% 2000 ==1999:\n",
        "      print('[%d, %5d] loss: %.3f' %\n",
        "            (epoch+1 ,i+1, running_loss / 2000))\n",
        "      running_loss = 0.0\n",
        "\n",
        "print('Finished Training')\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ClPbcDn0TOjY",
        "outputId": "7d2b5ead-0351-42b1-cb4b-489bdca11c20"
      },
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "[1,  2000] loss: 2.304\n",
            "[1,  4000] loss: 2.304\n",
            "[1,  6000] loss: 2.304\n",
            "[1,  8000] loss: 2.303\n",
            "[1, 10000] loss: 2.304\n",
            "[1, 12000] loss: 2.304\n",
            "[2,  2000] loss: 2.304\n",
            "[2,  4000] loss: 2.305\n",
            "[2,  6000] loss: 2.303\n",
            "[2,  8000] loss: 2.303\n",
            "[2, 10000] loss: 2.304\n",
            "[2, 12000] loss: 2.305\n",
            "Finished Training\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "outputs = net(images)\n",
        "\n",
        "_,predicted = torch.max(outputs,1)\n",
        "\n",
        "print('Predicted:', ','.join('%5s' % classes[predicted[j]] for j in range(4)))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "usnVpthYV6ea",
        "outputId": "b4f34ad2-1a8b-4ecb-efbb-a68b074bbbde"
      },
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Predicted: plane,plane,plane,plane\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "correct = 0\n",
        "total = 0\n",
        "with torch.no_grad():\n",
        "  for data in testloader:\n",
        "    images,labels = data\n",
        "    outputs = net(images)\n",
        "    _,pretdicted = torch.max(outputs.data,1)\n",
        "    total += labels.size(0)\n",
        "    correct += (predicted == labels).sum().item()\n",
        "\n",
        "print('Accuracy of the network on the 10000 test images: %d %%' % (\n",
        "    100 * correct /total)\n",
        ")\n"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "bzatTzAOWYfH",
        "outputId": "b5d4c94f-8ee9-45c8-c2b6-41a4b51a23e5"
      },
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy of the network on the 10000 test images: 10 %\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "class_correct = list(0. for i in range(10))\n",
        "class_total = list(0. for i in range(10))\n",
        "\n",
        "with torch.no_grad():\n",
        "  for data in testloader:\n",
        "    images,labels=data\n",
        "    outputs=net(images)\n",
        "    _,predicted=torch.max(outputs,1)\n",
        "    c= (predicted==labels).squeeze()\n",
        "    for i in range(4):\n",
        "      label = labels[i]\n",
        "      class_correct[label]+=c[i].item()\n",
        "      class_total[label]+=1\n",
        "\n",
        "for i in range(10):\n",
        "  print('Accuracy of %5s: %2d %%' % (\n",
        "      classes[i], 100 *class_correct[i] / class_total[i]\n",
        "  ))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-6MXAfglXeIX",
        "outputId": "26e358a3-7f38-445f-ff25-ab63b9cb887b"
      },
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy of plane: 100 %\n",
            "Accuracy of   car:  0 %\n",
            "Accuracy of  bird:  0 %\n",
            "Accuracy of   cat:  0 %\n",
            "Accuracy of  deer:  0 %\n",
            "Accuracy of   dog:  0 %\n",
            "Accuracy of  frog:  0 %\n",
            "Accuracy of horse:  0 %\n",
            "Accuracy of  ship:  0 %\n",
            "Accuracy of truck:  0 %\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# 在GPU上跑\n",
        "device = torch.device(\"cuda:0\" if torch.cuda.is_available() else 'cpu')\n",
        "print(device)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "t-Pehe3qY37Q",
        "outputId": "1f191db7-fe96-4f36-e09a-f157e5c56608"
      },
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "cuda:0\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "net.to(device)"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "osXnJUOCaHId",
        "outputId": "373ed47d-25ec-4ac0-ac22-9b2dd13af23f"
      },
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Net(\n",
              "  (conv1): Conv2d(3, 6, kernel_size=(5, 5), stride=(1, 1))\n",
              "  (pool): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
              "  (conv2): Conv2d(6, 16, kernel_size=(5, 5), stride=(1, 1))\n",
              "  (fc1): Linear(in_features=400, out_features=120, bias=True)\n",
              "  (fc2): Linear(in_features=120, out_features=84, bias=True)\n",
              "  (fc3): Linear(in_features=84, out_features=10, bias=True)\n",
              ")"
            ]
          },
          "metadata": {},
          "execution_count": 11
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "inputs, labels=inputs.to(device), labels.to(device)"
      ],
      "metadata": {
        "id": "sGbSmIZ8aLJk"
      },
      "execution_count": 12,
      "outputs": []
    }
  ]
}