4-3_CIFAR_CNN_Augmentation.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "''' 1. Module Import '''\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "from torchvision import transforms, datasets"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Using PyTorch version: 1.6.0+cu101  Device: cuda\n"
     ]
    }
   ],
   "source": [
    "''' 2. 딥러닝 모델을 설계할 때 활용하는 장비 확인 '''\n",
    "if torch.cuda.is_available():\n",
    "    DEVICE = torch.device('cuda')\n",
    "else:\n",
    "    DEVICE = torch.device('cpu')\n",
    "print('Using PyTorch version:', torch.__version__, ' Device:', DEVICE)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "BATCH_SIZE = 32\n",
    "EPOCHS = 10"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Files already downloaded and verified\n"
     ]
    }
   ],
   "source": [
    "''' 3. Data Augmentation이 적용된 CIFAR10 데이터 다운로드 (Train set, Test set 분리하기) '''\n",
    "train_dataset = datasets.CIFAR10(root = \"../data/CIFAR_10\",\n",
    "                                  train = True,\n",
    "                                  download = True,\n",
    "                                  transform = transforms.Compose([\n",
    "                                    transforms.RandomHorizontalFlip(),\n",
    "                                    transforms.ToTensor(),\n",
    "                                    transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))]))\n",
    "\n",
    "test_dataset = datasets.CIFAR10(root = \"../data/CIFAR_10\",\n",
    "                                train = False,\n",
    "                                transform = transforms.Compose([\n",
    "                                    transforms.RandomHorizontalFlip(),\n",
    "                                    transforms.ToTensor(),\n",
    "                                    transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))]))\n",
    "\n",
    "train_loader = torch.utils.data.DataLoader(dataset = train_dataset,\n",
    "                                            batch_size = BATCH_SIZE,\n",
    "                                            shuffle = True)\n",
    "\n",
    "test_loader = torch.utils.data.DataLoader(dataset = test_dataset,\n",
    "                                          batch_size = BATCH_SIZE,\n",
    "                                          shuffle = False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "X_train: torch.Size([32, 3, 32, 32]) type: torch.FloatTensor\n",
      "y_train: torch.Size([32]) type: torch.LongTensor\n"
     ]
    }
   ],
   "source": [
    "''' 4. 데이터 확인하기 (1) '''\n",
    "for (X_train, y_train) in train_loader:\n",
    "    print('X_train:', X_train.size(), 'type:', X_train.type())\n",
    "    print('y_train:', y_train.size(), 'type:', y_train.type())\n",
    "    break"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x72 with 10 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "''' 5. 데이터 확인하기 (2) '''\n",
    "pltsize = 1\n",
    "plt.figure(figsize=(10 * pltsize, pltsize))\n",
    "\n",
    "for i in range(10):\n",
    "    plt.subplot(1, 10, i + 1)\n",
    "    plt.axis('off')\n",
    "    plt.imshow(np.transpose(X_train[i], (1, 2, 0)))\n",
    "    plt.title('Class: ' + str(y_train[i].item()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "''' 6. Convolutional Neural Network (CNN) 모델 설계하기 '''\n",
    "class CNN(nn.Module):\n",
    "    def __init__(self):\n",
    "        super(CNN, self).__init__()\n",
    "        self.conv1 = nn.Conv2d(in_channels = 3, out_channels = 8, kernel_size = 3, padding = 1)\n",
    "        self.conv2 = nn.Conv2d(in_channels = 8, out_channels = 16, kernel_size = 3, padding = 1)\n",
    "        self.pool = nn.MaxPool2d(kernel_size = 2, stride = 2)\n",
    "        self.fc1 = nn.Linear(8 * 8 * 16, 64)\n",
    "        self.fc2 = nn.Linear(64, 32)\n",
    "        self.fc3 = nn.Linear(32, 10)\n",
    "        \n",
    "    def forward(self, x):\n",
    "        x = self.conv1(x)\n",
    "        x = F.relu(x)\n",
    "        x = self.pool(x)\n",
    "        x = self.conv2(x)\n",
    "        x = F.relu(x)\n",
    "        x = self.pool(x)\n",
    "        \n",
    "        x = x.view(-1, 8 * 8 * 16)\n",
    "        x = self.fc1(x)\n",
    "        x = F.relu(x)\n",
    "        x = self.fc2(x)\n",
    "        x = F.relu(x)\n",
    "        x = self.fc3(x)\n",
    "        x = F.log_softmax(x)\n",
    "        return x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CNN(\n",
      "  (conv1): Conv2d(3, 8, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "  (conv2): Conv2d(8, 16, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "  (pool): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "  (fc1): Linear(in_features=1024, out_features=64, bias=True)\n",
      "  (fc2): Linear(in_features=64, out_features=32, bias=True)\n",
      "  (fc3): Linear(in_features=32, out_features=10, bias=True)\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "''' 7. Optimizer, Objective Function 설정하기 '''\n",
    "model = CNN().to(DEVICE)\n",
    "optimizer = torch.optim.Adam(model.parameters(), lr = 0.001)\n",
    "criterion = nn.CrossEntropyLoss()\n",
    "\n",
    "print(model)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "''' 8. CNN 모델 학습을 진행하며 학습 데이터에 대한 모델 성능을 확인하는 함수 정의 '''\n",
    "def train(model, train_loader, optimizer, log_interval):\n",
    "    model.train()\n",
    "    for batch_idx, (image, label) in enumerate(train_loader):\n",
    "        image = image.to(DEVICE)\n",
    "        label = label.to(DEVICE)\n",
    "        optimizer.zero_grad()\n",
    "        output = model(image)\n",
    "        loss = criterion(output, label)\n",
    "        loss.backward()\n",
    "        optimizer.step()\n",
    "\n",
    "        if batch_idx % log_interval == 0:\n",
    "            print(\"Train Epoch: {} [{}/{} ({:.0f}%)]\\tTrain Loss: {:.6f}\".format(\n",
    "                epoch, batch_idx * len(image), \n",
    "                len(train_loader.dataset), 100. * batch_idx / len(train_loader), \n",
    "                loss.item()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "''' 9. 학습되는 과정 속에서 검증 데이터에 대한 모델 성능을 확인하는 함수 정의 '''\n",
    "def evaluate(model, test_loader):\n",
    "    model.eval()\n",
    "    test_loss = 0\n",
    "    correct = 0\n",
    "\n",
    "    with torch.no_grad():\n",
    "        for image, label in test_loader:\n",
    "            image = image.to(DEVICE)\n",
    "            label = label.to(DEVICE)\n",
    "            output = model(image)\n",
    "            test_loss += criterion(output, label).item()\n",
    "            prediction = output.max(1, keepdim = True)[1]\n",
    "            correct += prediction.eq(label.view_as(prediction)).sum().item()\n",
    "    \n",
    "    test_loss /= (len(test_loader.dataset) / BATCH_SIZE)\n",
    "    test_accuracy = 100. * correct / len(test_loader.dataset)\n",
    "    return test_loss, test_accuracy"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "c:\\users\\justin\\101\\lib\\site-packages\\ipykernel_launcher.py:26: UserWarning: Implicit dimension choice for log_softmax has been deprecated. Change the call to include dim=X as an argument.\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train Epoch: 1 [0/50000 (0%)]\tTrain Loss: 2.305369\n",
      "Train Epoch: 1 [6400/50000 (13%)]\tTrain Loss: 1.932201\n",
      "Train Epoch: 1 [12800/50000 (26%)]\tTrain Loss: 1.581616\n",
      "Train Epoch: 1 [19200/50000 (38%)]\tTrain Loss: 1.436576\n",
      "Train Epoch: 1 [25600/50000 (51%)]\tTrain Loss: 1.747659\n",
      "Train Epoch: 1 [32000/50000 (64%)]\tTrain Loss: 1.379740\n",
      "Train Epoch: 1 [38400/50000 (77%)]\tTrain Loss: 1.632054\n",
      "Train Epoch: 1 [44800/50000 (90%)]\tTrain Loss: 1.155150\n",
      "\n",
      "[EPOCH: 1], \tTest Loss: 0.0424, \tTest Accuracy: 50.91 % \n",
      "\n",
      "Train Epoch: 2 [0/50000 (0%)]\tTrain Loss: 1.183661\n",
      "Train Epoch: 2 [6400/50000 (13%)]\tTrain Loss: 1.264468\n",
      "Train Epoch: 2 [12800/50000 (26%)]\tTrain Loss: 1.477151\n",
      "Train Epoch: 2 [19200/50000 (38%)]\tTrain Loss: 1.120275\n",
      "Train Epoch: 2 [25600/50000 (51%)]\tTrain Loss: 1.553942\n",
      "Train Epoch: 2 [32000/50000 (64%)]\tTrain Loss: 1.377848\n",
      "Train Epoch: 2 [38400/50000 (77%)]\tTrain Loss: 0.961657\n",
      "Train Epoch: 2 [44800/50000 (90%)]\tTrain Loss: 1.553190\n",
      "\n",
      "[EPOCH: 2], \tTest Loss: 0.0378, \tTest Accuracy: 57.22 % \n",
      "\n",
      "Train Epoch: 3 [0/50000 (0%)]\tTrain Loss: 1.093908\n",
      "Train Epoch: 3 [6400/50000 (13%)]\tTrain Loss: 0.905866\n",
      "Train Epoch: 3 [12800/50000 (26%)]\tTrain Loss: 1.269535\n",
      "Train Epoch: 3 [19200/50000 (38%)]\tTrain Loss: 0.810739\n",
      "Train Epoch: 3 [25600/50000 (51%)]\tTrain Loss: 0.864472\n",
      "Train Epoch: 3 [32000/50000 (64%)]\tTrain Loss: 0.559747\n",
      "Train Epoch: 3 [38400/50000 (77%)]\tTrain Loss: 1.348746\n",
      "Train Epoch: 3 [44800/50000 (90%)]\tTrain Loss: 0.959024\n",
      "\n",
      "[EPOCH: 3], \tTest Loss: 0.0356, \tTest Accuracy: 59.70 % \n",
      "\n",
      "Train Epoch: 4 [0/50000 (0%)]\tTrain Loss: 1.046205\n",
      "Train Epoch: 4 [6400/50000 (13%)]\tTrain Loss: 0.929529\n",
      "Train Epoch: 4 [12800/50000 (26%)]\tTrain Loss: 0.840712\n",
      "Train Epoch: 4 [19200/50000 (38%)]\tTrain Loss: 1.383940\n",
      "Train Epoch: 4 [25600/50000 (51%)]\tTrain Loss: 0.947381\n",
      "Train Epoch: 4 [32000/50000 (64%)]\tTrain Loss: 1.163367\n",
      "Train Epoch: 4 [38400/50000 (77%)]\tTrain Loss: 1.128562\n",
      "Train Epoch: 4 [44800/50000 (90%)]\tTrain Loss: 0.942198\n",
      "\n",
      "[EPOCH: 4], \tTest Loss: 0.0331, \tTest Accuracy: 62.30 % \n",
      "\n",
      "Train Epoch: 5 [0/50000 (0%)]\tTrain Loss: 0.905439\n",
      "Train Epoch: 5 [6400/50000 (13%)]\tTrain Loss: 1.051276\n",
      "Train Epoch: 5 [12800/50000 (26%)]\tTrain Loss: 0.964778\n",
      "Train Epoch: 5 [19200/50000 (38%)]\tTrain Loss: 1.056428\n",
      "Train Epoch: 5 [25600/50000 (51%)]\tTrain Loss: 1.059727\n",
      "Train Epoch: 5 [32000/50000 (64%)]\tTrain Loss: 0.727895\n",
      "Train Epoch: 5 [38400/50000 (77%)]\tTrain Loss: 0.787058\n",
      "Train Epoch: 5 [44800/50000 (90%)]\tTrain Loss: 0.854833\n",
      "\n",
      "[EPOCH: 5], \tTest Loss: 0.0317, \tTest Accuracy: 64.76 % \n",
      "\n",
      "Train Epoch: 6 [0/50000 (0%)]\tTrain Loss: 0.936069\n",
      "Train Epoch: 6 [6400/50000 (13%)]\tTrain Loss: 0.710714\n",
      "Train Epoch: 6 [12800/50000 (26%)]\tTrain Loss: 1.168326\n",
      "Train Epoch: 6 [19200/50000 (38%)]\tTrain Loss: 1.110862\n",
      "Train Epoch: 6 [25600/50000 (51%)]\tTrain Loss: 1.109888\n",
      "Train Epoch: 6 [32000/50000 (64%)]\tTrain Loss: 1.210671\n",
      "Train Epoch: 6 [38400/50000 (77%)]\tTrain Loss: 0.947488\n",
      "Train Epoch: 6 [44800/50000 (90%)]\tTrain Loss: 0.843470\n",
      "\n",
      "[EPOCH: 6], \tTest Loss: 0.0311, \tTest Accuracy: 65.35 % \n",
      "\n",
      "Train Epoch: 7 [0/50000 (0%)]\tTrain Loss: 0.982925\n",
      "Train Epoch: 7 [6400/50000 (13%)]\tTrain Loss: 0.945703\n",
      "Train Epoch: 7 [12800/50000 (26%)]\tTrain Loss: 0.923167\n",
      "Train Epoch: 7 [19200/50000 (38%)]\tTrain Loss: 0.828612\n",
      "Train Epoch: 7 [25600/50000 (51%)]\tTrain Loss: 1.258147\n",
      "Train Epoch: 7 [32000/50000 (64%)]\tTrain Loss: 1.198558\n",
      "Train Epoch: 7 [38400/50000 (77%)]\tTrain Loss: 0.748312\n",
      "Train Epoch: 7 [44800/50000 (90%)]\tTrain Loss: 1.061050\n",
      "\n",
      "[EPOCH: 7], \tTest Loss: 0.0302, \tTest Accuracy: 66.16 % \n",
      "\n",
      "Train Epoch: 8 [0/50000 (0%)]\tTrain Loss: 0.904049\n",
      "Train Epoch: 8 [6400/50000 (13%)]\tTrain Loss: 0.937751\n",
      "Train Epoch: 8 [12800/50000 (26%)]\tTrain Loss: 0.979157\n",
      "Train Epoch: 8 [19200/50000 (38%)]\tTrain Loss: 0.786449\n",
      "Train Epoch: 8 [25600/50000 (51%)]\tTrain Loss: 0.603784\n",
      "Train Epoch: 8 [32000/50000 (64%)]\tTrain Loss: 0.812389\n",
      "Train Epoch: 8 [38400/50000 (77%)]\tTrain Loss: 0.885803\n",
      "Train Epoch: 8 [44800/50000 (90%)]\tTrain Loss: 1.081109\n",
      "\n",
      "[EPOCH: 8], \tTest Loss: 0.0303, \tTest Accuracy: 66.23 % \n",
      "\n",
      "Train Epoch: 9 [0/50000 (0%)]\tTrain Loss: 0.733635\n",
      "Train Epoch: 9 [6400/50000 (13%)]\tTrain Loss: 0.707936\n",
      "Train Epoch: 9 [12800/50000 (26%)]\tTrain Loss: 0.651994\n",
      "Train Epoch: 9 [19200/50000 (38%)]\tTrain Loss: 0.773648\n",
      "Train Epoch: 9 [25600/50000 (51%)]\tTrain Loss: 1.683995\n",
      "Train Epoch: 9 [32000/50000 (64%)]\tTrain Loss: 0.738951\n",
      "Train Epoch: 9 [38400/50000 (77%)]\tTrain Loss: 0.878403\n",
      "Train Epoch: 9 [44800/50000 (90%)]\tTrain Loss: 1.042976\n",
      "\n",
      "[EPOCH: 9], \tTest Loss: 0.0298, \tTest Accuracy: 66.60 % \n",
      "\n",
      "Train Epoch: 10 [0/50000 (0%)]\tTrain Loss: 0.907567\n",
      "Train Epoch: 10 [6400/50000 (13%)]\tTrain Loss: 0.965306\n",
      "Train Epoch: 10 [12800/50000 (26%)]\tTrain Loss: 1.070564\n",
      "Train Epoch: 10 [19200/50000 (38%)]\tTrain Loss: 1.250733\n",
      "Train Epoch: 10 [25600/50000 (51%)]\tTrain Loss: 0.730824\n",
      "Train Epoch: 10 [32000/50000 (64%)]\tTrain Loss: 0.759369\n",
      "Train Epoch: 10 [38400/50000 (77%)]\tTrain Loss: 1.060894\n",
      "Train Epoch: 10 [44800/50000 (90%)]\tTrain Loss: 0.703006\n",
      "\n",
      "[EPOCH: 10], \tTest Loss: 0.0306, \tTest Accuracy: 65.85 % \n",
      "\n"
     ]
    }
   ],
   "source": [
    "''' 10. CNN 학습 실행하며 Train, Test set의 Loss 및 Test set Accuracy 확인하기 '''\n",
    "for epoch in range(1, EPOCHS + 1):\n",
    "    train(model, train_loader, optimizer, log_interval = 200)\n",
    "    test_loss, test_accuracy = evaluate(model, test_loader)\n",
    "    print(\"\\n[EPOCH: {}], \\tTest Loss: {:.4f}, \\tTest Accuracy: {:.2f} % \\n\".format(\n",
    "        epoch, test_loss, test_accuracy))"
   ]
  }
 ],
 "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": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}