02-PyTorch/08-dnn-fashion-mnist.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "0bd10b24",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import torch\n",
    "from torch.utils.data import Dataset\n",
    "from torchvision import datasets, transforms"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "6984328a",
   "metadata": {},
   "source": [
    "- `torchvision.datasets` 에서 데이터 로드\n",
    "- 아래 링크에서 built-in datasets의 목록을 확인해 볼 수 있습니다.\n",
    "  - [PyTorch Built-in Datsets](https://pytorch.org/vision/stable/datasets.html)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "504ced26",
   "metadata": {},
   "source": [
    "`torchvision`의 `Image Transform` 에 대하여 생소하다면 **다음의 링크를 참고**해 주시기 바랍니다.\n",
    "\n",
    "- [torchvision의 transform으로 이미지 정규화하기(평균, 표준편차를 계산하여 적용](https://teddylee777.github.io/pytorch/torchvision-transform)\n",
    "- [PyTorch 이미지 데이터셋(Image Dataset) 구성에 관한 거의 모든 것!](https://teddylee777.github.io/pytorch/dataset-dataloader)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "046aeed5",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Image Transform 정의\n",
    "transform = transforms.Compose([\n",
    "    transforms.ToTensor(),\n",
    "])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1d0a2725",
   "metadata": {},
   "source": [
    "`Fashion MNIST` 내장 데이터셋을 로드하여 실습을 진행합니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "fa434082",
   "metadata": {},
   "outputs": [],
   "source": [
    "# train(학습용) 데이터셋 로드\n",
    "train_data = datasets.FashionMNIST(root='data', \n",
    "                                   train=True,        # 학습용 데이터셋 설정(True)\n",
    "                                   download=True, \n",
    "                                   transform=transform                \n",
    "                                  )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "00d964a5",
   "metadata": {},
   "outputs": [],
   "source": [
    "# test(학습용) 데이터셋 로드\n",
    "test_data = datasets.FashionMNIST(root='data', \n",
    "                                  train=False,        # 검증용 데이터셋 설정(False)\n",
    "                                  download=True, \n",
    "                                  transform=transform\n",
    "                                 )"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "9e8bb61b",
   "metadata": {},
   "source": [
    "`FashionMNIST` 데이터셋 시각화\n",
    "\n",
    "- 총 10개의 카테고리로 구성되어 있으며, `Label`은 아래 코드에서 `labels_map`에 정의되어 있습니다.\n",
    "- 출처: [zalandoresearch/fashion-mnist](https://github.com/zalandoresearch/fashion-mnist)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "1695f09d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 30 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "\n",
    "labels_map = {\n",
    "    0: \"t-shirt/top\",\n",
    "    1: \"trouser\",\n",
    "    2: \"pullover\",\n",
    "    3: \"dress\",\n",
    "    4: \"coat\",\n",
    "    5: \"sandal\",\n",
    "    6: \"shirt\",\n",
    "    7: \"sneaker\",\n",
    "    8: \"bag\",\n",
    "    9: \"ankle boot\",\n",
    "}\n",
    "\n",
    "figure = plt.figure(figsize=(10, 10))\n",
    "cols, rows = 6, 5\n",
    "\n",
    "for i in range(1, cols * rows + 1):\n",
    "    sample_idx = torch.randint(len(train_data), size=(1,)).item()\n",
    "    img, label = train_data[sample_idx]\n",
    "    figure.add_subplot(rows, cols, i)\n",
    "    plt.title(labels_map[label])\n",
    "    plt.axis(\"off\")\n",
    "    plt.imshow(torch.permute(img, (1, 2, 0)), cmap='gray')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "e726f4b1",
   "metadata": {},
   "source": [
    "## torch.utils.data.DataLoader\n",
    "\n",
    "`DataLoader`는 배치 구성과 shuffle등을 편하게 구성해 주는 util 입니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "0507e65f",
   "metadata": {},
   "outputs": [],
   "source": [
    "batch_size = 32 # batch_size 지정\n",
    "num_workers = 8 # Thread 숫자 지정 (병렬 처리에 활용할 쓰레드 숫자 지정)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "d4d3317a",
   "metadata": {},
   "outputs": [],
   "source": [
    "train_loader = torch.utils.data.DataLoader(train_data, \n",
    "                                           batch_size=batch_size,\n",
    "                                           shuffle=True, \n",
    "                                           num_workers=num_workers)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "d08219f4",
   "metadata": {},
   "outputs": [],
   "source": [
    "test_loader = torch.utils.data.DataLoader(test_data, \n",
    "                                          batch_size=batch_size,\n",
    "                                          shuffle=False, \n",
    "                                          num_workers=num_workers)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "472d11be",
   "metadata": {},
   "source": [
    "`train_loader`의 1개 배치의 shape 출력"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "ec40f259",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(torch.Size([32, 1, 28, 28]), torch.Size([32]))"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 1개의 배치 추출 후 Image, label의 shape 출력\n",
    "img, lbl = next(iter(train_loader))\n",
    "img.shape, lbl.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d180f133",
   "metadata": {},
   "source": [
    "배치사이즈인 32가 가장 첫번째 dimension에 출력되고, 그 뒤로 채널, 세로, 가로 순서로 출력이 됩니다.\n",
    "\n",
    "즉, greyscale 의 `28 X 28` 이미지 `32장`이 1개의 배치로 구성이 되어 있습니다."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "7b8369f6",
   "metadata": {},
   "source": [
    "## 모델 정의"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f7d88005",
   "metadata": {},
   "source": [
    "CUDA 설정이 되어 있다면 `cuda`를! 그렇지 않다면 `cpu`로 학습합니다.\n",
    "\n",
    "(제 PC에는 GPU가 2대 있어서 `cuda:0`로 GPU 장비의 index를 지정해 주었습니다. 만약 다른 장비를 사용하고 싶다면 `cuda:1` 이런식으로 지정해 주면 됩니다)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "25cb458b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cuda:0\n"
     ]
    }
   ],
   "source": [
    "# device 설정 (cuda:0 혹은 cpu)\n",
    "device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")\n",
    "print(device)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f89d3d4a",
   "metadata": {},
   "source": [
    "아래의 모델은 DNN으로 구성하였습니다. 추후, 모델 부분을 CNN이나 pre-trained model로 교체할 수 있습니다."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "ff786d7e",
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "import torch.optim as optim\n",
    "\n",
    "\n",
    "class DNNModel(nn.Module):\n",
    "    def __init__(self):\n",
    "        super(DNNModel, self).__init__()\n",
    "        self.fc1 = nn.Linear(28*28, 512)\n",
    "        self.fc2 = nn.Linear(512, 128)\n",
    "        self.fc3 = nn.Linear(128, 32)\n",
    "        self.output = nn.Linear(32, 10)\n",
    "    \n",
    "    def forward(self, x):\n",
    "        x = x.view(-1, 28*28)\n",
    "        x = F.relu(self.fc1(x))\n",
    "        x = F.relu(self.fc2(x))\n",
    "        x = F.relu(self.fc3(x))\n",
    "        x = self.output(x)\n",
    "        return x     "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "b1b7caac",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "DNNModel(\n",
       "  (fc1): Linear(in_features=784, out_features=512, bias=True)\n",
       "  (fc2): Linear(in_features=512, out_features=128, bias=True)\n",
       "  (fc3): Linear(in_features=128, out_features=32, bias=True)\n",
       "  (output): Linear(in_features=32, out_features=10, bias=True)\n",
       ")"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model = DNNModel() # Model 생성\n",
    "model.to(device)   # device 에 로드 (cpu or cuda)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "0e8b2270",
   "metadata": {},
   "outputs": [],
   "source": [
    "# 옵티마이저를 정의합니다. 옵티마이저에는 model.parameters()를 지정해야 합니다.\n",
    "optimizer = optim.Adam(model.parameters(), lr=0.0005)\n",
    "\n",
    "# 손실함수(loss function)을 지정합니다. Multi-Class Classification 이기 때문에 CrossEntropy 손실을 지정하였습니다.\n",
    "loss_fn = nn.CrossEntropyLoss()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "82586144",
   "metadata": {},
   "source": [
    "## 훈련(Train)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "18b4a1f8",
   "metadata": {},
   "outputs": [],
   "source": [
    "from tqdm import tqdm  # Progress Bar 출력"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "d229193f",
   "metadata": {},
   "outputs": [],
   "source": [
    "def model_train(model, data_loader, loss_fn, optimizer, device):\n",
    "    # 모델을 훈련모드로 설정합니다. training mode 일 때 Gradient 가 업데이트 됩니다. 반드시 train()으로 모드 변경을 해야 합니다.\n",
    "    model.train()\n",
    "    \n",
    "    # loss와 accuracy 계산을 위한 임시 변수 입니다. 0으로 초기화합니다.\n",
    "    running_loss = 0\n",
    "    corr = 0\n",
    "    \n",
    "    # 예쁘게 Progress Bar를 출력하면서 훈련 상태를 모니터링 하기 위하여 tqdm으로 래핑합니다.\n",
    "    prograss_bar = tqdm(data_loader)\n",
    "    \n",
    "    # mini-batch 학습을 시작합니다.\n",
    "    for img, lbl in prograss_bar:\n",
    "        # image, label 데이터를 device에 올립니다.\n",
    "        img, lbl = img.to(device), lbl.to(device)\n",
    "        \n",
    "        # 누적 Gradient를 초기화 합니다.\n",
    "        optimizer.zero_grad()\n",
    "        \n",
    "        # Forward Propagation을 진행하여 결과를 얻습니다.\n",
    "        output = model(img)\n",
    "        \n",
    "        # 손실함수에 output, label 값을 대입하여 손실을 계산합니다.\n",
    "        loss = loss_fn(output, lbl)\n",
    "        \n",
    "        # 오차역전파(Back Propagation)을 진행하여 미분 값을 계산합니다.\n",
    "        loss.backward()\n",
    "        \n",
    "        # 계산된 Gradient를 업데이트 합니다.\n",
    "        optimizer.step()\n",
    "        \n",
    "        # output의 max(dim=1)은 max probability와 max index를 반환합니다.\n",
    "        # max probability는 무시하고, max index는 pred에 저장하여 label 값과 대조하여 정확도를 도출합니다.\n",
    "        _, pred = output.max(dim=1)\n",
    "        \n",
    "        # pred.eq(lbl).sum() 은 정확히 맞춘 label의 합계를 계산합니다. item()은 tensor에서 값을 추출합니다.\n",
    "        # 합계는 corr 변수에 누적합니다.\n",
    "        corr += pred.eq(lbl).sum().item()\n",
    "        \n",
    "        # loss 값은 1개 배치의 평균 손실(loss) 입니다. img.size(0)은 배치사이즈(batch size) 입니다.\n",
    "        # loss 와 img.size(0)를 곱하면 1개 배치의 전체 loss가 계산됩니다.\n",
    "        # 이를 누적한 뒤 Epoch 종료시 전체 데이터셋의 개수로 나누어 평균 loss를 산출합니다.\n",
    "        running_loss += loss.item() * img.size(0)\n",
    "        \n",
    "    # 누적된 정답수를 전체 개수로 나누어 주면 정확도가 산출됩니다.\n",
    "    acc = corr / len(data_loader.dataset)\n",
    "    \n",
    "    # 평균 손실(loss)과 정확도를 반환합니다.\n",
    "    # train_loss, train_acc\n",
    "    return running_loss / len(data_loader.dataset), acc"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d4275f37",
   "metadata": {},
   "source": [
    "## 평가(Evaluate)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "1138f413",
   "metadata": {},
   "outputs": [],
   "source": [
    "def model_evaluate(model, data_loader, loss_fn, device):\n",
    "    # model.eval()은 모델을 평가모드로 설정을 바꾸어 줍니다. \n",
    "    # dropout과 같은 layer의 역할 변경을 위하여 evaluation 진행시 꼭 필요한 절차 입니다.\n",
    "    model.eval()\n",
    "    \n",
    "    # Gradient가 업데이트 되는 것을 방지 하기 위하여 반드시 필요합니다.\n",
    "    with torch.no_grad():\n",
    "        # loss와 accuracy 계산을 위한 임시 변수 입니다. 0으로 초기화합니다.\n",
    "        corr = 0\n",
    "        running_loss = 0\n",
    "        \n",
    "        # 배치별 evaluation을 진행합니다.\n",
    "        for img, lbl in data_loader:\n",
    "            # device에 데이터를 올립니다.\n",
    "            img, lbl = img.to(device), lbl.to(device)\n",
    "            \n",
    "            # 모델에 Forward Propagation을 하여 결과를 도출합니다.\n",
    "            output = model(img)\n",
    "            \n",
    "            # output의 max(dim=1)은 max probability와 max index를 반환합니다.\n",
    "            # max probability는 무시하고, max index는 pred에 저장하여 label 값과 대조하여 정확도를 도출합니다.\n",
    "            _, pred = output.max(dim=1)\n",
    "            \n",
    "            # pred.eq(lbl).sum() 은 정확히 맞춘 label의 합계를 계산합니다. item()은 tensor에서 값을 추출합니다.\n",
    "            # 합계는 corr 변수에 누적합니다.\n",
    "            corr += torch.sum(pred.eq(lbl)).item()\n",
    "            \n",
    "            # loss 값은 1개 배치의 평균 손실(loss) 입니다. img.size(0)은 배치사이즈(batch size) 입니다.\n",
    "            # loss 와 img.size(0)를 곱하면 1개 배치의 전체 loss가 계산됩니다.\n",
    "            # 이를 누적한 뒤 Epoch 종료시 전체 데이터셋의 개수로 나누어 평균 loss를 산출합니다.\n",
    "            running_loss += loss_fn(output, lbl).item() * img.size(0)\n",
    "        \n",
    "        # validation 정확도를 계산합니다.\n",
    "        # 누적한 정답숫자를 전체 데이터셋의 숫자로 나누어 최종 accuracy를 산출합니다.\n",
    "        acc = corr / len(data_loader.dataset)\n",
    "        \n",
    "        # 결과를 반환합니다.\n",
    "        # val_loss, val_acc\n",
    "        return running_loss / len(data_loader.dataset), acc"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bac24e4f",
   "metadata": {},
   "source": [
    "## 모델 훈련(training) & 검증"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "2b9a7089",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100% 1875/1875 [00:04<00:00, 445.34it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "val_loss has been improved from inf to 0.42429. Saving Model!\n",
      "epoch 01, loss: 0.55653, acc: 0.80107, val_loss: 0.42429, val_accuracy: 0.84950\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100% 1875/1875 [00:04<00:00, 460.48it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "val_loss has been improved from 0.42429 to 0.38282. Saving Model!\n",
      "epoch 02, loss: 0.37441, acc: 0.86393, val_loss: 0.38282, val_accuracy: 0.86280\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100% 1875/1875 [00:04<00:00, 453.84it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch 03, loss: 0.33538, acc: 0.87767, val_loss: 0.38931, val_accuracy: 0.86290\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100% 1875/1875 [00:04<00:00, 455.05it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "val_loss has been improved from 0.38282 to 0.35129. Saving Model!\n",
      "epoch 04, loss: 0.30863, acc: 0.88853, val_loss: 0.35129, val_accuracy: 0.86890\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100% 1875/1875 [00:04<00:00, 447.22it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "val_loss has been improved from 0.35129 to 0.34075. Saving Model!\n",
      "epoch 05, loss: 0.28994, acc: 0.89215, val_loss: 0.34075, val_accuracy: 0.87460\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100% 1875/1875 [00:04<00:00, 450.35it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch 06, loss: 0.27176, acc: 0.89980, val_loss: 0.36094, val_accuracy: 0.87150\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100% 1875/1875 [00:04<00:00, 443.07it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch 07, loss: 0.25868, acc: 0.90415, val_loss: 0.34475, val_accuracy: 0.87800\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100% 1875/1875 [00:04<00:00, 447.50it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "val_loss has been improved from 0.34075 to 0.32414. Saving Model!\n",
      "epoch 08, loss: 0.24608, acc: 0.90673, val_loss: 0.32414, val_accuracy: 0.88970\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100% 1875/1875 [00:04<00:00, 448.74it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch 09, loss: 0.23623, acc: 0.91130, val_loss: 0.32660, val_accuracy: 0.88340\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100% 1875/1875 [00:04<00:00, 450.33it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch 10, loss: 0.22619, acc: 0.91510, val_loss: 0.32550, val_accuracy: 0.88620\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100% 1875/1875 [00:04<00:00, 450.43it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch 11, loss: 0.21585, acc: 0.91827, val_loss: 0.33025, val_accuracy: 0.88710\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100% 1875/1875 [00:04<00:00, 444.86it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch 12, loss: 0.20716, acc: 0.92187, val_loss: 0.35040, val_accuracy: 0.88430\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100% 1875/1875 [00:04<00:00, 441.60it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch 13, loss: 0.19787, acc: 0.92447, val_loss: 0.35395, val_accuracy: 0.88630\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100% 1875/1875 [00:04<00:00, 445.40it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch 14, loss: 0.18987, acc: 0.92833, val_loss: 0.34877, val_accuracy: 0.88420\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100% 1875/1875 [00:04<00:00, 450.62it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch 15, loss: 0.18462, acc: 0.92942, val_loss: 0.32642, val_accuracy: 0.89690\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100% 1875/1875 [00:04<00:00, 445.37it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch 16, loss: 0.17759, acc: 0.93293, val_loss: 0.36149, val_accuracy: 0.88150\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100% 1875/1875 [00:04<00:00, 446.42it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch 17, loss: 0.17055, acc: 0.93517, val_loss: 0.33631, val_accuracy: 0.89640\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100% 1875/1875 [00:04<00:00, 450.76it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch 18, loss: 0.16626, acc: 0.93688, val_loss: 0.36313, val_accuracy: 0.89320\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100% 1875/1875 [00:04<00:00, 452.99it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch 19, loss: 0.15961, acc: 0.93843, val_loss: 0.38574, val_accuracy: 0.88100\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100% 1875/1875 [00:04<00:00, 450.54it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "epoch 20, loss: 0.15411, acc: 0.94178, val_loss: 0.36963, val_accuracy: 0.89350\n"
     ]
    }
   ],
   "source": [
    "# 최대 Epoch을 지정합니다.\n",
    "num_epochs = 20\n",
    "\n",
    "min_loss = np.inf\n",
    "\n",
    "# Epoch 별 훈련 및 검증을 수행합니다.\n",
    "for epoch in range(num_epochs):\n",
    "    # Model Training\n",
    "    # 훈련 손실과 정확도를 반환 받습니다.\n",
    "    train_loss, train_acc = model_train(model, train_loader, loss_fn, optimizer, device)\n",
    "\n",
    "    # 검증 손실과 검증 정확도를 반환 받습니다.\n",
    "    val_loss, val_acc = model_evaluate(model, test_loader, loss_fn, device)   \n",
    "    \n",
    "    # val_loss 가 개선되었다면 min_loss를 갱신하고 model의 가중치(weights)를 저장합니다.\n",
    "    if val_loss < min_loss:\n",
    "        print(f'[INFO] val_loss has been improved from {min_loss:.5f} to {val_loss:.5f}. Saving Model!')\n",
    "        min_loss = val_loss\n",
    "        torch.save(model.state_dict(), 'DNNModel.pth')\n",
    "    \n",
    "    # Epoch 별 결과를 출력합니다.\n",
    "    print(f'epoch {epoch+1:02d}, loss: {train_loss:.5f}, acc: {train_acc:.5f}, val_loss: {val_loss:.5f}, val_accuracy: {val_acc:.5f}')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "514ee1de",
   "metadata": {},
   "source": [
    "## 저장한 가중치 로드 후 검증 성능 측정"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "191eaf7b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<All keys matched successfully>"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 모델에 저장한 가중치를 로드합니다.\n",
    "model.load_state_dict(torch.load('DNNModel.pth'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "aa1db76e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "evaluation loss: 0.32414, evaluation accuracy: 0.88970\n"
     ]
    }
   ],
   "source": [
    "# 최종 검증 손실(validation loss)와 검증 정확도(validation accuracy)를 산출합니다.\n",
    "final_loss, final_acc = model_evaluate(model, test_loader, loss_fn, device)\n",
    "print(f'evaluation loss: {final_loss:.5f}, evaluation accuracy: {final_acc:.5f}')"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.7.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}