Colab_Py37_CIFAR10.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Colab_Py37_CIFAR10.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "authorship_tag": "ABX9TyNck+idJMGAdukCHpYyCIhl",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU",
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "4a9994d26cd746169b5740b7372dc35f": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HBoxModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_f6addf9985ac49ee98e88bdd575f2213",
              "IPY_MODEL_222a533302d04cfb8995eff1a59a96e1",
              "IPY_MODEL_1fa7378c119f4b629ecaa63d2e317f6f"
            ],
            "layout": "IPY_MODEL_b96c4ef9195a4b73a1a6c785d807c844"
          }
        },
        "f6addf9985ac49ee98e88bdd575f2213": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_a2e232f7c4fc470ba46747ecb3d0bf9d",
            "placeholder": "​",
            "style": "IPY_MODEL_046d5c6eaaab41b5840a86709e55dd00",
            "value": ""
          }
        },
        "222a533302d04cfb8995eff1a59a96e1": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "FloatProgressModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_5e441e1957fa4c18920d90a6b2c50b5d",
            "max": 170498071,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_9dcb0d34999843a681b9ddfc2388dac8",
            "value": 170498071
          }
        },
        "1fa7378c119f4b629ecaa63d2e317f6f": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "HTMLModel",
          "model_module_version": "1.5.0",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_187a3b7697b74c66bdc516bcf1d72822",
            "placeholder": "​",
            "style": "IPY_MODEL_c77653a6d45347b296ebea2cc6fb71d6",
            "value": " 170499072/? [00:06&lt;00:00, 27944737.91it/s]"
          }
        },
        "b96c4ef9195a4b73a1a6c785d807c844": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "a2e232f7c4fc470ba46747ecb3d0bf9d": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "046d5c6eaaab41b5840a86709e55dd00": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "5e441e1957fa4c18920d90a6b2c50b5d": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "9dcb0d34999843a681b9ddfc2388dac8": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "ProgressStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "187a3b7697b74c66bdc516bcf1d72822": {
          "model_module": "@jupyter-widgets/base",
          "model_name": "LayoutModel",
          "model_module_version": "1.2.0",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "c77653a6d45347b296ebea2cc6fb71d6": {
          "model_module": "@jupyter-widgets/controls",
          "model_name": "DescriptionStyleModel",
          "model_module_version": "1.5.0",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        }
      }
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/yeahxne/Deep_Learning_2022/blob/main/Colab_Py37_CIFAR10.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "ffLdcWV6nqN6"
      },
      "outputs": [],
      "source": [
        "import numpy as np\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "import torch.optim as optim\n",
        "import torch.nn.init as init\n",
        "import torchvision.datasets as dset\n",
        "import torchvision.transforms as transforms\n",
        "from torch.utils.data import DataLoader\n",
        "from torch.autograd import Variable\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "# (8)learning rate decay\n",
        "from torch.optim import lr_scheduler\n",
        "\n",
        "batch_size=16\n",
        "learning_rate=0.001\n",
        "num_epoch=30"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "cifar_train=dset.CIFAR10(\"CIFAR10/\",train=True, transform=transforms.ToTensor(), target_transform=None, download=True)\n",
        "cifar_test=dset.CIFAR10(\"CIFAR10/\",train=False, transform=transforms.ToTensor(), target_transform=None, download=True)\n",
        "\n",
        "# (2) data augmentation \n",
        "#cifar_train=dset.CIFAR10(\"CIFAR10/\", train=True,\n",
        "#                         transform=transforms.Compose([\n",
        "#                            transforms.Resize(36),  \n",
        "#                            transforms.CenterCrop(32),  \n",
        "#                            transforms.RandomHorizontalFlip(),  \n",
        "#                            transforms.Lambda(lambda x: x.rotate(90)),\n",
        "#                            transforms.ToTensor()\n",
        "#                         ]))\n",
        "\n",
        "# (4) Data Normalization \n",
        "cifar_train=dset.CIFAR10(\"CIFAR10/\", train=True,\n",
        "                        transform=transforms.Compose([\n",
        "                           transforms.ToTensor(),\n",
        "                           transforms.Normalize(mean=(0.5,0.5,0.5), std=(0.5,0.5,0.5)),\n",
        "                        ])\n",
        "                       , target_transform=None, download=False)\n",
        "\n",
        "cifar_test=dset.CIFAR10(\"CIFAR10/\", train=False,\n",
        "                        transform=transforms.Compose([\n",
        "                           transforms.ToTensor(),\n",
        "                           transforms.Normalize(mean=(0.5,0.5,0.5), std=(0.5,0.5,0.5)),\n",
        "                        ])\n",
        "                        , target_transform=None, download=False)"
      ],
      "metadata": {
        "id": "B46B06lWntDd",
        "outputId": "e1667c8d-26a4-4445-cb96-2044346fb5e9",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 103,
          "referenced_widgets": [
            "4a9994d26cd746169b5740b7372dc35f",
            "f6addf9985ac49ee98e88bdd575f2213",
            "222a533302d04cfb8995eff1a59a96e1",
            "1fa7378c119f4b629ecaa63d2e317f6f",
            "b96c4ef9195a4b73a1a6c785d807c844",
            "a2e232f7c4fc470ba46747ecb3d0bf9d",
            "046d5c6eaaab41b5840a86709e55dd00",
            "5e441e1957fa4c18920d90a6b2c50b5d",
            "9dcb0d34999843a681b9ddfc2388dac8",
            "187a3b7697b74c66bdc516bcf1d72822",
            "c77653a6d45347b296ebea2cc6fb71d6"
          ]
        }
      },
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz to CIFAR10/cifar-10-python.tar.gz\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "  0%|          | 0/170498071 [00:00<?, ?it/s]"
            ],
            "application/vnd.jupyter.widget-view+json": {
              "version_major": 2,
              "version_minor": 0,
              "model_id": "4a9994d26cd746169b5740b7372dc35f"
            }
          },
          "metadata": {}
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Extracting CIFAR10/cifar-10-python.tar.gz to CIFAR10/\n",
            "Files already downloaded and verified\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "print(\"cifar_train 길이:\", len(cifar_train))\n",
        "print(\"cifar_test 길이:\", len(cifar_test))\n",
        "\n",
        "# 데이터 하나 형태\n",
        "image, label = cifar_train.__getitem__(1)  # 1번째 데이터\n",
        "print(\"image data 형태:\", image.size())\n",
        "print(\"label:\", label)\n",
        "\n",
        "# 그리기\n",
        "img = image.numpy()  # image 타입을 numpy로 변환(3,32,32)\n",
        "\n",
        "# (3,32,32) -> (32,32,3)\n",
        "r, g, b = img[0,:,:], img[1,:,:], img[2,:,:]\n",
        "# img = img.reshape(img.shape[1], img.shape[2], img.shape[0])\n",
        "img2 = np.zeros((img.shape[1], img.shape[2], img.shape[0]))\n",
        "img2[:,:,0], img2[:,:,1], img2[:,:,2] = r, g, b\n",
        "\n",
        "plt.title(\"label: %d\" %label)\n",
        "plt.imshow(img2,interpolation='bicubic')\n",
        "plt.show()"
      ],
      "metadata": {
        "id": "B1pLCxSJnxm-",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 372
        },
        "outputId": "b21530ed-7de1-4fda-d904-19612e90d3b6"
      },
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "cifar_train 길이: 50000\n",
            "cifar_test 길이: 10000\n",
            "image data 형태: torch.Size([3, 32, 32])\n",
            "label: 9\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def ComputeAccr(dloader, imodel):\n",
        "    correct = 0\n",
        "    total = 0\n",
        "    \n",
        "    for j, [imgs, labels] in enumerate(dloader):  # batch_size 만큼\n",
        "        img = Variable(imgs,volatile=True).cuda()  # x\n",
        "        # label = Variable(labels)  # y\n",
        "        label = Variable(labels).cuda()\n",
        "        \n",
        "        output = imodel.forward(img)  # forward prop.\n",
        "        _, output_index = torch.max(output, 1)\n",
        "        \n",
        "        total += label.size(0)\n",
        "        correct += (output_index == label).sum().float()\n",
        "\n",
        "    accr = 100*correct/total\n",
        "    accr = float(accr.cpu().numpy())\n",
        "    print(\"Accuracy of Test Data: {}\".format(accr))\n",
        "    return accr"
      ],
      "metadata": {
        "id": "rgSe57-rqiJE"
      },
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "# === 3. 데이터 로드함수 ===\n",
        "train_loader=torch.utils.data.DataLoader(list(cifar_train)[:], batch_size=batch_size, shuffle=True, num_workers=2, drop_last=True)\n",
        "test_loader=torch.utils.data.DataLoader(cifar_test, batch_size=batch_size, shuffle=False, num_workers=2, drop_last=True)\n",
        "\n",
        "# === 4. 모델 선언 ===\n",
        "class CNN(nn.Module):\n",
        "    def __init__(self):\n",
        "        super(CNN,self).__init__()\n",
        "        self.layer=nn.Sequential(\n",
        "            nn.Conv2d(3,8,3,padding=1),  # 변경x\n",
        "            nn.ELU(),\n",
        "            # nn.PReLU(),\n",
        "            # nn.LeakyReLU(),\n",
        "            nn.Dropout2d(0.1),\n",
        "            nn.BatchNorm2d(8),\n",
        "            nn.Conv2d(8,32,3,padding=1),  # 변경x\n",
        "            nn.ELU(),\n",
        "            # nn.PReLU(),\n",
        "            # nn.LeakyReLU(),\n",
        "            nn.Dropout2d(0.1),\n",
        "            nn.BatchNorm2d(32),\n",
        "            nn.MaxPool2d(2,2),  # 변경x\n",
        "            nn.Conv2d(32,64,3,padding=1),  # 변경x\n",
        "            nn.ELU(),\n",
        "            # nn.PReLU(),\n",
        "            # nn.LeakyReLU(),\n",
        "            nn.Dropout2d(0.3),\n",
        "            nn.BatchNorm2d(64),\n",
        "            nn.MaxPool2d(2,2)  # 변경x\n",
        "        )\n",
        "        self.fc_layer=nn.Sequential(\n",
        "            nn.Linear(64*8*8, 100),\n",
        "            nn.ELU(),\n",
        "            # nn.PReLU(),\n",
        "            # nn.LeakyReLU(),\n",
        "            nn.Dropout2d(0.3),\n",
        "            nn.BatchNorm1d(100),\n",
        "            nn.Linear(100,10)\n",
        "        )\n",
        "\n",
        "        \n",
        "        # (3) weight initialization\n",
        "        for m in self.modules():\n",
        "            if isinstance(m, nn.Conv2d):\n",
        "                init.kaiming_normal(m.weight.data)\n",
        "                # init.kaiming_normal(m.weight.data, nonlinearity='leaky_relu')\n",
        "                m.bias.data.fill_(0)\n",
        "            if isinstance(m, nn.Linear):\n",
        "                init.kaiming_normal(m.weight.data)\n",
        "                # init.kaiming_normal(m.weight.data, nonlinearity='leaky_relu')\n",
        "                m.bias.data.fill_(0)\n",
        "                \n",
        "    def forward(self,x):\n",
        "        out=self.layer(x)\n",
        "        out=out.view(batch_size,-1)\n",
        "        out=self.fc_layer(out)\n",
        "        \n",
        "        return out\n",
        "\n",
        "model=CNN().cuda()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EnMMKRmPqnUM",
        "outputId": "32e62146-37e1-4963-b62a-279165eba062"
      },
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/ipykernel_launcher.py:45: UserWarning: nn.init.kaiming_normal is now deprecated in favor of nn.init.kaiming_normal_.\n",
            "/usr/local/lib/python3.7/dist-packages/ipykernel_launcher.py:49: UserWarning: nn.init.kaiming_normal is now deprecated in favor of nn.init.kaiming_normal_.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# === 5. loss, optimizer ===\n",
        "loss_func=nn.CrossEntropyLoss()\n",
        "# optimizer=torch.optim.SGD(model.parameters(), lr=learning_rate)\n",
        "\n",
        "# (6) Adam optimizer\n",
        "optimizer=torch.optim.Adam(model.parameters(), lr=learning_rate)\n",
        "\n",
        "# (7) learning rate decay\n",
        "scheduler=lr_scheduler.StepLR(optimizer, step_size=20, gamma=0.2)\n",
        "\n",
        "model.train()  # train mode로 바꾸기 -> dropout, batch normalization에 영향을 줌.\n",
        "\n",
        "# ===6. 학습 ===\n",
        "for i in range(num_epoch):\n",
        "    for j,[image, label] in enumerate(train_loader):\n",
        "        x=Variable(image).cuda()\n",
        "        y_=Variable(label).cuda()\n",
        "        \n",
        "        optimizer.zero_grad()\n",
        "        output=model.forward(x)\n",
        "        loss=loss_func(output,y_)\n",
        "        loss.backward()\n",
        "        optimizer.step()\n",
        "        \n",
        "        if j%1000==0:\n",
        "            print(j,loss)\n",
        "\n",
        "model.eval()  # evaluation(test) mode로 바꾸기 -> dropout, batch normalization에 영향을 줌.\n",
        "ComputeAccr(test_loader, model)\n",
        "\n",
        "# 학습된 파라미터 저장\n",
        "netname = './my_net.pkl'\n",
        "torch.save(model,netname, )"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "sSoA2GNiqviW",
        "outputId": "c8f13e0c-59b6-415e-e25b-98aaabd3f16d"
      },
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/torch/nn/functional.py:1320: UserWarning: dropout2d: Received a 2-D input to dropout2d, which is deprecated and will result in an error in a future release. To retain the behavior and silence this warning, please use dropout instead. Note that dropout2d exists to provide channel-wise dropout on inputs with 2 spatial dimensions, a channel dimension, and an optional batch dimension (i.e. 3D or 4D inputs).\n",
            "  warnings.warn(warn_msg)\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "0 tensor(3.1241, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(2.0352, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(1.6969, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(1.7286, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(1.6225, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(1.0582, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(1.2662, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(1.2547, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(1.4869, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(0.9894, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(1.3539, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(1.2913, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(1.0558, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(1.8733, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(1.2904, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(1.1909, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(1.7895, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(2.1052, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(1.2709, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(1.1105, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(1.0961, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(1.8262, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(1.1934, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(1.4729, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(1.2635, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(0.8867, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(1.2157, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(0.9808, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(1.2127, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(1.0979, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(1.8078, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(1.4998, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(1.1230, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(1.5656, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(1.4016, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(1.8738, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(1.2952, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(0.9413, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(1.4802, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(1.4652, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(1.0454, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(1.6221, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(0.9026, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(1.2984, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(1.3272, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(1.4020, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(0.9918, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(1.3531, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(1.6735, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(1.4324, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(1.0070, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(0.8398, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(1.6055, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(0.9927, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(1.0736, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(1.1260, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(0.4958, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(1.2495, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(0.5819, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(1.5680, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(1.1487, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(1.3396, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(1.7288, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(0.8682, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(0.7281, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(1.0532, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(1.4087, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(0.7157, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(0.8985, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(0.9705, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(1.2133, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(1.3882, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(1.3340, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(1.0647, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(0.9102, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(1.2514, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(0.7685, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(1.3427, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(1.2735, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(1.3022, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(1.4829, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(1.2054, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(1.4878, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(0.9554, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(1.5362, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(1.6677, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(1.5097, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(1.0944, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(0.9204, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(0.9422, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(1.0661, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(0.9100, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(1.4305, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(0.7313, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(0.9355, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(0.6863, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(0.9141, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(1.0917, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(1.2704, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(1.3284, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(1.0740, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(0.8475, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(1.0803, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(1.3724, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(0.5019, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(0.9606, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(1.0372, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(1.4062, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(0.9023, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(1.1229, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(1.2453, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(0.7676, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(1.2529, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(0.5091, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(0.6775, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(1.0552, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "0 tensor(1.3239, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "1000 tensor(1.1643, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "2000 tensor(0.8623, device='cuda:0', grad_fn=<NllLossBackward0>)\n",
            "3000 tensor(1.4616, device='cuda:0', grad_fn=<NllLossBackward0>)\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/ipykernel_launcher.py:6: UserWarning: volatile was removed and now has no effect. Use `with torch.no_grad():` instead.\n",
            "  \n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy of Test Data: 76.02999877929688\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "netname = './my_net_final.pkl'\n",
        "model = torch.load(netname)\n",
        "\n",
        "# 성능 확인\n",
        "ComputeAccr(test_loader,model)"
      ],
      "metadata": {
        "id": "zdyef4r7q0ks",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "ff2c2c33-75a1-44f5-cfd0-f77e36a88c7e"
      },
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/ipykernel_launcher.py:6: UserWarning: volatile was removed and now has no effect. Use `with torch.no_grad():` instead.\n",
            "  \n",
            "/usr/local/lib/python3.7/dist-packages/torch/nn/functional.py:1320: UserWarning: dropout2d: Received a 2-D input to dropout2d, which is deprecated and will result in an error in a future release. To retain the behavior and silence this warning, please use dropout instead. Note that dropout2d exists to provide channel-wise dropout on inputs with 2 spatial dimensions, a channel dimension, and an optional batch dimension (i.e. 3D or 4D inputs).\n",
            "  warnings.warn(warn_msg)\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Accuracy of Test Data: 76.69999694824219\n"
          ]
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "76.69999694824219"
            ]
          },
          "metadata": {},
          "execution_count": 7
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        ""
      ],
      "metadata": {
        "id": "lmH5sFJfKVBb"
      },
      "execution_count": 7,
      "outputs": []
    }
  ]
}