cnn_mnist_classification.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "825c742f-0ae7-4e3f-99b7-e8420c9c5152",
   "metadata": {
    "canvas": {
     "comments": [],
     "componentType": "CodeCell",
     "copiedOriginId": null,
     "hasOpendPDC": false,
     "headerColor": 0,
     "id": "fb702cab-f44c-4821-8dbf-9d9c24e01df5",
     "isComponent": false,
     "meta_id": "7c86f5bb-8057-4271-8f68-6ec787835b4e",
     "name": "",
     "parents": [],
     "preDefined": null
    }
   },
   "source": [
    "# MNIST classification with PyTorch\n",
    "\n",
    "## Fully connected layers vs Convolutional Neural Network (CNN)\n",
    "\n",
    "Adapted from: https://github.com/Atcold/pytorch-Deep-Learning/blob/master/06-convnet.ipynb"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "844dfc02-cd39-4058-89e3-32ae25f8c21d",
   "metadata": {
    "canvas": {
     "comments": [],
     "componentType": "CodeCell",
     "copiedOriginId": null,
     "hasOpendPDC": false,
     "headerColor": 0,
     "id": "811c6c65-84ca-4617-aa07-d99040e51aed",
     "isComponent": false,
     "meta_id": "56fd98fc-18cd-44d1-8f01-24bccf1dc503",
     "name": "",
     "parents": [],
     "preDefined": null
    }
   },
   "source": [
    "## Cell color\n",
    "- Pink: Related to FCN (Fully-connected network) model\n",
    "- Yellow: Related to CNN (Convolutional Neural network) model\n",
    "- Skyblue: Related to visualization"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a8c0bad2-bec9-4be9-91d4-31135a2bbfd8",
   "metadata": {
    "canvas": {
     "comments": [],
     "componentType": "CodeCell",
     "copiedOriginId": null,
     "hasOpendPDC": false,
     "headerColor": 0,
     "id": "faf50dcf-5099-418e-91cf-4258ead57c7f",
     "isComponent": false,
     "meta_id": "d9586192-8fe1-4ebe-bf04-6ce2bdf40a02",
     "name": "",
     "parents": [],
     "preDefined": null
    }
   },
   "source": [
    "### Required Python Packages\n",
    "- `torch`\n",
    "- `torchvision`\n",
    "- `matplotlib`\n",
    "- `tqdm`\n",
    "\n",
    "Run the following cell to install the packages. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "2c835093-5db3-461c-a06d-7973904ddf5e",
   "metadata": {
    "canvas": {
     "comments": [],
     "componentType": "CodeCell",
     "copiedOriginId": null,
     "hasOpendPDC": false,
     "headerColor": 0,
     "id": "a723a6b4-76fc-4188-92ba-3a0ca990a013",
     "isComponent": false,
     "meta_id": "2c835093-5db3-461c-a06d-7973904ddf5e",
     "name": "",
     "parents": [],
     "preDefined": null
    }
   },
   "outputs": [],
   "source": [
    "#\n",
    "# Required Packages\n",
    "#   Run this cell to install required packages\n",
    "#\n",
    "!pip install \"torch>=1.9+cpu\"\n",
    "!pip install \"torchvision>=0.10.1+cpu\"\n",
    "!pip install \"matplotlib>=3.5\"\n",
    "!pip install \"tqdm>=4.64\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "0fc9bc91-150f-4753-a15e-9178fe1fea54",
   "metadata": {
    "canvas": {
     "comments": [],
     "componentType": "CodeCell",
     "copiedOriginId": null,
     "hasOpendPDC": false,
     "headerColor": 0,
     "id": "341de75f-bacb-4af8-9269-f974f4547ca8",
     "isComponent": true,
     "meta_id": "0fc9bc91-150f-4753-a15e-9178fe1fea54",
     "name": "Import packages",
     "parents": [],
     "preDefined": null
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "from torch.utils.data import DataLoader\n",
    "from torchvision import datasets, transforms\n",
    "from tqdm import tqdm"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "798532a2-3909-41ec-a2ff-f1ccc070870c",
   "metadata": {
    "canvas": {
     "comments": [],
     "componentType": "CodeCell",
     "copiedOriginId": null,
     "hasOpendPDC": false,
     "headerColor": 0,
     "id": "f0e2a43a-d5bd-4f6c-bb5b-8bfa3fae2dd9",
     "isComponent": false,
     "meta_id": "966666b0-62e8-4a7c-b8c7-a255896f6dcc",
     "name": "",
     "parents": [],
     "preDefined": null
    }
   },
   "source": [
    "### Load the Dataset (MNIST)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "9de0ed7d-df92-4ed5-9edc-1cb789eec6d5",
   "metadata": {
    "canvas": {
     "comments": [],
     "componentType": "CodeCell",
     "copiedOriginId": null,
     "hasOpendPDC": false,
     "headerColor": 0,
     "id": "c1c9d3af-d77f-40cd-b145-3976174d3032",
     "isComponent": true,
     "meta_id": "9de0ed7d-df92-4ed5-9edc-1cb789eec6d5",
     "name": "Load MNIST dataset",
     "parents": [
      {
       "id": "341de75f-bacb-4af8-9269-f974f4547ca8",
       "name": "Import packages"
      }
     ],
     "preDefined": null
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/jovyan/.local/conda/lib/python3.8/site-packages/torchvision/datasets/mnist.py:498: UserWarning: The given NumPy array is not writeable, and PyTorch does not support non-writeable tensors. This means you can write to the underlying (supposedly non-writeable) NumPy array using the tensor. You may want to copy the array to protect its data or make it writeable before converting it to a tensor. This type of warning will be suppressed for the rest of this program. (Triggered internally at  ../torch/csrc/utils/tensor_numpy.cpp:180.)\n",
      "  return torch.from_numpy(parsed.astype(m[2], copy=False)).view(*s)\n"
     ]
    }
   ],
   "source": [
    "input_size = 28 * 28  # images are 28x28 pixels\n",
    "output_size = 10  # there are 10 classes\n",
    "\n",
    "train_loader = DataLoader(\n",
    "    datasets.MNIST(\n",
    "        \"mnist\",\n",
    "        train=True,\n",
    "        download=True,\n",
    "        transform=transforms.Compose([transforms.ToTensor(), transforms.Normalize((0.1307,), (0.3081,))]),\n",
    "    ),\n",
    "    batch_size=256,\n",
    "    shuffle=True,\n",
    ")\n",
    "\n",
    "test_loader = DataLoader(\n",
    "    datasets.MNIST(\n",
    "        \"mnist\",\n",
    "        train=False,\n",
    "        download=True,\n",
    "        transform=transforms.Compose([transforms.ToTensor(), transforms.Normalize((0.1307,), (0.3081,))]),\n",
    "    ),\n",
    "    batch_size=1000,\n",
    "    shuffle=False,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "f0fdb60f-e4c3-4446-a472-b95d4d11bf9b",
   "metadata": {
    "canvas": {
     "comments": [],
     "componentType": "CodeCell",
     "copiedOriginId": null,
     "hasOpendPDC": false,
     "headerColor": 5,
     "id": "ee9b8b42-0765-4acc-af7e-fccc4a70491f",
     "isComponent": true,
     "meta_id": "f0fdb60f-e4c3-4446-a472-b95d4d11bf9b",
     "name": "Show some images",
     "parents": [
      {
       "id": "c1c9d3af-d77f-40cd-b145-3976174d3032",
       "name": "Load MNIST dataset"
      }
     ],
     "preDefined": null
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1152x432 with 10 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# show some images\n",
    "plt.figure(figsize=(16, 6))\n",
    "for i in range(10):\n",
    "    plt.subplot(2, 5, i + 1)\n",
    "    image, _ = train_loader.dataset.__getitem__(i)\n",
    "    plt.imshow(image.squeeze().numpy())\n",
    "    plt.axis(\"off\");"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f62a44cf-6cbe-41c9-8d0e-e4ed17c91a7f",
   "metadata": {
    "canvas": {
     "comments": [],
     "componentType": "CodeCell",
     "copiedOriginId": null,
     "hasOpendPDC": false,
     "headerColor": 0,
     "id": "a298fe26-4ff8-4c03-89c6-a1351a6253d2",
     "isComponent": false,
     "meta_id": "3dafeeab-f2ae-49bb-9844-749281590f76",
     "name": "",
     "parents": [],
     "preDefined": null
    }
   },
   "source": [
    "### Train and test functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "abe73dae-f8a5-4d36-9fbc-c6cb5e51295e",
   "metadata": {
    "canvas": {
     "comments": [],
     "componentType": "CodeCell",
     "copiedOriginId": null,
     "hasOpendPDC": false,
     "headerColor": 0,
     "id": "e9f3371a-e347-4564-80c1-f4e1a17da0d3",
     "isComponent": true,
     "meta_id": "abe73dae-f8a5-4d36-9fbc-c6cb5e51295e",
     "name": "Define train and test function",
     "parents": [
      {
       "id": "c1c9d3af-d77f-40cd-b145-3976174d3032",
       "name": "Load MNIST dataset"
      }
     ],
     "preDefined": null
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")\n",
    "criterion = nn.CrossEntropyLoss()\n",
    "accuracy_list = []\n",
    "\n",
    "\n",
    "def train(epoch, model, optimizer):\n",
    "    model.train()\n",
    "    print(f\"Train Epoch {epoch}\")\n",
    "\n",
    "    pbar = tqdm(train_loader)\n",
    "    for batch_idx, (data, target) in enumerate(pbar):\n",
    "        # send to device\n",
    "        data, target = data.to(device), target.to(device)\n",
    "\n",
    "        # permute pixels\n",
    "        data = data.view(-1, 28 * 28)\n",
    "        data = data.view(-1, 1, 28, 28)\n",
    "\n",
    "        optimizer.zero_grad()\n",
    "        output = model(data)\n",
    "        loss = criterion(output, target)\n",
    "        loss.backward()\n",
    "        optimizer.step()\n",
    "\n",
    "        pbar.set_description(f\"loss: {loss.item():.4f}\")\n",
    "\n",
    "\n",
    "def test(model, data_loader):\n",
    "    model.eval()\n",
    "    test_loss = 0\n",
    "    correct = 0\n",
    "\n",
    "    with torch.no_grad():\n",
    "        for data, target in data_loader:\n",
    "            # send to device\n",
    "            data, target = data.to(device), target.to(device)\n",
    "\n",
    "            # permute pixels\n",
    "            data = data.view(-1, 28 * 28)\n",
    "            data = data.view(-1, 1, 28, 28)\n",
    "            output = model(data)\n",
    "            test_loss += criterion(output, target)\n",
    "            pred = output.data.max(1, keepdim=True)[1]  # get the index of the max log-probability\n",
    "            correct += pred.eq(target.data.view_as(pred)).cpu().sum().item()\n",
    "\n",
    "    test_loss /= len(data_loader)\n",
    "    accuracy = 100.0 * correct / len(data_loader.dataset)\n",
    "    return test_loss, accuracy\n",
    "\n",
    "\n",
    "# get number of parameters in the model\n",
    "def get_n_params(model):\n",
    "    return sum(p.numel() for p in model.parameters() if p.requires_grad)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "fff47fe1-adcd-4d8a-8524-bf5fafb28d85",
   "metadata": {
    "canvas": {
     "comments": [],
     "componentType": "CodeCell",
     "copiedOriginId": null,
     "hasOpendPDC": false,
     "headerColor": 0,
     "id": "768dc3ec-38d3-4174-8e2d-0cd0eab16883",
     "isComponent": false,
     "meta_id": "88aa0b10-382f-4f51-bd35-0b63fc877692",
     "name": "",
     "parents": [],
     "preDefined": null
    }
   },
   "source": [
    "### CNN model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "fcc627b7-2f68-4432-8468-ce664704f7cc",
   "metadata": {
    "canvas": {
     "comments": [],
     "componentType": "CodeCell",
     "copiedOriginId": null,
     "hasOpendPDC": false,
     "headerColor": 3,
     "id": "0e7afb99-966a-4372-ab33-d8579b44fa4e",
     "isComponent": true,
     "meta_id": "fcc627b7-2f68-4432-8468-ce664704f7cc",
     "name": "Define CNN model",
     "parents": [
      {
       "id": "e9f3371a-e347-4564-80c1-f4e1a17da0d3",
       "name": "Define train and test function"
      }
     ],
     "preDefined": null
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "class CNN(nn.Module):\n",
    "    def __init__(self):\n",
    "        super(CNN, self).__init__()\n",
    "        self.conv1 = nn.Conv2d(in_channels=1, out_channels=cnn_hidden_channels, kernel_size=cnn_kernel_size)\n",
    "        self.conv2 = nn.Conv2d(cnn_hidden_channels, cnn_hidden_channels, kernel_size=cnn_kernel_size)\n",
    "        self.fc1 = nn.Linear(cnn_hidden_channels * 4 * 4, 32)\n",
    "        self.fc2 = nn.Linear(32, 10)\n",
    "\n",
    "    def forward(self, x):\n",
    "        x = self.conv1(x)\n",
    "        x = F.relu(x)\n",
    "        x = F.max_pool2d(x, kernel_size=2)\n",
    "        x = self.conv2(x)\n",
    "        x = F.relu(x)\n",
    "        x = F.max_pool2d(x, kernel_size=2)\n",
    "        x = x.view(-1, cnn_hidden_channels * 4 * 4)\n",
    "        x = self.fc1(x)\n",
    "        x = F.relu(x)\n",
    "        x = self.fc2(x)\n",
    "        return x"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c0cfdabf-bd6d-4332-94d7-83fe12695584",
   "metadata": {
    "canvas": {
     "comments": [],
     "componentType": "CodeCell",
     "copiedOriginId": null,
     "hasOpendPDC": false,
     "headerColor": 0,
     "id": "67675843-5e88-41dd-b3e4-0fd8a69f0abb",
     "isComponent": false,
     "meta_id": "afd030ef-d19f-48e4-bea0-c59bb61057aa",
     "name": "",
     "parents": [],
     "preDefined": null
    }
   },
   "source": [
    "### FCN (Fully Connected Network) model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "e256ca5a-7ab8-4add-882d-6826eb012e87",
   "metadata": {
    "canvas": {
     "comments": [],
     "componentType": "CodeCell",
     "copiedOriginId": null,
     "hasOpendPDC": false,
     "headerColor": 9,
     "id": "0e159b4a-5123-486a-90f2-cb62ceb58095",
     "isComponent": true,
     "meta_id": "e256ca5a-7ab8-4add-882d-6826eb012e87",
     "name": "Define FCN model",
     "parents": [
      {
       "id": "e9f3371a-e347-4564-80c1-f4e1a17da0d3",
       "name": "Define train and test function"
      }
     ],
     "preDefined": null
    },
    "tags": []
   },
   "outputs": [],
   "source": [
    "class FCN(nn.Module):\n",
    "    def __init__(self, input_size):\n",
    "        super(FCN, self).__init__()\n",
    "        self.input_size = input_size\n",
    "        self.network = nn.Sequential(\n",
    "            nn.Linear(input_size, fcn_hidden),\n",
    "            nn.ReLU(inplace=False),\n",
    "            nn.Linear(fcn_hidden, fcn_hidden),\n",
    "            nn.ReLU(inplace=False),\n",
    "            nn.Linear(fcn_hidden, 10),\n",
    "        )\n",
    "\n",
    "    def forward(self, x):\n",
    "        x = x.view(-1, self.input_size)\n",
    "        return self.network(x)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "85c5dbbb-532a-47f0-846d-7b7aef911881",
   "metadata": {
    "canvas": {
     "comments": [],
     "componentType": "CodeCell",
     "copiedOriginId": null,
     "hasOpendPDC": false,
     "headerColor": 0,
     "id": "a2e18508-101e-4f6c-927b-a1175e5b0244",
     "isComponent": false,
     "meta_id": "695c7b84-0c69-4f3f-91f6-fbc995b565fb",
     "name": "",
     "parents": [],
     "preDefined": null
    }
   },
   "source": [
    "### Train and test CNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "d4086386-af5d-43a5-b3f7-a20388d3b573",
   "metadata": {
    "canvas": {
     "comments": [],
     "componentType": "CodeCell",
     "copiedOriginId": null,
     "hasOpendPDC": false,
     "headerColor": 3,
     "id": "01d4c6ec-e129-4397-b816-402c0b292d88",
     "isComponent": true,
     "meta_id": "d4086386-af5d-43a5-b3f7-a20388d3b573",
     "name": "Train CNN",
     "parents": [
      {
       "id": "0e7afb99-966a-4372-ab33-d8579b44fa4e",
       "name": "Define CNN model"
      }
     ],
     "preDefined": null
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of parameters: 6274\n",
      "Train Epoch 1\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "loss: 0.1957: 100%|██████████| 235/235 [00:40<00:00,  5.84it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train loss: 0.1825, Train accuracy: 94.60%\n",
      "Test loss: 0.1658, Test accuracy: 95.00%\n"
     ]
    }
   ],
   "source": [
    "model_cnn = CNN()\n",
    "model_cnn.to(device)\n",
    "optimizer = torch.optim.Adam(model_cnn.parameters(), lr=1e-3)\n",
    "print(\"Number of parameters: {}\".format(get_n_params(model_cnn)))\n",
    "\n",
    "for epoch in range(epochs):\n",
    "    train(epoch + 1, model_cnn, optimizer)\n",
    "    train_loss, cnn_train_acc = test(model_cnn, train_loader)\n",
    "    print(f\"Train loss: {train_loss:.4f}, Train accuracy: {cnn_train_acc:.2f}%\")\n",
    "    test_loss, cnn_test_acc = test(model_cnn, test_loader)\n",
    "    print(f\"Test loss: {test_loss:.4f}, Test accuracy: {cnn_test_acc:.2f}%\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c758e55e-1cbf-4378-9fdd-ce88c0d4ae46",
   "metadata": {
    "canvas": {
     "comments": [],
     "componentType": "CodeCell",
     "copiedOriginId": null,
     "hasOpendPDC": false,
     "headerColor": 0,
     "id": "1af576ec-c96a-4de7-b61f-da0b05ed50a2",
     "isComponent": false,
     "meta_id": "109aaca7-07b7-4f94-b59e-90c6dfee7f96",
     "name": "",
     "parents": [],
     "preDefined": null
    }
   },
   "source": [
    "### Train and test FCN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "f73fcf6d-bf2f-4fc4-bd4c-57bf368df5db",
   "metadata": {
    "canvas": {
     "comments": [],
     "componentType": "CodeCell",
     "copiedOriginId": null,
     "hasOpendPDC": false,
     "headerColor": 9,
     "id": "05d3595a-0216-4508-ab71-b1c3fc00940c",
     "isComponent": true,
     "meta_id": "f73fcf6d-bf2f-4fc4-bd4c-57bf368df5db",
     "name": "Train FCN",
     "parents": [
      {
       "id": "0e159b4a-5123-486a-90f2-cb62ceb58095",
       "name": "Define FCN model"
      }
     ],
     "preDefined": null
    },
    "tags": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Number of parameters: 6442\n",
      "Train Epoch 1\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "loss: 0.6298: 100%|██████████| 235/235 [00:31<00:00,  7.54it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train loss: 0.6959, Train accuracy: 78.56%\n",
      "Test loss: 0.6915, Test accuracy: 78.79%\n"
     ]
    }
   ],
   "source": [
    "model_fcn = FCN(28 * 28)\n",
    "model_fcn.to(device)\n",
    "optimizer = torch.optim.Adam(model_fcn.parameters(), lr=1e-3)\n",
    "print(\"Number of parameters: {}\".format(get_n_params(model_fcn)))\n",
    "\n",
    "for epoch in range(epochs):\n",
    "    train(epoch + 1, model_fcn, optimizer)\n",
    "    train_loss, fcn_train_acc = test(model_fcn, train_loader)\n",
    "    print(f\"Train loss: {train_loss:.4f}, Train accuracy: {fcn_train_acc:.2f}%\")\n",
    "    test_loss, fcn_test_acc = test(model_fcn, test_loader)\n",
    "    print(f\"Test loss: {test_loss:.4f}, Test accuracy: {fcn_test_acc:.2f}%\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f339db04-ec79-486f-8b24-feb1a32fedc1",
   "metadata": {
    "canvas": {
     "comments": [],
     "componentType": "CodeCell",
     "copiedOriginId": null,
     "hasOpendPDC": false,
     "headerColor": 0,
     "id": "8143115c-9975-42d6-8329-f21ceef65b1a",
     "isComponent": false,
     "meta_id": "3dd27c72-2327-4f98-bfa2-1f4cb06c2ece",
     "name": "",
     "parents": [],
     "preDefined": null
    }
   },
   "source": [
    "### Visualize results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "9e9da9b0-22fd-4acf-8d18-f6a874388a71",
   "metadata": {
    "canvas": {
     "comments": [],
     "componentType": "CodeCell",
     "copiedOriginId": null,
     "hasOpendPDC": false,
     "headerColor": 5,
     "id": "1286907d-29ae-47f7-8c32-2e2aae713ea6",
     "isComponent": true,
     "meta_id": "9e9da9b0-22fd-4acf-8d18-f6a874388a71",
     "name": "Visualize FCN result",
     "parents": [
      {
       "id": "05d3595a-0216-4508-ab71-b1c3fc00940c",
       "name": "Train FCN"
      }
     ],
     "preDefined": null
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x864 with 36 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12, 12))\n",
    "for i in range(36):\n",
    "    plt.subplot(6, 6, i + 1)\n",
    "    image, label = test_loader.dataset.__getitem__(i)\n",
    "    pred = model_fcn(image)\n",
    "    _, pred = pred.max(dim=1)\n",
    "    color = \"g\" if label == pred.item() else \"r\"\n",
    "    plt.title(f\"Ans: {label}, Pred: {pred.item()}\", color=color)\n",
    "    plt.imshow(image.squeeze().numpy())\n",
    "    plt.axis(\"off\");"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "31964213-84d1-45ac-81f0-ca4cbdc77f7c",
   "metadata": {
    "canvas": {
     "comments": [],
     "componentType": "CodeCell",
     "copiedOriginId": null,
     "hasOpendPDC": false,
     "headerColor": 5,
     "id": "157c99d5-cc73-43e2-b2c2-b2e7dc0584c7",
     "isComponent": true,
     "meta_id": "31964213-84d1-45ac-81f0-ca4cbdc77f7c",
     "name": "Visualize CNN result",
     "parents": [
      {
       "id": "01d4c6ec-e129-4397-b816-402c0b292d88",
       "name": "Train CNN"
      }
     ],
     "preDefined": null
    },
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x864 with 36 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12, 12))\n",
    "for i in range(36):\n",
    "    plt.subplot(6, 6, i + 1)\n",
    "    image, label = test_loader.dataset.__getitem__(i)\n",
    "    pred = model_cnn(image.unsqueeze(0))\n",
    "    _, pred = pred.max(dim=1)\n",
    "    color = \"g\" if label == pred.item() else \"r\"\n",
    "    plt.title(f\"Ans: {label}, Pred: {pred.item()}\", color=color)\n",
    "    plt.imshow(image.squeeze().numpy())\n",
    "    plt.axis(\"off\");"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "53b8c628-c60e-4d9f-ae9d-a6a39b1d2132",
   "metadata": {
    "canvas": {
     "comments": [],
     "componentType": "CodeCell",
     "copiedOriginId": null,
     "hasOpendPDC": false,
     "headerColor": 0,
     "id": "7f6e7b88-4fda-4663-94f6-042d43a87e45",
     "isComponent": false,
     "meta_id": "d25fbb6c-9da1-40bf-b44c-3ac0a3ef7057",
     "name": "",
     "parents": [],
     "preDefined": null
    }
   },
   "source": [
    "### Compare the performance of FCN and CNN, which is better?"
   ]
  }
 ],
 "metadata": {
  "canvas": {
   "parameters": [
    {
     "name": "fcn_hidden",
     "type": "int",
     "value": "8"
    },
    {
     "name": "cnn_hidden_channels",
     "type": "int",
     "value": "8"
    },
    {
     "name": "cnn_kernel_size",
     "type": "int",
     "value": "5"
    },
    {
     "name": "epochs",
     "type": "int",
     "value": "1"
    }
   ]
  },
  "kernelspec": {
   "display_name": "Python 3 (Link)",
   "language": "python",
   "name": "mrx_link"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}