MNIST_model_level_3.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "MNIST_model_level_3.ipynb",
      "provenance": [],
      "authorship_tag": "ABX9TyOnlVgYbEmspK2hRjtmt+pT",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/varunkumark1997/EVA5-Assignment/blob/master/MNIST_model_level_3.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "NpafuThxCZxI",
        "colab_type": "text"
      },
      "source": [
        "Target:\n",
        "\n",
        "    Add GAP layer\n",
        "    Change Convolution layer to keep parameters below 10k\n",
        "\n",
        "Results:\n",
        "\n",
        "    Parameters: 9,800\n",
        "    Best Train Accuracy: 99.205\n",
        "    Best Test Accuracy: 99.33\n",
        "\n",
        "Analysis:\n",
        "\n",
        "    We have achieved one of the tasks by bringing the params to under 10k\n",
        "    GAP layer has given us better result than conv with 7x7 kernel as at the last layers all the pixels are important bcz the kernel has extracted this for us,\n",
        "    so GAP helps in generalizing and hence better result\n",
        "    But still to achieve the required accuracy and we are almost there..\n",
        "    We can leverage LR Scheduler for achieving our task."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xpR51_htB408",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from __future__ import print_function\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "import torch.nn.functional as F\n",
        "import torch.optim as optim\n",
        "from torchvision import datasets, transforms"
      ],
      "execution_count": 126,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4DlehbUsCeKC",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "train_transforms = transforms.Compose(\n",
        "                    [\n",
        "                        transforms.RandomRotation((-7.0, 7.0), fill=(1,)),\n",
        "                        transforms.ToTensor(),\n",
        "                        transforms.Normalize((0.1307,), (0.3081,))\n",
        "                    ])\n",
        "                    \n",
        "test_transforms = transforms.Compose([\n",
        "                        transforms.ToTensor(),\n",
        "                        transforms.Normalize((0.1307,), (0.3081,))\n",
        "                    ])"
      ],
      "execution_count": 134,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4-ftP18sChNv",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "train = datasets.MNIST('./data', train=True, download=True, transform=train_transforms)\n",
        "test  = datasets.MNIST('./data', train=False, download=True, transform=test_transforms)"
      ],
      "execution_count": 135,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "91lnEYZfCjK5",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "cb12f8ba-bc9f-42ab-cfce-34cd315785fb"
      },
      "source": [
        "SEED = 1\n",
        "\n",
        "cuda = torch.cuda.is_available()\n",
        "print('is CUDA available', cuda)\n",
        "\n",
        "torch.manual_seed(SEED)\n",
        "\n",
        "if cuda:\n",
        "  torch.cuda.manual_seed(SEED)\n",
        "\n",
        "dataloader_args = dict(shuffle=True, batch_size=128, num_workers=4, pin_memory=True) if cuda else dict(shuffle=True, batch_size=64)\n",
        "\n",
        "trainloader = torch.utils.data.DataLoader(train, **dataloader_args)\n",
        "\n",
        "testloader = torch.utils.data.DataLoader(test, **dataloader_args)"
      ],
      "execution_count": 136,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "is CUDA available True\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SamGp0coClTo",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# train_data = train.train_data\n",
        "# train_data = train.transform(train_data.numpy())\n",
        "\n",
        "# print('Train Data')\n",
        "# print('- Numpy Shape', train.train_data.cpu().numpy().shape)\n",
        "# print('- Tensor Shape', train.train_data.size())\n",
        "# print('- min', torch.min(train_data))\n",
        "# print('- max', torch.max(train_data))\n",
        "# print('- mean', torch.mean(train_data))\n",
        "# print('- std', torch.std(train_data))\n",
        "# print('- var', torch.var(train_data))\n",
        "\n",
        "# dataiterator = iter(trainloader)\n",
        "# images, labels = dataiterator.next()\n",
        "\n",
        "# print(images.shape)\n",
        "# print(labels.shape)\n",
        "\n",
        "# %matplotlib inline\n",
        "# import matplotlib.pyplot as plt\n",
        "\n",
        "# plt.imshow(images[0].numpy().squeeze(), cmap='gray_r')"
      ],
      "execution_count": 138,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ff3_yey3CnnZ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 246
        },
        "outputId": "62dc756d-84a2-4962-a589-fe7af61bcf5d"
      },
      "source": [
        "figure = plt.figure()\n",
        "num_of_images = 60\n",
        "for index in range(1, num_of_images + 1):\n",
        "    plt.subplot(6, 10, index)\n",
        "    plt.axis('off')\n",
        "    plt.imshow(images[index].numpy().squeeze(), cmap='gray_r')"
      ],
      "execution_count": 139,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 60 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Shp0t8rbCpuj",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "class Net(nn.Module):\n",
        "    def __init__(self):\n",
        "        super(Net, self).__init__()\n",
        "        # Input Block\n",
        "        self.convblock1 = nn.Sequential(\n",
        "            nn.Conv2d(in_channels=1, out_channels=8, kernel_size=(3, 3), padding=0, bias=False),\n",
        "            nn.ReLU(),\n",
        "            nn.BatchNorm2d(8)\n",
        "        ) # output_size = 26\n",
        "\n",
        "        # CONVOLUTION BLOCK 1\n",
        "        self.convblock2 = nn.Sequential(\n",
        "            nn.Conv2d(in_channels=8, out_channels=16, kernel_size=(3, 3), padding=0, bias=False),\n",
        "            nn.ReLU(),\n",
        "            nn.BatchNorm2d(16)\n",
        "        ) # output_size = 24\n",
        "        self.convblock3 = nn.Sequential(\n",
        "            nn.Conv2d(in_channels=16, out_channels=32, kernel_size=(1, 1), padding=0, bias=False),\n",
        "        ) # output_size = 24\n",
        "\n",
        "        self.pool1 = nn.MaxPool2d(2, 2) # output_size = 12\n",
        "\n",
        "        # TRANSITION BLOCK 1\n",
        "        self.convblock4 = nn.Sequential(\n",
        "            nn.Conv2d(in_channels=32, out_channels=16, kernel_size=(1, 1), padding=0, bias=False),\n",
        "            nn.ReLU(),\n",
        "            nn.BatchNorm2d(16)\n",
        "        ) # output_size = 12\n",
        "\n",
        "        self.convblock5 = nn.Sequential(\n",
        "            nn.Conv2d(in_channels=16, out_channels=16, kernel_size=(3, 3), padding=0, bias=False),\n",
        "            nn.ReLU(),\n",
        "            nn.BatchNorm2d(16)\n",
        "        ) # output_size = 10\n",
        "\n",
        "        # CONVOLUTION BLOCK 2\n",
        "        self.convblock6 = nn.Sequential(\n",
        "            nn.Conv2d(in_channels=16, out_channels=8, kernel_size=(3, 3), padding=0, bias=False),\n",
        "            nn.ReLU(),\n",
        "            nn.BatchNorm2d(8)\n",
        "        ) # output_size = 8\n",
        "        self.convblock7 = nn.Sequential(\n",
        "            nn.Conv2d(in_channels=8, out_channels=16, kernel_size=(3, 3), padding=0, bias=False),\n",
        "            nn.ReLU(),\n",
        "            nn.BatchNorm2d(16)\n",
        "        ) # output_size = 6\n",
        "\n",
        "        # CONVOLUTION BLOCK 3\n",
        "        self.convblock8 = nn.Sequential(\n",
        "            nn.Conv2d(in_channels=16, out_channels=16, kernel_size=(3, 3), padding=0, bias=False),\n",
        "            nn.ReLU(),\n",
        "            nn.BatchNorm2d(16)\n",
        "        ) # output_size = 4\n",
        "        self.convblock9 = nn.Sequential(\n",
        "            nn.Conv2d(in_channels=16, out_channels=16, kernel_size=(1, 1), padding=0, bias=False),\n",
        "            nn.ReLU(),\n",
        "            nn.BatchNorm2d(16)\n",
        "        ) # output_size = 4\n",
        "        self.gap = nn.Sequential(\n",
        "            nn.AvgPool2d(kernel_size=4)\n",
        "        ) # output_size = 1\n",
        "\n",
        "        self.convblock10 = nn.Sequential(\n",
        "            nn.Conv2d(in_channels=16, out_channels=10, kernel_size=(1, 1), padding=0, bias=False),\n",
        "        ) \n",
        "\n",
        "    def forward(self, x):\n",
        "        x = self.convblock1(x)\n",
        "        x = self.convblock2(x)\n",
        "        x = self.convblock3(x)\n",
        "        x = self.pool1(x)        \n",
        "        x = self.convblock4(x)\n",
        "        x = self.convblock5(x)\n",
        "        x = self.convblock6(x)\n",
        "        x = self.convblock7(x)\n",
        "        x = self.convblock8(x)\n",
        "        x = self.convblock9(x)        \n",
        "        x = self.gap(x)\n",
        "        x = self.convblock10(x)\n",
        "        x = x.view(-1, 10)\n",
        "        return F.log_softmax(x, dim=-1)"
      ],
      "execution_count": 189,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "n67mUyHoCqpJ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 764
        },
        "outputId": "ed848420-53f8-4b29-883c-3e4221aa5815"
      },
      "source": [
        "!pip install torchsummary\n",
        "from torchsummary import summary\n",
        "use_cuda = torch.cuda.is_available()\n",
        "device = torch.device(\"cuda\" if use_cuda else \"cpu\")\n",
        "print(device)\n",
        "model = Net().to(device)\n",
        "summary(model, input_size=(1, 28, 28))"
      ],
      "execution_count": 190,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Requirement already satisfied: torchsummary in /usr/local/lib/python3.6/dist-packages (1.5.1)\n",
            "cuda\n",
            "----------------------------------------------------------------\n",
            "        Layer (type)               Output Shape         Param #\n",
            "================================================================\n",
            "            Conv2d-1            [-1, 8, 26, 26]              72\n",
            "              ReLU-2            [-1, 8, 26, 26]               0\n",
            "       BatchNorm2d-3            [-1, 8, 26, 26]              16\n",
            "            Conv2d-4           [-1, 16, 24, 24]           1,152\n",
            "              ReLU-5           [-1, 16, 24, 24]               0\n",
            "       BatchNorm2d-6           [-1, 16, 24, 24]              32\n",
            "            Conv2d-7           [-1, 32, 24, 24]             512\n",
            "         MaxPool2d-8           [-1, 32, 12, 12]               0\n",
            "            Conv2d-9           [-1, 16, 12, 12]             512\n",
            "             ReLU-10           [-1, 16, 12, 12]               0\n",
            "      BatchNorm2d-11           [-1, 16, 12, 12]              32\n",
            "           Conv2d-12           [-1, 16, 10, 10]           2,304\n",
            "             ReLU-13           [-1, 16, 10, 10]               0\n",
            "      BatchNorm2d-14           [-1, 16, 10, 10]              32\n",
            "           Conv2d-15              [-1, 8, 8, 8]           1,152\n",
            "             ReLU-16              [-1, 8, 8, 8]               0\n",
            "      BatchNorm2d-17              [-1, 8, 8, 8]              16\n",
            "           Conv2d-18             [-1, 16, 6, 6]           1,152\n",
            "             ReLU-19             [-1, 16, 6, 6]               0\n",
            "      BatchNorm2d-20             [-1, 16, 6, 6]              32\n",
            "           Conv2d-21             [-1, 16, 4, 4]           2,304\n",
            "             ReLU-22             [-1, 16, 4, 4]               0\n",
            "      BatchNorm2d-23             [-1, 16, 4, 4]              32\n",
            "           Conv2d-24             [-1, 16, 4, 4]             256\n",
            "             ReLU-25             [-1, 16, 4, 4]               0\n",
            "      BatchNorm2d-26             [-1, 16, 4, 4]              32\n",
            "        AvgPool2d-27             [-1, 16, 1, 1]               0\n",
            "           Conv2d-28             [-1, 10, 1, 1]             160\n",
            "================================================================\n",
            "Total params: 9,800\n",
            "Trainable params: 9,800\n",
            "Non-trainable params: 0\n",
            "----------------------------------------------------------------\n",
            "Input size (MB): 0.00\n",
            "Forward/backward pass size (MB): 0.64\n",
            "Params size (MB): 0.04\n",
            "Estimated Total Size (MB): 0.68\n",
            "----------------------------------------------------------------\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3KuLQaDkCsav",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from tqdm import tqdm\n",
        "\n",
        "train_losses = []\n",
        "test_losses = []\n",
        "train_acc = []\n",
        "test_acc = []\n",
        "\n",
        "def train(model, device, train_loader, optimizer, epoch):\n",
        "    model.train()\n",
        "    pbar = tqdm(train_loader)\n",
        "    correct = 0\n",
        "    processed = 0\n",
        "    for batch_idx, (data, target) in enumerate(pbar):\n",
        "        data, target = data.to(device), target.to(device)\n",
        "        optimizer.zero_grad()\n",
        "        output = model(data)\n",
        "        loss = F.nll_loss(output, target)\n",
        "        train_losses.append(loss)\n",
        "        loss.backward()\n",
        "        optimizer.step()\n",
        "        \n",
        "        pred = output.argmax(dim=1, keepdim=True)\n",
        "        correct += pred.eq(target.view_as(pred)).sum().item()\n",
        "        processed += len(data)\n",
        "\n",
        "        pbar.set_description(desc= f'loss={loss.item()} batch_id={batch_idx} Accuracy={100*correct/processed}')\n",
        "        train_acc.append(100*correct/processed)\n",
        "\n",
        "def test(model, device, test_loader):\n",
        "    model.eval()\n",
        "    test_loss = 0\n",
        "    correct = 0\n",
        "    with torch.no_grad():\n",
        "        for data, target in test_loader:\n",
        "            data, target = data.to(device), target.to(device)\n",
        "            output = model(data)\n",
        "            test_loss += F.nll_loss(output, target, reduction='sum').item()  # sum up batch loss\n",
        "            pred = output.argmax(dim=1, keepdim=True)  # get the index of the max log-probability\n",
        "            correct += pred.eq(target.view_as(pred)).sum().item()\n",
        "\n",
        "    test_loss /= len(test_loader.dataset)\n",
        "    test_losses.append(test_loss)\n",
        "\n",
        "    print('\\nTest set: Average loss: {:.4f}, Accuracy: {}/{} ({:.2f}%)\\n'.format(\n",
        "        test_loss, correct, len(test_loader.dataset),\n",
        "        100. * correct / len(test_loader.dataset)))\n",
        "    test_acc.append(100. * correct / len(test_loader.dataset))"
      ],
      "execution_count": 191,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1j1TBdvdCwtp",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "b3551ad8-dec0-4d8b-d062-d6d94556e8b7"
      },
      "source": [
        "model = Net().to(device)\n",
        "optimizer = optim.SGD(model.parameters(), lr=0.01, momentum=0.9)\n",
        "for epoch in range(10):\n",
        "    print('\\nEPOCH:', epoch + 1)\n",
        "    train(model, device, trainloader, optimizer, epoch)\n",
        "    test(model, device, testloader)"
      ],
      "execution_count": 192,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\r  0%|          | 0/469 [00:00<?, ?it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "EPOCH: 1\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "loss=0.11535421013832092 batch_id=468 Accuracy=91.29166666666667: 100%|██████████| 469/469 [00:12<00:00, 36.34it/s]\n",
            "  0%|          | 0/469 [00:00<?, ?it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Test set: Average loss: 0.0554, Accuracy: 9859/10000 (98.59%)\n",
            "\n",
            "\n",
            "EPOCH: 2\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "loss=0.15072090923786163 batch_id=468 Accuracy=98.05166666666666: 100%|██████████| 469/469 [00:13<00:00, 35.52it/s]\n",
            "  0%|          | 0/469 [00:00<?, ?it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Test set: Average loss: 0.0391, Accuracy: 9885/10000 (98.85%)\n",
            "\n",
            "\n",
            "EPOCH: 3\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "loss=0.04443027079105377 batch_id=468 Accuracy=98.46666666666667: 100%|██████████| 469/469 [00:13<00:00, 35.37it/s]\n",
            "  0%|          | 0/469 [00:00<?, ?it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Test set: Average loss: 0.0335, Accuracy: 9901/10000 (99.01%)\n",
            "\n",
            "\n",
            "EPOCH: 4\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "loss=0.0343194380402565 batch_id=468 Accuracy=98.69166666666666: 100%|██████████| 469/469 [00:13<00:00, 35.58it/s]\n",
            "  0%|          | 0/469 [00:00<?, ?it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Test set: Average loss: 0.0334, Accuracy: 9905/10000 (99.05%)\n",
            "\n",
            "\n",
            "EPOCH: 5\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "loss=0.034649137407541275 batch_id=468 Accuracy=98.84833333333333: 100%|██████████| 469/469 [00:13<00:00, 35.46it/s]\n",
            "  0%|          | 0/469 [00:00<?, ?it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Test set: Average loss: 0.0306, Accuracy: 9905/10000 (99.05%)\n",
            "\n",
            "\n",
            "EPOCH: 6\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "loss=0.01967504248023033 batch_id=468 Accuracy=98.99: 100%|██████████| 469/469 [00:13<00:00, 35.89it/s]\n",
            "  0%|          | 0/469 [00:00<?, ?it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Test set: Average loss: 0.0295, Accuracy: 9905/10000 (99.05%)\n",
            "\n",
            "\n",
            "EPOCH: 7\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "loss=0.00414412934333086 batch_id=468 Accuracy=98.99166666666666: 100%|██████████| 469/469 [00:12<00:00, 36.20it/s]\n",
            "  0%|          | 0/469 [00:00<?, ?it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Test set: Average loss: 0.0280, Accuracy: 9907/10000 (99.07%)\n",
            "\n",
            "\n",
            "EPOCH: 8\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "loss=0.07000923156738281 batch_id=468 Accuracy=99.095: 100%|██████████| 469/469 [00:12<00:00, 36.33it/s]\n",
            "  0%|          | 0/469 [00:00<?, ?it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Test set: Average loss: 0.0260, Accuracy: 9914/10000 (99.14%)\n",
            "\n",
            "\n",
            "EPOCH: 9\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "loss=0.09036801010370255 batch_id=468 Accuracy=99.165: 100%|██████████| 469/469 [00:12<00:00, 36.15it/s]\n",
            "  0%|          | 0/469 [00:00<?, ?it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Test set: Average loss: 0.0201, Accuracy: 9933/10000 (99.33%)\n",
            "\n",
            "\n",
            "EPOCH: 10\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "loss=0.03632470965385437 batch_id=468 Accuracy=99.205: 100%|██████████| 469/469 [00:13<00:00, 35.25it/s]\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "\n",
            "Test set: Average loss: 0.0233, Accuracy: 9917/10000 (99.17%)\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "9G1NCZFfCy9Q",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 624
        },
        "outputId": "82586086-58a4-4d96-cfea-d6856e39655c"
      },
      "source": [
        "fig, axis = plt.subplots(2, 2, figsize=(15,10))\n",
        "axis[0, 0].plot(train_losses)\n",
        "axis[0, 0].set_title(\"Training Loss\")\n",
        "axis[1 ,0].plot(train_acc)\n",
        "axis[1, 0].set_title(\"Training Accuracy\")\n",
        "axis[0 ,1].plot(test_losses)\n",
        "axis[0, 1].set_title(\"Testing Loss\")\n",
        "axis[1 ,1].plot(test_acc)\n",
        "axis[1, 1].set_title(\"Testing Accuracy\")"
      ],
      "execution_count": 193,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "Text(0.5, 1.0, 'Testing Accuracy')"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 193
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1080x720 with 4 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    }
  ]
}