Captum/Interpret_your_PyTorch_model_with_Captum.ipynb

{
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Pp1fXMH8oLz_"
      },
      "source": [
        "# Interpret your PyTorch model with Captum\n",
        "\n",
        "This notebook demonstrates how to use the methods available in Captum. It isn't focused on gaining information from the results. Rather it aims to give you information about how to implement Captum in your own project.\n",
        "\n",
        "<table align=\"left\"><td>\n",
        "  <a target=\"_blank\"  href=\"https://colab.research.google.com/github/TannerGilbert/Model-Interpretation/blob/master/Captum/Interpret_your_PyTorch_model_with_Captum.ipynb\">\n",
        "    <img src=\"https://www.tensorflow.org/images/colab_logo_32px.png\" />Run in Google Colab\n",
        "  </a>\n",
        "</td><td>\n",
        "  <a target=\"_blank\"  href=\"https://github.com/TannerGilbert/Model-Interpretation/blob/master/Captum/Interpret_your_PyTorch_model_with_Captum.ipynb\">\n",
        "    <img width=32px src=\"https://www.tensorflow.org/images/GitHub-Mark-32px.png\" />View source on GitHub</a>\n",
        "</td></table>"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "id": "yADE36K_ksI9"
      },
      "outputs": [],
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "\n",
        "%matplotlib inline\n",
        "\n",
        "import torch\n",
        "import torchvision\n",
        "import torchvision.transforms as transforms\n",
        "import torchvision.transforms.functional as TF"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 101,
          "referenced_widgets": [
            "87f16cad17374bd7ad6d41777b83d751",
            "facbadef58b84e798dfdd965abac0acc",
            "f5ed5675b79041d18d90215eaefd9490",
            "84baf7dd22f4400481d0387b3c8ebc86",
            "aa41c5e6ec354c259f78a4f69527a625",
            "78c4c247c9734fa7b6d4eb7928fd384e",
            "1adcbb4b34f54525b4cea6e11fef5502",
            "2346d728f302413f9bfd5a89d17aba6a",
            "793d6f5a901140579982778020495f8c",
            "9659711f0c8a43a1b4214778bce676a0",
            "b11fd8dd3eec4778bbfa7c8360098fdb"
          ]
        },
        "id": "vD7xfql8kvR0",
        "outputId": "33701263-9838-4444-b1e1-516f192783ca"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Downloading https://www.cs.toronto.edu/~kriz/cifar-10-python.tar.gz to ./data/cifar-10-python.tar.gz\n"
          ]
        },
        {
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "87f16cad17374bd7ad6d41777b83d751",
              "version_major": 2,
              "version_minor": 0
            },
            "text/plain": [
              "  0%|          | 0/170498071 [00:00<?, ?it/s]"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Extracting ./data/cifar-10-python.tar.gz to ./data\n",
            "Files already downloaded and verified\n"
          ]
        }
      ],
      "source": [
        "transform = transforms.Compose(\n",
        "    [transforms.ToTensor(),\n",
        "     transforms.Normalize((0.5, 0.5, 0.5), (0.5, 0.5, 0.5))])\n",
        "\n",
        "trainset = torchvision.datasets.CIFAR10(root='./data', train=True,\n",
        "                                        download=True, transform=transform)\n",
        "trainloader = torch.utils.data.DataLoader(trainset, batch_size=4,\n",
        "                                          shuffle=True, num_workers=2)\n",
        "\n",
        "testset = torchvision.datasets.CIFAR10(root='./data', train=False,\n",
        "                                       download=True, transform=transform)\n",
        "testloader = torch.utils.data.DataLoader(testset, batch_size=4,\n",
        "                                         shuffle=False, num_workers=2)\n",
        "\n",
        "classes = ('plane', 'car', 'bird', 'cat',\n",
        "           'deer', 'dog', 'frog', 'horse', 'ship', 'truck')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 155
        },
        "id": "mxy9Kn_joFzc",
        "outputId": "2d5e144d-580b-498f-f662-9f338a0c4757"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "  cat  deer  ship   car\n"
          ]
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ],
      "source": [
        "import matplotlib.pyplot as plt\n",
        "import numpy as np\n",
        "\n",
        "# functions to show an image\n",
        "\n",
        "\n",
        "def imshow(img):\n",
        "    img = img / 2 + 0.5     # unnormalize\n",
        "    npimg = img.numpy()\n",
        "    plt.imshow(np.transpose(npimg, (1, 2, 0)))\n",
        "\n",
        "\n",
        "# get some random training images\n",
        "dataiter = iter(trainloader)\n",
        "images, labels = dataiter.next()\n",
        "\n",
        "# show images\n",
        "imshow(torchvision.utils.make_grid(images))\n",
        "# print labels\n",
        "print(' '.join('%5s' % classes[labels[j]] for j in range(4)))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "id": "KYWC3lHboKhF"
      },
      "outputs": [],
      "source": [
        "import torch.nn as nn\n",
        "import torch.nn.functional as F\n",
        "\n",
        "\n",
        "class Net(nn.Module):\n",
        "    def __init__(self):\n",
        "        super(Net, self).__init__()\n",
        "        self.conv1 = nn.Conv2d(3, 6, 5)\n",
        "        self.pool1 = nn.MaxPool2d(2, 2)\n",
        "        self.pool2 = nn.MaxPool2d(2, 2)\n",
        "        self.conv2 = nn.Conv2d(6, 16, 5)\n",
        "        self.fc1 = nn.Linear(16 * 5 * 5, 120)\n",
        "        self.fc2 = nn.Linear(120, 84)\n",
        "        self.fc3 = nn.Linear(84, 10)\n",
        "        self.relu1 = nn.ReLU()\n",
        "        self.relu2 = nn.ReLU()\n",
        "        self.relu3 = nn.ReLU()\n",
        "        self.relu4 = nn.ReLU()\n",
        "\n",
        "    def forward(self, x):\n",
        "        x = self.pool1(self.relu1(self.conv1(x)))\n",
        "        x = self.pool2(self.relu2(self.conv2(x)))\n",
        "        x = x.view(-1, 16 * 5 * 5)\n",
        "        x = self.relu3(self.fc1(x))\n",
        "        x = self.relu4(self.fc2(x))\n",
        "        x = self.fc3(x)\n",
        "        return x\n",
        "\n",
        "\n",
        "net = Net()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "id": "kccm0izFoLe3"
      },
      "outputs": [],
      "source": [
        "import torch.optim as optim\n",
        "\n",
        "criterion = nn.CrossEntropyLoss()\n",
        "optimizer = optim.SGD(net.parameters(), lr=0.001, momentum=0.9)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "4l1ybxHkoPGv",
        "outputId": "14003653-0216-46ce-99dd-40fa96ae62d9"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "[1,  2000] loss: 2.221\n",
            "[1,  4000] loss: 1.921\n",
            "[1,  6000] loss: 1.700\n",
            "[1,  8000] loss: 1.578\n",
            "[1, 10000] loss: 1.526\n",
            "[1, 12000] loss: 1.482\n",
            "[2,  2000] loss: 1.402\n",
            "[2,  4000] loss: 1.375\n",
            "[2,  6000] loss: 1.362\n",
            "[2,  8000] loss: 1.336\n",
            "[2, 10000] loss: 1.314\n",
            "[2, 12000] loss: 1.273\n",
            "[3,  2000] loss: 1.240\n",
            "[3,  4000] loss: 1.212\n",
            "[3,  6000] loss: 1.223\n",
            "[3,  8000] loss: 1.198\n",
            "[3, 10000] loss: 1.192\n",
            "[3, 12000] loss: 1.168\n",
            "[4,  2000] loss: 1.098\n",
            "[4,  4000] loss: 1.103\n",
            "[4,  6000] loss: 1.117\n",
            "[4,  8000] loss: 1.112\n",
            "[4, 10000] loss: 1.134\n",
            "[4, 12000] loss: 1.116\n",
            "[5,  2000] loss: 1.033\n",
            "[5,  4000] loss: 1.010\n",
            "[5,  6000] loss: 1.046\n",
            "[5,  8000] loss: 1.044\n",
            "[5, 10000] loss: 1.060\n",
            "[5, 12000] loss: 1.057\n",
            "[6,  2000] loss: 0.955\n",
            "[6,  4000] loss: 0.961\n",
            "[6,  6000] loss: 0.996\n",
            "[6,  8000] loss: 1.000\n",
            "[6, 10000] loss: 0.976\n",
            "[6, 12000] loss: 0.996\n",
            "[7,  2000] loss: 0.899\n",
            "[7,  4000] loss: 0.923\n",
            "[7,  6000] loss: 0.940\n",
            "[7,  8000] loss: 0.946\n",
            "[7, 10000] loss: 0.968\n",
            "[7, 12000] loss: 0.956\n",
            "[8,  2000] loss: 0.859\n",
            "[8,  4000] loss: 0.901\n",
            "[8,  6000] loss: 0.899\n",
            "[8,  8000] loss: 0.903\n",
            "[8, 10000] loss: 0.915\n",
            "[8, 12000] loss: 0.908\n",
            "Finished Training\n"
          ]
        }
      ],
      "source": [
        "for epoch in range(8):  # loop over the dataset multiple times\n",
        "\n",
        "    running_loss = 0.0\n",
        "    for i, data in enumerate(trainloader, 0):\n",
        "        # get the inputs\n",
        "        inputs, labels = data\n",
        "        # zero the parameter gradients\n",
        "        optimizer.zero_grad()\n",
        "\n",
        "        # forward + backward + optimize\n",
        "        outputs = net(inputs)\n",
        "        loss = criterion(outputs, labels)\n",
        "        loss.backward()\n",
        "        optimizer.step()\n",
        "\n",
        "        # print statistics\n",
        "        running_loss += loss.item()\n",
        "        if i % 2000 == 1999:    # print every 2000 mini-batches\n",
        "            print('[%d, %5d] loss: %.3f' %\n",
        "                  (epoch + 1, i + 1, running_loss / 2000))\n",
        "            running_loss = 0.0\n",
        "\n",
        "print('Finished Training')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 155
        },
        "id": "-Ad6Apnboo6N",
        "outputId": "223534da-5df0-4deb-a8ff-68199bd7e87d"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "GroundTruth:    cat  ship  ship plane\n"
          ]
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ],
      "source": [
        "dataiter = iter(testloader)\n",
        "images, labels = dataiter.next()\n",
        "\n",
        "# print images\n",
        "imshow(torchvision.utils.make_grid(images))\n",
        "print('GroundTruth: ', ' '.join('%5s' % classes[labels[j]] for j in range(4)))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "id": "pZVCRWcHo9SH"
      },
      "outputs": [],
      "source": [
        "outputs = net(images)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "0wY-Op60pGag",
        "outputId": "d24c55b0-988c-409e-80ec-4abf57c6ba5d"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Predicted:    cat   car  ship  ship\n"
          ]
        }
      ],
      "source": [
        "_, predicted = torch.max(outputs, 1)\n",
        "\n",
        "print('Predicted: ', ' '.join('%5s' % classes[predicted[j]]\n",
        "                              for j in range(4)))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uJehSifKpHLf",
        "outputId": "7eeeb69f-64ae-491e-cb48-01e31e06d6c4"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Accuracy of the network on the 10000 test images: 60 %\n"
          ]
        }
      ],
      "source": [
        "correct = 0\n",
        "total = 0\n",
        "with torch.no_grad():\n",
        "    for data in testloader:\n",
        "        images, labels = data\n",
        "        outputs = net(images)\n",
        "        _, predicted = torch.max(outputs.data, 1)\n",
        "        total += labels.size(0)\n",
        "        correct += (predicted == labels).sum().item()\n",
        "\n",
        "print('Accuracy of the network on the 10000 test images: %d %%' % (\n",
        "    100 * correct / total))"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "03wDlK4bIjNU"
      },
      "source": [
        "# Install Captum"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yVmCZHm4wiVc"
      },
      "source": [
        "Installing Captum is super simple. All you need is a simple pip install."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 11,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BeYzu7iFLLX1",
        "outputId": "83a8f5cb-9fd4-4294-d779-15dce7f82fad"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Collecting captum\n",
            "  Downloading captum-0.5.0-py3-none-any.whl (1.4 MB)\n",
            "\u001b[K     |████████████████████████████████| 1.4 MB 32.0 MB/s \n",
            "\u001b[?25hRequirement already satisfied: numpy in /usr/local/lib/python3.7/dist-packages (from captum) (1.21.6)\n",
            "Requirement already satisfied: matplotlib in /usr/local/lib/python3.7/dist-packages (from captum) (3.2.2)\n",
            "Requirement already satisfied: torch>=1.6 in /usr/local/lib/python3.7/dist-packages (from captum) (1.11.0+cu113)\n",
            "Requirement already satisfied: typing-extensions in /usr/local/lib/python3.7/dist-packages (from torch>=1.6->captum) (4.1.1)\n",
            "Requirement already satisfied: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.1 in /usr/local/lib/python3.7/dist-packages (from matplotlib->captum) (3.0.9)\n",
            "Requirement already satisfied: python-dateutil>=2.1 in /usr/local/lib/python3.7/dist-packages (from matplotlib->captum) (2.8.2)\n",
            "Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.7/dist-packages (from matplotlib->captum) (1.4.3)\n",
            "Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.7/dist-packages (from matplotlib->captum) (0.11.0)\n",
            "Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.7/dist-packages (from python-dateutil>=2.1->matplotlib->captum) (1.15.0)\n",
            "Installing collected packages: captum\n",
            "Successfully installed captum-0.5.0\n"
          ]
        }
      ],
      "source": [
        "!pip install captum"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "VWc6UC-mogVW"
      },
      "source": [
        "# General Attribution Techniques\n",
        "\n",
        "First we can take a look at what features are actually important to the model when making predictions. Captum has multiple methods for this. For more information about the theory behind these methods take a look at the [official documentation](https://captum.ai/docs/algorithms#general-attribution)."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 12,
      "metadata": {
        "id": "VYrI5NMiJRl9"
      },
      "outputs": [],
      "source": [
        "from captum.attr import IntegratedGradients\n",
        "from captum.attr import Saliency\n",
        "from captum.attr import DeepLift\n",
        "from captum.attr import NoiseTunnel\n",
        "from captum.attr import visualization as viz"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "S9Fh3vwDI3Pb"
      },
      "source": [
        "Pick a test image to analyse."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 13,
      "metadata": {
        "id": "SWSoBEpuIwzE"
      },
      "outputs": [],
      "source": [
        "ind = 3\n",
        "\n",
        "input = images[ind].unsqueeze(0)\n",
        "input.requires_grad = True"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 14,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "iugOHOPBI8pJ",
        "outputId": "6e19963d-594f-4f74-db92-a4f9d30f45ae"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "Net(\n",
              "  (conv1): Conv2d(3, 6, kernel_size=(5, 5), stride=(1, 1))\n",
              "  (pool1): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
              "  (pool2): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
              "  (conv2): Conv2d(6, 16, kernel_size=(5, 5), stride=(1, 1))\n",
              "  (fc1): Linear(in_features=400, out_features=120, bias=True)\n",
              "  (fc2): Linear(in_features=120, out_features=84, bias=True)\n",
              "  (fc3): Linear(in_features=84, out_features=10, bias=True)\n",
              "  (relu1): ReLU()\n",
              "  (relu2): ReLU()\n",
              "  (relu3): ReLU()\n",
              "  (relu4): ReLU()\n",
              ")"
            ]
          },
          "execution_count": 14,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "net.eval()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {
        "id": "Wipl-zAMJC8k"
      },
      "outputs": [],
      "source": [
        "def attribute_image_features(algorithm, input, **kwargs):\n",
        "    net.zero_grad()\n",
        "    tensor_attributions = algorithm.attribute(input,\n",
        "                                              target=labels[ind],\n",
        "                                              **kwargs\n",
        "                                             )\n",
        "    \n",
        "    return tensor_attributions"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 16,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "kpwnQuXbKm-6",
        "outputId": "15f522d2-197e-4b9f-f9f1-2626b25e251f"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "(32, 32, 3)\n"
          ]
        }
      ],
      "source": [
        "saliency = Saliency(net)\n",
        "grads = saliency.attribute(input, target=labels[ind].item())\n",
        "grads = np.transpose(grads.squeeze().cpu().detach().numpy(), (1, 2, 0))\n",
        "print(grads.shape)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 17,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "7a3chfDxKsUB",
        "outputId": "3631e486-af27-4aad-85d8-9594cd79a1e6"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Approximation delta:  tensor([0.0151], dtype=torch.float64)\n"
          ]
        }
      ],
      "source": [
        "ig = IntegratedGradients(net)\n",
        "attr_ig, delta = attribute_image_features(ig, input, baselines=input * 0, return_convergence_delta=True)\n",
        "attr_ig = np.transpose(attr_ig.squeeze().cpu().detach().numpy(), (1, 2, 0))\n",
        "print('Approximation delta: ', abs(delta))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 18,
      "metadata": {
        "id": "8OypjhbDKwr7"
      },
      "outputs": [],
      "source": [
        "ig = IntegratedGradients(net)\n",
        "nt = NoiseTunnel(ig)\n",
        "attr_ig_nt = attribute_image_features(nt, input, baselines=input * 0, nt_type='smoothgrad_sq',\n",
        "                                      nt_samples=100, stdevs=0.2)\n",
        "attr_ig_nt = np.transpose(attr_ig_nt.squeeze(0).cpu().detach().numpy(), (1, 2, 0))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "h2B81GG7Kxdx",
        "outputId": "2a9f8c51-1dcd-4947-bb91-6142b9ef9567"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/captum/attr/_core/deep_lift.py:339: UserWarning: Setting forward, backward hooks and attributes on non-linear\n",
            "               activations. The hooks and attributes will be removed\n",
            "            after the attribution is finished\n",
            "  after the attribution is finished\"\"\"\n",
            "/usr/local/lib/python3.7/dist-packages/torch/nn/functional.py:749: UserWarning: Note that order of the arguments: ceil_mode and return_indices will changeto match the args list in nn.MaxPool2d in a future release.\n",
            "  warnings.warn(\"Note that order of the arguments: ceil_mode and return_indices will change\"\n"
          ]
        }
      ],
      "source": [
        "dl = DeepLift(net)\n",
        "attr_dl = attribute_image_features(dl, input, baselines=input * 0)\n",
        "attr_dl = np.transpose(attr_dl.squeeze(0).cpu().detach().numpy(), (1, 2, 0))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 20,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "eGZAg3rcK0aj",
        "outputId": "e2bc2432-3049-4ab4-a23c-fead1749d9ec"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Predicted: horse  Probability: 0.8272429704666138\n"
          ]
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 432x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        },
        {
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAU0AAAF1CAYAAABs2bgEAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3deZRcZ3nn8d9bS++b1Fosy7Jk2TK2LBsDtrETMITVZrEJi0M2IAFmINvsZ7JNTiY5Z7JMViZkIXOGQDKEbbI4EA4QwDF4xeBFNlheZVvW2upW7921vfPHvS1Vt1r9PK+qrZbs74cj3F319HPfe+vW07e63qfeEGMUAMCnsNIDAIAzCUUTABJQNAEgAUUTABJQNAEgAUUTABJQNE9DIYQYQrjgFG/zfSGEb53KbS4yhi35vpfy778UQnjvSo7pVFru/Q0h3BJC+MBy5UOGoumQF5SdIYSpEML+EMKfhRAGVnpcp1IIoS2E8GshhF0hhMkQwrP5k/wNz9U2Y4zXxxg/0Woezy+EvMDEEMKLF9z+9/ntr251HJbm/T0dfolhcRRNQwjhP0n6HUn/RVK/pKslbZb01RBC2zJvq7Sc+ZbZ5yXdKOk9klZJOk/SH0t682LBp/m+nMgjyvZPkhRCGJR0jaRDKzYinH5ijPw7wT9JfZImJN204PYeZU+kn5Z0tqRpSaub7n+JpCFJ5fz7n5b0fUkjkr4saXNTbJT0s5IelfRk020X5F+/WdK9ksYkPSPp15t+9ouSfn7B2B6Q9MP51xdJ+qqkYUm7mvdD0qCkm/O8d0v6TUnfOsFxeF2+j+cYx2u3pP+aj2FWUknSL0p6XNK4pO/NjS2PL0r6vfxYPZEfhyiplN9/i6QPNMVbx/FD+XE8IumjkoKkiyXNSKrnj+WRE4z9Fkm/JmmPpGJ+289J+rP8tlfnt10l6Y58G/sk/YmktqY8b8iP9aikP5X0r3P7IOl9kr6V7/OIpCclXb9gDB840ZgXOR7va37MJL1e0sP5tv+kedtLHb/8OP2hpIP5+bBT0o6Vfv6drv9WfACn8z9J10mqzT2JF9z3CUl/m3/9dUkfbLrvf0r68/zrGyU9lj8RSpJ+VdLtTbFRWWFbLamz6ba5ovlqSZcqe1VwmaQDkt6W33eTpLuacr1Y0mFJbZK6lRXZn8q3O1fIt+exn5b02Txuh6RndeKi+duSbnEcr92S7pO0qWlf3qXsF0tB0o9ImpS0Ib/vQ/mTfFO+/9/QCYqm8zh+QdKApHOV/VK7Lr9vXnE5wdhvUVawvqK8kCn7ZXKN5hfNlyl7tVGStEVZEfr3+X1r8qLz9vz+fyepqvlFsyrpg8p+YXxY0l5JYZH9PW7MWqJo5tsel/ROSWVJ/0HZuWseP0lvlPSd/NjN/aLZsNLPv9P1Hy/Pl7ZG0lCMsbbIffvy+yXpU5J+VJJCCEHSu/PbpKww/FaM8ft5nv8h6fIQwuamXL8VYxyOMU4v3EiM8ZYY484YYyPG+ICkv5X0qvzumyVdGELYln//k5I+E2OsSHqLpN0xxo/HGGsxxnsl/T9J7wohFCW9Q9KvxRgnY4wPKvslsNRx2D/3TQhhdQjhSAhhNIQwsyD2IzHGZ+b2Jcb4uRjj3nz8n1F2JXhVHnuTpD/K44cl/dYSY/Acx9+OMR6JMT6trABfvkS+E/mkpPeEEC6SNBBjvKP5zhjjd2KMd+bHdLekv9Cxx+NNkh6KMf5dPsaPqOm45Z6KMf5ljLGu7JhvkLT+JMa50Ny2Px9jrEr6owXbXur4VSX1KntlEvKYfcswpucliubShiStOcHf5zbk90tZMbomhLBB0rWSGpK+md+3WdIf50XmiLKXykHSxqZcz5xoACGEl4cQvhFCOBRCGFV28q+RpBjjjKTPSPqJEEJBWeH+66btvnxuu/m2f1zSWZLWKrvaaN7uU0sch8P5/irf7nCMcUDZVVf7gth5+xJCeE8I4b6mMezQsV82ZyeMwXMcm4vElLI/o6T6O0mvUfbS/K8X3hlCuDCE8IX8DcExZcVn0f2J2WXcngUp9jfdP5V/eTLjXGixbTcf2xMevxjj15W9nP+opIMhhI+FEPqWYUzPSxTNpd2h7G9zb2++MYTQI+l6SV+TpBjjiLKXdT8i6cckfTo/aaXsxP23McaBpn+dMcbbm1Iu9VFTn1J2Rbkpxtgv6c+VnexzPqGsGL5W0lTTldEzkv51wXZ7YowfVvbStabsZfGcc5cYw9ckXRlCOGeJmOP2Jb+K+UtlBWgwL7QPNo1/X8IYPMfRHJMZmBWyLyl76Xxc0VT2N86HJW2LMfZJ+mXN35+jxyh/1eE5Zt4xT0rqavr+rKav5x3LfNvNx3bJ4xdj/EiM8WWStku6UNkbn1gERXMJMcZRSf9d0v8KIVwXQiiHELYo+1vgHs1/Un1K2Tuv79Sxl+ZSVuR+KYRwiSSFEPpDCO9KGEavpOEY40wI4SplRbl5jHcou7L9/QXj+YKyl+4/mY+7HEK4MoRwcf7S8O8k/XoIoSuEsF3SCecHxhi/ouzl7j/kV75tIYSysr/tLaVb2ZP/UL7vP6XsSnPOZyX9QgjhnBDCKmVvGp1IK8fxgKRzEmY7/LKkV+UvvxfqVfZ3y4n8JfyHm+77oqRLQwhvy1+d/KzmF7YUi435Pklvzx+zCyS9f8G2LwkhvD3f9i8s2PYJj19+Xrw8f0wnlb0J1TjJcT/vUTQNMcbfVfYk+j1lT5a7lP3Wfm2McbYp9GZJ2yTtjzHe3/Tzf69sytKn85dzDyq7SvX6GUm/EUIYV/bu7mcXifmksjeL/qZpu+PK3sl9t7I3G/bn45h7Of1zyl4W7pf0V5I+bozjh5UV4r9R9s7xk8qucN94oh+IMX5PWTG/Q1kRuFTSbU0hf6nsXdz7JX1XWSE/Ua5WjuPXJT0kaX8IYcgKzv8Ge6I5kv9Z2S+u8Xz8n2n6uSFlb3z9rrI/aWyXdI+yVyupFhvzH0qqKDuWn5D0fxfZ9m/n296mpmNtHL++fF9GlP2J5LCyNzOxiLl37XAGCyG8R9K/iTG+YqXHgmPyvzPvkfTjMcZvrPR4sDy40jzDhRC6lF2NfmylxwIphPDGEMJACKFdx/7eeecKDwvLiKJ5BgshvFHZ3wsPaP7fUbFyrlE2mX9I0luVzak9bioZzly8PAeABFxpAkACiiYAJGjpk2h6enrj4ODgkjHFUtGVK8ybr724er3uyuWJ6+7pduXq7Ow0Y6rVqivX7GzFjOnu9o3Ls81yuezKVSws3+/O6J9HfuZazl20T3vXc+N0NjU9ZcY06ss3LbSjs8MV98D99w/FGNem5m+paA4ODuqXfuVXl4wZWLXKlcvzxB0dHXPlGh0bN2OuvvoqM0aSdlxyiRmzf/8BV67HnnjSjLnqKt+49u5b2NJ8vE3nbDRjJKnLdZL5KoXnb+T+P6M7cjkzecpOw5vMs4/OVFnjztJKReeFhyOX9z2M5cy1c+cDZsz4+KQrl+eX8iXbL3bl2rBuzVJtuyfEy3MASEDRBIAEFE0ASEDRBIAEFE0ASEDRBIAELU05ioqq1ZaeEzk7vXA1hMVNOuZyFYKvxleq9nzI/fsPunIVS/Zcx5HhYVeuAwfsqUkHhw77ch20x9/V7ftA8JJjLm1Yxnbb0enFVg85XrnomLvb8M3dLXim0DjnQ3qiSiXfudpwTE/ssacKS5Jmq3aymnNeVbloj39qcsKV66DjXN39jG91jXrNfm7Pzjy3rf5caQJAAoomACSgaAJAAoomACSgaAJAAoomACSgaAJAAoomACSgaAJAgpY6gjzqzi6LatXuEuns6vJts253PczO+roG9j77rBkzNDTkyjU+NWvG7HrkUVeuRt0+XkPDI65chWjn8n6YdHB0bXW2+X5X1xzdMpOTvsexXLJP9XrwPR3aynYHVSH6zvv2sn0s6s4ung5HrornoEraN2SfO0894TtXDx+2c01O2h2BklRo2M+hRx99zJXrZHGlCQAJKJoAkICiCQAJKJoAkICiCQAJKJoAkICiCQAJKJoAkKC15S5iVK1WXTJmamrSmcvxUf3Vpbc1p1iwJxYfdk78lkbNiIOHj7gyeaYoDwwdcuVq67KXsnh6j28Jgakx+1hsafM1Fqx2rM0w65xgXXXEzcz4llM5HO3rg63rV7tyfer2XWbMi3p9E9I3n73ejBkY6Hflqjbs87673feUHzu834zZtesRV67geG73dNrLykjSzIy9vEmj7lsC5WRxpQkACSiaAJCAogkACSiaAJCAogkACSiaAJCAogkACSiaAJCAogkACVrqCGo0oqaml/74+aJjGQtJKhXs+j3d8HV/TM5UzJi2dl93Ua1udzNMzPpyDXTZXQ9Hpny5zhvsMGNufXSPK9fWNb1mzPot21y5ZhyP97Tj8ZGkcUdjx76Dh1259ozZy2KMHjzgyrVx1u60evzpg65cu5+wj/25Gze4ctVL9jlx4ZZzXbkmJibMmOlp3/Oxs91eHmR62ndOhGB3PTUcHUit4EoTABJQNAEgAUUTABJQNAEgAUUTABJQNAEgAUUTABJQNAEgQUuT26WoGJeegVyt+iaa1oJdv9vKvo/E7+q0J/k2Gr5xTUzZE3jX9dtLPEjSPU/bE7Fv2HqBK9fwhD2uiYlxV65Gt6OxYMy3pMdItd2MqVR8E5k9E/2ffvJJV67usr38xEO77aVNJOl/f/V7ZswNL/FNSJ+oDZkxI84lUCYcp3S55lt+ZmJiyoyZcjZ1hIYdV3UuZVMq2RPlHfPfW8KVJgAkoGgCQAKKJgAkoGgCQAKKJgAkoGgCQAKKJgAkoGgCQAKKJgAkaLEjSGpYjRbR7sSQpELRjqvVfEtnlMv2bk1VvOOyc834mhlUL3WZMZUZ3xICN+/cb8a89eIBV64nR+ylDXbt2uXKJUdnV1G+bqw9h8fMmEJt6eVW5hyu2mtnfHu3b7mLH/vB81xxHnc/Ym/z3NX2khiS1OHocvuDf7jdleu9P2h3pvW0+1pvajX72Jt1JDflWGKjzfH8bwVXmgCQgKIJAAkomgCQgKIJAAkomgCQgKIJAAkomgCQgKIJAAlamgUao71sRKnkW6KiVrcnwCr6JrePTtkTnvt7fEtUFBwTZSdnHWOXtHnAPhYjI/aSGJK0sc+OmZj1zRje2Gcfi+mxEVeuSsOe8Oxd2qDhyNUotblyFYP9GL18q2+JitmqfR7un/RNur/+ko1mTMPRMCBJ9w7by4jcePm5rly3PX7QjHnxhn5XrolJe+mMKF9TR6lonxOdHfZyN63gShMAElA0ASABRRMAElA0ASABRRMAElA0ASABRRMAElA0ASABRRMAErTUEVQIQW2lpVOMOdeC6FjGT6jvLNu/C54d83UgnDvQbcb0d/m6ntpn7eUIvvmYb8mF1+/YYsaUSkVXruBYfuLhYV+Hy0Xr7KUZ2tp8x6sr2CdF1bkEyoHRaTPmzsdHXbneevnZZkyx09dxNuPomDurw/c4nlWxz+mBfl8Xz4vb2s2YYsF3zbWh125f62zz7aNn6Qz5VuE4aVxpAkACiiYAJKBoAkACiiYAJKBoAkACiiYAJKBoAkACiiYAJGhpSnkjRs0aSxe0Ocvy2JQ9SbngnLQagz1Rdk23b/LxniP2R/Wv6+1y5YpF+2P4r9q2yZWrHuxJ0UecSy58dedeM+adP3ChK1elaE+K7gy+ZTgmJyfNGO/k9vufGTZjrr1wjStXsWifX/1l3z5OB8fyIBVfg8jBIXsfv171Lc3yhovtpT9qzmPf65icX3YsKyM5H2/foT9pXGkCQAKKJgAkoGgCQAKKJgAkoGgCQAKKJgAkoGgCQAKKJgAkoGgCQILWlrsoFNTZufRyEJWGr41nsMceSm+Hb5mEWLA7ECZ8TRa6YK3dOdTfY3f6SFJne5sZMzzj67Jod7Q9VIt2N5MkXX3pZjOmo80eu5SdE5ZnJ3zjWt9hH9dvPnrYlet1juVBOsu+JReOjI2bMV966FlXrr07v2/GVKPv2mZg80Yz5oZNA65c//TAHjPmuovXuXI1on1cp2YqvlwNR0dTw16+pRVcaQJAAoomACSgaAJAAoomACSgaAJAAoomACSgaAJAAoomACRoaXJ7Z2eHLrlk+5IxwTHZWZI8U+AnZ30Tvzvb7Mm0dz9zxJVr25qlJ+9LUt35u6eny14Korvb1wwQHEtGHKj6cr103VozZr93QnqX3Qzw4MMjrlybd/SaMW+5eNCV65bH7Unw0zXf02Fjj30e7thsH1NJ2n6+3Vjw4B33uXLdcPVWM2bSWJ5mzo9ebS9vUnCuP9NesuNqdd+E9OiYuO5YQaQlXGkCQAKKJgAkoGgCQAKKJgAkoGgCQAKKJgAkoGgCQAKKJgAkoGgCQIKWOoKkoGJp6RQF7/R8R1ybc+mMRrS7ZSamZly5btv5tBkzuH6DK9crLrK7PyanfeMaG5s0Y+Ksr4unILuLZ6biWGZA0t76hBnzukvXu3L90ee+Zcbse+heV6633fQGM+b+bz/iynXJ9VeYMV+8/UlXrre9cpsZc/519vYkaXWn3Qk3OzTqyjU1M2vGdHX4lkCZnrW7eOoN+zkr+br9ouP53wquNAEgAUUTABJQNAEgAUUTABJQNAEgAUUTABJQNAEgAUUTABK0NLm9WCiop6tjyRjPx9NLvsmtM1VfrqJjiY32ypgrV2zYk7ov3bzRlatStZdJ+OQ3d7lyPXj/Y2bMTT9kL1kgSXvL9mnwjXueceW67opzzJhH9g27cvWvGjBjLn3n9a5cGwfsCfzvveEqV65bn7LPnR0X+ybwjxzcb8b847/4lrvYdrU9Cf41W3pcuUZG7X0slVa5cnWV7edjueS8fnM0wXhrzsniShMAElA0ASABRRMAElA0ASABRRMAElA0ASABRRMAElA0ASABRRMAErS43IX90fLOT7GXJ2zaueRCiHZHwL79B1y5LjjX7nBpb2t35Xpm714z5sr1vodk6jy7W2Z0etqV6+EnhsyYD7z+IleukVF7uYvtq3z7uP3a88yYirNLLDaqZszQmG95kO6yHfOFL3/HletdrzzfjLnpLVe7cvX1dJkxX3vUd96/66p+MyZE3/Ox6uiEOzDqW+ZlTX+3GdPh6EBqBVeaAJCAogkACSiaAJCAogkACSiaAJCAogkACSiaAJCAogkACVqa3B4V1agvPbm47vzo+dmqPVF2oMsxq1jSPbvtJQRW+VKp6Ph4/bGRQ65ctz/ytBmzumBPwpakSweXXmZEkorB11nwmq32RObPffsJV66Hb7vLjPnxt73ClatUtH+nH3Y2PCjYudZ2trlSfe2f/tmMOeeiza5cbf12k8Kmjb6lMybGRs2Yw84J/BXH8/HIxKwr19p++1wd6Cy6cgVHk0K5rdeV62RxpQkACSiaAJCAogkACSiaAJCAogkACSiaAJCAogkACSiaAJCAogkACVrrCIpRtfrSH2VvLYcxp250FknSiLMDYU273e4z7FpgQzo0OmbGdHfaHQ+StKHT7i6688lxV64rN9gf+//IwSOuXJI9rssHfUt6rH7V5WbMn37+dleu9VvOMmNuvGyjK9cz0/b59eW7d7tydW+yl0B5x5UvcuXauPFsO8jRzSRJfevtjqburoOuXAeGRsyY2/fanXeS9IEf2m7GjE1VXLlW9dvdPn09Pa5cJ4srTQBIQNEEgAQUTQBIQNEEgAQUTQBIQNEEgAQUTQBIQNEEgAQtTW4P+f+WUnAsWSBJRcen3Ze09ET6Oe3B/kj8B/bZk9YlqafPnvi9fpW9ZIEkbejvNGNeuXXQleu7Tx8wY87v8026f+SgPZF5Vbdvcvu+0RkzZsdl57pyrXc0A3zqjsdcuQZXrzJjDjzrm6z9k6+3J2tv3LjBlavqaOqoVn0Tv2/da5/T773WHrskffyWB+xcP7DNlWtyym5K6e7wlaKBvud24roHV5oAkICiCQAJKJoAkICiCQAJKJoAkICiCQAJKJoAkICiCQAJKJoAkKD15S5qS3fpVKp2d44kjU9MmTH/+sSQK9ezh0bNmE29vg6X9YNdZkx70e5ckaSRsUkzZnJy2pWru2h3khSCb0mP3YcmzJgQfd1YvUV7m+et63PlOjRqH69rtvq6sTq67G6sq87a6sp17oY1Zkyj4Tv2jXrdjJmu2TGStLbXXu5i94j9WEtSCPZ5X273PYema/ax6Ov2da+VSnbr4KEhX504WVxpAkACiiYAJKBoAkACiiYAJKBoAkACiiYAJKBoAkACiiYAJGhtcrui6sbk3Lrj4/wladYxf/cVW1a7cmlzvxnyN1/Z40q1etCeuD5d8+3j+Pi4GdPZXnbl2rZpkxlTGRt25TpvlX0aTM74mhSeckyUtxoi5mxyLNfx5V2HXbmqdTvuXVduceVq67In57d32JPpJak2Yy9lMVPzNU9cfo593leqvonyP/Uae3mQGHzj6u+wr80axrI5c2Zm7KUz9u63l4JpBVeaAJCAogkACSiaAJCAogkACSiaAJCAogkACSiaAJCAogkACSiaAJCgpY6goKBCYem6W2n4ZvoP9nebMaNjY65cni6kwS57aQBJuvuZI2bMlb1294Qk3fLoITPm7Vde6Mo11mUvuXDRWYOuXHd/c6cZc8V633IELz3bXn7i28/6Hsd7DtqdQ2+59BxXrrv2jJgxO0d91xAvG+gxY4J8y13MVOxtXjBoby/LZR+v3k7feR8K9vP2wJjdnSNJBUeuirNzsDpmd5wdPuzrhDtZXGkCQAKKJgAkoGgCQAKKJgAkoGgCQAKKJgAkoGgCQAKKJgAkaHm5i1pj6Um8XW2+ulxv2JNbvUsI7B6yJ09/+2nfBOs3X77ZjLl4vT2hW5Jua7cn8J+9zjch/bI1juUInL8Tt269wIy5YptvH2er9uN4wUW+cY1W7FxTQ/tduaaesJe7+OAVW1y5Ko7lTTrLvqaONb3tjijfRPm2or1UysFx34T0dX32c21tj2+ifKVmL7Ex61yGY6Zqj396etqV62RxpQkACSiaAJCAogkACSiaAJCAogkACSiaAJCAogkACSiaAJCAogkACVpe7qJkfJT9bNXXzVAq2vU7BF+NP6vX7mZob/PteptjXB/9x++6cl1/md3t410e5OCRSTOm5lxC4NXn2+OqOn+/dnTY43esfiBJ6nassHHXUzOuXG/evt6MOTzl60rZ3GPvQM3RNST5zvsjU1VfLsdDtM7ZxVOr2UtnVOu+53bd6BpMUavax6LR8D2OJ4srTQBIQNEEgAQUTQBIQNEEgAQUTQBIQNEEgAQUTQBIQNEEgAQtTW6v1WsaGRleMiaUHTOUJXW0+ybdehTLdq7rtp/tyvWlh+zlFN7zigtduYZGRs2Y7nbfQ9LdacdVnZPbY7TjisGXq1KxY4YmfUsulIpFM6av7Pu9/5VdB8yYD5xnL/shSZWqPfHbu0RFpW7n6u2wj4MkeXoGJqd9x77Dcx46lqiRpCnHNouOx1qS6nV74ronphVcaQJAAoomACSgaAJAAoomACSgaAJAAoomACSgaAJAAoomACSgaAJAgtY6gmp1DY8cWTJmzZo1rlxBdhfPTMXTiSHVHB/Df9iZ6/UvWmvG3Pe03W0iSW1ydFAEX2fEdNXOVa35OiPKBft4Tc06ly0p2adUt3OpkUNTdifJmCNGkq4+Z8CMCcG3Dke1arc9eXN5urGqjqUnJOeSMdF3TszO2HEHj0y4cnmW4fCe96rZj3fB2Y11srjSBIAEFE0ASEDRBIAEFE0ASEDRBIAEFE0ASEDRBIAEFE0ASNDS5Pa2clkbz96wZEwMy1eXCwVfrkbDngz88EHfpOhL1tuTlLcM9LhyDR+xl7vYNzblyrVpsM+M6XIuneGZh133TMyXVHMsNdBo+CYfdzuWQKl1+ZZJmXU0AwxPV1251vfY25yq+CaR1x3LXXjP+yA7l7fhwbPFrg7nEjWOCfyl4DsnxhxLZ8TI5HYAOG1QNAEgAUUTABJQNAEgAUUTABJQNAEgAUUTABJQNAEgAUUTABK01BEUCkFt7e1Lxsw4OyMqNbtroObsJJl2dKUcGZ525Tp722o7yPlR/at7O82YXucjUnB08Yw7O1yCo+Ok7FhKQZIqjqVGetp9x6vL0U22v+5bVqJYsLfZVfTlGncssTEx6zv27Y7jOtDtOymKjvHP+nbR11Xj7PbzdPt4OqMkqVKxj2vd8fxvBVeaAJCAogkACSiaAJCAogkACSiaAJCAogkACSiaAJCAogkACVqa3B5jVK269KTUovNj7GdcH8Pvm5m7qqNsxrzjVS9y5SrXK2bM4VnfxNz+jm4zJjgnDE/O2s0AXW2+SeQFxwTrmmPSuiQNeJafcCx/IEmTM/ZE5oGupZsr5lQq9uPYXvKdX6WSvY/lku9xfHD/ETPmEufj2LB3UW0F3+NYdTzeZWf1qDXs41oo+pLVG/a5w3IXAHAaoWgCQAKKJgAkoGgCQAKKJgAkoGgCQAKKJgAkoGgCQAKKJgAkaKkjSNH+aHnnahfyND04V7tQzfFx92v6el25esr2RkPF97unX3aHy8jkjCvX+ja7K2Vsynfw+zrtDqrY8OWqOZqjip61OiSNTNvHoiBfd9HMrL1ExdSMHSNJHWX78a7XfePa2mc/jlOO4yBJ4zP2wV/T1+HK5XmM6q4uvmxZHEv0lqLnuNvHgytNAEhA0QSABBRNAEhA0QSABBRNAEhA0QSABBRNAEhA0QSABK1NbpcUtfRk00LBV5fLjiUXqs7JtG1FezJtT6dvmYTZmj1JeWOvY4kHSVPT02ZMaXbSlev7+4fNmBcN2strSNKRCXvSfYfzTNk7bI9/VbfveHUb55YkDc3aY5ekesWeIH7g0GFXrlLJbgbocR6woUl7jYr+Tt/xWuVYamRi2ne8VvXYz4+as+Gh4FjuotbwLRlzOuBKEwASUDQBIAFFEwASUDQBIAFFEwASUDQBIAFFEwASUDQBIAFFEwAStNQR1IhRs5WlOwzay3b3RMZe76Lk6BqSpHrD7uIZ6PKNa7ZidyrM1uyuDknq7+0xY6ozU65cZ5XtbozJKV+uDsdaIzQ3NkMAAAlkSURBVNPTvuUbBrvtU6oYfEsWVBwdYKt6u1y5PAts9PX4loIole3OG++SHud22LkKRd/TtKPNjut2LhdRi/b4O9sda9RIKhTsuOgcV7ls72Mh+I79yeJKEwASUDQBIAFFEwASUDQBIAFFEwASUDQBIAFFEwASUDQBIEFLk9tDsJep8M4zLTgmAzcavgmwivZE7Ebd91H9Zcf83eiYTC9JNcdSFl0dvmU4Co6Hznu4PE0D7W2+B9JzKIJz4ndHhx3nPb+6ux2T4IPvGqJYso+995yIjnPV2yDimSDecE4i7263t+lM5Wo2KTqXxfE83NYSPK3iShMAElA0ASABRRMAElA0ASABRRMAElA0ASABRRMAElA0ASABRRMAEgTvx8wv+sMhHJL01PINBwBOmc0xxrWpP9RS0QSAFxpengNAAoomACSgaAJAAoomACSgaAJAAoomACRo6ZPb3/jG6+LQ0JAkHf2s5ND09WJic3D+RQhh3qdON/98IYT5nzadfxnC4p9MXljk9rkfLxTmPll8fkChEI5+Kvyi216YT1KxEFRfZAALb5/b32JBqi+y7RPlipJKhaBaI847XtntBdUW+TTs4+6Lx/5TLgZV61HNd8x9Vy4WVK0f+5nFbm8eXbkYVK0d//nYUVHtxYJm6415eSSpvVTQbG1Brnjs9uNzSR2lgmZqjXkfET7v9qYbo6TOckHT1caCPNnPdpaKmq7WF9wndZabbm8ac1c5u56YqjaOjnhuGF1tRU1Vjs/VVV54+7Hj3N1W0mSldty5NJdr/sy/qO72kiZna8duabq/+b55t3cUNTlT17yNRKm7o6TJmdoiZ57U017WxEz1uOPf01HSxEztuI9n7+koa3ymujCNpKiezrImpqvz74pSb2dZ49NVHbfzMaq3q03jU5XjxtXb2abx6cpxteJY/PHPy+y+2ePG1dvVofGpmXnb6O3q0N6nH/tyjPE6JWqpaA4dHtJtd96jo6dmXhxqjfyWuSdtzL4vzT1xjz6Zo2LMnoSVo7dnueYeq7a5J1ucf3t7uaCZ6rEn59z2OuaeOPHYE2au7h07Qeff3t1e1ORs/bhc3e1FTRy9ff7P9HaUND5TOy5XX2dJY9O143L1dZY0Op3FZ7dn/9dQ1EBXWSNT1aPbmTteq7rbNDxZOZp77mcHu9s0NFHJYo8ek6iGpLU97To4PpttuWkM63radWB8dl6eue2c1dehfWMz2fcxWxIhSjq7r0PPjs4cHdNc/T6nv0PPHJk+Ft+Ua/OqTu0ens5vj0djzlvdpScOTykqHs0To3T+YJceG2q6PR4b37a13Xrk0GT2fTy2jRet7dbDByeP24+L1vXo+wcm5uWY++8lZ/XqwX3jR79Xvp87NvRq597x7PgdPQWjLj27T5J0/7NjTdvPfvbFG/t1357Ro7cr385LNvXru08fmXfb3LhftnlA33lq5FiePOZlmwd0z+6Ro3HZ4xZ15XmrdfcTw8ce28ax+6/aulp3PT583PavvmBQdz56OB9ntjMxRl2zbY3ueGQofxyP3S5JP3DhWt226+Cx2/Of+8GL1ulb3z9w7GIm/5lXXLxe3/ze/mPP76PHJeqV28/SrQ/tO3Z7dqeu3bFBt+7cm+/A3AOZ7dC1l23UrQ/s0dEDELNfUtdedq5uvf+pY7fP5bp8s269d/fRn2++/9qXnKdbv/v48T/zsvN16z2PNd3W0LVXbNN/+/Bb1ixW1yy8PAeABBRNAEhA0QSABBRNAEhA0QSABBRNAEhA0QSABBRNAEhA0QSABBRNAEhA0QSABBRNAEhA0QSABK0u4fugpBkz8PlljaShlR7EKfZC2+cX2v5KL8x97ogx7kj9oZY+Gk7STIzxihZznFFCCPewz89vL7T9lV64+3wyP8fLcwBIQNEEgAStFs2PLcsozizs8/PfC21/JfbZraU3ggDghYaX5wCQwFU0QwjXhRB2hRAeCyH84iL3t4cQPpPff1cIYctyD/RUc+zzfwwhfC+E8EAI4WshhM0rMc7lYu1vU9w7QggxhHDGv9Pq2ecQwk354/xQCOFTp3qMy81xXp8bQvhGCOHe/Nx+00qMc7mEEP5PCOFgPj1ysftDCOEj+fF4IITwUjNptqLdif9JKkp6XNJWSW2S7pe0fUHMz0j68/zrd0v6jJX3dP7n3OcfktSVf/3hM3mfPfubx/VKulXSnZKuWOlxn4LHeJukeyWtyr9ft9LjPgX7/DFJH86/3i5p90qPu8V9vlbSSyU9eIL73yTpS8pWH79a0l1WTs+V5lWSHosxPhFjrEj6tKQbF8TcKOkT+defl/TaEEJw5D5dmfscY/xGjHEq//ZOSeec4jEuJ89jLEm/Kel39PxoaPDs8wclfTTGOCJJMcaDp3iMy82zz1FSX/51v6S9p3B8yy7GeKuk4SVCbpT0yZi5U9JACGHDUjk9RXOjpGeavt+T37ZoTIyxJmlU0qAj9+nKs8/N3q/st9WZytzf/GXLphjjF0/lwJ5Dnsf4QkkXhhBuCyHcGUK47pSN7rnh2edfl/QTIYQ9kv5Z0s+fmqGtmNTnessdQS94IYSfkHSFpFet9FieKyGEgqQ/kPS+FR7KqVZS9hL91cpeSdwaQrg0xnhkRUf13PpRSX8VY/z9EMI1kv46hLAjxthY6YGdLjxXms9K2tT0/Tn5bYvGhBBKyi7rDy/HAFeIZ58VQnidpF+RdEOMcfYUje25YO1vr6Qdkm4JIexW9refm8/wN4M8j/EeSTfHGKsxxiclPaKsiJ6pPPv8fkmflaQY4x2SOpT1pT9fuZ7rzTxF89uStoUQzgshtCl7o+fmBTE3S3pv/vU7JX095n9lPUOZ+xxCeImkv1BWMM/0v3Utub8xxtEY45oY45YY4xZlf8O9IcZ4Ur27pwnPef0Pyq4yFUJYo+zl+hOncpDLzLPPT0t6rSSFEC5WVjQPndJRnlo3S3pP/i761ZJGY4z7lvwJ5ztQb1L2W/ZxSb+S3/Ybyp44UnZgPyfpMUl3S9q60u+aLcO7btY+/4ukA5Luy//dvNJjfi73d0HsLTrD3z13PsZB2Z8lvidpp6R3r/SYT8E+b5d0m7J31u+T9IaVHnOL+/u3kvZJqip75fB+SR+S9KGmx/ij+fHY6Tmv6QgCgAR0BAFAAoomACSgaAJAAoomACSgaAJAAoomACSgaAJAAoomACT4/5uGcgx2tIkEAAAAAElFTkSuQmCC",
            "text/plain": [
              "<Figure size 432x432 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 432x432 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 432x432 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 432x432 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ],
      "source": [
        "print('Predicted:', classes[predicted[ind]], \n",
        "      ' Probability:', torch.max(F.softmax(outputs, 1)).item())\n",
        "\n",
        "original_image = np.transpose((images[ind].cpu().detach().numpy() / 2) + 0.5, (1, 2, 0))\n",
        "\n",
        "_ = viz.visualize_image_attr(None, original_image, \n",
        "                      method=\"original_image\", title=\"Original Image\")\n",
        "\n",
        "_ = viz.visualize_image_attr(grads, original_image, method=\"blended_heat_map\", sign=\"absolute_value\",\n",
        "                          show_colorbar=True, title=\"Overlayed Gradient Magnitudes\")\n",
        "\n",
        "_ = viz.visualize_image_attr(attr_ig, original_image, method=\"blended_heat_map\",sign=\"all\",\n",
        "                          show_colorbar=True, title=\"Overlayed Integrated Gradients\")\n",
        "\n",
        "_ = viz.visualize_image_attr(attr_ig_nt, original_image, method=\"blended_heat_map\", sign=\"absolute_value\", \n",
        "                             outlier_perc=10, show_colorbar=True, \n",
        "                             title=\"Overlayed Integrated Gradients \\n with SmoothGrad Squared\")\n",
        "\n",
        "_ = viz.visualize_image_attr(attr_dl, original_image, method=\"blended_heat_map\",sign=\"all\",show_colorbar=True, \n",
        "                          title=\"Overlayed DeepLift\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "82WxLVehdU4e"
      },
      "source": [
        "# Layer Attribution Techniques"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 21,
      "metadata": {
        "id": "v2VvCc6jgxYB"
      },
      "outputs": [],
      "source": [
        "from captum.attr._core.layer.layer_conductance import LayerConductance\n",
        "from captum.attr._core.layer.layer_activation import LayerActivation\n",
        "from captum.attr._core.layer.internal_influence import InternalInfluence\n",
        "from captum.attr._core.layer.layer_gradient_x_activation import LayerGradientXActivation"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 22,
      "metadata": {
        "id": "3nMu5xWPrjCa"
      },
      "outputs": [],
      "source": [
        "def visualize_importances(feature_names, importances, title=\"Average Feature Importances\", plot=True, text=True, axis_title=\"Features\"):\n",
        "    if text:\n",
        "        print(title)\n",
        "        for i in range(len(feature_names)):\n",
        "            print(feature_names[i], \": \", '%.3f'%(importances[i]))\n",
        "    x_pos = (np.arange(len(feature_names)))\n",
        "    if plot:\n",
        "        plt.figure(figsize=(12,6))\n",
        "        plt.bar(x_pos, importances, align='center')\n",
        "        plt.xticks(x_pos, feature_names, wrap=True)\n",
        "        plt.xlabel(axis_title)\n",
        "        plt.title(title)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 23,
      "metadata": {
        "id": "1X19ZN8FdaR-"
      },
      "outputs": [],
      "source": [
        "layer_cond = LayerConductance(net, net.fc3)\n",
        "attr_lc = layer_cond.attribute(input, target=labels[ind].item()).detach().numpy()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 24,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 595
        },
        "id": "TpLVblIZr4Mo",
        "outputId": "29f2a23c-6606-4dcc-a068-56f87c3fbd0b"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Average Neuron Importances\n",
            "0 :  0.000\n",
            "1 :  0.000\n",
            "2 :  0.000\n",
            "3 :  0.000\n",
            "4 :  0.000\n",
            "5 :  0.000\n",
            "6 :  0.000\n",
            "7 :  8.733\n",
            "8 :  0.000\n",
            "9 :  0.000\n"
          ]
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 864x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ],
      "source": [
        "visualize_importances(range(10),np.mean(attr_lc, axis=0),title=\"Average Neuron Importances\", axis_title=\"Neurons\")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 25,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 545
        },
        "id": "E5dxQ4vCr63T",
        "outputId": "58048656-4ea9-49fd-9d38-013d96bcab3b"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ],
      "source": [
        "plt.hist(attr_lc[:,0], 100);\n",
        "plt.title(\"Neuron 0 Distribution\")\n",
        "plt.figure()\n",
        "plt.hist(attr_lc[:,7], 100);\n",
        "plt.title(\"Neuron 1 Distribution\");"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 26,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 404
        },
        "id": "fjIT0V1ttl_E",
        "outputId": "11139602-06fe-4d1e-81cf-b941aa44ba4f"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 864x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ],
      "source": [
        "layer_cond = LayerConductance(net, net.fc2)\n",
        "attr_lc = layer_cond.attribute(input, target=labels[ind].item()).detach().numpy()\n",
        "visualize_importances(range(84),np.mean(attr_lc, axis=0),title=\"Average Neuron Importances\", axis_title=\"Neurons\", text=False)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 27,
      "metadata": {
        "id": "kQdtGktmfhIG"
      },
      "outputs": [],
      "source": [
        "layer_act = LayerActivation(net, net.fc3)\n",
        "attr_la = layer_cond.attribute(input, target=labels[ind].item()).detach().numpy()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 28,
      "metadata": {
        "id": "aAayuns3f5kS"
      },
      "outputs": [],
      "source": [
        "layer_int_inf = InternalInfluence(net, net.fc3)\n",
        "attr_lii = layer_int_inf.attribute(input, target=labels[ind].item())"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 29,
      "metadata": {
        "id": "RHzNNOSWgCAK"
      },
      "outputs": [],
      "source": [
        "layer_ga = LayerGradientXActivation(net, net.fc3)\n",
        "attr_lg = layer_ga.attribute(input, target=labels[ind].item())"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "fEJk3lGCdbIc"
      },
      "source": [
        "# Neuron Attribution Techniques"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 30,
      "metadata": {
        "id": "ptFZSNj8wIfu"
      },
      "outputs": [],
      "source": [
        "from captum.attr._core.neuron.neuron_gradient import NeuronGradient\n",
        "from captum.attr._core.neuron.neuron_integrated_gradients import NeuronIntegratedGradients"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 31,
      "metadata": {
        "id": "k9tSRDihunXM"
      },
      "outputs": [],
      "source": [
        "neuron_g = NeuronGradient(net, net.conv1)\n",
        "attr_ng = neuron_g.attribute(input, neuron_selector=(4, 1, 2))\n",
        "attr_ng = np.transpose(attr_ng.squeeze().cpu().detach().numpy(), (1, 2, 0))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 32,
      "metadata": {
        "id": "f7_YS-MEweE8"
      },
      "outputs": [],
      "source": [
        "neuron_ig = NeuronIntegratedGradients(net, net.conv1)\n",
        "attr_nig = neuron_ig.attribute(input, neuron_selector=(4, 1, 2))\n",
        "attr_nig = np.transpose(attr_nig.squeeze().cpu().detach().numpy(), (1, 2, 0))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 33,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "Pcye2X_2xlJv",
        "outputId": "d6758874-a022-485a-ad98-3c3825b519a7"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 432x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        },
        {
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAU0AAAF1CAYAAABs2bgEAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAfC0lEQVR4nO3dfZBdd33f8c93nx+0WkkrWZYNtjHYGNshPBpoGydt0sAwFDKkZUhLmqQOGZw27aTNtGkZKEM7DUnbZKYZZ1IyJZBMk9C0TcbNQ9uQ4KjQGOoANgQDlm3JT3qWVrvaB+3Tr3/cs9qr1WrP56t7tCt5368Zwe65v/s7v3PuuR+f3f197y9KKQIAeLo2ewAAcC0hNAEggdAEgARCEwASCE0ASCA0ASCB0MSWEhE/HBGfa/v+bETcupljwrWF0HyRi4iDEXEsIobbtv1oRDy0icNaU0Tsi4hfiYgXqjB7KiI+GRF3XKl9llK2lVKe6rSfapz/uokx4epGaG4N3ZL+0ZXcQUT0dPj8MUn/V9KQpO+QNCLpdZL+VNJfvxL7BC4Hobk1/FtJPxURO1Y/EBF3RMQfRcSpiPhmRLyn7bGHIuJH275f/aNtiYi/HxFPSHqi2vb+iDhQ9fdgRNywqv0HIuKJiBiPiAciIqqHf1LShKQfLKU8WVrGSym/Wkr5xer5t1R93BcRz0j6k2r7b0fEkYg4ExH7I+Kutn2OVeOYiIgvSnr5quMvEfGK6uv+iPh3EfFMRByNiF+OiMHqse+KiOci4p9Ud+6HI+JHqsd+TNLfkfRPqzvk/3EZrxGuEYTm1vCIpIck/VT7xupH9j+S9BuSrpP0Xkm/FBF3Jvr+PklvknRnRPw1ST8j6T2S9kk6JOm3VrV/h6Q3Snp11e6t1fbvkfQ7pZQlY5/fKelVbc/9Q0m3VcfwJUn/ua3tA5Jmq/H8verfpXxM0u2SXiPpFZJulPThtsevlzRabb9P0gMRsbOU8vFqnz9X/bj/N4xjwDWK0Nw6PizpJyJiT9u2d0g6WN3NLZRSvizpv0n6W4l+f6aUcqqUMqPW3dYnSilfKqWck/TPJb0lIm5pa/+x6g7yGUmfVSugJGm3pCPLjSLindXd6GRE/O9V+/xIKWWq2qdKKZ8opUxW+/yIpG+PiNGI6Jb0/ZI+XLX/mqRPrXUQ1R3vj0n6yep4JiX9G7X+Q7JsXtJHSynzpZQ/kHRW0isT5wovAoTmFlEFxu9J+um2zTdLelMVTuMRMa5W8F2f6PrZtq9vUOvucnmfZyWdVOvObNmRtq+nJW2rvj6p1t3g8nMfLKXsUOvH9r5L7TMiuiPiYxHxZERMSDpYPbRb0h5JPavGeEhr26PW71P/vO1c/M9q+7KTpZSFS4wfWwShubX8S0nv10qIPSvpT0spO9r+bSul3F89PqVWkCxbK0zbPybrBbWCWNL5H//HJD1vjO2PJX1fRDjXZPs+/7akd6n14/2opFuWdy/puKQFSS9ta3/TJfo8IWlG0l1t52K0lOKGIh8XtkUQmltIKeWApE9L+ofVpt+TdHtE/GBE9Fb/3hgRr6oe/4qkd0fEUPXHkvtqdvGbkn4kIl4TEf1q/Xj7hVLKQWN4Py9pp6Rfj4iXR8uIVn58v5QRSefUulMdqva5fLyLkv67pI9Ux3CnpB9aq5Pqd6m/IukXIuI6SYqIGyPirWu1X8NRScz33AIIza3no5KGJan6vd33qvV7uxfU+tH5ZyX1V21/QdKcWoHwKV34B5aLlFI+I+lDav1e9LBaf6l+73rPaXvuCUlvVuuPNp+TNKlWaI9Iun+dp/6aWj9yPy/p65IeXvX4P1DrR+gjkj4p6VfX6eufSTog6eHqR/3PyP+d5X9S649h4xHxu+ZzcA0KPoQYAHzcaQJAAqEJAAmEJgAkEJoAkEBoAkBCR58Ss21kpIyNja2/g+7mPohmcXHRare0VN9ueHi4to0kDQwO1rZZmJ+3+jo3N1fbxh3XvLHPvt7VhTRri66ob7QpkyycnRpjb3R/TTPG3+QhboKZ6ZnaNs571jUwMGC1e/TRR0+UUvbUt7xQR4k2Njamf/HBD63bZufOnVZfKx92c2kTExNWX2fPnq1tc88991h93XXXXbVtjh49avV18ODB2jZvetObrL6OHDlS2+aGG26obSNJ/f39tW38qWnNBU9ZMvoyrht/h97Ym4xW57rv7u5ubH/u6+iMy+3rscceq20zNTVl9eW44w7v41f37hm7VEntuvjxHAASCE0ASCA0ASCB0ASABEITABIITQBI6GjKUSlFCwsL67aZmamfoyVJx09P1raZnvPmcnUvzta2OfTcYauv6Ko/RRNnTlt9HTt2rLbNqVOnrL6OHz9e22ZkZMTqy5nSsrTkLN2z8ZocV1eXdw/hTLVxpuy47Tbjk8i6jLm7k5P1U/sk77p32kjS7Gz9e3t6etrq63JxpwkACYQmACQQmgCQQGgCQAKhCQAJhCYAJBCaAJBAaAJAAqEJAAnNfaz6JbgfoDpzrv6TyMdGt1l9TYyfq29z1vvQ0wNPP1PbZm76jNWX82nrTzzxhNVXXSWWJI2Pj1t9nTtXf7527dpl9eVU1TRZLeNUiEjedeh8GLPU8AcHO9U+DX44snvuJybqK/Qef/xxq6/Tp+sr5tzKQefDig8cOGD1dbm40wSABEITABIITQBIIDQBIIHQBIAEQhMAEghNAEggNAEgoePlLuombDuTUSVppL8+vx97wVsK4vSZido2x5+qn7wrSa+/pX5S975+b8kFZ+L30aNHrb5GR0dr27hLCDjjcvbn9uUu3+C0cye3DwwMWO0czmRtd9mSsbGx2jbuue9xJ9Qbnn322do2TRZiuJPue3rqI8vZXye40wSABEITABIITQBIIDQBIIHQBIAEQhMAEghNAEggNAEggdAEgISOK4Lm5ubWbbO4uGj1dWxi/X4k6fd/Z7/V14//wL21bXr7B62++nrq/7uysOBVBDnnoqvLq4zo6+urbeNUrrh9uRUbS0v158JZ9kPylp9wl/Rwqovca/WZZ+qXQDl+/LjVl7OMyN69e62+tm2rXw7mhhtusPqanKyvmHOWSZG8c+9eE03trxPcaQJAAqEJAAmEJgAkEJoAkEBoAkACoQkACYQmACQQmgCQ0NHkdpVSOyHYnWj60IH6ZR7e9+76SeuSpO7e2ibPnD5rdXXi7Extm3tu3m315XwM//Bw/QRlyTuvZ40JypLUbSwhcOLECauv3t76c19XELHMmVD/5JNPWn05+3SXZnGW2HAnfjv7dJdAcYoUbr31VquvmZn6694tBnDOvTu53VnugsntAHAVITQBIIHQBIAEQhMAEghNAEggNAEggdAEgARCEwASCE0ASOisIsjgLH8gSe9+9b7aNl1dXsb39tYf1iuHd1h9vWJxpLaNO67BwfolNpxKDMmrOJkzqyy6jMqOb3zjG1Zf3d3dtW2cyijJO0a3isepXnH7co7Rqc6Rml3mwam8efrpp62+hoaGatu4r6NToeO+h5xrwu3rcnGnCQAJhCYAJBCaAJBAaAJAAqEJAAmEJgAkEJoAkEBoAkBCR5Pbi+onrjrLH0je5GN3IrAzod6ddO+O3+Ec48mTJ62+BgYGats4SwNI3uRjd7kLZ5/u6+ic++HhYasvZ5/upGjndXSW6pC8ifLu67hr167aNs5SHZI3iXzHDq9AZHp6uraNW9ThXBNuYcHl4k4TABIITQBIIDQBIIHQBIAEQhMAEghNAEggNAEggdAEgARCEwASOqoICkVtRYNbeeO2czgVLk4lhuSNy60kcasxHM7SGW5VinOMzv5cboVLf39/bRunOkeSFo2lGZwlHprmnFenOkfyzpdb4dbkMhyjo6ON9eUsseFe95eLO00ASCA0ASCB0ASABEITABIITQBIIDQBIIHQBIAEQhMAEjpc7qJYk02tvowJ6S5n4nqT+3Mn046MjNS2cSYoS94kZXfit3Mu3CUEnMna7iT/qamp2jbuMc7NzdU3Ml9HZ6kRd3K4M35r7PImwbsTzbdt21bbpsnlZ9yCBydvmnxvr4U7TQBIIDQBIIHQBIAEQhMAEghNAEggNAEggdAEgARCEwASCE0ASOioIqirq0vDw8PrtnE/Xt+pCHCXqHD66uvzxtXX19wSAk5VjXuMTpVIk0sWuONyqjHc6g9n+Ql3qRFnXG4liVOp5C5R4bRzX0en0sq9VjdakxVUTVUpXgp3mgCQQGgCQAKhCQAJhCYAJBCaAJBAaAJAAqEJAAmEJgAkdDS5fXBwUHfddde6bdzJxw7nY/Mlb/kJd4kKZ9KtO2HYmSDuni9nWYwmP/a/yXPvTpSfnp6ubXPmzBmrL2dczhIPkjcuZ2K+JG3fvr22jVsMcOjQodo2ztglae/evVY7R5OFBc51yHIXAHAVITQBIIHQBIAEQhMAEghNAEggNAEggdAEgARCEwASCE0ASOioIigUtdUKbuWN067JqoGZmRmrr+eee662zY033mj1tXPnzto2zjIWkrfkglvFU7dkieSf+yarMU6dOlnb5ujRY1ZfTuWNq6ur/lp1K4LOnj1b28atCBodHa1t41YEOZVw7ricJSquJdxpAkACoQkACYQmACQQmgCQQGgCQAKhCQAJhCYAJBCaAJDQ2eT2rlBfX9+6bdwJ1g63L2fJiGPHvEnRziT4sbExqy9nwvDk5KTV1+nTp2vbuBPNnfPlLunhHOO5c+esvoaG6ifd79m92+pryTgXbsGDs9TI5MSE1ddR4zrcbR5j3XtR8l4fybsOncn0riaLYBYWFjodzrq40wSABEITABIITQBIIDQBIIHQBIAEQhMAEghNAEggNAEggdAEgISOKoJUml3eoI5bNeBUBBw/ftzqa9/119e2cSoxJOnkyfrlG06cOGH1NTs7W9vGHZdzvtylDZzlOtwlPZx9DgwOWn3NzdVXIc3Pe5Uk88b4TxivteQtBeEsbSJ5FXNuVZ1THeUskyI1W6HjVqZdSdxpAkACoQkACYQmACQQmgCQQGgCQAKhCQAJhCYAJBCaAJDQ0eT2olI7Oded/O60c5ZlkLzJ0+7SBs4yCUePHrX6cia3uxOZnQnD7rl3Jso7k7Al77y6y104r7c76b67u76v7u5uqy9nqRH3ddy3b19tG3dZifHx8do27kRz59pxrwnndXQLHpzJ+du3b7f6ulzcaQJAAqEJAAmEJgAkEJoAkEBoAkACoQkACYQmACQQmgCQQGgCQELHy100VRHkcD+q36lUcMc1PT1d22ZiYsLqy6m8cauenPE7+5O8SqX+/n6rL2dcbkWQ05e7BMrQ0JDVzuFU++zevdvqa8+ePbVt3GtibGysts2pU6esvpxr2h3Xjh07atu47+2BgYHaNu4yHJeLO00ASCA0ASCB0ASABEITABIITQBIIDQBIIHQBIAEQhMAEjqb3K76ycVNTtZ2l6hwJqQ3uRSEu4SAcy7c8+VM4HfH5XD7ciYpu+d+fn6+to17vnp7e2vbuMUAg4ODtW2uv/56qy+Hcx4kb6K/MwFe8pZwcZe7cJeycFzpiesO7jQBIIHQBIAEQhMAEghNAEggNAEggdAEgARCEwASCE0ASCA0ASCho4qgolJbKeJWkjjVPk1WILgfr++0c/tyKk7cqpQmqyyc8bvjcl6j7u5uqy+nEqanx7uEnSU23Iqzffv2We0cTVZtORVBbjXWrl27Gtmf5FVtudeE05ezfEsnuNMEgARCEwASCE0ASCA0ASCB0ASABEITABIITQBIIDQBIKGz5S5KqZ3M7E5Idya3upPInQnP7sRcpy93yYXJycnaNu4kX2c5hSNHjlh9ORO/3fPV5DIcTl/uZO2pqanaNn19fVZfzsTv/v5+qy/nGJtcMsZ9DzlLejTJvb6cgodjx451Opx1cacJAAmEJgAkEJoAkEBoAkACoQkACYQmACQQmgCQQGgCQAKhCQAJnVUEKWqrFXp7e72BGJU37hIPS0v1VRZuBYLDrSRxjI6OWu1uuumm2jYjIyNWX4cPH65t41TUSM1WUDlVNe414VTLXHfddVZfAwMDtW3cypsmr0OHe+6ddk1Wibl9Odfh6dOnrb4uF3eaAJBAaAJAAqEJAAmEJgAkEJoAkEBoAkACoQkACYQmACR0OLm9njvJ15l8PDQ0ZPU1N1d/WO6yEs7SDM6EbsmbbL5jxw6rL8fu3butdtu2batt4yyJIXmTlN0J1k7RwKFDh6y+JiYmatu4Szw416q7DEdT+3O5k8iddk2Oy82J2dnZ2jYzMzOdDmdd3GkCQAKhCQAJhCYAJBCaAJBAaAJAAqEJAAmEJgAkEJoAkEBoAkBCxxVBTVUFNFlJ4lT7uH053GoZp+rBrVSan5+vbeOOyzkXbjVWk9UyTjtnKQWXW9nlvI5NLnfhVvG4+3Q0We3TJOe6d6r4OsGdJgAkEJoAkEBoAkACoQkACYQmACQQmgCQQGgCQAKhCQAJHU1uX1hc0OnTp9dtMzw8bPXV29vbyVAu4CyT4O7PmeQ7Pj5u9eVMNncnMjuTut0Jyk5fc3NzVl9NFg04fTmTnaVmJ5E3OXnamZDe5Dl1J8A7+2zy+nKLOpzxN1nwsBbuNAEggdAEgARCEwASCE0ASCA0ASCB0ASABEITABIITQBIIDQBIKGjiqDFxUWdOXNm/R2YSwg47dwKBKeywx2XU/0xNTXVWF/uuJrU5DIJblWNo8nlLpy+3LG7VUgOZ/xuRZAz/iarZebmvOVUFhfrry+3Qs95r1ERBABXEUITABIITQBIIDQBIIHQBIAEQhMAEghNAEggNAEgoaOZ1L09vdq7d++6bdyPsXe4k3ydiczOkhgud1zOpGh3onmTk+CdvppchsM9X8643OvLadfk9eUWYjTJuXbccTl9dXd712BEfV/uEiKzs7O1ba70uedOEwASCE0ASCA0ASCB0ASABEITABIITQBIIDQBIIHQBIAEQhMAEjoqK+nqCg0MDKzbZjMqI5zqFbciqMklKpyP9D93zltCwBm/e+6dSpgmq2Xcqienush9HZ19ulUpTl/uMTb5/mhyqRGnL7cay3kdS/HOl1NVx3IXAHAVITQBIIHQBIAEQhMAEghNAEggNAEggdAEgARCEwASOprcXkr9hOAmJ++6E6ydibl1k/KXNTm5fXR0tLG+nGNs8ty7k7Wb5Jx795pw+mqyGMDtq8liAIc78du5vprsq7fXK1K4GpYa4U4TABIITQBIIDQBIIHQBIAEQhMAEghNAEggNAEggdAEgARCEwASOqwIKrWVFu5H4jdZ9eBUBAwNDVl9OeN3Kn0kaXx8vLaNu+SC87H/7jl12rnVH865d5dlcCpv3PM1NzfXSBup2aqUK700w2ru+3Gjl+FwK+E2ozJtNe40ASCB0ASABEITABIITQBIIDQBIIHQBIAEQhMAEghNAEjoaHK7o8mJzE1OBO7t7W2snXuMg4ODtW0mJyetvppc7sLpy51E7rRzz5czkdkd19TUVG2bU6dOWX319XlLMzTFnZDuvIe6urxzH1HflzvR3Hm5zZfRunbc6+tycacJAAmEJgAkEJoAkEBoAkACoQkACYQmACQQmgCQQGgCQAKhCQAJnS13ofrlLlxONYM703+jlyNwlp6QpOHh4do2bkXQxMREbRu3kqTJ6qImOfvs7++3+nKWJHGXXLCqxKyeJOesutVrTb6OzrXTZKWSy93nlcSdJgAkEJoAkEBoAkACoQkACYQmACQQmgCQQGgCQAKhCQAJV3y5iyaXXGhyQnqTS2csLHiT2ycm6vfZ5ERmV5OFBY4rvRzBWoaHhmrbdJkTp53XyF0KwrkON+N8OdeE+350zkWTE+CvdCEGd5oAkEBoAkACoQkACYQmACQQmgCQQGgCQAKhCQAJhCYAJBCaAJAQncyej4jjkg41NxwA2DA3l1L2ZJ/UUWgCwFbDj+cAkEBoAkACoQkACYQmACQQmgCQQGgCQEJHn9z+1re+rZw4cUKStDxxKdq+Xktpb1x9EREXfNpy+/O7IrTUPi2q+jJCWlpjR11rbF9+eleX1PoQ6QsbdHWFlqonrbnv1f1J6u4KLa4xgNXbl4+3u0taXGPfl+qrSOrpCi0slQvOV2t7lxbW+DTsix4rK//X2x2aXyxqf2D5u97uLs0vrjxnre3to+vtDs0vlIte56Ki/u4unVtcuqAfServ6dK5hVV9lZXtF/clDfR0aXZhaeUFXL29bWORNNjbpZn5pVX9tJ472NOtmfnFVY9Jg71t29vGPNTbup+Ynl86P+LlYQz1dWt67uK+hnpXb185z8N9PZqaW7joWlru68KZf0XD/T2aOrewsqXt8fbHLtg+0K2p2UVdsJMiDQ/0aGp2YY0rT9rW36uzs/MXnf9tAz06O7ugVQPTtoFeTc62rVTQ9mJuG+zV2Zn5Cx8q0shgryZn5nXRwZeikaE+TU7PXTSukcE+Tc7MXZQVK+0vfl+2Hjt30bhGhgY0OT17wT5Ghgb0wjMH/lcp5W1K6ig0T5w8oc8//IjOX5pVOCwsVVuW37Sl9X3P8hv3/Ju5qJTWm3Du/PZWX8uvVd/ym61cuL2/t0uz8ytvzuX9DSy/ccrKG2Y591Yu0Au3D/d3a+rc4kV9Dfd36+z57Rc+Z2SgR5OzCxf1tX2wRxMzCxf1tX2wR2dmWu1b21v/s6SiHUO9Oj09f34/y+dr53CfTk3Nne97+bljw306cXau1fb8OSlakrRnW7+OTZ5r7bltDNdt69fRyXMX9LO8n+u3D+jwxGzr+yItVY/fsH1Az5+ZPT+m5fx+yeiAnh2fWWnf1tfNOwd18NRMtb2cb/OyXUN66uS0isr5fkqRXj42pAMn2raXlfHdtmdY3zo+1fq+rOzjlXuG9Y1jUxcdxx3XbdPjR89e0Mfy/991/Yi+dnjy/PeqjvPufSP66guTrfN3/hIs+rYbtkuSHn1+om3/red++42j+spzZ85vV7Wf1750VF96ZvyCbcvjfv3NO/Tnh06v9FO1ef3NO/TIwdPn27Vet6I3vmyXvvjUqZXXdmnl8Xtu3aUvPHnqov2/+RVjeviJk9U4WwdTStFbbtutP/vWiep1XNkuSX/p9j36/DePrWyvnveX77hOn3v86MrNTPWcv/Kqvfo/Xz+y8v4+f16KvuPO67X/Lw6vbG89qHvv3qf9X32hOoDlF7J1QPe++kbtf+w5nT8BpfUfqXtffZP2P3poZftyX6+5Wfu/fPD889sfv/e1L9P+Lz158XNe/3Ltf+RA27Yl3fuG2/Sh+9+xe61cq8OP5wCQQGgCQAKhCQAJhCYAJBCaAJBAaAJAAqEJAAmEJgAkEJoAkEBoAkACoQkACYQmACQQmgCQ0OkSvl+TNFvb8MVlt6QTmz2IDbbVjnmrHa+0NY95oJRyd/ZJHX00nKTZUsobOuzjmhIRj3DML25b7XilrXvMl/M8fjwHgARCEwASOg3NjzcyimsLx/zit9WOV+KYbR39IQgAthp+PAeABCs0I+JtEfHNiDgQET+9xuP9EfHp6vEvRMQtTQ90oxnH/I8j4usR8VhE/HFE3LwZ42xK3fG2tfv+iCgRcc3/pdU55oh4T/U6/0VE/MZGj7FpxnV9U0R8NiK+XF3bb9+McTYlIj4REceq6ZFrPR4R8R+q8/FYRLyuttPWinaX/iepW9KTkm6V1CfpUUl3rmrz45J+ufr6vZI+Xdfv1fzPPOa/Kmmo+vr+a/mYneOt2o1I2i/pYUlv2Oxxb8BrfJukL0vaWX1/3WaPewOO+eOS7q++vlPSwc0ed4fHfK+k10n62iUef7ukP1Rr9fE3S/pCXZ/OneY9kg6UUp4qpcxJ+i1J71rV5l2SPlV9/V8lfXdEhNH31ar2mEspny2lTFffPizpJRs8xiY5r7Ek/StJP6sXR0GDc8zvl/RAKeW0JJVSjm3wGJvmHHORtL36elTSCxs4vsaVUvZLOrVOk3dJ+rXS8rCkHRGxb70+ndC8UdKzbd8/V21bs00pZUHSGUljRt9XK+eY292n1n+trlW1x1v92PLSUsrvb+TAriDnNb5d0u0R8fmIeDgi3rZho7synGP+iKT3RcRzkv5A0k9szNA2Tfa93nFF0JYXEe+T9AZJ37nZY7lSIqJL0s9L+uFNHspG61HrR/TvUusnif0R8W2llPFNHdWV9QOSPllK+fcR8RZJvx4Rd5dSljZ7YFcL507zeUkvbfv+JdW2NdtERI9at/UnmxjgJnGOWRHxPZI+KOmdpZRzGzS2K6HueEck3S3poYg4qNbvfh68xv8Y5LzGz0l6sJQyX0p5WtK31ArRa5VzzPdJ+i+SVEr5M0kDatWlv1hZ7/V2Tmj+P0m3RcTLIqJPrT/0PLiqzYOSfqj6+m9K+pNS/Zb1GlV7zBHxWkn/Ua3AvNZ/17Xu8ZZSzpRSdpdSbiml3KLW73DfWUq5rNrdq4RzXf+uWneZiojdav24/tRGDrJhzjE/I+m7JSkiXqVWaB7f0FFurAcl/d3qr+hvlnSmlHJ43WeYf4F6u1r/lX1S0gerbR9V640jtU7sb0s6IOmLkm7d7L+aNfBXt7pj/oyko5K+Uv17cLPHfCWPd1Xbh3SN//XcfI1DrV9LfF3SVyW9d7PHvAHHfKekz6v1l/WvSPrezR5zh8f7m5IOS5pX6yeH+yR9QNIH2l7jB6rz8VXnuqYiCAASqAgCgARCEwASCE0ASCA0ASCB0ASABEITABIITQBIIDQBIOH/AwCjJCCQfqEFAAAAAElFTkSuQmCC",
            "text/plain": [
              "<Figure size 432x432 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 432x432 with 2 Axes>"
            ]
          },
          "metadata": {
            "needs_background": "light"
          },
          "output_type": "display_data"
        }
      ],
      "source": [
        "original_image = np.transpose((images[ind].cpu().detach().numpy() / 2) + 0.5, (1, 2, 0))\n",
        "\n",
        "_ = viz.visualize_image_attr(None, original_image, \n",
        "                      method=\"original_image\", title=\"Original Image\")\n",
        "\n",
        "_ = viz.visualize_image_attr(attr_ng, original_image, method=\"blended_heat_map\",\n",
        "                          show_colorbar=True, title=\"NeuronGradient\")\n",
        "\n",
        "_ = viz.visualize_image_attr(attr_nig, original_image, method=\"blended_heat_map\",\n",
        "                          show_colorbar=True, title=\"NeuronIntegratedGradients\")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "U5brves8P4lK"
      },
      "source": [
        "# Captum Insights"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 34,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "iAKtIFKjTG-i",
        "outputId": "76df35f9-3b1b-41ad-dbf4-dc6a5c45f686"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Looking in indexes: https://pypi.org/simple, https://us-python.pkg.dev/colab-wheels/public/simple/\n",
            "Collecting flask_compress\n",
            "  Downloading Flask_Compress-1.12-py3-none-any.whl (7.9 kB)\n",
            "Collecting brotli\n",
            "  Downloading Brotli-1.0.9-cp37-cp37m-manylinux1_x86_64.whl (357 kB)\n",
            "\u001b[K     |████████████████████████████████| 357 kB 39.6 MB/s \n",
            "\u001b[?25hRequirement already satisfied: flask in /usr/local/lib/python3.7/dist-packages (from flask_compress) (1.1.4)\n",
            "Requirement already satisfied: itsdangerous<2.0,>=0.24 in /usr/local/lib/python3.7/dist-packages (from flask->flask_compress) (1.1.0)\n",
            "Requirement already satisfied: Werkzeug<2.0,>=0.15 in /usr/local/lib/python3.7/dist-packages (from flask->flask_compress) (1.0.1)\n",
            "Requirement already satisfied: Jinja2<3.0,>=2.10.1 in /usr/local/lib/python3.7/dist-packages (from flask->flask_compress) (2.11.3)\n",
            "Requirement already satisfied: click<8.0,>=5.1 in /usr/local/lib/python3.7/dist-packages (from flask->flask_compress) (7.1.2)\n",
            "Requirement already satisfied: MarkupSafe>=0.23 in /usr/local/lib/python3.7/dist-packages (from Jinja2<3.0,>=2.10.1->flask->flask_compress) (2.0.1)\n",
            "Installing collected packages: brotli, flask-compress\n",
            "Successfully installed brotli-1.0.9 flask-compress-1.12\n"
          ]
        }
      ],
      "source": [
        "!pip install flask_compress"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 35,
      "metadata": {
        "id": "w5ZoCrStP6Qe"
      },
      "outputs": [],
      "source": [
        "from captum.insights import AttributionVisualizer\n",
        "from captum.insights.attr_vis.features import ImageFeature"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 36,
      "metadata": {
        "id": "2fDQsWq1P7OF"
      },
      "outputs": [],
      "source": [
        "def baseline_func(input):\n",
        "    return input * 0\n",
        "\n",
        "visualizer = AttributionVisualizer(\n",
        "    models=[net],\n",
        "    score_func=lambda o: torch.nn.functional.softmax(o, 1),\n",
        "    classes=classes,\n",
        "    features=[\n",
        "        ImageFeature(\n",
        "            \"Photo\",\n",
        "            baseline_transforms=[baseline_func],\n",
        "            input_transforms=[],\n",
        "        )\n",
        "    ],\n",
        "    dataset=testloader,\n",
        ")"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 37,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 17,
          "referenced_widgets": [
            "13ed2c1fd01b4ff88f580127eb748fca"
          ]
        },
        "id": "JWBjSE6ZQoHW",
        "outputId": "c931cdfb-4f7c-44de-c534-aa22cf710650"
      },
      "outputs": [
        {
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "13ed2c1fd01b4ff88f580127eb748fca",
              "version_major": 2,
              "version_minor": 0
            },
            "text/plain": [
              "CaptumInsights(insights_config={'classes': ('plane', 'car', 'bird', 'cat', 'deer', 'dog', 'frog', 'horse', 'sh…"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "visualizer.render(debug=False)"
      ]
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "collapsed_sections": [],
      "name": "Interpret your PyTorch model with Captum.ipynb",
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.7.4"
    },
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "13ed2c1fd01b4ff88f580127eb748fca": {
          "model_module": "jupyter-captum-insights",
          "model_module_version": "^0.1.0",
          "model_name": "CaptumInsightsModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "jupyter-captum-insights",
            "_model_module_version": "^0.1.0",
            "_model_name": "CaptumInsightsModel",
            "_view_count": null,
            "_view_module": "jupyter-captum-insights",
            "_view_module_version": "^0.1.0",
            "_view_name": "CaptumInsightsView",
            "attribution": {},
            "config": {},
            "insights_config": {
              "classes": [
                "plane",
                "car",
                "bird",
                "cat",
                "deer",
                "dog",
                "frog",
                "horse",
                "ship",
                "truck"
              ],
              "method_arguments": {
                "Feature Ablation": {
                  "perturbations_per_eval": {
                    "limit": [
                      1,
                      100
                    ],
                    "type": "number",
                    "value": 1
                  }
                },
                "Integrated Gradients": {
                  "method": {
                    "limit": [
                      "riemann_left",
                      "riemann_right",
                      "riemann_middle",
                      "riemann_trapezoid",
                      "gausslegendre"
                    ],
                    "type": "enum",
                    "value": "gausslegendre"
                  },
                  "n_steps": {
                    "limit": [
                      2,
                      null
                    ],
                    "type": "number",
                    "value": 25
                  }
                },
                "Occlusion": {
                  "perturbations_per_eval": {
                    "limit": [
                      1,
                      100
                    ],
                    "type": "number",
                    "value": 1
                  },
                  "sliding_window_shapes": {
                    "type": "string",
                    "value": ""
                  },
                  "strides": {
                    "type": "string",
                    "value": ""
                  }
                }
              },
              "methods": [
                "Deconvolution",
                "Deep Lift",
                "Guided Backprop",
                "Input X Gradient",
                "Integrated Gradients",
                "Saliency",
                "Feature Ablation",
                "Occlusion"
              ],
              "selected_method": "Integrated Gradients"
            },
            "label_details": {},
            "layout": "IPY_MODEL_dc617123d2e845439bf2100779e6a9c9",
            "output": []
          }
        },
        "1adcbb4b34f54525b4cea6e11fef5502": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "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": ""
          }
        },
        "2346d728f302413f9bfd5a89d17aba6a": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "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
          }
        },
        "78c4c247c9734fa7b6d4eb7928fd384e": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "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
          }
        },
        "793d6f5a901140579982778020495f8c": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "ProgressStyleModel",
          "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": ""
          }
        },
        "84baf7dd22f4400481d0387b3c8ebc86": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "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_9659711f0c8a43a1b4214778bce676a0",
            "placeholder": "​",
            "style": "IPY_MODEL_b11fd8dd3eec4778bbfa7c8360098fdb",
            "value": " 170499072/? [00:13&lt;00:00, 14927072.35it/s]"
          }
        },
        "87f16cad17374bd7ad6d41777b83d751": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HBoxModel",
          "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_facbadef58b84e798dfdd965abac0acc",
              "IPY_MODEL_f5ed5675b79041d18d90215eaefd9490",
              "IPY_MODEL_84baf7dd22f4400481d0387b3c8ebc86"
            ],
            "layout": "IPY_MODEL_aa41c5e6ec354c259f78a4f69527a625"
          }
        },
        "9659711f0c8a43a1b4214778bce676a0": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "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
          }
        },
        "aa41c5e6ec354c259f78a4f69527a625": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "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
          }
        },
        "b11fd8dd3eec4778bbfa7c8360098fdb": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "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": ""
          }
        },
        "f5ed5675b79041d18d90215eaefd9490": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "FloatProgressModel",
          "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_2346d728f302413f9bfd5a89d17aba6a",
            "max": 170498071,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_793d6f5a901140579982778020495f8c",
            "value": 170498071
          }
        },
        "facbadef58b84e798dfdd965abac0acc": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "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_78c4c247c9734fa7b6d4eb7928fd384e",
            "placeholder": "​",
            "style": "IPY_MODEL_1adcbb4b34f54525b4cea6e11fef5502",
            "value": ""
          }
        }
      }
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}