ResNetに層を追加するときの書き方.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Pytorch_転移学習_test.ipynb",
      "provenance": [],
      "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/ykitaguchi77/Colab_Scripts/blob/master/ResNet%E3%81%AB%E5%B1%A4%E3%82%92%E8%BF%BD%E5%8A%A0%E3%81%99%E3%82%8B%E3%81%A8%E3%81%8D%E3%81%AE%E6%9B%B8%E3%81%8D%E6%96%B9.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "U4HYzU9ckXbS"
      },
      "source": [
        "#PyTorch : Tutorial 初級 : 転移学習チュートリアル\n",
        "\n",
        "http://torch.classcat.com/2018/04/29/pytorch-tutorial-transfer-learning/\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "oyzcteflxK14"
      },
      "source": [
        "#自作データ、Early stopping"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "F4oE6PwpkNUb"
      },
      "source": [
        "# License: BSD\n",
        "# Author: Sasank Chilamkurthy\n",
        "\n",
        "from __future__ import print_function, division\n",
        "\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "import torch.optim as optim\n",
        "from torch.optim import lr_scheduler\n",
        "import numpy as np\n",
        "import torchvision\n",
        "from torchvision import datasets, models, transforms\n",
        "import matplotlib.pyplot as plt\n",
        "import time\n",
        "import os\n",
        "import copy\n",
        "\n",
        "#Advanced Pytorchから\n",
        "import glob\n",
        "import os.path as osp\n",
        "import random\n",
        "import numpy as np\n",
        "import json\n",
        "from PIL import Image\n",
        "from tqdm import tqdm\n",
        "import matplotlib.pyplot as plt\n",
        "%matplotlib inline\n",
        "\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "import torch.optim as optim\n",
        "import torch.utils.data as data\n",
        "import torchvision\n",
        "from torchvision import models, transforms\n",
        "\n",
        "#ukiyoeから\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "plt.ion()   # interactive mode"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "5sZw9awMlMpC"
      },
      "source": [
        "#Google driveのデータをマウント"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fbsPFIplk0Zb",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "d226f1ec-b9f3-4244-e21d-e1a16625fed8"
      },
      "source": [
        "\n",
        "'''\n",
        "grav: 甲状腺眼症\n",
        "cont: コントロール\n",
        "黒の空白を挿入することにより225px*225pxの画像を生成、RESNETを用いて転移学習\n",
        "－－－－－－－－－－－－－－\n",
        "アップデータするデータの構造\n",
        "gravcont.zip ----data---- hymenoptera_data----train\n",
        "                         |            |    |--val\n",
        "                         |            |-- imagenet_class_index.json\n",
        "                         |            |-- __init__.py\n",
        "                         |            |-- bennetts-wallabu-01.jpg\n",
        "                         |            |-- goldenretriever-3724972_640.jpg\n",
        "                         |\n",
        "                         |-- utils---- __pycache__---- __init__.cpython-36.pyc\n",
        "                                       |                        |-- dataloader_image_classification.cpython-36.pyc\n",
        "                                       |-- __init__.py\n",
        "                                       |-- dataloader_image_classification.py\n",
        "－－－－－－－－－－－－－－\n",
        "'''                                     \n",
        "\n",
        "#google driveをcolabolatoryにマウント\n",
        "from google.colab import drive\n",
        "drive.mount('/content/drive')\n",
        "\n",
        "#Advanced.zipを解凍\n",
        "\n",
        "!date -R\n",
        "!unzip -qq drive/My\\ Drive/gravcont.zip\n",
        "!date -R\n",
        "!ls"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount(\"/content/drive\", force_remount=True).\n",
            "Fri, 08 Jan 2021 04:37:08 +0000\n",
            "unzip:  cannot find or open drive/My Drive/gravcont.zip, drive/My Drive/gravcont.zip.zip or drive/My Drive/gravcont.zip.ZIP.\n",
            "Fri, 08 Jan 2021 04:37:08 +0000\n",
            "checkpoint.pt  drive  sample_data\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XGMZmsSGkjOC"
      },
      "source": [
        "#データをロードする\n",
        "データをロードするために torchvision と torch.utils.data パッケージを使用します。今日解いていく問題はアリ (= ants) とハチ (= bees) を分類するモデルを訓練することです。アリとハチのためにおよそ 120 訓練画像を持ちます。各クラスのために 75 検証画像があります。通常は、スクラッチから訓練される場合、これは一般化するには非常に小さいデータセットです。私達は転移学習を使用していますので、合理的に上手く一般化することができるはずです。\n",
        "\n",
        "このデータセットは imagenet の非常に小さなサブセットです。"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "2Xcl55uJknDK",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "ef08f922-cba1-43e2-d233-c4a522f44a00"
      },
      "source": [
        "# 入力画像の前処理をするクラス\n",
        "# 訓練時と推論時で処理が異なる\n",
        "\n",
        "\"\"\"\n",
        "    画像の前処理クラス。訓練時、検証時で異なる動作をする。\n",
        "    画像のサイズをリサイズし、色を標準化する。\n",
        "    訓練時はRandomResizedCropとRandomHorizontalFlipでデータオーギュメンテーションする。\n",
        "\n",
        "\n",
        "    Attributes\n",
        "    ----------\n",
        "    resize : int\n",
        "        リサイズ先の画像の大きさ。\n",
        "    mean : (R, G, B)\n",
        "        各色チャネルの平均値。\n",
        "    std : (R, G, B)\n",
        "        各色チャネルの標準偏差。\n",
        "\"\"\"\n",
        "\n",
        "data_transforms = {\n",
        "    'train': transforms.Compose([\n",
        "        transforms.RandomResizedCrop(224, scale=(0.5,1.0)),\n",
        "        transforms.RandomHorizontalFlip(),\n",
        "        transforms.ToTensor(),\n",
        "        transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])\n",
        "    ]),\n",
        "    'val': transforms.Compose([\n",
        "        transforms.Resize(224),\n",
        "        transforms.ToTensor(),\n",
        "        transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])\n",
        "    ]),\n",
        "}\n",
        "\n",
        "data_dir = '/content/drive/My Drive/Deep_learning/gravcont_seed_20200815'\n",
        "image_datasets = {x: datasets.ImageFolder(os.path.join(data_dir, x),\n",
        "                                          data_transforms[x])\n",
        "                  for x in ['train', 'val']}\n",
        "dataloaders = {x: torch.utils.data.DataLoader(image_datasets[x], batch_size=20,\n",
        "                                             shuffle=True, num_workers=4)\n",
        "              for x in ['train', 'val']}\n",
        "dataset_sizes = {x: len(image_datasets[x]) for x in ['train', 'val']}\n",
        "class_names = image_datasets['train'].classes\n",
        "\n",
        "device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")\n",
        "\n",
        "\n",
        "\n",
        "print(class_names)\n",
        "k=0\n",
        "for i in class_names:\n",
        "    print(class_names[k]+\"_train:\"+str(len(os.listdir(path=data_dir + '/train/'+class_names[k]))))\n",
        "    k+=1\n",
        "k=0\n",
        "for i in class_names:\n",
        "    print(class_names[k]+\"_val:\"+str(len(os.listdir(path=data_dir + '/val/'+class_names[k]))))\n",
        "    k+=1\n",
        "\n",
        "print(\"training data set_total：\"+ str(len(image_datasets['train'])))\n",
        "print(\"validating data set_total：\"+str(len(image_datasets['val'])))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "['cont', 'grav']\n",
            "cont_train:266\n",
            "grav_train:266\n",
            "cont_val:67\n",
            "grav_val:67\n",
            "training data set_total：532\n",
            "validating data set_total：134\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cQowdDzHmTdT"
      },
      "source": [
        "#少数の画像を可視化する"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "E1YMGAaimaLK",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 190
        },
        "outputId": "b653e1e9-3c28-47b2-a964-4ff435fe2179"
      },
      "source": [
        "def imshow(inp, title=None):\n",
        "    \"\"\"Imshow for Tensor.\"\"\"\n",
        "    inp = inp.numpy().transpose((1, 2, 0))\n",
        "    mean = np.array([0.485, 0.456, 0.406])\n",
        "    std = np.array([0.229, 0.224, 0.225])\n",
        "    inp = std * inp + mean\n",
        "    inp = np.clip(inp, 0, 1)\n",
        "    plt.imshow(inp)\n",
        "    if title is not None:\n",
        "        plt.title(title)\n",
        "    plt.pause(0.001)  # pause a bit so that plots are updated\n",
        "\n",
        "\n",
        "# Get a batch of training data\n",
        "inputs, classes = next(iter(dataloaders['train']))\n",
        "\n",
        "# Make a grid from batch\n",
        "out = torchvision.utils.make_grid(inputs)\n",
        "\n",
        "imshow(out, title=[class_names[x] for x in classes])"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "o9q7wSBWaQ_t"
      },
      "source": [
        "#Define the Early Stopping Class"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "U78WhRfxaRH7"
      },
      "source": [
        "import numpy as np\n",
        "import torch\n",
        "\n",
        "class EarlyStopping:\n",
        "    \"\"\"Early stops the training if validation loss doesn't improve after a given patience.\"\"\"\n",
        "    def __init__(self, patience=7, verbose=False, delta=0):\n",
        "        \"\"\"\n",
        "        Args:\n",
        "            patience (int): How long to wait after last time validation loss improved.\n",
        "                            Default: 7\n",
        "            verbose (bool): If True, prints a message for each validation loss improvement. \n",
        "                            Default: False\n",
        "            delta (float): Minimum change in the monitored quantity to qualify as an improvement.\n",
        "                            Default: 0\n",
        "        \"\"\"\n",
        "        self.patience = patience\n",
        "        self.verbose = verbose\n",
        "        self.counter = 0\n",
        "        self.best_score = None\n",
        "        self.early_stop = False\n",
        "        self.val_loss_min = np.Inf\n",
        "        self.delta = delta\n",
        "\n",
        "    def __call__(self, val_loss, model):\n",
        "\n",
        "        score = -val_loss\n",
        "\n",
        "        if self.best_score is None:\n",
        "            self.best_score = score\n",
        "            self.save_checkpoint(val_loss, model)\n",
        "        elif score < self.best_score + self.delta:\n",
        "            self.counter += 1\n",
        "            print(f'EarlyStopping counter: {self.counter} out of {self.patience}')\n",
        "            if self.counter >= self.patience:\n",
        "                self.early_stop = True\n",
        "        else:\n",
        "            self.best_score = score\n",
        "            self.save_checkpoint(val_loss, model)\n",
        "            self.counter = 0\n",
        "\n",
        "    def save_checkpoint(self, val_loss, model):\n",
        "        '''Saves model when validation loss decrease.'''\n",
        "        if self.verbose:\n",
        "            print(f'Validation loss decreased ({self.val_loss_min:.6f} --> {val_loss:.6f}).  Saving model ...')\n",
        "        torch.save(model.state_dict(), 'checkpoint.pt')\n",
        "        self.val_loss_min = val_loss"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cSq-oHh4mobp"
      },
      "source": [
        "#モデルを訓練する\n",
        "さて、モデルを訓練するための一般的な関数を書きましょう。ここで、次を示します :\n",
        "\n",
        "学習率をスケジューリングする\n",
        "ベスト・モデルをセーブする\n",
        "以下で、パラメータ scheduler は torch.optim.lr_scheduler からの LR scheduler オブジェクトです。"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VO7mosiCmrq6"
      },
      "source": [
        "def train_model(model, criterion, optimizer, scheduler, patience, num_epochs):\n",
        "    since = time.time()\n",
        "\n",
        "    best_model_wts = copy.deepcopy(model.state_dict())\n",
        "    best_acc = 0.0\n",
        "\n",
        "    # to track the training loss as the model trains\n",
        "    train_loss = []\n",
        "    # to track the validation loss as the model trains\n",
        "    valid_loss = []\n",
        "\n",
        "\n",
        "    # initialize the early_stopping object\n",
        "    early_stopping = EarlyStopping(patience=patience, verbose=True)\n",
        "\n",
        "    for epoch in range(num_epochs):\n",
        "        print('Epoch {}/{}'.format(epoch, num_epochs - 1))\n",
        "        print('-' * 10)\n",
        "\n",
        "        # Each epoch has a training and validation phase\n",
        "        for phase in ['train', 'val']:\n",
        "            if phase == 'train':\n",
        "                scheduler.step()\n",
        "                model.train()  # Set model to training mode\n",
        "            else:\n",
        "                model.eval()   # Set model to evaluate mode\n",
        "\n",
        "            running_loss = 0.0\n",
        "            running_corrects = 0\n",
        "\n",
        "            # Iterate over data.\n",
        "            for inputs, labels in dataloaders[phase]:\n",
        "                inputs = inputs.to(device)\n",
        "                labels = labels.to(device)\n",
        "\n",
        "                # zero the parameter gradients\n",
        "                optimizer.zero_grad()\n",
        "\n",
        "                # forward\n",
        "                # track history if only in train\n",
        "                with torch.set_grad_enabled(phase == 'train'):\n",
        "                    outputs = model(inputs)\n",
        "                    _, preds = torch.max(outputs, 1)\n",
        "                    loss = criterion(outputs, labels)\n",
        "\n",
        "                    # backward + optimize only if in training phase\n",
        "                    if phase == 'train':\n",
        "                        loss.backward()\n",
        "                        optimizer.step()\n",
        "\n",
        "                # statistics\n",
        "                running_loss += loss.item() * inputs.size(0)\n",
        "                running_corrects += torch.sum(preds == labels.data)\n",
        "\n",
        "            epoch_loss = running_loss / dataset_sizes[phase]\n",
        "            epoch_acc = running_corrects.double() / dataset_sizes[phase] \n",
        "            \n",
        "            # record train_loss and valid_loss\n",
        "            if phase == 'train':\n",
        "                train_loss.append(epoch_loss)\n",
        "            if phase == 'val':\n",
        "                valid_loss.append(epoch_loss)\n",
        "            #print(train_loss)\n",
        "            #print(valid_loss)\n",
        "\n",
        "            print('{} Loss: {:.4f} Acc: {:.4f}'.format(\n",
        "                phase, epoch_loss, epoch_acc))\n",
        "\n",
        "\n",
        "            # deep copy the model\n",
        "            if phase == 'val' and epoch_acc > best_acc:\n",
        "                best_acc = epoch_acc\n",
        "                best_model_wts = copy.deepcopy(model.state_dict())\n",
        "      \n",
        "      # early_stopping needs the validation loss to check if it has decresed, \n",
        "      # and if it has, it will make a checkpoint of the current model\n",
        "        if phase == 'val':    \n",
        "            early_stopping(epoch_loss, model)\n",
        "                \n",
        "            if early_stopping.early_stop:\n",
        "                print(\"Early stopping\")\n",
        "                break\n",
        "        print()\n",
        "\n",
        "\n",
        "    time_elapsed = time.time() - since\n",
        "    print('Training complete in {:.0f}m {:.0f}s'.format(\n",
        "        time_elapsed // 60, time_elapsed % 60))\n",
        "    print('Best val Acc: {:4f}'.format(best_acc))\n",
        "\n",
        "    # load best model weights\n",
        "    model.load_state_dict(best_model_wts)\n",
        "    return model, train_loss, valid_loss"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "URMZLxsvmyZS"
      },
      "source": [
        "#モデル予測を可視化する\n",
        "少数の画像のための予測を表示するための一般的な関数。"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ga3kWm5Nm1Mj"
      },
      "source": [
        "def visualize_model(model, num_images=6):\n",
        "    was_training = model.training\n",
        "    model.eval()\n",
        "    images_so_far = 0\n",
        "    fig = plt.figure()\n",
        "\n",
        "    with torch.no_grad():\n",
        "        for i, (inputs, labels) in enumerate(dataloaders['val']):\n",
        "            inputs = inputs.to(device)\n",
        "            labels = labels.to(device)\n",
        "\n",
        "            outputs = model(inputs)\n",
        "            _, preds = torch.max(outputs, 1)\n",
        "\n",
        "            for j in range(inputs.size()[0]):\n",
        "                images_so_far += 1\n",
        "                ax = plt.subplot(num_images//2, 2, images_so_far)\n",
        "                ax.axis('off')\n",
        "                ax.set_title('predicted: {}'.format(class_names[preds[j]]))\n",
        "                imshow(inputs.cpu().data[j])\n",
        "\n",
        "                if images_so_far == num_images:\n",
        "                    model.train(mode=was_training)\n",
        "                    return\n",
        "        model.train(mode=was_training)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "BhvTS0Y8m3HD"
      },
      "source": [
        "#convnet を再調整する\n",
        "事前訓練されたモデルをロードして最後の完全結合層をリセットします。"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6s7odcUUm-hT"
      },
      "source": [
        "model_ft = models.resnet50(pretrained=False)\n",
        "num_ftrs = model_ft.fc.in_features\n",
        "model_ft.fc = nn.Linear(num_ftrs, 2)\n",
        "\n",
        "#GPU使用\n",
        "model_ft = model_ft.to(device)\n",
        "\n",
        "#損失関数を定義\n",
        "criterion = nn.CrossEntropyLoss()\n",
        "\n",
        "# Observe that all parameters are being optimized\n",
        "optimizer_ft = optim.SGD(model_ft.parameters(), lr=0.001, momentum=0.9)\n",
        "\n",
        "# Decay LR by a factor of 0.1 every 7 epochs\n",
        "exp_lr_scheduler = lr_scheduler.StepLR(optimizer_ft, step_size=7, gamma=0.1)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4BYMjdsES3Sw"
      },
      "source": [
        "#ResNetに層を追加するときの書き方"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OzgxHd9SMicI"
      },
      "source": [
        "class Resnet(nn.Module):\r\n",
        "  def __init__(self):\r\n",
        "    super(Resnet,self).__init__()\r\n",
        "    resnet = models.resnet50(pretrained=True)\r\n",
        "    self.resnet = nn.Sequential(*list(resnet.children())[:-2])\r\n",
        "    self.maxpool = nn.MaxPool2d(kernel_size=7)\r\n",
        "    self.fc = nn.Linear(2048, 2)\r\n",
        "\r\n",
        "  def forward(self,x):\r\n",
        "    x = self.resnet(x)\r\n",
        "    x = self.maxpool(x)\r\n",
        "    x = torch.flatten(x,1)\r\n",
        "    x = self.fc(x)\r\n",
        "    return x\r\n",
        "\r\n",
        "model_ft = Resnet()\r\n",
        "\r\n",
        "#GPU使用\r\n",
        "model_ft = model_ft.to(device)\r\n",
        "\r\n",
        "#損失関数を定義\r\n",
        "criterion = nn.CrossEntropyLoss()\r\n",
        "\r\n",
        "# Observe that all parameters are being optimized\r\n",
        "optimizer_ft = optim.SGD(model_ft.parameters(), lr=0.001, momentum=0.9)\r\n",
        "\r\n",
        "# Decay LR by a factor of 0.1 every 7 epochs\r\n",
        "exp_lr_scheduler = lr_scheduler.StepLR(optimizer_ft, step_size=7, gamma=0.1)\r\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fT333lNjRBUp",
        "outputId": "a41f9d7b-76ae-4c92-8b77-b894da02571d"
      },
      "source": [
        "print(model_ft)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Resnet(\n",
            "  (resnet): Sequential(\n",
            "    (0): Conv2d(3, 64, kernel_size=(7, 7), stride=(2, 2), padding=(3, 3), bias=False)\n",
            "    (1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "    (2): ReLU(inplace=True)\n",
            "    (3): MaxPool2d(kernel_size=3, stride=2, padding=1, dilation=1, ceil_mode=False)\n",
            "    (4): Sequential(\n",
            "      (0): Bottleneck(\n",
            "        (conv1): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "        (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (relu): ReLU(inplace=True)\n",
            "        (downsample): Sequential(\n",
            "          (0): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "          (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        )\n",
            "      )\n",
            "      (1): Bottleneck(\n",
            "        (conv1): Conv2d(256, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "        (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (relu): ReLU(inplace=True)\n",
            "      )\n",
            "      (2): Bottleneck(\n",
            "        (conv1): Conv2d(256, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "        (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (relu): ReLU(inplace=True)\n",
            "      )\n",
            "    )\n",
            "    (5): Sequential(\n",
            "      (0): Bottleneck(\n",
            "        (conv1): Conv2d(256, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
            "        (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (relu): ReLU(inplace=True)\n",
            "        (downsample): Sequential(\n",
            "          (0): Conv2d(256, 512, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
            "          (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        )\n",
            "      )\n",
            "      (1): Bottleneck(\n",
            "        (conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "        (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (relu): ReLU(inplace=True)\n",
            "      )\n",
            "      (2): Bottleneck(\n",
            "        (conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "        (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (relu): ReLU(inplace=True)\n",
            "      )\n",
            "      (3): Bottleneck(\n",
            "        (conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "        (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (relu): ReLU(inplace=True)\n",
            "      )\n",
            "    )\n",
            "    (6): Sequential(\n",
            "      (0): Bottleneck(\n",
            "        (conv1): Conv2d(512, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
            "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (relu): ReLU(inplace=True)\n",
            "        (downsample): Sequential(\n",
            "          (0): Conv2d(512, 1024, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
            "          (1): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        )\n",
            "      )\n",
            "      (1): Bottleneck(\n",
            "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (relu): ReLU(inplace=True)\n",
            "      )\n",
            "      (2): Bottleneck(\n",
            "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (relu): ReLU(inplace=True)\n",
            "      )\n",
            "      (3): Bottleneck(\n",
            "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (relu): ReLU(inplace=True)\n",
            "      )\n",
            "      (4): Bottleneck(\n",
            "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (relu): ReLU(inplace=True)\n",
            "      )\n",
            "      (5): Bottleneck(\n",
            "        (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (relu): ReLU(inplace=True)\n",
            "      )\n",
            "    )\n",
            "    (7): Sequential(\n",
            "      (0): Bottleneck(\n",
            "        (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
            "        (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn3): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (relu): ReLU(inplace=True)\n",
            "        (downsample): Sequential(\n",
            "          (0): Conv2d(1024, 2048, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
            "          (1): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        )\n",
            "      )\n",
            "      (1): Bottleneck(\n",
            "        (conv1): Conv2d(2048, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "        (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn3): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (relu): ReLU(inplace=True)\n",
            "      )\n",
            "      (2): Bottleneck(\n",
            "        (conv1): Conv2d(2048, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
            "        (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
            "        (bn3): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
            "        (relu): ReLU(inplace=True)\n",
            "      )\n",
            "    )\n",
            "  )\n",
            "  (maxpool): MaxPool2d(kernel_size=7, stride=7, padding=0, dilation=1, ceil_mode=False)\n",
            "  (fc): Linear(in_features=2048, out_features=2, bias=True)\n",
            ")\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "XaLCTzTLnBSL"
      },
      "source": [
        "#訓練と評価\n",
        "CPU 上でおよそ 15-25 分かかるはずです。けれども GPU 上なら、1 分もかかりません。"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8ijHctNfnD_i",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "2439deb8-5dbd-400b-b344-9d357d539e81"
      },
      "source": [
        "model_ft, train_loss, valid_loss = train_model(model_ft, criterion, optimizer_ft, exp_lr_scheduler, patience=10, num_epochs=10)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 0/99\n",
            "----------\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/torch/optim/lr_scheduler.py:136: UserWarning: Detected call of `lr_scheduler.step()` before `optimizer.step()`. In PyTorch 1.1.0 and later, you should call them in the opposite order: `optimizer.step()` before `lr_scheduler.step()`.  Failure to do this will result in PyTorch skipping the first value of the learning rate schedule. See more details at https://pytorch.org/docs/stable/optim.html#how-to-adjust-learning-rate\n",
            "  \"https://pytorch.org/docs/stable/optim.html#how-to-adjust-learning-rate\", UserWarning)\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "train Loss: 2.0077 Acc: 0.5620\n",
            "val Loss: 0.7933 Acc: 0.7687\n",
            "Validation loss decreased (inf --> 0.793254).  Saving model ...\n",
            "\n",
            "Epoch 1/99\n",
            "----------\n",
            "train Loss: 1.5250 Acc: 0.7256\n",
            "val Loss: 6.1747 Acc: 0.5299\n",
            "EarlyStopping counter: 1 out of 10\n",
            "\n",
            "Epoch 2/99\n",
            "----------\n",
            "train Loss: 1.8804 Acc: 0.7312\n",
            "val Loss: 4.4128 Acc: 0.6269\n",
            "EarlyStopping counter: 2 out of 10\n",
            "\n",
            "Epoch 3/99\n",
            "----------\n",
            "train Loss: 1.1517 Acc: 0.7857\n",
            "val Loss: 1.1257 Acc: 0.7910\n",
            "EarlyStopping counter: 3 out of 10\n",
            "\n",
            "Epoch 4/99\n",
            "----------\n",
            "train Loss: 0.9165 Acc: 0.7763\n",
            "val Loss: 1.7797 Acc: 0.6119\n",
            "EarlyStopping counter: 4 out of 10\n",
            "\n",
            "Epoch 5/99\n",
            "----------\n",
            "train Loss: 0.4852 Acc: 0.8327\n",
            "val Loss: 0.5085 Acc: 0.8284\n",
            "Validation loss decreased (0.793254 --> 0.508504).  Saving model ...\n",
            "\n",
            "Epoch 6/99\n",
            "----------\n",
            "train Loss: 0.2988 Acc: 0.8835\n",
            "val Loss: 0.4124 Acc: 0.8284\n",
            "Validation loss decreased (0.508504 --> 0.412401).  Saving model ...\n",
            "\n",
            "Epoch 7/99\n",
            "----------\n",
            "train Loss: 0.2450 Acc: 0.8891\n",
            "val Loss: 0.3779 Acc: 0.8209\n",
            "Validation loss decreased (0.412401 --> 0.377914).  Saving model ...\n",
            "\n",
            "Epoch 8/99\n",
            "----------\n",
            "train Loss: 0.1904 Acc: 0.9248\n",
            "val Loss: 0.4111 Acc: 0.8358\n",
            "EarlyStopping counter: 1 out of 10\n",
            "\n",
            "Epoch 9/99\n",
            "----------\n",
            "train Loss: 0.2144 Acc: 0.9267\n",
            "val Loss: 0.3790 Acc: 0.8358\n",
            "EarlyStopping counter: 2 out of 10\n",
            "\n",
            "Epoch 10/99\n",
            "----------\n",
            "train Loss: 0.2149 Acc: 0.9173\n",
            "val Loss: 0.3572 Acc: 0.8284\n",
            "Validation loss decreased (0.377914 --> 0.357195).  Saving model ...\n",
            "\n",
            "Epoch 11/99\n",
            "----------\n",
            "train Loss: 0.1546 Acc: 0.9436\n",
            "val Loss: 0.3695 Acc: 0.8358\n",
            "EarlyStopping counter: 1 out of 10\n",
            "\n",
            "Epoch 12/99\n",
            "----------\n",
            "train Loss: 0.1957 Acc: 0.9060\n",
            "val Loss: 0.3691 Acc: 0.8284\n",
            "EarlyStopping counter: 2 out of 10\n",
            "\n",
            "Epoch 13/99\n",
            "----------\n",
            "train Loss: 0.1587 Acc: 0.9286\n",
            "val Loss: 0.3536 Acc: 0.8582\n",
            "Validation loss decreased (0.357195 --> 0.353645).  Saving model ...\n",
            "\n",
            "Epoch 14/99\n",
            "----------\n",
            "train Loss: 0.1669 Acc: 0.9436\n",
            "val Loss: 0.3694 Acc: 0.8284\n",
            "EarlyStopping counter: 1 out of 10\n",
            "\n",
            "Epoch 15/99\n",
            "----------\n",
            "train Loss: 0.1829 Acc: 0.9211\n",
            "val Loss: 0.3531 Acc: 0.8507\n",
            "Validation loss decreased (0.353645 --> 0.353123).  Saving model ...\n",
            "\n",
            "Epoch 16/99\n",
            "----------\n",
            "train Loss: 0.1620 Acc: 0.9323\n",
            "val Loss: 0.3616 Acc: 0.8433\n",
            "EarlyStopping counter: 1 out of 10\n",
            "\n",
            "Epoch 17/99\n",
            "----------\n",
            "train Loss: 0.1732 Acc: 0.9229\n",
            "val Loss: 0.3568 Acc: 0.8358\n",
            "EarlyStopping counter: 2 out of 10\n",
            "\n",
            "Epoch 18/99\n",
            "----------\n",
            "train Loss: 0.1455 Acc: 0.9361\n",
            "val Loss: 0.3622 Acc: 0.8358\n",
            "EarlyStopping counter: 3 out of 10\n",
            "\n",
            "Epoch 19/99\n",
            "----------\n",
            "train Loss: 0.1716 Acc: 0.9417\n",
            "val Loss: 0.3463 Acc: 0.8582\n",
            "Validation loss decreased (0.353123 --> 0.346308).  Saving model ...\n",
            "\n",
            "Epoch 20/99\n",
            "----------\n",
            "train Loss: 0.1460 Acc: 0.9417\n",
            "val Loss: 0.3504 Acc: 0.8582\n",
            "EarlyStopping counter: 1 out of 10\n",
            "\n",
            "Epoch 21/99\n",
            "----------\n",
            "train Loss: 0.1644 Acc: 0.9229\n",
            "val Loss: 0.3505 Acc: 0.8507\n",
            "EarlyStopping counter: 2 out of 10\n",
            "\n",
            "Epoch 22/99\n",
            "----------\n",
            "train Loss: 0.1508 Acc: 0.9455\n",
            "val Loss: 0.3529 Acc: 0.8358\n",
            "EarlyStopping counter: 3 out of 10\n",
            "\n",
            "Epoch 23/99\n",
            "----------\n",
            "train Loss: 0.1483 Acc: 0.9511\n",
            "val Loss: 0.3597 Acc: 0.8433\n",
            "EarlyStopping counter: 4 out of 10\n",
            "\n",
            "Epoch 24/99\n",
            "----------\n",
            "train Loss: 0.1713 Acc: 0.9305\n",
            "val Loss: 0.3582 Acc: 0.8358\n",
            "EarlyStopping counter: 5 out of 10\n",
            "\n",
            "Epoch 25/99\n",
            "----------\n",
            "train Loss: 0.1457 Acc: 0.9474\n",
            "val Loss: 0.3591 Acc: 0.8507\n",
            "EarlyStopping counter: 6 out of 10\n",
            "\n",
            "Epoch 26/99\n",
            "----------\n",
            "train Loss: 0.1508 Acc: 0.9474\n",
            "val Loss: 0.3455 Acc: 0.8582\n",
            "Validation loss decreased (0.346308 --> 0.345521).  Saving model ...\n",
            "\n",
            "Epoch 27/99\n",
            "----------\n",
            "train Loss: 0.1878 Acc: 0.9342\n",
            "val Loss: 0.3543 Acc: 0.8358\n",
            "EarlyStopping counter: 1 out of 10\n",
            "\n",
            "Epoch 28/99\n",
            "----------\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "error",
          "ename": "KeyboardInterrupt",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-32-ca7cf89a33b5>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mmodel_ft\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mtrain_loss\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mvalid_loss\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtrain_model\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmodel_ft\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcriterion\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moptimizer_ft\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mexp_lr_scheduler\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpatience\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m10\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnum_epochs\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m100\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m",
            "\u001b[0;32m<ipython-input-10-02352cf7c85e>\u001b[0m in \u001b[0;36mtrain_model\u001b[0;34m(model, criterion, optimizer, scheduler, patience, num_epochs)\u001b[0m\n\u001b[1;32m     50\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     51\u001b[0m                 \u001b[0;31m# statistics\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 52\u001b[0;31m                 \u001b[0mrunning_loss\u001b[0m \u001b[0;34m+=\u001b[0m \u001b[0mloss\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mitem\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;34m*\u001b[0m \u001b[0minputs\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msize\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     53\u001b[0m                 \u001b[0mrunning_corrects\u001b[0m \u001b[0;34m+=\u001b[0m \u001b[0mtorch\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msum\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpreds\u001b[0m \u001b[0;34m==\u001b[0m \u001b[0mlabels\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mdata\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     54\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WrWRhUFoGZYS",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 585
        },
        "outputId": "93f77eb4-78bc-4aa9-edaf-a5bd052556aa"
      },
      "source": [
        "# visualize the loss as the network trained\n",
        "fig = plt.figure(figsize=(10,8))\n",
        "plt.plot(range(1,len(train_loss)+1),train_loss, label='Training Loss')\n",
        "plt.plot(range(1,len(valid_loss)+1),valid_loss,label='Validation Loss')\n",
        "\n",
        "# find position of lowest validation loss\n",
        "minposs = valid_loss.index(min(valid_loss))+1 \n",
        "plt.axvline(minposs, linestyle='--', color='r',label='Early Stopping Checkpoint')\n",
        "\n",
        "plt.xlabel('epochs')\n",
        "plt.ylabel('loss')\n",
        "plt.ylim(0, 1.0) # consistent scale\n",
        "plt.xlim(0, len(train_loss)+1) # consistent scale\n",
        "plt.grid(True)\n",
        "plt.legend()\n",
        "plt.tight_layout()\n",
        "plt.show()\n",
        "fig.savefig('loss_plot.png', bbox_inches='tight')"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsgAAAI4CAYAAAB3OR9vAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 720x576 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TzP0fvLCoWNS",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 581
        },
        "outputId": "fac574cb-f845-464c-a777-103e2b2316c0"
      },
      "source": [
        "visualize_model(model_ft)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAGkAAABeCAYAAAAg/TovAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nZwAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAGkAAABeCAYAAAAg/TovAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0\ndHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nO2deaxlyV3fP7WcOutd39Lv9fJ6nR57\nPJ4ZY2MM2CwWYglESggKUSKEE0ByFAUliiIShYgoMcGJEpEIkqBESZAgQBBBiCQyQhaE1TYGGxts\nvMzS03u/7e5nr6r8ce/09Exm7DH0wzzlfaWre8+pOr+q+n2rTlX96ld1hfeeE/zZhvxiZ+AEnx8n\nJB0DnJB0DHBC0jHACUnHACckHQN80UkSQvy4EOI9q9/vEEJ8+k8pXS+EuPKnkdafFF90kh6E9/43\nvPePfr54Qoh3CSF+808jT0eFL6QMD5UkIYR+mPJOsIL3/nN+gGvAPwQ+CYyA/wpEq7CvAW4C3wfc\nBX5idf9bgN8HxsBvA088IO9NwEeAGfDfgZ8B3vOgvAfingN+HtgDDoAfBV4PlIAF5sB4FTcE/hVw\nHbgH/BgQPyDr7wN3gNvA3wA8cOXzlX/17HBV7tsrHfzCA2HfAzwNHAK/CJx+IMwD7wY+u9LFvwPE\nq5XhVdN/jST94UphQ+C3XqbUFvgXKyXFKxJ2gS8DFPCdKxkhYIDngb8LBMC3Ac0rkbR69mPADwMp\nEAFvX4W9C/jNl+Xzh1dKGgId4H8CP7QK+8YVcY+vZP3UgyQBfxX4+OfQwf9eVajBKt9fvbr/TmAf\n+JJV+X4E+PWXkfS/gD6ws6ps3/hqZfiTkvTuB67/HPDMA0qtWbWs1b3/APyzl8n4NPDVwFetaqN4\nIOy3X4WkL18VSr9Cnl5SQJa1cwFcfuDelwPPrX7/F+C9D4Rd5TW2JGAbcMDgFcL+M/AvH7jOVpXu\nwgMkvf2B8J8F/sEXStJr7UNuPPD7eeD0A9d73vvygevzwHcKIf72A/fM6hkP3PKrXD4g75VwDnje\ne9++hvxtAAnwe0KIF+4Jlq2RVdq/9xrSfLV8HHrvR68QdprlqxsA7/1cCHEAnGFZuWHZDbyAnCWR\nXxBe68Dh3AO/d1i2hvt5e1ncG8APeu/7D3wS7/1Ps+wTzogHNLmS90q4Aey8ymDk5WnuAwXwhgfS\n7HnvX1DInVcow2vFDWAohOi/QthtlpUSACFECqwBt16D3Ne8/PBaSfpbQoizQogh8I9Yvp9fDf8J\neLcQ4svEEqkQ4puFEB3gAyz7sO8VQgRCiG8F3voqcn6HpXLfu5IRCSG+chV2DzgrhDAA3nu3SveH\nhRCbAEKIM0KIb1jF/1ngXUKIx4QQCfADr7HceO/vAO8D/r0QYrDK91etgn8a+OtCiKeEECHwz4EP\nee+vvQbRLynD58JrJemngF8GngWeAd7zahG997/LcsTzoyxHQk+zfP/iva+Bb11dHwLfznL09kpy\nLPDngSssR2w3V/EBfgX4BHBXCLG/uvd9q7Q+KISYAu8HHl3Jeh/wb1bPPb36vg8hxF8TQnzic5T/\nO1j2NZ9iOSj6Oyu57wf+MfA/WFaoy8Bf+RxyHsQrleEVIV7aPbxCBCGuAd+9ytAJvgj4M2VxOMEr\n44SkY4DP+7o7wRcfJy3pGOCEpGOAI7FaCyH+v36Heu/F54/12nHSko4BTkg6Bjgh6RjghKRjgBOS\njgFOSDoGOCHpGOCEpGOAoyFJCrTWSATyfiLic3+kZunL8WCulmFS6pfIfmmuV2HBC88tr6UUy7gI\nJOL+rF0iXkxDgpYgUcs0pAItHpR6/5kXZbx8nvpQ562viCMxsL5gcVgqxC+/5Uvrg3Puxd+rlWR9\n/1rgUGipEHgaV7OUB+5lRhKJwtEipQMkzmmQ1UviBEbRlPalmZQC6ZZOHsucGESgaLyDtkXikNKs\nSgBCSmzbgPQvyfvSjeIF2ctyPGyLw5GYhZI4IQw0nTRDC0ld11zYOU2v3+fevXuMR2O6a31s0yI9\nPPnkkzxy8QLeF9y4u8fBqCJNz/Ls859mtHeDp974epxUfPIzT3P5wlV63QGf/OSnePzxJzm9OWCR\nz7h8aYcgiHjmmeuc7oW86ckncE2FFB58y/7ePofTBR/66McYrm0isojFeJ8r585wZ3+fu7tj0CHj\nwqG0pppNGB8eUlQVOgwYT6cYYyirBTvnzvHsc88xnUzpDTYJAs3e7h38a3db+IJwJK87LQRaKpzz\ntK0lUAqNI/CWfHLI5rDPRq9LWxREAlIEzXTK6O5dbjzzLEZK7ty6gbWeSGtEVfHcZ59GSoPyLYvp\niDgwKA+ZgoiWkIatfsbozg2unDlH0LYY25Lh6StJKuDM2pDp7j75ZEo1njGMYrphwGaWsN4L2b11\nDVvWVPOcuinY3tzANTWTw0MQnqIqCHXI889dZzqaYrQhiTskcYZUhu3tsyj18Ov9kZAUmACtA8JA\nEyhFaAzKOzbX1uhlGZ04YjGdgm0IlORw9x51UXC4d8jW+gazw0Pu3r5BPp+xORhQlwWzyQQtJXWz\nQAjP+fNnuX3rGmkkGXYT+p2E2zeeBVfw2d3rHNQL6gBExzAXFTNKVDfijV/xZm5M9wkRXPvsM9y7\neZ3EaKQt2Rx2ONy9Sb4Ys39wl2J2wM7ZDeJUIWRL2xa41uJbh3eO2WzK7Vs3uHf3HloGjEeTI1Hp\nEQ0cFE1TMuh12Bh2wRasDQe0bU1ZzFlfH1IUFVoE9Lpdts5t00hLZWu01tR1RZoqIj8j66bMihol\nIKHGaENeNDRNSRpW9HsxW5tDjNY0ZcH6sEPsPZmKMDImn5VgHd0oQtcVPRNQ52Pm8ylOCMIsw7qK\numi4fPYcV3bWqeb7FHmBloI40tTFjPl4QpU7RuNd4iwg6UYoI3HNnLVBgqBhOhtTFMXDV+dDlwgE\nCoyS3Lx5g6rM0VKhtaKuSk5vb7OxPkR6Rxwb4igkiULasuDCuR2sUNy+d5fp6B5Xr1xCSyiKORLL\n5vqQYjZl2E+4d+saTzz2OvZ2d9lYX+fZp59mNpty/tw5ur0udV0jpaRpGkyY4KynWORsra9xatBn\nMh2htEBKSbc7IIoCpHAkoSZUnnYx5XA0xpgAoySBgMwI1gcDZpMRdVlglCBJQm7fvon1lm6ny5Ur\nlx+6Po+EJNvW1GXFqY11zm5vkaURO2dOc/rUBr5tGB8eILBc2DmHwDEdj+hkKcJLdvdHTKZz3vDY\nI7RtzfXr1yiKBVFoqOuStX6P3dvX6XVCTg2HrA+GNGVFN804s7VNbELyRc7W1hZlWRJGEW1rSdMO\ng14fgWNn+xRVtUBpyd7+AUJI+t0OSnmeevz1bA77nBr02ds/YG/3HhfPn2Nz0KGpZizmU4S3KAnO\ntYBHB4K6rqiaksVi/tD1eSQkhVqTpiGhUTz//LMMBwOUhPlsysUL59k5ewaBJ0tirly6xNbGGhfP\nnSVLUkbjBXHaQQrH/r07dNMYLeDU5jpxGBAqBbbgzU+9gbXhkFMbmzR1g5aK9bV1OmlGUzeAoN/v\nE0URnV6fzVOnMcawMRzQS2K2T2+RpjG3b99hPi2I45DIaFxd8eiFHTqRQRvD7du3CENDL0sYZgne\nWqIwpJdlJEmMCRXOWsDStjX3dm9/PvV8wTgSkjYHPeJAYZxjs9/j6oUd+onm3q3nSWOD95a3PfEl\npNqw0euD9Ywnc27u7dNMx/QDRT5ZkKQpxoS0bcuVC5eItSEAzgxPcXrtNFcuX8U5x/7uHleuPkpn\nsI5VEcp78qamt7GJ0obdu/sgA3Sa0en3efzxx3n03A7FeMrW5ilMHGJ0h1B3qCvL6e0zPPbY69je\n6KKF4vqzNzA65Kmn3oRSEEcxtrQE1pBPa5IgRTlNpEICGT90fR7JPMnaljNbWwjnePTyZQb9HlnW\n4a1f+lY8kryuWdvoI6WkbEpkoEEsLQRrGxv0ej3WNzrcu7fHfD7nne/8Okb7B1y59Ag3nr/JE48/\nyVd++TuQSrK/v8/2mTOowHD2zA4HByM2T21y6tQpwjBk4j3b29tIJTEmpG48UmrW1jf4+m/4Bn73\nox/hgx/+Xb7lm74ZnMNLwXyRc+bsDjf3xkipSbOM69evMxpNUAjy6Yw4inEWHJa2qQnDgI1Tp7h1\n67W4gX9hOJKWdOnCebqdDo8++gjbm+sEUjKfL+h0unjvWRuukXQSyrYiSlN0GHP34JCP/v7HOX/h\nIuPxmKJsuHz5ClcfuYrWAWdOn2aRF1y5fIVzO+dRShOGIdqE7Fy8yNmdi9RNw3BzkzhJUEohEKRp\nSts6BIokySjykjTpYKKIbq/P1tY2m1tbfOBDH2Z/NONwmrM/nnI4W7CxsYnzcPPmTS5evESn0+XS\nxSt46zizvUUcabIsJk4MwtbEYbC0RD1kHElL2lxfY//ePap8QXbuDFIInHUURU2vt0bdWmSYMprv\nEqaObrbG9T98hrXNLWRgOHvhIkU55c7tXbRSTMcL1gdD1odDvvadX4cQAiEVN27c5szOBaKkgwoC\n8tGY8xcuMb2rMIGhqCrKsiTQhqIoMTagk/XY3b1HnKR4KYmSFGYzdBzzkU/8EXGSIAON8xCamPW1\nTZTQzCZzzp49z/Xr1+h1My5dOMtwkDJZ5Eync2Z5ye6dG4iX7AJ6ODgSkrSSrK0NWOt1KYocnCdN\nMxbznDTtUpY5vVMbSGUQIqSsPNYq3v72d3B+5wof+OAHiFPDYjRldHjI1auP4pqWtbU12rYlyzLm\nszlKKWazOXXdcHg4otPtg1BIpfB4ZtMpYWDQQnOwf0inmzFfzKiqisPZgrIsGQzXyauaIEhJki7d\nQR9tDIeHh2wOBxRFQafT4blnn2U+n/OGxx/n4x/7KHmxQAjIkpgwDAkmc7TWNPfqh6/Phy4RkHi6\nWYb3gqJqQQhmozm9bpfxIkdqTS/KWO/0SALJYj7iK97yFGHcJZ9OeOTCOa7ffJ4rj11lOUob4Jwn\njhNmszFawcH+PhunTpGPRrSDAVGakAWCyWzM2sYmi/EeolmQ9TosyhJlHLfvXKNtLUmSIPM5vX6H\nssw5vbmGjmLKKiVOuswmM06vbdJUE15/6TxxErO91kMIgY40r3vdZUwUUVYl1559nu3tbeqqIksG\nDPrpQ9fnkZD0oJW4LEuEEHQ6AybjMUkcc+XKI/R7XYa9Dud2dhiPJnS6PaK0w927dxh0OmRXH2Vz\n0CUIFIs8B+9JswRvGz7z6U9y/sIFppNDxgd7uJ1zKOFZ5FO8szQNHBzuE4chB4cHKKXxzhKGBu9K\n6rokMpI0C0ljjVIKE8bU1rKYz+l3QoqipDvss1uVKAnbG+vs7u9B2zLodmmdI4563DEBZVnQ1hVZ\nt4t19nNo5o+Ho7E4BAEAeZ7jvWc4GFAt5lw+e5ZTwwH3bl2nKWdcOLdNFgWkoaabhKx3It7yxGNs\ndDMubm0T64DURJw9tUkcCOIAwkAw7GdMR3v4tiTLQpJEY5sc15SkiaauKrSWDAZd2qaiyGe4pkIJ\nCLWiWMzp9xICJUiTFNt4lAfjHDtbp0iU4/RaipGOYT8loKWbhZzd3iAAhllKGmgyE7C9dQqjFf1u\nF9tUhCZ46Po8mtedlFRVhVaKLE0JwxCZOiQeAZzZ2sLbBufg5t41Hn309ZRFQVPOEa7FSMnW2gZ7\n+yNiE2FtRb/bw3lLmIRkSYTzjvl8zs650zjXYEIDWOq6IAgDer0us/kUrSVN6whNgHUOgWd9bYj1\nDqkCjE4gNsQqIOslTKYjIh0wGY3xwORgnzhNqIuSQGsGvS57+3s4Z+l2M3pZStvWWN/ivELrh6/S\no3Ez9pY0jpAesjReXichxki01mgF1hmca9nc2GA2HuGFpW7A+YA46zJaTIj7hiTOsE2CbRaoQJDP\nJ4RhQGstw36GDCRSW4SWFFVNLDsoHdC2+XItq23ACZq2JQiWJhy8RJGQpUMcAZ2OYm9/j4Nb95Cy\nIQwDyqKhamocJfOiIDAxUZxhdMBkJImzmPlihtaObhZRHRRkaYptH/6a0tGM7rSG1fEDSimkVIRB\ngJSSxXzBcl9zS9taXKMITICSitZ61gYReV4hlEWohDjpEOmIopgwnR4SJ+nq+QprHU44vAfrS8Io\nRQcaFWjCKKJazFBKUdUV3rW0lmW+nCcIguUgIsvIiwVKWjpZyK1b90jSmLSbYUcLwqCLbVuUjZju\nFQhZkfW6IATee5q2Xm7jX7Wgl+7Zfjg4GgOrtQQrUqbTKYvFHOcc0+kUExoW8wXlYkyxGIPwTOdz\nFmWNCFNaFEEUUiwWlEXBZDRld++QcuFQMqVY1DSVAyGp2xIpBEKI5cRWa4QQWGtxzmJbi8fjXEtr\nG8qyQKsApSLadml5sG1LUcxB1pRNwaVHH8OqkFqCSgAtEIGkqucU+T4HkxHz+Zwiz3FNS9O2y1Hf\nKu1jQ5IxhrZtkVIuR07G3F86yPOcwATkdUPaHzKeV0SdIdHaBqrTQScJJkuJehmIlhu3nmM+n1A3\nFXXdIDD0XhiSRxHWOeTKf0KsajdA27QEgQYEQoCQnqX+BEmSIYRGKk3d1BijmSxmBJ0O0wpUOsAm\nGSMf0MRdbNIhlzWlKVEmIAgNgda4ZllG7z1Syvv5eNg4soFDaEKE8wwGAxaLBUhFWZZ0Ol3qqibt\nD8HEfPxjn+L52x/k9u4e+9N9EhUTCM+bn3odf+Ebv4ZQ95gvxkQmoG0qnBWMRxO0UeTFHI/AWY8K\nYpR1hIHGsnzVVlVFXddUdYm1NYExWAdlXpFlfbQKyG3DwcEeH/nEH/DLv/pb3N1rqGyDzlrSzlO8\n+7u+i9/54K+yd/eTvPVL30BcV5RFQYAEv3RKeaElWc+R9ElHMwRfeeFIAdPpBGctQiqyNKFoCxqj\nOZxO+Klf/DX23Q6nLr6DqjQkQcaoGWI2nuSjf7TL9//r/8Zzdye0WIq2orN1mrS/SR2n+HqMzXOo\nHK5sqcuW1nq8kiAFURzTOktVVeAlTVnjmhZtUtZ3HiXd2OBgOsE6+MgnPsVP/MJvsJv3+Evv+l6E\n6dFRMVQHTErPqUtvQydneP/7P8DCStAaqy2Fski39OeIjEZJhz8uZiFYtiZrLUoprLUgLM5L8rzl\nxrPXmFUpX/XOv0zn9CN0B11++yMfhKJGeMncNoxzx+ksZON1b2Pj/Hmq639As/8M4eAR4k6X8bXr\nmDCiKCu8AO9ainyBFAoRSMrFDOEdVbG0cLRti3eOPM+5urbG3YM9zlx9HZ++fpdzX/YtyF/6DWaj\nu/zkj/8IgQKlQ+Rgnbe89W1M9ma89eIZDm8+zQc+/PNcPL1JnHQQzuMokFZR25bG2SPpk45mqcI5\nnHPE8XJtxRiDEJKyrCjzhijK+Ppv+nZ81CfcOEvS7xCEMVGjKCZ7LO7sM0xSVP8S59/4lbS+pY5v\n0ukNyLJ1dL/D5HpI63OUltRNRVAWKGUhiGmsp21q8A7hHW3TEkYRUiniOGZ3d5egEzOejDh38QKv\nP3WWH/u3a5g2x3uH8wKlM777u/4JvaRP7m9im+c52P8YVy9dIgkkSoBw0NDgBNS2RSn1uRXzx8SR\nkBQac3/A4L0nDEKk00jtUKFlcvMu/+eXfpI3v/Vrse2C9f4T/OA/fS8/9APfT99dxzU5G4M1vvvd\n38PhJ36T5z7zB4SyZu3NT5J2Bc89c50o7lBVCxbFDCUC8sWMNO7QFDmqG+OammI+o21r6tYBjrLI\n0dEGbdNSLOZsDmN+7md+lE53i+/8tr/I4eEuv/Yr76O3dprv/XvvpWxnfPrDP8ft5z5BO9unm6UU\nbU1dNfQ7PXwLHo2XgtA7qrY5kpZ0JB6s7/mb3+G7nQ5KSiITIYVg2F+nKnPmeU7Z1IS6w+F4htMJ\n2WADk/aIu31+70O/TqA8WZSylqWcPb2ObXMef+z1CBTds4+ihePpj/4K89ke1tXYxhGGHeIwI44y\nuus9xqMDRge7VOUC55YexAhJ1ttiY+sSZm2L/b2bVJMDAjQHk5qD0SF7ezeZzHPmhSfNWu7eusFj\nj1wlCROU1NR2SttUSCRCBrS2xuEpm4bRdMZkMuM//vz7/ux7sDrvCE0I3oNzCKWYz8e0rUVJweZw\ngNAh3X6GtQ4lLXV9j+rWTb75HW9hY32Ts2fOcu3aXawWPP6mJ5ju3iOfTFFKcufZT5LGmjKXFHkN\nYmWG0hGht4z391jkc+q6pm6a5WRUS1SgcW3FZP82W2unWd84x1SlxFIzPfwswyhm+5HXIYSjqgtm\nszFfcuUSdV2itKRtW2ZzRRhlOL98pVsHvnVIodAywB0XiwOAdRZvHSaMCExAGicURUGSJOR5ThYa\nnHLIQKG1JIyGBNKQpgkHd+8yG03I1i+wtbXB7Vv3CIxm7fIjzPbvoLDsjQ7I88VSSc4iA8diPsVb\nCwKsWw4UALwTtHUDQJFPkVJTjGesnd6mGM2Zjqac397h5o1r+LohjAOksiRrfZxtEWlEvpjR4HBx\nQtu2OOuoV36Czi3dxV7qI/7wcCQkee/J8xyjA3Sgaa3FrSad1lriOMZIiJMMkIQmItAGazVYyc75\nRyhrR9hJl62xasm21pkWM5r5IfgSXAsSbCvAC5zyKOURymOCiOm0wDlHVVW0ZYsSHnCEYYgSFuNL\nFvt3GPQ6qNYSOMGZs+eZLUaUxZw07dK6BQhwtsGYiG63x3i+oKlryrKkLFoa/+KE1tqHv0wBR7g/\nSWtNlmYsFgsCrTk8PKRpGuIkQQhBHCZINKGOCIOITtYjjDLCpMvWmQuE2ZDZaB9ZVShvObh7G2MU\n+XzCfDamLHPqugLU/RosJbRtQ9O2S7vhylQTBAF4B87iXctiMaHKRxSLEaODO6RpiBNQVDU6MDgh\nSLodkrSH8wqhI4Ioo7EC7xx1XaOUIokTgiCgruv71o5jYxYKA4PwUFclw/4AJSQbm2tkWYKtS4yS\ntLZBCE9gNFIK2rZh0OvSNgX37j6PcAuuXjyPiAL6Z87Q6WSouoSixDUtVd0iMTjbLPsGawGJ9wJj\nNEoJwCKEp3ElItIUtaWsHVJopqObVLNdlG+w0nM4P6SRDVGy3CMVqRCBp61LmqpAS4+WYJQiiiO8\n8DgFRirCwKCkxOIpmoe/fH4kJC1Wi31t21LkBU3T0DYttmmwrSU2IU1dY4whjl70U5NKYkLDbDJh\n0O8xm43J0oQynxMbw97tWwgJdf2iz3jb1kgpUErivCOKQrxzaL20hEspEVLS1A1KK/BLC7hUmjiK\nEDiq2ZTtjQ0uXjhHkc/ZXB8ymRwyn00ItKaTpUtDLh7bNigBJtAYre631LJa7ok6ivWko9n6ohRV\nVZGlGdYtLeLz6WzpYhUnYB29Xg/btjRts5oEepqmotNJ6Q+7TKYjZCgpFmPK+ZjR7Rv4ImexmNE0\nBXVTYm1D01Q43yBXfZKnxRiD0hopxP0JptKKpmkJgoCmbbHWsre3T7WYY6RlOt6lWEzYWO8TR5ok\nNgRKEihwbU2gBXiHVoIkDMA2aMWyj1OKOIpwq1fhw8aRkFRW1bKGmYAwDBmNRlzY2cEoja3qpYVc\nKUwYEoUhcRzT7XRBwGRywPrakLW1IYFwzGcTFA7aBoNDa4lzlrZtqOuKwGi0lggBOtCAx7rlAEFK\niV4tmdR1g1bL/su2LQhJt9tBa4lvaxQeo+DWjWuUxYKmygnDgCBQSzlSIuXSyca1DVKAaxqc92il\nHzyG+qHjSEhqrcM6x2w2o1jMCbVkNpuitEQbzXDQpykrhIeiLBlPxlRNTRAGzOYznLPsH9xlMj4g\nijTjyXLl1sSGqm2QQUDTtFRVTdMUgCOQijhKSbpDAhOigwhjUpSKUDqgsS2tXc5qgzAkTmKsbbFt\nRVnOMaHkYH+fMDSERq2cVQzCOZRwLOYzfNuQJtFSaV4gvAShcLbFCFAO8Mdkf5IxhsAEOOcIgoA0\niSmKHK0USRyxyBfgPW3T0O/32d7aRgfLLY+XLl/m6ac/QxLHpJ0uUmo6vS4qDGnxhHFE27bYxqFV\ngEOihMYog5Yao5ZrSGbVgowOwC8XIuumoakbnIeyrji1vU2cJagwYDKdEscRJghw1pJEEd47hJQs\nFnPC0KC0Il/MEVIiV0sk1jvatsE1zaolHROS6rp+cTOV97RNi1RQNSVNU1HXJRsbGwwGA5qmYTKb\n0uv38av1mTCKqMoK70AFBus8Wdql0+vTNC15XlLVDVVlwUWYoEMWZsTCEMuAtNvFBAGBEgTCEhmz\nnNC2jsWipC4bkqxLXlXoIKKoGrQOEGK1xNHa5errfEGZ5yRhBM7hWksURZRlSb6avFblgqJYMK9K\nauex4uEbWY/M7+6FDVwVHgKN8ZKqzFl4hzEGvGexWCB0QBQn2JUVOUlS8rImSRLqpmbYH3A4m7Ko\nKppiQds4dnf3aasKiSE0ejlaMwmNUBidoqIEFbUIFdDUNbUDpQKKomIxz8nSEqUDTBijlWEtiAiQ\nHOztoZRcbnY7c4bD/XsID01ZUVUFCKjaikBrdKApqhrf1gjhUTrAK0d9BEcGHBlJbdtS1zWpCZaz\ndmcZDgYMe33wrF4vKVGSsH9wiFIBVVUzOjykLCsGw3UCEzGeTOl0B4SBY38x5dbN24wOx6z1uwRB\nRBpDlgboNCXbOEu8trnczl/OcCiCIidy0O32qcoDqqrm1q07nL1yhcAkpIM+HoFoLEppnGtJ04zp\ndIoWkiSKCQLJeGyxziKVomns/SmGx1LVBUVdUTWW+ggGD0fyumtESOsUeJDeYxAYbWiKmoO9A4p5\njtOSGseiyImiiMVsxsXNNbS3nD19lnlVcetgFzRLC0HtiHvr7B/cI0k76DAm7qREYQI6QKYRVnmM\nUYi4C3GGCwKEkXjlUEYQpgppLI1fmn2kNqjAgFTMm4Z0bZ3N7TMEJiEJUzq9AVXdggxoWo/zAu0k\ndVXT2obaV4iqRQlD0QgqL1Hm8x6Y/wXjSFqSCQMaa5mVLXhLJ4kJs4TaOrQMsF7QWVjSVBIQLA2u\n3T5tI5FCIR24skWrgDhOaZvlMvh8PsM1LSYIOL21TRSHhK1CVgp3MKGeFkxHY2QQIYRDFHPErCJw\nio3hBpGJOdg/YD6fsX8w4jaCKIoAAAHdSURBVNRWyDwv8R4kGomnXFRLt+bWEilPFicU0zmRCZlO\nxwitaBuL9wIs1LillUEs51PmCOr9kZA0zhes9Qd4E+BDQ6UM96Y5URDCIKO1AploWqOZz6asDTeY\naUXVWEySMrMNIgqIwuXWkiCKEYDxjhJNf20TkaRUePYdxEqQeItxFWUDyua0zmOdZdLU1I1DS0OQ\n9QkbWFjBYV7wxnPnqcuWqmmoy5xAS/rrG5Q4AiUox2OsDjE9w2j3LqY74O7okFaFVC2UTlNgyX3L\n8q9pxMpD6eHiSBb9lJJeSUnT2v+nFiynm9CsKtwL1v0XzyBangvkVmEOvzq4w92P98LxMEKCXZ5l\n80JxcM7fl+NeJvflcPfDXujsX1kXDw4FloNscf+ondVZOvdJAoX3zUMdPRzR2ULBUqiU3C+iXP0N\nklupxlle/IujVUmlXf30S1ejdvXtPGi1ejZYxlWAbdFuqWz3kr9LWvqFg0XSLsPky4fG7Yol8SLJ\nUi3z4PxK/qqKSV6871b33ANyECDN/Xvel8eBpJMjpx+mvJPz7o4BTkg6Bjgh6RjghKRjgBOSjgFO\nSDoGOCHpGOCEpGOAk7+LOwY4aUnHACckHQOckHQMcELSMcAJSccAJyQdA/xf2IdrgfWPkrUAAAAA\nSUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAGoAAABeCAYAAADLyoEsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nTvYAAAAASUVORK5CYII=\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAGoAAABeCAYAAADLyoEsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nq69lGE8vRwAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAGoAAABeCAYAAADLyoEsAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAGkAAABeCAYAAAAg/TovAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0\ndHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nO2de6xt213XP+M1X+u1H+fs87zn3Ntz\n7+2T9ralFGyhFA22gBoIRFMlFKNQQED/MGiMNkZUNCSYiEhiFBMTHhWLSglSG8RCoe1tS0u5bW/v\n69x73o/9WmvNNeccT/+Y65zu1nvbU9mbdsf9TVb2WnOM+Rtj/L7jN8ZvPLdIKXGEr27Ir3QGjvCl\ncUTSIcARSYcARyQdAhyRdAhwRNIhwFecJCHEfxRC/OTy+zcKIR7/U0o3CSEe/NNI60+KrzhJe5FS\n+t2U0ou/VDwhxNuFEL/3p5Gng8KXU4Z9JUkIofdT3hGWSCl90Q9wEfj7wKeAbeAXgGIZ9s3AZeAn\ngOvAf1o+/w7g48AO8PvAK/fIezXwMWAG/Arwy8BP7pW3J+59wLuBW8Am8LPAS4EWCMAc2FnGzYGf\nBp4DbgA/D5R7ZP1d4BpwFfjrQAIe/FLlX767tiz31aUO/uuesL8JPAlsAf8dOL0nLAHvAJ5Y6uLf\nAOKFyvCC6d8jSX+8VNga8IEvUKoH/sVSSeWShJvA6wEFfN9SRg5kwLPA3wEM8N2Aez6Slu9+AvgZ\nYAAUwBuXYW8Hfu8L8vkzSyWtASPg14F/vgx7y5K4Vyxl/eJekoC3AX/0RXTwG8sKtbrM95uWz78F\nuA28Zlm+fw28/wtIeg+wApxbVra3vFAZ/qQkvWPP728DntqjVMvSspbP/i3wT75AxuPAm4BvWtZG\nsSfs91+ApG9YFko/T54+r4D0tbMGLux59g3AM8vv/wH4qT1hD3OPlgScAiKw+jxh/x74l3t+D5eV\n7v49JL1xT/i7gL/35ZJ0r33IpT3fnwVO7/l9K6XU7vl9Hvg+IcSP7nmWLd9JwJW0zOUeec+H+4Bn\nU0r+HvJ3HKiAjwoh7jwT9NbIMu2P3kOaL5SPrZTS9vOEnaZvugFIKc2FEJvAGfrKDX03cAcLeiK/\nLNyr43Dfnu/n6K3hbt6+IO4l4J+mlFb2fKqU0i/R9wlnxB5NLuU9Hy4B517AGfnCNG8DDfDyPWlO\nUkp3FHLtecpwr7gErAkhVp4n7Cp9pQRACDEA1oEr9yD3npcf7pWkHxFCnBVCrAH/gL59fiH8O+Ad\nQojXix4DIcS3CyFGwB/Q92E/JoQwQojvAr7uBeR8mF65P7WUUQgh3rAMuwGcFUJkACmluEz3Z4QQ\nGwBCiDNCiD+/jP8u4O1CiJcJISrgnfdYblJK14DfBH5OCLG6zPc3LYN/Cfh+IcQjQogc+GfAh1JK\nF+9B9OeV4YvhXkn6ReC9wNPAU8BPvlDElNJH6D2en6X3hJ6kb39JKVngu5a/t4C/TO+9PZ+cAPwF\n4EF6j+3yMj7AbwOPAdeFELeXz35imdYHhRBT4H3Ai5eyfhP4V8v3nlz+vQshxF8VQjz2Rcr/vfR9\nzWfonaK/vZT7PuAfAv+FvkJdAP7KF5GzF89XhueF+Pzu4XkiCHER+BvLDB3hK4CvqhmHIzw/jkg6\nBPiSzd0RvvI4sqRDgCOSDgEOZNZaCPH/dRuaUhJfOta948iSDgGOSDoEOCLpEOCIpEOAI5IOAY5I\nOgQ4IukQ4IikQ4AD2oKl6TfC7BnTStH/jnt/fwHiMr5WQEQu61CMsf8WQUu1fBaIpLviPgeBBjyp\nT0OCVJLYhT7ZO0lJjZSJ6ANS9rKlVMQY+0gRFIogId7JlxTI6D8/zT+Fan4gE6xCmHSHJMmdMgsk\n/cYD6Olb0ta/IyUiQiS+gPI/H3eV/UXClVQIQCCIMZGAXsXyc5Uk+ruyJIK4J2cS1cuXwDJM74kT\nSegY7pJ6J2EX9nfG4UAs6ad/5HtYXRlT5hlGCaxzjIsKIyVlXiIS5GWBkpI8z0GAUpqsyFHGIJSk\n845yOKaxLavr68wXC06c2KBtOxKJEAJaabRRNIuWPMuw3rG6fozpzjZ2VtNM51z67FMQ4djJk/zP\n97+fWJZIo7l+/Rq2bZEkgrMYLVjUCzY3tzh58gRlVTEarXL16lV8imzv7rCzu0PbOowxtM5Stw3S\nTHCdZdE25PmXXAn/f8KBkLRRRIpYk7kO5RMVkNuOqijRrsMoxYDYNym2QQqBVIrMlEg0RV7iCAg7\nZbUs8O02a0bRbF0mzzPysmI+nWKixpChVQTfkgvBzrOfwRjD8dGAuYNu1TDIMhb1Td7wygvc2plz\n9cYN4u4tcI52UbMymZAax5hINS7YvvIMncnZVoqTp07yzHOXmZQFmMC080TXERcN0gbycoiRUBYl\n0+msby73GQdCUlRDnFB0zlNlBikFWTmgjpGBKsjLkqAz2rZjVI1AKaQxZGtjFm2L1BUyV7jO01nQ\nVYkqcqrBkN35DjZIxqfux9qO6c4UUAwGY5q2ZXxijXk9Y2th0YMJ9z3yOma3bnDuRUOefuIZcis5\nc9+QLSu5fPkyW92CzevbJO84trbGuTPnMMNjtE3D9mLBJx9/ltXjx5kcW+Py1qfQw+NsT3dJZY4o\nEpK+T3M+MB4WWHsvO9C+PBwISUZGjBCUg4rgLEomRLJUJiNLHuUtAFWmCK4leIHUJXaxA0lR1zPW\nj29QjUZE51lYR3IRbGI8HiHQvTJERj6uIEHdNmipmNczvLcMjMY6TxsbzErJYjFnUCgqYTl39gSj\nyvAp5fms8Fy8/CwqK7k93yVeDbz0gfsZVWf4+Cc/RSwykmt45olPc2J1wubOFNfs4pEIpVk0kaoa\nIGKkaz2dtfuuzwMhqShLku9rlNIKoxVKZrgk0XlOh6DQEqkUItNIo0lak43XGQ8H7M5rZsFCY8mz\nAjUeoPKcoA0mr3AukA9GhAT5oiImjxQLunZGPsio1ITZzi5CGaJLDNaOE03NmJyqGrN56ybjLONb\nv/nN5L/7AdqdKddn20QFV27OuHztMvedOcObvv6NXLx4kRs3b2KE4caNG3gCFx64wBPPXCSFiO0c\nUKO1QRtNXuT7rs8DcSCD6wlKKZFiBBI+CITKUFmJkxozmOCkwZMxbwJRF3gk27M5NgVUkZGXJeVo\nSF6VJCWJAlwAaTJi6L03jAFlCEmQ5wNAEqIgZTnlZIVBNUGKjGp1g83dOeP1Y9x/4SGqqmRjY4PX\nPvIq3vLmb+HYeESpFEYqrA9cunaTd7/nPRw/fZqvedWrOH3qNFIIXv3Ia6iKAlJCSYkxmhQT/dY/\n2Nx8vo2ufzIc1KIfkFBKEYkopREqx0XJrLHs7O6wud1SlCWzRc1gNCIbSJg2mMIQcBglMUiMUiys\nJR8MQfRK7KyliY7xZII0iaQUUq+gpMYtFohcUA4GSB+5dfU5jp8+SwKK4ZhUGGrhWd1YY9HVuGAp\nyowf/4Ef5NOPf4Zf/633EnxiazqnKUp+8dfezbG1db71zd/CyrHjfPBDf0DnAw+96AI70xm3trcx\nxhBjZHVlFSH2v94fyDjpV/7RDyQlBEZrlFQUWU5yBpcCT127ymA8QTqHUgoXBdZ5UoKtmzd5/eu/\nlo0T6xgj2SgHeC0o11ZY4ClXVvqxVOqtMi+GYDpUVEx3WvKiQAYLpiSGiBSJK1cusbo+ZjSokCTq\nzZvUsyllXiBdIjaWemvGbFFz6fpVMJr3/u/f4bHHHwNpSDGijWE8nnD+/HmOj1e4ePkKrQ/Ubcfm\n7Zvs7E4RyhBCJAIfvbL71T9O8iFg8gKlDcREiIm6rblx8wbdoqYgcvLsaW7cvA1CYpRgd1bz0le8\nnCs3bvCRj3+UY8fWOD4eYa3j9Pn7GIyGnLtfkWV9298sppgwR2YlqIKVyQqubUhGghJIkSB4Tm2s\nIVNE1jU721sI75CthaCwwdO2jpaA1Yaoci5du875Bx7CS8N0Pmdzc5P5fE53e5vt7SkPveQCL3rw\nAlev3WC6fYkXXbjAk08+Rd1aus4itPrSCvoycSCW9K53/nBSCPI8J9MaHwJRwO7mJi97+CGeffpp\n1k5uYIMnRMH1W7dxLiBUPxgcjkYs6jnjwQAtFbdv3eLsmTMYpTl2ao1Tp06QDwxJBqKZgMgwpqRY\nWcHWO0Sje6vZ2SV0DRUSN59DhMVsRlHkbO9MmS9qXBJ8+rNP4KOkLCsef+oJfIqcOnWKetHw4Ucf\nxdqOru1ICZy0ZFLz5je8CS0Mn3jiUwiheOriRYiSxrZ8/Pp8Xy3pQEj6z+/80SRSIs8yiAlrLWfP\nHGdYleQmYbQiaYkxOYu2o6iG3Lp9C4FkNquZT+dU1ZDZbBclFYt5jZSS4XCAVHD23DkmaxOq4ZCk\nc7JiBLrAi0RZZWitaLY3mW7ephCKdj7DtS0pCqbTOePRhIvXr/OJP/ojLl27wdmzZ7m9s0ORF6xM\nxkynM27dvs2FF93P6uoqW1ubfPrxx3nu4kW2fUs9bxFSMxlPeOSRV7FYtFy9foOdnV2m8/m+k3Qw\njoNUjAdDrLVUVclqZtDBIVNEG4mQASkg15I6OkRynNhYo5lNGRYT0uoEk5V4u8ZsPkee2mBWz3sX\nN1N0rmNROzpbU40iKUiGawOyQcF8toW2He1siomJ0M4QztE1C1rrsC7x+BNP8NS1q5TDEVLe4uyZ\nc1hr0Vpw3+mTuPU1Hn7gPJOVIWfOnGFne5vXvvLlfPjDH+aDj32GKzduEAGv4NFHH2VQDckHQ4SQ\nnDhxYv/1eSCW9I9/JGXaYKQiM4Ysy5hUGZlW5JmG4FGV7k+xKU1KEa0ULgRijAwHQzrbEWNkPq+p\nygHOOpwNaJMRYqQcDvv+qSqRUqNMjiwHUFWEnRvgHF09R1hLtJG6bvAhILXk1uYmKRiEgkW3AC0o\nyoL1tVWqvGQwqGjrBvBY5wgBQNIsFkw7x1MXn+U3f/v92BC5dO06WmUEQChFUVa8//ErX/2WlGLv\nfgsEWmuEFCilMJnBe4cQiWZrSp7nBOHouo7JygSh+njPXHyGB+5/gESiLCo6a8mLghQFMQRCjJhM\nUTdzhrlCK0WKAe8cgogxhrquSTERvGe6OyWmxHgyIS8LJuurzHemCJFArbC9s8VkdYIgkdwCW3tC\n10EKiJTABpwLBGvBwrnTZ/jOt34bH/3kHxNCZN42OJeIAsS9nw27ZxwISToz2K5jWA36idMsxwaP\nbz2hbREikglF23R9fK1xzqOVIZHY2NhgOpsigKIoGa5MaKxF6xy7s0Mi4INjUBWQAt62mEwhjaCZ\n7SJiQAiBVBKT56ysTjB5RucsdVuzdmydKCryPCN6y7HjY6azOaPhkKtXrjAcrNMsOjQSbx2283Rt\nhw8RawM2JLQUvOzhBzl2coMr167z9LPPsWha3GGZFmLZhHrvsdZCijgREESC7TBGMreeQVVRVhXW\ne/CeMkoyYwgIQgjEEBFS4oMnKyuatie1axuqTKO06pc5JMTgKfMcJQR2uo1zDpUS3nukkjRdQ2M7\nTJbhnEMMCuq2hRDogkflObPOMlxZwybBYLJKN2/xKeJJtF7gXaTzHa0XdCEQFeSZ4f5zZxlPJmzv\nTpnO6n1X54GQlGmDTx3Od8RgaRsYVhXSe7quIY0LFiLw7MUrnDx+vvemppu85nWvorIdpUw0sy2K\ncsh8+xZ6mjM6fh9aDGncLlWekceI8ImQLEE6qBSLriUN1vDxNoOiAtvRKodNHi0lq3qEyA3FaEzX\nBYa5ICx2WNQ1UWh2t3c4f/4B5i5w9daU7e0WdKJZzBlJgVAQbCTGRNO0NC4SvEcpxYm1FdbHQ0Lf\nge0rDoQkJQGlCDFigyczhvlijkTwgY/8IQ++9CXsOse1K7d5/x8+zbFjx7h46SJb1nP+1HEmVcZk\nkCPnm0yqkqpUCG8ZjQYspKRtGwgBfKSsCqRWtF3LyBjalJAIrO1IXUcKASHAWUvrOgZ6zM7WNoqI\nj57pbk0SmlYPuWYt/+1X38fN7SlbuwvcdMpLXv5iBlXGUEk2r1/npS+5QG0bnI9474kxUJYFUkqk\nFISw/9NCB7RUIdBZRmct1gqa6DEqsds46mzMz7/rN+iS4CWveDUvf90b+dVf+WXKTPPYb/0BRgSM\nhPXVMS8/MeQ7/tw34WKLtc8xWZlTDsYsfIf3HXjHPDikUpjxhBAC1bhigcBZh6GfcoJIDBGkYDgY\nUjct9c4280VDJ4e870Of4H+8/wP9rIEwSCFRyYMakG4v2Lz9DMY71sZj5KXbXL95jQvnz+OdQ2eK\nsshIKWFTQLCvjh1wQLPgmRYYo8nynNvbOyxs4OJz17l49Saffeo5ynyFd3z/O/jRH/4x/vixxxhW\nJckHopB0SUE2QJUThmce5plbNdu1xYfI1tYmt7e3ianfO5GiRyHQUqKFYjGbgw8MqhIlJCytKnQW\ntVzfunrtKrZusC4gdMkHPvpJzj78CD4EBJ6YIov5DsrPybKCv/a27+WHfuhvUVUDPv3Y4/zae3+H\n7bnns08/Q4yRqsgZViWZUigBav85OqA+SQpEluFCi0uJTz/+WZ55/DnioOTYynHe+t1v4cSxFTJv\nEe2MKllS7HCywHUW20UyfYI/89bvJut2+dhv/zpvevXLOHlsjViu4utN6Dpi8Hjreu/QdpjBCvPZ\nDHanSCnp6pboPYSIjYEuJQbDEc32jFu7LZev3eDr3vBmri9gkTIaIOarRHJk1qLaBapreNGp47z1\njW/k6rmzPPrUJW5sbXHuax5Ga836yipGaTCR4Bwuun3X54FYklEFRkhGecaDJ9d47Yvv55FXPszD\nZ45zblKgpjfIjMDojG94w5tpU8RJyzBMWckUZ1c3eM0rXssDJwuOGc3L73uIYjhErJRELHmu8dYi\nXCBEj4sRshyZK2CBCL27HJB412G0wBhFXuZEIifvP8tnPvYYx0YbbKwe5/6zJzm5NmZFJdbsbY6H\nmsIrrNSYwZBmVjNGsqFzzhjPN77sQdZzzcqgYCWXDHOFkqCUIUqz7/o8EJJi6gezRVGwfuwY6+vr\nPHThPCdPrHN8bcLFZ57k0x/6ILefepw/+/qv523f+TaknuDKCcV4zP0PnePbvv1NhFubXP7sZ5jt\nbPO/3vs+7LyhqnK0BO88SWpIiugThEBYtITFAu89wTls0/a7rUIkpv4zqAbcunULT+DypWfZuXaJ\n3Dd8z1/6i4zHE3RVwqggrVT8+A++nczWXLv4BE8+9SmevPgEaysjikJz4vgqJzaOkWUZWimyLOsX\nIQ8ABzIt9LGfe2fKihwbAghYLBbszGq2d3YILpFJzdaNmzRBcuGVX0s5HLK7O+X25hb3ndxg89Y1\nmvkuG2tjSl2yvjph4XZ46CX3I6TEzXah6wBNQKGzjGJlgq4qvEwo5+imLbZeoEJLiBYzrBisrWGy\ngotPX2Q2tzSzlms3trk9nTFZP4HtHPOuZXO6w+u+7nV0TcMnPvoo7WzO8dURk8mYum0ZDUtOnzzG\ncFiQKY3QhtZHtrZm7NQ1P/jTv/DVPy0E/WZHLQRIQZbnjINDyglN3ZEJxfjCiyhyw2gQKDLHqY2K\nuoSXvPLF5KOvZfv6FXSe89zF5xgMh5xcPUk20CRnMUVB3VkCAuEFWZXjW4uQfc/dNi22dRD6qSJE\nv7Q9nc1YW8sZjUZMjlU898QzPHTfKU4tVjl27j7WN07ysY88yqsH56lyxaZ1vOnVr4AQ2W6mzJoF\nJ0arjMcVea7IckUmM5KQiJAoygJzaGYchEbKDKX6pk9LS2ZKVnVBZTqC9xilic4hYwchkZkMOdRs\nXr/EzhNzhNb41nHq/vupVsfszLcxwqBjh9QGUj836APU84bx2gqutWRFhhKaFFpSSBAkQklEEqyM\nBgzWxyyiZ7HTcfrMaZrNObjA7PZNumbKww+cJnpH1zSsDDK62tO4Dq1hUGasDcdorSiyjAyNEoIA\n/SyLuLN1YH9xMN5dOQChKIoc5yyyGjAsJCEG5vWUGD2F6jvYsiyJMZJlGfN5i1KaUTUkL0rWHz5F\np2DWzTl+cqNf5U2WtrGEKHEhYl2gNBrnHUJLovMkB8H3YyMfoDAaENim4emnnuT0hYcwasF0Z8bG\nfWOG45rd3W1sZ2nDFEQiBYv2/aSpGRSMdYUNgUwapJRkWYYxBoikZSVr3d177fYVB2NJIZLnBSIm\nREwMyqpXWgogKkLwyNjXuJgSUimUMawdH6DzgpUsJwDbs20Ga2NWVyd07YJca9qmJUaIUtJZT9N1\nqMywmufEEGgXCySGEAK26zDG4BOApGstg/GYdragGFUEErubu8hMsHHqFPPpDEQ/HiuKgkGmiCIR\nU8KGQNM0mCzHGIPRmhAjSfSHC2IIgESp/ffuDoQkKfolhRQTRZ73AzwtkEkgRIaQBpE0Whu892Ta\nYIqcRQhkowHlYMB80TKuNKKQTOstVgYDomtJQfRLFkLhhCAKKKqSGEJ/C2GC6D1GSFoSLkUKneN8\npCwNofOYpFAyYYxkdW0CUSCsYKwMzrWcKkukEijrsN6RBFjv0VohhUZrjRQCISWd8wTpsbaDuOcE\nxn7qc98lAsT+SImWCiUEJBAiIaXEGIUxBiElMUWyzIASCCXJigJhFFduXMMTUYXGB8tkPOrXiaQg\nhIQ2BpSk7TpQktDfU0pvYonoPG3bElMiiYSQfVPbNhZvHbELpM4RbEeWZUDsN2EqQ4iRvCiIEeRy\n46bMDCo3lMMBg6Ik14bcZIiYQIjeepLEWo9g/93wg7EkAhKJSAklZD8XRgIpCVH2ywWqRirVN00x\n4XwiG5QQI8OqoCwSKkBygazIaWcdyTrGozGbt7eIEUQSZEaiNcjMYK0lGENyLUkIOhf66aHGMswL\nQFJmA1y7QKmSQTmktf3iXtNNQUqKPEdLhcxymtAgM40AdEqkJDCZQQpBiJEYI8r2FUAmgIjz3b7r\n84Bc8H5FVkpJCJGQAlluSFKgM0VnO2RWkFJCZQYRwIcIeEIIZPnSypQitIm8HELnQWuuXbmFEIKm\nbUkxYlSJFjlERWEG1K4mLhWYUuzXjgR4k6EEtIuGwWRIU3fEtiMvSqQyYALOOZIMOG/7Q2Wqbz6l\nUihA5oLkVJ9vKbHJIwT98oQUhBRRev9VeiAktY1FCg1JkpmCGCPeC0yeIdGURU5UkFLEh35/Q1Uq\nbLBUgwFd8Eih0KpgUMJ0Z0ZsLTpBipEYErs7O/3+aymRAlII2GAhRKTsW/EUE3W7oLGKsighOLJc\noxFIkxNkv2lGREiK/ixV06KUIVs6HNE5YkworUg+cucYnF0+T0oS+hadkPqlmf3GAR0mlGidIYTi\njsejpEFKg1YGrTISGT5qqsEInxLOB6Q0ZGYAyVDkK3Rtg9aq966sI8W+f6kXi7uubgotrpsjsEjh\nyU3f92V5BkJgO8t0NmVaz3qlhkgzn2OtZzReBSQpQDmYIMjpr1LNsA6IGUYPESknxQzI7lY+rbJ+\nA0oIqOU+dZ8i+3vG7442DwBWCLwL/ZnT1veW4CL1bo2WGd5GYlRoVeKcoshXUHqEtYJZbanKFaa7\nNUhD1/Urn0WW0bYLgnNsbe/ifSSXGq0MJOjaFkREaIGU/Q3rA6MZVCXeJ3Y3t+m6GYt2TpAC3zVE\nZ+lmC4wxNLUDlYMsyIoJSo1JQuJCQilDioKEAhQxCHwXCDaC1NRNR72wBC9pFvs/43AgJLXO9tbR\nWZxzDIdD0rJZm8+mvfsawLYdEk1MEi0zynJAUZSEFMirgrwoEVIuN1h2eOfY3tomLD1HKUBIMJnG\naI1MkClFVuTkWUZZlmQmI89LmqahaRtQ/YJkNSxZTKcMBxUAJjcYbRgNRiip8T6ihEZJQ28ekhQS\nwfdDC0Tf38YQcc73GUGy/4vnB0RSEhBJNLZ3kZu2RUqFcx7XOZzzVCZjVAxAQl4UqCLHFDlCyX5j\nihR9Wy8EWisUAls31HXL7dtbCCUwmcDTojIwucJ5h1QajCKrCsqqosgzytwwWzTc3K6ROsd1HU3b\nIpREao2Pkc45Yoy44DHGMBwOwRvcIqFESfKaaBUkgQ+RznuaGJgtahZtv6fPBtdvvNlnHAhJLngW\nbbMcBDqSFP3JvCSRymBtP45xzqKUonX9/oNIwhPRRlOWJT74fpwi+iMwxMhsOqcqK4xSKAWT4Qgh\nFEEpzOoaYrIKRhGlQEpJnuVUeY7JS7ZmC1wSZLkBIUhS0HmH1Gp5SNoghMC5fi+gSPQzFp1DxP48\nVNc6vI+4BCrLcd6htMY6h/OeedPsuz4PprmzluA9ISWS6LcLehdRymB0jlE5KUSCD3TWIo1Gmeyu\nZzRYWelrdFUSU+hJn9dcfvY5UpTkRUlWZOR5hk4ZUmSI4SrFybMMHnqYajxGGYNPESUlmdGYvKAc\nrVAvWozWZEWB1gapFYHUj3WEQBuDVn3zGWNHDB3eLZDCk5JdjpdFPyXlPChJEiC0xMVAvdj/LV0H\n4zioxMII6uQIJLx1ZEoSbYdrG/AW4xOVUExETtElcpcohO77g+CgyFjMGnKhUQFsl6gXgqEeMJCa\nkckJIZG1NRUBiWB84jzBnCRfX0UiyFMC3yJSYpJpxjIy29ymaSDST+XkuqDSQ6LU1MnjU6SxHW3T\nUgSQ1pE6i2tbonPoKNAIMqlI1uNpcHR0rsPFwP4PZQ9onFTOAgOZUHg0nkFeEIRESEFSmigl7RBC\n7IjOkZuSTBkkidwIfNuiCsNwpGnqKZLAu3/tPTz0wEs5tj5CaRBKsljUXKsKxiJSZo7kNnH1Fu31\n2wgSjbNs1lOkzlldX6Wua3anUxAwWVldWkUgxsAQieoiofNEH/Gto/aRhABl7tbmQEPwHhsaOtuQ\nkkBiSKmlqReIcEiuCLgpFLqo+uXkoqA1OSOnek9NSBCG+uo248m437PWaowqmBqP94K8GlK3LSFY\n8nJIE1sYH4PJBkYJBlUJJGadBVHB3KC3PPP6OUJUpFow33XcvrXAxZy1wRqFzpAhp5l6olVMye46\nM0r11xgoIfFSY1uPT7DuC+bzOcFzd1zmnCAlBU4hg8YXIxa+oXaCBkk4gCX0g7m2Rom7985I2V8H\n4+Uynb331uxFBP0ClVAAWuIIN7kAAADrSURBVApCBJunz915AxAViPDF/4fKHTPYc0+O7j4/+E7S\nn8uCIGbp/5Z7R1bqP5nr3xGyv6gHBcHu75D2AO8W+nyU9DqK3ClURiT0HfYyjtdmOZN9560MpAdc\nP5EWoYx7RyKKDkWvnkBEARkFvYcll7cD9effJenuDUeRBQruVBzupHnn9qPeGgrsXTl3cPcuJPrb\nheo7rxF6eUKQbDwMJB1dOb2f8o7uuzsEOCLpEOCIpEOAI5IOAY5IOgQ4IukQ4IikQ4Ajkg4Bjv5d\n3CHAkSUdAhyRdAhwRNIhwBFJhwBHJB0CHJF0CPB/AMJN8iONjf/IAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Ub7wROrhWFO3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 176
        },
        "outputId": "6e9a8219-0f55-4493-d0a0-7b2d64660855"
      },
      "source": [
        "print(class_names) #labelsとの対応を確認\n",
        "\n",
        "#Dataloaderをbatch size=1で再定義する\n",
        "data_dir = '/content/data/hymenoptera_data'\n",
        "image_datasets = {x: datasets.ImageFolder(os.path.join(data_dir, x),\n",
        "                                          data_transforms[x])\n",
        "                  for x in ['train', 'val']}\n",
        "dataloaders = {x: torch.utils.data.DataLoader(image_datasets[x], batch_size=1,\n",
        "                                             shuffle=True, num_workers=4)\n",
        "              for x in ['train', 'val']}\n",
        "\n",
        "\n",
        "total, label_positive, label_negative, TP, FP, TN, FN = [0]*7\n",
        "\n",
        "with torch.no_grad():\n",
        "    for data in dataloaders['val']:\n",
        "        inputs, labels = data\n",
        "        inputs, labels = inputs.to(device), labels.to(device)\n",
        "        outputs = model_ft(inputs)\n",
        "        _, predicted = torch.max(outputs.data, 1)\n",
        "\n",
        "        total += 1\n",
        "\n",
        "        if labels == 0:\n",
        "            label_positive += 1\n",
        "            if predicted == labels:\n",
        "                TP += 1\n",
        "            elif predicted != labels:\n",
        "                FN += 1\n",
        "\n",
        "        if labels == 1:\n",
        "            label_negative += 1\n",
        "            if predicted == labels:\n",
        "                TN += 1\n",
        "            elif predicted != labels:\n",
        "                FP += 1   \n",
        "\n",
        "accuracy = (TP + TN)/ (TP + TN + FP + FN)\n",
        "precision  = TP/(FP + TP)\n",
        "sensitivity = TP/(TP + FN)\n",
        "specificity = TN/(FP + TN)\n",
        "f_value = (2*sensitivity*precision)/(sensitivity+precision)\n",
        "\n",
        "print('Samples = '+str(total))\n",
        "print('label_positive = '+str(label_positive) +', '+'true_positive = '+str(TP))\n",
        "print('label_negative = '+str(label_negative) +', '+'true_negative = '+str(TN))\n",
        "\n",
        "print('accuracy = '+str(accuracy))\n",
        "print('precision = '+str(precision))\n",
        "print('sensitivity = '+str(sensitivity))\n",
        "print('specificity = '+str(specificity))\n",
        "print('f_value = '+str(f_value))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "['cont', 'grav']\n",
            "Samples = 130\n",
            "label_positive = 65, true_positive = 30\n",
            "label_negative = 65, true_negative = 41\n",
            "accuracy = 0.5461538461538461\n",
            "precision = 0.5555555555555556\n",
            "sensitivity = 0.46153846153846156\n",
            "specificity = 0.6307692307692307\n",
            "f_value = 0.504201680672269\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "jSDw-HEoowRk"
      },
      "source": [
        "#固定された特徴抽出器としての ConvNet\n",
        "ここで、最終層を除くネットワーク総てを凍結する必要があります。勾配が backward() で計算されないようにパラメータを凍結するために requires_grad == False を設定する必要があります。"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fDzcc8m4o0wS"
      },
      "source": [
        "model_conv = torchvision.models.resnet18(pretrained=True)\n",
        "for param in model_conv.parameters():\n",
        "    param.requires_grad = False\n",
        "\n",
        "# Parameters of newly constructed modules have requires_grad=True by default\n",
        "num_ftrs = model_conv.fc.in_features\n",
        "model_conv.fc = nn.Linear(num_ftrs, 2)\n",
        "\n",
        "model_conv = model_conv.to(device)\n",
        "\n",
        "criterion = nn.CrossEntropyLoss()\n",
        "\n",
        "# Observe that only parameters of final layer are being optimized as\n",
        "# opoosed to before.\n",
        "optimizer_conv = optim.SGD(model_conv.fc.parameters(), lr=0.001, momentum=0.9)\n",
        "\n",
        "# Decay LR by a factor of 0.1 every 7 epochs\n",
        "exp_lr_scheduler = lr_scheduler.StepLR(optimizer_conv, step_size=7, gamma=0.1)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KbhM5pn3o-g7"
      },
      "source": [
        "#訓練と評価\n",
        "これは CPU 上で前のシナリオに比べておよそ半分の時間で済むでしょう。ネットワークの殆どに対して勾配が計算される必要がないためです。けれども、forward は計算される必要があります。"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fBPFSO6QpGOS",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "80b92833-7f64-4177-9ed1-2af9c4d5f827"
      },
      "source": [
        "model_conv = train_model(model_conv, criterion, optimizer_conv,\n",
        "                         exp_lr_scheduler, patience=5, num_epochs=25)"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 0/24\n",
            "----------\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/torch/optim/lr_scheduler.py:100: UserWarning: Detected call of `lr_scheduler.step()` before `optimizer.step()`. In PyTorch 1.1.0 and later, you should call them in the opposite order: `optimizer.step()` before `lr_scheduler.step()`.  Failure to do this will result in PyTorch skipping the first value of the learning rate schedule.See more details at https://pytorch.org/docs/stable/optim.html#how-to-adjust-learning-rate\n",
            "  \"https://pytorch.org/docs/stable/optim.html#how-to-adjust-learning-rate\", UserWarning)\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "train Loss: 0.7169 Acc: 0.5371\n",
            "val Loss: 0.6409 Acc: 0.6385\n",
            "Validation loss decreased (inf --> 0.640896).  Saving model ...\n",
            "\n",
            "Epoch 1/24\n",
            "----------\n",
            "train Loss: 0.6403 Acc: 0.6387\n",
            "val Loss: 0.6971 Acc: 0.5615\n",
            "EarlyStopping counter: 1 out of 5\n",
            "\n",
            "Epoch 2/24\n",
            "----------\n",
            "train Loss: 0.6205 Acc: 0.6191\n",
            "val Loss: 0.5932 Acc: 0.6538\n",
            "Validation loss decreased (0.640896 --> 0.593230).  Saving model ...\n",
            "\n",
            "Epoch 3/24\n",
            "----------\n",
            "train Loss: 0.6210 Acc: 0.6426\n",
            "val Loss: 0.5900 Acc: 0.6769\n",
            "Validation loss decreased (0.593230 --> 0.589977).  Saving model ...\n",
            "\n",
            "Epoch 4/24\n",
            "----------\n",
            "train Loss: 0.5997 Acc: 0.6719\n",
            "val Loss: 0.6424 Acc: 0.6077\n",
            "EarlyStopping counter: 1 out of 5\n",
            "\n",
            "Epoch 5/24\n",
            "----------\n",
            "train Loss: 0.5570 Acc: 0.7012\n",
            "val Loss: 0.6804 Acc: 0.5923\n",
            "EarlyStopping counter: 2 out of 5\n",
            "\n",
            "Epoch 6/24\n",
            "----------\n",
            "train Loss: 0.5720 Acc: 0.6914\n",
            "val Loss: 0.6077 Acc: 0.6692\n",
            "EarlyStopping counter: 3 out of 5\n",
            "\n",
            "Epoch 7/24\n",
            "----------\n",
            "train Loss: 0.5566 Acc: 0.6992\n",
            "val Loss: 0.6252 Acc: 0.6692\n",
            "EarlyStopping counter: 4 out of 5\n",
            "\n",
            "Epoch 8/24\n",
            "----------\n",
            "train Loss: 0.5449 Acc: 0.7305\n",
            "val Loss: 0.6114 Acc: 0.6692\n",
            "EarlyStopping counter: 5 out of 5\n",
            "Early stopping\n",
            "Training complete in 0m 28s\n",
            "Best val Acc: 0.676923\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AmVfI2NupnAi",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 347
        },
        "outputId": "d45b8605-2421-4485-8bd2-96d675bb978a"
      },
      "source": [
        "visualize_model(model_conv)\n",
        "\n",
        "plt.ioff()\n",
        "plt.show()"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "error",
          "ename": "AttributeError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mAttributeError\u001b[0m                            Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-58-b590b3d1873b>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mvisualize_model\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmodel_conv\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      2\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      3\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mioff\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      4\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshow\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m<ipython-input-42-7070053f367a>\u001b[0m in \u001b[0;36mvisualize_model\u001b[0;34m(model, num_images)\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mvisualize_model\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mmodel\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnum_images\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m6\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 2\u001b[0;31m     \u001b[0mwas_training\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtraining\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      3\u001b[0m     \u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0meval\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      4\u001b[0m     \u001b[0mimages_so_far\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      5\u001b[0m     \u001b[0mfig\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mplt\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfigure\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mAttributeError\u001b[0m: 'tuple' object has no attribute 'training'"
          ]
        }
      ]
    }
  ]
}