vgg19.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "vgg19.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/krishnaaxo/project1/blob/main/vgg19.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "uhdDbHB9UrTX",
        "outputId": "6843f19d-1ed4-4d08-81a7-28ba32db5e49"
      },
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')"
      ],
      "execution_count": 1,
      "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"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qE9d-6TTrTLO",
        "outputId": "f69c0a8f-63c9-4e47-a25d-81c650cbc1cf"
      },
      "source": [
        "!ls '/content/drive'\r\n"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "MyDrive  Shareddrives\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FDMjs551vIXl"
      },
      "source": [
        "from tensorflow.compat.v1 import ConfigProto\r\n",
        "from tensorflow.compat.v1 import InteractiveSession\r\n",
        "\r\n",
        "config = ConfigProto()\r\n",
        "config.gpu_options.per_process_gpu_memory_fraction = 0.5\r\n",
        "config.gpu_options.allow_growth = True\r\n",
        "session = InteractiveSession(config=config)"
      ],
      "execution_count": 3,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "C6HotZI8sOeE"
      },
      "source": [
        "from tensorflow.keras.layers import Input, Lambda, Dense, Flatten\r\n",
        "from tensorflow.keras.models import Model\r\n",
        "from tensorflow.keras.applications.vgg19 import VGG19\r\n",
        "#from keras.applications.vgg16 import VGG16\r\n",
        "from tensorflow.keras.applications.vgg19 import preprocess_input\r\n",
        "from tensorflow.keras.preprocessing import image\r\n",
        "from tensorflow.keras.preprocessing.image import ImageDataGenerator,load_img\r\n",
        "from tensorflow.keras.models import Sequential\r\n",
        "import numpy as np\r\n",
        "from glob import glob\r\n",
        "import matplotlib.pyplot as plt\r\n",
        "\r\n",
        "#import matplotlib.pyplot as plt"
      ],
      "execution_count": 4,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "j3rvFOZIvNh1"
      },
      "source": [
        "IMAGE_SIZE = [224, 224]"
      ],
      "execution_count": 5,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wG6Kij1OvsKw"
      },
      "source": [
        "train_path = '/content/drive/MyDrive/kaggle2pneumonia/chest_xray/train'\r\n",
        "valid_path = '/content/drive/MyDrive/kaggle2pneumonia/chest_xray/test'"
      ],
      "execution_count": 6,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fdhm34qPwSQi"
      },
      "source": [
        "vgg = VGG19(input_shape=IMAGE_SIZE + [3], weights='imagenet', include_top=False)"
      ],
      "execution_count": 7,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "c2WHVtdDwVuq",
        "outputId": "4ee6dbd6-ddc6-477e-b039-437a6563bac7"
      },
      "source": [
        "vgg.input"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<KerasTensor: shape=(None, 224, 224, 3) dtype=float32 (created by layer 'input_1')>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ov4BTAP3wYQp"
      },
      "source": [
        "for layer in vgg.layers:\r\n",
        "  layer.trainable = False"
      ],
      "execution_count": 9,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "hlJbf_zmxRvJ",
        "outputId": "338c5524-a2fc-487c-a209-f1b024b35253"
      },
      "source": [
        "folders = glob('/content/drive/MyDrive/kaggle2pneumonia/chest_xray/train/*')\r\n",
        "print(len(folders))"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "2\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "PbqqvRSgwzuQ",
        "outputId": "baae1eb5-7597-4f10-8b89-10b9ad4d5fec"
      },
      "source": [
        "x = Flatten()(vgg.output)\r\n",
        "prediction = Dense(len(folders), activation='softmax')(x)\r\n",
        "model = Model(inputs=vgg.input, outputs=prediction)\r\n",
        "model.summary()"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"model\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "input_1 (InputLayer)         [(None, 224, 224, 3)]     0         \n",
            "_________________________________________________________________\n",
            "block1_conv1 (Conv2D)        (None, 224, 224, 64)      1792      \n",
            "_________________________________________________________________\n",
            "block1_conv2 (Conv2D)        (None, 224, 224, 64)      36928     \n",
            "_________________________________________________________________\n",
            "block1_pool (MaxPooling2D)   (None, 112, 112, 64)      0         \n",
            "_________________________________________________________________\n",
            "block2_conv1 (Conv2D)        (None, 112, 112, 128)     73856     \n",
            "_________________________________________________________________\n",
            "block2_conv2 (Conv2D)        (None, 112, 112, 128)     147584    \n",
            "_________________________________________________________________\n",
            "block2_pool (MaxPooling2D)   (None, 56, 56, 128)       0         \n",
            "_________________________________________________________________\n",
            "block3_conv1 (Conv2D)        (None, 56, 56, 256)       295168    \n",
            "_________________________________________________________________\n",
            "block3_conv2 (Conv2D)        (None, 56, 56, 256)       590080    \n",
            "_________________________________________________________________\n",
            "block3_conv3 (Conv2D)        (None, 56, 56, 256)       590080    \n",
            "_________________________________________________________________\n",
            "block3_conv4 (Conv2D)        (None, 56, 56, 256)       590080    \n",
            "_________________________________________________________________\n",
            "block3_pool (MaxPooling2D)   (None, 28, 28, 256)       0         \n",
            "_________________________________________________________________\n",
            "block4_conv1 (Conv2D)        (None, 28, 28, 512)       1180160   \n",
            "_________________________________________________________________\n",
            "block4_conv2 (Conv2D)        (None, 28, 28, 512)       2359808   \n",
            "_________________________________________________________________\n",
            "block4_conv3 (Conv2D)        (None, 28, 28, 512)       2359808   \n",
            "_________________________________________________________________\n",
            "block4_conv4 (Conv2D)        (None, 28, 28, 512)       2359808   \n",
            "_________________________________________________________________\n",
            "block4_pool (MaxPooling2D)   (None, 14, 14, 512)       0         \n",
            "_________________________________________________________________\n",
            "block5_conv1 (Conv2D)        (None, 14, 14, 512)       2359808   \n",
            "_________________________________________________________________\n",
            "block5_conv2 (Conv2D)        (None, 14, 14, 512)       2359808   \n",
            "_________________________________________________________________\n",
            "block5_conv3 (Conv2D)        (None, 14, 14, 512)       2359808   \n",
            "_________________________________________________________________\n",
            "block5_conv4 (Conv2D)        (None, 14, 14, 512)       2359808   \n",
            "_________________________________________________________________\n",
            "block5_pool (MaxPooling2D)   (None, 7, 7, 512)         0         \n",
            "_________________________________________________________________\n",
            "flatten (Flatten)            (None, 25088)             0         \n",
            "_________________________________________________________________\n",
            "dense (Dense)                (None, 2)                 50178     \n",
            "=================================================================\n",
            "Total params: 20,074,562\n",
            "Trainable params: 50,178\n",
            "Non-trainable params: 20,024,384\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "8jQgJXONw4dU"
      },
      "source": [
        "from keras import optimizers\r\n",
        "\r\n",
        "\r\n",
        "adam = optimizers.Adam()\r\n",
        "model.compile(loss='binary_crossentropy',\r\n",
        "              optimizer=adam,\r\n",
        "              metrics=['accuracy'])"
      ],
      "execution_count": 12,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gmFtmy-Dw68b"
      },
      "source": [
        "train_datagen = ImageDataGenerator(\r\n",
        "    preprocessing_function=preprocess_input,\r\n",
        "    rotation_range=40,\r\n",
        "    width_shift_range=0.2,\r\n",
        "    height_shift_range=0.2,\r\n",
        "    shear_range=0.2,\r\n",
        "    zoom_range=0.2,\r\n",
        "    horizontal_flip=True,\r\n",
        "    fill_mode='nearest')"
      ],
      "execution_count": 13,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pA4zTMgdw8cA"
      },
      "source": [
        "test_datagen = ImageDataGenerator(\r\n",
        "    preprocessing_function=preprocess_input,\r\n",
        "    rotation_range=40,\r\n",
        "    width_shift_range=0.2,\r\n",
        "    height_shift_range=0.2,\r\n",
        "    shear_range=0.2,\r\n",
        "    zoom_range=0.2,\r\n",
        "    horizontal_flip=True,\r\n",
        "    fill_mode='nearest')"
      ],
      "execution_count": 14,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1HSnzwkDw-8R",
        "outputId": "80b616a2-1c7e-4abc-adfb-e2195abbeceb"
      },
      "source": [
        "train_set = train_datagen.flow_from_directory(train_path,\r\n",
        "                                                 target_size = (224, 224),\r\n",
        "                                                 batch_size = 32,\r\n",
        "                                                 class_mode = 'categorical')"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Found 5216 images belonging to 2 classes.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8W0XVmITxA58",
        "outputId": "80144e14-acc8-4c9d-e9fe-2fe8cdc70fb4"
      },
      "source": [
        "test_set = test_datagen.flow_from_directory(valid_path,\r\n",
        "                                            target_size = (224, 224),\r\n",
        "                                            batch_size = 32,\r\n",
        "                                            class_mode = 'categorical')"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Found 624 images belonging to 2 classes.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "W9qXlJsrxDkY",
        "outputId": "7b490812-cb2b-4812-ac81-830f6485e6c1"
      },
      "source": [
        "from datetime import datetime\r\n",
        "from keras.callbacks import ModelCheckpoint\r\n",
        "\r\n",
        "\r\n",
        "\r\n",
        "checkpoint = ModelCheckpoint(filepath='mymodel3.h5', \r\n",
        "                               verbose=2, save_best_only=True)\r\n",
        "\r\n",
        "callbacks = [checkpoint]\r\n",
        "\r\n",
        "start = datetime.now()\r\n",
        "\r\n",
        "model_history=model.fit_generator(\r\n",
        "  train_set,\r\n",
        "  validation_data=test_set,\r\n",
        "  epochs=5,\r\n",
        "  steps_per_epoch=5,\r\n",
        "  validation_steps=32,\r\n",
        "    callbacks=callbacks ,verbose=2)\r\n",
        "\r\n",
        "\r\n",
        "duration = datetime.now() - start\r\n",
        "print(\"Training completed in time: \", duration)"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/tensorflow/python/keras/engine/training.py:1844: UserWarning: `Model.fit_generator` is deprecated and will be removed in a future version. Please use `Model.fit`, which supports generators.\n",
            "  warnings.warn('`Model.fit_generator` is deprecated and '\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/5\n",
            "WARNING:tensorflow:Your input ran out of data; interrupting training. Make sure that your dataset or generator can generate at least `steps_per_epoch * epochs` batches (in this case, 32 batches). You may need to use the repeat() function when building your dataset.\n",
            "5/5 - 509s - loss: 7.8592 - accuracy: 0.6438 - val_loss: 3.6839 - val_accuracy: 0.7099\n",
            "\n",
            "Epoch 00001: val_loss improved from inf to 3.68391, saving model to mymodel3.h5\n",
            "Epoch 2/5\n",
            "5/5 - 107s - loss: 3.3953 - accuracy: 0.7750\n",
            "WARNING:tensorflow:Can save best model only with val_loss available, skipping.\n",
            "Epoch 3/5\n",
            "5/5 - 107s - loss: 1.4902 - accuracy: 0.8438\n",
            "WARNING:tensorflow:Can save best model only with val_loss available, skipping.\n",
            "Epoch 4/5\n",
            "5/5 - 105s - loss: 1.3750 - accuracy: 0.8562\n",
            "WARNING:tensorflow:Can save best model only with val_loss available, skipping.\n",
            "Epoch 5/5\n",
            "5/5 - 107s - loss: 0.9927 - accuracy: 0.9125\n",
            "WARNING:tensorflow:Can save best model only with val_loss available, skipping.\n",
            "Training completed in time:  0:15:37.377737\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 295
        },
        "id": "lJtr_Cn4xGcw",
        "outputId": "7fee8504-a613-4e92-fbc7-d79e8812dbeb"
      },
      "source": [
        "plt.plot(model_history.history['accuracy'])\r\n",
        "plt.plot(model_history.history['val_accuracy'])\r\n",
        "plt.title('CNN Model accuracy values')\r\n",
        "plt.ylabel('Accuracy')\r\n",
        "plt.xlabel('Epoch')\r\n",
        "plt.legend(['Train', 'Test'], loc='upper left')\r\n",
        "plt.show()\r\n"
      ],
      "execution_count": 33,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYgAAAEWCAYAAAB8LwAVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3dd3xV9f3H8deHAAk7TBlhbweCRBC3VRHtz2pbreCoVquidddtq9bWUUfdVVFxoaDiwoFb6h4ge08hzBBImAkZn98f96CXeAM3mJuT3Lyfj8d9cMb3nPM5J9zzud/v9wxzd0REREqrFXYAIiJSNSlBiIhITEoQIiISkxKEiIjEpAQhIiIxKUGIiEhMShBSI5hZJzNzM6sdR9mzzOzzyogrGen4JQ8lCNkpMzvVzCaa2SYzW2lm483s4GDezcFJ9w9R5WsH0zoF408H4wOiynQzszJvwDGzJWa2zcxalJo+OXrdIpJYShBSJjO7ArgPuA3YA+gA/Bc4IarYOuAfZpayk1WtA/5Vzs0vBoZFxbIPUL+c60g68dSARCqKEoTEZGZNgFuAv7j7q+6+2d0L3f1Nd78qqui7wDbg9J2s7hmgj5kdVo4QngP+GDV+JvBs6RjN7FkzyzazH8zsb2ZWK5iXYmZ3m9laM1sE/DrGsk8GtaLlZvavXSS56GVfNrNVZpZnZp+a2V5R8+qZ2T1BPHlm9rmZ1QvmHWxmX5pZrpktM7OzgukTzOzPUevYoYkmqDX9xczmA/ODafcH69hgZpPM7JCo8ilmdr2ZLTSzjcH89mb2sJndU2pfxpnZ5TH28REzu7vUtDeCHw2Y2bVR659lZr8t41j9rGkvxv6ebWazzWy9mb1nZh2D6WZm95rZmmA/p5vZ3jv/60hFUoKQsgwC0oDXdlHOgb8DN5lZnTLKbCFSC7m1HNv/GmhsZr2DE/dQYFSpMg8CTYAuwGFEEsqfgnnnAv8H9AMygZNKLfs0UAR0C8oMBv5MfMYD3YFWwPfA81Hz7gb6AwcCzYCrgZLgpDc+iLkl0BeYEuf2AE4EBgJ7BuPfBetoBrwAvGxmacG8K4jUvo4DGgNnE/kbPAMMi0qiLYCjguVLGw2cYmYWlG1K5BiNCeYvBA4hcvz/AYwyszbl2B+C9Z4AXA/8jshx+SzYNsH2DgV6BNv5A5BT3m3I7lOCkLI0B9a6e9GuCrr7OCCbnZ9gHwM6mNmx5Yhhey3iaGA2sHz7jKikcZ27b3T3JcA9wBlBkT8A97n7MndfB9weteweRE6elwU1ozXAvcH6dsndRwbbLABuBvYNaiS1iJyML3X35e5e7O5fBuVOBT5099FBTSzH3cuTIG5393XuvjWIYVSwjiJ3vwdIBXoGZf8M/M3d53rE1KDst0AecGRQbigwwd1Xx9jeZ0SS//aayUnAV+6+Itj+y+6+wt1L3P1FIjWbATHWsyvDg32bHfxfuw3oGyTUQqAR0AuwoMzK3diG7CYlCClLDtCiHG3efwNuIFLr+JngJPnP4BOv54icWM+iVPMS0AKoA/wQNe0HoF0w3BZYVmredh2DZVcGzT25RBJYq10FFDTf3BE0r2wAlkTF04LI/i+MsWj7MqbHK3pfMLMrg2aZvCD+JsH2d7WtZ/ipOfB0Isf4ZzzyFM8x/NQPdCpRNSUz+6OZTYk6fntHbb88OgL3R61nHWBAO3f/GHgIeBhYY2YjzKzxbmxDdpMShJTlK6CASNPGLrn7B8AC4MKdFHsKSCfSnBDPOn8g0ll9HPBqqdlrifzC7Bg1rQM/1TJWEjlRRs/bbhmRfWvh7unBp7G778WunUqkk/4oIiflTsF0C2LKB7rGWG5ZGdMBNrNjB3zrGGV+vOor6G+4mkgtqam7pxOpGVgc2xoFnGBm+wK9gdfLKAeRpp6Tgl/zA4FXgu13BB4HLgKaB9ufEbX90vvGTvZvGXB+1N8h3d3rufuXAO7+gLv3J9K01gOI7v+SBFOCkJjcPQ+4EXjYzE40s/pmVsfMjjWzO8tY7AYiJ66y1lkE3ARcU45QzgF+5e6boye6ezHwEnCrmTUKTlpX8FM/xUvAJWaWEbSfXxu17ErgfeAeM2tsZrXMrKvF14neiEhyySFy0rstar0lwEjgP2bWNqhtDDKzVCK/vo8ysz9Y5FLg5mbWN1h0CvC74Bh3C/Z5VzEUEWnWq21mNxLpa9juCeCfZtY96OjtY2bNgxiziPRfPAe8sr3JKhZ3n0wk6T0BvOfuucGsBkQSVjaAmf2JSA0i1jqyiSTt04PjcTY7Jq9Hgess6OgPmupODob3N7OBQd/WZiLJt2QXx0YqkBKElClo276CSPNRNpFfexdRxq9Od/8C+HYXqx1N5Nd9vDEsdPeJZcy+mMiJYxHwOZHO1pHBvMeB94CpRDqSS9dA/gjUBWYB64GxQDydrM8Saa5aHiz7dan5VwLTiZyE1wH/Bmq5+1IiNaG/BtOnAPsGy9xL5Eqw1USagJ5n594jcvXYvCCWfHZsgvoPkQT5PrABeBKoFzX/GWAfymheKuUFSnVku/ssIv09XwUx7wN8sZN1nEvkl38OsBfwZdS6XiNyjMYETXYzgO39VI2J/B3XB/uZA9wVR8xSQUwvDBKpWczsUCI1rY6uE4DshGoQIjVI0FxzKfCEkoPsihKESA1hZr2BXCJNafeFHI5UA2piEhGRmFSDEBGRmJLmwV8tWrTwTp06hR2GiEi1MmnSpLXu3jLWvKRJEJ06dWLixLKuhhQRkVjM7Iey5qmJSUREYlKCEBGRmJQgREQkpqTpg4ilsLCQrKws8vPzww4l4dLS0sjIyKBOnbJeySAiUj5JnSCysrJo1KgRnTp1InjvSVJyd3JycsjKyqJz585hhyMiSSKpm5jy8/Np3rx5UicHADOjefPmNaKmJCKVJ6kTBJD0yWG7mrKfIlJ5kj5BiIgkK3dn/PSVjPl2aULWn9R9EGHLycnhyCMjr/9dtWoVKSkptGwZuWHx22+/pW7dumUuO3HiRJ599lkeeOCBSolVRKqXbxev4/bxs5m8NJd+HdI5Zf/2Fd6SoASRQM2bN2fKlMh76W+++WYaNmzIlVde+eP8oqIiateO/SfIzMwkMzOzUuIUkepj/uqN/PvdOXw4ew17NE7lzt/34ff9MxLSzKwEUcnOOuss0tLSmDx5MgcddBBDhw7l0ksvJT8/n3r16vHUU0/Rs2dPJkyYwN13381bb73FzTffzNKlS1m0aBFLly7lsssu45JLLgl7V0SkEq3ekM+9H8zjpYnLaFC3Nlcd05OzD+pMvbopCdtmjUkQ/3hzJrNWbKjQde7ZtjE3HR/Pe+53lJWVxZdffklKSgobNmzgs88+o3bt2nz44Ydcf/31vPLKKz9bZs6cOXzyySds3LiRnj17csEFF+ieB5EaYEN+IY/9byFPfr6Y4hLnrAM7c9GvutGsQdlN1BWlxiSIquTkk08mJSWS9fPy8jjzzDOZP38+ZkZhYWHMZX7961+TmppKamoqrVq1YvXq1WRkZFRm2CJSibYVlTDq6x948OP5rN9SyG/2bcuVg3vSoXn9SouhxiSI3fmlnygNGjT4cfjvf/87RxxxBK+99hpLlizh8MMPj7lMamrqj8MpKSkUFRUlOkwRCUFJifPW9JXc9d4clq3bykHdmnPtkN7sk9Gk0mNJ6GWuZjbEzOaa2QIzuzbG/I5m9pGZTTOzCWaWETXvTDObH3zOTGScYcrLy6Ndu3YAPP300+EGIyKh+nLBWk54+AsuGT2ZBnVr88zZAxh1zsBQkgMkMEGYWQrwMHAssCcwzMz2LFXsbuBZd+8D3ALcHizbDLgJGAgMAG4ys6aJijVMV199Nddddx39+vVTrUCkhpq9cgNnPfUtpz7xDTmbCrjn5H15+5JDOKxHy1Bvgk3YO6nNbBBws7sfE4xfB+Dut0eVmQkMcfdlFjkKee7e2MyGAYe7+/lBuceACe4+uqztZWZmeukXBs2ePZvevXtX9K5VWTVtf0Wqu+W5W/nP+/N4dXIWjVJrc9GvuvHHQZ1Iq5O4K5NKM7NJ7h7zmvpE9kG0A5ZFjWcRqRFEmwr8Drgf+C3QyMyal7Fsu9IbMLPzgPMAOnToUGGBi4gkUt6WQv47YQFPfbkEgHMP6cKFh3clvX7ir0wqj7A7qa8EHjKzs4BPgeVAcbwLu/sIYAREahCJCFBEpKLkFxbz7FdLePiThWzIL+S3/drx18E9aZdeL+zQYkpkglgOtI8azwim/cjdVxCpQWBmDYHfu3uumS0HDi+17IQExioikjAlJc7rU5Zzz/vzWJ67lcN6tOSaIb3Ys23jsEPbqUQmiO+A7mbWmUhiGAqcGl3AzFoA69y9BLgOGBnMeg+4LapjenAwX0SkWvl0Xja3j5/D7JUb2LtdY+48qQ8HdWsRdlhxSViCcPciM7uIyMk+BRjp7jPN7BZgoruPI1JLuN3MnEgT01+CZdeZ2T+JJBmAW9x9XaJiFRGpaDOW53HH+Dl8vmAt7ZvV4/6hfTm+T1tq1ao+j+ZPaB+Eu78DvFNq2o1Rw2OBsWUsO5KfahQiItXCsnVbuPv9ubwxZQVN69fhxv/bk9MO6EBq7cq7MqmihN1JndR+yeO+ASZMmEDdunU58MADEx6riPwy6zdv48GPFzDq6x8wgwsP78rww7vSOK36PjNNCSKBdvW4712ZMGECDRs2VIIQqcLyC4sZ+cViHvlkIZu3FXFy//ZcfnQPWjdJCzu0X0xvlKtkkyZN4rDDDqN///4cc8wxrFy5EoAHHniAPffckz59+jB06FCWLFnCo48+yr333kvfvn357LPPQo5cRKIVlzgvfbeMw++awJ3vzmVA52a8e9mh/PukPkmRHKAm1SDGXwurplfsOlvvA8feEXdxd+fiiy/mjTfeoGXLlrz44ovccMMNjBw5kjvuuIPFixeTmppKbm4u6enpDB8+vNy1DhFJLHfnk7lruGP8HOat3sS+7dO5f2hfBnZpHnZoFa7mJIgqoKCggBkzZnD00UcDUFxcTJs2bQDo06cPp512GieeeCInnnhimGGKSBmmLMvl9ndm883idXRqXp//nrYfx+7dOtTnJSVSzUkQ5filnyjuzl577cVXX331s3lvv/02n376KW+++Sa33nor06dXcG1HRHbbkrWbueu9ubw9fSXNG9TllhP2YtiADtRJSe5W+pqTIKqA1NRUsrOz+eqrrxg0aBCFhYXMmzeP3r17s2zZMo444ggOPvhgxowZw6ZNm2jUqBEbNlTsW/BEJH5rNxXwwEfzeeGbpdRJqcUlR3bnvEO70DC1Zpw6a8ZeVhG1atVi7NixXHLJJeTl5VFUVMRll11Gjx49OP3008nLy8PdueSSS0hPT+f444/npJNO4o033uDBBx/kkEMOCXsXRGqELduKeOKzxTz2v4XkF5Vwyv7tuezI7rRqnBydz/FK2OO+K5se913z9lekohUVl/DixGXc9+F8sjcWcMxee3D1kF50bdkw7NASJqzHfYuIVAvuzvuzVnPnu3NYmL2ZzI5NefT0/ejfsVnYoYVKCUJEarRJP6zjtnfmMOmH9XRp2YARZ/Tn6D33SNork8oj6ROEu9eIP3SyNBWKVJYFazZx57tzeH/Walo2SuW23+7DHzIzqJ3kVyaVR1IniLS0NHJycmjevHlSJwl3Jycnh7S0mtWBJrI71mzI576P5vPid8tIq12Lvx7dg3MO6Uz9ukl9OtwtSX1EMjIyyMrKIjs7O+xQEi4tLY2MjIywwxCpsjYVFDHifwt5/LPFFBaXcPrADlx8ZHdaNEwNO7QqK6kTRJ06dejcuXPYYYhIiAqLSxj97VLu/3A+OZu38es+bbhqcE86tWgQdmhVXlInCBGpudydd6av4q735rAkZwsDOzfjyeN607d9etihVRtKECKSdL5elMPt4+cwdVkuPfZoyMizMjmiZ6uk7otMBCUIEUka81Zv5N/j5/DRnDW0bpzGnSf14ff7ZZBSjV7zWZUoQYhItbcybyv3fjCPsZOyaJBam6uH9OTsgzqTVqf6veazKlGCEJFqa0N+IY9MWMjIzxfjDn86qDMXHdGNpg12/jpfiY8ShIhUOwVFxYz6eikPfTyf9VsKOaFvW64c3JP2zeqHHVpSUYIQkWqjpMR5c9oK7npvLlnrt3JwtxZce2wv9m7XJOzQkpIShIhUC5/PX8sd785mxvIN9G7TmGfP3odDe7QMO6ykpgQhIlXarBUbuOPdOXw6L5t26fW495R9OWHfdtTSlUkJpwQhIlVS1vot/Of9ebw2ZTmN0+pww3G9OWNQR12ZVImUIESkynB3Vm8oYOQXi3n6yyUAnHdoFy48rBtN6tcJN7gaSAlCRCpNUXEJqzbks3z9Vpbnbv3p36jhgqISzOB3/TK4YnAP2qXXCzvsGksJQkQqzNZtxaVO+FtYkftTQli1IZ/ikh3fXdKiYV3apdejV5tGHNm7Fe3S6zGoawt6tm4U0l7IdglNEGY2BLgfSAGecPc7Ss3vADwDpAdlrnX3d8ysEzAbmBsU/drdhycyVhHZOXdnw9YisnK3lFkDyNm8bYdlUmoZrRun0S69HgM6N6Ndej3aNa23w7/qU6i6EpYgzCwFeBg4GsgCvjOzce4+K6rY34CX3P0RM9sTeAfoFMxb6O59ExWfiOyopMTJ3lRA1vodawDbk8CK3Hw2FRTtsExq7Vo/nuj3ats46sRfn3ZN67FHo1S9oa0aS2QNYgCwwN0XAZjZGOAEIDpBONA4GG4CrEhgPCI12raiElbl5ZdZA1iZm8+24pIdlmlSrw7t0uvRsXkDDuza4mc1gOYN6uoJqUkskQmiHbAsajwLGFiqzM3A+2Z2MdAAOCpqXmczmwxsAP7m7p+V3oCZnQecB9ChQ4eKi1ykGtpcUPTjST8r+uS/fgvLc7eyZmMBpV9d3qpRKu2a1mOfdk0YsndrMqJqAG3T02iUpiuHarKwO6mHAU+7+z1mNgh4zsz2BlYCHdw9x8z6A6+b2V7uviF6YXcfAYwAyMzM9NIrF0kW7s66zdvKvPJnee5WcrcU7rBMnRSjTZPIr/1Durfc8dd/ej3apKeRWlvt/1K2RCaI5UD7qPGMYFq0c4AhAO7+lZmlAS3cfQ1QEEyfZGYLgR7AxATGKxKa4hJn9Yb8HU76WVE1gBW5+WwtLN5hmfp1U3486fdtn/7jyT8jqAG0bJSq9yDIL5LIBPEd0N3MOhNJDEOBU0uVWQocCTxtZr2BNCDbzFoC69y92My6AN2BRQmMVSSh8guLWVH6V3/Ur/9VefkUlbr8s1mDyOWf3Vs14vCerXaoAWQ0rUeTenXU/i8JlbAE4e5FZnYR8B6RS1hHuvtMM7sFmOju44C/Ao+b2eVEOqzPcnc3s0OBW8ysECgBhrv7ukTFKpIoU5blcsNr05m5YofWUWoZkcs/m9ajf8emPzv5t02vR/26YbcAS01nXrrXqprKzMz0iRPVAiVVQ35hMfd9OJ8Rny5kj8ZpDN2/Q6TpJ0gCrZukUUeXf0oVYGaT3D0z1jz9RBGpYJOXrufKl6eyMHszQ/dvz/W/7k1jXQ0k1ZAShEgFyS8s5t4P5vH4Z4to3TiNZ84ewGF6X4FUY0oQIhXg+6XruSqoNQwb0J7rjlOtQao/JQiRXyC/sJj/fDCPJ4Jaw7NnD9BbziRpKEGI7KZJP6zjqpensWjtZk4d2IHrju2lO48lqShBiJTT1m3F3PP+XJ78YjFtm9Rj1DkDObh7i7DDEqlwShAi5TBxyTquGjuNxWs3c9rADlx3XG8apuprJMlJ/7NF4rB1WzF3vz+XkUGt4fk/D+Sgbqo1SHJTghDZhW8Xr+PqsVNZkrOFMw7oyDXH9lKtQWoE/S8XKcOWbUXc9d5cnv5yCRlN6/HCuQM5sKtqDVJzKEGIxPDNohyufmUaP+Rs4cxBHbl6SC8aqNYgNYz+x4tE2bKtiDvfjdQaOjSrz+hzD2BQ1+ZhhyUSCiUIkcDXi3K4euw0lq7bwlkHduLqIT31RFWp0fS/X2q8zQVF/PvdOTz71Q90bF6fMecdwAFdVGsQUYKQGu3LhWu55pVpZK3fyp8O6sRVx6jWILKdvglSI20uKOKO8XN47usf6NS8Pi+eN4gBnZuFHZZIlaIEITXOlwvWcvUr01ieu5WzD+rMVcf0pF7dlLDDEqlylCCkxthUUMQd42cz6uuldG7RgJfPH0RmJ9UaRMqiBCE1whcL1nL12GmsyNvKnw/uzF8Hq9YgsitKEJLUNuYXcvv4ObzwzVK6tGjA2OGD6N9RtQaReChBSNL6bH42174ynRV5Wznv0C5ccXQP0uqo1iASLyUISTob8wu57Z3ZjP52GV1aNmDs8APp37Fp2GGJVDtKEJJUPp2XzbWvTGPVhnzOP7QLl6vWILLblCAkKWzIL+TWt2bz4sRldG3ZgFcuOJB+HVRrEPkllCCk2pswdw3XvTqd1RvyGX5YVy47qrtqDSIVQAlCqq28rYXc+vYsXpqYRbdWDXn1woPo2z497LBEkoYShFRLn8yJ1BrWbMznwsO7csmRqjWIVLRdJggzOx54291LKiEekZ3K21rIP9+axdhJWfTYoyGPnXEQ+6rWIJIQteIocwow38zuNLNe5Vm5mQ0xs7lmtsDMro0xv4OZfWJmk81smpkdFzXvumC5uWZ2THm2K8np4zmrGXzv/3ht8nL+ckRX3rz4YCUHkQTaZQ3C3U83s8bAMOBpM3PgKWC0u28sazkzSwEeBo4GsoDvzGycu8+KKvY34CV3f8TM9gTeAToFw0OBvYC2wIdm1sPdi3dvN6U6y9tSyC1vzeKV77PouUcjHv9jJn0ylBhEEi2eGgTuvgEYC4wB2gC/Bb43s4t3stgAYIG7L3L3bcGyJ5ReNdA4GG4CrAiGTwDGuHuBuy8GFgTrkxrmo9mrOfre//H6lOVc/KtujLv4ICUHkUoSTx/Eb4A/Ad2AZ4EB7r7GzOoDs4AHy1i0HbAsajwLGFiqzM3A+0GiaQAcFbXs16WWbRcjtvOA8wA6dOiwq12RaiR3yzZueXMWr05eTq/WjXjyzP3ZJ6NJ2GGJ1CjxXMX0e+Bed/80eqK7bzGzc37h9ocBT7v7PWY2CHjOzPaOd2F3HwGMAMjMzPRfGItUER/MWs31r01n/eZtXHJkdy46oht1a8dV2RWRChRPgrgZWLl9xMzqAXu4+xJ3/2gnyy0H2keNZwTTop0DDAFw96/MLA1oEeeykmRyt2zj5nEzeX3KCnq1bsRTZ+3P3u1UaxAJSzw/y14Goi9xLQ6m7cp3QHcz62xmdYl0Oo8rVWYpcCSAmfUG0oDsoNxQM0s1s85Ad+DbOLYp1dT7M1dx1H8+5a1pK7n0yO6Mu+hgJQeRkMVTg6gddDID4O7bghP+Trl7kZldBLwHpAAj3X2mmd0CTHT3ccBfgcfN7HIiHdZnubsDM83sJSJ9HEXAX3QFU3Jav3kbN785kzemrKB3m8Y8c/b+7NVWiUGkKognQWSb2W+CEzpmdgKwNp6Vu/s7RC5djZ52Y9TwLOCgMpa9Fbg1nu1I9fTujFX87fUZ5G7ZxuVH9eCCw7uqr0GkCoknQQwHnjezhwAjcmXSHxMalSS1dZu3cdO4mbw5dQV7tW3Ms2cPYM+2jXe9oIhUqnhulFsIHGBmDYPxTQmPSpLWuzNW8rfXZ5C3tZArjo7UGuqkqNYgUhXF9bA+M/s1kbua08wMAHe/JYFxSZLJ2VTATeNm8ta0lezdrjHPnTOQ3m1UaxCpyuK5Ue5RoD5wBPAEcBK6okjK4Z3pK/n76zPYkF/IlYN7cP5hqjWIVAfx1CAOdPc+ZjbN3f9hZvcA4xMdmFR/azcVcNMbM3l7+kr2adeEF04+gJ6tG4UdlojEKZ4EkR/8u8XM2gI5RJ7HJFKmt6at4MY3ZrIpv4irjunJ+Yd2obZqDSLVSjwJ4k0zSwfuAr4ncr/C4wmNSqqttZsKuPGNGbwzfRV9Mppw10n7qtYgUk3tNEGYWS3gI3fPBV4xs7eANHfPq5TopNpwd96atpIb35jB5oJirh7Sk/MOUa1BpDrbaYJw9xIzexjoF4wXAAWVEZhUH9kbC/j76zN4d+Yq9m2fzt0n9aH7Hqo1iFR38TQxfWRmvwdeDR6DIQJEag3jpq7gpnEz2bKtmGuP7cWfD+6sWoNIkognQZwPXAEUmVk+kbup3d11EXsNtmZjPn97bQbvz1pN3/bp3H1yH7q1Uq1BJJnEcye1vvWygzenruDvb8xgy7Zirju2F38+pAsptSzssESkgsVzo9yhsaaXfoGQ1AwfzlrNxaMn069DOnedtC/dWjUMOyQRSZB4mpiuihpOI/Ju6EnArxISkVRZi9du5vIXp7BPuyaMPvcA0uqkhB2SiCRQPE1Mx0ePm1l74L6ERSRV0uaCIoY/N4naKcYjp++n5CBSA8T1sL5SsoDeFR2IVF3uzjWvTGP+mo08e/ZAMprWDzskEakE8fRBPEjk7mmIvKK0L5E7qqWGePLzxbw1bSXXDOnFwd1bhB2OiFSSeGoQE6OGi4DR7v5FguKRKubLhWu5ffwcjt27NcMP6xJ2OCJSieJJEGOB/O3vhDazFDOr7+5bEhuahG1F7lYufmEynZrX566T92X7u0BEpGaI55bXj4B6UeP1gA8TE45UFQVFxVzw/PcUFJXw2BmZNEzdne4qEanO4kkQadGvGQ2G1UuZ5G4eN4upy3K5+2Td6yBSU8WTIDab2X7bR8ysP7A1cSFJ2MZ8u5TR3y7lwsO7MmTv1mGHIyIhiafd4DLgZTNbQeQ5TK2BUxIalYRm6rJcbnxjJod0b8FfB/cMOxwRCVE8N8p9Z2a9gO1ni7nuXpjYsCQMOZsKuGDUJFo2SuWBof30fCWRGm6XTUxm9heggbvPcPcZQEMzuzDxoUllKiou4eLRk8nZvI3HzuhP0wZ1ww5JREIWTx/EucEb5QBw9/XAuYkLScJw13tz+XJhDrf+dh/2btck7HBEpAqIJ0GkWNQF8GaWAujnZRJ5e9pKHvt0EWcc0JGT+meEHY6IVBHxdFK/C7xoZo8F4+cD4xMXklSmeXZrOQ8AABIZSURBVKs3ctXYqezXIZ2//9+eYYcjIlVIPDWIa4CPgeHBZzo73jhXJjMbYmZzzWyBmV0bY/69ZjYl+Mwzs9yoecVR88bFtztSHhvyCxn+3CTq163NI6f3p25tvSpURH4Sz1VMJWb2DdAV+APQAnhlV8sFTVEPA0cTeQLsd2Y2zt1nRa378qjyFwP9olax1d37xrsjUj4lJc5fX5rK0nVbeOHcA9ijcVrYIYlIFVNmgjCzHsCw4LMWeBHA3Y+Ic90DgAXuvihY3xjgBGBWGeWHATfFuW75hf47YQEfzFrNTcfvyYDOzcIOR0SqoJ21Kcwh8ta4/3P3g939QaC4HOtuByyLGs8Kpv2MmXUEOhNpytouzcwmmtnXZnZiGcudF5SZmJ2dXY7QarYJc9dwzwfzOLFvW846sFPY4YhIFbWzBPE7YCXwiZk9bmZHErmTOhGGAmO3PzE20NHdM4FTgfvMrGvphdx9hLtnuntmy5YtExRaclmas4VLx0yhV+vG3P67PnpCq4iUqcwE4e6vu/tQoBfwCZFHbrQys0fMbHAc614OtI8azwimxTIUGF1q+8uDfxcBE9ixf0J2w9ZtxQwfNQl359HT96NeXb02VETKtsvLVtx9s7u/ELybOgOYTOTKpl35DuhuZp3NrC6RJPCzq5GCx3g0Bb6KmtbUzFKD4RbAQZTddyFxcHdueG06s1dt4P5h/ejYvEHYIYlIFVeu6xrdfX3QrHNkHGWLgIuA94DZwEvuPtPMbjGz30QVHQqMcXePmtYbmGhmU4nUXu6IvvpJyu/Zr37g1cnLufyoHhzRs1XY4YhINWA7nperr8zMTJ84ceKuC9ZA3y1Zx7ARX3N4z5aMOCOTWnoIn4gEzGxS0N/7M7ozKsmt2ZDPhc9/T/tm9fnPKX2VHEQkbnqPZBLbVlTChc9/z6b8IkadM5DGaXXCDklEqhEliCR269uzmPjDeh4c1o+erRuFHY6IVDNqYkpSr0zK4pmvfuDcQzpz/L5tww5HRKohJYgkNGN5Hte/Np0DujTjmiG9wg5HRKopJYgkk7tlG8NHTaJZg7o8dOp+1E7Rn1hEdo/6IJJIcYlzyZgprNlQwEvDB9GiYWrYIYlINaYEkUTu/WAen87L5vbf7UPf9ulhhyMi1ZzaH5LE+zNX8dAnCxi6f3uGDegQdjgikgSUIJLAwuxNXPHSVPbNaMLNv9kr7HBEJEkoQVRzmwqKGP7cJOrWrsUjp/cnrY6e0CoiFUMJohpzd64eO5WF2Zt4aFg/2qbH9apwEZG4KEFUYyM+XcQ701dx7bG9OLBbi7DDEZEkowRRTX2xYC3/fncOv96nDece0iXscEQkCSlBVEPLc7dy8ejJdG3ZkDtP0mtDRSQxlCCqmfzCYi4YNYnCohIeO6M/DVJ1K4uIJIbOLtWIu3PjGzOYlpXHiDP606Vlw7BDEpEkphpENTL622W8NDGLi3/VjcF7tQ47HBFJckoQ1cTkpeu5adwMDuvRksuO6hF2OCJSAyhBVAPZGwu4YNT3tG6Sxv1D+5Ki14aKSCVQH0QVV1RcwkUvfM/6Ldt49cIDSa9fN+yQRKSGUIKo4u4YP4dvFq/j3lP2Za+2TcIOR0RqEDUxVWHjpq7gic8Xc9aBnfhtv4ywwxGRGkYJooqau2oj14ydxv6dmnL9cb3DDkdEaiAliCoob2sh5z83kYZptXn41P2oW1t/JhGpfDrzVDElJc4VL04ha/1WHjltP1o1Tgs7JBGpoZQgqpgHP17AR3PWcOPxe5LZqVnY4YhIDZbQBGFmQ8xsrpktMLNrY8y/18ymBJ95ZpYbNe9MM5sffM5MZJxVxSdz1nDfR/P43X7tOOOAjmGHIyI1XMIuczWzFOBh4GggC/jOzMa5+6ztZdz98qjyFwP9guFmwE1AJuDApGDZ9YmKN2w/5Gzm0jGT6d26Mbf9dh89oVVEQpfIGsQAYIG7L3L3bcAY4ISdlB8GjA6GjwE+cPd1QVL4ABiSwFhDtWVbEec/Nwkz47Ez9NpQEakaEpkg2gHLosazgmk/Y2Ydgc7Ax+Vdtrpzd657dTpzV2/kgWH9aN+sftghiYgAVaeTeigw1t2Ly7OQmZ1nZhPNbGJ2dnaCQkusp75YwhtTVnDl4J4c1qNl2OGIiPwokQliOdA+ajwjmBbLUH5qXop7WXcf4e6Z7p7ZsmX1O7l+syiH296ZzeA99+CCw7qGHY6IyA4SmSC+A7qbWWczq0skCYwrXcjMegFNga+iJr8HDDazpmbWFBgcTEsaq/Ly+csLk+nQrD73/GFfaukJrSJSxSTsKiZ3LzKzi4ic2FOAke4+08xuASa6+/ZkMRQY4+4etew6M/snkSQDcIu7r0tUrJWtoKiYC56fxJZtRYw+dyCN0uqEHZKIyM9Y1Hm5WsvMzPSJEyeGHUZc/vb6dEZ9vZT/nrYfx+3TJuxwRKQGM7NJ7p4Za15V6aSuMV6euIxRXy/l/MO6KDmISJWmBFGJZizP44bXZ3Bg1+ZcNbhn2OGIiOyUEkQlWbd5G+c/N4kWDery4LB+1E7RoReRqk1vlKsExSXOJaMnk72pgLHDB9G8YWrYIYmI7JJ+xlaCu9+fy+cL1vKvE/amT0Z62OGIiMRFCSLB3p2xkkcmLOTUgR34w/7td72AiEgVoQSRQAvWbOKvL02lb/t0bjp+z7DDEREpFyWIBNmYH3ltaFqdFB45fT9Sa+sJrSJSvaiTOgHcnStfnsqSnC2MOmcgbZrUCzskEZFyUw0iAR7530Lem7ma647txaCuzcMOR0RktyhBVLDP5mdz93tzOX7ftpxzcOewwxER2W1KEBVo2botXDJ6Mt1bNeLfv9drQ0WkelOCqCD5hZEntBaVOI+e0Z/6ddW9IyLVm85iFcDdueG1GcxYvoEnz8ykc4sGYYckIvKLqQZRAUZ9s5RXvs/i0iO7c2TvPcIOR0SkQihB/EKTfljPLW/O5IieLbn0yO5hhyMiUmGUIH6BNRvzufD5SbRpUo/7Tumn14aKSFJRH8RuKiwu4aLnJ5O3tZDXLhxAk/p6baiIJBcliN102zuz+XbJOu4f2pfebRqHHY6ISIVTE9NueGPKcp76YglnH9SZE/q2CzscEZGEUIIop9krN3DNK9MY0LkZ1x3XK+xwREQSRgmiHPK2FHL+c5NoUq8OD53ajzp6baiIJDH1QcSppMS59MXJrMzbypjzBtGqUVrYIYmIJJR+Asfpvo/mM2FuNjcdvxf9OzYNOxwRkYRTgojDR7NX88BH8zmpfwanDewQdjgiIpVCCWIXFq/dzGUvTmHvdo3514l76wmtIlJjKEHsxOaCIoY/N4natYxHT+9PWh29NlREag51UpfB3bnmlWnMX7ORZ84eQEbT+mGHJCJSqRJagzCzIWY218wWmNm1ZZT5g5nNMrOZZvZC1PRiM5sSfMYlMs5Ynvx8MW9NW8lVx/TikO4tK3vzIiKhS1gNwsxSgIeBo4Es4DszG+fus6LKdAeuAw5y9/Vm1ipqFVvdvW+i4tuZrxbmcPv4OQzZqzXDD+sSRggiIqFLZA1iALDA3Re5+zZgDHBCqTLnAg+7+3oAd1+TwHjisjJvKxe98D2dmtfnrpP7qFNaRGqsRCaIdsCyqPGsYFq0HkAPM/vCzL42syFR89LMbGIw/cRYGzCz84IyE7Ozs39xwAVFxQwf9T0FRSU8dkYmjdL0hFYRqbnC7qSuDXQHDgcygE/NbB93zwU6uvtyM+sCfGxm0919YfTC7j4CGAGQmZnpvzSYm8fNYuqyXB49vT/dWjX8pasTEanWElmDWA60jxrPCKZFywLGuXuhuy8G5hFJGLj78uDfRcAEoF8CY+XF75Yy+tulXHB4V4bs3TqRmxIRqRYSmSC+A7qbWWczqwsMBUpfjfQ6kdoDZtaCSJPTIjNramapUdMPAmaRIFOX5fL3N2ZycLcWXDm4Z6I2IyJSrSSsicndi8zsIuA9IAUY6e4zzewWYKK7jwvmDTazWUAxcJW755jZgcBjZlZCJIndEX31U0XK2VTABaMm0bJhKg8M60eKXhsqIgIkuA/C3d8B3ik17caoYQeuCD7RZb4E9klkbNvVMmPPto257KgeNGtQtzI2KSJSLYTdSR26pg3q8sSZ+4cdhohIlaNnMYmISExKECIiEpMShIiIxKQEISIiMSlBiIhITEoQIiISkxKEiIjEpAQhIiIxWeRm5urPzLKBH37BKloAaysonIqkuMpHcZWP4iqfZIyro7vHfG1m0iSIX8rMJrp7ZthxlKa4ykdxlY/iKp+aFpeamEREJCYlCBERiUkJ4icjwg6gDIqrfBRX+Siu8qlRcakPQkREYlINQkREYlKCEBGRmGpUgjCzIWY218wWmNm1MeanmtmLwfxvzKxTFYnrLDPLNrMpwefPlRTXSDNbY2YzyphvZvZAEPc0M9uvisR1uJnlRR2vG2OVS0Bc7c3sEzObZWYzzezSGGUq/ZjFGVelHzMzSzOzb81sahDXP2KUqfTvZJxxhfKdDLadYmaTzeytGPMq9ni5e434EHkv9kKgC1AXmArsWarMhcCjwfBQ4MUqEtdZwEMhHLNDgf2AGWXMPw4YDxhwAPBNFYnrcOCtEI5XG2C/YLgRMC/G37LSj1mccVX6MQuOQcNguA7wDXBAqTJhfCfjiSuU72Sw7SuAF2L9vSr6eNWkGsQAYIG7L3L3bcAY4IRSZU4AngmGxwJHmplVgbhC4e6fAut2UuQE4FmP+BpIN7M2VSCuULj7Snf/PhjeCMwG2pUqVunHLM64Kl1wDDYFo3WCT+mrZir9OxlnXKEwswzg18ATZRSp0ONVkxJEO2BZ1HgWP/+S/FjG3YuAPKB5FYgL4PdBk8RYM2uf4JjiFW/sYRgUNBGMN7O9KnvjQdW+H5Ffn9FCPWY7iQtCOGZBc8kUYA3wgbuXebwq8TsZT1wQznfyPuBqoKSM+RV6vGpSgqjO3gQ6uXsf4AN++oUgsX1P5Pky+wIPAq9X5sbNrCHwCnCZu2+ozG3vzC7iCuWYuXuxu/cFMoABZrZ3ZWx3V+KIq9K/k2b2f8Aad5+U6G1tV5MSxHIgOstnBNNiljGz2kATICfsuNw9x90LgtEngP4Jjile8RzTSufuG7Y3Ebj7O0AdM2tRGds2szpETsLPu/urMYqEcsx2FVeYxyzYZi7wCTCk1KwwvpO7jCuk7+RBwG/MbAmRpuhfmdmoUmUq9HjVpATxHdDdzDqbWV0iHTjjSpUZB5wZDJ8EfOxBb0+YcZVqo/4NkTbkqmAc8MfgypwDgDx3Xxl2UGbWenu7q5kNIPL/POEnlWCbTwKz3f0/ZRSr9GMWT1xhHDMza2lm6cFwPeBoYE6pYpX+nYwnrjC+k+5+nbtnuHsnIueJj9399FLFKvR41d7dBasbdy8ys4uA94hcOTTS3Wea2S3ARHcfR+RL9JyZLSDSCTq0isR1iZn9BigK4jor0XEBmNloIle3tDCzLOAmIh12uPujwDtErspZAGwB/lRF4joJuMDMioCtwNBKSPQQ+YV3BjA9aL8GuB7oEBVbGMcsnrjCOGZtgGfMLIVIQnrJ3d8K+zsZZ1yhfCdjSeTx0qM2REQkpprUxCQiIuWgBCEiIjEpQYiISExKECIiEpMShIiIxKQEIVIOZlYc9QTPKRbj6bu/YN2drIwn1IqEocbcByFSQbYGj2AQSXqqQYhUADNbYmZ3mtn04F0C3YLpnczs4+Chbh+ZWYdg+h5m9lrwcLypZnZgsKoUM3vcIu8heD+4k1ckFEoQIuVTr1QT0ylR8/LcfR/gISJP3YTIg++eCR7q9jzwQDD9AeB/wcPx9gNmBtO7Aw+7+15ALvD7BO+PSJl0J7VIOZjZJndvGGP6EuBX7r4oeDDeKndvbmZrgTbuXhhMX+nuLcwsG8iIeuDb9kdxf+Du3YPxa4A67v6vxO+ZyM+pBiFScbyM4fIoiBouRv2EEiIlCJGKc0rUv18Fw1/y0wPTTgM+C4Y/Ai6AH19O06SyghSJl36diJRPvagnogK86+7bL3VtambTiNQChgXTLgaeMrOrgGx+enrrpcAIMzuHSE3hAiD0R6WLRFMfhEgFCPogMt19bdixiFQUNTGJiEhMqkGIiEhMqkGIiEhMShAiIhKTEoSIiMSkBCEiIjEpQYiISEz/D5r8rQcw6II+AAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 282
        },
        "id": "JO8eEg9-DFAT",
        "outputId": "97ffc5dc-d1ef-4c16-9a06-c9e263ee963e"
      },
      "source": [
        "# loss\r\n",
        "plt.plot(model_history.history['loss'], label='train loss')\r\n",
        "plt.plot(model_history.history['val_loss'], label='val loss')\r\n",
        "plt.legend()\r\n",
        "plt.show()\r\n",
        "plt.savefig('LossVal_loss')"
      ],
      "execution_count": 34,
      "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"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 0 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "mnM735grDIjY"
      },
      "source": [
        "import tensorflow as tf"
      ],
      "execution_count": 35,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dDAGeJw1DJFu"
      },
      "source": [
        "from keras.models import load_model"
      ],
      "execution_count": 36,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_iNDMKh7D-pO"
      },
      "source": [
        "model.save('model3_vgg19.h5')"
      ],
      "execution_count": 37,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SqRjCL8J8sv6"
      },
      "source": [
        "from tensorflow.keras.models import load_model\r\n",
        "from tensorflow.keras.preprocessing import image"
      ],
      "execution_count": 38,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lEC5f6KaEBCs"
      },
      "source": [
        "model=load_model('model3_vgg19.h5')"
      ],
      "execution_count": 39,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "L-1TcZHMEPHM"
      },
      "source": [
        "img=image.load_img('/content/drive/MyDrive/kaggle2pneumonia/chest_xray/val/NORMAL/NORMAL2-IM-1427-0001.jpeg',target_size=(224,224))"
      ],
      "execution_count": 40,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JWaBXxq5EPgU"
      },
      "source": [
        "x=image.img_to_array(img)"
      ],
      "execution_count": 41,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "YU_VOdsMEiOX"
      },
      "source": [
        "x=x/255"
      ],
      "execution_count": 42,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "yOtGWWYKElaQ",
        "outputId": "369cf5df-ea4a-4b90-af13-9cc6f7004b76"
      },
      "source": [
        "import numpy as np\r\n",
        "x=np.expand_dims(x,axis=0)\r\n",
        "img_data=preprocess_input(x)\r\n",
        "img_data.shape"
      ],
      "execution_count": 43,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(1, 224, 224, 3)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 43
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7AOqdWUHEoQi"
      },
      "source": [
        "c=model.predict(img_data)"
      ],
      "execution_count": 44,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uit9e8OpEq9c"
      },
      "source": [
        "a=np.argmax(model.predict(img_data), axis=1)"
      ],
      "execution_count": 45,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wgGRep5NEvu6",
        "outputId": "c8065188-3f69-47c6-8c7d-1b81f92a69bb"
      },
      "source": [
        "a==1"
      ],
      "execution_count": 46,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([False])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 46
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 35
        },
        "id": "znzlnR78FIP0",
        "outputId": "1825aa5e-5257-44e2-ec87-ff931160b270"
      },
      "source": [
        "str(np.round(c[0][0],2))"
      ],
      "execution_count": 48,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "application/vnd.google.colaboratory.intrinsic+json": {
              "type": "string"
            },
            "text/plain": [
              "'1.0'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 48
        }
      ]
    }
  ]
}