pretrained_classifier.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BtzBPEgP0kST",
        "outputId": "6d7bedd9-631e-4438-e4b4-995525b62851"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Collecting preprocessing\n",
            "  Downloading preprocessing-0.1.13-py3-none-any.whl (349 kB)\n",
            "\u001b[K     |████████████████████████████████| 349 kB 8.2 MB/s \n",
            "\u001b[?25hCollecting sphinx-rtd-theme==0.2.4\n",
            "  Downloading sphinx_rtd_theme-0.2.4-py2.py3-none-any.whl (1.4 MB)\n",
            "\u001b[K     |████████████████████████████████| 1.4 MB 38.0 MB/s \n",
            "\u001b[?25hCollecting nltk==3.2.4\n",
            "  Downloading nltk-3.2.4.tar.gz (1.2 MB)\n",
            "\u001b[K     |████████████████████████████████| 1.2 MB 44.9 MB/s \n",
            "\u001b[?25hRequirement already satisfied: six in /usr/local/lib/python3.7/dist-packages (from nltk==3.2.4->preprocessing) (1.15.0)\n",
            "Building wheels for collected packages: nltk\n",
            "  Building wheel for nltk (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for nltk: filename=nltk-3.2.4-py3-none-any.whl size=1367724 sha256=cd58cd00679840222e48f13711d9e8d85fe7d0240cc800aa87b4b24392a58570\n",
            "  Stored in directory: /root/.cache/pip/wheels/90/5e/9e/4cb46185f2a16c60e6fc524372ba7fef89ce3347734c8798b6\n",
            "Successfully built nltk\n",
            "Installing collected packages: sphinx-rtd-theme, nltk, preprocessing\n",
            "  Attempting uninstall: nltk\n",
            "    Found existing installation: nltk 3.2.5\n",
            "    Uninstalling nltk-3.2.5:\n",
            "      Successfully uninstalled nltk-3.2.5\n",
            "Successfully installed nltk-3.2.4 preprocessing-0.1.13 sphinx-rtd-theme-0.2.4\n"
          ]
        }
      ],
      "source": [
        "#Import the needed modules\n",
        "!pip install preprocessing\n",
        "from collections import Counter\n",
        "import cv2\n",
        "import os\n",
        "import glob\n",
        "import skimage\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "import seaborn as sn\n",
        "import preprocessing\n",
        "from keras.preprocessing.image import ImageDataGenerator, array_to_img, img_to_array, load_img\n",
        "from tqdm import tqdm\n",
        "from io import BytesIO\n",
        "from PIL import Image\n",
        "from os import listdir\n",
        "import matplotlib.pyplot as plt\n",
        "from imageio import imread\n",
        "from skimage.transform import resize\n",
        "from collections import Counter\n",
        "from sklearn import metrics\n",
        "from sklearn.utils import shuffle\n",
        "from sklearn import metrics\n",
        "from sklearn.utils import compute_class_weight\n",
        "from sklearn.preprocessing import MinMaxScaler,LabelBinarizer\n",
        "from sklearn.model_selection import train_test_split\n",
        "import tensorflow as tf\n",
        "from tensorflow.keras.utils import to_categorical\n",
        "from tensorflow.keras.metrics import AUC\n",
        "from tensorflow.keras.preprocessing import image\n",
        "from tensorflow.keras.models import Model, Sequential, load_model\n",
        "from tensorflow.keras.applications import * \n",
        "from tensorflow.keras.layers import*\n",
        "from tensorflow.keras.callbacks import EarlyStopping, ModelCheckpoint\n",
        "from tensorflow.keras import regularizers\n",
        "from tensorflow.keras.optimizers import *\n",
        "from tensorflow.keras.callbacks import EarlyStopping, ModelCheckpoint\n",
        "from tensorflow.keras import layers\n",
        "from skimage import io, transform\n",
        "from sklearn.cluster import KMeans\n",
        "from keras.layers import Dropout\n",
        "%matplotlib inline"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "zpNiopa704gn",
        "outputId": "12d990b6-3da5-4a1d-8c27-d039ceaf4466"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Mounted at /content/gdrive\n"
          ]
        }
      ],
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/gdrive')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "id": "cOwIDBxW1RQd"
      },
      "outputs": [],
      "source": [
        "images=[]\n",
        "labels=[]\n",
        "feature_dictionary = {\n",
        "    'label': tf.io.FixedLenFeature([], tf.int64),\n",
        "    'label_normal': tf.io.FixedLenFeature([], tf.int64),\n",
        "    'image': tf.io.FixedLenFeature([], tf.string)\n",
        "    }"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "id": "OgEdCQIa1ZCX"
      },
      "outputs": [],
      "source": [
        "# Create the function to resize the dataset.\n",
        "def _parse_function(example, feature_dictionary=feature_dictionary):\n",
        "    parsed_example = tf.io.parse_example(example, feature_dictionary)\n",
        "    return parsed_example\n",
        "def read_data(filename):\n",
        "    full_dataset = tf.data.TFRecordDataset(filename,num_parallel_reads=tf.data.experimental.AUTOTUNE)\n",
        "    full_dataset = full_dataset.shuffle(buffer_size=31000)\n",
        "    full_dataset = full_dataset.cache()\n",
        "    print(\"Size of Training Dataset: \", len(list(full_dataset)))  \n",
        "    feature_dictionary = {\n",
        "    'label': tf.io.FixedLenFeature([], tf.int64),\n",
        "    'label_normal': tf.io.FixedLenFeature([], tf.int64),\n",
        "    'image': tf.io.FixedLenFeature([], tf.string)\n",
        "    }   \n",
        "    full_dataset = full_dataset.map(_parse_function, num_parallel_calls=tf.data.experimental.AUTOTUNE)\n",
        "    print(full_dataset)\n",
        "    for image_features in full_dataset:\n",
        "        image = image_features['image'].numpy()\n",
        "        image = tf.io.decode_raw(image_features['image'], tf.uint8)\n",
        "        image = tf.reshape(image, [299, 299])        \n",
        "        image=image.numpy()\n",
        "        image=cv2.resize(image,(100,100))\n",
        "        image=cv2.merge([image,image,image])        \n",
        "        images.append(image)\n",
        "        labels.append(image_features['label_normal'].numpy())"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "I5T3KA1A1d9-",
        "outputId": "70586eff-c759-426c-9066-415fe6288766"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Size of Training Dataset:  11177\n",
            "<ParallelMapDataset element_spec={'image': TensorSpec(shape=(), dtype=tf.string, name=None), 'label': TensorSpec(shape=(), dtype=tf.int64, name=None), 'label_normal': TensorSpec(shape=(), dtype=tf.int64, name=None)}>\n",
            "Size of Training Dataset:  11177\n",
            "<ParallelMapDataset element_spec={'image': TensorSpec(shape=(), dtype=tf.string, name=None), 'label': TensorSpec(shape=(), dtype=tf.int64, name=None), 'label_normal': TensorSpec(shape=(), dtype=tf.int64, name=None)}>\n",
            "Size of Training Dataset:  11177\n",
            "<ParallelMapDataset element_spec={'image': TensorSpec(shape=(), dtype=tf.string, name=None), 'label': TensorSpec(shape=(), dtype=tf.int64, name=None), 'label_normal': TensorSpec(shape=(), dtype=tf.int64, name=None)}>\n",
            "Size of Training Dataset:  11177\n",
            "<ParallelMapDataset element_spec={'image': TensorSpec(shape=(), dtype=tf.string, name=None), 'label': TensorSpec(shape=(), dtype=tf.int64, name=None), 'label_normal': TensorSpec(shape=(), dtype=tf.int64, name=None)}>\n",
            "Size of Training Dataset:  11177\n",
            "<ParallelMapDataset element_spec={'image': TensorSpec(shape=(), dtype=tf.string, name=None), 'label': TensorSpec(shape=(), dtype=tf.int64, name=None), 'label_normal': TensorSpec(shape=(), dtype=tf.int64, name=None)}>\n"
          ]
        }
      ],
      "source": [
        "#Import and read data\n",
        "filenames=['/content/gdrive/MyDrive/CSE676_PROJECT/training10_0.tfrecords',\n",
        "          '/content/gdrive/MyDrive/CSE676_PROJECT/training10_1.tfrecords',\n",
        "          '/content/gdrive/MyDrive/CSE676_PROJECT/training10_2.tfrecords',\n",
        "          '/content/gdrive/MyDrive/CSE676_PROJECT/training10_3.tfrecords',\n",
        "          '/content/gdrive/MyDrive/CSE676_PROJECT/training10_4.tfrecords']\n",
        "for file in filenames:\n",
        "    read_data(file)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "id": "PDEjtsoC2ULe"
      },
      "outputs": [],
      "source": [
        "#Normalize the data\n",
        "X=np.array(images)\n",
        "Y=np.array(labels)\n",
        "x_train, x_test, y_train, y_test = train_test_split(X, Y, test_size=0.3, random_state=42,shuffle=True)\n",
        "x_train_normalize = x_train.astype('float32') / 255.0\n",
        "x_test_normalize = x_test.astype('float32') / 255.0"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "#CNN model for one layer\n",
        "filename='cnn_onelayer.csv'\n",
        "history_logger=tf.keras.callbacks.CSVLogger(filename, separator=\",\", append=True)\n",
        "cnn1_model=tf.keras.Sequential([\n",
        "tf.keras.layers.Conv2D(64,(5,5),input_shape=(100,100,3),activation='silu'),\n",
        "tf.keras.layers.MaxPooling2D(2,2), # First Convolution and Pooling Layers\n",
        "# tf.keras.layers.Dropout(0.2)\n",
        "])\n",
        "cnn1_model.add(Flatten())\n",
        "cnn1_model.add(Dense(8,activation='silu',kernel_regularizer=regularizers.L1(0.0001)))\n",
        "cnn1_model.add(BatchNormalization())\n",
        "cnn1_model.add(Dense(1, activation='sigmoid',kernel_regularizer=regularizers.L1(0.0001)))\n",
        "cnn1_model.add(Dropout(0.3))\n",
        "opt = Adam(learning_rate=0.001)\n",
        "cnn1_model.compile(optimizer=opt,loss = 'binary_crossentropy' , metrics = ['accuracy'])\n",
        "history1 = cnn1_model.fit(x_train_normalize ,y_train,epochs=30,batch_size=32,validation_split=0.2,callbacks=[history_logger])\n",
        "loss_value , accuracy = cnn1_model.evaluate(x_test_normalize,y_test)\n",
        "print(loss_value,accuracy)\n",
        "plt.plot(history1.history['accuracy'])\n",
        "plt.plot(history1.history['val_accuracy'])\n",
        "plt.title('model accuracy')\n",
        "plt.ylabel('accuracy')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train','validation'], loc='upper left')\n",
        "plt.show()\n",
        "# summarize history for loss\n",
        "plt.plot(history1.history['loss'])\n",
        "plt.plot(history1.history['val_loss'])\n",
        "plt.title('model loss')\n",
        "plt.ylabel('loss')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'validation'], loc='upper right')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "CzNKT-tsGyPO",
        "outputId": "e82ea6f2-5086-4025-c785-6cdad1202f01"
      },
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/30\n",
            "978/978 [==============================] - 34s 22ms/step - loss: 1.7547 - accuracy: 0.7945 - val_loss: 2.2641 - val_accuracy: 0.8705\n",
            "Epoch 2/30\n",
            "978/978 [==============================] - 21s 21ms/step - loss: 2.2386 - accuracy: 0.8707 - val_loss: 1.4766 - val_accuracy: 0.8684\n",
            "Epoch 3/30\n",
            "978/978 [==============================] - 21s 21ms/step - loss: 1.8562 - accuracy: 0.8765 - val_loss: 1.2767 - val_accuracy: 0.8854\n",
            "Epoch 4/30\n",
            "978/978 [==============================] - 21s 21ms/step - loss: 1.6186 - accuracy: 0.8772 - val_loss: 1.0281 - val_accuracy: 0.8707\n",
            "Epoch 5/30\n",
            "978/978 [==============================] - 21s 21ms/step - loss: 1.4406 - accuracy: 0.8788 - val_loss: 0.8287 - val_accuracy: 0.8736\n",
            "Epoch 6/30\n",
            "978/978 [==============================] - 21s 21ms/step - loss: 1.2980 - accuracy: 0.8770 - val_loss: 0.7712 - val_accuracy: 0.8705\n",
            "Epoch 7/30\n",
            "978/978 [==============================] - 21s 21ms/step - loss: 1.2396 - accuracy: 0.8795 - val_loss: 0.7100 - val_accuracy: 0.8765\n",
            "Epoch 8/30\n",
            "978/978 [==============================] - 21s 21ms/step - loss: 1.1501 - accuracy: 0.8792 - val_loss: 1.5362 - val_accuracy: 0.8705\n",
            "Epoch 9/30\n",
            "978/978 [==============================] - 21s 21ms/step - loss: 1.0834 - accuracy: 0.8820 - val_loss: 0.5425 - val_accuracy: 0.8719\n",
            "Epoch 10/30\n",
            "978/978 [==============================] - 21s 21ms/step - loss: 1.0381 - accuracy: 0.8792 - val_loss: 0.6041 - val_accuracy: 0.8705\n",
            "Epoch 11/30\n",
            "978/978 [==============================] - 21s 21ms/step - loss: 0.9754 - accuracy: 0.8786 - val_loss: 0.4792 - val_accuracy: 0.8705\n",
            "Epoch 12/30\n",
            "978/978 [==============================] - 21s 21ms/step - loss: 0.9451 - accuracy: 0.8782 - val_loss: 0.4565 - val_accuracy: 0.8707\n",
            "Epoch 13/30\n",
            "978/978 [==============================] - 21s 21ms/step - loss: 0.9019 - accuracy: 0.8791 - val_loss: 0.4434 - val_accuracy: 0.8636\n",
            "Epoch 14/30\n",
            "978/978 [==============================] - 21s 21ms/step - loss: 0.8912 - accuracy: 0.8809 - val_loss: 0.3664 - val_accuracy: 0.8764\n",
            "Epoch 15/30\n",
            "978/978 [==============================] - 21s 21ms/step - loss: 0.8872 - accuracy: 0.8794 - val_loss: 0.3905 - val_accuracy: 0.8712\n",
            "Epoch 16/30\n",
            "978/978 [==============================] - 21s 21ms/step - loss: 0.8439 - accuracy: 0.8814 - val_loss: 0.3378 - val_accuracy: 0.8805\n",
            "Epoch 17/30\n",
            "978/978 [==============================] - 21s 21ms/step - loss: 0.8391 - accuracy: 0.8806 - val_loss: 0.4404 - val_accuracy: 0.8705\n",
            "Epoch 18/30\n",
            "978/978 [==============================] - 21s 21ms/step - loss: 0.8372 - accuracy: 0.8796 - val_loss: 0.4752 - val_accuracy: 0.8705\n",
            "Epoch 19/30\n",
            "978/978 [==============================] - 21s 21ms/step - loss: 0.8682 - accuracy: 0.8777 - val_loss: 0.3421 - val_accuracy: 0.8735\n",
            "Epoch 20/30\n",
            "978/978 [==============================] - 21s 21ms/step - loss: 0.8566 - accuracy: 0.8766 - val_loss: 0.3380 - val_accuracy: 0.8756\n",
            "Epoch 21/30\n",
            "978/978 [==============================] - 21s 21ms/step - loss: 0.8514 - accuracy: 0.8796 - val_loss: 0.3611 - val_accuracy: 0.8723\n",
            "Epoch 22/30\n",
            "978/978 [==============================] - 21s 21ms/step - loss: 0.8663 - accuracy: 0.8750 - val_loss: 0.3160 - val_accuracy: 0.8810\n",
            "Epoch 23/30\n",
            "978/978 [==============================] - 21s 21ms/step - loss: 0.8315 - accuracy: 0.8812 - val_loss: 0.3573 - val_accuracy: 0.8717\n",
            "Epoch 24/30\n",
            "978/978 [==============================] - 21s 21ms/step - loss: 0.8251 - accuracy: 0.8822 - val_loss: 0.5112 - val_accuracy: 0.8704\n",
            "Epoch 25/30\n",
            "978/978 [==============================] - 21s 21ms/step - loss: 0.8532 - accuracy: 0.8788 - val_loss: 0.5140 - val_accuracy: 0.8705\n",
            "Epoch 26/30\n",
            "978/978 [==============================] - 21s 21ms/step - loss: 0.8676 - accuracy: 0.8779 - val_loss: 0.3617 - val_accuracy: 0.8705\n",
            "Epoch 27/30\n",
            "978/978 [==============================] - 21s 21ms/step - loss: 0.8591 - accuracy: 0.8795 - val_loss: 0.3477 - val_accuracy: 0.8764\n",
            "Epoch 28/30\n",
            "978/978 [==============================] - 21s 21ms/step - loss: 0.8538 - accuracy: 0.8777 - val_loss: 0.3677 - val_accuracy: 0.8705\n",
            "Epoch 29/30\n",
            "978/978 [==============================] - 21s 22ms/step - loss: 0.9167 - accuracy: 0.8762 - val_loss: 0.5091 - val_accuracy: 0.8708\n",
            "Epoch 30/30\n",
            "978/978 [==============================] - 22s 22ms/step - loss: 0.8269 - accuracy: 0.8823 - val_loss: 0.3427 - val_accuracy: 0.8751\n",
            "524/524 [==============================] - 5s 9ms/step - loss: 0.3555 - accuracy: 0.8687\n",
            "0.3554789423942566 0.8687224388122559\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "# CNN model for two layers\n",
        "filename='cnn_twolayer.csv'\n",
        "history_logger=tf.keras.callbacks.CSVLogger(filename, separator=\",\", append=True)\n",
        "cnn2_model=tf.keras.Sequential([\n",
        "tf.keras.layers.Conv2D(64,(5,5),input_shape=(100,100,3),activation='silu'),\n",
        "tf.keras.layers.MaxPooling2D(2,2), # First Convolution and Pooling Layers\n",
        "tf.keras.layers.Conv2D(16,(2,2),activation='silu'),\n",
        "tf.keras.layers.MaxPooling2D(2,2), # Second Convolution and Pooling Layers\n",
        "# tf.keras.layers.Dropout(0.2)\n",
        "])\n",
        "cnn2_model.add(Flatten())\n",
        "cnn2_model.add(Dense(8, activation='silu', kernel_initializer='he_uniform',kernel_regularizer=regularizers.L1(0.0001)))\n",
        "cnn2_model.add(BatchNormalization())\n",
        "cnn2_model.add(Dense(1, activation='sigmoid',kernel_regularizer=regularizers.L1(0.0001)))\n",
        "cnn2_model.add(Dropout(0.3))\n",
        "opt = Adam(learning_rate=0.001)\n",
        "cnn2_model.compile(optimizer = opt,loss = 'binary_crossentropy' , metrics = ['accuracy'])\n",
        "history2 = cnn2_model.fit(x_train_normalize ,y_train,epochs=30,batch_size=32,validation_split=0.2,callbacks=[history_logger])\n",
        "loss_value , accuracy = cnn2_model.evaluate(x_test_normalize,y_test)\n",
        "print(loss_value,accuracy)\n",
        "plt.plot(history2.history['accuracy'])\n",
        "plt.plot(history2.history['val_accuracy'])\n",
        "plt.title('model accuracy')\n",
        "plt.ylabel('accuracy')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train','validation'], loc='upper left')\n",
        "plt.show()\n",
        "# summarize history for loss\n",
        "plt.plot(history2.history['loss'])\n",
        "plt.plot(history2.history['val_loss'])\n",
        "plt.title('model loss')\n",
        "plt.ylabel('loss')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'validation'], loc='upper right')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "VAtn3_R_HnOX",
        "outputId": "697c131a-2656-4565-a838-ad5426157d7c"
      },
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/30\n",
            "978/978 [==============================] - 34s 24ms/step - loss: 1.4325 - accuracy: 0.6018 - val_loss: 0.7383 - val_accuracy: 0.6424\n",
            "Epoch 2/30\n",
            "978/978 [==============================] - 22s 23ms/step - loss: 1.0010 - accuracy: 0.8575 - val_loss: 0.3534 - val_accuracy: 0.8804\n",
            "Epoch 3/30\n",
            "978/978 [==============================] - 22s 23ms/step - loss: 0.9022 - accuracy: 0.8804 - val_loss: 0.3366 - val_accuracy: 0.8949\n",
            "Epoch 4/30\n",
            "978/978 [==============================] - 22s 23ms/step - loss: 0.8783 - accuracy: 0.8801 - val_loss: 0.3168 - val_accuracy: 0.8805\n",
            "Epoch 5/30\n",
            "978/978 [==============================] - 22s 23ms/step - loss: 0.9032 - accuracy: 0.8753 - val_loss: 0.3493 - val_accuracy: 0.8801\n",
            "Epoch 6/30\n",
            "978/978 [==============================] - 22s 23ms/step - loss: 0.8426 - accuracy: 0.8783 - val_loss: 0.3258 - val_accuracy: 0.8894\n",
            "Epoch 7/30\n",
            "978/978 [==============================] - 22s 23ms/step - loss: 0.8761 - accuracy: 0.8790 - val_loss: 0.3099 - val_accuracy: 0.8797\n",
            "Epoch 8/30\n",
            "978/978 [==============================] - 22s 23ms/step - loss: 0.8224 - accuracy: 0.8784 - val_loss: 0.3375 - val_accuracy: 0.8673\n",
            "Epoch 9/30\n",
            "978/978 [==============================] - 22s 23ms/step - loss: 0.8421 - accuracy: 0.8727 - val_loss: 0.3502 - val_accuracy: 0.8765\n",
            "Epoch 10/30\n",
            "978/978 [==============================] - 22s 23ms/step - loss: 0.8632 - accuracy: 0.8732 - val_loss: 0.2909 - val_accuracy: 0.8854\n",
            "Epoch 11/30\n",
            "978/978 [==============================] - 22s 23ms/step - loss: 0.8411 - accuracy: 0.8773 - val_loss: 0.4297 - val_accuracy: 0.8767\n",
            "Epoch 12/30\n",
            "978/978 [==============================] - 22s 23ms/step - loss: 0.8146 - accuracy: 0.8801 - val_loss: 0.2834 - val_accuracy: 0.8856\n",
            "Epoch 13/30\n",
            "978/978 [==============================] - 22s 23ms/step - loss: 0.8314 - accuracy: 0.8743 - val_loss: 8.1216 - val_accuracy: 0.1775\n",
            "Epoch 14/30\n",
            "978/978 [==============================] - 23s 23ms/step - loss: 0.8434 - accuracy: 0.8789 - val_loss: 0.2691 - val_accuracy: 0.8947\n",
            "Epoch 15/30\n",
            "978/978 [==============================] - 22s 23ms/step - loss: 0.8621 - accuracy: 0.8777 - val_loss: 0.3023 - val_accuracy: 0.8814\n",
            "Epoch 16/30\n",
            "978/978 [==============================] - 22s 23ms/step - loss: 0.8694 - accuracy: 0.8736 - val_loss: 0.3937 - val_accuracy: 0.8489\n",
            "Epoch 17/30\n",
            "978/978 [==============================] - 22s 23ms/step - loss: 0.8560 - accuracy: 0.8745 - val_loss: 0.3007 - val_accuracy: 0.8847\n",
            "Epoch 18/30\n",
            "978/978 [==============================] - 22s 23ms/step - loss: 0.8837 - accuracy: 0.8758 - val_loss: 0.3269 - val_accuracy: 0.8792\n",
            "Epoch 19/30\n",
            "978/978 [==============================] - 22s 23ms/step - loss: 0.8797 - accuracy: 0.8747 - val_loss: 0.7823 - val_accuracy: 0.6485\n",
            "Epoch 20/30\n",
            "978/978 [==============================] - 22s 23ms/step - loss: 0.8712 - accuracy: 0.8744 - val_loss: 0.3488 - val_accuracy: 0.8786\n",
            "Epoch 21/30\n",
            "978/978 [==============================] - 22s 23ms/step - loss: 0.8811 - accuracy: 0.8755 - val_loss: 0.3251 - val_accuracy: 0.8809\n",
            "Epoch 22/30\n",
            "978/978 [==============================] - 22s 23ms/step - loss: 0.8538 - accuracy: 0.8799 - val_loss: 0.2864 - val_accuracy: 0.8935\n",
            "Epoch 23/30\n",
            "978/978 [==============================] - 22s 23ms/step - loss: 0.8526 - accuracy: 0.8758 - val_loss: 0.3894 - val_accuracy: 0.8772\n",
            "Epoch 24/30\n",
            "978/978 [==============================] - 22s 23ms/step - loss: 0.8712 - accuracy: 0.8793 - val_loss: 0.4105 - val_accuracy: 0.8765\n",
            "Epoch 25/30\n",
            "978/978 [==============================] - 22s 23ms/step - loss: 0.8531 - accuracy: 0.8763 - val_loss: 0.4649 - val_accuracy: 0.8769\n",
            "Epoch 26/30\n",
            "978/978 [==============================] - 22s 23ms/step - loss: 0.8513 - accuracy: 0.8782 - val_loss: 0.5003 - val_accuracy: 0.7790\n",
            "Epoch 27/30\n",
            "978/978 [==============================] - 22s 23ms/step - loss: 0.8821 - accuracy: 0.8692 - val_loss: 0.3715 - val_accuracy: 0.8769\n",
            "Epoch 28/30\n",
            "978/978 [==============================] - 23s 23ms/step - loss: 0.8732 - accuracy: 0.8733 - val_loss: 0.3660 - val_accuracy: 0.8770\n",
            "Epoch 29/30\n",
            "978/978 [==============================] - 22s 23ms/step - loss: 0.8390 - accuracy: 0.8792 - val_loss: 0.2928 - val_accuracy: 0.8862\n",
            "Epoch 30/30\n",
            "978/978 [==============================] - 22s 23ms/step - loss: 0.8546 - accuracy: 0.8796 - val_loss: 0.2829 - val_accuracy: 0.8857\n",
            "524/524 [==============================] - 5s 9ms/step - loss: 0.2929 - accuracy: 0.8852\n",
            "0.29293861985206604 0.8851842880249023\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "id": "w_E0Md7h3Fk3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "2a0d25e0-ad97-45a7-a7a4-c11dbbc06957"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/30\n",
            "978/978 [==============================] - 34s 21ms/step - loss: 0.9221 - accuracy: 0.8045 - val_loss: 0.2976 - val_accuracy: 0.8758\n",
            "Epoch 2/30\n",
            "978/978 [==============================] - 19s 19ms/step - loss: 0.8427 - accuracy: 0.8794 - val_loss: 0.2907 - val_accuracy: 0.8787\n",
            "Epoch 3/30\n",
            "978/978 [==============================] - 19s 19ms/step - loss: 0.8081 - accuracy: 0.8817 - val_loss: 0.2851 - val_accuracy: 0.8842\n",
            "Epoch 4/30\n",
            "978/978 [==============================] - 19s 19ms/step - loss: 0.8138 - accuracy: 0.8887 - val_loss: 0.2438 - val_accuracy: 0.9011\n",
            "Epoch 5/30\n",
            "978/978 [==============================] - 19s 20ms/step - loss: 0.8323 - accuracy: 0.8773 - val_loss: 0.3877 - val_accuracy: 0.8758\n",
            "Epoch 6/30\n",
            "978/978 [==============================] - 19s 19ms/step - loss: 0.8791 - accuracy: 0.8691 - val_loss: 0.3850 - val_accuracy: 0.8758\n",
            "Epoch 7/30\n",
            "978/978 [==============================] - 19s 19ms/step - loss: 0.8727 - accuracy: 0.8691 - val_loss: 0.3825 - val_accuracy: 0.8544\n",
            "Epoch 8/30\n",
            "978/978 [==============================] - 19s 20ms/step - loss: 0.8317 - accuracy: 0.8766 - val_loss: 1.2014 - val_accuracy: 0.5144\n",
            "Epoch 9/30\n",
            "978/978 [==============================] - 19s 20ms/step - loss: 0.7915 - accuracy: 0.8790 - val_loss: 0.2726 - val_accuracy: 0.8822\n",
            "Epoch 10/30\n",
            "978/978 [==============================] - 19s 20ms/step - loss: 0.7965 - accuracy: 0.8856 - val_loss: 0.2659 - val_accuracy: 0.8873\n",
            "Epoch 11/30\n",
            "978/978 [==============================] - 19s 20ms/step - loss: 0.7984 - accuracy: 0.8877 - val_loss: 0.2759 - val_accuracy: 0.8868\n",
            "Epoch 12/30\n",
            "978/978 [==============================] - 19s 20ms/step - loss: 0.7954 - accuracy: 0.8832 - val_loss: 0.2643 - val_accuracy: 0.8910\n",
            "Epoch 13/30\n",
            "978/978 [==============================] - 19s 20ms/step - loss: 0.7928 - accuracy: 0.8848 - val_loss: 0.2684 - val_accuracy: 0.8934\n",
            "Epoch 14/30\n",
            "978/978 [==============================] - 19s 20ms/step - loss: 0.7970 - accuracy: 0.8857 - val_loss: 0.2504 - val_accuracy: 0.8910\n",
            "Epoch 15/30\n",
            "978/978 [==============================] - 19s 20ms/step - loss: 0.8032 - accuracy: 0.8866 - val_loss: 0.2701 - val_accuracy: 0.8862\n",
            "Epoch 16/30\n",
            "978/978 [==============================] - 20s 21ms/step - loss: 0.7985 - accuracy: 0.8873 - val_loss: 0.2550 - val_accuracy: 0.8908\n",
            "Epoch 17/30\n",
            "978/978 [==============================] - 19s 20ms/step - loss: 0.7975 - accuracy: 0.8871 - val_loss: 0.2959 - val_accuracy: 0.8820\n",
            "Epoch 18/30\n",
            "978/978 [==============================] - 20s 21ms/step - loss: 0.7656 - accuracy: 0.8883 - val_loss: 0.2376 - val_accuracy: 0.8969\n",
            "Epoch 19/30\n",
            "978/978 [==============================] - 20s 20ms/step - loss: 0.7924 - accuracy: 0.8888 - val_loss: 0.2274 - val_accuracy: 0.9075\n",
            "Epoch 20/30\n",
            "978/978 [==============================] - 20s 20ms/step - loss: 0.7807 - accuracy: 0.8872 - val_loss: 0.2327 - val_accuracy: 0.9027\n",
            "Epoch 21/30\n",
            "978/978 [==============================] - 20s 21ms/step - loss: 0.7727 - accuracy: 0.8861 - val_loss: 0.3288 - val_accuracy: 0.8805\n",
            "Epoch 22/30\n",
            "978/978 [==============================] - 20s 21ms/step - loss: 0.7693 - accuracy: 0.8912 - val_loss: 0.2402 - val_accuracy: 0.8961\n",
            "Epoch 23/30\n",
            "978/978 [==============================] - 20s 20ms/step - loss: 0.7934 - accuracy: 0.8898 - val_loss: 0.2364 - val_accuracy: 0.8997\n",
            "Epoch 24/30\n",
            "978/978 [==============================] - 19s 20ms/step - loss: 0.7633 - accuracy: 0.8907 - val_loss: 0.2476 - val_accuracy: 0.8916\n",
            "Epoch 25/30\n",
            "978/978 [==============================] - 19s 20ms/step - loss: 0.7767 - accuracy: 0.8922 - val_loss: 0.2441 - val_accuracy: 0.8963\n",
            "Epoch 26/30\n",
            "978/978 [==============================] - 19s 20ms/step - loss: 0.7847 - accuracy: 0.8898 - val_loss: 0.2487 - val_accuracy: 0.8979\n",
            "Epoch 27/30\n",
            "978/978 [==============================] - 20s 21ms/step - loss: 0.7839 - accuracy: 0.8904 - val_loss: 0.2878 - val_accuracy: 0.8822\n",
            "Epoch 28/30\n",
            "978/978 [==============================] - 19s 20ms/step - loss: 0.7889 - accuracy: 0.8841 - val_loss: 0.2435 - val_accuracy: 0.8923\n",
            "Epoch 29/30\n",
            "978/978 [==============================] - 19s 20ms/step - loss: 0.7515 - accuracy: 0.8914 - val_loss: 0.3540 - val_accuracy: 0.8429\n",
            "Epoch 30/30\n",
            "978/978 [==============================] - 20s 20ms/step - loss: 0.7908 - accuracy: 0.8840 - val_loss: 0.3334 - val_accuracy: 0.8764\n",
            "524/524 [==============================] - 5s 9ms/step - loss: 0.3553 - accuracy: 0.8691\n",
            "0.35534125566482544 0.8691399097442627\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYkAAAEWCAYAAACT7WsrAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3deZiT5bn48e+dTDKZfQYYQTZBBAERQUe04taqlWrd6r60Yk/F02rVbqfaX09r7XI8PdbaxVq1pbW2iBatUou12LqvQMENFBBRhnVYZp9JJsn9++N5ZwjDLJklhEnuz3XlSvIuyfMmM++dZ3nvR1QVY4wxpiO+dBfAGGPM/suChDHGmE5ZkDDGGNMpCxLGGGM6ZUHCGGNMpyxIGGOM6ZQFCWMAEfm9iPwgyW3Xi8ipqS6TMfsDCxLGGGM6ZUHCmAwiIjnpLoPJLBYkzIDhNfN8Q0TeFJEGEfmtiAwVkSdFpE5EnhaRsoTtzxaRd0SkWkSeFZFJCeumi8i/vf0eAkLt3uvTIrLC2/dlEZmaZBnPFJHlIlIrIhtE5JZ264/3Xq/aWz/bW54nIj8RkQ9FpEZEXvSWnSwilR18Dqd6j28RkQUi8kcRqQVmi8gMEXnFe4/NIvJLEQkm7H+YiCwWkZ0islVEviUiw0SkUUQGJ2x3pIhUiUggmWM3mcmChBlozgdOAyYAZwFPAt8CynF/z9cDiMgE4EHgRm/dIuCvIhL0TpiPAQ8Ag4A/e6+Lt+90YC5wDTAYuAdYKCK5SZSvAfgcUAqcCXxRRM71Xvcgr7y/8Mo0DVjh7Xc7cBRwnFem/wLiSX4m5wALvPf8ExADvgIMAT4GnAJ8yStDEfA08HdgOHAI8E9V3QI8C1yU8LqfBearakuS5TAZyIKEGWh+oapbVXUj8ALwmqouV9Vm4C/AdG+7i4G/qepi7yR3O5CHOwkfCwSAO1W1RVUXAEsS3mMOcI+qvqaqMVW9Hwh7+3VJVZ9V1bdUNa6qb+IC1Une6suAp1X1Qe99d6jqChHxAZ8HblDVjd57vqyq4SQ/k1dU9THvPZtUdZmqvqqqUVVdjwtyrWX4NLBFVX+iqs2qWqeqr3nr7geuABARP3ApLpCaLGZBwgw0WxMeN3XwvNB7PBz4sHWFqsaBDcAIb91G3TO75YcJjw8CvuY111SLSDUwytuvSyJyjIg84zXT1AD/iftFj/ca73ew2xBcc1dH65KxoV0ZJojIEyKyxWuC+lESZQB4HJgsImNxtbUaVX29l2UyGcKChMlUm3AnewBERHAnyI3AZmCEt6zV6ITHG4Afqmppwi1fVR9M4n3nAQuBUapaAvwaaH2fDcC4DvbZDjR3sq4ByE84Dj+uqSpR+1TOdwPvAuNVtRjXHJdYhoM7KrhXG3sYV5v4LFaLMFiQMJnrYeBMETnF63j9Gq7J6GXgFSAKXC8iARH5DDAjYd/7gP/0agUiIgVeh3RREu9bBOxU1WYRmYFrYmr1J+BUEblIRHJEZLCITPNqOXOBO0RkuIj4ReRjXh/IaiDkvX8A+DbQXd9IEVAL1IvIROCLCeueAA4UkRtFJFdEikTkmIT1fwBmA2djQcJgQcJkKFV9D/eL+Be4X+pnAWepakRVI8BncCfDnbj+i0cT9l0KXA38EtgFrPW2TcaXgFtFpA74Di5Ytb7uR8AZuIC1E9dpfYS3+uvAW7i+kZ3A/wI+Va3xXvM3uFpQA7DHaKcOfB0XnOpwAe+hhDLU4ZqSzgK2AGuAjyesfwnXYf5vVU1sgjNZSmzSIWNMIhH5FzBPVX+T7rKY9LMgYYxpIyJHA4txfSp16S6PST9rbjLGACAi9+OuobjRAoRpZTUJY4wxnbKahDHGmE5lTDKwIUOG6JgxY9JdDGOMGVCWLVu2XVXbX3vTJmOCxJgxY1i6dGm6i2GMMQOKiHQ51Nmam4wxxnTKgoQxxphOWZAwxhjTqYzpk+hIS0sLlZWVNDc3p7soGSMUCjFy5EgCAZuHxphskNFBorKykqKiIsaMGcOeCT9Nb6gqO3bsoLKykrFjx6a7OMaYfSCjm5uam5sZPHiwBYh+IiIMHjzYambGZJGMDhKABYh+Zp+nMdklo5ubjDEdqNsKbz8CJSNh1DFQNDTdJTL7MQsSKVZdXc28efP40pe+1KP9zjjjDObNm0dpaWmKSmayTsN2eOlOeP03EG3avbz0IBcsRs1w9wdMBr+dGoyT0r8EEZkF/AzwA79R1dvarT8INyNXOW6ilStUtdJbdyVuFi6AH3iT0Q841dXV/OpXv9orSESjUXJyOv/4Fy1alOqimWzRuBNe/jm8dq8LDodfBMd/BcJ1sOE1d/vgOXjLmx8pWAgjjvICxzEw+ljILez6PUzGSlmQ8ObivQs3C1YlsEREFqrqyoTNbgf+oKr3i8gngP8BPisig4DvAhW4+XuXefvuSlV5U+Wmm27i/fffZ9q0aQQCAUKhEGVlZbz77rusXr2ac885hw0bNtAcDnPDDTcwZ84cYHeakfr6ej71qU9x/PHH8/LLLzNixAgef/xx8vLy0nxkZr/XtAte+RW8ejdE6mHKZ+Ckm6B8wu5tRh0NXAeqUP0RbHjdCxyvwgu3g8ahZDRctQhKR6XtUEz6pLImMQNYq6rrAERkPnAOkBgkJgNf9R4/AzzmPT4dWKyqO719FwOzgGQmou/Q9/76Dis31fZ29w5NHl7Md886rMttbrvtNt5++21WrFjBs88+y5lnnsnbb7/thpBGGph7238xqKSApoKRHD3zZM4//3wGDx68x2usWbOGBx98kPvuu4+LLrqIRx55hCuuuKJfj8VkkOYaePXX8MpdEK6Byee44DB0cuf7iEDZQe429UK3LFwHH7wAf7kGHjgXrvo7FHaaB67/NO6Ebatg20oXuMaeBOM+Dj5/6t/b7CWVQWIEsCHheSVwTLtt3sDNNfwz4DygSEQGd7LviPZvICJzgDkAo0eP7reCp9KMGTNcgGjcAdUb+Plv/8RfnvwnABsqt7BmzZq9gsTYsWOZNm0aAEcddRTr16/f18U2A0G4Hl6/B176OTRXw8RPw8k3wbDDe/d6uUUw8Qy47GF44Dz443lw5ROQ10/9ZOE6qHrPBYPWoLBtFdRv3b2N+F1TWfFImHYZTL8cysb0/b1VXTBt2ulqXI273OPGnXveHzAJZt4I/uy9eDTdvVNfB34pIrOB53ETvceS3VlV7wXuBaioqOhy9qTufvHvKwUFBVBTCQ1VPLvkHZ5+5Q1eeekl8ps2cvIFV9Pc2LjXPrm5uW2P/X4/TU1Ne21jstxHr8KjV7tf3hNmueAwfHr/vPZBH4OL/wgPXgLzLobPPgrBgt69VjQCz/7Ija6q/mj38kA+lE+EQ051J+YDJrkO9PzB8N6TsPwBeP7/4Pkfw9gTYfrnYNKnIZBEs2ukETYt95rRXnePG6pAOzvVCIRKIFTsyrnmabhgLpTs9Ts1K6QySGwEEhsxR3rL2qjqJlxNAhEpBM5X1WoR2Qic3G7fZ1NY1pQpKiqirs6bCTIWg0iD+wMtKKdGCygbNIj80iG8u34dry57Axq2uV85pv+puhpcTaW71W+BvDIoGeWGgxYOHXhNGrGoO3E+/39QOhquehIOOq7/32f8qXD+fbDg8/DQFXDpfMjJ7X6/RLs+hAVXwcZlcOgZcOSVLhAcMMmNsPJ1ctnWYee6W00lrJjnAsajX3An8sMvgulXwPBpu7ev2bg7IGx4Dba8CfGoWzd4vGu6Kh4OeYMgf5C7zytLeFy6++/grQXw1xvg18fDZ+6F8af1/LMb4FIZJJYA40VkLC44XAJclriBiAwBdqpqHLgZN9IJ4CngRyJS5j3/pLd+wBk8eDAzZ85kymGHkRcQhg4Z5P6Z8wcz61Pl/Pqee5k0aRKHHnooxx59lKuC129Ld7EHJlWo3QTbV+8OBDWVUJvwONrF1eK+HHfyKB7pgkbbbRQMHgeDDnZt9/0hXOfKGyru/WvsXAePzoHKJXDEZXDGj10TUaocdp5r0lp4HTzyBbjgd8kPlV31BDz+JVTjrP/Er3h/yKmU5AcoyQtQmhOgOKaEuru0t2QknPRfcMLXYf0LsPwB9N9/QJbcR/SAKYSLDyaweSnBhk0ARH0hthRNZv0Bl/FuYDJvy3gqw/nUfxhlcGGQ4SV5jCjLY3hpHiP8eQzPz+PAYIhQ4g+Fwy+AA4+AP8+GP10Ax38VPv7/ejdEeNNyePFO1+9z2q1dbrq1tpnXPtjJkaNLGVmW3/P36kcpneNaRM4A7sQNgZ2rqj8UkVuBpaq6UEQuwI1oUlxz07WqGvb2/TzwLe+lfqiqv+vqvSoqKrT9pEOrVq1i0qRJ/XpMvdK0y1WtxQ+Dxu5RVW/9/EXEnTR2rXftyYMOdr+U9kP7zedat8X9421a4d0vdzWxNgJFw6B4xJ4n/BLveeEw993UVELNBqjdmBBcNriA0/oLFFzTR+L1BMOnJ9fcEY/DjrVQ+bo7oW9Y4trf/UF3Eppxdc+ahlTdL+on/8v94v30T2HK+d0UQWmJx4lE47TElGgsTlEoQF6wFzWnV34FT90M0y6Hs3/ZeQ0ACIeb2PXYzQxb9TvWBcbzxfB1vBfpuPM7N8dHqRc43C1IUSiHSDROYyRKQyRGYyRKYzhGYyRGQyRKTqSGT8tLXOh/jkFSx/L4ISyLT2BZfAKrdDRRcijMzXHBKN/d8oM5bK8Ps6m6iW114b0q7kMKcxlRGmJ4aR6FuTmEo3HikUYuqPolJ9cvYmVgCv9b8HU2xQcRjsYJR2OowlEHlXHKpKF8/NByBhcm1LI+etXV9NY+DYj7oXH9ChcsEuxqiLDo7c389Y1NvPbBzrZyHT2mjLOnjeDMww9kUEGw599XN0RkmapWdLo+lUFiX9ovg4Qq1G12HXGBfHfi9wdQVRojMWqaWqhpaiGuygFFIQYXBvFpHHasgWgYhkxI7iS0j/XL5xprcSf1mg3uZOkPus/GH6S2xcfOJmV7k1LVBFWNcXY1Rjg8uJkp8gHldSvxbV7hPlsA8cGQQ92Jdvg014RROgqKhkNOH/6p4jFXq6vZ4E7qrc0XO9a69b4c9yuzNXCMnIEWDyfSsIvoh0vQyiX4Ny4huGU5/kgNANFAEdWDjmB76VSC4R2M3rCQnFgTuwYdwUeHXMG2UbPw5eTi8wk+EfwiiEBjJEZdcwuR+p1UvPU9Dql6mnUF0/jDsJv5KDaY+uYotc0tNLfEiETjRGJKS6w1KMSJxjv+P8/N8VGWH6SsIEhZfoCy/CCl+QEGFQQpzXfLSvICFOS6k21hbg4FuTmUvX47OS/8GI75Isz6n7YaVmMkyvKPqnlt3Q7WrVnJnG3fZ6q8z++ip/Po4GuYfvBQjhk7mBFledR6f/973BpbqG6KeM+j1DW3kJvjoyA3h/ygn/yguy8I5pCf6+7zgn4Kgn4KQwHKvEBQkhdsCzgBf+dBLBKNs7W2mcpdTWyq9m41TWysbmbjrkYaIzFCAT9Bv4/cgI9TWp7lmtpfEJEQvx92M+uKjyE3x09LLM5L729na20YEZg+soQrD/yI07b/gfxNr7gfGB+71jWz3X0cHHc9nPY96ppbWLxyKwvf2MSLa7YTjSsHlxdw9hHDOf6QIby6bgePrdjE2m315PiEEyeUc8604Zw6aSgFuf3TEGRBIl1BIh51bbDhWsgfjBaPpLEl3vbP0BKLIyIU5eYQV6U+HCXo9zG0JERpEGT7e7tPfvvZ1a+9+lxbml1b9IcvwfoX0colSMvenfTdiavwAQdSmTeRcPnhFB08gzGHHcuw8uQTOaoqTS0xdjW20BCO0hRxv0ybW9x9U0uMpkg04bG7b33ub9rJyIa3Gdv0DhMiKzk0tpoQEQB2aBGDpa6trKt1JP+OH8JyHc+/4+NZpweiCSnTimjkM/4X+Jz/H4zzbaZKS5gX+wTzoqewlUF7lPtY30ruCPyKcmr4WfwiHgqcS0FeLoWhHIpyAxSG3Ak06PcRyPER9PsI5vgI+IWA30fA7yM3x937fEJdcwvVjS3saoiwq7GFXY0RdjVGqG5soboxQidxpfVT5LuBB7jK/3d+67+Yhwovx+/zsWZrHdG48in/6/xf8D5yfMK7M/6HMSdcQml+//8KTouq1fDw56DqXTjx63DyzeDzo6q8XVnD+y8/ysQ19zAx+i5btZT5wc9QP/lyTjjsII45eBA5f/4csQ9e4Osj5/PU6hrC0TgjSvP49BEHctbU4Rw2vHiPv2VVZdXmOh5fsZGFb2xic00zeQE/p00eyjnThnPihPIuA2F3LEjsyyARj7uLliJ10FSNxlpoKTiQHVpMTVMLkYTAUJIfoDiUg9+rqtc1t7Clppmmlhh5AT8jCpT82nWuaWrwOBcw9hNJfa6RBte0sv4lFxgql0IsjCJ8FBjLs83jeSU2iU2B0Qwr9DO0QCjP8zEkTxiSB4NCQllIKQ1CcSBODnG2h0axLDyKpZujLN9QzVsba4hE4wAMLc5l+qgypo0uZXhpHtWNEXY2RNjVEGGndyLc2eBOgjsbIoS9/ZKRm+Nr+xUbCvjID+aQF/CTF/S7X7U5ysGxDxjX/A4jmtdQXzCKXWVHUDt4Kjl5JYQCfkIBX9t9bo6fUMBPjk+IqaKqxGJxQhuep+zt31P00T9BfOw66HS2Hvo5aodMY+xbP6P8zV8TKzuY+Hn3Ehzd6f90v4jHldrmFnY1tlDb5IJpXThKg3erC0dpaI5w6pofMH3H3/jz4C/yt8LPMGVoiEt23cvINQ/A8CPhwt/1z5DV/U2kERZ9A1b8Ecac4Dq1K5e4ZqUtb0HJaGqOuo4ncz7B4tXVvLh2O+FonIKgnxnyDr+TW/me70vEj7ics6cNZ/qoMny+7n/kxOPKkvU7efyNTSx6azPVjS2U5Qf4zJEj+e9Pd3EdTBcsSKQySKhCSyMarkOb65CWBgRFESK+EFu0jJpY7u7AkBegOG93YNj75ZTqpha21jQTicU5MNBEeWwL5A/Zr6527fJzjYZh3kWw/kWIR1HxUVU4kZejE3miZixL4ocyeMhQTp8yjFmHDWPqyJJeZ5aNROOs2lzLig3VLP9oF8s3VPPhjj1rJyV5rumkLL/1PujuvWWtzRihwO6mjNYAkBdwt2T+efvVrvWw5Dfw7wdc/1Ruibso7qjZcPqPej/8NBViUTdiadVC+MR/w6q/wuYVcOyX4NTv9a25byBYMQ+e+CrEIm5I7aBxcMLXYOpFe1xb0RSJ8fL72/nnu9vQeJxvf/Qf5OflIdc83+vBEJFonBfWVPH4ik0E/D5+ctERvXodCxL9GCQi0Rj1DQ34I/UEYg3kxhvx436RNmmQevKo1zwaCIG4dtTWwJDTRedee/G4sqMhwra6Zg7QHZRLDdGiEeQUHdBvx9IXnX2uqkr9xlUU/eZjvDf4VB5onsljO0ZRTz5TRhQz67BhnH7YMA45oDBlKcd31IfZ2RChrCBIaV6AnD5Uw9Mu0ghvL4DVT8ERl8Cks9Jdoo5Fw+4aivf/5QZbnHs3TDwz3aXad7atgudvh0M/5UaAJTOMeslv4W9fhf9Y7Pqz0siCRE+DhKrrVI2F3R9/6300QjzajA/3eUXIodmXT4u/gFigEH9OwGv3de2/fp/0+UQYjcWpqgtT2PARBTSyK280xSVlfWp/7Iu4ulExK1etYl1LKZtrmtlYndDhV93MuMi7PJ77HT7f8g3qR53C6VOG8cnJQxk1KL3D+EyKRRpc7WfyuXuN2jEdCNfDHZPddRcX/LZvr9XS7K5Z6eX5prsgsX/1iKZDLOpGHyUGBBIDp0BOEPXnsisaRAJ5DCobRDAnl1RUpAsLC6mvr2fTpk1cf/31LFiwgEjBwcS2r6akqZLqpmrOu/Byfnzrtzn26Gl09mdx5z33M+faG8kvcE0TvU093hKLt404aYy4K1Sr6iJ8deEbAAwuCDK8NI+xQwo4btwQjo7vgBVwxxUnUDrpY73+HMwAEyyAmTekuxQDR26hSzHy+r1Q90M3VLu3nvgKNG6HSx/qcjhyb1mQEHF59nOCLhrnFrv7nNzdQzNFaAxH2dhUz+iC/H3Szjp8+HAWLFgAQDAQgPJxxHesY1CsHh9RfJFa4g07ENw1FnsGC+XOX9zFFReeQ/7YKUDPUo/H4kptUwvVTS3UN0dRlLyAnwOKQgRzfMR3BvnX105ieGkeoUC7qvXKtbACSssGd/zixhjn6C/Aq7+CZffDyd/s3Wt88Dy8Mc/1g6QgQEAWTF/aLZ8fDpzqUgMMOthdaFUwxF25mlCFqw+7i6oKezg2+aabbuKuu+5qe37LLbfwgx/8gFNOOYUjjzySww8/nMcff3yv/davX8+UKe4E39TUxCVXzOawk8/n/GtvoTnmp7F4LB8GxnHJN3/GlNOvZMLHL+amXzyMDjucnz/8LJu2bufjZ5zHx0+cCbjU49u3bwfgjjvuYMqUKUyZMoU777wTgHUffMChEydy6Wev4tBJkznjU7Ooqa2nvCjIhKFFjB9axLCSEIMKguQG/BxcXrh3gAB3JTGk9spfYzLB4HEuV9XSuS6nVU9Fw64WUTYGTvxGvxevVfbUJJ68yQ1N66XilhjFQE7iiXHY4fCp2zrdB+Diiy/mxhtv5NprrwXg4Ycf5qmnnuL666+nuLiY7du3c+yxx3L22Wd32odx9913k5+fz6pVq3jzzTc58sgjKQoFOLi8kJ/95H+R3EK21zVz1YVn8fi/XuGi2dfwkzt+yuLH5zOs0EekejMKNIajvPTKa/x27lyeef4lVJWTTziOSUcegy+3kPfXruVHP7+P//v5XVx/9ZW89dI/mPbZz/bsg2r20rH3Jd2EMdlixjUw70J496/dXjW/lxd/6i7svOKRlF50azWJJChKLK74ezEUcvr06Wzbto1NmzbxxhtvUFZWxrBhw/jWt77F1KlTOfXUU9m4cSNbt27t9DWef/75tvkjpk6dytSpU9vWPf7oI5zx8ZlcdsZJfLD2PT5at5qqumai8Tirm0up1gKCjVuIx2J8sKOBhf94huNPPYMtjcrWJjjhtDN59tnnyQv6OWjMWM477XhGluVzzNEVfPjhhz3/sFprEkGrSRjTrUNOdTWB1+/r2X7b18ALP4EpF7jXSKHsqUl084u/KzWNET7a2ci48kKCvbgU/sILL2TBggVs2bKFiy++mD/96U9UVVWxbNkyAoEAY8aMobm5i8Rznfjggw+4/fbbWbJkCWVlZcyePZviABw6rJgcn49RZfn4Bg2mpX4DfmKMKwi73C/hAGOGFCBAaX6AoSUhRpblkxfKxefVZnqdkjxcC4GC/e4qcWP2Sz4fHH01/OP/weY3XdN3d1RdM1NOnrtuJtVFTPk7ZID65ih+n5Dfm2RouCan+fPns2DBAi688EJqamo44IADCAQCPPPMM93+Yj/xxBOZN28eAG+//TZvvvkmALW1tRQUFFBSUsLWrVt58sknAQjm+CguLoJoM8V5QQJDxoL4CDVt4bSPHcGTT/yVnHgLvliERX9dyEknntir4+pQuNaamozpiemXu9xur9+b3PZvzHdZcE+7BYqGprRokE01iV5SVerCUQpzc3p93cNhhx1GXV0dI0aM4MADD+Tyyy/nrLPO4vDDD6eiooKJEyd2uf8Xv/hFrrrqKiZNmsSkSZM46qijADjiiCOYPn06EydOZNSoUcycObNtnzlz5jBr1iyGDx/OM88845LRBYs4cmwOsy+7kBkz3AU8X/jCF5g+fXr/zXbXXGud1sb0RF6Zu0L7jfkuhXj+oM63bdzpah0jZ8CRs/dJ8exium40t8RYvbWOkWV5DCro4SQr+5t43M1BEKlzk7x09cfYhS4/1wc+46aFvPqffSioMVlm6zsuO+xp34eZ13e+3ePXumByzfMwtH9m2+zuYjprbupGXXPr0NcMmOPW5/PmsyiE6g/dXAr9LWw1CWN6bOhhcNDxsOQ+l6K+I+tfguV/dCnH+ylAJMOCRDfqmlsI5fgJ5mTIR+Xzu+tBggUukVxTdf++frjO+iSM6Y0Z3hzla/6x97poGJ640c1qeVIvL7zrpZSe+URkloi8JyJrReSmDtaPFpFnRGS5iLzpzWSHiIwRkSYRWeHdft3bMvSlOS0eVxoiMQpDGdZ14/O7bJUBL1BEkp/XodvP0/okjOmdiWe6ibI66sB+6eduWt4zfrLPswCnLEiIiB+4C/gUMBm4VETaJzz/NvCwqk7HzYH9q4R176vqNO/2n70pQygUYseOHb0OFPWRKKpKUaYFCfACxRhAXXK2JKgqO3bsIBQKdb5RuM6ltjbG9Iw/AEd/3mXT3b5m9/Id77t5KiafCxM+uc+Llcqz3wxgraquAxCR+cA5wMqEbRRobZsoATb1ZwFGjhxJZWUlVVVVvdq/urHFzaNbG0pZauu0UoWabRAKQ2h7UruEQiFGjhzZ8cp4zHWKW03CmN45cjY892N3cd0ZP3b/o3/7qksRNKv313r1RSqDxAhgQ8LzSuCYdtvcAvxDRL4MFACJlw6OFZHlQC3wbVV9of0biMgcYA7A6NGj9ypAIBBg7NixvT6AU37yLCPK8vnD53s349OA8MNToeIqOP2HfX+t1qutrU/CmN4pLHdzUqyYB6f8N7z3d1j3LJxxOxQfmJYipbs39lLg96o6EjgDeEBEfMBmYLTXDPVVYJ6I7HXmUdV7VbVCVSvKy8v7tWCVuxp5v6qBkyb07+vud0LFbshqf7Dkfsb03YxrXI381V/DUzfDiKOg4vNpK04qg8RGIHHOzZHeskT/ATwMoKqvACFgiKqGVXWHt3wZ8D4wIYVl3cvzq13zy0kThuzLt933QiVu2Gp/aH2dXKtJGNNrI49y84M/8wN38dyn70xutrsUSWWQWAKMF5GxIhLEdUwvbLfNR8ApACIyCRckqkSk3Ov4RkQOBsYD61JY1r08v7qK4SUhxpUX7su33fdCJVaTMGZ/M2OOuz/2i8nlc0qhlPVJqGpURK4DngL8wFxVfUdEbgWWqupC4GvAfSLyFVwn9mxVVRE5EbhVRFqAOPCfqrozVWVtryUW56W12/n0EQdmZod1olAJNKzNGH4AABuJSURBVPSuY38vbWnCbXSTMX1y+IUgPph8drpLktrcTaq6CFjUbtl3Eh6vBGZ2sN8jwCOpLFtXVmyopi4c5cTxGd4fAa5paMf7/fNabc1NVpMwpk/8OXDExekuBZD+juv90vOrq/D7hOMOyfD+COjn5ibrkzAm01iQ6MBzq6uYPqqUkrwMyNfUndYg0R+JHq1PwpiMY0GinZ0NEd7aWMOJmT70tVWoGOIt0NKLCYbaa6517aj7OG2AMSZ1LEi088KaKlTJoiDhdTL3xzDYsHe1daZ39huTRSxItPPc6irK8gMcPiJLRui0Bon+6JcI11reJmMyjAWJBKrKC2u2c/z4cvy+LPk13K9BwtKEG5NpLEgkWLW5jqq6MCeOz4JRTa1af/k390NzU3ONdVobk2EsSCR4fo27qCzj8zUlaqtJ9MPkQ+E6G/5qTIaxIJHgufeqmDisiAOKu5gvIdP0e5+E1SSMySQWJDwN4ShLP9zJSYdmUS0CdvchWJ+EMaYDFiQ8r67bQUtMOSkbUnEkygmBP9g/Q2Bt6lJjMo4FCc9zq6vIC/g5akxZuouyb4n0T2qOaBhiYeuTMCbDWJDwPL+6iuPGDSY3J31529OmP4JEW0oOCxLGZBILEsCHOxpYv6Mxe66ybi+3uO9DYFuDjPVJGJNRLEjgahGQRak42uvXmoT1SRiTSSxIAM+t3s7oQfmMGZyf7qKkR78ECUsTbkwmSmmQEJFZIvKeiKwVkZs6WD9aRJ4RkeUi8qaInJGw7mZvv/dE5PRUlTESjfPK+9s5ccKQzJ+FrjOhYqtJGGM6lLKZ6bw5qu8CTgMqgSUistCbja7Vt4GHVfVuEZmMm8VujPf4EuAwYDjwtIhMUNVYf5dzZ0OEqSNL+cTEA/r7pQeOUEnfh8C2TV1qNQljMkkqpy+dAaxV1XUAIjIfOAdIDBIKtJ5VSoBN3uNzgPmqGgY+EJG13uu90t+FHFYS4sE5x/b3yw4soRJoaYRoBHKCvXsNG91kTEZKZXPTCGBDwvNKb1miW4ArRKQSV4v4cg/2RUTmiMhSEVlaVVXVX+XOPqFSd9+X2kTYa66y5iZjMkq6O64vBX6vqiOBM4AHRCTpMqnqvapaoaoV5eVZOjKpP+T2Q2qOcB34cyEnt3/KZIzZL6SyuWkjMCrh+UhvWaL/AGYBqOorIhIChiS5r+kv/ZHkr7nW+iOMyUCprEksAcaLyFgRCeI6ohe22+Yj4BQAEZkEhIAqb7tLRCRXRMYC44HXU1jW7NYfQcLShBuTkVJWk1DVqIhcBzwF+IG5qvqOiNwKLFXVhcDXgPtE5Cu4TuzZqqrAOyLyMK6TOwpcm4qRTcbTH5lgLU24MRkplc1NqOoiXId04rLvJDxeCczsZN8fAj9MZfmMp7Um0aeOa0sTbkwmSnfHtdkf9FefhDU3GZNxLEgYCBaC+KxPwhizFwsSxs0p0ddMsOEa65MwJgNZkDBOX5L8qVqfhDEZyoKEcfoSJCINoHGrSRiTgSxIGKcvQcLShBuTsSxIGKcvmWAtTbgxGcuChHH6UpNoSxNe0n/lMcbsFyxIGCe3DxMPtTU3WU3CmExjQcI4oRLXbBSP93xf65MwJmNZkDBOqATQ3vVLWJ+EMRnLgoRx+pKaw6YuNSZjWZAwTl8ywbbWJIKF/VceY8x+wYKEcfqSCTZcC8Ei8Pn7t0zGmLSzIGGcvjQ3hW1WOmMylQUJ4/Rlnutmm3DImEyV0iAhIrNE5D0RWSsiN3Ww/qcissK7rRaR6oR1sYR17ac9Nf2trSbRy9FNNvzVmIyUspnpRMQP3AWcBlQCS0RkoTcbHQCq+pWE7b8MTE94iSZVnZaq8pl2+lKTCNdCqLR/y2OM2S+ksiYxA1irqutUNQLMB87pYvtLgQdTWB7TFX+O63zu7egm65MwJiMlFSRE5FEROVNEehJURgAbEp5Xess6ev2DgLHAvxIWh0RkqYi8KiLndrLfHG+bpVVVVT0omulQqJepOaxPwpiMlexJ/1fAZcAaEblNRA7t53JcAixQ1VjCsoNUtcJ73ztFZFz7nVT1XlWtUNWK8vLyfi5SFgqVuBnmeips81sbk6mSChKq+rSqXg4cCawHnhaRl0XkKhEJdLLbRmBUwvOR3rKOXEK7piZV3ejdrwOeZc/+CpMKvckEG4tCS6MFCWMyVNLNRyIyGJgNfAFYDvwMFzQWd7LLEmC8iIwVkSAuEOw1SklEJgJlwCsJy8pEJNd7PASYCaxsv6/pZ73JBBu2lBzGZLKkRjeJyF+AQ4EHgLNUdbO36iERWdrRPqoaFZHrgKcAPzBXVd8RkVuBparaGjAuAearqibsPgm4R0TiuEB2W+KoKJMioRLY/l7P9rHkfsZktGSHwP5cVZ/paIXXb9AhVV0ELGq37Dvtnt/SwX4vA4cnWTbTX0IlPb9OwtKEG5PRkm1umiwibQPhveagL6WoTCZdWvsk9qjUdcNqEsZktGSDxNWq2nY1tKruAq5OTZFM2oSKQWMQaUh+H0sTbkxGSzZI+EVEWp94V1MHU1Mkkza9yQTbVpOwIGFMJko2SPwd10l9ioicghuu+vfUFcukRW8ywbZeV2FBwpiMlGzH9TeBa4Aves8XA79JSYlM+vQmf5P1SRiT0ZIKEqoaB+72biZTtSbp60mQaK4FXw4E8lJTJmNMWiV7ncR44H+AyUCodbmqHpyicpl06E268NY04bu7rIwxGSTZPonf4WoRUeDjwB+AP6aqUCZN2oJEddfbJQpbcj9jMlmyQSJPVf8JiKp+6F0Ad2bqimXSItTLPgkb/mpMxkq24zrspQlf46Xa2AgUpq5YJi1yciEn1LMhsM2WAdaYTJZsTeIGIB+4HjgKuAK4MlWFMmnU00yw4RoLEsZksG5rEt6Fcxer6teBeuCqlJfKpE9PM8GG66xPwpgM1m1NwpsI6Ph9UBazP+hpTaK51vokjMlgyfZJLBeRhcCfgbbEPqr6aEpKZdKnJ0FC1WoSxmS4ZINECNgBfCJhmQIWJDJNqBiqP0pu22gzxFusT8KYDJbsFdfWD5EtelKTsJQcxmS8ZK+4/h2u5rAHVf18N/vNwk1z6gd+o6q3tVv/U9zFeeBGTx2gqqXeuiuBb3vrfqCq9ydTVtNHoZLkh8C2pQkvSV15jDFplWxz0xMJj0PAecCmrnbwRkXdBZwGVAJLRGRh4jSkqvqVhO2/DEz3Hg8CvgtU4ILTMm/fXUmW1/RWqMQ1I7U0QyDU9bZts9JZTcKYTJVsc9Mjic9F5EHgxW52mwGsVdV13j7zgXOAzuaqvhQXGABOBxar6k5v38XALFyKcpNKrf0L4doeBAnrkzAmUyV7MV1744EDutlmBLAh4Xmlt2wvInIQMBb4V0/3Nf2sJ5lgrU/CmIyXbJ9EHXv2SWzBzTHRXy4BFnjXZCRNROYAcwBGjx7dj8XJYj3JBGtTlxqT8ZJtburNT8WNwKiE5yO9ZR25BLi23b4nt9v32Q7KdS9wL0BFRcVeHeumF9qS/CWRCdamLjUm4yXV3CQi54lIScLzUhE5t5vdlgDjRWSsiARxgWBhB689ESgDXklY/BTwSREpE5Ey4JPeMpNqPZnC1Dqujcl4yfZJfFdV284aqlrN7k7mDqlqFLgOd3JfBTysqu+IyK0icnbCppcA81VVE/bdCXwfF2iWALe2dmKbFGsNEskMgw3XQiAf/IHUlskYkzbJDoHtKJh0u6+qLgIWtVv2nXbPb+lk37nA3CTLZ/pLT2oSzTbhkDGZLtmaxFIRuUNExnm3O4BlqSyYSZNAPog/+dFN1h9hTEZLNkh8GYgADwHzgWb27Gg2mUIk+dQcNnWpMRkv2dFNDcBNKS6L2V+ESpIfAmvDX43JaMmOblosIqUJz8tExEYbZapQkhMPWZpwYzJess1NQ7wRTQB4OZS6u+LaDFQ9am6y5H7GZLJkg0RcRNouaRaRMXSQFdZkiGQzwVpNwpiMl+wQ2P8HvCgizwECnICXDsNkoGRqEvG4CxLWJ2FMRku24/rvIlKBCwzLgceAplQWzKRRbhJBIlIPqNUkjMlwySb4+wJwAy6H0grgWFwajU90tZ8ZoEIlLgjEouDv5E/E0oQbkxWS7ZO4ATga+FBVP46bHCiJDHBmQEomNYelCTcmKyQbJJpVtRlARHJV9V3g0NQVy6RVWybYLpqcLE24MVkh2Y7rSu86iceAxSKyC/gwdcUyaZVM/iZLE25MVki24/o87+EtIvIMUAL8PWWlMumVVHOTF0AsSBiT0ZKtSbRR1edSURCzH+lRTcL6JIzJZL2d49pkslzrkzDGOBYkzN6SrkkIBAv3SZGMMelhQcLsLbcYkK4zwYZr3XYi+6xYxph9L6VBQkRmich7IrJWRDpMNS4iF4nIShF5R0TmJSyPicgK77bX3NgmhXw+19fQXU3CmpqMyXg97rhOloj4gbuA04BKYImILFTVlQnbjAduBmaq6i4RScws26Sq01JVPtON7vI3NddYp7UxWSCVNYkZwFpVXaeqEdyMdue02+Zq4C4v9Tiqui2F5TE90V0m2NbmJmNMRktlkBgBbEh4XuktSzQBmCAiL4nIqyIyK2FdSESWesvP7egNRGSOt83Sqqqq/i19tuuuJmFpwo3JCunuuM4BxgMnA5cC9yXMgHeQqlYAlwF3isi49jur6r2qWqGqFeXl5fuqzNkhtxiau0jPZVOXGpMVUhkkNgKjEp6P9JYlqgQWqmqLqn4ArMYFDVR1o3e/DngWl1TQ7CtWkzDGkNogsQQYLyJjRSQIXAK0H6X0GK4WgYgMwTU/rfPm0M5NWD4TWInZd0IlyQ2BNcZktJSNblLVqIhcBzwF+IG5qvqOiNwKLFXVhd66T4rISiAGfENVd4jIccA9IhLHBbLbEkdFmX0gVOwCQTzuhsQmikYg2mxBwpgskLIgAaCqi4BF7ZZ9J+GxAl/1bonbvAwcnsqymW6ESkDjbvKh9n0PrXmbrE/CmIyX7o5rs7/qKhNs26x01idhTKazIGE61lWSP5u61JisYUHCdKyrJH+WJtyYrGFBwnSsqyBhacKNyRoWJEzH2oJER30SNnWpMdnCgoTpWJfNTdYnYUy2sCBhOpZMx7U1NxmT8SxImI7lBCGQD+FO+iT8QcjJ3fflMsbsUxYkTOdyizsf3WRNTcZkBQsSpnOdJfkL19rwV2OyhAUJ07nOgoSlCTcma1iQMJ3rLBOsNTcZkzUsSJjOhTrrk7A04cZkCwsSpnPWJ2FM1rMgYToXKnEBQXXP5dYnYUzWsCBhOpdbDDFvgqFWqjZ1qTFZJKVBQkRmich7IrJWRG7qZJuLRGSliLwjIvMSll8pImu825WpLKfpREepOVoaQWPWJ2FMlkjZzHQi4gfuAk4DKoElIrIwcRpSERkP3AzMVNVdInKAt3wQ8F2gAlBgmbfvrlSV13QgMUgUDXOPLU24MVkllTWJGcBaVV2nqhFgPnBOu22uBu5qPfmr6jZv+enAYlXd6a1bDMxKYVlNR0Kl7j5xGGxbmvCSfV8eY8w+l8ogMQLYkPC80luWaAIwQUReEpFXRWRWD/ZFROaIyFIRWVpVVdWPRTfA7s7pxOYmq0kYk1XS3XGdA4wHTgYuBe4TkdJkd1bVe1W1QlUrysvLU1TELNbW3FS9e1lrwj/rkzAmK6QySGwERiU8H+ktS1QJLFTVFlX9AFiNCxrJ7GtSrTVIhBOam6wmYUxWSWWQWAKMF5GxIhIELgEWttvmMVwtAhEZgmt+Wgc8BXxSRMpEpAz4pLfM7EsdzSlhU5cak1VSNrpJVaMich3u5O4H5qrqOyJyK7BUVReyOxisBGLAN1R1B4CIfB8XaABuVdWdqSqr6UQgD3wB65MwJoulLEgAqOoiYFG7Zd9JeKzAV71b+33nAnNTWT7TDZG9U3PY1KXGZJV0d1yb/V37TLDhOggWgs+fvjIZY/YZCxKma+0zwTbXWFOTMVnEgoTpWkfNTdbUZEzWsCBhutaaCbaVJfczJqtYkDBdy23f3GRpwo3JJhYkTNf2am6ymoQx2cSChOlaqNSlB4+1uOfWJ2FMVrEgYbrWlr/J65cI11mQMCaLWJAwXWvLBFsN8RhE6q1PwpgsYkHCdC1x4iFLyWFM1rEgYbqWmAnWUnIYk3UsSJiuJWaCtZqEMVnHgoTpWmJzk6UJNybrWJAwXeuwT8KChDHZwoKE6VqwEBBXi7A+CWOyjgUJ0zWfb3cm2LYgYX0SxmSLlAYJEZklIu+JyFoRuamD9bNFpEpEVni3LySsiyUsbz/tqdmXWlNzWJ+EMVknZTPTiYgfuAs4DagElojIQlVd2W7Th1T1ug5eoklVp6WqfKYHWjPBhutA/BDIT3eJjDH7SCprEjOAtaq6TlUjwHzgnBS+n0mV3JLdzU25RW5aU2NMVkhlkBgBbEh4Xukta+98EXlTRBaIyKiE5SERWSoir4rIuR29gYjM8bZZWlVV1Y9FN3tIbG6ypiZjskq6O67/CoxR1anAYuD+hHUHqWoFcBlwp4iMa7+zqt6rqhWqWlFeXr5vSpyNWoOEJfczJuukMkhsBBJrBiO9ZW1UdYeqhr2nvwGOSli30btfBzwLTE9hWU1XQsW7h8BakDAmq6QySCwBxovIWBEJApcAe4xSEpEDE56eDazylpeJSK73eAgwE2jf4W32ldaO6+ZqG/5qTJZJ2egmVY2KyHXAU4AfmKuq74jIrcBSVV0IXC8iZwNRYCcw29t9EnCPiMRxgey2DkZFmX0lVAIo1G6C8onpLo0xZh9KWZAAUNVFwKJ2y76T8Phm4OYO9nsZODyVZTM90Jqao3GH1SSMyTLp7rg2A0FiP4T1SRiTVSxImO611iTAahLGZBkLEqZ7iUEi8bExJuNZkDDdS7yAzmoSxmQVCxKme6HS3Y+tT8KYrGJBwnQv12oSxmQrCxKme/4cb/IhLHeTMVnGgoRJTmttwmoSxmQVCxImOa2jmnJtdJMx2cSChElOW5CwmoQx2cSChElOqBhyQpATTHdJjDH7kAUJk5xQidUijMlCKU3wZzJIxedhzAnpLoUxZh+zIGGSc9Bx7maMySrW3GSMMaZTFiSMMcZ0KqVBQkRmich7IrJWRG7qYP1sEakSkRXe7QsJ664UkTXe7cpUltMYY0zHUtYnISJ+4C7gNKASWCIiCzuYhvQhVb2u3b6DgO8CFYACy7x9d6WqvMYYY/aWyprEDGCtqq5T1QgwHzgnyX1PBxar6k4vMCwGZqWonMYYYzqRyiAxAtiQ8LzSW9be+SLypogsEJFRPdlXROaIyFIRWVpVVdVf5TbGGONJd8f1X4ExqjoVV1u4vyc7q+q9qlqhqhXl5eUpKaAxxmSzVAaJjcCohOcjvWVtVHWHqoa9p78Bjkp2X2OMMaknqpqaFxbJAVYDp+BO8EuAy1T1nYRtDlTVzd7j84BvquqxXsf1MuBIb9N/A0ep6s4u3q8K+LAPRR4CbO/D/vubTDseyLxjyrTjgcw7pkw7Htj7mA5S1U6bYlI2uklVoyJyHfAU4Afmquo7InIrsFRVFwLXi8jZQBTYCcz29t0pIt/HBRaAW7sKEN4+fWpvEpGlqlrRl9fYn2Ta8UDmHVOmHQ9k3jFl2vFAz48ppWk5VHURsKjdsu8kPL4ZuLmTfecCc1NZPmOMMV1Ld8e1McaY/ZgFid3uTXcB+lmmHQ9k3jFl2vFA5h1Tph0P9PCYUtZxbYwxZuCzmoQxxphOWZAwxhjTqawPEt1lqh2IRGS9iLzlZdZdmu7y9JSIzBWRbSLydsKyQSKy2MsKvFhEytJZxp7q5JhuEZGNCVmQz0hnGXtCREaJyDMislJE3hGRG7zlA/J76uJ4BvJ3FBKR10XkDe+YvuctHysir3nnvIdEpMuJ67O6T8LLVLuahEy1wKUdZKodUERkPVChqgPyIiARORGoB/6gqlO8ZT8GdqrqbV4wL1PVb6aznD3RyTHdAtSr6u3pLFtviMiBwIGq+m8RKcJd/Hou7lqnAfc9dXE8FzFwvyMBClS1XkQCwIvADcBXgUdVdb6I/Bp4Q1Xv7ux1sr0m0ZdMtSZFVPV53MWVic5hd26v+3H/wANGJ8c0YKnqZlX9t/e4DliFS8I5IL+nLo5nwFKn3nsa8G4KfAJY4C3v9jvK9iCRbKbagUaBf4jIMhGZk+7C9JOhrSlcgC3A0HQWph9d52VBnjtQmmbaE5ExwHTgNTLge2p3PDCAvyMR8YvICmAbLonq+0C1qka9Tbo952V7kMhUx6vqkcCngGu9po6Moa6NNBPaSe8GxgHTgM3AT9JbnJ4TkULgEeBGVa1NXDcQv6cOjmdAf0eqGlPVabgkqTOAiT19jWwPEhmZbVZVN3r324C/4P44BrqtXrtxa/vxtjSXp89Udav3TxwH7mOAfU9eO/cjwJ9U9VFv8YD9njo6noH+HbVS1WrgGeBjQKmXgBWSOOdle5BYAoz3evuDwCXAwjSXqU9EpMDreENECoBPAm93vdeAsBBonev8SuDxNJalX7SeTD3nMYC+J69T9LfAKlW9I2HVgPyeOjueAf4dlYtIqfc4DzdAZxUuWFzgbdbtd5TVo5sAvCFtd7I7U+0P01ykPhGRg3G1B3AJHOcNtGMSkQeBk3Epjbfi5jt/DHgYGI1LCX9Rd5mB9yedHNPJuGYMBdYD1yS05+/XROR44AXgLSDuLf4Wrh1/wH1PXRzPpQzc72gqrmPaj6sQPKyqt3rniPnAIGA5cEXCvD57v062BwljjDGdy/bmJmOMMV2wIGGMMaZTFiSMMcZ0yoKEMcaYTlmQMMYY0ykLEsbsB0TkZBF5It3lMKY9CxLGGGM6ZUHCmB4QkSu8HP0rROQeL4FavYj81MvZ/08RKfe2nSYir3rJ4f7SmhxORA4Rkae9PP//FpFx3ssXisgCEXlXRP7kXQVsTFpZkDAmSSIyCbgYmOklTYsBlwMFwFJVPQx4Dnc1NcAfgG+q6lTclbyty/8E3KWqRwDH4RLHgcs8eiMwGTgYmJnygzKmGzndb2KM8ZwCHAUs8X7k5+ES2MWBh7xt/gg8KiIlQKmqPuctvx/4s5dXa4Sq/gVAVZsBvNd7XVUrvecrgDG4iWKMSRsLEsYkT4D7VfXmPRaK/He77Xqb6yYxf04M+/80+wFrbjImef8ELhCRA6BtPueDcP9HrVk1LwNeVNUaYJeInOAt/yzwnDfrWaWInOu9Rq6I5O/TozCmB+yXijFJUtWVIvJt3Kx/PqAFuBZoAGZ467bh+i3ApWH+tRcE1gFXecs/C9wjIrd6r3HhPjwMY3rEssAa00ciUq+qhekuhzGpYM1NxhhjOmU1CWOMMZ2ymoQxxphOWZAwxhjTKQsSxhhjOmVBwhhjTKcsSBhjjOnU/wcNZsh00tf5uQAAAABJRU5ErkJggg==\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "# CNN model for five layers\n",
        "filename='cnn_fivelayer.csv'\n",
        "history_logger=tf.keras.callbacks.CSVLogger(filename, separator=\",\", append=True)\n",
        "cnn5_model=tf.keras.Sequential([\n",
        "tf.keras.layers.Conv2D(64,(5,5),input_shape=(100,100,3),activation='relu'),\n",
        "tf.keras.layers.MaxPooling2D(2,2), # First Convolution and Pooling Layers\n",
        "tf.keras.layers.Conv2D(64,(3,3),activation='relu'),\n",
        "tf.keras.layers.MaxPooling2D(2,2), # Second Convolution and Pooling Layers\n",
        "tf.keras.layers.Conv2D(32,(2,2),activation='relu'),\n",
        "tf.keras.layers.MaxPooling2D(2,2), # Third Convolution and Pooling Layers\n",
        "tf.keras.layers.Conv2D(16,(2,2),activation='relu'),\n",
        "tf.keras.layers.MaxPooling2D(2,2), \n",
        "tf.keras.layers.Conv2D(16,(2,2),activation='relu'),\n",
        "tf.keras.layers.MaxPooling2D(2,2), \n",
        "])\n",
        "cnn5_model.add(Flatten())\n",
        "cnn5_model.add(Dense(8, activation='relu', kernel_initializer='he_uniform',kernel_regularizer=regularizers.L1(0.0001)))\n",
        "cnn5_model.add(BatchNormalization())\n",
        "cnn5_model.add(Dense(1, activation='sigmoid',kernel_regularizer=regularizers.L1(0.0001)))\n",
        "cnn5_model.add(Dropout(0.3))\n",
        "opt = Adam(learning_rate=0.001)\n",
        "cnn5_model.compile(optimizer = opt,loss = 'binary_crossentropy' , metrics = ['accuracy'])\n",
        "# history = cnn_model.fit(x_train, y_train,validation_data=(x_val,y_val),epochs=10)\n",
        "history5 = cnn5_model.fit(x_train_normalize ,y_train,epochs=30,batch_size=32,validation_split=0.2,callbacks=[history_logger])\n",
        "loss_value , accuracy = cnn5_model.evaluate(x_test_normalize,y_test)\n",
        "print(loss_value,accuracy)\n",
        "plt.plot(history5.history['accuracy'])\n",
        "plt.plot(history5.history['val_accuracy'])\n",
        "plt.title('model accuracy')\n",
        "plt.ylabel('accuracy')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train','validation'], loc='upper left')\n",
        "plt.show()\n",
        "# summarize history for loss\n",
        "plt.plot(history5.history['loss'])\n",
        "plt.plot(history5.history['val_loss'])\n",
        "plt.title('model loss')\n",
        "plt.ylabel('loss')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'validation'], loc='upper right')\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "-cRayTc_9u31",
        "outputId": "7e29471b-c46a-46a6-9505-847303600e6f"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/resnet/resnet50_weights_tf_dim_ordering_tf_kernels_notop.h5\n",
            "94773248/94765736 [==============================] - 1s 0us/step\n",
            "94781440/94765736 [==============================] - 1s 0us/step\n",
            "Epoch 1/30\n",
            "978/978 [==============================] - 113s 108ms/step - loss: 0.4505 - accuracy: 0.8338 - val_loss: 0.6348 - val_accuracy: 0.8758\n",
            "Epoch 2/30\n",
            "978/978 [==============================] - 106s 108ms/step - loss: 0.2999 - accuracy: 0.8845 - val_loss: 0.4919 - val_accuracy: 0.8434\n",
            "Epoch 3/30\n",
            "978/978 [==============================] - 107s 109ms/step - loss: 0.2461 - accuracy: 0.9032 - val_loss: 0.3404 - val_accuracy: 0.8924\n",
            "Epoch 4/30\n",
            "978/978 [==============================] - 107s 109ms/step - loss: 0.2303 - accuracy: 0.9093 - val_loss: 0.3713 - val_accuracy: 0.8685\n",
            "Epoch 5/30\n",
            "978/978 [==============================] - 106s 109ms/step - loss: 0.2123 - accuracy: 0.9184 - val_loss: 0.4682 - val_accuracy: 0.8797\n",
            "Epoch 6/30\n",
            "978/978 [==============================] - 107s 109ms/step - loss: 0.1979 - accuracy: 0.9241 - val_loss: 1.1363 - val_accuracy: 0.8634\n",
            "Epoch 7/30\n",
            "978/978 [==============================] - 107s 109ms/step - loss: 0.1821 - accuracy: 0.9289 - val_loss: 1.4568 - val_accuracy: 0.4772\n",
            "Epoch 8/30\n",
            "978/978 [==============================] - 107s 109ms/step - loss: 0.1749 - accuracy: 0.9352 - val_loss: 1.3749 - val_accuracy: 0.6501\n",
            "Epoch 9/30\n",
            "978/978 [==============================] - 107s 109ms/step - loss: 0.1591 - accuracy: 0.9409 - val_loss: 0.3061 - val_accuracy: 0.8930\n",
            "Epoch 10/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.1524 - accuracy: 0.9448 - val_loss: 0.3273 - val_accuracy: 0.8969\n",
            "Epoch 11/30\n",
            "978/978 [==============================] - 107s 109ms/step - loss: 0.1429 - accuracy: 0.9479 - val_loss: 1.8537 - val_accuracy: 0.5037\n",
            "Epoch 12/30\n",
            "978/978 [==============================] - 107s 109ms/step - loss: 0.1376 - accuracy: 0.9519 - val_loss: 0.3588 - val_accuracy: 0.8436\n",
            "Epoch 13/30\n",
            "978/978 [==============================] - 107s 109ms/step - loss: 0.1268 - accuracy: 0.9545 - val_loss: 0.7640 - val_accuracy: 0.6775\n",
            "Epoch 14/30\n",
            "978/978 [==============================] - 107s 109ms/step - loss: 0.1174 - accuracy: 0.9591 - val_loss: 0.2253 - val_accuracy: 0.9344\n",
            "Epoch 15/30\n",
            "978/978 [==============================] - 107s 109ms/step - loss: 0.1053 - accuracy: 0.9640 - val_loss: 0.7440 - val_accuracy: 0.7335\n",
            "Epoch 16/30\n",
            "978/978 [==============================] - 107s 109ms/step - loss: 0.0981 - accuracy: 0.9673 - val_loss: 0.2585 - val_accuracy: 0.9131\n",
            "Epoch 17/30\n",
            "978/978 [==============================] - 107s 109ms/step - loss: 0.0878 - accuracy: 0.9709 - val_loss: 0.3244 - val_accuracy: 0.9076\n",
            "Epoch 18/30\n",
            "978/978 [==============================] - 107s 109ms/step - loss: 0.0797 - accuracy: 0.9759 - val_loss: 0.7220 - val_accuracy: 0.8981\n",
            "Epoch 19/30\n",
            "978/978 [==============================] - 107s 110ms/step - loss: 0.0696 - accuracy: 0.9785 - val_loss: 0.6865 - val_accuracy: 0.7707\n",
            "Epoch 20/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.0616 - accuracy: 0.9815 - val_loss: 0.3662 - val_accuracy: 0.9004\n",
            "Epoch 21/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.0560 - accuracy: 0.9839 - val_loss: 0.3666 - val_accuracy: 0.8762\n",
            "Epoch 22/30\n",
            "978/978 [==============================] - 107s 110ms/step - loss: 0.0504 - accuracy: 0.9861 - val_loss: 4.2620 - val_accuracy: 0.4211\n",
            "Epoch 23/30\n",
            "978/978 [==============================] - 107s 109ms/step - loss: 0.0530 - accuracy: 0.9852 - val_loss: 1.0260 - val_accuracy: 0.6842\n",
            "Epoch 24/30\n",
            "978/978 [==============================] - 107s 109ms/step - loss: 0.0430 - accuracy: 0.9896 - val_loss: 0.3579 - val_accuracy: 0.9039\n",
            "Epoch 25/30\n",
            "978/978 [==============================] - 107s 110ms/step - loss: 0.0414 - accuracy: 0.9896 - val_loss: 1.0958 - val_accuracy: 0.6787\n",
            "Epoch 26/30\n",
            "978/978 [==============================] - 107s 110ms/step - loss: 0.0414 - accuracy: 0.9890 - val_loss: 0.6284 - val_accuracy: 0.9076\n",
            "Epoch 27/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.0354 - accuracy: 0.9909 - val_loss: 0.5224 - val_accuracy: 0.8496\n",
            "Epoch 28/30\n",
            "978/978 [==============================] - 107s 110ms/step - loss: 0.0379 - accuracy: 0.9910 - val_loss: 0.7991 - val_accuracy: 0.7848\n",
            "Epoch 29/30\n",
            "978/978 [==============================] - 107s 110ms/step - loss: 0.0332 - accuracy: 0.9924 - val_loss: 0.4311 - val_accuracy: 0.8871\n",
            "Epoch 30/30\n",
            "978/978 [==============================] - 107s 109ms/step - loss: 0.0307 - accuracy: 0.9920 - val_loss: 0.6140 - val_accuracy: 0.8560\n",
            "524/524 [==============================] - 17s 32ms/step - loss: 0.6518 - accuracy: 0.8433\n",
            "Test_loss_value = 0.6517839431762695\n",
            "test_accuracy = 0.8433138728141785\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYMAAAEWCAYAAACEz/viAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO2deXxcdbn/30/2TJZm676Xtlgoe9lBQUQWZRNFELzu9briel2uonL1qj+9XtTrhsq9uAByEZCryKagohSosnahLd2StE3SJpk0+zLf3x/fc5LpdCY5M3PObHner1deM3PmzJnvZJLzOc8uxhgURVGU6U1RthegKIqiZB8VA0VRFEXFQFEURVExUBRFUVAxUBRFUVAxUBRFUVAxUKYZIvI/IvJlj/vuFJHXBL0mRckFVAwURVEUFQNFyUdEpCTba1AKCxUDJedw3DOfFJHnRaRPRH4qIrNF5PciclBEHhGR+qj9LxWRDSLSLSKPiciqqOdOEJF/OK/7FVAR816vF5Fnndf+TUSO9bjG14nIMyLSIyLNIvLFmOfPco7X7Tz/dmd7pYj8h4jsEpGwiDzubDtHRFri/B5e49z/oojcJSK/EJEe4O0icoqIPOG8x14R+S8RKYt6/dEi8rCIdIpIm4h8VkTmiEi/iDRG7XeiiHSISKmXz64UJioGSq5yJXA+sBK4BPg98FlgJvbv9sMAIrISuB34iPPc/cD/iUiZc2K8F/g50AD8r3NcnNeeANwCvBdoBH4E3Cci5R7W1wf8E1AHvA54n4hc7hx3sbPe7zprOh541nndN4GTgDOcNf0LEPH4O7kMuMt5z18CY8BHgSbgdOA84P3OGmqAR4AHgHnAcuAPxph9wGPAVVHHfStwhzFmxOM6lAJExUDJVb5rjGkzxrQCfwGeNMY8Y4wZBO4BTnD2ezPwO2PMw87J7JtAJfZkexpQCtxkjBkxxtwFPB31HmuBHxljnjTGjBljbgWGnNdNijHmMWPMC8aYiDHmeawgvcp5+i3AI8aY2533PWCMeVZEioB3AtcbY1qd9/ybMWbI4+/kCWPMvc57Dhhj/m6MWWeMGTXG7MSKmbuG1wP7jDH/YYwZNMYcNMY86Tx3K3AdgIgUA9dgBVOZxqgYKLlKW9T9gTiPq53784Bd7hPGmAjQDMx3nms1h3Zj3BV1fzHwccfN0i0i3cBC53WTIiKnisijjnslDPwz9god5xgvx3lZE9ZNFe85LzTHrGGliPxWRPY5rqN/97AGgN8AR4nIUqz1FTbGPJXimpQCQcVAyXf2YE/qAIiIYE+ErcBeYL6zzWVR1P1m4CvGmLqon5Ax5nYP73sbcB+w0BgzA/gh4L5PM3BEnNfsBwYTPNcHhKI+RzHWxRRNbIvhHwCbgRXGmFqsGy16DcviLdyxru7EWgdvRa0CBRUDJf+5E3idiJznBEA/jnX1/A14AhgFPiwipSLyBuCUqNf+GPhn5ypfRKTKCQzXeHjfGqDTGDMoIqdgXUMuvwReIyJXiUiJiDSKyPGO1XIL8C0RmScixSJyuhOj2AJUOO9fCnwOmCp2UQP0AL0i8grgfVHP/RaYKyIfEZFyEakRkVOjnv8Z8HbgUlQMFFQMlDzHGPMS9gr3u9gr70uAS4wxw8aYYeAN2JNeJza+cHfUa9cD7wH+C+gCtjn7euH9wI0ichC4AStK7nF3AxdjhakTGzw+znn6E8AL2NhFJ/B1oMgYE3aO+ROsVdMHHJJdFIdPYEXoIFbYfhW1hoNYF9AlwD5gK3Bu1PN/xQau/2GMiXadKdMU0eE2ijI9EZE/ArcZY36S7bUo2UfFQFGmISJyMvAwNuZxMNvrUbKPuokUZZohIrdiaxA+okKguKhloCiKoqhloCiKokDeNbtqamoyS5YsyfYyFEVR8oq///3v+40xsbUr4+SdGCxZsoT169dnexmKoih5hYhMmkKsbiJFURRFxUBRFEUJUAxE5BYRaReRFxM8LyLyHRHZJrZv/YlBrUVRFEWZnCBjBv+DLfP/WYLnLwJWOD+nYptunZpg30kZGRmhpaWFwcHBVF6eN1RUVLBgwQJKS3UGiaIo/hKYGBhj/iwiSybZ5TLgZ0574XUiUicic40xe5N9r5aWFmpqaliyZAmHNqgsHIwxHDhwgJaWFpYuXZrt5SiKUmBkM2Ywn0P7s7c425JmcHCQxsbGghUCABGhsbGx4K0fRVGyQ14EkEVkrYisF5H1HR0difbJ8Koyz3T4jIqiZIds1hm0YoeQuCxwth2GMeZm4GaANWvWaP8MRVFyAmMMxtipQ5Hx+/Z2YHiM/pEx+odG6R8eo294lP6hiW19w2MMDI8yPBqhqEgoFrG30fcFiotk/Pk1SxpYPqt6ynWlQjbF4D7ggyJyBzZwHE4lXpALdHd3c9ttt/H+978/qdddfPHF3HbbbdTV1QW0MkXJb4wxtB8cYntHHzsP9LG7s5/h0cj4ODcRazGP28wCgiACRQLFRUWUFgklxUWUFgslh9wvoqTYnnz7h8foGRihZ3DUuR2hZ2CUg4OHbhsYHrMnfSAbbd2+fPnq/BMDEbkdOAdoEpEW4AvY4eQYY34I3I8dALIN6AfeEdRagqa7u5vvf//7h4nB6OgoJSWJf8X3339/0EtTlIwxPBqhe2CYcP8IXf0jhAdGEKCspGjip7iI8pjHZSVFDI1G2Lm/j+37+9i53574d+zvZ9eBPvqHx8bfo7RYKC8ptlfkMH4lzvh9xoeDjhnDWCT5M3ZNRQm1FaXUVpZSW1HCgvpKaufWUlNRQqismCKRQ0TICs/EfXGerywtJlRWTKishKryYipL7W2ozB6nqqyEyrJiykqKiETM+Hoj7m2Ew7bVVgaXSRhkNtE1UzxvgA8E9f6Z5NOf/jQvv/wyxx9/PKWlpVRUVFBfX8/mzZvZsmULl19+Oc3NzQwODnL99dezdu1aYKK1Rm9vLxdddBFnnXUWf/vb35g/fz6/+c1vqKyszPInU6Yr/cOjdPYN09U3Qmf/MF19w3Q6P139w3T3j9A94Nz2j9DdP0xf1Ek7HUqKhIUNIZY2VXH6skaWNoVY0lTFksYq5tVVUlzkPXZmjGE0YhgdM4xEIoyOGUbHIodsG4sYKkuLqa0spbq8JKnj+0VRkVCEUFqc8bceJ+96E03Fl/5vAxv39Ph6zKPm1fKFS45O+PzXvvY1XnzxRZ599lkee+wxXve61/Hiiy+Op4DecsstNDQ0MDAwwMknn8yVV15JY2PjIcfYunUrt99+Oz/+8Y+56qqr+PWvf811113n6+dQpjdDo2N0HByirWeI9p5B2g8O0dYzaB8fHHRO/sN09g8zOBKJe4wigbpQGXWhUupDZcypreDIOTXUh8qoqyylLlQ6/vwM5yp2eDTC8GiEobHI+P3h0QjDUY9Li4XFTVUsbaxiQX0lJcX+5LaICKXF9iRbSRbPtHlAwYlBLnDKKaccUgvwne98h3vuuQeA5uZmtm7depgYLF26lOOPPx6Ak046iZ07d2ZsvUr+MjgyxoG+YQ70DnGgd5j9vUOHPO7oHaLdOdl39Y8c9vriImFWTTmzasqZXVvBK+bU0lhdRn2ojIYqe8KfeFxGbUUpRVm4claCp+DEYLIr+ExRVVU1fv+xxx7jkUce4YknniAUCnHOOefErRUoLy8fv19cXMzAwEBG1qrkNsYY9oYH2drey9a2g2xr7+Xljl7aeoY40DuU0DVTXlJEU3U5TdVlLGoMsWZJPbNrK5hdW86smgpm1dqTf0OoTE/uClCAYpANampqOHgw/vTAcDhMfX09oVCIzZs3s27dugyvTskHRscitHYPsLWtl20dvfa23Z78o0/4jVVlHDGrmhMX1dFQVU5jdRlN1WU0VpXTUF1Gk7MtVFasdSlKUqgY+EBjYyNnnnkmq1evprKyktmzZ48/d+GFF/LDH/6QVatWceSRR3LaaadlcaVKtugbGmVP9wAt3QPs6R6gtcu5de7v6xkkOvFldm05y2dV86Y1C1k+q5oVs6pZPquaxuryxG+iKGmQdzOQ16xZY2KH22zatIlVq1ZlaUWZZTp91nyhf3iUveFB9oUHnduBQx7vCQ/QHeOvLykS5syoYF5dJQvqKplXV8nChkqWz6ph+azq8eCroviFiPzdGLMm0fNqGSjKFBhjONA3PO7C2dZ2kJ0H+p2T/QA9g6OHvaahymbazJlRwYmL65hfF2JeXQUL6u2Jf1ZNRVZSGBUlESoGiuJgjGFfzyBb23rZ2t7LtvYJv310Jk5VWTFLZ1axqDHEqcsa7BX+jErmzKhg7owKZtdWUJHNhHFFSQEVA2XacXBwhB37+9ixv4/tHbbqdcf+XnZ09B0SrK0LlbJiVjUXrp57iN9+7owKDc4qBYeKgVJQjEUMXf3DHOi1ufb7+4bZFx5gx/4+Xu6wAtBxcGh8fxFYUF/JsqZq1ixu4IjoYG1VmZ70lWmDioGSFxhj6OwbZndnP81dAzR39tNxcKLAar9TZNXZPxy3gVhjVRnLZlZx7pEzWdpUzbKZVSxrqmJhQ0hdOoqCioGSQxwcHGFveJDdB/pp7uqnuXOA3Z39tHT1s7uz/5CGZQA15SU01ZTTWFXG0qYq1ixpoKmqzNk2kYM/s6ZCs3MUZQpUDHwg1RbWADfddBNr164lFAoFsLLcYHQsQkev7YmzLzxIW88g+3oGaQvbW/d+bDVtqKyYhfUhFjaEOP2IRhbWh1jUYB8vqK+kqlz/fBXFL/S/yQcStbD2wk033cR1112X12IQiRj29w6NX803d05c2Td39bM3PHhYK+GSIhlvj/CKOTW8auVM5tRWMLeukoX1lSxsCKnPXlEyiIqBD0S3sD7//POZNWsWd955J0NDQ1xxxRV86Utfoq+vj6uuuoqWlhbGxsb4/Oc/T1tbG3v27OHcc8+lqamJRx99NNsfZVJGxiK83NHLxj09bNrbw5a2Xpq7+mntGmBo9NAul03V5SxsqOTERfUsbKhk7ozK8bz72bUVNFZpTxxFySUKTwx+/2nY94K/x5xzDFz0tYRPR7ewfuihh7jrrrt46qmnMMZw6aWX8uc//5mOjg7mzZvH7373O8D2LJoxYwbf+ta3ePTRR2lqavJ3zWkS7h9h494eNu61J/5Ne3vY2tbL8Jg96ZeVFLF8ZjUrZ9Vw3itmsbAh5Lh0KplfF6KyTIOyipJPFJ4YZJmHHnqIhx56iBNOOAGA3t5etm7dytlnn83HP/5xPvWpT/H617+es88+O8srnSDcP8Lzrd0819zNcy1hNu7pobV7omtqU3UZq+bW8o4zl3DUvFpWza1lWVOVbz3nc4Ztf4C5x0NV49T7KkqBUXhiMMkVfCYwxvCZz3yG9773vYc9949//IP777+fz33uc5x33nnccMMNGV/f4MgYG/b08FxzN8+32JP/jv19488va6rixMX1XHfaYlbNreGoebXMqqnI+DozzlAv/PKNcPbH4dWfy/ZqFCXjFJ4YZIHoFtYXXHABn//857n22muprq6mtbWV0tJSRkdHaWho4LrrrqOuro6f/OQnh7zWbzeRMYa2niFe7rD97zfvO8jzLd1s3nuQUSeYO7u2nOMW1PHGkxZw3II6jlkwY/qmYIabwUSg46Vsr0RRsoKKgQ9Et7C+6KKLeMtb3sLpp58OQHV1Nb/4xS/Ytm0bn/zkJykqKqK0tJQf/OAHAKxdu5YLL7yQefPmpRRAHhodY+f+fl7u6GV7Ry8vd/RZAYjpg19TUcKxC2aw9pXLOG5hHcctqGPOjGlwxe+VcIu9PfBydtehKFlCW1jnCe5g7xc3bORvByrYsMf69nd39h/SB3/ejAqOmFXNETOrOWJmlb2dVc2smnJN05yM9bfAbz8KJRXw2b1QVGDxEGXaoy2s8xBjDEOjEQZHxhgYGWNgeIzBkQijkQj7e4f5xoO7WNQQYtXcGi49bt74yX9pU5UWYqWKaxmMDlqXUf3i7K5HUTKMnjlyhOHRCAcHR+gZHKVvaJSIY7GJCBUlRdRUlFBZVkykq5znv/haaiumqW8/KFwxADiwTcVAmXYUjBgYY/LKDWKMYWBkjJ6BUXoGRxgcsf798pIi6qvKqCwtprK0mPLSIoqcz2WMoaOkSIUgCLqboXG5FYID22D5ef4cNzIGUmTboypKDlMQYlBRUcGBAwdobGzMaUEYixh6h0Y5OGAtgNFIBAFCZSXMnVFBTUUp5SVFcT+DMYYDBw5QUaFB30AIt8Ci06C33YqBH4wOw03HwOgAzF4Ns492fo6BWa+Asip/3kdRfKAgxGDBggW0tLTQ0dGR7aUcwljEMDwWYWQ0wvBYhKHRCMZAkUBFaTEVpUVUlBQzXCTsB/ZPcbyKigoWLFiQiaVPLyJj0NMKdQutdbB/qz/HPbAVevfBsnNgZACevQ2Ge50nBRqPcMRhtf2ZdzzUzvPnvRXvDHRBRd20t94CFQMRuRD4NlAM/MQY87WY5xcDtwAzgU7gOmNMy2EHmoLS0lKWLl3qw4pTp61nkOdbwrzQGuaFlm5eaA2zv3cYgOIiYcWsas44oonXrJrFyUsbKC206t185uA+MGMwY4EVg91P+HPc9k329oJ/tyf9SAS6d0HbBufnBds6ZeN9gJMSNvc4eMXr4RWvg1lHTfsTVOAMhuFbR8E5n4EzP5zt1WSVwMRARIqB7wHnAy3A0yJynzFmY9Ru3wR+Zoy5VUReDXwVeGtQa/KbR19q55frdvF8S5h2Z3pWkcCKWTWcc+Qsjl0wg2Pmz2DV3FodoJLLhJvt7YxF0LQfXrjTXsmXVqZ33PZNIMVWYMCmqzYstT+rXj+x31Cv3XfXX+Gl++HRf4dHvwL1S+DI11lhWHQaFE3zv6HRYbj9anjlJ2Hx6f4c88DLMNIPf/kmnPhWqKz357h5SJCWwSnANmPMdgARuQO4DIgWg6OAjzn3HwXuDXA9vrEvPMiNv93A/S/sY35dJWctb+KYBTM4doE98YfKCsL7Nn1wM4lmLIChHnv/wMswZ3V6x+3YbF1BJeWT71deDQtPtj9nfQQOtsGW38Pm38HTP4Z134NQI6y8yArDEeemL1TRRCLwp6/B4jOsSytX6dwOL/8B5h7rnxh07bS3g2H423fhvMy3iMkVgjxrzQeaox63AKfG7PMc8AasK+kKoEZEGo0xB6J3EpG1wFqARYsWBbbgqRgdi3DrE7v41kMvMRoxfPKCI3nP2csoK1GXz2EM9UL3bvtzcC8ceTHUzM72quIzbhnMhzHr2uPA1vTFoH1TaseomQ0nvd3+DB2EbY9YYdj0f/DsL6A0BNfcActeld76XDo2w5++brOeXvNFOOPDueme6trh3O708ZjOsVZeCOt+AKf+M1TP8u/4eUS2L2E/AfyXiLwd+DPQCozF7mSMuRm4GWwFciYX6PLM7i7+9Z4X2bi3h3OOnMmNl65mUWP+DqRJG2NsoLV7l/2Hck/83bvsbf+BQ/fv2QOv/tesLHVKwi02gFheY6/kIf2MopEBeyV7zJvSO055DRx9hf0ZHbaupLveaYPRfomBGyNZ+kp4+AbY+xxc+t3cy3bq3HHorR907YRQk43r/NfJ8JdvZb3ZZbYIUgxagYVRjxc428YxxuzBWgaISDVwpTGmO8A1JU14YIRvPLiZXz65m1k15fzg2hO5cPUcm/654R544vvw+m/ZmQfTiaduht//y8Tj4nKbjVO32LaBrltkC7fqFsMv3wR9uZXpdQjhFrt2sCfA2vmwP00x2L8FMDaF1C9KyqyLaMlZsPtv/h23+Umong1vvRf+ehM88iXbsO/qX9q4Ra7Qud3e+m0Z1C+xFwEnXAvrfwqnf2Di72EaEaQYPA2sEJGlWBG4GnhL9A4i0gR0GmMiwGewmUU5gTGG3zy7hy//biOdfcO844ylfOy1K6l22z307be9bAa64CfnwyXfhuPenN1FZ5J9L1g/9tW32xN/9ezE/XyqZh5uKeQS3THtJxqXWzdROrRvtrczA+iZtfgM2HQfhFutaytddj8BC0+1rqGzPmovbO56J9x8DrzxFjji1em/hx+4bqLBbvt/50ewt3sXzHfa9bzqU/DcHdZldtl/pX9sLxhjf3KgF1ZgKzDGjAIfBB4ENgF3GmM2iMiNInKps9s5wEsisgWYDXwlqPUkw/aOXq79yZN85FfPMr8+xH0fPIsbLjlqQggAHvq89ee+7f9g/klwz1q4/19gbCR7C88k3buh4QhYdCrUzp38jznUCP2dmVtbsoRbbPDYpWmFdROl08SxYxMUlU64nfxkkRM89SMFNtxqv8tFUQHZ5a+BtY9BzVz4xZXw12+n97vwi84dNl4C/lgHY6POhcAS+3jGAjj53dYF51etyVQ89jX49rHpW6I+EKgcGWPuN8asNMYcYYz5irPtBmPMfc79u4wxK5x93m2MGQpyPV5o7xnksu/9lRdaw3z58tXc/b4zWD1/xqE77fgLPHcbnPEh62f9p3vh9A/CUz+CWy+xeeuFTvduaxF4IdSQu5bBYBiGwoeKQeNyu71vqjLASWjfZEWlOIDWIXOOgbIa2OWDq6h5nb1ddNqh2xuWwbsehlWX2DjCr98Fw32Hvz5TjI3aq/glZ9nHfsQNelpsfUm0K+ysj9nOtY/+e/rHnwpj7Hkk3GzPG64bLEtk3zbJMb76+80MjUS49wNnct1piymOHdo+OgS/+5j1hb/S8ZkXl8IFX4Erf2qDbz96Fex+Mvk3HxuFrQ/D0z+xVkeuEhlz/OxJiMFAjloGYSeMNSPKR9y4wt6m4ypq3wQzfYwXRFNUDAtP8ccy2L0OSqtgzrGHP1deDW+6Fc77Arx4N/z0An/99cnQ0wKRUVh2rn3sxzrcY0SLQfVMOP39sOFu/2epx9Kx2V5UnfJe27Lk1kvt4ywxfcRgbBRa/zHpLk/t6OSeZ1pZ+8plHDGzOv5Of/2ODQ5e/E0oi8kmOuaN8O5H7Pb/uRie+rE387ptAzz4r/CfR9nRi7/7OHz7eHjyZptBkmsc3AeRkSTEoNFaBrngaohlPK00SgyanCKxVDOKhvvsVeysAGdsLD4d2jem737b/QQsWAPFCcKHInD2x+DauyC828YRtj+W3numgmsJzDnGZv90+WAZxBMDsFZ+xQz445fTf4/J2PKAvT3rIzZ4P9hjLYRw6+SvC4jpIwZ/+jrcckHCP+TRsQg3/OZF5tdV8oFzl8c/xoGX4c/fgKMug5Wvjb/P7KPhPY9av+v9n4B732fTDGPp22/zmn94NvzgDHjyhzaQ9eZfwDsfsieS338SvncyvHCXLQzKFdyrl2TEIDI6UdCVS4yLQZSbaMZCmx2Vqt/YHZ0ZlGUAsOgMe9ucggXqMthjL0QWeSjgWvEa+3ddNRPuelfq75kqrgvFreD2yzIoKj28H1RlHZz5EXuyTsXC98qWh6xFVjvP9qV6693QdwB+dmlWXM3TRwxOe5/1Bd/+FmhZf9jTP1+3i837DvL516+isixO2b8x9uReXAYXfn3y96qss1k253zWZif89Hz7hzc6bPvQ3H4N/MeR8MCnbaHPhV+Hj78E19xmfbSLTrWB6Wt/DWXV1l/743Pg5eTHYgbCuBh47PkfarS3uRg3CLfYE0J1VEFcUbH1madqGbg9iWYdlf76EjH/JPu3mE7coOVpO/c5Nl6QiMYjbN1E//7MJ0p07bACXTPPXsl37vThmDttCmm8Nh+nvheqZsEfbgzGou3vtPGalRdObFuwBq79X+jZCz+7LL2YVQpMHzEINcBb77HVhb+4Eva9OP5Ux8EhvvXQFs5e0cQFR8+J//oNd8PLf4TzPm+zZ6aiqAjO+RS85U578vzRK+E/VsKdb7XuqtPeD+97At77Jzjtn6Gq6dDXi9irsff+Ba64Gfq74OeXw88uhz3PpvGL8AFXDGZ47KBa2WBv+7uCWU86hFvslVlsNlTT8tTFoGOTPXE1BNg8sbQC5p2YXtxg9zrbO2lBwkmIh1NRZ28HMlwO1LnDikBREdQvtTGEdF2obo1BPMqqbA+kXY/D9gAuwrb9wQpxtBiAdf+95Q67tp9dltEsvOkjBgA1c+CffmPT035+xfjw86/+fhODo2N86dKj489DGOiGBz5ji6lOfndy77nytTZNb8EpNl/72l/DRzfAa/8NZnu4ciwqsvULH1oPF3zVBqhvfhX8+t3+VmImQ/cueyVd6nG2Qq5bBvHcXY3L7e93bDT5Y7ZvhqaVwTeWW3w67HkGhvtTe/3uJ2y7jPIa76+pdMUgw8LeuWNCXOuX2BNpuHnSl0zJZGIAcNLbbPPCIKyDLQ9Yl9u8Ew5/bukr4erbbGzy55dnTHinlxiALS76p3utD/tnl/Pchg3c/Y9W3n32MpYlChr/8cu2gvaSm1L7B29YBtfdZQt4VrwmcbBuMkrKbZbD9c/C2Z+ATb+15fNbH07+WOmSTFopWKsMclMMupvjWziNK2yQvHtX8sfs2Oxv5XEiFp1h/45bD3d7TsnYiHWXeokXRONaBoMZtAyMsSfuhmX2sSsK6QSRB5zCtcnEoKTcWvd7nrG9ofxibBS2PQwrXpu4Pmf5eTZ+2LbRejIGg4+3TT8xAJh5JLz1bsxAF/W/voqjaof40KsTBI1b/m5TPU9ZG1/FM03FDOuq+vAz9o912yOZX0PSYpCjlsHYKBzck0AMUswoGuyxV6xBZhK5LDwFENiVgqto7/M2ndFrvMClMgtuot52GOmz7iGYOIGnYxm7Ij9Vu41jr7YXBn/8sk2p9oPmJ20dy8oLJt9v5QXwpv+2YnTbVbb5Y4BMTzEAmHcCDx77bWaNtXF76BuEInEKasZG4bfXW/fSuTnWZK12bnbaPCRbYwBWwKQ492oNDu617oZ4YtDk1Bokm1E0nkmUATGorLPZa6n0KXJjDQuTFINsWAbRmUQA1XNsYVg6GUWJ0kpjKS6xDRY7NtmsPj/Y8oBNWnBrJiZj1SVw5U+sgNx+deouQQ9MWzHY3zvEJ9dX892mL1DbswVui/OLfupmW3hy4degojY7C52MqqbMi0GyNQZgg+G5WIU8PscgTlOyUIMNfCdbeNbhZhJlwE0E1s3T/HTysY3mdcumdmAAACAASURBVPZE6CUZIhq3H1AmLQPXHeS6iYqKbCZbWmLg0TIAWHWZTQF97N/9qfvZ+hAsOdP7OWX1G+DyH8LOx20KekBMWzH4+u83MzgyxhVvfgfyhpvtldKd/zTxZYdb7bSpFa+1dQW5iFvMlUmSrTFwycZapyJewVk0jcvHkww8074ZSiqhbklaS/PM4tOtC2Xfc95fY4zNJEo2XgATbqKMWgY7bAp29PeUbq1B104rbBUzptyVoiI79KZrJzzz89TfE+xn6dh8eBbRVBz3ZnjbfXbWREBMSzH4+64u/vfvLbzzrKUsn1UNq6+0weFtD9uGc5ExeOBT9vbib+TmoA/ITgO4ZGsMXHKxWV30UJt4NK1I3k3UvhFmrsxcF0q3+CyZuEHndpsQkWy8AGzrldKqzFoGndutK6+kbGJb/RJ7Yk01y2eqTKJYlr/Giuefv2Fb0qTK1ofs7YoERauTsfSVqSWfeGTaicFYxPCF+15kTm0FH371ioknTno7nP9vdkbBrZfYqVKv+pfc6uceS6gh44UpSdcYuFTW56AYtFhXUKIhLo3LoXdfcn2iOjYHW2wWS+1c+zeaTL1BqvECl8q6zFoGXTsmXEQu9UutRZTq33+yYiBizwcH98Lzv0rtPcHGCxpXBNPNNk2mnRjc9tRuXmzt4V9ft4qq8hiVPfPDNm1z119tAPCMD2VnkV4JNdqMkACDSocxXmOQ5AxeP91EkQj88Kz0A3rRQ23ikWxG0UC3PVkE2YYiHovOsCd4r1fJu5+w4ty0MrX3q6jLsGWwYyKTyMU9kaeSXhoZsxc1yV7oLTsX5h5nW3qnklk0dND6/afKIsoS00oMDvQO8c0HX+KMIxp5/bEJAmev/hxc/gOb4xtE+2E/yUbKZrg5+XgB+NusbqDTBvbdRl+pEm5JHC+AqIwij2LQ4Qy0yURaaTSLT7e/2/1bvO2/e521ClJ1ZWXSMhjott93bDX3eK3BzuSP2bPHJkEkKwbu8J8D22Dzb5N/3+2P2RnbycYLMsS0EoP/98BL9A2NcuNlCSqNwX7hx79lonNlLpMNMUi2xsAl1Gh7xw+G019Db7u9bduQ+jGMSVxw5lK/FBDvlkH7RnubDcsAvPUp6u2wnyeVeIFLJi2D2EwiF/dvMBUxcF+TbNwLYNWldi2P/2fyFzZbHoTyGen97gNk2ojBM7u7+NX6ZidonET5fS6TaTGIROwJNCUxcKqQ/ag1cOcp79+SejBvMAzDBycXg9IK+1m9ppe2b7aNBSezNoKg8Qhbc7J73dT7ul1OU8kkcsmkZeAWlsW6iUorbdO6VArPvNYYxKOoGM683haC7fiz99dFIjZ4vPzVOetxmDZisGFPDwvqK/nweSum3jlfGBeDDAVme1OoMXDxc62uGERGJ1wzyTJeYzBFIDyZjKKOTba6PdPzbEXsyd1L8dnuJ2wTvXnHp/5+mbQM3IKzeCfu+iWpWwZSnHwShMuxV9u42eP/6f01e5+F3racdRHBNBKD605bzCMfe9Whc4zznUxbBqnWGIC/a3XdRHBI99mkGBeDKT6LW2vgxSXQvjkzlcfxWHyG/X6mGoyye51tf11Snvp7VdbZTJ5MtLHu2mFPvOVx+oY1LE0tgNy10wpBqlfopRW26/D2R62F4IUtDwICy89P7T0zwLQRA4CK0oC7SGaaijpbjJNxMUjB1+pns7q+dlvOX1KRetwg3lCbeDQutye+g3unWNMBu65MVR7H4rp9JksxHe63V6jp+qwz2ca6c+fhLiKX+iX2e4k3PGoykk0rjcead1r//+M3edt/ywO2l1RVY3rvGyDTSgwKjqIimyefMTFwSvhTMa/9dBP1dlgf+axV0JbinNpwsx0OUzVz8v289igab0ORJctgzjFQVjN5ELn179a1lq4YZLIKuXN74rkQrkh0JdlZ1g8xqKiFk98FG38zdZX6wX1WhHM0pdRFxSDfCTXayVOZoHt3ajUGAOW1UFTin2VQPRNmr7ZuolTSVcMtUDt/av++11oDd7pZttxERcX2ynMyy8ANMC88Jb33GrcMAp5pMDJgu8rGZhK5jNca7PR+zKFe+//iRzHpae+zFxR/+87k+7lVxzkcLwAVg/wnk20eunennikj4p8V09dhRxLOOcZmJ03lwonHVAVnLjXz7DCkqcSgY7MVvNh5uplk8ek2vTXR30PzOlsd7TabS5VMtbEebyY3iZsIkosbeG1d7YXqWXDCtfDsbZPPLN7yINQuyGxlegqoGOQ7mewGmmqNgYtfVci9HfYfcfZq+ziVuEF3szdhKyqyqZtTuYnaN9v6gmz2sXLrDZrjDHGPjEHzU/7kuGeqjXVs6+pYqppsKm8ylkE6aaXxOOND1vW27vvxnx8ZtLPLV16Quz3OHFQM8p1MdQNNp8bAJdSYvmvBGOsmqmqaGBu6L8m4wdiItSa8xj4ap5iHbIy9Is9W8Nhl/ok2sB4vbtC+EYZ60qsvcMlUG+tEBWcuIhMN6zwfc6e99UsMGpbB0VfA07fE/33setwmIOS4iwhUDPIfP9s8TEY6NQYuofr0hWswbEv6q2bZk9KMhdCWZHppzx7AJCEGK6x7IVGBW1+HdVdl2w1QWmkFIV7cwI0X+GEZZCqA3LnDZuxM5tZKttaga+fUx0yWMz9iCxjX//Tw57Y8aFuaLz3bv/cLiEDFQEQuFJGXRGSbiHw6zvOLRORREXlGRJ4XkYuDXE9BEmq0ZupQwDNS00krdfHDinELzqpn2Vs3iJwMkw21iUfjcjsRLdEV6HjwOMuWAdgr/z3PHN68cPcTNmDuR3V0ptpYd26HhiWTu1dcMYhEvB2za6edg+6ny2busbbF9bofHprmaoxNKV12TmpJFxkmMDEQkWLge8BFwFHANSISe+n0OeBOY8wJwNVAAsebkpBMFZ6lU3Dm4ga707Fi3IIzNyV0zmrbLmJk0PsxkhWDpikyirLVoC4ei8+wFwet6w/dvnsdLDzVv5NgJlpSxGtdHUvDUhgbsparp2PutGLgN2d+xLovn71tYlvHS/b/ZmUKswuyQJCWwSnANmPMdmPMMHAHEDsyzADu7LcZwJ4A11OYVDXZ26AzitwsDC8ZOInwo1ldnyMG0ZaBiUzk+Xsh7M5kSDDUJpbx9NIEQeT2jTaoWj3b+xqCYuGpgBw67Ka7GXpa/YkXuATdkmJs1GkznSB47OL6/r3EDSIRm6EUxIySJWfB/DU2zdQdQep21V2R2/UFLkGKwXygOepxi7Mtmi8C14lIC3A/EHeAgIisFZH1IrK+o6MjiLXmL25lb9BDbrp3Wz99OuZupQ9VyO7nrIoSA0jOVRRugVCT989SMcO+X6JW1u3OQJtcyBaprIPZRx/ap8jPeEH0+wRpGYSbrYWTKJPIZbzwbOfUx+zdZ62IIMTAbW/dtRM23mu3bXnQpj97vejIMtkOIF8D/I8xZgFwMfBzETlsTcaYm40xa4wxa2bOnKJidLqRSTdROi4i8KcKubfdtuBwRbBhqa0DSCaIHG5Jvoq6aUV8N5Ex1irJdiZRNItOh+anJ65Qdz9hq5NnH+3fewRtGUyVSeQyY6H9e/BSa+B3JlEsR15sBwY9fpP9G29elxdZRC5BikErEO1TWOBsi+ZdwJ0AxpgngAqgKcA1FR55KQbpWAbt9jhFTp+pomJ7VZ5MrYHXgrNoGpfHdxMd3GfdXtmqPI7H4tNtOuO+5+zj3ets1XGRj725grYMxruVTmEZlJRZYfdiGYyLwRTHTJWiIhs7aHsBHviMdV+qGADwNLBCRJaKSBk2QHxfzD67gfMARGQVVgzUD5QMZdW2JD5IMfCjxgD8mWnQ2zHhInKZs9rWGngJTI8PtUlBDPoPHG7VuANtcsoycIfdPGHrOto3+hsvgOAtg84dthFhTYKJhNF4rTXo2glIsPMmjnmTzdp6/g7ripx3YnDv5TOBiYExZhT4IPAgsAmbNbRBRG4UkUud3T4OvEdEngNuB95uTNAJ8wWGSPCFZ37UGIA/nUvdvkTRzF5tr1J7pmjfDPbkONKXmpsIDm9KNp5JlEOtBmrn2hPk7iesuwjj/3StoNtYu83kvMyGqF/q3TKYscBaE0FRUganf8DeX/HazM+2SINAm/sbY+7HBoajt90QdX8jcGaQa5gWBN2fyI8aA/CnWV1vu5MxE0V0EHmqk7zXoTaxNLpisBUWnjyxvX2TvQKsyjHv5qIzYOuDVsSKSuwMAz+JbmMdK85+0LnduzunfoltPjd0EMonmWIYVCZRLCe+zc47XvOO4N/LR/JHtpTEBN2fyI8aA/DHiunrmEgrdXEDo17aWSdbY+BSv9ieVGODyB2bc6O+IJbFp9vf83O/grnHQVnI3+MHWYVsjL2Knyp47NLgMaMoqBqDWMqr4dr/Tb87bIZRMSgEgnYTjYuBD77WdKyY4T4Y6T98BkFFrbVavASRUxWD4lJ7VRndsM6YiQZ1uYYbNzi4x/94AQQ74Ka3zX7PU6WVunipNRjut+7OTFgGeYqKQSGQCTFIt8bApbIhdTGIrT6OZs4x3moNwrvtDOBU3DqxDevCLbYnTS4Fj10aj5j4PfkdL4CoNtYBzDTwmknk4qXWwL2gCSqTqABQMSgEQk32nzIyFszx/UgrdUnHpRXblyia2UdD58uH9+SJxa0xSKVAzJ2H7PbBycXgsYvIhEWwMAAxCLKNtXuF79UyqKyz65lMDIKuMSgAVAwKgVAjYIKbPOWrGKRhxUxmGbhtKdqnaEuRSo2BS9MKW8Hqzk/OpQZ18Tj9A3DOZ4IJ8AY54KZrB0hxcn9zDUsnLzxTMZgSFYNCwM9h87FEIvbk56cYDHR57zIZTWxfomjmuINupnAVpVJ97BLbo6h9k+1H5P7+c41Fp8E5hzUL9odALYPt9jsqLvX+mqlaWXfttJ1W3cJH5TA8iYGI3C0ir4vXKkLJAYKsQu5ts/MD/Agegz1xmjEYSqFZXa/jJopnGdQtsQV4k4nB6LCtGE616KgxptagY1PuWgVBU1IWXBvrTg/dSmOpX2otWLcFRyxu3UIu9I/KUbye3L8PvAXYKiJfE5EjA1yTkixBioFfNQYu6fQn6uuwQ0niXTEWFVnf/WRB5J5WkhpqE0v1LFsrsX+rtWw6XsrNeEGmCKolRdcO7/ECl/oltrFdosJDVwyUhHgSA2PMI8aYa4ETgZ3AIyLyNxF5h4gkYcspgZARMfDRTQSprbWv/fBWFNHMWW3TSxMVsadacOYiYrN0Dmy1WUkj/bmZSZQpgmhJMdBlf5LN+hmvNYgTN3DrFlQMJsWz20dEGoG3A+8GngG+jRWHhwNZmeKdIGMG7hwDv/q5jK81BcugtyO+i8hl9mrrfgo3x38+1RqDaBpXWDfRePA4BwvOMkUQlkGnx26lsbgn+nhxg952GB1QMZgCrzGDe4C/ACHgEmPMpcaYXxljPgRUB7lAxQOlldZ/G0RLiu7d9gTsVwVrOjMN4vUlimbOMfY2kavIFYPaNPrLNy63YrPnGft45jT2mAZhGXQlmVbqUjsfikrjF55pJpEnvFoG3zHGHGWM+aoxZm/0E8aYNQGsS0mWUGMwA278TCuF9NxE8TqWRuP67xMFkcNu8VxF8u/t4o7A3Pw7ewJyUyynI4FYBm7B2ZLkXlfkpKLGswxUDDzhVQyOEpHxv3oRqReR9we0JiUVqgKqQvZbDMpr7BVcsmsdGbQuoMksg/Jq62vel6BHUTpppS5uRlHbi9M3k8glCMugcydUz4GyquRfW78kfsxgfGSrj3/HBYhXMXiPMWb8WzfGdAHvCWZJSkoE0ZLC7xoDmGhWl+xMA7f6eDLLACaCyPFIp+DMpfGIifu52KAukwTRxjqVTCKXhgStrLt2Qs289CzCaYBXMSgWmUjQFZFiIMCm4ErSBCEG4zUGPl9RhVLoTzRZK4poZh9jXQ3DfYduN8axDNIUg7KqiZiDWgb21k/rIJnW1bHUL7FT52L/tjSTyBNexeAB4Fcicp6InIcdRPNAcMtSkiaImQZ+1xi4pCJcfZMUnEUzZzVgoG3jodv7O20qaLpuIpioRFbLwN76FTcYGYCDe5PPJHJJ1LBOxcATXsXgU8CjwPucnz8A/xLUopQUCDXYDpqjQ/4d0+8aA5dULIPJ+hJFk2i2gZtu6ocYuFPPpnMmEfhvGbgn8VTdROPppVFxg5FB6NmjYuABT5POjDER4AfOj5KLRFf21nqYG+sFv2sMXFKyDCbpSxRN3WJbJRwbN0i34CyaU99nJ4dNNlVrOuC3ZZBs6+pY4tUahJsBk5mhNnmOJzEQkRXAV4GjsEPrATDGpGjPKb4TnbLpmxj4XGPgUtlgA8iRiPcZsb0dUFYz9UwFEWsdxNYajIuBD1ZO0/KJFNPpzLhl4FO33GRbV8dSXm3/XqNrDTSt1DNe3UT/jbUKRoFzgZ8BvwhqUUoKBNGSwu+0UpdQo203ncwV5VQFZ9HMdjKKojujhpuhpDJ3O4zmI363se7aARUz0vuO6mMyilQMPONVDCqNMX8AxBizyxjzReB1wS1LSZpxMfCx8CxIMYDkrih7p+hLFM2c1TZ+4rq5wIpBqkNtlPj43cY6nUwil9hW1l07oaTCthpXJsWrGAw57au3isgHReQKtA1FbpFON9B4BFFj4JKKFdO3PznLAA6tRPaj4Ew5FL/bWKfSujqWhqX2ux4dto+1dbVnvIrB9di+RB8GTgKuA94W1KKUFEin5088gqoxAAjV29ukxKB96kwil1mrADk0iOxHwZlyOH61pBgbsRcfqcYLXOqXAGYiE07TSj0zpRg4BWZvNsb0GmNajDHvMMZcaYxZl4H1KV4pLrFmu19iEFSNASRvGYyNWovHq5uorMpWCrttKUaHrLj5nRWl+NeSItxs5xGk7SaKqjXQ1tVJMaUYGGPGgLMysBYlXfysQh7Pyw/gBJqsS6t/P2CSm+U7e/WEm8jPtFLlUPyyDFJtXR1LdK1BfycM96oYeMSrm+gZEblPRN4qIm9wf6Z6kYhcKCIvicg2ETlsGKuI/KeIPOv8bBGRAMYmTSP8FIPx5l4BiEFZNRSXeV/reMGZR8sArBh07YTBHhWDIPHLMki1dXUsNXNswLhrp2YSJYmnOgNsbcEB4NVR2wxwd6IXOO6l7wHnAy3A0yJynzFmvE+AMeajUft/CDjB+9KVwwg1Qk+LP8fq3g2hptS6R06FiI1xeBUDrwVn0cxxgsjtm/wZaqPEp7IO9vpkGZRU2I6l6SAykVHkCoyKgSe8ViC/I4VjnwJsM8ZsBxCRO4DLgI0J9r8G+EIK76O4hBph3/P+HCuotFKXZHopuXMavAaQISqj6AXoOwAI1M5LaomKB/yyDDp3WH+/1yLEyahfao/nWgZBxL0KEK8VyP+NtQQOwRjzzkleNh+Inj/YApya4PiLgaXAHxM8vxZYC7BokfYkT0jIudo2Jv1Uuu7dEyfUIAg1eG9j7bUvUTQzFtgCpn0v2sBk9WwoKU9+ncrkRLexLk5jHHo6ratjqV8CO/5sj1k92/8K+gLFqwz/Fvid8/MHoBbo9XEdVwN3OcHqwzDG3GyMWWOMWTNzZhInhOlGqBFGBw9v35wskQh0B1Rj4JJMfKOv3boQkukFJGLbWbe9OFFwpviPH83qjJmwDPygYakVqJb1ahUkgVc30a+jH4vI7cDjU7ysFYh20i5wtsXjauADXtaiTEJ0ymZ5GjWBfe0wNhSwGCQRM3DHXSZr7cw+Gp75BdTMhjnHJr9GZWqim9Ulk+0VzcF9dmC9n5YBQMdmOOYqf445DUjVQbcCmCqa9zSwQkSWikgZ9oR/X+xOIvIKoB54IsW1KC5VTfY23YyiIGsMXEKNth1FdP+gRCTTlyiaOavtFWLndi04Cwo/LAO/Molcoi0MDR57xmvM4CCHxgz2YWccJMQYMyoiHwQeBIqBW4wxG0TkRmC9McYVhquBO4wxh8UklCTxqyVFUHMMooluVjdVY7LejtTcPNExD80kCgY/2lin27o6lrpFgGBbVy/x55jTAK9uopQatxtj7gfuj9l2Q8zjL6ZybCUOfnUuDbLGwCVauKYSg752mJ9C1vGsVSBFVnQ0ZhAMflgGnTtAiv27+CitsJljPa0qBkngyU0kIleIyIyox3Uicnlwy1JSIuRTf6IgawxcvK41ErGppckUnLmUVk6MqFQxCIbxNtZpzDQIN9u50ulkI8XiioCKgWe8xgy+YIwJuw+MMd1oTUDuUT7DXmH5IQZBuojAe2O9gS4wY8mllUbjuorUTRQMfrSxDqKjbP1SKC6HGp8GPU0DvFYgxxMNr69VMkVRUXJZOokIusYAomYaTBHfGK8+TlEMVl1iextV1qf2emVySsqgNJSemyjcAgtP8W9NAGdeDytf608R2zTB6wl9vYh8C9teAmwa6N+DWZKSFqHG9AbcuDUGR17s35ri4TW+kUpfomhWv8H+KMFRWZ+6ZRCJ2IH1flsGM1faH8UzXmXzQ8Aw8CvgDmAQrQvITZJp8xCPTNQYgI1HFJdPLQZ9HfY2mb5ESmZJpyVFXztERmzMQMkqXrOJ+oDDuo4qOUioAfZvTf31magxAFtA5sWllUorCiWzpNPGWpsI5gxes4keFpG6qMf1IvJgcMtSUibdNtaZqDFwCTVC/xRZKH3tUFSiPv9cJh3LYFwM1DLINl7dRE1OBhEAxpgupq5AVrJBqMm6ibxU9sYjEzUGLl4sg74OaxXoDNvcxRfLQFN/s41XMYiIyPiloogsIU4XUyUHCDXaVMyh8NT7xiMTNQYuXqyY3g51EeU66VgGPa1QWjWRoqpkDa/ZRP8KPC4if8LWeZ+N01JayTGiK3tTca10785cHx8vA2762jV4nOuk08ba7Sirll/W8WQZGGMeANYALwG3Ax8HBgJcl5Iq6bakyETBmUuo0boXInE7l1vcjqVK7pJOS4pwq8YLcgSvjereDVyPbUP9LHAatsvoqyd7nZIF0mlJERlzagwu8ndNiRhvVheO35/ImNQ7liqZI5021uGWiRGlSlbxGjO4HjgZ2GWMORc7q1iH1+cirmXQl0LhWddOW2PQdKSvS0rIVFbMYBjGhtUyyHVStQxGh6zY12rwOBfwKgaDxphBABEpN8ZsBjJ0xlCSIh03UdsGezv7KP/WMxkhJ6aRaK2uoGnMILdJtY11jzPrSjOJcgKvAeQWp87gXuBhEekCdgW3LCVlvFb2xqN9IyAwc5Xvy4rLVPMX3L5E7tAeJTdJ1TLQGoOcwmsF8hXO3S+KyKPADOCBwFalpI5I6i0p2jfaaVOZGiA+lRWTbl8iJTOkahmEXctAq49zgaQ7jxpj/hTEQhQfqUqxCrltI8zKkIsIphYD7UuUH4xbBknONHAtg9p5/q5HSQnt71qIpNKSYmQAOl/OrBiUhiZ3afW220llrmgouUmqbax7WmyBY2llMOtSkkLFoBBJRQw6XrJpnpkKHsOESyvRTIO+dvt8UXHm1qSkRiptrMMtGi/IIVQMCpFUxKB9o72ddbT/65mMyeIbWnCWP6TSkiLcqvGCHELFoBBxK3vHRr2/pm2Dddk0LAtuXfGYrFldX4dmEuULqTSrC7foHIMcQsWgEPE6UjKa9o0w80gozvA000nFQPsS5Q3JWgaDYRg+qDUGOYSKQSGSSkuKto0wO8MuIlA3UaGQrGWgratzDhWDQiTZKuT+Tujdl9lMIpdQo01JjG1WN9xnO2FqX6L8IFnLIKzVx7mGikEhkqwYZLoNRTShRsAcfiLRgrP8IrqNtRfCzfZWxSBnUDEoRJIVg2xlEoGdaQCHr1ULzvKLZFtS9LTacabVs4Nbk5IUgYqBiFwoIi+JyDYR+XSCfa4SkY0iskFEbgtyPdOGVCyDynqomRPcmhLhxjdig93jloG6ifKCZFtShFugZp7WkOQQgaWOiEgx8D3gfKAFeFpE7jPGbIzaZwXwGeBMY0yXiOhloB+UlENZjff+RO0brVWQjWlTiYTLtQxUDPKDZC0DHWqTcwRpGZwCbDPGbDfGDAN3AJfF7PMe4HvGmC4AY0x7gOuZXngZNg8QiUD7puzEC0DFoFBI2jJo1nhBjhGkGMwHmqMetzjbolkJrBSRv4rIOhG5MN6BRGStiKwXkfUdHR0BLbfACDV6G3AT3g3DvdnJJILEabC97fZqs6Qs82tSkicZyyASgZ49WnCWY2Q7gFwCrADOAa4BfuzMTTgEY8zNxpg1xpg1M2fqlaInvLakaN9kb7NRYwC2wVlJxeEuLS04yy+SsQz62iEyopZBjhGkGLQC0Y1HFjjbomkB7jPGjBhjdgBbsOKgpIvXmQZuWumsDA20iSXR/AUtOMsvkrEMtMYgJwlSDJ4GVojIUhEpA64G7ovZ516sVYCINGHdRtsDXNP0wbNlsBHqFkF5TfBrSkS8+EZfuxac5RPjbaw9zDTQGoOcJDAxMMaMAh8EHgQ2AXcaYzaIyI0icqmz24PAARHZCDwKfNIYk8JUFuUwQg22CGhkYPL9Mj3QJh6V8cSgQ4PH+UaFx5YU7uxjjRnkFIF2JTPG3A/cH7Pthqj7BviY86P4SfR84UQpfKPDcGArHHlR5tYVj1Aj7Ht+4vHokG1kpm6i/KKy3qObqAVKq+z+Ss6Q7QCyEhRu6+fJXEX7t0BkNHvBY5dYl9Z49bFaBnmF12Z17lCbbNS1KAlRMShUvFQhj7ehyLKbKNRoryjd+Qvalyg/8dqsLtyi8YIcRMWgUPEiBm0boKgUmrKcwBVqAMzEVaX2JcpPvFoGPa0aL8hBVAwKFa+WQdNKKC7NzJoSER3fAO1LlK94sQxGh6C3Tcdd5iAqBoVKRR0gU1gGG7PXhiKa2CrkPhWDvMRLG+uePfZW+xLlHCoGhUpxif3nTCQGA93Q05L9eAEcbsX07YeyaigLZW9NSvJ4mcniPAAADNJJREFUKTzTCWc5i4pBITNZ4Vm221BEEzvToLddrYJ8xEtLivEaAxWDXEPFoJCZVAzcNhQ5ZBm4Mw20L1F+4skycKuP1U2Ua6gYFDKT9Sdq2wjlM3LDXC8LQUlllGWg1cd5iRfLINxi/y5LKzOzJsUzKgaFzGQzDdo32eZ0uVL4Ey1cahnkJ54sg9bcuABRDkPFoJAJNVkxMObQ7cZYN1EuZBK5hOrtWsdGrShowVn+4dUy0HhBTqJiUMiEGmFs2A6viaZnj+39kwvxAhfXMujfD5iJdhpK/uDFMuhRyyBXUTEoZNzAbOzEM7cNRS5kErm4wW6tPs5f3DbWiSyDwTAM9WjwOEdRMShkYit7XbI90CYerhhoX6L8pqIu8UwDHWqT06gYFDKJWlK0b7S9YXKphXBlg72iPLjPPlbLID+pnKQlhVtwpjGDnETFoJBJNGy+bWNuWQUwIVz7X7K3mlqan1TWJ3YT9Wj1cS6jYlDIxLMMxkbsCTeXgscwIVztm6GkIrtjOJXUmaxZXbgFpBhq5mR2TYonVAwKmYoZ9p8vWgwOvGwzjHIpeAwTwtWxycYLcqX+QUmOydpYh1uhdh4UFWd2TYonVAwKGZHDW1LkUhuKaFwx6N6taaX5zFSWgc4xyFlUDAqdWDFo22ithZlHZm9N8XDdRKDB43xmsjbWPTrhLJdRMSh0qppiLION0LgcSsqzt6Z4VEaJgQaP85dEhWeRiNOKQi2DXEXFoNCJ7U/UlmNtKFzKQrZgCdQyyGcStaTo64DIiE44y2FUDAqdaDfR0EHo3gWzcix47OLGDbTgLH9JZBmM1xioZZCrqBgUOqFGWxEaGbNpm5CblgFMFMFVq5sob0lkGWiNQc6jYlDohBrBRGxfGLcnUa5lErmMWwYqBnnLVJaBikHOEqgYiMiFIvKSiGwTkU/Hef7tItIhIs86P+8Ocj3TkujCs/aNUFoFdYuzu6ZEqJso/0lkGYRbbUwol1qgKIdQEtSBRaQY+B5wPtACPC0i9xljNsbs+itjzAeDWse0J7olRdsG24aiKEcNQlcMNICcvyS0DJqtVaDFhDlLkGeFU4Btxpjtxphh4A7gsgDfT4lHdBvr9o25Gy8Au7baBRMnFCX/SNTGuqdVg8c5TpBiMB9ojnrc4myL5UoReV5E7hIRzTvzm+g2D/0HcjeTCODEt8FHX8xdy0XxRrwq5LAWnOU62f6v+z9giTHmWOBh4NZ4O4nIWhFZLyLrOzo6MrrAvCfktHbY8Rd7m8uWgYi6EQqBypiZBqND0NumYpDjBCkGrUD0lf4CZ9s4xpgDxpgh5+FPgJPiHcgYc7MxZo0xZs3MmZppkhRlISiphOYn7eNctgyUwiC2jXXPHnurYpDTBCkGTwMrRGSpiJQBVwP3Re8gInOjHl4KbApwPdOXUCOMDkL1bKhqzPZqlEIn1k3U41wDaswgpwksm8gYMyoiHwQeBIqBW4wxG0TkRmC9MeY+4MMicikwCnQCbw9qPdOaUIMt+sm1gTZKYVJZB3ujxGC8xkBDgrlMYGIAYIy5H7g/ZtsNUfc/A3wmyDUoTASR1UWkZIJYy2C8FcW87KxH8US2A8hKJnDFIJeDx0rhENvGOtxi/wbLQtldlzIpKgbTgXHLQMVAyQCxhWdaY5AXqBhMB+qXQFk1zHxFtleiTAdiW1KEWzRekAeoGEwHTn4XfOApNdOVzBBrGYRbdKhNHqBiMB0oKdd/RiVzRFsGg2EY6tEagzxAxUBRFH+JtgzCWmOQL6gYKIriL9GWgVtwpjGDnCfQOgNFUaYh0ZZBkdOrUt2UOY+KgaIo/hLdxnp0EKQYqudke1XKFKgYKIriP24VcmQUauZCsZ5qch39hhRF8Z/KuolsIs0kygs0gKwoiv9UODMNws0aL8gTVAwURfGfynorBj171DLIE1QMFEXxn8o66NwOY8N2rrWS86gYKIriPxV1NpMI1DLIE1QMFEXxH7fwDDRmkCeoGCiK4j8V0WKg1cf5gIqBoij+41oGJZU2mKzkPCoGiqL4j2sZzFgAItldi+IJFQNFUfzHtQw0XpA3qBgoiuI/0ZaBkheoGCiK4j+uZaA1BnmD9iZSFMV/qmbCOZ+FY96Y7ZUoHlExUBTFf0TgnE9lexVKEqibSFEURVExUBRFUQIWAxG5UEReEpFtIvLpSfa7UkSMiKwJcj2KoihKfAITAxEpBr4HXAQcBVwjIkfF2a8GuB54Mqi1KIqiKJMTpGVwCrDNGLPdGDMM3AFcFme/fwO+DgwGuBZFURRlEoIUg/lAc9TjFmfbOCJyIrDQGPO7yQ4kImtFZL2IrO/o6PB/pYqiKNOcrAWQRaQI+Bbw8an2NcbcbIxZY4xZM3PmzOAXpyiKMs0IUgxagejetQucbS41wGrgMRHZCZwG3KdBZEVRlMwjxphgDixSAmwBzsOKwNPAW4wxGxLs/xjwCWPM+imO2wHsSnFZTcD+FF+bqxTaZyq0zwOF95kK7fNA4X2meJ9nsTEmoWslsApkY8yoiHwQeBAoBm4xxmwQkRuB9caY+1I8bsp+IhFZb4wpKMuj0D5ToX0eKLzPVGifBwrvM6XyeQJtR2GMuR+4P2bbDQn2PSfItSiKoiiJ0QpkRVEUZdqJwc3ZXkAAFNpnKrTPA4X3mQrt80DhfaakP09gAWRFURQlf5huloGiKIoSBxUDRVEUZfqIgdcOqvmCiOwUkRdE5FkRmbQ2I1cRkVtEpF1EXoza1iAiD4vIVue2PptrTIYEn+eLItLqfE/PisjF2VxjsojIQhF5VEQ2isgGEbne2Z6X39MknydvvycRqRCRp0TkOeczfcnZvlREnnTOeb8SkbJJjzMdYgZOB9UtwPnYHklPA9cYYzZmdWFp4FRtrzHG5G2hjIi8EugFfmaMWe1s+39ApzHma45o1xtj8mJkVoLP80Wg1xjzzWyuLVVEZC4w1xjzD6fD8N+By4G3k4ff0ySf5yry9HsSEQGqjDG9IlIKPI7tBP0x4G5jzB0i8kPgOWPMDxIdZ7pYBl47qCoZxBjzZ6AzZvNlwK3O/Vux/6h5QYLPk9cYY/YaY/7h3D8IbMI2nMzL72mSz5O3GEuv87DU+THAq4G7nO1TfkfTRQym7KCahxjgIRH5u4iszfZifGS2MWavc38fMDubi/GJD4rI844bKS/cKfEQkSXACdjZI3n/PcV8Hsjj70lEikXkWaAdeBh4Geg2xow6u0x5zpsuYlCInGWMORE7POgDjouioDDWh5nvfswfAEcAxwN7gf/I7nJSQ0SqgV8DHzHG9EQ/l4/fU5zPk9ffkzFmzBhzPLYh6CnAK5I9xnQRg6k6qOYdxphW57YduAf7B1AItDl+Xde/257l9aSFMabN+UeNAD8mD78nxw/9a+CXxpi7nc15+z3F+zyF8D0BGGO6gUeB04E6p2EoeDjnTRcxeBpY4UTXy4CrgZQa5eUCIlLlBL8QkSrgtcCLk78qb7gPeJtz/23Ab7K4lrRxT5gOV5Bn35MTnPwpsMkY862op/Lye0r0efL5exKRmSJS59yvxCbKbMKKwhud3ab8jqZFNhGAkyp2ExMdVL+S5SWljIgsw1oDYJsN3paPn0dEbgfOwbbbbQO+ANwL3AkswrYqv8oYkxdB2QSf5xys68EAO4H3Rvnacx4ROQv4C/ACEHE2fxbrZ8+772mSz3MNefo9icix2ABxMfYC/05jzI3OeeIOoAF4BrjOGDOU8DjTRQwURVGUxEwXN5GiKIoyCSoGiqIoioqBoiiKomKgKIqioGKgKIqioGKgKBlFRM4Rkd9mex2KEouKgaIoiqJioCjxEJHrnB7xz4rIj5xGYL0i8p9Oz/g/iMhMZ9/jRWSd0+TsHrfJmYgsF5FHnD7z/xCRI5zDV4vIXSKyWUR+6VTFKkpWUTFQlBhEZBXwZuBMp/nXGHAtUAWsN8YcDfwJW2EM8DPgU8aYY7GVre72XwLfM8YcB5yBbYAGtlPmR4CjgGXAmYF/KEWZgpKpd1GUacd5wEnA085FeyW2EVsE+JWzzy+Au0VkBlBnjPmTs/1W4H+d3lHzjTH3ABhjBgGc4z1ljGlxHj8LLMEOJFGUrKFioCiHI8CtxpjPHLJR5PMx+6XayyW6P8wY+n+o5ADqJlKUw/kD8EYRmQXj834XY/9f3C6QbwEeN8aEgS4ROdvZ/lbgT84UrRYRudw5RrmIhDL6KRQlCfSKRFFiMMZsFJHPYSfJFQEjwAeAPuAU57l2bFwBbHvgHzon++3AO5ztbwV+JCI3Osd4UwY/hqIkhXYtVRSPiEivMaY62+tQlCBQN5GiKIqiloGiKIqiloGiKIqCioGiKIqCioGiKIqCioGiKIqCioGiKIoC/H/7kyglVzIq9wAAAABJRU5ErkJggg==\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "earlystop = EarlyStopping(monitor='val_loss', mode='min', patience=10,restore_best_weights=True, verbose=4)\n",
        "filename='ResNet50.csv'\n",
        "history_logger=tf.keras.callbacks.CSVLogger(filename, separator=\",\", append=True)\n",
        "opt = Adam(learning_rate=0.001)\n",
        "base_model =ResNet50(input_shape=(100,100,3), weights='imagenet', include_top=False)\n",
        "model50=Sequential()\n",
        "model50.add(base_model)\n",
        "model50.add(Flatten())\n",
        "model50.add(Dense(8,kernel_initializer='he_uniform',kernel_regularizer=regularizers.L1(0.0001),activation='relu'))\n",
        "model50.add(BatchNormalization())\n",
        "model50.add(Dropout(0.3))\n",
        "model50.add(Dense(1,activation='sigmoid'))\n",
        "model50.compile(optimizer=opt,loss='binary_crossentropy', metrics=['accuracy'])\n",
        "history50= model50.fit(x_train_normalize , y_train,validation_split=0.2,epochs=30, batch_size=32,callbacks=[history_logger])\n",
        "loss_value , accuracy = model50.evaluate(x_test_normalize, y_test)\n",
        "print('Test_loss_value = ' +str(loss_value))\n",
        "print('test_accuracy = ' + str(accuracy))\n",
        "plt.plot(history50.history['accuracy'])\n",
        "plt.plot(history50.history['val_accuracy'])\n",
        "plt.title('model accuracy')\n",
        "plt.ylabel('accuracy')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'test'], loc='upper left')\n",
        "plt.show()\n",
        "# summarize history for loss\n",
        "plt.plot(history50.history['loss'])\n",
        "plt.plot(history50.history['val_loss'])\n",
        "plt.title('model loss')\n",
        "plt.ylabel('loss')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'test'], loc='upper left')\n",
        "plt.show()\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {
        "id": "VGZJSXt2-AoR",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "8bf2d179-50b9-4510-f830-8ec1946f215d"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "WARNING:tensorflow:`input_shape` is undefined or non-square, or `rows` is not in [128, 160, 192, 224]. Weights for input shape (224, 224) will be loaded as the default.\n",
            "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/mobilenet/mobilenet_1_0_224_tf_no_top.h5\n",
            "17227776/17225924 [==============================] - 0s 0us/step\n",
            "17235968/17225924 [==============================] - 0s 0us/step\n",
            "Epoch 1/30\n",
            "978/978 [==============================] - 43s 40ms/step - loss: 0.3669 - accuracy: 0.8755 - val_loss: 0.2734 - val_accuracy: 0.8753\n",
            "Epoch 2/30\n",
            "978/978 [==============================] - 38s 39ms/step - loss: 0.1667 - accuracy: 0.9467 - val_loss: 0.1674 - val_accuracy: 0.9443\n",
            "Epoch 3/30\n",
            "978/978 [==============================] - 39s 39ms/step - loss: 0.1412 - accuracy: 0.9547 - val_loss: 0.1778 - val_accuracy: 0.9378\n",
            "Epoch 4/30\n",
            "978/978 [==============================] - 38s 39ms/step - loss: 0.1270 - accuracy: 0.9586 - val_loss: 0.1439 - val_accuracy: 0.9509\n",
            "Epoch 5/30\n",
            "978/978 [==============================] - 39s 40ms/step - loss: 0.1181 - accuracy: 0.9621 - val_loss: 0.1156 - val_accuracy: 0.9611\n",
            "Epoch 6/30\n",
            "978/978 [==============================] - 39s 40ms/step - loss: 0.1045 - accuracy: 0.9688 - val_loss: 0.1309 - val_accuracy: 0.9544\n",
            "Epoch 7/30\n",
            "978/978 [==============================] - 39s 40ms/step - loss: 0.0980 - accuracy: 0.9709 - val_loss: 0.1254 - val_accuracy: 0.9605\n",
            "Epoch 8/30\n",
            "978/978 [==============================] - 39s 39ms/step - loss: 0.0863 - accuracy: 0.9745 - val_loss: 0.1193 - val_accuracy: 0.9609\n",
            "Epoch 9/30\n",
            "978/978 [==============================] - 39s 39ms/step - loss: 0.0747 - accuracy: 0.9778 - val_loss: 0.1441 - val_accuracy: 0.9592\n",
            "Epoch 10/30\n",
            "978/978 [==============================] - 38s 39ms/step - loss: 0.0635 - accuracy: 0.9826 - val_loss: 0.1913 - val_accuracy: 0.9594\n",
            "Epoch 11/30\n",
            "978/978 [==============================] - 39s 39ms/step - loss: 0.0597 - accuracy: 0.9840 - val_loss: 0.1381 - val_accuracy: 0.9601\n",
            "Epoch 12/30\n",
            "978/978 [==============================] - 39s 40ms/step - loss: 0.0523 - accuracy: 0.9864 - val_loss: 0.1692 - val_accuracy: 0.9571\n",
            "Epoch 13/30\n",
            "978/978 [==============================] - 39s 40ms/step - loss: 0.0444 - accuracy: 0.9881 - val_loss: 0.3269 - val_accuracy: 0.9038\n",
            "Epoch 14/30\n",
            "978/978 [==============================] - 39s 39ms/step - loss: 0.0384 - accuracy: 0.9899 - val_loss: 0.1875 - val_accuracy: 0.9556\n",
            "Epoch 15/30\n",
            "978/978 [==============================] - 39s 40ms/step - loss: 0.0386 - accuracy: 0.9897 - val_loss: 0.1830 - val_accuracy: 0.9600\n",
            "Epoch 16/30\n",
            "978/978 [==============================] - 38s 39ms/step - loss: 0.0348 - accuracy: 0.9909 - val_loss: 0.1483 - val_accuracy: 0.9604\n",
            "Epoch 17/30\n",
            "978/978 [==============================] - 39s 40ms/step - loss: 0.0328 - accuracy: 0.9924 - val_loss: 0.1959 - val_accuracy: 0.9542\n",
            "Epoch 18/30\n",
            "978/978 [==============================] - 39s 39ms/step - loss: 0.0298 - accuracy: 0.9928 - val_loss: 0.1567 - val_accuracy: 0.9533\n",
            "Epoch 19/30\n",
            "978/978 [==============================] - 39s 40ms/step - loss: 0.0319 - accuracy: 0.9924 - val_loss: 0.2391 - val_accuracy: 0.9610\n",
            "Epoch 20/30\n",
            "978/978 [==============================] - 39s 40ms/step - loss: 0.0326 - accuracy: 0.9922 - val_loss: 0.2147 - val_accuracy: 0.9601\n",
            "Epoch 21/30\n",
            "978/978 [==============================] - 38s 39ms/step - loss: 0.0334 - accuracy: 0.9924 - val_loss: 0.1560 - val_accuracy: 0.9600\n",
            "Epoch 22/30\n",
            "978/978 [==============================] - 39s 40ms/step - loss: 0.0256 - accuracy: 0.9939 - val_loss: 0.3456 - val_accuracy: 0.9199\n",
            "Epoch 23/30\n",
            "978/978 [==============================] - 39s 40ms/step - loss: 0.0255 - accuracy: 0.9949 - val_loss: 0.5170 - val_accuracy: 0.9480\n",
            "Epoch 24/30\n",
            "978/978 [==============================] - 39s 40ms/step - loss: 0.0248 - accuracy: 0.9944 - val_loss: 0.1722 - val_accuracy: 0.9606\n",
            "Epoch 25/30\n",
            "978/978 [==============================] - 39s 39ms/step - loss: 0.0244 - accuracy: 0.9941 - val_loss: 0.2838 - val_accuracy: 0.9536\n",
            "Epoch 26/30\n",
            "978/978 [==============================] - 39s 39ms/step - loss: 0.0358 - accuracy: 0.9905 - val_loss: 0.2095 - val_accuracy: 0.9402\n",
            "Epoch 27/30\n",
            "978/978 [==============================] - 39s 40ms/step - loss: 0.0296 - accuracy: 0.9935 - val_loss: 0.2560 - val_accuracy: 0.9321\n",
            "Epoch 28/30\n",
            "978/978 [==============================] - 39s 40ms/step - loss: 0.0292 - accuracy: 0.9936 - val_loss: 0.2345 - val_accuracy: 0.9479\n",
            "Epoch 29/30\n",
            "978/978 [==============================] - 39s 40ms/step - loss: 0.0237 - accuracy: 0.9949 - val_loss: 0.2194 - val_accuracy: 0.9555\n",
            "Epoch 30/30\n",
            "978/978 [==============================] - 39s 40ms/step - loss: 0.0219 - accuracy: 0.9951 - val_loss: 0.1839 - val_accuracy: 0.9569\n",
            "524/524 [==============================] - 5s 10ms/step - loss: 0.1752 - accuracy: 0.9546\n",
            "====================\n",
            "Test_loss_value = 0.1751578450202942\n",
            "test_accuracy = 0.9545508623123169\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ],
      "source": [
        "filename='MobileNet.csv'\n",
        "opt = Adam(learning_rate=0.001)\n",
        "history_logger=tf.keras.callbacks.CSVLogger(filename, separator=\",\", append=False)\n",
        "base_model2 =MobileNet(input_shape=(100,100,3), weights='imagenet', include_top=False)\n",
        "modelmo=Sequential()\n",
        "modelmo.add(base_model2)\n",
        "# modelmo.add(Dropout(0.3))\n",
        "modelmo.add(Flatten())\n",
        "modelmo.add(Dense(8,kernel_initializer='he_uniform',kernel_regularizer=regularizers.L1(0.0001),activation='relu'))\n",
        "modelmo.add(BatchNormalization())\n",
        "modelmo.add(Dropout(0.3))\n",
        "modelmo.add(Dense(1,activation='sigmoid'))\n",
        "modelmo.compile(optimizer=opt,loss='binary_crossentropy', metrics=['accuracy'])\n",
        "historymo = modelmo.fit(x_train_normalize, y_train,validation_split=0.2,shuffle=True, epochs=30, batch_size=32,callbacks=[history_logger])\n",
        "loss_value , accuracy = modelmo.evaluate(x_test_normalize, y_test)\n",
        "print(\"=\"*20)\n",
        "print('Test_loss_value = ' +str(loss_value))\n",
        "print('test_accuracy = ' + str(accuracy))\n",
        "plt.plot(historymo.history['accuracy'])\n",
        "plt.plot(historymo.history['val_accuracy'])\n",
        "plt.title('model accuracy')\n",
        "plt.ylabel('accuracy')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'validation'], loc='upper left')\n",
        "plt.show()\n",
        "# summarize history for loss\n",
        "plt.plot(historymo.history['loss'])\n",
        "plt.plot(historymo.history['val_loss'])\n",
        "plt.title('model loss')\n",
        "plt.ylabel('loss')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'validation'], loc='upper left')\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "filename='VGG16.csv'\n",
        "opt = Adam(learning_rate=0.001)\n",
        "history_logger=tf.keras.callbacks.CSVLogger(filename, separator=\",\", append=False)\n",
        "base_model3 =VGG16(input_shape=(100,100,3), weights='imagenet', include_top=False)\n",
        "modelv16=Sequential()\n",
        "modelv16.add(base_model3)\n",
        "modelv16.add(Dropout(0.3))\n",
        "modelv16.add(Flatten())\n",
        "modelv16.add(Dense(8,kernel_initializer='he_uniform',kernel_regularizer=regularizers.L1(0.0001),activation='relu'))\n",
        "modelv16.add(BatchNormalization())\n",
        "modelv16.add(Dropout(0.3))\n",
        "modelv16.add(Dense(1,activation='sigmoid'))\n",
        "modelv16.compile(optimizer=opt,loss='binary_crossentropy', metrics=['accuracy'])\n",
        "historyv16 = modelv16.fit(x_train_normalize, y_train,validation_split=0.2,shuffle=True, epochs=30, batch_size=32,callbacks=[history_logger])\n",
        "loss_value , accuracy = modelv16.evaluate(x_test_normalize, y_test)\n",
        "print(\"=\"*20)\n",
        "print('Test_loss_value = ' +str(loss_value))\n",
        "print('test_accuracy = ' + str(accuracy))\n",
        "plt.plot(historyv16.history['accuracy'])\n",
        "plt.plot(historyv16.history['val_accuracy'])\n",
        "plt.title('model accuracy')\n",
        "plt.ylabel('accuracy')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'validation'], loc='upper left')\n",
        "plt.show()\n",
        "# summarize history for loss\n",
        "plt.plot(historyv16.history['loss'])\n",
        "plt.plot(historyv16.history['val_loss'])\n",
        "plt.title('model loss')\n",
        "plt.ylabel('loss')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'validation'], loc='upper left')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "l8KSBBvL2so9",
        "outputId": "af626be1-203c-4093-f207-2227998e282d"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/vgg16/vgg16_weights_tf_dim_ordering_tf_kernels_notop.h5\n",
            "58892288/58889256 [==============================] - 0s 0us/step\n",
            "58900480/58889256 [==============================] - 0s 0us/step\n",
            "Epoch 1/30\n",
            "978/978 [==============================] - 113s 112ms/step - loss: 0.4007 - accuracy: 0.8499 - val_loss: 0.2595 - val_accuracy: 0.8997\n",
            "Epoch 2/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.3126 - accuracy: 0.8795 - val_loss: 0.3747 - val_accuracy: 0.8760\n",
            "Epoch 3/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.2804 - accuracy: 0.8906 - val_loss: 0.2431 - val_accuracy: 0.9077\n",
            "Epoch 4/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.2689 - accuracy: 0.8924 - val_loss: 0.2360 - val_accuracy: 0.9045\n",
            "Epoch 5/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.2650 - accuracy: 0.8934 - val_loss: 0.2316 - val_accuracy: 0.9078\n",
            "Epoch 6/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.2598 - accuracy: 0.8964 - val_loss: 0.2533 - val_accuracy: 0.8975\n",
            "Epoch 7/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.2563 - accuracy: 0.8969 - val_loss: 0.2626 - val_accuracy: 0.8926\n",
            "Epoch 8/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.2501 - accuracy: 0.9004 - val_loss: 0.4282 - val_accuracy: 0.8762\n",
            "Epoch 9/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.2508 - accuracy: 0.8999 - val_loss: 0.3285 - val_accuracy: 0.8820\n",
            "Epoch 10/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.2452 - accuracy: 0.9020 - val_loss: 0.3379 - val_accuracy: 0.8636\n",
            "Epoch 11/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.2399 - accuracy: 0.9054 - val_loss: 0.2368 - val_accuracy: 0.9073\n",
            "Epoch 12/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.2333 - accuracy: 0.9069 - val_loss: 0.2660 - val_accuracy: 0.9024\n",
            "Epoch 13/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.2284 - accuracy: 0.9113 - val_loss: 0.3753 - val_accuracy: 0.8406\n",
            "Epoch 14/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.2216 - accuracy: 0.9124 - val_loss: 0.2113 - val_accuracy: 0.9150\n",
            "Epoch 15/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.2162 - accuracy: 0.9158 - val_loss: 0.2332 - val_accuracy: 0.9013\n",
            "Epoch 16/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.2129 - accuracy: 0.9168 - val_loss: 0.1956 - val_accuracy: 0.9260\n",
            "Epoch 17/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.2094 - accuracy: 0.9176 - val_loss: 0.2083 - val_accuracy: 0.9214\n",
            "Epoch 18/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.2059 - accuracy: 0.9201 - val_loss: 0.3938 - val_accuracy: 0.8377\n",
            "Epoch 19/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.2023 - accuracy: 0.9222 - val_loss: 0.2659 - val_accuracy: 0.8942\n",
            "Epoch 20/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.2009 - accuracy: 0.9239 - val_loss: 0.5735 - val_accuracy: 0.7850\n",
            "Epoch 21/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.1957 - accuracy: 0.9246 - val_loss: 0.1776 - val_accuracy: 0.9316\n",
            "Epoch 22/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.1942 - accuracy: 0.9255 - val_loss: 0.1864 - val_accuracy: 0.9302\n",
            "Epoch 23/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.1919 - accuracy: 0.9264 - val_loss: 0.2131 - val_accuracy: 0.9156\n",
            "Epoch 24/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.1881 - accuracy: 0.9282 - val_loss: 0.1928 - val_accuracy: 0.9252\n",
            "Epoch 25/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.1871 - accuracy: 0.9283 - val_loss: 0.2605 - val_accuracy: 0.8990\n",
            "Epoch 26/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.1837 - accuracy: 0.9300 - val_loss: 0.2160 - val_accuracy: 0.9133\n",
            "Epoch 27/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.1793 - accuracy: 0.9328 - val_loss: 0.1782 - val_accuracy: 0.9347\n",
            "Epoch 28/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.1760 - accuracy: 0.9341 - val_loss: 0.1767 - val_accuracy: 0.9344\n",
            "Epoch 29/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.1750 - accuracy: 0.9348 - val_loss: 0.1732 - val_accuracy: 0.9347\n",
            "Epoch 30/30\n",
            "978/978 [==============================] - 108s 110ms/step - loss: 0.1748 - accuracy: 0.9349 - val_loss: 0.1950 - val_accuracy: 0.9224\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "filename='VGG19.csv'\n",
        "opt = Adam(learning_rate=0.001)\n",
        "history_logger=tf.keras.callbacks.CSVLogger(filename, separator=\",\", append=False)\n",
        "base_model4 =VGG19(input_shape=(100,100,3), weights='imagenet', include_top=False)\n",
        "modelv19=Sequential()\n",
        "modelv19.add(base_model4)\n",
        "modelv19.add(Dropout(0.3))\n",
        "modelv19.add(Flatten())\n",
        "modelv19.add(Dense(8,kernel_initializer='he_uniform',kernel_regularizer=regularizers.L1(0.0001),activation='relu'))\n",
        "modelv19.add(BatchNormalization())\n",
        "modelv19.add(Dropout(0.3))\n",
        "modelv19.add(Dense(1,activation='sigmoid'))\n",
        "modelv19.compile(optimizer=opt,loss='binary_crossentropy', metrics=['accuracy'])\n",
        "historyv19 = modelv19.fit(x_train_normalize, y_train,validation_split=0.2,shuffle=True, epochs=30, batch_size=32,callbacks=[history_logger])\n",
        "loss_value , accuracy = modelv19.evaluate(x_test_normalize, y_test)\n",
        "print(\"=\"*20)\n",
        "print('Test_loss_value = ' +str(loss_value))\n",
        "print('test_accuracy = ' + str(accuracy))\n",
        "plt.plot(historyv19.history['accuracy'])\n",
        "plt.plot(historyv19.history['val_accuracy'])\n",
        "plt.title('model accuracy')\n",
        "plt.ylabel('accuracy')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'validation'], loc='upper left')\n",
        "plt.show()\n",
        "# summarize history for loss\n",
        "plt.plot(historyv19.history['loss'])\n",
        "plt.plot(historyv19.history['val_loss'])\n",
        "plt.title('model loss')\n",
        "plt.ylabel('loss')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'validation'], loc='upper left')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "id": "b5PpkbRR-DOr",
        "outputId": "d871e275-5415-4343-e01f-92d5b3c23748"
      },
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/tensorflow/keras-applications/vgg19/vgg19_weights_tf_dim_ordering_tf_kernels_notop.h5\n",
            "80142336/80134624 [==============================] - 0s 0us/step\n",
            "80150528/80134624 [==============================] - 0s 0us/step\n",
            "Epoch 1/30\n",
            "978/978 [==============================] - 90s 78ms/step - loss: 0.4107 - accuracy: 0.8363 - val_loss: 1.7344 - val_accuracy: 0.4847\n",
            "Epoch 2/30\n",
            "978/978 [==============================] - 75s 76ms/step - loss: 0.3130 - accuracy: 0.8764 - val_loss: 0.2669 - val_accuracy: 0.9032\n",
            "Epoch 3/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.2858 - accuracy: 0.8878 - val_loss: 0.2770 - val_accuracy: 0.8946\n",
            "Epoch 4/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.2779 - accuracy: 0.8921 - val_loss: 0.4892 - val_accuracy: 0.8783\n",
            "Epoch 5/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.2716 - accuracy: 0.8935 - val_loss: 0.2816 - val_accuracy: 0.8882\n",
            "Epoch 6/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.2686 - accuracy: 0.8937 - val_loss: 0.2528 - val_accuracy: 0.9054\n",
            "Epoch 7/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.2639 - accuracy: 0.8956 - val_loss: 0.4298 - val_accuracy: 0.8785\n",
            "Epoch 8/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.2613 - accuracy: 0.8961 - val_loss: 0.5443 - val_accuracy: 0.7786\n",
            "Epoch 9/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.2574 - accuracy: 0.8980 - val_loss: 0.2417 - val_accuracy: 0.9094\n",
            "Epoch 10/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.2555 - accuracy: 0.8995 - val_loss: 0.2598 - val_accuracy: 0.8976\n",
            "Epoch 11/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.2480 - accuracy: 0.9021 - val_loss: 0.3195 - val_accuracy: 0.8862\n",
            "Epoch 12/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.2403 - accuracy: 0.9041 - val_loss: 0.4922 - val_accuracy: 0.8002\n",
            "Epoch 13/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.2370 - accuracy: 0.9050 - val_loss: 0.3402 - val_accuracy: 0.8544\n",
            "Epoch 14/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.2264 - accuracy: 0.9121 - val_loss: 0.2065 - val_accuracy: 0.9183\n",
            "Epoch 15/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.2307 - accuracy: 0.9078 - val_loss: 0.2051 - val_accuracy: 0.9205\n",
            "Epoch 16/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.2239 - accuracy: 0.9123 - val_loss: 0.2241 - val_accuracy: 0.9110\n",
            "Epoch 17/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.2213 - accuracy: 0.9127 - val_loss: 0.2193 - val_accuracy: 0.9116\n",
            "Epoch 18/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.2202 - accuracy: 0.9132 - val_loss: 0.1904 - val_accuracy: 0.9275\n",
            "Epoch 19/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.2175 - accuracy: 0.9148 - val_loss: 0.2109 - val_accuracy: 0.9179\n",
            "Epoch 20/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.2154 - accuracy: 0.9148 - val_loss: 0.2088 - val_accuracy: 0.9228\n",
            "Epoch 21/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.2095 - accuracy: 0.9186 - val_loss: 0.3293 - val_accuracy: 0.8574\n",
            "Epoch 22/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.2078 - accuracy: 0.9188 - val_loss: 0.2194 - val_accuracy: 0.9141\n",
            "Epoch 23/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.2064 - accuracy: 0.9187 - val_loss: 0.2082 - val_accuracy: 0.9196\n",
            "Epoch 24/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.2029 - accuracy: 0.9208 - val_loss: 0.2085 - val_accuracy: 0.9211\n",
            "Epoch 25/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.2040 - accuracy: 0.9223 - val_loss: 0.1985 - val_accuracy: 0.9252\n",
            "Epoch 26/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.2005 - accuracy: 0.9231 - val_loss: 0.2465 - val_accuracy: 0.9040\n",
            "Epoch 27/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.1994 - accuracy: 0.9223 - val_loss: 0.2602 - val_accuracy: 0.9070\n",
            "Epoch 28/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.1959 - accuracy: 0.9238 - val_loss: 0.2134 - val_accuracy: 0.9149\n",
            "Epoch 29/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.1956 - accuracy: 0.9253 - val_loss: 0.1978 - val_accuracy: 0.9238\n",
            "Epoch 30/30\n",
            "978/978 [==============================] - 74s 76ms/step - loss: 0.1924 - accuracy: 0.9264 - val_loss: 0.2018 - val_accuracy: 0.9224\n",
            "524/524 [==============================] - 13s 25ms/step - loss: 0.2012 - accuracy: 0.9254\n",
            "====================\n",
            "Test_loss_value = 0.2011735588312149\n",
            "test_accuracy = 0.9253847002983093\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "loss_value , accuracy = modelv19.evaluate(x_test_normalize, y_test)\n",
        "print(\"=\"*20)\n",
        "print('Test_loss_value = ' +str(loss_value))\n",
        "print('test_accuracy = ' + str(accuracy))\n",
        "plt.plot(historyv19.history['accuracy'])\n",
        "plt.plot(historyv19.history['val_accuracy'])\n",
        "plt.title('model accuracy')\n",
        "plt.ylabel('accuracy')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'validation'], loc='upper left')\n",
        "plt.show()\n",
        "# summarize history for loss\n",
        "plt.plot(historyv19.history['loss'])\n",
        "plt.plot(historyv19.history['val_loss'])\n",
        "plt.title('model loss')\n",
        "plt.ylabel('loss')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'validation'], loc='upper left')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 644
        },
        "id": "ApBFk4kuqyln",
        "outputId": "43530c67-68ae-4801-9e66-a799592d482f"
      },
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "524/524 [==============================] - 13s 25ms/step - loss: 0.2012 - accuracy: 0.9254\n",
            "====================\n",
            "Test_loss_value = 0.2011735588312149\n",
            "test_accuracy = 0.9253847002983093\n"
          ]
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "id": "6afwMUcUYYvu",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "75066b5d-69b0-43e3-e19a-e5d8ac95ad8c"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Collecting tensorflow-addons\n",
            "  Downloading tensorflow_addons-0.17.0-cp37-cp37m-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (1.1 MB)\n",
            "\u001b[?25l\r\u001b[K     |▎                               | 10 kB 23.6 MB/s eta 0:00:01\r\u001b[K     |▋                               | 20 kB 28.6 MB/s eta 0:00:01\r\u001b[K     |▉                               | 30 kB 12.9 MB/s eta 0:00:01\r\u001b[K     |█▏                              | 40 kB 10.0 MB/s eta 0:00:01\r\u001b[K     |█▌                              | 51 kB 5.9 MB/s eta 0:00:01\r\u001b[K     |█▊                              | 61 kB 6.9 MB/s eta 0:00:01\r\u001b[K     |██                              | 71 kB 7.5 MB/s eta 0:00:01\r\u001b[K     |██▍                             | 81 kB 6.0 MB/s eta 0:00:01\r\u001b[K     |██▋                             | 92 kB 6.7 MB/s eta 0:00:01\r\u001b[K     |███                             | 102 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |███▎                            | 112 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |███▌                            | 122 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |███▉                            | 133 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |████▏                           | 143 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |████▍                           | 153 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |████▊                           | 163 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |█████                           | 174 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |█████▎                          | 184 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |█████▋                          | 194 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |██████                          | 204 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |██████▏                         | 215 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |██████▌                         | 225 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |██████▉                         | 235 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |███████                         | 245 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |███████▍                        | 256 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |███████▋                        | 266 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |████████                        | 276 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |████████▎                       | 286 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |████████▌                       | 296 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |████████▉                       | 307 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |█████████▏                      | 317 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |█████████▍                      | 327 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |█████████▊                      | 337 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |██████████                      | 348 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |██████████▎                     | 358 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |██████████▋                     | 368 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |███████████                     | 378 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |███████████▏                    | 389 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |███████████▌                    | 399 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |███████████▉                    | 409 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |████████████                    | 419 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |████████████▍                   | 430 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |████████████▊                   | 440 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |█████████████                   | 450 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |█████████████▎                  | 460 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |█████████████▋                  | 471 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |█████████████▉                  | 481 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |██████████████▏                 | 491 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |██████████████▍                 | 501 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |██████████████▊                 | 512 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |███████████████                 | 522 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |███████████████▎                | 532 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |███████████████▋                | 542 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |████████████████                | 552 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |████████████████▏               | 563 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |████████████████▌               | 573 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |████████████████▉               | 583 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |█████████████████               | 593 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |█████████████████▍              | 604 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |█████████████████▊              | 614 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |██████████████████              | 624 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |██████████████████▎             | 634 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |██████████████████▋             | 645 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |██████████████████▉             | 655 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |███████████████████▏            | 665 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |███████████████████▌            | 675 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |███████████████████▊            | 686 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |████████████████████            | 696 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |████████████████████▍           | 706 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |████████████████████▋           | 716 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |█████████████████████           | 727 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |█████████████████████▏          | 737 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |█████████████████████▌          | 747 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |█████████████████████▉          | 757 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |██████████████████████          | 768 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |██████████████████████▍         | 778 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |██████████████████████▊         | 788 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |███████████████████████         | 798 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |███████████████████████▎        | 808 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |███████████████████████▋        | 819 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |███████████████████████▉        | 829 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |████████████████████████▏       | 839 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |████████████████████████▌       | 849 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |████████████████████████▊       | 860 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |█████████████████████████       | 870 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |█████████████████████████▍      | 880 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |█████████████████████████▋      | 890 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |██████████████████████████      | 901 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |██████████████████████████▎     | 911 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |██████████████████████████▌     | 921 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |██████████████████████████▉     | 931 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |███████████████████████████▏    | 942 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |███████████████████████████▍    | 952 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |███████████████████████████▊    | 962 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |████████████████████████████    | 972 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |████████████████████████████▎   | 983 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |████████████████████████████▋   | 993 kB 7.4 MB/s eta 0:00:01\r\u001b[K     |████████████████████████████▉   | 1.0 MB 7.4 MB/s eta 0:00:01\r\u001b[K     |█████████████████████████████▏  | 1.0 MB 7.4 MB/s eta 0:00:01\r\u001b[K     |█████████████████████████████▌  | 1.0 MB 7.4 MB/s eta 0:00:01\r\u001b[K     |█████████████████████████████▊  | 1.0 MB 7.4 MB/s eta 0:00:01\r\u001b[K     |██████████████████████████████  | 1.0 MB 7.4 MB/s eta 0:00:01\r\u001b[K     |██████████████████████████████▍ | 1.1 MB 7.4 MB/s eta 0:00:01\r\u001b[K     |██████████████████████████████▋ | 1.1 MB 7.4 MB/s eta 0:00:01\r\u001b[K     |███████████████████████████████ | 1.1 MB 7.4 MB/s eta 0:00:01\r\u001b[K     |███████████████████████████████▎| 1.1 MB 7.4 MB/s eta 0:00:01\r\u001b[K     |███████████████████████████████▌| 1.1 MB 7.4 MB/s eta 0:00:01\r\u001b[K     |███████████████████████████████▉| 1.1 MB 7.4 MB/s eta 0:00:01\r\u001b[K     |████████████████████████████████| 1.1 MB 7.4 MB/s \n",
            "\u001b[?25hRequirement already satisfied: packaging in /usr/local/lib/python3.7/dist-packages (from tensorflow-addons) (21.3)\n",
            "Requirement already satisfied: typeguard>=2.7 in /usr/local/lib/python3.7/dist-packages (from tensorflow-addons) (2.7.1)\n",
            "Requirement already satisfied: pyparsing!=3.0.5,>=2.0.2 in /usr/local/lib/python3.7/dist-packages (from packaging->tensorflow-addons) (3.0.9)\n",
            "Installing collected packages: tensorflow-addons\n",
            "Successfully installed tensorflow-addons-0.17.0\n"
          ]
        }
      ],
      "source": [
        "!pip install -U tensorflow-addons  "
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "id": "Ex8vF1ZwaD8b"
      },
      "outputs": [],
      "source": [
        "import tensorflow_addons as tfa\n",
        "import keras"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {
        "id": "CskvaDXaaKmM"
      },
      "outputs": [],
      "source": [
        "num_classes = 2\n",
        "input_shape = (100, 100, 3)\n",
        "learning_rate = 0.001\n",
        "weight_decay = 0.0001\n",
        "batch_size = 32\n",
        "num_epochs = 30\n",
        "image_size = 100  # We'll resize input images to this size\n",
        "patch_size = 4 # Size of the patches to be extract from the input images\n",
        "num_patches = (image_size // patch_size) ** 2\n",
        "projection_dim = 32\n",
        "filename='transformer_vision.csv'\n",
        "history_logger=tf.keras.callbacks.CSVLogger(filename, separator=\",\", append=False)\n",
        "num_heads = 4\n",
        "transformer_units = [\n",
        "    projection_dim * 2,\n",
        "    projection_dim,\n",
        "]  # Size of the transformer layers\n",
        "transformer_layers = 1\n",
        "mlp_head_units = [2048, 1024]  # Size of the dense layers of the final classifier"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "metadata": {
        "id": "q6fAFRi8dIqn"
      },
      "outputs": [],
      "source": [
        "def mlp(x, hidden_units, dropout_rate):\n",
        "    for units in hidden_units:\n",
        "        x = layers.Dense(units, activation=tf.nn.gelu)(x)\n",
        "        x = layers.Dropout(dropout_rate)(x)\n",
        "    return x"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 11,
      "metadata": {
        "id": "3xqh19BNddza"
      },
      "outputs": [],
      "source": [
        "class Patches(layers.Layer):\n",
        "    def __init__(self, patch_size):\n",
        "        super(Patches, self).__init__()\n",
        "        self.patch_size = patch_size\n",
        "\n",
        "    def call(self, images):\n",
        "        batch_size = tf.shape(images)[0]\n",
        "        patches = tf.image.extract_patches(\n",
        "            images=images,\n",
        "            sizes=[1, self.patch_size, self.patch_size, 1],\n",
        "            strides=[1, self.patch_size, self.patch_size, 1],\n",
        "            rates=[1, 1, 1, 1],\n",
        "            padding=\"VALID\",\n",
        "        )\n",
        "        patch_dims = patches.shape[-1]\n",
        "        patches = tf.reshape(patches, [batch_size, -1, patch_dims])\n",
        "        return patches"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 12,
      "metadata": {
        "id": "i72vjco9d0ti"
      },
      "outputs": [],
      "source": [
        "class PatchEncoder(layers.Layer):\n",
        "    def __init__(self, num_patches, projection_dim):\n",
        "        super(PatchEncoder, self).__init__()\n",
        "        self.num_patches = num_patches\n",
        "        self.projection = layers.Dense(units=projection_dim)\n",
        "        self.position_embedding = layers.Embedding(\n",
        "            input_dim=num_patches, output_dim=projection_dim\n",
        "        )\n",
        "\n",
        "    def call(self, patch):\n",
        "        positions = tf.range(start=0, limit=self.num_patches, delta=1)\n",
        "        encoded = self.projection(patch) + self.position_embedding(positions)\n",
        "        return encoded\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 13,
      "metadata": {
        "id": "FFPxI3Sgd8fc"
      },
      "outputs": [],
      "source": [
        "def create_vit_classifier():\n",
        "    inputs = layers.Input(shape=input_shape)\n",
        "    # Create patches.\n",
        "    patches = Patches(patch_size)(inputs)\n",
        "    # Encode patches.\n",
        "    encoded_patches = PatchEncoder(num_patches, projection_dim)(patches)\n",
        "    # Create multiple layers of the Transformer block.\n",
        "    for _ in range(transformer_layers):\n",
        "        # Layer normalization 1.\n",
        "        x1 = layers.LayerNormalization(epsilon=1e-6)(encoded_patches)\n",
        "        # Create a multi-head attention layer.\n",
        "        attention_output = layers.MultiHeadAttention(\n",
        "            num_heads=num_heads, key_dim=projection_dim, dropout=0.1\n",
        "        )(x1, x1)\n",
        "        # Skip connection 1.\n",
        "        x2 = layers.Add()([attention_output, encoded_patches])\n",
        "        # Layer normalization 2.\n",
        "        x3 = layers.LayerNormalization(epsilon=1e-6)(x2)\n",
        "        # MLP.\n",
        "        x3 = mlp(x3, hidden_units=transformer_units, dropout_rate=0.1)\n",
        "        # Skip connection 2.\n",
        "        encoded_patches = layers.Add()([x3, x2])\n",
        "\n",
        "    # Create a [batch_size, projection_dim] tensor.\n",
        "    representation = layers.LayerNormalization(epsilon=1e-6)(encoded_patches)\n",
        "    representation = layers.Flatten()(representation)\n",
        "    representation = layers.Dropout(0.5)(representation)\n",
        "    # Add MLP.\n",
        "    features = mlp(representation, hidden_units=mlp_head_units, dropout_rate=0.5)\n",
        "    # Classify outputs.\n",
        "    output= layers.Dense(1,activation='sigmoid')(features)\n",
        "    # Create the Keras model.\n",
        "    model = Model(inputs=inputs, outputs=output)\n",
        "    return model"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 14,
      "metadata": {
        "id": "hmIg7E2QeBS_",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "a6a8b3ab-e0b3-4f53-8321-76e7363387f1"
      },
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/keras/optimizer_v2/adam.py:105: UserWarning: The `lr` argument is deprecated, use `learning_rate` instead.\n",
            "  super(Adam, self).__init__(name, **kwargs)\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/30\n",
            "978/978 [==============================] - 48s 42ms/step - loss: 0.6337 - accuracy: 0.8557 - val_loss: 0.2810 - val_accuracy: 0.8795\n",
            "Epoch 2/30\n",
            "978/978 [==============================] - 39s 40ms/step - loss: 0.3025 - accuracy: 0.8765 - val_loss: 0.2644 - val_accuracy: 0.8856\n",
            "Epoch 3/30\n",
            "978/978 [==============================] - 39s 40ms/step - loss: 0.2888 - accuracy: 0.8821 - val_loss: 0.2730 - val_accuracy: 0.8832\n",
            "Epoch 4/30\n",
            "978/978 [==============================] - 39s 40ms/step - loss: 0.2767 - accuracy: 0.8885 - val_loss: 0.2587 - val_accuracy: 0.8891\n",
            "Epoch 5/30\n",
            "978/978 [==============================] - 39s 40ms/step - loss: 0.2737 - accuracy: 0.8897 - val_loss: 0.2514 - val_accuracy: 0.8998\n",
            "Epoch 6/30\n",
            "978/978 [==============================] - 39s 40ms/step - loss: 0.2754 - accuracy: 0.8912 - val_loss: 0.2552 - val_accuracy: 0.8947\n",
            "Epoch 7/30\n",
            "978/978 [==============================] - 39s 40ms/step - loss: 0.2712 - accuracy: 0.8904 - val_loss: 0.2629 - val_accuracy: 0.8931\n",
            "Epoch 8/30\n",
            "978/978 [==============================] - 39s 40ms/step - loss: 0.2708 - accuracy: 0.8928 - val_loss: 0.2461 - val_accuracy: 0.8989\n",
            "Epoch 9/30\n",
            "978/978 [==============================] - 39s 39ms/step - loss: 0.2630 - accuracy: 0.8944 - val_loss: 0.2467 - val_accuracy: 0.8983\n",
            "Epoch 10/30\n",
            "978/978 [==============================] - 38s 39ms/step - loss: 0.2616 - accuracy: 0.8970 - val_loss: 0.2484 - val_accuracy: 0.8926\n",
            "Epoch 11/30\n",
            "978/978 [==============================] - 38s 39ms/step - loss: 0.2545 - accuracy: 0.8993 - val_loss: 0.2423 - val_accuracy: 0.9061\n",
            "Epoch 12/30\n",
            "978/978 [==============================] - 38s 39ms/step - loss: 0.2540 - accuracy: 0.9001 - val_loss: 0.2468 - val_accuracy: 0.9001\n",
            "Epoch 13/30\n",
            "978/978 [==============================] - 38s 39ms/step - loss: 0.2573 - accuracy: 0.8979 - val_loss: 0.2500 - val_accuracy: 0.8884\n",
            "Epoch 14/30\n",
            "978/978 [==============================] - 38s 39ms/step - loss: 0.2474 - accuracy: 0.9007 - val_loss: 0.2461 - val_accuracy: 0.9035\n",
            "Epoch 15/30\n",
            "978/978 [==============================] - 38s 39ms/step - loss: 0.2501 - accuracy: 0.9003 - val_loss: 0.2484 - val_accuracy: 0.8926\n",
            "Epoch 16/30\n",
            "978/978 [==============================] - 38s 39ms/step - loss: 0.2478 - accuracy: 0.9015 - val_loss: 0.2420 - val_accuracy: 0.9038\n",
            "Epoch 17/30\n",
            "978/978 [==============================] - 38s 39ms/step - loss: 0.2451 - accuracy: 0.9023 - val_loss: 0.2382 - val_accuracy: 0.9044\n",
            "Epoch 18/30\n",
            "978/978 [==============================] - 38s 39ms/step - loss: 0.2424 - accuracy: 0.9049 - val_loss: 0.2387 - val_accuracy: 0.9015\n",
            "Epoch 19/30\n",
            "978/978 [==============================] - 38s 39ms/step - loss: 0.2429 - accuracy: 0.9019 - val_loss: 0.2360 - val_accuracy: 0.9012\n",
            "Epoch 20/30\n",
            "978/978 [==============================] - 38s 39ms/step - loss: 0.2405 - accuracy: 0.9063 - val_loss: 0.2428 - val_accuracy: 0.9025\n",
            "Epoch 21/30\n",
            "978/978 [==============================] - 38s 39ms/step - loss: 0.2385 - accuracy: 0.9050 - val_loss: 0.2351 - val_accuracy: 0.9081\n",
            "Epoch 22/30\n",
            "978/978 [==============================] - 38s 39ms/step - loss: 0.2408 - accuracy: 0.9056 - val_loss: 0.2472 - val_accuracy: 0.8966\n",
            "Epoch 23/30\n",
            "978/978 [==============================] - 38s 39ms/step - loss: 0.2387 - accuracy: 0.9044 - val_loss: 0.2479 - val_accuracy: 0.9018\n",
            "Epoch 24/30\n",
            "978/978 [==============================] - 38s 39ms/step - loss: 0.2371 - accuracy: 0.9068 - val_loss: 0.2364 - val_accuracy: 0.9029\n",
            "Epoch 25/30\n",
            "978/978 [==============================] - 38s 39ms/step - loss: 0.2336 - accuracy: 0.9069 - val_loss: 0.2421 - val_accuracy: 0.8976\n",
            "Epoch 26/30\n",
            "978/978 [==============================] - 38s 39ms/step - loss: 0.2312 - accuracy: 0.9092 - val_loss: 0.2413 - val_accuracy: 0.8993\n",
            "Epoch 27/30\n",
            "978/978 [==============================] - 38s 39ms/step - loss: 0.2334 - accuracy: 0.9071 - val_loss: 0.2435 - val_accuracy: 0.8993\n",
            "Epoch 28/30\n",
            "978/978 [==============================] - 38s 39ms/step - loss: 0.2414 - accuracy: 0.9069 - val_loss: 0.2475 - val_accuracy: 0.8956\n",
            "Epoch 29/30\n",
            "978/978 [==============================] - 38s 39ms/step - loss: 0.2316 - accuracy: 0.9085 - val_loss: 0.2391 - val_accuracy: 0.9029\n",
            "Epoch 30/30\n",
            "978/978 [==============================] - 38s 39ms/step - loss: 0.2322 - accuracy: 0.9091 - val_loss: 0.2428 - val_accuracy: 0.9022\n",
            "524/524 [==============================] - 7s 13ms/step - loss: 0.2388 - accuracy: 0.9014\n",
            "====================\n",
            "Test_loss_value = 0.2387504130601883\n",
            "test_accuracy = 0.9014075994491577\n"
          ]
        }
      ],
      "source": [
        "def run_experiment(model):\n",
        "    optimizer = tf.keras.optimizers.Adam(\n",
        "        lr=learning_rate\n",
        "    )\n",
        "\n",
        "    model.compile(optimizer=optimizer,loss='binary_crossentropy', metrics=['accuracy'])\n",
        "    checkpoint_filepath = \"/tmp/checkpoint\"\n",
        "    history = model.fit(\n",
        "        x=x_train_normalize,\n",
        "        y=y_train,\n",
        "        batch_size=batch_size,\n",
        "        epochs=num_epochs,\n",
        "        validation_split=0.2,\n",
        "        callbacks=[history_logger]\n",
        "    )\n",
        "    loss_value , accuracy = model.evaluate(x_test_normalize, y_test)\n",
        "    print(\"=\"*20)\n",
        "    print('Test_loss_value = ' +str(loss_value))\n",
        "    print('test_accuracy = ' + str(accuracy))\n",
        "    return history\n",
        "vit_classifier = create_vit_classifier()\n",
        "historyvt = run_experiment(vit_classifier)\n"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(historyvt.history['accuracy'])\n",
        "plt.plot(historyvt.history['val_accuracy'])\n",
        "plt.title('model accuracy')\n",
        "plt.ylabel('accuracy')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'validation'], loc='upper left')\n",
        "plt.show()\n",
        "# summarize history for loss\n",
        "plt.plot(historyvt.history['loss'])\n",
        "plt.plot(historyvt.history['val_loss'])\n",
        "plt.title('model loss')\n",
        "plt.ylabel('loss')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'validation'], loc='upper left')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 573
        },
        "id": "ovQj7py4TVqs",
        "outputId": "cf9e501b-2773-4f39-d294-aef1a351dcb6"
      },
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "num_classes = 2\n",
        "input_shape = (100,100,3)\n",
        "weight_decay = 0.0001\n",
        "learning_rate = 0.001\n",
        "label_smoothing = 0.1\n",
        "validation_split = 0.2\n",
        "batch_size = 32\n",
        "num_epochs =30\n",
        "patch_size = 2  # Size of the patches to be extracted from the input images.\n",
        "num_patches = (input_shape[0] // patch_size) ** 2  # Number of patch\n",
        "embedding_dim = 64  # Number of hidden units.\n",
        "mlp_dim = 64\n",
        "dim_coefficient = 4\n",
        "num_heads = 4\n",
        "attention_dropout = 0.2\n",
        "projection_dropout = 0.2\n",
        "num_transformer_blocks = 1  # Number of repetitions of the transformer layer\n",
        "filename='EAN.csv'\n",
        "history_logger=tf.keras.callbacks.CSVLogger(filename, separator=\",\", append=False)\n",
        "class PatchExtract(layers.Layer):\n",
        "    def __init__(self, patch_size, **kwargs):\n",
        "        super(PatchExtract, self).__init__(**kwargs)\n",
        "        self.patch_size = patch_size\n",
        "\n",
        "    def call(self, images):\n",
        "        batch_size = tf.shape(images)[0]\n",
        "        patches = tf.image.extract_patches(\n",
        "            images=images,\n",
        "            sizes=(1, self.patch_size, self.patch_size, 1),\n",
        "            strides=(1, self.patch_size, self.patch_size, 1),\n",
        "            rates=(1, 1, 1, 1),\n",
        "            padding=\"VALID\",\n",
        "        )\n",
        "        patch_dim = patches.shape[-1]\n",
        "        patch_num = patches.shape[1]\n",
        "        return tf.reshape(patches, (batch_size, patch_num * patch_num, patch_dim))\n",
        "class PatchEmbedding(layers.Layer):\n",
        "    def __init__(self, num_patch, embed_dim, **kwargs):\n",
        "        super(PatchEmbedding, self).__init__(**kwargs)\n",
        "        self.num_patch = num_patch\n",
        "        self.proj = layers.Dense(embed_dim)\n",
        "        self.pos_embed = layers.Embedding(input_dim=num_patch, output_dim=embed_dim)\n",
        "\n",
        "    def call(self, patch):\n",
        "        pos = tf.range(start=0, limit=self.num_patch, delta=1)\n",
        "        return self.proj(patch) + self.pos_embed(pos)"
      ],
      "metadata": {
        "id": "8UUjMNvg64zB"
      },
      "execution_count": 30,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "def external_attention(x, dim, num_heads, dim_coefficient=4, attention_dropout=0, projection_dropout=0):\n",
        "    _, num_patch, channel = x.shape\n",
        "    assert dim % num_heads == 0\n",
        "    num_heads = num_heads * dim_coefficient\n",
        "\n",
        "    x = layers.Dense(dim * dim_coefficient)(x)\n",
        "    # create tensor [batch_size, num_patches, num_heads, dim*dim_coefficient//num_heads]\n",
        "    x = tf.reshape(\n",
        "        x, shape=(-1, num_patch, num_heads, dim * dim_coefficient // num_heads)\n",
        "    )\n",
        "    x = tf.transpose(x, perm=[0, 2, 1, 3])\n",
        "    # a linear layer M_k\n",
        "    attn = layers.Dense(dim // dim_coefficient)(x)\n",
        "    # normalize attention map\n",
        "    attn = layers.Softmax(axis=2)(attn)\n",
        "    # dobule-normalization\n",
        "    attn = attn / (1e-9 + tf.reduce_sum(attn, axis=-1, keepdims=True))\n",
        "    attn = layers.Dropout(attention_dropout)(attn)\n",
        "    # a linear layer M_v\n",
        "    x = layers.Dense(dim * dim_coefficient // num_heads)(attn)\n",
        "    x = tf.transpose(x, perm=[0, 2, 1, 3])\n",
        "    x = tf.reshape(x, [-1, num_patch, dim * dim_coefficient])\n",
        "    # a linear layer to project original dim\n",
        "    x = layers.Dense(dim)(x)\n",
        "    x = layers.Dropout(projection_dropout)(x)\n",
        "    return x\n",
        "def mlp(x, embedding_dim, mlp_dim, drop_rate=0.2):\n",
        "  x = layers.Dense(mlp_dim, activation=tf.nn.gelu)(x)\n",
        "  x = layers.Dropout(drop_rate)(x)\n",
        "  x = layers.Dense(embedding_dim)(x)\n",
        "  x = layers.Dropout(drop_rate)(x)\n",
        "  return x\n",
        "def transformer_encoder(\n",
        "    x,\n",
        "    embedding_dim,\n",
        "    mlp_dim,\n",
        "    num_heads,\n",
        "    dim_coefficient,\n",
        "    attention_dropout,\n",
        "    projection_dropout,\n",
        "    attention_type=\"external_attention\",\n",
        "):\n",
        "    residual_1 = x\n",
        "    x = layers.LayerNormalization(epsilon=1e-5)(x)\n",
        "    if attention_type == \"external_attention\":\n",
        "        x = external_attention(\n",
        "            x,\n",
        "            embedding_dim,\n",
        "            num_heads,\n",
        "            dim_coefficient,\n",
        "            attention_dropout,\n",
        "            projection_dropout,\n",
        "        )\n",
        "    elif attention_type == \"self_attention\":\n",
        "        x = layers.MultiHeadAttention(\n",
        "            num_heads=num_heads, key_dim=embedding_dim, dropout=attention_dropout\n",
        "        )(x, x)\n",
        "    x = layers.add([x, residual_1])\n",
        "    residual_2 = x\n",
        "    x = layers.LayerNormalization(epsilon=1e-5)(x)\n",
        "    x = mlp(x, embedding_dim, mlp_dim)\n",
        "    x = layers.add([x, residual_2])\n",
        "    return x\n",
        "def get_model(attention_type=\"external_attention\"):\n",
        "    inputs = layers.Input(shape=input_shape)\n",
        "    # Extract patches.\n",
        "    x = PatchExtract(patch_size)(inputs)\n",
        "    # Create patch embedding.\n",
        "    x = PatchEmbedding(num_patches, embedding_dim)(x)\n",
        "    # Create Transformer block.\n",
        "    for _ in range(num_transformer_blocks):\n",
        "        x = transformer_encoder(\n",
        "            x,\n",
        "            embedding_dim,\n",
        "            mlp_dim,\n",
        "            num_heads,\n",
        "            dim_coefficient,\n",
        "            attention_dropout,\n",
        "            projection_dropout,\n",
        "            attention_type,\n",
        "        )\n",
        "\n",
        "    x = layers.GlobalAvgPool1D()(x)\n",
        "    outputs = layers.Dense(1, activation=\"sigmoid\")(x)\n",
        "    model = Model(inputs=inputs, outputs=outputs)\n",
        "    return model\n",
        "modelean = get_model(attention_type=\"external_attention\")\n",
        "optimizer = tf.keras.optimizers.Adam(\n",
        "        lr=learning_rate,decay=weight_decay)\n",
        "modelean.compile(\n",
        "    loss='binary_crossentropy',metrics=['accuracy'],\n",
        "    optimizer=optimizer,\n",
        ")\n",
        "\n",
        "historyean = modelean.fit(x=x_train_normalize,y=y_train,batch_size=batch_size,epochs=num_epochs,validation_split=0.2,callbacks=[history_logger])\n",
        "loss_value , accuracy = modelean.evaluate(x_test_normalize, y_test)\n",
        "print(\"=\"*20)\n",
        "print('Test_loss_value = ' +str(loss_value))\n",
        "print('test_accuracy = ' + str(accuracy))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "p9eXMaPmCkGa",
        "outputId": "8d7b146b-bf9a-4ce0-e5dd-065c48ff9dfd"
      },
      "execution_count": 34,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.7/dist-packages/keras/optimizer_v2/adam.py:105: UserWarning: The `lr` argument is deprecated, use `learning_rate` instead.\n",
            "  super(Adam, self).__init__(name, **kwargs)\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch 1/30\n",
            "978/978 [==============================] - 72s 71ms/step - loss: 0.2843 - accuracy: 0.8766 - val_loss: 0.2687 - val_accuracy: 0.8861\n",
            "Epoch 2/30\n",
            "978/978 [==============================] - 69s 71ms/step - loss: 0.2520 - accuracy: 0.8895 - val_loss: 0.2638 - val_accuracy: 0.8889\n",
            "Epoch 3/30\n",
            "978/978 [==============================] - 69s 70ms/step - loss: 0.2378 - accuracy: 0.8983 - val_loss: 0.2815 - val_accuracy: 0.8956\n",
            "Epoch 4/30\n",
            "978/978 [==============================] - 69s 70ms/step - loss: 0.2305 - accuracy: 0.9031 - val_loss: 0.2444 - val_accuracy: 0.9059\n",
            "Epoch 5/30\n",
            "978/978 [==============================] - 69s 70ms/step - loss: 0.2242 - accuracy: 0.9058 - val_loss: 0.2516 - val_accuracy: 0.9070\n",
            "Epoch 6/30\n",
            "978/978 [==============================] - 69s 70ms/step - loss: 0.2204 - accuracy: 0.9092 - val_loss: 0.2995 - val_accuracy: 0.8997\n",
            "Epoch 7/30\n",
            "978/978 [==============================] - 69s 71ms/step - loss: 0.2152 - accuracy: 0.9125 - val_loss: 0.2585 - val_accuracy: 0.9054\n",
            "Epoch 8/30\n",
            "978/978 [==============================] - 69s 70ms/step - loss: 0.2122 - accuracy: 0.9136 - val_loss: 0.2594 - val_accuracy: 0.9073\n",
            "Epoch 9/30\n",
            "978/978 [==============================] - 69s 70ms/step - loss: 0.2106 - accuracy: 0.9139 - val_loss: 0.2447 - val_accuracy: 0.9099\n",
            "Epoch 10/30\n",
            "978/978 [==============================] - 69s 70ms/step - loss: 0.2089 - accuracy: 0.9144 - val_loss: 0.2708 - val_accuracy: 0.9054\n",
            "Epoch 11/30\n",
            "978/978 [==============================] - 69s 71ms/step - loss: 0.2061 - accuracy: 0.9164 - val_loss: 0.3026 - val_accuracy: 0.9032\n",
            "Epoch 12/30\n",
            "978/978 [==============================] - 69s 70ms/step - loss: 0.2037 - accuracy: 0.9161 - val_loss: 0.2494 - val_accuracy: 0.9085\n",
            "Epoch 13/30\n",
            "978/978 [==============================] - 69s 70ms/step - loss: 0.2028 - accuracy: 0.9170 - val_loss: 0.3061 - val_accuracy: 0.9012\n",
            "Epoch 14/30\n",
            "978/978 [==============================] - 69s 70ms/step - loss: 0.2004 - accuracy: 0.9180 - val_loss: 0.2579 - val_accuracy: 0.9080\n",
            "Epoch 15/30\n",
            "978/978 [==============================] - 69s 70ms/step - loss: 0.1989 - accuracy: 0.9189 - val_loss: 0.2667 - val_accuracy: 0.9080\n",
            "Epoch 16/30\n",
            "978/978 [==============================] - 69s 70ms/step - loss: 0.1977 - accuracy: 0.9193 - val_loss: 0.2633 - val_accuracy: 0.9043\n",
            "Epoch 17/30\n",
            "978/978 [==============================] - 69s 70ms/step - loss: 0.1956 - accuracy: 0.9208 - val_loss: 0.2503 - val_accuracy: 0.9085\n",
            "Epoch 18/30\n",
            "978/978 [==============================] - 69s 70ms/step - loss: 0.1944 - accuracy: 0.9206 - val_loss: 0.3123 - val_accuracy: 0.9048\n",
            "Epoch 19/30\n",
            "978/978 [==============================] - 68s 70ms/step - loss: 0.1933 - accuracy: 0.9205 - val_loss: 0.2781 - val_accuracy: 0.9047\n",
            "Epoch 20/30\n",
            "978/978 [==============================] - 69s 70ms/step - loss: 0.1908 - accuracy: 0.9219 - val_loss: 0.2616 - val_accuracy: 0.9075\n",
            "Epoch 21/30\n",
            "978/978 [==============================] - 69s 70ms/step - loss: 0.1899 - accuracy: 0.9234 - val_loss: 0.2736 - val_accuracy: 0.9063\n",
            "Epoch 22/30\n",
            "978/978 [==============================] - 68s 70ms/step - loss: 0.1902 - accuracy: 0.9242 - val_loss: 0.2618 - val_accuracy: 0.9099\n",
            "Epoch 23/30\n",
            "978/978 [==============================] - 69s 70ms/step - loss: 0.1879 - accuracy: 0.9238 - val_loss: 0.2720 - val_accuracy: 0.9078\n",
            "Epoch 24/30\n",
            "978/978 [==============================] - 68s 70ms/step - loss: 0.1869 - accuracy: 0.9226 - val_loss: 0.2894 - val_accuracy: 0.9063\n",
            "Epoch 25/30\n",
            "978/978 [==============================] - 68s 70ms/step - loss: 0.1850 - accuracy: 0.9247 - val_loss: 0.2676 - val_accuracy: 0.9084\n",
            "Epoch 26/30\n",
            "978/978 [==============================] - 68s 70ms/step - loss: 0.1854 - accuracy: 0.9252 - val_loss: 0.3245 - val_accuracy: 0.9017\n",
            "Epoch 27/30\n",
            "978/978 [==============================] - 68s 70ms/step - loss: 0.1838 - accuracy: 0.9250 - val_loss: 0.3342 - val_accuracy: 0.9002\n",
            "Epoch 28/30\n",
            "978/978 [==============================] - 68s 70ms/step - loss: 0.1819 - accuracy: 0.9253 - val_loss: 0.2661 - val_accuracy: 0.9077\n",
            "Epoch 29/30\n",
            "978/978 [==============================] - 69s 70ms/step - loss: 0.1815 - accuracy: 0.9262 - val_loss: 0.2564 - val_accuracy: 0.9082\n",
            "Epoch 30/30\n",
            "978/978 [==============================] - 69s 70ms/step - loss: 0.1800 - accuracy: 0.9271 - val_loss: 0.2689 - val_accuracy: 0.9073\n",
            "524/524 [==============================] - 12s 23ms/step - loss: 0.2584 - accuracy: 0.9103\n",
            "====================\n",
            "Test_loss_value = 0.25835055112838745\n",
            "test_accuracy = 0.9102946519851685\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "plt.plot(historyean.history['accuracy'])\n",
        "plt.plot(historyean.history['val_accuracy'])\n",
        "plt.title('model accuracy')\n",
        "plt.ylabel('accuracy')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'validation'], loc='upper left')\n",
        "plt.show()\n",
        "# summarize history for loss\n",
        "plt.plot(historyean.history['loss'])\n",
        "plt.plot(historyean.history['val_loss'])\n",
        "plt.title('model loss')\n",
        "plt.ylabel('loss')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'validation'], loc='upper left')\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 573
        },
        "id": "nr0C780WbPHl",
        "outputId": "7a8df499-19dd-49da-8cf2-f162cda23598"
      },
      "execution_count": 35,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ],
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYkAAAEWCAYAAACT7WsrAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nOydd3iV5dnAf3c22RsCCXuFPQICCi60uHBvrbO2jlprl639Omz92qq17laq9utQqdJqUbRuRASRocywAgESRkImELKf74/nnHCIOck5yTk5OYf7d125Ts77Pu97bkLy3s+9xRiDoiiKorRFWKAFUBRFUXouqiQURVEUt6iSUBRFUdyiSkJRFEVxiyoJRVEUxS2qJBRFURS3qJJQFB8gIv8nIr/2cG2hiMzu6n0UpTtQJaEoiqK4RZWEoiiK4hZVEsoJg8PN8wMRWSciR0TkeRHpLSJvi8ghEXlfRFJc1s8VkY0iUikii0Uk1+XcRBFZ47jun0BMq886X0S+dFy7TETGdVLmb4jIdhEpF5GFItLXcVxE5A8iUiIi1SKyXkTGOM6dKyKbHLIVi8j3O/UDUxRUSSgnHpcCZwHDgQuAt4GfABnYv4e7AURkOPAycI/j3FvAGyISJSJRwOvA34FU4FXHfXFcOxF4AfgmkAY8CywUkWhvBBWRM4DfAFcAWcAuYL7j9NnALMe/I8mxpsxx7nngm8aYBGAM8KE3n6sorqiSUE40njTGHDDGFAOfACuMMV8YY2qB14CJjnVXAouMMe8ZYxqAR4BewAxgGhAJPGaMaTDGLABWunzGbcCzxpgVxpgmY8xfgTrHdd5wLfCCMWaNMaYO+DEwXUQGAg1AAjASEGNMvjFmn+O6BmCUiCQaYyqMMWu8/FxFaUGVhHKiccDl+6NtvI93fN8Xu3MHwBjTDOwB+jnOFZvju2Pucvl+APA9h6upUkQqgRzHdd7QWobDWGuhnzHmQ+Ap4GmgRETmiUiiY+mlwLnALhH5WESme/m5itKCKglFaZu92Ic9YGMA2Ad9MbAP6Oc45qS/y/d7gAeNMckuX7HGmJe7KEMc1n1VDGCMecIYMxkYhXU7/cBxfKUx5kIgE+sWe8XLz1WUFlRJKErbvAKcJyJnikgk8D2sy2gZsBxoBO4WkUgRuQSY6nLtn4FvichJjgBznIicJyIJXsrwMnCTiExwxDP+F+seKxSRKY77RwJHgFqg2REzuVZEkhxusmqguQs/B+UER5WEorSBMWYLcB3wJHAQG+S+wBhTb4ypBy4BbgTKsfGLf7tcuwr4BtYdVAFsd6z1Vob3gf8B/oW1XoYAVzlOJ2KVUQXWJVUGPOw4dz1QKCLVwLewsQ1F6RSiQ4cURVEUd6gloSiKorhFlYSiKIriFlUSiqIoiltUSSiKoihuifDnzUVkDvA4EA48Z4z5bavz3wLuBJqAw8BtxphNLuf7A5uAXxhjHmnvs9LT083AgQN9+w9QFEUJcVavXn3QGJPh7rzfsptEJBzYiu2TU4RtW3B1KyWQaIypdnw/F7jDGDPH5fwCwGBzw9tVEnl5eWbVqlW+/4coiqKEMCKy2hiT5+68P91NU4Htxpgdjrzy+cCFrgucCsJBHFYhACAiFwE7gY1+lFFRFEVpB38qiX7Y9gROihzHjkNE7hSRAuAhjnXgjAd+BPyyvQ8QkdtEZJWIrCotLfWZ4IqiKIol4IFrY8zTxpghWKXwU8fhXwB/cDQ0a+/aecaYPGNMXkaGW5eaoiiK0kn8GbguxjZEc5LtOOaO+cAfHd+fBFwmIg8BydieNLXGmKe8EaChoYGioiJqa2u9uUxph5iYGLKzs4mMjAy0KIqidAP+VBIrgWEiMgirHK4CrnFdICLDjDHbHG/PA7YBGGNmuqz5BXDYWwUBUFRUREJCAgMHDuT4hp1KZzDGUFZWRlFREYMGDQq0OIqidAN+UxLGmEYRuQt4B5sC+4IxZqOIPACsMsYsBO4SkdnYISkVwA2+lKG2tlYVhA8REdLS0tD4j6KcOPi1TsIY8xZ27KPrsZ+5fP8dD+7xi67IoArCt+jPU1FOLAIeuFYURQlqGo7CyuehvibQkvgFVRJ+prKykmeeecbr684991wqKyv9IJGiKD6juRlevwMW3Qvb3g20NH5BlYSfcackGhsb273urbfeIjk52V9iKYriCxb/BjY65k1V7Wl/bZCiSsLP3HfffRQUFDBhwgSmTJnCzJkzmTt3LqNGjQLgoosuYvLkyYwePZp58+a1XDdw4EAOHjxIYWEhubm5fOMb32D06NGcffbZHD16NFD/HEVRnKz9Jyx5CCZcB9GJUBmaSsKvgeuexC/f2MimvdUdL/SCUX0T+fkFo9td89vf/pYNGzbw5ZdfsnjxYs477zw2bNjQkkL6wgsvkJqaytGjR5kyZQqXXnopaWlpx91j27ZtvPzyy/z5z3/miiuu4F//+hfXXXedT/8tiqJ4wa7lsPAuGDgTzv8D7P0iZC2JE0ZJ9BSmTp16XI3BE088wWuvvQbAnj172LZt21eUxKBBg5gwYQIAkydPprCwsNvkVRSlFeU74Z/XQlIOXPE3iIiC5By1JIKdjnb83UVcXFzL94sXL+b9999n+fLlxMbGctppp7VZHR4dHd3yfXh4uLqbFCVQHK2El66A5ia49lWITbXHk3Jg9/LAyuYnNCbhZxISEjh06FCb56qqqkhJSSE2NpbNmzfz2WefdbN0iqJ4TFMDvHqDtSSu/AekDTl2Likbaqug1rcu7Z7ACWNJBIq0tDROPvlkxowZQ69evejdu3fLuTlz5vCnP/2J3NxcRowYwbRp0wIoqaIobjEG3voB7FgMFz4Ng2Yefz7Z0aauqghiRnW7eP5ElUQ38NJLL7V5PDo6mrfffrvNc864Q3p6Ohs2bGg5/v3vf9/n8imK0gGfPQOr/wIn3wMT20gaSepvX6v2QO/QUhLqblIURWmPLW/DO/dD7gVw5s/bXuO0JCp3d59c3YQqCUVRFHfsWwcLboGs8XDxPAhz88iMy4TwqJBMg1UloSiK0haH9sPLV0GvZLh6PkTFul8bFgaJ/WxMIsTQmISiKEpbvHM/HK2Am9+BxKyO14dorYRaEoqiKG2xfz0MOQOyxnm2Pqm/upsURVFOCJqboGInpA72/JrkHOuiaqz3n1wBwK9KQkTmiMgWEdkuIve1cf5bIrJeRL4UkaUiMspx/CwRWe04t1pEzvCnnD2J+Ph4APbu3ctll13W5prTTjuNVatWtXufxx57jJqaY/3ttfW4onhBVRE01R9fMNcRSTmAgepiv4kVCPymJEQkHHgaOAcYBVztVAIuvGSMGWuMmQA8BDzqOH4QuMAYMxY70vTv/pKzp9K3b18WLFjQ6etbKwltPa4oXlBeYF9TvVES2fY1xFxO/rQkpgLbjTE7jDH1wHzgQtcFxhjXGvY4wDiOf2GM2es4vhHoJSLRBCH33XcfTz/9dMv7X/ziF/z617/mzDPPZNKkSYwdO5b//Oc/X7musLCQMWPGAHD06FGuuuoqcnNzufjii4/r3XT77beTl5fH6NGj+fnPbQ73E088wd69ezn99NM5/fTTgWOtxwEeffRRxowZw5gxY3jsscdaPk9bkiuKg/Id9tUbS6KlViK0lIQ/s5v6Aa4/rSLgpNaLRORO4F4gCmjLrXQpsMYYU9fGtbcBtwH079+/fWnevs8GonxJn7Fwzm/bXXLllVdyzz33cOeddwLwyiuv8M4773D33XeTmJjIwYMHmTZtGnPnznU7P/qPf/wjsbGx5Ofns27dOiZNmtRy7sEHHyQ1NZWmpibOPPNM1q1bx913382jjz7KRx99RHp6+nH3Wr16NX/5y19YsWIFxhhOOukkTj31VFJSUrQluaI4KdsBEb0gwYOsJieJ2YCoJeFrjDFPG2OGAD8Cfup6TkRGA78Dvunm2nnGmDxjTF5GRob/he0EEydOpKSkhL1797J27VpSUlLo06cPP/nJTxg3bhyzZ8+muLiYAwcOuL3HkiVLWh7W48aNY9y4Y9kWr7zyCpMmTWLixIls3LiRTZs2tSvP0qVLufjii4mLiyM+Pp5LLrmETz75BNCW5IrSQnmBDVq72bi1SUQUJPRRS8ILioEcl/fZjmPumA/80flGRLKB14CvG2MKuixNBzt+f3L55ZezYMEC9u/fz5VXXsmLL75IaWkpq1evJjIykoEDB7bZIrwjdu7cySOPPMLKlStJSUnhxhtv7NR9nGhLckVxUFYAmSO9vy4pWy0JL1gJDBORQSISBVwFLHRdICLDXN6eB2xzHE8GFgH3GWM+9aOM3cKVV17J/PnzWbBgAZdffjlVVVVkZmYSGRnJRx99xK5du9q9ftasWS1NAjds2MC6desAqK6uJi4ujqSkJA4cOHBcs0B3LcpnzpzJ66+/Tk1NDUeOHOG1115j5syZX1mnKCcsTY1QUehd0NpJUk7IKQm/WRLGmEYRuQt4BwgHXjDGbBSRB4BVxpiFwF0iMhtoACqwmUwAdwFDgZ+JyM8cx842xpT4S15/Mnr0aA4dOkS/fv3Iysri2muv5YILLmDs2LHk5eUxcmT7O5bbb7+dm266idzcXHJzc5k8eTIA48ePZ+LEiYwcOZKcnBxOPvnklmtuu+025syZQ9++ffnoo49ajk+aNIkbb7yRqVOnAnDrrbcyceJEdS0pipOqPdDc4F3Q2klyDmx+E5qb3fd5CjLEGBNoGXxCXl6eaV07kJ+fT25uboAkCl3056qENNs/gH9cAje+BQNP7ni9K5//Gd76Pnxvi41PBAEistoYk+fufGioOkVRFF/RmfRXJ8mOLMsQCl6rklAURXGlrAAi4yC+d8drWxOCBXUhryRCxZ3WU9CfpxLydCb91UmSc4ypKomgICYmhrKyMn2w+QhjDGVlZcTExARaFKUtDmyEnZ8EWorgp6wA0rxo7OdKTCLEJIWUuymk50lkZ2dTVFREaWlpoEUJGWJiYsjOzg60GEpbfPAAFK2CH2zv3C5Ysemvlbtg1IUdr3VHiLUMD2klERkZyaBBgwIthqJ0D+U7oeYgHNwGGcMDLU1wUrkLmhs7F7R2kpwDFe3XPgUTIe1uUpQThuZm+4AD2BX09aeBo3ynfe1MIZ2TpOyQGmOqSkJRQoHDB6DR0ZJl9/LAyhLMOFuEd8WSSMqBuiqorfKNTAFGlYSihAIVhfa1VwrsUiXRacoKICoB4rrQMDTEWoarklCUUMDpahpzGVTtDil3R7dSXgCpg7oW+E9yFNSFSPBalYSihAIVhYDA+Kvse7UmOkdZQddcTeBSUBcailqVhHJi0twMn/0RjlYEWhLfUFEIiX2h70SITtTgdWdoaoDK3V0LWoN1VYVH23uFAKoklBOTA+vhv/fBulcCLYlvqCiElIEQFg45UzV43Rkqd4Np6rolERYWUnMlVEkoJyZl2+1rSX5g5fAVTiUBMGAGlG6GmvJAShR8lDkym7pqSYANXmvgWlGCGGenz1BQEg21cGjfMSXRf4Z9VWvCO3yR/uokhIYPqZJQTkzKHEqiNB+CvbeX0/ftVBL9Jlmf+K5lARMpKCkrsPGc2LSu3yspx1G7Utf1ewUYVRJKz+dwKTwywrfN65y7xtoquwsPZpw1EskD7GtENPSbrErCW7rS/bU1zlqJEMhw8quSEJE5IrJFRLaLyH1tnP+WiKwXkS9FZKmIjHI592PHdVtE5Gv+lFPp4ez4CA7vh8KlvrtnWcEx33Owu5ycSsJpSQAMmA771kLd4UBIFJz4Iv3VSQi1DPebkhCRcOBp4BxgFHC1qxJw8JIxZqwxZgLwEPCo49pRwFXAaGAO8IzjfsqJyM6P7WvZNt/cr7bKNsLLPd++L93sm/sGispdENEL4jOPHRsww2bqFK0MnFzBRGO9faD7ImgNIVV17U9LYiqw3RizwxhTD8wHjuu/a4ypdnkbBzidwxcC840xdcaYncB2x/0UTzDG1gGECjuX2NeDW31zP2cWS/ZUm9Nessk39w0UzswmVzdJ9lSQMA1ee0rlLjDNvrMkEvsBopZEB/QDXH9CRY5jxyEid4pIAdaSuNvLa28TkVUiskpnRrjw0YPw59MCLYVvqCi0gdnoJPtw94Xyc51hnDESSoLckqgohJQBxx+LSYQ+YzUu4Sm+TH8FCI+EhCyNSfgCY8zTxpghwI+An3p57TxjTJ4xJi8jowsNuUKNwqXWHx0KefJOK2LC1dBQA9XFXb9nWQEgkDIIMkdZd1OwZjgZc3yNhCv9Z1h3U2N9d0sVfPgy/dVJck5IVF37U0kUAzku77Mdx9wxH7iok9cqTow55j7Z+0VgZfEFO5dAXCaMPM++90VcorzAVsRGxkDmSKg/HLxugZoyK39bSmLAdNs+fN+X3S5W0FFWYMeO9krx3T1DpFbCn0piJTBMRAaJSBQ2EL3QdYGIDHN5ex7gfAIsBK4SkWgRGQQMAz73o6yhw6H9x/rYB/vDwRirJAbNgvQR9tjB7V2/b5mj0ydYSwKCN8PJOQHNnSUB6nLyhHJHtpsvx74m50BVcdDHB/2mJIwxjcBdwDtAPvCKMWajiDwgInMdy+4SkY0i8iVwL3CD49qNwCvAJuC/wJ3GmCZ/yRpStARhBfYGuZI4uNUWJA2aZTN3ohN9E7wud0l/zRhpX4NWSTgmqbWlJOIzIG2YBq89oWyHb11NYC2J5gabvh3E+HXGtTHmLeCtVsd+5vL9d9q59kHgQf9JF6I4H3aDZgW/knDGIwbNsju8tKFddzfVlNvOr84HQq9kSOgbvGmwLYV0/ds+P2A6bPqP3c2GBTwE2TNprHOkv17j2/smuRTUJfb17b27Ef2tCTVK8q0Pf+iZdvhMMAevd35sB7g4d8npw+FgF5WEM7PJNYslc2TwpsFWFNr/76i4ts/3n2Hdj8H67+sOKgoB43tLoqVWIriD16okQo3SfPvQy5pg3wdr8Lq52bbhcFoRAOlDbXZTV6qIy9rIYsnIhdKt0ByEHk13mU1OBmizvw7xdfqrkxCpulYlEUo0N9uc/8xRkDXeHgvW4PWB9VBbaZWEk/Th9rWsC8Hr8gJbZOb6YM3Mhcajx1w3wUTlrvaVRHJ/W9ilwWv3tKS/DvbtfaPjbbZUkFddq5IIJap2Q8MR+9DrlWzrAII1LtESj5h57FiaIxmuK0qizJH+GhF97Fhmrn0NtrhEU4P1d7enJESg/3RrSQRrLYi/KSuwD3Nfpr86CYHhQ6okQgln0NqZ1tl3QvBaEjuXWKXgGvBLHQxI1+ISrplNTjIc6bXB5rev2mNbSbSutm7NgOm2060zE0o5nrZ+J3xFUv+gr7pWJRFKOJWE86GXNcEGzYIteN3UYN0jrq4msMVvKQM6nwZrTNupjtEJ9o852NpztNX9tS1a6iU0LtEm5Tt9H7R24pxQF8RWnCqJUKIkHxKzbeUoWEsCgs+a2PuFrSJurSTAWhedTYOtKYO6qrZ3jZkjg8/d5KmSyBhpXSm7Qzgu0dzcuaK1hlq70/ebJZED9YdsfC1IUSURSpTkH/Ovw7HgdbDFJZytwQfO/Oq59OG26rozD4S2MpucZOZaC6Wp0fv7BoqKXRAeZRvJtUdYmI1LhLIl8e9vwN/mdryuNRU78Uv6q5MQaBmuSiJUaGqEg1uOVxK9UuwuM9gsiZ1LoPdYiGtjjGT6UJuJ1JlGf23VSDjJyIWm+mNrgoGKQpu9FObBqJX+063v/dABv4vV7ZTkw4YFUPgJHPAyrtSS/urjzCYnSdn2NYjjEqokQoWKnfYh56okAPpODC5LoqEWdq9o29UELhlOnXA5lReAhLcd6HX+3IIpeF1ReGxkaUe01EuEoMvp08chMhbCIuHLF727ttzfSsJRCR/EGU6qJEIF58OttZLImmBz6YMleF30OTTVuVcSzlqJzmQ4lRXYnXd4pJv7SnDFJToqpHMla7x9kIaay6lyN6x/FSbfCMO/ButesYkPnlJWALFpNmXcH8Sl26mBQVx1rUoiVCjJB+RYt1QnLcHrtd0uUqfYucTu9p0739a0NPrrpCXhzvccFWs7wwZLo7+jlTYY6qmSCI+E7CmhZ0ksexIQmH4nTLgWjpTA9vc9v758h/+C1mDrVIK8VkKVRKhQssk+5KJijz8ebJXXO5dYF1lMYtvnRSB9mPdpsM701/YeCBm5waMkKttpEe6OATNg/4ZjreSDncOlsOZvMO5K+yAedpYdR+uNy6ncD91fW+NMgw1STnglcfBwHT/7zwZW7woSd4w7Sjbbh1xrnMHrYOjhVHcIile7dzU5SRvmfdX1kVKbitjeAyFzpLU2Guu8u3cgaEl/9TAmATZ4jbExn1BgxZ/s/9Up99j34ZFWYWz5Lxwp6/j6esekQ39aEuCwJDRwHbTERoXz0ordvJ9fEmhROk9jnX1oto5HOMmaEBzB692fQXNjx0qiM43+PGniljnKfn5X2n50F57WSLiSPQXCIkLD5VRbDZ//GXIvsJalkwnX2BkO61/t+B7OCnRf92xqTVJ/6wZrqPXv5/gJVRJREYzPSWZ5gQc7j57KwW1gmtwrib5BErze+bHN+885qf11nWn050kTt2AaQFRRaK1EZ+GkJ0TFWldeKASvV//FFkae8t3jj/cebTdFX/6j43v4q/tra5Jd5koEISe8kgCYPjiN9cVVHK4LokIqV1r3bGpNVpAEr3cugeypX42rtKYzjf7KCuwuOsnNcB6wO1IJDxIl0UH3V3f0nw571wTtrhawsi9/GgafBv0mffX8hGth/3rYt679+/g7/dVJS8vw4Mxw8quSEJE5IrJFRLaLyH1tnL9XRDaJyDoR+UBEBrice8gx2jRfRJ4Q8eXw2eOZPiSNpmbDykIf7bSPlMEz06HgI9/cryNK8+0DMG1o2+eDIXhdU27/qDtyNYH9o5Yw74LX5QW2piC8nWGMEdE2ZhEMabDepL+6MmCGracpXu1ribqPtS/Zsban3Nv2+bGXWYt07cvt36eswAa63SVJ+ApnQV2QBq/9piREJBx4GjgHGAVcLSKtt7pfAHnGmHHAAuAhx7UzgJOBccAYYApwqr9kndQ/hajwMD7zlcvp83k222jzm765X0eU5FsFERHV9vnYVPuA7MlxiV2fAsYzJREZY+sdvEmD9XSGcWYQZDg1N9m8e08L6VxxuvKCdb5EU6Mtnus32f3vSmwqjDgH1v0TGuvd36t8p/9dTWA7GUuYupvaYCqw3RizwxhTD8wHLnRdYIz5yBhT43j7GZDtPAXEAFFANBAJ+K2fQK+ocCbkJLN8hw+URP0R+PxZ+33Ryq7fzxNKNrmPRzjp6W3Ddy6xxV79Jnu23ptGf8Z4ng+fkWvXNhz17N6BoHqvDc52xpKITbVuyWANXm963VpRp3z32MTCtphwrW3ouO1d92vaq5vxJeGRdo56kNZK+FNJ9ANcfypFjmPuuAV4G8AYsxz4CNjn+HrHGPOV7Z2I3CYiq0RkVWlpaZeEnTYkjQ3FVVTXelGt2RZf/AOOVtjmdPs3WKXhT+qP2D8ad/EIJ1kT7LqjFf6Vp7PsXGL95e6sodZ40+jv0H47jMkjS2IkYDrfjrw76ExmkysDZsCez4OrmSFYZb/0MVswOuK89tcOORPie7uvmag/Ymds+Dse4SSIayV6ROBaRK4D8oCHHe+HArlYy6IfcIaIfKUlqDFmnjEmzxiTl5GR0SUZpg9Oo9nAyp1diEs0NcCyp6xJP+PbNuPI3y6e0i321RNLAnpm8PrQARsH8MTV5MSbRn/eBCidyrYnu5w6U0jnSv/pthX7/g4Cuz2Nbe/Zsban3GM727ZHeIStmdj6DhxuI73d2cixOywJsMFrDVx/hWIgx+V9tuPYcYjIbOB+YK4xxlnFdDHwmTHmsDHmMNbCmO5HWZnYP5moiLCupcJufN3+Ipx8D/TLs8eKPveNgO5oGTTUgZJwZjj1xLhE4Sf21Ssl4ezh5MGOv70W4a1JHWwbxfVkJVFRaLOwnAFRbxl0qg3srvmrT8XyO0v/YOeljLnMs/UTrrUbtXWvfPVcd6W/OknKdrgJm7rn83yIP5XESmCYiAwSkSjgKmCh6wIRmQg8i1UQrup+N3CqiESISCQ2aO3Xv9qYyHAm9e9CXMIYG1BLHwHD59g216lDoGiVbwVtTckmCI+2LTnawxm87olxiZ1LIDrpWBaWJ3iTBlteYB+KSTkdrw2PtAqopyuJpH5tNyr0hPgMmHg9fPFi8LhAdn9m4ygzvu25SzJzpI1xffniVyfDtViXHfzd+IrkHFuoeWh/93yeD/GbkjDGNAJ3Ae9gH/CvGGM2isgDIuKcDvIwEA+8KiJfiohTiSwACoD1wFpgrTHmDX/J6mT64HQ27aumsqadjAh3FHxgTeGT7z5mCmdPsb5ff44uLMm340o9mSnQt4dWXu9cAgNP8ezf4KSl0Z+HlkTKQM/vnznSphX3VDqb/uqKswht6aNdlaZ7+ORR26110te9u27CNXYj1XpzVL7DxiyiE3wnY3sEcctwv8YkjDFvGWOGG2OGGGMedBz7mTFmoeP72caY3saYCY6vuY7jTcaYbxpjco0xo4wxbhKifcv0IWkYA593Ji6x9DE7IWzs5ceO5Uyx5fj+bBNcurnjoLWTrAm2FcHRHjRKsXK3lckbVxO4NPrzIMPJ206fmblWLm/afnQnnS2kcyU5ByZeB2v+3vNTM/dvgG3vwEm3d1xo2Zoxl1pL+8uXjj/eUbNHXxPEE+p6ROC6pzA+J4mYyDDvXU7Fq61ffdodtiDLSfYU++qvVNijlTZwmznSs/U9MXi9sxPxCCeeNPprbva+06czvuNMCuhJ1B+xG4+uKgmAmY6919I/dP1e/uTTxyAqHqbe6v21vVJg5Hm2l5Nr48byAv/3bHKlZUJd8AWvVUm4EB0RzuQBKd4Hrz993PrUJ994/PHM0Tb3319KwlkZ7I0lAT2rI+zOJRCb3nF2VlukD+u40d+hvdBY612qo1OWnuhyquhiZpMryf2tO2bN36CqE+NgXaneB6/cYHf9vqR8J2z4F+TdZB/4nWHCtTb1e8vb9n3dIVux3Z2WRFQc9Ert+RRpn4gAACAASURBVFZbG6iSaMX0wWls3n+I8iMexiXKCmDTQphy81fL+8MjoO8kG5fwB+6m0bkjNtU+GHpK8NoYqyQGzWq/MMod6R4Er73JbHKSMhAiYnpm8NpZI5E80Df3m/k9MM12t95ZjIGFd9lCtwU3+7YQcdmTtuXMtDs7f48hp1tXsNPl1N3pr06CtFZClUQrpg9JA+DznR5aE8uetFkmJ93e9vmcKTYf3R8VvCWbrRnuSdaOk57UNryswO70O+NqgmMZTu3FJZwPBG92jWHhPTfDqauFdK1JGQDjr4bVf7XWQGdY/X92GtzYK+DgFnj/l76RraLQFqdOuAYSszp/n7BwGH8VbH/PZhd1d/qrk6QcDVyHAuOyk4mNCvfM5XS4xO5Oxl8NCb3bXpM9xaa++SMOULLJtrf2ZhfetwcFr3d+bF87qyScjf7aa89RXmCtgsT2iv3bIHNUz2z0V7kLohKsVegrZn7P/o52xpoo3wnv3G9rLy5+FqZ+E1b8sevNLetr4J/X2f87d438vGHCtdZiWvfP7k9/dZLc31oS/sx29AOqJFoRGR5G3sBUz4LXK/5kO2rOuNv9Gn8Gr0vyvffl96S24QUf2uKozrZG8KTRX9kOSBnUcYVuazJH2nhHT1CmrjjTX33ZFDl1kMOa+D/v8vibm+D1O+xO/aJn7M949i+sFfb6HZ1vAWMMLPy2jW9c9rx30/fckT7MtqH/4kX7O5GQZeME3UlStm0P01Nb47hBlUQbTBucytYDhzl4uJ0xlnWHYOVzkHu+bRHhjvhMW8Tm67jE4VKoOeh50NpJ34n2NdBxifoa2P6B7dbZlQde+vAO3E2dbOLm/Ln2tAynikLfPDRbM+t7tq3Mp497fs1nz9gCt3N+dyx7JyrWWhRHSmDR9zsny/KnYMMCOOOndm61r5hwjXWHbX27+11N4DJXwscuJ0/6l3UBVRJNjfDK12HHxy2Hpg+2cYkVO9qpl1j9VztQ/uR7Ov6MnKm+r7x2Zt54a0nEptrCnkDHJXZ8ZHsvjeygUVtHONNg2/pDaW52tIPuhKXSMqVuU9fk8yXG+KaQri1SB9teR6tesL20OqIkHz74lW20N/7q48/1mwSn3mcf9OsXeCdHwUfw3s8gd651g/mSMZdY99XRiu5Nf3Xir1qJ9/4H/nOX35SFKomq3faB+be5NjOjeh9j+yURHx3B8h0H276msd7uogacAtl5HX9G9hQboPVl+ltJJ5UEQN/xgbckNi+yozcHntK1+7Q0+mvjZ1tdBE11nbMkknIgMq5nxSUOH7DpvP5QEgCzvm/dp8ueaH9dUwO89k2IjocLHm/bEjzlu/b3ftG9nqfXVhTCgptsa5uL/uhblxrY37fcC+z3AbEk/FB1vXuFndIXEe29S9VDVEmkDoY7V9idT/6b8FQeESueYdqABPfB6w0LrL/6FA+sCPBPXKJkk80bj3cTMG+PvhNt1k9tle/k8YamRpuzPnxO5/sPOWlp9NeGy6kli6UTu8awMBuX8NaS8GdQ0pc1Em2RNsRmKK18vu3OqU6WPGJjWuc/ZvtAtUV4hHU7NTXA67d3vMutr4H519ng8lUvWgXkD5xtPXqP9s/92yM21dZN+cqSaDgK/7nTbmhm/8I392wDVRIAkb3g9B/DHcttG+V37+d3Zd8m7eAqSqpbzQJuboZPn7CFckNne3b/3mOsmbvHl0oi3/rNO7PbCnTwes9ncLS8664maL/RX3kXUx0zcm2asafsXAK/7Q9/vxg2v+X7jp++Tn9ti1k/sNaXO2uieA0sedi6pkbNbXuNk7Qh8LX/tVlsn89zv85ZZ3FgA1z6gn/rFwbNgm996vnfri8RsbEbX1VdL/6Nzeyb+7hfe1CpknAlbQhc+ypc+SIJcpRXon9F3avfOH5Xte1dGw84+TueP6Ajouzu3VeWhDGdy2xy4gxeByousXmR7acz5Myu3ys+01a7t9Xor2wHRPSymSydITPXBmCPeJDpVrwGXr7a7hZLNsP8q+HxCbYx3RE3bktvcSoJb+pivCV9qG3FvfJ5mxzhSsNReO1b1no95yHP7jf5Rhj2NXj/5+4V7rInbVX1mf8Dw7rh4d1njO9dWZ6SNtSOju2qNVG82v7cJn0dhpzhG9ncoEqiNSKQez7hd33OPC4mq+gteHIyrHj22HzdpBwbBPOG7DwbB2hsJ2PKU6r3Ql31seCqtziD14GISxhj3XpDTveNS0HEPtjacjeVFzjmQ3Ty19zZE6uj9hylW+HFy+zP9aa34Z51cMXfbBbSB7+ER3Ph37dZS7Ir7qiKQjsGMzKm8/fwhFk/sAph+ZPHH//w1zY76MKnoFeyZ/cSgblP2nTT12776szpgg+tAhl1oW/qIXo6Z/7cuuBeutJmSHaGxjp4/U67+Tn7176Vrw1USbghPCaezwfdyc29HrfZGm//EJ45yab8Tb/Te1969lQbFNy/vuvCtQStvUx/daXv+MBYEvvXW3N75Pm+u2eam26wZV1s4ubJlLrKPfD3i+wQoOtft0PvwyPtQ+/GN+GOFXY3vfkteH42zDvVdl6tr3F/T7ef5YPur56QMdx2T/38uWNWVOFSGyDNuwWGemkBJvS2Ae59a+Hj3x07XlFok0UyRsKFzwRud9+dZI6Ey//PJkQsuKVzLsklD9uNywWP22C8n1El0Q7TBqexpDyF/XPn2//Y+hqIy/C+pz0cC177ol7C255NbZE1we60uzt4vXmRrZIecY7v7pk+zGaPuTb6a2q0D6GuZLEkZFlXljslceSgjT/UHYbr/922Lz1zJJz7MHwvH877vd1JL7zLWhfLnvJOHn+lv7bFqT+EhhprTdQdssHnlIFw9q86d7/cC2DCdXZ+xe4Vtpvt/Gv9H6juiQw9E859yLY/f+d+767dt9a6MMdf49saknaI6JZPCVKmOeolPttZzkUTL7bZOPU1navUTMyybipfxCVK8iG+T9daM7i2De9sW4zOsHkR5EyDuHTf3dO10Z/z31W1B5obuhYEFXEMIGrDl15bDf+4xKY1X/8a9Bnb/r2iE2DKrXYnvutT+OT38O791kodMKNjWRpqrZuxu5RExggYfTF8/mebVVW5B27+b9eqlOf8BgqXWLdTn3FwYCNcu6DzFffBzJRb4eB228Ikfah93xGN9dbNFJcOX3vQ/zI68KslISJzRGSLiGwXkfvaOH+viGwSkXUi8oGIDHA5119E3hWRfMeagf6UtS1GZSWS1CvyWCpsZC87lrSzZOf5RkmU5ns+Q8IdWQEIXlcU2ul9vshqcqWtNNiuZjY5ycy1lptrLKHhqA1SH9hoYw8DvBi/LmJrQ678h20p8sZ3PItTVe0BjH+qrd1x6g/tjn/jv+3Exf7Tuna/mESbFluxC/IXwpk/655AdU/law/aoP5bP7QNEjti6R/s38/5f/Bt764O8EhJiMh3RCRRLM+LyBoRObuDa8KBp4FzgFHA1SLS2on+BZBnjBmHHVnqmjLxN+BhY0wuMBVoJ3HbP4SFCScN8rCPkydkT7V/7J3ttgk2BbfEi2l07ohLs5ZNdwavNy+yryPP9e1922r0V+ajdtAZubZC15nh1tRo/ei7PoWL/gTD2/0zcE9UHJz3B5uV5cnQn+5If21NZi5MvsH+3p7upVvEHQNm2MyoGXcfG6F6ohIWbntTZebCqze1H/s6sNHGIsZc5vtNVgd4akncbIypBs4GUoDrgd92cM1UYLsxZocxph6YD1zousAY85ExxhnB+wzIBnAokwhjzHuOdYdd1nUr0wansbu8huJKH7T69kVRXWWhrTDuSjzCSVY3B683L7L1Jb52L0REOxr9uaTBlhfYNuqdKTZ0xXUAUXOzjSdsecvGGcZd3v61HTFstv2j/+T3NkOqPQKhJMAWzN3y7vETF7vKSbfZ2MaJEKjuiOgEuHq+raN66Yqvph2D3Zi8fofNKPM09diHeKoknP+b5wJ/N8ZsdDnmjn6AazJwkeOYO24BHKOjGA5Uisi/ReQLEXnYYZkcL5TIbSKySkRWlZa28cP1Ac75Ep95O62uLbLGQXhU15SELzKbnPR1E7yurYIDm2DruzZf/v1f2hTOxb9r+z6ecOQg7F5uGyL6g/Th1sfrpKzAdjft6oPIqSRK8m0MYe3Ldlc99Rtdu6+TOb+xbsw3v9t+emxFoX2QdFXpeYuIPsz9TXKOVRSHS2D+NTb+5Mqyx63Ff+4jXXN3dxJPA9erReRdYBDwYxFJAHzWTUpErgPygFNd5JoJTAR2A/8EbgSed73OGDMPmAeQl5fnl34II3onkBIbyfIdZVw6ObtrN4uItrt3XyiJjBFdkwWOxSVev8Om51YV2a+66uPXhUVATDLU/BMGnty5fktb/2szWfxlKqcNs/Oym5ttXUR5gQ2OdpW4DIhNsy6hwwfscKlZP+j6fZ3EZ8JZv4I37rYDdiZd3/Y6f7QIV3oO2ZNtvObVG2yrjUufs//XpVtg8W9tw8PRFwVENE+VxC3ABGCHMaZGRFKBmzq4phhwLQ3Ndhw7DhGZDdwPnGqMcUbwioAvjTE7HGteB6bRSkl0B2FhwrTBad7PvXZH9lRY9bwtqOlM36KSfFsI54sy/OzJdr707uV2KE/KIBg407YOcP2K722VyJOT4d2fwq0fel+gtnmRjYH44sHdFq6N/hKybHB09MVdv6+IjUvsWgrjrrJtJnz9oJ54Paydb3+2w+e03Q+pYpdtOa+ELqMvgrL/gQ9/ZTP2Zv3AKoyoeJs+HSA8VRLTsQ/tI45d/ySgo8bzK4FhIjIIqxyuAq5xXSAiE4FngTnGmJJW1yaLSIYxphQ4A/Bxr23PmT4kjbc37GdPeQ05qbFdu1l2Hnz2tC0q6zfJ++u70o6jNb1S4AfbPXvohfWy2SivfdM2OBx3heefU3/EVtZOvtF/O2HXDKemBjBNvuv0Oel6W2B2zkP+6bQZFgYXPAZ/OgXe+bHdRbribBHuSaqsEtzM/J51lS7+Dez9wnodLnnOWpwBwtPf+D8CNSIyHvgeUIDNPnKLMaYRuAt4B8gHXjHGbBSRB0TE2RnsYSAeeFVEvhSRhY5rm4DvAx+IyHps/OPP3v3TfIezXsInWU45U+1rZ+ZLNDXY4KyvlAR499Aee4V1l33wgHczu7d/YFtc+zMrw3Xeta8H3Y+/yqYddrVjbXtkjLBtKda/+tV0yKMVUH+o+4PWSvcjYjcM/WdYF+2Ic2HsZQEVyVMl0WiMMdjspKeMMU8DHfo7jDFvGWOGG2OGGGMedBz7mTHGqQxmG2N6G2MmOL7mulz7njFmnDFmrDHmRkeGVEAYlhlPenyUb4LXif2sO6SoE5XX5TtsgZgvlYQ3hIXZXjFVe+zoVk/ZvMhaLf39uBN2Nvor2xa4QfddZea9Vtm9ee/xbTsqdtpXVRInBhHRtgp95vfdz+voRjxVEodE5MfY1NdFIhIG+HFb1bMQEU4anMbyHWWYrs4LELGpsJ1pz+GLdhxdZdAsGH6O591Nmxrsjmj4OXbGgL9oafS31QatoxN9W9XdHURE211k5S742CXDPFDpr0rgiE21XXED6GZy4qmSuBKow9ZL7McGoR/2m1Q9kOmD09hXVcuuMh+Ua2RPsQ+C9ga7tEVJvi0ac/rfA8VZv7Rxho89SIndtQxqK7unAChtmE2DLXN0fw3GTKCBp9hA9rKnjjWDdCqJ5P4BE0s5cfFISTgUw4tAkoicD9QaY9qNSYQaLX2cfBqX8DIVtmSTffhF9uq6DF0hY4QNQq96oe3uq65sftPOdPBzz3vgWKO//ev9O7jG35z1gN1JvvEd2yW0YpdNxT2RmuApPQZP23JcAXwOXA5cAawQkcBGU7qZIRlxZCRE+yZ4nTXe1h54rSQ2B9bV5MppP7YP//d+7n6NMTYeMeQMiOpiVpgnOBv9HSkJ7qZxsanwtd/YwTIrn+ve7q+K0gpP3U33A1OMMTcYY76ObbnxP/4Tq+chIswYksaSraUcrmvs2s0ie9l6AW/GmdaUW197Rg9REvEZMPO7sGWRnTXQFvu+tLPAu6vXjKsbLtiC1q0Ze5lVrh88YMd6qpJQAoSnSiKsVR1DmRfXhgw3zBhIRU0Dz3zUxjxlb8meAnvX2L4sHVGxC174mrU+hs/p+mf7iml32Gytd3/a9qB75+yI7pLZ2egPgtvdBDaect6j1t1UU6aFdErA8PRB/18ReUdEbhSRG4FFwFv+E6tnMql/CpdM7Mdzn+xkd1cD2DlT7VCXko3tr9u3Fp4/y7aEuP51WyXdU4h0FNjt/cLOKG7N5kUw4OTu6zfjbPQHwW9JgO09ddp9x75XlADgaeD6B9geSeMcX/OMMT/yp2A9lR+dM5KIcOHXizZ17UbZefa1vbjE9vfhL+fapoA3v2v7JvU0xl5hXWcf/PL4AruyAhto7+a2xqQPtyMdu7Hfvl+ZfpfNlc+d2/FaRfEDHruMjDH/Msbc6/h6zZ9C9WR6J8Zw5+lDeXfTAZZu86BOwB3JAyAu031c4ot/wItX2H5Kt7zX9SFD/iIszA5PaV1g55wdMcLHsyM64uTv+Ke/UqAIj7CZZDGJgZZEOUFpV0mIyCERqW7j65CIVLd3bShzyymD6J8aywNvbqSxqZPNcJ1Fda0tCWNsS+7/3GkL1256y44+7cm0VWC3eZEd6dmdk9TAUWdwXfd+pqKEMO0qCWNMgjEmsY2vBGPMCbu1iYkM5/7zctl64DAvrtjd+RvlTLEZS0ccabVNDbDw27D4f+2g82tfDZ4dpGuB3eES2LMCRvppdoSiKN3GCZeh5CvOHtWbU4am8+h7W6k40sm2Us5JdcWroO6wnZv8xd9h1g/homf821DO17gW2C17AjCqJBQlBFAl0UlEhJ9dMIrDdY08+l4Hoyfd0XciSLh1zfzfebad9gWPwxn3B6dP3Vlgt+xJG3PpPTrQEimK0kVUSXSB4b0TuH7aAF5csYvN+zsRoomKsw/SNX+1jemuftnuxoOV+Aw45R77/cjzg1PRKYpyHKokusg9s4eR2CuSXy7c1LkOsSPOhfg+cOObMPxrvhewu5l+J0z5Bky9NdCSKIriA1RJdJHk2Ci+d9Zwlu8o452N+72/wak/gnvzoV8PKpLrCpG94LxHgrt3kqIoLaiS8AFXT+3PyD4J/HpRPrUNTd5dHBbmn5GYiqIoPsCvTycRmSMiW0Rku4jc18b5e0Vkk4isE5EPRGRAq/OJIlIkIk/5U86uEhEexs8uGEVRxVGe+2RHoMVRFEXxGX5TEiISDjwNnAOMAq4WkVGtln0B5BljxgELgIdanf8VsMRfMvqSGUPSOWdMH57+qIB9VV7Mf1YURenB+NOSmApsN8bscMynno+dkd2CMeYjY4yzU95n2Il3AIjIZKA38K4fZfQpPzk3lyZj+N3bmwMtiqIoik/wp5LoB+xxeV/kOOaOW4C3ARwztH8PfL+9DxCR20RklYisKi0t7aK4XScnNZZvzhrM61/uZfWu8kCLoyiK0mV6RMRURK4D8jg2N/sO4C1jTFF71xlj5hlj8owxeRkZGf4W0yNuP20IfRJj+OUbm2hu7kRKrKIoSg/Cn0qiGMhxeZ/tOHYcIjIbO/lurjGmznF4OnCXiBQCjwBfF5Hf+lFWnxEbFcGPzx3JuqIqHnl3S+dqJxRFUXoIEX6890pgmIgMwiqHq4BrXBeIyETgWWCO6+Q7Y8y1LmtuxAa3v5Id1VOZO74vn24/yDOLC6ioaeBXF44mIrxHGG2Koihe4TclYYxpFJG7gHeAcOAFY8xGEXkAWGWMWYh1L8UDr4pt4bDbGBP001VEhN9dOo7MhBie+mg7Bw/X8eTVE4mJDA+0aIqiKF4hoeIOycvLM6tWrQq0GF/hr8sK+cUbG5nUP4Xnb8gjOTYq0CIpiqK0ICKrjTF57s6rD8TP3DBjIE9dPYn1RVVc/qfl7K3UGgpFUYIHVRLdwHnjsvjrzVPZX1XLJc8sY+uBQ4EWSVEUxSNUSXQT04ek8c9vTqfZGC774zJWFmodhaIoPR9VEt3IqL6J/Ov2GaQnRHPdcys61zVWURSlG1El0c3kpMay4FszyM1K5PZ/rObFFbsCLZKiKIpbVEkEgNS4KF76xkmcOjyD+1/bwKPvbtHqbEVReiSqJAJEbFQE876exxV52Tzx4XZu/utKKo7UB1osRVGU41AlEUAiw8P43aXj+NVFY1i2vYzzn1zKl3sqAy2WoihKC6okAoyIcP20ASy4fToAl/9pGX9dVqg9nxRF6RGokughjMtOZtHdpzBzWAY/X7iRb7/8BYfrGgMtlqIoJziqJHoQybFRPPf1PH44ZwRvrd/H3KeWauGdoigBRZVEDyMsTLjjtKG8eOs0qo82cuFTn/LvNe2O1VAURfEbqiR6KNOHpPHW3acwNjuJe19Zy09eW09tQ1OgxVIU5QRDlUQPJjMxhpduPYlvnTqEl1bs5rI/LWNdkWY/KYrSfaiS6OFEhIdx3zkj+fPX8yiqOMrcpz7luudW8On2g5oBpSiK39F5EkHEodoGXlyxm+eX7qT0UB3js5O4/bQhnD2qD2FhEmjxFEUJQgI6T0JE5ojIFhHZLiJfGT8qIveKyCYRWSciH4jIAMfxCSKyXEQ2Os5d6U85g4WEmEi+deoQPvnh6fzvxWOpPNrAt/6xhtl/+JhXVu2hvrE50CIqihJi+M2SEJFwYCtwFlCEnXl9tTFmk8ua04EVxpgaEbkdOM0Yc6WIDAeMMWabiPQFVgO5xhi3DvkTwZJoTVOz4a31+/jj4gI27aumT2IMt84cxNVT+xMX7c/x5YqihAqBtCSmAtuNMTuMMfXAfOBC1wXGmI+MMTWOt58B2Y7jW40x2xzf7wVKgAw/yhqUhIcJF4zvy6K7T+GvN09lYHosv16Uz8m/+5CnPtym2VCKonQZfyqJfsAel/dFjmPuuAV4u/VBEZkKRAEFPpUuhBARTh2ewfzbpvPvO2aQNyCVR97dyrmPf8KygoOBFk9RlCCmR2Q3ich1QB7wcKvjWcDfgZuMMV9xuIvIbSKySkRWlZaWdo+wPZxJ/VN47oY8/nbzVBqbDdf8eQXff3Ut5dphVlGUTuBPJVEM5Li8z3YcOw4RmQ3cD8w1xtS5HE8EFgH3G2M+a+sDjDHzjDF5xpi8jAz1Rrkya3gG7353FnecNoTXvyjmzN8v5l+rizRtVlEUr/CnklgJDBORQSISBVwFLHRdICITgWexCqLE5XgU8BrwN2PMAj/KGNLERIbzwzkjWXT3TAZnxPO9V9dy7XMr2FF6ONCiKYoSJPhNSRhjGoG7gHeAfOAVY8xGEXlAROY6lj0MxAOvisiXIuJUIlcAs4AbHce/FJEJ/pI11BnRJ4FXvzmdBy8ew/riKuY8/glPfLCNukYNbCuK0j5aTHeCUXKolgfe2MSb6/YxNDOe/714LFMHpQZaLEVRAkRAi+mUnkdmQgxPXTOJv9w0haP1TVzx7HKumrecV1bt4VBtQ6DFUxSlh6GWxAlMTX0jf/m0kFdX7aGwrIaYyDC+NroPl07K5uSh6YRrqw9FCXk6siRUSSgYY1izu5J/rynijbV7qa5tJDMhmosn9uOSSdmM6JMQaBEVRfETqiQUr6hrbOLD/BL+taaYxVtKaGw2jO6byCWTsjl3bB+yknoFWkRFUXyIKgml05QdruONtXv59xfFrCuqAiA9Ppox/RIZ3TeRMX2TGNMvieyUXoioa0pRghFVEopP2HbgEEu3H2RDcTUb91axreQwTc32dycxJoLRfZOs4uiXxJh+iQxKj9eYhqIEAR0pCW0VqnjEsN4JDOt9LDZR29DE1gOH2FBczYa9VWzcW83fP9tFnaNdeVxUOGOzkxifk8yE7GTG5ySTlRSjFoeiBBmqJJROERMZzrjsZMZlJ7cca2xqpqD0COuLq1hXVMnaoir+srSQ+iarONLjo5mQk8R4h9IYl51EcmxUoP4JiqJ4gCoJxWdEhIcxok8CI/okcNnkbMAGwjfvO8Taokq+3FPJ2j2VvJ/f0oGFEb0TOHdsFuePz2JIRnygRFcUxQ0ak1C6neraBtYXVfHlnko+3lLKyl3lGAO5WYmcPy6LC8b1pX9abKDFVJQTAg1cKz2e/VW1LFq/jzfX7eWL3Xb44LjsJM4fl8V54/rSL1nTbhXFX6iSUIKKoooaFq3bx5vr9rG+2KbdTuqfzPnj+nL26N5kp6iFoSi+RJWEErQUHjzisDD2kb+vGoCRfRI4MzeTM3N7MyE7mTBNs1WULqFKQgkJdpQe5oP8Et7PP8CqXRU0NRvS46M4fUQmZ+ZmMnNYBnHRmoehKN6iSkIJOapqGli8tYQP8ktYvKWE6tpGosLDmDYkjdm5mZw0KI0+STEkxkRoXYaidIAqCSWkaWhqZlVhBR9uPsAH+SXsOHik5VxMZBi9E2PonRBD76QYeidE0zsxhsxE+5qVFEP/1FhVJMoJjSoJ5YRiR+lhNuytpqS6lgPVtRyornO81rK/upbahubj1vdJjOHM3Exmj+rN9MFpxESGB0hyRQkMAW3LISJzgMeBcOA5Y8xvW52/F7gVaARKgZuNMbsc524AfupY+mtjzF/9KasSGgzOiGewm6I8YwyH6hodCqSO3eU1fLyllNe+KObFFbuJjQpn5rB0Zuf25oyRmaTFR3ez9IrS8/CbJSEi4cBW4CygCFgJXG2M2eSy5nRghTGmRkRuB04zxlwpIqnAKiAPMMBqYLIxpsLd56kloXSW2oYmlu8o44P8A7y/qYT91bWIwKT+KczO7c3s3EyGZsarW0oJSQLmbhKR6cAvjDFfc7z/MYAx5jdu1k8EnjLGnCwiV2MVxjcd554FFhtjXnb3eaokFF9gjGHj3mrezz/A+/kH2FBsU2/7JMYwqm8iuVkJ5GYlkpuVyMC0OO10S6SiWAAADflJREFUqwQ9gXQ39QP2uLwvAk5qZ/0twNvtXNuv9QUichtwG0D//v27IquiACAijnbnSdwzezj7qo7yfn4JqwrLyd9XzcdbS1tapPeKDGd4nwRGuSiOkX0SSIiJDPC/QlF8R49ILBeR67CupVO9uc4YMw+YB9aS8INoyglOVlIvrp82gOunDQCsa2p7yWE27atm875D5O+r5u0N+3n582N7msyEaAamxTEgLZaB6Y5Xx3tVIEqw4U8lUQzkuLzPdhw7DhGZDdwPnGqMqXO59rRW1y72i5SK4gUxkeEtloYTYwz7q2vJ31dN/r5D7Co7QmFZDR9vLeXV1UXHXZ8WF9WiNIb3SWD64DTG9EtSt5XSY/FnTCICG7g+E/vQXwlcY4zZ6LJmIrAAmGOM2eZyPBUbrJ7kOLQGG7gud/d5GpNQeiI19Y3sKqtpURy7yo5QeLCGwrIj7KuqBSAhJoKTBqUxfUgaM4akMaJ3grYbUbqNgMUkjDGNInIX8A42BfYFY8xGEXkAWGWMWQg8DMQDrzoyR3YbY+YaY8pF5FdYxQLwQHsKQlF6KrFRES3xitaUHKrlsx3lLC84yLKCMt7PPwBAalwU0wenMc2hNAanx2lmlRIwtJhOUXoIxZVHWV5QxrKCgywvKGuxNDITohnRJ4FB6XEMTItjUEYcg9LiyE7pRUR4WIClVoIdrbhWlCDEGENhWQ3LC8r4fGcZBaVHKDx4hEN1jS1rIsKEnNTYY8oj3QbK+6fG0je5F5GqQBQPCGjFtaIonUNEGJQex6D0OK45yaZ3G2MoO1LPzoNH2HnQKo3CsiPsPGiVydGGppbrw8RmZvVPjbVfabHkOL7PSelFalyUurAUj1AloShBgoiQHh9Nenw0UwamHnfOGMOB6jp2lR1hd3kNe8pr2O34+mBzCQcP1x23PiE6gulD0pg9yrYgSdcWJIobVEkoSgggIvRJiqFPUgwnDU77yvma+kaKKo6yu8wqjm0lh/l4SwnvbjpwXAuSs0ZlMiRDW5Aox9CYhKKcoLhrQTIwLdb2rBrVm7wBKRocD3E0cK0oikc4W5C8v+kAywvKqG9qJqlXJFMGptA3uRdZSb3omxxDn8QY+ib3ondiDFERqkCCHVUSiqJ4zeG6Rj7ZWsp7+QfYWFzNvqqjVNc2HrdGBNLjo+mbFENWUi96J0ZjgLqGZmobm6hraKausYlax2tdYzO1DfY1NS6KWcMyOG1EBhNyktVaCSCqJBRF8QmH6xrZX3WUvZW17HN53VdVy76qWkqqawkPE6IjwomODCPG8RodEUZMZDjREWH2XEQYu8trWLO7gmYDiTERzByWwanDMzh1RAa9E2MC/U89odAUWEVRfEJ8dARDMxMYmpngk/tV1TSwdPtBPt5awuItpSxavw+AkX0SOG1EJqeNyGDygBSt9wgwakkoihJwjDFs3n+IxVtK+XhrCasKK2hsNsRGhbcUB/ZNjmmJi2Ql9aKfxkV8gloSiqL0eESkpcfV7acN4VBtA8sKylheUMae8hr2VtWyZncFlTUNra47FhcZkhHPpAEp5A1MYXimNkn0FWpJKIoSNNTUNx6LhVTWUlx5tCU+snn/oZaiwYSYCCb1TyFvQAp5A1OZkJNMr6jwAEvfM1FLQlGUkCE2KoKhmfEMzYz/yjljDLvLa1hZWMHqXeWsKqzg91tLAdvnanTfRCYPSGV8jp3fcbS+iaMNTRytb6KmvonaBvvqPHa0oYm46AhGZSUyqm8io7ISyUg48SrT1ZJQFCVkqaypZ83uClYVVrBqVwVr91RS19jc5troiDBio8LpFRlOryj7VXGkgeLKoy1rMhKiGe1QGE7FMTAtLqhdW2pJKIpywpIcG8UZI3tzxsjeANQ3NlNQepjwMDmmDCLtl7sHfVVNA5v2VbNxbxWb9lWzaW81S7cdpNEx6zw2KpzcrEQm5iSTNzCFyQNSQ8riUEtCURTFS+oam9h24HCL0thQXMW64irqHVbKgLRYJg9IIW9AKnkDUxiaEd9jrQ21JBRFUXxMdMRXZ53XNTaxobi6JR7y8ZZS/r2mGLAFg5MG2ED6sN4JxEdHEBcdQXx0OHGO7+OiInrkrHO/WhIiMgd4HDu+9DljzG9bnZ8FPAaMA64yxixwOfcQcB4QBrwHfMe0I6xaEoqi9CScg6NW7zoWSN9Wcrjda2Kjwh3KI4LEmAiGZMYfF/9Ijo3yuZwBsyREJBx4GjgLKAJWishCY8wml2W7gRuB77e6dgZwMlZ5ACwFTgUW+0teRVEUX+I6OOqyydmADaQXVRzlcF0jR+oaHa9NLt83cqS+kcN1TVTW1LN028EWawSgb1JMi8IY1dfWleSkxPrVleVPd9NUYLsxZgeAiMwHLgRalIQxptBxrnW6gQFigChAgEjggB9lVRRF8TvJsVFeWwOlh+rI31dN/r7qlhjIh5tLcMTNiY+O4LQRGTx1zSQ/SOxfJdEP2OPyvgg4yZMLjTHLReQjYB9WSTxljMlvvU5EbgNuA+jfv3+XBVYURelpZCREk5GQwazhGS3Hahua2LL/EJscyiM+2n+P8h4ZuBaRoUAukO049J6IzDTGfOK6zhgzD5gHNibRvVIqiqIEhpjIcMbnJDM+J9nvn+XPzljFQI7L+2zHMU+4GPjMGHPYGHMYeBuY7mP5FEVRlA7wp5JYCQwTkUEiEgVcBSz08NrdwKkiEiEikdig9VfcTYqiKIp/8ZuSMMY0AncB72Af8K8YYzaKyAMiMhdARKaISBFwOfCsiGx0XL4AKADWA2uBtcaYN/wlq6IoitI2WnGtKIpyAtNRnYRO61AURVHcokpCURRFcYsqCUVRFMUtqiQURVEUt4RM4FpESoFd/9/evYVYVcVxHP/+GrvYGKlkImqaFnTD7IJQWkhRVC8amGUp1ks9GCi9WFFkQhDR7SXUIkHJUkst6SkTsXzIa1OadhERUswhzGqCLui/h72mpsk9M8cZ3bP3+X1gmH3W2bNZf/5z9v/stc9ZqxuHuAD4sYe60xtULR6oXkxViweqF1PV4oH/xzQiIgbl7VyZItFdkrZ1dIe/bKoWD1QvpqrFA9WLqWrxQO0xebjJzMxyuUiYmVkuF4l/vV50B3pY1eKB6sVUtXigejFVLR6oMSbfkzAzs1y+kjAzs1wuEmZmlqvui4SkOyR9I2mvpMeL7k9PkLRf0k5JTZJKN+uhpMWSmiXtatM2UNI6Sd+l3wOK7GOtcmKaJ+lgylOTpLuK7GMtJA2XtEHSbklfSZqd2kuZpw7iKXOOzpG0RdIXKaZnU/vFkjanc96KtJRD/nHq+Z6EpAbgW+A2suVVtwLTImJ3h3/Yy0naD1wfEaX8EpCkm4EWYGlEXJXaXgCORMTzqZgPiIi5RfazFjkxzQNaIuLFIvt2MiQNAYZExA5J5wHbgcnAg5QwTx3EM5Xy5khAY0S0pHV5NgGzgceA1RGxXNJCsqUYFuQdp96vJMYBeyNiX0T8CSwHJhXcp7oXEZ8AR9o1TwKWpO0lZC/g0siJqbQi4lBE7Ejbv5KtGTOUkuapg3hKKzIt6eGZ6SeAW8jW7IEu5Kjei8RQ4Ps2jw9Q8n+MJICPJG2X9HDRnekhgyPiUNr+ARhcZGd60KOSvkzDUaUYmmlP0kjgGmAzFchTu3igxDmS1CCpCWgG1pEt5nY0LQoHXTjn1XuRqKoJEXEtcCcwKw11VEZkY6RVGCddAIwGxgKHgJeK7U7tJPUDVgFzIuKXts+VMU8niKfUOYqIYxExFhhGNnJyWa3HqPcicRAY3ubxsNRWahFxMP1uBtaQ/XOU3eE0btw6ftxccH+6LSIOpxfxceANSpanNM69ClgWEatTc2nzdKJ4yp6jVhFxFNgA3AD0l9QnPdXpOa/ei8RW4NJ0t/8s4D5gbcF96hZJjenGG5IagduBXR3/VSmsBWam7ZnABwX2pUe0nkyTuylRntJN0TeBPRHxcpunSpmnvHhKnqNBkvqn7b5kH9DZQ1YspqTdOs1RXX+6CSB9pO1VoAFYHBHPFdylbpE0iuzqAaAP8HbZYpL0DjCRbErjw8AzwPvASuAisinhp0ZEaW4E58Q0kWwYI4D9wCNtxvN7NUkTgE+BncDx1Pwk2Th+6fLUQTzTKG+OxpDdmG4guyBYGRHz0zliOTAQ+ByYHhF/5B6n3ouEmZnlq/fhJjMz64CLhJmZ5XKRMDOzXC4SZmaWy0XCzMxyuUiY9QKSJkr6sOh+mLXnImFmZrlcJMxqIGl6mqO/SdKiNIFai6RX0pz96yUNSvuOlfRZmhxuTevkcJIukfRxmud/h6TR6fD9JL0n6WtJy9K3gM0K5SJh1kWSLgfuBcanSdOOAQ8AjcC2iLgS2Ej2bWqApcDciBhD9k3e1vZlwGsRcTVwI9nEcZDNPDoHuAIYBYw/5UGZdaJP57uYWXIrcB2wNb3J70s2gd1xYEXa5y1gtaTzgf4RsTG1LwHeTfNqDY2INQAR8TtAOt6WiDiQHjcBI8kWijErjIuEWdcJWBIRT/ynUXq63X4nO9dN2/lzjuHXp/UCHm4y67r1wBRJF8I/6zmPIHsdtc6qeT+wKSJ+Bn6SdFNqnwFsTKueHZA0OR3jbEnnntYozGrgdypmXRQRuyU9Rbbq3xnAX8As4DdgXHqumey+BWTTMC9MRWAf8FBqnwEskjQ/HeOe0xiGWU08C6xZN0lqiYh+RffD7FTwcJOZmeXylYSZmeXylYSZmeVykTAzs1wuEmZmlstFwszMcrlImJlZrr8B05PJ5IZIU88AAAAASUVORK5CYII=\n"
          },
          "metadata": {
            "needs_background": "light"
          }
        }
      ]
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "collapsed_sections": [],
      "machine_shape": "hm",
      "name": "DDSM.ipynb",
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}