Pretrained Network for image Classification.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### In this file, we will apply two techniques of proceeding to use a pretrained models\n",
    "1. Feature extraction without using data augmentation\n",
    "    A. Running the Conv base over new dataset, store its output to numpy array, \n",
    "    B. Feed this Numpy array data as input to standalone, densely connected classifier\n",
    "2. Feature extraction using data augmentation\n",
    "    A. Extending the model by adding Dense layer on top of it and run the whole model on the input data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "import keras\n",
    "from keras.applications import VGG16 #VGG16 is a Conv neural network architecture developed by OXford geometry lab"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "# the arguments are available at keras applications website\n",
    "wt_imagenet_file = \"C:\\\\Users\\Ramy\\\\.keras\\\\datasets\\\\VGG16\\\\vgg16_weights_tf_dim_ordering_tf_kernels_notop.h5\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "conv_base = VGG16(weights=wt_imagenet_file, include_top = False,input_shape=(150,150,3))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"vgg16\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_1 (InputLayer)         [(None, 150, 150, 3)]     0         \n",
      "_________________________________________________________________\n",
      "block1_conv1 (Conv2D)        (None, 150, 150, 64)      1792      \n",
      "_________________________________________________________________\n",
      "block1_conv2 (Conv2D)        (None, 150, 150, 64)      36928     \n",
      "_________________________________________________________________\n",
      "block1_pool (MaxPooling2D)   (None, 75, 75, 64)        0         \n",
      "_________________________________________________________________\n",
      "block2_conv1 (Conv2D)        (None, 75, 75, 128)       73856     \n",
      "_________________________________________________________________\n",
      "block2_conv2 (Conv2D)        (None, 75, 75, 128)       147584    \n",
      "_________________________________________________________________\n",
      "block2_pool (MaxPooling2D)   (None, 37, 37, 128)       0         \n",
      "_________________________________________________________________\n",
      "block3_conv1 (Conv2D)        (None, 37, 37, 256)       295168    \n",
      "_________________________________________________________________\n",
      "block3_conv2 (Conv2D)        (None, 37, 37, 256)       590080    \n",
      "_________________________________________________________________\n",
      "block3_conv3 (Conv2D)        (None, 37, 37, 256)       590080    \n",
      "_________________________________________________________________\n",
      "block3_pool (MaxPooling2D)   (None, 18, 18, 256)       0         \n",
      "_________________________________________________________________\n",
      "block4_conv1 (Conv2D)        (None, 18, 18, 512)       1180160   \n",
      "_________________________________________________________________\n",
      "block4_conv2 (Conv2D)        (None, 18, 18, 512)       2359808   \n",
      "_________________________________________________________________\n",
      "block4_conv3 (Conv2D)        (None, 18, 18, 512)       2359808   \n",
      "_________________________________________________________________\n",
      "block4_pool (MaxPooling2D)   (None, 9, 9, 512)         0         \n",
      "_________________________________________________________________\n",
      "block5_conv1 (Conv2D)        (None, 9, 9, 512)         2359808   \n",
      "_________________________________________________________________\n",
      "block5_conv2 (Conv2D)        (None, 9, 9, 512)         2359808   \n",
      "_________________________________________________________________\n",
      "block5_conv3 (Conv2D)        (None, 9, 9, 512)         2359808   \n",
      "_________________________________________________________________\n",
      "block5_pool (MaxPooling2D)   (None, 4, 4, 512)         0         \n",
      "=================================================================\n",
      "Total params: 14,714,688\n",
      "Trainable params: 14,714,688\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "conv_base.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### feature extraction"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os, shutil"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "original_dataset_dir = \"C:\\\\Users\\Ramy\\.keras\\datasets\\PetImages\"\n",
    "#few images\n",
    "base_directory = \"C:\\\\Users\\Ramy\\.keras\\datasets\\cats_dogs_small\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "os.mkdir(base_directory)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_directory = os.path.join(base_directory, 'train')\n",
    "validation_directory = os.path.join(base_directory, 'validation')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_cats_dir = os.path.join(train_directory, 'cats')\n",
    "train_dogs_dir = os.path.join(train_directory, 'dogs')\n",
    "validation_cats_dir = os.path.join(validation_directory, 'cats')\n",
    "validation_dogs_dir = os.path.join(validation_directory, 'dogs')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "original_dir_cat = os.path.join(original_dataset_dir, 'Cat')\n",
    "original_dir_dog = os.path.join(original_dataset_dir, 'Dog')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "TrainImages_Count = 200\n",
    "ValImages_End = 300"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "fnames = ['{}.jpg'.format(i) for i in range(TrainImages_Count)]\n",
    "for fnames in fnames:\n",
    "    src = os.path.join(original_dir_cat, fnames)\n",
    "    dst = os.path.join(train_cats_dir, fnames)\n",
    "    shutil.copyfile(src, dst)\n",
    "\n",
    "fnames = ['{}.jpg'.format(i) for i in range(TrainImages_Count, ValImages_End)]\n",
    "for fnames in fnames:\n",
    "    src = os.path.join(original_dir_cat, fnames)\n",
    "    dst = os.path.join(validation_cats_dir, fnames)\n",
    "    shutil.copyfile(src, dst)\n",
    "\n",
    "    \n",
    "fnames = ['{}.jpg'.format(i) for i in range(TrainImages_Count)]\n",
    "for fnames in fnames:\n",
    "    src = os.path.join(original_dir_dog, fnames)\n",
    "    dst = os.path.join(train_dogs_dir, fnames)\n",
    "    shutil.copyfile(src, dst)\n",
    "\n",
    "fnames = ['{}.jpg'.format(i) for i in range(TrainImages_Count, ValImages_End)]\n",
    "for fnames in fnames:\n",
    "    src = os.path.join(original_dir_dog, fnames)\n",
    "    dst = os.path.join(validation_dogs_dir, fnames)\n",
    "    shutil.copyfile(src, dst)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "200"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(os.listdir(train_cats_dir))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "100"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(os.listdir(validation_cats_dir))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature extraction using numpy( not DA)\n",
    "note that: the final feature map has (4,4,512) from the info of the summary of VGG16"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "from keras.preprocessing.image import ImageDataGenerator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "datagen = ImageDataGenerator(rescale=1./255)\n",
    "batch_size = 20"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "def extract_feature(directory, sample_count):\n",
    "    features = np.zeros(shape =(sample_count, 4, 4, 512))\n",
    "    labels = np.zeros(shape= (sample_count))\n",
    "    generator = datagen.flow_from_directory(directory, target_size=(150,150), batch_size=batch_size, class_mode='binary')\n",
    "    i = 0\n",
    "    for input_batch, labels_batch in generator:\n",
    "        # features extracting by calling the predicting method of the model conv_base\n",
    "        features_batch = conv_base.predict(input_batch)\n",
    "        features[i * batch_size : (i + 1) * batch_size] = features_batch\n",
    "        labels[i * batch_size : (i + 1) * batch_size] = labels_batch\n",
    "        i = i + 1\n",
    "        print(\"i: \", i)\n",
    "        if i * batch_size >= sample_count:\n",
    "            print(\"loob broken \")\n",
    "            break\n",
    "    return features, labels"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 400 images belonging to 2 classes.\n",
      "i:  1\n",
      "i:  2\n",
      "i:  3\n",
      "i:  4\n",
      "i:  5\n",
      "i:  6\n",
      "i:  7\n",
      "i:  8\n",
      "i:  9\n",
      "i:  10\n",
      "loob broken \n"
     ]
    }
   ],
   "source": [
    "train_img_count = 200\n",
    "trian_features, train_labels = extract_feature(train_directory, train_img_count) # 2000"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 200 images belonging to 2 classes.\n",
      "i:  1\n",
      "i:  2\n",
      "i:  3\n",
      "i:  4\n",
      "i:  5\n",
      "loob broken \n"
     ]
    }
   ],
   "source": [
    "val_img_count = 100\n",
    "validation_features, validation_labels = extract_feature(validation_directory, val_img_count) #1000"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Flatten\n",
    "The extracted features are now of shape(4,4,512)\n",
    "Before feeding them to a desne classifier, first we flatten them to (samples, 8192)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "trian_features = np.reshape(trian_features, (train_img_count, 4*4*512))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "validation_features = np.reshape(validation_features, (val_img_count, 4*4*512))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### defining and training densely connected classifiers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras import models, layers, optimizers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = models.Sequential()\n",
    "model.add(layers.Dense(256, activation= 'relu', input_dim = 4*4*512))\n",
    "model.add(layers.Dropout(0.5))\n",
    "model.add(layers.Dense(1, activation = 'sigmoid'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "dense (Dense)                (None, 256)               2097408   \n",
      "_________________________________________________________________\n",
      "dropout (Dropout)            (None, 256)               0         \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 1)                 257       \n",
      "=================================================================\n",
      "Total params: 2,097,665\n",
      "Trainable params: 2,097,665\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(optimizer=optimizers.RMSprop(lr=2e-5),\n",
    "             loss='binary_crossentropy',\n",
    "             metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "10/10 [==============================] - 1s 42ms/step - loss: 0.7189 - accuracy: 0.5661 - val_loss: 0.6687 - val_accuracy: 0.5700\n",
      "Epoch 2/10\n",
      "10/10 [==============================] - 0s 28ms/step - loss: 0.7050 - accuracy: 0.5726 - val_loss: 0.6320 - val_accuracy: 0.6100\n",
      "Epoch 3/10\n",
      "10/10 [==============================] - 0s 25ms/step - loss: 0.6220 - accuracy: 0.6488 - val_loss: 0.5952 - val_accuracy: 0.6900\n",
      "Epoch 4/10\n",
      "10/10 [==============================] - 0s 29ms/step - loss: 0.5652 - accuracy: 0.7027 - val_loss: 0.5690 - val_accuracy: 0.7300\n",
      "Epoch 5/10\n",
      "10/10 [==============================] - 0s 26ms/step - loss: 0.6227 - accuracy: 0.6133 - val_loss: 0.5508 - val_accuracy: 0.7400\n",
      "Epoch 6/10\n",
      "10/10 [==============================] - 0s 26ms/step - loss: 0.4885 - accuracy: 0.7676 - val_loss: 0.5259 - val_accuracy: 0.7900\n",
      "Epoch 7/10\n",
      "10/10 [==============================] - 0s 28ms/step - loss: 0.4429 - accuracy: 0.8112 - val_loss: 0.5021 - val_accuracy: 0.8200\n",
      "Epoch 8/10\n",
      "10/10 [==============================] - 0s 33ms/step - loss: 0.4397 - accuracy: 0.8227 - val_loss: 0.5245 - val_accuracy: 0.7300\n",
      "Epoch 9/10\n",
      "10/10 [==============================] - 0s 31ms/step - loss: 0.4553 - accuracy: 0.8001 - val_loss: 0.4762 - val_accuracy: 0.8400\n",
      "Epoch 10/10\n",
      "10/10 [==============================] - 0s 29ms/step - loss: 0.3893 - accuracy: 0.8790 - val_loss: 0.4568 - val_accuracy: 0.8300\n"
     ]
    }
   ],
   "source": [
    "Model_fitting = model.fit(trian_features, train_labels, epochs=10, batch_size = 20,\n",
    "                          validation_data=(validation_features, validation_labels)) #the iteration will be fast because of np"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "acc = Model_fitting.history['accuracy']\n",
    "val_acc = Model_fitting.history['val_accuracy']\n",
    "loss = Model_fitting.history['loss']\n",
    "val_loss = Model_fitting.history['val_loss']\n",
    "epochs = range(1, len(acc) + 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x2c3d6ace670>"
      ]
     },
     "execution_count": 49,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(epochs, acc, 'bo', label= 'Training Accuracy')\n",
    "plt.plot(epochs, val_acc, 'b', label = 'Validation accuracy')\n",
    "plt.title('Acc Iteration')\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x2c3d6b1e880>"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXoAAAEICAYAAABRSj9aAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4yLjIsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+WH4yJAAAgAElEQVR4nO3deXxU1f3/8dcHwhZARMENhICKiOwGZFHEXYplE7+CVEQsiLvYUlGqoJa2Vn7WWqUaUXGhInWrCy7FiohLZRERBBUxQAoKYmURlcXP748zwBADmcAkdzLzfj4eeczcM/fe+cwon9yce87nmLsjIiLpq0LUAYiISOlSohcRSXNK9CIiaU6JXkQkzSnRi4ikOSV6EZE0p0QvKcPM3MyOjD2/18xuTGTfvXifAWb26t7GmcrS+bPJ3lOil6Qxs1fM7JYi2nua2RdmlpXoudx9mLvfmoSYcmK/FHa8t7tPcvcz9vXcRbxXVzMriNuebma/TPb7xJ2/zD6blG9K9JJME4ELzMwKtV8ATHL3rWUfUvllZhWjjkHSgxK9JNOzwAHAidsbzKw2cDbwiJm1N7N3zOwbM1tlZnebWeWiTmRmE83sd3HbI2LHrDSzwYX27W5m75vZejNbYWZj4l6eEXv8xsw2mllHMxtkZjPjju9kZrPMbF3ssVPca9PN7FYze8vMNpjZq2ZWp7gvwszGxr6Hu2Pve3esvamZ/cvMvjazj83s/wp95r+Z2VQz+xY4ORU/m5RD7q4f/STtB7gfmBC3fQkwL/b8OKADkAXkAIuAa+L2deDI2POJwO9iz88CvgSaA9WBvxfatyvQgnDh0jK2b6/YazmxfbPi3mcQMDP2/ADgf4S/OrKA/rHtA2OvTwc+A5oA1WLbf9zNZ+8KFMRtTwd+GbddHVgBXBR7r7bAV8CxcZ95HdA59lmqpspn00/5/tEVvSTbw8C5ZlYttj0w1oa7z3H3d919q7vnA/cBJyVwzv8DHnL3Be7+LTAm/kV3n+7uH7r7j+4+H3g8wfMCdAc+dfdHY3E9DiwGfh63z0Pu/om7fwdMAVoneO7Czgby3f2h2HvNBZ4C+sbt8093fyv2Wb4vR59NUpgSvSSVu88E1gA9zawx0I5wBY6ZNTGzF2I3ZtcDvwcS6So4jHAlvN2y+BfN7Hgze93M1pjZOmBYgufdfu5lhdqWAfXitr+Ie74JqJHguQtrCBwf67r6xsy+AQYAh8TtE/85y9NnkxSmRC+l4RHClfwFwKvu/mWs/W+EK8qj3H0/4Aag8I3boqwCDo/bblDo9b8DzwGHu3st4N648xZXnnUlIQHHawD8N4G4ilP4vVcAb7j7/nE/Ndz90j0ck6qfTcoRJXopDY8ApwFDiHXbxNQE1gMbzawpcGkRxxZlCjDIzJqZWTYwutDrNYGv3f17M2sPnB/32hrgR6Dxbs49FWhiZuebWZaZnQc0A15IMLY9+bLQ+74Qe68LzKxS7KedmR2zh3Ok6meTckSJXpIu1v/+NuHm43NxL/2akKg2EG7aPpHg+V4C7gT+DSyJPca7DLjFzDYANxF+MWw/dhMwFngr1l3SodC51xL6zn8FrAV+A5zt7l8lElsx/gL0NbP/mdld7r4BOAPoR7ja/gK4Daiyh3Ok6meTcsTctfCIiEg60xW9iEiaU6IXEUlzSvQiImlOiV5EJM0lXE2wLNWpU8dzcnKiDkNEpNyYM2fOV+5et6jXUjLR5+TkMHv27KjDEBEpN8ys8CzoHdR1IyKS5pToRUTSnBK9iEiaS8k+ehEpW1u2bKGgoIDvv/8+6lCkGFWrVqV+/fpUqlQp4WOU6EWEgoICatasSU5ODj9dCVJShbuzdu1aCgoKaNSoUcLHpU3XzaRJkJMDFSqEx0mToo5IpPz4/vvvOfDAA5XkU5yZceCBB5b4L6+0uKKfNAmGDoVNm8L2smVhG2DAgOjiEilPlOTLh73575QWV/SjRu1M8ttt2hTaRUQyXVok+uXLS9YuIqll7dq1tG7dmtatW3PIIYdQr169HdubN2/e47GzZ8/mqquuKvY9OnXqlJRYp0+fztlnn52Uc5WVtEj0DQovLFdMu4jsm2TfEzvwwAOZN28e8+bNY9iwYQwfPnzHduXKldm6detuj83NzeWuu+4q9j3efvvtfQuyHEuLRD92LGRn79qWnR3aRSS5tt8TW7YM3HfeE0v2AIhBgwZx7bXXcvLJJ3Pdddfx3nvv0alTJ9q0aUOnTp34+OOPgV2vsMeMGcPgwYPp2rUrjRs33uUXQI0aNXbs37VrV/r27UvTpk0ZMGAA2xdgmjp1Kk2bNuWEE07gqquuKvbK/euvv6ZXr160bNmSDh06MH/+fADeeOONHX+RtGnThg0bNrBq1Sq6dOlC69atad68OW+++WZyv7A9SIubsdtvuI4aFbprGjQISV43YkWSb0/3xJL9b+6TTz5h2rRpVKxYkfXr1zNjxgyysrKYNm0aN9xwA0899dRPjlm8eDGvv/46GzZs4Oijj+bSSy/9yZjz999/n4ULF3LYYYfRuXNn3nrrLXJzc7nkkkuYMWMGjRo1on///sXGN3r0aNq0acOzzz7Lv//9bwYOHMi8efMYN24c99xzD507d2bjxo1UrVqVvLw8zjzzTEaNGsW2bdvYVPhLLEVpkegh/A+mxC5S+srynti5555LxYoVAVi3bh0XXnghn376KWbGli1bijyme/fuVKlShSpVqnDQQQfx5ZdfUr9+/V32ad++/Y621q1bk5+fT40aNWjcuPGO8en9+/cnLy9vj/HNnDlzxy+bU045hbVr17Ju3To6d+7Mtddey4ABA+jTpw/169enXbt2DB48mC1bttCrVy9at269T99NSaRF142IlJ2yvCdWvXr1Hc9vvPFGTj75ZBYsWMDzzz+/27HkVarsXGu9YsWKRfbvF7XP3qyfXdQxZsbIkSOZMGEC3333HR06dGDx4sV06dKFGTNmUK9ePS644AIeeeSREr/f3lKiF5ESieqe2Lp166hXrx4AEydOTPr5mzZtytKlS8nPzwfgiSeeKPaYLl26MCl2c2L69OnUqVOH/fbbj88++4wWLVpw3XXXkZuby+LFi1m2bBkHHXQQQ4YM4eKLL2bu3LlJ/wy7o0QvIiUyYADk5UHDhmAWHvPySr/r9De/+Q3XX389nTt3Ztu2bUk/f7Vq1Rg/fjxnnXUWJ5xwAgcffDC1atXa4zFjxoxh9uzZtGzZkpEjR/Lwww8DcOedd9K8eXNatWpFtWrV6NatG9OnT99xc/app57i6quvTvpn2B3bmz9XSltubq5r4RGRsrNo0SKOOeaYqMOI3MaNG6lRowbuzuWXX85RRx3F8OHDow7rJ4r672Vmc9w9t6j9dUUvIhJz//3307p1a4499ljWrVvHJZdcEnVISZE2o25ERPbV8OHDU/IKfl/pil5EJM0p0YuIpLmEEr2ZnWVmH5vZEjMbWcTrI8xsXuxngZltM7MDEjlWRERKV7GJ3swqAvcA3YBmQH8zaxa/j7vf7u6t3b01cD3whrt/ncixIiJSuhK5om8PLHH3pe6+GZgM9NzD/v2Bx/fyWBHJQF27duWVV17Zpe3OO+/ksssu2+Mx24dh/+xnP+Obb775yT5jxoxh3Lhxe3zvZ599lo8++mjH9k033cS0adNKEn6RUqmccSKJvh6wIm67INb2E2aWDZwFbK80VJJjh5rZbDObvWbNmgTCEpF00b9/fyZPnrxL2+TJkxMqLAah6uT++++/V+9dONHfcsstnHbaaXt1rlSVSKIvat2q3c2y+jnwlrt/XdJj3T3P3XPdPbdu3boJhCUi6aJv37688MIL/PDDDwDk5+ezcuVKTjjhBC699FJyc3M59thjGT16dJHH5+Tk8NVXXwEwduxYjj76aE477bQdpYwhjJFv164drVq14pxzzmHTpk28/fbbPPfcc4wYMYLWrVvz2WefMWjQIJ588kkAXnvtNdq0aUOLFi0YPHjwjvhycnIYPXo0bdu2pUWLFixevHiPny/qcsaJjKMvAA6P264PrNzNvv3Y2W1T0mNFJAVccw3Mm5fcc7ZuDXfeufvXDzzwQNq3b8/LL79Mz549mTx5Mueddx5mxtixYznggAPYtm0bp556KvPnz6dly5ZFnmfOnDlMnjyZ999/n61bt9K2bVuOO+44APr06cOQIUMA+O1vf8sDDzzAlVdeSY8ePTj77LPp27fvLuf6/vvvGTRoEK+99hpNmjRh4MCB/O1vf+Oaa64BoE6dOsydO5fx48czbtw4JkyYsNvPF3U540Su6GcBR5lZIzOrTEjmzxXeycxqAScB/yzpsSIi8d038d02U6ZMoW3btrRp04aFCxfu0s1S2Jtvvknv3r3Jzs5mv/32o0ePHjteW7BgASeeeCItWrRg0qRJLFy4cI/xfPzxxzRq1IgmTZoAcOGFFzJjxowdr/fp0weA4447bkchtN2ZOXMmF1xwAVB0OeO77rqLb775hqysLNq1a8dDDz3EmDFj+PDDD6lZs+Yez52IYq/o3X2rmV0BvAJUBB5094VmNiz2+r2xXXsDr7r7t8Udu89Ri0ip2dOVd2nq1asX1157LXPnzuW7776jbdu2fP7554wbN45Zs2ZRu3ZtBg0atNvyxNuZFdVjHFasevbZZ2nVqhUTJ05k+vTpezxPcXXAtpc63l0p5OLOtb2ccffu3Zk6dSodOnRg2rRpO8oZv/jii1xwwQWMGDGCgQMH7vH8xUloHL27T3X3Ju5+hLuPjbXdG5fkcfeJ7t4vkWNFRAqrUaMGXbt2ZfDgwTuu5tevX0/16tWpVasWX375JS+99NIez9GlSxeeeeYZvvvuOzZs2MDzzz+/47UNGzZw6KGHsmXLlh2lhQFq1qzJhg0bfnKupk2bkp+fz5IlSwB49NFHOemkk/bqs0Vdzli1bkQkZfTv358+ffrs6MJp1aoVbdq04dhjj6Vx48Z07tx5j8e3bduW8847j9atW9OwYUNOPPHEHa/deuutHH/88TRs2JAWLVrsSO79+vVjyJAh3HXXXTtuwgJUrVqVhx56iHPPPZetW7fSrl07hg0btlefa8yYMVx00UW0bNmS7OzsXcoZv/7661SsWJFmzZrRrVs3Jk+ezO23306lSpWoUaNGUhYoUZliEVGZ4nJGZYpFRGQXSvQiImlOiV5EgOJHmUhq2Jv/Tkr0IkLVqlVZu3atkn2Kc3fWrl1L1apVS3ScRt2ICPXr16egoADVmUp9VatWpX79+iU6RoleRKhUqRKNGjWKOgwpJeq6SbJJkyAnBypUCI9x8zJERCKhK/okmjQJhg6F7TWIli0L2wADBkQXl4hkNl3RJ9GoUTuT/HabNoV2EZGoKNEn0fLlJWsXESkLaZXoP/0Ufvwxuvdv0KBk7SIiZSFtEv3XX0OnTnDKKaFvPApjx0J29q5t2dmhXUQkKmmT6GvXhttvh7lzoWVLeOQRKOu5HwMGQF4eNGwIZuExL083YkUkWmlXvTI/HwYOhDffhD594L77oE6d5MYnIpJqMqp6ZU4OvP46/OlP8MIL0Lw5TJ0adVQiItFJu0QPULEijBgBs2bBQQdB9+4wbBhs3Bh1ZCIiZS8tE/12LVuGZD9iROgrb9MG3n036qhERMpWWid6gCpVQjfO9OmwZQt07gw33hiei4hkgrRP9Nt16QLz54cbtb/7HXToAIsWRR2ViEjpy5hED7DffvDQQ/D002G2atu2cNdd0U6yEhEpbQklejM7y8w+NrMlZjZyN/t0NbN5ZrbQzN6Ia883sw9jr6XEit+9e8OHH8Jpp8HVV8MZZ0BBQdRRiYiUjmITvZlVBO4BugHNgP5m1qzQPvsD44Ee7n4scG6h05zs7q13N8YzCoccAs89F27SvvsutGgBjz8edVQiIsmXyBV9e2CJuy91983AZKBnoX3OB5529+UA7r46uWGWDjMYMgQ++ACOOQbOPx/69QvlFERE0kUiib4esCJuuyDWFq8JUNvMppvZHDMbGPeaA6/G2ofuW7il44gjYMaMUJPmqafC1f2rr0YdlYhIciSS6K2ItsJ1E7KA44DuwJnAjWbWJPZaZ3dvS+j6udzMuhT5JmZDzWy2mc2OYt3KrCy44Qb4z3+gVi0480y48sqf1pcXESlvEkn0BcDhcdv1gZVF7POyu3/r7l8BM4BWAO6+Mva4GniG0BX0E+6e5+657p5bt27dkn2KJGrbFubMgWuugbvvDtt7WXZHRCQlJJLoZwFHmVkjM6sM9AOeK7TPP4ETzSzLzLKB44FFZlbdzGoCmFl14AxgQfLCLx3VqsGf/wzTpsG330LHjnDLLbB1a9SRiYiUXLGJ3t23AlcArwCLgCnuvtDMhpnZsNg+i4CXgfnAe8AEd18AHAzMNLMPYu0vuvvLpfNRku/UU8MwzPPOg9Gj4YQT4JNPoo5KRKRk0q5McWmZMiUURvvhBxg3Ljy3ou5eiIhEIKPKFJeW//u/cHV/wglw2WWhIuaqVVFHJSJSPCX6EqhXD15+OdyknT491Lp/8smooxIR2TMl+hIyg8svh/ffD+Pvzz03FEpbty7qyEREiqZEv5eOPhreegvGjIG//z1Msnr99aijEhH5KSX6fVCpUhiN8/bbYUjmKafAtdfC999HHZmIyE5K9EnQvn3oyrn88jD+Pjc3bIuIpAIl+iTJzg43aV9+ORRFO/54+OMfYdu2qCMTkUynRJ9kZ54ZhmH26gXXXw8nnQRLl0YdlYhkMiX6UnDggfDEE/DYY7BgAbRqBQ8+CCk4N01EMoASfSkxgwEDwjq17drBxReHla1Wl4tK/SKSTpToS1mDBqE42v/7f/DSS2EY5vPPRx2ViGQSJfoyUKFCGHY5Zw4ceij06AFDh8LGjVFHJiKZQIm+DDVvHhY2ue46mDAh9N2//XbUUYlIulOiL2NVqoRhl2+8AT/+CCeeCL/9LWzeHHVkIpKulOgjcuKJYVHyCy8Ma9V27AiLFkUdlYikIyX6CO23Xxh2+fTTsHx5WLbwrrvClb6ISLIo0aeA3r3DJKtTT4Wrrw6TrgoKoo5KRNKFEn2KOOSQMOzyvvvCDdoWLWDy5KijEpF0oESfQszCsMsPPoCmTaF/fzj/fPjf/6KOTETKMyX6FHTkkfDmm3DrrfCPf4Sr+2nToo5KRMorJfoUlZUVhl2+8w7UqAGnnw7XXAPffRd1ZCJS3ijRp7jcXJg7F668Ev7yFzjuuLAtIpIoJfpyIDs7DLt85ZWwNu3xx8Pvfw9btxa9/6RJkJMTSi/k5IRtEclcCSV6MzvLzD42syVmNnI3+3Q1s3lmttDM3ijJsZKYM84IwzB794ZRo0Kt+88+23WfSZPCDd1ly0JZ5GXLwraSvUjmKjbRm1lF4B6gG9AM6G9mzQrtsz8wHujh7scC5yZ6rJTMAQfsrHW/cGGolzNhws5a96NGwaZNux6zaVNoF5HMlMgVfXtgibsvdffNwGSgZ6F9zgeedvflAO6+ugTHSgnF17pv3x6GDIGePeHLL8MM26Lsrl1E0l8iib4esCJuuyDWFq8JUNvMppvZHDMbWIJjZS9tr3V/xx3w6qthGGadOrvfV0QyUyKJ3opoK7woXhZwHNAdOBO40cyaJHhseBOzoWY228xmr1mzJoGwBMIN1+HDYfZsOOwwWLMGKlbcdZ/s7FA4TUQyUyKJvgA4PG67PrCyiH1edvdv3f0rYAbQKsFjAXD3PHfPdffcunXrJhq/xDRvDu+9ByNHhqJoWVmhvWFDyMsLXT0ikpkSSfSzgKPMrJGZVQb6Ac8V2uefwIlmlmVm2cDxwKIEj5UkqVwZ/vAHmDED6tcPffnt24cbtplMw00l0xWb6N19K3AF8AoheU9x94VmNszMhsX2WQS8DMwH3gMmuPuC3R1bOh9FtjvhhFAv5/rr4eWXQ999377h5m2m0XBTETD3IrvMI5Wbm+uzZ8+OOoy08PXX8Oc/hwlX69eHMfg33QStW0cdWdnIyQnJvbCGDSE/v6yjESk9ZjbH3XOLek0zY9PcAQeE4mj5+SHB//vf0KZNGI6ZCaUUNNxURIk+Y9SuDTffHBL+zTeHfvzjjoOf/zyM2ElXuxtWquGmkkmU6DPM/vuHK/v8/HCl/9Zb0K4ddO8eRu2km7Fjw/DSeBpuKplGiT5D1aoVyiDn54ek9+67oVjaWWeF0sjpYsCAMLy0YcMwCknDTSUT6WasALBhA4wfD+PGwVdfhfr3o0dD585RRyYiidDNWClWzZpw3XXw+efwpz/BvHlhmOZpp4XVrkSk/FKil13UqAEjRoSEP25cKIvcpQucfDJMnx51dCKyN5TopUjVq8OvfhUS/p//DIsXh2R/0klhiGYK9viJyG4o0cseZWeHtWqXLg1LGS5ZAqeeGq7yp01TwhcpD5ToJSHVqsFVV4UVre6+O1zpn3566Md/9dWiE75qzIikBiV6KZGqVeHyy0PCHz8eVqyAM8+ETp3gpZd2JnzVmBFJHUr0sleqVIFLL4VPP4V774WVK+FnP4MOHeDFF+GGG7SkoUiqUKKXfVKlClxySUj4998Pq1fD2WerxoxIKlGil6SoXBl++Uv45BN44IGdC58UphozImVPiV6SqlIlGDwYJkwIyT9exYrQrx9s3RpNbCKZSoleSsWFF8KDD+68gq9ePST+224La9tedlmooPnjj9HGKZIJlOil1AwYsHPUzcaNYRGUZ56BU06BiRPD5KsGDeDaa0PlTI3JFykdSvRSZqpWhV69YPLkcNP273+H3Fy4555QOfOII8JonfnzlfRFkkmJXiJRowb07w/PPgtffgkPPQRNmoSCaq1awbHHwi23hJu7IrJvlOglcvvvD4MGhYXMV62Cv/0N6taFMWPg6KOhbdvwC6CotV9FpHhK9JJS6taFYcPgjTfCmPs77ggjea67LpRR6NQpLHS+alXUkYqUH0r0krLq14fhw+E//wklF/7whzC79uqrw2unnBJWi1q7NupIRVKbEr2UC40bw8iRYUGUjz4KyyD+979hVu4hh4TyC488AuvXRx2pSOpJKNGb2Vlm9rGZLTGzkUW83tXM1pnZvNjPTXGv5ZvZh7F2rQ8o++yYY+Dmm0ON/Llzw/DMhQvD2P2DDoI+fWDKlJ/W2hHJVMWuGWtmFYFPgNOBAmAW0N/dP4rbpyvwa3c/u4jj84Fcd/8q0aC0ZqyUlHtY4Hzy5JDkv/giTNLq2TPMxj3jjFCXRyRd7euase2BJe6+1N03A5OBnskMUGRfmUHHjmFxlIKCsArWgAFhJE+PHnDwwaHa5pIlUUcqUvYSSfT1gBVx2wWxtsI6mtkHZvaSmR0b1+7Aq2Y2x8yG7kOsIgmpWDEse3jffeHKfurUkOwffDAM1+zXD95/P+ooRcpOIoneimgr3N8zF2jo7q2AvwLPxr3W2d3bAt2Ay82sS5FvYjbUzGab2ew1a9YkEJZI8SpVgm7dwo3a/Pyw8PnUqWFsfrduYRinZuFKuksk0RcAh8dt1wdWxu/g7uvdfWPs+VSgkpnViW2vjD2uBp4hdAX9hLvnuXuuu+fWrVu3xB9EpDiHHgp//GMYn//734cbuV27QufO8NxzKrAm6SuRRD8LOMrMGplZZaAf8Fz8DmZ2iJlZ7Hn72HnXmll1M6sZa68OnAEsSOYHECmp/feH668PV/j33BMmX/XsCS1bwqOPwpYtUUcoklzFJnp33wpcAbwCLAKmuPtCMxtmZsNiu/UFFpjZB8BdQD8Pw3kOBmbG2t8DXnT3l0vjg4iUVLVqoVzyp5/CY4+FG7oDB8JRR4UF0DU8U9JFscMro6DhlRIF97De7R/+AG+/DXXqhFm4l18OtWtHHZ3Inu3r8EqRjGAW1rt96y14881QOvnGG0PN/BEjwgLoIuWREr1IEU44AV54AT74IAzNvOMOaNQIhg4NXT0i5YkSvcgetGwJkyaF5H7xxWGYZtOmcN55Gosv5YcSvUgCGjeG8eNDTfzf/CbMuG3bFs46C6ZP11h8SW1K9CIlcPDB4Wbt8uVhTP68eWEWbseO8M9/aiy+pCYlepG9UKtWWAzl88/DilirV4f1cFu0CN07GosvqUSJXmQfVKsWVsT65JOw2HlWViiXfOSRYSUsjcWXVKBEL5IEWVlhsfN588JY/IYNwxj8hg3h1lvhf/+LOkLJZEr0IklkFla7mjEDZs6EDh3gppvCWPxf/zqsiiVS1pToRUpJ587w/PMwf36opXPnnWH0zsUXhxWxRMqKEr1IKWvRItTS+fRTGDIEHn8cmjcPV/6vvZaZQzOXLg1/4RxzTPhlKKVLiV6kjDRqFIqlrVgR+u3nzoXTTgvj8R97LP1H6vz4I7z6Kvz85+Fm9V/+At9+C717w8SJUUeX3pToRcrYgQfCb38byiRPmACbN8MFF4Rundtvh3Xroo4wudavh7/+NVy9n3kmvPfezs//0Udwyilw0UXhs0vpUKIXiUjVqqG//sMPw6pXTZqEWbf168O114ZZuOXZokVwxRVQrx5cdRUccED4y2X5crjlltBeo0aoKXTeeeGzjxihSWelQYleJGIVKoRlDV97LXTn9OwZroCPOCKsb1ueKnZv2xZmCJ9+OjRrBvffD+ecA7NmwTvvhAXbq1TZ9ZjKlcMchCuugHHjwtV9undjlTUlepEU0qZNuOpduhSGD4eXXoJ27eCkk8JNy1S92l27Fv70p/DLqVcvWLw4LNdYUBD633OLrJK+U4UKYYLZrbeGmcW9e2uyWTIp0YukoMMPD33WK1aEEsn5+aFccrNmcN998N13UUcYzJsHv/xl6G667rpww/nJJ0NpiOuvh5Is/2wW+u7vvTf8gjv9dPj669KLPZMo0YuksP32C1f2n30WhmXWqBFKLjRoAGPGhBo7ZW3LFnjiCTjxxPAXyOOPh7IP8+fD66+HrpqsrL0//yWXwJQpocuqSxdNMksGJXqRciArK/TXz5oVyiJ37Ag33xwS/iWXwMcfl34MX3wRbqI2bBhiWbUq/LVRUBCuwlu0SN57nXNOKAW9fDl06lQ2ny+dKdGLlCNmob/+uefCqJYLL4SHHw6LofToAX8aHSIAAA0sSURBVG+8kdwJWO7w7rvhJmqDBjB6NLRqFUbKfPJJ+GujtNbTPfnk8Hm+/z6s+DVrVum8TyZQohcpp5o2Df31y5eHBPzOO9C1a7h5O3kybN269+f+/vvwC6R9+/DXwwsvwGWXhSvrl16C7t3DDdTS1qZNWMO3Zs2Q+P/1r9J/z3SkRC9Szh10UOivX748dKFs2BAqaR5xBPz5z2HCUqJWrIAbbgg3gwcNCiNfxo8P3TN33hnG+pe1I48Myf6II8IvmCeeKPsYyjslepE0Ua1a6K9ftCiMZc/JCROvDj88TEYqKCj6OPfQ73/OOeGY224LXSWvvQYLFsCll4Yr6igdemjoxunYMfwSu/vuaOMpbxJK9GZ2lpl9bGZLzGxkEa93NbN1ZjYv9nNToseKSHJVqLCzv/6998JkrDvuCEMff/GLnYuaf/tt6Ppp2TJ0i0yfHmamLl0KzzwTShOYRfpRdrH//uEGbY8ecOWVobsqEwvC7RV33+MPUBH4DGgMVAY+AJoV2qcr8MLeHFvUz3HHHeci6eaxx9wbNnQ3C4+PPVZ27/355+7XXONeo4Y7uHfs6F6rVnjepo37gw+6b9pUdvHsiy1b3AcPDrFfcon71q1RR5QagNm+m5yayBV9e2CJuy91983AZKBngr9H9uVYkbQxaRIMHRrq17iHx6FDQ3tZyMkJ/fUrVoSumXXrwpX+W2/BnDmh7EC1amUTy77KygrF4EaODH+RnHce/PBD1FGltkQSfT1gRdx2QaytsI5m9oGZvWRmx5bwWJG0NmrUT6f0b9oU2svS/vuH/vqFC8NEp06dUqt7JlFm8Ic/hC6pp54Kv7RKctM50ySS6Iv636Bwz9hcoKG7twL+CjxbgmPDjmZDzWy2mc1es2ZNAmGJlB/Ll5esXRIzfDg8+ii8+WYYWvrll1FHlJoSSfQFwOFx2/WBlfE7uPt6d98Yez4VqGRmdRI5Nu4cee6e6+65dUtSIEOkHGjQoGTtkrhf/CJMIFu8OIwW+vzzqCNKPYkk+lnAUWbWyMwqA/2A5+J3MLNDzMIfgGbWPnbetYkcK5IJxo6F7Oxd27KzQ7vsu+1lnteuDd1R8+dHHVFqKTbRu/tW4ArgFWARMMXdF5rZMDMbFtutL7DAzD4A7gL6xW4EF3lsaXwQkVQ2YADk5YU6MWbhMS8vtEtydOwIM2eGm7VdusCMGVFHlDrMU3Agam5urs8uT6stiEjKWL48LFn4+edhFm3PDBnnZ2Zz3L3Iyv+aGSsiaaVBg3BztlUr6NMHHnww6oiip0QvImmnTp3QZ3/66WFd3ttuy+xZtEr0IpKWatQIo3H69w+Tq37969RdirG07cM6MCIiqa1y5bAGb506YXLV6tWhK6dSpagjK1tK9CKS1ipUgL/8BQ4+OKxJu3Yt/OMfUL161JGVHXXdiEjaMwvlJvLy4JVX4LTTMmvhcSV6EckYQ4aEq/n33w+Lm++uRn+6UaIXkYzSp0+oa79iRZhFu3hx1BGVPiV6Eck4XbuGhVk2bw71cd57L+qISpcSvYhkpO0Lj9eqFbpx+vaF55+HLVuijiz5lOhFJGMdcQS8/TYMGxZq4/ToAYcdBldfHRZkSZdJVkr0IpLRDj44DL/873/DFf3JJ8O990JuLjRvHmbVlvebtkr0IiKESVRnnw1TpsAXX4RlCmvXDrNqGzQI5RQefRQ2bow60pJTohcRKaR27bCm78yZ8OmncNNN8NlnMHAgHHIIXHhhqKWzbVvUkSZGiV5EZA+OPBLGjAmJ/s03Q+2cZ58Nk65ycuD662HRoqij3DMlehGRBJiFoZj33x+6diZPhpYt4fbboVkzaNcO/vpXSMUlr5XoRURKqFo1OO88ePHFcKP2jjtg61a46qowaqdnT3jqKfjhh6gjDZToRUT2wSGHwPDhoazCBx/ANdeECVh9+8Khh8Kll8I770Q7VFOJXkQkSbZ35axYEcosdOsGDz8cSi00aQK33hqWOCxrSvQiIkmWlRXWrZ00KfTnP/gg1K8fRu80bgwnnQQPPADr1pVNPEr0IiKlaL/94KKL4PXXIT8ffve7kPx/+cvQ7dO/P7z0UujjLy1K9CIiZaRhw1AXf/FiePddGDw41Mf/2c/CFf+vfhUKrSWbEr2IRGLSpDAOvUKF8DhpUtQRlR0zOP54uOceWLUKnn4aOnYM9XYqV07++yWU6M3sLDP72MyWmNnIPezXzsy2mVnfuLZ8M/vQzOaZ2exkBC0i5dukSWHm6bJlYTTKsmVhO5OS/XZVqkDv3vDMM2F0TmkoNtGbWUXgHqAb0Azob2bNdrPfbcArRZzmZHdv7e65+xiviKSBUaNg06Zd2zZtCu2ZLKuUVvFO5Iq+PbDE3Ze6+2ZgMtCziP2uBJ4CVicxPhFJQ8uXl6xd9k0iib4esCJuuyDWtoOZ1QN6A/cWcbwDr5rZHDMburs3MbOhZjbbzGavScU5xCJpIhX6xhs0KFm77JtEEr0V0VZ4jtedwHXuXlQtt87u3pbQ9XO5mXUp6k3cPc/dc909t27dugmEJSIllSp942PHQnb2rm3Z2aFdki+RRF8AHB63XR9YWWifXGCymeUDfYHxZtYLwN1Xxh5XA88QuoJEJAKp0jc+YADk5YXhhmbhMS8vtEvymRdTgMHMsoBPgFOB/wKzgPPdfeFu9p8IvODuT5pZdaCCu2+IPf8XcIu7v7yn98zNzfXZszVARyTZKlQouuaKGfz4Y9nHI8ljZnN2N+Cl2Hu87r7VzK4gjKapCDzo7gvNbFjs9aL65bc7GHjGzLa/19+LS/IiUnoaNAjdNUW1S/pKaDCPu08FphZqKzLBu/uguOdLgVb7EJ+IJNHYsaFPPr77Rn3j6U8zY0UyiPrGM1MpDc8XkVQ1YIASe6bRFb2IZKxUmFNQFnRFLyIZafucgu33K7bPKYD0+4tHV/QikpFSZU5BWVCiF5GMlEn1dpToRSQjZVK9HSV6EclImVRvR4leRDJSJs0p0KgbEclYmTKnQFf0IiJpToleRCTNKdGLiKQ5JXoRkTSnRC8iErHSrrmjUTciIhEqi5o7uqIXEYlQWdTcUaIXEYlQWdTcUaIXEYlQWdTcUaIXEYlQWdTcUaIXEYlQWdTc0agbEZGIlXbNHV3Ri4ikuYQSvZmdZWYfm9kSMxu5h/3amdk2M+tb0mNFRKR0FJvozawicA/QDWgG9DezZrvZ7zbglZIeKyIipSeRK/r2wBJ3X+rum4HJQM8i9rsSeApYvRfHiohIKUkk0dcDVsRtF8TadjCzekBv4N6SHht3jqFmNtvMZq9ZsyaBsEREJBGJjLqxItq80PadwHXuvs1sl90TOTY0uucBeQBmtsbMliUQWyqrA3wVdRApQt/FrvR97Erfx0778l003N0LiST6AuDwuO36wMpC++QCk2NJvg7wMzPbmuCxP+HudROIK6WZ2Wx3z406jlSg72JX+j52pe9jp9L6LhJJ9LOAo8ysEfBfoB9wfvwO7t5o+3Mzmwi84O7PmllWcceKiEjpKjbRu/tWM7uCMJqmIvCguy80s2Gx1wv3yxd7bHJCFxGRRCQ0M9bdpwJTC7UVmeDdfVBxx2aIvKgDSCH6Lnal72NX+j52KpXvwtyLvDcqIiJpQiUQRETSnBK9iEiaU6JPIjM73MxeN7NFZrbQzK6OOqaomVlFM3vfzF6IOpaomdn+ZvakmS2O/T/SMeqYomRmw2P/ThaY2eNmVjXqmMqSmT1oZqvNbEFc2wFm9i8z+zT2WDsZ76VEn1xbgV+5+zFAB+By1fbhamBR1EGkiL8AL7t7U6AVGfy9xGbTXwXkuntzwqi8ftFGVeYmAmcVahsJvObuRwGvxbb3mRJ9Ern7KnefG3u+gfAPuciSD5nAzOoD3YEJUccSNTPbD+gCPADg7pvd/Ztoo4pcFlAtNt8mmwQmU6YTd58BfF2ouSfwcOz5w0CvZLyXEn0pMbMcoA3wn2gjidSdwG+AH6MOJAU0BtYAD8W6siaYWfWog4qKu/8XGAcsB1YB69z91WijSgkHu/sqCBeOwEHJOKkSfSkwsxqESp7XuPv6qOOJgpmdDax29zlRx5IisoC2wN/cvQ3wLUn6s7w8ivU99wQaAYcB1c3sF9FGlb6U6JPMzCoRkvwkd3866ngi1BnoYWb5hPLUp5jZY9GGFKkCoMDdt/+F9yQh8Weq04DP3X2Nu28BngY6RRxTKvjSzA4FiD2uLmb/hCjRJ5GFqm4PAIvc/Y6o44mSu1/v7vXdPYdwk+3f7p6xV2zu/gWwwsyOjjWdCnwUYUhRWw50MLPs2L+bU8ngm9NxngMujD2/EPhnMk6qxcGTqzNwAfChmc2Ltd0QKwMhciUwycwqA0uBiyKOJzLu/h8zexKYSxit9j4ZVgrBzB4HugJ1zKwAGA38EZhiZhcTfhmem5T3UgkEEZH0pq4bEZE0p0QvIpLmlOhFRNKcEr2ISJpTohcRSXNK9CIiaU6JXkQkzf1/HnClLteZw+gAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(epochs, loss, 'bo', label= 'Training loss')\n",
    "plt.plot(epochs, val_loss, 'b', label = 'Validation loss')\n",
    "plt.title('Validation Iteration')\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Feature extraction using Data augmentation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [],
   "source": [
    "New_model = models.Sequential()\n",
    "New_model.add(conv_base)\n",
    "New_model.add(layers.Flatten())\n",
    "New_model.add(layers.Dense(256, activation= 'relu', input_dim = 4*4*512))\n",
    "New_model.add(layers.Dense(1, activation = 'sigmoid'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential_1\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "vgg16 (Functional)           (None, 4, 4, 512)         14714688  \n",
      "_________________________________________________________________\n",
      "flatten (Flatten)            (None, 8192)              0         \n",
      "_________________________________________________________________\n",
      "dense_2 (Dense)              (None, 256)               2097408   \n",
      "_________________________________________________________________\n",
      "dense_3 (Dense)              (None, 1)                 257       \n",
      "=================================================================\n",
      "Total params: 16,812,353\n",
      "Trainable params: 16,812,353\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "New_model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Num of trainable weights before freezing the conv_base:  30\n"
     ]
    }
   ],
   "source": [
    "print('Num of trainable weights before freezing the conv_base: ',len(New_model.trainable_weights) )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "conv_base.trainable = False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Num of trainable weights after freezing the conv_base:  4\n"
     ]
    }
   ],
   "source": [
    "print('Num of trainable weights after freezing the conv_base: ',len(New_model.trainable_weights) )"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_data_generator = ImageDataGenerator(rescale=1./255,\n",
    "                                         rotation_range=40,\n",
    "                                         width_shift_range=0.2,\n",
    "                                         height_shift_range=0.2,\n",
    "                                         shear_range=0.2,\n",
    "                                         zoom_range=0.2,\n",
    "                                         horizontal_flip=True,\n",
    "                                         fill_mode='nearest')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_data_generator = ImageDataGenerator(rescale=1./255)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 400 images belonging to 2 classes.\n",
      "Found 200 images belonging to 2 classes.\n"
     ]
    }
   ],
   "source": [
    "train_generator = train_data_generator.flow_from_directory(train_directory,\n",
    "                                                          target_size=(150,150),\n",
    "                                                          batch_size=20,\n",
    "                                                          class_mode='binary')\n",
    "validation_generator = test_data_generator.flow_from_directory(validation_directory,\n",
    "                                                              target_size=(150,150),\n",
    "                                                              batch_size=20,\n",
    "                                                              class_mode='binary')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [],
   "source": [
    "New_model.compile(loss = 'binary_crossentropy',\n",
    "                 optimizer = optimizers.RMSprop(lr=2e-5),\n",
    "                 metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/2\n",
      " 19/200 [=>............................] - ETA: 6:45 - loss: 0.5935 - accuracy: 0.7289"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-64-51d8426fef32>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[1;32m----> 1\u001b[1;33m New_model_fitting = New_model.fit(train_generator, steps_per_epoch=200,\n\u001b[0m\u001b[0;32m      2\u001b[0m                                  \u001b[0mepochs\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;36m2\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      3\u001b[0m                                  \u001b[0mvalidation_data\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mvalidation_generator\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      4\u001b[0m                                  validation_steps=50) # the model will train verrry slow! numpy is much better\n",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\tensorflow\\python\\keras\\engine\\training.py\u001b[0m in \u001b[0;36mfit\u001b[1;34m(self, x, y, batch_size, epochs, verbose, callbacks, validation_split, validation_data, shuffle, class_weight, sample_weight, initial_epoch, steps_per_epoch, validation_steps, validation_batch_size, validation_freq, max_queue_size, workers, use_multiprocessing)\u001b[0m\n\u001b[0;32m   1098\u001b[0m                 _r=1):\n\u001b[0;32m   1099\u001b[0m               \u001b[0mcallbacks\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mon_train_batch_begin\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mstep\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1100\u001b[1;33m               \u001b[0mtmp_logs\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mtrain_function\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0miterator\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m   1101\u001b[0m               \u001b[1;32mif\u001b[0m \u001b[0mdata_handler\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mshould_sync\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1102\u001b[0m                 \u001b[0mcontext\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0masync_wait\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\tensorflow\\python\\eager\\def_function.py\u001b[0m in \u001b[0;36m__call__\u001b[1;34m(self, *args, **kwds)\u001b[0m\n\u001b[0;32m    826\u001b[0m     \u001b[0mtracing_count\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mexperimental_get_tracing_count\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    827\u001b[0m     \u001b[1;32mwith\u001b[0m \u001b[0mtrace\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mTrace\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_name\u001b[0m\u001b[1;33m)\u001b[0m \u001b[1;32mas\u001b[0m \u001b[0mtm\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 828\u001b[1;33m       \u001b[0mresult\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_call\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwds\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    829\u001b[0m       \u001b[0mcompiler\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;34m\"xla\"\u001b[0m \u001b[1;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_experimental_compile\u001b[0m \u001b[1;32melse\u001b[0m \u001b[1;34m\"nonXla\"\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    830\u001b[0m       \u001b[0mnew_tracing_count\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mexperimental_get_tracing_count\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\tensorflow\\python\\eager\\def_function.py\u001b[0m in \u001b[0;36m_call\u001b[1;34m(self, *args, **kwds)\u001b[0m\n\u001b[0;32m    853\u001b[0m       \u001b[1;31m# In this case we have created variables on the first call, so we run the\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    854\u001b[0m       \u001b[1;31m# defunned version which is guaranteed to never create variables.\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 855\u001b[1;33m       \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_stateless_fn\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m*\u001b[0m\u001b[0margs\u001b[0m\u001b[1;33m,\u001b[0m \u001b[1;33m**\u001b[0m\u001b[0mkwds\u001b[0m\u001b[1;33m)\u001b[0m  \u001b[1;31m# pylint: disable=not-callable\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    856\u001b[0m     \u001b[1;32melif\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_stateful_fn\u001b[0m \u001b[1;32mis\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    857\u001b[0m       \u001b[1;31m# Release the lock early so that multiple threads can perform the call\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\tensorflow\\python\\eager\\function.py\u001b[0m in \u001b[0;36m__call__\u001b[1;34m(self, *args, **kwargs)\u001b[0m\n\u001b[0;32m   2940\u001b[0m       (graph_function,\n\u001b[0;32m   2941\u001b[0m        filtered_flat_args) = self._maybe_define_function(args, kwargs)\n\u001b[1;32m-> 2942\u001b[1;33m     return graph_function._call_flat(\n\u001b[0m\u001b[0;32m   2943\u001b[0m         filtered_flat_args, captured_inputs=graph_function.captured_inputs)  # pylint: disable=protected-access\n\u001b[0;32m   2944\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\tensorflow\\python\\eager\\function.py\u001b[0m in \u001b[0;36m_call_flat\u001b[1;34m(self, args, captured_inputs, cancellation_manager)\u001b[0m\n\u001b[0;32m   1916\u001b[0m         and executing_eagerly):\n\u001b[0;32m   1917\u001b[0m       \u001b[1;31m# No tape is watching; skip to running the function.\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1918\u001b[1;33m       return self._build_call_outputs(self._inference_function.call(\n\u001b[0m\u001b[0;32m   1919\u001b[0m           ctx, args, cancellation_manager=cancellation_manager))\n\u001b[0;32m   1920\u001b[0m     forward_backward = self._select_forward_and_backward_functions(\n",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\tensorflow\\python\\eager\\function.py\u001b[0m in \u001b[0;36mcall\u001b[1;34m(self, ctx, args, cancellation_manager)\u001b[0m\n\u001b[0;32m    553\u001b[0m       \u001b[1;32mwith\u001b[0m \u001b[0m_InterpolateFunctionError\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    554\u001b[0m         \u001b[1;32mif\u001b[0m \u001b[0mcancellation_manager\u001b[0m \u001b[1;32mis\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 555\u001b[1;33m           outputs = execute.execute(\n\u001b[0m\u001b[0;32m    556\u001b[0m               \u001b[0mstr\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0msignature\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mname\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    557\u001b[0m               \u001b[0mnum_outputs\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_num_outputs\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\anaconda3\\lib\\site-packages\\tensorflow\\python\\eager\\execute.py\u001b[0m in \u001b[0;36mquick_execute\u001b[1;34m(op_name, num_outputs, inputs, attrs, ctx, name)\u001b[0m\n\u001b[0;32m     57\u001b[0m   \u001b[1;32mtry\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m     58\u001b[0m     \u001b[0mctx\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mensure_initialized\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m---> 59\u001b[1;33m     tensors = pywrap_tfe.TFE_Py_Execute(ctx._handle, device_name, op_name,\n\u001b[0m\u001b[0;32m     60\u001b[0m                                         inputs, attrs, num_outputs)\n\u001b[0;32m     61\u001b[0m   \u001b[1;32mexcept\u001b[0m \u001b[0mcore\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_NotOkStatusException\u001b[0m \u001b[1;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "New_model_fitting = New_model.fit(train_generator, steps_per_epoch=200,\n",
    "                                 epochs=2,\n",
    "                                 validation_data=validation_generator,\n",
    "                                 validation_steps=50) # the model will train verrry slow! numpy is much better"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.3"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}