{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<h1>Dog Project Work Book</h1>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Brief Overview of Work Book"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "This workbook outlines the process of training a dog breed classifier for Udacity's Dog Breed Classifier project in the AI Nanodegree program.\n",
    "\n",
    "The things I tried out include:\n",
    "<ol>\n",
    "<li>\n",
    "<p>I used a data generator rather than loading in all the data. Because of this I was able to test out data augmentation relatively easily. I was also able to implement a preprocessing function and integrate it into the data generation. Additionally, I was able to adjust the image size easily, which had a noticable effect in improving the models.\n",
    "</p>\n",
    "        \n",
    "<p>Ultimately, data augmentation did not improve the models very much while increase the time to train the models. Although, when training a model from scratch it did show a benefit.\n",
    "</p>\n",
    "</li>\n",
    "<li>\n",
    "I tried out multiple pretrained models, using the conv layers to generate the bottleneck features of the data. Then trained a small classifier on top of each of those. I also tried to finetune the vgg19 network, but it didn't noticably improve the accuracy. Ultimately, the Inception based models performed the best, acheiving around 90% on the validation sets.\n",
    "</li>\n",
    "<li>\n",
    "Once I had several models performing well I organized them into an ensemble, averaging their predictions. This was able to produce a 93.6% accuracy on the test dataset.\n",
    "</li>\n",
    "<li>\n",
    "Finally, at the end of the notebook I have a short insight into the error the models ran into and also some visualization of what the model is seeing when it looks at an image. This section could be organized into a more detailed analysis which I may or may not do in the future.\n",
    "</li>\n",
    "</ol>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Import Statements"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using TensorFlow backend.\n"
     ]
    }
   ],
   "source": [
    "import keras\n",
    "from keras.preprocessing.image import ImageDataGenerator\n",
    "from keras.utils import np_utils\n",
    "from keras.utils.np_utils import to_categorical\n",
    "from keras.preprocessing import image                  \n",
    "from keras.layers import Conv2D, MaxPooling2D, GlobalAveragePooling2D\n",
    "from keras.layers import Dropout, Flatten, Dense, BatchNormalization\n",
    "from keras.models import Sequential, Model\n",
    "from keras.optimizers import Adam, SGD\n",
    "from keras.applications.vgg16 import VGG16\n",
    "from keras.callbacks import ModelCheckpoint\n",
    "\n",
    "import sklearn\n",
    "from sklearn.datasets import load_files\n",
    "from sklearn.utils import shuffle\n",
    "import numpy as np\n",
    "from glob import glob\n",
    "import random\n",
    "import matplotlib.pyplot as plt                        \n",
    "%matplotlib inline\n",
    "\n",
    "from PIL import ImageFile                            \n",
    "ImageFile.LOAD_TRUNCATED_IMAGES = True\n",
    "from tqdm import tqdm\n",
    "from keras.callbacks import ModelCheckpoint  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Organize the Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Data file locations\n",
    "train_data_path = 'dogImages/train/'\n",
    "valid_data_path = 'dogImages/valid/'\n",
    "test_data_path = 'dogImages/test/'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Data Augmentation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Augmentation and Preprocessing for Models"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The pretrained models that will be used were trained on images processed in two different ways. ResNet and VGG models expect images to be centered around zero and have BGR color channel order so below is a preprocessing function that can be applied to the data generator. This will apply the preprocessing to the image as it is input into the model. \n",
    "\n",
    "The Inception based model seem to perform best when the image values are normalized to be between 0 and 1 so each image is divided by 255 as it is input into the model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "code_folding": []
   },
   "outputs": [],
   "source": [
    "def imagenet_prep(image_tensor):\n",
    "    '''\n",
    "        Takes the image tensors for one image. Applies the preprocessing need for the Resnet and VGG models.\n",
    "        \n",
    "        Inputs: Image tensor of shape (width, height, colorchannels)\n",
    "        \n",
    "        Returns: Tensor of same shape with color channels flipped and values centered around zero for each color channel.\n",
    "    '''\n",
    "    # imagenet averages for RGB\n",
    "    image_net_mean = np.array([103.939,116.779,123.68])\n",
    "    image_tensor -= image_net_mean\n",
    "    \n",
    "    # Flip color channels\n",
    "    image_tensor = image_tensor[:, :, ::-1]\n",
    "    return image_tensor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "batch_size = 32\n",
    "\n",
    "# Augments the data for the Inception models\n",
    "inception_generator_da = ImageDataGenerator(rescale=1/255,\n",
    "                                                  rotation_range=10.0,\n",
    "                                                  width_shift_range=0.1,\n",
    "                                                  height_shift_range=0.1,\n",
    "                                                  shear_range=0.05,\n",
    "                                                  zoom_range=0.1,\n",
    "                                                  horizontal_flip=True,\n",
    "                                                  vertical_flip=False,\n",
    "                                                  fill_mode=\"reflect\")\n",
    "\n",
    "# prepares data for the Inception models without augmenting the data\n",
    "inception_generator = ImageDataGenerator(rescale=1/255)\n",
    "\n",
    "\n",
    "# Augments the data for the ResNet and VGG models\n",
    "resvgg_generator_da = ImageDataGenerator(preprocessing_function=imagenet_prep,\n",
    "                                               rescale=1.,\n",
    "                                               rotation_range=10.0,\n",
    "                                               width_shift_range=0.1,\n",
    "                                               height_shift_range=0.1,\n",
    "                                               shear_range=0.05,\n",
    "                                               zoom_range=0.1,\n",
    "                                               horizontal_flip=True,\n",
    "                                               vertical_flip=False,\n",
    "                                               fill_mode=\"reflect\")\n",
    "\n",
    "# prepares data for the ResNet and VGG models without augmenting the data\n",
    "resvgg_generator = ImageDataGenerator(preprocessing_function=imagenet_prep)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data Generators"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Here are the generators for the data. There are a few things to note:\n",
    "<ol>\n",
    "    <li>The target image size is 350x350px. The larger image size produces a better accuracy compared to say 224x224px or 299x299px which is standard for VGG and Inception respectively. One possible explaination of this is that to the model you are zooming in on the image which may make it easier to pick up details.</li>\n",
    "    <li>Shuffle is set to False. This is done because we will be passing the data from these generators through the conv layers of the pretrained models in order to obtain the bottle neck features. The data needs to stay in synch with the labels so to shuffle the data, it needs to be done with the bottleneck features and the labels. A step outline below.\n",
    "</ol>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "code_folding": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 6680 images belonging to 133 classes.\n",
      "Found 835 images belonging to 133 classes.\n",
      "Found 6680 images belonging to 133 classes.\n",
      "Found 835 images belonging to 133 classes.\n"
     ]
    }
   ],
   "source": [
    "inception_train_data = inception_generator.flow_from_directory(train_data_path,\n",
    "                                                       target_size=(350, 350),\n",
    "                                                       batch_size=batch_size,\n",
    "                                                       shuffle=False,\n",
    "                                                       class_mode='categorical')\n",
    "\n",
    "inception_valid_data = inception_generator.flow_from_directory(valid_data_path,\n",
    "                                                           target_size=(350, 350),\n",
    "                                                           batch_size=batch_size,\n",
    "                                                           shuffle=False,\n",
    "                                                           class_mode='categorical')\n",
    "\n",
    "resvgg_train_data = resvgg_generator.flow_from_directory(train_data_path,\n",
    "                                                       target_size=(350, 350),\n",
    "                                                       batch_size=batch_size,\n",
    "                                                       shuffle=False,\n",
    "                                                       class_mode='categorical')\n",
    "\n",
    "resvgg_valid_data = resvgg_generator.flow_from_directory(valid_data_path,\n",
    "                                                           target_size=(350, 350),\n",
    "                                                           batch_size=batch_size,\n",
    "                                                           shuffle=False,\n",
    "                                                           class_mode='categorical')\n",
    "\n",
    "train_samples = len(inception_train_data.filenames)\n",
    "validation_samples = len(inception_valid_data.filenames)\n",
    "dog_names = [item[20:-1] for item in sorted(glob(\"dogImages/train/*/\"))]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Labels will be the same for both types of models; needed to match back to bottleneck features\n",
    "train_labels = to_categorical(inception_train_data.classes)\n",
    "valid_labels = to_categorical(inception_valid_data.classes)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Breed Classifier for Dogs"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## CNN from Scratch"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Building a CNN from scratch we can make a generator that does shuffle the data because we won't be computing any bottleneck features."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Data Generator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 6680 images belonging to 133 classes.\n",
      "Found 835 images belonging to 133 classes.\n"
     ]
    }
   ],
   "source": [
    "scratch_train_data = inception_generator_da.flow_from_directory(train_data_path,\n",
    "                                                       target_size=(350, 350),\n",
    "                                                       batch_size=batch_size,\n",
    "                                                       shuffle=True,\n",
    "                                                       class_mode='categorical')\n",
    "\n",
    "scratch_valid_data = inception_generator.flow_from_directory(valid_data_path,\n",
    "                                                           target_size=(350, 350),\n",
    "                                                           batch_size=batch_size,\n",
    "                                                           shuffle=True,\n",
    "                                                           class_mode='categorical')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Build Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The architechture is based off the VGG architechture, but with just 1 conv layer for each block."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "model = Sequential()\n",
    "# Conv layer1\n",
    "model.add(Conv2D(32, 3, strides=(1,1), padding='same', activation='relu', input_shape=(350,350,3)))\n",
    "model.add(MaxPooling2D((2,2), strides= 2, padding='same'))\n",
    "\n",
    "# Conv layer2\n",
    "model.add(Conv2D(64, 3, strides=(1,1), padding='same', activation='relu'))\n",
    "model.add(MaxPooling2D((2,2), strides= 2, padding='same'))\n",
    "\n",
    "# Conv layer3\n",
    "model.add(Conv2D(128, 3, strides=(1,1), padding='same', activation='relu'))\n",
    "model.add(MaxPooling2D((2,2), strides= 2, padding='same'))\n",
    "\n",
    "# Conv layer4\n",
    "model.add(Conv2D(256, 3, strides=(1,1), padding='same', activation='relu'))\n",
    "model.add(MaxPooling2D((2,2), strides= 2, padding='same'))\n",
    "\n",
    "# Conv layer5\n",
    "model.add(Conv2D(256, 3, strides=(1,1), padding='same', activation='relu'))\n",
    "model.add(MaxPooling2D((2,2), strides= 2, padding='same'))\n",
    "\n",
    "#Flatten Layer\n",
    "model.add(GlobalAveragePooling2D())\n",
    "model.add(BatchNormalization())\n",
    "\n",
    "#Fully Connected Layer 2\n",
    "model.add(Dense(133, activation='softmax'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.load_weights(filepath='saved_models/weights.best.from_scratch_da.hdf5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(optimizer=Adam(), loss='categorical_crossentropy', metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Train Model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "After training it for several cycles(5 I think) of 10 epochs the model get to about 50% accuracy which is pretty good since there are 133 classes and only 6680 training examples.\n",
    "\n",
    "Pure guessing would only be 00.75%."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      " - 218s - loss: 0.2442 - acc: 0.9242 - val_loss: 2.9442 - val_acc: 0.4766\n",
      "Epoch 2/10\n",
      " - 214s - loss: 0.2494 - acc: 0.9245 - val_loss: 2.7488 - val_acc: 0.4778\n",
      "Epoch 3/10\n",
      " - 206s - loss: 0.2488 - acc: 0.9224 - val_loss: 2.7818 - val_acc: 0.4743\n",
      "Epoch 4/10\n",
      " - 209s - loss: 0.2389 - acc: 0.9243 - val_loss: 2.7629 - val_acc: 0.4754\n",
      "Epoch 5/10\n",
      " - 202s - loss: 0.2670 - acc: 0.9183 - val_loss: 3.2061 - val_acc: 0.4778\n",
      "Epoch 6/10\n",
      " - 211s - loss: 0.2631 - acc: 0.9129 - val_loss: 2.9207 - val_acc: 0.4778\n",
      "Epoch 7/10\n",
      " - 216s - loss: 0.2320 - acc: 0.9223 - val_loss: 2.6750 - val_acc: 0.4778\n",
      "Epoch 8/10\n",
      " - 223s - loss: 0.2335 - acc: 0.9244 - val_loss: 3.2669 - val_acc: 0.4443\n",
      "Epoch 9/10\n",
      " - 218s - loss: 0.2186 - acc: 0.9285 - val_loss: 2.9923 - val_acc: 0.4850\n",
      "Epoch 10/10\n",
      " - 205s - loss: 0.2538 - acc: 0.9219 - val_loss: 2.6119 - val_acc: 0.4958\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x1cd5cdf9b00>"
      ]
     },
     "execution_count": 47,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "checkpointer = ModelCheckpoint(filepath='saved_models/weights.best.from_scratch_da.hdf5', \n",
    "                               verbose=0, save_best_only=True)\n",
    "\n",
    "model.fit_generator(scratch_train_data, steps_per_epoch=(1+train_samples//batch_size), epochs=10,\n",
    "                    callbacks=[checkpointer], verbose=2, \n",
    "                    validation_data=scratch_valid_data,\n",
    "                    validation_steps=(1+validation_samples//batch_size))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Build Classifiers Using Pretrained Models"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The same process will be used for 6 different models and follows these steps:\n",
    "    <ol>\n",
    "    <li>Import and load the conv layers of the pretrained model.</li>\n",
    "    <li>Compute the bottleneck features by using the pretrained model to predict on the data generater.</li>\n",
    "    <li>Shuffle the bottleneck features and labels corresponding features and labels stay paired.</li>\n",
    "    <li>Construct the model that will go ontop of the pretrained conv layers.</li>\n",
    "    <li>Train the model with the bottleneck features and labels, saving the weights.</li>\n",
    "    </ol>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Inception Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.applications.inception_v3 import InceptionV3\n",
    "inception = InceptionV3(include_top=False, weights='imagenet')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "inception_bottleneck = inception.predict_generator(inception_train_data, steps=1)\n",
    "inception_valid = inception.predict_generator(inception_valid_data, steps=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "shuf_train, shuf_train_labels = shuffle(inception_bottleneck, train_labels)\n",
    "shuf_val, shuf_val_labels = shuffle(inception_valid, valid_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "inception_model = Sequential()\n",
    "inception_model.add(GlobalAveragePooling2D(input_shape=(9,9,2048)))\n",
    "inception_model.add(BatchNormalization())\n",
    "\n",
    "inception_model.add(Dense(133, activation='softmax'))\n",
    "inception_model.compile(optimizer=Adam(), loss='categorical_crossentropy', metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "checkpointer = ModelCheckpoint(filepath='saved_models/weights.best.inception6.hdf5', \n",
    "                               verbose=0, save_best_only=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 6680 samples, validate on 835 samples\n",
      "Epoch 1/10\n",
      " - 7s - loss: 0.9024 - acc: 0.7671 - val_loss: 0.6431 - val_acc: 0.8599\n",
      "Epoch 2/10\n",
      " - 6s - loss: 0.2002 - acc: 0.9346 - val_loss: 0.3820 - val_acc: 0.8814\n",
      "Epoch 3/10\n",
      " - 6s - loss: 0.0947 - acc: 0.9714 - val_loss: 0.3923 - val_acc: 0.8970\n",
      "Epoch 4/10\n",
      " - 6s - loss: 0.0552 - acc: 0.9847 - val_loss: 0.3996 - val_acc: 0.8970\n",
      "Epoch 5/10\n",
      " - 7s - loss: 0.0322 - acc: 0.9925 - val_loss: 0.3705 - val_acc: 0.8994\n",
      "Epoch 6/10\n",
      " - 6s - loss: 0.0233 - acc: 0.9955 - val_loss: 0.3947 - val_acc: 0.9078\n",
      "Epoch 7/10\n",
      " - 6s - loss: 0.0239 - acc: 0.9951 - val_loss: 0.3784 - val_acc: 0.9102\n",
      "Epoch 8/10\n",
      " - 6s - loss: 0.0207 - acc: 0.9952 - val_loss: 0.4408 - val_acc: 0.8862\n",
      "Epoch 9/10\n",
      " - 6s - loss: 0.0206 - acc: 0.9964 - val_loss: 0.4234 - val_acc: 0.9006\n",
      "Epoch 10/10\n",
      " - 6s - loss: 0.0209 - acc: 0.9942 - val_loss: 0.4817 - val_acc: 0.8898\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x22ee5dc1ba8>"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inception_model.fit(shuf_train, shuf_train_labels, \n",
    "                        validation_data=(shuf_val, shuf_val_labels),\n",
    "                        epochs=10, batch_size=32, callbacks=[checkpointer], verbose=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### VGG16 Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.applications.vgg16 import VGG16\n",
    "vgg16 = VGG16(include_top=False, weights='imagenet')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [],
   "source": [
    "vgg16_bottleneck = vgg16.predict_generator(resvgg_train_data, steps=1)\n",
    "vgg16_valid = vgg16.predict_generator(resvgg_valid_data, steps=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {},
   "outputs": [],
   "source": [
    "shuf_train, shuf_train_labels = shuffle(vgg16_bottleneck, train_labels)\n",
    "shuf_val, shuf_val_labels = shuffle(vgg16_valid, valid_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "vgg16_model = Sequential()\n",
    "vgg16_model.add(GlobalAveragePooling2D(input_shape=(None, None, 512)))\n",
    "vgg16_model.add(BatchNormalization())\n",
    "\n",
    "vgg16_model.add(Dense(133, activation='softmax'))\n",
    "vgg16_model.compile(optimizer=Adam(), loss='categorical_crossentropy', metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "checkpointer = ModelCheckpoint(filepath='saved_models/weights.best.vgg16.hdf5', \n",
    "                               verbose=0, save_best_only=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 6680 samples, validate on 835 samples\n",
      "Epoch 1/10\n",
      " - 3s - loss: 0.0647 - acc: 0.9952 - val_loss: 0.6923 - val_acc: 0.8000\n",
      "Epoch 2/10\n",
      " - 4s - loss: 0.0521 - acc: 0.9954 - val_loss: 0.6716 - val_acc: 0.8000\n",
      "Epoch 3/10\n",
      " - 3s - loss: 0.0461 - acc: 0.9973 - val_loss: 0.6840 - val_acc: 0.7892\n",
      "Epoch 4/10\n",
      " - 3s - loss: 0.0392 - acc: 0.9970 - val_loss: 0.6905 - val_acc: 0.7952\n",
      "Epoch 5/10\n",
      " - 3s - loss: 0.0342 - acc: 0.9972 - val_loss: 0.7044 - val_acc: 0.7976\n",
      "Epoch 6/10\n",
      " - 3s - loss: 0.0310 - acc: 0.9976 - val_loss: 0.6888 - val_acc: 0.7976\n",
      "Epoch 7/10\n",
      " - 3s - loss: 0.0290 - acc: 0.9981 - val_loss: 0.7023 - val_acc: 0.7988\n",
      "Epoch 8/10\n",
      " - 3s - loss: 0.0245 - acc: 0.9984 - val_loss: 0.7132 - val_acc: 0.7964\n",
      "Epoch 9/10\n",
      " - 3s - loss: 0.0246 - acc: 0.9972 - val_loss: 0.7263 - val_acc: 0.8024\n",
      "Epoch 10/10\n",
      " - 3s - loss: 0.0235 - acc: 0.9978 - val_loss: 0.7079 - val_acc: 0.7988\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x2045144ddd8>"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vgg16_model.fit(shuf_train, shuf_train_labels, \n",
    "                        validation_data=(shuf_val, shuf_val_labels),\n",
    "                        epochs=10, batch_size=32, callbacks=[checkpointer], verbose=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### VGG19 Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.applications.vgg19 import VGG19\n",
    "vgg19 = VGG19(include_top=False, weights='imagenet')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 85,
   "metadata": {},
   "outputs": [],
   "source": [
    "vgg19_bottleneck = vgg19.predict_generator(resvgg_train_data, steps=1)\n",
    "vgg19_valid = vgg19.predict_generator(resvgg_valid_data, steps=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 86,
   "metadata": {},
   "outputs": [],
   "source": [
    "shuf_train, shuf_train_labels = shuffle(vgg19_bottleneck, train_labels)\n",
    "shuf_val, shuf_val_labels = shuffle(vgg19_valid, valid_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "vgg19_model = Sequential()\n",
    "vgg19_model.add(GlobalAveragePooling2D(input_shape=(None, None, 512)))\n",
    "vgg19_model.add(BatchNormalization())\n",
    "\n",
    "vgg19_model.add(Dense(133, activation='softmax'))\n",
    "vgg19_model.compile(optimizer=Adam(), loss='categorical_crossentropy', metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 88,
   "metadata": {},
   "outputs": [],
   "source": [
    "checkpointer = ModelCheckpoint(filepath='saved_models/weights.best.vgg19.hdf5', \n",
    "                               verbose=0, save_best_only=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 90,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 6680 samples, validate on 835 samples\n",
      "Epoch 1/10\n",
      " - 3s - loss: 0.0581 - acc: 0.9954 - val_loss: 0.7184 - val_acc: 0.7844\n",
      "Epoch 2/10\n",
      " - 3s - loss: 0.0472 - acc: 0.9970 - val_loss: 0.7210 - val_acc: 0.7964\n",
      "Epoch 3/10\n",
      " - 3s - loss: 0.0398 - acc: 0.9978 - val_loss: 0.7322 - val_acc: 0.7880\n",
      "Epoch 4/10\n",
      " - 3s - loss: 0.0402 - acc: 0.9967 - val_loss: 0.7499 - val_acc: 0.7832\n",
      "Epoch 5/10\n",
      " - 3s - loss: 0.0332 - acc: 0.9972 - val_loss: 0.7548 - val_acc: 0.7820\n",
      "Epoch 6/10\n",
      " - 3s - loss: 0.0285 - acc: 0.9981 - val_loss: 0.7546 - val_acc: 0.7868\n",
      "Epoch 7/10\n",
      " - 3s - loss: 0.0232 - acc: 0.9987 - val_loss: 0.7600 - val_acc: 0.7916\n",
      "Epoch 8/10\n",
      " - 3s - loss: 0.0240 - acc: 0.9975 - val_loss: 0.7543 - val_acc: 0.8036\n",
      "Epoch 9/10\n",
      " - 3s - loss: 0.0219 - acc: 0.9976 - val_loss: 0.7726 - val_acc: 0.7952\n",
      "Epoch 10/10\n",
      " - 3s - loss: 0.0213 - acc: 0.9982 - val_loss: 0.7824 - val_acc: 0.7880\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x2049b8caf98>"
      ]
     },
     "execution_count": 90,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vgg19_model.fit(shuf_train, shuf_train_labels, \n",
    "                        validation_data=(shuf_val, shuf_val_labels),\n",
    "                        epochs=10, batch_size=32, callbacks=[checkpointer], verbose=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Fine-Tune VGG"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Use functional API to fine tune following the process outlined <a href=https://www.kaggle.com/chmaxx/finetune-vgg16-0-97-with-minimal-effort>here.</a> Unfortunately after finetuning the model there is only about a 1% increase in accuracy while drastically increasing the training time.\n",
    "\n",
    "And further, VGG is the only model we seem to be able to fine tune in Keras. Attempts to fine tune any of the models with Batch Normalization give much lower training accuracy than what has been shown with the current training process. After some investigation this seems to be bug with Keras or at least several people have run into the same problem and no resolution to the issue has been found."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Keep same architechture as before\n",
    "x = vgg19.get_layer('block5_pool').output\n",
    "x = GlobalAveragePooling2D()(x)\n",
    "x = BatchNormalization()(x)\n",
    "x = Dense(133, activation='softmax')(x)\n",
    "\n",
    "vgg19_finetune = Model(vgg19.input, x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Freeze Convolutional layers to train top model first\n",
    "for layer in vgg19.layers:\n",
    "    layer.trainable = False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "vgg19_finetune.compile(optimizer=Adam(), loss='categorical_crossentropy', metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 6680 images belonging to 133 classes.\n",
      "Found 835 images belonging to 133 classes.\n"
     ]
    }
   ],
   "source": [
    "finetune_train_data = resvgg_generator_da.flow_from_directory(train_data_path,\n",
    "                                                       target_size=(350, 350),\n",
    "                                                       batch_size=batch_size,\n",
    "                                                       shuffle=True,\n",
    "                                                       class_mode='categorical')\n",
    "\n",
    "finetune_valid_data = resvgg_generator_da.flow_from_directory(valid_data_path,\n",
    "                                                           target_size=(350, 350),\n",
    "                                                           batch_size=batch_size,\n",
    "                                                           shuffle=True,\n",
    "                                                           class_mode='categorical')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      " - 253s - loss: 0.1324 - acc: 0.9723 - val_loss: 0.7208 - val_acc: 0.7916\n",
      "Epoch 2/10\n",
      " - 252s - loss: 0.1177 - acc: 0.9759 - val_loss: 0.6880 - val_acc: 0.7940\n",
      "Epoch 3/10\n",
      " - 236s - loss: 0.1015 - acc: 0.9776 - val_loss: 0.7105 - val_acc: 0.7964\n",
      "Epoch 4/10\n",
      " - 223s - loss: 0.0929 - acc: 0.9833 - val_loss: 0.7500 - val_acc: 0.7868\n",
      "Epoch 5/10\n",
      " - 229s - loss: 0.0861 - acc: 0.9833 - val_loss: 0.7915 - val_acc: 0.7832\n",
      "Epoch 6/10\n",
      " - 244s - loss: 0.0834 - acc: 0.9821 - val_loss: 0.7618 - val_acc: 0.7832\n",
      "Epoch 7/10\n",
      " - 242s - loss: 0.0731 - acc: 0.9845 - val_loss: 0.7543 - val_acc: 0.7844\n",
      "Epoch 8/10\n",
      " - 224s - loss: 0.0654 - acc: 0.9865 - val_loss: 0.7502 - val_acc: 0.7904\n",
      "Epoch 9/10\n",
      " - 221s - loss: 0.0619 - acc: 0.9885 - val_loss: 0.8143 - val_acc: 0.7653\n",
      "Epoch 10/10\n",
      " - 235s - loss: 0.0652 - acc: 0.9868 - val_loss: 0.7650 - val_acc: 0.7952\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x1cbba481a90>"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vgg19_finetune.fit_generator(finetune_train_data, steps_per_epoch=(1+train_samples//batch_size), epochs=10,\n",
    "                    verbose=2, \n",
    "                    validation_data=finetune_valid_data,\n",
    "                    validation_steps=(1+validation_samples//batch_size))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "17"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "layers = vgg19_finetune.layers\n",
    "layers.index(vgg19_finetune.get_layer('block5_conv1'))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "for layer in vgg19_finetune.layers[:17]:\n",
    "    layer.trainable = False\n",
    "for layer in vgg19_finetune.layers[17:]:\n",
    "    layer.trainable = True"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "vgg19_finetune.compile(optimizer=SGD(lr=0.0001, momentum=0.9),  loss='categorical_crossentropy', metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "checkpointer = ModelCheckpoint(filepath='saved_models/weights.best.vgg19ft.hdf5', \n",
    "                               verbose=0, save_best_only=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      " - 236s - loss: 0.0524 - acc: 0.9901 - val_loss: 0.7355 - val_acc: 0.7976\n",
      "Epoch 2/10\n",
      " - 232s - loss: 0.0421 - acc: 0.9929 - val_loss: 0.6979 - val_acc: 0.8024\n",
      "Epoch 3/10\n",
      " - 244s - loss: 0.0403 - acc: 0.9922 - val_loss: 0.7089 - val_acc: 0.8024\n",
      "Epoch 4/10\n",
      " - 251s - loss: 0.0373 - acc: 0.9936 - val_loss: 0.7016 - val_acc: 0.8060\n",
      "Epoch 5/10\n",
      " - 245s - loss: 0.0349 - acc: 0.9940 - val_loss: 0.6989 - val_acc: 0.8024\n",
      "Epoch 6/10\n",
      " - 242s - loss: 0.0313 - acc: 0.9945 - val_loss: 0.6767 - val_acc: 0.8216\n",
      "Epoch 7/10\n",
      " - 233s - loss: 0.0312 - acc: 0.9953 - val_loss: 0.6871 - val_acc: 0.8084\n",
      "Epoch 8/10\n",
      " - 254s - loss: 0.0287 - acc: 0.9949 - val_loss: 0.7443 - val_acc: 0.8060\n",
      "Epoch 9/10\n",
      " - 246s - loss: 0.0283 - acc: 0.9955 - val_loss: 0.6948 - val_acc: 0.7928\n",
      "Epoch 10/10\n",
      " - 248s - loss: 0.0284 - acc: 0.9957 - val_loss: 0.6728 - val_acc: 0.8180\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x1cd5e94c550>"
      ]
     },
     "execution_count": 38,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vgg19_finetune.fit_generator(finetune_train_data, steps_per_epoch=(1+train_samples//batch_size), epochs=10,\n",
    "                    verbose=2, \n",
    "                    validation_data=finetune_valid_data,\n",
    "                    validation_steps=(1+validation_samples//batch_size))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### ResNet Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.applications.resnet50 import ResNet50\n",
    "resnet = ResNet50(include_top=False, weights='imagenet')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [],
   "source": [
    "resnet_bottleneck = resnet.predict_generator(resvgg_train_data, steps=1)\n",
    "resnet_valid = resnet.predict_generator(resvgg_valid_data, steps=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 74,
   "metadata": {},
   "outputs": [],
   "source": [
    "shuf_train, shuf_train_labels = shuffle(resnet_bottleneck, train_labels)\n",
    "shuf_val, shuf_val_labels = shuffle(resnet_valid, valid_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "resnet_model = Sequential()\n",
    "resnet_model.add(GlobalAveragePooling2D(input_shape=(None, None, 2048)))\n",
    "resnet_model.add(BatchNormalization())\n",
    "\n",
    "resnet_model.add(Dense(133, activation='softmax'))\n",
    "resnet_model.compile(optimizer=Adam(), loss='categorical_crossentropy', metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "checkpointer = ModelCheckpoint(filepath='saved_models/weights.best.resnet.hdf5', \n",
    "                               verbose=0, save_best_only=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 6680 samples, validate on 835 samples\n",
      "Epoch 1/10\n",
      " - 3s - loss: 0.0211 - acc: 0.9978 - val_loss: 0.8427 - val_acc: 0.7796\n",
      "Epoch 2/10\n",
      " - 3s - loss: 0.0158 - acc: 0.9982 - val_loss: 0.8752 - val_acc: 0.7665\n",
      "Epoch 3/10\n",
      " - 3s - loss: 0.0179 - acc: 0.9978 - val_loss: 0.9030 - val_acc: 0.7701\n",
      "Epoch 4/10\n",
      " - 3s - loss: 0.0180 - acc: 0.9979 - val_loss: 0.8756 - val_acc: 0.7796\n",
      "Epoch 5/10\n",
      " - 3s - loss: 0.0121 - acc: 0.9982 - val_loss: 0.8919 - val_acc: 0.7737\n",
      "Epoch 6/10\n",
      " - 3s - loss: 0.0142 - acc: 0.9984 - val_loss: 0.8654 - val_acc: 0.7916\n",
      "Epoch 7/10\n",
      " - 2s - loss: 0.0129 - acc: 0.9987 - val_loss: 0.9034 - val_acc: 0.7760\n",
      "Epoch 8/10\n",
      " - 2s - loss: 0.0218 - acc: 0.9972 - val_loss: 0.8933 - val_acc: 0.7904\n",
      "Epoch 9/10\n",
      " - 2s - loss: 0.0223 - acc: 0.9970 - val_loss: 0.9916 - val_acc: 0.7605\n",
      "Epoch 10/10\n",
      " - 2s - loss: 0.0396 - acc: 0.9906 - val_loss: 1.1564 - val_acc: 0.7449\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x2048e8e5f98>"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "resnet_model.fit(shuf_train, shuf_train_labels, \n",
    "                        validation_data=(shuf_val, shuf_val_labels),\n",
    "                        epochs=10, batch_size=32, callbacks=[checkpointer], verbose=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Xception Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.applications.xception import Xception, preprocess_input\n",
    "xception = Xception(include_top=False, weights='imagenet')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 83,
   "metadata": {},
   "outputs": [],
   "source": [
    "xception_bottleneck = xception.predict_generator(inception_train_data, steps=1)\n",
    "xception_valid = xception.predict_generator(inception_valid_data, steps=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 84,
   "metadata": {},
   "outputs": [],
   "source": [
    "shuf_train, shuf_train_labels = shuffle(xception_bottleneck, train_labels)\n",
    "shuf_val, shuf_val_labels = shuffle(xception_valid, valid_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "xception_model = Sequential()\n",
    "xception_model.add(GlobalAveragePooling2D(input_shape=(None, None, 2048)))\n",
    "xception_model.add(BatchNormalization())\n",
    "\n",
    "xception_model.add(Dense(133, activation='softmax'))\n",
    "xception_model.compile(optimizer=Adam(), loss='categorical_crossentropy', metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 111,
   "metadata": {},
   "outputs": [],
   "source": [
    "checkpointer = ModelCheckpoint(filepath='saved_models/weights.best.xception.hdf5', \n",
    "                               verbose=0, save_best_only=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 112,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 6680 samples, validate on 835 samples\n",
      "Epoch 1/10\n",
      " - 43s - loss: 0.8361 - acc: 0.7840 - val_loss: 1.0529 - val_acc: 0.8659\n",
      "Epoch 2/10\n",
      " - 34s - loss: 0.2171 - acc: 0.9301 - val_loss: 0.3965 - val_acc: 0.8731\n",
      "Epoch 3/10\n",
      " - 30s - loss: 0.1255 - acc: 0.9581 - val_loss: 0.4012 - val_acc: 0.8850\n",
      "Epoch 4/10\n",
      " - 33s - loss: 0.0898 - acc: 0.9753 - val_loss: 0.3887 - val_acc: 0.9030\n",
      "Epoch 5/10\n",
      " - 32s - loss: 0.0650 - acc: 0.9798 - val_loss: 0.4081 - val_acc: 0.8970\n",
      "Epoch 6/10\n",
      " - 33s - loss: 0.0450 - acc: 0.9856 - val_loss: 0.3987 - val_acc: 0.9066\n",
      "Epoch 7/10\n",
      " - 34s - loss: 0.0345 - acc: 0.9903 - val_loss: 0.3981 - val_acc: 0.9054\n",
      "Epoch 8/10\n",
      " - 33s - loss: 0.0270 - acc: 0.9931 - val_loss: 0.4116 - val_acc: 0.9042\n",
      "Epoch 9/10\n",
      " - 33s - loss: 0.0346 - acc: 0.9913 - val_loss: 0.4382 - val_acc: 0.8958\n",
      "Epoch 10/10\n",
      " - 33s - loss: 0.0281 - acc: 0.9939 - val_loss: 0.4461 - val_acc: 0.9162\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x2a403db34e0>"
      ]
     },
     "execution_count": 112,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "xception_model.fit(shuf_train, shuf_train_labels, \n",
    "                        validation_data=(shuf_val, shuf_val_labels),\n",
    "                        epochs=10, batch_size=32, callbacks=[checkpointer], verbose=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Inception-ResNet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.applications.inception_resnet_v2 import InceptionResNetV2, preprocess_input\n",
    "inceptionresnet = InceptionResNetV2(include_top=False, weights='imagenet')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 120,
   "metadata": {},
   "outputs": [],
   "source": [
    "inceptionresnet_bottleneck = inceptionresnet.predict_generator(inception_train_data, steps=1)\n",
    "inceptionresnet_valid = inceptionresnet.predict_generator(inception_valid_data, steps=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 121,
   "metadata": {},
   "outputs": [],
   "source": [
    "shuf_train, shuf_train_labels = shuffle(inceptionresnet_bottleneck, train_labels)\n",
    "shuf_val, shuf_val_labels = shuffle(inceptionresnet_valid, valid_labels)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "inceptionresnet_model = Sequential()\n",
    "inceptionresnet_model.add(GlobalAveragePooling2D(input_shape=(None, None, 1536)))\n",
    "inceptionresnet_model.add(BatchNormalization())\n",
    "\n",
    "inceptionresnet_model.add(Dense(133, activation='softmax'))\n",
    "inceptionresnet_model.compile(optimizer=Adam(), loss='categorical_crossentropy', metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 147,
   "metadata": {},
   "outputs": [],
   "source": [
    "checkpointer = ModelCheckpoint(filepath='saved_models/weights.best.inceptionresnet6.hdf5', \n",
    "                               verbose=0, save_best_only=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 149,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 6680 samples, validate on 835 samples\n",
      "Epoch 1/10\n",
      " - 6s - loss: 0.0600 - acc: 0.9801 - val_loss: 0.4551 - val_acc: 0.8946\n",
      "Epoch 2/10\n",
      " - 6s - loss: 0.0524 - acc: 0.9843 - val_loss: 0.5055 - val_acc: 0.8898\n",
      "Epoch 3/10\n",
      " - 6s - loss: 0.0539 - acc: 0.9831 - val_loss: 0.4982 - val_acc: 0.9006\n",
      "Epoch 4/10\n",
      " - 6s - loss: 0.0658 - acc: 0.9802 - val_loss: 0.5701 - val_acc: 0.8922\n",
      "Epoch 5/10\n",
      " - 6s - loss: 0.0656 - acc: 0.9804 - val_loss: 0.6207 - val_acc: 0.8910\n",
      "Epoch 6/10\n",
      " - 6s - loss: 0.0579 - acc: 0.9811 - val_loss: 0.5076 - val_acc: 0.8886\n",
      "Epoch 7/10\n",
      " - 6s - loss: 0.0565 - acc: 0.9826 - val_loss: 0.5696 - val_acc: 0.8862\n",
      "Epoch 8/10\n",
      " - 7s - loss: 0.0450 - acc: 0.9853 - val_loss: 0.5189 - val_acc: 0.8922\n",
      "Epoch 9/10\n",
      " - 6s - loss: 0.0445 - acc: 0.9862 - val_loss: 0.5866 - val_acc: 0.9018\n",
      "Epoch 10/10\n",
      " - 6s - loss: 0.0613 - acc: 0.9831 - val_loss: 0.6137 - val_acc: 0.8994\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x2a46f02f7b8>"
      ]
     },
     "execution_count": 149,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "inceptionresnet_model.fit(shuf_train, shuf_train_labels, \n",
    "                        validation_data=(shuf_val, shuf_val_labels),\n",
    "                        epochs=10, batch_size=32, callbacks=[checkpointer], verbose=2)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Organize Models into an Ensemble"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Hybrid Ensemble"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "To organize all the models into an ensemble The predictions for the test data from each model are averaged and compared to the labels.\n",
    "\n",
    "This process is outlined in the fast.ai <a href='http://course.fast.ai/'>courses</a>."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 75,
   "metadata": {},
   "outputs": [],
   "source": [
    "feature_extractors = {'inception_model': inception,\n",
    "                      'xception_model': xception,\n",
    "                      'inceptionresnet_model': inceptionresnet}\n",
    "\n",
    "resvgg_extractors = {'vgg16_model': vgg16,\n",
    "                     'vgg19_model': vgg19,\n",
    "                     'resnet_model': resnet}\n",
    "\n",
    "models = {'inception_model': inception_model,\n",
    "         'vgg16_model': vgg16_model,\n",
    "         'vgg19_model': vgg19_model,\n",
    "         'resnet_model': resnet_model,\n",
    "         'xception_model': xception_model,\n",
    "         'inceptionresnet_model': inceptionresnet_model}\n",
    "\n",
    "weights = {'inception_model': 'saved_models/weights.best.inception4.hdf5',\n",
    "         'vgg16_model': 'saved_models/weights.best.vgg16.hdf5',\n",
    "         'vgg19_model': 'saved_models/weights.best.vgg19.hdf5',\n",
    "         'resnet_model': 'saved_models/weights.best.resnet.hdf5',\n",
    "         'xception_model': 'saved_models/weights.best.xception.hdf5',\n",
    "         'inceptionresnet_model': 'saved_models/weights.best.inceptionresnet.hdf5'}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 76,
   "metadata": {},
   "outputs": [],
   "source": [
    "for key in models.keys():\n",
    "    models[key].load_weights(weights[key])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Found 836 images belonging to 133 classes.\n",
      "Found 836 images belonging to 133 classes.\n"
     ]
    }
   ],
   "source": [
    "inception_test_data = inception_generator.flow_from_directory(test_data_path,\n",
    "                                                           target_size=(350, 350),\n",
    "                                                           batch_size=batch_size,\n",
    "                                                           shuffle=False,\n",
    "                                                           class_mode='categorical')\n",
    "\n",
    "resvgg_test_data = resvgg_generator.flow_from_directory(test_data_path,\n",
    "                                                           target_size=(350, 350),\n",
    "                                                           batch_size=batch_size,\n",
    "                                                           shuffle=False,\n",
    "                                                           class_mode='categorical')\n",
    "test_samples = len(inception_test_data.filenames)\n",
    "test_labels = to_categorical(inception_test_data.classes)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 78,
   "metadata": {},
   "outputs": [],
   "source": [
    "# get bottle neck features on test set\n",
    "predictions = []\n",
    "for key in models.keys():\n",
    "    if key in feature_extractors:\n",
    "        features = feature_extractors[key].predict_generator(inception_test_data, steps=1)\n",
    "        prediction = models[key].predict(features)\n",
    "        predictions.append(prediction)\n",
    "\n",
    "# get model predictions on bottleneck features"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Take weighted average\n",
    "# Evaluate predictions\n",
    "predictions = np.stack(predictions)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 80,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_pred = np.mean(predictions, axis=0, keepdims=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 81,
   "metadata": {},
   "outputs": [],
   "source": [
    "import keras.backend as K"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 82,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.90909094"
      ]
     },
     "execution_count": 82,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "K.eval(keras.metrics.categorical_accuracy(test_labels, test_pred)).mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Inception Based Ensemble"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The Hybrid Ensemble gets up to about 91% which is ok, but the inception based models were already up close to that, while the other models were only reaching about 80% accuracy. For this reason I decided to try an ensemble with multiple version of the inception based models.\n",
    "\n",
    "This boosted the accuracy of the model further to 93.6%."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "inception_features = inception.predict_generator(inception_test_data, steps=1)\n",
    "xception_features = xception.predict_generator(inception_test_data, steps=1)\n",
    "inceptionresnet_features = inceptionresnet.predict_generator(inception_test_data, steps=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "inception_weights = [\n",
    "    'saved_models/weights.best.inception.hdf5',\n",
    "    'saved_models/weights.best.inception2.hdf5',\n",
    "    'saved_models/weights.best.inception3.hdf5',\n",
    "    'saved_models/weights.best.inception4.hdf5',\n",
    "    'saved_models/weights.best.inception5.hdf5',\n",
    "    'saved_models/weights.best.inception6.hdf5'\n",
    "]\n",
    "\n",
    "xception_weights = [\n",
    "    'saved_models/weights.best.xception.hdf5',\n",
    "    'saved_models/weights.best.xception2.hdf5',\n",
    "    'saved_models/weights.best.xception3.hdf5',\n",
    "    'saved_models/weights.best.xception4.hdf5',\n",
    "    'saved_models/weights.best.xception5.hdf5',\n",
    "    'saved_models/weights.best.xception6.hdf5'\n",
    "]\n",
    "\n",
    "inceptionresnet_weights = [\n",
    "    'saved_models/weights.best.inceptionresnet.hdf5',\n",
    "    'saved_models/weights.best.inceptionresnet2.hdf5',\n",
    "    'saved_models/weights.best.inceptionresnet3.hdf5',\n",
    "    'saved_models/weights.best.inceptionresnet4.hdf5',\n",
    "    'saved_models/weights.best.inceptionresnet5.hdf5',\n",
    "    'saved_models/weights.best.inceptionresnet6.hdf5'\n",
    "]\n",
    "\n",
    "predictions = np.zeros((test_samples, 133), dtype=np.float)\n",
    "\n",
    "for weight_path in inception_weights:\n",
    "    inception_model.load_weights(weight_path)\n",
    "    model_prediction = inception_model.predict(inception_features)\n",
    "    predictions += model_prediction\n",
    "for weight_path in xception_weights:\n",
    "    xception_model.load_weights(weight_path)\n",
    "    model_prediction = xception_model.predict(xception_features)\n",
    "    predictions += model_prediction\n",
    "for weight_path in inceptionresnet_weights:\n",
    "    inceptionresnet_model.load_weights(weight_path)\n",
    "    model_prediction = inceptionresnet_model.predict(inceptionresnet_features)\n",
    "    predictions += model_prediction\n",
    "\n",
    "predictions /= 18"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "import keras.backend as K\n",
    "scores = K.eval(keras.metrics.categorical_accuracy(test_labels, predictions))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.93660289"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scores.mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Looking at the Model Errors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.datasets import load_files\n",
    "data = load_files(test_data_path, shuffle=False)\n",
    "test_files = np.array(data['filenames'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "wrong_images = []\n",
    "for i, answer in enumerate(scores):\n",
    "    if answer == 0:\n",
    "        correct_dog = dog_names[np.argmax(test_labels[i])] + ': ' + str(predictions[i][np.argmax(test_labels[i])])\n",
    "        dog_prediction = dog_names[np.argmax(predictions[i])]+ ': ' + str(np.max(predictions[i]))\n",
    "        \n",
    "        wrong_images.append((correct_dog, dog_prediction, test_files[i]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Below is a list of the misclassified images with the prediction value for the correct breed, followed by the prediction value for the predicted breed, followed by the file of the actual image.\n",
    "\n",
    "It is reassuring that the wrong predictions appear to be for breeds that look quite similar to the correct breed. We also see that the prediction value for the correct breed is often still pretty high, even if it wasn't the highest value."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[('Affenpinscher: 0.0723859019502',\n",
       "  'Havanese: 0.592816216043',\n",
       "  'dogImages/test/001.Affenpinscher\\\\Affenpinscher_00058.jpg'),\n",
       " ('American_foxhound: 0.116203999127',\n",
       "  'Beagle: 0.881220284435',\n",
       "  'dogImages/test/007.American_foxhound\\\\American_foxhound_00524.jpg'),\n",
       " ('American_water_spaniel: 0.250119773949',\n",
       "  'Boykin_spaniel: 0.649206562175',\n",
       "  'dogImages/test/009.American_water_spaniel\\\\American_water_spaniel_00624.jpg'),\n",
       " ('Australian_cattle_dog: 0.250413881149',\n",
       "  'Canaan_dog: 0.67557344834',\n",
       "  'dogImages/test/011.Australian_cattle_dog\\\\Australian_cattle_dog_00761.jpg'),\n",
       " ('Australian_shepherd: 0.270271796759',\n",
       "  'Border_collie: 0.463511261675',\n",
       "  'dogImages/test/012.Australian_shepherd\\\\Australian_shepherd_00868.jpg'),\n",
       " ('Black_russian_terrier: 0.300719369048',\n",
       "  'Portuguese_water_dog: 0.442908981983',\n",
       "  'dogImages/test/026.Black_russian_terrier\\\\Black_russian_terrier_01827.jpg'),\n",
       " ('Bluetick_coonhound: 0.0372440269372',\n",
       "  'Pointer: 0.332300558402',\n",
       "  'dogImages/test/028.Bluetick_coonhound\\\\Bluetick_coonhound_01968.jpg'),\n",
       " ('Brittany: 0.0886546824375',\n",
       "  'Welsh_springer_spaniel: 0.460033175432',\n",
       "  'dogImages/test/037.Brittany\\\\Brittany_02607.jpg'),\n",
       " ('Cane_corso: 0.209250953913',\n",
       "  'Boxer: 0.623807925731',\n",
       "  'dogImages/test/044.Cane_corso\\\\Cane_corso_03114.jpg'),\n",
       " ('Cardigan_welsh_corgi: 0.0800161400552',\n",
       "  'Pembroke_welsh_corgi: 0.723006093668',\n",
       "  'dogImages/test/045.Cardigan_welsh_corgi\\\\Cardigan_welsh_corgi_03202.jpg'),\n",
       " ('Cavalier_king_charles_spaniel: 0.135604400486',\n",
       "  'Gordon_setter: 0.81196939076',\n",
       "  'dogImages/test/046.Cavalier_king_charles_spaniel\\\\Cavalier_king_charles_spaniel_03294.jpg'),\n",
       " ('Chihuahua: 0.104299604071',\n",
       "  'Norwegian_buhund: 0.54928533663',\n",
       "  'dogImages/test/048.Chihuahua\\\\Chihuahua_03403.jpg'),\n",
       " ('Doberman_pinscher: 0.195383851623',\n",
       "  'Beauceron: 0.509023196167',\n",
       "  'dogImages/test/059.Doberman_pinscher\\\\Doberman_pinscher_04194.jpg'),\n",
       " ('English_cocker_spaniel: 0.0479021549035',\n",
       "  'Gordon_setter: 0.664858317313',\n",
       "  'dogImages/test/061.English_cocker_spaniel\\\\English_cocker_spaniel_04322.jpg'),\n",
       " ('English_cocker_spaniel: 0.11893645782',\n",
       "  'English_springer_spaniel: 0.771950207651',\n",
       "  'dogImages/test/061.English_cocker_spaniel\\\\English_cocker_spaniel_04323.jpg'),\n",
       " ('English_cocker_spaniel: 0.235211091174',\n",
       "  'Cocker_spaniel: 0.760298713628',\n",
       "  'dogImages/test/061.English_cocker_spaniel\\\\English_cocker_spaniel_04354.jpg'),\n",
       " ('English_springer_spaniel: 0.0103569810697',\n",
       "  'Belgian_malinois: 0.737163090871',\n",
       "  'dogImages/test/063.English_springer_spaniel\\\\English_springer_spaniel_04486.jpg'),\n",
       " ('English_toy_spaniel: 0.0772671297853',\n",
       "  'Cavalier_king_charles_spaniel: 0.921342336469',\n",
       "  'dogImages/test/064.English_toy_spaniel\\\\English_toy_spaniel_04513.jpg'),\n",
       " ('English_toy_spaniel: 0.495178226019',\n",
       "  'Cavalier_king_charles_spaniel: 0.497961629182',\n",
       "  'dogImages/test/064.English_toy_spaniel\\\\English_toy_spaniel_04532.jpg'),\n",
       " ('Field_spaniel: 0.208788413606',\n",
       "  'English_cocker_spaniel: 0.66440030891',\n",
       "  'dogImages/test/066.Field_spaniel\\\\Field_spaniel_04616.jpg'),\n",
       " ('Finnish_spitz: 0.265328018719',\n",
       "  'Icelandic_sheepdog: 0.679710914691',\n",
       "  'dogImages/test/067.Finnish_spitz\\\\Finnish_spitz_04661.jpg'),\n",
       " ('Flat-coated_retriever: 0.20809178034',\n",
       "  'Newfoundland: 0.469552640966',\n",
       "  'dogImages/test/068.Flat-coated_retriever\\\\Flat-coated_retriever_04745.jpg'),\n",
       " ('German_shepherd_dog: 0.205309801701',\n",
       "  'Norwegian_elkhound: 0.667916487488',\n",
       "  'dogImages/test/071.German_shepherd_dog\\\\German_shepherd_dog_04960.jpg'),\n",
       " ('German_shorthaired_pointer: 0.324851795542',\n",
       "  'Pointer: 0.59252109628',\n",
       "  'dogImages/test/072.German_shorthaired_pointer\\\\German_shorthaired_pointer_04971.jpg'),\n",
       " ('German_shorthaired_pointer: 0.197621030101',\n",
       "  'Pointer: 0.719650773538',\n",
       "  'dogImages/test/072.German_shorthaired_pointer\\\\German_shorthaired_pointer_04981.jpg'),\n",
       " ('German_wirehaired_pointer: 0.0938460723068',\n",
       "  'Wirehaired_pointing_griffon: 0.774244668273',\n",
       "  'dogImages/test/073.German_wirehaired_pointer\\\\German_wirehaired_pointer_05043.jpg'),\n",
       " ('Great_dane: 0.184377800156',\n",
       "  'Anatolian_shepherd_dog: 0.464541401507',\n",
       "  'dogImages/test/078.Great_dane\\\\Great_dane_05326.jpg'),\n",
       " ('Great_pyrenees: 0.211387050752',\n",
       "  'Kuvasz: 0.325479442344',\n",
       "  'dogImages/test/079.Great_pyrenees\\\\Great_pyrenees_05424.jpg'),\n",
       " ('Icelandic_sheepdog: 0.0537254184326',\n",
       "  'Belgian_malinois: 0.466034144474',\n",
       "  'dogImages/test/084.Icelandic_sheepdog\\\\Icelandic_sheepdog_05733.jpg'),\n",
       " ('Icelandic_sheepdog: 0.1622289858',\n",
       "  'Cardigan_welsh_corgi: 0.595246114665',\n",
       "  'dogImages/test/084.Icelandic_sheepdog\\\\Icelandic_sheepdog_05746.jpg'),\n",
       " ('Irish_red_and_white_setter: 0.273842169533',\n",
       "  'Brittany: 0.547208773244',\n",
       "  'dogImages/test/085.Irish_red_and_white_setter\\\\Irish_red_and_white_setter_05766.jpg'),\n",
       " ('Irish_setter: 0.25493700914',\n",
       "  'Golden_retriever: 0.410134021827',\n",
       "  'dogImages/test/086.Irish_setter\\\\Irish_setter_05816.jpg'),\n",
       " ('Kuvasz: 0.3787450949',\n",
       "  'Great_pyrenees: 0.61878208785',\n",
       "  'dogImages/test/095.Kuvasz\\\\Kuvasz_06409.jpg'),\n",
       " ('Labrador_retriever: 0.116727667123',\n",
       "  'German_shorthaired_pointer: 0.860924628874',\n",
       "  'dogImages/test/096.Labrador_retriever\\\\Labrador_retriever_06465.jpg'),\n",
       " ('Lakeland_terrier: 0.087537272383',\n",
       "  'Airedale_terrier: 0.893322198031',\n",
       "  'dogImages/test/097.Lakeland_terrier\\\\Lakeland_terrier_06558.jpg'),\n",
       " ('Lhasa_apso: 0.275518269265',\n",
       "  'Havanese: 0.477407136104',\n",
       "  'dogImages/test/099.Lhasa_apso\\\\Lhasa_apso_06620.jpg'),\n",
       " ('Lowchen: 0.0423980039246',\n",
       "  'Miniature_schnauzer: 0.823359356986',\n",
       "  'dogImages/test/100.Lowchen\\\\Lowchen_06685.jpg'),\n",
       " ('Lowchen: 0.232889006639',\n",
       "  'Lhasa_apso: 0.325140198279',\n",
       "  'dogImages/test/100.Lowchen\\\\Lowchen_06694.jpg'),\n",
       " ('Lowchen: 0.362578515584',\n",
       "  'Havanese: 0.362647805777',\n",
       "  'dogImages/test/100.Lowchen\\\\Lowchen_06699.jpg'),\n",
       " ('Mastiff: 0.199180126203',\n",
       "  'Bullmastiff: 0.705845062931',\n",
       "  'dogImages/test/103.Mastiff\\\\Mastiff_06827.jpg'),\n",
       " ('Mastiff: 0.309837102373',\n",
       "  'Bullmastiff: 0.56120170984',\n",
       "  'dogImages/test/103.Mastiff\\\\Mastiff_06847.jpg'),\n",
       " ('Norwegian_buhund: 0.0010708936197',\n",
       "  'Canaan_dog: 0.775444792377',\n",
       "  'dogImages/test/108.Norwegian_buhund\\\\Norwegian_buhund_07120.jpg'),\n",
       " ('Old_english_sheepdog: 0.365514464055',\n",
       "  'Bearded_collie: 0.592918662561',\n",
       "  'dogImages/test/113.Old_english_sheepdog\\\\Old_english_sheepdog_07390.jpg'),\n",
       " ('Old_english_sheepdog: 0.378392549025',\n",
       "  'Bearded_collie: 0.592285710904',\n",
       "  'dogImages/test/113.Old_english_sheepdog\\\\Old_english_sheepdog_07391.jpg'),\n",
       " ('Pointer: 0.00494844107713',\n",
       "  'German_shorthaired_pointer: 0.936935348643',\n",
       "  'dogImages/test/122.Pointer\\\\Pointer_07812.jpg'),\n",
       " ('Pointer: 0.0940788569531',\n",
       "  'German_shorthaired_pointer: 0.818107194371',\n",
       "  'dogImages/test/122.Pointer\\\\Pointer_07821.jpg'),\n",
       " ('Pomeranian: 0.0369520934606',\n",
       "  'Japanese_chin: 0.92184251878',\n",
       "  'dogImages/test/123.Pomeranian\\\\Pomeranian_07857.jpg'),\n",
       " ('Portuguese_water_dog: 0.196315182585',\n",
       "  'Poodle: 0.557517687687',\n",
       "  'dogImages/test/125.Portuguese_water_dog\\\\Portuguese_water_dog_07971.jpg'),\n",
       " ('Silky_terrier: 0.328433410342',\n",
       "  'Yorkshire_terrier: 0.635486269991',\n",
       "  'dogImages/test/127.Silky_terrier\\\\Silky_terrier_08055.jpg'),\n",
       " ('Silky_terrier: 0.160753734478',\n",
       "  'Yorkshire_terrier: 0.640046057188',\n",
       "  'dogImages/test/127.Silky_terrier\\\\Silky_terrier_08074.jpg'),\n",
       " ('Silky_terrier: 0.488844433958',\n",
       "  'Yorkshire_terrier: 0.501412550466',\n",
       "  'dogImages/test/127.Silky_terrier\\\\Silky_terrier_08084.jpg'),\n",
       " ('Smooth_fox_terrier: 0.113431369666',\n",
       "  'Lakeland_terrier: 0.512756992131',\n",
       "  'dogImages/test/128.Smooth_fox_terrier\\\\Smooth_fox_terrier_08116.jpg'),\n",
       " ('Yorkshire_terrier: 0.255900710349',\n",
       "  'Silky_terrier: 0.669482510123',\n",
       "  'dogImages/test/133.Yorkshire_terrier\\\\Yorkshire_terrier_08326.jpg')]"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "wrong_images"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Visualizing Model Classification"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Using the process described in <a href=https://github.com/alexisbcook/ResNetCAM-keras>this</a> project, Iwas able to visualize what regions of the image produce the highest activations for a given dog breed. Essentially, this shows what the model is picking up on in an image when it looks for a given dog breed."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Feature Map"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "def path_to_tensor(img_path):\n",
    "    # loads RGB image as PIL.Image.Image type\n",
    "    img = image.load_img(img_path, target_size=(350, 350))\n",
    "    # convert PIL.Image.Image type to 3D tensor with shape (224, 224, 3)\n",
    "    x = image.img_to_array(img)\n",
    "    # convert 3D tensor to 4D tensor with shape (1, 224, 224, 3) and return 4D tensor\n",
    "    return np.expand_dims(x, axis=0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "test_image = path_to_tensor('wrong_image.jpg')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "conv_features = inception.predict(path_to_tensor('wrong_image.jpg')/255)\n",
    "prediction = inception_model.predict(conv_features)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'Lhasa_apso'"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dog_names[np.argmax(prediction)]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 206,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "98"
      ]
     },
     "execution_count": 206,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.argmax(prediction)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 207,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Get the weights from the node associated with the predicted class\n",
    "weights = inception_model.get_weights()[4][:, 98]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 208,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Get the \n",
    "feature_map = conv_features[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 215,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Scales the features to the original image size then takes a \n",
    "# weighted sum using the weights associated with each conv layer output for the predicted class node.\n",
    "\n",
    "import scipy\n",
    "# make into a function\n",
    "fmaps= np.zeros((350,350), dtype=np.float32)\n",
    "for i in range(len(weights)):\n",
    "    fmap = scipy.ndimage.zoom(feature_map[:,:, i], 38.9)*weights[i]\n",
    "    fmaps += fmap"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Below the feature map for the predicted breed is shown with the original image shown below that. The model appears to focus in on the ears for this breed."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 221,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x2a48592ecf8>"
      ]
     },
     "execution_count": 221,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAQsAAAD8CAYAAABgtYFHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4wLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvpW3flQAAIABJREFUeJzsvX2sdctZH/abd62z1z7nvcauw0cc\n28g0OFUJUkx6hZGQKgppCyiSGylQqMRXrdy0AqWRogrDH4V8IDlVAqIiQr0RFDtKYiwShIWcUEKC\nUKTwYbskGBxaB1y4tYuBGufee85e+6x1pn/MPGueedYzH2vvtc+778t5pHPWrJlZs2bNnvmt3/PM\nM7OMtRYP8iAP8iAlefSkK/AgD/Igrwx5AIsHeZAHqZIHsHiQB3mQKnkAiwd5kAepkgeweJAHeZAq\neQCLB3mQB6mSk4GFMeYrjTG/boz5qDHmHae6z4M8yIPcj5hT+FkYYxoA/yeA/xzACwB+CcDXW2t/\nbfWbPciDPMi9yKmYxRcD+Ki19jestXsA7wHwthPd60Ee5EHuQdoTlft6AL/Nzl8A8NZUZtN9psXj\nN4WIFNmR8UvPa8WsfF5zj2PSlt4vV9/ae6/RBoeIXRjm51qeXB/R0o4l4of8fof8Xlr8Jz74e9ba\nzyrcLSmnAgutqlEzG2OeA/AcAODqc4Gv+ECce1BKkHGl81RcSWSrrH2eisvFL72mzaQfkrbGM+fi\nNSn9nqWwjCvlyZWZi1siNX1hyfmSdv4b5v/O1Kwop1JDXgDwRnb+BgAf5xmstc9ba5+11j6LTgG7\nNRo1FZeTteFzDaBoM+VobbBkkNemrSXHDrZjRGuX+3jm+5YTPdOpwOKXALzZGPN5xpgNgK8D8L7F\npdQ89NPyY68BaqUylgDHK0FOXf81Xj5PkZzk0a21gzHm2wD8FIAGwA9ba3+1WBPtrUM1HBLn8tpU\nuoxL3edYWTpg7yNuKQNbS+XQZGle+Zvl+klN3lKcTF9Sh/uQJwhWJ7u1tfb9AN6/6KLcoK75EWvL\nl/c4hpof24KnBIWaPKd+vqdVju1/r0A5z+qXBrXGImoYSOoeh9atlHasoWotG81aQHPMdaeSWqZQ\nG0dhJNLkfZBIS+VdKk/a5nY/RS8Qg3lN5A+nxdX+0PL6Gll7ENTS/LUBZkncoflPKTVqAB/8wDJw\nSF0j71NivU9C7vm+57s2pMVxA+eYDr9GvlTdap9BPr92LtNKcVo5pfxanUvxKbmvt2vuuqVqV237\nHSpLynrCr/ZXHrNInZfSSA554qV0/D7DNeel+DXzrAUsqetr2eUhrEILy/JS9VlL1mKvJ5DzAAtN\nUgM9Z4Mo2SfObY6fh9cGjFTcIXlq895HbyqpCjnjeAlQSmEo5ef6XUnuk6GscK/zAAuNWQB5tpDL\nq10H5dpDjJ+HqEK14JA61t4nda7JWr/8Gp00l2fpzJjGGlLxKTAosQxZr9p+lcqfij+EwS3Ns0DO\nEyyOUUFyPzCPk/ElWTJAa4BAxi29pqZOpfhTylpvuprfrrYvpF4mspwls22yXjKtVg7tXzXnK8l5\ngIWUHHAsoYu5t8CxMyM1ebXBnovX0mqBI1Xefcsp6XCNKrBGuNTXZB0O6VOy/ofKPf7m59C9nCxV\nQ5YYpUqs5Zg6HqJi1ALC2gwjJ6e05xzCIrjk3tzH/J50fanvldThY6bpS31Knp8SjJ9c0SsIH/Dy\n/D6ZBb+u9pzH5ZhELm8N+JTunZIa6nwsgNT0rkPylH7PNViE7Ec5Q3uun8l4LT0nZ6B+3FPxlZIy\ncB4qhxg7a6UG9ZccS4zjEHWkRnID8JRyyCApAdsSQ/gSqS1nLdvYksG/5GW1kpw3WByqhtTSR4rL\nyRKaeAw4LAWSXLhGaqYAtTjtWk3WYhVL8t2XlGblkEhfwjBqVZF7ZBfn8TMQWKTeIGsYr7Rzfo9a\nKQFFKn/uxz4ETHLhQ6TEwA6dRZKyBojU9JPUeW34WPvYWqqHPH+CI/Y8wIKkhl1ocipmUYP6uUFc\nGvyltFrGkqurlJrZJe18abklOTTfEhUj9wyl51vy/PfNYEtpubgj5DzAIuVnAeQ7NX8DpN4iazKL\n0ltdO6biDgUS7SjDa0iN3YfLGozh0DJqZr0OMWbW9q8Ss0nVOyfHAMWJRvV5gIWUJT96CsHlD5nK\nf2i9DgGLFBAsZR2p+5ak1LG1gSLDx0qurof+NrnBWTNjVgMehzCttdnFEwKJeyp+gaTeGKnBXgMa\nx1DD2h9uCVDwMI/bJtJT56m6lGTIhFNv0aV6eU5OARRaGSk2mWMWOWDMAcaaQMrLzMU9MAtF5A+x\n5G2Xu5bH1dYjF65lFPI8VdapwEKTkmU/Bc7H3jd17bHPIsuofeGk8vO4nBG41FZavZBIy8U/QXZx\nHmBRs0S9ZLNAIczLWiJLgCIVVwKLFo5dlK5J3bdWBnaUbaOxixqWsVROyS5SAzl1nupLqTjtBaSV\nVZIcQNfE1wDECUb2eYCFJrWsojYMLOvgpR9nKavIxaXUkLXB4hgpvZGXyimYBUmpfrX1TzEKng6R\np8YwvOQZH5iFEM4slhipaoxX8vpaybEIGacNdHmeilt6DQC0/ntN7QL0G1rWHv4bULKNcuwi1661\ncszbtEZSagiPKzHRFOvK2SqOMYSWpIY1nBJ4T1fcCpL6oQ8BhkPVjxSrKL3pj2URyWsZOLTjVJ1H\nLAwADTsfhyZKuxtYXkob/M0GM+/8dKxR8w6RmkFQiudSevtreWSaPGoAII+ynjnQOQZIniCjWOUW\nxpiPAXgRwAhgsNY+a4x5LYAfBfAmAB8D8LXW2k8dV02kAaP2mtp75MLHgoUGHNl8c5AggGhmQFH/\noHezGKpQ5qObNUCxxIC8BCxK6ccCV84Iekg5mu0jdZ+U1ALnPbGKtYr8z6y1v8fO3wHgZ6y17zTG\nvMOff/vBtTvGbkHnJKl4La4EECWwWAIcEYDoANFMIOEeomVgIYHDyT5iFwMLj0Pr0vx1d8Po2EaL\noK6k2IYM8/OlHXcpYOTAIiU19ZZ5aoy+JSZWwyxkuoxPyVrttlBOQV7eBuDLfPhdAH4Wh4LFoZJi\nFEvQWgMDnnYIowAKaojFZIdoxyxINEWwqJNxaPCoHQ9jG0DdW/jYtyQQWFaqnOFUn21PSM6uo+XT\nwiQ1o3AJQJxiVK9QrAXwvxtjLID/1Vr7PIDPsdZ+AgCstZ8wxnz2sZU8SFKAwdNz4RyjWAUY6M+q\nLKJphwgYInBofHxBxxqb8GBjG5hFMzQYvQ2DGMfo7RfERiYbR8q+QVLSfpIdWhn8miq1BAiFncbF\niQpwUKlhFvK4hFVo6kjOCJuSQ1jDCQDj2CK/1Fr7cQ8IP22M+be1FxpjngPwHADgNZ97ZDWgg0Pt\nG0+CxRIWwe0ORRsE/QWA4OAAOObQcJAQwNCCgQbyA2lk6UPD1JCmncCjmcCiEjxcYfObyUGZs6EI\nAJCGWuB4thRkDwC443ESAAFEM0QlsNDiUgAij4cyi5QcyupOcKusWGs/7o+fNMb8OIAvBvA7xpjX\neVbxOgCfTFz7PIDnAcC84dkMxzyRlGjdoSyiGHebZBESJCRANAuAoijKeK+VucqCagZQAwxLjLWa\nNO0wgZ0mU/3bkQEfU7tKrJQkp4ItNZLWGuzXyHOgHFy0MeYxgEfW2hd9+L8A8NcAvA/ANwF4pz/+\nxMG1SxmkjpEalaMGILQ/Yhb8KNSMi+1+xiIkg+Dg0GCucrQLgGLwqDCyB2/QTKyDGAepLCXG4eKW\ndxsNANoZSBzHMDRGkTPy8vSs2gXMWUUuLFlGSiWBcn6snCNYAPgcAD9ujKFy/oG19p8aY34JwHuN\nMW8H8FsAvub4aq4kKaDQ1BGeVg0M8k9XNzhIAJhYRBuBw5xZUPxJJcE4mnacBlbbjtHAK4kEBSov\ne94sf066ZhyZfYbVW3sGYiFTWjs6lWViHYxxrC0a+zh0qv8e5ODbWWt/A8CfUuJ/H8BXHFMptcGO\nYRk1bIKHaxhEMj4GiM22d0ULFqExCBkXH+MHTrGLQYx2YhQNix8xzhhHSN9jRAM0rixuJEUXBuKG\nylJAI8UGJADUPlMNQI78ubl5wj+Ly9POnsHVP0wzD0MjWMfegQdnHJJtlGwaqePa9osTy3lUzSI9\n+GuAIxWnMQV5TIGFxhh43CzdAtt+xiA2W0eHN40HDYzYeIp8CJsoDZxmBhbMyMkAYgMaYPtpoOkA\n4uk8ldsIQCoQDAkAsv7z8/RbQAMTCY4k9AwTqPFnbdi1TQwiTQI8VHWFqyqHgoYMa+dnIucBFktk\nrYasUTlS8Zq6wYCi82xis93PVAzOIHIgwe0VdF4jzhpxhPVyJdHqWwIGDQzK4Mhc3NlzU1kEJg0G\njGjz7dM49jOOjaqycFtNbORtscgwykWqIiWjaUnO1GZxWqllD6n4EqugcEnN0MIKm9hs+0nV2HT7\nCRCIRXRwAJJSOVJqR+nNDEDt/PyNS2/ZcRo4xCBi1pFUSZQyayVX/xpWUWvM1epGDMqF5yyJnnnG\nsrgKpjCOItuonYblRwprqgnJEx6t5wMWOdRcAhKa1NonssDA4+YgsdnuJ1Wjw34CBB7m7AKYM4rU\nQCo5X5HIGY8QH79led4YQHRw0Cn9XFIsIDWTU1JJcmVKkXXi4MDTZRuM0zPqADKBR+Pix7aZbB0E\nHEBwoc+qKe4G6WOrhCHyPUE5D7DIeVkcCxQpSdktZNwMOJwBk1SOlhktacBvEECDh4EaZnHYFCnl\nl7Sby6EqCi9X1rF0Hb+3rEs6bfmPy68hdcOFqT3cs8s20NSTuB1F/iZuV66ucJmpKS1wkEt6amr1\nCYDHeYCFlENYBpfU7McSUFDDzqHqYrtHt+0jlYOAgVQProakmUV69kOGc8I78yY6Lxkv0wZQWfYh\nrCIHAocYcCWISDCUdZZh+XwTcxDtNGMWLH7K3zRZtsFnVJJswxU8P6YYhjZrco+gcV5gUXrwNRtG\nU0ly+VqL4C8xTMbLDv3EJkjlIPvEJsEsuP2Cjjm1QzIMqZ9T/gAEaQYxZwnzvDmGUm9oXQ4MS1lV\nql14e3CGQc9Kz8efjYDB5UkzC60OBBrTc7QjxmHuizI3igKASasdOcNnyZnrBCByPmBxCnUjZdSU\nxyKziBkF2SdajLjE9cQgAnBIZiFnQ+qYRUo24jzHJFy6bp/gcZp9gpc7v1cseTXjEOA4zPksxSpS\nz6mxBxnveCFnFnHcZO8RbEMzigIZtgGkGYecacnNvCx1Na+U8wELkrVsEVr8EkMnUz0ebfdo2hFX\nz1w71aPpcYWbSP1wx6CSEIPgAKFNoboqHOahKQeu9KMIedLqCFczZF5ZbklSMx8lw23p+XNtooFX\nSfWQR55HgoeMV0GiAjhyRtHI8asFqrZABPKAcQI5D7DIOWVJkflKTyDTczYL9S84WZEhs4uYQz91\nJa6SBGOnxizmMyBLbBXcaCfPOc3WJGXgk3ly99HypuJKDOPYGSBp2JT1DeFY9eDtINuNwCGlgnBV\nJVb94jh2QXxaZRSFBw2y/idAQ7NdnGhUnwdYpKSmv6QoVwoU5LmmhrRw06PtMFM9iEkQs3BHBxRc\nJZkzi7kqciizSPk+8LcnZxZl1STtT5EyIkopqRq1jGIps3B11KZN4/BS9UMzbKZYSIpt9OgiQ+nQ\nNNGCvZyaArAdzFxjIdrFTEqJZazAQM4TLE7lpUlxVUAyeFYhGUVgFtyHYoO92n04y5DAAaQHSc7y\nr70VXdXnhjueJ2esy5UJpFWRGqerHKNYwiz4vVJTuSmjZsq4mTNqUlrP6tcwIJHPLZlFhz4qk+5E\nbubkLarJbAezoXF9MsU2XAPF53HjHS3nARZL1JClUsMoZgZOO9kpuu0elx0xhn46EqO4wvV01GwW\nqalTDTBkmEvu7Zla80H5tLeqnAotMQtpVOWyjFXMjbupvLLsnGjTvTwsGQGl5ewTFLfBHnvfAtLY\nyfPt0eWZBWMffIsA2oyIbwugbUIUr4YFpo6dYhsry3mARU6W2ih4Ho1B8POkx+YwrfHYdPuIURBQ\nkFGTwOIS1zOQiG0XsXcnMKfjS+h2brajZMzkHVxL11ykS5JjFKWVtPXqSNnPIgd+ORVEAohUTYgl\npNQPeo4YONwz9BPQtBPozICjAfqmc/kSwEG2DnXzIWoLue3hinJeYLHmTEgunJsZAWL1w/+s8n0S\nVJBRxAc2kZ46nS8k4yLPNUMjD0tqXat2pK7neVLXp+qaAoClACLL0ETzs5A+JyUVhMqRhkmpmlBc\nj2AU3mMT5dsACGpKP92jYzaheEo6lunZfTJfyObO22j5/x3AbBpCRTkB0zgPsFg6G5KrdWr2o0YF\niYya+4lVkKGSgMAdw3QpGTcJQIhtcEetYw2csoPVTAlSvhrmoakvS0RXOdKzPku8V2tUEU0NkepI\nimFoBk4et8eGDXYOAnm2QSyiRzeVzeOozYl5BFvJGKknAKZNlmebDtFmPVwiu8Z6ch5gsYbkZkSk\nCiLjorDGKkiVCGxCTpdyFiHzxMbQZQZOOR3o4tJvS8kq+JsvlUdjFjWsIlXvGrVjnpZWTbT78LaQ\n+TVQaPx/d655sQZDLqkQnDFw1aXFOAEAf+69iKPruKFz41kGj8syqAaREVTuOjYZQrk9g7YwPGD7\nw5y88sGiNONRnCJl534lacc8NIlFSAPnpZ86DUwjMBCNWcxnRdYxcNbYJbTVlWnDX55ZDB5cpCwF\nhhLrIFmymK7WjqNNk/J2ILWB2x/oegIJDqYUJ/00NKYR+3FQKQTizQRQEWAxlqH5Z3CZzZ6sKOcB\nFik1pFS7ma1BCcu45JZ4t9Nmuk0b2yTiKdMRV7iZAcMVric/C+5vwReXAZh5dQLLBkdpQKRBIE/F\nS67iJHJWpGSklGpHCjDm+Y9XQ3g4B57aAKd2kODBmQEHDqonL4fSneqxn6UHUAggEXw14K91oNG7\nG7jp1gRozFgGyYL9UnNyHmCRkiU+7ppBszQbMv0F/a4VO23HNoZYNdmIgd+Ko1RFuFcnle+qEgaJ\ni58DBnVgSsupIC5fw8J59UJTXfg9AN3PIjeLUcMe5uBRniVJSWr1LH8mqYakvC9TKgcNWlI5SJUI\n6sa+Kj2Wfqa+0H0HxKoiPROAyNApZ0seSTvGEd9giev0SpOScTPFNlJqSAugHfCIeWpyAyZXQ2JD\nZ/DmJNWEqyTENuJl6zFoyIHjwsunT1NTgpQmmUfK6FkznSpFYwCSKaTsGFraUtuFpi4d4q2Z8rNw\nxt/9FEcGTjfwXauQmsKZh6aS9JNq0UTXknCW4Z41ZhzAxoFXMweXqE08YACYGz+PkFceWJBo7EFL\n0/5maWO8NT9S7CDNIlrWvQhQNO2VqyS5KVSNenPvRJcnvCkBiLdlM+u0dI/IKQjNVG4oL2YtdG9N\nclOhKWaRAonStKp+/zl48Lfx3HA7f3ayHwCIvDVpcHIbwjRg0bEp0R6kPtD17jcOeQFEqks3AdDI\nWEP6949Z0nw9D/88A197MmMZR0gRLIwxPwzgzwL4pLX2C33cawH8KIA3AfgYgK+11n7KuI+IfD+A\nrwZwDeCbrbUfWqWmWq1rp0VTzMLvfMVtFZsmXmZOjGAzMYZhCnM2cYnrmVdnxwyhtJ+FZuhsWIcB\nyuwC0AyU8zcjpddMFWqMJLVMXYoEuTkIzIFDsqyQf4jKrGVcvA1y7UJxmjMWZwZObQhTozS4iVlw\nr87eMwuuwpEPBtV9ziy6KT8f/ARM7n5zoCHA8gWHZx+amT8GBwwg8SW5BVLDLH4EwA8AeDeLeweA\nn7HWvtMY8w5//u0AvgrAm/3fWwH8oD+uI7mZj1rgkHtrbns07TDNgAS/iiFSQ/h6EL4sPagmsUri\nDJ3BECp30dJYBwC0o6Ddg3hzsg8cT18Ty6oh8Z1SgyRn8OTlzZs/NQUaA4TGpniNKJ+qivg2kW0h\nZUnbjKIduJel1lbSZ2KPzQQcVN8+AoghClM+6VMRfDCC6sHrQfXlNhCaRyGjZ42UZlFqpAgW1tqf\nM8a8SUS/DcCX+fC7APwsHFi8DcC7rbUWwM8bY15D3z09uqY1ohk5tfTpaPHI76PpoudqCMWn4nKq\nScrnQlVPxIBoBv09wONJbRr9ke/YJBc8adNzmoGPW+ypQ4c8c8OoDOdYBCD3Ii2DZjOMs7ZolMmR\nsQ1tM7aPpnbMtc0AcoJqp3rR21yqHNI/wg35DVMhGlZGG4UpT7wSmINKMERTu8l2HZXfooVTYppG\nX/IetRlflHagHGqz+BwCAP8B5M/28a8H8Nss3ws+7niwWDJNmjNo8h26yflq2kszzHBwNSOoIsGA\nyVUSbuDkqou2LyffJKcdx2kw0ABoR8DwwZCaNWwBIpa2dcehucXo22RsH2Fsm7ABiwJfACDXKvCO\nLA2FOckBRErtkmDJgaEZFrQFgIuobahdXBu5ceTahtrFtZfeNnNg2M9qv8eGveljZsFVEm5jcKrJ\n3DGO7CKurQkoiIEEICF3cj6jMqltbQwW9GlGAAd9m1aTtQ2cmkO66ndqjHkOwHMAgGc+Vy8tVbsc\ngyhNl0bHceatyQ2VAJ8+HWbhVrCNmGLHg2K2ma8Aimlg0IDgjFtbTEfpTRhQFy3YosQ7V3Y7YmzD\nPpHyrSrXKmjMgndm/efQVZDwRt3PmEUU13u2Mdyh8yNhAomBPWutj9EYl3HROuBoR2AY7jB6MCm1\nDWdikm0AZEyNnbkozMHExaVBgMIESkNUbmCyBNr8/pMRVyx3X/o92ho5FCx+h9QLY8zrAHzSx78A\n4I0s3xsAfFwrwFr7PIDnAcB89rPLHdlTzEKeZ2wVpIJwb01iBNw+EYddOp8O3fi0FoF9NIhXn07X\njj02u9s5QIz+KMEiN0D4r9e4cwM3OC58P7HbOwzNHcb2FmP7CADQd5sk25DMgjpoagFUrIroBssU\nk9jsbgEgAMROPHcKKLJsS4R9lY3vBxfwbdM6RjY0d5CsQzKOHhuVbZBHJqXtGbMgg2Q0oCcoamcA\nzA2cwZ4U4sKjN9N0K589IXVkOvcGT84wbhPNViuHgsX7AHwTgHf640+w+G8zxrwHzrD56VXsFSlQ\nSKXl1JAWk61i4/0q4oEfGyxlmKspcmEZV11IHYn8MfwA6XoPEsQnaZBwsOADBdDZBUnD4nh7NIDp\nGXi0d7AtsNntsN/m1RRgPrPiqhEDRm7KlNsnpiX6/X7Opvqp8DicA4ySCLBQ26Z1bQMAFx0A3OG2\nc8AKAPutU242TZjt4P4R9JzkZ0H9gvJxVYNAI7YDBRvEnjGOYJQNbIS4Do/n06thijuePl2TYRTB\nwhjzD+GMmZ9pjHkBwHfBgcR7jTFvB/BbAL7GZ38/3LTpR+GmTr9llVqmRKoVFJZ/UVqwVZC3ZuoN\nCEhHIa6mDOp1lB4fR6Z2MDbBB8PO15GzDGA+QEa4AcDjW5bGd03i6oqPNwPQbgHsHB0nKk6OPg2C\nxZ5EM27ytPg4zOIIKGdqF39mya4kSPDnLc2i8vYptY3wEr5AUOU2O8fGOvSRgdSxid4XG/wsgHid\niBzUMZD2CIzOgQwBcYc+au/Qz+aGUK6akDoCxLMfcvHZoVIzG/L1iaSvUPJaAN96bKUi0QBBi682\ncgZvzZStQk6dxuGghgSVJVY/gjril7Az1ePiZbi3Jx8ckllIlgGk36oaUDYsvvPlt/6+DWBGxzQu\nGBVvtuPENLiRbpjeZnPKDMQDwVVBmQFhbOKCmAO1QY5hSdAsAYbEM62PNIlz+tsFxhHY2B36zgNH\n64CDfCFILSBDJmcWFN743kD9SzIL93gBOMJ56JW83QOb405zYXYEQOQOvsa0KW/GV5bkjJpa3ghQ\nmI7dxiChTY3KcBMNBu3NIQn9MDGKdkQYAMQmJLOQNByIB40UZuic3pL01mxYngEOMKmcrYs31Dy7\nO+y3bqqy79wzaWsWpPA2ittAN+JGg18+a007AOm2kGyLs7CWHXk5kmHIc99GHYJxtGlH98FkhEVf\nBABUbc4sXNsQE3GdkTML8vYkO4h0bAOcSxzvZ1wdJNZBgIUGkbEztZv4UjlfsNBqdqhRcwrHtgpS\nQTg7oONGxMXsgS9N7xHbN+K4rt9js7tzb1RiERSWx9QAovNSW6Xemlsf7hEGDSvXtE5nb0c3zdgM\nO4ztI6fvspkCEkmR42NipoeDgARLjW1IhiXbodbwW2obzsAojcLbkDaxjc6xMW73Ieq/965aNBsS\nBr9cDOaG9p6pewEIiCnQTEy8kMDlCT4dVF7J2NmsoIqcD1gI3bFKalWQCTD6yFZBAzuoHkEF4WEO\nEny2hANN8PYMsybOoOmBokcACqLbLyMGidQgKakjqUFAYc4m6JwPOh9nOq+zb+O3KHdsAnQ1RDpR\nzWZ7cqpXr8QdAxa8TXhYAgQ/JzDbYlLXJnDdYQ4cHXDROxVl41WUpgtqRtyH4tW8wzTow0M4VUYH\niLAOJaS7x5rbRejek7PWuA5Q8GZ8ZYnWESrzcL+KKY6huvypNDVFqixztcSfD2Og3rzz87eqxixq\n9fWGpUuaLSm3BAkex8tq3QC/mPIpfn+M0XL39CRQjNAHfw4cUjMk/MjbQtaP8vB2ke2qqSA7FpYG\nUDqye5ohqCjAflJPXDZ3JLdwmmalAd9EPWlk6z8G0ZfI82IIqkZCSEWRgLGGvDLAQqob/MjTs3/B\ntZtsFdI/IjhQpcP0Q/LpUc46JqYy9tjs7oLvAGcWxCgoTaomYGmArrdzkdOD0mhHHX7HziXAcJuG\nj6OZAfROPRnbu8iVerp9yuuSP4tkTqlnlmCpMaxcW8g2oXbhR9k+Mo7UEK6OENtgDIPUOzIYt+Pd\npMLFLONqGuQ3CPaGPlJDuBqhGzfJ9kHqCJ9B4S7ibVRWrJIcI68MsEhJrfrR2kkFoQ8GxUbINBjE\n4fwMybS4zM9+TOoHt1XsEOvqHES0ty5QBgsgrYN3CIDA1ZKR5ZEqCQMN4/O0W0wekOpP4eulOpnR\nMwIxQHCw1FQSnl8Di1x7UJtMFUQMDDyO2zO2mM0eoUdoK273EeqK6YFtB6C9Q/N4h3Y7TrNLfJEZ\nX5PjzgNb4OpE7PNCBsyYlYSU2uUZAAAgAElEQVQNd7gdKSxA4zaQY+W8wIJTvkNsGFImsPA/hHDt\nBmJ3bndJOlyaIZnUFD/7kdS96ai9Rektm5tC1J6T55ODgvJsRX6eBpYu05hLeeonmblna3WX6og8\nluw2shwu8pw/I1cp+ENIda1BAIqUKrdD3AiyTC/OK/XW2Xy6sMSdBjdnCW4gb6I+RDMkUv0NNgkC\nAud+zqdcpUGVg8Yxcl5gURJNBZHpkl62ANoxqCBNIHjcX4K7e2thOSuSmiFpMGCzu/P0HfGbVLKI\nlBoiBw4NGCANFsCcXnMmIW0avCwNOGQbcuYg218OYq6GcBCUR8ksqB2AOdsA4vag+8k6yOcgkbMh\ndNSMnlzVAGK2wRlGh3n7+vqaHtg+pkV+127xWhM7+fHl8w2GaaBzRjGwI2cZBDRy+r6NGiX4X6wh\n5wcWtYxC5tGuUeIkwnJ1JJznDJfzBWWUBgDtOMZ6+yD++GDhg4EPGHmdHIRStDdozs5DQNWKOC1/\nbT8blXCKUWgsSwMCGbcTaSRau8g3vga49Py8PAIGKRrDkM/I83r1xAzApnEzS8Njsje4IzEMYhGc\ndUy+Kp41SId8IO6r3FEr9sGo+6RDjZwfWCwRySK0OG/YbJhhk095SpqnsQc+BUbqCNknuKGUbBXT\negfJLMheoU2l5phFSQ2h5+a6NRnl6PqGHUmog3PfKw0wOE2nOvBBCBbWACE360HAxZ9Zsg2tDXLs\niIvsG2DPoRk4ySahTZ1yNaVj6T2Le4xYLWyAi5HWnuyAx3ObBYWDCtFGYSDYMFw4fA0t3kELE7DQ\nPhxrynmABX0KYA0bhTxOe1YM0R6bcwOn5k/Rz87JYYvrltPiMr9ITDVkcpCg2ZCXWToPcxYCzI2d\nqWfnU4YSGDQVLjXg6XoeR0ZPKlfWQxpjef35oJdqmQSQGtCUACHZD+9LPYvnIEFHTYWjfFwd4YBA\ncdxvhashLRxocJDeAdvBObztt4+ALrABvr8Fff2MwuTROVdDgtGTryXha0j4Evk15DzAQooEjhog\nqXgSPgviLgngIaeeuLqirXfgasg0yTUkZgO081L8UmahtRkQU2pNvZC0n8fRAOIu02DXNpiDCq+z\njCMpqSMaKORUFfnMMixVLQ54BBD8HhTHjZn8XhoIy+fjRmNiJiOcLWt3h7HtgcZNn3J2sMF+CnND\nJ7etuSqEF507D+oI31VrLXsFf9TzEK5LSp2T0mtBowXk90CA2C1Zt0GEH4W7fcfxfF9Or6b0++DS\nLZmEVD/kW/NlJY4b+0qAkbPuyxkO+SYFYj19QPwG5gOGDyRZj9TATs141IKkNHpCaRdZBymUv2Vh\njWWkpp+3LI0zDLofTaHy9h/g2EXPjv5l4lSSWzTbcboHMYseG1yBpkmHadBznwru2BU8PAPgEKOo\nWduzRM4LLDThjX9Ibb0KAugW4zmjCGhNIkGFdMLoGvI/0AbBmIivYRxAGEBU/hLh7cYBgxvrZNty\nVsDfwNxmAczLlfWWb36tbhJMeB208rRreR3kM3AWIF9Gmi2GD3jKq02nQpQFzAG3x7z9fV7aJqDv\ngjPfJtorQxo1g7FTehzTFCpfUObS12MV9KhPXqTN4lBgyEiwVwTAkD4TnF1I+0WLuSF0+hTA2M9t\nFTn7hGbcTNksSoOCRHZmOfi4odI1hDAEI2YRMo2DmDYAgRj4wMLaHzd6pmZFUqAq71MzO0LPAcRM\nI8UwgMDK5HQqNya34kjx9IKgsrh7/9bvaNYBV+0O4+OgZLTge3iG6VTqge5xuT1jjNaNkPCNmdeS\n8wELIK2GJND5UFCRVmipZnAWwReIhZ2w2BfLaAZEUzWkTwUdNYDQfA5SeromXEWQ9gjqL1QXeusB\nMcWXRlBuHOTlkGiGRa4apFQRzZ4hVZGSg5b0veDPm2qjkrGTjry95MwI3YMMvhxkyajJf3f6PbhG\nwJ5/2wLAS9hvLzA2DiBoN3HArWQF4GPC5spAUE3kwjOCm7DidR2GcR5gAaQHPsVL6nuAtApI1EqS\nYZC3JlA2ao6I/SuWvGVzsyEcIEKF5wZJAopRiec6uBT522i/lbxWvulTxksSCST82hSjkOFcG0nh\nA5mrWhx0eT5eLk2hUhyxDs5YCES4PUjOvvhrms4vwpu2TOgnYNgw5yueNop80heIZljWlPMBCyA9\nEHIMYgWWkbJJhKnR/Qy7ua1i8tZMGTUl43gJAThyzCJFuVNtIJ9/FPEac6MO3bN8FMcHh1RpUuCQ\nAkwpKVtFSv2AiCttQ1gzTsgwyZ+LjtLjk/JyA6eMIxWE2k4aOEm4c1kbFqGN7bVbJYrgUxF/qqGJ\nWMeIZmbs5LuJU29eyzHrPMAi52ehqSClt1jmqWqmkuQ0KcXNnbfEdKns5EuMmtqbV4blswJzI50E\nCH4NXzjG4ziz4G3dYA408t6pNCnyeVPpKdEAQTIRCVg1wp9ZM3Zq9hlZtlThOJMYEYMw2T8IPDr3\nZwBsultgC1w3l+CzceSyHb/UmqiPclucVDvWYhnnARYkOWaxlPpmpMQmpP2CvDeD81Y/7YJ1wRmD\ntFe8DN2OIVmEZBby7ek7lk0MMpN702sDnqstHIQki9NsR6ER43PNDlEatCnVI8dONDDF/Jzaiq+m\nvxV1oZ29qX3pOz1G2jG4gZPPjHBvTtlPqS78Oro/t7e04dptCzTDLfpX30yqB2cW3NjJDaHBkSsw\niMAqmhl4HCrnAxbaG5PieUcdRFgDk9JbKiGxk8t8wxvu4k27YKkqR0r94OcDO2ru3qzz2yHu9FLo\nsQ3/NbkBT6PmxBxkmmxfKkcrT0rqzZuyw0ipBRdZX5ZGbcWBYWBhHn/jjxes3VpSJ3z8BCDc2Cln\nQ3g/5DYQ5l8RGTil5ycLXwzA1fYaYxcvbd/7TfoAZ+wEYt8L6q1hF64wG7LG8nTgXMBCU0NS4CFl\nEOFZOXW6WjxDMsziZdy0AQx1kJT6UPrT1BVABQr5ZgRchx5G36k5AHDg5OpHw+I0Y2JKUoZAea38\nPU4pSltFyQmQkMA7tZ/Px8GDzluwz+2lVB8SAoBe5GkRDJ70m/Cwb9PN7g5NF9gufSog53vB/X6I\nSZD6spYUwcIY88MA/iyAT1prv9DHfTeAvwDgd32277TWvt+nfQeAt8M131+y1v5UsRYEFqkayrdd\nDbMAgMFET0hULv4qhgSCwd92rpKQ92aDcb4Jb45RaHGSRXCQ8cdb39luh7jjT03Ths48jL5D86m/\n8OBOOEDIQS9BBZgb+bj+zfNpsmRmYkk+mdffR4LqMPgjV0Ogh3k5FwBuKMzUk4vWtff0USJqN75A\nj/dDCpP6QeoIEBgJXcenV0dn8LzaXqPphohZXOEaOd8Lvj9nvPL0/lad/giAHwDwbhH/fdbav8Uj\njDFfAODrAPxJAH8MwD8zxvwJa20dvGlvLR5H5xowyDQWPw4txmGcBkPKMsydtPj5/DiE+3HdX7Pw\nl9iGvEa8JTU2MT3uEANGnJi+bgIG7VjqEfwtmUpbW7RyEwyGAwWXW8wBorq6HIQ8aAD+Y8xaexCo\n9uycjJt8PwxgvnoVAB47djG2I7rGOVxxZsE9OZ0NDaAVpnwh2Vq2CpIiWFhrf84Y86bK8t4G4D3W\n2h7AbxpjPgrgiwH8q/xNUP7lSgxD6zwD3K+rSGkqKbZdxNvvtePo3jrUCaTXpbZtnnSy0vwt/B/X\nu4lRJFWQIXTe6NlTakgNGDSYd/6U2jFjc1im3pDwwaKllQADoY04o+AgQdnleeqWk00DmLYNvBgd\n42gb4NIXcDkwI7NmsyDWQJKxWRALofUj/avJOYuYRVh67gybN5OBs/dHAJGdYi0vzmNsFt9mjPlG\nAB8A8FestZ8C8HoAP8/yvODj8lIDFimQ4HGSgQDAYDAODYZhbugpGX5IJ5SbkTTDOB/kOdAg1iEN\nnBJUCkBBA0DTr2+HEJ8VDh50LtUQGa+xuJpp06VM44g+PZmQOMvwfwMOBwzAgcYFQnPcwIHH8HK4\nZ9t60CCwkFPY/EYjdLDgLK8Lxk5SRwgkuDrCZz9c0aSWcBvcOmDx6MDrfhDAHwfwFgCfAPC3fbxR\n8qpfSDfGPGeM+YAx5gPY/25IqGUOqTdN6hovJYDga0Y0mRaNpVQOTR2Rf0DSaUga3yRQaHnuRWRb\nnkrlWMHkPow6GKSA4jbxx3+yWziQuGFp1xTfAzc7d7Saiz9/IYwsrRd5FJXUOf0F701pQ4teYpNq\nMncNWEMO+mmstb9DYWPM3wXwk/70BQBvZFnfAODjiTKeB/A8AJhXPWtnG6GSpFhELk6E74bG2S3G\nBmjKKgigrR9hP4L246aMlbzjjEpeBijcTsH17hpwmGwWsg1L51ykAXOpaCMzJZwFamtQ2kRa7vZS\nLWFHyS5STIMLV0NklVsEtnEJAJ5pEMu4osIlw+DT1cQiaI3Qlh29sXPr1ZGrxzfTkvOeTZ0Gb85g\n6OTnLrzOgrKDmIUx5nXs9M8B+LAPvw/A1xljOmPM5wF4M4BfrC5YgoCMk29nLU5eJyQ0ZsPi0iOE\ng0ZyGoqzhRyLkPXkYOMlZ9CUMjNqhkoHSz2XQ4Ggpk5LmIash7SRyLxaeqvEKaIZNSVo8HPtjzMN\nzix43JSPgJ7btCTD4Oopf3Hs9Ap0fWw7m3tzaksRhJ1tBXZRM3X6DwF8GYDPNMa8AOC7AHyZMeYt\ncCrGxwD8RQCw1v6qMea9AH7NP+q3Vs2E1NgsSHjDtiIu+ddg9H9ogIEBhgQKjsDavhcNhvgrYxrL\n0UBDmyHhTTBA1bs1kbYJAg2T+zXbinDp5TOI/PI8d+9BCWv5+OwSkF4pWuhVNJCpmnKAyzy5aVVi\nFjeIWcalj6PjLYDb0RlBp0dqvG4ufTP41CnZKmh6lZgFALzsP8zc79F1jkWEFanBUUu+BMOMSOvP\n7wEsrLVfr0T/UCb/9wD4nkW14GBR6oA5NYSv8uNf1xpa3A3uS9JjFzzd+EYhowhXiVRFJGAAMXDw\nuuZYkZC2QeQ4BIR5f54HQMwqWszfwNrCqZxoAMHLyQ18ni7PpQ8HHaVqksqnqCdtmwfaW3HUmIUm\nPJ7yXiA0jdpEvWMZl1vgioMbZSTVg9QQsCPfHczPnGx2d7jsrgE4AydAaojLKD+ZKBeUrSErmJNW\nkCUenFJF4VOqmvV+ADAYkCen5pSVAofFjSxUihmYHwHuuZmOlr9xUyKp/BJbxtqSYxck/Jk0u4UA\nq3Yo+KQox9y06iHCVZIWjmGQSjL5ZNCN+CY6A/SZNNa32xHO0NnEKkjscxE+PAQ87TtlpaAaCI0n\nmYNo1BnLYOyCpk+DFhcvvAHmhqKkSFqZEplW8fvRG/KijQeAXMMwUz8km5D6Pk+fCkK+FxzbQ6TK\nwH9fqXJwAys/tpl8B4h83wD1oCG1qVsElnELp5JETIR54V61cOzBT40CCE5ZQMwmhKOW2bpVqcNj\nt/2eW0C2n9SQvXfNor7doPEr6Iepzx8r5wEWXDhgAGnwkOBSQevHIQBBbOBMqx9rNPJU3wNEM2Jy\noJixCk0FSRkIV3q0ovA1JdzHgKcD899eS08wTpNiIEykOpEzfsr8dKsS89BMMxPgU/3lGhsex93G\nRZr78PQINPMd3hqmbvAlCyHuHmwW9yIWSE6dcuYAzA2bOWax49c456xxbDA0c1ty6mMumG4lRtbS\ngcYNWkqaQd2PwUFiYhTaLtQaaPB4VOTjddfC8pzbIyQoUFhLo3hqI79tPgCdbfCXB7MVTW0zBCOj\nHNyp8wFldpGzV/DrW5ZvUkdGoH3Zswsgdvd+WcSRzaJlR6+KbHa36B47NtHgclI/uO2CGzlp4foa\ndovzAYvSm1eqIbxDa8xCAY87mhVpWvBdLPjkE991yBXdiDDXBcRREz4oJAsQqoxpC0Wx39toA71B\ncCWWhk6NbZQkNfWaulY+a0ldkO3HjcQd5itouZGQT0f7c7+J++QfURJpv9DCqeskbg1wKsg1wozJ\nBBa+zHYHXPqNbqK9SridguIIJGhNSQ+YnWuWjf8MYifUELJd0O5ZwD3PhtyraL9OSg0B4l9KU0Mk\nbSWwaB1gyBkQ17D7aJbaFRODxNCEDjGbXdDotXyeXpyz5zNw06gpg2a0MUuKRQBzYEixCA1UjhFN\njcjlBWKQ0PJIPxX+myv3pFW4wBw4LpAHgRrbhSyD1BPeRS/ZtZFqM/oVqynvXq3/Kv44fO+VcO4M\nnnvwfK6wp8dmcYf0LyjVDBLNwCkNnzwMgIyczdBgbIK2t0cHvvNQvMumsodW6ykvvTm1QTiIeFIV\n5Nt2wOxjvBMg8MFAIlUIQFdDOsT1ybEMTe2Qb32tp6R6j6Zy5ET+9kPiqKkhLI2YWbSBDbu0Fsck\nUJSuIf+LGxEnwzcjcOF/5wvJIihM/YoYhQAVs3M+F2PXgnZtAzD1YQ4ia+5lARy+NmR9kYiaSkvl\nl2EVqY1frh77vLnLQhyACBxIIqtyanABc98AOeBSvgM0kBvMmYKMa5W43GxIDhy0esv6LhXtWq1N\n6Dx3D/4c8noRJqBtOfs7QHJAobGOku8GZy23nC3Im/A+OyphL41Ss3hNiGQetXQvLefBLIC44fib\nWX5+jxA5xyxm06aYXGlvdxs0rVuxt0eHAc20wQjF9eiiJqeP01LcfnuBbXMbpsCorszrbjLU8d2S\ncoCo+UrUMAs+0CSzoLhO5K1RYVIqlFZP2YukesBnMxoln5RGHEmoTelabghlnzK4aDHtTEBLy2Xz\nU/U0UDjU90KO+YlR8LjeryWhttghLGVnxkxVTfFri5rhDm3rfC6CA4DmbbzudNd5gEXOwCnVEAkS\nQNwLNPWD8u4ADA32uw77Zzq0zYg9OtAn7InK7f1yYNo7YJxAxIFK33Sw21tnqKIOK+fOqV7SUEf1\nbBAPJu1Nw5eMa29iObj54Odf0JJf0+LgQfeQ/hi8XO3+WljG8WdOgQ9/BlpsB3GUYMlVt45dz1jF\ntEHNGFeJ35KDQK2KQsL9K4BgxOTlkc8F/RHjaAdnl4rUzQ4xKAiA4AU1A9AMI1rhoOV2yxoA5nZI\nIPL02Cw4WHBWQTIocUA8uDQqJ/MOAHf9Hpo5cdPUEDlLMqJxRk7tzczffj3i2RsggAsNDi7atCK1\nCc9DcVKt4YNfYwkp9qCpSiVAyMXzdA0wNOahTacC8e8+snOu31Ocz2fa4D3JBzEVI42enGEA84Gf\nEgIMCsvmoDIuc4WkmOYBInd0406HT89sSIpZ0BsELF37upOmgkg/iymfAVo3K7JvO/RNN33mrWdH\nvvfhjVtwPKkjPTa4ecZ1qQu+TRrdhxsztUECxIOEg9uo5KXnBcpgUWIZgK6upK6VQLUETDSAlyKZ\ni/xd+T1kf+Asg9idz9sObl3G7QBcjXET86lNbkuQoFErHA9TdhK+kI38Li44UGjfdOFH1i/aEdgj\nnhFx6kgLORuyppwfWEibhWQP/EtZHBi4qiJVkUgNAbC7QL/zW5M9jncgajHiBle4wg1ajLjGFS5x\ng8aHAaDDHlfNDfAMcNH77kV6p2RGEgjorTggXg8gwSIlOTbQsbDGNqQ6kgMZeY+UoTInuXSuhgFz\n93kNZLkdisKJFcBmcMVPjlrjvNocKKaBjHjQU3gJiEhiq7GOSeSCQ6nCanFwdgsKrz3rkZLzAAsg\n/RaS4CHzSjDh+bVB6/9oqz25GwD36iR7Bd/fMDJ6NgNsewvDbQFAPBXGd3KWnRqYD5oULdXe5Joq\nIQe8BBDtOl4PiLypOkjR8qf6MH9eropwgJB5eHn89+X2IAFyBohsF3SJ9vaXXUtWLyfnM4hOK+f1\nnKlfhdNPySaAmDlIVQTi6MN3O7crwP6ZDn3jDELEHK5wM4Uds7iO0jtvDh3QoH31iMvm1hnSeE/j\nhsOeHQlUJHhIx6RUW8jXY2pKMTfbwVWTVDq/Vt63BlSGQhoPb5U4YA6qdMz1EyCaGbns/J4Tg9tr\ngjtLkS1BGia5cBdvmTelcsgmOrWstQS9JOcFFiTaW1W+ZWS+lNqS/AuGzn0TnLIGhM/B9dPqvjBL\nQjYLCg9Ng/12xEV/p9sGUm9XGkzyLcmfWZOS38LSgZ9L1+5Tqot2zp+Nz25obFJbUyJnSDjTHMWR\nswxfxkUbHKdydgUAM+9Mro7wc+261E92jL9HJOwGY+tcpOSq6RRwPD2zISnR3iCaAYgzixb6MnaI\n8M4A6NDvNui7YNgE5syCdiaiY4c9Os8sGozYd3s0jz+NrdQvaQA+hvO9oCM9A9kstGnTnBqqDXoZ\nv4R5aPlyjKIWJHIi1S9+T8k2+D1rmAUzdhpvSLwcMO3GTULMgj/iDcKj1q4x4WDQinNtnUjkys+f\nK6Vq8nwtpo/sSQDgSxfiTfXWYR7nDRbAvGPQgOedVwLIgBgYpLETdDSTzwUaoPPGhmtcTmH6CPKI\nZjrST7DBxjOOPfAYeA0+ja738+dUR5rxoCPtX6AZ55aqIjKszVaUgCQXVypH3h8iDz+XO0UNLE0a\nBjQ1RmMasg6a+uLLuRzgNtNl6gige1qmikoJ3f4CMThciHTwuNzMUqrtI2bBp/j5l8niI8nTs+q0\nJCXjJg8X1Y/4725osN9t0DwOHpzkxemmTjumhoTdlenr1j2Lu368xdju8Xi4c3WhwS+/eQkE0NAM\nnFz31hiG/N1TnY7y1g74pWyiZLvQ6iwHekk4gNI1nRJHzzmydFZ/02LmqAXMB/TkOAV9FkNiGi9H\nNpuMm4CkZZ+Z1NpaPhfifGNLddGZhfQVerq21QMQf15E+fxIChiA8ua9ioETL9Fxi37r1YvHPdyn\n4vppgQ7tc0iMgtiE++7pHuSLQbLv9hj/yDU2u1u3jTuBAu1RwJmFZBe1zEKTHNuQ8SkAqOmsNeXX\n1JOYA7c1pGw4nJnw357AgdqO20N4fq+6XA7A0PqPM41zvMl1MapWbiaFqxo8fMnOJ+9S7a9hR81A\nzdqa78ky3wZHrqheZ5ifCVjITTgtZoAhfzWISwbEHa3mz/td8GnU4KlJLuCb2fcaes88AOAGl15F\ncTlHNGiaEdgCzXAbPqBLdecDgm92olHppb9OauDKcnIAoAFBDkS08kv15r8VBwxNOLPKqU2NEiYD\ns29fWpWq+V7w7rXUIMnZiQYYYEf18w3ShpFRP9AE42aNrDlTciZgQcJ/sgRgSFCQC8l4WPPoVKZT\n7166wm5ocLN1S383Hg4IBBqMuIL7fP3oHbSAeJdlAMyu0WLT9OhffeM+PzfcOZZBACGdsejcFZLW\nvWsk9YuWBvQaQFOqg0yXz8mBlL9ZOZhK5y0g3h9EswcxMT3wKrgFXbQWRTIMgG1Y44vgKomminCQ\noHjJLK7gWMUlMU76axHW8GhMg8I+fWwbtqapm/kAxUsY1tl/U7aPKsaYN8J9Qf2Pwu088by19vuN\nMa8F8KMA3gT37ZCvtdZ+yhhjAHw/gK+G2zjom621H8rfxUL3kWtRBAwOHJSe87uQYPESgNbNjFy/\ndIVxaNA97iM1gxbpAMFGAWCaYiVGQeE9OmzQO3tHt0HTjdg87tH1e2x2d+6jyrJDy9WpgO76vUQ1\n0aQWTIC0bSQXrnXm4mqGVCH525XPZPEjhXNTsTsR58MGbHu7ndtnQqsmTbdK20XObkGG00sEkKD4\ny87vlCXVD+5RS6uEtyzcYnLRty2iNU2kdjjAaCdHQjpf025RwywGuA8ff8gY8yoAHzTG/DSAbwbw\nM9badxpj3gHgHQC+HcBXwX2J7M0A3gr3XdS3Hl1TWSOqvXx7HKKODAB24QPKPTpc4hoNmkgN2bCv\nV5OdgliFW0tyFU1fkUy7FXUtxrbH2N6iGdzXuCcbBu/Qkl1wgx5fd7K21Az0GpCo6VW5PAT0FNbs\nFXKFrzR60gDjG8kI1YUvZb8VS9kpzGc3bhNhsLCcDeEMA0hsh8j/SDiQimnT+acK08d7NXBaaz8B\n9/FjWGtfNMZ8BO7L6G+D+1IZALwLwM/CgcXbALzbWmsB/Lwx5jXGmNf5clJ3QfxLS586xi5SbEJj\nFtI+UfLq9OpI245oHg/YT8zCGTYBTCDBGcUGPWgmZYMePTp0/rj3syo3uEKDEZfN9bTh6ga9+xbE\nMGKzczMo7ejWNUygQc8EJawBRg3zSOWpVR9ISjMkMk+pXMk26LeWLDI1myNtHzuRn/L446Qy+P4x\njPrt6Kh5f5JINYSYRQunflx2bmFbxBQ0NUTGsby2BfbbR97kvplsZ3wLhT02GNi5e9Qn8CkAY8yb\nAHwRgF8A8DkEANbaTxhjPttnez2A32aXveDjMmDBRbPsCfuFlkUbABwsUmwjUk3IUWuPy8c3GP1M\nx34CChrkXcQoAOAGtBFJMx1JiFnw/RAbn6dpRmyaHmPrQGP0C4Tctu/u+gk8IJ5bPrMcGO5mc8m1\nXalH1Ezb5tSXQ0SqIDSA+H4WklkoBs5ZXRvgooP74pu/ls+SLBECCn6LKc7PgLQpNiHrJu0W/rqh\n4f4VgV1oO7rx8L07ZRljngHwjwD8ZWvtv3emCT2rEmdnmYx5DsBz7uyNmTtryIA8swDmxk4g/lG4\nPstZxmBw217ixXbAfruZOmPw4mTTpB7FO+zR+3AL5wlKcc4DtJ+YxcbzjQYjOvSBuTQ90ABtF5uo\nAEzsA4BbbQgHJi6NtWktA5GjoUvEk5TsHAkVxLLwkOmv5P9guMrFfyfOKMjBjd83pRoNIh+PY/4e\nZPQcRrc35kUffwxZfoWdFyNtFhw0ruAA4jOecUfzGDFrSP1tRdjn328fYWj4MoRg2ORsgrOLe/ez\nMMZcwAHF37fW/mMf/TukXvivqn/Sx7+AePS/AcDHZZnW2ucBPO/Kf4uNtxKh3sLjmEijlQYWnDm0\n7MgNnRT/EitzC+ClC+zaK4xDi8ZvX0aGzY5Ru46pHGRocmtHXNw1rnCFaz+bcj0DCwKEcD6gxRws\nmibsioQO02wMBxMA9V3QjWEAACAASURBVICSAoeUExiXDGsgcODAMIoepk37jeTE5hkVOqaONdCd\n14BIpdDqUwsWaAHTOFZBLGAY/BfRxxg4SLQFZQQUxCYut84ucsHVCQkEW7hlAHTUQGML2C3Qdxvm\nNEhqSIgj+5oEintTQ/zsxg8B+Ii19ntZ0vsAfBOAd/rjT7D4bzPGvAfOsPnpvL0iJRIRCqDB42S6\nLEoeObAAHlA63ALY7zboH5NbdzBqhn0O4yOl8R+HXMdJ9SBQ6fz3HQh4HG+hD8jEwDFOqkzY7GT6\nNkTj69AMaMcRY9s4lcY/88jBo2Hg0Qo2AswHpCYJ9gAEoOAgwQFizH20FYCbcAu3MfyEby5EdeTq\nCIGLVNtIHSGVRF5DsnMGyEt4wyfdz7eXtusWFw4UF637m1QP2l6R2C1tQMQdrjQVxB/JsDl3vEp9\n0gKrAkXqmaV8KYBvAPArxphf9nHfCQcS7zXGvB3AbwH4Gp/2frhp04/CTZ1+S311EkwCQNZuIZkF\nt09IZsHZRQs/dQoFWAwwbHHtR9ZmuwcaRNOkk90BtG+nUz/4zlvEGohZ0LuAMwl3LlmFzjQ4SEmG\n0cAxEMCxkSlfl2Eg02YqoWnbml4BnUFowDA0uj49SUOHwbMKp3btfR27nrEMroZwA6h8GSAR1yF8\nb/Rllk4D2n/Mh1gG4HwyyKaR+/gyfdmeQMIQINDxGcTsgbMJYhH8nMVJw2ZwGAwMgz6YJdWQewML\na+2/hG6HAICvUPJbAN96ZL3qJcUuUgxDU1u4nUP4ZdztNug9YDSPw+Akt+9rXHnbw+DBo/V5XN7A\nGtx1RA0JJOjTcp03n5KbeWAdHCQ2CeBopntyBtJOdfAlNY7RDE0TMRAn4Y1eKxp74MyBQCI2uJU7\n7ti0E0sCgB53aBr2GiHAHxDPGCmb96psQxpFW1EOqSb+fGIawPTVs0EB14lJgAEFgZAwVkbMQe7A\nTmDC8o5trFIM7E+yCp7vKV0bAix2tJVAoYEEj5dsg/5eQjAkDSyuBdBeOJ1152dE2g2GLqgOAxrc\n4DKaJiVmsfdqRYMBV97zMzCLwBroXQHwZfDB+CkZBoBZfCkuVl9G7BsPaE38RicJIJKXHHvQfE5K\nHZcD4L5xDKlr/Xc8yT+FfjdtloT/SYcuILzl6fd2FQysha/dIcPrLiw+u+BlzSsfWA8BAODYAoGA\ntE9wpiFZhmcitx38XrHdxC6CnSLYKvqJbQTHLPd4T91X1CVQLKgagUAKOKQ6Ir066ToOHrwKwwWw\nvcBLAB61I8ZnGoyP24nu0deqCRzkkVav5tSQsCTeAckNLiOVhNs9UgCyQT9TUzTVRYIHMZYJQFzm\nSb3JiQQGHuarH3lemY/EtROFA0MaPRANj3vPiG7Rkb2Fq5hSvaC3NTm9UZgbTSVIcOctAg1u29B8\nX8Cu4cxB+ksQAEjQ0ADE57vtggqy97NrMWjEhk5SUUgNoXZ+ipaoJ6dh15WccZMAhW8KC5HWdrjb\n9uh3HZo2zEzIj7y4y7ifRVBfUmoIqQo8jQMCpYe4WEUhQBr9vWiGhucDMMVxN3Z+XzoPefMMI8ce\nJEhogCLL0u7Xe5VvRIuxaf2OZLfo4AFD7u6tebymjNzy9w6VccJnYbg6q409zTjJgUNzypLAIq4f\nWzcLQlPzBPfOGatlcbHaMf+49/FD/UzAQgpVi09KVQKKZouQDEN+iV1+v0NWYwobYLfF7TMNrgH0\nuw3GZ5xj1R4bXOIaNxhx6XcJJ49OPviJNPIp0w49bphKAgB8OjU9tZo2gpIqtERdcdvLx7Mu7tF1\nwOCAqLEHDiRxXp1h8HtO08Ls2ejzDGPTYHjcY791+5+2o5v6BDcmAnOVQzNk04IyiiMXfLKHDEo4\nJdxWwuvCfCVUFsH/WJzdAtePt5O6cYMr9Og8uwgqh/PnuQT/WBYBCrXvU6aGAHp1FqojNdnlW0Kz\nZ6QAZgtgd4Hb1hkI23acsQwg2B4ATHH8c3IUprc5gQp/uwYGMZ9aDTYNyTDmjGNuKKV6tVOcfKtT\nmgvrb/yYKeh2isAyyuwix2pcG4XwiBZogP3Web1uJUvgg5v/NeLIRfpjUL5WhFPCncRSaggBSIph\nMGCj6VJtVWm8wjR8vFvuVh/a66kBC4O4KitscZpiFhIEgJhd8LfDVlxH1zwDYNjibmvx0tDgUTvi\n6plrjNvAMmi6lFgExZE1A0DEOsgzNLCIS3DnLblwLcUcOJhQ/hQjcY8rDaccUDYqSEiRKkgJOFzc\nnIHIe0n2Q6oSMbYGI8buBk03YGz3uHr5LtgxuC2Ds8gUsyA2wZauJ4EnJ5oqQmBArCJns/B/u8fA\nfnuBa1xi79mEYxbBwHmNK1z7xYvkzcm/1Ut95ilTQwzSBs6E+sEHdUlqVBNenmZFp2s5gOwM8MwW\nd60DjYvtHo13E990tMiMwGIOAjT96vJcgvbzpMEchwNYcPDhxs+NdzHnagh5k1I6EFgJoBlCN1Ea\nScrYWWfgDPqzZuNICQc3fs20yRBjRGN3g7Ht0QyjAw16e8uVvRpY8I2UKT+g72gG5AFDs1lwZkEz\nIwQQn4Fg9Hy1O9rHQf0ggLj2Rx7HgSM+0u70T7UaAsRVqrRTtOLIJaWaaD+8BipyFyvOQKZ0twBt\n9P4Y4+CZADk8IagcG1C/pQEQVJQxGsDhGvk1bMpLRkwXN4BWrpDaQmDl9hN16QQgnEFIQ2jK2El1\nIykBBI+vZRnhWQKTCDaL8Bw8HxFxNM6LtRl2Ti2RklJDeBydc3dw6g9y3Yom0rApbRekZnD1Q/hV\n9F3aW5NUkqB6SE/OWB2hNn/KZkMWGDFnhsdK4SAgAUADHB6XSp/eVAZ3w2PctRa32w0etSM22x77\n7QZNO2LTBBZBb3bq/JwlxIwhhMNO4l3ENng5HQMGySxyRlFg7l5O39IkKTELF04DwxLfiza6b6gj\nZ0gB0Jxthew+HfYYH7cYW79DGU2n0sDkzGKLwCjI3kFqyAD39h9YPBD6j6adSZuFVEP4uhBiGK9G\npIaQUZOrH2TApDhn5LxkHpxhCpV8fhyzeGr9LA5gEUuBohXnQF0ZJdtHFGeAocEdgHFoMXrnpj26\naQ0HwAen+xE7T60pjQ98593ZgRspOUPZsGtDmc5hjJfDDZ6Uzm0C7jGaabBqKoA8J8kZNktGTynB\nC3aI6iiZhTQcuzy+fbsBbTsCuHXu4vQbccCgOM4owOK5UVNeR3m4EEAAMWDw2RDuys2Nmh0tFnPP\nT74UZIeQ+8LyXbH4FKo0fFK7P2VgUSGpN/0h9gs5EwLEDAOJML9Ovqmm8AXQXuB2aDD6BRSbbY+2\nHTFum8nbkWwX1CEoDMC7dPWzMF3DGQNnE5spPfgmaNOrsYGU2yxikJCMgr/1cwbLpWpJSnhdeBsM\nUzuRMbmZnnVPzKMZsXm8x37bY9PdYku/E7dlcEZBLJN/uBrsHMpRiuyXUg3hxk5y0voMF3f9+BH6\nboMX8UzELPaeWRBr4GtDyMB5Mxk6wzoRvlHTU6SGZKRWPdDOpcg3Ch/kMp+Mz6kk8p4MNO6GFmgH\n7PysyTA02LdONRm7FvEgD4OYPEIbP+D54AeC0xUQZlXoOqnn51zHc67k7tFilWCPWDR1QgOGEnik\nhNcluNF3k1oWqyHBmEz2lg167JtNAI3dbViY1iOoITQDQkACzGdFwM6Beb/h/YFPs2rTpMyn4tYv\nFHuxexV6bHAzzXx0E2jc+JkPUkm4isIBgs6fcjWEiaxVChDWAA0tblDO+bFV4rjtI4ozQHsBbN1S\nrf2uA7bEGDZo2gZonA8FqQcurWPhYLgEMIEAhUlF6dAr6opUQ2LjKJUrpylDzlglSIlGeWvZBc9D\nIo2vfPqUq2WubnP/FQ4yE+tpgOFxA2AXL0zjvxf3qyCRRu6OHaVogEF/3Hbij3brgIK8NIkFBPYQ\ndu+ODZtz42d8fFrVEDeZoEuKUfC4nB0j9YRyFkR2EK6a8C+4U3nyy99aXchwxlSTu+3txDI2HjTG\nbYNNFz4nQD+1DMfpQxSm6VLpVyGZBr8WgGLP0Pwt5j4QXOSgX+qglerIcgo3PGfwjJX+K2QY5ioJ\ntR/5ZuwfO5Xvqr/GZnfnPEBp4x/OJmjmK/WpBoiw7JccLBQ1xHrV48XuVSBVoscGL+JVk8oR1JDA\nLII64sLEQglgpBrylPlZMNFqVLIjpK7TJDeVKm0QSIRTBk4UwgCwi1lG047TVGvTjhiaBtIwycOY\nBgc3cIbp0j1jKGFwjlM5xCbi6dvAaAKjCFOp7jGCATU8Wt7JqmY2JMdW+D0D4+GepbGthdcvGHLn\nqtQU12Hyzdg0d8FtXO7MRUcCCm3tCZdWpG3Zubdf3HaI1n2Q2kDGTP1LY41IC8vR5x6dDTg4Pz3M\n4hFCg0opqSSp45InSwEFRDj1QSMOJqk6cOu331PpbnuLu3bEODQRy2hat/cEuXhrRkrOOghIJOsA\nYhWDG1GD0TBsPKyxC/co0rNynHU+DThKtgpZBj+Pp23na1r4Kt5gmyEgDSt+ZZwzhoYtDvfNxm2a\n3Pmd1h+PgW3Qbyu9OQG9j5SYRRtWkfad+12d4ZIYg7NTvIhn8CJe5e0SwT5BLOMGV6CtEfYTw7jy\ngKOvOl1DzgMsDNJgQVIDGrmwJhp7gIjTrN6aaqLFaXWO4gJokGri9v0c0G33bnOaRt8chwND51Pi\ntR6tyBdUEBpIlA5glsdVt04N4eUsccoqLVnn95WzNdLtfTOxqnmYyDnfwUz6vGzQu70zmj2azrdr\n77c49Lt1AWwbwpS/BZs65XuSjq1z3+4b/hGgJjJcctDgKokDlavIz4LUjusJJAL4yM9trqGC+Ec6\nU0nVLAUaJZA41OCZO8ryNWDJsZXp6G7O96rqd3CAMTTYdEFtCOwg9NKQxvcej52yaLYASjoQHJ7I\noOjCeTWEx7v8aYBYsqCMS0otkg5psr68rpyGS89ZIJ4eHj0wN2im2aq2Hf2nGsJnGtAh2ghZCgEE\n4EACoA1s4i3v4r0ogv9E+KJY7Fsxgq8DkYbPJnrWqa3Hp4lZ5NQQLofYM5baNSTb4AM9N8fO98Lg\nR7krl1oHP2PStrhrB6eatM3EMpya4plFM2cWKfVEqiQ0YOXu4sB8YZprgvkaEX6eUiNyKofmkKUB\nh66GhHpRXQnkyBclGED5upthehPLrQE2LOzKvIKcap52V+/mU8spkc9ONgUOEiMav0Ds0qsVV5MX\nppw6jZnFZsYwAghtphfAvvfMIvcdhgXyygILTZbYNHIi9c6UoTPFIID5KscSgPB9NFoAbQI0vH9G\n68+lTYMvENNAZP7GD0ZC2s5PAwptwKZEG/A1C8mkgTO1TkQ6jHGVStosaHYkBgsOJg4grtmivugb\nLsxvRbZlqMPA6phaLxNKAcA8LjlYXDIbRFBDpGqyR6ymxNs5buKjB4l+R2DxtKshmuRqmwONQyXl\nxAXMwWOJGiPrOQOfsGQ/t41u38wXiEn/iTBDErtLk+8CnxUBYv8KudgMmIOGZuisXURW45gl6xNP\nBQfVJIAJr/PcD2Pv709AAvH8ganEHqMjdHaTq7dsD3r7k2pBTEN+zFiqJjwPzXyE74OEHbQGNBjH\nZmISBBJPF7OoMXDWSi1oaEBAYX5tbgpVU1HkPei4ZdfwcPIvzTIAoNv2btMdbCYfDc4s5MrR1J4W\nNFDk90pcWFdDUlILCjmVJCWp/Tfi5fnBZ2S+lD+oF9rWABvBNlyZsYt93CbxlK0mkllxQOC2i2Cs\ndLtdaTMkN7iM2IYEmmtcYRwb952bnVNDaJPp7OfgFsj5gIX2Zq6VWmPoMVLDMuS9NXBB4bqZMJYx\nNMDWT3XuOnQ03ep32O4bnVnEj9FMgyE2Gm6ms1EMiBSrIOHMYClA1Bg6uS1GTiOTfk75ghcs+ZvE\nLIAbQ6ktKBwW5QUXcu2erpwGe3bfVJtIe420WXAVIhgzY2Mnt3kQ8HCVZmSMYmDrkSaQuC81xBjz\nRgDvBvBH4Vjx89ba7zfGfDeAvwDgd33W77TWvt9f8x0A3g43sfSXrLU/lb+JUpMcIyjJktmQFMOQ\n96oZ8KXBTwyDRC5cS7INzzKAaa0J+Wc0zJ7RkE2jyTMLbrOQbAPIU+2SGlLLKjSbRYphpOwWIRxP\nB0vbhYsLNgu+3ykP59bMSCZWw7r4s/JNdCXDiNgBGsYsusj4yb05KX+PDfZjh/1u49Yd7TrcTYyi\nBQa/mnvJyzchtaa/v2Kt/ZAx5lUAPmiM+Wmf9n3W2r/FMxtjvgDA1wH4kwD+GIB/Zoz5E9badKtq\nYLGmLAEKCRA8r1buUsYgfTAkQGhxPC2hnuwBNO04sY3Gg0fTaB1f/zxAaiDMAYPezvOGXWqn0FiJ\nFM3fQ1OjtP1B6Nu05KYFpLYuvAJXYVw+/fMLubbJtQUHDB0sghpC6gkHBkrni8Zu+iuMQ4N+12Ec\nGgcUBBIpP6EDpThE/XdKP+HDLxpjPgLg9ZlL3gbgPdbaHsBvGmM+CuCLAfyrg2qi0ftc+pOQHAPh\n6cfEzUQ3grrPFIiL2Bh0TIOrH/M4yUroyMuQoqkTtSoJz6OJtr8FlSPVqNTmxOE5g7s3N3zyL8Cl\nWBn/IhzVh8rSRHv+fgKvoI4Ef4k2UkdI9eDGTZ4+oo1Uj7uhmQPFfYIFF2PMmwB8EYBfgPsG6rcZ\nY74RwAfg2Men4IDk59llL0ABF2PMcwCeAwBsPzdvd8gZD0sgU8udtHway0gxDO2ammu1OnBmAcTs\nImIdfu/S9gJorWoIbdphUlEk09A6f+7NmfMtKK0N0fJI0CntuZDbb4MzguDVqm9wTGzDGTUdo7iZ\naqR90ClmZ4e0Syil8fUJMyBz34vYn0IzgI5ocfPy5ZxR7BSgWAkwqsHCGPMMgH8E4C9ba/+9MeYH\nAfx1uC8W/3UAfxvAfwt9yys7i7D2eQDPA4B59bM2O1hTMxUyDkq6JjlwOAR8lgBX6plSoKipRuoz\nx4bQu3bEI8XF0Nk5GoyNZBbxQq3cPpw5SbGGnNqhqjPe67BpOKOJ6wZg4kRONtN/7tVKhlBusI3V\nmZiNaHYeMmhSPMAZhb4LugaWMfzMN6shBjGfNRGMYqxgFE8CLIwxF3BA8fettf8YAKy1v8PS/y6A\nn/SnLwB4I7v8DQA+nr+BUhONyuco+6ENkvOD0Aa3vFarQ85mUTUDotRPhlXmEQyhKZsGZxrAxh8x\nMQ5X3PE2Cx6e0fGR/ABY3srpvam+iXpLYMgZLEvTpBR3zPJ92Q4cMKTPBTd0UlxY+9FFjCIyZkpG\nwXenv081xBhjAPwQgI9Ya7+Xxb/O2zMA4M8B+LAPvw/APzDGfC+cgfPNAH4xfxPkgQEijWpe0xAl\ng2UtA6lVL7R6LQET7f4SJOhYAg6voty1dppBuYXbTLgRaooLu/g9whtdUmxSX0IVNftFGRQG4Tgk\n82nSMKZEdhlZ91jNgqqacPDYT3FzdYR8MOTsC5WnrcaN20FnURq74GChzZDwWY9I9dhdBIDQwIIf\nj5QaZvGlAL4BwK8YY37Zx30ngK83xrwFTsX4GIC/CADW2l81xrwXwK/5an5rdiYkV6sSLU+FSbTz\nnPoh2UWpdXKgptVDq28NC0kdq8XAbb7TOtAQqbQGxYW9OzkdEasDJQeqJQAhweGuABaUrqlX/F6x\nmhWrJkDYHDleTDZXR6TqoflZpFgXkLbjcFYhZ0a40ZPvUcGnR2PVA2nVY2WwMNbOzAn3Lua1z1p8\nxQdCRG7Qaw2Qi0tJq4Szb+pMXC5dK+/Ux1w9pnjHNtx5sG9wxgGENzdJI841NpBjDRMg8Ou409Cg\nmbyo3qyvsroDKLIlOYUM5L86r/lyAIBm5KxdVObCuqGTM4uZ0ZN5Zs6MmcQoiE3QOTBXQ/68+aC1\n9tlsZTNSbeC8VznEDrEUPVNv+GMMnDxds2eUmMLRDGKJENtwA/BuaIB2xN20rwYxjDaaih0Kb/8s\nQABQvQo5QOSel/K1CHUnJzVm1B0HNxsk2RJnHLQ6VXqr6utNXDyAWX4n+c88SnWEG3zlehGeztd6\nLDJmypfnfRo470UONQrWqB6pMim9ViXg5aXUlZLRM3Uvfp4Dr1qQoTKLYbJtwKkpAO78ILsDAG/n\nAPL0XwWE6dzfTHoT1v5WmrSAW9qPUH94+4x/hlvBmAg8Ou8yr9k4co5rLqwbgWWYizT6SvuF9L0Y\nvNrhnK02GIfWrfPYda4NpX1CU0VWBgr3zOcktW/SGuTMdcTa+ywBAXlNzpZSsrVw0ewoKdtKjUqi\nhaejHHyYAARgzl/c6Utbd6AxhaVqo9Ym8laqKsmfYQ4eQNj7VKordEwBiAsfDxbu8fSZEVI5HJPw\nIEFsgmY8asACyvFIOS+wOFZqbRc5lULLx/Om1IslZdcM/CET5pJiKLUgVEofzDzfUPjKfel3OBQs\ntLZOqYVTOximtniVBY4Jac5rJCOa6QtyMXDM1RGSkienu66dyqej3KyGpkXdiQCKWgYxKHFHynmB\nRUnVyHW4HJqu2WhLmUWunCX1SoGJBjwQx1Iaj5PpMq62njXhpWCRqkuOWUVH8kNxzOnOe70CmFhH\nzlAaHZl/R6hGWk2bOamxWaNgk2iZXYKBBOAYhDRi5v6Adfs9zg0sNKkBgRrAyJWdEu0tlnqLyzJr\nWEuOOSxRb5YCzynlULDI5V9y75IdZ8Y6ghNbzlA67VQWTS2LFbdNGiyA+bQyBwk6n4CC/CdSIFAL\nFCv2i/MBiyUdRWs4ZI5auUufPGeQlPm0e+fKO2Sw5wAmZ/CkYy5NA7wl7bUEGNZmFzxcc+SsYzL0\nXnjjrsUdgNt2AJ+iJZHTzDUifUwmcACQnOWgP84ogDnDgBJeUc4HLAC9s2gAoCGszKMd15LUQM/Z\nHrRjTfkk2gBecq8aAJHHQ8GshjHUMJBUnFafFLAtAQ46RgCCWHWZyrUeUAbcRvVIsIvcDFGJMQBp\nYNC8NZ9qsLCoYxYpBAXmDSTTtbKA9CDS8qTCNepE6T7afYH5L6SlHcNQuKQGYW5gyjppcaVwzW9f\nul+tpH6DXB7tmoEZTknI7yN7/8T0ceolmAOQ3LXyeVaQ8wALIP2AORZRE06Vn3obayJ3tKLrS+xA\nK19jFocO8hLolO6fYxsak6lRsZaARG24RlJqSY5laHGpPDXl8wXXJUCtHfApkKgBjtS9D5TzAAvp\ncV5qABmubTTtHjnVINc6OXDJlV9iMKlyUvUqgU4JoErPv5S15DrqEoaRKm+p5ABUi1NZBObgkGpT\nXl6qPjJc+2KsfXGW6nCgnA9YpBox15C1zCLVaLkBTNcd8z2TEmDwvLl6SikBlZY3xUJyg0PGyfrK\n++TiSp255m2Yq0MONEv2DB6ujdPqkIrT6sjPS6ChhXNxqfsdKecHFhpr4OESSNQCBaWl3jD8u6X8\nA0E5yYHAmipI6prU21Cml44yL38WoFzfGnZxDGjk4msY4hLAqMmfqoeUHADm+r085trwDxVYAPkG\nghKXAxHgsI6VGzg8n6xT6k2ce8OnykuJ9jzyvho41TAcmVeGU/cr1XNpx64FDVmnNQaI9hul+gW/\ntyyjdI9c+NBjzb0PlPMAC+A4ZpFC4lyHK71RS2yi9IOUKL/Md8wPnBskGsNIpeUGxZLBUMMseLjU\n0WvbphbQcuxg6XkqrkaWMLBj2NlKch5gscRmASVNi0tdJ+8B6HRbG+Clt3Wp/LWYhZQc+NSAVM1z\n1bCqXNwx4VxcSlIAWmIfOQDldUipZaXfsaY/rsnIVpTzA4vcUbIIba9BDURyQvn5oJBsgqZOa1Ee\n0DucNiBlOfclOcZRyldT1yVvzVye2vuV7g/U2TNScYfE18gSkD3kfEU5D7C4Q/xF8RyzGDAHiWqA\nsNA3H2dltEgbNXkdWnHN0h9pLf06V+Yx95AAurSnHAMWSwdADQCk8qXKLj3vqUZODfM4JO8Kch5g\nAeQ7Uwo0qgEida4AR07tyIHEIYNp7dbPUWgtXGI8mhqSq/OpVZHaQbDGYCn9Pqf6/Q5Nvwdmeh5g\nQWpICjBSqkiygaQR5FbJQ266LWagIZlFTv3g4LFEaimxvO/SMo/Jt1TOBSxqJVfesW10LBM7VB7U\nEFSAhGUZAQcQJSiWZvHEDk/SZiHTgbqOd2q7hVRBKC43K7Bm51obLEppNfevlUPsMqeQ8xiRqpxH\n1WpnQ7JAwUECyrmUC5aHwmxRkBzcvMictVwTrawl12vlccnNTuRUpZKaotXvUFVgTYaxpB618qTA\ngcta4H2iUV0s1hizBfBzADqf/8estd9ljPk8AO8B8FoAHwLwDdbavTGmA/BuAP8JgN8H8F9baz+W\nvYlFmlnQscgmOJOQgMHjLlhayhResbHsEopcM315jOTAo5ZZ1ISPqVPq/NRgcQ4gUJJSP1hqbD3R\nMz+qyNMD+HJr7Z8C8BYAX2mM+RIAfxPA91lr3wzgUwDe7vO/HcCnrLWfD+D7fL68cJtF6i97kQQK\nHk9/YOkcTHh+QL1ZChxqKXIqLflsR0otkOUGZxWjS5RZYhXHAEWqLtX9pvK6c/hL1TH3DCeUIrOw\n7itEL/lTvwsILIAvB/Df+Ph3AfhuAD8I4G0+DAA/BuAHjDHG5r5mJJkFkHlwqXJoYMGPsrCWpcuN\nZ1t2FOyCh7W4mmk4yS6WzDTI++Xuo5W5hEGkHLEOfZsvAYKl+Wril+ZZ45payRm6l/SLHDtZSS2p\nKsYY0wD4IIDPB/B3APw7AH9graWqvQDg9T78egC/DQDW2sEY82kAfwTA74kynwPwnDv73IofhEAi\npW5IxsDTuWggQXKDYL/IqCN80GcISVJSneHYTql1Lu0etQByaN1Sb781wkvjlqTX5lmSLyUpQKa4\nWrtYDbCsBHZVYOG/VfoWY8xrAPw4gP9Yy+aPmtfTjFVYa58H8DwAGPNs5huKObsED+eAQrYkAQZv\nRQ4gt8g2DYHETrC7EAAADJ1JREFUUkPnIXkPEa18CRI1/hh0TlJjyC3FHQsUS0DiWPBYA3ykHGI/\novy1BugnZeDkYq39A2PMzwL4EgCvMca0nl28AcDHfbYXALwRwAvGmBbAqwH8f5V3EOcEEBQGYtWD\nnyORl0uOVUhR2IX2Y5bUkBTN1N7wwHo/dA1gIHFeosWl++biTqWGLAWM3LOsrebkZpNqAEArrwQs\n8r4rSNHAaYz5LM8oYIy5BPBnAHwEwL8A8Od9tm8C8BM+/D5/Dp/+z7P2ikk0K6dmrDwUKJDIl4pX\n/DR41aQs6WAcaGoGwqGyFlUvPXcqXcblwqk2SeWriTskXrtfKT7XPrn8sv4ynItL3Ue770pSgz2v\nA/Aub7d4BOC91tqfNMb8GoD3GGP+BoD/A8AP+fw/BODvGWM+Cscovq58izs4ewEQP520SVBcCSAO\nbSFiHqSG0FFoVtoPmLptLbPI2RnWlhKrSb39tLSl8adgFrWgeEzcIfGa1Kp/qWsHlNnfiaZSa2ZD\n/g2AL1LifwPAFyvxOwBfs6waJbDIgUIto5BCYEByIc75r3IR34L/MDxOhnPGwhJg8PtBSauVVCfM\n3e/QzlYzmE4FEDWD/r5AJSfai6EWJLR4rZyaMg+QU5rZFkhqLYdmpKxVPXKteyvOKY4AhDtuKTYO\nAonSj5UakPf8IyclByQkhxg1U2mHgMYa57Vxa4KLJrkp6ZqjJvfYf84ELO4AXLPzlCpScy6vI+Gz\nH6VfjYCDgIItbZc/Tg2zWGLEWhp/rJRUnmNYRSnuSQPGqUAoJYf+hiVVtualtIKcCVgAOkDw+Fxc\n6lqIeM4YgLqZEWUadSmz0DDqnqa7qmUtNpFKXzIQ1x7Upz4vxefsUTUs4p7AoCRPuot64TYLkpRK\nUQKI3C8mp05rFplREyU2zlnKLGRefgtepow79Jc65Lqlb8va+GMAo5S+5vnaYHGsLFEzTqiWnDFY\nAGXAkPHy2guRTxo1U48vWYhodal+APqPpFHHlKz5w679q9YMgkPUkPs+XyttSVxJalVWnjcVPrH6\neiZgQaK96WtBIuXaDZRXmh4hJRVDy4tEnmOrt9aj1Xb6Y1iFFvek2MEpWIaWR1M9l6qsT1DOpBqS\nWSwBDSTiNXVD2ij4LyN9K/g1yvXaj6oxCq0DyOsP7RBr/nqlTp9LX/KWPWTw3TcorMEypKT6Rc5e\nkWIXWtox/ahSzgQsLOYAoP0Cmvt2SnIAcQJ2cajx8lD6eF9AcSqQ0OIOzVOT9xCgWINVyLy1hs1c\nn3pCcgZVANI2C5IcKKR+MXo0bn9IrQ3Rfjl+TcbImUJ4XgWtqktafi1jpybHgsHaqkjtPdZmCffB\nLHje3MwHlHiZ5wmM3DMBi1pmsUR4i2rGTqB+UVmieCD9puB5IPLmri9JKk+pcy2RNVSF3H3XZB/3\nBQyHgGTtlCmX3DX8upwsZbeVciZgAZR7dK0btzYDQtcvBQfOKjLL1lNAcei05dLrSvm1TlkrawHH\n0vi1VZIaOQRccvcqDfYzUS9q5UyqKpnF0l+aP8YpZkAyZaSoZM7SDcQdhp/n5Bh1pAYwDn1ra+ep\nuFLaGmrJUnZwLAMpyRIWeWwf0WSlUV6zB+c9iFyevlS062qZyFJhq+1LbxwsSF8qT3L25Fh2cSzD\nyJVX85vwa9b+XZY+35I6HGovWUnOlFksFenGvcZjaWpL4pUgmUPOyp0TLW8rjjKci6MyZb6aDpYa\nTMfq8TXxqbS17Bel/Gs/c41/BbC8r8iyTyxnAhap2ZCSjYHbE3j+e1QGl6ohtdSyNv4U9opjBlYq\n7pi0pcBxCNNYeu0aakgJKM7MpnEmakhKbpU/LpLzLfHDWFmOoZKHyhl1pEkOUT9y1x2b95Rl3Iec\nUT3PpLvJ/Sw0kSwCuB8mMWDRLIqsyrF0MyWH+GlINakkxxgAU3E1aUvKPJZhaOElakgu/syZwlI5\nk6pLm0XKcQq4v/nII2XNW2p2i1Qefv9TyjEUfS2gyKWvoXode+2TAocT3fNM1ZCc+sH5bO4TAGuP\nlkx58o2UekMtlWPsFm3m+iVSevMuMfjVqCKlPDX3yuVbCiilZ10DGI/Jf49yRswi1UpnYMQ8VjRV\nRKYdIrW+GUvVD02WGD1zcbVlL7l3ri6nMnAeKzm19Ey79pkyCy7yV8otRX8CsHxftzx153kSBtol\nLGJtOfa+Z8wATiVnAhZks6hRPSDiNamZBTl0XUjinqW3Vk5FWdLxTm0YrVWhjqHnx6oaawPWqeVJ\nAsuKL5majwxtjTG/aIz518aYXzXG/FUf/yPGmN80xvyy/3uLjzfGmP/FGPNRY8y/Mcb86fWqS3KP\nU6JL5Q/hG2eR/GFvn1fw89fgTg/gy621LxljLgD8S2PMP/Fp/6O19sdE/q8C8Gb/91a4L6u/9bDq\npbbGQyGuJGeoEC6VnBen7JA5W8UxdgztXsdOlx6S/xTlntJe8QqVIrOwTl7ypxf+L/c5wrcBeLe/\n7ufhvon6uuOr+gqTp4R6riavhAH3SqjjE5SqbuU/XfhBAJ8P4O9Ya3/BGPPfA/geY8z/BOBnALzD\nWtsDeD2A32aXv+DjPiHKfA7Ac/70JeCv/j6A3zvmYY6Qzzzq3qSD757AvY+Th3v/4br3f3TMxVVg\nYa0dAbzFfyD5x40xXwjgOwD8vwA2AJ4H8O0A/hrU/fLnTMRa+7y/DgBgjPmAtfbZxU+wgjzc++He\nf1jufcz1i2ZDrLV/AOBnAXyltfYTXtXoAfxvCN89fQHAG9llbwDw8WMq+SAP8iBPXmpmQz7LMwoY\nYy4B/BkA/5bsEMYYA+C/AvBhf8n7AHyjnxX5EgCfttZ+Qin6QR7kQV5BUqOGvA7Au7zd4hGA91pr\nf9IY88+NMZ8Fp3b8MoD/zud/P4CvBvBRuA+YfktlXZ4vZzmZPNz74d4P9y6IsTY3sfEgD/IgD+Lk\nTDw4H+RBHuTc5YmDhTHmK40xv+49Pt9xD/f7mDHmV7zX6Qd83GuNMT9tjPm//PE/WOleP2yM+aQx\n5sMsTr3X2p6viXt/tzHm/2Fet1/N0r7D3/vXjTH/5ZH3fqMx5l8YYz7ivX7/Bx9/8mfP3Pvkz57x\ndv48Y8wv+Of+UWPMxsd3/vyjPv1NJ7j3ep7W1ton9gegAfDvAPyHcFOw/xrAF5z4nh8D8Jki7n+G\n8xMBgHcA+Jsr3es/BfCnAXy4dC84O88/gbMBfQmAXzjBvb/7/2/nbEKqiKI4/jtIGaQEBoVLjaBa\nRElJULQoCFpJ4MKdi1Z9LFq0E9q1rFaREEVSkX1Y5LLIoF1GZWZIFrUrdBF9rCrqtLhn9PWYeY2+\nufMIzg9krjPX+c85bzjv3jv/ETie0neT5b4Z6LDPpKkO7Xagy9qtwIxpRI+9hnb02O36W6y9DHhs\n8dwA+mz/IHDI2oeBQWv3AdfriDtL+xLQm9J/0Tlv9MiiG3irqu9U9QcwTHCAlk0PMGTtIcLTnbpR\n1UfAp5xahTpfM7Sz6AGGVfW7qr4nLE53/+Nvaml/VNVn1v4GTBOMedFjr6GdRWGx2/WnuZ33AMlr\nEdVxJ/m4Bey1p4tFamex6Jw3ulhkuT1josA9EXkqwUUKsFbt8a5t10TUz9IqKxdHbdh5sWK6FU3b\nhtZbCd90pcZepQ0lxC4iTSIyAcwB9wkjlc+qmpjJK88/r23HvwCri9JW1STukxb3GRFprtZOua5U\nGl0scrk9C2anqnYRXng7IiK7I+vlpYxcnAPWAVsI9vtTMbVFpAUYAY6p6tdaXYvWT9EuJXZV/aWq\nWwhmxG5gY43zR9WWBaf1BmA70EZwWi9Ju9HFonS3p6p+sO0ccIfwgc7KgsmsnVCZY5GlFT0Xqjpr\nN9Rv4DwRXbcS3lAeAa6q6m3bXUrsadplxm56idt5B2GIn3iaKs8/r23HV5F/6phHu1CndaOLxRNg\nva0WLycs8ozGEhORlSLSmrSBfQTn6SjQb936gbuxrqGGVnTna9Wc9AB/u277bHW+g/DvBcbr0BHg\nAjCtqqcrDkWPPUu7jNgl3e08DTwEeq1bddxJPnqBMbXVx4K0i3VaL3X1tagfwqrsDGFuNxBZq5Ow\n8v0CeJXoEeaJD4A3tm0rSO8aYcj7k1DJD2ZpEYaFZy0PL4FtEbQv27kn7WZpr+g/YNqvgf11au8i\nDGknCe7eCfuco8deQzt67MBm4LlpTAEnKu67ccLi6U2g2favsN/f2vHOCNpjFvcUcIWFJyaLzrk7\nOB3HyUWjpyGO4/wneLFwHCcXXiwcx8mFFwvHcXLhxcJxnFx4sXAcJxdeLBzHyYUXC8dxcvEH/Y0k\nz9U7tooAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x2a4858e1320>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(fmaps, cmap='jet', alpha=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 218,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x2a485873f60>"
      ]
     },
     "execution_count": 218,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAQsAAAD8CAYAAABgtYFHAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nCJvKYhOb2MQNYVNZbGITm7gh/H9dmXd7HS370gAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x2a485510550>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.imshow(test_image[0]/255)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "heading_collapsed": true
   },
   "source": [
    "## Plots Weights For Each Class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "hidden": true
   },
   "outputs": [],
   "source": [
    "# Planning to use the weights for each class node in the final output as a vector describing the dog breed.\n",
    "\n",
    "# Using PCA and t-SNE I would like to try to reduce the vectors down to 3 dimensions and plot them.\n",
    "\n",
    "# Hopefully similar looking dog breeds will be grouped together."
   ]
  }
 ],
 "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": "3.5.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}