classification_using_DL.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Deep Learning #\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "jYysdyb-CaWM"
   },
   "source": [
    "# Basic classification: Classify images of clothing"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "FbVhjPpzn6BM"
   },
   "source": [
    "### Training a neural network model to classify images of clothing, like sneakers and shirts."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:03.661948Z",
     "iopub.status.busy": "2022-02-05T02:23:03.661380Z",
     "iopub.status.idle": "2022-02-05T02:23:06.352001Z",
     "shell.execute_reply": "2022-02-05T02:23:06.352451Z"
    },
    "id": "dzLKpmZICaWN"
   },
   "outputs": [],
   "source": [
    "# TensorFlow and tf.keras\n",
    "import tensorflow as tf\n",
    "\n",
    "# Helper libraries\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "yR0EdgrLCaWR"
   },
   "source": [
    "## Import the Fashion MNIST dataset"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "DLdCchMdCaWQ"
   },
   "source": [
    "We will use the `Fashion MNIST` dataset which contains 70,000 grayscale images in 10 categories. The images show individual articles of clothing at low resolution (28 by 28 pixels)\n",
    "\n",
    "Out of 70,000 we will divide  `60,000 images to train` the network and `10,000 images to evaluate` how accurately the network learned to classify images.\n",
    "\n",
    "we can get  the Fashion MNIST directly from TensorFlow. \n",
    "Import and **[load the Fashion MNIST data]** directly from TensorFlow:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:06.357055Z",
     "iopub.status.busy": "2022-02-05T02:23:06.356498Z",
     "iopub.status.idle": "2022-02-05T02:23:08.243268Z",
     "shell.execute_reply": "2022-02-05T02:23:08.243667Z"
    },
    "id": "7MqDQO0KCaWS"
   },
   "outputs": [],
   "source": [
    "fashion_mnist = tf.keras.datasets.fashion_mnist\n",
    "\n",
    "(train_images, train_labels), (test_images, test_labels) = fashion_mnist.load_data()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "t9FDsUlxCaWW"
   },
   "source": [
    "**Dataset contains four NumPy arrays:**\n",
    "\n",
    "* The `train_images` and `train_labels` arrays are the *training set*—the data the model uses to learn.\n",
    "* The model is tested against the *test set*, the `test_images`, and `test_labels` arrays.\n",
    "\n",
    "The images are 28x28 NumPy arrays, with pixel values ranging from 0 to 255. The *`labels`* are an array of integers, ranging from 0 to 9. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:08.248025Z",
     "iopub.status.busy": "2022-02-05T02:23:08.247468Z",
     "iopub.status.idle": "2022-02-05T02:23:08.249506Z",
     "shell.execute_reply": "2022-02-05T02:23:08.249073Z"
    },
    "id": "IjnLH5S2CaWx"
   },
   "outputs": [],
   "source": [
    "class_names = ['T-shirt/top', 'Trouser', 'Pullover', 'Dress', 'Coat',\n",
    "               'Sandal', 'Shirt', 'Sneaker', 'Bag', 'Ankle boot']"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Brm0b_KACaWX"
   },
   "source": [
    "## Explore the data\n",
    "\n",
    "> Let's explore the format of the dataset before training the model.\n",
    "By seeing shape we get to know that  there are 60,000 images in the training set, with each image represented as 28 x 28 pixels:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:08.255326Z",
     "iopub.status.busy": "2022-02-05T02:23:08.254721Z",
     "iopub.status.idle": "2022-02-05T02:23:08.257282Z",
     "shell.execute_reply": "2022-02-05T02:23:08.257646Z"
    },
    "id": "zW5k_xz1CaWX"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(60000, 28, 28)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_images.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[[0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        ...,\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0]],\n",
       "\n",
       "       [[0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        ...,\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0]],\n",
       "\n",
       "       [[0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        ...,\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0]],\n",
       "\n",
       "       ...,\n",
       "\n",
       "       [[0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        ...,\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0]],\n",
       "\n",
       "       [[0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        ...,\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0]],\n",
       "\n",
       "       [[0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        ...,\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0]]], dtype=uint8)"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_images"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "YSlYxFuRCaWk"
   },
   "source": [
    "Each label is an integer between 0 and 9:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:08.267860Z",
     "iopub.status.busy": "2022-02-05T02:23:08.267274Z",
     "iopub.status.idle": "2022-02-05T02:23:08.269783Z",
     "shell.execute_reply": "2022-02-05T02:23:08.270118Z"
    },
    "id": "XKnCTHz4CaWg"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([9, 0, 0, ..., 3, 0, 5], dtype=uint8)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_labels"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "TMPI88iZpO2T"
   },
   "source": [
    "There are 10,000 images in the test set. Again, each image is represented as 28 x 28 pixels:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:08.273959Z",
     "iopub.status.busy": "2022-02-05T02:23:08.273317Z",
     "iopub.status.idle": "2022-02-05T02:23:08.275794Z",
     "shell.execute_reply": "2022-02-05T02:23:08.276131Z"
    },
    "id": "2KFnYlcwCaWl"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(10000, 28, 28)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_images.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "rd0A0Iu0CaWq"
   },
   "source": [
    "And the test set contains 10,000 images labels:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:08.280012Z",
     "iopub.status.busy": "2022-02-05T02:23:08.279383Z",
     "iopub.status.idle": "2022-02-05T02:23:08.282069Z",
     "shell.execute_reply": "2022-02-05T02:23:08.281611Z"
    },
    "id": "iJmPr5-ACaWn"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[[0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        ...,\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0]],\n",
       "\n",
       "       [[0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        ...,\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0]],\n",
       "\n",
       "       [[0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        ...,\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0]],\n",
       "\n",
       "       ...,\n",
       "\n",
       "       [[0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        ...,\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0]],\n",
       "\n",
       "       [[0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        ...,\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0]],\n",
       "\n",
       "       [[0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        ...,\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0],\n",
       "        [0, 0, 0, ..., 0, 0, 0]]], dtype=uint8)"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_images"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([9, 2, 1, ..., 8, 1, 5], dtype=uint8)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_labels"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ES6uQoLKCaWr"
   },
   "source": [
    "## Preprocess the data\n",
    "\n",
    "We must be preprocess this data before training the network,by doing so we can attain few insights\n",
    "> If you inspect the first image in the training set, you will see that the pixel values fall in the range of 0 to 255:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:08.288632Z",
     "iopub.status.busy": "2022-02-05T02:23:08.288086Z",
     "iopub.status.idle": "2022-02-05T02:23:08.491137Z",
     "shell.execute_reply": "2022-02-05T02:23:08.491535Z"
    },
    "id": "m4VEw8Ud9Quh"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure()\n",
    "plt.imshow(train_images[2])\n",
    "plt.colorbar()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Wz7l27Lz9S1P"
   },
   "source": [
    "## scaling\n",
    "\n",
    "- We should scale these values to a range of 0 to 1 before feeding them to the neural network model. \n",
    "- To do so, We  willdivide the values by 255.\n",
    "\n",
    "\n",
    "**We do Same for both train and test pixel data**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:08.495092Z",
     "iopub.status.busy": "2022-02-05T02:23:08.494541Z",
     "iopub.status.idle": "2022-02-05T02:23:08.649262Z",
     "shell.execute_reply": "2022-02-05T02:23:08.649716Z"
    },
    "id": "bW5WzIPlCaWv"
   },
   "outputs": [],
   "source": [
    "train_images = train_images / 255.0\n",
    "\n",
    "test_images = test_images / 255.0"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Ee638AlnCaWz"
   },
   "source": [
    "## Verifying\n",
    "\n",
    "- We must  verify that the data is in the correct format or not.\n",
    "- Then you're ready to build and train the network.\n",
    "- Displaying the first 10 images from the *training set* and display the class name below each image."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:08.719166Z",
     "iopub.status.busy": "2022-02-05T02:23:08.718438Z",
     "iopub.status.idle": "2022-02-05T02:23:09.508634Z",
     "shell.execute_reply": "2022-02-05T02:23:09.509037Z"
    },
    "id": "oZTImqg_CaW1"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 720x720 with 20 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10,10))\n",
    "for i in range(20):\n",
    "    plt.subplot(5,5,i+1)\n",
    "    plt.xticks([])\n",
    "    plt.yticks([])\n",
    "    plt.grid(False)\n",
    "    plt.imshow(train_images[i], cmap=plt.cm.binary)\n",
    "    plt.xlabel(class_names[train_labels[i]])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "59veuiEZCaW4"
   },
   "source": [
    "## Building the model\n",
    "\n",
    "Two steps for building the neural network:\n",
    " - Configuring.(structure)\n",
    " - Compiling .(optimisation)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Gxg1XGm0eOBy"
   },
   "source": [
    "### Set up the layers\n",
    "\n",
    "The basic building block of a neural network is the *layer* \n",
    "\n",
    "**Layers extract representations from the data fed into them.**\n",
    "\n",
    "Most of deep learning consists of chaining together simple layers. \n",
    "\n",
    "Most layers, such as `tf.keras.layers.Dense`, have parameters that are learned during training."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:09.515540Z",
     "iopub.status.busy": "2022-02-05T02:23:09.514954Z",
     "iopub.status.idle": "2022-02-05T02:23:11.016295Z",
     "shell.execute_reply": "2022-02-05T02:23:11.015745Z"
    },
    "id": "9ODch-OFCaW4"
   },
   "outputs": [],
   "source": [
    "model = tf.keras.Sequential([\n",
    "    tf.keras.layers.Flatten(input_shape=(28, 28)),\n",
    "    tf.keras.layers.Dense(128, activation='relu'),\n",
    "    tf.keras.layers.Dense(10)\n",
    "])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "gut8A_7rCaW6"
   },
   "source": [
    "### Interpreting Model\n",
    "\n",
    "1. The first layer in this network, `tf.keras.layers.Flatten`**:**  \n",
    "    - It transforms the format of the images from a two-dimensional array (of 28 by 28 pixels) to a one-dimensional array (of 28 * 28 = 784 pixels). \n",
    "\n",
    "    - We can think of this layer as unstacking rows of pixels in the image and lining them up.\n",
    "\n",
    "    - This layer has no parameters to learn; it only reformats the data.\n",
    "\n",
    "2. After flattening pixels, the network consists of a sequence of two `tf.keras.layers.Dense` layers **:**\n",
    " \n",
    "    **`These are densely connected, or fully connected, neural layers.`** \n",
    "    - The first `Dense` layer has 128 nodes (or neurons).\n",
    "    - The second (and last) layer returns a logits array with length of 10 (**`Each node contains a score that indicates the current image belongs to one of the 10 classes`**).\n",
    "\n",
    "### Compiling model\n",
    "\n",
    "Before the model is ready for training, it needs a few more settings. These are added during the model's [*compile*](https://www.tensorflow.org/api_docs/python/tf/keras/Model#compile) step:\n",
    "\n",
    "* [*Loss function*](https://www.tensorflow.org/api_docs/python/tf/keras/losses) —This measures how accurate the model is during training. You want to minimize this function to \"steer\" the model in the right direction.\n",
    "* [*Optimizer*](https://www.tensorflow.org/api_docs/python/tf/keras/optimizers) —This is how the model is updated based on the data it sees and its loss function.\n",
    "* [*Metrics*](https://www.tensorflow.org/api_docs/python/tf/keras/metrics) —Used to monitor the training and testing steps. The following example uses *accuracy*, the fraction of the images that are correctly classified."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:11.023505Z",
     "iopub.status.busy": "2022-02-05T02:23:11.022906Z",
     "iopub.status.idle": "2022-02-05T02:23:11.028191Z",
     "shell.execute_reply": "2022-02-05T02:23:11.027758Z"
    },
    "id": "Lhan11blCaW7"
   },
   "outputs": [],
   "source": [
    "model.compile(optimizer='adam',\n",
    "              loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),\n",
    "              metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "qKF6uW-BCaW-"
   },
   "source": [
    "## Training the model\n",
    "\n",
    "**`Training the neural network model requires the following steps`**:\n",
    "\n",
    "1. Feeding the training data to the model. Here, the training data is in the `train_images` and `train_labels` arrays.\n",
    "2. The model learns to associate images and labels.\n",
    "3. Then we can make model make predictions about a test set—in this example, the `test_images` array.\n",
    "4. Then verifying that the predictions match the labels from the `test_labels` array.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Z4P4zIV7E28Z"
   },
   "source": [
    "### Feed the model\n",
    "\n",
    "To start training,  call the [`model.fit`](https://www.tensorflow.org/api_docs/python/tf/keras/Model#fit) method—so called because it \"fits\" the model to the training data:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:11.032563Z",
     "iopub.status.busy": "2022-02-05T02:23:11.031996Z",
     "iopub.status.idle": "2022-02-05T02:23:45.127248Z",
     "shell.execute_reply": "2022-02-05T02:23:45.126760Z"
    },
    "id": "xvwvpA64CaW_"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/15\n",
      "1875/1875 [==============================] - 13s 4ms/step - loss: 0.4913 - accuracy: 0.8279\n",
      "Epoch 2/15\n",
      "1875/1875 [==============================] - 7s 4ms/step - loss: 0.3741 - accuracy: 0.8651\n",
      "Epoch 3/15\n",
      "1875/1875 [==============================] - 8s 4ms/step - loss: 0.3354 - accuracy: 0.8760\n",
      "Epoch 4/15\n",
      "1875/1875 [==============================] - 7s 4ms/step - loss: 0.3130 - accuracy: 0.8856\n",
      "Epoch 5/15\n",
      "1875/1875 [==============================] - 7s 4ms/step - loss: 0.2947 - accuracy: 0.8919\n",
      "Epoch 6/15\n",
      "1875/1875 [==============================] - 7s 4ms/step - loss: 0.2802 - accuracy: 0.8960\n",
      "Epoch 7/15\n",
      "1875/1875 [==============================] - 6s 3ms/step - loss: 0.2678 - accuracy: 0.9011\n",
      "Epoch 8/15\n",
      "1875/1875 [==============================] - 6s 3ms/step - loss: 0.2550 - accuracy: 0.9060\n",
      "Epoch 9/15\n",
      "1875/1875 [==============================] - 7s 4ms/step - loss: 0.2475 - accuracy: 0.9078\n",
      "Epoch 10/15\n",
      "1875/1875 [==============================] - 7s 4ms/step - loss: 0.2387 - accuracy: 0.9115\n",
      "Epoch 11/15\n",
      "1875/1875 [==============================] - 7s 4ms/step - loss: 0.2304 - accuracy: 0.9139\n",
      "Epoch 12/15\n",
      "1875/1875 [==============================] - 7s 4ms/step - loss: 0.2244 - accuracy: 0.9155\n",
      "Epoch 13/15\n",
      "1875/1875 [==============================] - 6s 3ms/step - loss: 0.2149 - accuracy: 0.9183\n",
      "Epoch 14/15\n",
      "1875/1875 [==============================] - 6s 3ms/step - loss: 0.2092 - accuracy: 0.9207\n",
      "Epoch 15/15\n",
      "1875/1875 [==============================] - 7s 3ms/step - loss: 0.2027 - accuracy: 0.9245\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x29b63db48e0>"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(train_images, train_labels, epochs=15)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "W3ZVOhugCaXA"
   },
   "source": [
    "* As the model trains, We can observe the loss and accuracy metrics. \n",
    "\n",
    "* This model reaches an accuracy of about 0.94 (or 94%) on the training data."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "wCpr6DGyE28h"
   },
   "source": [
    "### Evaluate accuracy\n",
    "\n",
    "Next, we compare how the model performs on the `test dataset`:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:45.131545Z",
     "iopub.status.busy": "2022-02-05T02:23:45.130989Z",
     "iopub.status.idle": "2022-02-05T02:23:45.773332Z",
     "shell.execute_reply": "2022-02-05T02:23:45.773676Z"
    },
    "id": "VflXLEeECaXC"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "313/313 - 2s - loss: 0.3455 - accuracy: 0.8815 - 2s/epoch - 6ms/step\n",
      "\n",
      "Test accuracy: 0.8815000057220459\n"
     ]
    }
   ],
   "source": [
    "test_loss, test_acc = model.evaluate(test_images,  test_labels, verbose=2)\n",
    "\n",
    "print('\\nTest accuracy:', test_acc)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "yWfgsmVXCaXG"
   },
   "source": [
    " - We can see that the accuracy on the test dataset is a `little less` than the accuracy on the training dataset.\n",
    "\n",
    " - This gap between training accuracy and test accuracy represents **overfitting**.\n",
    " \n",
    " - Overfitting happens when a machine learning model performs worse on new, previously unseen inputs than it does on the training data. \n",
    " \n",
    " - An overfitted model \"memorizes\" the noise and details in the training dataset to a point where it negatively impacts the performance of the model on the new data. \n",
    " \n",
    " \n",
    " For more information, you can see the following:\n",
    "*   [Demonstrate overfitting](https://www.tensorflow.org/tutorials/keras/overfit_and_underfit#demonstrate_overfitting)\n",
    "*   [Strategies to prevent overfitting](https://www.tensorflow.org/tutorials/keras/overfit_and_underfit#strategies_to_prevent_overfitting)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "v-PyD1SYE28q"
   },
   "source": [
    "### Make predictions\n",
    "\n",
    "With the model trained, you can use it to make predictions about some images.\n",
    "Attach a softmax layer to convert the model's linear outputs—\n",
    "\n",
    "[logits](https://developers.google.com/machine-learning/glossary#logits)—to probabilities, which should be easier to interpret."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:45.778940Z",
     "iopub.status.busy": "2022-02-05T02:23:45.778310Z",
     "iopub.status.idle": "2022-02-05T02:23:45.792392Z",
     "shell.execute_reply": "2022-02-05T02:23:45.792727Z"
    },
    "id": "DnfNA0CrQLSD"
   },
   "outputs": [],
   "source": [
    "probability_model = tf.keras.Sequential([model, \n",
    "                                         tf.keras.layers.Softmax()])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:45.796585Z",
     "iopub.status.busy": "2022-02-05T02:23:45.796041Z",
     "iopub.status.idle": "2022-02-05T02:23:46.217066Z",
     "shell.execute_reply": "2022-02-05T02:23:46.217474Z"
    },
    "id": "Gl91RPhdCaXI"
   },
   "outputs": [],
   "source": [
    "predictions = probability_model.predict(test_images)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "x9Kk1voUCaXJ"
   },
   "source": [
    "Here, the model has predicted the label for each image in the testing set. Let's take a look at the first prediction:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:46.222583Z",
     "iopub.status.busy": "2022-02-05T02:23:46.221954Z",
     "iopub.status.idle": "2022-02-05T02:23:46.224500Z",
     "shell.execute_reply": "2022-02-05T02:23:46.224882Z"
    },
    "id": "3DmJEUinCaXK"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([2.0369461e-08, 4.1206804e-12, 1.0320026e-11, 1.0398885e-11,\n",
       "       5.6118195e-09, 2.5575163e-04, 3.7448431e-09, 4.6442347e-03,\n",
       "       5.4625957e-09, 9.9509996e-01], dtype=float32)"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "predictions[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "-hw1hgeSCaXN"
   },
   "source": [
    "A prediction is an array of 10 numbers. They represent the model's \"confidence\" that the image corresponds to each of the 10 different articles of clothing. You can see which label has the highest confidence value:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:46.229202Z",
     "iopub.status.busy": "2022-02-05T02:23:46.228668Z",
     "iopub.status.idle": "2022-02-05T02:23:46.230873Z",
     "shell.execute_reply": "2022-02-05T02:23:46.231256Z"
    },
    "id": "qsqenuPnCaXO"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "9"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.argmax(predictions[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "E51yS7iCCaXO"
   },
   "source": [
    "So, the model is most confident that this image is an ankle boot, or `class_names[9]`. Examining the test label shows that this classification is correct:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:46.235327Z",
     "iopub.status.busy": "2022-02-05T02:23:46.234746Z",
     "iopub.status.idle": "2022-02-05T02:23:46.237338Z",
     "shell.execute_reply": "2022-02-05T02:23:46.237702Z"
    },
    "id": "Sd7Pgsu6CaXP"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "9"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_labels[0]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "ygh2yYC972ne"
   },
   "source": [
    "Graph this to look at the full set of 10 class predictions."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:46.244767Z",
     "iopub.status.busy": "2022-02-05T02:23:46.244181Z",
     "iopub.status.idle": "2022-02-05T02:23:46.246387Z",
     "shell.execute_reply": "2022-02-05T02:23:46.245908Z"
    },
    "id": "DvYmmrpIy6Y1"
   },
   "outputs": [],
   "source": [
    "def plot_image(i, predictions_array, true_label, img):\n",
    "  true_label, img = true_label[i], img[i]\n",
    "  plt.grid(False)\n",
    "  plt.xticks([])\n",
    "  plt.yticks([])\n",
    "\n",
    "  plt.imshow(img, cmap=plt.cm.binary)\n",
    "\n",
    "  predicted_label = np.argmax(predictions_array)\n",
    "  if predicted_label == true_label:\n",
    "    color = 'blue'\n",
    "  else:\n",
    "    color = 'red'\n",
    "\n",
    "  plt.xlabel(\"{} {:2.0f}% ({})\".format(class_names[predicted_label],\n",
    "                                100*np.max(predictions_array),\n",
    "                                class_names[true_label]),\n",
    "                                color=color)\n",
    "\n",
    "def plot_value_array(i, predictions_array, true_label):\n",
    "  true_label = true_label[i]\n",
    "  plt.grid(False)\n",
    "  plt.xticks(range(10))\n",
    "  plt.yticks([])\n",
    "  thisplot = plt.bar(range(10), predictions_array, color=\"#777777\")\n",
    "  plt.ylim([0, 1])\n",
    "  predicted_label = np.argmax(predictions_array)\n",
    "\n",
    "  thisplot[predicted_label].set_color('red')\n",
    "  thisplot[true_label].set_color('blue')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "Zh9yABaME29S"
   },
   "source": [
    "### Verify predictions\n",
    "\n",
    "With the model trained, you can use it to make predictions about some images."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "d4Ov9OFDMmOD"
   },
   "source": [
    "Let's look at the 0th image, predictions, and prediction array. Correct prediction labels are blue and incorrect prediction labels are red. The number gives the percentage (out of 100) for the predicted label."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:46.285608Z",
     "iopub.status.busy": "2022-02-05T02:23:46.285071Z",
     "iopub.status.idle": "2022-02-05T02:23:46.364057Z",
     "shell.execute_reply": "2022-02-05T02:23:46.364410Z"
    },
    "id": "HV5jw-5HwSmO"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x216 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "i = 0\n",
    "plt.figure(figsize=(6,3))\n",
    "plt.subplot(1,2,1)\n",
    "plot_image(i, predictions[i], test_labels, test_images)\n",
    "plt.subplot(1,2,2)\n",
    "plot_value_array(i, predictions[i],  test_labels)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:46.395590Z",
     "iopub.status.busy": "2022-02-05T02:23:46.386856Z",
     "iopub.status.idle": "2022-02-05T02:23:46.477744Z",
     "shell.execute_reply": "2022-02-05T02:23:46.478119Z"
    },
    "id": "Ko-uzOufSCSe"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x216 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "i = 12\n",
    "plt.figure(figsize=(6,3))\n",
    "plt.subplot(1,2,1)\n",
    "plot_image(i, predictions[i], test_labels, test_images)\n",
    "plt.subplot(1,2,2)\n",
    "plot_value_array(i, predictions[i],  test_labels)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "kgdvGD52CaXR"
   },
   "source": [
    "Let's plot several images with their predictions. Note that the model can be wrong even when very confident."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:46.495208Z",
     "iopub.status.busy": "2022-02-05T02:23:46.494656Z",
     "iopub.status.idle": "2022-02-05T02:23:48.120493Z",
     "shell.execute_reply": "2022-02-05T02:23:48.120884Z"
    },
    "id": "hQlnbqaw2Qu_"
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA2AAAALICAYAAAAZqpOTAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAC7Y0lEQVR4nOzdd7wcVf3/8feH9N4boaQTIKQR6b0JiBRFQZRm/2IBLD/4WgD1q6LYRUAExAZIB5XeQUJJ74FUCOm99/P7Y+Ymez67d3dvcu/cktfz8biP3M/s2Zmzm53PnbMznzkWQhAAAAAAoObtVdsdAAAAAIA9BQMwAAAAAMgIAzAAAAAAyAgDMAAAAADICAMwAAAAAMhI4yw20rlz59CrV68sNoUaNGfOHC1dutSqa318LjB69OilIYQu1bnO+vC52r59exSvW7cuitu0abPb21i/fn0U77VX/H1b8+bNd3sbNam6841UPz4bqFnVnXPqw2dqzZo1Ubxo0aIobtmyZRRv2bIlips1a5a3Tp/Dtm3bVrQPmzdvjuK+ffsWbZ+1upxvxo+Xtm4t3a5xY2nIkN3eHKpRsXyTyQCsV69eGjVqVBabQg0aMWJEta6Pz0XD0L275P6e5+nWTVq4MH+5mc2t7v7Uh8+VPyB66623ovjkk0/e7W2MGTMmilu3bh3FAwYM2O1t1KTqzjdS7X02dmcfQfWq7pxTF/KNn07ILB5HPP/881H8u9/9LoqHDh0axQvdB7Ffv35521y7dm0Ur1ixIoobN44PL2fPnh3FjzzySN46a1NdzjdW5rBw61apjv/p2+MUyzdcgghgt5Q6sCy3DdBQsY8AAHJlcgYMABqyjRs3RvFvfvObKL733nuj2H9bvGTJkihu0aJF0fbl8JcY+th/Q33cccdF8Re+8IUoPv3006vcBwA1r9QZsOuvvz6K//vf/0bx448/XnT9bdu2zVvmL3He6q6R8zlsw4YNUfzvf/87is8666yifQAaGs6AAQAAAEBGGIABAAAAQEYYgAEAAABARqgBA4Aquuaaa6L49ttvj+LVq1dHsb/Ns6+P6NChQxT7eolWrVpFcaFbPvtbRftt+DqRTZs2RfF//vOfKPZ1IUceeWQUv/LKK3l9AJA9P8WEN378+Cj2+aZLl/gu2X5ajEL5pmPHjlHcpEmTKPb5ZsaMGVE8bdq0KKYGDHsazoABAAAAQEYYgAEAAABARhiAAQAAAEBGqAEDgBJ8jdfPf/7zKO7evXsU+5otPy+Pr4/YsmVLFJeaw8uvT8qvA/Hz8nh+na1bt47iRo0aRbGfO+ijH/1oFP/rX/8quj0AtWPt2rVR3Llz5yj2Navbt2+PYl9fWqiN30ah5+R6//33iz4ONHScAQMAAACAjDAAAwAAAICMMAADAAAAgIxQAwYAJXz/+9+P4rZt20axr8ny8+YsXLiw6Prbt28fxb4+q3HjOFX7egtJ2rhxYxR36tSpaJ/8Ov28YL5OrVu3blHs5wFbunRpXp98rQmAmrdo0aKij/t9v1BNaa5C9aR+3i9fM+rX6XPm4sWLi24TaOg4AwYAAAAAGWEABgAAAAAZYQAGAAAAABlhAAYAAAAAGeEmHABQwqpVq6LYTzLqb1jhb7rxP//zP1H8pS99KYqHDx8exX4i53nz5kVxmzZt8vq4//77R7EvxPd99uvs2bNn0fZr1qyJ4g0bNkTxrFmz8vrETTiA7E2aNKno402bNo1ivy/7G2r4m3ZI+RMx+xxYajLnQjftAfYknAEDAAAAgIwwAAMAAACAjDAAAwAAAICMUAMGACX4SYr9RMm+/sH76U9/GsXt2rWLYl8vsX79+ig+4YQTovjFF18suj1JOvDAA6N42rRpUbx69eoo/u1vfxvFfvLpLl26RLGf2Pm1117L68Nhhx1Wsp8Aqtf48eOj2Nd8+fzl842f1N3XwEr5E737iZd9TvQ51Ne5AnsazoABAAAAQEYYgAEAAABARhiAAQAAAEBGqAGrw3yNxV57xeNlf82156+5lvLn4nj33XejuH///lXpItDgbN68uWQbv+8V2tdyXXLJJVH82GOPFW2/YsWKKPY1X9ddd13ec9q2bRvF9913XxQvX748iufOnRvFF1xwQRT7GjCfj/zcQOPGjcvrE4Dsvf3221Hsjx18zZffl33Nl5+nUMrf3zt06BDF/ljDb3PffffNWyewJ+EMGAAAAABkhAEYAAAAAGSEARgAAAAAZIQasF3k57goNA+Qv+76gw8+iOKRI0dG8RlnnBHFuztPhr8Gu5CHH344iq+55prd2iZQ382fP79kG79vb9iwoWj7efPmVakPDzzwQNHHL7744rxlLVq0iGJfszVkyJAoXrBgQRS3bt26Kl3M4+tJAdSOqVOnRnGTJk2i2OevtWvXRnGPHj2i+I033sjbhq+D9XMZ+njr1q1R3LFjx7x1AnsSzoABAAAAQEYYgAEAAABARhiAAQAAAEBGqAGrJv6a6kJeffXVKH7zzTej2NeefP3rX9+tPi1evDhv2dNPPx3Fbdq02a1tAA3NkiVLqvwcX9/gay78vu3rI7zjjz++6OMf/vCH85bNnj07in2NxZNPPhnFJ5xwQhT7GjFfE+b73KhRoyheuHBh5R0GkBk/j5ffV0vVgH3sYx+r8jZ9DmzZsmXR9uXMtwg0ZJwBAwAAAICMMAADAAAAgIwwAAMAAACAjFADtov8HDuNG+e/lW+//XYU+7k5unXrFsV+Hp3zzjsvijt06BDFGzdujOL9998/ipctW5bXp9WrV0dxz54989oAezI/X18hheb9y+XrH3x9lK/B8OubPn16FPv5+WbNmlWyjwceeGAUT5s2LYrfe++9KL7lllui2M/94/OPn2ewnPcNQM1btGhRFFd1TtFPfepTJdv4/X/58uVR3Llz56LPX79+fZX6BDQ0nAEDAAAAgIwwAAMAAACAjDAAAwAAAICMUANWJj8Hjq/5WrduXd5zHnzwwSj210z7Gq41a9ZEsa8LKRVPnjw5ivfZZ5+8Pvk6Dl/LBuzpypkHzM+r4+fA8bGfU+s73/lO0fbPPPNMFI8fPz6K/b4u5dd3+povX0d2wQUXRPG4cePy1pnL50Azi+ItW7YUfT6AbGzYsCGK/Xyfpf7un3jiiSW3ceSRR0bxyJEjo9jnNK9Tp04ltwE0ZJwBAwAAAICMMAADAAAAgIwwAAMAAACAjDSYGjBfD+XrE0rVL/jYXyPtaz682267LW+Zn+erefPmUTx37two9jVh/vn+mmrfZz/Xh685k6RVq1ZF8aZNm6LY17JVdf4QoL5bsGBByTal5vHy+2q7du2i+Kc//WnR9fv2PhdMmTKlZB+7d+8exUuXLo1in49KKWfuw1LPKZVHAdQ8X6/p9+VCxw5er169ovi1116L4lJzJfocB+xpOAMGAAAAABlhAAYAAAAAGWEABgAAAAAZqRc1YIWuJS5Vw+X5mg2vqrUK9957bxQvXLgwr82wYcOi2NeFrFy5Moo7duwYxX6eDF/DsXbt2qLrL8S/l+vXr4/id999N4qHDh1acp1AQ1LOPGBe06ZNo/ikk06K4ldffTWK/Rx9Pt/42kyfn/y8YoX4fODryPw2/Drbt28fxX6eMJ+vCpkzZ04U9+3bt+RzAFQvf3y0efPmKN6V/dLnMJ+jSh2TAXs6zoABAAAAQEYYgAEAAABARhiAAQAAAEBG6kUNWDnXEvt5vnzsayz8OkvVfN11111R/M4770Txvvvum/ecZcuWRbGvv9qwYUMU9+zZM4rXrFkTxb7PLVu2jGI/j1g5tXPe008/HcXUgGFP42szC/H7pt93L7vssih+8skno9jvu16pfFYOv6/7mjBfA+bnAvrYxz4Wxb4GrBy+bpUaMCB7ft/2830efPDBVV7nmWeeGcU///nPo3hXchawJ+EMGAAAAABkhAEYAAAAAGSEARgAAAAAZIQBGAAAAABkpE7chKNUsWahG0f4G0z4iZZLTbzszZ8/P4offvjhKPY3zOjfv38U+0mRpfwid39TjiZNmkSxf51+kmTPv8ZmzZoVfVySWrVqVXSb//3vf4tuE2jo/H5aiM8HXbt2jeIOHToUfb7f9/0NMvx+WdV8VmgdpSZK9fnq8MMPL7p+36fmzZvntaEQH6h9ft/3x099+vSp8jqHDBkSxX5yZ5/TPH8sAuxpOAMGAAAAABlhAAYAAAAAGWEABgAAAAAZyawGLPcaZD/pcXXUN3hLliyJ4jlz5kTx9OnTo3jBggVR3LRp0yhu27ZtFPvJWlevXp3Xhy1btkSxr7Hw74Pvo7+Gun379kX7WOo6b0lq0aJF0ee0bt06iidNmrTjd1/3AjREhSZi9vVNftJzX88wderUotvwE6P6XOGVMxm95/d/vw4f+9ddapu+vqtQvvETMQOoefvss08U+4mX/THX3nvvXeVt+BzmUQMGFMcZMAAAAADICAMwAAAAAMgIAzAAAAAAyEhmNWC+3inXokWLonju3LlR7K9fLrTM1yfNnj07iv2cWv765TZt2kSxr29YtWpV0e0Vuh7ab9PXX/l5u/w8Gj169IhiX2fm1+/nHio0N9ny5cuj2Nd8LVy4sNL2vl4MaIh2Ze6qAw44IIpnzpxZtL2vr/Lb9I8Xqq8qxa/D12T4/OO34ec288qpAfO1uABqnt93Z82aFcU+F7zzzjtV3oavQfdK1YiVmucUaOg4AwYAAAAAGWEABgAAAAAZYQAGAAAAABnJrAYs13PPPRfF8+fPj2J/7XChOgJfj+RrzErVePn6KF/75OsZ/Bxevt6qUN2I34bvs58Hw9dj+Xm/qlpP4fso5c//4WvZfB1aqeu4gYam0JxcpfYDXwP28ssvF21fao4cn398filn7kS/Dh8Xq8uV8ucS8nE5c3wVqkMFULMOO+ywKPbzEvr6z3HjxlV7H/wxk+f7AOxpOAMGAAAAABlhAAYAAAAAGWEABgAAAAAZyaTAZ/Xq1XrmmWd2xHfeeWf0+MCBA6PYz3/l67ek/JoIPyeFr7fy9Q9+nb72yddYrFmzpuj6fC2VlD8Pj++Trzvz86FNmTKlaB9Lzcvla8qk/PnTmjdvXvQ5ufOJNGnSpOj2gIbAz9cnla6X8vli2rRpUez3nZqYU8+v0+cfH5eqa5sxY0YUd+/ePYp9/iqUH5jrB8jecccdF8V//vOfo9gfL40dO3a3t+lzYKk613LqWIGGjD0AAAAAADLCAAwAAAAAMsIADAAAAAAykkkNWKtWraJ5Kd54443o8YkTJ0bxa6+9VnKdvt7A13R17NixaNyuXbso9vVVvsZr2bJlUTx9+vQoLlTrsHr16ij2NRjjx4+P4sGDB0dxr169ovjZZ5+NYj/PRjnXVPu6j7333juK27ZtG8W5tW81UbcC1DWFaqNKffb93GHLly+P4pYtW0ZxoXkDq8LnknL4OrZSNRqPPfZYFPt8NGbMmCgulH9WrFhRhR4CqA5HHXVUFPtab58Lcmu9d5U/dvDHUN7u5kCgvuMMGAAAAABkhAEYAAAAAGSEARgAAAAAZCSTGrBGjRqpffv2O+LrrruuaPu1a9dG8ZtvvpnXxtdgvf7661E8Z86cKJ4wYUIU+/mw/PXKvsbC1zf4mrJDDjkkr4+nnHJKFJ955plR7K/LLuXss8+O4vfeey+KO3XqFMX+mmwpv1bO17s0a9YsigcMGLDj96r2F6iPCtUybdy4sehz/Lxfvj7T71e+ZszXZJSqnyj0eKkc5pWqwfA51NeoPvjggyW3518ngJq3//77R7E/FvD5yee3WbNmRXGfPn1KbtPX5Zfa96kpx56OM2AAAAAAkBEGYAAAAACQEQZgAAAAAJCRTGrAqqp169ZRfPLJJ+e18cuuuOKKGu1TXfD444/X6vbLmWcMqO98vZZUul7Kz3flayr8On3Nl+f3NR8Xqrfyy3zsa8R87OdGHDlyZBTn1oMWUqhPGzZsKPocADXP13z5OQD9PKi7UgPWo0ePKPY1pB06dIhiasCwp+OIGgAAAAAywgAMAAAAADLCAAwAAAAAMlIna8AAoLb4+WwkqWXLllHs5yr8xje+EcXPPfdcFPtaqKrWU5aq75JKzx3m69h8H1atWhXFJ5xwQhSfddZZUfyDH/wgigvVtfnaEwDVr9QcgOedd14U33PPPVHsc8Nrr70WxX5O00J8jizVR18TBuxpOAMGAAAAABlhAAYAAAAAGWEABgAAAAAZYQAGAAAAABnhJhwAkGPdunV5y/wNJvyNOrZs2RLFXbp0ieJ33303iv3EpqUmet4VpQrz/Wvwk0l37do1ijt37lx0e4VuLDJ37tyS/QSwe0rt6+ecc04U/+Uvf4nipk2bRvFDDz0UxTfccEPJPviJlUvdOKjQhPfAnoQzYAAAAACQEQZgAAAAAJARBmAAAAAAkBFqwAAgx9FHH523bOTIkVHcvHnzKB4wYEAUv/POO9XfsVo2a9asKG7Tpk0UF5p0+bDDDqvRPgEoPcn6GWecEcV+EmS/71Z1onhJGjRoUBRPnDgxin3OXLBgQZW3ATQknAEDAAAAgIwwAAMAAACAjDAAAwAAAICMUAMGADkK1S1t2LAhiv28ObtSM1Hf+LnOfN3I5s2b857TqlWrGu0TgPx5CkvZf//9o/iNN96I4vXr10fx66+/nreOo446Kor9PGAbN26MYp8fli5dWl5ngQaq4R81AAAAAEAdwQAMAAAAADLCAAwAAAAAMkINGADk6NmzZ96yYcOGRbGf06ZUrdPWrVuj2NdshBCq0sUa4fvg+9ivX78o/shHPhLFK1euzFvnkUceWT2dA1ApM6tS+y984QtRPHDgwCi+8MILo9jXexVy8cUXR/GqVauiuHXr1lF87LHHllwn0JBxBgwAAAAAMsIADAAAAAAywgAMAAAAADJiWdQemNkSSXNrfEOoafuHELpU18qq+LnoLKkqE4fQvn60r9bPlES+aUBq87NRV/YP2ld/+9r8O4a6q7b/FtWlfaS+t69Lfan0c5XJAAzYXWY2KoQwgvYNsz1Ql9S1/YP21dseqGvq2j5Sn9vXpb4UwyWIAAAAAJARBmAAAAAAkBEGYKgvbqd9g24P1CV1bf+gffW2B+qauraP1Of2dakvlaIGDAAAAAAyUtYZMDOda6ZgpoGlW0tmmmOmzgWWr61K56ravsh6LjPT3pU89gkzTTbTdjONcI/9r5lmmGm6mT6cs/z0dNkMM12bs/wfZppgpp/kLPuemc4t0rdhZrrTLXvUTG+U+dpOMNO/K3nNN5ezjl1pX2Q97c10RU7cxUxP7e56gWLMtM1M48w0yUwPmKllifYvVezvleWrmmKmr6a5I+Ru10xmpt+lj00w0/Ccxy4107vpz6XpsmZmeip9zbn73O25zy2w/XPNdF36+w1m+iDnvTu7RN935JvqyhkltneWmX5Yk9sAdoeZOqX7zzgzLczZn8aZqWkd6F+dOMYx0+U578tmM01Mf7+xml9y2Tg+QW0q9xLET0l6Lf23PrpMKjwAkzRJ0sckvZK70EwHSbpQ0sGSTpd0i5kamamRpD9IOkPSQZI+ZaaDzDRY0oYQNFjSh8zUzkw9JB0egh4t0rfvSPpdznbbSzpUUjsz9anqC60D2ks7DwZD0BJJC8x0dK31CHuCDSFoaAgaJGmzpC/XdoekHYMqn2f/K+kU5d+e+AxJ/dOfL0q6NV1HR0nXSzpc0mGSrjdTB0kfVpKXB0u6OG07RFKjEDSmSLf+n6RbcuJfh6Chkj4h6a4C/a0VZmos6T+SPlpqQA3UlhC0LM09QyXdpnR/Sn82p5/jzKTHKLnqxDFOCPpzzvs0X9KJaZw7wPN9rzFmaszxCWpTyT+0Zmot6RhJn1Oys1YsPyH9FvlBM01Lvxkx99wWZnrSTF8osN5vm+nt9NuUHxTZ/q/Tb2+eN1OXdNlQM72RPveR9GCk4HIznS9phKR/pN+2tMhdfwiaGoKmF9j0OZLuC0GbQtBsSTOUHPwcJmlGCJoVgjZLui9tu0VSi/TgpYmkbZJ+qOTAqbLX1kbS4BA0PmfxxyT9K11v7vt9tyXfjr9uplnp6/Lr+5CZxpqpr1vexUwPpe/320WSzb7p/+m7Zjv7baZvpN+OTzLTVSWW3yipb/pe35Que1TSpyt7H4oxs9PNbLqZzTCza8tof5eZLTazSWW03dfMXjSzKWY22cyuLNG+uZm9ZWbj0/aVfm7d8xqZ2VgzyztTWaDtHDObaGbjzGxUGe3bm9mDZjbNzKaa2ZFF2h6QrrfiZ7WZXVVi/Venr3WSmd1rZs1LtL8ybTu51Lpr0KuS+qU5asd7bqabzXRZsScW+kyb6UYzfSWnzQ1m+lb6e14eM1Ov9Nvjvyo5+Nk3dxshaGwImlNg8+dI+msICiHoDUnt0wOcD0t6NgQtD0ErJD2r5IBpi6SWSvJNRe79kaTvF3l9AyRtCiF/DpMQNFXSVkmdLT5D2NmsYH9z19vLTC+k78PzZtovPUCbWzGgM1MrM71vpiZm6mvJ2bvRZnrV0qsr0jx3m5nelPTzEBQkvSTprGLbr05VyTlVyTdpe3JONeacOpJv8vjPcZFjloL7mZkONtNb6d/RCWbqny7/TM7yP1o6YDHTWjP90kzjJUX/H3XwGMe3ifpeSQ7uZaZJOc/5lpluSH//upmmpO/TfemyVma6K32vxprpnHT5ZWZ63EwvSHo+Xd2j2sXjk+rAMU7J9nUm36Ttqy/nhPSvfWU/Uvi0FO5Mf39dCoemv58ghVVS2EcKe0lhpBSOSR+bI4VeUnhOCpfkrGtt+u9pUrhdCpY+999SOK7AtoMUPp3+fp0Ubk5/nyCF49PffyiF35RY/pIURpR4nVEbKdwshc/kxHdK4fz0546c5Rfn9Os3UhgnhW9KYWjF+1ZkmydK4SG37FkpHCuFAVKYmLP8bik8kL5fB0lhRs7/w7+lcJQURkthv3T5ZTn9uifn/2Y/KUwt0JfLpLBACp2k0EIKk6QwQgqHSmGiFFpJobUUJkthWJHlvaQwya27Z+5rKfdHUiNJMyX1kdRU0nhJB5V4znGShkuaVMb6e0ganv7eRtI7xdav5CC3dfp7E0lvSjqijO18Q9I9kv5dRts5kjpX4T36i6TPp783ldS+Cu/tQiWTBFbWpqek2ZJapPH9ki4r0n6QkgFHS0mNJT0nqV9V/9935ScntzSWwmNS+J+KfSOnzc1SuCz9fcf+nuarzkU+08Ok8HLOeqZIYd/K8li6D2yXQtHPRsV2c+J/V+ynafx8ug9+Swrfy1n+/XRZ43TfHiuFi6RwthRuKLHNy6Xwy5z4Bil8K/39cCnMT19P7vvTWQpz0t93vKcux/xLCpemv39WCo+mvz8mhRPT3y+oyJ3pa+ufs90X0t/vTt+HRjl9/LQUfp/N56hqOacq+SZtT86pppxTm/mm8v4n+5P/HKuMYxO3n/1eO499mqZ/kw9M97Mm6fJblB5fSSFI4ZMl+rZjW2lcK8c46fIduS+370VycC/lHFek7/EN6e/zpdAs/b19+u9PKl6bFNpL4Z10nZdJYZ4UOuasa5eOT6rn88IxThnt60S+SR+v1pxTzqUmn1LyDYjSf3MvQ3wrBM0LQdsljZPUK+exxyT9OQT9tcA6T0t/xkoaI2mgkstuvO2S/pn+/ndJx5ipXfIfoJfT5X+RdFxly8t4fdUmBF0VklPqv1T6TbSZvmum+63AWUAlO8eSisBM3ZS8D6+FoHckbTHToJz2j4ag7SFoiqRuOcsPVHJXlo+GoPcKbOcUSTebaZykxyW1teTMpvdsSC6n2CDpYSVnPo+R9EgIWheC1qbLjy2yvJDFqvwS0GLSb+LCrBBC7jdxlQohvCJpeTkrDyEsCCGMSX9fI2mqkh2ysvYhhFBRl9gk/QnFtmFm+0j6iKQ7yulTVZhZOyWf8TvT/m0OIaws8+knS5oZQvCXwXmNJbUws8ZKks78Im0PlPRmCGF9CGGrpJeVnNHNQov08z1K0ntSXFdZpoKf6RA0VlJXM+1tySV+K0LQ+yqex+aGUF4d564KQVtD0EUhaJikByRdJemXZvqVJVcmFKrninJO6ur0vfuFpAtCKP6ZrsSRSv4AS9LflLyXUpK/L0h/v1DSP9Pcc5SkB9Lt/jHtV4UHQtC2nHhX88euqFLOqUq+SduTc6ov59RmvinHAyFo2y4em4yU9B0zXSNp//Rv8slKyhPeTvebk6UdZQrbJD1U3S/A291jnErk9r0qxxUVJii5wukzSs7gS0levjZ9n16S1FzSfuljz4YQ7bNZ5hePY5zi665L+Uaq5pxTdABmSe3BSZLuSE+Nf1vSJ23npYabcppvk6Jrnf8r6XRzlyVWrFrST8PO66T7hVDWAdOuHBjsqg8UXzq0T7qssuU7pKe7R0tqLalvCPqkpPMtv45hg5LEUOGTkjpImp2+370UD3hz3+/c93WBpI2ShlXyWvaSdETO+90zTW6ef3+r6/1uruS1VlVPSe/nxPNUJHnsDjPrpeT9e7NEu0ZmNk5J0n42hFC0vaTfKKm52V5mV4KkZ8xstJl9sUTb3kr+uP05Pf1/h5m1KnM7F0q6t2hHQvhAyUH5e0o+Y6tCCM8UecokSceaWSczaynpTLnL72rQhpzP99dCcunMVsU5ruilBSU8IOl8JYOJii+FiuWxdbuwjV3OOUrqLv8q6QhJq9J+frPANnzOkXbWrBwbgl5Nl+W+d7vzvj2u5O9ARyUHjy+k612Z874NDUEH5jzHv3e7mj92BTmnuLqUc2oz35SjnBxQcD8LQfdIOlvJ5/4JM52kJN/8JWefOSCE5DI8SRvdlxblqI1jnELK6XuxXP4RJTVrw5UMThsrea8+nvNe7ReSS6yl2s0vHvmmuLqUb6RqzjmlzoCdL+lvIWj/ENQrBO2r5HRdqW8kJOk6SSuU7Bje05I+W3EWxkw9zdS1kv5V1DpdpOTM0CpJK8x29OFiSS9Xtjz9fY2S069V8bikCy2501hvJd9svyXpbUn9zdTbkjscXZi2Vfpamij5Jvrnklpo5yCmkZR3R6SpkvrlxJ+SdHr6XvdScsByoUpbqSQJ/dRMJxR4/BlJX8vp49BK1nOqmTpaUid3rpJB9KuSzjVTSzO1knReuqyy5YXe6wHSzuu36xoza63kG7irQgiri7UNIWwLIQxV8kfpMDMbVFlbMztL0uIQwugqdOeYEMJwJQXQXzGzYt+UNlbyR+fWEMIwJX9YyrmGvKmSP+4PlGjXQcm3cb2VfEPYysw+U1n7EMJUST9T8nl7SslZ8aoeFFSnuZIOSvfh9kq+ESumss+0lAy6LlSSjyret3LzWLkel3SJJTfuOELSqhC0IN3OaZbUtHZQ8u3u0xVPSpedpWQA1lLJH8IgxfWuKZ9zKjNHSf6RlF9vWsDr2pmrPq30fUu/6Hlb0m8l/TsEbQtBq5V8yfSJtP+WnlmsTJ3OH7uCnFNpu7JzTh3MNwWVODaZowL7mSU34JoVgn6n5GqiwUpqls6vyDHp3+r9d6NrtXGMU0plOXiRkqsQOpmpmdKa0LQebd8Q9KKkayS1UzIofFrS1ypOAJhV+uW01ADzi0e+qbRdrR7jlBqAfUrSI27ZQyr/bohXKrk06Oe5C0PQM0ouVxlppomSHlThAdI6SYelxZcnSTtuR3yppJvMNEHS0DKW3y3pNitwEw4znWemeUouofmPWXJgE4ImK7kedIqSN/or6cHDVklfVbKDT5V0f9q2wleUfEu1Xsmp8Zbpaxwdgla692GakrsdtjFTL0n7SzsvWwpJYewqMx1e4L2JhKBFSpLSHwq0/7qkEWmR6hRVfoe4t5T8/06Q9FAIGhWSu6ndnT72pqQ7QnITgcqWL5P037SAtuImHCcquZtZVZXzzf9uMbMmSl7zP0IID5f7vPQ0+ItKboZQmaMlnW1mc5RcWnCSmf29xHo/SP9drGTfO6xI83mS5uV8Q/WgVPntx3OcIWlMCGFRiXanSJodQlgSQtii5HKQo0r0/84QwqEhhOOUfAHzThn9qRHpZYL3K/njer+SSwWLtS/4mU4fm6wkR32QDoqqksciadH4PCWf5wlmOy7deELSLCXF8H9SejfR9HKZHyk5MHpb0g/dJTTXSfpxein400q+IJuo5FJA7xVJwyq5MiHXLyT9j5nGSmXdov9rki5Pc+/FSnJ/hX9K+ox2njmUkkHa5ywpvJ+s4pfd7Gr+2BXknHqUc+pSvimhsmOTyvazT0qalF5CN0jJzXmmSPqepGfS9Tyr+NLdgurKMU4Z71GlOTgEbVHynr2Vvu5p6VMaSfp7uv2xSu62uFJJvmyiJL9OTuPKZJlfPPJNPco3ab+rL+dUFIPxUzs/UrhaCp+v7X7U8Gt8RQodqv48NVZyQNpbOwtUDy7jeb1UXoGqKTlr8Jsy+9NFaQGokm/+XpV0VpnPPUElClQltZLUJuf31yWdXuI5r0o6IP39Bkk3ldGX+yRdXka7w5UcHLdM36u/SPpaied0Tf/dT8kfyfa1/fnjx/8fhd9K4ZTa7keZfe0mheez217Vc065+SZtS84p3q5KOYd8U/d/6voxzq4en1TPtjnGqU/5Jn1OteWcTOenQEG3Kpl/p0GyZOqAX4Xk9tlVEkLYamYV38Q1knRXCGFyseeY2b1KEkFnM5sn6foQQmX1hUcr+bZ+YnrNsyR9J4TwRCXte0j6i5k1UnL2+P4QQsnbrlZBN0mPmJmUJOZ7QgilJon8mqR/pKfcZ0m6vFjj9PrpUyV9qVRnQghvmtmDSm4wsVXJN4y3l3jaQ2bWSckti78Syi+YRXZ+IpU+q15H7KfCtWw1oqo5p4r5RiLnFLULOYd8U/fV2WOc3Tk+qQ4c49S7fCNVY86xdCQHAAAAAKhh5dyGHgAAAABQDRiAAQAAAEBGGIABAAAAQEYYgAEAAABARhiAAQAAAEBGGIABAAAAQEYYgAEAAABARhiAAQAAAEBGGIABAAAAQEYYgAEAAABARhiAAQAAAEBGGmexkc6dO4devXplsamybd26NYqXLFkSxY0aNYrivfYqPlb17csRQojixo3j/442bdpEsZlVeRvVac6cOVq6dGm1daK6Phfjx0vuv7Ogxo2lIUN2e3OoRqNHj14aQuhSneusi/mmlHXr1kXx9u3bi8bl8M9p0qRJFLdu3brK68xSdecbqX5+NlC9qjvn1IfP1PTp06PYH0v42B+bNG3aNG+dvs2WLVuiuNQxk39+//79i7avaXU533CMU38VyzeZDMB69eqlUaNGZbGpsvkB1x//+Mcobt++fRS3aNGi6PratWuXt8wntW3btkXx5s2bo7hr165RfMIJJ0RxoSSYpREjRlTr+qrrc1HuuHTrVqmOfQz3eGY2t7rXWdXPlR+oFDpw8AcL3u5+OTJy5MgoXr9+fRT7XOFzSSGbNm2K4i5d4r8Bxx13XFW6mLnqzjdS3fxbhGxVd86pD58pfyzhvzBu1qxZFG/cuDGKCw0ifJtFixZFsf8C2ecsHz/xxBN528hSXc43HOPUX8XyDZcgAgAAAEBGMjkDVhc98MADUfx///d/UdyhQ4co7tGjRxTPnj07inv27Jm3jQEDBkTx1KlTo7h58+ZRfMopp0Sx/0bp4osvztsGgN1T6vKbcp7jrVmzJopfeOGFKB4zZkwUP/nkk1F8wAEHFN3e2rVr87a5bNmyKO7UqVMU+2+sf/zjH0fxRz/60Sg+++yzo3i//fbL2yaAumf16tVRPHny5Cj2Z8O9DRs2RPHMmTPz2vjjF3/lQMuWLaPYn8Uv1QegoeMMGAAAAABkhAEYAAAAAGSEARgAAAAAZGSPrQHzd0H0d/kpdQvV7t27R3Ghu5L5moxVq1ZFcdu2baP4gw8+iOKBAwcW7QOA3VdODVipmq/bb789iv1tn/2dFv2+fcEFF0TxuHHjotjfpcxPoyHl1435u5C1atUqin0OnDs3vlnT1VdfXfT5N954YxTvvffeeX0CkD1f71nqjsz+Dss+9jXxhdbh6878MZQ/xip1Z2mgoeMMGAAAAABkhAEYAAAAAGSEARgAAAAAZGSPrQHz9Vl+Tgo/70XHjh2j2M/z4+stJGnlypVR7GtLSl0zfcghh+StE0D18vtlqXovSbrllluiePny5VHcu3fvKG7SpEkU+/qJrl27RvHxxx8fxQ8//HAU+xpUKb9uo1Q+8XOP9e/fP4rbtWsXxb5G7Hvf+14U33XXXXl9ApC9hx56KIr98c4+++wTxT4f+ZpVX4NaqI2fO8zXqfoa+Pnz50fx6NGjo/jQQw/N2ybQkHAGDAAAAAAywgAMAAAAADLCAAwAAAAAMrLH1oDtv//+UTx+/PgobtSoUdHYz4nj6y+k/Ouqfd3GihUrorjUXEEAql85NWDvv/9+0bhPnz5RvHbt2qLb9Plj0aJFUdy3b9+i8bvvvpu3Tl+nevjhh0fxK6+8EsV+3i4/d9D69euj2M/bs3Dhwij+29/+lteniy++OIp3pd4OQNXccccdUdyjR48o9jWnPv80bhwfGvp8J0ktW7aMYn+M1Lx586LrXLx4cRS/9dZbUUwNGBo6zoABAAAAQEYYgAEAAABARhiAAQAAAEBGGIABAAAAQEb22Jtw+OJvP0mpL5L3xeN+omZ/Qw0p/6YaAwYMKNonX2jvi1YBVD8/YXEhM2bMiGJfcO4nHW3dunUUb9q0KYr9DXp8ez+J+xlnnBHFr732Wl4f/U0yfJ987G8KtG7duij2k81v3rw5iv3krGPHjs3rk78JBzfdAGre9OnTo3jEiBFR7CdN3rJlSxT7Yxefn6T8fODzi5/I3cc+7/qJmYGGjjNgAAAAAJARBmAAAAAAkBEGYAAAAACQkT22yMhff7zvvvtG8UEHHRTFvnbhgQceiOLly5fnbWPy5MlRfNxxx0Wxn2iwZ8+eUeyvsfYTHwLIht+X/SSjvsbL14z6fdfXWPiastWrV0exn0j1tNNOy+tjqcnj+/XrV7SPfmJlX9PhJ2r2/ESqALKxYMGCKPY1pn7iZT8Jsj8eatq0aRQXmojZ50BfJ+brynw+8c/3NaVAQ8cZMAAAAADICAMwAAAAAMgIAzAAAAAAyMgeWwN24IEHRvHzzz9f9HF/ffLBBx8cxYcddljeNr74xS9G8X777RfF++yzTxR36NAhiv28PgBqx7x586K4bdu2UexrwLxu3bpF8fr166PY10c0adIkin0Nmp+3UMqfi3DvvfeOYj/Pjp9rbNGiRVHs5wnzfejdu3cUd+rUKa9Pvo7V15YA2H2+frNUvbiv//THGkuXLo1iP4+YJE2aNCmK165dG8W+JszXpfkaVV8TBjR0nAEDAAAAgIwwAAMAAACAjDAAAwAAAICM7LE1YL4Go1WrVlHsr6n29Vmer+GQ8utC/Nw//prnxo3j/w4/7w7zZAA1z9dCFeLrHXw91eDBg6PY13T5egjP10f4fd9vT8qvt/J1Hn5eHj93kN+GX1+hbeby+U2SJkyYEMWFakkA7J533nknin2+8cc3np/n1OeGmTNn5j1n2LBhUTx9+vQo3n///aPY13/64x2Ob7Cn4QwYAAAAAGSEARgAAAAAZIQBGAAAAABkZI+tAfPXRPuasL32isemfg4dX/M1dOjQvG3466o3bNgQxb7GwteF+Ou4AdS8WbNm5S3zc9r4+sx169ZFsd/3ly9fHsW+Hsuvz/P1Vb5GrNA2Fy9eXPRxv03fJ58T/Xvga1x9TYckzZ49O4qpAQOq37Rp06LYzwPm85PPH76+s0uXLiW3ecQRR0TxuHHjotjnG58v/OPMEYg9DWfAAAAAACAjDMAAAAAAICMMwAAAAAAgI3tsDViLFi2i2Nd8+XoHzz/u58QoxNdU+D74eTCoAQOy9/777+ct83P2FZrzKtfcuXOjuFevXlHs6x18/aevMW3Tpk0UF8oNfpu+j75Gy78m3wc/F6Kvm/V9KNQnPzcQgOo3Y8aMKG7Xrl0U+3pzv6/6GvfLLrus5DY/+9nPRvFtt90WxaVypK9DK1TXCjRknAEDAAAAgIwwAAMAAACAjDAAAwAAAICM7LE1YP56Y39NtJ+jwselasSk/BovP8+Or6ngmmig9vl6CCm/RrRt27ZR7Oe4WbNmTdHn+xovv6/7x/3z/fak/JoLXze2YsWKKPY1YH6eQv8alyxZEsW+zqRQzcf48ePzlgGoXqtXr45if+zhj1/8sYiPr7rqqpLb/NCHPlR0G6XmLvQ17xzvYE/DGTAAAAAAyAgDMADAnq17d8ms+E/37rXdSwBAA8EADACwZ1u0qHraAABQhj22Bqxz585RXOr6ZT+Phq+fKMTXYIQQiq6jZ8+eUezrPgDUvLVr1+Yt8/N2dejQIYr9HFznnHNO0XX6fONrUH2Nl499zYaUP8+Xb7Nx48aiffD5aODAgVH82GOPRbHPT4XmAfN1ZQCqn9/3fY2639f9ftndnd3t06dPlfvgj6n8MVTHjh2jeNmyZUX7BDR0HOEDAAAAQEYYgAEAAABARhiAAQAAAEBG9tgasB49ekSxr/Hy9Vrr16+P4kI1GJ6fy8fP++Xn2Sk0jw6AbPlaKSl/Xh1fc+EddNBBUfzqq69Gcal5BH191cqVK6PY16AVeo6vyfJ99jnOGzBgQBT7Gg3/fD+vjyStWrWq6DYA7L5OnTpFsT/28HxN6umnn77bffB1ZH5eL18jtnz58ijm+Ad7Gs6AAQAAAEBGGIABAAAAQEYYgAEAAABARvbYGrCWLVsWjX2Nhr8+2V+/XIiv+fJz+fiaCX8dN4Ca5+slCtV3btu2LYp9/ZOvt9p7772Ltvd8jamv11q3bl0UF8oVfq4fH/u5zDz/Gvv161e0j759offN15r4uFQtHIDS/H60YsWKKPY5bsaMGVH8y1/+suj6C9Vn+ZrT3r17R/G8efOiuEuXLlHs84dvDzR0nAEDAAAAgIwwAAMAAACAjDAAAwAAAICMMAADAAAAgIzssTfh8JME+htm+KJTX8TqC0oL6d+/fxT7iUx90XqhCWAB1KylS5dGcaEbZvgbWvgCcn8TDp8vfOxvsuEngvdF9f4mQYVueOHzS9euXaPY5zz/Ov3j/kYivuje85NVS/nv28KFC6PY3+gDQNX5G3r5Ywl/8xu/7/uJ4z2f76T8fHDwwQdH8ezZs6O4TZs2UbxkyZIoLjS5PNCQcQYMAAAAADLCAAwAAAAAMsIADAAAAAAyssfWgHm+RsNPtOwfL+d6ZX9d9fvvvx/Fq1evjuJCNRQAatbKlSuj2O/rktS8efOiz9lvv/2i2Nc7+ImUu3XrVnSbvgbV12cVqhf1NWD+Ob7uzNeRrVmzJop93Yjvs19/oToRX2uyePHiKKYGDNh9hxxySBS/+eabUezzha9P7969e9H1l6r/lKQzzzwzin/3u99FsZ/I3deDduzYseQ2gIaEM2AAAAAAkBEGYAAAAACQEQZgAAAAAJARasBSy5Yti2J/jfSTTz4ZxV/60pdKrnP48OFR/NZbb0Vxz549o9jXfQCoeX6uKj8Hl5Q/z8706dOjeODAgUXX4euvPF8/5euzfB99f6T8GlJf9+G34dfp61793Ii+zsTXjBWqi/Xb8HVlqPu6d5cWLSrdrls3yZX1ICMXXHBBFP/5z3+OYp9/fP35Cy+8EMWnnXZaFBeaG9HzOXDfffeNYl9H5tfp8wnQ0HEGDAAAFFTO4Ksq7QAADMAAAAAAIDMMwAAAAAAgI9SApV5++eUonjFjRhT7GrC//e1vJdc5aNCgKPY1FjfffHMUDxkyJIoPPfTQktsAsHt8/Wehei0/x9aqVaui2O+7S5YsiWJfc+Fro3zN16ZNm6K4ZcuWJfvoayx8n32NaZMmTaLYz+v13nvvRXHfvn2j+PXXXy+6PSm/LsS/DwB2n993/b7tay99e38842vAStWwSlLnzp2j2M/zNXfu3KJ98nMtAg0dZ8AAAAAAICMMwAAAAAAgIwzAAAAAACAje2wNmJ+Dws+R42vA/Lxg5Vyv7K+b9nUjfl6wrVu3llwngOo1ZsyYKC5Uy+SXLXL33PZzYI0aNSqKfQ2Xr9fysc9HTZs2jeJCucI/x8d+7jAf+3w1fvz4KG7btm0U+3nHCr1v69evj2L/vpx//vl5zwGwe3x9ld83/fGLPxapDn4ewtGjR0exr3stlD+AhowzYAAAAACQEQZgAAAAAJARBmAAAAAAkJE9tgbMz8OzefPmKPbXI/t6iXL4dfprnn1NmH8cQM1r1apVFPvaBUn64IMPonjNmjVR7OcB8/VT7du3j2JfG+X5GlU/L5iv75Ly5/Zp3bp1FPs6Mt/e58Q5c+ZE8dlnnx3Fn/vc56L4k5/8ZF6ffO1bjx498toAqF5HH310FN9zzz1R3LFjxyj2uaE69OrVK4pXrFgRxeXkNKAh4wwYAAAAAGSEARgAAAAAZIQBGAAAAABkZI+tAfP8NdCrV6+OYl8nUo4mTZpEsZ9nx9d8de/evcrbALB7Lr/88pJt/Lw6s2bNiuK+fftG8cMPPxzFfp4wv77t27dHsa8ZW7p0aRT7+lIpv3bNzxXmYz/3WNeuXaP4jTfeiOIvfelLUbxkyZIo9jVnUnnzJQKoXl/96lej+MEHH4xiv++vXLkyin1+69OnT5X70KZNmyj2dbM+5/kcCTR0nAEDAAAAgIwwAAMAAACAjDAAAwAAAICMUAOWatGiRRT7eopdqWXwdWV+bh9/DXRNzMUBYPf5+qbBgwdHsa9vWLZsWRT7eXdK1X/6ecL8+nwukfLzh6/z8PPulMo3vg/jxo2L4jPPPLPo8wHUjp49e0axryn1Nai+pvStt96K4l2pAfP5xdeg+nxUqK4VaMg4AwYAAAAAGWEABgAAAAAZYQAGAAAAABmhBiy1cOHCKN62bVsU+3qtcvi6EV+T4bfh69AAZK9QfZXf/xs1ahTFr732WhT7Of+8li1bFl3/jBkzoricGgyfw/w6fV2rn9vQ5x9fR/LKK69Esa8BK/S+mVmRHgOoDn7f8/vdqaeeGsUPPfRQFPt6rcceeyyKL7zwwir3yR//zJ8/P4pL1cQDDR1nwAAAAAAgIwzAAAAAACAjDMAAAAAAICMMwAAAAAAgI9yEI9WtW7coXrx4cRT7ovtydOjQIYpLTYzatWvXKm8DQPUqdOOIUvv/9OnTo9hPfOr3dX+TDv/83r17R7G/YcYHH3yQ1we/DV/UvmHDhij2RfC+EN/H/iYfXqH3rdTNAQDsvlI3CfI3zHnwwQej2N+AZ968ebvdp3bt2kWxn2jZHx8tX758t7cJ1CecAQMAAACAjDAAAwAAAICMMAADAAAAgIxQA5Y644wzonjUqFFRvCs1YG3atIlif020nxh1//33r/I2ANQ8P2m6zwdz586NYl+PNWDAgKLPHzhwYBR37NgxiqdMmRLFhWqptmzZEsW+zqxUPvI1Gv41rF+/vujjzZo1y+sTNWBAzfP15d4xxxwTxX6S9ZUrV0axr/ccP3583jqHDBlSdJtt27aNYp8/mjRpEsW+bhZo6DgDBgAAAAAZYQAGAAAAABlhAAYAAAAAGaEGLNW8efMo9vVZu1ID5vl5ePw10fvss89ubwNA9StVu/STn/wkim+66aYofvLJJ6PY11z4eb98/ZbPHYXmDFyxYkUUr169uujjvs7D12B07tw5ir/61a9GcaGaL69UbQqA3VfV2sr99tsviseNGxfFvj7r2WefzVtHqRqwNWvWRLHPYd6iRYuKPg40NPx1BAAAAICMMAADAAAAgIwwAAMAAACAjFADlrrkkkui+LXXXotiP0/Yrjj77LOLPn7IIYfs9jYAVL9StUwtWrSI4uuuu65o+/feey+K/Txfvh7C13Nt37696Pql/DoOH/s6kKOPPjqKW7duXXIbAOqf7373u1HcvXv3KPa54vjjj6/yNi644IIo7tatWxT7mtOTTz65ytsA6jPOgAEAAABARhiAAQAAAEBGGIABAAAAQEYshFDzGzFbImlujW8INW3/EEKX6lpZFT8XnSUtrcLqaV8/2lfrZ0oi3zQgtfnZqCv7B+2rv31t/h1D3VXbf4vq0j5S39vXpb5U+rnKZAAG7C4zGxVCGEH7htkeqEvq2v5B++ptD9Q1dW0fqc/t61JfiuESRAAAAADICAMwAAAAAMgIAzDUF7fTvkG3B+qSurZ/0L562wN1TV3bR+pz+7rUl0pRAwagYTLrJOn5NOouaZukJWl8mELYXMnzekn6t0IYVOCxH0p6RSE8V+CxyyQ9oxDm5yy7UFJfSf+VtFkhvL5LryV/W/tJukPSvpKCpDMVwhyZ9ZZ0n6ROkkZLulghbJbZ1yR9SdJ7ks5Nlx0j6eMK4epKttFC0lOSTkq38Zuc3zdK+qRCmC2ztQohf9Zmsy9LWq8Q/lrgsXMlvaMQpqTxLyQ9oRBe2KX3AwCAeqTGz4CZqZOZxqU/C830QU7ctKa3X0b/PmGmyWbabqYR7rH/NdMMM00304dzlp+eLpthpmtzlv/DTBPM9JOcZd8z07lFtj/MTHea6fKc92WzmSamv99YzS+5bGbqYqanamv7wG4JYZlCGKoQhkq6TdKvd8SVDb5Kr/O6SgZfjSRdJmlv98gZSgYxJ0g6ape2WdhfJd2kEA6UdJikxenynyl5nf0krZD0uXT5pyUNlvS6pA/LzCR9X9KPimzjs5IeVgjbJF2g5LUNVgiHSDpP0sqiPQzhtkoGX40lnSvpoJylv5d25lKgvuAYh2McYFc0rukNhKBlkoZKkplukLQ2BP2i4nEzNQ5BW2u6HznbaxSCtuUsmiTpY5L+6NodJOlCSQcrOfB4zkwD0of/IOlUSfMkvW2mx5W8lxtC0GAzPWumdpJaSjo8BP1fkS59R9L/haDxkv6cbnuOpBNDiG9zWaDvNSb9f1lipgVmOjoE/TeL7QKZMjtYyX7XVMkXUh+XtEVSI5n9Scmg6QNJ5yiEDTK7W8nZsQdlNkfSP5Xkgl9JGiHpHzLbIOlIJWeJhkpaLunLkrbJ7DOSvibpfUl3Kbmd7RJJlyuE99L1b0zX1VbSNxTCv12fD5LUWCE8K0kKYW263JScoboobfkXSTdIulWSSWqiJCdtkfQZSU8qhOVF3p1P56yrh6QFCmF7us15rk8/lnSWpA3pe7VIZjdIWqsQfiGzlySNk3SMpEcknS3peJl9T8lZuJky6ySz7gphYZE+AXUKxzgc4wC7olZqwMx0t5luM9Obkn5upqFmeiP9ZuURM3VI271U8Y2NmTqnO63MdLCZ3kq/PZlgpv7p8s/kLP+jmRqly9ea6ZdmGq/kwGiHEDQ1BE0v0M1zJN0XgjaFoNmSZij5pvkwSTNC0KwQtFnJ5T7nKDmoaWGmvZQc6GyT9ENJ1xd5H9pIGpwmpsraRH030zfMNCn9uSpt08tMk3Ke8630D4HM9HUzTUnfp/vSZa3MdFf6Xo010znp8svM9LiZXtDOS7ceVXIgVivM7HQzm25mM8ys5DfkZnaXmS02s0lltN3XzF40sylmNtnMrizRvrmZvWVm49P2PyjzNTQys7Fm9u8y2s4xs4lmNs7MRpXRvr2ZPWhm08xsqpkdWaTtAel6K35Wm9lVJdZ/dfpaJ5nZvWbWvET7K9O2k0utu474sqTfpmfJRig54JCk/pL+oBAOVnKm5+OVPH+ZQhiuEP4uaZSkT6dn2DZIGiZpvEKYrfgM3KtKzvj8RSEMlvQPSb/LWWcvJXnmI5JuU/57PkDSSpk9LLOxMrspPQPXSdJKhVBxsDdPUs/095slvSFpPyWXQ16u5CCrMLOmkvoohDnpkvslfVRm42T2S5kNy2ndStIbCmGIpFckfaGStTZVCCMUwo8lPS7p2+n7MTN9fIykoyvtU0aqknOqkm/S9uScasw5dTXfcIyz433gGKcEjnFKtq8z+SZtX205pzZvwrGPpKNC0DeUXE5zTQgaLGmiiuzQqS9L+m0IGqr0oMlMByq5TObodPk27dypWkl6MwQNCUGvldm/nkq+pa5QcTBTcHkImqrkm+wxkv4lqZ+kvULQmCLbGCGp1E60o+9Kvl2+XNLhko6Q9AUzDSv2ZCWX9QxL39svp8u+K+mFEHSYpBMl3WSmVuljwyWdH4KOT+NRko4tsY0aYclB5R+UXMZ1kKRPWfLtfzF3Szq9zE1slfTNEMJBSt7Pr5RY/yZJJ4XkQHOopNPN7IgytnOlpKll9kmSTgwhDC1znonfSnoqhDBQ0pBi2wkhTE/XO1TSoZLWKzkbUZCZ9ZT0dUkjQlIP1UjJN6aVtR+k5OD7sLQvZ5lZvzJeQ20aKek7MrtG0v7pwEmSZiuEcenvo5UMigr5Z5F1ny7pyUoeO1LSPenvf1NyZqjC/Qphu0J4V9IsSQPdcxsr2Se/JelDkvooufyxciH8TSEMUwifkXS1kgHfGTJ7UGa/lpn/W9BZuZcYJme8DpD0v5K2S3peZienj26WVPGHd1ffKym5jNJfwpmpXcg5d6v8fCORc6ot59SDfMMxDsc4RXGMU3/yTdq+WnNObQ7AHghB29LT2O1D0Mvp8r9IOq7Ec0dK+o6ZrpG0fwjaIOlkJW/422Yal8Z90vbbJD1U3S/AC0FXhaChIeiXSmorvm+m75rpfrOC3wr30M6bAlQmt+/HSHokBK0LQWslPazSiWOCpH+Y6TPSjssgTpN0bfo+vSSpuZJvxiXp2RCUe1lSbR4Upd/EhVkhqdmp+CauUiGEVyQVu6wqt+2CEMKY9Pc1SnbsnkXah1BxuVfyDWATJTckqJSZ7aPkTMYd5fSpKsysnZJ95c60f5tDCCvLfPrJkmaGEOaWaNdYUgtL6nZaSppfpO2Bkt4MIawPyVmYl5Vc+lJ3mJ2XnsUZJ7MRCuEeJZfDbZD0hMxOSltuynnWNlV+ufa6Ils7TdIzu9BL/5ny8TxJ4xTCrPRs16NKDiqWSWqv5P9KSg4AP4ieaba3khuQPCrpm0oO6FYq+Tzk2qAkL+T0ImxSCE8qhG9L+om0o+5ji3bezWlX3yul29tQok1Nq1LOqUq+SduTc6ov59T1fMMxDsc4pXCMU3zddSnfSNWcc2pzAFbqj7GU7EwVfdxxMBCCooMmM52kpMbhL2lyGBqCDgghOUUtaeMuXFf8gZI7jFWoOJipbPkO6enu0ZJaS+obgj4p6XwztXTbyD/IyVdO33PfJ7l1fkTJNyzDlSTuxkreq4/nvFf7pd9uSfn/L7V5UFTZN3TVzpI73w2T9GaJdo3MbJySpP1sCKFoeyV3jvt/Ss4alCNIesbMRpvZF0u07a3kj9uf09P/d5hZqxLPqXChpHuLdiSEDyT9Qsmd8xZIWhVCKDagmCTpWDPrZGYtJZ2peF+pfSE8knMjjlEy6yNplkL4naTHlNyoYletkdRGkpT84WisEJblPZZ4XTu/afu0pFdzHvuEzPaSWV8lB1j+8qG3lQy0uqTxSZKmpIOgFyWdny6/NH1NuX4k6br09xZKPm/bJZebQlihpA6uefp6hqeDN6VnywZLKvWHrRj/fkjJpZVlXcpXg8g5xdWlnFPX8w3HOBzjlEK+Ka4u5RupmnNOrc8DFoJWSVphtuNbjoulHd8UzVHyjY+086BCZuojaVYIyj1oel5JAuiatulopv13o2uPS7rQTM3M1FtJTchbSg5++puptyV3OLowbVvRtyaSrpL0c+08wJGSU5v+jkhTlZzGL9erks41U8v0dPp56bJFkrpacjemZkqK4ZVeq71vCHpR0jWS2ilJmE9L+pqZLG1X7BR/XTgoqlFm1lrJN3BXhRBWF2sbQtiWnt7eR9Jh6SnpytZ7lqTFIYTRVejOMSGE4UouSfiKmRX7prSxkj86t4YQhin5w1LONeRNlfxxf6BEuw5Kvo3rreQbwlaW3ESioBDCVCV34XtGyZ3/xknZFFTvhk9KmqTkD84gJZcK7aq7ldRsjVPy/ubeLfFfkirOvh2r5EYcl8tsgpKcl3tt/ntKcs2Tkr6sEDZGW0nuSvgtJZcBTlRysPGn9NFrJH1DZjOU1ITdueN5FXVb6TeiSi6BnKik7qrQncCe0c5LI7tK+peSuoMJSg6Ibi7xfhRzn6RvpzVsfWXWREkuLFkT0BCQcyptV3bOqS/5hmMcjnFqG/mm0na1eoxT43dBLNOlkm5Lvz2ZpeQaYCkZmd5vpi9K+k9O+09KuthMWyQtlPSTELTcTN+T9Ey6U26R9BWV+JbWTOcpKYjvIuk/ZhoXgj4cgiab6X5JU5QcbHyl4lsaM31VyQ7eSNJdIWhyziq/ouRbqvVmmiCppZkmSnoihPi2zSFompnamalNCFpT6k0KQWPMdLeSJClJd4SgsWmffpgu/0DStPTxRpL+nl4CYZJ+F4JWmulHSr61mJC+V7OVJrQCTlT83mep5Ddxu8uSA7+HJP0jhPBwuc8LIaw0sxeVXItdWfI+WtLZZnamkm/Z2prZ30NSh1PZej9I/11sZo8ouUThlUqaz5M0L+cbqgdV3q28z5A0JoSwqES7UyTNDiEskSQze1jJXQH/XqT/dyo96Dezn2jnTS1qVwg3VLL8RinvNsjLlQzGKtr8Iuf3y3J+7+XW9ZAqLqUxu0O5l2SE8I7yz66dpMKeUwhfruSxivU9W2B9UgizlHxmCj1nrHbell4K4TdK8kBl/qCkXuw5hfCUCg/SpNw5wEJ4UMnnMH7PQzjBPee/yr0Nvdl5kh7MuYFIbSHn1KOcU2fzTT6OcTjGKYR8U4/yTdrv6ss5IQR+avFHCldL4fO13Y8i/XtFCh1qZ9tqrOSPVW8l36yNl3RwGc/rJWlSGe1MyRmP35TZny6S2qe/t1DyzdxZZT73BEn/LtGmlaQ2Ob+/Lun0Es95VdIB6e83SLqpjL7cJ+nyMtodLmmyksvTTEntwtdKPKdr+u9+Sv5Itq+Nz069/ZHuDtL5td6Pnf35bJAaZbCdT4Q68FnZlZxTbr5J25JzirerUs4h39T9H45xim2bY5z6lG/S51Rbzsn8A8eP/88MzaVwcW33o5K+dZHCubXbB50p6R1JMyV9t4z29yq5lneLkm8mPlek7TFKLp+YoORU8jhJZxZpP1jS2LT9JEnXVeF1lJOc+qQJeHyaFMp5vUOVXLY1QcnNGIr+IUmT3jJJ7crs9w/SJDNJyd36mpVo/6qSb1THSzq5Nj87/PCzKz9VyTlVyTdpe3JO6fWXnXPIN3X/h2OcUn3gGKfEc+pMvknbV1vOsXSFAAAAAIAaVus34QAAAACAPQUDMAAAAADICAMwAAAAAMgIAzAAAAAAyAgDMAAAAADICAMwAAAAAMgIAzAAAAAAyAgDMAAAAADICAMwAAAAAMgIAzAAAAAAyAgDMAAAAADISOMsNtK5c+fQq1evLDZVbbZt2xbFjRo1iuJNmzZF8datW/PWYWZF4xYtWuxOFzM3Z84cLV261Eq3LE99/Fygeo0ePXppCKFLda6zIXyuli1bFsXr1q2L4hBC3nN8jmrevHkUd+7cuZp6l43qzjdSw/hsYPdUd87hM9UwkG9qzvjxUoFD5EjjxtKQIdn0J0vF8k0mA7BevXpp1KhRWWyqUv6AxQ+GvBUrVkRxhw4donjmzJlRvHTp0rx1+AOiZs2aRfEhhxxStA91zYgRI6p1fXXhc1ETuneXFi0q3a5bN2nhwprvT11mZnOre5114XO1ffv2KPb5x+cG769//WsUjxw5MooLfeHjc9TAgQOj+LOf/WzRbVY1R5Z6/q6sI1d15xupbnw26puGls+qO+fwmWoYyDc1p5w/A1u3Sg3xrSqWb7gEEahm5RysVKUdANQW8hkAVL9MzoDVhlKXEPpva/3ZqS1btkSxv1xww4YNUdy+ffu8Pvh1NGnSJIq/8IUvRPHPf/7zvHUAqH/22qtq321NmDAhii+99NIoPvLII0uu3+eXX//610XX6XOiP1tV1TNiu3O2CwCAPQlnwAAAAAAgIwzAAAAAACAjDMAAAAAAICMNtgas1F3G/vnPf0bxddddF8W+JuOBBx6I4m9/+9tRPHbs2LxtPPfcc1F8yimnRPEVV1wRxf7OZo0bx/89u3uXMgC1Y9q0aVG8yN2xoGvXrlH85ptvRvH1118fxatWrcrbhq9TveOOO6L4lVdeieLXXnstiq+55poobtq0ad42AADA7uMMGAAAAABkhAEYAAAAAGSEARgAAAAAZKTB1oCV4uur9t577yj+3ve+F8VnnnlmFD/11FNRPHv27JLbvOWWW6K4V69eJZ+Ti5ovoG4aPXp0FD/66KNRPH/+/Cg++uijo3jlypVR3LFjxyg+4IADonjx4sV5ffA1YEOGDInizZs3R3Hbtm2j2M9DePzxx0fxgQceGMWdO3fO6wMAACiNM2AAAAAAkBEGYAAAAACQEQZgAAAAAJCRelED5ue/kvLroXx9w5gxY6LY11hs3LgximfMmBHFkyZNiuInnngiitu3bx/FPXr0yOvjO++8k7cs1/Tp06N406ZNUezr0rZs2RLF3bp1i+K99mI8DWTBz5l18sknR7Gvj/I1XIMGDYriOXPmRPHf/va3KD700EOjeMCAAXl98vnj8ccfj+IPf/jDUexrut54440o9vMY+sfPPffcKO7fv39enwAAQD6O2AEAAAAgIwzAAAAAACAjDMAAAAAAICMMwAAAAAAgI/XiJhzlTEA8ZcqUKH777bej2BfB+yL2oUOHRvEHH3wQxWvXro1iP9HqsGHD8vq0dOnSKN6wYUMUt2rVKoqXLVsWxe+++24UN23aNIqbNGkSxUyMCtSMiRMnRrG/wcXPfvazKPaTrPuJ3/v06VO0/YoVK6L48ssvj+JZs2bl9XH9+vVRPG7cuCg+/PDDi7b3N/3p2bNn0fX96le/iuJbb701r08AACAfZ8AAAAAAICMMwAAAAAAgIwzAAAAAACAj9aIGrBy+ZqJfv35R7Gu4unTpEsWrV6+O4k6dOkWxr68aNWpUFL/11lt5ffKTrS5ZsiSK16xZE8UdOnQo2gc/0bKvKQNQM0aPHh3FTz31VBTfddddUfzYY49Fsd+X/STI06ZNi+J//etfUezzk5+4WZIWLVoUxb6G1E/c7ieC93VlHTt2jOKDDjooij/ykY/k9QEAAJTGGTAAAAAAyAgDMAAAAADICAMwAAAAAMhIva0B8zVdvp6qR48eUezn7TnkkEOieOPGjUW317p16yjevHlzFBeqx/LzdG3bti2K/fxmLVu2LBr7eXt8DKBmvPDCC1Hcu3fvKPbzCLZr1y6Kff7w9aFz586NYp+/TjrppCieOXNmXh+3bNkSxX7uMl/H6mvGfI2YX583b968KPbzHhbaJgAA4AwYAAAAAGSGARgAAAAAZIQBGAAAAABkpN7WgK1cuTKKN23aFMXdu3ePYl/v4OfkatWqVRQ3atQoips3bx7Fbdu2jWJf7yVJIYQo9nMB+TqR7du3F4193Zmv0fDvQbNmzfL6BKDq/Dxc77//fhSPGDEiin0Nl68xbd++fRT7eQx9Punfv38Ur1q1Kq+PvmbUz/Pl62R9H3yOO/7446P4oYceimI/z9iyZcvy+kQNGAAA+TgDBgAAAAAZYQAGAAAAABlhAAYAAAAAGWkwNWBNmzaNYl/f0KFDhyj29VL+cV+/tdde8VjV13S0aNEir4++psI/x88d5ms4fN3a1q1bo9i/Rl+n0qVLl7w+Aai6UjVbTzzxRBT7fc/v675Gdc6cOVWKp02bltfHjh07RvGsWbOi+POf/3wUz58/P4rHjRsXxS+//HIUv/7661Hs85XPqQAAoDDOgAEAAABARhiAAQAAAEBGGIABAAAAQEbqbQ2Yr6nwNWB+Hi/ffunSpVHsazZ8zZeZFe1P48b5b+W2bdui2M/r5efp8uvwNWSFtlGsPYDqceihh0bxpZdeGsW+PsrXXy1fvjyKFyxYEMW+pmzt2rVR7Gte/ZxeUn6+8fNyzZs3L4r9PF7r16+PYp8j/Vxnvi7O16ABAIDCOAMGAAAAABlhAAYAAAAAGWEABgAAAAAZqbc1YH5OLV/z5euh/JxavubC1zP4+onNmzdHsa/H8tuX8uvSfF2Zrwlr27ZtFPt5eIYNGxbFvi7Nz10GoOomTpyYt+zee++N4k996lNR7PdlP2dfu3btorh169ZFH/f5xsdbtmzJ66PXqVOnotvwOcznJ5+/Tj/99CheuHBhFL/44ot5fbj44otL9hNA1fjjE19z6us533vvvSgeNGhQFN9+++1R7PfbvffeO68PPp/4uVQ9nyN9vinFH9+UqssH6jrOgAEAAABARhiAAQAAAEBGGIABAAAAQEbqbQ2Yr4lo2bJlFPvrhVevXh3F3bt3j2I/z46/vthfr+zrJ/z1zYXW0aRJkyj2dSLegw8+GMUDBgyIYn9dtq+LA1B169aty1vm653uvvvuKH7iiSei+Prrr49iv+9269Ytin1N1wcffBDFRx55ZBQXqp/o2rVrFPt5ufr371+0va+LPe+886J46tSpUTx+/PgoHj58eF6fqAHDnqBY/XWpWiU/f5+UX1P+wgsvRPHvf//7KJ45c2YU+xzm6zn79u0bxb5G/vjjj4/im2++Oa+Pzz33XBQ//vjjUXzEEUdEcamaL39M5/tMzRcaGs6AAQAAAEBGGIABAAAAQEYYgAEAAABARuptDdimTZui2M9B4a/JnjZtWhT7ecKaNWsWxRs2bIjiQtdpl3q81Lxffi4g75FHHonib37zm1Hsr5Feu3Zt0fUBKO2ggw7KW/bTn/40ik877bQo7tKlSxQ/9NBDUeznzNlnn32i2OeKe+65J4r79OkTxb7mQ5IWLFgQxa+++moU+xz5/vvvR/GaNWvy1pnrzDPPjOITTzwxigu9b8CepqrzXRWaQ3TMmDFR/Jvf/CaKDzjggCi+4IILovjQQw+NYj/Pqa9ZHTlyZBT/6U9/iuI2bdrk9dHXsfqa0d69e0fxtddeG8Vnn312FPvjGaCh4wwYAAAAAGSEARgAAAAAZIQBGAAAAABkpN7WgPk5Idq2bRvFvkZszpw5Ueyvafbt/Zxafg4vf113oeu8C13bnavU3GV+rjI/N9DgwYOjuNBcZACq5t13381b9s4770Sx37cXL14cxX6OP18j6mtM/fp8fdbkyZOj2Ne0Svk5zOcTP9fYe++9F8XLly+P4oMPPjiKfc2Hf58mTJiQ1yefo4CGKPd4pNTf/XL4Gq5ly5ZFsZ/jr6ouvfTSorE3e/bsvGX/93//F8Xjxo2LYl+T7uto/Tp79OgRxT4f+XxWqO7eHwOVyoEnnXTSjt/Xr1+ftz6gJnEGDAAAAAAywgAMAAAAADLCAAwAAAAAMsIADAAAAAAyUi9uwuFviCHlF2D6iZRXr15ddJ2+4LJVq1ZR3Lhx/Nb4m3D4Ys5CfCG+v3GIn3jQ32Rj/vz5UTxv3ryi2+MmHMDuK3QTDj9xu9//77///ii+8cYbo9jf0MJPjOr3XX+DnosuuiiKx44dW7KPvsj9jDPOiOIjjzwyin3R+9VXX110mz6H+hwpSStXroxi/7qB+m7z5s3R32Z/wx6/L7do0SKKC93A66qrropif4Od119/PYr9fuaPmXy+8sdPb731VhQvXLgwiv1NziRp4MCBUXzqqadGcf/+/aPYTz7/6KOPRrGfON7fKM3nl0LHO/6Yy7fx79OHPvShHb8XuqkHUJM4AwYAAAAAGWEABgAAAAAZYQAGAAAAABmpFzVg5Vyb669R9tdhe/46bF9D5rfpJxX0ky0Wuh7ZX9vt2/jrlXv27BnFpSY+9XyNWaH3rTomiQQastGjR+ct8xOf+olRp0+fHsW+hvSFF16I4gMOOCCKfX55+eWXo3jYsGFRXCi/+foG38fjjjsuikeOHBnFviZ1v/32i2JfA+bz1dKlS/P6tGTJkiimBgwNTaNGjdS6desdsa+/8hOe+1ruQjVghxxySBTfeeedRfvga8T8vuxr3Lt27RrFn/zkJ6O4d+/eUewnSa4OX/rSl6LY1+37HOqPlwrxEy/72MvNR77mDKhpnAEDAAAAgIwwAAMAAACAjDAAAwAAAICM1IsasEL8ddN+ro0xY8YUfb6vAduwYUMU+1opPwdFObVU/hpm3+dS1yfnXlcu5deZeKVqzCRqwIBS/PxYknTEEUdE8aRJk6L4mGOOieIOHTpE8cSJE6N48+bNUez3XZ8bfD2nz3dSfr2Vf47PB35uIF834vOPr9HwdSpr1qzJ65OvNQEamkaNGkW1RGeeeWbtdaYeKzTXGNCQcQYMAAAAADLCAAwAAAAAMsIADAAAAAAyUi9qwHytgpRfM+HnwFq+fHnRdfo5H9atWxfFvkbD11P4eoly5qjw9Ve+7szXmXXq1CmKC70PuapaYwYg37hx4/KW9evXr2gbPyfWggULoviDDz6IYj+vjq+fKjV30OzZs/P66NusX78+ihctWlR0Gz7fDBgwIIp9Ttxnn32ieO7cuXl9WrFiRRS3a9curw0AAHsazoABAAAAQEYYgAEAAABARhiAAQAAAEBG6kUNmK89kPJrwPycW76+yvO1CAsXLoxiX0+1du3aKN60aVPR9lJ+XZqv4fKvwc+D4fvoazY8/x749QMo7d///nfeMl9P+dvf/jaKP/zhD0fxoYceGsU+PwwfPjyK33///Sg+7LDDovjggw+O4kL7ts8Xvi51yJAhUezrZP3cZX5esW984xtR7Ocl9HVukvSd73wninv16pXXBgCAPQ1nwAAAAAAgIwzAAAAAACAjDMAAAAAAICP1ogas0HxWvr7K8/Pw9O/fv+jz/Rxdvp6qVOznCZNKz8NVqG4s14EHHhjF06ZNK9qeGjBg9/3iF7/IW3bkkUdGsa8J7du3bxSvXLkyin39Z/PmzaO4ffv2Udy9e/co9vOMFdq358+fH8WrV6+OYp/z9t133yjeuHFjFPva289//vNRfMwxx5Tsk28DAAA4AwYAAAAAmWEABgAAAAAZYQAGAAAAABlpsDVgfs6sffbZp+g6/bxevqbLzyvm6x0K9adUm1JzlbVu3bpon3zs69j8PEAASps1a1beMl+z5fe9Aw44IIqff/75KH744YejeMyYMVHs67fuvvvuKF6xYkUU+3nDJGnq1KlR7Gu6/DbGjRsXxcuWLYvi0047LYr9vGCLFi2K4kLzNfpauC5duuS1AQBgT8MZMAAAAADICAMwAAAAAMgIAzAAAAAAyEi9qAErxM+r4/n6qn79+kWxr5dq1qxZFJeaJ8w/Xqj+wfPrKKVVq1ZR7F/T+vXro9jPA1ZOnwDE1q1bl7fM1z/5eMSIEVE8fPjwKPbzEPr5scaPHx/FvubswgsvjOLJkyfn9dFv09epXXTRRUX7vHz58ig+/fTTi27Tz4VW6H0rVecKAMCeiDNgAAAAAJARBmAAAAAAkBEGYAAAAACQkXpRA+bns5FK11PNmTMnio866qgonj17dhQvWLAgilu0aBHFHTp0iGJfg+brLaT8ebj8c0rVsfk+rFq1qug2fQ0YgKpbs2ZN3jI/79aMGTOiuGXLllH89NNPR7HfV31uWLhwYRQfdNBBRfvotydJhxxySBT7+czat28fxV27do1iP6+Xz4l+XkI/12Kh961Q7gYAYE/HGTAAAAAAyAgDMAAAAADICAMwAAAAAMgIAzAAAAAAyEi9uGuDL1iX8icq9UXuvvjbTzoaQojipk2bFl2fn6TUF6Rv3749r49+YlJfpL7XXvH41/fZT6zavXv3KPY3BjjggAOiuNRNPgDk8zezkKQjjjgiit95550obtKkSRSvXr06in1+8TfUGTlyZBR37tw5ip977rko9pMgS1KfPn2i+M0334ziU089NYp9/vA3LhowYEAUH3/88VE8ZcqUKG7btm1en/r27Zu3DACAPR1nwAAAAAAgIwzAAAAAACAjDMAAAAAAICP1ogbMzEoumz9/fhRv3rw5is8///zq71gVderUqUrtfd2ar/t44YUXotjXrvgaNACl7bfffnnLnn/++Sj2kxD7es4JEyZE8d577x3F69evj2Jff9WxY8eifSxUF7thw4aisa9B9X3wNWG+TrZZs2ZR7Cdu7tmzZ16f/AT2AACAM2AAAAAAkBkGYAAAAACQEQZgAAAAAJCRelEDNnfu3Lxlfh6dlStXRvH3v//9muxSrbjyyiujuHfv3lG8cOHCKC40Nxk1GUBxheYB+/3vfx/Fb731VtF1XHLJJVH8xhtvRHGjRo2i2Nd3+nrRmTNnRrGfd0zKr+nysa9T83WyPjcMHDgwin1dm4979eqV16dC9bsAAOzpOAMGAAAAABlhAAYAAAAAGWEABgAAAAAZqRc1YK1bt85btmXLlihu27ZtFJ9wwglV2oaf86Yu1i58/OMfj+KmTZtG8bZt27LsDtAgNW6cnxY/9rGPRXH37t2LrmPQoEFFY++zn/1sFB966KFR7POdn1dMyq/B6tGjRxQfdNBBRdt/9KMfLdpH3ydfU7bvvvvmPacu5lEAAGobZ8AAAAAAICMMwAAAAAAgI/XiEkQAO1177bUl29x4440Z9ARAbSsnH0jkBACoS8zXPtXIRsyWSMqfzAv1zf4hhC7VtbIqfi46S1pahdXTvn60r9bPlES+aUBq87NRV/YP2ld/+9r8O4a6q7b/FtWlfaS+t69Lfan0c5XJAAzYXWY2KoQwgvYNsz1Ql9S1/YP21dseqGvq2j5Sn9vXpb4UQw0YAAAAAGSEARgAAAAAZIQBGOqL22nfoNsDdUld2z9oX73tgbqmru0j9bl9XepL5UII9eJHCt2lcJ8UZkphtBSekMKAXVhPeylcUeTxK6UwSQqTpXBVzvIhUhgphYlS+JcU2qbLj5bCBCmMkkL/nG08I4W9imznQSn0SX9vLYU/5ry2l6Rw+C6+T9/J+b2pFF6RQuPa/v/jh5/68iOF76b7/wQpjKvYF6UwRwqdC7Q/WwrXVrKuE6RwVCWPdZDCI+l23pLCoJzHKstDP0vb/zVn2Wdy2xTYTg8p/Dv9vaUU/pHmsUlSeC3NP72kMKmS5/9QCqdU8thlUtg7J76vIg/yw8+e9lNZ7qiG9b4khRG70kYKX5XCDCmE3PwlBZPC79LHJkhheM5jl0rh3fTn0nRZMyk8leaNK3La3p773ALbP1cK16W/H5D2c5wUpkrh9mp6f06oyHHltJHCWVL4YW1/XvjZs3/qxRkwM5mkRyS9FIL6hqBDJf2vpG67sLr2kq6oZDuDJH1B0mGShkg6y0z90ofvkHRtCDok7cu30+XflHSmpKskfTld9j1JPwlB2yvZzsGSGoWgWTnrXi6pf/raLldyl5Vd8Z2KX0LQZknPS7pgF9cF7FHMdKSksyQND0GDJZ0i6f1izwlBj4egvHt8m6mxpBMkHVXJU78jaVy6nUsk/TZ9XsE8ZKZ2Of3abKZDzNRCSb74Q5EufkPSn9Lfr5S0KAQdEoIGSfqcpC0lXt91Iei5Aq+vkaTLJO2ds/hWSf+v2PqAhmhXckdG/qukL/5ufGdI6p/+fFHJviszdZR0vaTDleSg683UQdKHJb0mabCki9O2Q5Qcy4wpsv3/J+mW9PffSfp1CBoagg6U9PvdfnW75j+SPmqmlrW0faB+DMAknShpSwi6rWJBCBofgl41k5npJjNNMtNEs2SwYabWZnreTGPS5eekT71RUl8zjTPTTW47B0p6MwStD0FbJb0s6WPpYwMkvZL+/qykj6e/b5HUMv3ZYqa+kvYNQS8VeT2flvRY2s++ShLd9yoGbCFodgj6T/r4N9LXNslMV1WswEyPmmm0mSab6YvpshsltUhf2z/Spo+m2wNQWg9JS0PQJkkKQUtD0Pycx7+Wk1MGSpKZLjPTzenvd5vpNjO9Kel+JV/KXJ3uk8e6bR0k6YV0O9Mk9TJTN1Weh7ZLapJ+IdVSSe75lqTfh1B0EPVxSU/lvL4PKh4IQdMrXqukRmb6U5pTnkkHdxWv6fz09zlm+pmZxkj6lKQRkv6Rvr4Wkl6VdEo6+AT2JJXmDjNdZ6a307/jt6f7sMz0Uro/vWWmdypyhJlamOk+M0010yNSsi+mj91qplHpfvqDUp0KQWND0JwCD50j6a/pl/FvSGpvph5KBlrPhqDlIWiFkuOd07XzWKeJlPRf0o8kfb+ybZtpgKRNIey4ZXcPSfNy+jYxbdfLTK+muXWMWfKllZlOSN+jB800zUz/yHnvTk+XjdHO4zSZ6TAzjTTTWDO9bqYDCrwnQdJLSgbMQK2oLwOwQZJGV/LYxyQNVfJN8SmSbkqTyEZJ54Wg4UoGcL9Md9xrJc1Mv4H5tlvXJEnHmqlT+s3ImZL2TR+bLO0YxH0iZ/lPJf1VyRm5myX9WMkZsGKOznk9Byv5Fnybb2S242zY4ZKOkPQFMw1LH/5serZshKSvm6lTCLpW0ob0tVUMuiZJ+lCJ/tRZZna6mU03sxlmVnLGUTO7y8wWm9mkMtrua2YvmtkUM5tsZleWaN/czN4ys/Fp+5J//NLnNTKzsWb27zLazjGziWY2zsxGldG+vZk9aGbTzGyqmR1ZpO0B6Xorflab2VUl1n91+lonmdm9Zta8RPsr07aTS627jnpG0r7pwdAtZjrePb40zSm3Khn8FLKPpKNC0Mck3aad3/i+6tqNV3rgYKbDJO2fPrdgHgpBayQ9IWmspAWSVkk6PAQ9WtmLMVNvSStyBll3SbomPUD5PzP1z2neX9IfQtDBklZq55dM3rIQNDwE/V3SKEmfTl/fhvRLpBlK8nG9VJWcU5V8k7Yn51Rjzqlj+aZY7rg5BH0oPevcQvGBf+MQdJiSq2iuT5f9j6T16Vmi6yUdmtP+uyFohJIzUcebafAu9ren4jN089JllS1/VlIvSW9I+p2ZzpY0xn1B5R0tRWfHfi3pBTM9aaarzdQ+Xb5Y0qlpbr1AyZmyCsOUvDcHSeoj6WgzNVdyVv+jSt6b7jntp0k6NgQNk3SdpJ9U0rdRUt6XYpnjGKdk+zqTb9L21ZZz6ssArJhjJN0bgraFoEVKvi3+kJJvaH5ipgmSnlOSQIpeshiCpkr6mZJE+pSkcdKOgdFnJV1hptGS2kjanD5nXAg6IgSdqCQ5LJBkZvqnmf6efqPt9ZC0pMzX9kgIWheC1kp6WDsTxtfNNF5JMtxXig6kcl/TNiWXK7UpY3t1ipk1UnJp1RlKku+nzOygEk+7W8m3deXYKumbIYSDlAxwv1Ji/ZsknRRCGKJk0H+6mR1RxnaulDS1zD5J0okhhKGhvHkmfivpqRDCQCUHvZVuJ4QwPV3vUCV/tNYruZy2IDPrKenrkkaEEAZJaiTpwiLtC1w6Z/0qa18XpfvZoUouyVki6Z9muiynycPpv6OVHIwU8kChL1QKuFHJt87jJH1NycBqW7E8FIJ+ng52vqnk2+frzPR5M91vVvCLnyjXhKBxSvLUTZI6SnrbTAemD89OHy/1+v5Z4nUtVnxZYr2xCznnbpWfbyRyTrXlnLqWb0rkjhPN9KaZJko6SckXrxUK5ZTjJP09Xe8ESRNy2n8yPeszNl1Pqb+J1SIEbQ1BF6UDmweUDIp+aaZfpWeozi7wNJ9//qzkDP8DSi7PfsNMzZScVftT+v48oPg1vRWC5qVf7oxT8h4NVJKv3k3PZv09p307SQ+YaZKSAV/ue52r1vMUxzj1J9+k7as159SXAdhkxd8AlePTkrpIOjQEDZW0SFLRka0khaA7Q9ChIeg4SSskvZMunxaCTkvPOt0raWbu89Kza99TclB0vZLrnv+k5D/X25DTl8mShlhST1EWM52g5GzfkSFoiJJEXOy1NVNyRrC+OUzSjBDCrBDCZkn3aedZyIJCCK8oqacrKYSwIIQwJv19jZIdu2eR9iGEsDYNm6Q/RWcyN7N9JH1ESZ1ftTKzdkr+UN+Z9m9zCGFlmU8/WdLMEIKvC/AaS2phZo2VXH5S7NvO9NK5sD6E4C/hrTfSL3NeCkHXS/qq4jNBFWeStkmVXma3rsztrA5Bl6f56RIl+WpW+ljBPFQhPRNukqZL+kQI+qSSS6v9FzG5uaZiu2tD0MMh6AolBy5nute2u6+vebrd+qhKOacq+SZtT86pvpxT5/JNodyRnq25RdL5aQ35nxTvk+XkFEk7zmh/S9LJaZ3Zf1TGcU0lPtDOK3mk5Oz7B0WW57pCyZU/Ryg5E3+Bknp4r1D+mR+C7gpB5ygZIAySdLWSY7QhSq7qaZrzlHLzUoUfSXoxPdv4Ub/9HHUhT3GMU3zddSnfSNWcc+rLAOwFSc0qap0kyUyD0+ulX5V0gZkamamLkv+st5R8C7I4BG0x04lKLu+RpDVS5WeDzNQ1/Xc/JW/sPW75XkoGWre5p14i6YkQtFzJf+L29KdQkedUKbm5RwiaqeRU+A9yrm3uZaaPpK/tXDO1NFMrSeely9opuaxovSV1KLnfUGwxU5Oc19NJyWVTRQvt66jKLoWodmbWS8mlDm+WaNfIzMYp+fbs2RBC0faSfqNkMF7whiwFBEnPmNloM/tiiba9lXy7+Of09P8dZtaqzO1cqOSLhMo7EsIHkn4h6T2ll7yFEJ4p8pT00jnrZGb+Et56wUwHuEHMUOUXr1dFpfnGTO3NdhxofF7SKyFodfpYwTyUo6L2oom048ubQvnmHeWcyTLT0ZYU1Cvd9kGq/tc3QMlnoT4i5xRXl3JOnco3RXJHxQBgqZlaS0k9ZQmvSLooXe8gacdlhm2VfAGyKr265ozd6PLjki6xpI7+CEmrQtACSU9LOs1MHdJccVq6TGl/Oii5hPKv2nmsE5RTp5Zjx7FO+tzTK45PzNRdUiclg7t2khakZ7kulkp+IV1RM9s3jT+V81g77RwwXlZkHXUhT5FviqtL+Uaq5pxTLwZg6Snm85QUd88002QltVcLlZxenKCknuIFSf8vBC2U9A9JI9JT2pco2WEVgpZJ+q8lxbD+JhyS9JCZpkj6l6SvhKCV6fJPmemddD3zJf254glpncZl2nknsl8pqdX4jfIHalLyrdUJOfHnlVweOSM9bX63ksHjmPT3t5TsNHeEoLFKLktqbKapSi5jeiNnXbdLmmA7b8JxYro9VMLMWkt6SNJVIYTVxdqGELalp7f3kXRYekq6svWeJWlxCKGy+sVCjgkhDFfyh/UrZnZckbaNpaQeKYQwTMkf5nKuIW8q6Wwll3oUa9dBybdxvZVcqtHKzD5TWfsQQrFLeOuL1pL+YqYp6eXLB0m6YTfW9y9J51nhm3AcKGmSmaYr+f/OvT6/sjwkM50raVT6TfJKSePSPNc8BI3P3UAIWidppu28m2tfSS+n7ccq+fLnod14fXdLui19fS3Sg8INaQ5GJcg5lbYrO+fUwXxTMHek++iflBy8PS3p7TLWdauk1unf+B8qrRlP9++xSo5D7lFyh8OizPR1M81T8vmZYLbjTMUTSs64z0j7d0W6jeVKvuB5O/35YbqswnWSfpwOlp5WUhYxUdLfCmz+FUnDKr5cVjKYm5SWTzwt6dtprrhF0qXp8oEqcZY9BG1Ucqnnf9LLMRfnPPxzST8101gVP1u2xxwbkW8qbVe7xzhVvW89P7v/I4UWUnhDCo0y2NbD2oX50urCj6QjJT2dE/+vpP8t43m9JBWc06hA2yZK/hB8Yxf6d52kbxV5/KdKvtGao+TLgvWS/l6F9d9QYv3dJc3JiY+V9J8y1nuOpGfKaPcJSXfmxJdIuqUK/f+JpErn3OMnmx8pnCeF/8toW1dL4XO1/Zp3vf9VzzlVyTdpe3JO5e12OeeQb+rmjxR+W9k8grXYp25SeL72+8ExTn3NN2n73co59eIMWEMTgjYoqROrkVPNFdJLjB4NIa4fqUfeltTfzHqn32hcqOSyiWphZqbk2uKpIYRfldG+i5m1T39vIelUpWdWCwkh/G8IYZ8QQi8lfX8hhFDptytm1srM2lT8rvTbwiLrXyjpfTOruM3uyZKmlHodSi7XKHpqPvWepCPMrGX6Xp2sEoW2ZpZeOmeVXTqHjIWgR6SCt6GuCSsl/SWjbdUEck49yjnkm3rhJypcilGb9lPhmrWskW/qUb5J+11tOYe5WmpJCDuvqa7BbWxWcp12vRRC2GpmX1Xy7U0jSXeFECYXe46Z3avk8s7OZjZP0vUhhDsraX60kuvNJ6bXPEvSd0IIT1TSvoekv1hy56K9JN0fQih529Uq6CbpkSQPqLGke0IITxV/ir4m6R9p8p6lZNqCSqVJ71RJXyrVmRDCm2b2oJLbCG9VcunL7SWe9pCZdVIyZ8xXQvkFs6hBIVR/gXQl2/lz6VZ1V1VzThXzjUTOKWoXcg75po4Lyd2pq21QUR1CKOsy0BrHMU69yzdSNeYcS0+jAQAAAABqGJcgAgAAAEBGGIABAAAAQEYYgAEAAABARhiAAQAAAEBGGIABAAAAQEYYgAEAAABARhiAAQAAAEBGGIABAAAAQEYYgAEAAABARhiAAQAAAEBGGIABAAAAQEYaZ7GRzp07h169emWxKdSgOXPmaOnSpVZd6+NzkY3x46WtW4u3adxYGjIkm/7kGj169NIQQpfqXGd9/Fy9//77Ubxhw4Yo7tixYxRv3749bx1m8a65YsWKKO7WrVsUt2vXrsr9zFJ15xupfn42UL2qO+fwmWoYyDc1py4fg9S0YvkmkwFYr169NGrUqCw2hRo0YsSIal0fn4tsWBl/UrZulWrjv8LM5lb3Ouvj5+rKK6+M4okTJ0bxxRdfHMVr167NW0fjxnE6f/jhh4tu46yzzqpSH/2gb6+9avYCiurON1L9/GygelV3zuEz1TCQb2pOXT4GqWnF8g2XIAIAAABARjI5AwYAe7KXXnopim+55ZYobtasWRQvX748ir/+9a9HcaNGjfK20bJlyyg+4ogjovj++++P4scffzyKb7zxxij2lz3W9BkvAAD2FPxFBQAAAICMMAADAAAAgIwwAAMAAACAjFADBgC7afr06VH8s5/9LIrfeeedKB48eHAUT506NYpbtGgRxZ07d47ipUuX5vVh0KBBUexvQ+/vkujrzq666qoo7tevXxR/+ctfjuKuXbvm9QEAAJTGGTAAAAAAyAgDMAAAAADICAMwAAAAAMgINWAAkGPbtm15y/y8W7feemsUv/HGG1HcqlWrKD7ssMOiuHXr1lG8cePGKJ42bVoU+5qwQvVXvt9vv/12FH/uc5+L4g4dOkTx6tWro3jBggVR/KUvfSmKb7vttiju1q1bXp+2b98excwlBgAAZ8AAAAAAIDMMwAAAAAAgIwzAAAAAACAj1IABQA5f71XIxIkTo7h79+5F1+Hn4PJzdJ199tlRPGXKlCj29Vi//OUv8/r0wx/+MIpPO+20on3ydWctW7aM4rZt20axr+e65557ovjqq6/O6xM1XwAA5OOvIwAAAABkhAEYAAAAAGSEARgAAAAAZIQBGAAA9VT37pJZ6R9XpggAqEXchAMASvA3zfA3sOjSpUvR9lu3bo3iNm3aRPGSJUui+IQTTojiRYsWRfH999+f18fevXtH8cCBA6N43bp1Ubx58+Yo3rJlSxT7yZ/9jUbmzZsXxeVMYI3q5z4au90OAFDzOAMGAAAAABlhAAYAAAAAGWEABgAAAAAZoQYMAEqYPXt20cd9TdimTZui2NdCtW7dOorfe++9KF69enUU9+jRI4p9vZckLVy4MIrnzJkTxb7urFu3blFsZlHsa7rWrFkTxf41r1q1Kq9PHTt2zFsGAMCejjNgAAAAAJARBmAAAAAAkBEGYAAAAACQEWrAAKCEDz74IIp9/ZOvv/JzZvmarqlTp0bxypUro3jBggVR7Ofk8u0laezYsVHcuXPnKPbzgr3//vtR7Gu+1q5dG8X+NXnTpk3LW3bUUUcVfQ4AAHsizoABAAAAQEYYgAEAAABARhiAAQAAAEBGqAFLhRCKxnvtVf1j1VdeeSWKjzvuuGrfRlWtW7cuilu1alVLPQHqDl8D1qxZsyj2+83WrVujuFOnTlE8d+7cKF6xYkUUN2/evOj2unbtmtfHAw88MIqbNGlSdJ2+jm3AgAFR/Nxzz0Wxn7vM15hNnjw5r0/UgAENkz9G8nWre++9dxT7HPmrX/0qir/61a9GsT/2aNq0ack++TpWP/8iUJdwBgwAAAAAMsIADAAAAAAywgAMAAAAADJCDVjKzIrGpXz961/PW/bee+9F8bHHHhvFzz//fBT37t07ivfdd98q9cHXnTRuXPq/96abboriBx54IIpfeOGFHb9v3769Sv0BGgpfL+XnyJoxY0YUb9iwIYp79eoVxb4mzNdrLVu2LIp9jdj69evz+rhmzZoo7tOnT9Ft+PqIVatWRfHIkSOjeNCgQVF82mmnRbF/DwDUD76eyx//zJo1K+85V111VRR/+ctfjuIxY8ZE8ZVXXhnF//znP6P4P//5TxTfc889UXzWWWdFsa85k6SWLVtG8Re/+MUo9nnXv24gS5wBAwAAAICMMAADAAAAgIwwAAMAAACAjDTYGjBfr7S7NV7+GugPfehDUXzRRRflPWf48OFR7Gsu/PXIX/va16L40UcfrVIfy6n5+tvf/hbF9913XxT72pZp06bt+N3XwQB7itWrV0exn9PG7ze+ntM/3rdv3yj283y99dZbUbxkyZIoPuigg/L66LexZcuWKPZ1ab5ewvf5zjvvjOLvfve7Uezr0Px7AqB+KHU85OtJJenxxx8v+pyHH344ik899dQo9vMGbtq0KYp9DfzLL78cxX5ew0LKOSYCagtnwAAAAAAgIwzAAAAAACAjDMAAAAAAICN18gLZUnNSlNNmr72Kjy03b94cxQsXLoziYcOGRbGf8+Kaa66J4sGDB+dtY86cOVHsayQOPPDAKH7uueeiuEOHDlH8ne98J4rPPffcKPbz/Lz22mt5fbrllluKPmfIkCFR3LNnz0rbAnsKvy/7mi1f3/npT386im+88cYo9vuSz1e+5szPC7Z48eK8Po4fPz6KfU5q2rRpFPt5A/08Yn7uMl8z5mvOmFMHaJhy5wOtMHPmzCjeb7/9ovjuu++OYn+842vcW7VqFcX+mM7P+3XMMceU7NO//vWvKP7MZz4Txdu2bdvxO/kLWeMMGAAAAABkhAEYAAAAAGSEARgAAAAAZKRO1oCVM0dXqTavvvpq0cevv/76KM6tdZLy58Dx84rNmzcviv28PYX4eXj8Nccf+chHorhdu3ZRfOutt0bxXXfdFcVt2rSJ4qVLl+b1wV+nfeSRR0bxm2++GcW5tSi510sDexJff9C5c+coXrlyZRT7fb1///5R7Ouvcufbk/JrVH0u8DVpkjR//vwoPvroo4uuY+7cuVHs84ef+9DXiPl5eHxNmJQ/V5ivIwMagmL1Q/5Yxbf1xxZSfk1pKT6f+DkAS63P14f+9Kc/jWKfC6T8/b979+5R/Mc//jGK/dypPhecdNJJUdyxY8co9jXtfm5EKb/O7KGHHopiXwOWO09YVeeGBXYXZ8AAAAAAICMMwAAAAAAgIwzAAAAAACAjdbIGrBwzZsyIYl+Dce+990axr7H4/ve/H8V+ji4/L5h/3F9z7a+HlvJrpvy13hs3boziTZs2RfEnPvGJKD777LOjePr06VHs58DYd9998/p0yimnRLGvC/nnP/8ZxbnXhnONNPYEvv6q0DI/b5evZyg1Z5bPV/vvv3/Rx/28X4XqrfzchT6/+Of4bfq5x1q3bh3FvibD15j6GhApP4/26dMnrw1Q31Xlb2M5bUvNSeXXkVvLVCguxc/Z5WtMDznkkLzn+GOiTp06RXGPHj2i2NfNX3HFFVG8aNGiKB44cGAU+2OXtm3b5vXps5/9bBT7vPn3v/89in1NGJAlzoABAAAAQEYYgAEAAABARhiAAQAAAEBGMqkB27Rpk959990d8X333Rc93rVr1yj2tQt+/hkpf54LX59w4oknRrGfg8LP2+VrOvz1xX4eDV/ftXz58rw++voH/zr8XEG+Bsw/7msyDjjggCg+5phjorhDhw55ffL9fPTRR6PYX8c9efLkHb/7/xegIfL1pZLUrFmzKPb5Z9WqVVHs6x98TYavIW3RokXR9S1btiyKfX6TpHfeeSeKC80DmMvXqfkc5/vo5wnzse+jVDh3Aw1NqZqtYqo651chfl+97bbbonjs2LFR7OcxvOyyy6LYz8l1zz335G1zypQpUexz4lFHHVV5hyX94Q9/iOKrr746in2f/TGZn+dQyp/n1MejRo0q2icgS5wBAwAAAICMMAADAAAAgIwwAAMAAACAjDAAAwAAAICMZHITjsWLF+vWW2/dEY8fPz563Be4e4UmFfQTCC9ZsiSKfRG7v9FHq1atonj27NlRPGnSpCj2kwj6iVL9DTOk/JtW+Bt3eP598IX5I0aMiOK33347im+++eYo9jcWkaSDDz44iv2Ejv45/fr1q7R/QEPk922p9E04Bg8eHMV+UmKfP/wNdUpN1Oz300KTHufe6KhQH/2NAvzEy76Qv0uXLlHsc0OpmwRJ+XkYaIiqMhGzV+i4wN+Yw98sy+cTf8MKn08uvfTSKH755Zej+MADD4ziWbNmRbE/vpLyj2/8MVYp/j3zkyb792X9+vVR7CePlqTTTjstin1O8jfleO+993b8vnnz5uIdBqoZZ8AAAAAAICMMwAAAAAAgIwzAAAAAACAjmdSAdejQQeeff/6O2E+a/P7770fxihUrorjQZJ7z58+PYl8TNmfOnKKP+5qvdevWRbGvO/P1D359flJTSTrkkEOi2E+e6icuffjhh6P4mWeeyVtnMf598tdMF+Jr4Zo2bRrFudeSF6rFAxoaX4sglZ442ddk+forXy/RrVu3KPaTsPt9zbd/4YUX8vroJ0bt06dPFPuJ2X2f/GvydSg+N/gaDv+apcK1sUBDVmpS5u3bt0dxORMxjxs3Lor9vtukSZMo/va3vx3Fw4YNi2L/d3/q1KlR7Os/fY2ZlP86//73v0fxl7/85bznFOPzx9y5c6N4wIABUezrbiXpkUceieKLL744iocOHRrFEydO3PG7f0+BmsYZMAAAAADICAMwAAAAAMgIAzAAAAAAyEgmRT0tWrSI5p/af//9o8d79OhR9PmF5snw1+v6eSt8HceTTz4ZxZdddlkU++uJO3XqFMW+/qEmfPSjH43ip556KoqHDBkSxb4uzV9LXmheHn/dtq99W7BgQRTn1pEVqvEAGpqlS5fmLWvTpk0U+32hd+/eUexrKvy8PL7my9eQ+bpYX0vl62il/BouX+fhH/d1baXm+fOv2bcvVPtCXQX2BLmf/VLzffr6Tj8fnyTNnDkzin09lK8v9/Wf11xzTRTff//9Rbe57777RrE//nnxxRfz+vihD30oiv0xl69TPemkk/LWkcsf3yxatCiKL7jggij2x0uSdMYZZ0TxRRddFMW+Tj43P/naPKCmcQYMAAAAADLCAAwAAAAAMsIADAAAAAAykkkNWKNGjaJ5s/z1x88//3wU+9oEP8eFJLVv3z6KBw0aFMV+DqyvfvWrUeznyNm8eXMU+zqQQnOR5So0341f5msk/LXgPXv2jGJfY/Hqq69Gsb9m2tdoFJpfxF+f7t8nX6virwUHGrpC+3Lz5s2LtuncuXMU+3oIP2+gr89cuXJlFPvc4Gs1fc2YJC1fvjyKff3VwoULo9jn0FI5zudlHxeaJ9DnVaAhyp0Tr6rzZRaqnXzssceiePr06VHs920/T9ikSZOi2M85umTJkih+/PHHo/iqq66K4pdeeimvjz/4wQ+i2OeXH/3oR1Hsa8BWrVoVxV27ds3bRrH1F+L77fn5zHJr/32OB2oaZ8AAAAAAICMMwAAAAAAgIwzAAAAAACAjmdSAeX7OCR97M2bMyFvmazDefffdKPY1FX7OLH9NtZ+Xp23btlHs69Byr/mWCs+55efq8TVa/tpvf41zly5divbBz1vh17dixYq8Pnl+fiPf5759++743c8jBOwp/L5bqh5q8uTJUezzj499vvL5pUOHDkX7I+XnB7+/+ppSXyPq67V8PvJ1s16h2hdfYwo0NOvWrdPIkSN3xLfddlv0uK+r9vuJzwWF2vi/077G1Ndz+vk833jjjSj286L64x+vUM2pr+HyfN3Z4YcfHsX+GO7UU0+NYp/z7rvvvii+8sor87bZv3//KB4+fHgU+/nUfvvb3+74vVAtHlCTOAMGAAAAABlhAAYAAAAAGWEABgAAAAAZqZUasKrq169flZ9zyCGH1EBPADR0vv5Kyq+v8jWfU6dOjeKjjjoqigcOHBjFvr7K12P5eXp8TYifz6/QMl8n5ms2fM1D06ZNo9jXmJaaZ7DQPDq+Ng5oaFq0aBHNJ/X5z38+etzvy742u5w5RP28X/5xvy9+73vfi2K/L/sadz/fp58vy9eYSdI3v/nNKPY1675uzOefH//4x1E8b968KO7Ro0cU+3zlH5fy615btWoVxT4P5+YnasCQNc6AAQAAAEBGGIABAAAAQEYYgAEAAABARupFDRgAZKVQLYCvwfJ1Yn7+vP/5n/+J4lmzZkXxmDFjotjXT0ycODGKp0yZUnR7Un4NmK/B8HVs8+fPj+JLLrkkio844ogo9jUcvo+FFJrjCGhI9tprr6jW6Nhjj63F3tQffi6y2tayZcva7gL2MPx1BAAAAICMMAADAAAAgIwwAAMAAACAjFADBgA5Cs0D5vl6q2OOOaZo+z59+hSNveOPP77o435eH0natGlTFPs5cXaXr1Mr530q1E8AAPZ0nAEDAAAAgIwwAAMAAACAjDAAAwAAAICMUAMGADmaNWuWt6xUvZOfY8vzNWONGjWKYj/3WKntFZpfa3drvkr1oU2bNlHsX0Oheq/NmzfvVp8AAGiIOAMGAAAAABlhAAYAAAAAGWEABgAAAAAZYQAGAAAAABnhJhwAkGPp0qV5y7Zs2RLF/gYUjRvvXir1N7yo6k05qoO/iYZ/jf4mHH7iZ/+4VPrmJHuKa6+9tmSbG2+8MYOeAADqAs6AAQAAAEBGGIABAAAAQEYYgAEAAABARqgBA4AcftJkKb+WaevWrVHco0ePau3DrtR8laobK/V4qRowP9Gzr4vz74lUuC4MAIA9HWfAAAAAACAjDMAAAAAAICMMwAAAAAAgI9SAAUCOvfbK/15qzZo1Ubxy5cooLlQ3lqtUfVV1KFU3trtzifm5zvxrLjTnV6tWrXZrmwAANEScAQMAAACAjDAAAwAAAICMMAADAAAAgIxQAwYAOS6//PK8ZaNHj45iXwN26KGHFl2nr5+qiwrVvuXyc535uNBrbN++/W73CwCAhqbuHxUAAIBace2115bV7sYbb6zhngBAw8EliAAAAACQEQZgAAAAAJARCyHU/EbMlkiaW+MbQk3bP4TQpbpWVsXPRWdJS6uwetrXj/bV+pmSyDcNSG1+NurK/kH76m9fm3/HUHfV9t+iurSP1Pf2dakvlX6uMhmAAbvLzEaFEEbQvmG2B+qSurZ/0L562wN1TV3bR+pz+7rUl2K4BBEAAAAAMsIADAAAAAAywgAM9cXttG/Q7YG6pK7tH7Sv3vZAXVPX9pH63L4u9aVyIYRq/5HCd6UwWQoTpDBOCodX8/pPkMK/q2ldA6UwUgqbpPAt99jpUpguhRlSuDZneW8pvJku/6cUmqbLvyaFSVJ4ImfZMVL4dZHtt5DCy1JoJIW9pPC7dB0TpfC2FHpX83t3mRRu3o3n/0IKJ9XE54YffurKjxS6S+E+KcyUwuh0nx6wC+tpL4Urijx+dZorJ0nhXik0d4//Tgprc+LdyjFpPCB9/rtSGCOF+6XQbRdeWy8pXJQTHyKFu2v7/44ffmrrp6aOfaTwkhRG7EobKXw1PVYJUuics9zS/DIj7e/wnMcuTfPDu1K4NF3WTApPpfnnipy2t+c+t8D2z5XCdenvB6T9HCeFqVK4vZren5LHhLltpHCWFH5Y258Xfvbsn2o/A2amIyWdJWl4CBos6RRJ71f3dnaVWd7k08slfV3SL1y7RpL+IOkMSQdJ+pSZDkof/pmkX4egfpJWSPpcuvzTkgZLel3Sh81kkr4v6UdFuvRZSQ+HoG2SLpC0t6TBIegQSedJWrkLL7NGpO/J7yWVNzMnUA+l++0jkl4KQX1D0KGS/ldSt11YXXtJV1SynZ5Kcs+IEDRIUiNJF+Y8PkJSB/e03coxZmou6T+Sbg1B/UPQcEm3SNqVu3/1knRRRRCCJkrax0z77cK6gHqtDh/7/FdJX/zd+M6Q1D/9+aKkWyXJTB0lXS/pcEmHSbreTB0kfVjSa0ryz8Vp2yGSGoWgMUW2//+U5BhJ+p2SY6ehIehAJccTteE/kj5qppa1tH2gRi5B7CFpaQjaJEkhaGkImi9JZppjph+YaYyZJpppYLq8lZnuMtNbZhprpnPS5b3M9GrafoyZjvIbM9OH0uf0NdOhZnrZTKPN9LSZeqRtXjLTb8w0StKVuc8PQYtD0NuStrhVHyZpRgiaFYI2S7pP0jnpAc9Jkh5M2/1F0rkV3ZHURFLLdH2fkfRkCFpe5P36tKTHct67BSFoe9q3eSFoRfoa1prpx2Yab6Y3zJKDQTN1MdNDZno7/Tk6XX6YmUam783rZjqgwHv3kbRNZzOdlv4+xkwPmKl1zv/Zz8w0RtInQtBcSZ3M1L3IawLqsxMlbQlBt1UsCEHjQ9CrZjIz3WSmSWkOu0CSzNTaTM/n5LZz0qfeKKmvmcaZ6aYC22osqUX6xVBLaUeubCTpJiUHL7l2N8dcJGlkCPpXzmt7KQRNMlNzM/057f9YM52Y9qWyPHyjpGPT13Z1uuxfyhlEAnuQYsc+16V/nyeZ6fb0OKLi2ORn6bHPO2Y6Nl3ewkz3mWmqmR6R1KJiI2a61UyjzDTZTD8o1akQNDYEzSnw0DmS/pp+Gf+GpPbpMdOHJT0bgpanxx/PSjpdSb5pqST/WLqOHyn5AqggMw2QtCmEHbfs7iFpXk7fJqbtCuYYM52QvkcPmmmamf6R896dni4bI+ljOdsseewTgoKkl5QMmIHaUd2n1KTQOj29/I4UbpHC8TmPzZHC19Lfr5DCHenvP5HCZ9Lf26fPbSWFlhWX5EihvxRGpb+fIIV/S+Go9PKg/aTQRAqvS6FL2uYCKdyV/v6SFG4p0e8blHMJohTOr+hfGl8shZul0FkKM3KW7yuFSTltxkrh71JoI4UXpNCkyDabSmFhTrxP+h6Nk8IvpTAs57EghY+mv/9cCt9Lf79HCsekv+8nhanp722l0Dj9/RQpPJT+fln6Os6TwqtS6JC+plek0Cptc03OJQNzpPD/XL//JIWPZ3GKVkniny5phqRry2h/l6TFkiaV0XZfSS9KmiJpsqQrS7RvLuktSePT9j8o8zU0kjRWUsnLZiXNkTRR0jhJo8po317JlwHTJE2VdGSRtgek6634WS3pqhLrvzp9rZMk3SupeYn2V6ZtJ5dad139kcLXVcklfVL4uBSeVXLJcDcpvCeFHlJoLIW2aZvO6WU9puQyvUo/i1K4UgprpbBECv9wy69Of8+9BHF3c8yvpFDwcy6Fb+bkzIHpa2teKg+7dRwthX/V9v/h7v3/l59zqpJv0vbknGrMOXUp35Q49umY8/vfcv6WvySFX6a/nymF59Lfv5GzLw6WwlallxdWrCvNQS9JYXDOuiq9TDH9W557CeK/K44d0vh5KYyQwrcqji/S5d9PlzVOjzfGSuEiKZwthRtKvCeXV7y+nHiVFJ5Ucvl1+3R5sRyzKj022ktJucgxaV56P21rSi6jrri8sLJjnyhfSeHTUvh9bX5m0s8wxzjF29eZfJO2r7acU0MfqNAo/bD/QAoLpXBZunyOFHqmvx+ek2xGKbmueFz6854UDpRCuzRZTUyXr0/bn6CkNmOSFPZOlw2SwuqcdUyUwjPpYy/lJsNK+nyDdnMA5tZ3nZJrn8+WwoNS+LUU9nJt9pbCNLesmRTOkMJNUlguhZPT5ZukkM7bFi7QzsHr4pzXPE4KHyj5Q7CvFB7RznqyaWn7y6QwRQpvaOcB41lSWJqzjilSuDPn/2x/18cfKx1I1+RPulPPlNRHUtM0KRxU4jnHSRpeZnLqIWl4+nsbSe8UW7+Sb/1ap783kfSmpCPK2M43JN1TheTUuVS7nPZ/kfT59PemktpX4b1dqGSSwMra9JQ0W1KLNL5f0mVF2g9KE1NLJWd2npPUr6Y/J9X/uSs6APu1FD6bE/8t3cebpPmhovZjg5I6sl6F8kP63A5KBlBd0uc/KoXPpHnhtZyDiLWVPL/KOUbFB2CPKKe+U8kXNINL5GE/AOsvhdG1/X+46//3Vcs5Vck3aXtyTjXlnLqYb1T5sc/HldSNT0z/Rl+bLn9JCkenv3erOLZIc0HuvjhGOwdgX07jCUq+uLkwZ101NgBz62qS5q42aU55UApnF9jmd5RTP58u21sKn5XCY1KYlh7zFMsxz+Y899Y0Rw6Vwis5y8/WzgFYZcc+Ub6SwqlKB2e193nhGKeM9nUi36SPV2vOqZG7IIagbSG5rOV6SV+V9PGchzel/25LX4CU/Kd/PCTXBQ8NQfuFoKlKRqaLJA2RNELJm19hgaSNkoblrGNyzjoOCUGn5bRfV8WX8YGSbw8q7JMuW6bkVH1jt3wHM+0t6bAQ9Kikbyqp7Vop6WS3jQ1KvnHYIQRtCkFPhqBvS/qJdl7euCUEhfT33PduL0lH5LzuniForZJLA14MSW3JR912ZirZGQdUdFnJJQcV6zgohB11bVL+e9c87XtNSy8DDbNCCDsuAy32hBDCK1LRy7Fy2y4IIYxJf1+j5NuVnkXahxDC2jRskv6EytpLkpntI+kjku4op09VYWbtlCTjO9P+bQ4hrCzz6SdLmhlC8HUBXnqJnEWXyFXiQElvhhDWhxC2SnpZOZeG1COTJR1axed8Wkkd1aEhaKiSvNW86DOSuozZIWhJCNoi6WFJRynJaf0kzTDTHEktzTQj94m7kWN25bUVy8NeVrmhplQp51Ql36TtyTnVl3PqXL4pdOxjSd3lLZLOD0lt958U75OFjokKMlNvSd+SdHJI6sz+o9J5pjKVHeNUtjzXFZL+KukISauU5J9vFthGoWOc+SHorhB0jqStSg5qi+WYTTm/l3yPVPzYJ1ddyFUc4xRfd13KN1I155yauAnHAWbqn7NoqPKLP72nJX0t59reikFVO+2sibpYyYi2wkol/+k/NdMJSk7hdrGkEFZmamKmg3fjpbwtqb+ZepupqZK6hsfTQdCLks5P212qnfUVFX4k6br09xZKPsDbpbjgMyTXVzdKE7TMNDw9sJKZ9lJS7FrqvXtG0tcqAjMNTX9tp51J8zL3nLlKBsV/Td+jNyQdbaZ+6TpapdduV2aAkm8BalpPxUXM81QkeewOM+ul5MD3zRLtGpnZOCWXADwbQijaXtJvlNTxbC+zK0HSM2Y22sy+WKJtb0lLJP3ZzMaa2R1m1qrM7Vyo5HR75R0J4QMlN6d5T8kXHqtCCM8UecokSceaWSczaynpTMV/yOuLFyQ1M9OO999Mg9P6jFclXWCmRmbqouSPw1tK9rfFIWhLWju1f/rUNUq+7CjkPUlHmKllmvtOljQ1BP0nBHUPQb1CUC9J60Nyw59cu5RjlHxLeZSZPpLz2o4z06D0tX06XTZA0n5K8mplebjQa8sqN9QUck5xdSnn1Kl8U+TYp2LfW5rWVp/vn1vAK0pvcJPum4PT5W2VfCG6ypI68DN2o8uPS7okrWs9QtKqELRAyfHYaWbqYMnNN05LlyntTwcltVN/VZJvtiv5DLXwG1BywN8v57mnm6lJ+nt3SZ2UHKcUO9YrZJqkXmbqm8afynms2LFPrrqQq8g3xdWlfCNVc86piTNgrSX9xUxTzDRByR0EbyjxnB8pGWlPMNNk7byj1y2SLjXTeEkD5c7EhKBFShLBH5R8sM6X9LO0/Tgp/6Ydnpm6m2meklOo3zPTPDO1DUFblXyD9bSSJHJ/CJqcPu0aSd9Iv5XupHR0nq5vWNq3irsC3aPketejJT1VoAvPSDom/b2rpH+ZaZKkCUq+Hbq5xEv4uqQRZppgpimSvpwu/7mSwelYFfjGKARNU3Kw9YCSpH6ZpHvT/7ORSt7vPGny7CdpVIl+1Rtm1lrSQ0qu511drG0IYVsIYaiSbwUPM7NBRdZ7lqTFIYTRVejOMSGE4Ur+sH7FzI4r0raxkksRbg0hDFOyf5S8Q6WZNZV0tpL/+2LtOij5Nq63krtztjKzz1TWPoQwVckdQp9R8lkfp+Qby3ol/ZLlPEmnmGlmmpN+quRyhkeU7JvjlQzU/l8IWijpH0r2w4mSLlFygKAQtEzSf9Pi+5vcdt5Ucm37GCU5Yi+VMb/I7uSYELRBSc78mpneTXPGFUr+yN0iaa/0NfxT0mUhuaFAZXl4gqRtltwYqOImHCcq+VYeRZBzKm1Xds6pg/mm4LFPCFqp5KzXJCXHE2+Xsa5bJbU201RJP5Q0WpJC0HgltTbTlOz3/y21IjN9PT3G2UfJMVbFmYonJM1SUnv0J6V3aw3JDX1+lPbzbUk/DPFNfq6T9ON0sPS0pGOV5J+/Fdj8K5KGVXy5rmQwNynNJU9L+naaP4se63khaKOSOzf+J70Jx+Kch4se++TYY3IV+abSdrV7jLOr1y7yUz0/Uhguhb/Vdj+q0N/zpPCjbLalIyU9nRP/r6T/LeN5vVR+TUYTJX8IvrEL/btO0reKPP5TJd9ozVFy8L5e0t+rsP4bSqy/u6Q5OfGxkv5TxnrPkfRMGe0+IenOnPgSSUVvZuOe/xNJlc6BxU82P1nlmLSW4w2ltWv18WdXck5V8k3anpxTebtdzjnkm7r5I4XfSuGU2u6H61M3KTxf+/3gGKe+5pu0/W7lnBqpAUP5QvIt9otmJU+51xWNJf0yo22ll4Fa7/QbjQuVXDZRLczMlJy9nBpC+FUZ7buYWfv09xaSTlV6pqOQEML/hhD2CSH0UtL3F0IIlX67YmatzKxNxe9Kvy0ssv6Fkt43s4rb7J6s5G5HpXxKJU7Np9JL5Kxl+l6drORscKXMrGv6735Kro2+p4ztoAZlmGP2k3RtSK4eqK/IOfUo55Bv6oWfSHVuvq39VLhmLWvkm3qUb9J+V1vOKVXM+P/bu/M4qaoz/+PfQwNNA42sAoKAorgB4hKXyahxHEdnMhPjaCYaHUfzyySZ0ajZZkhM4pJEyZg4iSZmcUtmYhL9JdFoxGgWjWjigsqqIKBoBETBAALNIj7zxz2NdZ6qrlsN3ber9fN+vepFPbdOnXuquupwT93z3IMCmOnGrm5Drcyqn9Lt2H3Z6yGE1mmgDZJuNLP51Z4TQvixpHdJGhpCeFHSxWZ2QxvF36lsvvncOOdZkj5rZtPbKD9S0g9CCA3Kpovdama/bM9ryjFc0m1ZP6Cekn5kZpWmlJX6mKSbY+f9rKRzqhWOnd7xkj6S1xgzeySE0DpF7nVlU1/ypsj9LIQwRNmaMeda7Qmz6ERF9DFmWiRpUWfvpzO1t89pZ38j0edUtQN9Dv1NnbMsVaTDBhUdwaymaaCdjmOcbtffSB3Y58TLmgMAAAAAOhtTEAEAAACgIAzAAAAAAKAgDMAAAAAAoCAMwAAAAACgIAzAAAAAAKAgDMAAAAAAoCAMwAAAAACgIAzAAAAAAKAgDMAAAAAAoCAMwAAAAACgIAzAAAAAAKAgPYvYydChQ23cuHFF7AqdaOnSpVq1alXoqPr4XODxxx9fZWbDOrLO7vi52rx5cxI3NjZ2+D5aWlqSuKmpqcP30ZE6ur+RusdnY/Zs6fXX88v17CkdeGDnt+etpqP7nO7wmcqzatWqJH69hg9gjx7p7/e9e/dO4oEDB+50u4r0du1v0Lmq9TeFDMDGjRunmTNnFrErdKJDDz20Q+vjc4EQwvMdXWd3+Fxt27YtiZcuXZrE48eP3+k6Gxoaknju3LlJPHHixCQOoUOPPXZaR/c3Uvf4bNT6Z3j9danOX0pd6ug+pzt8pvJcd911SbxmzZokrjQg69+/fxKPHj06iU8++eSOaVxB3q79DTpXtf6GKYgAAAAAUJBCzoABAN60devWJP7Tn/6UxHlnwMysbJs/4+UtX748iSdNmlS1PIDOV+m7nHc22j/Hn6Hq1atXEvuz4z17pod+fspzLWfDfRk/xfnEE09M4rvvvrtqff41+DYCbzWcAQMAAACAgjAAAwAAAICCMAADAAAAgIIwyRYACtanT58kvv7665PYX8J5ypQpSVxLjsYvfvGLJP7GN76RxCeccEJuHQA6Vy05YG+88UYS+0vA+5wv77zzzktin/M1cuTIJPaXlN+0aVNZnVu2bEni5ubmJJ41a1bVNnk+5yvvqq5Ad8cZMAAAAAAoCAMwAAAAACgIAzAAAAAAKAg5YABQML8O2IwZM5L4scceS+LJkycn8TnnnFNW52WXXZbEPm9j4sSJ7W4ngM7l87mk8v4hL8dr+vTpSfzVr341iZcsWZLEgwcPTmKfczZq1Kgk9msISuU5Wr4On9vm88w+/elPJ/GFF16YxOR84a2OM2AAAAAAUBAGYAAAAABQEAZgAAAAAFAQcsAAoGA+p2PEiBFJ/PrrryfxggULkvjcc88tq9OvLTZo0KAkHjZsWLvbCaBz+TW+pPycr9NPPz2Jb7311iTu379/Evft2zeJfb7W+vXrk3jFihVV9y9JLS0tSdzU1JTEPkds8+bNSXzRRRcl8ZVXXpnE11xzTRKfeuqpSez7SKl8LTGgnnEGDAAAAAAKwgAMAAAAAArCAAwAAAAACsIADAAAAAAKQsYiAHQxn8C+bNmyJG5ubk7igQMHltXR2NiYxH4h5n79+u1ECwF0lfvuuy+Jb7/99iQeO3ZsEvuFnCtdsKLUli1bknjp0qVJvP/++5c9x19UY82aNUnsLwrkY98f+TZ/8IMfTOIpU6Yk8V577VXWJr/4s7/YCFBPOAMGAAAAAAVhAAYAAAAABWEABgAAAAAFIQcMALqYz7FYsmRJEuctzFqpjM8BGzVqVNXnkz8BFK9Hj/zfwb/73e8mcUNDQxL7HC+/CLL/bvvFn33f4ePly5eXtcnnnOb1H/5x32a/T/++fPzjH0/iO++8s6xN9FnoTjgDBgAAAAAFYQAGAAAAAAVhAAYAAAAABSEHDAA6WV5+hF8Tp2fPtGuuJT9r+PDhSbx69eqqdQCoT/67+uCDDyZx3759k9ivoZWXf+XL+3wun2Pmc8YkacOGDUns1zL0+8zrf3xO2IABA5L4gQceSOK5c+eW1TFp0qSq+wDqCWfAAAAAAKAgDMAAAAAAoCAMwAAAAACgIOSAAUAny1ufZvHixUmctzbQ5s2by7a99tprSTxkyJAkfv7556vWyRo6QH245ZZbkvjVV19NYp8f5XO0/Hd5l112SeKNGzcmsc8J8+uI+RzVSvv0fVKfPn2qtikvJywvh+xrX/ta2XO+//3vV60THWPECGnlyuplhg+XXnqpmPZ0V5wBAwAAAJArb/BVa5m3OwZgAAAAAFAQBmAAAAAAUBBywNpw7bXXJvG8efOqPl6LWtbyAfD2c9999yXxmDFjkrhXr15JXGldHs/3LwsWLNjB1gEo0h/+8Ick9uty+Zwtr3fv3knc0tJS9fm+f/Frcg0cOLDq/qTy4xufR+bzWvOOh3wb/HswY8aM3DYB9YwzYAAAAABQEAZgAAAAAFAQBmAAAAAAUJAuyQHz85GbmpraVV4qn+Ocx88f9n75y18m8fLly5N41113TeKzzjorib/85S+X1bn77rsncV7Ol58z7eW9BgDdw6JFi5J42LBhSdzY2Fj1+X5dH6m8f/HxihUr2tNEAF3kiSeeSOK8/Cl/POS/+5s2bUpiv0aXz7eqpe/wZfKOybZs2VK1vG+Df42+T+zbt2/V/QH1jjNgAAAAAFAQBmAAAAAAUBAGYAAAAABQkC7JAfP5U+edd14SH3PMMUmclyPWEfy6XocddlgS+/nKo0ePTuJbbrmlrE6fN3byyScncXNzcxL7HC+fE+bnRO8I1h4Dup7P8fD5Ef57mrduj1Se5+HzRl588cV2txNA8ZYsWZLE/tjAHwv4dQH9d79nz/RQLy/fypf39UnlfZKv0/N15JX3xz++TevXr6/6fKDecQYMAAAAAArCAAwAAAAACsIADAAAAAAKUkgO2BtvvKENGzZsj30uwh133JHEGzduTOKJEyeW1Tl48OAk9mtC+DnRL7zwQhLfdNNNSTxixIgkHjp0aBLfeeedSXzSSScl8Zo1a8raOH369CResGBBEu+5555JfPzxxyfx2LFjy+psj0rriuXNFWetMaDzPfLII0nsv4d5+Z+VcjJ8GZ9HNnLkyCRevHhxEu+1115VWgygKCtXrkxifzyys/lXeWsG+vr8cUOl5/h9+Of4vFX/Gtqbn7506dKybevWrUviAQMGtKtOoEicAQMAAACAgjAAAwAAAICCMAADAAAAgIIUkgPW0tKiefPmtfl4aX6YJN18881JPHny5LLn+HW5fOzzG+bOnZvEft2do446Kon9Oj0nnHBCEvucM79/STrxxBOT+OWXX07iZ555Jon/+Mc/JvF+++2XxAcccEASH3rooUk8bNiwJK6Uz0WOF9D15s+fn8Q+f8L3J37Nm0r5Ej5vI28tsdWrVycxOWBAffD5nP7/7bx1A30OaV7Ol+fzt3wOmlSeq+9j3+ZKeWSlfJvzyleycOHCJH7HO97R7jqAonAGDAAAAAAKwgAMAAAAAArCAAwAAAAAClJIDti2bduSdbJeffXVtBE902asXbs2iW+77bayOgcNGpTEPr+hubk5iY888sgknjBhQhL7HAy/9tiqVauS2M939uuSSeWv0+eNjRkzpmrs17SYMWNGEj/22GNV6x84cGBZm/zaYrvuumsS77vvvknc2NhYVgeAnePXsPE5Xz6fy8e+z5TK80I8X8eiRYuS+PDDD6/6fAAdb9myZbllfM6WzxHraL7+SvlYvj/xx2CV1iqsxj/f94m1vObnnnsuickBQz3jDBgAAAAAFIQBGAAAAAAUhAEYgNSIEVII1W8jRnR1KwEAALolBmAAUitXdkwZAAAAlCnkIhw9evRQv379tsd+AeJzzjkniceNG5fE/mIWkrRp06Yk9hec6NOnT9Xyc+bMqdrm/v37J7G/wIVPeH/ppZfK6vBJpAMGDKhah7/ohl/8sNKFPkr51+gXfpak5cuXJ7F/XV/60peS+Mwzz9x+v/RCKgB23AsvvJDE++yzTxL7hHSv0kKq/sIcPnHeJ8X7xekBFM8vHlyLvO/2zvKLKPtF2yuV8cdcvo15bfYX9fDHT7UszLxixYrcMkC94AwYAAAAABSEARgAAAAAFIQBGAAAAAAUpJAcsDVr1uiOO+7YHo8cOTJ53Ocl+VyoPffcs6xOv2ixnz/s69y8eXMSb9u2LbfNpfzi0L169Upiv6CxlJ8D5vl8rOHDhyexb7PPIfNzsH0slb+3/n3yuSVXXXXV9vsrufACsEP8d9fna/p8iLxFlSvlQ/jvru/zfM5GpbxVAMV69tln2/0cn+/pFyn2fYHvX/LKe42NjWXbfJ/m+6T2tsnHvnwtOWCvvPJKbhmgXnAGDAAAAAAKwgAMAAAAAArCAAwAAAAAClJIDtjmzZu1ePHi7fH48eOTxydOnJjE8+bNS+IXX3yxrM689azy5gv7x33OhY/9fGQ/J7rS3GNfpqmpKYl9Hpm3atWqJPZtfu2115LY5635x6Xy9c18LsqiRYvarDMvLwU7ZurUqTWVmzZtWie3BJ3l+eefr/q47782bNiQxL6vqNS/+ZwLH/ucUL8WGYDiVVqvM48/HvH5WD7fs71831FLf+PblNdGf3zkc8D8Woh5x0tS5TVjgXrFGTAAAAAAKAgDMAAAAAAoCAMwAAAAAChIITlgPXr0SHIcHn744eTxvPWsKuUebdy4MYn9GltDhw5N4vXr1ydx3jpgfg61X3fDx37+slS+Dpjn5zT7/Cw/N9y/Zr+ml8/n8nOopfJ2+/XTfB2XXnrp9vsXX3xxWX0A8i1YsKDq43n5D74vqdTf+D7N52j47/6yZcuqtglA51uyZEluGf9998cnLS0tSVxLvlQ1Pudrt912KyuzevXqJPbHDj4HzPc//rhu0KBBVev3r8nXJ7EOGLoXzoABAAAAQEEYgAEAAABAQRiAAQAAAEBBCskBGzNmjK655pokLjV48OAk9utfVcoB8zkRPj/KrwfR3NycxD73yc+x9vOL/ZxoP+faz3eu1Eb/OvL2mfe4f98GDhyYxD6XrtJz9tlnnyQ+/vjjy57T6uqrr27zMQBta2++le+fvFrW5fF5ZL6PqrROIIBi+eMdqfz/ev9d9t9/X973BV7emoH+2GPFihVldeSttZp3/LJ27dokPvbYY5P4rrvuSmLfJ1bKc/N5Y0A94wwYAAAAABSEARgAAAAAFIQBGAAAAAAUpJAcsIaGhmSNh8svv7yI3QJAXfD5Vu3N8fD5D5XyLyrloZbyORl5eWYAOl+lXEyf3+Rz3MeOHZvEPsf9kUceSeJRo0Yl8ebNm5M4r+/Ie7wS30f5/sevzerlrQtWaR2wvPVdgXrCGTAAAAAAKEghZ8AAAAAAdK6pU6fmlpk2bVoBLUE1nAEDAAAAgIJwBgwAOplfB8znePh8CZ/LUEsOhs+x8LHfh88D8XloldbZAdCxKuWANTU1JbFf13TKlClJ7POhHn744ST263zl9Se+fC35or7OvNjvw+d8TZgwIYl/85vfJPHQoUPL2pC3NhlQTzgDBgAAAAAFYQAGAAAAAAVhAAYAAAAABSEHDAA62bp165K4sbExiX0+hNfQ0JBb3udY5OWEeT7PZPjw4VXLA9h5ldauysu/PPbYY5N4/vz5Vcvnffc935f4dcak8rXJdnZdwSFDhiSxz/HyOWCVXlNePwrUE86AAQAAAEBBGIABAAAAQEEYgAEAAABAQcgBA4BOtn79+iRu7xpbPrehUq6DzxPL24df92vNmjVJTA4Y0Pl8PqhUOS+s1EknnZTEs2bNqlref9f9ell5awZW6m+2bNlStQ7/HL/uoNe7d+8kPvroo5P4iiuuSOJKa5kNGDCg6j6AesIZMAAAAAAoCAMwAAAAACgIAzAAAAAAKAgDMAAAAAAoCBfhAIBOtmnTpiTu169fEvukex/7pPhKi576ZH5/UQ6fNL/HHntUbSOAzucvPlFJ//79k9gvUrxhw4Yk9heo8P2Hj/O89tprZdv8RTd8n+XbUOmiGaX8BTR8f+b7vEqvIe/iJUA94QwY8BY3YoQUQvXbiBFd3UoAAIC3BwZgwFvcypUdUwYAAAA7jwEYAAAAABSEHDAA6GQPPfRQEjc3N1ct39TUVDWulDfiF172ORd+YVSf87Vw4cIkPvDAA6u2EcDO8/mgUvnC7Xn5mf677/OnfL6Wj32+aF4OWaUyPvb76NkzPdzs06dPEq9bt65q7FVawHrIkCFVnwPUE86AAQAAAEBBGIABAAAAQEEYgAEAAABAQcgBA4BO9tGPfjSJr7jiiiT2a3T5dXdWrFiRxIMHDy7bx9atW5PY54n5vLONGzcm8aBBg8rqBNC5pk+fXrZt1apVSdzS0lK1jsWLF7drn3nrDvp8UZ+/JZXnfPk8Mr9ul6/TmzNnThJ//vOfb9fzge6GM2AAAAAAUBAGYAAAAABQEAZgAAAAAFAQcsAAoJNddtllSTxp0qQkfuqpp5LY53xMmDAhiadMmVK2D5/T1bdv3yT263ydfvrpbTcYQJcZOnRou8r7fE+/xpZfJ8zHPn/U51v5+qT8vLG88j4ndd999636fOCthjNgAAAAAFAQzoABAAAA6HJTp07NLTNt2rQCWtK5OAMGAAAAAAUJRaytEEJ4RdLznb4jdLaxZjasoypr5+diqKRVuaUo393Kd+hnSqK/eQvpys9GvXw/KN/x5bvy/zHUr67+v6ieviPdvXw9taXNz1UhAzBgZ4UQZprZoZR/a5YH6km9fT8o37HlgXpTb9+R7ly+ntpSDVMQAQAAAKAgDMAAAAAAoCAMwNBdfI/yb+nyQD2pt+8H5Tu2PFBv6u070p3L11Nb2mZm9XOTtpk0y6TZJj1h0l90cP2nmGQmHRrjXib9wKS5Jj1t0mfi9mEmPWjSPJPeW/L8X5i0W5X6LzTprHj/CJMeia/naZMu6YT3a6lJQ3fwucNM+lWX/825ceuCm2TbJJsl2XzJZkv2Scl6FLTv98X9viHZoe6xz0i2WLKFkp1Qsv3EuG2xZFNLtt8s2RzJLi/Z9jnJ3ltl/wdJdkO8f7Zkr0j2pGSLJLtHso7td9N9D5OMfocbt5KbZCMk+4lkSyR7XLLpkk3YgXoGSvbvbTy2T+zzWm/rJLswPnalZAtiX3KbZAPj9nfGbTMl27tkH/dW6y8l+6lke8b7/SX7bslru1+yw3fwffpsyf3ekj0gWc+u/vtx47Yjt3o7A9YisykyO1DSZyRd0WE1h9As6QJJj5RsfZ+kRplNknSIpI8ohHGSTpf0HUmHSbowPv8fJD0ps+Vt1N9T0gcl/Shu+YGkD8tsiqSJkm7tsNeys0LoKbNXJK1QCO/s6uYAXaDFTFPMdICk4yX9raSLfaEQOmWtxHmS/lHSA25f+0s6TdIBkk6UdG0IaghBDZK+Fdu4v6TTQ9D+IWhyfB2TJb0jBO0SgkZKOtxMt1fZ/2clXV0S32Kmg8y0t6Rpkn4egvbzT+qI98JMr0haEYLodwBJIShIuk3S/WYab6ZDlB3/DN+B6gZK+vdKD5hpYezzpig73tkY9ytJv5Y0MfYlz8T9S9InJf2dsuOgj8Ztn5N0uZneaOP1HCCpwUzPxk3XS3pV0t7xtZ2j7CpyO+KzJa9ni6TfSnr/DtYFdKl6G4CVGiDpz5KkEPorhN8qhCcUwlyFcNL2UiF8XiEsVAgPKoQfK4RPtVHfFyV9RdKmkm0mqV8cPDVJ2iJpnaStkvpKapS0LT5+oaT/qtLev5L0hMxej/GuklZke7FtMnsqtvcShXCjQrhfITyrEM4veS1nKoRHFcIshfBdhdAQt39bIcxUCPMVwqVlew6hSSHcrRD+VSH0i/U/qhCe3P5ehXC2QrhDIfxOWaclSbdLOqPKawLe8sz0sqQPSzovBIUQdHYIuiME/U7Sb0NQvxB0Ywh6NAQ9GYJOkrIDjbhtVgiaE4L2jmXvCkGzQ9C8EMoPDsz0tJkWVmjKSZJ+YqbNZnpO0mJlPwIdJmmxmZ6NBx0/iWW3SmoKQT0k9ZK0TdJlqjCQbBWCmiVNNtPsNt6L+5RNr/hwLH9/CPp6CJop6YIQdEgI+n0IejwE3RMHfApB54egp+L78JO47Zj43syK71tz3M3tot8BWh0raauZvtO6wUyzzTQj9kdXxr5kbmt/EoL6h6DfhqAn4vbWY6JpksbH79yVVfZ5nKQlZtkl0s10r5laj10eljQ63m89FuoraWsIGi9pdzPdX6XuMyT9IrZzvKTDJX2udcBmpufMdFd8/BPxtc0LIf7YnW2/PfYx80PY3hdNU9bfzQpBN8eit4u+BN1VV5+CS25vTkFcYNJakw6J23uaNCDeH2rSYpOCSe+I5fuY1GzSIpM+VaHeg036Wbx/v5uC+BOTXjFpg0kfjtt3Mekuk2aadJxJ55t0dk7bLzXpYyXxF0z6s0m3mfQRk/rE7ZeY9AeTGuNrWR3bsZ9Jd5rUK5a7tmQ64+D4b0Ns/+QYLzVpnEm/KSl7uUlnxvsDTXrGpH4mnW3Si9vryh4fZdLcLv+7V7kpOxOwUNnB6NQayt8o6WVJ82oou7uk+yQ9JWm+pAtyyveR9Kik2bH8pTW+hgZJT0r6ZQ1ll0qaK2mWpJk1lB8o6aeSFkh6WtKRVcruE+ttva2TdGFO/R+Pr3WepB+r9XPcdvkLYtn5eXV37efK1lfYtkay4XFa3ouSDY7bL5fszHh/oGTPSNZPsmskOyNu7y1Zk2SnSHZdSZ27VGnD/aVTECX7Zut+YnyDZKfG2/Ul2/9Zsm/G+1+P04k+KdmU1qmFVfZ5rGQ/K4nPbq2rZNt7Jbu7pI3Xxvu9JPuDZMNi/H7Jboz3l0vW2PoexX/vlOyd8X7/1qlCko2SrG77nfb0Oe3pb2J5+pwO7HO6S39T/fXa+ZL9dxuPnSLZryVriH3TC5KNlKynZANimaHKpiYHycZJVsP/fXajZOe18didJf3dFMkeluw+yUbHaZJ759T9e8kmxfvvkey2NsodItnc2Jf2VzYt+6D4WGvf2yTZPMmGxHi9q6NBsle6+m+4c39/jnFyytdNfxPLd1if0+UfvuQmrS+5f6RJ8y0baPUy6ZsmzbFswNVi0gjLcq4uLXnOVeYHYFKPOGgZF+PSAdg7Tbo51r+rSQtN2tM9f5BJvzapv0nXmfRTq/QBkL5n0mlu23iT/s2k35t0f9x2iUkXlZR52qTRJp1n0vL4+mbFtlwSy3zUspy4OZYNFk+L25dali93Rkl9My3LXWut5wXLBndnm3STa18vk1Z3+d+97Q96g6QlkvaU1Dt2CvvnPOdoSQfX2DmNlHRwvN+sbOpFm/VLCpL6x/u9lE1nPaKG/XxC2dTUWjunmvP6lE11/VC831vSwHa8ty8pWySwrTKjJD0nqSnGt6rKDxHKptrOU/ZraU9Jv5G0V1d/jiq3NXcAdlPJ9pnxIGBWvL0g2X6SfSAeNPyn3syPmCDZUsm+ItlROW24f2cHYK6+OyXbTbKLJLtVsn+tUOYDkn2nJD67wgDsZDcAOyben6gsb6T1fZgr2b3xsV8py/s4U7L4HbGpkj0SDzBHl9TfS7K67Hfa2+e0p7+J5elzOqjP6U79TfX3peoA7L8l+2BJ/L9xUNMr9hdz4nexRVke2bi8AZiyH4tWSTa8wmMXKcsBCxUeOzq2Z4Jkt0j2wzbqeEayEfF+tQHYBZJdVhJ/UbLz4/1LlOXmzpZsrWRHxO2V+u1lkjV39d9xx/72HOPUUL4u+pv4eIf2OfU7BdHsj8rmCQ9Tdop5mKRDlOVUrVQ2Uq9Fs7I37X6FsFTSEZLuUAiHSvqApF/JbKvMXpb0kCS/uNrnJX1ZWV7Yg5L+RdIlFfbTUtYmsyUy+7ay0/0HKoQh8ZHNJaW2KftDBkk/UJYDN0Vm+8jsEoWwh6RPSTpOZpMl3eX285CkExVCiHGQdEpJPWNk9nR8bINrc5/Y7noVp17Zs2ZWOvWqTWb2gLL55rnMbIWZPRHvv6bs15VRVcqbma2PYa94s2r7CCGMlvRuZfPgO1QIYRdlnfENsX1bzGxNjU+PU1Ds+ZxyPSU1hWwabl9JlXMgM/tJesTMNlo2Fff3ynKd6l4I2lPZd/HluKn0uxIknWIxf8JMYyybRvgjSe9R9h2aHoL+ykzPKPvPca6kL4WgL7SjGcuU/WLZanTc1tb20vafJOlxSf0ljTfTP0k6NQT1dfso76fKHaTsu9Cq9b0IkuaXvA+TzPQ38bF3K8tTO1jSYyGop5mmSfqQsundD4WgfWPZeu532tXntKe/ieXpczquz+m2/Y0zX1lOVntsPyayLKerPcdEfyvpCTOtLN0Ygs6W9PeSzjBLP2MxT+1zylI5Lpb0H5Kuk3S+ypX2MfMlHRjzWGsSgt4l6a8lHWmmA5WdWan22hqVppZ0JxzjVK+7nvobqYP7nPodgIWwr7IR7GpJu0h6WWZbFcKxksbGUg9J+geF0Ech9FfWeaTM1spsqMzGyWycsvnN75HZTEkvKMvdkkLop2xwtqCkDXtLGi2z+5X9Yd5Q9mFsqtDipyXtVfLcd5cMivZWdnC3psor/q2kUxXCrvH5gxXCWGW5cBskrVUIw5V1nqW+oCxX7lsxvkfSx7bvO4SDquxzgrLRfL0aJelPJfGLqtJ57IyQXXzlIKUXaalUriGEMEvZgfqvzaxqeUlfV/afVcWE5QpM0r0hhMdDCB/OKbuHpFck3RRCeDKEcH3IPse1OE3Z6fa2G2K2TNJXlX1PVkhaa2b3VnnKPElHhRCGhBD6Kkve3r1K+boQgoYpu+jON/2BR3SPpI/FgxCFoIPiv3tKetZMVyvLeZgcgnaTtNFMP5R0pbIBSa3ukHRaCGoMQXso6zcelfSYpL1D0B4hqLeyv90dJe3vpTdzVJv05n+YDcp+MSyV9lNOCDpGWf7XdRUeXihpWAg6snW/MQ+uh7K8kPsk/aey/rp/CBpvprlm+kp8Da0DsHrud+hzqqunPqdb9jcV/E5SY2uukySFoMkh6ChJMyS9P16MZ5iyg9FHFY+JzLQ1BJUeE70mbc+1bMvpcn+HEHSiss/Me8y0scJzzpI03Uyv6s1joTfifW97H2OmJZJmSrq0pP8cF4LeHV/be0NQ3xDUT9LJcdsukv5spo3xR5sjSureGvu71nYPkbTKTFtzXnO9or+prp76G6mD+5x6G4A1xQtQzJJ0i6R/kdk2STdLOlQhzFXWEWSDJLPHlB2IzJF0t7Jfnde2Y3/fktRfIcxXdoBwk8zmlDz+ZUkXxfs/lvRvsdw3KtR1t7LOsdU/S1oYX8v/SjojvpbKsot0fE7SvQphjrKrEo2U2WxlvwAtUHaK96EKz75A2Xv3X8p+oeolaU58XV9sc59Z8u9dVR5/WwjZ4P1nyubzrqtW1sy2WXYWdrSkw0IIE6vU+/eSXjazx9vRnL80s4OVDbTPDSEcXaVsT2UH+N82s4OUDdSn5u0ghNBb2Zmb/59TbpCyX+P2kLSbpH4hhDPbKm/ZmdavSLpX0q+UzcFu+zPftVqTuecrm0Zwr6TyC9xktn+nYvnW79Q/SZoXgmYpO8v+P5ImSdmFOZT9UvwlX1kIOjkEvSjpSEl3haB7JMlM85VNgXhK2ft3rpm2WZYcf56ygeDTkm6NZVudK+kH8cBpjqS+IWiupMfN0h99zLRA0i4lF8SQsgO8WSHoGWVXGTvFLDkD1vrcLZJOlfSVEDRb2d/3L5QN9H4Y9/mkpKvjfi+MyfVzlCXz3x2retv3O/Q5bZaruc/pZv1Nm+KPPidL+usQtCT2MVcomz51m7Lv9GxlA7X/MNNLisdE8Tu3/ZjITKuVnW2eV+kiHHGgc7ykn7uHvqls4Pbr2Bd8p+Q5fSWdrTd/5L1K0nRlB97fUbm7JL2rJP6Qsis6Lg5B8yR9X9ng8Yl4/1Flg4LrzfSksr9lzxD0tLKLijxcUtf3lPXDrRfheNv3JbWgv2mzXNce4+zo3MW6ucX5qib1tSz/6eAubMttJlVNUK2rm/SASYO6vB1t3JQdoN5TEn9GrWu1VX/eONWek9FL2YHtJ3agfV9QpYu+vPn4Fcp+0Vqq7D/TjZJ+2I76L8mpf4SkpSXxUZLuqqHekyTdW0O590m6oSQ+S9K17Wj/5ZIqrknDretukn1csg914f4fkKwu+50d6XPa09/E8vQ5bZfb4T6H/qY+bsounPGwZA0F7Ovn2oH10urlxjFO9+1vYvmd6nPq7QzYjvhePMv0hKSfKc537SJTlSU91r8Qhkm6SmZ/7uqmVBGnXoU94i8aydSrnRWyaZo3SHrazK6qofywEMLAeL9J2S+JC9oqb2afMbPRlk19PU3S78yszV9XQgj9QrZeneJp9r9RlalaZvaSpD+FEPaJm45TdvYkT9kUlDa8IOmIEELf+F4dJ5WfGSkV4hTaEMIYZXOjf1StPLrEt5XmoRYmTqO6ykz12u/Q53SjPof+pv6YqUXZ2f9OmUrXKk7Hvt2yvNvuiv6mG/U3sd0d1+fs6MiNG7cibsrm2D6j7EpBF9VQ/sfK5vJuVfbLzP+rUvYvlc1HnqM3L1v6d1XKT1Y2xWqOsk7jC+14He9SzhWClF0JabbevARsLa93irI59nOUrYlS9cyCpH6KeZU1tvtSZR3wPGVTaRtzys9Q1kHOlnRcV39+uHFr7609fU57+ptYnj4nv/6a+xz6G27d/cYxTm69ddPfxPId1ueEWCEAAAAAoJO9FaYgAgAAAEC3wAAMAAAAAArCAAwAAAAACsIADAAAAAAKwgAMAAAAAArCAAwAAAAACsIADAAAAAAK8n9AvOjkJ6qnywAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 864x720 with 30 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Plot the first X test images, their predicted labels, and the true labels.\n",
    "# Color correct predictions in blue and incorrect predictions in red.\n",
    "num_rows = 5\n",
    "num_cols = 3\n",
    "num_images = num_rows*num_cols\n",
    "plt.figure(figsize=(2*2*num_cols, 2*num_rows))\n",
    "for i in range(num_images):\n",
    "  plt.subplot(num_rows, 2*num_cols, 2*i+1)\n",
    "  plot_image(i, predictions[i], test_labels, test_images)\n",
    "  plt.subplot(num_rows, 2*num_cols, 2*i+2)\n",
    "  plot_value_array(i, predictions[i], test_labels)\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "R32zteKHCaXT"
   },
   "source": [
    "## Use the trained model\n",
    "\n",
    "Finally, use the trained model to make a prediction about a single image."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:48.124975Z",
     "iopub.status.busy": "2022-02-05T02:23:48.124437Z",
     "iopub.status.idle": "2022-02-05T02:23:48.126215Z",
     "shell.execute_reply": "2022-02-05T02:23:48.126580Z"
    },
    "id": "yRJ7JU7JCaXT"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(28, 28)\n"
     ]
    }
   ],
   "source": [
    "# Grab an image from the test dataset.\n",
    "img = test_images[1]\n",
    "\n",
    "print(img.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "vz3bVp21CaXV"
   },
   "source": [
    "`tf.keras` models are optimized to make predictions on a *batch*, or collection, of examples at once. Accordingly, even though you're using a single image, you need to add it to a list:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:48.130399Z",
     "iopub.status.busy": "2022-02-05T02:23:48.129816Z",
     "iopub.status.idle": "2022-02-05T02:23:48.132539Z",
     "shell.execute_reply": "2022-02-05T02:23:48.131963Z"
    },
    "id": "lDFh5yF_CaXW"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1, 28, 28)\n"
     ]
    }
   ],
   "source": [
    "# Add the image to a batch where it's the only member.\n",
    "img = (np.expand_dims(img,0))\n",
    "\n",
    "print(img.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "EQ5wLTkcCaXY"
   },
   "source": [
    "Now predict the correct label for this image:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:48.136943Z",
     "iopub.status.busy": "2022-02-05T02:23:48.136102Z",
     "iopub.status.idle": "2022-02-05T02:23:48.174620Z",
     "shell.execute_reply": "2022-02-05T02:23:48.175042Z"
    },
    "id": "o_rzNSdrCaXY"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[7.8192167e-04 1.4889931e-12 9.9545813e-01 2.3544904e-14 2.7742677e-03\n",
      "  1.0530308e-15 9.8571123e-04 1.0706052e-16 1.0197677e-13 2.7946701e-12]]\n"
     ]
    }
   ],
   "source": [
    "predictions_single = probability_model.predict(img)\n",
    "\n",
    "print(predictions_single)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:48.209829Z",
     "iopub.status.busy": "2022-02-05T02:23:48.188654Z",
     "iopub.status.idle": "2022-02-05T02:23:48.261300Z",
     "shell.execute_reply": "2022-02-05T02:23:48.261992Z"
    },
    "id": "6Ai-cpLjO-3A"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_value_array(1, predictions_single[0], test_labels)\n",
    "_ = plt.xticks(range(10), class_names, rotation=45)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "cU1Y2OAMCaXb"
   },
   "source": [
    "`tf.keras.Model.predict` returns a list of lists—one list for each image in the batch of data. Grab the predictions for our (only) image in the batch:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "execution": {
     "iopub.execute_input": "2022-02-05T02:23:48.267791Z",
     "iopub.status.busy": "2022-02-05T02:23:48.266710Z",
     "iopub.status.idle": "2022-02-05T02:23:48.270407Z",
     "shell.execute_reply": "2022-02-05T02:23:48.271167Z"
    },
    "id": "2tRmdq_8CaXb"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "2"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.argmax(predictions_single[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "id": "YFc2HbEVCaXd"
   },
   "source": [
    "And the model predicts a label as expected."
   ]
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "name": "classification.ipynb",
   "toc_visible": true
  },
  "kernelspec": {
   "display_name": "Python 3.9.12 ('base')",
   "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": "3.9.12"
  },
  "vscode": {
   "interpreter": {
    "hash": "edb042ae9075e3557312c0e21ddc2bba4f09b37b77c6722de2d98e505549e3df"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}