Autoencoder_with_Sonnet_and_TensorFlow_Datasets.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Autoencoder with Sonnet and TensorFlow Datasets.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/jarrydmartinx/generative-models/blob/master/Autoencoder_with_Sonnet_and_TensorFlow_Datasets.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "yF4mttKxUPlF",
        "colab_type": "text"
      },
      "source": [
        "# An Undercomplete Autoencoder with Sonnet with TensorFlow Datasets \n",
        "* Let's make an (undercomplete) autoencoder with Sonnet using TensorFlow Datasets"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gggYOmxiTwsU",
        "colab_type": "code",
        "cellView": "both",
        "colab": {}
      },
      "source": [
        "#@title Install\n",
        "! pip install -q plotnine"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hAITsMZSVeNq",
        "colab_type": "code",
        "cellView": "both",
        "outputId": "ff4dcb4a-2b3a-4acd-e0ff-a22f0dfb727f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 136
        }
      },
      "source": [
        "#@title Imports\n",
        "\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt\n",
        "import pandas as pd\n",
        "import plotnine as gg\n",
        "import sonnet as snt\n",
        "import tensorflow as tf\n",
        "import tensorflow_datasets as tfds\n",
        "from collections import namedtuple\n",
        "\n",
        "\n",
        "from typing import Text, List, Tuple"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "\n",
            "WARNING: The TensorFlow contrib module will not be included in TensorFlow 2.0.\n",
            "For more information, please see:\n",
            "  * https://github.com/tensorflow/community/blob/master/rfcs/20180907-contrib-sunset.md\n",
            "  * https://github.com/tensorflow/addons\n",
            "If you depend on functionality not listed there, please file an issue.\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "P3Y5Lh2HwAKh",
        "colab_type": "text"
      },
      "source": [
        "## Getting our Data using TensorFlow Datasets"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "eMHamshhgIH-",
        "colab_type": "text"
      },
      "source": [
        "Here's some friendly stuff about tfds"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UjpjCHlXV4nK",
        "colab_type": "code",
        "cellView": "both",
        "colab": {}
      },
      "source": [
        "#@title Preparing our data\n",
        "\n",
        "def get_data(dataset_name: Text, batch_size: int = 25) -> tf.Tensor:\n",
        "  \n",
        "  # This downloads and prepares the dataset, or retrieves it from the cache.\n",
        "  dataset = tfds.load(dataset_name, split=tfds.Split.ALL)\n",
        "\n",
        "  # Pipelining operations: repeat the dataset infinitely and batch it.\n",
        "  dataset = dataset.repeat().batch(batch_size, drop_remainder=True)\n",
        "\n",
        "  # Make an iterator: this iterates through the batches.\n",
        "  iterator = dataset.make_one_shot_iterator()\n",
        "\n",
        "  # Data op, when run in a tf.Session, spits out a batch of data.\n",
        "  data = iterator.get_next()\n",
        "  \n",
        "  input_data = tf.cast(data['image'], dtype=tf.float32)\n",
        "\n",
        "  return input_data"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gxmGnGJIXnz9",
        "colab_type": "code",
        "cellView": "both",
        "outputId": "0e2dbc5c-83d6-4214-f559-4cef0dac2a57",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 523
        }
      },
      "source": [
        "#@title Visualising our data\n",
        "\n",
        "def show_img_sample(dataset_name: Text):\n",
        "\n",
        "  # Get a batch of images\n",
        "  data_op = get_data(dataset_name)\n",
        "  with tf.train.MonitoredSession() as sess:\n",
        "    batch = sess.run(data_op)\n",
        "  \n",
        "  # Print the shape of the batch of images\n",
        "  print('Image batch shape: {}'.format(batch.shape))\n",
        "  \n",
        "  # Retrieve and show a sample image from the batch\n",
        "  image = batch[0].squeeze(axis=-1)\n",
        "  plt.imshow(image)\n",
        "  plt.axis('off')\n",
        "  plt.show()\n",
        "\n",
        "show_img_sample('mnist')"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/ops/control_flow_ops.py:423: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Colocations handled automatically by placer.\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "WARNING: Logging before flag parsing goes to stderr.\n",
            "W0502 20:51:54.932442 140091387574144 deprecation.py:323] From /usr/local/lib/python3.6/dist-packages/tensorflow/python/ops/control_flow_ops.py:423: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Colocations handled automatically by placer.\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:Graph was finalized.\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "I0502 20:51:55.014315 140091387574144 monitored_session.py:222] Graph was finalized.\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:Running local_init_op.\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "I0502 20:51:55.695707 140091387574144 session_manager.py:491] Running local_init_op.\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "INFO:tensorflow:Done running local_init_op.\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "I0502 20:51:55.702118 140091387574144 session_manager.py:493] Done running local_init_op.\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Image batch shape: (25, 28, 28, 1)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAPsAAAD4CAYAAAAq5pAIAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "htksKIusgMXl",
        "colab_type": "text"
      },
      "source": [
        "## Making Models out of Sonnet modules\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ZRO3j-IjfW_j",
        "colab_type": "text"
      },
      "source": [
        "Here's some friendly stuff about Sonnet modules\n",
        "\n",
        "Sequential takes a list of callables that are functions from tensors to tensors.\n",
        "Think of Sequential as a scaffold to connect a bunch of pipes together. Each sonnet module is a pipe that alters the tensors flowing through it.\n",
        "\n",
        "sequential(List[Callable[[tf.Tensor], tf.Tensor]])\n",
        "\n",
        "From Conv2DTranspose: https://sonnet.dev/sonnet/#class-conv2dtranspose\n",
        "\n",
        "output_shape: Output shape of transpose convolution. Can be either an iterable of integers or a callable. In the latter case, since the function invocation is deferred to graph construction time, the user must only ensure that output_shape can be called, returning an iterable of format (out_height, out_width) when build is called. Note that output_shape defines the size of output signal domain, as opposed to the shape of the output Tensor. If a None value is given, a default shape is automatically calculated (see docstring of _default_transpose_size function for more details)."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6oJVi89Ba15v",
        "colab_type": "code",
        "cellView": "code",
        "colab": {}
      },
      "source": [
        "#@title Models\n",
        "def get_model(model_name: Text):\n",
        "  \n",
        "  conv_layers = snt.nets.ConvNet2D(output_channels=[16, 32, 64], \n",
        "                                      kernel_shapes=[3, 3, 3],\n",
        "                                      strides=[2, 2, 1], \n",
        "                                      paddings=['SAME'])\n",
        "  deconv_layers = conv_layers.transpose()\n",
        "  \n",
        "  models = {\n",
        "      'mlp': snt.Sequential([\n",
        "          snt.BatchFlatten(),\n",
        "          snt.nets.MLP([50, 25, 10]),\n",
        "          snt.nets.MLP([25, 50, 784]),\n",
        "          lambda x: tf.reshape(x, shape=[25, 28, 28, 1]),\n",
        "      ]),\n",
        "      'conv': snt.Sequential([\n",
        "          conv_layers,                       \n",
        "          snt.BatchFlatten(),\n",
        "          snt.nets.MLP([25, 10]),\n",
        "          snt.nets.MLP([25, 392]),\n",
        "          lambda x: tf.reshape(x, shape=[25, 7, 7, 8]),\n",
        "          deconv_layers,\n",
        "      ])\n",
        "  }\n",
        "  \n",
        "  return models[model_name]"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "iVcQop3lgeF3",
        "colab_type": "text"
      },
      "source": [
        "## Training our Models"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "24Nc3e0fJsrf",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Make a container for a batch of true/reconst. images for each trained model\n",
        "Reconstructions = namedtuple('Reconstructions', ['true_images', 'recon_images'])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "OJ7Ds88qtfAJ",
        "colab_type": "text"
      },
      "source": [
        " * All of the graph construction and calls to sess.run occur within this train function below"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vZWnZcAPmQbx",
        "colab_type": "code",
        "cellView": "both",
        "colab": {}
      },
      "source": [
        "#@title Collecting our Results\n",
        "\n",
        "# List the datasets and models we want to use\n",
        "datasets = ['mnist', 'fashion_mnist']\n",
        "models = ['mlp', 'conv']\n",
        "\n",
        "# Train each autoencoder model on each dataset\n",
        "results = pd.DataFrame()\n",
        "images = {}\n",
        "\n",
        "for dataset in datasets:\n",
        "  images[dataset] = {}\n",
        "  \n",
        "  for model in models:\n",
        "    print('Training {} model on {} dataset...'.format(model, dataset))\n",
        "    \n",
        "    result, recons = train_autoencoder(dataset, model, num_training_steps=2000)\n",
        "    \n",
        "    # Label the result by model and dataset and add it to our DataFrame\n",
        "    result['model'] = model\n",
        "    result['dataset'] = dataset\n",
        "    results = pd.concat([results, result])\n",
        "    \n",
        "    # Store the true and reconstructed images\n",
        "    images[dataset][model] = recons\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "CTaSaOotiLL_",
        "colab_type": "code",
        "cellView": "code",
        "colab": {}
      },
      "source": [
        "#@title Our train function\n",
        "\n",
        "def train_autoencoder(dataset_name: Text, \n",
        "                      model_name: Text,\n",
        "                      num_training_steps: int = 1000,\n",
        "                      batch_size: int = 25,\n",
        "                      learning_rate: float = 1e-3,\n",
        "                      log_every: int = 10) -> Tuple[pd.DataFrame, \n",
        "                                                    Reconstructions]:\n",
        "\n",
        "  tf.reset_default_graph()\n",
        "\n",
        "  model = get_model(model_name)\n",
        "\n",
        "  input_data = get_data(dataset_name) \n",
        "\n",
        "  # Make the output op, which connects our model to the data ops defined earlier\n",
        "  y = model(input_data)\n",
        "\n",
        "  # Define a loss op. We're using mean squared reconstruction error.\n",
        "  loss = tf.reduce_mean((input_data - y) ** 2)\n",
        "\n",
        "  # Pick an optimizer. Adam is a popular choice and usually works well.\n",
        "  optimizer = tf.train.AdamOptimizer(learning_rate=learning_rate)\n",
        "\n",
        "  # Create the stochastic gradient descent op. Computes and applies gradients.\n",
        "  sgd_op = optimizer.minimize(loss)\n",
        "  \n",
        "  # Make and initialize a session so we can run our graph ops\n",
        "  sess = tf.Session()\n",
        "  sess.run(tf.global_variables_initializer())\n",
        "\n",
        "  # Training Loop\n",
        "  results = [] \n",
        "  for step in range(num_training_steps):\n",
        "    sess.run(sgd_op)\n",
        "\n",
        "    if step % log_every == 0:\n",
        "      \n",
        "      result = {\n",
        "          'step': step,\n",
        "          'loss': sess.run(loss),\n",
        "      }\n",
        "      results.append(result)\n",
        "  \n",
        "  # Collect a batch of true/reconst images using the trained model\n",
        "  true_images, recons = sess.run([input_data, y])\n",
        "  \n",
        "  images = Reconstructions(true_images=true_images.squeeze(-1), \n",
        "                                  recon_images=recons.squeeze(-1))    \n",
        "  \n",
        "  sess.close()\n",
        "  \n",
        "  return pd.DataFrame(results), images                      \n",
        "                      "
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Z7qGOz_svIgJ",
        "colab_type": "text"
      },
      "source": [
        "## Plotting the Loss and Viewing the Reconstructed Images"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "93gPX2h9NdpV",
        "colab_type": "code",
        "cellView": "both",
        "outputId": "33f436c2-0e18-4baf-cbd0-8372bb2d6a36",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 449
        }
      },
      "source": [
        "#@title Plotting the Reconstruction Loss\n",
        "\n",
        "#Plot our reconstruction loss results\n",
        "p = (gg.ggplot(results)\n",
        "     + gg.aes(x='step', y='loss', color='model')\n",
        "     + gg.theme(figure_size = (8, 4))\n",
        "     + gg.facet_wrap('dataset')\n",
        "     + gg.geom_line()\n",
        "     + gg.ggtitle('Reconstruction Loss')\n",
        "    )\n",
        "\n",
        "p"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAyYAAAGfCAYAAACqbcamAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 800x400 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<ggplot: (8755565938664)>"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "r24f5CkJvOYp",
        "colab_type": "code",
        "cellView": "both",
        "outputId": "93ef1fb9-e08c-4ed5-b642-6d76bd617d25",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 441
        }
      },
      "source": [
        "#@title Comparing the Reconstructed Images\n",
        "\n",
        "# Make a figure with a (2 x (datasets * models)) subplot grid\n",
        "num_combinations = len(datasets) * len(models)\n",
        "f, axarr = plt.subplots(2, num_combinations)\n",
        "f.set_figwidth(14)\n",
        "f.set_figheight(7)\n",
        "\n",
        "# Populate the figure with a true and reconst. image for each model and dataset\n",
        "column = 0\n",
        "rand_idx = np.random.randint(24)\n",
        "for dataset in datasets:\n",
        "  for model in models:\n",
        "    axarr[0, column].imshow(images[dataset][model].true_images[rand_idx])\n",
        "    axarr[0, column].set_title('True image from {}'.format(dataset))\n",
        "    axarr[1, column].imshow(images[dataset][model].recon_images[rand_idx])\n",
        "    axarr[1, column].set_title('Reconstruction by {}'.format(model))\n",
        "    column += 1"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAzgAAAGoCAYAAACKdUDoAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDMuMC4zLCBo\ndHRwOi8vbWF0cGxvdGxpYi5vcmcvnQurowAAIABJREFUeJzs3XmcZHdZ7/HvU0vvPUvPnlkyWSaE\nQDYYYsB4jQRCkEhQEIlXWQQCClwQXBC8wlWUeBUBBa9EiQMSlrCERI1ACGAIgSSTHbIvM8lMZl96\neu+uqt/9o85kenqeX/ep7urt1Of9es1rup86dc7vnDpPVf1q+baFEAQAAAAAWZCb7QEAAAAAQL0w\nwQEAAACQGUxwAAAAAGQGExwAAAAAmcEEBwAAAEBmMMEBAAAAkBlMcFIys38xsw/M9jgmYmavMbNt\nZtZrZqfP9nimYr4ccxxrvtx29AsmMl+O62yfy2b2C2b2aLL9i6ewnm1mdn7ksvPN7GeTHuQMmy/n\nTj3Nl32mX+aeep87Nl/+Do6Z9Y76tU3SkKRy8vvbQghXzfyo5h4z2yrp90II/znbY5kpZrZN0m+F\nEH4w22OZK+iXdOiXuY9zOZ3ZPpfN7L8lXR1C+PQU1zOvzs96qOc+0y/p0C/zV9p9LszMcKYuhNBx\n+Gcz2yLpLSGE78aWN7NCCKE0E2ObK8wsJ2mtJHfG3ojHpFHRLxOjX+YHzuWJzZFz+fjY9jFz6JeJ\n0S8NIoQw7/5J2iLpJWNqH5H0FUlfktQj6Y2SviDpw6OWeYmkLaN+XyPpGkl7JD0h6R3jbPOZdR1e\nj6Q/Sa77tKRfkXSxpEck7Zf0R6Ou+0JJP5F0UNIOSX8vqTjq8pdLelhSt6R/kPQjSW8cdflbJD0o\n6YCk/5K01hlfu6ReSUFSn6SHkvo2SX8o6T5JQ0ntOZL+OxnPfZJeMWY//0HSt5P13SRpRVI7KOkB\nSWdGjlEh2f7vSnosuR0+JGlDsv+HktunOOY4/tGo4/j6yDFfLun6ZAz7Jd2U1L8kqSJpIBnve2f7\n/Jxr/+gX+iUr/cK5PGfP5S1jzqt8MvYHktvkMVWfaB9e3j0/R437vcn4upPbtTlyO060P3+fHLce\nST+WdMIE51emelL0C/3SwP0y6w9YdW7aYVWbJyepVeM0bbLM3ZI+IKlJ0snJei9I2bQlSR+UVExu\n3N3JMh2SzpA0KGldsvwLJP1ccjKcqGqDvnPUjdgj6ZJkXe+VNKKkaSW9WtJDkp6VXP/Dkn44wcm2\nfszJf4eqd1Ctyb4+kZxgxWRfeiWdPGo/d0s6W1KLqs3whKTfVLUJL5d0wwTb/4akzuQ4DEu6QdJ6\nSYtVvfP5n2OO44eSsbxS1TucBc4x/xtJn0qWa5L0P8bs4/mzfV7O1X/0C/2SlX7hXJ6b57J3XiW3\nx4mSTNKLVX1CckbK8/MnklZKWpIcs7c4t2Oa/dkraWNy+VckfWGC8ytTPSn6hX5p4H7JWsjAzSGE\nfw8hVEIIAxMs+0JVD+pfhRCGQwiPSvqspNel3NagpMtDCCOSvixpmaSPhxB6Qwj3qtpoZ0hSCOH2\nEMKtIYRSCOFxSVdI+sVkPRdLujuEcG2yro+repId9nZJfxVCeChU3zL9iKRzzGx1ynFK0idDCNuS\nY/Lzqp4sfxNCGAnVt67/a8x+fz2EcFcIYVDSNyX1hhC+GEIoq3rSnz3B9v46hNCTHIcHJH0rhLAl\nhHBA1Vc7Rl9/UNJHkrFcp+rnhU9x1jki6ThV7wiHQwg31bD/8NEvPvpl/uFc9s30ufyM5PZ4PFR9\nT9KNkn4huXii8/MTIYSdIYR9kv5D0lnOJtLsz9dCCJuT43tVZD2erPck/eKjXzLUL1mb4DxVw7LH\nS1pnZgcP/1N1Zrsy5fX3JiexVJ1pS9KuUZcPqPoKhczsVDP7TzPbaWaHJP25pKXJcseNHncIIag6\nOx09zk+PGuNeVd+eW5N2R3X0cTlO0pPJdg7bKmn0ncDY/XD3axy1XH/0cZSk/sj6L0/GeaOZPWZm\nfzjBGDAx+sVHv8w/nMu+mT6Xn2FmF5vZrWa2Pxn/hTqy7xOdnztH/Rw7x9PsT5r1eLLek/SLj36Z\neD2eOdkvWZvghDG/96maInLY6IZ8StIjIYRFo/51hhB+ZRrG9RlJP1X1rcAFkv5M1bchpernTJ9p\nQDMzHX3CPSXpzWPG2RpCuLWG7Y8+Lk9LWpts57B1krbXsL4ZF0I4FEL4/RDCekmvkvTHZnb4lZ2x\ntzvSoV989Mv8w7nsm5Vz2cxaJX1N0kclrQghLJL0HSX7PsH5mda86s051pP0i49+mSPq0S9Zm+CM\ndbekV5jZYjNbJel/jbrsx5KGzex9ZtZiZnkzO93Mnj8N4+hU9ctffWb2bElvG3XZf0h6npn9ipkV\nJL1b1bdwD/snSR9MriczW2Rmr5nCWG5R9TOQ7zOzopm9WNIvq/p26pyVHJ+TkubsVjX2spJcvEvV\nz65iauiXY9Ev8xPn8rFm8lxuVvXjMHskla36dz4uOHzhBOdnWvOqN+d4T9Ivx6JfZlE9+iXrE5xN\nqn4ecKukb6n6+U9JUvIZzV+WdI6qX5jbq+qrBwumYRzvk/QGVb8g9xmNOqFCCLsk/Yakv5O0T9JJ\nku5S9XOJCiF8Nbnsq1Z9y/ZeSS+b7EBCCEOqfpntElX3+e8l/WYI4ZHJrnOGPEvS91T9UtyPVP2s\n7A+Ty/5K0v+x6lvT75mtAWbAJtEvR6Ff5q1N4lw+ykyeyyGEg5J+X9Xkrf2SXqPqE9TDxjs/025j\nvvXmXO7JTaJfjkK/zLop98u8+UOfjcLM8qq+lfiaWk9goNHQL8gKzmUgPfoFE8n6OzjzgpldlLyd\n2izpf6uaHnHbLA8LmJPoF2QF5zKQHv2CWjDBmRvOk/S4qp+/fJmkX03eTgRwLPoFWcG5PMPM7Hwz\n63X+HZztsWFC9MsMm8/9wkfUAAAAAGQG7+AAAAAAyIzCVK5sZhdJ+qSkvKR/CSFcPt7yTdYcWtQ+\nlU0C065HB/aGEJZNvGRt6BdkEf0y/ayp6NZHFjb5y0cCZAvdg249lMpuvVZWyLv10sIWf7vmllXs\nHvaXHxmZ1LjmikH1aTgMRfZ6amrpF3pl1PL0ypw11ceWSU9wkgSLT0t6qap/TfZ2M7suhHB/7Dot\natfP2QWxi4E54bvha1vrvU76BVlFv0y/wkr/j7E//cp1br3pkP/R8yXXP+TWy/v2T25gY+QXL3Hr\ne19xiluv+M85teJ6/5QqbX96UuOaK24NN07LemvtF3rlCHpl7prqY8tUPqJ2jqRHQwiPhxCGVc1N\nv2QqgwEyjH4B0qNfgPToF2CMqUxwVkt6atTv25LaUczsMjPbbGabR0TYBRoW/QKkR78A6U3YL/QK\nGs20hwyEEK4IIWwMIWwsqnm6NwfMa/QLkB79AqRDr6DRTGWCs13S2lG/r0lqAI5FvwDp0S9AevQL\nMMZUUtRul7TBzE5QtZFeJ+k36zIqIHvoFyC9zPbLU3/6ouhlr3r1zW59XfNdbv2slmvdemfOT1ba\n8qHFbv3C1j63nlNtgV83DLS69Q1Ff5w7y21u/e7fP96tbx/yx3/NtedFx7Tuw7dEL8uQTPYLvXIE\nvVK7SU9wQgglM3unpG+rGkt4ZQjhZ3UbGZAh9AuQHv0CpEe/AMea0t/BCSFcL+n6Oo0FyDT6BUiP\nfgHSo1+Ao017yAAAAAAAzBQmOAAAAAAygwkOAAAAgMyY0ndwAABoZP2/+nNu/bUf+ZZbf17rp6Lr\n2jKyzK3vKXW69e/2PsetLy8ecutntWx167vK/h9+rLhVaST49a582a3/oP9kt/7Y4HK3vrjoJ1Wd\n2vq0W//M6//RH5Ck+197zN+HlST924cudusdX701ui5MDb1yBL0y/XgHBwAAAEBmMMEBAAAAkBlM\ncAAAAABkBhMcAAAAAJnBBAcAAABAZpCiBgDAJL31o1936/2VZrf+nZ7To+tqtpJbb8v7yU0d+UG3\nvmVwqVu/v/84tz5QLrr145q73Xre/MyovSMdbj1mebHHrVeC/9rrk8P+fj02tCK6jXVN+9z67/z5\ntW796q+ujK4LU0OvHEGvTD/ewQEAAACQGUxwAAAAAGQGExwAAAAAmcEEBwAAAEBmMMEBAAAAkBmk\nqAEAMIGdv/8it35q8yfd+nXdz3PrXYW+6DZykcSloYqf3FS0sltf37LXrbfYsFtfEEmYOlRuceux\n1Kt1Tf52K5HXUp8eXuzWR0LerceSsGLHR5KeGFrm1l++4B63vvPdv+bWV37ylug2cDR65Qh6Zfbw\nDg4AAACAzGCCAwAAACAzmOAAAAAAyAwmOAAAAAAygwkOAAAAgMwgRQ0AgAmsveQJt95T8dOTYglQ\nZVl0G/V6xTGWlPS9g6e69ccPLnHrxbyfPPXqNXe79a9s3+jWn71ol1s/pW2nW+8utbn1A6V2t95s\nJbcuSUuLPW59MPjH6Phfe9ytD/kBYHDQK0fQK7NnShMcM9siqUdSWVIphODfYgDoF6AG9AuQHv0C\nHK0e7+D8UgjBD/QGMBb9AqRHvwDp0S9Agu/gAAAAAMiMqU5wgqTvmNkdZnaZt4CZXWZmm81s84iG\nprg5YF6jX4D06BcgvXH7hV5Bo5nqR9TOCyFsN7Plkm4wswdDCDeNXiCEcIWkKyRpgXWFKW4PmM/o\nFyA9+gVIb9x+oVfQaKY0wQkhbE/+321m10g6R9JN41+rsex+x4vc+l0f/Ee3/tnulW796686z62X\nH3p0cgPDjKNfJka/4LC51i+bTvqaW7++73i33pwbcet5VaLb6C77iUgjIe/WVxUPuPWPP/QSt97e\nPOzWf2HlY279kd7lbv3Td/+iW3/Zsx5w6995+Nlu/aEV/vp//bg73PreUqdbz1v8mLbl/H3eWVrk\n1j9zgn87v1H+fcpcMZf6hV45gl6ZPZP+iJqZtZtZ5+GfJV0o6af1GhiQJfQLkB79AqRHvwDHmso7\nOCskXWNmh9fzxRDCt+oyKiB76BcgPfoFSI9+AcaY9AQnhPC4pDPrOBYgs+gXID36BUiPfgGORUw0\nAAAAgMxgggMAAAAgM6YaE41E4cT1bv1f/uATbj2W9PH6Bdvd+l/97lK3fvJ75lYqVGHNare+95fW\nRa+z6Krb/Asq5XoMCXMQ/VJFv8w9lfPOcuududv95SOvE3bmBtx63uIJvbFkqI78oFtvj6Qefei0\n/3DrPzx0ilvfMbjQrW/o2O3WH3n4RLf+jvO+79bfvuwHbv0j217h1gdD0a2Xg3+sVxUPuvXx1hXT\nlW926+GF/ifA7Mf31LT+LKFXjqBXjpgrvcI7OAAAAAAygwkOAAAAgMxgggMAAAAgM5jgAAAAAMgM\nJjgAAAAAMoMJDgAAAIDMICa6Tvo/48cZntHkx9vGXNG93q2f+vc73XqpprXXzgr+KbLnGj/68GOn\nfc2tn9PsRzdK0q9d/QtuPQwRe5tV9EsV/TL37DjPj5/97wG/3mIjbj1nFbe+stAd3fYTQ8vc+snF\nXW79q3s2uvUHvnaqWx9Y7vddftDc+l/8zv9165e/zY8B/vFQq1u/au8L3fqdP9ng1i++2I+T3T28\nwK1vaPb7XZL2lPzrxKLnvz/Q4da3XdDu1tf+OLrpzKNXjqBXjpgrvcI7OAAAAAAygwkOAAAAgMxg\nggMAAAAgM5jgAAAAAMgMJjgAAAAAMoMUtRrl2v10iOd1PVWX9X9s80vd+obH76zL+mu1/aunuPU7\nn/dvbn1bacCtn/+nfxjdxuKhBo6hyTj6pYp+mT8GT/dvk/1lPzFoOJIwlI+8frjAhqLbLofaXnPc\n1rPIrR86fditt3T6225r8etv/613ufW9Z/oJUMVeP3lq6c1+clP5T/xcwzOat7v1bcNL3HosnWsy\nDpb9BLChU/3zopHRK0fQK0fMlV7hHRwAAAAAmcEEBwAAAEBmMMEBAAAAkBlMcAAAAABkBhMcAAAA\nAJkxYYqamV0p6WJJu0MIz01qXZK+Imm9pC2SXhtCODB9w5w79v36GW798pWfqmk9D4/4yR2n/IOf\ncOHnbdRPYc1qt/7B0/6rpvX8ybZXuvXFmxoj+Yl+ORr9Mj76Ze71yxlrt7n105ufdusHK81uvRx5\n/fC4gp+GJEldhT63vra4z62/Zf3Nbv2fdZ5b33OgM7ptz2Ov8fdtzak73Pr6Bf443/uhG9z6a255\nu1v/ycCJbn1Fsdutr83H07YGwyG3XomkcK3I97r109f6t/9M5kXNtX6hV46gV46YC70ipXsHZ5Ok\ni8bU3i/pxhDCBkk3Jr8DoF+AWmwS/QKktUn0C5DKhBOcEMJNkvaPKV8i6XPJz5+T9Ko6jwuYl+gX\nID36BUiPfgHSm+wf+lwRQjj8vttOSStiC5rZZZIuk6QW+X8UCMg4+gVIj34B0kvVL/QKGs2UQwZC\nCEHjfOQ9hHBFCGFjCGFjUf5nFIFGQb8A6dEvQHrj9Qu9gkYz2QnOLjNbJUnJ/7vrNyQgc+gXID36\nBUiPfgEck/2I2nWS3iDp8uT/a+s2ornALHpR98v95I5avfKHv+fWT779rrqsv1aPv/l4t/7qjutq\nWs+t953s1k/RbTWPKUPolymiXxrKrPZL7/uPc+sX/8673Pofv8hPznv7ou1u/e/2nx7ddlvOTzi6\n9EeXufU3neGn7Z27bItbf/HJ97v1X2w56NZf9P33uvXKs/yeP6HNT4b6yLZXuHU93eKWP/OjS9x6\n2yt3uvWWE/w0RUkarBTden8k0etNP7rQrW/4Vz/J0bQruu0ZMmv9Qq8cQa8cMVd6ZcJ3cMzsS5J+\nLOlZZrbNzN6saiO91MwekfSS5Heg4dEvQHr0C5Ae/QKkN+E7OCGESyMXXVDnsQDzHv0CpEe/AOnR\nL0B6Uw4ZAAAAAIC5ggkOAAAAgMxgggMAAAAgMyabopZpueZ4RvxPz/vXmtZ1TV+XWz/5t2cn/Smm\n4wV767Kepbfl67IezB/0y+TRL3OP3XKPWz/lFn/5a7Sspvp4rt7mJz39Y8f/cOufvePn3frqVQfc\n+re/eq5b71tXcuvL9/p/gmj4y/7fXv1u/3K3Xok801gcuevoX+nXP3nKN9z65c//Jf8KksoH/GMR\nc4pur2n5RkavHEGvzD28gwMAAAAgM5jgAAAAAMgMJjgAAAAAMoMJDgAAAIDMYIIDAAAAIDNIUZtm\n77/hdW59g26d4ZFUFVb6iR6ffs5VkWv4c+Bd5QG33vnk8GSGBUiiXzAH5CLJdpXytG/6rqF2t/7e\nU29063953avd+o4DfkJTU5O/3dygf952X9zr1pt/1OnWR9rNX89p/rH78AVfc+sL8oNufUPBH0+t\n6U+TMovnxZxFrzyDXhlljvQK7+AAAAAAyAwmOAAAAAAygwkOAAAAgMxgggMAAAAgM5jgAAAAAMgM\nUtTmufyihW594NxT3Hrx/dvc+tlNtc11X3f/6916+3fvqGk9wEyiXzChWUzFumrvC936K7rucetv\nefl33fpI8FOMHuxd6dabcyW3flbnU279xec86NZ/MnCCWz+pabdbv7X/JLe+q1Rx6985+Fy3LvlJ\nUnXVyGlpMfTKM+iVUeZIr/AODgAAAIDMYIIDAAAAIDOY4AAAAADIDCY4AAAAADKDCQ4AAACAzJgw\nRc3MrpR0saTdIYTnJrUPS3qrpD3JYh8IIVw/XYOcz975ize49X+95ly33nKdn/J08KUDbv0vnn+t\nW391x/dSjG7ydt+1wq2foMendbtzHf0yNfRLY6FfjtZTanbrg5WiWy8H/zXKk5t3ufWNbU+49fsG\n17j1e3v9+o8O+olOy5t73Hp3uc2t91ea/PUUD7n1x3uWuHVpe6SeLfTLEfRKsh56JSrNOzibJF3k\n1D8eQjgr+Zf5ZgJS2iT6BUhrk+gXIK1Nol+AVCac4IQQbpK0fwbGAsx79AuQHv0CpEe/AOlN5Ts4\n7zSze83sSjNbHFvIzC4zs81mtnlEQ1PYHDCv0S9AevQLkN6E/UKvoNFMdoLz/ySdJOksSTskfSy2\nYAjhihDCxhDCxqL8z0wCGUe/AOnRL0B6qfqFXkGjmdQEJ4SwK4RQDiFUJP2zpHPqOywgO+gXID36\nBUiPfgF8E6aoecxsVQhhR/Lrr0r6af2GNPtCuRK97F8PrXXrb1rwlFt/1+JH/Po5fp27puyhX45F\nvyAm6/0ynvv3+Gl7v73CTwWM2VNa4NYfHfLXvzDvr//cBY+59XLktdHucqtbjyVb5RTcemfOH8/W\nfV1ufV0DJUON1aj9Qq9U0StxaWKivyTpfElLzWybpA9JOt/MzpIUJG2R9LZpHCMwb9AvQHr0C5Ae\n/QKkN+EEJ4RwqVP+7DSMBZj36BcgPfoFSI9+AdKbSooaAAAAAMwpTHAAAAAAZAYTHAAAAACZMakU\ntawLI8PRy775cye79a98Y6Nb/5cNX3Trawp+gkbMHZG/y/XWe37brTf910K3/oUP+H9S4uRibbn4\nK28t17Q8sot+mRj9gjQWtQ669b6Kf77lzU8wbM6NuPUOmVsfCXm33lNpcev95drO/2Ku5G+37G+3\nJTL+45fsd+t+vhSyjF6polfieAcHAAAAQGYwwQEAAACQGUxwAAAAAGQGExwAAAAAmcEEBwAAAEBm\nkKJWo/KhQ2698BK//vZzftet7z27o6btLv+3e9z6cf33++M5fq1bb/mgnyRSq47vPejWyYrCaPRL\nFf2CNM7oetqt90eSoWIqwU+AqgT/Nc1cJGGqxfyEpkrOX085kjzVYpFkKPOfgsT29+TOvW79EbeK\nLKNXquiVON7BAQAAAJAZTHAAAAAAZAYTHAAAAACZwQQHAAAAQGYwwQEAAACQGaSoTbfb7nPLS2+r\nbTW1Zjltv8RPhVpTaK1pPRf89DVuvbX3yRpHBKRAv6CBrW3Z79aHg/9QXTQ/h69c42uXscSoQRXd\n+kjI+9uNrCefC/52I0lSgxV/uyubu936I2px68gueiXZLr0SxTs4AAAAADKDCQ4AAACAzGCCAwAA\nACAzmOAAAAAAyAwmOAAAAAAygxQ1jOvADavcemvliRkeCTD30S8NKOenJKnipzaN576e1W79hYse\n8zdtfl5gLOkpliQVE1v/UCQxqpgr1bZ++YlRI5EkrAd7V0bWdLCm7VY3Xr/bDSnRK0fWT69Muwnf\nwTGztWb2fTO738x+ZmbvTupdZnaDmT2S/L94+ocLzG30C5Ae/QKkQ68AtUnzEbWSpPeFEE6TdK6k\nd5jZaZLeL+nGEMIGSTcmvwONjn4B0qNfgHToFaAGE05wQgg7Qgh3Jj/3SHpA0mpJl0j6XLLY5yS9\naroGCcwX9AuQHv0CpEOvALWp6Ts4ZrZe0tmSbpW0IoSwI7lop6QVketcJukySWpR22THCcw79AuQ\nHv0CpEOvABNLnaJmZh2Svi7pPSGEQ6MvCyEEyf8GVAjhihDCxhDCxqKapzRYYL6gX4D06BcgHXoF\nSCfVOzhmVlS1oa4KIXwjKe8ys1UhhB1mtkrS7ukaJGbPytsHZnsI8w790rjol9rN936xnLn14Icq\njaurqc+tjwQ/xSiWAKVIolOtiqrPesryj1Fsv2LLL23udet7JjGmet5uM4VeOYJeGX/5Ru8VKV2K\nmkn6rKQHQgh/N+qi6yS9Ifn5DZKurf/wgPmFfgHSo1+AdOgVoDZp3sH5eUm/Lek+M7s7qX1A0uWS\nrjazN0vaKum10zNEYF6hX4D06BcgHXoFqMGEE5wQws1S5D0w6YL6DgeY3+gXID36BUiHXgFqkzpk\nAAAAAADmOiY4AAAAADKDCQ4AAACAzKjpD30CAIBR8n58q0olf/FFC6OrOrF1q1uPRdzm6hRxG4/Q\nrW09ef9PsEQjbotWrmk865r3u/WHF62Njql8sNu/wHh9d8bRK8+gV6bf/Bw1AAAAADiY4AAAAADI\nDCY4AAAAADKDCQ4AAACAzGCCAwAAACAzSFEDAGCSzPz4JD8jSbKuxdF1deV/5tZ3jPjXydUa3RQR\nS3TKWazuJ1KVaxxPvsb1dOX73LotXhTfSCQZyvL+67thJL4qTA29cgS9Mv14BwcAAABAZjDBAQAA\nAJAZTHAAAAAAZAYTHAAAAACZwQQHAAAAQGaQopZRx31rl1vf8QcDbn1VvtWtr/vrR9z60y9uc+uV\n/v4UowPmFvoFk5ar7XXC8uL26GX5SBLTYKXo1tvyQ/56IklPRSu79eacH4dUCX5C00jFf+pQiSQ6\nteWG3Xpsv2L1E5v2uPXy0gVuXZL0RKRe4+2GOqBXjixPr0y7+TlqAAAAAHAwwQEAAACQGUxwAAAA\nAGQGExwAAAAAmcEEBwAAAEBmTJiiZmZrJX1e0gpJQdIVIYRPmtmHJb1V0uGohg+EEK6froGiNuWH\nH3Prl/zfP3Lr73zHN9z6nbvWuPWVlS2TGlfW0S/zE/0yOxqxX4aW+Yl6ktRifkLTUCSJqTM/WNO2\ny5HkpqFIElPOKm69mCu59VhiVK3r6R1pdusxg8taopfVtqa5i145Gr1SRa/EpYmJLkl6XwjhTjPr\nlHSHmd2QXPbxEMLfTt/wgHmHfgHSo1+AdOgVoAYTTnBCCDsk7Uh+7jGzByStnu6BAfMR/QKkR78A\n6dArQG1q+g6Oma2XdLakW5PSO83sXjO70swWR65zmZltNrPNI/L/0BKQRfQLkB79AqRDrwATSz3B\nMbMOSV+X9J4QwiFJ/0/SSZLOUvVVhY951wshXBFC2BhC2FjM3Cf8AB/9AqRHvwDp0CtAOqkmOGZW\nVLWhrgohfEOSQgi7QgjlEEJF0j9LOmf6hgnMH/QLkB79AqRDrwDppUlRM0mflfRACOHvRtVXJZ8J\nlaRflfTT6Rki6mn5p25x61d/aqW/vB50637OB+iXbKFfplcm+sX8tKWYnjXxh932nP/RoeZIglLR\nym59JORrqvdHEp0qkSSpnEJNy/eX/XcMYslWxZy/X7Fkq97V8WMafa+ixtttttErR6NXquiVuDQp\naj8v6bcl3Wdmdye1D0i61MzOUjWucIukt03LCIH5hX4B0qNfgHToFaAGaVLUbpbcqWAmctaBeqJf\ngPToFyAdegWoTU0pagAAAAABIqGnAAAgAElEQVQwlzHBAQAAAJAZTHAAAAAAZEaakAEAAOAp+ylG\nMU09fqqSJG3uP8Gt91ea3Hq+7OfzxZKb8uYvH0to6ir0uvWecotbL0deMx2qFN16bL8GI8sfLLe7\n9WJv/JhGhUlcB1NDrzyDXpl+vIMDAAAAIDOY4AAAAADIDCY4AAAAADKDCQ4AAACAzGCCAwAAACAz\nLMxgOoKZ7ZG0Nfl1qaS9M7bxuaHR9nm+7u/xIYRlsz2IBu+XRttfaf7uM/0y+xptf6X5uc/0ytzA\nPs8PU+qXGZ3gHLVhs80hhI2zsvFZ0mj73Gj7O50a7Vg22v5KjbnP06XRjmWj7a/UmPs8HRrxOLLP\njYGPqAEAAADIDCY4AAAAADJjNic4V8zitmdLo+1zo+3vdGq0Y9lo+ys15j5Pl0Y7lo22v1Jj7vN0\naMTjyD43gFn7Dg4AAAAA1BsfUQMAAACQGUxwAAAAAGTGjE9wzOwiM3vIzB41s/fP9PZngpldaWa7\nzeyno2pdZnaDmT2S/L94NsdYb2a21sy+b2b3m9nPzOzdST3T+z3d6Jdsnjf0y/SgX7J33tAr04d+\nyd65Q78cMaMTHDPLS/q0pJdLOk3SpWZ22kyOYYZsknTRmNr7Jd0YQtgg6cbk9ywpSXpfCOE0SedK\nekdy22Z9v6cN/ZLp84Z+qTP6JbPnDb0yDeiXzJ479Etipt/BOUfSoyGEx0MIw5K+LOmSGR7DtAsh\n3CRp/5jyJZI+l/z8OUmvmtFBTbMQwo4Qwp3Jzz2SHpC0Whnf72lGv1Rl7ryhX6YF/VKVqfOGXpk2\n9EtVps4d+uWImZ7grJb01KjftyW1RrAihLAj+XmnpBWzOZjpZGbrJZ0t6VY10H5PA/qlKtPnDf1S\nN/RLVWbPG3qlruiXqsyeO43eL4QMzIJQzebOZD63mXVI+rqk94QQDo2+LMv7jemT5fOGfkG9ZfW8\noVcwHbJ67tAvMz/B2S5p7ajf1yS1RrDLzFZJUvL/7lkeT92ZWVHVhroqhPCNpJz5/Z5G9Iuye97Q\nL3VHvyib5w29Mi3oF2Xz3KFfqmZ6gnO7pA1mdoKZNUl6naTrZngMs+U6SW9Ifn6DpGtncSx1Z2Ym\n6bOSHggh/N2oizK939OMfqnK3HlDv0wL+qUqU+cNvTJt6JeqTJ079MsRVn2nagY3aPbLkj4hKS/p\nyhDCX87oAGaAmX1J0vmSlkraJelDkr4p6WpJ6yRtlfTaEMLYL77NW2Z2nqQfSrpPUiUpf0DVz35m\ndr+nG/2SzfOGfpke9Ev2zht6ZfrQL9k7d+iXI2Z8ggMAAAAA04WQAQAAAACZwQQHAAAAQGYwwQEA\nAACQGUxwAAAAAGQGExwAAAAAmcEEBwAAAEBmMMHJIDP7mZmdPw3r/YGZvaXe602x3Tea2c0zvV1k\nV9Z6BJgpWesdHl9Qb1nrkflqXk5wzGyLmQ2YWa+Z7TSzTWbWMdvj8pjZh83sC9O4/k1m9pHRtRDC\nc0IIP5iubWLuo0eOWj89gtTonaPWT+/gGPTIUeunR+aoeTnBSfxKCKFD0lmSzpb0J7M8nkmxqvl8\nO2DuokeAyaF3gPHRI5jT5v2NGkLYKenbqjaZJMnMms3sb83sSTPbZWb/ZGatoy6/xMzuNrNDZvaY\nmV2U1I8zs+vMbL+ZPWpmbx11nQ+b2dVm9nkz60negtw46vI/NrPtyWUPmdkFyXo/IOk3klc67kmW\n/YGZ/aWZ/UhSv6QTk1dEXjJme18Y9ft5ZnaLmR00s6eSt9Uvk/Q/Jf1Rsv5/T5Z9Zl3JsfiEmT2d\n/PuEmTUnl51vZtvM7H1mttvMdpjZmyY45CeZ2W3JsbvWzLqSdf2nmb1r9IJmdq+Z/erYFZjZejML\nZvamZF8OmNnbzewFyXUOmtmnYgNIrvu/zOxxM9trZn/DHVQcPTL/eiS2P0l9YXKM95jZVjP708Pn\nf7LPNye37QEze8LMXp5c9htmtnnMNn7fzK6bYH8aFr0z/3rHeHyZUfTI/OuR2P4k9ew8voQQ5t0/\nSVskvST5eY2k+yR9ctTlH5d0naQuSZ2S/l3SR5PLzpHULemlqk7wVks6NbnsJkn/KKlF1WbdI+nF\nyWUfljQo6Zcl5SV9VNJPksueJekpScclv6+XdNKo631hzPh/IOlJSc+RVJBUHL1PY68n6XhJPZIu\nTZZdIums5LJNkj4yzvH5c0k/kbRc0jJJt0j6i+Sy8yWVkmWKyb71S1ocOe4/kLRd0nMltUv6+qgx\nvlbSraOWPVPSPklNznrWSwqS/ik51hcmx/abyThXS9ot6ReT5d8o6eZR1w+Svp/cvuskPSzpLbN9\nXs6lf6JH5nuPjLc/n5d0bXK7rU/O/zeP6pURSW9NboPflfS0JJPUlqxzw6jt3C7pdbN9vs6lf6J3\n5nvvrBePL/QIPdLwjy+z3ihTaK7e5GAGSTdKWpRcZpL6Dp/cSe2Fkp5Ifv6MpI8761wrqSypc1Tt\no5I2jTrZvzvqstMkDSQ/n6zqHeZLJBXHrPeZJhlzkv55rCGc5voTSddEjsVEzfWYpF8eddnLJG1J\nfj5f0oCkwqjLd0s6N7KtH0i6fMwxGE5O9BZJBw6f3JL+VtI/RtazPrndVo+q7ZP0G6N+/7qk9yQ/\nv1HHPgBdNOr335N042yfl3PpHz0y73vE3Z9kPcOSThtVe5ukHyQ/v1HSo6Mua0tu/5XJ71+Q9GfJ\nzxuS86Ntts/XufSP3pn3vbNePL7QI865PuZca+QeaYjHl/n8tuurQgidqp4gp0pamtSXqXrQ70je\nejso6VtJXao20WPO+o6TtD+E0DOqtlXVVxcO2znq535JLWZWCCE8Kuk9qjbEbjP7spkdN8H4n5rg\n8tFiY07jOFX347CtSe2wfSGE0qjf+yWN92XB0ePeqursf2kIYVDSVyT9VvJ25qWS/m2Cse0a9fOA\n83st45joeDcieiSdudgjsf1Zmqxv7Hjd2yCE0J/8eHi8X0y2K0m/Kembo5bBEfROOnOxdw7j8WV6\n0SPpzMUeaYjHl/k8wZEkhRD+W9UZ9N8mpb2q3nk9J4SwKPm3MFS/DCdVT46TnFU9LanLzDpH1dap\n+nZgmnF8MYRwnqpv/QVJf334othVxvzep+qdwmErR/0cG/N46z/s6WRMh61LapO1dsy6RlQ95pL0\nOVU/j3qBpP4Qwo+nsJ1axzGVfco0emRe9khsf/Ym6xs73lS3gaQbJC0zs7NUfSD6YsrrNSR6Z172\nTj3w+JISPTIve6QhHl/m/QQn8QlJLzWzM0MIFUn/LOnjZrZcksxstZm9LFn2s5LelHwBLZdcdmoI\n4SlVPxv5UTNrMbMzJL1Z1bfcxmVmzzKzFydfHBtUtbkrycW7JK23ib+keLek15lZMfni3GtGXXaV\npJeY2WvNrGBmS5IT6PD6TxxnvV+S9KdmtszMlkr6szT7NI7fMrPTzKxN1c+Nfi2EUJakpJkqkj6m\niV9dm6o/NLPFZrZW0rtVfeUCcfRI3FzsEXd/kvVcLekvzazTzI6X9N604w0hjEj6qqS/UfXz8TdM\nch8bCb0TNxd7px54fKkNPRI3F3ukIR5fMjHBCSHsUfWLUX+WlP5Y0qOSfmJmhyR9V9UvoSmEcJuk\nN6n6JbhuSf+tI7PVS1X9/O7Tkq6R9KEQwndTDKFZ0uWqzn53qvplssORiV9N/t9nZneOs47/reqM\n+oCk/6NRM98QwpOqfvnsfZL2q9qIZyYXf1bSaVZ9K/ibzno/ImmzpHtV/SLgnUltsv5N1Vdrdqr6\nmc//Nebyz0s6XVNr4DSulXSHqsfiP1U9DoigR+ZXj0ywP+9S9dXGxyXdrOpxuLKG8X1R1c+qf3XM\nRyPgoHfmV+/UCY8vNaBH5lePNMrji4Uw0btrQHpm9npJlyVvFU/XNoKqX6R7dLq2AUyXmegRIIt4\nfAHGx+PLEZl4BwdzQ/KW6e9JumK2xwLMRfQIMDn0DjA+euRoTHBQF8nna/eo+nlUvrgMjEGPAJND\n7wDjo0eOxUfUAAAAAGQG7+AAAAAAyIzCVK5sZhdJ+qSqf/30X0IIl4+3fLG5PTS3d01lk+OLvRll\n82gbta4/9g6cRa5Q8xt2Na5/MmL7kKvTPtR4TPsObNsbQljmXzp5tfZLU6EttBYX1nsYQF0dGtw5\nN/ol3xpa8wucFUWuEL1fqPG+LbZ4JbKBfPx1xUox79eb/I1Uiv56QmRMuZHIkIZi9/OR9ef9C0ot\n/vKxl1ItkrEUG09uuOLWLXaspfjtUOt5EbughsejgdIhDVcG6vmM5Bm19Eu0V+qlUvbruciJUPCf\nilaaxumVyDkY/BZSiKwqVre0fz3nmSv45UrkWXZs/bHV52KHNNLTuZK/JivHT1gb8fur5ueftfZc\nxKHh3VN6bJn0BMfM8pI+LemlkrZJut3Mrgsh3B+7TnN7l854ybuPvaDGJ6yxEzJ2AsSWj65/nBsh\nto1KIfZIUNsEITfiL18p+svnh/wTMjaeaNNGxJqh1OIf1Nj6x3sAijVidBuxHoxsI7Z87Hb+8df+\ncKt/yeRNpl9aiwv1wpPfXO+hAHX17Z/+5dzol/wCvWjlpcdeEH2xJ3KfVIg8Q4qtJ1Yf9p95hM42\nty5Jg6v9J529a5rcet+qyJO8yGNe225/nxc+PuyvJ/IkcmiR/9ThwCn+hsut/nab9/vrX/SYP/Np\n2+b/UfRc35BblyQb8C8LkdvZypEHjFLsCUbsmeqx9Vt2fdlfdopq7Zdor8TU+FWG0NPrj7PdP/cr\nK/wXvvvXdLh1SRpa6J9rwwsik++2WN1ff3Ti4LeKypHJ/VCXfz7lSpHejdydNHX7F7Tt8G+btr3+\n+dp0ILIDkoq7eyIX+P0e663YfV/0vjLyYva3tnx8So8tU/mI2jmSHg0hPB5CGJb0ZUmXTGUwQIbR\nL0B69AuQHv0CjDGVCc5qSU+N+n1bUjuKmV1mZpvNbPPIkD+rBxpAzf0yXPZfrQQaQO39UhmYscEB\nc8yE/UKvoNFMe8hACOGKEMLGEMLGYnP87UYAR/dLUz7+URYAY/ol1zrbwwHmLHoFjWYqE5ztktaO\n+n1NUgNwLPoFSI9+AdKjX4AxppKidrukDWZ2gqqN9DpJvznRlbwvUEXTJCJfcIx98T325axav2Q+\nbtJDNFmjti+4W41f2qs1fCC2fOyYxkIJQmz94yRx1Koc2UbsWMdCCWLrydUeHTcdJtUvQIOqvV9M\n/pdYS5GorqZIBFnki7CxL6VXFrW79b1n+J9Y2PeCyJfVJV34vPvc+u8u+4Fbv3fomE/tSZJGgv/Q\nfk/fWrd+516/3tHkf4n4pPaDbv2szifdemfO/0jUykK3W99Z8tMjP7ftRf7yN65x65K0/C7/C8+t\nT/rbrvXRwvoiH/fyQm+mJT9NUq39EuuV6PKxiLDIE5y1q9zy0DK/V/ac7X9D/9CzI19Wl3T2qY+5\n9fO6/HpHftCtryz453Il8vr/Q4P+vuUiT/Ze1PaIW49piUQLbhlZ6tbv6Fvv1r+34xS3vvOueCjZ\nsjub3XpTtz+m5p2R86LTf4fQdu7z662x+MWpmfQEJ4RQMrN3Svq2qrGEV4YQfla3kQEZQr8A6dEv\nQHr0C3CsKf0dnBDC9ZKur9NYgEyjX4D06BcgPfoFONq0hwwAAAAAwExhggMAAAAgM5jgAAAAAMiM\nKX0HZzJiqWL+srXlmUTT2GpMLgm5+BUqhUg6WeQ6wQ/eiY4pFwnYiS4fSRQbWuBvOJZ+Fktda+rz\nb7BSqz83jo2n1BKfS8dS8aK3Z2RVtSa71XpeAJjDguKpTp7Y40vkjqey0P+7VFtf3unWz3vFPW79\nPSu+Gx1SV+QBoCdyZ7Ugkgx1/4Cfrrag4C//nK4dbr0YecBeVPT/CPGBkp+S1S3/2LXk/JSs1y/Y\n69ZffMoX3frm41e6dUn6g9t+3a2v+tpit9623d+33GAk0SuSxmcl77acIw86QX7yazme8Oeupt1P\ny+o70U/B2/FC/3nJOeff79ZXtfhJd5L0go7H3fq6wn63vrvs9+lDQ34qWj6Sp9dfaXLrq4p+Gttj\nw8sj6/ETy7oKvW49ljh4SutOt37OBv/43LN6nVuXpKtOfIFbb77L7+ul9/pPxlof92+DmNDqH4up\n4h0cAAAAAJnBBAcAAABAZjDBAQAAAJAZTHAAAAAAZAYTHAAAAACZMbMpaiESUFNjWlY0iCQWoBNZ\nTyzRLZYENh7zEknGUSn4OxFLjqs0+csPtcXSzGrbbixFbbijtjS2mPxQfPlak+aiKXeR2yDkI8e6\nxn0AMA/F7hdqTO458Gw/hen0Cx9y6+9acaNbH4k+sEm7yv5l9w75qWi39Zzk1ocq/kN7e2HIrQ9H\nlh92q1JzJP2st+wf06L56Vy7Rha59Z8M+ilMC3P+8Tm9yU+SkqR3nvV9t/7p7S936+u+7adkNQ1H\nEsYs8njkJvrNocccpy9CJIXQIsddef/Bu2eNX197zna3/gervu3W+4KfUCdJT40sceu39G9w6205\n/9zvLvkJf7FzdqQS2edKi1u/q9dPLWuOPElb3tTj1g8W/SSzmDbz9/f1i26LXqfrjD63/vE9F/nL\nPxB5btU34G+gxb9/KHf4x26qeAcHAAAAQGYwwQEAAACQGUxwAAAAAGQGExwAAAAAmcEEBwAAAEBm\nzGyKmkWSsSLTrFiyV7k5lhzmJ4CESBxXuVhbkpkk5WJBKrFErki5MOCPtdTq71slkgRW8QNfVI6E\nUkSCQVTJ+9stDkRS3WLjidRjSWmSVBj0txE7ppHAn2gq3ni3J+aHWBJezcmJkXMqRJKQbCQSRxjp\nl0qrn/ozstBvyKb9ftqMjUQaFeMIfnJVW6u/eOQ2Ly/0lz/wHH81f7byR269JXJne/uQn6okSWuL\n+9z6ony/W+8r+w8AuRrTuhYW/fNwoOyfzwORB57OwqBb7yv56Un9kfUcrPjJVk+Wutz6uoKfuiZJ\nL21/wK1/48yz/THdvcKt5/v9+4J8JKXPSs7tH0tcmw3OWCw2vib/PBhcu9CtH3y2f8f8plX3uPWT\ni/4x/ErPcf54JO0v+6lisfSzJ4f81LWDkRS11pyfIbi6+aBbX1o45NYrked0sXO/u+Tf/0RT3SJP\nriqR573tkTQ5KZ40t/xk/36p/+5l/nqW+emIuZ3+emzAT2WcKt7BAQAAAJAZTHAAAAAAZAYTHAAA\nAACZwQQHAAAAQGYwwQEAAACQGVNKUTOzLZJ6JJUllUIIG8e/glRxkstypVhaVm0pZ6VIulqtKVq5\nSHDSeAq9fsKFRTYdS4IrtUSSfSKpaLFjlB+KpEVFprQhciYMLoqt319ekRCW2HgkqRJJWIslZ8XS\n0nLlSDJdi7/T441pOtTcLw2o0uSfiJUWvz6ywE/3KbX7J1XxkN/cTd1+Ys7IUj9hZ6TTH0/Pan+7\nQ4v9c3npff7y7U90u/VGUrd+8ZLVJIWifxv2r/Zv88LJPW799Ka9bv3uoeVuvRxJN5KkPaUFbv3e\ngbVuPZZyNlz2z6tc5AGpr+QnOg2W/WO0uMlPXZP8B6rYPu8Y9lO4jm/yb4OWnJ+2tK/iJ2pJ0sJI\nMtSvrL7XrW86/mVuvXWvf6xzg5EnDGXnOUEkca0eauoXk5/oVvT3MdYrg0v95dedutOtv37hg269\nI+cnh8WS0qR4qlisXo68nj8UOcd7Iz0Rs2vE793YORsb51AkJjbWK90j/rF7TH7C2YZm/7aRpNXF\nA279vSff4NY/8JxL3Xrrvg633tHv96J5vVIH9YiJ/qUQgn8PD2As+gVIj34B0qNfgAQfUQMAAACQ\nGVOd4ARJ3zGzO8zsMm8BM7vMzDab2eaRwb4pbg6Y12rql+Gy/8f9gAZRW79UYh+bAhrCuP1y9GML\nvYLsm+pH1M4LIWw3s+WSbjCzB0MIN41eIIRwhaQrJKljyVr+pDwaWU39srB1Ff2CRlZbvzQtp1/Q\nyMbtl6N6pXkFvYLMm9I7OCGE7cn/uyVdI+mcegwKyCL6BUiPfgHSo1+Ao036HRwza5eUCyH0JD9f\nKOnPx71SJagweGyiTaXGtKxIyEQ0ISy2nsKQf0FuOP7iRizZq9wSSU9a6A+q+yS/PrzYH1OlOZLS\nVvLH03TAH0/rbrccPaaxVLRSJNykdbd/7MrN8eSg2DYKg5F0vdht4CT0SeOl9MWHVG+T6pcsKPnn\n7fBxfuLMjnP9FKbB5bFe9W/zjq2R1LI9flpa7pD/kQ1riyRtLfX769BJblnlJf5294/4ST2Ffj+F\nRpJatuxz66G1OXqd+WZy/WJSzrlfjSVXRR4vhjv9C567codbX5Gv7biPxCIrJQ1GxtpsflLXoWG/\nX7qa/Y+3xlLXWvN+0lNvyd+32PIxhZx/P7C44I/zYNlPUTsu56c85RV/zG6JPAFYW9zv1vuP9491\n5W7/PiWW8Jjzkse85LI6qNfjS2j2zw/l/fu73tV+r7x0yZNuvcP886m3MujW+8vx3toz3OnWFxT8\n+/K9Q/596p5IvSkSp7tH/vL9NaauLWmu7SsblcgT3Pa8/9gSS2N7amRJdBvLCofceuw+q9zl3w80\nH/R7zg71+hturu3YpTWVp3grJF1j1YYtSPpiCOFbdRkVkD30C5Ae/QKkR78AY0x6ghNCeFzSmXUc\nC5BZ9AuQHv0CpEe/AMciJhoAAABAZjDBAQAAAJAZTHAAAAAAZMYM5khJMlOlcGyKSCzlLPjBHYoE\nsqgSCQCJiaVujafU7M8Je9f49Z5n+SkTq9f5cWYbl/npI/fsX+3Wt2xb6tbzOyIHL7LLTT1+Ck3v\nmsgVIlPjEEmJiQR9VLe90992bsSvW+T2t4q/fMhF0m+apifRBkeUF/tpSPuf5afjHH/hFrf+oiWP\nu/XP3+8noeYf9WP+ivv85Brr9dOcKoWFbn1ocSTJb+mQW29b4KcE9UbSFDuejt+ZtWyJXtTYzCQv\nuSqSTBaa/Ie/wSX+bXvWgm1ufWvJv3MbDP5tuLJ40K1L0iNDK916f8VPGcqZv2/DFf/+fzCSotZW\n8PdhWbOfenQokq6Wj4wnF0k5i41/abHHrbfn/P4aP0XNr68sdLv14iK/V0vtrW69yV+Nfy5OU4ra\npDiPixZJvQyRlKvBZf5xP7F1j1vfF/ljvJHQreh5IEm7hv0kzlivxFLF9g/4j1Hl4N9WS1r9x4oV\nLf5Yt/UtcuuLm/xjUYqkpcXGU4w8IT6hda9bHx4nxbEof10bmna69eec8LRb37NuvVtfstc/Fnao\ntkS5tHgHBwAAAEBmMMEBAAAAkBlMcAAAAABkBhMcAAAAAJnBBAcAAABAZjDBAQAAAJAZMxsTHRFJ\nxVMs+bEciffNlfzlSy3+BnJlfwOl1vi8r3+Fv+2Bs/zIv9eedpdbf/mCe936nQPr3fq3Djzbrbds\n8eM6Fz7u5y6OtPvjj9UHV/kH1cqRmNwm/9i1b4vHY5YiOZ4F/5AqV4pkSkaMtDOPr4tI3K5GIo0n\naXhhp1s/8Hz/Oh9cfbNb/9enf96td37fj4Nefst+t27dfuytIjGoxYN+ZGznNj9ud/hcP273khPv\nc+v/nnuuPx75EaiSFJprzMNvJE4Mb/DieiVVmvwo5RE/NVbFyANMfyR+diQSx7onErEsSYvyfgTt\nPT1r3XosZrl7yI80bi74+3BouMWtd7b5sczbe/2411Vth9x6LPo2ptjmx9XuK3e49Vh8tCQdrPjb\nbjH/TzisXerHeB9a4N/XLPiZf5vNabmcQptzm0f+1EJpgX9+xMK5nxxa4ta/lTverS8v+BHLnbnI\nkwBJnQX/vnn3kP+Yc2iktn3oaPLvy0uR82l7v/8nBWLn/o4B/z5+QdHfr0WRWOlC5O9mdJf9+4AT\nm/0/USJJC3L+tiuRfVjW4j+ePtUVeZ7Z5d+5NhETDQAAAADjY4IDAAAAIDOY4AAAAADIDCY4AAAA\nADKDCQ4AAACAzJj5FDUvsiISsBXy/gUVP/xGuVgcRmT9wx3+BaVIopgk9Z7mp7W8/JQH/HokLW17\nabFbv+qJjW69eJufDLLsbn88zTv9VIrBNf56drzQT2aydj91p2OBn+jR+5SfDFJpjtxokkIkhMYi\niS4hFzkvipF0vZHYeqJDgqfkp7VYv5+8Ikm9q/3z6pLn3e7WF+X98/bBO/z0nVO+H0mE2eOnqGmB\nn8IUWv1kKxvx9zl27izu9E/mczsedetX9z/frS/dF0+mQ4TJTVGLL+8vGwk/09aBpW79+soZbn0o\nsqJcNLdJGgn+/WTsOh0F//5/uFzbeppy/nneE0l8K0eSpGIJUANl/36gOe+f57EEqP6Kn3bYlY+k\nI0raOeInvrXk/BS1niF/nyNheVLOPxZuYmO5tgTQaVMuyw4em1wW2v3jXjjo368Ve/xkuc371rn1\nO81PA7xwxf3++iMJYZL07Nan3Xpf6QS3vmW4y63HzuWY5a1+4ttg5BxvyfvnWXOk54YiT3Db836v\n7x/2b4No8lnBTzqUpHLkyXJfxe+JpkiyZNkPrFO+30+mmy48xQMAAACQGUxwAAAAAGQGExwAAAAA\nmcEEBwAAAEBmTDjBMbMrzWy3mf10VK3LzG4ws0eS//1vzAMNhn4B0qNfgPToFyC9NClqmyR9StLn\nR9XeL+nGEMLlZvb+5Pc/nnBNQbJaQkRCJP2q2Z+XlSLJDZFgGg0v8hMjhhfGU266lvoJGq/u8lOh\n+oKf+nLN7rPdeu89S9z6Cd/vduu5x7a5dWv101BKpyx068PH+wkdFz3bTzdpzflpGN/Y5adClZvi\nKWq5kn+8S63+7RxiQUm1JChJ0fNrijapXv0yx1gkRa2yxE/Ok6T9p/vH+HWLb3Xrn9jxUre+wl9c\n5Ycfc+uFNavdemjyk25icj1+WmC56D+HsMjJ+Z2Dp7v1/Bb/Tqtll9/vkmQDfq/GEpDmuE2q2+NL\nkEaOTfWxsn+b5Pv8+y0e8bQAACAASURBVLCmQ21u/ZYd69364jb/HBks+Q+vbUU/VUmSzli03a2f\n1LbHrd+87yR/GwV/37qH/XNkJJK61tXip2fF0s/2D/vHrj1fW3pSPpL29nD/SrdeGCdt64Fu/zqr\n2vwe27PdT11bszfy5GUosm8l7xhN+TFnk6bx8cUGI/vi9JUkNR/0F3/ioVU1bffK/f4xb2vx7+sk\n6WVrHnTr7ZFkwVhSYOw5Wj7yZLW/5C8fq8dS1CqRx4piZLt9ZT/JrKvJTx7dPuAf0489+TK3LkkL\nmvwUxAODfl8//Mhxbn3NI/6xzg3E7/umw4Tv4IQQbpI0NnP1EkmfS37+nKRX1XlcwLxEvwDp0S9A\nevQLkN5kv4OzIoSwI/l5p6QVdRoPkEX0C5Ae/QKkR78AjimHDIQQgsZ539XMLjOzzWa2eWQo/se4\ngEZQS78MlyN/ARVoELX1i/9RMaBRjNcvR/VKhV5B9k12grPLzFZJUvJ/5M+JSyGEK0IIG0MIG4vN\n/l8RBzJuUv3SlPc/9wpk3CT7ZV5+BwmYqlT9clSv5OgVZN9kJzjXSXpD8vMbJF1bn+EAmUS/AOnR\nL0B69AvgmDBFzcy+JOl8SUvNbJukD0m6XNLVZvZmSVslvTbNxiwE5UaOffe00lRj+lVE3lm3JA23\n+OsvRwKVRhbGo95OWrzPrT845KdJPBRJfbnzLj/9Zs1mP30iv/OAWy/3+W8159vb3Xr/Un9Oe/zq\nvW79jPan3PpP+9a4dWvxx2+VeHpVU49/vPNDtaWrxT7JEiKL2zSEqNWzX+aa8kL/nOrZ0Bm9zpnP\n91POTi76aS13b/fTz1Z1+yk++WXL3HpYWJ93i8td/np61vn3KSe3+x/DfbLfT13LD/rrCcV46mBo\n9tN65qO69ouZVHQe0iIJUNbr33d2bPPP510P+7dh73BXquEdVmmK3/Ese2FtH+PuHvJT+PZX/Hd/\nCzn/vnY4kqJWiCRPNRf88zaWttSb9xOgVrT4qaRPDvnH9M49/uPOrm3xZOR8j79vj5t/X9O633/A\naN0V+chwt3+bhSEnzatSS4zsserWL2ZSIU2IbqLsj3vxQ5HEskORZLLIc7TBRf79bP/y+GPLj1tO\ncOuvWX2nW3+saalb7xvxx9qU98/94UokHTGSXNg74p/7LZEkws7IY2Mu8vwmlq527y7/OengE/Fj\nGns8ajrg19c96O9D2xY/odAivaLc9PxJzgnP8BDCpZGLLqjzWIB5j34B0qNfgPToFyC96Zk2AQAA\nAMAsYIIDAAAAIDOY4AAAAADIDCY4AAAAADKjhhiNqQs5U6k1fWJaJMAlmsQRCbeQxdbjB4AoNMWT\nTtojSRm7Rxa49R9u99PSOh/zk12a9/pJLeW9fnpbGPHHU1m2yK33rHfLOr/LT0t7UZufgvXIgP/H\nksNgJI1nnL9ZaZHDnR/yb7hKMZKKV2Man1WmIUYtA4L5x3FgjZ+itud58eP+1uX3uPWfDvtJLiPb\n/G20bvFT/qzJT+eL3bK276Bbj/bXxtPc+sD6Ebf+smU/c+u3dvuJP4U+tywb8teP8ZhC4dj7n9jZ\naZEEymKff4fUvt0/19p2+cvn/IAh9S+Pv67Y/wI/0Wmg4tcHR/wHvXzO74Chkr98IZIYFUuG6h+M\npGRFUtpqVYw8aO960k9XW3Rf/KlM+y5/Xf3L/Nuh6VAk1bMv8lh76JBbz3U492WR+9Y5LzLu5t3+\nHVjxkH+scv3+/VrbUj9972DJTwiT4sl/3SV/XZ1F/wnfvsi5/6zOXW49F4lffaJviVuP9dD2Pv85\nWn/J76117fvd+qKCn7rWt9c/DmtvGuf57VY/5Wxwhb+u1i3+42nw0iwlKfJ4rVLkSfoU8Q4OAAAA\ngMxgggMAAAAgM5jgAAAAAMgMJjgAAAAAMoMJDgAAAIDMmNEUNQU/GS3UOM3KRVLUrOInfVT8sA2V\nW/y6tcQTHQbKfgrEnkgq1OCwv3xTZJ/71viDam4/3a3nh/xEjN3Pjazn2X7qxcsW3ufWi/LXv2e4\nw63nBvyD3dQTTyyL3f6xNK9KoT5JNJX8PE20qZPY8VXkuPSt8G/bjtP8dBdJKkdu3Ct3n+fWmw75\nyw+tWejWi93+eW6lSFJMh58Gk1vgn8/bXuTXLzzzLrf+650PuvVYilqsL6w8TsKfkxQGSQpSOPa4\nhbx/TlnOrzfv99OW2nb6x71zq5/GlhvwY9RyJf+ckqSeET81qjOSlFTM++d535CfxLS0w0+9Wtjk\nr78UefAcidR7h/3tthX99KxKJONuVVO3W7eyv/yCrZHIOkmtu/zbp9TsJzYWBiM92euvx5ojSV9e\nSmfGgjtzB/zUrdxI5Pbu8Zdv7lvq1heXFke33dzi3x7PatnhbztyrsWe0y1v8tPxdg/7ibnDZf/p\n9GBk/YVIfOxw7AlrjVqe9rfbvtV/DihJemSrW27bF7kdSpG+a2/16yOR5acpXZB3cAAAAABkBhMc\nAAAAAJnBBAcAAABAZjDBAQAAAJAZTHAAAAAAZMbMpqiZn5hVbvITFAoDkTSTSBJJxQ+NGHc8nlwh\nksAkqb/kp8R0NfW79ROW7nPrD57pr6fnBP8mad7nL19q9Q9G26kH3Prlz73GrV/Y5qee/GjQT6l6\n+MByt54fjBzUcdJj8oPx4+0p9Pspd+VWf75eLvpjysWDdxpDJC2t3Oafa8ML/eVP7do7zib823bv\noJ8kNbwoktp3pp9U1LLfH2slcs8W8n4a28Ayf9/CmT1u/ZVdfopa2UnxkqQH9q10680Hazv3MQHv\n+B/wE7lid0m5/mG33rbbf4ApPuXfx6vi37bNCyOpW+OMqRL88zOf87exuM1PmIqmpdUYZbq42X+8\ni4mNP5Ya15H3xxly/hFq2e3vrxRP+mrb4287N+wf09ASud3K/uORNTnnS6MEd0ZSsazJv78OLX49\nPxRPtF3U5N/mT410ufUn+v2ktpjukp+4ua7ZTw3d2u9v99Cw/xxqSYufaBizpOgvfyAyzliv2IB/\n/yZJtsBPA66VDfrbCJF0vdh5MVW8gwMAAAAgM5jgAAAAAMgMJjgAAAAAMoMJDgAAAIDMYIIDAAAA\nIDMmTFEzsyslXSxpdwjhuUntw5LeKmlPstgHQgjXT3YQuRE/7SEW7BJLXYulqxWGImkSpUiKVF88\njm1Zs5/IckrbTrf+/I4tbr1nuZ+scdvBE9z6nkjqVC6Su3Pu0ifc+vqCn662t+wnx2wvHefW9x30\nx1PsiSSWRdYvSeWWyA0duUHLLfWJooklbU3FTPRLvZQ6/VSgweV+fWixf3u0F+KpLMsKh9z6z3Vt\ncetNZ/upOY8dv8StD0bOkc6WIX/5Ef9GX9vmp0JdvPI+t/6KNj/l6bo+P13wwIN+ws7x+/1UmUZR\n134Jkjn3M6EQafQh/xyxQf82KfT59cqeSIpgLvb6oZ+oJ0mtBX8bzZHIx2IkRa2Y9/touJJ367GU\ns1zsQTVSjt0XFHL+ePYNtrv1dUU/mS435B/Twl4/7VCSNOD3avGQP9ZSm//4n+v17yMqkRQ15b1j\nPbXHrnnz+BLpOeuI9OKA34uh03+eJEn37PKfm7Tm/R5qL/jbGIk80Xy0b5lb39/kn7MxQ+VIMm7e\n7+lCJHm0txxPX/TEnj/boH8cqoOKpJnFnr/F7lsjpistLSbNOzibJF3k1D8eQjgr+TfrT9aAOWKT\n6BcgrU2iX4C0Nol+AVKZcIITQrhJkh/8DeAo9AuQHv0CpEe/AOlN5Ts47zSze+3/t3c3MXbd5RnA\nn/ec+z0znvF4HNtxjA0UCbmVmggLEHQRqRvKosAGlUqIRaWwoBJIbBAbuqnURaHtokIKSkQqIQoV\nqCC1qqARVUsXqCFEDcGNEnCCP8Z27Bl7Pu7cr3PeLuaiOPh5r++1770zc+7zk6KM37lzPv/vOec/\nd+YZs6fN7HD0IjN7wsyeM7Pnum3+410iM2Dkfulko/0xPZECGb1fcvWLzKx79stb7y3xH0YVKYr7\nneB8FcA7ATwKYBXAl6MXuvuT7n7O3c+Vq/z3NkQK7r76pZLyv1AsUnD31y+J+kVm0lD98tZ7S32a\n2yeyJ+5rguPu19w9c/ccwNcAvHe8myVSHOoXkeGpX0SGp34R4e4rR8rMTrj7av+fHwPw86G+0AEj\ngSMWBIvkJf6JNEhdC8InQpXbvN5Ziw/LS2s8AWepzH884mzjCq2fKvMfo11tLNH6SpDedr29QOtZ\nEKGxHKTZrKT83bVXW3x/e1s8aaYR/BRilJQHAGmLnzhP+Pn3YMD0gnS1YJcBj7dpnO67XybMMr7/\npW1+wCq3eALTf194R7iOKNHmdI2nJD22dJHWz8zz10fpT3kw/s/fPkbrW12e7nK68gatR/5l/fdp\nfe5ykP60GSfQzar7v7/kAEkI8q1t+vJ8m9fTh1aGWt1vWI2nG2W3+A3GBlx3PrD8K1qPEpTaGe/J\n+QpPSurlfBx60Ee9oI+2unx7dnr8vnCswVPOovV2PEh7m+fJU54O+F5tj1/PovtLeSNImRqQBMoY\nS9EbTwDoWxe5T+8vo/Bt/iNzXubpkwAwV+Vf80iNJ8U2c36NbwdxquXggXKjy5PdosTBaIxvBj0U\n9Va7znvi3XPX+HqDx1ivxCnBFiQFjpqWtl8MExP9TQCPA1gxs0sAvgTgcTN7FLthka8B+PQEt1Hk\nwFC/iAxP/SIyPPWLyPDuOcFx90+Q8lMT2BaRA0/9IjI89YvI8NQvIsN7kBQ1ERERERGRfUUTHBER\nERERKQxNcEREREREpDCmG41gQF6+O1HC8igVLUiliNK1gulalBZVu8VTMroL8bzvxnmesPODNk/o\n+Onc28JlMRstnqBxa42nnHmHb+uV04u0/seLz9N6Bh5/dmGH72+yxRM90g4/1gkPvwEA9Ob4sqLz\niSCEKErRS9v8E73abM/vSxutkeqHa3xMNW/Ff3/khxcf4594hCfg5Dnv7WqNp7G1tnnfeZNf2pJW\nkCK1zNPM/n3ld2n9fOsWrf/ghd+j9VMXeAMkLb5fch/c4V1yPFmaFYCkxtOQ8rmgXgqWE6RrJQs8\n4XJ7hY9ZAGikPMHrdsb/bkmtxMfVdpAKeGqeJ0ylQQJUJbhwR+mIl7Z5CmgSXLRX6vy+U2FxqwDm\nloO0rXp8TJPgnpq0+b5l83xZaTl4XCrzVKqcpfeNmMQ2KyxIwcuq/NkAALIgEXCxxJPAXlo/QevH\naxu0Ph/04oXtI7Teyvg4iJILmz0+zrp58GwV9Og7qtf5eo/ze5oH1zcAsCBxMEoi3O9m+wlPRERE\nREQKRRMcEREREREpDE1wRERERESkMDTBERERERGRwtAER0RERERECmO6KWoA4HcnQTgPTgIsSEsL\nXl9u8oSSdIcnQOQVPr+L0rgAIA1SmJo9njB1sXyI1pMu34n6NV5favIEjazGX3+pskzrV9/Bt7Nm\na7R+MUjFqa7x41BbCxKFekH0GQak5QXnvzsXnbfRUvdKQbravmEGT6PmIILdMdJz96NxmafT1K7H\nSTfzV4KkmPk4yYXpBK9fvs13uhRcC0otniyz+n6+/J138+1vJHw55Zv8klq/ShKVANgOXw4qPJFH\nBnAAvbuTsazBE8isxFPOsiAtLUz7DNLYfHOT1lmS6G80M574lQc3vcUKTzzsBElMGx1+LA5VglTD\nIMpyO0iAKif8XtvKeF/Ml3lS1a2MJzPWK0HqYJBUBYCOCQBhul6UlseeXYB4fGGHnJsRLuf7Sj7i\nvTLj4yBb5+mT3uHXwe77ToerOL3In1nWe3O0fqrOEwRXW/yZ6ETtNq33PL7fMYeqQVppEh1TPl53\ngpS2lvN6fYGvd9C59E2eamj1YIzvc3oHR0RERERECkMTHBERERERKQxNcEREREREpDA0wRERERER\nkcLQBEdERERERApjuilqDiQkICIIagnTn6LXR+k0SZd/QRqkaNXW4tQpt+CQBYksKQ+nQRqEJ9XX\neIJGusO3tXmMJ2hsGt+HK93DI9VvNnkiSbT9UcCIZQOOaZByFp3PtDPauAjHVxCuc1CNKy0tkuzw\nBKOoDgDpJk9JijJZRt6HDl+3NXmCjNd5SlX13cdp/Vh1g9azIA6pus7r1uapQihPP8iy0Gz479l5\nMHaSJr+4WZiuxa/NHqR3tZbiKK1GyvtlvcdTxQ5XebJhZL3Nl9PJ+TisBqloUSLVZeOpmwl4X5eD\nyNLtnPfp25du0vpG/WFaBwDf4Tdh6/Dzk7T5TSy6dmCLnwObI8e6eUC/pxwkmobS4BgGaWkWvH5Q\n4uCNnXla787zZbVz/qz0fzcfovU35vjyd7p8OUnwzFUv8evMyUaU0jbaGHm9vULrWcaX45X4mCZs\nzAJAkCC53x3QbhMREREREbmbJjgiIiIiIlIYmuCIiIiIiEhhaIIjIiIiIiKFoQmOiIiIiIgUxj0j\nfMzsFIB/AHAMgAN40t3/zsyWAXwLwBkArwH4uLvzaJW3LJCUeJAKki5PbsiqQepWkODVWeDzuMb1\nIEYrSAgDgDoPcUHaDVYeCUIpqus8cSMr833o1fhykjI/qL9o8rSZXhB/duMiT8U5ss53IG3z+qA0\nlLwUpKUFy4rOfzSORk3dexBj75dgnw66sSW+Rek+QaphuJjgOB8r8xS1zYw3XmkrWH4WpKiNmk5U\nMGPtlySBNe7O54vS0qzEr3lZjd8WrRddYIKxHCRDtY7E5/xs9TKtLyQ8FfBn2WlaT4IB3Q1uknMl\nnt42H6S6RYlRK5VtWq8mcdIic7LMT/VCsJ23ooQ7ABac/0jSGa1XLZ3O94nHfm8Zh2jfg9StdIUn\nfiHnx7wzFx/bc0cu0fpKObgIB47O8TF7tM6Xs1XiaXqdoLcaJf5A+UaLp7RFvbtY4deAZl6h9aV5\nnh7YPsKf6QCgfuEKrVsjSFfb54bpzB6Az7v7WQDvB/AZMzsL4AsAnnX3dwF4tv9vkVmnfhEZnvpF\nZDjqFZER3HOC4+6r7v58/+NNAOcBnATwEQDP9F/2DICPTmojRQ4K9YvI8NQvIsNRr4iMZqS/Mmdm\nZwA8BuAnAI65+2r/U1ex+7Yp+5onADwBAJV6/NaYSNE8aL/Uyocmv5Ei+8QD90u6MPmNFNkH1Csi\n9zb0D4+a2TyA7wD4nLu/5QfT3d0R/FaJuz/p7ufc/Vy5yn/mUKRoxtEvldLcFLZUZO+NpV+Su3//\nRqRoxtIrqXpFim+oCY6ZlbHbUN9w9+/2y9fM7ET/8ycAXJ/MJoocLOoXkeGpX0SGo14RGd4wKWoG\n4CkA5939K3d86vsAPgXgr/r//979bkQebEXaCRK5gsSUINgF5S2e0LGzwtMnajfj1JWkyxMu5i/w\nxIqkFSS1RXpBgstR/u5Xb44nu3jO61d2Fmn9epO/ZV3aCNLhgmPtaZAQlMWpWdF56zWC8xwktIzK\ngkP9QMscZ7+4jy9tbI9MfPuj5fd432UnDtN68xgftw8HaU4XsczXG3zLyAaM/1k23vuLAyytzoP0\ns5Rf/8Nt3QniNUsj/aQ3uovxWDhe2qT1lpf566u3aX29y1OPjlV5KmCUilZNeB/lHiSZjpiWdqq2\nRutHU34cTtT4/r66wpOtAKCR8/NvHb5vFiUwBvdmX+DvtFuLjJcHSE2cxrPYgJXz+qj34qP8+mvr\nfFwGYZUA4rS005UbtB4lX0ZpaUtl/kwXpaXNl3nCXz3lPXG0wsd4OeHjbKvHx3g5eJA5PseXf/1w\nkGQHoFEZ7Zq43w1zZf4ggE8CeNHMXujXvojdZvq2mf0ZgNcBfHwymyhyoKhfRIanfhEZjnpFZAT3\nnOC4+49B/3oNAOAPx7s5Igeb+kVkeOoXkeGoV0RGM52/UCUiIiIiIjIFmuCIiIiIiEhhaIIjIiIi\nIiKFoQmOiIiIiIgUxmj5lg/KgLx09+/IJV0eNegJ/326KCY4ihvOqnwex7YFAHqNIBoZQNoOIkfD\nGMUgonKHRwrmh3i8Z3eRx4R6MEXN23wffr3BYxo3mzxCMVp+VuH7u3OEv75+c/S5dJBEGv+aZbCK\nJIjo7dU0vz/wor6r8rhLL/O+6C7yPk2N13+1c5Qv//5TYOVBucNJPHh+m8elJh0e35osjvYHqb3J\n42QtiI/2NI7Wvdrjcf15cHE7XNqm9STI8b/U5tf/pVKT1qPx3wjioxdL/Fg0c96PabCdv+zy/opE\n93IA8CBKPmm2aN2q/F5rLX7PDqOS2b1/VtLig2e3UJ0/f7SX4uVkwcPJy60TtH6tc2ikTaoGkecr\nVd5zUQ81Uh4vHy1/K4izjuKml4NrwKEK78VL87Nzk9ITnoiIiIiIFIYmOCIiIiIiUhia4IiIiIiI\nSGFogiMiIiIiIoWhCY6IiIiIiBTGdFPUECegMVl1tLSHKMGo1+CfCIJgkJfj9VrO54Sto3Var0Zp\nIgtBasgRXt9+iJ+q9uEgUa7E029ubfCUtnKFH4w0CI4Jgj5Q2eD1PA6mQ9rh+5AG6WeRKHUvC85n\ntF45QMpBUlWJD7jOEk9IwnE+0KNEnpdu8XqU5OhR2lvAguQnGcCMJ5dF14Utnj4UXqrWbtFyvs3T\nk5IgGWpQ0l43WPtCwhORUizGC2PLKfHksCQYuO+sXqP1zZzf76K0tMWUb38SpLRtZnz5XefHJ0r7\nBABkGf+a2/xmFS6qzdOwUA6uKbMsC9JjN3jPIbhe5wOeULeyKq0fDtLMlst83Tfac7R+KOiVuvM0\nsyhxMHKxtUzrR4LtbAYH43bQK7/e5MvvLgy4AAXXLHSDh+V9Tu/giIiIiIhIYWiCIyIiIiIihaEJ\njoiIiIiIFIYmOCIiIiIiUhia4IiIiIiISGFMPUWNJchEmQ5R2syIYRUo7/CEmLzEVzBw+VG6UcqX\n1TrGEy6i1JfmUZ4m0l7iy8/LYWwTLfe6fPm9NZ6eUQq203gwTXjsomMNAEmP70OWjjj/Ds5NEmzr\noG2SYooSErNtfin8t6tnaf31l4/T+sOrwWAb9VtJwWJkEANIilq6fJi+2ltBRGSUlpXya2dS4Sla\nVuMpT3kjvsFc7vLkoyidbCuIs3x95witV4Po0HrK93nUtLR2zo9FM+Ovj1LR3jP3Gq3nwX2tsxA3\nmAcpamysAIAH5w3Red7iqV2ek/M8I+mI3uVJY+gFaVyHeRpg1oiP12KJJ/O1nZ/XaKxVggeE9R5P\nnI3sZHx8RKlo80FEbdRbjaBHE/Bj9J4jv6b1H5QepnUAyJbnaT1d2+Jf0AnO84ipoZOid3BERERE\nRKQwNMEREREREZHC0ARHREREREQKQxMcEREREREpjHtOcMzslJn9yMx+YWYvmdln+/W/MLPLZvZC\n/78PT35zRfY39YvI8NQvIsNRr4iMZpgUtR6Az7v782a2AOCnZvbD/uf+xt3/epQVGgl8yCo8cSHt\nBAkaQUBDlEwW5XCUdoI0mwFBJ70GT+KIWJCakgVpTlE6WRCggcYVvpydILWGHX8ASFt8OZXb0ev5\ngtI2r0dJdsCA8xam6I2WRBMtJ0pve0Bj7Re5T0GKSxr0/OKLvF9uvHKS1k++yht17jWeNuNBAld0\nfUCU/FQ84+sXw0jpPdbgCWHZAk9PSnvBOQnOlVWDNK5ynKJ2JOXjJwXfpuudQ7S+VObJXlGS1EaP\np7FFr0+Dm+SV9hKtR0lPpSDB6lbG93e1xdO2BkkaQRpW0JNe5Y9Fyc0N/vooMSxY/gM4OPeWqCfK\nPGksW+TnKKsOSFFLeYpaFjwgrnb42Jwr8TSzKBUtqi9XeM9lwfsI0XKipMNy8HAY9xa/zqQtWt5d\nVjN40IzuR9H9a5+kqN1zguPuqwBW+x9vmtl5APyuLzLj1C8iw1O/iAxHvSIympF+B8fMzgB4DMBP\n+qU/N7P/NbOnzYz/sQGRGaV+ERme+kVkOOoVkXsbeoJjZvMAvgPgc+6+AeCrAN4J4FHsflfhy8HX\nPWFmz5nZc9128MeCRApmHP3Syfhb3iJFM55+4T+yIlIk6hWR4Qw1wTGzMnYb6hvu/l0AcPdr7p65\new7gawDey77W3Z9093Pufq5c5X8lVaRIxtUvlXS0v6QschCNr1/479SIFIV6RWR4w6SoGYCnAJx3\n96/cUT9xx8s+BuDn4988kYNF/SIyPPWLyHDUKyKjGSZF7YMAPgngRTN7oV/7IoBPmNmj2M0cew3A\np4da4wjBVXmwdUHIRLzKKKUrCDmJkswGipLdgjSJIDwG1Q2efFFq8+V054K0qIvB9iT89aUg5ayy\nxevRMY3Ob5RkNmibLAvWnUYHm78+CBOJ9+HBjLdfZKyqN3mEzPLL/PXR2Kze4D/ikbR5opJXgotZ\nHidqzYjx9Ys70CM3hyRo9DI/J1GynVeDZMoWT2HypQVaT+vxDexM+Qat/06D3zC687+k9SsZv7m9\n0nmI1ucSvg+nSrdovRXcPM/WLtN6JEq8Oppu0vrPtt5G6/mAwDKrBOctqOdVnm7lc8G7Hm1+7KxE\nxteDpUsdmHuLzc/xT7R5SldW58d8kLUeX8f75nhPfKDxCq2/1l2h9WtdnrrWzPm4SYxfyxsJ3+c8\neAB5uLxO69G14aetM7T+wsYjtF5qxg/h1uLb6sG1EhV+3qLlTNswKWo/Bn98/9fxb47IwaZ+ERme\n+kVkOOoVkdFM5nvYIiIiIiIie0ATHBERERERKQxNcEREREREpDA0wRERERERkcIYJkVtvMivyKWd\nKC2LLyJKV4sSSiyPksD468PlDxAEaMSpcUGYSlbln4hSyOo3+YqzSpRMxl8fpbpFSWNR0lyYWDdg\nKp30RojWA4AgXS08ByJ9yQ5POatd5vWRRdeg7v1EM8pYREl1UeritTX+8uVFXg+SoewmTyA78U88\nnQkA/vTCZ2m9uxTsw3wwbkcM66o1+D5UyzzxbafN05N6PX7Tzrv8BuBt/vpkm9cbl/lyTpzfpnUA\nQBqsuxakqwXXPvmQTwAABCZJREFUCOsG6Xd1nq7mLF0vGHOFEyQL9q5eo/VKkMZ18j9O0DoAfGv9\ncVr/+pkP0rqVeQ95HjRLh4+bZCd6KOJlHAlSFtf5+PN5fq9IykH9So3W69f5Bi1fGXCvC8a4RSlq\nvr8fuvQOjoiIiIiIFIYmOCIiIiIiUhia4IiIiIiISGFogiMiIiIiIoWhCY6IiIiIiBSG+RRTPczs\nDQCv9/+5AuDG1Fa+P8zaPh/U/T3t7kf3eiNmvF9mbX+Bg7vP6pe9N2v7CxzMfVav7A/a54Phgfpl\nqhOct6zY7Dl3P7cnK98js7bPs7a/kzRrx3LW9heYzX2elFk7lrO2v8Bs7vMkzOJx1D7PBv2ImoiI\niIiIFIYmOCIiIiIiUhh7OcF5cg/XvVdmbZ9nbX8nadaO5aztLzCb+zwps3YsZ21/gdnc50mYxeOo\nfZ4Be/Y7OCIiIiIiIuOmH1ETEREREZHC0ARHREREREQKY+oTHDP7kJm9bGavmtkXpr3+aTCzp83s\nupn9/I7aspn90Mxe6f//8F5u47iZ2Skz+5GZ/cLMXjKzz/brhd7vSVO/FHPcqF8mQ/1SvHGjXpkc\n9Uvxxo765U1TneCYWQrg7wH8EYCzAD5hZmenuQ1T8nUAH/qt2hcAPOvu7wLwbP/fRdID8Hl3Pwvg\n/QA+0z+3Rd/viVG/FHrcqF/GTP1S2HGjXpkA9Uthx476pW/a7+C8F8Cr7v4rd+8A+EcAH5nyNkyc\nu/8ngLXfKn8EwDP9j58B8NGpbtSEufuquz/f/3gTwHkAJ1Hw/Z4w9cuuwo0b9ctEqF92FWrcqFcm\nRv2yq1BjR/3ypmlPcE4CuHjHvy/1a7PgmLuv9j++CuDYXm7MJJnZGQCPAfgJZmi/J0D9sqvQ40b9\nMjbql12FHTfqlbFSv+wq7NiZ9X5RyMAe8N1s7kLmc5vZPIDvAPicu2/c+bki77dMTpHHjfpFxq2o\n40a9IpNQ1LGjfpn+BOcygFN3/PuRfm0WXDOzEwDQ///1Pd6esTOzMnYb6hvu/t1+ufD7PUHqFxR3\n3Khfxk79gmKOG/XKRKhfUMyxo37ZNe0Jzv8AeJeZvd3MKgD+BMD3p7wNe+X7AD7V//hTAL63h9sy\ndmZmAJ4CcN7dv3LHpwq93xOmftlVuHGjfpkI9cuuQo0b9crEqF92FWrsqF/eZLvvVE1xhWYfBvC3\nAFIAT7v7X051A6bAzL4J4HEAKwCuAfgSgH8G8G0AbwPwOoCPu/tv/+LbgWVmfwDgvwC8CCDvl7+I\n3Z/9LOx+T5r6pZjjRv0yGeqX4o0b9crkqF+KN3bUL2+a+gRHRERERERkUhQyICIiIiIihaEJjoiI\niIiIFIYmOCIiIiIiUhia4IiIiIiISGFogiMiIiIiIoWhCY6IiIiIiBSGJjgiIiIiIlIY/w9ncQej\n/xlvhgAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 1008x504 with 8 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OTHXpRo24IPm",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}