VE-HTR_part_1and2.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "aadc0547",
   "metadata": {},
   "source": [
    "# Text recognition task 1\n",
    "-----------------------------------------\n",
    "*José Carlos Aradillas jcaradillas@gmail.com*\n",
    "    \n",
    "-----------------------------------------\n",
    "In this notebook, I document all the steps followed to complete the task for the Senior ML Research Engineer Test Task at Veriff.\n",
    "\n",
    "I will split the statement of the problem and the several steps asked by Veriff in the following format:\n",
    ">### Title of something asked by Veriff\n",
    "> Statement and description of the problem\n",
    "\n",
    "I document every step with comments and references preceding the different cells of code. I write also some inline code comments `# comment in the code` to clarify their porpuse.\n",
    "\n",
    "Let's start:"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f9365413",
   "metadata": {},
   "source": [
    ">## 1. Prepare the dataset\n",
    ">\n",
    ">We will start with preparing our dataset. For this step, you will need the [EMNIST](https://www.westernsydney.edu.au/icns/reproducible_research/publication_support_materials/emnist) dataset. This dataset is a set of handwritten character digits derived from the NIST Special Database 19 and the dataset structure that directly matches the MNIST dataset. Torchvision provides a built-in `torchvision.dataset` [module \n",
    "for the EMNIST dataset](https://pytorch.org/vision/main/generated/torchvision.datasets.EMNIST.html)."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ac6eda99",
   "metadata": {},
   "source": [
    "First of all, let's import all the neccesary libraries that will be use in the code to solve the task."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 428,
   "id": "4fcbc7d7",
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import torch\n",
    "from torch import nn\n",
    "from torch import optim\n",
    "import matplotlib.pyplot as plt\n",
    "from torchvision import datasets, models\n",
    "from torchvision import transforms as T\n",
    "import random\n",
    "from tqdm import tqdm\n",
    "import os\n",
    "from PIL import Image\n",
    "from torch.utils.data import DataLoader, Dataset\n",
    "from torchvision.datasets import VisionDataset\n",
    "from typing import Any, Callable, List, Optional, Tuple\n",
    "import time"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0a474ed0",
   "metadata": {},
   "source": [
    "Once the libraries have been imported, let's initialize the random seeds of every library I will use to randomize things, with the aim of achieving reproducibility. \n",
    "\n",
    "Completely reproducible results are not guaranteed across PyTorch releases, individual commits, or different platforms and furthermore, results may not be reproducible between CPU and GPU executions, even when using identical seeds [Reproducibility in Pytorch](https://pytorch.org/docs/stable/notes/randomness.html).\n",
    "\n",
    "However, let's limit the number of sources of nondeterministic behavior for a specific platform, device, and PyTorch release:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "88526178",
   "metadata": {},
   "outputs": [],
   "source": [
    "random.seed(60)\n",
    "np.random.seed(60)\n",
    "_=torch.manual_seed(60)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "17e341f8",
   "metadata": {},
   "source": [
    ">* Generate random sequences of random length (constrained to minimum 3, maximum 10) using characters from the EMNIST dataset.\n",
    ">\n",
    ">* Use the bymerge split that consists of 814,255 characters with 47 unbalanced classes. [The built-in implementation](https://pytorch.org/vision/main/generated/torchvision.datasets.EMNIST.html) provides an argument that specifies which one to use.\n",
    ">\n",
    ">* Apply RandomAffine transformations on each character with the following specifications: degrees=20, translate=(0.2, 0.1), scale=(0.7, 1.15) while creating sequence images.\n",
    ">\n",
    ">* There should be at least 50,000 generated samples in your dataset."
   ]
  },
  {
   "cell_type": "markdown",
   "id": "823511b9",
   "metadata": {},
   "source": [
    "**Solution:**\n",
    "First of all, let's download the EMNIST dataset. I create a trainset and testset objects setting the path where to download (if not previously downloaded) in `data_path = '~/.pytorch/EMNIST_data/'`, the bymerge split setting `split = 'bymerge'` and apply the RandomAffine transformations on each character in the `transform` argument.\n",
    "\n",
    "I also set a `trainloader` and a `testloader` objects with `batch_size=1`and `shuffle=True` to create the random sequences from this dataset. The generation of the random sequences will be done in the next cell.\n",
    "\n",
    "*Note that in the `transform` argument I also indicate that the images are converted into tensors, with the aim of concatenating the sequences using tensors (for me is more confortable than concatenating PIL Images).*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 187,
   "id": "0d521e90",
   "metadata": {},
   "outputs": [],
   "source": [
    "# Define a transform to load the data as Tensors\n",
    "data_path = '~/.pytorch/EMNIST_data/'\n",
    "split = 'bymerge'\n",
    "transform = T.Compose([\n",
    "    T.RandomAffine(degrees=20, translate=(0.2, 0.1), scale=(0.7, 1.15)),\n",
    "    T.ToTensor(),])\n",
    "\n",
    "# Download and load the EMNIST training  data\n",
    "trainset = datasets.EMNIST(data_path, split=split, download=True, train=True, transform=transform)\n",
    "trainloader = torch.utils.data.DataLoader(trainset, batch_size=1, shuffle=True)\n",
    "\n",
    "# Download and load the EMNIST test data\n",
    "testset = datasets.EMNIST(data_path, split=split, download=True, train=False, transform=transform)\n",
    "testloader = torch.utils.data.DataLoader(testset, batch_size=1, shuffle=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "d02c0e8b",
   "metadata": {},
   "source": [
    "In this cell I would check that the dataset in the **bymerge** split consists of 814255 characters with 47 unbalanced classes as it is commented in the statement:\n",
    "\n",
    ">* Use the bymerge split that consists of 814,255 characters with 47 unbalanced classes. [The built-in implementation](https://pytorch.org/vision/main/generated/torchvision.datasets.EMNIST.html) provides an argument that specifies which one to use."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "70a9f3d5",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The trainset has 697932 samples.\n",
      "The testset has 116323 samples.\n",
      "The total number of samples is 814255.\n",
      "The total number of classes in the set is 47.\n"
     ]
    }
   ],
   "source": [
    "train_size = len(trainset)\n",
    "test_size = len(testset)\n",
    "n_classes = len(np.unique(torch.concat([trainset.targets, testset.targets])))\n",
    "print('The trainset has {} samples.'.format(train_size))\n",
    "print('The testset has {} samples.'.format(test_size))\n",
    "print('The total number of samples is {}.'.format(train_size+test_size))\n",
    "print('The total number of classes in the set is {}.'.format(n_classes))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "146206be",
   "metadata": {},
   "source": [
    "We also can check that the classes are unbalanced if we plot the histogram of the targets in the training set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "fa98bff1",
   "metadata": {},
   "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": [
    "plt.hist(trainset.targets.numpy(), n_classes)\n",
    "plt.title('Histogram of samples per class')\n",
    "plt.xlabel('Class label')\n",
    "plt.ylabel('Number of samples')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "80eab2b2",
   "metadata": {},
   "source": [
    "As you can see in the previous histogram, the trainset is unbalanced, the classes regarding the numerical digits (from 0 to 9) has around 35k samples and the rest of characters vary from few thousands to 30k samples.\n",
    "\n",
    "Now we check the type and the sizes of the loaded images using the trainloader (remember batch size of 1 and shuffled in order to build later our sequence dataset)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "8a702f7a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'torch.Tensor'>\n",
      "torch.Size([1, 1, 28, 28])\n",
      "torch.Size([1])\n",
      "tensor([46])\n"
     ]
    }
   ],
   "source": [
    "dataiter = iter(trainloader)   #To iterate through the dataset\n",
    "images, labels = dataiter.next()\n",
    "print(type(images))\n",
    "print(images.shape)\n",
    "print(labels.shape)\n",
    "print(labels)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3333ca4d",
   "metadata": {},
   "source": [
    "In the previous output, we can check that the images ar torch tensors and their size is 28x28 and has 1 channel, therefore they are gray images. The labels are an unique digit that indicate the class.\n",
    "\n",
    "Now let's show some somples (in this case 10, setting `n_samples_per_class=10` in the following code. This code looks for the images in the training set and built an image concatenating all the images for the same class in the same column. The samples of the classes are ordered."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "e15304ea",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1120x800 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "dataiter = iter(trainloader)\n",
    "n_samples_per_class = 10\n",
    "image_list=[[]]*n_classes\n",
    "counter = 0\n",
    "while counter<n_samples_per_class*n_classes:\n",
    "    new_image, new_label = dataiter.next()\n",
    "    new_image = new_image.mT # Transpose the images, the original ones are not in the correct visual position.\n",
    "    if not isinstance(image_list[new_label],torch.Tensor):\n",
    "        counter+=1\n",
    "        image_list[new_label] = new_image\n",
    "    elif image_list[new_label].shape[2]<28*n_samples_per_class:\n",
    "        image_list[new_label] = torch.concat([image_list[new_label], new_image],2)\n",
    "        counter+=1\n",
    "plt.figure(figsize=(14, 10), dpi=80)\n",
    "plt.imshow(torch.concat(image_list,3).squeeze().numpy(), cmap='Greys_r')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3142180f",
   "metadata": {},
   "source": [
    "Generating the random sequences of random length (constrained to minimum 3, maximum 10)..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "9c83a032",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['ses_mnist.png', '56', '7', '9']"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "'ses_mnist.png 56 7 9'.split()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "683532e9",
   "metadata": {},
   "source": [
    "**EMNIST images are transpose with respect to MNIST ones**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "560edd78",
   "metadata": {},
   "outputs": [],
   "source": [
    "def generate_random_lengths(min_length=3, max_length=10, n_sequences=1):\n",
    "    assert max_length>=min_length, 'max_length={} should be >= min_length={}'.format(max_length, min_length)       \n",
    "    return np.random.randint(min_length, max_length+1, n_sequences) #max-length+1 so length 10 is also included\n",
    "\n",
    "def create_folder(path):\n",
    "    if not os.path.isdir(path):\n",
    "        print('Creating root folder '+ path +'...') \n",
    "        try:\n",
    "            os.mkdir(path)\n",
    "        except OSError:\n",
    "            print ('Creation of the directory {} failed'.format(path))\n",
    "        else:\n",
    "            print ('Successfully created the directory {}'.format(path))    \n",
    "\n",
    "def save_sample(root, image, seq, idx, dataset='demo'):\n",
    "    transformToPIL = T.ToPILImage() # transformer for saving Images as PIL    \n",
    "    if isinstance(root, torch._six.string_classes):\n",
    "        root = os.path.expanduser(root) # In case we use '~/.pytorch' path for example\n",
    "    create_folder(root)\n",
    "    database_path = os.path.join(root,'SEMNIST')\n",
    "    create_folder(database_path)\n",
    "\n",
    "    set_path = os.path.join(database_path,dataset+'set') # Check if set is train or test\n",
    "    create_folder(set_path)\n",
    "    \n",
    "    set_img_path = os.path.join(set_path, 'img')\n",
    "    set_labels_path = os.path.join(set_path, 'labels')\n",
    "    create_folder(set_img_path)\n",
    "    create_folder(set_labels_path)\n",
    "    \n",
    "    img_path = os.path.join(set_img_path,'semnist_' + dataset + '_' + '{:0>6}'.format(idx) + '.png') \n",
    "    seq_path = os.path.join(set_labels_path,'semnist_' + dataset + '_' + '{:0>6}'.format(idx) + '.txt')\n",
    "    imagePIL = transformToPIL(image.squeeze())  \n",
    "    imagePIL.save(img_path) # Saving image into path\n",
    "    seq = ' '.join(list(map(str, list(seq.numpy())))) # Convert seq tensor into string\n",
    "    \n",
    "    f = open(seq_path, \"w\")\n",
    "    f.write(seq)\n",
    "    f.close()\n",
    "    \n",
    "def generate_sequences(dataloader, sequence_lengths, save = False, root='~/.pytorch/SEMNIST_data/', dataset='demo'):\n",
    "    num_seq = len(sequence_lengths)\n",
    "    num_source_characters = len(dataloader)\n",
    "    num_target_characters = np.sum(num_seq)\n",
    "    assert num_source_characters>=num_target_characters, 'There are not enough characters in the source dataset to generate {} sequences.'.format(num_seq)\n",
    "    dataiter = iter(dataloader)\n",
    "    if not save:\n",
    "        image_list = []\n",
    "        label_list = []\n",
    "    for seq, seq_length in enumerate(tqdm(sequence_lengths)):\n",
    "        generated_image = []\n",
    "        generated_seq = []\n",
    "        for ind in range(seq_length):\n",
    "            char_image, char_label = dataiter.next()\n",
    "            generated_image.append(char_image.mT) # Transpose the image because the original are not visually align\n",
    "            generated_seq.append(char_label)\n",
    "        generated_image = torch.concat(generated_image,3) # Concatenate (w dim) the characters images to build the sequence image.\n",
    "        generated_seq = torch.concat(generated_seq)\n",
    "        if save:\n",
    "            save_sample(root,generated_image, generated_seq, seq, dataset=dataset)\n",
    "        else:   \n",
    "            image_list.append(generated_image)\n",
    "            label_list.append(generated_seq)\n",
    "    if not save:\n",
    "        return image_list, label_list"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "cef18cba",
   "metadata": {},
   "source": [
    "With the following code, we will generate our dataset, which we will call `SEMNIST` (from Sequential EMNIST) and we will save the data following the next structure:\n",
    "\n",
    "```\n",
    "root/SEMNIST/\n",
    "    trainset/\n",
    "        img/\n",
    "            semnist_train_abcdef.png\n",
    "            semnist_train_tvwxyz.png\n",
    "        labels/\n",
    "            semnist_train_abcdef.txt\n",
    "            semnist_train_tvwxyz.txt\n",
    "            ...\n",
    "    testset/\n",
    "        img/\n",
    "            semnist_test_abcdef.png\n",
    "            semnist_test_tvwxyz.png\n",
    "        labels/\n",
    "            semnist_test_abcdef.txt\n",
    "            semnist_test_tvwxyz.txt\n",
    "        ...    \n",
    "    validset/\n",
    "        img/\n",
    "            semnist_valid_abcdef.png\n",
    "            semnist_valid_tvwxyz.png\n",
    "        labels/\n",
    "            semnist_valid_abcdef.txt\n",
    "            semnist_valid_tvwxyz.txt\n",
    "        ...\n",
    "        \n",
    "```\n",
    "In the structure showed above, you can notice that there are two main directories in the `SEMNIST` folder, the `trainset`folder and the `testset` one. Inside the folders, you can find all the images form each dataset in a subfolder called `img` with the name `semnist_train_abcdef.png`in the case of training samples in `.png` extension and `abcdef`is the index of the image. There is also a subfolder called `labels` with `semnist_train_abcdef.txt` a `.txt` file with the same identificator. This file contains a string with the sequence of labels. In the `testset` folder we have the same structure.\n",
    "\n",
    "There also exist a `demoset`in order to try different modules of the code. This demoset has 10,000 samples.\n",
    "\n",
    "In the statement of the task, they tell us the following about the sequences:\n",
    "\n",
    ">* Generate random sequences of random length (constrained to minimum 3, maximum 10) using characters from the EMNIST dataset.\n",
    ">\n",
    ">* There should be at least 50,000 generated samples in your dataset.\n",
    "\n",
    "To generate the lengths of the sequences, we will use the method `generate_random_lengths(min_length=3, max_length=10, n_sequences=1)` which generates `n_sequences` integers between `min_length` and `max_length` (included). This method returns an array with the `n_sequences` lengths.\n",
    "\n",
    "As showed before, in the **EMNIST** training set (from which we will create the **SEMNIST** training set) we have 697,932 samples available. If we take sequences of lengths between 3 and 10, the mean length of a sequence is 6.5. It does it mean that we can generate the minimun number of 50,000 samples and even 100,000 since 100,000 sequences of mean length 6.5 characters means that the total number of expected characters in the SEMNIST train set is 650,000 that is less than 697,932.\n",
    "\n",
    "So, the total number of sequences (`n_training_sequences`) in the new SEMNIST train set will be 100,000.\n",
    "\n",
    "Following the same analysis we will generate 16,000 number of test sequences `n_test_sequences=16000`.\n",
    "\n",
    "*In the task they tell us that the number of generated samples should be at least 50,000. We fullfill this constraint and we consider that generate more samples is better while having a bigger training set is it always in deep learning tasks.*"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 185,
   "id": "6ed2deda",
   "metadata": {},
   "outputs": [],
   "source": [
    "n_train_sequences = 80000\n",
    "n_test_sequences = 16000\n",
    "n_valid_sequences = 20000\n",
    "train_seq_lengths = generate_random_lengths(min_length=3, max_length=10, n_sequences=n_train_sequences)\n",
    "test_seq_lengths = generate_random_lengths(min_length=3, max_length=10, n_sequences=n_test_sequences)\n",
    "demo_seq_lengths = generate_random_lengths(min_length=3, max_length=10, n_sequences=n_demo_sequences)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "53f0dd80",
   "metadata": {},
   "source": [
    "Once we have the arrays with the sequence lengths for the 100,000 training set sequences and the 16,000 test set sequences, we have to generate the sequences and we will save it in disk to have it available to be loaded with a dataloader and save space in memory. It is not reccomended to have all the sequences loaded in the machine memory when training since it can saturate. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 188,
   "id": "8ad84d5b",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  4%|█▍                                      | 36/1000 [00:00<00:04, 202.81it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Creating root folder /Users/jaradillas/.pytorch/SEMNIST_data/SEMNIST/demoset...\n",
      "Successfully created the directory /Users/jaradillas/.pytorch/SEMNIST_data/SEMNIST/demoset\n",
      "Creating root folder /Users/jaradillas/.pytorch/SEMNIST_data/SEMNIST/demoset/img...\n",
      "Successfully created the directory /Users/jaradillas/.pytorch/SEMNIST_data/SEMNIST/demoset/img\n",
      "Creating root folder /Users/jaradillas/.pytorch/SEMNIST_data/SEMNIST/demoset/labels...\n",
      "Successfully created the directory /Users/jaradillas/.pytorch/SEMNIST_data/SEMNIST/demoset/labels\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████████████████████████████████| 1000/1000 [00:02<00:00, 413.15it/s]\n"
     ]
    }
   ],
   "source": [
    "SEMNIST_folder = '~/.pytorch/SEMNIST_data/'\n",
    "generate_sequences(trainloader, train_seq_lengths, save = True, root=SEMNIST_folder, dataset='train')\n",
    "generate_sequences(testloader, test_seq_lengths, save = True, root=SEMNIST_folder, dataset='test')\n",
    "generate_sequences(trainloader, demo_seq_lengths, save = True, root=SEMNIST_folder, dataset='demo')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "76a044eb",
   "metadata": {},
   "source": [
    "**Now we have our dataset created!!**\n",
    "\n",
    "The next step is to build the Dataset class which help us to load the data during training, we also need a Dataloader which loads the Dataset and batch the data during batch training."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 448,
   "id": "3308aa42",
   "metadata": {},
   "outputs": [],
   "source": [
    "class SEMNIST(VisionDataset):\n",
    "    def __init__(\n",
    "        self,\n",
    "        root: str,\n",
    "        dataset: str = 'train',\n",
    "        img_type: str = 'GrayScale',\n",
    "        transforms: Optional[Callable] = None,\n",
    "        transform: Optional[Callable] = None,\n",
    "        target_transform: Optional[Callable] = None,\n",
    "    ) -> None:\n",
    "        super().__init__(root, transforms, transform, target_transform)\n",
    "        assert os.path.isdir(self.root), 'Couldn`t find {} folder'.format(self._root)\n",
    "        self._img_path = os.path.join(self.root, 'SEMNIST', dataset + 'set','img')\n",
    "        self._labels_path = os.path.join(self.root, 'SEMNIST', dataset + 'set','labels')\n",
    "        assert os.path.isdir(self._img_path), 'Couldn`t find {} folder'.format(self._img_path)\n",
    "        assert os.path.isdir(self._labels_path), 'Couldn`t find {} folder'.format(self._labels_path)\n",
    "        self.data_list = os.listdir(self._img_path)\n",
    "        self.id = lambda id_extended: os.path.splitext(id_extended)[0]\n",
    "        self.data_list = [self.id(path) for path in self.data_list]\n",
    "        assert img_type in ['GrayScale', 'RGB'], 'Image type {} not supported'.format(img_type)\n",
    "        self.img_type = img_type\n",
    "    \n",
    "    def __len__(self):\n",
    "        return len(self.data_list)\n",
    "   \n",
    "    def __getitem__(self, index):\n",
    "        img_path = os.path.join(self._img_path, self.data_list[index]+'.png')\n",
    "        labels_path = os.path.join(self._labels_path, self.data_list[index]+'.txt')\n",
    "        if self.img_type == 'RGB':\n",
    "            img = Image.open(img_path).convert('RGB')\n",
    "        elif self.img_type == 'GrayScale':\n",
    "            img = Image.open(img_path)\n",
    "        f = open(labels_path, \"r\")\n",
    "        target = torch.tensor([int(x) for x in f.readline().split(' ')])\n",
    "        f.close()\n",
    "        if self.transform is not None:\n",
    "            img = self.transform(img)\n",
    "        return img, target"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bc21025d",
   "metadata": {},
   "source": [
    "When performing batch training, Pytorch needs that all the input images to be tensors of the same shape (height and width). The next custom `collate_semnist` method is created to pad the smaller images in each batch to have the width of the widest image in the batch. It also pad the targets to have the same length.\n",
    "\n",
    "The `collate_semnist` function will be an argument of the `Dataloader` which will load our batched data."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "c7e22798",
   "metadata": {},
   "outputs": [],
   "source": [
    "def collate_semnist(data):\n",
    "    '''  \n",
    "    We should build a custom collate_fn rather than using default collate_fn,\n",
    "    as the width of every image is different and merging images (including padding) \n",
    "    is not supported by default. It also pad the labels sequences to have same length. \n",
    "    Args:\n",
    "        data: list of tuple (image, label)\n",
    "    Return:\n",
    "        padded_img - Padded Image, tensor of shape (batch_size, channels, height, padded_width)\n",
    "        width - Original width of each image(without padding), tensor of shape(batch_size)\n",
    "        label - Padded sequence, tensor of shape (batch_size, padded_length)\n",
    "        length - Original length of each sequence(without padding), tensor of shape(batch_size)\n",
    "    '''\n",
    "\n",
    "    # sorting is important for usage pack padded sequence when using lstm models. It should be in decreasing order.\n",
    "    data.sort(key=lambda x: len(x[1]), reverse=True)\n",
    "    images, sequences = zip(*data)  # unzip the images from sequences to edit them more comfortably.\n",
    "    width = [img.shape[2] for img in images] # save all the original images width\n",
    "    length = [len(seq) for seq in sequences] # save all the original sequences length\n",
    "    padded_img = torch.zeros(len(images), images[0].shape[0], images[0].shape[1], max(width)) # initialize the batch tensor with zeros\n",
    "    padded_seq = torch.zeros(len(sequences), max(length)).long() # initialize the sequence tensor with zeros\n",
    "    for i, img in enumerate(images):\n",
    "        end = width[i]\n",
    "        padded_img[i,:,:,:end] = img # insert the images in the first width part of each place \n",
    "    for i, seq in enumerate(sequences):\n",
    "        end = length[i]\n",
    "        padded_seq[i,:end] = seq  # insert the sequences in the first length part of each place \n",
    "    return padded_img, padded_seq, torch.from_numpy(np.array(width)), torch.from_numpy(np.array(length))"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "72f0c0e3",
   "metadata": {},
   "source": [
    "Now we have our dataset preparated and all the methods to load the dataset to feed our models, following the **Pytorch** Dataset framework. The steps for load the dataset are the following.\n",
    "1. Instantiate the EMNIST trainset, with the command `trainset = SEMNIST('~/.pytorch/SEMNIST_data/', dataset = 'train', transform = transform)`. We apply `ToTensor()` and `Normalize` transformations in the images to get directly the inputs of our models.\n",
    "\n",
    "2. Instantiate the EMNIST testset, aplying the same command sentence that for the training case but the argument in the call is `dataset='test'`.\n",
    "\n",
    "3. Instantiate both dataloaders in shuffle mode to randomize the batches, we choose the batchsize and set the `collate_fn` argument with our custom `collate_semnist` explained above."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 242,
   "id": "34f301c0",
   "metadata": {},
   "outputs": [],
   "source": [
    "transform = T.Compose([T.ToTensor(),\n",
    "                              T.Normalize((0.5,), (0.5,)),\n",
    "                              ])\n",
    "trainset = SEMNIST('~/.pytorch/SEMNIST_data/', dataset = 'train', transform = transform)\n",
    "testset = SEMNIST('~/.pytorch/SEMNIST_data/', dataset = 'test', transform = transform)\n",
    "trainloader = DataLoader(trainset, batch_size=64, shuffle=True, collate_fn=collate_semnist)\n",
    "testloader = DataLoader(testset, batch_size=64, shuffle=True, collate_fn=collate_semnist)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8c9776ab",
   "metadata": {},
   "source": [
    "In the following code we show some imformation about the **SEMNIST** dataset we have just loaded. We check that the train set has 100,000 samples and the testset has 16,000 samples. We also show some samples of pairs image, target in the training set."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 243,
   "id": "790f96d7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The train set has 100000 samples.\n",
      "The test set has 16000 samples.\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAxsAAAItCAYAAABRimfNAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/YYfK9AAAACXBIWXMAAAsTAAALEwEAmpwYAAEAAElEQVR4nOy9d3hdV5nv/1mnF+mo915tucQ9jkt64vSQhE7oEBgYyGVol99chnJhhgGGuXSmwFAGGGoSkkASJyZOsePE3ZItq/fey+ll/f7Y2juSLNuSrOZkfZ5Hj6R9dlm7nHPe73qbkFKiUCgUCoVCoVAoFAuNabkHoFAoFAqFQqFQKF6bKLGhUCgUCoVCoVAoFgUlNhQKhUKhUCgUCsWioMSGQqFQKBQKhUKhWBSU2FAoFAqFQqFQKBSLguVCLwohXpelqoQQ7N69mxtvvBG3243ZbOYHP/gBDQ0Nyz00hULxOkVKKZZ7DAqFQqFQzJULio3XK0IIkpKSKCgoIDExEbPZTGJiIhaLhUgkstzDUygUCoVCoVAoLguU2JiGEAIhBOXl5dx2223Ex8djNpv5wx/+QHNzM0NDQ8RiseUepmKR0O+/EAIppbrXCoVCoVAoFJeAEhvnwWQyYTabjd8lJSVs2rSJgwcP4vf7mWszxLi4OJKSksjLyyMpKYm6ujrGxsbo6elRBu0yoAsKh8OB2Ww27qnNZsNkMmG1WgkGgwQCAYA532+FQqFQKBQKhRIbs0IIwc6dO7FarZw6dYpgMEg0Gp3TPtLT09m8eTO7du2itLSUv/zlLzQ1NbF//37DoFUsHbqgSEpKwm6309vbSywWIy4uDovFgtPpZGRkhHA4bIhBJTgUCoVCoVAo5oYSG7NECC03MxqNztoTYbFYjO22bdvGPffcw5o1a0hKSqKxsREpJRaLugVLjRACu92O3W7HYrFgsVhYtWoVLpeL7OxszGYzFouFhoYG6urqCAaDRCIRgsEgsVhMeaIUCoVCoVAoZomydGeBHruvC43ZzHCbTCbsdrshNkpKSrjmmmtISkrCarVSVlbG0NCQEhtLjBACk8mEzWbD6XRiMpmwWCyUlZWRmJhIfn4+JpPJuOe9vb14vV5DaEz2dCgUCoVCoVAoLoyydC9CY2MjfX19/PznP+fkyZOMj49fdBur1UpmZibve9/7CAQCNDY2snHjRqOyVSQS4c9//jP79u2b1f5WArpomozZbAZY8RW6zGazIR6sVivx8fEUFRWRn58PaPdr27ZtZGRkUFpayuDgIEePHsXpdFJQUEBTUxPBYNDwhIyPj6uQKoVCoVAoFIpZoMTGNCwWCwkJCbhcLkwmE/39/Zw8eZKjR49SV1d30VwNPek4NTWV3bt34/V6sVqtZGRkIIQgFAoRCoWorq7m7NmzS3RW80cXGSaTyUiWn040Gl3RxrfJZDJ+u91uEhMTKSgoIC8vj3A4jMViobi4mPT0dLKysjCbzVitVlwuF0lJSXR2diKEmPHcFa9PpovvuT7/k4tP6B6zlfweUigUCoViviixMY3du3fz9a9/nZycHDweD06nc9bb6gbqjTfeSEVFBVu2bMFqtXLddddRV1fHD3/4Q44fP05NTQ3V1dWLeBYLg8lkwuPxGMZ6dnY2q1atIhqNEolEOHr0KENDQ4bhtdKMJV0gxMXF4XK5KCgooKioiLVr17Jx40YKCgro7+8nEomwevVq3G43DoeDuLg4otEotbW1OBwOhoeH8fl89PX1EQwGV9x5KpYOPQzP6XQa5ZH1CYS57GPTpk3k5+eTkpJCIBBg7969jI+P4/V6F3H0CoVCoVAsPfMWGyaTacrs3vlmfaWURCKRy8ZAS0pKIj8/3+ivMZvZbH29vLw8MjMzWbNmDWVlZZhMJmKxGH6/n46ODk6ePMnBgwdpampagjO5NPTSsCaTCYfDQWJiIqtXr6aiooJwOEwkEqG9vZ1QKEQkEplzda7FGO/0Z2yyN8blcpGXl0dOTg5FRUVkZWWRlJSExWIhFosRHx9v/C2EICEhgcTERFwuF3a73biXKl9jZSCEwOl0IqXE7/cv2XH15yE9PR2AUCjE8PAwg4ODs9pefx5LS0spLi4mJSWF8fFxzpw5Q29vLz6f77L5rFQoFAqFYjbMW2y43W7sdru2E4tlygz4ZKLRKG1tbZdFeVc9tGG264ImppKSkkhLS+Mf/uEf2Lp1KxkZGUgpqa2t5ZVXXuFf/uVfGBsbw+v1zmkGdDbookBnoYxhi8Vi9J/Izs7mrW99K1dddRUbN240KjO53W4qKyvZu3fvss7ICiGwWq1Eo1FD9OjL9PuUnJzMtm3bKCwspLy8HLfbjRCC7OxsrFYrJpOJUChEVVUVNpuN5ORkXC6X8RqcK7AVS8vkkD6bzcaWLVsIh8McPXp0SSY0zGYzmzdvJi8vj127dhGJROjp6eHo0aM89dRTs9qHw+EgPj6et771rWzbto1YLIbX6yUxMZFTp07xi1/84rKanFEoFAqF4mLMW2xYrVbcbjdbt24lLi6O+Pj4cwwxIQSxWIz+/n6Gh4d55ZVXLmhsSylX7MyeEAKbzUZubi4ej4fc3FyjfKoerlNSUkJycjJ+vx+v18uJEyc4c+YMnZ2dRCKRBZ0V12dI09LS8Hg8gDbL2t3dTTgcBjAqaM0H/R7Y7XZSU1MpKysjMzMTt9tNNBrF4XCQnp5OcnKyUeJ3ue6bbnzq446Pj8fpdJKbm0tcXBwJCQmkp6dTWFhIVlYWHo8Hu92O1Wo1RJWeQB4KhfD7/YyPj1NTU0NbWxu9vb34/f4V79VYqeFsC4nZbOa6664jKysLk8nE8PDwkghAm81GXFwcmzdvprCwkI0bNxIMBrFarTQ1NRmer9mM3263k5iYSFxcHKFQiHA4bHhMYGYvneLCCCGuA56dtKhGSrl6eUajULx2EUKMA+5Ji4qklM3LNBzFZcK8xYbdbiczM5NPfepTZGVlGTHMk5ls/AwODvLggw/S0dFx3n1GIhGampoMY3klIYQgPj6et7/97ZSXl7Njxw7i4+NJSEigt7eXzs5OysrKcDgcPP3005w+fZo//vGPdHd3L7g3A7Tr73a7ueGGGyguLkYIwdjYGA8//LDhZfD5fPOudqVXmEpMTKSwsJBt27aRkpKCw+FASkk0GqWkpITBwcEpJX6XAz0vQxd/GzduJCsri927d5ORkUFaWpohiNxutyHO9G2FEEZomJSSxsZGDh48SGNjI9XV1YyNjREMBpft/GaLXkZ5Jb5/FgJd8H/1q19lzZo1/OY3v6GyshKr1bro3oDk5GTy8/N5//vfT35+Pm63G5/Ph8PhoKGhwfCgXUxw6HldcXFxOBwOotHorL2pilnxHLAf6J+8UAhRBtwH3AKUARnAEHAI+LaU8llmQAhRDPwfYM/ENoNooubLUspZV/gQQmwE7gFuBoqBFKAPeB74ppTy2Hm2MwMfB94/MW7/xJi/KqU8OIfjXwM8AGwCstCMxS6gEviOlHLfDNtcCdwLbJzYLgPokFLmzva4k/ZVAbxt0r7yJl6ySinPW85QCLEb+AywAcgEeoEq4LtSyifnOo5J+y0HjgMu4FdSynfOsE4zUHCeXfRIKTPne/xpx0kHPg3cPnG8ENAM/Ab4Nynl2Cz3sx74X8AWIBfwoF2vGuCHwMNyhg/JeTxj/wTY0J7nDbM8TcXrnHmLDb/fz+DgIPX19Xi9XoaGhohEIkQiEdasWUNRUZERdhKNRnG73dx7770Eg0Hi4+MJBoNTQm/00JyzZ88yMjJCR0cHg4ODFxQnS4HFYsHlclFUVERqairXXHMN2dnZpKSkYLFYpvTgOHHiBGNjYzz55JN0dnbS2dm5aKVtLRYLbreb0tJStm/fjt1uJxQKkZWVRXd3N4cPH6atrQ2v1ztvI0z32Fit1inhRPr5Dg4OGvddSrlsM7L6uNxuN9nZ2eTl5ZGfn09cXJwREqUb4mazmWg0isVimXI+4XAYr9dLc3Mzra2tdHd309fXRyAQMKpt6T8XQwgxpaGjxWLBbrcTCATmlV+g55BcyCi12WzceuutAPz2t79d0jyG+ZCcnMzatWvp7+9ncHDQEHQzeeJ0r8KePXvYuHGj0QulpaWF1tbWRa/kZDKZuPXWW9m0aRNpaWmGF81sNpOYmGiE5F0MvZmk3qVeSsno6Ch9fX10dXUxPDw86z4+ivOyX0r5pRmWfwV4K3AG+AuaaFgF3A3cLYT4X1LK707eQAixGU1YeIC/ohl/ecAbgbuEEDdJKQ/Nclz/BmwHjgIPAeNohvfbgDcJId4ipXx42vHFxDHfhGYwfh9InjiP54UQb5RS/mmWx79h4ufliXPxAvkT53+XEOKrUsp/mLbNO9CM1zBQjSY25sstwBeAKFAHBADHhTYQQnwEzUj2Ag8D7WhG9H3AbUKIz0sp/3GuAxFCWID/Bmbjqh4Bvj3D8gX5YhdCFKLdk3Q0kfwE2nXZA3wDeKcQ4iop5Ww+0LegCYBDwEG0sWcCdwF/BH4JvGva8ef8jEkp/2nS2JXYUMyKeYuN8fFxhBCcOHGC06dPc+zYMbxeL16vlw996EOUlJQYBp2UEpfLxVve8hbDcJguNoLBIIFAgMrKSjo7O6msrKSurm7ZxYbNZiMhIYGCggKEEGzZsoW4uDijX4YeBhEKhfjNb37DsWPHOHPmDD6fb9HHFR8fz6ZNm9i8eTMOhwMhBDt37qSzs9MwNltbWy/JgLHZbMaPbuxGo1HC4TA9PT309vYaXpDlDuNJSkoiLy+P0tJSysvLjefP4/EYYiMSiRAIBIzO4bqXJhAIMDo6aoRONTQ0MDw8TCAQMBr5zfa8hBBG6WTASLDXxctcr4/JZCItLc0Is5kJj8fDhz70IaSUPP744/M6zlKSmZnJnj17aGpq4syZMzQ3N5/X2Ha5XGRmZvKe97yHm266CZPJRCAQoLa2lhMnTiyq2NCLJNx1111cf/31U+6ryWQiISEBj8djiNiL7cdutxv5bVJKent7aW1tpa2tje7ubiU2Fo8nga9LKY9PXiiEuBZ4GvimEOL3UsquSS//BE1ofFJK+f8mbbMDzSPxCyHEWinlbFyJvwLeKaWsn3b8+9GMwP8UQvxZSjnZDf42NCPwIHCjlDIwsc2/AS9ObPPXWc58//NMIkwIkQMcA/5eCPHDaef/M+DnwGkpZUgIcSkP5hPAS8ApKaX/Il4DhBBW4GtoomSLlLJm0mv/hOaV+D9CiH+RUs7V7fz3aELvM8B3LrLu8HnE60LxGTSh8SUp5Zf1hRPehr1oAvHNwC9msa/fSCl/Nn2hEMKDJkDeKYT4npTylUkvL+QzplCcl3mLjWg0itfrZd++fVgsFjo7Ow0j9A9/+AM1NTWUlpaSlpZGe3s7sVjMaKS2YcMGwyA7efIkr7zyCnfeeSepqals3bqVYDDI5s2bGRwc5I477qC/v5++vj727dtHU1PTksbOZ2ZmYrfbycvTvL4OhwO/38+xY8dob2+nrq6Onp4e+vr6qKysZGBgYElDbh577DFOnz7NW9/6VlJSUrDb7SQnJ3PVVVcRi8U4efIkPp/vkkJrdCNpcqhIOBymu7ubzs5OkpKSsNvt9PX1LaqhNNlzYrFYiIuLY+vWrSQmJpKamorb7SY+Pp7S0lIj8dtms01pPqgbhMPDw/j9fjo7O+nv7+fgwYOMjIwwPDxMb28vQ0NDeL3eOQuNxMREPB4PN998M263Ftbq9/sZGRmhqqqKkZERoxP9hc5Tr6Rls9lISUnhwQcfNHq1TEfvC1NcXIzf7+cNb3gDtbW1HDx4cEUYr06nE5vNhsPhIC8vj/e85z0UFhZSVlZGf38/7e3tHD9+nLa2Np566imjspPVaiUxMZE77riD9773vaxevdow0qWUZGRkkJubS3t7+4JXQ5tcNtntdpOeno7T6TSepUAgwPj4OI2NjXR0dBAKhc57T/ViBS6Xi507d3LVVVfhcrnw+Xw8/fTT1NfXc/ToUQYGBpb9Xr1WmckIm1j+nBBiP1p40060GWA9fGojWhjKd6Zt85IQ4k9oHo5bgcdmcfzvnWf5r4QQX0QLX1mP5vnQ+cjE78/rRuDENoeFEL9Fm6V+E/DTWRx/xgotUsoOIcRBtBnxYrTQKv21Exfb72yZLBZmSTKQgCZOpmwrpawWQtSiXa84YNZfuEKIrcA/AF8GTs1xTItB8cTvRycvlFJGhRB/RhMbabPZ0QXu8agQ4imgAu05myw2FuwZUyguxLzFhl5fvrKyEpgaq7x//36ef/55rrnmGvLz86mrqyMWi7F9+3bGxsZYu3atYQjW1dXxxBNPcOONN2Kz2UhNTUVKSXp6OuFwmHXr1tHV1UVTUxNNTU10dHQsaa8DPXEzOTkZ0Axvr9fLX/7yFxoaGjhw4IBRaWqpiUajHDx4kMrKSvbs2WOECzmdTlatWkVzczNOp5NwODxvsaEbXXpuA2AkUvt8PkZHR40Qn76+voU8vXPGMbnyls1mw+PxsG7dOpKSkkhPTzfC9lJTU0lISDAEkm6g6qVrpZSMjIzQ3d3Niy++SEtLC4899hjBYJDy8nJGR0cZHx8nFArN2Yh1u91kZmayZcsW0tLSEELQ19dHXV0dfX19tLS0GM/v+Z5h3TjVq7zl5uZy0003kZqaOmNelF41KxqNYrPZ2LBB82y//PLLsw79Wiz08blcLlJSUli7di1vetObjByb7OxsioqKSEhIoKqqipdeeomhoSFAy0vKyspi69at7N69e8rzF4vF8Hg8pKSkLEq+kNlsxuFw4PF4SE9PJz4+HpvNZoSFhkIhxsbGaGlpYWBg4ILPiclkwmq14nQ6KSws5Morr8Rms+Hz+aiqquLYsWNGOJhiWdAv/OTcAT0ev1lKOZOKbJz4fSOzEBtzPb4Qwo4mfnzACzNs8wSaIXgDl2AITuQLbEcz2OcqCBaTXrSclnIhRJmUsk5/YSLfogw4IaUcmO0OhRBONA/BCeCfgd2z2MwuhHgnWsiZF02gPC+lXKjZjdNogvUONG+NPlYTcBtaqNdfL+UAQggX2nMCWo6OvnxJnjGFAhagqd9Ms3l6aMrkkCJdXNhsNoLBoGHEdnd3c+zYMR588EFycnK4++67yc7OprCw0PiSzsrKIjk52RAq//M//0NPT8+lDn1W6KE47e3t9PT08Oyzz9LW1sbTTz/N+Pi4MVO91IyMjOD3+0lLSyM+Pt5YrhuqWVlZRmPCQCAw57AuPb589erVZGZmTjHo9KZ+999/PzfccAPf+c53DANxIZmeo6Abrk6nk3Xr1lFSUsK73/1uhBCMjo6SlJREcnIyHo/HiKuHV59Hv99vJM1XV1dTU1NDVVWVIQCCwSDV1dVGWNVcjXQhBAUFBZSVlbFlyxbS09NJTEwkEAgwMDDAvn37cLvd1NTU0N/fz+jo6DnPjp4jU1ZWRnFxMe9+97spLy8nLy/PCAWbji4E9QplTqfzgiFXi8nkccTFxREXF8euXbvIy8vj7W9/O2lpaSQnJxueG93rccUVV5CZmckf/vAH2traEEKwfft2vve975Genj4lRE/PcTh48CBnz55d0Pefnm+zevVqbr75ZlatWsXq1auNMEo9X2lsbIyhoSG6u7svONGgf87p3pHMzEzS09OxWq1G+Jfu6VJejaVHCFGAJhh8aKFROnqCeYEQQsyQWKvPSF9SxSshxHZgDdCBlvisUwqYgcbzJFDrxnf5HI+3FbgT7bs/Fy1nwwN8XErZf6FtlxIppRRC/C1aiNlRIcTDQCeQg5a0fhotBGgu/DPafdsspYzMcpIiEy2/YzJNQoj3SSmfm+PxZ+IbaPfjK0KI69FC2mxoORuZwAenh/5dDCFEKfBOtOcnA03IZANfk1JO9uYsyjOmUMzEonYQHxkZYWRkBNBCIvr7+xkfHzeMVd1IHBwc5LnnniMuLo7MzEyKiopwOp04HI4pRmNycrIR1rTYycj6jLAeetPb20tNTQ0vvPCCkUS8nKVQ9WT86bOh+uy/0+nE5XJNybWYC7po0cOCJnsVQBMCGRkZxr3Q8zYWCt1o1Y01fVliYiLx8fHk5OSQk5NDZmYmUkrMZjMpKSlGoz6z2WwYcPp1GhkZwev1Mjg4SHt7O11dXXR1ddHX12dUo7rUxGo9F8Rutxs9FfRnuby8nNraWkZGRoycpZnEhj7jr4cc6iFh0+/B5KR1vceIXmJ5OTwaujdAD5lKTU0lLS2N8vJy48fpdJ6zjS5MIpGI0cU9LS2N1atXU1JSYpQmnvxebGpqorW1dcFDj/TnPi0tjVWrVnHFFVdQUFCA3W43wkRDoRD9/f1GEQH9M+5C+3M4HKSkpJCQkGBcA/18VnpJ5dcqEzO7vwLswGellMaMiZSydiJUpxytUs93J223HXjDxL9Jl3D8JF41ZD85bbY8YeL3+R4ufXniHA+7FfjipP/HgPdJKacb1MuOlPL3QohO4H+Ad096qQdtpr1xxg1nQAhxI9p9/JyU8swsN/sp2oz/abTrVAx8DPgQ8IQQYoeU8uRsxzATUspeIcRVwH+hiSjdAyGB/wSemcduS5l6j0NouSHfmrbeYj1jCsU5LKrYmIwQwij3aDKZ6OrqorKykurqakDzkIyNjfHjH/8Ym82G2+0mKSmJnJwcw8gKBAJGIq/JZFpUj4IeKnH48GFeeuklDhw4YIRyBYPBFW0gTA85mg8Wi4X4+HjWrFnDmjVrpuQ+6KFlJ0+e5MyZM7S2tjI8PLxgRp8+u+92u3G73aSkpACawNm1axdZWVnYbDbS0tKM2fGkpCTMZvOU2X9dZOheqX379jE4OEhPTw+BQIBgMEhTU5ORCH6p91RKydGjR2lqamL16tVs3ryZ3bt3G/klFRUVRgWjhoYG9u7dy9DQEFJKwxuQnJxMTk4O27Zt48orryQ1NfW8pYWDwSB+v9+o5BQIBBgeHqaqqorm5uYlnS23Wq2kpKRwzTXXkJGRQVFREatXrzaqg13I26KLpXA4TEZGBldffTU//OEPjYpv4XCYQCBAY2MjZ8+e5cknn6Suro6urq4FDT0ymUzGvcrMzGT16tWkp6cTFxeH3+9neHiY06dPU1tbyy9+8Qv6+voYGhq6YMNSm81Geno6q1evZuvWrUZDye7ubqPAgi46lWdj6ZhIwP1vYBfwW+BfZljtw2iJ5d8RQtyFFn6jV0M6A1yBVl1pPsd3o8XplwHfkFL+bq67mPg9p4dGSvlvwL8JIRxAEfA3aInuu6SUfzPHMSwqE+FL/4lWvesrQAtaUvk/oFVNuhZ4yyz2k4gmHF7mXIP7vExO2J6gCvgbofWZ+BTwJTSBMG+EVtHpUcCJVvr2AFo53jdMjPUNE6KmaQ7jflLbtbCihX/dj1au9lqhVZeabS3+eT1jCsVMLJnYsFgspKSkGJWB9GZ/k0uzxmIxIxxHzwEYGBgwjE99ncWIbXY6nWzYsIHS0lLsdrvRpEuP7dfzRfx+/7KETc2HS6nf73Q68Xg8ZGdnGzkZusGr71ePWw+FQvP2bOhhK2az2chlsFqtRtWelJQU8vPzjRn7rKws8vPz8Xg8RmK6xWIxSvPqHq9YLIbP52NsbIzW1lba29sZGBhgYGCA5uZmwwswnyTwC6E3BNR/9H2aTCbi4+PJzc0lIyODgYEB3G63Ea6l52cUFxezatUqI3RK92hMRvdajI2NGSWix8bG8Pl8DA4OGt6apfRuCCHweDysX7+e7OxsSkpKyMrKIikpibi4OOP9qxMOh/H7/YYg1O/72rVrDdHhcGiVMYPBID09PVRXV3Py5ElqampoaWlZ8P41ujcqNTWV1NRUkpKScLlcWCwWhoeHGRkZ4ezZszQ2NlJfX288++dDr2S3du1aioqKjD484XCYoaEhOjo6jH0oobF0TAiNX6JV+fkdWpWoc26AlHK/0HpNfB7NsL0WaAO+CpwE/oSWWzDX47uBP6PlDPyrlPJ/z7CaPqucMMNroIU+TV5vTkwkA1cD/2vCw/NhIcQzUso/zGd/C81EXsZ/oeVIvGtS3sxZIcS70MoWv1kIcZ2Ucv9FdvevQCpw8wLlWvwbmti4ZgH29TO0RPcNk0KcRoF/nxCE30bzUrx3rjueqJLWAPxfIUQIrbrXg7wqrBf1GVMoJrMkYsNkMuFyuVi/fj3r1683wkouhO7pGB8fx+l04nQ6jYTdhZiFnk5eXh4//elPSU9Px+PxGDOtOsPDw8Ys9OWE7o2YCyaTiezsbCoqKozeAjMJFz1JPBwOz0uA6SJSN+o6OzsJBoNG1+/169eTlZVFaWmpkYS/Zs0a8vPzKS8vnzFETA+3iUQiHDt2jJqaGo4dO0YkEsFut+Pz+eju7mZ8fByfz7eozeB00aN7LhITE0lISKC7u5uhoSEjXFBKSVpaGuvWrWP79u1cf/31JCcn43Q6z8nTmOwFeP755/nZz35m9HPRz3toaOiC1ZEWA5PJRHl5OX/zN39j5GHo4m966JceCvXII4/Q3d1Nd3c3H/nIR1i1ahWf/vSnDQ+DTkdHBz/5yU84cuQIJ0+eNM51Ie+b2WwmIyODjIwMtm7dylVXXUVhYaHRY6a1tZXDhw/zwgsv0NTUxMjIyAUFttlsJisri/Xr1/PhD3+YnJwccnNzGR4epq6ujmeffZa6ujpaW1sZHR1d0Z7S1xJC67HwazSh8Wvg3RcyQCcMwHNmz4UQ+qz34TkePx5NaFyN5tGYSWgA1KN5TYqFEJYZYurLJn7XzuX45+EJNC/OdcCKEBtoOQtW4LnpCfpSypgQ4nm0vhJb0PpTXIjNaJ6Ds+fx9t8vtBLEJ6WUG2cxNl1gui+41kWYeBauBQan5VLo6I0mt1zKcSZ4Ak1sXMerYmMpnzHF65wlERt2u93IIdA7jc82xEdKSTgcRkx0eV6M0BA9L8ButxszyXpYS0FBAdu2bePgwYNGOdTLQXBYLBaKioq49957OXHiBCdOnKCnp2dWOQlms5n8/HxSUlLo7e01cjemG/Z6/Pp8r4deDjQ9PZ2UlBQyMzMxm80UFhaSmJhIWVmZkZSbmppKeno6GRkZhndMT4zWjXrdW6HH0r/00kt0dXXR0dFBNBrF5XLR3d2Nz+czvDGLcS8jkQitra3ExcUZZU6dTqeRFJ2dnc2VV16J2+1maGiI5uZmEhIS2LBhA2vWrCE5OdmopKU/73pOxtjYmBEuVVlZaYSB6aF9sVhsScP89JwEPb/C6XRitVqnVC/TvVLBYJBIJGJUcWpoaCA1NZXdu3eTmppqGPaT72kkEjEE7fj4OF6vd1Hum9lsZvXq1WRkZJCXl0dmZqbxjAEMDQ3R2dlp5Ghc7Ph6b5T09HSys7MN8TgwMMDg4CD9/f0MDw8veudzxasIIWxonow3oFUlet95Kk1dbD92tByCGFpDtNlul4AWlnUV8I9Sys+fb10pZVBoJWmvnviZ3uH8tonfl1SpaIKcid8Lm3h3aegxl+cr+6ovn4178yHgyAzLs9BClxrQBEvrLMe2Y+L3rHNGzoOekOoRQthmCG+ayzlejHPu8RI/Y4rXOYsuNvQKQgkJCbjd7ilG12y/ZC+ldOtsmD6Tqi+z2WyUl5eTn5/P/v376e/vp6WlZcGToRcavYpXRUUFFRUVPPXUUyQnJ/PUU0/R2dl5weuub7tq1SqysrI4ePAga9euJTc3d0rXbcDIoZmvsaSHrJSXl5OVlUVhYSGpqamsX78ej8dDRkYGnZ2dPPnkk+Tn55OVlUV2drbRVHFyhaJYLEYgEGBkZISHHnqIAwcOcObMGaOD++R1xsbGFtUYDwaDPPfccwwNDbFr1y4yMjLIzMw0ytkWFhaSl5fH9ddfTzAYZO/evcTFxbFu3To8Ho9RXUwX2pFIhPHxccbGxoz+LjU1NZw6dYrm5mYjDGly2NRSGbB6eNgtt9zCtm3bjNwe/TmRUhqVvtra2hgeHubo0aO0t7dz+PBhPv7xj/PGN75xSpgevBouOTQ0xOjoqNE0b6FDp3Tsdjvbt28nNzeXgoKCKUn5gNHosb29nZGRkYs+PxaLhZKSEoqKisjJyTHynEKhEI2NjXR2dtLa2qpCqJaICYHwEJpx+RPgQxcTGhPhToHJno+JOPgfAYXAD6SUDbM8fhJak7atwBellP93Fpv9CM0I/KoQYnLDtW1oHZ77mOgLMovjXwu8MP2chRAlwP+Z+PfPs9nXEqGXYn2T0Br3GTP/QoiNaL0fJLMwhM93rYUQ16E9D4eklB+c9tpaoEtKOThteQFavghooXjzRko5IISoRut/8Q8TP/pxHGjhewD7ZrM/IcRu4GU5rcmkECINrRIXnHuPF+wZUyguxKKKDd2IX7VqFaWlpUZyrx7qEggEVoThrnsxphs88Gp1oDe96U0UFRXxzW9+02g6tlLR8yBAM/ZSU1PJzMw0uh9frCcAwPHjx+nt7TVm4Cdfm8nhQcC8K+q43W4KCgrIz8+nsLCQiooKEhISSEtLM0KI0tLSuOmmm3C5XDgcDqOnhD77rc/6T84b8Xq99PX1MTY2ZuQFAMZM+WIKDX08w8PDdHd309HRgd1uJyUlxXg/6M+aLrivuOIKoyKT3tVc99QMDw/j8/no6+ujt7eXI0eO0N3dTVVVFT09PUa41HSDdTkMWH3ck4WGlJKGhgYaGho4efIkQ0NDtLa24vF4uPHGGykpKTnn2YpEIkZ+zQsvvEB9fT1Hjhyhq6vrQoe/ZPRwz+zsbKMCm+4ta2lpMcZ/sT4/eqnb9evXs2nTJtxut5FTNDQ0RFVVFY2NjQue4K64IP+GZlj2o5WZ/cIM3vX90+L/rwd+LIR4Bi1XwzOxj0I0o+3Tczj+Q2hCowEwCSG+NMM6j8ipjfR+g5aM/ibguBDiMSAFzQg0Aw9IKUdnefw/AcNCiJcnzsUClKD1eLAA35NSPj15AyHEauBz0/aTJIT42aT/Py1nUTJXCJHK1CT81InfPxGvdib/ZynlWQAp5StCiJ8C7wMOC630bQvatb8HzSvwbSnl6Ysde568GficEOJZoAmtGlUJWhlZB/AXZi4qMFceRHuWPi+EuBmtk7cTzatQgBbq9PVZ7uv7QKYQ4gCalyaKdr1un9jnI2h5MJNZyGdMoTgviy42bDYbpaWlVFRUsGrVKiNGPRwOG2EVy41exeh8YsNsNnPVVVdRVFTEf/zHfzA8PLxiY6z1/ACr1WoYex6Ph6ysLFwu10XD13Qj/ujRo7S1tXHNNdecs51uVE7+mY9xGx8fb/RUqaiooKioCLfbbXgu9IpYk2f6J48TMIzyYDBozHr7fD4GBgYYHx8nEAgsqUGni43R0VF6e3tpb28nKSmJcDhslHGenEdjs9nIz8+fkkiu70PPv+jv76empoauri6OHTtmhCCtNCZ7VyZ7WI4ePcqBAwc4ePCgkWh9ww038IEPfIDs7OwZPRqtra1UVVXxm9/8hvb29kXpEj4ZXSDp3jaXywVgTIq0t7df9Jrr4aE2m434+Hi2bNlCSUmJUYFPCEF3dzfHjx+nqalpxj4rikWjaOJ3KvCFC6y3f9LftWjVga4F0gE/WmL4l4FfzDEESz9+CVPLkk6mGa3iFWD0mng7mgH6frTSrQG0fiBflVIenMPxv4iWB3EVcBeaIdmDZoD+WEr51AzbZALvmbbMNW3Zl3i1J8mFiJthXzC1pO3PgLOT/v8A2rm+F7gFiEdLnn4R+E8p5axD2ObBs2hJ6JvQwqbcwPDEsf8b+O+ZigrMFSnlMxNehM+gPWcfQxMJjWg5Ft+QUg7PcnffQhNim9Culw3t3vx1Ysy/mz7mBX7GFIrzsqhiIz4+nqSkJK699lrWrl1LMBhkbGyMpqYmDhw4wB//+Efa2tqWtaGVXkUnLS0Nt9s9xSMQjUYNI8Rut5OZmcmPfvQjXnnlFb761a8uWkjHfNErMFkslikJ+B6Ph5KSEtLT02lpablgQqpegjMnJ4dVq1axadMmcnJypgix8fFxBgYGOHToECdOnLhg6c8LsX79et74xjcangq9CdzknhL6c6EnoeuhOAkJCUQiERoaGoxwroyMDFJSUozqTHofg+VAz7GY3LBtpmd8shdK/1/3gAgh8Pv9dHd3s3//ftra2jh79uycGzQuJvozd/jwYQKBAGvWrDFycPR7uGHDhikdzt/2treRnp5Obm7ulOc0Fovx8ssvc/bsWfbt20dHRwenT582PKCL+Rmh55R4vV5CoZAxrrNnz/L4449z/Pi5fbX0e6c3LC0tLaWwsJB77rmH0tJSysrKsFqtRCIRgsGgkRze19dneNxUCNXSIKW8bh7b1AJvXKDjF85zuwjw/yZ+LuX43wG+M8dt9vNq+dNLQkrZPNd9TRjGP5v4WXAudH5Sa9i3EE37ZjOOU2idui91P//NuQ0IZ7PdgjxjCsWFWFSxYbPZ8Hg8lJeXk5OTY8zUHj58mMrKSk6dOrXsnXP1nJL4+HgjsRUwKv7oRrbFYsHpdLJt2zacTif//M//vGLEhh4XHwgEiEajRglP3Vi32+0kJSXh8XjO27NBRxch8fHxJCcnk5qaSnx8/JQwl/HxcXp7e2ltbaW+vn7enoPExESKi4sZHx8nGAwaCfrTK2jpnpRQKERbWxtNTU1kZWURCAQ4ePCgEfdeWFjI6tWrGR0dNbxmy/Vs6d67QCCA3+838ipmYvr56oIDMJoRnj592ijfu5IMVCkloVCIhoYG4uLiaGhowGw2G+WSTSYTqamprFq1isbGRpKSkti1a9eUZp06sViM06dPc+TIEfbu3XvR0rKLcR7BYHBKdbW2tjb2799Pb2/vOY1ETSYTdrsdu92Oy+WitLSU0tJSbrjhBtLS0ow8E13EdHd3G+W+Fzuc73XMF4UQXwRqpJSX1N1boVCcy0SfkUuqxKV4/bGoYsNisRgx0MnJyTQ3N9PT00NHRwd9fX0rIoTAZDIZITyTE1sHBgY4efIkg4ODBINB7rnnHqM8a319/YoxFGKxGKOjo3zjG98gNzeXBx98kNTUVON1IYQRrpScnExcXByDg4MXvPa6sNB7o+ghJbq3p6enh6NHj1JbW0tPT8+872NXVxdHjhxh586dUxJydQPTZrMZeRbDw8MMDg4yOjpKOBzm4MGDjI6OUllZyejoKENDQ0YpXT1fY7kbpfX19fGf//mfbN26lb6+Pq655hpKS0unrHMh4SelNJrJDQwMMDo6uqKEhk40GqWrq4uRkRGampq44oor2LlzJ7t27aKoqMioylReXm6E+E033PVE+IqKCsLhMA899NCShr/FYjF6e3uJi4ujr68Pi8VCQkICCQkJ5Ofn09zcPCUx3Gw2Ex8fb1RtS0tLY8eOHaxfv564uDjjcyQWi+H1ehkYGKCmpoaOjg4GBgaW/dl8DdKMFt6kM5vQHoVCMXf+iVcraYEWXqZQXJBFFRtZWVmUlJQYnYP7+/vp6OhgeHjYqBK03AghSExMJDExcYrh5/P5qK+vJxQKGYmi+vpOp5Ps7GyGhoaMJoTLhd4cTU9eHR0dJT4+HimlEcaiJ7nrHpwLGbh6snxqaiqJiYlTPA16A7b+/n7a29vxer2XJLpGR0eNpF+bzWZ4k3w+n1FtKhKJ4Pf76evro62tjd7eXkZHR+np6WFoaIienh7Gx8cZHh5e9Kplc0XvXh4fH09ZWRmFhYUkJycbycJ6LP+Fmi/qYmu+vUyWilAoRDgcNmbt9fdIYmIiaWlp2Gw2nE7nlG0mJ/mHQiH8fr/R2X2pjfFYLMbw8DD9/f10d3cTFxdHLBbD5XJRUFBAYWGhkYyvl/tNTk5m7dq1JCQkkJKSQmlpqVEyV99nJBJhZGSEgYEBowHjSshTe60xEabzpWUehkLxmkdK+U/LPQbF5ceiiQ2TycTHP/5xbr31VkBrzPXjH//YmNnr7Z1z49VFwWw2k5eXR1FR0RSjb2RkhDNnznDfffdRUVFhVArScx/i4uLYt28f3/nOnMJgF5yhoSGGh4eN8KgTJ04QjUaNRnyThYXeqK+pqWmKUT7ZqHO5XCQlJXH33XezZcsWbDabIbZGRkZoaGhg7969/PWvWsXBjIwMOjo65mVAdXR0cPLkSdatW4fD4aCvr4/R0VFeeeUVEhMTuemmmxgaGqKyspLnnnuO/fv3k52djcvloqWlBZ/PN2W2eSXOFEejUdrb23nyySd5/PHHCYVCXHvttZSVlXH//fcbPTVm23dmJaOHjjU1NdHb28vIyAiVlZW85S1vITs726hqNn39cDjM6dOnqaqq4uGHH+bs2bOMjY0t6diDwSDPP/88Ho+HpqYmbrjhBnJycigvL6esrIxt27ZRXV1tVNtyuVzk5uayceNGbDabEbqoe230xPLe3l5+9atf0dbWxsmTJ2lra1vS81IoFAqFYrlZFLGhz6jrYTt9fX3GbHhHRwd+v39WzeWWCj3JE7TeER0dHRw/fpzW1lbOnj1rJI7rtfLdbjdlZWXU1NQY1YOWy9DVDW29wd7o6Og5vS8mJ7La7XZKS0sZGBiYscKPw+EgPj6evLw8UlNTDeNQD2U6ffo0AwMDRCIRI8Z9vudutVpxOBwMDg7S3d1NW1sbAwMDNDc3G8Zcd3c3Z86coaGhgd7eXqSUOJ1ORkZGjOPr57gSxQZohmxfXx8+nw+/309raysOh8PI5dA9Ha8FwQFMybNwuVxGsvtk9Hvl9/vp7e2lqqqK06dP09jYyMDAwJKHKUopjcR7PcxzfHyc+Ph4nE4npaWlUzyGNpuNxMREkpOTp5TNjsViRnlck8nE4OCg8bmnN5RUKBQKheL1xKJ5NvRwI7PZTGtrKydPnqS2tpaBgYFLMlAXEyEEbW1t3H///XR2dtLT08PJkyfJysriwx/+MFdccQXr16/H6XRSXl7O5s2bjUZdyx3iolfTGRwcNErz6gaQHkalV6l6y1veQn9/Pz/84Q/x+/1T7kVSUhIFBQWsWrXKqEIVjUYJBAJUV1fz29/+lmg0SnJyshESN997mZaWRklJCR0dHXR0dNDZ2cnIyAhVVVXEYjEOHz5shEyNjo7i8/lob283zne6oFqJzxRoBnVnZ6fxf319PUIIBgcHiYuLm9IITxfqlzO6xyI/P5+rr76a1NRUo2DB5HVisRh1dXX87ne/4+WXXzY+H5YrFE4P4evq6qKzs5P29naKi4uNIhfl5eVT1p9+Pvokyi9/+Us6OjrYs2cPPT09nDhxwtinQqFQKBSvNxZFbGzevJnt27eTk5MzJaRAD5lYSUahxWJh06ZNrFmzhkgkwujoKN3d3UaX6eHhYaLRKMeOHcPv97Nq1Sqj90ZeXh4f+tCHOHToEEeOHFnWylp6f4KOjg6ysrKIRCJGVSO9lOqOHTvIz8+nq6vrnMaE+roFBQXGLK7u7YFXmwNu2bKFuro6amtrL7n7cX9/P42NjUaJ2N7eXrxer1FVS++T4fP5DAP0fCFTK+mZmonJ4xsfH6e9vZ3KykrC4TDFxcU4nc5zmicKISgsLDQqpl2sIeNKYd26ddx2223s2bOH3Nxc7Hb7jB3CQ6EQHR0dVFZW0trayvj4+LKfny7a9YprehPSyWOfXJltcs8Zr9fL4OAgra2tNDU1Aa/mJS11WJhCoVAoFCuFRREbN9xwAx//+MdJTk4GMMqy6omfKwmr1cr27dvJzs42EjkHBgaMUJCxsTG8Xi+PP/44jY2N3HvvvUbSdF5eHp/97Gf54Q9/yKlTp5bVY6PPrJ45c4akpCT8fr+RyKrPmm/atIl169bx2GOP0dzcPGV7vZJTaWkpGzZsIDEx0cjX0MnKyuLuu+/mpz/9Kb29vZdclrSrq4sXX3zREBNerxfQ8kBisRj9/f1GcvRKzsuYKwMDA3i9Xp5//nlaW1vZs2cP6enpRm8H/ZqbTCYKCwtJTU0lNTWVrq6uy6I3w/XXX88Xv/hFrFarkSw93VsTi8Xw+/1GT42VcF56tbVAIEBfXx8NDQ2UlJRM8RLq6+m/9STwaDTKwMAALS0t1NbWcuLECeO8VlLRAoVCoVAolppFERs2mw2Hw4GUkkgkYsQrr7QqLLt27WLNmjW43RcuGR2LxRgYGKC2tpZnnnmGiooK1q9fD2i9IvLz8ykpKaG5uXlZq2xJKQ1xpPetgHMNvZlCdRwOBx6Ph9LSUq644gpsNpuRr2EymXA6nUaeTSgUWpBZaJ/PZwiKaDRqiBc9Zl/PPVnuXiwLjX6u9fX1BAIBnE4na9asISMjwwipmozNZuOzn/0s1dXV7N27l/b2dmPmfCURHx/Pxo0bWbVqlSFyzxcSZjabcbvduN3uFVNGGl4NAWtububFF180cjaSkpKm9AaJRqP4fD66u7sZHR1lcHCQEydOGHleo6Ojl+z5UygUCoXitcCCiw29jr4eghOJRIwE8ZVkVAghWLt2Ldu3b8dqtU6ZrZyJ8fFxpJS88MILjI6OUl5ejt1ux+l0kpmZSV5eHt3d3csuNvS4cb/fj8vlmlICVz+3yU3jdOx2O4mJiRQWFpKTk2Mk9U5OiLVarUb4i9/vv+T7qTes08euj09f9lpFN2jb2toYHx837sVVV111jtAALdTvpptuYt26dXR3d2OxWGhpaVlx7yePx8PGjRspKSmZ0hV9JnRv2/RyuCuBaDRKX18fL774ItnZ2eTl5VFcXEx8fLyxTjAYZGBggEOHDhm9g6qqqmhubmZwcFB5MxQKhUKhmGBBxUZRURF33nknu3btwmKxGPHPx48f57nnnjMqB60EpJR0d3dTW1vL4OAgHo/HaBx3vpn0YDDIoUOHCIfDXH311WRkZJCWlobf78fr9S57iJheCeeFF17gM5/5DJs2beKWW26htLSUpKQkYz1dfEz+SUtLY/369WRnZxMXF2f0OdCFWDgcprOzk2eeeWbBZtWXs4rXSiAQCNDf389LL72E3+9nzZo1RoPJyfdGN8rT09O5+eabycnJweFwcPbsWZqbm5f9GjqdTm688UZWrVrFO97xDrKysoALNyzUe6pkZGRw6623cvToUbq7u5dqyBclFAoxNDTEo48+yrFjx0hISMDpdBoCPBKJGB3Eh4eH6e7upquri+Hh4RXnwVUoFAqFYjlZMLEhhCAlJYU1a9YYJVP1cJGOjg56enoW6lALRn9/Pz09PXR3dxMMBo0mcecz3vQ8gv7+fsbHx0lMTCQajWKxWHC73TgcDiwWy7IZG1JKgsEgPT09PPXUU8RiMUpKSsjMzJzStFCPpdfDXPRZ6dTUVCNURG9epu83HA4zPDxMU1PTOcnlirmjN7LTSwo3NzdTW1tLRkYG+fn5U0KQ9NLFdrudvLw8o3xub28vLS0tyyo2HA4HCQkJlJeXs27dOgoKCi7qrdDzHHw+H1JKUlJSjJC/lYKe9N3a2kpbW9uMjRfj4+MpKipifHyc0dFRI3RKoVAoFArFqyyI2DCZTLjdbsNQio+Px2w209/fT1dXl5H4u9I4fvw4p0+f5sCBA1itVsLhMOPj4+dNVjWZTCQnJ5Oenk52djZ2ux2fz8fatWv55Cc/ye9+9ztOnTrFqVOnlqWPiC429Nlwv9/PyMjIFI+LyWQiJyeHwcFBEhMTAS0JPj8/nyuuuILm5maGh4e58sorDQMwGAzS1NTE4cOHefHFF1dcSNzlyujoKKDdt/b2dqOpXHl5+ZQGcfqPxWIhKyuLWCxGe3s7bW1t54TILSVms5l7772XsrIy3vWud5GUlERCQsJ5PRp6ArbP56Ojo4M//elPHDt2jOeee47h4eGlHfwc0Mc9ndHRUaqrq4lGo4Y4USgUCoVCMZUF82xMbv6m/3R1dVFVVbViyz7quQ1jY2OGwXax0J5gMIjX66Wjo8NIoh4bG2NkZGRFlPWdnHuij1VPVNVFSHx8PGlpaaSmphpVoPx+P0NDQ7jd7inrghbyMjQ0RH9/PyMjIysqHO5yZrJxGggEaGlpoaGhgebmZlJTU/F4PEavGl1U2Gw2bDYbTqdzSmnipcZsNuN0OikqKmLdunVGLw0952SmJn56Zab+/n5aWlo4c+YM9fX1DA0NLXsI4nzQixgoFAqFQqE4PwsiNnTDtr+/n4aGBmw2G1JKfvjDH/LHP/5xRXULn4nZzkhGIhGamppoa2tj3759xnJdoAQCAUNorQQGBwdpbGxkaGiIUChk9DvIzs7GbDZz5ZVXUl9fz+joKAcPHuTo0aN861vfoqKiwih7q89EHz9+nJaWlnM8JYqFYXx8nKqqKlpbW3nooYe4/fbb2bJlC3fccYfhgYJXk/tnSvJfSlJSUsjIyOCuu+6ivLycuLi481af0j0DIyMj9Pf38z//8z/U1dXx5JNPGl3UFQqFQqFQvDZZMM9GNBplcHCQuro6hoeHaWxspLa2dsWGUM0XvRrT0NDQOa9drKLVUqEfPxqNEolECAaDhMPhKbkaDoeD1NRUQqGQ0Xl8ZGRkSslSPa9gbGyM9vZ2+vv7jcRxxcITjUbxer1IKWloaMBsNlNcXExubi4ZGRkrqrN4KBQiEAjgcrlwOBznNO3T0YV4NBqlrq6Ouro6ampqaGpqWhFFFRQKhUKhUCwuC+bZCIfDtLS08Pvf/55wOIzf718xM/yLwUqPz55cUWtsbIyxsTHsdjsWiwWr1Yrb7aawsJDMzEw2b95MdXU1VVVVRs8Rvb/F6OgonZ2dHD58mObm5hV/3pc7kUgEr9fLgQMHOHbsGG1tbRQVFfHJT37SCHFbCej5Ji6X67wJ4ZOFRigU4sc//jHHjh2jrq6OUCikPBoKhUKhULwOWNDSt3rIjR5KpAzT5WVsbIyuri6am5uNilR6tSmbzcauXbsYGxujr6+PWCyG0+kkOzsbt9uNyWQiEolQU1PDkSNHaG9vX7G5N68lJnexllLS1taG3W5nZGQEs9mM3W5HSklnZyfBYJDExERGR0eXxXCPRCJUVVVhsVjIzc01emvoIsPn8+Hz+YyS0oODg0a3eOXRUCgUCoXi9cGiiA3F8iOlZHx8nIaGBpqamkhLS6OkpMSYhbbb7ZSVleH1eg1PR15eHpmZmTidTkNsPP/885w5c4bu7m5V1nMJ0A11PfSto6MDq9VKX18fdrud+Ph4Q4T4/X4SExMJBALLIjai0SgHDhzA7/dz7733GsnhemWmoaEho9ldR0cHvb29BINBNRGhUCgUCsXriAXvIK5YOehJ+8FgECklVVVVeDweVq9ebSSA2+12CgsLCYVChEIhEhISDKHh8/mora2lpqaGYDCoDMQlRA+DGxoaIhaL8etf/5rs7Gy2b9/OiRMnqKysZGhoaNl6O+gemJaWFpxOJzfffDMjIyMcOnSI/v5+ent7aW9vp6enh97eXsbGxmhrayMQCKjnSKFQKBSK1xFKbLyG0cvaBgIB/H4/Z86cwe12U1ZWZqyjJ4Trs9EWi8WoQhUMBuns7KS3t/e8XdUVi4eUEr/fTzQa5YUXXiAnJ4dgMEhLSwsdHR34/X6CweCyhSTFYjE6OzuJj49neHiY8fFx/vznP9PV1UVNTQ0DAwMMDw8rcaFQKBQKxesYJTZew+i9DZ544gleeeUV7rzzTtasWXOOaNArCU2uQuX3+xkeHiYSiWA2m7FarUSjURVrv8Tooq+trY3BwUG8Xi99fX0MDg4aSfzLIQJ1z8bJkyeNClORSMTwXui5GUpoKBQKhULx+kaJjdcwevx/X18fAwMDU/qd6I37dKaXLQ2Hw1PyAEwmkzIcl4lYLIbP50NKSUdHB16vd0VUetPzgsbHx+nr65vSzVw9KwqFQqFQKACECo1RKBQKhUKhUCgUi8HytSBWKBQKhUKhUCgUr2mU2FAoFAqFQqFQKBSLghIbCoVCoVAoFAqFYlFQYkOhUCgUCoVCoVAsCkpsKBQKhUKhUCgUikVBiQ2FQqFQKBQKhUKxKCixoVAoFAqFQqFQKBYFJTYUCoVCoVAoFArFoqDEhkKhUCgUCoVCoVgUlNhQKBQKhUKhUCgUi4ISGwqFQqFQKBQKhWJRUGJDoVAoFAqFQqFQLApKbCgUCoVCoVAoFIpFwXKhF4UQcrEHUFRUxIYNG/jud79LUlISTqeTmpoaDhw4wIMPPkgoFCIWiy32MBQrDJvNxrZt29i+fTsPPvggSUlJ2O129u7dy/79+/n+979POBxGykV/RBWKCyKEwGQyEY1GF/U4UkqxqAdQKBQKhWIRuKDYWGzMZjO5ubkUFhbicrmwWLThjI2NMTo6it1uJxKJKLHxOsNkMpGcnEx5eTlbt24lPj4es9lMKBSira2N7u5u9UwoVgRCCDIzMykqKqKlpQWv18vo6Kh6PhUKhUKhmGDZwqjMZjNOp5OysjI2b96M2+3GarUipaSvr4/+/n4cDochQBSvH3QRumnTJq699lpDbIyMjFBTU0NlZSXRaFR5NRTLhu7NMJvNVFRUcNddd7Fx40ays7Ox2WyYTCpCVaFQKBQKWGLPhtVqJT4+nj179rB69Wre+ta34nQ6sdvtxGIx6urqePzxx3n00UdpbGxkYGBg0UMTFCuPaDRKU1MTJ06coKCggBtvvJGRkRF+97vf8fLLL9Pa2qqEhmLJEUJgsViIj4+nrKyM/Px88vLyuOaaa9i5cyc33XQTp0+f5uc//zmVlZX09/cv95AVCoVCoVh2llRs2O120tPTKSkpYdOmTRQXFxONRolEIgwODnL27FleeuklqqqqGB4eVgbl6xQhBHa7HYvFghBamHo4HGZgYACfz6cEqGJZcDgcOBwOMjIyKC8vp7CwkPXr17N69WoSExMBTShnZmZSX1+PEEJ9hikUCoXidc+Sio3U1FR27drF1q1bKS0txWKxEA6HGRsb48c//jHHjx9n7969BINB9SX9OkUIgdPp5Prrr2fPnj3s2rULu92O3W4nJyeHkpISfD4f4+Pjyz1UxQJjMpkMcRmLxVbcZ0BmZiZZWVmsXbuWXbt2sXbtWsrLy42wqfj4eNLS0sjMzMTpdCqxoVAoFAoFSyw24uLijITwtLQ0w7DQw2aampoIhULqC/o1xOTYdSnlRe+tEAKbzUZeXh7Z2dm4XC5AMz4DgQDhcFh5NhaIlWQMCyHIysqiuLiYq6++mpaWFn7729+uiNwci8WC0+nkfe97H2vWrDFER0JCAna7ndHRUTo6OkhISADg3e9+N5mZmdTW1jIwMMCZM2eor69XSeMKhUKheF2y5GIjPz+f9PR04uPjkVISiUTw+Xy0trbS19envpBfQ+gJtDabDdBE5cXEgi42MjIySEhIMIoGRCIRxsfH8fv9SmzMEj0czeFwEAgEiMViWCwWQ9A7HA5DAAYCAcOo1ycBltLI18XGunXrePOb38yJEyfYv38/fX19hEKhJRvHdMxmM3FxcaSkpLBnzx7DI2uxWDCbzUQiEbq6unj22WfZsGED2dnZVFRUIISgpKSEs2fPEgwGaWhoWLZzUCgUCoViOVlSsbF+/Xre8Y53YLVaicVijI6O8stf/pI//vGPvPzyywSDwaUczmsOm81GLBYjEoks91AMQzc+Pp4777yT+Ph4urq6OHLkCG1tbYTD4Rm3M5vN2O12PB4PVqsVgJGRERobG3nuueeorq5meHh4Cc/k8kQIQVJSEu985zv54Ac/yE9/+lOCwSAVFRU888wztLa2cs8992C1WolEInz/+99nYGAAIQRxcXEIIRgaGlrS8RYUFLBu3TrKyspYvXo1b3jDG3jHO97Bk08+uWTj0NFFcmlpKdu3b+eWW24hOzsbn8/HiRMnCAQCDAwM8IMf/IChoSGCwSDbtm1jx44dPPjgg5SVlVFYWMjAwABut3vJx69QKBQKxUphUcSGEAKz2YzD4SAYDBKNRtm2bRsVFRVYrVZMJhOxWIxgMMjQ0BAdHR2qed8lYLPZcDqdbNiwgaGhIc6ePUskElnW8BOTyYTNZiMuLo6ysjKysrIoLy8nJSWF+vp6Dh48SCAQmOKlEEKQlpZGSUkJpaWlxMXFAdDX10d9fT1tbW2Mj4+fV6isFPRzv/LKK4lEIrz00ktLei8sFgsej4d3vvOd3HLLLeTl5XHTTTcRDocpKCjA4/EwODjItm3bMJvNhkBtb2+nvr6erKwsHA4HJpOJ48ePc+bMmSVpoKhfN93Qt1gsFBUVkZOTQ2dn55J7WsxmMzk5ORQXF7Nhwwb6+/vp7+/n2WefJRAIMDY2RkNDA+Fw2CjRreedmM1mpJREo1H1ubYECCGuA56dtKhGSrl6eUajUCh0hBBbgcOTFrVIKQuXaTiKZWJRxIbFYsFut5ORkcHQ0BB+v5/777+f3bt3GyEaAKFQiFAotOyG8eVOfHw82dnZvOlNb6KpqYn29nbGxsaW1cNhtVpxuVykp6ezZcsWiouL8Xg8XHnlldTU1FBXV0dfXx9er9fYRgjBqlWr2Lx5M6tXryY+Ph6AqqoqTp06RUdHhxEOtJKx2WwkJCTwwQ9+EL/fz6FDh5b0+Xa73RQVFfH5z38ej8eDzWbj+uuvN8a2YcMGgCnXcfv27fT19fHII49QXl5OcnIy2dnZ/Pu//zudnZ1LVoZ6ciiXyWRi7dq19Pf388gjjyzp54Re5jYvL48NGzZQUlLCT37yE1588UUeeughQqEQ0WiUaDSK2WzGbDaTlZVFZmYmFovFEBo+n095bJeW54D9wEXrDgshfgK8f+LfMill/XnWiwP+DngTUAJIoBU4APytlHJBZj+EEFcDnwB2AsnAIFAJfFtK+ZdZbO8G7gHuADYDeUAMqAH+B/ielDI0bZsc4D7gdqACyALGgWPAj6SUDy3Aqc001lSgCsgADkgpd89h21zgPcBGYBNQDAgucA8ntrsW+AywA4gD2oFHgK9IKYfncx7T9v+ziXFdiL9KKW+cxb7mfF+EEInAA7x6XcoBM3CzlPKZ2Z7HbBBC3At8HO05cwBNaM/YN6SUgWmrdwJfnvj7Ews5DsXlw4KJjclx3ldeeSUVFRW88Y1v5Omnn+bQoUNcffXV5OfnT6k2MzY2Zsxmt7W1rXgjcqUghMBqtZKSksK73/1udu/ezapVqzh16hQWi4UHH3yQRx55hMbGRsbGxi66P7PZjBCCSCSCEGJK4vB8jDuTyUR+fj5r1qxhw4YNrFq1ipSUFMxmMy6XC7PZPON2sViMmpoaPB4PcXFxWCwWIpEIhw4d4tixY5eF0ABYtWoVe/bsISUlhZaWliU//oMPPsi9995rNEPU0Q34yfdUv98AycnJvO1tb8NisWAymbBYLDzwwAPceuut3HXXXfT09CzamGOxGNXV1cTFxfH2t78du91uiM/Ozk4j9HKxBY8uMtatW8cVV1zBAw88QEZGBl6vl/j4eFJSUnA4HEQiESOXJBaLEQqFGBsbIxgMYrPZ6O/vp62tjf3793P69OnL4rl9jbBfSvmli60khLgLTWiMoxme51uvEHgaKAVeAH6EZtgWoomPTwKXLDaEEJ8HvoImkh4HuoBUNKPxOuCiYgO4Gvglmkh5Fs2QTgbuAv4FuE8IceM0Y/DjwP9GMxafBbqBAjRD9yYhxP+TUn7yEk9vJv4dmG984Vbgq2iirwkYARIvtIEQ4oGJY0aAh4A2NEP5k8CdQohdUspLbYzzCNB8ntfehSaKnpjlvuZzXwqBb0z83Y72LGXM8nizRgjxFeDzaO+dPwIDwG40QbFHCHGzlNKvry+l7AS+NLHtexd6PIrLgwUVGxaLBZfLxYYNG9i8eTNXXHEFY2NjhMNhMjIycDqd52yTmZlJRUUFNTU1+P1+Y0ZwNpWLXo/oBmN8fDw5OTns3LmTTZs2kZmZSXd3N5FIhEAgYISrzQaXy4Xdbsfn8+FwOHC73YyPjxMMBvH5fHMan8lkwm63s3r1akpLS1m/fr2Rf6GXOe7t7SUcDs9ogAUCAbxe7xQjWD/ny8Vgs1gsOBwOQqHQks9qCyHIy8sjNzeXSCRCLBbDZDIxMjJiJIjrpVr116WUJCQkGM+Vvh+TyURCQgKxWMxI8l8spJSMjIzQ3d1tjEkIQWpqKjk5OTgcjiWpRGY2m0lOTmbNmjVs3LiRgoIChBD09fXR19fH+Pg4UkrjWdRDpqxWq1Gi2Ww2G+Kjv7/f+FxTrAyEEGnAfwK/BTKBa8+znhV4GM3Ie4OU8tFpr5vRPAeXOp43owmNZ4D7pJRj0163znJX3cA7gd9P9mAIIeLRvD07gb8FvjVpm1eA66SUz007ZgVwCPg7IcSvpJRH53RSF0AI8W40o/mjwA/nsYsjwDXASSnlqBBiP+e5hxPHywS+C0SB3VLKVya99hk0A/1fgPfOYywGUspH0ATH9OMnAp8FQsDPZrm7+dyXFuAm4LiUcnCWnpY5IYTYBPwfYBjYIqVsnFgu0K7xx9BE0pcW8riKy58FExs2m42srCx2797Ngw8+SFFREVarlTvuuIMbb7zRMGZ09LCDt7zlLdx+++20tLTQ3d1NOBymt7fXKHOqmIrJZMLhcLB161bWr1/PVVddRXx8vBFyMj4+zte+9jXq6+unhChdaH8bN24kPz+fpqYmVq1axdq1azl+/Dj19fW8/PLLcxqfy+UiNTWVd7zjHaxatYq8vDxcLheRSITa2lp+85vf8NRTT9Hd3T2j4ZiWlkZ6ejqjo6OGSLn//vtJTU3lhRdemNNYlotTp07R0NDAxz72sWXpGdPS0kJtbe2UEJ/HH3+clpYWKisr2bNnD9nZ2dTX19Pd3U1fXx8//vGPSUlJwWQyYbVajZyrpqYm9u7di8/nW3TBFwwGGR0dnXK9ioqKkFKSl5dHa2vroiatm0wmPB4Pt9xyC29/+9u58sor8Xg8NDQ08G//9m889thjdHV1TRHgVquVxMRE0tPT2bZtG+vWrTNEncvloqioiJGREVVpb2XxHxO//xZtZvZ8vAstJOVfpgsNACnlJStfIYQJ+DrgA94xXWhMHGdWX4RSyhPAiRmWjwkhvgX8Cs1L8q1Jr80YJiWlrBZC/BYtLOc6YEHEhhAiH80o/Qmzn+WfPrZ2tJn72XI7WqjPHyYLjQm+hRZa9Q4hxCellIPzGdNFeBfgBH4zW+/JfO6LlHII2HfJo70w96J59n6sC42JY0shxN+jvac+IoT4ykK8PxSvHS5ZbDidTjweD+95z3soLS1l3bp1xMfH4/V6GRgYICkpyahuMxndaAatM+8Xv/hFenp66O3tpaqqip6eHk6ePHlOXP9KQZ/J9Hq9S2JMmkwmVq9ezdq1a7n66qvx+XwkJSXhdruN5FQ9sVYPQZnO5FjypKQkcnJyuOWWW/B6vcRiMd71rndRVlZGdnY2DQ0NRnjVXM7P7XaTnZ1NaWkp6enpuFwuTCYT0WiU9vZ2uru76e3tnWJ46bPt8fHxRtNHt9ttlGTVjfeV1BfiQkSjUbxeLzU1NUuSWD0ZKSXPPvusEZaoJyzX1dXR399PX18fzz77LImJifT19TEyMoLf7+e73/0ueXl5lJaWUlxcTEpKCjk5OUua4KwnquteTd27NbnZ32IipcTtdrNjxw4KCwtxOp2EQiG6uro4dOgQAwMDUzxVVquVnTt3UlFRwfr169m5cyfp6ekIIWhpaeGFF16gsrKS3t7ey+K5fT0wEcZxD3CvlHLgIs/VOyZ+/2winOo2tHCdVuBJKeXAAgxpJ1AE/AEYEkLcAawDAsArUsqXFuAY8Gqo11wS+eazzXmZmP3+GVrY0yfRQryWgsyJ343TX5BSxoQQzcA2NG/JI4tw/Acmfv/HBdeaPQt6X+bIha7lmBCiH0gH1jOD8FW8frlksWG320lJSeGNb3wjRUVFJCYmMjw8TG9vL6+88gq7du0iOfnczxS9NKq+j2uvvRa/38/AwAD5+fnU1NQQjUY5duwYPp9vRX1ZWywW4uPj8Xg8BIPBKYnYizFOPURtw4YN7Ny5k3e84x1UV1cTDoeNsA09tEMIgcfjMQTI5H04nU5isRjhcJicnBy2bNnC/fffz/79++np6eG2224jKSkJh8OBw+EgFothNpvn1FjN7XaTlZVFenq60fRM75PR2tpKd3c3IyMjhiEphMDtdhsiZdOmTWzfvh2bzWaEXZ06dYr6+vrLRmzEYjFisRiVlZXLMpt98OBBDh48OGWZ7lWMxWIcOXIEmNrU70c/+hGFhYXcfvvthMNhysrKyMzMNJKglwI992Gm522p7rvT6WTjxo2kpqZitVoZHx+no6ODyspK/H7/lPtps9nYvHkzV199Ndu2bSMhIQGLxUI4HKaqqorDhw9TX1+/LN4txbkIIQqA7wC/nAh5uRjb0Iz+24CvMfX70iuEeFBK+V+XOKxtE7970JJ/108b8/PAm6SUfZd4HD0RflZ1pIUQHuCNaHkRey/x2DqfQJuN3zMR/rRUYkP3JhRNf2HCs1Q48e+CVy8TQuxAu6e1UspnL7b+LPa3GPdlLlzoWsaj5RmBdi1PLNGYFJcBlyw2wuEw4+Pj9PT0GI36fvnLX3LkyBFeeeUVvvzlL5OdnW0IC32byQmWOlarlaysLG6++Wauu+463vrWt/L1r3+d3/72t4yNjRlG3HJ+cTudTu644w7Wr1/PunXr+OY3v0lPTw8Wi4WxsTECgcCC94FwOBwkJiZyyy23sG7dOgA2btxoJPLq3oq9e/fS0dHBzp07aWxsZHR01DCOrFYrd999t+H1eOCBB8jNzSUtLY1Vq1YhpTSEQTQapaenh5GREfLz8+nu7p517kZFRQXbtm0jMTHROJYev3727Fl6enoQQlBYWIjL5SIuLo7rrruOnJwcw3OTlJSE1Wqlvb2d559/nieeeOKyLCBQX3/ewihLzkzXbvL7yGq1sm3bNj7zmc+QnJxsJOd3d3dTW1u7JMn54+PjdHd3Mzo6itPpxOl0GjlIS1EaOy4uDo/Hw9jYmNH40Gq1kpSURFpaGr29vQSDQUMg5+Xl8eCDD5KcnGzko3V2dvLVr36VQ4cO0dbWpoTGCmHCqPw5WlLrg7NY3w540OL8vznx8/2J7d+AFgr0YyFEs5Tyr5cwtPSJ33+Dlgx8E/AyWp7It4BbgN+jGenzQgjxMeBWNOPvouJowgPxY7Tk4h9KKavne+xJ+1wD/BPwbwtdGWkWPIXmBbhHCLFVSnlk0mufANIm/k5ahGN/aOL3f17qjhbjvsyDx4H/D/igEOKHUsrmSa99FS3EChbnWiouYy5ZbESjUaPBVWJiIlJK6uvrjVCooaEhIpEINpvNCIUYHh5maGiItra2KcmWhYWFFBYWYjabMZlMpKamUl5ezubNm2lpacHr9eL1eg3hsVQ4nU6ysrKoqKggIyODjRs3sm7dOnJzc3nzm99MV1cXvb29HDlyhO7u7gWZgZ88E22320lMTKSgoICUlBRsNhtWqxUhBH6/H4vFYvytX1e/348QAofDwbp161i9ejVXXnklDocDq9VKfn4+Ho8HKaWxPbxanSgtLY20tDTq6+vndK17enpobm6mu7ub9PR0PB6PEQbj9Xqx2WykpqZy11134Xa7sdvtbN++nZSUFNLS0ow8DSklw8PDnD59moGBAfx+/8UPvsJYCc0VZ4PZbOb2229nx44dJCUlYbfbiUQinDhxgpdeeon6+vol6eKtl4p95pln2LlzJ2vWrMHr9TI8PIzf71/066k3k9Q9cvp7IiEhgSuvvJLu7m5DbOgFEJKTk3E4HIZ3MRgMcuzYMTo7O41kcsWK4O/QkojvmIhtvxjmSb//KKX87KTXfiq0crjfRUuGvRSxoR9HoHkwTk78f3qivGgtcK0QYsd8QqqEEPcB30ZLHn/jLPM/vgW8Ga361iVXoppIcP9vtApbn73I6guOlLJFCPEFNLFzQAjxEFrOx0Y0cXcKuAJNWC4YQogE4C3MLTH8QizofZkPUsqDQoh/Bz4MnBJC/BGt+tkuNC/daWAtC3wtFZc/lyw29Fnr1tZWotEobreb06dP09DQQCgUwufzEQqFcDqdxpd3VVUVBw4c4E9/+pMxgwjw5je/mY9+9KPGl7fD4WDHjh0A1NTU0NXVRV1d3ZIZPzqpqanceeedvOtd76K4uBgpJU6nE4fDwd/93d8xMjLCwYMH6evrM2buL9XIsFgsmM1m/H4/LpeL3NxccnJySE1NxeVyEQ6HCYVCdHd3G8vcbjcDAwP8+te/JhKJYDabSUxM5AMf+AB33XUX8Gpeh151aHx83DiWXlnHYrGwZcsWfD4fBw8enFMYzbFjx6ipqaGwsJDrr7+erVu3GvktwWCQpKQk0tPT+dSnPoXL5UIIYZTD1ccAmuHZ2trK888/bwhWxcKjhzN+7nOfo6ioyDCyx8bGePDBB2lqamJgYCFC0y9ONBplZGSEj3zkI/zt3/4t3/72txkYGKChoYHR0dFFf8+7XC7S0tLIz8+fkmuUnp7Ovffey+joqPEc3nHHHeTl5QFTyweHw2HOnDlDMBi87Dxxr1WEEGXAPwI/lbPoVwEgpfQJIUKADa0i1XQeRhMbV17i8HTh0zhJaOhj8AshngI+MHGcOYkNIcQ9wG+AXuD6yQm9F9jmm2jC7Hk0YbYQ5fT+P7QSvtdLKccXYH9zRkr5NSHEGTRPxu1o9/U08HZgA5rY6F3gw74TcDGHxPDzsUj3ZV5IKf9GCPEKmtfmLROLj6J54T6AJjYW+loqLnMuSWzoFZC2b9/OzTffTF9fHz//+c+pr68nIyODb3/722zYsIG4uDij/OapU6f4yU9+wrFjx+jo6JhS4vaxxx4zcjb08J6MjAzuv/9+AoEAfr+fkZERPvShD1FbWzvnsqxzwWaz4XK5cLlcrFmzhh07dpCZmWkYIUIIotEop06d4vTp0+zbt4/u7m7DWLtUwbFnzx5KSkqQUrJ582a2b99OVlaW4SGKRCKMj49TV1eHzWYzwjp0QTI2NobH4+HNb34z27ZtIzk5eUovlKqqKmpqanjsscdYs2YNRUVFFBUVUVBQQEZGBtu2bSMUCvHnP/+ZwcFBAoHpfXrOj5QSv99vhL6YzWYSEhL4yle+YvTySE1Nxe/309/fT0JCAg6HA6fTicViIRQK8dhjj/Hkk0/S0NBwWQsNt9vN2rVrqa+vZ3BwMQqdXBpbtmzh6quvprCw0KhqVlVVxZEjR85JiF5KfD4f4+PjFBcXY7PZSExMJBQKLWqFusnPoP4edrlc5OTkkJKSMiWXxGaz4fV6sVqtWCwWI+zMarVyww03cOTIEZUYvnJYC9iB9wkh3needeomPh/vnZTPUYMWbz88w/q6SHDO8NpcqJn4PdMx5n2ciXK6v0bzaNwgpaybxTb/D80Yfxa4U0q5UF+wm9E8N/vPk5C/SwghgREpZeICHfMcpJR/Av40fbkQ4iMTfx6e/toloieG//ul7GQR78u8mchVOickTwjx44k/F/paKi5zLtmz4XK5SEhIID4+nv7+fiorKxkbGyM5OZkrr7yShIQErFatTLjf76eqqorjx49TV1d3zox5XV0do6OjVFRUkJqaSl9fH/fddx+JiYlG7H9SUhLFxcX09vYuitgwmUwkJSWRkpJCbm4uSUlJFBUVUVpaaoimQCBg5EqcOHGCU6dOcfLkScbGxi7ZuNCTwUtLS9m+fTtWq5UrrriCwsJCw9iKxWJ4vV7Gx7VJIl2wJSYmkpWVRX5+Pj09PaSnp7Np0yZSU1OniJRQKERVVRUvv/wyf/rTnwiHw4TDYdLT040E3cTERHJzc4mLi2NkZGTW49fHMjIywvj4uCEULBYLBQUFRr5ONBplbGzMyAdJSEggMzPT6P9w8OBBzpw5MyXv5HIjMzOT1NRUSktL6enpWZFiIzs7m9WrVxtGth4GefLkSXw+35Ilh08nEokQDAZJSEggPT2djRs3cuzYMYLB4KIJDj0ZfnL1KyGEEbY4mYGBAUZHR8nI0Hpmmc1m4znVu7YrVgzNaKVWZ+IOtAo7vwdGmdqUbR+a2FgH/Hnadusm7ftSeB4tn6BMCGGT0zp8z+c4Qoh3AL8AOpiFR2MiF+D7aH0vnkbrKbKQcatPM3NH9zjgrWjJ8Y+jlf9dUoQQq9Ea0jUxR8/RRfa7Hc1jUiul3D/PfSz2fVlQhBB70HKNnpNSdiz3eBQri0sSG3plm9OnT7Nq1SpGRkaMGv3zYWRkhNHRUaPsqtPpNBKj7XY7ra2tHD58mM2bN+PxePjFL36xoIaoxWIhISGBT3ziE1xzzTVs2LDBaICml+D0+/184QtfMBrWPfXUU1RXV9PY2Gg0SbsUA83hcJCdnc2NN97Irl27cDqdRnjRyy+/zOjoKAkJCXR3d+N2u7n11luNClO5ublGWdzW1laSkpK4/fbbcbvdxnUaHh7m7NmzPPTQQ5w6dQqAxMREysrK2L17t9HZXe8TYLfbZ90cEDDK1b7wwgvEx8dTXl5OcnKyUZL39OnTHDt2jGg0Sl9fH52dnfh8PsrKyvjgBz+Ix+MhFArx0ksv0dDQcNn2WhFC8H//7/+lsLCQmpoaGhsbl6Wb+IXQQ+b0WfxwOIzP5+Pw4cO8/PLLK6Y3REJCAj/4wQ/4j//4D371q1/R0NCwKCKoubmZ1NTUKR5APT9s+iTCD3/4Q/bt28fXv/51SkpKyMjIMMLAnn/+eQYHB5VXY4Uw0X/igzO9JrSGcJnA30spp1d0+He0JmV/J7QGau0T2zjQwrJAC1O6lLH1C61vwv3AF9A6M+tjuxktNGWE2VeReg/ajHMLmtC44IfOhEH7H2jX5wm0poKzd2PPAinlD85z7EI0sVEvpZzx/iwUQgiPlHJ02rJ0NO+PCfjfUsqF/LDTE8PnVe52Ke7LfDnPtSxBG28U+NyyDEyxolmQBPFQKGRUYKqoqDCM1fl82erVkPTKU3rMfiAQoLu7m/b2duPvhf4yv+eee4ywkry8PBwOhyE0pJTU1tZy6tQpXn75ZRoaGjh27BgnTpxgcHDwkqvOmEwmrrzySioqKrjttttYtWoVJpPJuI66R6K5uZkzZ85QXFxMeXk5JpPJ+IlGo6SkpHDTTTcZXbj1Cll6Cdv6+nr+8pe/cPbsWQYHB42cGV1MgSa6jh8/zoEDB4wKPHMhFovR2dnJyZMnefjhhykqKiI5OZm0tDSefPJJjh07BmAk+2/evJns7GyjZ4jNZuP222/n0KFDPP/88yuu9PFsEEIYIW319fVzEmxLhS42HA6HIThisRjBYHBFJTcLIbBarVit1kW9jrFYjLa2Nv7rv/6LsrIyUlNTefHFF43nf/L16OjoMESGy+VCSsnAwADNzc2MjIxcNL9ED7+y2+1GXptiZSGlPCuE+N9oibknhRCPAF40AVCOVjXq6wtwqE8C24H/I4S4Bq17dAFaA7Uo8ICUcvhiOxFCXI8mNExoITfvmyFsaVhK+e1J/38BzaD1o1Wr+twM25yYZangJUNo3bF19JK1XxdC6E0RfyylfHHSOl8QQtyK5r3oA3KBu4EE4AtSyt8v4Ng8aCIqhFYBbT7M674IIf6FV8vP7p74/RkhxDsn/n5kAe7lT4RWRvooWphfKXAXYAU+KKU8dIn7V7wGWZAO4lJKfD4fcXFxlJSU0NzcbCT/6ugNuyYnhF9sn5FIhKeffpqXXnqJjo4OQqHQosTv64b2DTfcwD333IPT6TRCJ/SmbFJK9u7daxjLC9msTU/M3r59O9dffz133333lBAWveJXIBCgvb2dX//613zwgx+krKwM0ISK7k2Ki4sjISHBCDs7efKkIdz8fj8vvvgijz76KK2trUYexuTZa31fzz//PIcPH57X7LaUkr6+Pg4dOkRnZ6dRtaekpISnnnrK6AGhJ47ffPPNFBUVGeLOarWyZ88eRkZGOHr0KH6/f8UYvnMhLi4Op9N5Sd6+xUIP17Pb7Uauhv5+DYVCl+01vxSklPT09PDP//zPbN++nby8PH7zm98YYndyftmePXvYtm0bqamp2O12YrEY3d3dNDU1XbDR5+RcED1HREqpxMYKRUr5r0KIGuBTwJvQcj8a0YzBf1mIsBYpZe9E2M3n0QTGVcAYWujW1+ZgvBWgCQ14ta/GdFrQqlPp6P0SnGiJ3DPxcxan2d2l8J4Zlt036e/9wGSx8Sxa7sgb0BozDqFVEft/UsoXFnhs9wNuLi0xfL735U1oz8Fk9kz6u3mGbebK47yaHB6Plgz+R+AbUspTl7hvxWuUSxYbCQkJZGRksHbtWlatWkVubi733XcfoVDIqNcfi8XYv38/f/3rX/npT3/KwMDArMMgent76evrm1NjubmOPzk5mauuuor169cTHx9vNLTz+/186Utf4vTp07S0tDAwMMD4+PiCCg3dELVYLKSlpRkNEO12uyEW9u3bx5NPPkljY6PhqZh8/XQDRi8Zq48tEAgY4VJnz541OluPjY2dV7Tp1apKS0uNMr7zSXaPRCL09/czPDxMTU2NcY7j4+PG2O12O263m5KSEjIzMzGbzUafkoceeogjR44wNDS0IkJ55kosFuPd734311xzDV/5yld4/vnneeGFhf5Omz8mk4m0tDR2797Nrbfeis1mM56J3t5e2tralnmEy0M0GmVwcJBnn30Wu91OVlaWUVJ68rN7+vRpAAYHB0lLSyMuLs4QcMnJyYyOjp4TAmi1WomPjyc/P5/du3djtVr53e9+d1kXQLjckVJeN4t1/sy5ORsLPY5BNA/HvEuaSil/xhxLrEop3wu8d77HvBSk1qPhgi3cL7DtnLZbins46Vg/An50ift4L/O4L1LKwks57iyP8XPm77FRvE5ZsDCqnp4eI/QoJyeHWCxmxIPHYjH6+vro7+83hMNc9r8YxMXFkZaWxurVq0lPTzeS0nVxNDg4SGNjIydOnKCystIoaQsL181YTHT7drvdeDweioqKSEpKMsRMOBzm+PHjVFZWcuLECUZHR4mLi2P9+vVUVFRQVFTEdNeqLgz03ic1NTVUV1fP2BTP7XazevVq1qxZQ0lJieFB0Q2lybPd80HvCK0bXZOvmxCC+Ph4CgsLyc/PNxpCer1euru7aWhooKOjY9kSlBeC9vZ2urq6jNyjhISEOSXbLyZOp9NoEqk33AwEApw+fZr29vbXtQEci8Xw+XwEg0Fyc3ONnjaTsVgsRlK9Xiyivb2d9vb2806MWK1W4uLiyM/PJzc3F5PJRDAYvKyf8RXIF4UQXwRqpJQL3hFaoVDMDSHEVqZWp1pZyYuKJeGSxYbX6yUYDPLcc88RCoUoLS0lMTHREBpyomnfcnf+nk5+fj633nort956KykpKUZittVqZXx8nH379vHlL3+Zvr4+I9xoocdvMpmMcrN5eXlGczufz0csFmNkZIQ//elPVFZWMjQ0ZIz7gQce4I1vfCNpaWkzCiAhBJ2dnRw9epSnn376vCEaOTk5/N3f/R3XX3896enpeL1ezp49y7PPPguwYGVPz2d4FRYWsmvXLiPMSkpJa2srBw8eZP/+/UvevHGxkFKSkZHB6tWrOXLkyIowLlNTU/nCF75AUtKrjV7b2tr42Mc+RlNT0zKObGWge/RmyhExmUwUFBRQUlJCQkICFouFaDTKX/7yF1544QVGRkZmfOadTifp6emUlpZSUlKC3+83Pj8Vl0wz8OVJ/19SXwOFQrFgdDL1vTm8TONQLCOXLDb03IqXX36ZUChEfHw8d911F6mpqVPWOXbsmBHKs5yiw+Fw8KlPfYqKigrWrFlDdna2kSvQ2dlJe3s7f/jDHzh9+jRdXV0EAoFFM3jj4uL49Kc/TVFREWlpafT29jI2NkZhYSHNzc1GlavR0VcLP0QiESOfQ+9hEYlECIfDHDx40Oi2vX//fmpqaozmYtOvuR62tWHDBuLj4w3RMjg4yMmTJ6mrq1u08DWTyURKSgpbtmzhnnvuITEx0SgpPDw8bCSuvxaEBmjnq4eMrQSSk5PJz8+fMlvv8/lWjNdFn5xYTvRCFc3NzUYZ3lgshtVqJTMzk61bt3LDDTcYIZd6Y9OZClfo4VU33ngjGzdu5O677yYpKQmv18vdd9/NkSNHaGxsXFGTMZcbEyE5X1rmYSgUimlIKTtR783XPQuWIN7V1YXFYiE9PZ2bb77ZCC0IBAKMjIzQ3NxMW1vbsn6hmkwmXC4X119/PQUFBaSlpeFwOACtB0hjYyPPPfccv/3tb43cjMXEYrGwdetW0tPTcTqdVFdX4/f7cbvdVFVVcfLkyXMa6oVCIUZHRwkEAkaDPH3Zvn376OnpwefzsW/fPvr7z53c06sPASQlJZGWlmb0END7Y+gVr/z+xSvpbbFYSElJIS8vz+hJEAqFGBwcZGhoyBBTrwX0Lt16XP9y43Q68Xg8U5aFw+EVM8OuF5zo7+/H5XJhsSzIx9S8xjFdgFksFuLi4igoKKCwsNB4//X399PT02P0vpmO2WwmNzeXNWvWUFhYaLwPS0tLaWtro6mpSYkNhUKhULwmWbBv8UgkQl9fHwcOHMDr9RKLxQgEAvz85z/nW9/6Fu3t7cveMyEtLY3S0lLDk6B3A+/p6eEjH/kIJ06coLOz87xhR3pohZ6HcqnGge5dsNlshEIhDhw4wLFjx4wGgcFgkEgkMuU4AwMD/PWvf2XDhg1Eo1HS0tKora3l0KFDPProo/T39+Pz+WZseCiEoKioiJ07d3L27Fny8vJITk7GZDIZceqDg4MMDAwsasx+LBajvb0dn8+H0+k0urF7vV5OnDjBgQMHLrmU8EoiFouRmJhIRkbGihAblwOPPPIIzzzzDI8++ihr165dMc+CXoBBFw2BQICqqioefvhh2trazvu+0XOUEhMTcTgcxnpZWVlG9TiFQqFQKF6LLJjY0L0Yvb29/PWvf2V4eBifz8exY8fo6elZ0ApO80HvBOxyuYwmefrMpT6zODIyYggivTN6cXExKSkppKWl0dTUxNDQEKmpqbS0tNDe3n5JJUL1mPBwOMzAwAANDQ2cPXuWoaGhGcOI9DGPjo7S0tKCzWbjySefpKWlhba2NsbHxwmFQgQCgRnHZDabSUtLo7i4mN27d1NWVmb0Fmhvb+fgwYOcOHFiSSpASSk5c+YMjz/+OHfddZdRZngxO0QvNbFYjOHhYZqamnC5XJSWlmKxWJb9/Hbs2MH27duJj4+f4jUIBoP09vauiDKs0WjUqP6kl4q99tpricVifO1rX8Pr9S7LuBITE7ntttvIyMjAbDZjtVpxuVzExcVdsLxxNBqlv7/f6FszNjZm9A96PSfjKxQKheK1z4LGJ4TDYfr7+/nTn/7EkSNHiEQiHD16dNkMg8mYTCZsNhtut9swCqLRKG1tbVRWVhpGlt4gLyUlhZKSEq655hoKCgpYt24d+/fvp729ndzcXKM61KVUk9FnSX0+H83NzZw8edLoMn0+sRCLxfB6vbS2tjIwMMCPfvQjwuGwEaoTjUbPa7xYrVZSU1MpLS3lrrvuwul0Gt2+9+7dyyOPPGIYmkuRxPzyyy/T1dXFli1bjMT4cDj8mgmf0hu9vfzyy5SVlbF+/fophROWi82bN3PTTTcZnj3QhNH4+Di9vb3LNq7zYTKZcDqdbNiwgZycHL73ve8ty2eKEIK0tDTe8IY3GJ3G9XK2KSkp5w330otkdHV10dLSwtjYGG1tbXR3dxufIQqFQqFQvFZZlGDol156CZvNZhjSK4WNGzeye/du3G43fr+ftrY2/v3f/50TJ04wPj6O3W4nPT2dT3/60xQXF5OZmUlOTg52ux2bzcaqVasIhUJEo1HKy8txOBw8+uijDA8Pz8s4Hxsb4xvf+Abt7e0899xzdHR0XNAIjUQieL1eAoEADz/8MBaLhdHRUWObC3XaNplMpKamEhcXZywLhUI0NDRw+PBhXnrpJePclsoQ1kvjDgwM0NHRwfe//32qqqro7+9/zQiO9vZ2vva1r/GNb3yDTZs2GWVUl0tsCCHo6uqitbWVtWvXGiGB//iP/2g0W1zJCCGMMs0z5SQtJunp6axZs4b169cbTUt9Ph/19fX89a9/nVLIYTqRSIS9e/dy6NAhfv/73zM2NkYoFMJsNs+5HLhCoVAoFJcTiyI2JucMrIRYa71J3dq1a9m8eTNCCAKBAH19ffT09BCJRLj77rsxm814PB527NhBUlKS0QNDn422Wq1GJ/SioiLD2+H3++clqsLhMNXV1TQ0NMy63Kheqae/vx+TyTQrI0Wv73/dddexdu1acnNzAU1sNDU10d/fb+TZLOX9CofDRp7G+Pg4x48fZ2RkZEpC/FKgd9K22+0MDw8vqNAJh8N0d3dz9uxZEhMTufHGGzl58iS1tbULdoy50tbWxunTp7nhhhuM3iaNjY10dHQs25jmgt1uN4oaLBUmk4mrrrqKjRs34nQ6MZvNSCkNL0VdXd1Fw8/Gx8fxer309fURi8UQQuB0OpVnQ6FQKBSvaRZFbKwEgTGZlJQUHnjgAe68804KCwsZGxtjeHjYqLi0YcMGvve97xkVkGw2m5GwqTe2k1IaIVYWi4Xi4mLi4uJ4+umn5y02otEox48fP28Fmwsxl+OlpKSwevVqvvKVr5CQkGAYan19fTz11FM0NzcDS3/f/H4/nZ2dfPvb3yYcDtPZ2bmkxwft/iYkJJCdnU1+fj7PPvvsonjjfvCDH/DEE0/w8MMP87vf/Y7PfvazC36M2SCl5PHHH+f06dO86U1vorOzk8OHDzMwMLAs47kcMJvN2O12PvGJT7Bu3TrDaxuJRGhoaKCyspLq6upZvX/0Zp06Y2Njizl0hUKhUCiWneWpKbnEhEIh2traGBkZIRKJGKU/k5KS+OxnP0t6evqUTsCgzUL29fVhs9kMY8PlcmG1WjGbzVRXV/PrX/+a06dPXzB84kLoMf1LmZCr9w949tlnOX78OIcPH6a9vX1KONZSoXtS+vr6ljWsKD4+nk2bNrFt2zaOHj26KGLD7/czOjpKJBJh/fr1fOxjH+NXv/qV0axxKYlGo/T19fGd73yHwcFBWlpaqK2tXVEhjyuJWCxGOBzm2LFjxMXFkZSUZISf/fGPf+SVV15ZcRMsCoVCoVCsFF4XYiMajTIwMIDX6yUajRrCIT09nY0bNxrx13pYks/no7e3l9raWqO7eFxcnBFvHwgEOHXqFE888QT9/f3zDoPQQ1iWOj9BSsnZs2epra2lv7+f8fHxJQ9dmjyW5TRyhRA4HA7S09NZu3YtDofjknIqrFar8XwNDg4ayfp6M8aBgQE8Hg/XXnstjz32GMPDw0tuqOrhP48//jher5fR0dFlrxa3kpmc4D25aV8kEqG6upq2trZlHqFCoVAoFCuX14XYCIVCtLS00N/fTzgcxuFwkJyczO7du40yuKB5M3p6evj85z/P6OgoQgje//73k52djRCCpKQkgsEgn/zkJ6mtrTVK5V6KkbaUZS+llFgsFpxOJ+9973tpa2sjNTWVp556allzCJYTKSWhUIhIJGLci/mKDYvFwo4dO9ixYwd33HEHH/7wh6murgYw8mw++tGP8ra3vY23vOUtFBUVMTIywvDw8EKe0qyIRqO0trYCKy/scaWh99YBjPCpQCDA4OAgJSUlDA4O0tXVtcyjVCgUCoViZfK6EBvhcJjm5mb8fr8RKqXnXkxGn+W+/vrrje7ZFRUVOBwOYwZYT6bVuwUvdVL1XNGT3tPT06d0Y9Zr/ff09OD3+5e9HOtyofdaaWxs5MCBAxes6HUxLBaLEY5VWlrKG97wBnJzc3n66aeN69vf3080GsXpdBIXF4fdbl/gM5o9K/l+696E6WNczjH39vZy9uxZCgoK6OjooKmpidraWvr6+pZtTAqFQqFQrHReV2JjbGzM6JY9ebZSx2Kx4PF4eNvb3mbMdCcmJhKJRAzPSFdXF11dXQwODq6I/iEXQgiB2WwmOTl5itjQZ9nr6uqMfiGvV7ERi8WMXhidnZ2XlLtisVjYtm0bV1xxBRkZGbzjHe8gLy+PZ5991hC3kUjEyAFabrGx0tErv+nlmGOx2LKWiG1oaCASieD3+6murqampoaamhpVTUqhUCgUigsgLmRYCSFeU9ZnUVER69ev51Of+hTZ2dmkpqYyODiIxWIhNzfXmEmdPKPq8/kYGhri8OHDVFdX09LSwu9///tLaua3lOjdl1NSUvjsZz9Lf38/lZWVNDQ0MDg4SHd3N5FI5DXT12K+WCwWrFbrebuvzwaz2cyqVat4z3vew0c/+lFMJhM+n4/jx4+TnJyMzWajoaGB1atXU1hYyOc+9zleeuklXnnllQU+m9cO5eXlbNq0iU984hP09PTQ1NTE97//ffr7+xkZGVmycZhMJlwuFzabDafTic/nIxAIXNLzMleklOLiaykUCoVCsbJ4XXg2dFpbWxkcHGTHjh2UlJSQn59Pa2srNpuNQCBAfHw8drudsbExY0bV6/UyMDBgJFQ3NjZeNkIDXk1Cj0ajvPLKK/T09PDKK68wOjp62ZzDUjA5Z2O+RKNRmpqaaGtrY3R0lLS0NDweD+vWrSMuLg6z2Yzb7TZ6K4TDYXUPLkJdXR3j4+Ns2rSJgYEBo6DBUlZwg1ffR16vd8F7sSgUCoVC8VrmdeXZAG2mPyUlBbfbjcfjYXh42ChL+s53vpO1a9eyd+9eOjs7GRgYYPfu3cRiMU6ePElXV9e8emKsFPRGZMpQWjyEENx333387d/+Ldu3b8fpdE6Z+ZZSMjQ0RENDA//wD/9AZWUl3d3dyzjiywO93w3wuhVoyrOhUCgUisuR153YAHA4HJjNZqxW65RKRCUlJWRmZlJXV0cgECAYDJKWloaU0hAlS1k9aqF5veZlLDWrV6/mmmuuISMjA6fTid1un3LtR0dHGRgY4E9/+hP9/f1GMQLFhZlcEer1iBIbCoVCobgceV2KjfOhz56+XmdOFQtDYmIimZmZRuJ9UlLSOevEYjHq6+uXPBxIcfmixIZCoVAoLkcuKDYUCoVCoVAoFAqFYr6YlnsACoVCoVAoFAqF4rWJEhsKhUKhUCgUCoViUVBiQ6FQKBQKhUKhUCwKSmwoFAqFQqFQKBSKRUGJDYVCoVAoFAqFQrEoKLGhUCgUCoVCoVAoFgUlNhQKhUKhUCgUCsWioMSGQqFQKBQKhUKhWBSU2FAoFAqFQqFQKBSLghIbCoVCoVAoFAqFYlFQYkOhUCgUCoVCoVAsCkpsKBQKhUKhUCgUikVBiQ2FQqFQKBQKhUKxKCixoVAoFAqFQqFQKBYFJTYUCoVCoVAoFArFoqDEhkKhUCgUCoVCoVgUlNhQKBQKhUKhUCgUi4ISGwqFQqFQKBQKhWJRsFzoRSGEXKqBKBSKlYfZbEYIgcViIT8/nz179rBhwwauuuoqEhIS2L9/P9/85jeRUtLR0cHQ0NCSj1EIgZTL/1FlNpv57ne/yx133AFAVlYWQgjC4TB2u/2c9dvb2/nqV78KQGNjI6dPn6avr49YLDbj/qWUYvFGr1AoFArF4nBBsbFUmEwmXC4XDoeDSCTC8PDwcg9JoXhdYjKZcDgcmM1mTCYTu3fvxuPx4Ha7WbVqFTt27CAnJ4ekpCQsFgvXXHMNKSkp9Pf38/vf/56//vWv+Hy+RR+n2Wxm586dfPSjH+XPf/4z1dXVNDc3MzAwcN5tlkKU2Gw2nE4nbrfbEGr6sfW/AaSUJCUl8f73v5/6+nrWrl3LVVddxV/+8hcaGhoYGxtb1HEqFAqFQrFULLvYsFgsuN1uiouLycjIoL+/n1OnThEKhZZ7aIpZYLFYsFqtmEwmAoEA0Wh0uYekmCVCCENceDwe4uLicLvdJCYmAuDz+bj22mvJyMggOTmZ0tJScnJycDgcxux7cnIymzdvpqenh+eeew6z2bzo47ZYLBQXF3PTTTdx3333YbFYyMzMpLKykqeeeuqc9fVzcrvdeL1efD7fonlgYrEYQghcLteU8QLniA2Xy8WGDRtITExkbGyMwcFBjh07Rnt7uxIbCoVCoXjNsOxiIycnhy1btvCRj3yEgoIC6urq+PKXv8yxY8eIRCKLfvyVEoJxOSKEIDc3l8LCQjweD4cPH6a/v59wOLzcQ1NcACEEZrMZh8NBQkICO3fu5K1vfSu7d+8mOTkZs9nM+Pg4x48fZ82aNcTFxWG32433SiAQ4PTp00SjUWpqajhw4AAvvfQSbW1ti24kW61WCgsLefzxxyksLMRisfDGN76Ru+66i56eHoqLi6cIXiEEd999Nw888ACrV6+mqqqKffv28fWvf33BxxaLxWhoaKClpYXU1NRzXpdSGoLDZDIhhMDpdFJeXk5TUxOHDx/m6aefVoJdoVAoFK8pll1sBINBTCYTW7duJRAI4Pf7aWhoWPQvXCEEN910E7t27WJoaIhDhw5RX19/wTAMhYbJZKKkpIQbb7yRt7/97QwODlJTU8Mrr7yy3EN7zWC1WrFardjtduLj46fMio+NjREMBuflSXK73WzcuJH169eTlZVFfn4+V111FRkZGbjdbmMW3ul0snnzZux2O4ODg3R3d5OXl0d3dzfV1dX86le/IhaL0dvbS1tbG0NDQwSDwQW9BtO56aab2L17N29729vIzc3FYrEghDBEz8MPP3zONl/5yle4+eabWbNmDTabjW3btuFyuRZFbEgpefrppxFCsGHDBmKxGNFodMqkiclkIi4ubsp2JpOJzMxMbrjhhin3WaFQKBSK1wILKjasVisJCQn4/X5CodBFZ7itVivx8fEkJibicrloamri9OnTjI6OLrq3QQhBfn4+V199tTFjK6WcUWwkJCTgdrtxOp309PQQCASWxOuyUhFCkJKSQnl5OVdccQWDg4NYLBacTidms3nFeDamh63oWCwWLBaLMdZoNLqks8mzMSj1sKbExETy8vKM8CQpJU1NTQwMDBCLxfD7/bM+rsViISEhge3bt3PNNdeQnZ1NSkoK6enpmEwmIpEI4XCYUCiE2Ww27qff7+fs2bO0tLRQU1NDTU0Nzz//PLFYjFAoRCAQOG9S80IhhKCwsJCdO3dSWlpqeAaklIyMjHDs2DGee+45pJRGaGZOTg4333wz5eXluN1uAOx2+znG/kLS2dlJfX09PT09+P1+fD6f4e3JyMggNTXVeH/oHibQcj2SkpJwOBx4vV7lbVUoFArFa4YFFRsej4ddu3ZRV1dHf38/AwMDSClnNESEECQlJbFx40bWrl1LJBLhBz/4AU888cSSGasej4fc3FyKi4uN3IOZZufXrVvH2rVrKSws5Pe//z3Nzc0MDw+/rg0Cm81GfHw8TqeTjIwMpJQUFBTQ399PIBBY7uEZM/Q6sViMWCyG2Ww24vfj4+MZHBxkbGwMn8+3JILDZDJhMl284nRRURH5+fmkp6dTUVFhzOLHYjGOHj1KfX09Z86cIRAIzPo5dLvdlJSU8OY3v5krrrjCqJA0MDDA2NgY4+PjhmcxNTWVHTt24HQ6CQaDtLW18fnPf55wOLwsz73JZCI9Pd2o8KQLjWg0ygsvvMBf/vIX9u/fjxACj8fDunXreNvb3kZFRcWU/InR0VHa2towmUyLJpC6urr46U9/Snd3Nz09PVRXV2MymfjUpz7FnXfeyfj4OEIIrFYrbrfbuJ5CCJKTk4lEIiviPXQ5IYS4Dnh20qIaKeXq5RmNQqFYKQgh3gT8ftKi56SU1y3TcF63LKjYCAQCNDQ08Pd///c0NDTwwgsvcOrUKYaHh6ckfJvNZuLi4njggQe4/fbbKS0tJRQKkZubS0lJCd3d3UsyU+rxeEhKSjKET3p6+jnrmc1m7rrrLu655x6ysrLYunUrzzzzDL/97W9pa2tb9HHOBt34mj6Tr/8sNjabjZSUFK655hoaGxuXLbnVYrHgcDhITU3ltttuIykpCbvdjtfrpbGxkZ6eHvLy8rj++uspLS0lPT2dxsZGDh8+zKlTpzhx4oQxI72Q6InYVqvVmL2eLoam86//+q8UFhZis9lwu92GcR2LxUhJScHtdtPT08Pw8PCsRZLVasXj8eByuaiqqqKqqoonn3wSs9nMyMgIIyMj+P1+hoaGuOmmm7jqqqsM78X4+PhCXIpLIisri6ysLEMoNDY28rWvfQ2bzUZ9fb3xPv6bv/kb9uzZw7Zt26bkmvT19XH27FmOHj26aO/bvr4+hoaGqK6uJhKJGGFUxcXF/PznP+fPf/4zWVlZfPaznyUrKwvAKC3scrkoLCxkbGxMiY358xywH+ifvFAIkQf8f8AWoABIAgaABuC/gF9KKS84yyWEsANHgbVAh5Qyd7aDEkIkAg8AG4FNQDlgBm6WUj5znm1m8+H9binlf892HOc5zgeAKyfGth5wAv8opfz8pex3Yt/vAn4x8e8DUsofn2c9AbwbeB9wxcQYuoHDwOellLULMJb1wOeA7UAOMAjUAv8G/F5KOa8PBSFEKlAFZAAHpJS7Z1hnP3DtBXbjlFJe9E0vhLACd078bAfy0ey4BuBh4JtSynO+gCeu7/uAD6E9v2agBvgp8AMp5YLPtF3K++UC+0wB7gXuQHtWc4AQUIl2Lj+d4T6eAb488fcXL3UMivmxoGIjGo0yPj5OSkoKmZmZ5OTk0N3dbYRVTcZsNnP11VdTWFiIy+Wivr6etrY2+vr6lmzm1Gw2G2EMlZWVHD16dMb1kpOTSU5Oxu12s3btWoLBIK2trXR2dq4IsaEbshaLBbvdbsz4hsPhJSlDqh9/1apVxMXFLUvSvd4Hori4mFWrVnHnnXeSlJSEzWYjGAzS3NzM4OAgCQkJrF27luTkZBITE/F4PCQnJ5OXl0d2djZPPfUUHR0dC2bs6eE/5eXlxMXFkZeXR0JCgmEEn4+ysjISEhIwm81YLBZDaEQiEWw226y8I9MJBoO0t7fz9NNPMzw8bEwI6Ndocg5IKBTCYrEQjUYZHh6mvb2dWCy2rN68hISEKV4KPbwrEonQ3d2N0+nkvvvuY/fu3RQWFuJ0OgFNeEciEY4dO8aRI0eorq5etDHq4mz6593g4CDDw8PYbDYqKioIh8PG/dcTx81mMwUFBTQ0NCxLv5LXCPullF+aYXkJcD/wMvAImqGZAtyGJjbeLYS4WUp5ofjYf0ITKvOhEPjGxN/taGIo4yLbfPk8y+OATwER4Ol5jmcy3wISgCGgE+1aXTITAu97wDjamM+3ngNt5vlONAP418AYkA1cjSbMLklsCCHuAh4CYsCjwB+AVDTD9TfATWhicD78O+Ce5brnu6ezjcsuQTsPL5on789o1/YW4B+Atwohdkkp+6dt93PgXUAv8NuJ7W8CvgNcI4R4s1z4D/dLeb+cjzcDPwK60M6/Fe19dB/wY+C26ecipTwDfAlACKHExjKx4GIjEAjg9XrZvHkzW7duZe/evQwODjI+Pn5O3Pz69etJTk42QiEOHTpEY2Pjkho0+hf+3r17+d3vfjfjOjabzTDwMjMz2bBhA3V1dTz66KMrIj9BCGGEBWVmZhoiY3x8nGAwuGjhQZPDPywWC2VlZSQmJmIymZY0B8JsNuPxeNi+fTtXXHEFV199NRs2bMButxuz4OvWrSMUCuH3+xFC4HA4sNlsZGRkkJSURElJCeXl5fT29uL1eunp6bnk51CvNrRjxw6uv/56EhISWLduHQkJCRcVDHFxcVP6NOh5JYFAgKGhIbxe75zzhrxeL3V1dfzgBz/A7/fj9Xpn9EJlZ2eTnJyMxWIxvEJVVVXLWiXJ4XCQkpKCw+EAMMRXLBajtraWQCBAcnIyH/jAB8jLy8Pj8UxZLxKJ8Ic//IGzZ8/S1ta25OPv7e0FtM+StLS0c66lfn8zMzNxu92LGub1OuUgkDR91nNipngvcB2awTLjl8BEmNbfAR9FM3bmSguacXdcSjkohPgZ8J4LbXAe0YQQ4sMTfz4mpeyex1im8zagWkrZIoR4L9oM8SUxMZP+UzTv0UPApy+w+rfQhMbX0LwYM92jS+Wf0eyd66SUz03a9+eBk8AHhRBfkVK2zmWnQoh3oz03HwV+eLH1z3dP58AY8LfAz6X8/9k77zi7zvLOf9/by9wy7U7voynqzZIl2XLDHTfAoRm8JNmEDZsCCZuEJOCQDQtZdkOS3QAbQiAQDAb3IrCNLcm2rGa1GUkzmt773Lm937N/3DnHMyPJVpkm6f1+PvqM55Zz3nPL+Pm9z/N7HiU0Yx0mMq/zvWR2739/xn0PkhEaXcAWVYhMv65PAB8m81n8wWWuTWMevi/n4wxwP/DizM+JEOJLwEEy1/Ih4Ml5PKdkHphXsZFMJpmYmODLX/4yH/7wh/mN3/gN/u3f/o2//Mu/5Ic//CFerxej0UhOTg6NjY3EYjHS6TSJRIJf/OIX9PX1XRHzNWYaO5cL9913HzfffDO33HILkUiEsbExmpqa+OpXv8rk5OS8v65qaY0akOv1esrLyyktLeXkyZP4fL55Pd/5UA3Ujz76KLfffjulpaVasKYoColEAqMx8/+qqakp/vVf/5WOjg4cDgdr167lAx/4AEVFRdq8iEcffZSSkhJ+/OMf4/V6LyngE0JgNptxuVw88sgj3H333axatQqj0YjNZtP8F++V2QgGg1pAGo1GaW9vp7m5mZaWFt588036+vrw+/0XJQDS6TThcJju7m6Ac4opIQSPP/44dXV1nD59mhdeeIG33nqLpqamJctqmEwmvv71r7Nq1SotC+D3+zEajfze7/0e1113HY8//jivvvoqjY2NWK1WLRukZvgmJyd54oknCIfDS3YdBoMBt9vNDTfcoJXGwbtdrP76r/+ao0ePLpkv5mpGUZRz/gFUFCUhhHiGjNhYca7HCCGcZAKxXyuK8h0hxEUHT4qieIFfX+zzzsPvTP/87nwcTFGUX87HcebwB8CtZF7XW8/3ICFEDfBZMuVSf3Gu3fX3K2+7QKoB/0yhMX3sYSHEAeA+IJ/MTvkFIYQoB/4R+Fdg1zys8X1RFGWAc4gaRVHiQoivkREbN8+5+0PTP//XzIzH9Gf/r4AHyYiTH8zHGufj+3I+FEV57Ty3DwshvgP8LZnrl2JjmTGvYkMtVxgaGqK3t5eBgQHq6+v59Kc/TWNjIwcPHuTw4cMUFBSwYcMGcnNzMRgMJBIJNm7cyJkzZxa93j8ej3PixAmGhoYu6PFqABOJRJY8INDpdJhMJoqKirSd+dzcXMLhMOl0muzs7As2JF8MiqIQCoUYHR0lmUxqwbM6u8FqtS6K2FBLt2688Ubuu+8+PB4PJpOJRCJBT08PU1NTBINBcnJyGBkZ4cyZM7z66qsMDw9jNpvp7u4mJyeH7du3Y7fbMZvNNDY2MjExwbPPPnvJk+zVTk75+flcf/31VFdX43a7Ne8GoNXyn+8z1N7eTiiU2bjy+/0cO3aMM2fO0NHRQV9fH8Fg8JI7op3vnKpIcjqd6PV6enp6OH36NJ2dnQve1va9SCaTPPfcc+Tm5nLTTTdRVFSEzWajtLQUl8uFy+XiU5/6FHfeeac2/Vwt5YtEIhw9epTHH3+cSCRy2dkCNStmtVpxuVysWbOGrq4umpqaZk0LB7Rz6fV6NmzYwA033MD69eupqanB4XBox1QUhVgshs/nuyI2W64mhBB64J7pX0+c52H/SMbj8VuLsqj3QAixEdgIdDM/JVTzjhCikUwm4R8URdkrhDiv2AA+DujIlPk4p8udyshkRF5TFKV9npZ1EtgkhLhBUZQ3Z6zVQ8avMkimtv+CmM7c/ADwAV8Aci7weR8Fqsj4DE6Tucb5+uOqirK5/2MonP7ZeY7nqLdtFEK4FUWZmod1LNX35XzXL1kGzPucDUVR8Pv9jIyM0NPTQzqdpqGhgdLSUoqKikgmk7jdbrZu3YrNZtNqnPPy8jCZTPO9nPcllUpx4MABRkZGLvjx0WiUycnJJRcbaovPqqoqiouLNY+CKgD0ev15u4FdDqrYmJiY0AJmNcAymUxaJmGhcblcNDY2cuONN1JRUaFlybxeL7t376a3t5fJyUmKiooYHBykt7eXY8eOafXy/f39rFmzhtraWgoLCzGZTOTl5VFVVYXNZrvkmQdqVqOyspKGhgby8vK0zk+qyFAzQ+fLTBw4cICxsTEgU/7U0dFBS0sLo6Oj+Hy+9xQql4pOp8PtdmM2m0mn0/T392uibSlLetLpNK+88gqVlZUUFBRQVFSE2WzGZDJpHqHa2lpqa2u15ySTSSKRCD09Pbzwwgt85zvfuex16HQ6HA4HRUVFFBYWah4ht9uNz+dDCDHr89/V1UUikcBqtbJlyxYeeugh1q5dq4k6eLeRQyKRWBYlmVc702be/woIMjvZtwO1ZHwCL5zj8Q+RKTH57YstsVkg1BKqf7lUQ/NCIoQwAD8ikyH40gU85brpny4yJufcGfcp07vifzAPBubPk3l/XxVCPEsmyM4js6s/BXxCUZSL6QzyR2R20O9QFMUvhLggsUHGHzKTUSHE5xRF+cVFnPt8/Ob0z7mZKjWbUXWO51TP+O8GYP/lLGCpvi/Tn7tPT/+6EJk6yWWyIEP91EClqakJv9+Pw+HA6XRyww03aIFYXV0dOp2O0dFRTpw4wY9//GMtuFrOqK1Bd+3ateS7kA6Hg7KyMtatW0dOzrt/6+LxOD6fj76+PiKRyLzX2quCsre3l0gkgsVimffsyfuh1+v5wAc+wH333cfOnTvR6XT09PRw6tQpXn/9dZ588kkmJiY0YaEGdWrQrO56P/fcc2RnZ2O326murkZRlMsukSsvL2fDhg3ceOONVFdXa4ElZEqi/H4/Q0NDNDc3n3e3/X/+z/9Jf3//rNvUrNpCIITAZrNx5513YrPZSCaT+P3+SxocuFA88cQTDA4OcvPNNwPvdmE7F2NjY3z5y1/mqaeempcsm8FgwOl0cu+993LPPfewcuVKSktL+cUvfkF1dTX/5b/8FwA8Ho+WtfjoRz9Ka2sr9fX1Whtep9N51prj8Ther3fJurhdY+QxuyONAnwT+NLcEh4hRAGZUqVdiqL86+It8dwIIbLIZAKSZEzty5Evk+m2dcMFBu9qC8ivAq+S8XZ0k8k2fJdMzf8Y0wbfS0VRlDeEENvIeBR+Y8ZdATLekqYLPZYQYiUZ8/N3ztdJ7Bw8S+ZzdpRM1qaCTFD+x8DPhBAfVBTlkkuxhBD3kxGi/bzbiEDlBTKfmy8IIX6qKMrk9HMMzDasZ1/q+aePt5Tfl68Dq4GXFEX51SKfW3IBLIjYUBSFzs5OgsEgZWVlfOQjH6G4uBi73c7WrVvR6XRaN56xsTGOHj3KwMDAopdquFwuzUh6PtRJzmqAkEgkCIVCS96a0mAwUFxczObNm9m2bRt1dXVkZ2dr9f2HDx/m4MGDVFZWajv880VhYSEf/OAH+Z3f+R3N4D88PMzXvvY1fv3rX19wluhSUYfT3XrrraxZswar1Uo0GmXPnj289tprvPPOO0xMTBCLxbSsy/myAGoGrru7m5KSklkelEulr68Ps9lMcXExbW1tVFZWap+zUChEX18fL730EqdOnWJ0dBSv13vW+kZGRs4K8hc6k2YymSgvL2d4eBiv18vrr79OT0+PVs611ASDQU6cOMF3vvMdduzYQUlJCXl5ebMeo/pSenp6eOWVVwgGg5eUlXE4HJjNZqxWK/feey9WqxW73c6HP/xhSkpKsFgsJBIJ7r//fq2BxEwvVzAY5O/+7u9QFAW3261lY2YKjWg0Sn9/P6lUiq6uLvx+/+W9QJL3RVGUFjJVMHoybTMfIhPo3iCEuFcNxKb5F8DIpXcpmm8+DjiAp+bJGD6vCCG2kMlm/C9FUd6+wKepf2iHgIdmCJTXRGY+whEyQfLXzue7ucC13U4mq3CYzA54C5nyov9Kps7/XiHETe/TjUw1Vf9oer3/7ULPryjK38+5qRX4khBikEzHrq9xib4PIcR2Mpm5EPDhaX/QTH4KPEKm89opIcRzQJhMw4IaoI2MX+lyd5WW5PsihPgDMqKthYwRXrIMWRCxARAOh+nv7+f1119nx44deDwezGbzWbXKiqIQj8c1n8Fiok6Sfi+sVqtm6gW0ErGlnPKr0+lwOp00NDSwadMmGhsbyc/Px2KxaOU1qVSKdDr9vlkHtf3v+UqfFEU5q32uzWbTfCJGo5F0Ok0sFqOzsxOv17vgGR+1fExtD2swGIhEIgwODtLW1sbQ0JAmNNRrUJm7qxyNRgkGg0xMTBCPx7UWwg6H45KzNer1OxwOLZui+plGRkZoamri0KFDtLS0MDU1hd/vP+uztNjfBbUEqKGhgfb2dlpbWzl16hTBYHDJM3gqqVSKiYkJXnrpJRRFYceOHeTm5s4yWquteru7uxkcHCSVSl3S91Qtp6uvr+e+++7DZrNhMpmoqanBYrFoWRX1c6JONFfXaTabaWho0IaFzrxf/UyoGwPd3d20t7cvm9f5WmC6LKcX+AchxAjwOBnR8V9B6zJ0H/DotCl3OaAaw//fkq7iHMwonzpDpgXrhaIGxr+cmwlRFOW4EKKLTEDcSKZr1KWsLYdMu9cwGUGj/g+tk4yQqSJTTvUI72+S/nMymZtbFEWZj+FD3wP+HlgvhHAo55iR8V5MZ2t2kWnpe7eiKGdNJVYUJT2d+fhDMsH4p8j4G/aRya78HzJiY/RSL2Kpvi9CiM+Rad97CrhtzmaBZBmxYGIDMnXTv/zlL/nwhz9MTU3NWXXwqVQKp9NJZWXlopdqzPyf/3s9Jj8/n9zcXM1P0tLSwokTJxgdHV0SsaGasKuqqti2bRsPPvig1qY0nU4TDAapqMi0tg4Gg9ocgrmogbTNZtNmTZyLZDJJS0uL9rtaZjQzENfr9TgcDjZu3MjJkycv2Vh9oeh0Omw2mzYgb6Yx93zvqXrfXAGh0+m0bFU8HsdsNmvm42PHjl1SAOhwOGhsbOTOO++kurpa89EEg0FeffVVXnvtNV5++eVlU56kYrPZ2LFjB//9v/93XnnlFQYGlkuM9S7hcJiXX36ZcDiMEELzQADaxkV3dzfNzc2X7IHQ6XSsWbOG2267jUceeQSXy3XOvxdZWeceHaB+H86HEIJUKoXf7+fgwYP84z/+46J1b5OcE3VH+eYZt22c/vlDIcQPz/GcEvHu0L3seTLWnhchxHpgM5n2pS8v5LkukSwy8zAAouf5O/wvQoh/IWMc/6Pp21qBO8j4Js6FKkasl7G27WRKhF6fITRm8joZsbGJ9xcbG8n4fXaf5xp3TH8ufIqiuN9vYYqiRIUQgen12cmUdV0QQogbyczZSAN3KopyXr/FdMbmf03/m3kMK5lhjhEyJvpLZdG/L0KIPyIj1JrJCI1LFkuShWdBxYbKl7/8ZR5//HHuuecefvd3fxeLxaINxnviiSfYtWtRusZp6PV69u7dS319PZDZxSwsLMTtduP3+7VdZYPBwIMPPkhxcfGsshq1b/9SoJbnbNiwgYaGBrKysmZ1g3I4HESjUQKBAJ2dnWdNb1cRQnDjjTdyww03sG3bNqqrq88ZrIfDYT73uc9pGZKBgQG2b9/O6tWrtbIQtaXnH/3RHzE4OMi+ffvOGUiHQiEikchlB1axWIyBgQH27NmD2WzWhOz9999PcXEx+/btY9euXUxNTZFIJDCbzaxcuZL6+npWrFgxq1tQV1cX+fn5KIpCNBrFZDIRiUS0TlsXi8lk4pZbbuHmm2+moaEBm81GIpEgGAzS1tZGc3Mzp06dWnZCQy1tVGeTLFf0ej0ul4sNGzawatWqWWtNpVI0NTXxve99jz179rzHUd4b9XugCvEL2Zi4WPR6PXl5eTzyyCPodDpee+01du/ePa/nkFwwJdM/Z37h3+b8g+h+i8wu+ePTvy9G/a9qDP/eudrDLgNiZFrAnouNZLIBb5IRFzNLrH5Npu3q6rlPEpkJ1Go74u7LWJtqmss/z/3q7Reys/QKc6bTT5MFfBQYIeORuKBpukKIejJCI3Ce457vebeSGU4YJyM0Dl3oc+fwKcBCZm7H5XSoWNTvixDiT8n4NI4BtytnDzGULDMWRWyMjY2RnZ1NSUkJOp2OeDzOxMQER48e5dSpU7S3z1d3uwunoqJCm8ewbt06Pvaxj7Fx40ZtinkymcRms3HrrbeSnf2ub2oxuy3NRa/Xk5WVRX19PXV1dRQUFGA0Gmd1gxJCEAqF6O3t5fTp0/j9/nOKDUVRsNvtuN1u8vLyZpWjzMTpdPKJT3xCExYTExM0NDRQV1envQ6q0HG5XNx33300NDSclfVRFIXBwUH6+vrYs2fPLFF3saglKEeOHKG4uBi3263VxW/btk17vyYnJ4nFYmRlZbF27VoaGxs1UaUeZ3BwEKvVqjUxUEWHz+e76MyV2vK2srKS2traWYF7NBrljTfe4MyZM8tyOrRa4qiW+CzXgXJGo5G6ujpuuukmamreHXSsdrX72c9+xqFDh7QheheLmt0pKysjJydn1mDFS2Xm93Pm7yaTCY/HQ15eHgUFBeTl5WmzXZZnPHnlIoTYCjTN3dmeNl3/w/SvL6q3K4ryMzKlN+c61m8BXkVRfnuBljv3fHbgEyxjY/h0CdQ5Xw8hxGNkxMYPFUX53py7d5EpZ7pTZCa4z2zn+1dkulTtuUyPyttkXrsdQog7FEXRMkMiM+VcFXLvOwdFUZT/e67bhRCVZMRG+9zPhRCiGojNLS+a7oymDlD86fv5RWY87w7gGTIB/O2Kohy9gOc4FUXxz7ntOjIBe5BMCeEls5jfF5GZDfJV4B0y3cBk6dQVwKKIDZ1Oh8fjYfv27RiNRkKhEAMDA7zxxhscO3aMiYmJxVgG8G6ffJvNpgXLtbW1lJWVEQ6H6erqorW1VSup2bJlC3a7XXu+2lJ2KTAajbhcLmpqati4cSMej0dbi+p/SSaTDA8P09rayvHjx4lGo+cMXNSSk3g8TiwWIxqNnnNHWwjBJz7xCW13V80UzDXOQ+a1+cAHPqB1C5pJOp2mt7eXAwcOcPr0aUKh0CUHtGqr0N27d+NwOHC73WzevBmn04nT6aSoqAi3283Y2BjRaBSn08maNWsoKCjQhqmpr8nKlStnmcLHx8fxer34fL6LWp9Op8NoNOJ0Oqmrq6OkpETLOKkCZteuXZw6dWrBy8wuBZfLRUlJCZFIhFgstmSZu/fCYrGQn5/Pli1b2L59u2YOnzm475/+6Z8uOWukeqFU03dRUdElC42ZfqF0Oq15vtTb1YxgVlYWdrsdj8fD6tWrOX78OKFQSCsBk6Jj3vhz4GYhxB4yXo0wmXkOdwNuMvXr/2MhFyCE+CaZblgAN0z//KIQ4pHp/35GUZRnzvHUjwFOFsgYLoT47RnrUftH3yeEKJ3+7xZFUb4+3+cFbRjdo2RKw3YJIZ4mM239OmAnmU5Uv/Meh7iQcwwKIf6GTOelXUKIF3jXIP4hMjvyTyuK8tLlnOc92Al8b/qz1wFMAuVk5ru4yJjWL8hsPp0JeZZMNuIl4AEhxANzH6ecPan8FSFEhEzJUQBYNX3+GPAhRVHONYNj2TH9WfkqGTP7G8AfnONvdLeiKD9Y5KVJ3odFERsVFRU0NjZqZkq73U59fT179+5d9LrwFStW8KlPfWpWZxhFUbS+/evWrWPt2rXa7uNcA/nBgwc5efJyShsvnZycHDyeTKfATZs2YbPZUBRF8x1EIhG6u7v51re+xTvvvAOcP1hRFIXXX3+dI0eO8L3vfY/8/PxzBlbZ2dl8//vf14SZKmjU10YVKOogPXWC99xzpdNpiouLWblyJQ0NDfT391/2XIG+vj7+4z/+g1//+tfcf//9PPDAA1RUVGCxWNi+fbsmkFSvhrrWme+7xWIhGo0SDofp6+vjhRde4J133mFoaOiixIbdbic/P5+bb76Zm2++mdLSUm3OyeDgIK+//jrDw8PL1gT82c9+loceeohvfetb7Nu3b1lmX77xjW/w8Y9/HJfLNevzePLkSb7//e/zgx/84LLK02aWR37oQx+aldG8GNTvnDqob2hoiFWrVp23GcVHP/pRPvjBDzIxMcGuXbvYs2cPzc3NhEIhRkZGlqXwuwL5FzLdeq4j482wkfEDvEOmHer3L3Rn+TL4CJmWpzO5Y8Z/d5PZsZ7LQhvDbyBjFJ7J2ul/AHvI7IAvCIqivCmE2EymJfEtZMTfCJnr/RtFUfrf4+kXeo6vCiGOk5lWvp3MpO0wmZa3P2JhTffvAD8m4wlZT0Y4BqbP/QTw3YvotFVERmgAfHj637l4bM7vvyAjWh8h438ZJGNO/7qiKN0XeO7lgDorRE9m1sm52MM8TUOXzB+LIjYmJyfx+XxYLBZtvkFvby/j4+OL3u7WZDLhcrnOeZ9aDjT3NhW1jeX5nr9QqJPCt23bRkVFxax2nzqdTish8fv9dHZ20t3djdfrfd9gPplM4vP5iEQi5w0uLRYL3/3ud2cFSjk5OWzYsIG1a9ditWZ8e6lUivb2dm3y9cjIyCyhE4/HGRoaYnBw8JKN13NJp9MEAgGSySSvvvoqPp+P0tJSCgoKeOCBB3A6nbNK3uaKKTXr4PP5aG9v55VXXmHfvn20trbO6mZ1IdjtdoqKiqioqNB8NOoaI5EI4+PjWser5Yg6rLKtre28Pp+lQgihDdBT31M1yxYKhdi/fz9nzpyZl9ax0WiUY8eO8eEPf/iCvSszy6RmZoXUzGJWVhZnzpwhNzdXm84+U6gbjUaysrIwm8088MADbNu2jaNHj/K3f/u3uFwuIpHIWR3hJBeHoigvMqNM6jKPdUnpLkVRKi/xeVsv5XkXcfz/BPynBTz+Y7zPnAxFUU6RKUNaMBRFeZZMVmAhjt1Nxjh+rvuamKfXV1GU3ec7z/s8738C/3M+1nAJ5543w9uFfJYky5NFERvJZJJkMoleryeVSuH1ejl48KAWKC4maptXNdBUg0F1B1wIod0/0zCt9u9f7CBMbU1rNpvxeDyUlJSQk5OjlU0JIbRpyWrLz/HxcUKh0AW9tup7c75gRqfT8e///u+zAvW6ujqSySQNDQ1YLJlNFtVs3dzcTF9fH2fOnJm1y5xMJpmamiIcDs/rpORkMkkgEODEiRMMDw/j8XhYsWIFt956qzZd+r1QFIVAIEBLSwsHDhzg6NGjWt38haLX6ykoKKCmpoa6ujqsVqsWSKbTaUKhEOPj4wQCgWVnDFeJRCKzBjUuJ8+GTqejoaGB/Pz8WeIxkUgwNDTE4cOH6erqmpeSo3g8rmUVUqmUtvkwMxumkk6ntfdZmW676/f78fl8xONxfvnLX6LX61m/fj29vb2sXLmSsrIyHA4H2dnZ2mwO9W+O0WikvLycsrIyCgsLeeaZZ0gmk/T09NDW1nbZ13aN8BUhxFeAVkVRGpZ6MRKJZGkRmXktP1/qdVzrLLjYEEKwYcMGVq5cqfXsf/nll/n85z9PJHIhA0bnD3VHUUVtR/qtb30Lo9GI3W6nsLCQG264Abvdjl6v1yb+BgIBvvKVr/DUU08xPLw485QMBgNWqxWr1YoQgn379pFMJvmd3/mdWcbVUCjE4OAgLS0tNDc34/f7503EpdNpOjtnl3MqisLmzZtnTeNOpVIEAgFef/11Ojo65lVQXAiJRELzQng8nks22V6sIVin05Gdnc3dd9/NBz/4QRobG7XGA6pAbW5u5q233iIejy/LGnwhhCZULyQjttgYjUauv/568vLyZr03U1NTPPPMM/ziF7+Yt9axyWSS7u5uent7ycnJweFwYDAYtPdNFYupVIpwOKz9fUilUoyPj/PMM8/w2muv8dJLL5GVlaWJlUAgQE5ODkVFRaxZs4avfOUrFBcXn3cdRUVFPPnkkwSDQZ5++mm+8IUvzMv1XcV0M3sasuxOI5FIIDODY+bfhu4lWsc1zYKLDZ1OR1lZGStWrMBut2O1Wtm4cSMf//jH+eEPf7ik9cipVIpgMMg777zDwMAA8XicrKwsXnnlFS0w+MM//EP0ej3RaJSXXnqJ8fHxRVuzWo5jNBppbGykoaGBdevWUVpaislkQq/Xk0wmefrppzl+/DhdXV00NTURCAQWNKide+xkMkkoFKKlpeWijdXzhbr7vWbNGs08rHbqmrnmmeZwdVe5oKCAm266CbfbzdGjR2lvb+fNN99keHj4Pa9F9YJkZWVRXFxMcXExNptN82qoXbPU3e5LHTC3GAwPD3Pq1CltncuJeDzOrl27uP3226murgYywzW7uro4ePDgvGZi0uk0Xq+X7373u6xcuZKNGzdy++23A2jtut966y36+/uJxWLcddddhEIh2tra2L9/Py0tLQwNDRGNRmeJtlQqRTwex+v10tfXx5YtW9ixYwfl5eWYzeZZIlf9XJlMJhwOBw0NDedsuiB5l+kylseWeBkSiWSZMV2i99hSr+NaZ1HKqKxWK263G4PBQDKZ1GqU57t3/cWimp0HBwfp7u4mEAig1+vp7++noKCAxsZGbTr28PAwfX19F13Lf7mo2RbVuK4a7dUMTTKZ5MiRIxw7doyenh6tbe9CoNPpsFqtbNu2jcbGRi2YD4VCWunUUkyCB7SJzevWreO6667T5oKoXoRkMkkqlcJoNJJKpTSDu8ViwWQyUVBQoHX4KioqIhAIMDU1RSQSOe/7rc6mKCgooLCwUHtf1NK2cDjM4OAgQ0ND55wSvpzo6uoilUoRi8WWVQkVZASAKgA9Hg/l5eU0Nzfz2muvcfLkyXnPxMTjcd5++20CgQBWq5X169cTjUYZGRnh0KFD7N27l46ODvR6PfF4nFAoRFdXF/v37ycajWqv39zXUe3+FgwGefPNN4lEImzdupWysjLNs6EOD51ZXuV2uykvL5/Xa5RIJBKJZLFYFLExk1gsxtDQEEeOHFk2O6iq6EilUqRSKQYHB4lEItrcgfb2dnbt2kU0Gl3UdQUCARoaGti+fTsf+chHqKmpweFwYDZnZhSpHahOnz5NV1fXght7TSYTVVVV/Omf/inV1dXY7XaSySRnzpzhqaee4uDBg0uSqdLpdFr729tvv53KykpMJpMWPJ8+fZre3l58Ph/l5eVMTExgtVopLCykoaEBs9lMVlYWWVlZlJWVUVdXh06no7m5meHh4fMGs2qzgcbGRq0Ll06n01rdtrW18cwzz3DgwAHGx5dvVYeiKDQ1NdHU1LTUSzkvPp+Pv/iLv+DHP/4xr7zyCn/2Z3/GG2+8sWDn6+vrIysri1QqxdNPP01raytHjhyhq6sLr9dLMpnEbDbT1NQ0q5zwQnniiSf41a9+RX19PR/84AfZsmULK1asoKioaNbjhBBYrdb3LLmSSCQSiWQ5s+BiI5VK8eSTTxKJRLj77ru1WRHl5eXs379/UXdR9Xo9FosFp9MJvDub4lyYTCbtcfF4fMm6wZSXl7NmzRpKSkqw2WxaRkMty1CHz8Xj8UUXQ/Du7r7NZlv0c6uYzWZuueUWtm/frrW+hUwXtIMHD/K///f/ZmBggGg0isPhIBKJYDQayc7O5uabb2bnzp2sWrUKj8ej7STff//9NDc38+abb3Ly5Mlzfk6j0Sijo6McP36cZ599llAoxC233IJOp8Nms7Fq1SqOHz9OX18f/f39dHV1LfZLc9Wgbky0trayc+dO+vr6Fvyc7e3tfPOb30QIoc2jSSaT2mfhfDNs3g9FUYjFYoyNjeHz+Whubub3f//3qa6u1rwnagne1772NeLxuMxsSCQSieSKZVEyG36/Xwu41HIqi8Wy6GVUdrud4uJiamtr0el0mrF6amrqrB15dVAYvDsnYrFQa7bVadQbN26cNVsgFovh9/sZGxujo6MDj8eDyWRakjKdmXMslgK1XXFJSYnW4Uen02ltfU+cOMHp06fx+XzaQMJkMolOp2NkZAQhBOl0mkQiwfr16/F4PBgMBrKzs9m0aRMTExN0dnYSCoXOOrdqjI9Go4RCoVkm5bltlJfq9bnaSCQStLe3L8pnXR0UeD4uZw3qRkc0GiUajXLgwAEcDses7m7RaJTXXnsNvV7P2NjYJZ9LIpFIJJKlZNHKqEZHR3nttde45557ALRyk8XE6XRSXFxMVVUVOp2Oqakpjh8/zvj4+FnlR6lUatG7ZUHmdVEzMHl5eaxdu5YVK1ZobTLVDlqtra00NTXR2tpKfn6+Nu9iqVhK/43RaKS4uBi73T7LyzIxMUFrayuTk5OacJyZoYpGoxw6dIhYLKb5dbZv347D4cBqtbJlyxZ6enp48803zyk2VM43oV1thap6RiTzw3L2vlwqv/rVr3jllVfOul3N6DQ3Ny/2kiQSiUQimRcWbc7G+Pg4e/fuZd26dRw7dowXX3xxyQIwdbJ0IpHQ2vEulwCmtLSU6upqVq1axa233sq6deu0zIE6E+TUqVO8/PLL2iyNt956i4mJiSVbs8lkWtIyKuCsYYzxeJzBwUFOnDjxnt6gVCrF6dOn6evro7W1FbPZzJo1a8jPz6e6upqbb76Zzs5Ofvazn53386qW2QWDQS0DFg6HaW1t5bnnnqO1tXVZTuSWLB9UcSqRSCQSydXGoogNdXDa/v37sdlsDA0N4fV6Fz3AT6fTJJPJWR2lzreG66+/ns985jOLUv6iliIZDAbq6upobGxk48aNrFy5kpycHC2QVkVRSUkJQggmJycZGhpieHh4USexz33NHA4HlZWVGAyGJWvvOvec6us5cwjc+VCnr7e0tPDCCy8QiUT4wAc+gF6vJzs7m+LiYk2gnuva1LK20dFRrUQrHA7T1dXFwMAAgUBg2c2ukEgkEolEIlkMFq2MSu3OMzU1RTQaXZISpUQiobU0zc/Pf8+guLi4mPXr1583wJxPdDodRqMRq9VKVVUV9fX1NDY24vF4sFqtWmlQLBYjHA6j1+sJhUKMjIzQ1taG3+9flMnmM8WaugsrhNC6MhmNxiXJEqm176p5V6fTafMvCgoKtKzQ+Uin08TjccbGxjhw4AAul4sbb7xR8xWpdfTvde6RkRGGhobw+XwYjUbGxsZoaWnB6/USjUblrrVEIpFIJJJrkkUTG4qikEgkGBwcXKxTnsXExAT79+/ne9/7Hl/96lffMwA9fPgw3//+9/nyl7+stcRdKLKysnC5XBQXF3PjjTeybt06ysrKZnkQIDN4rampiTfffJOf/vSnWgvOxQru1YFnTU1NZGdnY7fbtQzL6dOnycnJIRaLLYrwUVE7cbW0tDA8PIzL5dLmFVRVVXHbbbexf/9+fD7fecug1MwSwNjYmPa69vX1cfjwYY4fP/6eGZvJyUmOHTvG2NgYTqcTg8HA4OAgL7/8Ml6vVwoNiUQikUgk1yyLPmdjKUkmk/T39/Pss8/y6KOP0tbWRnt7O+l0+qxAsqOjg6eeeoqqqiqOHTvG0aNHF2RNBoMBnU5HcXEx3/jGN6irq8PhcGhdlSAT5IfDYWKxGD6fj6NHjxIMBhdVaEDG3+D1ennhhReIxWI89NBDZGdna52xlmr6dDwe58SJE+zatYuuri6cTifXX389brebe+65h9HRUU6dOkVbW9s5WxgXFBRQXFyMx+OhuLiY1atXYzAYaGpq4sCBAxw4cOA9henM7MaePXvQ6XRMTk4yPj6+7AbkSSQSiUQikSwm15TYUBSFcDhMX18fr776KgMDA7S1tZ0zIAwGg3R3d/Pyyy/T0dFBf3//vK9Hp9NRVFRESUmJ5tFQd8bV8i114KDX62ViYoLh4WF6e3uXrFwpmUzS0dFBRUUFnZ2dWK1WDhw4oAXySxFcp9NpzQw+OjqK1WqlpKSEvLw88vLyuOuuuygqKqKuru4ssRGJRDSxUVZWRnZ2Ni6Xi1AoxIkTJ2hvb2d8fPx9X+tUKkUoFKK1tRVAmyy9XBoPSCQSiUQikSwF4r2CISHEVRspqa13lyoYFEJgNpt5+OGH2bJlC5s3b2bjxo2a0IB3PRLRaJTDhw+zb98+mpubee6552aZ3BdzzXq9noaGBurr66msrGTXrl0MDAzMmjGxFOj1eqxWKxaLBavVymc+8xluvvlm1q1bh81mI5FIaFPiVRRFYWJiArvdrk1mV9vVtrS08Nhjj3Hq1Kkl7fQlkagoirJ0/aUlEolEIrlErlmxsdTodDocDgd//ud/zg033EBlZSX5+fmzshrxeJxAIMDg4CD/9m//RktLC52dnfT19S2J2FAxGo2YTCbMZjOBQGDRy7nOh+q9EELg8XhYvXo1GzduZMeOHVRVVZGTk4PRaNTEnFr+lE6nSaVSjI6Osn//fjo6Ojhx4gRHjhxZsmyNRDIXKTYkEolEciVyTZVRLSfUKeqlpaUYjUYSiYTmHVF77k9NTdHb28uJEyc4fPgw/f39+P3+JQ/uE4kEiUSCcDi8LESGysxM1djYGE1NTYRCISYmJqisrKSgoOCseSBqF6tYLMbAwACtra10d3fT398vhYZEIpFIJBLJZSLFxhKgzoCw2WyUlZWRTqfx+XxaZkPNavT393Pw4EHNpLxUMyzOx3Jay1ySySTDw8OMjo5y9OhR3G43WVlZmM3mcz5W9Vj4fD7i8bgUGRKJRCKRSCTzwHuWUUkkEolEIpFIJBLJpbLw47ElEolEIpFIJBLJNYkUGxKJRCKRSCQSiWRBkGJDIpFIJBKJRCKRLAhSbEgkEolEIpFIJJIFQYoNiUQikUgkEolEsiBIsSGRSCQSiUQikUgWBCk2JBKJRCKRSCQSyYIgxYZEIpFIJBKJRCJZEKTYkEgkEolEIpFIJAuCFBsSiUQikUgkEolkQZBiQyKRSCQSiUQikSwIUmxIJBKJRCKRSCSSBUGKDYlEIpFIJBKJRLIgSLEhkUgkEolEIpFIFgQpNiQSiUQikUgkEsmCYFjKkwshlAt4DDqdDrfbjdFoRK/XU1lZyfr167nvvvvYsmULLS0ttLW18aMf/YjW1laGhoZIpVIoioLFYiErK4uamhqqq6ux2+2YTCbGxsYIBoMMDw8TDAYZHx8nFosRDocX49KvOnQ6HTqdjlQqBYDdbsdoNKLT6TAajTidThRFIZFIkJubS05ODqtXr6a4uJiNGzeiKAo//vGPicVi5Ofn43a7KS0tpampiWPHjtHe3s7U1BSKopBOp0kmkyiKgqK870dIIlkUFEURS70GiUQikUiWG0sqNi4Eg8FAVlYWt9xyCxaLBYC8vDyqq6upqKjA5XJRV1eHx+MhHA7T3t7Oj370I6ampkilUiSTSXw+HwBZWVls2LCBsrIyAGKxGKdOncLr9dLR0UFPTw8nT57UhIrkwrFYLNhsNkKhEMlkko0bN5KTkwNAJBKht7eXFStWsGnTJiorK7V/TqeTvr4+4vE4X/3qV7FYLFgsFoQQCCF49tlnOX78OAUFBbhcLhKJBMlkkvHxceLxOIlEYomvXCKRSCQSiURyPpa92FB3saPRKE6nk4aGBjweD1VVVTidToQQ2mPS6TQAVqtV+10NWuvq6ti6dSuNjY1kZ2cDkEgk0Ov1DA8PE4lECAaDUmRcIkII9Hq9JgC3bdtGXl4eZrOZRCLB5OQkeXl5rFq1itzcXNxuNy6XC5PJRE5ODslkEpfLhcFgQK/XaxmMiooKdu7cSTweJ5lMEo/HCQaDPP/881JozCMGgwGdTkcikZDfAYlEIpFIJPPGshcbqVSKWCzG0NAQpaWlfPjDH8blcpGVlaUFpeFwmKmpKSYmJggEAuTk5GCz2bQSLCEEt912Gx/5yEe0UizICJmioiImJiZIpVJEo1EZaF0mt9xyC6tWreLuu+8mJydHy1LMRC25UikqKgIygkUVGYlEgng8TkNDA3V1dQAoikIsFmN0dJS9e/cSCAQW78KuYoQQWtnb1NQUyWRyqZckkUgkEonkKmHZiw29Xo/ZbKakpIS8vDxsNhsmk0kTEQBGoxGz2Ux5eTlZWVmUl5cDYDKZgEwwtW7dOk1o6HQ60uk06XQan8/HxMQEo6OjBAIB6QO4QHQ63SwhUVxcTEVFBevXr2fdunVYrVZN7CmKQjweJxqN4vV68fv9BINB1q5dq5VGKYpCKpUikUgwPj6O2WwmKytLy0wZDAaEEBiNRlKpFDfddBOtra0cOnRI829ILg71dTWZTGzZsoWioiL27t2L1+slHA5rmUKHw0FhYSFTU1P4/X70ej16vZ5EIkE6nZYiXSKRSCQSyXlZ1mJDDS6tVivFxcUUFxdjs9lmZScgIzbsdju1tbUoikJJSQkmkwmz2aw9Rt25nUk6nWZ0dJSenh76+/sZHh6WQdMFotfrtTI2gLq6OhoaGrj++uuprq6etTueTqcJBoMMDQ2xf/9+Tp06xcmTJ/n7v/971qxZo3kv1AzVvn37qKurY926dbPOpwoOvV7PBz7wAfLy8mhqaiISicjd+ItEFYJqA4UdO3awbt06JiYmOHXqFKlUSvMu5efns3XrVtrb2+no6MBisaDX64lEIsTjcWKxmPzeSCQSiUQiOSfLVmy4XC5sNhtVVVWUlZVx3333UVZWhtlsRq/Xa0Guoihahym145HJZNLuV4PTmWU7kCnPisfjtLW10dnZSV9fH2NjY4t+nVciatnN6tWrqa2tZfPmzVRUVJCfn09paSl6vZ7+/n4CgQB6vZ5gMMihQ4cYHByktbWVgYEBRkZG+D//5/9QU1NDMBgkHo8Tj8cJh8P09fVRUlJCU1MTyWSSVCrFpk2bKCgooKysjGQyycGDB2lpaSEWi2kdsCTvj06nw2q1AhkRmJWVRW5uLiUlJVRUVLBjxw7tPTGbzRQWFlJaWsqaNWtob2+ntbUVl8uF2WxmcnKS8fFxfvCDH+D3+6XgkEgkEolEchbLUmwIIcjKyiIvL4+GhgYqKipYs2aN1v5WLa1RS57MZjMmkwmbzYaiKJo5PJlMotPptB3xmcFQOp0mHo/T29tLX18fAwMD+P3+JbzqKwc1YK2oqOC6667jgQcewGq1YjKZtHKooaEhent7CYVCjI6O8uqrrzI8PMzg4CDxeJxUKsW///u/YzAYZpn7ISME8/LyaG9vJxaLaSJkw4YNFBYWkkqlaG1tpbm5WRqaLxI1I5VOp4nFYjgcDvLz8yksLCQvL4+NGzcyMTFBJBLB7XbT2NhITk4O+fn5lJWVUVtbi9vtxmQyMTk5SX9/P0899ZTWhUwikUgkEolkJstObGRlZeF0OnnwwQdpaGhg3bp15ObmkpubqwkNeNcsPDfYTKVSTExMEI/HCQQC2Gw23G43brcbg8FAOp2eZSpvaWnhzJkzDA0NEY1Gl+qyrxj0ej35+fk0NjZy22230djYiN1uJ51OEw6HOXbsGM3Nzezdu5eRkRFSqRSBQICenp6zSm7UrIX6u+rvAPD5fDQ3N2tiMRwOk0qlNK+Ox+OhsLCQsbExreRH8t4IIVi1ahV/93d/Rzgc1toRl5aWkpeXh9Vqpbq6mpKSEgwGA0ajEZfLpbWcdrvd2O12LbOYk5NDaWkp3/nOd3jnnXf4xje+QTQaXdL3Yu6mgkQikUgkkqVl2YkNi8WC2+2mpqaGjRs3UllZid1ux2w2zyqFUgfEzcxGpNNpUqkUIyMjRCIRvF4vOTk5GAwG7Hb7rMeEw2F8Ph+Dg4OMjIwQDAZlwPo+qEb8oqIiysrKqKysJDs7G6PRSCQSIRQK8c4773Dw4EH27duH1+sFMgIwEomcdbyZZny13A3QDN/hcFjz6KjHgYwoyc3NpaioiM7OTqLR6DmPL3kXIYTWROG6664jGo1SUVFBcXExOTk52vwStUzRZrNpgkMIoWUJLRaL1k1MzSquX79ea+SgZq0W+9rU9er1ei0TNjNbJpFIJBKJZGlYNmJDp9Oh1+spLCxk5cqVrFmzhsrKSm32gio0UqmU1klq3759PP7447OCVjWzYTAYyMnJoaKigm3btmGz2XA6nYyMjDA2NsbBgwfp6+ujra0Nr9crOxq9B6pR/6GHHqKmpoZPfvKTGI1GfD6f1kGqra2N3bt388orr3Ds2DGmpqaIx+MX1N1LLYHbunUrqVSK9vZ2qqurue6661izZg0NDQ2Mjo4SDocRQmC1WvmTP/kTBgcH+clPfsKpU6f41a9+tUivxpWHEAKn08lDDz3Epk2bMJlMmqhXM0WDg4P09fXx3HPPEYvF+OQnP4nBYGB0dBRA+/4JIbjuuuvIzs4mmUxqgzFPnDiB3W4nkUgs+vwTj8fDt771LcrLy/F4POzdu5e//uu/xuv1yvbIEolEIpEsMctGbKhtONV/gDaUT91FVf0AapDT3t7OgQMHtNIoyIiNUCiExWKhuroah8OhleCofo5oNMrY2BgjIyOayJBC473R6XSaj6akpAQhhFYyEwgEGB4epq+vj/7+fk1ozHxfZpa3zDTsq+U4DoeD3NxcIPO+19XVsX79elavXk15eTl6vZ7JyUltV119bH5+Pi6XawlekSsHNWuUm5tLTk6O9l1TM0lqa+KpqSkGBgYIBAJ0d3ej1+vp6uqaVeomhKC+vh6HwzHrO6n6rGKxGKFQaFGzCiaTicbGRoqLi8nOzmZycpKioiISiYQUGxKJRCKRLDHLRmzo9XrsdjtTU1M0NTXx0ksv0d7ezoMPPqjNYlCFxNjYGE1NTbz99tt0d3fPCoYgEzxZrVZsNhs1NTWa3wMyu6BOp5PDhw9r3XZMJhMDAwOy7OI8qCVrTU1NpFIp/H4/DodDK2P6wQ9+QDgcJhwOayZhNRBVxaJer9dM+3l5edogP5PJxM6dO3E4HPT39+N2u/nUpz5FSUkJVVVVGI1GjEYjK1eu1MRLPB4HIBaLMTIygtfrnSVIJbNR37/R0VGGh4dJJBLa+6LicrmoqKigsrKS3t5e/uEf/oFAIMDk5OQsoaHT6SgpKeGmm26iqKgIu93ODTfcQHV1NadOneLMmTP4/X7t+7oY6HQ6rUkEZL7jDz74ILt372ZoaGhR1iBZWoQQNwOvz7ipVVGUhqVZjeRaRAjxEeDnM27aoyjKzUu0HIlkWbFsxIa6QxqNRkmn0wwNDWE0GrXgKJlMEo/HGR8fp6+vj6amJgYGBrQd9JmoJnKbzYbdbsfhcMyq+1d3caempojFYrKE6gJQFIWpqSkCgYAWRKqD95xOJ263GyEEExMTQEYIQGa+icFgwGq1at2LampqWLlyJSaTCZPJxMaNG2cdp7KyErfbPavNsfr+hcNhEokEY2Nj9Pb2Mjw8zPj4+BVhDFYzDGoZoHrbzCwDsCCNClKpFF6vl8nJSUKhEIBWQiWE0ISc2WzGaDQyNjaG3+9nampKO4Zq1u/r62NoaIicnBx0Oh0mkwm3282aNWu0bIg6uHGxmFlmqU6fX+xyLsmyYA+wGxi/kAefQ6Scj3JFUfoudVEXcJ5vKIryZ5d6/Dnnegj4fWAjYAG6gMeBv1MUZV7+uAghPMBfAB8ESoEA8Dbwt4qi7J+Pc0yf5wbgi8A6oBAYBZqBf1QU5ZeXcdw64ChgA/5DUZRH5tyfCzwE3AusAUqAONAE/Bvwb4qizN2dPAX89fR/f+VS1yaRXI0sG7GRSCSYmppCr9djNBrZt28fp0+fJhqN4nK5tOnFLS0tTE5OMjY2pgmTuahzIBoaGqivr6eiokKbLTA8PExXVxf79u2jqamJyclJ2T71AlAUhfHxcS0DpLa/ra6u5j//5/+M0WjEYDCwefNmTpw4QTqdxmQysWLFCk2UHDx4kB/84Ac8/PDDPPDAA2RnZ2Mymc5qTawGwXNno6im8eHhYb785S/T2tpKV1fXFSEW1UDd7XYTiUS0GSRqeZPVasVgMJBIJOjr65vXNrJq57Z33nmHVCpFc3MzhYWFlJSUaAMy/X4/k5OTWjMF9f1QRcjM9+Ltt99mYmKC3/qt36KwsBCXy4Xb7eazn/0sJ0+eJBqNcuzYMU6ePLmo74u6WTEwMMCbb75JR0fHop1bsmzYrSjKYxfx+G7eDRDnsgb4EHDycoTGHFQxNJc35+PgQoi/Af4SCAJPAhPADWSu8Q4hxO2KolxWNw0hRAXwFpkA/CDwDJBH5rW6WwjxsKIoT1/OOabP81+AfwZCwNNAPxlho57nLxVF+dtLOK4B+BHwXqUMDwPfBobIiMReoGD63N/j3evU/sApinIKeGz6HFJsSCQzWDZiA9CCm2Qyqe2+9vf3Mzk5SSAQwO/3Mzg4SDgcJhQKnVWmodbzW61WcnNzqampobKyEovFQjqdJhQK0dnZybFjxxgcHCQUCpFIJK7a8ik1wwNcdtCnBqzBYJCpqSlsNhs2m03rUKT6MMrKyrR5G+pMB7XEqba2lrvuuou1a9ditVqJRqNEo1FycnK0nf2Za56J2kVsYGCA9vZ2enp6tBbH6rUtF8ExUzipZWB5eXm4XC4aGxsZGBjg+PHjWiZnzZo15ObmapPWn3nmGUKh0IJcTyQSob29Xes8pfpl1IGY69evp7CwkFAoRF9fH2+88YZ2HWazGbPZrJUyDg4OYjAYsNlsJJNJuru7OX78OIODg4s+s0b9/KXTaRKJhPbdlkjeC0VRupkOEOcihHh8+j//3zye8mLF0AUjhNhAJtswBWxSFKVz+nYB/CPwX4E/5TzXexH8Axmh8Y/AH6kB97TQeQf4nhBij6Iok5d6AiGEEfgfQJTMtbTOuO9rZLISfyGE+KaiKLGLPPyXgPVkMib/cJ7HnAHuB16cmcEQQnyJjMD6MBnh8eRFnlsiuSZZVmID0HZS/X4/4XCY/fv3YzQateFuahnPuQSCwWDAZDJRWFjI6tWrueOOOygtLcVmszE+Pk5vby/PPfcce/bsobu7m0gksmwC1IVgbqvgyxFVQgjC4TDj4+O0tLRoJm11d16lsrKSyspKzT+hztYYHR2lrq6OrVu3asLh1KlTDA0Nceedd+JwOLTzqOtVPR+A1hTgV7/6FcePH6enp4dAILDshOLMtrCKouB0OnE6ndxwww2UlZVx5513cujQISYmJli5ciUVFRU88MADlJeXE4lEGBkZYc+ePUQikXn1PAghsFgsBINB9u/fT29vL16vl+uvv15rI1xQUEBVVRWxWIzGxkYOHTrEgQMHAHA4HDgcDtxuN4FAgPb2dt566y2Gh4e1Lm9f+tKX6OzsPMtUvtCong217W00GiUQCGjeHonkYplRRhMhswt+JfAQIIDvqUIDQFEUZTpI/hzwX4QQf6MoyiX9cRFCWIB7yGQF/nLOzn67EOJfgD8GPgn806VfCjmACzgxU2hMn+e0EOIMmcxTFnDBYkMIsRn4KzKZnhPne5yiKK+d5/ZhIcR3gL8FbkaKDYnkglh2YgPe9W8oiqKVm6hTps8XxOj1esrKysjPz2f9+vXU1NRQWlqKw+EgmUwyMTHB8ePH6evrY2Ji4qopnVIFxcxOTxs3btS6Dqk4nU7y8vI4cOAAhw8fvqRzZWVlYbfb8fl8DA8Pa5kNIQS5ubk4HA7NjxAOh4lEIvT29pJOp7HZbFqXI3WYX3NzMz6fj3Xr1pFOp3E4HGeVTkUiEcLhMM3NzZw+fZpDhw7R0dGx7AJJtezL5XLhcrnw+XzE43HKysooKipi5cqVVFZWUl5ejsViwWg0UlRURGFhIZWVlTidTk38rl27FqPRSFdX17x9RtPptLYmRVEYHR1lfHycYDDIpk2bqK2tJSsrSytpq6ysJBqNUlZWht1uZ8WKFZSVlVFRUUEwGCQej9PQ0EBOTg6Tk5MMDg4yPj6uZRwX47ul1+u56667aGxsJCsrCyGEVg4mJ5pLLpP/BJiBf1cUxTuPx60VQvxXwAkMA28oitI2T8cunP7ZOfcORVECQohxwEMmSD92iefIAYzAqKIo52r1pp77Ni5PbIwCY0CdEGLFzNdo2m+xAjimKMrEhR5QCGEF/p3MtX+dTHnZpaCmTOUfGInkAlmWYgPQgtYLaV2pGm9Xr15NWVkZt99+O+Xl5RQWZv72JpNJent7eeuttzh58iRjY2PLbkf8UlCve2ZgZzKZuOWWWygvL9duUxSFiooKVq5cyXe+851LFhtut5vc3Fx8Ph/t7e20t7drBuNbb70Vh8OhGXQHBgYYHR3l+eefJysri7vuuovOzk5ef/1dj6Tf7yeVSrF+/Xqi0Sj19fWaEXzmY7q6uvjWt77Fiy++qN0+V5QsNaqJPS8vj5qaGjo6OggGg9TX11NcXMzWrVspKirS/q1evVobmgdo5mydTqfd19PTM2/ZjXQ6zeRkpqpBLX8ym800Nzdz8OBB/tt/+29ayZsqhNLpNCtXriQvL4+NGzeydetWamtr8Xq92hDFZDLJ0NAQbW1tmthYLBFvMpn40Ic+xKZNm7DZbCQSCSYmJrSZLLKMSnIZ/Pb0z+/O83E/Of1PQwjxJPCf50HUqKb4qrl3CCEcZHwVAA1cutjwAikgTwiRpSjK3C4Q1TPOcclMZ2M+B/wYeEcI8TQwSKZ86yHgJPCxizzs16fXt1FRlOT5Snbfi2m/x6enf71kg7pEcq2xbMXGxWC1WsnKymLt2rU0NDTQ2NiIw+FACEEkEmF4eJjTp0/T09OzqMHQQqDT6cjPz+eOO+7AarVitVq1LMKGDRuorKykrq4Om82GoiiEQiHa29uJxWKcOHGCsbEx7Ha7NiPjQlEUBa/Xq5Uymc1m7T71db711luZmJjA6/Xyzjvv4Pf7SSaThMNhfvKTnzA1NUV3dzfr1q2jsbERl8uF3W6nuroal8ulZUlUf4h6nNdee43Tp0/PWu9yEovqrJC6ujq2bdvGddddp62vsrKSrKwsCgoKsFgsGAyGWUb4mcdQOzvl5uZqA/cWon3szM5vvb29JJNJurq6yM3Nxel0akM0c3Jy+MM//EPNA6XT6RgfH2f//v20t7djMpk0IT86OsrU1NSiTnIXQlBUVKR1xVK72I2OjmrZF4nkYhFC3EQmWD6pKMq+eTrsGPBnwItkTOkWYDPwNTL1/4VCiJ3n6HB0MbwA/Dnw20KIf572o6j8dzIlVgDZl3oCRVEiQojXgNuBrwJfUO8TQlTzrki75HPMONfPhRCDZDppfXrGXSNkOkKdlcE5H0KI28h06PqzaSP3pfJ1YDXwkqIocpKsRHKBXBViw2QyaaUejY2NeDweTCaTNniur69PGzgXDoevSLGhZhBUT8qtt96K2+3G5XIxPj7OxMQE999/P4WFhbMGHKo+i6GhIdrb2/F6vVgslkuagxAMBgkGg1p725lrs9lsGI1Genp6mJycZP/+/cRiMXbs2EEsFmPfvn3akMYtW7awatUqysrKyMrKApg1JV4VG0NDQxw8eJCXXnqJwcHBs2apLAfU1rU5OTk0NDRw/fXXs337du31ULMVc836M/0oKuo8ErfbrYnlhWrpq2YOvV4v0WiUrq4uqqurtVbF6nu6detW7RrHx8fp7+/n4MGDHDp0CJvNRjqdpr29nVAohN/vX9T3RQiBy+XSXqt0Os3IyAhTU1MXLaYlkhn8zvTPectqKIpyksxuvEoQ+KUQYh+ZLMMO4D7g2cs4xz4hxHeB3wVOTGdMJqePfd30+VeRyUxcDn9EpnvW54UQ24B9QC4Zw3QXsHYezoEQ4hHgX4CngL8BeoAKMp6L/wPcBPzGBRzHTUacHAD+12Ws5w/I+FFagE9d6nEkkmuRq0JsOBwOCgoKKCsrIycnZ5bQGB8f1+r8h4aGFnXn9XJQzbixWIxUKkV5eTkVFRX85m/+JsXFxZSVlWkBuioCjEajVqvu9/v59re/rQ298/v9jI2NMT4+rmUcLpWZWQU1IH777bfp7OwkEAgQjUbx+Xyk02n27NlDKpVicnKS7OxsamtrWbFiBTU1Nbjdbu29gow/QzX7ql2T+vr6GB0dXZDZE5eDXq8nPz+f6upqHnzwQerq6lizZg1ut1vzsqj/AM0rMVMAZmdn43K5tDkbailWfX09Pp8Pg8GwYG19bTYb2dnZ3HrrrWzdupVt27bhdrtnmf1VcasSi8UYHh5mcHCQzs7MpqJa6rgU7YeFEDgcDiwWy1mNBSSSS0EIkUMm07AoxnBFUfxCiJ+Q6SK1k8sQG9PH+6wQ4iAZwaQG4u8AdwK/RUZsjF7mOU4JITaRCfrvIJMxGCXTEvZxMt2aLusc076M75MxcX9qRsanRQjxKaAeeFgIcbOiKLvf53D/m0wJ2e2XYYz/HJnOVaeA2y6n05ZEci1yxYsNNUDLysrSWnOq3YBUo/Lo6KhWWrHcUQfg5efnU1hYqAmDuro6Kisr2bFjh7ZrrhrmIbMrrg4zi0ajjIyM8POf/5yRkRGMRiOJREKbSzIfbXDn/j4yMsLw8PBZ96kiIZ1OYzQaKS4u1mYzmM1mTTCpQasaoIdCIQYGBvB6vQSDwWW1S636GrKzs6msrGTnzp0UFxdTUlJyVrZC/RzG43GtdEktJ1MUBZPJhNVq1UQbgMVi0W5bqPVbrVaKiopobGxkx44d5OfnzxJ+c0u9ZpZehcNhrUxJbVW92Oj1ekwmk/YZgsxnTBXnM8XeuQZ/SiTn4VEyxvAfKooytUjnHJv+aZ+PgymK8n0ygfoshBDfm/7PQ/Nwji7gN89xjs/M0znuIGNE3zO3tExRlLQQYi+wafrf7vc51kbASkaonOv+TwohPgkcVxRl/dw7hRB/BPw9mWGCtymKcllCSiK5FrmixYY6ADCVSuH3++nt7cVut2v17mazmezsbMrKymhrm6+GHwvLnXfeyRe+8AWsVitms1kTSevWrcPhcGjtPWOxGC0tLfzrv/4ryWRSC/gURdHamvb395NMJolGo5fd+vZ8qEHo+VADQYDrrruOz33uc5pxX/UpqELopZde0mZndHR08Oqrr2oZkuWwW20wGDAYDBQVFeF2u1m1ahX19fWUlJTgdDpniQN1zWpw/stf/pLh4WH6+vqIxWJEo1Hy8vLweDzcfffdlJeXY7PZiMVi/PCHP6S5ufm8QysvZ/1GoxGbzca2bdt44IEH2LJlC9XV1bNERiqVIh6Pa0J+ZsexdevWsWLFCrq6uuju7iYcDs/b+i6GDRs2zPJmxeNxxsbGeOGFF/B6vdTV1WmthV988UVaW1uXlWCVLFv+8/TP+Zyt8X5cP/3zgj0IF4sQ4g4yJUh7FEUZWKjz8K5n4z8u8ziqKTD/PPert1+IMesp4FxdUYrItPHtICNYeuc+QAjxp2R8GsfIZEYuaDK9RCKZzRUtNtTJ1NnZ2eTk5GidfVRvgLrrmkgklnUbTFUYeTwe6uvraWho0CZyW61W/H6/Nm3b5/NpGYympiZ2794961ipVIrBwcFlUXakluHYbDY8Hg81NTWUlZXN2slX36dYLEZ/fz/Dw8OMjY3R0dFBT0/PUl/CLIxGIxaLhaKiIjweD5WVlVRUVGC328/qopVKpUilUtogxDNnzjAwMEBHRwc2m42cnBxisZg2fE5t7ZxIJOjv79daBs8X6mfMbDbjcrkoLCyksbGR3Nxczeitzq9R59wYDAacTqf2fNXnAcwqt1oKysvLqa6u1hoVqF3Q0uk02dnZ1NTUUFdXR319veb3icViJJNJvF6vJnAlEhUhxI1AI9A8j8Zw9dg7gLfn7tJP+xI+SiZofmIezuNUFMU/57YaMuIpRcakfrnnMAPMHKY3PTjwr4HtZAbh7b7M07wx/fMj04P7tJkYQoj1wEcABTjnPIyZKIry1XPdLoS4mYzY2K8oym+f4/6/ImOCfwe4Q5ZOSSSXzhUtNurq6ti4cSPr16+nqqqKrVu3YrfbNc+CamZ9+umn6e09a9NiWaDOZti+fTvf+MY3tE5E6kTk3t5e+vr66OrqIhQKsW/fPsbHxzlx4oQ2zXsuy2UH12AwUF5ezo4dO3jsscdwOBxYrVYAbcK4oiiEw2F8Ph+pVIpAIMDrr7++7LqGCSHweDwUFRWxc+dOampquP/++7Hb7bM8A5C5tvHxcW3ew+TkJEeOHNH8J/feey+f//zncTqdOBwOTCaTlrHy+/3z3rJVNapXVlaSn5/PihUr2L59O6tWrcJgMJBKpbROUupk95/85Cd4PB4+/vGPa8KiubmZ5557jtdff522tjZisYsd3Ds/CCG46aabuOuuu7Sua8FgEIvFwpe+9CUtm2kymTAYDNx7770kk0mmpqaYnJzku9/9Lu3t7ezdu1eWV0lmohrDFyKr8R+AbtoQ3k+mG9V1wBYy8xp+d073qEvlX4UQFWQCZC9QS8Z4bgR+W1GU/fNwjhXAG0KIV8h01jKR6U61kkz51KfP/9QLQ1GUg0KIfwM+Axyabn3bA1QCD06f81vTxvt5RwjxKBmhkSIjfP7gHCVY3Yqi/GAhzi+RXG1csWJDCEFeXh6lpaVUVlZSUlKC1WpFr9cTCoUIh8MMDw8zPDzM6OjosjSGi+mpzps2bWLVqlXk5uZisVg0E3EoFKKtrY2WlhYikQihUIhTp04xOjpKT0/PsgrG52I0GnE4HKxatYra2lot8zTTTxONRkkkEkxNTeH1epmamiIQCBAMBpcskD0Xapeo0tJSqqurqa6upqamRvMJqdekZjNSqRShUAifz0d/fz8jIyOEQiHMZjP19fVUVVXh8Xiw2WyzDNjvV5J2KahdpWw2G2VlZXg8Hmpra6msrMRsNhOLxfD7/XR3dzM2NobD4SAajTI6OorBYCAcDmO1WjEajZpPQs1GLSWqUFNLvNTysNLSUrKysrDZbFo5mF6vx2AwkJ+fj9Vqpbq6mlgspmVrJBIhRDaZ3fKFMoZ/G/gAmc5QeWTa0A4APyATNB+fp/O8wLvmcAcZo/aTwN/NzA5cJiPAS2SyGPeRGXLXAnwe+GdFUear5/RvAXvJDFi8k8z1+Ml0wvoXRVF+Ok/nORfqrBI9me5b52IPmfdPIpG8D1ek2NDpdBgMBiorK9m+fTs1NTVamZGiKExMTDA8PMyhQ4c4c+YM4+PjJBKJBWsjeqkYDAYKCwv52te+hsfjASAQCBAKhXjhhRd45pln6O/v13a71XKj5d5xRwihTcb+9Kc/TW1trSY04F3TdFNTE83Nzdp064MHDzIwMDDvXoXLxWKxkJWVxW233cb27dtpaGjQMhLqbpdalhMIBPD7/QwPDzMyMsKTTz5JX18fFouFtWvX8uijj1JWVkZeXt6s1rYLdb06nY7i4mJKS0vZuXMnlZWVbNmyBbfbjdlsprOzk5dffpnnn3+e48ePYzAYsNls7Ny5k+zsbLq7uykuLqa4uJiKigruvPNOurq6GBwcxOv1LkkXKgC73Y7Vap3VLhjQSiej0SgdHR0cP36c2tparbQqOzubO+64A7fbzZNPPrlo084ly5vpgXrWBTz+N4BvLNTxZ5znh8APF/gcY8wZTLhA51HIBPM/WKDj7+bd2SNz73sMeGwhziuRXItcsWLDZDKRnZ1NUVERDodDq91WTbnJZBKLxYLdbic3N5epqallt1u+adMmVq5cicfjwW63EwqFGB8f5+TJk5w+fZre3l78fj/hcFirpV9OQfhMVLO+x+MhJyeHHTt2UFpaSlVVFW63e1aZUSQSwe/3c+LECY4fP651DRsaGiIQCCy7a1S9QQUFBRQWFmqfN1UoxONxIpEIY2NjCCE0/43D4WDnzp3a5O7CwkJKSkq012NuWj4ajRIKhYhEIvNSSqXu+K9du5by8nI2b96My+VCURSmpqbw+/00NzfT0dHBwMAAfr9f697k8/nwer14vV5ycnK0LFx2djZ2u13rArUUwbr6GhuNRu01FEKQSqW0z9bAwADHjh3j+PHjhMNhKioqqKys1Hwo2dnZC9btS7KkfEUI8RWgVVGUy5piLZFcDEKIjwA/X+p1SCTLkStSbKitbquqqqipqdHM1KrJVRUbBQUFRKNR1q5dy8mTJ/H5fEu9dA2z2cxnPvMZNm3aREFBAalUiqGhIQ4cOMD//b//VxvUt1w6Mb0fFosFh8PB7bffTmNjI5/85CfJzs4MkZ05GRxgeHiYgwcP8tRTT/HWW28RiUSW7Q6zOsuhtLSUhoYGysvLZw2+U8vAhoeHOXz4MBs2bKC6ulorPdq0aRPpdJpgMKiVNM0d8qeeZ2pqiqGhIbxeL5FI5LJfD71ej8Vi4e6772bDhg3U1dWRTCZpb2+ns7OTN954g+7ublpaWhgaGprV0ezMmTMoikJHRwc5OTkoioLFYqGwsJDc3FwcDgdTU1Naxm2xUDuYOZ3OWSVsquhrb2+npaWFX//61xw/fpzm5mZuvvlm6uvr2bp1qzYNvbi4WHsPl+PnTnLRdJMxKKvIrkGSxeYUsz+D3Uu0Dolk2XHFiQ01+FuxYgUej0cbiAbv1tYXFBTgdDpxOp2kUilaWlrO6ha01CQSCQ4cOEAoFKK+vh4hBHq9nlWrVvG5z32OQ4cO8cYbb2jG3eVWWnQ+cnNzqays1EpcZvoY1KBuYmKCzs5OxsfHtbkIyzHgUwNbm81Gfn6+NkBOLQeDzGfObrfj8XjYsGEDhYWFWCwWbUif+r5aLBbt8TMzIrFYTDum3+9nZGSEQCBwWVk4q9WKzWbj1ltvpaqqinXr1uHxeNDr9SQSCbxeLwMDAwwMDDA4OKh1OJuJyWTCaDRqXg3IvG+nT59mYGCASCSiddBaTO644w5uuOEGqqqqNKGRTCa1OS/PPfccAwMDnDx5ksHBQRKJBJ2dnej1egKBAA6HA0BrwCCzG1cH0+bqx5Z4GZJrGEVRTiE/gxLJObnixIZOp8PtdtPY2EhRUdEsg63q5SgoKCCZTGrtRd944w0MBsOyCmiTySRPP/00ra2tPPLII1itVgwGAytWrOD666+nqqqKYDBIS0sL4+PjV8RgMrVjU0VFhebRULNN6gRtgIGBAdra2jQvzXJ6X+ai7qLn5OTgdrs10TDzfrUjldvt1rJswKxA1mKxaNeplsOFQiHGxsYoKSnBbDYzMTHBwMAAwWDwstqy2u12ioqKePjhh7n++uvJzc3VhiemUil6e3vp6enh5MmTmjF/pthQxZHaKlcVTj09Pfz85z/n9OnThMPhRc9qANx777189rOf1daktrw9cOAAR48e5Z/+6Z80IaTS1dXF1NQUExMTWkmYXq/XBMdy6d4mkUgkEsnVyBUhNoQQWK1WrFYrxcXFbNy4kVtuuYWioqJzPl41iI6Pj9PT00NfX9+yKqFSiUQidHR08LWvfY3CwkJqa2tZsWIFeXl5rFmzhs9//vM8//zznDp1in379jE1NbVssgBqwKbicDi07mD5+fnafclkEp/Px7FjxxgaGqK9vZ2Ojg6ampqYmJhYFtcyF3UmRV1dHTfffDMbNmxg7dq1FBQUAO+KCEVRtP9Wp4rPzHrMRX2sOn+jvb2d119/nVQqRTKZ5NSpU/PqWzGbzZhMJm2SthBCE7RqRiOVSmGxWMjJycFms3HXXXeRl5enCZbt27fjdDpJp9Pa96+/v5+JiYn3vNaFQhUJanbI6/UyNjbGr3/9a86cOUM0Gj2veFAzVSaTCafTSV5eHqlUCq/Xu8hXIZFIJBLJtcMVIzbMZjNOp5Oamhqqq6tZvXo1LpfrnI9XfRtqV6rBwUFCodAir/r9iUajDA8P861vfYuKigp+8zd/U6uPV83IExMTxGIxTp48SSQS0Vr4LnWQrpqP1VIhu91OQUEBBQUFuFyuWWLD7/fz5ptv0trayq9+9Svi8fiy3k1WzdD19fV88IMfpKGhgeLi4rNM3TPr/dWA/v1Q29vGYjFaW1vZvXs3TU1NjI2NaUJyvt5btQxq5toMBgMejwe32000GtW6TxUVFVFYWMiDDz5IUVERoVAIl8tFWVkZkOnyZDabycvL08rJlmKw38zSJ3WeSXt7O2+//TZ9fX3nNNbPfc/UYZlutxu/3y/FhkQikUgkC8gVITZUg62iKPT09OB0Ojlw4ADbtm3DZrNpJSJq28uJiQlGR0fZs2cPXV1dS738C2J0dJSf/vSntLe3MzIywoYNG6ioqGDr1q2sXLmS1atX09XVxZNPPsnY2NiSZwXUuQWVlZWsXbuWLVu2sGHDBoqLizUPgrq+aDRKa2srZ86c0Vr3LmdUP0UoFGJiYkIrATtffb86sf5896tlU6rI6Ojo4D/+4z84ffo0x48fx+fzLUgLWdUro6IKxKKiIm677TYSiQR5eXl4PB5cLhc2m42SkhIMBgPRaFQTk8lkklgsRjgcJhaLkUgklqT8baZoUtf005/+lCNHjtDT03PeDYWsrCxyc3NxOp1apkcikUgkEsnicEWIDXh3joHf72dqaorBwUECgYAWTKmDxlQD7ODgIF1dXYyMjCyb0qP3IhKJcPr0aVKpFG63G4/HQ3FxsWZ037p1K4WFhRw+fBjIvB6qwXgp5m7o9XrsdjulpaWsW7eO6667jtraWu1+1Tys1tQPDg4yNTW17H0n8G72QW2jqprY54oJ9XVX71N33WfuvKuopVKRSISenh5+9atfaaJxIT6fqnF65rHV0jer1UpBQQE7duzQJtarmQo1EFdLlQBtvkskEiEcDmv+ocV8L00mE1lZWZpnRvUBnTp1irfffhu/33/e9ZhMJux2+yyzu/p+SSQSiUQiWViuGLEB7+7MqnMPVFN1a2srPT09hMNh/H4/e/bsYWhoiFOnThGNRrVJ1VcCvb29PPHEE1orWTVAKikpoaCggN/93d8lEongcrk4cuQIP/vZzxgeHmZ8fGk6PRYWFnLLLbeQl5eHXq9HURQSiQTf//73GRwc5J577qG3t5eJiYllOcX9XKjX0NPTw/79+7WsWUVFBXa7XRNR4XCYSCTCyMiINl1bLSObSSKRoL+/n0gkQiAQ4MyZM0xOTmotf+c7aFdLCI8fP47D4eD666+f1UhBLXtbuXLlLD+HulbIlGCpGZ6pqSk6Ojp49dVXeeGFFxgbGyMUCi1qKdwDDzzAX/7lX1JSUoKiKExOTjI4OKgNvXyv19BqtZKdnT3LU6O+Bup8HolEIpFIJAvDFSU2AG1HVTWBx+NxBgYGaG5u1oK55uZmxsbGGBsbu2LmVKio4qivr48zZ85gMpmwWq0UFhZiNps1U3xhYSGpVIrDhw9rpS6hUGjRRZXFYsHpdGqthdXd/ra2Njo7O1mxYgUjIyPzNqhusVAHDY6MjNDf38/g4CD5+fmYzWYSiYRmfFcH8akdqOYGvaowGRsbIxAIMDExwcjIiFaOtBCfTbVca3BwkJaWFmpra8nJycFqtc7KzswciqeWeqlrUtft8/kYGxujtbWV3t5e+vv7icVii9b2Vs3GFBQUUFtbi16v1ybOq0Lj/USPukExc2K7RCKRSCSSxeGKEhvqALUzZ85gNBoJBAK0tbXx2muv0dLSopUV9fX1vWdXmiuB5557jtdee00z7sbjcUpKSojFYrjdbgoKCti0aROKotDb20tfXx/PP/88p0+fXtJ1J5NJzQA+MjLCl770JSKRCD6f74oooVJRFAWfz8eJEydIp9NMTk6i1+spLy+nra1Nm0ifk5PD6tWrtXI3tUxHFcXRaBSfz8dLL73EwMAAnZ2dDAwM4PV6F0wIq2Lgqaee4pe//CW/+MUvqKqq4rHHHsPhcGilX4lEApvNpgmoRCKhzfno7++nu7ubF154gfHxcUZHR7Xp5jNb+C40qsfE4/FgsViIx+MEg0F27drFnj17GBwcfM/Pldqa2OFwaFkNtcRMHZ4okUgkEolk4biixIZa1hEMBhkdHcVqtQLQ19fH4OCgVhevThC/kgmHw4TDYa1c5/jx44yOjmKxWCgrK6Oqqgqj0UhVVZX2+LkzIBYSq9VKfX291tVIPbe6O67u+I+Ojl5RGY2ZJBIJQqEQg4ODmM1mjh49it/vp7W1lWg0Sl5eHm63G5PJpA3Bm/ncRCLBxMQEY2Nj9Pb2MjAwQHd39yyv0UKgZiZ8Ph/BYFAzd7e1teF2u2d1xMrKysJut2vm756eHrxeL319fQwNDdHS0kIwGCQYDC7Yet8Ls9lMVVUVubm5CCGIx+NaCVVXVxfxePy8zzUYDJhMJqqqqqiqqtJKyVS/RygUes/nSyQSiUQiuXyuKLGhDkJLJpOEw2GtM1V3dzdTU1Pa466mMonJyUm8Xi/f/OY3MZvNlJeXs27dOlwuF4WFhTQ2NqLT6ejr69MGsC10RkcIQXl5OZ/97Gex2+0MDg5iNBqxWq1EIhG8Xi+jo6OaOf9KJZlMMjU1RSAQ0NrUmkwm7bV/5JFHcLlcs9rAqmVkY2NjDA8Ps3fvXs0QrpZdLfTnUxUTavYiFovh8/n44he/qHkUVFGotsf1+XyEw2EtAI9EIlp2ZikzUnl5edxzzz1s2LABgOHhYXbv3s2RI0fo6Oh4z+GHDoeDnJwcHnjgAW644QYtqxOJRJicnGR4eJhAILBYlyKRSCQSyTXJFSU21B3beDyuTQ9WPQ5XusBQ+/9nZWWRnZ3N1NQUPp+PsrIysrOzZ7X8BDSDvDrkbLFRxZ7L5cLlciGEwO/3a8biaDR6RbS5fS/UtavlTup8hrKyMoqLi6mtraWgoACz2axNolbFRjAYZHx8nOHhYcbGxohEIks2BV41vA8PD2veEtWYrtPp0Ov1mqdGLcFSP2fna+e7WJjNZjweD1arVWtr3draytjY2Pt6XtRBi+qE95nXov4tuZJK+yQSiUQiuRK5osQGvDs7IB6PX1W7knq9HqfTSW1tLdu2bePIkSMcPnyYu+66i9WrV2sB7CuvvILT6dTmWcxsT6oGUIvBxMQEu3btori4mPr6eoaGhhgdHeUTn/gE5eXlBAIBbTbKlY4qBMvKyigvL2f9+vVUVlaybt06rFYrJpNJaw8LmUB2eHiY06dP09nZSVdXF+FweElL+1KpFENDQxf9vKV8/3Q6HVlZWVRVVZGVlaW1ut21axf9/f3vW56ndq9T36OlFk4SiUQikVyLXHFi42pFnZJeVlbGxo0byc7Opq6ujg984APU1NQAEI/HycnJobCwUNtJj0aj9Pf3c+TIEUZGRhZlrWopSmtrKxMTE1qpl9fr5emnnyYvL4++vr4rftdYne3gdrvJy8tj586dVFVVUVlZSU5ODhaLRcsUxONxxsfHGRwcpKOjg6NHjzIwMEBLSwter/eKLidbClQ/0sqVK7UObF1dXbS3tzM2NqYJu/ciFosxOTnJoUOHMJlMrF27FqvVSjAYJBQKXfGfT4lEIpFIrgSk2Fgm6PV6LBYLxcXFXH/99axdu5ZoNMqKFStwOp3a0MKGhgYURUGn05FMJkkkEhw/fpxdu3bh8/kWbSc6HA7T0tKiDUdTMysHDx5clPMvBurwu/r6esrLy7n//vuprq6eNe9FvfZIJEJHRwfPPvssP//5zwkEAtowwKshu7OYCCGwWCxs3ryZNWvWUFhYyMjICHv27OH06dMXPFMmHA6TSCTYu3cvg4ODuFwuPB4PoVBoQbuBSSQSiUQieRcpNpYJ6tyGt956S/OkCCFYt24dxcXFDA8Pax4Iq9VKRUUFAwMDnDhxgiNHjhAIBBa965MaqF1tu/bqbAe3282qVauorq5mxYoVFBUV4XQ6tUF4AD6fjxdeeIHu7m5OnjxJS0sLfr9f82fIYPbSUE3u4+PjHDt2jObmZp5//nlaWlou+BiqCV71eJjNZtxuN7FYjKGhIeLxuHx/JBKJRCJZYKTYWCakUikCgQAnTpzg6NGj2O12srKy6O3tJS8vj+bmZoLBILm5ueTl5VFXV0dbWxvPPPPMkq57ZrB2NdXE63Q6zRtTWVlJXV0dLpdLy2aoDQomJyf50Y9+RHt7O93d3TJ4nUfUoYL79++nubmZV1555aJKn1TBMjg4yMjICFNTU1r726uhPbZEIpFIJFcCUmwsE9RBY2o3I7UE5ODBg5jNZm1S8sjICEajURMfy4mrKdBWFIXi4mJuuOEGysvLKSgoIJlMMjk5SXNzM36/n+rqanp6ehgYGCAUCl1V17+UqJ6gN998E7PZjMPh0MqeLvV4apZEzUjJ8jaJRCKRSBYHKTaWETODKbXr1vnmCIyOji7Wsq5JFEXBYrGQn5+P3W7HaDQSiUSIRqPs27dPM3339/drg/Mk80cqlWJ4eHjejqcKGIlEIpFIJIuLWMrdPSGE3FqULEuEEBQXF7N161ZtOngikSCZTNLc3EwsFqO+vp5AIMA777xDIpG46rwrkotDUZSrp45QIpFIJJJ5QmY2JJJzoM41UQ3JOp2OeDxOMplkbGxMG4iXTCaXbFifRCKRSCQSyXJnSTMbEolEIpFIJBKJ5OpFt9QLkEgkEolEIpFIJFcnUmxIJBKJRCKRSCSSBUGKDYlEIpFIJBKJRLIgSLEhkUgkEolEIpFIFgQpNiQSiUQikUgkEsmCIMWGRCKRSCQSiUQiWRCk2JBIJBKJRCKRSCQLghQbEolEIpFIJBKJZEGQYkMikUgkEolEIpEsCIalXoDk4hBCyJHv1yBCCIQQbNy4kdraWh5++GEMBgN/+Id/SCAQoKamhpqaGtauXcvevXt5+eWXSafTKMqlfVwURRHzfAkSiUQikUiuQaTYkEiuEHQ6HW63m+zsbPLy8jCbzezcuZNYLEZBQQEej4fa2lrOnDmDw+EgkUiQTqdJJpMoikIymVzqS5BIJBKJRHKNIcWGRHIFoNfrMZlMNDY2snXrVlavXk1WVhbf/va3EUKg1+u17IcQgjNnzhAMBonFYvj9fmKxGF6vl3Q6vdSXIpFIJBKJ5BpCig2JhEzWQK/Xk5+fTzqdZmpqimQyueTZgJKSEq6//nqysrJwOBzs3LmT2tpaLBYLer0evV6vrV+lrq6O+++/n2g0SjKZJBAI4PP5+PnPf04wGFyqS5FIJBKJRHINIi61pluyNEjPxsJgMpnIysrijjvuIJFIsG/fPoLBIIFAYEnX9clPfpK/+Zu/IScnh6ysLC1z8V7M9GooikIikWBsbIxbb72Vnp4eUqnU+55XejYkEolEIpHMBzKzcY0zN3BVFAWj0YjdbicajRKNRpdoZReHEAK3243b7aagoACbzcY777yD3+8/67FzBbbRaOSTn/wk5eXl1NfXMzk5SSqVor29nebm5sW6hLMQQqDT6Wb9U4nH48TjcU6dOoXBYKC8vFx7L202GxaLRXusoihaVuT06dM0NzcTj8dJJBKLfk0SiUQikUiuLaTYuMbR6XRakKooCoqiYLPZyM/PZ3x8/IoQG6pnobCwkFWrVtHQ0IDb7aanp4dQKDTrsXM7NAkhsFgs/PZv/zYNDQ0YjUYmJycJBAKYzeYlFRuqwDhXJiMejzM+Ps4//uM/YrVa+cxnPoPBkPk6V1VVYbVatccajUasViu33HIL+fn59Pf34/f7pdiQSCQSiUSy4EixcQ2j1+u58cYbycvLAzKBt8FgwG63k5eXR29vL+3t7bS3t+P1eoGzswJLgdVq5aabbkJRFFpaWsjJyaGmpoba2loaGhpobGwkOzubVatWaWVQIyMj7N27l/b2dpqamkin06TTaVwuF7m5ueTk5GCz2RBCkEwm2bNnD21tbUt6nel0mpGREZqbm1m1ahU2mw2dTkcqleLFF1/kyJEjHDp0CIPBQDQaxWAwoNPpKCkpoaSkhHvuuYfCwkL0ej06nY7rr7+egoIC+vr6aG9v59ixY6RSqWXxnkokEolEIrk6kWLjGkan07Fy5UpKS0u17IDVasXhcFBQUEBTUxNms5nx8XF8Pt8sH8BSrtlms7F582aSySTxeJySkhLWrFnD6tWrqa6upqSkBJvNRnV1tfY8r9dLJBLBbDbT1tamlRE5nU6Ki4ux2+2YTCYURSGVSrF///5zlmAtJoqiMD4+TnNzMyUlJRQXFwOQSqV46aWXeOWVVxgdHSWdTtPS0oJer8doNOJ0OnE6ndTX15Obm4uiKAghKCkpwWQysWLFChKJBM3Nzdr1SiQSiUQikSwEUmxco+j1egwGA5OTk2RlZXHvvfdqtf5msxm73U5OTg4lJSWMjo4yOTlJMBhc9MBU7biUSqUwGo385m/+JvX19dx1112YTCYeffRRTSg5HI5ZoiEWi5FOpxFCYDQa+ehHP8q9997Lb/3Wb/Hcc8/x7LPP8sd//MezsjuJRIJIJLLkO/4Wi4Xi4mI2b97MBz7wAYqLi9Hr9QwPDzM4OMjQ0BDRaHSWAEylUqTTabxeL8FgkG9961tUVVVx1113UVJSwooVKzCZTPj9fvx+vzZ/QyKRSCQSiWShkGLjGkVt9aqKB4/Hg8PhwGQyYTQaMRqN2lC47OxszGazFtwupuDQ6/WYzWYt67Jp0ybWr19PaWkpRqNRK3sKh8MYjUb0er0WRAeDQZLJpDajwuPx4PF4qKys1MqTtm/fTmNjo2a+TiaTxGKxJQ/C9Xo9LpeLnJwc8vPzsVqtCCEIBAIMDQ0RDAbPWqPquVHN42+99RYtLS1UVFRgNpupq6vDYDBo4sxgMEjfhkQikUgkkgVFio1rFIvFgtvt5sYbb2TdunVa8K7T6bSp01arldzcXJxOJzabDbPZrLVRXeidf7UTU2FhIRUVFTz88MNs2rSJ+vp6rFYrRqORWCxGR0cHx48f5/nnnycajRKLxUilUuh0OvLz8zGbzVitVkpKSnjooYfIzc0lLy+Pe+65hxtvvJHs7Gz0ej3pdJpEIkFnZydNTU1LvusfiURobW2lvLycpqYmVq9ejc1mo7y8nLy8PFavXo3P5yMSiZBIJM651ltvvZXVq1fz4IMPkp+fr5VY/d7v/R5HjhzRDPCdnZ1LLq4kEolEIpFcnUixcQ0ihMBms+F2u/F4POTk5GAymbRd71QqRSqVwmAwYDabyc7OpqioiMnJScLh8PvOeZgPdDodJpOJ/Px8qqurWb9+PStWrMBut6PT6YhGo0xNTXH69GmOHTvG7t27iUQiRCIRrXSqrKwMu92O3W5namqKtWvXotPpyMvLw263k5WVpZ0vkUgQi8Xo7u6+4FkUC0k6nSYcDjMyMkJnZyfFxcXk5eVpBn6Xy4XL5cJoNJJOp88p/iwWC06nE4fDgcVi0RoAFBQUUFlZSVFREQMDA7O6kUkkVyJCiJuB12fc1KooSsPSrObqQAjRDKyacdMtiqLsXqLlSCSSKxgpNq4xVA/Exo0bqa2txeVyAcwaFqfX67U5DalUirvvvpu6ujr+3//7f4yPjy+Kn8FiseDxeLjrrrt48MEHqa6uJisri2AwyNTUFI8//jgdHR388pe/JBwOEwgEtDIiyATOaiCt0+lobW3l4MGDfOxjH+OLX/wiZrMZs9msBeodHR10dHTwF3/xF/T39y+b8qIzZ87wox/9iO7ubrZt28bOnTtxu90UFhZSXV3N4OAgwWCQaDRKIpHQ1q3T6WhqaiIej3PXXXdpPhy9Xo/NZqO4uJhNmzbh9Xo5efLksigdk0jmgT3AbmD8Yp8ohFgDfBG4BfAAPuA08K+Kovz7BR5jC/AQsB7YABQAA4qilJ7n8bnTj78XWAOUAHGgCfg34N8URUlf7LWc51zVwF8Ad0yva5KMQPtrRVFazvGUfybzOtwM3DQfa5BIJNcmUmxco8RiMS1APZd4EEKQTqe1siTI+BkWesdfCIHZbCY/P58NGzZQW1tLQUEBJpOJdDrN6OgoQ0NDtLa20tHRwcjICOn0uf9fPHOtqshS28CqwiqRSBCPx+np6eHUqVMMDAwwNTW1oNd4McRiMcbGxvD5fPh8PuLxOOl0mhUrVmjlYpFIhFAopL0uZrMZi8VCaWkp+fn56PV67XiKopBOp4nH41qHrqXO4kgk88huRVEeu9gnCSH+E/A9IAy8AHQDbmA1cA9wQWID+ATwh0CCjFApeJ/HPwx8GxgiE/j3Tj/nQ9PruVsI8bBymTsBQoiN08d3Aq8BPwXKgA8D9wkhPqAoyv6Zz1EU5Z+nn/sYUmxIJJLLQIqNaww1MD9x4gR9fX1s3ryZ/Px8LSsws0TK7/fT1dXFoUOH6OnpYXh4mFAodN7gfj4wm80UFxdz66238uijj1JdXY3H4yEWixEIBPjJT37CyZMn2bt37wWvRa/XU1BQwMMPP8xNN92E3W4H3m0tOzg4yE9+8hOOHz9OMBhcsGu7FNLptGbMN5vNhMNhIpEId955J3fddRfRaJRUKkU8HufNN9/kS1/6EpWVlZSUlHDvvfdSX19Pfn6+NvAvlUoRDAZpbW3liSeeoL+/n3g8vsRXKZEsHUKI68kE9s3AXYqiDM+533gRh/sB8EPgpKIocSHE+4mEM8D9wIszMxhCiC8BB8mIgQ8BT17EGs7Fv5IRGl9QFOXvZ5xnG7AX+HchxCpFUZZHSlcikVxVSLFxjWE0GrHZbOTl5VFUVERxcTE5OTmzyqggE4j7fD6OHz9OR0cHAwMDCy40IFM+VVdXR3l5OcXFxZopva+vj4GBAdra2mhrayMcDl9UkKz6Feb6UsbGxmhra2NgYEArEVtOJJNJIpEI/f39HD9+HKvVSmVlJQ0NDVqpmypEGhsb+djHPqZ1sGpoaCAvLw+TyaR120qn04RCIXw+H9FolGQyuZSXJ5EsB/4O0AOPzBUaABcTgCuKcuxiTqwoymvnuX1YCPEd4G/JlDFdstiYLp9aD4wC/zDnPG8LIZ4lI2ruAp6/1PNIJBLJ+ZBi4xrD6XTi8XjYsGEDpaWlrFmzZtbON7xbanP69Gl++tOf0t3dzdTUlGa+Xuj1bd26lR07dlBSUkIsFsPr9fKTn/yEX//61zQ1NZ2z7et7oWZt1DkU8O48jf3797Nv3z6ampqYmJhYqMu6ZBKJBFNTU7z++uvs37+fpqYmamtr+au/+issFgsmk0l7bGNjI1/5ylc00XguI38qlWJkZISxsbFFuwaJZLkihCgFbgQOAyeFELcAmwAFOAa8Pl+eiUtAFTmXuyNQOP2z+zzX0jn98zak2JBIJAuAFBvXGE6nk9raWiorK6moqMBqtaLX62d1JFID+ZqaGh588EFefPFFjh07ppmIFUVBp9NhNptJJpPzZqZW5z9YLBatDW84HGZoaIjBwUEGBwfP2+b1vVCzNO+88w4ej4ctW7YwOTnJ8PAwra2ttLW1EY1G5+UaFopUKkUkEqG3t1d73eeKibnZKfU2FdUMPzExoXUWk5kNyTXOddM/28h4GW6ec3+TEOJDiqK0L+aihBAG4NPTv/7yMg+nmuUrhBDiHP6P6umfsnuXRCJZEHRLvQDJ4iGE0Iba1dfX09jYqIkNlXQ6TTqdRlEUqqqq+PjHP87KlSuxWCxaCRJkuh3Z7Xatpep8oNPpMBqN2o69TqcjGAzS0tJCT08PIyMjl+QvUBQFr9fLrl27OHLkCH6/n+7ubt544w2OHDlCU1MTkUhkXq5hoVB9GQMDAzQ3N1+ywEsmk/T19TE+Pk44HF42XbckkiXCM/3zN4BGMv4IF1AL/IhMh6gXhRCmcz99wfg6GXP6S4qi/OpyDqQoyhky3pAC4Pdn3ieE2Ao8MP1r9uWcRyKRSM6HzGxcI6hZg7GxMTo6Orj++uvJzc1Fp9PNyhTodDrNzxCJRJiamsLv9xOJRBBCYDKZsFgs2Gw2ampqtDkQqkiZD2auR11HPB6fl0F7+/bt43/8j//B+Pg4Q0NDtLW1aR2ergTC4TDj4+N885vfpLq6mu3bt6PT6bT3aWxsjA0bNlBdXY3VasVgMGizOHp6eujq6uKXv/wlnZ2d+P1+aQ6XXOvoZ/z8bUVRXpj+3S+EeJSMANlMxtPw+GIsSAjxB8AfAy3Ap+bpsL9LJkPyD0KI+8iUiJWSEVengLXA8jKsSSSSqwYpNq4R1HkTwWCQzs5ODAYDWVlZ5y27UY3JPp+PcDhMLBZDp9NhMBhwOp3k5ORQWVmJ0WhkYGBAC9jnmswvFtUvogb/aptWVWhcrtg4c+YMZ86cuaxjLCWJRAKfz8e3v/1tsrOz+f3f/32MRiOdnZ2Mj49z6tQpHnnkET70oQ+Rm5uLzWZDr9eTTCZpa2vj8OHDvPbaawSDQZnVkEjAO/0zBrw08w5FUZRp8/RmYAuLIDaEEJ8jY+I+BdymKMrkfBxXUZTd0zNA/pJMG9ubgD7gvwPHgWfJGMglEolk3pFi4xpBDeLVYN1isWCxWLS5Eyrq5Oru7m7a2to4efIkgUAAj8eDzWbDbrdTX19PVVUVDz/8MJFIhL6+Pnbv3s0rr7yiCYOpqSni8fhFBbTpdJpgMEhXVxcrVqwAwGAw4HK5yMnJweVy4fP5pM+ATMbH7/fz85//HCGENoMjEonQ09NDW1sb2dnZ2nusKAo1NTWEw2FycnIA8Hq973MWieSqp3X6Z+A85mn1S2Jd6IUIIf4I+HsyLXhvUxRlXoN/RVFOkCkXm3vev57+z0PzeT6JRCJRkWLjGkM1Easm7JmmYrVjUyQSYWhoiK6uLnp7e0kmk7hcLgoKCnC5XJSXl1NfX09tbS2pVIry8nLGx8dpamoiEAgQjUY1w/XFlD4pikI8HmdiYoLx8XESiQRCCJxOJ263G7fbrXXEulLKnhaSeDzO6dOnNSGpZq/8fj+jo5k4RfXZ6HQ6HA4HeXl52iR2iUTCCTIG6jwhRIGiKCNz7l89/bN7IRchhPhTMj6NY8DtiqJc9AT0SzyvmYwRPU1m0J9EIpHMO1JsXCPodDpMJhM2m43c3FxcLpdW06+KDXWA3PDwMPv37yccDmOz2SgvL6e0tJQPfehDFBUVkZ2dTVZWljbVG6CqqorGxkba29sZHR1l8+bNpFIp3nrrLW0o3YXg9/t5/fXXCQQCBINBduzYwc0334zL5eL666/nxRdfpL29nb6+vivKa7FQzJwLopaZpdNpksmklslSBaU6WVzt9CWRXOsoipIUQnwX+Avg74QQn1EzHEKINcB/ItN69hcLtQYhxF8BXwXeAe6Yr9KpOeewA1FFUVIzbjOSmV5eCfxfRVE65vu8EolEAlJsXDOYTCacTif5+fkUFhZiNpvP8muk02lisRjhcBifz0cikSCRSGCz2TCZTFRWVlJQUKCJFDVgNRgMZGdnU1paqu2Yezwe0uk0OTk5+P3+C56NkUwmmZycpK2tDY/HQ11dHXV1dXg8HtasWUNrayt6vZ6JiQmEELMyJzLjkREciURilsBTW94GAgG8Xi+JROKaf50kkhl8jcyMiU8Da4QQu4F8MqZwC/DHF9r6VgjRAPzZnJuzhRA/mPH7n6iZi2kT+lfJmLPfAP7gHN39uhVF+cHcGy+SW4DvCSFeJePVcAL3kBEaLwJ/cpnHl0gkkvMixcZVjjo5u6ysjBtvvJG6ujoaGhpwOByEw2GysrK01rfqwLeJiQkMBgMDAwOcOXOG2267jfr6eoqKinC5XNpcDrXDlV6vp7GxkaKiIrq7uxkcHKS3t5dwOExpaSnDw8P87Gc/u6BMhBos9/b2MjExwejoKLt37+bTn/40GzdupKysjJGREb7+9a/T399PIBDQumd5vV58Pt9ivKzLFkVRGB4eprOzk3g8rs0qmZyc5LHHHqOlpYXTp09L34tEMo2iKGEhxG3AfwM+BnwOiAL7gP+lKMquizhcIfDonNtsc257jHdnX1RN/9QDf3SeY+4BfnARazgXZ4C3yBjDPUCEjDH8r4F/X8LBhRKJ5BpAio1rBHVOg16vx+VynbeUJplMEovFiEajmrnbYDBgNpsxGo2auFBRd+FMJhNut5uioiIt8+H3+5mcnCQUCl30LA5VmPT29uJwOBgfHycvL0/rhrV+Qg1BjQAAwAVJREFU/XoqKirw+/0kk0mSySQtLS20t7dr4kMtI5ppjL+a0ev1mEwmPB4PJSUl2nRx9bX3er2Mjo5e0mBEieRqRlGUMBkR8NhlHmc3cMF/7BRFuexzXuB5zpDJ1EgkEsmiI8XGVY4abI+MjLB7925yc3PZsGGD1llqpuBQA9BgMEh/fz/pdJqysjJtcJ/BYMBgOPdHxmw2azM4ysrKqK6uJhgMsnv3bkZGRi5abKgtb4eGhlAUhaeeeoqamhrWrFlDdnY2X/ziFwG07lSxWIznn3+ep556Cq/XSyQSIRAIkEgkCIfDmvi4mrHZbJSWlnL77bfz8MMPk52dmdFlNptxOBxkZWVht9sRQkixIbla+YoQ4itAq6IociL2ZSCEaAZWLfU6JBLJlY8UG1c5qjHcbreTl5eHx+MhLy8Pk8l01kwMte2tTqejpKREu6+srIyysrJZGQ2VuSJCfYzaUvdyp4tHo1EmJydpb28nFArhcrlQFIWpqSmMRqM2tC4rK4uNGzfi9/vx+/3EYjFtSvbQ0BB+v5/+/n4SiQSpVOqqDLZ1Oh16vR6j0ai9vzPL3QoLC5mYmODUqVNLvVSJZL7pJlMSpLIo3Zyucv6ZdyeswwJ35JJIJFcvUmxc5agtT8vKyli3bh2bN2+mpqYGo9F41mNTqRTDw8OYTCZuuOEG7fbVq1eTm5t7zufMRQ1sVS53EF8wGCQYDDIyMoLFYsFqtTI6Okp/fz+FhYU0NjZqu/Zbtmxh7dq1RKNR4vE4/f39TE1N0draSl9fH0899RSBQIBAIHDNlFap6PV6Nm3aBMCvf/3rWV2sJJIrHUVRulmEcqRrCUVR/nmp1yCRSK4OpNi4ypnZ9tTtdmsTpc+VdTAajTQ0NBAMBvH5fJhMJu15M0uu3i9bMbML0tjYGLFYjPr6ekZHRxkcHLykIF8tqzp27BiDg4Pa7I2+vj4qKys1AaV2yTKbzZSVlVFcXExhYSHd3d10dHQQDAYJh8MMDw8zOjp6VbbPtdlsWK1WdDqdZvofHR1l7969nDlz5qq7XolEIpFIJMsXKTauYmYO7LPZbOTn52O328+aGq4+1mw2U1VVhd/vp7u7G5fLhd1ux+12YzabL7osSg10I5EIGzdupLOzU/NgXCxql6o33nhDW6vdbqempobVq1dz2223UVNTQ3FxMSaTCZPJRG5urtaJq6SkhGPHjhGLxVAUhebmZkKhkCZiriYcDgdWq1Ub0nj8+HHeeecdfvrTnzI1NbXUy5NIJBKJRHINIcXGVYxqBA6FQkxOTjI2NkZfXx+5ubnk5eVhsVi0x6rlToqiYLVaqaio0DpQmUwmrd3thZxTp9NhMBiwWq0UFRWhKAodHR1MTk7OS+mSKjxCoRDd3d1aJsbj8ZCfn4/JZMJoNGrCKp1O43A4uPfee4nFYoyNjTE6OkooFLoqWsDqdDpsNhtr167l7rvvpq6ubqmXJJFIJBKJRAJIsXFVo4qNaDRKIBDA5/PR39+Px+PB6XTOEhsqiqJoQ/p0Op3272IyGqpvw2w2k5eXh8/nY2xsDJ/PN28+CbW9bTQaZWJigq6uLpxOJ1lZWeds67tt2zbuvvtu4vE4TqcTp9NJNBqdl7UsNarYqKys5OabbyY/P3+plySRSCQSiUQCSLFxVZNOp9Hr9TgcDm1yeElJiVZeNBM1I6HOZlAFxqV0k1IUhXg8jtfrZe/evXR2dtLR0UEoFJqX65qLOvnc6/USDAbPWrMQglAoxB/8wR+QTCYJBAK0tLQsyFqWglQqpRnhX331Ve6///5ZguNy3kuJRCKRSCSSy0GKjasYRVEQQmAymXA4HLhcLlwuFxaL5bxtbNXb1cD0UsWGOudCFRpTU1MLVrKkln/F4/Fz+i9UsfGzn/3ssrtjLUfUa5+ammJsbEzL2KjtjGOxGPF4/Kq7bolEIpFIJMsfKTaucvR6PXa7ncLCQtavX09xcTFOp/OcYgNmi4uLERpqEJ9IJEgkErz55ps0NTXR2dmJ1+td0mD3ahQY5+K6667jC1/4Arm5uQghiEajhEIhnnjiCd544w0CgcBSL1EikUgkEsk1hhQbVzk6nQ673T5rRsb7TZC+nHIbNbAPh8OEw2FisRiJROKaCfiXEpvNRk5OjlYKp3baGhgYYGBgQM7WkEgkEolEsujo3v8hkisVIQRZWVmsXr2aVCrF888/z5EjR7R2tPOdbVDLsIxGI7fccgsPPfQQVqv1qp3YvdwIBoOMjo4SiUS011y+7hKJRCKRSJYSmdm4ylEN0QAGg4Hx8XFGRkYwmUwoiqK1qb1U5gazsViMSCSC3+/XJnVLFofx8XFOnTpFeXk5Ho+HaDRKMBgkEoks9dIkEolEIpFcowi583llIYS4qDdMCIHRaGTlypXcfvvt2m735s2bqaysZPXq1VgsllnTwS+0jOpcn53Dhw9z9OhRWltbGRsb48UXXyQQCJBIJC5m2ZJLwGKx4HQ6Wb16NfX19eTk5GAwGPjhD39IX1/fRZVRKYoiW1dJJBKJRCK5bGRm4ypHNW37/X7GxsZIpVKk02n6+vowGo3U1dVhNBq1QFRtgXs+A/nM4yaTSU1wqCJmYGCA9vZ2+vv7GR8fJx6Py+zGIqF2pOru7gYgOzsbvV5PMBiU5VQSiUQikUiWBCk2rgEURWFsbIx9+/ah1+vR6/XodDqmpqbYvHkzJpNpVstYs9mMzWbTAtRzZTpSqRTBYJB0Oq0Jj0QiwaFDh3jjjTcYHh4mEolo/gHJwqMawru7u+nt7dXeGyk0JBKJRCKRLBVSbFwjxGIxJicntQFvp06dYnx8nCeffJKioiI8Hg9Wq5W8vDzcbves6eLxeJxUKjWr1CoWi9Hb26tlN9SWt319fQwODhIIBGRWY4mQIkMikUgkEslyQYqNa4R4PM74+Lj2++joKDqdjoMHD+Jyubj11lspKSlh586dCCFwu92aMAmHw4RCIUwmk3abz+fj0KFDWkcrdXDcqVOnZJvVZYAUGhKJRCKRSJYDUmxco6jTpaPRKGazGUVRmJyc5Omnn6ayspLt27djsVgwm82Mj48TDAax2WxadqOvr48XX3yRWCyGoiikUimSySR9fX0ymyGRSCQSiUQiAaTYuKZRFIV4PE4sFiOdTjM1NcXhw4epq6sjHo/jcDhwOp1MTEwQjUaxWCyacbynp4dXXnmFWCw265hqCY9EIpFIJBKJRCJb315hXGzr2ws4HkajkdLSUhRFwefzYbVa8Xg85Obmkp2dTV9fH36/H5PJpIkNv99PZ2fnWVkM+Xm6OpCtbyUSiUQikcwHUmxcYcy32FBR/RgzxYPb7cbhcDAyMkIwGNT8GvBuGZbk6kSKDYlEIpFIJPOBFBtXGAslNs7V3lav12MwGGRXqWsQKTYkEolEIpHMB1JsXGEslNiQSGYixYZEIpFIJJL5QIoNiUQikUgkEolEsiDolnoBEolEIpFIJBKJ5OpEig2JRCKRSCQSiUSyIEixIZFIJBKJRCKRSBYEKTYkEolEIpFIJBLJgiDFhkQikUgkEolEIlkQpNiQSCQSiUQikUgkC4IUGxKJRCKRSCQSiWRBkGJDIpFIJBKJRCKRLAhSbEgkEolEIpFIJJIFwbDUC5BcHEIIOfJ9DkajEbvdzoc+9CHuuusu2tvbmZiYQK/XI4RAr9djs9morq6mubmZX/ziFzidTtxuNy6XC6PRiNFoJJVKcfr0aXw+H8PDwyjK2S+1oijEYrFz3nc1oSiKWOo1SCQSiUQiufKRYkNyxaMoivYP4O677yY7Oxur1YrBYMBisSCEQAjBli1bKCgooKCggKKiIrKzs7HZbJhMJvx+P3/yJ39CLBajoqLinOdKpVL09vYSjUYX8xIlEolEIpFIrkik2JBc8Qgh0Ol06PV69Ho9VqsVm82G1WpFr9djNBq1x7rdbjZv/v/s/Xd4XOd54P1/z5neMANg0HthBXunqGY5kmzJjmVJVpzEjjeJ340vZzfFSdabTbJxkjfvZnfjXxI7zm4cx7Fjx3GTZcuyZFkiTcqqlNgJEgABEr1jem/n98fwHAHsBAECpO7PdfECMfXMDEA+93nusg2Px4PL5cLlcmG1WjGZTFitVvbs2UNHR8dFz6FpGqdOnWJgYACXy2U8ZqFQIJ/Pk8vlyOVyN+01CyGEEELcCiTYELc8PVDQ/9hsNmw2G3a7HZPJZNxO0zR8Ph87d+4EMHY7dC6Xi09+8pPk8/mLnkPTND772c8yPDxMVVWVcXk2myUajZJMJgmHw4v4KoUQQgghbj0SbIhblqqq2Gw22trauOeee1i9ejUulwuz2YyqqnMCCd2Fl2maZlymKAoOh+OytRr3338/fr9/zmWpVIrJyUl6enp44YUXFvgVCiGEEELc2iTYELcsVVVxuVxs3ryZj370o8blerBxIT2o0IOJ2UGFvstht9sv+3y7du1i06ZNxvd6sDE6OsqPf/xjCTaEEEIIIS4gwYa45eVyOeLxOB6PB4/HY3yfyWSMTlSKoqBpmlFfoQcaFosFs9l80W6GpmkUCgVyuRxmsxmz2YzVasVsnvsrM7seRAghhBBCzCXBhrilaZpGNpslEongdruxWq1kMhmy2SyqqmKxWLDZbKiqagQPqVSKQqGApml4PJ5Lpk3pt02n0yiKgtlsvmRgUSgULpmuJYQQQgghJNgQt7B8Pk80GuXgwYNkMhljZwOKqVR33nknVVVVDA0NUSgUiEQiTExM0NvbSyaTIZfL8ZGPfIS2tjagGLjoux6apqGqKqqqGrfV07PMZjOKohi3raqqmlPLIYQQQgghiiTYELcsTdPIZDJMTk7y6quvGl2oFEXBZrNRWVlJNBoFIJ1OMz09zcTEBMePHyedTpPL5bjzzjupqKgAirsUmUyGQqEAgNVqxeVyGSlV+twOvcOV3rVK73wlhBBCCCHmUm73Sci3G5kgfjF9Srhe5O33+42uVPoOhF6rkU6niUajlJWVUVZWRnNzMyUlJcDbKVm5XI5YLMaaNWt4+OGHjcuam5vx+XxUVlZisVjIZDJkMhkmJiZ45pln+NSnPrXE78TCkQniQgghhFgIsrMhAIxWsSaTyahX0BfvF3Z2UhTFGGa3HOjpT7pMJoPZbGZ8fNwIFGZ3oCoUCtjtdlwuF6dOnTLup7+mfD5PPB5HURRWrFiBzWbDarWSSCSw2+0X1XjIQD8hhBBCiEuTYEOgKAolJSU4HA5qa2sJBAIMDg7icDiMRbnVajVubzKZSCQSjIyMGClHy2mHLBAIEAwGr3hswWDwskP4zGYzDoeD7u5uUqkUDzzwAO9+97sxm82kUqll9VqFEKAoyr3AT2dd1K1p2uqlOZrbg6IoJ4GOWRe9S9O0/Ut0OEKIW5gEGwKz2UxlZSVVVVV0dHQwPT0NQFlZGaWlpZSXlxu1EFAMTiKRCNls1tjdSCQSpFIp8vn8ki/G9SDjQqqqGsXfejcqnd7SNpfLoaoqdrsdh8OBw+HA6/VSWlo6J10L3p7NobfGFUIsuQPAfmD6Wm6sKIoLeAR4GNgCNAAFoBv4d+DzmqZlbvSgFEVZD/w2sBWoB0qAyfPP8w/AU9o1/sOpKEoD8IfnH6sJKAVmgD7gy8DXNU3LXua+HwN+E1gL5IEjwF9rmvbMJW7+D0AlcC9wz7UcmxBCXIqskN7hVFXF7Xbznve8h9WrV/O+972P0dFRfvKTn9DU1ERzczNVVVW4XK45Q/HGxsb44he/SCZT/H+4t7eXnp4eIpGIcdlyoaeC2e12stmsMX9Db2WrKAo+nw9FUZicnMRqtVJXV8eqVavYtm0be/bsobGxEZPJNCelTJ/PoQcjQoglt1/TtM9cx+3vAr4OBCjujHwfKAPeD/w18KiiKO/WNC11g8e1lWJQ8zrwKhAGqs8/z5Pnj+Gjl7vzBdqAXwbeOH+8AaAceC/FYONXFEW5X9O0ObmdiqL8NfB7wDDwT4AV+DDwQ0VR/rOmaX8/+/aapv3D+ft9Bgk2hBA3QIINgaIoxjwKi8VCeXk5u3fvpqysDJ/PZ6RRFQoFCoUCyWQSu93Ozp07jVoFr9eL3W6ns7OT6enpJa/nmL37YLVaje+tVitWq5WKigqjZa3T6cTpdJLJZNi3bx8ul4tNmzbR3NxMR0cHpaWlRqAxu+A8lUqRyWSIxWJG16vFYjKZMJlMxs7R5XZvhBDXZRz4CPCd2TsYiqJ4KO6Q3EFxJ+CzN/g839Q07SsXXqgoSgnFAOQjiqJ8XtO0g9fwWK8CpZqmzflHQFEUC/ATijsRjwLfnnXdHRQDjT5gu6ZpwfOX/2/gEPDXiqI8o2la//W/NCGEuDIJNoRBX8RWVFTQ2NhopAnpi+t0Ok0+n2d6ehqbzcaHPvQhYxHf1dXFK6+8QjabJRqNGoPzloIeFEBxke52uwFIpVI4nU58Ph979uxhy5YtbNq0Cb/fj6ZpBINBAoEAXq+XJ554gurqaurq6nA4HEa7W/1xEokEZ86cMXZxRkdHF+31KIqCw+HA6XSSSCSMoYVLna4mxK1O07SjwNFLXB5VFOWzwL9RXLzfULBxuZ0RTdMiiqI8D6wBVgBXDTYul9alaVpWUZTvUzzeFRdc/YnzX/9SDzTO36dfUZQvAH8C/Crwp1d7fiGEuF4SbLzDaZpGOp2mu7ubaDSKpml0dHSwYcMGI+VofHycUCjExMQEkUiEkZERysvLefe7343H48Hr9eL3+9m2bRtdXV2EQiF6e3tJJpM39bXoU75bWlqoqqqipqYGRVHo7OwkkUiQzWbx+Xw0NDTQ1tbG+vXrqampwe12E4vFcLvdfOQjH8HtdrNy5UqcTqcxW0N/rwqFAhMTE0xNTfHaa68RiUQIhUKcOXNmwV+Pqqp4PB7Ky8tZtWoVTU1N9Pb2Mjo6Sn9/P4lE4pL3mz3RXAISIeZNr3tYtFZziqI4gfvOf3viBh/LBDx0/tvjF1ytP8ePL3HX5ygGG/chwYYQYhFIsPEOp2kayWSSV155Bbvdzssvv8wjjzxCe3s7wWCQmZkZDhw4QF9fH6dPnyYYDBIKhVi5cqWxCPb5fPj9fnw+Hxs3biQejzM6OnrTgw2r1YrT6WTHjh2sXLmS7du3o2kaX/ziF5mZmQGgoqKC+vp6Nm7cyIoVKygpKTG6TLndbh577LE5g/tm01sC9/T00NnZyd69exkaGqKrq2vBd3H0YvTS0lK2bt3Kjh072L17Ny+++CI9PT1MTk6STCYvCib0XR29PTFIwCHEPP3a+a+XWqDPi6Io7RTTtkxAFcXC9Frgf2iadmGAcLXH8gP/CVCACuB+oB34BvDMrNu5gDogpmna2CUeSj9TsvK6XowQQlwjCTYEhUKBeDxOOp0mlUpx4MABTCYToVCIUCjEwMAAU1NTBAIBMpkMuVyOTCaDz+fD6XRiNpuZmZmhv7+fw4cPc+jQIeLx+E1/HVarFY/HQ0dHB3fccQeNjY2oqsrHP/5xUqmUMenb6XRSU1ODz+czCr69Xi9Q3B3Ru1bl83my2Swmkwmz2UwwGCQYDPKzn/2Mvr4+zpw5QyQSWdDXoKdM+Xw+tm/fTlNTE5s3b6ajo4OmpiaGhoZIJBI4HA4URTECCUVR8Hq91NbWUldXR0VFBUNDQ0xPTzMwMHDZXRAhxMUURflPwHsoplh9eQEfup25uwcZ4A+YX5qW/4LH0igWtf+3Czpbec9/vXSv77cv983jGIQQ4qok2BAAxjTsRCLBoUOH6O3tJR6Pk0gkjHaxOrPZTC6Xw+l0YrPZMJlMRCIRjh07RmdnJ2fOnFmSAnF9PkZ7ezutra1GYfeuXbsAjO5T+m0tFgu5XM4Y8qd3rdIX8XoRuD5jZGZmhnPnznHq1ClOnDjBxMTEgtdO6MFGVVUVmzdvZsWKFWzZsoXKykpKSkqor69naGgIi8UyJ11KURQ8Hg+rV6+mpaWF1tZWTp48ycjICOPj45fcBRFCXExRlEeBv6VYPP7Y5drIzoemaT8uPoViARopdpX6/4B7FEV57Hra7Gqa1nX+sUwUdy4+CPw5cKeiKA9rmha43sO7ztsLIcQ1kWBDXCSTyRAMBsnn85dMD9I0jUQiQX9/PyUlJdTV1aGq6pyOTbPPui82PcgACIfDHD58GJvNxp133onT6cRutwPMaVur/13v8qQfqz5BfPZOwJkzZzh16hQvv/wyvb29Rn1LJpO5aF7HjXK5XHzwgx9k9erVfPCDH8Rut2O324nH44yPj/PMM8/w2muvMTU1ZXw2JpMJj8dDY2Mj69at46677mLt2rVUV1dz8uRJTp48SSwWI5tdsDWTELclRVEeAb5JcQbGuzRNO7sYz3M+gOkD/lxRlAzwP4Dforgzcb2PlQcGgb9TFGWC4nyQP6eYYgVv71x4L3H32ZdfbudDCCFuiHr1m4h3mkKhQCaTueLuRCaTIRwOk06ngbfb584eenezqKqKxWIxit2npqY4d+6cseswO/iZHRjoOxmzAyV9R0Pf1clkMgwMDPDyyy9z4MABXn75ZaampkgkEpcNxuZLb9O7YsUKNm7cSF1dnTFQMZVKMTAwQF9fH6dOnZqzU6GqKjabDb/fT2NjIw0NDVRUVNDS0kJzczNer3dO+18hxMUURfkQ8B1gArhH07Tum/TUz53/eu9iPJamaXFgBHArilJzifvonat6FuD5hRDiIrKzIeZND0YURcHtdtPe3k51dTUej4dwOHzTUqmy2SzhcNgIdvTUqNHRUUKhEGNjY2QyGdLpNGaz2VjQV1dXX7QA14MPgOHhYV566SWOHDnCq6++yszMjBFcLQaz2YzL5aKtrY3q6mpUVSWdTjM5Ocnx48d59dVX6erqIhKJzHlvTSYTDoeDVatW8cEPftDooNXa2kpZWRlvvvkmTqeTQ4cOkUrd6GwyIW4/iqL8EvCvFBfli7ajcRl1578uRNeryz3WPopDA98D/MsF17131m2EEGLBSbAhFoTZbMbpdGK1WjGbzTf1LLq+G6FLJpOEw2GGhoaw2+10dXWRSqVIpVK4XC78fj+1tbXk83lj12P2EEC9E5Te9ndiYoJgMEg6nV601DBFUXA6nXi9XioqKnC5XEBxB2l6eprBwUFGRkYIh8PkcrmLjkNPpfJ4PEawpBfMu91uPB6P7GwIcQmKonyMYhH4AMVAY2ARnuNO4I0L6z8URakA/ur8tz+6xsfaCZzQNC1xweVu4O8u81j/l2Kw8UeKonx/1lC/ZopDC9NcHIQIIcSCkGBDXLfZ6Uc6k8lkTCBfilQqwEhreumllzh+/Dhf/epXjY5Sei3Gnj17eOyxxwiFQkxPT+N0OrFYLFgsljkpVXpQ0tLSwszMDAMDA0xOThKLxRb8uPXgZuvWrXR0dLB27VpcLhfZbJaRkRG+853vcPToUd58803i8fhFqVt6CpveSUt/7/VWvbFYzKjBEUK8TVGUd1EMNFTgp8CvXuLfrpCmaX97g0/190C1oiivUKyvyAPNFOdiOIDvc+1dr/4QuFdRlAPnHysBNFDcofBRnDD+P2bfQdO0VxVF+f8BnwKOK4ryXcAK/AJQBvxnmR4uhFgsEmzc4vSFqtlsNgqW9bPz+jA6vYh5IYqZ9ce12Wy4XC6jU5OmaUb60oW1ETeTpmlEIhEymQyxWMxYYBcKBQqFAoODg/T19aFpGqlUivLycpxOJ6WlpVgsFmw2m/GeulwuVqxYwczMDIFAwGgNnM/nF/z1qapKRUUF5eXlRn1FPB4nEAgwMjLC6OgosVjssjUiF+7O6PSAYzF3ZYS4hTXxdu3ir13mNgMUu1PdiM8CjwCbgQcpLvSnKaYufQ34tnbtv6D/BMSB7RRrM5xAEDgEfBv4sqZpF6VkaZr2e4qiHKdYOP4fgQJwGPjfmqY9c+HthRBioUiwcQvTF8U+nw+v18vk5CTxeNxIZSovLwcglUqRTqeNxeqNFDUrioLL5aKyspL29nZ8Ph+KopDJZIhEIiSTySVf2IbDYSKRyCWP4fjx45w9exav10t5eTlr1qyhsrKShx9+mJqaGtra2jCZTFitVqqrq6mqqqKtrY2tW7fy1a9+lQMHDhAKhcjlFm6osF7g3tjYyIoVK4wUrq6uLl5++WVefvllQqHQZXcm9J+DC8/IFgoF8vk8yWTSaGEshHibpmlfAb5yE57naxSDioV4rB9xjSlXl7jvV4GvLsRxCCHEtZJg4xZms9lwu92sX7+eFStWcPLkScbGxqiqqsLr9dLY2AhANBolFosxNTXF0NAQY2MXD5G90kJUTy1qaWnB7/ezevVqWltbqa2tRVVVJicn6enp4ZVXXmFwcJBMJrPkC9vLPb+eVlQoFEin05hMJqLRKNFolIqKCqODlaqqxmI9HA7T3d1NOByekzq2UMrLy/H7/axcuZLW1lajtfDJkycZHBy8attai8VCWVkZTqdzzuXZbJZEImEEm0v9mQhxE/ypoih/CnRrmrZ6qQ/mVqYoykmgY6mPQwhx65Ng4xbmcrlob2/noYce4qGHHuLZZ5+lp6eHDRs2UF1dzcaNG9E0jZmZGYLBIOfOnWPfvn384Ac/uOixrpTPr3dJ+uAHP0hHRwf33Xcfbrcbl8vF2NgYP/jBD9i3bx8//OEPjVSu5Upv65vJZAiFQoyPj+NyuXj00UeN+Rr6lPFcLkcikeDgwYN8/etfJxgMAswZ/Hcj9NSndevWsWbNGu677z4qKyvJ5/NMTEzwox/9iN7eXkKh0GWfS1EU7HY7ra2tVFVVzTmuaDTK2NgYgUCAeDwuwYa4nfUDfzbr++klOo7byT8AlbO+71+i4xBC3OIk2LgF6Z2HVqxYwd1338369espKytjz549RktXh8OB3W4nn89jtVopLS3F4XAwMzNDMpk0Hks/+93b20sikbioAFpRFGpqaqirq2PDhg2sWrUKt9uNxWIxUnTGx8cJBoM3PE1bL9DWi8ytVivJZHLO8S40TdPIZrMcPXqUeDzOzp07KSsrw+FwzCm81gvN9fkWCxVsmEwmqqqqqK+vN4YPjoyMcObMGQYGBggGg1cMNBwOBxUVFaxdu5b6+vo5xxWLxZicnCQcDpNKpSTYELet88XNn1niw7itaJr2D0t9DEKI24MEG7cgi8VCeXk5HR0dPPzwwzQ1NeHz+diwYYOxoNQH3GUyGaOY2+FwYDKZ8Pv9xm3i8bhRUK3XfOiPoXedam1tZfXq1WzZsoWamhpjWnc2myUajTIyMsLMzMwNF6CbzWYsFgsulwu73U5JSQnj4+OLHmxkMhl++tOf0t/fj8lkYs2aNdTW1hrF8HqhvV4gDhhpVjdCD6yqq6tpa2sz0qCOHTvG0aNHGRwcnDPJ/EJ6u9z6+np27dpFbW3tnNcVCoUYGBgwAkwJNoQQQghxs0mwcYtRVRWv18sdd9zBpk2baGlpoVAoMDIyYhQInz17lpmZGYaGhkgmk4RCIWPw2+rVq9mxY8ecFrWpVAq/38+ZM2f40pe+RC6XQ1EUvF4vXq+XrVu3smnTJkpLS7FaraiqSjabZWpqit7eXk6fPs3k5OS8F7OKomCz2ejo6KC6upqVK1dSVlZGU1MTe/fu5Uc/KtZC6rUM+g6K3mFrvu+jvkPkdrvZvHkzlZWVRlct/bhUVaWsrIzW1lZjang8Hl+QNrJ6QNXY2EhbW5vRQevIkSP09PSQTCav+DwXzuZwu93A253BotGoseMkhBBCCLEUJNi4xaiqakyZXr16NX6/n/HxcYaGhow2tPv372d4eJiTJ08Sj8dJJBLGWfRPfvKTvOtd78Jmsxln7QuFAlarFZ/Px7/+67+iaRpmsxmv10tTUxOrV6+mo6MDl8tlDOzT6wr6+/s5e/YskUhk3q9JT1dasWIF9fX13HPPPdTU1NDY2EgoFOKNN94AMFrOqqp6w7UheppWaWkpFRUVtLS0GDUP+uPqNRUejwe/34/X6yUWiy1Y/YPVasXtdlNXV2cUp0ciEfr6+uju7r5qWpoepDmdTkpKSrDZbMDbrY5jsRiRSOSSQwCFEEIIIW4GCTZuQSaTCafTaSy6u7u7eeaZZ4zC52PHjhEIBAiHw8YwO/0s/Q9/+EMGBwepqamhrKyM7du34/V6qaurA+C9730v2WwWq9XKpk2b2L59u9HiVn98RVGYnp7mxRdfpLOzk3A4TCaTmdfr0IutHQ6HcYa/tbXVSNW6++67aWpqoq+vj2AwSCwWIxwO89RTTxGNRo3uULNne+hBwqWCEVVVMZvNtLe3s2bNGhwOBw6Hg1WrVlFXV0dzc7Px3uqPZ7VaKSkpwe12L+h0dFVVsVqt2O12LBYL4XCY6elpxsfHCYfD1/QY+ucqhBBCCLEcSbBxi7Hb7bhcLtxuN6qqkkwmGRoa4vTp00bXodHR0Tk5+vriWNM0Ojs7mZqaYu3atVRWVtLS0oLVaqWqqgq/309bWxv5fB673c727dvp6OjA4/FgNpsJBAJGW9upqSn6+voYGBggnU7Pa5dBH0bocDjw+Xz4/X4qKyuNAnRFUaiqqqK8vJzS0lKjfmNqaooXX3zRSA/SB9ddWDOiF0vrC3I9dcxms1FbW0tdXR12ux2Hw0FVVRVlZWX4/X6jRmP29HE95WwhJ6ObTCYsFgtWqxWTyUQymSQSiRCNRq+pVa1eU6IX1s8+Nj34kh0NIYQQQiwlCTZuMZ/85CdpaGjg4YcfZnp6mieffJLjx48TiUQYHx8nEomQzWaNdrV2ux2/3084HGZkZIRsNkssFjPSd1pbW/H7/cashk9+8pMARvqQ3W4nEokQiUR4/vnnGRsbIxqNMjMzw4EDB4hGo/OuX/B4PJSXl/O7v/u7rF69Gp/Ph91uN9KDXC6XsWC22Ww0NzcbNRPJZJJoNAoUuy5NTEwQDAYJBAIMDg4SiUSMuoyKigrq6uq4++67jcV5Op0ml8vxcz/3c8b8EH0YYqFQIJVK0dPTw1e/+lWCwSDBYJChoSEikciC1GuoqkplZSXr1q2jrKwMq9XKzMwMvb29zMzMXNQV7EKKomC1Wlm9ejUNDQ3Y7XZMJhOAUcsSj8eN+hIJOoQQQgixFCTYuMVUVlZSVVVFSUkJ0WiURCKBzWajvr7e2NlwOp1YrVbKy8spKSmhoaGBiYkJEokEXq8Xv99PRUUF1dXVuFwuLBYLgLEwn93atVAoEAgEGBsbo6ury5jbEAgEmJ6enlf6lE6vm6ivrzfqJUwm05w/erDhdDqx2WzkcjksFgtr1qwhGo2SzWYJBoOYTCZcLhclJSWkUikj0DCbzTQ3N9PS0sKGDRuMYCIYDBKNRqmpqaGiosJYrOuvOZvNMjMzw6FDh8jn88bMjRtt7wtvp09VVVXR0NCAx+PBZDKRTqdJJpPkcrmrBjR6nYvX68Xn882ZIK63JA6FQsRisQWddi6EEEIIcT0k2LjFdHd3k8lk2LlzJ6WlpXzkIx8hk8mQTqf53ve+R2dnJ01NTXi9XlauXElpaSn19fUMDQ3xwgsvUF9fT3NzM+3t7ZSWluJyuQCMInK9VkLTNIaHhzl9+jTf/OY3ee2115iamjJSpvQF+I0svJPJJMFgkAMHDnDmzBncbjeNjY3s3r0bk8k0p4uSnm6USCSwWq3s3LmTVCrFxMQE0WjUGIZXKBTYunWrEXQ5HA46OjqorKxkxYoVFAoFkskkDofD2MnQa0f0dsHZbJZ4PM7MzAwzMzOEQiGj0HohBha6XC6qq6t54IEH+NCHPkRlZSUWi8Vop3stz2Gz2SgpKaGjo4N169Zhs9mMYGl0dJQ333yTF198kZdffplYLCY7G0IIIYRYEhJs3GImJiawWCz09/dTUlJCaWmpUQDe0tKCqqq0tLRQUlJCZWUlTqcTt9tNVVUVGzduNGoT7HY7mqYZaVDBYBC73U59fb2R/6+fgdcH36VSqQUdDpfP50mlUgwNDZFKpWhqaiIYDDI9PW2kBOn8fr+x+2C1WrFardhsNvL5PC6XC5fLRS6XI5PJ4PF4SKVSxmyRhoYGSkpKsNvtFAqFObUS+uJ+divdXC7H9PQ0k5OTJJNJUqkU6XQaYEHqIFRVNdLEXC7XnF2Jq9F3fKqrq2lubqapqYnS0tI5NRt6MKjPWVnOE92FEEIIcXuTYOMWs3fvXjweD6dPn6apqYn169djs9mw2WzceeedPPjgg0ZalF6XkE6nKS8v5+GHHzYW1xMTE4yMjBgF44cOHaKlpYX/8l/+Cw6HA5vNRllZGevXr+euu+7CZrPx/PPPMzk5eU3Fy9cikUiQTCZ56qmn8Pv9/MZv/AaDg4O88cYbFxV6/+Zv/iYtLS3G4DtdaWkp8PaAwnA4PKeLlF6APns4nx5k6Avy2WlGZrOZSCTCU089ZcwP0YvEF9p8Cs7dbjclJSU8+uijrFy5kl27dlFSUjKnXsPj8dDS0kJlZSUej4dsNiupVEIIIYRYEhJs3GLS6TSKojAwMGDUS9TV1VFbW0s4HDbScfQ5GOl0mkgkQlVVFU1NTSQSCWKxGN3d3YyOjnL27FlCoRA9PT0AdHZ2UlNTQ01NDVDsfrV69Wqy2SxDQ0NYLBYGBgYWbPGqD6CLx+OcOXMGs9nMxMTEnE5aJpOJ8fFxKioqjAW63q0KMLoxWa1WPB6P8dj6joFeFD6bngYWi8WMnYtsNksgEGBkZISzZ88yNDRENptd1J2Baw029Nfd2NhIa2sr69evp729HY/HY8xX0W9ntVrxer1GnctCdtASQgghhLgeEmzcYvQFciwWo7+/n4MHD/LII49QW1vL4cOHAYwuRG+99ZbRoWnLli089NBDnDt3joGBAY4fP87Q0BChUIhMJkM+nzcW1zt37uTxxx+nvLwct9vNxo0baW9vJx6PU11dzfj4+IKfKY/FYnzrW98CmLO419Om7rrrLmPHxmq1UlNTY9Ra2Gw2HA6H0cZWD1Rmt8Kd3f5Xr82IxWKMjY0RDAbp7+9nYmKCr3/968buiF6svRzoqV933XUXW7du5f7775/TvWs2u91OZWUl5eXleDwepqenZYq4EEIIIZaEBBu3MH1Rfvr0aWMhDpDJZEilUnR3d5NKpUgmk5w6dQq73c7MzAzj4+MMDw8TiUSMQEPTNKPdqz73YvPmzbS2thr1Bdu3b8dut/Pss8+STCYX7Iz/7JkQF6Zn6UXTR48eJRaLGbUW+jwMRVFYtWoV69evNx4rm82SzWYJhUJAsYOXvsuRSqWM7lpnz5413od0Ok00GiUcDhsF1YsVaOh1Iblcbk53Kz1ocrlcJJNJksmkEWRUVlZSW1vL1q1b2b59u9HBSv/s9Mnq+i6O3sq4rq6O0dHRBa21EUIIIYS4VhJs3OIKhQKnTp2iq6trzuWaps2pMwgGg3R3dxv5+5daeCaTSXp6epiZmWFgYICZmRne85730N7ejtPppKOjA5/Ph9vtJhgMLmiwMfvrpa57/vnn+elPf0pJScmctCGAD3zgA7S1tRmLd33S+LFjx9A0jfe85z3Y7XYAxsfHef311zl58iRHjhxhcnKSVCpFVVUVAOFwmHQ6vSg1GrNfUz6fJ5PJXBRsuN1ufD4fqVSKTCaDzWbD4/HQ0dHB6tWrufPOO2lpaTE6T+kF4HpL4NnBht5a98SJE0QiEQk2hBBCCHHTSbBxG7hwcvblbqMXdl/utvp1druduro6mpubqampQVVVcrkcdrudsrIympqaSCaTjI+PL/oCVi/i1gft6TM0oJhOVVZWRldXF//8z/9s7GhEIhGjy5WiKExPT6OqKsFgkFAoxNDQEGNjY0xOThpdnMbGxoxuVovdvSmdTjM9Pc3x48fx+/3cd999VFRU0N7eTnV1NdXV1cYui9Vqxe12U1tbS1lZGdXV1cZuTz6fJxAIGHU5FRUVOBwOY8ZIVVUVtbW1RoH8jcxEEUIIIYSYDwk2bhNXW/RfuNNxpdvpBcb68EBFUcjn81gsFlwuFxUVFYyNjTE+Pr5Qh3/V49aPPZFIGNfp6V6nT5/m9OnTxjC+VCplBB2qqjI8PEwmk6G/v98ISHQ+nw+bzUYgELhpdQ25XI5wOMypU6ew2Wxs2LCBsrIyysrKKC8vp6amxtj50NOozGYzFovlki16o9EoExMTqKpq7NCYTCZ8Ph+VlZWUlJQwPT19yRkeC9HKVwghhBDiciTYEHMUCgWmp6c5dOgQK1eupLGxkaqqKmw2G11dXfT29nL8+HHGxsaWfJGaTqcZHh42vtePRx/uVygUUBSFc+fOzdkhmS0Wi5FIJG5qIbi+wB8eHkbTNA4fPozFYjHqY/RBg3odjqIoZLNZEokEQ0NDTE1N8bOf/YzJyUkGBwdJJBKkUinWrl3LAw88wPr162lsbGTVqlU0NTWxa9cuwuEwp0+fNmpBMpkMiUTCSCeLRqOy8yGEEEKIBSfBhrhIMplkdHSUyclJAoGA0XJ2dHSUrq4upqenicViS32YFAqFKx6HniIVDocve5ul7DYVjUYZGRlhcHCQ+vp66uvr5wz4m91BK51OEwqF6Orqoru7mx/84AcMDQ0RCASMAGp6ehqHw0FpaSnV1dU4HA6jBiSbzVJRUWG0Po5GowSDQYLBIKdOnZqzYySEEEIIsVAk2BAX0VOQxsfHOXv2LI2NjXNayt4qlvvkbH2H4emnn+bkyZMMDg4aNTKz5fN5Tp06RU9PD2fOnGFkZMSYoTL7NU5MTPDiiy+Sy+UYGRlh5cqVlJaWGhPG9Rkkg4ODjI6O0tfXx9mzZwmHw7KrIYQQQohFIcGGuIheD5DJZIjH44vamemdTC/s1+eWrF27lnQ6PSfY0FOu+vv76ezsZGBggFAodMkp7ul0mkAgYKRXlZaWYjKZcLvdxjR1s9lsTJWPRCJEo9E5HbGEEEIIIRaSIosMIYQQQgghxGJQr34TIYQQQgghhLh+EmwIIYQQQgghFoUEG0IIIYQQQohFIcGGEEIIIYQQYlFIsCGEEEIIIYRYFBJsCCGEEEIIIRaFBBtCCCGEEEKIRSHBhhBCCCGEEGJRSLAhhBBCCCGEWBQSbAghhBBCCCEWhQQbQgghhBBCiEUhwYYQQgghhBBiUUiwIYQQQgghhFgU5itdqSiKpv9dVVUaGhp44okneN/73kdbWxuVlZWoqoqiKBQKBePPBY9RfCKzGUVR0DTtktfncjlyuRzZbJbBwUECgQDd3d1885vf5JVXXrns44vbi9PpxGq1omkauVyOQqFAOp1GURSsVivpdHrJfwbcbje1tbWYTCa8Xi/bt29ncHCQvr4+Tp8+TT6fX9Ljuxyr1Yrf7+e3f/u32b59O8FgkIMHD/Lqq6/y5ptvkkqllvoQxRVomqYs9TEIIYQQ1+uKwcZsmqaRTqeZnJwknU4bQYOmaWiaRiaTIZ/Pk06n59xPVVVUVcXhcGAymYzgQqd/b7FYsFgsOBwOVq9eTTabpb29nYMHD/LKK6+Qy+Vu9LWKW0B5eTler5dMJkM6nSaVShEMBjGbzXi9XoLB4JIviisqKrj33nvJ5XJUVFTw0Y9+lFdffZWDBw/S29u7bIONXC5HOBxm586d7N69G5PJxNq1a6msrOT06dNL/r4KIYQQ4vZzzcHGbIODg3g8HhwOB8lkklgsRqFQIB6PMzo6Omf3wmq14nQ6aW9vx263Y7FYUNW3s7f0IGP2ZbP5/X5aW1s5c+bMsl3EieujqirV1dWYzeaLPvf77ruPuro6EokE4XCYUCjE8ePHWbNmDY899hidnZ0MDw/z/PPPE4/HSaVSF+2WLbbm5mbe97730dLSgtfrpby8nL179/LGG2+QzWZv6rFcK0VRKCsr484778Tv92MymYyTBMlk8qa/h9equroau90OQKFQIJfLEQwGjV0vfZdruR6/EEII8U53XcFGNpslFApx9uxZCoUCiqIQi8WYmJgAIBKJMDw8PCcosNlsOBwOCoUCTqcTm802Z5Hp9XrxeDx4PB5jl0NRFBRFwWQyUVpaSnV1NX19fRQKBVlU3OIURcFsNtPY2IjD4cBqtc65vqOjg5UrVxKPxwkGgwSDQfx+P83NzWzfvh2fz8eZM2fo6+sjHA4bAW4+n78pPxt2u52KigoaGhqor6/HbDYTCoUYHR1lYmJiyVO8LkdRFEpLS9m5cyder9dIaUwkEoyNjS3qzqHFYmHVqlXG734ikSCbzZLJZPB6vdhstsvet6GhwQg29DTLM2fOkE6nSafTJBIJUqkUkUhE/m0QQgghlqHrSqMKBoM8//zz/OxnP8Nut1NWVkYymSSZTAKQyWQu+k/fbDZjNpuprKzEZrNhs9nweDyYzcWn3rJlCxs2bOAXfuEXsFgsQPHMt8ViwePx0N7ezurVqzl06BCJREJ2N25hJpMJk8lESUkJ999/P16v11hI6h555BGqq6uBYnCbz+ex2+1GINrU1MSOHTtwOBxks1kSiQR/+7d/SyAQWPRdBUVRaGxspLm5mbq6OhwOB8FgkF//9V/n2LFjTE5OLurz34hCocDGjRv51Kc+ZdRPmUwmurq6+Jd/+Rfjd3ihmUwmqqur+dKXvkRFRQVOp5ORkREikQhjY2Ns27YNn8932fvPTrvUU+sOHTpEOBwmkUjQ2dnJ0NAQe/fuNXY7hBBCCLF8XNfORqFQIJPJUCgUSKVSJJNJo7AbIJ/Pk8vl5gQb+XyebDbL5OQkJpPJCCTKysp48MEHKSsrI5/PE4vFjOJgRVFQVRWr1UpHRwfZbJaf/OQnTE1NEY1GF/YdEItK36XSF7dut5u6ujoeeeQRPB4PJpPJuK2qqpSWlhq7XhaLxajzmd1owOPx8MADDxCNRgkEAlRUVJBOp4lEIou22FQUBYvFwh133EFHRwc2m804zunp6YtqlZab9evX09raauwOZrNZ3nzzTY4cOYLVal20dLSSkhJqamooKyvD5/PhcDiw2WxGTZbP5zNOMlyNnja1e/duY5fD5/Nx7NgxXn75ZWkgIYQQQixD112zoQccmUyGRCJx1dtrmkahULgoSFBVldraWlRVJRaLMT09jd/vx2w2GwtMs9lMdXU169evp7q6mng8LsHGLUQPGPWgQf88V61aZaTVzA4kLmweoNNT9lRVNQLWxsZG4vE4LpcLv9/PzMzMov5s6MFGW1sbzc3NWCwW43chFost21oNKB673j0un89jMpnIZrM899xz9PT0zAn4FprD4aCsrAy73Y7NZsNut2O3243A5nKf+ZU4nU6jMUUwGGRmZgaLxSIF7kIIIcQyNK8C8YUwNTXFZz/7WWMRd/DgQR588EHe//7343Q6jQWQzWajtLSUtrY2QqGQUR8ilr/GxkbuvPNOWlpa8Hg8uFwu7rvvPurr6+cEGvqC88Iz6yMjI3R2dlJTU0N5eTk1NTVGq2VVVTGbzcYi9nINBhaaHgRbrVbeeOMNXn75ZcbGxhYtDWkhaJrG/v37KSsrw2azkU6nCYfDZDIZY3doseod9F3QUCiEy+XCZrNd1JXuep9b/5kpFAqEw2Gmp6eJxWKSYimEEEIsQ0sWbOTzeYLBIBUVFZSVlZHJZIyzw5c625nP5yVF4hahqipNTU1s2LCBlStXsnv3bkpKSow5D2az2Ui90zSNiYkJstmssQOi6+np4a233mLFihW0t7fj9/sBjJkbZrMZh8NxUwINPcDRaZpGNBolGAySzWaXdXGyXhxeUlKCoihEo1FGRkYYGBhgfHz8omO3WCwoikImk7nh506lUkxMTPD666+zceNGWlpacLvdcwrU9bbZen2XXntxYfCgp+JVVlbOmc+jv//L+TMQQggh3qmWLNjQh7a1tbWxevVqAFwul1E4fuHt9IJQcW0ut1twM5jNZu677z5WrlzJ1q1b2bFjB06nE8Co9dEXkoVCgX/5l38hGAxSW1uL3W43OlR1dXXR2dnJ5OQkqVSKtrY20uk0ZrOZ8vJyrFYrbrfbaJ2sn/FejNes72joAYemaUZzhGw2u6wDYVVV2bhxI62trUBxx+j1119n7969F6WeKYqCx+NBVVVmZmZu+L2MRqNkMhm+9KUvce+99/Lggw+ycuVKzGYz2WyWnp4eRkdHicfjlJeXU1paytTUFMlk8qK0KEVRcLvdPPHEE1gsljmDHyXQeGdTFOVe4KezLurWNG310hyNEO88iqL8PfCbsy76M03TPrNEhyOWmSULNhRFwWazUVdXR3t7OwA1NTVGpxydfoYzHo8v67z45cJiseB2u3nsscdwOBw8/fTTTE9PE4/Hb8rzq6qK0+lk8+bNrFu3jubmZqPd6RtvvMGxY8cYGRkxCpILhQLHjh0jk8lQWlqKzWajUCgwMDBAJBIhkUhw7tw5o6bHZrPh8/koLS01znR/5jOfYe/evXzhC18wApOFXvy7XC6qqqq4//77qaqqIhKJcOjQIY4dO3bT2u7eiF/4hV9g27ZtqKpKW1ubsdjXd5jMZjPvfve7+e///b9z5swZY8fmxz/+Mfv27Zv38+bzeVKpFKdPnzZmpujpcNlsluHhYcLhMOl0GpfLhdPpJBqNGq1xZ7PZbNTX1/Poo48a9+/v7zdm+yz3z0DcFAeA/cD0tdxYURQX8AjwMLAFaAAKQDfw78DnNU278S2+i59XAX4C/Nz5iyyapl1T/2lFUVYAjwIPAiuAKiAIvA78raZpP73EfeqBjwGbgM1AK6AAKzRN672hF/P2c/wH4F+ucrOCpmlXLRKb7+eiKMqvAzsovs71gAP4S03T/vhaX8e1UhTFD3waeD/QBKSBc8BPNE379A0+9meAP73Kzc5qmtZ2DY9VDnyQ4nu5HqgDMsAJip/Xv2iadtF/mIqi2ICPU/y5aQXswBDwAvBZTdMGLrjLsxR/75rP30cIw5IFGyaTCZ/PR1VVFS0tLdhsNvx+v3GGWqcvVmafDReXZrFYqKmpYfPmzdx55504nU5mZmY4evQoAwMDJBKJRV+QmUwmrFYr5eXllJSUYLfbyWQyhEIhXnvtNU6fPs25c+cIhUJGapx+dn1mZgZVVSkUCgwNDRnHmkql6Ozs5ODBgzQ0NBjBKRSDm+bmZlatWmXU+syn6PhqLBYLXq+XsrIyLBYLgUCA4eFhhoaGlv2uht1up7a2Fq/XCxQ/I322haqqqKrKunXr2LlzJxs3bsTv9xMOh8nlcvT09PDaa6/dULcq/TPO5XJMT08b6Vz5fJ5wOEwymaRQKGCxWIwg6FJpk06n0+h2p/+7MDExwdTUlAQaQrf/Os+m3gV8HQhQ3Bn5PlBGcQH518CjiqK8W9O0he4+8J+AdwEpiou46/EXwC8Apygu8ALAKuDngZ9XFOW3NU373AX32Qb8v4BGcUEcBnzzPfjLOAr82WWuuwu4D3juGh9rvp/LZwEvxeBrFLjqYnw+FEXZDDwPlFMMGr9P8XNsBZ6gGITciP1XuO79FAOwa30vPwT8H2CM4ns5SDFAfRT4EvBeRVE+pM36R1RRFDOwF9gDdFEM8NLAduA/A7+iKModmqad0u+jadqzwLPndxkl2BBzLEmwoaoqXq+X+++/n/e9731s2bKFkpKSi3Lv8/k8k5OTnDp1isnJySVJo5pdwLycF5X6DIiPfvSj/OEf/qFRQPue97yHb37zmzz55JPs27dvUXeH9N0ql8uF1Woln8+TSCQ4evQonZ2dfO973zOGsU1OTpJOp6/peFKpFCdPnuTUqVN8+tOfnhNs3CxOp9OoGQkEArz00ku89dZbDAxceHJnefF6vdTU1DAyMkJlZSV+v5/+/n7eeOMNo2Db7Xbz3e9+l+rqahwOB+3t7RQKBSKRCJs2beLkyZMcOnTohn929LSzQCBwyeuvNliwtLSU1tZWFEUhmUxy5swZDh48yJEjR5b176ZY1saBjwDfmX2mXFEUD8UF3x0UU0M+u1BPqCjKKuB/Ulw0f5jiWfHr8WPgf2qaduSCx72H4lnn/60oync0TRubdfVbwN3AMU3TIoqi7AfumedLuCRN045SDDguoijKa+f/+sVrfLj5fi4fBk5rmjZwjTst101RlFLgh4AV2KNp2usXXH9tvbyvQNO0/Vwi4FAUxQT8+vlvr/W97KEYiP5o9g6Goij/DTgIPEYx8Hhy1n0+SDHQ2As8cMH9/gz478DvA792jccg3uFuTguf8/S8d5/PR3NzM7t37zYmSc8ONAqFgjGwTU+JWcgzl/rZXp/Pd9W2n01NTWzcuJH3vve9NDU1GbUHy5HeGlZvD5tOp3nyySfZv38/p0+fXvSdIU3TSKfTBAIBvvvd7/Lcc8/R39/P008/zXPPPcfQ0BDj4+PMzMyQSqWu63j0NBm9TbJeC6BfvlhntlVVpaysjA984AN86lOfwuFwGOlfuVxu2e+2rVy5kkcffZTNmzdTVVUFwNmzZzl06BAAO3fu5GMf+5hRAwNv73bY7XZaWlpYu3at0T1sKehDQfWaD03TmJ6e5rXXXmN4eFhqucS8aZp2VNO0f7swJUfTtChvL2TvXajnO3/G+GsUdxeuliZzSZqmfeXCQOP85XoKmZXiYnz2dcOapv1M07TIfJ7zRiiKsg7YBYwAP7qW+8z3c9E07ceXSO9ZaL9LMRXpjy4MNM4fw2Lmez8E1AOva5p2/FruoGnaPk3TfnhhqpSmaePA/z3/7b0X3K31/NcfXXg/4Afnv1Zc81GLd7yburNhNptxOp1UVVXR3t7O9u3b8fv9xiJHXzTqdRqxWMzoVrOQC0qTyYTL5TKGwV2ubamiKLS3t9PQ0GAUJ+spXbdC2kYikeCrX/0qp0+fvmktg7PZLOFwmCeffJI9e/bg8/l47rnnOHfu3A09rt4NKplMEgqFiEQixuyOK7XPvVFms5nGxkbe/e53c9dddxGLxYzrboXCZL39sD7xHGB4eJi+vj5UVWXlypU89NBD2O32OSloekvq8vJyamtrjQ5VN/v16p3HGhoa2Lp1K3fffbex43ns2DGmpqZIp9M31EpXiMvQF43XVEtxjf6YYs3Ebk3T0osQwC/GMd+o3zj/9Z81TVuIszNL/Rp/CcgDX1MUZS3wbsAJ9AE/1jQtdqU736D/eP7rte5qXM3l3svO81/fqyjK310QcLzv/NcXF+gYxDvATQs2rFYrzc3N3HXXXezYsYPVq1ezdu1aYxGjd5w6d+4ca9asMXYc9u/fz969exkbG1uQFCBFUaiuruZd73oXjz/+OH/yJ39CZ2fnJVM4FEVh48aNbNu2jXvuuYeysjKqq6v593//92V5Rnv2YlHv1DMzM7Mk060zmQxvvvkmZ8+eZWxs7Op3uApVVbHZbMa8DZfLZbTKdblceL1eI9d/IT+b+vp6vvGNb1BXV7do9SCL6eTJkzz55JNs374dl8sFwBNPPMEdd9zBX/3VX9Hc3GzUUMym70Lqnavi8fhNT1VSFIWSkhJaW1v5vd/7PXbu3ElVVRXd3d288sorvPLKK2SzWVwul9E+W58hstwCQb2ZgV5vIm4JeorIjxfiwRRF2Q78EfBXmqa9tRCPecHjN1Fc+CaAlxb68edDURQHxXSoAsX6gIWwoJ/L9TifQtVGMTXpM8DvUCy0180oivIr5+sXFvq564D3Uqy3+dYCPJ4Z+JXz3174Xv4I+B7F9KoTiqK8SLGofCtwJ/B54O9v9BjEO8eiBxv6IsbtdlNXV8eKFStYvXq1cbZUVVVjUZDL5Ywz15qmMTg4SF9fH319fUZh6EIcj8ViobS0lKamJioqKnA4HJedPq2nW3m9XrZs2UI2m+W73/3uslsw6Ok+Ho/HqNfQi2iX6liTySSTk5MLEiSaTCYcDgc+nw+fzzdnkF8ul8NqtdLR0cH09PSCDtjT0/70oBiK7VxPnDhxS0yzDwQCdHd3Mzg4aNRKORwOSktLcbvd+P1+fD7fJWeVaJpGWVkZTU1NWCyWBZm7cTWzd1ZMJhNut5va2lpWrlyJz+fDbDaTTqexWCysWrWKpqYmXC4XDQ0NxGIxYrEYR44cIRgMEgqFbkqApO+6eTweHA4Hfr+fyclJ4vE4JpOJ0tJSKisrUVWVWCzGyMgIfr8fp9NpzBVJp9P09vZetW5F3ByKovwn4D0UaxC+vACP56CYPnUK+PMbfbxLPL4N+DfABvwXTdOCC/0c8/QExUL0H2maNnSjD7bQn8s8VJ7/2kaxUPrTwL9SDDg+AvwP4ElFUbZomnZ6gZ/744AJ+LqmaQuRO/pXwDrgWU3Tnp99haZpmqIoj1OszfgTYO2sq/cC31igXSrxDrGowcbs2Qfl5eW0tLSwa9cuWlpa8Hq9mEymi85AappGX18f09PTHDhwgNdff53h4eEFXTSYzWZKSkqorKykqqoKn8932YWjnsZhs9nYuHEjNTU1/PEf/zHpdHrZnDnVA6jW1lYjLz+fz5PJZEgkEku2gMnn8wu28LdYLPj9fqqrq42Fmp7WE4lEsFgsvPvd7+b48eMEg4v7/+zAwABf+tKXll3AeSlTU1OEw2Geeuop7rvvPu6+++45C/mGhgZjITybfga+sbGR97///TzzzDOL3s1MrxWB4u+ozWajurqauro6mpubjW5jmqbR0tLCQw89hNvtpqKigjVr1hAKhZiZmeFb3/oWx48f5/Dhw4s+cFHfAbJYLLS1tdHa2srWrVt5+eWX6enpwWazsXPnTjZv3oyqqgwPD7N3717uuOMOqqqqcLvdxONxQqEQn/vc5+ak6YmloSjKo8DfUixSfmyBcvD/F8U8+B0LndN/vmj4axQLer9FsfB8udDTfv7xRh9okT6X62Wa9fWzmqb971nX/bWiKDXApyjuePwGC0RRFJW3d3RuOIVKUZTfAn6PYqepj17iejvFIOq9FAvxf0Bxx2wP8DngpfMdrH5w4X2FuJRFCTbMZjNut5vt27ezdu1a2tvbcTgcrFixgvXr12O32+cM79MLf4eGhvjZz37GuXPniEajTE1NMTk5uaCL5UKhwNjYGN3d3Rw+fJjW1lYKhQLf/va3Lwp8CoUC3/rWt0gkEtx9991Gy9CrFZXfbPqC64knnqCjo8NYTFqtVpxOJ5HITa8JXDR9fX1UVlaycuVKcrkcmUyGffv2cerUKY4dO3bZTkfzoZ917uzspKOjg4qKCo4fP87Ro0cX7DkWkx7o6z/X+le9NfEnPvEJqqqqrpgadurUKb71rW/R0tJCZWUl8Xic0dHRiwbuzee4zGYzHo8Hu91OSUkJbW1tRkcsi8WC0+mksbGRtrY2I21OVVXWr19PLpcjl8sZTRFsNhtlZWXU1dWhaRoul8uYz7KQQWFlZSWPPfYYGzZsoLGxkbNnz+L1eqmtrTWO3e1286EPfYh0Ok0ikcDr9Rqpatlslo997GPYbDZjYvrBgwfZu3fvsjl58U6mKMojwDeBSeBdmqadXYDHvIfigu0z5zs2LZjzgcbXKbY3/TbwEW2Z/CCdr2e4Axim2KL3Rh7rERb4c5mn2WeynrrE9U9RDDZ2LPDzvhdo5DoKwy9HUZTfBP6O4i7buzVNu9R/mv+V4s/Ub2uaNjtQfO78jsfR848hwYa4JgsWbOhnJfW0k6qqKjZt2sTOnTtZs2YNUGzD6Xa7jcW63tEnm80yNjbGmTNnOHPmDMeOHTMWyIuRK55MJpmammJwcJDKykpSqZSRJnLhv9N9fX0MDQ2Ry+UuKkheJv+mY7Va8Xq9tLS0UFZWNue6W63G4HLMZjMOh8PYrdEVCgW6urro7u6ms7PzhhbBF9IX6sPDw7S0tJDP5xkdHV3QgGYxmEwmLBYLLpeLTCZjpOnond1UVcVqtdLY2DgnPWw2fWcjFosRj8epqqoiGo2Sz+fn9TOlpxrpQx8tFgsej4f6+npKS0vx+Xy0tLTQ3NyM2+02AmU9zWv2753L5ZoTPOn1SbN3bBZj5oo+G2jbtm3ccccdVFdX09LSgsvlwufz4XA4jOesqKgwurPp/27Mfi/0nVGr1UplZSUul+uSqWzi5lEU5UPANyieOb9P07QzC/TQmymm2fzZ+bahl5I9/7O6+VoDkvM599+guCj8BvAryyy1ZUEKwxfxc7lumqaNKYoSAUqA0CVuogcjjgV+6gXZIVIU5XeAvwFOUgw0Ji9zU70I/KIBkZqmHVMUJQA0KYpSrmnazI0ck3hnuOFgQ/8P3ufz4Xa7qays5F3vehcdHR088MADuN1u7PbizKLZHaU0TSOVSjEyMkJfXx//+I//SE9PDwMDA3NSlBYj53p216u1a9dSUlKCz+czBozNps8bCAQClJaWAhhnWJdLb399IT57xyifz5NOp4nH47dEus+VqKqKz+dj69atbNu2jZUrVwLF1221Wunv76ezs5PBwcEFfa16Klo0GiUQCODxeG6JKfbV1dU0NDSwe/duzp49yyuvvMLg4CBnzpzh7rvvNnbnLtfGWV+8Dw8Ps379elpaWvja177G8ePHOX369LyCbJvNRklJCWVlZWzZsoWamhrWrVvHxo0bKSsrw2az4XQ6jVqc2bUbl5p1k81mjc5wkUiE48ePE4/HjXqazs5OAoHAgv486CdQ3vWud1FZWYnT6TQGJV5If4/0f/tmX14oFBgZGSGTybBy5UpaWlr4+Z//eb785S8Ti8WWzUmMdxJFUX6JYtrICAt/5vwk8M+Xue4XADfF+gMNuKaFm6IoVoo7GR+geNy/eokWpUvmfBrORykWhl/utV/L4yzm5zJf+yhON1/H212bdOvOf+1fqCdTFKWW4vTvMMXPfL6P82mKdRpHgfs1TZu+ws1t579e1N72fH1QyflvF7+QT9wWrinYUBTFWMhaLJY5l1VWVtLQ0EBLS8ucHY3KykpKSkqMM5L6ffQzpqlUihMnTvDWW2/R1dXFiRMnCAaDpNPpm9pJprS0lHw+j9PpvGy+dDAYpK+vj3Xr1mGz2fjd3/1d9u3bxwsvvHBTjvFqampq2LBhAx6PB7PZjKZpTE5O0tPTQywWuymFvYvJZDJRUlJCbW0t7e3t1NbWztlZ2rx5M6lUiq6urkV5/sWe5bHQ9CLkQ4cOMT4+bgxWTKVSmEwm46y8PmfmwvbBswue9cVxT08P/f391/0eWCwW3G43d911F01NTbS0tLBixQrKy8upqKigrKwMq9Vq7Hqoqko+nzfSo2af7deHAs7MzDA5OUlvby/BYJBwOExfX58xMHJwcJCJiYkFr9eYvSOqn0SZnVKpaRqJRILu7m7i8TjpdJpYLMbq1atpbGw0gqHh4WFee+01stksH/zgBxkcHDR25W6Vn7HbiaIoH6O42B+guKBd0DkNmqa9yGXahCqK8nMUg43f0DTtmvKFzy/2vkdx5sI/A/9xOQUa530IKAWemW9h+GJ/LjfgCxSDjT9WFOV5TdNCAIqi+CgWU0Mx5Wuh/DrFGpGvzbcwXFGUP6HYmOAQxSF9V9ue/xnFwOm/KYryiqZps1tafobi2vHN83NPhLiqqwYb+s6Fnlutt880m834/X5WrFjBqlWr2L59O6WlpZSXl+P1erFarcZ/xLP/A1UUxUjPeOGFFzh8+DDHjh1jbGxsTsepK6U/LNR/yHprzUKhgMvlmpPqMFsoFOLIkSO0tLRQUVHBRz/6URKJxJIHG8r5id16LYz+GjRNY2hoiGPHjhGNRpfNDsx8mc1mXC4XVVVV1NTU4PV6jUVwoVBg8+bNjI6OLtpC7cJFLyzvWQ7RaJRYLGa8J3qqWV9fH6+//jqf+MQnuOuuu1i7di1OpxOr1WrMutGDOEVRjHqfSCTCiRMnmJy83I775dlsNnw+H7t27WL79u2sWrXK6O41O4Urn88bJyLS6TRms/mihXw0GmVsbIyDBw/S1dXFwYMHmZmZIR6PMz09TS6Xm5NatdCfkR7EHT58mLq6Orxe75xdGP02Tz/9NKOjo4RCIQYGBvjlX/5lHnvsMYaHh+np6eGnP/0pb775JgAOh4P+/n7OnTtHPB5f0OMVV6coyrsoLmhViikjv3qJ/3tCmqb97U0+tCv5vxQDjWmKZ/z/+yWOeb9WnEJtUBTlK7O+XX3+6/9UFEVfMH5J07SXF+gYb2gexHw/F0VRPk6xNStA+/mv71cUpf7837s0Tfur+RyTTtO0FxVF+TzFblQnFUX54fmr3kdx4N73Ke7G3LDzheHXOzH8wsf4GMVAI08xiPitS7yX/ZqmfWXW938JvJ9iK+UuRVF+DCQpFojvOP/3357P8Yh3pqsGG3rx5q5duygvL6e6utpon2kymVixYgVNTU34/X4sFotxlnI2/Qd79n/+VquV1atX43K5WLlyJW+88QaRSOSK3Vj0xWUymTTOGt7oLsjl2n7OFovF6OnpIRwOU1ZWRklJiVHwuVSLTlVVaW5u5mtf+xqNjY2UlZUZKRv5fJ5AIEAgEFjWi+JrMbvj1NmzZ420mGg0yszMDKOjo/zhH/4hZ88u/O56oVDA4XDw7ne/G4/HY1y+3GclXC64zGazTE1N8fnPf55//dd/5Y477kBVVSwWC5/+9KepqqqipKSEcDhMOBzm1KlT/OQnP+HAgQPMzMwsWNCaTqfJZrOoqko2myWZTNLT00MoFCIcDhMIBFi/fj133nmnkVaVyWQ4fvw4zz77LD/84Q8JhUJGiqD+78Ji/6xns1lGR0f59re/jcVi4e6776ajo2NOUGSxWKitrWVoaIipqSnGx8f5m7/5Gz73uc+RzWbJZDLE43Gjtuiv/uqvjLq1WyFF7zbURHFBC293+7nQAMUuSMtFy/mvfoqtSS9n/wXff+wSt3n0gtvfcLChKMoaigv+GykMn+/ncicXv84N5/8AHKCYSnRDNE37LUVR3gI+STFdzESxs9P/Av5hAXeaHqT4XryuadqJeT6G/vNiotgl61IOAF/Rv9E0bURRlC0UW/s+DPwqxc9j7Pzt/qemaYuTSiBuS1cMNqxWq5Gis3PnTiorK6mpqcHj8RhnQsvKyoxUCP0M8OyUjNn0haO+wNEHilVUVBAOh+f8J3wpeqFrV1eXURB6o0wm01W7S2UyGaampoxidYvFQk1NDatWrWJoaIh0On3T28vq7XubmpooLS3F4XCQy+VIJBIMDw9z6NAh+vv7b+oxLTSLxYLNZqOyspK1a9eyceNG7Ha7keYTiUTo7u6mv7+fmZmFr1HTf171ANhkMhln4G9V+XyemZkZIpEIb731ltHaubu7m0gkgsPhYGJigsnJSY4ePcrJkyc5d+7cvH++9fbHfX19mEwmUqnUnGLoZDJJNBrl6NGjRupReXk5mUzG2AVNp9OcOXOG1157jZMnTzI2NmYUu99s6XSagYEB+vr6qKmpMQZMOhwONE3DYrHQ0dFBJpPB5/MZO0zT09OXTGecni6mTS+nhhPvJOfP5n5lCZ+/eR73uXeez3VTuoVoxfkSN/Rc8/1cNE37D8B/uJHnvo7n+lcWaAfjCs/xHDf+Xn6GYurT9d5vCvj983+EuCFXDDZKS0u58847+cVf/EW2bdtmpOnM3gnQA4vZQcKFqQWzaZqGyWTCbrezatUqo8jT7XYTjUbndBq6kN5Kcmho6KYOVEsmk5w6dYpgMEgul8PpdLJ+/Xo+/OEP841vfIOJiQnC4fBNOx7A6NSj12nox3nkyBH+7M/+jCNHjtz0Y1poHo+Hqqoq7rjjDt71rnfxxBNPGIGhw+EgFArx1ltvGV2SFprZbCabzbJv3z5jHstCdrtaStlslq6uLuN38cknnzSmvuvXT05OGt2s5kuvs/i3f/s3SkpK8Hq9RoMFKAbymUyGyclJo1PV7/7u7+L3+7FarUYTiU9/+tNGoLGU9M55+gyPdDrNli1bWLFiBVD8vdyxYwcbNmwgEAiQTqc5deoUhUKBqampy/6cSqCxYP5UUZQ/Bbo1TVt91VsLIRaEoih/T7HFsxAXuWKwoU/DbWxsNNpVXi6QmJ3zfbmdDX1XQ7+d0+nE4XDg9Xq55557jJSIC2UyGV544QXS6TSqqrJx40YqKyuN1Iv5nOW8MO3iwjaZ+muYPfxMHyQH0NjYyOOPP27URhw+fPimnmnV37fZtTGFQoFsNks4HL4tJhGXlpbS0dHB5s2baW9vx2Qykc1myeVyTE5O8rOf/Yz9+/eTTqev/mDzkMvlCAQCPPPMM8bP//j4+LxqF5arQqFAKpVi79695PN5I0DV27Yu1M90JpMxUp8u7C6lP9euXbvYuXMnjz/+OH6/Hygu7vWajMX6nOejs7OTkZER+vv76enp4b777jMmtDc3N2OxWKisrORTn/oUkUiEcDjMkSNHOHbsGF/5yleW+vBvR/3A7LayV+q0I4RYeM8y9/du/xIdh1iGrlqzoac86d1hrnS72fQi09nfz6acn3qts9lslz27l81mqaiowGazkUqlWLlyJWVlZaiqyunTp40F0/XQO2LpZxo9Hg82mw2LxYKmadjtdux2Ow6Hg6qqKtra2vB4PEZQ4nQ6aWhoYMWKFYTDYY4dO3ZTOxbpBbazpy4DRv737XCm1Gq1YrfbjZoUPa0mmUwyPj7OwMAAAwMDi1Y/oWkayWSSkydPGmku4XD4li+4n02vPxkamlfDmGtWKBSMXYzLqa6uZs2aNdTW1mK1Wo3f0VgsRiQSWVY1DYFAgFAoRCgUolAoGKl9NTU12Gw26urqcDqdNDc3GwMIq6qqMJvNfP3rX7/l0/GWG03T+plHqogQYmFomvYsNzi8Udy+rhhsTE5O0tXVxf79+3n88ccvGk4126X+47ye/0yvdFuLxcKDDz5ofK+f7RweHubzn/88P/3pTxkeHr7m59NTGs6cOUM0GsVkMrF7925KS0uNIuR169axbt06KisraWlpYcuWLcawMf2YVFXl/vvvx2w2s3//fsLh8E1bEF1q9+hmtgxebIqiMDw8TDqd5rd+67fw+Xzk83kikQjT09McOXKEkZGRi+aiLLRCoUAwGLz6DcUNm/0zncvlSCaTnDlzhoMHDzI5Obnon/X10IO0mZkZfvKTn7B3715jqKnT6eSpp55ix44dxskai8VCe3s7qVSKpqYmxsbGrpgyKoQQQtwurhhsZLNZpqenjQna8PZQKj1lJ5lMGoXTN9qlx+fzXTQBW3/O2fRUrkwmY6TVXG9gE4/HGR8fJ5VK4fP5ePjhh9mwYQNTU1NAMU2qtbUVh8OBy+XC7XbPadWZSqWIRCK8/PLLHD9+nGQyedPOeCuKQmtrKy0tLcauRqFQoLu7m8OHDzMzM7OszgLPh57Cpu8k6T9biUSCqakpDhw4QHd391IfplhAZrPZ6KiWy+WM+TYjIyNks9llu6OUz+eNf/dUVSWVSvGNb3yDM2fOsGrVKsrLy40p6eXl5Tz++ON861vfuuUbOAghhBDX4orBRqFQIBQKGf/Z64GGvsAPBoOMjY3R09MDvD3V93oWBbPPZuqzOi48Bn2i8eyp4qlUiunpaaODzfWKx+P09/cTi8Woqqpi27ZtbNy40QiqHA4HDofjsvePxWKcO3eOZ555hlOnTt20Hvn6Iry9vZ3Vq1fPmVOwf/9+Dh06xMzMzC1fs6G/TpvNRj6fN9LD9Ba43/ve92751yjmMpvNxu9cJpNhbGyMs2fPMjw8vOjBxkJ1g9JPxHzhC1/A5XLxoQ99iK1bt7J9+3Y2bdpEaWkpH/rQh3jzzTcZGhpa1i2UhRBCiIVwTRPEodiiUVVVrFYrwWCQUCjE0aNHOXPmDH19fUDxbOT1zr7YsmULra2tAJw6dYqRkZE516dSKVKpFENDQ2QyGdLpNJqmEQqFOH78OH19ffPquqTng4dCIRKJBB6PZ04NiaIoRmCl12Loix2Hw0E4HKazs5Pe3t5Fabt6OXa7naqqKjZs2MD69euNRVKhUODkyZOcOHHitqnZ0FPlvvzlL1NVVUVZWRmHDh2it7dXFmm3ofr6ejZs2EChUGBiYoKnn36a1157bVHrcnw+nzGsdHR01EjV0mswoFg75HQ6CYVCxnGYzWasVitVVVVEIpGL/g3QZ4P8+Mc/Nmo5Vq9ejc1mY82aNfzRH/0R99xzD3/xF38hQbMQQojb2hWDDU3TSCQSTE5OcuTIEerq6rBarYyNjTE9PW0EB7ODjVQqddWFrsPhMGo/JicnjdSJS0mlUmQyGcbHx0mn0ySTSXK5HJFIhHPnzs277Wk+nycej3Py5ElsNpsxPEw/o64HGtFo1DirnsvlsFgstLS0YLFYqKurM3Zdbhan08maNWtob2/H7/ejKIpxfFNTU0Sj0dsi0NDf11QqxZkzZwgEAng8Hjo7OxkbG7stXqOYy+Vy4fF4KBQKRKNRhoeHGR0dJRKJLMrzqapKe3s79fX11NfXEwwGiUajnDlzhtLSUqNNr94oYt++fUSjUcrKyvB4PLjdbtasWUMymSQQCNDf308ul8Nms1FdXU1JSQl2u52amhrjJAkUm2HozSUWYlaQEEIIsZxdNY1qZGSEyclJTpw4gc1mw2w2E4vFSKVSBIPBy7arvZKmpiZjIvORI0c4cuTIZW+by+XI5/PGoj+fzxu7J8C8z3jqHY3+6Z/+iQMHDnD//ffjdDqx2WxUVFQQDAYZHx+nr6/PmP+Rz+cpLS3lj/7ojygtLeXee+/F7/czOjo6r2OYj5qaGh566CE2bdpERUUFUJw/MjU1RX9//21TzKyno+RyOY4dO4bVasVmsxEIBG6bWRfibYqiUFJSgtvtJp1OEwgE6OzsNFIlF4PZbObxxx9n9+7dbN682eh2961vfYt169ZRX1+P3W7HYrFgNpv5nd/5HQYGBrjvvvsoLy/H7/eze/du48TJ008/TSQSoaqqiu3bt1NZWYnD4aC/v5+33nrLeJ2qqlJaWkpjY+OivC4hhBBiOblqGpXesnJqaso486/n0F9vYbZuamqKUCh0TbedXbMx++83Sn+coaEhQqEQk5OTxq6Gw+EwBggGg0Gy2SzZbBZVVamsrOTo0aNs3LiRqqoqysvLcbvdBAKBGz6mK1FVlSeeeILt27fzyCOPUFJSYqRPTUxMcPDgQSKRyG2ZkqF3DNNTU5ZrobCYP0VROH78OC+99BKtra3GQLzFbBFbKBQYHx+ns7MTq9VKfX09mqZRVlZGRUUF5eXlxr95iqLw+7//+6RSKcrLyzGbzZjNZjwej7E78d73vpd8Po/NZsPlchkd62pqarjvvvtwOByoqko6neZrX/sa+/fvvy1/X4UQQojZrhps6PUKC9mmcbm0fNQ0jWg0SjQaZWJi4qq1JvpCobe3l7a2Nqqrq3G5XNhstkU/VlVVWbduHbt27aK2ttYI+NLpNAMDAxw+fJhEInFbphfpwZ64vZ09e5aXXnoJm81mpE0uZjtnTdOYnJzEZDKRy+XIZDLYbDZj8KDZbDaev1Ao0NTUBLzdvGB2J7h8Po/X6zXq2vRdWD19yul0kslkjJ3Z/fv388wzz9yWv69CCCHEbNdcIH67u9Z0rEKhQDweJ5PJoGnaFWePLCQ9v7yyshIoduuZnJzkv/7X/8rx48fp7++X9CJxyyoUCrz11lv09PTw3HPPEYlE5nTBWwz5fJ7vfOc7RuBgtVpRFIVsNsvv/M7v8PM///OMjY2RTCaNwvFcLkcikeDOO++krq6OfD7P6Ogox48fp6enh7q6Oh599FGOHDnCqVOnePPNN4nFYiQSCQKBgLFDOzw8LDt0Qggh3hEk2LgOmqYZOxvj4+O0tbWRzWYXvTOSfia1qqoKt9ttdMrSC+dDoZAEGuKWp7fN1ptCzDdN83rM3jXTAwpN03j99ddJJBKEQiGjAQNgNI4YGhqiuroap9NptOgdGRmhvLyceDxOX18fk5OTnDx5knQ6bQQp+uuZT7tuIYQQ4lYkwcZ10DSNZDLJm2++aczl0FM9FpuqqpSVleFyuYC3Fz3xeFxSjMRtQV/436yZNReaHdi8+OKLvPjii1e8vaqqNDU1EY/HCQQCRsrVk08+Oae+TQghhHgnk2BjHjKZDIlEwphCvlitOWfTNI1AIEBZWRkWi4Wuri7efPNNJicnZVdDiCWgd+vTazZmByuSIiWEEEIUSbBxnTRNI5VKMTk5ycjICOFw+KbsLGiaRl9fH06nE5PJxOnTp+nu7padDSGWUCaTWepDEEIIIZY1CTauUz6fZ2pqin379jE+Ps7U1NSit6/UNI1sNsvTTz/N8ePH2bx5M08++SSHDx9e9Ja7QgghhBBCzJcEG/OQy+UYGBggGAySTCYXvYjVarVSUlLCo48+ysqVK6moqKC/v598Ps/09LTR4UYIIYQQQojlRIKNeSgUCgSDwZs2rdtiseByudiyZQt1dXXYbDZaW1sZHh7GYrFIEaoQQgghhFiWFBkqJYQQQgghhFgM6lIfgBBCCCGEEOL2JMGGEEIIIYQQYlFIsCGEEEIIIYRYFBJsCCGEEEIIIRaFBBtCCCGEEEKIRSHBhhBCCCGEEGJRSLAhhBBCCCGEWBQSbAghhBBCCCEWhQQbQgghhBBCiEUhwYYQQgghhBBiUUiwIYQQQgghhFgUEmwIIYQQQgghFoUEG0IIIYQQQohFYV7qA7hdKIqiLcJjomkL/rBCLBuapilLfQxCCCGEWDyys7FMmUwmTCaT8b2iKCjK3HWZqqqYTKaLLhdCCCGEEGI5kJ2NZWT2Tobb7cblcrF7924UReHYsWMoikJpaSnxeJxIJILb7cbpdKJpGplMht7eXlKpFIDsiAghhBBCiCUnwcYyoaoqqqqSz+cBqKysZM2aNfyv//W/UFWVv/u7v8NqtbJ27VpGR0fp6+ujtrYWv9+PqqokEgm+8IUvMDIyAhSDjUKhsJQvSQghhBBCvMNJsLFMmEwmbDYbHo8Ht9vNhz/8YdavX095eTmKovDoo49iMpkoKysjHo8TCoVwuVzY7XYA8vk8LpeLiYkJTpw4QTKZJJ1OMzg4yNmzZ5f41QkhhBBCiHciRdJtFsaNFog7nU48Hg/t7e00NTXx53/+5zQ2NmI2Xz4e1DTtonSpZDLJU089RTAYJJFIcOTIEb797W/fyKEJsWikQFwIIYS4vcnOxhKzWq34fD78fj8NDQ08/PDD7Nixg8rKyjkF4npKlF4Mrmka2WyWfD6P1WrFbDajaRoWi4W7776bbDZLLpfDarWyb98+4vE4yWRySV6juD0oioKqqlgsljlBsM1mw2QykU6nyWazJJNJqRkSQgghBCDBxpKzWq2Ul5fT1NREc3Mz9957Lx0dHXOCCk3TjFoOVVWNy7LZLOl02gg0NE3DZDJRV1dn3Hfjxo2UlpaSy+Uk2BA3RFEUrFYrNpsNh8NhXFZSUoLZbCYcDpNOp0mn08bPqxBCCCHe2STYWCKqquLxeGhububOO+9kz5497N69m4qKCiPQyGQy/Pu//zvBYJBVq1YRDAY5deoULS0trF+/nlQqRTKZJBQKkc1maWhoMAIPt9tNXV0dk5OTS/xKxe1AURR8Ph+bNm2isbGRtrY2amtrqa6uxmQyUSgUOH36NCMjI3zxi18kFost9SELIYQQYhmQYGOJqKqK3W6nsrKS9vZ21q1bR1NTk9FFSm9ne+DAAcbHx5mZmWFmZoYXXniB3bt3A5BKpUin0wwMDBCPx2lvb8dut6MoCpWVlTgcDgKBgJxlFjdMURQcDgdNTU20tbWxY8cOGhoaKC8vN3beXC4XHo8Hm81GPB6XVCohhBBCSLCxVFRVxe12s379ej72sY8ZaSmBQICZmRkmJiYYGRnh0KFDjI6Ocvz4cXK5HJFIhGeffZZXXnnF2AFJJpPk83lsNht2u52GhgYsFgtf//rXGR0dZWpqypi/sdj0XP5cLnfRdfpgwsu15J2dOiaWD0VRsNlsVFdXs2nTJjo6Oli1ahVerxe73U4ikTBS9FRVpaSkhEwmQzwel/bLQgghxDucBBtLwGQy4XK5jM5TJSUlxkC/aDTK2NgYJ0+eZGRkhJmZGSKRCNls1rh/KpUiEAgYBbt6vYZeIJ7L5YzHymQyNzWH3uFwYDKZjDPbswMHi8WCzWYjnU6TyWRQ1eIAez0IsVqtFAoFEomEcZ+rBShi8emfjcfjoaKiAq/Xi9vtNgrF9Z/BWCxGJBLBYrFgtVqJx+NzBlUKIYQQ4p1Hgo2bzGQyUV1dzbZt2/iLv/gLKisrUVWVXC5HLpfj9OnT7N27l7179zI0NEQ4HL5soDC7cFyXzWYZGhqaE4As9mJP35EwmUxs3bqVsrIyTp06RTKZnBMkNTQ00N7ezqlTpxgaGjIWrCaTCbvdzqpVq5icnOTll1+mUCgYqTs2m41wODznscTNYzabKS8vp66ujqamJsrKyozPTZfL5fj+97/PqVOnMJvNuFwuwuGwkRIoxGJSFOVe4KezLurWNG310hyNuFaKojwOfGfWRQc0Tbt3iQ5HCLFIJNi4yfSajFwuRyqVMtKN9AV7NpsllUoRCoWIRCLzOqN/qRSmxaLvPJSVlVFaWsrGjRspLy/H4XCQSqXIZrOYzWZsNhtVVVXU19djt9spLy/H7/djs9lQFAW73U5jYyPj4+MMDw9TKBRQVZXS0lJKSkro7OwkFAoBbwdZhUJB6lEWmR44rFq1ivr6esrKynA4HMbP62zxeJxgMDjn51qIm+wAsB+YvtY7KIqyH7jnCjdxaJp2Q3mo1/AcAF/WNO3X5/n4/wz82vlvV2ia1nvB9XcD/w+wGagBXMAYcAL4O03T9s7neS9xHI9TfJ2bgI2AB/g3TdM+cpm7nAL+7Pzf/3QhjkEIsfxIsHGTFQoFgsEgR48e5R//8R+57777+PCHP2zML9AniDscDiwWC3DpHYzlRFVVHnzwQbZu3coTTzyBz+fj3LlzpFIpozOWHlhYrVaGhoYYHR2ltrbWWLiaTCYcDgdTU1M0NzejaRqqqhpFyF/+8pd56623gGIwFY1GSafThMPhm7J7806kKIrRMe2jH/0oq1atoqWlxQiYL7wtFIPlsbExCTbEUtmvadpn5nnfP7vM5Qvxw/wVikHQpfxnoAx4bj4PrCjK+ykGGjHAfZmb3Xf+zxvAPiAONAI/D7xfUZT/V9O0P5nP81/gjykGGTFgGLji7pKmaaeAzwAoiiLBhhC3KQk2bgJVVVFV1Ugp0WdeTExM0NPTw7lz5ygrK6OkpITa2lp27NjB8ePHjbkF2WyWRCJxyUXecqB3zkokEqTTaXK5HG63G5vNRiKRIJ/PEw6HsVgsRj5/Q0MDbrcbq9VqBBZWq5XS0lJ27NhhXFZSUoLT6WTXrl1UVFQAxWBjenqaQCDAiRMniMVihMPhJX4Xbj96YbjX66Wurg6fz2fUZ+g7WvrtFEXBbDZjsViMOg1JoRK3khsIUq7lsb9yqcsVRVlF8Yz+BPCD631cRVEqgH8CvgVUc/ndk7+61OtTFKUOOAz8N0VR/kHTtLHrPYYL/C7FIKP3/LH89Mo3F0K8E0iwcRPYbDbMZrNRqJ3L5UgkEnR3dxOJRBgeHuaXfumXuPfee2lvb6e5uZkjR44Qj8cJhUIkEgnGx8fJZrNkMpmlfjlz6LsKIyMjeDwezp49SyaToaysjHw+TzQaJRAI0NvbayxAH3zwQRobG8nlchQKBaM+w2Kx4PP52LVrF8CcVJ1f/uVfNjpqZbNZRkZGOHv2LE899RQ9PT0cO3ZsSV7/7UxRFFwuFxUVFTQ1NeHz+YBibY5e3K/fTq+70etw9BokIcQV/cfzX/9F07T5FKV98fzX3wSevNyNLpcGpmnaiKIorwKPAK0UU6vmTdM0I7i4VKqlEOKdSYKNRbRy5Uo2bdrE1NQUkUiEc+fOkUgkKBQKZLNZAoGA0S1qeHiYXC5nLNwefvhhVq1axdjYGJFIhL6+PkKhEIODg0Zr3OVC0zRGR0cBeP7556mpqWHt2rXkcjlOnTpFMBhkeHiY6upqWlpajNcaCoWMzlQul4uOjg6j6Fh/H/RdIT1g0zSNdDpNNBplfHyc3t5epqam5n3sDocDq9VKIpEwunjpZ/RNJhMWi4V8Pm8EiXqtyDvhjL2qqlRVVVFaWmrsXqiqetEiQg80amtrCYfDjI2NoSjKsguMhbgSRVF+AWgBMsBpYJ+maelFfD4r8CuARnF34nrv/x8oBgkf1DRtZj6Le0VRKoGdQBrovu4HEEKIayDBxiLatWsXH/7whzl48CD9/f1zgot8Pm8M6hsaGqK/v59sNmu0E921axc7duwwApVjx44xMjJCZ2cnJ0+eXHbBxvDwMOPj45w9e5aSkhLuvPNONE3j8OHDBINBxsbGeOCBB2hpaeH48eMcPXqUgYEBIpEI8Xic6upqfu/3fg+73W608J3dWtVmswEYaWSjo6MMDQ3R1dVltNGdT4qZXiOTz+eNFDdVVXG5XFitVpxOJ7lcjnQ6TTKZND6/2zXYmJ0eZTabqa+vp6qqyrhs9o6GTm8A0NraSjKZpLu720idu13fJ3Fb+uYF308qivKbmqZ9d5Ge7zHAD7ygadrZ67mjoihNwN8BX9c07fvXcb9twPso/t9fT7FmowT4z5qmXXNRvRBCXA8JNhaB3W7H6/XS0NBAc3MzpaWlhEIhTp06xfDwMPD2ok5PQ3r++edJp9Ps2bOH1tZWGhsbsdlsRs2Cw+EwdjnS6TSBQIDJyUmjQ9NS0xfrev3G4cOHyeVyjI2NGcHViRMnjNsBzMzMGFPQy8vLqaqqwmq1ksvlaGlpMTogeb1e4PLb8vOpZXE6nTidTt7//vezYsUKnn32Wc6dO8fU1BQlJSX83M/9HH6/n+bmZtxuN263m2QySTqdZmxsjEAgwHe+8x1mZmaWdfH+fKiqyo4dO6irq8Plchk7PfrAxtmfw+zuasFgkHA4TDweN3aJhLgF/AD4a+AIMAM0AR8Dfg/4lqIo79M0bV7F21ehp1B98Yq3uoCiKCrwVYpF2L91nc+5jbldn6LAr2qa9rXrfBwhhLhmEmwsApvNZiye9a+5XM6oY9DPDutf8/k8r7/+Oq+//jof//jHue+++6isrMRmsxmTxfWC65qaGuLxOCtWrCCZTC6bYEPvmKVpGtlslp6eHrLZ7JxF59DQECMjI3Puo183u91tKpVix44dAHi9XmPo4aWec75sNhulpaVs27aNXbt20dPTQzQaJRqN4vV62bhxI62traxbtw6fz0dJSYlRhzA6OsrY2Bj79u0zumLdLh2x9FSptWvX0tzczMTEBGazGavVesldDcBoQRyJRIxgWGaiiFuFpml/c8FF3RQLpkeBzwP/H/PsFHU5iqKsoFhAPZ/C8N89f9+HNU0LXs8dNU37v8D/VRTFTjFl7BPAvyqKskfTtE9c53EIIcQ1kWBjEej5/fF4nFgsZsySeN/73sfq1as5d+4ciqIYZ44rKio4fvw4Bw4coK6ujtraWvL5PJlMBpvNZizySkpKeOCBBygtLWVoaMhojbuchEIhVFUlm81eVNtw4Q7E7OvS6TQ9PT1GUNHa2kooFJqzaNXTnNLpNP39/UxOTs57ga+qKmazmcrKSiorK6mvryeRSNDQ0EB9fT0PPPAAZWVllJWVYTabjU5LmqbR3NyM3+/nwQcf5OzZs5w5c4ZIJMLk5KRR03Gr0t8XfaDir/3ar1FaWordbp8zxG82vQZpZmaG8fFx4vE46fSipboLcbN8CfgbYJOiKB5N06IL+Nj/EVC4zsLw80HKX56/37PzffLzBeOngd9WFMUG/IaiKC8uYsqYEOIdTIKNRTB7UZxOp43F6vr16ykvLzeCj7a2Nvx+P/X19QAMDAxQUVFBSUkJwJz0HL21aEVFBdXV1UZh83JztaLgywUHhUKBUChkzNvI5/Nz6gdm3y6TydzQ0MPZ9CL00tJSKioqyOfzNDQ0UFNTg8vlMuaAzD4O/fNcuXIlJpMJRVGYmJggmUySSqVIJBI3dExLSd/ZMJlM2Gw2mpubcblcxnWXou9qJRIJYrEYmUxGOlGJW56maSlFUaJAKcUheAsSbJwvDP8Y8ysM7wBswK8qivKrl7nNmfO/qx+8xnqO54DfAO4FJNgQQiw4CTYWQSqVYmJigsnJSaanp/H7/Xi9XrxeL5s2beKBBx4w2t/qdRltbW384i/+ojF7Qi+MvjA/Xt/tWLFihXFW/XaoGdBb365cuZIPfehD3HHHHWzfvh273W68B3qKVjweZ2BggL6+vjk1INcjnU4TiUT4wQ9+QFdXF7t372bPnj0cOnTIOJOvn+G/1CLbbDbzS7/0S2QyGVKpFDMzM7z22mu88sorfPOb37zlOlbpOxr6LprH46GystIIxi73PsDbwUYymSQSiRhdu4S4lZ2fgVFKMchYyOLpDwIVzKMwHOgH/vky1z1McdbGd4DI+dtei7rzX+UMgRBiUUiwsQgKhQLpdJpQKMTw8LAxwE5PiVIUhXw+b3RbstvtOBwOoxAa5p5F1gtwM5mM0b0pk8kYCzq90PxWphchV1RU0NHRQUNDAx6Px3gf9NS0YDDIxMQEExMTxGKxeb9uPfVneHgYVVV54IEHqKiooL6+HrfbbexY6LtUs2tP9OtcLhculwtN0/B4PCSTSQYHB3E6nUbdwq3yueithvUdDY/HYwzx0wONy/2czS4S11/zrfK6xTuboiitQFrTtJELLvcD/3L+229qmraQC3G9MPwfr/eOmqYdBT5+qesURdlPMdj4b5qm9V5w3T3AzzRNK1xweRvwR+e//dH1Ho8QQlwLCTYWgb5A/clPfsLx48d58MEHWbduHR/4wAfwer3GQsxqtc4ZkKbPeJj9OPl8nnQ6TVdXF2NjYxw6dIiuri5eeuklwuHwbXMG2Wq10tzczJo1a9i0aRPl5eVzdjTGxsYYHR3lc5/7HCdOnKC3t5dMJjPvNKp0Ok2hUKC3t5dsNovb7cbv97N7924jlUtP2erp6WHv3r3GLI67776bqqoqGhoasFqtWK1WY05IMBjkxIkTnD592ujEdSvVcJSXl1NdXc327dtZvXo1Vqv1omnhF8rlcqRSKbLZ7C1fsyLece4GvqQoygGgDwgAjcBDgBd4C/gvC/VkiqK0A++iWBj+9EI97jX4ARBSFOUNYIji//1twHvO//3zmqa9cKNPoijKIxRnf0Ax8AHYrSjKV87/fVrTtN+/0ecRQtxaJNhYRLFYjJGREXp7e8nn86xcuZKamhoj4IjFYgDGGWWTyWScQfd4PKiqSiAQIBwOc/jwYSYmJhgYGGBgYICZmZnbKi/ebDZTW1tLeXm5MYUa3u5YFY/HmZmZ4ezZswwNDZFKpW7o7LkeECaTSYLBYkMXPciYvbguFApEIhH6+/uJRqNkMhmqq6spFAqUlpbicrmMlrBWq5WamhrWr18PFHdjpqamSCaTy/5Mv6qqWK1WysrKqKurM17bpdKnLiz6TyQSRstbGeQnbjGHgK8DW4FNFGdORIETwLeBf9Q0bSF/qP8f5lEYvgD+FHgA2AW8HzBRDHi+D3xJ07TnF+h5NlGsR5mt9fwfgAFAgg0h3mEk2FhE+gyJ5557DovFwle/+lV8Ph+f+MQnsNvtnD17llAoxLlz56ipqaGmpoaRkRFmZmb4+Mc/Tl1dHZ///Oc5c+YM/f39WCwWKioqjOGAtwuTyYTb7Wbr1q1s2bIFr9drLOChuKCdnp6mu7ubyclJ4vH4DS/e9dSfQCBgpEkpinLJGpHx8XFOnz7N+Pg40WjU2BGxWq1G6pXFYsFms7F69Wp+53d+h3379vHSSy/xwgsvMDQ0tOxTqpxOJ+Xl5WzYsIFVq1ZRW1trpJPNpgd/+u6FpmmcO3eOw4cP09fXRzAYlJ0NccvQNO0E8B9u4vN9Gvj0Ij32vVe47u8oDgFcVJqmfQb4zGI/jxDi1iLBxiLTF6z6WfRsNsvx48exWCyMj48TCoUYHR0lEokYf49Go7z++uuUlpbS1dXF5OQksVjMKBy/lbsdzeZ2uzGbzaRSKSwWizE8T08r04us8/k80WiUYDBotNS9UfrCX09TO3z4MIqi0NraauyqFAoFotEoU1NTjI6OEgqFSKVSnD17lkQiwf79+6mrq2PPnj14vV4j9cvtdtPa2srMzAydnZ3MzMwQiUSWdbBhNpuNGhSPx4PFYpmT3jdbPp83Wt3OzMxw7Ngx+vv7jd2m5fw6xW3vTxVF+VOgW9O01Ut9MOLKFEV5nGJBuxDiNibBxk0wewEWi8X4xje+ARRTV/SzxIODg3MKcD//+c+jKMqchXUymSSZTN78F7AIVFWltraWkpIS+vr6sFqtlJaWXhRs6IXhQ0NDDA4OkkwmF+zMuR7IxONx/vqv/5r169fzuc99zhgimM1mGRsbo6+vz+j6VSgUOHjwIIqi8Oyzz1JXV8cnP/lJ1q9fz65du7Db7TidTlavXk1FRQUnTpxgdHSURCKxrHej9Pe/tLQUv99/xSF+uVyOUCjEd77zHb73ve8ZKWn6rs/tUkckbin9wJ/N+n4hu0eJxXOKuZ9b/xIdhxBiEUmwsQT0gOJSQ+9m//12PUNstVqx2Wxs27aN2tpa7r33XsrKyrjjjjvwer2kUimjlmJiYoLR0VFeeeUVOjs7SSQSC/6+6HUHqVRqzuW5XI7p6Wmj69WFn08+nycWi9Hb24vdbmfNmjXA23M4fD4f27ZtQ1VVvve97zE9Pb0sU4z0AX4VFRXGDs3s9KnZaWVQfF9isRihUIhAIGDs2KVSKQk0xJLQNK0fSd+55Wiadgr53IS47UmwsYRu12Diamw2Gz6fj61bt7J161aamppwuVyUlZWRzWYJBAIMDw/z5ptv8sYbb/DGG28wMjJCPB5ftGNKp9MXFXJnMhkGBweJRCKXvI8epBw5cgSTycTu3btRFMVId7Nardx1113U1NRw4MCBZVvPoCgKHo+H8vJyKisr8Xq9RnvfS3WfymazBINB48/tVkMkhBBCiIUjwYa46fTAoq6ujqqqKnw+HxaLhVwuRyQS4bXXXqO3t5eTJ0/S1dVFKBQim72ZjVvePs6dO3fS19d32dtkMhn6+/vJ5XKYTCbWrFnDjh07aG1txefzGdPef//3f5/u7m6+8IUvLGrQdK1UVcXpdBoTw1euXMmOHTvo6OigpqZmToH+bIVCgVgsRl9fH6FQ6OYetBBCCCFuORJsiJtKT9kpKyszAg2Hw4GqqmSzWcLhMK+//jpDQ0McO3aMQCBgTKW+GfR0KX2ieWNjI1VVVZe9fT6fZ3p6mkgkwvj4OPfccw9Wq5XKykrKyspwu904HA7e+973snbtWr7yla8sm2BDr48xmUzU1NSwbt06amtr53QDuzB1rFAokEwmGRoaMlo3CyGEEEJcjgQbYtEoioLZbDaKhl0uF16vl5/7uZ9jy5YttLW1GQtbvcg4FosxNDTE5OQkgDE5fTHTjzRNI5lMEg6HSaVSuFwurFYrwDU/r74rMzw8TG9vL7t37zaCFpPJhM/nI5vNsmvXLnp7exkZGSGdTl9UJ3Iz6NPPt27dyqpVq/jABz6A3+83ZpxcqjhcnxCeSCSIRqNAsWFBJBKROg0hhBBCXJYEG2LR6LsDUFys+nw+6uvrWbFiBRs3bqSkpASLxWIUWycSCSKRCDMzM0SjUaP7082oc8hkMiQSCeLxuNH6VT+Tn8/nr1pfo2ka6XSaYDDI1NQU4XCYdDqN2WxGVVWjtW9tbS3JZJKZmZklXaRbLBYqKytZvXo1W7ZsMYILs9l8Ub3G7PchmUySTqfJZrOk0+lbbkK6EEIIIW4uCTbEorHZbDQ3N9PY2Mj69evZvn0727dvp6SkBLvdjtlsJp/P09/fz/DwMF/+8pcZGxvj7NmzZDIZUqnUTTvzH41G6e3t5S//8i/ZuHEjn/jEJ0ilUvT39zM+Pn7V7mB6wHTu3DkmJiY4fPgw9fX1/OVf/iWrVq3CYrFgsVjo6OjAarUyPDw8Z4r8zaSqKg6Hg+bmZpqbm42dDD3AuLAwXJ9zoqeyTUxM8MYbb3Du3DnZ1RBCCCHEFUmwIRaF2WzG4XDQ1NREU1MTHR0drFmzhrq6ujlzNHK5HGNjY3R2dvLKK68Yi+98Pk8mk7lptRr6zsrRo0dxOByk02lisRjDw8OEQqFr7hymz0IJBAJ0d3fzB3/wB+TzeWOHo7a2lkAgYOweLAU9tcvlcmG321FV1Qg29Ot1epCVyWTIZrNGJ6rF7g4mhBBCiNuDBBtiwZlMJmpra2lvb+dXfuVXqK2tpb6+nrKyMkwmE9ls1kjJiUajfOc73+HUqVNMTEyQzWbn7CLczPbA+Xye8fFxzp07R3d3N+fOnePf//3fOX369HUfh16DkkqljE5Vqqqye/duPB4PX/7yl5esXsNiseBwOCgpKcFms10U9Oi1JvrnMHsg5eHDh3nrrbfo7e1dkuMXQgghxK1Fgg2xoMxmM3a7nZUrV9Lc3Izf76ekpASPx2MUgqdSKeLxOJFIhFAoxOjoKENDQ0YQslT0Y5uYmODIkSMMDw9z9uxZwuHwvB8vEokQj8eNBb2+m3Mzd20uZLVasdvtcwrhLxVwzA74EokEwWCQ/v5+ZmZmSKfTkkIlhBBCiKuSYEMsKI/HQ0VFBR/+8Iepq6vDYrFgt9vxer1GOs6ZM2c4duwYo6OjBAIBDh06xNTU1JIvXjVNIxwOc+LECf7gD/6AQqFAKpWa9+6KpmmcPn2a8vJyGhoaUBSF0dFRTp8+TTQaJZ1OL/AruDpVVfH5fNTV1dHQ0IDP57tkOpdeFK+qKrlcjs7OTrq6uvjZz35Gb2+vDPETQgghxDWRYEPcMEVRKC0txev1smfPHurq6tiwYQMej8coRs5ms0ZnqdHRUbq6upieniYajRKPx5dkaN/l5PN5ox7hRtK4NE2ju7ubfD5PW1sbZrOZ4eFh+vr6yOVyS9LFyWKx0NLSQkNDAxUVFbhcrkset17wrs85CQaDxGIxwuHwsvqshBBCCLG8SbAh5mV2xyJFUdiwYQNr167lt37rt6ivr8dms6GqqtESdmZmxli4Hzp0iB//+MdGe9hwOLzkuxoXWohaEU3T+OY3v4nJZDJ2eeLx+E3tsjWbqqrY7Xa2bt3Khg0baGxsxGKxzNnZKBQKxmvP5/O8+OKLjI6OUigUCIfDRCKRJdmREUIIIcStSYINMS9msxmPx2PUILS0tLB27VpKSkowm80oimIUgQcCAY4dO8b09DTT09McOnSIyclJ0un0kp3hv1n0xXsoFEJVVTKZzJIFVvq0cLfbTUlJidGF6lKi0SihUIiTJ08yOjpKKpUiEAgQCoVIJBI3+ciFEEIIcauSYEPMi81mw+/3k8lkSKfTrFixgu3btxsD8aB4ZnxiYoK+vj6efvppurq6OHz48KJPBF9u9GBjqenDBUtKSvB6vXPa3c6maRrj4+P09/dz6NAhzpw5w+TkpPGZLVVhuxBCCCFuPRJsiOum12hs2bKFfD5PNpulo6PDSBVSFMWYOD08PMzJkyc5fPjwnNa24uaYnermdrvx+/2sXLmSmpqaSwYbuVyObDbLoUOHOHz4ML29vczMzJDJZIzP7Z0UKAohhBDixkiwIa6bqqp4PB4aGhqMyxobG/H5fJjNb/9I5fN5hoaGGBwcpLu7e0lTiN6JFEUx0qQURcHhcOD1eqmpqcHn810yhSqfz5NOp+ns7OTgwYNMTExI2pQQQggh5k2CDXFdLBYLfr8fn89HJpPhzjvvZPv27ZSVlRkzG/RuTqFQyEiZ0lupipuvpKQEh8NhfD5X2pnI5XIkEgmi0SjBYFBSpoQQQghxQyTYENdFURTsdjsmk4lkMklFRQV1dXVAMddfH8wXiUQIh8MkEgnS6bQEGktE0zRj1onVajXqaS6nUCiQyWRIJpMkk0n53IQQQghxQyTYENclm80yNjZGIBDg7NmzNDc3s2rVKsLhMNFolOeff57JyUncbrfxfSgUIpfLycL1JtPnZUQiETKZDHfffTdNTU3GDoemaUbNhqZpRnvbkZERwuEwyWRS6jOEEEIIcUMk2BDXRdM0UqkU2WyWZDJJKBQilUoRiUSYmZmhp6eH/v5+6uvricfj9Pf3S1H4EsvlcmQyGex2Oy6X65K1GnpgkslkjN0oqa8RQgghxI1SZBEohBBCCCGEWAyXnuglhBBCCCGEEDdIgg0hhBBCCCHEopBgQwghhBBCCLEoJNgQQgghhBBCLAoJNoQQQgghhBCLQoINIYQQQgghxKKQYEMIIYQQQgixKCTYEEIIIYQQQiwKCTaEEEIIIYQQi0KCDSGEEEIIIcSiMF/pSkVRtJt1IGL+FEWhtraWuro6HnvsMdrb29mxYwfDw8O89NJLrFy5kpqaGurr63G5XDgcDjRNIxQKYbVacTqd5HI5crkchUIBTdNQVRVFUTCZTGiaRjabpVAoAPDnf/7n/P3f//0Sv2oh3lk0TVOW+hiEEEKI63XFYEPcOlwuF6WlpVRVVeHz+Uin08RiMQKBACaTCZ/Ph6qq5PN5otEomUyGyclJPB4PFouFcDhMMBgkk8mgaRr19fXYbDY0TUPTNCPwUFUVp9OJy+UinU6Tz+fRNIlJhRBCCCHExSTYuA2oqkpbWxsrV67kjjvuwGQyMTIywvHjxzl06BAPPvggzc3NxONxYrEYr732GmNjY/T19bFixQre8573cODAAfbt28fk5CQAn/3sZ2loaKBQKKAoCqqqYrFYsFgsNDU1sWbNGs6ePUssFiOTySzxOyCEEEIIIZYjCTZuA5qmMTw8jMPh4PDhw+TzeTo7O+nv72dgYID9+/ejqiqjo6MEg0FOnjxJOBxmbGyMWCyG3W7n6NGjdHd3EwqF0DSNZ555hoaGBmPnw263U11dTV1dHevWreOXf/mXef755+np6WFoaIhsNrvUb4MQQgghhFhmlCulwEjNxq3DarXicrnYuXMnkUiEN954g3w+D4Df78fr9RIIBEgmk6TTaSP1ye12U1tbSygUIhgMksvljOssFgstLS14PB5KS0vZs2cPH/jAB2hqasLtdvMP//APvP766zz77LPEYjFJpxJiEUnNhhBCiFvRstzZMJlMeL1eysvLWbVqFadPn+bcuXNGgbK4WC6XI5lMcurUKTKZzJz3Kh6Pk8vlSCQSF9VYpNNppqamSKVSF12Xz+eZmpoiGo2STCaJxWLG4yqKQnV1NTU1NZjNZgk0hBBCCCHERZZlsGGxWCgvL2ft2rXs2rULVVUZGRm5aBEt3lYoFEilUgwODl50XTKZJJlMXvJ+2WyWYDB42ccMhULYbDYURSGVSlEoFIyOVeXl5fj9flRVOigLIYQQQoiLLas0KlVVqaiooKmpiccff5xVq1axceNGotEooVCI//N//g979+5lZmaGXC53Mw/tHcdiseD1enn/+9+P2+3GZrOxZ88e9uzZw6uvvsqrr77KsWPHGBwcpLe3V2o2hFhkkkYlhBDiVrRsdjasVis2m436+npjTkRdXR01NTXU1NRQKBRYtWoVR48eJRKJSLCxiBRFwW63U1ZWxoYNGygtLcVisVBXV4fFYuH48eN897vfZWxs7LI7JkKIIlVVjRbSQgghxDvNkgcbZrMZm83G448/Tnt7O/fffz9+v5+amhpjsFw+nyebzbJt2zZyuRz//M//TDKZNNJ35D/yhWM2m7Hb7dx33320tbXx0EMPUVJSgtPp5Pjx4/zhH/4hr732GsPDw7KbcRMpSvGktvyc31qsVitVVVUkk0kikQj5fN5o3CCEEEK8EyxpsKEoCm63m5KSElpbW9mwYQNNTU24XC6sViv5fH7OmXO/38/q1avZtm0b586dMwqa0+k0yWSSaDRKKpWSXY/rYDKZjCni+mdRUlJCR0cHzc3N2O12FEUhnU4zMjLC0aNHGRsbk9kai0RVVWNq++w/JpMJ4JYfoqgoivFHHzJ5O9ZhKYpi/C5t2LCBUCjE8PAwgUCAaDS61IcnhBBC3DRLVrNhMpmwWCzs2rWLtWvX8mu/9ms0Njbi8XgAjAnXr732GuvWraO5udk4u5vNZslms0QiERKJBMFgkEOHDvHSSy9x9OhRRkZGFuuwbytmsxmv10tLSwt33303e/bsob29ndLSUrxeL4VCgTNnznDmzBmeeuophoeHGRgYIBwOk0gklvrwb0sejwe/3086nTY6iBUKBRwOB4VCgWQySS6XuyUDapPJhMlkMk4mOJ1Oo+Xy7cbpdPLII4+wYsUKHn/8caampnj99dd58cUX2bdv37weU2o2hBBC3IqWZGdDURScTid+v581a9awadMm/H4/Dofj/9/emwfHdZ73ms/pfUd3A42F2BcCBMEFXESKJilRUiRKSaxYsRXJpYqTazk35di544krM5UbT+zrjOM75Zm5nusq74mVxVsiRUlsxRIjW6IomrRIigtIAsSObhBAY+l9R3ef+aN5PoEUKS4ACJD6nioWiEb36e90nwbe3/e+7+8lk8mQTqcZGRnB7/dz5swZbDYbLpeLfD5PsVhEp9Oh0+mwWq3o9Xrm5+cxGAzC2lVyfbTXr62tjUceeYT169fT2NhIeXk5drudfD5PMpmkp6eH/v5+zp07RyKRIJ1Oy/KpJUQT0BpGoxG73Y7P58NoNBIKhVAUhS1btgCIwYzT09PEYjESicRKLPuG0DIYJpMJvV6P2WzGYrHQ1NREWVkZlZWV4vPe09PDzMwM8E5ZZLFYJJfL3XFlkna7nfLycjo7O1m/fj2VlZWYzWY2bNjAwMAAbrebRCJxRwrG1YqiKPuA1xbcdEFV1XUrsxrJakNRlLNA14KbHlBV9fUVWo5E8r7jtosNrXyitraWe++9lyeeeIKtW7fidDpRFIWBgQHOnz/PV7/6VcLhMOl0mkgkQi6X48KFC8zNzWGxWKioqOC3f/u3gVJwEovF6O3tJRaL3e5TuiMxGAz4fD4efvhh/vzP/5z5+Xmy2SxGoxGAgYEBhoeHee655/D7/QQCARHw3UmB32pGK42Cd0SH1WrF6/XS1dVFTU0N09PTmEwmPv3pT6PX6zl//jwDAwMcP36ckydPcvbs2VX7fuh0OoxGIz6fD7vdjsvlorKykj179lBfX09XVxdWqxWz2czXvvY1fvaznwGl6yufz5PJZJibmxOZzDsBRVGor69n3bp1fOhDH6Kqqory8nJcLhder5dAIEBvby99fX1EIpGVXu7dyEHgdWD2aj9UFOUDwOeAewELMAj8DfA1VVVveKdKUZRngR1AN7ARsAJfUlX1czezWEVRyoEngN+4dJxaIAf0AN8Dvqeq6g3XGSqK8jpw/3vcxaqqauYGjtMEjNzAU96nquqh6xxLAfZTOse9QCOl12sM+BnwZVVVgzfwXNc6vhk4QUlMXFRVte4qd/s6UAns471fH4lEsgzcNrFhMpnYsmUL9fX1bNy4kfLycurq6mhtbcVqtaKqKrlcjoMHD9Lb28vExASZTIZcLkdvby86nY6pqSnC4TAWi4Xq6mp2794tpluXlZVhsVhW9U7vSqPX63E4HHz0ox+lvLwcr9dLd3c3hUKB6elp/H4/wWCQaDTK+Pg4s7Oz+P1+otHoHbe7vBpZmMXQavo1waGZIezatYsHH3yQpqYmvF4vkUgEg8GA1+tFVVWqqqpIJpN4vV7sdjuKoqy690XrA6qoqMDr9bJnzx4RdJeVldHW1obT6cTtdmM0GjEajTz++ON0dHQQj8eF8E0kEoyNjREKhZiYmGBycnLVl1ypqko4HGZubg6Hw4HVahUmF9FolHA4TDQalVmN5eN1VVW/cLUfKIryW8ALQAb4MRACPgj8D2A38ORNPM//A5QBYWACaL3F9T4JfAOYpJSZ8QNVwG8D3wUeUxTlSfXmP+T/7Rq33+iFF3mPY9QDHwfmgLdu4FhmSqIiB7wBvArogQeB/wV4WlGUvaqqDtzg2q7krygJmGuiqurXARRF+QJSbEgkt53bJjbsdju/9mu/xj333MOjjz4qyiSMRiMGg4FsNksymeQnP/kJp0+fJhQKiSDqxIkTnDt3jkwmw/z8PHq9npqaGh588EHa2tpYs2YNVVVVOBwOotHo7TqlOw4tAPzMZz5DdXU1er1eDAPs7+/n3//933nttdcYHh7G4XBQLBYJh8OyNG0J0MSEJjgURcHj8WCz2cR99Ho9O3fu5KMf/ShmsxmDwSAawi0WC/Pz8/h8PrLZLJWVlZSVla3U6bwnRqMRl8tFa2srzc3N/M7v/A7Nzc243W4MBsNlDfBaplOb4RKPx8nlcmSzWeLxOKdPnyYQCDAwMMDhw4dvq9i4Vcva2dlZHA4HdrtdGCzk83mCwSDT09OEQiFpsHCbURTFBXwHKAD7VFU9fun2/wP4BfARRVGeVlX1Rzd4yKeBXlVVxxRF+X1KWYhboR94HHhpYQZDUZT/SimQ/zAl4fHCzRz0WoLrJh4fAa56DEVRvnzpv3+nqmr2Bg5XoJRN+rqqquIDrCiKjlLG4Q+B/5eS8LspLpXP/a/AH1ESbRKJZBVyW8SGXq/HZrOxdu1a6urqhGWtFngVCgVmZ2e5ePEiU1NTJJPJy/7Az8/PUywWxW6g3W7H6/VSV1eH2+1e0rXqdDpMJtOSNOHq9XrcbjdtbW1MTEwwPj6+IrvQW7ZsYcOGDWzYsIHq6mri8TjJZJLp6WmmpqYYGBhgaGiIc+fOcfHiRbLZrHAIuhudguAd1yeXy4XBYGB2dpZCofCuaegL369bfe8MBgN2ux2DwYDRaERRFKxWK5/61KdoaGgA3ulvaGpqEte65kKlZT+KxSKJRIJQKMTs7OyqdTXSzrempobGxkZqamrweDwUi0VSqRQTExM4HA6qqqoAxLlC6Ry1z43NZmPLli14PB7S6TSNjY3MzMwQjUaXpaxKURS8Xi+PPPIIBkPpV+PJkyc5d+7cLb33msCEktBvbGzE4XDIOUErw0cAH6UA+bh2o6qqGUVRPgf8HPgkcENiQ1XVl5diUaqqXtUtQFXVKUVRvgl8iVLpz02JjeVCURQD8PuXvv32jTxGVdV5Sudx5e1FRVG+SEls7LuFtbiA54Cfq6r6TUVRpNiQSFYpyy42FEVBr9djsVhoaGjA4/FcNh+jWCyiqirT09MMDQ0RiUTIZC4vKV3oTa81nDocDhwOh9g5XCp0Oh0Wi4VcLifWdiuBhnbeLpeLzs5OAILBIMVicVHHvZnn1wLqrq4udu3axb59+3A6nQwNDTE1NcXJkyfx+/289tprJBKJy0rQ7qZsxsLrTQvqzWYzJpNJNO9mMhmy2Xdv0i18r27VdtZgMGCz2bBardhsNgwGA263m4ceeoj6+nqxLm2N2mtfKBSwWq1ih11zo4rH46RSKbLZrDif1VRKpc1qcTqdlJeXiwn0yWSSZDLJuXPnqK2tpby8XAirha+tXq8XGZDq6mrm5+epqamhsrISr9dLKpVath4Ol8vFvffei8lkAiCTyTA0NHSZAL8RFopWVVWF85vZbJaDMFeGBy99vZpIeANIAR9QFMV8g7v1twPtIr9pZaooylNAM6XSpV7gF0t0Xr8FVANvqKratwTH01J8t6K+/yfgAZ5dgnVIJJJlZFnFRnV1NW63G51Ox9q1a2lvb8fj8aAoCslkkqmpKQ4fPsyJEye4cOECk5OTzM5eta9PoKoq8Xgcv9/PgQMH6OrqYu/evUsSbOl0OjFjQgsyJycniUajN7UbqQkNu91OU1MTu3fvZtu2bXzoQx+iv7+fiYkJ0cyu9UMsFZrTV319Pbt27eKxxx4TLlPxeJyZmRm+//3vMz4+Tl9fH8lk8q4tlTIYDFitVrZv306hUGB0dJTW1la6u7tpb2+nra2NqqoqLBYL+Xyeubk5nn/+efL5vMhuZbNZhoeHmZiYYG5ujlQqRSKRuKn3zOPxsHPnTjo6Oli3bh0VFRW4XC4xkX1h4DwzM8PAwIC4hrQd/tbWVgwGA2vXrsVqteL3+4nFYoRCIYLB4KqyIs5ms4TDYQYHBykUCuzatQu9Xs/k5CRTU1O8+uqr5HI5FEXBYDBgMpnE502v16PT6XA4HGzYsIEnn3yS5uZmfD4fXq8Xt9vNv/zLv+D3+5d83Tqdjrq6Op544gksFgsA9913H3/4h3/IX/3VX92wZW15eTk+n4+5uTkMBgMul4tsNksgECAWi+HxeIjH4zK7cXvpuPS1/8ofqKqaVxRlhFKDcQul4HxFuZRB+Nilb28li3JlhmZaUZRPqar6/OJWxn++9PVbizyOhiYUbuocFUV5Avg94BOqqi79LwOJRLKkLJvY0GrS6+rqRGDldDoxGAzk83ni8TiBQIBjx47xyiuvMDk5ecMBUy6XI5lM4vf7RXnGUqHtQGo7k7FYjHQ6fVPZEy2rYDabKS8vp62tTew+a7drZTvxeHzJAn1FUbBYLJSVldHa2kpnZyfbt2/HbDZjNBqZmppibGyMvr4+RkdHuXjx4l1bJqUoCna7nbKyMjo6OsRrvHbtWrZt20ZXV5e4JjUHrmQyydDQEPPz88KGOZvNYrFYKC8vJxAIEA6HGRgYuOFAUWvKr62tpb29nc2bN+N2u7FarRiNRmGMoGVPQqEQ58+fF+dQWVlJRUUFjY2NImNgt9ux2+1YrVYcDsd1BfrtplAokM1mmZmZwWq1MjY2hsFgwO/34/f7GRkZYXp6munpacxmM1arVTxWKxvzer0YDIbLslBVVVWi7G25MBqNOJ1OLBaLEKsVFRWsWbMGvV5/3c/qwutufn5eXCf5fJ5QKCQc364s15MsO1qD07Wa+rTb3cu/lBvivwMbgH9XVfWVm3jcvwL/N3CSUgN3I6Wg/LPAjxVF+U1VVX92Kwu65FD1a5eOu+iyLkVR7gE+D8Qp9XTc6OOqKImdn6mq+teLXYdEIll+luWvtrZbuXPnTjZt2sTjjz9OeXk5FouFbDbLwMAAR44c4YUXXmBwcJDJycmbCrjz+TyJRIJTp06JRualCJq1wE/z+weIx+Nks9mb2snWymbq6uro7Oxkx44dFItF5ufncbvdTExMYDKZ8Pv9/Md//AepVGrR2Q2tD2Dnzp2sXbuWJ554gsbGRnw+H9PT04yMjPAP//AP9PT0cObMGdLp9F0rNAwGA2azmQceeIDm5maeffZZ9Ho94+PjVFdX09jYKFyQoPS+ZzIZFEXhySefFLdpX7UyPm0K9FNPPcXU1NR112E0GqmpqWHHjh18+MMfprGxkerqahFojo+PE4lEGBsbY35+HlVV+eUvf8nzzz8vdv0feOAB2tra2LlzJxaLRawnmUySzWZv+tq8HeRyOSKRCKlUipGREU6ePInZbL5sh18LxBeW+2klYRaLhcrKSvL5POFwGLvdjs1mo7+/nwMHDhCNRt+VEVoqJiYm+Ld/+ze6u7vZuHGj6LvQBN/4+Ph7NncrikJ7ezsdHR14vV4hpDKZDAMDAwSDQdmzsTrRdpNW/MOkKMp/oSQO+oDfvZnHqqr6P6646QLwXxVFmQC+Rsm56ZbEBvAHgA7428WWZCmK0g78BDACT6uqOnQTD//Opcf9wWLWIJFIbh9LLjYURcHhcOB2u6mvr6ejo4OKigosFguxWIxwOMyFCxcYGBhgcHCQcDh8S394C4UCqVRK/OGfn58nmUwu+o+4FuDodDrh1KQ1qN8IC0tD3G43TqdTlIkoinKZVa/D4Vh0v4kWqGllJm1tbbS0tNDQ0CCG801NTXHq1CmGhobw+/1iCvXdiKIolJWV4Xa7Remez+cTO+ZlZWVYrVZRIpVIJITzkdFoZM2aNUDpetLeG51OJ+xnC4UCVVVVpNPp6zqfGY1GKisrcblclzULQ0nEBAIBxsbGGBkZEQMrR0ZGmJmZEc+p9Q5pAqVYLJLNZolEImLtq1E0auvM5XIkEgnxvgDvEteamFv4+iQSCeLxuLD+tdlsQmhpPR0LG8uXilwux9zcnOir0ASQ2+2mrq6O6enp6zpJzc/PUygUMBqNQkQZjUaqq6sxm81CWEpuK9qH9VoWbq4r7rciKIryKeD/A84DD6mqGlqiQ3+XksVvt6IoTlVVb8pd4lJZ13+69O0NNYa/x7HWUrL59VISGv92E4/9GCXXqt9TVfXiYtYhkUhuH0sqNrQAqaOjg23btvHYY4/R3t4uGq7/+q//Gr/fT39/P2NjY4t2Z1rYVBqLxRgdHV3U8VRVJZFIMDT0ziaLVkt+I2h19jabjbKyMurq6vD5fJft3lqtVlwuFx6Ph1AoJJ73VrHb7TidTp555hmampro7OzE5/PR0tJCJBLh3LlzfO973+OFF14QYuxuDXQ0QXHvvfeybt06Pvaxj1FTUyOCfc25TFEUZmZmGB4e5sUXX2RoaIhoNEpdXR2f/vSnyefzjIyMiNdJm3hdV1eH1Wpl7969+Hw+XnvttWteHzqdDqfTyY4dO/B4PJw5c0aYG2i78t/4xjf41a9+RTQaFdeyFqhq/Qtut5vKykpMJhOKopDNZpmamuLYsWMMDw8zOzt71cb21cBCEwRVVd9ziN1CwZROpzl37hyKovD222/T3d2Nx+OhsrKSdevWkUqlRKnWcl7L2pr0ej0bN25kfn6eoaEh0un0NTOxqqoyODiI1WrFZDJhMplQVRW3282+ffs4ceLEXdkfdQdwAdgOtFMaACe4FEg3U2pSHr79SxPr+AwlQXCWktCYXqpjX3LdilNqqLZTKl26GT4I1AAHVVW9cKvrUBSlk5LzVznwpKqq/3qTh9h66evfKoryt1f5ea2iKNovBc8lC1+JRLLCLKnYMBgMOBwOWlpa6O7uprKyEovFIlx0BgYGxE7uYhujtVruaDTK0aNH6e/vX7LA41aDAc02t7Ozk7q6OrZs2UJra6s45vz8PPF4nLm5OeG3f6u70pp46ejoEGU2mhWwtnM/NTXFgQMH6OvrI5FI3LKb0mpGc/xqbW2lUCiQz+fZsmUL27Zto6KiAqPRyPj4OJlMhlAoJGyTz5w5w8mTJ7lw4QLDw8PC5emVV16hUCgwNzcnnmN0dBS32013d7couQLeMyulqirZbJaRkRFh6zw3NydmmBQKBfE5SKVSV3UoKxQKTE5O4nA4CIVClJWVYTAYqKioYNeuXZjNZiKRiBAodwua6DIYDNTW1mK324GSU5TP58PlchGNRpfUhU573oXiCN75TDudTqxW62VT39/rOJoYslgsGI1Gcrkc4XCYWCwm5tvcbZ/FVc4vgGeAR4EfXvGz+wAbJYelFVHuiqL875T6NE4BD6uquqSNWIqidFASGnGuMV39OmiN4bec1VAUZSOlgX5lwIdVVf3pLRzmCOC4xs+epeQqpr2/q3MXRiJ5H7JkYkOn04lmys2bN7N//35RPpVMJolGoxw6dIhAILAk1o86nQ673c7Y2Bhf+MIXLuuzWCm0zMW+ffvo7OwUE861XpBYLIbf76e3t1fMtrjVunOTyYTNZuPBBx9kz5497N27F5vNJuaDxGIx3njjDb74xS8u8VmuHjTBuXbtWp5++mni8TjRaJQPf/jDrF27FrPZTDab5fvf/z69vb387Gc/o7u7m/379/PWW29x5MgRwuGw2CEPBAKcOnVKCDkNp9OJ3W7n3nvvxWazCevZ97Kd1XbyX3nlnd5OvV6P0WgU/UsTExNkMhkR5F4506NYLHL48GFGR0d58MEHaW5upqWlhaamJj7xiU/wwx/+kAsXLpDL5e660pxCoYDb7aajowOHoxRbuN1uWltbGRkZIRwOMzMzs6Qi62q9X0ajUYgNm832rnK4q6GVvEWjUUwmE3q9nkgkwsGDBwkEAphMplXZa3OX8zzwf1GaVv21BUP9LMD/eek+KzKn4dJgwS9Syrg8cqulU4qitADZK8uLFEWp4J2hgz9SVfWmamgVRWkEHmERjeGKonRTEho24LdusuldoKrqjylNf7/aczwLhFVV/cStHFsikSwfSyI2dDodNpuNrVu38swzz7BlyxbcbjeKopDL5bhw4QLnz58nFostSVOnwWCgvLycp556img0yosvvrgEZ3Hr6HQ6dDodHo8Hn89HV1cX7e3twvY3l8sxNjbGoUOH6OvrY2Jigp6eHlE+c6vPqfWFVFRUYDAYUFVVDEb8xS9+QW9vL16vVwTPNTU1mEwmzp49+65ZJhqrsf7/amgN8R0dHXR3d7Nr1y4RBHq9XrLZLIODg0xNTXHixAlR/jI0NMTBgwcZHBwU1+OVQd+Vr4E21+H06dOYzWZSqZQopblewHjlsVRVJRaLiX4j7edXHsdsNmM2m2lsbKS1tZXa2lrcbrcwGtBmVmSz2bsuY+V0OnnyySfZvHmzEH2qqpLP58lkMszMzAg3t6UmEonwi1/8ArPZTHd3t5j34XA4WLNmjRiKGQqFrvmam81m0dSulb8VCgXC4fB7lmBJlg9VVWOKovwBJdHxuqIoPwJClKZ3d1y6/apB7NVQFOUTwJ5L37Zd+vpBRVHqLv2/T1XV/34Dx/k9SkKjABwC/stVxOyoqqrP3cCy7gO+qyjKQWCI0vk1AL9OKZtwHPjfbuA4V/IJFtEYriiKh1LplPfS112Kouy6yl2/KsueJJK7kyURG9pshw0bNvDUU09hMBhEA2cul+Po0aOcPn16ybzltWFsjz/+OBMTE/zrv95s2efSogX+Ho9HNMXX1dVhs9lEI/vZs2f5x3/8R4aHh0UD6mICe22XXGuG1uv1ol78rbfe4itf+QoWi4WKigrRJ7B161asViujo6PXFH3LPWxwqdAyaV1dXWzYsEH0BplMJhKJBNFoVJSQvfnmm8RiMXK5HIFAQFj+vtd5LnxvMpkMmUxGDD281ddH2zmPx+MiAL0WRqMRm81Gc3MzjY2NVFVV4XQ6ReN1NBoVDeJ3W7O/Jjbq6+uF/a1WWpXJZAgGg6LfaamJRqO89NJL1NXVkc/nhVB3OBzU1NRQXV3NzMwMkUjkmu+f2WzGYrEIe2PtvY7H4zcsUiVLj6qq/6Ioyv3AnwMfBizAIPAnwP9Ub+5N2UPJUnYhmy79AzhIqSzqejRf+qoHPnON+xykNCn7epwA/gHYBnRTanqPAz3APwLfUlX1vd0NrkBRFD3w8Uvf3moJVRkloQHw0KV/V+M5IHKLzyGRSFYxSyI2tD/Gdrv9MgeWWCzG7Owshw4d4q233rrmbvr1sFgswjbXYDDw+c9/nu7ubmw221Isf9FoJU1r1qxhzZo1uFwuzGazqN3OZDLCWUdzt1pssKHNAOjo6KChoUE0ou7YsYOamhrm5ubw+Xx0dnZisVjEzA+9Xs/u3btFPbnmuhUKhYhGo/z4xz/m1KlTS/PCLBM2m40/+ZM/ob29nXvuuQe73Y7BYCAWixGPxzly5Ajnzp3jrbfeEoPUtIBcK0+6ldd/qQJEzZnsvdDEhtfrFVPOFUUhkUgwOztLX18fk5OTSzqnZaXR6XRs2bKF9evX09bWhtfrvcyF6vz58/z85z+/rJ9mOdBK4CYmJvD5fDidTiorK7Hb7WIi+rXePy27qJVPavfzeDw8+uijjI2NyX6NFURV1cOUdvoXe5zfB35/CY7zBeALiz3OpWP1sARruuKYBaB2kccY5R1r4WVDVdVlfw6JRHJrLGkZldVqvWzgViwWIxAIcPbsWUZHR2/p2Fq5jNvtJpvNYrVaeeihh2hqahKWtCu9G282m3E4HGIAm8ViEQPA8vk86XSaRCJBMpkkl8stemdTczWy2+1iwrSGVr62fft2mpqaaG1txWg0YjAYxFwJj8eDqqoiaCoWi2Ia9YkTJ+jr67vq82r9ILeLa/VDGI1GPvCBD7Bu3ToaGhrEbn8qleLixYscO3aMo0ePMjQ0RCKReFcWZzUEetdbg7ajru2SL9zdTyQSTE1NiTKw1Vj6dq33TvsZXP4aaE5umnWzx+PBZrOJ5up0Ok0gEOD8+fMkk8llX380GmViYgKr1YrdbhdD/ioqKvD5fFitVrLZ7Ls2DhRFEWYDC98Xk8lEbW0tZWVlq+L6u4v5vKIonwcuqKq6bqUXI1kdKIpyltKEeIlEsgIsWmxoYmD9+vXU1NQApZ3bfD7PD37wA370ox/dkNC4WgCiNWY+9thj3HfffWzdulUE9ADDw8P09fURDodvOWuyGDSr3w984ANs27aN++67jzVr1uDxeFAUhUAgQDAY5OjRoxw/fpypqSkymcyidqI1x6/169ezefNmbDabqP0vFotCyNx33304HA5cLpcQY1NTU0QiEfx+P263mx07dohmVofDQVVVFc888wwdHR2XTTnW5o2cPn2aV1555V2NzEuNNqvE6XQKVx/NcclgMKDT6ejp6cFkMlFTU4Ner8dkMlEoFBgdHWVoaIje3l4ymcyqDMRvhEgkQjKZ5NSpU4TDYerr66mrq2Pt2rVYLBYikYiYCzIzM0MqlVrpJQtMJhMOh4NUKiWGJS6cP1NRUUEqlWJ6+h1nT6/Xi9frZffu3eK61jYuTp06xV/8xV8wMDBALBa7LZmcN998k9HRUf74j/+Yj3zkI7hcLoxGI5/97GeZm5vjxRdfZGBggJdeekn0z2iT3Tdv3kxTUxNTU1MUi0UqKiqEO9ndVvK2ihgF/tuC75fUzUlyx/N1oHLB96MrtA6J5H3JosWGTqfDbDZTWVmJ1+sVswDC4TCDg4P09fVdNyDVBIvRaMRsNos+B6fTSX19Pa2trWzatInm5macTudlO50XL14UpUm3G61Xo7Kykvb2durq6igvL0en05HP55mcnGRkZISxsTEmJydFM+9i0AI3eKfZOB6Pi4zJQncjk8kkJjXncjkmJiaYnJxkeHgYn89Hd3f3Za5Lqqri9XppaGgQvR5aCUsqlRJWrtpzaYPoEomEyOAsFr1eT1VVFVarFZ/PRyaTIRKJEI/HxY62JpzGxsbYvn07ZrNZZG4cDofIKr2XW9RqR3ttNZGl9WdoAjCZTDI/Py9KFlcLmiPbQgtmbQieVpJUW1tLIpEgEAgQDoeJRCI4nU58Ph8NDQ1UVlaKHqRUKsXw8DBvvvmmeE1uh4CMx+P09vbS19fH+Pg4zc3NKIoiMi7bt2/HaDQK44tkMimyux0dHdTX14sSRUDYf6/E76n3A5dKdb6wwsuQrFJUVf36Sq9BInk/syixoSgKLpeLxsZG9u/fT3t7O4qi8Ktf/YqvfOUrnDp16rqBnmYtuWPHDmpra9m0aRPpdJpXXnmF9vZ2fuM3foNNmzbR0NBAOp0mGAzy05/+lOHhYZEtmJ2dve072FqfSnl5OV1dXezatQufz4fJZCKXyzEzM8N3v/tdQqEQqVQKv9+/JL0a8/PzRCIRjhw5wttvv83c3Bytra2Ew2FSqRRTU1MiQG1ra2Pz5s0MDAwwPj7O2NgY0WiUcDgsGmCtVutlx9csXbdt2yaEnZYx2b17N3/0R3/E3NwckUiEWCxGJBLhzTffJBAIcOjQoUWViCmKQnl5OX/5l39Jc3MzFRUVhEIhzp49y+nTpzl9+jT9/f3E43Fefvll/H4/LS0t1NbW0tTURE1NDb/+67/OyMgI09PTXLx4kXQ6TSQSueMyHAaDAYvFwo4dO+js7OSee+4R8zmi0SiDg4OMjIwwNze3aob6aVPWa2tr2bhxI/X19VRVVdHf3082m+V3f/d3he0vlETjP/3TP/HNb36TTZs2UVtbKzKkBoOBaDTKCy+8wOHDh0W52O1+H//u7/6On/zkJ/zZn/0Zu3fvpqWlBYvFwv3338/mzZtpaWkRPVvT09Mkk0n27duH1+vFbrcLcRiPxzlz5gzBYFDO2ZBIJBLJ+4pFZzby+bzIZGgTqsPhsMg4wDv152VlZWInXCupaG1tpbGxkQ0bNlBVVUVXVxfpdJp0Ok1tbS0tLS2YzWYymQwXL14kGAxy9uxZxsfHGR0dFbXRt/MPtzYR3GAwYLPZRE23Xq8Xu5nz8/NiFoA2l2Gp1qiVZORyOfr7+8nlcmI43MTEhAj4tR3l0dFR/H4/oVCITCZDMplEURROnz4t+jgWHttoNLJ582bheqQNJNQyBSaTiaqqKpHB0XpCtMffCoqiUFNTQ2trK+vWraOyspKysjIcDgdGoxGj0YjT6SQWi4lsx9jYGCdPniSTyYg5CGazWcw4CQQCzM7OCnOCO0lwaJ8Pt9tNdXW1ODft2kokEktmNrBYTCYTO3fuFKVPa9asYf369dTX1+Pz+chmsyQSCaqrq3G73cKhCWDbtm088sgjVFZW4vF4cDgc6HQ6ZmdnmZiYYHBwkGAwuGLBufbZPX36tCjb09ZuMpmE2LBYLJSXl5PJZLDb7ej1euFmZTQaL8sASiQSiUTyfmJRYkNVVZLJJBMTExw8eJB0Ok1DQwORSEQE5FqZldPpZN++fWQyGY4dOyasJD/+8Y/zwAMP4PV6xQCsYrHInj17hJtLMBjk5MmTvPjii5w6dYre3l6SyaQoGbrdaAG3zWajvLycsrIyITa0sp18Pk8ikWB6eloIgKVEK5U6duwYx48fv+x2Db/fz+HDh686NC4YDPLtb39b7LxqFItFjEYj0WgUl8slpo9rok6n0/HpT3+a+++/n4mJCSKRCOl0etFiSqfT8dhjj7Fp0yY6OztFWZ3X66W+vp5NmzaRSqVEI7s2jfk73/kOXV1dPPTQQ+zYsYOOjg727NnDjh07GBoaYmBggOHhYaampoT4Ws1ogk6b1VBXV0d9fT1Op1PMbNGyNel0+rJZHSuFz+fjq1/9KjabjUAgQFVVFU1NTeLzoNfrmZmZwefzCQc0LWDfvXs399xzD5FIhGw2i9vtJp/P89xzzzE2NkZPTw+BQGDFzrFQKFAoFPj2t7/N888/z8aNG0WW1Wg00tbWdpn5QrFYJBKJiPOxWCysWbOGYrEoSt8kEolEInk/sSRlVA0NDezfv5+2tjYsFgvV1dW0t7djt9tJJBJih3P//v0A7N27VwQhmzdvvmy3M5fLkUgkOHr0qBhaNjU1RTAYpKenh7GxMeHqtBJCQ3PeampqoqmpSVjPms1mEbjHYjHm5uaIx+Nks9ll3ZW9nhPXtUSOJoiudnuxWOTs2bPY7XZGR0fJ5XKXleocPHiQmZkZAoEAkUiECxcuMDc3t6iAUFVV+vv7KRaL7Ny5E4/Hg9VqFfMKVFXFYrHw+OOP09jYyM9//nOCwSDZbJZAIMBbb70lsh9Op1PMPWlqamLPnj2Mjo6KpvGl6C1ZLjQHqoqKCurr60U5mclkErvjWsncSmc1FEWhs7OTDRs24PV6sVgsFItF7HY7UBK009PTvPHGGwSDQaLRKPX19WzevFm4bGkN/x6PR2QDNSeoYDAorItXmnw+Tzwe55//+Z+pra0VM27q6+vxer2Ul5czMTHB7Owsk5OTIttWU1NDeXk5ZrOZnTt3cvbsWbGhIpFIJBLJ+4FFiQ2dTkd5eTkbNmzgN3/zN8UsgLq6OtatW0dtbS3z8/Ns3LiRlpYW7r//fkwmEwC5XI5UKiWm7Gp9AbFYjOHhYT772c+K4VlalmQ11DmbTCbKysrYunUrra2t7N69m8bGRiE2isUiwWCQwcFBMTF4Naz7alwr4Mnn8/zyl7+85uOee+454OoOYotZy6FDhzh16hSVlZVUVlZSXV1NXV0dzc3NOBwOLBYLjz76KA8//DB6vZ6zZ89y+PBhRkZG6OvrE9bI27Zto7KyErfbjdlsZv/+/cLJaGZmZlWLDU0waRawjY2N+Hw+zGYz2WyW2dlZgsGgENwreW3p9Xr27dvHxo0b8Xg8GI1G0Y+Rz+c5duwYL7/8MocOHSIYDALQ1dXFN7/5TVEaphkUWCwWCoUCR48e5fz584yNjYnyv9UQmGvN6l/60pcwGo00NDRQXV3N+vXrWb9+PTt37uSHP/whr732GrlcDrPZzNatW+ns7KStrQ2Hw8HmzZtZu3atKLFajb8TJBKJRCJZahYlNrSSAa35uLy8HI/Hw5o1a3jiiSfEH1Sfz4fNZhM1/lovwMTEBB6PB7vdzszMDNFolFdffZWxsbHL7DwXM4htKdHpdLhcLtrb2/ngBz9IfX099fX1uFwuFEURu7Lnzp1jYGDgsgzMSq99OViOc8pkMrz++uu4XC5cLhdut1uUEjU0NNDR0YHdbqerq0sIkNnZWc6ePcvc3By//OUviUajtLS0sHXrVioqKti7dy8dHR2YTCZ6enp48cUXV+0U51wuJ+bTGAwGTpw4QTKZpLOzE4PBQHNzM/l8Hr/fz7Fjxzhx4sSKBK4LrWwXupYlEgmGh4d55ZVXGBwc5MKFCxQKBXw+Hw8//DBr164Vvw8WopVMtrW1UVZWxvnz51FVlZGREVEutlooFArMzs6STCYJh8MMDQ1x/Phxzpw5w8TEhCg3zGazor/KarWiKAqPPvoodrudb33rW5w4cWKlT0UikUgkkmVn0T0bsViMixcv0t/fT2trqyhj6ejoEPfTMh6aXWo0GiUYDNLb2ysmbp8+fZpAIMDf/M3fMDMzsyoHsSmKgsPhoLa2lq1bt+J2u0UtPcDMzAxjY2MMDAyIBvmVKve6U5mfn+f48eOXNeB7vV42bdpEY2MjFouF2tpa6uvrcbvdqKqK3+9nenqaRCLBuXPnhBvY1q1bsdlswpJYE7pGo1EEx6sNzWY4FAphNBrp6emhUCjQ3t6O0WjE7XbT2NjIrl27CIVCnD59ekWEk9ZbotfrhTlAsVgklUrx9ttv841vfEOsyev1Ul1dzf33309bW5uYWaH9XBssCVBeXo7VaqW2tpZ4PI7NZqNQKKwqsVEsFolGo0SjUaampjh//vxV7xcKhfB6vaTTaex2O6qq0tjYSE1NDa+//jpvv/22/N0gkUgkkrueJXGjmpyc5O///u9pbGxk7969xGIxotEofr+faDTKunXrMJvNjI6OkkwmmZ6eJhqNMj09jdVqxWQyCTer2dnZVevYotPpcDqdYkd94eA7gNdff5033niDnp4ecT6rMaBd7SwcUKjN74jFYpw8eZKTJ09SU1PD5z73OWpra2lrayOVSvHEE0+Ix2t9MvF4HChNVbdYLHR1dTE7OysGroXD4ZU6xRtm4VwVKL02mpjS3I1WImDVen6GhoYwGo188IMfpFgsEggE8Hg8fOpTnxJuYpqZwtatW3E6ncKdKRaLCec57VwaGhowmUw8/fTT9Pb28uqrrwo3skKhsCpKqm6Gvr4+/vRP/5Tm5ma2bdsmenCsViuNjY1cvHhRNo1LJBKJ5K5m0WJDc6Q6fvw4oVAIs9ks5i/09PQQDAaZnZ3FZDJx+vRp4SakDZu7U9AsLF0ul3BK0lx1tMBYs2INBAKLnhT+fkYLnrXyufn5edLpNDqdDr/fj9fr5TOf+QwVFRU4nU7sdjtlZWXi8dr07Xw+TyqVwuPxoNfrsdvtVFZWUlVVRSQSWaGzuzmuHOKoiY1IJEIqlVrR4FsbrujxeEgkEkBpTovb7Wbbtm2Ul5fjdDpFWZHH4xElV7lcjunpabExEQ6HyeVyeDwe3G43lZWVpFIpampq0Ol0qKpKJpMRpUl3CtFolAMHDtDS0iJEc1dXF/l8XvSsSLEhkUgkkruZRYsNKJW+jI+PMzMzw7lz58RQuXQ6TaFQ4MCBA2KyuBaY30kBA0BlZSUVFRXs2LGDTZs24XA4MBgMKIpCJBJhYmKC8fFxotHoirsE3Y0sLHuam5vj6aefpra2lr1799LY2Mg999xDWVkZTqcTl8uF0+kUfQULS348Hg+bNm0iGo0yOTm5wmd1bRZaR2tio1AoEI/HGRsb48iRIwwMDKyYoNUyG2fOnGFwcJDp6Wm6urr45Cc/iclkEgJBVVVmZ2eZn58XzeMXLlzg+PHjfP3rXxcCoqWlhaqqKpxOJ01NTbS3t1NbW8tzzz1HT08PL7/8Mj09PQwODhKLxe6YjQpVVZmfnxf2y2+88QbV1dVcuHCBcDgshYZEIpFI7nqWRGxAqZxKK3m5siFa2/W8U1EUhYqKChobG6mrq6O2tlYEgoDYaV4490CKjeWjUCgwPDzM9PQ0Ho+HeDyOy+WiqqoKr9cLvOOUpTUwA6LxWrNUXs1o2ZxUKiVmymhD/QwGAx6PRwzRW0nxnk6nmZ+f5+TJkxiNRoLBIBaLBYvFIrKXExMTom+hUChw5swZenp66O3tJZ/Pi7K5dDqN3+/HbrfT2NiIwWCgurqabDYrbGUHBwdX5DwXg5bJyeVy5PN5YrGYyORIJBKJRHK3s2RiQ+NOq6m+EXQ6Hffeey/r1q3jkUcewev1ioFlUCoduXjxInNzcyKbI8XG8pNMJnn55Zex2Wy89NJLop9mIT6fj2effZbJyUm+/OUvE4vFiMfjq35nPBQKCRvoYrFIe3s7Pp+Puro61q5dyyc/+UleeOEFQqEQkUiETCazYmstFArMzMxw4MABjh49itfrpaWlRcw0CYfDIoNRKBREidvCNfv9fqampqipqWFsbAybzUZlZSW1tbW0t7fT2tpKMBjkpz/96R1dnphIJEgmk/L3g0QikUjeNyy52LhbMZvN2Gw20Ri+sJZeG/iVyWRWfRB7N6GVqCSTSaAUyJnN5svuEwqFOHr0KOFwWAwBvBOCVS07aDQaRY+QNgyuUCgwPz8v/q30+WglbsVikbm5ObGDr60tlUqJ/plisXjVz4jWizI0NEQ+n+fo0aM0NjYKF7tMJsPU1NRdsZkhhYZEIpFI3k9IsXEDKIqCyWQSQZ829VgjnU4zMzMj3KdkMHH70ATHtRq+dTod3/nOd0Swe6ewsMdEKw+z2+3k83kxy2JmZmbF+xcWNvNDKcsRCoUIhUKi1FATCNcTCsVikZMnT3L27Fl6enpoa2tj7969zM3N4ff7efvtt5f3ZCQSiUQikSw5UmxcBy2DkcvlSCaTJJNJ9Ho9RqOReDxOIBDgzTffpK+vj0gkclfsvN5NqKpKIpG4Y98Xo9GI2WwWIle7HrVMh/b9ahS4C4dx3sz6CoUC4XCY/v5+jEYjsViM6enpO8ZBTCKRSCQSyTtIsXEdtB3mbDZLKpUiHo+LgXODg4P8+Mc/Znh4mP7+fkKh0KoM+t7PqKp6R2U0rsRoNIrSPU1s6PV6zGYzBoNBNIivRm5WZGgUi0VCoRDhcBi/3y+OdacKRolEIpFI3s9IsXEdtHr0qakpjEYjZWVluFwuKioqOHPmDL29vYyPjxMKhaS7jGTJ0HoYDhw4wODgIHa7nbq6Ourr67HZbOzatYuqqiq2bt3KD37wA06dOnXX9QstFBhSxEskEolEcmcixcYNUCwWmZmZwWw2Mzg4iM1mw+l0Mjw8zMjICNFoVE4Llyw5xWKR8+fPMzExwfbt28nn83i9XhwOBxUVFRiNRsrLyzl8+DBnzpxZteVUi+FuOx+JRCKRSN5vSLFxHbTMxrlz5+jv7+fIkSMYDAaMRiPpdJpQKEQ+n5eN4ZJlQ5uePTo6Sm9vL06nkzVr1hAMBpmZmeH8+fPk83l5/UkkEolEIll1SLFxg2iDCcPhMAB6vV7WkUtuG5qbViAQwG63k8lkmJubY25ujmQyKa9DiUQikUgkqxJF7oZKJBKJRCKRSCSS5UB3/btIJBKJRCKRSCQSyc0jxYZEIpFIJBKJRCJZFqTYkEgkEolEIpFIJMuCFBsSiUQikUgkEolkWZBiQyKRSCQSiUQikSwLUmxIJBKJRCKRSCSSZeH/BzS+D3fKbetsAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 1008x720 with 18 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print('The train set has {} samples.'.format(len(trainset)))\n",
    "print('The test set has {} samples.'.format(len(testset)))\n",
    "num_plotted_samples = 9\n",
    "fig, axs = plt.subplots(num_plotted_samples, 2,figsize=(14, 10))\n",
    "for ind in range(num_plotted_samples): \n",
    "    img, label = trainset[ind]\n",
    "    axs[ind,0].imshow(img.squeeze().numpy(), cmap='Greys_r')\n",
    "    axs[ind,1].text(0,.5,str(label.numpy()), fontsize=20)\n",
    "    axs[ind,0].axis('off')\n",
    "    axs[ind,1].axis('off')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "92f4fae9",
   "metadata": {},
   "source": [
    "Now we have our dataset created, let's go to the second part: designing a model"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "feb9d9cf",
   "metadata": {},
   "source": [
    ">## 2. Design a text recognition model\n",
    ">As the next step, design a text recognition network that can identify images with different lengths of texts we generated in the previous >step. We will use this network as the baseline for the next steps of this test task. You should choose an architecture that has an adequate >capacity for the given complexity of the dataset. Please determine and motivate your evaluation criteria (i.e. what metric you will use to evaluate the performance of your model and why). Feel free to make reasonable assumptions as you see fit\n",
    "\n",
    "After try several models, and given the time available for solving the task, let's try a CNN consisting in 3 convolutional layers with batch normalization and pooling steps.\n",
    "\n",
    "The most important part of all the models that we want to solve this tasks is the flatten step, due to we want process the whole sequence for having at the output a tensor of shape [B, P, L] beeing B the Batchsize, P the 47 probabilities and W the length of the sequence (we have to compute the 47 softmax probabilities for each character in the sequence) "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 270,
   "id": "1b2801d9",
   "metadata": {},
   "outputs": [],
   "source": [
    "class CNN(nn.Module):\n",
    "    def __init__(self,height,nlabels,prob): #Nlabels will be 10 in our case\n",
    "        super().__init__()\n",
    "        \n",
    "        self.height = height\n",
    "        self.nlabels = nlabels\n",
    "        # convolutional layer (sees 28x28x1 image tensor)\n",
    "        self.conv1 = nn.Conv2d(in_channels=1, out_channels=6, \n",
    "                               kernel_size=3, stride=1, padding=1)\n",
    "        \n",
    "        # convolutional layer (sees 12x12x16 tensor)\n",
    "        self.conv2 = nn.Conv2d(6, 16, 3, padding=1)\n",
    "        \n",
    "        self.conv3 = nn.Conv2d(16, 32, 7, stride=7, padding=0)\n",
    "        \n",
    "        # Max pool layer\n",
    "        self.pool = nn.MaxPool2d(2, 2)\n",
    "\n",
    "        # Linear layers\n",
    "        self.linear1 = nn.Linear(32,256)\n",
    "        \n",
    "        self.linear2 = nn.Linear(256, nlabels)\n",
    "    \n",
    "        self.activation = nn.ReLU()\n",
    "        \n",
    "        self.logsoftmax = nn.LogSoftmax(dim=1) \n",
    "\n",
    "        self.BN_1 = nn.BatchNorm2d(num_features=6)\n",
    "\n",
    "        self.BN_2 = nn.BatchNorm2d(num_features=16)\n",
    "        \n",
    "        self.BN_3 = nn.BatchNorm2d(num_features=32)\n",
    "\n",
    "        self.BN_4 = nn.BatchNorm1d(256)\n",
    "        \n",
    "        # Dropout module with 0.2 drop probability\n",
    "        self.dropout = nn.Dropout(p=prob)\n",
    "        \n",
    "    def forward(self, x):\n",
    "        # Pass the input tensor through the CNN operations\n",
    "        input_width = x.shape[3]\n",
    "        output_width = int(input_width/self.height)\n",
    "        x = self.conv1(x) \n",
    "        x = self.BN_1(x)\n",
    "        x = self.activation(x) \n",
    "        x = self.pool(x)\n",
    "        x = self.dropout(x) \n",
    "        x = self.conv2(x)\n",
    "        x = self.BN_2(x)\n",
    "        x = self.activation(x)\n",
    "        x = self.pool(x)\n",
    "        x = self.dropout(x) \n",
    "        x = self.conv3(x)\n",
    "        x = self.BN_3(x)\n",
    "        x = self.activation(x)\n",
    "        x = self.dropout(x)\n",
    "        # Pass the tensor through the Dense Layers\n",
    "\n",
    "        # In the following lines we change the tensor shape to get te result asked above.\n",
    "        x = x.squeeze(2)\n",
    "        x = x.transpose(1,2)\n",
    "        x = x.reshape(-1,32)\n",
    "        x = self.linear1(x)\n",
    "        x = self.BN_4(x)\n",
    "        x = self.activation(x)\n",
    "        x = self.dropout(x) \n",
    "        x = self.linear2(x)\n",
    "        x = x.view(-1,output_width, self.nlabels)\n",
    "        x = x.transpose(1,2) # Let's have [batchsize, nlabels, outputwidth] at the output\n",
    "        x = self.logsoftmax(x) \n",
    "        return x"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 271,
   "id": "a2e8e827",
   "metadata": {},
   "outputs": [],
   "source": [
    "my_net = CNN(height=28, nlabels=47,prob=0.25)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 272,
   "id": "730e0fdb",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([252, 252, 224, 224, 224, 196, 140, 140,  84,  84]) tensor([9, 9, 8, 8, 8, 7, 5, 5, 3, 3])\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "torch.Size([10, 47, 9])"
      ]
     },
     "execution_count": 272,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "demo_dataloader = DataLoader(testset, batch_size=10, shuffle=False, collate_fn=collate_semnist)\n",
    "dataiter = iter(demo_dataloader)\n",
    "img, target, width, length = dataiter.next()\n",
    "print(width, length)\n",
    "out = my_net.forward(img)\n",
    "out.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "93a57d11",
   "metadata": {},
   "source": [
    "The following class will take a model (as the example above) and will performs the trainloop. Note that sequences has different widths (between 84 and 280 pixels (3 - 10 characters)). The dataloader has padded the sequence (see dataloader). In the loss function, we should'nt take into account the padded elements, so the solution we provide is to mask the lost function (multiplying the computed loss function by 0) before sume up and compute the loss of the batch. So the gradients are not affected by that."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 330,
   "id": "0f8f9567",
   "metadata": {},
   "outputs": [],
   "source": [
    "class CNN_extended_GPU(CNN):\n",
    "    \n",
    "    def __init__(self,height,nlabels,prob,epochs=100,lr=0.001):\n",
    "        \n",
    "        super().__init__(height,nlabels,prob)  \n",
    "        \n",
    "        self.lr = lr #Learning Rate\n",
    "        \n",
    "        self.optim = optim.Adam(self.parameters(), self.lr)\n",
    "        \n",
    "        self.epochs = epochs\n",
    "        \n",
    "        self.criterion = nn.NLLLoss(reduction='none')             \n",
    "        \n",
    "        # A list to store the loss evolution along training\n",
    "        \n",
    "        self.loss_during_training = [] \n",
    "        \n",
    "        self.valid_loss_during_training = []\n",
    "        \n",
    "        self.device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\n",
    "\n",
    "        self.to(self.device)\n",
    "        \n",
    "    def trainloop(self,trainloader,validloader):\n",
    "        \n",
    "        # Optimization Loop\n",
    "        \n",
    "        for e in range(int(self.epochs)):\n",
    "            \n",
    "            start_time = time.time()\n",
    "            \n",
    "            # Random data permutation at each epoch\n",
    "            \n",
    "            running_loss = 0.\n",
    "            \n",
    "            for images, labels, widths, lengths in tqdm(trainloader):\n",
    "                \n",
    "                # Move input and label tensors to the default device\n",
    "                images, labels = images.to(self.device), labels.to(self.device)  \n",
    "        \n",
    "                self.optim.zero_grad()  #TO RESET GRADIENTS!\n",
    "            \n",
    "                out = self.forward(images)\n",
    "\n",
    "                #Your code here\n",
    "                loss = self.criterion(out,labels)\n",
    "                \n",
    "                mask = torch.ones_like(loss)\n",
    "                for ind, length in enumerate(lengths):\n",
    "                    mask[ind,length:] = 0\n",
    "                 \n",
    "                # Apply a mask to solve different lengths sequences\n",
    "                loss = torch.mean(loss*mask)\n",
    "                \n",
    "                running_loss += loss.item()\n",
    "\n",
    "                #Your code here\n",
    "                loss.backward()\n",
    "                \n",
    "                #Your code here\n",
    "                self.optim.step()\n",
    "                \n",
    "                \n",
    "            self.loss_during_training.append(running_loss/len(trainloader))\n",
    "            \n",
    "            # Validation Loss\n",
    "            \n",
    "            # Turn off gradients for validation, saves memory and computations\n",
    "            with torch.no_grad():\n",
    "                \n",
    "                self.eval()\n",
    "                \n",
    "                running_loss = 0.\n",
    "                \n",
    "                for images,labels, widths, lengths in tqdm(validloader):\n",
    "                    \n",
    "                    # Move input and label tensors to the default device\n",
    "                    images, labels = images.to(self.device), labels.to(self.device)                    \n",
    "                                   \n",
    "                    out = self.forward(images)\n",
    "\n",
    "                    #Your code here\n",
    "                    loss = self.criterion(out,labels)\n",
    "                    \n",
    "                    mask = torch.ones_like(loss)\n",
    "                    \n",
    "                    for ind, length in enumerate(lengths):\n",
    "                        mask[ind,length:] = 0\n",
    "                        \n",
    "                    # Apply a mask to solve different lengths sequences   \n",
    "                    loss = torch.mean(loss*mask)  \n",
    "\n",
    "                    running_loss += loss.item()   \n",
    "                    \n",
    "                self.valid_loss_during_training.append(running_loss/len(validloader))    \n",
    "\n",
    "            # set model back to train mode\n",
    "            self.train()\n",
    "\n",
    "            if(e % 1 == 0): # Every 10 epochs\n",
    "\n",
    "                print(\"Epoch %d. Training loss: %f, Validation loss: %f, Time per epoch: %f seconds\" \n",
    "                      %(e,self.loss_during_training[-1],self.valid_loss_during_training[-1],\n",
    "                       (time.time() - start_time)))\n",
    "                \n",
    "    def eval_performance(self,dataloader):\n",
    "        \n",
    "        loss = 0\n",
    "        accuracy = 0\n",
    "        \n",
    "        self.eval()\n",
    "        # Turn off gradients for validation, saves memory and computations\n",
    "        with torch.no_grad():\n",
    "\n",
    "            for images,labels, widths, lengths in tqdm(dataloader):\n",
    "                # Move input and label tensors to the default device\n",
    "                images, labels = images.to(self.device), labels.to(self.device) \n",
    "                probs = self.forward(images)\n",
    "                top_p, top_class = probs.topk(1, dim=1)\n",
    "                equals = (top_class.squeeze() == labels)\n",
    "                mask = torch.ones_like(equals)\n",
    "                for ind, length in enumerate(lengths):\n",
    "                    mask[ind,length:] = 0\n",
    "                equals*=mask\n",
    "                accuracy += torch.sum(equals.type(torch.FloatTensor))/torch.sum(lengths)\n",
    "                \n",
    "            self.train()\n",
    "            return accuracy/len(dataloader)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 331,
   "id": "94847d88",
   "metadata": {},
   "outputs": [],
   "source": [
    "demoset = SEMNIST('~/.pytorch/SEMNIST_data/', dataset = 'demo', transform = transform)\n",
    "demoloader = DataLoader(demoset, batch_size=64, shuffle=True, collate_fn=collate_semnist)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 344,
   "id": "c9342322",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████| 1563/1563 [05:04<00:00,  5.14it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:13<00:00, 18.55it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 0. Training loss: 1.183364, Validation loss: 0.744315, Time per epoch: 317.621748 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████| 1563/1563 [03:25<00:00,  7.59it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:13<00:00, 18.47it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1. Training loss: 0.916602, Validation loss: 0.645137, Time per epoch: 219.382169 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████| 1563/1563 [03:24<00:00,  7.65it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:13<00:00, 19.10it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 2. Training loss: 0.858073, Validation loss: 0.639641, Time per epoch: 217.482807 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████| 1563/1563 [03:24<00:00,  7.63it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:13<00:00, 18.00it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 3. Training loss: 0.825386, Validation loss: 0.614321, Time per epoch: 218.748411 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████| 1563/1563 [03:27<00:00,  7.53it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:13<00:00, 18.82it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 4. Training loss: 0.805930, Validation loss: 0.610093, Time per epoch: 220.858059 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████| 1563/1563 [03:24<00:00,  7.63it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:13<00:00, 18.95it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 5. Training loss: 0.790907, Validation loss: 0.614778, Time per epoch: 217.990107 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████| 1563/1563 [03:21<00:00,  7.75it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:13<00:00, 18.99it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 6. Training loss: 0.783140, Validation loss: 0.611029, Time per epoch: 214.728812 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████| 1563/1563 [03:17<00:00,  7.90it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:12<00:00, 19.35it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 7. Training loss: 0.774072, Validation loss: 0.607949, Time per epoch: 210.809834 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████| 1563/1563 [03:20<00:00,  7.79it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:12<00:00, 19.37it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 8. Training loss: 0.767742, Validation loss: 0.623321, Time per epoch: 213.432980 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████| 1563/1563 [03:19<00:00,  7.84it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:12<00:00, 19.47it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 9. Training loss: 0.761119, Validation loss: 0.635046, Time per epoch: 212.187643 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████| 1563/1563 [03:17<00:00,  7.91it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:12<00:00, 19.91it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 10. Training loss: 0.756484, Validation loss: 0.622928, Time per epoch: 210.259603 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████| 1563/1563 [03:15<00:00,  7.99it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:12<00:00, 19.65it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 11. Training loss: 0.752934, Validation loss: 0.635951, Time per epoch: 208.305642 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████| 1563/1563 [03:15<00:00,  7.98it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:12<00:00, 19.62it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 12. Training loss: 0.748355, Validation loss: 0.625452, Time per epoch: 208.683239 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████| 1563/1563 [03:15<00:00,  7.98it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:12<00:00, 19.24it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 13. Training loss: 0.745670, Validation loss: 0.626214, Time per epoch: 208.754853 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████| 1563/1563 [03:15<00:00,  7.98it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:12<00:00, 19.63it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 14. Training loss: 0.742549, Validation loss: 0.623819, Time per epoch: 208.510983 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████| 1563/1563 [03:17<00:00,  7.93it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:12<00:00, 19.24it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 15. Training loss: 0.737951, Validation loss: 0.633786, Time per epoch: 210.215110 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████| 1563/1563 [03:17<00:00,  7.93it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:12<00:00, 19.43it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 16. Training loss: 0.736875, Validation loss: 0.648170, Time per epoch: 210.074216 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████| 1563/1563 [03:17<00:00,  7.89it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:12<00:00, 19.51it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 17. Training loss: 0.735045, Validation loss: 0.642328, Time per epoch: 210.804376 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████| 1563/1563 [03:16<00:00,  7.96it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:12<00:00, 19.40it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 18. Training loss: 0.733452, Validation loss: 0.639849, Time per epoch: 209.252223 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████| 1563/1563 [03:17<00:00,  7.91it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:12<00:00, 19.57it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 19. Training loss: 0.731166, Validation loss: 0.653227, Time per epoch: 210.426390 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "my_net = CNN_extended_GPU(height=28, nlabels=47,prob=0.2, epochs=20)\n",
    "my_net.trainloop(trainloader,testloader)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 352,
   "id": "4e2375ec",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████| 16/16 [00:02<00:00,  7.55it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:14<00:00, 17.21it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 0. Training loss: 2.226022, Validation loss: 2.331234, Time per epoch: 16.651561 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████| 16/16 [00:02<00:00,  7.98it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:13<00:00, 18.00it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1. Training loss: 1.805569, Validation loss: 1.999347, Time per epoch: 15.896919 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████| 16/16 [00:01<00:00,  8.08it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:13<00:00, 18.04it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 2. Training loss: 1.488909, Validation loss: 1.568286, Time per epoch: 15.839876 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████| 16/16 [00:01<00:00,  8.02it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:13<00:00, 18.58it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 3. Training loss: 1.228361, Validation loss: 1.272163, Time per epoch: 15.453488 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████| 16/16 [00:01<00:00,  8.02it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:13<00:00, 18.46it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 4. Training loss: 1.030923, Validation loss: 1.146094, Time per epoch: 15.542779 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████| 16/16 [00:01<00:00,  8.02it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:13<00:00, 18.73it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 5. Training loss: 0.887962, Validation loss: 1.082083, Time per epoch: 15.350451 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████| 16/16 [00:01<00:00,  8.05it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:13<00:00, 18.71it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 6. Training loss: 0.772632, Validation loss: 1.005794, Time per epoch: 15.352922 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████| 16/16 [00:01<00:00,  8.07it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:13<00:00, 18.59it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 7. Training loss: 0.686654, Validation loss: 0.955472, Time per epoch: 15.436615 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████| 16/16 [00:01<00:00,  8.05it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:13<00:00, 18.74it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 8. Training loss: 0.609461, Validation loss: 0.923406, Time per epoch: 15.332375 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████| 16/16 [00:01<00:00,  8.08it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:13<00:00, 18.67it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 9. Training loss: 0.547682, Validation loss: 0.925390, Time per epoch: 15.375755 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████| 16/16 [00:01<00:00,  8.08it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:13<00:00, 18.54it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 10. Training loss: 0.495248, Validation loss: 0.931703, Time per epoch: 15.470349 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████| 16/16 [00:02<00:00,  7.93it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:13<00:00, 18.67it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 11. Training loss: 0.441823, Validation loss: 0.913137, Time per epoch: 15.412693 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████| 16/16 [00:01<00:00,  8.05it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:13<00:00, 18.14it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 12. Training loss: 0.395336, Validation loss: 0.891189, Time per epoch: 15.774478 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████| 16/16 [00:01<00:00,  8.00it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:13<00:00, 18.59it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 13. Training loss: 0.356033, Validation loss: 0.909158, Time per epoch: 15.448979 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████| 16/16 [00:01<00:00,  8.08it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:13<00:00, 18.68it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 14. Training loss: 0.320406, Validation loss: 0.941558, Time per epoch: 15.370599 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████| 16/16 [00:02<00:00,  7.99it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:13<00:00, 18.56it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 15. Training loss: 0.289350, Validation loss: 0.930182, Time per epoch: 15.474928 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████| 16/16 [00:01<00:00,  8.04it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:13<00:00, 18.70it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 16. Training loss: 0.263833, Validation loss: 0.983713, Time per epoch: 15.361374 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████| 16/16 [00:01<00:00,  8.05it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:13<00:00, 18.81it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 17. Training loss: 0.234163, Validation loss: 0.980185, Time per epoch: 15.285073 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████| 16/16 [00:01<00:00,  8.06it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:13<00:00, 18.70it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 18. Training loss: 0.215379, Validation loss: 0.983982, Time per epoch: 15.362872 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████| 16/16 [00:02<00:00,  7.81it/s]\n",
      "100%|█████████████████████████████████████████| 250/250 [00:13<00:00, 18.77it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 19. Training loss: 0.196771, Validation loss: 1.009724, Time per epoch: 15.368498 seconds\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "my_no_dropout = CNN_extended_GPU(height=28, nlabels=47,prob=0, epochs=20)\n",
    "my_no_dropout.trainloop(demoloader,testloader)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 355,
   "id": "00e8d1f3",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|█████████████████████████████████████████| 250/250 [00:13<00:00, 18.21it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor(0.5912)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████| 16/16 [00:00<00:00, 19.16it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor(0.6004)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|█████████████████████████████████████████| 250/250 [00:14<00:00, 17.20it/s]\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor(0.5720)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|███████████████████████████████████████████| 16/16 [00:00<00:00, 18.79it/s]"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor(0.9299)\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "print(my_net.eval_performance(testloader))\n",
    "print(my_net.eval_performance(demoloader))\n",
    "print(my_no_dropout.eval_performance(testloader))\n",
    "print(my_no_dropout.eval_performance(demoloader))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5fd3270a",
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 356,
   "id": "54cbb662",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<matplotlib.lines.Line2D at 0x7fc4c17b9060>]"
      ]
     },
     "execution_count": 356,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(my_net.valid_loss_during_training)\n",
    "plt.plot(my_net.loss_during_training)\n",
    "plt.plot(my_no_dropout.valid_loss_during_training)\n",
    "plt.plot(my_no_dropout.loss_during_training)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 361,
   "id": "16c76b90",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([ 9,  5, 39, 21,  3, 28,  1])\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "img, label = demoset[56]\n",
    "plt.imshow(img.squeeze().numpy(), cmap='Greys_r')\n",
    "img = img.reshape(1,img.shape[0], img.shape[1], img.shape[2])\n",
    "print(label)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 362,
   "id": "91e8626d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[[ 4,  5, 39, 21,  3, 10,  1]]])"
      ]
     },
     "execution_count": 362,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "probs = my_no_dropout.forward(img)\n",
    "top_p, top_class = probs.topk(1, dim=1)\n",
    "top_class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 412,
   "id": "a38df25c",
   "metadata": {},
   "outputs": [],
   "source": [
    "def eval_performance(model, dataloader, mode='seqCER'):\n",
    "    \"\"\"\n",
    "    accuracy = eval_performance(model, dataloader, mode='seqCER')\n",
    "    \n",
    "    Computes the eval performance of a given model over a given dataset.\n",
    "    \n",
    "    Args:\n",
    "    \n",
    "        model\n",
    "    \"\"\"\n",
    "        \n",
    "    loss = 0\n",
    "    accuracy = 0\n",
    "\n",
    "    model.eval()\n",
    "    # Turn off gradients for validation, saves memory and computations\n",
    "    with torch.no_grad():\n",
    "\n",
    "        for images,labels, widths, lengths in tqdm(dataloader):\n",
    "            # Move input and label tensors to the default device\n",
    "            images, labels = images.to(model.device), labels.to(model.device) \n",
    "            probs = model.forward(images)\n",
    "            top_p, top_class = probs.topk(1, dim=1)\n",
    "            equals = (top_class.squeeze() == labels)\n",
    "            mask = torch.ones_like(equals)\n",
    "            for ind, length in enumerate(lengths):\n",
    "                mask[ind,length:] = 0\n",
    "            equals*=mask\n",
    "            if mode == 'totalCER':\n",
    "                accuracy += torch.sum(equals.type(torch.FloatTensor))/torch.sum(lengths)\n",
    "            else:\n",
    "                accuracy += torch.mean(torch.sum(equals.type(torch.FloatTensor),1)/lengths)         \n",
    "        model.train()\n",
    "        return accuracy/len(dataloader)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.10.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}