ee046211_tutorial_09_self_supervised_representation_learning.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "# <img src=\"./assets/course-icon.png\" style=\"height:50px;display:inline\"> ECE 046211 - Technion - Deep Learning\n",
    "---\n",
    "\n",
    "#### <a href=\"https://taldatech.github.io\">Tal Daniel</a>\n",
    "\n",
    "## Tutorial 09 - Transfer, Representation and Self-Supervised Learning\n",
    "---"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "### <img src=\"https://img.icons8.com/bubbles/50/null/checklist.png\" style=\"height:50px;display:inline\"> Agenda\n",
    "---\n",
    "* [Pre-trained Models and Transfer Learning](#-Transfer-Learning)\n",
    "    * [Transfer Learning](#-Transfer-Learning-Definition)\n",
    "    * [Pre-trained Models](#-Transfer-Learning-with-Pre-trained-Models)\n",
    "    * [LoRA & DoRA - Low Rank Adaptation Fine-Tuning](#-LoRA-&-DoRA---Low-Rank-Adaptation-Fine-Tuning)\n",
    "* [Representation Learning and Self-Supervised Learning](#-Representation-and-Self-Supervised-Learning)\n",
    "    * [Autoencoders](#-Deep-Unsupervised-Learning---Deep-Autoencoders)\n",
    "    * [Contrastive Methods](#--Contrastive-Learning)\n",
    "* [Recommended Videos](#-Recommended-Videos)\n",
    "* [Credits](#-Credits)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "outputs": [],
   "source": [
    "# imports for the tutorial\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import time\n",
    "import os\n",
    "import copy\n",
    "\n",
    "# pytorch imports\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "from torch.utils.data import DataLoader, Dataset\n",
    "from torchvision.datasets import ImageFolder\n",
    "from torchvision import models, transforms\n",
    "import torchvision\n",
    "\n",
    "# scikit-learn imports\n",
    "from sklearn.manifold import LocallyLinearEmbedding, Isomap, TSNE\n",
    "from sklearn.decomposition import PCA, KernelPCA"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## <img src=\"https://img.icons8.com/dusk/64/000000/back-sorting.png\" style=\"height:50px;display:inline\"> Transfer Learning\n",
    "---\n",
    "* Training deep neural networks has come a long way in past years, enabling learning good mappings from inputs to outputs, whether they are images, sentences, label predictions, etc. from large amounts of labeled data.\n",
    "* However, it is usually the case that our models **struggle with generalization to unseen data during training**.\n",
    "* The traditional supervised learning paradigm **breaks down** when we don't have sufficient labeled data for the task or domain we want to train our model for.\n",
    "    * For example, if we want to train a model to detect pedestrians on night-time images, we could, in theory, **apply a model that has been trained on a similar domain**, e.g., on day-time images. In practice, however, we often experience a deterioration or collapse in performance as the model has inherited the **bias** of its training data and **doesn't know how to generalize** to the new domain.\n",
    "* **Transfer learning** allows us to deal with such scenarios by leveraging the already existing labeled data of some related task or domain. \n",
    "    * We try to store this knowledge gained in solving the source task in the source domain and apply it to a new target task or domain. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<center><img src='./assets/transfer_learning_setup.png' style=\"height:300px\"></center>\n",
    "\n",
    "* <a href=\"https://ruder.io/transfer-learning/\">Image Source</a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## <img src=\"https://img.icons8.com/nolan/64/math.png\" style=\"height:50px;display:inline\"> Transfer Learning Definition\n",
    "---\n",
    "* Transfer learning involves the concepts of a **domain** and a **task**.\n",
    "* For simplicity, we assume a binary classification setting.\n",
    "* We denote the following: a domain $\\mathcal{D}$ consists of a feature space $\\mathcal{X}$ and a marginal probability distribution $P(X)$ over the feature space, where $X=x_1, x_2, ...,x_n \\in \\mathcal{X}.$\n",
    "* Given a domain, $\\mathcal{D}= \\{\\mathcal{X}, P(X) \\}$, a task $\\mathcal{T}$ consists of a label space $\\mathcal{Y}$ and a conditional probability distribution $P(Y|X)$ that is typically learned from the training data consisting of pairs $x_i \\in \\mathcal{X}$ and $y_i \\in \\mathcal{Y}$.\n",
    "    * In the binary classification case, $y_i \\in \\{True, False\\}$.\n",
    "* Given a source domain $\\mathcal{D}_S$, a corresponding source task $\\mathcal{T}_S$, as well as a target domain $\\mathcal{D}_T$ and a target task $\\mathcal{T}_T$, the objective of transfer learning is to enable us to learn the **target conditional probability** distribution $P(Y_T|X_T)$ in $\\mathcal{D}_T$ with the information gained from $\\mathcal{D}_S$ and $\\mathcal{T}_S$.\n",
    "    * We assume $\\mathcal{D}_S \\neq \\mathcal{D}_T$ (\"domain adaptation\") or $\\mathcal{T}_S \\neq \\mathcal{T}_T$ (\"transfer learning\").\n",
    "* In most cases, a limited number of labeled target examples, which is exponentially smaller than the number of labeled source examples are assumed to be available.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## <img src=\"https://img.icons8.com/nolan/64/movie.png\" style=\"height:50px;display:inline\"> Transfer Learning Scenarios\n",
    "---\n",
    "1. $\\mathcal{X}_S \\neq \\mathcal{X}_T$ - the feature spaces of the source and target domain are different.\n",
    "    * Example: cross-lingual adaptation - in a document classification task, the documents are written in two different languages. \n",
    "2. $P(X_S) \\neq P(X_T)$ - the marginal probability distributions of source and target domain are different.\n",
    "    * Example: domain adaptation -  in a document classification task, the documents ($X$) discuss different topics.\n",
    "3. $\\mathcal{Y}_S \\neq \\mathcal{Y}_T$ - the label spaces between the two tasks are different.\n",
    "    * Example: documents need to be assigned different labels in the target task.\n",
    "    * Usually happens with scenario 4 (as it is rare for two different tasks to have different label spaces, but exactly the same conditional probability distributions).\n",
    "4. $P(Y_S|X_S) \\neq P(Y_T|X_T)$ - the conditional probability distributions of the source and target tasks are different. Very common in practice.\n",
    "    * Example: source and target documents are unbalanced with regard to their classes. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## <img src=\"https://img.icons8.com/bubbles/50/000000/car.png\" style=\"height:50px;display:inline\"> Transfer Learning Applications\n",
    "---\n",
    "* **Sim2Real** - transferring from simulation to real environments. For many machine learning applications that rely on hardware for interaction, gathering data and training a model in the real world is either **expensive, time-consuming, or simply too dangerous**. It is thus advisable to gather data in some other, less risky way.\n",
    "    * Learning from a simulation and applying the acquired knowledge to the real world is an instance of transfer learning **scenario 2**, as the feature spaces between source and target domain are the same (both generally rely on pixels), but the **marginal probability distributions between simulation and reality are different**, i.e. objects in the simulation and the source look different, although this difference diminishes as simulations get more realistic. \n",
    "    * At the same time, **the conditional probability distributions** between simulation and real world might be different as the simulation is not able to fully replicate all reactions in the real world, e.g. a physics engine can not completely mimic the complex interactions of real-world objects.\n",
    "    * Common applications include autonomus driving and robotics (where gathering data can be slow or dangerous)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<center><img src='./assets/robot_and_simulation.png' style='height:200px'></center>\n",
    "\n",
    "* <a href=\"https://arxiv.org/abs/1610.04286\">Image Source</a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "* **Domain Adaptation** - transferring between domains that share some properties. \n",
    "    * In **vision**, we commonly have many labeled data for some domain, but for the actual data that we care about, there are very few labels or none at all. Even if the training and the test data look the same, **the training data may still contain a bias that is imperceptible to humans but the model can still exploit it to overfit the training data**.\n",
    "    * In **NLP**, models trained on news data have difficulty coping with more novel text forms such as social media messages and the challenges they present. Even within one domain such as product reviews, people employ different words and phrases to express the same opinion. A model trained on one type of review should thus be able to disentangle the general and domain-specific opinion words that people use in order not to be confused by the shift in domain."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## <img src=\"https://img.icons8.com/nolan/64/system-task.png\" style=\"height:50px;display:inline\"> Transfer Learning with Pre-trained Models\n",
    "---\n",
    "* One of the fundamental requirements for transfer learning is the presence of models that perform well on source tasks.\n",
    "* The two most common fields that build upon pre-trained models to transfer between tasks and domains are computer vision and NLP."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### <img src=\"https://img.icons8.com/fluent/96/000000/image.png\" style=\"height:50px;display:inline\"> Using Pre-trained CNN Features\n",
    "---\n",
    "* Evidently, lower convolutional layers capture **low-level image features**, e.g., edges , while higher convolutional layers capture more complex details, such as body parts, faces, and other compositional features.\n",
    "* The **final fully-connected layers** are generally assumed to capture information that is relevant for solving the respective task, e.g., classification.\n",
    "* Representations that capture general information of how an image is composed and what combinations of edges and shapes it contains **can be helpful in other tasks**. This information is contained in one of the final convolutional layers or early fully-connected layers in large convolutional neural networks trained on ImageNet.\n",
    "* For a new task, we can thus simply use the off-the-shelf features of a state-of-the-art CNN pre-trained on ImageNet and train a new model on these extracted features. \n",
    "* In practice, we either **keep the pre-trained parameters fixed or tune them with a small learning rate** in order to ensure that we do not \"forget\" the previously acquired knowledge."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<center><img src='./assets/transfer_learning_vgg.jpeg' style=\"height:300px\"></center>\n",
    "\n",
    "* <a href=\"https://medium.com/datadriveninvestor/what-you-must-know-about-transfer-learning-4a6e4cb9fbad\">Image Source</a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### <img src=\"https://img.icons8.com/cotton/64/000000/olympic-torch.png\" style=\"height:50px;display:inline\"> Transfer Learning Example with PyTorch\n",
    "---\n",
    "* We will follow examples by <a href=\"https://pytorch.org/tutorials/beginner/transfer_learning_tutorial.html\">Sasank Chilamkurthy</a> and <a href=\"https://pytorch.org/tutorials/beginner/finetuning_torchvision_models_tutorial.html\">Nathan Inkawhich</a>.\n",
    "* We will train a classifier to distinguish between **ants** and **bees**.\n",
    "    * The data can be downloaded from here: <a href=\"https://download.pytorch.org/tutorial/hymenoptera_data.zip\">Download Link</a>.\n",
    "* There are two major transfer learning scenarios:\n",
    "    * **Fine-tuning the ConvNet**: Instead of random initializaion, we initialize the network with a pretrained network, like VGG for example (which is trained on ImageNet 1000 dataset). Rest of the training looks as usual (except for usually using a lower learning rate).\n",
    "    * **ConvNet as fixed feature extractor**: Here, we will freeze the weights for all of the network except that of the final fully connected layer. This last fully connected layer is replaced with a new one with random weights and only this layer is trained. The advantage is very quick training, but several parts of the model are not adapted to the new objective."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cuda:0\n"
     ]
    }
   ],
   "source": [
    "# Data augmentation and normalization for training\n",
    "# Just normalization for validation\n",
    "data_transforms = {\n",
    "    'train': transforms.Compose([\n",
    "        transforms.RandomResizedCrop(224),\n",
    "        transforms.RandomHorizontalFlip(),\n",
    "        transforms.ToTensor(),\n",
    "        transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])\n",
    "    ]),\n",
    "    'val': transforms.Compose([\n",
    "        transforms.Resize(256),\n",
    "        transforms.CenterCrop(224),\n",
    "        transforms.ToTensor(),\n",
    "        transforms.Normalize([0.485, 0.456, 0.406], [0.229, 0.224, 0.225])\n",
    "    ]),\n",
    "}\n",
    "\n",
    "batch_size = 4\n",
    "data_dir = './datasets/hymenoptera_data'\n",
    "image_datasets = {x: ImageFolder(os.path.join(data_dir, x), data_transforms[x]) for x in ['train', 'val']}\n",
    "dataloaders = {x: DataLoader(image_datasets[x], batch_size=batch_size,\n",
    "                             shuffle=True, num_workers=4) for x in ['train', 'val']}\n",
    "dataset_sizes = {x: len(image_datasets[x]) for x in ['train', 'val']}\n",
    "class_names = image_datasets['train'].classes\n",
    "\n",
    "device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")\n",
    "print(device)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "outputs": [],
   "source": [
    "def imshow(inp, title=None):\n",
    "    \"\"\"Imshow for Tensor.\"\"\"\n",
    "    inp = inp.numpy().transpose((1, 2, 0))\n",
    "    mean = np.array([0.485, 0.456, 0.406])\n",
    "    std = np.array([0.229, 0.224, 0.225])\n",
    "    inp = std * inp + mean\n",
    "    inp = np.clip(inp, 0, 1)\n",
    "    fig = plt.figure(figsize=(5, 8))\n",
    "    ax = fig.add_subplot(111)\n",
    "    ax.imshow(inp)\n",
    "    if title is not None:\n",
    "        ax.set_title(title)\n",
    "    ax.set_axis_off()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZcAAACJCAYAAADgzXVSAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/NK7nSAAAACXBIWXMAAA9hAAAPYQGoP6dpAAEAAElEQVR4nOz9eaxl2VnfjX/WsKcz37nmqq6u6uq5PQ8YjInzAjbmJSQOipxAmyEoxhZCECGSSDFJjEgUQLackAhFwUrCkKBE/gVeYnhtMAHsDm7Hs3vumoc7n/nsYQ2/P9Y+51bbBtqkjXnNfaRbde8+++y91h6e4ft8n2cJ773nUA7lUA7lUA7lBRT51R7AoRzKoRzKoXztyaFxOZRDOZRDOZQXXA6Ny6EcyqEcyqG84HJoXA7lUA7lUA7lBZdD43Ioh3Ioh3IoL7gcGpdDOZRDOZRDecHl0LgcyqEcyqEcygsuh8blUA7lUA7lUF5wOTQuh3Ioh3Ioh/KCy6Fx+QrLW9/6VoQQCCG4//77n/PZmTNn+Mmf/Mmv6Pl/8zd/8898jkuXLiGE4MMf/vCf6ft/2eb3vve9DyEEjz766J/peF9pmY/vzyJf6lq99a1vpdVqvUCje+HlrW99K6973eues63X6y3ex3e84x1fnYH9JZFD4/LnIEeOHOGjH/0ov/zLv/znfu7f/M3f5J/8k3/y537ePy/5Wp/fobyw8sEPfpCPfvSjX+1h/KUQ/dUewF8GSZKEV73qVV/tYRzKofyll5e97GVf7SH8pZHDyOUvkGxvb/NDP/RD3HvvvbRaLdbX1/krf+Wv8Pu///vP2W8OUfzMz/wMP/dzP8cdd9xBq9Xi1a9+NY888shiv7e+9a3863/9rwEWUIAQgkuXLgHwa7/2a7zyla+k2+3SaDQ4e/Ys3/d933c4v/9D2d/f53u/93tZXl6m2Wzy7d/+7Tz77LNftN8HP/hBXv/619PpdGg0GrzmNa/hQx/60Bft99RTT/GWt7yF9fV1kiThnnvuWcx7Ls453vWud3HhwgWyLKPX6/Hggw/ynve85ys2z7l87nOf4/Wvfz3NZpO1tTXe8Y53MJ1On7OP956f//mf50UvehFZlrG0tMSb3/zmP/N12d7e5gd/8Ac5efIkSZKwtrbGa17zGj74wQ9+Red6KF+G+EP5isrDDz/sT58+/bz2ffzxx/3b3vY2/6u/+qv+wx/+sP+N3/gN//3f//1eSul/93d/d7HfxYsXPeDPnDnjv/Vbv9W///3v9+9///v9Aw884JeWlny/3/fee//000/7N7/5zR7wH/3oRxc/eZ77j3zkI14I4f/W3/pb/jd/8zf97/zO7/hf/MVf9N/93d/9FbgKfznm94u/+Ise8CdPnvTf933f5//H//gf/hd+4Rf8+vq6P3nypN/f31/s+x//43/0Qgj/1/7aX/P/7b/9N//rv/7r/k1vepNXSvkPfvCDi/0+97nP+W636x944AH/H/7Df/C//du/7X/sx37MSyn9T/7kTy72++mf/mmvlPLvfOc7/Yc+9CH/gQ98wL/73e9+zj4vtDz88MM+jmN/6tQp/1M/9VP+t3/7t/1P/uRPeq21f9Ob3vScff/u3/27Pooi/2M/9mP+Ax/4gP/lX/5lf/fdd/uNjQ1/69atL/u6fMu3fItfW1vzv/ALv+A//OEP+/e///3+H//jf+x/9Vd/9XmNHfBvf/vbX5gLcShfUg6Ny1dYvhzj8oVijPFVVfnXv/71/ju/8zsX2+fK94EHHvDGmMX2P/qjP/KA/5Vf+ZXFtre//e3+S/kQP/MzP+OBhaL+asjX2vzmxuX2uXjv/R/+4R96wL/rXe/y3ns/mUz88vKy//Zv//bn7Get9Q899JB/xStesdj2Ld/yLf7EiRN+MBg8Z993vOMdPk1Tv7e35733/k1vepN/0Yte9JWY1h8rDz/8sAf8e97znuds/6mf+ikP+D/4gz/w3nv/0Y9+1AP+Z3/2Z5+z39WrV32WZf7Hf/zHvfdf3nVptVr+R37kR/7MYz80Ll95OYTF/oLJv/23/5aXvOQlpGmK1pooivjQhz7EY4899kX7ftu3fRtKqcXfDz74IACXL1/+U8/z8pe/HIDv+q7v4r/8l//C9evXX6AZ/MnytT4/gL/9t//2c/7+uq/7Ok6fPs3v/u7vAvCRj3yEvb09Hn74YYwxix/nHN/6rd/Kxz72MSaTCXme86EPfYjv/M7vpNFoPGffN77xjeR5voAJX/GKV/CpT32KH/qhH+K3fuu3GA6HX7X5vuUtbwFYzPc3fuM3EELwd/7O33nOHI4cOcJDDz20YKA93+syn+/73vc+3vWud/HII49QVdWf23wP5fnJoXH5CyQ/93M/x9ve9jZe+cpX8l//63/lkUce4WMf+xjf+q3fymw2+6L9V1ZWnvN3kiQAX3LfL5TXvva1vP/978cYw/d8z/dw4sQJ7r//fn7lV37lhZnMl5Cv9fnN5ciRI19y2+7uLgCbm5sAvPnNbyaKouf8/It/8S/w3rO3t8fu7i7GGN773vd+0X5vfOMbAdjZ2QHgH/yDf8DP/MzP8Mgjj/CGN7yBlZUVXv/613/FadFa6y+6T/P53z5f7z0bGxtfNI9HHnlkMYfne10A/vN//s88/PDD/Lt/9+949atfzfLyMt/zPd/DrVu3vqLzPZTnL4dssb9A8p/+03/ida97Hf/m3/yb52wfjUZfkfN9x3d8B9/xHd9BURQ88sgj/PRP/zRvectbOHPmDK9+9atf8PN9rc9vLl9Kwd26dYtz584BsLq6CsB73/veP5ZFuLGxgTEGpRTf/d3fzdvf/vYvud8dd9wBBCX/oz/6o/zoj/4o/X6fD37wg/zDf/gP+ZZv+RauXr1Ko9F4Iab2RWKMYXd39zkGZj7/+bbV1VWEEPz+7//+wkG4Xebbnu91me/77ne/m3e/+91cuXKF//7f/zs/8RM/wdbWFh/4wAdeuAkeyp9ZDo3LXyARQnzRy/fpT3+aj370o5w8efLPdMzbvf0sy/7Yfb7xG7+RXq/Hb/3Wb/GJT3ziK6J8v9bnN5df+qVf4m/8jb+x+PsjH/kIly9f5gd+4AcAeM1rXkOv1+Pzn//8n1jIF8cx3/RN38QnPvEJHnzwQeI4fl7n7/V6vPnNb+b69ev8yI/8CJcuXeLee+/9P5vUnyC/9Eu/xA//8A8v/p7Xc80LGN/0pjfxz//5P+f69et813d91x97nOd7Xb5QTp06xTve8Q4+9KEP8Yd/+Id/tkkcygsuh8blL5C86U1v4p/9s3/GO9/5Tr7xG7+RJ554gn/6T/8pd9xxB8aYP9MxH3jgAQD+xb/4F7zhDW9AKcWDDz7Iu971Lq5du8brX/96Tpw4Qb/f5z3veQ9RFPGN3/iNf+zxLl26xB133MHDDz/M+973vsP5fQl59NFH+YEf+AH+5t/8m1y9epV/9I/+EcePH+eHfuiHAGi1Wrz3ve/l4YcfZm9vjze/+c2sr6+zvb3Npz71Kba3txfR3Xve8x6+/uu/nm/4hm/gbW97G2fOnGE0GvH000/z67/+6/zO7/wOAN/+7d/O/fffz8te9jLW1ta4fPky7373uzl9+jTnz5//Y8f6vve9j+/93u/lF3/xF3nrW9/6Zc81jmN+9md/lvF4zMtf/nI+8pGP8K53vYs3vOENfP3Xfz0QjMYP/uAP8r3f+708+uijvPa1r6XZbHLz5k3+4A/+gAceeIC3ve1tz/u6DAYDvumbvom3vOUt3H333bTbbT72sY/xgQ98gL/+1//6lz2HQ/kKyVeZUPA1L18OW6woCv/3//7f98ePH/dpmvqXvOQl/v3vf/8XHWPOpvqX//JfftExAP/Od77zOcf8gR/4Ab+2tuaFEB7wFy9e9L/xG7/h3/CGN/jjx4/7OI79+vq6f+Mb3+h///d//08c42c+8xkP+J/4iZ94XnP6yzS/OVvst3/7t/13f/d3+16v57Ms82984xv9U0899UX7/97v/Z7/tm/7Nr+8vOyjKPLHjx/33/Zt3+Z/7dd+7Tn7Xbx40X/f932fP378uI+iyK+trfmv+7qvW7DPvPf+Z3/2Z/3Xfd3X+dXV1QU9+Pu///v9pUuX/sQxv/e97/WA/8AHPvBlz/fhhx/2zWbTf/rTn/ave93rfJZlfnl52b/tbW/z4/H4i/b/9//+3/tXvvKVvtls+izL/J133um/53u+xz/66KNf1nXJ89z/vb/39/yDDz7oO52Oz7LMX7hwwb/zne/0k8nkeY2dQ7bYV1yE995/lezaXwp561vfyoc//GGefvpphBDPYT/9f1F+/ud/nh//8R/nmWeeWeDfX0vytT6/L5Tv+q7v4uLFi3zsYx/7ag/lz0WstXjviaKIt7/97fyrf/WvvtpD+pqVQ7bYn4NcvnyZKIp46KGHvtpD+T+W3/3d3+WHf/iHv2YV79f6/G4X7z0f/vCH+amf+qmv9lD+3GRlZYUoir7aw/hLIYeRy1dYLl26tKBaZlnGfffd91Ue0aEcyl9e+eQnP7nI762vr3Pq1Kmv8oi+duXQuBzKoRzKoRzKCy6HsNihHMqhHMqhvOByaFwO5VAO5VAO5QWXQ+NyKIdyKIdyKC+4PO8iyn/8428jTWKKWU6z1ajXzpA4b3HGYUNvWrx3ICXOepzzNBspeT5hNJnhvUAKhfMeIUBKWTens6RJwvJSByEs1lmsFcRRRFiUVQCesshRShOnCcZUeG/RQjAYjjBOkqQJSRTjnEcqgfeeOI5RSuKtYzydoKRE4BmPZ+GwCJI0ptVqkucFOooQQjCdTkAIpBBEOgI8pjJ4L1BKI5XEWROO7T1CSDweqSQegffgnMdZR1mVxEoiBJRliZASPAgpiOMYay1laSiqCqUUUoSxC6kwxlLkM1rtFtYYcGaxj/NhHQ/rAr2SOnsmhEBIiRTyYFnb+n8hBFIKPOCsw1iLcw6g3h/whPsI4XrVGz0gEHgRmEZxmqJjhTUVvjLhvBDm7zwudN2+bWldh3MewqHDNcIjmH+v3ohHyrDVe4/DM88Mzo/vvEOEAeG9x1hLaR3eQxwpkjhCax2uk7NYazHW3XaNQEqBlIEa7rzHuzBWKQQegXUO7xxKCZSUeKGw3uOtDYeZX9r5Lxzcg/qvg3l9odz2gUAsjnX75+KLNi5u48ExWDzGIATi9gl+ye/c9oc4+D+MwT9nM4txeb54gHzBvrcPZP6k/HFfE1/it+f+EWZycN75UU3pMaUF7xbPBoTnbP681VsW19ff9u9zb5G/bR2gcBIpBLHWNLKMJKo7InxhWroeVniuw/vuXBjPfKxCSqTSB8dn/p3wS3hfff2d+QjDdu9ceCddvf3244pwkPlxnzuo2373HoSvzysW79hiooLFuwZ+fusPnikh8EKGsdb/z7VAfzjmn/zzn+dPk+dtXBppxte95H5uXnmG+x+4m9nM4o1HZ4rxXsVgOiNLNcLliHab2dggnWBlKeUjj/wOj115hqpIaGRLlHjiSJGlEbt7Qwb7E17zshfzzd/wMrIs58mnnwLWuP/COdx0hkw67O/fYufaVVaOrLF++gSbN65w5dKTaFfxh597ijJa59TdZ7nz5HHSLCFKNVEcsbG+wdJyj/7lq3z6qSdoZjFUI/7gf21iEOR5zrmXP8BrXvpSnr54idWjR8kaER/7+CNMJiWR1pw8dpzldo+rV26SzywCOHPmGLPJLqPJHknWYWX1BEU1RccKEXcYjAuqwjEaznjq6cd58flTxJFmc2uT3HhSFbO03CJppexs7fKpzzzFze19HnzwIe44foLt7V1u3Nxh++Y2Tz/7ON/x176ZzRs30JMtlpZ7NLMmxjsGozGD4ZDxeMpsWiHQNJtt2p0OjSxBKI2QEh0nxElMmsYkcUplKvYHe9zc2WE4mmKNI000sVJIB1WeAxXtRkaaxOFBFQpkhPWWojIce+lDrJ3osXX1Wcqb+8RRFJQ9kllhyfMKU3mEtPiqpCgK8mmJ8AJNjAOshyiKibTCVBVFkaNwRJHEeciriqI26nGkiLUmr2AwnVIUU7w3TPKCW4MJu+OKKIm46/Qy95w9RW91Ge8Nw8GY4e6Q8WiGt8ERiCTEWUSaNXBSMisdprJksSCLI4yTjCYFVTFiqRnTabXxrSX6M8uk3w8voQxvo/QSATjswevtbzMurlaTQuBxC8XgawslxYEClVIGp0bKsL027EIKhAQlqX/3CAlogRIglUIojxQglUQoATIcW0of/gaEkAgpQIJX82MJpBJIFZw+IeXif2TQOlLKoIjmygmP9M8dO1LihQcJSkgWKk2Ea7HYIoKDIhDI2niJ2jjK2pn5YoMbrtetZyybT+1TFVNKV1K4ksqWVMaQm5JpVVIai/U2GJva8HjvcD44F857cB4vgvMURRqVCOJEsN7t8or7H+Cld59mud3BVw5nqzDuhcIO45ZaIeIGlY8pi4qiqHDGIvEkjSat9jJZo42OFEp7okigIg1eY4zBW4Mp8uCsCIFDUlaGfDZjNhpTjqfY0ganB4dSMrwrSUoUpwglaxvhcDjwHuEt3jnwFUKUKOWIpCcSEiUFkZAIrevnwYG3eG+QOCThOZIqBpXg4iZGZXjdwMsYIQTOWH7p137jedmM521cpBK02zHjXpvPfe4xTh27gyTRaJfT7S1z+tzdKDNgNtllr5zRWpU88/lLdNt3IaMMLTNKoYhbXWItaCUxiRasr2zQajR52cvOEcuK5WyFRI85euwUWxefIJ/0OfXgK8gaGotnVo7Z3L/M7t4mn/jMs6TZKjRPQ1mR6YxbN7ZJmoITJ9YQKqXbSRnt3mR/tMv+/oA4XkMaOHZihd1RzvZgl06nxXg2ZW1jjdFkzPUbO5jSkU+mDIqK5d4GrU7G0tFT7O5sYadTympIVRUUM8Hy2iqb27cAS6ORMS6mONUgnxUMxlOazYjl5RamMFy4+y6G0ymNLGW/P+HS9W3GwzFPPHWdY6eO8PWvfhlLvSYmP8H2tS0++IcF09kxVnvLbF6/jikLqmJCpT3Wg6IkFp5YamTWoNno0Fvu0G01kBKsKRFA0oxpNps0GglRnGCcIEkk1jiscQwGQ6bjGVY5IhmjgCLPGdkSZ1NiHSGVRCqDMRVSpXSWm1hbMNrew0xLCgPWCbyUVMZSlRaMxRU5xXSCKQ0KQSPNSJNooUh1FGGdY1gZrLVYoKqjHCEEjSQjkgrwVKZkOhlzfXuPzcGY0oVIWemIpaUW3VbC2lKHRiPCe0thFIVNqdwE7y1KGNIoQUcpSRIhtcbJhBTwsUUB1hnyfEKZz7BVwUSUJJEnabaDkluEXkHpOYLTv1Ckc+9vbl3k7UFCjUT7xW+LYGKueOfqO+znUbWHHOIpjxD1Tx2dKhwKj0cga+ODCMZoHqEhDzxdIcFJgvGRtRIX7rboB6QAIT2LwdSGtB7oIqrywjP3a13t/srbDIqfRwf1RXEi3C/5nChrfh1qI1NH1guPv77UskY86rg7KNHaww+RQx09OI/0Eo8NRn4R1YDzNow2WEOcg6qyODStLOWldz/Iy+57Eb1mE1uVOGPwztXXPRgjWf9orRBKIaVCiuBQWESIPGxBVU2ITYTWKQKJFB4l59FxhJMSJQS2LPHW4IVAJlGI2ZzFWYuzsxCEKI3SMTrN0EkDEcUgJNZZnPPhmfQWfIX3JcIrhIuDY6MsDo+SFU5YtHUoZRHKoThwOKQCoWJE3IKkC3ELKZs4mSCkDjdBW5DPL5vyvI2LEILh/oTNGwOSRFG5Abs3xywvt/Fxh3NnTtO/+knktKAY7bI/LEiyNirNWF5ao9PeotntgI6xruTsHce589RxiskeyytNlpdbrPQSkjji6JF1ZsM+W7c2icgp9m6QRymjUrJ9cZvJExPy4Yy9Xct9Dyxz7mwHW1U0dYOd7T794Zh2K2Pt/ApbN65w7eo1Pvf4s8yspFKCkxs9+rMp1zavkURdlptL9PvbtHurDPpjrl7b4ZOf/DRSamZTz9LynWyciBiPxwynlrMnj5NEnlnhsMKTtnpUpkRYx3hvxNgKcjNhMhqztb3NA/edw1qD1OBsxWBnkzxrcPXWlIs3J2StHmdO3MfXv+Y+1ja6VEUfqQ2TYpNOu+TFLz7FkaNtPvMZKGaem2YIu2MQYI0nUhlpc5lWI6PbatPtZGglcKag8BZjCtx0ROEKfKHIGi2SZoeVdodqpWLQ32evnDEaT5He04wztFYBBpwaTOVoZilSAz4o09XTp4iziMl4h9HOkEgopHYINMIKfGGo8imitIjKkFpLI4nptpq0Wg10pDHGMitKcltQVZaqrCiNwSEQTqKlJJUS4S2jfMLecMRWf8j2KGdcOZI04djKCqsrPZY6bbJGTK/dZG15mXYzRSlBaQSt4QQlPDMNOEcSx8RZSpqlxDrBoiiNx7gSvMdVFd5ZqsIwLUumM0eiZkTdCoihhvLwBx72ASRY/3MbxCRrb/8AIpx78PXOC9gkRAKL6GGhyR1IFxSAFMw/Cd6mRygFqlYO8uCcYg6LCrd4j6WoFT4cGJs5LESNlCyMSogwuE3Jz6MM4eeGVOC8QIjaq59HZB4QDlEPyM+hHxG88BC6zE9a/4k8gEblwTnFl7iewoPCoxBYKYlUrcocGBTGh2p8U0cwzjucB4NdwFwewNVQlPGcWjnCA3eep9toYYsSUxY4ZxC+vicqwNmS4EAJnaC1wgtXQ/CC0oM3HuE8VT6h0AIhHEKkOB2eGVlfJ0SInqQMBsl7gxbgtcemCmtjnK0oJxUhHIzwOsHrCGSMx+OkwItwPb2lhgltuIfOIz1YAxaLVhqrBJEPBibCIqULxlJJhFSgYnzUxOsWTjZwRHjj8DYP8LIPkfjzkedtXExVgbc8+MC99EfXefzzT9NtnWB1pQO6R9zo0MoabD09Y3+z5MnLt7jnRS/m6auf4fL1G1TOo2PP6nKHqhgRiwn58AathiKLHUudFkpbokzRa2csN1doaEEkpoxvXWUQtbjrvgtcv3KZY/Ex1leWWfrsUxw/sc7OzWtkOmM8GJDECYUpsRby2ZTZKKfftyxvnEPGMUW1z9LRY8TPbrLaTbn3wj0oJUlaHZzz2MpwZP0U992tKEpDkRsajQZPP/0U3ZU2s2LKYHwTk2Q4DEdPLiH1jJXlDoP9AbMyJ8oiOksNorV17r/7NN5UVNawurrGcDRCxTFStUE6mq0KG1Xc/eAJzp5rgJngCsXm5jbb+zP6g4KymvKb1/8ncRwhJJSlwziPFJJm2mBldYPlpVU6nYQ0joiVDCG31cRZA1MVlLMcbz351FKVQ9pCkGRduq0GS60Ww6yBzS1VZahMhfMCrVM8Dus9eVkRORkUv4pZO30U70r6tzYZ7Q9oNTO0lChpKPKCYjzFFyVKKCIv0VLQzhSdZkyUhpc/wGKOojKUpcFUFXkxY1ZWVMZiTEnlDFVlGBeGwkqiOGJpeYVjWczq6hLHjx1jY32ZdjMjTRWddpss7aKVoyimVKUljSWpWqdaCffYIRBahnycAFtVGOuwVlOWhgoHSYyvKsp8RllZQCJEVDvxkjkXZm4f5p76HLueK+95Pg4ODFCAxw6MzgFMdpADkHM4TABCgXALw4MISIKUKkR/UuClR0qQah51zCOPGj6bJ9MWx5gbNM/tgLu4TecHl/nArMyNUA023QZ7HWRHvD+AuA5yAsH4zA0ozHNtB9HePHx5jkHBH3w2N+TeI6VCSYXzCo3HO1BSodFoFFaFMVbOYqylqvOS1nusCIYmwGQeJ8JcpBCkcYM4TvHeYaoCYyoQoKRG6hilU1SkUTJEiEgFaKRwaC3xPuT3jA8GzPmKYjaeX0ykiJDSIjQIEdXYaW1sBeAc0hk0jlR7fCrxVuOx2DIYEifALYy1qJ+NkOP0wuOcJcBdjoM8U23trZi7Izgp8U6AckjhkNKDjiDKsDrDyQRvJc5WtaHNsaYI7XPs81uY7fnDYhKaDUlVjVjqrdJszFhdXWZrd4vVY0fwdoCd7TPY32OaW5yK+PgnnsCaMVHSIWmU4C0vvv88xXgX5YYstTylNdy4scepUxcQwlLZiqqaEEWOY2fXySfbDG+UDG9eo7u8TLsb00hTdJyj4illOSCWCa6QeAsqFqRphPMhJM6abY7f2SH3gsH+LuNbFZ/61OdZai1x//l1spYiTlsI6TDGIjA0U8VSV5AmPVZW14iSmKeefZbPfvIKt65P8XmDY2sn2d/do9XVNFsa4yQzm+Nihc0ta6sNhjt9pDMM9/sQabJGm/GswEcJR8+scHN3G+VKvGlwY3efS5cMZ0/G7A1mjEZTpqVjOHEMRlPSRoMoshRlgbGKNNK0soyVbpf11Q7LKy1a7TZKQJn3GU1HoDRJnJLEKZEOr7K14L1FOIHGkSlHN4XlVkwq2lTGMstLKmuIU0Wz2SFLYiIpsKZkMs1RjQ6d1Q5lOWHvxi6usqQipoFiNu6ztbnLaDjFWkukIyKtyLTE2DQk3QeSqh6LMZa8KBjlM7YHE271x/SnOXlpMdajtaLbTtlY6nBkdYXeUg8dJwgF3aUuGxsbLHdbJDFI6UAInJ+SVwYlNY1mQhxFNBLNbDoLOVIlUToJ+LstwJYIFKbyjIcDRi7HKKipBCglSLMUpTTitihgDg/dngudkyWC811reQ7yL/OE7u2G5rkRhFjkN27fvoh2oPaEPVLqoOTqPEeIaggeqJwHB3WUNVfUIuQL/DwqkWGIcgFNHUxvkZZfjGOegOY2w3H7324x1i9MMHt/m6GYb/V1JDOPguYEkPm5ZRiBr6+xJBgRKSRKRXgvkaJCY7HCYYUjEjoQXbzHeIfRBms9xhmst1gB1jkMLnxuDdY6BLC7s8fe1k1aiQrwmZRIHaOjBjrO0CpBClnnsQRKiYVx1hIEtobnDM5YwGOtpypG9QxShNCIzCBlDF4jcHhha/gwGBuJI1EOoR0qC7BjrhzGzDCGelwi5NkQCzgwoLT1DfWyvl8WIYKDGNBEj8GF++9iXDWPdCwqSkCleBHjLDhbUhYlJp9gizHO5PjwwZ9kKhby/I2LkEgr2NvbZKl7jIdedj/7e9eZ9ROWVo9g7ZTcVFzfG5CT4oXilS95kG435frOTZKnHYPxjBubt7jjaIvrT08Y9PusrB3h/F3n8H6CVss4V+BKGI4H3LhxmdMnj5L1jrIuCvJJwcrKCpPxhOHQ0G2uMd7rk6kGk9KCirFYlPTs9wecvuM4jTRGegdVQbzUQvuEa89e4UUP3cfm1jNc7+9z7sL9bG4PaWYd1td7DEd9nC+Yzix3rd5BVZRkkeT4sbOsrsSkcoIArEu4fGmXNGuzsbaOrxTddI1x1ccVBVI6vLXc3Nri3vseII0jKt0kHxZUkwlLnRRnFJdu7DAaTXhMjFheP8LEOAoLw3HBpCyIG0021rtMhnsYo+j1llhbXmVtqUe7kZCmkkgaMBNKZxgN9hgNA9tNRwpZP+RCxXjCy+FKQUHFeDqlmuakShA1A9Nuqj2TqSeKJUtLHVZXN2ikMdPplJ2dHbK1o+hIMNgdMNkZ0kxT1rKMbizZnxRcmQzZHORMTfASBZ5YQDuNiLSkclDWbDLnPdZ5KucpKs8oN+SVQKuYlU7K+nKLoyttjq506XU6qCTGColOErq9VZa6LeLIozAYKzDe4bGkcUKaxCSxwmUJaZaQ5xVCEFiIQlFURYDunAEnsMYQ+QJhC1xlmMSKKIpCQtVbvCvwXs9V7iJpP3e+hRDzbAyyblAaMiX+OfmDLzQo3Lb94Hd5kIsRByw2KQ9yKeFHIqTDz3MnUuDlPG8T8jIH4ZQIVmS+r/J46WtDUxs8eZAngTnrytcMxtvH+FwDMo/QDogK/rbjuAOld9u3g4cNXs6jwKAoESooY+9ruMyDkGjZQiuHFKBrBWu9RKvA9DPOBfaq83WE6jDeYJTDEJL8pmZiVd5gnKGSEqPmhqBkuLtDubZBnKQIYREqDrkOFaJEpSKUUmFe3td5nRAlCBkia4+l8hZnHQKDNTZEw75EEiNkRpx4BGaBY0rhA3ypAkHEO08sBTIO101Jy2xWUBpLVUKEB59AzVadR0HB+AeDLGp2LvPosoYGhQ9MVis9wkFZKTySKIqQPhhnbytsZXGzGa4Y4aspwpcL+PP5yPM2LkppkAotG4zGJa0lw2Q6YTAUzKbblAPYHAyYeM+trW3O33WeE8clOzuXUJXjobvP89FH/jeqGLPcPEp694NEqeD0sXW0qOhvbiKONEhiCSK8wFUJuRUcufdeViclV29eZefmdVqtDt1OjyRqshNBMTFYWyC1Dt6Y11gjaLV6RFRQeSKtmOZT1noRnfPHaUfw7GDGRz71OUS2zoW770XLkitPXWN/d4v+/j4qySgrg0TT7a1w4uw59vbHXL9ykazV4FjcpdFocvrUeaaTfVZWVtEqI++m5PmIIrf08xJHjIwDAyNpxqylTfb29knjBkc34OLly+xs7tDvV6hWhDWKwf6UvZ0Rx1aPsb27x+7OFp1WixMnznDm9GmOrK7QamiELyjLGWVpGI/HGFtRGZBKU05nVLkMbCIdh5fXBdbMxFjcuGQ6mVJVJuRMhMX4Ai2ht9Qj7fToLPXo9Bo0spQo0eRVyfLxVawt6W/vUE1zlPaU0y0iJ1jVBfesN2hlCSOvcFJjrWWaF0zziqkDH8UILWkkMc1GSjtLaaQaX3lG4wmDaYlUmkYSs9xO6LVbJFGE9SVVYWh2VlheWWep10MpS1GMyG2JipukWRZyKpFES9DKorQmy5q0Wg68QUqFMR6VF1Q+5A+EE3gbkcgukbRoafC+jfeCvCgw1pOXlvq1vk1BHkBJB5HFgfGQUn4Jb38e8XyJ7YvIhfr79fHqJH1w3+voZJ6ymJ+vjkjmifhFTqc+7jxacTVUJucfyzqCqA0N+EA/lXPIbpFEuQ2++1IKZk4ZFwfJeebKjgDViLlprs3fbVHZ4ig1Df7ARAbFqLRGKlezpxRaBQXmCcYlXuRWPNYHiMg4hfMO4yzGWawMil57iXUqQGY44jjmzjOnWDt1gbR3BOk9tpohla4ZVBFSRQghaz1ug8J2vg4KHWCRAmIt8ImmLDze2GA4bIktLLnNkb5C+gYyShCoOscWxo2fkwckQukAowmLFBYhDH5WUpoca3V4Bpyur/EcBvsCuKy+T95b8AbhwDqoqQfYOevECYSXKAfeWlzlMEWFK6d4myOo6oj54Jn60+R5G5e8yLly9SpZlNHUsHPxKleuXCddOsZg7xpPjq/wxJMXmUw9J9aPcMfxHrvbN/nfjz7NA/c/wInjHSYX7uDMuiZ1z3Lk+HlykdLuZGivESyTZRG7e32U1vTSJiLyPHv5Oq1uB1FUqFgQKU+axuzu79Frd2i2M6piQpJElMZgHRS5IW0lIBR5ldPotrn89NM8+eRjeOc5e/ws9FbZ3hmx37d0VjZwquCxJz6NnUxpZMtIb5lNHUIkrK72KKyjLMdMpjvoNGJSzGilmrvuOUZeDNjb22N1ZZ3JeIs0jekttcmyjLz0tHtNlBIYGzyn0kK3u0GkHZcvP85LHzyPN57CGlZbGxTjHKEV5x64i3vuuoM//Oj/4pOPP8qJo0c5u7LM0aNrdNsNtAZbJcRFSpHP8CMTamFkhNAWmTRI45g4AqEjrBPYqqAsKowxFPmMyWRGURbYKigsEWUsLS3RWztO1szwbop0JbbwmCoHrYiyCFPN2L22TVmWmNywPe2TtTwrjYgTHUkvdQyNoRLgVExpEgqnEI0Gje4SjW6TLMvIogiFBWeY5SWj0YTJzGKspDIlSgmSJMVYzzi3JFlGu9uh284Qogi0zekQ7y1NqYmlpZFI6tKk2lG3SOmRWFwoyAr1DJEmlgnSu5Ak944q90gyBCVSCqJIsTOYsD8YURQVxMltL9c8l3CgHBd1DXVC/ovrEQ6MiRQB/vL+QIkfRCu1LZEHhkPo4OEKqVBzgzanJgu5UM/iOXkPFkZIUuc6RP2dRU7jQC0FozSnCrOAxOYkgINWhHVtV/DmbjuPQwgVzuaDgXB1lOPrzL2HhXEKwVb9uedgHt7h/Zzu5hHeY605KDnxB/mjMEQZxuwVynsi7zFeoZ3GekvkagPjDdpXVF7gvA4sO+E4eWSdBx54iCOnLxDHGeO9m4z290mzhEazjZQh33nASnNzs4cgPFvemzpZD7GOEV5Q+TIoaxeUv6ks07HBe0PWbCGiBIkM99HLQJH2LjxREpSXSBGDCnVtpTNYDFDhsYg5FdGzMOygakp3fd+Fx/sK6cALWweMATEIGz3ShbkJZ6AqcKWHqgQ3Q/gCIVxwdA4epz9Vvoyci6Q0Jd4Yrm0/y6c++ySrpy5wx3LG7i4kaxnr3XXuu3eV4dZFmpPrjG5u87J7T+GTMZ/4zGeR5QoSxf7eNT79zCbf8M3/N1KUzMoZUbPFNB9jrMcWM5ybUVYlzmuuXL7GPWfvoLkU8/lrV9BZk73xmGaakKSBsz3n69vSMs3HPPjSC8zyGVU+xRQFz3zqSfr9KRde+hJIU565cYOrW/usrh8njQWlqXj64hBTTnnda+4iNzFFUdDIEopyRpIIqqKkqhxeavb3Rggb0WzNwDu63Q5ZIwVhqMoC6yv29nYRXuKt56nHr/PiF70Spzxx2iAiY39/k6Vuh7Kc8vKHzuJJWD26TFGMyNImp07fyWhnwJkTR7i6tUIaa5bbmiTy6DgiazZw1lAVM4SO8ULi3D7WTqmQxFmEjgRCBO9+Nsspiwpc8I6r0uC8RMUt0m6HVrtHo9Gg2W6RZjGmypkMcsx0ihUKW1SkzSVk5BhsbzHYGoHUCO+pXMQ4d7S0pZF6lhJIvGGaTzHe4VQMWZt4eYn20WXaS13SNMZ7T1XMKKYC4Ry+GRMnUDkwRhHrBmnWZFYWOA2R1lhb0O9vYa0J+K+pQtQlJbaR4m2EimJUFJKm1pU4W9a4dFCSOopIkqymbBpwdWLfFAgNKkmJM1CzAuMs+5OcNLY0E+oExUHNxhdGKl+4bS6394g9gMVkbUC47TP/nGNRGxEp/CI3EhL1wagEBtYBGaDmRdfRgqjhMbGIGgTBw0ceIGQHUMrcIM61yAED7IDtNveG54OpDSSLg9Vqd25WFlokjEe4xXeFCEpNClnnqr7QKw50Yw84V+G9WsBuop5fIFvcbvTCHLSTOBGgSSt9YEU5gUKgMDjhiTwkacQdJ06wurJOHGU46+kPhjz2+NM47zh+9Bhnz56j101rJR6udzAYDkSoMwnX1SB9nYOJJBJNVTq8McGwS4k1kE8sYEmzFBGnOELBpVIqJOxDhTESiZdBUUcxREZRGodzwcAIoYIxl3VS3xOMce2YhLFaAmtPhWNKj3MSyZyeXWEcaDN/l0IRvLcl+AJBcM5UDck+X+vy/NlixjIYVSyvL/Gxxy9TqQ3USPKylWOcPLfKtWefppkK2o0Sr6dcefwWVno+8eQVStmhtAXrRzIu7sLlJzbpT5d5+WtLtm9exrqIlaVVZDNGUrE/HHL85BoqjphOZ6ycuYP140vM8ilPPn2ZYycvUJUjRJSRxQlFsQeRJh8PaXZanL1wH8vrbSyC/l7Jkko4v7TOsTilnSn61ZRrN26SLHd51UP3UPR3uHx1wiAvedHdF9jbuc6RExs0lMKYnMFwQpwleFMhhUIqyaSccefaOq0kppjNGOcDZmmEUDE3tq4yLSfs7A7Y3dwGK8kaknvunYFoUdoclcFw2me11+apJx/nyImT6ESRV7OQyHYl+Ipup8WxIxvccfwsKjakjYwoUigNOo5AZMg4RqcdGu02URwxG09wIiFqNsEbysk+1dYN8umEojQIH17myoTEXNZs0Vvu0el20Frh3ITpaJuyNNjSYywUxZjcaFonOkDBzo0tyqIiiQRRlCB1AyNhLGYIU4VixCxGigpTTfCuQrgZkR2QVG2UbSCI8DikVERJiDTRitTLwKLxnjjWSJUxnEqqKsdWhnw8YmoMHk8sFQqLVII4NVgT6mOcd+EFFzFKRUjvcM5SlTnWlEjliaIM4SXOgJeBYFBYx7QyVEhE3CBqSnQmaLQdUZY+F/q6LVq5Hcp6DkuKL34Zbzc+BzDDAU15DrXVh6gjoVqhc1CMOD+wvO0kQszV7oGyhaCspAzefa2qF1GCDxYFia+PHb43ZyQdGJHFDOpjzmGS241MOLZ3c2bZ7VInj7mtK4Ovq8AJhnaRN6gNJLUXDwG2Cc/F4oqF+dcjcPMosvbMEb6up6kjqNq4aiFRQmJlyOmcOL7BkSNHyJIMISVxpNk4cZq008U6SZY2yNotpNZgC4SdQ40BZsUbwOBFgMDCJfFoqYK+EAkVAmcLnK1ACkxlmU0M3lc08OgorSNBdUBDF+HaC1ehEERaoXWAJK21CPK65qb+rlbhui+emzkcdhAVBvpzgOLwhES/U0gs1jmEdSBMKOy0BXiDEqHjiVbqNljzT5fnbVyctaSxZq+/TykanDx9F699zatZb0suP/lpXCm42h8xLQfc+uxltGixMx4yKjqcfeACRTmg21qhmQkaq5LVzglcZfnMZ56k3T6DcBltAiNiPJvgxAm8z+k1m7Sbkp0blyDWTApJYRT7gz5b+x0Els7SCsfWj2KLCTrRTKsJ4+EeqytLHG1nXHrqKiJTSN3hs48/xR0P3k/hHQ++6DwrqymT0ZhulHB+4ygbzS5VYek2uhTTitIo4kYbBFTOBpy2sszGOSutLsuthKvDMZ///LPc82CbJIYsa3P15hbDsaG0DXCSho4wtiTShlRH3LxyDe8dFknSWaPVOcbmrWt0u44kSZkMJ1jrkTqm3cl4+Ysf4LHLT3P0zJ1oLYm0QkVReFWFQslQfNhqJiSJJmktk7Z7OFcwuAVmvE+V5oydY5rnVLZiVuTMTElUTimrEf3dCHCUeUXhDFInpHGDosjJx1OaK8dY7qSU+S753oxIC5SCTq9Lr9cj8pKyGrE33iErcpqJptlO0F7hTR6USD7E71/D+JIyX0ImDaRKkEqhE0lzHm1UJc5UOGvI8z7D/ojpcESSNEh1ExE3QxshGeCGKFKkzTZJmqA0OO+oqhytHZGKUFHoPGAqhbN5SFbKQNl01uC9xBkDWNJIkUQxXqZEaUmFpAKcTijLuu4BsaiA/1KMr5DwDS9zYO2Ig7zMbZDVwfa6Bcdt+ZPnQGL1sZHBEHoxjzKCkpgblQXYNK89qZWUkEH5+jp/I0QgVM/bpMyVop+fuz7OIhqpD+WcQylVJ95dPQC5wNSeG80EJVsXy9THcvMylpCDAfAh+rLeIG9TXLKmR8zbRimVAuWiHZB0ISJbgJT+wJrNoxq3MJGgUEgtcEIGjx9L2ow4urFEt9mAyoZnQEU0Gz3arR5Sh5wlOGxVsqD4egsYfF2zE0xzYPF5RMCoXIWQGh1LIMKUFmurAD8Jgass5Sy0F0KC0nGIOAgRRrjpoe2Q8gE+jJRHKY8zhrLMw71PAwEmsARrA1yP2ePrilkBc8ozDulUfcss3kksoUWSthVOzaPFkGdRknAdtLzt4v7p8vzbvzQzWu0mj/7Rx9FJj5e+5EXc/8AFblz8NOO9PuNhwfZwRGftFD5tcvn6DbbHFQ++9Os4dcdxHv3YRe4/fx+2GrA7nnHvS46A8Zw/91LiVo9B/yZdsYFyFVmyxF5/SiOxxEnExaevsD+akLQykmSJorS0203SOKLV7HHz5iaOEu8r0iwlaiyzv7UHpSDvT/noo58gXl/hRXffz2OPfpIyibjj3J2s9LoIIXGkWDngngdOQFWxvLTE1atP0uxtsN4+yu7eFh7HcDBiNLTs9yesd5dpxgmmMgwnM3TcJklbYC2TccV4ZDFWEGUREtjt77C9t0uvDdZ6nnzyce5/8B6evXiRM+cvsNQ9xvWr12i3MooivGRFUUCsSRoRDRGj44jWUjdUCQuBsxWunCFsAbZmZWVtUi1Jmh2iNKXIQwsIpTVJppnMDJWdkhuLkRqdtUHFzErNJC+ZjIbk0xl55XAiJDOttUg0d9/RwwvDbH8fWVSkWhOnkqXlHqtrq0g846FibD3T8YCyytHK0Ew1UdrAmSoUto32scUMO+pTNJeRrSVU2kTqCKUkZZmTz2aYvMJ7TWlAkrLca9NqNWh22ug0wxOiEecrlDA00pg0UcRRyBlYLxHWI0WFMwZjHHleUpUVRWGpillItgaNHeoaXEmkA15U2gqYIbwhiSNKoQBTvxHzRLVfwGPAcw2NXGj9BVTD4lv1eZkbiDmcRu21ywNDwzwv4eoog5odFrxxVRurORR2e+wRlKzHPSfXEX5cnVgJQ/Q4IWpFuoh3wmjnMFStM6039TkCWUEtdPrtEVPN/JqTGTyh2K8+5oJV5+puA67uEzYPyHzoHwceJQLzLlINBCGh753DSRFyEv7AKAYLXcNut7HNyqqgKgtaWU6cdZlZjULQaUZ02jGSnNHgFo4KOl2SOIFIIayoOwooVCRAJwSGmMbaKOQvvMObCu9nWF8i68JNgUO6AikjtJbgwwqYvqoWLEpXGfLJFICk4dG6vgGSUNBCgLO8B+0lkRZECsCQ50Xof+cFpKB1oErPDUjdbAip5l0hZGDmOYEgJPf9nE1WR/bWlChU+L6wAY6t6d/UMOfztC3P37gUZcnF65vsDgyNhubIapdWQ5J0e6TtHtc2n6E/mODEGg+8eoXVK5f48B98HnyFGU+5//y9rK82+IM/+l/0Gj0SZ0DGrBxp8r8/+1lslXDsuEHaCcOdXdaOrHFz8zHiJOPKfp/tzRzPHi996UtYWeqiKei1m+AF95w7g7WWdHmFq9ev0Oy26S61QQluDUc0lo5z9PhJ9oY5d124l/vuO8toNEHKgrwwtDpddvoDRnsDmo0I4zU3r17idK9HplO62TJ5PqPXdESR5djaMsutFFPNmBpDZ2WVB1bWiZKE7RubHNk4SX9kuXLjCpGSRMpz/WqfT33uaR58MOXysxfprLZoNBTtLGZ1pcGNG9coTYGXFWma0Gq22bp1g7NnzzGyJf3J6ABPn7fuEIrIJ+goCQ0YhcAbEzwOGSNjhSwVUdqk0e5hvcBoT6E80jji5hKdzjpZs4mWmlk+ZXdrkyuXLjMpxzgrQvQkBM12j5Vjq2ALptsDlDOomuFiXYExOVoJ4jiit7pO3uxQFhP27AhsQUdV4aVwFu0NUTUlHlcU5ZSqyvHdI0TdFeI4JtEJncYSWiU0Om2iRkpzpU2UpvjCUQ5GDPf3mdoK6x3OKpy1dZPKQLtGBiqqF56qMBTFrG52KgKtXulaGbvQOLQqyWcFRVFirce6CuMUo5nHeR2cEHfAmJpDSs9pDsoc56aGyWrlWifuoX6PEQtqr5Q1zbdmYc17xtyGkAVdKwMTac5YRQWFKmqQS/hQLxJsRzAe3od6hgB1+brw/nbV4BemYB7BeO8CsaOGkzyuZo5F9NrH6LbXsK5gd/cyZTVlrsj97efEgQ8GYZ5CEbXSxwdjbDmAzaS3CF978w4WmGCd47HCIZVGiTaCHCF8yD/JAwDQh1seFO3iubQMxyOK2QQnPEnWpBJthK2dM0Ah0JFHipyiGDHY3UP4I8jldaK0i05SlArU9QBTBbqzt6EmTmBC5IsLNHXvauMLoLDOhOJIIYLh8BrjPd5VeO+DcShL8hqqlE0fWGkyXHPhFXMSBM4R6TBeFVWhMDQ39fOiwEu01rUjEwyzvC0XJ2p4zEtVfx5IE6GsKsJ7g3UG4WsCjBfhB5Deh5KOGtZ8PvK8jct0lrM7m4HuEicZVDn5qKDd7rA/HDNzMJtFrPVWabdGzHqrvPbr/yoby4ZBaTh+bJlZ0eeZi7c4e7yHdh6H5I8efYRnr0649+57mYx2uPz4JW5t7fHgq15CZ7zMxz/5OEsra9zzQJdOIyWLJLGo6DYbbN7c4ujxdfJqRqRirKvqUN2wvXeT1aUldKp4zavvpTAj4rSJEquUsylPP/0MO90G9959V6AHb2zw8Uc/ztFja/QabR6/NKF1dMq50yVpS7E32MNLR6/ToRE1Ea5EJxkVBZnwRHGL/mjIiWMbpK1lUDGDvX12R/tsDrfY3JmwcUJy6dImu9t9Hrj3PsajIUdW1rl5+SZ/9LHHiJVHnz8aWF04yqLi5s2b7Ax22dqbBgVRe4xCSFACJRVCapTW4AzldJ9yNkSqlIgGUglUkpAtdSFWpJ02q0fvwAlJFDfJWl2iOELiKWdTOs2YopxSmJAsTOMErWD52EmiDGbDfeysQscR2ltK79gfzUAPabUyWo0WzTQD7yiLgmI6ZTQbU1YD0mpCQ0CqS7S0COFJVATNJeIjx2gcPUK61EElDXSq8UKHimgtUbHGGRBI0o0uWb/D7Pomg0kfm6ZYpylKS+Usrgi0TYEnwPSh84CUnkamSZOkzi1J8lnFdDpmNMqZTAqqqsQ4Qp6pgtxYRtMZRZXjZbaIQELCVC7yLYu6ED+PUOqXumYYhXyCQPi66E0+R4fOcS8OIqJFZoM6FRGUdJ2B9y7Ue4SAJhxvHhGIRTYiGB0v6jxNnYwOXvNtp6shrQDfgXcWL6NgOLEolfHgPd/M+eMvpdFsooTi5s4lHvnEf2V/dCXg9lDbjjkMOI/wDoyZEAGmmeeWZA3RBeM4V4A1vCUOqNpSeFSUolUTwR7r7X0SP6OV5FSVo3SK0azDzDUwVpKVV7mxXzEsYhqtDr2VDWIVmqo64bGLnI9lNi0oZmNkU6ETSz6bMJuOaPWOIqMWUdRESl03vQwNVAPr1yF8aH3jhQIRASZUyLuqvhMO4Q0e6vc0GBghdCiGNFW4Pw6qsgr30TuSrEEUhSgeoUKuCY0QBqlDNwClFUJJqqJgNpsEFM4FSF1rFdCxOQFkgZ06JLXxFyEH452GOevMh3fFewnSIzUhvyRue/6ep2GBL8O4gKbdPsZoOuKuO89x6uQGUlhuXr/EdFIxHpScOHGEOBkxGPVJGg3Obazy5Gf/iJmVdJcyPvvZpxj2Fe17esyqGbubYy5eHbG/V3DsaJubN64idMbd9z3IbBbwxCNHNqicZX21zcmNYwz7+xTFGK1jVtaWEMqwubNFb+kIFx+/xN3nz7G9s8mszGk2WnRaHRKRorwn0Q0qA2UFvfUTjPZ2iHRGOZ0y2Nvn+pUtvNGs3r/MG7799QhXce3qJZrtJaIkpb+7hVhKcLoBs9BOQkhHFmlKU7G81GZ3+zoq02SdmOWNFa5u93n2mQFn7jjPXaeOs7Tc5PwdGyQ6Q6URWdbBTSacOn6kVogpN29sc/r0CTo9zfWbN8jzgpWlHtvbfVw5DS+kEIuCsVDcFWjG+XRCVXma7QYIidIpzc4qKspIW2NMlRNHLaROQToqY7EWymJCWY7wJke6HKUEzSylmaY0s4TO8TUQJeVgjHceHSdEePJyRuk8pQ2KRceQZgqJJI4FSRZTlS2qvMtkNqYq+hTVBBVpspPn2LjnLrKNNZyO0GmKUFGoNFeAkOgocHvwEqkEwju8FiRrSyTdNuKJZ9jZ3wTVRAiPseEFVXJeLS/RUUaaRkQ6NO8DG+qBbMSsUIxnglkpUDojjptUxtEfjdkZ7DOYzIJyEBokVIvI5cAQHCjPL2SIhRxEgMF9HYXc9h1fmwHhbz8K1IZgkb24ve5jTsH1NZQV9ghJbV/nN1wwqsJBnVgJRoP687knXFfDzztv+hqOCrkLixUWLzUnj72Iu068jEbSBCGRUnNs/U5e96q/w//8X/+Zrf6zeB+6hbs6QDmYy7zqPvz4OcuthrK8rWcmBAdZoHlOSNaRE0gRoUUMQnD1yg5Ff49XXii49/SM2bBk12S4xgafu7HE//rsNS7ccyfZytG6oaesza3FO7DWoJRAKMFsZtm+tRNQCR1jlKKYTZj0d0nSNlpG6Dhc47Asg8Pb4H14U4KtgrPgKrwrca7AuwqJQQgTKtqFw9maVqwUSqcIFJZQZC1qZpg1hmKah35nqSNOQGmPkKFGTfpAWVZK1D+SSoC1FWUxq+FyjxAJCH1AZhDzpqfh+sraIUGEfnXzUl/pAmMN5/HqIBe4WBbgj61v+uMsxvPdUWsaSYPTx9c4eWQVwYTKGHb3ttjc3CMWipe+9G6yjuL6lSpcbL1DKRW9bo9HP/kEl57Z5rWv/WbOnlplPLoemr/5BsfWOzz7+OeYTmacO383S0srpHHKvi1o9zI+9dnHOLaxzGwaqMq7+wNWlteYjAv2+1M++/mnOH9vQqPbpbIOrSKWug36gxF3njlLMcgxxrLRW8fh2Hxih7Pn7+bSE5+nnOVkWUSnEXFkrcOLH7qbo0eWmEy32by5Q6fT4eq1m9y8tUuvlfKi48ep7JROq4UQntFoRCNr0uq0GE4H9Hoddvc2mdqYrBNz3/2nuHDXCe65+yg7l6/QzWKG+YgyErRaG8h0g5aYcHSjpN1LaTRaHPeAyGm0YiqTs7q8SpRJdkSAN7ypcAgqE15UF/p54FxMkqU0OwlRloEKVedeBEKBTTO8rZBSIYCiKkLzyMmA7a2r7O/sMNkbMRmMCOk+g7FTrO6iWxEmH1DsT5FCopOYSkCEI440kaqxe++pbFljuA6lIc1iZDdjOsnYGzUZRDEn7zrJ0fPHiWOF1CFZO2+eJ4UPL6GUSBXhjA9N+PS8SlwjpUOkMe2jR9m+coU+M5KkidYB344jTRxFRFGCiuqKaluEdXPKmlXmBZXTSB3RyDJiHbrk7vT77Ax22dnbxTpoN5sI78mtWeRXgrN/UPvxxUn9+Zszh7kW6fNFnuYgJ1L/I+ZRaR39LNIXId/i5sdd5Mh9DYcFSMvNITPp6v/rRLsLRqiGzYPSrw3eoheYIhhwQhpYeo9Ds9I7zfmTL0ErjXEW6QW2ZhsudY7zonu+ld/+yC9A3eo9VIurhXf83JYudcQkZd1Ak9vyVTIo2/qaCRyuruEIfnuKIAYEv/fYLpP9XZ65FfHAuuGVd3uOL41Jswm6G3OrKWikmgJPZULLpFlRMstLjDUgodvO6HYzrPdsbU5YW3GsLy2jhaIqS4a7z6KEQdqTZK02UmmcdfVPhbMF3pZ17U1tVHyFsxUeA74isMjqmhgZclzKS7QqkbqBkGBLjbdhP+cBYyjyPFTJO4hih4oqlFRAzSRTIkTzWqPq9bG8NdgqxyqwWtbrWSkW9HIpDxpbhxMtHBQhDujqQqhF1CkECC2pF+vCi7k78wLDYgKH1oZWK+XkiWWqYsB0pLh86Soez0MPnqKyBa32eZwacXPzOraS3POyl/LpRz/Ndr/ir/5ff4UzR5cpJ32EcayurnJ6fciZMxtcfPIZ8C0+/djjvOZVy+zu3CDJYv6f3/ht7rzwItqtFt5WdHpNhtOE4aRPq92kPyi4774HSbOMbrdBWeasrq0y6A9xDi5fvUxUz2Gcb2PyHF8MmOzd5Ph6Fx15BuMhk+mEV77yJawf6VFMh+QFxHGDXneZT3/mGTrdddaOriELixvtEZ1qcuvmLpcuXufs2btwfsy0Cq1UnIrYunGLSCnMdMyZUyu0I8uVieXypS2SVsLGUovW8jrHT9/DrYuXSUa7mGLM7nRMq9ljNBqyt7uNtBWNpmBWeaSOiLM21hqcB5UGZRwql0ObCDHHOURYwMi5sCBYrWUo8yHTyRBnPUpFFEXF7vYNnnnqcXZ2BpR5aMnRbKREUYRWEY3VdZzLqfoDtBXIKEZISStKaXcSWq02jUaK84ayqjCmREmI44RGo0EriYibGftIXNbi5Il1zpxZCi1SRIUQEa4IFdHpcoZAIp0B45FZEhIN1iBziUsUUgmoHLZyiE6T1ZOn2X3qCYgTmmlKkqTEcYxOIpRK8N5jTFE3p/QYqykrFzoxGxMWWxOOWTFhf3fEk1eu8fizV5hOcrJUkXdbxHGKbq4CjeDdz1/IhSEJ133epHLujc/7Ps2ZymKeWJ+DXnPD9EWAQ2BiiXkZvZg3e6wbUc6jmLphoZ2na6iVh/R1q3gZ2F21Bypva40273gcDGXQ9sKHio0sXebUiRdz5/EH6bVWQm5ABK/du7CYm5Ka9ZXTRLLBtJwGo4ev6a9uYVTw/mDOguB41LMVdeGErA3vIvKrba0MiSmUjBFoBIJpbhlMDdclbPUd/Zng7IrkwnFPu5nz4lMJV8YDbo1zhrlESUVcLwjYyCIifZtxlp5WKyJNMrR0GDcJHro3eDeiKgfISVVDSOCcxXuLcxXOG7wrDyKWOh8hvQ3FivgDNqGs167xFu9mYe0dnQAKZ+Si0NLZABc7ozAIvHWhXker0EtMxCgFSkmUkgdRYN3+SFmFtgbnQslEQPACEOjraFAIHSIWXyEoEdh5GnfxEIWIi7prc53Tu72v3vOQ521c4lhTVVN0UrK+0kGxx82b1ynHBffedYZIOU4eO4JMMlrLS7jr19ibjDhi4MmnbrC2cYK7z6/j82HoWmpihFE8cO9ZTDXm3vsusN0fsjfYZ6XT5tbmM9zY3ePmzSFnzwmeeOLzHFve4MwddzLa2ydrJjgZY6UnUppEBVy03e1Q5BYvMxIpiSPH3mCHi89c4+4LZ8iEJELy9Oc+y6tf+RKm+YTPP3mNhk6g8ownI6IsYdTPufPUEZa7bR546CGipE0aC7ATptOci5ev0251uPOuu4h0gygRKKfZmXkMBbvb27z2FS8Ho7l8+ToaOHfhJDpuglT0Vo9y/NR5TLXL3v5NtI7wxiOdIE4i/FBy/fIuRTkgbThUsorHY+bMH1GzOLQEEeAx5PwV9lCvTuldeMCrsmLS3+HapcfY2dxFSU2WxBhbsb+7y2R3QpUb8tKQxintLKaRxoikR7bcBDtATgxxkgTnVwpiGdHINL1OkyiOKKqwcFdlDJFOWOkts3RkCZEmXN0bUCSK86eOc+xIiPpsUVHlBZ0jHaaXx+h26CkljMEMc0QOaglU1sQbh1UCISVKxDirQJXISNI5eZTOtWuhlsAHGEVoVUdCHmdD9bT1jso4ZnlFnhvyogodBkzFdDZlZ3uTW7c2ubE1YjqrQERINLFusNLtops9doblHO0JrCUZsKd51CYX+ZF5K5R5ROfqezM3JLX2rJX6Iu8y16xQG6aQXxM+9LPF1Qn2EDKF4j1XN6f0IhAZpATnapq2q0kAAX6r0yshwnBi0acstP8KkWESLfHAPd/M3SceQEcpRVlijWFOt0XKen2U8DykcYfh7FaAzFyN2zM/R5jcfOEzP78ewRItDKVbzL+mLHvCfIVH+Qjlo6AYEbQSKFor7EwnGDtD3BQ8sSV46qbgFWclW3uGdHXI8fVzrNMOz8ztdUjSY7E44YhjwcZqj6XuGqmqcJUBXyBkiSkHzCaKskgQQh2QZjzBSPiqNqTB4EAgGQSIMIw9rCg7P/G84BIgRwmBiiKk8tgKMCFa8c5ibYmoc6tVEWqFdSRAWAQBElO6ph47CB2ZDdaUOBPjo4gDCeMI7qWo+5iFAs/g5JTgDfN1gpDhfXHeEPkQKQViwUHj0ucjX0bOxbF7c5OXvepB4lRgpwrlNRvdVU4eW6PZiUhiB36KtBM++anPkCQrVDPL6tIaZ04cJZ8M2d68TqPd4cjpdW5ubnLkyAbXLu2xurGBYEq7GRNpGE4Krt3c547TF9jf2iRea9Do9rh2cxvrIqajgvHkBuOipNNeRnmFVwIvwmqMm1tb3HXmJFLMkGqJNF6icgXXd/eJoyZOj9nJp5xaWue+c8tcfeZZ0qzBXjmmp1JWViJ6nRjjLc1uQjUdkYiEaT4lXV7jmau3WFnukSQaKWA6GjMeDsm6y3zsk48RCY0Ukm4MT+3ssnFkiXY3I0kkTgraa22qvKIYDJDW4ohI4hbT0T7TyYg4U6SdFJFHRFEPZ0MFrrHzpKvDupKytBiT4xzoKAketg5K2llLWeTMJkOmowHjvW3yYUUiMmyVM5ruUeQ5+bQgUmEFRiU1rWaTZqMZIKulFYrZPpGf0VAa0ZKUxtfNBgNebCpLFEc0Gw0aSUycpCwdXSZba0MSc+3mLlv9KWdOrXLy5BK2KikGhmq3QMcS15lhxhOyZhNGUwQEWvKShlYavLrSIrIYn1tESyIbMcJGuKpAd9ocOXmU7f4YL2OM8ZiyRAkHqk5xO0lVWgajnL39IePJhCKfURZTZnnBcGIZTSocTc6eWiNJ4rDEsbC0GwntVoOJz9gZFgvlvyjPWyicQMSYb5s3nJwXG/p6LZw5JDR/UedsLnGbgRG1MhbzXAHUNSQyEHakw8t5J10WUUNddhGUW53/sLWiCyhZMELSCTx10Z+XCKdqb1tz/MgF7jx6P2ncxliHMSGalcKCqBPxNY1ZKc1S5yS3dh9HCE+sMqQUzMpZ7eWHAfkFEeK2yARC7kJQ1+3clnma52QkoT1QraQFoD20Uk9RacYT2Jl4RlpQWEWr2+LCkRlZo88mltHcWIn5/QoMwnnOzOPQOiOKV4giT0KMEwPKcsZotMVsto+SEbImzUg1b61SLyMuRQ1VyZoBOPf4Q25DyXofEdriCAKLLhAzqgDv6gABGymwVYgMnTE4UeBsqKEJtTQgVAReo2TIH0opw3o6CPAOawymKtBxjHfxog1MMOhu7pOwwGO9AKEC5FWzUaVQoQW/r2n+DoQM0Y57nmu5wJcFi8HdZ05x99mT7O7v0I4ykjTjrnvPk7Ut1248y+buDuvcyZNPPctwlHOiucby0jp3HI9opCnDwQhjU4rSk2SWwpTc2rzGlSuXWT9ynNMn7uTi5atcv7VJN0t59f13UxnN088+y8cf/Tyt3gZb27scWz1CI0r4xP9+lGHpufOOGBqCZtZFesnO3jZHjqzQ6sZcu3IJ5yIajTY3t/aIWy1GwylohS8dQjZYParZHWyzub/Pykqbtc4STzz+SVpJG5V22N8fccfxo3zq0Y/jFVx48cvY+dyT5NMx1565wqmz51AuZvfWLt2jMWePrLHcFSw3I57a3KaYzMgrS2VAFBWlKdnbucGJo12ET1lZO8KgnwfefcdjjaU/HdE7tsrq0mlirdje3sb5aajrcBZrw/onVWUxNjRjTH1EpB3OKbwzTEcDtjevsXnrCtN+n5iIZpwSxZ5RkZNPRkymFZW1TGY5pRFonRBJRaQkOumge03MbAvTH2O9Jo5TIg3GVRjnkASlnWWCRhLRPXWMdLWFimLKsmR7b58bt3Y5fWqFo0c6mHyGs5bprsMNgRa4wtA500aMZ8jKo7oZUiQgGkgUxeYI5wWNlVbAgL1ACLOoDRNRTPfIEWx5DdFqEilB1R8wHA2IGk2EDAuTDUcTdvdH7O33w4qepSPPPbMiNEJcW2rSaTdpZQm2KpkMRxSzKqykKTymXraAumPxvMhx3sxx4akuemQcdBK+3ePzsMiTeO9D36g5hFLv42rK7jzJPl/wy98GwQXGkscJGUgMtRJ1deQkF/mYGiytIwfp3TyHXyNrQengBFna4ujyWRKVhXbxi7yCI5ISa+uZCYFG46Rnbfk0TzwbcXrjAmdPvhQdKZ569lGeufqJ0CJFyHnHnKDE6jDF+0B3FfXYYA6/1On/urjSRyFCcN7gcCSRQIoS181oJjHHWiOE8FSVYlgKxmVMJx2zrp/E8AA5zfpugJMeL0LU4qiofEV/mlP6mGbUIDYCEwmsDQ1XrRE1u6xEytBjS8kQVYm6bY/SYWXKeRsqJRVCqzoXKlFehvV2BHWX6rmhVYtVLYXSSK0wSmDLef+yUNyo0KG7sQ+NZyFQm5UO0ZSQoi56DM+EqcpQhBzFaBmc3HkvtIPcXl0EO4dza2MeLH2dQF2gmBWICmojehBhv0DGxTmLFxWjyZDRzh73nD5L1gThpmxfvcL+1VsMxzmjqeWzn3qcVzz4AM3GCutLGesrbfa2b9Ef9FnfOMtotEciI7rNFf7gD36HI2sraK3QUnPuzjNM7QQzzsmSiMtXb1GUYdGsxx+/yJ1njyJ8Tllabty4xcRldKJd1s80uX7pCsfPHWNjrcfu7pDp1HP9xojV1RViXaJtzqkzR0myM+SzglSlrJ0/R779FOvrKXv7M46snwVgmksGuWe1kzCdjHHGc2Ozj2xqznrHZDKhmhlE7hlNCoqZxbuE6xef5uzJIzTbCRcvXeGxp67Qn8yYDCZEx4+wt7dPaSzXb26y+q0btI6dR81mFNU21WyAziLMBPZ3ppx/8B6Wl1fDsr/7BXAzRCzGYCqLtcFLTpIGkdah+M9ZXGWZjftcv/wEl5+5yN7OPpEQdJsdVGaoqgnj0YTxZMJkZiiNZVra0GfLVxSzCXkE8doJRAxJJbBOEjcyGmmTylRMixmxjokiTaoV7U5M7/wqSa8FeKqypN8fcOnSNqurTY5udLCVZ7o7JW0lxE1Q3QizW2G2JyRdRdRS6NUWvtCUtyaQCeLVLkSaxol1VJriMXhjQSqEKNFpFGCf5Q5ddwzRDIs9xbFi9sxl9qsZeEtZ5kzGOflsEkgIcYxVtfKFmh2X0chSqqpkuL/PeG+IsY4sy4gUeH3AapJy3s3WP8d4+AV04Bb5GDFnaM0r2uflL6LuUuwPaj4OcCvqflFzDI6Aec+VtD1gd3kJyDpXP++pzu0de8HXXqeEus1+UOyhUC8cwzmIVYMsagUiiA8Lttkq1HFUlOAM3iq0UlQatBO0Gz2Odc/wkgfexNLKOlUxJS9zbt58nHE+hrpWRtSdeJ2fc5dCa/x5Yb1E1Kstz0kTDq8CcSUsUhWgp8I4Kqt58Utfx5HTp/no/+99VGbAeGroTyo2+4LZULDc3mZj7WlucjclitABLLDgjDRYX1F6y7W9LU5NCjqdZXTSIzIW4SWVMVQmrGjpcWDrjmki5B8CpOiQlqDwpa2XRagQ1EsiKxF+IoWOYyIdo6MYIUIkNF+JQAoQkULFEleEJcK9M2GZB+mxPgnPDFWIZESoT1EqRDwLcgg+wGpVgTMRLtIB/qrJJ4LbapLmoS1u8czM2WEhr6bwIiWQhcoAsy6KcF9A41JWlnEhmOQTsAnOWlQ14dozT+JnjhURo9hj86kn+L//6uuZTqfk05IkNpTVkNlwl6efeIb1YyeZTvuMJxmnzpzh7M27WO42UVoglWG/v4fVCbe2xpw9s8ascGzd2mc8qjCFoJNlNOOYmzcuMhjuoNJ1GonGmILcFjSbTbZ3blGZiiRrct+DL2F3d5vrN66wu3WVlaUOS3d0EV7RXVshaUVsf/win/29P6Js9rh8a5tWQ3Puwj1U1lPZCefPniCKJfe+6CFkqpnlBUVVYYViOHH40YyXvuLr0Z96nE986lk2txTDq2POn32QB17xCmZVwdXrTxHFiqWlLuNpSZZk3LryDCebG+i0Abpif7jNylKL8c6M/u4Y5TRaC/LpLNxc78BV4KsQckdhGQQpwsqT+WiMq2a4qmI8HLBz9QblYEriwGMZjfqMR2BMGRSwStHdiFJAXFSYokAiiKMEHTXwaUyZD2kSs3TqHI0kQB6TUZ8oSclaXdJmhlIlzeMJPoowVaBgDoYznrm4xfJSm1PHe1STinxicTNLpB3lIGfpVAsrNcIaooYi7qygmsu42KOKiHitAzpCtTNUnAAghAYl8H4WCuG8weSOKhfQaCN0wJRVMwk4uxeMxjPGkz5VYYhcWICsKguqsiSRnrSp0UoRxRJTFYz6Q/o7Q2xlgoIAokjhdIyUJVGkF4l7UVNr6/e6rpgOHqFzfmFc5tGMq3MvLJhfITE/f90FB6tQBkTJs2h866kjgTprI8A5sSge9cIHD9lSrxclgofug+qeFzuGwri5N+sW8BlSUFZ5aCBrcqRXWFNhqgIhFLnJieIIFUuqqlpAYxvrJ3jwntezvnYqsNDynCOrJ7jz1Mv53JMfpnQOgb2NNVZTqOvIbd7KxopgHAPsFqId50PbpaLIMSZ0Re5kgtlozOal/83FJz7OtZ0+6z1AydCixxtotRj7gnS2QxZfJ083qFwVugr7sFpkRUVpDZNhwTPXn6W3ssJSo0vsauPtg2MiPBgX7i2IGlZT9X2dExEI91uCcPVibi40HZYVqApMYai0QUWSOE1IRIs4yfBK4aRHKonC4+IKWZXYYhqMvLM4DELEC5qxqGFOWbcgup0kAjWjzZRYo5FRIAOKOmoRwgdK/5wgwvw5nT/GB5TjMLe6jscbDnorvIDGReuE4ydPM9i9xPH1U5TjG2w++Rj967voJMMpR5XPEDZjrddg0xnOnjpDYXbpj4dMZxXrG6dQcUyj3aYyhqtXn0VIwdr6Ufb39lnqdel0Ojz68c+zurzCoD9CeXjlyx7i3N13c/zEMYQZ0262ad55gU98/Amu3hjggSMnN2gtt7i1eQuPotdtMJ3uMplN+fznH6OVxLSay9gitHGZVQVpr0taFPRzQdk8wtRphmPBU08+wze+9hu4/vTTrC2fY7i9hcksp0+cYjgd8cyVS2wsbTAqPLOkwYmVZZwtePbSk9x//z0M9kdkQtDtJLgyp1VlPPuMYjIpmM4KkmZGo9mmHAzZe/ZTWN1gMhwihaM/GJAkCXedO81sss94TzCblNjJbn2jJVrGNTsMqqqgmEwY7+8w2NmmnE0QAqrSYPKQS/F4ZrMZeWlCvgRPmmnavTa95SPoOGI2HTMe9jHGhHYVS2tYaRBFzlL3CI0oJdKKoihxQJJour0WcaawDY9sRFhv2LvYZ+/WmKHI6a20OX6khUQwzQuqUpIPHOQljW5EtV8Q9SRJEiGSNj5N8c6g4oToxGp4EWRcJ0Jrw4IHQmM9aw1mYiiHFTLVRHEjtBGZjNh87CK3xlOm1YjJNKcqA/XSGBcaeOYTtFREKsY7T1FWjIY5k8mMyTgH44l0hJISax15UeIiE9qASB1w9toIyBpXx4O0NlRl4+r27GIBZYVCxoPE9pztJaRYwGAh2X6w0t9cCc9bb4RgxKG8Dy1gICwjoPzimBaPcGKhyKWo26xAiFDqA0sROvsKKRBOIrwPXRX2b7HaPUaWNXDOYKoSOIB0pAp1F5gC5yBrNDlx+h6iOAp1RgK0TLnzzpey3b/C1RtPYUVgroUiUlvDLQFystRRlA8VKb6mbVtqo+Mc09mYsgpO0VpbI2JP6a8iXczxtZjxtCQWnjQRZO0maZpjTMmnrwua7QFmLWEaCUpXUREg3YKwhLZ3ik8+/TnazR4P3PMimq1lhIPpaBshQucJIQjdvV2IYuaRp5gjpb52NmoiRUiMi9p4siBQSOOQ5YyqHFMWAxqdVbL2ColqIJVG4NHa42ILWYYrJ7hyijMz8DO8S0Ndm0gCi0sd5OekCOd3PnRNtmWJ0sFwCakWTUnnlJJ5lT7UAfFtuZh5dCMIxbrBkN1WzfvCGhfFai/F5FNMMaISHjut2Nwe0exVNHTKdFxRUBJpy2w8plop8FQ89eQzHFk7zerqEs5Dq9VACUlRGk6dOoVzCmcFOtJs3rrFiaPLzIqCzVs5zVZCsxORtjMkBd1ukyjSXL16k1d9/TeRfvoxllY6dNdazKqCsszpdnqM9vuMBoZ8NubYeoeqKEh0kzhLiLRmMByQTyYsH0nZNxn7RnB8Y4UzGxuIqqDR6nDy9ElUFBM3myRZg8rkpFFGIiJefM899MdTdvOSe3TCk489zmg0YmnpHtK0xZPPPoU1M1otRV4I/ur/9UqajS6Pbz7L1auXueP0SdabHjnc5ObeiK29IXHWZGV9jWk+5dSJo/RHuwy3NxmPp+xv7UCc1E0TLa6cUZYzitmM6XjKaDBmVkq8bGCqKWUxY5bneBMWTbKVxZQGh8VY8FIRG0tc5UgFkXAsd7p4FVE6j++2MUzpxg2yJMVWlulwj2KWU5ZTtIY0kVROQ5Lhco3SnivP7PDo/7zIK990jo3VFlLE5FPLdK8gaibEMXhfoRNH1E7IOt3QjiNLqYY5tBWqpRFS1+1u69xCVUBUu1fe4I2iGueYWYFqRugkKDw7GdN/+mk+/+wV9qYTrHUkUUIWJXhgkhdMpzl4iXSK4XjMYDKlqhSVERhj0RKaSUSSRugoQipNFCsKUdM/E33wmoqQ0J23zXdO4pzEGIOxJnRoDqsz4eaNLucvlfeB3jsHNPxiM8+pxbThpffSHVCPZcDf5wrO1wtx4QRehcWwahZA3UNqHi0IpAtJdKfC5dSuzkP4ELns9W8yHPdRMrTBr5ylqqrQFRdBpBPiOAnG01V4W4WuHUKBq8KaQsaRxU1OHX8xW7cuMasK3JxFRSAnVHO6sicklet52pp+LURtZIUjL6aU5QxnHVpYmrFneTVDxDH9sWc41YxGJcsNQ6chkc6yM4SLmyVyd0B7d8rShQ6DSmFEBUrhhcKiwXr2xyMe+czHaKYt7r7rXpLeeljCYrCDNePQRl/J0Pxz7vWL25QyBw6BcB5q8kON94X9rQdlQkRqJbYoKEYVWkqSThpyNjXkpnEgNSQNfCRxZoarQi2NlyXKR2it0VpTyVCfFU4+z5NYjCnQpcDKkJcM3ZbnsJZbpO5C9GXrHExYCkCGsHLBIJsvre2/EpFLFClaWcKk6uJUi+39GziRMLCOST/nSAPGeYlrpIzGOVpopPRs3bzJH/3hx7j3oYgXv/wovV6LybCAsmJa5kRJGyktx08d4/EnnyaOMpJIkhcVJ04fJa9Kmu0uERrrJYn23Lx5ExkJVjtNvukbHmKp12ZS5vSHEypToiOH9yVplPI7v/VRxsMhr3zVi2j0lsnLkqoyKKGoihIzDYnsZuK45+w6zSymt3o3u9s3iOKIwSgk+Z01NNIE7xocO3qSstznxPEuO3u9sASqh5e/+lUoEfH5xz7NYDjiSnaFlfUj6KSBVzlCWppZxur6BZKoYqkt2Lq+xZVrOwymBXecb5MkEf1ixHQ0ob+7z/raKutnT7C9u0deVUyHe1hThrVlSoOzDus1aatH0pah1mTaZ+gcVWlDIz2bIFNHLCUxYU0UUOSzgsptE0URSRyRZQ0aSYpMMiYKVGVYX10hSxNmZhKYZ8MRxWwKAhrtFt1TRxGpwFczcgvPPHGTpZUWJ06ugojYvz5Ba49SIG1JOc7pnGmTLTXJljoonYbV/urCTBVHYUVEyoU3iNA4W9ZKVGMLR7G/i8Oj0rh+Qj2uLKn29rh28Sr9cc5olOOsp9QluR6BF1gDziic9QxnOYP+CGshijRSKrJIksSSKFboJCJrNEiThLShGfoEOXagojqhP2cPqbrtfegs671HKYusSowx2HopXO9rb1yEZHvIh6hFzcccy57Tmufawnsf8JW6yaOviQMIgVBBcc25vF74GuJyNYNtXgPjcPM2+IGTUHd3YBEhADhfMRhu0x9u00ibKKVx3lCVM6wMreorXaCVxhOS2aEGUmCtoSxyyrwIJAAkq0un6XaOMd18Cu9FsJN16/x5HzLn5+3wa2VnA1Qmb1sus5hNMeUMj6eZCPKBZTSYcOZ0xNnjGaaC6V7FcDRCzSyjwjEbenpaIDKBiCOkKcltioxC9KlUsIdWhiWJbw42+b1H/ydRFHPh/L00V46hopTx3g2KyU5ofS8OGhD4ORFi/q+QdUoiJFG89SDq3m9CgBQIXwXGmAuOhSlnFJM9srSL1qEVC74CVxD6HYWIT0dLyChCyAjrI6gkpXUksadUU7ydRyPz6+jwtqQqbehUIQRe67p5aIi+5o5OgCvDvZCYRYdssVge4cDZeY7T86fI86cie0FROBAJy0dW2bt2C6ssvUaD69tDnMjZ2x/wV1/1EtrdHknSwVnPoD9lub3Ck088y8bGGiu9c0ghmUyHbO3tc+7YadxkzO7uDtuDCXefO4YvBlQWtgZbKGLWVtZoxDFaR4z6Q9bW1rhy/TKZHlEO96Bzips7OfmsIIos+/v79PcHbBxpkqYrxK7DaNeQRICq2NndQUrJtetX0UmXzc1bnDh+jH5/TFFZnMoZDcY8eN8DCOfYvHmF8WiPlY1jHL3jfvaGfbqtHo995pPcfeZummmTG8MbtIQkbbW497570VIz3L9GljUYDBSPP32DFz+kkWmJ1JokdjQbGTpJ6PTWyBnS7bW5dvUKJ4/dyXhvSDULRYk7u9tEiWJ/Z8Dm9WtEOsHjKMoJk+mUWZnjRUSv0yaJIuIopreyRtps4SqLMwWz8YjZaIj0FoSi9J5qmofq9NYSUgsG0wGll0RZh6rq0yhzTD6llBJTzRDeoVSEF4q0k9G68xjEEi8txhkuX+pT5jHHjvbIxxZjLOWsot1RxJkm357SOdKivd4m6zaxRYmO20gdLWoIQsZZ4H1+kDYUMV5KfOmx1RQz3MIjiBotQEEUGlcKD3Y0Yn8a1s2w5Yy8sEycIFIRjUYLgWKWD5iMppjKEqsYFYflsb1zaO3rvk6haE1rTZxq0lgzMwKlIyKhUVqFpb+pIwgp0Cowf5xzlGWO9y4saIZfGAxRGwI3B+q5rVG+qJP2tcJakAMEITkvDvadd7D0XuI1ddgToBjpA20dFaKi0DNy3lIfDlof+1rhz3H44GlPZ332+rfottZoNlr1dk9V5uRQtyAJbevjJAkMIsCbiiqfkRdVUKjSEccpx489wM72dfIqD7Rob2vnObDdfJ2Dmkdxqk762zmdTQjyckZRzPAuLP0xKASRNkymjsJZ8mHBeCixhaPXEqx0NaopuWaaTJotZkDfaUSsiGONjhVS1yBrVVGUJVVRcWnvCv/vR/9fJIK7LtxHY3UDGScMtyNmg02cm9b3oPbuFys+Hqyys7h3ePAS513dZFOAiOp8ja/zbRJsicnHaCnr5L4AGSNkE6FihEzqPNQ8qhBESOIYdGRROsJWBeCC0wLMS26dA2tmwbkT8SKHF6Avc5CgrxEC6iUsgi2c9yYLz6CoV618vvbleRsX78Nys6u9COcnjEYjvBP0eg0G/T7j8Yy9oeT0iWNE1R75eMJk4mipFdaW1mkQc/nZayytLrPclhhnuXz9Jj57hpaOaKUZ91w4TxZJxrOK61tbXNm6wZmNM6wurSNangyHcwXj2ZTd3RFVaWEwJG4PkVGHwdYO3XbMYDTiwrlzDPpD7nnoXnpJyvbmTaZFRSo7tJeXGe4PyWRKNdzi3LEWuYf1jaOQhyUBlltdIuO4dfMWpiiZjaZcGj7NYFry9DPXuePEadK0S9aI+eynn2D95El6nZSGrsjaCbGMoBow3OtTmowTx9bxIsH6CJMXaCmRpacocnQk6PTaxGnKcncJZ2b0VrosrbbQsWI0mWCMYTQacPHZXRrNNh4oq4LJrGScT8iyGFfOSJMGWSMjy7qsrxxDUjEZbFGUA8blFGs9UdzAC0nlPdKWCOnRUmArT9XuoLTHj3PMbMau26TZGKOlqhOLJVknYe3uOyBJ8TIsl3xzq8/gSkU3brF+ZoUbT085csaTdWQ4pxZ01tosHe8RN1O8t+hmCxlHSB0tsGAp6xbgJKH62SuwoTeTNzNc2UdoiU7aqCzBVFWIBHSMcIbJ/j6zylFVPrTVLyFJM9q9JdrtZcrCMpnkCJ/TyIIRsc6HjgVRgMLiNELriCiOyJoJSRKBCB1j4zgJzTYBpRRpHFqSWGspyhJj85ALsgaEINK6po4fRC+i7sw8T5neDoPNiwlvh84O/gpt6Re1IgI8BheWWwzHrVu9hN5ismak+UXVPk7gRCi+lIuCGxcq+hEI5yiqMcPJDpPZkCQObdyVklSVqw1niRAzWs1uvQaJwlaWajojn+UYH/J80oGSmrW10zSbXYr9KWJOSnChvsYh6/nX/LG6jEfX2fxQbOqoiinj6T7WSbSKyWWDjY7Hxxk7YoV+uswoH1LJlM3EE0eSURRRyBikxDqPUJI40egkJkl0aAukBNY6dFFS5CVlXnJlcIP/5w8/QFUa7rvvIZq9VVSkiaKYcf8WtpjgKYMCdzbkIwQslklALPIbcx9iXjyLl4FOXOdgQCN8ilIdlF4JeRfl6/zWHHq97flAAKEAUisbCCcytIWZY2MHjUvrJ6xmM3pnQ7PqORHBuxoOMwuSSIhS3GLFU1HnW5RwYbnkr0TkEkWKUyeXsfmUvMgpyhzng5FotjTbVx3HN86zlkpEIehv9YlFyp1rp8hHBY2jG8wqzxPPXOH8qRUiIZhNPGncpj/Y4dozz/ANr/tGbBWqQ6dFiSKj358wHo+I4kAXRZVcuX6FSmhWN07SPHaMmSvp728jpOKpZ67QaHmQJ4kT0IlDZB6XaIrcInSC8RKpYtppSmQllz/3OdbvvouTp5fZeWoLNyuYlhZjV+ktd5gVMZdv3OSTn/oM585Pueuue4nSlNFIkiYNZnnBZDaikwUuhReeWVmikwaqhLWlZa5cu8LlS7e478LLGG1dZCWNcXmFNY4Tp07RH4WW8LNpQTvNmObjwBQrPPnM8vjnLyO059Zgnyjpo+KIZmuFZnud7rIMy7aaCZubV7FesL5xgjXh0NIyHPXZG4zYGuRUzpPEYcnS2XRGYQ3TWU4UKaJ4mU6vS1XuIwsLZIwmI0w5pZlkgCTKIrp33IHstKgKgxIVW7tTItWgGVewIf7/rP33k2Vnet8Jfl53zPXpq7IsUPBAA90Au9kkm2xxtKPh7iik3VgTsX/g/jARuxvakeEwNBRFiSKbVDs0GigAhfKVWenz2mNfsz+852aBGsWoGdE3ooBCodLck+c8z/t8n6/BuxgdazLF6qSm10/JJ4bx9hBbw9EXl+y8vUk6yNA6oZOHR/jL+wgLBEXwMp7MnQO3QPpVFHPlA1TaAxGQtjtFSkk7v+T5s2POzs45v5hFVl7ep5f3SdOMIMFicXhkYiIlVMaHPUkFvV5K3svIsgSVJBiTdh5lkuAtsqmRIsYKZGmO0Ya2baiqOKWkqWFoMpqqpKpKCu8obdutPmJRWBccf4VpxwJytW9BdCkvHcxFtHGJjeSV6j+6MXTQF+s8d4FQHe4V1vCIuNq5BAehS7QUrM+2a2gEhHQEAt7WzBYnLJczhvmEJE27JtJSNzVSK7RWJGkSg9lcQ1vVFOWS1jVXjKJY0KCXDdjff4fLyyOEdx2roEtIFKHre9HVTHTN1HnxaucCNKFhtTzH2zFtW1EXJU9Pc7xpKIYVSxWoNz21k8x9QFiNNppEabQ2CBUbiTaaJDOkSUqaxpA66wNpamnSmiotqcuKk+KMP/3rP2VVrvjoO58wHo8Z70UtyvLiKCbhBtsp1123T3IdXbz7KYsorIzrjRBhSbn+qapIH7ceQUs+CKQD1QXBxQlVrnd6HWU9zisCUJ1Dh++sYCJjtPvpsRbvCiEibCtUtxzqmIrd64poQJxa5NX90EFmwneNMVw1nCD+3qf43339A5IoW04OvmF6Maf1jtl8xiDbYGWhaCVepvzhH/8utl4we3nG0csThhs7VCcnmCTh/OyYO++8z1/89O8wwnJn/zqT8R51UeOqkoMHX/Lk1m12926wsbPL5uY5e1s77GxO0InkcjplOl9SlQuePD/hR3/4R+TGM1/MWNYl0+k5t27eY7GccnZ6zMGLU8ajIePRhOW8BDlAywq8I0sNrpJkk4x6EXjt7U/Yf+8O1fSci+mMlyfnvPfB29ShwTrHxWzKZdGgB1vcvHWX3Z0trJQsFhohHDs7GwTXcHF5yrPFnNfu3sJ2VOWNjV2SNEeJhDfvvtlRjQVpL+Xk6JiQGOrWUVcl2cCwKgtsGNPULVpklKsSbTJuXLvN0+NHTBcLZCEZjsZsbORsbYzJ8wTXNhydLjg7ndJaS1UUFPNT0iRCBTiJMYqLyyWX84JECfIkZZDkKCsJGHo3bmBDhS8KNvpDpDRUhQcbw4ySrE+yew0z6eOkpyhLZmcz+v0MIzV2qPDSkeTxZm+mltEkZ+vmiF4/Q6TxTLf91ia9jXFnrhd5+9GS3oEX2DKm/AmV4v0cg0XUFu8MrlGRgo2IJoLeE4LF1SWLw1NeHE95eXjKclWR9Pox9Cl45pdnVI2lrlrqoiDRmmACSgRUoki1ZjBIGA17ZFmK7Bb5CBGtV6Qk0ZqNjTGbwzHL5SomdNYldBkfbeVYzBvatqaqKto2ZnZIITFZnHasjQylb7/W/mLrPYtfP/Ai2odGcVt32g3rJtOp9jvKK0qAclgnkaioIZHRTijITra/fgm65Xl3zdef3xEZecJRlhcsihM22z2EiQHIUsYz87DXZzAYYNIE7yy2qSlWK6q6uoL71vsTFwIow53b3+Pp0y+Yzo6IViOqK2qdgI914Vs321gq4/cbm9RqNcPInCQxaO05OKuopWBTaMqeYNlG/7iAJNFRWyJltAFSWqGMisLDrujqtfJeBFKpqaSKmURaU5qKi+WUf/O3f8rR+RG/9/EfcP36PsOdm2iTsrp4SVNc4n0NQkPQsWF3TLirbXlnZS+JmirZsQJFCAgsiJq6LZjRIsSK3mQflaSR0CI62HU9iQQ6M1LRQZMqGliqOPF7t2anhQ4h7dT5nUuM6OBVIQAVUKFT/wsZG4oM3a9InJEqoDrRqBAhMtPW+uDf4PUPgMXiFwk2cHR8BB6ePH3G9tY+3339fazQbN3IeXTwnI1eH5dqzlczmuWMo7MpW7d2afwSkyZ88+glw/42g/EmvX6PUbKNeONt/v1f/oQ//pP/I1tG8fzxN7x+53XaesrpyYLxZJNrt27TesNb73xE8C2HB0/pDYZcnJ9jG8uqnGGdY76wpMkY7ySz2Yxhb4jNLdlwyOt3b1DbitGgx8OvvuHG9Vvs3BjicJSLAmdyTqYN1+aOn3/6N7i25u7du7x27x5vffARPQRHR0cMdzbY3plQVQt8XfD0yQG37uxhReByeo7RitUyYMySURjx2t0bVKslOnU4o3l4MkXpMVnaY1XWTC+mOBpq20ZOvYPz0xnPXz7ku9/7iBu3Nnl6/JjWg5aKft5jPBwwyHOUcMyaeFrWiencUqMyuG8G6ASMlBTVirOpwHqQRrE1HjMZDaKtR3+HMOnj6ynX+kMmw0m8lmGFrx1pliO3t7C5wdcllW1YrhqG/RxlWyyQDiVKGfKegcayMc7Z2h+Sb/WwtUOphN7eMA72UiK7QuZrQTtfkkx6uKYm+JhV4ZoVUjUIYQg6w68srg7Ijc5DDBBS49sG4VouDo84u5zinWNzMmE43oyFoiqZzlaUZRXNLgGTy5gI6CVGGBIT6BnFuJfQHwwIMsYoNG0b9zkShAxURcVFY2mbhlW5oO3osc61NE3L2kNMSkme510kcMBaS9M0OOf/vlo/cFVk4+rh76v61x5c3dqDdd0KCJztJhopUNJfeYUFHyOvpeMKXxcy7leu7GEkMS/dR5Gf8IbjF5aNHchGktbVzFYnlNWKNMmQMpIeBoMRg9E2SZYjRKCpSuqioKnr2P/W7UFI1maZQkr6/RF3737IZ78+6ZrxelRbM5ji+42rpVfXIv6dOGNV9Qp0xcZI88aWoScaeoNAnUFhK2oXEEF1yZXd53MeL2OqolQC5yLmaK2lkS06BFQ3vSkh0FITtCOkGuEzqqLiP93/CUdnR/zhJ3/I66+/QX+8jTYpxbRHtTzFtUUHLxEnmU5LAvE9CHycAq6ahAPhWDshB6EoyxXi/AShh+TjHCFM11xCB1t1asvOLjl05qlGK3QSG6f3Ch/igSESJDzSR+guBPWKmec7tf4Vuhr3Z0FG6DQKg11sTFLGqUUGpIqQ6m/6+s3tX4QiTcZsb2VMF1Om80vKCt589x3K5SXHpwuKFwV2USBPTvjm4SOW0xVSaTZ2rnFva4OyuGB5McXbnP077zDZNBT1kuVMILdu8v3rbzDeGpPIgg8/+A5C99DKU5Zn9LKG1WJJnmVsb2ygZMX9z8+5vP+MVdGSDxICjtmqRCaSyjbYoAkioW4a9vc3ET5Q1SVKJ5yfnzMZ7/Hi8Ijt3S2kOyV3lulyznJ5wcnzJ4T5S7Z3b7K/f5PPv/6KW7dusTXcojKK47NTPI5iuuDZw+eclQWvv3ULt2p48NUDer2ErckNUH0upzPyvMdossHTF0csrUYpy97GJrPzFd56lEqpahhNNiiailVZcnI8Z+/GXVSSMxhClmrGwx7OxYLobUO5muGDpaoL8mTA/rUU15akRrKzsUGWaOp6hQ8xvndr1KOpPfkgZe/6LjsbY6xtWeZblKJioAK74z4mkTR1Sy3BmRSzex3GKcWqoChLAophluEqx/mZYzhRGKnwyxYz9mzd6JO5jjqtDL2NAXgf3W1ld8rzFoJgeRLNQr0DqXvxofMXMXEvJIh0gOz3wLQoKQg6FlUpu8WoEzSrFccvz7DO0e/3GY/G9Po5VdXQNJG5JHWOJYYyOSTWxyxB1zp0TVQ1+wbvHLb1VPUK6x1aGaSMrLSmrtEmxua6tsK2Lbbbp0TvrmivGHygaZqrfUu07XkFQ7zqL2t8PP7TrZf+6/Pv2pcsfhQCEaeBtQ4hZvxGrUuILsKhw/j9WtD/ra4UVDSdlAFUIC7RvaSeTxDhlDRzRLsYz2J1znJ1wbA/AQRJkpP3RnESVxrbhcFVdYPzceJaf49CiquvLTttz+2b7/Ps6adcXLzsYLlXkOCakC3CWq9D/D5C6Aw4Ja2t0KJiubTYIDBpQi9PCDpAE6+Z7HJOpIiQvW89rbdorzHeohOFDwqPo/EWIzVaxswdH4gQrG8RPrKmtAZvPF+fPODs353wyYuP+eiD77Kzs8dw9xbJYEg5PaVZXeB9FQWhHfAJHZu+W5rHSWL9s1WvYhFE3D1VRcH8/AghUrKBQmnf7VrixwjhEaFzCxABicdoSBNFmxiCdfHQ5cPVpIS3BC/wTqCEvkq2EYFXDc9H5uNaCSM6C6PIFotfUxF3/SL4b91P//uvf1DM8cWiRKuIu9etp2wszrckiebm/g7nRc3XL57SrhZcv36PfFwidMJ4YwRScX54wlZ/TNFCf5JQ+wUHx084Op4xSHL+8MPXEM2cJy+OWRQtr9+7xmSSUTcrNgYTNrY2eXn8kkQ7lkXB22++x+nJJU+fnVCVDQ++fIRVBq1HDMfbqERx49Y+54fPyFKNcx4nNeWsIDc5+9e3mW70qXzL+ckZy7rh8OCcQZ7TzI+5PpSUdYFUCePBNrns4d2K5eKUFw9fopLokzZtJDUxk+Tho0dcHJ+zf33C5sAyGY1BC/Bt1IdISWgcVrRIBLu728xnl8ymCU+eHHD9zn5EZI3i9p197tzax9oa50qMUuxOJiBSkizHVTVFWyMkGJPQG/URoaYu4s2jg8U1MamubS1JkrOzkWDblqyXs7MxYmM0ZOEVMx9QZclEKdqqxLUVwQmSNIWtLcIwpWkrVnWFqwPDscHWgdmZxZYeMzFICzu3M67fmGBEirOObKeHSRO0yUCEK/grhkopbN2gMo3upfHkaxdgCnSSgDeIRiOaqHCXmY5wiYtK7Sgi7CJtVwWtheFoiKlqTKKo6oLz2ZLFskFrQ6/fIw2KpqlxvqFCIOgiC7QhiISqFrhQ4NoG6z0mzSNDTEmUBOsarAw0bYl1rmMCRdKCtQ7n2iv4IeC6KIz1KTaeZH1nr/EqYr5zDQ5XNfUKAusge3xHZxYdZ3cdQhW8JGA7Vb/DoxAhakakDLgQRZ5BWIICFUSUwgfweJTICcUuZ+ePufGa7ijMCukFTb1gVpyxWV+PUdgmRSmDFODalqap4qTt3PqbvhL0XaVvsp5MNIPhBvfufZ/59N/Q+qjwD+u46NhJY3PqgP0guinPA8JH5pN0XM5aTuYai0KuFDe3M5YhcOw9SsX0Uq0UPsQI67ZxNG2LNgoTFCoopFdo19Ao1bU4H9dYUnTGkVEM60VAaNBBcNFe8hef/wceHj/id979Pm+99hbj0YSByWl6A6rFGbYuOgv+zsBSym69H2LOzqtwATq8qWNmRceEpppSLjKUlpDmCBmxqHWyajy+xPYkvUWLyIJtE0VwBicsrnUd7uljxk4bCRxOJghtohdylwoajSpDt+fUV1NjJKhHavLV6Czodn2/2es3t3+pGw6OjplMRlzOK86nJTdv3eHy8pT9vW3yoDk5OaVYlawuV3z37Zzx3Q3OLhvuvfMBvZ4jVTXX/vA2Ly/OmV08Y1XNOTo+4/HDI/7H//4fU61mfP7pz7gsLO+/9wmpCARn2d3Z5ejghNNZyebmgOV8BlbSy1Im4wk3fu8OL48u+Ou/mzE9X/LjP/qAzckGt9++y+zyFKENdQuXl1NGozGjyRjfWKbLC54fHnByfkGiM969d4+NrT1+8YufsjEUtLVH6RypJVlqqFYz2rLlxctjbOPZ373O3/z0Z/g05bXbt8lNikoSvM4Yb+6xs7fL0eEB+XjMta0exewMVxeosiFknTivn7I4WjK9XHJ6vMBywPvv3KPX7zHsDWjKJZ7AxcUFUir6gxFZPkZIhXcVwVUIL6O9fOvwbYUrW7y3rJzDJBotU4Z9jVQNVdPGQms00je0bUuZDXDlJVsmYauXE3BY5xCJxiUJ9TCnrOcUywYjU3ojQ1042qrBSEnW7zOcJPRGgu3rE1RrcI1gfGscGTkI1hko3sWo2NCqaEIpNUlPI0SDb2Yx1jjpo0wGrcAdrmjnl8hRhtrv49QCV4HKxyDB2ppgLW6xRGrBeNyn1+vhvGK+LKnqwLJqkbKh3wOtE4SE2gmE9eSpI00zhqMRvdEAk6VRkW3yCC3q6Bjb+hbrHW3bsCgLjNEoqTBJgnOWuo56DYmKVizERhpjZWOMbHTIjZND3dY418ZrwXrh373WaFB3ugydky90aYisi3JceoduD8OapdXpL4IXXa6LxXfLXKs8OsSmijS45QZnZ8/Zu606PD0WnCA8rVsxX51QVgvyvIfuTBLjctpibYu1sfiErnt+OzANIqyy1rNoKXntThRVPn32Wbc/8ASvoqZjPZ8F1zUYeQXvh3D1f6mtZJgpdjYTTCZR3rGZJUzxKG3QiY5FmYBro3WMd/HrhBCfB52YLg44NkXfOSvEMLCY0+K8jy7AISAVaBXL5dPZU07+9oQvvvkVH7z5HW7fvM2gNyAzCa5aYstFZJT59W5tXbC7iftql7R2TO4OHjIgcDhb09ZFt0hXVxNPDJALIGUX7dDlJhlIUom1scEEbwkuBsaFEEXUwnUCXwReRfLEeonv8RGeDLYjVnA1tdDtWoJRBN3N2b+h2OU3Z4slBiSslisupy2jyS7vvfceUiwQMvDwm69Zlg2XyyXa5KR9SeNXNEHy1lv3uDx7yjIzoD11dcLJsyVB5Rw8O0Nazc54hG1n9Lb3SZ3lxrWd6JtjBEcvXvDs6SHns4K3336Nu3f2aYoWKTQbGzmVrbnxxiYf1O8xm9XsbuTc//Wn7OxvcHFyzPnlAtcEemmKrSvEcEg6GnD0suDnnz5lczJh8/YG0ij6omXo51yb7PB3vzzld37/Y1arBeNRisFStJLNndu88Z0bCJ1xt1xgcbx+bQ/XNgQRqIXjxckJd25dhxDI8ozDw0MOD57EAB+bkYgB/cmEi9kUGwLjjU3Mi1NGvSHbk00kkGlFUS5onSPJTNQTGEUbGlxjaduS4OJCPHM1Wig0nra1NLal9TDUJmZGhID1lrqpEc7jlaAuS2y+SeFqdFuwO8ro9xQuKKqmobGWpZTM5hcgPJlOSRJFtWq5OKzY3O2hc8VoU7K5lzDoDUhkgh5nXU4FnX23ASFxtsI1Ko7WEqSLy+jAlGDP0DpFJntInSPI8EWJfmMX6SqCbWhX5xT1Ofl4myALvI0UYm8bipXDeh8X9UHQVJamaYFAqgVKGnKtyFONTyVVE0/XgzxhczxkOB6RD4dkWRaV2Aisa/GuwbUtVeuomliohLXUZXzY814frRJ0P43TiIhQltGKtm1ZLhZooxn0RxitCXiatuH09JTWN3T+xMRVQCfCY+01toaJusp6tYOIU8orivF6AooTiXO+06EE6HJLghRoumTCrsDZ5QZnFwfs3onuCoToP7Ym0foQWCxPWZQXjAYbcYpYa29Cdyr2/op0sC45r5yiu4JKtLiR0pDlio9/50+4nB4wW0yJrptXm5pIkWVtqRizaNYcu7UC6MN3Num1FXlP06qEM3LaVtOXLSZLMUkPqSTWt7RlhQSaJiawaqExOiPLctI0/VbDB+8cra1p2wrblDSUONeyHiGVkt1EBFVd8OmLX/P1i29488Y93n/7HW7t32SQ9TBJhqkHuHqFaypwEYpd76G4Ur2/Ii1LoVA6xegMJbL4fLsyMr46B4h1Uqt3cdm+1iYpJWOwnyFOLc50tj50gtVukrYtQXoCnSXM1a5vTeqQr66ycB1bTMednIqTuG+i1ui32lxCgKJsuHPrLn/0xjscvHhBrlqQmhfPn7Eol2RqxGa+zc2bNyjKBb/8/Et++KN/wnz2FQfPzsn6+yzKJb1kxGeffU7a2yETCbs3dlG64fjoJXtbPX76qy/YGk8YZGN6eUIqFFJoTl+e8sMffMx42ONkdcB8meAKicwTdrdHZKnkne+9h7UF440bfPH5l+AtwuRUZcFIKHa2NljNF4i0YTpf0jq4vrvBzrDH2dkZt7dzBpnnyeMH9Ac32Nm5zmB3ggS+/PRTKmvZ3L7GYJBxfrlgenHCO2/dJjWeo9Mlh08O0GnOZGNCXTdMdrYplkuC7yHZYmtrC4dktZxxenJKW3o2Nq+RmgU/+MF3UBp6eUq9Klgua1orGYwHbOucz+/f5+Iy5kv4ELoHw9JaR+tTBmmPEAS1c9RtTSIdtVVo2siisR7bRLsJoSRB9yiTPnV5yUaH7bcu4Dw4L5gHz/GyxDeW7WsD2lqwWlkWJw39nkFJwWhXce3WkOG4T2gkykQYSelOxS5MxNTbhhBMd9oG4SwihIiP6xzRXoPa4wtLcCv8SCD6OUErsBK3cKjeNr6uOD18Qm+YolONlDk4sNWcqq6prWO6rLicLihXLc47BqlhNOjR76UkJsUHaBpFEIJBv8d4PKY36JHlKcYk+E4gl0qFUBphBM4qLqsuU6eu8c4jpKS1np29axhjKMuiEz0qnPdIadjZ2SPv9RiNNgDBannJdHZJksQI67ZtO3KCQKDjYTEC4t86rcelcEcBgtDZdKgolYgaC3H1nEaEbc0xiiCKDHGpTJA4oSjnPWazI67dFQgdjSU74/+uQUUcvq4XTBfHbE9udHYvsaFdkRLEq6nr2/EDrxI6r7bGQEApxWi4zRtv/i6/+MWfReKBoIsN6JrqOl0r+Cs5yHovEwLc3ssZ9d/ABcdFCZXN0ZVjaEpk2qeXb5GZBItjkZwjwhk4H1X/BDSgpaZnYmKp0AkgcdbiXIO1FWVTUKkEwoomVAQCWutY6LtJ3DvHqi744vAbTssL7p7e5s7+ba6NtximOUk+IUk9rinwTUFwTTeCdfYxPkKW2uQk6RCT9LpnRyOUJsTTGQhiAJzo4tI6uLFjEKBkwGiH0Q6rJCgNwXY9I9Bx0WOKaBND85SOjdx31vsirGnv634fpy3ZGXP6psEXpxSnD3DNtd9uc0FAbWuKcoY9O2d3c4JwkulsRZqkXNu7SV00fPe9Ifu3dvjz//WnpGZMpmNntd4zGPZZVnNM3ud0uSKzhj/6vd/DtZ7nzx4xSA3nl1MOjma8+VrCeKzJhhm8jHkSQUmyXsLF9IJvnjwmyA3OZguGGyOu37pJ6x3/4S/+I3/049/jjfff5ejkkM8//4JWpKwWBddHt7k8OuXL+5/T2xoz3t3jj3/8fXbGPfAlYlHx4KsvODs94tnJku/+/vcwmSIoRV013H/whI3xiNf2bzM/ecnzpy94/eY1XFPjsxzvFD/+vT+mKBYUrqRoa/Y29mGuaG3J7rVN+uOE4KGvEiZ9w7RsOD89I8sUW7sj6mKF8Jav7n/Ou++/D6HFtzXexuyW2nl6mSTVnTmeMoDCmAykpm4rFmXBqlyQuxQhA1ke41S1hEGadE+qo04GVE2Nqks2hhltK2jaaLNSBcuZbTFa4ypLa1tcSGiXllRJUqPZ2M3Y2+/Tn/RQOnqwad3rGFIuFmLbELwm+BgsJYON+RRYVJaCyRBK4/0Kv5ghdzeR2kTsu23wVY3s5aiNAULCUO2TDAe01TTixN7DomJ2cEbrAz7oaLliWxIlUEYzzDSTcUZvMCIgaZuGTEmU0aR9Q5ZpklSTZQYlFdZF/F3LDrkhELTGpGpdI6Ezm/Su4eXLA0QnNJxMxkxGm5HG3J0qk9R0DKUGYzLSJCcxCS7NcdZ1dOrYMEToWFYdzLWOH7tqM4HYWLzAC4HQNoJuQV5Bj0j/LT3Cehnc7UIIzM8FVTFj745Bmuh0G7yM+TjdtBADqjzOt0wXhxT1m4SwFyeVjqm3tr4J7pV+4u8FgsH/5vcxZlfw5puf8PTJF5yePY/vXYirfJn4l9eOAgHf7Z1EiDuKo/67zFMP9ZzjRFDgyXKL8UNM2qffn2BUihOg04wQwNmTqM1zFm9bcC0ieLTISGQS9w5a40OC8xlJlSJFB3F2GUrGGITQHX1eEDxYXxKCY9GUPD5+zsninElvyM54gzdvvsGtvdfIlaJdTuPSvy3Btp2g1hG8Bpmj0mE8mHXOB3INh7GGxNaRctG6yQfRBc8GRGjReLSO5SD4rmB3NHoCV6mkcccfl/1C0cGm8S5BWeJX7ODI4LGuhnpBU57SFsfYZgr8lptL3TQcn5aMexO8L3g8PeT6jTucn15SVwuOzwt+55Mf8OXnv+D6/oSPvvs9kmSHtphRVkMG4wGNrVitlvzkP3+K9Rlbm7tsbw9RCA4PvyLZ2uTwxRHTs5rjo2N6pqZuBqSDAR99/CFvvPMeTVvQWsPO7ndovCUfDgHHw68f8uv7D5FtxqJY8fzJr3n46DGX04b+5g2C0AzHEy4PDvj6y0dM6xW/+6Mf8OG7H3Jydsn9z3/Bs0ePKadTrm0PefPD3yUfDTk8fM6mEBw+f8nFZck7b76PEobhqIfkKWeHhzw+PORHf/hjbt3aZ2c8jJx/kZAPEpaLC4RWpHnKV7/+Fd/54HWU6yajjU0ODy55cXKCdRW/+8PfIetnrFYl81XJ5fKStqwp65bh1uTq4KpUwGiNkp2iXQm0SUhMQqssOk3QNkEgsc7TtCE6ACfdqcmB0Dlt0qOu5uxnUY3eOkfT1BAsJ41HjlOMtKhBjm0kdhUXj/kw0s6v3x6R9xN0kmFMhlIpUoWYL24DrhUEFCpEt2AtW4R3SBuQvaxrjFE8GQoBMkMIQf3zJ4TKod7aQl3bQiRJpFq6BhkEqRliko7i6QS2eMLFYnkF1yjvGZoEmSiUcPQSzSg3DEdDZBcLQAgoLUnylP5oTNYfxWbtLfg0JgHaOor5iAposXa5XcNWcbHRscFiC2jbhtWyIMtyAEySMRpPKMsVZbGkqSuKcoW1bWef3/mSOYeXAS98FyVwZfHY7Vk8SoguhZLOxLIrxhJciFRa6IxIRASS4iE3ZhAGIZkd1ZxcXPLW+9sYsxZhik4HsSYVdA0ixByWorhkPn+J37vbFdvosqu0xiQJoW7wfMv85Aq+C3/vv6NHGmitGelN3n7rd5ieP8cH4tpYrNltAlyMa/bCdxEBa6AwcN4oTl2gatO4kFYSGSqETBj2RqRpD7RB+EBfDGj7I1xVgoi2PFXlUVpSGY1WCVJ6UmMwSWe+GTxKJREabRpcXWKtQymBURqCxMpoowTQNg1NU1AQ3QeKasXZ4pSsP+b69XeZjHepkyFWKHx5iXA1wntoG3DdXqipQBg8BiXiuObl2sgyTqOyg/C61k9QcTGPkAilSIzEp5ImSKzQ0QDAO+g85OR6Jx/5yBECDfFzSdGZqsqAoCW4GusKvL3AtQu8rxAKVH8DpPrtNpcsy/neB+9B66hryf6tCUnPsC0GnJ9NMavArddvslwd4rE4YZnPp7z/1h73P/+S80sHUvPgm4ecX5bc3H+N77z1Nkpazk9P2dqeMJsvePL4GYkYkOiGP/83/xKV7/GP/sk/om4W3Lyxy6e//jWv3btL2vcMe5t89stfsrm1weXRcybS8/r79/BNxfMHJzz4/An9rTuozDEaTJAqCvDS3oTzg3Me3n+CaBQPnx/w4z/6Q+7e/oD/8O/+kuH1m7z+4XdoipLQthy/OODy5JJ333mdyXbCsp2yWMan4cWLl6BS2hZOTp/TSzcZbm7RMzu8eHHI6cFz9q5fY3M84q033mBnPOD84IiLsznzRcV0vsA2nrp2PH36gtu3r9O2nsHWNlvX91FOYr3Fq0hBLJsVQkKi+ygR1ddGS1IJWjqMkPh+Tq66GzR46qImJAElNWmao4UijG+wNJDVLZuDEUmvhxGSgZDM5jNIarRwOOsRDpRMCAIUnp07Gbu3hiTaoIUiSXK0TAFNW1W4VkbNhbCo4FBSoPMEhY76gUThqhb3coaXFtFPULZbKPtACBJxe4jc3kRmPQguMnCamtWzY3p3NpBSUC8ucacLHv7dfU7mFW1QVEVJsyijM61yKB1x9sQkZIkmHyQI0ih2E5DmGcPNEVlvgvN1JFDYGCjVQsSckx7a9EjnXbjXulB+6yVlhHWauqauK6SIOHnW6+OCYzlfsFzOaNsK51qctx18tG5VxKUqXZ589wreorVHCo/3mjYElO+cgzuPDhleZbcrJa8Qk5j1EfWVAcViYVlcFLzx1pDEONZRnjEEKmJPHWs4zjsy7ma8rbicPcO6Twgh4LxHdZG9SmmUdgT7969HRMT+/vL3KtZZRkLsO+/8gOdPP+P5wZeRPBUU3gucXJOUifeYUgTraGwLAeZVQUuNEDEWQcroDu29p/UNqXBokXVNx9FPhrhhEw8uahVX2N5T1xatim7n0EOYjNTo7pkSZImlMilNqhEyxv8KIePEpg3KeBJjqJuaqiwp6xVt25AkGnpZt7DXSJ1hcgOriuVsSbANieoIBToy4kJVEbwgMRlBG7z0kWofr1y8bkqjdYwu9ldoYwevCUeiABNjDaSUOCnB6ji9iKhdkSL2RCVj4JvRSYw09iukXRDaJQ0LgltBiMQGgYv5L6HFvqI4/jdf/6CY42CbKAgDRJLiQuDk8pJEgm1XrMpzRApFXTObLlguVnz43ms8e3TCKgy4fnefbLLg99+9wdBo7r2+zWJ1hg+esqp4/PglJy+n3Nif0M8MT795wtsf7DOfLlBas2wKkjQnCZ7cwJdffckvf/WA7/3gE+688RYXB0d88dnfsbe7zTvvvslke5Ns3KOXSl67tkUvVRSpYWd3m+HJCdd2tjk5O+H6jZu8PDri+uYIBjkvzk/YevgAHxy3b73JL37yc3ZGOzTKcnR6ysvnB2yOdglV4OJ8iRykTDZ6HL444ec/PeDNt+8ybZ/z4vmSxfkL7t69CSKQGMPl5QlfPX3Gy2XBzvmSpnJRf1MGjo8v2N4YY5IMkYIQkqPzM/I8oyxqBBIjNakSaBEIrsX7gBImCvkaS922NGVJW1UYnaBEVIaLoOnlfYyRoBLK/pB2dcZEe3QiSFJBmhqUgIuVwzeWtrboTOIaRX+owVh29lN2b42QjSasQO8PEUjatsaWLQGFEBbZ1qgkLkCVkEidEg0tdMS/PahRD5XJaJu/2YfKEhJJ+sltRJog9Dq/RURcX0KyJ3E2hijZ2YqHP7nPg4MLvEhAKLyNOwOlFEmWkec5k0mP0WRCv5+QpJEhtFb3Syy+WeATgNgQvJSotI/KNW1wMeZVukjDfOXT0kFX4dWOAa5yXZx3NK1jVZacnZ3GB9kkSKFjKJYQ8fOGEJelMnQiQkFRR3ffYTLn5mbJjY3A9gZUPuXB4SbTqo/3kRjhCYRuXxG/rbV1x/rBjZbr0ht0Ebjx9gYmi/63620OuKiF6Py91hYg3fmYEBzL1QWrYsFwvI11r06uSmqM8ghvozDyyvfs7zPH1tdo/d+xsA1574M/5nx+itQ9dHd9WtvQFEvatmSzZxhIwaysOVlGeu+ymFP7lsxoQtLFDOuExEiE0HjnQJQoIfG0aKMxSUqWZxgTSQ1SJSRpjOOWKjYArVKUSOLEJzp6t5AIlaC/RSsIxARJIyVKxWwlARS+oKoLnFMYo+K9AdH12Qusl8yrltV8gXQBYzR5quilkjQ00ZfYtyQh7dT3JmYHCQEd69CFNhLGgiK0ooMPBVJGoa82AqlirEijaryNfnGCgJA60upNSpqlpFkfYzKkFNhqTjN/hl2scHWLbwqwS0JoEMJCiL+CSPh7p5/fRnMJ3sc8ARRCKKazFf3csLt7m2p6wjCD6dmSPOnT7w+4uFjSNo7ZomQ0GaNqSU84fvyDH7CzMWE2fcbx0SPKOqBkwuHRKYPJNh9+mPGrX/yck5ePuX77Pb588jUbt7f53nc+wlYVb929zWjQY7FYsLt3m08+Trg8OWO1MQRjmJ+umJ2d06xmPHxxyMe//0fcuX6T2clLVpdHXL9xne9+8h77d7a5/9mnnE4v+T99/2OKRYFJMt59+32mZyc8/PIJrRQ0ZYaWOYUFKVMuTguubd8kIeGvf/pzHjw55q3vvs355SXz80uKiwV/efATThcl23t32Oj3+PqrB6gk4ebuBi8PLRdLyfd/+Ef0h9vMLwqs9TTOM19NMXnKydE5d67fpl21vHzxku9897s8efY1WkuGKqGfKDT2iimkdQDX0voWZ0EJFZf2IaC0wEhNkqb0shSUoE4mMUujKsnzWJKksGBjc1rUBa71rE5rJjdi8wjeMblm2L45xMgUe25J9jOEllSrmraG0LakmUAa0L000n5rD04geim+trj5goAFpdEbA8KqxQx6YBQkGmEMIPA4aFqa6Qo9SECLeApM+nGasBUHnx/xq/vPCFrR72UkMkUOPE5bEmPI+0MGg5zRKGc07pHlSWSxiQRrW1zbIE1kstnWEoTAek9QKToZoLMxsq5YXD7E1mfUZRmBmSsG1ysrj7Up5fr/r1k+hPherI20435vSLAVrnYdfh73Nm1oOV+WnF6uODu7pChL7m55RsC2Aquhb1a8OVlwXvd5NrvGvNHUVUBqhTECoSDNdCdcjdCZVgYC6DAmm6wQedGJMOM0JAHhBV6Gbj8cIbjQUXTXMpO2LVktz3H2Jl4obIjLeSljEQQRkyJ5NaH8l/uX9Ss29ch+unnzHh998N8jTUqWj+PJnUBZFxSrS7Rd4Mo5ej6lOD/GlYZ8dgGZwnrwbYsWhkyrGPedJCit8YAKjoBFKNCJJg29aPhoeuT5AKN7CKUxypCaBK2jM4QXNbgK6xp8sJEhtoYJQ4hqQqlQKkUhkCLSyrXV0bTUWpq2jjqoLtveB4H1AuehcQFnPdI21LWkSGLMQ6KbaD9jFIlRMaxOR/saKUKcFtfMMSG7hiExaU6ejpGiRaoRQqUI4bB1ia0LvCsRQmGSnLTXJ0l7V/otZMAHhx6MycY7uPptbDXDFhe46gTfzAh2BX6JDBUy2USq/LfbXLz3FMUKo1Ma55nOVty6dg+FplESY4Zc29nD+gGXl1O2Nm/w+p0ttHFsX7tOb75i0tfoMGN5OefFi0MQCe+88yaX0zO2Nj6iWS34D//6z/CLijCeMNy9w16es5wGVosFWq24OA9cXCYMh0PGoxR1PWV+uuDBr+8zyKMfFsowu2iplp7F5RQRCgY9yc9/8Su2r+3S+gp8Regogk+ffs3uxoSzs0uu7W5yY/ce25sf8dPP7jPZHPLe+/f4u5/+go3emK1xj2He46v7T/ny0QuKNpBnGc8efsOHb7/BdHBJIzSzX3zG6zevcXOjh3AV56sFUm2xfeN1dvbvcn1/kxcHp9T1MuokqoL+OGWxWlIWK0I74eTyhH4/BqtNNrYw5ilpiJTe0Mq49NMmJusFi5SKLM0Y9Ic438ZMDR9xdaU9ra2xNsf2x7h2RRo8aT7EJHmXuNiyKhpqB6EIsFRINKanyAaGvVt9esMBxhvUSEFqWMxWeFshaNBGoswAnQ2hDSANahxvRB+isEukIHs9VG+M0JqgFK5YEqoWNR4gIPLypcKHQDmv6OdRsxBP1X3wnulXD/jp33xG1Vp6SY6UADVGO/rjHv3BkLzXI08Vw2FO1k9RSiOUiqf0zkxQKAVS0bqIe1vb4kOCMAVpW4Lz2MZG362rE9uaKdUxbET3R2sdQve8RHV5XEQrIXDBsixnkSLb2KgxqltmyyXzsuLw9ILFqomFR8K0l/DyomV/HJEubx1nZ57nZy2nq4rTZkhRCZRR9PoJ/V5K3jOMRxnWtoAkDY4k19h6he6XOB9FeKwt9teYfujYaYKracyHSI0OIhB8S1meY9uWRCZ4L5EyoLqCjIgK/fXu6b98fbvJBB9oQ4sMgTTJePPed/BBkPdHJCae2n0I2LbF24qmLlmWK+7M53zxq0/55skDkoHG7u4gjAIjccJjXYPpSHUCEwkPIkYQZCo6IafpgDQdk2cjEpNFV2YfiR/O1eDqiAL4CueWIB3SJLHpuo5EA3gkQudoaRAywQRHYluC9bSixYUQnRsQHSNMYEyKUhkQPQmViJYqocvCKb3vmAsCJUS3J4nsMCE6ZqUgfiwR9syHI7Z3xggsCoWXI7TWKC3I8i2Cawi+6r6HFGlE16BS1iapQliQDcIkKLNB2h/hN6/h7Rvg2uhO7mukr6I413zzG/WMf5C32PTyApAk/QllZcmSFONbxrsb1L5mOj9mWcw5fHHKVw9OIQj+6T//E27cfA3lW14cvaSpPGcXU7Z2r1M3UciTJpI8y/j6V38dcxVIOL88ZOK2+OCDe9SzFudqZrOS2XTJk8Nj3nv/LTYHOSeHzxn1Jmxv3ebJg8/I8wG333qN85NnTJqMjVGf84szPvv5T9m/cRfrHctVSVM27F+7Sete8vDBN3y5rLh37w1GozFVO+X0fMrBk4eMJ30W5ZLL6QWXJwdsfe895m3gP/3nX/Dk6BiTZRw/e0heDzgYKCZbO2z0N/jRD3+AHvSQWcLy5JKL00t2rt3ChhXDXk5wsJy3pMkGs+aIcS/ju+99wP7eHqmH5XzKZblkMhli64Ivv/gcpTS9/gisJ7SOtmlpmkDrBIlJGPfHDMfDjp3kKYsFdbG8Cm+q6oaQZ2ilaC8L9iYj+uMhJsuQRhKkic62jcUuPaGRuMrRGwgm2z3yXobRKb6R1ElDOTtF6Ljoy9IRaTJEBIUIEmEEIVi8bwjLBvtgQf1yhZkY1CTF7FtkPzLnQhuwixZfTzHbk8jJj6wDRje3vl2hILSsnj3mJ3/211zMl6R5SiIhtDVrr6X+aMBgsNHBIAKZKqy3WN9AayBImqpBSIewAR8aGhubS2stngznJcYsMUqCaJDKXyE+34ac10aN/+UORnTE3lfmix3lNsQkwtZZzpdLnh2ccTZdUjZttH5XJhYc77hcOc41rCqB1IFEGYR1nJ97jpcFL6qGoo2n9mSmyRLDxkaf+aJGKEeaJYxGGdWqpp+C6gwKxXoqkXQMpHDFGI60ZdCe+HdEZ2bioGlWtE0DyaBbChMXMwSU0gQdm8s6MO2/NrkIITraq8B7h9aa4WCC95Ik75EkCVp1FvJEbzDnWja95Vrb8vLZC+Q4Q2aQJYGgY4NMjaSXRhdroSJM6BAxb8sFVJBIbTCmh9EZRsV8qNBBrl51Ops2pmuqEEi1gbSPD7Kz9IrwkkNgg0SIGIOtVdKx8TRKLqjbGqMThIwQHEJiTEKe9UjTPkqZzgpJIHVcpId4ueM1D3HviIuiValUdIhQMezLugovBIPhiNHOLoPMkYmAV7sI2T1/mDjxaAMii0aXUnc/Z9/Rk6MFTzxYRNp6vF9bwCPV2vFSgcjj5FZX3c/8v/36jZuLFApjTDxtFSUahW8bpLY8evQckoye7rEsz3j49ITLouWrB4/48T/5MRvjAThJYxt2Nrf56oun7Fzbp5UFJlMkXiBTx4uLcx6eHFP6Hu9/8C7SWZbTUwbJAJUIxlvXGYxaHh6d07rAs28e8Hd/87doevzwk2ii9+LhM7bGm7hmwbVrE7LMcP/XT2jbPgeHR5j8C27cuEGepHxz/hXLywWip8iGW5CN+LuffkaxnMc4VSOZzwuk6KHUkC++ecDp4RHSpHz54BFZNqIsF3z5xSEjeYNbt/b5xS+/4Pd+/3uM9jJmS0APOVk2YPqMRhOsbxEqYEPDjRs77G1v8uRxwv6d65yfHaKCpW09y8sFP7v/K1Ij+JP/7h+zfe0GxXLO3bdugG1pioLlsqStWySa3iBh2M/J+oogPW3bYDoTbiEVTR2/LqMt6sYivWVrayMKyXRXIAW0okGbgDdQeo9JYGNT0e+n4GA1K6nKFVKUZCZydRM9Qvs0Fl0jsVVJmNX4RRMt9SuBnGQkk02ary8g1fgXM+yiwdwYYrY3mH72IlIpxzPMXk5+exIhCm8RIlKpcTXFs2/4q//fn/Pw2RFaBrTSkQatdYxe7vcZDIekWYbUGq8SaucihhxiKp9WBmEibu6AtqlprcdZSfBpPLW7FutadKowiUQJkF2U7bfZT+vSecWg7RbL63TJyCqLRck6R9vWTJcFp/M5zw7Pmc0aWmtBxgCuLMviBOUdrXNoJdAqCt6UifqbshU8O4eZ9QhVI0WCI7AsG1Z1S6oji3Bzq8diukBYydvvpwQZQ9F8iNYw0X8qLofF+r10/4gZhGs7FtWd5n2EDWU8ba8zQqSISYdKqatr819OMFL+/YKkZJyUpJKgBdbGfEQX6KAmhVYarwIypCjv0ImLOiQRsNoj/ApCEjVVMlrXCKXQaJzwtLbB2ZbWWpyzSC2gtYSwwtsGSkFnyYbQAqlBe4/3DQBGpx31OBIjCETYK0DoJpaoK3WdM7HESEnuLMjIpIsGrbExJGlGmkcbHYG9giVxDrte2gv1rV2PQut1SmxOr9dHaoVQgTzXDEZD8lSjVyegtqP2TcfmEnzMQJIqvWLAXUG5kV/fHX8c6xjSIHXkMTsFoYG1t5ivYjMCgvqvT6b/tddvDouFgPWS1kvmq4KPP/kE0zfM5wsqGxiPxzS+4fR4Sr3UNJ3BVdUs+KuffMadO68zGPSZTi+48+YbTLY2CAuJkp4XB0949mLBuDcmMym///t/yI//+Hc4Oz/l/OiYl0cnyONt7t3b4ez4JXdv73Nzf4eZMtx+q+I//sWfM18u2N3ZpSpWfH3/a7Z2x9zc3UcZg/WeYrGknJ8yffmS3X/2PzLc2iQbjtiYbDBdnJEOPYkKrGZz7n/2Ja0X7N3YYe9Ww40bQ37vR3/A3duv82f/4l/x9OVzBv1N0jTh4vKcsqz56a+fE2SPwWDAf/7JL9m5cZ0bd95kZ2eTXv4hWZIghOTs5BKdas4vZ2xt7KFNC6rtcPMBUqT0Bht8880hZSPpyZatvT3Gt3b4xc/+lmt37tE2BW1Zs9FYbFPircMYTZJJlAxY1+AEpIOM3nAEQVKuLmmbllW2weJiybCnMRqCL/EtoHSnUC7QiSC9ntEWls2bE/obOcEmzC6WCNmSGo12CdpFryltcpROUWkCraN5tojFcmNAMDLm2uSGYC0qjDFbfdRkQOY6tlOWsvXj15h+fox1AZMmON/d88HELPOmpHh8n7/6l3/Blw+fIURgOB4xGm/QHwwjjGAEw/Em+XAMiOgD5i2+rRF4jE4jNJH38G0DIsGjo4EmFiFiJgduzXKL34Nw0efKu06ZvFY3droTH7oHbv1n8TB9pVoPXWMpqprz2ZLnL895eTalrGsIolvaRsijLEpcsHEKU4JBFtgYw6Av0Zmn34vssGUDrZcYBE0VkKol0RmD4Qa+WdHUBSeHlixXDPsGnSq88xGrlxYZVOeQHNk/kq5hrLNfAlcJlBDwPjo7O9e+omIT8M5GqDFyXK+YSutEzm/HCYjuhC5EB791PmvRMcDT1EVkJZFghCcIH6OkpY4FUrYIIRju71KypKXbF3iFchbZVnSqH1pnaZuauqlpXNNFOktkteqco33nfwa6i2+OmfRxSpDSdzuojtTQ0bulMmjncT4ANt53gkhcMQkmDLDeIWS0+6HLV5FSxOkl75PlPWpqRGiv7p0QumA3KeM+R2uMWTeVHnm/T683IM0STKoxaUD5AJePEOkmXvfjJOIcSZIhRbS08dIhQ9d86TJnwprWbYmU5BAbSefAIIQCkeJDtZ5rI+OMVw3nt9pcgJgoaDLee+cG+zsjpHaobIvRuMfW5DoHL5/w8OvH1IXi8uySYhF4+NVLinLJo+c/40c/+Ahf1og8Y7o4p1gVzHXNNw+f8fK4QdVL9rZ3ee+9e8wvT3hw/yvKqqWsag4Pj0gSSSIzPnj7O2xv9Dg8OODx0QHTJbRVzbUb2yRmRVUJqtawsXWTg8MnFMWM6dkRqq1pS8G//p//Ndu3bvDh+z/g+TfPGA9GGNXnP//NpxgUeM3J8RmDjTGXZ0e8zAUffPIxf/vkPkcXZ8g8oz/uI70nyxKkNhR1zfnJGT/8+APuP3xJNWqQweJDBQRa22BMiklSdJLiEbw4eMH1a5vUtgQJWa+PVILp8ozz6Tk6MYwSRX9jzOl5hH1UmhCkQOoEbEO1rGnKSKeUSqOTBGElznuC0iRJQggSqTOKKmPpE5q6ZjIONE2Bsw1KK5IkjTenB5V4TCLYfnub3nhM07YEd4GUgcTorpElsFKIREWbCA1CeoKWiA2FTFLMOEclBiXjwpdE0n9z0InwLDJLo77D1YieYvzxNYJ12FpCyGLAkwBb1iwfPeDv/uwvOTg+YWPcj+FNkw36oxFKJ7RtRRABnaWYrI8HbGlp64IQLP3eiP5wRJr1SPMhi4uX1FURBXye7rQIV/HBOCSxQLZ1PGUG/2piWfv7hTV01P2BDxA6vygVQky59IH5suDw7JyzacH5xZyqbhCISJVWuqPngvMNwXdgVfcMSylQOjDswXgg0XKtTwlYF/+SSQy9vEe1WLGxqWmspGoC1bIhSQzLeU1voJkVS4oqGnk6766Kq9SSLE8YDjOEttHg8kq74mmdo6zmtPUqMhRlwDs6q1wBWnY+Yh0pIIQrenYs4h1TjDjdKRH/rnMeIzSZSSjrhqqu0WENJgqMVMgQ4o5BxMloXlfUocX5QJAKZdYJndFHy/qGqqlYVXOKYh73T9J0z18edUS+wrmK4H2EtjpigtJJXKSL0FGxiUVXSJTsYKKu6XofDUdVZ34pg0BL1RmCiivxZejID2maMhxNKMst5tTYpogHluAjo7KDvowxmDQhz3v0ez3SLMWkGVkvJ8sj8UBJaI++QesxMrkGMkF6j20bgqshTVA6IeZ3xHsqqBgmJ0KMIIivAL6N971vOkaYYh0dIaQhhLbDfxMCzbcg1N9Sc9Fa8d3vvMdgMGIy6dNUS4rWMp1dsDHe4NpOjm0VD77+Eu+2ScYJ4/GQ0WDCnVs3+NnnX3F60XDv2nUKW7OcXlA3Bb98fMRy4XnjzTegWrG4XLKxNebpg695+fwlpYuwRa9ccHH0jO9+8jvcvjFhOT2NE0lV0esbNgdDXr93h42tMcdHJ3jrWM1LtBf8zkcf8UApvvzsC56/vECfLXhxtuLBFwcMjSbvC77/wcdczEsuT47Yu3mNo/MpZ8fHPBae6fEZtmo5ePyUVVNzUZUUTUUmNd61TIZ9pPXs7444fvGYclnwi5+d0h9nKBUY9idIoSibFTpRFGWBVIaqcVRNzf6N61jXEGjwQlAsHacHZyxCxSffe4tAwBCV0bZpu5uxxYUl3paR1muitb3SeaQ9exPFfwJsUyCFB51Rzy2+daTaX3keWQuIJhYZ63BF1IdkmwOqcoFRNVLU9LJxFG9qh8o9etiP5SKAWxSozRHSSMx2D2wgiBbQeN8gpUagcE0BpUNNhlFPsiqRSRJ/aYXQOaa3vi0Dvlxy8enn/Pwv/4qL5ZK93W0mozGDwYBsOEJnPZyHslzS2thInF3igsO2S4JvuodWdJTKkrooCT5aejjn8M7FLA+tQSqssLg2YH03xRAjgV1nR3J1av8WvZb1CXydrIjAeUvrauaLFeezFWfTBYfHc2wbG1WE6HSEnFw0C/UEULEYuwCzSnB2CXd2A4kUJMax2Q8YIShcIM8MW5tD5rMCJRM2tja5PD+krqFxNXXd0tY1Gsnudh8bKgrr0IkC4WKDEwKdaPp5hkYwmBgsr/JHApFkUddLmqaMENO3rF6CAOXFFR3bfauZ+PX+RetuEvXx9x1dOjaj2OCGvZyybmhti+0KfhAOqWVHnog06cW0ohYNUkX/Ou8CEof3DUoIrKupqwWL4pJlsYSgSROFkhIVomNyIGpiatdQ15YgQqTp6ujSsDYcDR2JQ8mYaCml6HYancFjN7nE44DHuQYXHFIqVuUC21YROiQezEbDIcFfI00E5XJK00YCgZICbQymi1NO0pQ8y0jTBGMMSRpF0lrFvZmvVsjQQrKPkGln1OlRosaHFh1UJLkYE/egtB3k1cVeS9c1Fhshs2/lzfjQAFFo7PEIX3d+Yj4u334zJvI/wLhSK5r6kk+ffMP3P/6Q86Nn5KMxN/d3YsZ5c85Xn/5n5quSZTvlxmCPja0ebVtjBnuMxyOOTi756J3b9PyKrz5/yOz8gsfPn/P08BJUSmprXnvnbcqmQpucy9mCuoWzsxPynuLybMGH3/0uZ2fH/Om/+Jf4fIdbN28Q6oZJPmJVnrN/bRMjIrb9/OsH1D7iq0+ePaERgdI6mukqLra1Yjpdsm0HfPngc4RuqOqKqq756PvvcXTwksXKcrqcgzlgPp/j0GTZBpeXp4z6GcEFilXB3pbh9dev86tP7zPaucFk0Oev//3f8NY7x/zBH/4B6DFNVfPs2XOEVGxsTdicDNEi4cXRATdv3mQ86HF6fMzmZIu2dVwf93n/ex9x+uIAoXr4AG3bIkKFoiZ4S5YPSHtxpJZK4oPFO4eUBqm6icFJhHIEmyCixweJAa0i5m2DZG0p0U8M9awg9Ey88UWD7kmMyBC1QBiBSvJ4WpXdzdiCbSzatkjlIp4sDcLETAi/qAkqwVuLNIZ2WtIbD4CAzNNYm7tAI6EUrmiQucZNTzn+2a/59aefU1rHZHPCxmiTPI1JhIoGTTxxykGPIHoIAd5VNPUyCkCVQQjZWX+U2DXlVCryrEfTBsqypLY1YDE6Q5NFC4zgcATStA8+I4izbmp55a3lfbTcF10VXu8pXAg4BavKcTxbcnRyyaqqaVv7Ld2CiNMDa/1E7Fyio5l6YN7AydQzX0TGn9KCmzuS3Z5nMQ8M0oTXb93jfFwjJVxM59Re07oogLUWFo3jdD5n5/qYN958DZI24u1iTaeOL+HAKBVZcSJCR+FKUSk6ZwNeFZdvQV749SjX4fdrOnYXwbt2NyDILpb3imIXp7QQc3M2xyOKpsX6QHCOoNasu6itCUJQVS2NcKQmHkqEMtjW43xNwOFsQV0vqOoS52IR9d6xbgkyeBAaL1OkaBG4CGGKSEJxzuGCw/mGtmnwrrPyl9F6RYrofr22zledzsiHOIX7EGO3p/0RVbXAuSZOwlrQ62Vovc1wkFNXa6cGh5QBbeIEq2ScaI1WkfmlVJcC6sFVON/gZi8wpo+UKYi4o0KBV51BrBQImcZDnbAE2o4aD8K1YF1HrY5WNKLbtUWCg+/uvijODJ44rYk4iYffsLv8xs1lvljyH//j3/Lk6SHVYsm9127z7KtnvHHvbcrilEffLPjJr1+QbN5g+vyY2cNveO3uTY4PD9jf36A4f85gvMdf/81fc/b4AdNFydeHp5yeLfDWMT3/Gd/9zruM5xZ1OkelgtHWFp/94iusazk8OiHPh/zVX/0MGSSnc8cwFPTynJ1rN3nz5h22twYQAunAsDya8c3XDxhu9CjKBefnlyTZkP5oTDFbMl+WpGmP1kvy/iaf/eo+ja1QQXF6OsMCb735Or/+9AuOphXL1Yy2XZENNhBNy/6tPeazOTZPKJuai4uao4NjaCVff/WUrWv3uHnnA54fHHB8OuejT96gfXnMeDRmPp+zmi0Yj0Z888XXTBen3Lmzj2sDi1WNSiU/+NH3GOYS0hQdLEfPn3ZAdRQiSiFI8l6HYweca7FtjGUG8EHi2uhL5TwRy1X5lQVHmiUI5cBHHB4hcd5hBkM2dM4qmbAslngH/cwgUo9I4gmSucCWFm8WiFQh8xTRT7CtQwWHTiKkBUSDx14e2VJIpDJkg4zgLc7aLmAvjSckoaNdeCgpnh1y8Mv7vDy8ADyj0YD+cEye9cCVNOWC4GsCDTrNSbI+QiY4X1OUMXrYmAyhU4L3NG1N29Q4W5GlCpMOMSZSmF2rWFWWyluEjFCi0gKNQJocoXKKRUVVNUQ2T5xgrowa12yrDlmT3tO6hrK1zOdLLmYrlrVjOivjA+yjPiaErjmJ6KQQxXCxIHvnaBrPvAgsloLLRSDTgA/kCWz0BGoF27sTbtx+jZ0q8PDZY86Ozun1UrwPtK0k+Ci6my8ajk+X7F0f0p+AVLEwIrtwKxEhk/X7u2KRXTEV1ni8fNVErqa2rhHJV2LO9QfG/ctaCBg/RknVUaFDPAnj4gK7K/69LMV6j7WxibsQUOvvEUFlGxrfgFR4D9618VtqI9zqfIW1FikkiVJEgXpL266osbjOar91dff3RFT6axW9vaSA4AiddhZsZ9rpumkGvA84F61XZPf+bedbtp7ULwZTlqsFbVvjfZQKKAl5lpKlBgbDCIMGGynO4tV1E6IjanUlP/ga52pcU2PrGlmXJINJt6uLzgHxCTN42ca5y9k4LQoB0kdquXAgLEjZMdLi14z5SgGcje/dRYjMehtPjwAyijh/6yJK6+Gb55fMLyr+X//T/8Lb777D7Tt3OLn4Nc8e/5zACMcGy/kMX1VUdUFTlqhEQbNCNFOePbwk62le3H/ExaomTPbIxgnGVYyGQ2Q2od/fRIqEJ8+/Jklz0sRwenjEbNmwbCoeHZzz0cqzsbXD6vKYulJYC6+9fofr18b89Gc/4+T8mJfHxzRB8NXXj7h9d5vBcMLTF6dcLEvKBlQLQRhW1YzHTw/IEsP2xg6PnzxDZwNmi5KqqklSzUY/4eDlMSozTEzB1rDPRj/n5PCY0lmkUYzHWxSV5/U7rzOsId+5zXsffMiX9+Hzzx8zm88ZZD2kFLz1xut89eVX/Mef/ZLVvOBkesCb775FWxf88tNf8/q7b3L3zRv0E8nl7JLdyYRRr8/L5ZzgLFJLtASl4wnYBx9tU2SnCpYp3sZgq+j1kCKlIFiDt21kJfV7KNUtEqUidBRS5wVqc5PV4pxq1rCaegY60B9KUA4RNMXZCqYSrQL1WYtMDYwk/dt9sr0MTPy8wVuCBYFFKME6DljpLD6o1iJMdLwWIiU4h13OmH3zgOdfP2K5jDYqo/GYNE/I+0PSROPagLfx6dNJilQgqK+arBAqOjlkPZAa2zqch7opkNQY0ycVUWugRcAo3y1fo89XpGXquJj3UFcL5pfn3Y5mfThfFwERH15BVyS7pL9U0taOqmmpGktdN9B5fXkXvlV/uxyYNZsqRLsWj4DOnbr1krIIrJLIzKqbQOOBIHlxeMl0+hc0tmW5rGhaR1mvGI5SEh0pqb6bkoIN2NKhBpGS6mU8n8qOfhzf0Cvh5FrjIrtlUoS1uiIv14r1bi+x/vCrEW6dsSwIwXVajc5uZv37Tl0uhOiU7kmM6fUCI2I8h4dIcAggiAeWxjVUvsQ3YGjRPmpCuDJnjFCPkRJtTEfEaHGtowkN1sfmYl2L8z7CTipCotpES37rPIESrVw8lDRV1Ea5ENlnfi2YdQTnYyJp0+LbSHhIjGSxKFiWC9qmxLsKIbMuNjh69MkQ0Oi4p7ta7Afi1V+nTkbYNtgG1zS41hKsRHmBby3ChJgouq4B3qOQ0TpGmKjHiZ+4mxBj3QOLCG209gsx1TKIaA4sZUYIJl53XxGC7naM0RXkyrX6v/H6jZvL1s4uf/JP/x/85Z/9W14+v2DrxnWG17fY3B7x6y/u09aOa7eu83/7v3/Mv/1f/5TDJ0+4vrnFWbng6yePSBLNl7/+OflwyMe/83s8OXzJ7u5tslTz9MmXXNu7xnsfvMHdO9fwruQYzacPHrK1M+F8+pJFE9i+cYM2S5ktF+h6Rr2aclSAs54kkfzqV5/y8ukTvvnqIU2HmqaZ4fnzUzbHA9JeSj2fc+POHS4uL5guFtjgKdua1+7coi4LlivHys2QOtB7eoAtW1xVIiQsyxLpWuxqSTkYQdpjmKaEYLFty2Vpqeozdu/c5vW3r3P35oA37/0P/OpnX3L0/Btees/BwQH/6I9/H6k95xeXfPnrJ6TDhIePnvL666/z2r3b9LKUzckGJ8+fcHL6knp1Dec0wbd4WxJjURtCE29E70UMNnKR1RLorNiVQpkUhMH7FtVojJExN0UKTJIidB5PksFjvWZmM6YXpyRtyWvboxjBrAuU1JFuqTzJ9ZS6bamPLOm9PtZaLj6bM3+04Nrv7aBeSxE6ELxFhCTi2U7hlhWzF0vyvZxsM4u7IpcSMNi6xJ2fcvbgGZfTAicThhs90iTmbiSpwCQSrcBZSdOqqDRHdnoa8F6BMKRpilCRYeR8VEcLaRC6jzADdNrHJFnn1xX9lrI0TjiRMt/ibPxzpZuowUgy0rQGZusyelVku2ByvHBIEUkA1nqqsmE2W1AUVYw8Dq9OpbBGm7qsd/8qjdK28aQoZIxAmBaOiwUYJIkJzIoIlwUZqOsyekMBSnqUcjSNo2k0GxsjVssFrQetDVli0KnAqY6ZhEOth66rMLLuvfFfwB9CYNcY/LppxFYRf/+txf3aABMfQK0tUEKkc0vVNZeuySiFIqBVt+vo4C8hJbqbpn3w0eoouAjcyI5UIaJI0Yf49YR38RoH27WhTsmuO8BHxGvmidn1XkikUXFS7J4cITSIBKVAB4W3NV6JmImDw+HiNmptSx8ihdoSYr68iF8zCEljLYtyQdWscL6KCZAo1imRMULg1a94/ToL/A4qDD5GZbgGfKvBddfIS9p2gfYefLz/CbbzmOv2rR3DLRpsxnt9rXERdM/Fuung8USChwshYqTrLKCOco5zOPutafa/8foHLPQNv/P9P+DeG28xn9eMxkO2tkYE4NaN13jw1QMePPiGP/rjPybNFf/uf/7/YhXcv3+fxcjQrGY4KXhxOOWD7+f86Ec/YjQeYuuSVC5Jsx6JkaSZ5vLlBZ98/AHL8xXz5YyXecb5vGUyTDGJ5qvP71OdHlAHi5UJ17fH/Ps//zcEF7g4uaR1Eik8g0wzyre5OLmgrgLHszkkhqZZsNHvUSxWNHWDN/Dg+VNyqZEomrrE6JRvnhxye2eL4caEWXuGrRzjwYjNjQ2mRaQAK+O4c/sG+IJ5WfBHf/KP+du/+xtK19IsTnn9nfe5+9Z1Et3wV//pr1ksVvyrf/Gv2L+zyx/8d7/L9Vs3WRWBr74+5cnT57zz7pvcuPE2F6dnFEXDqgocfvWId955t3tYYqa4D55gLbZjvITgcUF1ylsZixWhM63zCBkfImOiF5Xz8UGPiXYW5wWHlw22nXNzs08vHxGCZzU7Z+EW0ApEFm/Daloj+4rsXk5+vYdMDf2bQ9ppA4nEueiXFrrFphACLz3et5w/m1F/eczr/+ge2XCAQGFnl8wePufk2TFeSXSaopMEicMYiUoUOjVIHWIjkZKg0q4YSJw32MZh2zbuYozBtjVl42i9wKR90nxMkg9jHrgBIVQMhapbEBqTSLxr4/7QiwjXaNCmH63Qc0my8MDxq3oLcSkursiqeG9pfct8vmK+KFiVjqq2OBdYZ8K/kvR3DUbKiOsb1YkQ1/RPaD3MK8HxzGME5Cks28gGFB6IKbgYBZN+j7J1nJ0v4q4gOMaTIcvFEiEllbNkWqKEw/tuD9IlT/o1PXXN6rqCveLWXfiAs220rI+V7GpK8d5dNZU1DTkIFfPog+s+rergxKtZqVuGRxaW6j4u3i+xsUm65g1X6Z4hdPsPFErL6AqhRDdxeYT3BGexwaNC3F9IRRdlnZKkMR6CNby3hjk70kYIMd00Qs8KL7t9mIyQNCKghEaxZoP5TjcSNToRmgKlDRbB5XLKsrykbUckSWBNkQhBdPeDiGmXHWFn7ZYQKfAqcm5aT7ASERTKmE7rcwtbP47WLOkAaRukKBEmJ6h+3FF1k3GMNIg/3rgL853rQgAcrmswPti4g/GxqTnvCZ0Hn/AeieFVyPJvsblE8Y1g7/o+e/vrJVwMofnkhz/kez/4ARdnU0ajAdvXtphsDvkX/5//iaZY8mIFCQ4vwOjA40f3+eSdu2z0DY+ef0WvbVguCuz160CgLBY8enxGsTzm5bMTlvOad964R2Yky9kFZa9H4QM6eHJhSXA8f/yIft5jvqpJB2OOnj9Db4w4nV/QOAjOc1HUjPMcXzrGW9tc28l59vQFs/mUum4ZjFKWMmojesmQpnbUwaNl1ASM84yknxByxVD1MASKcsHJwRGv3dqnaQqePHzE5eWKZwenLC6m3H/4lD/4R3/Mi4MXzC6neAdpknB5XvP4yQtUAh+99TpPn13gG8vXv3rI1tYm1npWzlK4gAN0liCXgsRolIz5G9ZZ2rqOUIoySBmLk3UN3jZIHCFEGAqlkQKS1KBMRuM8fQTOxkCtZel5+uiE77xzDSksdeOQImaS+FUgbICzNUHnJBsS2S3SvbEoozFjjRknVGcltrUoYxCk3c2oCcKhxgPu/GGPtmzivqhuWD55xuy0Yj6vKNqWXMWYYe/a+DDpyCILxIYYo3t7mExGbBgP1sbcCVd31OGCVVVR20DW32A0zGMccWNpqhmhXlLVgbJq8V6SZv1IPPA1WkfGVNtY2rqNMKHxVC2UVc16oy2v7NBfxdQGAhZJ03gu5gXn8xIvI9wS9zJroWJsij54nHcxPljERb73jjwfYG1LU9e0beCshKO5IFWCvIJVGxgMBEMnKW20DlFoQtCkqWIw6nXWzZ5Rr0+iJYvFkrYsyFwUktYdvTbgkd0EKLplrhA+CvFEp6QnwjjWtjRtjcMChisR6bccCtZNxq+niC4fJjYp0TWXWEuUlFEwiejifqNvlvMu7iiEBqK4VykRoagQSExCSDxZlmES1Z3WPd4GrG9oQsBZe/WehIx+Ymk+Is8GmDQncq466rKLS3zWp3Q6F27vuucj0s3jsNrpmBCd9Xy3szLRw80JCD5GHHgf9WxnFy/ZmvRQpiVJIsFEhISwtty52t91TTeAdAHfOnwb73mpNErHWAutDCIZUp6d0q4OMVm0UhKyD1JGsXRw0TtMxFRKgsXT7U6cJ4QmwtbEg4HzPurnmgW2qSN0K2SXvyNRKsUohcN+a6/2W2oukc1SE3zEDF0nShJCXJ06dvd3ERKyYcaf/LP/K/MXT/jL5V/h0xG7o4TL2THPHz1leXHMl7/4KT/80Q9ZLeY8ffyUog2o8ZDJxpDz82MOXh7x7MlDDl+WNE5STFe0ekGSpPTyhK8fH/PB3Rv0jOHk+BQfPIOewsX2jEoHLNqAV4bT01OyNGcyntA2DbU0zL3l+sYAwnWePQu4YEmUZNskTIYJVdGwv7dHfzzgcnrOaDSirSrm5Yr3PvkObl7QliVffDlltljy6f3HfO+jN9G9jOv7N7j48hvOzpYMvOLRw2fUdUU/S1msSspWkJExn1UY7ZlPT/n+J+9TzBacvjTc/8XnjPb2WFaCi8tz3nnrTeazWXz4lEaIFh9a2nZJax1SJlHw5gPWVrRNiez46J4oQtNakRqBSQ1SShYrSy+z3bnBo0SgZzzTyzlyo9cFQcWH29Ue33jIY4FO8zwuPZWLi0Fc1KR4SW9niHee8tySb/aQUl+Z7wUfMHnUK9QvD7l4fEIZUqTKCDIu2aVRnR2+Isn6pL2cICzWR8w8akNMx0CK1u1etggl0IkhuJq6XNC0Bb18xPbWkH4/xbqGZbGIDsiu6kRmAS9yqsbi/QJ8i5aO0Aps3VBVNY0r8GiqJrCYlq+gBum7sKX4fHgJOkBTLbg8vODo5Zxl01DbWOTptDHQZWdIibfrrPtYNMuyRgnB7s4WjY1RFE1bM13CcSrIDAwNeBX1SMNen1sbW5ycn1E0DQ5IVUqeRpZXliukatnd6aONwnvL6XnFm9sDTrxHOhdd9wMRRupU2AiNTFO0yTFZRpLkJMYwyIboPIsQjqADXr618+/cCeIF8Yg1nTkIhOtYdR3cE79ix17qvocI5Ytuqulgz8iNjeLMrnFFGyIViRdKdnBijO6uvKB1liAsXsQMG6ESlOmTmCFZ0idPe0itsN0+w1uL95bGxoOlc9CGBu9bgojZLwiFcN0UJ7pESSGjB8Z61yEEUYDY2eDUDReXCx4fPmUyNGjtkP0RUmfx74V4DYTQcQcjVLxezuFbi7OBEBSqs6qROkVIHWFJEUhHH9JMf0W9eE5/fBOk6QgagBCRvdYRJoJvrxqjFx15xEd6Nz7GcQTTi++/bmnbAmwTNWAyumBYZeIu9bfdXOqq4Gc/+XPu3vkO27t7kamhYmDVVeSuiBcZgLbFpBP+6J/8n7m+v0t59pyDg2csz1b0xiMUmj/7X/6cpy+esj0aMd7a4PDglF7vMa/vb/HNVw94fjBn5TRvvPU2vrY8efGILEsQScIw7XN0sWB7a8TRxYy6ttQWsl6Pp4+fM1/V7F3bpmzb6GHmG5qyIellbE/G7F6bcHJywmCQo43n+OSS/f410iSnWl4i8CxXS0yeoKTi6PwMLyUjmXPx7JhUQ6IC+zd2eX4wZTwacXpxwWiS8cM/+CHzsuHi5ISq9LS1o65qlsUqxh47hdCCyWTIsD/ivXc/YLZacXj8gg/fv4d+mPLo4JSQSD7+6CNu7W7z9ePnBNl5IYkYWAQ6wgG+w4J9ZB8p4UlUdL11Xd66EA4tG5TOqVdLplKwOdJXTSRLAu+/cw3X1igZ2S5Sa4IZkF5k+JrOViJQlTVCKpIs7mF822kWmhpkFGx620ZFuJbRJsMF2tWKUCwpD89YzWpaFWGxEAKeeE1Ch0VrqRHK4zrOvQvhqvh4fOfzpEFHSxMhLUpE23HhNWk2oNcbkKeWYKfUyxXFYk7TBhKTo3UgU4HWCxpbUDUltmmxVRHNOjuDQVsXeFzMdbI+IiNKItaZKcITlGcsKzbdS46mFzQzj7Me79RVI5FSgve8ojB3VjKIK0hKCUGmFa5YkJmko4qC9YFFG9CZAelYlOC8pG0b6qZlMt7Czy5wtqXyMdMlENCyJcscTSXIjaBqG1q7YiwSLqSk8RJFFEF6QJk+veGY7e0bbEz2GA236OUj0iSP+gulSZN+LHTrwrAWq8CrotPtIuJuJf6720ohQ9SJ4DsLhvUC6urjo+DQW39FHxbdjmNd0vr9Hs7YeO8i4nX1LVZItJIkOomixC4dVKoUoRJQuvvZxQNxgM7XS+BcwPn1Hs+jZBonLd/GiaZrGkFKvFcd+cViHfHrI6O4uUM5cJ6iLji/POfRc9gbZ0yGA3pJPzYIaYikERkJLaw/zuGtwzsQQqMTjVRJ51IQ91xuLfw1ffTkbWz1DVV1RCp2QSVdCF+cEr2vWWOssak7vG8jqkGkpPvgo8+eaEilQoo+tdI0RYmtypg/5GI8RDR9dfwmr3+Q5f6Tr77k8nCOyVM2ru3x+pvvMRlP8L6lXtWotOuwMpD3B/zBP/1nrBYtuSj59D+94J23P2A6h/e++10effUF9954n5qcg4cP+OG7HyLTHkcvHvPg0+ckSrC7dx2djamLOZPxBKTkYjrnclogg+f9D7/D/v42ZVFwdnLGfLHgbLrgcrbEoXh28JLd69ts7m5zcnBMbR25B1G1zA+PMUpycbYkNZLbe9ucXi5pASUsIUBd15w+enq12pz0c4ZacvjkOeQ5d25tIZUkyxQ7mxPq1ZyXLw74MsvQJvDOd9/n2bNDzo5PuD7ZZLE1Y7i1wV5vg/n5IXdf22fY3+LLB19x97U7vHHvfS7OL2nbOKb+6He/y81rt7g4OePGzRs8eX6IqxegO9iIqJaPN3SNbWOAkjYxBChCEhqhAiLUGKnp9RTXrl1jWa0ic0uK7jQuUTJBiQQfWqT0JEmKSnqom3e5OH0EvQTVj/GxIQjaWnQIc0AbCUohiDY3+U6P4CKW64qSdjZjdXAJQSP0BNFroW7RgkgzDXEJLgixKOg1Vi5RugdKdg+YjCfLpqWxNTjXFXCB8C7SrPN+9KaSgXp5jmsr6qqBxqL1hN5okzzPseUlbnkOvkL4QFk6ltMSHGRZg/RQFQU2BGyIDtERGwkgonOxlIJBGngne87Z0xecH4FtcwaZiU1WKFZCxPwZ56It/bcU7ELIuC/z0Rcqy1JwnqYTeSqlccJRNpJFIxjt5VxMS85XsChqCjfl2t4GGxsDphczglMoaTBJtMA/O5egBOOBZGLgxo7morSEJGHNKNb5iK2dG1zbe429ndtsjncZ9AdkSY/EZBilrlCK4D1N3WCt6xT54UrTcvUSEZAKr7KW426HEG10WOsoXLSc4dVCe91k6romeB/ZXjJCarIjDvSzHq2K07ttW5wtOxq+xTqLUiLa4ctorSN1Gl0QOgjTB4tzvoMlo6q9tQ2NtTjvOtv8qD/pCNVEr5z4vpQ0CJ12eymPt44QPFpLlFAx4tp76rKgthXz5YK69nhrIGSIkCLIiGR3wL+Cx+LDpUCpzjgy/lqnrNJphSL5okWaHlLcoawPCPKUJNtlPQ3FBkOH5EYoDCIUFlX3LZ5ofURbIXFIbLyvM4mUAxqdYqsaa7t4ZmevDhO/teYiEFzfucZkY8Ljx7/m4uA+s0cP2L7zGjvXt3j5/JDh5h12r9+kP8gRUnPzzk2wEru64Oj1t5Fec+97Ge98/BHv//Bjrl3f44df/wE//Y9/w+cPP+fjD6+xOmx4eXbBwWxJf2NMeT7DtQ3T+YLlvKBuW4SJBSkJnsvDI+qyxmhJsLAsSryAoqwISrGYrXjz9k3uXd/mbDrn5HxKWyyRImG5ailWK9IsqqRTIchNzELJ+yPOLpc0UrNYLbm2s0WqBNNlgUUj6hWL+Tmv371BlvV49PQJmRS8vXuXr+9/jRnk7O1t4nxDJlNWVUmWZyQiMD1+ipSatiro7exwdHIKEt679wbSjbn39i0abRBCsJgvqCoLWmGbmnJ2EYOIRJcImA5Isj4ylFRYQvdwSR0jW5VY59k3kRWlWzYmG7ipomqn5JlEaI0LPcq6T104Do/PUKnl3h3JJAErBH/9k6+Z9Gre+d679K/tIhMDQdMWbWe/YrGFRSUSnza0paUtSlQbCG2gLgWYfrTYCBIvFEqFuD+y7sr9VqlYqIwxZL0+SWaQShFER1YIULclrqypyiJCOh1MJoPvsPJ4YhO+QosGiUU4F8PT8oxef4RWKeXiiGJ5TlUV1M7QtIrGSVbLFX66QliL8JY0S0mSPEIHwuNV3Ft4BWkC7++eoA5fcn4RqBtBsDVtE5iWDUEZpBCM+n3KpqGxbRQj8oopFpxAKNW5OCc0AUprozW68Ogud+ObY8tFFbg4h8ulQ+iYdNg2niSV9PIELWIQV5IG8t4QbRRKREfo61sjZvPAs5ll77Ucow2TnZu8/tqH3Lj+Opsbewx6Q/KkH3d7WqNkgpRrunXc0cmwomhXf3/Psj6EEinLaybcuunIDkKLtCYHUuG8ZZ2m6V1cOiupYnF3NhpZ+uhXBq+aS2ISvPA0dUVdTinrKEYMAqSI+Sw6MWiVYHQeBb1KEWhpWmIMt3dYH7qG39C2DU0bdUyvHOFgzbLyHRlBIFAmJVUZAoOztqMad35w2kRn6wCJTqnqAqEELki801ivMRjwmk4W2UGm3bSHQHYOBgh5ZS0UzzOxA60JBoroFybNBiIYqvYFqEuU2kBIf3VwodvtIGJ+UAjuSuMSiPuY0FG9pRBIVZElFUorlE5pTA9Z5rhqReOK35Qs9ps3l6ZpWMzmZCZglzOaOnApTkE2HD4qefLkBTVD9vb3uXXnNh989H2CbBgONsGkfO/3fsx8XvDuJylaOrwveHj/l7x89pK3PvqY3/2T/4H5+RNOXjxjsfyccgU69SRGEpINvEoxWUXjC5o2jmjnl+esiiVN8JSlJcXEBbSHJMuxIeCaBhUcezsTXnvzFl/8+gHKeoY9w8XTM+rGghLcuLHLpG65vJgynow5vZyzWi2praCoGlZFiQuBWRPQSWC/p/He4KxCBMWdu7c4O3jJ8dE5ZVXw9v71GL5VLil9Td7vc/P1W5weHlIuZrw4PuP1d25HIqfMaVrB84NDdnf3ODh+HgODkpy818c18PzkPDJS2hgS5gGhDVrG6FO8R5sUF4NH8CF0fljxhOatByxSzkizffpZTtXUUcnvE6bTmtYu6GUDRmkPlaUsC4XSBQ/uP+DgxZJfLx7x8OyMm+NNtoYbZIMe6SBBBElRN9RFzWK6pKxnrKopiRjxnY9/RN4fgzIoF5e1kgbnGryzWA911VCs5gjpUFlClmSYLCftDUmyOOI7F7DWRxzZtgjhUNISnEN6DyKJqvSmIQSHdw3eVWSJITM5QTaIoAnes5odUddzVotT6qqmaRWenLy/gUlzVsVDTs4OcE3UNA3HI/p5xrKOJ00hohuzMI4fvNmyUzzj/ryltgHrZNRn6QFD4fAd7dYHD0riC9+ZCyY456PQTSpECGRSkhvDoqqoqvrKKNN3XmJLFyhOLa6NhLNEa8BT1hVBKLIkY2gcm6ME3U/ACHavbbKZQ11XmKTHi6MFW9tDBuMdbtx6i9fufsD+7l02hhN6+RCto8uwlPKqoVxpcDpln2gcQlSEYLtSLL7VgHzct/xvTreuE1PGKSgyGQM4j+ugXQngE2wItCHgfLTEWU94km6nI4h2OU1J1VS0TRVdmbVBS0OiU9JsSJL0MUoTUbgAPlC3DUUdTS2ttbQ2wluuO5GvlfhrgoYPtoOBop+a7iyW0nSAUX0E0FRFVNu7Bq10l+6o0GmPJMmBipP5jLP5jLy3hZJ5jDjuxMtXuqmuSa8vnbgylxOd9CVOT6iOULDGZgXoZBOvUprmGUoeo0wPqXKEMJGGH2SExcR68omMMSFkt/vxCCyIKu5rfYkRnmRyEy82qRaeYlUSllOEnPGbvH7j5uKsAx8oVjOKEgoraJKKpFTYpuKNt99mXjnq8ox//f/+t/zy735CUxf88f/hn/P6++/SH/TZUAphW7RKmV/MmR4fk6UKqQV3774B9+5yenpGsYTVbMV4e4Ovv/6MrWt3+ZP/y/+TP/2X/4InTx5yNjvn0cPH/OLBE0aDHqO8z6rwhAxkmuLbltbHzr1alRwdnfLy+CUqzxlmKcvFjL3d2/Rz6A8nHByfMJ6W7O5s4GrL5WLJycWUykqSpMfGJMG2Fis1k0GCcC2DVHN0ueD5yzNC69mc9OkPMi5XS/ZvX+fBo8c8ffacdJRhzJisP+G1N97gyaMXPHh0zOatHbwy1K2mbmpuv/Yuj756QOk821tb2FKzuJyjrI3FWevIysoSqrrCA7kZoE2CcwW+WeCaFYTI/sLZuKwlQFBYp6LnlaqQumI4DiRqSJ71kVZyc0dE00gL2UZOUdY0K8tFU/HZ/Yccnh/SuIpwOuPlwQLOH3N8eIxDoJygCRZrQGhPaiy7t/r86Ef/nGB6FFWN0o4kxMLsbYt0Nb7LmCmqGmc9g35ClhmSTEeK6Roy8R7rWlrbEpwnBIc2CQOp8S4uHZsmsCoavHUMe4oky6P4m5i7IWVnJhhqmtWCcnWOdQ79/6ftT2IszdL0TOw50z/c0WYzn4eYIzIyI4di1kBWkSx2cewmJaAJCFo00EBrI617L/SiN62FIGjR0ILirgU1CJFCFckim2SxqnKoyozMGDImjwifzc1tunanfzqTFuc3jyhSiySUMsDTI93NzM3uvXa+833f+z6vGTGa7qCybZw3zGbL1Pk4x2CQsbE9YTQqky/GR6KKCJNGY7f24J1rT7j/5yuIgixPkcXWgckyQrWmsw4XUoZKZx1dlxzzRmlEdOmkiCBDkoEGL8gJNHi6pNEgdP5rSqX+Fi/7BXp0WCeQFgYDGJWarVIw2CvxmSHPIR9l5KOCWdMyuTnl4OAOr9z9Ljevv8bWxjaT0ZSiKDE6T6PN/t94YRDlhZK196akhXL0SeIu5GUAVaIPXHYt/auvH0fpfrkuey9L2h8kmazoO5hkKLxUhcXw1fOfooeTbNcHS9ste3yKQwuDNimBUmtDYUYUWYKUaq0JIdC6jjYkSnLTVLRt1V9WUkHRxqTuQH4NgRLTDi/PsxQ7HBNdQOgSowcMdIkWmkZmCCnp2jWKmGCbUhGQyDzgYuDJ7Iz8wad4D1fajvFokzwfpAC7vnj+xZbga1k4PT28b2Z69V3Av+gYexm4GSPkNXDPCH6F94skEBAqpbgKRYwOGT0y+KTgFIKgHCK0IBwidkQsOp8gRzeRekIMMq1ySofKp0j1+FdbXFyP6lhUa5oQKLIMEzxn53OuXrvGfDbD5EOCECxOz3HNZ0QR+D/99/8df+9/9b/mW9/6Fuuq5vZrrzGajBlMRrz8yiv8we//vylG51y/dYudvU1eeuVlnn76OdYLbr76Ol2wnJ4v2djd5X//3/63fPTez/jH//j/xpeffUlQGi1yfGURUnK8XoOAb771Ok3rOTs5J7Qtq7ZLL9pqRZx4VuuODz/5gkwKVFQMs4JCCm7duI6Mgflhx/7BVZ6dznuESiTKmJalLqJsy/lFYN200EWmg4Knjw7JygIP7F3ZZ3dvn5/+9D22x0P2tnfZ37tKawNvv/1tnjx+zlvffIPMDDg9PqddzXhw/3O29g7YmI7Yno7whefdP/9zXn75CjF2dG2V2vHRlKpZYx3kIbGMXNfg2xWEdcJ9BE/ok+eUMikh0NPneRisWCGGKZDJZQ7fWap5wyDLKIYlI6PJSslqtmQ+q1jNLc8Ojwm2wtaesigoipxWepaVxaDwBKyLaBEZ7UqaVcm/+pfvcePGQ7Z3S0YHA0wYsrU1ZZjlDE2B1prhZMpkMyeGGqMi5SBHGYMyBoT/ylcRAgLfwyMjSiWytG089XLBfD5jsZyhpUaxTz4dU+QDfJewPeljZMrsEC51xHi00UwnU1Q25WI+o1o+xndLtjZLdve22ZpMyXIFMaOwDiVlet2ZyG+9LTDtETKHwTgyaAVlFlkvamb1ksYF6rqjtYlP5ZzvC4NgtV5TFkUa8QSXiLoq3eyF8hS5hi7FX0fJiwP/q7Onz43xvUxWK3IRMEagYstAeapyCEpRhZTv0UbJwZVX+Mabv87Nq6+yvbnDaDBKGByd4su5VGz9hX/qqwIDfQYL0M9xUo6JMekc/FqOyyWv66sP7seAMZACqPp9X/zq/aKgh0OmgnOZQRISK4dLP1HE40Ia50iVRlFGabTOMabAKJ38M8IgVJL1uiDwDpz1yYEfBZkx5CZP4pUQ0usreEJMwog8KymygiJLvLnoQzKlItLlIAYMkYFKY7joPcJ5XGxfWC5dTKzDTx59zsXigtsHx1zbvcX2xgHj4YQ8z3sV6GUReVEygEvDabpQpMJNUq1dFhVE8qsGi/DLZJxWu0ShCL4l2lUvVwclPJlsieGErl3S+ZoQHXTzNOXRisHWbeTW2wg5QgiNkIKsTHJoqbO0C/ol3n7p4pIXOdFI6mVKKFusl+zt7jIuR0htaJdzPv/kfZbLimKwST4asrs3IYiKf/77/xN/+M//CX/lr/5N9GDCq2+MOHz8GBEzbtx+k83dA85n5zx8+AmuWbCzvcvzoxmz+ZLvfu/X+fMf/oB7H/2cv/Z3/jZvf+fb/M1nv8fq5Izj41PqusZrzXg0Ynk2oxwahhp83TAqFSIfojPD02fPWSxWZGoHrzOWy4qyMChhmYzHPDo54/Ef/ZAYPFXlGG5s4p2nqqskpTWGkEnWXWRvOKLtGqQuMSZDq8BqXZGj2NzdxLWWuy+9yvsffsrBwXWyQjEaGz7/9HO65ZqD/RFGBI6ePKetPufW7bs01jIsDNf2tslyjTOezekWjx895ltXdhiVBatVi+0MrRW0rcdkDUJ6pHTJAOV6VzzpOVIqR5mMaEFJT8ATg6YOCcq3Xla4KlCWI4aDjIvVOVXdJmxI6xBdxPuOt958jbpzfPzxh7R1S1hWyEIS6oD34ERimAkJOged5TStpJ4dI5ZzeGWC3NpkcfaY86ZAC4GtNZsbm1zfu8uV3SsMhiVKR0yuESpL0E0RoJ9zgyU4S722LJdVipB1Edc1uK5jvpxzMTsnxMB82bC9s89kOGRgcrxTROHQKlBIhTKKQmik6PBeUC1nRFGzXq4wIrKzOSXGjuGwQGqZVE7KIUzK5tAi8ObNMS/fOsY/smzsBKoW6g42RgJzGrDhEhGSrqMpTz39LIWYMDtZFjAKlAQjcwqtKLXECYNSCmPgYt30JruvArekTPBH2zmUkhgTGciO65sjbt7agOUZg+BYRA9R0NUOISS7V+/y1pu/zu0br7OzscdoOE5+JJkOENGPWL5WDf6jt0ujpNZJWHH5/79ih31l1AP5F0x3QoQkALlEnLxY1CRhSBRJdSWkQJPIxSnmIPajzpQ5opVEmwKl88Rp68UA1jlcjHQeWh/IfURrRxARax3Bh55AIPulfPKNGZ0uWkGIF8U60zlFPiDPB+RZTq4zFBLXNAjf4m1Nayuij6g0p0N7R4wB5zo6V9P4ji6kGDAhBV0U3F+teX5yzP7WfW4c3OLKzg22pjsMywl5XqK16cGel96ny+iCyy4v9q8DgUo4Zi4V29EtwS8IsgSV0C8xAlojiSip0JlDGIEQ15CLp7SnH9MunxLWz/BiwGA0Idu6nURAogORumaERueC4nIa90u8/dLFRWlNADoXEEqws7PDfL6kyAecnZ0y2tjk7b0Rv////AOuvf5rjLc2mE4yIo5v/9oOH/7sPX7wr/+An/3kJ/z1v/W3GIynvPnmO/zaX7nBuq4ph4b33/13fPBnP+bs8Jzv/vpvMxhNeH6xIM8y7n/wc+x6wZU7L3Gwv8+d61fIguWl19/g4NoNnj56zMUf/zGRluV6ztnpBXWMlMWAwkeKwhDFiPN1g5SRW1tTpILziznTvGDVRhaLFXlpqJuO+ek5MQaKMifEJF+tm4aD3U2u7Wzw6PCQamHZGZWsVudsbQzRecZ6XXHvo3vImFEt1jz89B7f+7W3+NY3XuKLKPiDn/wEmQeKi02kKLmYL7n4+AFXrm2wtzHGNpZP7t3jys4ug8mIo/Nn3H9wyrzynM/Oee+nZwTnKYoiuWblmCwn3TqVAQJKSXRWoMwAITOk7BDREmOLc5E2Co7n53z80S+4OGzZP3iF8Tjn4MqEZxfnSKXZmpQMc411ETnKufP6S4iiYHX6iKPDY1zu2L+dkS8t685jbcTWgVxJzCBDWMvQRLq5xTeGsCoYlIJcJ79D17Scnj3l05/eJ3aGd75/h3e+/RrTchNjgGhx3aqXUwrqdcWDh+d88NFnWLlAGvALuLq/y9bOHpaQ2nvvcG7FyWlH67bY3bqOlCOWyxmCSNFEjE4eDNtB21na0+d0KYgFISSDcUkIKsUYyJB2ECpDqg4hJWWm+MbLGdqtEMMJW9MZrg50TeDKtuDqRWB2AUHnWBdRIaB1TAfNpQtbpFTDrE/GVEKS5RnTccrPUW3DMEJZFlys2mTg7BMuk6lNEbPIsMgZZAGDpwuWk4s1RRTYpSUMPa5fCG/uH/DGK9/n9tXX2NrcYTgYYUz2FR3xRT25PMzji2Lz9cCvNOoXaJ3TdRYiL6KNLz8ufZ6eDkF4oUbEC4KI/aw/4WC+IkR/1ZOIKEFKVBQ9dyv9SkKNiEJSZDlNlqfdnU1elWSAFETRpqW0qdCmSLHHvWEyeJtyWsQlmtOn8RVJZm6ERMmCIi+TBFtrFBLhAp1v6WxH1zRYW+NdiwwRLXraMummEFzyqFiXhAlCCKRW/X5FUtctT48eM1uc8eTkCbubB2xPd5kMtxgNpgyKIUVeYnTWF5sAUqQ9FfQqRdmrxwSCgIgWGRpQBikyCMm7F0NAExHKkFTbnhAViAIzugN6SDRDWpUTqjOkUbj2FJNpUENiTIyypH32CaUjAr/M23+CWgxW83N85xiWY9bzC+rFjGo0ZDQeEAI0naOJ8N57P+e3f/d3KcsN9J5kMTthfzKmenrM8/uf8tMfGHQ54pOPfsHewdXE1Lp1jd3NKfP5EqcNrXPsT7d4/Vtv8Xh7k1+8+xP+6A//Nb/5e5KIozSSkY6cHj3l+t2X+Yf/9X+NH+T8wR/8M9poMMWQ5WLF83pOJiRV05EXAxZNxcu3rzEw4GzLZDJEy4B3LUZJSqDt3cijIkvQOKG4sjVhuRQYIj4K5nWHFZFyWOCajNH2FlvTEcuLJVvTKT/7ybvUizUjozl58oQf/cmfsj3apGordsbb7GzvMt3YZ36xx49+8j5GV8xOZnz46Ze8/OYdTFZSlI7r117lwRdPODo/ZTA2fH54ilGasjQ06218t8d4Y0xRKJTM+luoIgRFW1uCt8ltG2T6wU5IRIgtpV+wHhpWbk11vuaiXiF0JC8K7s9XxODZnI4ocwmZwEaLkobNrYyj4wu28ngp9kEZKDPwjWSxtOxf38FpyfLhMcvTisGJptgZ0FaerNT4Lo2IVqdrFhcdq/aCh4eH3L15ne9853U2plPwFtd1LBcVP/nZJ7z704/ZuC4ZlgXrEwsq57PHX3DdLrm6t89kb5tMCmKQ1BaMFinq1rcs1wuWq1VihkkSqDImaGQIl74MgVQRrRVZZsiE6f8+gQ+jT0VBKc3stGIxGDPN99jYdkRXkY54y6JtWdqKe01Klww+YHRSFoU+90TJ9HxcQguNMgihKIYjMmdBQOccAyEQkxQg1XZtEgbE/mvUcOPKlKyE9dEJx6cdkw0orh/g+tGWQDCYbPL6K9/lzvVX2dncSqOwLH8BIEzk3P4H/VIB9kLD+tXu5PJ9lZCoFGPad2NpIS8ud3yCdOi9aH4uiXeREEDK2I925ItdSi9UBizKB4RQhGgS3Tc4vLN0LiU3igi5zijKgs41QJLuX2biONcS2ppGrdEmSyF6SqONRmcKIzQxJOVb2ue1aATGDMizEYUpyXRC2UfvWbc1dVvhugprW7xte4hkl9hcQpCpnFKXyaNDMgrnMkcLg1D9rkYl9Z/ocS9N23B0csjp7AyT5WRZzjAbMh1N2RzvsDHZYjSYMCiHaJX3UxKdkjNlitiQUvapl55MBrQs8NEkv0xvI4rSgEzx0EJ0hCiJPiJRmOKAyZUd3Pbb2NVjQn1IdXaMXp1iiimqmCCzEqGKxOsLfVH7VRYX2zXkYkzdduTDDQZb23z47o+ZL2tu3L7FoChRwvOd73+PJw+fcPjoY7Y3v82gkDxfL/jgFx+hdUE0gY8/eJd3vvMd9jc1Z88/42d/+i/5O3/77/KDP/4Bw+kWvrJ476hbSz7awGUDZDFma2ePZrXChY4i03RtQ91YPvnoI/7+P/yHfOONN/HzNc61fHD+C4ajCdX5jFnVUFmHshC8p6oqyv0tojGIrqZaV8iiZGs4YX1xQfCe7a0JG8OSuqrwncU1C6T0rBv46MuHCVUyLHl8+AQVIk+fPefi9Iit0RTiBoNywPnZiq5pOTs+58ezd3nlldt865tvc3464/mzI27dep37nx9xcvKcX/vOd3nw+Amn5zO+t/ktWufY2rlCtT7EW8ViNmd77ybbxTbVusZ2jvnZnK7tGBznlGWOyS7T9HSC83mL80nhkmc5Wa7RWpMVQ7amm/ibV9l1Bc4OiDpjNBpCCDgfOTvtODqdM7+oKXLBajHjxs0btJvb/OyHR+RmRKaWhEzQeNFHzsK88qx9w+b+ElFsMLp2nS6sqNYWRg3eWYYMaVtLcJFV1RBlRBr48qMj7n90yKOj+7zx6m22N3ZYr+Bn7/+Cx198Rj6KlPmQZrFisjskRoP3OWfrBXoR2R1sprFHFERR0DSetjlG6QKpJyAc84vDFzdxKUGpLBnkAKPTuEcqhTYGrfPUYYSEabEu4fFdl/Hexx3XdzfZ3QoYucL5TaySdPaQW3PF8cxx/NyzVir5KXy6fQoRKXNNJgWOSNQq3WpjSno9vThnVBRoo9PPsY+MyhKTD6nrKmHdfUBKz+a04PZLu8gi8mRxSukDOxsZvvfKIAJZNuCll9/mpZvfYHfrgNFwTG4KhJCk8t7vNfoicSkrjl9xSfqq8rUFPSItooXAuw6UQl0WlMtG6HJMdlle4leoHC4tlSH0IW4S8aKQCULfzSBs2qt5n1Rd1hFDTJlGKjDKB4DAWot1HdJ2iK5DOEk/Eezlwx6tTY+6H2CkxAdH07V0rsWHpPLKM0NhNJlKRO/OBaq2Zl0taOpVMtq6Lsmee0IyMaKloFSBgKQQRR+VbJAqPbZCyhd5NFIGRH9hiTiCizS+ZlktsS6JVpSMjAZDNqabbIy3mQw3KPMxRT4kNzm5ycl0QZFllHnaCQ0HClMKnDAgZP88fOVhQ5KiAGIkkJMeYUVqEhVGTzDFLtG/Bb4muCXOLbDzBYJzEDZR1GWWcPy/yuISQkQGSde1nM/PuHrtGu9893v86E/+HOSAl+7eYDSRjArJwd6U6OGH/+aP+M4734SqxncNw61tJmVGN1/TrmrOT56xt7/Ho/aCf/x//T/zxdMTvNQolfH0iy/Yv/8IKwLffecdvv3d71MvZvzpv/oX3HzlJc6fPWHz5iusnz7jFz9/n//hv/s/sj2dsFFkyXypO4KQNPM1D0/O+ODzxzRNy6TQ4DqWqzXz1Tp5SZZrmqblYGeLTIyQlaFtOy6IbIwGECvm64oLnxhc1no2RgXLixXOe7anE7SLCCV4fjJDD6aMt7bZlyXN+oxl7cic5vDwlL3xBloZRhsTvvzyPodPn1MOBCYbMF8sGGSSn/35n3P71degbTg6nvPs5BSpIoPBmJs7uxw/O6Ra1WTDIbpU2LCiWyySIkqbdNPQGqV0YollApl5skKRGcBIohyysXNA0zjubu4jhOBi3XIxX3Mxn/P6/gbfvD7EBsHpvOaf/Pwz9m5+i8w45vFHXLs54q13rrKqT/ni0wtOHgacFahsRds4ZrMzopth2hHTyZiY5URTJIev8gTpsN7juhaUwAVN01h2rg9puoY/f/cTzk4jF2cLxkVDrjpGm2Oq85rR7gi8YrgtKTeHrM9bzk6XHB/Nmeqcve19hPJ0vqHr2sTu0jnVcgVEtFLkWXLAexKbK880WxsjhqMRZVmQFwXG5PgoaZuarmv6sZimrmBnJ2MwSa5uGaeowRWKdo3RpyhRsTWMbJk1J02JUYrWWpSUZFry+vUDbtw44N0PP2ZhAz6AlWl85p2l6eiz2SNegBKRXMJ0a4qUiqZdk+eegysjpnslyMh4ULA3lEw2BpzHtExXUrN/9SVevvk2+9t7jIbDJJMVSa31Yi0fk3brUjLwFwPRvupgLmnGUQiU0ZRlQbdo8c4jVB8DjOjf/avC8nWD5eVyWvSRxZfjLill30HJtITqfR8yneIEb/F9qmXXWJQSZEXOcJCC0Zx1WFcnGKlNGPwQQeqUXV9kQwblgGFRkun03NddzUW1oO3oc1zAY6mdw7rAuqqo6hVNU6duyF/uhDSCfvkPOAKdcBjp0TogL8ecJu/lzKFXzPVjxShAKLzztK6lCx3hcmQq0/vU3ZLuouZ8eYZWKRDtMgZBCU2mDaNixHgwZWs84bXbVyn2D0BsIkU/ylMClHxxeRLBItUlx7rnvHmZxpRSgEj5R0KNkHoK0fZkc0cIXWL3eUtk8astLkIq5k3LcGPCaDrm2ekpw9EG77zzOh8/PuVoPsV6z89++i6HhyeYbMjO7g28Kjk8nrF37TqTnW3axiI3dyi05P13f87B9Zs8uH+IjppxliNMehDPl2c8+vGf8N57P+XOrTv87b//D5hOhpzMjnn2szO+99d+j6eHZ9w9P+OH/+7fcPjwMaex4/j0nKggRo9UBdf2D/jON95msUx4DyWTA3fRtjxfrigHA2JRErzg5OKCIpNYH6lc5KxeoIuSPAom5Zh6vaSRkWIwIM8LXKjZmZTI0OF8pLNQDsc8eX7MdHubdVPRtC3b21N87Tl/fsb6dI4sSmxuyHPHyfkhv/79t1lfVHz52T3yXEJ2g52dPUKV0awk1jZMpyOS495htKAYGkYbAzY2hmTZJloKpLyUrGbJX6EURZ6TGZUKj0osIxslvl7Tnh7hsim4jkwLdkrBdjllMZQcHS+48dI1ijLn/qNf0AbN/s4+v/9P/19EFOczyf37QwIZ0uwy2puxPjlmMApoGemalIzosZi4pGwHdOeG4WBCiIZyUjCcluxfV1hfgbSMpoHBOFItO06PLOfHSza3PLkCM5QEW1BOR4zGO+zevUK7fs7yqMJkIw5uT6kXCxZHc85OTpls7tK2gbZtEr4iCJxPt0itTco6jwkUWWSarc0R21vbFOUQ3Sf/BQTOJaNd6IPa1nXH0fM5sW04OdtiHCNZKNLBGtZIbVG5YFpGXtkJrK1NoV1C42zs+VGgiUyLnLVvk8JNJB+CjwoXegqvTYYW7x1ZZtjd2GVna4fOzcmHjnJq0Lkmxsh0d8SkyHFFgQ8J07OxscPdO9/gYPcW49GEPM+QUr0YQb2wPl4Wk76wpJ7Df03ZlQZb8vJY6vWwphyQdS1ttSL0WgB1iST6msHy8vevupevfn3dmS/7jJjLr65HaUJM6q7Qj2PaxpJpjdaCwaAkiEjbWbpOYnSGM65Pd0y7KS0NeV4yzEuGpkBJjQsWqyy5NsSYJ/GAcNiQ1ImdtfiQSBVFkSFFlkQHpE40Wktru8QxCzZdBjTILMVcKKl6TJEjWNuTo9OYlH6X5GKg8x0+WoQImEz1exbdb4MiIQha5wgupM7Gtmm/FDxSKXKTMRkN2Bx9h2sbI7QeJ+ZdgsWlQqa+8sREIjImGCgAIRmUY+y7x8sZqdBEmiRNjxlSjhLeRgQEH/1qiwtEytEYIeDZs+c8enzE63dusX/tGnK4wapqkUIzyjP2dw84byw+z/mzn/0EKsfWlT2Wy4aB1Ny9fYMf//jPGG9uce+Tj5nPFwzyMTduXOXiYs58MUdLiTEwGhjmx4/5v/wP/z3bWxv4rqVVhsU68Pf+7n8OwTI/PuTB559RnR6RZZrPnp6ws3/AdDRh3jquyYKXb76MDBZTgM4LPv/sCxyas/maqmkph2M2BxvJkBVafOdQIXL47JSt0oCzeA+jvEBpRSY8o+mYpqoQCgbjYYo9jgrvoW4dy9mMXHvGRcbVW9d5fnzM4ydPCSz46Msv+cZbr7M1HfD4yy/5xtvvcP3GdT67d4/sYs3FxZyDzQO0cVy/ecDtO3scnZzRNUmF46JgVVdILdnaGFFOB+R5ep68T+oYiMg+Rc97RecC3jniIGNQCpbkKTgoRtbLRZqfyxTg6m3N85Nzbl3f5871HV591dG4CLblzs1dzpZr3v3ZU/av3uDl17/FxcUHtMIyvXqF4wdPiLFBaAHFkDZYohmye+MAYwq6pqKuWoyByTWD74YQLcUoSUKrOahmyZUrko39AS5IRhub5MUeq/MHFPGY4y/OqU7OGF4fE+IG7ekGWRf55q1JEnfMADFMP5De4VxAq4R2b7sGa9MoJCtyBoNEyg1C4GPEdT0OpHO01mG7DiECXauo1p7lsmGZS56frNmfjMgn1wmrTynzQ8Yl7Ewjro10EV7uOs5aSe1SjK3INE9nK2z01E3KVtdaUmYZWiYjYK6TVsr1eR6Xh3BTV7TtnOFUs7E3BBNxwkPoMOWAthV4kQLjMlNw48YbXN+/y8Zkg+LSU9HvN7hUH0EvmnBf61b6sVRfbC5/D7EnAouv5LFFUeBsQ/ARHfuO4/KGflkqvu7k77uVHtv8tfMlmZ8vFVKpJsn+UtRj4vvPU1U1TiTkvil0ujwZl4qGF1iX8P5aKXRvitQqR8qeJO4drbNU1mJ94CvpdUALidKa0hjiIFEoREiSrFR8BS6G5NFqapp2SWc7tIDcFBR5kZD+gRTwFTp8kpsk2GqEy4hpJxxRBZSQaJ2RZyVaZwhBej58AHWZV5OMu4SAjalrUj0jLYkTFCH2qZtavJB5X6pH0qjSEGKDiLFf1Cvoja39s8wlLybx3y65f+kkIXz1XP4yb/9JPpdPPv6Yg/0rbIw32P/ONVSwlMOMwbrChZZnDw6Zn8w4OW/It/fZHo0pyDhcPmI8yinznEcffc6/v/8Fd197k6s3b7K7PebHqwVt1XL47CghIWJgPV8z7zzXrx2guhV1FdkZ5CyiZzFf8If/9J9w/OhLNiYjtJCMd3YpM8XqwUMCktFgSAyWpqs5On5E19Wo4HEBXrp+k2ExQXz2GVVdpcWeq6hOj9mdTrEiMNCSGA1tFMzWFZvjAWMzYLFaI41KDnoVaaJLrmCTEWLEe49tAlW7YmQ0O+OSQkueHj7mvG4Y725w88p1Pv/sC16+dZP5xRJlIp989HOUKNjeOWBvfz9Rc2NHOYr81b/+myzmMx49PWI+syxWFyyqFcVghAiWMg8MBwotMmKIdJ2lbRMyXCqFzgp0liSeIQiyQYlygmJ7m/p8Qb1eYNsm0WelQuuMm1d2OF1UXKwce3sH/Mbr0Fzc4x/+jVcp9+Dx0Yzf/4P3iN0ZTx/8gvn8mNFkn93bL/P8pOV8dsju9Zzp1LM6DoiomGxtIJTCtDnjLYUxCrubMDHr8yVtKygHJVdemzKcZgTfELOW9XrJfLEmzp5gisjipICmY7g14fDnZ/jmGa+8ccz3f+3bbO/cYJjX/OSD5zjXoUy6CWZaJcVMlIToyTJFnuUUg5KyyFEyElxLFyNN11FVDU2bOlKfIgipY4FtOwSCuo08O+m4vuUZFZpSL5DZGaNhYHMHGieoHTQ+ctY6YjSsQp93rmDZODySTJmUIBgDShVppp7n6BjxomfIKUHrI1W9Yj63DCe75CYlNV4Ot4ajIXls6XqJ6s7ONW5df52djV0GgxKtEyQxHeN/cewVeqDmV38W+j1M7Mc68aui0Xc2l6Y/rSS5MTShfVFALovG11c2l38XeyPtV2/p88YXXxdfK0D/cacDUK1r2gg6k5g8MfGMThBW5x1KO0Sgd8CnjtN5S2c7gk9eKR8SEscHSwJVKrTKyI2m0Dq5/OUYMD2LzKfojphc+7ZzSCV7jH0SGUitEyVcRHzs6HzTdz/0uTWXLLlEUhaSJIMWksyUaf9HMuKH2O9oVPo46Tw2OhC+3yMl0YWUCqNSDlWMqVjEvqikmAD1ooYLVRBCjQ4tgjL9ofpKp4f4WggaAiEzCAnwebnHSd/f/3eZ+n/49ssv9K0n10OU0rgYKCRE29FUNWdnh3zy4Wecncx6JUzG3tYmp6eH7ExHzE4viO99SF4IDp8ds2gFenrGbL4EO8doWNqOLgqyQqXMCBcx0rC+mGOUY9E0+ONjbu9vMcxS4fjRD/6IIi9oW8f+jVsMM0mnFAc3bnK+WrK/s83s9BmZEOzu7HP/i0RV/sEPf8Cdl1/lpVs3wHuaaoaKgYcPnhNCR5EJFusOpQXD0rBaKearlvFY0HQtjbWcOsd0c0QgMsoUy8U8LQXHmxQDMHmkFIGmavjFpw9wSnH37h0kK6ZjxWSYsVosWa0WFEXO/mCLp08OieWauze/Q1NVDA6usX9VUw6HLBZL1qslT2dPIXjG0w22JlNGpUF2lmq+oK0MPkQ6mzhIyXBlkY0nywOFyVCDESrXdHWFxZJi2RVCDYh0KYwM0Kbk+tVNkAopLS8fTKmbmmJnzPPZc5ST7G6XiHyJ90uu7s5BFDz+5F0kcwZDieg6Mp2hc8H584qH7x8z3C7Z2J2gMsXhJ0eUE83urT2KyYDz5+cMN4dkwwHFVsvq3DHdnYCAWXNGRw2hwF60aFlw/u5zNncHXLk25PaoIJsdU9eSottHzZ/Q5haZlalDyTKENHgfUVpSFHnaSZl081XB4pqaJnZUTeBitaTtbIpRlhrXOhos67rmbD7DxZz7R0Oub1uuHxxjihrpYDAK7ErBsoKNOtJ2kpf2I8XI8OlJTlW7tBBHoDJNJkg30hhxwWFQyGjQmSbPBoSuTcbGtsMHT2s9y3XFeJkxmCQTJjKjUVXyWqDIx1PuvvRNru3fZDqekGd5PxK97ByS+5z+ME+sqfhi/5F4U6F3xvcF4MX+JXU0gUvwC2gj086RPu0x9uOsFwj/rwrD5bl0uWe5lC7HkN7/qwIkX4guLqnSsicUrOsKfBojXZIcijIpqYxW6KjxPvm6fPDEkECV1iU1WcCljlEpjCr6fQMYpSiNocxzBmZArsfEKLDO09gWT50uGoEUtiVS+mjEfy14L9HAvXBE0ZOfY3L8fKWNSCM/LRWIjBglEY2PiuA8zreI6MmMSZcjk0GEpq2JweNsi6MnM3Qt1mV0tlcS9jyyBA7qsbIhopTody8F3tcoGYnkpB2C6qnP4sWKLca0qxHqEtYK0bv+dfPLvf0nhIUJtjZ3mYwHqFyRaYFrPR9++D5ttaRtKrb2r7NYX2CtZ7GeIYTk/Z/fxznPxaMTlIyY8YjKd3zw8Udc2Zyyv1GyXNU0zuEEuLVnWgxoSo+OGtfFBKZrPWsCD5+fo4RFRk+R58znKwZZxrMvPyfPM+Z1Sx0TimL/ygFeZjx+dky1XrO3v0tVOV66cZ3p1hb1+QznI5vTLar5BbtbW3Rdxdq16FVDkWlESC/0xjmKEMkHJauqYTAYUK8aBBGzWSCDZzAomC0XuNjx+is3yETg8HBG5yNBKp4+fca0lJwWFxydrdjYsRw/m7Fe1eiRZjzU7I+HPH56n739a1hn+cUH93n+9Al3XznAOocVgWu3r7G9u8ug1ITQpRt5PkxYjhDJTMpvUSriXIoz1ULiXAdGE7oVdbugqlq2p7vk+YjoO7pmDsEjjU6Jd1oj8Hhvic6jC8G8XvDxh0+5/9Fz2lVDOWp5+61bHD1WdM+e4eeObORQCvJc0TaecqyolzU/++N7vPUbN5nsjCgGmu0bE0YbQwaTksGoZDgesjxb896//gVHXx6zeWD41rU3GU0P2Ng94PnTR8zPT3C1ojuZMx3nZM5x1YzYUhlh1UE248qbr/KWPefH7z+l7RJZQpkMJUHJDK1SXojsqbneQ9VY2rZFyII2aBoncSiULAhR03Y1bXRY27FerdBS895Hp0zzEftXInpPk5kcMfBkHqZbgnwgGU0j5YlgpxLMaskxJW3TJkNrppEu9LuQFDegIZEK8gTjJFOpg0OxXDvOFyu6flSx7SZMt6boUmF0iRcN2zvXuXrnLV678xa7m3uU5RCl8xeHNaTDv1/dpwIjeiNiTEWD6HspcugLBST5le8vtX3Hc3k2SJG8GKFGkAKrvn4I9RmSfceSvgbvw4sdDKSOmX40laKrQz+uS3J3LSVRpYPZ+0C7XicKsVK9GiuSlSqNPoXESXA+9liuVCiVjn0sdUauNVmPXkGIHjsvMEZTZEOMGvR7k6Qcc6Ei+pbgkn+kc5bWVgTfJBOskj1fLWGnFBKR5QSvcLal9R241EVpZYjhMh657xgiyBCI0dK5mhAcQo/IVEZp0khTy+wF2dk2TQJu+oZ1A4t6mUbhacp3qYdIhVr2gTkAcoT3kRCq9LmiTs+kjP1dQCBThkXC9fQjWR9aCDWdXRKj+9UWlzwzZDLifEuzsGSbE5R0jIzmwdNj5kuHCQ07e1dACHywDHTG1uuv8vP336OJkfn5kqzq8MIgjKZar1lJcFailCFXkq5tqVYN2UAjgud8eY6MAiUidbVmPNzgYG+T86PndHVFbtJ8VGiDFRHnPM5ZtJIcPXnMcrGkNILFekUwJaNywltvvcVPfvJnZAg2t3ZQBm5cvcaXX9zns3unzFZLRsMBZZZ03bPqnLwoGBaJbJvnOT4EVnVDOShoXCRWDVII6rqhKA2i81zM1+hiDG2aE89mM1bGsFo7ZquGzx+ldErfWaQ1mGKH6dY+i0VDns2Yz87QBJ4+eMDJ6UNGu3u8ducNbt2+yXBc4oKlay0IRV4MEMJR1Q3OCTKtMTqFBxmdoYWk6RrcsMDWS1b1Cre2FIMhOsuwdYePyWRlVJ666WiBmDJ7gie6wNGzc7745Ixya8DtKxN0XnH6yQkH7LNqFMIsiMqxGEucDJw8bRhONVku2diLRL9gPT8nzyTjzU2MEWm2GzwCOLl/xtnTYzauFLzy/ZvMni54/MET3vrdN7j+0ssIoXjy7DHRS7pFx5WDCS/d2KOQmu6ioz4+wz97yuP3HmKDJlcJP+5DBB8xmv6GljwubZd2K4vFmq5rUKrCDMfoIsM1kaa12G5Fu76AYkhZZphMUzctmcj45FHH5D2BejNjb7KJFJJyYrmaOZoqMJpAMJL545QDgvd03mO0Ju/5Ys515JlGy0SxlhKKQmC0AidZrvsuQni89ywu1hiTwrLKvEQoTecjg+1t3nj1L3Hjzqvsbu4zHI7IsqIHRv7F/YcUmhD6ZMIoCcKjVIPyEds7u0V8kdf4VSGKl7HAl7OW9Hm1knT+GBkL6FM2L7NsYry8tX/V+aQo40vd0uXivo/URSRmmVB9iBZoGQnSv9j5RCHonGNVNchcI4xiIC1lnuTkpk9e9d4ThUQrRYgR12eRpJ8PnVAz9DwzEpU5jQI9Lkh8tDR2ifN1Gg0JAcETQ0v0FQpBZlJ4npQCKQ1aFUil8dHTdhVt11C1Ld7ZfgRWoKXp0zdT8ZUyoGXa+SWsZCDGJBZwwROlSh2SIMnUvXkhN/YRjuczLlZLJuMudarpUeYyTjn0FUeiQZVgWyK+j4/uCEIh+nwZxGUCqQDSTs+6FV19ymr+GGfrX21x0VqSK89qvebGtesUheFsfoQKEWcFLma4usG2La+8+hrzxYKPfv4zNqZDCiU5amqsMegg8cGTKY1tW9ZCUpQjfNfx+NnzlGgYJQNGhFgRVMbmxibt4pQigvSOenaBtIHpcEDtYDoYYa1lva6RAnbHJZPBgMJoBtMx1rcIJbj36CnTwQz1R/+G5WxGUQ44nS8YjkvWVcWirag6R9U6hlmgaWogoxyWiCjYHI556c4O5xcLnj47Ylk3zOsWf3TCS7vbLFZrFk3HWdfQfdGiYoE0HYGI6zxSK4IyNCGSDwuciLQKzEgxzkpG+Sb37h9z8/YBjx894er2Lrtbm7z55l0+efCU6caEu6/ssrk5ISqNcwOU7og4ZK/QkUoi+8WoUoosUxgdCb4l6AQ6bOo1s/MZmY+sL54TwgSlc5QuCd5hu4CIDU7J3txVEkJHvapRpWK0YTh+OGdjf8rdt67y7Cxyfu+QN779Ggt/RlFYPnnvhKdhRTlMUcnzmWPvwCBiS7u64CJahhsbDMdD2qZFxpzgPft3t7j+xj5ag4+W9z/8mHIzSygqJPs3b3L2eEazXrG/Ybi6VSKiRYxzRld2UONdXCwpZjmDiwFSZSnYLkSEDwThkEIQRIa30HU1q/WKuk77FVPA3mADk01w7jnVak50DiWSvFcrTdum+IJKCD5/ZGm6nPUq4+WrgWv7GwyHa0oSpt25FmszFm2Bi4IYLQ5H6ARKC3KdbvlKKWR05FpSDgQqsxiZcDhta3BaMSo1ounofGCxbMjyCqJgaN5WTwAAWidJREFUV25w6423uXX7Li/dfIOdvmPRWYHqxxr/oVor+ooQGhA5Jgp09pTMzIk4rBtQ1Tt4ilSIgkyekV7DdLlWScqy1I1IVWCoIJ5A3Ef3SBjBV8ok0f9PjKmjiC8KVoowliIVm7Ru6QtR74FRKqLC10ZlIjnenfesq46oNUFoZJSQG7S+NMrGFzn3od8X+Z6a7WKSNKddiO4LmsB5RwwVUqrkyLdNeh0pgzagnMSFpDQTvh8hXe6rYiCIFBomQyq+npTX43yKkJYykbBl6Ps0KZHRpRGWBKUVeIX3HU27xAWXpMu2xdqE95cyYnSeZMpIjuZrPnryJSafsK81pdTIXtacHs5+ZNm7m6TI0bIgRknbfYFSBvQAokoCApmYAxHbF9KaGB1ZMUHK019tcVkul3z08JA3v/c9Blrx9P4XHD5/xsPHh+xeP6D+4jltUzF/1vC5rdk5uMbb33yTz+7dQxUlk+GAi1WDJeKcRdqAMzmz1ZpF3RGUT9W29TgjGA0KmjqwvbPP7WvXOX0WmS8XnJyvEXaAEIa2bolSsLQNWmuWXcNgUHAwHdGtKxbLGRQ566ajLHPGWnFlPOLo8RNq2+EvFgQp2ZoM+OTTT9nbnLK1vYMeDCmHBc16xcnZBVvb27x25y7tukJrzaTMmd65yeik4GxVUS9WIAUra7moPUEERmVOoQJnpxegI1uF5M7ePvePz3nz1hV2tjdZrTzLkzmTnZJpkRHcmpdee51vfOMun/3iM977+Qf87t/8Xd763nfpdM5gMmL/xm1su6apO6z1PXZD9qmT6dWphEeJQG4EeaaIztK2NbVT+Lpl3a6xa8+WzrDrGUpIykmGipG6ddiqpdOSLCsIBeCTVl8VmlIO2L+5QaYNeaa5cXOHs4cznprAXnbEMNvh6M++YOsiwmTMs1BxslghtKAcaTyBxWxBtVjTNZZuXZGVJcPRBlIYcIpiM0doB53l6qubDLaGaANtvaTcGLBzbZNiGvnuOy9x9coOmztTzGCMzifE/BbnT98jTnPyZgBk0BcFrfvY5hBwPj1+bdv0UNZE6/UelvNThLhguZy/6Ki8dQQVWFVrjNLYGFh2DW7tmF0sOToe8toNw1svC64fSEa6ol53nJ9lHJ4XPDyTdB68FGncFSONtRhhUDJBNYeFoiw05VhRlCotpY2kKBPZV/vk1F+3LTYozs8rXNtRDgdsb+xx58pdtqZbFMUQbfJ+HHI5HokvDvgQOlan/xNG3MMM/zJZcROT3Us8N6XQpkEpx2p1HRdzREyIFETPtfwPFEMRT14+Ix98QTUfYSlBbrwwZCYMzF/snF5IYfv+hX6Mcyk6iPJywR+4FD1JpfquBmQ/+kYJfAxUdZtC0i4XzyiE9MToCf0+KWkHIt5aGl8jJeRZRq4zpBapYMUkqI7o3tTYEpEYUSRMj5JYrel8i7YljXVY2yAFGKFf3PY1KbVSaYPRBUUWcaoP4lMpPdVHn1YyPsmOVb9XKvIcyF983953dK3FuproLMF78qKkLEYYM0yduQ989OQh1gu+FSJX924xHutUOGPPaBO9JDmGFyb7ZGzfJoRHZPmApGEWxCAQyiDIiFlB5kvK4QbWWpR68kvVjF+6uBRFjtgULFZrDmPN+cUZTavZuXKXs7Pn7O1NuTiqOJ3NOVvMyAcTXnntFbwPfP7ZpzjraZ1gZ3eLDMvACGxT0XpPZztE0OxsbEBnWViLowHhOHrygG55iowt49EEP9YM8oKmqWmFpaoblMlxXcu4LGhcx8o6qrYFrek6T9Urc4SQPDg+QiAZjQZ4H+jqNUscznouLlasqzO2bt7m2q1bPHtwj9GqwzaW+w++ZGs44uzZE/b29si0YWM0SuDA6QTftriQln9ZWRBkQe0dXgmu7m5wUEraJtmGldJc2dzmw6MvCL6jWXmuTze5fvMa8/kJh49hMp3yk59+ymw+ZzLNuXnjGhdtQ5Zv0NQX1OsF3l0qSySCy6WnR2KRpH1J9CIdjF4RpaSen7BOH5SUM12DrefkuQYRMAJsdAQrCSI5mO3SsYiSfDLCrQK5zNnYlrz59m2KsWbn6ojV/IAf/fvnqOoJr926wY1fn3Jb1Rx/ec7HlWC94dC5xrWe5VmL0YLZUcNwmjPdGrLemFOUI5rzGqE3KKYDIpHdW7uEIHBdzWA6Sly7zQFjHMuPVhx+0lJPF5giI9u6io2eh48/4aQO5IMxMYL1Fhc6CAKlEoalqVbYrk2wzSxDkSXJsm1ZnD3Hx14lRoofFiRsSSSlYApjaNqWtrMUUvH0ZM3JPPDpE8PL1yPbhcUIR2sVJ8uM4wuPixIfUuyuFBGhJSY3iBiQUjGaDBlNMnSRcCUqJpXRYJRm5LYTqCiIImfdepwLNE3g9HTNs0eHvPON76FMQZTqcnPxYu0eiQQh0uHazqjP/xiRPUfHJdJ8u5+xbyPCkCjAZAvybE7s9lKcOT36Xsrep9Hf0glk+Tll8Zz18vuI7CkmPMKiII7TLV4ohHCpa4E+/Ct1PQlYeblIToBKcYn/lamTCZE+LE724oDkh7lUgwkEBEHrPLULKJv2HlKGpK4i7W7SriHRqRsv0CoinUOQXPlGZkiyFxgbb1POS0ThcCnrRmpCcGRak2cF1jmkVCip0p+ZnCIrkwjERwIFsUgesxjS3kn0Um9rHV2XjJ8+SLQSDMoheT5KYYIEOtci2rTA9z4QbErPzIVBmREqG6CFoula6uqC9+69z3q94LtvNty++jLj8RbaJDXZ1wu84rIzAWP2aX2gs6fk2QZIgxCS4Dt8uJRuQAiCrw1Ef3XFJS8Kdve2ycoSIwO37r7K9Vhw+Pw5ezubZK7mxxcz7rxxnZOTI548OcQiEbZGhUAuBYWWbE5G5NLj25q2t5xeasBX1YqhgulowGw5J8ZApjNOzs5RSjBftIwHQxbVihig9h4hFXVVE33AlBllUWAj5EVGXuYsmxRSVFUNeS7oQorkXbUdMTg2RmXK4NEwX1bYGLj3xZcEaxlrwXA0ZrZe0CzWeGdpneejz79kWBS8+c1vUNcVsbEcHj1ntLHJRXOKEnB6PKMcGASRrqo4XkWqLrJ2HffuP+LZ42M651g2LdtXrnKxWlDMjtm5cpV107Ax2UKqIfcfHFKvHlLXNXfeusvq4hHr+RzXOaIXBO+xXboZFrkhy2WfRCnx3mHbhq71eFkgRjlyMGT9/IjCO0IMZEXBZDJiNB0hhKCWAdcmGWNWjBFEmtU5MsLFfMZgo+Cq2+TJlyfYRct0a8jr37qDrSRnh5aVlBx3Z9wyWwzDJpvNBd/Z36Y5KDhcn/L4ouLi1DGeCpYXlmZt8balWq0Zba4oRzmr5QV1s8IMFFmWYzvJ5z854uqdLcpJhjyraY8sRgoy1XFx3qClobxzk8c/+CM+ePyAwW+9ynAEEIje40KKcFUywwdPiBUmE5ispHWBzlaImJh1uszwSKwHvKXM8rRnUQMkc6xLQWdExXgwoutaltUav/ZcrFuOTgW7I8nW1FCUmnWtma/S1+BjROGZjgxaSKLWxGDTzTnXDKdjtFJfjZyCwBSRrGkJNkEQh3m/1LUCbXI2ipLP3/s5f0jGzvY+N+/c5eDaVTa3dyjKEVqpNPKh33UIcG1MTvjunLp5yijfhtwTZUFkGxFrEBXyhblO9IUFIEvM4tAhQmQwCKzXr2DDFMUmuf4MET6lC6+DGIO0SNmR1EuXpGRIORABXsz60+jzxaL/0pDT58CIF34UXozlYl+EpEpdh7WBTguMlr1hNUUphz4QLEZFjEmSHoPHe4g6mYtFTOM3QcRHT/SWGB3et3jpQIbkYwsOSSBXmpAVuB73X2a9z0VlhJCEBwILSJTO07fSzxS9axBU2K7F+3T5o1BIVWLMEC37iUO0Pd7GYW3Adh4h0ygxJcz65JlqVyzXK+q6Yja/oO0aQvDcvv4qk/EGpkcZiV5OnjpRmbCiArLsNt5quvYxuhgjVZn8Ot4TbI33yXzuXYKR/kqLiw+ewUAxnhiWywV1k+PiAq08e1s7rJ4/Ybp7gB9OeWt/h08++pzDo+fYxRkGT2EUN8cT3OoCl2WcnS0hSobjIXLdsKzXTEqFzFJewOZ4wrprMLJkvnZoY1BS0LqOWdWwt7dP9C0yRsqiYL5coWO6ZYbWsjXJqao1dmUpypJiMgLhsa3FO4cocpTQjIcli0VF21iE1CglKLzl4vQZtZBUXYfPFM46orSs246r166yt7XLsnEIJ9jZ2ub56Rm//pe+z9n5jC/vfcpqVdG2a/anW/hgOV/XrJ0nKsmjoyXOzShKTTkYsfKBqzf2KUc5G3sHKCWoas+Va7d5/4N7uNojYuDq3QPOj08IPrX3Iva3QuWJviMTnkxmSClTV2Yb6tbS1h3kAjnQiYtkLROjKVS6NAwGGqNIs1zhkSrQp9mn1tgUZE5Qbk/IN3K29sfs396gXtZ426LliI3JBjdfckyu3uLs0QmfPj6iOPXcffkqr755wMM//ZCXfuM1Ht074/xKxfPlnDKraV3AdYFONqylx7ucEOZAUnhFL2gqePr5CeVAYrzhrbe3OPhrrzMaleS56ZP9dghygjw45PGfXCDHQ1zbJkVY39UpFMKl77HMMqQ2WCdom4a6qjFKko2HSFOkEKkYMEoyGAwoRwW4HEFKj/TCUGQC27VIrRmNxiwXc9oWbAfLuuPZhUj7rjij1DnRS5wTGBG5uSnZ2oSji8jFWqGFYmAUuVHIPuAqhMTsFZmmHA7wzYLoI0UuKQY5nQPnA2dnRyybwBdP/iVSKHSWM52OuXLlCtevX+f1N9/izW99m53dPZRQZPk21rxK050ico1u1oRyC6Ua0McIocE9RYTv9XTipA0LxER+4HLUpVC6RWDxbhOBJIiMjmto8RMy9Qsc30JKh5IVkYwY8/7+3OfD0xv0LoOQL3UC/XtddihaaoIU/U4k4EJAR9C9SVDppBRDCDrvKYKmNAqjUkaNl5pGKlqXik6mAkoElExdoxIqdVTRkkIfHQGbxmrB4l1HcB1WJzGIDz4p3gQI1ad1Cpm6rJBihH2I/aGcfiXqABAD1ls6a+mcp+6S9DvLHdpahGxAwrpesaqW1NWapm0ShNVZtBJkzRpjDFnW4F1gWa1YrFe0TUsMnh99+C5V2/CbneOlG6+zMdnEZNkLlH+Mhojpi7UCPErvI+MQ2z1ByhOUzAg9oTr1kC4VFvHL9S7/CQ59cG2gazxCBM7OjynMiMlwk2AMam+PybxCC8PVvQknz56xc/Mmh599xqeff0ndRW5MtpDOQfAsqorBeERbN+AjuzvbDAeKXGpOzxaUA4NvHFkRyPN028mzjNZ1SCGomoZMwmg8oLH+EuyNUZLOOTQwNgVz1yF8pG5q8kxipCIvMrwU5NJQrS1HZ3MaJONcQAgcXNlJHVdUWG8Z6ozKwcXFCgkYZbhoOkxwxKZiZ3eXO6++znJZ83t/4z/j3ynJ8vQEEQPDYcHxxZx8U/HkwSNW3Zo6Jm9Ds7SMSoX3kVXbcuPKDR58ccy6XnN+8pyoAq+99hpPP/6SwXiCs5bl/AIpFZlOQWBSSnSm0pRYayKauo3UtgEUzgnqLhBjwygk9EUmDZPhEElAKnDe4hanuM5hrcMLaDuPWz5HqqQ8QxUYWTA7rBhsJRc4WnL+vAbRkU8KvvFbL1OOC55Phzz9QPHs5BhfPePZH53wjb/6OqOh4pvvjBgdDFjWloe/eMST0xVHT2YE4WnXHdG1IAPrZUezimxdyVgvApMDiW2WDJsB5/cE1UPY3JqwsT9i69Y2+eYBSq7IbubovQH1oqYJa7JckhcJx2G7Gk9EKAVE6qqirj3r2tI0LRGo2oDJa6SSyeBmFM5ZbKtxIQkkJI7ORpyMtK0jtKCNwRgNQlDVFZ0NrOtIblIg1SiT2M5CsGzvRd58OSHrHx5XOJ9TjhTFRCNUwvxL/IuDIEiBHmaIUNIsakyuyTaGdD5gjCF4xePH55ycLwkOqsWcJ8fP+fSLL8mNYvqv/oDf+Su/zX/13/wf2NzaQ8qcjb2/wezBv0XrBULntP5tcn+Ack8I4mk6yPV9VNgjhiItQfpSEOkXxbLPHeEUKQ8QUWJjJIRNrL9JXv4ZGkkU1xHC9fJYmRbsgjT/R/XGQvHiZn9JEbgsL1oovFCI3jwYfIKJxuD7ELlUmqRKRuHettIbI00KuRKQKUXtoPORJlhkdBgp0UIn9/ylhLf3+lxGCbRdk4jU9NRsmdRn3kesS4WuDYJlu0qeJSEhaIKHznXYrqW1yacEXyWrJrNz+twAeq2JeNbNEh89TVNR1w22sziX4g2Sux+6bsWyCqhOE12katKI1vn0mCzdmnc/eZ/GdtRdw2s332RzY4ciHyCUIEhJ6Hdogr6BFBKhNzBiRHDn+OYJPp4RlSJG3e+jxH+0c/v/ubjUVcfj+oKtg0CeSVrr2d0do2TgyYMvufvya9y98xLbGyO++PTnrM5PWTcNGYa8nCCEJTM5r71yk1988Avu3rzOeDrh6OEjCq0QwTMebODqBoVkuahouw6dZUkR0rSIkBNCwOgEfQtWUAeP9QmF0LQ1WZ7a6vNFReg6TFHiY1K71dZBplmvawZlDjjUcIDKMoYmY1KWVKsVmZSMp5v4xmKMxscEmV1ULePRgPPTY9adY1Lk1PMLrI/85//l/5bDw6d88O5PyWPkld/4TbqLFVK0XPUOVUz553XN/FmLCFDmGaURjIYFJmge33vC/PEZ8yYyX6fkxArHs9kSua64c/t2cub6Gi2SRv+F81eZpCYRmsYq1k2Dd5LMZOmWR8ZgaxNBxAZLJjJGky28q5LcMUZs29A2NSEKhNFEqRNqPqTDWGiJCJFnX56zvmd569cOCKHj/PmCq6/skI/HPP38jC/eO+LN799E5zkXFxZXFszNgg8ffMlUBd747l2y7R3G5yteefUGt27Dh4f36GzgcFnjxw65LanbJddvalQZkcKDDUwKhRznUGSoMtKKFYf3lzz/cklo7uGpqaYLqrWirRuUDJSDgqJI2Iy6a3GdQ2U5Uit819E0HVVlkSpHyJzlaomuYTodIXVJ11msXWJWDTYbM50O2JxkPD1tWFaO4NMIp3O2X7721N5e9RRDYFDmKJn2N0UueOflgrZeIbKSrmuRKmc8KBgPCrIiGWFj//7pNp38LjIraW3NIIfhhiZITZCRatEwHQlWleP+oznWX2amKIwUVNUFP/i3/4w3Xtvm+3/tbzEeX2Nz9zvY1X9JvfonCH9BGwxK/SbBNwQBMWii/xn54BOsfQVnhxB0f8CHvhAEiBnR12Rqjgv7aEkCOorrBA4ZlofYLuLZwYf0POD7kVw/IorR949X2q9IkXaIqSkRCWkiIvJSDOD7/Y8PKRIjJIZXjBopdU+3AOsjUkiMUESV6NO5AKU8WJ8O7ZjifqMLhCAhCEKwuJgSROtmTd1UiVFHQCnITIbWOe4yoMy6FKdAz0cTEoF5keliO0vbpShv+tx6QSRcCmWU6nl2yU/ibItzPWXDe6QUlHnyacm+sEsZwbfYYAlBEKNLnZjWyQzbd5lfPrkP0dN0NS/dfI29jSsMBkNMVhDVV/k7se8ak4E1R+h9hJ5CqIl+RfALvFjju/qX3rr80sVFqjRq0VmG945yUDIYFjx9+DnPHz+gWtZ4KdG3rzKdTNnf3ed0WXF4fMzV69cYGcMXn9/jYiMnCsH8Yk7wkcFgyGJds1ysCdbj25bheNJXeliuGxD0iXGa5apK88gIUhhWdc0l5M0oiescG+Mpy/WKTqYApfT+gXXbMVUZmTJsDsZ4WlpXMx6VhBDZno5oqhUX8zlbw33K0ZDjJ0/RmWF7c0qICYinRSAXDt8FyiLDrs/5/f/HP6LMNKprqL3gjW+/w2s3XuLf/7t/wWw54/R8ycXsjKZpky8ndnzj1df4z37vbzM7mfP88AllOWC+WrOua9774CNivebx4yMUnnJjihSRQQ65EWjjk0u4H1+5qLA2qb1CkBiTp2WmcJhigCkzfLWgrSoyCbapem29BgFehh58KQm9L0KJ5C9SMjl+oxKMN0fkweEbQTEYsnfDYDtBngvqdcvzx2eMpgaTZ2SDCd2sId+Y4vOceVjxox/dZ/TJQ8Yq5+Yb1zFyzN7dV5DSsnGy4mg24+2/+ir/+h99wHdu77J7q2TpI5/+wQP2tgdMhxO6xnNyb8Vka8hLv3WFfHvIyf2aH/+zezy7WHNw9y4ys2jlyfMcEQVNVePqihglwgxxzlC3LVUVWK5qTJGwLHXTkGmFtQFtPMSkmqh9gxgZBmVGPjC4sKSzySsjIqBSfrzsVWcx+H53EuisS2gPaem6wGql2L8hODpa451mczdne3+IGeXp1h0FMV4evslcKCIMhmuu6+dMy8BoLJGFwmWbNFXDaT4nix3HzzuOZwIbBVIJrNLkCjaF4n/+R/+IL7/8mL/7D/4S480Dxld/g66+RaEDkREXy3/LdPK7wBhCQIrfIYYvybJP0GpCZ2/ifYEIKYdFRU+UEmdfocg/piMj0hF9jfMDonsToZ6Rlee01tC2Q5zXKYuEPub4xS24X/L3euVLDAyClFAZU9499F42nw5tpRVed0k2rzVSG0TwWAcVCusVykXMpeESiURghMIJQW1roovoF8tuifUdnW+xztK1LZ3vCM4lOCwKLzwxdnifluuJZqwwwqCVRoqUV+8UaBkwsUWJDK8dQoAWClQiQ8eYLiG5ysjzDKEE3jdU9RIZUkijUgalsvQQCdebGrt+ZJf4ckYpijIBNkEmM2yIKKWYrRf87POfMa8WvHTtVQ52rnJ9Yzep5NAv1ISi97fEXt2XXPsF0gyQZg9NwBiHlPd/tcWlKDLuXt+mbloiju3pmGq54uqVq4TlBe/94h7lxiafffQeucyoKs9oa4MbN0vyyRA7nzEeab64f58uKJa1Zb06ZTgoqFrPou4oBkOK0YCszCijQeaG+bLFxEipMwYmpzEOpTTz5boHf3pa2xGjxIzHWBtZzCvmVYXKJOumZlgMGOY50npcDKzbBr8SlIVCqshESwqd06yXeBGpu47ZcgUBmhiIdcXmoGRSas6XNUYKdve3ODk9RwnJxjCjKBTTyZjFRcfTZ2f8i3/yP7OxuYXrGpbLOSEq9gY5MkwwuWRUaK7ubeFtw9UbV/itv/47BDI+++gjNoclb73yKn/8wx9w//AJnRJ0aJSEPNMoncYnIIheYl1H20HnBdYLlC7QKhmojDToskTFpLNv65aRLOnatlfeJP+G89Ch8D39N6l0BDF4mrYjKEEuIpOdnLpO48PgoF41fPzuId/8zRtMd0bcfTuwfWWT++8fE51nc28DqQQXjy649t0rrE8nPHsy42wbHv/ZF/hzS/WoYef2JtWJhUmJGQ7Yu3YDk43Y3Njg1ksTivmI1b0Ft1+7jlCe1c3Iz//FQ47+75/wnf/dS1wwZyZazHDCetXiupqyCHSNpq1dAoxGjzQ5q8azqivW6wXr2tJ2nqpZ9I9HJAbDunIgLWWW/BRKCqTJEUakXY9O4xURE4FYCUkXUs5FDAGp0gjF+TTgcR68C2RS4aPk4TN49FQyGmZc2R4mJI7uuVCksKx0mPQRzTFgqmdMRmvGpcfg0FaQZRXsX2F7Y5PBZI408LNfRC7WEW8lZS65c7Ukio6ff1Hx5ckPWZ97/ov/zd9l98pjhKjwagutr6A6xXL2A8rROyi9mRRk8WW820XwmMx8SlBDfJiCLyEaXFjR+KcYfcpo8Gfg1gj7gCC3WfvvsFrtMxpritwhxZxqrQgMeyPr5UDr8jXXH2q91i09mJC87LGXFAec9VjrkcoipECppNbyosUJQYgZOqhEVvBJfZZpKDNBptIoTsmcwgRCsNTtgqVrcDHtUZxN4yXvbT8CVGipyXSGMqYnS/eeH6kptUl8wWxIpgdoXSbyg3O0naVpaqp2hXctQgrybECWFWiRFGQ2BrTOGJRDQojU3RIhMhRzlJDkOkU6h+CxweJ86oy9v4TTdmhtKPQAozMQEuvTRVjI5FVrrOWzx/c4uXjO3YO7DF//NcbDMeisJzakop4KTOomiaEXIPRxAeEycvyFjvxXU1y0VmwMBwTnmG5MWM9XFINNilzQhBad52QmQwjBYr0myhEIydHTh7xx5TfIcsXJ0VO0KKmcReUZdl1zsV7TKokVkaprGOYFzXJBFBZrBS5GvHNsFiWudbTWYhRkRYYIjqIs0I2irlsigmVdMyzTiwabWEGqf8BGwwIvFcZ7ll1HYyVXDiaMhKRet9TWo1VSEyE15UBTtg1CZlRdxWpeMR6PGI1zrh3sMioLrPMs5hdEFXFzy3g0ZVBWaN9nbbuOYAOrakEUku1RCVrT+A5ZJgT6k0ePefMvfZcbt9/ksw/v8W/+8N8xHWW8+fJdfuuv/zbvfXaPetkglUCZpMH3UacFahDYrsM6cEEkVZTQFMMymcDqGptLgmvoQoOzFpHlND65ga1PP0jJlBb6rIr0w+Mi2M5huwqpPaODCdPdEeFkRTHN6ZqayY7h5W9uITN4+MUC11n2rgsGA9j/9gHbt7Z4/PEp2VhhWo3OS9ax4eGHM67enXL85JSm8rShJe4Lqvkx/+oP1jy/v6AqBjx+rtk5mvLskxp73LH+BBAWaws+OnxO05yx+F9aqrkBOUEoWM7nZCoSjEkO7qhTAqLK6Kxi3VYJDaJz8kKicwi9kQ9ShrqXkSg8Uqd5vUIlZZfw7G2PONsaIueBzsO6rpI8GYGzCY0hk0sQCTgHi3WDloK69Tw67kcvFm5vG66+tEM+yZFB4kQqWGlh3rO/rKVbzTl/Irgo71LuD1HeIrEw86Byhhsluzd2+S9e0XzrHctibvBWsnPtDjvbhv/xf/ynLK1kU+U8e7rggx894Hf+zt9nY3ebQE2wSxBTYtykWi0wqgNZJCGE0BBv4fEgGqBBiJogW5S0aK4SGeBCgzIeqUcIf8HAH3G2KlmthwzKOXmuUMrSVBbnLwGIPQ/rhQCp9+PIvtBE+iLkX5ABOtvR1F16vYrkEbrEooQYMT6ANkQTkCYhXlLImsPnObkyJF2aIlMFrezwvsY6i3OJEB0iGD3AmILMjCh0TiYVSpcIqXDBAslDY4hk2mDyIUoNMKpAIrHOstbrdPwHl2TbpmRQThmXA5SSOJuAmlpr8qx8geCpdYtSHUYKZFYSRYYP7QvKd3BJmpziICKD3GBMhtEFCEEQEmcbIAXVqehoo+XxesX5+TFvXLnJzYOXLs1FySQrZR+rcBm78JVgj/CVH+mX1SL/8moxF1hc1Fw5uElwjmxrTDmcMr94QqTg1TfucOv6Ff7wX8wI3YrF4hxXCgiO+w8e8e03XsKUI3Kp6FanaCMoBjlBKGbHJ+RaYm1HaxXTomDReAqlaKNHFQWqyDg7u8AHj8lHVE2Lax06M0wmQzIj0ljKGWLbsT0coY2mc56z1RLnM5ROi28jIMsNLoD1gpO6JjOa0HVsT4ZUVUvbJP/M3uaQvDDcunGTTz78jKgyFIJHjw+Z7O7wxjdf590/+3N8U7E5GrA4v+DOzRtMxhNUNqBpVnzZ3Ee0DhUM1lkuVkv29/d4+OCQOy99k52D27z35x+yeLbg0/f/jOFUcnTykJ29bb73G9/jt/763+ZP/5d/i2CGsx0hgDI5xgyJWqHyNo3K8Iim6Z27E7SBrq1BQfSermlp5zVua0iQkiizPgxrRqYUmUnoDmcdnXN0XUvbtFycz4lyzcHdPTIjGG+VfPqzx7R9uNe1l/dZzjoWZyt0Jjl9uqJ1EdmAPOt4djTj1iub1OctvhZMO4XLcnaubdFUntk9B+NtRuMhT58pmnPJ7jd3+OKLczIZiJ/OqVeO3AgOP5hhyozl+Qp2C8aTuxw+03R1YKxblPKITKO0wAdYVy25Fkid0XaWulnS+YA0JRHV53XkKJ3jnKVuO5CSolCMRorJcIRRGhE7aqWJncWUGXs7E4g1i6qjaTzKpBGWDySZa+/QFwist6zayN60AOE5v7DUnUUJePsbEwZbI1DgVdpx0S+ViQHXtayfP4ULz2i6hxsNWIeEs/HRJzqHA3/h+RTLZBR4486Yb77tmQxLVHaBAP7B793i/qMbvP69X+P6zTtMt7dYtx3Zao1WmihGxDhM/ickbRQIkngH0j5JEpNmn3FfAGIvOJii5I0UTxzWGO1QmUfINZuFB2EQYoAgT3HcIlCtO1zIEr/qa1HH6S28KDMp1Cp1LN6nWGLbWpq661VrvXgpJlZZcIHgPD7L0d6gQ0BohRSCzgVwnpjnvbJMIkVGZkqMbwkojJJomaFNzrAYkZkBWpUJZYPEyIIoPNY7bAyEUKFwaQ+TDdEiQ6CRiP5QT1HJ3jXgHTpPNO48S5c/Rw14LjUSicuW2HdG5yjS68FHS+vSct7aOhWXXtquVdqRCqmROu2IbWiw3vUvjoBXafxlfSI6t22dVHExbfUv8TXxa499wt3EhH6K4UVX8ytXiwkhmEw2mM1OE89rNGK+fg7Ss7m1i8ojq/U5WkFZ5EyGmsVyxa2b11mGguW6JirF4fMjTDFgNDBcnJwTSG1lMRpgjObKtX2Mt7DSVMuK3emQk3XN+XxBXmbkMqN1nnXdIEizcTXSbEzH4AJWQ4um3Jgwu5jTtE3CrkjFctUwKQyZVhSjQc+V8tQ+0GHZmY7RGlwr2d3fSaFXixnjcsx4NGA8GTJfW8qspKo6qvkFi+cn3Ny/wvHTJ0gyJiNBOci5cucmZ89OODk5YbKzwVtvv8TZ4QU///gzkJG2njM7X/Hg4ae8/e3f5MbNO7z/kz9ja29KYz07KsPGhidHR2TaMTs7Z3cDgktQQZHiFogEMqP7rAqJsB1ds2QV7xOix5MhC4lrLevlGtfYpPJRiiA0ih43EdOS1naOuu2om46msSyXK46PT4hS8I3uNYwLmNxw7aUtzo5mHD+dM5xmKAUvv72Nj5LPPzhiczPjYHdMPhQMp5JyVCCFZP3gguvf3ad4rrlxcxuN4ckXMz7+wSOKQUlTBRYLx+bBkNHWBt2yomojsswY7Q9RWaCzEjNSFJMRT76oscuGMoN86FEyOepDJEE7AxidOElV19G1NSrT5OU0mdJai48e361puw4hYVAOGQ8LRmVBYQT0DC6RnJSoLOPWrV02xguODs8pZEnnLU0XsP2StnO8CFU0KhWx6MFoTdckievtvQ1efnUfoVQKzIrJeZ4Q/NA2LfPDZzTPL9jav0LYnKTC45IUWElPrhzKdIQYqZ1mWkZ2JwtyuSZ2EYIhL27ze3/v77MWb0C2T16U6aatVD8ZsF8TAMWv/Z6Kh+xBlf6FhitF9YrQL6iRRCnwaIQYYUUaCRo9RkqLkDYVIqkIPqK0pygtbWOxocfpx35/+MLfknaJX4WWeYJPaY1N3VGta0LweG+SGdh5vI24Iu1iTOHIXY5xDpWZtDOWki54OtdSZD3AFNAoRtmU3ASCh7IYMxxsMDQTtNIvvDcutIhe3ONiUmUhQKkcbYo0gpYmkZS9w4eOLqT9DcGSZQWDomCYFeS6SMt0KV50D4QEEKVXxUmle45Z4iU2bUXXVXRt8yJ6WyrJICsodU5mcqTSOJeSLUM/GiPSe7tiKnjS09pEWO4rC6lL/goXk8aW8quCf9m2XBIxf5XFBWC1vmBVXVAvLZtXrjIYapTySOPY3r0FwrN/8xmfvf85YjgiSsP21ibzp3Pq1YqyUFy9ssWjZ6dprt12RBkQSuJj5GBri6vXb+AWM0xRssjmCcS4XDHIc8bFgHm94HyxIvTIa2sFs3PHoCzQEopBSag6YghMx+MUf+qTCSkIkZTmQhGkZmtrysXJjExKOtthW4/3kqgE89WSa5tTlg0cPp/j3EMePj5kUXsOdja4dWObrb19nj0+ZHtrG7RgWa3Z2ZniNQw3hzTnh0wmI5zOuHb1LhvjFQ+Pj2G1pqkjB9NNPn/3Q7rFkvXb3+W9n/2U7d1tvvnO9zl5fs4nn3zM8fMz/vLvvEMh/zLz5z8g0ybdzNwSHzqEytDagCOFXHVtMnB1LQJPyCR4j28ttrM0jeX8YsVwNMLkKYTJKEVwMc2ynaOuGhbLmvPZknW9QGvBcDhIyPLWUrUtdWW5dneHbLjEWonJYbwx4OOfP2Yxn3HzpZuMNwsuzpbcfXuf6CVRlUyveybXx4yvD1lfNCxnS67f3uLs6Jw737jK6fM5V1/eYrlwPHh6Ql5EsqEmyyTZQIHMicKhVOT4cct81rC1UVDkEqlikmRLiRYJM25MjjFJgSWlICsN040NRuNdfHAsV4K2c7S2oywzNjc2KMsBwTVEa7moWubLirZzmM0DwiQHpciNYUtEBrlienJO8B2LecO66vBR0ViHC8kjr4VAG8FAwLKN+E4y0JHXvnEA08lXlOIYIPSQSh9p1pa2he29gpdvHKOyExq5wVm9ifOCl3cfMxWnaOVQBrwYgxaUyiNtQDLGlO+gxr+NHrxOke+nDPSQlr0xuBfnWrogX54sPhWNy7yXS0WX7AtJkm+9MASCRwSJELZXkEmEs8SQij0iIHvJMaI3dMpAPgzINgW3pfPtMu9dJqBicMSeBhBC6Hcukab2VKsG5zy2czgbcDbircB2jjxvya0jWE/IM7RzaK2RWqUxmdLJs3FJD6ekUJ5cCpTQmGJIUWwy0BO0TEmprWtoXUPnaqzt+qx7QW4UmVIYaVAykawDEY+jCzW1rWjcGrQmK0qKvEwxCUAXE8pFxCSYCr3/xdmOGGwyffbFpbUNtuv/bZ/CyqKAXOWU+Si5+lWWiqB3xJCUdN57ogSpYi+tDggVaLqqh1YmhRxCXIZVps6kJyS/GINdgk//E+rFL11c0gy+5fG9z/j80yOm1+/wvV97k0km2Ng5wGjD+XzJjWvXOHo6J8qSoC33HhyxXFQMC4NrHKGrubK9i+s8aiS4eusKx8fHXFysWa9WXBw/Z9MoJjrDFSk8aT2vGQwzzudLHB6MYFTkLNYNG4OcGCMtgdmiZlgUCBk5PDxiZ3cHoVKErIiiz33Ima8rupOG0e2cq9c2mF/McY1kUBiqUKeVYutYzlc46zg+u+DKjet889vf5IfvfkixOaYNMJiO6Z4+Y/fKAdduXePLTz7n6o3r3H/8lGqxhDyjWS5YrBzvdo7MCHZ2tpnuTgjrFhlq8kxx+8o13vvhn+J8Q5lNWK8eIITnt//Kb7IKHda23HrjLh+e/AhlMnxM6IjgHELJFNcbI03rcEGA7H+QYySoDFyHdQ3eR1preXb4lKwYMJluMBkXaJleYDF6vI9U647T01PatmJnOuZgb5uyzKFzRO/oqhnRlixnkZ0rUy7OVpwdNezd2EAAr35zj+F4xHppefzROd/47dt4GXnw4RnTPYO1HaePZhx+Oefg7iajacatVycMt7dwnafMS54cnzEaaIqBYLl0xNpTXtlidrQA3zLcH2JyKEpFVohkptUipfkplZb3SqC1IUbwtkaIFNNgdIaIFZkRbEwK2qalaZN6aZCXaKmovWNdtZycVRydXNB2lr1bQzbGhqgEXksKU6J0ZFDuInxLt6homhVNk1Iuu87jXFo4twgWC0dtwYjI3ZtT9l/ZJwqN6BlTod91RZEoxSYv2Lp6FRNnVNUzirqiGHgOJiU6nzIyU/IClBYIcqLUKDrw5+jhyxSbfwU9+ja6uILIpgnobzuCSz/RL0jJokft0zOlXoSCJSvtZXCXfLFj9y8+XsrLALKAkOnGHaNNHYeQyXsSE93YO594Wj3/S6lE5oCMFxkz0ScxQ7w0JNKjWxzJEyOwTWC9DHSNpc0dTWtpmpyythRlRlFmlDbgXcA7T+YDxoRUSHxEZL2MoJd5G6377ikk020MSN8RlSUES4gtzne0tqGqVzRdiyAyyHJMNqAwEzJTIIROX2+0eOepu5a6XRFiR5GPmJZjynyQQs2Cxfo2qe60QUmFC57GNTTdgq5dpUyaS69N26VJi3U4l5A2eVZSllOyYowyBUiFdTVt19E6i+tHhfSECh98/5wVzJuGZLvpC0gfBJa8Po5LcFDoKdJCyP6a8f+H4pKoqg7nI1tXr2Jj5Pz5MaOr+2T5IBnvqg6lckwBTdsxnoyRwbNcLnjy5CnXru4yW65omxqtJONpztOnR0hpsK7lYj5LuvyiZN1V2JiAEXmuCN6yqlu8EGxNN3Gdpcwkuc5YVxWrJnUeuQ7sbW0go2Dd1qxdg1CaYVFQDgq6pkGFyPZoxOzoFL2/iTaGcVHStg2j0SjhvCtL1TRc1DWT6ZhCK4xWvPLyTYzRdE3Hycmc/d09Tp8/Z3tzzMt3bnJ6fs7OaMS7P/wJv/6Xf5NvvFVycjLDbI3Y29zk4YOnTHc2sfMV57MjHh4+4d/8+Mdsbm+yc7DFm2+9jckNIXg+ff89jk/n/PRP/oRX33iDvYmk9TlOQBc6fBAoYQhO4m3AeoEL8atlMJKgM6JdY6Ons46DzQl1s2ZRzVmvl5yaPM2AjSbTmnXTcnR8grU1+9sbXNvfYWtzSmYMrUzz+ChG4BxPPz3l5e/tUBYd1VxSzR2zoyX7d7aBjGdfzhhtTBDC4G3Har6mGA1pqgEqy5nulJRDg2sznj9dc/+zh2ztDPn05+c0qxVvfGuXKA2f/PwZN1/ZIssFV25ucn64pFsLgg+MJxnlSKJqB717O4okdHCtA2wqst6mGTUSt64onWU8HqNV0R+mNbZ1zOwZURg655gvVpwcX9As1uQRMtulG74Pid6mFSovKEYjROwwWxvEZonv1n1OjMPbNM8OMufocMZk0eJt5LW3b2HK0QuarogpR+MyoCsGMMbgVMDZbZ7GTYJt4QLcqWIydFzZGTD0vaQXQcCjZcnW1ttMd38LNbqDNNsEVYALeJ86+jQzl30BSaUhXmatREEUl0XFXw7HXizL++3ui3PBe98XCPEXRmsp1UCmmT4JLpkCwdJcX/R/F4TsO5bY+1s84NPYjcsZf98l9TssvCC00HaBroZmHagzTzG0DIaWQZuQSKmbiXgX8UUkyw3mBZus9yGJJCHXRvbxAA5HhYygCUSlscFjQ5dwSj7tOjKZIpSNGaDVCNHbI2Kff9Q1a6p6SYgNZV4yLCYMyjGFGRCI2JAKlogiedaESNLnrqJrVzjX9oVV4HykdR1t76eJPqBM1o/vRhTFEKmSaKG1iQeYup+QYrv7kZiWKTVUac1ivcI6S94vey6J1OHF64O+wLj+v2U/FZP8sm+/fOcSA6enz5kva8Z7d5mOB+xujxlMN6i6Dtt2DAZD1q3jytU93v/5J7z8ylvMR8c0zTnrJbz0yuvosuDBo0ecnZ/go+TktEntpIoYEeisZR4EO1ubzGdzZhcLZB4ZlgPKvKNuLAWaKjqk90nZEwJd1VIUBSJ6uqpmOpmwcA1ZcJRFxjgvaOuGeVMlkqr0tDhWq4ZbG5ucVxd0wbFR7jIdD5ldXDA7n1OMhsjoCW1LYyV3b95gczLl00/vcfLsiN2NKd/93nd48ughH37wAaPxmPVyRdd0fPT+L9gYj8lHJVF4TKb59e9/j9npnHtH58hM8drbbzGrOvYPdtnd2mBdd0z1iKaZc/3GLut6xf2PPsO5io3vv8FimZARbdfhhUTJ0M+eLTEIXB9EnukMmZUYQcLsh4ambtgaTylHJWVnWa1WLJYLVrMZ3iVSqvOBzEiuHxywuzVhOMiRQuGJBOtApAViFyODQU4gQ2cTtvY9937xHJ0ZMlPw8z99QFev+dbv3CEvFUpmXL+7waP7Cx7eO2M61ezd3uKLD2ecH83Yujbk6p0xZ0ctZ8dzJtOCcpgTZMagyCiKAlMOuDitmOyPOTtdMdkquDhqyAqFbyzr1qZRzv+nvfPqkuy67vvvhJvqVujqOAkzGAADkAQIihJpaik46UX2N/EX8Jsf/OYv4G8kW7JEUqQpRgwGmMGknulQ6caT/HBuVQ9FWRyuhcfea/XqdLu67r1VZ5+99z+IHus9vXFXPATcrp+shKMsEzqn8EHSNi3O9CRisCRwLcZ7euORaYoex5ajHcBNOIcHjIykPasSiixuQHBzMu8g9AgfZxOdaVl3lo1bU5meH/z7T0kODyI5kBA5HyF6iGwXhK3ifFSkDggHozRwa7bk5SJnuc7JUrh5otHa4oJHBEM5eoe9W39JWn6ITPcIXuP6gc2+43LAYIgeU8ubOl67AyKPKjLzfeSdvJFYtlVP/Cu/mxn44JEyIKUm+NjiGibV0X9kNwzewo8DQrhBLVlGbtagxxX/12DpS4hk4R2OiSHRR1Mw21m6xtFUhnZqh6oxzlBciIZ5caYzPKoIWBEiYTQ4nFeDJz0xCYYOH0AlSUxErsP7fhhsxzmlTlPQCQaLMW6wGK/ZtAtW1QWdqdEyZ5xPKPOSPCmQUmJsR296bG8QwdMKiZeCtm1Z12uqZoPw8b67EOitpe07jOkIISLTxsWYaVGSZwVCR4Rrb3q6rsEaE6kCQxssDJYDOklJkug1s6pretfFLahQ7Lj6AyIvdsc8wvexuAkpERn0Zvv0X4+3Ti5KKd65fRfakvnRMUkWsD4SE1cXl0zzEclIsOxzDvbf4Xj/ktenCx589C7nZ1/SNR3/9Mtfc3gw4dvf/IjPHiX86jdPOdifkSSCIjuhSANV1VD1azbnBtv09H1LZwJ13TMeElqzqUiUIR0luGCZzEqcjPa1Skj63mBV3Al858FH2GbNer2hE56syBiPR2glkETbTqWg7Xsa21NVaz64fxdjeurGkLnAuJC0tuGd++/x6uUp2gXu3jzi0RefUa0vuNwseff9j3j18py9gwNa+4L5CIS3vHz9EuujIGGWFhwc7HF29oyzs9eM55rvfvJtPnv0FfPZnOOjQ0xV8+zJCtKE+++e8JH0zGYFTe9omp5nT1dUbY0PgizPI9PcOQIaOWDhg7dIZchmKZPQYUVH3bV4C61XCBk5RKXQJGlO37V0XQ8ekixjUpbsTWfkWQIhYPEE4/FOYrvYbpCZYL2pqX7acuvBHKngzoP9AbVjuHGvZDyfkeQSZxz5NOfWgyPWq5p3v3GETuD8Wc38uGB2kuGDYu94hEobTp+sSHSCC4q+a7n1/oSnD9f05pL9GymbStNsPJMDTZ5Bojydjex7Zy1BRbmg3no6awh28Igfds2KQGstq9rhh3Uq0wojBrVfRBx4JylJklJ1HeuqodMJgYD3g9KWjG2UTGgQCTYI0nQcCXJYMA6tBNZaNl++wDQN3/7j+yTHBwNDm20BwUCrHqCoA1rMBYLzwwLqCUHjesP9A0d1rDmeSxANq41ltZGkao/x/AcI/T5SzQk+4Lpmt7Be6ZsMSWW7UA9rxdatMhLqhmvmr7CnIlz5wWxjm5diWyxOu0NQEVU2VDRbf3a5NdsayKFy4BXFRGfo2k0k7iqBCHagAkaRxugcqnc2Ar8TAbwJtIO4YnS63CKctokl4EVKsh0kBI8hRL8f4YbKSaCVILEebT2i72Co4vDRY0frlDKdUKRTUqlxztB2HU1bU7VLNs0S4xq01qRJTpZmpErjsHSmpelbNtWa3vQEEZXVnI9yL1W9puvbmKh9TC5d30UFbwJZmjIejZmWM5J8FG0ibI/pa9puQ9NWO2HVuLGKqgFZklOkI7I0Q0jFumvoumZIJkPV5bf+PFFFACRelmwT/LaKfdt46+RiraPtHFmuGWfQBUcICts3VHXFuCjQOiHLMoLzfPLJt3h+1nN2ueHuvfdQ9hnPTisOZgWPvnrObDrnxsEmZnxAKcnR3ojDccKXZ0uyIqc2NVkhmZXTWApqRdvUzA8mjJyPez4XfbHLImW+f8D5+TJiwaua2XyPTCvqtme9jqTKQmn2ZyW5ECyspwuBnz15wnxaorxkVa94+OhzkkRx/84xq4tLRnsTbr93h/39PUxd8fzZM05unfDdP/leHDJKxWoddxuzg0NElmL7louz1xydHFGUB+gi5fhoRlmOmB3tw8PPGRczfvajn/Dy9ILPnOev/vo/cnA44fTp5wipeaEN2ju+9Y33WbeO50+fcnp5jgs9eR45CMJFqfjeGDwdW+KTFBI5nWNDh3GOzXJFt+lYy2ywQI6S41mmSJOC8XiM1oo0SSmy6M0uZdRW63qDtxYbAnmVRUTNZMR8v2S9qnj22SX7N0uSIiHIiD7yOGRS8uppTVO3zI9nICTvfHQDawwqCdx4b0S1sKyXHa+fXnB8a8J4mjE/LlmcbVhebNg7KUmzDC86vBKcn/eE3jKea7TMKEfgqhrhB0JekpEXWVyU6g2dcfTG0RsPUkbwg5Y01g+eHopUK1rjsTbu8PMkJUliO0tJSZkmpKlmMhkNq39cEL1xSC2x1lKLDqWgCAmQIFSKkx1NCHz1q8ecPnvFt77/MfnBXhQDVgML3Q2WWdtB/jC8Dj7CZr0LeC9xDmwv+bw5pMh6ZnuOcd5SeMm4zDiYaaazD5mc3CXPJjtJH+fFdh86bEyvloet3woMVUmIO3kgAgvwQ8uKYYbH7/Dntslm+6F0NPIScmvsJXfJJYJprjzZ2RH0BDotSbM9gjM4u8DbFYJmSHIJMELI5PcvVB5M49nQDfBaiR+U1/3Q5hMhoELcmROiy6MVIopXCkkvBLqXaG0HiaX4wF44tFYkSUaW5SQ6A8BZS93VLDYXrKtzeluhlCBN98jSqJjssFRdEwmVbRNVuL1DC4UlGpQZU+N8D4QIGd6ROSMqLNU5ZTFmMpqS5QOPzdS0fUtb13S2iRpoLpJOAVKdkGclRVqS6hStFTY4ltWaTbvBB4cKg4+N2s7uYwL3A3pMoIbKcrspebt4e/mXwa1t/3APFwLT6ZzlukPSMR7nTMcp62WDTBLUOOf14pIgU6omcHx0xO13A2teR4z7ouXifIEMDokjy8c4IUmSMV3bgpOkZYluakYqZT6JctWr9YZ37rzD4Y1Dfvmr39BZx0hClmas6p7FumZUlGz6Bd4JlB5R95Y0S5nMpxjfYbxlsV5xUEzoHOwdztGblEwr8izFBYdKNJu+4dsPvk1bHVL1hjTP+PSTj2kWDQ8/+4K8Kvi3n/4V69UGkLx8/IRbt2/x81//ig+/8SGnzxbMZjNUmnL33TuMJiXTUcEXDx9hkQgvWJ9vKPKM1eUZrYf/86Mf8m//3V/yyccf8+Mf/j2nT1sef/k5d96/x8d/9L3Iw2grsjzHB6iqliAsQYjoDWEcvTMQBFpq9t67i3cWa31M4kHQNR0i06g0tnSEipwZrVKUiIgq4cCGHoTCuTgwbZo4KBxrzY33jpBa0tmGvVslj37ximI64vWzBZPDnMl+zmrpuTi/4NXjS47uzBjvW1xnGB/kNMsOrQRuA/WXC2rRcXA0ol0HJgeaW3en5FJRjkdcPvOoItC2jk3VM9/PUCPJ4rJn4gTjUvLy8ZqubUmTjHyckuVZlMWwhtyBtU1UOE4SRqOCvIgoIWMcxvQMk01kohAiDkaDEiRpgtQa4T1KQ14WtMNuWgQI3uEdUTDQR9ENmfnIM7fQrSqe/OYJ7ariw2/dJzncx0jQIRBciDMWPA7wg7xJcB6cix7rkaIQd+EWrA8YK6nbhFWl6HqNlA5rBMf7JX/2J7fIRYEICmtc9PuBKBMUIqdcbuceYitfH2HAYvCuCXEEMsi/cJVQQkDIq2SwSxy7of5QnQBSbY+TkfMx6G1tWyrxIeNAfWv1G9cYBUoj05zg9un7JbZdIEJAkuLEti32eyKA6zxtbckzg9LxOSol0apFsZ1JxtadkzFJSmLnQwqJHaosKSRShmjoJhWqkCgVkKmPTqNBRHHIrqVqa/q+xfmeTJcUSUGms1hdeIfpO6q2ousbpBDk6ZiEFBMCQXSI3iCiyTQ+WIy14ANZkpKnGXlWUBQTsmyMEBLrepq2YbW5pGtbnLWR3BkAKSiSjFE2Is/HaJ1ET5kQsLanDYa2r9kJkgoPQiOEBrZt1oBE4AeAR0wtbz/U/4OgyARYrjYcHBwjVYKQhtlsTlfXLF4+5cf/+Gs++tO/IJsIRKapLlZY53j5tCPBc/edm/jVObfffZevzl4STE27uKBtK1QukHlOu3QkCdR1jdQp5XjKanOBc5513dGdvkJlGfPZlE3bkzhHCFCUI9Z1S8ByY39KVVe0y1dcmBHzScbebEzvE5T2BOG5WKwJSrJaLBiXBYmX3L1/j9fPn2OWNfffu8dsNufj73zKV89fslcW/OKffk3d9nz3e3/E+eIVl4tzbtx6F4nm4uUrRrdL9q1hvDfl7p1/w5PPHlJ1LXuzMW3b8fR8wapuePb8ObW1nC0WlOOc0XSPQidMZzeoNpJi3PPxd77F3//tP+K9omodv/nNQ5SUqHyEF7CuW+qmwwcxCFeKKDXRGxCS8WSKyhVght2wwPUWE1okWezt20FVV3pMqIdF1iNklOYOQaJ0LNnrpqVtO8Sioq6nZGXC8XsHpKMEkUge/fycpIA0z9FacOPOiMVZS72XYk3N+lKjk4SXX26YnxQ0qxqpYHxUcHjrmJ/+r+e8fPaMT/7sDmkxJpv0dEbw/MmScqqZTDMSpRBesllFpesUwfKs4vJySZJo8kIiVULXe4wx9H1EYKkkJZWRJJrlKUWRotMM1TXDbESSZAUqyfBBE3wg0VDmOTLREeVkXUw6IVYTYeCwyEE7LBUCRELX9Jw9fcbp40sSAWWZ8eD7DxCTfEDeSKzwu8QS+zkh0oy8x7sIQ3YWQohEvNgaswg3QHRDoDGOL74UBAfCB8rsJp28R6n3omqDs9jtHGVodRECW0/IbYIUQ9KBCD0OPs5U5C4JMCjmhogkk0OFyBXIaAtZDYShYhk0vKRGDsrdUsgBTuwiKT/WljsDLaxD6GxIRIDQJMmcQE7fdXiXopxCDO7wv3ep8uBMtNjwaUowAXqPTxxOGbQcJkhesS3KwtA+3CItY60VE2NUyRYEawi2ibIt1iKkoO062naDdz1SS1JRUBZjCh1VyIWQBBHtodNEo9SYXBeM0hIZJFXf0zfRwsA5gemjb4v3kKU5k9GUcTGKPBadAYq+b2hMT910NE1P3zc7IUwgoiLTLELxB2uDKPMf0adGBNb1KoI2gozAiggBih4+waNsBcEQkjGQxB2TcG+dLv4Ahn6EqW6qisncUqhAXa84PJjQ9Zb9vX1UKvjii8d88I13ef+9dynHF3RNyy9/8hP250eEoEnLjJPJHDHNyKTj9bMnmM6zagzL1YL3PvoW6cuX1JsFfdvQNjXlaBKtaFE0Xc+jR4853C8ZZykTnfPV02cEmWKtR+pASKAcaVRvCK7DmRSdKMpiQqI8q+Vrbt+Yk6UZp6fnmMaRaMlnnz/ixtEhp89e8vTRYz746AGjbMR8PGN/NqHfVORlwf7hnI8//ojleoOkB+F454N30Jni8MYxXkhGeUryjQ85vzxnMk756vFjeuPJypzb926yNy/IsoKLiwXSC3rT8slH99nb2+Pi7IzpJGP/cMLBrUPS8ZhbJ8c8efwYHwRNa6Kabxshg2Fw+TO9wTlI0oRiOo2JwRlMZ6guK+y6w+dJZAwnGiEShPQ439F3NcHHayxVFq2ArWVUjkkSRdvG/qzpDG3Vko01xTTDh8B4r+TuRwnFVKHzuGBMDkf44Dh9CvOTGcFbVhc9Fy/XfPVQ8cGnB4xPcuykZ1N76rbl4HbBxWnNerEhm0iqZcN4T7JetOg8LmJtbdEq7iS7dcNytUBJybgcMd2bxjlg09A3BucMQkTpfCk8Ug5S51ahlEcLRTrAQJN8QhCavu8Q0pFlI7JRgQiCxnUILXaIID+0xQKSICTIgXzaNsyMJV/C3ixj/v476FQS0gTvIo9A+EGccdgJhqHL5p3febgEp/Au+q47GwgurpbBDTwTL1GOOLQNkkTCrLwB+RFBZdg+DrHDdqTyBu9NbKU+tii1XSUQVYeFFwgFDFXs1o94W6UE7yPRUQ8qUwOz/srbXsYEIAdP+t2HjsrQzgy1mhvmc1dYAucMkKCkZqserGSJVlnkdgTD1Zn8nhBxM46DYCCYgDeeYAPBeIKIiT4E8NuEFqJVuPAMbpkChkomSIXXYMwwf7QB1RlIBQZLoEElFmUVqSooknFMrGJAxQVQSjMuJkiZUugChabte6yLM8+mbqnbqE7uvCHPcibjPSbljFE+jsRa22H7DU3bULc1XV9jTBcRcU6AFwgZQSJaJkipcSJgXUfXtTGxWIMMguX6FO87FMlQEJoomirifXGqgFBEqwHJsAmSb30P/gCzMM+6qjk4OmZcFjSbBdNiRNPUTGYz1l2DCZYUsLXh0q9IVUpre9LRmGwyZTyZcTTZ48mTNbf35jz8zS/wzvLggwf84teP+PLxM+aHN/n+D/6Cl1/+nJ/+7Fe8PDvj1uEB0lk+eOcmv/7iMXiFsJ5FdQbljCQrOD1bkpUl+/tjkjxncW6p2wYpA3kWyKSmraP7pPBR7vpw/4BEF1ysalJ6qk1D3TvuPHifzXrJ+XLBo4ePEEFwsn/AjZu3mM4Puby8xHvBbDxDOEdvWqaTCX3fUmYpJkjauqXpen78wx+hpGC2P+f8csX9++9g+o5U3cMbwy9+9SVJlvLhg3c4OT5BkPHo889YrhwPPv4QZE6el6wXF9F6oOnpDXTG0zmHHkiC3lukHkiRiWB2PEcnluCg7mqqZY1rHL31WAuIjrbtMS7qxhVFilSarqnwdoN1UU48W69IU9BSkBUpxiRUq4piniN6j/eC9UVDs6qZHs5QAnoP1cJSrQL1uuflkxXvfvOItHDMjjJOv6p4+sUKlSeYNnD6eMHBnREHRxMun684vjdhszK0TUOSJeis5+jujGcPl0wOFCooXO1wwUT740xxuL/H7OAEGyzZasFGrOn7QJIkKK3w3g36YhlJlg+JQpAMvIPoZOhJlUTp6PmjdYq1ASFMVNWVCcG2cVbi497bCYcUnhAkeZ7QSk1xZ495EdtIXkhwoDw4NczDtrVDYFC4tREk4KKiQLAD4c3H9pn3AZxGeo8FEulR2qGlQQhBOZvx4ONjyrEm+ATrza477oPfJYj4PyOcd+tlz4ACE8NgP2xbZcTnIgWD0rO8aoNth/oSokSI2v1eCYXc+rzs/FWu5jaCJEoPeRmr5+2IOMS5khBxziWQEQrmIQwcrJgt3mJhE4OiPwzeVmGY3wcw0ao58vYG2RURhhZg3BhFtNgAV961xhxB6ah0PbQcdWERmcArixKGVDqCFpRpQaITtEoBPcgASbSOHKpE5YgA1ljqrmFTr1ivF6yrJVVTIwmUo5LpeI/xaEqSFyAlfddS1xvaZkPVrqPcv41ClltbkOBiXRlh3iKiPJ2ltx1N19D3Bm9j9bFYnRN8B4yAyBXb6oUKwIskJvlwJQkTrkTgfm/8AWgxiVKaNMtYLi8xXYXwJelIkmcZXgmOT27QGI9pag73DknSHILizr27nJ8+Y6IFr9cLHv76SYSZ9i1tY3nx+oxus2FUZJy+OiVPBcvXL6i7nsODA+q6pXUec/oC63vKZELTGmSa0yJ5vWpIUs24SFDecHm+RgnNKM9jSZ8qbt4+4dWz54REUDWa/ePbvL5cM5tMuDkbo2gRrwTG1Ny7/YAXp4J20/L61WsOD+ZR2wjY35vTVg0/+8UvOTw4YjrfRycZQmcEa8lnJaq1jEclh/MDlt/8BNMH7r13jxu3LX21RGs1vHE9s/mUYjJlPN2nqnu6rkYoz8nJCaPJHiIknJ2e0fZNZCQ3HdZFN8Q00WRZig8OG1wc0ieKvCyYHowHBz1HW7fQbf0xBptb52i7lt5YxpMxo8mUJEmisyMK7y2b1RJrOlKVkGcanWpCcNTLJdWlopiWOAvWNJy8OydNFCoFRCDNE1ZnDbO9HBs0v/i7F9y8O+Xo3pTJrKdtOnCe9euO4Hpu3J8BMJqmKCm5fNFw+WLFwa0cLRUXTyvyLKpCN5ct5VTSNZajk5LpbI/5/glZMUHKwHya01YlddvgGCTEXcAZj/PRrwYpkHKEUtGNT4m4pEYhwGivraTGYWKvGoHzA/PZhmjqRRx6Rlt5gd+2vJRHeYlGRuBuAJSP1cbwhoetVW3A2tgWc27Q0XIutsKGxWIL8RXS8aeftHx4b8E0aykLh6fkrJuxdpKqTZiWWxmPgU/Cln9y1f4aSqFh9b1iXl9VMSCGdpAQkQQKQ+KRcTcvZKzY4txkW53EpCBFgtQD/2ZISFFGRhPk8F920iTDYGg4R+8cnYvukCAG2ZJhHrV9/v9KCBlJpVoqtFQRceUj4i7YqD0WhMd5i3AeYQNCyx2SzvuAtyIqAAuJFLFK1hKkclgVkEahnAPnEX0gpBa0QeJRWuJU3HSAiz5BIkGHeFWFiFySzhjW9Zrl6pKLi5cs1xe0XYuSKZNyyricMCoKdBIVFZq+YlOtWK4WNM0GY9qr+xrAOxGfe3CREO2j/FF00IW+35qTmVjFEThfrjDOkQTAC5wCFQJy0BCTYlulhyGpWLZy/G8TfwDPZWgFuMCozHi5vmQ+PWI8znh9uqJrDQ8++iZnixoRoJAJZ6evCVrz3t13UW7B3//N33BydIPJZJ/Hzz7D15757fvcun+bs6+e0i4rNjrDHR8xHs2QyYLbRzc5O31JvVzRp4rOO2yzYn9+wHsfvseL5y+Yzyc0mw3zMqfrWyQB27fkOkEoydnFGR+Ge2jRY4Xm4GDG69cvKLMc27eMRhP+/C//Ay+ePuWLzz7j+dOHnJ2tWF1u+PjTb3BwOKft1iCgbmsODg95PzgOjo7pnUEnGqRHJZokjcJ1m+WKUZ7z0YffZLneYKoWpQzBGoSUPHr4mMOTE+7evY2QkraOQ76iGPHd730PIQTWdtjO0PUN52fncTcZzPCmDNFlTzis7XHWkihNmmTMDk/IUkVoLb2zbBYNzgYSrRiVI/KiwHmH0lFIsxjPyIoJ23lLiMAq8nKMdylS+EEhOSCMo7U1o70xo2lOORtRTBO62rFZGcbzjICnbzxaaw4fzCOaaiOpLmoe/rBi7/YYERIuntcEH4+rFoZipHA9mDrQN4G949g3TrVhNNJUvWZ13jLaDxSpYDzaIytKRuM5OslRMpAmAi1S/CSPUhp+gE96gel6TG/pncciUCpFyZh8vIuWtDrROzJg1LIyKBEdO9FxgXLBIYMCR7Q2cICMkG2Vxl1x9E8fPDt8XBadEIOU1DAaDbHlFZwcIOSxetnKv4Qg8MPCGyHQkrMzyST1vBaB9Sbh2YWm7uHTHyhmJ7Ha8Ntd+nYIK67gxzto8E7LS/LmchF3/ANcOGYEtl4fkdVOhE+/0dpTw3RCiVixyDeTzXZaL8SAPNqSjVSsioIEEfDCgzdxcQxxoSTIK1jxQC7dTf//hZAqggmUjolF6eg4ao3DG0fQGqf8FuC8g1pHRF28RMFdgS3iY2l0kMggCYNMSkAwuCsgnYikziS+BrzwCG8gaJAOL+OgXA4qGi44ur6jqtes1wsuF2csVmc0bYVSKeWopCxKMq0J3tN2FX3fsWnWbKoVVdVgTAdBxXslB2WHiMIY2owDFdUZehM5WX3fY0yUm0FEd89ltabvO0bb+VrYMoogoBngjAQ3kGq9fVvNSuAPlNyXUrC6XFGmY+7evE1v4OL1GcYIxrMpaSIZTwUvn7/iK9MDnnExo1pdsrhcI7IR58sN6/Ul+WhC5yNr2lvHh598xOoffhbJRQju3n6PF6dnvD4/RaaKG0cnZJMR98f3WZ1foFSB7QJ3b9/mSfMFKYKuMyw3DdPJGCsNUmtGRcFyXfHo4WP2xilHR4e89+Ae//CjH2PWFVrkjIsRjz97RJol/PV/+s/89P/+BCces6ksPgikSHEukKQKF6J3x/HhMS4EXr98xXx/hvSO4AWd9UzmByilMG2HLlLGckyWZvzyn37OO/c/YH9/n8tFTdX2sNhw48YBWqc0VY0UlpcvnpClBdPZBO8N40nOepNhW0NZTjDW0tZrTF8jZE6SapIsjW9xqSj39+ibDe3FCjf1tJsOqRXpaMR4f580LzAOQloj25bg4fJiFeW/2wqCZzTJGU/3kFLS1hvqehPd9Hwb3S/zgunxHmkIaC2RZcDlEu881WXHr/7uKUIYju+cUE4jPFigcLbi4U9fEazl9v05xgouXm6AuGPSCSRFghA9N96ZsXzVM7+dIZVic1pR7FsyBJlOSdKSfDQmTRO0DGRakCgAjyw0pcpwXkb9JpVjrKCpa9q+pxuQOFKA8wJjWjyGIhuTJHnUu/IOWeQgAlondCLnrF6zK0fkwGh2CgHIMHAWQlykPTJyNeJGGAQ4MZACtwN2HwYCq9/Nc/yuNRMGhntAhLjYfP444/Mnt3aMdacFk5HC+X5n8hgtg6OUy27oIraLx5BowjDAF8MMSIjhtGICVGFgx3OVXLYRhr/dQZCJSC8lo8yLkAIZ2F3fILfVURxobAmZYpu8RJSakULgnRy4Rj4myi08O8RjCP8CHpqYWNQ2sShFonTstsiY+LyLMxcvweKQgThHGyrFq+s8XCK5lYgJV7DpHcdm8JZxYtcixSlCGgjKo3qD0FGKpQ8BKTMEEucZ+HMbNpsFy9WCxeKMqqoI3pMlGiUEzhma3mOco+lquram63r6vqfvYwtVythqi3fJx3suQCtJmmqUijMq7/SuUs7SjCJTaCXRUrM2Patmw3johkkfIJEorWMyJoCPZFYfTLzqOwjhW+SMtzoKrm4uDoXEW0fTNFhrGE/mpLnmxfPnjCZ7KK0o5zMSrSjShB/+7f/mq1cvmB7d5Jv37/Py9BlfPHlIOkrY259h+pa9ozl337/Nq7OKh48ecfr0KTOdxYXUePbKlL43fPrRt9ksLnjy+AWLxQWn1Yp20yCSnIvVGhcEzknmewc8e33KnXv30EFT95bzxZKDkxucn59z6/iY0/4rNk1F8/Q5l+eK/Zs3eV/lfOePv49B8+rVBUme45H0VYN1CpmNaDYVykco4mQ8pRxNMD7Kj0gpWSwW1FWFJKKOxtMxXzx6TJKP+Icf/5g///M/4+adGzx/eRYRNkLRNVEMbyQycqWYjnLyNBlYwwLn4PHTGp2kg5+2I80SprMZSZLRWhd71lKSlSmL5z/HuAB6D99JdJqj8wmegqoVVFVFtV5jux5vevquxZqIaMmLFJmmyM7hg6WpeowRdD10nY1vkucX3Hz/mLTY+qDHCqBatNQrw95BSjkf0/ees+cLsonCmo7NqkUqmMzHCCXIM8e7n044/6ph8bojGxUsX3Xs3xizeN2jM4NKC0wryUYC2SgKkYFM8ErjQuSZKBXVmvM0i0ijIqcoJ3gnqasFvRUY2yNEXMSkUKAFKklIlIjtTakI3mK6jkR50iSPqDId9ekWtb9izjPAagU4L6Nn0MCnUDssVNyFsx2iy5hUBGJg4YfBsnfokfvt44uBUBnVr7czEAEENcjECIFQCg2IoPE2EJwlBDso4r7ZBhvQYAw980HOnzcW+BixTx/EAD4Yqg0/JJjgA1LJ4TlG101UNDgLQcaHHVBnXoVht8swJBZsz0Rsz13EdmIcFA9+Ldvc56OPk3URWOKdBemHNmScy4Q31nylY/tOa0WiNUootJZxIVUCPbA9nQtYKwbHz9h3VAPiQQkR79EwPwpyGN5IufssRETEhWGgE5AEF7XQ4pzGRUFdZfC+JsgeFyQuCIx1dF1L09RU1ZLVakVdb/DWoQd74t70ccNBoO0auq7GGT/oqzmkgETHc0NKBvAeDBvLOB8aroGUyK2oZqHQOiXTadwESEWQCWerc06OB4tmxK5luBM3DZ6AJXgTN0fOX83rfk+I8JZH/rf/+l9IVMxackBRvLmHkEpijEWpiHRRSqKTyPCu1mta00e3tbzA+sibCC5ECQViy8Z7j+ntzrRpACjgfTQEMs4yKkdopWKJ5wN2+Bz7s1ECWyKuji8KCHHG4KwhzzJc8CgRVUh7YxFAmiZ4IRiNRujhd03TopQiL3LauiHN4g46GkLFN4iUMvop7PqScSfnvIvXwju0Tqg2NWmW0rYNRVGwdXiTw3BR7HDkUVZfKYUY4JLORYkXa6M/gx9uesTu60F5I7ZEhBSkWYKzTbxHQeD64XkNu8oQoibUFnZK2LY9w5Agr3rlW1QRbDe8sT0hhaQo0/giF0OrhdjWcTZcHSejBphKomx531q8i+2LSHIbXjtdFBmMfX6QWmCNJ0mHUj1soatbscMrvgUIpAg7yCvD8DhqRTHIqsSefhhkNbYDb7mbCQxaVsEPkOHtzlzudufOB9q+j2ZyvNGh2S2c22/f+P6Nr998v+zedG988VvvxC2ESlz95QADeOOg+GhSSIrRiCSNr8+wg2D9zj+5+kkIb4xcfrcSEG/8fPfbN3+2G/xvv/lt5v72eBF440L9f2I3E/rtn4Xtfdrdr0BdVbw+O48ckH/2XMW27TZ8vurIXSHbEFfHvfm0tufxW/dMvPGzf3b8v3hO2wpxR/vZVmzbYXjYvQ79IM/yW+3KN7hHu7bdG8CL7Rdvthp/+3Uldjfrt6/DGx9Dy3N7KyfFmDwrr+73m6e1fQ3+s1nXcrnmv/+P//m75/87l+Nt09B1XMd1XMd1XMdbxttLXF7HdVzHdVzHdbxlXCeX67iO67iO6/ja4zq5XMd1XMd1XMfXHtfJ5Tqu4zqu4zq+9rhOLtdxHddxHdfxtcd1crmO67iO67iOrz2uk8t1XMd1XMd1fO1xnVyu4zqu4zqu42uP6+RyHddxHddxHV97/D+T2GKI6fQ1NgAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 500x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Let’s visualize a few training images so as to understand the data augmentations.\n",
    "# Get a batch of training data\n",
    "inputs, classes = next(iter(dataloaders['train']))\n",
    "\n",
    "# Make a grid from batch\n",
    "out = torchvision.utils.make_grid(inputs)\n",
    "\n",
    "imshow(out, title=[class_names[x] for x in classes])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "#### Set Model Parameters’ `.requires_grad` attribute\n",
    "---\n",
    "* The following helper function sets the `.requires_grad` attribute of the parameters in the model to `False` when we are feature extracting. \n",
    "* By default, when we load a pretrained model all of the parameters have `.requires_grad=True`, which is fine if we are training from scratch or fine-tuning.\n",
    "* However, if we are feature extracting and only want to compute gradients for the newly initialized layer then we want all of the other parameters to not require gradients."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "def set_parameter_requires_grad(model, feature_extracting=False):\n",
    "    # approach 1\n",
    "    if feature_extracting:\n",
    "        # frozen model\n",
    "        model.requires_grad_(False)\n",
    "    else:\n",
    "        # fine-tuning\n",
    "        model.requires_grad_(True)\n",
    "        \n",
    "    # approach 2\n",
    "    if feature_extracting:\n",
    "        # frozen model\n",
    "        for param in model.parameters():\n",
    "            param.requires_grad = False\n",
    "    else:\n",
    "        # fine-tuning\n",
    "        for param in model.parameters():\n",
    "            param.requires_grad = True\n",
    "    # note: you can also mix between frozen layers and trainable layers, but you'll need a custom \n",
    "    # function that loops over the model's layers and you specify which layers are frozen."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "#### Initialize and Reshape the Networks\n",
    "---\n",
    "* Recall, the final layer of a CNN model, which is often an FC layer, has the same number of nodes as the number of output classes in the dataset. \n",
    "* Since all of the following models have been pretrained on ImageNet, they all have output layers of size 1000, one node for each class. \n",
    "* The goal here is to **reshape the last layer to have the same number of inputs as before**, AND to have the **same number of outputs as the number of classes in the dataset**. \n",
    "\n",
    "* When *feature extracting*, we only want to update the parameters of the last layer, or in other words, we only want to update the parameters for the layer(s) we are reshaping. \n",
    "* Therefore, we do not need to compute the gradients of the parameters that we are not changing, so for efficiency we set the `.requires_grad` attribute to `False`. \n",
    "* This is important because by default, this attribute is set to `True`. Then, when we initialize the new layer and by default the new parameters have `.requires_grad=True` so only the new layer’s parameters will be updated. \n",
    "* When we are fine-tuning we can leave all of the `.required_grad`’s set to the default of `True`."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### <img src=\"https://img.icons8.com/nolan/64/download-from-cloud.png\" style=\"height:50px;display:inline\"> Torchvision Pre-Trained Models\n",
    "---\n",
    "* The `torchvision.models` subpackage contains definitions of models for addressing different tasks, including: image classification, pixelwise semantic segmentation, object detection, instance segmentation, person keypoint detection, video classification, and optical flow.\n",
    "* You can see all the available here - <a href=\"https://pytorch.org/vision/stable/models.html\">Models and pre-trained weights</a>.\n",
    "    * In code, you can use <a href=\"https://pytorch.org/vision/stable/models.html#listing-and-retrieving-available-models\">`torchvision.models.list_models`</a> to see a list of all available models.\n",
    "* Examples:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "def initialize_model(model_name, num_classes, feature_extract, use_pretrained=True):\n",
    "    # Initialize these variables which will be set in this if statement. Each of these\n",
    "    #   variables is model specific.\n",
    "    model_ft = None\n",
    "    input_size = 0  # image size, e.g. (3, 224, 224)\n",
    "    # new method from torchvision >= 0.13\n",
    "    weigths = 'DEFAULT' if use_pretrained else None \n",
    "    # to use other checkpoints than the default ones, check the model's available chekpoints here:\n",
    "    # https://pytorch.org/vision/stable/models.html\n",
    "    if model_name == \"resnet\":\n",
    "        \"\"\" Resnet18\n",
    "        \"\"\"\n",
    "        # new method from torchvision >= 0.13\n",
    "        model_ft = models.resnet18(weights=weights)\n",
    "        # old method for toechvision < 0.13\n",
    "        # model_ft = models.resnet18(pretrained=use_pretrained)\n",
    "        \n",
    "        set_parameter_requires_grad(model_ft, feature_extract)\n",
    "        num_ftrs = model_ft.fc.in_features\n",
    "        model_ft.fc = nn.Linear(num_ftrs, num_classes) # replace the last FC layer\n",
    "        input_size = 224\n",
    "\n",
    "    elif model_name == \"alexnet\":\n",
    "        \"\"\" Alexnet\n",
    "        \"\"\"\n",
    "        # new method from torchvision >= 0.13\n",
    "        model_ft = models.alexnet(weights=weights)\n",
    "        # old method for toechvision < 0.13\n",
    "        # model_ft = models.alexnet(pretrained=use_pretrained)\n",
    "        \n",
    "        set_parameter_requires_grad(model_ft, feature_extract)\n",
    "        num_ftrs = model_ft.classifier[6].in_features\n",
    "        model_ft.classifier[6] = nn.Linear(num_ftrs, num_classes)\n",
    "        input_size = 224\n",
    "\n",
    "    elif model_name == \"vgg\":\n",
    "        \"\"\" VGG16\n",
    "        \"\"\"\n",
    "        # new method from torchvision >= 0.13\n",
    "        model_ft = models.vgg16(weights=weights)\n",
    "        # old method for toechvision < 0.13\n",
    "        # model_ft = models.vgg16(pretrained=use_pretrained)\n",
    "        \n",
    "        set_parameter_requires_grad(model_ft, feature_extract)\n",
    "        num_ftrs = model_ft.classifier[6].in_features\n",
    "        model_ft.classifier[6] = nn.Linear(num_ftrs, num_classes)\n",
    "        input_size = 224\n",
    "\n",
    "    elif model_name == \"squeezenet\":\n",
    "        \"\"\" Squeezenet\n",
    "        \"\"\"\n",
    "        # new method from torchvision >= 0.13\n",
    "        model_ft = models.squeezenet1_0(weights=weights)\n",
    "        # old method for torchvision < 0.13\n",
    "        # model_ft = models.squeezenet1_0(pretrained=use_pretrained)\n",
    "        \n",
    "        set_parameter_requires_grad(model_ft, feature_extract)\n",
    "        model_ft.classifier[1] = nn.Conv2d(512, num_classes, kernel_size=(1,1), stride=(1,1))\n",
    "        model_ft.num_classes = num_classes\n",
    "        input_size = 224\n",
    "\n",
    "    elif model_name == \"densenet\":\n",
    "        \"\"\" Densenet\n",
    "        \"\"\"\n",
    "        # new method from torchvision >= 0.13\n",
    "        model_ft = models.densenet121(weights=weights)\n",
    "        # old method for torchvision < 0.13\n",
    "        # model_ft = models.densenet121(pretrained=use_pretrained)\n",
    "        \n",
    "        set_parameter_requires_grad(model_ft, feature_extract)\n",
    "        num_ftrs = model_ft.classifier.in_features\n",
    "        model_ft.classifier = nn.Linear(num_ftrs, num_classes)\n",
    "        input_size = 224\n",
    "\n",
    "    else:\n",
    "        raise NotImplementedError\n",
    "\n",
    "    return model_ft, input_size"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "# Models to choose from [resnet, alexnet, vgg, squeezenet, densenet]\n",
    "model_name = \"vgg\"\n",
    "\n",
    "# Number of classes in the dataset\n",
    "num_classes = 2\n",
    "\n",
    "# Batch size for training (change depending on how much memory you have)\n",
    "batch_size = 8\n",
    "\n",
    "# Number of epochs to train for\n",
    "num_epochs = 15\n",
    "\n",
    "# Flag for feature extracting. When False, we fine-tune the whole model,\n",
    "#   when True we only update the reshaped layer params\n",
    "feature_extract = True"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "VGG(\n",
      "  (features): Sequential(\n",
      "    (0): Conv2d(3, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (1): ReLU(inplace=True)\n",
      "    (2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (3): ReLU(inplace=True)\n",
      "    (4): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (5): Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (6): ReLU(inplace=True)\n",
      "    (7): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (8): ReLU(inplace=True)\n",
      "    (9): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (10): Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (11): ReLU(inplace=True)\n",
      "    (12): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (13): ReLU(inplace=True)\n",
      "    (14): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (15): ReLU(inplace=True)\n",
      "    (16): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (17): Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (18): ReLU(inplace=True)\n",
      "    (19): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (20): ReLU(inplace=True)\n",
      "    (21): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (22): ReLU(inplace=True)\n",
      "    (23): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (24): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (25): ReLU(inplace=True)\n",
      "    (26): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (27): ReLU(inplace=True)\n",
      "    (28): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (29): ReLU(inplace=True)\n",
      "    (30): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "  )\n",
      "  (avgpool): AdaptiveAvgPool2d(output_size=(7, 7))\n",
      "  (classifier): Sequential(\n",
      "    (0): Linear(in_features=25088, out_features=4096, bias=True)\n",
      "    (1): ReLU(inplace=True)\n",
      "    (2): Dropout(p=0.5, inplace=False)\n",
      "    (3): Linear(in_features=4096, out_features=4096, bias=True)\n",
      "    (4): ReLU(inplace=True)\n",
      "    (5): Dropout(p=0.5, inplace=False)\n",
      "    (6): Linear(in_features=4096, out_features=2, bias=True)\n",
      "  )\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "# Initialize the model for this run\n",
    "model_ft, input_size = initialize_model(model_name, num_classes, feature_extract, use_pretrained=True)\n",
    "\n",
    "# Print the model we just instantiated\n",
    "print(model_ft)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Params to learn:\n",
      "\t classifier.6.weight\n",
      "\t classifier.6.bias\n"
     ]
    }
   ],
   "source": [
    "model_ft = model_ft.to(device)\n",
    "\n",
    "# Gather the parameters to be optimized/updated in this run. If we are\n",
    "#  fine-tuning we will be updating all parameters. However, if we are\n",
    "#  doing feature extract method, we will only update the parameters\n",
    "#  that we have just initialized, i.e. the parameters with requires_grad\n",
    "#  is True.\n",
    "params_to_update = model_ft.parameters()\n",
    "print(\"Params to learn:\")\n",
    "if feature_extract:\n",
    "    params_to_update = []  # override the initial list definition above\n",
    "    for name,param in model_ft.named_parameters():\n",
    "        if param.requires_grad == True:\n",
    "            params_to_update.append(param)\n",
    "            print(\"\\t\",name)\n",
    "else:\n",
    "    for name,param in model_ft.named_parameters():\n",
    "        if param.requires_grad == True:\n",
    "            print(\"\\t\",name)\n",
    "\n",
    "# Observe that all parameters are being optimized\n",
    "optimizer_ft = torch.optim.SGD(params_to_update, lr=0.001, momentum=0.9)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "\"\"\"\n",
    "Training function\n",
    "\"\"\"\n",
    "def train_model(model, dataloaders, criterion, optimizer, num_epochs=25):\n",
    "    since = time.time()\n",
    "\n",
    "    val_acc_history = []\n",
    "\n",
    "    best_model_wts = copy.deepcopy(model.state_dict())\n",
    "    best_acc = 0.0\n",
    "\n",
    "    for epoch in range(num_epochs):\n",
    "        print('Epoch {}/{}'.format(epoch, num_epochs - 1))\n",
    "        print('-' * 10)\n",
    "\n",
    "        # Each epoch has a training and validation phase\n",
    "        for phase in ['train', 'val']:\n",
    "            if phase == 'train':\n",
    "                model.train()  # Set model to training mode\n",
    "            else:\n",
    "                model.eval()   # Set model to evaluate mode\n",
    "\n",
    "            running_loss = 0.0\n",
    "            running_corrects = 0\n",
    "\n",
    "            # Iterate over data.\n",
    "            for inputs, labels in dataloaders[phase]:\n",
    "                inputs = inputs.to(device)\n",
    "                labels = labels.to(device)\n",
    "\n",
    "                # forward\n",
    "                # track history if only in train\n",
    "                with torch.set_grad_enabled(phase == 'train'):\n",
    "                    # Get model outputs and calculate loss\n",
    "                    outputs = model(inputs)\n",
    "                    loss = criterion(outputs, labels)\n",
    "\n",
    "                    _, preds = torch.max(outputs, 1)\n",
    "\n",
    "                    # backward + optimize only if in training phase\n",
    "                    if phase == 'train':\n",
    "                        # zero the parameter gradients\n",
    "                        optimizer.zero_grad()\n",
    "                        loss.backward()\n",
    "                        optimizer.step()\n",
    "\n",
    "                # statistics\n",
    "                running_loss += loss.item() * inputs.size(0)\n",
    "                running_corrects += torch.sum(preds == labels.data)\n",
    "\n",
    "            epoch_loss = running_loss / len(dataloaders[phase].dataset)\n",
    "            epoch_acc = running_corrects.double() / len(dataloaders[phase].dataset)\n",
    "\n",
    "            print('{} Loss: {:.4f} Acc: {:.4f}'.format(phase, epoch_loss, epoch_acc))\n",
    "\n",
    "            # deep copy the model\n",
    "            if phase == 'val' and epoch_acc > best_acc:\n",
    "                best_acc = epoch_acc\n",
    "                best_model_wts = copy.deepcopy(model.state_dict())\n",
    "            if phase == 'val':\n",
    "                val_acc_history.append(epoch_acc)\n",
    "\n",
    "        print()\n",
    "\n",
    "    time_elapsed = time.time() - since\n",
    "    print('Training complete in {:.0f}m {:.0f}s'.format(time_elapsed // 60, time_elapsed % 60))\n",
    "    print('Best val Acc: {:4f}'.format(best_acc))\n",
    "\n",
    "    # load best model weights\n",
    "    model.load_state_dict(best_model_wts)\n",
    "    return model, val_acc_history"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 0/14\n",
      "----------\n",
      "train Loss: 0.2460 Acc: 0.9057\n",
      "val Loss: 0.1525 Acc: 0.9542\n",
      "\n",
      "Epoch 1/14\n",
      "----------\n",
      "train Loss: 0.2114 Acc: 0.9139\n",
      "val Loss: 0.1093 Acc: 0.9477\n",
      "\n",
      "Epoch 2/14\n",
      "----------\n",
      "train Loss: 0.1623 Acc: 0.9221\n",
      "val Loss: 0.1752 Acc: 0.9412\n",
      "\n",
      "Epoch 3/14\n",
      "----------\n",
      "train Loss: 0.2186 Acc: 0.9262\n",
      "val Loss: 0.1530 Acc: 0.9477\n",
      "\n",
      "Epoch 4/14\n",
      "----------\n",
      "train Loss: 0.1423 Acc: 0.9467\n",
      "val Loss: 0.1667 Acc: 0.9412\n",
      "\n",
      "Epoch 5/14\n",
      "----------\n",
      "train Loss: 0.1301 Acc: 0.9508\n",
      "val Loss: 0.1108 Acc: 0.9673\n",
      "\n",
      "Epoch 6/14\n",
      "----------\n",
      "train Loss: 0.1602 Acc: 0.9303\n",
      "val Loss: 0.1001 Acc: 0.9412\n",
      "\n",
      "Epoch 7/14\n",
      "----------\n",
      "train Loss: 0.1888 Acc: 0.9385\n",
      "val Loss: 0.1254 Acc: 0.9542\n",
      "\n",
      "Epoch 8/14\n",
      "----------\n",
      "train Loss: 0.1432 Acc: 0.9508\n",
      "val Loss: 0.1194 Acc: 0.9608\n",
      "\n",
      "Epoch 9/14\n",
      "----------\n",
      "train Loss: 0.1139 Acc: 0.9631\n",
      "val Loss: 0.1230 Acc: 0.9608\n",
      "\n",
      "Epoch 10/14\n",
      "----------\n",
      "train Loss: 0.2273 Acc: 0.9467\n",
      "val Loss: 0.1154 Acc: 0.9608\n",
      "\n",
      "Epoch 11/14\n",
      "----------\n",
      "train Loss: 0.1507 Acc: 0.9385\n",
      "val Loss: 0.1096 Acc: 0.9608\n",
      "\n",
      "Epoch 12/14\n",
      "----------\n",
      "train Loss: 0.0876 Acc: 0.9590\n",
      "val Loss: 0.1767 Acc: 0.9608\n",
      "\n",
      "Epoch 13/14\n",
      "----------\n",
      "train Loss: 0.3315 Acc: 0.9344\n",
      "val Loss: 0.1371 Acc: 0.9608\n",
      "\n",
      "Epoch 14/14\n",
      "----------\n",
      "train Loss: 0.0936 Acc: 0.9672\n",
      "val Loss: 0.1540 Acc: 0.9673\n",
      "\n",
      "Training complete in 3m 39s\n",
      "Best val Acc: 0.967320\n"
     ]
    }
   ],
   "source": [
    "# Setup the loss fn\n",
    "criterion = nn.CrossEntropyLoss()\n",
    "\n",
    "# Train and evaluate\n",
    "model_ft, hist = train_model(model_ft, dataloaders, criterion, optimizer_ft, num_epochs=num_epochs)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## <img src=\"https://img.icons8.com/pulsar-color/96/split.png\" style=\"height:50px;display:inline\"> LoRA & DoRA - Low Rank Adaptation Fine-Tuning\r\n",
    "\n",
    "---\n",
    "Based on the great blog post \"<a href=\"https://magazine.sebastianraschka.com/p/lora-and-dora-from-scratch\">Improving LoRA: Implementing Weight-Decomposed Low-Rank Adaptation (DoRA) from Scratch\" by Sebastian Raschka</a>."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* Low-rank adaptation (LoRA) is a technique to fine-tune a pre-trained model by adjusting only a small, low-rank subset of the model's parameters.\n",
    "* It allows for efficient fine-tuning of large models on task-specific data, significantly reducing the computational cost and time required for fine-tuning."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### LoRA: Low Rank Adaptation\n",
    "---\n",
    "* Introduced in <a href=\"https://arxiv.org/abs/2106.09685\">\"LoRA: Low-Rank Adaptation of Large Language Models\"</a> by Hu et al.\n",
    "* Given a weight matrix $W \\in \\mathbb{R}^{d_i \\times d_o}$, in regular training or fine-tuning using gradient descent, the weights are updated according to: $$ W_{\\text{updated}} = W + \\Delta W. $$\n",
    "* LoRA proposes an alternative method for this update, by learning a low-rank approximation of $\\Delta W$: $$ \\Delta W \\approx AB, $$ where $A \\in \\mathbb{R}^{d_i \\times r}$, $B \\in \\mathbb{R}^{r \\times d_o}$ and $r << d_i, d_o$ is the *rank* (very similar to SVD--Singular Value Decomposition).\n",
    "* In practice, we usually keep $W$ frozen and only learn $A$ and $B$.\n",
    "* LoRA can potenially save a lot of memory; for example, if $d_i = d_o = 1000$, then a full fine-tuning requires training $1000 \\times 1000 = 1,000,000$ parameters. However, if we consider $r=2$ then $A$ and $B$ have $1000 \\times 2$ parameters each, resulting in only $4,000$ parameters ($250$ fewer parameters!).\n",
    "* <a href=\"https://github.com/microsoft/LoRA\">PyTorch Code</a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<center><img src='./assets/lora.jpeg' style=\"height:300px\"></center>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### DoRA: Weight-Decomposed Low Rank Adaptation\n",
    "---\n",
    "* Introduced in <a href=\"https://arxiv.org/abs/2402.09353\">DoRA: Weight-Decomposed Low-Rank Adaptation\"</a> by Liu et al.\n",
    "* DoRA build upon LoRA and extends it by considering the principle that any vector can be represented as the product of its **magnitude** (a scalar value indicating its length) and its **direction** (a unit vector indicating its orientation in space).\n",
    "* The paper found that LoRA either increases or decreases magnitude and direction updates proportionally but seems to lack the capability to make only subtle directional changes as found in full fine-tuning. Hence, they propose the decoupling of magnitude and directional components.\n",
    "* In DoRA, we apply the decomposition into magnitude and directional components to a whole pre-trained weight matrix $W$ instead of a vector, where each column (vector) of the weight matrix corresponds to the weights connecting all inputs to a particular output neuron.\n",
    "* In practice, DoRA takes the directional matrix $V=\\frac{W + AB}{|| W+AB||}$ and applies standard LoRA, for instance: $$ W_{\\text{updated}} = m \\cdot \\frac{W + AB}{|| W+AB||}. $$\n",
    "* <a href=\"https://github.com/catid/dora\">PyTorch Code</a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<center><img src='./assets/dora.png' style=\"height:500px\"></center>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<center><img src='./assets/lora_vs_dora.png' style=\"height:200px\"></center>\n",
    "<center><img src='./assets/lora_vs_dora_2.png' style=\"height:250px\"></center>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### <img src=\"https://img.icons8.com/cotton/64/000000/olympic-torch.png\" style=\"height:50px;display:inline\"> Low-Rank Adaptation in PyTorch\n",
    "---"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "# example\n",
    "class LowRankLayer(nn.Module):\n",
    "    def __init__(self, linear, rank, alpha, use_dora=True):\n",
    "        super().__init__()\n",
    "        # rank: controls the inner dimension of the matrices A and B; controls the number of additional parameters introduced by LoRA,\n",
    "        # a key factor in determining the balance between model adaptability and parameter efficiency.\n",
    "        # alpha: a scaling hyper-parameter applied to the output of the low-rank adaptation, \n",
    "        # controls the extent to which the adapted layer's output is allowed to influence the original output of the layer being adapted. \n",
    "        \n",
    "        self.use_dora = use_dora\n",
    "        self.rank = rank  # low-rank\n",
    "        self.alpha = alpha  # scaling hyper-parameter\n",
    "        self.linear = linear\n",
    "        self.in_dim = linear.in_features\n",
    "        self.out_dim = linear.out_features\n",
    "\n",
    "        # weights\n",
    "        std_dev = 1 / torch.sqrt(torch.tensor(self.rank).float())\n",
    "        self.A = nn.Parameter(torch.randn(self.in_dim, self.rank) * std_dev)\n",
    "        self.B = nn.Parameter(torch.zeros(self.rank, self.out_dim))\n",
    "        \n",
    "        if self.use_dora:\n",
    "            self.m = nn.Parameter(\n",
    "                self.linear.weight.norm(p=2, dim=0, keepdim=True))\n",
    "        else:\n",
    "            self.m = None\n",
    "\n",
    "    def forward(self, x):\n",
    "        lora = self.A @ self.B  # combine LoRA matrices\n",
    "        if self.use_dora:\n",
    "            numerator = self.linear.weight + self.alpha * lora.T\n",
    "            denominator = numerator.norm(p=2, dim=0, keepdim=True)\n",
    "            directional_component = numerator / denominator\n",
    "            new_weight = self.m * directional_component\n",
    "            return F.linear(x, new_weight, self.linear.bias)\n",
    "        else:\n",
    "            # combine LoRA with orig. weights\n",
    "            combined_weight = self.linear.weight + self.alpha * lora.T \n",
    "            return F.linear(x, combined_weight, self.linear.bias)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "VGG(\n",
      "  (features): Sequential(\n",
      "    (0): Conv2d(3, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (1): ReLU(inplace=True)\n",
      "    (2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (3): ReLU(inplace=True)\n",
      "    (4): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (5): Conv2d(64, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (6): ReLU(inplace=True)\n",
      "    (7): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (8): ReLU(inplace=True)\n",
      "    (9): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (10): Conv2d(128, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (11): ReLU(inplace=True)\n",
      "    (12): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (13): ReLU(inplace=True)\n",
      "    (14): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (15): ReLU(inplace=True)\n",
      "    (16): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (17): Conv2d(256, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (18): ReLU(inplace=True)\n",
      "    (19): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (20): ReLU(inplace=True)\n",
      "    (21): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (22): ReLU(inplace=True)\n",
      "    (23): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "    (24): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (25): ReLU(inplace=True)\n",
      "    (26): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (27): ReLU(inplace=True)\n",
      "    (28): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "    (29): ReLU(inplace=True)\n",
      "    (30): MaxPool2d(kernel_size=2, stride=2, padding=0, dilation=1, ceil_mode=False)\n",
      "  )\n",
      "  (avgpool): AdaptiveAvgPool2d(output_size=(7, 7))\n",
      "  (classifier): Sequential(\n",
      "    (0): Linear(in_features=25088, out_features=4096, bias=True)\n",
      "    (1): ReLU(inplace=True)\n",
      "    (2): Dropout(p=0.5, inplace=False)\n",
      "    (3): LowRankLayer(\n",
      "      (linear): Linear(in_features=4096, out_features=4096, bias=True)\n",
      "    )\n",
      "    (4): ReLU(inplace=True)\n",
      "    (5): Dropout(p=0.5, inplace=False)\n",
      "    (6): Linear(in_features=4096, out_features=2, bias=True)\n",
      "  )\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "# let's fine-tune vgg's penultimate fc layer\n",
    "weigths = 'DEFAULT'\n",
    "model_ft = models.vgg16(weights=weigths)\n",
    "# turn off gradients\n",
    "model_ft.requires_grad_(False)\n",
    "# replace the penultimate fc layer (a 4096 x 4096 weight matrix) with low-rank adaptation\n",
    "rank = 4\n",
    "alpha = 8\n",
    "model_ft.classifier[3] = LowRankLayer(model_ft.classifier[3], rank, alpha, use_dora=True)\n",
    "# change the last linear layer to the correct number of classes\n",
    "num_classes = 2\n",
    "num_ftrs = model_ft.classifier[6].in_features\n",
    "model_ft.classifier[6] = nn.Linear(num_ftrs, num_classes)\n",
    "input_size = 224\n",
    "\n",
    "# Print the model we just instantiated\n",
    "print(model_ft)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Params to learn:\n",
      "\t classifier.3.A\n",
      "\t classifier.3.B\n",
      "\t classifier.3.m\n",
      "\t classifier.6.weight\n",
      "\t classifier.6.bias\n"
     ]
    }
   ],
   "source": [
    "# print and collect learnable paramters\n",
    "print(\"Params to learn:\")\n",
    "params_to_update = []  # override the initial list definition above\n",
    "for name,param in model_ft.named_parameters():\n",
    "    if param.requires_grad == True:\n",
    "        params_to_update.append(param)\n",
    "        print(\"\\t\",name)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 0/14\n",
      "----------\n",
      "train Loss: 0.4624 Acc: 0.7541\n",
      "val Loss: 0.2203 Acc: 0.9477\n",
      "\n",
      "Epoch 1/14\n",
      "----------\n",
      "train Loss: 0.2490 Acc: 0.9139\n",
      "val Loss: 0.1516 Acc: 0.9542\n",
      "\n",
      "Epoch 2/14\n",
      "----------\n",
      "train Loss: 0.2523 Acc: 0.9098\n",
      "val Loss: 0.1460 Acc: 0.9804\n",
      "\n",
      "Epoch 3/14\n",
      "----------\n",
      "train Loss: 0.1845 Acc: 0.9344\n",
      "val Loss: 0.1168 Acc: 0.9739\n",
      "\n",
      "Epoch 4/14\n",
      "----------\n",
      "train Loss: 0.1578 Acc: 0.9508\n",
      "val Loss: 0.1487 Acc: 0.9542\n",
      "\n",
      "Epoch 5/14\n",
      "----------\n",
      "train Loss: 0.1988 Acc: 0.9426\n",
      "val Loss: 0.1284 Acc: 0.9673\n",
      "\n",
      "Epoch 6/14\n",
      "----------\n",
      "train Loss: 0.2539 Acc: 0.8934\n",
      "val Loss: 0.1873 Acc: 0.9477\n",
      "\n",
      "Epoch 7/14\n",
      "----------\n",
      "train Loss: 0.2276 Acc: 0.8934\n",
      "val Loss: 0.1416 Acc: 0.9542\n",
      "\n",
      "Epoch 8/14\n",
      "----------\n",
      "train Loss: 0.2035 Acc: 0.9221\n",
      "val Loss: 0.1160 Acc: 0.9608\n",
      "\n",
      "Epoch 9/14\n",
      "----------\n",
      "train Loss: 0.2706 Acc: 0.9016\n",
      "val Loss: 0.1437 Acc: 0.9412\n",
      "\n",
      "Epoch 10/14\n",
      "----------\n",
      "train Loss: 0.1976 Acc: 0.9262\n",
      "val Loss: 0.1392 Acc: 0.9673\n",
      "\n",
      "Epoch 11/14\n",
      "----------\n",
      "train Loss: 0.1820 Acc: 0.9303\n",
      "val Loss: 0.1148 Acc: 0.9608\n",
      "\n",
      "Epoch 12/14\n",
      "----------\n",
      "train Loss: 0.1943 Acc: 0.9180\n",
      "val Loss: 0.1133 Acc: 0.9542\n",
      "\n",
      "Epoch 13/14\n",
      "----------\n",
      "train Loss: 0.1434 Acc: 0.9426\n",
      "val Loss: 0.1223 Acc: 0.9542\n",
      "\n",
      "Epoch 14/14\n",
      "----------\n",
      "train Loss: 0.2694 Acc: 0.9016\n",
      "val Loss: 0.1216 Acc: 0.9477\n",
      "\n",
      "Training complete in 6m 23s\n",
      "Best val Acc: 0.980392\n"
     ]
    }
   ],
   "source": [
    "# hyper-parameters\n",
    "device = torch.device(\"cuda:0\" if torch.cuda.is_available() else \"cpu\")\n",
    "batch_size = 8\n",
    "num_epochs = 15\n",
    "model_ft = model_ft.to(device)\n",
    "# optimizer\n",
    "optimizer_ft = torch.optim.Adam(params_to_update, lr=0.0002)\n",
    "# loss function\n",
    "criterion = nn.CrossEntropyLoss()\n",
    "# train\n",
    "# Train and evaluate\n",
    "model_ft, hist = train_model(model_ft, dataloaders, criterion, optimizer_ft, num_epochs=num_epochs)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### <img src=\"https://img.icons8.com/cute-clipart/64/000000/language.png\" style=\"height:50px;display:inline\"> Pre-training for Natural Language Processing\n",
    "---\n",
    "* One of the greatest challenges in NLP is the shortage of labeled training data.\n",
    "* Because NLP is a diversified field with many distinct tasks, most task-specific datasets contain only a few thousand or a few hundred thousand human-labeled training examples. \n",
    "* As big companies like Google and OpenAI have shown, **modern deep learning-based NLP models see benefits from much larger amounts of data, improving when trained on millions, or billions, of annotated training examples**.\n",
    "* Large pre-trained models on the enormous amount of **unannotated text on the web** can then be fine-tuned on small-data NLP tasks like question answering and sentiment analysis, resulting in substantial accuracy improvements compared to training on these datasets from scratch.\n",
    "* **Bidirectional Encoder Representations from Transformers (BERT), Google** - a Transformer-based machine learning technique for natural language processing (NLP) pre-training developed by Google. The idea is to mask certain words and then try to predict them. The original English-language BERT model comes with two pre-trained general types:\n",
    "    * (1) the $BERT_{BASE}$ model, a 12-layer, 768-hidden, 12-heads, 110M parameter neural network architecture.\n",
    "    * (2) the $BERT_{LARGE}$ model, a 24-layer, 1024-hidden, 16-heads, 340M parameter neural network architecture. \n",
    "    * Both of which were trained on the BooksCorpus dataset with 800M words, and a version of the English Wikipedia with 2,500M words.\n",
    "* BERT uses the straightforward technique of masking out some of the words in the input and then condition each word bidirectionally to predict the masked words and also learns to model relationships between sentences by pre-training on a very simple task that can be generated from any text corpus: Given two sentences A and B, is B the actual next sentence that comes after A in the corpus, or just a random sentence? "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<center><img src=\"./assets/bert_1.png\" style=\"height:50px\"></center>\n",
    "\n",
    "<center><img src=\"./assets/bert_ill_1.png\" style=\"height:350px\"></center>\n",
    "\n",
    "* <a href=\"https://ai.googleblog.com/2018/11/open-sourcing-bert-state-of-art-pre.html\">Image Source 1</a>, <a href=\"http://jalammar.github.io/illustrated-bert/\">Image Source 2</a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<center><img src=\"./assets/bert_2.png\" style=\"height:100px\"></center>\n",
    "<center><img src=\"./assets/bert_ill_2.png\" style=\"height:350px\"></center>\n",
    "\n",
    "* <a href=\"https://ai.googleblog.com/2018/11/open-sourcing-bert-state-of-art-pre.html\">Image Source 1</a>, <a href=\"http://jalammar.github.io/illustrated-bert/\">Image Source 2</a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "#### Pre-trained Models for NLP in PyTorch\n",
    "---\n",
    "* <a href=\"https://huggingface.co/\">HuggingFace</a> is a copmpany that is dedicated to publishing all of the available pretrained models and it works with PyTorch as well - <a href=\"https://github.com/huggingface/transformers\">HuggingFace Transformers</a>\n",
    "* <a href=\"https://pytorch.org/hub/huggingface_pytorch-transformers/\">Examples with PyTorch</a>\n",
    "* <a href=\"https://huggingface.co/docs/transformers/training\">Tutorial: Fine-tune Transformers for NLP Tasks with PyTorch and HuggingFace</a>."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<center><img src=\"./assets/huggingface.PNG\" style=\"height:500px\"></center>"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### <img src=\"./assets/torch_icon.jfif\" style=\"height:50px;display:inline\"> TorchTune Library - Fine-tuning and Experimenting with LLMs\n",
    "---\n",
    "* <a href=\"https://github.com/pytorch/torchtune\">TorchTune</a> is a native-Pytorch library for easily authoring, fine-tuning and experimenting with LLMs.\n",
    "* Native-PyTorch implementations of popular LLMs with support for checkpoints in various formats, including checkpoints in HuggingFace format.\n",
    "* <a href=\"https://github.com/pytorch/torchtune\">TorchTune on GitHub</a>\n",
    "\n",
    "<center><img src=\"./assets/torchtune_gh.PNG\" style=\"height:200px\"></center>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "## <img src=\"https://img.icons8.com/color/96/000000/self-esteem.png\" style=\"height:50px;display:inline\"> Representation and Self-Supervised Learning\n",
    "---\n",
    "* How do we learn rich and useful features from raw unlabeled data that can be useful for several downstream tasks (e.g., classification, reinforcement learning...)?\n",
    "* What are the various general tasks that can be used to learn representations from unlabeled data?\n",
    "* **The way we represent the data has a great impact on the performance and compelxity**.\n",
    "* Such representations can be learned in typical *unsupervised* settings, or in a *self-supervised* manner."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<center><img src=\"./assets/rep_data.PNG\" style=\"height:250px\"></center>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "* **Deep Unsupervised Learning** - learn representations without lables, subset of deep learning, which is a subset of representation learning, which is a subset of machine learning.\n",
    "* **Self-supervised Learning** - often used interchangeably with unsupervised learning. Self-supervised: **create your own supervision through pretext tasks**."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### <img src=\"https://img.icons8.com/color/96/000000/code.png\" style=\"height:50px;display:inline\"> Deep Unsupervised Learning - Deep Autoencoders\n",
    "---\n",
    "* Most of the natural data is high-dimensional, such as images. Consider the MNIST (hand-written digits) dataset, where each image has $28x28=784$ pixels, which means it can be represented by a vector of length 784. \n",
    "    * But do we really need 784 values to represent a digit? The answer is probably no. We believe that the data lies on a low-dimensional space which is enough to describe the observasions. In the case of MNIST, we can choose to represent digits as one-hot vectors, which means we only need 10 dimensions. So we can **encode** high-dimensional observations in a low-dimensional space.\n",
    "    * But how can we learn meaningful low-deminsional representations? The general idea is to reconstruct or, **decode** the low-dimensional representation to the high-dimensional reperesentation, and use the reconstruction error to find the best representations (using the gradients of the error) . This is the core idea behind **autoencoders**.\n",
    "    * **Autoencoders** - models which take data as input and discover some latent state representation of that data.The input data is converted into an encoding vector where each dimension represents some learned attribute about the data. The most important detail to grasp here is that our encoder network is outputting a single value for each encoding dimension. The decoder network then subsequently takes these values and attempts to recreate the original input. Autoencoders have **three parts**: an encoder, a decoder, and a 'loss' function that maps one to the other. For the simplest autoencoders - the sort that compress and then reconstruct the original inputs from the compressed representation - we can think of the 'loss' as describing the amount of information lost in the process of reconstruction.\n",
    "        * Illustration: <center><img src=\"./assets/tut_xv_autoencoder.png\" style=\"height:250px\"></center>\n",
    "        \n",
    "* The basic architecture of an autoencoder:\n",
    "<center><img src=\"./assets/tut_xiv_autoencoder.png\" style=\"height:250px\"></center>\n",
    "    \n",
    "    * Image from <a href=\"https://towardsdatascience.com/applied-deep-learning-part-3-autoencoders-1c083af4d798\">Applied Deep Learning</a> by Arden Dertat\n",
    "    \n",
    "Let's implement it in PyTorch using what we have learnt so far!"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "class AutoEncoder(nn.Module):\n",
    "    \n",
    "    def __init__(self, input_dim=28*28, hidden_dim=256, latent_dim=10):\n",
    "        super(AutoEncoder, self).__init__()\n",
    "        \n",
    "        self.input_dim = input_dim\n",
    "        self.hidden_dim = hidden_dim\n",
    "        self.latent_dim = latent_dim\n",
    "        \n",
    "        # define the encoder\n",
    "        self.encoder = nn.Sequential(nn.Linear(self.input_dim, self.hidden_dim),\n",
    "                                     nn.ReLU(), nn.Linear(self.hidden_dim, self.hidden_dim),\n",
    "                                     nn.ReLU(),\n",
    "                                     nn.Linear(self.hidden_dim, self.latent_dim)\n",
    "                                    )\n",
    "        \n",
    "        # define decoder\n",
    "        self.decoder = nn.Sequential(nn.Linear(self.latent_dim, self.hidden_dim),\n",
    "                                     nn.ReLU(),\n",
    "                                     nn.Linear(self.hidden_dim, self.hidden_dim),\n",
    "                                     nn.ReLU(),\n",
    "                                     nn.Linear(self.hidden_dim, self.input_dim),\n",
    "                                     nn.Sigmoid())\n",
    "        \n",
    "    def forward(self,x):\n",
    "        x = self.encoder(x)\n",
    "        x = self.decoder(x)\n",
    "        return x\n",
    "    \n",
    "    def get_latent_rep(self, x):\n",
    "        return self.encoder(x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "# hyper-parameters:\n",
    "num_epochs = 5\n",
    "learning_rate = 0.001\n",
    "\n",
    "# Device configuration, as before\n",
    "device = torch.device('cuda:0' if torch.cuda.is_available() else 'cpu')\n",
    "\n",
    "# create model, send it to device\n",
    "model = AutoEncoder(input_dim=28*28, hidden_dim=128, latent_dim=10).to(device)\n",
    "\n",
    "# Loss and optimizer\n",
    "criterion = nn.BCELoss()  # binary cross entropy, as pixels are in [0,1], can also use MSE\n",
    "optimizer = torch.optim.Adam(model.parameters(), lr=learning_rate)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "slideshow": {
     "slide_type": "fragment"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch [1/5], Step [100/468], Loss: 0.3832\n",
      "Epoch [1/5], Step [200/468], Loss: 0.3317\n",
      "Epoch [1/5], Step [300/468], Loss: 0.3360\n",
      "Epoch [1/5], Step [400/468], Loss: 0.3037\n",
      "Epoch [2/5], Step [100/468], Loss: 0.3093\n",
      "Epoch [2/5], Step [200/468], Loss: 0.2943\n",
      "Epoch [2/5], Step [300/468], Loss: 0.3223\n",
      "Epoch [2/5], Step [400/468], Loss: 0.3077\n",
      "Epoch [3/5], Step [100/468], Loss: 0.2953\n",
      "Epoch [3/5], Step [200/468], Loss: 0.3018\n",
      "Epoch [3/5], Step [300/468], Loss: 0.3083\n",
      "Epoch [3/5], Step [400/468], Loss: 0.3133\n",
      "Epoch [4/5], Step [100/468], Loss: 0.3082\n",
      "Epoch [4/5], Step [200/468], Loss: 0.3007\n",
      "Epoch [4/5], Step [300/468], Loss: 0.2898\n",
      "Epoch [4/5], Step [400/468], Loss: 0.2976\n",
      "Epoch [5/5], Step [100/468], Loss: 0.3038\n",
      "Epoch [5/5], Step [200/468], Loss: 0.2896\n",
      "Epoch [5/5], Step [300/468], Loss: 0.2972\n",
      "Epoch [5/5], Step [400/468], Loss: 0.2990\n"
     ]
    }
   ],
   "source": [
    "# Train the model\n",
    "total_step = len(fmnist_train_loader)\n",
    "for epoch in range(num_epochs):\n",
    "    for i, (images, labels) in enumerate(fmnist_train_loader):\n",
    "        # each i is a batch of 128 samples\n",
    "        images = images.to(device).view(batch_size, -1)\n",
    "        \n",
    "        # Forward pass\n",
    "        outputs = model(images)\n",
    "        loss = criterion(outputs, images)\n",
    "        \n",
    "        # Backward and optimize - ALWAYS IN THIS ORDER!\n",
    "        optimizer.zero_grad()\n",
    "        loss.backward()\n",
    "        optimizer.step()\n",
    "        \n",
    "        if (i + 1) % 100 == 0:\n",
    "            print ('Epoch [{}/{}], Step [{}/{}], Loss: {:.4f}' \n",
    "                   .format(epoch + 1, num_epochs, i + 1, total_step, loss.item()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "shape: \n",
      " torch.Size([128, 1, 28, 28])\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 6 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# let's see some of the reconstructions\n",
    "model.eval()  # put in evaluation mode - no gradients\n",
    "examples = enumerate(fmnist_test_loader)\n",
    "batch_idx, (example_data, example_targets) = next(examples)\n",
    "print(\"shape: \\n\", example_data.shape)\n",
    "fig = plt.figure()\n",
    "for i in range(3):\n",
    "    ax = fig.add_subplot(2,3,i+1)\n",
    "    ax.imshow(example_data[i][0], cmap='gray', interpolation='none')\n",
    "    ax.set_title(\"Ground Truth: {}\".format(example_targets[i]))\n",
    "    ax.set_axis_off()\n",
    "    \n",
    "    ax = fig.add_subplot(2,3,i+4)\n",
    "    recon_img = model(example_data[i][0].view(1, -1).to(device)).data.cpu().numpy().reshape(28, 28)\n",
    "    ax.imshow(recon_img, cmap='gray')\n",
    "    ax.set_title(\"Reconstruction of: {}\".format(example_targets[i]))\n",
    "    ax.set_axis_off()\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [],
   "source": [
    "# let's compare different dimensionality reduction methods\n",
    "n_neighbors = 10\n",
    "n_components = 2\n",
    "n_points= 500\n",
    "\n",
    "fmnist_test_loader = torch.utils.data.DataLoader(dataset=fmnist_test_dataset,\n",
    "                                          batch_size=n_points, \n",
    "                                          shuffle=False)\n",
    "X, labels = next(iter(fmnist_test_loader))\n",
    "latent_X = model.get_latent_rep(X.to(device).view(n_points, -1)).data.cpu().numpy()\n",
    "labels = labels.data.cpu().numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "PCA time: 0.001 sec\n",
      "KPCA time: 0.013 sec\n",
      "LLE time: 0.077 sec\n",
      "Isomap time: 0.12 sec\n",
      "t-SNE time: 2.7 sec\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x576 with 6 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(15,8))\n",
    "\n",
    "# PCA\n",
    "t0 = time.time()\n",
    "x_pca = PCA(n_components).fit_transform(latent_X)\n",
    "t1 = time.time()\n",
    "print(\"PCA time: %.2g sec\" % (t1 - t0))\n",
    "ax = fig.add_subplot(2, 3, 1)\n",
    "ax.scatter(x_pca[:, 0], x_pca[:, 1], c=labels, cmap=plt.cm.Spectral)\n",
    "ax.set_title('PCA')\n",
    "\n",
    "# KPCA\n",
    "t0 = time.time()\n",
    "x_kpca = KernelPCA(n_components, kernel='rbf').fit_transform(latent_X)\n",
    "t1 = time.time()\n",
    "print(\"KPCA time: %.2g sec\" % (t1 - t0))\n",
    "ax = fig.add_subplot(2, 3, 2)\n",
    "ax.scatter(x_kpca[:, 0], x_kpca[:, 1], c=labels, cmap=plt.cm.Spectral)\n",
    "ax.set_title('KernelPCA')\n",
    "\n",
    "# LLE\n",
    "t0 = time.time()\n",
    "x_lle = LocallyLinearEmbedding(n_neighbors, n_components, eigen_solver='auto').fit_transform(latent_X)\n",
    "t1 = time.time()\n",
    "print(\"LLE time: %.2g sec\" % (t1 - t0))\n",
    "ax = fig.add_subplot(2, 3, 3)\n",
    "ax.scatter(x_lle[:, 0], x_lle[:, 1], c=labels, cmap=plt.cm.Spectral)\n",
    "ax.set_title('LLE')\n",
    "\n",
    "# Isomap\n",
    "t0 = time.time()\n",
    "x_isomap = Isomap(n_neighbors, n_components).fit_transform(latent_X)\n",
    "t1 = time.time()\n",
    "print(\"Isomap time: %.2g sec\" % (t1 - t0))\n",
    "ax = fig.add_subplot(2, 3, 4)\n",
    "ax.scatter(x_lle[:, 0], x_lle[:, 1], c=labels, cmap=plt.cm.Spectral)\n",
    "ax.set_title('LLE')\n",
    "\n",
    "# t-SNE\n",
    "t0 = time.time()\n",
    "x_tsne = TSNE(n_components).fit_transform(latent_X)\n",
    "t1 = time.time()\n",
    "print(\"t-SNE time: %.2g sec\" % (t1 - t0))\n",
    "ax = fig.add_subplot(2, 3, 5)\n",
    "scatter = ax.scatter(x_tsne[:, 0], x_tsne[:, 1], c=labels, cmap=plt.cm.Spectral)\n",
    "ax.set_title('t-SNE')\n",
    "\n",
    "bounds = np.linspace(0, 10, 11)\n",
    "cb = plt.colorbar(scatter, spacing='proportional',ticks=bounds)\n",
    "cb.set_label('Classes Colors')\n",
    "\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### <img src=\"https://img.icons8.com/cute-clipart/64/000000/task.png\" style=\"height:50px;display:inline\"> Self-Supervised Learning\n",
    "---\n",
    "* A version of unsupervised learning where **data provides the supervision**.\n",
    "* **Idea**: withhold some part of the data and then task a neural network to predict it from the remaining parts.\n",
    "* It requires us to decide what proxy loss or pretext task the network tries to solve, and depending on the quality of the task, good semantic features can be obtained without actual labels.\n",
    "* For learning good representations, it turns out that there is **a strong signal from a classification objective** (e.g., cross-entropy), guiding us to formulate proxy tasks that involve creating labels in a self-supervised manner. For example, BERT creates labels by masking words.\n",
    "* Advantages over supervised learning:\n",
    "    * Large cost of producing a new dataset for each task (prepare labeling manuals, categories, hiring humans, creating GUIs, storage pipelines, etc).\n",
    "    * Good supervision may not be cheap (e.g., medicine, legal).\n",
    "    * Take advantage of vast amount of unlabeled data on the internet (images, videos, language)."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<center><img src=\"./assets/ss_learning.PNG\" style=\"height:300px\"></center>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### <img src=\"https://img.icons8.com/cute-clipart/64/000000/console.png\" style=\"height:50px;display:inline\"> Self-Supervised Learning Methods\n",
    "---\n",
    "* *Reconstruct from a corrupted (or partial) version*\n",
    "    * Denoising Autoencoders, Masked Autoencoders\n",
    "    * In-painting\n",
    "    * Colorization, Split-Brain Autoencoder\n",
    "* *Visual common sense tasks*\n",
    "    * Relative patch prediction\n",
    "    * Jigsaw puzzles\n",
    "    * Rotation prediction\n",
    "* **Contrastive Learning**\n",
    "    * Word2Vec\n",
    "    * Contrastive Predictive Coding (CPC)\n",
    "    * Instance Discrimination\n",
    "    *  Simple Framework for Contrastive Learning of Visual Representations (SimCLR), Momentum Contrast (MoCo), Bootstrap Your Own Latent (BYOL)\n",
    "    \n",
    "\n",
    "**Code Demos** - <a href=\"https://colab.research.google.com/github/taldatech/ee046211-deep-learning/blob/main/extra/Self_Supervised_Learning_Demos.ipynb\">Self-Supervised Learning Demos</a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### <img src=\"https://img.icons8.com/external-wanicon-lineal-color-wanicon/64/null/external-mask-brazilian-carnival-wanicon-lineal-color-wanicon.png\" style=\"height:50px;display:inline\">  Masked Autoencoders (Vision Transformers)\n",
    "---\n",
    "* While the field of self-supervised has been taken away by storm with the introduction of contrastive methods, the recent developments in Vision Transformers (ViT) have resurrected the simple idea of masked image modeling.\n",
    "* <a href=\"https://arxiv.org/abs/2111.06377\">Masked Autoencoders Are Scalable Vision Learners, He et al. 2021.</a>\n",
    "* **Masked Autoencoders**: during pre-training, **a large random subset of image patches** (e.g., 75%) is masked out. The encoder is applied to the small subset of visible patches. Mask tokens are introduced after the encoder, and the full set of encoded patches and mask tokens is processed by a small decoder that reconstructs the original image in pixels. After pre-training, the decoder is discarded and the encoder is applied to uncorrupted images (full sets of patches) for recognition tasks.\n",
    "* Masked AEs show competative performance to contrastive methods while being much simpler to implement and understand.\n",
    "* Code:\n",
    "    * HuggingFace: <a href=\"https://huggingface.co/docs/transformers/model_doc/vit_mae\">ViTMAE</a>\n",
    "    * GitHub: <a href=\"https://github.com/facebookresearch/mae\">Official PyTorch implementation (FAIR)</a>, <a href=\"https://github.com/EdisonLeeeee/Awesome-Masked-Autoencoders\">Awesome MAE Models</a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<center><img src=\"./assets/masked_ae.PNG\" style=\"height:300px\"></center>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<center><img src=\"./assets/masked_ae_sample.PNG\" style=\"height:220px\"></center>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<center><img src=\"./assets/masked_ae_table.PNG\" style=\"height:150px\"></center>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### <img src=\"https://img.icons8.com/plasticine/100/000000/protect-from-magnetic-field.png\" style=\"height:50px;display:inline\">  Contrastive Learning\n",
    "---\n",
    "* Contrastive learning is an approach to formulate the task of **finding similar and dissimilar things for a ML model (basically what classification does when given labels)**. \n",
    "* Contrastive methods, as the name implies, learn representations by contrasting **positive and negative** examples. \n",
    "* Using this approach, one can train a machine learning model to classify between similar and dissimilar images.\n",
    "* More formally, for any data point $x$, contrastive methods aim to learn an encoder $f$ such that: \n",
    "    * $x^+$ is a data point similar to $x$, referred to as a *positive* sample.\n",
    "    * $x^−$ is a data point dissimilar to $x$, referred to as a *negative* sample.\n",
    "    * The **score function** is a metric that measures the similarity between two features: $$score(f(x), f(x^+))  >>  score(f(x), f(x^-))$$"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<center><img src=\"./assets/contrastive_1.PNG\" style=\"height:150px\"></center>\n",
    "<center><img src=\"./assets/contrastive_puzzle.gif\" style=\"height:200px\"></center>\n",
    "\n",
    "* <a href=\"https://analyticsindiamag.com/contrastive-learning-self-supervised-ml\">Image Source</a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "* The most common loss function to implement the score paradigm is **InfoNCE** loss, which looks similar to softmax.\n",
    "\n",
    "<center><img src=\"./assets/infonce_loss.png\" style=\"height:100px\"></center>\n",
    "\n",
    "* The denominator terms consist of one positive sample, and $N−1$ negative samples.\n",
    "    * But how do we get negative samples if we are in a self-supervised setting? The answer might not surprise you -- usually the negative samples are just all other samples in the batch.\n",
    "* <a href=\"https://github.com/RElbers/info-nce-pytorch\">InfoNCE Loss in PyTorch</a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### <img src=\"https://img.icons8.com/pastel-glyph/64/000000/qr-code--v2.png\" style=\"height:50px;display:inline\">  Contrastive Predictive Coding (CPC)\n",
    "---\n",
    "* <a href=\"https://arxiv.org/abs/1807.03748\">**Contrastive Predictive Coding (CPC)**</a> learns self-supervised representations by **predicting the future** in a learned *latent space* by using powerful autoregressive models. \n",
    "* The model uses a probabilistic contrastive loss which induces the latent space to capture information that is **maximally useful to predict future samples**."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<center><img src=\"./assets/cpc.png\" style=\"height:300px\"></center>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "* First, a non-linear encoder $g_{enc}$ maps the input sequence of observations $x_t$ to a sequence of latent representations $z_t = g_{enc}(x_t)$, potentially with a lower temporal resolution ($g_{enc}$'s architecture usually depends on the data type, e.g., CNN for images).\n",
    "* Next, an autoregressive model $g_{ar}$ summarizes all $z \\leq t$ in the latent space and produces a context latent representation $c_t=g_{ar}(z \\leq t)$.\n",
    "* A density ratio $f$ is modeled which preserves the **mutual information** between $x_{t+k}$ and $c_t$ as follows: $$ f_k(x_{t+k}, c_t) = \\exp(z_{t+k}^T W_k c_t) \\propto \\frac{p(x_{t+k} | c_t)}{p(x_{t+k})} $$\n",
    "    * Note that the density ratio $f$ can be unnormalized (does not have to integrate to 1).\n",
    "    * $W_k$ are learned weights.\n",
    "* Any type of encoder and autoregressive can be used. \n",
    "    * For example: strided convolutional layers with residual blocks and GRUs.\n",
    "* The encoder and autoregressive models are trained to minimize an **InfoNCE** loss.\n",
    "* <a href=\"https://github.com/jefflai108/Contrastive-Predictive-Coding-PyTorch\">PyTorch Code</a>\n",
    "    * <a href=\"https://github.com/rschwarz15/CPCV2-PyTorch\">CPCv2 - PyTorch Code</a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### <img src=\"https://img.icons8.com/nolan/64/collapse-arrow.png\" style=\"height:50px;display:inline\">  Simple Framework for Contrastive Learning of Visual Representations (SimCLR)\n",
    "---\n",
    "* <a href=\"https://arxiv.org/abs/2002.05709\">**Simple Framework for Contrastive Learning of Visual Representations (SimCLR)**</a> is a framework for contrastive learning of *visual* representations. \n",
    "* It learns representations by maximizing agreement between differently augmented views of the same data example via a contrastive loss in the latent space."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<center><img src=\"./assets/simclr.PNG\" style=\"height:300px\"></center>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "* A **stochastic data augmentation module** that transforms any given data example randomly resulting in two correlated views of the same example, denoted $\\tilde{x}_i$ and $\\tilde{x}_j$, which is considered a **positive pair**.\n",
    "* SimCLR sequentially applies three simple augmentations: random cropping followed by resize back to the original size, random color distortions, and random Gaussian blur. The authors find **random crop and color distortion** are crucial to achieve good performance.\n",
    "* A neural network base encoder $f(\\cdot)$ that extracts **representation vectors** from augmented data examples. The framework allows various choices of the network architecture without any constraints. \n",
    "    * For simplicity, ResNet is used to obtain $h_i = f(\\tilde{x}_i)\\in \\mathbb{R}^d$ where $h_i$ is the output after the average pooling layer.\n",
    "* A small neural network projection head $g(\\cdot)$ that maps representations to the space where contrastive loss is applied. \n",
    "* MLP with one hidden layer is used to obtain $z_i=g(h_i)$.\n",
    "* **The authors find it beneficial to define the contrastive loss on $z_i$’s rather than $h_i$’s**."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "* A mini-batch of $N$ examples is randomly sampled and the contrastive prediction task is defined on pairs of augmented examples derived from the mini-batch, resulting in $2N$ data points. \n",
    "* Negative examples are not sampled explicitly. Instead, given a positive pair, the other $2(N-1)$ augmented examples within a mini-batch are treated as negative examples. \n",
    "* A NT-Xent (the normalized temperature-scaled cross entropy loss) loss function is used: $$\\ell_{i,j} = -\\log{\\frac{\\exp(\\text{sim}(z_i, z_j)/\\tau)}{\\sum_{k=1}^{2N}\\mathbb{1}_{[k\\neq i]}\\exp(\\text{sim}(z_i, z_k)/\\tau)}}$$\n",
    "where $\\text{sim}(z_i, z_j) = \\frac{z_i^Tz_j}{\\left\\Vert z_i \\right\\Vert \\left\\Vert z_j \\right\\Vert}$.\n",
    "\n",
    "* <a href=\"https://github.com/sthalles/SimCLR\">PyTorch Code</a>\n",
    "* <a href=\"https://colab.research.google.com/github/taldatech/ee046211-deep-learning/blob/main/extra/Self_Supervised_Learning_Demos.ipynb#scrollTo=YB_cqJagEXbw\">Colab Example</a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<center><img src=\"./assets/simclr_anim.gif\" style=\"height:350px\"></center>\n",
    "\n",
    "* <a href=\"https://ai.googleblog.com/2020/04/advancing-self-supervised-and-semi.html\">Image Source</a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### <img src=\"https://img.icons8.com/officel/80/000000/gyroscope.png\" style=\"height:50px;display:inline\">  Momentum Contrast (MoCo)\n",
    "---\n",
    "* <a href=\"https://arxiv.org/abs/1911.05722\">**Momentum Contrast (MoCo)**</a> is a self-supervised learning algorithm with a contrastive loss.\n",
    "* Contrastive loss methods can be thought of as **building dynamic dictionaries**. \n",
    "* The **\"keys\" (tokens)** in the dictionary are sampled from data (e.g., images or patches) and are represented by an encoder network. \n",
    "* Unsupervised learning trains encoders (by minimizing a contrastive loss) to perform dictionary look-up: **an encoded “query” should be similar to its matching key and dissimilar to others**.\n",
    "* In MoCo, we maintain the dictionary as a queue of data samples: the encoded representations of the current mini-batch are enqueued, and the oldest are dequeued.\n",
    "*  The queue decouples the dictionary size from the mini-batch size, allowing it to be large.\n",
    "* Moreover, as the dictionary keys come from the preceding several mini-batches, a slowly progressing key encoder, implemented as a momentum-based moving average of the query encoder, is proposed to maintain consistency.\n",
    "* <a href=\"https://github.com/facebookresearch/moco\">PyTorch Code (v1, v2)</a>, <a href=\"https://github.com/facebookresearch/moco-v3\">PyTorch Code (v3)</a>\n",
    "    * <a href=\"https://colab.research.google.com/github/facebookresearch/moco/blob/colab-notebook/colab/moco_cifar10_demo.ipynb\">Colab Demo</a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<center><img src=\"./assets/moco.png\" style=\"height:350px\"></center>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### <img src=\"https://img.icons8.com/external-flaticons-lineal-color-flat-icons/64/external-average-sales-flaticons-lineal-color-flat-icons-2.png\" alt=\"external-average-sales-flaticons-lineal-color-flat-icons-2\" style=\"height:50px;display:inline\">  Exponential Moving Average (EMA)\n",
    "---\n",
    "* Generally, moving average is a smoothing technique that is commonly used for reducing the noise and fluctuation from time-series data.\n",
    "\n",
    "* Exponential moving average (EMA) is a method that increases the stability of a model’s convergence and helps it reach a better overall solution by preventing convergence to a local minima.\n",
    "\n",
    "* Instead of using the optimized parameters from the final training iteration (parameter update step) as the final parameters for the model, the exponential moving average of the parameters between the current weights and the post-optimization step weights. To avoid drastic changes in the model’s weights during training, a copy of the current weights is created before updating the model’s weights.\n",
    "\n",
    "* Mathematically, the EMA for the optimized parameters $\\Theta$ is computed as: $$ \\text{EMA}_t = \\begin{cases} \\Theta_1 && \\text{if} & t=1 \\\\ \\alpha \\text{EMA}_{t-1} + (1-\\alpha)\\Theta_t && \\text{else} \\end{cases},$$ where $\\alpha \\in [0,1]$ (usually $[0.9, 0.999]$).\n",
    "\n",
    "* EMA is broadly used in generative models such as Diffusion models and GANs, and more recenetly in contrastive and self-supervised learning methods such as MoCo and BYOL."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# EMA code example\n",
    "# https://github.com/lucidrains/byol-pytorch/blob/master/byol_pytorch/byol_pytorch.py\n",
    "class EMA():\n",
    "    def __init__(self, beta):\n",
    "        super().__init__()\n",
    "        self.beta = beta\n",
    "\n",
    "    def update_average(self, old, new):\n",
    "        if old is None:\n",
    "            return new\n",
    "        return old * self.beta + (1 - self.beta) * new\n",
    "\n",
    "def update_moving_average(ema_updater, ma_model, current_model):\n",
    "    for current_params, ma_params in zip(current_model.parameters(), ma_model.parameters()):\n",
    "        old_weight, up_weight = ma_params.data, current_params.data\n",
    "        ma_params.data = ema_updater.update_average(old_weight, up_weight)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### <img src=\"https://img.icons8.com/officel/80/000000/gyroscope.png\" style=\"height:50px;display:inline\">  Bootstrap Your Own Latent (BYOL)\n",
    "---\n",
    "* **<a href=\"https://arxiv.org/abs/2006.07733\">Bootstrap Your Own Latent (BYOL)</a>** builds on the momentum network concept of MoCo, adding an MLP $q_{\\theta}$ to predict $z'$ from $z$.\n",
    "* Concretely, in BYOL we keep a momentum network copy of the representation (backbone) and projection networks (the bottom part in the figure below).\n",
    "* Rather than using a contrastive loss, BYOL uses the $L_2$  or cosine distance error between the normalized prediction $p$ and target $z'$.\n",
    "    * The negative cosine similarity loss also works well in practice.\n",
    "* BYOL produces two augmented views $v \\triangleq t(x)$ and $v' \\triangleq t'(x)$ from $x$ by applying respectively image augmentations $t \\sim \\mathcal{T}$ and $t' \\sim \\mathcal{T}'$.\n",
    "* BYOL tries to convert both augmenetations of an image into the same representation vector (make $p$ and $z'$ equal). \n",
    "* The **$L_2$ or cosine-distance loss functions do not require negative examples!**\n",
    "* There is an *implicit* contrastive loss, by using **Batch Normalization** in the first layers of the MLPs (MoCo does not require it).\n",
    "* <a href=\"https://github.com/lucidrains/byol-pytorch\">PyTorch Code</a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "<center><img src=\"./assets/byol_arch.PNG\" style=\"height:300px\"></center>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "* One purpose of negative examples in a contrastive loss function is to prevent *mode collapse*. \n",
    "    * An example of mode collapse would be a network that always outputs $[1, 0, 0, 0, …]$ as its projection vector $z$. \n",
    "    * If all projection vectors $z$ are the same, then the network only needs to learn the identity function for $q$ in order to achieve perfect prediction accuracy!\n",
    "* If **batch normalization** is used in the projection layer $g$, the projection output vector $z$ cannot collapse to any singular value like $[1, 0, 0, 0, …]$ because that is exactly what batch normalization prevents. \n",
    "* Regardless of how similar the inputs to the batch normalization layer, the outputs will be redistributed according to the **learned** mean and standard deviation (and scaling-shifting). \n",
    "* Mode collapse is prevented precisely because **all samples in the mini-batch cannot take on the same value after batch normalization**.\n",
    "* Said another way, with batch normalization, BYOL learns by asking, **“how is this image different from the average image?”**.\n",
    "    * The explicit contrastive approach used by SimCLR and MoCo learns by asking, **“what distinguishes these two specific images from each other?”** \n",
    "    * These two approaches seem equivalent, since comparing an image with many other images has the same effect as comparing it to the average of the other images. "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<center><img src=\"./assets/byol_bn.PNG\" style=\"height:300px\"></center>\n",
    "\n",
    "<a href=\"https://generallyintelligent.com/blog/2020-08-24-understanding-self-supervised-contrastive-learning/\">Image Source</a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# BYOL example\n",
    "# https://docs.lightly.ai/self-supervised-learning/examples/byol.html\n",
    "class BYOL(nn.Module):\n",
    "    def __init__(self, backbone):\n",
    "        super().__init__()\n",
    "\n",
    "        self.backbone = backbone  # e.g., resnet\n",
    "        self.projection_head = BYOLProjectionHead(512, 1024, 256)\n",
    "        self.prediction_head = BYOLPredictionHead(256, 1024, 256)\n",
    "\n",
    "        self.backbone_momentum = copy.deepcopy(self.backbone)\n",
    "        self.projection_head_momentum = copy.deepcopy(self.projection_head)\n",
    "\n",
    "        deactivate_requires_grad(self.backbone_momentum)\n",
    "        deactivate_requires_grad(self.projection_head_momentum)\n",
    "\n",
    "    def forward(self, x):\n",
    "        y = self.backbone(x).flatten(start_dim=1)\n",
    "        z = self.projection_head(y)\n",
    "        p = self.prediction_head(z)\n",
    "        return p\n",
    "\n",
    "    def forward_momentum(self, x):\n",
    "        y = self.backbone_momentum(x).flatten(start_dim=1)\n",
    "        z = self.projection_head_momentum(y)\n",
    "        z = z.detach()\n",
    "        return z\n",
    "    \n",
    "# initialization example\n",
    "resnet = torchvision.models.resnet18()\n",
    "backbone = nn.Sequential(*list(resnet.children())[:-1])\n",
    "model = BYOL(backbone)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# calculating the loss example\n",
    "criterion = NegativeCosineSimilarity()  # can also be MSE as in the original paper\n",
    "optimizer = torch.optim.SGD(model.parameters(), lr=0.06)\n",
    "\n",
    "for epoch in range(10):\n",
    "    total_loss = 0\n",
    "    for (x0, x1), _, _ in dataloader:\n",
    "        # call to EMA updater\n",
    "        update_momentum(model.backbone, model.backbone_momentum, m=0.99)\n",
    "        update_momentum(model.projection_head, model.projection_head_momentum, m=0.99)\n",
    "        x0 = x0.to(device)\n",
    "        x1 = x1.to(device)\n",
    "        p0 = model(x0)\n",
    "        z0 = model.forward_momentum(x0)\n",
    "        p1 = model(x1)\n",
    "        z1 = model.forward_momentum(x1)\n",
    "        loss = 0.5 * (criterion(p0, z1) + criterion(p1, z0))\n",
    "        total_loss += loss.detach()\n",
    "        optimizer.zero_grad()\n",
    "        loss.backward()\n",
    "        optimizer.step()  "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "slide"
    }
   },
   "source": [
    "### <img src=\"https://img.icons8.com/color/96/dinosaur-egg.png\" style=\"height:50px;display:inline\">  DINO: Emerging Properties in Self-Supervised Vision Transformers\n",
    "---\n",
    "* **<a href=\"https://arxiv.org/abs/2104.14294\">DINO</a>** (self-**di**stillation with **no** labels) is a self-supervised framework that utilizes Vision Transformers (ViTs) combined with momentum encoder, multi-crop training, and the use of small patches.\n",
    "* The authors observe that self-supervised ViT features contain explicit information about the semantic segmentation of an image, which does not emerge as clearly with supervised ViTs, nor with CNNs.\n",
    "* The model recieves two different random transformations of an input image to the student and teacher networks. The output of the teacher network is centered with a mean computed over the batch. Each network outputs a $K$ dimensional feature that is normalized with a temperature softmax over the feature dimension. Their similarity is then measured with a cross-entropy loss. The teacher parameters are updated with an EMA of the student parameters.\n",
    "* **Momentum encoder**: similarly to previous methods like BYOL, a teacher network is updated with an exponential moving average (EMA) of a student network's parameters.\n",
    "* **Multi-crop training**: the self-supervised training is formulated as learning to predict the same representation for different crops of the image. The crops are categorized as global views (crop covers more than 50\\% of the original image), or local views (crop covers less than 50\\% of the original image). The teacher network receives only global views, while the student network receives all the crops. The idea behind this is to encourage “local-to-global” correspondence in extracted features.\n",
    "* **Centering and sharpening of teacher outputs**: to avoid collapse towards learning trivial features, two operations with opposite effects are applied to the teacher’s outputs:\n",
    "    * **Centering**: subtracting a bias term from the teacher's output, $t \\leftarrow t - C$, where the centering term $C$ is computed on batch statistics and updated using an EMA (similarly to BatchNorm). It prevents one dimension from dominating the softmax, but encourages collapse to a uniform distribution.\n",
    "    * **Sharpening**: using a low value for the temperature $\\tau$ in the softmax: $\\text{softmax}(\\frac{t-C}{\\tau})$ . It has the opposite effect of centering, i.e. it “sharpens” the softmax, at the risk of letting one dimension dominate.\n",
    "* DINO uses cross-entropy as the loss function.\n",
    "* DINO's features are excellent $k-\\text{NN}$ classifiers, reaching 78.3\\% top-1 on ImageNet with a small ViT.\n",
    "* <a href=\"https://github.com/facebookresearch/dino\">PyTorch Code</a>, also available in the <a href=\"https://docs.lightly.ai/self-supervised-learning/examples/dino.html\">Lightly library</a>."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<center><img src=\"./assets/dino.PNG\" style=\"height:350px\"></center>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### <a href=\"https://arxiv.org/abs/2304.07193\">**DINOv2**</a>: Learning Robust Visual Features without Supervision\n",
    "---\n",
    "* Intorduces technical improvements aimed at accelerating and stabilizing the training at scale.\n",
    "* Large focus on the *data* -- an automatic pipeline is proposed to build a dedicated, diverse, and curated image dataset instead of uncurated data, as typically done in self-supervised training.\n",
    "* DINOv2 models produce high-performance visual features that can be directly employed with classifiers as simple as linear layers on a variety of computer vision tasks; these visual features are robust and perform well across domains without any requirement for fine-tuning.\n",
    "* <a href=\"https://github.com/facebookresearch/dinov2\">PyTorch Code and Pre-trained Models</a>, also available on <a href=\"https://huggingface.co/docs/transformers/en/model_doc/dinov2\">HuggingFace</a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "### <img src=\"https://img.icons8.com/clouds/80/000000/performance-2.png\" style=\"height:50px;display:inline\">  Performance Comparison\n",
    "---\n",
    "* **Benchmark dataset**: *Imagenette*, a subset of 10 easily classified classes from *ImageNet*.\n",
    "* Images are resized to $128 \\times 128$ resolution during training.\n",
    "* At the end of every epoch all training images are embedded and used as the features for a $k-\\text{NN}$ classifier with $k=20$ on the test set.\n",
    "* The reported $k-\\text{NN}$ Top 1 is the max accuracy over all epochs the model reached.\n",
    "* All experiments use the same ResNet-18 backbone and the default ImageNet1k training parameters from the respective papers.\n",
    "* The following experiments have been conducted on a system with single A6000 GPU.\n",
    "* Training a model takes three to five hours, including $k-\\text{NN}$ evaluation."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<center><img src=\"./assets/ssl_table.PNG\" style=\"height:500px\"></center>\n",
    "\n",
    "<a href=\"https://docs.lightly.ai/self-supervised-learning/getting_started/benchmarks.html\">Source and More Benchmarks</a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### <img src=\"./assets/lightly_logo.PNG\" style=\"height:50px;display:inline\">  Lightly - A Framework for Self-Supervised Learning\n",
    "---\n",
    "* <a href=\"https://docs.lightly.ai/self-supervised-learning/index.html\">Lightly</a> is a computer vision framework for self-supervised learning implemented in PyTorch.\n",
    "* Can be installed with `pip install lightly`.\n",
    "* <a href=\"https://docs.lightly.ai/self-supervised-learning/examples/models.html\">Examples and Tutorials</a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<center><img src=\"./assets/lightly_simsiam.PNG\" style=\"height:500px\"></center>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### <img src=\"https://img.icons8.com/external-gradients-pongsakorn-tan/64/null/external-clip-gdpr-gradients-pongsakorn-tan.png\" style=\"height:50px;display:inline\">  CLIP - Contrastive Language–Image Pre-training\n",
    "---\n",
    "* **CLIP** is a neural network which efficiently learns visual concepts from natural language supervision.\n",
    "* CLIP can be applied to any visual classification benchmark in a **zero-shot** manner by simply providing the names of the visual categories to be recognized.\n",
    "* **Training data**: text paired with images found across the internet.\n",
    "* **Self-supervised task**: given an image, predict which, out of a set of randomly sampled text snippets, was actually paired with it in the dataset, which is similar to the matching pardigm we introduced eariler.\n",
    "* **Loss function**: cosine similarity between pairs and normalized temperature-scaled cross-entropy loss.\n",
    "* At inference time we can, for example, classify photos of dogs vs. cats by checking for each image whether a CLIP model predicts the text description “a photo of a dog” or “a photo of a cat” is more likely to be paired with it.\n",
    "* <a href=\"https://github.com/openai/CLIP\">Official Repository (PyTorch)</a>\n",
    "* <a href=\"https://colab.research.google.com/github/openai/CLIP/blob/main/notebooks/Interacting_with_CLIP.ipynb\">Colab Example - Interaction with CLIP and Zero-Shot Classification</a>\n",
    "* HuggingFace Demos:\n",
    "    * <a href=\"https://huggingface.co/openai/clip-vit-large-patch14\">CLIP-ViT-Large</a>\n",
    "    * <a href=\"https://huggingface.co/spaces/taesiri/CLIPScore\">CLIPScore</a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<center><img src=\"./assets/openai_clip.PNG\" style=\"height:400px\"></center>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<center><img src=\"./assets/clip_zero_shot.PNG\" style=\"height:250px\"></center>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# clip usage example\n",
    "import torch\n",
    "import clip\n",
    "from PIL import Image\n",
    "\n",
    "device = \"cuda\" if torch.cuda.is_available() else \"cpu\"\n",
    "model, preprocess = clip.load(\"ViT-B/32\", device=device)\n",
    "\n",
    "image = preprocess(Image.open(\"CLIP.png\")).unsqueeze(0).to(device)\n",
    "text = clip.tokenize([\"a diagram\", \"a dog\", \"a cat\"]).to(device)\n",
    "\n",
    "with torch.no_grad():\n",
    "    image_features = model.encode_image(image)\n",
    "    text_features = model.encode_text(text)\n",
    "    \n",
    "    logits_per_image, logits_per_text = model(image, text)\n",
    "    probs = logits_per_image.softmax(dim=-1).cpu().numpy()\n",
    "\n",
    "print(\"Label probs:\", probs)  # prints: [[0.9927937  0.00421068 0.00299572]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# loss function skeleton\n",
    "image_features = clip.encode_image(image)\n",
    "text_features = clip.encode_text(text)\n",
    "\n",
    "# normalized features\n",
    "image_features = image_features / image_features.norm(dim=1, keepdim=True)\n",
    "text_features = text_features / text_features.norm(dim=1, keepdim=True)\n",
    "\n",
    "# cosine similarity as logits\n",
    "logit_scale = clip.logit_scale.exp()  # temperature parameter\n",
    "logits_per_image = logit_scale * image_features @ text_features.t()\n",
    "logits_per_text = logits_per_image.t()\n",
    "# shape = [global_batch_size, global_batch_size]\n",
    "\n",
    "# lables are the indices of the correct pair, which is just the index of the sample in the batch\n",
    "labels = torch.arange(batch_size) \n",
    "\n",
    "loss_i = torch.nn.functional.cross_entropy(logits_per_image, labels)  # applies softmax to logits inside\n",
    "loss_t = torch.nn.functional.cross_entropy(logits_per_text, labels)\n",
    "loss = 0.5 * (loss_i + loss_t).mean()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### CLIP Extensions\n",
    "---\n",
    "* <a href=\"https://github.com/yzhuoning/Awesome-CLIP\">Awesome-CLIP</a> - a repository that collects CLIP-based applications\n",
    "* <a href=\"https://arxiv.org/abs/2107.07651\">ALBEF - Align Before Fuse</a>\n",
    "    * <a href=\"https://github.com/salesforce/ALBEF/\">PyTorch Code</a>\n",
    "* <a href=\"https://arxiv.org/abs/2204.05991\">ReCLIP - A Strong Zero-Shot Baseline for Referring Expression Comprehension</a>\n",
    "    * <a href=\"https://github.com/allenai/reclip\">PyTorch Code</a>\n",
    "* <a href=\"https://arxiv.org/abs/2303.15343\">SigLIP - Sigmoid Loss for Language Image Pre-Training</a>\n",
    "    * <a href=\"https://huggingface.co/docs/transformers/main/en/model_doc/siglip\">Model on HF</a>\n",
    "* <a href=\"https://github.com/mlfoundations/open_clip\">OpenCLIP - an open source implementation of CLIP</a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "subslide"
    }
   },
   "source": [
    "### <img src=\"https://img.icons8.com/bubbles/50/000000/video-playlist.png\" style=\"height:50px;display:inline\"> Recommended Videos\n",
    "---\n",
    "#### <img src=\"https://img.icons8.com/cute-clipart/64/000000/warning-shield.png\" style=\"height:30px;display:inline\"> Warning!\n",
    "* These videos do not replace the lectures and tutorials.\n",
    "* Please use these to get a better understanding of the material, and not as an alternative to the written material.\n",
    "\n",
    "#### Video By Subject\n",
    "\n",
    "* Transfer Learning - <a href=\"https://www.youtube.com/watch?v=yofjFQddwHE\"> Transfer Learning (C3W2L07) </a>\n",
    "    * Transfer Learning in PyTorch- <a href=\"https://www.youtube.com/watch?v=K0lWSB2QoIQ\"> PyTorch Tutorial 15 - Transfer Learning </a>\n",
    "* LoRA - <a href=\"https://www.youtube.com/watch?v=dA-NhCtrrVE\">Low-rank Adaption of Large Language Models: Explaining the Key Concepts Behind LoRA</a>\n",
    "    * <a href=\"https://www.youtube.com/watch?v=rgmJep4Sba4\">Insights from Finetuning LLMs with Low-Rank Adaptation</a>\n",
    "* DoRA - <a href=\"https://www.youtube.com/watch?v=m7KQdGSr0Dg\">DoRA: Weight-Decomposed Low-Rank Adaptation</a>\n",
    "* General Self-Supervised Learning - <a href=\"https://www.youtube.com/watch?v=dMUes74-nYY\">Lecture 7 Self-Supervised Learning -- UC Berkeley Spring 2020 - CS294-158 Deep Unsupervised Learning</a>\n",
    "* SimCLR - <a href=\"https://www.youtube.com/watch?v=APki8LmdJwY\">SimCLR Explained!</a>\n",
    "* MoCo - <a href=\"https://www.youtube.com/watch?v=LvHwBQF14zs\">Momentum Contrastive Learning</a>\n",
    "* BYOL - <a href=\"https://www.youtube.com/watch?v=YPfUiOMYOEE\">BYOL: Bootstrap Your Own Latent: A New Approach to Self-Supervised Learning</a>\n",
    "* DINO - <a href=\"https://www.youtube.com/watch?v=h3ij3F3cPIk\">DINO: Emerging Properties in Self-Supervised Vision Transformers (Facebook AI Research Explained)</a>\n",
    "* CLIP - <a href=\"https://www.youtube.com/watch?v=u0HG77RNhPE\">Connor Shorten - CLIP: Connecting Text and Images</a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "slideshow": {
     "slide_type": "skip"
    }
   },
   "source": [
    "## <img src=\"https://img.icons8.com/stickers/100/null/prize.png\" style=\"height:50px;display:inline\"> Credits\n",
    "---\n",
    "* Icons made by <a href=\"https://www.flaticon.com/authors/becris\" title=\"Becris\">Becris</a> from <a href=\"https://www.flaticon.com/\" title=\"Flaticon\">www.flaticon.com</a>\n",
    "* Icons from <a href=\"https://icons8.com/\">Icons8.com</a> - https://icons8.com\n",
    "* <a href=\"https://ruder.io/transfer-learning/\">Sebastian Ruder - Transfer Learning - Machine Learning's Next Frontier</a>\n",
    "* <a href=\"https://ai.googleblog.com/2018/11/open-sourcing-bert-state-of-art-pre.html\">Jacob Devlin and Ming-Wei Chang - Open Sourcing BERT: State-of-the-Art Pre-training for Natural Language Processing</a>\n",
    "* <a href=\"https://sites.google.com/view/berkeley-cs294-158-sp20/home\">CS294-158-SP20-Deep Unsupervised Learning</a>\n",
    "* <a href=\"https://paperswithcode.com/method/contrastive-predictive-coding\">Contrastive Predictive Coding</a>\n",
    "* <a href=\"https://paperswithcode.com/method/simclr\"> Simple Framework for Contrastive Learning of Visual Representations (SimCLR)</a>\n",
    "* <a href=\"https://leimao.github.io/blog/Exponential-Moving-Average/\">Exponential Moving Average</a>\n",
    "* <a href=\"https://paperswithcode.com/method/moco\">Momentum Contrast</a>\n",
    "* <a href=\"https://untitled-ai.github.io/understanding-self-supervised-contrastive-learning.html\">Understanding self-supervised and contrastive learning with \"Bootstrap Your Own Latent\" (BYOL)</a>\n",
    "* <a href=\"https://creatis-myriad.github.io/2022/06/01/EmergingPropertiesSSViT.html\">MYRIAD - Emerging Properties in Self-Supervised Vision Transformers</a>\n",
    "* <a href=\"https://openai.com/blog/clip/\">OpenAI - CLIP: Connecting Text and Images</a>"
   ]
  }
 ],
 "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": "3.8.18"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}