diff --git a/Copy_of_Part1_TensorFlow.ipynb b/Copy_of_Part1_TensorFlow.ipynb new file mode 100644 index 0000000..d9ec08a --- /dev/null +++ b/Copy_of_Part1_TensorFlow.ipynb @@ -0,0 +1,748 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "metadata": { + "id": "view-in-github", + "colab_type": "text" + }, + "source": [ + "\"Open" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "WBk0ZDWY-ff8" + }, + "source": [ + "\n", + " \n", + " \n", + " \n", + "
\n", + " \n", + " Visit MIT Deep Learning\n", + " Run in Google Colab\n", + " View Source on GitHub
\n", + "\n", + "\n", + "# Copyright Information\n" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "3eI6DUic-6jo" + }, + "outputs": [], + "source": [ + "# Copyright 2024 MIT Introduction to Deep Learning. All Rights Reserved.\n", + "#\n", + "# Licensed under the MIT License. You may not use this file except in compliance\n", + "# with the License. Use and/or modification of this code outside of MIT Introduction\n", + "# to Deep Learning must reference:\n", + "#\n", + "# © MIT Introduction to Deep Learning\n", + "# http://introtodeeplearning.com\n", + "#" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "57knM8jrYZ2t" + }, + "source": [ + "# Lab 1: Intro to TensorFlow and Music Generation with RNNs\n", + "\n", + "In this lab, you'll get exposure to using TensorFlow and learn how it can be used for solving deep learning tasks. Go through the code and run each cell. Along the way, you'll encounter several ***TODO*** blocks -- follow the instructions to fill them out before running those cells and continuing.\n", + "\n", + "\n", + "# Part 1: Intro to TensorFlow\n", + "\n", + "## 0.1 Install TensorFlow\n", + "\n", + "TensorFlow is a software library extensively used in machine learning. Here we'll learn how computations are represented and how to define a simple neural network in TensorFlow. For all the labs in Introduction to Deep Learning 2023, we'll be using the latest version of TensorFlow, TensorFlow 2, which affords great flexibility and the ability to imperatively execute operations, just like in Python. You'll notice that TensorFlow 2 is quite similar to Python in its syntax and imperative execution. Let's install TensorFlow and a couple of dependencies.\n" + ] + }, + { + "cell_type": "code", + "execution_count": 2, + "metadata": { + "id": "LkaimNJfYZ2w", + "outputId": "a3cb04c9-6ab2-4597-fbb3-696b4b5b6bd1", + "colab": { + "base_uri": "https://localhost:8080/" + } + }, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "\u001b[?25l \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m0.0/2.1 MB\u001b[0m \u001b[31m?\u001b[0m eta \u001b[36m-:--:--\u001b[0m\r\u001b[2K \u001b[91m━━\u001b[0m\u001b[91m╸\u001b[0m\u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m0.2/2.1 MB\u001b[0m \u001b[31m4.3 MB/s\u001b[0m eta \u001b[36m0:00:01\u001b[0m\r\u001b[2K \u001b[91m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[90m╺\u001b[0m\u001b[90m━━━━━━━━━\u001b[0m \u001b[32m1.6/2.1 MB\u001b[0m \u001b[31m22.3 MB/s\u001b[0m eta \u001b[36m0:00:01\u001b[0m\r\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m2.1/2.1 MB\u001b[0m \u001b[31m23.1 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n", + "\u001b[?25h Preparing metadata (setup.py) ... \u001b[?25l\u001b[?25hdone\n", + " Building wheel for mitdeeplearning (setup.py) ... \u001b[?25l\u001b[?25hdone\n" + ] + } + ], + "source": [ + "import tensorflow as tf\n", + "\n", + "# Download and import the MIT Introduction to Deep Learning package\n", + "!pip install mitdeeplearning --quiet\n", + "import mitdeeplearning as mdl\n", + "\n", + "import numpy as np\n", + "import matplotlib.pyplot as plt" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "2QNMcdP4m3Vs" + }, + "source": [ + "## 1.1 Why is TensorFlow called TensorFlow?\n", + "\n", + "TensorFlow is called 'TensorFlow' because it handles the flow (node/mathematical operation) of Tensors, which are data structures that you can think of as multi-dimensional arrays. Tensors are represented as n-dimensional arrays of base dataypes such as a string or integer -- they provide a way to generalize vectors and matrices to higher dimensions.\n", + "\n", + "The ```shape``` of a Tensor defines its number of dimensions and the size of each dimension. The ```rank``` of a Tensor provides the number of dimensions (n-dimensions) -- you can also think of this as the Tensor's order or degree.\n", + "\n", + "Let's first look at 0-d Tensors, of which a scalar is an example:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "tFxztZQInlAB" + }, + "outputs": [], + "source": [ + "sport = tf.constant(\"Tennis\", tf.string)\n", + "number = tf.constant(1.41421356237, tf.float64)\n", + "\n", + "print(\"`sport` is a {}-d Tensor\".format(tf.rank(sport).numpy()))\n", + "print(\"`number` is a {}-d Tensor\".format(tf.rank(number).numpy()))" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "-dljcPUcoJZ6" + }, + "source": [ + "Vectors and lists can be used to create 1-d Tensors:" + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "metadata": { + "id": "oaHXABe8oPcO", + "colab": { + "base_uri": "https://localhost:8080/" + }, + "outputId": "6a29b94a-8c11-437d-c350-7ff7bfcaa9f3" + }, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "`sports` is a 1-d Tensor with shape: [2]\n", + "`numbers` is a 1-d Tensor with shape: [3]\n" + ] + } + ], + "source": [ + "sports = tf.constant([\"Tennis\", \"Basketball\"], tf.string)\n", + "numbers = tf.constant([3.141592, 1.414213, 2.71821], tf.float64)\n", + "\n", + "print(\"`sports` is a {}-d Tensor with shape: {}\".format(tf.rank(sports).numpy(), tf.shape(sports)))\n", + "print(\"`numbers` is a {}-d Tensor with shape: {}\".format(tf.rank(numbers).numpy(), tf.shape(numbers)))" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "gvffwkvtodLP" + }, + "source": [ + "Next we consider creating 2-d (i.e., matrices) and higher-rank Tensors. For examples, in future labs involving image processing and computer vision, we will use 4-d Tensors. Here the dimensions correspond to the number of example images in our batch, image height, image width, and the number of color channels." + ] + }, + { + "cell_type": "code", + "execution_count": 4, + "metadata": { + "id": "tFeBBe1IouS3", + "colab": { + "base_uri": "https://localhost:8080/" + }, + "outputId": "aa1293d8-3059-441d-95bb-fe2d0f002432" + }, + "outputs": [ + { + "output_type": "stream", + "name": "stderr", + "text": [ + "/usr/local/lib/python3.10/dist-packages/ipykernel/ipkernel.py:283: DeprecationWarning: `should_run_async` will not call `transform_cell` automatically in the future. Please pass the result to `transformed_cell` argument and any exception that happen during thetransform in `preprocessing_exc_tuple` in IPython 7.17 and above.\n", + " and should_run_async(code)\n" + ] + } + ], + "source": [ + "### Defining higher-order Tensors ###\n", + "\n", + "'''TODO: Define a 2-d Tensor'''\n", + "matrix = tf.constant([[1,2], [3,4]])\n", + "\n", + "assert isinstance(matrix, tf.Tensor), \"matrix must be a tf Tensor object\"\n", + "assert tf.rank(matrix).numpy() == 2" + ] + }, + { + "cell_type": "code", + "execution_count": 6, + "metadata": { + "id": "Zv1fTn_Ya_cz" + }, + "outputs": [], + "source": [ + "'''TODO: Define a 4-d Tensor.'''\n", + "# Use tf.zeros to initialize a 4-d Tensor of zeros with size 10 x 256 x 256 x 3.\n", + "# You can think of this as 10 images where each image is RGB 256 x 256.\n", + "images = tf.zeros(shape=[10,256,256,3])\n", + "\n", + "assert isinstance(images, tf.Tensor), \"matrix must be a tf Tensor object\"\n", + "assert tf.rank(images).numpy() == 4, \"matrix must be of rank 4\"\n", + "assert tf.shape(images).numpy().tolist() == [10, 256, 256, 3], \"matrix is incorrect shape\"" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "wkaCDOGapMyl" + }, + "source": [ + "As you have seen, the ```shape``` of a Tensor provides the number of elements in each Tensor dimension. The ```shape``` is quite useful, and we'll use it often. You can also use slicing to access subtensors within a higher-rank Tensor:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "FhaufyObuLEG" + }, + "outputs": [], + "source": [ + "row_vector = matrix[1]\n", + "column_vector = matrix[:,1]\n", + "scalar = matrix[0, 1]\n", + "\n", + "print(\"`row_vector`: {}\".format(row_vector.numpy()))\n", + "print(\"`column_vector`: {}\".format(column_vector.numpy()))\n", + "print(\"`scalar`: {}\".format(scalar.numpy()))" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "iD3VO-LZYZ2z" + }, + "source": [ + "## 1.2 Computations on Tensors\n", + "\n", + "A convenient way to think about and visualize computations in TensorFlow is in terms of graphs. We can define this graph in terms of Tensors, which hold data, and the mathematical operations that act on these Tensors in some order. Let's look at a simple example, and define this computation using TensorFlow:\n", + "\n", + "![alt text](https://raw.githubusercontent.com/aamini/introtodeeplearning/master/lab1/img/add-graph.png)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "X_YJrZsxYZ2z" + }, + "outputs": [], + "source": [ + "# Create the nodes in the graph, and initialize values\n", + "a = tf.constant(15)\n", + "b = tf.constant(61)\n", + "\n", + "# Add them!\n", + "c1 = tf.add(a,b)\n", + "c2 = a + b # TensorFlow overrides the \"+\" operation so that it is able to act on Tensors\n", + "print(c1)\n", + "print(c2)" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "Mbfv_QOiYZ23" + }, + "source": [ + "Notice how we've created a computation graph consisting of TensorFlow operations, and how the output is a Tensor with value 76 -- we've just created a computation graph consisting of operations, and it's executed them and given us back the result.\n", + "\n", + "Now let's consider a slightly more complicated example:\n", + "\n", + "![alt text](https://raw.githubusercontent.com/aamini/introtodeeplearning/master/lab1/img/computation-graph.png)\n", + "\n", + "Here, we take two inputs, `a, b`, and compute an output `e`. Each node in the graph represents an operation that takes some input, does some computation, and passes its output to another node.\n", + "\n", + "Let's define a simple function in TensorFlow to construct this computation function:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "PJnfzpWyYZ23", + "scrolled": true + }, + "outputs": [], + "source": [ + "### Defining Tensor computations ###\n", + "\n", + "# Construct a simple computation function\n", + "def func(a,b):\n", + " '''TODO: Define the operation for c, d, e (use tf.add, tf.subtract, tf.multiply).'''\n", + " c = # TODO\n", + " d = # TODO\n", + " e = # TODO\n", + " return e" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "AwrRfDMS2-oy" + }, + "source": [ + "Now, we can call this function to execute the computation graph given some inputs `a,b`:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "pnwsf8w2uF7p" + }, + "outputs": [], + "source": [ + "# Consider example values for a,b\n", + "a, b = 1.5, 2.5\n", + "# Execute the computation\n", + "e_out = func(a,b)\n", + "print(e_out)" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "6HqgUIUhYZ29" + }, + "source": [ + "Notice how our output is a Tensor with value defined by the output of the computation, and that the output has no shape as it is a single scalar value." + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "1h4o9Bb0YZ29" + }, + "source": [ + "## 1.3 Neural networks in TensorFlow\n", + "We can also define neural networks in TensorFlow. TensorFlow uses a high-level API called [Keras](https://www.tensorflow.org/guide/keras) that provides a powerful, intuitive framework for building and training deep learning models.\n", + "\n", + "Let's first consider the example of a simple perceptron defined by just one dense layer: $ y = \\sigma(Wx + b)$, where $W$ represents a matrix of weights, $b$ is a bias, $x$ is the input, $\\sigma$ is the sigmoid activation function, and $y$ is the output. We can also visualize this operation using a graph:\n", + "\n", + "![alt text](https://raw.githubusercontent.com/aamini/introtodeeplearning/master/lab1/img/computation-graph-2.png)\n", + "\n", + "Tensors can flow through abstract types called [```Layers```](https://www.tensorflow.org/api_docs/python/tf/keras/layers/Layer) -- the building blocks of neural networks. ```Layers``` implement common neural networks operations, and are used to update weights, compute losses, and define inter-layer connectivity. We will first define a ```Layer``` to implement the simple perceptron defined above." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "HutbJk-1kHPh" + }, + "outputs": [], + "source": [ + "### Defining a network Layer ###\n", + "\n", + "# n_output_nodes: number of output nodes\n", + "# input_shape: shape of the input\n", + "# x: input to the layer\n", + "\n", + "class OurDenseLayer(tf.keras.layers.Layer):\n", + " def __init__(self, n_output_nodes):\n", + " super(OurDenseLayer, self).__init__()\n", + " self.n_output_nodes = n_output_nodes\n", + "\n", + " def build(self, input_shape):\n", + " d = int(input_shape[-1])\n", + " # Define and initialize parameters: a weight matrix W and bias b\n", + " # Note that parameter initialization is random!\n", + " self.W = self.add_weight(\"weight\", shape=[d, self.n_output_nodes]) # note the dimensionality\n", + " self.b = self.add_weight(\"bias\", shape=[1, self.n_output_nodes]) # note the dimensionality\n", + "\n", + " def call(self, x):\n", + " '''TODO: define the operation for z (hint: use tf.matmul)'''\n", + " z = # TODO\n", + "\n", + " '''TODO: define the operation for out (hint: use tf.sigmoid)'''\n", + " y = # TODO\n", + " return y\n", + "\n", + "# Since layer parameters are initialized randomly, we will set a random seed for reproducibility\n", + "tf.keras.utils.set_random_seed(1)\n", + "layer = OurDenseLayer(3)\n", + "layer.build((1,2))\n", + "x_input = tf.constant([[1,2.]], shape=(1,2))\n", + "y = layer.call(x_input)\n", + "\n", + "# test the output!\n", + "print(y.numpy())\n", + "mdl.lab1.test_custom_dense_layer_output(y)" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "Jt1FgM7qYZ3D" + }, + "source": [ + "Conveniently, TensorFlow has defined a number of ```Layers``` that are commonly used in neural networks, for example a [```Dense```](https://www.tensorflow.org/api_docs/python/tf/keras/layers/Dense?version=stable). Now, instead of using a single ```Layer``` to define our simple neural network, we'll use the [`Sequential`](https://www.tensorflow.org/versions/r2.0/api_docs/python/tf/keras/Sequential) model from Keras and a single [`Dense` ](https://www.tensorflow.org/versions/r2.0/api_docs/python/tf/keras/layers/Dense) layer to define our network. With the `Sequential` API, you can readily create neural networks by stacking together layers like building blocks." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "7WXTpmoL6TDz" + }, + "outputs": [], + "source": [ + "### Defining a neural network using the Sequential API ###\n", + "\n", + "# Import relevant packages\n", + "from tensorflow.keras import Sequential\n", + "from tensorflow.keras.layers import Dense\n", + "\n", + "# Define the number of outputs\n", + "n_output_nodes = 3\n", + "\n", + "# First define the model\n", + "model = Sequential()\n", + "\n", + "'''TODO: Define a dense (fully connected) layer to compute z'''\n", + "# Remember: dense layers are defined by the parameters W and b!\n", + "# You can read more about the initialization of W and b in the TF documentation :)\n", + "# https://www.tensorflow.org/api_docs/python/tf/keras/layers/Dense?version=stable\n", + "dense_layer = # TODO\n", + "\n", + "# Add the dense layer to the model\n", + "model.add(dense_layer)\n" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "HDGcwYfUyR-U" + }, + "source": [ + "That's it! We've defined our model using the Sequential API. Now, we can test it out using an example input:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "sg23OczByRDb" + }, + "outputs": [], + "source": [ + "# Test model with example input\n", + "x_input = tf.constant([[1,2.]], shape=(1,2))\n", + "\n", + "'''TODO: feed input into the model and predict the output!'''\n", + "model_output = # TODO\n", + "print(model_output)" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "596NvsOOtr9F" + }, + "source": [ + "In addition to defining models using the `Sequential` API, we can also define neural networks by directly subclassing the [`Model`](https://www.tensorflow.org/api_docs/python/tf/keras/Model?version=stable) class, which groups layers together to enable model training and inference. The `Model` class captures what we refer to as a \"model\" or as a \"network\". Using Subclassing, we can create a class for our model, and then define the forward pass through the network using the `call` function. Subclassing affords the flexibility to define custom layers, custom training loops, custom activation functions, and custom models. Let's define the same neural network as above now using Subclassing rather than the `Sequential` model." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "K4aCflPVyViD" + }, + "outputs": [], + "source": [ + "### Defining a model using subclassing ###\n", + "\n", + "from tensorflow.keras import Model\n", + "from tensorflow.keras.layers import Dense\n", + "\n", + "class SubclassModel(tf.keras.Model):\n", + "\n", + " # In __init__, we define the Model's layers\n", + " def __init__(self, n_output_nodes):\n", + " super(SubclassModel, self).__init__()\n", + " '''TODO: Our model consists of a single Dense layer. Define this layer.'''\n", + " self.dense_layer = '''TODO: Dense Layer'''\n", + "\n", + " # In the call function, we define the Model's forward pass.\n", + " def call(self, inputs):\n", + " return self.dense_layer(inputs)" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "U0-lwHDk4irB" + }, + "source": [ + "Just like the model we built using the `Sequential` API, let's test out our `SubclassModel` using an example input.\n", + "\n" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "LhB34RA-4gXb" + }, + "outputs": [], + "source": [ + "n_output_nodes = 3\n", + "model = SubclassModel(n_output_nodes)\n", + "\n", + "x_input = tf.constant([[1,2.]], shape=(1,2))\n", + "\n", + "print(model.call(x_input))" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "HTIFMJLAzsyE" + }, + "source": [ + "Importantly, Subclassing affords us a lot of flexibility to define custom models. For example, we can use boolean arguments in the `call` function to specify different network behaviors, for example different behaviors during training and inference. Let's suppose under some instances we want our network to simply output the input, without any perturbation. We define a boolean argument `isidentity` to control this behavior:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "P7jzGX5D1xT5" + }, + "outputs": [], + "source": [ + "### Defining a model using subclassing and specifying custom behavior ###\n", + "\n", + "from tensorflow.keras import Model\n", + "from tensorflow.keras.layers import Dense\n", + "\n", + "class IdentityModel(tf.keras.Model):\n", + "\n", + " # As before, in __init__ we define the Model's layers\n", + " # Since our desired behavior involves the forward pass, this part is unchanged\n", + " def __init__(self, n_output_nodes):\n", + " super(IdentityModel, self).__init__()\n", + " self.dense_layer = tf.keras.layers.Dense(n_output_nodes, activation='sigmoid')\n", + "\n", + " '''TODO: Implement the behavior where the network outputs the input, unchanged,\n", + " under control of the isidentity argument.'''\n", + " def call(self, inputs, isidentity=False):\n", + " x = self.dense_layer(inputs)\n", + " '''TODO: Implement identity behavior'''" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "Ku4rcCGx5T3y" + }, + "source": [ + "Let's test this behavior:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "NzC0mgbk5dp2" + }, + "outputs": [], + "source": [ + "n_output_nodes = 3\n", + "model = IdentityModel(n_output_nodes)\n", + "\n", + "x_input = tf.constant([[1,2.]], shape=(1,2))\n", + "'''TODO: pass the input into the model and call with and without the input identity option.'''\n", + "out_activate = # TODO\n", + "out_identity = # TODO\n", + "\n", + "print(\"Network output with activation: {}; network identity output: {}\".format(out_activate.numpy(), out_identity.numpy()))" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "7V1dEqdk6VI5" + }, + "source": [ + "Now that we have learned how to define `Layers` as well as neural networks in TensorFlow using both the `Sequential` and Subclassing APIs, we're ready to turn our attention to how to actually implement network training with backpropagation." + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "dQwDhKn8kbO2" + }, + "source": [ + "## 1.4 Automatic differentiation in TensorFlow\n", + "\n", + "[Automatic differentiation](https://en.wikipedia.org/wiki/Automatic_differentiation)\n", + "is one of the most important parts of TensorFlow and is the backbone of training with\n", + "[backpropagation](https://en.wikipedia.org/wiki/Backpropagation). We will use the TensorFlow GradientTape [`tf.GradientTape`](https://www.tensorflow.org/api_docs/python/tf/GradientTape?version=stable) to trace operations for computing gradients later.\n", + "\n", + "When a forward pass is made through the network, all forward-pass operations get recorded to a \"tape\"; then, to compute the gradient, the tape is played backwards. By default, the tape is discarded after it is played backwards; this means that a particular `tf.GradientTape` can only\n", + "compute one gradient, and subsequent calls throw a runtime error. However, we can compute multiple gradients over the same computation by creating a ```persistent``` gradient tape.\n", + "\n", + "First, we will look at how we can compute gradients using GradientTape and access them for computation. We define the simple function $ y = x^2$ and compute the gradient:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "tdkqk8pw5yJM" + }, + "outputs": [], + "source": [ + "### Gradient computation with GradientTape ###\n", + "\n", + "# y = x^2\n", + "# Example: x = 3.0\n", + "x = tf.Variable(3.0)\n", + "\n", + "# Initiate the gradient tape\n", + "with tf.GradientTape() as tape:\n", + " # Define the function\n", + " y = x * x\n", + "# Access the gradient -- derivative of y with respect to x\n", + "dy_dx = tape.gradient(y, x)\n", + "\n", + "assert dy_dx.numpy() == 6.0" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "JhU5metS5xF3" + }, + "source": [ + "In training neural networks, we use differentiation and stochastic gradient descent (SGD) to optimize a loss function. Now that we have a sense of how `GradientTape` can be used to compute and access derivatives, we will look at an example where we use automatic differentiation and SGD to find the minimum of $L=(x-x_f)^2$. Here $x_f$ is a variable for a desired value we are trying to optimize for; $L$ represents a loss that we are trying to minimize. While we can clearly solve this problem analytically ($x_{min}=x_f$), considering how we can compute this using `GradientTape` sets us up nicely for future labs where we use gradient descent to optimize entire neural network losses." + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "attributes": { + "classes": [ + "py" + ], + "id": "" + }, + "id": "7g1yWiSXqEf-" + }, + "outputs": [], + "source": [ + "### Function minimization with automatic differentiation and SGD ###\n", + "\n", + "# Initialize a random value for our initial x\n", + "x = tf.Variable([tf.random.normal([1])])\n", + "print(\"Initializing x={}\".format(x.numpy()))\n", + "\n", + "learning_rate = 1e-2 # learning rate for SGD\n", + "history = []\n", + "# Define the target value\n", + "x_f = 4\n", + "\n", + "# We will run SGD for a number of iterations. At each iteration, we compute the loss,\n", + "# compute the derivative of the loss with respect to x, and perform the SGD update.\n", + "for i in range(500):\n", + " with tf.GradientTape() as tape:\n", + " '''TODO: define the loss as described above'''\n", + " loss = # TODO\n", + "\n", + " # loss minimization using gradient tape\n", + " grad = tape.gradient(loss, x) # compute the derivative of the loss with respect to x\n", + " new_x = x - learning_rate*grad # sgd update\n", + " x.assign(new_x) # update the value of x\n", + " history.append(x.numpy()[0])\n", + "\n", + "# Plot the evolution of x as we optimize towards x_f!\n", + "plt.plot(history)\n", + "plt.plot([0, 500],[x_f,x_f])\n", + "plt.legend(('Predicted', 'True'))\n", + "plt.xlabel('Iteration')\n", + "plt.ylabel('x value')" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "pC7czCwk3ceH" + }, + "source": [ + "`GradientTape` provides an extremely flexible framework for automatic differentiation. In order to back propagate errors through a neural network, we track forward passes on the Tape, use this information to determine the gradients, and then use these gradients for optimization using SGD." + ] + } + ], + "metadata": { + "accelerator": "GPU", + "colab": { + "collapsed_sections": [ + "WBk0ZDWY-ff8" + ], + "provenance": [], + "include_colab_link": true + }, + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "python3" + }, + "language_info": { + "name": "python", + "version": "3.9.6" + }, + "vscode": { + "interpreter": { + "hash": "31f2aee4e71d21fbe5cf8b01ff0e069b9275f58929596ceb00d14d90e3e16cd6" + } + } + }, + "nbformat": 4, + "nbformat_minor": 0 +} \ No newline at end of file diff --git a/L1_micrograd.ipynb b/L1_micrograd.ipynb new file mode 100644 index 0000000..5172cf4 --- /dev/null +++ b/L1_micrograd.ipynb @@ -0,0 +1,560 @@ +{ + "nbformat": 4, + "nbformat_minor": 0, + "metadata": { + "colab": { + "provenance": [], + "authorship_tag": "ABX9TyMXO/7E2HPH13jXZKn4jz3f", + "include_colab_link": true + }, + "kernelspec": { + "name": "python3", + "display_name": "Python 3" + }, + "language_info": { + "name": "python" + } + }, + "cells": [ + { + "cell_type": "markdown", + "metadata": { + "id": "view-in-github", + "colab_type": "text" + }, + "source": [ + "\"Open" + ] + }, + { + "cell_type": "markdown", + "source": [ + "Value class is a scalar class - 0 D tensor." + ], + "metadata": { + "id": "v_WdNgouGQzc" + } + }, + { + "cell_type": "code", + "source": [ + "import numpy as np\n", + "import math\n", + "import matplotlib.pyplot as plt\n", + "\n", + "class Value:\n", + " def __init__(self, data, _children=set(), _op='', label=''):\n", + " self.data = data\n", + " self._prev = set(_children) # has to be set to revent cases like b=a+a\n", + " self._op = _op\n", + " self.label = label\n", + " self.grad = 0.0\n", + " self._backward = lambda: None\n", + "\n", + " def __repr__(self):\n", + " return f\"Value(data={self.data}, label={self.label})\"\n", + " def __add__(self, other):\n", + " other = other if isinstance(other, Value) else Value(other) # support other as an integer\n", + " out = Value(self.data + other.data, (other, self), '+')\n", + " def _backward():\n", + " self.grad += 1.0 * out.grad\n", + " other.grad += 1.0 * out.grad\n", + " out._backward = _backward\n", + " return out\n", + " def __mul__(self, other):\n", + " other = other if isinstance(other, Value) else Value(other) # support other as an integer\n", + " out = Value(self.data * other.data, (other, self), '*')\n", + " def _backward():\n", + " self.grad += other.data * out.grad\n", + " other.grad += self.data * out.grad\n", + " out._backward = _backward\n", + " return out\n", + " def tanh(self):\n", + " x = self.data\n", + " t = (math.exp(2*x) - 1)/(math.exp(2*x) + 1)\n", + " out = Value(t, (self,), _op='tanh')\n", + " def _backward():\n", + " self.grad += (1 - t**2)*out.grad\n", + " out._backward = _backward\n", + " return out\n", + " def exp(self):\n", + " out = Value(math.exp(self.data),(self, ), _op='exp')\n", + "\n", + " def _backward():\n", + " self.grad += out.data*out.grad\n", + "\n", + " out._backward = _backward\n", + " return out\n", + "\n", + " def __sub__(self, other): # self - other, in such case, otehr can be an integer, instead a Value obj\n", + " return self + (-other)\n", + "\n", + " def __neg__(self):\n", + " # return (-1) * self\n", + " return self * (-1)\n", + "\n", + " def __rmul__(self, other): # other * self, in such case, otehr can be an integer, instead a Value obj\n", + " return self * other\n", + "\n", + " def __radd__(self, other): # other + self, in such case, otehr can be an integer, instead a Value obj\n", + " return self + other\n", + "\n", + " def __pow__(self, other):\n", + " # why Integer other, not Value other?\n", + " # it is only for simplicity, because if the power is a Value, _backward(derivative) is going to include self.grad and other.grad f.e.\n", + " # self.grad = other.data * (self.data ** (other.data - 1)), not much different.\n", + " # other.grad = (self.data ** other.data) * ln(other.data), since other is a Value, we also need to upgrade its grad in _backward.\n", + " assert isinstance(other, (int, float)), \"only support int/float power\"\n", + " out = Value(self.data ** other, (self,), _op=f'** {other}')\n", + "\n", + " def _backward():\n", + " self.grad += other * (self.data ** (other - 1)) * out.grad\n", + " out._backward = _backward\n", + "\n", + " return out\n", + "\n", + " def __truediv__(self, other):\n", + " return self * (other ** -1)\n", + "\n", + " def backward(self):\n", + " # backward gradient\n", + " visited = set()\n", + " child = []\n", + " # collect child points\n", + " def find_topo(v):\n", + " if v not in visited:\n", + " visited.add(v)\n", + " for nxt in v._prev:\n", + " find_topo(nxt)\n", + " child.append(v)\n", + "\n", + " find_topo(self)\n", + "\n", + " self.grad = 1.0 # !!!don't forget to init grad to 1\n", + " for c in child[::-1]:\n", + " c._backward()\n", + "\n", + "\n" + ], + "metadata": { + "id": "uznqt_WRGkxK" + }, + "execution_count": 84, + "outputs": [] + }, + { + "cell_type": "markdown", + "source": [ + "Function for drawing dataflow between Values" + ], + "metadata": { + "id": "bLM0-KjzGnK9" + } + }, + { + "cell_type": "code", + "source": [ + "from graphviz import Digraph\n", + "\n", + "def trace(root):\n", + " nodes, edges = set(), set()\n", + " def build(n):\n", + " if n not in nodes:\n", + " nodes.add(n)\n", + " for child in n._prev:\n", + " edges.add((child, n))\n", + " build(child)\n", + " build(root)\n", + " return nodes, edges\n", + "\n", + "def draw_dot(node):\n", + " dot = Digraph(format='svg', graph_attr={'rankdir': 'LR'})\n", + "\n", + " nodes, edges = trace(node)\n", + " for n in nodes:\n", + " uid = str(id(n))\n", + " dot.node(name = uid, label = '{ %s | Data %.4f | Grad %.4f }' % (n.label, n.data, n.grad), shape='record')\n", + " if n._op:\n", + " dot.node(name=uid+n._op, label=f'{n._op}')\n", + " dot.edge(uid+n._op, uid)\n", + " for e_l, e_r in edges:\n", + " dot.edge(str(id(e_l)), str(id(e_r))+e_r._op)\n", + " return dot" + ], + "metadata": { + "id": "FFRUWPKDGpj4" + }, + "execution_count": 85, + "outputs": [] + }, + { + "cell_type": "markdown", + "source": [ + "Build forward dataflow and also calculate backward gradients." + ], + "metadata": { + "id": "hfTO9zUOGu0c" + } + }, + { + "cell_type": "code", + "source": [ + "# mini neural network\n", + "\"\"\"\n", + "x1 ->|\n", + " |X1W1+X2W2+b| -> tanh -> out\n", + "x2 ->|\n", + "\n", + "\"\"\"\n", + "x1 = Value(data=2, label='x1')\n", + "w1 = Value(data=-3, label='w1')\n", + "x2 = Value(data=0, label='x2')\n", + "w2 = Value(data=1, label='w2')\n", + "\n", + "x1xw1 = x1*w1; x1xw1.label = 'x1*w1'\n", + "x2xw2 = x2*w2; x2xw2.label = 'x2*w2'\n", + "\n", + "j = x1xw1 + x2xw2; j.label = 'x1*w1 + x2*w2'\n", + "\n", + "b = Value(data=6.8813735870195432, label='b')\n", + "h = j + b; h.label = 'h'\n", + "\n", + "# use Value.tanh directly\n", + "# o = h.tanh(); o.label = 'o'; o.grad = 1.0\n", + "# o.data is 0.7071\n", + "\n", + "# construct tanh manually\n", + "o = ((2*h).exp() - 1) / ((2*h).exp() + 1); o.lable = 'o'; o.grad = 1.0\n", + "o.backward()" + ], + "metadata": { + "id": "-FsIPVCWOv85" + }, + "execution_count": 86, + "outputs": [] + }, + { + "cell_type": "code", + "source": [ + "draw_dot(o)" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/", + "height": 390 + }, + "id": "kk4iqILbqaN7", + "outputId": "1d9acaee-1786-4b63-c3ef-0a0c0fa3699d" + }, + "execution_count": 87, + "outputs": [ + { + "output_type": "execute_result", + "data": { + "image/svg+xml": "\n\n\n\n\n\n%3\n\n\n\n140710486501424\n\n \n\nData 2.0000\n\nGrad 0.7523\n\n\n\n140710486504544*\n\n*\n\n\n\n140710486501424->140710486504544*\n\n\n\n\n\n140710486499408\n\n \n\nData 5.8284\n\nGrad 0.1464\n\n\n\n140710486498448+\n\n+\n\n\n\n140710486499408->140710486498448+\n\n\n\n\n\n140710486499408exp\n\nexp\n\n\n\n140710486499408exp->140710486499408\n\n\n\n\n\n140710486493744\n\n \n\nData 2.0000\n\nGrad -0.5320\n\n\n\n140710486500080*\n\n*\n\n\n\n140710486493744->140710486500080*\n\n\n\n\n\n140713403091600\n\nx2\n\nData 0.0000\n\nGrad 0.5000\n\n\n\n140710487012784*\n\n*\n\n\n\n140713403091600->140710487012784*\n\n\n\n\n\n140710486500080\n\n \n\nData 1.7627\n\nGrad -0.6036\n\n\n\n140710486498976exp\n\nexp\n\n\n\n140710486500080->140710486498976exp\n\n\n\n\n\n140710486500080*->140710486500080\n\n\n\n\n\n140710486500128\n\n \n\nData 1.0000\n\nGrad -0.1036\n\n\n\n140710486500704+\n\n+\n\n\n\n140710486500128->140710486500704+\n\n\n\n\n\n140710486496096\n\n \n\nData -1.0000\n\nGrad 0.1464\n\n\n\n140710486496096->140710486498448+\n\n\n\n\n\n140710486504544\n\n \n\nData 1.7627\n\nGrad 0.8536\n\n\n\n140710486504544->140710486499408exp\n\n\n\n\n\n140710486504544*->140710486504544\n\n\n\n\n\n140710486498448\n\n \n\nData 4.8284\n\nGrad 0.1464\n\n\n\n140710486499024*\n\n*\n\n\n\n140710486498448->140710486499024*\n\n\n\n\n\n140710486498448+->140710486498448\n\n\n\n\n\n140710487002272\n\nx1*w1 + x2*w2\n\nData -6.0000\n\nGrad 0.5000\n\n\n\n140710486506896+\n\n+\n\n\n\n140710487002272->140710486506896+\n\n\n\n\n\n140710487002272+\n\n+\n\n\n\n140710487002272+->140710487002272\n\n\n\n\n\n140710487014608\n\nb\n\nData 6.8814\n\nGrad 0.5000\n\n\n\n140710487014608->140710486506896+\n\n\n\n\n\n140713403090208\n\nx1\n\nData 2.0000\n\nGrad -1.5000\n\n\n\n140710487008944*\n\n*\n\n\n\n140713403090208->140710487008944*\n\n\n\n\n\n140710486500704\n\n \n\nData 6.8284\n\nGrad -0.1036\n\n\n\n140710486507376** -1\n\n** -1\n\n\n\n140710486500704->140710486507376** -1\n\n\n\n\n\n140710486500704+->140710486500704\n\n\n\n\n\n140710486506896\n\nh\n\nData 0.8814\n\nGrad 0.5000\n\n\n\n140710486506896->140710486500080*\n\n\n\n\n\n140710486506896->140710486504544*\n\n\n\n\n\n140710486506896+->140710486506896\n\n\n\n\n\n140710487012784\n\nx2*w2\n\nData 0.0000\n\nGrad 0.5000\n\n\n\n140710487012784->140710487002272+\n\n\n\n\n\n140710487012784*->140710487012784\n\n\n\n\n\n140710486498976\n\n \n\nData 5.8284\n\nGrad -0.1036\n\n\n\n140710486498976->140710486500704+\n\n\n\n\n\n140710486498976exp->140710486498976\n\n\n\n\n\n140710487008944\n\nx1*w1\n\nData -6.0000\n\nGrad 0.5000\n\n\n\n140710487008944->140710487002272+\n\n\n\n\n\n140710487008944*->140710487008944\n\n\n\n\n\n140710486499024\n\n \n\nData 0.7071\n\nGrad 1.0000\n\n\n\n140710486499024*->140710486499024\n\n\n\n\n\n140713403090784\n\nw1\n\nData -3.0000\n\nGrad 1.0000\n\n\n\n140713403090784->140710487008944*\n\n\n\n\n\n140710486507376\n\n \n\nData 0.1464\n\nGrad 4.8284\n\n\n\n140710486507376->140710486499024*\n\n\n\n\n\n140710486507376** -1->140710486507376\n\n\n\n\n\n140713403090832\n\nw2\n\nData 1.0000\n\nGrad 0.0000\n\n\n\n140713403090832->140710487012784*\n\n\n\n\n\n", + "text/plain": [ + "" + ] + }, + "metadata": {}, + "execution_count": 87 + } + ] + }, + { + "cell_type": "markdown", + "source": [ + "Below is create scalars using tensor from Pytorch. And replicate the same dataflow with tensor and also calculate the backward gradients of these tensors.\n" + ], + "metadata": { + "id": "J8brBQog_rrb" + } + }, + { + "cell_type": "code", + "source": [ + "import torch\n", + "\n", + "x1 = torch.Tensor([2.0]).double(); x1.requires_grad = True\n", + "w1 = torch.Tensor([-3.0]).double(); w1.requires_grad = True\n", + "x2 = torch.Tensor([0.0]).double(); x2.requires_grad = True\n", + "w2 = torch.Tensor([1.0]).double(); w2.requires_grad = True\n", + "b = torch.tensor([6.8813735870195432]).double(); b.requires_grad = True\n", + "\n", + "h = x1*w1 + x2*w2 + b # note tensor h doesn't require gradient, h.grad is None. But its gradient was definitely calculated as part of the chain rule.\n", + "e = h.tanh() # so does e\n", + "e.backward()\n", + "\n", + "print(e.data.item())\n", + "print('----')\n", + "print(\"x1.grad\", x1.grad.item())\n", + "print(\"x2.grad\", x2.grad.item())\n", + "print(\"w1.grad\", w1.grad.item())\n", + "print(\"w2.grad\", w2.grad.item())\n" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/" + }, + "id": "URY-HwRa_01_", + "outputId": "cb2977eb-22de-4e22-a564-5d26262d1f22" + }, + "execution_count": 88, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "0.7071066904050358\n", + "----\n", + "x1.grad -1.5000003851533106\n", + "x2.grad 0.5000001283844369\n", + "w1.grad 1.0000002567688737\n", + "w2.grad 0.0\n" + ] + } + ] + }, + { + "cell_type": "markdown", + "source": [ + "Now lets get back to Value, and use it to create Neuro, Layer, MLP." + ], + "metadata": { + "id": "4YYSZ6vpFzpV" + } + }, + { + "cell_type": "code", + "source": [ + "import random\n", + "\n", + "class Module:\n", + "\n", + " def parameters(self):\n", + " # overwrite!!\n", + " return []\n", + "\n", + " def zero_grad(self):\n", + " for p in self.parameters():\n", + " p.grad = 0.0\n", + "\n", + "\n", + "class Neuro(Module):\n", + " \"\"\"\n", + " computation unit:\n", + "\n", + " inputs -> |weights*inputs + bias| -> activation func -> out\n", + "\n", + " so three things from one neuro: weights, bias, activation func\n", + " \"\"\"\n", + "\n", + " def __init__(self, nin):\n", + " \"\"\"\n", + " define weights and bias base on input number:\n", + " one input one weight, one neuro one bias\n", + " nin: number of the inputs that go into the neuro\n", + " \"\"\"\n", + " self.w = [Value(random.uniform(-1, 1)) for n in range(nin)] # a list of Value objs\n", + " self.b = Value(random.uniform(-1, 1)) # a Value obj\n", + " def __call__(self, inputs):\n", + " \"\"\"\n", + " give neuro output by first of all performing inputs*weights + bias, then a non-linear activation func(tanh here).\n", + " inputs: a list of Value objs/ or just a list of numbers, because in Value.__mul__(other), we can have \"other\" in non-Value type.\n", + " out: a Value obj\n", + "\n", + " kick off the computation\n", + " \"\"\"\n", + "\n", + " act = sum((w*x for w, x in zip(self.w, inputs)), start=self.b)\n", + " out = act.tanh()\n", + "\n", + " return out\n", + "\n", + " def parameters(self):\n", + " return self.w + [self.b]\n", + "\n", + "class Layer(Module):\n", + " \"\"\"\n", + " Layer is a series of neuros, f.e.\n", + " neuro_1\n", + " inputs neuro_2\n", + " ...\n", + " \"\"\"\n", + "\n", + " def __init__(self, nin, nouts):\n", + " \"\"\"\n", + " construct layer: a list of neuros.\n", + " \"\"\"\n", + " self.neuros = [Neuro(nin) for _ in range(nouts)]\n", + "\n", + " def __call__(self, inputs):\n", + " \"\"\"\n", + " kick off(call) all the neuros in this layer\n", + " \"\"\"\n", + " outs = [neuro(inputs) for neuro in self.neuros]\n", + " return outs\n", + "\n", + " def parameters(self):\n", + " return [p for n in self.neuros for p in n.parameters()]\n", + "\n", + "class MLP(Module):\n", + " \"\"\"\n", + " Multi layer perceptron, f.e.\n", + " neuro_11 neuro_21\n", + " inputs neuro_12 neuro_22 ...\n", + " ... ...\n", + " ----------------------------------\n", + " a network needs to know the # of input, # of layer and # of neuro per layer.\n", + "\n", + " \"\"\"\n", + " def __init__(self, nin, nouts):\n", + " \"\"\"\n", + " nin: number of inputs\n", + " nouts: a list of integer, f.e. [1,2,3]. len(nouts) - number of layers in MLP. nouts[i] - number of neuros(the number of outputs) in layer i.\n", + " \"\"\"\n", + " layers = [nin] + nouts\n", + " self.layers = [Layer(layers[i], layers[i+1]) for i in range(len(nouts))]\n", + "\n", + " def __call__(self, inputs):\n", + " \"\"\"\n", + " kick off(call) the layers in this MLP\n", + " \"\"\"\n", + " x = inputs\n", + " for layer in self.layers:\n", + " x = layer(x)\n", + "\n", + " return x # return MLP final outputs\n", + "\n", + " def parameters(self):\n", + " return [p for l in self.layers for p in l.parameters()]\n", + "\n" + ], + "metadata": { + "id": "A4E_rSQ4GBU0" + }, + "execution_count": 89, + "outputs": [] + }, + { + "cell_type": "code", + "source": [ + "## lets try init a 2 inputs, 1 hidden layers neuro network\n", + "import string\n", + "import random\n", + "\n", + "nint = 2\n", + "input_labels = random.sample(string.ascii_lowercase, nint)\n", + "inputs = [Value(random.uniform(-1, 1), label=input_labels[i]) for i in range(nint)]\n", + "print(inputs)\n", + "nlayers = [2]\n", + "\n", + "nn = MLP(len(inputs), nlayers)\n", + "outs = nn(inputs)\n", + "\n", + "\n", + "draw_dot(outs[0])" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/", + "height": 376 + }, + "id": "mBvPXaDOWymW", + "outputId": "a766720f-fdd7-4f11-8caf-069301c1aa6b" + }, + "execution_count": 90, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "[Value(data=-0.4749878447805571, label=q), Value(data=0.28584922745581, label=t)]\n" + ] + }, + { + "output_type": "execute_result", + "data": { + "image/svg+xml": "\n\n\n\n\n\n%3\n\n\n\n140710489791488\n\n \n\nData -0.5353\n\nGrad 0.0000\n\n\n\n140710489791488tanh\n\ntanh\n\n\n\n140710489791488tanh->140710489791488\n\n\n\n\n\n140710489789952\n\n \n\nData -0.4572\n\nGrad 0.0000\n\n\n\n140710489793456+\n\n+\n\n\n\n140710489789952->140710489793456+\n\n\n\n\n\n140710489789952*\n\n*\n\n\n\n140710489789952*->140710489789952\n\n\n\n\n\n140710489791584\n\nt\n\nData 0.2858\n\nGrad 0.0000\n\n\n\n140710489786304*\n\n*\n\n\n\n140710489791584->140710489786304*\n\n\n\n\n\n140710489790096\n\n \n\nData -0.5976\n\nGrad 0.0000\n\n\n\n140710489790096->140710489791488tanh\n\n\n\n\n\n140710489790096+\n\n+\n\n\n\n140710489790096+->140710489790096\n\n\n\n\n\n140710489798832\n\nq\n\nData -0.4750\n\nGrad 0.0000\n\n\n\n140710489798832->140710489789952*\n\n\n\n\n\n140710489796816\n\n \n\nData -0.0952\n\nGrad 0.0000\n\n\n\n140710489796816->140710489793456+\n\n\n\n\n\n140710489796912\n\n \n\nData -0.1579\n\nGrad 0.0000\n\n\n\n140710489796912->140710489786304*\n\n\n\n\n\n140710489793360\n\n \n\nData 0.9626\n\nGrad 0.0000\n\n\n\n140710489793360->140710489789952*\n\n\n\n\n\n140710489793456\n\n \n\nData -0.5524\n\nGrad 0.0000\n\n\n\n140710489793456->140710489790096+\n\n\n\n\n\n140710489793456+->140710489793456\n\n\n\n\n\n140710489786304\n\n \n\nData -0.0451\n\nGrad 0.0000\n\n\n\n140710489786304->140710489790096+\n\n\n\n\n\n140710489786304*->140710489786304\n\n\n\n\n\n", + "text/plain": [ + "" + ] + }, + "metadata": {}, + "execution_count": 90 + } + ] + }, + { + "cell_type": "code", + "source": [ + "# model params\n", + "inputs_shape = (4, 3) # 4 is batch size, 3 is epoch size\n", + "layer_shape = [3, 4, 5, 6] # 4 layer, each with 3, 4, 5, 6 number of neuro" + ], + "metadata": { + "id": "7LbxUeK4mD27" + }, + "execution_count": 91, + "outputs": [] + }, + { + "cell_type": "code", + "source": [ + "# Now, lets build a more complex NN and test it with multiple inputs\n", + "def build_model(nin, layer_shape):\n", + " return MLP(nin=nin,\n", + " nouts=layer_shape)\n", + "\n", + "# due to generate random inputs and outputs, sometime cannot converge...\n", + "def create_batch(batch_size, input_size, output_size):\n", + " inputs = [random.choices([3,-1,3,0.5,1], k=input_size) for _ in range(batch_size)]\n", + " ys_expected = [p for p in random.choices([1, -1], k=output_size) for _ in range(batch_size)]\n", + "\n", + " return inputs, ys_expected\n", + "\n", + "\n", + "def train(model, inputs, ys_expected):\n", + " # forward pass\n", + " ys_hat = [p for input in inputs for p in model(input)]\n", + " # loss\n", + " loss = sum((y_h - y_e) ** 2 for y_h, y_e in zip(ys_hat, ys_expected))\n", + " # zero gradients\n", + " model.zero_grad()\n", + " # backward gradient descent\n", + " loss.backward()\n", + " for p in model.parameters():\n", + " p.data = p.data - 0.1*p.grad\n", + "\n", + "\n", + "model = build_model(inputs_shape[1], layer_shape)\n", + "inputs, ys_expected = create_batch(inputs_shape[0], inputs_shape[1], layer_shape[-1])\n", + "print(f\"Expected: {ys_expected[:layer_shape[-1]]}\")\n", + "\n", + "for _ in range(50):\n", + " train(model, inputs, ys_expected)\n", + "print(model(inputs[0]))\n", + "\n" + ], + "metadata": { + "colab": { + "base_uri": "https://localhost:8080/" + }, + "id": "CsqUxi2U6Ij-", + "outputId": "9bd3a712-791b-4be6-90a4-672416a64760" + }, + "execution_count": 112, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "[[3, 3, 3], [-1, -1, 3], [1, 3, 0.5], [0.5, -1, -1]]\n", + "Expected: [-1, -1, -1, -1, 1, 1]\n", + "[Value(data=0.060032904940682356, label=), Value(data=0.02798170941222967, label=), Value(data=0.05716545513303159, label=), Value(data=0.03363076737453407, label=), Value(data=0.9901239017092056, label=), Value(data=0.9883133857466925, label=)]\n" + ] + } + ] + } + ] +} \ No newline at end of file diff --git a/Part1_TensorFlow_done.ipynb b/Part1_TensorFlow_done.ipynb new file mode 100644 index 0000000..efa024d --- /dev/null +++ b/Part1_TensorFlow_done.ipynb @@ -0,0 +1,1057 @@ +{ + "cells": [ + { + "cell_type": "markdown", + "metadata": { + "id": "view-in-github", + "colab_type": "text" + }, + "source": [ + "\"Open" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "WBk0ZDWY-ff8" + }, + "source": [ + "\n", + " \n", + " \n", + " \n", + "
\n", + " \n", + " Visit MIT Deep Learning\n", + " Run in Google Colab\n", + " View Source on GitHub
\n", + "\n", + "\n", + "# Copyright Information\n" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "3eI6DUic-6jo" + }, + "outputs": [], + "source": [ + "# Copyright 2024 MIT Introduction to Deep Learning. All Rights Reserved.\n", + "#\n", + "# Licensed under the MIT License. You may not use this file except in compliance\n", + "# with the License. Use and/or modification of this code outside of MIT Introduction\n", + "# to Deep Learning must reference:\n", + "#\n", + "# © MIT Introduction to Deep Learning\n", + "# http://introtodeeplearning.com\n", + "#" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "57knM8jrYZ2t" + }, + "source": [ + "# Lab 1: Intro to TensorFlow and Music Generation with RNNs\n", + "\n", + "In this lab, you'll get exposure to using TensorFlow and learn how it can be used for solving deep learning tasks. Go through the code and run each cell. Along the way, you'll encounter several ***TODO*** blocks -- follow the instructions to fill them out before running those cells and continuing.\n", + "\n", + "\n", + "# Part 1: Intro to TensorFlow\n", + "\n", + "## 0.1 Install TensorFlow\n", + "\n", + "TensorFlow is a software library extensively used in machine learning. Here we'll learn how computations are represented and how to define a simple neural network in TensorFlow. For all the labs in Introduction to Deep Learning 2023, we'll be using the latest version of TensorFlow, TensorFlow 2, which affords great flexibility and the ability to imperatively execute operations, just like in Python. You'll notice that TensorFlow 2 is quite similar to Python in its syntax and imperative execution. Let's install TensorFlow and a couple of dependencies.\n" + ] + }, + { + "cell_type": "code", + "execution_count": 2, + "metadata": { + "id": "LkaimNJfYZ2w", + "outputId": "73e87438-2408-462d-8f7f-81d9c0142222", + "colab": { + "base_uri": "https://localhost:8080/" + } + }, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "\u001b[2K \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m2.1/2.1 MB\u001b[0m \u001b[31m7.0 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n", + "\u001b[?25h Preparing metadata (setup.py) ... \u001b[?25l\u001b[?25hdone\n", + " Building wheel for mitdeeplearning (setup.py) ... \u001b[?25l\u001b[?25hdone\n" + ] + } + ], + "source": [ + "import tensorflow as tf\n", + "\n", + "# Download and import the MIT Introduction to Deep Learning package\n", + "!pip install mitdeeplearning --quiet\n", + "import mitdeeplearning as mdl\n", + "\n", + "import numpy as np\n", + "import matplotlib.pyplot as plt" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "2QNMcdP4m3Vs" + }, + "source": [ + "## 1.1 Why is TensorFlow called TensorFlow?\n", + "\n", + "TensorFlow is called 'TensorFlow' because it handles the flow (node/mathematical operation) of Tensors, which are data structures that you can think of as multi-dimensional arrays. Tensors are represented as n-dimensional arrays of base dataypes such as a string or integer -- they provide a way to generalize vectors and matrices to higher dimensions.\n", + "\n", + "The ```shape``` of a Tensor defines its number of dimensions and the size of each dimension. The ```rank``` of a Tensor provides the number of dimensions (n-dimensions) -- you can also think of this as the Tensor's order or degree.\n", + "\n", + "Let's first look at 0-d Tensors, of which a scalar is an example:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "tFxztZQInlAB" + }, + "outputs": [], + "source": [ + "sport = tf.constant(\"Tennis\", tf.string)\n", + "number = tf.constant(1.41421356237, tf.float64)\n", + "\n", + "print(\"`sport` is a {}-d Tensor\".format(tf.rank(sport).numpy()))\n", + "print(\"`number` is a {}-d Tensor\".format(tf.rank(number).numpy()))" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "-dljcPUcoJZ6" + }, + "source": [ + "Vectors and lists can be used to create 1-d Tensors:" + ] + }, + { + "cell_type": "code", + "execution_count": 3, + "metadata": { + "id": "oaHXABe8oPcO", + "colab": { + "base_uri": "https://localhost:8080/" + }, + "outputId": "6a29b94a-8c11-437d-c350-7ff7bfcaa9f3" + }, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "`sports` is a 1-d Tensor with shape: [2]\n", + "`numbers` is a 1-d Tensor with shape: [3]\n" + ] + } + ], + "source": [ + "sports = tf.constant([\"Tennis\", \"Basketball\"], tf.string)\n", + "numbers = tf.constant([3.141592, 1.414213, 2.71821], tf.float64)\n", + "\n", + "print(\"`sports` is a {}-d Tensor with shape: {}\".format(tf.rank(sports).numpy(), tf.shape(sports)))\n", + "print(\"`numbers` is a {}-d Tensor with shape: {}\".format(tf.rank(numbers).numpy(), tf.shape(numbers)))" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "gvffwkvtodLP" + }, + "source": [ + "Next we consider creating 2-d (i.e., matrices) and higher-rank Tensors. For examples, in future labs involving image processing and computer vision, we will use 4-d Tensors. Here the dimensions correspond to the number of example images in our batch, image height, image width, and the number of color channels." + ] + }, + { + "cell_type": "code", + "execution_count": 7, + "metadata": { + "id": "tFeBBe1IouS3" + }, + "outputs": [], + "source": [ + "### Defining higher-order Tensors ###\n", + "\n", + "'''TODO: Define a 2-d Tensor'''\n", + "matrix = tf.constant([[1,2], [3,4]])\n", + "\n", + "assert isinstance(matrix, tf.Tensor), \"matrix must be a tf Tensor object\"\n", + "assert tf.rank(matrix).numpy() == 2" + ] + }, + { + "cell_type": "code", + "execution_count": 6, + "metadata": { + "id": "Zv1fTn_Ya_cz" + }, + "outputs": [], + "source": [ + "'''TODO: Define a 4-d Tensor.'''\n", + "# Use tf.zeros to initialize a 4-d Tensor of zeros with size 10 x 256 x 256 x 3.\n", + "# You can think of this as 10 images where each image is RGB 256 x 256.\n", + "images = tf.zeros(shape=[10,256,256,3])\n", + "\n", + "assert isinstance(images, tf.Tensor), \"matrix must be a tf Tensor object\"\n", + "assert tf.rank(images).numpy() == 4, \"matrix must be of rank 4\"\n", + "assert tf.shape(images).numpy().tolist() == [10, 256, 256, 3], \"matrix is incorrect shape\"" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "wkaCDOGapMyl" + }, + "source": [ + "As you have seen, the ```shape``` of a Tensor provides the number of elements in each Tensor dimension. The ```shape``` is quite useful, and we'll use it often. You can also use slicing to access subtensors within a higher-rank Tensor:" + ] + }, + { + "cell_type": "code", + "execution_count": 13, + "metadata": { + "id": "FhaufyObuLEG", + "colab": { + "base_uri": "https://localhost:8080/" + }, + "outputId": "b93d4d93-9081-43b4-e2ae-4b8011a25a4a" + }, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "`row_vector`: [1 2]\n", + "`column_vector`: [2 4]\n", + "`scalar`: 2\n" + ] + } + ], + "source": [ + "row_vector = matrix[0]\n", + "column_vector = matrix[:,1]\n", + "scalar = matrix[0, 1]\n", + "\n", + "print(\"`row_vector`: {}\".format(row_vector.numpy()))\n", + "print(\"`column_vector`: {}\".format(column_vector.numpy()))\n", + "print(\"`scalar`: {}\".format(scalar.numpy()))" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "iD3VO-LZYZ2z" + }, + "source": [ + "## 1.2 Computations on Tensors\n", + "\n", + "A convenient way to think about and visualize computations in TensorFlow is in terms of graphs. We can define this graph in terms of Tensors, which hold data, and the mathematical operations that act on these Tensors in some order. Let's look at a simple example, and define this computation using TensorFlow:\n", + "\n", + "![alt text](https://raw.githubusercontent.com/aamini/introtodeeplearning/master/lab1/img/add-graph.png)" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "X_YJrZsxYZ2z" + }, + "outputs": [], + "source": [ + "# Create the nodes in the graph, and initialize values\n", + "a = tf.constant(15)\n", + "b = tf.constant(61)\n", + "\n", + "# Add them!\n", + "c1 = tf.add(a,b)\n", + "c2 = a + b # TensorFlow overrides the \"+\" operation so that it is able to act on Tensors\n", + "print(c1)\n", + "print(c2)" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "Mbfv_QOiYZ23" + }, + "source": [ + "Notice how we've created a computation graph consisting of TensorFlow operations, and how the output is a Tensor with value 76 -- we've just created a computation graph consisting of operations, and it's executed them and given us back the result.\n", + "\n", + "Now let's consider a slightly more complicated example:\n", + "\n", + "![alt text](https://raw.githubusercontent.com/aamini/introtodeeplearning/master/lab1/img/computation-graph.png)\n", + "\n", + "Here, we take two inputs, `a, b`, and compute an output `e`. Each node in the graph represents an operation that takes some input, does some computation, and passes its output to another node.\n", + "\n", + "Let's define a simple function in TensorFlow to construct this computation function:" + ] + }, + { + "cell_type": "code", + "execution_count": 19, + "metadata": { + "id": "PJnfzpWyYZ23", + "scrolled": true + }, + "outputs": [], + "source": [ + "### Defining Tensor computations ###\n", + "\n", + "# Construct a simple computation function\n", + "def func(a,b):\n", + " '''TODO: Define the operation for c, d, e (use tf.add, tf.subtract, tf.multiply).'''\n", + " c = tf.add(a, b)\n", + " print(c)\n", + " d = tf.subtract(b, 2)\n", + " print(d)\n", + " e = tf.multiply(c, d)\n", + " return e" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "AwrRfDMS2-oy" + }, + "source": [ + "Now, we can call this function to execute the computation graph given some inputs `a,b`:" + ] + }, + { + "cell_type": "code", + "execution_count": 24, + "metadata": { + "id": "pnwsf8w2uF7p", + "colab": { + "base_uri": "https://localhost:8080/" + }, + "outputId": "7d2792e2-0018-4345-cd47-c163c2758a7c" + }, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "tf.Tensor(\n", + "[[ 3 5 12]\n", + " [ 3 5 12]], shape=(2, 3), dtype=int32)\n", + "tf.Tensor(\n", + "[[0 1 2]\n", + " [0 1 2]], shape=(2, 3), dtype=int32)\n", + "tf.Tensor(\n", + "[[ 0 5 24]\n", + " [ 0 5 24]], shape=(2, 3), dtype=int32)\n" + ] + } + ], + "source": [ + "# Consider example values for a,b\n", + "a, b = tf.constant([1,2,8]), tf.constant([[2,3,4], [2,3,4]])\n", + "# Execute the computation\n", + "e_out = func(a,b)\n", + "print(e_out)" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "6HqgUIUhYZ29" + }, + "source": [ + "Notice how our output is a Tensor with value defined by the output of the computation, and that the output has no shape as it is a single scalar value." + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "1h4o9Bb0YZ29" + }, + "source": [ + "## 1.3 Neural networks in TensorFlow\n", + "We can also define neural networks in TensorFlow. TensorFlow uses a high-level API called [Keras](https://www.tensorflow.org/guide/keras) that provides a powerful, intuitive framework for building and training deep learning models.\n", + "\n", + "Let's first consider the example of a simple perceptron defined by just one dense layer: $ y = \\sigma(Wx + b)$, where $W$ represents a matrix of weights, $b$ is a bias, $x$ is the input, $\\sigma$ is the sigmoid activation function, and $y$ is the output. We can also visualize this operation using a graph:\n", + "\n", + "![alt text](https://raw.githubusercontent.com/aamini/introtodeeplearning/master/lab1/img/computation-graph-2.png)\n", + "\n", + "Tensors can flow through abstract types called [```Layers```](https://www.tensorflow.org/api_docs/python/tf/keras/layers/Layer) -- the building blocks of neural networks. ```Layers``` implement common neural networks operations, and are used to update weights, compute losses, and define inter-layer connectivity. We will first define a ```Layer``` to implement the simple perceptron defined above." + ] + }, + { + "cell_type": "code", + "execution_count": 34, + "metadata": { + "id": "HutbJk-1kHPh", + "colab": { + "base_uri": "https://localhost:8080/" + }, + "outputId": "2f593761-3c5a-4647-e85a-2931898d0ea3" + }, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "[[0.86002684 0.46097362]\n", + " [0.971696 0.990455 ]]\n" + ] + } + ], + "source": [ + "### Defining a network Layer ###\n", + "\n", + "# n_output_nodes: number of output nodes\n", + "# input_shape: shape of the input\n", + "# x: input to the layer\n", + "\n", + "class OurDenseLayer(tf.keras.layers.Layer):\n", + " def __init__(self, n_output_nodes):\n", + " super(OurDenseLayer, self).__init__()\n", + " self.n_output_nodes = n_output_nodes\n", + "\n", + " def build(self, input_shape):\n", + " d = int(input_shape[-1])\n", + " # Define and initialize parameters: a weight matrix W and bias b\n", + " # Note that parameter initialization is random!\n", + " self.W = self.add_weight(\"weight\", shape=[d, self.n_output_nodes]) # note the dimensionality\n", + " self.b = self.add_weight(\"bias\", shape=[1, self.n_output_nodes]) # note the dimensionality\n", + "\n", + " def call(self, x):\n", + " '''TODO: define the operation for z (hint: use tf.matmul)'''\n", + " z = x @ self.W + self.b\n", + "\n", + " '''TODO: define the operation for out (hint: use tf.sigmoid)'''\n", + " y = tf.sigmoid(z)\n", + " return y\n", + "\n", + "# Since layer parameters are initialized randomly, we will set a random seed for reproducibility\n", + "tf.keras.utils.set_random_seed(1)\n", + "# initialize output number\n", + "layer = OurDenseLayer(2)\n", + "# initilize W and b for this layer\n", + "layer.build((1,2))\n", + "\n", + "# give inputs.\n", + "# for shape(n, m, p), n*m*p = # of scalar in input\n", + "x_input = tf.constant([[1,2.], [9,8]], shape=(2,2))\n", + "# compute outputs\n", + "y = layer.call(x_input)\n", + "\n", + "# test the output!\n", + "print(y.numpy())\n", + "# mdl.lab1.test_custom_dense_layer_output(y)" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "Jt1FgM7qYZ3D" + }, + "source": [ + "Conveniently, TensorFlow has defined a number of ```Layers``` that are commonly used in neural networks, for example a [```Dense```](https://www.tensorflow.org/api_docs/python/tf/keras/layers/Dense?version=stable). Now, instead of using a single ```Layer``` to define our simple neural network, we'll use the [`Sequential`](https://www.tensorflow.org/versions/r2.0/api_docs/python/tf/keras/Sequential) model from Keras and a single [`Dense` ](https://www.tensorflow.org/versions/r2.0/api_docs/python/tf/keras/layers/Dense) layer to define our network. With the `Sequential` API, you can readily create neural networks by stacking together layers like building blocks." + ] + }, + { + "cell_type": "code", + "execution_count": 95, + "metadata": { + "id": "7WXTpmoL6TDz", + "colab": { + "base_uri": "https://localhost:8080/" + }, + "outputId": "69675036-e288-44be-d540-e8839136caa6" + }, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "Model: \"sequential_38\"\n", + "_________________________________________________________________\n", + " Layer (type) Output Shape Param # \n", + "=================================================================\n", + " dense_39 (Dense) (None, 3) 9 \n", + " \n", + "=================================================================\n", + "Total params: 9 (36.00 Byte)\n", + "Trainable params: 9 (36.00 Byte)\n", + "Non-trainable params: 0 (0.00 Byte)\n", + "_________________________________________________________________\n", + "1/1 [==============================] - 0s 63ms/step\n", + "[[ 4.539036 -5.6524134 -1.4490504]\n", + " [ 3.3795772 -4.0207777 -0.8153119]\n", + " [ 5.6984944 -7.284049 -2.082789 ]]\n" + ] + } + ], + "source": [ + "### Defining a neural network using the Sequential API ###\n", + "\n", + "# Import relevant packages\n", + "from tensorflow.keras import Sequential\n", + "from tensorflow.keras.layers import Dense, Input\n", + "\n", + "# Define the number of outputs\n", + "n_output_nodes = 3\n", + "input_var = tf.constant([[3,4], [2,3], [4,5]])\n", + "\n", + "# First define the model\n", + "# if you specify an `Input`, the model gets built here,\n", + "# meaning don't need to use model.build() in later steps.\n", + "model = Sequential(\n", + " [Input(len(input_var.shape)),\n", + " Dense(n_output_nodes)]\n", + ")\n", + "\n", + "model.summary()\n", + "model_output = model.predict(input_var)\n", + "print(model_output)\n", + "\n" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "HDGcwYfUyR-U" + }, + "source": [ + "That's it! We've defined our model using the Sequential API. Now, we can test it out using an example input:" + ] + }, + { + "cell_type": "code", + "execution_count": 59, + "metadata": { + "id": "sg23OczByRDb", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 1000 + }, + "outputId": "d45fc1b7-5ad5-4a8d-ab45-430308f881a2" + }, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "[-5. -2.5 0. 2.5 5. ]\n", + "[-5. -3.88888889 -2.77777778 -1.66666667 -0.55555556 0.55555556\n", + " 1.66666667 2.77777778 3.88888889 5. ]\n", + "[[-5. -2.5 0. 2.5 5. ]\n", + " [-5. -2.5 0. 2.5 5. ]\n", + " [-5. -2.5 0. 2.5 5. ]\n", + " [-5. -2.5 0. 2.5 5. ]\n", + " [-5. -2.5 0. 2.5 5. ]\n", + " [-5. -2.5 0. 2.5 5. ]\n", + " [-5. -2.5 0. 2.5 5. ]\n", + " [-5. -2.5 0. 2.5 5. ]\n", + " [-5. -2.5 0. 2.5 5. ]\n", + " [-5. -2.5 0. 2.5 5. ]]\n", + "------\n", + "[[-5. -5. -5. -5. -5. ]\n", + " [-3.88888889 -3.88888889 -3.88888889 -3.88888889 -3.88888889]\n", + " [-2.77777778 -2.77777778 -2.77777778 -2.77777778 -2.77777778]\n", + " [-1.66666667 -1.66666667 -1.66666667 -1.66666667 -1.66666667]\n", + " [-0.55555556 -0.55555556 -0.55555556 -0.55555556 -0.55555556]\n", + " [ 0.55555556 0.55555556 0.55555556 0.55555556 0.55555556]\n", + " [ 1.66666667 1.66666667 1.66666667 1.66666667 1.66666667]\n", + " [ 2.77777778 2.77777778 2.77777778 2.77777778 2.77777778]\n", + " [ 3.88888889 3.88888889 3.88888889 3.88888889 3.88888889]\n", + " [ 5. 5. 5. 5. 5. ]]\n", + "******\n", + "[[ 0.70886129 -0.63885987 -0.95892427 -0.63885987 0.70886129]\n", + " [ 0.05110034 -0.99602027 -0.67965796 -0.99602027 0.05110034]\n", + " [-0.534056 -0.56094357 0.35584199 -0.56094357 -0.534056 ]\n", + " [-0.84827672 0.13653875 0.99540796 0.13653875 -0.84827672]\n", + " [-0.9497436 0.54853247 0.52741539 0.54853247 -0.9497436 ]\n", + " [-0.9497436 0.54853247 0.52741539 0.54853247 -0.9497436 ]\n", + " [-0.84827672 0.13653875 0.99540796 0.13653875 -0.84827672]\n", + " [-0.534056 -0.56094357 0.35584199 -0.56094357 -0.534056 ]\n", + " [ 0.05110034 -0.99602027 -0.67965796 -0.99602027 0.05110034]\n", + " [ 0.70886129 -0.63885987 -0.95892427 -0.63885987 0.70886129]]\n" + ] + }, + { + "output_type": "stream", + "name": "stderr", + "text": [ + "/usr/local/lib/python3.10/dist-packages/ipykernel/ipkernel.py:283: DeprecationWarning: `should_run_async` will not call `transform_cell` automatically in the future. Please pass the result to `transformed_cell` argument and any exception that happen during thetransform in `preprocessing_exc_tuple` in IPython 7.17 and above.\n", + " and should_run_async(code)\n" + ] + }, + { + "output_type": "display_data", + "data": { + "text/plain": [ + "
" + ], + "image/png": "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\n" + }, + "metadata": {} + } + ], + "source": [ + "import numpy as np\n", + "import matplotlib.pyplot as plt\n", + "# from mpl_toolkits.mplot3d import Axes3D\n", + "\n", + "# Create data\n", + "x = np.linspace(-5, 5, 5)\n", + "print(x)\n", + "y = np.linspace(-5, 5, 10)\n", + "print(y)\n", + "x, y = np.meshgrid(x, y)\n", + "print(x)\n", + "print('------')\n", + "print(y)\n", + "print('******')\n", + "z = np.sin(np.sqrt(x**2 + y**2))\n", + "print(z)\n", + "\n", + "# Create a figure and a 3D Axes\n", + "fig = plt.figure()\n", + "ax = fig.add_subplot(111, projection='3d')\n", + "\n", + "# Plot a 3D surface\n", + "ax.plot_surface(x, y, z, cmap='viridis')\n", + "\n", + "# Add labels\n", + "ax.set_xlabel('X Label')\n", + "ax.set_ylabel('Y Label')\n", + "ax.set_zlabel('Z Label')\n", + "\n", + "# Show the plot\n", + "plt.show()\n" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "596NvsOOtr9F" + }, + "source": [ + "In addition to defining models using the `Sequential` API, we can also define neural networks by directly subclassing the [`Model`](https://www.tensorflow.org/api_docs/python/tf/keras/Model?version=stable) class, which groups layers together to enable model training and inference. The `Model` class captures what we refer to as a \"model\" or as a \"network\". Using Subclassing, we can create a class for our model, and then define the forward pass through the network using the `call` function. Subclassing affords the flexibility to define custom layers, custom training loops, custom activation functions, and custom models. Let's define the same neural network as above now using Subclassing rather than the `Sequential` model." + ] + }, + { + "cell_type": "code", + "execution_count": 104, + "metadata": { + "id": "K4aCflPVyViD" + }, + "outputs": [], + "source": [ + "### Defining a model using subclassing ###\n", + "\n", + "from tensorflow.keras import Model\n", + "from tensorflow.keras.layers import Dense, Input\n", + "\n", + "class SubclassModel(tf.keras.Model):\n", + "\n", + " # In __init__, we define the Model's layers\n", + " def __init__(self, n_output_nodes):\n", + " super(SubclassModel, self).__init__()\n", + " '''TODO: Our model consists of a single Dense layer. Define this layer.'''\n", + " self.dense_layer = Dense(n_output_nodes)\n", + "\n", + " # In the call function, we define the Model's forward pass.\n", + " def call(self, inputs):\n", + " return self.dense_layer(inputs)" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "U0-lwHDk4irB" + }, + "source": [ + "Just like the model we built using the `Sequential` API, let's test out our `SubclassModel` using an example input.\n", + "\n" + ] + }, + { + "cell_type": "code", + "execution_count": 105, + "metadata": { + "id": "LhB34RA-4gXb", + "colab": { + "base_uri": "https://localhost:8080/" + }, + "outputId": "8a9370dd-1c4d-47f1-d6c1-f97f8d14d720" + }, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "tf.Tensor([[ 1.9564508 -0.1900602 1.3396393]], shape=(1, 3), dtype=float32)\n" + ] + }, + { + "output_type": "stream", + "name": "stderr", + "text": [ + "/usr/lib/python3.10/random.py:370: DeprecationWarning: non-integer arguments to randrange() have been deprecated since Python 3.10 and will be removed in a subsequent version\n", + " return self.randrange(a, b+1)\n" + ] + } + ], + "source": [ + "n_output_nodes = 3\n", + "model = SubclassModel(n_output_nodes)\n", + "\n", + "x_input = tf.constant([[1,2.]], shape=(1,2))\n", + "\n", + "print(model.call(x_input))" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "HTIFMJLAzsyE" + }, + "source": [ + "Importantly, Subclassing affords us a lot of flexibility to define custom models. For example, we can use boolean arguments in the `call` function to specify different network behaviors, for example different behaviors during training and inference. Let's suppose under some instances we want our network to simply output the input, without any perturbation. We define a boolean argument `isidentity` to control this behavior:" + ] + }, + { + "cell_type": "code", + "execution_count": 106, + "metadata": { + "id": "P7jzGX5D1xT5", + "colab": { + "base_uri": "https://localhost:8080/" + }, + "outputId": "06ca9aec-b503-456e-ffec-ef4000dc9e2a" + }, + "outputs": [ + { + "output_type": "stream", + "name": "stderr", + "text": [ + "/usr/local/lib/python3.10/dist-packages/ipykernel/ipkernel.py:283: DeprecationWarning: `should_run_async` will not call `transform_cell` automatically in the future. Please pass the result to `transformed_cell` argument and any exception that happen during thetransform in `preprocessing_exc_tuple` in IPython 7.17 and above.\n", + " and should_run_async(code)\n" + ] + } + ], + "source": [ + "### Defining a model using subclassing and specifying custom behavior ###\n", + "\n", + "from tensorflow.keras import Model\n", + "from tensorflow.keras.layers import Dense\n", + "\n", + "class IdentityModel(tf.keras.Model):\n", + "\n", + " # As before, in __init__ we define the Model's layers\n", + " # Since our desired behavior involves the forward pass, this part is unchanged\n", + " def __init__(self, n_output_nodes):\n", + " super(IdentityModel, self).__init__()\n", + " self.dense_layer = tf.keras.layers.Dense(n_output_nodes, activation='sigmoid')\n", + "\n", + " '''TODO: Implement the behavior where the network outputs the input, unchanged,\n", + " under control of the isidentity argument.'''\n", + " def call(self, inputs, isidentity=False):\n", + " x = self.dense_layer(inputs)\n", + " '''TODO: Implement identity behavior'''\n", + " if isidentity:\n", + " return inputs\n", + " return x" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "Ku4rcCGx5T3y" + }, + "source": [ + "Let's test this behavior:" + ] + }, + { + "cell_type": "code", + "execution_count": 108, + "metadata": { + "id": "NzC0mgbk5dp2", + "colab": { + "base_uri": "https://localhost:8080/" + }, + "outputId": "b3eed6b7-d3f6-4e23-f44c-2c3922100e17" + }, + "outputs": [ + { + "output_type": "stream", + "name": "stdout", + "text": [ + "Network output with activation: [[1. 2.]]; \n", + " network identity output: [[0.79819345 0.33411932 0.2267358 ]]\n" + ] + } + ], + "source": [ + "n_output_nodes = 3\n", + "model = IdentityModel(n_output_nodes)\n", + "\n", + "x_input = tf.constant([[1,2.]], shape=(1,2))\n", + "'''TODO: pass the input into the model and call with and without the input identity option.'''\n", + "out_activate = model.call(x_input, True)\n", + "out_identity = model.call(x_input)\n", + "\n", + "print(\"Network output with activation: {}; \\n network identity output: {}\".format(out_activate.numpy(), out_identity.numpy()))" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "7V1dEqdk6VI5" + }, + "source": [ + "Now that we have learned how to define `Layers` as well as neural networks in TensorFlow using both the `Sequential` and Subclassing APIs, we're ready to turn our attention to how to actually implement network training with backpropagation." + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "dQwDhKn8kbO2" + }, + "source": [ + "## 1.4 Automatic differentiation in TensorFlow\n", + "\n", + "[Automatic differentiation](https://en.wikipedia.org/wiki/Automatic_differentiation)\n", + "is one of the most important parts of TensorFlow and is the backbone of training with\n", + "[backpropagation](https://en.wikipedia.org/wiki/Backpropagation). We will use the TensorFlow GradientTape [`tf.GradientTape`](https://www.tensorflow.org/api_docs/python/tf/GradientTape?version=stable) to trace operations for computing gradients later.\n", + "\n", + "When a forward pass is made through the network, all forward-pass operations get recorded to a \"tape\"; then, to compute the gradient, the tape is played backwards. By default, the tape is discarded after it is played backwards; this means that a particular `tf.GradientTape` can only\n", + "compute one gradient, and subsequent calls throw a runtime error. However, we can compute multiple gradients over the same computation by creating a ```persistent``` gradient tape.\n", + "\n", + "First, we will look at how we can compute gradients using GradientTape and access them for computation. We define the simple function $ y = x^2$ and compute the gradient:" + ] + }, + { + "cell_type": "code", + "execution_count": null, + "metadata": { + "id": "tdkqk8pw5yJM" + }, + "outputs": [], + "source": [ + "### Gradient computation with GradientTape ###\n", + "\n", + "# y = x^2\n", + "# Example: x = 3.0\n", + "x = tf.Variable(3.0)\n", + "\n", + "# Initiate the gradient tape\n", + "with tf.GradientTape() as tape:\n", + " # Define the function\n", + " y = x * x\n", + "# Access the gradient -- derivative of y with respect to x\n", + "dy_dx = tape.gradient(y, x)\n", + "\n", + "assert dy_dx.numpy() == 6.0" + ] + }, + { + "cell_type": "code", + "source": [], + "metadata": { + "id": "xeOb0-enLIUW" + }, + "execution_count": null, + "outputs": [] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "JhU5metS5xF3" + }, + "source": [ + "In training neural networks, we use differentiation and stochastic gradient descent (SGD) to optimize a loss function. Now that we have a sense of how `GradientTape` can be used to compute and access derivatives, we will look at an example where we use automatic differentiation and SGD to find the minimum of $L=(x-x_f)^2$. Here $x_f$ is a variable for a desired value we are trying to optimize for; $L$ represents a loss that we are trying to minimize. While we can clearly solve this problem analytically ($x_{min}=x_f$), considering how we can compute this using `GradientTape` sets us up nicely for future labs where we use gradient descent to optimize entire neural network losses." + ] + }, + { + "cell_type": "code", + "execution_count": 51, + "metadata": { + "attributes": { + "classes": [ + "py" + ], + "id": "" + }, + "id": "7g1yWiSXqEf-", + "colab": { + "base_uri": "https://localhost:8080/", + "height": 905 + }, + "outputId": "6f3011e8-45f8-4b9e-e28e-da0a56cb2150" + }, + "outputs": [ + { + "output_type": "stream", + "name": "stderr", + "text": [ + "/usr/local/lib/python3.10/dist-packages/ipykernel/ipkernel.py:283: DeprecationWarning: `should_run_async` will not call `transform_cell` automatically in the future. Please pass the result to `transformed_cell` argument and any exception that happen during thetransform in `preprocessing_exc_tuple` in IPython 7.17 and above.\n", + " and should_run_async(code)\n" + ] + }, + { + "output_type": "stream", + "name": "stdout", + "text": [ + "Initializing x=[0.5]\n", + "Initializing b=[0.5]\n" + ] + }, + { + "output_type": "display_data", + "data": { + "text/plain": [ + "
" + ], + "image/png": "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\n" + }, + "metadata": {} + }, + { + "output_type": "display_data", + "data": { + "text/plain": [ + "
" + ], + "image/png": "iVBORw0KGgoAAAANSUhEUgAAAZgAAAGPCAYAAABh6MZ2AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAEAAElEQVR4nOz9d5hk6V3eD39O5dhdnXOO05Nz2JUNQmKNbFkCA/oJjAIgbPBifl64XklGAbBQsEBeXiEsjK0XGywjIwQGKyEtrNgwu6ud6Z7u6Zxzrk6Vq8457x8950xVdXV3VdfprprZc1/XXNJ2Vz/nqfTc55vuW5BlWUaHDh06dOjQGIZcb0CHDh06dDye0AlGhw4dOnQcC3SC0aFDhw4dxwKdYHTo0KFDx7FAJxgdOnTo0HEs0AlGhw4dOnQcC3SC0aFDhw4dxwKdYHTo0KFDx7FAJxgdOnTo0HEs0AlGhw4dOnQcC3SC0aFDhw4dxwKdYHTo0KFDx7FAJxgdOnTo0HEs0AlGhw4dOnQcC3SC0aFDhw4dxwKdYHTo0KFDx7FAJxgdOnTo0HEs0AlGhw4dOnQcC3SC0aFDhw4dxwKdYHTo0KFDx7FAJxgdOnTo0HEs0AlGhw4dOnQcC3SC0aFDhw4dxwKdYHTo0KFDx7FAJxgdOnTo0HEs0AlGhw4dOnQcC3SC0aFDhw4dxwKdYHTo0KFDx7FAJxgdOnTo0HEs0AlGhw4dOnQcC3SC0aFDhw4dxwKdYHTo0KFDx7FAJxgdOnTo0HEs0AlGhw4dOnQcC3SC0aFDhw4dxwKdYHTo0KFDx7HAlOsN6HhjQZZlRFEkGAxiNBoxmUwYjUYMBgOCIOR6ezp06NAQgizLcq43oeONAUmSiMVixGIxQqGQSiiCIGAwGDCbzSrpCIKgE44OHY84dILRceyQZRlJkohGoygft0gkgsFgQJZl9feyLKvEohCNEuHohKNDx6MHnWB0HCtkWVajFtiNVmRZJhqNpiSMgwhHiXCUlJoOHTryGzrB6Dg2KLUWJQWmkIISzQCHRiU64ejQ8ehC/1bq0BxK1BIMBnnuueeQJGkPAaR7X6OQk8lkUglFEARisRiBQICdnR22t7fx+XyEQiGi0SiSJB3H09KhQ0eG0LvIdGgKJf0limLaJJIJlAhGISwlwolGo0QiEfX38fUbPcLRoSM30AlGh2aQJIlIJKJGLCbT7sfrOLOwqQhHScEpaTjlfwsKChK61HTo0HG80AlGR9ZQZluULjFlpkUhlpMs8yn1mfi9LSwssLGxQVdX154IJ75LTYcOHdpCJxgdWSE+JQYkDEwq/5vLPhIluhEEAbPZnBDhKCk1g8Gwp2lAJxwdOrKHTjA6jgwlalFSYsmH8n4Ek4vDW9lDqggnFeEk13B0wtGhI3PoBKMjYyTPthwk8xKfKkv+eT4gnnCUfSq1pHA4rBOODh1ZQCcYHRlBudtXWoEP687aj2BOEumSgfI4nXB06NAGOsHoSAvxqaT9UmKpkA8Ec1SkIhzlXzgcJhKJAOiEo0PHPtAJRsehOKiQfxjy5aDVguTi9dCMRuMewomPcOKFO3WlaB1vVOgEo+NAKFGLKIpHOigf5QjmMBxEOKFQSH2M8nubzaZbE+h4Q0EnGB0pocy2xGKxjFJiyXicCSYZ+xHO+vo6Y2NjXLlyJaVStE44Oh5X6ASjYw9kWcbr9RIKhSguLs7qAMwHgsnV4R2vMqAMdyq1LCXCUURAdcLR8ThCJxgdCVBmW1ZXV9nY2KC0tDSr9fKBYCC3w57wkGxSRTiSJBEOhwmFQjrh6HisoBOMDiD1bItWhfFcH+65xn7PP9lETSEcURQRRXHfpgHdfE3HowKdYHTsmW1RDjWdYE4W+ylFK7Uw3e1Tx6MGnWDewEi2Mo4/qLQihnwgmHw4fI/aIJGKcGKxmOoIul/TgA4d+QCdYN6gSJ5tSb4LNhgMmhh3HSQVc5LEk2uS0wKZEI7u9qkjH6ATzBsQ6cy2aEkAj8Phng2O6/kfRjiQWmVAJxwdJwWdYN5AyGS2RSuC0aqW86jjJNJ0B7l9jo+PYzabqa6u1glHx4lBJ5g3CDKVe3mcIph8qMHkAvGEEw6H1Z/Hu30m20vrbp86tIROMG8AJFsZp3OAHHcN5qSRyz3ky/NPxwtHd/vUoSV0gnmMsZ+VcTp4nLrI8gG5PqQVgolHuuZrutunjqNCJ5jHFJIkEYvFjqSADMdLMLIss7q6itVqxe126wfWCSAVwSRDd/vUoTV0gnnMcNBsSyY4rhRZOBymt7eXnZ0dRFFEEASKioooKiqiuLgYu92u+YGV6wMwHyK4dAgmGbrbp45soRPMYwRZlvH5fHi9XsrLy7Oa8D6OCMbr9XLv3j08Hg83btxAEAR8Ph8bGxusrKwwNjaG2WxWCaeoqAibzZb1HiD3h3yuD92jEEw8dLdPHUeBTjCPCZSoZXNzk5GRESorK7NaT0uCkSSJ8fFxJiYm6OjooLa2Vm2VLiwspLCwkMbGRkRRZGtri42NDebn5xkcHMThcCQQjtlsznpPJ41ck5sCLQ963e1TRzrQCeYRR/Jsi9Fo1CS1pdX8iizLzMzMIMsy169fp6CgYN91jUYjxcXFFBcXAxCLxdjY2GBjY4PJyUnu37+Py+WiuLiYoqIiCgsLMZkO/wjrB1r2EcxhSMftMxQKEYlEKC8v190+3yDQCeYRRqrZFuXLnS2UyCMbeL1eNjY2cDqdXL9+PS0yiIfJZKKsrIyysjIAIpGIuubw8DDhcJiCggI1uiksLMzbocFcH6KSJJ3oHlIRztbWFmtraxQUFKiPUZSidWuCxxM6wTyi2G+2RQtiUNY5KlHJsszExAQTExM4nU51ejxbWCwWKisr1fRfMBhkY2MDr9fL/Pw8oiji8XhUwonvUHujz8FAbklOubZCKPvZS+uE83hBJ5hHDIfNtuRaZj8SiXDv3j2CwSDXrl1jcnIy673sB7vdjt1up7q6GlmW8fv9akptampK7VAzmUyIonjsaaJ8Rj48d+VmCA5OqcUTjm4v/WhDJ5hHCOnIvWjVXqwQVSYHk9IlVlRUxIULFzCbzSc2aCkIAi6XC5fLRV1dHZIksbOzw8bGBsvLywQCAV566SW1fqNlh1q6+8sl8iGKOihNtx/h6G6fjzZ0gnlEoEQt6YhUapUig/TufONTYh0dHdTV1WnuK5MpDAaD2qHmdDqZnJyktbVV7VAbGhrCZrOp8zcejweLxXIse8mHw12Jdh+VPSS32B/k9qkTTv5CJ5g8Ryor44O+QPFKulrMPRx2OEYiEXp7e/H7/Vy7do3CwsI96+TDAQvs6VDb3NzE6/UyOTmJ3+/H5XKp0Y3H49GkbpQvyIf3IJtGg/2UonW3z/zG4/MNegyRbGWczt2f8hilZfmoSGed+MHJW7dupZxRSUUw+fClN5lMlJaWUlpaCuwSpVK/GRkZ0bxDLdfPNx9qMFpGUYd54eiEkx/QCSYPES/3kokCMmhHMAdFMPEpsfb2durr6w9M2eX67jmd185isVBRUUFFRQXwsENtY2ODhYUFYrHYvh1qhyHXz1/ZQ64P1uNslc6EcHS3z5ODTjB5hkx9W5KhVVvufusclhJLtU4+HLCZ4qAOtenpaQA8Ho/aNOBwOA58n3J9uOcDwciynNVNTyY4iHDGxsawWq26+doJQCeYPEI6VsaHIT6CyQapCGZjY4Oenp4DU2KpkA8Ek80eUnWoKZpvq6urjI2NYTKZEkQ7T7JDLR3kA8Gc9LBnPOIJJxQKqR2OilI06PbSxwGdYPIA8YX8TH1bkqH8nRYEo3SkybLM5OQk4+Pjh6bEUq2T6nDP9WGXDQwGAwUFBRQUFKgaatvb22o6bXh4WO1QKyoqUt0jc4l8IJh86GSDh+ljpe0ZEtPS+7l96oSTOXSCyTGUlFh/fz+CINDZ2Zn1QaDlsGU4HGZgYCDtlFiqNbRom84Gx32wGo1GlUyam5vVDjUlnebz+TAajYyOjuasQy0fCCZ+0DLf9qHUZxQc5vap20unB51gcojk2RatUghaDVsCdHd3U1RUlFFKLB77RTD5kDY7LiR3qE1NTbG6uoooioyOjhIKhXC73Qmincd98OYDweTDHiA9okuHcHS3z8OhE0wOkGq2xWg0qv+dLbKNGpSUmCRJ1NTU0N7enlXKLh/IJJd7MBqNWK1WOjs7gdQdaoWFhWr95jhcPvPhcM+XCEapcWaCdAlHtyZIhE4wJ4zk2Zb44qNWUUc2KbL4LjGTyURlZWXWA5v5QDC5RvxrmNyhFggEVNHOmZkZgISWaKfTmfVBlS8Ek+s9KPvIlujiCUc3X9sfOsGcEA6zMtaaYI6yVnKX2IsvvqhJu3OuCSafv9SCIOB0OnE6ndTW1iLLsqqhtr6+zvj4eEKHWlFREXa7/cjXyiXyqciv5T7iNdRAJ5x46ARzAkiebUk1TawlwWSaItuvS0yLAv3j/gVKB5kQrCAIaodaQ0MDkiSpLp+Li4sMDw9jtVoTRDvT0VDTI5jEfRznPM5BhBPv9qm8Jw6H47ElHJ1gjhnpzrbkKkV20OCkFt1o+dBFlg846sFhMBhUIgH2dKj19/fjdDpVwtmvQy0fCOZxjWAOQzzhxCtFr66uMjU1xeXLl4GHXjmPk9unTjDHhGQr43REKk86RXbY4KQW6a2DJGROErlO02mFVBpqCuHEd6jFa6gph1quD6t8imBySXRKdkCWZTVddpj5mtIIlA+vXybQCeYYcBS5F4PBoD4+WxxGDOkOTmqVIsv14Z7rL+VxPn+LxUJ5eTnl5eUAhEIhtUNtYGBA7VCTZRmfz4fVas3Z4ZoPEYxy45frfcBuN1uyAOd+5mvveMc7+MAHPsB73vOeXG45Y+gEozH2szI+DFoNRypr7UcMkUiEvr4+fD7foYOTWqXIck0w+YCTIjmbzUZVVRVVVVUJHWobGxsMDg4iy3JCw4AWHWrpIh8iGOWzmE8Ek4xUhLOzs3Pk5o5cQicYjaDcGe1nZXwYTiJFtrGxwb179ygsLExrcFKrFFk+SMXkkuRydW2lQ81utzMyMsL169cJh8PH0qGWDvIhglG+FyclunkQ9iOYVPD7/TidzmPekfbQCUYDSJJELBY7sgKy8jfH1UUWnxJra2ujoaEhrf0dJ8HoODnE37Wn06EWL9qppctnrmsfyh4gPyKYTLrZ/H4/brf7mHekPXSCyQKHzbZkAi1rMPGprUxSYsl4XGowb3Ts9/qn6lBTCGdmZoaBgQGcTmdChJONhlo+pMjiB5xzjUxqQYFAQI9g3khQiqbb29sUFxdnRS5wPCmyTFNiqdY5zgjmpIgn14dJrju4lNf5sD2YTCZKSkooKSkBIBqNqvWb8fFxgsFgyg61TPaR68ghGyuM49iL1Wo99HGKH5HL5TqBXWkLnWCOACVqWVtbY2pqilu3bmW9ppYEA7C+vs7Y2FhGKbFkHFeKTJZl5ufnCQQClJaWHov2VjLeyFFUugSTDLPZvG+H2uDgINFolIKCAnUGx+12H0gg+RLBHDvJyT7ABsLBx2u6NZhIJEIsFtNTZI87kmdbTCZTzuVdkqF4y4uiyNWrV/F4PEdeS6shyfjDPRaLcf/+fTY2NnC73czNzQGoOX+l0Jzrg+hxwlEJJhnJHWrBYBCv16um1GRZTnD5jO9Qy5fureMmGEGaQ5C3kYxdae0lHYLx+XwAegTzOCPVbIvRaNR8+j6bdIqSEjMYDFRWVmZFLvF70moNn89Hd3c3VquVa9euqYZPOzs7eL1elpeXGRkZSZBCKS4uPpJNQDxyTVaPSoosEygSJw6HQ9VQ8/l8CR1q8T45Su0v1+/FccrECNIoRvEVYuafSevx6UYwfr9ffb0fNegEkwb2m23RsjCvfNCO8gWQZZmpqSk1JaZ8ILOFVgV6WZZZXFzk/v371NfX09bWpqYZ47W3FHfIzc1NvF4vU1NT9Pf3q3n/4uLijPP+Oo6HYJIhCAJutxu32019fT2SJKkun0tLS4yMjAAwOjpKSUkJRUVFadUftMZxRTAG8R7m6P8kbP1k2n+TCcGc5LySltAJ5gAcNtuidQQDmROM0iW2s7OjpsSGhoY0c7TM9vkpqZT+/n7Onz+v5vP3W9doNCYUmpW5Da/XmzCZXlxcTHFxMS6XK60v3hu9BnPSh5PBYMDj8eDxeGhqaiIYDHL79m3MZjOzs7N7OtQ8Hk/WkWo6OA6CMcZexhL5HEH7Hx1ad4lHul1kPp9PJ5jHDenIvWjd+QX7H7ypoKTECgoKeOKJJ9QvqCAImkRW2UYwoVCIsbExRFHkySefPFKIb7VaqayspLKyUp1MV/L+U1NTaqutQjg2m+3I+z1O5DpFluvDSfl8t7W1IQjCng61QCCwx+XzOCJVrWVijLG/xRr+BCHbH4JQlPFe0nmOj2qLMugEkxLJVsb7fTmNRiOiKGryBVb+Ph2CUVJio6OjtLe37+kS00p2Jpt11tfXuXfvHm63G1mW95DLUbvaFO+Uuro6NQ3j9XrVQUG73Z7QMJAPvum5jp7ygWCUDjJlH8kdauFwWL1xGBwcJBKJ7HH51IIYtIxgTNGvYYl8irD1N5GMpzL++3QJxufz4XA4cv4eHgU6wcQhlZXxYQrIyt9pQTDpRETxKbFr166lLORrFVkdJUUWrxpw6tQpbDYbAwMDWe8lFeLTMLDboZZqbsPpdCJJUk4nyXN9OOT6+oe1KFut1j0dasp7OTc3hyRJqstncXHxkVNGWhX5zZE/xhL9PFHTv0Q0ve1Ia6S7l0d1BgZ0glGRbGWczkGkfDi0CrsPI4b4lNitW7f2lfHQkmAyufuORqP09vYmqAasr69nvY90YTKZKCsro6ysDNhN0Xm9XlZWVojFYrzwwgv7ttE+zsiHCCaTIcv4DrWampqEDrWNjQ0mJycTVAgyaW3P+iZDljFHP48l+t8RDdeJWH75iMvIGUUweorsEUW83MtRFJAhs7rJYeulWuuwlFgytOr+yiRFtr29TXd3Ny6XK0E1IJeGYzabjerqalwuF9vb21y8eHGP0GN8O/RxdTXpKbLshiwP6lBTWtstFkuCy+d+72VWN4OyiCXyKcyxv0QS6gjZPpVRUT8emYhu6jWYRxRH8W2Jh/JYLTXEkg/jdFJi6axzFKRLDnNzcwwODtLc3Exzc3PCa5gPhmNK7j/+kBJFUa3fKPtPxxkymz3kCvlAMFrKxCR3qCmt7RsbG2qHmsPhSHgvlRueI0cwcgRr+KOYxO8i4yRk+08gpK/rl4z4M+cw6CmyRxDpWhkfBEEQNG9Vjl9rc3OTnp6eQ1Nih61zVBxGMKIoMjg4yPLyMhcvXlRdFpPX2O8OPpeHXvwQYEtLi9rV5PV6GRkZIRwOH0uRORfIB4I5TpmY5Nb2aDSqEk58h1pRURHBYDDzdmg5iDX8a5jEV5ARCFt/G9nQlNWeMyEYn8+nE8yjgkytjA/DcYhUxqfE2traaGxszDiy0mJPB6XIAoEAPT09CILAE088sW97cL6oKR+2h+SuJkUGxev1Mjs7C5AgY5+JnE2un3+ur6/s4aQI2mw2J9TilFkq5QZCFEUCgYD6fhYUFOy/N3kbW+hXMEq9AETNv4xoelPWe4x3szwMfr8/IxX0fMIbimBkWWZzcxNJknA4HJqoqiqtylrAYDAQiUS4e/duRimxVOtoNWiZap2VlRV6e3uprq6ms7PzwIMjXwgmU9jtdmpqatQisyJns7KywujoqOqboqRhDosuc50iy3X0lcsOvuRZKpPJhNPpTNmhVlRU9HB4V1rDFn4aozQKQNT0dqJmbSyLM+lmCwQCVFdXa3Ldk8YbhmCU2Zbx8XGsVivt7e2arKtlBCNJEkNDQ3g8noxSYsnQKoJJZVw2OjrK9PQ0Z86coaqqKq01ck0wWrSQp5Kz2djYYHp6mv7+flwulzrsmW9yNvmQIsuHPSj7sFqtCTcPfr9fncFROtQqSqN0Vv1XjOZdchENF4hY/j1o9BwyaTZ4VN0s4Q1AMMmzLSaTSbOIA7TRI1NSYj6fj8rKSs6fP5912k7rCCYcDtPb20soFOLmzZtp54TzgWC0RnLOPxKJJAwJRqPRBDmbXDs55sPrn+vXQEHywS4IAi6XC5fLpXaoBXd6sYu/j908DEAwXMzQygdwF65r1m2YqV2yXoPJQyTPtihF+Ugkotk1si3yRyIR7t+/z/b2NoWFhZSUlGR9p6dVVKUQ1cbGBj09PRQVFXHx4sWMuqvyhWCOcw8Wi2WPnI2S75+amkKWZWw2m6q9dZy+96mQD9FDPnjBKPs4iOhMcj/Flv+CSeoGQMbJpvAZjOZCtdvQ4XAkzOAcRUMtE4IJBAI6weQTDrIy1rLrC7I7zJUuMbfbza1bt7h//37OJvD3g9/v5/XXXz+ycVm+EMxJIV7Opra2FkmS6O/vJxKJqHI2NptNjW5OQuQxHwgmH+pAcDDBGMTXsIR/F4O829QhYyRs/QwFzksUFKF2Gyrp0cnJSe7fv5/g8unxeNIijkwjGD1FlidInm1JtjLWsih/1PVkWWZ6eprR0VFaW1vVLjGtI49sEIvFVNfJa9euqb7tmSIfCCaXh6vBYMBisWC322ltbSUWi6l2BMk2xEr9RuuDOB8IJp8imFQHuzH2d1jCnwQsCIQBiFg+iGi6mfC4gzrUhoeHCYfDCS6f+3WopUswj7JdMjxmBJPObIvWBJNpDSYajdLX18f29jZXrlxJOLi1JJhs1lGMwQB1FuSoiHc0zIcDJldQnrvJZKK0tFSdGVJEHr1eL/39/cRisYR2aC3kbPLhtc/nCMYU/Wsskd9GElowyrt1l6jpp4mZ/8Wh68V3qAEJLp9Kh5pSj4vvUMskgvH5fI+kXTI8JgSTyWzLcRBMuod5ckosuUssHwhGMQZraGjA6XSqlsbZIteHXK6jqP2QLPIY39E0MTGByWRKsCM4SoE516895FcEE08wpuifYo38J2KGS5ikuwDEjD9ExPL/Hmn95PZ2v9+foKEmCAJFRUUqwaTz3uhSMTlEpnIvWrpQQno1nf1SYqn2lguRStj94g0PDzM/P68agy0uLma9n3jF6Tcq0n3uqTqatra28Hq9zM/Pq3I2CuGkK2eTDwSTLxGM2kUmy5ij/xlL9L8hGs6q5CIazhK2/hYI2e81/v1U7CV2dnbY2Nhgfn6ecDjMyy+/nBCxJt9A6CmyHGI/K+ODcBxF/oMI66CUWKq1chHBhEIhenp6EEWRW7duqd4tWrU7Q24JJteH61H3EK8YDCSYdI2OjhIKhdR8/0FyNvlAMPkVwYAl8mnMsa8iCQ0YpN20mCTUPNAYOx7TOoPBQGFhIYWFhYRCIbXdXSGcoaEh1c9IUftW9vyopshyf0txBCizLeFwOGO5l+Mo8u93mG9ubvLSSy8hyzK3bt06tJahdQSTzqG+vr7Oyy+/jNPp5MaNGwnGYFoU6ONrMG9UaPXcFTmbjo4Obt68yY0bN6isrMTv99Pb28sLL7xAb28vc3Nz+P1+9br5QDD5EsHIcoQi4ycxx76KjAdkPwIRZAoJ2X4/Y1fKo0IURcxmM8XFxbS0tHDlyhXe9KY30dLSgiAITE5O8uSTT/Lkk09SVVXFyy+/jM/ny/q6X/jCF2hsbMRms3H9+nVee+21Ax//7LPP0tHRgd1up66ujn/37/4doVAo7es9chGMJEnEYrEjKyAfRw0mea5GSYmNjIxkpCWmVfouHSM0WZaZmJhgYmKCU6dOUVtbu+cxjxPB5Pr6x4FUcjYbGxusrq4yNjamHmDp3mwcJ/Ji0FIOcrnlS9gYQsaEJJRglMeRsRKy/R6yof7EtpKqyJ/sZ/Q3f/M3/Pmf/zm/+7u/y6/+6q8yOzvLjRs3+MhHPsJTTz2V8TW/8pWv8Mwzz/DFL36R69ev8+yzz/LUU08xPDysavDF48tf/jIf+tCH+NKXvsStW7cYGRnhfe97H4Ig8LnPfS6taz4yBHPQbEsmOO4IJj4ldvXq1Yw6sLRMkcH+X+p4Y7Dr169TUFCQch0t5mlyfeecLzju1yFezqahoQFRFNX6zfLyMuFwmNdeey1Bwv4k5WxyHsFI21hDv4bTMwSAaDiNSbqHjIGw9ZNIxrMnu500tMhqa2t561vfyu/8zu8wNjbG9PQ0zz33nEpAmeJzn/scH/jAB3j/+98PwBe/+EW+/vWv86UvfYkPfehDex7/8ssv88QTT/BTP/VTADQ2NvLud7+bV199Ne1rPhIEkyz3clRygYdRglZpg3hSOKxLLJ21tJJ4gdRGaPsZgx3Xfg7ay0kh1ySXi+jBaDSqtRmLxcLm5iYVFRV4vV6GhoZUOZt4O4LjfJ0kSdLUXycjyOtYwp/GJN8BICpcwCz1ABCxfAjR9AMnvqV0tcgUqX5BEGhqauLnf/7nj3S9SCTCnTt3+PCHP6z+zGAw8Ja3vIXbt2+n/Jtbt27xp3/6p7z22mtcu3aNiYkJvvGNb/AzP/MzaV837wkmfrZFGUbMBspdg1be3EajkVgsxtTUVMYpsWRoHcEkH2wHGYOlgpYpMh25gyzLGI1GKioqqKioUD3vlfmb6elpBEFIcPfUWs4mV0V+QVrAHPldTNLfAwIbvlo8rh4AIuZfSGvW5TiQ7hxMIBBIqIseFWtra4iiSEVFRcLPKyoqGBoaSvk3P/VTP8Xa2hpPPvmkepP/r//1v+bf//t/n/Z185ZglNmW+fl51tbWOH36tCYfUOVNzWTQ6SDIsszW1hY+ny/jlFgytKrBJEcNoigyMDDA6uoqly5dUkUa01nncanB5Bq5ngFKdhlVPO8VORvFjmBpaYmRkRFsNluCHUG2cja5SJEJ0jjm8CcwSsMICIjU4LavIABR048RNf/Cie4nHumeP4pMTC4+P88//zyf/OQn+YM/+AOuX7/O2NgYv/Irv8J/+A//gY9+9KNprZGXBBM/2yKKIn6/X7MXOD6CyRabm5uMjY2pXWJHlddXoGUXmVI/CQQCdHd3YzQauXXr1r7GYMe9n2SCUSyLDzR70hC5JLhck+th6eD49tmmpqYEOZt4va14O4JM37OTjmAEsR9L+IMYZAkDYWQKgTAmY4SY8c1ELB/STHr/KEiXYLRysywtLcVoNLK8vJzw8+XlZVWFIBkf/ehH+Zmf+Rk1LXf27Fn8fj+/8Au/wK//+q+n9RnIO4JJnm3RWl5fSbNls2b84GRFRQU7OztZkwto7465trbG6OgoNTU1dHR0ZHwoaKUjlryOQnp+vx8gYVLd4XBofhC90dN0mb6H+8nZbGxsqHI2Ho9Hfc/SucM+yQjGIH4fS+gZZEM9BnkIGcuDjrEJ1rdbsVV+AoTc+vWkm6LXasjSYrFw+fJlnnvuOd75zneqe3juued4+umnU/5NIBDY854pe073M5U3BKOkxJQuMaX9WKlxaIlsOsmSBydjsRjb29ua7EsrglFsl0dGRtI2BkuF4yCYtbU17t27R2VlJRcvXiQUCuH1ellfX2d8fFxtrS0pKdEkNZMvyHWKLJvDPZWcjWJHMDk5idFoTEinpYqSTyqCMcT+Hkv4Q0iGs5gfSO5LQjtG+T4RuZ3uyZ/jZlX2fi7ZIpMUmRY1GIBnnnmG9773vVy5coVr167x7LPP4vf71a6y97znPdTU1PCpT30KgLe//e187nOf4+LFi2qK7KMf/Shvf/vb0y4v5AXBHCT3onVbsbL+Udbc2tqip6dH7cCyWCyqx7dW+8qWYMLhMPfu3UOWZc6ePXtkctFqP/Cw3XliYoLx8XG6urqoqqoiGo0mSKMoTpHKwdXf36+mZkpKSvadVM935HuKLBOkkj/Z2traM40eb0dgMplOZA7GGP0bzJHfRDJ0PZB+EVSNMUloZC3ySWQWj3UP6UC5ATzJCAbgXe96F6urq3zsYx9jaWmJCxcu8K1vfUst/M/MzCS8Rx/5yEcQBIGPfOQjzM/PU1ZWxtvf/nZ++7d/O+1r5pxglKhlv4l8rVNkkDlpxafEWlpaaGpqUvepZVorWxmbeGMwi8WSUb0lFbSU2h8eHsbv93Pt2jUKCwvV5xl/+CU7RSqpmfX1debn55FlOSGdlmmnUz5MtOcCx0lw8XI2zc3Nql+K1+tNkLMJhUIEg8FjIxpj9H9iifwOklCPURrdLeobzj8glypCtj8gFjRhMCwfvtgxI/5G+jBorUP29NNP75sSe/755xP+22Qy8fGPf5yPf/zjR75ezggmebblMHl9LQ+HTA7yaDTK/fv32draSqklpmWEdVSyiifA9vZ26uvreeGFFzQbkszmtff7/epNRHIjxGHrJqdmlE6n5eVlRkZGUt4pH/Q8colcp8hO6vrJfinBYJCNjQ3GxsaYnZ1ldnY2Qdwx65qbLGOK/mfM0T9CphCDHMJABFHowCD1IVFCyPYHyIYKJGn1RIdL94NyXqQbwRQXFx/3lo4NOSGYVCmx/aBIWmtNMOmQQqqUWDK0LsxnulYsFuP+/ftsbGwkEKCWU/hHnRlaWVmht7cXg8HAqVOnsmqEiJ9Ub2xsJBaLqXUA5U5Z8d0oKSlRh9PyAfmQIstVatFut2O325mbm6OpqQmbzYbX62VtbU2tucVHpRl9RmQJU+RZzLE/QcYIQjkGeRxJqMMgzyJRRsD6/8XwQAImL+RqeDhkmc7nMxAIUFdXdwK7Oh7khGCUFzadiXzlrjQWi2nSqQWHE4wsy8zMzDAyMrInJZYMhRS0IMBMSUExBrNardy6dStB6luLKfyjSu3Lssz4+DiTk5OcOXOG4eHhrPaRCsm6Tcrg4Pr6OtPT0xgMBvXQUlIMb+QUWa6ft3K4u91u3G73Hjmb2dlZBgYGcLlcCXYE+97YyFGMkc9jiv4PEARkwzlM0j0kSkAOEqMEr/l3cZla9uwh18jkhs3n8z2yXjCQwxRZugeg8oE4KZtjJSW2ubl5qLy+shZoowygpO7SORDijcHa2tr2PF6LyOooQ5JKl93Ozg43btzA7XYzPDx87Hfx8cKPkiSxvb2t+qgoXX7j4+OUlJRQWFh44qmSN0qK7KA9pGp5VW4CYFfORIlKFfthJSpNkLORQxjDn8QU++6uO6ThEmapBxkHYCeClUXjb1NuaUm4XrryLMeNTPbxKHvBQB4U+Q+D0qp8EjbHSkrM6XTyxBNPpBUxxQtLZntopaOCLEkSQ0NDLCwsqMZgqaBliixdcvD5fNy9exeHw8HNmzfV109LX5l0YDAY8Hg8eDwempub8fv9vPrqq0SjUQYHB4lGo+ocR0lJybHM3sQjH1JkuSaYdKIHi8WSUs5mY2ODmZkZAEqKrXTUfBGzeRmDEEQSTmGU7j1QR64lKJuZ4SM02NqOtIeTQCYqIo+ymyU8AgQDx6OAHL9eJimxVGvBQ3+HbHCYCnIoFKK7u1tVDjioP17LFFk6RLW0tERfX9++EVXyXk7ywFPel87OTgRBIBAIqDpcExMT6uyN8u9xmb1RkA8Ek+keUsnZ+HZmcEnPIMbAZFkkGC7BZp7GYJCJCafYkUxM8mt0uNpTrvkoEowewRwRmXzYTCaTpsOW8V1kmabEkhEvy5ItDjrQlQHFiooKTp06degHVMvmg4OISpZlRkdHmZ6e5ty5c3vE9GD/dueTPvSUNI3T6cTpdFJXV5dQB5ienlZnb0pKSiguLtZMyuaNniLL9nA3sEyJ6d+CUIbJchcZNxazCaMhxIa/leVwjNeWf5QLRWa8ES8ej2fP9bQSuM0W6RKMMtT6qLpZwhs8gjlKSiwVtDYKiyeGdIzBUkGrFNlBszCRSITe3l6CwSA3b97c905Ly3karZFcB1Bmb7xeL319fUiSpBadS0pKjqQynA/PPdcEkxXJSVOYg/8KWSjBKN1FFkwgVGE2jCEaLrFpdbNkfC9P1O7aEQwMDCTI2RQVFeFyuR7JCCYQCOgRzHHjOGowSgE405TYfvvTWqQS0jcGSwWt6h77RUKKr4zb7ebmzZsHen3k+nDL5PrJszc+n4/19XVWVlYYHR3FZrOphFRUVJQ7j5MM8ChHMII4gCn4SyDYMUozGASQDGcwSb2IhnMMhIrZNLyXqyVdAFRWViLLckIadHJyUtU1tNvthEKhrIeQs0EmkZSipvyo4pFIkWlJMNFolOXlZQKBAFeuXNFkiOk4ZmGSZ3AyrQtolbZLFX0sLCzQ399PU1OT6iF+Ens5aQiCoLbVKrM3ypT6+Pg4wWCQgoICNZ12kGnXGzlFdtQ5NiH2fUyhXwFkBJwYBD+S4TImqRvR0MVr/lpi5p/iUuGpxL8ThIQ0qNJVODIygt/v5/bt29jt9gT9tJO8UUi3i0ySJL0GcxLQqgajHNqKvIVWE7JaT/MvLi4yNTWVVXSlpTumso4kSYyMjDA3N8eFCxfStm7N9d2zgmxfj2SV4XjTrpmZGQRBSEinKXNJuU6R5QPBQHrSKCpiL2EK/RoQRDB0YZQHEQ3nMEl3EYU2vrvdisv+Li4VdB66lNJVqJBObW2t2g6t3CjE2xEct4VEJmZjsizrNZjjRrYHeHKXmNlsZmlpSbP9aRXBKP4309PTGRmDHeeelOgjEonQ09NDJBLh5s2bGYXt+VyDyQbJszc7Ozusr6+zsLDA8PAwDoeD4uJi1X4iV8g1wSjPPd09CNFvYgr/PgIBJOOVB3pirRilISShga9tnqHW9S7OFaTuFjtoH0ajcc+QrqLovbGxodbd4u0ItG5jz4RgAD1FdhScVIosvkvs8uXLFBcXs7i4eCJzNZlA8UgBOHPmTFbkouxJq862nZ0denp68Hg8XLp0KeN0Qq4J5kRk4uNMuxTRR+UuORAIMDY2xtramppOO0mXwlwTTCYRjCHyvzDEvo7AHJLhPCbxDpKhGoO8iiyU8SfrVzhd9P/Q5Wo5dK1k7FcHstlsVFdXU11drdbdNjY2VAsJk8mUUHeLV8w4CkRRTKuhyO/3YzKZsr5eLvFYRzD7dYkdR1daNoe5otlVU1ODKIqa5IO1qnuIosjg4CBtbW00NjYe6aDKNcHkAmazmfLycsrLywkEAmqeX5m9iT+0MtbgyhC5fu3TimBkGUP0DzHEXtsVqRSaMUpDIBRikGVkrPzh6i1ulv4U7c7GI+0jndpHfN1NsZBQ7AgUORun05kgsppp63MmbpZOpzMvOt+OikeGYEKhUNqPj0+JNTc309zcnPDhPillgMMgSRKjo6PMzMyoxmBer1ezyCOb56goBkSjUdrb22lqajryWvlCMLnagyAIWK1WqqurEzxU1tfXmZmZYWBgIGtL4oOQLxHMvnuQJYyR30EQ7yBIw8hCKUY2EQQDCGVI8ja/t/wDPFX5XpocNUfex1E62eLb2FtaWhIi02Q5m6KiIgoKCg59rTNxs3yU02PwiKTIMvGESZUSS4ZWcysKjhLBKMZg4XA4YYZEy9RWNBo90t+GQiF6enqQJAmHw5F1kTFfCCaXiP+8x3uowO48kdIs0N/fjyiKe3xvsiGIXKopw8ODPeVzkGMYw7+BQXwNZD9gRRBcCPIcsuEUkrTF7608xT+v/pfU2vYO8R5lH9kgPjKF3dS2QjiKnE28HUGq9y7dLjKdYE4ImcrrOxyOPerCyetpWXTNlLDijcGSaxpaF+czhbK3kpISTp8+zSuvvJI1OeSaYHLdxXbYc7dYLFRWVqozHD6fD6/Xy+rqKqOjo1it1gQb6UxTqLkm930jKDmMKfT/QRBfQRbKMLAMhrMY5X4kw0Vi0ja/v/oj/Ivqd1Fhza4mCccjFaPI2dTU1CR4FilzU1arNaEd2mKxZGSXfJK1uuNATgkm3YMnHXn92dlZhoeHU6bEMl0vU6RLWKmMwY5DBVlZJ5ODJf41jN+bFuRwkFRMrg+/fEN8DUCRtN/c3FQLzsrsTXxL7WEHUK5TZCkPdmkHU+jXMEivIBnO7tZdjJcxS91IhsuExCB/tP42/p+aH6fInP6A8WH7OE6pmGTPongL8HgZomAwSCAQOJRofD7fIz0DA49QBLPfHEw0GqW/v5+NjY19U2Kp1tPKwwXSI4X9jMGOspZWe1IgiiIDAwOsrq7ueQ21kpzJByJ5FKOoZBtppaVW8VABEpoFUk2o55pg9lxf3sAY/uQDcrmMQbqDZLiISbyLZLzEdkzkTzbexk/V/Cgu0/6CrpnipKVikt87JRU6NDTE7OwsU1NTCXYEySZ5j7qSMjxCBLOfvP69e/ew2+0HpsRSrQdo1rF1WESktPnabLZDNc+0qg+le6gHg0F6enoAuHXr1p4DSouBzXwhmFxBy+ee3FKr+N4sLi4yPDycYCNdVFSkOsLmTQQjLWEI/0cM4ncfRC53kAwdGKX7yMYzrEZt/NX2W/mZ2n+K1aBtZ12utciUVOjo6Cjnz5/HaDSq8zdTU1Nqbc7tdhOJRDQ1G/vCF77AZz/7WZaWljh//jyf//znuXbt2r6P39zc5Nd//df52te+htfrpaGhgWeffZa3ve1tGV33kUiRJRf5M02JJUNrgjmooK7IqjQ2NtLa2nroPk8ygvF6vfT09FBeXk5XV1fKL59W5PAoRg/5DkEQ1NmbpqYmotGomk4bGRkhHA7j8XiIRCKEQqGcEY0kSbvXlaYwhD6OQRoCoR5BGkMWajBKC2BoZiZSwff8b+Gna34Yo6B9KiufDMdMJhMOh2OPnI3X6+Xll1/ml37plygvL8fj8fBXf/VX/OAP/iCFhYVHut5XvvIVnnnmGb74xS9y/fp1nn32WZ566imGh4dT+klFIhHe+ta3Ul5ezle/+lVqamqYnp7G4/FkfO1HLoKJTzWlmxJLhvIl01LeJXmtdI3BUq113EX++FpQZ2fngZ7fWqTITtpwLB9xEge72WxWJ9TjDbu2trYYHR1lcnLyxGZv4iHLMm7HHMbgf0CQJQQsIIcBGwIxMJTSH2phJPY2fqLqyWN5rRQ9tFzL9cuynLIWlGyS99RTT/HBD36QgYEBPvzhDzM6Osr169f5+7//+4zft8997nN84AMf4P3vfz8AX/ziF/n617/Ol770JT70oQ/tefyXvvQllegUDcTGxsYjPd/c03kaUGow29vbvPzyy0SjUW7dunVkLTGtXTKTi/zBYJBXX32Vzc1Nbt26lTa5wPEX+UVRpLe3l8nJSa5cuXIguRy0TqbIB4LJ1R5ycV3FsKu2thaz2cy5c+c4ffo0VquV2dlZXnzxRV577TXGxsY0m73aD0Z6ONP4uwhCIQIryEIlsIVgKEYQjLzmP88iP8Hbyt90bESsPL9cRzDKmXPYPpQW56eeeorBwUGmpqb4tV/7tYzJJRKJcOfOHd7ylreoPzMYDLzlLW/h9u3bKf/mr//6r7l58yb/5t/8GyoqKjhz5gyf/OQnj3Re5jxFlg6UQ+6VV16hpaUl45RYKmjZqhxPCpkagx20llZ7UqDI0ZhMprRrVsfZRfZGQq6L7EajkcLCQoqKimhpaUmYvVH8U+Jnb7TS3xKi/0CJ5YPsBJrwGAeQDBcQpHsIhtMI8jJ/t/OPcTp+kjcVdGjwTPdHvhFMum3KyoBzbW1t2l5Q8VhbW0MUxT1GgBUVFQwNDaX8m4mJCf7u7/6On/7pn+Yb3/gGY2Nj/NIv/RLRaJSPf/zjGV0/71NksVhMfSEuXLiQUTRwELSOYGKxGOPj4xkbgyXDYDBoohydTDAK8VVVVdHZ2Zn2F+24usgkSWJxcRGbzab55Hqq67/RkfwaJM/e+P1+vF4va2trjI+PqzbSyuzNkWykY9/BEP4oEamJQucwkuEKBukOsuESgjzB1zZ/iArnv+DCMZML5BfBCIKQ1j5y1UUmSRLl5eX8l//yXzAajVy+fJn5+Xk++9nPPl4Es729TU9Pj+oimInh1mHQkmAUJd1wOJyxMVgytKzBKHnnyclJxsfH6erqoqYmM6mN44hgotEoPT09+P1+RFFEkiT1MNuv1fZRRq6jt8OK+4Ig4HK5cLlcqv6WMr8xOTmpzm8o75Hb7T5c0yvyVQzR/x8IxVgN0/hDTThtd5ANlzHKw/zx2lPUu36UK54urZ9uSigHe65vNjI1G8t2Dqa0tBSj0cjy8nLCz5eXl6msrEz5N1VVVZjN5oR9njp1iqWlJSKRSEZpurxMkaXqEvvOd76jyZ29Aq0IRimgAty8efNod3px0DJFJooiPT09bG1tce3atSN1oWjdpuzz+bh79y5Op5Nr164hCIJ696y02jocDpVsUnmrHxVv1E62TLvHkuc3wuEw6+vrqgusLMt7pGziLoYQ+SME8R9A3kLAgSiVYLcsIBvOY5DG+PzKD9NV+KPcKjmr9VPdF7luUVaQiV2yFlIxFouFy5cv89xzz/HOd74T2H0tnnvuOZ5++umUf/PEE0/w5S9/OeE1GxkZoaqqKuMaUN5FMPFdYvGeKPkiUKlAlmXm5uYYGhqisrKSjY2NrMlF2ZcWBKO0piozQkftGNIyglFUo+vr62ltbSUWiyFJUsL0syImuL6+zsDAQIIuV0lJSeJhlsH138jItj1ZEepUZm8UOZTl5WVGRkbiZm88lDr/GEEeRJD6EQytCNIaghAmHK3Cbljms4s/zNXid3Kr5IyGz/Bw5BPBpLsPrdwsn3nmGd773vdy5coVrl27xrPPPovf71e7yt7znvdQU1PDpz71KQB+8Rd/kd///d/nV37lV/jlX/5lRkdH+eQnP8m//bf/NuNr5xXBxKfEkovQmQhepoNsCCt+8v3SpUsIgsD6+rom+9KCYFZWVrh//z4Gg4ErV65k9cXSSvZ/c3OTubk5VTV6P9KKFxNUagPr6+uqtpNymJWUlBxJKj0XyPcUWSZIlkOJxWIPxB5XIfBRAuIMbsc8oehpHOZxZEMVkhRDFmL85uw/5m1VP84FT2ZGYVrguGVi0kW6EYzy2dfCzfJd73oXq6urfOxjH2NpaYkLFy7wrW99Sy38z8zMJJwRdXV1fPvb3+bf/bt/x7lz56ipqeFXfuVX+OAHP3jotaamphKU1/OCYNIZnMwXDxe/309PTw9Go1GdfN/a2tKsIy2b7jZZlhkfH2dycpKWlhYmJyezvmvLNkUmiiJra2uEQqGM61PxtYGGhgb1MFtfX1etBDwej5rOyVZ1+DjxKKXIMoHJZKKstIAK139CkCZA9hOOncZuHmArUI/VtM1OrIBPzF7kPU0/wTlP27Hs4zDkUwSTLtFpWeR/+umn902JPf/883t+dvPmTV555ZWMr1NXV8fi4qL63zmvweyXEkvGcRBMpustLy/T19dHTU0NHR0d6gdWq7RWNmtFo1F6e3vx+XzcuHEDgPHx8az3k02KLBgM0t3drbZJZtukEW91K8sygUCA9fV1VQjSYrGoZOPxePaoNOQ6ksgVjnWCX/YjhD8L4ndAqACDB5tpGAyXKHROMR+q5RMz7bw5fJm1e7O85tpI8L05qagin6b4T7LIf9IwGo1q80AoFMotwezs7PDaa6+lpSV2kODlUZAJwcQbg509e3ZP94WW5HcUglEK5w6Hg5s3b2KxWPD7/ZocqEdNkSkyNBUVFZjNZiKRSNZ7Sd6X0+nE6XSqnU+KL8fo6CihUEiNbo46kKsVcklsh5p9ZbX4JkLotxDE74HQCawhSBtgOI8gjTDo6+APlk7zXs9bcIpWWlpa1NmbwcFBNQJVUp5azd6kQr5EMOmm6iRJIhAIPHIEE4+f/dmfzS3BmEwmamtraWpqOvSDpXUNJt0iv2IMFolEEozBktdSWoKz/YJkSjBLS0v09fXR0NBAW1uben1lnWz3dJQIZnZ2lqGhITo6Oqivr2dsbOzYD1mj0UhpaSmlpaXAbnrB6/Wyvr7OxMQEsBvRlZeXU1xcrIkGXSbIVYrs2AhGWkYI/SqCtPCAUIYQKABDI4I0wPNrp/nLrUv81pmfZWlqAckgYTabqaiooKKiIiECVWykldkb5Z8WTTPqdvOEYDIxGwM0qcHkAp/4xCf49re/nVuCcTgcNDc3p/XY40iRHXZXrZhvFRcX7zEGi4fygdFCPDNdgpFlmdHRUaanpzl37tyeSV3lQMmWYDKpwSj6a4uLiwk6cbmY5FeMoGpraxFFke9973sYjUYmJibo7++noKBATacly6Q/TjgWghGnEUL/GkFwIhgqQepGMLQBIrI4zl8snOLVyE0+dfZ9mA3mh2KXcUgVgW5tbSV4pyT73mRDEI9akV8hmEcxgvmLv/gLfuu3fotvfvOb+VHkTwcnWeRPxxgseS1As/mVw9aJRCL09vYSDAYPjKog+/RMuimySCRCd3c3sViMmzdv4nA89PFIRTAnOfRmNBoRBIGGhgYcDocqArm+vs709LTqu66k07S8c4b8SJFpBnF4N3IRKhDkVZBnwXAOQV4lIm7xp7OnmDW8ld8+8+6Em5zDDlXlPVBuSsLhsJpO6+vrQ5KkrNrV8ymCSZdgLBaL5p/F48b9+/d5z3vewwc/+EFOnz79aBHMSdRgYrEYfX19bG1tcfXq1bQkqpUP7kkQzPb2Nt3d3bjdbm7evHloVJXtnVs60cf29jZ3797F4/Fw+fLlPXvKNy0yu91OTU0NNTU1SJLE1taWSjYDAwO43W41unG73dpocj0OKbJYN0LkWRA8II0CDjB0IUiTbEck/nS+g4j9R/n3rW9P+LOjHO5Wq5Wqqiq1pd3n8yW0q9tstgTfm8MyB/lU5E9HA1DxgnnUIuvXX3+dQCDAJz7xCT7xiU/kvossXZhMJsLhsGbXTkUw8cZgmQwnKtpCWkRYBxGM4i2TjgeOVqR3WIpMqQEdtKd8IJj9XivF5ElxGI2fWp+dnUUQBPUgKykpOTGJe62gGcHEXkSI/DHIXgR5FUGoAcEG4hDzATt/sdKC0/X/8IGmt+7501QpskwQbyOtzN4oUjbxNtJKBJrqpuBRjGDiswCPCt73vvfxvve9T/3vRyqCOc425UyNwZJxnCrIkiQxPDzM/Pw8Fy5coKys7NB14tMT2WA/cpBlmbGxMaamplLWgNJZIx8RP7UebwI1NzfH4OBggiZXunWBfEiRZUUw0W8iRP8YpEUETGBoB3kHMTbO6LaLv/G2U+v5SX6y/gf23YOWh7vJZEpo6FBSnl6vl5mZGQRBSEinWa3WvCGYdDMKSovyoxbBJCPnBJPu4XNcUjHxhelMjMGOa3/JBJPcxZbu4JVS49BCCTl5jVgsljBzc1iny37v8Ul/eTI96JNNoBSJ+/X1dfr6+pBlOSG6OSj1keuD4sjXj/4fhPAfAEEEoRSEQmRpjmDUx+Cmh29un6Kr9Md5R80T+y6RbQRzGJJTnjs7O6yvr7OwsMDQ0JCaarLb7RnNoRwHMukiy4WSstbIOcGki+OowcRiMV599VVkWd5TmM4UWhuFKX7r3d3deDyeA7vY9sNxSO0HAgHu3r2L1Wrlxo0baaeMHpUI5iAkS9zHH2TDw8M4nU6VbOItCHL53JX3P+MDXpYh+icIkf8Jgnm3qE+UWGwYb8jIuK+Yv/Wd5WrFj/FU5f7e7rtLaRvBHASDwaDaSDc3N6vadhMTE3i9Xl544QUKCwvVdNpJ1znSJTifz/dIdpAl45EhGK3nYHw+H4FAgNra2iMZgyVDyxoMoErntLa20tjYeKQvgRZKyPFrKJ4y1dXVCUoGR93HSR68Wh8i8ZpcTU1NRKNRNbrp7+9XRTpLSkpyniI7CrkIkf8EsZfBUAqYQPISFleZ89tZCLj4jv8iP1T9L3iy7Pyhyx13BHMQFG27tbU17HY75eXlajptYmICk8l0ojbS6RJMIBB4JGswycg5wZx0ikzR65qYmMBoNHLmjDaqrlo6ZAKMjo5y8eJFNc98FGgRVQmCgCiKTE1NMTo6emQztcchgjkIyUOEStfT8vIy4XCYgYEBdcjzpEU6MzncZTmGEP4dkAZBcAIBpNg8/kiM2YCbraiNb/iv8qP17+JyUXpGYScZwewH5WBXZm/q6uqQJEltFpiZmVE7COOlbLTedyZFfj2COUFoQTDK/EggEODcuXP09fVptDttIphQKERPTw8Aly9fTqtF+iBokSJTJq4nJye5cuWK2m2V6T7ygWBOag/JXU8vvvgi1dXVhMNhVaRTiW4Ue+LjQiYRjCxHIPw5EHuBKLK8Qyiyw07EynzQSQwbf7lznZ9r+Ze0u+vT3kM+FNhT7cFgMOyZvVHEVO/fv58we6P43mQbiekpsjxFtjWYra0turu7KSgo4ObNm0SjUURR1EwIMNsIRlENKCkpYXNzM61e+cOQbYosFAoxMTGBKIq86U1vOrLTZL4QTK6gdDV5PJ4Ee+LV1VV1pkMhm6KiIk2jm3Q/37Lsh9BvIkijCHKUiOhnOxDDLxewHpExGq38L+8tnml/PzWOzBphjlVsM02kQ3JWqzWhxubz+RLeJ6vVmmAjfRTVjnS7yHJll6w1ck4w6X7wjhrBxBuDtbS0qLpnChlo9eE/ajoq3qpAUQ1YWlo69pmaw7C1taUKaBqNxqxsjPOBYHJ9wCnXT7YnVmY61tfXGRkZIRKJqEVoLQQg0/l8y9IGhH4dQV5CliNs+gN4/SBaPARjYSwmG19a+8d8tOsDeCyZa2PlSwSTCXHHR6ENDQ0JYqrxszfxUjaHvs6ynFEX2UGt/48Kck4w6cJkMmUs3iiKIv39/aytre2xAlA+bFpN+B4lRRZvXBav3aVFcR6OniJTZoJaW1txOByqJXQ2+8g1weQSBz33+JkOWZYJBoN7BCAVsjnKXfNh3xdZWoHwpxDkLXwhkeWtMKGoGbPDgiwHiQqF/Onam/ntsz+PzXi0qPpRiWAOQrKYaigUUps6ZmdnARKaBVLdkCnfRb0Gk4eIJ4R0vmSpjMH2W08LvZ9MU2TBYFCttyTvT+uW53QhyzLDw8PMzc2pA51ra2uaDGtm8vPjQr6TnCAIqkhnXV0doiiq0Y1y1xzfYpvOIN5Bh7sszkDoU4Rj2yx4d9gOhTEazNhdYDEEGPFV8ELoKT597j1ZC03mQwSj5R5sNluCjbQykLu4uMjw8HBK51XlBlQnmBNEJiky2B3yO4xgFGOw2tpa2tvbU36wtJR3gcwimPX1de7du0d5eTldXV0pi48nJZypIBqNcu/ePVVAU8n/ahF9vNEjGDgamRqNRjV6ARKim6mpqYTfFxUVpbxR2o9gZHEEMfgZVrfWWNsJIiNjt4DNEcNikHnBW8MoP8JvxIlWHhW5bFNWcJxaZIIgqLM3Ssu6kk4bHh4mHA7j8XjUgeR0Xgt90PKEYTAY1JbZ/aAYg83OznLmzJk9xmDJ0FIdIJ0IJl6lubOzk7q6upSP04r40j3YFcMyp9PJjRs3Eg6q4xjWzAVybVmsBex2O7W1tdTW1ia02E5OTibI28eLdKacP4p2s7r+u6z7lkEWsRhlbFYwW2SMgoG/Xmlky/pP+PX2n9Rk3/nQpnySUZQye1NeXq6mPb1eLysrKwC8/PLLh87e+P3+R9YLJh6PDMHAwcOW4XCYnp4eotEoN27cSCu81NqJ8qC1RFHk/v37eL3eQ1WaTzKCWV1d5d69e9TV1dHe3p7St+O4Iphc39U+yohvsW1tbU2oCczMzKi/t1qtCa/zxMI38O38ITHRj8MsYTaK2CwCRjPEJCNfXW7D5Hobv9byzzXbaz5EMLnyg4lPexYUFHDv3j1Onz6tiqkODAzgcrnU99Lj8WAwGPQIRitk8sHbjxDijcFSycXvBy1TZAcZmAUCAbq7uzGZTIdaQyv7Om6CkWWZyclJxsfHOX36NNXV1fuu8bikyHK5h+M+YONrAopIpyJvHwwG+eZ3X2BTvk1jxfOYDCI2ExgNMlaLAYPZwE5U5v94T1Pm+ae8t+EpTff2Rotg9oMyA6Ood7e0tKj6dl6vl4GBAf7kT/6EyclJLBYLGxsbWTdIfOELX+Czn/0sS0tLnD9/ns9//vNcu3awtA/An/3Zn/Hud7+bd7zjHfzVX/3Vka+fe3nRDJA8C6Mckq+//jotLS2cO3cuoy4bLafv9zvM19bWuH37NkVFRVy9ejWt+RatCGa/9JYoivT29jIzM8O1a9f2JRdljceFYHKFk37uikhnS0sLpRW1fLd7nYXI39JQ/veYjSIGQUYQwGw2IhrNeEMCX9u4SGPxvzgWcsl1BKPsIV8IJh6Kvl1XVxdPPPEEv/iLv8gTTzzB0tISv/ALv0BTUxO/8Au/wD/8wz9kfL2vfOUrPPPMM3z84x/n7t27nD9/nqeeekpN1e2Hqakpfu3Xfo03velNGV8zGY8cwSgRRywWo6enh+npaa5evXqo6+Rh62WL5GhIlmUmJibo7u6ms7MzZTH/oLWOq4ssGAzy6quvqsX8wsLCA9c47hrMSU7Xv5GwsR3kD//8Nr/6n/6Wls7bdNa8jsUoIggyBgGsVjMBDKyFDHzVe45afogfcp7TVO4IHr6/uTzcleeUjwQTD0EQuHDhAh/72MfY3t7m5Zdf5g//8A9xuVyMjY1lfL3Pfe5zfOADH+D9738/XV1dfPGLX8ThcPClL33pwD3+9E//NL/5m7+Ztp39Qch5iiwTKDWYnZ0duru7sdvtGRmDJeO4ivzxrpjXrl079BBPxnGlyDY2Nuju7t63e22/NY7TU+aNgpMiuI3tAH/+7V7+7z8MIMkS7/6xPhqKxzAKMkajjAEBh8vCjhglErXwbf9Frhc8RX24WJVHibcgyGbAFnSCiUe6MjGiKBIKhSgrK+Py5cs89VTmUWUkEuHOnTt8+MMfVn9mMBh4y1vewu3bt/f9u9/6rd+ivLycn/u5n+OFF17I+LrJyDnBZFqDUXKVRzUGS15P6wjG7/fT3d2NxWI5MvkdR4psdnaWoaEhVS0g3dftcUqR5WoPJ3Hd5fUdvvq3vXzzxSEiURG328S733GXCvcUJqOExRDDaTFhtJrYjEaIyg7+r+8SP930Hk4VNKr7VEQ6l5aWGBkZweFwJMxzZHpIH9kuQEPkC8Gk22jg8/kAspqDWVtbQxTFPWoAFRUVDA0NpfybF198kf/23/6bOp+nBXJOMOlCkiT8fj/r6+tcvHgxLVfHw6B1kT8cDnP79u0D52/S3ZeWEczAwACLi4t71AzSgUIO2RQbDyIYURTVmSQdmWN8do0//9tevvf9cURp9zUuKbHw4297kTLXAiaDjEGQKXRYwSyxHQsjCgX81dZVfrnj56ixP/weJYt0KvMc6+vrDAwMEIvFVJHOkpIS7Hb7ofvLlwgmHz5jmcjEQHYEkyl2dnb4mZ/5Gf7oj/4oKwX3ZDwSBKNMvUejUWprazUhF9DWAmB5eZlAIMD58+epqqrKaj2tCEaWZRYXFzGZTEc2VIs3zdKSYCRJSqinKddS/mmNXNdgtLy+JMm82jfN1757n3vDCwm/q6218s9+8O8pca5hFMBpNWAQJGRzDH9MJEYRX928xoe7/hVFloIDr5M8z6Hc4K2srDA6OppyWn3vXvMjgsk1uUBmUv02m+1IYpoKSktLMRqNLC8vJ/x8eXk55Xzg+Pg4U1NTvP3tb1d/prx3JpOJ4eFhWlpaMt5HzgnmsA/e6uoqvb29VFZWau7PoEUXWTQapbe3l+3tbWw2W9bkAtoQzM7ODktLS5jNZm7cuHHkD6vy/mST5oknGCUaUsjFYrEgSZKqM6eQjXLHmQ93ntlCqxTZ9naQ7353kPtzK7x4f2bP71vbrPzwjb/F49jEYoJCu4lQJILdBSFJIoiHr209wUdPfwC3ObMZi3iRzoaGBmKxWMK0eiQSwePxJFgQKO+7Yt+dKzxqBOPz+bJ22rRYLFy+fJnnnnuOd77zncDu6/Dcc8/x9NNP73l8Z2fnHvuSj3zkI+zs7PB7v/d7+w6FH4acEwykvsNVjMEmJyfp6uqipqaGkZERotGoZtfNNoJRJuAdDgdnzpyhv79fk31lSzDLy8v09vZSUFCA0+nM6k5IC4JR/l4hFmUtZV/Kl04hGuVf/HujRXSTD3WgTCHLMvd65/jOdwZ49bUpGi9W0ju5vOdx586bePLcNyh0+PE4zYhilHAshsUFsiCzFi3gW/4f4DdP/zx2U3aFe9h978rKyigrK1M9g5RBz/HxcSwWCyUlJXkxLHicMjGZ7iMd3UO/36+JR9AzzzzDe9/7Xq5cucK1a9d49tln8fv9vP/97wfgPe95DzU1NXzqU5/CZrPtMV9UhsGzMWXMC4JJRrwx2I0bN1TJBKPRSCgU0uw6Bw1HHoalpSX6+vpoaGigra2N7e1tTes5R7UmUEj57Nmz7OzsZP16KQSTDeEpNxAKcSiRSTLiCUSJaBRCetSjm0zvRufmN/je8yM8//wwK6s7mMxGai9UpCSXy1dkrnV+nbKCGHaziWAkgt0CBjuYDAYGfYV8P/Zm/sOZn8VsyF7YNRmCICQ4RcZL209PTyPLMt3d3Wo6Ldu780yRqyn+ZIiimFZXniJ0me1r9K53vYvV1VU+9rGPsbS0xIULF/jWt76lFv4VxYfjRN4RTLIxWDzjZ2s6loyjHOSyLDMyMsLs7Cznzp1T3yythzYzjdSU1ujt7W2VlP1+f9Z37fE1mKNCIZi5uTnKysrSHjaFxOhGIZujRDe5HvRLBysr27z08jgvvDDK+Piq+nObw0xxRzGD06t7/ubKFZHL7X9BbUkRwbBIUJZw2QWwgcUAd7aL6fHf4NNXfh6j4WQO2Xhp+8rKSnp6eigrK2N9fZ3JyUnMZrNKNsXFxVlF2OkgX1Jk6RKdljIxTz/9dMqUGMDzzz9/4N/+8R//cdbXzxuCiTfeam1tpbGxcc+hcJAW2VGQaRdZJBLh3r17hEKhPXpnylpaeF9kmiILBALcvXsXi8XCzZs31dZoLWo52abIJEnCbDbT1NSkGr8VFBSoB5AiyngY9otu4lNqyuOUnH/yoZKPKbKZiVVevz3ByMw6t1+d3PP7giI75hon4/PePb/rvGLj2qm/psjpwBcOYDFIOB0mTDaZqCTx8nYtc9Fb/Izt0omRSzJkWcZoNKoinaIosrW1pZKNItKpdKZpceeejHwhmEy6yB4HqX7IE4KRJIm+vr6UxmDx0HJuRVkv3QN4e3ub7u5u3G43N2/e3HPXpdyZnDTBrK+v09PTQ1VVFZ2dnQkfYK2m8CHzFFlyMb+lpYXW1lbC4TBra2usra0xNTWVYLiV7t3sftGNQjzxnUv7peNOGsoegoEI97tn6Pn+FN2vTLC5GaD6dDWj43ujk9IqN6ECI3MrW4lrGQTOXa+jo/J/Yzf5kSQjJiO4XBYsNpntcJR7oVaw/wjvc59le3v7RJ5jKiTLxBiNRnWQE1CVhtfX15menlZ/r0Q3Wng15VMNJt0IRosaTD4gLwjmzp07iKKY0hgsHsdBMOmspzg8Njc309zcvG/9ALS5W0qHYGRZZmZmhpGRkX2l/7VyxjyKcVl8MT++i8hqtVJTU0NNTY0qOa94ni9ML3P/b0axmizUtdXSfLaRtstNlNcf3Jd/UHQTX8NRfqb8zUkg4AszPbrJ7NArDN1fZHxoCVHc3YOzwE5JW3lKcqluKmaVCNsb/oSflxY5KGotBOlrVBWOYjIVYjLKFBaYwCSyFY5xN9xBWeGP8RO1/5jp6em87uCy2+0Jnwclupmenk6IbtK1JT7KHk4KmRCMHsFoiK6uLmw226EfgpOuwUiSxPDwMPPz86rD40FrQfqOmwfhMIJRhidXVla4cuUKRUVFR1onXWQyiR9/kB8WPcRLzuMzcvv5XuQweL1bLE2v8Np37mI0GHEXuKjrqKGmvZra9kqazzXgLk79BUyObqLRKAMDAwiCgN1uP9ZGgXAwyuz4ChNDS0wMLDI+uICt2MVg/8KexxaXu8FjZ2Z2Y8/vGjvLmfTvEAo//KwXumw01hWzaomwtv087746gtVUiNko4yk045cjCDGBV0KnaS/+Cd5WdQPIvV1xJkrKBoNBVRqGXQsOxWBtdnYWQRASZGzSVcnQCSZ3yAuCcbvdaUUSWtdgDiKYcDjMvXv3iEQiCQ6P+0GLbisFBxFDOBymu7sbSZK4efPmgdPUx63KnIxMyCUeL3/9+9z+2uuI4RgWsxWjw0jUZCXgCxIOhlkPeVmeX+W5L79EVUs5qzNe6tqrcBW5aOiqobS2hNqOKkpriymtKcZs2f1YKzUzWZa5fv06JpMp6zZoSZTwrmyzMr/B0rSX7a0go/2LzE2ssjS3gfxgmt5iN1PTVZ2SXKoaitkGNld29vyu9XwVgyteYg+iHJfDQlNtCYvebWYNQdZWRvn5f9SHy+rCbjXgLjCyFQthM5n42+0ublS9izeXX0p4T3IdwRz1+larNcGCYGdnh/X1debm5hgcHMTtdqtkU1BQsO979ygSjJ4iywG0TpHtV+RXOtk8Hg+XLl1KKyLR0oJ5P2JQ9lVUVMSZM2cO/bBqpQGWToos/uDOhFz+5o++w/DLI8TCMaJRETEqEQ1HiYQixMISFrMVu9vOzPACRRUFbCxvgkFmemieglIXr3/7Hl232hl6bQJZkjn9ZDsLY8sUlrsprHWCbKC+qZbpf1jFXerGaDJgsVkwmQ2Y7RZikRiSLCPGRKKRGLK8m9YKByIEdkL4NoMYzUZmxldZX9pmc20HMbb73jSdqWFxcYdQILHV3eWxU1BbxPjI3pbi+o4KFraDBIN7uwQ7rtRyb3YZWQa300pTTTGj02ssbu4QLDGw7J3i/ddeocRtwuO0YnTIbMUCOC02/tJ7lrc3vJtrxV0Ja+a6sUErLxiDwaDaEjc3N6s+Kuvr6/T19SHLcoKMTXynYr60KWfSRVZTU3MCOzp+PJIEo9VdWaoiv3J3tF8nW6brHQWpCEapA7W0tNDU1JR259Vxp8iSi/npkosoinz5M/+HheFFpKhILCISi8aIhmLEIjFC/jDRcIyiyiLW5jewWi3YHDa2Vn0UlLgIByIsT65Tc6aSgdtjmCxG6s7W0P/SCACOUhu9fzuGLMncN4/RdrmJoe9PqNfvutXO8N1ppAeRgmA00Hm9haHXpxL22XS2loXZTcJJJNJxtYmxoWWVbBSUVBUiOyzMTe/t+mq9UMvo3AaxpL9BgI5rdfRMLlFUYKO2wsPo9Bq9I4tU1xayYo2yuR3in3XdpaVMosBpx2CX8MVC2M1O/tx7jp9q/pecLdwr5ZFrs6/j8oJRfFQqKyuRZVmNbhYWFhgeHsbhcKhkk09Ffr2LLAdI9wOoZZ1DWU85GCVJYnBwkKWlJS5evHgkwbfjiGDi527Onz9PeXn5kdbJBgfJ7SspMaWQn8576fV6+fNP/w1bi9uIUQkxFiMaFhFjuzcP0UiMoC9EQUkBQX+InQ0fDV21jL4+QcPpWhYnV5BEibbLTYzemcJV7MRV7GCqdw4EgarOUuYHdgvnzkI7ZQ1lKrkIgkDXEx0MvDqu7sdiM9Nwpm4PuXRca2bs/sIeEjl1o5XBe3N7nldNaxlboRjbKVJf7VfqGRhfJfllNBgNNF2uZnknwOnWSoanVugbXQKgsbmECclHIBDlRv1r/KP2VWTBhWQTicoxDEY7X928zL9qex9NrtSmcblOkZ0EwQmCQEFBAQUFBTQ1NRGNRtXopr+/n2g0is1mY35+nuLi4rREOrWG0nySTgQTCATU4fJHHXlBMOlCIRUtCUaSJEKhED09PWpd46j5T60jGEXnzO/375m7SQdapchS1WCSySXdQ2R6ZJq//vx3CWwEiIkisUgMMabMsey2GPu3ApjMZix2CyOvT3Dqeiv3Xxym61Y7I69P4PI4cJcUMHpnioqmMqSYyPLEGmabieI6D4tD6wC4SuwYrSam+nfJwGQx0XKxKYFc3MUuPJUexnpnE/Z5+lY7/UmEIxgNtF9pSkkuTWdrmFvaJhxKSn0JUNHuoX9sb6eY2Wqi80YdG4EQS+vbLK49bCduPVVO/84GUVHkeu0U/+zcFDHZga1AwGCQ2YzYeD58g2c6f5ZyW/G+r3euU2S5qH+YzWYqKiqoqKhAlmXu379PNBpleXmZkZER7Ha72pm2n0in1lBuPNPVItNrMDmAcoespSQLwMsvv0xpaSmnT5/O6sOmpcx+LBbjlVdewW6371E0yMV+4g+qoxbzu//hPs//6QtEAjE1JSZGJaKxGLIoYTAaCOyECPoj1HdWM/z9cVovNtL/0ghn3tRJ/0vDVLVUEA5EmRtZpOlsPSuzawS2gzg8dmxuK8tju6mp2o4qtr1+tpZ3vTWsLgvuMjcjd6fU/VTUlyAKBubGHtZKjGYjbZeb9pCLxW6m/nQdw73ze55X2+V6xsbX1PZj9XUzGmi+XM9gUi3G6bTQ2FLGuiHKa4N7yarjXCXda6uYjAL/qDHK5cbXMJrsmN1gMcJCyMIr0TfxoVM/R4H54IMoHyKYXF5fEASMRiMul4umpiZVpHN9fZ2hoSGi0ShFRUVqs8BxHeyZEIwewWiMTMyvtJTYX1jY7fBpaGjYd74lE2iVItva2iISiVBTU0N7e3tWMvlap8iOSi7f/fKLdH+3l1g4upsWi4oPivkxYpEoFpuFWERkfWGTmrYqpgfmqGquYKJ3hvarzfS/NEzLhUYWxlcI+kKcutHKaPcUYlSkqKqAaFTCO7cbAbRdbmJmeJFwcLduUlpbjNFiZmX2YV2kpL6ArZ0gYf/DiMPmtFDdVs3Q3USlYleRg6LqYsZSdISdutlM//29P7fYzFSeqkwgl4aGEhx2M2tbAeZjQRbX9qbSOi9V8/riEl01ZaztzNFe+U3KPA4sBQICEvNhO6+K/4iPdL0fm/FwyZ18qMHkuv4RX1xPJdK5vr7O2toaY2Nj2Gw2lWyKioo0i24y+b4oasqPA/KCYDKBFrMwoigyMDDA6upu2qKqqurYmgYygSzLTE1NMTo6isFgoKOjI6v9aDVoqRBVfGtvJuTy1c/9DeM9uwX1aERCjMaIhqJEQlHCgTAGowGD2cTC8ALFlR62Vrex2q1srm5R1VzOyOsTu0X5708gxiTOPNlB/8u7xfzy5mK2VgKE/btkcvqJdgZfG0d60C5c11nNljfAzvJDcum40sTk8BLRuDkTe4EFR5GbiYFEsiit8SBYrcxOrCW+JkaBjmvN9N/fG9G4PHacNR7GJ9coLnZgt8tEwkamp9eprC3E5wRvMrkI0H65mgW/j47qUgbmJ3n3xRdpqTLgLDDhjwXZopju2A/yG6feg8mQ3lc31xHEcRX5M91DKpKLF+msr68nFouxubnJ+vo6IyMjhMNh1YJAiW6O+lzSbVFWSE8v8ucI2UYwinkZwK1bt3jhhRc0Fak86t5EUaS/v5/19XXOnTvHvXv3NNmPVhGMKIrqc0u3mB8MBvnKZ/6a5fEVxJhELCoiRmJEo7spskgwQiQYpaiqCO/CBkaTEZPVxPriBoUlbhBgdniR00920v/SCCaLkVM3WlVyqT5VxvLYBmJMQjAInLrZRv8rY+r12640MzO8RCSuLnL6iXYGX59KIN7KxlJCEYn1xcRDv7yhmEBYxreYKNVitpmoP1vLQApyKaksxFHlxuGwYjAKzMw8HKSsbSlmIRbCv51YpzGZDLRcrEQ0wvK2n4VNH29tucv55gBWp42tSJANoZIh6Qf5WNdPZ3TI5Zpgch1BQfrdW/GyRbIsEwwG1UHPiYkJzGazSjZFRUUZ1YEz6WTTUuwy18gLgsnkC5DNsOX6+jr37t2jvLycrq4uDAaDprM1R41gQqEQ3d3dANy8eVNt/c32cNAiRabsYXl5GYvFgsfjSWtPKwsrfO13v8nmyjZiREQUJWLhGJIoggSxaIzAdhBXkYtIKMLm6jaNZ+oYfX2C+tO1LE+tEo3EaL/STP9LI7gKHZTWlzD02jgIAjVdZcwP7EYVFruZhtN1DMSRy+knOxj6/qQayQgGga6bbQx8P1FQsvF0DStLOwR2Em0N6rsqWZzfIRpJ/Gw4C2wUNZYyOrSU8PPSCjdVjaVshGNMTa+r11XQcqaSkY1NItHE9axWA+2XauhbWsUX2o3CLlfd5YcuzCNaXYTFKHNSLWvmt/KRjh8/9HVPRq4JJp8jmIMgCAIOhwOHw6FaEGxubuL1ehkfHycYDFJYWKim0w4T6Uw3goFdgtFrMDnCUSX2p6enGR0d3aPbpSXBHCVi2NzcpLu7W20yMBgMhMNhdd/ZfDmzTZEp9Zbm5mYWFhbo6enBYDCod3klJSUp7+KG+4f51he+R2g7vBu1RGPEouLuzIkEkizj3/RjMBqxu+0MvzZG5402+uM6xZweOyU1JYy8PrlbQzEZmO6fw2g2UNZSopJLQamLwrJCRrungAdEcqudgVcfzrxYbGYaztbtIZeOq02MDywRSzr0T11vYbh/AUlMfO2cxTYEp4WZyTVMJgO1DSU4XTbWVrexexwMTK8TSu4iAzou19I7v5JAOgLQ1VrBii3C7amHkVBb8Tg//uQUAcFBgdHAvVAVJufb+dWmf5rem5YCb/QIRos6kNFoVKOXtra2jEU60yWYaDRKJBLRCSZXyLQGI4oi9+/fx+v1cvXqVdWlLX69XBmFKUOdbW1tNDQ0qAeBVsKZCsEchajii/klJSWUlpaqYoSrq6uMj4/T19dHUVERZWVllJaW4nA4ePn/vsbtv7xDNBx70PcvIEkCon93Ut5sNRHcDuLfCVN/qoaR1ydovdjEwEsjam2lsqmcaERkdmiBhq4aNpa32NnwY3NbcZU4WRrebUOuaCwjJorMjiwCYLaZaTpbn0Au7mIXRVUexu4ltiF33Wpj8M7MHgI+fauN/u7ExwJUNZfirixEEKDAE2BpboupByKVtZ2lTCxv7+kiA+i8Xkf35MNoRwA6m8oRBZg1h1hYf5iWu1jv40ev9hPAQLHNym1fJZWeH+Mn634go/cuHvkQweR6iv449pAs0qlEN1NTUwwMDFBQUKASjiKFlW6LMqCnyHKFTA7xQCBAd3c3JpOJW7dupTS6ykUEo4hoLiwspBzqVEgl23mfeLOwdA8ZhZBSyb7EixG2t7cTCARYW1tjdXWVkZER+v92hLWJLcyCBQEBMSbtFvODEURJwlngwGIzI0tQWF7IZO8MVU3lTPTN0H6tmf6XR2g6V8/y9DqB7SAd11qY7JshGo5RUO5CMBhYm94EoPFsHStzXgLbQUAhkiJGex52gJXXlSAbDcyOPuzkEowGOq+1MPD6dOJrZTLspuO6ZzEYBcqrPXhKXBiMAma7laHRZeb79tZcWi7VMjy5vmeAEkGg/Xot3RMPyEWGzqYyAoEom6EQXpfI+ubu3p1WMx2VUX7k4otIRgmPzcL3fE10lLyDHym/pt5QHeR1c9D7qUcwx9vJFi/a2traSigUUqMbRaTTZrMhyzKRSORAkU6/f1c9WycYDXEcNZi1tTXu3buX0iclHlq1Fqe7liLAGA6H9x3qjI9gskG8AGc6X7B4YlH2cdB743A4qK+vp6amhv/16b9kc9JPNBRj27eDJMoYZANIAmJERDAISA6RsbtzuIrcxKIxnIUObC4rXbfaEKMSF998muUZL7FojNNPdjB4exRZlilt8ODfjBDcDgDQeb2Vsd4ZYpHdg7eioRQJA7MjDyOF+q4aNlZ9+LYC6s8sNgv1p2sZvLNLLg63FU9pAS6PHUexi6A/Sk1jCSuLWyzNbrA0u0HHlQYG++f3TPMjQOf1ZvqGFve8LgajgKfVyb2JJQSgrb6EYDDG8MQq9U0lLBlC7PgedL3VlrOyPceT517AYY+CYOF7gVPcrP0J3lx+OS2vm4Pe21wTzKNag8kGNpstQaRze3ubyclJdnZ2ePHFF3G73Wq6ze12J+xNEbrMddSnFfKCYCD9qfPDIg5ZlpmcnGR8fJyurq5DReO0TpEdZHW8s7NDd3c3LpeLGzdu7BudKHep2RJMJkSV7OGS7hdyY2OD//3pr7O5uInZaEGwGrAIVsKhMEF/iMCOn5A/TGGZm5W5NTAYsNjNeJc2KShx4V3cZP3OBqef6KD77/oxGI103mhldnCeiqZSbEVmiBmoqN29o7O6bPi3gzSdqUWKSVgcFkQJIqEoLs9uG6mr2Ek0JuEsdCAYBAyCgMFsRBIMBHxhKutL2Fr3EdgJY7IEEWwWRl6e2PPcTt1sYaB/fk90YjQbabpYl5JcbA4zJZ0lTC9t0FFfwprXx+jUbkqvqsHFmLRNOCJS6nZQUehibGmNn3rzHeoKAvgkA9+PnOZH6v+lKlp5mNeNgvjHxb93uSaYfIhgcqlFZjAY8Hg8eDwerFYrra2tamea0imqRD8Wi0UlGC3esy984Qt89rOfZWlpifPnz/P5z3+ea9eupXzsH/3RH/E//sf/4P79+wBcvnyZT37yk/s+Pl3kDcGki4MIQfGl39ra4tq1axQWFqa13km0KS8vL9Pb20tjYyOtra2HfoC0tDtOx7zsKMOTs6Oz/OXvfZvgVgAxIhGNicTCIrIkYjSaMJvNyFEoKvWAATZXNiht9DB1f5bKtnI2F7aJRWN0XG2h/6Vh7AU2qluqGHzQDeaucjLdvYQsg9FkoP1qC3ef61ev33GthfG+eaKRhzW507fa6H9lIuFmpbS2GCxm1hY2E/ZfXleCaDQy/4AA1NfNINB5o5n+FCkxm8NCeWfFnul8ALfHTlFrETabGafFwkjcui2ny+jb8iLLMk2FNhZ8IXzhMG970z3aSqcISB5eil7gXza9l1OFTSlf78OcPFN53eSaYPJl0DLXe1BqMBaLhaqqKqqqqpBlme3tbdWC4Md+7MdUrbTvfe97PPHEE0d29PzKV77CM888wxe/+EWuX7/Os88+y1NPPcXw8HBKPcPnn3+ed7/73arp42c+8xl++Id/mP7+/qyUnXMvMZoh9ivy+/1+XnnlFaLRKLdu3UqLXJT1jrNNWZZlxsfH6e3t5ezZs7S1tZ2YErJynYMiw/jIJRNyuffCff78d75BYDNANCISjcYQIzHEWAxJkpElCf/mbj7ZVexmfmiZ9svNLAyu0nKpgeXxNWQkCipdDH9/nMJyN55yDxO9MwhGA9WnylkYXEOWwe620XCmjsHXHmqInX6yg9GeWZVcBIPA6VttDLw2mfB86zqrCMXkPeRS31mFPyKyvpxoJ2yyGGm90piSXAqKHBQ2lTCeNHQJUNdYTGVXGZNzXnqHFtjyPWx77rxczb2tdVori6kuLmBqK4TFZuQfXxygo+gu4VgR39o+w78o+gnanHudSfeDwWDAbDZjtVqxWCyYzeYEQdhYLKYSkFY3UZki1ykyJa2YLwQTD0EQVPuBa9eucefOHd7xjncQDAZ517veRUlJCT/2Yz/G9PT0Pqvuj8997nN84AMf4P3vfz9dXV188YtfxOFw8KUvfSnl4//n//yf/NIv/RIXLlygs7OT//pf/yuSJPHcc88d6fkqyJsIJt0UWaoazMrKCr29vdTW1tLe3p7Rh+k4i/yxWIz79++zubnJ9evXKSgoOPJaR4FyJ5tqnYOK+Yfhuf/1Ane/04sUkYhFRMSwSDQWIxqOIUZjWOwWQr4wO14/DafrGLkzScvFpl1NsQedYuX1JciyzOq0l4rmUnxbQZYmVzHbjBTVeFgY2j3Ei6s8WJxWJvp2O7sEo4FTN9oSOsXMVtNu99hriW3IbZcbmR5fSxi0BGi71MDUxHrCJD+Aw22jtKWM4cG9qa/S6kKiDgtz85vqzywWA22tFQgWI6ObW2yNJP2dAB1Xq5ne2eF0TTn9cyu7axU7ONc1yemS17Bb6/hu6ALvKngLofktnh99nqKiIrUVPN10SXIqDWB6epqdnR1VgwuOx8nzIOQ6RaacKbmuaYiieKgDZ0VFBefOneP27dvcvn2be/fu8c1vfnPX8TUDRCIR7ty5w4c//GH1ZwaDgbe85S3cvn07rTUCgQDRaDTjaycjbwgmXcQTghIdTE5OcubMGaqqqjJeT8sif/zegsEgd+/eVTvY0rV3jd/XcblRZlrMj8fXfu9vGL87RSwqPRCsFImGY0SCYSLhKBarBUmUWZ5eobq1itnhBSoby5jsnabjWgv9L4/QeLaOtdl1fJsB2q40Mzu0qxvmLHJgcZhZmdidfi+qKSASjuKd3I0yrA4LdV21DL72kFxcRU6Kk7rH4EEbcs+s6jCp/vxmK4O983t+7ilzYylxqq3H8ahpKcMbibG95kNAprTMhstpZ3Fhh4AYY2p1i2A4kcQMRgNtV6uQjLDlD7G6vRvNNdUVIXmGuVj6d1jsNfx99CYfOftzeCy7cw9KZ56ijWW1WiktLaWsrCxt5V9BEBgfH2dubo7Lly/jdruzdvI8KnIdwcR/xnOJTMzGnE4nBoOBixcvcvHixYyvtba2hiiKVFRUJPy8oqKCoaGhtNb44Ac/SHV1NW95y1syvn48HlmCUaTsfT4fN27cOPJgktFoJBKJHP7ANKCQgtfrpbu7+9AOtnTW0mJPqZSQMy3mB4NBvvKp/8PK9BqxqLQbtUR3o5ZoOEooEEYSJVweF/NjS3gqitjZDGC2mNlc3aa6tZLh74/vdoB1TxGLxDj9ZCdDr44hSTJFNYVEQjE2FnbnQlovNjAzukTkgfOjvcCKs8TNeNw8S1ltMRgTu8cQBLputTFwJymtIAh03WxlIMWMS2VjCUFgKS46UdB8pprp1R1KSlxUVbmZn/eythJijRBt56sZWPHuMRAzWwx0Xa/j/tIqW4HdVJnFZOT8qSr6I938ctsrxIyVvB77AX7jzPuwm2zq3yqdefX19YiiiNfrZXV1VfU1UWaSSktLsdlsJEOWZQYHB1lbW+PKlSuqplWqRgElij3O6CbXEUy+EEwmZmO5blH+9Kc/zZ/92Z/x/PPPp/yMZYJHkmDC4TC3b9/G4XBw69atIxfClPW0LPIHg0Hu3LmzRzEgV/tKNi87SkpsaXaJr/3uN9nx+h+oIO+mwsSYhCzuzrqEfOFdocqVLZDB6rSyOL5MQakbg9HAzOA8p5/soP/FYQSjgdNPdjJwexSA8pYSNpd8RAK7ZNJ1q42h1ydVt8nKpjIiUYn1uYeaYKUNHvzbYYI7YfVnJquJlouNe8jFaDbSeqkpJbk0nKpieSNAwB/e87tTVxqQbWY8kRgL8xsJv+u4Use9meWEDjMB6GqvxOuM8dLEw2u1VhTjKrPxSriHD59+jaCxkGHDW/h413swGff/ChqNxgTlX5/Px9raGouLiwwNDeF0OtXoprCwEFmW6evrw+/3c+3atZSHw36NAspnQ+voJtcRTLx+Xi6R7qClFm6WpaWlGI1GlpcTG1GWl5eprKw88G9/53d+h09/+tN897vf5dy5c1ntA/KIYNL9AGxvbxMIBGhubk67YH4QtKrBSJLE3Nwc4XCYq1evZp271Fpq/6idYqN9o3z9888RCkQQow8EK6MiovhgzZiIb9OP3W1DkmVW5zdpuVDP8GsTNJyuYXl6jXAwQuf1VvpfHMbqsFDfVaeSS/WpcpZG13dlWQSB00+00X/7oaZY09k6VuY3E7TC2i41Mj20mKATZnVYKG0qY7gnkUTsLhsVrRUM9+71XWm7VM/E1DrRB1IxVpuJ+sYyzBYTRreV7vt7/wag43o9PROJWmSdzeWIgsyCNcTsym5Kz2Ex01ZVQsQpMy5M8Ittr+IzFTFreiu/kaFopSAIuN1u3G636tqopNJ6enrUSMFoNHLx4sW07zwPaoNOlUbNNLrJhwgmk8/7cSFdgtFCSdlisXD58mWee+453vnOdwKoBfunn35637/7j//xP/Lbv/3bfPvb3+bKlStZ7UFB3hDMYVCsg2dmZjCZTLS3t2uyrhYEEw6H6enpIRwOY7FYsiYX0N687Cjkcvvrd3j5a68Si0oICAhGASm060IpyzKCQcC36UeWZApKChh8ZWxX0fjFYbputTF6ZxKbw0pdRxVDr45RWObGXVLAWPcUCFDTVaFaG5utJprONySQy6nrrYz2zSHGaYWdvtnG4J2phBpKYUUBRruF+fHE7q7CUheOkgKmhve2FJ+60czIyDJVtUW4C+wEfGFmp9YZG1qi43ojvSnIxWA00HylJoFcOpvL8QcieANBfB5Y8e7WWk7XlrO06SPmgYh9k7cVvIRoLWbT+jZ+vfMn03r9D4LZbFbbXcPhMK+//rpqyXv79m0KCgpUCZ/DhBjV53dIG3SmQ57KGrmuweS6wA/pE4xWXjDPPPMM733ve7ly5QrXrl3j2Wefxe/38/73vx+A97znPdTU1PCpT30KgM985jN87GMf48tf/jKNjY0sLe1+xl0uV1aE90gQjDL9HgqFOHfuHH19fZqtnS3BbG9vc/fuXTweD62trfT29mqyLy0IRpmB8Pv9FBYWZkQuX//Sd+h/YQQxEiOmeLiERaLhCCaLCbvLjhiLYTSbqGwuYvTuJK0XG3c1xd7UQf9LI5TVFSMIAlP356hurSAUjDI/uoTRbKSsuVglF1exi+IqDyN3HnaBnXmyg/64TjHBINB1o42BuAI/QHVrBf5gDO9SotR+UZWLsCiwOPPQB6aw2ElpVSHuMjfejSACAjNxpGSyGGm5Ws/9FAOUZquJ6rMV9E0sIwhwqrmCLV+IoYkVauqLWLeKbO2EKHM7KS1w0D+3wtlLVQQtIuecz2F22REKf5xnWv5ZWq9/ulCaSdxuN2fOnMFgMBAKhdToZnJyMkGGvri4OG35oWyHPJW/y4cIJtfIJEWWLB11FLzrXe9idXWVj33sYywtLXHhwgW+9a1vqYX/mZmZhNflP//n/0wkEuHHfzxRsfvjH/84v/Ebv3HkfeQNwex38G1vb9Pd3Y3b7ebmzZtEIhFisZhmA2TZdJEtLi5y//59mpubaW5uZmdnR7N6jhbmZaIoUlZWxujoKJOTk2q+vri4eN8PuyzLfPkzf8nC0CKx8G46LBYViYaihIMRoqEoJquILMlM989R1VLJyuw6LecbMVqMXP4n51if99JyoYHl6TV8G35aLzayMLFC0BfGXmDDWexgaWR3CLG8vgRZEJgZ2jX6MpqNtF9tSSCXh23IieTSfL6ehdkNQoHEJo3G09UYrBbMFhOSLOLfCbGx6mN7K4CrysXo9/fOFdhdVsrayhhKMUDpdFtxNxcxueDlbHsVy2s7DIzvPq6pvYyx6BaxsMTFxioG51dZ8/m5fL2WIb+Xf1L0KrgjlJT8K97d8OYM3sHD4fP5uHv3LuXl5XR0dKjfB5vNRm1tLbW1tUiSxMbGBmtra4yMjBAKhSguLk5og04HRxnyVBpMcnnA53KKPx6ZdJFpZTb29NNP75sSe/755xP+e2pqSpNrJiNvCCYVFhYW6O/vVw/w+LqEVqHvUQ5yWZYZHR1lZmaG8+fPq5OxWuuaHZVg4ov5ra2ttLS0sLGxwerqKsPDw4TDYbUbqaysTBUB3djY4H//x6+zOb8RZ2scIxbZ7RQLByPEIjFcxU4Wx1coKCvAvxVAjIrMjSxSXO2h/8VhOq61MHFvGpPFxKW3nsW3GaDhdC2BcACj0QQxA7Y2G1a3jR2vn43lTWD3kK9uq2YoTlbf5XFQXF3MeN8sDrcNm9OCzWGlpLaYaEyi+XQ1ALGISNAfprC8kKH7C3sk+O1OK7Vd1UxPJE7tAxQUO7BVFTAxtXeA0lPmxN3gwWYzYd400Rc369J2poL72+u0VZbg9QXpnlrEbDJw5lo1r6zP8466YcLOHc6V/zL/vP7JI72X+0Gxeaivrz/Q7ttgMKi6Vx0dHfj9/gSBUrvdrpJNUVFR2odxcnSzXxu0QkC5Qj5FMOns43Fys4Q8JRhFbXh+fp4LFy5QVlam/i5+UlkrgsmEFGKxGPfu3cPv93Pjxo2ED4NCVlpEV0clmFTFfEEQ1ENGlmX8fj+rq6ssLCwwNDSE2+0msi3x/b/oJuyLIT9wn4yFY4iihCTJRMJRgjshiioL2V7bQRQl7G77bqdYiRuj2cjMwG6n2MBLI8hA+9VW7n53V9qlrKkY30aI0IOur45rLUzcn1WHHUuqi3AUOlmaXsNZaAegqKKQWExmYWIFSZQJ7IQI7IQ4/UQ7fa9O7nnuXbfa6O+Z3aMdVljqxF5akJJcPOUOQlYD3qRJf9iNTii2MTy1ukeKv/NiFctikOaKYgbmd1N9DeUe5FoDU/5NfqhigkhBkDfX/QpvrryU0Xt4GNbW1ujt7aWtrS3jTkXFIrihoYFYLIbX62VtbY3+/n5isVhCG3Qq9fFUSJVKkySJxcVF9XsaiUROfMhT2U+uCUZ5PTKZg3lckDcEoxzI4XCYe/fuEYlEuHnz5p4XO17KXgtkQjB+v5/u7m6sVis3b97c0x4d/yXLlvyOQjDxd5H71VsEQVALd01NTUQiEV78xm3+7k9fJrDtBwmMBjNGgxGj8CAdIkr4Nv3YHFYEQWBlxkvLxUaGXxunvquGlZk1woEwpx6YhpmtJprPNdL/0q61cVVnGatTW2qx/vSTHQy8Mqbe2Va3VhAMRJkbe5iequ+sYmPNj2/zoRqy4YEeWf/rU4nP6YFdcqo25PK6YqJGAwuzG3t+V9tWzloojM/3sEXZbjdSXGzF6LAwFfQTGN/a83dnr9cRMIoszO4gSjJmo4HzzZXMOv24zSbaLAMYyny8s/HfcqW0c9/36yhYXFxkYGCA06dPH9pyehhMJhPl5eWUl5cjyzI7Ozusra0xPz/P4OAgLpdLjXILCgoyahRYXFxkbGyMc+fO4XK5cjLkCflBMMpzTrfI/7iYjUEeEQzA1tYW3d3deDweLl26lLIYKQhCxqZjByFdglHk/2tqavaVo4nPT58kwcS3ISt/m24E9Q9fe42eb9/HZXViLbQR9AcJBYP4/T4kScZqtxLcChOLiBQ2ehi8Pap2ip262cbY3Uksdgv1XbUMvjKKu8RFUYWH4dd36yU1ZypYGNxNPRmMBjqut9D/oEUZoOVCAwtT64Ti5lDaLzcxPbyUIPFidVio6axm6G5i/cRsNdF4viEludS1V7K2HcLvDez5Xcu5aqZWtgmHYxiNAmVlNux2B7OzG5gLXIxtbhNNGqC0WY20XKqme3GZUHT389dSUYzJatglF6sVozCKtTzKe1t/lVOe1KKVR8XMzAxjY2NcuHCBkpISTdcWBIGCggIKCgpobm4mEomwvr7O6uoqd+/eRRCEBCfTg2bPpqammJyc5NKlSwkGfyc95KlcM9ddZOkSjCzLBAIBPYI5Dni9Xl577TVaW1tpbGw88IBM1xMmHSgEs19aK95u+TD5//joKpvhT2Wto8jsKymxdPDnv/d/meye3u0Ue1BzESQBg2zCYrbtuk/6g2wsbVLSWMzw98eo7apK6BQrrS7CaDEx2TtDRWMZYkxiZnABg9FAZWeZSi42p5WajioGX30oWNl1q52R7ukEr5Wum20MJbchl7pxlRYw0b+QsH9noZ3SxnJG7yf+HKD1Qh2T096EWRkFnVcbGJpapb6hBFmOsTC/xdJSEAjScamW3vlV4u2NzSYDdZVOtpwir87simA6zCbaqkpYDgVwVdqJhESixjk8pT5++dQzNLur03oP0oEiiaRIv6Qr5JoN4lV/FSdTpSvt/v37FBYWqm3QTqdTnbeK32ey9t5JD3kq18iHCCZd4tSyyJ8PyBuC8Xg8XLlyhaKiokMfq7UCMqT2zRBFkf7+ftbX11PaLSdDuQPTan7lIG8ZOLrMfiAQ4H9/5q9ZmVpDjIhEo+Lu/z4o5IcDYUyWXbn9+fklattrWZ/3UlpbzPzQIpWdpfS/NEJFcyn+jQC+hQ2aztazPLPrRGl1mvFUF7L4QLDSU16Aw+Ng/N6uXpggCLtT/a88JJv92pArGkqJyjA/magRVlJViMnlYHp0Zc/z67zWxNDgUgJJAJjNRk7daCQkydjtFiaSdMc6r9c/dKBkl1hONVewvOEjXG5jenm3htNa6mZpO8Ds9jquWhuLO1u4CnYQSif52Jlfp9yhXXShSL8on8Fc3N0aDA+dTBU/eqUNenx8HIvFQmlpKeFwmK2trQSJmsPWPc4hT2XdXBNMJlGUXoM5JhiNxrTIRXmslt1asLfLIxQK0d3dDcDNmzczmozWWuIlFY5KLsvzy/zlf/o6W6u7Hi5iRCQaiT5oR44RDoSJhKK4S9wsTa3gLi4g6AthNBvxbwSp66hmemCe5gv1zAzME4uIVJ0qZ3poASkm4S51YLZbVcHK6tYKgv4wC+O7RGA0m2i73JRALmariaZze9uQG07XsLayg387lPDz6pZy/GGJ9RTaYaduttDf+1Bqv7TcTXlVIdGoiMlj5+79vTL8gkGg7WqdSi42q4n2xjJmFjeZXN7A2uBgfHmdao8bt93K8OIaLU0lbBVEEAUBF0tg3eFnY/+UrXkv5lKDOneUDSRJUqVfrl69mrUulFaw2+3U1dVRV1en6qWNjo4SCAQQBIGRkRE1urHb7WmteRxDnpAfbcqZ7MHv9+s1mFxD6xoMJKa1lBbQkpISTp8+nVEOV6tW5YMIJp1ifioMvT7Ct//b8wR3wsSiEsgSBrMBm9FCiAjBB11aRRUF+DZ8RMMingoHC6NLFJTudopND8xz+ol2Bl7etTM+/WSnWswvqi0g7I/hW9iVSmk8V8vy9DrBB0V0Z6Gd8sZyhu9MqXtyeRyU1BQz2p2ohtx+tYnJoeUEMzGA5nN1LCxs75l9MZgMtF1uZHFhi46zNRgEgdWlLdZWdtjaDFJ/oYa+wb2pNKPJSMOlKu5NLOF2WmmuLWF8dp3e4UVKypzEyows7/i42FhF3+wyi5s7XDhbTU94hQ57CZvmReqKHPz26acJ+Xa78xSnwpKSEsrKyg6tWaSC0q0oiiJXrlzJWI37pCAIAouLu63bTzzxBKIosra2xvLyMsPDwzgcDrV24/F4jtwGfdiQ537r5ksNJp09KDN+eorsGJBJW6+WEYxyJ6SsNz8/z8DAAG1tbTQ0NGTcbnwcIpUKsinmv/KNu7z4tVeJhmLIUfGBOZhILCaCLCDGdluSK5vKsDltbHt91HcVMvDyCA2na1mdXSfoC9F1s43+l0Ywmo20XW5WyaWivZSN+R2ioV1CqDtbyczQ4q7GGFBUWYjVZWdq4OEhX1pbhMFkYmY4Uder61Ybg3dn9sxPdF5rZnRwWW0ZtjstVNQU4Sy0IVsszM962Vj3413zqX/jcNsoaSlheHTvAKXNbqH01O4My9n2KkYmV7g3vLu/qtpC1h0x6gpdhGIxuqcWqSpyUdNcxHh0k7qCAjYtizR4Inzu0r/BZDJR6HRRUVGBLMt7ahYej0ftyDrM4yUSiXD37l0sFsu+zS75AFEU6e3tJRwOJ5Cgy+WisbGRaDSqtkH39fUhSVJCG3S6pHnUIU/lsfkQwaTbQQboBJNraFnkh4cR0dDQEHNzc1y8ePHIcg3HFcEke7hkUsz/+peeY+CFoYeT+ZFdt8NoKIokSVisZvybAXa8PlxFLu5+9z5dN9u5/8IQ59/cRWA7SMuFBiw2M6tzXsrqi3EWuhh6ULCvOV3O4oh3tzAvCJx+oj2hU6ysoRj/TpjN6YdzKHUdlWx5g+wsPWwfFgwCp2607lFDLihx0nS2jkhEovN8HeFwlI3VHdaXt1lf2SZiMjI3sJdAPGUuzGVOJqf3zr+4PDYazlez5Q+ztLbD4upDqZmGlhKEUjOuUJh7M0sYBIGLjVXEimDbHMElmIm5VrlW4OLDV/7VngNMEATVh721tXVPzcJqtaoppOThxlTSL/mI+Ajr8uXLKSM0s9lMRUWFSrrb29usra0xOzvLwMAAbrdbJZt026Ah/SFP5buYbcNNtsiEYARBSFtd4VFAXhFMuq6WWkYwsPtBHBgYIBaLpZy9yQTHJbOv3LFlUugURZE/++xfMT+4QDQi70rsP5B9iYQiRMIxbE4rkVCMtXkvjWcbGFM0xV4e4cybOul9fpDiag/+zQCLEyuU1pZgMJqYH1umuKYQd4UTs9FK141iBIOAzWVleyNAy/l6YlERV6ET304Qk8WCq1BCkkRshVaCoTD2AhOF5eXY7FYsVhO2AiehYJSOSw1EQlH820G21v3UdFRz79WpPc+vrMaDZLMyN+3d87uK+iICJoG1xcQ5lsJCO42tZcwEA7w+kEJh+VQFITcMPnCfbCzzICMjlgrsCGFsghlj4QZPuEv5f6+kJ1qZqmaheLwow41lZWXY7Xb6+vr2SL/kG6LRKN3d3RiNxrQjLMUeuLCwkJaWFsLhsNoGrehixbdBZ6OXFk84ig5fNBo98SFPBUcxG3tckFcEky60rMH4fD4ikQh2u52bN29mnY7Qush/1GL+xsYGX/30/2VjaYtYREKK7Bbxo6EokXCUkC+EwWjEZDYxPTxLZXMli+PLlNWWMNk3szuv8tIwdR3VbHt3WFxYob6rlo3lLXa8fkxWI1a3lZl7u5GDy+OgpK6E+3FqyMm+LgBdT7QzdGc6oQ3Z6rTgrnCz1pNYgLfYzdSfrmMwxYxLbVsFG4EoO8vbe35X31HBsj+Ef2O39iMI0NhYjNlsxh+NMejdZMuX2DhQXuyitr2Yl+bnEbdl7BYTHdWl9M+tcO5KNeOBDaqcbsKuJX6kqI73nntbWu9DMpI9XpThxqmpKfx+P1arFYvFgs/nS1sF+SSh2PHa7XbOnj175PqG1Wqlurqa6upqJElic3NTjfD6+voySikqiE+lybLM/fv3iUQiVFfvtoyf9JCngkyELtN9ro8KHlmC0SKCWVlZobe3F7PZTFNTkya5bq32poT3RyGX2dFZ/s/vfwffmg9RfKgpJsZERFEiEogQCUUprXGzPL2K0+MkHAgjGAQ2V7epba9i+LVx2i83Mz0wRzgYoeNaKxN9s0RDURweOw6PnaXR3dRTWV0JgsnA9MAuQQiCQNetdgZeeUg2giDQ9UQ7A68lSryU1hQhWMysJnWEOdxWCmuKGevfW5hvOVfLzMIW4dDem4zW8zVMLG0SiYhUVrkxGiW2NmNMTnpp7CxnPhYiGDfAWVrspLLEjVRo5B9mdyOas3XlzG/sMLK4RsvFMhajfsodTsSCJX6itJ0fPfWDab0Ph0EZbgyHw0xNTdHa2orFYlEJx2Qyqam0gwRKTwqhUIg7d+5QUFDA6dOnNTucDQYDxcXFFBcX097erqYUV1dX1ZRivF7aYa+DJEn09/ezs7PD1atXsVqte4Y8tWqDTgePkpul1sgrgkk3RWYymQgGg0e+jizLTExMMDExwZkzZ5ientZMBVlLmf1gMMjMzAzl5eVpt3v2vtTPd//7C0SDUcSYRCwSIxYT1YghFonh3w5QWFaAfztAOBChoqGMuZFFCkvdmK1mtVNs8PYokrTbKaZ0jRVWuZElgfXZ3dRT/alqNlZ32NnY9UHZlYlpSCAXs9VM84WGPeRS11HJ5laInSQdsOKqAiSTicXpvfIundeaGB5e2aMNtvu7BvyxGG1tFczMrLG0+LCu0na+hoGVddXeuKLYRVmxi6HJFUpaC7kzv0htSQF2s5m+2RU660rxlYrIRig22xCKV/nJ4vO8tf1GWu9Dukgl/VJTU6OqIMcLlBYXF6uEc9Ity4FAgDt37lBSUsKpU6eO9S47VUpxbW2NwcFBIpHIga+D0todCAQSGg8yaYNWiEYrssnUC0aPYHKMbKIEURTp6+tjc3OT69evU1BQwNzcXE60zVJB+bAXFhbS0NDAysoKo6OjuFwuysvLKSsr2zd18p0/e56eb/cjRh4U8hX3yZiorr3j9WG2mLHYLIz3TNN+pZnBV8bUTrHATpCuW+30vzSCwWjYFZB80ClW3lLM9mqQsH+3Rbj9ajOT/fNEw7sRgdJyPHJ3St3TbuqslJGkNuTWiw3MTqwRDiUOk9a1V7K5E2Znba+8S9WpEgYGEjvObHYzdU1lOMocjE2vs7W198aj40odvTMrSLJMbaWHAqeVoYkV1jYDNF6uYGTdy4WGSvpml3FYzFxuq2ajMILZYMRiNGAp8fLusqs82ZS9hWw8DpJ+iVdBVgRK4+2Sj6ITdlT4fD7u3LlDVVWVJi6ymSA5pagItSbbRpeWluJ2u+nv7ycYDHL58uUDu9ROYshTwUm6WeYb3lAEEwwG1eLkzZs3VbVYrQc3s5HZV0J4o9FIU1OTKkippAympqYwm83ql07pQvqLZ/+GiZ5pxKiMDBhNuym2UDCMFJOwOiz4NvwEfSEauurof3lEbTk+daONsZ6pXcOt8w0MvDyC3W2jqqWSgQc1laquMlbHN1VZl9NPtjPwyrgacZbVFiOYjMzEmXWV1hZjtFj2tCGfutnK8L25hNoMQOv5emZnNvaQjmAU6LjWwkDfPIIBSitcmK0QDUtsrIUQHQbupLBEhl1743uTS7TWl4IMo9O76gJWq5HKc6UYLLuHRs/0EmfqKghIESJlIEZlCs1WHCVe3l95k/P1HUd6T1MhU+mXeIFSpf1X6Uq7e/fukQvk6UAx1KurqzvQFuAkkCzUGo1GWV9fV3UCY7EYRqORtra2jNY9riFPBaIopqVM7fP5HqsOMsgzgkn3w3uUIv/Gxgbd3d2Ul5fT1dWV8AHR2sflKGsdVMy3WCxqQTS+C+n+/fv4/X76/3qcwHoYk9GEFJOIRmJqp5gsg6vIidlsJByIUN9Vx2j3FK0XGhl4eZQzb+qk/6VhPJWFOFw2xrqnKK4qwuqwMvFA2qX2bAXzA7sHs2A0cOpGa4K1cf2pGjZXd9jZeNixVddRxdZmiJ2VuA4vQeD0rbY9asgAp663MNy/oM7NKHC4bbRfayYcidHcXsH8jJfVxd15AbPFSFlHKSPje31cBINA+/U6RANUlRYwGuf14nCZab1cxfjaJsuLPsoLnNQWF7IZDWGpMxNDothmo7hohw80/hCN5bUZvJMHQwvpl3i7ZEUnTKlX9PX1UVRUpKaQsjmwNjY26Onpobm5mYaGhiOvc1wwm81UVlZSVlZGT08PoVCIsrIy5ubmGBoaoqCgQI3yMmmY0GrIU0G6sziPmw4Z5BnBpItM52BmZ2cZGhqio6OD+vr6Pb/XWtvsKAZmSuRyWDE/PmWwsrjC//rE19ha9hH0B4mGowgYdmX2JQj5IggC2J02xrunKa0tYX18mfrOasxWE1f+yTk2lrdoOlfPxtImC2PL1LZXsfOgJdloNlLRVqqSi9Vhoe5UTUJ9pf1KM1MDC2qaDKDtUhOz4yuEgw9/ZrQYab3YlJJcTj/RxkDPLCUVhXhKnFitZqJRkWAgjGy3cTeF94ujwEZRUzHTM3tblAs9Flx1LoZm1giFE29E2lrL2HRLvDw+h8mwO98yMLtCWYWTkFPGZTJiMxqpLgryc80/QHVp1f5vXIYQRVG9KdBK+iVeJ6y9vZ1AIJBgKHbUSfr19XXu3btHe3s7tbXaEazWEEWRnp4eRFHk2rVr6sxLKBRSoxulYeKottGQ+L1Od8gzfo8n7WaZL3gkCSZdQpAkiaGhIRYXF7l8+TLFxcX7rqdlkT9dslIm848k+3JnhG/91+cJ+2I4LHbMgoWIMUwoFCIUjBDYCRINxiipLmJpegVHgZ1IOAqS/GCepZjBV8Zou9TE3Mhup9aFHzpDcCdEQVkBhdVuTFYTUZ9IaW0xyDI2t52xnodDkKdvtTP42kRCY0aqNmSby0L96To2Vndo6KzEardgMu9+4Yx2KysLWxiNRtYWt9S5leKqQgwuOyuTe6OTogo3xiJ7ArlYrUZaWiqIxkS2rTC2mNggUFvuxFXkYMoQZG01QGd1KduBMJPLG1w4W43XGqbU4MBpMtJRZuCnW36IksLUn5ejIBaL0dPTgyRJxyr94nA4qK+vp76+nlgsph6y8ZP0SnSz3wDiysoKfX19dHV1UVWlHcFqDVEU6e7uRpblPfM4NpuNmpqahIaJtbU1RkdHCQaDapRXUlKSdhSpEEe6Q57Kv0wIRu8iywOkQzCRSISenh6i0Sg3b948MFWgdQSTTvpuv1bJdHD7G6/z0l98n2go+sDaOIYYkxAEAzabHUEyEPAGKSx2E/QF2dkMUFRVwNrcOu5SN3arien+uYROsa4nOuh9fhBJknGXOTFZzGwu7XZhVTaVEYmILE6u4PQ4sDut1LRVsr64RW17JQaDAAIUVRXjXd6muqmMWDS2O8jpsCKbTIz0JM6y7Pq71DDx2tSe51fdXMZOVGIraUgSoKqxmB0BVpe3EQQoL7fjdjmZnd1kYXkLodzOwuLD2ZiW+hKi0RghovSHN7AYDTQWORmaX+NsXTnl9W427RFsRhNWk4HTJQZ+tuvNabs5poN46ZeLFy+eWLuxyWTaM0m/urrK9PQ0/f39KSX3FxcXGRwc5OzZs6oVeD4iFovR3d2NIAhcunTpwNc0vmGio6NjT5SnlW30fl43sVhMVVk/aO3HTegS8oxgtKrB7OzscPfuXQoKCtKaNFYsXbVAOkX+ZA+XTAqG//e/fYehF0eIRnbnW2Lh3TZkMSKCsPsabq/7MJlMuIvcjN6dov1qM0OvjFHRUsrG0hZhf4SG8zX0vzSCYDDszqe8vCvtUlLvIbQTZdO7Sy5N5+pZnlkjsLM7mBgNR6luqaT3hRF1T0ob8r24n8Gu6nEgFGNzIZEoCkpcuMsLmRhcJBlNZ2pYWPERDOx9PxpOVbDsC1FeXkBRkZXFhW2Wl4IsE6S81oPPDuuru+TS0VRGOBJjfGad5s4yNonRWVJB3+wypnCM2gI7W8IWgZiAJ2zH7LJwqdzOe7p+QNMieTAY5M6dOxQWFmo6O5Ip4ifpW1tbCYVCe2ZNbDYbW1tbnDt3LsGmPN8QryRw4cKFjAk7OcpLZRutaKZloqIOexsF1tfX1QFa5czar3bj9/vzOmI8CvKKYNLFQTWYpaUl+vr6aGpqoqWlJS3S0rqL7KC1jjqZL0kSf/bZv2KmfwE5utuCHIuKxMJRwqEIYkzC5rTh3wrg2/TTdLae/pdH6brZTv9LD9wnu6cwmgw0n6tn4t4MJquR4roilVwqO0pZn90h9qBucepGKyM906rVcUGJm8LygoQ0mdNjp6y+fE8bcvO5OhbmNgj5E4mivL4Y0WBifmqvPljHlUZGxlYTDMgA7A4zTWeqiZkMsBPa4+NS11rKQjRE2BfjTFsl65sBhh/4x7SfrYRCI7ElH30zS5ytr2R82Uv1KQ8hYwy7KBIVd/DIO7RsX2B8fJyysrKM6hX7QWnvraioyDvpF5vNRm1tLbW1tYiiqKaSTSYTfX19CaKUWkZz2SIajXL37l3MZjPnz5/POhpMto32+Xysrq6ysLCQ0A5eWlpKYWFhRo0CGxsbDAwMcOrUKYqLiw9tg37c3CzhESUYRQoiPuSUZZmxsTGmpqY4d+4cFRUVaa+nZRfZQfWco5KL1+vlq//x62zMb8KDIUzBKCAGREKBMCF/BJvTgixKLE+t0nCmjtHuaVovNNL/0vDDTrHyApyFDibuzez+f4+L+dHdFuKa0xUsDK/Bg9JJ580WBuP8WcobSpFEmbmRhy3HpdVFGO1WpocSI5HO6y2M3l9Q528U1HdWsb4RxL+9V96l62Yr9/seKAEYoLahhAKPA/9OGHuJg/ujy3sMxACaz1SyGo3QUlnK5LyX+w+ejwCcvlLLZGCL5Uk/HdWlbAfD+CMRWs6XEjKKuA0WTPYobynr4J923FC1sZR6hXKwHFSv2A+K5UNDQwNNTU15RS7xkGWZyclJVldXuXbtGm63Wz1k5+fnGRwcVEUpy8rKcLvdOXsu0WiUO3fuYLVaOX/+vObRoCAIuN1u3G53gm302tqamo5L1zba6/Vy7949Tp06pUYlh7VBz87OcuOGtoO8uUZeEUwmKTJ4KMEQi8Xo6+tje3ubGzduZJzHPIki/1E9XKZHZvjrz3+X4FYA8UELciy8K1oZi+3WOFxFTowmE9ODc5Q3lLM0tUZpdRGT92d2W4pfGqa6tZygL8z86BLVrRWE/FHmR5cQjALVpypYGNy94zeaDFR3VTAUN3Vf21nJ5rIP//bDIcbajip2tkKszyZ2cXU90c5Aik6xtksNTE2sE03q6hIMAqdutrDji3D6Qh3hUJSFWS+zk+vAOqduNtGXIpUG0HWljqAJ1qc2WfE+lOjvbC7HXGHl9sw8ZQUOTteVM7O6SWtVCYYqE0EhisNoxu4S+fH6Tm7UnQJIuJNV6hVTU1P09/fj8XjU7r3DWn8Vkmpra6Ouru7Ax+YSsiwzMjLC8vJyggtl8iGrpNKmp6cTurFKSkpOrJ6kaKA5HA7Onj17IqnGdGyjldcivg3a6/XS09NDZ2dnypRXqtrNt771LXp7e3n7299+7M/rJJFXBJMulA91LBZTQ2ar1crNmzeP1J1znG3K2Xi49L44wHf/+/P///bOO76q+v7/z3P3yN57kRAIKyFsEdyIg6C11dq6asVfv2LrrB2OVttaq9+Kq9Wv3yr22+ECBBdVGU5ESNiBhGySkORm3CR3r/P743KOuSFAAjdL7/Px4PHQmzvO5yT3vM/nPV4v3HYvbrfHP6Hv9uLz+hBFHw6ri/auTuLSYjG3tiGgwBChxxhlQKlSkJAZR82eeiYUZtJc1Yrd4mBCYRbN1W3YLQ7UehWxGTFycPEbgsVTu//rocWMaSk0VZrw9Ulb5RZl0VhjCmhDVqgU5M+eMGBwKZify8G9TYg+EUOYlvjkSIzhehAEfBoVB/cfPU76RaFUkDs747jgEhmpJz09BiFKw1cVgeKYk3ISsDlceONUHDjaSlFWMgca20iMDCfCqMWRIKJS+jAq1BjDPVydOZWilJzjjrd/vcJut2MymeQuJIPBIAeb/mkTSfpl6tSpQ9pFjzSiKFJeXk5XVxezZ88+oRRR3xmsvt1YlZWVOJ3OgJmbwcoZDRUpuBiNxlGzMOjbDp6XlxdQw6qpqZFto3U6HTU1NUyaNEkW2TzV+27evJmbbrqJl19+mR/84AcjsJqRQxAHI/41Qvh8vlP60Et8+OGHTJ48mYqKClJSUsjPzz/tPzxJjuWss846rdf3pbW1lerqahYsWHBGxfxN//6Usg/24jkms+85ZhDm8/pA9Bfb25s6MUYZUamUtDS0kzohiYZDzUTGhiECpiMdTDlrIo2VR4lOiiI+PQ6H1YnL4aLb3INap6alqhOv20tsShQqnYbW+q/bgqeeNZHy7YFtyGlTk2iu6gpUQzZoSJucSvX+JhRKgYiYMMIj9eiMWsLiwnE4PNitLsztvZg7/JplxnAdMdnx1Ncc34as1atJLkii6litRamE9PRo1Go1NbXtTJiVxp7qY6kwASZPSKS710FbZy9pM+JRaJQ0tHejVSuJMuhQahQoUlQIStAqlERFCPx4QhGZ8UmD/8UeQ2r9lQIOIKePbDYbtbW1zJgx4zjpl7GEz+dj//79WCwWZs6ceVrzOKIoYrPZ5PNgNptl2ZaBAu/p4nQ6KS0tJTw8fFSbJE6G1+ulq6uLpqYm2traEASBmJgY+VycLPB+8sknfPe73+Xpp5/mxhtvHLOp1NNlTAUYURQH1c0liiIfffQRoihSUFBwxoNgHR0dHDhwgEWLFp3R+4A/PXLo0CEWLlwo51qHsmsBWPf021SV1eNxef2ClU4PHrcXt8sffJUqFZ3NnbhdHuIz4ji8s5b8Obkc3FZJRkEa7Y0dWHvsfuOvzyuPmYDlf21tnBqBxyVi6fTrfU0ozMRhd6FQKtDo1Kg0KoyRBpx2FyAgAKIA+gg9Dpvbr8jsdOFyuvAJPkSFCpfNi9Pmxtbr9EvdqJXkzsymYgAJl+jECNRRRlr6KSiD3wgsIj2G5qNm0tKicLocdHa4cDo9sr3xgdo2tGolE7MTaG3vpa3Tgt6gJm9WMg3mHtq6rUxJS6Cxs5v0pCh8CQIo/MElPlLgp5PPIiII7aCiKGI2m+WisNvtJiIiguTk5FNeWEaLvi6UM2fODNo8Tl/Zlv6BNyYm5rRMvyT1ZqkDbyxffKWa24QJE4iNjZXPQ1dX1wmHXb/44guuvPJK/vSnP3HrrbeO6fWdLuMuRebz+SgvL8fr9TJ58uSgTBkPR5H/dIr5LpeLfz36Jq1VHfi8ol9i333sYu5w43F5MITrsffa6e7oJXu63xhs8vw8DnxewaR5uVSV1aFQCkyc5bcz9qsbZ30tWJkbQ3eLDdex9FZecTYNFUePBZNj0/r5KezaclA+LpVGxYSibMq2VAQcb0JGLB4UdLaaAx7XGtQk5CQOGFySs+OweqFjgOASmxRByqRE7A4PSqWCurqv6zuSvXFzZy/T85OpOdLJvkp/+iwnKw5HjMAX1Y1MSU/A5fbgE0WiY/Q4Yn3olWo0ImRG6vnZ7MVB+yJLzpVHjx5FqVQydepUrFYrbW1tVFZWYjQa5VTacAtSDgZp2FMUxRO6UJ4ukmxLUlJSQL2ir3yNdJEdTKeUw+Fg586dREdHU1BQMOrn7mR0d3eza9cucnNz5Zqb0WgkMzPzuGHXsrIyNm3aRHFxMS+88AK///3vv7HBBcZZgHE6nezatQufz4fBYAjaHaJKpQpakV8QBJxO55Bl9tua21jzxPv0tvf6G7mOzbS47P6dgsPqRKPVIAJHa1rJmJxO9e46JhRmUv55JVMX5rP/swoi48OJjIugcmcN4bFhRCdGUbHT3w2WOiWBlsOdst5XwVkTOfRVtdydFRUfjjE6jOq9Xw9Fft2GHGhjnFmQSrspsPAPEBkXhipcx5EB9MGyClJo6bBiO9a6LAiQmhFLZIwBhUZFTYuZXf0GMgHCo3SkTU/G7vLQY3Gyt8IfWLJTYwiP0VHt6UXrUZGXHEu3zUF0mAFNhAohUolSpUDjE5keE8uNs+YG9Ys8kPRLXFwcmZmZJxSklKbHR9rbpe/syHAPe/avV/SvYUmDjSdqB5dmh2JiYobdGuBMkcRAJ0yYMGBDR/9h19jYWCorK/nHP/6B1Wrl//7v/+jo6KCkpISioqJRWMHwMqZSZOAPIgMh3SVER0czdepUvvrqK7KysoIymGS1Wvn888+56KKLTvs9pGK+2+2msbERk8lEd3c34eHhxMfHk5CQcEKvh8qyKt59cQvOXgeeY7bGkty+UqWQZVUQoHpXPZHxkfSarRjCdLQ1tJNblM2h7VUk5yTgdnnoaO4iMSser8dHe1OX/0I+NYmmcn9Nw9+5lRegKZack4DT4aGr9euhSKkNua1fp9jEWdnUVrTidgV2hCVlxeH0Qle7hf4kTIiivdNNbHwY0XFheD0izQ2dWC1OsgqSaOqyYrcH1t/CwnRkT0yg2WWnsc9x5aTHoFAosIkenNEKIg06qlrayU+Nx+pwYYzV4osW0CgUaHw+LsnN5/zcoSnsnoq+0i+FhYUnTTVJjo0mkwmTyTTi3i5Op5OysrIzdqEMBtJgoxRwJPkaaXfj8XgoLS0lPj5+zM0O9aenp4fS0tIhiYHu27ePSy65hHvuuYcf/ehHbNy4kffee4+oqCheeOGFYT7ikWdcBJjm5mYOHDhAbm4uWVlZCILAjh07SE5ODkqKzOFwsHXrVpYsWXJaf9D9i/nS4JTU4tnW1kZHRwdarVZuhZWKoNveK+XztTvwOFz+mov7604xn0/0p8NMPcSlxdDTYUGhVJKQEYcoijisTlQqFQe+qGBCYSZHa9uw9zrImpZBW0MHth67LFh5tMK/o9Do1WRO8SsqS+RMz6ClvgN7HxthqQ25p9MasNaCBXkcLGs4zhgua2oqba0WbBb/70+hFEhIiSIyJgx9tAFTp5WWRjNuV2AqcsL0FKqbzbIRmCDAhAkJKJUKrG43bbjpPqYikJ8dj8vto7axg+y8WDSJeg40tpGXHIvZYic2wog6SoknXESjVGJE4Nop05iakjrk3+nJ6Cv9MtRhv77eLtJNSFhYmJxKC/acyXC5UAYDqR1cOhe9vb2yy+fkyZPHpGW0RG9vL6WlpWRlZZGVlTWo15SXl7N06VJuu+02HnrooTG7tmAy5gKMy+WSL15Sn/6RI0eYMWNGgHxFWVkZsbGxQZERd7vdbNq0iQsuuGDIMiHSoNSpivmSbERbWxvt7e0IgkDllnpaK9oQBA0+txevx4PX5cPr8YAg4HV7MTV2og/XodFpOFrTRtrEZBrKm4iMCwcB2ho6mLZ4Ej3tFoyRBgyReqw9Dno7eunptqAP12Oq8ws/hseGEZUYyZGKr1t/J8/L5fDuBnlaHyB3ZhaN1YFtyPKup08bsqAQiIwNI2VCIqJCgUIp4HZ56TXbMB3txuPxMnl+Lgf2BrYTS2RMjaem0S9JExenQ29Q09XpwmJxkjUpgTqrFbfby6ScBDq7bTS39aBVK8mfnsy+7g5iw/SolAr0GjWiKGJM1OEx+FALAjFKNbfPmUd0kLWdgi390nfOpKOjI6g2ySPpQnmmWK1WduzYQUREBAqFgo6ODrn1V/I9Gm3LaAmLxcLOnTvlIdrBUFFRwdKlS/nRj37E73//+zH9uwgmYzbAuN1u9uzZg91up6io6DgZ6z179hAWFsaECRPO+DN9Ph8ffPAB55577pBkMU53Mt/j8fB/D79G1d56HDaHPxUmKFEISvD6d0AarYrOFjMOq5Ok7EQqdtQwaa6/Uyy9IJWOpi6sZhtTFk6UC/hTF05i/7H/jkgMQ6FQYut2EhkXRmJWnF8Q81ggAAF9uA671d/1JfoARHQRBpwODwLSTsxv+KXUaXHY/Y0GDrsLa49/dzNp7gQO7j5C/78ipVpJTlEmFeUDDEkKMHleDja3G5VKoKG+A7v96wCXkhtJJyJpiVE0HDXT3etAo1aSn5MAEUoO95pJiYmg02IjQq/FoFGjiFUgGgSUQF5YBP9v7vygX5Akjbvhkn7pa5NsMplwu90BqbSh/G2OpgvlUJGONTU1VZZ3klp/peDrcrkCUmkjbRnd91h37txJenr6oK891dXVXHzxxVxzzTU8/vjjY2oXOdyMyQDT29vLrl27MBgMTJ8+fcBul/3796PRaJg4ceIZf6YoinzwwQecffbZgzZoOt3JfLPZzJuPvUNXsxnRB263B2uvDXuPDZvFjsvlwRCmRyEItNV1MqEoi4Pbqpg0L5fyzyuZNDeXql11CALkFGZSuaMGpVpJXnEOh7ZXAxCXHY2ty4m915+uypqWRtuRTmzHCvJKlZKJsydwaMfXUjCCIPin8PtM8IPf8jgmPZaGytaAxwWFwOR5uZTvOr4orw/TkpCbRG1VW8DjUTEGktOiESJ1HK5qO67mApA/K5Uej4u65m58ImhUAlmpUbR22EkriMWlgbo2M1nxUXi8XtRqJeoEFaIWVD4f56ZmctnUqUG/oI609EtfXaz29nZ6enqIiIiQg83J0kdSvTIjI2NMy9TA16mmkzlm9rVKbm9vD0grxsXFjViHntVqZefOnaSlpQ06uNTV1bF06VKWLVvGU0899a0KLjAGA0xTUxO7d+8mPT2diRMnnvAP5+BBfxvt5MmTg/K5H374IfPnzz+l4U//yXyp3jIY6ivreefZzVjMVv9UvsuL2+WfNVFr1cem8524nE4aypuJTouis7GbpOx4jpQfZcrCfA58VkF4TBgxyVE0HGwiLNJAXEYcdcem75MnxdNe143nWMpr0twJVO1pwHOsIC9ZIfed1ldpVP6ZlbLATrHYlGiUOg1tTYHeKmqtiqzpmRzef3zqKzIuDF1cBEcbuwiL0JOSFoVKo6SjzUK32UbipESqawLFKqNjDKSmRiNGqNl5zJsmwqglIyWamiPtON1eEnP1NDpcpMWE4/KKRIUb0aiUKBIVCGrQiXDdtBkUJJ96enqojAXpF6fTGZBK02g0x9lmw9h3oeyL1IGVkZFBTs7xigonQtIIk85F3w69oZiJDQUpuPTdZZ2KI0eOcPHFF7NkyRL+8pe/fOuCC4yxACOKIl9++SVJSUmnlFk4fPgwTqeTqVOnBuWzN2/efEp/9BMV8wfD7k/2seWfX+CwOPG5fXg8Hjwur9/i2O3G2m3HaXOSkB5HZ4uZ2JQYFEoBl9vNkYPNRCSGcfSQiZi0KARRoPOombi0WBRKJW0N/gJ+2rSvO8UAf/rsmFIyQExyFFqjjpY+9sHGSD3xWYnUH2wOON60iUn09Djp6Qos8hsj9cRlxlN/OHB3olIrSJuQQERyNA6nm06ThbY+fi4R0QYMqZE0NvqDlVotkJUdh9cL9fUdZBensq+mleT4CGKjjVTWtvl9anKTsEaK2H0e7E4XeiV43G5USgWqVA1avYbsqGhuKz49++FT0dzczMGDB8eU9Etf2+z29na8Xi+xsbHodDqOHDlCfn7+mHahBP8uq6ysjOzs7EEXyQeib4dee3s7drt90FP0g8Vms7Fz506Sk5PJzc0d1Hf+6NGjXHzxxZx99tm8+OKLY6Z+NNKMqQADgUX+k1FTU0NPTw+FhYVB+dyPP/6YadOmndD1sm+9ZSBb1JOx+bXPKf3PbrzOYxL7Lg8+j5RiE7FbHJhbuolNicbaY8dhc2GI0NPbacXr8hCXEUu3qYfU/CTsNjt2mwMRcHa76Trag6BUkDo5gaY+gpUTZ+Vw8Ktq+RhS85Kw9jrpbu+VH4tNiUZt0NLaz3Z4QlEmjbUd8vCl/PzkSLRR4fh8PsIj9Wi0KjxuHz1dVrRhOlo7bVgtx3cBxqVE4jWo6bU4SEw04nS56Wh34nZ7UWtVpExLxKcAUYSqehNqlZJJOQm0dVmInRiN2elArVSACHq1CqVagTvai83WS7rXx9kJSfLdfHx8fNCm0+vr66murh7T0i+iKNLb20ttbS1tbf6g31eYcyzKv0vpxuHYZUkden2n6KVUWmRk5JB3EVJwSUxMPGlGpS+tra0sXbqUWbNm8corr3xrgwuMwQDj8XgGNVVfX19Pe3s7xcXFQfnczz77jPz8/AGNlk63mA+w5rl3qC2rx+3w4HH5ZE0xr9tvECaKIm317WgNWgzhepoOt5I+KYX6A40BnWKT5+dRVVaLx+1l4pwJ1O49gtvpwRitJyLJiAIlhjA9OoMOjUGH1WzD7fLgdrkxRhro6bRh67Xj9YiIPh8xqTG4PT7sFgcqtRKVWoVGqyImJRqH0z9/o1AIiKK/TqTVa2lp6aG95Xip/byiDGrqO49rQQZImxBHVHo0FouT2loTXu/Xf24R0ToyZ6RQe9SMqdOCUa9mQmYc9U1d6AxqjFlh9Lpc6DVqDBoVKoUSpRp8CQr0CpGrJhZQlJ4hT89Lra6SU+PpXmAl64empiaKiopOuqsdC0jeJdOmTSM8PFy+m+/s7JSHP4PlcXOm9JVUycjIGNbPcrvdATs9QLaMPpXcPvg7Bnfu3ElCQsKgg0t7ezuXXHIJU6ZM4Z///OewpOvGE+M2wDQ2NtLc3MycOXOC8rnbtm0jOzubpKRAAcTTDS52u53XHn2LtvoOPC4fvmO2qQgCDpsTW7cNnUGLua0bS7ed1NxkDm6vYvL8PA5+UUnG5FTam6VOsXwOfO6XaZm6cBIHvjiMKIqExerR6rV0HnOMDIsxoNSp6G79Oq1VsGAiFaV1fpHMY+TOzKKxph1nX9dIQWDKgjwODKCGnDM9neambhwDuExOnpvDwYMtshqAVqciLTMWnV6NUq/mQK0Jp+PrgUylUmDChAS0Rg1NLjtNbT3ERBpIS4rkcH07EUYtyalRdGrd/iAngFqpRK1QoNIq8MRAilHPfxXPwThAQ0ZflVvpApuQkDBoAUafz8ehQ4fo6Ohg5syZY3IH0JcjR45w+PDhAXdZUmt836HGvmoCwZSKGQySjP3EiRNHPIUniqIsX2MymbBarURFRcnnw2AwBPxtSO3ocXFxg+4Y7Ozs5NJLLyUnJ4fXX399xM/vWGTcBpiWlhZqa2uZP39+UD53+/btpKeny7WfMynmtza1sva/38fc3osoF/M9uJ0eXA4XTruLiJgwDBF6nDYXap2GvZ8cZGJxDuVf+DvFqnfXIYqiv/j+VTUKpYJJc3Mp3+afvo9Ji8Rl92Dt8gtWpk5MwtLjoNt0LAUmQGpBAs0V/fxaFkykYldDQMBRao4JU/ZzpQSYNCeHwwdbjnOZRIDJ83Opr+skOSUSrU5Fb7eDpoYOvB4fE4szOFTfIcvwJyQaiIuLovFIJ7pwHRYDhIfpMGjVVNSZSIgJIybSQJfDTnhWOA63G7VSiU6t9O9c9AJug4dFWRksnzp9UL+HgZSPpZ3NQHItfaVfTldleCSpq6ujtraWoqIioqKiTvrcgS6wkj7YYDxuzhQpuOTn55OaGtzB19PBbrfLqbTOzk60Wm2A7YA0Zzdp0qRBfe+7u7u5/PLLSUpKYs2aNWPKBXQ0GXMBxuv1yt7VJ8NkMlFRUcHChQuD8rlSnjU9PV0enpT0yYYSXCrLqnnvfz7CYXXJ/i1+NWQPLocbe68dQakgMi6cmj31pE1MpfHwUdImJuNyuIlPj8XWbcPaY0OlVVGzu0Hu/KrZ4w8AiXkxmJssuBz+Nt/cmdkcOdwi70j8XWFZVPTdjQiQNjWZpopAq2J9uJakCUnUDmDqNWVBHgf6tCFHxRqJTYxAZ9CgMGhpOtJFR1vvca+bND+H/RVHSU2NAsFNb4+H7m5/bSZ9QhzG1HBMXRYaW7rJTInGoFPT1NZDXn48RwUHeo0ajVKBWqVEKShQ6iA8XsP3p0wlJyFhUL+H/kgCjFIqzel0yumSuLg4lEqlLP1SVFQ0pu8+RVGkurqaxsZGiouLh2ywB8j6YCaTKaBWEUypfYmOjg727NkzaI+UkaZv04Q0c6PT6cjJyRlUTa+3t5eSkhIiIyNZv379mL8xGUnGbYDp7Oxk7969nHPOOUH5XEnnLDMz87Q9XD5/Zydfri/F6/QPJLqPyez7vD68bi+Wbhv2XjsJGXE0V7dijDTisDrQ6DW0NXQwsTibg19WkZARBwJ0NHWSNTUdrVGHoFBgt9vwCV7cvV4sZhu9HVa/YOWOWnlHYowyEJ8RR335111hfjXkLCr77VDCYgzoo8MwNZoB0OhURMSEYYzUE5sag9PpwePxYet10NHag83iRGfUkDwxmZp+czHgb1+esmACVreHpsZOuru/lp7JyIwhNjWK3UdasTvcTMxJwO320t5pITstFrdGxB3ur/uoFAJqpRKVUgCtyMSsaH40O3hWsn3nKiS5FoVCgU6nk+sYY3V2RBRFKioqaGtro7i4OCgpvJN53MTGxp5RHcFkMrF3714KCgqCohs4nDidTnbs2EFYWBjh4eG0t7fT29tLRESEfD76zx9ZrVauvPJK1Go177zzzrDvBMcb4zbAdHd3s3PnTs4///ygfO7evXsxGAxkZWWdVjH/3Rc/pPzzSrxun+w+qVQJoFDg8/jo7bDQ3tRJTFKUvwbT4yAs2khPey9et5e49Fjq9h8ha1o6HU2dWMw2vx5Yl41uUy+CQiBtSgKNB461BwsCUxbm01jZgiFch1avITwuDJ8PRJ/f1x4ElGolSq3ab1Usggj4vD5UWiXmbht2qxOXw4PL7sHl8KDRq8mYkk7Vgebj1hgRayQsIZKmY11nguKYGnK0AYfDjRCp52BFi/z8qGg9aakxmEy9RKZGUNPZTXZaDC3tvXg8XtISozB1WkieEE2v0oNKqTgWWBQoBAG9XqBk7iQK04avGCwVcjUaDRqNRq7bSHfzUVFRYybY9HWhLC4uHha/GWmnJwUbm80WoCYwlM9sa2tj3759Y6rF+0RIxmaSZpv0O5fmj9rb22UpH0ne57zzzuOmm27C6/Xy3nvvnXKG7tvImAswg3W1DIYCcl/2798PQF5eHkqlctAXFZ/Px78fW0dTxVG8Tr+OmMflb0V2uzwICn9Lc8fRTqLiIzFGGbF220EBNbvriYwPRxAUtNW3M3l+LlW76vG4PH6floN+nxaVVklCTixHK/xtyGqtiuwZWVSWfj11nz4phZ5Oa4A4pb8NWUdrQ2BaLHtaGi1N3ditgS3F+jANujgjXa2BEvwA8WnRKI06tHo1xjC/bExzQyd2mwu9UUtcfgI1te1ERGgJj1CBqKKpyd98MG1BFjbBR1V9O7GRBqIjDJg6LSTGhSNEq3CofccK+UqUAgiCj0nxkfzwwnnD2uI5kPRL38K4yeQ/3yer24wUwXChPB36CnNKrpWD8bhpbW1l//79TJs2jYTTTGuOFC6Xi507dxIeHs7Uk6hASFI+GzZs4He/+x2dnZ0YDAYeeughrrrqqjE/ezQajNsAc6YKyBJSMf/o0aMcPHgQpVIpy+v3nZAeCLPZzJtPvEPnETMetweP0+d3nnS4cDlceD0+DOF6ejotdLf1kD45hYNfVDL5rEkc/KKSSfPy/LsJjRKtQUvDwSY6mjqZvCCfQ19W4fOJGKJ0GKMMtDf4BxTDYsKISY6m4dDXO4yJs7KpP9Qi12TAPyjZ2+ukpyNwUHLSnAkcPtCM1xPYSBGXGg1aDe3HhiONkVr04Sp0eg3GKCNdZjdtLT0BVsng39UkTUpEUCgwmcy0HutgCwvTkpkZixiloayymYlZ8Yg+EYvNRWy0AbfbS1i6EbvCh0pQoFaCAgUGLVxePIUZE4b3yypNvJ9M+qWvY6XJZMLhcMh38/Hx8SNWyB0uF8qhInnc9J2gl3Y2fYNvS0sLBw4cYPr06QO2/Y8lXC4XpaWlGI1Gpk6dOqiUuNPp5Nprr6W9vZ3LLruMzZs38/nnn3PVVVfx6quvjsBRjx/GbYA5EwVkif7FfOmC0tbWRltbGz6fTw42/e9e6yvqefe5j+jtsuN1e/H5vAgKJaLXh6XLQk+XFYNRh0KloG7/ESYUZXPoy8NMmpdH+ReHmTR3AtW76/H5RCbOyuHQ9ioEhYIpZ02ko9mMPkKHR3D771TdYLc6EQQBt9tHS12faf2z8jm0o0ZuEwbILcqisbZfGzL4LZR31qFUKQiPMmCM0KMzajFE6ECtxml3Yz1Wb5FakjMKEmlq7sXj/rqLTG/UkJYZiyFKz5FOK62t/tkYpRJychIQFAL19Z1kFyXjVQi0tvdi0GnQalUI+FuVVYk6vCpQKxWoBAFB8DI5LZ7vnz1z2HcJkvTLUNplJQ96qUmgrzaYNG8zHKm0vi6UhYWFY6b54EQeNxqNhpaWFmbMmEFcXNxoH+ZJcbvd7Ny5E4PBwLRp0wYVXNxuN9dffz319fVs2rRJbg3v7OyksbGR6dMH1+H4bWHcBhhJAfmcc845rXRBf9mX/n9cUlunFGycTidxcXEkJCTQfKiNj/+9DZfVjdcjdYq58bh9uJ0ubD12jFFGImLDsHRZUaqV1O5tICkniZo99bKmWFikgfjMOOr2H/FbFU9Oo6qsDoD4nGgsJjuOY2mszClptLd0YzXb0Bo0GCL0ZBSk0dXag1Kl8Kf1FBCZEEGv2YEogujz4fX6EH0iEQmRNNW1Y7M45fcEyC3M5Eh9J07H8ed88pxsDh1qwRimIzElEkEBXR0WOlotRCUb6UWB2+0lIcGAQqmgvd2Bw+4mOzcedbyBynoTORmx2B1u9Dr/hVGjVUKcGlQCSkGBApGoMC0lc/LJTx3+DqNgSb/01waT2lwlr59gDDS63W7KyspQq9VD9p0ZSaSmierqallNoG9hfCw2TbjdbkpLS2UTtsH8vjweDz/60Y84dOgQW7ZsGfO7s7HAmAswoijich0/0DcQH3zwAWedddaQO2mGOjwpKdu2tbXxwb8+5vAXVWiUOtRqDQqfAp9PRPT68Lp9WLtt9JqtJGbG0VpnQmvQ4nK4iIiLRKvXEBZtpLPFjMvhwu3w0NbQTlRCBGHRYTRW+gvkKQXxtFZ14js2e5I/ZwI1+xv9hXpArVORPS2Tw31sjE+khqzVa0gvSKVqAGHKyfNyqdjfJFsoA4RF6olPiSIyMRKb3U2HqZf21sDp/QnTUxGNamw2O0eP9uJxi0RFaUlIiMDu9KJKMKDWKGnrsJAQG4bT5UGvV6NQKvzBRSGiFBRoFCJz8tJZOntkPNeHS/ploLrNmXZhSS6UQ7m7Hk2ampqoqKigsLCQsLCwgOCrVqvHlK+LFLi1Wi3Tp08f1Ln1er3ceuut7N69m82bNx83kH06PPfcczz++OPyju+ZZ5454eD46tWruemmmwIe02q1OBxfd2qKoshDDz3Eiy++iNls5qyzzuKvf/0reXlfu7l2dnZy++238/bbb6NQKPjOd77DU089NWwNCuNax0CpVA5qKLMvoijKXWqD7RQTBIHw8HA++NvH9FbZiImIx2a1YTFb8Xg9aDRq1GoNHreH7o4eohOjsHRZ8Xp8qDQqrN122urbSciI5dD2KjKnpNHT3gsITFtc4E/VeX2kKkXUeiXtNWY5uExZmE/5l9XyTkuqwfQNLkqNiryZ2ccFl4jYMCISo44LLsYIPXnFWVgtLiYVZiCKInari862Hiy9DrSRBnbtqJOfr1YrScuKxRCmRRWho2x/Ez6fiDFMRXp6FC6nj+bmbnRhdlxGAUeHmbiYMMKNGgRBwGjQIKgFiFEhqAQUokh6jJHvnVVITOTwd970lX45laDp6aBUKmWnUmnnazKZqKqqYv/+/UO2R7bb7ZSVlREZGUlBQcGYDy6SmkBRURHR0dEApKSkkJKSgs/no7Ozk/b2dg4ePIjb7ZZ9XYKpGzdYPB4Pu3btQqPRDCm43H777ezcuZMtW7YEJbi89tpr3HXXXTz//PPMnTuXVatWsWTJEioqKk7YFBEREUFFRYX8//2vXX/60594+umneeWVV8jOzuaBBx5gyZIllJeXy393P/jBDzh69Cgffvghbrebm266iRUrVvCvf/3rjNc0EON6B7N161amT59+QoHK/u/bt+YylDZkm83Ga49vwFTT/vXwpMeDx+HB6XDh8Xjw4cPc2o0x2khkbARumxuVRk3NnoaATrFJc3Op2VuP2+lhQmEWzdVt2C0OFCoFKfnxNB30pxh0YVryinPo7rCg1WlQqpXowrSAgMftPWbwJSIIArpwA3abEwEBv1MYaLRqvChwuzx4PF5cDje2XgeWHgdZ09I5tKfxuHVq9WpSC1JobeomMTUKnV6NtddJU32H31lyfg4ms43IKC3tpm46O50kJ0cRE2PE4fWgiNFhd7rweb34vB5E0YtapUQbpkedqEelVBITruXi6blMnTAysvc+n4+DBw/S2dk5KtIv/edtwsPD5VTaQHUbyYUyLi5u0FPko0lDQwPV1dWDVhOQPG4k3biRqGNJeDweysrKUKlUg045+nw+7rjjDjZv3szWrVuDpp82d+5cZs+ezbPPPit/Tnp6Orfffju/+MUvjnv+6tWrueOOOzCbzQO+nyiKpKSkcPfdd3PPPfcA/lGOxMREVq9ezTXXXMPBgwcpKChgx44dzJo1C4CNGzdyySWX0NjYOCxDsGNuBzOUPzCVSjWoHUz/Yv5QgktrYytr/vw+FpMFj9d/YVeqlf68s92Oy+nGGGHA1m3H6xAxhhk49HkVKQWJNO8+QlZhGiqlCo1WQ/qkFGr2NeB2evySLTtq8Hp86MI1RCVGfB1cjFpSJ6aw77NK+Tgyp6TRVGPCYrbJj8UkR6Ex6mktDRygzJicQkd7B9aewHZjnUFLyqTkgOASFRdGTEI4hnAtXqUSU0svll4HlkP+dJ1WpyJjQjyGpHCONHdjMvVitapJTo5Co3XhdntQhanptrqJFgS0GjUatQ6Fwu+G6caDRe2gx9TF9ORoLi7IJiEmDFEUh/3i6fV62bdvHzabjdmzZ4/KhLXRaMRoNJKVlYXL5ZIvrrW1tXLdRpq3kYLLeHChBH/KsaamhpkzZw5qVyhlAsLDw8nJycHpdMrzNjU1NWi12oBUWjB3btLORalUDim43HfffXz44Yds2bIlaMFF6lz75S9/KT+mUCi44IIL2LZt2wlfZ7FYyMzMxOfzMXPmTP7whz8wZcoUAGpra2lpaeGCCy6Qnx8ZGcncuXPZtm0b11xzDdu2bSMqKkoOLgAXXHABCoWC7du3c8UVVwRlfX0ZcwFmKAwmRXaqYv7JOLy7hndf+OhrDxeXB7fTjdvlwWV34/F4iYwNR63XYLc4yCxI49D2w0yeN5HybYeZMDOT+v2NeL0+UiYl0PxFGwgCUxdNpruth+wZ6bi9LrR6HR67j4RMv3qxWqehes/XQSN/9gRqDzThdn09gJqal4TV4jpuxmVicTa1h1vleo0hXEt4lBFDpJ7IpCjcDg/5M9Kw9jhob+nGbOpFpVbS4/Biau1Bq1eTPTERg1GDpddB69FuxEgdLe0WIiM1eDwanC4RlVqNVusjMikciyASFaFHqRDQqFWojtkyi1oBTYSGOWnpXDZ7Cr29vZhMJvbs2QMg38mfqff8QEiW2z6fj9mzZ4+J7iuNRkNqaiqpqakB8iT79u2Ta4IJCQlj3oUS/Be0uro6iouLiYiIOK330Gq1pKWlkZaWFnA+9u/fj8/nC5DyOZPfn9frZdeuXSgUCgoLCwcdXO6//37Wr1/Pli1bhmSIdiokD5/+TSaJiYkcOnRowNfk5+fz0ksvMX36dLq7u3niiSdYsGABBw4cIC0tjZaWFvk9+r+n9LOWlpbj0m8qlYqYmBj5OcFmTAYYQRAG5QlzqgBzuikxgM/f3cm2tV/hcXjxeD0oBAVKjRKv14fX7sJudfpTQVFGDm6rJCk7gSMVR5k8Px8RKF4yncaKo2j1GhKzEqjd14BKoyQ+J5b9n/rzqLEZkTh63di6jxX3cxOxW12Y2zuITohAF6YjeUIiPZ1WsqelyWvQR+hxONxEJIgkZ8cdW6dIWEw4nW09xCVGYLc6sXTb/YoBUQa6e5zUVlUFnj+VgqwpqYTFh+M55s3S0tRFbWUreqOG9Jw4soszaGwyY7M5EGJ1REdHYLO7UQgC0alR2PChEkRUSiUqlYBC4f8nan1kZ8Vw+ewCYqP9FyC9Xi/XKaR28IqKCrlDT7qbP9Ng4HQ65Tx7UVHRqBeVB0Kat4qPj5eFIKOiorBYLHz88ccB8zZjTduqpqaGhoYGZs2adVo6aAPR93yIokhPTw/t7e3U19dz4MCB07ZgkIKLIAiDDi6iKPLII4/w6quvsnXr1oAi+Wgxf/78AGHfBQsWMHnyZF544QUeeeSRUTyykzMmA8xgUSqVJ5SVORMPl3f/dxPlXxxE9OKfzHd68Hr87b4+rw+lUkFMUgR6o462xg4iYsLpMfViiNBzcHsVE4tzKP3PXuLSYohJjsbj8jBt8WTUGjW9XVZi0zyojUra67r9vjBAxpQU2o6Y5RZic4eFgrxkdm05GHBsk+fnsfuzygA1ZEEhUDB/Irs+raQ/mZNTsDm96PVq8pOjUKmVeNxeerqsGGOMNDR246zvRG/QkJIRw+RpafT02LHaXLj1akSviFYn4HIpiIgMx+X0EhNjRBunx+bzolIoUCsFlEq/jpgg+shIMnLhedNIiY0e8PwKgkB0dDTR0dFMnDhRzss3NDRQXl5OdHS0fDEZqhyKzWaTC+RTpkwZ8wXy9vZ29u7dG6AyLNVtWlpaqKiokOs2A2lhjSSiKFJTU8ORI0dOW2RzMAiCQGRkJJGRkUyYMAGHwyGnFquqqtDr9fLO5mQeN16vV54hGuyNhiiK/PGPf+Sll15iy5YtTJo0KdjLk8VVW1sD9fxaW1sH3UCgVqspKiqi6thNo/S61tbWAM231tZW2ZQxKSlJbiOX8Hg8dHZ2BqVxYSDGfYAZaAcj7VqGGlx8Ph+vPrGOI/uP4vX4U2I+rxeFUkCpVuPzgq3HSmutiYi4cHo6LJhbe0jKSUAQFGiNWlLykjhcWkvG5BS623vpaO4iMSueo9VttB/ztk+blkhzuQlDpJ7waCPRaZH0mq3EZISjUkWjVquJjImk12wje2qaXyzT6yMuPZb2o2ZScxP8OwWlAqVKSWRCFHabk4I52QB4PT5cDjcRceEcPtiCrZ/LpEqtZOKsLESlkpx8HeZOK61NZo7UtpOeHUdsUgTOHjsdXVaMRgWiz0dycjQKQYFeJ6BNNODCh1apQqESUAoCiB6SEyO4aMFkMtMGPx/QPy/fV+W3srKSsLAwOZV2qourJP2SlJQ0aIOo0USSU5kyZUrAF7x/3UZq+a2rq0OtVsvnYyQNxKROvObmZmbNmjWiuls6nY709HTS09PlC6IkoimK4oAeN16vlz179uD1epk5c+agWsVFUeTJJ5/kL3/5C5s2bQqaHXt/NBoNxcXFbNq0ieXLlwP+a8+mTZtYuXLloN5Dqi9ecsklALKX1aZNm+SA0tPTw/bt2/nJT34C+HdBZrOZ0tJS2ahx8+bN+Hw+5s6dG9xFHmPMdZGBP38uFeRPxv79+9FqtfIW9kw8XBwOB6t/8zrdTd3+uRafCIj+6fZjnisavQZLVy+GSAOGcD2dR81oDFq5U0ypUNBa307+nAl0HO0iMi6CqKRI3A4PvV0W2ps7iE6NpHH/sTuXY4KVB744LB9HWLQBXaSOjsavZ0+UaiW5xdlUlNYFHLMxUk9cRgINlcfnTwvm59FQ205UjBFDhA61WiXvXGLSY9i/pwljmJaktGh0ejWWHgdHm7pIn5iIJtpAZ6cFr9eFRqPCYDCgUatAIaCK0+FFRKFQoFSCAkiND+OCxZNITwru4Jnb7ZaDTXt7e8AwY38RysFIv4wl+rpQDnZgr7+sfF8Dsbi4uGFzTxRFkcOHD9PS0hI0Bedg0LclvL29Xfa4iY2NlQ3WhhJcnn32WR577DH+85//MHv27GE99tdee40bbriBF154gTlz5rBq1Spef/11Dh06RGJiItdffz2pqak8+uijADz88MPMmzeP3NxczGYzjz/+OG+99RalpaUUFBQA8Nhjj/HHP/4xoE157969AW3KS5cupbW1leeff15uU541a9awtSl/Y3YwZ1LMb6htYMN/f4Ctx+5v9UXE6XDhtDhx2l2oNSrCosNwu9zYLS7CosMp+2AfBQsnUbmzhoJjmmIej5fUickc+KICl91NfHoce7cc9Nc3dCriM2M4WtFObEo0kXHhhMdH4HZ6mLIgD9EnIigUKNUqLN029AYDbpcbt9uNPkZHW1M78elRqDUaNFoVaq0KY3Q4LqeHKXOyERHwef07l7C4CA4faJbrMHBsgDIpgqTcRKx2NwnJkbS39uCwu9Dq1Kg1SiZMTcWnV2OzOfB5nRgMOowGA2q1AkGlgEgNCAJqBQiCSF56DOfOn0RiQlSwf7WAPw0gzVNIF9e2traAJgEpZ3/gwIFRcUo8HaS5kcLCwkG12EsMVKcwmUzU1NQEzNsEs24jiiKVlZW0trYya9asMSVHLwgCUVFRREVFkZeXh91up62tjdraWtxuNwaDgdra2lN63IiiyP/8z//w6KOP8v777w97cAG4+uqrMZlMPPjgg7S0tFBYWMjGjRvlIn1DQ0PANayrq4tbbrmFlpYWoqOjKS4u5osvvpCDC8DPf/5zrFYrK1aswGw2s3DhQjZu3Bjwt/DPf/6TlStXcv7558uDlk8//fSwrXNc72AqKytxuVxMmTIFn8+H1+sdcr1l75f7+PClz/E4POAT8bq92C1OPB4vGq0ahUpx7MLtonpXHTkzMjG39ZAyIQm71YlKreBwaS1Ou5vJ83Ip33YYhVLB9MUFWLptKNUKrFYLeqOeXpNfel+lVhGbHkt9+dcDkNnTMmhr6sLWp7XY34aspaU+sFMsJi0Cl0vA2sdvBSA82kDuzGwsvU7UWiWi7+sBSrvdTd6sbJwON1q9GrvNRaepl9iECLQ6NdpoPRanB6/Pg8ViISI8DINRj0IQ/EOSETp8eNGqBKZOTOS8BZNG7WLTt0ng6NGjuN1uIiIiSE9PJy4ubtTEIAeD1H01mLmRoWCz2eSdjdlsHlJq8USIosihQ4dob2+nuLh4TAWXgfD5fOzduxeHw8GMGTMCHDwFQQgQ5pR2NaIosnr1an75y1/yzjvvsGjRolFexTeLMRlgBmubXF1djcVikQPMUIPL5jc/Zse7+/B5RASfvxPLYXOBV0StV+FxebB129HoNRw51Ex6fgoKlQK1Vk1laS2Zk1M58HklYdEGMgvS8PlEFCoFeqOOhoPNmJo7iUoKw+sS6T2mahyXFoPGoMXt9KAP06HRqYlKiMBud+M/cv8OSqVRISoVchOA9LjGoMVideK0O3DYXX5rAKcPn08kPiuRuspWjOE6YhLCMUboUCj8UjZ2H/T02IlLjEClUmK1OFGpFOj0anSxYVicLrxeDzabhciICPQGPYICRKUARgXGcDWzpmdxdvHEIf42h4+6ujpqamrIz8/H6XTS1taGxWIhKiqKhISE02oSGC76qwkMV4EcCKjbSFIt0s5msPMloijKA6rD5T0TTHw+H/v27cNut1NcXBzQidjX48ZkMmG32/nPf/5DbGwsRqOR3/3ud2zYsIFzzz13FFfwzWRcB5ja2lqam5spKCgYsqDe2ufeo+qrahQocB2bL/F5fbjsLlQa/92NxWxFoVBgtzgxRupRqpR43D6qdtcxYUYmrXUm0vKSUWmUWHvsWM02NEYttXv9NsOJebH4nCLh0WFoDRr04Xq8PhGn1YXb5cZldxObHktdeRMuuxu3043X42NCYSZNtSYcNheC4Hek1GjVZE1Lp6vDivrY/6vUCnyiD5/oxer0YLe6sVv8HW/xyZFERBvRGbVY3T4sPQ7CInRotCq8Hh+CQkClUqCOMeD2idjtdhwOGzHR0Wj0GhSAUuUjNS+BecUZ5GSMvo+6RN+L9cyZMwPmMPpbAQfjTj4YxxtsF8rB4vV66erqks+J1+s95XxJX2OzWbNmjbk26f5IXjlWq5Xi4uJT7mCtVivPPfccb7zxBgcPHiQ7O5sf/OAHXH755RQXF4/5zsPxxLgMMFIx32q1UlFRQWdnJ3q9nsTExFNeSFwuF//601pMVe34RBG84PP48Pp84BPxieATfYheHwgC1i4rXq+PqIRIXHY3bUfaCY8Jw2lzk5gZh9fjxdzeg8fpxu0S6TjaRUSckai0cLQqHSqV6lgNRkNXq3+w0dptQ6FSkD8nl0Nf1QQc3+T5eRze24DP6280kH49Uxfmc2BHHQqlgDFCjzFcjz5Miz5ch9qow+n0YOt10N1lAVFEF6bCGGXAp9SgUKrQatUoFCAICgQBlGoFQrgWnyBgtfbicjmJjopCpVISFxvGlKlpzD9r4pj7sg1F+kXyL2lrazuuSSBYiseDPd7hdKEcLH3rNiaTSS6K920J9/l8lJeX093dTXFx8ZgPLqIosn//fnp7e5k1a9ag06Nr167l1ltv5W9/+xter5e3336bjRs38qc//YkVK1YM81F/exiTAeZktsn9i/mSC2HfC4lGo5GDTV/XPYvFwiu/+zfWNif48Ot5eUW8XhGfx4N0IhwWJ6JPRBeuxdzSTVRCJA67G0EAj8tLa52JnMJM3E43LocbQYTWI5143B70kVrC4g3o1HoQwWl3ER4bTu3+Rno7LYgiGCJ0pE1MpaO1+9hORIlCJRCbEoO5wyKv0+f1B5jIhCia69qx21yIPh9hkXqMEXrCYsJQ6jT+tmSnB4/HPxCq02tQGVVYHG5cLhei6EOtUaPTatHqtajVSgjX4hOhu6cbj9tBVnYyE/NTKS7OIT4+uGKQwaKv9MtQXR37NgmMlFOllLaR7qxHyqBssPTf7RmNRvn7NVrSOkNBau7o6ekZ0vl95513uOmmm/jnP/8ptwmD/4bE4/GM+XTgeGJcBZjBDE9K0unShUSlUslKt1v+8xnl71eiUegQfT4QBfk1otffeuv1+FNMCpUCp82FzqDD0mVFF6bD4/QLR+rDdHS2mjGE69GH67F02ehsNePxuolOjkSn0aFU+af+DZEGWmpN2Hod2HrsxKZGowvT43K4UWtUKFVKlColhigDNovj2Dr9Xi4KlQKVTkdXey+IoFAIqLUqtHoNhugwPF4Rj8eLoBD8P1cKKJVKdOE6HB4fKrUSBBBFv8eO0+nA5XGjjNCh1WkQcZCaGsllly0kMyv5uHM5lugr/VJUVHRG0/59nSolr5/Y2FgSEhKC1iQgzWG4XK5RdaEcLE6nk927d2Oz2RBFEZVKNWhn19FASuOZzWZmzZo16OCyceNGrr/+el5++WW++93vDvNRhhg3AeZ0JvMlqXDJNEwQBBAVmA6101HfjbmlF8En4vWJCCiO7VA8KJVKXE4Pao0Kp9WFxqDFaXXi9XoxROr9wpZeH+HRRnq6bIBIr8WCMUyP1ykiIKAN06IzaHHa3Xi8XjxOD4JSiaBUYLc45BkdtVaFNlyPrdfuT18BKECtU6M16P0qyYKASqVEoVIgIKCLMuB2eREBn09EpVaiVAiIgMqgxac4Nv8jCCiUoEDABygEF5FpMYRHKtHpegkL8w/0SQF4LBpDQaD0S7CNt6RUq/Q3EowmgbHqQnki+nZfSXMjfedtPB5PwLzNaK9HakCQ0o6D3Wlt3ryZa665hhdeeIFrr702KH/rQ/F06curr77K97//fUpKSnjrrbfkx1tbW7nvvvv44IMPMJvNLFq0iGeeeSZArqalpYV7772XDz/8kN7eXvLz8/n1r3/Nd77znTNeT7AZkwGmv6vl6U7m93/PvnbIXq+X8PBweuuttNV3Yapro7fTjlKlRPT4QKHC6/GgVAp43H6ZGJVWjdvhwuv1odFpcLu82C12bA4rYeFhqFVqfwFdENAatTgdHkBE9Pnl990uLz6vz79DEUWUKgUqvRaHzekPLHLQUYNKhc/r88uvHAsYoiiiNurwekWEY/UU8O9sEESUBh0+QYFCAFEQEUQfUdEGUjKjyJiQxLSZOXR3d7Nr1y5SUlLIzs6Wd3vt7e2o1Wo52PQfZBwtJOmXqKioEfFGOdMmgfHiQinh9XrZu3cvTqfzuO4r8P+dSiKlbW1tA9ZtRpK+3W1DaUD45JNP+O53v8vTTz/NjTfeGJS/7ddee43rr78+wNPljTfeOKmnC/i7HxcuXEhOTg4xMTFygBFFkQULFqBWq/nv//5vIiIi+POf/8zGjRspLy+X640XXXQRZrOZZ599lri4OP71r3/x0EMPsXPnToqKis54XcFkTAeYM5nMPxl97ZBbW1txu93ExcX5nfiqzBw93ELrkQ7M7b3g8uETQaFU+LXIxGMaaC4PNpsVm81GeEQEGrUaUTxmN6AU5EAjiv4uMLfTgygKCIIIgoBKq0IUBNlNUpKvV2tV+ATFsff6es2CAEq91r/bUghyQAIRRBGlXgtaBbHR4SSmRZCaEc/EqckBX0BJuTc3N/c46fH+NQppbkBSOx6NFMloS7/0bRKQ2n2lnc1AMi3jzYVSSuN5PJ5Bpx0HqttIwaZvvXM4kOZyOjo6hhRcvvjiC6688kqeeOIJbrnllqAd41A9XcB/zhctWsSPfvQjPv30U8xmsxxgKisryc/Pl+WDpPdMSkriD3/4Az/+8Y8BCAsL469//SvXXXed/L6xsbE89thj8nPGCmN2kr+/h0uwgov0Xn0ngHt7e2lra6OlpQW70k7WwlTmJBQRFxdHc+1RGmtNtNWY6G3vocvUjaPHjdVhwea0Exsfi1qlBEGJKIgoBAGfDxQK0b+rUKvxen1oDVpExGO1FAUoBX+wUoG0KkGtQqFQ4BepAeGYoZggCAgaNaIgoFL5a0f6MC1RcXqiYsNJzI4jLS2OpLQT2wA3NjZSWVnJlClTBvSi7zsl3ne3J7kQSsHmdC2Ah4ok/ZKVlUVWVtao7KbUajXJyckkJycHBOC9e/cCgU0CLpdrXLlQSkKQQ9HqAr8idkZGBhkZGXIANplMlJWVBfwNBfumRGr1bm9vH1Jw+eqrr7jqqqv4wx/+ENTgcrqeLg8//DAJCQncfPPNfPrppwE/czr9moF916ZQKNBqtXz22Wdy8FiwYAGvvfYal156KVFRUbz++us4HA7OOeecoKwtmIzJAONy+V0ipbv64fyyCoJAREQEERER5ObmYrFYaGtro66ujgMHDhATE0PW1CRmnzsNjUYju+K11LeREJaE1Wynu8uGrceG3ebEaXHgcvvwOF2IohJBJaBSKPwbDRQolMhpLNWxHQ6AVwGCwj/XotFr0Om0aA0KjOF6wqKMRESFEZUQTkJyNIknCST9kRRwGxoaAixtT4ZCoSAmJoaYmBjy8/PlAFxdXS1Lkkh38sNRvJZ2WmNJ+qW/TIvUJFBZWSlfGCIjI8nLyxvzwaVvjWgowaU/fQOwz+eT5236WiMHw4JBkqsxmUzMmjVr0Gm5srIyrrjiCh588EFuu+22oN6knI6ny2effcbf/vY3du/ePeDPJ02aREZGBr/85S954YUXMBqNPPnkkzQ2NnL06FH5ea+//jpXX321fLNnMBhYt24dubm5QVtfsBiTAeamm26ivLyckpISli9fTn5+/ojdwYaFhREWFkZOTg42m422tjaampo4ePAgEREROBwOdDodS5cvOeWXxuFwYO21+QcrHV4EpYDD5gaFArVK4ReM1CjQG1SodRqMYYag5uz7zozMnj37tBRw+wdgqSDe2NjIwYMH5YJ4QkJCUNpam5ubOXjwIFOnTh1wpzUW6Gs3kJSURGlpKWFhYXi9Xj799NMxqSQgITk7KhSKoHrlKBQKYmNjiY2NlW9K+lowREVFycFmKJIzktCmpIU22PO5d+9eSkpKuO+++7jzzjtHvZ7Y29vLddddx4svvkhcXNyAz1Gr1axdu5abb75ZNuG74IILWLp0aYA/1gMPPIDZbOajjz4iLi6Ot956i+9973t8+umnTJs2baSWNCjGZA2mq6uLt956izVr1vDRRx8xYcIEli1bxhVXXDFq6QfJt0OpVMraV9KszVi7iID/QiIVb4uKioZlpsHhcMhNE2azmfDwcDnYnM60el1dHbW1tcyYMWNIIpCjhdQwkZmZKafxJO+Stra2MaMkIOF2u2Xb4MGabwWD/ufEYDDIAfhkdZv+FgGD/ZsqLy9n6dKlrFy5kgcffHBYzrnL5cJgMPDmm28GzNLccMMNmM1m1q9fH/D83bt3HxfQ+1q4V1RUMGHCBPln3d3duFwu4uPjmTt3LrNmzeK5556jurqa3NzcgDoN+K2Pc3Nzef7554O+1jNhTAaYvnR3d/P222+zdu1aNm7cSGpqKsuXL2f58uXMmDFjRIJNe3s7+/btk6XgJW/1trY2Ojs7CQsLk4PNWJAyl2YaVCoV06dPH5G20v7nRHKvHEz788mkX8YqnZ2d7NmzhwkTJpzQq32oTQLDidTdptFomD59+qh1t7ndbjo6OmSJfYVCEVC36Xtc0t/EUIJLRUUFS5cu5eabb+Z3v/vdsAb0uXPnMmfOHJ555hnAHzAyMjJYuXLlcUV+h8Mhm4NJ3H///fT29vLUU08xceLEAdPNhw8fZtKkSbz//vtcdNFF7Nu3j+nTp1NeXs7kyZPl5y1ZsoTMzEz+53/+ZxhWevqM+QDTl97eXt577z3Wrl3Le++9R1xcHMuWLWP58uXMnj17WL6wjY2NVFRUUFBQEOAUJyF5lrS2tgZcWBMTE0fljtVqtbJr165RdXT0eDwBw659TbKio6MDzslQpF/GCgO5UJ6K/l4uklGW1DgxnBd8qQFBp9Mxffr0MVMj6lu3MZlMAXUbi8XC0aNHKS4uHnRqt6qqiqVLl/L973+fP/3pT8O+zqF6uvTnxhtvDOgiA3jjjTeIj48nIyODffv28bOf/Yzi4mLWrFkD+K830rXoiSeeIDY2lrfeeot7772Xd955RzYgGyuMqwDTF5vNxn/+8x/WrFnDO++8Q3h4OMuWLaOkpIT58+ef8RdWyv02NzczY8aMQRXHPR7PoCRrhgspZZOamkpubu6o553h+GFXIEAPrLy8/LSkX0aLE7lQDoW+bfLDpSQgIXU7jfXWaVEUZevsxsZGnE4n4eHhJCUlkZCQcMq6TV1dHRdffDHLly9n1apVI7bOZ599Vh60LCws5Omnn5bdIc855xyysrJYvXr1gK8dKMA8/fTTPP7447L18fXXX88DDzwQ8Ddx+PBhfvGLX/DZZ59hsVjIzc3lnnvuCWhbHiuM2wDTF4fDwUcffcSaNWvYsGEDGo2Gyy+/nOXLl3PWWWcNOUXk9XrZv38/FouFwsLC07qr7i9Zo1Qq5Z3NcAwxnmzGZazQ18eltbUVp9OJWq0mLy+PxMTEEWl/PhNOx4XyVPRVEjCZTPT29soF8TOt7zmdTrkBYerUqWM2uPSltraW+vp6pk6dKs/cdHZ2YjAY5FRaf/OwI0eOcPHFF7NkyRL+8pe/jIt1flv4RgSYvrjdbrZs2cKbb77J+vXr8fl8XHrppVxxxRUsXrz4lHeHUv1CoVAwY8aMoNxNDiRZEx8fT2JiYlB0nk414zLWkC58arWaqKgoTCYTNptNvosfrvbnM6GhoYGqqqohu1AOlf4F8b5SPkNJuUrnODw8fNRSpUOlrq6Ourq64/xypJSrlEpTKBQ0NzcDMGfOHK644goWLVrEiy++OOaVE75tfOMCTF88Hg+ffvopb7zxBuvXr8dms3HppZdSUlLC+eeff1xKxmKxsGvXLqKioobtSzmQZI0UbPoXOU+FKIpUV1fT2NhIYWFhUB0Sh4sTSb/0vYvv6ekZU62+kgvlzJkziYwcOaXp020ScDgclJaWynW4sZAqPRX19fXU1NRQXFx80iYP6fvzt7/9jeeff56Ojg6SkpJ46KGHWLZsWdB2liGCwzc6wPTF6/XyxRdfsGbNGtatW4fZbJZzthdeeCEffvgh69at4/7772fChAkj8qU8kWRNYmIicXFxJw02fYvjRUVFpzXjMtIMVvploFbfvnfxI0XfNtmZM2cOqwvlqRhsk4Ddbqe0tJTo6GgKCgrGRXBpaGigurp6SAG8vb2dpUuXkp2dzdy5c3nnnXcoLS1l5cqVrFq1angPOMSg+dYEmL74fD6++uorOdi0tLTg8Xi48cYb+e1vfzsqFxJJVFDa2djtdmJjY+Vg07eOJM24uFwuioqKxpzPyECcrvSLZP8r3cXrdDo52Axn44QkTWIymcZcd9uJmgSioqJoaGggPj6eSZMmjYvgcuTIEaqqqoYUXDo7O7n00kvJycnh9ddfl78bzc3NdHR0jLlhw28z38oAI+Hz+bj//vt59tlnKSkpYceOHdTX13PBBRdQUlLCJZdcclxBcaSQJGskCXlJniUyMpIDBw7Iar1jvTAOwZN+6W8sJzVOSOrPwUppSq6OZrN51F0oT4XUJNDU1MSRI0cQRTFAXWEsH7tUO5w5c+ag07tms5nLL7+c5ORk1q5dO+ZqdSECGfuVv2Hk97//Pa+++irbt2/n//7v/zh48CBfffUVM2fO5KmnniI7O5urrrqKv//973R0dDCSsViSq5k3bx4LFiwgJiaGI0eO8OWXX+J0OomLizuh6+dYorm5mX379jFlypQz1hVTKpUkJiYybdo0Fi9eTEFBgewa+cknn3DgwAHZd/50kd6vt7eX2bNnj+kLNHwtAtva2kp6ejoLFy4kKSmJ9vZ2Pv/8c7Zt20Z1dTU9PT0j+vd7KpqamqisrKSoqGjQwaWnp4crr7ySuLg43nzzzaAFl+eee46srCx0Oh1z587lq6++GtTrXn31VQRBCJjkB38r+4033khKSgoGg4GLL76Yw4cPH/f6bdu2cd5552E0GomIiGDRokXY7fZgLGnM8K3ewUi57IG8G6QUyZo1a1i7di379u3j7LPPZvny5Vx++eXEx8eP6M7GbDaze/duEhMTMRgMmEwmzGYzERERcvvzWLsYjpT0i5Qyam1txWQy4XK5AuZKBtumLsnXu91uioqKxsXdscViobS0lJSUlONmn/qqHUt+P1L780grCfRFavceSkeexWLhyiuvRKPR8O677wbtb320PF22bdvGxRdfzC9/+Usuv/xyVCoVe/bsoaSkZFykvAfLtzrADBZJkfjNN99k3bp1lJaWMn/+fJYvX86yZctITk4e1mDT1tbG/v37ycvLIz09XX7c5XLJabSxJFnTd0h1pKVfpIE96bxYrdYA9ecTfXklEUiAoqKicZF6lIJLamrqKRtTRlNJoC9Hjx7l4MGDQwouNpuNq666ClEUeffdd4Pa6DFani7z5s3jwgsv5JFHHgnaWsYioQAzRERRpKGhgbVr17J27Vq2bdvGnDlzKCkpoaSkhPT09KAGmyNHjnD48GGmTp160jsqSbJGKoaPlmTNWJN+kRSx29ra6OnpITIy8rj6hMvlYteuXePGhRL8HXmlpaWkp6cHiCQOhhM1CUiDjMO1c2tpaaG8vJwZM2YQGzs4ywmHw8HVV1+N1Wpl48aNQb1ZGapgpcRDDz3E3r17Wbdu3XHT+JJWWFVVVcDvJT09nfPPP5/Vq1fT1tZGYmIiTz/9NP/+97+prq5m0qRJ/P73v2fhwoVBW99YYOzfpo0xBEEgMzOTO++8kzvuuIPm5mbWrVvHmjVruP/++yksLJSDTU5Ozmlf2PsLQJ4qT61Wq0lJSSElJSVAsmbHjh1oNBo52Axn55XX62Xfvn3Y7XZmz549JqRfDAaD3LnWt/358OHDhIWFERMTQ1tbG2FhYWNKp+tk9PT0UFZWJouvDpX+hntWq1WWaDl48KAchIcqrX8yWltbKS8vZ/r06YMOLk6nkx/+8Id0d3fzwQcfBH0nPFqeLjU1NQD85je/4YknnqCwsJC///3vnH/++XKm4ptCKMCcAYIgkJqaysqVK7nttttoa2uTbQYefvhhCgoKZE+boVj+Sl1MXV1dzJ49e8i7AJVKRVJSEklJSQGSNZLr4HBI1rjdbvlLN2vWrBFRcB4qOp2O9PR00tPTcbvdNDc3U1VVJcumV1dXj5hu3OnS3d1NWVkZ2dnZZGVlnfH7CYIgeyBlZ2fLQdhkMnH48GFZSSA+Pv6UqtgnQtJvmz59+gm9UPrjcrm44YYbOHr0KJs2bRoTQ8TB8nSR/t5uvfVWbrrpJsCflt20aRMvvfTSCcUxxyOhABMkBEEgMTGRW2+9lRUrVtDZ2cn69etZs2YNjz32GLm5ubKnzeTJk094p+zxeORC85w5c8644Ne3lbevZM2ePXuCJlkjedFLar3jIcXkcrloaGggJSWFvLw8+bz0tf6V1J/Hyq7GbDaza9cucnJyyMzMHJbP6B+EpSaB+vr602oSkOqH06dPH/SUvcfj4cc//jG1tbVs3rx52BpEpGHm1tbWgMdbW1sHFDKtrq6mrq6Oyy+/XH5MChYqlUr2dCkuLmb37t0DeroAsip7QUFBwPtPnjyZhoaGoK5xtAnVYEYAs9kse9r85z//IS0tjZKSEq644oqAtExrays1NTVotVqmT58+rIXmYEnWnEj6ZSwjKQoM1HklSchL58Xn88kX1ZEshvdHGlTNzc0NaPQYKfrenJhMpkGdF5PJxN69e5k2bdpJ64d98Xg83HrrrezZs4fNmzeftmL1YBkNTxdRFElLS+NHP/pRQJG/qKiIpUuX8oc//GEYVjo6hALMCNPb28u7777L2rVref/994mLi6OkpITp06fzy1/+knvuuYef/OQnI3qh7l/0dblcsmSN5Pt9orUMRvplLCGlmLKysk5ZvxioGC51Xg2l/flM6ezsZPfu3Wc8qBospPMi1bMcDkdAp55Go6G9vZ09e/YMyfra6/WycuVKtm3bxtatW0lJSRnmlYyOpwvAqlWreOihh/jb3/5GYWEhr7zyCk888QT79+8fctPGWCaUIhthwsPDueaaa7jmmmuw2Wxs3LiR559/nmeffZbp06dTV1fHtm3bmDdv3ojdLfcv+kqSNdXV1ezfvz9A5Vi6qJ6u9MtoIl2oB2tp0P+8SO3PdXV1HDhwYFDtz2dKR0cHe/bsGZK52XDT97zk5uYe1yRgNBqxWq1MnDhx0MHF5/Nx55138tlnn7Fly5YRCS4AV199NSaTiQcffFD2dNm4caN83A0NDUO+2Tt69Ch33XXXcZ4ufbnjjjtwOBzceeeddHZ2MmPGDD788MNvVHCB0A5m1Fm3bh3XXXcdjz76KFlZWbKnjVar5fLLL+eKK67grLPOGrW5jIEka/R6Pc3NzeTn54+JO+rBIMnVTJo0KSgXL6n92WQy0d3dTWRkpJwyClbnleScGaxjHgmOHj3KgQMH5CBjNBrl83KiJgGfz8fPf/5z3n33XbZu3XpanXEhxiahADOKuN1uzjrrLO6//36WLVsmP+5yuQI8bQDZ02bRokWjNmFus9k4fPiw7EwZHR0tNxCMhZbkE9Ha2sqBAweGzS/H6XTK6aLOzs7T9nDpi1S/OJFV91hE2iFKAXEwSgI+n49f//rXrFmzhi1btnyjWnRDhALMqOPz+U66Bfd4PHzyySeyp43D4eDSSy9l+fLlnHvuuSN6Ye8r/WIwGOSdzViWrGlqaqKioiKoLpQno6+HS3t7O1qtVg42gxVObWtrY9++fUOqX4w2XV1d7Nq164S7rf5NAo8//jg6nY7IyEg2b97Mxx9/zKRJk0bhyEMMJ6EAM47wer18/vnnss1Ad3c3S5cuZfny5VxwwQVBS83051TSLwNJ1kjBZjQn+UfKhfJE9LfNVigUcrA5UfuzNDMylM6r0cZsNlNWVjboOpEoimzevJlnn32Wzz//HI/Hw5IlS1i+fDmXXXZZyDTsG0QowIxTfD4f27dvl4NNa2srS5YsoaSkhIsvvjhoek1DlX4ZK5I1o+VCeSKk9mfp3Eht4fHx8fI8RktLCwcOHBjSzMhoI83m5OXlDboeJ4oif/7zn1m1ahWbNm1Cq9Xy1ltvsX79eq644gruu+++YT7qECNFKMB8A/D5fOzatYs333yTtWvXcuTIkQBPm9OdTPd6vezduxeHw0FRUdGQ03F9JWva29tHRLJmLLlQnghRFOnp6ZF3fQ6HA6PRiMViYerUqcM++xEspJbvoczmiKLIM888w5/+9Cc++OADefiw78/HQ0diiMEx9qfiQpwShUJBcXExjz76KIcOHWL79u0UFhby5JNPkpWVJXvadHZ2DtoTxO12U1ZWhsfjYdasWadV65Eka6ZPn87ixYuZOHEiLpeLsrIyPv30Uw4dOjSkYzoVoihy6NAhWlpamDVr1pgMLuBv842MjCQvL48FCxaQnZ2NxWJBp9Oxf/9+SktLOXLkCA6HY7QP9YRIemgTJkwYUnB54YUX+OMf/8h77713XHABTiu4jJafi7SmpUuXIghCwCxMCD+hHcw3GOmCK9kM7N+/n0WLFsmeNnFxcQN+oYdb+qVvwbetrS0okjXjyYWyL5Kro1Qnstvt8nnp7u6WmyeC2f58pvT09FBaWjokyRpRFHn55Zf51a9+xbvvvsvZZ58dlGMZLT8XiSeffJIPP/yQ999/n3Xr1h0XrL7thALMtwRRFKmurpaDTVlZGQsWLJA9bZKSkhAEgfLycurr60lPTx8R6ZcTSdYMRZpFcqG02WzMnDlz3Bg2SVYMRUVFREdHH/fz4Wh/PlMkmwBpwHYwiKLIP/7xD+655x42bNjAueeeG7TjGS0/F4Ddu3dz2WWXsXPnTpKTk0MBZgBCKbJvCYIgkJubyy9+8Qu+/PJLDh8+TElJCWvWrCE/P5+LLrqIhx56iAsvvJCdO3cyZcqUEZGrUSgUxMTEMGnSJM4++2zZSbKyspKPP/6YvXv30traekJ7aK/Xy+7du3E4HMyaNWvcBBepw23mzJkDBhcArVZLWloaM2fOZPHixWRlZWG1WtmxYweff/45FRUVdHV1jZgVshRcMjMzhxRcXn/9de6++27WrFkT1ODicrkoLS3lggsukB9TKBRccMEFbNu27YSve/jhh0lISODmm28+7mdOpxMgICWsUCjQarV89tln8mM2m41rr72W5557btzUzEaDkFTMtxBBEMjKyuKuu+7izjvvpLm5mccee4xVq1aRk5PDRx99RFhYGCUlJWRnZ4/YnfJQJWv6ulAWFxePCxdK+HqeaCgdbmq1muTkZJKTk2V3yv6q2AkJCcTExAzLjYHknpmRkTGkSft169Zx++2389prr3HRRRcF9ZhGy88F4M4772TBggWUlJQEbT3fRMbHNzLEsCEIAjt37uSll17ir3/9K5dddpnsafPb3/6WKVOmyAZqIyloKQgCERERREREkJubK0vWNDQ0UF5eTlRUFHa7Hb1eT1FR0biwCAB/+3R9fT3FxcWnbaAl2QnEx8cHpBjLy8vxer0BgpzBOC9Wq1V2z8zJyRn0695++21uvfVW/vnPf3LppZee8XGcKcHyc9mwYQObN2+Wb25CnJhQDSYEd955J4sXLw7IH4uiSGdnJ2+99RZr167lo48+Ii8vT7YZmDx58qi1k5rNZnbv3o0oini9XqKiosaFZE11dTVHjhyhuLh4WDrcBmp/HkiodChYrVZ27txJamoqEyZMGPTv/P333+eGG27g5Zdf5rvf/e6QP3cwDNXyePfu3cfdjEh+LgqFQvZzkRjIz+W5557jjjvu4Omnnw7YKXq9XhQKBWeffTZbt24dlvWOR8ZVgHnuued4/PHHaWlpYcaMGTzzzDPMmTPnlK979dVX+f73v09JSUlAK6Eoijz00EO8+OKLmM1mzjrrLP7617+G9JD6Icmzb9iwQfa0ycjIkIPNtGnTRsxewG63U1paKvvP9FURGKuSNVKDRWNj47AFl4HoL1QqacfFx8cPKhDbbDa5gN3fN+dkbNq0ie9///v8z//8D9dee+2ZLuOkjIafS0tLC+3t7QHPmTZtGk899RSXX355SKyzD+MmwAS7HRHgscce49FHH+WVV14hOzubBx54gH379lFeXj6m74RHm56engBPm4SEBNkauri4eNiCjZSqSUhIID8//7gL3liUrOk7+FlcXBw0hYWhYrfb5Y40s9lMeHi4vOsb6NxIwSUpKYm8vLxBB5dPPvmE7373uzzzzDPccMMNw77LHS0/l/4IghDqIhuAcRNggt2OKIoiKSkp3H333dxzzz2Af0ucmJjI6tWrueaaa0ZkXeMdq9XKxo0bWbNmDe+++y5RUVEsW7aMkpIS5s6dG7TayMlcKAdiLEjWiKJIZWUlra2tFBcXj6ouW19cLlfAuTEYDHKwCQ8Px+FwsHPnThISEoZUd/v888/5zne+wxNPPMEtt9wyYinUZ599Vs5sFBYW8vTTTzN37lwAzjnnHLKysli9evWArx0owDz99NM8/vjjx/m5nEzFPBRgBmZcBJih5lolHnroIfbu3cu6deuO+0OqqalhwoQJ7Nq1i8LCQvk1ixcvprCwkKeeemoYV/TNxG638+GHH7JmzRrefvttdDqd7GmzYMGC0+7ykvSuBuNCORCjJVlTUVGByWSiuLh4zAxJ9qf/uVGpVHg8HmJiYgLsvE/FV199RUlJCb///e+57bbbQnIvIYBx0kU2HO2ILS0t8nv0f0/pZyGGhl6vZ9myZSxbtgyXy8XmzZt58803ue666xAEgcsuu4wrrriCs88+e9CeNpLHSF5e3ml70UuSNUlJSQEKx2VlZSiVygCF42BcGCUFhfb2dmbNmjVmakED0ffc2Gw2duzYgVarxWw288knn8g1m9jY2BMGm7KyMq644gp+85vfhIJLiADGRYAZKoNpRwwxvGg0Gi6++GIuvvhinn/+eT7++GPeeOMNVqxYgdPp5LLLLqOkpITzzjvvhMORwXahBAICSl/Jmr179wZlnkQURVl9eqwHl744HA7KysqIj49n8uTJiKKI2WzGZDJx6NAh3G43cXFxJCYmEhsbK+9G9+7dy7Jly7jvvvu44447QsElRADfyBTZYNoRBUEIpchGAa/Xy2effSbbDPT29gZ42kgX5A8//BBBEJg2bdqImG6JokhXV9cZSdaIosiBAwfo7u6muLh43DSKOJ1Odu7cKXfm9Q8SoijKQ69tbW1s2bKFjz/+mIULF/K///u//PSnP+WBBx4IBZcQxzEuAgwEvx1RrVaTkpLCPffcw9133w34u6MSEhJCRf4Rwufz8eWXX8rBxmQysWTJEmJjY/n73//Oe++9JxdrRxKpLVu6oLpcroDhxYFqST6fjwMHDtDb28vMmTPHVXApLS0lIiKCKVOmDCpIHDp0iL/+9a+8++67tLa2smjRIq688kqWL19+2mnMEN9Mxk2AGY52xMcee4w//vGPAW3Ke/fuDbUpjwI+n4+ysjLuu+8+PvnkE7Kzs5k8eTIlJSUsXbp01EzDRFHEYrHQ2tpKW1sbdrv9uOFFn8/H/v37sVgsFBcXjxs9NJfLxc6dOwkPD2fq1KmD3oFUVVWxdOlSrr32WlauXMnbb7/N2rVrqaqqor6+/lu3kznnnHMoLCxk1apVo30oY45xU4O5+uqrMZlMPPjgg3I74saNG+X0SUNDw5Bz5j//+c+xWq2sWLECs9nMwoUL2bhxYyi4jAKCIPD++++ze/duPvvsM3Q6HW+++SZ//vOf+a//+i/OO+88SkpKuPTSS4NWjB/scYWHhxMeHj6gZE1MTAwulwufz8esWbMG3bww2khCkWFhYYPeuYB/ruyyyy7jO9/5Do899hgKhYKVK1eycuVKnE7nty64hDg540pNeeXKldTX1+N0Otm+fXtA+mTr1q0n7HUHWL169XGGQIIg8PDDD9PS0oLD4eCjjz5i4sSJJz2GYJobud1u7rvvPqZNm4bRaCQlJYXrr7+e5ubmQb3nNwm73c6nn37K1q1bmTt3LjNmzOCRRx5h3759lJWVMW/ePJ5//nlycnJYvnw5L7/8MiaTacSUhCXCwsLIyclh3rx5zJ8/H4fDgdVqxWazsXfvXhoaGsa0URh8HVyMRiNTp04d9I3ZkSNHuOSSS7jkkktYtWrVca873Z3bSBuGdXZ2cvvtt5Ofn49erycjI4Of/vSndHd3n9bxhzgx4yrAjDavvfYad911Fw899BBlZWXMmDGDJUuW0NbWdtLX1dXVcc899xxnsmSz2SgrK+OBBx6grKyMtWvXUlFRwbJly4ZzGWMSg8HABx98wLRp0wIeFwSBgoICHnzwQcrKyjhw4ADnnXcer7zyCrm5uVx66aW88MILHD16dESDjdfrpbKyEqVSyaJFi1i4cCEJCQm0tbXx2Wef8dVXX1FXV4fdbh+xYxoMklOpwWAYUnA5evQol156Keeffz7PPfdc0NQagv2dEkWR5cuXU1NTw/r169m1axeZmZlccMEFWK1WAJqbm2lubuaJJ55g//79rF69mo0bNw4o3z9YPB4PK1euJDIykri4OB544IERv/kZi4ybGsxYINhqAgOxY8cO5syZQ319PRkZGcOxjG8EoihSX1/PmjVrWLt2Ldu3b2fevHmyikBaWtqwpWu8Xi979uzB4/FQVFR0nIjkiSRrJKOw0UIKLlqtdkhDlC0tLSxdupS5c+fy8ssvB1W5ejQNw/ryxhtv8MMf/hCr1TrkgeBzzjmH0tJSbr75Zn7yk5+wc+dOVqxYwapVq7jllluG9F7fNEI7mEEyHOZGA9Hd3S37ooQ4MZKnzd13381nn31GXV0d3/ve93j33XeZOnUq5513HqtWraK2tjaod5KSwZnX62XmzJkDKhRrNJoAo7CMjAx6enrYvn07X3zxBVVVVfT29o7oHa4UXDQazZCCi8lk4vLLL6eoqIiXXnopqMFlNA3D+iPZU5+u2kR6ejpPPvkk+fn5/OAHP+D222/nySefPK33+iYRCjCD5GRqAiea/JfUBF588cVBfYbD4eC+++7j+9///ml7hXwbEQSBtLQ0fvrTn7J161aOHDnCjTfeyObNmykqKuLss8/m8ccfp7Ky8owu6h6Ph7KyMkRRpKioaFAXI6kdvrCwkMWLF5OTkyNPzH/++edUVlZiNpuHNdhIxmxqtZoZM2YMOrh0dHRw+eWXk5+fz//93/8F3dBtOL5TfQ3Durq6cLlcPPbYY8cZhvU/jkceeYQVK1ac9lrmzZsXsGOeP38+hw8fxuv1nvZ7fhMIBZhhYqhqAm63m+9973uIoshf//rXETjCbyaCIJCUlMRPfvITPvzwQ5qbm7ntttvkppB58+bxhz/8gfLy8iFd1KWLtEKhGHRw6Y8kyzJ9+nQWL17MxIkTcblc7Nq1i08//ZRDhw7R2dkZ1GAjHbdKpRpScDGbzZSUlJCZmcmrr756Wl4ywWYohmGVlZXExMRgMBjYsmULS5cuHXDtPT09XHrppRQUFPCb3/xmmFfw7WPctCmPNpI7YGtra8Djra2tA3pyV1dXU1dXx+WXXy4/JqkJqFSqAHMjKbjU19ezefPm0O4lSAiCQFxcHDfffDM/+tGPMJvNsqfNf//3f5OZmSnbDJzM08btdgdcpIORJjqZZA0g/+xMLJC9Xq8cFIdy3D09PVxxxRUkJCTwxhtvDFvr9XB9p4qLi9m9e/eAhmF96e3t5eKLLyY8PJx169adURDdvn17wP9/+eWX5OXljRun1eEitIMZJBqNhuLiYjZt2iQ/5vP52LRpE/Pnzz/u+ZMmTWLfvn3s3r1b/rds2TLOPfdcdu/eLU88S8Hl8OHDfPTRR8TGxo7Ymr5NCIJAdHQ0N9xwA+vXr6e1tZUHH3yQ6upqLrzwQmbMmMGvf/1rdu7cKV+0wJ8m2rFjh5xeGo4LhkKhIC4ujoKCAhYvXizXSMrLy/n444/Zv3+/LF8zWKTgIggChYWFgz5ui8XCVVddRVhYGOvWrRvWmbDh+k5JREZGEh8fz+HDh9m5cyclJSXyz3p6erjooovQaDRs2LDhjNfZ0NDAXXfdRUVFBf/+97955pln+NnPfnZG7/lNILSDGQJ33XUXN9xwA7NmzZLVBKxWKzfddBNAgJqATqdj6tSpAa+XCvfS4263m6uuuoqysjLeeecdvF6vnHuOiYkZN0N745GIiAiuvfZarr32WqxWK++//z5r1qzhsssuIzo6Wr5w3XfffVx22WU88sgjI+LaKQgCMTExxMTEkJ+fL0vWVFZWDkqyBr5uRBBFkZkzZw46uNhsNr773e+iVCpZv379iAh1Bvs7BQMbhi1fvpyLLroI+Dq42Gw2/vGPf9DT00NPTw8A8fHxp3UTcf3112O325kzZw5KpZKf/exnZ1TT+aYQCjBDINhqAk1NTWzYsAEgQHATYMuWLZxzzjnBOvQQJ8FoNHLVVVdx1VVXYbfb+eCDD/jnP//JtddeS0ZGBlarlc8+++yMPG1OB6mbMCoqiry8PFmypqamhgMHDhwnWQNft1D7fL7jBF9Pht1u55prrsHtdrNx48YRa6ceDoWOo0ePctdddx1nGCZRVlYmp7Ryc3MDXltbW0tWVtaQPm/r1q3yf4fqp4GE5mBChOhHa2sr559/PpMnT+a6665j/fr1bNiwAYVCEeBpM5qFb0mypq2tDYvFQkxMDPHx8XIqbebMmYMOhk6nk2uvvZaOjg4++OCDUIt8iKARCjAhQvTjxz/+MXa7nVdeeUW+SLvdbj7++GPefPNN3nrrLdxut+xpc+65546qwKXNZqO1tZW6ujo8Hg+RkZEkJSWRkJBwytqCy+Xiuuuuo7GxkU2bNhETEzNCRx3i20CoyD8OCbZ2U1/+3//7fwiC8K1Whn3qqaf4+9//HrADUKvVXHDBBTz//PM0NTWxZs0aIiIi+OlPf0p2djY//vGPefvtt0dFGkan09Hd3Y1er2f+/PkkJSUdJ1ljs9mOe53b7ebmm2+mrq6ODz/8MBRcQgSd0A5mnPHaa69x/fXX8/zzzzN37lxWrVrFG2+8QUVFBQkJCSd8XV1dHQsXLiQnJ4eYmJgB5WrWrVvHb3/7W0wmE/feey933HHH8C3kG4LX6w3wtGlvb2fJkiUsX76cJUuWYDQah/XzfT4f+/btw263U1xcHJC2G0iyxul0Eh4ezvTp07n11lvZs2cPW7ZsGRFTtxDfPkIBZpwxXHpoTU1NzJ07l//85z9ceuml3HHHHaEAM0R8Ph+lpaW8+eabrFu3jqamJi644AKWL1/O0qVLgz7fJPnQWK1WiouLT9p16Ha7MZlMvPjii6xatYrY2FicTif//ve/ufDCC0My+yGGhVCKbBwxXHpoPp+P6667jnvvvVcWCAwxdBQKBbNnz+axxx7j0KFDfP7550ybNo0nnniCrKwsvvvd7/KPf/yDrq6uM57Wlxw0BxNc4GvJmgceeIBrr72WqKgozjrrLL7zne+Qm5vLr371q5D6b4igEwow44jh0kN77LHHUKlU/PSnPw3q8X6bUSgUFBYW8rvf/Y79+/dTWlrKnDlz+Mtf/kJ2djZXXHEFq1evpr29fcgXdlEUKS8vp7e3d1DBRcLn83HvvfeyZcsW3n33XTZs2IDJZOLJJ58cURO3EN8eQnMw32AGo91UWlrKU089RVlZWegCM0wIgsCUKVOYMmUKDz74IIcPH+bNN9/k5Zdf5mc/+xkLFy6kpKSEZcuWkZiYeNLfgyiKHDhwgJ6eniEHl1//+te8/fbbbN26lezsbMDfIPBt9B8KMTKEdjDjiDPRblKpVKhUKv7+97+zYcMGVCoV1dXVfPrpp7S1tZGRkSE/p76+nrvvvnvIA2chTo0gCEycOJFf/epXfPXVV1RWVnLppZfy+uuvM3HiRC6++GKee+45Ghsbj9vZSDuX7u5uiouLB90aLYoiDz/8MK+//jofffTRccOFZ8pIO1KCX3n8tttuIzY2lrCwML7zne8c970IMQYQQ4wr5syZI65cuVL+f6/XK6ampoqPPvrocc+12+3ivn37Av6VlJSI5513nrhv3z7R6XSK7e3txz0nJSVFvO+++8RDhw6N5NK+1fh8PrGhoUFctWqVuGjRIlGlUolz5swR//CHP4gHDhwQu7u7xWuvvVZ8+umnxY6ODtFqtQ7qn8ViEX/961+L8fHx4v79+4N+3K+++qqo0WjEl156STxw4IB4yy23iFFRUWJra+tJX1dbWyumpqaKZ599tlhSUhJwHubNmyeeffbZ4ldffSUeOnRIXLFihZiRkSFaLBb5ef/v//0/MT09Xdy0aZO4c+dOcd68eeKCBQuCvr4QZ0YowIwzXn31VVGr1YqrV68Wy8vLxRUrVohRUVFiS0uLKIqieN1114m/+MUvTvj6G264IeALPRCZmZnik08+GcSjDjEUfD6f2NzcLD733HPieeedJyqVSnHixIliQkKC+M4774gWi2XQweW3v/2tGBsbK+7evXtYjnXOnDnibbfdJv+/1+sVU1JSBrzhkfB4POKCBQvE//3f/z3u77GiokIEAoKh1+sV4+PjxRdffFEURVE0m82iWq0W33jjDfk5Bw8eFAFx27ZtQVxdiDMllCIbZ1x99dU88cQTPPjggxQWFrJ79+7jtJtOZKwUYnwgCALJycn813/9Fx9++CE33XQTHR0dFBYWcuWVVzJ//nweffTRk3raiKLI008/zapVq9i4cSMzZswI+nGOliNlaWkpbrc74HMlo7GTfW6IkSdU5B+HrFy5kpUrVw74s77CewOxevXqU75/XV3d0A8qxLDwi1/8gg8++ICdO3eSmZmJ2Wxm/fr1rF27Vm5/ljxtpk6dikKhQBRFnn/+eR577DE2btx4nA9KsDhZV+OhQ4cGfI3U1bh79+4Bf97XkfKFF17AaDTy5JNPBjhStrS0oNFojtNMO1k3ZYjRIbSDCXHaDIdkzcGDB1m2bBmRkZEYjUZmz55NQ0NDrCN53QAACxJJREFUkI98/LBo0SI2b95MVlaW7Glz4403smHDBlpaWrj//vs5fPgw559/PoWFhdx///385je/4be//S3vvPMO8+bNG+0lyAyHI2WIMc4op+hCjFOCXdwVRVGsqqoSY2JixHvvvVcsKysTq6qqxPXr15/yPUOIYm9vr/j666+Ly5cvFwVBENetWzfsn+l0OkWlUnncZ11//fXismXLjnv+rl27REBUKpXyP0EQREEQRKVSKVZVVQU832w2i21tbaIo+ms9//Vf/yWKoihu2rRJBMSurq6A52dkZIh//vOfg7fAEGdMKMCEOC2CXdwVRVG8+uqrxR/+8IfDdcjfGrq7u0fss4Ld1TgQlZWVokKhEP/zn/+Iovh1kf/NN9+Un3Po0KFQkX8MEtpzhhgyw1Hc9fl8vPvuu0ycOJElS5aQkJDA3LlzBxTlDHFygq15djLuuusuXnzxRV555RUOHjzIT37yk+McKX/5y18CyI6Uff9FRUURHh7O1KlT5aHRN954g61bt1JTU8P69eu58MILAxwpIyMjufnmm7nrrrvYsmULpaWl3HTTTcyfP39MpQRDhIr8IU6D4SjuSsZZf/zjH/nd734nF6ivvPJKtmzZwuLFi4O9jBBBYDQcKQGefPJJFAoF3/nOd3A6nSxZsoS//OUvQVtXiOAQUlMOMWSam5tJTU3liy++YP78+fLjP//5z/n4449lO1qJ3t5epk+fzl/+8heWLl0KwI033hig6iy95/e//33+9a9/ya9dtmwZRqORf//738O/sBAhQgSV0A4mxJA5E8kaCZ/PB4BKpaKiooL09HRUKhUFBQUBr508ebI8/xAiRIjxRagGE2LIaDQaiouL2bRpk/yYz+dj06ZNATsaiUmTJrFv3z52794t/1u2bBnnnnsuu3fvJj09HY1Gw+zZs6moqAh4bWVlJZmZmcO+phAhQgSf0A4mxGlx1113ccMNNzBr1izmzJnDqlWrjivupqam8uijj8rF3b5IQ3J9H7/33nu5+uqrWbRoEeeeey4bN26U1X9DhAgx/ggFmBCnxXAUd6+44gqef/55Hn30UX7605+Sn5/PmjVrWLhw4XAsIUSIEMNMqMgfIkSIECGGhVANJsS4J9iSNRaLhZUrV5KWloZer6egoIDnn39+GI48RIhvNqEAM4555513iIqKwuv1ArB7924EQeAXv/iF/Jwf//jH/PCHPxytQxx2XnvtNe666y4eeughysrKmDFjBkuWLKGtre2kr6urq+Oee+7h7LPPPu5nd911Fxs3buQf//gHBw8e5I477mDlypVs2LBhuJYxaoxGcG5paeG6664jKSkJo9HIzJkzWbNmTbCWFGIsMbpCAiHOBLPZLCoUCnHHjh2iKIriqlWrxLi4OHHu3Lnyc3Jzc2UfjW8iwyFZM2XKFPHhhx8OeGzmzJnir3/966Ae+2gzHHpyt9xyizhhwgRxy5YtYm1trfjCCy+ISqVSXL9+vfycCy+8UJw9e7a4fft2sbq6WnzkkUdEhUIhlpWVDccyQ4wioR3MOCYyMpLCwkK5y2rr1q3ceeed7Nq1C4vFQlNTE1VVVd/YKfjhkKwBWLBgARs2bKCpqQlRFNmyZQuVlZWyVMk3hT//+c/ccsst3HTTTfJOw2Aw8NJLL53wNV6vlx/84Af89re/JScn57iff/HFF9xwww2cc845ZGVlsWLFCmbMmBGwM/riiy+4/fbbmTNnDjk5Odx///1ERUVRWlo6LOsMMXqEAsw4Z/HixWzduhVRFPn000+58sor5eHEjz/+mJSUFPLy8kb7MIeFk0nWnMgXRJKsefHFF0/4vs888wwFBQWkpaWh0Wi4+OKLee6551i0aFFQj380Gc3gvGDBAl577TU6Ozvx+Xy8+uqrOBwOzjnnnKCtL8TYINSmPM4555xzeOmll9izZw9qtZpJkyZxzjnnsHXrVrq6ur6xu5fTYTB+JOAPMF9++SUbNmwgMzOTTz75hNtuu42UlJSAC/J4Zjj05MB/7lasWEFaWhoqlQqFQsGLL74YEJxff/11rr76amJjY1GpVBgMBtatW0dubm5Q1hZi7BAKMOOcs88+m97eXp588kk5mJxzzjn88Y9/pKuri7vvvnuUj3D4GA7JmpSUFH71q1+xbt06Lr30UgCmT5/O7t27eeKJJ74xAWaoBDM4P/DAA5jNZj766CPi4uJ46623+N73vsenn37KtGnTRmpJIUaCUa4BhQgChYWFolKpFP/617+KoiiKHR0dolqtFgHx0KFDo3x0w0uw/Ui6u7tFQHzvvfcCXrtixQrxwgsvHPb1jBTDYRZms9lEtVotvvPOOwGvvfnmm8UlS5aIoug3lQPE/fv3Bzzn/PPPF2+99dbgLjLEqBPawXwDWLx4Mbt375Zz2DExMRQUFNDa2kp+fv7oHtwwE2zJGo1Gw+LFi7n33nvR6/VkZmby8ccf8/e//50///nPI7q24aSvnpzUaizpya1cufK450t6cn25//776e3t5amnniI9PR2Hw4Hb7T5OwUGpVMo7RZvNBnDS54T4BjHaES5EiDPlmWeeETMyMkSNRiPOmTNH/PLLL+WfLV68WLzhhhtO+NqB2pSPHj0q3njjjWJKSoqo0+nE/Px88b//+79Fn883TCsYHV599VVRq9WKq1evFsvLy8UVK1aIUVFRYktLiyiKonjdddeJv/jFL074+oHO3eLFi8UpU6aIW7ZsEWtqasSXX35Z1Ol04l/+8hdRFEXR5XKJubm54tlnny1u375drKqqEp944glREATx3XffHba1hhgdQgEmRIhvMaMRnCsrK8Urr7xSTEhIEA0Ggzh9+nTx73//e7CXFmIMENIiCxEiRIgQw0JoDiZEiCAyFOmV1atXIwhCwD+dThfwHFEUefDBB0lOTkav13PBBRdw+PDh4V5GiBBBIRRgQoQIEqejixYREcHRo0flf/X19QE//9Of/sTTTz/N888/z/bt2zEajSxZsgSHwzHcywkR4owJpchChAgSc+fOZfbs2Tz77LOAvysrPT2d22+/PUCAVGL16tXccccdmM3mAd9PFEVSUlK4++67ueeeewDo7u4mMTGR1atXc8011wzbWkKECAahHUyIEEHgdKVXLBYLmZmZpKenU1JSwoEDB+Sf1dbW0tLSEvCekZGRzJ0796TvGSLEWCEUYEKECAKno4uWn5/PSy+9xPr16/nHP/6Bz+djwYIFNDY2AsivG8p7hggxlggNWoYIMUrMnz+f+fPny/+/YMECJk+ezAsvvMAjjzwyikcWIkRwCO1gQoQIAkPVRRsItVpNUVERVVVVAPLrzuQ9Q4QYTUIBJkSIINBXekVCkl7pu0s5GV6vl3379pGcnAxAdnY2SUlJAe/Z09PD9u3bB/2eIUKMJqEUWYgQQWIoumjg91aZN28eubm5mM1mHn/8cerr6/nxj38MgCAI3HHHHfzud78jLy+P7OxsHnjgAVJSUo6zKg4RYiwSCjAhQgSJq6++GpPJxIMPPkhLSwuFhYVs3LhRLtI3NDQEiDx2dXVxyy230NLSQnR0NMXFxXzxxRcUFBTIz/n5z3+O1WplxYoVmM1mFi5cyMaNG48byAwRYiwSmoMJESJEiBDDQqgGEyJEiBAhhoVQgAkRIkSIEMNCKMCECBEiRIhhIRRgQoQIESLEsBAKMCFChAgRYlgIBZgQIUKECDEshAJMiBAhQoQYFkIBJkSIECFCDAuhABMiRIgQIYaFUIAJESJEiBDDQijAhAgRIkSIYeH/AwbZhQWHaR7EAAAAAElFTkSuQmCC\n" + }, + "metadata": {} + } + ], + "source": [ + "### Function minimization with automatic differentiation and SGD ###\n", + "\n", + "from mpl_toolkits.mplot3d import Axes3D\n", + "\n", + "# Initialize a random value for our initial x\n", + "w = tf.Variable([0.5])\n", + "b = tf.Variable([0.5])\n", + "print(\"Initializing x={}\".format(w.numpy()))\n", + "print(\"Initializing b={}\".format(b.numpy()))\n", + "\n", + "# learning_rate = 1e-2 # learning rate for SGD\n", + "learning_rate = 0.0001 # learning rate for SGD\n", + "history_w = []\n", + "history_b = []\n", + "# Define the target value\n", + "x_f = 4\n", + "x_x = 9\n", + "\n", + "# We will run SGD for a number of iterations. At each iteration, we compute the loss,\n", + "# compute the derivative of the loss with respect to x, and perform the SGD update.\n", + "for i in range(500):\n", + " with tf.GradientTape() as tape:\n", + " '''TODO: define the loss as described above'''\n", + " loss = (w*x_x + b - x_f)**2\n", + " # loss minimization using gradient tape\n", + " grad = tape.gradient(loss, w) # compute the derivative of the loss with respect to x\n", + " new_w = w - learning_rate*grad # sgd update\n", + " with tf.GradientTape() as tape:\n", + " '''TODO: define the loss as described above'''\n", + " loss = (w*x_x + b - x_f)**2\n", + " grad_b = tape.gradient(loss, b) # compute the derivative of the loss with respect to b\n", + " new_b = b - learning_rate*grad_b # sgd update\n", + " w.assign(new_w) # update the value of w\n", + " b.assign(new_b) # update the value of b\n", + " history_w.append(w.numpy()[0])\n", + " history_b.append(b.numpy()[0])\n", + " # history.append(w.numpy()[0]*x_x+b.numpy()[0])\n", + "\n", + "def loss_function(w, b):\n", + " return (w*x_x + b - x_f)**2\n", + "\n", + "# Generate data\n", + "w1 = np.linspace(0.3, 0.5, 500)\n", + "w2 = np.linspace(0.3, 0.5, 500)\n", + "w1, w2 = np.meshgrid(w1, w2)\n", + "loss = loss_function(w1, w2)\n", + "\n", + "# Create a figure and a 3D Axes\n", + "fig = plt.figure()\n", + "ax = fig.add_subplot(111, projection='3d')\n", + "\n", + "# Plot a 3D surface\n", + "ax.plot_surface(w1, w2, loss, cmap='viridis')\n", + "\n", + "# Add labels\n", + "ax.set_xlabel('w1')\n", + "ax.set_ylabel('w2')\n", + "ax.set_zlabel('Loss')\n", + "\n", + "\n", + "# # plot 3D loss\n", + "\n", + "\n", + "fig = plt.figure()\n", + "ax = fig.add_subplot(111, projection='3d')\n", + "x, y = np.meshgrid(history_w, history_b)\n", + "z = loss_function(x, y)\n", + "ax.plot_surface(x, y, z, cmap='viridis')\n", + "ax.set_xlabel('w')\n", + "ax.set_ylabel('b')\n", + "ax.set_zlabel('zoo')\n", + "plt.show()\n", + "\n", + "# # Plot the evolution of x as we optimize towards x_f!\n", + "# plt.plot(history[:100])\n", + "# plt.plot([0, 100],[x_f,x_f])\n", + "# plt.legend(('Predicted', 'Observed'))\n", + "# plt.xlabel('Iteration')\n", + "# plt.ylabel('x value')" + ] + }, + { + "cell_type": "markdown", + "metadata": { + "id": "pC7czCwk3ceH" + }, + "source": [ + "`GradientTape` provides an extremely flexible framework for automatic differentiation. In order to back propagate errors through a neural network, we track forward passes on the Tape, use this information to determine the gradients, and then use these gradients for optimization using SGD." + ] + } + ], + "metadata": { + "accelerator": "GPU", + "colab": { + "collapsed_sections": [ + "WBk0ZDWY-ff8" + ], + "provenance": [], + "include_colab_link": true + }, + "kernelspec": { + "display_name": "Python 3", + "language": "python", + "name": "python3" + }, + "language_info": { + "name": "python", + "version": "3.9.6" + }, + "vscode": { + "interpreter": { + "hash": "31f2aee4e71d21fbe5cf8b01ff0e069b9275f58929596ceb00d14d90e3e16cd6" + } + } + }, + "nbformat": 4, + "nbformat_minor": 0 +} \ No newline at end of file