conditional_neural_process.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "oaNVFIQZ-HLL"
   },
   "source": [
    "Copyright 2018 Google LLC\n",
    "\n",
    "Licensed under the Apache License, Version 2.0 (the \"License\");\n",
    "you may not use this file except in compliance with the License.\n",
    "You may obtain a copy of the License at\n",
    "\n",
    " [https://www.apache.org/licenses/LICENSE-2.0](https://www.apache.org/licenses/LICENSE-2.0)\n",
    "\n",
    "Unless required by applicable law or agreed to in writing, software\n",
    "distributed under the License is distributed on an \"AS IS\" BASIS,\n",
    "WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n",
    "See the License for the specific language governing permissions and\n",
    "limitations under the License.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "X6mXxhKedaRW"
   },
   "source": [
    "# Conditional Neural Processes (CNP) for 1D regression.\n",
    "\n",
    "[Conditional Neural Processes](https://arxiv.org/pdf/1807.01613.pdf) (CNPs) were\n",
    "introduced as a continuation of\n",
    "[Generative Query Networks](https://deepmind.com/blog/neural-scene-representation-and-rendering/)\n",
    "(GQN) to extend its training regime to tasks beyond scene rendering, e.g. to\n",
    "regression and classification.\n",
    "\n",
    "In contrast to most standard neural networks, CNPs learn to approximate a\n",
    "distribution over functions rather than approximating just a single function. As a result, at\n",
    "test time CNPs are flexible and can approximate any function from this\n",
    "distribution when provided with a handful of observations. In addition, they\n",
    "learn to estimate the uncertainty of their prediction from the dataset and as\n",
    "the number of observations is increased this uncertainty reduces and the\n",
    "accuracy of their prediction increases.\n",
    "\n",
    "In this notebook we describe the different parts of a CNP and apply the\n",
    "resulting model to a 1D regression task where a CNP is trained on a dataset of\n",
    "random functions.\n",
    "\n",
    "Any thoughts or questions? We'd love any feedback (about this notebook or CNPs\n",
    "in general) so just contact us at garnelo@google.com."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "Q1nCi4-vkpdM"
   },
   "source": [
    "## Implementing CNPs\n",
    "\n",
    "We start by importing the necessary dependencies. We will make use of numpy,\n",
    "tensorflow, and matplotlib."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "cellView": "both",
    "colab": {},
    "colab_type": "code",
    "id": "enLH-GEtVr_J"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/usr/local/lib/python3.5/dist-packages/h5py/__init__.py:36: FutureWarning: Conversion of the second argument of issubdtype from `float` to `np.floating` is deprecated. In future, it will be treated as `np.float64 == np.dtype(float).type`.\n",
      "  from ._conv import register_converters as _register_converters\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import collections\n",
    "from tqdm import tqdm"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "KDIziZYOLTgX"
   },
   "source": [
    "## Training data\n",
    "\n",
    "A crucial property of CNPs is their flexibility at test time, as they can model\n",
    "a whole range of functions and narrow down their prediction as we condition on\n",
    "an increasing number of context observations. This behaviour is a result of the\n",
    "training regime of CNPs which is reflected in our datasets.\n",
    "\n",
    "![](https://bit.ly/2O2Lq8c)\n",
    "\n",
    "Rather than training using observations from a single function as it is often\n",
    "the case in machine learning (for example value functions in reinforcement\n",
    "learning) we will use a dataset that consists of many different functions that\n",
    "share some underlying characteristics. This is visualized in the figure above.\n",
    "The example on the left corresponds to a classic training regime: we have a\n",
    "single underlying ground truth function (eg. our value function for an agent) in\n",
    "grey and at each learning iteration we are provided with a handful of examples from this\n",
    "function that we have visualized in different colours for batches of different\n",
    "iterations. On the right we show an example of a dataset that could be used for\n",
    "training neural processes. Instead of a single function, it consists of a large number of functions of a function-class that we are interested in modeling. At each iteration we randomly choose one from the dataset and provide some observations from that function for training. For the next iteration we put that function back and\n",
    "pick a new one from our dataset and use this new function to select the training\n",
    "data. This type of dataset ensures that our model can't overfit to a single\n",
    "function but rather learns a distribution over functions. This idea of a\n",
    "hierarchical dataset also lies at the core of current meta-learning methods.\n",
    "Examples of such datasets could be:\n",
    "\n",
    "*  Functions describing the evolution of temperature over time in different cities \n",
    "of the world.\n",
    "*  A dataset of functions generated by a motion capture sensor of different humans\n",
    "    walking.\n",
    "*   As in this particular example differents functions generated by a Gaussian process (GP)\n",
    "    with a specific kernel.\n",
    "\n",
    "We have chosen GPs for the data generation of this example because they\n",
    "constitute an easy way of sampling smooth curves that share some underlying\n",
    "characteristic (in this case the kernel). Other than for data generation of this\n",
    "particular example neural processes do not make use of kernels or GPs as they\n",
    "are implemented as neural networks."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "IBp9NkWGw-n_"
   },
   "source": [
    "## Data generator\n",
    "\n",
    "In the following section we provide the code for generating our training and\n",
    "testing sets using a GP to generate a dataset of functions. As we will explain\n",
    "later, CNPs use two subset of points at every iteration: one to serve as the\n",
    "context, and the other as targets. In practise we found that including the\n",
    "context points as targets together with some additional new points helped during\n",
    "training. Our data generator divides the generated data into these two groups\n",
    "and provides it in the correct format."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "cellView": "both",
    "colab": {},
    "colab_type": "code",
    "id": "SI188jyyJvHl"
   },
   "outputs": [],
   "source": [
    "# The CNP takes as input a `CNPRegressionDescription` namedtuple with fields:\n",
    "#   `query`: a tuple containing ((context_x, context_y), target_x)\n",
    "#   `target_y`: a tesor containing the ground truth for the targets to be\n",
    "#     predicted\n",
    "#   `num_total_points`: A vector containing a scalar that describes the total\n",
    "#     number of datapoints used (context + target)\n",
    "#   `num_context_points`: A vector containing a scalar that describes the number\n",
    "#     of datapoints used as context\n",
    "# The GPCurvesReader returns the newly sampled data in this format at each\n",
    "# iteration\n",
    "\n",
    "CNPRegressionDescription = collections.namedtuple(\n",
    "    \"CNPRegressionDescription\",\n",
    "    (\"query\", \"target_y\", \"num_total_points\", \"num_context_points\"))\n",
    "\n",
    "\n",
    "class InstancesReader(object):\n",
    "    def __init__(self,\n",
    "                 batch_size,\n",
    "                 max_num_context,\n",
    "                 testing=False):\n",
    "        self._batch_size = batch_size\n",
    "        self._max_num_context = max_num_context\n",
    "        self._testing = testing\n",
    "        \n",
    "    def generate_instance(self):\n",
    "        num_context = tf.random_uniform(\n",
    "            shape=[], minval=3, maxval=self._max_num_context, dtype=tf.int32)\n",
    "        \n",
    "        query = ((context_x, context_y), target_x)\n",
    "        \n",
    "        # If we are testing we want to have more targets and have them evenly\n",
    "        # distributed in order to plot the function.\n",
    "        if self._testing:\n",
    "            num_target = 400\n",
    "            num_total_points = num_target\n",
    "            \n",
    "            x_values = tf.tile(\n",
    "              tf.expand_dims(tf.range(-2., 2., 1. / 100, dtype=tf.float32), axis=0),\n",
    "              [self._batch_size, 1])\n",
    "            x_values = tf.expand_dims(x_values, axis=-1)\n",
    "            \n",
    "            y_values = tf.tile(\n",
    "              tf.expand_dims(tf.range(-2., 2., 1. / 100, dtype=tf.float32), axis=0),\n",
    "              [self._batch_size, 1])\n",
    "            y_values = tf.expand_dims(y_values, axis=-1)\n",
    "        # During training the number of target points and their x-positions are\n",
    "        # selected at random\n",
    "        else:\n",
    "            num_target = tf.random_uniform(\n",
    "              shape=(), minval=2, maxval=self._max_num_context, dtype=tf.int32)\n",
    "            num_total_points = num_context + num_target\n",
    "            x_values = tf.random_uniform(\n",
    "              [self._batch_size, num_total_points, self._x_size], -2, 2)\n",
    "\n",
    "        return CNPRegressionDescription(\n",
    "            query=query,\n",
    "            target_y=target_y,\n",
    "            num_total_points=tf.shape(target_x)[1],\n",
    "            num_context_points=num_context)\n",
    "\n",
    "class GPCurvesReader(object):\n",
    "    \"\"\"Generates curves using a Gaussian Process (GP).\n",
    "\n",
    "    Supports vector inputs (x) and vector outputs (y). Kernel is\n",
    "    mean-squared exponential, using the x-value l2 coordinate distance scaled by\n",
    "    some factor chosen randomly in a range. Outputs are independent gaussian\n",
    "    processes.\n",
    "    \"\"\"\n",
    "\n",
    "    def __init__(self,\n",
    "               batch_size,\n",
    "               max_num_context,\n",
    "               x_size=1,\n",
    "               y_size=1,\n",
    "               l1_scale=0.4,\n",
    "               sigma_scale=1.0,\n",
    "               testing=False):\n",
    "        \"\"\"Creates a regression dataset of functions sampled from a GP.\n",
    "\n",
    "        Args:\n",
    "          batch_size: An integer.\n",
    "          max_num_context: The max number of observations in the context.\n",
    "          x_size: Integer >= 1 for length of \"x values\" vector.\n",
    "          y_size: Integer >= 1 for length of \"y values\" vector.\n",
    "          l1_scale: Float; typical scale for kernel distance function.\n",
    "          sigma_scale: Float; typical scale for variance.\n",
    "          testing: Boolean that indicates whether we are testing. If so there are\n",
    "              more targets for visualization.\n",
    "        \"\"\"\n",
    "        self._batch_size = batch_size\n",
    "        self._max_num_context = max_num_context\n",
    "        self._x_size = x_size\n",
    "        self._y_size = y_size\n",
    "        self._l1_scale = l1_scale\n",
    "        self._sigma_scale = sigma_scale\n",
    "        self._testing = testing\n",
    "\n",
    "    def _gaussian_kernel(self, xdata, l1, sigma_f, sigma_noise=2e-2):\n",
    "        \"\"\"Applies the Gaussian kernel to generate curve data.\n",
    "\n",
    "        Args:\n",
    "          xdata: Tensor with shape `[batch_size, num_total_points, x_size]` with\n",
    "              the values of the x-axis data.\n",
    "          l1: Tensor with shape `[batch_size, y_size, x_size]`, the scale\n",
    "              parameter of the Gaussian kernel.\n",
    "          sigma_f: Float tensor with shape `[batch_size, y_size]`; the magnitude\n",
    "              of the std.\n",
    "          sigma_noise: Float, std of the noise that we add for stability.\n",
    "\n",
    "        Returns:\n",
    "          The kernel, a float tensor with shape\n",
    "          `[batch_size, y_size, num_total_points, num_total_points]`.\n",
    "        \"\"\"\n",
    "        num_total_points = tf.shape(xdata)[1]\n",
    "\n",
    "        # Expand and take the difference\n",
    "        xdata1 = tf.expand_dims(xdata, axis=1)  # [B, 1, num_total_points, x_size]\n",
    "        xdata2 = tf.expand_dims(xdata, axis=2)  # [B, num_total_points, 1, x_size]\n",
    "        diff = xdata1 - xdata2  # [B, num_total_points, num_total_points, x_size]\n",
    "\n",
    "        # [B, y_size, num_total_points, num_total_points, x_size]\n",
    "        norm = tf.square(diff[:, None, :, :, :] / l1[:, :, None, None, :])\n",
    "\n",
    "        norm = tf.reduce_sum(\n",
    "            norm, -1)  # [B, data_size, num_total_points, num_total_points]\n",
    "\n",
    "        # [B, y_size, num_total_points, num_total_points]\n",
    "        kernel = tf.square(sigma_f)[:, :, None, None] * tf.exp(-0.5 * norm)\n",
    "\n",
    "        # Add some noise to the diagonal to make the cholesky work.\n",
    "        kernel += (sigma_noise**2) * tf.eye(num_total_points)\n",
    "\n",
    "        return kernel\n",
    "\n",
    "    def generate_curves(self):\n",
    "        \"\"\"Builds the op delivering the data.\n",
    "\n",
    "        Generated functions are `float32` with x values between -2 and 2.\n",
    "\n",
    "        Returns:\n",
    "          A `CNPRegressionDescription` namedtuple.\n",
    "        \"\"\"\n",
    "        num_context = tf.random_uniform(\n",
    "            shape=[], minval=3, maxval=self._max_num_context, dtype=tf.int32)\n",
    "\n",
    "        # If we are testing we want to have more targets and have them evenly\n",
    "        # distributed in order to plot the function.\n",
    "        if self._testing:\n",
    "            num_target = 400\n",
    "            num_total_points = num_target\n",
    "            x_values = tf.tile(\n",
    "              tf.expand_dims(tf.range(-2., 2., 1. / 100, dtype=tf.float32), axis=0),\n",
    "              [self._batch_size, 1])\n",
    "            x_values = tf.expand_dims(x_values, axis=-1)\n",
    "        # During training the number of target points and their x-positions are\n",
    "        # selected at random\n",
    "        else:\n",
    "            num_target = tf.random_uniform(\n",
    "              shape=(), minval=2, maxval=self._max_num_context, dtype=tf.int32)\n",
    "            num_total_points = num_context + num_target\n",
    "            x_values = tf.random_uniform(\n",
    "              [self._batch_size, num_total_points, self._x_size], -2, 2)\n",
    "\n",
    "        # Set kernel parameters\n",
    "        l1 = (\n",
    "            tf.ones(shape=[self._batch_size, self._y_size, self._x_size]) *\n",
    "            self._l1_scale)\n",
    "        sigma_f = tf.ones(\n",
    "            shape=[self._batch_size, self._y_size]) * self._sigma_scale\n",
    "\n",
    "        # Pass the x_values through the Gaussian kernel\n",
    "        # [batch_size, y_size, num_total_points, num_total_points]\n",
    "        kernel = self._gaussian_kernel(x_values, l1, sigma_f)\n",
    "\n",
    "        # Calculate Cholesky, using double precision for better stability:\n",
    "        cholesky = tf.cast(tf.cholesky(tf.cast(kernel, tf.float64)), tf.float32)\n",
    "\n",
    "        # Sample a curve\n",
    "        # [batch_size, y_size, num_total_points, 1]\n",
    "        y_values = tf.matmul(\n",
    "            cholesky,\n",
    "            tf.random_normal([self._batch_size, self._y_size, num_total_points, 1]))\n",
    "\n",
    "        # [batch_size, num_total_points, y_size]\n",
    "        y_values = tf.transpose(tf.squeeze(y_values, 3), [0, 2, 1])\n",
    "\n",
    "        if self._testing:\n",
    "            # Select the targets\n",
    "            target_x = x_values\n",
    "            target_y = y_values\n",
    "\n",
    "            # Select the observations\n",
    "            idx = tf.random_shuffle(tf.range(num_target))\n",
    "            context_x = tf.gather(x_values, idx[:num_context], axis=1)\n",
    "            context_y = tf.gather(y_values, idx[:num_context], axis=1)\n",
    "\n",
    "        else:\n",
    "            # Select the targets which will consist of the context points as well as\n",
    "            # some new target points\n",
    "            target_x = x_values[:, :num_target + num_context, :]\n",
    "            target_y = y_values[:, :num_target + num_context, :]\n",
    "\n",
    "            # Select the observations\n",
    "            context_x = x_values[:, :num_context, :]\n",
    "            context_y = y_values[:, :num_context, :]\n",
    "\n",
    "        query = ((context_x, context_y), target_x)\n",
    "\n",
    "        return CNPRegressionDescription(\n",
    "            query=query,\n",
    "            target_y=target_y,\n",
    "            num_total_points=tf.shape(target_x)[1],\n",
    "            num_context_points=num_context)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "AqK9Xco1fuNz"
   },
   "source": [
    "## Conditional Neural Processes\n",
    "\n",
    "We can visualise a forward pass in a CNP as follows:\n",
    "\n",
    "<img src=\"https://bit.ly/2OFb6ZK\" alt=\"drawing\" width=\"400\"/>\n",
    "\n",
    "As shown in the diagram, CNPs take in pairs **(x, y)<sub>i</sub>** of context\n",
    "points, pass them through an **encoder** to obtain\n",
    "individual representations **r<sub>i</sub>** which are combined using an **aggregator**. The resulting representation **r**\n",
    "is then combined with the locations of the targets **x<sub>T</sub>** and passed\n",
    "through a **decoder** that returns a mean estimate\n",
    "of the **y** value at that target location together with a measure of the\n",
    "uncertainty over said prediction. Implementing CNPs therefore involves coding up\n",
    "the three main building blocks:\n",
    "\n",
    "*   Encoder\n",
    "*   Aggregator\n",
    "*   Decoder\n",
    "\n",
    "A more detailed description of these three parts is presented in the following\n",
    "sections alongside the code."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "lxl8wMT_Io5B"
   },
   "source": [
    "## Encoder\n",
    "\n",
    "The encoder **e** is shared between all the context points and consists of an\n",
    "MLP with a handful of layers. For this experiment four layers are enough, but we\n",
    "can still change the number and size of the layers when we build the graph later\n",
    "on via the variable **`encoder_output_sizes`**. Each of the context pairs **(x,\n",
    "y)<sub>i</sub>** results in an individual representation **r<sub>i</sub>** after\n",
    "encoding. These representations are then combined across context points to form\n",
    "a single representation **r** using the aggregator **a**.\n",
    "\n",
    "In this implementation we have included the aggregator **a** in the encoder as\n",
    "we are only taking the mean across all points. The representation **r** produced\n",
    "by the aggregator contains the information about the underlying unknown function\n",
    "**f** that is provided by all the context points."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "cellView": "both",
    "colab": {},
    "colab_type": "code",
    "id": "jw0VYpkQWGhq"
   },
   "outputs": [],
   "source": [
    "class DeterministicEncoder(object):\n",
    "    \"\"\"The Encoder.\"\"\"\n",
    "\n",
    "    def __init__(self, output_sizes, aggregator_method=tf.reduce_mean):\n",
    "        \"\"\"CNP encoder.\n",
    "\n",
    "        Args:\n",
    "          output_sizes: An iterable containing the output sizes of the encoding MLP.\n",
    "        \"\"\"\n",
    "        self._output_sizes = output_sizes\n",
    "        self.aggregator_method = aggregator_method\n",
    "\n",
    "    def __call__(self, context_x, context_y, num_context_points):\n",
    "        \"\"\"Encodes the inputs into one representation.\n",
    "\n",
    "        Args:\n",
    "          context_x: Tensor of size bs x observations x m_ch. For this 1D regression\n",
    "              task this corresponds to the x-values.\n",
    "          context_y: Tensor of size bs x observations x d_ch. For this 1D regression\n",
    "              task this corresponds to the y-values.\n",
    "          num_context_points: A tensor containing a single scalar that indicates the\n",
    "              number of context_points provided in this iteration.\n",
    "\n",
    "        Returns:\n",
    "          representation: The encoded representation averaged over all context \n",
    "              points.\n",
    "        \"\"\"\n",
    "\n",
    "        # Concatenate x and y along the filter axes\n",
    "        encoder_input = tf.concat([context_x, context_y], axis=-1)\n",
    "\n",
    "        # Get the shapes of the input and reshape to parallelise across observations\n",
    "        batch_size, _, filter_size = encoder_input.shape.as_list()\n",
    "        hidden = tf.reshape(encoder_input, (batch_size * num_context_points, -1))\n",
    "        hidden.set_shape((None, filter_size))\n",
    "\n",
    "        # Pass through MLP\n",
    "        with tf.variable_scope(\"encoder\", reuse=tf.AUTO_REUSE):\n",
    "            for i, size in enumerate(self._output_sizes[:-1]):\n",
    "                hidden = tf.nn.relu(\n",
    "                    tf.layers.dense(hidden, size, name=\"Encoder_layer_{}\".format(i)))\n",
    "\n",
    "            # Last layer without a ReLu\n",
    "            hidden = tf.layers.dense(\n",
    "              hidden, self._output_sizes[-1], name=\"Encoder_layer_{}\".format(i + 1))\n",
    "\n",
    "        # Bring back into original shape\n",
    "        hidden = tf.reshape(hidden, (batch_size, num_context_points, size))\n",
    "\n",
    "        # Aggregator: take the mean over all points\n",
    "        representation = self.aggregator_method(hidden, axis=1)\n",
    "\n",
    "        representation = tf.reshape(representation, (batch_size, size))\n",
    "        return representation"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "ZE7hj8dnz5X2"
   },
   "source": [
    "## Decoder\n",
    "\n",
    "Once we have obtained our representation **r** we concatenate it with each of\n",
    "the targets **x<sub>t</sub>** and pass it through the decoder **d**. As with the\n",
    "encoder **e**, the decoder **d** is shared between all the target points and\n",
    "consists of a small MLP with layer sizes defined in **`decoder_output_sizes`**.\n",
    "The decoder outputs a mean **&mu;<sub>t</sub>** and a variance\n",
    "**&sigma;<sub>t</sub>** for each of the targets **x<sub>t</sub>**. To train our\n",
    "CNP we use the log likelihood of the ground truth value **y<sub>t</sub>** under\n",
    "a Gaussian parametrized by these predicted **&mu;<sub>t</sub>** and\n",
    "**&sigma;<sub>t</sub>**.\n",
    "\n",
    "In this implementation we clip the variance **&sigma;<sub>t</sub>** at 0.1 to\n",
    "avoid collapsing."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "cellView": "both",
    "colab": {},
    "colab_type": "code",
    "id": "WGzI381UV7FJ"
   },
   "outputs": [],
   "source": [
    "class DeterministicDecoder(object):\n",
    "    \"\"\"The Decoder.\"\"\"\n",
    "\n",
    "    def __init__(self, output_sizes):\n",
    "        \"\"\"CNP decoder.\n",
    "\n",
    "        Args:\n",
    "          output_sizes: An iterable containing the output sizes of the decoder MLP \n",
    "              as defined in `basic.Linear`.\n",
    "        \"\"\"\n",
    "        self._output_sizes = output_sizes\n",
    "\n",
    "    def __call__(self, representation, target_x, num_total_points):\n",
    "        \"\"\"Decodes the individual targets.\n",
    "\n",
    "        Args:\n",
    "          representation: The encoded representation of the context\n",
    "          target_x: The x locations for the target query\n",
    "          num_total_points: The number of target points.\n",
    "\n",
    "        Returns:\n",
    "          dist: A multivariate Gaussian over the target points.\n",
    "          mu: The mean of the multivariate Gaussian.\n",
    "          sigma: The standard deviation of the multivariate Gaussian.\n",
    "        \"\"\"\n",
    "\n",
    "        # Concatenate the representation and the target_x\n",
    "        representation = tf.tile(\n",
    "            tf.expand_dims(representation, axis=1), [1, num_total_points, 1])\n",
    "        input_ = tf.concat([representation, target_x], axis=-1)\n",
    "\n",
    "        # Get the shapes of the input and reshape to parallelise across observations\n",
    "        batch_size, _, filter_size = input_.shape.as_list()\n",
    "        hidden = tf.reshape(input_, (batch_size * num_total_points, -1))\n",
    "        hidden.set_shape((None, filter_size))\n",
    "\n",
    "        # Pass through MLP\n",
    "        with tf.variable_scope(\"decoder\", reuse=tf.AUTO_REUSE):\n",
    "            for i, size in enumerate(self._output_sizes[:-1]):\n",
    "                hidden = tf.nn.relu(\n",
    "                    tf.layers.dense(hidden, size, name=\"Decoder_layer_{}\".format(i)))\n",
    "\n",
    "            # Last layer without a ReLu\n",
    "            hidden = tf.layers.dense(\n",
    "              hidden, self._output_sizes[-1], name=\"Decoder_layer_{}\".format(i + 1))\n",
    "\n",
    "        # Bring back into original shape\n",
    "        hidden = tf.reshape(hidden, (batch_size, num_total_points, -1))\n",
    "\n",
    "        # Get the mean and the variance\n",
    "        mu, log_sigma = tf.split(hidden, 2, axis=-1)\n",
    "\n",
    "        # Bound the variance\n",
    "        sigma = 0.1 + 0.9 * tf.nn.softplus(log_sigma)\n",
    "\n",
    "        # Get the distribution\n",
    "        dist = tf.contrib.distributions.MultivariateNormalDiag(\n",
    "            loc=mu, scale_diag=sigma)\n",
    "\n",
    "        return dist, mu, sigma"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "JfMjcMHf019M"
   },
   "source": [
    "## Model\n",
    "\n",
    "Now that the main building blocks (encoder, aggregator and decoder) of the CNP\n",
    "are defined we can put everything together into one model. Fundamentally this\n",
    "model only needs to include two main methods: 1. A method that returns the log\n",
    "likelihood of the targets' ground truth values under the predicted\n",
    "distribution.This method will be called during training as our loss function. 2.\n",
    "Another method that returns the predicted mean and variance at the target\n",
    "locations in order to evaluate or query the CNP at test time. This second method\n",
    "needs to be defined separately as, unlike the method above, it should not depend\n",
    "on the ground truth target values."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "cellView": "both",
    "colab": {},
    "colab_type": "code",
    "id": "P3LJYP1Qh-jO"
   },
   "outputs": [],
   "source": [
    "class DeterministicModel(object):\n",
    "    \"\"\"The CNP model.\"\"\"\n",
    "\n",
    "    def __init__(self, encoder_output_sizes, decoder_output_sizes, aggregator_method=tf.reduce_mean):\n",
    "        \"\"\"Initialises the model.\n",
    "\n",
    "        Args:\n",
    "          encoder_output_sizes: An iterable containing the sizes of hidden layers of\n",
    "              the encoder. The last one is the size of the representation r.\n",
    "          decoder_output_sizes: An iterable containing the sizes of hidden layers of\n",
    "              the decoder. The last element should correspond to the dimension of\n",
    "              the y * 2 (it encodes both mean and variance concatenated)\n",
    "        \"\"\"\n",
    "        self._encoder = DeterministicEncoder(encoder_output_sizes, aggregator_method)\n",
    "        self._decoder = DeterministicDecoder(decoder_output_sizes)\n",
    "\n",
    "    def __call__(self, query, num_total_points, num_contexts, target_y=None):\n",
    "        \"\"\"Returns the predicted mean and variance at the target points.\n",
    "\n",
    "        Args:\n",
    "          query: Array containing ((context_x, context_y), target_x) where:\n",
    "              context_x: Array of shape batch_size x num_context x 1 contains the \n",
    "                  x values of the context points.\n",
    "              context_y: Array of shape batch_size x num_context x 1 contains the \n",
    "                  y values of the context points.\n",
    "              target_x: Array of shape batch_size x num_target x 1 contains the\n",
    "                  x values of the target points.\n",
    "          target_y: The ground truth y values of the target y. An array of \n",
    "              shape batchsize x num_targets x 1.\n",
    "          num_total_points: Number of target points.\n",
    "\n",
    "        Returns:\n",
    "          log_p: The log_probability of the target_y given the predicted\n",
    "          distribution.\n",
    "          mu: The mean of the predicted distribution.\n",
    "          sigma: The variance of the predicted distribution.\n",
    "        \"\"\"\n",
    "\n",
    "        (context_x, context_y), target_x = query\n",
    "\n",
    "        # Pass query through the encoder and the decoder\n",
    "        representation = self._encoder(context_x, context_y, num_contexts)\n",
    "        dist, mu, sigma = self._decoder(representation, target_x, num_total_points)\n",
    "\n",
    "        # If we want to calculate the log_prob for training we will make use of the\n",
    "        # target_y. At test time the target_y is not available so we return None\n",
    "        if target_y is not None:\n",
    "            log_p = dist.log_prob(target_y)\n",
    "        else:\n",
    "            log_p = None\n",
    "\n",
    "        return log_p, mu, sigma"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "4eaNM3-HWqkW"
   },
   "source": [
    "## Plotting function\n",
    "We define a helper function for plotting the intermediate predictions\n",
    "every `PLOT_AFTER` iterations. The ground truth curve will be shown as a black\n",
    "dotted line and the context points from this curve that are fed into the model\n",
    "as black dots. The model's predicted mean and variance is shown in blue for a\n",
    "range of target points in the interval [-2, 2]."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "NunDvti3VVwL"
   },
   "outputs": [],
   "source": [
    "def plot_functions(target_x, target_y, context_x, context_y, pred_y, var):\n",
    "    \"\"\"Plots the predicted mean and variance and the context points.\n",
    "\n",
    "    Args: \n",
    "    target_x: An array of shape batchsize x number_targets x 1 that contains the\n",
    "        x values of the target points.\n",
    "    target_y: An array of shape batchsize x number_targets x 1 that contains the\n",
    "        y values of the target points.\n",
    "    context_x: An array of shape batchsize x number_context x 1 that contains \n",
    "        the x values of the context points.\n",
    "    context_y: An array of shape batchsize x number_context x 1 that contains \n",
    "        the y values of the context points.\n",
    "    pred_y: An array of shape batchsize x number_targets x 1  that contains the\n",
    "        predicted means of the y values at the target points in target_x.\n",
    "    pred_y: An array of shape batchsize x number_targets x 1  that contains the\n",
    "        predicted variance of the y values at the target points in target_x.\n",
    "    \"\"\"\n",
    "    # Plot everything\n",
    "    plt.plot(target_x[0], pred_y[0], 'b', linewidth=2)\n",
    "    plt.plot(target_x[0], target_y[0], 'k:', linewidth=2)\n",
    "    plt.plot(context_x[0], context_y[0], 'ko', markersize=10)\n",
    "    plt.fill_between(\n",
    "      target_x[0, :, 0],\n",
    "      pred_y[0, :, 0] - var[0, :, 0],\n",
    "      pred_y[0, :, 0] + var[0, :, 0],\n",
    "      alpha=0.2,\n",
    "      facecolor='#65c9f7',\n",
    "      interpolate=True)\n",
    "\n",
    "    # Make the plot pretty\n",
    "    plt.yticks([-2, 0, 2], fontsize=16)\n",
    "    plt.xticks([-2, 0, 2], fontsize=16)\n",
    "    plt.ylim([-2, 2])\n",
    "    plt.grid('off')\n",
    "    ax = plt.gca()\n",
    "    ax.set_axis_bgcolor('white')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "colab_type": "text",
    "id": "YFL6HWKoOpgw"
   },
   "source": [
    "## Running Conditional Neural Processes\n",
    "\n",
    "Now that we have defined the dataset as well as our model and its components we\n",
    "can start building everything into the graph. Before we get started we need to\n",
    "set some variables:\n",
    "\n",
    "*   **`TRAINING_ITERATIONS`** - a scalar that describes the number of iterations\n",
    "    for training. At each iteration we will sample a new batch of functions from\n",
    "    the GP, pick some of the points on the curves as our context points **(x,\n",
    "    y)<sub>C</sub>** and some points as our target points **(x,\n",
    "    y)<sub>T</sub>**. We will predict the mean and variance at the target points\n",
    "    given the context and use the log likelihood of the ground truth targets as\n",
    "    our loss to update the model.\n",
    "*   **`MAX_CONTEXT_POINTS`** - a scalar that sets the maximum number of context\n",
    "    points used during training. The number of context points will then be a\n",
    "    value between 3 and `MAX_CONTEXT_POINTS` that is sampled at random for every\n",
    "    iteration.\n",
    "*   **`PLOT_AFTER`** - a scalar that regulates how often we plot the\n",
    "    intermediate results."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "colab": {},
    "colab_type": "code",
    "id": "5qSrrfzpPMPz"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Iteration: 0, loss: 1.4917353391647339:   0%|          | 0/200000 [00:00<?, ?it/s]/usr/local/lib/python3.5/dist-packages/ipykernel_launcher.py:36: MatplotlibDeprecationWarning: The set_axis_bgcolor function was deprecated in version 2.0. Use set_facecolor instead.\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX8AAAEDCAYAAADdpATdAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvNQv5yAAAIABJREFUeJzt3Xl8U1X6x/HPSZvSjQIqKgoKVBRxYRUZWRRQkMUFZBQVZHQUFBR1EByXn8so6qCIivuCooA4Io6ggrigMgIqKlQQQUA20QKytXRJ0pzfH2lLutECTW6afN+vV7i5W86TNDz33HNPzjXWWkREJLa4nA5ARETCT8lfRCQGKfmLiMQgJX8RkRik5C8iEoOU/EVEYlDIkr8xZoAx5h1jzAZjTK4xZpUx5mFjTO1QlSkiIlVjQtXP3xizGNgIvAdsBloD9wE/A2dZa/0hKVhERCoVyuRf31q7rdSyq4DJQHdr7WchKVhERCoVsmaf0om/0LeF02NDVa6IiFQu3Bd8zy6crgxzuSIiEiRkzT5lCjLmWOAHYJm19rwKthkKDAVISUlp27x587DEJiISLb777rvt1tr6lW0XluRvjEkFPgeOAdpbazdXtk+7du3skiVLQh2aiEhUMcZ8Z61tV9l28WEIJAmYDTQFzq5K4hcRkdAKafI3xriBGUA74Dxr7Y+hLE9ERKomZMnfGOMCpgLdgL7W2sWhKktERA5MKGv+zwB/BcYCe40xHYLWbVbzj4iIc0LZ1bNX4fQuYFGpx7UhLFdERCoRspq/tbZxqF5bREQOjUb1FBGJQUr+IiIxSMlfRCQGKfmLiMQgJX8RkRik5C8iEoOU/EVEYpCSv4hIDFLyFxGJQUr+IiIxSMlfRCQGKfmLiMQgJX8RkRik5C8iEoOU/EVEYpCSv4hIDFLyFxGJQUr+IiIxSMlfRCQGKfmLiMQgJX8RkRik5C8iEoOU/EVEYpCSv4hIDFLyFxGJQUr+IiIxSMlfRCQGKfmLiMQgJX8RkRik5C8iEoOU/EVEYpCSv4hIDFLyFxGJQUr+IiIxSMlfRCQGKfmLiMQgJX8RkRik5C8iEoOU/EVEYpCSv4hIDFLyFxGJQUr+IiIxSMlfRCQGKfmLiMQgJX8RkRik5C8iEoOU/EVEYpCSv4hIDFLyFxGJQUr+IiIxSMlfRCQGKfmLiMQgJX8RkRgU73QAIiLRxFqLBfyW4qkfsHbf8+B1FhuYFq6rl2BIjDMhjzPqkn+BtWTm2sCMgaKP0BQ+yiwzRc9N8TJKLN8nePvy1lFqfeltSiwv5/XLey1TwbwpXZCIVFlRgrYEkm5wsi5aFjxfbrJmX8IOntpDje0Q96+qqEv+AN6iT++APsVwfeTVxRYfjAzgKpoWLgvMm+JtXOVsS9DzkgcWHWRqMmsD3+UD/UYHb2/L2bmy16tovalgm9JllE7GBD23pZ5DIBEH9rMl1xUm4jL7Uz3JOVpEZfKPFUVfZoCCcr/R1fE1tyUOChB0NlLB2VGJZfs5w6lI2TMsU2J56QNXeQew0stcYTyIWWuLT+2DT/OLHmWTWWGyDk56hf8EJ7vSy8rbRqSqlPylUsGJpsyKsEVw6K8RfPZTfFAos8zsO3BQthmvKHkX2H3ttAWUbBYQqQmU/CVm+KFUNbo8yt4SG9TVU0QkBin5i4jEICV/EZEYpOQvIhKDlPxFRGKQkr+ISAxS8hcRiUFK/iIiMUjJX0QkBin5i4jEICV/EZEYpOQvIhKDlPxFRGKQkr+ISAxS8hcRiUFK/hFoRcZSBl3Qg1UrljsdiohEKSX/CFJQUEBebi7TX32Z/83/lKw9uwGY9fZ07ht9i8PRiUg0UfKPINl79tCuaQPy8/No2uxEfD4fe3bv5s6R19OybTs8Hg8AmzasByAvN5d7R41k9U8rHIxaRGoiJX+H7Nqxg7y8vBLLdvy5HZ/Px4plS/n0+5/o0Pls0urUwRjDR7P+S0JCAh6Ph86npDP3vZncP+ZWJr/wDFdd3IvsrCyH3omI1ERK/g4Ye+do2jdryOfz5rAiYykffzALgKMaHMNZZ3floaeewxTeOdxay51jx/HvZ18G4NdfVhMfH88PS77h1rvuZcRtdzB7wbek1q7NHTddz5nNGrF54wbH3puI1AzG2si8YXW7du3skiVLDni/AmvZnBOZ7wkgPz+f0xrUpWXb9rz89nu0bHg4AC+99S7n9bmw0v0LCgoYfOH5ZGft4Z1P/4fb7QYCTUDN66cC8MasuXTudl7o3oSIhMyRiYakOHPQ+xtjvrPWtqtsu/iDLkEOSkJCAh99k8Gqn5ZTp25dht/2T5Yt+Za/dOlapf3j4uJ4Y9bc4udFEpOS+O7XP8j8YwstTmsZkthFJHqo5h9GWzP/YMKD93HXQ4+SWrt2yMqx1rJyeQY/L8+g/+WDQ1aOiFQ/1fyj0IQH7iUraw85e7NDmvx37dhB77+0we12c27vC8nLzWHzxg20PuPM4msJIhLbdME3TPx+Pz8u/Z733/kPWXv2hLSseocfzpmdzqbNmWexe+cOhl0xgP7dOjKwV3f8fn9IyxaRmkHJP0xcLhezvvyaWV9+TfqJJ4W8vPEvvspVQ2/gqGOOpWWbdlxy5VWMfeIZ+nU9i7mz3sXn87FyeQZ5ubkhj0VEIo+afcLI5XJxeptKm+KqRcPjjqfhcccDcN9jTwJwee9zWfbdtyz49GNanNaSXh1ac+XfhzH2yWfDEpOIRA7V/ENob3Y2N18ziCH9erN+7Rqnw+HJSW8w7NbR3HT7XRRd6F++9HvW/bKa/Px8h6MTkXBSzT+EfF4vv2/ezDcLF/DFxx/ROP0ER+M58ugG3PHAI0CgR9Dny1ZR7/AjOPv0Zlx13XD+8X/3OxqfiISPav4hVKdePV6fNZf3//ctQ64f4XQ4JRhjaJx+At8uXEDW7t18+dnHWGvZvWsXt4+4Die6AP+y8idefXYiBQUFZdZ5vd4yw2GIyMFTzT/EEhMTObVVG6fDqNC5vS9gwYq1ZO/ZgzGGee//l7cmT6Jrz95gLVs2b2bI9SOIi4tj/rw51KqVSPuOnYmP3/fV8fv9uFyBekRBQQEzprxGau00CgoKSD+pOaec3qrcsrP27CElNZV/33MH1lpefHI8ABt+XcuwW26jwbENi7e9YdBf+X7xIj77YSV1DzsshJ+ISGxQ8g+RNat+Ji8vlxNOOpnExESnw9mvYxo2Kn6evSeLth3OIr3ZSfy5fRvXX/lXMn//jTse/Dfj7r2LlT8u49WZ79O1Ry8Atv7xO51aNOWuhx5j8NAb+GDm29w+YigNjm3I779t5oZRt5OaWptXn32KHX9up3HTE7jsb3/njZee4/13/sNHXy9j3vvv8euaX+h2fm8+m/sha37+ict7d2fetz+yYtkPJCUl88kHswHIzc2hLkr+IodKzT4h8srECfTt2I43XqxZPWkGXXc9Hc/uxp/bt3FUg2M4/Ij6JKekUlBQwGVXXcOJJ59Cnbr1ePCO29iWmcl9t92Mx+Phg3dnBJpmcnMAuGrYCNqf1Znnxv+bdb+s4rXnn2bW29N5+tGHWLLoK774+CN+27iBxQs+p2vP3vTpN4B7xz3Buj1evF4f69euYVDfHgwd2J8CfwHLNv/JW3PnlzgbEJGDp5p/NZv3/nsc07ARSSkpALRq197hiA6M2+0uceH36zWb8Xm9xMXFYYzh9n89xF03D2flj8to37EzE155g7sefoxjGjbCGEPLtu15/6slnNqyNWl16vDNwgVMefl5brvnAbxeL0+PG0v9o47mkadfwJ2QQIvTWtK9V98SMYy5fyxZe3Zz8zWDOPKoBnjy86lTty5nduoS7o9DJGpF1dg+eXnw5FOWHCxuN8S7wZ0Abjf75guXxcfvWxfvhoRS2wevdydA0BhqFfJ4PJzVvDHbt2by7Btv4fP5uGDAZVE3pMLmjRt49L67OK5xU0bd868Kt/t5+Y+89fokzuzYmfMv6g8EmokOr39kiUHpKrI18w/mz/2Ay4b8Ha/Xyy8//8S2zD84+9ye1fZeRCJNuMb2iarkv20bHHlkaOIxxpY4WLhLH0jc4IrL4s9tD7A3ewGnt1mIO8EUH0ji4yGu1LT0MnfxvK1gecl93QlQKxFq1SqcBj+vte95Qi1w1fAGvm2ZmbQ/4VhSa9fm5jvvZc+uXfTt/1eOa9KUxKQkp8MTqTZK/geR/Pfsgfvut+zMBa8HfL7A1OsFnzcw9XoL13nB69v33FN6+6DnXg9YeyB/DD+RdjklIcGWOVAkFD1PKHvwKDpguFyBsx7j2jfvcgXNGzAG/BZs0cO/77nfv29KqWXBy4P3L1pPqf1//OFmCgpy8Hn38vtv03G5EqmV2IBTWo6jyQmXUKsWJCYVvQdb/Dwxad97K16fFHieVGpaK7F6DpQFBZCfV/jID0zzcvc9z88vnC+13usN+nxNOZ93RfOm7PqiZUVTSs0byt/OmH2vHx9U2Yh3F1ZG3IHvRFFFqHhdXGDfWGHtvjzhyQdPYa4oeuR7Ss57Ch9FecWTX3I+8DoGtx+GXWdo1uzg4oqIUT2NMY2ACcB5BL5rnwC3WGs3hqK8tDR49DFCMqRzQYEteeDw7juobN+6HY/XkpJSP+hA4y8+ABX4CqcFgX2K5wunvhLzJjAtCLxO6f2Llvm8gS+bJ79kAinxvHDqyTd4PAaPB2r2zR4fA14D3gOS8fu7k5vzPksWrWbJotJZxwBrgfHAFCAbSAUGAaOA9ApLqZVoSUqGxEQC01IHCJ+vMFnnFX2+ZZO71xtDWTBIfLwtbj5NKGpGLTxjTih8nlBr31lzQlDTatF25VUCSi8repRXobCUrYSUrmyUXlb69XwFQYnbG/i7essk7tD9jc/txkEn/6oKWfI3xiQDnwH5wBACf5MHgfnGmNOttXtDVXYoxMUFHuX12pwx5Slmv/MWr838oBp+xVv9By6/3wZqHaUPFOUdMPICtZH8vMDBxvoD/zn8/sLaefC8f998cI3RmLK1SGMC6dgVtL50rTO4hlpebRTicbmuxecbjM+bS0pqHTZvyGBr5ga+/KQnJ582gFNb/Z28XFj901zmvPdXCgq8WOst/CSywLyMy0ymcfrbJNTqRV5uIGnn5hbVxE3h49A+c2MCZ1r7zkT2TUsvCz4zcScEJaTyPmtb9rMvb76gnDOt4iTnL7s8+OyreHnBvsqH17uv8lFUgfEGrfN5we83+HwGny/wWcaC+HhbfOBKcBce2BJKHtyKDnpFj1oJZQ+K+w6IlnrJhmbNQl95CGXN/zqgKXCStXYNgDEmA/gFGAY8HsKyw8Zay/x5c1i/dg2Zv29xfAiH8rhcgYNWYiJQx+loqkOtwgdAS6a+sphZb//IYUc0J2vPffi8Hj6ZMxGfL6fsrtaL33r5/be/MnfxUo5vWvIMwO+35OdBbk6gZl80zcvZd5CId1eeyN3u2GoCgcBn5yvVlOoNagoJbhYpsy5ovnQloExlovBzLW+5MZRo3irxKGeZK2if4OXx8WXPUILnixJ3VTqCHKgjEyEpBK9bWsja/I0xnwKJ1tqOpZZ/AWCtPXt/+9ekO3l5vV4WfDqPrj17R13Pnprgt00bOb9DKx54fCJPPfIg635ZTVx8PAU+X4X7xLvdXH71dTzw+MQwRipSuXBd8A3lVclTgOXlLF8BtAhhuWHndrvpdn4fJX6HHNvoOBav2kjfSy7jtNZtcSck7DfxQ2DQvXenTwlThCKRJ5TJ/zBgZznLdwD1ytvBGDPUGLPEGLNk27ZtIQyteqz+aYUGG4sQKampxMfH8+SkKfi83sp3APZmZYc4KpHIFVH9Ea21L1pr21lr29WvX9/pcPZr8YIv6NH+dDqd3ITPP57rdDgSJCU1tWrb1a7adiLRKJTJfyfl1/ArOiOoUY5tdByj7xtL02Yn0aptzRrCIdpdfNmVxLvd+90m3u2m38BBYYpIpKQN69Zy9y0jOLVBXZrUjufUBnW5acjlrFyeEbYYQnnB9zMgwVrbqdTyzwvLjZoLvhJZNqxby/kdWpGbU05vn0JJycnl9vYRCbX58+YwfNCleL3ecpsoH39yIreOvPGgXz8SLvjOAjoYY5oGBdUY6Fi4TiQkjm+azrNT/kNScnKZM4B4t5uk5GSenfIfJX4Juw3r1jJ80KXk5uRUeG3q7jtuZ+3atSGPJZTJ/yVgPfCeMeYiY8yFBH6auQl4IYTlhtziBV/w1uRX+GPLb06HIhXo2qMXcxcv5fKrryO1dhrGGFLT0rj86uuYu3hp8f0IRMLppacex1tJhwSv18uECRNCHktIx/YxxhxHyeEdPiUwvMP6yvaN5GafW6+9inenT+Wef0/gmhEjQ1qWVL8PZr7NEUcepSGiJexObVCX7KzKB1lJS0tj9+7dB1VGRIztUziGzyWhLMMJXc7tQc7evXTpfp7TocgBmjD2fp58+F+43W4++2EljRo3cTokiSF7s6vWvTi7itsdiojq6llT9Bs4iBfefIcTmp/sdChygPr0G8B5fS9k6gef8MITj5XobXH3LSPYsC70ba0Su6raDTm1itsdCiX/A7Bp/a9E6hDYUjUntjiFK64ZypCLezF98itkZ2VhrSU7K4vpk1/h/A6tmD9vjtNhSpSqSjdkt9vN4MGDQx6Lkn8V7d61i95nteHy3udW+dRNIs/+elv4vF5yc3IYPuhSnQFISFw38h+4q5D8b7311pDHouRfRatW/Eh+fj65uTlVPnWTyFPV3hYvP/1EmCKSWFLUDTkxsezd54q6IU99623S00PfDTmq7uQFoe3t4/P52L1zJ4dH+NATUrGq9rZITUtj+ZYa/0N0iUC//7aZLZs2MqR/H7L37AEC37d+Awdx7Y23cEaLE8IyqmdIe/tEm/j4eCX+Gq6qTXYa9E1C5f7Rt/DJh7MZ99zLpJ/YnGbNW5CckhL2ONTsUwX5+fls2rDe6TCkGmjQN3GStRaPJx+As87uRsu2ZziS+EHJv0pGDf0bA8/vysirr6SgoMDpcOQQaNA3cZIxhkkzZrNk3e8cfcyxjsai5F8Jay3NTzmN/Lw8el98CXGhuG+bhE1Velu4jIt+A68MU0QSC/x+P999vai4q3jdww5zOCIl/0oZY7hxzJ18vWYz51/U3+lw5BBVNuhbXFwcHk8+P2UsdShCiUb3j76FS7p3onubU9i+davT4QBK/hXKz89n/do1xfOq8UePEoO+paVhjKt40LfR94/ljL904uflPzL5+WecDlWixPX/GEO7v3Rk6M3/4Igjj3Q6HEBdPSs0YvBlLF7wBTfc9k/+PuJm3Z83huTn53Nag7p4PB6Wbf6TOnXrOh2SRAFrbZXySLhu4K6unhU4rmk6C7+Yz/atmUr8MaZWrVpccuUQkpKT8eTnOx2O1GC7duygdp06xMXFRVweUc1fRGqMDevW8tJTj/Pft6ayNzublNRULr7sSq4eMRJb4OfIBseQVqeO02ECsGnDeq67rB+HH1Gfp16dWuXfCIWr5q82f5EKrFyewT9vHMqz4//tdChC4PaH53doVWZAvqmTXqRHu9MYfFEvWjY8nBUZS8nZuxcIDOfRs31L1qz6OezxZnz3LZ78PDauXxf2sqtCyb8Uay3/eeNVdu/a5XQo4rBdO3Yw/bVXyM7a43QoMW9/A/JZv5+CggL+2LKZxKQkfvx+CS0bHs4vK39i5fIMVv20nEnPPMkdNw7j/jGhHzCtSJ/+f+WT71bw4cLvI3JkACX/Uj6a9S5jbriWlg0PZ/GCL5wORxzUql17GqefQPqJJzkdSsyryoB8cfHx9L/iKr78ZB5er5dvFi7gmhE3M/m/HzJk2AjefO1lPpj5Nj6fL0xRg8vlonZaWtjKOxC64FvKaW3acfcj44lzxdGh89lOhyMOSkpO5vNlq4DAGeHypd9zSsvWuFyqM4Xbf9+aWuENz4v4vF7e+880vl3zG01OOJEmJ5zIqS1bF68fcdsdXDbkGowxzHzzDTp360H9o46q9ljzcnOZ8vLzDPzbtaTWrl3tr19ddMG3kLWWee+/x9nn9iQxqexwqxLbLu1xDt8sXMDk/37I2ef2dDqcmNOkdnyVbqRkjItfs/Z/kLjnHzfx+ovPcs2Im7nn349XV4jF5n/0IdcMuJBmzVsw79uMA95fF3zD7PtvFjPs8kvo2+kM3a1Lyuh6fm/qH3U0e6swHLRUv+ockG/g1ddyXJOmtGwbyI93jryBTqekszXzD/43/xPycnMPKkaPx0NeXh5ffPwR3c7vzeMvvXZQrxMuavYp5HK5OKVlazqe0y3i+uOK866+4Sb+dv2NJCUnA1X/wY5Ujy7n9uTDd2fsd5uqDsjX4rSWzF/6M3Fxcaz+aQXTJr3IEUcexWvPTeTZxx6hWfMWpJ94Er0u7k+rdmeW27X0upH/4Pim+264sn7tGrq3acHlV1/HwKuvLdHcFKnU7BPEWktebm7xf3CR0rZm/sFDd45m6x9/MO2Dj50OJ2b8uuYXep7Zcr8/uktKTmbu4qUlknJV/PDt1/y6ZjU9L+hHl1NP4M/t2wBITk3F+v14vd4S1xvi3W7cbjfPTvkPKSmpHH3MsezetZMHbh/FNwsXMPOzr2jTvsPBvVHC1+yj5C9yAPJyc2nb5Gi8Hg+LV2/isCOOcDqkmDF/3hyGD7p0v8m4a49eh1SGtZbNGzfw5MP/Ytbb0ys92BxR/0g2bVjPu/MX0vqMM/H5fMTHH1qDitr8w+SPLb8x6Zmn+HXNL06HIjVAYlIS946bwKff/6TEHwarViznm68W4PF49jsg39zFSw858UNgFN9GxzcmMTEJv9+/3229Hi+uuDgaNW7Caa3bAhxy4g+nmK35b964gWmTXuSYho24+5YRnH9hP56ftv82RZFgP/24jCkvPc/frr+RE1uc4nQ4UemOm67nzVdfYsz9DzF81O1hK/dA7vWcsfnPau3+q5p/iL0z9XWefewRPp83h4suvZxeF2usfjkw0155kWmTXqTXX1rzxEP/cjqcqNTw+MY0a96Cjud0C2u5B3Kv55r6u4+oq/kXsdZiIfCwlHk+5fXJvP7aa4wYeTN9L7q43G1s8WsVzQeW+AtXBG9b3nzRMkq9nkSHH3/4jvdnvs0LEx4FYOHP6zmmYSOHo5LqcCA1/+VbdlZr2RrS+RAZYzDA2rVrGT9+PFOmTCE7O5uUlBR69OjBuHHjuPbqvx3oq1ZbfEUHp+J5AgeO4PnSy21586UOLOUdmCy25HalD26lCi19kCrvoGWKHoUfiatwYfByU/h5uUzZfYo+SZcpuaxoeUXz+xN8sC69zB/0nv1233zw1G8t/qD1Rc8rclrrtpzWui1Lv/2G3zZtYMf2bRUmf1fhe3AFvY/g91cUa4m/SWEMxets+X+LaPTOtNeJj4/nvD4XOXKD84svu5Lpk1/Z76+Ka/q9nqO25g8wZ84cBgwYgNfrLTMuSGJiIjNnzqRXr0O/SCTRzV90FlnBf5WiHh7Bq4MPbq5q/j1AcDxFB6gSB/tyzjTLPSMtd7kts7zEWXHQvD9EB6PNGzfQrVVzPB4Pb374KX/pck4IStm/DevWcn6HVuTm5FS4zcF2La2Mav6HaO3atQwYMICcCv54eXl5DBgwgIyMDNLTq/ePJ9GlOHlX8P8xPsHNzp07mTdvHr1796Z2iMdzqSyeQ3PgL+q3toIzqZIHp+CzqxIHrqAzHICkpGT+9fhEsrOzHUn8sO9ez5V1La3uxB9OUZv8x48fX+kogF6vlwkTJvD000+HKSqJVpmZmVx++eVcccUVTJkyxelwwsplTHHTXuUq3tBay85du8jK2cGoG64rc3ZR5hpbBdfmqGBZYHk5ZzaF/5Tet+f5vfj0m6W8OPEJ3nlzCtlZ2aTWTuWSywcx9KZbaByixB+uy8dR1+yTmZnJBx98wC233EJWFS7YpKWlsXv37oMJUaTY77//zjHHHMPAgQOZOnVqje0B4rSePXsyb9483nvvPS688EKnw6mRYrLZx1rLGWecwaZNm6q8T3YVu3SJ7E+DBg3Ytm0bR+iHXwfNWsvOnYGeM+3bt3c4mugXVdUTYwwDBw6kb9++JFdxfJ7UKo4WKFIZJf5DY4zhm2++oaCggKOPPtrpcKJeVCV/gHHjxjF79myGDBmC2+3e77Zut5vBgweHKTKJBTk5Obz88suMHDnS6VBqnKImaDWZhUfUfsqjRo2qUvK/9dbw3dNTop+1ltGjRzNx4kQWLVrkdDg1xuTJk7n11lt1L40wiqo2/2Dp6enMmDGj3H7+7sKuWjNmzFA3T6lWKSkpjB07lqSkJNq0aUNWVlbIu37WdDt37uSGG24gNzeXAQMG0KlTJ6dDiglRW/MH6NWrFxkZGQwdOpS0tDRcLhdpaWkMHTqUjIwM/cBLQmL48OE0bNiQevXq0aVLF/7880+nQ4poSUlJTJs2jUceeUSJP4yirqunSCTIyMigZcuWABx11FEsWrSIJk2aOByVxAKN6inioBYtWpCZmUnXrl3JzMxk4cKFTocUkTZv3ozH43E6jJik5C8SAvHx8Rx55JHMmzcPay1XXnml0yFFnC+//JJGjRrRvHlz1q1b53Q4MUfJXySEatKdncKtS5cujB07ll9//ZWVK1c6HU7MUfIXCYPMzEy+/PJLp8OIOP/85z/ZsWMHffr0cTqUmKNqiUiIrVy5krZt21KnTh3Wrl1b5V+fxwKXy0W9evWcDiMmqeYvEmLNmzenRYsWdOjQgdzcXKfDiQivvvoql1xyCR9++KHTocQs1fxFQswYwyeffELdunWdDiVizJ8/n5kzZ9KzZ0+nQ4lZSv4iYRCc+P1+f+A2o9V8h6+a5O6776Znz5507NjR6VBilpp9RMKoT58+1KlThy1btjgdiqNOPPFErrzySho3bux0KDFLyV8kjPbu3Ut2djYrVqxwOhSJcUr+ImH00ksvsXXrVnr06OF0KI7weDycc845PPTQQ/h8PqfDiWlK/iJh1KxZM+rXr+90GI6ZP38+X3zxBdOmTdMP4BymT1/EAdu2beOll16iT58+xQPAxYLu3bsza9Ys8vLynA4l5in5izjgkUc6/+WzAAAKi0lEQVQe4fHHH2fNmjVMmjTJ6XBCbtOmTYwePZpLL72U/v37Ox2OoOQv4ojrr7+ezMxM+vbt63QoYTF79mzeeustPB6Pkn+EUPIXcUCzZs2YMmWK02GETd++fcnPz6dFixZOhyKFdDMXEZEoopu5iES4goICvv76a1544QWnQ5EYpOQv4pC8vDw6d+7MiBEj2L59u9PhhIS1lubNm9OjRw9ycnKcDkeCqM1fxCEpKSl069aNU089lVq1ajkdTkhs2rSJVatWsWPHDpKSkpwOR4Io+Ys46JlnnuHTTz/F4/Fw55134vP5GDdunNNhVZtGjRqxbt06tmzZEtMD2UUiXfAViQDffvstHTt2xOv18sknn9C9e3enQzpkH3/8McnJybRv3x632+10ODFDF3xFapAzzjiDuXPnsmzZMrp16+Z0ONVizJgxdOrUiS+++MLpUKQcavYRiRDRkvQhcM+Czp074/f76dSpk9PhSDlU8xeJIH6/n/vuu4+uXbvi8XicDueguVwunnrqKZYtW0ZiYqLT4Ug5lPxFIojL5WLGjBl8/vnnLFq0yOlwJIqp2UckwjzwwAO43W7atav0ml1E2rhxI1999RWXXHIJCQkJTocjFVDNXyTC9OvXj759+5KSkuJ0KAflrrvu4oorruDmm292OhTZD9X8RaTaWGtp1aoV27ZtY8yYMU6HI/uhfv4iEei7775jypQpdO3alQsvvNDpcKQGUT9/kRps4cKFPPHEE7z55ptOhyJRSs0+IhGob9++/PHHH1xwwQVkZWWRmJgY0b+Szc/PZ8yYMWzdupUXXniBtLQ0p0OSSqjmLxKBmjRpwtixY5k4cSJ169aN+G6fv/zyC2vWrGH69Ok8+OCDTocjVaDkLxLB6tWrhzGGr776iuHDh5OWlobL5SItLY3hw4ezdu1ap0ME4NRTT2Xq1Kk8//zz3H///U6HI1WgC74iEWzbtm3873//Y9CgQXi9Xrxeb/E6t9uN2+1mxowZ9OrVy8EoJZLogq9IFNizZw+DBg0iJyenROIH8Hq95OTkMGDAAEfPAKZMmcKMGTMcK18OjpK/SAQbP358maRfmtfrZcKECWGKqKSMjAwGDx7MZZddRkZGhiMxyMFRs49IBEtLSyMrK6tK2+3evTsMEZU1atQojj/+eEaOHOlI+VJSVZt91NVTJIJlZ2dX63ahMH78eMfKloOnZh+RCJaamlqt21WXTZs2kZmZGdYypXop+YtEsEGDBlX64y63283gwYPDFFHAPffcQ8OGDZk8eXJYy5Xqo+QvEsFGjRpVpeR/6623himigJNPPhmfz0eHDh3CWq5UHyV/kQiWnp7OjBkzSE5OLnMQcLvdJCcnM2PGDNLT08Ma15gxY1iwYAEnnXRSWMuV6qPkLxLhevXqRUZGBkOHDi3xC9+hQ4eyaNEiUlNTcaLXnu7NW7Mp+YvUAOnp6Tz99NPs3r2bgoICdu/eze23387gwYPp0qUL06dPD3kMXq+XIUOG8PnnnztysJHqpa6eIjVUWloaPXr0wBhDv379Ql7eBx98wOuvv87ixYtZuXIlxpiQlymho+QvUkPVqVOHRx99FGttWBJx+/btGTt2LA0bNsTlUqNBTae/oEgNF+rEP2fOHNLS0li0aBF33nknV111VUjLk/BQ8hep4VavXs3gwYN5+OGHQ/L66enpZGVlcdVVV1U6zpDUHEr+IjXcrl27mDJlCu+++25IXj89PZ1rrrmGOXPmRPTdxOTAqM1fpIZr06YNAwcOLO5z7/f78fv9xMcf2n/vNWvWAIHk/8orrxxynBJZlPxFarj4+PjiG71PmzaNe+65h6uvvhq/309iYiKjR48+qNd9+OGHmTRpEk899RQ33XRTdYYsEUDNPiJRJDc3l7Vr1xIfH88999zDk08+eUB98nNycujTpw8LFy7k9NNPp1atWnTu3DmEEYtTVPMXiSIDBw7ktddeY8yYMWzdupVzzjkHv99PXFxcuduvX7+e2bNnM2zYMBISEtiwYQNz5szht99+4/vvv2fAgAEce+yxYX4XEg6q+YtEkZSUFL788kuMMYwfP54LLriAuLg4PvvsM3w+HxC4L/BHH32EtZakpCTmzp3LQw89BAQGbGvRogWTJk3C5XIp8UcxJX+RKFO6339mZia9e/cuHvkzPj6e/v37M3HiRFavXs2HH35Y4sxg0aJFtGnTJqwxS/gp+YtEuQ0bNlCnTh3OPPNM/H4/v/76K7Vq1SI3N5fOnTtz2223sWXLluLta9eu7WC0Ei4hSf7GmBONMU8aYzKMMdnGmN+NMbOMMS1DUZ6IVKx9+/Y8+uijLFy4EGMMrVu3ZtiwYZx11lkAPProozz33HMORynhFpIbuBtjbgSGApOB74G6wBigFdDJWvtdZa+hG7iLiBw4p2/gPh14xgYdWYwxnwHrgZsBDQ4iIuKgkCR/a+32cpbtNsasBtR9QETEYWG74GuMOQw4FVgZrjJFRKR84fyR10TAAE9UtIExZiiBawUA2caYVQdZ1hFAmbMPkWqi75eE2qF8x46vykZVuuBrjDkX+LgKr/eFtfaccva/A3gI+Lu1dlJVAjsUxpglVbngIXIw9P2SUAvHd6yqNf+FwMlV2C6n9AJjzPUEEv/d4Uj8IiJSuSolf2ttDvDzgb64MWYw8Cww3lo79kD3FxGR0AjZBV9jTD/gVeBla+1toSqnAi+GuTyJLfp+SaiF/DsWqh95dQHmASuAmwB/0Op8a+0P1V6oiIhUWah6+3QDagFtgK9KrdsANA5RuSIiUgUhqfmLiEhki+pRPTXAnFQXY0wjY8wMY8xuY8weY8xMY8xxTscl0cEYM8AY844xZoMxJtcYs8oY87AxJmRDrEZ1zb86BpgTMcYkA8uAfOBuwAIPAsnA6dbavQ6GJ1HAGLMY2Ai8B2wGWgP3EehleZa11l/x3gdZZpQn/yOAP0sNMFeHwABzs621GmBOKmWMuRl4HDjJWrumcFkT4BdgjLX2cSfjk5rPGFPfWrut1LKrCFRcu1trP6vuMqO62cdau92WOrpZa3cDGmBODsSFwOKixA9grf2VQGeGixyLSqJG6cRf6NvCaUhyVVQn//JogDk5CKcAy8tZvgJoEeZYJHacXTgNSa6KueRPFQaYEynlMGBnOct3APXCHIvEAGPMscC/gE+stSG5q1WNSv7GmHONMbYKj88r2P8O4ArgxuBTeBGRSGGMSSVw4dcHXB2qcsI5pHN10ABz4oSdlF/Dr+iMQOSgGGOSgNlAU+Bsa+3mUJVVo5K/BpgTh6wg0O5fWgvgpzDHIlHKGOMGZgDtgPOstT+Gsrwa1exzMBweYE6iwyyggzGmadECY0xjoGPhOpFDYoxxAVMJDI1zsbV2ccjLjPJ+/hpgTg6ZMSaFwI+8ctn3I68HgNoEfuSV7WB4EgWMMc8B1wNjgfdLrd4ciuafaE/+9wH3VrB6g7W2cfiikZqscCiHCcB5BHqLfQrcYq1d72RcEh2MMeup+PaL91tr76v2MqM5+YuISPmivs1fRETKUvIXEYlBSv4iIjFIyV9EJAYp+YuIxCAlfxGRGKTkLyISg5T8RURi0P8DNzdPCvgQmzAAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe61dd57e10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Iteration: 20000, loss: 0.4363040626049042:  10%|▉         | 19990/200000 [01:59<17:55, 167.31it/s]"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe61dca7ef0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Iteration: 40000, loss: 0.5281735062599182:  20%|█▉        | 39994/200000 [03:57<15:51, 168.21it/s]"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe61c102ba8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Iteration: 60000, loss: 0.6122265458106995:  30%|██▉       | 59990/200000 [05:55<13:50, 168.62it/s]"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe61c0c9f28>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Iteration: 80000, loss: 0.6348920464515686:  40%|███▉      | 79987/200000 [07:53<11:51, 168.79it/s]"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX8AAAEDCAYAAADdpATdAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMS4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvNQv5yAAAIABJREFUeJzs3XmczdX/wPHXuXPv7KtdZE+KEJIoJbKEtIgUlbKUlMhXlFK/UqmkBclSyb6UNaSFSkUhRJSd7OuY5c7MXc7vj49lVnNnufv7+XgUZu79fM6duff9OZ/3Oed9lNYaIYQQwcXk7QYIIYTwPAn+QggRhCT4CyFEEJLgL4QQQUiCvxBCBCEJ/kIIEYTcFvyVUp2VUl8qpfYrpaxKqX+UUm8qpWLcdU4hhBCuUe6a56+UWgscABYB/wHXA68AO4CmWmunW04shBAiX+4M/qW11ieyfe1hYCrQUmv9g1tOLIQQIl9uS/tkD/zn/XH+zwruOq8QQoj8eXrA99bzf2738HmFEEJk4ra0T44TKVUB+BPYrLW+I4/H9AH6AERFRTWsVauWR9omhBCBYsOGDSe11qXze5xHgr9SKhpYDVwBNNZa/5ffcxo1aqTXr1/v7qYJIURAUUpt0Fo3yu9xZg80JAJYAlQDbnUl8AshhHAvtwZ/pZQFmA80Au7QWv/lzvMJIYRwjduCv1LKBMwAbgc6aK3XuutcQgghCsadPf9xwP3ASCBFKdUk0/f+k/SPEEJ4jzunerY7/+eLwG/Z/uvlxvMKIYTIh9t6/lrrKu46thBCiKKRqp5CCBGEJPgLIUQQkuAvhBBBSIK/EEL4EIeHSu5I8BdCCB9hd2oyPLTTiQR/IYTwEakOz51Lgr8QQviIVLtnUj4gwV8IIXyC3alJ9+DmthL8hRDCB3gy5QMS/IUQwid4MuUDEvyFEMLrPJ3yAQn+QgjhdSl2z59Tgr8QQnhZisOzKR+Q4C+EyEOGU5Nk0zg9tOI0WGU4NTYPp3zAA3v4CiH8h82pSbVDquPSStNkO5QNB5NS3m1cgPJGygck+AshzktzaI6nabL38zOccCxNUyYcQuQCUOxSPDzL5wJJ+wgR5LTWJNtyD/wXZDjhiFWT5oXcdCBLc2i89SOVnr8QQcju1KQ5wOow/nQl5ezQxh1AqTCIMssdQHFI9lKvHyT4CxFU0hyaMxlFqxx5Mt0IWHIBKBqH1l7L94MEfyGCRpJNczqjeHqacgEouiSbd88vOX8hgoDVUXyB/4KT6cZUUFFwTq1J8mLKByT4CxHw7E59sade3E5naM7JBaDAUuzg9PKPTYK/EAHuVIZ2a6A5k2GMI7iTU2usDo3N2xGzGGitSfSBC6bk/IUIYOdsxmweT5zH7tSUDFPFuhjM5jQCZaqdi9NQzUoTbVbEWkD54bqDZDtem96ZmfT8hQhQ6Y7C9cjtdlixCLq0VtQqreh2p+LA3vyfl+qAw1ZdbKWJz9k0R6zGjJjMR7RrOHt+XYK/lZ7wlV4/SM9fiIDk1JpTBczzJ56ByR8p5nwBx49e6lH/9hN0bA4fT9M0ve3yx3BoOJGuiXVCfCF75lprTqRrrPncsaT54cpjX+n1g/T8hQhIZzKgIB3Ms6ehazvFR28rjh9VVLtKM+JtJz9ucdKynSbxjKJHJ8UrgxWJZ/I/3jmbEcAL2jN3as2xtPwD/wUZTjhmNVJOvs6Xev0gwV+IgGN16AKtHN2+FXp0UuzYagT92cudfL9R07MfwG7KXdEPS2gsDkcIn0+Io3GN/sz+fLcL7YATBUjNaG3MSiropiY2DUfTNA4fTwEl+VCvHyTtI0RAcSXd43TC5g3wzRLFyiWwZ6eRMrmyimbGUk35CsbjVq1cTr/uXbDZbNhtF1YkJZGePpmh/aeyb/dchr7W7rLnSnMaF4BS+aRmrA5NYkbhd7NyaDibASXDCvd8d3Nq35sSK8FfiAByOiPvQmFaw1ez4N1XFUcOXQrE8SU0He6F517SJJQ0vrZ/z276de+CNTU1lyPZABsTxtxPaOgaBr1U/7JtSnMaA8FxFogxZx0HSLVrzmboAqWo8pJs18RawGLyvfx/ks23ev0gwV+IgJFqz7tWTEYGDOqtWPqlERivqKhp3QFad9Q0bgZmMySePcuiucvpcF8XJn4wGpstv/oDVj4cdScxcYfp/czlH+nUxnqAJBvEWEBhBOui1BjKzdkMTelw3wr+vtjrBwn+QgQEp758+YYXBxiBPyZW89Iozf3dIXsWZuhTvdn1zw5AM3vqFBx2V6qOHWPkC+kcPxrOsNc1pnxGEe3nLwLukuowpriGhfjOBSDR5lrVVE+TAV8hAsDl0gpfzoR50xThEUZOv0uPnIE/JTmZFYsXsOuf7Sz5cq6LgR+MPnxpJn04i5bX96d2uXiqxpipUz6e4c8+xf49+Q8MFzdfmlFjd/pu/SMJ/kL4OafWnMtjds+enfDSQCPSv/qupm6DrN8/c+oUb740lF9//IFftu9l9MTPaH9PZ0whIS6dOzIqkjLlygM92Lt7EinJScbmMElJzJ46hbZN6rNq5fKivLwCszrwmU1nztry3iDH2yT4C+HnrI7ci4RlZMDTjypSUxQdO2u6PJzzMZs3/sEnY95h4vujuaLildzbrQf3PNCdB3v2xmyxXPa8ZouFlu06cOLYbozERta7BbvNhjU1lX7du3j8DuB0ukZ7eepnusO79frzI8FfCD+XV1rho7cU2zYrKlXVvPGhzpLq+f2Xn1mz6jsqVqrCM88P594Hu2d5bu9nBmHJJ/hbLBZMJhMh+dwl2Gw2Jo9937UXU0xs2si1e5O7i90VlQR/IfyYzZn73PhN62H8aFBKM/oTTUzspe9prXlj+PN079iGv/5cz6CXXqVbz95Znl+5WnUmTp9LZGRkjjsAs8VCRGQk46fP5fvlS7HnMz5gt9lYMHt6oV9jYSXavFcFNM1R+DULniLBXwg/lttuUGlWeK6PwuFQ9H4Gbmia9fsOh4OWbdtzVa1radPxnhzPtygoFabofk97tmzZQt8+fYiJjUUpE9GxsXTr2ZsVazfRonU7UpKTXWpnSpJrjytuBa1vVFzO+nivH0B5Oy+Wl0aNGun169cX+HkObdQu98WFHkIUJ6fWHErVOaYRvj5MMfkjRY2rNUt/0YSH5/58rXWOwmulwlSeWzM6tVEeOtl+qfZOnfLxJCcl5dvW6JhYth5xoSiQGySEKmItnosHVodRcbSwyoQrIoowVVUptUFr3Si/xwVkz/+cl3N9QnhCqj3n/PF1a2DKWAgJ0YyemDPwr/52Benp6UDOipuXC/wAJqWINCvKhJuIOx9M7+76UL4Dw2ChYuXu+TzGfc5kFF+ZaVck+kGvHwI0+KfY/a/OtxAFlX0P2JRkGPyEQmvFU4OhXsOsj3+sc0cevac9rRrWxum8dNkIN0HZ8MsH/uziQxUJocqlgWGwsGPrID7/2OXDF7uT6Zp0D0z/tPpBrv+CgAz+GqNuthCBKsOZszTC2yMUB/cprq2r6f98zkD3yjsf0K1nb/r/bxjhZhMJoYoKEYqyESbCC5FmiLUorq9Vg/HT5xJxmYHhx/vPAarz6hDFsgUFPk2x0MDxdPcPAPtDrv+CgMz5/5eqCVFQMTIgr21CcCo9a9nmTevhnhaKkBBY/LPm2utyf54ZTalwU7GWP0ixazZs38Xkse+zYPZ0UpKSiYqJ5p6uD1G6XDm+/nIe9RqNYe4XrQgN00xbpLnx5mI7fYFYFJSLKN6tJi9ItRt7GBSV5PyLyKHxaJ5PCE9x6qw5bIfDqN2jteLxp8kS+JctmM8PK75Ga02YCcpHFm/gB4gyKxpdU4PX3vuIrYfPsDfJxtbDZ3htzFhOnzzJP39vxWKZT4/emox0Re8HFNu3FmsTXGbTRolpd3R6z/poGYe8BHRht3M2TWQB8phC+AOrI+tA7/KFsG2zosKVmgFDLwWgs6dP069HVwCuvqY2Y8aMpl2bNm5pU6RZUUbBybSss49eHvUe9Ro2olOXB9Fac+wIrFyq6NwKHu8PPZ+8VEbaU9KccMiqMSmNWRn7DIQoMCkIOf+fSRk9Y5PCpbuEFLvG5ie5/gsCNu1zQX4zGITwN8esTtLOBxqt4a7mir/+VAx6aSfHj7zHwjkzSElOJio6muiYWI4ePgTAmDFjePbZZ93atgynMc0xr7HVtDQY8qRi8TzjMxkZpeneC54ZqomOcWvTisSEUQzPpIxSdiaA839XQLqz+Or1S9qnmBibW/jmBU6IgrI79cXAD7B1E/z1pyIqehkfv1uf2VOnkJx0qbjayRPHiYiI5H//+x8PPPCA29sXalKUCzd60tnt272LBpVjuf6GD5n7jZPmrTSpKYqJHyjuvV2xf4/bm1doTozgbnMa+wanOSHNYdyFpTp8b6MWVwR88Hdqo8iTEIEg+yy22Z8rYDfpafdjtaZm2m7RYLfZsFpTGTduHCkpKR5po9mkKBuusgSXNKuVTrc1wZqaSkxcLI2bwRcLNYt/clK9pubf7Yq7blV8t8wjTRQEQfAH48qcIoO/IgBknuGTmgKL5wGMRnP5lY02m40xY8a4t3GZWEyKUpl21AqPiGDaohVMmDmfu7s+dPHrdRvAwtWalu00iWcUvbqY6NVFkeidxcCFdmDvnnxrHPmaoAj+YKR/7F4q8iREcUhzZM2lL1sISecUJtN0HPb8g/+0adPc3MKsIkKyllWo26ARbTrejdmcdZ5JTCxMnK0Z/paTmFjNd8sU97dWHDnk0ebm61xiIieOHcvx9Y9GjeT2669h/vTPcTqdnDl1ygutK7igCf5ODaf8aAGGENllrw1vpHzA6XStaFqyi0XYilOchRz5/xPHjvHeayMYP3rUxa+FhECv/vDN75qrahlpoIc6+M4dgNaabne2pGXDa3ME94qVK+N0Otm7exdj336DBlXK8uqQgT5/JxDws32yKxGqiPFgkSchikP2Im67/oFWDU1ERWsgjpTk/IurxcbGkpiY6NZ25ibJlnV/4U3rf+fu224iLiGBdf8eJDwiIsvjz56GB9opdmxTNL1VM3WhJt8KEm6mtebqkpFkZGTwxaLlpFmtALTu0Akw0j6VqlYjJTmZ6yuXockttzJ57iKcTifh4eE4nU4G9X4Ea2oqE2bOz1FXKTOZ7eMmZyX9I/xQ9rn986YbwaHDfXDPA/kXV7NYLPTo0cONLcxbjEURminS1G/UmCefe56Js74iLJeSo/ElYMp8Takyml9/VLw8yDudtf17djNn6pSL1U+/3/g3244mUqtOXfo8cG+WfY4rVa0GQFR0NF9+9zMfT5/HTTUrUatUFMeOHGbvrp0snDOTW1regcPhYOumjWRkZPDzD996LU0UdMHfCVl6IUL4g8wTFmw2+HKG8fcuD2uXd90aOHCgO5t4WSVCswbw5199g4qVqzCo9yMM6vNojsdXuBImz9WEhWtmfaZY/a2HGppJtztb8vxTfVi2YD4AV1apSlR0NMu+Mv69L4+tKa+7viFR0dHceHNzwNg17Y9ffyYuIYFKVasx4rlnuKv5jaz54Vse7tSOrm1bkHp+Jtbh/w564JUZgi7tc4Gna3wLUVhOrTmY6T393TLo1cVE9Zqa7zYY2zOuWrmcft27YLfZsGWa7mmxWLBYLMyfP5927dp5o/kXZa9HdOrECRpWLUdkVBR/HT6T63aQn7wPbw43UamqZuXvmvCIHA9xC601DauU4/SpkwDc3vZOPvp8FlHR0aSmpLBt8580bNIUkynv/vOZU6f4e8smmt52O0opnE4n6Wlp3Nm0AXt37WTLodN0bduC7X9tZvPBk9jtdto2qU+9evVYvHABERGFe7GS9snHmQzNqXQp/Sx8X/aB3jlfGJ2WLg8bgf/A3j3ExsWzau0G+vTpQ2xsLCaTidjYWPr06cOWLVu8HvgBEkKNFbIXlCxdmnc/+ZSZX3+XZw78safg6ms1B/Yqxr/rmc5a4tmzJJ07x9qdB3lh5NtEx8SQZk0jMioKgMioKG5oevNlAz9AQsmSNGvR8uJrM5lMRERGsnztJmYsWUlsXBzjvpjNhr1HiUtIIDo2lttat8VsNhc68BdE0Pb8L7AoiAtVRIS4VsNDCE87anVerBF//BjcVNN4n/72r6Z0GU27mxqwY+sWHurenekens5ZUNkHf13xx69wf2sToaGa5Ws11Wu6qXEYi9HaNqlPy3YdeOmt0YAxO8mamnIxr+8udrudj0aN5Jl+falW8YpCH0d6/i6yaWOjh4OpmuNpTpJsUg5C+A57tg3aF80Bh0NxezsoU9bYj/fhPk/SolVrxo8b572Guij74O8F7776Er0fuIeD+/fl+N4NTY27nIwMxUsDFe78eP7+688c3LeXVd8sv7jhTemyZd0e+AHMZjMDXxxB+fLl3X4ukOCfhdVhDAb/l6o5ZnVyziYzg4R3pTqy/nvhHKPXf183jdPpxGw28+BjfVj5zQpiY2O90MKCSwjNeYe9d9e/fLt0Me+88mKuzxn6f5qEEsbsn+UL3de25i1b892GbXz42fR80zr+LrBfXRGkOY1xgUNWzRGrk7MZmgy5EAgPyzzLZ9cOo3RzTJzmtjbQvWMb2t5Ynz1/b8Fs8p+UZXiIIjpbpd0Bw15m8U/raHhjU554sDPnsq1HKFEKnnvZ+FmMfFGRZi3+dmVkZOBwOKha4yrq1G9Q/CfwMRL8XZDhhESb5ohV81+qkzMZntkPVAQ3e7atGhfNVUA69RosAm1l6+aN7Nj2F1UqeCZNUJwSQo3xtgtqXlubug0aUbpsWVYsXsC7rw7P8ZxuPaFWHc2hA4px7xTvxW7blk3Uv7IU3y5dVKzH9WUBvZmLOzi0sUnMORuYlSYs0+w0rY29QjXG/y5cHsJCFJEhFPsOSiKwWTOlfLSGRfMAxrJm1WD6PtiG33f+x87tW6lSvoy3mlhoJqUoGQZH07J2oqpddTW33tGGsuVzDniGhMD/jdZ0bQvj3oUmt0CzFsXTnkkfjCY1JYWkpHPFc0A/IMG/COwaXCnfke40LhYhShNjVsRYZGaRyF9qprvLjb/Dgb2KuPgqJJ6FBo1vIjwigqaNG1+2VIAvCwtRxFmMu+oLatW5jqkL8q7r3LgZPDUYxr6jeOoRWLRaU7kYxmLfGjeJOvUb0Lp9p6IfzE9I2seDHNrY5/OIVZMmaSNxGU6tScvU818898Lc/nvZm2RnwLCXAIjy8+5bnAUseUSh0ydPsnnDHzm+PnC45va2mrOnFb27KlKLsE3BxA9G0+3OVny7dBG9nh5IXEJC4Q/mZyT4e4Fdw7E0WWQm8paa6Y7SZoOlXxp/79RVX+zpW0z+n0pUSuWY/aO1Ztc/O+jatgVj33mDs6dPZ/l+SAh88Kmm2lVG9c8Rg13/GaRZrXw1axrHjhwGICY2lt9+WsV/B/YX/cX4GQn+XpRs1xy2aqzn7wLmzJlD3759ycjI8HLLhLdlTvmsWQWnTioqVtrIiaPLSTxj1DnOPmPGX0WEKMIyRaJ9u3dxZ9MGHNy/l3sf6E58iRI5nhMTC+OnGbV/5k1TzJiS/3nsdjuP3NOeQb0fZcRzzwDQsl0H+j47mIce71tcL8dvSPD3MoeG42maoykZDBgwgH///ddj2+0J35Q95bP0SyPIJ5ScQM/7OjB7qhHp/D3lk1nmOltVa1zF+j1HWLVpB2073Zvnc2rVgdffNy6SLw1UfL88//MMfHEEV1apyvDzq3fLlCvPsNdHERsXV7QX4Ick+HvZvt270Fqzeet2ziYmsm7dOhYtWkRaWhp9+vShT58++GoJDuEeKfZLM8UyMuDbrwGc3NqqNjfcdDM3NL2ZiBAI8dOB3txEmlWW3H9sXBzlK1REa32x0uXZ06c5fvRIlufd3x2eeV7jdCoG9lacuUx1ZLPZTJNbbuXnrbuoWKmyO16GX5Hg70Xp6enc3/pWBjzWnWo1r2bjvmNcU7c+PXv2JDo6mkmTJnH99ddfzPHKRSA4ZK58+ctqOHf2MCZTAnt3rWHetz/SoHGTgEn5ZBab7TVZU1O57ooEmtaqwp9/rKN9s4b069E1xw5ZA4drbm6hOXdWMS6P4m8pycns3bXTbW33RxL8vWjrpo2cSzzLv39vw2KxEBkVxa2t2lC5WnVad7iLUqVL06BBA+bMmUO9evUYPXq0t5ss3MzqyLqwa/lCBWzA6TzHuXPGqleTgoic1Y/9XpQ565aPEZGR3N62PXfefR8/f/8thw4eoMbVtXLsAawUDH1do5Tmi0/g4L6cx375uad5/P67LtbNFxL8veKtl4dx/NhRGt54E7/t2M/oSZ+jlEIpxdW16zD8zXf5eMZ8/thzhKr1biDDZmfLli0sWbKEa6+9ln379nn7JQg3OZep4qXNBiuXAtzF7GUHePmt9wCIDFF+O7f/cpTKWfbh3U8+o3b9BuzbvZM69RvQ99n/5Sj9AFCnHtzdFTIyFO+9bhzj37+3sX3rFmw2G2t/Ws2enf9ycN9ej7wWfxD0JZ097ejhQzSpWYm6DRqx6Me1Ln2Iz505w6F//mL6Z5OZMWMGd9xxBytXrvRAa4UnpTt0lhWvP30PD3cycVUtzbfrL329bLgi3M+neObFlc9v2xvr07jZLbzy7geYTCbOnDrF2TOnMVuuouX1CpsNlv2mGfVye35d/QOT5y3ipuYt2Lz+dxrd1MxDr6TwZA/fAJWclMQdHe6iao2rXO69xSYkcE2T5gx97S26d+/OK6+84t5GCq/IvNIVYOzbnwHNqFFr3sWvhSgCNvCDMYh9uVlMWmsO/XeA+TOmXszh/7L6e1rfcB3HDv/CQ71A6394ZfBuoqJjCA0L45o69bBYLH4R+D0pgCaL+bYjh/4joWQpalxdi0mzFxRq8DamzBV8NOULtm9YywsvvECbNm249dZbycjIIDQ01A2tFp6S4dRZavnY7fDXxkNAGlddfWkKS1QADvRmF2VWWaqZZqaU4uPpc7FYQqle82rAmDFXqWp1Fs+bzW8/9Qe2sG7NI8xZMZsPPrXnGCMQBvmpeMjQ/n1Y/9svTJq94OKenoWRaNN8vXwFb775JuvWrSMjI4NatWoxadKkYm6x8KRz2Xr9634Gq/UBTCEf0e2xDhe/HhmAA73ZRYQoQpQmrwooN7doleXfvQc8x7bNf/LFxPGZvrqGV4fYWPzT5Te2D2YS/D3Abrdjt9lJT0vj6trXFfl4TVq2ZZjdRo0aNXj88ceJi4vD6XQG/OYTgcrm1Fn26V25dBEzPw0HmvDYkzsoX8GoN2NR/l/OwVVRZpXjgpiXsLAwPp4xj0/ef5czp0/RukM3+j98DX9vCePTcU76DHBzY/2UDPh60N5dO6la46piOVZCqCLWohg+fDiDBw8mPj6+WI4rPO9Uus4yt/+OG+qyc/s2YC3LfruBa8/3F+IsivhcdsEKRBlOowBiYa1aCT3vNREeoflmXfFU/vQUGfANAHa7nfff+D8WzJ4OUGyBH4z0j1NrXn/9dQn8fszuzBr4tdZcfW074FYqVa3LNXUuPTYyiO7TQ00qy2YvBdWiNXTqokmzKoY97d59f/2VBH83+vqrubz/xqtMn/xJsR/bqbmYKtBa07dvX+bNm3f5Jwmfk5RtPwilFAkl3gZW0/7ecC4MDZmVERCDSWQRB7dfHnVp399504qpUQFEgr8bNbixKdc3bsLdXR50y/EvzIhYsGABEydOpEuXLiQnJ7vlXKL4ObUm+Xxee8PaX/n1x1U4HLBisfH99vdc6q4WNRD6o6Le6ZQsDS+/bfwMXx2i2PVPMTQqgEjwd4Mzp04x7Okn2Ll9Gwt++IUefZ50y3nSnUbaoE2bNvTs2ZMVK1YQHR3tlnOJ4pdsByeQePYsfR/szIPtW/H2iC85cUxRuZqmdr1Ljw2GWT7ZFTX1A8aq346dNSnJin7di7bxS6CR4O8GP32/klmfTWLK2A/cfi6rA6Kiovj000+pUqUKjzzyCIMGDXL7eUXRaK0vzWbRmu69nqBdp3tZv/ZuAB59QmdJ+QTLLJ/sIop4x6MUvDVWU72msfHLC89I/v8Cme1TzGZ/PpljR45gMpmoUr0GHTt3dev5wk1QNsK4hm/bto06depQsmRJjhw5gsUic5x9VYpdczI96/t0wzq4r6WJ2HjNbzs0Uedv4mItOXe7ChbZS14U1s7t0Ok2RWqK4rUxTnr0LobGuYnM9vFDRw8fYmj/vowZ+QrdevZ2e+AHSDuf+gGoXbs2o0ePZv369RL4fdw5m2brpo1079j64tcmf2h84Lv34mLgh+BM+VwQFqIojpueq66BNz8yPicjX1Ds2130Y/q7IJo85n5R0TE0b9Wa6jVrUapMGY+dN9UBsecv45Ly8X1p58s2f/ze22ze8AeHDh7AbqvEisVgsWge6XuppxsSxCmfCyJCVJbpsIXVqQusXqlZMFvxTE+Y+bUmOqYYGuinpOdfjGJiY/li4XJGvD3Go+fN7YNx7tw53nnnHRwORy7PEN6UdD7Xf931DYmJjUMpxZSxCq0VnbpC2fKXHhsMtXzyU5x7F4x4W1OxsmbLRkWrhorlCwnaMQAJ/gHA5jRWRGY2cuRIhgwZwptvvumlVoncOLQm9fz1+ImB/+OX7XuJiLiSuefnofd+OuvvMRA3bSmo8GL8GcSXgKkLNPUaao4eVjzZ3UTP+xQHgrDMvwT/YvLbT6t5dchA9u3e5ZXzp2RbLNS4cWMqV65M8+bNvdIekbvsvyelFDOmQJpVcesdmqtrX/peoJdvdpVJKcKKMVJVrwlf/aB5bYyTmDjN6pWKO25QjB9tbKATLCT4F5N50z/ns/EfsmjuLK+cP3sJ3Pvuu49//vlHgr+PSbFr5kydQq+ud/PbT6tJT4fPJxgBvs8z2Xv9EvgvKO6fRUgI9OgNP2zU3N1Vk56meHuEiQ7NFBvWFuupfJYE/yJKs1rZ/e8/dOnRk97PDKLDfV280g6HNgYSMwsLC/NKW0TubE5joPe//fv57uslHDl0kCXz4eRxRa06mqa3ZX18MNXyyU9xpn4yK10W3p+imbbYSeVqmn/+VtzXysSwpxWJZ9xzTl8hwb+nqfCyAAAgAElEQVSIVi5dRMsG1/LVrGm8+MY7FzeY8IbsKQWAvXv38umnn3q+MSKHsR9/wvTJEyh7xRXc3KIlLdt2ZMpYo0f7eP9Li7rA2KQ9XD6dF4WFKNxZ2uiW2+GbdZqnh2gsFs2szxQtGygWzQ3cAWF5exXR6ZMniYmNpXbd673dFFLtOssOYXa7nQYNGvD444+zc+dOL7ZM2O12lixexPBnn6JMufJMX7KSv/+KZ/tfilKlNR07Z318RAgBuUl7Ubh78Ds8Ap57WbPsV03jppqTJxQDHjPxcCfF/j3Fc460NDh+DA4dBKezeI5ZWBL8i+jRJ/vzx+7DdHm4p7ebghMuziQBMJvNPPDAA3Tt2hW7PZfbAuExDhXC/T16Yjab+X75UgCmjDOC+0O9IDw86+MjJd+fg6fGQK66Bmav0Lw93klcgubnHxStGyvGvQMZGQU/3u5/YcxIxe3XK2qVMtG4uolm15iod6XioQ6KUSMUKxbBkUOevcuQ8g4BJiIEyuSSL7Db7ezYsYNy5cpRqlQpL7QsuJ3JMGr5nD19mpi4OA7uC6FFfYXFAr9s15Que+mxJqBipJKefzZOrTno4c/2yePwxouKr2YZv4uramlGfqBpnM9e8P/+DRPGKP78A/buuvR7NJs1cfFGzaGTJ3L+fkuX1TRuDC8MVTRtWrg2S3kHD0hJTi7URuzuZHUYF8DsHn30Ua677joWLVrkhVYFt8TERFas/BabzUZ8iRKEhITw+QRjUdddXcgS+MFY2CWBPyeTUh4fBylVBt6bpJm51EnVGpqdOxRd2pgY2l9x9nTOx9ts8NEoaN/MuGDs3aWIidV07q75YpGTHSc1G/Zp1u/VrNvpZOJsJ/3/p7nldk1svObEMcXXSxRpae5/bRL8i+Dx+ztxQ/UK/PrjKm83JYu0XBb11q9fnypVqsiKXy9YsvwbHuzYhp73tgfgXCIXNxd57KmcF+oYKcuUp6JW+SysprfB8rWaZ4ZqQkM1sz83BoQXzL6Uqtm2BTrdqhj9mgmbTdHtUc3SX5xs2Kd5d4KmeUswZ5rBVbY8tO4Ag0dopi3WbD6oWbXJyWfTNDfc4P7X5NbJZEqpK4ExwB2AAr4DntVaH3DneT0hOSmJ40ePkJqSzHXXN/R2c7JItescZQEGDRrE4MGDvdSi4KW15ly6sXKobad7AZj7BaQkK25qri/uz3tBeAhYgmzHroKICAFvzcAMD4dBwzV33Q8vDoB1axQDeynmT9dc1wAmfwh2u6JiZc2osZpmLQp2fKWgag24sY5nVna7LeevlIoENgPpwHBAA68DkUBdrfVlt1Xwh5y/1po9O//16vTO3EjO2PucWnPg2Gk2/72d2g1uYMuGP7i+cROUMtGygWLPTsWkOU7uaJ/1eWXDlazqzccRq5MML8+U0Rq+nAkjhynOnL70+3qkr2bIqzpLVdaCCoSSzr2BasDdWuuFWutFwF1AZaCvG8/rdrt2bOfgPqMYiK8FfjBm/aTl8eFITU2lbdu2HDp0yKNtChYOrTmboTmUqpk4ZQp3t7yF154fRMMmTTGZTPz1J+zZaUzvbNEm63NDTVLOwRW+sPJZKej8EHy/UfPAo5rrG2vmrHDy6uiiBX5PcmfwvwtYq7W+WOxGa70X+AXo5Mbzut1bI4ZxS50aLJk/x9tNyVNqHjM7X3zxRb755htmzfJOGYpA5dCaM+eDfqJN4wSiY2KIL1GC21q3vfi4hXOMwNWhc9b8Lxibtoj8RfvQyucSpYydwhb8oLnxZm+3pmDc+WOsDeQ2tWQbcL8bz+t2be+6h5++/YZrr6uX/4O9xOrQGMMsWd1+++3YbDYaNvStcQp/5tCaI1ZNtuoaPPR4Xzp1eZDIqCgA7HZYPM/43t1dsz44RAX3pi0FYTYposw61xXtwnXu7PmXIPexmdNAQm5PUEr1UUqtV0qtP3HihBubVjSdH3qEhT+upfrVtbzdlDw5tLEFXnYdO3bko48+YtOmTUyYMMELLQs8p9NzBv4LomNiMJmMj9mvPxp1fKrW0NTLdu2NtcgYTUHEW1QuXRtRED411VNrPVFr3Uhr3ah06dLebs5lXXtdPZ//sFrzmNU5f/58Bg0axPbt2z3boACUatdZVlUD9Op6Nx+NGpnjsQtnG++XTl2y1fHBt1IZ/sBsUpImKyJ3Bv8z5N7Dz+uOwKdprXn31ZdYMn8OKcnJ3m6OS1Lz2PquXr161KtXjzIe3GoyEDm15nRG1p/xgb17WPfzj4wZ+Qq7/tlx8eupKfDNEuPvd2fb2jnaojD5eEfCF8VaKJb9fYOVO/sb2zDy/tldC/ztxvO6xZaN6xn7zhsADBj2MgNfHOHlFuXPpo0ywtnnjdesWZNNmzb53Opkf5NkI0e6p0z5K3hi0PNUr3k1NTKlBb/92pjbX/8GTZXqWZ8TI73+QjEpRZyFHBdg4Rp3vu0WA+8qpapprfcAKKWqAM2Aoe466cjXoUItuKk5xTrlql7DG1j77wFWr1zO/T28X8TNVVYHWPK4vzt9+jRr1qyhdOnSNC1sIZEg5dBGrZ7MtNaEh4fz1OCcb+9Fc40L8D3ZBnojQ4wUhiicGIsixa5J9/K8f3/kzrTPJGAfsEgp1UkpdRfG7J+DwCfuOOGhQzDiZUWvLibqX6l4rLNi8lhYsRj++hOSk4p2/HJXVOCBR3sREuI/0zKseaR+wMj933333YwbN86DLQoM52zGeorMXh0ykJtrV+erWdOyfP3UCfjpOwgJ0bS/N+tzYiRvXWQlw5RvDV76Cbf1/LXWKUqp2zHKO0zDmHf4PUZ5B7ckzc1meOllzdIVsHk9/LBC8cOKrBX1br4duvXUtGyXc551XrZu2khCyVKUr1Dx4swNf5HmNHLTueWUmzVrRosWLWjUKN/FgCITmzNrr//k8eNERUcTHR3Df/v3ERsXn+XxX39lLPtv0UZTKtMwi0X26C0WFpOiRBicTJf0T0EEbEnnk8dh5VLYsVVx6CD8tx92/QMOh/FhK1NOc+sdUO0qTdXqRk2NytVz1lUHaHZNVQ4dPMAPf26n2lU1i/rSPK5UmMpR60cU3ok0Z5YZPs8/1Zv506cyavwkHHY7XR5+LMtMsHtbKjauU3zwqZNOmXb5TAiVGSvF6ViaM9eihv7GU+UdAnaoqVQZePAxMEoKGU6fhK9maWZ+atRWMSorZr0zaHob3Hm3pnUHY/Wew+GgYuWq2O12rqxS1cOvonhYHTkLvYnCSXPknNqZlJiI1pqq1a+iYZOsYyf7dsPGdYrIKJ2ljo8CogL20+cdCaGKI1bf7Mz6ooDt+V+O1vDnH/D3FmOjhb27YO8u2L8HnE4jSIaEaJrcAu3u1rTpmLPmuj/Jr9DbqVOnmDBhAgMGDCA62k8Kk3iB1pojaRpbLoOLqSkphIaFYc6US0xOgiFPKpYtVHTubpT1vSDKDKXC/CuF6A+y35X5I+n5u5FS0KCx8V/2O4Nvv9YsW6j4ZRX8slrxy+oTvDSwNI2bGXcEbTsZdbhdYU2Ff7dD+QpQppxbXopLnBizfiLz+G137dqV77//nsjISAYOHOjRtvmTZDu5Bn7gYgmHjAxjcHfhHMV3yyDNqggN1QwYmrVDEiN3Ym4RH6pIld6/S4Iy+OelRCno+gh0fUSTeAamfrKJD0e1QnGIdWvCWbdGMWIw3HCT5oGexsyNC2MEdjvs3GEMNG9er9i0wdjK7cIYQ71Gmjvu1LS6E66uDZ5e05Ni10TmEXAGDBhAaGgoN954o2cb5UecWnM20yDv2dOn6fPAvfTo8yQdO3dlwzr4coZi2UI4m6nEb+OmmicHa66sculYFhOEyUCvW1ik7o/LgjLt46r/Pfk486Z9zmtjPiUm9hGWLVCs/hYy0o0PbkIJTcs7Yf9u2LoZrKlZP9AhIZqqNeDgfkhPu/S9Jrdohr2es76Lu1WIUDKnvJDOZhjVOi/YsnE9j3W+i5PHj3FD00/449c+F79Xq7amU1dNx85QsVLOY8lAr3vZnJrDftz791TaR4L/ZRzct5eZn02i68OPUaV6DcDI4379FXwxUbFtc9ZfUKWqRkCv10hTrwHUrgeRUUb6Z80q+G6ZYvkiOHfWeF7Hzprn/0/nGiDcIc6iiA+VoFNQDq05nKpzzOv/78B+Rr28miXzuxMaZqZnP2MRV606lz9exUhFiJRzcKvT6Zqky6xx8WUS/H0g+F+O1rDxd9i4DmrUgnoNjLRRfhLPwPj3FJ+NN+4gwsI1vZ+GJ59z/yYQIcro/ec18PvXX38xdepUnnnmGSpV8tAVyQ+cStck5xJItm+FTs0VGRmKT2Y5adMx/2OFh0BZT+9CHoScWnPIqnH6Zni7rEDYyctvbd+6hcQzl689pxQ0vBF6PwMtWrsW+AHiEmDYa5pVmzR33a9JT1OMfUfRor5izlQ4duTShtDFzaG5bC505MiRjB49mhkzZrinAX7I5swa+F8dMpC3X3mRQwePMbCXEfgffEy7FPgBIiXX7xEmpSghd7mXJQO+uZg28WP+3b6Njz6fSfkKFd1yjgpXwoefaR7pq/m/oYrN6xXPP2W8WWPiNBUqQnwJCA2D0FCwWMASCuERcFNzTdu7ICKy4Oc9Z9NEmcm199+7d29KlChB27Ztc3lmcDqTqWhY0rlzzJjyCbaMDJIS+7Jjq6JSVc2Lb7h+tc5rxpUoflFmRbJN57mlabCTtE8uura9nXVrfmTqwmXc2qpN/k8oIqcTFs2FmVMU/+6AxDP591hKldZMmKlpdFPBz1cyVBEtA475SnNojqVlfS9tWv87sz5dwbzpRlXXud+4/juQlI/n+ePgr+T8vRj8ramp7N29k0pVqhEdE1OkYxWU1kYhsGNH4ewpsNnBlnH+PxucPAELZim2blKEhmnGTjVWIxdEfrn/EydO8Pbbb3PkyBGmT59eDK/KPx21Okl3Qnp6OmFhYQCkJMOdTRX79yiefE7z/Kuuv9dKhCop5OYF2Wdq+ToJ/l4K/mdPn8YUEkJsXFyhj+Fudju8MlgxfbIxYPzJLM1tdxTsGJeb+ZOYmEjZsmWx2WwcO3aMUqVcHNAIIKl2zYnzhcLeeHEIqSkpjHjnfV4ZHMqMKYpadTSLftScvya4RKbaeofWmqNpmgw/Sf/IgK8XnEtMpFWjOvTr7tv7y5vN8NoYTbeexoDxo/eYGP6sIumc68dItOk8d/qKi4vj9ddf5/fffw/KwA9k6Sl+t2wJ0ydPYOonvzNjirFid8zkggX+MJPU7fcWpRSlwmTP3+xk+CmTP379mZPHj2Gr6bsbs1+gFIz8QFPhSs37bxh3AWtWwbhpmtp1XTvGiXRNjAMizEahMZvTKAJnMSmeGvgcwZqeTrFf6iVqrRk1bhI/rPiOT8c1AWDAC5pr8pnLn11eq6uFZ1hMioRQ2fUrsyD9eOeuZbsO/LJ9LyPeGePtprjEZIL+Q2DpGs0112n27Vbce7viy5muHyPJrjmeZgxsns7QWB3GjKDjacatstNH04LusmvXLvr160ed8vFUjTFz3RUJLJo7i2NHH+bwQQt16mv6Plvw40b6z/4/ASvGogiX38NFkvMPEGlWGDFYMWeqQinNZ18VfBwgu3VrfmLmlAm0aH4zA/r3L56G+rDly5fTuXNnMmw27Dbbxa+HhFhwOCyYTPNYsqaty3dWF4SaoHyE9LN8gT/M/pGcvwclnjnDkH69WLPqO283pdDCI2DUOM2AYRqtFQMeUxzcV7Rjnjh2lEXz5jD63feIiY3FZDIRGxtLv3792L17d7G021fs3r2bzp07k5qamiXwAzgcNiAVk+l+oqML/rqlpIbvsJiUVFQ9T4L/eVs2rKd7xzZ8v3ypt5tSJAOGaVq20ySeUTzxkCLNWvhj2e0OzGYzRw79R3JSElprkpKSmDx5MnXr1mX58uX5HkNrja/eXWY2evRobNmCfg7KxuSx77t0vDATxFoUpcOK1osTxS8+FCT+S/AHIDY+nhZt7yQqOpr4hBLebk6RmEzw3iRN5WqabZsVLw0q3Lt8/57dDHu6D3a7Hbs9a1C02WykpqbSuXPnHHcATq05k6E5anXyX6qTA6maA6ma/SlODqQYXzuc6uSY1cmpdE2K3TfGFb744ot8g7/dZmPB7NzXPYSZjGqdZcMVlSIV5SJMJIQqGej1QSaZ/QNI8AeMqWDPv/oGfx0+k2MbPn8UFw8TZmrCIzTzpikWzS34MSZ9+F6+wdBmszFmzKXBcafWnEw3NjdPdxq1hDLTGF+zaWNj+WS78fhD1qwbonva8uXLSUlJcemxKUnJOb4WYzaCfaxFER6S9+I54TvCQowLQDAL+uC/Z+e/TJ0wjk3rf8dkCpwfxzV14OVRRkAdPtDYxL4gFs6ZkSP3nZ3NZmPqtGkcS3NyPM3JYasxW6ignNqooXPM6sTm4TKMF3L9roqKyVp6NcoMJYI8iPirSHNwF38LnGhXSL/9tIoRg59hxpRPvN2UYtetJ7S6U5OUqHiuj8JZgBWOKck5e7i5Pi4pmTSHsU1k9p5+QaU54bBVczLdcxeBd0ePJiO/XP95ZouFex7ofvHf4SajTpLwXzGW4N1YJ+iDf7Wrrqbbo724uUVLbzel2CkFb43VlCqtWfuzYlBvhYtxjigXN3LP3hMuDil24yJwPM1JelGvKPmYMX16vnc4F1gsFnr1Nyb5W0xQOlxSPIEgIVQF5TqMoA/+NzW/jTfHfkKnLg96uyluUaoMjP1CExWtWThH0bvr5WcAOZ1GYbkWbR/CbLZc9tjZe8LFzeqAo2nGRcDhhkFhm1OT7OIdDsD46XOpXK06IQrKhClMEvgDRqkwRViQRcMge7lZ/fbTar+YhlhUTW6BmV9rSpTUrF6peOQeRXJS1sekpcF7rykaVFE0rGpiybzB2O2XD/6Ze8LuZHXAUavGXoypILvTWMUcGRXl0uOjoqNp0bodCigdJgXaAo1SitLhKqimgAZt8P9l1fd0u7MlfbvdFxQXgHoNYc4KTdnymnVrFA+2V+zcDmdPwyfvwy21FR+OUpw9rYhL0FStUQ2TaR4QCeS8CISGhV3sCXuCXRt3AcUxFpDuMAK/XePSHZ/ZYuHeBx8GzvcQZd5+QApRijLhimC5rgdt8D9y+D9KlSnLtXXrB03e9qprYN5KTcXKmi0bFXfcYKJ+JRNvDjdx4pjimus0M792svmgsc3k/O/aUu6KzUAfIBYwERoaTr2GNzB7+Q+0aN3Oo+13aDiWpsko5AUgxa45lOo0LiLnD1GxUqV8f/8X7nBk3n7gs5iMhXnBIKhr+zgcDmwZGYRHRBT4PP7s9El4+xXF8kWQmmykhR5/WnNrK2OQOLM0K3w6Hsa/q0hOUsTEaj5foGl4o2vncjiMu4vEs5c2pHE4jO0pS5aC0mVznjM/IQrKhbueenFozan0rNNQtdYopThz6hQvPPME3y1bjs1mBy4N/potFiwWC+Onz6Vdu3ayC1cQSbZpTnmpAqhs5iKF3TxCa9eC78njMPxZxYrFioQSmgWrNFVyyfisWAwzP1UcOmBcZM6eAa3zPkFMrKZ5K7i/R+4Xn7yYFZR14QKQYjeqlWa/WRj29BMcO3KI/70ykvJX1KVF/T2cOf0+YeHTyEhPJiommnse6E6v/s9StXp1yhfgYiMCw5kM7yw+lODvpuB/9vRpfvjmazp2fgCL5fIDmiIrux36PKD4YUU6pcrMoU793+n/vw+pea0iJtbYh3hgL5Uj2MeX0MQnGBvRmy0QEgIZ6XD8KJw5femxt7XWjHhbU7WGa+2xmIzBV0suQdmhNafTNamZevt2u51JH75Hq3YdGPB4D/7esonvNmzj4/eu4csZisZNNbNXaLKv9SsVpoiSdE9QOpHmzPIe8gQJ/m4K/s/0fIjF82bTsl17psxbXJQmBqXkJLivVRr/bKsMnARWoFRrwiPAmmq8YR99QvPgY5qEkpBQ0th5LC8H98PieTBhjCIp0dgl69EnodfTmjLl8m+PCWOeduYN6VPtxi179t7+8oVf8sXEj3n6+RepVfs6vpg4nusbv8wjd5sIDdOsWKupdlXW50SEQBlJ9wQtp9bnx5k8d04p6ewmrTt04qbmLXh5lH9s2OJromNg7jfhNG42ikpVx1OnfgvMZiPwlymnGf6WkxHvaGpea+TzLxf4Aa6sDE8NhlV/au7vocnIUEz8QNHsGsXgJxQ7tl7++U7gVIbmWJpRKO6I1cmJ9KyBX2vN2p9/5IM3X+O3n1ax9qfVlChVil5Pv8wLTxsfsoEv5gz8Jgjq5f/CKAJXOkwRiBO8gq7nL4pferqRxomOKfjgbXabN8DHoxXfLLk0VnBLS82w1zXXXle4Y04e+z6vD30OgEeffJr7ezxK7br1eWWw4vMJijr1NQtX6xwXqhKhipggXfovskp3GHcAnogs0vMXfmH71i08dl9rXh70cJEDPxjrESbM1KzerHmkryYiUvPz98b2lIWpTgrQ7dFetO7QibFTZ/HKO+9Tu259ln4JUz8Bs1kzanzOwB9mQgK/uCgQq4AGTfB/7/VXGNKvF//+vc3bTQkoFksov6z+npKly7B/z26GP/vUxf1v65SPZ/izT7F/T8F3v6pcDV4drflth5EOSrMqBjxm4qNRxgylgoiKjmbi7K/ocF8XkpNg8lgY8JgxMD1weO4b3pcMsA+6KLpIsyIugDoEQRP8Txw9wtwvPiM11bW67cEuROFSrZPqNa+m78D/kZqSQtsm9Zk9dcrFXb+Sk5KYPXUKbZvUZ9XK/Hf9yk18CXh7vGbE206U0ox+zcQLzyjsdteeP3/GVGw2G+cS4bWhiiY1Fa8PNeFwKJ4eoun3XM7nxFlyn0EkRHyoIiqfcSx/ETQ5/7//2szQp/qwcPVvAVW3v7iZlRH8osxGvZM0h7Ez1+VmO+zfs5u2TepjTU3N8zERkZGsWLupSOUgViw2euzpaYp7H9S8OyHntMzMVn2zjJ73deTGmx9i764vOH7UeHDjppqe/TRtO+UcozAruCJCqnWKvLl7BpDk/IvZNXXqSuDPR7gJykcY0yYvBL/wEEX5CBMlQ1WebxZXd/1ydf/bvLS9C2Ys0URGab6aqXjjRZUjBfTXnxsY0q8XK5cuwuksRVj4Faxbk8Txoyaub6xZ+ouTuSs17e7OfXA6IVQCv7i8QJkBFNCRcNeO7Tz50P0sWzAfpZQE/suICDHq0+dVpjjaoigfobDk8iN0Zdevy+1/WxCNbjIGhC0WzeSPFG+9pMjIuPT9X1b/wNwvPmPyhysZ8NiNpKcdICp6Af832sn8bzV16uV97Giz1O4RrjGfrwHkz++WgI6GE95/h+WLvuLn77/1dlN8WoxZUSbclG99erNJUSaXHk9Bdv2Ktxhb55UMNf6MtSgKuoaqeUsYPVED5/jk/fHc1TyVDeuMGtXVr+pIxcrv8fuv3UhOUrRqb+K7DfBwX2NlcV4sChJCC9YOEdzCQpRfTwwIkKGL3L323ljKX1GRR5982ttN8VlxFkV8ARYyGT0essx5joqOJjkp6bLPA4iJiSYu13MpMpxG8TVX86h33Q9fznyGH7+dyo6t/bmvJUREvog19VXgWsqU0/zfaCdt7nJt7UFJ2ZxFFEKUWZHhxCs1gIoqoHv+EZGRPPfy/1GydGlvN8UnWJSR2og73/uuEFGwwH9BWIiibLhxnDiLonO3hzDnUyfJYrHQo0ePPL8faro0thDqwrty6ZdzubKyUY21ctX7gDpYU9cQGubg0Sc0367PfUA3N3EWqdEvCi8hVBHth+nCgJzt06JVa96Z8Cnlrqjghpb5H5MyUjtxFtwymLl7927q1q1L6uVm+0RE0KRJE1q1asULL7yQ7zHTHZokuybl/JTOX1Z9z+efjKXu9Y04/N9BZn02CYCN+46RULIkf285QkRkKSpUCiUszPW2h5mgXERA94GEh6Q7jAqyRZ0FJLN9CumPP/7g5x++o/MdzcnIPBIYhEKU0Su50MN31yyW6tWrM3/+fCIjI3NUSrVYLERGRjJ06FBWrVpFhQquXZCNFZUmKkYadxe7//2Hb5cuZuw7b1wM/O3vvZ8SpUqhlKJ2vSuodlXBAr8JAm7VpvCesBBFuXBFvJ8sBAu44A9wR4e7GPH2GEJDg3cEL8asuCLCGFD1RC67Xbt2bNmyhT59+hAbG4vJZCI2NpY+ffqwZcsWQkNDadq0KbfffrtLxzt69Chaa0KUIv3McerWrMr/vfsBX/+ygb4D/8fkuQsZ98XsQrdXYcxukhr9ojgppYgLNT57rqQvvSkg0z7BXNhNYfRmfX3K4oWdtP744w8aNGiAUooVK1Zw5513snXrVurWrUuzZs346quvqFOnDsePH2fosGE8+9LrpBXxtjrUZBRtkzy/cCeH1hyxahwFDEeS9hGFUjrctwO/1pohQ4ZQsmRJdu7cSePGjalVqxZdu3alffv2fPzxx9SpU4dWrVqxZcsW4uPjGTlyJBUqVODBbt0oE174wbUYs5ECKx9hksAv3C5E+XYxOAn+ASTeUrQegyccO3aMSZMmcebMGWbOnEn16tVp2LAh7du3JyoqikaNjA7LrFmzmDt3rrF5eq9e/Pfff1x33XUopSgRCpGXmbOf3YWVyyXCJM0jPCs8xFjT4osk7RMg/GnHqR07dvDXX39x3333YTKZsFqtREREcPjwYa644gqXjqG15kS2TdmzuzDgLVswCm8ryH7Ankr7BPQir2BhUv6141StWrWoVavWxX9HRBjz9V0N/GAMrJUOg7O23BfYxFqMqa2ycEv4goRQYyFjmof3A74cCf4BoERocKYzlFIkhEKU2ehZObQxmCslmYUvKh2mOFyIAWB3keDv56LMBH1aI9RkrDgWwpeZzndWTqb7RvT3jySxyMGkjAHeUq7suCKE8AlRZt+Z/x/0Pf/MOzvn2ToAAAbGSURBVFY5NUWeQ+4JCeerYQoh/E9CqOJYmvd7/0Eb/LPvWHWBUxszSFLtxuCML10LTBjz+MN9fDqnECJv4SGKMJMm3cvBJeiCv8II+rF5FDkzKXUxj64vXAgcGqu94BeCKDNEhigcGtIcxi+7sIM9oSZj5a4MZArh/2IsinQv5/6DKvgXNIAqpYg0Q6RZoUONC0GKXZOaz9zyGLMi2mKs8Lsg5nyaJt2hSbZDst31X3ysRRHvpoqcQgjPiwwxxu2cXoz/QRP8w02X36YwP5kvBDansal55gVGESGubQMYFqIIC4E4CyTajItJXr//UJORH5Q0jxCBRSlFVAgkFaATWNyCIvhbVNECf47jmRRlwhXpDuMCEGWmwOkYs0lRMsy4CCTZIcl26SIQbjL2zA32KZxCBLLo8599bwn44K+AUsUY+DO70IsvCrPJmPsba4FUO4SHFPxCIoTwP6EmhcWksXlp4NdHZpy6T0KoItQPgmmIUsTIylQhgkqkF1O6AR38I0IuDbQKIYSvifRi7iVgg7/Cv4qdCSGCT6hJ4a2hvYAN/vFBWuxMCOFfvLX5UkAGf4uCmIAfyhZCBIKIIk4aKayADP7xoUoWRAkh/EKYyVjw5WkBF/xDlG/vYSuEEJkppbzS+w+44C+EEP7GG3tvS/AXQggvk56/EEIEIZNShHs4GkvwF0IIHxDh4bFKCf5CCOEDPJ36keAvhBA+wGJSeLIajQR/IYTwEZ5M/UjwF0IIH+HJ1I8EfyGE8BFhJs8FZamAI4QQPkKpom8Q5Srp+QshRBCS4C+EEEFIgr8QQgQhCf5CCBGEJPgLIUQQkuAvhBBBSIK/EEIEIQn+QggRhCT4CyFEEJLgL4QQQUiCvxBCBCEJ/kIIEYQk+AshRBCS4C+EEEFIgr8QQgQhCf5CCBGEJPgLIUQQkuAvhBBBSIK/EEIEIbcEf6VUTaXUB0qpLUqpZKXUEaXUYqVUPXecTwghRMG4q+ffGmgBTAU6Av2A0sBapVRDN51TCCGEi8xuOu5sYJzWWl/4glLqB2AfMAB42E3nFUII4QK3BH+t9clcvpaolPoXqOCOcwohhHCdxwZ8lVIlgDrAdk+dUwghRO7clfbJzUeAAt7P6wFKqT5An/P/TFZK/VPIc5UCctx9CFFM5P0l3K0o77HKrjxIZUrL5/0gpVoB37pwvB+11rfl8vxhwBvA41rrT11pWFEopdZrrRu5+zwiOMn7S7ibJ95jrvb8fwWuceFxqdm/oJR6AiPwD/dE4BdCCJE/l4K/1joV2FHQgyulegDjgdFa65EFfb4QQgj3cNuAr1LqHuAzYLLWerC7zpOHiR4+nwgu8v4S7ub295hLOf8CH1Sp5sBKYBvwNODM9O10rfWfxX5SIYQQLnPXbJ/bgTCgAfBLtu/tB6q46bxCCCFc4JaevxBCCN8W0FU9pcCcKC5KqSuVUvOVUolKqXNKqa+UUpW83S4RGJRSnZVSXyql9iulrEqpf5RSbyqlYtx2zkDu+Sul+mMsGpsKbATigSFAfeBmrfUGLzZP+AmlVCSwGUgHhgMaeB2IBOpqrVO82DwRAJRSa4EDwCLgP+B64BWMWZZNtdbOvJ9dyHMGePAvBZzKVmAuDqPA3BKttRSY+//27tg1iiiK4vDvCBaiFgpWNjGNEFsLwSKFplVrC8HOQtBKsDIqtrH0L7CzUksVy4B2GhsDRkinGATBRjgWM0EZZzFudnbcN+eDZeENy73V2Zm3M3fjryRdA1aA47bX67VjwHvghu2VPvuL2SfpiO1PjbVLVCeuZ2y/mHTNord9bH9249vN9lcgA+biX5wDVreDH8D2B6qbGc731lUUoxn8tVf1eydZVXT4t8mAuRjDCeBty/oasDDlXmI4Fuv3TrJqcOHPDgbMRTQcBrZa1r8Ah6bcSwyApKPAHeCZ7ddd1Jip8Jd0VpJ38Ho54vM3gYvA1d8v4SMi/heSDlD98PsDuNxVnWmOdJ6EDJiLPmzRfoY/6oogYiyS9gFPgHlg0fZmV7VmKvwzYC56ska179+0ALybci9RKEl7gUfASWDJ9psu683Uts84eh4wF2V4DJySNL+9IGkOOF0fi9gVSXuAh1SjcS7YXu28ZuH3+WfAXOyapP1UD3l959dDXneBg1QPeX3rsb0ogKQHwBXgHvC0cXizi+2f0sN/Gbg14vBH23PT6yZmWT3K4T6wRHW32HPguu2NPvuKMkjaYPTfL962vTzxmiWHf0REtCt+zz8iIv6U8I+IGKCEf0TEACX8IyIGKOEfETFACf+IiAFK+EdEDFDCPyJigH4CGoEBQJo5XV8AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe61c07ddd8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Iteration: 100000, loss: 0.01387718040496111:  50%|████▉     | 99984/200000 [09:51<09:51, 168.95it/s]"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe61c01c550>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Iteration: 120000, loss: 0.014888259582221508:  60%|█████▉    | 119996/200000 [11:50<07:53, 168.99it/s]"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe61c04e3c8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Iteration: 140000, loss: 0.183819979429245:  70%|██████▉   | 139993/200000 [13:48<05:54, 169.04it/s]   "
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe6140534a8>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Iteration: 160000, loss: 0.09254933893680573:  80%|███████▉  | 159994/200000 [15:46<03:56, 169.09it/s]"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe5f0323898>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Iteration: 180000, loss: 0.14224696159362793:  90%|████████▉ | 179991/200000 [17:44<01:58, 169.11it/s]"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fe5f02d6748>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Iteration: 180000, loss: 0.14224696159362793: 100%|██████████| 200000/200000 [19:42<00:00, 169.11it/s]\n"
     ]
    }
   ],
   "source": [
    "\n",
    "\n",
    "# Sizes of the layers of the MLPs for the encoder and decoder\n",
    "# The final output layer of the decoder outputs two values, one for the mean and\n",
    "# one for the variance of the prediction at the target location\n",
    "loss_values = []\n",
    "agg_method = tf.reduce_mean\n",
    "TRAINING_ITERATIONS = int(2e5)\n",
    "MAX_CONTEXT_POINTS = 10\n",
    "PLOT_AFTER = int(2e4)\n",
    "tf.reset_default_graph()\n",
    "\n",
    "# Train dataset\n",
    "# dataset_train = GPCurvesReader(\n",
    "#     batch_size=64, max_num_context=MAX_CONTEXT_POINTS)\n",
    "# data_train = dataset_train.generate_curves()\n",
    "\n",
    "\n",
    "# Test dataset\n",
    "# dataset_test = GPCurvesReader(\n",
    "#     batch_size=1, max_num_context=MAX_CONTEXT_POINTS, testing=True)\n",
    "# data_test = dataset_test.generate_curves()\n",
    "\n",
    "mnist = tf.keras.datasets.mnist\n",
    "(x_train, y_train),(x_test, y_test) = mnist.load_data()\n",
    "x_train, x_test = x_train / 255.0, x_test / 255.0\n",
    "\n",
    "encoder_output_sizes = [128, 128, 128, 128] # hidden layer sizes\n",
    "decoder_output_sizes = [128, 128, 2] # hidden and output layer sizes\n",
    "\n",
    "# Define the model\n",
    "model = DeterministicModel(encoder_output_sizes, decoder_output_sizes, agg_method)\n",
    "\n",
    "# Define the loss\n",
    "log_prob, _, _ = model(data_train.query, data_train.num_total_points,\n",
    "                       data_train.num_context_points, data_train.target_y)\n",
    "loss = -tf.reduce_mean(log_prob)\n",
    "\n",
    "# Get the predicted mean and variance at the target points for the testing set\n",
    "_, mu, sigma = model(data_test.query, data_test.num_total_points,\n",
    "                     data_test.num_context_points)\n",
    "\n",
    "# Set up the optimizer and train step\n",
    "optimizer = tf.train.AdamOptimizer(1e-4)\n",
    "train_step = optimizer.minimize(loss)\n",
    "init = tf.global_variables_initializer()\n",
    "\n",
    "with tf.Session() as sess:\n",
    "\n",
    "    sess.run(init)\n",
    "    pbar = tqdm(range(TRAINING_ITERATIONS))\n",
    "    for it in pbar:\n",
    "        sess.run([train_step])\n",
    "\n",
    "        #     if it % 1000 == 0:\n",
    "        #         loss_value, pred_y, var, target_y, whole_query = sess.run(\n",
    "        #           [loss, mu, sigma, data_test.target_y, data_test.query])\n",
    "        #         pbar.set_description(\"{}\".format(loss_value))\n",
    "\n",
    "        # Plot the predictions in `PLOT_AFTER` intervals\n",
    "        if it % PLOT_AFTER == 0:\n",
    "            loss_value, pred_y, var, target_y, whole_query = sess.run(\n",
    "              [loss, mu, sigma, data_test.target_y, data_test.query])\n",
    "\n",
    "            (context_x, context_y), target_x = whole_query\n",
    "            pbar.set_description('Iteration: {}, loss: {}'.format(it, loss_value))\n",
    "#             loss_values[i].append(loss_value)\n",
    "            loss_values.append(loss_value)\n",
    "\n",
    "\n",
    "            # Plot the prediction and the context\n",
    "            plot_functions(target_x, target_y, context_x, context_y, pred_y, var)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "colab": {
   "collapsed_sections": [],
   "name": "Conditional Neural Processes OpenSource",
   "provenance": [
    {
     "file_id": "/piper/depot/google3/learning/deepmind/research/imaginative_agents/neural_processes/Conditional_Neural_Processes_OpenSource.ipynb?workspaceId=garnelo:opensource_NP::citc",
     "timestamp": 1533652807835
    },
    {
     "file_id": "1SYohd8cNXwJ360Kjcibu5DUmk6_z2KVT",
     "timestamp": 1533652159348
    },
    {
     "file_id": "1F-D4ElWS4UDjxs_1vuUF1G28B_2wWZjB",
     "timestamp": 1533297559549
    }
   ],
   "version": "0.3.2"
  },
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}