{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# Linear Gaussian filtering and smoothing (continuous-discrete)\n", "\n", "Provided is an example of linear state-space models on which one can perform Bayesian filtering and smoothing in order to obtain\n", "a posterior distribution over a latent state trajectory based on noisy observations.\n", "In order to understand the theory behind these methods in detail we refer to [1] and [2].\n", "\n", "**References**:\n", "> [1] Särkkä, Simo, and Solin, Arno. Applied Stochastic Differential Equations. Cambridge University Press, 2019. \n", ">\n", "> [2] Särkkä, Simo. Bayesian Filtering and Smoothing. Cambridge University Press, 2013." ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [], "source": [ "import numpy as np\n", "\n", "import probnum as pn\n", "from probnum import filtsmooth, randvars, randprocs\n", "from probnum.problems import TimeSeriesRegressionProblem" ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [], "source": [ "rng = np.random.default_rng(seed=123)" ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "/tmp/ipykernel_125705/236124620.py:5: DeprecationWarning: `set_matplotlib_formats` is deprecated since IPython 7.23, directly use `matplotlib_inline.backend_inline.set_matplotlib_formats()`\n", " set_matplotlib_formats(\"pdf\", \"svg\")\n" ] } ], "source": [ "# Make inline plots vector graphics instead of raster graphics\n", "%matplotlib inline\n", "from IPython.display import set_matplotlib_formats\n", "\n", "set_matplotlib_formats(\"pdf\", \"svg\")\n", "\n", "# Plotting\n", "import matplotlib.pyplot as plt\n", "import matplotlib.gridspec as gridspec\n", "\n", "plt.style.use(\"../../probnum.mplstyle\")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## **Linear Continuous-Discrete** State-Space Model: Ornstein-Uhlenbeck Process" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Now, we have a look at **continuous** dynamics. We assume that there is a continuous process that defines the dynamics of our latent space from which we collect discrete linear-Gaussian measurements (as above). Only the dynamics model becomes continuous. In particular, we formulate the dynamics as a stochastic process in terms of a linear time-invariant stochastic differential equation (LTISDE). We refer to [1] for more details.\n", "Consider matrices $\\boldsymbol{F} \\in \\mathbb{R}^{d \\times d}$, $\\boldsymbol{L} \\in \\mathbb{R}^{s \\times d}$ and $H \\in \\mathbb{R}^{m \\times d}$ where $d$ is the state dimension and $m$ is the dimension of the measurements.\n", "We define the following **continuous-discrete** state-space model:\n", "\n", "Let $x(t_0) \\sim \\mathcal{N}(\\mu_0, \\Sigma_0)$.\n", "\n", "$$\n", "\\begin{align}\n", " d\\boldsymbol{x} &= \\boldsymbol{F} \\, \\boldsymbol{x} \\, dt + \\boldsymbol{L} \\, d \\boldsymbol{\\omega} \\\\\n", " \\boldsymbol{y}_k &\\sim \\mathcal{N}(\\boldsymbol{H} \\, \\boldsymbol{x}(t_k), \\boldsymbol{R}), \\qquad k = 1, \\dots, K\n", "\\end{align}\n", "$$\n", "\n", "where $\\boldsymbol{\\omega} \\in \\mathbb{R}^s$ denotes a vector of driving forces (often Brownian Motion).\n", "\n", "Note that this can be generalized to a linear time-varying state-space model, as well. Then $\\boldsymbol{F}$ is a function $\\mathbb{T} \\rightarrow \\mathbb{R}^{d \\times d}$,\n", "$\\boldsymbol{L}$ is a function $\\mathbb{T} \\rightarrow \\mathbb{R}^{s \\times d}$, and $H$ is a function $\\mathbb{T} \\rightarrow \\mathbb{R}^{m \\times d}$ where $\\mathbb{T}$ is the \"time dimension\". In the following example, however, we consider a LTI SDE, namely, the Ornstein-Uhlenbeck Process from which we observe discrete linear Gaussian measurements." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Define State-Space Model" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### I. Continuous Dynamics Model: Linear, Time-Invariant Stochastic Differential Equation (LTISDE)" ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [], "source": [ "state_dim = 1\n", "observation_dim = 1" ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [], "source": [ "delta_t = 0.2\n", "# Define Linear, time-invariant Stochastic Differential Equation that models\n", "# the (scalar) Ornstein-Uhlenbeck Process\n", "drift_constant = 0.21\n", "dispersion_constant = np.sqrt(0.5)\n", "drift = -drift_constant * np.eye(state_dim)\n", "force = np.zeros(state_dim)\n", "dispersion = dispersion_constant * np.eye(state_dim)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The _continuous_ counterpart to the discrete LTI Gaussian model is provided via the `LTISDE` class." ] }, { "cell_type": "code", "execution_count": 6, "metadata": {}, "outputs": [], "source": [ "# Create dynamics model\n", "dynamics_model = randprocs.markov.continuous.LTISDE(\n", " drift_matrix=drift,\n", " force_vector=force,\n", " dispersion_matrix=dispersion,\n", ")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### II. Discrete Measurement Model: Linear, Time-Invariant Gaussian Measurements" ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [], "source": [ "measurement_marginal_variance = 0.1\n", "measurement_matrix = np.eye(observation_dim, state_dim)\n", "measurement_noise_matrix = measurement_marginal_variance * np.eye(observation_dim)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "As above, the measurement model is discrete, LTI Gaussian. Only the dymanics are continuous (i.e. continuous-discrete)." ] }, { "cell_type": "code", "execution_count": 8, "metadata": {}, "outputs": [], "source": [ "noise = randvars.Normal(mean=np.zeros(observation_dim), cov=measurement_noise_matrix)\n", "measurement_model = randprocs.markov.discrete.LTIGaussian(\n", " transition_matrix=measurement_matrix,\n", " noise=noise,\n", ")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### III. Initial State Random Variable" ] }, { "cell_type": "code", "execution_count": 9, "metadata": {}, "outputs": [], "source": [ "mu_0 = 10.0 * np.ones(state_dim)\n", "sigma_0 = np.eye(state_dim)\n", "initial_state_rv = randvars.Normal(mean=mu_0, cov=sigma_0)" ] }, { "cell_type": "code", "execution_count": 10, "metadata": {}, "outputs": [], "source": [ "prior_process = randprocs.markov.MarkovProcess(\n", " transition=dynamics_model, initrv=initial_state_rv, initarg=0.0\n", ")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Generate Data for the State-Space Model" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Next, sample both latent states and noisy observations from the specified state-space model." ] }, { "cell_type": "code", "execution_count": 11, "metadata": {}, "outputs": [], "source": [ "time_grid = np.arange(0.0, 10.0, step=delta_t)" ] }, { "cell_type": "code", "execution_count": 12, "metadata": {}, "outputs": [], "source": [ "latent_states, observations = randprocs.markov.utils.generate_artificial_measurements(\n", " rng=rng,\n", " prior_process=prior_process,\n", " measmod=measurement_model,\n", " times=time_grid,\n", ")" ] }, { "cell_type": "code", "execution_count": 13, "metadata": {}, "outputs": [], "source": [ "regression_problem = TimeSeriesRegressionProblem(\n", " observations=observations,\n", " locations=time_grid,\n", " measurement_models=[measurement_model] * len(time_grid),\n", ")" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Kalman Filtering" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "In fact, since we still consider a **linear** model, we can apply Kalman Filtering in this case again.\n", "According to Section 10 in [1], the moments of the filtering posterior in the continuous-discrete case are solutions to linear differential equations, which `probnum` solves for us when invoking the `.filtsmooth(...)` method." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### I. Kalman Filter" ] }, { "cell_type": "code", "execution_count": 14, "metadata": {}, "outputs": [], "source": [ "kalman_filter = filtsmooth.gaussian.Kalman(prior_process)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "#### II. Perform Kalman Filtering + Rauch-Tung-Striebel Smoothing" ] }, { "cell_type": "code", "execution_count": 15, "metadata": {}, "outputs": [], "source": [ "state_posterior, _ = kalman_filter.filtsmooth(regression_problem)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The method `filtsmooth` returns a `KalmanPosterior` object which provides convenience functions for e.g. sampling and prediction.\n", "We can also extract the just computed posterior smoothing state variables by querying the `.state_rvs` property. \n", "This yields a list of Gaussian Random Variables from which we can extract the statistics in order to visualize them." ] }, { "cell_type": "code", "execution_count": 16, "metadata": {}, "outputs": [], "source": [ "grid = np.linspace(0, 11, 500)\n", "\n", "posterior_state_rvs = state_posterior(\n", " grid\n", ") # List of Normal Random Variables\n", "posterior_state_means = posterior_state_rvs.mean.squeeze() # Shape: (num_time_points, )\n", "posterior_state_covs = posterior_state_rvs.cov # Shape: (num_time_points, )\n", "\n", "samples = state_posterior.sample(rng=rng, size=3, t=grid)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Visualize Results" ] }, { "cell_type": "code", "execution_count": 17, "metadata": { "tags": [ "nbsphinx-thumbnail" ] }, "outputs": [ { "data": { "application/pdf": "JVBERi0xLjQKJazcIKu6CjEgMCBvYmoKPDwgL1BhZ2VzIDIgMCBSIC9UeXBlIC9DYXRhbG9nID4+CmVuZG9iago4IDAgb2JqCjw8IC9FeHRHU3RhdGUgNCAwIFIgL0ZvbnQgMyAwIFIgL1BhdHRlcm4gNSAwIFIKL1Byb2NTZXQgWyAvUERGIC9UZXh0IC9JbWFnZUIgL0ltYWdlQyAvSW1hZ2VJIF0gL1NoYWRpbmcgNiAwIFIKL1hPYmplY3QgNyAwIFIgPj4KZW5kb2JqCjExIDAgb2JqCjw8IC9Bbm5vdHMgMTAgMCBSIC9Db250ZW50cyA5IDAgUiAvTWVkaWFCb3ggWyAwIDAgNTU1Ljg0IDMwMy44NCBdCi9QYXJlbnQgMiAwIFIgL1Jlc291cmNlcyA4IDAgUiAvVHlwZSAvUGFnZSA+PgplbmRvYmoKOSAwIG9iago8PCAvRmlsdGVyIC9GbGF0ZURlY29kZSAvTGVuZ3RoIDEyIDAgUiA+PgpzdHJlYW0KeJzNvcvObltSHdjfT/E1oeGd835pQtlGQqqGcUrVQG4gSNKgk2BnUqYev8aIEWvNWOs/eJ+UaLhKyPtEfv+6zDVnXEeM+NV//M3/+oe//c1f/cWff/6v//rtV+e//vYP3/LnH7/96s/y57d/+KTPP+L//vWTP3/xef4oQf67b73376vhnz9d/6yp8v/5CT8I/+QP/vu3b/+Ea/8W//EXuOBvv30r9Xsvq639afN7GWOPjWvWWb7vUUcuQfxTFJe5vuc1JqTnEkGIO/39t//5uf63NXCRc61a8JRtf2+151J2qg2/TN933rOm3nP7/P43n//n80+fX/1ZwRp8S9/TKLP1tNcnfW875T3S5Ft/bxKnuj+//y3e6R/t3f6tP/j2+oNv3yqeImc8wqeM+X3wWRZX4Ihb/953nXXgtRou3FbPFLfvOY01JR0QFkrL9z7xk0Jx/l5LnpXi/D33kleWePdVG8S4y2qtdbsI/nK33bG2dX0frbZRKa7fS819UDy/97bwWBKvsfKkeHzvI7dh4va9pzYXxR0PtVK3J+nfc8t1U9y+42Xm3hLPuVLKJs4rbW6VNvDquByl5TveG5uI4vk98WtRnPHlyh4uxiKsmu0Llo7f26UX3h2vBnHZ30vKqyyJ98ipQ7qwF2pKdseNV5/YIKXM7/ibWvnQHZerrUxIx/fZclpb0jXTWibNM+fZKc1479m2XSFVW11IC167pU1xx+7q+ogQ461nopirse2E9IrXnjVLmnav08T4d8fKUVy/Yxv04WLsFfwPEONyM/kzd7z2zG2ZuCycyCoxVmt2ivH1sVLVxOP7SKkOicfOJXEndBwTrOL0X+NqpUi8cHt7dWy4lkcz8freM/Y3xBnriCWZJt5494HXhBi/KL22KTE+Fl4UYny5zm8H8cC5KClxsfP4vhOOIVdqZLz8wJObuMyR55TYzjDF2EI80vwIo+Dl9+rDxHVNLrGJ9xiNn5cXabskPuCoePmapn6dM74612Q0vPwei2LsSSyEHrDh5Qf2HMQ4EQPnxB4QZzNDgWBDZZyfuvOwX2Nr4ChlirneUysI8VzQXhLjhM1mv8amw76pEqdR+QuIF15+10YxNl1JeHuJcffdm/267Dy3Pff+ji3AzcFfY2foW0IMfZexlUri3h/Vrj3xAVOnBqM45Tptc09csOH1KZ44kF06AOI1d6IySjicWIfGl4c2xisOKqMExQ8NlO3XlbqmFIlxlHdrEs+1VpWYX7jy5WfDy/dGbZSwxtCNdtpnpwal9igwIY2aaUk89hrURgn7u0x9yznw8r1QGyVoOhxrvQ620syL2iglPNTsJsW+Swvb4JM3Fi3hiNvzQSE07GSKsWGhu6qL1+I5h3hCcdZlJ2pyqy+oVxNjc+HUQbx4m7Zrtmu3jYMh6cTpb9kujZsP22oLe6Cs3HUN3CXpGgWvjkU2aVlbnx3Sye8E6cBmqMXU3IJKqBNLQyk/jGl4SDeszDYpjkmyE7mgMnIuJsSFtp3H1fkDaF6JoS/s4C1+T2jCbQ+Bw1Fs6RZUPbZXoZgKLVc7BWtycUelGN9wTGgYiXEXHDiIK/XusFOwFrZ+Wn2ZmFpf19547QmVATE3FrTikBj2DLYTYhylCb3Di+z0vTaYP4k7TuMYEu85y5YY+2qZWcK/sU2hViDG4SirmhLFfUqHUYEYDwXvorcu8VozF4qx1lDrlQ+46UtAAVPMY9+rrg0tAUXSKIbuWHlNF0+cPCgjbBP8AgbfnhsnAlt1UFxoP4bZD6w9DELmR1881I0+iIknFmVRnGB8sUHtLSdevkJLfXD4v1Odmr3n/oQ1TRTzdfY0U80tl8eAMoL2/I6Fx06kGGYRWqJQjG9Zdjc3AOK19qoUwzil1G0LZxwgLDJ8BcjxMWkrlslx3gadJpNj1xdTMJTDyEPvQl6o/7e5Ktg8WAD6RpDze/rmzDjl+Bc2ncmx67DULocahw75ZOgyOm9mWjOUBTwU/D+Q42Xh0Izp8r1wGCvk3JabVplyaA7st0I5FtvOtMlxQjrULOUwA3V0cwkoXxsbgnLcKtE6mRxOEJRupxwbAts4D5PD9Rw4BDhfNDhQAXu5fCW8S7HfUw34e9H7xELr92PObuowU3lPPBDlWEK8iW0QyrElqe6w23D05rJTg6OJdcDbZJPDXMkCUQ5T2KigYHN5sOzxYaVwClOVGI4AtrjJYdT6GlRRMOjwO+HwuHzB1nXJJxah6PJw9Cr/lpdp/P62arSkY/Es4Ne4jd4VUjiC2C/QEnQEly4x6GtBjcGGQJWXkXUJ7FG4CJDyqGBxskvh1mw+Bg4WXTiJ+emhQbfJYbT6bC7fWBQqKvhVE4u4JMdSQN9VyRt2jVkbbGbY9Fyb5HA0Rrvka08qKD4k9hi+vcmhc2CZqawg79C8pqzwgeQTSt4SF9jlcIXmkryukbKuj4eA7dv+ezgSU3J4fbAHSfJJN2S7HO9Ci8ClpOoyrQL1jHWgE2FyuiRaY7qdO62qD4JVkx7P8F15Bv1DrUbjaXJGSA1PYdeBiYAb73KoAupY+4SIH3Sa4UVjPxfqLa4Drli0DvD+Zl1LcnjGpXWX47UaNVenKc80MZDDdcfD5CR5o7awdYCnz6en7oIcynVrnSuPRqIvY3Jo12zvi4ABynVXyaF9cQZcjiNJ9U55aoN+H+VwAvGJqL26Ke82dZ3GSHGObvsB1rhLizB64a5uJs9Ytm3rj2AHzh8d+UzvviEGsv1WGSOMviXHGVx+HWwmPFSSHM5uLroO44+NxzB5obeo38NJw+mk9oKc+lbHGfIpbx6HAHIYPPsuCOxq39zEGQEL9l2u1eVQyqVLDoVbdG4ZQdYCv+STGRbjtJhnis9DX7dRe0GO0+9akzFkgiNDOYNFuCm2zgwisQW25FiGpPdCuEg/NUle4NVKOzKMhAahV8JADjZr6/rwCrFXqb0YDmI/VMnx8dbOVF/4J0NDvz7CAMR+VF/8U6h5qTX8MYxNpfrCrXD0S7rk+FpwST58ZGj/of1jD0FDSDFWcFkEYc88Ecd89OId3rCLcbjwb4o7fB0ddoaTDetK6USsbiuM5YaVTrwhPghCTy0ww8lsX5nfZtHtsWvgazZ4NUVyaP9pLhW/ftrwKCjHAsPR0gHlrsPmogLDHsIqDpkT7FKatC45tn1vehwomARLKDnCXDlKlOMvCxUY5NigS+aQpwzRChUY9+66FqybG8ivyTOA9yoy56YVcdQph85CqJL0nAwrcNIpZ2oI/sJw+UJkRQUGObwEfScop4aFpf7CEcY+TFpiGB04BZ36qyqdofPG8LLC3kjeB9Sm3RamEf7BHJLDv1oyKwwwU6vUX7gtkz86z4wwEbgvyeF8UUVQzkOJMGLY88BMdVlbmvxsm4ty7Nutc0UXAYYvS15XU1RGlwJ+YqP+onwg7OkuX6WlKjn8m6L3hWsCnTWa5FRMlkkwVwauEvUXnhMOZNI5YaRZcQa6fS56I133xbeeppwoH5ORqcmhX+CALsn5xOWSQ4mMLXmDJZQdhStW4NEm//3sw98L+iVTrZvcYkZbH8SbvUNbVttWeKttUZRcvUTlRHmfDMhdvgqclmrbHGe6yJ5NLgq2fLXDhW0rNwCeZ9p7m/rC6YKC065izIntPyVHwODbCp4tHPls6ovBsiWp6AYnrCaVF88iPqE+LoPOhu3gcpztrN/DbYEDkSWHdsPGc/lGNFV09cx4Wr/HG0JvV+kA+D7L4go66wh5limvbPtUSpyhp4UC9nssx5JRhM+PgDNReVHPF6y+3ZfBZ8LPsi0a1OeQMmX4WXEDW2McuSkPHYEGU0LSALBjK+vX0DC5aOOsDo25XTrgfmV977K3Fh6hTVmd5qEyBE/+hAxAC2JUXRsGocprRIDUcUfTXvT+ih83xFM4Md20FxYbll1mhmEoYmzTXsobSoy/HWsOieti8GnySRenTimXDY9CSo2BKByNJTmsgMIoBX3woyWHrzx1ahmKQqMmyRle6bRBPhD1ZylBqJymhUdMycjFlBddXd7Y5QghUpUcXl6S3WDECj+9SZ6ZyZIc3iMiAtNeeN3dspQ7A1LYi+HyeVl5RKRQpW3qOmnWpo3GkDStZNqLXhPM/3Y5n3kPX2Z+UZMzW5araS8ceLygTj+jUuw2016MLXFuhssHTkGRvHkay9IVzAl0E+NUVdkaxqXQmq27UoD3abvE0ixwPSXHcrgzzsgUB9GVF3ZDmpcceh7KRtdBiGkhKHM7eITsymtD+dr1LRUEw2nKi35tT6YEKZ+NPppsInTj0nVMQ6dLeQ08qeTQLlAcRfLU5nQpNEsx1QXtDFtjlpJZLPg3y1RXsxjMtAXlUBGtS467KhPHZFgfiPIlp67YJh6Mn8aUGHq6t+lyPGOh6qp2p5z0spP5n7UlX1DZRQ8P5YIHTJKzuuC3heeI0N6UF/d4T6YaoZkZnkwqL7Nw/rYMS/GuVWL8vOVLDvd2U3dBDm9MyT6lJken7oLBXUpsSw5vKw/JJ7e+fg+rOfacLp8zFV0fCgZrs3T9hSfuy+U4cHvrttuzK0zZ4iwNuwgTvbon07uMIy2cgC1ty8WzayHxEtDL2k2Z8dm2z8ekdE+WWGK2Gro1mepKzBqaArQkth2kS6w8IHPeeLrSJYdD2k0zUo44ZA3JEYWkIjl8xml6gDFSQqBmR5yZdrwF6w2MteBw53HJV2N487GYbePZ7L4ITPGF+SqK8VLSWWBgCv2VJcc7bQtErHjQYf8o7wws/Dux1oCTRNXF2JI7SNdvTJ/NJjniUZ1xygecEqou1tzWfV/8MYtflNN3wga55DgknaqrZNuh+ijFvAlLLOAVE3OJ+j1eEnE1Py2WCgoq+/PQouFAUL54erq2PRarcbdQPhkjKenAYhBuOIrkjNW17SGHLUSY/WFeArpKpoM1JVx0UXdxy5Xhx4o7qSLsoLywqLB1rCAfiEapuyxpX5eOLQ8Nwk7qLpywjn9bBYxyHB7Y0A8zRYg4YHFNjkWEH0DdhZPNeE+Xx5HHXam60jBXIV1ypt+puqhPWKWz16JmWnNSdTG7jv9h6/LDvl2VHNtt7+lybEnb/Yku7VSukKW/CrtD5ZWYWEIgqOVZVCl5SA7PZ+kQ0YzMNKm9qJ2Z+9XvmXCfFWdh02NC4F5djHtiDSnGw+ubw+nBRsIJoTR3ryCxloljB43xgQHZPfkSQLx6peXiNaDe9ePCrArOH3+M8DtZWqswHi2T/0HjVLvCaYonYoZuYngbyiax1NqSnTBYPipWnR0Go3Uw3UuDiEh6dhdDbSP4obgxY2xSfC6sxaawYi39yrweDgeFy51laAG8N5QkXwTu0qi6LDRK9iWC21tncenC3rD1hAvcll6PtQdsT5Mu7H17BIagxfKhXE8oFctoUjwHbmLi5h+K8WcqLLtvnoypz8ros+5OLYXPWlgW3i5HTJKX5HPDbOj30CJYZ2oploLa9DvSOUO0LHHFlzf/CiuIFx9M5ZgcrlDVg3fLDxbJ6UDO7XJ4b9hsJkdcXnR4LJXWLQ2ZbLWhgEw+zdnqkq+1k0wz5HAmE5VUko/u12cKA8Z56nC2IVvI4LMVKmhmWfGqXYfK8qy46jA5gz1L2hUGn9kKK5TTIdFqMo+LUNF0FFO6cN31e7iHa80iOfZX075g8Im1q/57HGwLoihnBsF0FA4hYkwdFASfCMx7l5z+mqUiCoPP2vOQfOHxu4uZKKWKghlFKOYeB2NPbKIlOfwyi60KQ88GLdZMM9K50uJDDg9yJMmrhTkmx4mi3pYc3m9a3eWItFZxOTadLgPnEHFkdXFHTCQ5dj+PsV8eRlZOGuQLURo1VDYfX2IGnskcTmahE2u19jQMPOH7U0HxpXpL8hUmnY/Wli6DHTDNlS8MPLHqu5q+x9v5WzHybPRWTA57bnkCigdcQHpX2SxCkT5m4Al7XiSv0LVSUyy1wO5VyWm1ZT4nsR8Nn9Q/7dgy/ww9oUjpXmV6s7v46w6W31iYyuaZ9ZX1PNjWOLl0r2B+JhOpkjO4Tp3uFcvDrciaQwwXLG+Joe+qpRUKI89sx40RVVeFFF8DC0hdaqYKtkgabxGVYisDMeKIKf+YYSfsTNUZ5AecLt5Q9TJssHzyuhZBKYP61k7m0B5g1AnbRefKyjRTATnluGEakteZLJlbGHViQU1rQU1MOi0mJyoFkank8IiSFncRlrLXlrzDI9LeYNiJU+Nqi6VFXYaxTs2utTrMpw7IIjIFd5N8MnI0MZEpo9oq8iFnknJaLM8zM6WnYcJBlyc4xYpzlNMcWpBXtld7pplweKZFS7apMvDBJWdd1G7LmBPbd0mceHl7Wcaczcw/S23YAW5lN51/qKghOb61gifGnAhIqbRSs+NqFQU8OCO+RaWVmMSAj9VdPjZWfMjiw7Ozt2LIiUdFPEm73OZ1FQJVOg8Rq+qbtsbEBKrAAnQzqSPJGWK8WaEoIaXPU0edLl4MkSFmvFu6PFpGm5Z4opi5a9kcmkJoUqw167uJXouLEXKZCWXpZidTbpXODKww/mNT81elmeHh4uXxcxPjp6qOUDx2zzhTLFcTUlNNDF+wZipvFrexYeslxlbnyYAYm6523ZI4lcY/5Q6BUbEAqPLbtsyoAmLGqn5t4lRwuk08iDvRA7Jg3RAlmXgtGT5E/NCsmQEqNitOc8rDxbBjTK0zo4JDOPUkTOW0ASVFHEBnetzE8AFH5q5cw9JwtbmYGKIOMXd/UeqtMr6EUYCGWlYzqdOlUE7EPyxW/Voxi1epw6BTl8TdFBjFhKkg+st8G+hdU9HVbBm+VP4Y8CBpO1RGlpmakueRob6uTJviLw7vRmliSifBR/ZbFuZMSIQKHDqtUa5LUp5F7E97hEKYiUvxPTs2M8S4fhZukTFlnfwPppamh/4U47NB13CTUYVkXZoPCl/QxIRvLP2aAJWB9aWYZvgWMx0PrUQxzsbWakBtlESzwBMO3aU3LASoDILheO1dlvZHoddW6WfwSaDa9A0ZTRIjZ2KEDslOaS28EcK9yTNTCUEaLp5UQyZO9KckJkZlw5DycOxGzepiojlwUkxjVemdykASoTH00doKWPRr4lQ2D+ey993zEi88LLbtMjciZV2EOcpCPWzKeS+/pQECmDrHTs0wG6m7GIEljCt3Gc5frxITp8Lk6sfSlr0UXZs4FTis+PKsLCcF/BQjEME1KU7UI/byjB95ZCHGel/fjFjTuWmDV7YCqt6dwSN2KJ5qMXZc8rRqJUwFh71RnJk9uMR0YaGLaJY7vW4TV0v94gPSiruvWYmvxBlsJs1UXMXF+B5Y5o95CF3lvcqwsRr4iw4FdYEem6XcwvNB/4PFle1i7FGGcsyU912LLsK6Ume8b2CuLo+pKpNIN4zQr0XP0sREqRDE+iG8hCU23ZLBDmGfFOO/thSX5eet+D3NTnVtekaM8AKaieEiDIs3KtGvw6Ctkxq5DClzRoyJADaKMzG29iQMGavBQ4iSw1roLRsBKg26lWI4k0nqhSEjzi10EeEpCIb1IESnGLJ1MC5bQx+4MSFo0FYWVqD4LdCtjBgL33GwRr+6PzWRKURCEBeMzac93OjwQDVRmomjsrVmyFhZYqP3jTfUKjViUohrHUwmwlZ3l8J/ZmwCz56x+NSPsYmaVTOHVdqzRfjVysST6GqrTe1s9pFixCWEthIdAzNT9WuY/m7Q1kFnIpmNpRTnH5qK0s76uS00o8ZihYnBtc0CjlYCMEYiHIbAGJr15mJ4U4S2WhyUxtCvCUQxaOvg5ywqdCGi/85iGHTRYCi2/N0ZMSbmGT8MviYz2iZmMQIHy8R9li41QogFdsg2MS63X1LCMeCoXbf8t+T3xSVnRPx8FpNjUerr0SVPOk7nRU2M7znWc11MnqZi8rCMlHcWIsZz1U2OLZ2f38jEnRDs5xc1OZyA/NoAJk+tygKe/UI5SwMtP7eXyVlsqc/daPLEQPaxdymu8N7nfGx0E5c6d36cCorL5VOfI2Ri+I/ted4oxkHf83U6TQ4rLcfsHGbD5SCe1LY+R585z71Ke2sKk8Mf1gc/ioVyhD5KlgU9RDmski/lUVuUD8a0+6nlLNc6TH8/lKLhigZRFk8dSjnO00j7qXIpZ05J7t/R0JRjGfz5j0KnPDEzUp7633BOzBledsHNBeUTf1iexoVifJ5+myi3RZQjwO7SeMd0WQr58giOncvmj+zk9u82i8ws0xnu+WlFKR+IEHROjtGlvJlWftpoyguTd+Vp0ZmhxpkpL/tP8XSX+TgLlLZ+bYXjW2TzyOAcXh6KuyJMc8Otu5yO23OhHHvCXZTbz7GsODfFeLpFzIrzVEt1HS/KsuVQVlZ9CE4Xs+ILayvlf3w0yuFTumk6Lp1ly2HadciPB0g57qFCZXAYmYWB3mgKYo5/SXkbxBY/3VFDFCqtF3xX5mzg0Oqmx9OlOMOBl5d6HGOmcnhMtEGOH82UzeqzuvN/u92WyqlMlT69dKXvWCdwuTv1zKkglB06hicGMDmetxSXK2IwcYIO1mNe4YVBRXNSL8MJRSgmLN2DkStuoRiORV3PIEdpkj3eIRFzDTvX9g6gKMdr7aab3vGW5RqglDw0vcMzy02w3NRd7sGcXQcn3uV37Ec5tqiO4IkULcmzcfJ1+TuwtMdkZ4JCyDsOtRQSbKK+7Alb9VrTT8+Jcikfq/d5xcoeFBtMF+dZmuXE0Lb2zPcrGr1DbivWsA9C4fIdoRtMt+HLbpd7QE854tTxjv8Nvtut+vtTTBdYIpf5NS3bnV0w+byqFScZYQlhxOPKi5/che1kZn2UXLlTHVk9LFmZqpMZ4bEqTLrmzyORwmNLpZp1/TvtYsWynjwRdrI0lG+cBM8x3UkdqpfOZNLnkQKSlgryK2VkZzgLyxZTTLRj9PR0+ZOSMsU057jkVwqLLgSrn8qQnZQXcbM4RFu/Pxky9l+xCal/nhm1wrM6pmq8IQNHefeSbUjY8TI4q/WSXwk+c9TKVCk75AN5GYal/fPMHxamRMYlP+lGVktbFmQwpicLgdOE2X+e6cxC7HH2EmlIfxbLnq3+zpba73H5d3bVrp8UAMVcLO0LtKO/75255WMiTm6vPC/b50ZTT2LICluv3ShKw50cMsvcVGz26zvhTIPJHr3PMztdGOIOrxeGZDbtLuH47+Q3y7/YbFvZ7JMsNzstNyrm1nF5BOG1vHPxVv0tQuzH3D19CZr4/Hnm+olRgB5QfjLUBgwTSBTh51lKIMaPakoliVN6ICIjL2GsY6miWnPQUhY8lDaIReQl36WQatGxw2dC5YQoKOLjy+dZaSE2kig9ladOZYYQzpKzcrehkGPymbQRQuGH1/FGilgn4m1xMFSqCHUlQi87rcrnWYbi8kCjaIuEspUhUNZs7ypXNX+vL13nVMWINNlVuYNYRTN53l69O1U3u84QKCtW6Qg2YcVZz3OqepBnOvP986wCEm8CW9auauJVNSQiBKfQi6GnysgO1jmmSkihKGmIEFzpXcQkKGJ4J1csehb2rrDn2eSnSMqe1bFGexdVqRb2VntqLMLy+GfGliY/RVuqtXwhUUKRl1ETdoO2T6gJwy3EkXboTqgh01XGM+yrtnzVnCnf+Vqfu0JNeUWomF/lbHOt23Ck0ql+m2udvMvy1MophjMzdbruwrp50IP708WqwlOc8J29jn+V7M1QLnmCd3WfVtV8qelSIQEoZvuQ1MWBDWRhjdRpE1AG5oN39agFTALFfdI6m/yGMFCOCG04mOJGPJi8X7ixA5Cw6xD3qBe94RT2SrA3/vsbfWE2vmb3RQ5Yw9YRZlSf+0A7DNAzdlaZ6iBBKMfHy1+QIwQG4S91/YA0YfNNYgLB5AeZUlgJXVVGPiBZGAYOhN1v5AvlOE4O3DlIGWYWVvKKakDWFIYuNL0uv5A4OCbsRtZ3CcAdHk94UvkN9DHjNF0tBGAQm2PW8np2ABJV67YWAjoCj4hkxyo7hOYAlQxYt/2YB2ATXh0+lmMtAhCKS0g37w2cqmyRn46dCEAra0gYjs0IwCwDRxf3CgKQqxk+U/2mEfjVzMddDlw7QDGCoDuda5MfYBmbLC4EfQSiNcuMbfmsAbjWaDG6H/MAdGsMH/DBh8svYFwb1pu+3kC6xvXJDmkJwDvIJ1tQi8svoF5jA4Zvw4DrY0YL3t248IEXDtAyXduPdcANNnr9eFvhIw/OEIoMWkBA/whLZCtJch/0gBh7tjYlR2vekEdL611l/QOQ7JYBcWzRDaaklEpC0ht5yWuwM/gCZF5AzW5b1MvUAdhJOfNZwlMeIGjPRq8g7RiAo+yA2lmhbASaWq9LV+wYgam9Wl3bwb8Hx8pMaGf10OQH99oN9zpbdfmFkzULlRVZR1ytdXx5CSficCFH8K3aZMTtsqOssR3N5Afm25utoNY+wIKNvKE7pvmgiK1RrpU36Nj62wg4cfkFUWa/nbqSfnpAmiHP5BYQSPlAoHmdVRzOGiDT3fqu5OsEhDXbIHEGt5DdB5Ftboa5YU8Ed+fhqXJNI+Kb8gGv5o0Qt9+7kYiIcrv+Vm47AtDteUZRd1QArHdjEVmOHD8Ad7o3hT0nnycgHnJSD8zPEz/PxYcRVNdewNtDPo2V4PPE59vHUsUiwvl5dVa19VYH/k+nCl6Gw/ZPuwC3SO/evBTaCyCHP5m9+eK0I7BJsXj9OLYvdLWe62uFdgc2xS7vXQrdEdYD2braWEI3BbvCcIjVsxe6LyDf7O7Q70+zBksYzHDVz7O5g/LavU0uNIOwJxP2Sl0KoXmEvb6McS6595rYVl5ZrUWnM4VbYdRnE0snjEQqOjS8wIfGcVlFy3K3x3RjJvAG8NBNw87n4u3+sfuGrbPkG5H8dOuwg5p3Wp9ndw9/zya3z7MZiMUlxcTP5qEh/Lk2U2g2Yo2KgFo1CZ3mJMqpWPrn2czE68D66muE1if2i+80vZnptEp1cob06/entQryzvSnWqhOJ5YtA+tan2fnFnv6CGPZn2enF+VwsPq7M4wRDKt96mA7nWTsaC5JuIvYeWadyCVLJYRONXZMty12mtjZxm2DQ6r7hk44dhxjzfXZQ+ccdz17DN+ddtaBu5RSj515kG8qPT3n6eSjXUJEpV0cOv8ox/Hxzr/TKUi5m73YWIjLLza2XfKrEbEz91tdo4XGRQZOk4wen2ejI635dHRRbIyk/0DKFfV1nkZKmudBwgSXX42XvC/1ieSnUbMbDcS62j2vtk6qHGxDnfHQBtq5IN37IkPbKONKKHOpzNBmyq+ylj9+aEulrijDv0poY7X++a4SX2x75a5KV9NbaJPtVniUwxTbaqkESHm0XX614ZLoqNeivvHQtkteBFITXd2/V5sv4/2Vu6/D3RXcDWbdddjvDmL78VTuP7Qb82tWmLPiYm9OHsx/THHAhFZmZjY22VxMfDqfyYDEOm91+dUpTQokYirVKX06q21LTwFiYyc28QKIRaXuQuc2yZFKG+PqAL86vSEnH5M2TugMJ85hLSekCJ3kw0ISZ5gIneeDK7+Vboyd6pCThEodvqGznbgN1i6uDvmrE35SP7J92+VX5/xMVt/VgQid9pOOdvIdEjrz+fs2sp4ndPITdEIElolP4/9kH9Vo/SIEuIgCSHIH06BWx0AsMMkVsdR/G3gIJnNtzMu6/OItmMRqT9/3gedg0ivd8p4jLwLlCA7mm0cB8sy8r4kP7QLEMORKjUaahskTeZGuBFqH2Q1A5vQZhwaC1E2bAZjLL9qISRTwdOMRaCZIYdPJgWfyQ0sx2a0+3UYHGgvKEfNu0VIc2otp/qrTOgSajMlsJaLt4fKLVmMay9OWaxRoOKbhVGebLr9oOyiHK6fzHGg+7Dlz3hddyEULwmQ09pE6kQONCOTQKO7BBNoRcvYwmyuaj0NTwvVP7m4HVhOa65LEBhdZUIiqGkOptsiawq++xheSFYjXukgpDicLbgqvWBDUQ+AyrRyQ9WkP2wu/OP7hNDU3NcwkprQ7/8bhkZl09ceUU3NYZ/hr7DLnxjkkNZMRRlFcG0ltJt2GvZ1i5pDgkD9pb3e0A2fOYjqFbGkuvzh26CwuxzZHTp7Fkl5WcB85fFY2TiSp5MD5Q5YnprzENXQ4gpbFyAKSRE6hZTvFyR4CB9EyzHZ27qPDWbRYlFju6wSOIzrphP9IfjiRlkVUResZOJQQBDBK9+c/nEvEwuLyOv6Bo4kxBuyCdG/gdCIcFkZE+y9wQG2L/Z18InBGMaTqWfXJyDG1LfugZo3IScVIjjieF4XVNpy+UKGB8MoY0FZ9sWNtw85t/fhwaW3LjutkBuYt/pNMc3aRm6eLoTIjOlOvh9WLgTU+R91PDjDG55MUGPbrmzGM4TxeINkROfxiRpDHsNA4wG42MqYW6AmKX+zmLmMmojOLMT6R6cyI/epQDHZ40SgeSeD3Q6JWrC8J9qR+IuUakzQ4DWbwDz+bpXq45fInsrlZZgienJnFw/1meaRk9Z/IFMekU2Ht0sXOK0cxdvAw9XtY6CiG+6fy8+GsM/GaImI7DHcii6zq3rvY8CxVRm1gN7yI8yhdhDu7VCR7zLZhtZ297yLko7Rx/fcnkPdZxi7hrDyZ/kyMbZ2fvIAUd3J69U8kEbQbjqre4UM5aFm/7F/8EBQyeZgNN/WJdIbMQbJcp4vc5IfMWDKTbD8+RIk0JfjBcPFFq8h0KAICMSUeEkYCe0pyEsZD2Uj8FiMSo1s8BI/MwcIQ1BcdpMEQk8qCgTySxXD6p/Ykh2qSRSkRdD6IKWnykyPzAo2lpa23SPYC6SUcCpyMKn7QQ5HJem9iQfnzINSkW8L6jYkP/SbLcj2JZCOQdRoN7V521AO1J4uBF5VgIAIlfQ0zTC/aUJY4m2c+Asko/LhCsJo9yaEkpXs3lYgN/KXs98/1TXZKF7GoRhqYUY1LYIvlJvCoEpM3XJ0H1lWyXMDjN88+cLRWw+ssEY8eRleSLOCFzFYE/tdqLZUrOy3sxRZb2dTrCcDALUtC5M70xufBRMuaQxGvZ6CtNVKXVU31BZLblm23mpEIlLgMKRiPrM+DQLcV+0h6kEO321g/Fu9l4OZt9Py20p+ByZeliakEUWD9JdC0DbEzBo5gsiTv5kS+N6Ewea7gZ9sxOOTDrG3sIXqqQ1RMGCxUsTnZh9TY6hFebj4EyM06UbZlHwJbMvG7LPsZm/PhVuad8Vnm50HETM4xKhy7xqFtZuKee/rF8cwVICh6fR6M0Az0d9MeC/zRzNbghnZeAtt0pwLbIjQJ3NRMuRF8YOLDZM0UFzahfdrAez3MoVMuPLBkc+PD8RGT9eHUHqwTDB2kQMBtmZzmXRc/x+LtLOT/5fPvwUNeyUOeP3gq/F+5uMazEbNfwr/6i0/8xbf/yT/A/58/9/PVZB+Znjscs0nO08/f/u7zq/87ff7jP98/5zdjx21m0P0fMuP1RAwEVeAPf02iUuaNjRL8B78mOfKA+8ea1y+5crKuGPb6/fjXRp+FF0w//i0R91RRVhv74YXpTS9yKZIE4Qe/ZooeWwrK9BcstCXOEHsyj/Xjx9jM7yAYsf/88Rsyk5yJ1pw/vjS1PqNF801+8GtmAWFaWGL8BR+FME4oARrNH1232Dno5L36ZbuOIIlF+OoPf83MKDExcFzrL/gxs1vEgswf/5jI4W0aLI8fPzXxKY1EvQ2B8S84htjMRH+VH29/a4smW2afPz6EzOps2MXF8u0PfkxdQFY6GJFf8MjFqHwWCeh++GPSjRHSwkL7D0/VRHBBIqzyw0VmKMVCJhX7D5/BcqY4I4kohh8fV/g0g25G/eETs4WaXD+DQc6PVwL70uY1tB/vIdK3w8ewjuYf72Ujb6nk0Z8/1keM5+A9QIP9gqUzult4Wkwu/vj7pcUECd2uH+5OKOfFHhp2Fv/ox4nE4cSt5x9v+6acHDTuj5UR9CfHamzojPJjC7GNyZDdh/sXbE86MPS99i841Dyi8BDSL3gI5jRnJtFi+gUa0eoQc1vL8i/YcqwD4PCRxfiHn3qRhWJtRvo/+tSNHXGF+z5HK/VfzIeSD4S/ctcHFg3ej01b+fw2eGCPQTJHms8cmW9/jvv/67c///W3X/1nsoKxKxn/U2Zzxq///ltlNhi7HmqLAcmynO4WsPTXf/ftT9Kffn79j9/+06+/6RGYFMrDfP946yD95bfe1VBKMKeMJL/cubzubAX4YuTT8dZR/MvvzeIV2WoRIvGkfb17e92dCR82WSA0iXeP4l9+dwJ2ChNX1bhNv959vO/Osg1zcc+bH+kfce9BFNXOjWXI9TP3Xq97s3+rMPuen7stiP+I7VaISyjEHrb1c3fP94a7LsF2AOiAlpmg4qdjShHHtxm8LLOuz3sYeIxX+Jf32p2pS0QzEJ6GXTO/CH/47Hj5T2OTA/sIF/OEP3dAzu02OeYqk0Hxfkf6i264mTZjEchY2n/uXJwbZmuuLOwHjXcM4l90SyOW7Jzg05jz/tnTcO5JLoKZpg1yCvc84l92z2WQ3cXeL8SeP3sGzugrdklWY6CO9wziX3RPy2nRKleDtj32/r9jHFs+f2k6vFCB/6LZWn/1Hj4WQ26meMh6VfDuIUAnsqCxQPzMTjA9QLjreCRUmNKqhlOOyRcry6qxNOSRKsGT2ArlmXXiG6+hbpmQQyMqnkmU/sy4FYMuZDtlIXtYLHwWLXvMNSbTjq08U6e0mCVbaSLkWdk7yDvuZ9Y4TaZFxdwbcsykNmO6fzyT5aRK6luUZyG1TjExnP1ZKGBLG2dhlM9z2gpbGqdaxsMwE5YSqKH2s6RiRaLaVTAMRaJFLPpuQr7EQSEs9FwouVAsYwWd2PhrssVVHOR6Wu/BJ5YpCUfYrGx8niVTwibaEslNrLASCUR8mirBp3I8CAfeQ4zGodLMvph1jfcIFfTBGn5zQv1QcB/dkpdO2H9gBIPsKT7fIqAOmJYlHkYojoOm4PyGnZQwjegL7uOea3lS1Vs7/y77TQLfrB+zuvygbZrVbYtjqQ6IiClWHF4RWwfQUStXS+UTS9UIavKMe8Re8foZ2/PNBt6sN92RexGDRk4OQcwjtI5YaTzmvBB6NxIvGYzJ6Z5vYCH5AvKhtr5giASMpyzmoIiubAYs3s6+fdCYzO/uKcKOiDHlJL+h0XARkko1SqS8aIIPMlfZXHUVBbhxM34UjZGJ8GSmoBMzTZ8n7Lqx+JS85y3AtJkPdRKwiDon7HFW8TpGtlr2CxSNUYng+ypWEP0+gPWrBY/eKxWaEIrRyDuDZmhaKNbcuSQ+vRhs9M5widUzcno3MrNDfYuFNPSkEN7drs650MNCJVeH2tFPX04yGqDh0ruHJ2kY1Lo4L6++pGyNGNtJJk8fE2uWOYtzPvZncVlxnMeXfi4mq4pTBYY+NTawKIH17Gsz1oaR1SEQ2vjICsFoUPLTnliyFSPUnxXbGe00zBffnhVAOdDOxKfJk539MOz53bxqs/W68xmHZld28G8n6Ik9vLRGvTmPb+z5NQ58dVWETubEeQt7ful8TiRc6+vd0b3J+ts1jCo0gG8+r8NGQlv7ZhMb1lQN+XcXvM04K9Khoa+f7KmIAtfF4XXRAPBRUrsIqw63AZ7Khmxe8osLwfj5mgrjkW+CuIbmBZ/A9sFn5zOK3+bmWdlMWFQnljhkM7uwL7Jbif7fEH/7r/9OzltTEeIv79rDX3/wcRGKj89/QxT+dx/bm/Lf6pd5sQ+XbXGaB8u1T5fNur80wfMUYFhQwxlUAeuUawiIRZzSnrUdYcm9ynQqQQYPTRp9EOtGRiougGWsMrHntgquEWpSLGbh+eqrgsVZMdkJ7uJM02ETl/Q2YajpMmv5rqXB0s81VcA6dbdGgrW87TSEKl1jldd3YCjpJfYhSPmF8l+xdqfL371KhZX9LsNwKqGsaD3FcihDCZJwtC6URaxXstXKnKlY2zS9ohb5WAjNRMxludehbMo/nNNORqyxcvNOuUbPiiyeQh5WrN+y3Nj9SYKXTqqgMczliF66sUNr28RCMnF2TZ5qLDsvNqmLhjIWqelsrzZeI01pDBu52D/PArj5U9M2QiyXc1LU+jK7tNgraBpfjBfYMJMFHzjV/EL4pGb6hMo/G3TH1HCNABMwpvjpkcgBFbDLn0UnRyZcEASqC1baHJlwARZs3Kz9NIwtXUYvLABCGFtKIH4V0i+OLWUNYa83ysKggIh8xueByWDTK9whDSI9CA5aoewEIgHvwZbaMTX8KKBD+CR4PFvqgCUx09d8tGiIqooVkyUOYRWxuLMa9DSAXWj3YCUUbR1oDO0FjL65XQFIY2JhnAMWB6aIPfXmlEfkDlv9qjo9As7HmOL76G9UUDbeNnNqY3xHiATO9Pg8EUeFffYaxBzxSRwj1qT5ApiJqYhmXyzgnohtrDrqByOlBrs3nCqRfivFMaTFeH5xWhy7dcedbCj17swYd1ZGadO+dwCAJeJkmlymABejZV3bB4secFkqRhyjUPdA0Ti9FiarO57tAq4RZsnBDY5nu2ButObMHb0jYHYHk9Xn88TVsY90azZImDdqbPRNDEZh3ui2Bik9d5g3SsKVtNSREwaLkqsTgYv50mGwKO9+tTGFwaIk5WTVzJGJIRaf1KRveKM4476AIRk9+gzwCJ3kL6pIAJ84y2It3J8nWhOfhIPFNGUmoDsXC8+jlzcalOhX+7SfV2LAOA/2+II2zTZOV60vAZ1q01KnNxcGNCuPHGeX6PcHFQtVQVIaR6EeFO1khnX1N+iWjQc8JXrMMCiUvUZVLO4R7DtJTeiTayM4eLJ7rXs0GweCbk2lEcz6MfhzU4nuzxPEbHKoX8nD5E+OfNzZR3YekDRNFvnY6ueZN2EXCmvJAjefyZ88ndcgzhvHTXvD+ptueqO+GfOnft3zzr3Qb5s+DzTM8iQ2abQbfn6h0g23k4Y6UwKK3egZ5/YZliGFw9QbXJdLfqdwaMOKHKUIqh+cRpB9ilIA4Q/OBE3e7RlA+2ZEh1jJI8hfxJoiBIlNAcPGHGmETWwiYP0Bmlbo9NB0wImxfJ75eaWamBCZjk6PqSbjpJCejs0RdCM4IFfy00zBCbPbCJ6ezReD/hfisfJ5NmsQfAWLut/NHYPYxuLi0wsyjFtCuu+R+aIDcHWsxcwXnQ/CEFx+taawsZGgUP0+jOw06pTr+mFkJ7GCu0pdxJGd9AK7zyiLIzstMaRiQOzE4bYiMu0a5XmP7GQyoYoQ/TGy0+aRpPxO0dFXJQ+YRpGG0ZwMYJpPmYujOY2l9kv/UiNlhucR42DOZc6etEhoj7KBo1kkirGdytr2q/ysR8KQ/n732XMxYUju8iLE42Mw57apV95WdtrBiP7DXvPrnPaxbuG2yI5iu1m3aNrHQ4f2NGsWvpvc7m429hFye10tcd77xoZsnBdftLtTrtuhTOOV7aQPs8q7BY9e0PZ+tdixRzLmXeuXDj8GidZt/tOjI7CRLS/V+R7MafHb9km1oeOwkc/K2bpihyLXAnpjvHOvpHTAjaTuY2ckm/dL8cGioZOSyc7uI/hC52Vj+rwtHwh6OjWb2XgZ59DXydXjNIF3H2ij0undGzhP3ygTDHt4y01MEbPtcFRPBT9SxES/ji8DONlevH2Qchy0yWBrisbmMWizc78O7e84aLMZM2Jz8T1ncxnbg1YndPPawMgprP9jziYpx+q+u4LvjHUhNFsqNjQXW+K7KvKIzciNyPmuZpPYvMzFN5iJyUOzM9OQVcmER3M0vachceyltmKBQphH73W28o1UUcyrL1NFWsyYVyd3TvOp0bEXnM4pooqrR/zuHec00yyut9iDLr4EtXXEnnWq3ulDPE/nu3VGT9HRPtL5ywxEvsZj3ul8Um5mFZAfjfg8xNsbzUPfPl2bVJwtIfT5Z2No0eUDLQBJNK7LBBIBKpFZ+4twgNPeFFVFaoJCYyBi0khkAJXIKUrj8yo4bLZ9iCE58iQ01uSqO4aBV4GJHyyL8zmEoZk06dkpEcLQTEJTukgwH0MzsdLkf7voH+6hmYvNNz45OLBIMGVJUlg9TqiLcIRN9p69WBdhur8ki/geQzPZW0tuMZdfLBiWv+fUT5OHoZnsxdx5vFk2SHAEk6RKXhyaSc0ylDl9DM00rhmRcj6GZpJf7mo9j0MzGYYkH+gcCzichZWL1iEOzdyWQVGvahyayVR91gi1x9RMI1LwUcNxaiZLChcHSRiauTip1w1koFxpNlDPWfkCdQsUMtM55T0ck5lMJyt/lJ9sbp7PngpEMtT3dDFEJHMIaYxVe6rNKPLXtG3OeX9PwWRm+xpq+SDH6UzxXcMubzId6CROef5SDbOoTLmQB1mPdasoBHyQ+zAKbT74K5AE0Sql6bWUQCpEq7Shl/bnSUJkVbXtE0UDaRGPImJbkRDFaZfMbDelQB7TLrv5GSInitMuCUkqPjYqTLvkaW3TB1ve0y7ZD1avYt4Zd2lpg+4TOc+4S7Y5VGFC4rhLFpw4cv7z5JXi/E5mRy75Pe+SLWFMk32eFUQOGYbj5L8PFcRlhaj0nmtJ9kd2Cb/nWpKkkhS4otc6dFzVOG3E/vKYa8nouG+tcKD7Yj8WXBuf/3jowXj9kpXLfFQ0yW3eNMUm0pVRzsnSF+3ZRYfGoufGJxdt2KFPI4c8kwdvujWj0b92eKyMVuY8mw96DJVRkY3OL3MnzdVSEBjp4shKfr1WZJcjum/58gQ2OqOrxinQ5Q97HTdG3vJiI9sdGX85tvQ9d5JdkNS3YqMLFVwWNZca4R9zJ7dFBs7ud1j8jCDcR3RE1j+mWUka82YJZIqBI4DEBnhYBbP1TY3XfEnjN1/DyckiaSE7JK45h5HkkAs7lOl/kCLi626GxSYPJIrMHnn/74N0kaDIe1ZlIGm0oQDJC9mB1JGAn6VJGQ8OSJbYRtbQtMAZyeUkK80n1Lc37eq6eCRPeZtnzGdLP8ZKrjrEnxTYLy1lWjWT71BlcoQYts9+zZVkrm00hZyRiBNWpRCvLyLOx2jJtcUB9potya45bb3nDElsSQ2hfgyRpGlN1yzKMESSlbGqmPAxRHKR5jZ9HSLJ+L1cPKJhiKSRB/gixCGSNNFNHdaPIZJbj6z7BjxAsTkCum0YImluYt1vOtgsaIVOSKCVpe5k46Roa8OwSMKnVk1vOltaoZLdlMShkBxCMa/rhKGQnXSqwqk8pkKaRmhOuxumQpLLvTh/XxwLScqqpFk0j7GQFtdqOMVjLCSBVVOIpcdYSDr6Dqx6zIUkaXhzWtnAVkwYBTOQ0+UXuzGHK9Sp2WwPGAWL5sUBNYE9uRhPnxMzBrZlm++Q69fBkJYd0y6McyEH85erX/Mi78GQxknuCjMOhqT97mpIfwyGtGFgF9f1YzAkzY8PdAyDIa3mpYbjx2BIjlbtq725sdlrDq95XBMgLy5t4+KAS3gNjLy4t1n+6cvZh2+ibhLRZy/AxRGQg0nWsd4jIKmUnFktjoC0yos+VBwByfkBywknX0Mg65rlyxBI+ozbTUycAllsxpw2VJwCyVJpv6QXyToNWBrDcTCHlJ1VQoKs9ftD4p5F1n6JA2ymEc0lbvcAm7H0Xe3XcMiLU94GdNThOJhATc96wMWOGqnsjbTaya8D9T3JjlnvFsX9ocrPHASYqg+TPNT6NkZ7OrIqwni6jaCRXY4wHtbzSntPe+TyMAaREoyTAaiNuju8cZIAE1/Vx5uHyQOslzAffs12vCYV0BjUi301TDawcdlj+7qdSQisS9ITvSYn3BMVmH9us7wnMLAeg3hGg1DCxIakTuN8DX68Jjwke8V8T4q4JkJwyGfLYs2LEyRYLyyCwD4GTiyb8TmuaY5hQAVCY5UXH4MuqGhavwc33oMxGEdlH04TB2mwelGv69+DN7Ag9WIcCXM6tK4+NilO9chqyrpGOvp0EI74JIuZAFNhmEg2xmnd9Dl8BJ5AuYeS3MNKunXC+oDDM9tkcEKLT1M5g1AafUQl8OLYFCp+LLVmstxDVrKRffnIzTCSxSonCskfE1w4CdnHecaBL6bTh+ZARWQY/ed1jXA5A2UYVo9a3/Nnks0F8KmPZ2DjMoDde2Aj3SJOWPdZhmdEDr+iD4d4oM4YyvtQ8jixh94VvEPtpDCFkRo6KY0bBwKR0AY7xAfz3AOEtpE9Z5/Mc+YN8VOvfc14PPOJEusY28f+3NOMttnA5sOS7uFHnPKKLaUJYWdWEnMcpHrVtc9oJVv1okR8HMTEs3GNno9zm5IRZK91zXO65jxxtD39ttfUxc0Kb7+nSPl0qc0DubJmtp1hVJxbi/9Pg2TO6CoqOkcjnjlXrMLDedAanrFY2yOg18xFm6uLiFYoxTB0i0qr+rDxOKSLc3vPdK0z1MtGsQ+fbhqGgNnoIGLDTX6GhlH10LpeAxmvIWNmvKsvTBhKlgyxo5cNM8xYV69d2bM484xJHhaaNWvtzEijl578tIeRaonfpSvDHEazbbrxSbm2M+BtW4qstec0OFqRVZWRCqPjOLqelXlbx3vQ3LaikRTjGUrH7cyy0XTxNcOOip8lZ73kmXmXLIzsWeDNMyOP36iqmyKO1EuyxzpFYQSfeTRZOec4so9y6ktd/kz4o25E6HFPULwmAhJksxxUFScIUvHCR9K4zzBxkPRVffvU4DCgkPYMnsF4DTTc3Bl56WOc+Yfb8rayDgGpyq4MlZV+Xvp/InyVWpkDS57oVaKcfL5GgK8246kUL34AqiqZLy83AFUJTs1ZEMwAVMXNBxMRLzoU4iqxYvtFnsJ2NQLMHZF6AVWhYya0t5hMAlDVBvAKyBWAqsXeNzengrmAqsXwGKI7D0hVNkXB61oOSb2QqgQXkrClfCJStVhFYQpfeyNVOf+EHV+OSXWkqg2TqbJvoV+LMICloC30a9GHaMKgB6QqE/q99uriC6tajU9tWcAcsKoaR6NEZcCqcvGwJgKOHqwqxRyobRc5WFUyNRn5wgOqanecosCPUFW2l+e831BVNmxNkZAFqGrhgnlhPEJVqQuTUAwRqpqtGWq+oarMwky5WBGqyn4c51qOUFUSKlURGQUeKNr3VLWfIlSVENsqexyxqst4vswwRLDqIo9C0S0DWHUZ8+Abq0rv3hKIT6wqzavHzE+sauLGNgDqgaty+mZKqkPF1juLzdf+2npnWQx79wBXJYap7DeFWLZx9iIijHBVMg5i01mnXoCrEiik1G5Eq1IhC5odwaoEAzlCOdCkJdbXktLtAcNKZ3329MawJrbfVDWvBQxrsrSByDcjhpUQn62kaMSwUkVNx58e6rjMuSk+TyVgWzlCryo4D9DWZKTEyanqbmQrmbGmav2xc5ElguE0fwHZSqcW6zA/D2RrElBKCNKDbGV0z+neDmJ1ZGuyUfZWhjrIVsZ+O4nmNLD/0Q5788iTKRDq2somh1TQHGKNbIkEhNxVVTn1SFfYbcigOR6R3LAazFevF6gQOWXJh3pFEOtmBsmUQiRfpNdbFYVHqsZkUOH54nVkgAXlJJbKAG0l+qEJ1h6hrazps0j1eRJMFhs2XT9PNkoOyx12iCJ1pQVA/eK/vIkuifMTQiTCXRkfVBGRR7grAcdCAEe0KzH3XX20Ee3Kzmn5QxHsSpUGTds/T7ArlBRsu3pdA9hVa2rSgHW1mWr65JGBlIXTq0X1gYDF911CnTwRsDiqeptXCywdIwFmH0DXbSlolwega59ZyYsHbWumSXfg6mF5pSbYXjeP8NdJ5I0jxSL81RKdeb3xr+RshhXJb5Jacioz3tLvQ5+uQRKG9/sGXCyN0E06+8TFFmIRKQ+4WKjOPj2giaS80+Z/zXaR+9542U1T4FidgJc1wu/lnMKBI9gIHTWH6gGXHTbGLb05iA3emESWF+Cy7PeFInSS5BsYa9MKLgLpGxhLRAtb+LpLnWh5WHem+GwjYNYQklNHMQJmWU3oRW03L8As1Vd/00SzWEFeuOzyGzBro2zqF7xsUZfdRWYd8LKI4wXWf+BlKzVzEkIt4mVZJV6+QSKJdrLey3LhaG/SbUMxKZf3IOmmBi/tJu++Sb3x7esSF+uDBNx6Wea88LIXaTiexxi8istvHO1mhaLWLzjazp4FbziPOFqCMqBMpssvHO0wD84B1RFHq5k76wuQ1oZ2pPYFSMuu567xqBFIa8WlJALGB5C2WilCeOsApB3TClPCtAUg7bA5nz6bIQBpWbSBvZLCCEDaYUXUdANmbyBtZw66aTcHIO1gvbNeAN7Q6s6cabtI50OvOycKVHkYkZHfUH6cDmTyA7ClHJZNJzQgaTtZLprYMyOUths7tgNDI5S2M5zy2RURSmt9o86yHaG05KjFd7wgtjeUdvIoeJd9nJTAMLpqPvpjsoKVdhwQFKG0yyYNCKAUobSLRUVVAB9Q2kXEmQ+2CZjZYUgEDT0KmNlFTI8rqoOZHQZyatc8iws0a5+kyaRE0CybH7diyse4DANwLgephvEanYWVKXRwHMfBqhdUmq4TxnckDklR2eFBWEAgY1dU/SAsIO41DQflHtBs4ynuIveMoNlmdPSuaSNoloCQ7OMTImjWstTq5X2gZmGs64UJj6hZO2XeQxJRs7z+vuCoATVbKJeb+UDN0u5s79kIqNk2zSzfU2SuaTGWYHMEYQDT2uP7hPQImm30qJZK4hE1S7jUuqD0ETXbGNAo0n7wPHClLkMUUbODOc3uYOAAm232ubR9ImyWLTM5CxkaYbNECSyVHSNstglYIgv1xMeS1ULL88LHcq70Jb9nGTHfR8Ihkwd8LNvSitoyH/hYssAuha8PfCxHeG7X5GEWE39fUr/xsTduluHAFuztNetpcpjbhZy9Z0NVS9C34fJ7lhRhq2P5TKcwe8ra07ogiq9RVTSZxeU3otbSED53IUBqaYmaY70Dopa3rb2/ALW8affsZgTUDhZFfWLOAdQaxfsbT8vegKsxJuBpuUHXuKRhjBiOUf46doy5iawBmo8xZYxxdl5vOC01MFFNJj5w2saxwPCWdJkDpyUsFH6g3OcApzXY7FKG7wGnJZ3Drr6IAU7LBMT0iUEBTmsGFsrqPSSOzcmpu2oMcNpq63TBTgOclt0+Y+03nJZcUAR8vGff0YDDW5SKCrBZQt33pRMi7QkJS1bx2XoBNlvteF3o2As1awTbszq696BmWTJYuJMuc1CzTIGuoSHoETXLx4Hv8kbNNuKAmivSiJolScNWwiQONqScSlWA0QeatjIqvuYp3mhagnhx2t9sLgZIxcpqzuIDTosoSejBMMaRRDfWMe/yG2VLmz2FfX6gbA3r5RCrJ8qWdBrrjbNl/hxvmC+c7T3MkrnA7aiOgJttBMN4904cotkMB6fpRg88LUswq71ndNI+lL19qGmA05rFd5BJgM3q8lkzcSOpTTETta6RpDepTTUeIy3Dgc02+5fjTG7YLMs5uKsPML1hs43+D/zPS3yNTCXMrYrsPMBmqRRYKn3DZrlviGypLr9gs1ZYKj4NNxLvGFu/tN9jPCwtaho+TTWMk2VzWvfh2RFOu5jsUK39AZvlQc0XIc+BzVq2G0+VXX7DZjNxF7p6AM0ywwX7pasEuh8ejDR10zPKl8XYrAA/QmlZ0mVMop+fScFFcKB0TRy+JgszRcwa0hfobTdQZX1PLi4kYPIE0gN6azlMx9o8SYkQhwryEhC2WA0Sufqg5oCwLebp6zIRYkufNPkg4gixJfH9hWeKEFuYPhhxv2+A2DJJO9Vx94LYMtnzhSOJeYC1NGc2ciThsy0r4v/0mIdtzH9p+hzugLxl2qX5fOiIvC02REcAs4i85ZTP5jszIm9JCDnaPef7ht5WI6LyKegP6C0+RH1TOVmec6tL54G8tW7K9RV5O9ghoK6WB/IW25HQ/guReyNvsTxzq7XpgbzlaClVtR/AW8LXt8PC47B2Zq/9OD9Gu9Or6g88LlcSzr/yH3FmPAG5Yzv2+EyYN+Sb0l1xHn0jorFK9YXp9ZXXbuXC3t7D7q3hQcm0ByaXy+WAooDJZR6OWVjZuwDKNYqK5uOrAykWkTMtOa4qkmIZc43sV8DqshAyOJbe5Aery01J3IbkB6ubm+Y9XPILq+v46zdU1+YEFYd9B6huNvxF+gLV5eZgbkTY2wd1V8uOBo/MXezjskDwpwdUl7VpVisvqC4ixJ6c7ax5YugB4SVkh/+DyQ+Et1ijtrrkCeEtnLoghYYXGUIsB2iv6euqCkiE9jKH1jzX+GIYY8LuhvwGhrEFPyK/ob3M0fkGDMBeAxllh/AFYG+xAauOTQ/AXmbQJpnEXX4Be4km4lRkPeUB9rLAycGl4kc7wF72ay1v24/AXpb74bnLbERg7yBEyOHoEdhrB8uBrgHXW9n17jYv4HoL28G6BqNGYG9hCL8vuO8F66XtwbPI5gVYL41lu9poAqyX5o8pjTesl0aXA1resN5CfrS55ScEWG8hDU3a8w3r5Swmlqd13wPrhbEa19ThCOvFFsn4U+Ewb1hvMbZYR2GuZDMY9eO9zSP7KaJ9C21iXzrMB+1LKCpndQvqGpjm9ImlogKuN9tYQXG7RVwv/b1G4JTJD643Kw52KO3B9VIZ71r8cQKyl4UUOMrCBz+QvbhPkpMakb10z5eIix7QXs7A6BrWFKG9hCBWQgZcfkF7rVRXVV7EA2IdtrW2008gc709Z4D8UsV6WBsRvyysZktA//RA/BIFSuCjkK4H8cs2BgK8l8tvxC9hF7t9Je4jaHHlrd8/iPsmAbPZ5RfiN1Gtks3L5AfxS/AQ8ZTF5Y74JbsetY8ucwN+TbwcvX/wviz+GSrOxQ73JVYST55eLIIktbHBei52sO+2CTKCrgSs76apYv7784D0cjYhSW2Eib2Ru9sgpiIQC8jdTRUwr2vfwF0+SeEEis+TtpBeHbvJPk/awsIfCON7Y3lZqNw+TilAeYmmKlvaOSJ52QuoRu4I5OV4hlX8oW8g72Z53MeMEcdLFm4oE3vo7EitAO81XgpHjQVSxWJqf75JFWETcc7SBfp1cK9VcYcCxojtTQaW1ZMEcC950Xrvl1jgXn5oUugJZntAvHTuhiZJPUC8y9qgtVIRxYu9SJaoi/bxRvEybYxX+4Li7dykXXjNCOO1gQjDn53+IBxOnXOGSkmveuN7zddZWen7CPCl84LvLGxyAPiavoOzLJRsZsk/teXvtVWAjDyUAipozl5EBHOaI4IAIXwDIJinib21nwcgWJBzsgh8nohgfmLO1NVj3ohgq9Rib9cLKXxBggk2yEmtcg9IMEsLY5QL/OuYYP4e9kbsfwEUbGpwOcz1gIJNXH2y8AMVvFi08a8SUcGs+/tg5AcqmHWXLnX3QAU3pveWw2gDeyezDj7eNqCFhcjv8oICXpg7nB6ulvMAhg1DyQLBCzHM6zDBVOrnCRq2kq4fqwMatsuwI1ng5YAaplWu6usLqGFehomYCxx8g4Z5mJO3K0TQcLIEnc5WBA2Tsms7zP6Ahkn8vZyxLGCGqUFmF/NPgAxvFvazpucFyDBNKdSAzslBDFvKNDk49gCG1zD2Bp3yAxgm/KaL1KVadiiTzZpeAFSJXuegiG08bksO3yVAwqCQi1ujCNR1wMWLNOsp6wPd4OJloHMxVB9wMef9Ehsl6YUtXlYPH45xvsDF/G13TGfAEC+OcN6iiQ4Q4mnf6IL43hDiSYI5K009EMS8dlmaoxoAxIuVkSTq+oAfZhKeHYIS3/DhxQr0UGtSQA9b/9YcfssbPEzHjb7x5wEdJqyqlO6I5Rs5vPRnOuQHOUzXEvtwvaDDLClkPF/9t8X/J4KHjU+EIe0TPUzV0tQuGclvyeY39cUC+S3HNzqEPWKKjVZXzNIRU8wiqoetAVPMyyH0tW8QMcVUAZ67eJPfItIx3G8gv+UvurI9AVPMrDzMmumFyHLLKMidgggeNgCF4t4AHmalgTOznuDhShY1TmT7PMDDdBBFahCww2R1HeKWCthh0kpsJY+fsx4gtLMfWW6ZpHaVFbHDRHk45WnEDht1iPh4InaYvuYYX0DCwzCzVvGJKGEm4ixRGzHCTCI62WPECLM00EVhGulsyQjjmOkDEYYVHPiy5jdFiDDetpDt8fOECEOTseBi6N4AESZuRFwzESHcDHxmieE4KZbtoDgTjhAOc2VXTypdxSm0Ng1gWdoiAodZNp9Cv0XgMMunTe09EThcDKeiCRoBOUwUV1XV7QkRJoeh3TEghFlaa0IvRIQw64ujdZ/ZcSOEC5FMAnvHscGV2149dXHIMMthTaDkOJKYRdGqcxQHGFdLjlYHDt/jjjl8Lr8QwsXicnX2Rjpb5iIRgRmoNtDZ8iiKziKy2XI8cJEfHpHAU6UUE4ex0HzzNS/xPUSaeA/yQ32eE6ftSw/zVyPLLSsWeV3iMM26UBG/sMAMkpf3v0Y6WwIDmqhDI+iXZNtLLVYByduMqHM7c+2F5KUqMM89Inmp1rvYPZ4stRxP/x6OYtAmwagjSS0P0lQKNpLU8hMlbetIUtu4VdTmF/G9zKqXfQ1dv0lqC1n6XXhT1FL1kwX984T3slFsKe8Q4b3FmGtNlcXZLSRpLepljMy1lsUXKDQy17ILs4zynE3PcjVhHLrIDfBlVXqw7J0/D4TvYnFYCYQA8OVKVrKdfx4I32VRmuoAAeHLj9/qLC+I7+IwoL10kYPlFXZA+OOD5V1G+iNXMmB5mclCrCN4WwTzssbq6a4I5mVWLV+F+0Bny928iwKuCPKlfBW/fBhzs6xiWF18YXx5V7V1P6hsoX+GsncRyktFyKqTAMQHyjusfuBIuyeVLZEFwscEyO6w7IJqTRGay+unKZhQZKy1+fWOcnoic6Eih6gQAzLXsqvNWfEiMrcI5yxC2ADNzWyS7+sLk23nKfPxMRGaSxvvfSIRmlsNEjcvKltYN6NT5XCAXB2EezPcdkvrOWFlAPKyAr+dPe5Adju5+rrTFEbIrgfpXzhu6WYzSDN5gOwSdnthHCPHLc/3hbp4QnYnzpXQIRGzW61qt36GmxZ/KaB6pKal4cpfELWG6JzTGXEDopattdXJGgNydhKBBjUg5OxByGIjENrlQ5YCQpb7LBVBySNClie6yGWNCFk2xBARmV1+I2SXBdf1QsKGIUswRO0L0yxLOjn77KUHQJZYUAfUBoAsW2NF0hPxsTMbsF3wpICPJSAde0iotoiPJUBM5IARHcsapr5TxMba5JskwN8TG5tc1URkLJmq0/BfB2RsYiPpRRp7kLGk7C0MN0x+kLGEbMEdFm4tIGOJKIYmEtYoIGOHMaw2R5weZOwgUrM4ri8gY/mcBEoLSRuQsdQN83qegIwlMrI6Xi4iYInGcPrFCIBl79RSZi1wySJ6wNOuL1yy5uZNX5yDiyWiY0nXHFgswQ6zOz41csmywNO+cMkSualv+mSShfXaFyb2xsry4w0RAz2IZK0/rO43JpZJD6xuueQ3JpZu+i4CrgVMbDUmv+HPfjCxddnoN0FWDyRWNI3deWoDJBYeAYIfYfSekFh4Cj9DJJsN6SJkVoTEUt+zsOnyGxLLRgHn83lBYo0lwsQHEVutgCA2qAePLCGZNzT1IGKZ5GKaYbv8hsQ264nTJwz8spZxarN8gcoyrHecwQMqawFykUmOUFkCDjl4w+QPqCxtlI7UEyprpNaCvh6orNXoi4PNA1S2W2t5u5lnL6hsZ/bJh148GGaJdUi5fGGYhSomcOcLxSzjWOeBfEBoFW5MF18IWmHZFWQ+ELRU+l6HjAhaQ7KO6kjWg6Dll2NdWcjXg6AdNH3J4YERKcudUbcjaB8cs7yvP37gmGWQNVa+uGRvRCyxdoyGXH4jYjUFR9sqksxaJsuJYA8ilqTLVh92uUNiIWYoO18cs81YNty4R45Zbvw3ySwTJ9Cg+xpfdw21YyZwODo1cMwahFazlh4cs41F0SzLHklju0XH2jZP0lhDXV1ksoE0FkrJSVQDmrUa+i75KwU0qzVj650imNUmcOYb5Hpzw8KAQaplj2BWTZFxEGrghuVIgdSlAyM3LDueHBn4ALlaR3G/waw3Nyy1SPYYIIJciWfqAg0+QK4cWzbcFYgoVzIqDPffI8qVQ4E8F/NCuUK9OVVtALkylTWHP35AuXJIyUo+gjCgXEnn15UceaFciTETSiCiXOktpkseUK4ml12OKFdWALZDqiPKtVpnipfII5yVWZ5xyQOclT5WcRrVJ2y1ERqv3x96WLoMjX0oLg/DEke3MfM/PYclMhdVbzTrjXItLL9oUMID5Vq/W5b2C2vssHb08YU1lh86e6k9wlyLhVuqeUfWWEKwS/KpiwH+ShbIvpzY9QF/rWzAFlg0wF9tntZI16zHC/5Ka+cfJeBcqYT7Ul3+4FzlRTiBaAS00otps30BtDLb5ym+B6DVfK6Vu8tvQGu2bJ5QFAHQyioRG4AuQOvNA9uYnbx4YwOglR3kNX/hey0kVPNZNA/oKv+0+zcPfK+GNvLpm09Eay9OzvOAtNJNTAodIqY1U/Vv8dU8MK3M2W/xBDwwrRwWWASWiJBWpnmnGFUekFbuNKm1B6KVRwPvml1+I1rpPlXfsJFMlniJphbdB9SV9r60r2Syw7Km6wubLEGIc9U31JXIH1zRx3seqCvJcBEz7DfUlQlEwo2ayy+oq3K7ouZ+YF0ZV0MH6fkD1pWhwRDX0QPrmtmdn4S6j1hXUowutQw/sK5MqkynBg1YV2N9g/q96GRvrCvz/VV77Ql1rXS+Te1EqCtfd+/yhWWW4KbsBM4R68pNCO9OWM+DdSWywCCPJj9Y12RlYs28i1hXVnWx4gLVHKzrtpJO95/fWNdNsFf1locDdd3sz6ndAbk31JUTrjjPobvYR6viGsx6aAUO9yxR5XXmi2X25p6F5+ZnJHDPEjRSnKIwkM9SGdfULkTrTT5Lc1P2V6TrJHTHKSkD0pX5qGJpiJ+eY14tRhPqMbLPcv5FVTPyA9GaSWJX3+SzBKgagM/kD0ArqRT0qSOgdfAT+eDdiGhlaazX9CafzRaByo96IFqZ4hojfyGf5ShuBwY92GfN86wO/TzQVQ6Rr0s9shG6SrBeFznZA7lqrKUXeWtArpoH2W9E641ctf6QJQciIlcZVVQlhB7IVTILe8z7QK4yEZUvJG1Arm6bs6fHDNBVYcGkDwJ0lSqSvV16zANRpWojWki/PxBV4TE0puMBUZ2smMvdiAhVOnhdcy4jQpUaZqfiIMyDUOVliP0bLr8gqsbxg40ncOaBqFIzs3H/DVFVufVAVy+QKlvBoJ36G6RKzQyXob1BqtUSguq3iyBV/p7zuPX7A1JlzqvV63kOSJVkSdhgUqMBjMrZLMuHqgTQKTksk+ZLRNApT5BV7356gE65WRgKmfgGnRbLd1ZBtg/FLB2J5oPHIxaV9cChDE2Eotq29IJghKIWFh6u5oUARTWW1rzrm2LWyICv0cwBikqSdw5t0OMcKCrleFkH2B4oKh0PuIX+VgeKSn+AnqqW7EBRDdZ1X+dAUQkIh+GUtohQVHokWTOIIxTV/MGuCmjEohJHD10uuxnZZ4thuaQFn1hUgs/2F/ZZhr5LBFkPzKlFq90xpAFzapDiPt6YU6s15AtcelhmjY3VKSEfLLPQsttHjz1YZie7uFSMe7DMdvJkLcUhkWXWIj+B1B5TrTlaZTrENgy1JqJwOVFeJJ+1gTbqjXyQz7K4xNamz09P8tlmXaRCdwbuWX44HPb6eSJYmaNLWpyIYB2s5iut8KCepTvoZEEBwmqQMA4mEGz2YFhZSJjJEagHxTqsqTa/UaycTjodcXXwqlYWswD589ObeZYNhD7v+0as8veDKvANWeW/tkbeRmwqO3mbJrJFbGq3PIIDQi9sKj8vfCQxFAZoKjHGw+GtF78sB4yTyU1MrxcydRunnzqf4nzxyQCz1/554E83ec0QDwg2esNPt6FW5PMF9CnHgnPAaP88wKeblA9JGzpiT8kLzJSWQJqBXJavDmv1hVyWdr2qoeaBPTVsvqhWHjPQu5n+fLHLBpBpulGOEWTK6nnRiNAH7WwhQHLrZESQKVM9Pk7iwTvLBEAaYuo9INO97DG1RwPvLImupsKRF+/srKLRibyzpIlx8vKAMOWv8Vg38NQBptxbcBO66FJvfCm956wscECXMrME7aItcMClnOyauvNwBiZa4iRL124+ENJFPXNxSR8EKcwIYXov/CiMxb6QdAc+uprhScrFLevoUVyCWaTxAo/S4KQLc3jAo/x1dzaqgB1dzK+UJO10oKMc9gr7f9PKOnKUhXocTH2DgxxdNkJiaP0OchTmf3PiquhZb+TopEe2L3jnDRGdDEgVHQWEKPwc6KjiFK83RHRaqkx0YwEiahxAQ1C1ABFlwbz7st4IUWbRa9OkgIMQZc0dMZKjSS+EKEvTxYdMHIQomYgIFnwiRIfN287OvHuzz1rxNX8BjhJNxBlGF8nsBRxle2jfb+5ZvghcLlHmHoTosKENOb8YZieDn640ZkCI8hXndpzvQYiS1wfejTTbQYiyWLKGYz5vhChLMfAdpUoPQpSVxeIzVwPm0yIgn/Xxb4gdCvqrP6uf3/7hUwjlxBf5/OsnfYFrXnDRfaCin98Fqfmu6sjBVSvhCYi0sanjz4P4+fufv/rPXuX1+/s6/5urJ7xY+vz2xqni39/+/NfffvWfibvgoOCJrQP1/vn131sGG9szNaMx+tCDzfwSm6zvn1//3bc/+f/+9PPrf/z2n/D3f9a5bBlXTrgy1wz//obn2NdzyJskKAfP11l3aFH20yVjApi81Nqo5+8f8v/+7duff4vfCjctfB1C1WYzaktyuV+Y248lrSlmtvKv3l8UW53huz0m+310Gyzj3j8jJ6i2/qz8scDaOvcCM5CxmRCbGKrOBW6EUvAyRoRHOJddiGV4Lu9ff/7kf/zp5z/QMGHJWdfHs//JP//hX37z+3/45z/9vOS/x09p2/AuCJcTjfWf/O43f/NPf/r5b99+/Zf6SE0f6S//OHhyXJ1pvaBEAzxX58ifqxPlWB3/ZNci/fV1+z9qoZwQDkHK1EL9yR/+5nf/46ffnM047Ca/aC/8/tqx/rnOH3373/3Rt8eqUJ3CRFSjfA5vf+TPVSHF2CWP14lybPG/P+sVT+wfs1bXEyiywlpxVzHO/bnd8jdfN9Vv/+Gf/uanL7/Fiv/L3/3db/7X2VsZCzTwf4ULev7jr/7iE/+Xb489U9kvzbkSvpfoZuUR5S2Zh7QtC2ATbggQppTDRqauQSJJGKS//Ub5JqVOuaYxMgpsjJCuO1ab5ZzKW9p1P17jIcdS0H4/rky+ICOHejwH4R9WCtNz3E9NshUbMv54w3694WM9bunf2nog1mBx5/HrZfhWz67cVyaRfdk9P5+Dcg4JHDk+NaVf3/Al7ec51s+s3uPK90o/niN8l8dT31/x8YblZ9ejnPUwhZ/MvPwx298qF+yOK4y6qCqgEv/w//7+N7/7zT/9yx8uhfFfvv3/UHWH1wplbmRzdHJlYW0KZW5kb2JqCjEyIDAgb2JqCjIzNjIzCmVuZG9iagoxMCAwIG9iagpbIF0KZW5kb2JqCjE0IDAgb2JqClsgNTI4IDgxNSA3NjEgNTkyIDY1MiA2ODYgNzA3IDc2MSA3MDcgNzYxIDcwNyA1NzEgNTIzIDUyMyA3OTUgNzk1IDIzMCAyNTcKNDg5IDQ4OSA0ODkgNDg5IDQ4OSA2NDYgNDM1IDQ2OCA3MDcgNzYxIDQ4OSA4NDAgOTQ5IDc2MSAyMzAgMzExIDQ4OSA4MTUgNDg5CjgxNSA3NDAgMjcxIDM4MCAzODAgNDg5IDc2MSAyNzEgMzI2IDI3MSA0ODkgNDg5IDQ4OSA0ODkgNDg5IDQ4OSA0ODkgNDg5IDQ4OQo0ODkgNDg5IDI3MSAyNzEgMzExIDc2MSA0NjIgNDYyIDY1MiA2NDYgNjQ5IDYyNSA3MDQgNTgzIDU1NiA2NTIgNjg2IDI2NiA0NTkKNjc0IDUyOCA4NDkgNjg2IDcyMiA2MjIgNzIyIDYzMCA1NDMgNjY3IDY2NiA2NDYgOTE4IDY0NiA2NDYgNTk4IDI4MiA0ODkgMjgyCjQ4OSAyNzEgMjcxIDQ2OCA1MDIgNDM1IDUwMiA0MzUgMjk5IDQ4OSA1MDIgMjMwIDI1NyA0NzUgMjMwIDc3NCA1MDIgNDg5IDUwMgo1MDIgMzMyIDM3NSAzNTMgNTAyIDQ0NyA2NjUgNDQ3IDQ0NyA0MjQgNDg5IDk3OSA0ODkgNDg5IDQ4OSBdCmVuZG9iagoxNiAwIG9iago8PCAvQXNjZW50IDc1OCAvQ2FwSGVpZ2h0IDEwMDAgL0Rlc2NlbnQgLTI1MSAvRmxhZ3MgNAovRm9udEJCb3ggWyAtNjIgLTI1MSA5NzggNzU4IF0gL0ZvbnRGYW1pbHkgKENvbXB1dGVyIE1vZGVybikKL0ZvbnRGaWxlIDE3IDAgUiAvRm9udE5hbWUgL0NNU1MxMiAvSXRhbGljQW5nbGUgMCAvU3RlbVYgNTAKL1R5cGUgL0ZvbnREZXNjcmlwdG9yIC9YSGVpZ2h0IDUwMCA+PgplbmRvYmoKMTcgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyMTY2NiAvTGVuZ3RoMSA0Mjk0IC9MZW5ndGgyIDE5NTM0Ci9MZW5ndGgzIDAgPj4Kc3RyZWFtCnicjLcFVNVL9z5OgwjSSHMo6e6U7k4pyUNzqEN3CAoI0kiLNNJIdzciXdIh3R2/w33ve8H7ftf6/xdrAc+ePTPPntn7mf2hIlNWYxQxtTMGStqBwIysTCx8ADEFNTVWNgALCzsTCwsbMhWVuiXYBvhfOzKVJtDRydIOxPfEQ8wRaASG2MSNwBBHBTsQQNbZBsDKDmDl4mPl5mNhAbCxsPD+19HOkQ8gbuRiaQpQYALI2oGATshUYnb27o6W5hZgyD7//RdAY0ILYOXl5Wb4azpAxBboaGliBAIoGIEtgLaQHU2MbABqdiaWQLD7H0sAaAQswGB7PmZmV1dXJiNbJyY7R3MhWgaAqyXYAqAKdAI6ugBNAQ8xAxSNbIF/x8aETAVQt7B0+s+Amp0Z2NXIEQiAGGwsTYAgJ8gUZ5Ap0BEA2R6gJiMPULIHgv7jLP8fBwbA36cDYGVi/We5v2c/LGQJ+muykYmJna29EcjdEmQOMLO0AQKUJOWZwG5gBoARyPTB0cjGyQ4y38jFyNLGyBji8Bd1I4CkiArACBLi3/E5mTha2oOdmJwsbR5iZH5YBnLOEiBTMTtbWyAI7ISM/EBQ3NIRaAI5eXfmv6/XGmTnCvL8LzKzBJmaPcRh6mzPrAGydHAGyoj/7QMxIT/azIFgACcLCwsPCwcA6AAAuplYMD9soO5uD/xrkPXBDAnC29Pezh5gBokD6G1pBoT8QfZ0MnIBAsCOzkBvz6cDfyJkVlaAqaUJGGAMNLcEIT+uDjEDzf6DIRngaOkG0GWBJCArgOXh55//9CE5ZmoHsnF/dP/rjv8b6z9WUVE7N4AnIxcbgJGNkxXAy80D4ObkAXj/Of+fyP8b9V9WZSPLv1mxPC4oAzKzA/D+hzzk1P4OAMDs8ndO0PxdMbSAP7YAMCvaQXIZCKB5TH09Fk4WE8gv1v/fBfDXlP8r7R9W+f/M/P+hJOlsY/OXA81/PP7XwcjW0sb9bxdIMjuDIYWhYAcpD9D/+GoB/1PPCkBTS2fb/xmWARtBSkQEZG7z3+MEMFs6SVq6AU2VLcEmFv/Jor8HNB4q0MYSBFS2c7J8EB0AIysLy78HIXVnYg1RFifItf01BoSU1Z8XKwEysTN9qD82Ti6AkaOjkTsyZHdWCOQEeEL+QkoC6PZXggOYmUB2YMgkACRKb4CZnSPyw/WyAJiljGxtjR6sfxlYAcziQBvwo4ENwKxuAXxiYAcwyxvZGps+WjgAzG8s/0GcAGblR8QFYFazNH+yATckQntIqUOC/q+JBzLD4nEKLwQ6PUJWCEclW6D5E44QkmZmjxBC0eyJP4Sgmc0j5HhwfjLM+YCfjEMYmtpBXgenp3ty/2O0ejRCeJo7PpT/PxYIVSMTSNY8nhaErImR45Pg2CBkjR2BT2axQfjaGpn84QTh7Phwj/8YIKxNIJlmY/PkJiDMzYGOthBhMrZxejRDAjB6sjqEut0TCCFt52Rj5GTxaIKwFpF4vM+H830CIXyV/pzADiHs5Gxv7/iQi/8YIZQhiWVjZPtogpB2cLYDAyH8/iMB/4xAqIOcbY0fRMT8MWz2v84eEqPjowlC3x7oCHlsnsyGxGBkC7E6PTws/1h5/97tz604IOHYQ548kA3Q7ImV9W/rv5whkRk5PSzhZP1ohERmb+P8GCrHw2XYPa0SDkg8Fu72FsDHWDi4/iJuaffIkAMSy58nyQGJxAPoaPdogAQB6R4eiwfCHuz6OA4RT2awhSPwicdDtts5P54Y50O+Wz7JL04IXSfIy/IPhpB1giTgI1eIWDAD/zgITghVkOVTIjwPMT+tUk7eh2VsLf+0crH8Nw1MIQ/toxnCG+jgbPRYZ5AnCnJdQKcHufvTlf1x4NEIicHokR4XJASRRwShL/qIINTFHhGEt/gjgnB+TG1uCFfJRwShKPWIIPSkHxGEk8wjgpCRfUQQLnKPCMJF/hFBuCg8IggXxUcE4aL0qHkQLsqPCMJF5RFBuKg+IggXtUcE4aL+iCBcNB4RhIvmI4Jw0XpEEC5vHhGEi/aj4EK46DyivwTLyMQaCP6jhHjZHqv7zwH2fyb8WV28D3Vj6WjibGtmA3zMR17Ov8QV0j8+LXJerr+X/3NtSBCPZccLCcL4EUGCMHnyTECiMH0CHzLwCXwomycQwtn8CYQwtXgCIQyfPkAQak+eARYIJ+snEELqyXMC6XGYHyWR9eH1Aj2BEFZ2TyCElf0T+FAKTyCEleMT+FDHTyCEFfgJhLByfgIhrFyeQAgr1ycvJoSV2xMIYeX+BEJYeTyBD0oPMn2qY6wP7xPQ9l+2B1F0BpkbOTrb2hg5P+H28ESBLW1Mn9zHwzNlagmEPCmWT2J6eK3+56VhfdAYJ3sjkyfTH/TlX43Lg8D82bqwPkjLn80L64Ow/Kt9YX1QlycNDOuDvig/xdz/bmJYH2Tm320M64PaPG1kWB8E549W5kFznMzA/3o6WB/EB2T8rxAfNOhffQ83x5+Nz4MUPW1tHsToaWvzIEd/9D4PivRH8/MgSv/b/TyI0//R/rD+T//zoFR/NkCsD3L1ZwfE+qBZf7ZArA/C9a8eiPVBvv5oglgfJOzfXRDrg5L9H20Q64OmGT3tzVj+aIRYH3TtX50Q64OmPWmFWB+kTOkp5vifZoj1Qb7+N0cf5OuPfP6jUTdxdoT0HuC/vqogpfRf/NdXMxDoBjRBnp6wM+EPsioPargoFSF0ZVwbRFhYavrwSbvzPSf41WSGp4U8YpLUTwdRQ9MivP4E5an84JEJL0K6vYshN73WFP9f8dLpYCiJzhXRA8YvvZfHreZAElM7AvNh7oQPKxj7USrEaIa6cLrUMVoTN++429g3itezbpucNnR4twG2gJwzdLJ4QvBHkbSmbZgwbTLKJXCfth56BKajbqKzuozG8C1MmGXEuWhDOP6l+4puxOe++fmelK7OBPajTLn2Eq6iYCx8bA9ZsrT+6B/TZ+RhKvletOr2zwQy7dtrX6pYnB7mtyywamtHMDqacTDsA8omM+9wl6SiKpdD7tlIk0mv+NtNNiZ4ghmISEtjZUziqrZLUY4YbrjOc+qZpaGYo7jx0TmlvgXzs//G0j5wGu4tH3NWT1adSn/Ly7r80ypQqSb9HWe5AtFi4rc8jdISynCSkEgu6T6GK1gHZssFrYjp1p4tyzFT44YMCvvpzNtnk4dK43fxJJnPFMZsLjFiKuiGTb3LaF1r2AraO19DC7lJBmxjviJY2EFy/D2oTREMRGIHa/miLOlI9uYzoxaD7vhjFobmUL5827yBmTO1+UEAfT5jaxipt2Gu3icr+52CmIJReWHdeI3HFI2gNXrzbcVg4VVjAD6fFoaeYgsGP/EPepae46hTCgZE5ivlL/r539Oc1F8kHYRJ+ZGgdz2X0SHFzaBlG53wRSt7w0Qro+NCfZu9ViRrAIwuuc5Ft0mX9tg05lxnaDafHKCrTXl7Hz/utnfFmDn/bqG1R/UDnvtBXYatHUzMqAGvh7L378+4jGog0XzuGNlIHYF2np/DHmNmY2/zu60m1Nf5Mtsz/cp/dknCMhLihrT1HTMZH84WovFMryqom2oBYVV/VBgfb7RNRZ5BZ+XRp2tkdhVkKBX6EKWNJ1mJ8DyDqxNDWgF7uS6b49Y28BCIKtMF8ny/jJozUYDbfic4Xq1Pn/h+PkXQUvuqEqciFF6RDJpymsls9VVvPNWp2xAmhl/PWZNYXH5brba0nKw/qrvN63Ok3P45+5z3Mw3+CwfNo8Jh1oLaX2qDI81mv0xdgN4qovm6fnQdCaK/UI6giBHcjsJsubr/OfST2qKOfCKPMRjt1WUYcoymsR5vJXOeTQ/6yWnvzpBfwZBhd4c8ZwJ3mxCWDCFL+ZB5IQHt+VQKGCDSgQnjLN251BEvNDxopHeKg/RK8eym4LRs3y6LNW43g7Rw1ICNvsqLo9w9XkIhg60KmkMDsT2LdHqje3s8MfMjHWXbHHwua0TKzQX9F7ao69ff3ZcD0WjpOZ7Vle2UNY1oWWLG2oi/yABfXHJSahq8zPhYqDVIod1EbfI9N33MD+eVa9ptbJLsXiMmQ4MMWgSvDJWZSPEKr6ZMGK6iEFi/DlySy01vL8hx3M1dGwjUzwbMgFyHdnOq6kQ/0xQLuxGYvFT4NmAkEjp/N/Zy/lWeaOBRH5xLM3hFQpP5TKbQY/LCgsQhYRtzIRVvFkAdZsHnxKjBD/fWcNUHtRsOlaPN96WXU2CM/HjmjAntZPuODuOMB7v5uJISe5S5ctWvFE8i9N7fTFAuMj9NE866ip3rqE3WX7grkArugb52UL+4JuPE36GrCdZ3AMjMDXbPIyiFZWyqEp3dtiyGWkuMbvvq2WnNX8WMJcc7Mj43aqvUsSqo+DjD3Mhkblx4yVeU6Hc+/gtb/dPdGiUwZKPs8FdtskOT+yxy4wy+gl86J22sr/I7n5u0g47jOtI3ZMGSymWV43yBUDb1vqDj0/3XH+jQuFAZqWg5atnYwelXUHzVqC8rmcbTdFkJS1uKAdHpq86t7WiTI5SI6Be/t51/s+7zocfsW5XutfT1b4Ut/cxjSXbSBNGpPNdeDe/Xl+EKf25XO8eDDh1OG39sIUf5dsPvTfI39KNR4kI06bn3fHFzTrFauL2A2YaOC9WhHGCwaKvX9symic2wYMhAeTNzuSYWVdoPkU/Mh8bdOH4z1WNusJaBfW1vVTuapOhdoj9PJ8bREZ/gtFIOvUUPWVO9qx8RbGtYS1UUW3Z42avspFU6TzqHhCdyEpxeM/wBdrMyJ3HnWq4VX5nFmRyAsuMgF6G8MHdr+qosbDEtUEVBov5ri8/rprL9XPnhFHQWAy33DNQbJcUg2hxPin23ijsP8U0nNhEoxTYVsnZfC/PGxBFa0uvrH9/ghUIKNK6oPCjGatA/6Oz3JV7RXoxXb6P5OfmyDqeJ/2pPzheTqa9Hb1FDxq0fsfbslmPE9s2YL0tgLkUTfYGNvV651dJtA0TMbgLtuHfyn15sPpuUOSMvvWOQUZP8PfJDxpPAy9EGFyEz4gOrPFUEHG8fK4jVtisx8cXNEO8HnJdrY3iSrP4oz3eWzR25xE4zulXSuQFtNa0FFtVsnlB2fVLfnHWg9WQ+C4zSkfkPm+JTf2rpDwxM+XXcG00xuyRmsQoK4SOfJcs+uHD3bIVL+sxlUGDcFbsUorhJwJflxghH0guQjDbsZSAXYgoEBQ69nBUS64kpKnHYNhXJKx5UNuycNY6zrn3OpJWqnYEUOTuVhp+1JO4SZHchrG1fFnjfWolZOjwoPmVOuI4aohVfWdxsQCByRChn4Jokp3skG17M2FgusHjvEKDh6HrQx8K4pggfyI5Gfc8W+3KP8v0vHzHReY40OP2xXFLnL8trJws6vkoUutjJMa8zC4Qx+fxRHdCt/BHL5lg47nCjecu5mbwYxeeqOuwxuspgEy3mCIQyVW5j3RvjN2HLvxxN781WyU7Mx62hIKS+zokpl5NUijUR+jjKagNtyia3Uv9MdOFeTgGWIlzE4ZoLoaNsYlRDuYOxdWBfjr2nh5in4jC2HAGTf4w3vGZQfIeVJmLBcGOKv3eD1GvYesuxWSycrg81pAMFts3lbJf7tYCneRelM2qBowoAsWE9rFCs/iL9B5r3jb6NqM+IiNBRYiFzMC6q+l6MokbfudVYS4sgZ54yzcvVMNy92or0G/kdd1xi+YXZLL3KUZKc+1W0/IbJ9A9aryXz+ZnQZxxAFhsvD4eyUXxWgEHQec4XTdLMwFfuOzayeuvw3M+vkPKyejOqF8Wd5jGHjr86Y2B1zuFowqQeL1MmnpO5615yCjUL/upgDDW8UxsOj3id2zBQdAOIDt56NiNR8SEJ/H1iSLDXg0sg9avc9t7HTFWiPjW34K6xXis9ADrCgqJlWcB7SUpi63M+Rm0duNyTr8NAZpARejPZr2R27j40G6kmTzS6EEzCLyBE7cK5ifljHJFPacHH98Lkjj/myQ3e7hCb3Q2aDCSG1orbV4CBQX0V6ulufNc22INSxQLoFAamxiZY6D6paliaq2/kWLcUO8b5GJnSlRxNPrk4fOj5id6hjyvpy48L8LMMdnfoa3n5EUZ/EdVYZanYmGnV0HXXYnuyiW8Iv0ya8fQTPHLBR8M1Op1lokrU+j509haqmBoGbbDfdYybFuuiEW+jiuzWvWRWzYDzGbV18mvKDJ4cxQbnAJRO9VOHvJRd31lM9189aX2jb6LrndxSYOu4c51dmjvTKDJ22got4IC7/ZzfP5h7l94F5NEPzd3L5wjsNvDtDdS5QCmhMr7wtIKZeStTLQR/n7yfzYUTkzStHoF3XtijXu2UjlDL4Ne7UOe6FwBeHISR7kV1qTNCKfTT7S1kuDaDZUdjnTdioJxxlPHWjMnDiUUkNA589l774qO4hQ8WkiM5wZBGsL8vs4me7HicMrTcMpeeYaS0RJYNuVsM/LdgnR/fatOAN9m7zBXjUz7W3MS6k992zNrMmmdGvnfNBBBHit7eiUeu2yjRm0j5HARQB+CahatyNQLrwbLqnIsGn+HhNaodTvRV5cWnAkiTi0olqNPwhEosFiKqClITSnh1Ja/lLTcJe5NH5hrKGYhOkPy1YvjYasYlb6dwa3xvq9XCRJTfr5GC2AgAUkQKzvFVLPqwefWXz042szeM1p9JdjInVlE9j7dD7m80pkiG5WZi+y2bJutasbn2lkqPiyADlzKqNkAY1iPtU7ZaRu8a2usegHXGUp7sLork7zub69t9LrhxT9VvL1qo1LQuvwuWel0Rhw16SoIBnbQ0qdquJS6ZCblhrJ4SzUh1PIPoSHy5cwLlcRL5H0lWK6BkA34S2Jbp5ANl/GhKR0rrJs06m8vRqVBuiSwM343qHb5dXKEeZ5q4QylZBALg+ea9Ll8M5YF+rZhmzJKPeTbqVCkUw7blvRu88QrKvO1/rhnNgYzT+3ktUjPOAMP/c+tOdbQUfUbRIKIg1qGEpEbiWaEWsnd5AeP2yKsFac9EycHJpToH7tnySz0Kt1QkZqXJ/W85tbmgwtwq2oWoeOSVg9n4Vy2EMNal2jZoAd6WWKS/hE24y/DyPzi2v3ZLO06iQwNgdkQDFwdkx95EZt6BqowsWqyslthLPtFGB9hD6/Imf6nlKsMOnHfVJWAuCenUBiepCq9yeCN+zYMjV+wCHmQOtG5mYftTh37GyZ47ifmEVgiMCzNjtzEbRUjpdwkptswujQ3Tc09r2VoXoY74ZYa+WM9h4uv1GQtqFzFg0dwtFD0KtqfyIqApn0aUzgyB4ytuXgzaC+1RxMHTyY0+BWKoyvbnsHJGqwhfsN/1IHphTdPFg+AdpRmkG+rECI6jqj5Tyy8vpDD4SIJCdQOaM2djapL1yJMPKn64RrHlaTE0k1ZiQrOPc2B/vy0ShxlF8jaZohMtyVfwGpFbObFLRwucmQ254fLUdd7l28lmYu6brtwKgQcZOPe100gDNWW8lToM0QqXeMtLpJ+xzo1Lxnw4h+aWxP2Mx+uQ08+6gvZB/L5OsSMgxXewMK0oXpnOqIOE8lWPpUcWpVj00RshueB8uBgUol7j049HA/gHEdIoTaD1rrvf21LhA4M3IYpFHkRFDrfCAAaPdUO+5x73kmEhV9ctRE2WehgpEgeqo4TtZvpVgfde5+nlhJhJc9WFbFmzpPapw7kWKJw3MT4HYeh8V/S1J0o6L0aDhEivmWMZCquaUcC9PxA09Uqpu0XsMgNWjV9gkMTupIkbV3ylNVSa/oQ0Df0uu3zppjXSwnHMvv8Wa9puLwV/KuBlou2z5fqzYnOf0hpg4Cj/IYaAUibn12u0w24a5Bi/O3bUadUkOoMzyYvOYSpR298plIi2FLSHpTMxPlr6y3TRt4kx+Z6ZHwmRkbOcyN5LLDSbvh1MCydH0QgrGTvQewPXpdBv6VpznJ5TLMv3yvPA0yZfW62/IVdmiAqGK/NevDwnjrpsW8KE2VV8GI+V96BHDvxeiaGKZc5PCkxEvaQHFZVHFohMJqRCFB899GP13fykPvzF+Po3U0oNkKPgmOEO4QKSS8gOS90pzPlgmfL7rwyDn4GE2Z+ygwRZZHGTjZMZtNbenwfEsezFyQoYDD8/XSdHHUZX/UJ8/aO8nkn0C0ejg8knhjrF7/TXaeTGeVkmv4I7qGQUQ0nhb/odygtpMC3TEVSy0NNghF9hTy2T9AaDaLOOxexTctNbkTxr0hRPsX+iVX2CkQ5n+IyvJbvTzrWjo8oTtrJmpOrwWiEm5pWUWc7hmhvxrm0h8qZZo2pNY4GS0xce+RLWwR9VApN5OrSgrYjvpvg0428ctN/ixG57abwwmxRGILxSsX29Tij/zX4C4/yCfF5iUM9YaRRZKuaEUeQNtATOVY+Ya1CbY1vYz7SL9baMxru62t0kAzlRd6eNEuJeFWNCTvXujYqlXNwjJp5m3RDD+Rgtt4DCnuL78qvOUSTYaS6EqfDhBf6yIwUdf4s6/8uD0QHsewzGdtRWgyZDfOovOks1se58r7u0VwS4hy9sVCeH8QvPsMWOXDlamJcY2lHWvw9xbLWhWqrLR5aNVE+Xc9X97lvyJd/7fRKqNoMke7r4WoNZXyLO6FW9r+36bnU+eII+/h1eC079t+hetc3CgCrYBl/eH/wo0nRqW00iZOttmmpfSxJDfsHOQuWLNFzSHulTH93oma/SjVgznn8R9eB1N3VmOfxtgPZjpuFVKc8MR9Nr2xajOPI2zzJW6+SBL58td+y33VIFC6LLmkUzR0xxsaW5xUMGego/z6R+Q0DFeWFqC53u1+jOsFIfFLhEZaZ3726w7BClIW+yIPWGKaXlrKpgInPA//QlNomvnWZJStrnBovwpG38VktSC36JX97WJU0lTT+H4cgsZjVXSeROkZ49V7jQyDUdLN3CfvdmPb+CukZqrzaVTMGIYVsoEcOzPPbeV6OvJF9zzidkaYqaNXqbZXVs6I3Yblq/ghK6UqLPXP/Xfq56XenLYjFa0/DFdp0QH7hkBmseug8yq8smJoEsp1kKe799EZvbOPUuQYRwuXQb4wJcqKkSR65CkTsxxop0DVeuWlEu+930Mf0OWc9mdYBZBENjWxSH7DY72Tll1xmwvPbz9FKCsMRnQpUKo9SEIsSGNHa5cXIHryoM7p3mxFpG2NpNzlBqghk8iwdhv6anxPSxjnJgX207/6LxHc/4BO+OH5m7dlXrvLVnW0n0fL9UWTx5b1kb0Rx3MgU7ff34Ph5FWlKZ54pGvLuMFV1TW/ojp3Qngg3CqoTLOveKCN4euJqInZuMy/xVEnQC6IZti0Aj1D9a6L2NiW/RADzIHXcpDk3lwp++t7E7iuM89hshVI9V33hz+yfS2s2abS1Jnb6Pics13Cj+8uZIoyljYRSIqj6gc1fvdgrPtqLhpn5r/UubidDrOimcDQn1ZfVuHhI/sv6Xpy41GV7OPaK+LQOw3/w9CFG05Yel3CTCXTpCDtYyXngf4BPyGXkdvlBpq0mW91/CC9sjJnbbGGb1a5mk2u9rft3lQEgbPC3mRc+TR8zQZWYtHF9S1KAbQ4xYiQeVNtGcGJzIdGqRRmQ+7XdoiUUzN0JwEesyJZPQ3/ki6/T1q6vKeVppqaC2Z5hcdBZjuIQBr9d5PDjzgk3ExDz2Rd4WKarqtVBIM3UbSec1MMtyx4m2v+dPvW1EpPmxsibSepN+znBdbhDJPXa8+9NuGGpP/bfJ+7TONdj1VW56LNctm3hH39mamnAnXwBMcEbxVg32MAvNbN/u616T0/KsZYLmJX0E9dKTO/rvGs5eAD/6po/G5sjuawfQqRvqRysGGFmx2P7SMmN3XkiKO7hmx+SFuK/N+z2nxu7Y5UkJuCM3zFNSuysSGrZV4QxV5HjKSPUuDvZm4vzMWCxOM8oAfgLF9zxwsbns2jJVO61Ds947mAgy8WYqZuqKiC/lOHHKUD+jrBLFvnA0mKAJf2EVdZFIzcj4fkKIwDVkN7ACrBr2xbOZOeut6Fh8GSMlh0+exwScR4UOG9vUUlYOqebe4rqkeRjefEqnwUsVaTp8uxV9WQW/HX+L8TkbCvyI2CDuUfykJCyn1mih9BhE9fQYz2HxjdI8aJxDHqbE4N1JXM+obdJhmuyYzMxh/RT4mRwKo5cE6sfol4kzgbzyRbuSktgxPhfY6y6nbP7paPdhSxOW6grLQUXk61j7mdBYohFTkveLPhfkDZM5vg4vKidKwn9f2DK3vdK8xrNyLjehHN6Uq05cb/xxhMLwQbOArMUxsVYS5dWzfjDHTgMKj7g0cujBUJF2A20KW8m6DJOYYpF393jid7lDEhxwX1TSYkHZ1/MdYRtzXYSb+KOYXHAnzQyw54N7q4SgTV64FhR2oqHQa8n0ZXLvOJ2Zlsm98N92ngHFvvlhx6ZhYgFDMBo9HZjGIYJ6QtWsgXqLKWlpbVwGjLgxbbcVt/Ii/ffliWzxlf7iyPx0YRvKnzKFd5LRe1C3tnYROrcz2rmpCO2n2btfRr+m0qHYtArhZaya526C/TCyIMWYBbRvamnWSpyOtgjnGS8MMM8rcrKkfbtxA0dwT7V6YIZq6r5+qQyNHc9khOpZWP6SBc8AwZ+AsG+nH8haX7THXn42Kt2xUCJS7YMSpl0LEoJNYZjw4S4sWN2kP+589ktgjwlXrZbI9cx31r7PX5JURU18uW2sbent23blqPKGcSYiZKWZQf7nE40YX/SvsWdCfjkU8fGELyie9Q9YRkVAH0UTcaAbMyBKcXxZ5n5xb6sv8J6/cUnwuBgedP4rbLhibNgjalO0//RuJeW5bkYPBi9aHKV5jlZtxo0h/L1wAqNWIQeNM/KK2XRcU9cESXLRRFKzQU5JxZVUMR8KvwsoqJOkhK6k9433UYLF+jHJpv/wEqYS46cli98Nul/Z3oxkENl0oNbQOTRDuYsDVDPrbJKF7EOQaNa+ChMk8F/L151LIfhUGNvYl1nBc8Sdqp3vW67dlXMrXRQMovpailEw/UrHAGbN5T7f3mbquC7fjf6MaJvo3ghwOpNw5mwdzQ6rkvDGvPDH3/NXHfzeagkwXt6nnpefPgGVPs+JKfwMHrUX2nhVoAWVezpfZGw2zVYi3L2bwyCOED1RuRNVw5oe2XYsx1Az5Xkz8O5d5YJn4THLPRSz2waPLGoGzfbHl7h417S0krUNR1uzJr772XrWVJ/OCIwSf8YBmZY3DU1kVD/hsoD9g1JkxsyGA2+9ffVjr7PZErxNx2DzJvfx7A3GCW6+lnF5/8jjOEodZhchxP96OkvioYtvUGK5WiDMHb98PwPLzVwCYPkBXedN4TwNr3goIUN3OeRaH7b09d2ODtfVi/tYKESORkfqce73iaO8/XW1jQ5qOJlazlft9hgpccFj2xeZxm9NPkaNy/tffj+PktREHpxxGLTrPpl8A/clKEKX/X3s+98A/X5HXuYlMJGx8da7kppgrhgtc2/ovUUkFmJDpNfV5BXv4/KxO6RFlV8S8X6E8uvdvEDpDRVKkp5+Dg65+JJkM0AGVDTI3HrzG/F7XdeE+A/cLIWaBOuIYqgNClBsBzzLnq3ll/PMA5+9yUZxdTx/hlPFlTTDyJ03x21hHq2Vt7I193UVsgKJV78zMhv1Z7NdOdcAuflLBvdHRGdNVp64gLYW6oEI/w7imJDMKAeLgDbBV+1eMRiOp+eRXWrBPUSv5HdaRCo5HfNiP6iyxFX+bIImqRvjz7QbVvNXYGHS+mmMNUQ79prNdOb9iVTzy+bPNzs6JolHzj0f8Rwz9Avjbo/VqNZh7lHSGwtkh5PsUJm36gPhuN+GvHjWOjDbaOWorBQlG0pAOwV/kFHHU9fNl4+hY9dGXYkHMGHOJoRXgw5uSnmRmSueC0OQOWZMfIVOaL0AlQN2cLrgnADgG0jDkIbRccQd0YfO1EXMIG7rjBJ9y+naDvTf72o5k2xlE1H7jM/9E4+hhYSkbvrIIn1oFurMyedclHrCe8jELi0ATuI94Sdkq64zznsScnFrbuJ8gUpkZzhZUGMN2fvRBK/+0++MJptMviueMc3F4jkbPgl9kph7JXfyZ2IOnOtHidscCaRIb42diR1MzT5CPiNacLhJq14Foa24VBCXmYiWEPrzXAT/HhWt7gEx2GW+RdWqNCqi8K7yVOzU832Ff8i5rRSqO7+bxtGI5B8qcCUhRfS8pS3U+a5itK8HC1mkfTerxI6M+Cfy8agshyr1+zIawxGNN3sk0TqW8Ti3DeO3m1fbehbog7hQCcmmCIJxMXF7EwYKr5nF9MlBhJnjgdt3NfNrC6U1wZTmu8HmWu81+lmjo5kUZAQUGvq7X4k1WGMQjOWmyCSN7EeRHcMg/QjxjOrRH1MN/+5+UxQZCrXyafxMaM6uh/BYJt+FqNFIhira/QVFuuYJyOisqd5hhszfJHbpyja6qEtKbCNhrj+B0JpDS4jRcLCBX7r6Y7kv6YglM87gFsnWcR6soOhYj8gHN2isTkeJfe6XbUz0UDfft2+qN1kYJy0YFr31UpdiCL3v1gdiglvzfg06Eoau+nSmMXyJ2XMd2P9CWhdSW+gp8VEHiwALNbsC3Xa+mEBmcskiYfzq+Td2XBFv+U9R6xGo3Mk3Y1vXdqVNK7dd4StyDb+AES6C3Cl+sF+f9e14Jcwh5b92zY5VVFRFlSBgSeq1D87cq+FAC2uozBBwbB5A71JiHSbppWyZzfw2AdtcWonZM4KrPPsuoKQRAW7KMrs/16l8EoF4/BhhYcAav1SlTUwdh1H+06KUm0nhJyE/K5TzN2Js5Enqb3Dz38FVz7ibdySnGwcOf0620YiQOc/6CvyA8HOm0umK0d+gBkwHnLgpIe3vq9jS7e4keaZewnSy83q3X+A8J4AkPPirGpf76kjEayKbHorGxIz4cpsdxmdLRvT1ym5+7KAKjDM0z4w7N7Mxtd/z7+iyaKqGS7zgu+xQPl3HdX4MJIQrMA31TJ8Cnn3b3it+b7NEGd1RQIuZb1GVR1qz3cA2hQlya8GYrrXFqFrYOxgYTfxqmcV6q3ZA7RhC24TMRSrEjyqljKn6zEon0/dVF65kuSQxQD2e2xsn5XiEwmcHSccg+etbK2a1iMhLvzOUjuasQWjXZU9q3HI8D3o7KMsjZNHz8pHOXGBuYhbXBkXMa+sLpSiLsSOvdwwJQ0smWR8KJ0NrX3kyDQ12JrHQKPGTFnPXuKSSdx6hmMpluLu/FxCitfFOBlm5DLXR0myjvN3WY0OfdJyEPQMFoc2suBFQhZQYeHfqE1yt662zHZ6+rKILG3/XYqdSLM+34LFmPWaQ8C5giP69dgsepWPtW7vnL9/ykLVlnN9eySezX7HaYGnZ697WZNV+Ky9dbz3bMif3u1XulnbGMet3UsjaT77eRLP8NdS2s/ipFvdd89L3+sySqazPhSeJos9qo8UEe9z6KcQuoYZovAdPosyXP7t+/mi5IyYeMcZWDyd4oAnM90VB/flti0DIqLveQKGVFXEVras62YXb1V7mja6bHBFpvgYPzX7U1t1qLJVbKm9tH7gbDzqN68NC1DB/oyvV3sbz9iz7q+TbYJ+k8Mj4vRzdjy+Tv2W5/khU9d6sIGrD8aUEp1NsqhPsQmXGUcESdY+LfrWwtiKfIi2oKNQ0PdPfI1wboonMZdayzMmhyuxAN/jhI/mF5wZmj0GuCK7utG6VuNVwoDo1XO2nIhU/3VGTVTbH57OEQCsyKAS1QffdwJioLVmPZIYi2ROdicuiHMO39VT6bfCDkbf+v5/bKkR/vcnEQpkKX4jP2F9Wme27EPPhyDFwCo2kTxuo1rvNsvukfZWydB48ybhdhOssXIZZXbqt3EIT5PDGPeo1l0+yPObIb5Fc9kTVGcYzLp+hHeKNQte2zxtoULk0hKdmQ1B6EoxiAgayxQbwsO+fvagPfkPl3tGdlO7U63GOl1UjURo/dAAckAu/CaVj+j6d+Eb6XXaIE2u9Flu+D1dmQbR4Ttqb/ezvfTQZXuGxK5QOX8NEoCkwnF5fq/iXFKBrQGGQbNp+vKlfCojMHA8+z9HuuDDUpdaUxe54y94nxt9/xi6rPQFDcGuRUgLij0Ic/8jst5doK+YQ29hxt/zWGFr2PqYeZ+yTK33DR5LVLwpqd+swFvzMOtTf412cbjGmAo8jTrBIio6D/Z5JOFRiWM2+GevPPV42mLaodhapc7icL4CnNo1MZYvbOJPiYcszVaSFFx0KbFv0zmPs4F+TWhff9SKIfrGjRuSLmtaTstkUSMMI/iFKPnJhP9+fhRcB0JhQ4I2pIVUYuS/Gp9JEJFA6dxQ3CbgsHr4P5ArlnVLfDMLrw5X/epUUxO+nsyrIY+NIGhRQYWUg9hVWWcI/qCRo8tJ58ZIm2lCJCjkf3Y+HdyPUm4g+lXXSCQp5+cr08Dv8nK6dKiwZiXzDBMlLVXHQsWW/EHNvC4o4zHYx+V363nHNts6xIaXgD63Z9VY6nJgzurWZPadXLlDuJsKsrVFdeB7PkIN1r0eJVjqBsSG9szYaJwuf7zLVhaiul6BnL3LrOmD2Op4JUcbV+HPxq8QLICzaui6R+inkktNnuandPzcQcCxSyOWy6VSAUsOvg3vLbrn8WsjqwE09Vlo6pZ57M8RXXMHK1diQjWxfjh2GnP/KBvF+SoiUy+9IHqjFU9gidkSRiYnl4S6uQPkuMZCrgBy/YsXc0uTY+6b3cC4piMq48+uAePXbfguOAlNLdb/z5VfwUXIWuEpBxGOqY5irNzT91KrsJaQ6BB9GE0+bFoVewhX87PRtSB6rx4Auumv5KJ39wXVlnuqwxT+s+k75N9PnZLu3C5hZ7Z/I/XPm0vWd9cylMZDDKfXtD6Eg36LZpe/GVUZ5JHeY7/D954QYT7ICoFhfV3nqGMmYtyhhib4AlR4Wa+/ROclpFqQt6MyM0Ebb7RZ/JHmBjyvIuBJ/3soD/SwyWAtrVKTd6lPcVYM/gv4tDwtG/+h50wGdxxLT/YssyxvipKypRRKXZk7hayojGOaaqyw4R5eIl5SFRgIo/MbULUkfJce/bUx3j3wkIyLEXu5/mexJPpO9Hqo8HglebMQuAKnWldRHYoJfZqzIaXaJGUC/vR8EgumkuA81PZd3GnB84G1Rm2y5oRzquk4uhPjwdBxPvv80aVq3ApyAaaETe9mhCLzvP55IvaQg8rCT4fai2ny5mmPEVaaTZuHeKvP+kmdTNwKT+vrHC/uBoerWMWODZ3o18s9hrzHu59DvXn5WXWlTYi4Q2nbDw8vNHWnFSFR9hwrfePG+P9WXQyYPPzbimSdj8gGPTwwXYxVlT7ZzCJyUsBs/zzt5ytqPLGP7JX4S7XWEMDpVjCe8ZcR3w1jYn5ak6Kadv12rFpBWiaxuvzVScZPAt7X59o1/9iuL7Er2LvQCy29avGFV2Rkufx8k3e5u5h9XU2WJ83INglu1zj+bxpIzAAntPiLIJzBToq85acKnkJJioKauflHHOHw9/AAuJ61TFF5ceiflJQ4lpaV7w2qUKtJd53gUnthM6OZfO4xKvKkZ3dAy5g9T8CHpHTKCZrRnT1Rrvsjn5yXzrXOLC2Gfgc5BebNxArafV1LPONmqQvdck1YOPzLoo9QL9BLhpN195jqapVGR1ybhTcCuPZa7XDfxGjuElsrX+oUuHp2dvNDCua4I7+MgKEzm6Vq9Tp0LjzNEXsRtfiLv8VGrmQ/zYsL71Hrqlf4wO40zyhcmZOZTkQFvTBi7YX3LOl2VZILJyOyroS/g7os43asyh53poRP9YOrCCCYLUdA8+Ls7J0rh2zOXj0yxdYSpwN0IApYuyub0wdG3zwKTJI8TA8LTyZyJSQ/YjODQ5e/eGTowIYlXRwfzyBghxzo5AIsotRSQTtsWUz4YmDpV2wzYKHUMY8vdjVll0nTYcyQXe+0v/yTC5lUv3+N+p+lIlK739QKrQW8pczjB7pgeHgntW9Mho9clVIr1YApycocnA7BbplOqsFKSPPf3PBVeI+H3qTf5h19UW1ltp9z6MsPsjj709HgCaq9FNmRWRWHF9QLYPuRVH8uwn2L6Ybw1IvGkiEd11oC/8JxrIYOmNlmscmBpgJeeD0Ro02f8FEtrtIzjdjuqdS9ONxE67qmNST7MUTp6XYwct6CWD+W/B1B6tlrUlOvNj+x9YNKVdDSfJ0qQi0PIafJ7lgsOptoSeGlP1fNdD7EYCT6yn7uIJCbQmnyywcR9jMqkk7LXw/HtqyI9HeyIBuSQYCRhz+gWGB/2JoSP36Tg5/eyaDhl6uOj+NhC+y/yT/p77PYZsRNdy8yPSY0CHPudIF+KO1WqCD8nAr0jkwwNr/sdfySvFlM0BGlNedpwQmHBJ+ROiZQ0CS8J4CMroXX1oH4x8vS1TP/WNX1D7/UFQEavro13BOPf1+YvzFqvmuzXBU2MAGclmUtbTUaGvhQntdacNBRsSg5/dYWdMBIAfaJ86qyB+ybTpEKFQK8boGEoqBsRG2vYTBiwnItuX4BB3GZ/FRbqn5MvsBbwrZjGb11yISg2qTo414nvB9wVxaG0BTB91eV3hjo1Qblb77mgCU2OQuZFtgnlSJyT99klsXGexbJQ82Cjgkgv0nmyf8oBnHcrWeICwCi7HtNxYLJeKRBD8dcgdwzZWcB9F8Ew7BL/2SAuvYLNAubP29Vaf61Z6xy91G+CObVObDTNNIusyewMYhohpPQd45TT7LHTKEC45w56X+UwmxRRGV21aQaSDwWBO2lM+hKI5F77ppZwRsl8HzQbCApeVnfbf9HnRhNfbjzkXi4OkG2uW4hOhiZB1BBuWFWLld04SHYVmAKxkcDeu6kLrgJwo59xp61na4BOxr4HmBh/scJTFnLTlWcyYfs2LKoomeB9tfHdswps5ql3vbSSgXDfrIKpQAOC62T6gAPaqahRGjuznxYYP3T36r1pX4xqYSPlUHTAHgiPEn6/Mzv3zhd62dgLTEEaCdOG0Pypgr9tGDSNHljK/ZmKRXCwq9ztxaBQ8Xm/+XviT5bDb+KlRxh9K+uCZ9LZc4l14V2PSpoKesSLYMm37OlkPgh69diwVSG7DSnBMxi1xMEsBWyVbtgpm/qRrSveJ9S4i9LUOt9uoYdtdl1LAttxztcR0FMwtwOocAXUtH7UuyffVyx/fKcd+JWHzcJLVqj+2fts+J92lMvlMi34Q0rDklKzC3MOKzYWXBb4FQHxvwgzqPzYlGm6514jRRDzxdCGQ72pZUz/1hNoonZCsGCSErmLrgkYVKdzyF5s8eYqjkpnAlBOpxH37bb1EVFZlNc6XKi4oDEGUZW83PcL4fUTkKyYcKRaec5G+9GH0qC/cSkKmavFb4Iq7/wwdX9YL7NY66KIJ/Gci4k+zwN4EZu8RdlhwVc81R2/bleVJozb8Vuerj/Dp95JaLH31e+gG2ysEuDFtg1/3CqC/NicNiyNhqWDhxSHJNoZItRzjfgo5jXnAY9lloCwtdpuWRo/0vVXO1BBpyANbxHl4lcGi/OI+TFt5PG2l070EbD74wISDZJwvw6CHbMLvcqm1x18MIbCMDqI1wIv8jTu+zOK71xw3Tl7BGBwY340alsdKJvV0DVwDwmdUJFe9t9TQu06x2EzCnLK2chtCe2dbNX/jsPqciQwBgzjYnRgWeHxNTKnIiS0raO+N1QPYQslcX8tI9x33kQxdHZz5EqBysFtx3kDPdOfjJrtqq15mqoW9UYAB298RLSk9nU1lwZmOPulwEHKtgvPSI7G9DUMs11SHF+ny3t0LSFomve3Y4RK43OFr76hpkWmyCzqzVz5ScCaRPdE+IgdIywh9lK0FXo4uQpgWYZ+8WaGduu7zAnb56Zy6UP4bE+F3IcZimHl76p9+BMTU5GIcLjyoDvhTLsy90TTj7RhOIPfQJWvwLm0ys0kitgXvlNroqaLq4DUvRa1NlLQt2Us2mq+P/oLSYfg8zc/OeBUDrUu03Z8BLliRZI/7h3rjZlEXx5QpTnBHhuQyMLbHTODsjLlnitBpQgeaqPBjX5fnJsoL7DG6c7zrldosLatnpZjOv3aEvT2gPiHUVPadCjsHsM0c5zYZe7wV5yfX0qvSuf40l8ei0et5WgUh9FtLaqBO+Z/FkdRd0+q7IuH0nDjhFy1+rn17q+9RCM0g9u8qBtmj8GXIJmAQu+QIILSNJTXRoRSODffFbL3Ehz2jnqRzxt+73nUf37r4c8niMS/8OFO8rendKZPOzW37qkC5wAZ8K2hYYYduBx+iS6tAIv+UO9t+KyU1xX4Z8Yzj8XzmEX4EHHYdP6rFwMuZ1GGmorZpZ9bYmeKeN72JB0mkll7MAyF3OuZ687yqSa9C33DdZXJUEQi/SKgpg6mtjb9yKw9GAu6SL3bHjCQOvEc+kN6FCfZLzufjeViQLyptB4h7HHCuQ2H0GXzugjKniCUp+ndxmWjR4UlX2A3yH2GYXTp05bvckQTq3BA9hvJsdZxF5bRUic95NArx+ty7IO5z/RKKBVpBljOWS7WFrecKoGFnHso71bOTLmtNXfZ0/U1ArS23Uy122uyAX6c3lkzyYfzGDqo9fGFvV72LKaKpyIFqRfRduJJLM5EH1RSt5/dR1Rn6U1ZmxY4joVl+kQHuCQr7KdymHRH83MC3Iqm54q6EnH0YzsBfdPIJLDowJeOiFuUmw3NZ/u0LB/IOaljpJdRSiO+kizoxp8B4DoDs2acpfv56UobfJbHF1swXDa5RNi806emtmVDIpM4+8+uYQ22X/HIbUuN3bCeUaFaR8v1O15ooJuq8744bDj2HOJcS/DZVD8wmAjnrAygDPi9bT8LNO+E4vx55cdOD1sVWK2D2MnfxjwaCgvrv7korM7t/NpVPngvorlXP9M1qVFFaL+Vn5mb4jho7ndlTGDkUmU7aUvIb3UUZiP1morrfhTksPy6DFSHDiQvSsS2PTb7jHhT9cjTzdUPDMxkVqFVa+eI7dk/P6PXSTC252ZG+Zm7z9pW0vTbWixJgwtglcEnJXuf/dFwI/bxoM4ofg6NqBsGzNX3TEe3Aa7R8XFyfTyUJBeHUM0S5d40/Ra2MnFQNTHaZeBOuv9gO9XisnhzCVeOy8AtksT/1cV0+n6yWxkDkF+FwqlT9d1gv5ZHoYteoP3Ny9Tq7dDAesa5b0FSlAV3igdlk8ZDIeBwubTf1d/bzGtyc0YZe8EouDCg9wSwSLbB+xzzFiDE68tpEsSMQn+/WGAL43L+W7qaiy3JwPGsVatmO5sepSE7cre+8rTrgzz5E2SuM8V5d0Uz8krtwt0dUV/atvKM8B/HNwGKluwS+ponGlpmrmsju52Zpj/M1VB6Twje1DF5RS5m3+8bWbvZ0l1+K5RTeM6z+HaB0LJ2FPjjSHq4OVFm4cKJKXP6RIUj8fPnmTZr6dVMSVfOcan1Q+V6ip9Ti05jv7Zed6nE6x0PWiMcZFzJ22AEcIgIwL5kprNV6xK18D15MY6j8LxqqskCh5FPVGPTp0PycGHAo06UPT7rkzNLOYO0wnu7lSBYU/9+1svwrtR4p+u+qK+lUNcA95dmB5vPtr1BskyWTRam7ec/9L6zS5q8s44abs60pt+5nMMDPQ+zyJt1r68MC1zT2RyEdRuPDNoGbDglAdX4cEf6BJISp21vgY0Un+LHIh3EiMgq4jgNA9+IzasIazNxTSDj3O5NqHwdFTsPmw2V9VfstuC319xDj9zVgYlokgMnCN0Kr5BU4xmrAXuD7EZD3Fd6rIigl6d9K3hZJ7GTCuvntzkuUnuzuiSb2mln5poW7bANsD4taPo1dfrA/7Re7ZfBy8QV0P2xCtZ/PWKvwdrqiGZdzyFcP4jFEiMU/eJk6jWjx3VVCewkS/QPOVzKQ26TQidxQdW7pnPZyxP/Xdnxr6UVPOfCnMDKTBsbGh9Z3Jnud/LOl2y0BYpGeMdQH2odQ0iexX9Z3uADmXwBEeGwIrNqJTpRXll0Trw/zaUpHFcT7vQBLSjBJxMFIAZ8phCg1Yd+JV45zGEa3S9FrzJZaN0Rt8hgHjXsie6O/r1uCIZyvFANijhzHwqa+XQcKsIrOiw0I+Watgj1lNcy1QmVhDNn8tk4IekNy3hKUDl5uTsbBtJecpPL6Cde/S3TkNXMU9B+ek0WllAFTybS+3nTCbgQpwFKC2WP+aVL54JlxZa8bZwYGg7xIG5lBTcA3OCZclLKSJBI9jFXRV2vxw0YyyzQ504hzlP5pWfm++eS0PF+fe7W0MQRScHUudf5ARrvbBL4Xzp6aenJiFvLYTyDwtmuhkr9/vwHc5RzW0NBsDsLmjG3o6PUYDNjJp5K00STVpIffCaY4BSPP9lMWqI6Gt0KrLxvdGDz8Y23mZVgsZN9pl75noXtEVSOhnwWgrLYeEXrATFJlGZ+tP459SboThqd2VFnyN5HMMqJosD2S9gnVBxFtv04kvcvyHH34m2x99cndD7vZMfo3+KctvoT7b8G7y5cIyT6UXUtK+jNMSiJb9a045BQ+pKX4mVKUN6PAcOYgyUWCyPbdfKklFCQm5LK2yLQj+yswco/29GR07upQL11VBTR1RipWLjJDkFzFODbgttDJMtkpx25dfYU8EdKsTc2Xx17tF0tjFXdb9X1pDvcUV6xVP3q/7WJMjzEe5GDbnfo5NfmfY/elHlzKeYUR8uf8Xy7BHHbmvQ3123gFh8iKyPiFdICWsbwjCLjcOAvhp3+/mcnNqiqcB24ukZhJ1FiqkwxYHMou+qZxSu5TOKVVIxAPRYlWlZDH85+rpDfEdiNQwkGJ1ZphAT00MPMDWFXky8+5+vMLfUVhu+1i7pVa73uVMX2s+u4s46Pj3DHQw72rQtF8jS7RAFH3MikQeufn7O3j53KuTgKry2Rl9w8W8Uac92qy0ZCvjreFaWv2gN0czy7M8QOhQ6JtwZ0jrV9yGHuoTh0jhFz+IKsUH9FU2843q9UZTCS1BVnThyNoMvLU/gu0U84XtXOB2rsHPeq5nfUOD7sGw4AOuloIygjPT6tQY2X9cXhybctgl2Y6uL3eZ5kY2/PY+M6g57/HMntWCMmWhtxaAATfeHAQTosrHJxERRbspMPau29ckzmb7MumoKhDHt30C0qDyYR6FN7JmcqJPLz5HNXWjZed0VQavExHr0w774KsWz4YMYBrhbVs7JKMwn/kI/Vbb9+rI+XWvU0dt7zIB3JMbYzGY+jPhs1DVa1ShsUzaTUX/sEpfFSOyJeZyFmjLBlb26nUyYxKlnfEZiAhOfHQGwW/n3GTVltkkE5i5HH/UVbYvjWgFzhEKln+miFMphkl/8oGx1d1B3m2iR9MSFUqC/Hi/okNN6oYfC3auLxuziUFdHn4weIlwG4uIATVF3b5wXT6ev3JrTw9Bf4QWS1ohU6dKNMtoPdAQblX/tTict6WIw+s9kDbSWwrTwIPH+gaYpc7Cvp/5ZE4o5hWkpaT778SXjlWaey5o7dLdRS2pKuLDeDnhDhgLfgjzDqz0ch3IAUi+RZ4+KXonsq50oG2BiYQFCdhIlsUyBaSSh38/Zy6XSC4cIdbu1Cb+FGjWGQFQGguAVrGM0mypRmTp0STdDWuxbpl+AdEwrYOksnXXqbf1AVNTBDRQBIbZhM/safSiFRO0h1ggdvsCy2YjFAZzAi4Rqeeo2cz1LORtlTtpNyByKO9rQY+p1LkEd8S15co8+cdh2ty/QYULGNzRwWxw8oXVR8B2pL83iuE7I+8kbaXWtL5q0e+WrPe/yAtq1uq6+xJ0sBqXpwrF7VvZFYdFNTLLnypvM9cjFIcTPoqAtOrFGF20Ilks5DqvRmbVYsEgZtdtuSvWhWe8VTk/vv3UO+qyppLNBc8ZVUg1ddyCqdYpqB9KJNsNiOB1sKAYjHQ+MBqr22ehfDlnPSKgkx8QGcFzPWown1AZremvyc3eioXdir9m3HXDET5Z/WXuj2vkc/4U80O1uiPsj5xkSn4lQ6Vc+HumHGT5XEMLoHHD//dSgGo8tyPZZElNX3XYp03p36Fdxh0EaNYemSvDtX30/o6tjZMYWbHdlFU4IQuJvnQeG9cQJuPuvzMq/1Cz+UuU05u5bb3fzC2Pshy40YSS/g3rLX0w/1/IhP8/r6jj8mSiP3pOWATM6CsJDfOZkMlUXa1z55glBUSDfd36E2+iC5X8KGi0601ehVcNpoDFQazPPvV/7b7d3oBF+qS2RVc4IXQhq+OwKzZWc5QDLWfT3P3eR3rRtzoN8ThyLlhHvDwF2qogt7uawcrFvmEfGXiqX8+CL89Ysj7TfdIkHtQhCy5Ad0fZ/UiTkbnTVb4nsysIIy9htYdNmkiku9CuPKsfrxduzI+YPOUY09hbmzTZxXi47Q6/sFWp++7ALn06T0kl+tbJJ8dbYU4ut/VVlHNnUaTUk/jYpFh4zyW+otbgWYhhlstsVitoDlihsOcBTO80GQ7FrX9isOGcnbDZ8kF8oKMhfeqTeh863AyVfAHeC4DwSOr9w9D2ZChfsE6FeqN6AwY3wlFNv9W8q27T6Cv7HvwrddlT61Z4NDSn+igp5+rkmst/umGoRWYWHOvdN85BvnlpAw3MC3ji88XNRTsQ2FXrGBBzJhK7FvzE1gtfR+zTe+jJglTxWhglWaHujDZFsPBUu85iw6cmMvy+4m1qWUfYRp0+cTY5KRgYFBGDDrbSm615VFLbU3RaCpWxjJuEjUgn1AqNXAjUZlPK8SVMrErdCnDzusXHAzMArvxB+Gm3MAfpWGb/VCeobsTeDUHf9YU7gOkhDGOX5QmtVRe2h19/rEu1/LGaI/I1XzuBP88MtU5Qps5lqFhfLlk+MHGnr/XwCvEVDuTDE2BWFsSwEbNAodwHse91SOLeBklAj/dxNoyvRbDV3DKP0fI9psQ3lXohPHKGrqUlXzaLoIQsvE7fr9aTMJGhlR26J/gNdUjSNikJU5+LndBn5zxjN4VzynjFXBYfSR7uhO4E8+N2zy1Tc1JDAzIfsiqgmg4DQaObW+r7Qz+etYTA9QvrBmKGeWMsHoxCXcmWxYDDtWaVhSOZdRvnh27jPzqEOHlTnY6SDxssJC0+V/LzJwMrtlJCklZ5gRB8vFysrsc8/ZNIsqyWut+gpRB7+Ckt2EGrnPDGp7KMRAsjP+VpxfFjVuRip+gYAn90HsVMtCbV4FUv5R2ZpzyHsSt58etOxH/X4IiqW8xj1thigAqwXN3RlmxzrAsSsSbe/x7q9L2XZl2S1L8Pd2k0AkMRFuDYwR2fit6y0qKrmVWN2zR6r30QOnm9zbvPj6uBpHZNg1d/+1C8UkEO8u5NcJHI8ZOw51529pw7pK3BABKYLNJ2dM/9c8vAxxw/In09y9LW9zjvr0kLAKUDTA79W1lbkSzUpsOdXJQDsKffsTEiBVSHnCCFh3CfJeuYMdA55YhUVyxKJ795F13srdP1gVjzMgWxy81lLrv3YpnXiBVnO0T8771T8eia8uRd8ONAbLO7xu+qDu3U7Q/oxOsDo8BmyAuJnRn/JSN/w+cqiFcaXwph1PM+Fsg3Puqsd09VhJFVi86CCO64y2b2EwHnJl+ys0rbhIYcJ+ARthIno+Ro/2Mp9VjCOywp7XvT56Yej9PK1Tf15vgedip6DEh1tAaaSw09ilk+3rXxeWLTIu5mSennXpTEDPq+YXRikWh6NRoclGoH8eW0ULKMDnl+rqimt05ZV0K+58VDd3gLD9LGPzbEFOYpvkRy2RsFhy2+rS1kXJ0M7HYwbAC1jtVPVGpM4aZaNQD05pibW3qsWGxU3Uvby8ZbiKIupCsaTJfWUHrYQa0JfbLdcNIqiNyxNzDtzJd2hBFugjAwf1dlQoITXDJ+6PdSznaBwH8oAPNL0rox9BYOLh2B2OKOp4/BCQuih0RCnN2cD4bw9ectaDXIhV2Q2qgq6hRta97brviK8Y1LDzon26N3Y3aD2//cF2qHipQFubDFZrbUwc87W0+FGun+vyhF3hFJeeBAOovG6e4nLflTCjX1RXqy7NVm6y32VQxxV3GSOyFXa1B6FqsYa8INSmYWvP456wOlhiLWSqwq7xAUm77hxqna3u0k+TtRByIyGfotzkUjStJvhjOw4jVjISspnClkCZEnD+z+deEP5HT1p1bGNGL/ze/GaChzbr5XHbjzilGHnE2HLjp3JnhqH1McLhjQrJOZ95M4yv4Q3wrT/Q/9UL3ECbtG7zaFImgrxWScewO1iXdwyUpLHImyI9ugMzOvABMPhPWSfnQECwoGojdSwvMNmtYnOIgsHzzfMnCLJwjVxOnIpdZgFfMouhlc84D5VyJEXEWBd3vlJgyu+4u/yl+96k0nWj/lgy1i7uf8A/fEknQIylAyi7Ge12G+kjTqcXzK2aCrcSO9tNanVNNOr4+4jRwoWkamKygWyU4DaoJiVirB8hyiYyvgF53rrKxVpcwEvV97SBri0rul6SwVDoYG0ppKquB9rZzkJpcOz1ryYEA/3FtsYaPjpj65L4QsodiSD3spOp5KZ+psXsm9DFiWGjM2kazV5R1MVFtGbrJbS5HhAjg5mkyKPVoJ3tX31Ylf4fkENB3exdB3ld/OD18PFzOdsFrQ6SUib2slOVs+48F7aZLaWaQtnMkFtuLQ2LAEU5mShB1hDz3lptsalvZzzgxOOVZo9R+Y16n1N03TEjgszyLiXVpNvCcJESk4rPm7IDnY8TAG/XpClaGWwQybPfpxPp0ZPygNDuPDG6l1R+3nQk0w41GpgP15JtWdDkrnhBkdpqt/BRwLdrsyPpDe+z+u7091a5P81EGEVvfRamSNWdO4i5LevhBSEj9bi/IkoTtyr+vmUutEcZloXJGMQBUukYUA8OajOgvHvu1IX7WM/HmoDEtKQlqLDB4dd4RAjph78blhcnC/khpoulQk5o0Zon79mtEPYpYc32ixCA45n26UYw/2h2cFG584CaRTBraAe+O4mXok6WwjKO5eir35dJoJVnGkak2MTCa0z38X5K8IhRqOwzeLiWhGi7DDw3nedQ/qiS6blyyEvwIpKWiPRgabL2TrYZNRHnEu7+XEE0LA8yWr34P/YuMMvtsSszhh12HCsa537moaq2cmf5YNeDhA7bKSPvhBgwFN+/4xIbMIfqQOltTR09dRCmA8EftcL7bECQwSX1XApLLMDUAG747+/ipvPaf17UMyrdjVBgg5xyRLFn6GW4E1u3HTwj3G4hJNeIJ5ldNM5xqjwVYfcl0zk7dRHDcJWPvZDMKagnV2DVi3DzceDAgBgCa+WYBTfFjbhEIAtW0Nj7D/GjU9yNtvHVXqre/HG1H1ZXSKLHIBHjF5skaTSQJ/W5dMlc5mGsWCx/C/TChc6eMlNoE6uU2Ma6jtqe+Das7jxPTAThTIyMmVEW9vwDtnh2h71qwt+lSbsRliI9EgZBn0uTr1UChdqRGJubRUvL8K2uWF/lh7Y+qir1724quxR6pqxPlMJxkhNXeVXfHUy58NcQ2HaRNUs9fW1RRIxpZic2/Fy2vQV6snoz3IHb+2Igfp2YnKryBCF+li79CvkRL6STTB07wj1Y1+cqETublOUkIvKnfliHZxNLkfrdUJ1YKPuFFjsuLY+rZJ+c2kDPbVM9BXN7AbhNHPFFko0zSmTE1bBlI2peyfINySSnH6YcEwNwN01Z+CHX2wiRQeqQExoHRpJSCHQxKNR8Jy3os+9PMt4qr1C3Oeib2lmQHYuZFSjMHn6rRdoHmMDImDOXv4c8MIv/hz8V/gWvFtUaE1rOjJXZrGblzgH6gWE9Wt2HaFdFqhlzafY2E4Wa9HU7W2fDDg9k2ES/5UnzIef8Tw5i8fjNX0llmTV23STU62LswzAAvrAWkRCGjYXdNym52ghMWp+iWpfPexwzFFKZuPSPhbtruBvjS2rR/dwQURPrexFMMA+BmibZMZlmBJaiJH15EAL6kot7RreincryLrh3qvmT3FXlT7r4ynhJwMv74rdJwXaEkbrRV1BVRFN/Y1dxiKirDkkchxJTzD09CMSkFkjz6A00S1+u+E4PCqpKV9ClefMzTfSmvDBaE7TJYDFT8DJsNjNuPFVo4qSbVSg2SniElJkSWS+rVw7IOjV+iIRotwdhXof6YBALJM9NzJ/bnxhoCoFStz8eTN1JErNBSd5NspSbPuNxs6pOSBmFj3ngmxfJql07cXZ9pQU4WyFPVVxwRlo/2zdGTWLyCyMRrMXNiAwobyEuDj5huJvwCr1IDsLcDEKNFT8qjiZ9XR1in/UQtL4dEnBuLz0eX49HHpNNPkN6BKWockiEkvuPR9d4xP5QqJucbQUqLbHJ+f9qnTuGh9FCMhRsloTDWMlh849m+F53bmE98R0UOoRTRazAmpfmUtwBGMsmvKpJC16vAx9mfp4obl8goqR2f/W8e+XPfewA/XgjA4GwoRUR1tTgli/NYlekKGTQmeMGOPGZTZMTE/LQsEdW13AF8ZSsiByjrOy0lalJWoZo8N9frRPTZUF5CmuD9Q8UiNwzUHhmYfT1oKI/Tp53rEjgIKS+7NN2fzSukl4J/StCzuIcAGybYM4pBqchTEhVUqXzBIESCUDW8M69YgqpP/x3IhwIebjcBUz74rq/dqBbUQgksi87nIDQA+B4KoQFPLFVO3/n3h/J8fZd1SDV01+VZZOpXCx/nTVYUcQvRdgsBGvX/HorHpPZpZsXx7qJcvvBmBFI/oHcUjls/2vbZfmxvScy6HUSS7XQ6O+0SoIFvjdMpEvejF1g55M5JmASThQOMwzqqBnij5i31s+Z09x6S4weU3/sL37cNr4oxQG88sLZuCvpQU6kNmB+J5VBRENU7KWoMl/yWkIM6TIdAjabn8usguAOwdxGDZ3Mn81ifJOTXxF/24s6ZtRGJP+C0Wbdqn4uK8BFj1VtMDxIxs3sLqxaF/qUQbXU6aoIT4vVfrLZ077pIAwbf8d7av4mN1gGFOEJuDqb5cK7E0s8PhIoEjFanO7tb6xZA4G7A/mirXLsloqtZbZM7LZuEDxZA/ZRdihdTs4i/sFYs8MkzTPMr1Bgsa4LaPC1wsKvu53djXRepTplxKT29gZmYaqvUiDW7q/oP5qq+DGG4acFtNY/5WkMwVAHTVr+m9RQDsDzhn8/9DgqNB8DQflk7r5gXxqrBC+LwGG9GUsNSdyDXTPKkHe7A0eRt9fS1lgO2EI1n8UAJnb8MDwr6yAf+KN4Pz1uiDSpVcEj5j7K9YZ96Uc9iXHFHRWOYaYPDOa8ldb6S37h3HN2N9nxqFr7DWM42Gw6If93tJz4HNNiauLX+JnegNgB137heqUWmRfNIb6pRmJB8KD/S92fClXEEahm1G9f2g0mnpeyDJl5LCdtEgRyEveC8IyB/gkuSOVDcoqbXbOOzs7sOQvSPku+3UwQb8IC4eo4HML8bede2Ue3x3DhagvJ4SJSyi4XlnoLVykmZ1gb6X6dovOaXQrXlFdrmSfsI4Wp07al8KE4z2AXjK/B1tuWubGjq6e1UgL389813OFa6tGHrbkgULhuOWmCDvTNuRdGzAMnk+Iu+hQzNdiXyjK4XJSXmiHqXDRksCRe0YiAT0rBaG/VUqtmAHF1RHXv7+73DP7eKffb3vTkbZ+RC8gQm2T7rQTdS11ck8tmanLYFWJuUOHVuZSLdZZ7TxO2hDo5G1qmmdVUKUmWxG5pAp6g1nqDocvCcOcje016GdO8bG9o7OHRAWOpWyOWffM3rw1WjpNHx5dmhfiV8wVKspWA87HgN08pAyajIjLyZeyD3HM3gUHA1eZUtfB+dGJzJ/AukgGi8k36itj9O9BMUx6fZXSTnsF4NGkxSnrZWPX+yIV7TTtWxd+CRR4RBAdJwOq4OWIShu2xobuJf0TEDGIO6rM10+0RCjwoTMtHN8Vq5TfqxqMxjNQMzXMqNBDZno3fy24AsL8jvvB4K8ik49DnXWbhQgb8b/ZPBxP7q1l+KcPOPxYSkHtUgB/1fIDflGDUkIMU8EUBiPdvuNGCFw3tZL7mwGjiIJAZ8CTKKnrrEucuhIVLFZPXOTPI3nzRTooEvD1ZQLC9IOVUadtYULYe4PG5iYUvMp3u7jRki26IeDMj9XiK7PaIIcuK0gjN4l5kFkk7KyhtMMOZWZGZOxaaNTmGQp7/ZuHCyE52yMXxvtGBH+hNGysa68+apmUN0PGEc6j0ouC/CaMZ+uveT3ZW3e7c/czOPzQC8+k/g5bI70XltfRZOO3vC3pQr2gYtoHkK4qcIrVSblChP31o2ytfqwqXDEI39o8q3UcU74d5VXflesaRe4es5X8bwaqjP8+iLApKaq9sVeI5mUfRz+CeG6MOJbA20KZnpiWFkw+Wy0WwwbC4JeZAC403ORm0dK4KkqgOUvydQGoqv5bnJq7D6vbadGuKGWZIC7MB7UVyIgXgBz1Hxed4RHBNpvzuIhZpPqWAWRPu7DWIkMdm/kzSMkXdPZGeESr5sMydwXbxowBIUcD7O4Z/wHGGXrsVOhG+OvrUhD2Xz6la+6dsjtrwBWRVDjIvIVnhcTKExourbwziznU9cs8lE+EalTQZ7cM/rryTBofSpcYqImDPAe2A8phywCaiAZrH/k+B8ljSf+PRPq7U4XuyGFTfd2Qbvm8fA2SRvpuvbNZqeuYcgxSb2xahr+crppNYfP0Q1y6jE80QGoy2OxvEVZ8s0HFJn9C0P72jFsdRM34JT2ocDNh4/JKwjdlPXTZ0jAE7cxCsEU41ije9UyMOs88J5TPPZbNWKARV/PlNHsRkFj8BHJyqbJS1zZEFvnR9YrVudSTcsN1dmoLX67VSEixxA9RTZL6Q+LaINbEjKFU81f2huFTUbEZfzr/r+YDVI6MywXYuA0P/eHuLoOasMTpHovqLUNbcvkkNBj2CvgplbmRzdHJlYW0KZW5kb2JqCjE1IDAgb2JqCjw8IC9CYXNlRm9udCAvQ01TUzEyIC9GaXJzdENoYXIgMCAvRm9udERlc2NyaXB0b3IgMTYgMCBSIC9MYXN0Q2hhciAxMjcKL1N1YnR5cGUgL1R5cGUxIC9UeXBlIC9Gb250IC9XaWR0aHMgMTQgMCBSID4+CmVuZG9iagoxOCAwIG9iagpbIDUwOCA3ODMgNzMxIDU3MiA2MjYgNjYzIDY3OSA3MzEgNjc5IDczMSA2NzkgNTQ4IDUwMyA1MDMgNzY0IDc2NCAyMjIgMjQ4CjQ3MCA0NzAgNDcwIDQ3MCA0NzAgNjk5IDQxNyA0NTAgNjc5IDczMSA0NzAgODA4IDkxMyA3MzEgMjIyIDI5OSA0NzAgNzgzIDQ3MAo3ODMgNzEyIDI2MSAzNjUgMzY1IDQ3MCA3MzEgMjYxIDMxMyAyNjEgNDcwIDQ3MCA0NzAgNDcwIDQ3MCA0NzAgNDcwIDQ3MCA0NzAKNDcwIDQ3MCAyNjEgMjYxIDI5OSA3MzEgNDQ0IDQ0NCA2MjYgNjI0IDYyNSA2MDAgNjc3IDU2MSA1MzQgNjI2IDY2MyAyNTggNDQyCjY1MCA1MDggODE5IDY2MyA2OTIgNTk5IDY5MiA2MDYgNTIyIDY0MCA2NDMgNjI0IDg4NSA2MjQgNjI0IDU3NCAyNzIgNDcwIDI3Mgo0NzAgMjYxIDI2MSA0NTAgNDgzIDQxNyA0ODMgNDE3IDI4NyA0NzAgNDgzIDIyMiAyNDggNDU3IDIyMiA3NDUgNDgzIDQ3MCA0ODMKNDgzIDMyMCAzNjAgMzM5IDQ4MyA0MzEgNjQwIDQzMSA0MzEgNDA4IDQ3MCA5NDAgNDcwIDQ3MCA0NzAgXQplbmRvYmoKMjAgMCBvYmoKPDwgL0FzY2VudCA3NTggL0NhcEhlaWdodCAxMDAwIC9EZXNjZW50IC0yNTAgL0ZsYWdzIDY4Ci9Gb250QkJveCBbIC05NCAtMjUwIDEwMTkgNzU4IF0gL0ZvbnRGYW1pbHkgKENvbXB1dGVyIE1vZGVybikKL0ZvbnRGaWxlIDIxIDAgUiAvRm9udE5hbWUgL0NNU1NJMTcgL0l0YWxpY0FuZ2xlIC0xMiAvU3RlbVYgNTAKL1R5cGUgL0ZvbnREZXNjcmlwdG9yIC9YSGVpZ2h0IDUwMCA+PgplbmRvYmoKMjEgMCBvYmoKPDwgL0ZpbHRlciAvRmxhdGVEZWNvZGUgL0xlbmd0aCAyMjIzNyAvTGVuZ3RoMSA0MzA1IC9MZW5ndGgyIDIwMTA3Ci9MZW5ndGgzIDAgPj4Kc3RyZWFtCnicjLgHNBxa9z6sEwQhoiQY0XvvovfeexsMRhu9E9EF0Wu06L236L1Ei06iRu+iE+E/7nvfS+77W+v7lrWYZ599znn2OXs/Zw8KUiVVBmFTiDFIAmLrxMDCyMwLEJVXVZVm4QIwM7MxMjOzolJQqIGdrEH/DKBSaIAcHMEQW95HLqIOIKAT1CYGdIJ6ykNsATLO1gAWNgALJy8LFy8zM4CVmZnnv44QB16AGNAFbAqQZwTIQGxBjqgUohA7dwewuYUTdKP/fgRQm9AAWHh4uOj/mg4QtgE5gE2AtgB5oJMFyAa6ownQGqAKMQGDnNz/WAJAzW/h5GTHy8Tk6urKCLRxZIQ4mAvQ0ANcwU4WABWQI8jBBWQKuI8aoAC0Af03OEZUCoCaBdjxPyOqEDMnV6ADCAA1WINNQLaO0DnOtqYgBwB0f4CqtBxA0Q5k+x9nuf840AP+Ph4ACyPLP8v9Pft+IbDtX5OBJiYQGzugrTvY1hxgBrYGARQl5Bid3JzoAUBb03tHoLUjBDof6AIEWwONoQ5/cQcCJISVAUBojH8H6GjiALZzcmR0BFvfB8l0vwz0oMVtTUUhNjYgWydHVNR7gmJgB5AJ9Ojdmf57w1a2EFdbz3+gGdjW1Ow+ElNnOyZ1W7C9M0ha7G8nqAn1wWYOcgJwMDMzc7OwAkD2AJCbiQXT/RZq7nagvwZZ7s3QMLw97SB2ADNoJCBvsBkI+gfV0xHoAgI4OTiDvD0fD/yJUFlYAKZgEyeAMcgcbIv6sDrUDDL7D4YmgQPYDaDLDM1BFgDz/c8/n/ShaWYKsbV2f3D/65r/CfYfs4gIxA3gycDDDmBg5WAGsDCz8AC4OLgB3n+u8E/s/437L6sSEPw3L+aHFaVtzSAAnv/Qh57b3yEAmFz+zgvqv8uGBvDHFgAmBQg0oUEA6of812PmYDaB/mL5/10Ff035v3L/fpX/7/T/H04SztbWf3lQ/+3yvx5AG7C1+98+0JR2doKWhzwEWiS2/+OrCfpPWcuDTMHONv8zLO0EhBaKsK05NNkZ/j5jABPYUQLsBjJVAjuZWPwnl/4eUL+vRGuwLUgJ4gi+Vx/oJGbmfw9C68/ECioxjtCr+2sMBC2vPy9X3NYEYnpfh6wcnACggwPQHRV6oSxQyAHwhP6FFgbI7a80BzAx2kKcoJMA0Di9AWYQB9T7K2YGMEkCbWyA99a/DCwAJjGQtdODgRXApGYBemRgAzDJAW2MTR8s7AAmLfA/iAPApPSAOAFMqmDzRxtwQSO0g5Y8NOj/mrihMywepvBAoeMDZIFyVLQBmT/iCCVpZvYAoRTNHvlDCZpZP0D2e+dHwxz3+NE4lKEpBPpOOD7ek+sfo+WDEcrT3OFeBP6xQKkCTaB583BaULImQIdHwbFCyRo7gB7NYoXytQGa/OEE5exwf4//GKCsTaC5Zm396CagzM1BDjZQeTK2dnwwQwMAPlodSh3yCEJJQxytgY4WDyYoa2Hxh/u8P99HEMpX8c8JbFDCjs52dg73ufiPEUoZmljWQJsHE5S0vTPECQTl9x8Z+GcESt3W2cb4XkjMH8Jm++vsoTE6PJig9O1ADtBH59FsaAxAG6jV8f6B+cfK8/duf27FDg3HDvr02VqDzB5ZWf62/ssZGhnQ8X4JR6sHIzQyO2vnh1DZ7y8D8rhK2KHxWLjbWYAeYmHn/Is4GPLAkB0ay58nyQ6NxAPkAHkwQIOAthEPxQNl7+T6MA4VUCYnCwfQI4/7bIc4P5wYx32+gx/lFweUriP0ffkHQ8k6QhPwgStULJhAfxwEB5SqLfgxEe77mB9XKQfP/TI24D+tnMz/TQNT6HP7YIbyBtk7Ax/qjJP1/rpAjvdy96cr28PAgxEaA/CBHic0BOEHBKUv8oCg1EUfEJS32AOCcn5IbS4oV4kHBKUo+YCg9KQeEJST9AOCkpF5QFAusg8IykXuAUG5yD8gKBeFBwTlovigeVAuSg8IykX5AUG5qDwgKBfVBwTlovaAoFzUHxCUi8YDgnLRfEBQLloPCMpF+0FwoVx0HtBfggU0sQI5/VFCPKwP1f3nANs/E/6sLp77ugE7mDjbmFmDHvKRh+MvcYX2kY+LnIfz7+X/XBsaxEPZ8UCDMH5A0CBMHj0T0ChMH8H7DHwE78vmEYRyNn8EoUwtHkEow8cPEJTao2eAGcrJ6hGEknr0nED7HKYHSWS5f71sH0EoK8gjCGVl9wjel8IjCGXl8Aje1/EjCGXl9AhCWTk/glBWLo8glJXroxcTysrtEYSycn8Eoaw8HsF7pbc1faxjLPfvE8jmX7Z7UXS2NQc6ONtYA50fcbt/opzA1qaP7uP+mTIFg6BPCvhRTPev1f+8NCz3GuNoBzR5NP1eX/7VuNwLzJ+tC8u9tPzZvLDcC8u/2heWe3V51MCw3OuL0mPM9e8mhuVeZv7dxrDcq83jRoblXnD+aGXuNcfRzOlfTwfLvfjYGv8rxHsN+lffw8X+Z+NzL0WPW5t7MXrc2tzL0R+9z70i/dH83IvS/3Y/9+L0f7Q/LP/T/9wr1Z8NEMu9XP3ZAbHca9afLRDLvXD9qwdiuZevP5oglnsJ+3cXxHKvZP9HG8Ryr2nAx70Z8x+NEMu9rv2rE2K517RHrRDLvZQpPsbs/9MMsdzL1//m6L18/ZHPfzTqJs4O0N7D6a9vVtBS+i/+69szCOQGMkGdm4aY8AVYVgU0X1YIv3RlWB/hRzxMutRiZRjJMUBx6hEfN1yJUU1L/SZbJDHXzSJhYNmmIGJ/mb44c+K5Vk1S485+zkAisWVOYhw1fXcOOxnrefHq9TRmA8wnzQQRYt58hx4YJaIwrPYngqbmPaoUmF5UVV2zsa6UzwskuGQCmxVa2ysKUuVevOLR2Puh4vTZuF9tneMUnyrjh02k/2q4XmiiQsprWf7WK9QY13Ks5b4v/fPPuok6wt7JuPitH5EHOT0T5+t/JcaOXC5c/t6fMDAfZrTXgIag1Seh2TRPemcBp0DndemiDG1B+ceW+s+I8gEy0ggUo/0KPN88p5EoOL+kKlt6zp6ZSvf3ydBMfEAWCERC5ziIYlrmsOEtniOl5w1TByEMvEeKbahBRzCjcMVym54aEZJZU21+uYnEKBcwaK5Qk2blj4u1cAApRAjs8wMMJ/Pjrp4a0I8A4zuEZ7fuao49qRisvN0+H65T7VenOzsG0AvoBmuq8o8iCOYLVmKhyqx4ldwomSKnKKFkHYW3f2dUEbZrZlKrTuXAVVOaYM/jqZyK1bj++Eq69LXOhyhLld81mYerH2naGztUgKg2maxGHuJOL+Mlmk2JBhRWdn3uLL76l7Lt/25rSlKNNjMTLxQb/L1KKxIe/WY55OXWegUBhafLYW+3jvwMZ/lgxWlEVGOciWMioRyPP3Cev2fX+dhRwnlym5BQkHaXvkCGabxuj/h52yqa2YKk5XEiYhwd9kcxa/v2dp9+hKGy0vC+ubF6wp+eYzTRhI4yW3oM355aKNvH37ahmpphHiwxV4ME26S5ZWM0HEqq2rhXkmhxskr0qzGfXOLGMa+5DFd6jDdM7tOo+g5jKdCnMeAqaUqUBmr/XvrhY/8OnqzSg3URE4kF3dNvre78iHjt+FDiy3FTDtyHxq5kJh0cIsmlwlE+l34tsjKRGW9uJbXZtewfJWNvat69klloReAmjXzBbklLdJCngEdifqrqkmkGs7IkJXhAlOtosvSKvAa1o4waIdq6EsHMoXhurmVtPuTUQQ5Zm8S6RF5OVaISaXP4W93YiVYc13vuing2HsOJKZqCSJpTnpKZ8GosOgZ7AZ/QNKaDauXqo85JId2gpLumS/+I/FjJxXT5wiETYmkx1aO3DfZsJ8CPYUMBn/AgxvkdlJWs2GrkXNO+JAn6wV6armEtnmJj3ppPlcrodmkx6QXOovuQXFQ9txj0v1SmnJEzGLLnff36tadyJFHozZMV8rLb1DfxJYo9nzhLJVvxevqqRGyHEcskgOV53Z68VPg6W+d0IK5v/qhK+lF1r6gE8ZVApu9rETZHfcyDo+wMwZ5Dac9NTQd4uaXWo2h98vSIV2f5whBuDnvlscR67frFZysKnsfa/lobS/VnQBr65U2bZIjGHJb6O0uvKo8X1sqybEGnK3HYHfR636Vl9TICkBcbsz7Sm5qoHZbniCJj25fXN4d2rRQUcuae6W8MCJ4PJE9nCO8jytCx63DgIqZNqzLv7re4Q8354Je5sUu7yo/woXq11O9F7SW+h2EVinf7PChNu7Xm7pPkcNI1XCSPtHBaNSNuPeqAgNghV6kEtZm5Pl6N3xkBa8vyWywji67YFuMe8A3kPGRQfCsy8jIpYYqsoUYWe8iWMGQt4qQJvg7ilFr+aVJxHvILfcM3EGXsOUqABKB74+qYKu7AoCIrsTxJKLQEt7Vfsv/rRcybVcff4cE074FPN3zFdaXaRvic6Knwugp82w6lXgk3TBiz241dFwB4hcntsJNZJNNGT+clx8KPO8/foXdArtubPiD7gXlevZZQp6tIxLCyJ3jhEQIZqk26miyokMHOD87eXuYVDWryzowtmlY2/iFpJQE3TytUkuKN15/mFoPGJhaqJW/Vf86aVGH4fAchQzxAtBjEKH7hx/SSb4V9s8JMMPStJCFMm9BnopGWfe8jJzL7JutRegckdtV69KliUxLE6Q7khqssvgMNw0YJOvrLWzp0ZTRFBLpnqyhdbzS+1K5FR58Yu3NrJL30uA5702Em0FDAAfyZ+qtB0251E1jxa+l9crUXhggTrwOIdB57+ocirdLeu1KVokLHwZ5GARh6VPjPW8/ozHuGruR/fH95tU82KaSXbPARuekZBOI/Wpwas6AP7FDHnYLT7ZzUSYOYhzNI5SfmcnJTXVhPY8IxI3d5E2RQimf/vqSY7bqokhOZoORgWSdv7vkSHVTlHHqC2XOFn9vjR+WhsMPtTRoX932fcqsCd3KLxQcpSiaEOAZvjfi09pf95sVcgOcGToiMa0uvNSn/4XT1TVmFKf5dtVBVu9zKp/i+jUyGYMnowWHGbZfd2QrGpujYNdbaEeatKosDCZ0M6URikivvyoxSh95JfNiVjeZIOSMUt3xzqNAsGfAHfcio6Cmjr6vStJhKGiLOFoURiwt87x+A8+lWqgNOw6FGwvTOlCN75jv2wnC/smNw1NqTlKkxknyz97RG5Td29OKoT5hHkGS0dZZJUcTjQ2cyjAUwrF9693o028+qbU+NHuEZMD15Ts+bw8DIUicNnG5+UZIgovQTfFX7pj5xsycrwEyb7S79rSIuLprSbteB3jLl4RBqkZP6NXs+xlVa6/eMyW24HOP+N4EUYSg4aBo+2U+aFzJp240HM7LTI45FF2mr33fs2PfMSn0KgvsJLLkWdIGJR2EqPLFOeJ0wR0MbvVwXnGeeYD40Kpmj3VqkMe8XUT6awfCe/KS1pF/lImti5aNquNszK5jnONe32f7JCyUSVFvbTC9+19lpccdxdoAmJZkuZwlUDfHKUE/SKmYG88Jg8FFa1HI+CvCnnqFA0IR+XFGPFiql6szvyYS6UD57S2DSPTPEbDUdp/f0bnjLbS0iA0an8ds5kZCRKwMTTTtXVIRip4TrXPfbMATjuPXdjM+Cz4ySC6iRyUCyM7DhlD93qXPmdCteP2uxKIHFFH926N5RSHL6+cQqH42CtCPSW9bXJXSHZ+tarjNk4kaGlst21Xr1O0sX9W9Hktx0vkbDkbAf3zv1+DnSyE9MpSKftExviXAyKeTvu8QOSMmQvN9Kr+kThUFOhXV1eIfTx02p85o+wFSvwdHGoDxb6nv5zTDhuevICz/Znxiy9shPEfRbIsP1OyOZTBulp0sOINyf4L4yLL3rz9wok/UkDH3b63eTG9CWkz9D8EviWyimgWNApiWBnlQDXloRfrJCRRYdf+MBsvqBWGXwTHaEbUXehEHTMje/QNX8h7FbaoJFzzC8FcLLGPRxEYyeo17UnCtR8duSpBb9z64yzKqk13DYIgEtEnGeYs2fHYAUx+ttPwLpUfWqMRn9ttz9JygGbkMM4AZONGsycpC/6FO1it41klBItlxhWsZ8p3cp6fWNs05uj+/Wv9wQbFUVcAWbTxuM8MSmzMH0EVO5yCVLvphznjJMPPzWXO5EL76SnqNzxR2dlIdCWlQ5kum1/IQUUe/mFmf7aaVbcxhRH+o3u5MowTjS/XHQWUthAU13HgX+rqR7/C1mhiZaoKjv2AvWFZw5EfOkkPXRrkxWcp+iwR+IF0HjJfBG16tdJFX2dJS8FORxVC/aFJDlSZixcUiC4tpADBArGKvqDjIYFPCx8vCeYsjdSe1W8kCAcq8rGqor2ZK47o1VyoeDNNBtk+rcrDVScfQJ/OtwyclxUlparwlw6USNzWwROPbZ4s0SnYw/mgxrc02JPqlQzTjzuwP0HdH1XyXAsk4Sms8dhrZ+hvqvwghMRbMLniu43+Jdvg/AZInmr92oVfiSmLF/RpJFmweLH6o90J+291keP52t3gCgexvxy08z4czfmZi53SMmVFhpNynuhV2XVfw8o6BEdEAkAtdr2DdwfhUKxavLNr75VIxqFMVdG4lod0aaikpqG6KDzvpJJEhznHKZ6YdrKmpp+vrTE/tu5aafp+1vpBVh3L8PTjlIwOgEiXibG+yUOuEUVhWH0+fKIuo1NGxsalDyNgWbvUwfWlkK0/old8UalvdD0CdIK5l17ejJG80KyJThuu+udnQONbL7ob1KAGxmjvi+9YIXRhUTCYGJ6v5R3gi7LsdztyRpTAeM8lVKbn3fcLYXk5Xvhg7RZ25LA38npFV9rruDLQYNH80yWzSgkiHbSCkOvpYMV5nHZ2VNjdq/1vyasugbuodC4TATZYsbSHu5C6qajGdzMmToSk2swsU5SeSYuvTkoSPbw6wzQZATnjB1SpfGp+yK2e03kA2XddBLVfPKsk3QBU80+998uqOU4Y9uJk2CqDQZSZ8zRWHWP9GVPF/Tzb5Z92NIOf3C9mRA51SNOvlXEAdnwV5OrI7sJKkM9cRyjHoJ/zObw6P467jZAe58f/xEzWnmqePtBE3D96wHBAhFs0kppfBikRjB5JefXT0QXsBTjUapR1DFyeVwarBZcKpcqR4UTWmdQ4rLGoVQ1KTlce94s92UaBQnMZgbWcWkujhr2HT9h9ybMHsbGH2vjkYihtpkI56zLtMctTdvFUkXv/p1+oxD9B1br2hPezProFSavGVxmyYof/yFcr6tNkIG0ignFkpQ/MlrbGotnNb8iMzzfeV5942j7X2aEFqSBBiO57rVk5RE0+b1bDkhp99UGytDX3xGptseLsT1PUdcumDto5OZwjW2KC9XGnxXzFLvXBPLZ/tJZywr8Vzq64mAPVuw1bShbgR3fqFWM9vXUHO7IDDV98FW3bdfYIURadfg1UQsXb8DXYyQ5K6rsO7Yqwc6+NUtMV5gFafwxUh3Vi2V9OIavx5lIQ6nUSpenqMnDYK/Sw3xT7pBqWuRHozNspdi2f7SeEzLRrunHcWV2LgZYombs1H8Yvbww1LsKQRx661IBkJyYQMk0UywAVNVnWiOXgO5Oty3yexU7yppSCfHTrwRrRyevH6KxJizz8PCwToWvxn2O506+vqe1PONLKw3tb+74Zt4XIft+zjDkFLYWpPVFV/yKUyynCWKxI7uOOmyx8w4RLYBuJy9Ln+NiT6zOu4O532f8HqojHjHFtcv8XteeSKcVWacoKNBUfz4sIZnfcVAges3V3fDpR/XSaxPaG1FzMm5M1i4KZUvx3RfqhU5uE7/kOR/sxdrfw059iLtzB20sexQzKAIfBKAPukTEeryAd6XLdEdpkplhsFqDgM162JrqrKlBtFE1voAtyVSvBFW9JuXueebbE6hnUk9Nvd5SS8F1UVt4uO3oED58z2NrkpvzsJAFIdZbi44m93t29XS0DMkUf2+8nw6fpHqp9MU/tu3XjejnUk/OeXwbVG0sfkvRMnsJRwp2tEZpI0O++hfw+peXlc21Q3/jEQKXaQod052aDluUnaMReX7pVOD7NJSV9aFGuZFULJmTuckMhwzhzClSU0fKrC+j7Dxsw9MSz1ozPT1qByjyXWg0Jou5kZZiCDNaXBHpfPoglFdpGnJfOMH69jptnDKqM8QOZbpCdyrCfjsGhTzufzKY3eQ53ngs5FfT47kIE1wm0pj+DMVX36l3onNcLELJdaM+CsTfTp5mitublibg5hoJ1G/k8xdRGV4S/X9Y9An9vDdGtNYYmLil9Ie5JrppB6VsgoqzRZboxk9sbZkoku34yqnxcEX4V8768coMrvi07Qjry5Qw3fR5Ap/1Z0gfNjwN03PCJynpdTw8ghMDqQ0HoeQ0AlhxBAI4Mwb+MlqL8+Ag1qM9S4QIVT+/oovmyPYlwfzD0gJ2Sl51AzJombDh8WRWzZhJ7IPEjo+cMAFcF1UpOE7SOvPZ7cFicE2dP5sGh+o2VYeX6Zpl33zizJgnUfZTxrGy/16yP2iWmIRtQJmVmsrAWzVUGilzz+rVUjM0XmWT/zThd08w+03Rki98aiH+fPamAFBgj7hPHNeApbinjcsv/uSxV482RMnWtgvXFJRA1yrvQFUSgdTOomz6vJc9ETs+EpJWyEYEfZ7INlXCr+1Juz0wqtEKtk2VTty7rBo33sWrVzjWT1U8q0lfBVO6am7x+yi0Uk9p7qEp42KU1AD49EaiY226eK+4bDhkbBLkoW7UwYG2Wu6uBwRDj8dB5K1Ma64tGAlz7ljFHeu0BB7c7QfKc83RUT9B9GmVJKOxROZBtp+4s+U12NJGDdK/bx2oFcL7xIWC7l7NX5E7janHtwAenUn6mXRhxzvmLhzYEVD9qzJj10p8JSWx7/Gm8JA073rubZNsS+p0zusG5F3WcfuISrbIpKYO7QVDruBwyk140lFMCsdxHJwpkQeLrGfB+xXtGO5+/jJVLuwTD7sU01/PHELQHuiUt0eOhVpzZRW5luPo/JzRIVotVlNajgjNHhAQ8eq1DVtwEsTcWFwZgHNXtAmluX9eOWzHNcKRcM4VZY2fe8TvFj7NZmpY5avOdyIyqxfep7r/PgWEsLMpDdBwkrH8N3yyTERJ3tza8JLaZ5e7ySD1u9JTzOLDHoK90rHZDB/kWmGxfafSfgvk4MzrgrPzziTVMorUhD705BJIZBlZfkLW812W9sUzkLtKIoSZ7MwN/8xqoWScxytmzi3iskqdcMg4koNFdPiyDniwGvL1deYQDxeGY4QqhbRTIhw14pFLbw/Iu7c1x9vWNkdSnzzkVIGoklIUIywBJxX0z4m0raaTMR35jvhOQuWnwY1TIkZ1FyY+swg9bMnqH7qfUp2uarRfaWKlzWZejJldRqBKHzRZ3ytKOzxmUtklqKhc4EnOKn0wAqmkKRFFwezrS5gB8VBhc2iaGYKdg/5OpdaG7bQIp2NdVBnLt2e5q5bQMKHADQe2JT6sY/jpnPfZucksLekySbKAGuXP1ZD8ZWZFaRk0H/rlxle1K1vvqRquUDT6xOCJw4du/wH7A3Gi3PSm+gW5azfNflPlAbXsG7lGoFojFv2Y5kYdTQi/bUZEn3Sjootb94NgOg/kMNrz391bploBIxt7B1o/0TwwWTtSHn3i4pLWPfkdZHxFI+HGS9o/JIX1x8S3iP5evBcnTFgmZxGVZMsOQ3ZJUrrkNY9IvBirMWOLwDHU2OdSRCOHgkuIor3W/62IHv44Y/sGHQ+ioS1NPqL6CVT19rwlRw9nGpGH9Wg+ov3RPGSZ6nHZG+WJQXlg2sn0YhPMUKUu285nW2GPbxFRPF5Oiw0GxO7wgXg0ZXkNn98CoDhfjr4iQ+XsZG6ODHX9jxdeeiYF+OMXSyYLIHbLoVr+1nPxHNMiuVfg79xMo1WyvqS1H5LsqYWkuS4YuAjaelgVDRf4jWp47j1OF0tXcO83cDRsP9CsEaKaTFbNvvF6lRJRPLIWf829YJt/sOgh/DE0As1rAHf1CuZa3Z4uyM1U4vOW21wQ1PH4EZmOT1LAsFgHAwn4YUhsbbQ5+DSb2tPt8ap2zRPly07yp1yJvBujCLxXu0lOvF0S3Q5zdrBO+BIJ740inbbYSxQEHyrx34LzJZvU8P8yNUn+3IY0UWEUT1z1qv4+dR0m/4AjU8+/PHqh2zQj28p8OWzNImJWqOX2UaEGbLtxLX+KGHoCLWncK8TVqREfOmnc+lCTXvCKxo/F9B9Y2xvS8UWMMHnEsZQAOHx/CxmxChfeQfDd0V8bgFKIrLGSkuTucD8rf2WLii+P+712Igmw/EEONYQia6LAncFedWMgz6fxuJV8+kJFreIsrDdx6DjJ24ReZ0NSWnathClm3G5G5G31b3H+ckHuJ3oldqMJF1IoSmntiXvwAOq5MU+PxbV3jpHv9KNuuYkNkH00yfAhfckEHK13vhxgqYYrmBqGDV23YETCsuDg1KsuIGtXE8SYyUam9nBisLyjPCOcF3MXxV9yYqrTJqkrzpivXFYhJZqSAX2BhPnfG6SlA3MZ6A8pvNRC1VfNHXakdy8LXRrJ0LO+C6eWUFvpMmoGPNHlv6+CONThrOWleOv6Y6vrJnSGz4alTNZboHsk/cd7YfJwJPB6c9Y+9adsbbMcl+a7wh9SWILsV+FH9c8NCkqfbkP/ILB/gNTFo6tY7TP/PeIf3+wlEBWwGa4tU07gYCf+DGZo9AAoXa/RJYr2+gryurS2VZC93hqiYrSdBTalkL5uacK/ZxEb87CghmUEnjUnoTWmiwxH9mCF9VzZIw7Uw2r/RN4WQGMpDFzn43ksWvFMGiI8Zkk86sEu20FvqP81mtDzRBo78DH1T9ByKcOauu2STXbh21QZyfA49fM9hSVCrE8JZnhBj/5lADvLwl3xyHCvq38QndjEAf/vX6vRLHRV9/pFxkKF2l8eC/NJn9g7i3Bjb2qyYPTiZQ3yOafxI3Kygz2QSpX5hqyY9PS22IJjl/0NWIoNUfZLVchkCKdwvAONpHLflavjDS/znjKG7sBZkUKMacbGhYxOoI9RfD6aBFeUF7J3+oIhwJzaIIfts8WGSBYwF5McOvlMYUF8EhbO9deS87UPlGpCexs2sjpbllA+Zmf7Nr5LVxiqglEVJ+x3MASq1v36Upe4huscnQ/WXIXcudMZMozJ2u5D82oCpaidlpjONpx47nUJmA+YbQE+2xJNuXir8+Hudc0KdBLHSUt5paoTW9IjLtd0VT3YfDUMFEugqbUMRB//8ibNuCdVCXpb0b+8azWhsuIGE7gZCG42p2Xwth3W8wDB/zDZptL05+87NY5gMoAFccX/O7kNTusdpx034aV1Ngm8qtcjpSwiUmMlIT4uPcTJTRTwaAirgMbW5hnggpSbNpXdryghY5dDHqluxGh8jETBM+82n2yS6as4Kc/ZqpAffuqNZsdMZpEE0N938VnGD9nf3fk2GdsEx3KXpnRlbp2fpUBvvyKwrpRHCWKrnn8EdGEx8qaigTxWT0m2s1I/VgSoBMPPX4yfXPYzPQivrzkSx06bcYX4JSo3Uf0VHJwUOCrW2f/d7TeI4b5K1VKX3WG6nU1laLS1wtNKpbJnGKMObtIm4TiUb4cYaKxyC8faDOA9zjcacrP3rIzzocnTvVPIvBsXsBjGsfG6jj7AXVxwEeBp+s5HWDpsk7fzLbP4k9Snn2cAVcaMNO/t2DFCwtSts+nlRzfMue4mLmSM29vrNw5y+55FRYnynhVaG4HwxA1R76S86FM8Ae+Q3klJgFCMyopcYrICy7pYW3ecM2f5fSri/6nEMoq4zrKr8mJEb4NlR5ftEr8T5c6IyDrw/NrBBJR4s4Hl9XM+tO0Rrr1a1tz+DcnlzT+uly27FnUMjOsNldzHrcOBBNvDF+24W13ZVqoGLcVvXG9U7+MXxNakQRGhS7bwBc5nBF7iNzmCq8+982jeKsqB1O7JLf64dtUHa1AjXTZB57blvh51Tc19F5LJhyHvrsYEhUd32sVQSUO2RzxmvENtUcur4R/i50JlZUtbvLH2PJTHPu6yCf4Fq/0nVJiN/QEpikimofnWD4pL48n2njyclMrOvllgvTGIoUk+gZFQUUoekl984CeYzd64gjdj2IsEkrPGjwzoUSK9ZrK+ucRgymBeRpHkvLcqN/ZcBOlz5nRl0kz2Z1eaFioaQHrJ8OZdILo+NnrLlz2tvfWn+evmgTuP8/Ly+UtXgx6U6AEr4WkzSZ3ljFDVmgTKUDzezHeNGBMX53LQPBDsc7OTPGTV7pT9fJPcQ5lvVptFvt5SmlXqWoJaN8kfDIZ15ndSPBiXl8Ag3DP1SPKihxbznqs6eDPG+Uu5T7/MrCrk1vKa3t2S7WQ9PubXSWpUyugQ5tTHzWqW0Gwi399FL6DaCXsui83T63Y81CrIt2inl1eosDnQ2shZNKUIjyOpOrwhb1GB88lPoWhkk9s11wG83JvyfWnlfWPKWE9tm6Su8NMThqqlh9rks3VSM4qHYcYh6YyL8oMO5ikvPUkHTHlfuY5yBzDMW0qfZFeDLzhWusIZgklLvMPlh2ou/VeDPDc7b/Le3Ya4cj12puOZCApiN9SfnUS6HMRUVcQo7NCLT1W2DQ/YuzpVj9X1tN9N90OuS63xO0hZKq5snAUsPL9LH4UvS6iFV3zPMqkeH9mWt09WnQYx8F8MvVGjFIM4bDq48e7+WnY/vzcdktluAJdiMlt9wyN7WgzG+lzBxTPLyarRL+Vi9RlkYzCY+Pizz5tJd0OaRth4PCCzDqET+i/eo5/Ts1GwZKBa+K6WNI0+ELIJzG6H6x8csMSAvT86MXMbtz1OvX8yqXurZAXl6kusdCuogrr62pMC/3AucbzCokP38IIhHeuBp8EwGoKhiMjaB5+ab8qGyddGPHCTpPCbDRQEACSp/7Asia9c4mDKZhcHu0fI8ZJIa8Xv6mXszsyNb5d9x2DiRWl0x31jqVvAglFHGhnoPiFHfEYiTKm4+YtH5MxvR4C3bB9bgc/e52SubTjoIW4n9vTBX/po8cNO+yUuSDcpol79KTcqBPBp9YF93PH3ogLnk6c1fIVvy6Ary9pKS6miiy377U50qh+FcdWei36Pjc90RlwORaJUHjxhbioPqctz2D/yloJ7M+6gQBuaXRZRHElPvqQFKpv/OqtCALkY95roQ4BVXdXDTv95lxkP9lWdJSUQVdR46n8/HNuehqZ3a3xks+AZJ9BiEMX6cX5GOWXRHoipNHuJs2bLJ7ebzHIuaukqBDj8itk7bDrXU0QfiFi0yiIoOXFZJq9SvVZX7TB5UK2hcWMLX7dxM5+0VA07ordO6nNMyCid/I82bKDJSEF8aX+B4w+wwXhH3BjJwZAogjXH9URa7s3LrkoWTnG77QxBc1p+vTyKQwIKS8dxUUshpvOI0B3KYxHCxOQ4Ql9VwApfxXXC1kuJ5EaNNga4QUIvWxJRuXgnO2Y311wJcG6JMJbUTTXviHfH8uLcTW/NdIdeGYb4RyvdF1HPvpbWvXu7UfvcJhzaGsiKGGkmSTxJkbMMep5imzXEjqBazaruXpq+DQO4+eSo9jXvJ3uams2NEI0u3OJjrIE9OgACSTvfx/VfT/qzGWzpXz/yecQ+quSowSt8dXxRDXYJZFVSm08ZqXlVDMnW1k+Mq97jFghWdshXa73ZCHR4vTwkATUKlqWgms1VhB1Bxulf7PXfHPc4UfY7uCFzGLeFdluM7KXH0epydcg5/8uCHlv6awf3773u2/Lr9esJz+9kekQa3KHM56+wUXGPoarBmlmzfsqHwi94hwjMGJ9rbE//Iw5KsCW6v0nEkExXXovNOFVQjogNYGNNmDmPdU2jF4anVU+30Fzx/dI4+8ilMFAlAJp/qNOYD9ql7VBEXPmtmy3bVDiOZhB3y6JVhTeW6UZdXRwXHqW/fVC7ZaCzuH7cpslxbsZSH9zDnsupYLp3SuZkp3K+IAsu6bXSj9sGg7TRIWOR9yNTSM+M7Xkch5aFl4co3f0UcbAd6vb6l6CIn41q4hdes76IZZWxO2WV5CoDHEn6OlqigTAPNXucSfaq9BdrvSfDlZ3DdmaVNpumyeatxieSRChLg1ax/+qWXrk+gxve4JTQamlB5JbXnphvNck4+hNJpWYh9ZMTYv3FIg6RBCftxn+EV42+mv0otWWXDAPDJfZL4I3dGQm9sSsfOfYaA3sTWYVKiQ8LWcVHOM+35cFNZEqy1k8hr44LJ8MFHliTWkPmOrn9+tmv5RFPnruTgNtG/Ut0rg756XjfGBUbWkSpRainKfNjxxn3tTjfVgTgy2zT6+GZxLG/ubNCqkvEsaS+WI+yhaGG8vJkm54Q9W7K7nXrrKbW9jsWqwg/w2Rm3BYT7PbKuL8Kmouyda5l4W5C6G3GFZPskg5i1a3urzy5OJ3YCKEOT5lOPkEJa/eOJwN9mTKOwS94asc2fc5Zxqe5y1W39+aHj997a9+UcY+F2uCK70MegMjKHPhQQ8er0+I5ES5UDSOzY1OBg5V+eJkG03bQb8XNtC4moBoMrhcbL4MqXWLtc51HXqIkKKOF9mqsbzybQzwPniKqFTPRqGz+fOIm7FkKw0+Ob+ij5Y3wYlRNrQULBozNxqPF+zl74XXZeUytHjLmTX5yhPfhuHb0YjRtTtGBLOJrHT15dLdaW0ubvnSZ3l1ZaEQCT6n4pJtTyu/4wmECezR1z/NILJaDUSq3TcqlaclVwvyKoxXcHei0ch7ngfExXAaXGBnwp+zyGUNC8MWcgrsA/lydNYYakG/XTMnf2II+l0SDjkxENP8YhQGj3MBAW4zhxciVId3DP5sEt8a1O5G0XGx7TjEN0Y3hbanpzWE1UeW6zNfKmNE53HGgrgL7VaIAmgns5PYkF4xlbcTND5BbgYe0zIJYBHwsVqOytaH+9I2J84MvRDrW3Pob5iX856gD7RqpQqroXn1zf4VUo0jD9lkGT4aaPImEmRPp5ZXt/Xdopm8b/Yj7j6jyRcn/Sc0+rusnIIjfb6FVcG5uUvXzay0tILGHqYIucYvhynoCzwkBz7tvmtYl35vO/kdFaaLtBsm5fdozlO9MLWYZqaB0nLvW5+Zkl/PyZVbmeQ3aj37HRk+rNeXGrdZtEVFR9Y9TbfaOtMbrHzjtVbhKugwOvsm6HtD14Rv0RqWoynayFE4Eh3pz6bOhZRm1XAncjW1gVebzB1SG4W7odPeM6E/6VvYQJx2JvDI0zbav0sL0D2c6ZOclBIPGs/AL6/OYye1O/TFSUiiM1GHyKdTk/sU6zIIM4WRkasMJq2fTXzHUofFDE4uqhdPfjruzUdQ5n8x9AmZSSPlpihPVUvImi7BInkaDskyRL/TqRpsDcxc0TlymfxazwFuy87w0h8uEh6tIyUeEQnjGZfEfBPG4fc0TsCT6jw78VWzMiKNMvKTjmDpwgjsTgrTubovEfCy1eAOWuygsvHBq7VkoUuEo6FXoUl8qbnegjXDRsGUKF4xi+dL5+pL+hf7Wz0JuNNCh+DPFj4inVzEzSjdu/KGFy5ph9jiZN/bSb5ItpMqGcvFW6fmLb4SI1AE/fJrca57qy2mbu44B2lpNed8n0NMJ+1QlR+86u9ZjW7GY2BmBGr7bhM+cuSyIIomzrS09XHERvd6DyBkxXdbXJ3GWhDQB+l4Azv6wvOpZxNtFRBeWcAz0JyNwWOWaLKaHH7JE6aFHj6sP75gLD1AZIHDsuW08qSjUiWNlXtE4siuLQVXZsAMDJA5ckYOBizbbgYm7podUBkazPhL7wngvOqz0JD7Hpk0FVBPiNHfTCqIvV3e0c5KgbF/2IRFZ6Hs+9r1ZoYCr9FurGrG/J0Dblru04jW/U6MBkKTV3Z9AYSDP83oBuudOrJ3+O1DTP3Y6YcFtBf8CFb9U+ccmq965cq+tAW9KI2HpXAoslranPywAeJLfCdNIXeWeQazVv3me+tduwkx6EJJNw3SzsMwHVCxGrBXRG4eeRkcpzspJ/+Znc/pXdAHAa2NmfNqsCPTz+aWVPQ+51LrU0cV86bU6RuJL9JXgcdLtOSLwouNyB0CSDduVFl0HjnFDCOUm0ftWNQFco0cWVbwnj/pq6uXaMv0RK32LCN/u1in+xvA8FTSw698w7bH+PCW1Af5A81KFSZ7PMU8Ly24J1d7zHfDn5g7rRGO5W4VY1hjU84S0KzWQ41kSndlNaEgGIMN0HKWEWlaHVBveL/h8rtrOGl46CXpxdcoMpRZXkPr4PSS84Pn3xGpGjGGlw5D5DCvDuGJhOzFAyqofXairsLm+59eFZN9hUk4qMrJphKtqkCt5HDOxg6Ndq+r8RS0fX9H5A92zXdLtXKJFVCVA2svftadJe2iV36lYOlqjwr/Uu8WbBC08i5s/nqcOFmH6t2n0Gs7vuR9Pun3P+ljZrgFkZHCvusjCM2gXo5hPtUHgjadeKotevCzm+spXPLbJZcdVb8hvgSXdrU2tAtqcBoE69UiZU8pLl5z9r5o6dW76uDbKlD4bewU0WDK4Bt21UHzcV4SndBY8BU/0msXkRqnkbeCKi1PHfVrWDZYnBcTEMc7siwgYK6+4M8jxmtNNQ6KRvlYr1MbmQlKJ9CUxAqaVgqO4AxRcS5Jo9bhsY4k8TM9vrTid3bYsV6ryIFrW+F5gdovCVqkiVklrKSDIVR2ZCZVJ2CrICQgxjHH2AIe8B25OkpbRER0/nR6K0+aZ+OFrKNpFSczz2J9EIcHPWCbWziBKBu517kzOOoDLgEnm5vgeHxT6NOaUj/8/fSPPErtFfRLZbGSaIcS0y+t2fZuGZfXMoMtPqsEfpPmjRg1SQMCont12VkHuo6PHXRohCiASOR7RSR+BFQknTkb3vYIMqIcPWHv+L2ltwMSreeOEsfn6qN5lZ5jazL0dRIZ6g6qTVY9RzrvhLiVrYx6+Wdqz2pvw5tAiHtqzYzXkDNFzQgiQI4mg9JU8RNKvg1tXi3VeGhC30xUx7d/ocrMTZs7CzKvs7XzbZP0Y/H2Lu3EtvBJbePsWPTftKfYSuztRs2IfquoicLMwoF0E71gdpDREffpcipJGj5dfkvzaxN9QWrAxf7FV9jfY5r23+p8xD9IfPIJ3wqe//Eh/7t0XWa7qRezpWgp1u5Ska7CFhjWYlC2cWtDLyqOxs+pyxDhAmmgYvVVrg7toKCabyZ/5K6z1Pjh4pF3J6NQdgYcw4tKyV8NOxBUrD1ZvtxKZqu890JaN9mp2O+5ii2bCa5nCThGPkx0431sPkZ0zl6Yqfntk3jKvKQQzW5yQjf0a/hXqe4cn2qU+URk1ZCBz7si2esfGpdqGtMJBwkY5wDi4oa317LeILb1KQl+OaGBpea9kVP88ML3b7P69HmolCf6EovTTuPo3sEOo7z3oOxc7lulwqgtUFtOnZMm5YQ/iF+HU+fmaIMTnkjEtBbrbZv7MLLaj9pIpfW0xur9MadR/6lCK8lBdhnq26ADdqSz0Gn2lnYst7Xk+W1S7g8Xxa4I4vkNfQM0uk+8X5SR4IdevLmbXIFIRT9lmF0ZnkPUQbSntxoyl5HLjp+mY/fWAgx+OBx+uWNoLW3lNnmnm7GbjMAd9zrS1uRXXDf/hamqt816hlyWie10awy8Gy4vJ7Dt6MzZ8EWncGYGsrs80is1n0g8U0tu99Qzq1y+L1aoZFED9hj8MRu0Iq5G/u5jVU8jB0NmkejSO4aQdjcCN4WFGU5J5D652cf6vcziHAo01YQz1pBD+HnRK2d4Ezk1zwwX+u1TebUVNo3Xxu4syVefuEbC6cb+X9Fp1bVrtBa9FeqwK50EbtvMGjx2dQ3evwZK4hWpgg9C3BiLWvY6CZaNMykHXbBaS2lUfEwJTZYrSo58qQeQBxQg5cVUm0qVE1kwz4p1SQgJ1kDaR0suTyWrXt7seCXBsjjuG8Q/7yt/stIvfPKzotFNZ369e0+zWXpAGUUoGQhamJPzqMg3MVNT0nLFElmXG7SHWJdQvB9HbxzV39h1S5kMY5oq6pBGDlbnONazx7WTjOnq/YQW5OveT2DUBadHvPxj25n4/aQcvLTqXdgkQX7qTEwAuPysEPxJeJZGJgL7uLICqQj1DY+66OZmxvXlsIfvCN/El45QHlofHSSekFdKG5Tb4ySBop4TNSwTSOKzJkWvDuB/3hjv0dlIunb4Mt2q3oiK4YaOLDQ44JvXSX8LhK1+drWUYWrI62uijJbYNa/ZygpIX3rr4KqiqsotyH4nIAB/lTVsZmIfmWxtA6CN+FjHn+/mTtlqLT6GXiGNPZkezHmFsi9mPGeHg6rLJ9E0habSnWxi4oaDnk39PtCfKM+0LNueI9sY1EBnFJOY8R0mx5tM32+lrx/MVSMO2xtKgsIs1MmNeY4O9wQY1rv6ZuUrbEiZeozzm7Sz/TeVTuqo6nWDMiwWa5+2GoJikX43Hm/AFrzB3BLVEA+JxxNO1ebmm92zxphAl5ejfFu0Udep7Cg1+u2nhGHMWfiJZKCpLSpc7iJLa6/fLy+s33KE1oADaXBNHQ3O5rhzkcJ6nXTEgvM5jKaQOQYFGeW2sbQkldJxH3U2yERGpcG7hIJwtXF/UAbvc+jxQncE0vWnr0Rov8JOddjBXPPWNGf9594XxtHHfZIfQJMu4g35+KA3J44abRI/cPQH2q8YQUaRFM3UKJnJF/audHczRkHO6k9Zl1gZCRdtnCLnxEX49LSFRUxOWV5Iue3LN/V4WUxPLFZ5Ro4qJDawPVXmuN09gPem/35opJEo4vGx6P1CjtYPKfaIg8WRj3tZwG2tgRyEiLuOKk0xjzxT8V3fIwT8/qpZEaIARLmrKnV73ikfFqQRg/NmLMiWvR2Z4nPlTIfK28Lrgv3DJUadFEaEwa59xuH8mA8cMGEBGfgL3b3xNImsjHw6RWLZZg2ptWYU82jPa2abjchL6a6/oiUqmQx2PothYLDfg0/eDoAcYjbVvX/H4/Rp45V1qsetCzx5wVw6wk8is8YYg4GtjvQejbCvdU1uq0tmMtJf1ckp5yrRvTqejB/RZQlVRfwSdKYf1t3dXL95YTukR/FEka177duuCoA05/L9TAep8M/19NKZfjaRU5LOD69HJlCHeNIVCgWmEy2DA9RM+fZ0opzqbfth4mS4xYq/sq+Ei/M/rYA0NCYo5N8+tYy5ZKnoHTfAHpr+yMMT8lxp8VQ//gu3W/n+Trt7SgN+z1xTkb7j+7fYclwG1Bn2iX6+vc7M4kg/cdUiLSU0emG2K9cH1R3HPXHk9K9FmUfqZip+ep4ICcXqaAEj8wUtBHjC0VxlnzOHExPV9ld/6nW31KAen5GyqsE8bMb/mby4reKVt9SQ0E1L5EPc2Tistmze72HZ6ejWIAYWM+qzNcEd7emXUjlZiIgoopo+k/kV6K9iN3uRTa1wIv6br3vjJSa/jP5sbRQJwqjJvqHWhPp6rBme5AJKAIXoOR79Fjrf/ORIRX00hHRfc5ecNo8K8diM3eaU0H9pN6h9IKYD8tXWHmyv2wahxGtGzJgbVTQZYDjOahhQuBgBZsixtr9MQnCdLC12ufvp6NzHzYA29ymkplqkUqootuDtJn5nwYIW+ujXRaufOV+kGDMJHGN/bH8tTXitdNZBTsa47Jemphfav0hQRUf8aUl4c2VKT1q0PCjP9Jr/BaWTzy5YinenwZXRqNtjW7Loq1CWHH9ADc/V9oZby04w7p7VUGQEgmJS40B//pTXVYtj+YhDkxLlVrmjUq7tkkcwRoOhG9r8ovCUjzEviiQe5GKp1hZXVf3K2FuVMPZT4u5FO3XyVtj2aBEhtmuQN3ZCCNumCeUlvhNEJ3cVM+ZYJwv+KbznKHuvafR3dauE4y6ZWmADCyJ4+B1DU1t2jfpyql8VG3c4Fykh5jKCYgPTnl0d1W9/IJ9igNOg12jw6R1fJ4IG1UFZXIWFTaHFTaIY7DYVLVeifGutyEr8C2CONcJUhqCmcjzzd8zhDuOXLVvWDbX6NDovL2JvOKun3XRr6D4Ndgr5uxFFm23hEcoGb8fpb3DqqB8yKE4dFL1HtpaFcxKx31fRVPkG/1VYVCj1dLVtbw93h+kAhADnRxqNUEWauIPQPi/Hq80khnIzHtHHPc/NSfSTsGnWSy9d7fkw/uihwE4upeDaFLx5Mf2guX0T+HDnN6sWndIhPioW85nuev6d6tVzA+KPRGYMLF9EE9zDhC5bV4ipxcJMhc2UF+VpkNgyzSUlehXEJlq2UX+dlH6kt8RnTpk/3o0VZCsvbQUQ1vP1IzH5xSu3tSlT3GrRkNrzMAZRNHBJ58l8Kjt5vt7TZNeIS+/3AqtcHZC/ZmnBxG4Pur4lZwr22SJol5KVIJH+Gs8V19Kq3iVaX4HjITmN4ngeNgf2HLP/ukmkKuYKJkxKP5GfCk3tH3Q7Dira1ROlxSLNM4PYWbP7Pvm8OMe3MDAYTgUBZdqCVKmsYs4VCVyzv/QOstHR+8yYVU2rn6nu0fBU7sYWawoalzUP5UKCVmGjlHOo0f0NWEAhUqIYzXcotyV8TT55WKkTRr6xvZobQ0mlbIfMahDiATbBLxrNvzy+EA1Pw7IZb9ObMWIaQwKPJTZutApIG/eeRe89FQxUmIPXn9i8nXrhRnnGurroUTsaOgHqXThmLIZBPMAR/SRTZgVs060F2rPGFeoO8BrA568Gqa7h2dIVPK0h+oRnpcPMKBDtmizn6cbX2xxZQNnoloX7iy/5sJ7UU94kqMXGdRubtvxldZ9h7HBEPOhbynu+pLjx/eUz80M/HJZNGBgUuG3ewuLY2MPvJYmutxl8RXHUkhISXpjnmU4nfqf5ZH5LTDCByTOkxCE76RHwRyZSJsBPAxd4svm3BfilNEGLHzTyJwq/JbRzcfV+yI9FUtMvE2n8FElSrzKnj1rEwcm3UuiFY3HMg21ufxsV2tSTy0hKcDERnyNQ5uHb4Lfwzapv9hOnULCVK855xLpVtuv3n+w9ybC7XWAiPC/bNLwwPG5UBtMvEpwfCW5mqp1tjbB7NHRmv+VGL+8qZHBcYSvwR+1JI6ZgXLgrvriJQCUcC2QTXMHjZrd+2YBFAdCc39PcdREMHyggLDVpL5w4mnIPZ5Qhtx8fbeBY1MhwZXGd/ZqP996tCJ2wVUX0rf7vM+MUlndoVxIvdZEMzQ9JM4xTMNU5Wk9eiPz+PZplb+ng9g0j1sCITjqQ+TQvJyBmYoR1oqga0q5NSezdXu0Q3uGhVR+4y7JQl0w0L+2KyF+gpfIFue6qiSW17615ndpPn5RL/+WTcxl+4gtmtBenVixbiv14MD3xY0N4L3LK2n6gXdhsMYzhfNWKE7C/y2xW7otfVniW9pwmSpf8pvF8FGdK1f6CZCWqJte3jriJOY/VBbuJ6az2WZ/LTrgaTiSBUhCZB3z66wo5goF0CRo6QgPctE8imGoQ+V4EgoaKq1Uke1F1oAU4FEWhzTnvLKa5PNP+5LfLetOQ6XR4FmGqm3A3I36K1wetAEVHoc9mmtUQgxcDnqBjVvlIGUtHbAQ5RcfUcEe0aSZ0IlsMlachdjkfQke4F+odcQW+LyaIKRdDLJek3+cKNGuOHBmf1Ywa5vZtCBZRib9rzrdV7lIvCmkle6G/AIsROyCdFDyQfYdGVeg3GqScP4utYqu2sJlFtutgrf7TpJ/E+D3LVv4090x601DttflXCB282D//4NWDzeEtcGL7ql5lMIUgOPk5hmefIv0wdv7ifJP3qqm+L96ywqnZF1IpKuwBy7LGKjmGsiimzlFTUiRi12P5e+J8l5hp/ZcpOVKuhSCy60AYm37L+HDDxdp1+fHhZ+v+txcRUWbJg6nd5n5z1G4Cs1INIUPjErzj2ivB8uPhLXHU0nH1Jfixe71l5EmC9AcFcFIh3EKC9LKL3/1Iy1Ej6iO/DHTAM15bjXO931bWc/zhthk6VfYOv38yCrYsSMyYZIuwuIBgQGrq87BQl9u5ok2ckLPAwja1tz68IX1b97PzTpQsGY2nFiI7KWPfMvwCL2oZWL0b1eheI+zUwMadunJEllOVw1R6hnnxt5ECtBBqI499FAIGvcQK6IzJqHaYERbSRdLvvJwYB/D2GoPZPErpEav2pHnYn45lrdh9o9jR2HqStfkr56u+4IZoi/ShhMYOa6rsO37Km7y+PJTyNRxi1IJtJwjW+kvT5Ylqio7CfGNjiNPeLQbz3fpcX7xygF9BouYXu2RgrRipqdQe4gGqMHYx6na2R8YXjWEfdLo7cRt8nfP1KbiMu0xhBZonoYU5lOMZDvvIW+TZsck/LBwVblX2dIj7rdL3h0c1bpC6vgaEYX/6PUCbE9gWK+DfFqSlSd6rJqZ3fqZ+7PxJaEtRjQYNRKrOocEknqjN3nvCZtq5j+VNbtYvvNwp4/eWm2+dPalPrNjpXYmKoWYmPyQn/Hl4++3QudgHq3DYqEjql1tLN1U7oycdewy0fRsu0U5txqqlmd+ezauddMVyD7/tBeYJBhquE/s8O9GMmTf9zL5gwGkdaq6es1d63O0HDoRjQoubs0MMZXt6E3qBupGk2PF2tjOsjJPgnDzN+B0ekeG6rHy+ovoTiYAFZFyZMBe1PrL+H7hnoQpl3PWIllE0tG5wFYetbEDJYf2SrQ/jWUW41AlfDnVMU4LYNEj5SbTyvx2mkVDZcPodO/G1tgRWycEdZnWKNRTIqrVqWbK9y4ns59ihWS6qewOv1C2GncofEFb+0TLRJpikbdWm+CNbEAl2/2TwJdlG02HeOd1BkFm1OC2Vw8cEzuuPcC9Q5U3a8L3fx97aS35UTvi5KN2TDsrQo5ioAm4UUrFb8sz9vwDkExvsJqrlwCgXbfRTksUw1lt8VwbOZqmQAP+nCWap/AT1Bfl+nXbBcG/ree2vrI91Z1fcFeqUBYbKktAbAeGHfsZF3XFiqMghJJnljs/uv27bgCma3aY7a5BxqLwfK67CNbyGNoi0Pq2ZjzI3/87UuWoreRdvlwYl33+ULMEu3AZwKGcW7HxZx4FsaMOMbBTz8IJpSCG064Q+BxS77M91aBy41yBxT2ACtMKR9/HXBdTdAuwA5vlMYyGiyJ1GTZ8n4LWTWy+xqLr2kK3VyqAjAilxvZwt/PfdBcigWu2qZzcc1S+U/loCeFzIY/KRLC5oqKhFJcFWbkzzQ8xLfAb71xo0ueSZOUF7yH6++LNFr7LPnT5Wgfr2Dk+74HFvQiqk38cnvYuKyP9c7qqwixDcFeaN1YJ3De0rYzuQ9tecNudb40iExEYXOSVg1z1d2eWM/ETMyxYaeWRxczcEam9zwPGMOfnNjIPBjt0HEm2JCBtEq+jg8/8vb1QieFE5W3seN+LCONfh2gvFysXQ2eFhWx1kDHNI6Ya0nhBtJlfw3nOdlBy6WQIH+aJvwy+QIaixbWi2eJd6Y036S/VYtHrPgxO7VbhNbVZ1iuN2RtiFDb14VtywXYL7R44pQCsuaQxSh32yEFu9PZM3c/mWKt+2vd/vuL117XD4lNsusXI30PHBGuRoo3wjukyJPlnuG9pOMOR020TXtQN3w+yJIO/hb+Uo9X0y238r7RbBwaAMoVfaZXWBtpg3wz67D8xGl+l0wUDPDgEUrUyUaPeFH1dL+Lmt3FU1olqKX9RHhtb3GHqtQV5x+cuCAz9no/pf7XzoxVyF9ygxkz1rHlz9ggSMuoj6TePOqyuXtgZr6qDX5M4m6zVf6XOZ5vFypM3dN/WEXnkNJbweawgf7krR5++3hwjzracrMIKjDBAwCJo7TlKksoj/6dj9fY5E9fjZQC1zsef0bNtFciqZGmyJ30hlCFnwy6CsVpo0lG8Fw1ch4eF+DcrBe4dAIVQJPD2c/DJxuV3Ahp7oXQ4xUMaBe7ctlsHLMYBYiF8zEZMRuYiNC+mWw2AxsArkx2iSkzYiTxNPCPCwtWv5mgYnP6NrcL03NQTJMrXplXWPKrsPubbuwG0yay1zYBoIvhgYCxozB8NIFyxYHIZkElwbpOBsPUmKdSmlqXP687/nt9z/c0Q23Hr8CEM6N5w3wGTajznAQxeuyObhZvAafyMn9Q3pZQHY11SsDqV2m7o1BOnADZ097tgeJyi0YrN6BRLbPpnyAtW+7Gaz7ezCK/GP+Q087V/8AZuvUB/NWUsFvdli6zK65oih64U4ne6eQZtLBV07sAI52ljxdAfhxruJwnflfLJofYKrUdfdczJXlB6Lfb4WbJyueaYKypN9TZCId5VEcAoes3mWcSD7kra1OAzX6Oa7dEVoQ1pnmLBEAcuZjQj8vy12z6gdUBDUKK2GWd+ClsGfQE+3S5qSNwe9SaPAKHD+GmanGUO+lWeZ0KOldwnG4thZOyxBQitIckZJMnk5LfFt3wbOXSYkMo2Zqw3phrc61W9VCsb7Gz5xnX09EAG0AMU1JcaFHDWrpBVws8Mx6q6g2dNqhr7Tm1TYZjIIdzRt6csWXzRGdJKn76EawScuvJbN8AiU3Ws6qhmGINLvcZ1IZ7wmQIIMOk8H8DEziiS/4K+5CkYCldO2oaAfrRZkTtDnl8JPqXMTQ3KXCsm/fwig05Pfg+JYvozbOJwpR/jpqa3oRJiheP2Ubb1RDHIowdVVSkZkAPCNmhb2Y/zsgbF8+Wc2MPpO7dcvIcksFY/jjx20bCOksLp3ujAXLQTR/vHcieRok3fB7yZ9v2iQAcnUgL4UiaBzq+4kJ/D4c4/hzsMVJWXRDLIVlYdSbiXlW7j6ogxATqVhETQpRIsr9DsvW/21ZB695q+QxlFwvnD+34UX7rkWUOnUHWORnCItNexfNdWtB4yuxSMdSiZUobsxzzu90p9LPD+Rdh+8n/XIiraAOKcRm2Eyszrs5br4uS/2snGFGDw8+PUVbtx94z0b3iHQqQLyHF9QAMvYf6gWiJZpJgthbSMaWHjCy31h0Q0eOgVCN49XLRdowQUdzxX4zjEv9ejiiIGsbCSSxMgJpAG07Wb5LzkY4ylIdL63yS48ChUE8Hd5CGa+S+h1dLMhjt7s+s83dd97SL6XbiQ+YYZ8Rf4RgWATxfvpmHdTvKsnvwOJKz+S7M04RjNNoMXw3w/dI0rls+QLHpsKpVrweeqrKqqT5wun8olh8MpixRhJrUtmDW6H529XB+ou0IYSjRziCSOwMR6ZeLm8ZQ3HczOHMWd7LNn3NgVRJTi1GYgKwqu2hYxNpxPt+R0LyyEWze9F4upajT5KoVLlP1mrcXOMFx052pobCEBMr2IDLFQXlYKQUTonEjXU5efDYvcPnimLO7SCu2N9hSwRq+ODvFqle1fq9MC7hvPm8XdAKyrPdmyiCHO1RKyKYjqSBQYFy5npJw62zetc2lGgzghlWhtAi+lkWhunKSBRhfzDb8UcZaeaU0s2/sAaWT8RDTB9ag0aH0EzM+VukoIfA8y/9nURkE1kAGA/hmzi3W0Ymsc72Y3dXqiL6FCbncDCUT1EoHcNOEQaSGx76d4T4QCzbPRYfL/+SSsmSJPeuxkMR0AHsEISj+tDBGVzGG7zKX8V6ExmDh03WBJ9rxmXS8UEFcayFIcnGd33rEvGwbuOBrVCJhuwe11WbFbNASbe2oq6XzTMgCY1EHCVP2IMjinfKec51u5nDFu4E6EgDesMNEA3TZMn4I5uNxnEhXmBMrEf8yBpwEIEC6wHvY0qJY011MuUJ7HcXPrR9IbJxUxswrxfZeHgDjTXupagn4uir+z9FRAqYJ4Ozab74c1TX07UnBk0b5Hz5J8W9Af25f8B3rPdHKNKK08y0er2YN/uBPaynVba9JeyJUmqjNVwRNBfGLcQegXQvTF564yssm3L5lpuZKqDoAjgO2uNvoCU1o1cZqv0asDWhVzuwEOJnhm6f6nX+d8ObrBubSYvb2E+GTimKvKwW/jSVK5+HcQjmb6fdiPzCWDzBfaGtAtBWyFmLCaOcpgyTkUq83gqbkzphpWKbAPTFRwWN1/4E76vYfArscpoUw3JVbTX0umb6+ixokVpxJWpAwhl5fmwY+UY/bUKd8n8ocIguCzlVGo+yMKDwI6MlxlRgh7e8r4R0HBHZJFFk7sUqfRh0DVJPkZYNfEir4abPRKQB3gpZT2cjFfBjKH9t1/2YNRaFANoRT3moW0klhgGyMuInQ8dQ7WaPJSg3r2zKMOKTkHAIO0y/OWhDuhkj1AnyAMKbmhXt/pEi22YhDJyuc/HRKcb33Dayqfj2XSpuTbxJ/hIfGMidp8y+VuwIlP5uOID8w22W1Qw5t+lVwpoHADrJ4Dovj5PNNRK/vFMIe5nuoLfhP+qZ5w6Klpj8lnIEvNvPQsgXqmMWfOS4Ilj3ZrsIj5y2zXoZxFpkWtnABt24AzWx7xPTLhULgKemTV3ynOBJy4c08HoAVbaO8qmB3f0SJ58P9eP1VJtQ+8RARASwerVDlNSKt1oA/IRFlgVW9aRYRmopVqOsEt1r57kolZfJWyF6vOa4qFJcLhFws6PPGpgavpCx4kDIPfS1WYpb+kMLoNqwQdi/5qgkpCv8nJJZlIf0hq5lnbWFuvZZwU4PApJT1ifFlJWAcqUV3+bAnwHQoxJnJePcRurH968VzyRustnbL6lBZha+TvF1y7TWwY4wzkabYfJfynA11FGHyZDupO2ZQUR6dpFoui7gk8uA8781COuP6rnrwo6I087tw3GNPg0LNPDBZ8UBVr4AL9YKhHccZg93EQ6GAmqpjzzK0u+c67z52z3JEaLCs2v87156W/Ynv5chJh0LmquBy+wFd/Hgp7CEv3Ctyg1sX5NckXvZPLvlG4IQYzQwmSH9NZrnovZkH+uWGvZyCouZ/DtuxbAbvFibiRIBE+Mv2AvsQgcYtd3TqqVzS+5xjYqBasy/nrBm3phh3sPZHzzVMZnIZW6up8DCM1H5S/ctxCYPPx1WdbBI8pswyz2NtjBi9Ve8y45Uka7tcchgqFS+NhYY87BtsWEp34JAaV9E1Z49RdrLteGaPY8Lb+l5mqP8r6o79kHefLI5zwJiS4+Zev2F52J0EnMKWoT25PcWXplSCvUG6AQwk8zLysf+FdjzcQnpb/IB9DTWm5vijc+F5iCM32+0Hq5XwueiXvq4S2ebt8N+0BvK4PuejOfXXAvakdX1xW9V0eWXrnp1iqcYK1oZJO04nLTRO2BjR7qji1Vfu/G/5B1Y8oOWs7o1QnonXPanmZOsruca34iE8a7XX1wDeImyxZKZpsvBiQakNzmqwSm8YeIrNk4H/hcAjUaSR1cdaAJ1ZXOmIs9QHatEg53ALgLtyAfI4/G0NckfMwCQ+FD1/As2d5NldUzHQZAr9k4uPgz6sTEwREff+h3d0b8CSmrCx/FSPtCE0+5U+rEeqXeS+qi+PUXf25qfC1ChoVVl9flVwb7b3OLR8uyyGDSq8K00XxoESYRETM4IuB8kc3dqqg8Ao8mYGZBaSpJ2lOxYoyT+47My2loJk7P4gPzp9u4GqxgktwnQ6YcAmMWSwpWa9OczYz4RpB4E8fSoGd9ng86N6iTe5eET8JYbERT7yBpU4gHEtsBhR4oQK7PHrucWpgjC0j9rDu8snsni4DVNoHQRlSRCu05dHxTrSuPiUl7fbKehxXFERXcSSC0VkTmEKPALa8Co5z2rA9JA6P8wZQww47WjwVxKycpy5dxUp0/6cCE28oW42sECUqBJ6AvI2w+88cWGQNgfVFolfObiC4UR+Lyed929doVoZLTfARz5rHF0x2rR2yXDZrWHxzAlZrr4bdYqreKaSrVku+wDYgnis2Jr0Q7cRXM/3tBDlOX4MIvDRgWxvje7IV+2v+mW8JD3cz0f57+c9A9t/BUNFIUkP7O0XagTcPgDcxTp7qwUyJ5qJ1E5lA6dmC2R/W64JLoFvdHA9JomTmf+R9PH0870Dsp2A0BM7TLXGXzdpmAC4D0g3fSO0gp3m6TrMoc+fB9/WBf1Sh3i8OsUAClxatMEczatadTYdet8b7gcVTCG5FBMFq40bVtTT0giELSKv6IJsqzJgPoCsM05kFBVihWYngAr6XQ5SdxlJtt3Vko4+LtN+1fFZYtfb6+XIQJEaX6FvNny3vm+ERPatPRjL75MjYjQaRAp+/Mrkz5GQ6YzEt8dFqs1JmxXIbqw+I031m9iU0uzo++Ub3mxpyl6jy4hxQQACL1usTBK0mo3TJC9+K+wO4d8qzDEcdQQrjeeXDwkFZj4Xy87qcaDdzhSedaxbgyHZMjsQRarrSGMGqJn2f7SRMd5WPcTh5oKrmJtw29GVsPWCO+C5yCqdQYzRRXSG2EORaeI1JUY5mZsS+i6gsG0HiU4Cxd+XGTlIKF6MjjSPTeyvbQ/3gXmBmdVCUCz8yTs8XJ72SKsENFeCoXlUYmFPxr5gVPFk2YUWOwAJ+JZn8EXyw1/9gcfaE+6fGl9CS333GWXcfOLaZchfwysD8JRXhZ7a/W5ggKtzL053S/sYtKZAhLxJX/ClocNkuuLgcDFTQulMUFqye1HAVVzR406magyZ6L7j4JS3va/lB3NzWoi97qZFmH8RfIVGWGfLgXLdfnkbJIATt+yM5PyTnyvwcS8tzvy6npJ9Scc5qcjBzl0I23EywebIQR11lumjSB7M41X0bJH1Xtc3+lE1H8ViNXUWEIa0IhP9bxY/GRZkBABMtev22nKm1/+xG5kufjnxluHlQ/DN7FGtSheaBw7fLFd49RNcD7pG74dh0Zpen2JnV1AXV1GiSjNUgU/a7m5YCyDOJO2W5+ZG7VBasRKX+8RfqsW5lSfGdUXwBhw8Y+b7RbTlt3C9A9QJvTU/ou8j3Kp1RC/CuBX4nYjYDI6Z4ViDUs9IVyJ767MQg8s4o2h2U/tUj/V1kLksopqRWHvAVAwR+0HgOv5h7RcqGGGYm4zwY2TBjQseoTTN6TUbk98y3pHXNqwTleOiDoRZ3kBlSoWcI+ebrU+26SyZiqYXIHBLatZzcOOrYyE9ABWrBxG7UWWUexXvvAmc5ZNT4KiBn9K34Co8R9kyopP6nSiHq/udd09TnViBD0nKbQlNXo6wHpFp4lW/20QYnwTMbGaeIGGj0Qcd5IxVbUL7Ysi9eFHMMj9yfcpyn9hx1F9l4/VpfvbwKeZTtXjShxrhrqwB1hIQDIjkPMchy/a2JMoPKzkLhjDaYaFBs/wNEj8zBOVtYcIStjefvc/nbm/rSJ3XDK/qhSdiMEqrKwMXQZwEUBxCiWSrfWS0qfSGEMkkmymTZetY11R1I5LF+16F7C8JSrBtTbSY+YYT0WsB5iDn85LfdjsgPoDkkgnQMSTOFs5AMt6SqhA4tvibUmU5jZWgVplNojDACcuSoYfTlF5uc2+j4m9/2zi4Ulak3V/TbT0H+l4i3dQy+XWUDyRvfHa6CTQ2D46/IO7DnDgFkxI86JaZdBPfv1feD8K0jxAjqABRr20jShtO/p+YgXv/jFEUAOiclCjW10sFA7m49sb4mG7T7jmkiy2uKULi4l63I0NrQBkavCa3MG0CpGIF7DuoFcN928+9w4atIOhQV7eUsnntr+oAwU+5+Qd1O2k6SPax8TZv/XOVyrVf53AXYFTV5a52z1AgS+gXoe08tyyqJHqbnAmKER3dSxt+32/VKRPv66gw3D+GAj9tPzVr9NJxxaOVgpFPFJiFsKZW5kc3RyZWFtCmVuZG9iagoxOSAwIG9iago8PCAvQmFzZUZvbnQgL0NNU1NJMTcgL0ZpcnN0Q2hhciAwIC9Gb250RGVzY3JpcHRvciAyMCAwIFIgL0xhc3RDaGFyIDEyNwovU3VidHlwZSAvVHlwZTEgL1R5cGUgL0ZvbnQgL1dpZHRocyAxOCAwIFIgPj4KZW5kb2JqCjMgMCBvYmoKPDwgL0YxIDE1IDAgUiAvRjIgMTkgMCBSID4+CmVuZG9iago0IDAgb2JqCjw8IC9BMSA8PCAvQ0EgMCAvVHlwZSAvRXh0R1N0YXRlIC9jYSAxID4+Ci9BMiA8PCAvQ0EgMSAvVHlwZSAvRXh0R1N0YXRlIC9jYSAwLjIgPj4KL0EzIDw8IC9DQSAxIC9UeXBlIC9FeHRHU3RhdGUgL2NhIDEgPj4KL0E0IDw8IC9DQSAwLjc1IC9UeXBlIC9FeHRHU3RhdGUgL2NhIDEgPj4KL0E1IDw8IC9DQSAwLjggL1R5cGUgL0V4dEdTdGF0ZSAvY2EgMC44ID4+Ci9BNiA8PCAvQ0EgMCAvVHlwZSAvRXh0R1N0YXRlIC9jYSAwLjIgPj4gPj4KZW5kb2JqCjUgMCBvYmoKPDwgPj4KZW5kb2JqCjYgMCBvYmoKPDwgPj4KZW5kb2JqCjcgMCBvYmoKPDwgL00wIDEzIDAgUiA+PgplbmRvYmoKMTMgMCBvYmoKPDwgL0JCb3ggWyAtNi41IC02LjUgNi41IDYuNSBdIC9GaWx0ZXIgL0ZsYXRlRGVjb2RlIC9MZW5ndGggMTMzCi9TdWJ0eXBlIC9Gb3JtIC9UeXBlIC9YT2JqZWN0ID4+CnN0cmVhbQp4nG2QOQ4DIQxFe5+CC3xkiwBDO2WuQRNFmvu3E0MAEaexvP3nRdyb2D3pYyA+uovYh5IPjj1mn3MJWYPwkBLFiefEKWmmO5VmDaNdpZikBmp9CpzJr2hpOm9xV2W2bkS1lfY5MCJYMMxwbLthbIz9ENhL8ecfsF/DLxt9+RfRSTejy0j6CmVuZHN0cmVhbQplbmRvYmoKMiAwIG9iago8PCAvQ291bnQgMSAvS2lkcyBbIDExIDAgUiBdIC9UeXBlIC9QYWdlcyA+PgplbmRvYmoKMjIgMCBvYmoKPDwgL0NyZWF0aW9uRGF0ZSAoRDoyMDIyMDIwOTEyMzgyOSswMicwMCcpCi9DcmVhdG9yIChNYXRwbG90bGliIHYzLjUuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZykKL1Byb2R1Y2VyIChNYXRwbG90bGliIHBkZiBiYWNrZW5kIHYzLjUuMSkgPj4KZW5kb2JqCnhyZWYKMCAyMwowMDAwMDAwMDAwIDY1NTM1IGYgCjAwMDAwMDAwMTYgMDAwMDAgbiAKMDAwMDA3MDY1MCAwMDAwMCBuIAowMDAwMDcwMDAzIDAwMDAwIG4gCjAwMDAwNzAwNDYgMDAwMDAgbiAKMDAwMDA3MDMxMiAwMDAwMCBuIAowMDAwMDcwMzMzIDAwMDAwIG4gCjAwMDAwNzAzNTQgMDAwMDAgbiAKMDAwMDAwMDA2NSAwMDAwMCBuIAowMDAwMDAwMzM2IDAwMDAwIG4gCjAwMDAwMjQwNTYgMDAwMDAgbiAKMDAwMDAwMDIwOCAwMDAwMCBuIAowMDAwMDI0MDM0IDAwMDAwIG4gCjAwMDAwNzAzODYgMDAwMDAgbiAKMDAwMDAyNDA3NiAwMDAwMCBuIAowMDAwMDQ2NjE4IDAwMDAwIG4gCjAwMDAwMjQ2MDggMDAwMDAgbiAKMDAwMDAyNDgzNyAwMDAwMCBuIAowMDAwMDQ2NzUxIDAwMDAwIG4gCjAwMDAwNjk4NjkgMDAwMDAgbiAKMDAwMDA0NzI4MyAwMDAwMCBuIAowMDAwMDQ3NTE3IDAwMDAwIG4gCjAwMDAwNzA3MTAgMDAwMDAgbiAKdHJhaWxlcgo8PCAvSW5mbyAyMiAwIFIgL1Jvb3QgMSAwIFIgL1NpemUgMjMgPj4Kc3RhcnR4cmVmCjcwODY3CiUlRU9GCg==\n", "image/svg+xml": [ "\n", "\n", "\n", " \n", " \n", " \n", " \n", " 2022-02-09T12:38:29.517494\n", " image/svg+xml\n", " \n", " \n", " Matplotlib v3.5.1, https://matplotlib.org/\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "\n" ], "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "state_fig = plt.figure()\n", "\n", "ax = state_fig.add_subplot()\n", "\n", "# Plot means\n", "ax.plot(grid, posterior_state_means, label=\"posterior mean\")\n", "\n", "# Plot samples\n", "for smp in samples:\n", " ax.plot(\n", " grid,\n", " smp[:, 0],\n", " color=\"gray\",\n", " alpha=0.75,\n", " linewidth=1,\n", " linestyle=\"dashed\",\n", " label=\"sample\",\n", " )\n", "\n", "\n", "# Plot marginal standard deviations\n", "std_x = np.sqrt(np.abs(posterior_state_covs)).squeeze()\n", "ax.fill_between(\n", " grid,\n", " posterior_state_means - 1.96 * std_x,\n", " posterior_state_means + 1.96 * std_x,\n", " alpha=0.2,\n", " label=\"1.96 marginal stddev\",\n", ")\n", "ax.scatter(time_grid, observations, marker=\".\", label=\"measurements\")\n", "# Add labels etc.\n", "ax.set_xlabel(\"t\")\n", "ax.set_title(r\"$x$\")\n", "\n", "# These two lines just remove duplicate labels (caused by the samples) from the legend\n", "handles, labels = ax.get_legend_handles_labels()\n", "by_label = dict(zip(labels, handles))\n", "\n", "ax.legend(\n", " by_label.values(), by_label.keys(), loc=\"center left\", bbox_to_anchor=(1, 0.5)\n", ")\n", "\n", "\n", "state_fig.tight_layout()" ] } ], "metadata": { "celltoolbar": "Edit Metadata", "kernelspec": { "display_name": "Python 3 (ipykernel)", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.8.10" } }, "nbformat": 4, "nbformat_minor": 4 }