GAN.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from __future__ import print_function\n",
    "import tensorflow as tf\n",
    "import numpy as np\n",
    "from matplotlib import pyplot as plt\n",
    "%matplotlib inline\n",
    "from tqdm import tqdm\n",
    "from MyHelper import *\n",
    "import pandas as pd"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "downloadData(filename,loc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "X,y=getData(filename,loc)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "img_size,nchannels = X.shape[1],X.shape[-1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "batch_size = 16"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# TensorFlow Helpers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def createWeight(size,Name):\n",
    "    return tf.Variable(tf.truncated_normal(size, stddev=0.1),name=Name)\n",
    "\n",
    "def createBias(size,Name):\n",
    "    return tf.Variable(tf.constant(0.1,shape=size),name=Name)\n",
    "\n",
    "def lrelu(x, leak=0.2):\n",
    "    f1 = 0.5 * (1 + leak)\n",
    "    f2 = 0.5 * (1 - leak)\n",
    "    return f1 * x + f2 * abs(x)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Discriminator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "\n",
    "g = tf.Graph()\n",
    "with g.as_default():\n",
    "    img_input = tf.placeholder(tf.float32,shape=(batch_size,img_size,img_size,nchannels))\n",
    "    \n",
    "    W_conv1 = createWeight([5,5,nchannels,64],'d_conv1')\n",
    "    b_conv1 = createWeight([64],'d_b1')\n",
    "    \n",
    "    W_conv2 = createWeight([5,5,64,128],'d_conv2')\n",
    "    b_conv2 = createWeight([128],'d_b2')\n",
    "\n",
    "    W_conv3 = createWeight([5,5,128,256],'d_conv3')\n",
    "    b_conv3 = createWeight([256],'d_b3')\n",
    "\n",
    "    W_conv4 = createWeight([5,5,256,512],'d_conv4')\n",
    "    b_conv4 = createWeight([512],'d_b4')\n",
    "    \n",
    "    W_logit = createWeight([512*2*2,1],'d_W_logit')\n",
    "    b_logit = createWeight([1],'d_b_logit')\n",
    "    \n",
    "    d_var_list = [W_conv1,b_conv1,W_conv2,b_conv2,W_conv3,b_conv3,W_conv4,b_conv4,W_logit,b_logit]\n",
    "    \n",
    "    def D(img_input,var_list):\n",
    "        W_conv1 = var_list[0]\n",
    "        b_conv1 = var_list[1]\n",
    "\n",
    "        W_conv2 = var_list[2]\n",
    "        b_conv2 = var_list[3]\n",
    "\n",
    "        W_conv3 = var_list[4]\n",
    "        b_conv3 = var_list[5]\n",
    "\n",
    "        W_conv4 = var_list[6]\n",
    "        b_conv4 = var_list[7]\n",
    "\n",
    "\n",
    "        conv1 = tf.nn.conv2d(img_input,W_conv1,strides=[1,2,2,1],padding=\"SAME\")\n",
    "        h1 = lrelu(conv1+b_conv1)\n",
    "\n",
    "        conv2 = tf.nn.conv2d(h1,W_conv2,strides=[1,2,2,1],padding=\"SAME\")\n",
    "        h2 = lrelu(conv2+b_conv2)\n",
    "\n",
    "        conv3 = tf.nn.conv2d(h2,W_conv3,strides=[1,2,2,1],padding=\"SAME\")\n",
    "        h3 = lrelu(conv3+b_conv3)\n",
    "\n",
    "        conv4 = tf.nn.conv2d(h3,W_conv4,strides=[1,2,2,1],padding=\"SAME\")\n",
    "        h4 = lrelu(conv4+b_conv4)\n",
    "\n",
    "        h4_s = tf.shape(h4)\n",
    "        h4_flat = tf.reshape(h4,[batch_size,512*(h4_s[1])*(h4_s[2])]) #512*2*2\n",
    "\n",
    "\n",
    "        logit = tf.matmul(h4_flat,W_logit)+b_logit\n",
    "        \n",
    "        return logit\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Random Distribution"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def Z(n,batch_size):\n",
    "    \"\"\"\n",
    "    Uniform random distribution\n",
    "    \n",
    "    Parameters\n",
    "    -------------\n",
    "    n : int\n",
    "        size of the vector to return\n",
    "    \"\"\"\n",
    "    return np.random.uniform(-1, 1, [batch_size,n])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "# n = 100 #from paper\n",
    "n=100"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Random Class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  1.])"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "def one_hot(x):\n",
    "    r=np.zeros(10)\n",
    "    r[x]=1\n",
    "    return r\n",
    "\n",
    "def mat_one_hot(x):\n",
    "    r = np.ndarray([x.shape[0],10])\n",
    "    for i in range(x.shape[0]):\n",
    "        r[i,:]=one_hot(x[i][0])\n",
    "    return r\n",
    "one_hot(9)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "def C(batch_size):\n",
    "    x = np.random.randint(0,10,size=[batch_size,1])\n",
    "    return mat_one_hot(x)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  1.],\n",
       "       [ 0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  1.,  0.],\n",
       "       [ 0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  1.],\n",
       "       [ 0.,  0.,  1.,  0.,  0.,  0.,  0.,  0.,  0.,  0.],\n",
       "       [ 0.,  0.,  0.,  0.,  1.,  0.,  0.,  0.,  0.,  0.],\n",
       "       [ 0.,  0.,  0.,  0.,  0.,  0.,  0.,  1.,  0.,  0.],\n",
       "       [ 0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  1.,  0.],\n",
       "       [ 0.,  0.,  0.,  0.,  1.,  0.,  0.,  0.,  0.,  0.],\n",
       "       [ 0.,  0.,  0.,  1.,  0.,  0.,  0.,  0.,  0.,  0.],\n",
       "       [ 0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  1.],\n",
       "       [ 0.,  1.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.],\n",
       "       [ 0.,  0.,  1.,  0.,  0.,  0.,  0.,  0.,  0.,  0.],\n",
       "       [ 0.,  0.,  1.,  0.,  0.,  0.,  0.,  0.,  0.,  0.],\n",
       "       [ 1.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.],\n",
       "       [ 0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  1.],\n",
       "       [ 0.,  0.,  0.,  0.,  0.,  1.,  0.,  0.,  0.,  0.]])"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "C(batch_size)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/floptical/anaconda2/envs/tensorflow/lib/python2.7/site-packages/ipykernel/__main__.py:3: VisibleDeprecationWarning: using a non-integer number instead of an integer will result in an error in the future\n",
      "  app.launch_new_instance()\n"
     ]
    }
   ],
   "source": [
    "y_one_hot = mat_one_hot(y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[ 0.,  0.,  0., ...,  0.,  0.,  0.],\n",
       "       [ 1.,  0.,  0., ...,  0.,  0.,  0.],\n",
       "       [ 0.,  0.,  0., ...,  0.,  0.,  0.],\n",
       "       ..., \n",
       "       [ 0.,  0.,  0., ...,  0.,  0.,  0.],\n",
       "       [ 0.,  0.,  0., ...,  0.,  0.,  0.],\n",
       "       [ 0.,  0.,  0., ...,  0.,  1.,  0.]], dtype=float32)"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([ 1.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.])"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "y_one_hot[2090]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Generator"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "\n",
    "with g.as_default():\n",
    "    z_input = tf.placeholder(tf.float32,shape=(batch_size,n))\n",
    "    class_input = tf.placeholder(tf.float32,shape=(batch_size,10))\n",
    "\n",
    "    W_project = createWeight([n+10,4*4*1024],'g_project')\n",
    "    b_project = createBias([1,4*4*1024],'g_b_project')\n",
    "\n",
    "    W_conv1 = createWeight([4,4,512,1024],'g_conv1')\n",
    "    b_conv1 = createWeight([512],'g_b1')\n",
    "\n",
    "    W_conv2 = createWeight([7,7,256,512],'g_conv2')\n",
    "    b_conv2 = createWeight([256],'g_b2')\n",
    "\n",
    "    W_conv3 = createWeight([14,14,1,256],'g_conv3')\n",
    "    b_conv3 = createWeight([1],'g_b3')\n",
    "    \n",
    "    g_var_list = [W_project,b_project,W_conv1,b_conv1,W_conv2,b_conv2,W_conv3,b_conv3]\n",
    "    \n",
    "    def G(z,c,var_list):\n",
    "        W_project=var_list[0]\n",
    "        b_project=var_list[1]\n",
    "        \n",
    "        W_conv1 = var_list[2]\n",
    "        b_conv1 = var_list[3]\n",
    "        \n",
    "        W_conv2 = var_list[4]\n",
    "        b_conv2 = var_list[5]\n",
    "        \n",
    "        W_conv3 = var_list[6]\n",
    "        b_conv3 = var_list[7]\n",
    "\n",
    "        prjt = tf.matmul(tf.concat(1,[z,c]),W_project)+b_project\n",
    "        h0 = lrelu(tf.reshape(prjt,shape=(batch_size,4,4,1024)))\n",
    "\n",
    "        conv1 = tf.nn.conv2d_transpose(h0,W_conv1,output_shape=[batch_size,7,7,512],strides=[1,2,2,1])\n",
    "        h1 = lrelu(conv1+b_conv1)\n",
    "\n",
    "        conv2 = tf.nn.conv2d_transpose(h1,W_conv2,output_shape=[batch_size,14,14,256],strides=[1,2,2,1])\n",
    "        h2 = lrelu(conv2+b_conv2)\n",
    "\n",
    "        conv3 = tf.nn.conv2d_transpose(h2,W_conv3,output_shape=[batch_size,28,28,1],strides=[1,2,2,1])\n",
    "        h3 = lrelu(conv3+b_conv3)\n",
    "        return h3"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Get Variable lists"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[u'd_conv1:0', u'd_b1:0', u'd_conv2:0', u'd_b2:0', u'd_conv3:0', u'd_b3:0', u'd_conv4:0', u'd_b4:0', u'd_W_logit:0', u'd_b_logit:0', u'g_project:0', u'g_b_project:0', u'g_conv1:0', u'g_b1:0', u'g_conv2:0', u'g_b2:0', u'g_conv3:0', u'g_b3:0']\n"
     ]
    }
   ],
   "source": [
    "with g.as_default():\n",
    "    all_vars = tf.trainable_variables()\n",
    "    a = all_vars[0]\n",
    "    print ([v.name for v in all_vars])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Setup Train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "with g.as_default():\n",
    "    z_input = tf.placeholder(tf.float32,shape=(batch_size,n))\n",
    "    class_input = tf.placeholder(tf.float32,shape=(batch_size,10))\n",
    "    img_input = tf.placeholder(tf.float32,shape=(batch_size,img_size,img_size,nchannels))\n",
    "    \n",
    "    G_z = G(z_input,class_input,g_var_list)\n",
    "    D_x = D(img_input,d_var_list)\n",
    "    D_fake = D(G_z,d_var_list)\n",
    "    \n",
    "    \n",
    "    loss = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(D_x,tf.ones(shape=[batch_size,1]))+\\\n",
    "            tf.nn.sigmoid_cross_entropy_with_logits(D_fake,tf.zeros(shape=[batch_size,1])))\n",
    "        \n",
    "    loss_gen = tf.reduce_mean(tf.nn.sigmoid_cross_entropy_with_logits(D_fake,tf.ones([batch_size,1])))\n",
    "    \n",
    "    opt_d = tf.train.AdamOptimizer(learning_rate=.0002,beta1=.5)\n",
    "    opt_g = tf.train.AdamOptimizer(learning_rate=.0002,beta1=.5)\n",
    "    \n",
    "    d_grads = opt_d.compute_gradients(loss,var_list=d_var_list)\n",
    "    d_minimize = opt_d.apply_gradients(d_grads)\n",
    "    \n",
    "    g_grads = opt_g.compute_gradients(loss_gen,var_list=g_var_list)\n",
    "    g_max = opt_g.apply_gradients(g_grads)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Other"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From <ipython-input-22-f4afde653f64>:2 in <module>.: initialize_all_variables (from tensorflow.python.ops.variables) is deprecated and will be removed after 2017-03-02.\n",
      "Instructions for updating:\n",
      "Use `tf.global_variables_initializer` instead.\n"
     ]
    }
   ],
   "source": [
    "with g.as_default():\n",
    "    init = tf.initialize_all_variables()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Session Trains"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "sess = tf.Session(graph=g)\n",
    "sess.run(init)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "losses = []\n",
    "losses_g = []"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "3750.0"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X.shape[0]/(1.0*batch_size)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 3700/3700 [12:47<00:00,  4.82it/s]\n",
      "100%|██████████| 3700/3700 [12:48<00:00,  4.81it/s]\n",
      "100%|██████████| 3700/3700 [12:48<00:00,  4.83it/s]\n",
      "100%|██████████| 3700/3700 [12:47<00:00,  4.84it/s]\n",
      "100%|██████████| 3700/3700 [12:47<00:00,  4.83it/s]\n",
      "100%|██████████| 3700/3700 [12:47<00:00,  4.82it/s]\n",
      "100%|██████████| 3700/3700 [12:47<00:00,  4.82it/s]\n",
      "100%|██████████| 3700/3700 [12:48<00:00,  4.79it/s]\n",
      "100%|██████████| 3700/3700 [12:48<00:00,  4.80it/s]\n",
      "100%|██████████| 3700/3700 [12:48<00:00,  4.85it/s]\n",
      "100%|██████████| 3700/3700 [12:48<00:00,  4.82it/s]\n",
      "100%|██████████| 3700/3700 [12:48<00:00,  4.79it/s]\n",
      "100%|██████████| 3700/3700 [12:48<00:00,  4.79it/s]\n",
      "100%|██████████| 3700/3700 [12:48<00:00,  4.82it/s]\n",
      "100%|██████████| 3700/3700 [12:48<00:00,  4.82it/s]\n",
      "100%|██████████| 3700/3700 [12:48<00:00,  4.80it/s]\n",
      "100%|██████████| 3700/3700 [12:48<00:00,  4.83it/s]\n",
      "100%|██████████| 3700/3700 [12:48<00:00,  4.80it/s]\n",
      "100%|██████████| 3700/3700 [12:48<00:00,  4.82it/s]\n",
      "100%|██████████| 3700/3700 [12:47<00:00,  4.82it/s]\n",
      "100%|██████████| 3700/3700 [12:47<00:00,  4.80it/s]\n",
      "100%|██████████| 3700/3700 [12:47<00:00,  4.82it/s]\n",
      "100%|██████████| 3700/3700 [12:47<00:00,  4.83it/s]\n",
      "100%|██████████| 3700/3700 [12:47<00:00,  4.80it/s]\n",
      "100%|██████████| 3700/3700 [12:47<00:00,  4.84it/s]\n",
      "100%|██████████| 3700/3700 [12:47<00:00,  4.81it/s]\n",
      "100%|██████████| 3700/3700 [12:47<00:00,  4.83it/s]\n",
      "100%|██████████| 3700/3700 [12:47<00:00,  4.81it/s]\n",
      "100%|██████████| 3700/3700 [12:47<00:00,  4.81it/s]\n",
      "100%|██████████| 3700/3700 [12:47<00:00,  4.82it/s]\n",
      "100%|██████████| 3700/3700 [12:47<00:00,  4.83it/s]\n",
      "100%|██████████| 3700/3700 [12:47<00:00,  4.83it/s]\n",
      "100%|██████████| 3700/3700 [12:47<00:00,  4.83it/s]\n",
      "100%|██████████| 3700/3700 [12:47<00:00,  4.83it/s]\n",
      "100%|██████████| 3700/3700 [12:47<00:00,  4.80it/s]\n",
      "100%|██████████| 3700/3700 [12:47<00:00,  4.78it/s]\n",
      "100%|██████████| 3700/3700 [12:47<00:00,  4.83it/s]\n",
      "100%|██████████| 3700/3700 [12:47<00:00,  4.83it/s]\n",
      "100%|██████████| 3700/3700 [12:47<00:00,  4.81it/s]\n",
      "100%|██████████| 3700/3700 [12:47<00:00,  4.82it/s]\n"
     ]
    }
   ],
   "source": [
    "num_steps = 3700\n",
    "extra = 3\n",
    "num_epochs = 40\n",
    "for _ in xrange(num_epochs):\n",
    "    for i in tqdm(xrange(num_steps)):\n",
    "        img_batch = X[i*(batch_size):i*(batch_size)+batch_size,:,:,:]\n",
    "        fd = {z_input:Z(n,batch_size),img_input:img_batch,class_input:y[i*batch_size:i*batch_size+batch_size,:]}\n",
    "        l,_ = sess.run([loss,d_minimize],feed_dict=fd)\n",
    "        for _ in xrange(extra):\n",
    "            lg,_ = sess.run([loss_gen,g_max],feed_dict=fd)\n",
    "        losses.append(l),losses_g.append(lg)\n",
    "G_test = sess.run(G_z,feed_dict=fd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "fd = {z_input:Z(n,batch_size),img_input:img_batch,class_input:C(batch_size)}\n",
    "G_test = sess.run(G_z,feed_dict=fd)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7fb5c8c33d50>"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAfwAAAE6CAYAAADz4+C6AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\nAAAPYQAAD2EBqD+naQAAIABJREFUeJzsvWusbNtV3/mf9X7u93nZ1wInhgBKhwbjJjQ4kHZLJJYa\nnC8g3JLjbiHES6L5AJYl1HYACTWPiA7EEWrRJqgDktWBQBDcC00TiAHbbTsmcWhwG+61772+55z9\nqMeu92v2h9r/ecaatVa9dlXtVVXjLy3V3rWrVq1dvzXGmI8xxzTWWqhUKpVKpdptJe76AlQqlUql\nUq1fGvBVKpVKpdoDacBXqVQqlWoPpAFfpVKpVKo9kAZ8lUqlUqn2QBrwVSqVSqXaA2nAV6lUKpVq\nD6QBX6VSqVSqPZAGfJVKpVKp9kAa8FUqlUql2gOtLeAbY77PGPOiMaZtjPmIMeYt6/os1d1IGe++\nlPF+SDnvh9YS8I0x3w7gZwC8D8BXAfgzAC8YY87W8XmqzUsZ776U8X5IOe+PzDo2zzHGfATAR621\nP3DzuwHwMoB/Zq39Se+1pwC+GcBLADorvxjVMsoB+GIAL1hrL8NesAjjm78r5/hpKmdlvBNaqS0r\n41hqJmMna+1KDwBpAH0A3+I9/0sAfj3k9e8EYPWI5fHOVTBWzrE/Jjgr4507VmLLyjjWRyhjeaSw\nep0BSAJ44j3/BMDfCnn9SwBQKpXQ6XRQLBZvfQHNZlPPc4vzDIdDNBoN4IZNiBZl7M6VTCZRKpUW\nuNpwreq7WOW5tu08MzgvzVhtOT7nWYMtvwQ8Y7wKW240GkufR45Q+9/FeKAi/LW+5GsbjcbSbPzz\nbOL7GQwGsxg7rSPgL6oOAHQ6HQyHQ7RaLfeHbDaLbDa78AmNMUilbv+v7cN5ut0uut1u4O/CMFY5\nZNcBgNFoFGAMLMd5Vd/FKs8V5/NsiLPa8h2eZ1sZJxKJpb8LGcT973TZgH8bNvI8t/m/pOR5whiP\nRiP+OJPxOgL+BYAhgAfe8w8API56U7FYRKvVwsHBwRouSRWlMCMdDAao1WrT3rYUY2Dcw1fGm9cS\nnJdmrLZ8N9qkLZdKJWV8B4piXK1W53r/yrP0rbV9AJ8A8DY+d5ME8jYAf7Lqz1NtXvvGeIH8lYnf\nt1XKePcZA9vHOYzHqs+7y1rXkP4/BfBLxphPAPgYgB8EUMA4EUS1G1LGuy9lvB9SznuitQR8a+2H\nbtZw/ijGQ0OfAvDN1trzae9bZv5Hz3M351mWcTqdvvVnA6v7LlZ5rkwms5LzxIXzsoxX8dlxPc+u\nMQbUX+/LeYA1rcNf6AKM+WoAnzg8PFxZEpbqdhLzfm+21n5yFefcZs7L2oifNBQ3rZqzMo6f1sX4\n6OjoThgvwmmRpL1V6S7uBzGHP5PxdllljHWbG1GlUqlUq9U0nxx3fz3vioJFpZvnqFQqlUq1JbrN\nSIX28Ncgf22oane0rpa3Kj5SxqpdlfbwVyzfWdx1joRKpVKpVMCe9vCTySSSySQSiUToo/x71PpM\nY8zEc4lEAolEwvUC5HuHwyEGgwGGwyGGwyFGo1HgsNa6x31ZE7puSZbkycN/Tq6xpiRH+SjfByDA\nzedMrrNYq5bTPIxn2TIln5fn4d/UlqM16/+7zciI75tn+Wt+99Ou1RgTypjvnWXH8vcw37GMNnGP\n7F3AZ9nEbDaLTCYTevBv6XQao9HIGTXfz0dpyMYYpNNpd/Cm4dHpdNzR7XbR7/cxGAzcI2+uwWAw\ncWOpFpfkLHnykc9ls1mkUqnAd87385BOW3LOZDIBzoPBIMC51+uFcuahnG8nsvBtmWxm2TLPAWCq\nLQOItGUy5qG2vFpN89dkz0Pa8XA4DD0fOScSiUjG89rxNjLey4CfTqeRy+VQKBSQz+fdI38uFoso\nFArIZrMTrbxpgSCfzyOXyyGXywFAwBFcX1+7o9lsotvtuuDf6/XcYYzBYDAAEKiRrFpC5FwsFh1f\nMvI5S4dOh8BegGzRJxIJd458Pg9rLfr9vnMKZNxoNJTzmsVgQJ60Y7KZx5Z5HtrxNMZ0+D5j8mVw\nUMarlWQs7Tifz6NYLIba8WAwCDCmaMtRjH1/Pc2O+/0+jDGBwB93rTzgG2PeB+B93tN/Ya39ilV/\n1jJiwM/n8yiXy+4olUqB38vlMvL5fMDQZSCgk2DLLplMBm4+ay16vZ7b7KBaraJSqaBSqaBer6Pd\nbqPVaqHdbruWJIeleJPSMcVNcWcMhHMulUru8DnL1juAiYAvHQU5F4tFWGsd406n4zhXq9UJzmSd\nTCbdPQQo52VljEEmk0E+n8fBwcFMxr4tSwYy4Pu2DMA5emnLZNxqtdSWZ2jZZXDT/HWpVHLcDw4O\nQhnL801jHOWvw+xYjtQmEgn0ej13Xn9kYVUJoKtKFl1XD//TGNdi5lUO1vQ5C4stu+PjY9y/fx8n\nJyc4PDzE0dERDg8PA0NHyWQyMFTnD+vLnp8MDv1+H6PRyDmJfr8PYFylq1QqIZVKBVqKvMkYNK6v\nr1Gv11Gv1yd2RoqRYssYiObMQw73knNYMDDGBIbs5Jx/v9/HcDh0HAeDAYwxyGazKJfLSKVSjq/k\nTIehnG8nyfjBgweO8cHBwVRbpn1KyfnbMFuWvTrgmS0nk0kUCoVIzsr4dkokEigUCpF2nMvlnC3P\nYiztmLZN+41iXCwWHeNyuTzhs33GUVMJcdG6Av5gnvKbdyEOvR8fH+PRo0fOUfCg6ADkHJ2foOHP\n+9JJ9Hq9QCDg0E82m3VOSib8yJuo0+ng6dOnsNai3W671mMMFVvGQDTn4+NjHB8fTwzj+o5Ctqh9\nzslkcoIz3ys553I5x5ivk45Cco5xMIgtZzI+OTlxjMl3li2zxy2PWbYsGWcyGcdaTgf1+30X+Dud\nDs7Pz9WWb6FZ/tqfWp3G2M/hiGLc6/UAwDH27dhnvCX+GsD6Av6XGGNexXh/3j8F8F5r7ctr+qyF\nZIxxLcZHjx7hueeew9nZGe7du4ezszP0+/2J5DoGZHkD+Zm5srfPOR2+l8PEbInK3iMwLo3Iz2i3\n2+7mqVQq7ppjeBPFljEw6SjI+ezsDKenp44zgy+dBFv7UcFA/uxzlgE/l8sF5oiNMQFHQc6tVgtX\nV1d3/G1NVWw5yx7+w4cP8dxzz+HevXs4PT0NtWXZAOcc77QMe5+xnPKhLQPBRF7acr/fR7vdBoAJ\nWwZit1w3tozD7Ji++t69ewsxlkHbz8afhzEP6a87nc62+GsA6wn4HwHwbgB/CeARgPcD+CNjzN+2\n1jbX8HkLicN1zPCUiR/lchmdTicwzCNvBpn0w5vC7/FLRyF78VKJRCKQTczPYQZwrVZDPp9HKpUK\n3Dz+TXSHRUBizRgYfzfJZNJxlol65MzWvWy5y+kb8vMlA4Sf7CcZ8T6TWeK8n7rdLmq1GgqFAtLp\ndGi2uPxf7kix55xMJl0WNxP1OH8fxpjO2s+0lpply5IPe/v+6hx+Tr1eD7XlGCl2jP3vV9pxGOPh\ncIhutzvBmL18cg5rxM/y18zb8f01z0s7nsdfT/s/+b9G/W1VPmHlAd9a+4L49dPGmI8B+ByAbwPw\nwaj3NZvNiX+Cyy1WfH3o9/totVqo1Wq4urpCKpVyR7vdRqPRcNn0nU7HJWrIOfler+ecTdgaYIo3\nrEzasta6RgbnAZmcwh5iOp1255Q9Tp5zVeIQs/8dzfgOl2IM3B3ncrmMdDod4CyzcGUyjuwl9Pt9\nx5msZUKfNPBEIjEX50wm44b8Ob/MaYJd4LxJxr1ezzGuVCoTtkzGjUbDceEUisyoj7LlZDIZ+Mxk\nMhlIwpSMi8Wic/rpdBqJRCLUlsOWb62C9SYZNxqNgJ+z1m6MMRtXYf5a2p+/aiKRSARseBpjnmce\nxmwM0L553etk7E85zqu1L8uz1taMMZ8B8KZpr+MXuYHrCdxA7GHRSbRarcCSDGbg+pnWnU4nsAbU\nXyPqFwTp9/uo1+u4urrCYDDA8fExBoMBkskkcrmcuwa2ZKWTCGsxrioYhBmp2GFr3u90LsbA5jgD\ncJzr9fpEMJjFWa7D5QgBvyvZa/eDRK/Xcw3J4XA4wVn2FHiuVCq1U5w3acuyUcfPJRMG/Hq97hg3\nm03HmXzb7XbAliXjTCYzlbG0ZU4x8H0AAsFglxgz+Vi8b7GLnf960O/30Ww2Jxj7dnx9fT3BWAbu\neRknk8mpjGnHvr9mAyJstG/VjOX5xG55M7V2qzTGlDC+eX553Z81j3wnIYOtvIHq9Xqgd8CeIG+m\nZrPphoi5tEOu6ZfFIJitXavV8PjxY3S7XQyHQxcEOGTF65DBxF86dIfDu5GKG2MgnLPvKJhZy97B\nNM6lUinAmQcbe+w9cIjv8ePHbuiYTkKyZY0AOg72/njtynm22HhnMJjFmHzJWx7z2LIfDMJs2W8M\nykadtGWZKR4nxZWxb8f8TslYNuwkY+m/w+xY1m6Qgd9nHGXH/JnvizvjdazD/ykA/xbjYaHXA/gn\nAPoAfnXVn7WM7M266Uaj4ZKlOp0O6vU6Li4u0Ol0Ag7fX0ftDwt2Oh20Wi0UCgUcHBxgOBwG8gSM\nqBSVz+ddyziZTGIwGKDVarnf5RDRwcEBzs7OMBqNAsFnHcO9iyrujIFJztbaSM5yKDCMM7Prm82m\nywHgEjwGej5KzlyPPRwOA5w5ApTJZFAul3Hv3j3HmYdynq0oW67Vari4uEC73Q7YclivjwcZt1ot\nx5gOW9rmvLbMRiAZqy0vLk6X9Xq9UH99fn4+4a8lY/9xHsYcqvcZp1IpDIdDNJtN1/DjyI1vx+Qr\nay/EJeivo4f/HIBfAXAK4BzAhwH8XWvt5Ro+ayHxS+/1eri+voYxxhVZePLkCXK5nMu85FyJnLOX\nj0zu4g3UarVcsGcPjuu2M5kMCoUCjo6OMBgM3PASALRardAb6PDwEI8ePUIqlcLl5aXL6JZDgnd4\nE8WWMcVgQM6dTgeVSsW15sM4S7Y+506ng0wmg1ar5YK97Klzbr5QKODw8NBlaWezWceO88QM+Ol0\nGkdHR3j06BGSySQuLy+dw1DOs8VgcH19jUQigW63i0ql4qpdRjEOs2nacjqdRrvdxnA4DCRdhjGm\nLXP4PooxbZmM1ZZnS34ftGPJeJYdhx1M5mMjj4w5vWatdcyKxWLAX7OMtmRMf03GDx8+DDDm/D9w\n54yd1pG09x2rPucqJXsFg8HABQS55CKq3rl/yMSPfD7vHAKHipjhmU6nUSgUXKBg0GCvQLYqOVR0\neHjonmcQiMMNA8SfMRAM+P1+H7VaLZBstyxnOn5W/8rn8+41kjMA1wgkP+lc2FggZ9YBj9Myvbhz\n9m25Xq+H2jKztcPY8nlWTUskEs7WptkyudGWOYojGfu2nEqlXCBQW55Psxhzdc00O5b1TtrttrNj\nJlWyLHMUY/bW6a9nMR6NRmtlfJvz7l0tffYKRqORy6SWVdb8Ndjy5pKHPB+NOKxGP4d1eBMB49rQ\nzWbT9Rw5Z8QblkNELOByfX2Ny8vL2DiJbRA5D4dDtx7adwzzcpavo5HLGv2yyAcbAvyZQ43kXCgU\n3H0mORtjHGcm/qimSzKeZst+BT3JNsqWGew53yvrcPi2zCFcn/FwOJyw5UajobY8Qz6T2/jrsPPJ\nFQWSMZkBCPXXXPM/zY7J+OrqaibjaX9f1/2xdwEfgFsry+EW6RDCvuiwDFTOL/Fg64/FFzjUxHlE\nWdiBBwA3/McWYy6Xc9fHEQEe0lmpw5gtOnv587Q5tSjO/Bu5LMKZvf1FOEc1LlWT4hys3KTGZywf\n/cDgPx/FmMPA89oypxUkY0pteTEt6q/5HuAZ23kYMx+Lwd0vxsORPWnD22bHexfwpeNnMJgWCOQc\nDId8+LqwG8gY424eGQjkueUNzLkg3kQs1cmsUL4vTjfNtkiylrynBftZnNmwA4LBvtVqodFoTPQm\n+PlyCU8Y5zg6h20Qbcm3ZWlrUTYNTDbolmUMPAv4PmNWYFTGi2tRf8338FFuYCT/HuavyZj+mufn\n64HJhvs0O5Z1CuKivQr4fkuPz1HTWotRyRe8ETn/yiQiZn5zaYjce5lZ3cz69W8gFnuJe2sxzpLM\n5HNSYY5gWpKNDAbMHJbZweQs115z/bVyXo/CbNm3adq8P/zLdfEy6Pu2LBlLW/bX15Ox33jv9/vO\n3pXxcprXXxtjAiN6fAwb0pd2TH/tM57mr2WjTtoxryOujPcq4AOTPTkJya8cFfYoDZetPjoKOYRr\nrQ04c+67bIwJ3DS5XM7t1FcqlVAoFFy5SDYYuNaX16maX9JZTDPAWZzl3+UmKgwcvHeYxMfhPzoF\nyblcLkdy5v2jnGfLt2VgNuOouV7fluXeCEzK5DklYy7TI2Pu4ubbMoOJMp6tsKm1Wf7af92i/jqM\ncaFQCJRh50HG0o6ZqNdqteZmPKsDsg7tXcAHJluMMgs7qiUedg46fn+uGECgPjPhc8kd1+zLg3s6\n53I5VKtVdDodXF5e4vz8HLVazSWsqOaX38unk/BL40Zx5ns5Ryt7ibLXKGtrs6Hnc6aDIOt8Po9K\npYJ2u43Ly0s8ffoUtVoN3W5XOS+gaYyj+MoePcXledKW5fmlLdNnhNnyNMa0ZWW8mOR3Nc1fy2m4\nsHNM89d8Tu6fITP2fbaScbVa3RrGexnwKdnSZ01l/5A3FYDAfBKHBMM235A3kHQSvIGOj4/d7m2l\nUsmNAHBUQN5A9XpdA/4ttCxnHuzlSc5hjoJ/l8Hg+PjY7e7lc2bt/YuLC+V8S0nGPlf5swwMsq79\nNMZ+ow6AW3/t2zJXbrAiGxmzUaeMl5M/MhNmx35DALidv5YN96OjI7fTZqlUcktypR1fXV3FnvHC\nAd8Y81YAPwTgzRjvrvQOa+1veq/5UQDfCeAIwB8D+B5r7Wdvf7m3lxzy8W8gVlSSpXZ5M3ENrb87\nmkyu4/mBSSfh9wqOjo7w4MEDvO51r0OhUHCfwRtTOgkOE23qBhIG8bwx5gxbxhi4PWe5M2IYZ0py\n5mv9YPDgwQO8/vWvdw6CVbtkw46cN+UoWHb4Rh83xuwEYznX6h9Rtkz7jLJlv/c3zZa5a5pk3Ol0\ncHV1hfPz88DGSuvWLjCWirLjKM4AFvbX5AxggvH9+/edv+Zn0V/LEdlNMl5Uy/TwiwA+BeAXAfya\n/0djzHsAfD+AdwF4CcCPA3jBGPPl1tpe1Ek3OZ8lE3lk1SQ5V+MnZsgNbfieWq2Ger2OWq2GZrMZ\nqOLFak2sxc1a6cwG5RpO2erkMhC/5CvX6G/qexI36k8A+OmQ728pxqvWLIOah3PY4W+kQcas1S2r\ntGUymQBnltLtdruOHUcJJGfWgPc5bypXg9/JzX018UVuK2O5L0WYDctiKb4tS8ayCh+d/jy2DCCw\nnItJYGS8yXycbWE8S/ye2Kv37TiK86r8davVmvDXnP+XiZ13wXhRLRzwrbXPA3geAEz4f/MDAH7M\nWvtbN695F4AnAN4B4EPLX+pqJC9Z3kBy2ZQsnMOtTeXGGny8vLzExcUFLi8vUalUAsbN6ntyCJeZ\nodyqkw5e7s3NHb5kwRa5n/MmlE6n+eMfAtg6xsDkGmzpKLhOOoozf+ffrq6uJjjzYNUusmbRHJ8z\nh4t53DVn1nq/2U51pxiTcxhjeYTZ8sXFBarVasCJ+4yn2TIDQZgtd7vdQJGmdWsXGFNsTIXZMYfX\nwxhLzsv662w26xp2Uf6auzFumvGiWukcvjHmjQAeAvh9PmetrRtjPgrg6xCTm0gm8UQFglKphHK5\njMPDQ5ddfXx8jJOTE5ycnOD4+BhPnjzBa6+9htdeew1Pnz5FpVJxh7U2UH6RWaLcv1n2DPr9vqsF\n3Ww2AzcQh4YW2fN4ndoWxsAzzv4wvgwG5XJ5gvPR0ZFjLDk/fvw4wLlarcJaG9hm0xjjnEEmk5no\nubPuN3f04r3Q7XYDw4l3qV1gHGXLPJaxZck4ypZZPVMyZs+P2dxxCAS7wFjasc/46OgIR0dHOD09\ndawX8dcAnB1zfwU23KW/jjPjMK06ae8hxkNHT7znn9z8LVKb/oJk4odcJidbiuVyGcfHxzg9PXVJ\nOffv38e9e/dw//59HBwcuP2veQMCcC08uV5XJo/IeSUebG3WajXXw+h0Ouj1okfW7mi4aGnGm5bM\nxOdQr+wVsAcQxvnevXvuIGe5hzYD+2AwCKzZlYlgMg/A51ytVlGpVAI9RCB8LXnY/7VmbT3jabZM\nzovachjjabZcrVadLUvG8/5fa9bWM55mx2T84MGDuRnL5/05/XkZc0VPXBWbLP1mszlxk3MuZh2S\niT6ch+XaWZlhK7fDzGQybl6nWq3i+vraFW/gblvNZhP1eh0AUCwWXWKYdBhcppVOp93wLocTnz59\nii984Quo1WoLOYhlxZaq/92sS5vkLJdskTN79T5n9hbI2VrrtlqVDps9OHK21gaSeKI4sw73+fm5\n4/zaa6+hVqsFlv2tS5vkHFfGvi2TcZgt+4yLxeJMxrTlaYzXqU0ybjQaE8lvd8GYw/DcvlhOy2az\nWbc+Pspft1otXF9fu/P7jFOplGOcyWTcyE6cGC8y+rvqgP8Y47miBwi2HB8A+A/T3kiD2qT8bNvD\nw8PATmbyBpJOot1uYzQa4fr62gUCOnTeQHxvPp93CSGlUgmlUmnCSVxfX+P8/Bwvv/wyXn31VTx9\n+hTVanUjN1CYkQ4GA9Rqtai3LM0Y2Dxn6SjkLlgcnvWDAR0Fg4G11o22MBFLBgP2Oji3S0c0jfMr\nr7yCV199FU+ePEG1WnXrv9epBTnvJGMZDNh4j7Jl2Xjn/cEMfJnvEcb44uIiYMubCgabZMx94ql1\n37+L2DGH+WWjzlob6q8lY99fkzOnCsi40WjEjnG1Wp3r/Su1Smvti8aYxwDeBuA/AoAx5gDA1wL4\n56v8rFXI7+EfHh4ilUpNLO8JcxIM7Gwx+jcQb0prrXMSJycnOD09xcHBgZsr6vf7aDQaLuD/9V//\nNRqNhssGj5u2ibF0QrJncHBwECiXGRYMALj9EKb18JPJpNsdj5w5rHh4eIhcLuc4y4bdiy++iOvr\na1xfX29kJGcR7SrjsEbdvLYsGXO5JW1ZMqYtv/LKK7G25V1jLIf752Use/jk6vvrk5OTAGO/4R5n\nxmFaZh1+EcCb8Czr828YY74SwJW19mUAPwvgR4wxn8V4qcePAXgFwG+s5Irn0Kz5TzlswwQPGjCX\nXTAZi/M8XGoji3dwnWu1WsXV1RWur69dKU3OB+XzedfjY+a3HCngGl0O6V9cXKDX6wWWdmxa4vv7\n0pvHWDBetBch19kfHBy4ZDxy5jnZQwDgevEyUAwGAzSbTVSrVVxeXrqewmAwcPeRz9lvIJIzM4XP\nz8/d8NxdcOb8s9BOMqYt0x6BZ4xlMhhtuVKp4PLy0hVPkfP3uVwu1JajGN+1LW+SMZdGrkNRjM/O\nzgJVE/15+SjG9Nf1eh3tdjvAOJ/PB2zYn/7pdrvuHokD40W1TA//awD8AcYJHxbAz9w8/y8B/I/W\n2p80xhQA/ALGxRz+PYB/GJd1ncYYZLNZN2TDwig8gGfFGtjzGwwGqNfrGA6HKBQKLhHo/PwczWYT\nT548wSuvvOJKKnL4r1wuu0zRQqGAdDoNa61busGDGaMykHDP9ruQuHH/FbaQMRDkXCgUcHJygvv3\n7+P+/fsBzlxbm0ql0O/3Ua/XMRqNAku2WBTn8ePHeOWVV1CtVtHtdl3vvlQqOc6lUsk5iDDOdDSt\nVstxvouMXg5N3minGdOWfcZM6isUCpG2TMa05ZOTE5ebwyDAzsBgMIiVLW+S8bru4TDG9NX37993\nKybk/hVh/noextKO5ZQA7Zgs48R4US2zDv8PAUzd989a+34A71/uktYrrqk9ODhwJTHpJO7fvx9Y\nLz0cDgNOgokpmUwG5XIZiUQCrVYLT548wauvvuocO4eHSqWSWxLCVipvILnL2vn5ubuBWq1WoHrY\nXUjMzb3FWvvJsNfEmTEQzlk6Csl5NBq5xDomXDKX4+DgAIlEIuAoOCQogwE/gz3+MM50FAz4ZHwX\njoKjWzfzu6Gcd4Uxl8z5tsxev2T89OlTvPrqq46xtGUGA1lljcPCrVbLJXOFMb4LW951xg8ePAgU\nwRmNRkgmk1MZR/lr2vHh4SFOTk4cY9pxGOO4+OtFFJss/U2JLcZyuTzhIO7fv4/hcOgqMNHoORRk\njMHR0dHEDfT06VO88sorLoGE8/eyhy+XeHD9ZrVaRa1WC9xA7XZ7oua3anEtypmBgZyZpMOGnQz4\nMkvbH8mZxZlTP61Wa2ITH9Vimodxt9uNtGXJ2BjjgkEYYwb8k5OTUMas1CcZqy3fXmGMOYLz4MGD\nCTvmwT3tZ9mxnHpdhHGUv467Hccm4G/ii+J8jkzSY0IGS6LSKXBeho+sZZ9Op93SDC7HGAwG7pw8\nmEEqW5tcu8ulQLLyE+cM2dvj/KM6iuUUxZlz8iyLKTnzZ/b4WTGPyySZmyGH+4+Pjyc48+CyPsmZ\neR4+50X+L9VYMmObu5exLKosVR1my+zx+4xpy5Lx0dHRXLbM4/r62n0GoLZ8G4XZcT6fj7RjzrPT\nxqIYMxOfnNmR6/f7qNVqE3bsM47y14vY513YcmwC/roly3CGOQljTMA5yJuHldCYmNFsNnF1dYVX\nX33VzQfmcjnXQjw6OkK5XEY6nUav10OlUgnA5brQi4sLPHnyBLVazTmJbRkairsW4Sx5s9jRcDhE\no9HA1dUVXnnllQBn9ujpgOgoyJlHu912AZ9LLZXzajUPYxkEeHDP8kajgcvLS7z66quo1WoBxqza\nRsa9Xi+w/ImMGQzUllevWXYsq96RLRMoybjZbOLy8tLN249GI5cXIO04lUqh2+06xvTZ0l9zCd62\nMt6LgC+DrVzWwT3K0+l0ZCDgvE273QYAtFotXF5eIpPJOAOXToLbobI3SSchs76ZKUon0Wg0AqMI\nwOYrD+6i5HpdyVn2COgoyLjdbsMY45xEGOeDgwOXJexzlpvutFotF/AlZ7nuXnv3y4u2nM/ncXBw\ngFKp5LIGbv3XAAAgAElEQVS0ZzG21k4w9ht1ZMxNkfr9vrNlDuH6jfdms6m2vEL5/pqNL/prP9jT\nX3OXQHbO0um0s+PhcIhsNus6aLTjZDLpOmjSX7fbbZeZv+3+eicD/jTHSEOV8zcy01PO9TERxC+j\nyaDAuUAu0Tk9PXVlHBkEer0eOp3OxA3UbDZRq9VQqVTc0hC5/7oO/01q0eEyWTaZgYAMJWc+J0vi\nDgYD11P3OcuyrMlk0gXxTqcTCPhcw1+v11GpVNBqtdy84C5zvk3DZFHGzL6PsmXmZ8gyx7JcKm2Z\nj74thzGmHfN9XK9fqVTcUr59s2Wf26oCoO+vWSfF99cyAVf6a9bNAODsmBUYfcbz+Ourq6utZryT\nAT9KXCsqdyvzi3Nwfo7V01i5idvTSmM/OTlx2Zuj0Siw+QrX2cu9kXkj8mfpePgz5/q2bagojur3\n+wHOMrGGDGSFrnw+H9jWkqwlZ2ttYHMOyZmjQHL+j1wZcGSNbsl5nkC36BzhPohFcqJs2Vo71ZZl\nTz2M8eHhYaBmhtwchfO3krPa8uo0zV8nk0m3hS1tiYzZe+c5pL/mqpkof+0z3jV/vVcBn5KBgM6e\nvUHeTKzalMvlAsUbZA9OZoVaawNbcXIv7HQ67YYDmUQ0LRDIhoFqefmOwmcnD26GIznL13Oqh+zk\n2l5WVyNnmSwWtgGHcl6NwoKBZBpmy3wfgIn7wWdMO87n824euNlsRtqyMl6tZMNW+utZjHO5nHs/\nR/mi/LVkTH/darUCjNk4DGvUbSPjZSrtvRXADwF4M4BHAN5hrf1N8fcPAvjH3tuet9a+/TYXuipx\n3SZvICaFcP00axXzZzn0Lys+cR0uDwCBhkO73XY3JNdyskhHlJOQLcq7vIF4kwN43hhzhi1jTGch\ne3+ywiJ7ASyXyc03fM5k6TPhOThsz2BvjHE9BLb+/d22bst5Vb18Lk+70ceNMVvB2P/fOWTLzVxk\nFU3aMnfPk0vt5L1AxtJxk69kzGFb2jJHg8KCQRxseRcYL+OvoxhH+WsyzmQySCaTAIKMyXCVdnxX\nWqaHXwTwKQC/CODXIl7zOwDejWfld7sRrwtoEUd2my+Yc7gMyrxB2GrvdDruee6glsvl3CFbjf71\n+z1F2cqUe3VzW0dmh/Kzh8OhG666q+Fb8d3+BICfjnjZUow3JRqo5Mzvn983kzIZ9CVrLu9iIJer\nPMIOOSrApUCSc7lcdiMAYZzXNQc67fvh9WBchS1MsWYMhNuytCfWxKct+wdtU/Ll4yzGUbbM3qFk\nfBe6S8ar9F1R/pqjMvJ5ny8TLsnNv8YwxmH+OpfLOcZyJO+uGS+qZSrtPQ/geQAw0VS71trz21zY\nusTWuMzIl8PrsqUOILAel467VCq5ZA15SMllITwXh49ZlOf4+Nj1TuRyoaggsClx+BPAH+KZI/AV\na8YAXF0FWeua1RNlAh0TecjY5yzzNvwVH2TGIA4848yiPCcnJ4FNeOLAOZPJIJFIcKvNrWEsRzhk\nYpZkzIAfZss8uHNluVx2jGUjXYpLNWX5VGnLXLXR7XZdnQUeMqdg09pWxlJh/prM2WuPYiztuFQq\nBfjOYswRQZ5TMs7n8xN2fFeMF9W65vC/yRjzBEAFwP8N4EestVdr+qy5JYdmwpwEh+KZvNPv951T\nKJVKODw8dA0EtupkMJDDOzJDWN5AiUTCLSPiZ+dyOTQaDTdktSVDRLFkTNFRsCfPQjoyO5ucB4NB\nIACQM2uwyx6Cz4fZ237Al5w5lZPL5XB9fY1Go7EtnGPJmEF/VjDwGcsAcHh4iOPjYxfsfXv2bdlv\n1HE4eDQaBRhns1k31yyvN8aKJWMqzF+Tb5S/5iHtmCOtYYyBZ3bsN+oSiYQL+DxXNpt1/nrbtI6A\n/zsA/jWAFwH8TYyHhX/bGPN1dsqdL798fw5n1WIg6PV6bviVN1G9XsfV1ZVbLndwcOCOVqvlggDn\ne+RNJJM4uEaUToJOhUuHDg4OAufiPKK/bCimzmIpxpsUv8uoYBDF+fDwMMCZQ4KStc+50+lMcGZr\n3+fMQOD3QP2vLQbZ+Esz3sQtIEdxpC2TLzdQmceW52EsbVnafJgthzGOaWJX7O0YCPprDrUPBgMk\nk0m3tJmMy+VygDGX18q8HPKTeRv0ExyRlXYsGfNcbOyFJfHFWSsP+NbaD4lf/7Mx5j8B+CsA34Tx\nLnuhajabE06OCRm3lZx/45pdefhOWvYMWZNZlnG8vr7G5eXlRIvRn9uXNwGTfrjhCoeRT09PcXFx\ngcePHwPARLYwexTrEgtISM26aZdlDKyfs79uVyZuTePM5CY2Ejg8e3l5OdEDDONM58HP5VKw4+Nj\nx/n8/NxlEcs6AJvI9F2Uc5wZ+7bsM5bDq8zk7/V6jrG0ZVZTnIextGXZ4D8+PkY+n8fJyQkuLi7c\n/ykZy9GBdWmTjJkkKbUOO/b9NXNrZINL2jFH4dgQZB18+ms5WuczlnYY5q/DGEu+tOV1KowxRyTm\n0dqX5VlrXzTGXAB4E6bcRMViUe7StlJJwP7N4y/JAoItSmAcBFipi0GgUChMJOYxSYRZ32yNcg04\nn2MFOH4m14zKjHKZFbpOhRnpYDDgDltzaV7GwHo5k4N0EjIYzOIsg7/kLOd3wzj7yzrtzdpczvHy\n/ZIz8zbIWTqKdTiN23KOC+N5bZliL0zaMjOwGex9W5arOJj5LbO9GejkPD4wvv+KxaJblUPG/Nx1\nN+o2ybhUKq3djqMY8zXz+msyDrPjMMYM9hx5pb+myJj+mjYvVjetTVGMZcnnaVp7wDfGPAfgFMBr\n6/6sKdcQCAT+kh3f2cphPAYBueRHBnGeP51OB+aPuL6TBxPh2GKUySW5XM5t0HBxcREYjmbwj7Pi\nwPjmOgLZtXJJVjabDUwbRXGWGbqSM49UKjXBWbIO48ykT3Ku1WoTnHn9cWUdR8Z+IJCM+T3K3p9k\nzCIuPHjuRRnncrlQW/YZy2RAZTzzOuby1/KRAT+MMc81zY5ZYZGcyZjbI0tfHuavB4MBEonE2jto\nt9Uy6/CLGLcA2Yz+G8aYrwRwdXO8D+N5occ3r/tfAHwGwAuruOBlxGDNFh2H4+j05XCdLKsr53Nm\nLZ1KpVIol8tuDkneTJxH4g1srUU2m3WJQ/1+HxcXFzg9PXVDyGylApsPBOKzvvTmMfaMAbjRk2w2\ni3w+H5hPlZwZ6NfBWQ5DkjMzfCXni4sLN1fc7XYDjDfBO2S6aCsYc7pG1lHgMG4YY8lXrspYljHv\nkzDGx8fH6PV6gd0RJWM27DalTTK+zf3qs4hiPI+/5nD+sozZaCBjNiJ9xtP8dZy1TA//azAe7rE3\nx8/cPP8vAXwvgL8D4F0AjgB8AeOb53+21m72bhdiS5xLrjKZDIbD8a53si7ycDh0y2pkZqfs6fk9\nRP/mY3lVtjRZh1lmc/MG4tBQJpPBwcEB7t+/79YM8/ruQsJJ/CtsCWNgkrP8HuV+BavmzPrbrVYr\nlHOpVHKcDw8PHedMJoPRaDTBeRONO+4Ux4/EljCmE/ZtudFoIJlMBlZicL9yn7FM2pIjAbMYL2LL\n9+7dc7Y8Go3QarXQ6XQ22svfNcbSX/Oo1+uhjFdpx8zNKRQKU/013xdnLbMO/w8BTFtw+A+Wv5z1\nSNZX5jwqbyCZackSmjIQyGElP7OTPQc+588F00HkcjnXwmcrlFsz+k6CNxh3ebqr7+tGb7HWfjLk\nJbFjDAQ5l8tlAEHO5LVKznQSzWYT+Xw+sFlLIpFAuVx2pTylo7hrzul0GoeHh5zfDeMcS8Zs1JEx\ne7HSlvn9kzFtjiNAnOuXw/+zGLfbbTdCKHuViURiwpbL5TLu37/vsvRbrRaurq42PlK3C4yj/HUU\n41Xase+vacfs7cfFXy+ivailL3t+BwcHrmhCu90OOGiW6QxzEtwGVbYSacAcTpQ3iKzSxKpQbFik\n02k3NCQDAd/LbTvXlRSzq5KOgpz9vbHXwVkmFoVxZu0GnzO3WlbO8yuZTLpGnWQsbZl8yF3WR+BQ\nsb/0blHGDB6pVApHR0duq9QwxldXV8p4AcmGe5i/lnsZyBK4q7bjRRhvix3H/wpXILnEg3O6HM6h\n0/fn+jjsJ2szc34ImBySk8OD8ne2SGVhBzkkxSUVct5K1mZXzS+fM4ftOJy6Cc6yYlfY+lzlfDvJ\nLGvJWNqydNaSMZ25v43uoowLhUIoY78Kn1ytsw1V2OIinzEQ9Ncy4PP75zK929gxgFA79jkDCDQs\nmOEvC7DNo2mvW5dP2IuA70sO03H+h1ABuBYfDZdGCzxb8yiHA8OcBm8Otjz5Prn8Bwgu1wlbTylb\npXHN7o2TpKFwOdYinNlKvy1nuU6Yr/MzyFXLaRpj2aDma6MYS6a3sWX5OmW8GknGvr+WU3NMzqOP\nnseOw/hwaojnC/PXYXkAUdceV/57GfClk2DFLRq0nAPyd12SWa/z3EDyppGBQCbu+DeQOozVaVHO\ndBrW2kDhlnkcBUWOfsMuLLhISQen7KMll8Lyu5/G2OdLe17UlsMY+2W1w2zZv3Zlu7jCGJNzFOMo\nO/Z/pjjkT86c65eNOt4HMtjL87CxEGfGOx3wo754zt/IkquyBQ/ABQRZbEO2FqcFaOnk/cDPYSo2\nAmSvU95Ics2on22qCmoaZ1mDex7OzPSW55gVDHzOcjiS7MIYRzUM4u404iTWMYhiTLZhjKNGX2Yx\nltwkY9qy3/vzbZnn3CVNWwK37Gsln2UZU3IUNerz+HfZqJTTMT5jOSLr8531HdyVdjrgS9FAE4kE\ncrkcyuUyhsPxFqYsV8gCO8CzYMFHP9GDjzKYy8/gHA8P1ncul8uualO73cbV1RWGw2FgDpI3calU\ncnNWPFSzRQ7ZbNZxzmazkZzZmmfWNTnLbOwwzjI5kwmaZCw5M3GLGdscluS8crFYRKlUCpRVVk0X\nGdCWR6MRstmsS9SbZsu3YSxt+eDgAPl8HsAzWx6NRpG23G63HWNmiKui5TNexF/7jDkXP42x5Czt\nmKV1fX9NnmGM4+qv9yrgs8XGDFCC4u5lLKoBPLuBgMmEDj/pji1+WauZSTssHuHfQEDQScgbk9dV\nLBZddqq1NpY3UNwkDTmXyzmm3OGKRTLCemayFzeNs1/elZy5O15UMLDWBnbWM8a49cZ0IGx4xLF3\ncNfyR79oy+Vy2dlcs9lEo9FwiVzA7WyZjFnpzbdlv/FOO5WFYGjLZEx7V8aTimJcKpVC7dhvuMvy\ntmF8p9mxz5icl/HXAGJpxwsFfGPMewH8IwBfBqAN4E8AvMda+xnvdT8K4DsxLujwxwC+x1r72ZVc\n8S0ke36EJLM5WfEMeJbsEeYQ/Dk6OZ/Hwg8MACy5KQMBa3ezNyLPxdYpAwH/vqkyu+12WxaP+D1j\nzB9hyxhLR8HAL5diSc5yuN0fko3iLIt7ZLPZQGnVqJEccuR56KzYM5CcN6Ft4xw2RM5lmLQXMl6V\nLfuMac9RwUBWWpOMw2yZ+yisU7vImDUUmEkvGUs7lsF9ETsOYyztmNx4yGmFu/DXi2rRHv5bAfwc\ngI/fvPcnAPyuMebLrbVtADDGvAfA92NcweklAD8O4IWb19zZOJbs+XGuh9tiykI5fC0QrL4lkziY\nsS/nbfzNGGQN9WKxiJOTE9fr4/aO7OnxvJw2YM+vXC4H1o0Dk0axavX7fWQyGX4X34Mxx9gzDkuQ\nYoPOWotkMunW7XLXND+hzp9j5yiAnJuTG6uwjjo5l0olHB8fT3Dmwc+kI5Cc5dpx//9RzmP5c+zG\nmEASns9Y2hUZkzMVZsuyDLds0JFzmC1HMQYQYMxgz/+H2jXG0+bpZwVB/3thEl4ul5tqx7f115Lz\n8fGxa9BJfy1HEMIYh/nredmu4x7wtVDAt9a+Xf5ujHk3gKcA3gzgwzdP/wCAH7PW/tbNa94F4AmA\ndwCQ2zFuVNJRyM0UgGctNw63Sccih37oDOT72RuQ87fsxctNFzhkxGQSP3mLv7PFmMvlcHh4CGut\n26Z3EzcEC0rcOInPAng3toQxEHQWcrtTa60bUi0UCs4h8JCJmZKzPIfkTNYMAj7nVCo1kzMAN3zI\nHgunl9atXeDMIC8z5ck4n89PZOGTMW3XL7UrGct91f3NU7ieX9oyRxEkb9l4J2PuoqeMp0vaJpn5\ndpzP56f6a9mLn5dxLpdbyF+TMe2YI4ib8teL6rZz+EcY12i+AgBjzBsBPATw+3yBtbZujPkogK9D\nTJyEvxOavIH8YSHgWS+Aw8TS4OVWi2dnZzg7O8Pp6SlKpVJgi03ZQpTJI7wuf404gwYw3nuaQ1l3\noFgz9nsLvqOQ0zayKpss2iHnAnl/yKpbMvDn83nH+ezsbIJzWFGQqHljBgMuIWo0Gu5a70Cx5ez3\nDn3GstBKq9Vyjpt2JouzUHJZpmTMXQ5PT08d42Kx6KYMpjGWw7yyFoDPWG15tvzGOBPqADjGs/x1\nlB1Lf03O8zIG5rfjOGrpgG/Gd+zPAviwtfbPb55+iPEN9cR7+ZObv92Z/J4fk62SyaTLnG61Wi6p\nivNEciognU4HWvqyt18sFvHo0SO87nWvw+te9zqUSqVAZa9arYZarYZqtepK6soWo2wQyOErYwwq\nlcpd3kCxZyy/R/koOScSCRQKBecsEomE+86B4HBfJpMJcJZzfcVi0TF+/etfP8G5Wq2iWq2iVquF\ncuaaYsmZgahSqbj9vu9AseYcFfRlgqzPmN8zMJmVTcbs0clOgLTlR48eBRinUilnx2GMOZcsbZlD\nxgBwdXWltjxDYY26aYwX8dfTGMtS6OQb5a+5HNS3Y2MMqtXqwo26TTUAb9PD/wCArwDw9au4kLAh\nELa2bivp+OUQLEtkyhY6Syqy9johshUnh+plFn6pVMKDBw/w8OFD3L9/390AfL/MCucwEJ+Ty/jk\n/KLUum4ILm+REs71vVghY2A9nPm9knM+nw/sY07O0mh9zn6Oh3yv7MGXy2U8fPjQceaSHbm8h0OP\ncg6Rj2w8cEoh6v9ZtTbJeROMpS3zkbZMeyZjueVpGGPasmT84MEDPHjwIGDLcv6X+Ta+LZNxOp0O\nzPcC61+XvUnGzJSXWrW/zufzjq20Y9nbDrNjsvIT8WYx9vM4pL/2fbn01/Pa8SrsOoyxHC2epaUC\nvjHm5wG8HcBbrbWviT89BmAAPECw1fgAwH+Yds5isbi2zQfYyubmOdx2sVgsBnp47O1zY5Verxe4\nCXgO3nxyn+yDgwOcnp7i5OQEx8fHzqmH1dum45K1/WVSkVyPLXd4WkcgCDPSwWDAHba+AcDXroox\nsDnO5XI5wIe9ADau5MY683DmfP3h4SFOT09xenoa4OwX5PA5y1KfPMiYAWlaUZDbapOcV81YJjn6\ntky+vi0nEgnHl7UsFrHlk5MTx5jzuLMYS1umPbPBsWu2XCqV1m7HzJYnX8mYATjMjv2AL5Nqwxgf\nHR1NMA5bxicrrvqMmZAppxaA1TfeoxhXq9W53r8wsZtg/60AvtFa+3n5N2vti8aYxwDeBuA/3rz+\nAMDXAvjni37WbSV7BdJJMBgwq1IGgmaziVar5fY2poMA4NaD8uY7OjoKHHLpHXsY1lq3haK/1pdz\nTLLSGgC3DSRbrTIpaBNDP1ytAOC7toUxMOkoZBIle3hsnfuceS4ZDCTn4+NjHB8f4/DwEAcHB+6R\nSUSdTicQCHzOXDImJdeMy2DAa1m3tpGz3Ctd8i2VSgFbTiQSjq9vyz7jUqkUsGOy9W15FuNptsxC\nPLJRp4yDmsaYh884yl9Ps+MwxmTkMyYrJg5GMbY3ydWSsYwdcdGi6/A/AOA7AHwLgKYx5sHNn2rW\nWi5C/VkAP2KM+SzGyzx+DMArAH5jJVccfl2B32UviTcHYXF4hwFfDssxKavRaLiqXfI8/g10fHzs\nentHR0eBoSdZeCPKSfC6ZAYpANdQ4B7NbDBsQo1GQ1YBa8eF8SzJ71PWQGBr3v/eF+HMnsDZ2dkE\nZ/bcuAVuFGd/ZQcQ5NztdpXzDEXZMgODH0ybzaYbwo1izAb80dHR2myZvUAyjpq2W7XixHha4LuN\nv+YwfSaTcYyjAr7vr2WdlFmMAUQylmvv2dvflB0vqkV7+N+NcZLHv/Oe/x8A/DIAWGt/0hhTAPAL\nGGeF/nsA/9De4Rr8wWCAVquFSqXi4DIZhMNwXKZBxyxvIALndojMoGersF6vB1qI0sjDKmrx5uC1\nsWXLZJBKpeKSv66urlCtVjdSqAOAPz/0gvg51oyB8XfZbDYdZzpW9uqttUin065qF+cKp3FOJpOO\nba1Ww2g0CmT7s2VPzn5yGdfkyiViHO2pVquoVCqoVCqOM8+5bm0rZ253XK1WnS0Dz/YvGI1GAcbM\n7ub/6zNm9naULZMHh+b9aZd5bZmMa7Wa2vIMTWPM/BifcSqVmpjbTqVSyOfzkf46n8/P7a9px2GM\nmdxHf12r1dwS77hp0XX4c23qbK19P4D3L3E9a1G/30ez2XTZkwz23Jdc3kB0EJlMxiV6ybk6ueMW\nbyB+hnQSvDFlQiDFG4hZpZw3YnYvnUOlUkGtVsP19fXGbqDT01M57/cWa+0nw14XN8ZA0FFwnovM\ncrncBGeutQ3jzPuAc/SyN14oFBxrmdhDxvJRjgCQM3sCPud6va6cZ4i2LFc0kHGYLZOxrGg5ry2z\nZgPrM8j3S8aL2jIDzLq1rYzZQaO/BqIZ83nW2A9jzNUVZMxGOu1Y+uuwhhjzquQIgM+YDTtpx/75\n4jC8vxe19GUgkD27XC4H4FlWKB/5N5n1K4eReHCevdfruWUidBKsDsWsbSl/yIjDfd1uF51OB1dX\nV+64vr6ObHmqgmIPn5zZqIvizOx7nzOACc4sg8ykTvbwfc5hDTvJmedhXe7Ly0tcXV254Vc2PlTh\nko13nzGdPOdZJWMOsy5iy7TnVdmyMp5PizBmI4Br56MYy1GcbreLZDIZaNSx4cfRQCn25JmDRYZk\nzEadz3jre/hx1SzjYbUpWZCB8zD9fj+w3tq/mcJaZfw8mYgnayczQUyqXq8HWn+yYIQMBJ1Oxw0P\ncYjXn0tSBUVG5Fyv1wEgsFVmr9dzc3ZkTb5cFy1Zyx4cOXPtLZn4nK21Ac5y7+7BYOA485Br9lut\nlnKeQxxavb6+BhDOmPO6PuOwJCp+17IIE22Zdh1myzzImPYcZsuswaG2HC6fSZi/noexrGER5rf9\nvRQ40iKXbFI+Y/KlHbNxGMV42nK9u9ROBPxZkptpsIUIPOsRymxN4Fn9Yw4HyWptMshzSIeJOCzm\nws9rtVpoNptoNpuusA83tKDhy2FeHsw8ZYs1aqhJFRSHzn3O/E7JmMOvlF+RC0AkH75XciYvmTEs\nOcv7RS7F473h14JQRcu3ZS4NI2Nm7bOHF2bLPmMGdh58b5gtS77MDA87D1mrLS8ufudsaIX5aybw\nScZ+9UQ/AS+M8XA43upWsprGWPoDPvqM49izp/Ym4DNxikPjbEWyut5oNHKtRVkIh1nAxWLROQEe\ncviOa6l51Go1PH36FOfn5zg/P5/Ye1vONckeJK9RZnuqk5hPdBQ0Ss6ht9ttNJtN9zeO3shKiSzW\nUywWA1X42IiQnOU8bbVaxdOnTx3raZxlL5K9QeW8mHzGwLNef6vVwunpqbNlyde3ZTLm4TOWf6tW\nq86OLy4uIhnL4CIZyyCjjGeLjBmkZdKcDMB+nXwyJmfZAZtlx7VabSZj8vMZ+w2JODPei4Avh9s4\nnM5hQQ7BsDyizOrkDXR4eOiKM0gnzXlcDgfxJuDc3csvv4yXXnoJn//85ycSuvyf5e9hr1XNlqxt\nzXr57XYb19fXbshcLuXhkGAqlZrKma19VlPjPcR5+JdffhkvvvgiPve5z0WyC+OunBeXnEflktV2\nu41Go+ESpXxbBjAXYzlHz2AgGX/uc59TW96AZGNJ+msypr+WFTBlB42MOWXLg/7at2Ppr5dhHPa3\nuGovAj4Q3C2v1+u5BCs6fGZgc3McJoPI6ky8gdj7ky1Gztnx8fLyEhcXF7i+vp5YLhKHbM1dFlvh\n0zhzSRYTu2RlPtnDl5zJV/58cXGB8/Nz5bxBSVuW281yiRZHduQQfpgtL8JYbXm6ZgW8Rb4nP9hG\n2bFkLP016ytIf+338H3Wl5eXuLy8XCtj/zu6i3tnbwI+xaFVZsryS+/1em6XI1kOlRmgnBOSyXb+\nXI48Go2GS7pTbV7zcJa7cMmdsgBMTLFEcb6+vka1WnV7c6s2J58x1ev1cH19PcGYgX9RxmrLd6co\nO+Y2w3JDHLkMk3kafjKmXE63j4wXrbT3XgD/CMCXAWgD+BMA77HWfka85oMA/rH31uettW+/5bWu\nRJwP4iPn4huNBq6urgJLdeSGGbLYgr8Ux5+749wT54q2SUw2u9HvGWP+CFvGmGKmbBRnWTGLjP36\nCVGM/SU6ynnzkrYsM6i5Ba3kKzfO8WsnzGLMnBplvHlF+evr6+uF/PUsztvKeFEt2sN/K4CfA/Dx\nm/f+BIDfNcZ8ubVWNo1+B8C7Md6YAQCCYyR3LJlJySFBWSpRDrVIw+d7woavwuZ85M23LeIw2U1L\n93sAvAtbyBiYzplM+LgMZ/Y4lPPdSfLivL4yHitOjG8zx70Ofy2vi3+fl/G8ZYLjqEUr7QVafcaY\ndwN4CuDNAD4s/tS11p7f+uo2ILb8+LMPMyzrkq1HKelcws6zLeJGMzdO4rMYO4KtZgwEdzCjJKNF\nOEu+yjleUsbPtIuMw+wYeMYoinHUueTf/GC/rdynaa5SuVN0hHFt/Svv+W8yxjwxxvyFMeYDxpiT\nW37O2uQP08sWotzdzpff6peHP7S05dp6xlQYa+XstPWcyVcZR2rrGQOTdiz9th/sp0k26CTfHWEd\nqqWT9sz4G/lZAB+21v65+NPvAPjXAF4E8DcxHvb/bWPM19mYjnf4wzr+3+bVtJbkFmsnGAOY2vJf\nBQv9Y18AACAASURBVOct105wVsZTtdOM/WnVeYbed71H78ssy9QY8y8AfDOAr7fWvjbldW8E8FcA\n3mat/YOQv381gE9w61ApmVGrWo+4LEWKCTIA/k+Mh/9uxfjmNcr5DrUJzsr4brVpxn7JYcl4VW2F\nWUF4HW2SdXT6VqUwxqziCeDNNmKDJGqpHr4x5ucBvB3AW6fdPABgrX3RGHMB4E0AQoMBABSLRVcm\nU7U5hTliscPWNwD42lUxBpTzXWmTnJXx3WiTjLkboWqzimJcrVbnev/CxG6C/bcC+EZr7efneP1z\nAE4BTL3RVPGRWIv6Xcp4d6Wcd1/KWCW1UNKeMeYDAP57AO8E0DTGPLg5cjd/LxpjftIY87XGmC8y\nxrwNwL8B8BkAL8w6vz9Usaz0PMufh9s73qi9asarWue6qu9ilefapvOsm/M2fRe7eh5lrOfxtWiW\n/ncDOADw7wB8QRzfdvP3IYC/A+A3APwlgP8NwP8D4O9Za/uzTh63L2gfz+P97QWsmHG/P/Mlc0kD\n/u3Os27O2/Rd7Op57pLxtJUP/rGu78L/nGXPs6rruevzAIuvw5/aQLDWdgD8g1tdkepOdXp6Kuf9\n3uIngSjj3ZBy3n0pY5Wv267DV6lUKpVKtQXSNEuVSqVS3ZluU+jGX0c/7Tz7sM5+luIQ8HNAcKOT\n20rPc7vzDIdD/pi79Yc9U26ez55XqzrPKs+1bedZA2e15ZidZ12MB4OBXP99K8X9PMuuw9/U/yX+\nNpuxLEd5FwfGGf9Wj1ge71TOe3GshLMyjvWhjHf/mMl46Up7q5Ix5hTjin0vAejc6cWoqByALwbw\ngrX2chUnVM6x1Eo5K+NYShnvvuZmfOcBX6VSqVQq1fqlWfoqlUqlUu2BNOCrVCqVSrUH0oCvUqlU\nKtUeSAO+SqVSqVR7IA34KpVKpVLtgWIT8I0x32eMedEY0zbGfMQY85YF3/8+Y8zIO/58jve91Rjz\nm8aYV2/e8y0hr/lRY8wXjDEtY8zvGWPetOh5jDEfDLm+3/Ze815jzMeMMXVjzBNjzK8bY7500euZ\n5zzzXM+qpYzd63aW810xvnlvbDgr45nn2Hpb3kbGsQj4xphvB/AzAN4H4KsA/BmAF4wxZwue6tMA\nHgB4eHN8wxzvKQL4FIDvxbh4gX9t7wHw/QC+C8B/BaB5c22ZRc5zo9/xru87vL+/FcDPAfhaAP8t\ngDSA3zXG5Be8npnnmfN6ViZlHNBOcr5jxkC8OCvj2dp2W94+xnddae+mDsBHAPyv4ncD4BUAP7zA\nOd4H4JO3vI4RgG/xnvsCgB8Uvx8AaAP4tgXP80EAv7bg9ZzdnOsbbnk9YedZ+HqU8eoZ7xLnuDCO\nI2dlvHrOynjx4857+MaYNIA3A/h9PmfH/+H/BeDrFjzdl9wM0fyVMeb/MMa84ZbX9kaMW1Ly2uoA\nPrrEtQHAN90M2fyFMeYDxpiTGa8/wrj1eXXL6wmc5xbXs5SU8UxtPec4M765vrvmrIwntWu2HHvG\ndx7wMW7NJAE88Z5/gvGXNa8+AuDdGJd9/G4AbwTwR8aY4i2u7SHGX/xtrw0YD8e8C8B/A+CHAXwj\ngN82JnwLp5vnfxbAh621nNta+HoizrPw9dxSyjhCO8Q5zowBteVVaFWMgR2z5W1hHIfd8lYia+0L\n4tdPG2M+BuBzAL4N4+GQO5W19kPi1/9sjPlPAP4KwDcB+IOQt3wAwFcA+PpbfnToeZa4njvXDjIG\nlHNAcWcMqC2vQnHnvKuM49DDvwAwxDgZQeoBgMfLntRaWwPwGQATGZoL6DHGc1QrvTYAsNa+iPH/\nHpZB+vMA3g7gm6y1ry17PVPOs9D1rEDKOEQ7xjnOjAG15VVoLYyB7bblbWJ85wHfWtsH8AkAb+Nz\nN8MUbwPwJ8ue1xhTwvjLmPrFzbi2FzEGI6/tAONsyqWv7eY8zwE49a/vBvq3Avj71trPL3s9086z\nyPWsQsp48vp2jXOcGd9cn9ryLbUuxjfn2Upb3jrGt836W8WB8TBOC+M5ii8D8AsALgHcW+AcPwXg\n7wH4IgD/NYDfw3iu5HTG+4oAvhLAf4lxZuT/dPP7G27+/sM31/LfAfgvAPwbAP8fgMy857n5209i\nDPqLML4BPg7g/wWQFuf4AIAKxss0HogjJ14z83pmnWfe61HGq2e8y5zvknHcOCvj3bflbWS80pvh\nljfS92K8x3IbwJ8C+JoF3/+rGC8PaQP4PIBfAfDGOd73jTfAh97xv4vXvB/j5RUtAC8AeNMi58F4\nv+LnMW7tdQD8NYB/4RtJxPuHAN7lvW7q9cw6z7zXo4xXz3jXOd8V47hxVsa7b8vbyNjcnFClUqlU\nKtUO687n8FUqlUqlUq1fGvBVKpVKpdoDacBXqVQqlWoPpAFfpVKpVKo9kAZ8lUqlUqn2QBrwVSqV\nSqXaA2nAV6lUKpVqD6QBX6VSqVSqPZAGfJVKpVKp9kAa8FUqlUql2gNpwFepVCqVag+kAV+lUqlU\nqj2QBnyVSqVSqfZAGvBVKpVKpdoDacBXqVQqlWoPpAFfpVKpVKo9kAZ8lUqlUqn2QBrwVSqVSqXa\nA2nAV6lUKpVqD6QBX6VSqVSqPZAGfJVKpVKp9kAa8FUqlUql2gNpwFepVCqVag+kAV+lUqlUqj2Q\nBnyVSqVSqfZAGvBVKpVKpdoDacBXqVQqlWoPpAFfpVKpVKo9kAZ8lUqlUqn2QBrwVSqVSqXaA2nA\nV6lUKpVqD6QBX6VSqVSqPZAGfJVKpVKp9kAa8FUqlUql2gNpwFepVCqVag+kAV+lUqlUqj2QBnyV\nSqVSqfZAGvBVKpVKpdoDacBXqVQqlWoPpAFfpVKpVKo9kAZ8lUqlUqn2QBrwVSqVSqXaA2nAV6lU\nKpVqD6QBX6VSqVSqPZAGfJVKpVKp9kAa8FUqlUql2gOtLeAbY77PGPOiMaZtjPmIMeYt6/os1d1I\nGe++lPF+SDnvh9YS8I0x3w7gZwC8D8BXAfgzAC8YY87W8XmqzUsZ776U8X5IOe+PjLV29Sc15iMA\nPmqt/YGb3w2AlwH8M2vtT3qvPQXwzQBeAtBZ+cWollEOwBcDeMFaexn2gkUY3/xdOcdPUzkr453Q\nSm1ZGcdSMxk7WWtXegBIA+gD+Bbv+V8C8Oshr38nAKtHLI93roKxco79McFZGe/csRJbVsaxPkIZ\nyyOF1esMQBLAE+/5JwD+VsjrXwKAUqmETqeDYrF46wtoNpuxO0+pVHK/Lzuqsqn/azgcotFoADds\nQrQoY3euZDIZ+C6W1aq+i1Wea9vOM4Pz0ox33Za36TxrsOWXAGUcp/PMwdhpHQF/UXUAoNPpYDgc\notVquT9ks1lks9mFT2iMQSp1+39tVedJJBKB8ywb8Nfxf3W7XXS73cDfxfWtcsiuAwCj0SjAGFiO\n86q+i1WeK87n2RDnnbflOJ9HGe/+eW7LeB0B/wLAEMAD7/kHAB5HvalYLKLVauHg4GANl6SKUpiR\nDgYD1Gq1aW9bijEw7uEr481rCc5LM1Zbvhtt0paV8d1oScZOK8/St9b2AXwCwNv43E0SyNsA/Mmq\nPy8OmtVjD5kH22rtI+N9074xXiBHSW15B7XLjKXWNaT/TwH8kjHmEwA+BuAHARQwTgTZKckbZGwn\ne6O9YbzHUsb7IeW8J1pLwLfWfuhmDeePYjw09CkA32ytPZ/2vmXmf+J0nqigv+3/V5iWZZxOp2/9\n2cDqvotVnmvXzrMs41V8dlzPk8lkVnKeOP1f++qvo7SLjKm1rMNf6AKM+WoAnzg8PFxZEtYmFfb9\nbXtPX8wJvdla+8lVnHPbOe+iVs15mxnfJpE2zlLG82ka/11ivDvEYqRZziPuN5AqWosEBuWsUsVT\nd93RvSvp5jkqlUql2hvta7AHtId/K8kbJ+om0l7e7koyV867o20e3lUtpl2ckp0m7eGvWfvcmtxl\n+VyVs0oVf/nL7vZNe9nDTyaTSCaTSCQSU39OJBJT12P6S/ISiYQ7+DyPwWCA4XDoDmstRqNR4OBz\nu7oGdNMiT8mVh//8PN85OUvWfD6K8zyMlfPyWgVjY0zAlgFMteXhcBjgrLa8Xs3LOIwjAGezPot5\nGftst5nx3gV8YwzS6TQymQwymQyy2az7WT6XzWaRSqUc8OFwGHo+gpbn5dIz6fg7nY47er0e+v0+\nBoOBe5SHf1OpFhfLUZIl2YaxT6fTjvFoNAqcg/I58wCiOXe73VDOdCbK+XaSNrcMY/KVwUCelwcD\nwDRbDrNjZXx7LWLHqVQq8H3z/X7AH41Gc9lxt9udasfbyHgvA34qlUIul0OhUHBHPp93PxeLRRQK\nBWSz2QBY6RCAYIswkUggl8shn88jl8u53h6dQaPRwPX1Na6vr9FqtQI3U6/Xc4cxBoPBAAACwUe1\nuNLpNHK5HIrFIvL5fOAI4yx7bHQSQDhnHj5nMm40Gmg2m8p5jaItk+eyjP3en88YgOMbxlgGB2W8\nes2yYx7ZbHZiFDWRSLgRORmUjTET/loGdfKlv94VxisP+MaY9wF4n/f0X1hrv2LVn7WM2LIrFAo4\nODhAuVyeOPh8LpcLGLrfM5AtxmQyiWKx6BwMAOfku90uqtWqO+r1OtrtNlqtFtrtNtrtNjqdjrs5\neX7etHFT3BkDzzjn83nHtVQqBR555PP5gLGHBQM6i2Qy6QJJsViEtXaCc6VSCeXMXiGHJZXz7XRb\nxtLefFuWjQVrreMbxrjVainjNWlexqVSCfl8PtALBxAa8EejERKJxEw7lv6afCVnTiPEnbHUunr4\nn8a4FjPHRAdr+pyFRdDHx8e4f/8+Tk5OcHR0hKOjIxweHgaGjZLJJPr9vmvNyRacdBC8gXhTsXHA\n9/X7fRhjkMlkUCwWnUMpl8sBR8IW5PX1Ner1Oq6vr9HprHLDupUqtoyBMed8Po+TkxPH+fDw0B1h\nnOko/GF96SiMMUgmky4IsGfA9xpjkM1mUSqVXCOQ94FyXq3CbHkRxnJI3+/9+bbMYNDv9wGMq7GR\ncaFQiGRcr9cdY3+Xsxgp1oyn2bGcik2lUlMb7nI+nqNDtF+y5vsTiQSy2SzK5TJSqVTAjmVvX9px\nvV6PM2MA6wv4AztH+c27kDEG+Xwex8fHePToER4+fIiTkxN3yNb+cDhEr9dz8zxswclgLx/pOHq9\nHobDYWB0ABg7CQ4lydyAfr8fcBLn5+ew1robC4hlFnhsGQPPHAU5P3jwACcnJzg+Pp7gTGZyJEfO\n6frzgrMadtls1n2+HGJUzquVb8vLMKaiGNOWoxhns1lnx3wdGwedTgdPnz6FtRbtdjvOwSC2jMPs\nmHyPj48nGMvpNX+I3Z9rTyaTAOCmeuT8PP10MpmcsGM/6EvG7ATEVesK+F9ijHkV4/15/xTAe621\nL6/psxaSbDE+evQIb3jDG3Dv3j2cnZ3h3r176Pf7gaE5GnA6nXbz+LOyNpmUJW8+9vBZF1m2PAeD\ngXMo7XYbANBut1GtVt0NLTNQY6LYMgbG3y97fw8fPsQb3vAGx/js7Az9fj+QfCVHcmTmdVQGtj/n\nJzkzwUj2II0xgc+I4hxDxZazH/Cfe+65uRn7LGfZMt/LoWLflvkoP6PT6bhAUKlU7ux7mkOxZizt\n+LnnnnN8T09PMRgMJpJkeYR10OaxY9nDZw6HTOAdDAau4d5utycYx9iW1xLwPwLg3QD+EsAjAO8H\n8EfGmL9trW2u4fMWVjKZdJmfTPzgfFC73XZA2WL0bwY6AWByXae8sWTyiJ8FLDNNpWPpdruo1+vI\n5/NIpVKBfIEYKfaMOfTOhJ98Pu/m68h5OByGcpbJXVGc6TD8JCH5ukQi4TjLLHHJuVAoTHD2Wd9h\nIZBYc5aMacvTGMvg7S+fBIKMZWDwl1lS8zCu1WoBW5af4f8vd6StYUw75tx7uVxGp9Nxve5pjP0R\nHb9BH2XHwLiTKDP65UgCGUs7jiFjp5UHfGvtC+LXTxtjPgbgcwC+DcAHo97XbDYnvhD2lFZ8fej1\nemi1WqjX66hUKkin00ilUkilUmi3227OrdlsukQN2ePnwYZD2Pp9+b8kk8lAwoe11jkmzgMyOUU6\nEJ5nnTcQW6r+dzTjO1yKMXA3nGu1Gq6urhxjcmYmbqPRQKfTcSM7cr42ijMP2bsjZx6SM3M3yJkj\nAel02p2TwUQ2DFelTXLeBGPaxSzGzLQmY3KWdtzv90PXenMuX9rgKhjL4KOMp8tnXCqVJvw1bZgr\nJqJ6/FHr+Tm0T/mMAbgETjIGxqM8HAnwGfsdvFVpGcZSa1+WZ62tGWM+A+BN015XLBY3svsSh3B4\nA7FlxoMNAbmErtlsuuAvDw75+GtEmUAib65er+ec0mAwwPHxMQaDAZLJJHK5XKAFKd/P4WC/hbqq\nmyjMSMXuS/N+p3MxBjbPudlsOs7SUbRaLceYnGW2tZzW4XckOfucZnHmUi/ZG5SOQvLcds6bYgxg\nwpajGDcaDWfLfkY9bTmML88ngwR7dfMw5s+y+A+gjBe4nql2zIY7O2hk7Ntwt9sN1Fih7Ul/LTts\ntONKpeIYHx8fO8bSX/N39vDDRmXjwnjtxIwxJYxvnl9e92fNI7/FKGElk0kX8Hmw9cgbSrYmZcue\na0S5HpiGz/PyBnr8+DG63S6Gw6EL9hyy4k0c1fNjdmkchoak4sYYmM5ZNuzYuCNXspXOQy7f8dcC\nS87MEo7inEqlAkFkWg8/boyB+HH2G+9hjOVKiDDGPPwaHJIxnazfqFuEsR8M5CoBIB7DvUA8Gc+y\nY9lw9xnLhnwYYz6yoRfFWDbo5rFjThXEjfE61uH/FIB/i/Gw0OsB/BMAfQC/uurPWka8gRqNBq6u\nrgAAnU4HtVoN5+fn6HQ6AUfgr7+UPUDO4fD3crkcWLrFYT325pgrwEbAYDBAu912v/OGy2QyKJfL\nODs7w2g0ctch13ne5c0Td8YA3LraMM4XFxdot9uRjGXvgMO/5NxqtVAulzEcDic40/jDOLdaLfe7\nnLopl8u4d+8eRqNRYDRJOc8WGV9fX7tej2QsbdkfvfEPZlwzMEjGnMOVjJkvwABBxvzdZ+zbMhnf\ndRDYFsZRdkzGfs+e07Cyl8/ffcbAeJ6ejTIG9DDGzWbTjQQwyGcyGRwcHDg7lvEjblpHD/85AL8C\n4BTAOYAPA/i71trLNXzWwpJOgsNzlUrFVdVidq1fHU1m38ps7m636xzJcDh0ziGTybilHyz0c3h4\n6BwDHVSr1XI3GwN+Op3G4eEhHj16hGQyiaurK5cJKocE79BZxJox8KxhN42zXwWPjOUjk3+63a7r\nCfqcWYchk8k4zlztkclkAGAiGIRxvry8dIFfOc/WPLYs18X78/b+I1fjsHFN589ELTJn0S4yzmaz\nsNYGbHkaY742Jom4sWc8zY65EkMy9u1XHp1OB41GA4VCwS2/ox3ncrkAY/pr2nEU40wm4xinUilc\nXFw4fx236nvrSNr7jlWfc5WSTmIwGKBerwcSr/z1mP5SDlm8odfrucSPVqvlsu85VCSdRLFYdDdY\nq9VyGaTyfTIP4PDwMFAcIk5Le+LOGAj2DCRnuRQyqh62f8jkHiYrSc7M1paOAkAoZ472cEhQcmaw\nv+teHxVnzhwyXRVjVrrksl2y4rA+R+4kY9oyM8N5b0xjTFtWxvNpETuettGNtGMyBuAa6WQs7Zj1\nU8h4OBwG7F/666OjI/fe0WiEdrvtRiTipL2rpc8WI50F5wHDiq6wdyUPIHxTFbYCOQzEG4ZDRqz5\nzZ85BMUSjVzjOxqNkE6ncXBw4Ip7cDgrLk5iG0TO3AjDX287i7OfK8HX0cjDOFtrF+LM4V5yvr6+\nxuXlpXKeU7dlDETbMpOxGAx8xiyfnU6n0Ww23XpwVt7j633GjUZDGS+gddlxq9VyDbpCoeBYs9aK\nZJxKpdy0gc/Y99fGGGfHiUQictO1u9LeBXwAbv00f5aV8sIM0c+25GvljcZeHHvhnBrgXHHY8BIA\nN5QvM0h5Plnre9pNrAoXe2Vyc4tpc+NhQ6zzcmZgl9MAshSrz5kbdsi1/uTqc1dFa1nGcqrEZ8xp\ntzBbbrVagRFAWbc9jDEAx1gy9ZfuqqK1ajvmOvppdizZ8pENATKWG2ix8mbc/fVeBny/VTjtBpIO\ngkO3Ya+Rw7a+g/CTN/iZQLiT4LSCdEh0Eqr5RS5ylUPUnHgYZyA6GBhjAgmbPme+T47wyEYdOfuB\nIE7OYRu0LGOKrwtr1AHPAoFsvPuMZjHu9/vaeL+F1mXHtFeZkCuT7chG2jHzMsL8dRjjuGkvA74M\n9PwdiM6I5uv5c9gQEQMBE0xk0Z5msxlYtymLPjD5Q64PZQKR37iI840UR/nc5CMQ3ribxpmOJowz\nl2xKzlzbKyt1sWfgc+b1xN1hxE23YSxH6/g3zveSsb+8q9FoTNTbIKsoxsz1kdekfOfXqu1YNuqk\nHZMx/bVcekdfPI+/jrMd713A91uAshctjVLeUFHzQ/60gBze80cDmPzDm0b2Ag4PD92WvIVCwZWL\nZCDhWl9ep2q2fM7AdEfrO4hpnOUmKtbaQIEkcmaQl46BOzKWSiUUCgW3skMuI1LO8+s2jP3XzmPL\n9A9+oibzOsIYkymDibRlXoMqWrdlLIOwXNYcZse+vy4UCoHVOPTXZEx/TTtmY4Hr9nmtcdLeBXxg\nsgUoe9z+PBuA0JvLWuta72GjBTKrn/M/cunWwcGBO7in88HBAXK5HCqVCjqdDi4vL3F+fo56ve4S\nVlTzy/++yNQvmxrGWd4fnJeXnPk3WYe73+9PLO2RjOWjz/np06fKeQlJWwaiGUcFCb43zJbla/y9\nMaQtS/uVj/l8HtVq1WVsK+PlFMXYZz0PY/6+rB1LxgcHB8jn86hUKmi3286Oa7WaSwiPm/Yy4FOy\nN8dlV2F18fkIBLdYZKlMPxNTThnIpCy5Jv/o6AhnZ2c4OztDqVRyG0Pws3gDMeDH9QaKu/zRGZ9z\nWF1tObQrOZMjzwsgwDisYXd8fOw4s0ofqyuGcdZgsJwWYcyAIOvay7nasEYf/+bbcj6fD9iyrMTI\nz9TG+2p0G8ZM/GPCnjwnXxdmx/TXtON79+5NMDbGbI0dLxzwjTFvBfBDAN6M8e5K77DW/qb3mh8F\n8J0AjgD8MYDvsdZ+9vaXe3vJxA3/BpI19cMO4NneybwpmHQlzw8Ee/hhvYLj42M8ePAAr3vd61Ao\nFNwGLQwwnU4HV1dXOD8/d9WiNnUDCYN43hhzhi1j7CuKsyzRyYMBn4zJjqU15Tn56I/k+A07cuau\naalUyq3CkI6CQ76b4MyStDf6uDFm6xjPY8uSMYMC18PLHRH5e1hirF+HY5otS8a05SjG6x7u3XXG\nZClzZnzGcse8Re1YBnza8XPPPRdpxxcXF3j69KmbntuJgA+gCOBTAH4RwK/5fzTGvAfA9wN4F4CX\nAPw4gBeMMV9ure0tf6nza9YXLbMvZfUzOVcjDy6/8DdKqdfrqNVqqNVqaDQagSpPmUzGtQS5BpcF\nPliu1e89cimXrPXNsp+bnNsV399PAPjpkO/vzhnPI35XYc7BZyxZT+PcbDZncuaaYZ8zgMBWyz5n\nud/6ujnz3r/5vAmD2SbG0pbnYZzNZieCwzKMmXvBdeJkNs2W+dpNaFsYL+Ovw5LnpA37jOexY1lX\nX9oxGcva+LRhn7Gfp7H1c/jW2ucBPA8AJvy/+QEAP2at/a2b17wLwBMA7wDwoeUvdTWSlywDAW8g\nWWyDNdHlxily84XLy0tcXFzg4uIC1Wo1AJ4bLcghXGbyX19fB4xfVouS27by5pFFPzahdDrNH/8Q\nwNYxlvJ7A9JRcDMNMiVryZicyXoW50Qi4Tg2Go0AZ7lHNznLgh5yRGHd4taeN1ttbiVjmXALIBAI\nmClPO/Y3QAqz5XkZS1umnfqMWYhHMpa2vAntEmMgvOHuMw7z1zyurq4W8te0U9oxC+1IxnHw14to\npXP4xpg3AngI4Pf5nLW2boz5KICvQ4xuIh5hToLOv1wu4/Dw0B3Hx8c4OTlxj48fP8bjx4/x2muv\n4enTp6hWq6hUKqhUKrDWBrZfBJ5NB6TT6YkbiPWg5c58HMr3t8a9S20LY2Bxzsy+5VaYZP3kyRO8\n9tprePz48UzOzAZmJa9Op+N6BwwCkjOdCYcA41B7exsZ+8P4DAabYOw36iRj7rgpbVkZL6YwO5Y1\nDyTjo6OjmYzpr8mXjOVyS2OMa9RlMhlnx2yscR+OKH8dB8ZhWnXS3kOMh46eeM8/uflbbOT3/Hjz\nsLVYLBbdDXR6eorT01OcnZ3h/v/f3tnHSJaV5/05013V9dkf1TPTC8JysAnClmKSAFkRWLPJRoq9\nUsD+h8iOhDdRhIBYSvJHQEgoIBMJhQSJCLwWihLkKE4ky3aMFcEuxMZgQrCDseMPkhDYYXdmd2em\nv6q6qqv6q/rkj6rn9HNP3VtfXdV1q/o80lXPdFfdvl2/8573nPe85z23b+PWrVu4ffu2y8TV87MB\nuBGg7uP01wH9i/u4q9WqG31yUJAyzQ1jYHjOGxsbPZzJmtm4eoY2EM9ZE/40D2DOOM8NY83Y1qIo\nOvPzGTPBjnyHYaz1Fca15ZOT5Ch53I6SKWtuGANRO9al17j+mnyTGNNeGcnkjFwZa+K1X6//7OzM\nHcHLgaFGbalhliquWqnJ0mcFKxXXZKYhTQLR7TWFQsGt/xhzcRAGGwrX4avVKur1eiSMw6MaDw4O\nHGx/1rG8vIzV1VWsra0hk8m4sND29rZL+nj55ZdRq9WuxAnwlCn/s5mW0siZs0OG+VneuNVqufwM\nXWPvx1k7E+XM9nEdOF8lY92y5TNmeFYZc8mOjHnU6iDGhUKhJxHQZ0xbZtg4MJ6c1I55GBlrHShj\nv7+OY+yfgEnGxWKxx4559K3asfbX29vbjvEozn5cXZbxpB3+fXTWirYQHTluAfijfm/kh32Vd5zV\nogAAIABJREFU8h3B2tpapIKW7wjYgHjsIR0+kz9YKKder7uGmc/nXQP1w4vsJOr1OnZ2dnD37l28\n+OKLbi/nVXQScUZ6dnaGWq2W9JaxGQOz5ax7alksRTsKOgPtKKy1kdE7q2olceY6YqlU6nEG9Xod\n29vbuHfvHu7du+c495v5TUojcp4rxurw9TQ7vxqe5m74jNXh+4N39gPchqVrxWSczWbdzI+MU27L\nc8UYGL2/1oE7+2t/UNdsNnsY8+CccrmMUqnklnTp8JPsWPf5T0tj9NcRTZSYtfaOMeY+gCcA/AkA\nGGNWATwK4Bcn+bsuIwXiF8PRcqjagHSGz+zcg4MDl0XPmtvsJNhorO0cuFAsFlGpVFCpVFzhlWw2\n67K12YCee+45Fy5KWZgXwPwwBnrLb/qcdd+uOgMaFGuo+zMD7SiSOG9ubvZwZkdx9+5dPPfcc24N\nP22cF4mxFmdRW9YZPpPrkhw+IwPn5+eON5cHyJiDukajgZ2dndTb8iIx1v46aeAeZ8c6cKezj7Pj\ntbU1t9RzenqKg4ODHjuu1+uucFOaNc4+/CKA1+Ai6/OHjDGvB7Bnrb0L4JMAPmSM+S46Wz0+CuAe\ngM9N5IljNOqHrHtzV1dXsb6+jkqlgps3b0aKbvjrPCzDyDVDHsBQrVaxt7fnOg1d88vn85Hsb3/g\nwP32mvF/cnLSU4LzKiWf52u7X2fOeMBz9ogzOnJmwhY5615cvg644KydCJN3qtUqdnd3HWdd98vn\n825GwC1cSZwZDmR4bhacue9cNPeM42yZoWC1ZW6N1GptPmMO5n1bTmLMRL002XJaGV+mv46zY94v\nrr8eZMfcHcPwvc/Yj+zSjhnO397enqkdj6pxZvhvBPBldBI+LIBPdL//ywD+gbX248aYAoDPoFPM\n4fcA/GRa9u4aY1xmZ6FQQKVSwdbWFra2tnD79m0AF1WXOCvgqK7dbke2eWxvb+Pw8BAPHjzAiy++\n6MKzHC0yHLS5uRlpOFzz50iTGaM6YOCZ6bOQNNxfwRwyBvpz3traAnBRUIUzA3I+Pz+PbNnyOVer\nVZycnGBpaclxZnJnoVBwST/sTMj64cOHbmDYbDYjVb2uWpyNdrUwjG/fvh2xZSbSGWOcLddqNWfL\nfO/Dhw9xeHiI+/fv4969e6hWqzg+PsbS0hJyuZyz5UqlgmKxGMs4bba8qIx9O+ZJetpf1+t1nJ+f\no1gsuvez8NGDBw9w7949119zGYiJnayYqIz1WHPa8cHBwczteFSNsw//KwD6ntNqrf0IgI+M90jT\nFfdbrq6uOrjagKy1kf3S6vC5dSOTyaBcLuPGjRtoNpt4+PAh7t275wxfHT5Hoxyh0uHzMI1ms4nt\n7W3s7++7BqSVpWYhWZt7k7X2W3GvSTNjYDjO2hkvLy+7dXYALoS3uroKY0yko6DhkzOzg8mZHYXP\nmc5AOc9qy2Umk8Ha2hrX/mI5zytj7rCgLXP7a5wtLy8vO1vmoI6MOaijw+egTivpKWMmc6XFlq8L\n4zg7JmPasc/4xRdfdIx1zZ6/I44x6ynE2XFattQOUmqy9K9KHDGWy+VI46EjaLfbbk1eL56RzHWj\ncrkMY0ykATGBREP5nBXoNp7j42PU63VX9Wl7ezuyJOAf/BE0uuI469VutyOVtviVpUiZpBPnDPTo\nTCZw0RnoVq1+nJvNZuB8SQ1izGp3tGcypi0rY7VlMvaX5ZJsudFoDGXLQaMriTH7bNoxL066ms0m\njDFYX193mfaD7JhLQkl2XK1WB9px2nWtHD6LN7BUJo+l1QppbDDsKLg2x4IKrJ/carXw0ksvoVar\nuTUgrealWZ0ME3Mk2Gq13DoSr3q9juPjYzdK5NrUojmCq/h7xuGs1/n5OTKZjCuuopx5T7Jmh+Jz\n1u2bypkHa5Dz0tLS1D+PRZW/86JcLruyqMpXnf4gxu122yWF8VpfX3dJeZw5knMc48va8qLZ/GWU\nZMeshKclbpUzE+hYN6HVauHll192oX5dJigWi9jY2HA7LfgateNarRapxsj1f9+OfXZJkZ1ZMb42\nDl/LcMZ1EsDFEZnq5HlpNa3Dw0Ps7e25dft2u42VlRU302PDzGQyODk5wf7+fgSwOvwHDx6gVqu5\nTmKWofxF0CDOxhjXSaiTJ2Oy9znTmWt4dxjO7CjiOAPT3cJzHZRky+oAfFvWqmmNRgN7e3tuTbfd\nbiOXy7mtszwGNZvNxjJWhx9sefLyt1pySy3tOG7A7jNuNps9dqxLcbwvB+7VatX9bqDXjlmDZR7t\n+Fo4fDVQf1tHuVxGJpOBMSa2g+AWvFarBaBTcIKhvgcPHrgGpA7/5s2bbhR6cnKCarUaCfvQ4e/s\n7ODBgwc4PDx0v3feGlCalMR5bW0twtnvKLjNrtVqodVqwVqLw8ND7O7uIpvNuo6cDp9rfbdu3XIz\nSuXMixnB5NxoNCKcA+PLiYzz+TxWV1dRKpVclrY6fLVlXqyhnsS4VCq59dyVlZUeW+Y1aVsOs/te\n9bNjPxqrnLnmvr+/j2w2i/v377sIDdsMKy/Sjunwdctus9mMOHzaMaMIo9rxLBkvhMMf5QPUutta\nlIOZnlzr01KKDO+wrCKdP9d7We1Jy7Kyg2DEQM9pZmM8ODjA/v4+jo6O3O/hiUzzsB40a/Xjzs9Q\n1+niOJN1XLnUo6Mjl7BnjHFbsXgu9u3bt52zZ4ejzkD3cu/v77utfLo7IDj9/hqGsVZG8xmrPZOr\nlj3m8k8S41u3bmFpaSlSa8N3+MGWR9M4/fUgO/b5MhzPqJ0xxrFjhb4kOz4+PnZlfOPsuNlsujY0\nb3a8EA5/WHFfLqviseACjXdpaSmyJY+lV3O5XOSMZHYSlUrFZeien59HDm7g+p5/MIpmdKpz8c/a\nDjPA8UXOepqVX6CDDJiFTc7lchkAIhEZ5WytjRzCwpyMcThrZxFmduOJRXLiGPOz9xnzCGJ9rW/L\nyliPtdbDrHQGP4wtB8bjKcmOaefKmEmW3FrsD87Ujsl3fX09YseM8unFtqSTQL3mxY6vlcOntAFx\nhqDnJushDUD0RC7+TGfwLI3Kgxzy+bwrypPJZNxxmhyJAvGdhI5Mg7O/vHRgp5zZaZAjq3MpZ20P\nmuFtrXWMybnRaLjiIBo10M7C7yj8gUHQ6PIHdcoszpbJGEDP6zTxC4Djm8/ncXp6mmjLSYO6YMuT\nkT9Bi2PMi6fc6WBd+3aN+KgdFwoFx5h9gyb2+n31PNvxOJX2HgPwzwC8AcArAPyUtfa35OefBfBz\n3tuesdY+eZkHnZT8BsSCHOywWauYpVa5Xs/wv14KHECkcR0dHSGTybhwHg884MifSwR+GGrWe/AB\nuEEJgGeMMTcxZ4yBeM566Ak589QtDRlqZS89AY9MyHh5eRmtVss5AoYNgYviRXEdRRo4c+tSV980\nxswVYzpuljpuNBqug+cVx1htWTknMVZbZiY2bZl2Hzd4D4wvL5+xDtyVMctXk7Vuj46zY7+/ph3T\n2QMXx9/2G9SlgfGoGmeGXwTwxwD+HYDfSHjNFwA8hYvyu8cJr3O6ylAI924SMrfP6alm/H4ul4tc\nTODhpc/OGaJeOluMO499dXXVzRo4C+TMZFaSxvsxAP864WUjM75KcUDlc+b2Kq2XTh48UpX/9hkr\nZ/2eH/1Rzvl83nFmURAex0nOswgDWmtdu0OnClucxmI8S1tWR07GPLfet2XgwrHHseXXuOget3vF\nMdY1ZbXlq+Y8TcZXKdZT4MBLOethVmTBK8mOqbj+2mettRhYnEejPIPsOG0h/nEq7T0D4BkAMMl/\nzbG1dvsyDzYtafKd1rlmZ6yjOACR/bg8c7lUKrl1fm0oKib58J6MJPCeq6urqFQqOD4+Rj6fd2uE\nTBjhNQtxKQPAV3DREfhKNWMAbp8uOfuzbS27qoyVsyZeKWf+Dm7jo/EDcOFFbv0hZ+7d5Xv8e16l\nstksbty4wdnq3DEGkm2Z9qy8lTEL6SjjfsVTuMVLbXlpackdtqS2rIz9BL+r1rwz1iUxOnY6YD9f\nglzUhnlqZblcduvrSfZGxr4d6xkKlUolcghPGux4VE1rDf9xY8wDAPsAfgfAh6y1e1P6XUPLb0Da\nSZydnbnQDhM7Tk9PXaNhpS2+XrM4NYGEl1Z+0uIMN27ccJ0E77WysoJGoxGZacxBiCiVjCldO2dH\nwcNuWDiJ5TLPzs4inQM5MxlIWfuctYqbdhQ3btxwW3+Oj49xdnbmOHOJYQ5Cgall7OdHxNkyZ/es\nd65OYG1tDScnJ46x2rImdcbZMgfvPCp3dXXVhX9XVlbcWrM+a4qVWsbAxaDO7685y/cZcxBHO65U\nKi7SoXzj7Dhu4K79NRnncjl30uWc2LHTNBz+FwD8OoA7AH4YnbDw540xb7Z9PpWr/MB0VsCZvdZg\n3tvbc9uoyuWyK77RbDbdbCGbzfYkB+mavpZs9QcIxhisrq66ezFMxSQkHbWmtCGNxRi4Os5JHQU7\nC3Le29tDq9VyjNfW1hznGzduJHLmxbBxHGcAEc5cO+az6Qw0hZxTzxiAc/icATLESlve39/vseVy\nueyy8bkbZxBjDtr4njhb1vYyJ7Y8NuOrkkZxjo+PXaVT9tm1Wg37+/uoVqsRO15dXUWr1ephTHbG\nmAhj9hFcjvEjM+Vy2W3jJGMAPVHhlHxsiZq4w7fW/qr898+NMX8K4HsAHkfnlL1YHR4e9oREmHRz\nWem6DPfsakKPrtl0/wY3IGi1Wm7fJ79Xr9ext7cXGTX6GaO8Dy8m8PGENaCTCVypVLCzs4P79+8D\nuDjWUbPCpymObFWDfue4jIGr4cxBlM9ZefkRACY3cZDA8Ozu7m5PNEc560wzjjPDjJubm9je3u7h\nTMacVUxLo3JOO2O1ZV4snqIdtTLWTlpteXd3tyeS49uyDubZtrgVbGNjI2LLWrkzzbacZsb+/ntN\nyotjzIG3DraY7HdwcICdnZ3IIM3nrIl5zM3Qn1cqFWfHOzs7LgfE38GTNjv2NfVtedbaO8aYHQCv\nQZ9GVCwW9ZS2iUrB+R0Et3LoxREbt+iww2i1Wq4UZz6fj4SHNGOU99UMUT9LnM4AAEqlEgC4TFTd\nR87lgGkpzkjPzs54wtZQGpYxkB7OQDTSAyDi/OkICoVCZH2Xnbzu5ujHuVAo9HDmgSvct38VM8DL\ncp4XxkA0wVIdPnCR8a22TMZJtqyM6SistW5GT8bGGJRKJVhr59KW08KYDtmfnKkd83UaySNjTdak\ns1c7VsbkrDZMxtyZ0Y/xvNnx1B2+MeZVADYBvDzt39XnGSINSDuIlZUVB4hftZNQJ8B7aKNg48tk\nMpH1fk0QyufzLhGOMz9erNKnJTrZeGaZqT+K0sC4+xw9WfLaUQDR0bAfruUsUDnTiWtH4XPW+gs+\nZ78d8CAOcmb4kM/vP2NalEbG3HqljONsmTMwVk9cWlpyRVySGHMtWBkz81sZ53K5SGIvbVkZz4st\np5GxzvB5+WvmdPhcWouzYTLm/ePs2O+vufavdsxKjL4dc4KYZhsGxtuHX0RnBMh4zg8ZY14PYK97\nfRiddaH73df9SwDfAfDsJB54HDH8x20augajYRzb3bvtl9XlGixHlHFiJ8F1Qm1MNHiOKq3tnNPM\nKk8nJyfuFCaGkNk5zULyN762+zX1jIFkzn64jolYPuukbFtlnsSZSV1xnHmW9+npKXZ2drC5uek4\nc8Z51YoJPy4cYy2CNIixioy53k/GTPrT5QS1ZWW8vb2NjY2NmdryvDJmdvzKygry+Xxif821d+V8\nWTsulUqJdswkQGW8s7MTYXx6enqln9WoGmeG/0Z0wj22e32i+/1fBvA+AD8G4F0A1gG8hE7j+efW\n2pl9EhyJc7tGNptFu905DY21r+ncua2GTp4jRT+DXtdttYNhnX3NAeCBGprpyzCRtdad13zr1i1X\n0avd7pzypIdwXJWkk/gVzAljYDKcORPQmWKcI/E5M4SonMm2WCxGON++fdtxPjs7Q7PZvHKHwAqB\nXS0c43a7jYODA3d+QRxjIBoJ0AQ7Fq1hFIZLevw9WrGNR62ScblcjjCelS0vCmPWPKCdKOOrsmMe\njpZkx2TMuv38nWnTOPvwvwKgX3zqJ8Z/nOmIIzWO4AC4ToJGztEh91hqA2JYyd/KwZkDG5OuBdPZ\ncyaio1BjjDvf2e8kOCDg0ayz+ry6epO19lsxL0kdY+BynDVE7G/L4nuSOLOATz6fd52EcmZHrwM7\nZvVzHZm6qk4ik8lgbW2Na39xnOeaMUtbcytVEmPfljUCRKegzj6Xy8UyXl9fdwl9yniWtrwIjHm2\nhTLmZ6+Mp23HPAnVd/hkzCN4+0WB06BrUUtfR4zcrqFn3GttbC2pyPcyvORvvSNc3b7jdy68NOS0\nvLzsQvnagPjeZrOJ3d1d53jT3ojSonE568Aul8v1bMvi58+EzkGcOfvIZDLY2NhI5MzjWaeV/LSI\nGsSYbOio/dmfMlZnoAlgXI89PT11SZi6lqzLBMvLy9jY2OgZ1AVbHl9kXCqV3P53cvZD+CxwNG07\nrlQqrm5DEmMpWJZaXYuehhnWDM1rOIeNRR2yFlzRusyawU+j9cP8lIYK2+02CoWCczoaktKiPCzx\nyuzQeUj0SZPiODPSwvU1Pz9DM3J5cRsm0DvjjuNMx8F1PJ+z1u/m7yJnTRgLjmCw+jFutVqOLzts\n7sOOY0wm/laqOMYa8udhK35bSmIcbHk0+Uup7K8ZMtf+mlGZq7ZjnQjqWRpp17Vw+L7a7bYL07GT\nIFTgIouTTpgdMwA3QqQ0LKjibIEdFBsK/0/560r+vYIjGF+cqfmcOTIfxNnfRhXHWR0BnQjbiO7i\n8Pn6jP02FTSclDELJpEzcOF8kxjzaxIbIJmx7gXXQUE/zoHt6NIcijjG/eyYNqw84rbOkR37Bp+x\nH83176G7PNLM+No5fDpiv5OgQeshK3rpaUuc4Q/qJDjC9B2BNgzdu5nmhjKPiuPMzzuJM2cUfuRm\n0MCO0lmeX8Kz36AuaHjp5+Uz1qpnDNPquq7P2O/E45yBz1gdjTLWWWCw5fHl2wMZs4SuRkjjbFgZ\n64BsUH8dZ8c+436TM33+tPJfSIc/6MPmSE5LruoIXjsJ3XdPJ87f0Q+6vkadvV+9S0er2lHowGDY\nv+u6Kulz0XoKR0dHkSzscTknhQZ9zv7sTwcbcTMDXVPu9zddd+leaiDKOM6WAcQy9jv3URjz9/tV\nNWnLvtMPtnw5cb1dczKUsUZwxmWsDl3X+pMY63Ks3oPR3LQ6/YV0+HHSjnVlZQXlchntdhvZbNZ1\nFjz5CLjoSPiVIHXtyK+epRXZ/Bkk6zuXy2VXWrfVamFvb89lebZaLZcolM1mUSqV3Pd4BfWXOttc\nLuc4r6ysjM2Z68EaodGqbJrUtba2Fst5d3c3srVI63IrZyYlBfWX2jL3To9jy5oAps48iTFtmXu4\nWTGTttxutyO5QcGWx1NSfx1nx3Suk2Ssdswqe/0Ysx0yeVS3baZJ18bhAxej8lwu5wo08AQzNhC/\nkwB6E/D8JA6d1WnZTy3dqEU8fIdvbbRICDuJYrHoMlQBuC2AQcmiIS8tLbmOwudM9sBkOLMEK0/H\nYzGPQqEAIJ6ztr9isejyDNhxBc694szMdwalUsnZDGu8j8pYoy86s0uy5X6Ddx0sBlseXppTAVwM\n6jhwj7NjjZZOgjG3UeshPP7APc6OyZhLS3xN2jSSwzfGfBDATwN4HYAWgK8D+IC19jve634BwD9E\np6DDfwfwXmvtdyfyxGPKdwRsPNxff35+juPj40ioEEBsY9Fwnd84GVLiWdn5fN5tIdJZAQA3SuR9\nGDLSToLbfTjzn3Yn0Wq19HCGLxljvoo5YUzpwI6cV1ZWnCPQ4hjjcNbiHuTMK8kZKGcAPc5AOV+F\n5o1zXLsnY36OtGUm5fphfw0F+4yVsx7O49uyz1gHdaypzvv4g7pgy8ny7az7XJH+mgNrTs5G7a+H\nYczyuf4MH4jaMe/D/pqM2QauivGoGnWG/xiATwH4Zve9HwPwRWPMj1hrWwBgjPkAgJ9Hp4LT9wH8\nCwDPdl9zMqkHH0d0+kzCy+VyWFpacmuALJWpjUhHiTrD03U5dTC81AkUi0VsbGy40SKP4uWlz8eG\ny9kpX+M7DL5+0jo9PUU2m+VM5L3ocEw9Y3XI/KrH0S4tLbl1QJ6M53PWhEwy5et46cEqyplVwSqV\nSoQzE4400Yuc2VGUy+XIvmL9e/Q5J6l55ByX+8D1W7VlJnjxdUCyLY/KuFAoxNpy3LIP/x1seTgl\nJcAx4ZJ2rFn7/uetnClumRvUX9OG2V+vra0l9tdJjGnHjORMm/GoGsnhW2uf1P8bY54C8BDAGwB8\nrfvtfwzgo9ba/9p9zbsAPADwUwD0OMapKWlkxVEeGwBHinowRlwGJjsBnd1x3ycvVsvj6J8zglwu\n5+pB80xmf0+njkTVEayurrqRbNxxlNMQC0p0G+x3ATyFFDJOknLnCJ4JcQy35fP5ntm7hurJm6yV\ns+ZjkLMeuhHH2Z/1aXJQLpfD2tqa49xoNALnAVL7JCN+Zgyb5/P5yB5sZczXk3GcLfuM9cArRhM4\noNM8Hp1R6uBdGQdb7i+/71Y+tGNe/o4ptds4O2Yb8BnzUByyph0rY7VjZcz+ehZ2PKouu4a/jk6N\n5j0AMMa8GsAjAH6bL7DWHhhjfh/Am5GCTgK42JvLvZp0Avl8PlJh6fz83HUIdPZaiEc7jFwuh5s3\nb2JzcxM3b950J2dx2SCuuA+fyV9rAuCObLTWotFozLKKU6oZ+52Db/xMwgE6nAuFAprNZk/BFHYm\nAFwYkax1rY+ceRWLRTdLWFlZiSQJaWcE9K4pAnCDA+U8o44itZyVsR/61WIrxhhny7lcLrYoTpwt\n+4zz+bzju7m5iVKpFDmilbYcx5iDOl4c1NGWDw8P3cmNM1BqGftSO6aDJmM6e7Vj7a/Jwj8dlQMH\nteNbt245xjrjJ19G59Th67LQoP5aI8Zp0NgO33T+ik8C+Jq19tvdbz+CToN64L38QfdnVyad5Wvj\nYQNilSRjjOsguOar66j+qJAzdj1DOZPJoFgs4hWveIW7SqVSZO9vrVZDtVpFrVZzpVb9ToKVnXRN\nyRiDarUaOQf6ipVaxpTvBNRxk5kxxjl7rZVN1rpUk81mIzM6nSUUi0W88pWvdBc5k3WtVsP+/r7j\n7D+XctY8A2MM9vb2ZukMUs3Zd/o6w+/HmINsoNeWNQIXZ8txjJeXl1Gr1Zw9s5a6z5i/12e8v78/\ny0Fdqhn7irPjGzduOMaFQgHHx8duvZzZ+Qy384Ay2rHm3sTZMQd0mUwG1WoV+/v7qFarif11HGMA\ns7bjvrrMDP9pAD8K4C2TeJC4MBcBXFbaYPzQDesjU0dHR27LBxuQhvr8s7F5lUolbG1tYWtrC7dv\n33YdkK4Z6TqTfo+jToaPqKvoFPi3qqRz/SAmyBiYDmd+ruTMREnNo/CNlqUzyZmclLOecc413VKp\nhEceeSTCWYvscBahYWQNSSrnuJkr/55J6yo5XwVj2rJyOj4+jkTOyJiOX22ZuyqUMTmXy2VsbW3h\nkUceiTDuZ8vKn2VXNT8HmP7e+0VjnM/nI2fVJ9kxT7YjY3LipInthDYcxzifz0eW9JLs2Gfs27H/\n90xaAxgP1FgO3xjzaQBPAnjMWvuy/Og+AANgC9FR4xaAP+p3z2KxOLVDRBiiZba8JmhwhseOgPuk\nuZ9SO3M/I1vPUF5dXUWlUsHm5ibW19fdGq5fqIENRDt/ILqO7+8d1WSgSSvOSM/OznjC1lsBPDop\nxsD0ONMQlbOeY85ZAD93cqZT0MGZ31H4nDc3N7G5uYmNjQ3nvP3wMXnqrAKIlmBlJ6XHJk/LKVwl\n50kz1uSoOMZqy2SstsxzFHRQF8eY22bJuFKpYGNjo8eW4xgn2bIOOBbJlq+yvybjUqkU4ah2zBoH\nwEVCnR6xq4yZhc/+ulKpODv2t+IOa8dxjPksk9QAxgM1MrGus38HgLdZa1/Qn1lr7xhj7gN4AsCf\ndF+/CuBRAL846u+6rNhRaIicW+PYgespS0tLSzg8PESz2XRn2PudBBsPj8TUS7frcIbBCm+67nN+\nfu5yALLZbGRUaq11JUI5+/Srdk1bzDAF8O60M1YpZ2XBDGl1+IeHh461bu+hw1dHEseZ23bI2Vrr\nwrvaUTBJjNuJlN/h4SEajUZk21jgHC/flv2trsyQVntlKVZlrIM6Pyt7Y2Ojx5bJmFuyfFum81bG\n+rw80pVFWoItx0uXYOMY86tGXbW/JmPey7djHm9LzrRf2rL210l2zOJavh3Thrm8oDkjadOo+/Cf\nBvAzAN4O4NAYs9X9Uc1ayw2KnwTwIWPMd9HZ5vFRAPcAfG4iTzyiNDNX12aLxSLK5XJPQRWGfDKZ\njNuvHTfzoyPgbG99fT0y+9fCG9pw1OD5XBryBeD26zLkzOSgq1Cj0XBrzwBaaWeshqfOleFaztji\nOJM1t0nxXnHOwOfMgV8cZ+XN5/Iz/gG4gYIuLQTOyYqzZZ25+Z85z69XxrwPGdMhkPHNmzd7bFmP\nVI5jzAkDbZnPCcANEoItD6c4xuSkdgx0GOfzeTQajR7GAJzD58UZfRzjJDvW5QPdtx/HmOF2On3+\nPWnSqDP896CT5PG73vf/PoD/AADW2o8bYwoAPoNOVujvAfhJO8M9+Kenpzg8PES1WnVwAUTWaVj+\nklAzmUzPWgkT/bhdw1qL4+NjHBwcuFG8zuqTyitaa91+XH/rULvddglB1WoVe3t7qNVqkaIe05T3\nNz8r/041YwDuCM39/X0UCgX32dLhkjOrdilnTcjh0abkTLY+Z+4DZkhPOwfOWOgomLHNTqTdbkcS\ng/b29lCtVgPnASJjtWXOwGjLmUwGpVLJOWG1ZX62ZMwMfXKt1Wo4Pz+PbP2infuMeT/G/+L5AAAO\neUlEQVSGkrXSGxmT7/7+frDlIaWMt7e3+9oxGWez2Z76BmrHozCms9Y+we+v1Y739/fdddV2PKpG\n3Yc/1KHO1tqPAPjIGM8zFZ2dnbkGxOxJzdQH4DoJdhCsyw1E9+8zFK+OgEavjkATPlTqCJixzXV7\nZo1rJ1Gr1VCv190gYtra3NzUNaE3WWu/Ffe6tDEGOpw5sCNXMltZWYk4A3YSrM1NVhq+I2vlfHJy\n4hj7nJMGdnT2XANmcRh2EMq51WoFzn3EwTuz3ZXXysoKzs/PXdiVtryysuLsTBnr9lpWYCTjQqEQ\nYezvraeSGDMHJ45xsOV4sW9UxiyGpozp8DUS4Nsxt+ixLx+GsUZg4wZ1Onjn/09OThxfDugODg7c\nICJtuha19HXEqAVUcrkcgIssft2bm8vl3IheC+Lo1h46AlZvy+fzaLVaKBQKPfv0VWxAvDfDfawe\npZ0Ew3J8lqBkaSRHOavzT+KsYTzgIokzjjMZt1qtyAAwibMm9/A+R0dH2Nvbc1e9Xg+ch1AS41wu\nFxmUc/aey+UiMzeuv2tyn87++jH2M7eBaJXGQYyDLQ8nHbjrDintrzOZDPL5fOSMgyQ7Jrdh7Vhr\ncgAXg7o4xjxDYXd31zFmWD+NjK+Fw2e1KXbI2kGfnp5G9mnSGbCRAb3rMASpWZ1aO5l7M1UHBwfu\n8s909jsJ3bPPGd80M7gXReR8cHAAIJkzWZNz3L5oDefpDgvlfHp6GinNaq3ty5l8+ZXLNtVq1XHu\nt80nCC60Wq/XAUQZn5yc9Oy5BjpsWAhFOevn3I8x12opZdxqtSLFXzhwZ7SPdhxseTiZ7n56tWMd\ncNGOC4WC+x5rlCTVN9DBmd9fM7rqM+ZMvVarReyYs3pdvqUN87VpZnwtHL4epsFZAHAxktRMUOCi\nOpJWaqLz1xGkX1WNIT0dRfoXD7TQ+2gIkHXAuc0kLsQUFC+G3HzOnBUyK5fRGl3rY3hXD2BRB6D1\n0tkpMVFHM4WTOPuh3pOTE/c+jQIEzv3l2zKdutqyMgYuHL4yBuJtWcPz/DdtWTkzFDwq42DLg8XP\nnDNt5kv5jMmJ/bUeU62HotGW/QqJSYx9zmSs99EttXyPv2yURl0Lh++vvWgSBrdz6JoPw0CZTCaS\nlX/jxo3IIQp6LjP3UrMBMeFke3sbOzs7PXvrdd1XGxNn/Hxt2htQmkTDZcdLzizD6XPWAhrKmR0B\nOwhdclHOx8fHjvPDhw/7cva3c5EzWfslWoPilcSYtsxOtx/jQqEwtC2fnZ2NZMvKN9jyeNIdEUyI\nZI5UPzteXl52jHlk8iiMHz586Dgrt2EY8/VpZ3wtHL7WtWYI7ujoCI1Gw4XZWOyB60JM+CkUClhb\nW3PFdBSuJnzoaPLo6Aj7+/u4d+8enn/+ebzwwguRELF+VQ3zmqBk+ZzZSTAZjrN+llNl9i47ijjO\nJycnbsDAsJ/P+e7du/j+97+P559/PpZdXDJf3NegwWJny7DqMIwH2TIZc9mPtqxrtHfv3k20Zf/f\n+v/AeHQpY/bXrVYL9XrdhczJWLPwacespeD315yxM/LHQaMyvnPnTqId+/+fR8bXwuFTmiTHzpsN\nhet13IPPzE/dx8k6+5qxqeuyeu3u7mJ3dxf1er1ne1/a9mYumsiZzkCTufpxZkEmY0wkjB/Hmf/e\n2dnBzs4ODg4OAucrks6gyBjob8ssi8yLjHn59hvHONjy1UkZs7/27fjk5CRyyJVvx+yv1Y59G1bG\n29vbC8/4Wjl8IHqAiY7IeKQhQ0S6PU/LGfqJeroupP9uNBouGSvo6sVQLznTaI+Pj1Gv1yNZuSzW\nQ9a6thfHWXnX6/XAeUZKsuWTk5MextytQcbAxQxP197jOAdbnp18xv3s2O+vh7Xj09NTZ8fNZnPG\nf/F0NWqlvQ8C+GkArwPQAvB1AB+w1n5HXvNZAD/nvfUZa+2Tl3zWiYnFEzTrstFoYG9vL1IpS+vo\ns6H5VfP08hO9GE6aJzFJpasvGWO+ijlkDCByMp3PWfkqY5+zz1rLI5Mz1wTnSYvAmYM6OgVlzO2X\nPmN+1T3Xvh3r+iwjRcGWZyNlrHZcr9cjR94mMU7i6zPW5Z1F1qgz/McAfArAN7vv/RiALxpjfsRa\nq8PfLwB4Cp2DGQAgGiOZsXSfJtcCteSphnC0YYyy/s77qBOZB52eniKbzXI2814A78IcMgainBn6\nTeLsH5oBBM5dpZpzP1vWLH1gPMacVQbGs5PPmHlTgfHoGrXSXmTUZ4x5CsBDAG8A8DX50bG1dvvS\nT3dF4ihQpTM9v8Holi7/9YO+l3atrq66RBYA30WnI1goxhoapOKyawNnAHPMGegdvA/LWNvIPPIF\nAmP9eT87nnfOo+iya/jr6NTW3/O+/7gx5gGAfQC/A+BD1lr/NamQNgTf6fPncSPEpAIP/vcXoBHN\nPWMgWnzDN3bdcuMrcJ4fzkm2PA7juJ8FxrPXNBiP8pp519gO33Q+nU8C+Jq19tvyoy8A+HUAdwD8\nMDph/88bY95sU7pvIW5EqD8bVgvaYBaCMRDlnFRZbxgFzunlHGy5rwJjT36FvXHuMU8y4/5xxphf\nAvC3AbzFWvtyn9e9GsD3ADxhrf1yzM//KoA/5JGDKs2OD5qOuEVFxQQZAL+GTvjvUoy7rwmcZ6ir\n4BwYz1aB8eJrAOM32IQDkqixZvjGmE8DeBLAY/0aT/dh7hhjdgC8BkCsMwCAYrHoymQGXZ3ijFRO\n2HorgEcnxRgInGelq+QcGM9GgfHiawDjgRqZWNfZvwPA26y1Lwzx+lcB2ATQt6EFpUey3/jdgfHi\nKnBefAXGQaqhzrenjDFPA/h7AH4WwKExZqt75bo/LxpjPm6MedQY84PGmCcA/CaA7wB4dtD9/VDF\nuAr3Gf8+PMKzq9akGU9qn+ukPotJ3mue7jNtzvP0WSzqfQLjcB9fIzl8AO8BsArgdwG8JNc7uz9v\nA/gxAJ8D8H8B/FsA/xPAj1trTwfdPG0f0HW8j/ezZzFhxjwM47IKDv9y95k253n6LBb1PoFxuI+v\nUffh9x0gWGuPAPzEpZ4oaKba3NzUNaE3+UkggfFiKHBefAXGQb5GneEHBQUFBQUFzaGCww8KCgoK\nCroGSsO+ihwAV6u+u5/wUgr3udx9eLgQumwmpNwwv3tYTeo+k7zXvN1nCpyDLafsPoHx4t9nJMZa\njnAWFzoZ/zZcqbx+NnC+FtdEOAfGqb4C48W/BjIeu9LepGSM2USnYt/3ARzN9GGCqByAvwDgWWvt\n7iRuGDinUhPlHBinUoHx4mtoxjN3+EFBQUFBQUHTV0jaCwoKCgoKugYKDj8oKCgoKOgaKDj8oKCg\noKCga6Dg8IOCgoKCgq6BgsMPCgoKCgq6BkqNwzfG/CNjzB1jTMsY8w1jzJtGfP+HjTHn3vXtId73\nmDHmt4wxL3bf8/aY1/yCMeYlY0zTGPMlY8xrRr2PMeazMc/3ee81HzTG/IEx5sAY88AY81+MMa8d\n9XmGuc8wzzNpBcbudQvLeVaMu+9NDefAeOA95t6W55FxKhy+MebvAvgEgA8D+CsA/heAZ40xN0e8\n1Z8B2ALwSPd66xDvKQL4YwDvQ6d4gf9sHwDw8wDeDeCvATjsPlt2lPt09QXv+X7G+/ljAD4F4FEA\nfwtABsAXjTH5EZ9n4H2GfJ6JKTCOaCE5z5gxkC7OgfFgzbstzx/jWVfa69YB+AaAfyP/NwDuAXj/\nCPf4MIBvXfI5zgG83fveSwD+qfx/FUALwDtHvM9nAfzGiM9zs3uvt17yeeLuM/LzBMaTZ7xInNPC\nOI2cA+PJcw6MR79mPsM3xmQAvAHAb/N7tvMX/jcAbx7xdn+xG6L5njHmPxpjfuCSz/ZqdEZS+mwH\nAH5/jGcDgMe7IZv/Y4x52hhTGfD6dXRGn3uXfJ7IfS7xPGMpMB6oueecZsbd55s158C4V4tmy6ln\nPHOHj85oZgnAA+/7D9D5sIbVNwA8hU7Zx/cAeDWArxpjipd4tkfQ+eAv+2xAJxzzLgB/E8D7AbwN\nwOeNMSbuxd3vfxLA16y1XNsa+XkS7jPy81xSgXGCFohzmhkDwZYnoUkxBhbMlueFcRpOy5uIrLXP\nyn//zBjzBwCeB/BOdMIhM5W19lflv39ujPlTAN8D8DiAL8e85WkAPwrgLZf81bH3GeN5Zq4FZAwE\nzhGlnTEQbHkSSjvnRWWchhn+DoA2OskIqi0A98e9qbW2BuA7AHoyNEfQfXTWqCb6bABgrb2Dzt8e\nl0H6aQBPAnjcWvvyuM/T5z4jPc8EFBjHaME4p5kxEGx5EpoKY2C+bXmeGM/c4VtrTwH8IYAn+L1u\nmOIJAF8f977GmBI6H0bfD27As91BB4w+2yo62ZRjP1v3Pq8CsOk/Xxf6OwD8DWvtC+M+T7/7jPI8\nk1Bg3Pt8i8Y5zYy7zxds+ZKaFuPufebSlueO8WWz/iZxoRPGaaKzRvE6AJ8BsAvg1gj3+FcAfhzA\nDwL46wC+hM5ayeaA9xUBvB7AX0YnM/KfdP//A92fv7/7LH8HwF8C8JsA/h+A7LD36f7s4+iA/kF0\nGsA3AfxvABm5x9MA9tHZprElV05eM/B5Bt1n2OcJjCfPeJE5z5Jx2jgHxotvy/PIeKKN4ZIN6X3o\nnLHcAvA/ALxxxPf/Z3S2h7QAvADgPwF49RDve1sXeNu7/r285iPobK9oAngWwGtGuQ865xU/g85o\n7wjAcwB+yTeShPe3AbzLe13f5xl0n2GfJzCePONF5zwrxmnjHBgvvi3PI2PTvWFQUFBQUFDQAmvm\na/hBQUFBQUFB01dw+EFBQUFBQddAweEHBQUFBQVdAwWHHxQUFBQUdA0UHH5QUFBQUNA1UHD4QUFB\nQUFB10DB4QcFBQUFBV0DBYcfFBQUFBR0DRQcflBQUFBQ0DVQcPhBQUFBQUHXQMHhBwUFBQUFXQP9\nf6TY7CSBT3BOAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fb5e1a80c90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.subplot(241)\n",
    "plt.imshow(G_test[0,:,:,0],cmap='gray')\n",
    "plt.subplot(242)\n",
    "plt.imshow(G_test[1,:,:,0],cmap='gray')\n",
    "plt.subplot(243)\n",
    "plt.imshow(G_test[2,:,:,0],cmap='gray')\n",
    "plt.subplot(244)\n",
    "plt.imshow(G_test[3,:,:,0],cmap='gray')\n",
    "plt.subplot(245)\n",
    "plt.imshow(G_test[4,:,:,0],cmap='gray')\n",
    "plt.subplot(246)\n",
    "plt.imshow(G_test[5,:,:,0],cmap='gray')\n",
    "plt.subplot(247)\n",
    "plt.imshow(G_test[6,:,:,0],cmap='gray')\n",
    "plt.subplot(248)\n",
    "plt.imshow(G_test[7,:,:,0],cmap='gray')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x7fb5e1c6bad0>"
      ]
     },
     "execution_count": 61,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAigAAAFkCAYAAAAKf8APAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n4/8By/PFrConBrEAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fb5c8ae5c90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(np.arange(len(losses)),losses,np.arange(len(losses)),losses_g)\n",
    "plt.legend(['discriminator loss','generator loss'])\n",
    "# plt.ylim([0,10e])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/floptical/anaconda2/envs/tensorflow/lib/python2.7/site-packages/ipykernel/__main__.py:3: FutureWarning: pd.rolling_mean is deprecated for Series and will be removed in a future version, replace with \n",
      "\tSeries.rolling(window=100,center=False).mean()\n",
      "  app.launch_new_instance()\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fb5c8c15a50>"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAg4AAAFkCAYAAABIPLOYAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nP9S4mG0oDK0gAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fb5c882a450>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pdloss=pd.Series(losses_g)\n",
    "# pdloss.rolling(20)\n",
    "pd.rolling_mean(pdloss,100).plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/floptical/anaconda2/envs/tensorflow/lib/python2.7/site-packages/ipykernel/__main__.py:2: FutureWarning: pd.rolling_mean is deprecated for Series and will be removed in a future version, replace with \n",
      "\tSeries.rolling(window=100,center=False).mean()\n",
      "  from ipykernel import kernelapp as app\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fb5c30f8f90>"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAgUAAAFkCAYAAACw3EhvAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fb5c2fd1c50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "pdloss_d=pd.Series(losses)\n",
    "pd.rolling_mean(pdloss_d,100).plot(ylim=(0,100))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7fcebc5fc750>"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAWEAAAFfCAYAAACfj30KAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nVs2oFcKrVq1aNaNWCK9atWrVjPr/AatlNAoVGPE0AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7fcebc7ceb90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "z_test = Z(n,batch_size)\n",
    "z_test[0,:] = np.ones(n)\n",
    "fd = {z_input:z_test}\n",
    "G_test = sess.run(G_z,feed_dict=fd)\n",
    "plt.imshow(G_test[0,:,:,0],cmap='Greys')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "anaconda-cloud": {},
  "kernelspec": {
   "display_name": "Python [conda env:tensorflow]",
   "language": "python",
   "name": "conda-env-tensorflow-py"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.12"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}