du16_2-rbm.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "################i tested rbm from the guy in github, this code has implemented the exact pper from hinton.\n",
    "\n",
    "import glob\n",
    "import tensorflow as tf\n",
    "from tensorflow.python.ops import control_flow_ops\n",
    "from progressbar import ProgressBar\n",
    "import h5py\n",
    "import matplotlib.pyplot as plt\n",
    "I=0\n",
    "def _parse_function(example_proto):\n",
    "    features = {\"X\": tf.FixedLenFeature((257), tf.float32),\n",
    "              \"Y\": tf.FixedLenFeature((257), tf.float32)}\n",
    "    parsed_features = tf.parse_single_example(example_proto, features)\n",
    "    return parsed_features[\"X\"], parsed_features[\"Y\"]\n",
    "\n",
    "def rbm_layer(n_visible, n_hidden, num_epochs, num_cases, lr, ws, bs, layer_n, len_data, directories):\n",
    "    \n",
    "    Data_path = directories[0]\n",
    "    tfrecord_folder_parent = directories[1]\n",
    "    tfrecord_folder = directories[2]\n",
    "    \n",
    "    tfrecord_path_x = os.path.normpath(os.path.join(Data_path,tfrecord_folder_parent,tfrecord_folder))\n",
    "    sorted_names_x = natsorted(os.listdir(tfrecord_path_x))\n",
    "    trainfilenames_x = []\n",
    "    for i in sorted_names_x:\n",
    "        trainfilenames_x.append(os.path.normpath(os.path.join(tfrecord_path_x,i)))\n",
    "#     filenames_x = tf.placeholder(tf.string, shape=[None])\n",
    "    dataset_x = tf.data.TFRecordDataset(trainfilenames_x)\n",
    "    dataset_x = dataset_x.map(_parse_function)  # Parse the record into tensors.\n",
    "#     dataset_x = dataset_x.repeat()  # Repeat the input indefinitely.\n",
    "    dataset_x = dataset_x.batch(batch_size)\n",
    "#     iterator_x = dataset_x.make_initializable_iterator()\n",
    "\n",
    "#     x  = tf.placeholder(tf.float32, [None, n_visible], name=\"x\") #The placeholder variable that holds our data\n",
    "    W  = tf.Variable(tf.random_normal([n_visible, n_hidden], 0.01), name=\"W\") #The weight matrix that stores the edge weights\n",
    "    bh = tf.Variable(tf.zeros([1, n_hidden],  tf.float32, name=\"bh\")) #The bias vector for the hidden layer\n",
    "    bv = tf.Variable(tf.zeros([1, n_visible],  tf.float32, name=\"bv\")) #The bias vector for the visible layer\n",
    "    \n",
    "    W_adder  = tf.zeros((n_visible,n_hidden),dtype=tf.dtypes.float32)\n",
    "    bh_adder = tf.zeros((1,n_hidden),dtype=tf.dtypes.float32)\n",
    "    bv_adder = tf.zeros((1,n_visible),dtype=tf.dtypes.float32)\n",
    "    \n",
    "#     weight_decay = 0.001\n",
    "#     initialmomentum  = 0.5\n",
    "#     finalmomentum    = 0.9\n",
    "    \n",
    "    weight_decay = 0\n",
    "    initialmomentum  = 0\n",
    "    finalmomentum    = 0   \n",
    "    def sample(probs):\n",
    "        #Takes in a vector of probabilities, and returns a random vector of 0s and 1s sampled from the input vector\n",
    "        return tf.floor(probs + tf.random_uniform(tf.shape(probs), 0, 1))\n",
    "\n",
    "\n",
    "    def gibbs_sample(k,data):\n",
    "        #Runs a k-step gibbs chain to sample from the probability distribution of the RBM defined by W, bh, bv\n",
    "        def gibbs_step(count, k, data):\n",
    "            #Runs a single gibbs step. The visible values are initialized to xk\n",
    "#             hk = sample(tf.tanh(tf.matmul(data, W) + bh)) #Propagate the visible values to sample the hidden values\n",
    "            hk = sample(tf.tanh(tf.matmul(data+ tf.random.normal((len(data),len(data[0])),mean=0.0,stddev=1), W) + bh))\n",
    "#             hk = sample(tf.matmul(data, W) + bh)\n",
    "#             data   = tf.matmul(hk, tf.transpose(W)) + bv+ tf.random.normal((len(x[0]),len(x[0][0])),mean=0.0,stddev=1)\n",
    "            data   = tf.matmul(hk, tf.transpose(W)) + bv\n",
    "#             data   = tf.tanh(tf.matmul(hk, tf.transpose(W)) + bv)\n",
    "#             data = tf.truncated_normal((1,n_visible),tf.matmul(hk, tf.transpose(W)) + bv,1)\n",
    "            return count+1, k, data\n",
    "#         print('2')\n",
    "        ct = tf.constant(0) #counter\n",
    "        #Run gibbs steps for k iterations\n",
    "        [_, _, x_sample]=control_flow_ops.while_loop(lambda count, num_iter, *args: count < num_iter,\n",
    "                                             gibbs_step, [ct, tf.constant(k), data], back_prop = False)\n",
    "#         print('3')\n",
    "        #This is not strictly necessary in this implementation, but if you want to adapt this code to use one of TensorFlow's\n",
    "        #optimizers, you need this in order to stop tensorflow from propagating gradients back through the gibbs step\n",
    "        x_sample = tf.stop_gradient(x_sample) \n",
    "#         print('4')\n",
    "        return x_sample\n",
    "\n",
    "    ### Training Update Code\n",
    "    # Now we implement the contrastive divergence algorithm. First, we get the samples of x and h from the probability distribution\n",
    "    #The sample of x\n",
    "#     print('1')\n",
    "#     x_sample = gibbs_sample(5) \n",
    "#     #The sample of the hidden nodes, starting from the visible state of x\n",
    "#     h = sample(tf.sigmoid(tf.matmul(x, W) + bh)) \n",
    "#     #The sample of the hidden nodes, starting from the visible state of x_sample\n",
    "#     h_sample = sample(tf.sigmoid(tf.matmul(x_sample, W) + bh)) \n",
    "\n",
    "#     #Next, we update the values of W, bh, and bv, based on the difference between the samples that we drew and the original values\n",
    "# #     size_bt = tf.cast(tf.shape(x)[0], tf.float32)\n",
    "#     W_adder  = tf.multiply(lr, tf.subtract(tf.matmul(tf.transpose(x), h), tf.matmul(tf.transpose(x_sample), h_sample)))\n",
    "#     bv_adder = tf.multiply(lr, tf.reduce_sum(tf.subtract(x, x_sample), 0, True))\n",
    "#     bh_adder = tf.multiply(lr, tf.reduce_sum(tf.subtract(h, h_sample), 0, True))\n",
    "#     #When we do sess.run(updt), TensorFlow will run all 3 update steps\n",
    "#     updt = [W.assign_add(W_adder), bv.assign_add(bv_adder), bh.assign_add(bh_adder)]\n",
    "\n",
    "\n",
    "\n",
    "    ### Run the graph!\n",
    "    # Now it's time to start a session and run the graph! \n",
    "\n",
    "    incount=0\n",
    "    k=0\n",
    "    pbar = ProgressBar()\n",
    "    errsum = []\n",
    "    \n",
    "    for epoch in pbar(range(num_epochs)):\n",
    "        #Train the RBM on batch_size examples at a time\n",
    "#             for X_batch in da(batch_size, layer_n, ws, bs, len_data, name, data_name):\n",
    "        err=0\n",
    "        for x in dataset_x:\n",
    "            xx=x[0]\n",
    "            incount+=1\n",
    "            if incount%1000==0:\n",
    "                print(incount)\n",
    "            if layer_n>1:\n",
    "                for j in range(1,layer_n):\n",
    "                    xx = tf.tanh(tf.matmul(xx,ws[j])+bs[j])\n",
    "#             else:\n",
    "#                 x_0 = x[0]\n",
    "            x_sample = gibbs_sample(1,xx) \n",
    "            #The sample of the hidden nodes, starting from the visible state of x\n",
    "            h = sample(tf.tanh(tf.matmul(xx, W) + bh)) \n",
    "#             h = sample(tf.matmul(x[0], W) + bh)\n",
    "            #The sample of the hidden nodes, starting from the visible state of x_sample\n",
    "            h_sample = sample(tf.tanh(tf.matmul(x_sample, W) + bh)) \n",
    "#             h_sample = sample(tf.matmul(x_sample, W) + bh)\n",
    "            if epoch>5:\n",
    "                m=finalmomentum\n",
    "            else:\n",
    "                m=initialmomentum\n",
    "            #Next, we update the values of W, bh, and bv, based on the difference between the samples that we drew and the original values\n",
    "        #     size_bt = tf.cast(tf.shape(x)[0], tf.float32)\n",
    "            W_adder  = (m * W_adder) + tf.multiply(lr, tf.subtract(tf.matmul(tf.transpose(xx), h), tf.matmul(tf.transpose(x_sample), h_sample)))- W*weight_decay\n",
    "            bv_adder = (m * bv_adder) + tf.multiply(lr, tf.reduce_sum(tf.subtract(xx, x_sample), 0, True))\n",
    "            bh_adder = (m * bh_adder) + tf.multiply(lr, tf.reduce_sum(tf.subtract(h, h_sample), 0, True))\n",
    "            #When we do sess.run(updt), TensorFlow will run all 3 update steps\n",
    "            err= err+tf.reduce_sum(tf.reduce_sum((xx-x_sample)**2))\n",
    "            updt = [W.assign_add(W_adder).numpy(), bv.assign_add(bv_adder).numpy(), bh.assign_add(bh_adder).numpy(), errsum]\n",
    "#             updt = [W.assign_add(W_adder).numpy(), bv.numpy(), bh.numpy()]\n",
    "        errsum.append(err/100000)\n",
    "    \n",
    "    plt.plot(range(num_epochs), errsum)\n",
    "    plt.show()\n",
    "    print('zoomed in:')\n",
    "    plt.plot(range(4,num_epochs), errsum[4:num_epochs])\n",
    "    plt.show()\n",
    "    return updt\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  0% |                                                                        |\r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "imported\n",
      "1000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "  5% |###                                                                     |\r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2000\n",
      "3000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 10% |#######                                                                 |\r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 15% |##########                                                              |\r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5000\n",
      "6000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 20% |##############                                                          |\r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "7000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 25% |##################                                                      |\r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "8000\n",
      "9000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 30% |#####################                                                   |\r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "10000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 35% |#########################                                               |\r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "11000\n",
      "12000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 40% |############################                                            |\r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "13000\n",
      "14000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 45% |################################                                        |\r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "15000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 50% |####################################                                    |\r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "16000\n",
      "17000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 55% |#######################################                                 |\r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "18000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 60% |###########################################                             |\r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "19000\n",
      "20000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 65% |##############################################                          |\r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "21000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 70% |##################################################                      |\r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "22000\n",
      "23000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 75% |######################################################                  |\r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "24000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 80% |#########################################################               |\r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "25000\n",
      "26000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 85% |#############################################################           |\r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "27000\n",
      "28000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 90% |################################################################        |\r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "29000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      " 95% |####################################################################    |\r"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "30000\n",
      "31000\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100% |########################################################################|\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "zoomed in:\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# import libraries.\n",
    "import numpy as np\n",
    "from sklearn.utils import shuffle\n",
    "import tensorflow as tf\n",
    "import datetime\n",
    "tf.enable_eager_execution()\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.layers import Dense\n",
    "from tensorflow.keras import layers\n",
    "from tensorflow.keras.wrappers.scikit_learn import KerasRegressor\n",
    "# import keras\n",
    "from tensorflow.keras.layers import BatchNormalization\n",
    "from tensorflow.keras import optimizers\n",
    "from tensorflow.keras import regularizers\n",
    "\n",
    "from sklearn.model_selection import cross_val_score\n",
    "from sklearn.model_selection import KFold\n",
    "from sklearn.preprocessing import StandardScaler\n",
    "from sklearn.pipeline import Pipeline\n",
    "from sklearn.metrics import mean_squared_error\n",
    "from pystoi.stoi import stoi\n",
    "import h5py\n",
    "######################\n",
    "#import libraries.\n",
    "import matplotlib.pyplot as plt\n",
    "from tabulate import tabulate\n",
    "import time\n",
    "import os\n",
    "import librosa\n",
    "from librosa.core import stft, istft\n",
    "####import sounddevice as sd\n",
    "import time\n",
    "print('imported')\n",
    "# #######################\n",
    "Data_path = 'D:/studies/university/thesis/speech_separation_codes/du16/donesomestuff'\n",
    "# Data_path = os.getcwd()\n",
    "tfrecord_folder_parent = 'tfrecord_files'\n",
    "tfrecord_folder = 'mixed_10h_norm'\n",
    "tfrecord_val_folder = 'validation_mixed_norm'\n",
    "# ckpt_folder = '3'\n",
    "dirs = [Data_path, tfrecord_folder_parent, tfrecord_folder]\n",
    " \n",
    "# len_data = (684108, 257)\n",
    "# len_data = (100000, 257)\n",
    "# len_data = (2197278, 257)\n",
    "# val_len = (97278,257)\n",
    "len_data = (500000, 257)\n",
    "val_len = (100000,257)\n",
    "w=1\n",
    "#######################\n",
    "#define reconstruct function to reconstruct sound from framed signal.\n",
    "def reconstruct(wave,angle):\n",
    "    recon = np.sqrt(np.power(10, wave))\n",
    "    recon1 = recon*np.cos(angle)+recon*np.sin(angle)*1j\n",
    "    recon = librosa.core.istft((recon1.T), hop_length=256, win_length=512, window='hann')\n",
    "    return recon\n",
    "#######################\n",
    "I=0\n",
    "global batch_size\n",
    "batch_size = 128\n",
    "# epochs_num=50\n",
    "global datalen\n",
    "datalen=len_data[0]\n",
    "\n",
    "h = [512,512,512,257]\n",
    "seed = 7\n",
    "rate1 = 0.1\n",
    "rate2 = 0.2\n",
    "# from tensorflow.keras.layers import Activation\n",
    "# # from keras.layers import Activation\n",
    "# np.random.seed(seed)\n",
    "# model = Sequential()\n",
    "# act1 = layers.LeakyReLU(alpha=0.1)\n",
    "# model.add(layers.Dropout(rate1, noise_shape=None, seed=None))\n",
    "# # ,kernel_regularizer=regularizers.l2(0.001)\n",
    "# model.add(Dense(h[0], input_dim = w*len_data[1]))\n",
    "# model.add(BatchNormalization())\n",
    "# model.add(act1)\n",
    "# # model.add(Activation('sigmoid'))\n",
    "# act2=layers.LeakyReLU(alpha=0.1)\n",
    "# model.add(layers.Dropout(rate2, noise_shape=None, seed=None))\n",
    "# model.add(Dense(h[1]))\n",
    "# model.add(act2)\n",
    "# # act3=layers.LeakyReLU(alpha=0.1)\n",
    "# # # model.add(layers.Dropout(rate, noise_shape=None, seed=None))\n",
    "# # model.add(Dense(h[2]))\n",
    "# # model.add(act3)\n",
    "# act=layers.LeakyReLU(alpha=0.01)\n",
    "# model.add(Dense(len_data[1]))\n",
    "#############################################\n",
    "import os\n",
    "from natsort import natsorted\n",
    "\n",
    "# def _parse_function(example_proto):\n",
    "#     print('1')\n",
    "#     features = {\"X\": tf.FixedLenFeature((3*257), tf.float32),\n",
    "#               \"Y\": tf.FixedLenFeature((257), tf.float32)}\n",
    "#     parsed_features = tf.parse_single_example(example_proto, features)\n",
    "#     print(\"i was here\")\n",
    "#     print('2')\n",
    "#     return parsed_features[\"X\"], parsed_features[\"Y\"]\n",
    "\n",
    "# tfrecord_path = os.path.normpath(os.path.join(Data_path,tfrecord_folder_parent,tfrecord_folder))\n",
    "# sorted_names = natsorted(os.listdir(tfrecord_path))\n",
    "# trainfilenames = []\n",
    "# for i in sorted_names:\n",
    "#     trainfilenames.append(os.path.normpath(os.path.join(tfrecord_path,i)))\n",
    "# filenames = tf.placeholder(tf.string, shape=[None])\n",
    "# dataset = tf.data.TFRecordDataset(filenames)\n",
    "# dataset = dataset.map(_parse_function)  # Parse the record into tensors.\n",
    "# dataset = dataset.repeat()  # Repeat the input indefinitely.\n",
    "# dataset = dataset.batch(batch_size)\n",
    "# iterator = dataset.make_initializable_iterator()\n",
    "\n",
    "# # orig_path = os.getcwd()\n",
    "# tfrecord_path_x = os.path.normpath(os.path.join(Data_path,tfrecord_folder_parent,tfrecord_folder))\n",
    "# sorted_names_x = natsorted(os.listdir(tfrecord_path_x))\n",
    "# trainfilenames_x = []\n",
    "# for i in sorted_names_x:\n",
    "#     trainfilenames_x.append(os.path.normpath(os.path.join(tfrecord_path,i)))\n",
    "# filenames_x = tf.placeholder(tf.string, shape=[None])\n",
    "# dataset_x = tf.data.TFRecordDataset(filenames_x)\n",
    "# dataset_x = dataset_x.map(_parse_function)  # Parse the record into tensors.\n",
    "# dataset_x = dataset_x.repeat()  # Repeat the input indefinitely.\n",
    "# dataset_x = dataset_x.batch(batch_size)\n",
    "# iterator_x = dataset_x.make_initializable_iterator()\n",
    "\n",
    "#######################\n",
    "visible = w*len_data[1]\n",
    "hidden = h[0]\n",
    "visible1 = h[0]\n",
    "hidden1 = h[1]\n",
    "visible2 = h[1]\n",
    "hidden2 = h[2]\n",
    "visible3 = h[2]\n",
    "hidden3 = h[3]\n",
    "\n",
    "layer1 = rbm_layer(visible, hidden, 20, batch_size, 0.0001, [np.eye(visible, visible)], [np.zeros((1,visible))], 1, len_data[0],dirs)\n",
    "# layer2 = rbm_layer(visible1, hidden1, 20, batch_size, 0.00001, [np.eye(visible, visible),layer1[0]], [np.zeros((1,visible)),layer1[2]], 2, len_data[0], dirs)\n",
    "# layer3 = rbm_layer(visible2, hidden2, 20, batch_size, 0.00001, [np.eye(visible,visible),layer1[0],layer2[0]], [np.zeros((1,visible)),layer1[2],layer2[2]], 3, len_data[0], dirs)\n",
    "# layer4 = rbm_layer(visible3, hidden3, 20, batch_size, 0.00001, [np.eye(visible,visible),layer1[0],layer2[0],layer[3]], [np.zeros((1,visible)),layer1[2],layer2[2],layer[3]], 4, len_data[0], dirs)\n",
    "\n",
    "###############################\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(range(4,50), layer1[3][4:])\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[<tf.Tensor: id=5003141, shape=(), dtype=float32, numpy=48993.992>,\n",
       " <tf.Tensor: id=5085262, shape=(), dtype=float32, numpy=3299.5447>,\n",
       " <tf.Tensor: id=5167383, shape=(), dtype=float32, numpy=498.27252>,\n",
       " <tf.Tensor: id=5249504, shape=(), dtype=float32, numpy=156.02788>,\n",
       " <tf.Tensor: id=5331625, shape=(), dtype=float32, numpy=90.85684>,\n",
       " <tf.Tensor: id=5413746, shape=(), dtype=float32, numpy=70.39419>,\n",
       " <tf.Tensor: id=5495867, shape=(), dtype=float32, numpy=61.0814>,\n",
       " <tf.Tensor: id=5577988, shape=(), dtype=float32, numpy=55.887653>,\n",
       " <tf.Tensor: id=5660109, shape=(), dtype=float32, numpy=52.614014>,\n",
       " <tf.Tensor: id=5742230, shape=(), dtype=float32, numpy=50.322826>,\n",
       " <tf.Tensor: id=5824351, shape=(), dtype=float32, numpy=48.67836>,\n",
       " <tf.Tensor: id=5906472, shape=(), dtype=float32, numpy=47.40805>,\n",
       " <tf.Tensor: id=5988593, shape=(), dtype=float32, numpy=46.35858>,\n",
       " <tf.Tensor: id=6070714, shape=(), dtype=float32, numpy=45.55414>,\n",
       " <tf.Tensor: id=6152835, shape=(), dtype=float32, numpy=44.87195>,\n",
       " <tf.Tensor: id=6234956, shape=(), dtype=float32, numpy=44.29702>,\n",
       " <tf.Tensor: id=6317077, shape=(), dtype=float32, numpy=43.78311>,\n",
       " <tf.Tensor: id=6399198, shape=(), dtype=float32, numpy=43.36668>,\n",
       " <tf.Tensor: id=6481319, shape=(), dtype=float32, numpy=42.98594>,\n",
       " <tf.Tensor: id=6563440, shape=(), dtype=float32, numpy=42.64132>,\n",
       " <tf.Tensor: id=6645561, shape=(), dtype=float32, numpy=42.349915>,\n",
       " <tf.Tensor: id=6727682, shape=(), dtype=float32, numpy=42.077267>,\n",
       " <tf.Tensor: id=6809803, shape=(), dtype=float32, numpy=41.814823>,\n",
       " <tf.Tensor: id=6891924, shape=(), dtype=float32, numpy=41.58303>,\n",
       " <tf.Tensor: id=6974045, shape=(), dtype=float32, numpy=41.405346>,\n",
       " <tf.Tensor: id=7056166, shape=(), dtype=float32, numpy=41.18715>,\n",
       " <tf.Tensor: id=7138287, shape=(), dtype=float32, numpy=41.014854>,\n",
       " <tf.Tensor: id=7220408, shape=(), dtype=float32, numpy=40.85292>,\n",
       " <tf.Tensor: id=7302529, shape=(), dtype=float32, numpy=40.667492>,\n",
       " <tf.Tensor: id=7384650, shape=(), dtype=float32, numpy=40.52166>,\n",
       " <tf.Tensor: id=7466771, shape=(), dtype=float32, numpy=40.38993>,\n",
       " <tf.Tensor: id=7548892, shape=(), dtype=float32, numpy=40.263206>,\n",
       " <tf.Tensor: id=7631013, shape=(), dtype=float32, numpy=40.128876>,\n",
       " <tf.Tensor: id=7713134, shape=(), dtype=float32, numpy=40.024605>,\n",
       " <tf.Tensor: id=7795255, shape=(), dtype=float32, numpy=39.868366>,\n",
       " <tf.Tensor: id=7877376, shape=(), dtype=float32, numpy=39.777935>,\n",
       " <tf.Tensor: id=7959497, shape=(), dtype=float32, numpy=39.687572>,\n",
       " <tf.Tensor: id=8041618, shape=(), dtype=float32, numpy=39.56303>,\n",
       " <tf.Tensor: id=8123739, shape=(), dtype=float32, numpy=39.458515>,\n",
       " <tf.Tensor: id=8205860, shape=(), dtype=float32, numpy=39.34931>,\n",
       " <tf.Tensor: id=8287981, shape=(), dtype=float32, numpy=39.23616>,\n",
       " <tf.Tensor: id=8370102, shape=(), dtype=float32, numpy=39.143677>,\n",
       " <tf.Tensor: id=8452223, shape=(), dtype=float32, numpy=39.062065>,\n",
       " <tf.Tensor: id=8534344, shape=(), dtype=float32, numpy=38.950134>,\n",
       " <tf.Tensor: id=8616465, shape=(), dtype=float32, numpy=38.8681>,\n",
       " <tf.Tensor: id=8698586, shape=(), dtype=float32, numpy=38.75115>,\n",
       " <tf.Tensor: id=8780707, shape=(), dtype=float32, numpy=38.6819>,\n",
       " <tf.Tensor: id=8862828, shape=(), dtype=float32, numpy=38.582058>,\n",
       " <tf.Tensor: id=8944949, shape=(), dtype=float32, numpy=38.47524>,\n",
       " <tf.Tensor: id=9027070, shape=(), dtype=float32, numpy=38.40557>]"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "layer1[3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 105,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2019-07-16 19:19:48.115610\n",
      "initialized\n",
      "Epoch 1/50\n",
      "781/781 [==============================] - 4s 5ms/step - loss: 1.5745\n",
      "\n",
      "Epoch 00001: val_loss improved from inf to 1.57449, saving model to D:\\studies\\university\\thesis\\speech_separation_codes\\du16\\donesomestuff\\checkpoints\\3\\weights.01.hdf5\n",
      "781/781 [==============================] - 14s 18ms/step - loss: 2.0359 - val_loss: 1.5745\n",
      "Epoch 2/50\n",
      "424/781 [===============>..............] - ETA: 2s - loss: 1.4739"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Exception ignored in: <bound method EagerResourceDeleter.__del__ of <tensorflow.python.ops.resource_variable_ops.EagerResourceDeleter object at 0x00000204F1C09898>>\n",
      "Traceback (most recent call last):\n",
      "  File \"C:\\Users\\ASUS\\Anaconda3\\envs\\myenv\\lib\\site-packages\\tensorflow\\python\\ops\\resource_variable_ops.py\", line 139, in __del__\n",
      "    self._handle, ignore_lookup_error=True)\n",
      "  File \"C:\\Users\\ASUS\\Anaconda3\\envs\\myenv\\lib\\site-packages\\tensorflow\\python\\ops\\gen_resource_variable_ops.py\", line 300, in destroy_resource_op\n",
      "    \"ignore_lookup_error\", ignore_lookup_error)\n",
      "KeyboardInterrupt: \n"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-105-006aa6b2bfdd>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[0;32m    102\u001b[0m \u001b[1;31m# opt = tf.keras.optimizers.SGD()\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    103\u001b[0m \u001b[0mmodel\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mcompile\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mloss\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;34m'mean_squared_error'\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0moptimizer\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mopt\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 104\u001b[1;33m \u001b[0mhistory\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mmodel\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[1;33m(\u001b[0m \u001b[0mdataset\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0msteps_per_epoch\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0msteps\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mepochs\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mepochs_num\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mcallbacks\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;33m[\u001b[0m\u001b[0mcp_callback\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mearly_stop\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mverbose\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mvalidation_data\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mdataset_val\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mvalidation_steps\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mval_steps\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    105\u001b[0m \u001b[1;31m#     model.save(os.path.normpath(os.path.join(Data_path, 'models', \"model_3h_dataset.h5\")))\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    106\u001b[0m \u001b[1;31m#     tf.keras.models.save_model(model, os.path.normpath(os.path.join(Data_path, 'models', \"model_3h_dataset.h5\")))\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\myenv\\lib\\site-packages\\tensorflow\\python\\keras\\engine\\training.py\u001b[0m in \u001b[0;36mfit\u001b[1;34m(self, x, y, batch_size, epochs, verbose, callbacks, validation_split, validation_data, shuffle, class_weight, sample_weight, initial_epoch, steps_per_epoch, validation_steps, max_queue_size, workers, use_multiprocessing, **kwargs)\u001b[0m\n\u001b[0;32m    849\u001b[0m           \u001b[0mvalidation_steps\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mvalidation_steps\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    850\u001b[0m           \u001b[0mworkers\u001b[0m\u001b[1;33m=\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 851\u001b[1;33m           initial_epoch=initial_epoch)\n\u001b[0m\u001b[0;32m    852\u001b[0m     elif distributed_training_utils.is_tpu_strategy(\n\u001b[0;32m    853\u001b[0m         self._distribution_strategy):\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\myenv\\lib\\site-packages\\tensorflow\\python\\keras\\engine\\training_generator.py\u001b[0m in \u001b[0;36mmodel_iteration\u001b[1;34m(model, data, steps_per_epoch, epochs, verbose, callbacks, validation_data, validation_steps, class_weight, max_queue_size, workers, use_multiprocessing, shuffle, initial_epoch, mode, batch_size, **kwargs)\u001b[0m\n\u001b[0;32m    223\u001b[0m           \u001b[0muse_multiprocessing\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0muse_multiprocessing\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    224\u001b[0m           \u001b[0mmax_queue_size\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mmax_queue_size\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 225\u001b[1;33m           mode='test')\n\u001b[0m\u001b[0;32m    226\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    227\u001b[0m       \u001b[1;32mif\u001b[0m \u001b[1;32mnot\u001b[0m \u001b[0misinstance\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mval_results\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mlist\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\myenv\\lib\\site-packages\\tensorflow\\python\\keras\\engine\\training_generator.py\u001b[0m in \u001b[0;36mmodel_iteration\u001b[1;34m(model, data, steps_per_epoch, epochs, verbose, callbacks, validation_data, validation_steps, class_weight, max_queue_size, workers, use_multiprocessing, shuffle, initial_epoch, mode, batch_size, **kwargs)\u001b[0m\n\u001b[0;32m    175\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    176\u001b[0m     \u001b[1;32mfor\u001b[0m \u001b[0mstep\u001b[0m \u001b[1;32min\u001b[0m \u001b[0mrange\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0msteps_per_epoch\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 177\u001b[1;33m       \u001b[0mbatch_data\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0m_get_next_batch\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0moutput_generator\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mmode\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    178\u001b[0m       \u001b[1;32mif\u001b[0m \u001b[0mbatch_data\u001b[0m \u001b[1;32mis\u001b[0m \u001b[1;32mNone\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    179\u001b[0m         \u001b[0mcallbacks\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mmodel\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mstop_training\u001b[0m \u001b[1;33m=\u001b[0m \u001b[1;32mTrue\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\myenv\\lib\\site-packages\\tensorflow\\python\\keras\\engine\\training_generator.py\u001b[0m in \u001b[0;36m_get_next_batch\u001b[1;34m(output_generator, mode)\u001b[0m\n\u001b[0;32m    256\u001b[0m   \u001b[1;34m\"\"\"Retrieves the next batch of input data.\"\"\"\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    257\u001b[0m   \u001b[1;32mtry\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 258\u001b[1;33m     \u001b[0mgenerator_output\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mnext\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0moutput_generator\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    259\u001b[0m   \u001b[1;32mexcept\u001b[0m \u001b[1;33m(\u001b[0m\u001b[0merrors\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mOutOfRangeError\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mStopIteration\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    260\u001b[0m     \u001b[1;31m# Returning `None` will trigger looping to stop.\u001b[0m\u001b[1;33m\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\myenv\\lib\\site-packages\\tensorflow\\python\\data\\ops\\iterator_ops.py\u001b[0m in \u001b[0;36m__next__\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m    544\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    545\u001b[0m   \u001b[1;32mdef\u001b[0m \u001b[0m__next__\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m  \u001b[1;31m# For Python 3 compatibility\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 546\u001b[1;33m     \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mnext\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    547\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    548\u001b[0m   \u001b[1;32mdef\u001b[0m \u001b[0m_next_internal\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\myenv\\lib\\site-packages\\tensorflow\\python\\data\\ops\\iterator_ops.py\u001b[0m in \u001b[0;36mnext\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m    573\u001b[0m     \"\"\"\n\u001b[0;32m    574\u001b[0m     \u001b[1;32mtry\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 575\u001b[1;33m       \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_next_internal\u001b[0m\u001b[1;33m(\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m    576\u001b[0m     \u001b[1;32mexcept\u001b[0m \u001b[0merrors\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mOutOfRangeError\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    577\u001b[0m       \u001b[1;32mraise\u001b[0m \u001b[0mStopIteration\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\myenv\\lib\\site-packages\\tensorflow\\python\\data\\ops\\iterator_ops.py\u001b[0m in \u001b[0;36m_next_internal\u001b[1;34m(self)\u001b[0m\n\u001b[0;32m    565\u001b[0m             \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_resource\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    566\u001b[0m             \u001b[0moutput_types\u001b[0m\u001b[1;33m=\u001b[0m\u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_flat_output_types\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m--> 567\u001b[1;33m             output_shapes=self._flat_output_shapes)\n\u001b[0m\u001b[0;32m    568\u001b[0m \u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m    569\u001b[0m       \u001b[1;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_structure\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_from_compatible_tensor_list\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mret\u001b[0m\u001b[1;33m)\u001b[0m  \u001b[1;31m# pylint: disable=protected-access\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;32m~\\Anaconda3\\envs\\myenv\\lib\\site-packages\\tensorflow\\python\\ops\\gen_dataset_ops.py\u001b[0m in \u001b[0;36miterator_get_next_sync\u001b[1;34m(iterator, output_types, output_shapes, name)\u001b[0m\n\u001b[0;32m   1861\u001b[0m         \u001b[0m_ctx\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_context_handle\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0m_ctx\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_eager_context\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mdevice_name\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1862\u001b[0m         \u001b[1;34m\"IteratorGetNextSync\"\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0mname\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0m_ctx\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_post_execution_callbacks\u001b[0m\u001b[1;33m,\u001b[0m \u001b[0miterator\u001b[0m\u001b[1;33m,\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m-> 1863\u001b[1;33m         \"output_types\", output_types, \"output_shapes\", output_shapes)\n\u001b[0m\u001b[0;32m   1864\u001b[0m       \u001b[1;32mreturn\u001b[0m \u001b[0m_result\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m   1865\u001b[0m     \u001b[1;32mexcept\u001b[0m \u001b[0m_core\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0m_FallbackException\u001b[0m\u001b[1;33m:\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "I=0\n",
    "\n",
    "# epochs_num=50\n",
    "global datalen\n",
    "datalen=len_data[0]\n",
    "\n",
    "\n",
    "seed = 7\n",
    "rate1 = 0.1\n",
    "rate2 = 0.2\n",
    "buffersize = 1000\n",
    "from tensorflow.keras.layers import Activation\n",
    "# from keras.layers import Activation\n",
    "np.random.seed(seed)\n",
    "act1 = layers.LeakyReLU(alpha=0.1)\n",
    "model = Sequential()\n",
    "# model.add(layers.Dropout(rate1, noise_shape=None, seed=None))\n",
    "# ,kernel_regularizer=regularizers.l2(0.001)\n",
    "# model.add(Dense(h[0], input_dim = w*len_data[1]))\n",
    "# model.add(Dense(h[0], input_dim = w*len_data[1], kernel_initializer= 'lecun_uniform', bias_initializer = 'lecun_uniform'))\n",
    "# model.add(Dense(h[0], input_dim = w*len_data[1], kernel_initializer= tf.constant_initializer(layer1[0]), bias_initializer = tf.constant_initializer(layer1[2])))\n",
    "# , kernel_initializer= tf.constant_initializer(layer1[0]), bias_initializer = tf.constant_initializer(layer1[2])\n",
    "# tf.constant_initializer(layer1[0])\n",
    "# tf.constant_initializer(layer1[2])\n",
    "# model.add(BatchNormalization())\n",
    "# model.add(act1)\n",
    "# model.add(Activation('tanh'))\n",
    "# act2=layers.LeakyReLU(alpha=0.1)\n",
    "model.add(Activation('sigmoid'))\n",
    "# model.add(layers.Dropout(rate2, noise_shape=None, seed=None))\n",
    "# model.add(Dense(h[1]))\n",
    "# model.add(act2)\n",
    "# act3=layers.LeakyReLU(alpha=0.1)\n",
    "# # model.add(layers.Dropout(rate, noise_shape=None, seed=None))\n",
    "# model.add(Dense(h[2]))\n",
    "# model.add(act3)\n",
    "# act=layers.LeakyReLU(alpha=0.01)\n",
    "model.add(Dense(len_data[1]))\n",
    "#############################################\n",
    "import os\n",
    "\n",
    "def _parse_function(example_proto):\n",
    "    features = {\"X\": tf.FixedLenFeature((w*257), tf.float32),\n",
    "              \"Y\": tf.FixedLenFeature((257), tf.float32)}\n",
    "    parsed_features = tf.parse_single_example(example_proto, features)\n",
    "    return parsed_features[\"X\"], parsed_features[\"Y\"]\n",
    "\n",
    "tfrecord_path = os.path.normpath(os.path.join(Data_path,tfrecord_folder_parent,tfrecord_folder))\n",
    "sorted_names = natsorted(os.listdir(tfrecord_path))\n",
    "trainfilenames = []\n",
    "for i in sorted_names:\n",
    "    trainfilenames.append(os.path.normpath(os.path.join(tfrecord_path,i)))\n",
    "# filenames = tf.placeholder(tf.string, shape=[None])\n",
    "# dataset = tf.data.TFRecordDataset(filenames)\n",
    "dataset = tf.data.TFRecordDataset(trainfilenames)\n",
    "dataset = dataset.map(_parse_function)  # Parse the record into tensors.\n",
    "dataset = dataset.repeat()  # Repeat the input indefinitely.\n",
    "dataset = dataset.batch(batch_size)\n",
    "dataset = dataset.shuffle(buffersize)\n",
    "# iterator = dataset.make_initializable_iterator()\n",
    "\n",
    "tfrecord_path_val = os.path.normpath(os.path.join(Data_path,tfrecord_folder_parent,tfrecord_val_folder))\n",
    "sorted_names_val = natsorted(os.listdir(tfrecord_path_val))\n",
    "trainfilenames_val = []\n",
    "for i in sorted_names_val:\n",
    "    trainfilenames_val.append(os.path.normpath(os.path.join(tfrecord_path_val,i)))\n",
    "# filenames_val = tf.placeholder(tf.string, shape=[None])\n",
    "# dataset_val = tf.data.TFRecordDataset(filenames_val)\n",
    "dataset_val = tf.data.TFRecordDataset(trainfilenames_val)\n",
    "dataset_val = dataset_val.map(_parse_function)  # Parse the record into tensors.\n",
    "dataset_val = dataset_val.repeat()  # Repeat the input indefinitely.\n",
    "dataset_val = dataset_val.batch(128)\n",
    "# iterator_val = dataset_val.make_initializable_iterator()\n",
    "\n",
    "epochs_num = 50\n",
    "steps = len_data[0] // batch_size\n",
    "val_steps = val_len[0] // batch_size\n",
    "# You can feed the initializer with the appropriate filenames for the current\n",
    "# phase of execution, e.g. training vs. validation.\n",
    "# next_elem = iterator_val.get_next()\n",
    "# Initialize `iterator` with training data.\n",
    "\n",
    "if not os.path.exists(os.path.join(Data_path,\"checkpoints\",ckpt_folder)):\n",
    "    os.makedirs(os.path.join(Data_path,\"checkpoints\",ckpt_folder))\n",
    "\n",
    "print(datetime.datetime.now())\n",
    "# with tf.Session() as sess:\n",
    "#     sess.run(tf.global_variables_initializer())\n",
    "#     sess.run(iterator.initializer, feed_dict={filenames: trainfilenames})\n",
    "#     sess.run(iterator_val.initializer, feed_dict={filenames_val: trainfilenames_val})\n",
    "print(\"initialized\")\n",
    "checkpoint_path = os.path.normpath(os.path.join(Data_path,\"checkpoints\",ckpt_folder,\"weights.{epoch:02d}.hdf5\"))\n",
    "checkpoint_dir = os.path.dirname(checkpoint_path)\n",
    "\n",
    "cp_callback = tf.keras.callbacks.ModelCheckpoint(\n",
    "    checkpoint_path, verbose=1, save_best_only=True, save_weights_only=True)\n",
    "        # Save weights, every 5-epochs.\n",
    "#         period=1)\n",
    "early_stop = tf.keras.callbacks.EarlyStopping(monitor='val_loss', patience=30)\n",
    "opt = tf.keras.optimizers.Adamax()\n",
    "# opt = tf.train.AdamOptimizer()\n",
    "# opt = tf.keras.optimizers.SGD()\n",
    "model.compile(loss='mean_squared_error', optimizer=opt)\n",
    "history = model.fit( dataset, steps_per_epoch=steps,epochs=epochs_num, callbacks = [cp_callback,early_stop], verbose=1,validation_data=dataset_val,validation_steps=val_steps)\n",
    "#     model.save(os.path.normpath(os.path.join(Data_path, 'models', \"model_3h_dataset.h5\")))\n",
    "#     tf.keras.models.save_model(model, os.path.normpath(os.path.join(Data_path, 'models', \"model_3h_dataset.h5\")))\n",
    "#     model.save_weights(os.path.normpath(os.path.join(Data_path, 'models', \"model_3h_dataset.h5\")))\n",
    "model_json = model.to_json()\n",
    "with open(os.path.normpath(os.path.join(Data_path, 'models', \"model_\"+ckpt_folder+\".json\")), \"w\") as json_file:\n",
    "    json_file.write(model_json)\n",
    "# # serialize weights to HDF5\n",
    "model.save_weights(os.path.normpath(os.path.join(Data_path, 'models', \"model_\"+ckpt_folder+\".h5\")))\n",
    "print(\"Saved model to disk\")\n",
    "    \n",
    "print(datetime.datetime.now())\n",
    "%matplotlib inline\n",
    "# summarize history for loss\n",
    "plt.plot(history.history['loss'])\n",
    "plt.plot(history.history['val_loss'])\n",
    "plt.title('model loss')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train'], loc='upper left')\n",
    "plt.show()\n",
    "plt.savefig(os.path.normpath(os.path.join(Data_path,'images',ckpt_folder+'.png')))\n",
    "# model_json = model.to_json()\n",
    "# with open(\"model_10h_dataset.json\", \"w\") as json_file:\n",
    "#     json_file.write(model_json)\n",
    "# model.save_weights(\"model_10h_dataset.h5\")\n",
    "# print(\"Saved model to disk\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "def sample(probs):\n",
    "    #Takes in a vector of probabilities, and returns a random vector of 0s and 1s sampled from the input vector\n",
    "    return tf.floor(probs + tf.random_uniform(tf.shape(probs), 0, 1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "write_path = 'D:/studies/university/thesis/speech_separation_codes/du16/donesomestuff/10hdata'\n",
    "mixed_folder = os.path.normpath(os.path.join(write_path,'mixed_log_10h_norm'))\n",
    "h5f = h5py.File(mixed_folder+'.hdf5','r')\n",
    "ftr = h5f['mixed_log_10h_norm'][0:126]\n",
    "h5f.close()\n",
    "mixed_folder = os.path.normpath(os.path.join(write_path,'mixed_phase_10h_nozeroinsert'))\n",
    "h5f = h5py.File(mixed_folder+'.hdf5','r')\n",
    "ftr_phase = h5f['mixed_phase_10h_nozeroinsert'][0:126]\n",
    "h5f.close()\n",
    "mixed_folder = os.path.normpath(os.path.join(write_path,'mixed_log_10h_nozeroinsert'))\n",
    "h5f = h5py.File(mixed_folder+'.hdf5','r')\n",
    "ftr_raw = h5f['mixed_log_10h_nozeroinsert'][0:126]\n",
    "h5f.close()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(1, 512)"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "layer2[2].shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'layer2' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[1;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "\u001b[1;32m<ipython-input-11-a8d71fddf553>\u001b[0m in \u001b[0;36m<module>\u001b[1;34m\u001b[0m\n\u001b[0;32m      1\u001b[0m \u001b[0mhidden1\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mmatmul\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mftr\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mlayer1\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m+\u001b[0m\u001b[0mlayer1\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m2\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[1;32m----> 2\u001b[1;33m \u001b[0mhidden2\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mmatmul\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mhidden1\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mlayer2\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m+\u001b[0m\u001b[0mlayer2\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m2\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0m\u001b[0;32m      3\u001b[0m \u001b[0mhidden3\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mmatmul\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mhidden2\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mlayer3\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m+\u001b[0m\u001b[0mlayer3\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m2\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      4\u001b[0m \u001b[0mhidden3_b\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mmatmul\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mhidden3\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mlayer3\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mT\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m+\u001b[0m\u001b[0mlayer3\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n\u001b[0;32m      5\u001b[0m \u001b[0mhidden2_b\u001b[0m \u001b[1;33m=\u001b[0m \u001b[0mnp\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mmatmul\u001b[0m\u001b[1;33m(\u001b[0m\u001b[0mhidden3_b\u001b[0m\u001b[1;33m,\u001b[0m\u001b[0mlayer2\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m0\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m.\u001b[0m\u001b[0mT\u001b[0m\u001b[1;33m)\u001b[0m\u001b[1;33m+\u001b[0m\u001b[0mlayer2\u001b[0m\u001b[1;33m[\u001b[0m\u001b[1;36m1\u001b[0m\u001b[1;33m]\u001b[0m\u001b[1;33m\u001b[0m\u001b[0m\n",
      "\u001b[1;31mNameError\u001b[0m: name 'layer2' is not defined"
     ]
    }
   ],
   "source": [
    "hidden1 = np.matmul(ftr,layer1[0])+layer1[2]\n",
    "hidden2 = np.matmul(hidden1,layer2[0])+layer2[2]\n",
    "hidden3 = np.matmul(hidden2,layer3[0])+layer3[2]\n",
    "hidden3_b = np.matmul(hidden3,layer3[0].T)+layer3[1]\n",
    "hidden2_b = np.matmul(hidden3_b,layer2[0].T)+layer2[1]\n",
    "hidden1_b = np.matmul(hidden2_b,layer1[0].T)+layer1[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "def relu(x):\n",
    "    return x * (x > 0)\n",
    "# hidden1 = 1/(1+np.exp(-np.matmul(ftr,layer1[0])+layer1[2]))\n",
    "# hidden2 = 1/(1+np.exp(-np.matmul(hidden1,layer2[0])+layer2[2]))\n",
    "# hidden3 = 1/(1+np.exp(-np.matmul(hidden2,layer3[0])+layer3[2]))\n",
    "hidden1 = np.tanh(np.matmul(ftr,layer1[0])+layer1[2])\n",
    "hidden2 = np.tanh(np.matmul(hidden1,layer2[0])+layer2[2])\n",
    "hidden3 = np.tanh(np.matmul(hidden2,layer3[0])+layer3[2])\n",
    "hidden3_b = np.matmul(hidden3,layer3[0].T)+layer3[1]\n",
    "hidden2_b = np.matmul(hidden3_b,layer2[0].T)+layer2[1]\n",
    "hidden1_b = np.matmul(hidden2_b,layer1[0].T)+layer1[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(126, 512)\n"
     ]
    }
   ],
   "source": [
    "# hidden1 = 1/(1+np.exp(-np.matmul(ftr,layer1[0])+layer1[2]))\n",
    "def leaky_relu(x, alpha):\n",
    "    output = np.where(x > 0, x, x * alpha)\n",
    "    return output\n",
    "hidden1 = np.tanh(np.matmul(ftr,layer1[0])+layer1[2])\n",
    "# hidden2 = 1/(1+np.exp(-np.matmul(hidden1,layer2[0])+layer2[2]))\n",
    "print(hidden1.shape)\n",
    "hidden2 = np.tanh(np.matmul(hidden1,layer2[0])+layer2[2])\n",
    "# hidden3 = 1/(1+np.exp(-np.matmul(hidden2,layer3[0])+layer3[2]))\n",
    "hidden2_b = np.matmul(hidden2,layer2[0].T)+layer2[1]\n",
    "hidden1_b = np.matmul(hidden2_b,layer1[0].T)+layer1[1]\n",
    "# hidden1_b = np.matmul(hidden2_b,layer1[0].T)+layer1[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def relu(x):\n",
    "    return x * (x > 0)\n",
    "def leaky_relu(x, alpha):\n",
    "    output = np.where(x > 0, x, x * alpha)\n",
    "    return output\n",
    "# hidden1 = 1/(1+np.exp(-np.matmul(ftr,layer1[0])+layer1[2]))\n",
    "hidden1 = np.tanh(np.matmul(ftr,layer1[0])+layer1[2])\n",
    "# hidden2 = 1/(1+np.exp(-np.matmul(hidden1,layer2[0])+layer2[2]))\n",
    "# hidden3 = 1/(1+np.exp(-np.matmul(hidden2,layer3[0])+layer3[2]))\n",
    "hidden1_b = np.matmul(hidden1,layer1[0].T)+layer1[1]\n",
    "# hidden1_b = np.matmul(hidden2_b,layer1[0].T)+layer1[1]\n",
    "# hidden1_b = np.matmul(hidden2_b,layer1[0].T)+layer1[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[nan, nan, nan, ..., nan, nan, nan],\n",
       "       [nan, nan, nan, ..., nan, nan, nan],\n",
       "       [nan, nan, nan, ..., nan, nan, nan],\n",
       "       ...,\n",
       "       [nan, nan, nan, ..., nan, nan, nan],\n",
       "       [nan, nan, nan, ..., nan, nan, nan],\n",
       "       [nan, nan, nan, ..., nan, nan, nan]], dtype=float32)"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "hidden1_b"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[-0.16172186, -1.7788432 , -0.39018264, ..., -1.6597615 ,\n",
       "        -0.90431803, -0.45906612],\n",
       "       [-0.0668347 , -0.28278783, -0.29795218, ..., -1.984158  ,\n",
       "        -1.311973  , -1.1100224 ],\n",
       "       [ 0.1633877 ,  0.08443733, -0.4262945 , ..., -0.9554962 ,\n",
       "        -0.93476564, -0.80358255],\n",
       "       ...,\n",
       "       [-1.0645881 ,  0.14120764, -0.45629025, ...,  0.28538626,\n",
       "         0.11404049, -0.8666614 ],\n",
       "       [ 0.16425018,  0.00892877, -0.34537208, ..., -0.9353687 ,\n",
       "        -0.81345695, -0.4461429 ],\n",
       "       [ 0.16982023, -0.05551087, -0.20826632, ..., -0.288958  ,\n",
       "        -0.92982405, -0.9446395 ]], dtype=float32)"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ftr"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\ASUS\\Anaconda3\\envs\\myenv\\lib\\site-packages\\ipykernel_launcher.py:53: RuntimeWarning: overflow encountered in power\n",
      "C:\\Users\\ASUS\\Anaconda3\\envs\\myenv\\lib\\site-packages\\ipykernel_launcher.py:54: RuntimeWarning: invalid value encountered in multiply\n",
      "C:\\Users\\ASUS\\Anaconda3\\envs\\myenv\\lib\\site-packages\\matplotlib\\mlab.py:138: RuntimeWarning: invalid value encountered in multiply\n",
      "  return windowValsRep * x, windowVals\n",
      "C:\\Users\\ASUS\\Anaconda3\\envs\\myenv\\lib\\site-packages\\matplotlib\\image.py:397: UserWarning: Warning: converting a masked element to nan.\n",
      "  dv = (np.float64(self.norm.vmax) -\n",
      "C:\\Users\\ASUS\\Anaconda3\\envs\\myenv\\lib\\site-packages\\matplotlib\\image.py:398: UserWarning: Warning: converting a masked element to nan.\n",
      "  np.float64(self.norm.vmin))\n",
      "C:\\Users\\ASUS\\Anaconda3\\envs\\myenv\\lib\\site-packages\\matplotlib\\image.py:405: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_min = np.float64(newmin)\n",
      "C:\\Users\\ASUS\\Anaconda3\\envs\\myenv\\lib\\site-packages\\matplotlib\\image.py:410: UserWarning: Warning: converting a masked element to nan.\n",
      "  a_max = np.float64(newmax)\n",
      "C:\\Users\\ASUS\\Anaconda3\\envs\\myenv\\lib\\site-packages\\matplotlib\\colors.py:885: UserWarning: Warning: converting a masked element to nan.\n",
      "  dtype = np.min_scalar_type(value)\n",
      "C:\\Users\\ASUS\\Anaconda3\\envs\\myenv\\lib\\site-packages\\numpy\\ma\\core.py:713: UserWarning: Warning: converting a masked element to nan.\n",
      "  data = np.array(a, copy=False, subok=subok)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "mean = np.loadtxt(os.path.normpath(os.path.join(write_path,'mean_mixed_log.txt')))\n",
    "std = np.loadtxt(os.path.normpath(os.path.join(write_path,'std_mixed_log.txt')))\n",
    "hidden1_b_unnorm = (hidden1_b*std)+mean\n",
    "recon = reconstruct(hidden1_b_unnorm, ftr_phase)\n",
    "import sounddevice as sd\n",
    "sd.play(recon,16000)\n",
    "\n",
    "Pxx, freqs, bins, im = plt.specgram(recon, NFFT=512, Fs=16000, noverlap=256)\n",
    "plt.show()\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [],
   "source": [
    "import soundfile as sf\n",
    "sf.write('D:\\\\studies\\\\university\\\\thesis\\\\speech_separation_codes\\\\du16\\\\donesomestuff\\\\results\\\\rbm\\\\1\\\\3layer_sig_hid_lin_vis_noise_in_gibbs_x.wav',recon,16000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAD8CAYAAACb4nSYAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjAsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+17YcXAAAgAElEQVR4nOy9eYxs+XXfd27tVV3V+/r67fOW2WdIUSRFSqIWSpaERApgS1AQGJQsgDIgMYq3aIlhA44sKEgQQYbgAIRsgwoiUTKRREycWBFlyaElkiKHpDjL45u3r92v1+qu6tqrbv74nVPnc6eqhzNyxNcE+weQr+Z21V1+9/c76/d8TxTHsRyP43E8jsfx+OYYqcd9A8fjeByP43E8vn7jWOgfj+NxPI7HN9E4FvrH43gcj+PxTTSOhf7xOB7H43h8E41joX88jsfxOB7fRONY6B+P43E8jsc30XhLQj+Kor8TRdGrURS9EkXR70RRVIii6FwURZ+LouhaFEW/G0VRTr+b1/++rn8/i/P8oh6/GkXRX/ureaTjcTyOx/E4HoeNryn0oyhaFZH/UkTeFcfxsyKSFpEfF5H/TkR+LY7jiyKyKyI/pT/5KRHZjeP4goj8mn5Poih6Wn/3jIj8gIj88yiK0v//Ps7xOB7H43gcjzcbbzW8kxGRYhRFGREpiciaiHyPiHxC//4xEfnP9POP6H+L/v17oyiK9PjH4zhux3F8S0Sui8i7/+Mf4Xgcj+NxPI7HWx2Zr/WFOI4fRFH0P4jIXRFpisj/IyIviUg1juOefu2+iKzq51URuae/7UVRtCcic3r8szg1fzMcURR9WEQ+LCKSLmS+pXJmxu5E/z8afrdVy/t9Jp4kfDdX9SOdWVyjE84R51CNHPnnSC8R9yP83T+m0oPh50wqfE4lfu+fu/3gzPTauME0vmv3QvWLv6fafuFcNUx3Z9rPlfid/iwxFyy4ToX/SHX8nDGeK87jvlKjldoxvhzhHDYyTVyqE+alPe/f47wN+n7jdq144N9NN3Ctfvi3N+m/z2/6tdqz4VxRxv8e93B/OO9wblOHPGszvK+YPqifVlJ9GRlRuedfxRzZmuj3/GTpjJ/A1g5Hu5X1/+DLsTWFZ4lyeN6uzkF29JwiyTU5/A0/d/weU7lwjxm8rx7e1/Dd4ZxprNl+b/Tv0sNCTSydr8UIoF/mY2XG/IZbOT3+nDYH4+ZCRCQ73Mvj57Afv7mN3Bv43+3dpg+51l/1qF7d3IrjeGHc376m0I+iaEaClX5ORKoi8q9F5AfHfNWeblQahL8ddjx5II4/KiIfFRGpXFqOn/uNv5n4Oyd2v+VCv3530u/ZNsZ8e3isWOoMPzfq4Xe5Qnd4rL1THH4uzAXp1YGg5u/JXGEbI5/1jV/K+nkPOjkREdl6MDU8loaQyOjvcjk/1v3izPDz4Jn68HNnP9x36sA36KAIKWTPnR+/aLP6DL0OXvueC5kY9zW/UBMRkd39kp8eG9c2zqDl54qa/vdYN2Zq0ucii2cs5v24zR2FIJWofb5zc3F47NQ5l/o1XQd9CMnBZ3wOp757ffjZNmGn73PYhUBrtsP76t6oDI/15vxeTXilJvzY7MzB8HMu489od9Po+BwvV2rDz+VsWJ+bzfLwWL3ta3rnulsqpqgrT+4Mj/WxF+q1goiITFRaw2NU0va5BaWSgQJq7xb8Hk+Ha+TT/ncqs+16WBPt9vhzzU+G+eAcb236/kxt+e/6M2G+irdyw2Odpxv+92o4ntn3c8Wn/BkHaphRkKdoVGFNZdQo6OC+l+b2hp9t33b746POWZ2PFMTWAGKNv0vrdQ8711/1+NPv/+/vHPa3ryn0ReSDInIrjuNNEZEoiv5XEXmfiExHUZRRa/+kiDzU798XkVMicl/DQVMisoPjNvibsaPTS8u9jbB5U2qN9Zp+y+kihOeCL4S0vtwTM/5Cx03+wcaE/6bim7i1rQoAwr0Dqye+54KwMamL/RALS9TipKDvN/wZUpXw+w4EcfucK6sIQjVlz4vnTsNKLRaDUD+o+QbOF6HYmmGxxx1YLFAQWXzXLDsK33Ter9vZ0WvA6opmXTGmzEDr+rW6mz7fvUV/xk4h/K6Qw712fWOmdAOdv+DCe6vu56pVw/soT7uwiN7rbl4dxkER1xj+/sDnq6iGQN8fVTIl/49C0Z/RRqvr7yhhEaugrBT8Wffafq2v3lkJ5yz73znyJ13h29xUq/7cmawLWvOSapuuQKK8/z39MMxBvOr7pN30eaFXsXcQ1j8NGX42BZK+5YZSe86v1SyG52lg3tM5/3v5oitJ25fpOT8GX0fq62G+Sk/6+9x/xGcMa4NGeH/bFUgEb71bCN+N8D7XHk37D9UAisswpCAD0tVRcRnBo0yf8/e18PGwJtf++uh6edzjrQj9uyLy3iiKShLCO98rIl8QkT8Wkb8hIh8XkQ+JyO/r9z+p//0Z/fu/i+M4jqLokyLy21EU/Y8ickJELorIn7/plXspkS214rJh9kv3XXi3nvaXMz3ri+bJ2UciknStqJ1vRPMiItLt+rk6VV+g2Wo43p3y86euuaDv+5qS3Hb4bm7Pp/LgNBb4rfD3/He7AtrZdAsungirtQNBnIbbnoK1ZWECWsx5fJ4vhzlYRwjg5LRf9/6nTofneh6WKQR5A+EyUeG2MOOW6cNb88PP0UT4HUNg/JwrBSHV3vF546iUPRbUUkt4v+bf7VWxcfVakxCejau+WWee3hYRkSwU8/KE3/f1bb/vjY3gcUVU4hDUNVX42Qs+RxHm810r90REZLPlguf1NfdAGhT6qtwHC/BAGJ7Re2juuvBcWt0dfn5m1pVcrRfezVVY9y1YrLFeK9WC9KsjZKO6ro95nbzuf89939bw814t3E8Xhki1jXOpgTNYgNFVdmW6sxW8pPK0v+Mzc+6hFNLwijOjQjGDGNr0yesiIrLV9vl+WILXrAbBzRtL/vvG+FDS1FfDM1Sf8r9XbsMoUX3aaeMd0mvWd9ef8PvLb/m89G/6PbbVUZz4vL/bozLeSkz/c1EUfUJEvigiPRH5koTwy78RkY9HUfTLeuxf6E/+hYj8z1EUXZdg4f+4nufVKIp+T0Re0/P8TBzHYyKkGKlYBiX9ilqUjTNQvVyIEPD36sE7WCr5xqcLbSONjZ+d9MXXVUsm6vqKaS/6rTKe3dO8QGcR94Xf1S6EBd6lVYWwU18t4RgWcZ9hlEk/rwn7CAu59zkPYzS+szPyXOs1D1M0nwxWXiE7ftop/KprwR2vY14EysjCCHSra+t+rXQlPGPM/AQEUh3Wde8gCK9UAeY1NlvcCu+5BstxgHyMWb8FhOA6iKMfbLkyiXRDD7KY12m3frtjwgX9gZ/rdi0obIZOuvDGGHPP7Ibj9RSuf+Dfze2F77bn/X08eujKrIBQUaunCuTPEPp70T0by1FQSEW8x2nNm+B97D+NEFvD57a7Fz5np3ydpqquYAYm9LDk+1i/3Jc2Gl1XNtWWC8JKPlwjC0G/i7/31BNYqewPj61XfZ0N3wPWZq8yfn3vvlNj+rjX+hnkmSo638wBUeZbqqKEsNgJfAE/q9fD86awpI/KiI4ytXL+7Ml4+R9/REQ8lMNk3wBhkvKCW2b19SDgszO+mYuF0dAFLf1uA5alxswL6/AqFv3lppuwblV4DOYR80cIyoRMTIuBiUPbhOOSjSKSgSDM6zMw2cdwgx1PI6HZqsMtUWWSn0JMFNdlrD97N2z8Xhn3CmU1qIdrVV5HuO0Dbs1ZSKb5ObeyW1CcxI0N54bTMuHPnXstCIHec+4+d3dcadgmZv6Nwo1zHxXCPcQQvkzExiqwsmWfVybh0+thPntL/veI4boZnyPL0zShdPKzbv12dM3FsCwpvATCaWhJ34VAnPc1HalRFO3C+ucc63kn8b72n8MzcE3WwndiWLRRDYpNFWIaSrw3g5Cjxt/7E/4sM3/heyn6oe3h5z317tI3/bk6KwjB6Ryk4EkwGT64H34/KFADjYIAREQyCg7ozI0P3+Tmwr6IE8oM+TPb1wjl5id8Dpmwz7xeGrn+13Nc/Sd/96U4jt817m9vJbzz+EYqTsTtRZIhBNvAIiIH+y4Epk8Eq6CU9xey9Xl3/7LPhL+XS4i1riMMoe+0XxqP7ulDEFrCMr0BK3TBr1u5Gqa4doEJAiA85sJ3E8lhWOIzFcSp9R5SSE5lEf5JFRW1hE1BhNPCiRAX7WJx8lzVnseL4yfCdeMN34wlWNIrKyEcUD+DBBzO21HLtIc5ZPwzRi5gaOFD7/WRYJ58/4a8cTyCMkupO184754dY/dNJFI7ilKZmnMFsvXIk4ymEEsFf9YMUEP7pbDOVqb89/U2FCuGWaEDCHoKhpXlEMphzqHZ9HP1qz730WRYq9E5N25SQNyYMURlV4TR0tTlf7DqzzL9Bb9W94MeBhSdjqmS3/deA8pGjYfePV8vFLRTl4LyLyAPsN7yENhg3cMzQwPkks/nBIwWe18L0/73tWsApSzq/qHXX8NnItVUR6WgZGPkobqq3KMEsssvNQRQIKzVQz6HyshG8dHRM6qPttAfRNJX13+okbG4Mnv+cstPeSzUEmd5uMdL7waCQwUd0Q97MLDSBqOEl8qE0ADWmIWA+nAp6QofnFC3Gkm1/EOkqpYVuULr/IYLocZFF162iRjbJmJlqRw2xv0931RMzlq4wFxqEZE84qu7Ox4Cm9UcSR0xfyJTGppoLed802x3EcZQBdWdQxJ0zZ+byqCooaIJKOl41p/L7rcGZMvqSfcqGovhvGen/Vhn4Es7Axegpzu/2fN72S+7wTA5Ee5lG3NBZXRhJaCGDhCumIFwrGR9bl/+4rlwX885XqHdAyJM0SJchxwzz7ggtvncOfA5Thf9WvvXFPAAiG9zCRbptnq3UGDVd/jaOFtxoZrTUEsl5x7hyoSHV156/Wy41iHAlFQ0emz1qUfDz9MFny+L6fcOmYMvfvGCiIjMLDtaa/+kK/dFve9br60MjyWgx5gP29fdaSS4H/maMqucHkoCpTk3qmCiFmL6BD1Uwjn2nxgHWny842gL/Ugk0qz/cKJhIdK9bUPjrkyGBXrlqyeHxyZXfKEYbO4hECCDSbj4eg1ivVPwKqQOF9oQAUTvAM7XV5eQmOrOaRduabX8Ok0/Z7Tsm43Ij6qiKgBhl0rRv3t3N2z82Qn3DtYRCnr4leXwLCf9790GrotnMOsz/WlXILVv9d9VNzW2PO3PUgQKxdAmRC21zwGjDjRHQ+P7RJ7Q7S6eC5Z+Dl7NI8R1zXq+BwvtoOVCmaig3e1wjaUlF6gp/G53bzQ/sDjpAnGjrqHDQ/D2N3bmhp9LZ8M6vPPIj9EaPDEb7oGQUSJT8id97oaYfwjHJjH9y2Hue0CuMDzTKuqioRWLENdmza9rIT/OC3Mc5dmwDuoRwlZ492ZslXN+bLvhe62LZ3i0eUJERJYW/H1sbLvRs3QpCPt9oJ6YT7H5SISi4O30KqiZsWQ3QkX9cWFX1hQQ/jkYFeDR1ChiTUQk7igsvHos9N/2MLd1GIsFymVy3l1dusVPTQarfvYFF1I7bV+gdX0hXDxFJPNaD8MCjaAIBs3xU2WhCZ6LOGGDVxLPLiwisqQxY6qbbn3QCrUEbgvP2kQuYmE2KLN7dz2Ofvash0Y2VdA2D5C/aLqlkp/0TWp47sy3ubLsAcudmtNww5ofS0+5BVerBwUVH4Jk7dYQEtFHZzKdcXSrdTg96d4ck7o2th66glo66d+dgDeyux4EyqN1T5iWcN8ZlaOc49SUv5sPKJpkr+vhjq9supVJQIFBNvs1F87FJZ/jB18IAm9w2tfe+5+7Nvx8uuiey/1WuN+7NUd+PdjyZ+jrvmCSnx5KWgVddN/vO3/BrXfmt0zYlwt+XxydT4bwSurbAf/EmmpoaGx929/HPFBgF6fdav/Whbvh+Rr+LHNF37fPTz0I54fn9pWM13ROqKewNePvsIFQL4dBNRkizpaYK1D45yERGQvlpAhzzo3P1DaK4bvdyrHQf3ujH0lKY7upxbDA+gidMHb9zOm7w895TZm/Z+rW8NgfbT05/Pz8bHC3r6Q9zn9v01ERMq0bhNV0+MxksgntGMm8HoR6WhdaDgKV1aj2e1q+nRIsbkAeM9sqiM+65bk66xbSpnou73/GBcef/sUl/+65EIfvIBbchQeTQwzWnp1FN13gm22xd7ABGtcghKb0u4zd4xlTSNROaEiFBUBEMM2XgnKvdXwzX5xzwfGwHoTLu154bXjs2p7HfVcnHONdWw3KggV0fM8Pd8K5BnDbn5leG3426GAx7b+/OOtwx7946AJpmIOB53di0gXt3gtBYG2gcGkp739/1PbjC7nwzrMIz6zt+t+tDDuC55VQAJZfWHFB3cc6JALpzImQaK0ijs93U3tvuAbXxuy0G2Bz+r7eu+L1Qa/tLuMZXQGYkvzOOV+zVw5cidpervZ8H7w4c3/4+dW98F2DK4uI3IfCpgAfvlPs3y7rSPRzfgN7+SIQUnqvNEi68NAFBqkh/PrFQ6yexziOttCPZZhUNVgfFzJd+HrPLb8fnvqSiIhkkTp/4ZQvwD+qPSMiIruNs8NjxKubFRoB55xBQqg7C+GoeYVsDQKrCPilFSnRuoeCSGvFKjdQh4iCdV9UJ94dlNVC0YX+dsvd5p++9B9ERORze+eGx/7+B/7v4effePW7RERkquwbf27RkRT3qy60Ly0EobrV9PNTSLTUw3jyuXvDY7RyH2heoX7FlWkf8LYUk3VqQTEMk8X7sLjvf7LwleGxlw88dGcCaa3pQvAfnP+3w8+3O64AHhygGMeeBXH2Wa0mTSFRy3X0iyfCfNYGvvY+VX9m+Hl3FrUGGhC+fMa9LVq001GwTmdPu2CZyfrnn57/9PBzbRDWwSf33zE8ZvkHEZFtFT6smiZ00nJSMcKMXXikH7jsQne7Hd75czOei7i65waSva/2Q18b29uuIN7/3psiIvJq1YX3D5/wd/dC0fficjrM84Oev7un8w+Gn89mg8JOI9732daZ4ecni0Eh/95DB6r0WXwITzatCf8Eqik/ata3F4FEegDFp6FcIs8KmzCgJkaNxEHxOJH79kbkkEgZw2FBQfl981eGn89nw2ZYTE+M/EZEZCH1eRERSQHz/69vYDNdCRuISIc+cOERoGr5rfCZcEQiUzKb4VyEtJXu+rQ3LoXf9TLYoMD5d7AAKxojnc+7VdPp+7meyAXh0q64oljNepjjhRNhM7FQjRYrhf5S0S1OGw8eeGjhvAqyHPDV5CyxSkvOhcBL60/7decUmcEE3/W6oz2erIQkIIVBre+b8YGGPn5s4fN+fxkPjVDof9u8e3/D20IC+svbwVLnalvNu6ewoDUQLZSYzGT8fTw/4/dI69hGDsDtmoYZKVxXsn6taQiXmk7tXm98sY9ZtOkdn2PyF5XVOG4uwJB5lyu2s0VX/gwr2TgJb8mqireA9uqjijWtltrfPvXvh8dOZf38J9Lu9U6nMnoMdAgpf4auoim6iBO+t+BK43YvGBcTSKAv/6H/vnoJXrVxPOG1dJmMNi8Jxk0fC6F0P9xrcwW5EsJLx8Ft+8fhnbc30nGCD0bEE6ciIgMkhLqA2vy7RrACP1B0K7SSgqBVV7iQ8hdGeNrD53Rj4R0W7rpl11r2e2qe0XPAck0kkuwz3n3jTG/ku509lKwDD89Q0sZBCC1MZt3Ce6LsYY513QCf2vRQ1nejGnRfwyMslEkjCcmK3Fd2gpXG5GlU9zncVdf/5oYL5wKKebp3gkDol8e7t6wl2G+E+6LQLwJOa+/pWsdDBH9afWL4+X4tCP3urN/fzZ4rqN+89v7h5287cVtERNpQlvQQHt7QeDXw9nsrLmi/3A7X+oO954bHPrd51s+/6Erlyn64380DT5JOIk5uSvxh02PfVMJfzTvi7F43JINrXQ9xtUnkZugweCCdJV9nOwujwmcahtTdps/XgZac39j13NBqxYXy1m5IohPRxpDJp+5dFhGR5570MMwA5vUdhGr+fTfM93NQ6LOxz/0jvZfqwBXMes+NkzvtcI897P/1D2J/ob5g6OXQEOEz9MaQ1gHJY4WhhIqn8V1SjgxzkUdQ6B/p4qzSwqn48l//OyIi0ppTuBWy8XLOXeEnTzgkzOB0jMW+r3J9+PmP94JQNMEmIvIISaeBUj+kWwjZIM5e2PAF1p7RknSgBzKgZOhNhwVYQKFYC65wpIneLIqwOsCgkzmy+NWw4VPvBn0oxjuWwyb79BWP4z/3hG+81x+FDdauuuBgURtrHXIK9exBsPSZzNbFXJoDORYsJEt69fdheSJc1p914ZZSgcU4+vSCW6GGxpoEhPDajgukaUUwEZp5ecrXw8u7J4afLYdCsr4OwjuW5B4g73HmlMfsz00Gi/Va1b2Hh1dd8b34jhvDz9f1HjmHrJY2CG75EAitAQ5EPPn50rqHtQgeqK8FxcJaiMpNv27tnDJIwjDtT/qanT/hQt1i8mv/h4dR6qcB/zxV1+eC0YVEbkV5dOaAImN1/GrB1+/VWggbrZbgTeWAyW+FfXm16nP8/Jx7Rn90M6x1EhY2b7oSZ+W1hWU4RxyG6S9swtNGEn9Ct9LBCDdwGPldhID1Fhgh+HqOG7/4974xi7P65Vh23qsxb2Xmi4Fy4ca8tuGb8DvPBAFP7f8Hu88OP1vMcrfuFgd1X6yavIcYXdQYH7uzOB8x6CkwF/RLWlE4hkZXxGGOzFV0YDWlYJW0ntYYMBKutNQfNcJKm1/00EyCEVCF8uppd7WJgokhdFc0mWfIGRGRzZpbWLnp0ftmjHlT6QRSYN6UVeCzwbK5UAlCglht4tm7SoNwH9cnz47B9liQtd6aHPk7x+kpJHe7PgcbURCeFNR3bvvamr6s7wAIk+zTblxsNBxKekmTzY0e5hBWv1nq/YEr25NTLnxPVzw0Zwqtvu8GQ3kSXNaZMaELvxWvSsaeEQi/yTFIndUfvj38XIAH8kif8eE9h6JOQkmfmq6O/Ia5J1Ki2Lt5IP5uD5Cfu18PxxehNL6y7Ur87HwIRb1+zY9Fh0g1i+nTg+8Bsz95XcM3oFTpoBCro4YnjZcBErXdSZ/PiftaMLh19Cz9Iy30JYolrRZw6VIQDA1whBgZm4jI1IILujv14KoSP30P8eoptQyJv76z6xs7rRV9tO5TiLP3kMi1jdNC7D1R1q/WUBPMlxSEQ8gmXG0ifWg9p3Vjb627QCvD0t6qhs3EXAfDHFOVICSaqGlI2CGISZqwpzVJSGVehTaXdB38Qla9mCEvPu6L560ZnTGYUFlpbIyaZHqs1sHwqJ4LvRZCLhkqsrqHnZYr/K2aCyTD33MO6aUdqACnguKzcG5vVcPcL0zAmwL4wKCRxO6Tl2YTIQ2rdiZrKp/RQg5cR60nRukG+tnxYUgWfdn+oEGxBZz9plbUVq4Airrq92WY+oOU398UvLQeQj1WJFdGTD6PvMeyFoXVoZhpENh85/COjII8PAQq6XV5xUShYT77H9TPoCNJ4bPBqgcTMCJQOZ4G02+zp175mL4Tj3sc6fBO5dJy/I5//jcP/fud++7ik6PGNlHpj92iqD4DAT4ThBetb+LkbUryQJDQimy0R2PihI9ubbqJNTkThHIHwqBPulfjkCGzJqp3B91Ry6x4F7QCM7B4tSqYhFLpGhgDTyo0EsVliZgjilaG/EGk581TiCgEENYiYad+ASSN2b8Am8nYHImfLoDTxLhryFvTxsZOWziNcMT98dQI5trHIM3KohCrqwqE8xJB2Vm9RXYT9w9rMCJyRGGpCf545DiM6I25KYoIkpiZx0fMfwSKkqFHuAZD6LInZKtadDaur4SIJPsq6FpktSmVd+ZisLqbCBMyn2MKk7UzifoFPK89D0nvohtQdmpMZSskKYRxoJw93Sd8bRCJ10JtSVHXDxX6EiqR91qFkXtljYd5JVT4/Ey6j4PrQTEeltP6qx53f/q//sYM78Tik2phCJa85yAYGAfPKTKk9h3+3UVwd+xoYxBCF9ksxBYFrU1yptMCMstxMg9Lhl2hxnTrSUOOd3WDsYlE6wSJyeKRz30Qj02iFH9vT8mrEKuNZyD0VZCzd0AC3gbX3xrJsOimvwdBqhZjGlZTF4LWXODsabdykzh88MbsKEQWgrYFCuCZcyHMQYu6VwRb5Z3w3f5JxJ1h9VMhmzdBhd+E91i8p3DEJ33tMOk8vRzCL7WKC5MJ1liw8Y6S4VWW/R1tVN0QMYK7GLFxKtY8CofMUIgBSCMral8RZSlci30CShPh+LhuXSIi5VPuKZtQplLoliBobS2TxRNK3Oo9OMeViUMKvUpqqYPiI/dOf3dVDcHOVfwYLf0dzeczH5QCCWEahoxBgulRVpusRdDnO4SF1vZ9IePn5HweZH3N7k+MSXYfkXGkhX63l5b1nRDKsETNxr5vms62L+r50x6jNQHL8A0TZBaPXdtHwgcCb3AnLITdqfEvn7S/jWLYGNtZkJWxIORGuN/uAjJoXAcqdClkyO1BoWqLkXj2OgTW2ZUQh7+HSs1JELbZxmWfAbbEY9iopMqMlZppuMILimNnV6g8qi4fXQ9eWHvNlWmLzSnIvb/UTDyfSFJ4lvR56ZPyvmvPhTlamfbrU2FTSBg5GvmLavhu8x3hXBNkcoQ1Z6GaadBfMGdQnvXzWuiigYYwJPkr63yRlmB9zesaLqw4vt9QVOt10E8g97OjpHEUfulN3x91DUmyKI71IoOy34N5tWWgyCjcHtwO75bcVz2EUVZWggKp5lygLk74XkxYxzpHhP52QGVtjZAmsmCTxfucLoa1c2fDw5hUQLTq9zUkyFqGOubAPEkaF1t7Lm/SY7rkUYGQEmViyZXUURtHWujLIBq6f814tG2agHunCku9rFZFMwJeF9WF+7eCUCQePp5CnF4Pl+4BhQNmzR5jeorDb8G6tiYsIiBtg6BPAd1jfXrjMZhukWTTi1gLuSgQKQmN44Wl4ey0lL0VhEDueSQxMW+Mf1o4gKEgxoPXtHGIVUqLiOyxQYiG0OLd8ZQPDJnYONhD+Tw25oYEQVEj35sAACAASURBVEeLu5sAWIdx54EnFkl5SyZEg3vXkBOgMjHlT8XNQrLbmiMhL00TqCc2Z7GQYTrR4Wo0NMBWnwLFeHPDn6er4aw8wiAd0oFbIhdznCgYzGtO4D5Cg9Pj5/NgO6wJhlR438az38eeSm/7ea+2ArylvOKC/sorXjAYn3QDZ1Kb6TSQ62BjoGEhGDiiSrgvowtJgXKaqLsItla8oKHYPTZZ8ffYlnCOdgZhL4TYjHunyZAnqKwbS6PhsmNL/+2OfjRMCA40qUoGPDYr6SHOvq8vJ33PN2N3CWX3FtMnHhex0oFuoNbS+HxHIjmjayKPGC/5uiduh/tvnWKVH4pPtDUiPY1Ew2s2VFkPi4oNRBIUxeridwSLb8E3iDWCiV8FCgZz2D+BOdKFTXIqJs6taK6/48cy+xBoGsvkoi8+8r83ngZrqYYpKKgTYSWdgy5qGUhtMG6w53GiSYom1AcU6nxGxXWTXIt9iNsqHPP3IHDPInSBpWG0vANYgwzzGSiBrT6p/Pk7a5XJXrYks7NYfmETVdMLvo4y2gOijwrUy//MsfGv/TdeA2GKhe+ACU/Le2RZpQ42VUPHdF51GHQPeQ8B9HdfC/4GAE10QWJmezgGeKK57d5OcSM878FZeDAwDLlXIv0K54B/z1QNwEHSIv8ougxSwP7z9wJ48pl/E55n5zKbPx6NcaSFfjTwBJKFPPjChFA8llYb3zf58FFFG+u+iZh45Ca3a4xh1RMR6cGqN6TPgHSuEKR17ZwVQXj3YWFZwjRbQTKxNj4JOVR4CVItLGAliGNxCbn5Bzll9HQDMjnY5UrdeUJG+2A67BkZHbjJe5Oj8WLSUDeeRKcm9uNVQdqFB0NkiTFednAsj6SwEWwxHs4GO8wfGMojnx+PgomVSI7tGhPNXQww8BT62m64t5Rb9HBaW2PirLUoIE5vVBb5CbSBROvExPozz4KHEHLp6po8YOIQH/ujjePk6kcAOEeF8ZAPKj/eQzGF24W3Q8TZkFa7BPECbh+rUhcRiU+r1Q/jIebcn9X5BMfNAF75QUELqqCAEoYhsQc2tSyzJTOmVTVDVrCuoVseNQITTVrq/ru19ynlwxgmgcc9jrTQj1Mife2IY2ES8mZkDlAcgk5GYpj6idGQjYgM48kkTqJlN6R+QDyQNLQc1p6O+5MvP7+l2N8LKN7CohpYe74pKg0sOhaIGSaYdLBMxKp0YuvFLgp8hjTV5IAlIygEeEqL4DoQaDF7BqgyGRxCRDes3mUHK5y/jdCbFYLFMKvyYxqY95AopkWcWVMY5QmEI2CdMxRj98jkbSK8Y3PHeUGBT6yNutsQPCmQnLEvQt/gxojjE7U0MxPCHywOSwDtSW1sihwCsZdjNyv9pxuNHAs3o/+yBgSCMkLIMTVrjWQAQUQ/3qxat2znTiBFSXvjdpBoTiogGD1qzM284t+tPg1lpsI+AwSUIDzU62v9ziEdxxKelzWagSGTPmAodlSx0sjM1keNQPYnIHeAebWDo2fof22hH0XRZRH5XRw6LyL/SER+S4+fFZHbIvJjcRzvRlEUicivi8gPiUhDRH4ijuMv6rk+JCL/UM/zy3Ecf+xNL56OpW/JVJ3b7A4sOLwQco6YpmdBVekhWh/OW6gI1Ys3+HcNXZQQWqFTkB59+Vx0PbiHVqXHuHDidypQOuS1x7XIn2Iar8/+rsgfdM2bhqWf3eZ8aSK3Pd4qYqWiufYsvqIysvxAZw6/b1Ng6S3zEN4HnTQj/kr0/oVn1NMK5kQXJNQEdJdV6DJEhmsRdtrbVVptKKsUFKd5lrl9CES8j/ZM+F1hE12n0GieVNcGQSXvPQ0N460nBDHCvUY7fq5hiImtH0GLYQqC73AAWGpKY/0sbqR0s8pxEYelMnfESuK2hkKZoxEo2cZeeF/pXTwr9mqM9Wm5gOpz49dRyphlsQ8668ih6HnTdXqcmMP+qJJDLVyiuMpkzADGTZRokh6uxUY1adTc9EERY8Zp9gjmc7+m0I/j+KqIvCgiEkVRWkQeiMj/JiK/ICJ/FMfxr0ZR9Av63z8vIj8oIhf1f+8Rkf9JRN4TRdGsiPxjEXmXBOPqpSiKPhnH8a4cNgaRb17VzgkUDCQK3ct5RawQWTI4iY2tAqWPuG8X7m93XjcANkhqH8lAWmDWTBnhBPZkrV/UBcTWdhC0Zsn3C6MCUyRpKVjZPMNSg9OIw1sSkn1vV2D9aihpADeVQiZzD9QEqiDYfo+9gfujdPYSo09wyoQiw09MlkPQWiiGG4wtJW2+IrShzNxGviatcw8SN2LY+/TSTMHAImYz8TgbjneXD+mjqr1zyXsjZGKcHA0ZdtkzGbwtQ1peCE8qoPSyh4pi8xCQe0qEevR38Rw8XiJ5VGh2FyDQ9rnQMB9qfVOod+ldjot60tW16OgcqDYg6NljIaUJfaKxSKnQVg+DCfK4MRp668eoiIcgJj3K0MOgd8wmKdZnmDULiBZkNSzFOH4Pa5JGS8vqQQ7pLvY4x9sN73yviNyI4/hOFEU/IiLfpcc/JiJ/IkHo/4iI/FYcqr4+G0XRdBRFK/rdP4zjeEdEJIqiPxSRHxCR3zn0aul4yHxnMV7CGSWBffdFsacdppiIyqBgo2cCGL9PNAA3oc6kGiyCCDDKxP288ff2DJK0PBl6KN0I91gDIoiWUG8aikevS2KnPulzNcFkEEiRZJxcxngbORZMLVHohs/lGyCqAzWynNBCF+K+idu2+WLt12ugwj6PgiZVwsMyeUlaa0Mvj6gO5EXySm/bptBfR9IX2PVBbUw4gARc+p4S7wt/t6Qzq3wP5mFQPEL4xQwUhsB6Y5Q7z48iw4VJF/pW47DZd6RPgi9fPd04gVfHmp/U2DcNhoy/Wya7bZhnKPKGdW49JLBOc49QrHY6zA3RVqw65n1Pl8Mz9g+pcK4/CM87ecZtw/Y9wCiXwzrsZkZDXSPDFBP5ePhurBHNDIwX7OWusUsTkUOSNYaNzKh4c7zBYxlvV+j/uLiQXorjeE1EJI7jtSiKjBFpVUTu4Tf39dhhxw8fscenY8TvbURwX6cm0HxaF5AxPYqIyEmf/emZ4HMdIESQf921e21K4+zd0WuKSCJLb303B/S08aLNUieMrAdrrzkGIURuj3gRuG4tJiNOmFWAvXL4XQkY9L0+WtpZgQ6tJly3hwKcWD2PxvP+MIMDv9ZkiRHdMFqI+4q1kcSz1C9BKHdHlVwfNLURLNrcQhAMxPH3gYfvaoK5iIRoDs0v9sFXk9LCNCa4O+wIpo2yU2la+vAo1Xhg0+/CvNcHDJAkt1g9FSMx6pbwbCH8QyEz7ndMcLOLnAnS0i0/V/PEqCWfvQ9lCG+Kyt+ucABvhhW3rU0N3yAe3pkfrcHIHYKwqqCg0NqcVnd9rxKSaVXipK/InfSYyTBpTMQbvLgUvBUDFSRglNgABp8mBTjDRvFU+D37XRNanIAk61fSR5CG4S0L/SiKciLywyLyi1/rq2OOxW9y/I3X+bCIfFhEJDM5IxPXlfVQ30MivvqMLyr21ZyeCovife/3TkokvTLirjKKeTYm3XqwODhfWAcJm2xtVBkwv5CBgG/PKnkbIGvZDUA2zaLFZmeT9SysJYP4NQDJLKPSMadW4s66Q+WeuuCwvCu3A6toedoVJDtzMUac0sKcFHHnqOTd3wqbNIuNTSqLIS0GnpV5PdIg2GaJYWxm5/weTdgvomS+lvM52NKNx8pbI/0SEVnHceONmZrxc23SI1RhPhiMCnoRp+Am2Rgtv/mTfl2raKWgZom/KR5avsSo51GlGqsCYW4oQT6oc9ByMsqhByTiSDau08ReQoLZ+jVPTPoLYXVupmaesv8+C6VgFv4CqmhL2dG6DBE30Eg0xzF3Ipzj4YGvacJejYl0/Q56/EKqZYHJtxAqATUkR8wcqAEH2yWNv8cZhRbfg3GEOeDvjD7oGz2884Mi8sU4jo2z9lEURStq5a+IiJUP3heRU/jdSRF5qMe/6w3H/+SNF4nj+KMi8lERkcLqqXi4WHVflu/5LDex2U4s+ma7PB1uZT4/vgpwJhc2Lulas/5V6WmCl7Sosy/7tQ6c0E/6en/M7Ccy/ho37c1w0SNpqzFD8nL3t32D9Q/Q01QpbZvgvc9AGZ2bDhW5RVqh6dG4KkncilNAaAz8Wha+qFcBIYSLX5wPljRd8UQlsqGS4ErTahJY6sXZg5H7YlXlRD7MHTuGkf3T2CbPzTh7KEcF1MXtudHMWmXWjxmLJfscH2y4J/Cd3xKoky+hz+t9dOO6g7abU1OjoQt+toIowkcz8DDYXKWpKJX9pt9LB3MkWn1b2Mb7oG0SGU7fD9HSp/GQ1e5hrNymsmqfU8V40wUt2y0WlW+KVdEkVKMBdkEpsJt9xOkxiip1L2OdXtnzmgJrvs7kanYPe7EA6O/mqFDnvjf66fwO3hHCvpbf4ry1FgAj3gGiTJdUx3XVkRlvR+j/55KMv39SRD4kIr+q//4+jv9sFEUfl5DI3VPF8Aci8itRFNmu+H75Gl5DnHb+fEOc7F+EpQOh/3DDNx5pWG1kUOa9rU3S6yjoas9BOLVGERwNX2cS0xLQ5CZpmBOhHG2XNoDATNHi1TBHn8kjFlwRf6xCMQest3Gfi3gXra398R3DzKKkFVtBmCa6BTpi5eyZhkCsrvnfTdgzl8LR2dAKaiR/QXcvHYZ39FykW94m74s+dwfFShsbvptyiuNngxFSH5A3qXE1rJPKC863T4HWSGsxEO6V8W6jqj6A4NoGQ+XclM9XXoVebQxBn4jzzeyiofuJs87dT4phW7MHD7wwKbvg3tBAv9pBcS/DjGaVk+qXCoDEYvdVcXGdnF5y8rYtJQDqEmUDo8XQSvW8P/dCAXxRaAJkFNjshZDo5qbtJQugUUnju6+/EvoLxLMEJIBaoYq8XDb5r4hI/Sw8WV0GLSTxx0Fga6g3SSDSMtj3qg9nXj+EyuUxjrfEshlFUUlCPP58HMd7emxORH5PRE6LyF0R+dE4jncUsvkbEpK0DRH5yTiOv6C/+Vsi8kt62n8ax/G/erPr5s+vxid++WdExAV8okIV4Yips27pW9EK+VEoBHZrJf0eErmv+2Yyt5eLg24a3UODPBLlMhq08mrZN37X4ohMDjNnwGIfE9p8ZR0UrViuorrjFtjElAuGYbNzCJ46SLloaRtBF/HsTBpPz4VN3IZA5XzaO2i/7MqYypJ1FVYEFJH1FHvN+FNmkLdJ9AGwex6D7RdJWtdGUXyA5yJaxKznIki76M0Yv8oMunyRtpt0Bm2llaiAaz49pling9g58y1l5GbsGbZB2NYnRFXpADLwOBNrVtf0OINFRKR11q81oVZ1wnMb05Ogc8+NCwuNiIhkng7cO1TiVHZdrLOKenFsasP5Nprn7iFslhbS2/2SU6MzvMMmKqYraJR1wVJrAAvmoWQKuQ6TPYmiHKxZ7OvilfDu+4CHfj3HtX/0d//jWDbjOG6IyNwbjm1LQPO88buxiPzMIef5lyLyL9/KNUVEpB8NuV9M4xa2gKc/A8GBzfLOReaLwyBH95/HZ0REZGfPF213BphmhQ6ymjTCQkzgiK28nU2R8fIzilWmq53g5tfYdyIpNwl2RQgEI0/beeBCZhWWoUFUiVSigDe+mT7gpbPLHkvdg4dgsf5EcgqPmFkM5+304PYjITlMsAEiyPBOCp5NT1EirDnooOw+1hxL5wXQAr/uc9BTvpoDzNvqghsB9+57yMWstWwZlA/3XZCaQm6DwqALiKAps0fwNKLt8aGJzHIQWL0v+PVrp6CYbDpIKY1cRjRG6BMFQ+rlgc5BJ491yjWrHiNrNNqgCylgPqx7WOoBKrDZQ8JoTkgRDjRVR3H6EXrQz5fdA5rAvjILfxLCkZ66/X2zhZwb1rR5WVRwvcroXhbx4ix65Wwkk9YaoERNAJ5BxnTcSiHpS+hvQ2lXokM4tR7nONIVuZKSIc+8WShtV+hSghVbzPpmKmvjZWuYLSKymPeFdGkmxPw//egiroXYn1bTsRMOQxOJWKmhOxmzGfN3JmfpMhrPTjwG5ywi0t9F8whFWMyuukDrY1GV1Woi/XQRNLCvboWNs3DCf894995XXa+ntctVn9Y7FviOErlNI5xBVIYpoC4qNVPYNL0xENgcmAkHoEbIGM0zexIsAh+tOHwmXBmyIb2ueQ3xNRciqXOo8FSFyCIqdiJ7dj50ybpV8rm6K87wSNqKIcT0Rf99Fp7AuAY6/My2gNY97FOvef9jWpwzL4X7La/7Olt/D5OYtlD957ld4PhBS53VeeygxiMLqKfd96DhkppV6AOlsqB1z89kzLRc22TGvXJScPd0M12edMbRq/ueizOjpo4OWKRkKGwABqz5NyoI7mVrmgTAWxKdY0oWnEeDQ/rtmrGX3T8W+n/pYXwdjNFxoW6CcvlP++dFROS5OW9j95n1c8PPQ1QEMe7kkCmP2SBIDtF9tLhoKlHsg/CNHmalJIu3rICGFl4LiVo2VzFLmmyUtJgLp4Jio9eTEH4KFWUTmD2wTWbPehhiWC2KBuLREojBdIGT0YHFcDYyqAgmWiQCqZZBUVm5OgNlYv0JSAVAcraW3itpdK3blojIyXlXchavTj3h5yevjFFZ5w5B0axpS8r1qocDlxbdW6ojTFF/FNZkGoyi8+jrYGEdhjOqW76OP/Pw7PCzhTlIfMYmP/vfEa5bg3XO8M1wLZOhAO+D/EP5QrjfuTk0GKmjf7LOzQAMrW2ESSxPREQQw6ukI7Y+FDl44j1QK1uHs8+/7vs3jVoGe0KGR7MAQrRnff1bjo4eK/m5hrlkllJUUVVspHCs9uZgLs5e09GT+Udb6EedSHIP1WLUhG4TiBpySBnvtojIty8EhEU+8oXw1FlXAFtafvu/o0duC7C9ATauDSu/F3lDC0BdB/3i+Ji+fTNTo1LxvxdWgvBh0mwCkEqO83Oj6JQauHVm8iH8w97ATJp9Zj9sHHYBW6iMR8TUzIrDfc1CEBdUEDLWatzmIiLXb2klC4qsig9RwAOryIQ9Q1mkyjbumqeX14fHrm4Cm6gK24SsiMh7nrs+/Mz+rCVFqTC2TmViAovXZxT+mamwjroQTPst0AIQ8dIw2gx/7moGFMCqnNlKc3XVE6bft/JVf0S1ND6x8+LwWKJdol43T/QOZFNWY+5s9N2dckG9NDequNgMiH0A9nRuMhRyJNHU++J6eXrGE+cchkpiC8UTJb+Xs+Ww5r97+fXhsYfw4F/bDevs0Y5nsJPKjuib8C+NtiI8gdZs+F35PgAcqKPpqTcPsZKgaWbYyLzavL/OIzOOttAfOBTSCioIsdov+yb/gTNXhp9XssGyK6WAHcZmvNkMMaJ54Ijvr7lgmP+L8N3qRbzQQ4S6rVVaTfy7wbjq6FeaYgJtXfHuQGIcrPu9VE54WMqQI+zcswBYqm0AWpuMpRbVOn5+3sMGB4DKXTlYGn62hKSxiIokC2RmVDGVMK+LRb/XuafCdT/38oXhsYznpKXyZRd0+xdJVRVGItqlAicHCCATvTklCDu34EpxAgDrLzw6Pfw8bDeI37MHrlnfF9HAhHkRyw2dr/i1Xmqc9HuF1V44o20F16Hl0ay8YxQDSAayA9t219fB7YMQTlpGo5q1bVAXKxldGqEJGiJ2W5km/w4kEea2prDVAlorMl8Tawe1sZ6EyDBUOo1ktwl3EZHpnB8/WQyVtllsoKm0/32nF+agMfC1d33f26RaKKiPvg9p1HbFQNQM5wBJZyoAWxM8RgPPvH0i/Qqb8BQQykl3wvHSxnh02+McR1rox5GHTywm2UZiMIUs+8tVB8/fyIRFwdjgl66fGX4+fzps6Edw0VkstP5B/UCODrInriDjr7wqjCOShMkqbonY6bPQSzmDEg3QgRgwjLqISEtDLXf33NKh1b5+JwgGhorqbVcAK1Mhtnyz5vHohzsuOGYn0cDb0C2gMOAwyyyF3b7R9Pm0sBK5ZurnWcniH61Klj0FmDSb/WgQml/60PjK1f79IOluX3UheRMVuQsQlOv74dnzMwinoddrWhvNsxH4PGCx//beUyKSzC9QAVkDEhGgdlDU1sa1LAacwt9vbLhAs2I6EZHLZ4KX00/URcA7vK1srs8CxglKBssdscNV6pGvjUd7/u6M9jqxJhEynPxSOL77LPYB5qCoYbarr3vB/dJpN3mZR7qbDuG27SaaIAE+WlFPdQfeGr3LmtYtkFiw/iQAHmjuMkzwEoWJsCxJF220Fv0ZLT9A9M/BKjwBsIdaXU+PnFpHZBzpxuj506filZ//OREBVwrCAmlQ2jJ2bYUmbIlH63ijHoQIu/WQo2aiEn7P8vg0NksPybiekVM1xuvP9IzxkIBoi4VHGs9OuOoQaDNod/hoTSuJZ/1YHULG4tF90CVML47WLCQqGgHZnCCfjN0PdjO5byxvUAPFQakMygid+z0Ul3HeSIMQXwnvI3rKvZZxDa1ZlMaYvvHh8Px0FYhmMkFJuGOmwESvhlyQN8niWrOTowVXu0CBkZPHwhxM0pfRl9kG6StYqEXv1GL9s0gq15FIlWva6ey8C33OsfFQJRoHsWk9IKoGW+1xjpDI7T5SdA4KDvmMlsxOQD5BOjez5M9g72N2YpRnSMShnDsbHr7h+jcPpAvAQzpBdT2KkyedRx1emOXdcrvji7OG/SqQC4xh1NBItA5/CSj313Hc+nt//xuzMXqUHUhuKbxgE5TF0ii0TESkv+UbYLcSFkJty1EVPXb20fdAQiou6mEcnoRRsHT6UDxZ9UAGbDiBCs6CbiaGCBi7PtAE2QBcM2kkOR8BblhSK7S+6UJm6mX/3d6zel9YaHt33ZI3QrM+rHc+tzVWF/GKVDbqkC/4uZqTasmfdCHWvuYb80AhfolGNTB6qEC6TyguHIo3C9ZUE0LsadAlfbSdixTFQBoRhWX0EBRIpDb2zDv+jve1cTN4CmRfZLORxgHyQXoNdnBrX0AS0uCjEEJU/uwUZh4TYbUZVF4PLgRvpIc9wf7Iba3n4HPTQ6lv+XkXPx3mef8sCrmedoVsQIQ8Eqo53IsZAjReSGBXvYvaDVUQ+xOo/GaHNpWjlbuw7p/2rw7Zb4vj9yr3sM0tjzEsldEQa/oElCWa2hj/EA2SCXiMRCjVSuF3YwkZH/M40kI/nR7ItPbQPFDECUvP81/1F8IquoqyOdYQK45YHKUFG+z/GgOn39VNfhgtAJti2JphEVMKlLgm7A+wmclnEz/QzQZaYjZupsV66lfCvzd/1O+ldh5xWW3ZSERCgh7COoYBVx6TwhiN4A2qSeHWexI8P0ptHBPLzaTtbe0dfBmVktnx3o481HMt+bz2NkGSZthylNRbXFlEpF3U/ANcebbfK1337zbOaIJtHZxAOG93Jjxvfg1/R2zc4sEthAPZZWvqij/X3nMatgLvUv6Kh1Fa2p4yDYQI31cPocz87bC+yC7ayaJxifajzQD6uF8bVeIJ6xvrLI0m55vfGq6RBslhj4iy7Kj12iXVtCLKaP2T2iAPMvW9Z7WBDnops1+11Ym0EEfPs92oVewjRMbrcg8bKonGYgZewZR6YaTCIBLOPJDpJfeeWVjHaENs1z2k+97jHEda6Pf7KakqpNBiuB28MEHf2RRoeevnx5Q+08JRN468270eGmUrzpghAIGVWUNRTF4pEVgok2LlqcLTuozl0sXWBU63e4CK3Cz6fV792XCOqE7+d3+wjmGVcS/tIqxc3dgRhSsgn2T3HOh32HaQC7ht/XSJMQeao7Wsc8wYMhtxcD5m9Fy4VmF51EoVCAYjhBNBsdmqW11sudd+1s9lTxjNunDs0zr+chBu+y/4+fluWuaRwYjIQmA23u8W8YR6OQdbCME968iUSKkmiheQSwGSh2u2peFNxq67yP00bwdlMphB8RebnCjraYphNazpwfwoR1MEpSPksNcwRoueCK6V21EljGdtlGGgkZlrDJFt4bZfq7Ua7qu3iufC/ohUaPeRv+gTHgrIM5/dRiIEpgKc4bocwm2GsGMTpR6EfhNennkgLNg6KuNIC31ppSS+GmJuFlGZv+F/3n4RMcmzrn1NuzN2zdheXSsGi6sgG7vpoYlhWz6+L8bmsMDaxi1ODg4kIRuRMi1WXEGl18G5rnQE5OZhInhwyu/RKJW7cE8Z7hr+HsKVeYmMVpm2sIELQA21Cr5ojYahncEGwzVMgHeqcMUp4G0ztsfPS4o9CdTizCDezRL/8vkgPCYWXIjUQDzW6I/i9CeAeqIVWtTEOGm10+SY/0BAfkUQvkwa5+a1aA1W8uxZF9rMYVjnqYl5VzodGBJGIXyw57+pTCNfA2y8MWoWnvKagzSSya2OMUD6u+3MjvJUDVB3kYGyKp/wOLsZKrzX5RU3zw0wwMKlRO3JU+EZyHpaQhx/a933mtUdxAgVDZZARaGhO9JblGagJPU98l4JHkh4NnYOxN5b8Pb7Oh8DJGS7yHlZ8xTS8bCzXQpw8spNTfo+Jmrlu2/ytyMt9OO0SFdjx0Y8tvsMXChUxqXGVP9VkMh9sOaxceMWSSykB4gZXtJz0QrBms6xZaPF9PHCSZhmFhYT0IlmC2qJUDiyepch9fZWEA4Ty6N84iKu7AYJznb8Xi1mWt+JPD6s135Zz4WNndqAQNH2ehE2O4WnCcVExzE0ORm0R008VsGm6I2o0G6nQP+LTW5UvkVYZXz399f93cdjyuJ7SDLWDfqIgqrMV91Sz74zCL9z866Avvqam67zZx2lYsqmRn4jvuf1cN65C/6bfQoZ3JchgRr4O4dVgLaXfU+UbrtAa5weDcn0ZpBXQcFU56qSoF10xblDEj871SFJzP6tYKgVnnfY6+ZrXkp/9nmvmVmvhmvNAj49DnLMxPnariuNzr4aVTCUuL8KG1CMY7zTDndGLQAAIABJREFURMcwY4ymoObHMdw9JAclT39LH/dxce+82TjSQl+ieIjaMQY7vpAIbncNhTnLCg8jvO3UCd9YO8rX0UZCpn0JFX1ant5dHE/g1ZnHf9j6OQQGNhSqbS5KfFc7PEV8E1i0EWOpSu52sOEbkDzm0XUtmnnOrcEOLE+nEYTbD1ZGLtq29qVlT1Y2tDYlRlHS38e96iUYfmLXKL67YfwTyezsEkIy6lUkEoDsfKXn7SPX8hAY9sUFtzK3Xwm7MUZzbdJLVz4ZwiQb70OM+ZJ/t10Nc9wE2oReBfmc8q+E75be7VZyLwfuKLXet0GQx0YwXSaINWfE5i69R+iFUBgV6tYLWsTrLRJkf1hn3dd8IXSXw7qPEA9nD2erEifhISHPRmew/sCV7eozrgDISro0FeaO6Lop1CpYEdwOPDt6dB1rSwhPevolX4d7T/p3Z74U7nfvEuYAxlpf+YsiKLAeutxZ/UEfZReJRksEBxhUMzX6Xh73ONpCfxANk0kDW9SsB8HLOXPO+c2/df6OiHgVo0gyybJenkz8KyJyd82RPtao26gfRET6FQip/GhskDHNPEnhTozG9AZYSFm10PqIxdI6Jk/4QHv3suyf48z3BKK5O/u+2ZbOeIz5+p1QfFVaBAUBFGOzCHd+NQiqrV1g79fQU/iUduFijBghm4E1r+aegP5JFEdpmCM1NTpXIm7tffd7rw2P1XouBG7sBS3MYqClFRfERkssItJ/WimjUT18Z8Pf/c4P6PF9F/oTCDv9oBYBElb45ZQXZ62AwC6vBXm8/lbTJUanFM5LITfAxJw/5WR6Rhv92RtOR5BF3qO7oZMLoW4QQxEZTwcAgbXybre+80pjvNeG11GGQaBhpYMdV8ITp1yxLmuoda7g62yxMAodFhHZ6YS5WSn47/lup7Wirz3te+IrVcf/5+eCMbdW83W6fxIKG8po9x2qrEgiCCMyr+GdNruAEd2jCXcWtZFziAl/q9Sdvvp4Yvp33uRvR1voi0O2zLKLyZCA0AOpWUtajflk0RdyH9LHqgOLYGW896oT5ue1HVu/AosdFMZFtHNrapKOVhMLyHLT4bsdJL3IwW2NWlIIjWQYZ6Shr52MCuCCqeT8XoxUrlECjzkqdrfnlTmTfDywRDrAiy9oQRJd7SYgfEVFQjDeTQVyoPeQcJXZ9BsxXLP2WHTDcMO75kKEsoxSS3LNW1+Eb1/0hE8JFbnkvjfCtKnseFK6m9shXs26ijlY9Su54EUNQBtQRlJ5t+cC3gQ1ueTJC2+FR+SaoaHyfPn+8LNVpN6c98K6KniTRNcy38dhHbuGf7/p9zr/jN/jyVJ4xjRimk1UxN7Qiti7Z5FURt7k4mQwwLqYI55rMecK4OlSqA4vRP6+cqjObSnTYgs85+UMWjfq8Rub7n4zbFXAXrVE7AQgy9Wqe2YtaxfaHx/eMabRhCcAYzABdFAZsD6mHerXZfz24X862kI/HXs3+8iP2WBFbhmVq1dqQYBPgcPmbtutuYVcWOCv7KEF1rh3w96oJRbwoCPQUjgXuXOYxS+pcEy/7BZea4k1Afo7YJMJuaT7addg9WQKjdO/uBViyxvgXB8sI6SiG78JwrUsniXeRNOOOW3agYRm+iGIrDSMzZxAIqmsx2Pip9Fakd3B7naCIKPiY7HQg2awKMmhQyvUnutu09/xUt4tR3Zqqquy2IX1vXHg83Wwo/1f4bVwvm/OhPDQS1sexye3fh1cSBZiWgXhG5W03UsGQm7twL3Pp8tutDxshzmgh9FH56ohKglJZSoAa8PI5Kq84Pf96rpX/4raPxV0u2ojAXwb9S82cnh3660wX1PgfXrU8+cio+ZrDexBHfPgWrnTnNPfoAc2DDhT/tx/mcoo4ibcY5gDdtHLsnDP/g44Mft3DEOaBHUUfH1HCfy/hqMPjp6IPXp3hBH1IsltKW3unsGl/O/p93is9MGrzhvz/HcE62ENvcpOg/nokw+fFxGRu+ugxCXxpSZMydwXI3HYQXbG4J/ZKhYSnqGqzanzz4NDB/08++qCE2c8gECcuOPXrWvBB6l+SZK2OKFYbTwMOxIZbjvRFJx0Aou+Gc3Cn5/z+95k/FITkjlUq7KP6vCcSLBJCZ7CCd/YloAm6oj0Ey/9SaATvvj+28NjibCTPu+Dgr/vaset4Ad7fryk513f9GPLQAWZDua9NDehIFbCdZ+Y8tDLlW33EtuweI1wjEnIfSgrS0iWS8gp4LnX0QbLwj70popPujKpbeqa2ht9ByIiFiWMkAuZ/Q++dmb/hhOiffleCFeRnTSfJmhCKVEQ558ArfYXbwWuoydWPeR6csLPdbvp3op1yaLi2wN82pT3PkI+xkMkItLW3sF8X51EbgmerN1vZ1SQi4jMvRTe3f4T/mc2Upp7We/vCRhSzAkjfzWhLT06U8c4/bc3Bo7aMZZLZs5Z6zZ9EYlajRNO5/y7n616LNRIrS6d9IV+fRPgYbNSgZ4geoeQSrs/MuzNvOLX3VZuDtI80HpPafETC0oIf2ugccj8dBAis0UPN1CgnayEjcX2fAyZmLAvY4OyuKQNvHpRkTYJumSs31VNjJN2OIuWlMa7wkSurPtmbLmOHoZSBtDoDJOsfGuweEsZn4sTpEv+ajjZ3ElvAk/FRwVifRdeOOOhkzt7ngOpaA9dQl0ryA8sa+z5es3RKDnMAefDcgy3/sDXXu69QPdoX9npORf6O0hyVrsu/Cyc1Mc6qaGK1oQb6UAokIwZso/E/fY70Y8Xa6qi4Q9a9x28G6tI55ql97ikSnQKxGqvVzFfSEafrYT52G2DPA5hISNnm0Y4jnxPE+qNNLd9rqLe+OSqDebc6FXXzusx7GXy4e88a+c/JGTDCmfl+Iv6x4nctzVSPZGCGgsHKpPnXwZ64b2+mSfhNu+o6063/vSEbzaLm+5xU7GZuXZwYuMUsu3ld5Hx105FXcT/q08iofNQK4nRwzMPbv52WbtGwWKOYKlE5VEE0VTeN8DCslvMKb3hJw5pEN5Rq4gNZ9hjd38DbRaV9GoHFMO06k24JaoQiZbSxPrW/+thgxbK22PkLYqLlh/wCWdYyZT0iSLof7s+R5Uz4fhc3p+FIQRyta81gnCh4MnCSrTwB1svcr6spysbzs8AlpdoIK7zvfg9rowY3jF0GX9DDprZnH82Q+bEpHt5DFPU1oMHws5YZH6dvBn+bc4DxnkBVeBsxqOClPdFb8X6D6R38A7/HFXJP7ozcs45KJXzZfeSLL9mtOAiSe/UFAAT3JZzEEEOBBZ9GnTH7BM8rokVewYbqSmPldbgKajVnqv6sblXfc09/HbfP7la+E6vdGzpv60xyIi01Qiz8vftZxDHx0KkxfvD54IfdrvhbuD5oruaRq38qAGWTcIkdbPQq+AGai6OJnrA4pwopbe/sxCGLuPw94i/Fu8jZrni5zJB3IEFxo11v6ZNv8FiWEf4Z7UyKhzptl+NPb5aVYjcApp+bICz/ObNYF2fOuMbmF7FykQQTg/QdlAwh6w07s8pYRqLfVB9W1tWfndYm/fxvo1T5cquh1kosOaL/gw2Hw+r/iwFCPieJpDZfYzWt83XKYQrdpHK4PuoabdyNnSvttzQqO2G856e9nO9vuZ9AiyBLeKW7qtb/oxUTOlXFKeP9n4kC2vOq4IhizUgtLxvyyswV9HE3FsOhfmc3RdRfKgMpVcf+Hr6T7/ly/53uM32XGstuCCI2ZsCSOQU6sjdKJ13Coi6FAAJRI+ZMGefgc4kcma65LrwxNvotFm+F35fPw3Awirxm3yE8B9zr46HfT/O8ZaEfhRF0yLymyLyrASH6W+JyFUR+V0ROSsit0Xkx+I43tXG6L8uIj8koTH6T8Rx/EU9z4dE5B/qaX85juOPvdl144wnPTNKs0B3Kwtr7YNnrg4/WyKojMx9G5LWrDVuRsbvMxrWoWVADm3yiJuyaM+Np0bIqNdAOBetMfPBWc49QMa/gITmsgrS+RySdbCKFhUlsoF+omfL7uEY3/4TZVeAXeyACFBSc9cbQO8Q9fPCZUXUINnHhi1G30xlWr6J/AT6C3TVNc9O++8zC6Mhj/WCK2kWoFl1cAlt+GZhOX7mCoK0OvWnTro3RKFuo421QRbMd86EYO1mx+/FYuAiIpNlEnCFZ2fBFVk0s4pgYmP1BNMpEEqGhJmAoN/Y8/fcO6t8OQx/Qhl1pq3IEQ/JIiZ4LgaBJcvnOnIotmZpvJDCw0JQly963wYqYa65jCoAA1eIJPfqcj4YKg1UQc1i/d9pBAWQaDcKQZ8Cemw4B+xRjVDOEDSCOH/pvovI/Sd1jrC/CT5gIWRLYaH1c2N4Jr4e4/88/E9viVo5iqKPicin4zj+zSiKciJSEpFfEpGdOI5/NYqiXxCRmTiOfz6Koh8SkY9IEPrvEZFfj+P4PVEUzYrIF0TkXRIUx0si8i1xHO+Ou6aISP78arz6T0OP9b7ybZB7nP1Z90AutTQbhOPDB24RzLMM/F7YZFk0z+ju0EcfczOMj3ZHhT4LmxKVwlqEREpdNvUenFCSJyZXEVuPEHow1sI0uf37dLujkXOR6dAocYldTlAnozm2oWfYhOUu8OxyX5PdLFk464LWrttBAdzCKX/VDK/YHTKEwM822M1qetKv1dJkNBkqyZu0iLoGq9beR7FPAQyR21tBuMUoYEsBdWSl/GxbyOB5ATTNS9qVrIaeBsyRWNVwPwFVBWXzLsIFt8L99i64UiFT6ZBFE8l2Fk/Ze0oYJCBUI0Bi+bmQ62JeZNy4c8vj9E9d8hDWrnozzBdVE/TTvs4ssc7Q4O4V99D7WvmdRwEdk8bGOtpHYd9YJJ6IW+IoDExUjGutTHQIu655QzH3Ouo5iOpJK36flfpfz3HnI//gL0+tHEXRpIh8p4j8hIhIHMcdEelEUfQjIvJd+rWPicifiMjPi8iPiMhvxUGbfDaKoukoilb0u38Yx/GOnvcPReQHROR3Drt2qpmS/CthAVl3+cKr7h7vPYd4Ml6EYbxnFjwBRyt19VwISTy4jdJavLDINjxDPrAYBmzqrQogv00+Hj+tUQ8kmj1PwxXWw+SCTxSPMAE8FzYIrf86eVvQKN4GGQNrC1q5SkWBeRs8BLPlhbCxKLAoSGOjYSDPCZqg28ajgqTFOzkBSmYV2vubLuROnfaw0b0bQbhMr6Ky9rb73TNngjLJTGCDIpZ6gCSjrYPWNQ8PddEb2AQ4yfhSCAEsLwTPaa/hc8VGNkzkVpvhO+w4RnZRu0O+IyqgDIT6/HtCE5WHG4jfYAyVFEI2lduoqNXHpffaAbNsdh/5AfXu2CuBlcLGPJmGccP2kRtbwZO2vhQiyTXLvbj9ukIyz3uI69K7vLToys1RSGcPXlhFPasqhT72stFTiDhlChl3B4BcZjX/lkyAY9/O6Xwx11dn/sC/a2G0/phK6cc93kp457yIbIrIv4qi6AUJFvrPichSHMdrIiJxHK9FUWTByFURuYff39djhx1PjCiKPiwiHxYRSc/MSEtjwjb5RqUqkmyRxhddVxKmmfL4xgyWACuCCCvRNENj8jG6fufQkL09A0FpJEygKCjfBvPecrhWjAYjiapJY/wkyyYsjRgUwUaf0MwCkoZNbIRstIiJd7emGLRiyZ3Tn2bANwx2iEpYkRZDxZpm3YQpRnpAxskiIrKHQi+rZWBzjXXkDxZUqBskVSRJedtU67kElA5RLmzebpb+7IuOgd+D1d82LhcW5bQgOHTtHHYtKhjLFRAt1QZO3pqszCWapfu1mE+xpC/fQTc9yoCaQa6E4Rdr98lWgkShWYW1iLfAJFyX4RmD89K6viVuQFnnrTbCYlMIe9GqT50LXlgeyo6w1gWtcKZnyDEM0cKjZYtPWtoW+uIxMqRma7oX4cTl9rFXswq/3oGByZwAwCC27scljx/3eCtCPyMi7xSRj8Rx/Lkoin5dRH7hTb4/7jHjNzmePBDHHxWRj4qI5M+cjI0XPaeWdBetBhMhFzZLULx4FQuVMXOzNJg8TcNi7U3otOBSB+chEMdU7KVB7Vx7mkVKWqQBi718FdSxL2onJtI4c6bGxCrJ8Ei64XopPHcXnbPY/ah4NWymxhkkAB+B8RN8+Mb/zzmiFRlrYo2WPj0gG5kaYqpz2LhAKJlAoEVMJW18NvQUjHxORKR8U5OY3+7hIza02N7x0MK0MjQmipy4M3VuB+Byoee1rYVvtFZ5LhbumcKkdd8D4VpZiwcfbaDRDZQ/rWMjB4xaFOqYb31PRJk1F0atzATzJokBsc6sYK/WGM8WZkylqdfxXCX/nP4WRf9g7VBpdMjpo3PEepEO9kpew2VUCkQtDZvWHIKXJ/w5pdfIb+H6mI/G6VHunTqpmdVDODgzvgo3Rk4sbTUB36BC/76I3I/j+HP635+QIPQfRVG0olb+iohs4PtkzD4pIg/1+He94fifvOmVIxn2pu2poCJtADnsox1UmaqgaaxSYGJV6O9SB9hA2ORGMsbkbkL49kcFcQxoZeEauMNPhw3SRSFL72mYW3ovtOB4fi5AixP2UPCRQVK4fU8FEhZ9egsQvYvhulkKengShTUwEubDd3JoWCHgNC/cUSjqzCH0unYMuiw1ydi4/27rpibjJpCLgKAdkqsBls7N1FLh1ltHshFoDtIR7N3QsBCoE1Jr8PKsCRcS94zV9lShpkgFDGx8H4RqtmYIbUwtowBuJ3hkjCv3IYjrEwgFZcY0NkEIy5Rv68QoSCD8UD9jmTHfM//HviY2vy3MEUMbMXsKRDpfl/19EgSQ1tBZwoiI8fKorKxDG+PoaMzTzgeh3l5AeOkWOIGM8oREctzLpExYULpvB0glmGFTuu+4XmiqR1OKWvK/JpQZlX+kUFTCP4/K+JpCP47j9SiK7kVRdDmO46si8r0i8pr+70Mi8qv67+/rTz4pIj8bRdHHJSRy91Qx/IGI/EoURRaM/X4R+cU3vXgqlpQuZivgYeu6REETuu0Mmygw9LALXnjdeIm6CVIIq0Y3d05EJFVDmCM7KhAoBNrnwXypybSJVXeViUG3dmqkKEj0G32IzlTG+sk4IzZWpO0f4z4FAxa1brbu7Hhl1r7gAsn49DOznsjtVj1s1FrSmD4VFIWECi+D2ookIZmVGQ+tlZbCczHxx2G9jOmtpaBke3qtBM0DOmvl1/3d5d4RvIEGOp1lz/u7MS+RWzWRTN/Q0AaFFKupyZdvrfTQUo9WrPUUHkzjHcDb6bOEX2NU/cXxsfHhWqfeZRJRp56CPgbz6+a3+TqZXQ3xdfbgZXHhzEwIO+2suYfCe4lOhXdL75ZjHHaE7UIj9JCw83JeuzN4Hyof6OkPEsyX0cjnRKtMdg/TczBBTj4qGwkeo8N4rBQp1Brj/T7u8VZx+h8Rkf9FkTs3ReQnJSyj34ui6KckcPb/qH73/5KA3LkuAbL5kyIicRzvRFH034rI5/V7/8SSuoeOfjTsIjWIlCTtAayfdzgqg/HDnm0iJinBUW8Ve8U534zk6Og+DFYJLTC236NwGyqIFgQ5XrSx7XUXETu85wKns6gLDIuTQt/QPSIiRW0P2brgSoXhmdgQA4CHCtodptSC4lxkZ/xcOfCQNO4Fqzmz6HM0gURxXStZaAVPvYp+vU+Hc7GNJTdbgpzN4KHgLMoDBTOul2wX/NQT2juY6B0qhe5lNPDWUAuvH73k+YNYC5ZYLJfgYlFrMUL3JnqcZ846HHbvkyEJmfprnpRmLmBLkUIrS57EXGPnLHqn1r4S4Z2ZZc+BWC7gAIoigUwZ0+KQiccBvCwjIWOj+0TDdr2XLPpKZC+6ErdQzG4d2Hs8Swa9qYcNi7CO2HPAFG4in4QEsqGoEtTP9MThVRRvhmu1FrEm85gDRc0lENVk8NacVxbsu120GCVoYfUzihI7Nd6QeZzjLUE2H9fInzkZL//Sz4mISEqtkhx6bbbO+AbKghDNcNs9xAnZRLqnXZnSJ9HRiG0Y1SUkgoP4ZzLrWSy/x8YnCe589QRmUYkJKzSrLJxE1GRuw7pHgU08E543B7ggqROG54TXwRaFptgIK2RTDlpIGVUAxI3XwOpovO4x4qs9Jtb1udnJqXsecEMoGBPqh9FDWPKUjVOKeAaLB7Oil0KdCt2OW/JYRKSOOTA0CJ+V5xq3XRJwW0BUzQOYAFKJuYZ9vQbfx2F9mS2+T3oICuLGvl639eZCJgXkCkfupHt0lpcgfUUNncbM+u0hrJVHI3tTznwWAiUYPrEeC1l4acxDWY6D65i5ArPU++tYmxC+fSC6rPsYQ6Y02vObajQhIdtewTrVfZ0GYqc3jX0PebGopu3kb39WHsf4VPyJvzxk87EPnX+DPrZAd8B4XeFVjxnWntL4P919ChENBQ1Il3wDydVFbdwCi31AHivAvCy0wBgym4UMm8B0x2+2ocBg8Qp45RknT+u1zBMREcmtuOIygdRFkjRbgcWq7nadMFAmkHHfXb3fTAXshsB9pzShnqBphlfR13fDxDs7QQ0gKDOZUReYiT0TQtzs/LthqSncCYGl4rIkIJOBPWCtD8xgQLghRW/JLG4oggKEb7+C8KOylvaRkOV9232xLSIx7Gn8bijUMZq7KJzTd0fjJEELrGFAQhTZ+aqFZj0VDentI7zDMIcpV5K3Ee1kc0Ohn4iT4z1mx9CMZNiBbdj6lJBshF0t/FMGoICKeVxjcqLnsL+algvgXmUyWl2APhF1+DvvoXY6zFfjv3rf6PW/HuPXPnHon4620E+JiCWITOCwkw3eJxEz1qCbkM6IjU/Mkt9F1eUKhLZZBHQT2fUJlvwQmgh3n12dbGNQuD7xey5EbvwXhZFnoVJJuL02B2iO0b/ncfaBhh7SgO31uug4pM3AuwlufzwLzjv95SAIm3OcQyi7rXDeRPKW4Qi9b5LLMTnby0OBvB7uZ2v5kMS5zSfZEfk+LQGHJCgb7LgNi4Q63m0iP6DviR3J+lAgQ28G19qvHZJAVkOFcXwW1tU0R5LadKXTdJYFieERxqp4EjF5Ag3s/iZGkUjhyyNflQh5qvwOqqXtGlDygiR8X+sxSGzG2hTZC3OUQ5MXwbUEDd1tbmkQcG6t7wJbaTJ+b+8zAdNkzo15DfUaCMVOgCb03cVcx2xvqV5DhPCUULHhvqavheO1k0cvvHO0hf5Ahu6quVQxAm6J8Ays2/S6IixYeApMc09fHnvRsieqFSxl7iCRdRIW2Bpix7ooevOwWLjZdONMnvT4672/DUGqGPAi4qe0kDpw5yNdwLSq+kCZZNeUo2YFSAdYTRYn5wYaQEixZ0D3e4KFz5BJEcU2XRXatKgTgkWtvYQrjDlamPfkaU9hlEWGZPDZQiLVO16YVAINtLFc8u8CxUur3Fg0mchNxKvV0meOiEVGtXoQ8AwR5Me0dhRxYc/wD6GHOf1cRMVwExW7/Vc8ZHFwXgUOvI4Sm/lYs3tE2AbE5FuHNiTxY3yXCcey5m4i9GpIoIjV267voh6EZHza3J29gQlpHoBSwWhVOqeZd0FldjV8d/o1X7PV55Hv0e/Gi3hYCmISAqosofGS2vZJGCjiJpHEZ97D1lSiwg2hXHyuXgzXykDvHZVxpIV+4NMPk2cY7ygBnRwfMjGe8cOqVZsqSNn44QAxx/JrQSC0IFAzEPSJKjtDwrFfKCwVi+UzlhvfRHjmiYCEIOkX48kJ/LFaoXkoiCxC+n2N9afZGB1Wy3CTEmmR6Hzl51qaDEJ5F5WnTJR2a9pAHFZyH2GSodXPnBqUyi7K8heVNqMENkt2ztrYD1DUiZNQFAjXNVWIzJ8f32C8DdrdkhY8TRddkB9A0NrcUzF2UGm8dCZcgx3FKjgXKRWmlbStDGZNDpKv2ZjCuXbe48qgbLUleHdMCjcehjlie80U4ZnqqWYA4c3tAjX0rO8PU6JFKCjyXFlBVJOJc6wjewYW0PWfQqcyrHVT7tto/NPP+I0biqyz5PdKZ8bulXmZYatOkeRaN2MOiLLONMKnNQV4bPrf6+dGc3VkCKHXTw/FiNwG2W9AyObjHHE2ls6KLhCrcMPE9lGExNDCwa0AJcss+0KjhWVudQUJtoO2C6Gurj8yEnaxOAqP0FBF4+8JZBjucaBx1y7is4PTfl1TXKT3JR1CjESUJYo6+C49hLIK/e2bTlFg1A0i7ionqBcQu+7CxS5qc+yDDAqmSHJWHKVWZs1kRpEdTJRlv+Jz3LgI/pXsKFXFuIKnC7OOgmHnK7uHNqpZEzTP5x1RYxz3rPqcBaOmKRvmPcw7EPGCKwrBcs6fZQKc/0YHfB2t/F5ccY6ayVxYB62+vwM2EyGhWjE/OkdMdtuaI86fVmp6fzTH0jyJNQ1vZ2EiPC+5d3pjuJD6qFTGipayGjpsYcpK492HDvU0kr1EspzrcxCN/J20FaZYGLIUIJHoYVhv2z50beLvGquvn/G/M9QTGyoOiKBEFy0Mg3sPDuPef4zjSAv9qBdJ7lHYEGZdD2BlZwC5nAKh2eQZbTiNJg576KSUWw5YbfKk7CAc0F4yGCXuhZBOIGos08+Xb8yaIt6vk9bNwQN0fdIGGqQ76CMhlGKz8Ww47+ysl+fvg2huQhEpZKgkMdm2Engx5CNzTKj6Z2PMTIGFcO8Ac6hKdP+RC6bJZbfE93Mq4IEW6Z/0OVhFtyqjK272ENrAJj87E97XrV0n4jKysvA5vGfysE8uuoJga8Vp5ain0GcD721l3GTRDgXOlCoI0kgbva9IkuffKKx7c8DDgwHVOOTZx5i0wecXnAnUPIhbGz4HTFBbOIswTQr9biVJZyKSrN6dArJqUZ9nB4bQNHo4tPQ9WYGeiEjqef+7NT4RIDYPuv5der3LlXAtKlEqti2laW6A3mKxUh/57iZzPMgdMekaGU6fIR9CWe3vkAVZ5P0sSd5H45VB4vfA/sVfAAAgAElEQVSC79rf5ciNIy30JZZhsUlO2xGW1n1md55zIZQHT4kNNlGZAO2uCYd99DONUJ2bUTeP1aTkyO8D0lW8pTkHMC90UXzYL4dz7SCcYecXEYlnw3mbiCGTpjY76wI8xR60OoxRVMQtsxTYR8lZYjH9lWn/TQsW2CZcbOtHe2fbvQa68LbZFk46xjwx9L2RC4aCvDOL8I32qJ0q+LO2WJavVa7j4vwibh33EM9gl60GBM5GMyhccuxX276OLExQAVcMrVRr6JKgiYDXwfBOZ4AErw7+7uW1gOO3xvAiHsp643Fr3l6Gd8q+0LWdkM/IA9JMBlTywg/vbwGAANy3eR6vXnPK6KcuuIdiz0CvgeA2U3zkwGfz+RU0grH1SVpujnmlC6c3tN9BrsBqOFivMvZMXqmP/joJD932cGseXgcdHP3MEBqp3ingJ+6Hu+ig8O6ojCMt9DONWJa+EBbF2vs0jogXQhQAuxtt1YPAYhciDrPw6/tuirCF2uQthUaCZh3tdqXgRuRwM3Wxv8lZbix83QkIcvDx91V45VkFODe+8cLlhcB0QWW20/SbXCwFIVEDYqeCzWSblY1TGl3EeJnjUEFZhNtPF39S47ZstMFQT+pE+Lx/y2veI+ygzXWf0IJivKmgDoALt25Wzy06SdqffeHy8PMD9WzOLbpl3Bn40qYVaU29Ly15yOfujkvEKYWosgViozVaC0GBdv2adwd77qm7w8/XNgI7aAU9cNnq8vRs8GDWgP6Zr4xvcHPlahDAVi0rklTYw4I8SLzKPVq5iljD2qxN+PuYQIjq7n6Yj2cuoqXkrs+RxfIJD23s+DrcKYfPFPQcbEBztx7O+2e3vOfBi+d9Dgvp0TDijXWndH5iObzHhPDGfWUB3SpuhHPUTzEW6x9NaOeR6+iiR4pdIzXqHITjozaZTKx941bkPpbRL0Sy81S4xYLuZ4ZWSE714owv0PaUEishKE8L614zLDQmzR6gCCNOKdKBla0A/7ZBK289MKeu+TFaB8PvIk5JS6NVHxUoHC0kIbcmwwq8PL0xPLZa8jBJTZtusFsW+9aawGE4YaXoVtcX2k6ZZMyT9W3fzCygWV0InOuLOBdb2ll8HSSGkoG31AOe3cIAD9ZcsEyCpoHMlTZKJ/26lmBmeIgCx4wAEVeuD6DwSQ9h6BnGjVnIZXO3NvDfP3XZ1555LSIiC5Phu1x7TN4a3py24IPXvHnwB9//F8PP5WeCYryz54tvb0zCn9bmzjMwLhQGTIHIMAaH3e8u7pVKcKBVrGmsje40FLp6K5cXfZ2y2c4ejBJrmH76KSfLm8yga5m6zfQS33HayXo3m+FapCvP3vNJsN7aIsjVkUXTt4/09bbYWwC2g+QVJ0BUYHs807X0i4pKOrb0396II38BA42jMQ7ZRhLyK7vO0nxiIrzJHroBsZmyLWoKkyzi8MYzzq47WRJdIc5tm2j/PKxcYn9VKaSrPtX1J7Dq9LQ9QDOJjy6tuHCzzUTL9QAC3qomc6TvxWa1xFoHXget63002raGKhEUK5NltyohtvzlmocATsz5DtpR1E97AbFxJhlRb9HRhCALeNqoRs2poN6HsCCUdEhPDWswByuZw7pgZYAQScy9ro1GyueVvYObc0GQ0dLn4Jra0fMa0kkkyUpangzCLcEAu+QWAY2WnEoaa/ISLob4vc4BWSMZ08/talEa4tEMIz7YculleaAqcji8R0MzZZbHx7Pt3Txq+LxlkKdiu85tzRvMIP/G5zaefoZ/9jtUnOHvRQh6PiOr560nRgQDrlfEHlejhPu79ABInjOKoMLSSjQRwujrtRgiPirjaAv9XCxNowHW99BmwRSECGPPC8XR+D7Hte3gHrISUpAgNmFfAAXrwWk0t2Dxk3oe5OjOeChWctoEvQFBn96DW26hHKIEIBj6Y2CpVfC/s32eYcu5QQlls1J+WpsFKIXsBqouNeRCLHj/BmKpp8NmKh9CMdBUHHwC3cCP+FxSjHcbwjfRPlLnoF3BvI2BpT7ac4FImCST5ONonMkpZF4DG4gU0GHNFAtDL3d2xuc9/OD4KLNdq0uoK/s+QGKYF5UBfUW3QQrT0fMPgPxqGccMIxsoAiTctq51Baw5IDVCTg2CLikfAN21d9MCFJYKmb2MDU21C0HOFqAm7Knwd1ruYRh8lH02EpBp9rnujc4BKRcac1oAB8+vVkDMXpdRinThULKsSWn+f+y9aaxs2XXft07Nt27deR7f/Hpii4PIFjXFFCEwdmSECGJFih2DDhTziwIpUITIghJ4iBTYH2IngA3ZRBRAEiAzigBBjBNZkSwxki1K4hSK3c0e3vzened769ZcdfJhr3XW73Sd+9SM4X6XgA7QeNW3qk6ds8/ea6/hv/5/lTxl/e+yHJfa6EskCTwqr9V5kjHx4GLbOA/54pQMHV4nOUl4xAWkCzrqZTbmAVGkUZ6EgIbB1sjdP4sGHfXGyINeuea7ghnKLhkmUUu4OecFBMupbyE1sQkPLdafraJRhlhu5t/tqKOYXXzOUz3G395F89Xcd20nr48bwxhzomfevxZSHndHHa54vAU4B1rdG0oBQPgp2SbN4GzivpuUlFRcdu0Fz+m/hbwvPXnbRCm5ZwRjIiITE8HLJe8Maxl3T0KEsw/N2AEcETZyGYEdWRtbD/17I9dPUtckkpYA/Tcb15LXdSViSyGvSGds6+MQ4ATCII1iAOuHzLCyj3mgutSkjKgARGDXyw2dv2UHm7O+Ul/330V0uTUIa7WE2lHzCWocN8Mz5aaxs+1z3npWonXSkcBQ94Z3wxSNOYncdDzjPiCZE6yL6N/m8T61K+CsRbohExJ6WY7LbfRjSdrxE0w+uTAA06JX8nAjGJrR13wi18FhHxlRFLobS+NosNGFkSKGAl//IKO1m41iqUypVfynfNHQozXvm5vKHCBpTEsZrruAnoFtbFwmesGmGCJynr8aUhNcQCcFNKXBO75/oBz3hE6Oe/NT2bRLL0ijNBVtcY4GnpQ+QUb3Lpu/UvS2mqs/f+RGn+yfvQXFhcPIkH+oDMF1owhu5UGbAcNgDVVLVd8AOV5/+FYoOJZgBEvo8Ezp9Op5yyDIG7/hRcxzjYZS7KYXCNFbY94soK7dGmhE1ECfgbqBkbDBGKmW1S8grTQFau93XL+ISAsMqNRSTr5DcRdNmeQvIknDJmn6BOQ6IuR4T+mbC+icjdBx3lH+LBPHEXkHWR4zAzaXAVtN8fAbnQfSq2QfNfhxSmIRjJvsv4m0Cc4i/ct0XH6jbx1vOfxND3aTEgdf1zxh+0Mo3R8DVGZfo5oPw1fVymRusLKLqIESaaXhnTxHzL2Gd0UYHjJj1hUuODfvRoaG9u6h47INn3x47qHAMXRlE86hu/633qIv0Htn4VxNIHaY+iD2vWFFOozx3pSf1+oKJSQ4iRo6UarplOA8FssYePrNc0zx5eN3jRbgjGLlMCg13TipU5Dqb7gDgfFbYQxpkPLweB9uhjHqLoDRsziMQuk0/BmSnmIkg1yNheLWDmBgmhIhjTS7XI2CQEQSA74LGUl6xzZnmMdPOd9atO3jGaSEajpIc5TCWplG1NEe8XmSXCO8G1J022bFMaImwYDqYbPhN5qgxeiRh0dvooIu2XPwE9lRx/dJdU3WUUvvMK1FOg0z+vEF2PuE0pkNjbNwZHhfKp7SHw6In/lxyY1+JDl7KDr4xk8vItK5ifwm0yM6UeanPI1yBM/x2kzwWJnb3njoxjWvhpqFXNPqFUl3mSYC4PQesBis0EOvKcZiG18K18gc8yG8c6agDIVCz3Ns1hemiWqfVtGMBOSJFX2p6cpO5T1gxG/eCKmcxwceSj85pmhG+HcC3ax7SHkYjUEFAiakiW5WQGOgeXbm6cnpYymuA2xKrEUkVBF4BHPT/uzHljxFltcUFNNaFCO3Gt8ZPG4+7htrIVoiImhtMrtXwdBEhM12FsEeqtey3/Rzzdb8eW4V3MCbkPxY2ceAqKBH2yEyK2J99LHfFjS6pRh6Ho1zPcCEry2ElAqpMJgafPuRgiYg0MMiv8FW2QDHek9KqtLy/6R0QARvz3G3hNQgvr+kjZabQH5RqSylkWtgDIAyuC/2GcHb+4TiJc4iPsCNE9bUjH3x6eXFZ3JcaqMfDRwdYw0RRRRJ+/AeWLB8/wsPRUTkrT3P6zI3fv8wLJAZ4PhJwiTK2Mduu1wvO0wbvxv+pXRdD7t7V+cqSbfoN/bUy2VKpkEhjLHhxdIjTzm828mlsJntIbVBSmlDZRDdk2rlx5w3T504fW4QyflhqFnU7Sgh2/min//2L/n7+z/jr09Og9EbYIxH3vD7elwNGw/rE6wf2Nhdh3HfPUMERL4a9dqZp6/BkNbbbDPS94vD1ApMITygIAu+ZwaclBEHiNKsmY6KS6Q+IFTUjD25kNg0NtDa0gDBQeUABUeT+iPrKQLhtVWvh9jG2BV/dnvn2JhuhHl29HWv1/SXfQy3lZTOOsBFRF5Y9XrQNx67q76o6SqKz9++4p/d1h6G59f9b3d3/XdPFdRQILU5mORYd8hKtZAy3QTPO6lI3t83B47n4WeLGG8LgFd/zXsO3svj1ae8d6mNfq7tRrWjTiS9FxbAnlvZSV5be/unnvvj5G+1vBuZb1RDJ+SdM588qdyfHlTNmXrdXx+9iNfvCx+ihzX7pz4Rdj+kISO7SRvDRps5YnLg0/s1HpJR4sbRzGOt8nkINzPn/uabwUMroct3ety/f7TtBql3NfwGjePJfY+GchrOU81qsgZsvcJHKdrx1o/4CmNX8oJCPTfv+fOY+Zg3Ylkq6WrNawpbTff8zOu/PuZG/4UJNxJvnjj2/c7XA8T0+sveYcp73H8SNpiFdf8tNsDdmgrNQB958aHfV90b0BiFffHtq+G6V/26SPRmTYL0fPcQQXzPyv3kdV0b8raO4f2j6aulyCU6KudX/NlUdsJSz2+CLwe9Kexr6Kt3yw3qypTj6A90PPo1pMi2/dnOroWxXx/378xX3OW9+aI3xvUUkvn9S28mf7t77s7a+yY3w/UBfs3xOjCaBtQcUlYNy9p0Mopn2Q6afTiPjtvp11GvUT3oJjR2RwHpJDzTNoi7f5Ny4e/h8d9d/NalNvr9aiwHH1R2TTXKLESxANZFR8V8OYQDUwUyqQ8fbBghF3ZfQ90iCnB74JVPcbFrxb6PdMH2R/EjWpRq7/oCikgDq2iLHvDw1a/4TDy7hXtUL7A4CoIwGKwbtbCYrPlMRGQCqYWNxZCeYSGZmxF5+tcnQsriDFwy90f9uvqa1ywgXXB46gbLJB8HaJopIHfOPLYt4tn17G5TM/oLZaCLEHbbsz/p+vVdr7qhZV/D0vMhPUNitD2kV6z2srvvxvXasp/rgxPBc2ujCypFgoZunufUY2UvBAnVBnoPbGrbjd07vl31javRD8/hawXvR6GYiM3ZVA4bdY3ofcPd6SV8n9w4tnGR44a0FruDcI0xgBBlyHpatEI8/hg7EnHY2HGtft+URx37vWEqi++c8c3wyWiY67+z93zyt8qKnyutmhb+7c7COh+AFE6ppAcQVm+ushA8zF/Unkd4zMK5foY1g8tyvCujH0XRAxE5k0Ck2Ivj+MNRFE2LyP8mIldF5IGI/CdxHB9FQeLmf5agk9sQkb8Rx/FX9DyfEpH/Vk/7s3Ec/+JTfzgfS04l5yIdOxZpciRB+x89ZDz/u2Fhf/nsip8KHphhfl/b9vZ55gkXFoLnyUUlSF0w15lg4xfckLMIOaKkWC1sKilkiomN4PdbH/HFNgqIXvtOMETH+P6VafemNlrBSzUPU0Tk5evu0VpXZStG4xHUkSqATG6chQ0ii49cxJuMKJgdgdJ2oONRJF8Qc7kY2+O6iokQ51+HPoJ6aA8aHmkYN5CI59f5XOgZpjp1Nb3DzYz5+66pqWGe7Zy64bk/EbzQbxx79MAGuW6GuHuWHKOIp4iY+mMX6z64Pdq6mTCSODr0DaKsMN2RaTQkYjwtFcR51kPRl2Ng6T9uZndBdtdWIEAOjgo7do1+orCAJsIehHsQQtu6fNDyKG8JbbJnGtqf9oaVw0QcJTY142smizZDxGlGRkDO2CJMWOdyLiO6FpEkp08hHI4Rx9sADIwaLsvxzXj63xfHMVhn5G+JyL+K4/jvR1H0t/T/f0pE/pKI3NL/vkNEfl5EvkM3ib8tIh+WMHxfjqLoc3EcH8mFR5RM0rx2JPYhgNzHpHv4g6CJ1bQNYXeb516ENKbCFKkWjJeRo7FppjrrnhJTLl3NRbJphimPM0XX5MFH0Jwl6kJfoNBrtKzvPFfhWpjYFPi4Azz6jJLOLS66x0zOlLONsGmU5/1eWpBezM+iVV6vZwBDmiU0E1O6bsXHM9axizAurBl0Nvx3IxVEGQBVRJ3fw69oPP0h/z5RR8c7wThWIcTx1rGPy/a+P3uTIyz+jv+t9e+h2qaeGTtnGSm8ehgchW2gaDhPihgjMwhc9kd7bsitd6O44J7pn973DuezVWxG+hyIfc/Bs2wfhg0kN4scM2o39b0w3lSQo9FmCdOoKPZ5rez+tW7TWZDawdCa8M8bu74xUu/36oJ78kYT8uTcAQP/pnk9eW2cQGy8G0Ftx5wSbqxd2AiKsOe0PkVo7wDzt2fom3jY0IfvD2+GPKjBUN67vEmUf5sr+6SIfExf/6KIfF6C0f+kiPxSHBTX/yiKoskoipb0s78dx/GhiEgURb8tIn9RRP75xT8RJw/VPLA8GqPYeEFDZ+H+5193Uq6ra55HXJsL+wyhj5Rry1sXKvKjjSNPHVCIIlpRCUJw6PRhMArKV1OecUOb0ju16j8lEkGP2yKfvXnaSC9Ng+VyWxEM9K4pul1bCZsgi4WjXwez5kcxaXUMcoBk9h6iOGovlkDohusqWRfrWxCMeck34cYkytm2gcC77uXgMSv+mVEJjUhBoY8nKHJS2GQkg50095fcfyn8MQqSzyuSCEaOTWXFtTB3bi76fHpQdD4ceoYjGpGdU6ULHc6FqWHuGxpiRi7WecrnSeEfUe+Tqbs+nIeR6XBf9PRzX0WD2YeHxdmnoW52GPsY5KZNYxqAA8Cby/fCdRU/4E4AieRI8hep1NfsiL9/EsEo6/VemfEaCykwbA2bfKeISGHGxzjOKLIzeu2huco2RCLmOKfN0eHciJh5mEbUYAymly+7866Nfiwi/3cURuufxXH8GRFZiON4S0QkjuOtKIqsvLEiIo/x3Sf6t4v+njqiKPq0iHxaRKQwN5HIsLV0MhNfzf2WyjtFpRueW/AwkZS1Bg9LNY+MDCNTpjDpedRrgPPpdRE6eQ56hynFOh8gFE+p+WhelKmTAgq9MTy3ghZ4CxkNWSKOhaZhOIOknSEciO8u/IBvGgDFyel2MAj5MeTeIalX0mihA3F5Gu2uiV8s+fdHkQI4T3G1hH/n/sDPtfe9/r7hyRtAeNBjTRwDwDzJnUM4oI09DWZ0yzeI6tfDxtFYx8aNovPheNjETr/sG0UbOgEpgXuNogYgNmPGsG3IEqRZqvD62U9xonMqpRXbwPLVnxjFPKQkpPEXMQ3T+vYLmspU47YDPv3cHOg41NjXpuHIIB03WAnj0ccGRAK7w8fu1U+vhfn3tcfesVs4AvfU7fA+UUt1MHpaGq44h2Y9OARMxRj6i4izIsbLNgjeSyq9qQ8vX0P0gPtihDC4AO13GY53a/S/O47jTTXsvx1F0RtP+WzW3cZP+Xv6D2FD+YyISPn6SmzanwMVpO5TLAGe/sde9kuyLlbmdcug1nvYCJ7Zl9+6mvytCOoCwxx3kE8nzpgatWas2VxCD2j/OPzW6IpvIKkFYgsXHkV3wa9lGXz5V8aCl8ki5Dlolq0g+JXHniJgTt+8xecnHOnUQwH8X779QvKaxt6OSRRajcu9g65QojXuHAajyLD7aNc9ywKbp3bDZ/Y/As8XxjGnnhsJ3bbZE6DGJY7BKAo4Lhe5zafRCTf08Rc91dNcDNcwf81TEAznr0+Gv3cgeP3mnsM5WmhIKizpxkgmVWyMlvIbgM2VsNi50frQ640Tv9ZTOApmiOrgDKLBs3w1G+Da2LCX5hExxuE3qKomaF40Ej5uFKk8t9a0RuB8ELV07eV7flqtVXxgbjP5W5Pc+Vp/YxNgY2KYWpzpvnM8bxqYnT2tU+3jupewWegYxTTYiLxyag+4EfRT/UH+0qQVLyJke5bHuzL6cRxv6r+7URT9uoi8IiI7URQtqZe/JCJWgXoiIsQprYrIpv79Y+/4++ef+sO9nMRWXddGEDZe5FZ8YRtiR8SLRiwYkbnPJtrUnH+HhqGseWg+XBbQmGe3QmwPed3yLibC82GxnkOvN4aupnWODhBmklaY+egF5cufLvn7W11QE+j9Pr/kxUCyExpn+fvf7xtBEbMy1ahiLem4bxJoffvCE70WIDQAi63ohvuF11/y8wPL3YcRyamnO0CvBEVtjCLDBExERDrj/v7+45C/n3mfG64sNSwR7+JmB3YM+t3q1bDJkl9+8wg1gZlh2ULyBMVZxWpaHuxribHHBnt04BHhB+f9OZ2rDvB+EQo9BF6NhTmRUmBDWGHX1UO3NtsuJtbQN1ENc7XJxlSGKBrF9ciDxdSeRqRTaNyjuhib1WytElo8C7rurVaY30RFneDZGV34vVOPvCbGEIGgQN1WWxItZDe4FTeGO/F5DDQlybRy7gz0E2ASHdsO5y00vgUJ16IoGhWRXBzHZ/r6EyLy90TkcyLyKRH5+/rvb+hXPici/2UURZ+VUMg90Y3ht0Tkf4iiyCqLnxCRn376j0uKG0MkLX82PeaTKgsd8Ce7jt55aRpkYdqJe/QEQh4ID4u6GvJR9gNjk1BeO3XZMFV83c+bqwznk0emhr3oBuX58PcK2vq/+vo1ERHpPO9ZMnY9bh6E311EdMD315eDl2oLScRhgyIiq3PuqT/eNe4dvxbmizca4bdGIBl2DMik0d9Ovg/CJvCKOF5tvXdOxpSdfC1c74OXkRajt6WwUG7cjGBYeE8I+8jKAbihGUo2d7W30Fw1H4wuIY48iihc9+x7SAeQJM1grXnM8RyQWWSePFe0EdN5PKzXgMix0cowTDL3bW4QuUFQXcwI8GLUJ/IsAGvUW4B334TuQ0cj5a07jo6b/B5fX1sN1Af0Oe5FPt4TkGbcPg+f5YZPNFVNu7TTbK8+ButAt50o1JlzhzDjjkmHXsCKavW3LHI5EUltfPUrl9DF1+PdePoLIvLrAYkpBRH5lTiO/2UURV8UkV+NouhHROSRiPygfv7/kgDXvCMBsvmfi4jEcXwYRdF/LyJf1M/9PSvqXnjkYl8wtkCQFqD3/QcbXvG30J4qRVxAJpXXvYZcLR6+FTpZSGP3LjHk5Eqxo/F+NE/lh5WxehmQMBpX1gSWZj2lUfj9YGg3lnxTYRpjdTZ4utvHvijmJ9xrMqFtYvsPcN/8rMFKiZQg8sS6It/ccIRGSiNUoyGif4pv+TOofcQLqVl0xMzPxxotrYLugBQaxYXhBUZIJtE3ltZJkZkR+qv5YNJA1NZ9EzX4JzeYURicc6SzDJVUZJ4fqR5Lk/SB5S4D/UMuI+uS5XWRX6irBpz3QqZQa2CroAnrWNAPAh0AQwUVwDnE81JhzY7a/HBufGLdjTepJiiIYvQOO3VssijUVrUbmmud88DoNJrom+CzOZJU91W4fkJZgfSpzoV7aEAcJkI9Jq+kiz2g7wqAyA7wbJOrvYSp/SiOL1/4YUf5ymq8+DM/LiIitQWDKyJ8xQLIF+lBKdyQC4DYXDXU/R2fEBPX3SMwRa0O0RMw9CkvczD8VA933ZOxYlf92H+LqAjDave3/f3x6z6pqZ1r+cNxhK+HkB00BAfPf/QIXO+KfCoj+mAO+fihF9isa5gTvDrpi9gQSHksigl09+Z0WFjA234EyTFkC0q7arDggpSOfVyby0p5i9/KkfNEUVYR4ZI0jhCwMa1lqic1V/warUDdy8Dbi4gMtJmO3OkcI2oSmBjHyDaimmlEK3o7KXEZ5JOpBlXRLuo2jFSM2tGsoriOT91gGe0GjyNoNaeKlCxCHinQAYACdqcbJUrvlp+/cMfnb3s+w8vFsxsH2Z5tkvEuwBGIfKauhXVJ731/z9dX9RsalbzizlGq+x1gDaPF4LreAfS2nwHZrCHVanOCz6Bcze6LMArtZyWi8vDHfvLLcRx/OOu9ywsmFRHJxwk23Aql9IoEr5+77cVJoyCmgDI9hdcOA2iInXvMV5shTqFcMHlOdwBd1HA9R25xULP2J5RbB8LN7Qmw9GktIAeMPD2hNoq+z10N1ATs8Lwy4ZvVoYmZP/b85vXnnM4gSyO3gtf9NR/P00M1Dqg/sIBt3aaFC4i0TjStdHiGHDTGaOx1MI1e1XoNKBvM0IuIxDrGREidHQJZohs+FzuRW8dVGKRieB2DMjeCgpTdYwmRGY/ZW+G8bGY6xrV0l5D60/pT4yb+hg2oN673jbxwb8Kv6+ZNT0ne3w7PlCkXdtwaOozMs8SjG+KLEMQKDBY3/9O9cG8DpL3IjJnMb+TW28swfgpZHqBoPQlDTzZXS8+UsFEYjYqIyNceBVDCAEil8XlIdH40OCIsKvPZkXnVokcWX/tEhOkaL+D7da51va8IdcU2IbDsqt8adi4uy3G5jX4/J4MjXVzqAU3cgWH6iE9KFooMtTOaH86ni4gcL4es0j5SGz3kN438aZDhxYuIlCZhoHUC9SlxCM+uNR0myGASalhnbjBGZsKkbe67h9ZC1EKvyIw9W/mprmS6tlfXHUNOA29hM1vqJ0o+hmeAolraKoJ0HBkgzdizUMzfMqNPGmlu0mfPY0OtZawMtMeXTcUL9NkxPLDc74cIJfcxzxaShZMQvsJsuN4U/QQawWoqrjJTHbSuflIAACAASURBVE5XiDhfDnPM9PBIxNadCt8boVZDDU6Lzp0iuJCqML4cT4uimHJkNNJS54DGeYANINI5ZRKNIulCL+9h9GrY2Npf98ivdxUbgEkMIkkw8gDP+QOaDkP5gXrURMJNavqG3jfntHFqkRWVn7VGMjp19O55v5YCIw6/x/BS0Uw5YP5lfri5MEbBtvIADXQgRzROn/JwWeWZH5fb6A9Ecs0wwYxSub6Oh4tCEpkM324OUxBPghq2qQ+fhSjraBQRyWlDFaGXkgFJE3FMczQYTkfwN5h6oPqXTdASJPnozTE/aZN9t+XexyHCdTNOFC4hbM743QuQeKOnzqYZ48ZhOo1YaVu4W2dQOaq6IU4WI7s+sSh4j9UquIj06CHKMu+7Du+axfSz7wobPkucjJYYhVnKo4EcNo2QUVyTQ5/3bYaSnjFTcOSDMsw+jStTKlY3KfJcgP7Wq/56WpEwWycgmsMmWFCnon7s64AIqaaON+snbEwSYO5tTkbPuUddIkxZ75GOUvMqRGVsjPAMGQV2kSq1ouzKhEdmrz5YTl4vLYS0FdOrBAHkMsa408lOxSaUIbiXCJKTFmkmjVWSlkYtbig99QTOiTlN0XmjGsr1Ll/6/HIbfZEkv9ZXDVvmFsnD8+auY6WtGMfcN3VlLeSrgneDFXvj+Uk1USG87MFgFW1SEIoH+lqT+iuBvI2MnmZU45TmK1ARwH3vKVzvBBEKqSRsYbG4zJy6XdV+ww3DLgpo0W96/r/78bDgGQqfIu/6znO+81xmCFkyYt2lC4y4bUYs3rIBzeidSStA2l5TviKVBovlrDXsPQz3WJqBCwbbZ3PHit4i2fTUR8id07PkZlTMh++xV4FyhiMqek+DRehvA2nAdoY8JYVDDGNevg/PcxxNYRPaBEhdXfDhE+GUdV9Mb9omeQIqDUG6LPGIYejZ/Z4VYVCrYWISG5DRLDTR20KHQetT1scjItIH42YfOsG2xq2GJCKpDaCnENoI65Maum2tHRXOUXfBEPTQb+HswJevknv5jb4+9LEHYfC6o/Cor6C78BhIB216OTrFaq4MF5dSxToU0yxVExFeR61NThSdVJwoY3f9sw2FknbceZEYG8hALU552o03IxA0a8qhFbDwW8dn7vlF2lVMHdXjt714Gs+EBUIve+QuDPn3O0qlremm0h48oRfQC6GGrI97odcUjwdDXdhDpyfGuLCIwrrB/dg9SR6gByHNEM0MRwQi3jQzgGd7DqqLOp6d0XgQtdEvoSdAN5gOvHfmwVkoTa4VhXFGXsn1nPi5xu5hbtTDsyvAQ5QVdw72D8B9owaL85DOw+Rs2EBO5lHo5Zw31Sh0VbO5kesnMiZbsK7mIFeYGFK8X3mMnLrm3NvwuLvgeBpMIIrTNcpr7R6AAkNZXHOoLQ2wLjuGvWUkDk+drKPWdd9dzaY8Kai4yoCOSgM0DMaCAuNe2UMXOrj1zU7l23/u6X9TRzQQKeigG2Fm6TT7s/lR8oGHf8lLT6/daAoI5evdd8/OPPUBqVIxqVKWWD9Cmbo2QCoJCuUqcszHw3qgDJVFsNgQgUwqk2A/lWMepiPYA2KnuOxe04QimM7oIb4E2B/4hYxPfm/UDQ/zzU2tS6QIvAinVQ8uR4HzZXThwqO1dNo48q/kq5m6Hq6FjTYlXIt5nswbUwyFIinWqFWZQAEbm6Dl5MmEWkXh0VAk9GLbcB5mJ+v4rNJirPq5nix6nrz2B2HOTX4STKg5H88jwFLN06WXzM3GMPdEOJEuPNb3I7A+kl5i5AVHjC2Ohc19DPWe/WVfH5buOnvo3nn3NuaZpp2YAuvf8DFM1T1047K6jYhIeQGbiQwfVDqz+2YaNEWjAHBAQTu/Ozh/vjvsiZcP0NdRAcJJP1rEuHXQ2JeDubGmrNGdy4fXv9RGX2Lv1rQjpUqFFEEFYs2jylfPSccOy0SiDXnCYUIISWuIUlmLNl83iD686xK8h/N1zduS+ImetjZfMQ1TbwyjLkRETrXFnh2BPRB4WY736k1HMtH4WZ76sONe1/QkiK78tuT6RGiqmkAlimIihsZgATBFD605ffYw5U/AibLqRmJ9NiCQ2Dk7A+P5bTOhRZ9dmeRn/8ZJoNVeqvodDEZR8Afe/eWpcC52Im+13HgdjoZrYOGwjKKwiY1MlnyDWhkBxBYUAnawMMk60+gPhud03PZJvTx6gu/5PRjFBZvOeLz2hdCnEkHYhFPakD6MFLpwam4ip742qjBJXOvyiL//hU5oEiTibBR1GWuEpNANx3Mc32tqCouFd9aWjKqa/SgscFtN6u6+Uz/zHiMQ+3VKOn9Q42H6xqL15gJSWdgkS0fWOZ78KXFKRUQ6E4g2tAgy+kd35bIdl9rox0WR1qLC8dRryaE9X2A8W/vA3F9RzxAwzCwEBrnPU7k/TU1QyzbVGQwUSs5CYbxdRDQy+kRTTd+PiTiPZh71aOkhdoB4IddQAltjp+Qk8rYZeVmGqlYIngB+m14u6wMF9TjzKU52H48s8qo6aglJkxMw7OSdGSD/blxC+zG6I/FbRqFBkRTyKhk9bwM8RFvnbnCOz/17V0fDZnYK74Hfa2nzVRORF4vd99S4zI/78yJtQApFooUedj2TgmBREWc0+ndPHG5LfqH+tJJ94VpShzVCoWZQ3vJ7aK/p95ASKk5APwGcPmb0ebSx4ZrzcEycPz6bdNmee3Qwju5gjlGW6DznZLs7bKJYFE4kRDFfUmlGhAo5RYkNMmiiRUQi3UviVC0EehZ5bcxDlEjHkINwruIrp3/h5tD1vyfHr1381uVuzrq+Eq/83I+KiEhBvfZ2Bp2sSBolYvlPTq7qH7pBaXx3WKRsf+8+RMfgWHjQXBSE/UUZHaScSEWwBMa6GKkUdY6OvyQ9hLCcjJupQ6+XKZWpKTc4h/vBSOQwLtEOCmBzwbiOTbhxl897Kuj0OZB9rYeFz0VXR6ewnbcP+BpVgizdFSF3X0ilovwSDL3CDtJyfjgsZqcxv29prVT38AO/r+VbDmE91tQEU3unJCl7rLUK5G3zS75JdlV/ePQuOKC+040kC6JlNWhzo+gHgaBL0jcBL/fuG178MSpsEZHWGyEtVLzlf0sJoqh3y6IzcfI2Z6MjEPRhztWWfB6x+9wOduyayMr2A/euJ5f9uswrf7CN9yfQY4H0jnUYE2FVYPrG6j2sT+C+DQ6bQulgrfapU6E1lgi/RWZYSwXlUTvqnQDlNaGODrD9FUTaKRui9NPyjNg2H/3Nn/oWbc7q5ZJOvZ7KxZUQbg2uufGil2oTgIui+3HAx6wjFw+kQIUb1Q5N0QNgsRQWfVEksoCprkrcgk668x2U+bFB5RVBMThEez4c9hhFq7KijShW3vl/3DMsfjgs3MFj/BaETSqlYU3Wwvc5tj164Pnmva3g+VE7gL0G07fD986BqqAh7ujC6B+jNR3EYiwmdwvhs2dbkMbD0FvYvXrbieQ2XnP6h4EiR/qTyN+CS2n7wL1Yo0ygGAmhufFqGOMYfRfxjm/SIytqHCGqUweZXkqTVWkrDqdAWo05N7kQPP2Dnp+/soWGp1mkL58PKSRSdKQ8VmsSxIZPvpy+1lAqyFf3qoC1LoOPRjfGLjZ8it7YZsH+gpMTvwcTVGHPQK/PYnlv6DU34eN9kMIplHkUHFScv3VF4pERlNfFwwr+RO2x4F9+rJTSU9mG2grnjCTkEZhjX3SgQ/Qg3EPuytMlW5/Fcak9/ekX5uJP/K//kYi4196Ap/SnT5yOn8YlQeKwuQpeTRLiMk1PhIeKftN7SHnqj1HcXA1GYB3C0Wx+sjTD5hNUd3FZ46+G665DzJwGj0WlphYEI6oBEUa5bApUQHhwDNTgsOA6QFRReYSu4ZlwPcQep4SwFekQryJq4LGhna8QWWFTjMnJiXi7P+s3XWwwVgOpPvJr7Y4hL2vs1Mxnc8POWMOsqxROWCMxbgT/LLt3rcuWjKAxILpU3DKK3hxTdGQX1RpHqlUfzkMKfaPPNAKKhfnmhIgQ5G5VNIUZjz7n9PExDDkgjyZmXz7E3LsGtSrtYI6mYIjPhmsZRLQJdGfJMmt0HNSzYCTd116ClN7FHqJXPW/1Lubu+3xOEq7rnfZ+rqvT7vSsVMPGyrpMGYX1LMWsMxA9Mk1o9orNi+/l8c+/83/51vT02/2C3DsJIeLqWHggNPTTCBlPyf2hohY5SBQKBNUH6l2XIGnWWcGk1rwpG5tKVNNZ8VB4YfwsudbkfUyOXeP2wHzJoaBZXzPIp78/gGIX0QPWrl9+lP3YmiNh0o297fd9dguNYsoX00ddJJoGAmQGRl2b4sgLM3YH572h50UOufoYIiZr2rKOBVpY83HrUf3r61ZAY4Wc1AThus4B0SXT4eyfhPvaxxSvAIvdAlqjtK8Gi3URIEH7lfD3yj7gjPPI+xoGHRtBaQu1o7xf98hO+F59HXh3bGzWc8KmUPpgRaC8OgpdtOciIin+osJsMHTdswvw7FrkJ33EzLQ/j71zIL7Owvdac+gM/5obNHv29JibcD7sflKbVmfY0IuIVJ6Yd53dSbz8r8O17H4Y6wtzMqcRWe8D0JVGJFxEf0FvLzga5Nligfmtbuj12T72yOyFBafC2G2M6TnRQwICO6aT2woSIfX5ZTkutdEv5vqyOBqMqnXc9tFFeBxB7hB0ATbBqGpfQX7eqHI7s9l5QsPJX1Q0GyVUTh8uYZS10nD0RJm8Dnl6jvR+8FNUzurViC7Qv6EkkLoHnXSnL6LWAQPfVa6X0gG8+wfucZ++gAhCm866+P3GEu6rP+z1dCbhperCpIQc02WEuB6pdkuui7zsFJWrwrX0UgU0oFBqiok+B5Ru/IKCpx4LX/RzbXwfCvPqnZNTfQADX1TMfY+RBF52UfiLc7bB+PvFOqIZnbMFBEvlDWw2M9jwdU5U9gkNRhOg0V7jufQQvVphmog2HvTK7ZmPY5NHLVom3lLxF3TH0wk221faAU7/Omo7933OZZGzxRR/f05pTtBIlsec6mmEQvbRJiG4qNdYXwE7euernpIxPQpqax+0PBoy6g3CajsXoKke7IfI3mimL9NxqY1+LoqTVImpQrFY2D1z76Oy7l6/4cmp1sPdeU/fZy6WxR2DPrKIyQIdC5LtSKkNiPkvDFMn97dhqdHUEmXYpsIZFi7avCMtxBaw6RRZENVrYKFsdME9oFOVTpz+sHPcM9ca3Xdvr2Myh6SKYd+CjRdyyD1gloubOtkns43vALoIeU0xdadpHf115Va4hxHQDuyBybQzrr8FKt9JMpFueK2isxrGbusmMORYmNWH4Xm25rFxghBt6pUAs0wJs0PUm2nEvuXM8YyaSL9UtvW3rsCI3fR5NDflBsn0getVsGSiPlC+F+6htQYvFx3QBqFlZy3RaxOrDsk0fp+T6gV1KCWNy6OBjfc40Pk9WPBrofZF7aOeUrHf2tzxMVxecAhs/kaYU3TACENujYRrOX8EMSFE4jxMm4IpLhr1CYXhktCQ3esWNSzXfKxyKMAxrZtsQoVhB/BZH5fa6Dc6JfnqZmDZm1ABhPI8qE6xu1+bdUNmkDPykfPh2mYwDWWfrS+64MPpNYvhkTMFNIt489KVMMHoPbC4mRSzGJIibM8pZKwHdsZmNbtT2IrVS+PuiXThaRDfbAfxzbYxERPdpIs27os0r7/bZ1czrqUyGcaQiBnS3x4eaDEPnZ5descwjl3FRefI057BpTI/hhB+1fn4H7RDUfcqRGAYdjdm/HnYhk5Kh/1d3yRbL4YxJmUE03zPT4Vi8gkYXB8zgqHott4D/0YOGdFaNNSTpb7r8/TGVQjQjIXn/KWzq8nfCCk2j5kRUIoX3uoKWS65iNSAnbeNYfKqjzEj2aMxlSg89XHNz4M5UzH7y+PZnZSjRf9sVZ+TpW9FXO5UxJ09kifulThi4Xj9nht9khfm0Jx1NhHGnsXu0zchaq+KWmVQpixO+j0YPPm1LbcVbJAj+mxuOmzYWXWA9+J49JT3LrXRjweRtDRHaR5z5xhoESBbaPx+/2HAxn7X2n3/bAb5GoU4elfAIqhoDlIkp3hK5vzhthRTX0BxiDw6SfEUk4/efVe9IU7EuIjPwsOyCURsPGlo81EwRJRYvDLphtC8qsIFaSvys48vB2+mMwI+G6QLbBMmzprn3dfxKqKlnpxE7IsYW9IFwq7KTV/YS6r4NTfiRp+buKUmGHZXgONfnnLP7P5mQDt92xXvgm2u+8ZmvExjEEZh5GRarQ9PPCpKRZS4BkOPDeAGkyXTlM5YuFy+4oaeoAVrTDP+eJE0VLRxR3Vt531uMscc5Yc9TjbWjaFreUyN8tv7jgxbm4JwidYF2pgPfdRFClpQZ0cvn81re4vJ65fmQs6cRp2F1IniMFDg1YcOa/3QtWDeOJ9S3DmMqnUM2Nwos3BaCsPrYuNL/ltGuWJd9CIix5C3rFdYjA6fjTN6Z571camNvsSS5Lq76o3l8EDZJcumF/P6u0hE1uBdWH2A3iCP3JmGryia0SAOwNViCIfWHIw2NqMkPUIY5pxfS6TeYA0iL2dI2o8CU2/pDV73CRp7bGNbX/DwmY0/yTkRhvL1wbhPYMNqNxE9sAhoBp6LmcVso5/uED1EkWkU+azIyI154qbf92otGJyUQPmYe6HNG+EaOS583uyIvba8r39Dug7Gz/oeGK1dm0c6TJ2LMXh16zU3vjRY2+fDRTzi3ec1fTNaBIiAdCHoPDUDXv4ev+8R5OfPFBceAUUzoBEzcABqKUR5NQCKsHEeG/ExHIcB34zCfVFIJl70sTdH4wRqdeto+GLE9qQe0jqz2NCZXtlqhs1sDg1wZuhFfFyoLVB8CBpmCORYZJQbhYMGCu/SvqbIFrF+kV412ooB7quAFFtcRH+Ajjc5gy7LcfmNfi89aIyW2MTEDsyjs2A0m1PZHbmLoyFk46Ik5tnmXA7deqRbHX0ImoXrSiyGlE8/j8VgBU1AH+NtYK0nh7uDY9QtSLVrsLs9SM+1kFs2amPSS+xBGSjSxqPH48gbX/HFSNGY42aY2A3kbZtbvimcqidPvn/m3AcPwjVGIPiK80DMILds3bdESB2DimKnGFATd+47Nv/Kuhu/DYXDjl6jl4vFGg0XCzdPfVwYlpvGQgc1mCN4/Vagu7XgDV87TYfwMuI0A89I4OjQx3ByJdwD89VbuC42ck2qbiyjoTZTRdYMhzlbQo9FWxFMhWOgomZAWIjudWuMY0R559C9/jMTFkFdRuAUmaOxse95+vYM5ikM/N0nayIisnzTozHy5b/6aiDdunLbUTTcpG0M8pu+TlLgB1Kp6CYYo+eGDlprMU59TkSktI9itIIDYtRlcifDqdrwRY0qytlR9bM83rXRj6IoLyJfEpGNOI7/chRF10TksyIyLSJfEZG/HsdxJ4qisoj8koh8u4gciMgPxXH8QM/x0yLyIyLSF5Efi+P4t576ozlJcr+Rdn4ST8wwroiJ8IGVELqnmjigmfr/bgTYZxvwNmL6rUCWCCWLSJ4wR0TKOW3U6M0gHw5MscnIpVAKnAd6suYpO41REAXxl1EqN/A3huibvWAwUhTF7G5USoTb13wBkdSrDQPftgYbeDJETZhHSk1YJhD6i/pZ1ASY6ulivI+tUxJpDuLZ7+u10NA/2vZc7PRC2MRJJUDBlz2qd9m9YNNhBGN/j4H8Iq3vrOoIN2GQ7z2e83tEiN/VBrUiuFxIsb15pOkd1gHw7CZRc9o+DxsLc+vnZ9CA0PlJLpg2BECMFz4GHJjzsAFiPDt4XUxx5bXJ7iKJxK2jMA9JfU6qaq5L0+PdafjGOVPxZ2fGnhsQOZqMcJCpQxpaQnttjGg38qdAymmKl8XuPhsKNYpKzVPME0FNau63w7oY3X420lkPn/LeN+Pp/7iIfENEzBX5ByLyj+I4/mwURf9UgjH/ef33KI7jm1EU/bB+7oeiKHpRRH5YRF4SkWUR+Z0oim7HcXwxDV0UJ8U9M35d1HDyT7LlEO1g6uFqzVMeNoHudn2x5oG1Nlhdj3QIOH8KSqcefASw9YD0Dup1lHfB84P11dfuX0YtpHFo4XdN+5ae41l7GBJWP/RFcWXNDeWBFidpsBgB9cF8aQ1g0QFYMuG1Lyi/EUPxFO+MUSP8pnt7By/5NRYPASU1o5jaWJGG0LQQFzvzr4dPwm+U0InZhrfIo6FaxUVIBXZPWCcK981u1vEqctN6v7tn0CFg9PkY3bXXwgaxDw6dVOOcdXjCG82h4YmqZkYyljoXN2SjTkZgTGhu6VR/C9+pX/PnVasORyPFjBy1iEhPC7gR/BRDIomItIoqRDPh51yf9oiSc255NqTuGJ0yLbSgkEpKn3ZZ+9FoajAGWC2cC2YGbEOk8EmHiDF9NvEFinlJIxlBGYgSGYXtvRLOuycXUKr8uz6e4k6/K6MfRdGqiPyAiPyciPxEFEWRiHxcRP6qfuQXReTvSDD6n9TXIoH25x/r5z8pIp+N47gtIvejKLojIq+IyBcu/OE4coEMy5XC0+8iTWHhr4jIw9NQZKuBWXOv6YvUxD6YzugWfVK1ZtXbwwIineqAalDqOdGDisCyWT7S659gmgP3aDlYeCc9dJaOYuGMK8VvSgIxD+IyzQ0XkcesIl/8pKn5U3jBlAU8PnWulIGph2ExsahsBVxeyzSKyubNbX0QaS2OJ4UodMOLK1iAWHjWQUm900oV8DhlT6RSEyGVKZUrvQdyshTGWczTXgZEU7sQ4p7ShibmpbcogoKCfltZYOMMTncRkdgEdPDsqarGArA1ETHyayFCsYaqHvpV+vB+G/p3Gv3KDjYFQEVjjbKof5Di8dE8tbFOiqTV4qwmNnJBUZm1tnNNKxHVlOIUUkPKugfZPy01ViDf1QVpdHPAGA310dltqRry/af6UUylCxoSfbKqUg9iI//Ua3mWx7v19P8nEflvRMTcjBkROY7j2GbgExGxVtkVEXksIhLHcS+KohP9/IqI/BHOye8kRxRFnxaRT4uI5GcmE7Ub8+x6yMGRL5xNEuUMb4yG7uQohPup7kfYm968PnQqXAGDnvLs7DMs6CDX2VTBhhy8uVSKWXOxlR2fPK2b0FQlFLQTvJqUVwQGyY3DYLTLk/79vYGPgRlFLkDmjfu4bgtlUwVo5OQfPAjdi5PzjiVnPrqhtQgKfbN5ikbdJDFRB5YcGtS6yohZBJ9O/CpqFdPhN87LSDXx/ERG6YLNndOIseFJWV0pXwmD11TP8uQMjYFIRQkQX2Zgc8gA9JEaSBrn2KyK2s/jGY+STImsAyU1RgjNK4pBp0NCimHLZ/tPSasMBwoRhM37GN3U9H1NxKSJ1ctagon1HO5ld6OyN8Sup0bNAszPU93YbO6LiBweQzXNGFfgALLAnBCfiSTC51TOJtFbknaioe/SU9GmTRSNY/opvQwL/2wQm089/kyjH0XRXxaR3TiOvxxF0cfszxkfjf+M9572Hf9DHH9GRD4jIlK+thqbUTUxcfKzDOAZMp9b/D/DYjn9zux8WlHzqmQkTDVBNYwTBRj5132o2JlqTUYxFhtJyvpLSuBFtj96z7pwmyvZFAM89pVFk8pYTDuZt9OrYYPB93tvhM3u3hwIxKDYlep0tFQP6w9AIhjM8uSRj3sZRHR9bXnn75OStj9CDgH1Qkm9TNUm9Ry5AXZBEWDPqfw28PZL+K2M1B8hfn0CnNQod8GkWOAi12eTatCD55fizjEPntQLhOBOKWoJ38+fwntmjcVEa7iBAYUiGiGMvebGkXxO1lVMHD8Fvpm/F0ttAWbMtWY5kzw25j7ej/Va2JFbH0V6Zg+IMx2bM3Q9c062lKgwN+Vzs09KB1tXbBafvID62KKdIyCcwLdv900ngeye5kiwZyfF70XSxeFWn0tzvBtP/7tF5D+Moug/kCBkMy7B85+Moqig3v6qiGzq55+IyJqIPImiqCAiEyJyiL/bwe9ceCRFE53gzGOSz+MMgtDl5QyKgK86rjr/cij89bbQvILPVrUVvn4DkLiXUcREcdJgXL1lKHfVhiltU/C5DAoD4vgL+z6pWov+2fHZEK2wtbyAcL83p578LjRZKU23Otx8khKRxvUkaQCudaQhWtr2X1oA7XAn2/j5BWKxMmq2RUQseUZelZS2KQI9/drIDiKY1eyCZRJtUHweWO3k7/h56tYaUyhD/JFHQDhdAfxS0zqpCKcOR0ONE71k5pgjaA7YXKcRyR8izaAR1dk1GCn0ltTeVGK/aww70D+AuoYVfXtwTjiG1cfhd5srDFn9MA3aFEUINCIiFkc1pcJO+xaimbg8nG5rMbKycRnJdoTouJl0YgxDn98HSdpaWF/lLwF7j87tZE1gs42QjhNy/mj2L9f5FuzIjeP4p0Xkp0VE1NP/yTiO/1oURf+7iPwVCQieT4nIb+hXPqf//wV9/3fjOI6jKPqciPxKFEX/UEIh95aI/MlTf7wfJeiP4nzY/VN0x69CZOIFD8cnviPAv2ZYfFoEkuBthZ+RVxs5+brlN2lYWIADZKs7ohOB6RtMagv9C8dockIomhCaMaJEKimPyXymtBF5eIs9eE3FWeUWH2UA64dFNh0ojrFpp8/Ctf0uvHt64gmxF75fRU9BV5FPXXTk0nAUAR1MeHb4bLG5DxAhZL1fPArnOv1eNkmh47dF4zg83tRRtfvJoWjdJyWEtvLnAeM8zVMdG2kOfbZElqQYOa1wOJOx6Ui6KziyTMk2RL9nsiJZPC/cV/2mfja12cLRmMezM+URomDgtLReVJEhGjzOX/1oHh3erJ8ROdXaUl2GJXeqZmc8ZWjUIvtA/7BzO9LX/X0fF252zNsmcx3zjCyw0Z3wHJsfckemyEhanwdrDsUp/2wD8GrLBgzebQL9PTz+bS7pp0Tks1EU/ayIfFVEfkH/FKE5VQAAIABJREFU/gsi8staqD2UgNiROI5fi6LoV0XkdQkisD/6VOSOHra7dtXrIWqjt4YmI0yEWS1OMned4pB/LpzjpAnx7B3AgszZ46YwyDZIRqXL5i125w60Jby/gEIUxMJjFRlJ5YWZ5sBrE4oZB6/MAMbHKBe2wBJIcrijNwLMMYaxYKooBmxvcj2gKthBer7pm6zlNSka3oDGbpISSVEv4PU15HD12aS4YurQL1Bvb4qsqoCKljWVQ1jfGBBOVOEyhEUd369CNKOpWgvdI6CisPlbrwQpJ4qInKjr0NbGNqal2ufD0EgWpWtoiJqtDjdM3Y0ccUY6cBNfJ9Q0JUqv8M0sWmORtPRodT78LtWoUs/GFNJSoaG/HFsPkTTXH2GQJdCATH0wjJd1yYukpQ8bun5euO0d1A8OHK5rtBpHp4TEMRzCGra1hHQbNYW7yssUpVI2qBtquo3Zhl4RRXo4GjXFTBYbz8bTv/eU974pox/H8edF5PP6+p4E9M07P9MSkR+84Ps/JwEB9O6OwkByC2HxVr6iDVeLmDwowJUXnCPD6BXGoe/KjkIzZHvwHhiC1x6G1yfP4+GzyQKpi7wqSNF7p/ctamCJ4CCNQ2y4bC6gYraHZYW1I0zqkVE3Ese6ibW23fOcv+0okyPLRwOOSC94FNTI3RUN8dk0xgVkOUvcFxW9DKNdYAqCeXTWJZSWd7ACxSaeV40qqQJIoGe0AGwOmxnxc72944bSIsURaLqyMa5r/O2M1mBcTQXrHOK/XTK/Urhcnw1F3FsN4N0nwzxot/z7TLct1tzjNZgtTUhuF86D3kKcVUwUkbx+tse5R91mMr9G9r5//xx8NhaNMC1VwPrpjoeLIdSVz46HdYSTrGx10hu1IqV/YIc10Vi2+bJ4Sxx+bwqpGIusSE+BaCW2qiyK8V3WTWzTYHSMVC8zAMfPZ0SU7+Xxyxe/dQmDDxzdnMSbwZAVz8MgtrDLdsDiV8FEmB0JC7MFOAi5WqyLbxoCC4cNf785r9CuU1AQw3g1sPHYkWOBDOvOuFLObyIUx6TKqQQbI5j+thuGCpgjC1fDIqqWs9M3U+otlW64QSRePadi5BQdJ1JiI3YP6qbqBBAbX1nCeKuBzc/69wmrO2mH57bT8HOW90AQNgLe+fVwPSRB61R8MS1Phw2d0dqVMcd9v6n9FgswkuRJX54mK2K4XkZDSyDVatXCGDHCyeoBYVfoCze8NEUqiiyajyZQZKetMEZzo/48ts88mmLHrT2HKegbC17vb4f7yWFupdBY6rRU7oOi4LbPraVZH6NaKRjYlLQj1KS2tcub0VB30Z/9dcXks4ejBAgTcfjnKjxCWgx2KDd0c+UGxfk/ptd6fuQdw801GGp47VYfSCFy8FFDUxFBNUCncnVruNZHCGxhG/UBAwAiu3lZjstt9IsDEc31HdWGGSTzyJN3Fv31xllAlNTgXWQtXB5dUABHGbqWfXIYM0LW+dcDfzu54rvjihiA9z72GuCZ36k5SaYgyBvPgqJ6kWyZn0Aa41QN7fZDaJMugSWwP+wFMsXFXGnd4KH4LX7/VLtBWWMpIAXQ0XQcQ+VUVzKhchlFW/YnmLE/gZdMo2vXeJGQDZtmrMWfGww/u72nMpGPgDZZQ1QwFc6VwyUzosxFbugNY84NiILr1mFNo8+0Ebul7R7pPXfxPCKlESlD/IV0BCZaw/pCtOX3WARfU18NMJugyMya9DiwEeyJG7yTRXXU8IxoyNk0mVwfG64QRdnzYj8IHRV75s11ABaQwiLaL86Fv1NxrItIwOpYFGZnJNurDq+fsfv+t5MPwt4oQqh0fPmA+pdaLrF8ZTVe/JkfFxGRyoxxj+CBAd3AUNcmdpwhZiIiktdmHDbzMB9t3tKAPQFI7+QO4AFp+oYGk3n4vBbeonVQQjN/n4HpipD/HBCxYoYSSJ/SEVMynaHv00Oy0H/k3rAsoki6oJjLYGXMQ3ClMxsWy9hbIGR7Hp6tFUQxLtWv+Rg3P+BeZlnrDs06GCRJzhalezVEJGFfFXHt4AIF4ZlPztggWEw8P0Y6zgzGBXKLJmBfm4ROMsa48RDYdK0VUGGKzodh022jeefBZ1+shTHqYc7TICUYcjQpDaC6lhQ3gZIpQFmOhVRDhE2ABPAc9R7b6M9P4MYCyFBdCNFME810goZFbhaWfy8CqtpBRGn33QWNMyUpDV6Z6ibPKKaLiJOygZQumkBHrY5nacz/RmiuodNIw8BaRwNUKgnt9TPi03/0X3yrCqNH4l1wGTBH8okTyWDeMRc+m3UMbtdAdy8fToJJpnNPHP6ah+gdzcfS0LN9va9yiKRYTbV52z0Q54yiGqGctomRECq+4dcSK+wuhrBDCjWkXk0TQhupohdSAwPNZbJALeVhQ3gO4821bAaLGrzna8OMhyIiDUshMVeK381roxQNPZPbRo7WgfGlh9W8hZSIfqZN3iYUYrvmlTOaAz1ErM+jXnfjXgZsdfSqp0mMLI+keYQTWuRGCG1zF1zwYxnonFyGERNvOuM6SDWl6YafqwFaTC+YncD6vROIsBOFlWg0swEO19owcSJmQWn8Up3X4Z8O6MQ5D+x3i3Bu+qu47wy9a/YvcC0m54QgPCMr2yAY0abQghnZAjqhLJJbKrM3THL7zI/LbfRFksnSVewuOVHIUdNfca+kNhZen2OxkR/dml7a9FRgXK3AFtMgboHFbxmFXP17D51/HeYM7RyAhKYEyK0DFAtwuGVN/2yLGIiD+JHXIvI6Bn0w/zGvaymmCCIuccakF3EjwWJdjh6ULvzUBgZxepNJHMB4sxko5ZHqS/YBcBF3VCGN8NUIvEuJjjB4a3qoNfA5xuoRRqBeoAKbeb+DZQi6k2nRSPjoRKB42kK3qBW2o0OMC9lWdbNrI7cuuMfRMb+GuipDcYw4/625MMe5QeyBXksPRWd2ODOCsCioc4TU3ygK23YOtkIggjBPOw9U1AAbCGHAA50TpRmgrRp+jZXNcF2tq+h/YPTasLSWbwodaCnTUOe/EZ5ND419XOI2l4mka3efnp7psQCOCKSo0cTCn2TLU/67Pu4/5b3LbfQjcW9bJ1WqGw9FzglA5UyxZ24RHN0oJO21wsN/Mupt7vsHaEPXSVWCl9CGODYpgPtr4XeJYU/hl3UR0usavODXVdCJxkVXAoSPaQjLTbcQ6o7cdnELowjuTcBrw8I0COE5jPfknF8LYYwGxYzAFVMkV5GGsiUQl8kiNj5duKlQnCm4WURmeo+5Czx9824XZ9yLbv0eVNG+I/zGHPDdVYi7sG7R1R6GFhAz9J6LN0zQBTWBq+7Jv7QUlLNSLJsP5/1cGG9Lk/S8L1BG0MtgrKaDDBiniMj1Kc+zt3ROv/XE6aVjeNdlTTOQpbOIsTceIEuXiIj0djyqqF3xsbXrYlQwgjnZ0LFh9BujITCvndklpPbW37+bvObYWcGcdRHSU48odTnVtN7Y8fEeUy6k1gLpLxAVsBP+RQUMiB8cL9sf6KikGgbtZWofwBiQMuWqUoOsPSP4zv9x8VuX2+h3IynsKdbZ4H4XeMFM5WQVbfMZ7xNdwHSDUbO2wOnOQiw7T5MJAkNKJI9R3zDS7WCilayRiqkoejJf9Xxv/31h0s4CntrGBDeCOSJuKFn3jSdBsWh2EfDWV0GyhvTN+PWwmZxhI+B91yaD8SLEkF2TJnrDFFkfRHKpJqaBFVeResD7c6rfSm3Sx/8xOpWVD4eSlePIR+/XobKl4zwKyGYeDoNx63SwKawuOFJopRrGhWIp7RUACoD6MQbIMvo2qoCV2tiNQZOA9YfVqm/oRlL2GI4KUw/nlk+mjXmEwvcVRSXBiyYMmeujopt7D84NN8kEpjyNDR/9ATYC61M+boRM8/VsKdz7SD4bkdZVLCrHOwf03EDH5WvfuOJfIuS54c/DIo/ogY9LbQf1mBVNnwJanEOtwUAJeb986VWZFmbj3XBn92U5LrfRz6Eomx+GURagcGNICBEXE3nS8wVCdMCeErERUcCCqqUDcilqWXw0BZbW/2FtlvU1LQqx+NNBy7tFKK2yP4qzPfQPvOhe5ogaDxo3IlOMYI4i7USAjKs4NTcKFvDWl9yzNO+YBbIciNzMOOXg/XeATJm6GhZ8inQLKbQc0nTWnMTC38jMsEzeZt03wDoYJuuqBVud8u8cNuHFgszr8GSYWz+XUfhm/wN7Au6dBWjgScuvNSWkjYjR7osIrN17vsnOXw+KXGwqKwMl8+ape7SGpqKhp/hLs6DcVLiWCJQLA5sz7PhFiouspFa8pPORooTWnpSIeXqIEFW0rkAUzt65z+nxClI5uklSz5oqePb3PERx6NTZcyZTao+R9jhrffq9a76m6mu4L53rBTDqkqk06bCG1TT+JBFJFbPzRg/x58Lo/z8OnXfG9zFxzx/+6S20x8Nbe3gQ4mkukMHrvhjHPhQW2xkLbCyQmZYmkTNs6MDDLTSGHy6NvmGZx5c99dCtDDfgGNxSRGT9lnt4TEvtQmgiuWwuAPXwuZkdQkBkeix4VQ1I9q0uQsau6tdo+P7ZVb+WaTQ82cFQfVCB963h+uG+XzObZmhQLF1WmhruQOX17u25IZ8CFNUil6NTf3/jxDe75WXfzIxGoY73JybdwK9MhGiCylqPoQBlKJLJBR8rdgeXIKJeKQ0XYvOrft12j1QnY/fwERhUG3vhOV69sZP8jVKWVa1Z1Tf8utlMZ7UZbkB9JDq4fo6VhbaGVNTMjKcBrcDbB3cVkXKWHnqMTZz1iSNAb/c0CisjqmCkauIuY4jM9rd88zeqaqZBc6hZCaRJrRGR2tcpXqQsr5wkm+r1GwxbRFJNhJRGrG5r2rZy+Vz9y230c7EMFGrW1uaQ7e/GREYudHrejZO1Q7NhS15xg7WjgtQkzaIht5CQlLhFKOx0wY2f1znVASY6ZrHZIGfsbIXXZA04TE+xMYhG/1gXC/PNKdlBLUCxHZxt/8WJYKiMrkEkTVlLdbHkfUD12PdgjVj7DV+gFHdJjDZ52GH0I+T3e2oQuIHRPzIlMBr6U1AbL2iunxDcKhqXOJ52BUsL2FjhHVtEWIMXzd+a0g2G3jmhop1UETA8W1IQMH1jc4LPkEdKvGUpXCPTR9zc7b5y0HUugtytWxv+jTxglDFqVub0cP1QltPSVdO/6+fc/E+H61A9hL/nKOTmYeANYMHoleyiNrYpyCgicNvESdKWUs6CBm5eNz6ua/bXJOuWEF3Yhc6kZh3A7ROhcD4ooY/lRWtrlkt3XGqjnysMpDYXvL/BTBhEYrkHyI1Tjs28yB6MAA2KteDXsUsT5mWeQB/UCoOR7KfX0Tw4qZVT0EO9Rha1qnd9gt4ph27SQQo+inPB06gpuVPzCC3x+eEIpDOHFAC8NZMYpDdKz+8MsoDGM0LRjsd7XpG0cJ/t+2zkSkHd7ACag7KBbStiUzxjH5hnjaLyEJ8njUIzA3t/eIQ8/i4K39eD0WYE1L/rc6e3HM5rAunhXrCJdoeXTMJFIyLTkx6t7G+oRwrDMAaCLqO44FiRkoHoM6tBHMP7T3H6KBItQl2mTVrgilErI7WHjaAPOUS7h71Nj3DyQO/Y5vr43wcNCrl51ACPjfvcqz9w77y0js5prZvx2VHGNKdNehwXomu6xn0DwjUBTbP0mJ7RP6FuznVr6DCmb/Oj3Mwuho+LiOSQ0+9okTy64LPP8rjURn/QyyX5WmMMJMRwdNEXWIprXSfFKUi7Uru3TjBq2XJhd807ZbGRhWBSw6qh4qZSuotOSiWFq8HgHb7s1700FRbAXtENz7W5g+Q1uzkt3C+OAQpH/L8uPHqeXCykYU6O15D2esURFp3i8GdpnDrRcKMKUzbjI0oJAV575vdzWEwV3YToBbvJdYOX8tgzCt9U1iKRXG/Bx8vGo4taxeQLnuIyTz+lCZsaz2GkDZEtJ9CtrSlxWeOB3/eZIGrQlMk5NtsaoMXHqO2cKgSwgLlXxcY3r/rBT95wdA892oFGokxn9EACyFrA4XHYEMeA7CIHk80DKm+1aj4uIxmQ6Xga9YcMtFKnnb1WO+3i0Gcr46i36Dyt4/yMLvOoIxk/FscgRqHXwAOEj/afQCzHphzgvDTpPaYvE0bPoct/5selNvqSiyWnHkriaWCx09C3DrBIlXwpR60DUOVG9ATsXJjUlVktrsLjSHW5EpuuizFGwaj/vC8Wsn/aUQKawww5DeoBeIAYVht6hh29MQpokYadFEZP4ZR1InY2/Py520yDIBWj3zs6QgRFiUENsakjTMzy5l4oeNYAmx0HSuV036/BCrApcjdsIIYg6p9kd11WV8JvDIBB7x6CHx7SjLmb4bN5NhPBkCdYbXZK474NMpyH904jxrlhG2Jhmd27ftpT3VgokZgSM4ERyWkjFiOc0yMaJDVYS6iLcJ7qhjlAP4sQmogmIov+6tCoyGMet22MoB1QRc7eHBGCABjtFLb9OZk0aGHW7ytVQNbnTOeCKDKb61wHqUZL9lvoOeIaxpURaQYdSDzPXgOFslJYnc4F4Z+aOShkrP9nfVxuox9HCZKmqx4aIViVdRBl4eGZcg4RInnsuIZBHwHtcGPTjVt3OpyrCA+uv4kFBm6cvMIzyxvuZTafg8qPFruWvg0wyS8566MVlZtIjbAgyTZvepzJkaGidVFTTVFpEAY34WUTfQOCLzPA7BZtb6NbVD31fopSAim0uWDoSMdcACKnDG8taV/HYs8jbzvQvGx+OhvWZ7ntEgjdulOI3GYwhrqJpqQykQaxnDxTh/0M2cFUvwiOVBfswOauj2t1HF6i6mYOEIGlFLlgg4zp82zX52mhNpwiGzxG2grIEmOTZDd3Dh29pFZODCy7h4ni0t8l5n8MlNA7j0IakRtrH8XT+CqQWeqVD3bQIzI5nH5kg90AjYwVjXqp3MUUVkqxjrQV9j5lHrVRrEP0DxtydbxiNCH28ZAGUOyyDbt3PBzVPOvjchv9QZQIoed0cEsn/hBHb/mkq8OTKKtB4s7bJ+RMC7gdGNTSvHtja/8kvH/nP0OeEA+UXcF9TXN3BsQxwyO9otz+mD3VDziaxApY9MpGRsAHgr8n7e3kRCdVhIay4KaS3CFw9hOYwXZ8yXOto9/r6Z1zCffAYl5nani6MAJKCbLofbHzlV4V0Rjm2dEQswZyrlquKbUqwDMttTa4AWPCzQgGo28eIXK5WW33KYpiGPi8SmEyEGDhkF3BXU0vMh/O37Lokga1OAEKYlyjNcF1jhHBYP5WFZZ6top7pSHX55DHM+L3W6QDV4rg1Bjgexapcn2RkM3y2KQdJmUEu4JNMcuE3UVEOkXAP7fCb7RfwLM9Qkd7TqGsmC8s9FK8pWERNClPmD2y54zHmUKcZXTXj95DsRonS4Tqaxlr7hkfl9voR3FSaIm1kNrOA8+OTsuUso+mekbnPdTt5ofTEJSI42fv/A3L6ful5LBA+mfDecY+GzNS0JPwu/f3nGJ4cswn8MFh8NxYnKLxTGngaqNajzxDGd5LBUyLLRQ/jXEwB9QG9Ql6QKl01GD14JGzU7Gn3P6kkS6C76ZzRRthLiB/I8zRvFumd86PMkhLsGk0wQVjo5kSrCYBGJ6tAQGIoGohMsqf6ffQoMOu4r5tuCTyAmlXDINkaYAYRqgDXnrLN7P4SkPKdIP1dlA4pYX5a9FIigysOhwZsXDfQQTCjvLoQA0pUhuseVkhk5sGu54rSrjWAuhCCFRgJ6/137wPUFZ0pzdz2n+Ae0jRdWj6J2ZH+zS6eynhqYVapuboaMTKscTUIfWszXkgd39KnhIp5LigDYen2RHhszwutdGPOpGMPNRWdV1LZXeS5bTsoezNW1vJ6+fGg8eau6Ch4yv7ayIisl9B7vweVKH039ENeIAoKJ3ehEHQl+SlGcATGGgeug/DcPIV5/6O18PCYlE5pQWLzSB/LeSja/Be2HRmNMtHa24w18Y9p27omxIjHHjyp0gdzK0ESCONMz3mSA0CYZLFK0hN6LnGv+4L//Q2oG6TPh7jVrzEuY6wAVxZCCkw0uuyVX/rJBRKCUWtoeksVQxXo8jGJsIv61pU7aNGRC/yu155W0REGoCA7DfBt7OAusZhiKKITJmaAyGbpvRYyKXGw0fmHiWvDzthrr6+74XaJoqfnSfh/ZTBqgynAylmXjlGr8RH0ZinuXzSXjDbbXWowdc8Sjxe9/FcWwvPa2QKFCEQQZlARy47be2YLvlzNOrk877f62kHG772iTx8fcm/gzQkPfWSet+kVi5Qt0d1NEpHeP8+DXwYz/IhQCPw5ItbbM4K/67+C4+e38vj7ae8d6mNflwQaWnXbaFuXNj+PqlhR6G8c9gJD3227AZvs+kTdF0FONjcEh0A/jYbHu7pC2iugUdA1kXrGCZvfL7J2F9DXYSszDcnmGjAKBsoQlLOraf52upLPpHYPn91PCy2naKjReZGfAyseavdyX7seeSIl2vB80rho5GmMHw0pe9OUIuwc528nPlTKSOS6ozWYwEG5/pYuK9XDxeTv/WA5DHFrMVR9xZ5UJXJnv2TusMRWbQ91nsYW3RYISMUM0gtGKtr4462OsdmsKSNWOww3Wn6sxmovvL4LCgjIPZjhl5EZFzPQVbIiVGPGHdGgiGOLkj9JTnmCRQeyQKNruWXbwdRmCKcpjbaUIv5sJYeLIPsD1GWzcMyePMP2j43uGGPFcLv0kGbhCU+iMNvzOZ9M2TvyqkKsqSQSvBTSDrXUaI1kgjSEzdtjDY0qgfQ0bANpIM0acoeIQNl7JobfxG8TO/l8cbFb/2ZRj+KooqI/L6IlPXzvxbH8d+OouiaBFH0aRH5ioj89TiOO1EUlUXkl0Tk20XkQER+KI7jB3qunxaRHxGRvoj8WBzHv/XUH489T21zrj2NXRiDTJWsG7U9ERGZL/nC5evdTvDqZ6BBerAKVizz3hs07gzdkN88Tl+fiEh3DPBOnYwRC3+kdDDKBxg+hq85RACj88GAL11g3GyRnqC7l0bfjqUp/z4FLe6ksNThd6vl7Hy0qRexqWYG2PYDRf0UoTg2SNFXs8g+bPT3wFbZnghwxBsTblz5vK1TuZL3a63B0B9Dqcm+R0Wx4zoavVSlq4HCOgvceRm+VtZruBnMqdPBjZkKUXa9e4gUFqr+bFYr7inX+8ONc6Ml1EusMxypIjZqWZ/J1NfQ8fshn1vcMA/bwdA+N+bdv8ddHyNzlihuPyBZ39VwLXugAFkY8fNzPMyAd7CA7jU8El7RMRjBs93HZjhXCWPMNCNpFHpgmTWoJg19bxTpU0V5MUJqMc1n6VGq92EDYCe+TYn+t2hHbltEPh7HcT2KoqKI/Osoin5TRH5CRP5RHMefjaLon0ow5j+v/x7FcXwziqIfFpF/ICI/FEXRixJE0l8SkWUR+Z0oim4/VRw9EomLae6dFjsH8U2G8Ke6zXbxFDhpmhoqEhfOiZJ1lMhLj0YtSzstfsEv5skn4PXrBKOaVqHhr1vqIbHBJ5VGwZwxugJKPx6jpd3SG0xXbJy7ITdDzTRJShKPk12NF2kBSKiWsFgiOifc0NIoHKvyEaKpIjqBx8OzI2cQ+w/Mm6PHTu/c5P2YcqHxnS375m4bwK3JPT9/FeIuOh6kQOBx/zxw53z17SvJ3z7+krtVJEl70gjRBFNwJBszL5fpoS9trCev86sYT/WUOUYNRKoGSS6cDUehIu7dJtqtIlLc8zldv4mGKH2Oj5twhHAY79HYPf/byXN4P6OzuwfvfpBBiHgK7p0/fvN68vr7X/pGuG5sOg9P/brWxsJ40xFL9UPhdVHXeAcqedws+jqGJczTzjTG0JCEM2zpxboGQqlBycZLdvyZRj8O0lrmwhX1v1hEPi4if1X//osi8nckGP1P6msRkV8TkX8cRVGkf/9sHMdtEbkfRdEdCcLqX7jwx/OxDCbUuJiocYpBzy//FARcv3fndvg64G+EPhqBFgWWmb4xbnJqunbRrp2H0TbQzuMfgMdwCMiYXmIOmSKGojnlvu9SpBobBImuLPVwFI9mvm/t61VAUR/teAHZCrEnQL4QIphDvne7Hrzn0x03SIT4WSMU4WvlRQ/LBwkjod9LG9DJGHUPa+IhPW9uHimRhhLkUXWqBEF49SiPIvRq4LOv73paqKyblZHyiaR1ac80n81O5hnUCu4eBi/09tXt5G9vHnsIn6qxKP/QAJbnuO3XuKMRCqkwFiY8It3Ehm3EYmz+4sZozK6MpjIpAAgMwPO+u+ve9cJkuIb9xjA7afjdcD9HLxMF4+8/Pg3XTQryE+Th2WTX13lfQ9QyNetj8PpRqGGcolDMgv+BEuixRkTAQWkXUZpOqTy87+4sF2b4p42oPodO/X5BFzsbPbFuCdWuvaqF98sH3nl3Of0oivIi8mURuSki/0RE7orIcRzHNiJPRGRFX6+IyGMRkTiOe1EUnYjIjP79j3Bafoe/9WkR+bSISH5m0j1dhUmmYGR4ICTQWlZeFbIjEl1wrIgZIjyiDG6SFppPiN6hgTYUCNFDvSwHiUgM5GXJf5L8DYtxAG+uN6kLBK38dTboaKrl6MANdYX86eZ9YwxT6AZM5iNdTOwAJZbaCLYmr7hne7SHYrimpVjUjllYJBTOxuaCBbJ7EJLPRIvsxqBOsJb3aV90I+j4ZVqqrh4li6eExRYUjliHhGLjHIVWoyg4d4OY4kpCjcOI/0jTzPlrc4r9F6SPmASlg1FcpLR/OYYKKU5h1AEtNoMWoR7VAkttDATT9mC4IYpH8rxAp5zL4A/a3va6yTQMeRtz2pqrTvLoZAbm3+5wBOR1fYyxbXzHYPHk3KL0YkfrHeyoz4G+YTCu54XRT+H0KyZ4lG2DqC9Qf94o0+XSHe/K6GsK5gNRFE2KyK+LyAtZH9N/s24zfsrf3/nSw+qGAAAgAElEQVRbnxGRz4iIlNfW4rwSc1lxtDsGgiM0SV3VNnQRkTH1AskASV55y9H2R7ONX05FsWMY3zjyc03e8es9vRb+JR14ioXPrhVdwKn8/op6lPBO+os+6augXCBFcPK3peG/MSUzisVyrBsfGQ15HMGgWF8D5RILiEAmbwW0B43c4vXN5PVOPSzC08fwTKk5gPb4QUmvF5C3QQZ2nY1kp9tgHFXjMwGuF3bZEmFkzkH8Zygi5SrZYzSp9BJ1eOfHMNTnUC0zCF8E40vnwEQ3+Etko1yoeY3E5vLjXY/cUo1g1qvAnoIT1FN086VQeB/rZ2zWN5ikAYxRNek2FFGWB5CCaZJDRZ+VAPclLxMbuQblcD1TFX92HaSCjGBuHuLxD4/cq2paWhQbd/XRBWkWHYPiPcxJpMCKOufITspSimWD6b0zkmXaqKz9RIMLdLqf5fFNoXfiOD6OoujzIvJREZmMoqig3v6qiNiKfyIiayLyJIqigohMiMgh/m4Hv3PBD0pSVDWxgiKLNFBqIhLHjH4RRUpW/C3EP+8hj8lioj7H3mw2eufwg8PMkX18vQAcr7H4ceFTGjFrJ8zS9Qy/oV4NUDJTE8Cg68Ii304OtrHxdvC8IlA386igKayhGGummvrgsNl/Es41MutRB7uKLbTvgbQuZUTIc64GuIDCIPOm5s2dHrhxLU9D6u9hMFIneRDRoTGosEwh7fAg2UmZMsq64Etvu2Ho3PTfMs2CJjmNQKvRHUH0mCGXSA6a4hMVbJlD2kzc434TXEIryiLLHpHSlG8QhmYqUTZwFvelGy6N0MgDMJ1CKyHRgLiHvg2k5szqoR8xrYFrdB+IHqaveBGeNSdr6qKDtvG6w1KvvhxMBFNN56eE5wxHiew9IWSzrKkeGucIS610ougcD1BShxHUcVPoAQlEBF9FfVCify7L8W7QO3Mi0lWDPyIi3y+hOPt7IvJXJCB4PiUiv6Ff+Zz+/xf0/d+N4ziOouhzIvIrURT9QwmF3Fsi8idP/fGcJGpOhujiwFJSrz2K/L6iV5ZHHfbHwxYIc4MDwiRVQJydr6mUCxASCQMk1LLYqGUIiggwzUFr2MtkQxbrDx02qih7Ib2ufZCYjU0GA8wGn5TgiqZXuOh4ftINmORjad4NHtMkFeXm52ZJby6RXmT+E41kLLaZ99tbBKMhOyz1NxZW/Hlu7wNpdFVl8FCjGQd1MovdBuEjPJXL0tIUXejWsoB9TSUM95tuhDYeujBKCdz8HeXeJxorgtHuXVVDi3GZhqbAypjfr0FBKV/Z3ff7ssY4olG4YcemvYuBb677dRWZ8rOu38VspycRT8H6iMgZocZxBBq8rLGUIWVptQI+o+vftpG8tnQWO37pFCV6vg88vUN2XEbVHaVEJ2ljDmvBoJoEiPBcFds0UN/Lwej3AVoonerGeAlB8e/mkpZE5Bc1r58TkV+N4/hfRFH0uoh8NoqinxWRr4rIL+jnf0FEflkLtYcSEDsSx/FrURT9qoi8LqHX9UefitwREYn9AZiHwvQOOzTJWd41j7iTjcBomQYouwzphRqjJ4s45NA/AmpCF3QqjAN0K698Mz1634vuVSUUBEg39P/UDVp3FXWFirFoAuXCwpzWMMim2UfKxlgRGxPZrfypApimVC6SoTw3UXnkkItT4BxSI5IS1EbhnQ1PltNmbpu0ADlNI5w8h2iJJGhKrFSC59kEUV2KEM2UsVDsPt9FwdLSOqy7YG5tn2vxlayQ5AzCZw0dE91wQ84O0FGNrM4Lfq7Dtz19U3geVBT8PXsfnacD1ZsYkLyNnd0mIL4DR2kF9R7WdnROF2Bcy+gjMecjBkcOox076CRwY5xe9s3MIkKCKlgMH1EJzQ2opnH+G+ptAONcRtTS3cSz1UczAAyzx85q3RDZWc5O5URFi204F3Ti93V5dMYvn6cfxSntv8t1lNfW4pWf+K9EBOsq4vue50tRsG4Ho0boVe8lX3jJ02FRjPlsXTgkxyJzJUXMk5w3DBZzqbnl5vD5wT1SUKNYugAPzyKj/S6F14vIm3aVmiC1sNdRFFMjUHyM4tU1eGNYTGa8qF8wUsNv6Xh0IdJe2fD7alnqDYuqOg1Gz1QBefh5MFoxwXRO1Ro0cA1xQ86jPhY2JfNMwpDzpU+mUjP6SA9NXXHq5SMDAZC2gIaDpHj654kJdJhiM2LTlx0cF65MQ2nxGfCwqJVssayhJGyPVIDDz3O8jSqa/RNZEp3kladTU9OIk+RwljISESl/2Q1x/VZ4NhVsYKTFMOeD10dHxf7O+dDCsyXrqTkieYIyQL9upHSpqKbDkFTfpoQiHU/AZRNYdgZH4ntxvPV3f+LLcRx/OOu9Sxh8+BEN0vlxkbQQ8UWKQ1PPhRC8z+IOCpqb21oIAnImymBwpBoPNXJzxATrIo9gOEjl29aiFkNSFtt6mhjtwcsuVhH+MhTVaKSMhd84QMu5Qirz7/MNrobvW56++IIXtRntNMB/3lYagjzoIxpASJmnXp1xg1YBUiLfHa4vkIu+T84gNWh5LJoYRcakMQ7Pq07RDPU4yUlkghjv/C17puNfdMNy9h2+8ZmRIAFf7185nDGnCmwpdlFYCUZLtqEfUTKSXEQ6Z+b/yM9VB2Fa+aOeB7dzXDSG9jziDHpgEXckKIpDh4Islo2zYX4iNhTanGQKjhtyItkIiC+dntJfcNDFsnr4hGSyxmIUHQQnUJDFABLnEZBlMAuMHku6Brmxlraw+RvZKxBnpGywtp8ITW9c633YppL6CUt/6Kil9/J46ynvXWqjH+dFOlPhoVrujGkUesSvXH2QvM5lgGOZ6iksh4mwPw7unbueGzebTrrW0gEKZMjPF7V6TwFlVvwrD8P/tGfhwWHTyNeGw2JGEgvITVeRC7UjN+kG3LRFyckyDjy6MWZemXCeFXZCPhRHRRgDabsBqgoYrPmZ4a7gOXQ4b52F8ex93RcoC2gCdExSZFwYhuqJuHe+sOLG4uEW8uiaeuiAVZK8MWfo4TjXWkP3Y/7+AJuZwWGJlGp/3De+F1VDl5rGe1Dhovdu83MEESPnbEc3iP1PgPMdY7w66uNpvQKPD7zKmBL+UY81Qm2JOHxLk/IZVB/7nJ77mPcd7ByGZ1dFr0ILm3dfuW0qTyBOfxswSMXMz037HCnyWtl4p8aeAjjzc/69EZ3zE0A17eHZmN7E4WOMC5wqUidLXetro9kpGWtwS4EXynT79V9YTdbvUmmfUvifs2s+N97T4ynV0ktt9EVQfdd/ykAnxEs+ylsNN9pWqNzHYqRX09Y0CL3U/jiMrxplcve3locNrojIQO1FASmdrIakGEauvAnvV70mpmmI0NjZ8ck8o1hnCrqnjIx6Qwzr97vuZRo6hx2gPMjJY+OVSmMgf1/TWgMRUudd32xME6APZAphmrzunNIR01scIHLqFMLrjQPfQHhYmmIUxpU5ZPK6m4B4E2NY3PHxblbCPbBDOpVi07nVviCFkBK4Mbw7cuuEWRoCKVNaUtK9ADanu22mxYa/Q0hmD/UvS0M89xnfSN7+lM+N3qn/llF7c4OqgQr7ZCbcY72ClCoiBYsmWIcgTJhe+7iSBJ5j4z1tgNJB+Yn260gJ7fnrZN2Q5BDRDHtxElZT+B4kTLOsQg7ePZsyk3Oy0xlrjZuFce/UV/6cZfObOnJdkepWGLSGwrAoSk7EGPHL5mFcnXGPliRPD/vhs8xXE+drHi3JzlILjCG0Pv8cUq09NHoZ30evmL2wjbytK1DpAkKDOH0z9uOQ1Ns/gNyh4tTrEKFmAc4w9eSdyV/QMmgIIopfkKbhwU4wqkvwqOnBJXULRDVcLOncsiIdqMwF2oucImLYadyiNq9u4uTYp4j6yV2PYNoa5XHTyd/w2lBRnz2NOlNs09XwGw8PoT5G/eOx4boH0VgNqLG1NYdcm8xGSLEh6Uj5gZg+SiX9Tb+VtQxsMObA0NAzNZFCjOnmz82sQarrRAqQSCG/7pZGHcUpn8c09My5GzPs7XWPNNrgVTJuKOL0exM+52xjbHA+ANVEeKX1z9CoV/aIytN7gf/H9E4y3kjvkG+nsoe0j/qTTUefXprjUhv9QTlOCj125E9RUL3nHuvtVx4mr2eVhIksfyS6mlgNi2z73KODjXuety1sKSIAhnpkF9CuSRQnt8ODPnkRM4WG1IwL1mJ7FTl7C+dTnZoIhVHAWp0JqZ4WCMDmZ0FkpSF0FbTB46ArMCz0LEjYyFdzVoIwuilMsdaBsHl1LiQtDV0hkiZva82H36p81iOV7Y8Pc5OLvDM/Ho4eRGvGtci9gsX++NjP29Xc+VTNIzejQBARaa4iD27FPHiezPEO9HksIjWxsekOxXw1RFu16z6uu+gGPYG2rkE92Rmem+CzVbUrOhyYOyTWs+ale+IRDI/mntZ2YKOIMrNItsQiP/SL50aA+S9rVIENSGq+MR1rt3Z+28eVfFIjr4S5weayyRL6KrAWjaeHnETkgzpsKS0GqJlLuKxJ7Ys43MgG16eiS13PPTyD/gLmpEaX1EdI1dR0fbFoHaOuco5zxRYBXKCw9iyPS230o3ycdPVZV2V7HOyHEJHgpLGDXN1ffOxEVs8tBmpiYoOpptNTQz36wB9oYwU4+yroHXSvIO6bR6K5SrFmtsfrQcNHPv0KvMAEvnboaQ56/aKeNonkyLh5oiRi0xWQiOOYqoKq90Q9QuxflJ80TnRblCJp42a6tsefwAYHyCYpAmzh5DCG3PiMm2e/5N41eV2aaijZf5AqHAIOeG5kYJTRg8BHTSmVU/EPNihyuSdvk2IjQy+3ueHOReW6F/bMwDfQbDQ7l134s/shRcEZ0iCGHOnVAMGFCIqoAW8vZDsnh2AabWttJFUEBSTT7qu34ufvHQO5pQyrdVzf7IRvAKxNGUMpOYmeYENPHJmiG+J7u77xWWpsHFTYLPTGsxkSm71sQ1xSiU/WVeh0GWoppdKFcWGRPMkitIcdmmd9XGrI5sjCWnzzr/2EiIh0dN2w+aR8yz0hLpwRLfqkDANCSpso7ScQ/SYya9pUn5AnZK6WkPx9ZewEzfPUG/6B7e8JE2h6zQuyp1CoqmqulKE8KXNpyDY1p80UACdoAqWjd4HXpYw0yfGmG6TxJV84VgwvYgy5WC3HSs+UzW5mnE6QaiJEkNQIVmAmLJCHjcEZhLpnUewzoReORSOjOCviY8c8OjcrS0Ok2U2xaShnDx0OwhEJJ7TfYDMSn73VWAg8YOqjuwO01IzqF2Dc5iGQY4ybTPcR2msU3cTxF8FPNEDKb0y9+tNT/31yGVlk0rnvv9VHZLa4GtKqVLbrYP0wldS0Au5pNtDBeHJ43ayJDW5pQyJrHeSzQi+BoaXYsMX0jtkW2gJGMFYMZyBP8sTGEuo1qs+RAi+8h8f9//onvzUhm4NKLKe3w6haAZfoHRpf60YVca+IHaJc5Nbc0YKeKXOx7bcVeYKJnENBKNVyboVY0EMcUDhE/0xGRBoGM/Z7h76AOuN+L4UMo05Mcge4757mbctcoBV0WmoEcYIFQkN/Dk50M0jcLHs4l3XyUs+3ClSFeWg9KFClCl2z8EJrw8VVEniZZ1mESPvetntzZX123HTojZGULjrTtBVpIFho1ddlbIzk8bF5RBAAYZQdoEVGdUMnuRsPQ6wwhUAqjNIavNdD7UVAQXWThW21RLk9FKCRnox0k6Lf2QNQgTDFYwUSUEiGR0cLtANwJVGX1pgvC4gUaOi5OSewT1BGpJ69XhYb+9rziJQNsUVLjGinuAHQQ4bTTXvSVaRgESy57L43JA/pJwjaYHhomhrF+p+nd765I5LEwCZ5cGLcsciJV7dUC7HHLIBZ81SMyRED+x6thwUdwTshM1+E9Il5SOezKO4wfNRrOINB5QI41AVyEcMk5QrNqz5FEXHkHtIJ6+oVweBxMVivQAcNKU2gVEZgUAwhMfaGz/DGR90I2MZ69hjyS7ht6xYtzw/TLcv/196bx1ieXfd95759qX3t6uq9p6dn4YgccsShKEoiI0umFEeMYUmRkBikoICIYSWIEwSWHMMCJEEgECCBDCd2GIeJBCSUHC8x7UiRCEmMbFEUyeE2w1l77+qqrn17+3bzxz3n3c+v36uepoZTVTTrAoOuedvv/u7v3nPPPed7vl9JUghUNlUHFRtB9w6QK/qcXJHHasTkVSqQAiIUZE9wPihVd3oTcVlARS0MVwcvk1AT2TYLnvbuQ2v5HkJBj+t9wTB0p+Aaar9y8Fzr54lwih+16zUT0EmEJ3XsekieJuLJelIQ9DUDmgTqRVurLMVn6xHStPAkac49Cw43FRYLI0seKz47r/Qn+dMRVcRcXfGqcg4tx/WdeM461+nUUXi9swD9YqON4PLkWtGTQOc0voONMcG+ab9/ACKz9GpYV4Wt4xdJOd5Gv+fEaSLGDDS9xfZanPSjZ+Nxv3ozeECzVyOum5WQ7ZnwIHnkrEGwemw+HJs7yBglqmS79No1NOEQk4dxG50Ok5k45AaOzUaalagSxFG7Rb1P46gH13wTpfoZO3UgtjhWAi/967Phd8A93sEYZhbjb02eCuPZmUMoCfdt4Y8OsOCU4mvrwssCUcPVlp0Bq+IQg8N4dFqNSxmbEkMiKn2a4MghD08tHX9/VHMg+zkU87AISX+XtAI9eM/j58Pc4MmNDE9780hoapiRSKE60DvmdrcvxbFIoStl5GvsREhFr0aUCehvwvk8qTBAyaCOijtTG/o+OYNsQ85iU+Cc7+gmmJ8cXl9QL4SxzUGYnU72NHSAu9ODSePc90DrYCUgr8YuxVHew8nNHCj3KkK1V+BoYI0bTThDZERpWRfoLCYU2Q09juSt5wmDX9Ol0B45fjQMx9voe5GU7s7lG2Gg6xCezm/FAZ19Mk6kJ743JGrHs4PJXRGRpsI371VjwuhmK6J39tUo02NP8JQneHo0eVqhcYzXquoRdwy0v7vb0YhYfL0Kj5uJ2HNPRMm6fT1WF2HQ8uQW14V/61rEiZ15HCRlSkFcQMVvAWExioWbUd1GspsLxMI6/mKMK1OwvaBGam0D3iI2Q9IJZDRUQm+NIRkTNJkZic+YwiOjM8EbrAAXfm485lBebUWRE+MiOg2oKRlayRRqbfwdUZN4thj6QFZIoqW4xI0WuAGHY3IadMnaXxpM/pbpFItEwjUmqHktg+kSiZTI92hfeHrwu3G8Zi9FB6mrMf02T9IIYdmJiye37JdjqGn0/QG9wxwN6ZSni3HOmRoa+XbY3vPMkogkJRLvFuK6taKvu+NwjpaxoY+iMlvnVx3JeGM6FRHJm1wiyw9YfJXSEDPCu5nq8FBQfvv4efjWjrXRd12R3F54EE2FWmf340SrnkVsHNDB903cFBGRPOSqqG361UpA8lBuTmCIy3cVWw+m3iyoe5qTcYKWljWphfBqgunwfphgPVQUMqxUrQ16uYyjk4hqTGPmVMP63osRqmrIklMXINSNgilTJGKCewI85m8sR+NoeQcaareEkMxUmDqLZ2ItBL1vi+lnbyKZNw1CNRh4K49vEemA3zJ6iOmZaHxvvBHdXMtLjCKnUMB8sGI8EZGUFkSt7sQcSpn87nq/ZcT0uREYbPaV3bixWgWrSLKwzojFKjjZpcimqicTitpsAv75odNv9P9+Yz+c0u5tx4lWB3beUG6jfxCN495lie8rtrywjo31fHyf9Rop9Vo2rkWUzOj5uEka303ua7GvtWfiPJrQ7xOEcAn6xl/9/SjH8f4f/4aIiDTdcFNUVKGK8Wz8rbsSjb5t/kTUZTew8UE5q5fVRC5+39aviEhD/b7yShyL5sRgIrdTRLgOzJqZCvI5KypNusGT7vFox9rop5siYzeCodi5qsk8GP36VYhjQCP3i7sXREQkg9K7+Xz0mlbrYZGyNJwxy6bG56e/ER/o9hN8+IPFGRkcKnwKcEFdK3sTcYFktkFnrFzvpTLi6Vtxt9mHtzajNAdXTsP47UwPvN9BeOiJmZX+3y/cPyMiIpPATBMex2TZ7LnQcYYTahdiXy5NB29uG+LXORxxdpQSormIatjr0LDFJmneJxPco8CFW7jrhVsRdpudGAz10LcqQ9Vm/kwkTJtSL5PFevT0bZOt4dQwBblEOhLWHl+Iz2PqQvysibDPgk6B89QkH7lZ3oFASBMUGaeKYWOrTaGuAnUJyzeDxdq/GPtF56O4Fubk3pX42uhNVNw+D+Ee9b7X5+LGeHYinpxeb4QNiLTBHk7T7OJa4ndEkkL2P/ITX+r/3VPMfjkTrz+Xi7mjmupZZ7GWWb/w4ueuiIiIO4+wVgG5OAr3WDoGVr8R/Zy+Ua9Bh9vjHkv3tVKfHP27gOtCu2LnMR2PK4PsqIfSvnzwW8fa6PeyItUFLZ1Wh5eJE4/j4z4W6e2lSyIiMn8qTtT1fDS6G7VgkFg8ktmCIdZj2v0fAAQLHkNrCokopfWlchD5uOsXkxq/IklhkTFFrnDhj4CNkvHgO2qgSR63vha9zJSGvmZRvXh9N4atzunCpaDM6HRcLNckfrasuOgV8NafmkGoSHVlGXJh6MG4WtKgn8iQ6BQLzzxtcgYRtWSqZymIVLewsdX0NMPT3qvQrT03Fo3+MJx9EZBKuwc+j+lC7Pi/uR/cZyo9FdLx+zWEfUycncavAFH6ES04olIUc0ef+bP39P+evxzCLyy2GwF2Pa0J5hZCj2lUNe99T/hs4Q4Uv56J/c5j8x5RNNS7z93tv8Y5M62hlPuX4m8xMW73uNWIi5Xrcxk0yQZPHocjcqMS56GNnQnDiyRP6AvPB6emX1ciIm2QXxVX40Tbfyz0K0UdAHgKOatQxmvZzfg/dd0MWLGbQARhTpud+k7l0z+y5jMizSmDRCoMcwYwMRZJsBJRwwVrEBhZh4RU6l5Y+M3HkeWfG0R7kDLXPxFdeaJ6WtnwG8W7oA3ApOiY0YPxY6LV4qaZAyQMefTvKSlc+V3Rs5yfj4Z4dTUspvYMNsMqQxtaDNSG4ATCN4zR2sZIOGELlcAmV5iHHB09feOKJ2x2/9losDxkBXd6ir6BwWPCc/Rs+N7a75/pv7b4YzGsdX879MWzOAv3slONBs3CNixsInGZ5S1mJ6O3SYNjceo9eOx3tqJ33rkZx7a3EObMhYUY2iCHjCVnSVFAYfR19GvjG2ET270c32cIyloazJapEc5pzS/AI2aRIENYr23MJfon8kAy2j4L49hF7HxdBWZ4WkpAf3G/a9vK6ImwVxrIK8tLcDMslRiO06IzEgNCr6ICrz2jGyJPhN17pGwI/7ZxQkpTGUuHtgu/gdTKpGwwTp8h7NlH3o630U97aatX3RnX0YNtpDe4DQlBW8RdUiODfdDgYz28zwpQ92YnMj5ITdTWz7eHv6+ZflIvtyZxfNQNhguwN4S0S0QkrZjibRT4EAttv2Fl8iLJSkrDdbdR8WvFZSIicnpQ3IWthoVrJGY7+3HRJIRidEM+qKgtD/SOoUg8jBxpEkyisPjB9f5r1+/PxuuqtimNBatsM6hrqGpCkiph2SJRP+G6FPVmfUBb50wOtRDU7vUX4inL7nyHRUpI6vYlJbFZbR2QOO+qU5Li6Rb5IIOaeoQzhPTSCt/kPONmtwfu+6mFQcU5bkwmOsPx9rl43V29B6LjOhDF8ah+H7kdPtN8FoielfjZrhpwvwoq7AkgcvR+UpzHC/F5M4dic5oFhT28X1vQ8A2QmbtzUOmqaCIXCVsCOBogZ5t4RXN9E8fP6h9roy8p6WNj+waDOH0UGZVBA1u5HyYwy/pzmEh2eqBBbN0FLrxkEjvoCpKBHtwdxmrYG42/lVAsMmOPQjDydZhubIJs7IDS7vRU+Cx57zsU5bBus04AeWIzdEUUorWAPea1zCB4xGLpbdlnKbdYRLLaDFZqA9WTyJs3samYPkATFAYkemsXw9+VCthBcYsmnJ44q8MgUVvXqlBZBLW/Ae3dMVMMiz/VRr/s+zQcrBRmkaC9TmZMQnM3VsImTKK4BmkcSK6mxu0g8ja79/Gvx9dqp1FYpIVHXRYZAm6bQq1CTamoWWzH0J2poTFpTRhjTzdJQkKtPkJEJLWJYrbx0MdEnQC0FLK3wrptA29P99nyQI0prPXr0DeehZeoU53Spj2AKpyp4JFZFs5g19h3yY1VJm9SvNTOM8fXtB7fnlnTyWwiJnRAqZRUQXLl7KXgETJBl3t/9F62lHK5AUNO7h1jxhxWrSciCZ3etKIifHr4BLeWKNgila4iVhb/IP7m0o+iyrZApSW9PkW/oWFrp5UWrGsTOH+rumx0h28wzVXA3pR0ikVtrR2cnHSzy58DBLGK6kdbOBA2IecJaatbGkoSaMmaVyUi0tV+jyyAafFr0RPvqjwenYAROAHuz+N97Wm1dJeVQ5hUrRX1YmfjBjbyp/H7O+8O/87Mx2Ti1lLcNOqgG7ZiIIfQQ5vOgZ6GWHDFRtF5C/M1boBmAWyuGSUirJyHkRryPoVmGFAsAVNfXw6bayPHOAaMm9XKwGHgfttQ3hmKrOSwyWYAXzaJT9Y9cEMVPdARccb3rfqXzLcUhC8uDYZda4vDQ6n9UxLWZwK9ZgpuXMsUSoMNOoC89li0NzX6zrmzIvJbInJKwi1+0nv/G865KRH5HRG5ICK3ROSnvffbzjknIr8hIj8uIjUR+Zj3/iv6Wx8Vkb+rP/1r3vvffOi1206yatgNk0+YJnkx8vCWDLtONkqyLhoaZBSFSzsvxuRReyJcY+GP48Nd+RBgmIjzdZTKlnSt9FJzqreZWkOhDCZKXdkal370gMq/fUxaPS1MPxZjxOTp6d9LHtQHk4Cf6amAUoG9A/DRefXWWjweA6o2MRGO41tQhaJhKClRXOfVmFchLTbHS04N8dShrTuqMdxZJI3Xn42fLanhSI8Cdojf2npzaZUAACAASURBVHg+Gs+chgTJ/VPpobJ1CBle44Mxjp5TI8Dk79yTMey0uRNPI0bvTIx7gtJ5iPIb3ye3zpjO37vFuNkxTt7qho2JMebcepxTNvZUiuLJK38GusrzYbwYOuQpr7mmlBDYmHuzqPFY03DbObC9Yq3dvxdzIGasZ6Cby9qTXQ2NMb9AeOnsRBij5a24fnkKbCxgs9B1O/X1+IHqGaxLK/qfgvMBGuWePkbmHxLi81jCM18Pn5n+XEyGH2a7/ZD3HsXT74jIf+29/4pzblREXnDOfVZEPiYif+i9/4Rz7hdF5BdF5G+LyI+JyBX973kR+Yci8rxuEr8sIs9JyKW84Jz7jPd+e+CK2nza9w1wp6jGE1SpTTycCUyUld1gaBoghMo+DeoDXZhU9kmEsNVj3virwGHi2F0GJ4mFMXiUzWBS5vRYWtmBYSHplR2Fs4MGQERkZCoaOvOKSBC2B1UoS2ZTDJ1EWmZkZpCkpJHZzqAwwa4JDyuHJKHxzk+gf/TADPq490T8LTekoldEpLE/WKsgQFh49RKJ2uBmtaXhmVOLcSoRjssNwOLvCR1VHOfHZgAxsmthjCp6rX0kUQswSCOoojWcOovOHHb8xt0wP8cuRpQZE9/UPdjXxDHj8Nw0vMl2Yty6l4Ew0nBc8SIL8GRos8rmFgq9WMvQndL8AUM++H5XTxOTSLgSsy/kOlIHqUmyPlQVW9K3gnnOojEjBEyjyjwR3iRdhp5qa2fxDMANVbqla5kFXQjfmF1InNSprQFPf/250Me190XwwaG2v3nwW29q9L33KyKyon/vO+deEZFFEfmIiHxQP/abIvI5CUb/IyLyWz7Qd37BOTfhnFvQz37We78lIqIbx4dF5NMHXjztJaOxWfLdWEsRIojFUjd9VyR0Vu5H78LEq9e34aWeQfJHnynVgtJIBvbwzFtaVJJK8/gZ3zeRiOpyvFZ6EigWXQw9xEcz4PNm4s8I0erwUutbg+IWBahwtZC0NabC3hRi8+g3vTkrwOGRlTH3fR37DIi2yN3TVHQMTxXTE9FzXV0Gfa5uJr1NxnXjPZqqUhPhG46xhYruQy2LoZGE0dbKVcb029hwDZVEj585kPy4Pk9UUO+sxNMMk5sW86ZyVyKxrYnWHQqyoC8bNyPs1EKN1DdgmMGuUZ4fHoe3DZk1HAyZbO8OkshkUBi4s4O8iBpP0imQOsQqkMkbdZB2rwnzVJAXYc7KvsXNro6wbFNRcQl67HlAsfE9Fjr2+wowR/1JpXRgES6dsSHgBgIOCFToap3KQZxaR9m+pZi+c+6CiDwrIn8uIvO6IYj3fsU5ZzN0UUR4plnS1w56/cFrfFxEPi4ikp6ekK55YXp2IgFSAdwgTeij9pO/jMfhqNovQuJEJNWuUuXuQxCDmw5jxwYzTMP4EVJmOHfKJdJD6llRC7wfiqS7W0A9nA0GJ1FqXwYNgxroxFiA51yUFZEJQI+CLFLO9lko8Voizq7eUAebyj7+NjZLLqB9bFapIWglD71gLi/zxDvQSiNKJTVlFMXxOxyDHuL3KSXbq5UxhhBk7xphGbz/5h3AMHUM02DpTGgO3AHyZFZPcRiX5HgqFTb4WZhEJJqqP2f5vIjU0TqTBrxY8uXYOsqAdXJvLt5DBjj7nl63uTecHdQUv3axphJEcLax8YRD75hEcDYPUsNDc23NhySEZpg0troKJKXpHORw8toFdHdYGx8PjsLuLuivoTxXVQcrDVTUsBCdiEhbKZ1d5VsysYfSHrlHzrkREflnIvJfeu/3nDvgbCjDlDvFP+T15Avef1JEPikikr+06I1sa0LphsljUr8BD2sRVZMaMuEuzjY6p4RqOCbWQbjWVuwwJeBKy8j4zyLWqUaCR8oEOZt6INkJ5BSwmJx6aKktwEd3ManPA0Zpxg8OSxbqRT3NLzC526GgkG4KRIUkqouxgfShfcgv5O4Aa306XKO3hOQvwuEdY5PEZsfNKqFVYJel54oNoqOGzmPOJfDotjFeip5n6lo01K2LqLFQIrc0+FmYsLcx4maaXQJU9UIIjRG11MXXO6fB1mo6wxjjDE4CHdHYN04C9BxHEAffzyi3ToX5AcyzGVXG+gbuG3F6ew7MJxGF0lscwlPlh493V58HxdKHbdJZ1AlQkKUwGuenhWe4rieg4lUfzQ6837iHE4RqX3SxafRwEtjCWht9LfxW5SJgq5ODObEcDD1hpyPKEcXT8UH5Ggvbdg7QPz7K9khG3zmXlWDw/w/v/T/Xl1edcwvq5S+IiFUMLYnIWXz9jIgs6+sffOD1z73Ztc17tYdevYuQzAzjqvAcdbHZzv1gs4IlemUyQTri8A+98+oVGBm8bkaTgtd1MGoWpsMEZ/FIZmSIQQMNRG8MBR+gwvWo5LXWnof3rGOQfgWe6eNgSrQQEsWcpxBqQq6hq2RcxVnQCsyjAvNOMPYZ0AaTLdJ4dUsvQ9zmA5HUq4rF0jAvdB2oJeRubGGSJoI8QNmrITdTX4n3XXga+Zq1uDH1PUN40XlAVRsKHezCm+yeBRxQ74tGrEva7fTgIie00WoK2HqgMOhik91COM14k3aw2RRuIzRxVpFZV2G8iSbRfvXwjBIbPsjX/JBkdm8DXr96uvTISWB3W3mCWA/CQ0kPjJg9zTv01uI8qXpU7CoPVvWd2LixQVlSl0WGDM1l8Pf+OzTvgY96bIJmwDs4GZIaxJxE6j4QtFHGWrFahqmvHo1y1ltK5Coa538VkVe89/893vqMiHxURD6h//5LvP4LzrnflpDI3dWN4fdF5NedcxZc/1ER+aWHXTuV8lJQKFn/eEc4JHdW7L4Lk2HBk18lj/L3PcUfr+Jo5ghtVANcuIuk2mV43KRe1T+TSTUYQouNY1IyJGKJLKIAqMKVYPyz67KQDLjroiJmGmNxUtJzdBpOIAVyag2LGQksp55dYwWxXvKjqyffofg2la/0CN/Aqcgjvl+H0pnpBLvL0VPPUQdAF14KeZXmRBzDvmoa+d1fgkTh09ABMO8bz6PBk5dVbeI0lMHJa/5c2LiYLN/Pxu836oNGPTuFMA2ua8n/LhWuoPp09qkoFm40A3QC6ucQntHwTmcaIRWK/ZhsZ29w7oqICOU+9TnmwAflEYIyo9jZjM9wHRunURjTkJcvxE0hOx2fhzHGprBppBFKXZoP5qIE75tIHsuvbXEzJeUJ7vfM/xP6vfKTyHvgdGmIrl4CcAAHTOdpGvmgDvIEiWevHv7+B4c7nm97+0cHv/Uonv73i8hfF5EXnXNf09f+jgRj/0+ccz8vIndE5Kf0vd+VANe8JgGy+XMiIt77Lefcr4qIsS39iiV1D2q9noshGvO+hxgWkWSSZk91YadA4ZpCJMk0YvcLcQHtAJOc2VRo4ySQElhsVCTqv4ay/N7sIMlY7TEoIkE0o6Pl2t1JTOTccEoHKzaj9yGoVxpXjdvsY/F9IktWF/R36cWiqCUFIe0Jrcqsl+NE7rwRT1kdE9WGp1NYBfZeZeI4VrX1uIEwzGH85kS5MEY8MqrHahiDVi5ea0vnxNR83DSap2C0sSEbAopH8Y0eZP9qg0uCCmrG70NqZUIIqSRmxolc8YS45pTiemw0zlM3S6QP+H9Gw72tnYf3ztyOjtfkv439qi7A+dAiqCQEEbDXd0Tkk7GVMhFM4j2D/tZRL0IRI9NiSEMBbqIYPRmrsBYRKeqcTSPuRK6i6niYf5Tq5KZgCeT8H2M+PIk6FNjc5R/UDX0d4SHkkYwNlU5dgrLBNmmEfJi0Giae1MG1jkt7FPTOv5Xh8XgRkR8e8nkvBwCGvPefEpFPPXLv2imR9WBMi2uDnmMXcm174FI5OxkgcAvF6D3cq8Xg9q2tQE2c2Jl55MsOlmPTePkhMWDWDKSAha5daiU+JyLSRrk2wzr976MOoIvqRIsR0zhOjUZDd389HIvLqIxl+bzFWlOgHeBn9xvRIBkaqr4bxzXNMTBvCuGI5uOAzan3mkKdgbCoMv7ZF5FO6BfgBGEoq8vzMTy0thsNh1VWEztPHqD1vfhZC7MlaDcQI3Zajc38Bk+UPZ0oxJLfXY+oock5METqGBKNksZ4We6HGxC9WBYXGjXx8lqcx6VvIiRyPvRnYwqbAmIqptlaR5w/TQoBGHU7xVQg6E7EmW1G/H0PimH7rcXJuP64gZCx06i/V/bjtXgaGS+HTYEn9Xu7ED7Xy/Yeh3ePojQSNBr1MYu3WLdAmmVrrGou3Q/v104Bmrwdv1NdBIT81fD63qWBnzzydvxSyw80M7x94rVdIEygrkRP/0w5TKqRdHz/6mgUIzE6YCZpeiUs7EG4eiLeTaSNnTaaEHcp3YHHq8pAVhEpItKdR4LL4vDwRls4wTgYiaIet0m0RfzziBpwyimeWgB2XSkECEUl1I4ICjshpMGB352I1zIoJo03la2qqlbVo8oQ8xaUolSqBoc4O0NgJa2u3W2CwwZJ+q4ms1cuxf4RTsikq2khkx10DaEJR29N2ygQIGZ8OG4j0MvNJPRd3cD32Yxzh0VW/P5ZsINuKmMlT3nZ98eDslPECXMotSeREFUOmCyq2DuleK1CImQSPnv+8Rhe2gV/UMWE6OcRh4R0Y280fH+zGhfSCMRhRrNxnkzmw9gtluMGQZrlujKV9pC0XgMdhz2H5iVoLtP7JoLJHK9E3gVsrcq9QzplbiB1g1fDV6ycJ4tsCq9rV+T4teNt9FO+X0pufB+dFcaCAe3CIvzS/ZBHZsKFRkRM7QoFHUzEdpVuIAOIIr3+LgiX+nORsT+scUuWdU9hUoJhMq9JKYprlxCmIJba8NxMFFME3WgUyqcjTvn+cqxPMJgkhbrJ5UIIq+Gye9zsiLRTvhwWpZGDxvrVKDKnAEMPPVynmgKJpBmel3nEPM2xZS4rGgsJuDUImxCTb+3Oa1EEZfJFLNYfDmPPJOQWNAf2xkoDfSU+u4IQlmk1c4xJ/ZEuD76/sRy92M0txO40Gcxq0L1VFBadCyeM6gXMc5D8Gd8MjRgpBHZW4ynPUGuVMZ7ysGHrptCFFq1MAJyg4Z/tSpxnVczTNQdpQ/2tFvIqGVRjp69pjcYCNmZq7NqJE3H87DYKEgE0aCoPT26T78dbMB6gLjZDD2qQ5rz+TfoJnGTb07GP4y+FsZm4PkjJchjt+kPeO95GH8VZFkfrIAY9gsVMVI9BFhl74zGuT6GK2PkoKjFrRrdKQ0/ZNDSbFDzqGne5iMjsn2iV309GQ9wYGawsZSw5Dc8wIVNn3B+IozcrOJY46z+MRXEwfEQCMBqcqTPx2L19RzcLGJkE3FCNffY10A6fAqyvEDrjICdHxTDin2sW3sGGLmVUTd5R6mT05dTn43iv/OXwb4qb6Qbgn4jrVs/o88IGtv1OPGjVh/WEhAID3i3p89pCJTQIvPIEMI2pF4qEagF0HA3FuDlUhaYwz3qE+1k4Chtzgo5Y49yFZdSQ4LomqJLfwMYLP6R2FiELpW/obeFkBeIyq/pNAflVAHdP95u64YKiuE2BccbMs0ZjgvAM3u/MDQFVQLe5qHmkxhWcEolOw0nYogSWbxKRJBGirkGH8GcO9TcthcsyRJddoOYwwmUz4bfuTw/nVXrb278++K1jbfSdg2yfvtbDaq0jJk+DYmLfjL23z6K6diUs2Nx9CF7AEFtykwnXHutUaP/d4JGQwiEbWo49z+rHTZBm2QTHRCf+uY0iI4OaFe4Ds3weXqxNdlYO4ndL3wjjUluIk7aKUFKzhI3R7oE/hYrZ4nK4r9o5xL5hZPp4crIYHoBZ9opA6oEMzCW81NCJ9ny81urzSHZrAjq/hblBjDqaGfDsXrwxSl2abkNqHYgc0OtacRTzLjSuKYxnSh+Ng7dpiXuRCBgoI5a8fxGJXHIwqbHP34z9ItCgqUldbmbmuYpEj5f3kttiQjL+2dGNh/1KI2lcVwirh1PlEDs34ZCEgAgePTWmsxr2zEARr5dDeFO/R94mzmkz4Nm7oGkYRf0B+tAnoCN+FM2AEnk4JAQUTMwEx21vb3iRVyI3o/H/hN04Ju1YG33vYzzWknUlIB227qGUnwUuepTksZvkUfVZw4VjF74dZ63TUAwJlKh/SREFM7ApHElZvt7Mh/6vrsOyUI/UFjaget0yjp9M/FlRzBkcGVnBqcfe9BaMN/IeDdW17eHIygIauRnHoKRGsQ7CKh51a5fUY8ZROw/e+ZyOATdmK3oTEemU4IlrRSyBWeQpN9gqKzGpVGb9anFPYYUnwkpZfY5t1ELws0a2RQw7i/TEQoI8gbEorQSvft0l+i+SNMSWBGzMYCOoYp7NwevXP+ml8rqmA+ywETCH0tXYdYEKcPwtQIatX9Uz6BeBCuq194B8qa7C6muhGMeN2H9WY3e0j21w5DucsrxtorDTXLeGQGpN4/d5MkM4y/iJSii2q4P7rainFVI+dDLxt8zYZ8lMi1qKEdBq73xPeJ1cYcelHWujLz3Xp/btaoKLknqcqCyAyalICqtsOyyo0offRQhBsGlY/JTxzwwmEo1mnw8fXk8aR920ejIsHuEG4Y2uGMiYhHg2wi8mhMEjPg2a1/GgQSSnj7dYZXaIRy4iQvpoDd+kuYAY3jfjByhrowiYJGiQ44/GP9ugI05rv7swtISqWuiscxrjegOhBzWOPOAUUT1M42ZjQ+PKSl+j0M6DobIJtEe/LqNBJ2D4CcZOVKSiIETW+TD2Uy/FD6z+IK4FSKR5ylk4HwkvUk8bVG/qpmF01XloxVRHgum0A+84vh//TiNsVVRDVr3KjRXX0vmb3YuvtUnGSuSUrusUr88Bs5Am3u6M4yRgKncIyfRAc0496qzWWzARS1isOYkkmhtHbYgViLImiIV9Ca4j7Rcr+Y9LO9ZGP5PtyuzpEGe2pBa9xQz0XXsMj1gCiZWF1cFbTcDyCBe092Hc2+DbZxjDNpAEtpcQP036kgCsjhiuU371HDaKJhJgCR5z9UiJJOoB32/sncQek9HTl/UkwAVCbwz30L0UJjuLuzrYIPo5CIRv2NeqJpuZ8HTILxTmYg7FxotPKL8wGJ7ZvxZPdiPvivTSlhepvhzdtsbjcbGyDxldpDUUh1EZy2kyPIs9qwHYnlCERFtucTjTqFUlE9W0B6UxOR/mxPoZGAYUhaVI1KabdwvhmSxlBXUMO2OId4NLvquQSuo7p8hWyQKyOaXzwL1Ut6IDZaIvWSTOuzhJG4qsRsI15FtyS6ho1fVBjzwRPjUHKDW4EYjEDaRAJwCbPGHdFlpjXkMQxqtN6tjhGXcmBp93heI2B4SKnNqWbvZoKnIf1o610S9kOvL4ZOAqL84si4jIbjs+3G+unur/XQN1cR/Vw2M5Kg7NaBZGBlEdIpFOlUIbLFNPjTQHvtPbjv164vKd/t9fqyqSCKcObkYF/a0GjselU9Hi1LZhJMzDZ6y3OXgs9owr475t06ChzyJRRZ6Qs3MBDkgB8vvtGKKyzeAgPv6uLowME3xUBwM0cao8WLVYaaLYzbzcc9G4pil0oS7a+DNxI9iBfCYLpia1GKg3gpJ5YNT3FBrZ+wAkA9cQutBTUuGJmPSeKoFLBl6gFQSa3rCIyLn5CLO0zY6U0WXAee996XT/b+PWKQEeSpjy9npwpXmKbD1JjVvrU2wdwoFhtA1fT8+1PBPH9n4lXKsKXdGej/dgugexqkKkginbHlKbkvC4h9EVs+N49n3VtUvxAik4ZTzp9l+7EsewAE/9/FSAyI4AUjoK+GhTEwQ7rbgmM4D1Tefj/NzvhPm7e/bhJG9vV3vt1w5+71gb/WyqK4vFsLhswLOpOGFGwfFdX0JyVHfZPBKeHYQeusrZ09jARlEe9Cw9vTosEE5K844Xrkax8lPFWKDTR9xgUpP7w/DuWWjGsnI1QbmQG3IEhzfY54vB76dYeGSFLCAFY6KY7Z7q6XagpuXBLGn34OEp5bYRCjo/uKGSO6cFRbBb1wN8Mj8FXhjcgxXrENbXgLxe/p5ypiBc1wVP0R45hXSTopGh1oHlc4jtZ4w4tRN+a78V51sNtBc9nETtmRfnkIdai/GVnDodrNHYAsVxD0V8ljvh3GiPw/nQMUrEm0G7bbHlHgwu5xPhrrfXpuTB1iY80zbvAzQg7m3p3IEEaUKWkFoJuolmwHTKSmELFTE23gXnlsFDSSfC93mtPlIHz74EYr/FUtjsWCeQTw2GbxYKgxrCfF9EpNp9M6Hto2vH2ujXOln52lYQIVivhgnEIo/VO5icCMWUr4UBbwMpwcrWvsdMbhIeKdU7SCA0EIsd+3xcTJUPhAXNCtDlWvSIzbvOT4I3BtTG3Z5x8+D6B4g0mCpSbgnJUQjJxLAUMNWoKTCjTDHoHuLGnkZAdQISMWKqMr0axqC+CO6Rsxhj3QwZyqK3RiNjxqMFnh/KFXbU4DiQ6rF610IWKXDNJ5KI2LyrzXCNPKCTWa4C98C/ItKbQZhFr5vZgZGC0Abj1UasV8fzTmiu3g0baguoJSbuE3PSDApe67DATU9sNLQZQh/1GuRqYt1E12EeTCsQguLxOBV3V1Q5C5uOIeZERNoKSuhhzZAyxSPk0qcrmD4Az65Gm8ImXBNGVuegnUGaEXJ1GZVKz8XnUX5nJOZr6lrcb8SNeWmfNLWhnRuNRXO1TlyLu63oYFnd0EhuMCpw1O1YG/1WJyM314Nhb+vCqUIghGIJRIlUH9OYJFAAkyPRi9x7I8R+U0gEt6Fanza6Vno3WCz7H4BAh7IP3pUYT76zilimGtjWxHDEjYVkEkVQ2cFFISKSUm781hkYNxjtPgwRzJoUCGmYd40jb4ql9ExwGSsjvS7EMhvz3UT/RZIJ6D6JGFGB2DQ6PHYbvQQKthgKEktQg143QQGg3mIXeRt6sQk0iN5DE7hzhsCMe79wCxsrif2UIruH2HmPhX8kfbNcBYuJ4ielN2c6AIBWIoci9G6tYvyAEs9+HybjvZLKwi7hJoZvnMkqVr0IVbjAddTTuewZ/uQpUK9LfQNP54I5J/1sOj88NyTFwRN4ghbDcmoJbQx8FPUBVnRFiO3SRjTqt5eC5GIBojk88dk17t2LzqbJgopEckWRGPakLOdxacfa6PuOk7bGyg1Nkb4RJ08FniWra03yjjHPJg2GegVtMvORj9s8KC4aPHwqPFmxTcMjjDIbJ03KCo7WUFOAUTekQcKLpWAEDFlfNYkeXosbl/Z3HbTDEJppqcGjkUvDCPQQMrFNKA/aC+rlGjKjMwEEB6BuXY2Ts6ArkeBmTF6ND+mM6c2ZgU9jU+hQclJpGDqL8f0EJzp5382I0CARK6q3QESO4elFRNJq9Cl72D0gdNdTZBYpuhOhN33m3PBJzla8FL1Qq0pmyCaB7jG6Y1KMYzzthJFaivO0izBfgmbEjD76naC1tg0Vz6hwK86d5kWFPOP6WWg8JCqv1VC291GRPoTWgnQj1U0WJOpJ+h7yC3BUEn+rc1C8hzAjUEeliyEsyzxa4Q42f9Vg9jhVJGjUr6OGQusKGpnvQMK1o2yu46Swqp6Z4o894HMjt1Aw8k4scjXw+yiiSK0DeaI7PuOEjDnGH0IsF9WHXUiz1S/p9/BbmbvxQXdUI5cLgM1itdnbQCU9FRd7ncaRxsn6Uh40KKxuTMTsLZGL/mcA02QIKvvFELOunUGCDGgpp55++gZ4UMAuarKB9FxJ2SBY5OYdO4xhdp8QwPAvI2DpDSxGK3/HBtkcwpYpEhd8HRz5MoSLnWGxhGaq6SLQ48bJzSiOQ8fDPx249wwrNZXXiPkFMpU2nkSRkW30OEks/kH87Mpf036zvoDhS20dnAQSiLYh+aIcDGmbsFeLjiJpnKAe198leqiOfneGIFpSOI1RbKdllA8HnEpMKrN5Ls7NzPrwDcDCm/XTOA3htF9TsEVCmWvYBkKnbx1OEWokcjvh9fLScDjv291uPeS9Y230fUqkpcksKw7pIH5auRq9SNLumpj3FGCBvVlAOtULpbZp4roayqHxdLsYKhiBsp48ukBZdsA4mFNPIcEdkuNRNvzTWsDxGMZ9fi4mjfaV9IpVgol6JB0DyjkSnePzg7GBFkrt89PRG/PvD9fN4MjKDcQk8xzw1dTmtWdwkCA8WRtNsIXxV8/wi3p7HptGF0VnpjqWRkk8BeF5Ammop89QVA/Gz0Ix6bOsdAadh6I9qI6WBZqL1Mj961OuEdXQGTVuTJK6d8eQ5ThOqsaXxFzI6k8Nau8yZEMJxL6jg7AYaZYzQLTYvXVhiIlE7cfhcVLOMr6vjnjrKRTrZYafZDtD9CZS2ICMUTYNjYpmDRKi5vQwf0GxIdJ1az3HxIvxeVfOI7SmnD2EtSaK+IZo5DZQvEj4Zlc3iL3LA1858nasjb446XtUfWFoJGxJd0CDlHk5JLOWUbnqhnDQEDtf/GxEY2w9r2IKDAFk8T8wGI3pwdh3bhcTVI0X6Yx5nGdM0Fod3DkbxFK/HFZTl8IkQB1Z8VZ6fBCqJyLSVSbERHHZATFig1HuN+JutlsfRHAwaUxETZ9yGugiJooZ6yydDdciT/r6VnweJqi+DhK10fE4bqmJQcgnkRT83bQe0XOgRnYz8XtrSuOcCKPwd9UgJMQ18Ps0bqR6trYF5slRzTOVp/aGvk+N5qxuNm0gztowynbyIWChx+SmjkevTi8WhhKVpTa2k5OxX41ONBX7ugHVsPFRScx2iDLFRhCe4di1M6pd0RxuimxTSFGLFutvcjaEZLbvx7nBQi2PDd0oMnafwAaU4UlAYcY4ZXaAoLJitQQ4YpMFaAh1WtHYcDaQI23H2+invPTGtKRbJyuPXil4Fz2q2l8Iky27Bo84hZi8Mlsyxrz3QzAiau1JMZCDkaLotiEYusDpt1AxaN5zGfDSbbAnWsInC69u9BxiuYxlKjc/mf3SED4xY05D36ZOee+9iwAAIABJREFUsIl6gwKWFaJsxqk+Pxrhp/sotukOSfaNv4JE7w+E7zGkw1AP8f2Gw98FXJHNnhOrrjOgTh4zJbRdiLzAIyaXSloheNzM7PtsQ6umJeYH3CZi0JcjmqOHsTc+/CbGII+NoKOnkTRKX0tAp5GDybzvFE5ITAA7VdHqUgWMpxl7TpjzKSRfCRu1sOg4tGq5sVmcnQnPUxNxzt5+JdTPtM/E511lrQMpMoqDGtNN0DgbKomspzmcKO3UzjyA/0pEzzWu4BRmzhCeEU8w9dMKGEDFOinR7Xse0OAmHUskjZt2Wjh+BbnH2+i7tJesqgs5hYy5A5JmgoRM9l4YfPLGJGgUNKHZgBEpzbGqMvxbJ2c7vBpyzBjOPo2EJRej0SiwmGh6GsVX+v39jbgoyM/OhZ9R3vkEbI9HWfWk869Hb7ADpkODxfUY5mHFLjQFdqrhN6gKlYHRzqggOr2eVBtHaUvAI09QpHcP47a8pGgIhm2xgLazGmuFB0e8e2c6PLvGchxDku0RNdSXS4TxWwEvklEIs6iH+YWMIrPq5+O90CPvEipqITA4DzyhtHUzuL0cRVgS4ThKKy7reM/GcSuNDxaFNZEkpV5EUYEOVcpUpoZbP0u0Lq1ElEpmOa6FwlOqV1EcvgFllWff/Vkc1zFGy5CHbY0rURwMLTUzeoqnJdKuOQsHziijkatIPRnHJUeaBB3PNE/qPJ1axTqel5uI87dfrEnKFXIKoRDRnJpUIi52PNqxNvq+62K8Ux8UJ2oBHN1dJlw05MKJQj6arApw9y6gIAoZ+6x6edxU2sAhd4GY6SnKg6GLHpNWOkHG5+Pk2diLxmlMaZRT04OGSeSBEn4lTKMaFblxMkqm1QJkk2Xm/cQe5yEWABeI8Y9QC5beYM1go4ilbs/H943agHqj/C3mYBbPhCrVzf04LkUU2BjmmSG88mS8R6uyrU7jJAL6afYhrRsPjT7Vw/rxbHjBPYTL2oZ+wdxy9IIZOtNCKRoZhp1MuITPOyGmg9OjKE0BqX6bEF9pKKFMd0iyX0SkouGbPLxzkoV5KKQZrUS5FB2hDjYYM2i7SMbvkXJB74c5t5FXId4CR8S8avIAkUhubCn8u/0UxpU5AZWyTFOkCKGmJjQYUkrz3UE9B8nfLM/DXEU3j7lja5whU0KPEf8vfiXsbGN3hhewvd3tLfHpO+c+JSJ/RUTWvPfv0NemROR3ROSChETxT3vvt1VE/TckaOTWRORj3vuv6Hc+KiJ/V3/217z3v/lm105nu335OQtzMNxBucMeuXF04c0vxFJ5LraNG+rBbGCXJzWz7ujcYMpnYpiD4iyG3OAGVEK4oKZhhOVVMIJiU9jWkAb52QmzZBgkN6pIIIQeeOzNmteC+0qfQthKDUqLHt4QpSiRyB2+AhpCwgFNhYtUFcxVWDx392709vazCDW9Hn/XmCeZ7G70IOBhRUTwcuWL8XdvPx425JGX4+9XnkboDwLdtU74e/RmvO+dZ/G7ZlCQlCvgebTPqhfLMX41huta44OLvAOHpMH6AY0xewrNYEMvLkKDQQ10C2Lk1EqwEwIFYxo4qVqSnSflhAAP+tvUE1Oir3SgbK0dQCLYM0cNxrfy1HCtzOJ42HCpBtfBxlVVJJ6/C+EVrJ+ZL4T3d38kno6zuC7DprbRc/1yvCbUedgpRAePYS1DBWbhHFFxjOGw9e8Lr6+++4j49H/n4LcexdP/30XkH4jIb+G1XxSRP/Tef8I594v6/39bRH5MRK7of8+LyD8Uked1k/hlEXlOwiN/wTn3Ge/9tjykZdM9OT0WYoV2TNppxAey9hI0PK9Gr8TCEKyS5cOTS8Gz3P1azOBlsMAmy4PJVSaiRmDUbRMi1wubLTJPHiCeBBRaeP5342urPxdXBRNcduxmjJrH8vZVK4rBayBJy+sCOz0XN8M66hdMLUskis7kt4B6em8cI6/37YngAPrHJAATdQLoC1kbe/oYGNJh6ysZYQw7yLOmFaVSeQx5kRfjwq5cglFVY96Gzgwx3jZ0rcdiBxvot9VNpKZhMFGoRWhw/zt0UhHfH389fG//Ak5DqExtoELZnI52EeG417HZzGhR2anYr4XZiPyqKjst5/EUGCSXqdug/SWFAZE+HTWejiIqrwyGFLsFhLqAtuLJyoj5WLU8OkkDruG2xUGCPhGRjfeH+yrA0NMQ836bupZmJqMDV8gMZlovTUeeIXJPbdXDGqddGZasFxGZVe3qLOpgDrPdfMh7b2r0vfd/4py78MDLHxGRD+rfvykin5Ng9D8iIr+l4uhfcM5NOOcW9LOf9d5viYg45z4rIh8WkU8/7NqtTrovYm6C0ZsbMVk3+Wx8OJcnN+XBdr4Uya2ol/uZO+8QERF3BRW9K3EBlS8qNw8MYht/8wjeT8zB0O7tg9LZRDfIl08lJaVUuPNhIApAXVCajZP9spJeNYGkuL8fxyOXCYupAjnEC5fiuNxaDbHjsXxcdKcBN3wJtNX1Ga3+JdMFTwjKgXQGOqqjKDm3BWJoGJFkWGrnHWCLVEoDenCk1bBTBQXhm2fjR7MqhtNDzH/2x2O/Gqi67OjY1k7DiyXCTzcFeoB1cBi/63tuiIjI1++cid/BCShBF2CNzj/ua+eZQUqHDMI377l6o/+3GZ+v3TjXf80vwnvWMFtlKTpC2fNMMIf3m6/GE9L6pfg8mGvYt6Iy9JXotT4QAGGr+iLCJLpBkLq8jFOFOXIiInPz90JfsTNmwK9VVeKySjs+g2/cXuz/baffBhLYzSzyYIzp6xhWkMTfbAxKN5bKg+E+kTj/cjhhbVTixpwAOuhpJ106fvCdv2hMf957vyIi4r1fcc7N6euLInIXn1vS1w56/aGt1071s/5VM6rY5amLe7YYJ/hiPniyWTd8l50p1Qa+n0JRzGYnGEdDDj3Y+glVEalqWIh6uoU1EC+pcDLDTw40s00VfyBigIUy0yODcEQyOZKh0kS7yznINRbixnatHnVhrfFUkJjgmiAjvJM5jtPKxDhdiJsSGQmNYXL59dn+a5lT6Cs8s1YtLDwKl6QgBGORhcfOR3H72924G3XGw9iXEK+eB+ndkgN/itERcP8ivloRTqzmTgO2d7EcNtEU8N1fuRkNcVJVTQYa1dhMgpBJSEqAzuTi2N6qhvvN4JTXRhLSIMmUUNxajxuAbSYdnARSy9E5yTwG9k+FAdNR6UE31gABNB4TZyG1qTq/U4vxtStTkXMzMU/U2BMwMJuLz66SDvc4htDg+lw0tC11gDYwFmmE3jpFFieGf6pAAFKQpate+f4BleG2eVd83ChyACowvm/PI5U+Gk//Ye3bncgdhvr2D3l98Aec+7iIfFxEJD093o8VWgKLOzonyit7kWa5qBzz9S6MK+SLJgvB+FxfjeGd/LvAnGe5AtIWk1dmDh6WhiRyl6P3UstCMUITgiNjcXIkODoMivdGXICtSaBoAF/b1b+JBSfevNoK/d6tDI9JWhLv9laMp9OQE15pbIxp8KQw1LTXCH1Z3o6e49QoqI/12ZBPZ2oMykLoo3HQJEJgqP4V9Zq2UJTTWQcKRRNzzPGs1KLBY1jJHMqEXOJ5GCE9hZG2owa44cpsuN99kGuNfB05g1NADem1UqRdQszf4IApTKf2ZHwGy/V4D5VWmPdUgOvBKDuth+giTJIBR75RjmRBKdE7H9+vIOFZuq75g0kkXBPMmKEPJYAMzo1HA7+jkGSGYbYbcX5vShzPnXp4pozjn52IvzWWDX0spqGNnY0Ddk2RV0UwtFJTwA/ZGBMavWcBS9V/SQdiIVGRuFYaW8OpFVhVbFW7lz816LQdRntLidwD2qpzbkG9/AURMV7hJRHBwVvOiMiyvv7BB17/3LAf9t5/UkQ+KSJSfOy0N8FlM5QUAGFS6u5O9OZubQav6LnFyGu/1YwT7dpaMPakwS2ciw/HjnEVTNQ0k5g4MsqI9QX3AHSAEXydQiJ4EwU4Zlxqp+OknpqPGwiPjCYmzvBNczUa8D6lNI7Sd0CTa4lchkm4AY0D+769GX6X3DytnXjdjWYwSGSN3APPiCFSRinyDt56GgTb/tM7MEgwbqWbavTb8RmnuoPf74KL5vZOdALOfjbe78r7FOIHBImAlsKQHw1Q9QpOfC8shelN+GqThh6OXUbVtbqAteY3B0v8qeRUxNx6dS2ezGyuuxdAzYw9MnU5zI0G5P8IAjDBFeRjyYUn3SU4HecVEQO6Y6MVEBFppZRvCrrSje6gKdlGKHb/VcA/LyJBreG26RfitV7/ETwPpflw51Ftzbmj6DQHKm9WYzMEZbTTRHMxP2ZztpdlSAinX10/uQlwFrFgl4JH6uxc/ztHRLH8Hx381l/U6H9GRD4qIp/Qf/8lXv8F59xvS0jk7urG8Psi8uvOOZuSPyoiv/RmF+m10tK4q+IQuoCySChtL8bJMToXJ5KJQNyrRiPBOPioeigbqeg5bm/HI5slTMnSSZIoNkNV0NvzV/C98TAZr9+LYQ4PXpixVzXm/57oqVRh1HPYbAx1QA+rh0Sp5RpagPLlgYowjzUF7v42vPcteIlWoNajdjBitCmtm2D8Mw9DmNcTyB7ip/v3qZkX/7RzYGIfAKNnXaGguel4LXLrZ5bD2FNYfWQmbmBLPwK6jc6Qym4Uq3mn+QUkakdewmb2vjB2TWjkdrEpOKChOopKSiP01wG9hCWVGUb0fxqtdvmHYkhkW4ECPdSe5LbB06PPjogz1pMYnw1rBro4yabmUcSk90AenyaqTa3fdAJe20K0VsOTPB2nzuMUiJzY6QvhHt2F+PUSMf/PDQq6rO/HtdrQvb1ehY4BOLM4qQw+TQ6r3jfhgCkXUhqQbG7iFjhgCI1srwL1r5xyLPWyQwMaR9oeBbL5aQle+oxzbkkCCucTIvJPnHM/LyJ3ROSn9OO/KwGueU0CZPPnRES891vOuV8VkS/p537Fkrpv2nTMrAioMR0XCKkNqGu5qpOCMaUiMvobK+FISMpdIgosy98BJ0o3C69pSAENE4OZXSzynBp1LLaJ0zGU1NCEqa/Q0MffL6Hf5pWTRpphIW8T8CpQNFgMJrhCL7UDPdGR6bgwq8XwvS5OQyXo3lpYqQbjR6M+diqcbBLFZRwXxLEt5NCeGp5DsYWXwKDDIBkvUw/E+JUs+FnIcqkhJEIMBWNkxHs8uVUuw6hrwp8wy4R8H7O22t3EpkBeeQ0lMXZeuRL7VQcizGgrannkBM5jHqoBz4yivgFrwk6nPYQ8PcqHaYg7ylF/UH7CTpSUmSRKzGmNhB/F+qTUJtatyVYmqJF5XX1eXKuMs6eVvRZiV9K5gJBMetDo9qrYLOf5HPWSyF+4Fq2IvsZTJkF5W6wuHPjasWmPgt752QPe+uEhn/Ui8jcP+J1PicinvqXe+VhgZUdkxiQFRzoerbgBWGuTw0a91IR+K7yLjhkqikisMgE3JNGDh5xJwNvCvzkoGrVw6rCEYTONwqM8McskDtN4Mzm+ERay/jLRm80NLuwM4vw8NTCpa5zqLDpzQ6oLaSzKqGrOaB/a9Wi4HDwhbtgdIyHjYsLffUPMmD882obNA8Yu6MWSi2WIkH0GHDVjOh5M5LbH42eNpI+8T4kj/hA2SBqe9Dg8dT3FMcTGMWaOxTY8EpsxTJfXIqUKqH571LDVzdsfpMSG6xqbKjWTE+FLvUcS+yV0HzS04Q9wcqlTYZQKLSZiQf2dKoUfYWEeq2Atv9Yh3Je5ONxjZllFVJDMzoBiI7OvJxjQhScAFvpTXN8UasrixJhAbB2zdqwrcl1PJK3r0Yx/G6yOmdvRoLRPN+XBRo5wYsDNU2m3Br1gkXg8JLWCn+DuD+z8PWU/LCFOjj5mFenTgwfW2BkUmSbr404lvt+lWLkxEpLbHMdLNz5YMNUljFg/WsW4jEwNj5WmVWYxcxFMpV+M4bLtx5QWA0aEUn1940XlL3jHvF+raHXYoDoI72TUaJNDn961ccAU/hgxZAjd0MgYhC59F+gg0EdbQpGar4lYrX7/IMGWhMdqlA8wBtmZwcQ4FbDYPF7vpDU8A0NPkZROZzB0kaJTo4ayS8pp1nMgtGevF6DQ1roMBlbzvhF6ZPy/Z0iizvCTmVXEi4i09FqWtxERaT6DE4Q9OzoBOOmKrtt0nVxOmHP4XtsoS6geBkffjH12H99Ps6rf6g/idxI1GOTsqWotQ+/4ufzH2uj7dKQ2Nba6/BZIjS4i884qPD3KNigFBa52Z8YDBrMvdiIiclqpEeARMzlUAra9OqXVu/BCuzge5tSoN5dQDQRETH1Hr4v1QYNHDHhG//a410TIw+CIdIhxD50NvRYMJr1MepYd5cbPwzDUEE5IDyHKatUQDrPfotAGeExaqCS2St4sQi40aMal7kCtQPqJjvZl5z2g2iZH/bXYr9rTavxOgWwMdRPGcMrQBO/BuKCYaKa+Qgrxe8IcrZH6IGVkeTAWvjw8bNSXOeSpgvuLbjzkuGdMvqy5Fw/eM9aeNGko9VqNRXgM2CxS5cHq37of3Ow4d9m655DzUloLgOsSAAvLDzAvwufRVW7+HkRv6FxkwYeTzw+GD6sl5OJ03TmPTQFfMSoXbgpEONHxy2+q0T9+If3jbfTF+Yhvt122zHBD/GjCUN4M8dz84zEGLSWSpGnMEUc7UqhaHJHKQxRUqQJnb6XwiWMgnrQxclJNy+G6Vt5OitfE0RAFTV0LX/Ao3h30JDrQdHUbcVKXN8Jnq2fBdgkRlAR2/UwwEk3wv7AUP/9iOI005nDyamATNUJDJAAphs5N2sjNuvsos50cPHm5MRoh6AzbxoOhaCPJ2L4AI7OiJzMQ1ZVfis/DIJddePqJcJ3mbpgspO5sD/qvljdIcPdTEeyUSi9iA5z+s3jd/QvxfroapyackPUiRpPcYWIRp4K9tlp7OD/UiM6tDpoC1KRJD581jeUu4EPTMxGdtvtiqHNpA/lFo13YhNf+XECqNbAZ5rFW7cTYKmN9U6HNnCaEhBje6YCKumdoKuSO0ttYXzonODcSUps69j3oQvdwiqQi3b7VB6SOn9U/3kY/JX0lo5QaCbIfynacKGlqcmv8n4RVWYRXTDKOGN0uFrElahPFLyODXkL4kfAPK/9kSFyWi3Xq6/Fau1cVtgcedOre0tCaIS1sxO+TYsAQTgkWTlSIVvT4moEgTBuJ8YmvxuvuzOkJhglL3IMhHQhPbfM4r+NZAFKogQVIb8zw6h0ob+WQJBT9mr8OFk2Q5XkzmjR43As5HjOMd4W2/+yQEyNCK91pnCJvhbnTQi4lAylAhha66ojQ0A8zAhnkHDbfy+QQdn875cD4dhcAHWwMMnqyWdgpDVnRHk6nrUXEptU7JjMsNEz6tTIMe22BwiN9IYTWUpgPxPmzRmJnXUOZ2BibdIqMoRjPtk0dYF0faazVDvImgrCtnQoSFdSAuObuhfnfPse6DZ449XnidMxCSo5nv3/NwdeOuh1vo99xURxCed9TMM5peJYdLIDJqeDhF5CMY+l15Uz4jQTtKuLsfSFsxKv58BnSsCQd0QuJAhj1dkp3AI18HokkXfCEOzbhqfQQU+xq9W7mcRgp5BcsNp7hYkW/DUnDop3Ua9HT37+IcIIamTTYD+kBdfUoW0RNQBb3UAdqp38tFvhQ6cgWKd5vITRX1LBQ5ikQ6AGD3kcYjTJkg7AWToH2vFgIRinLjuLNs6BGaCPxN/nu9XB9Qgjvg0IYCKiUhlSqiEFT38CMTxcbzPz/F/9u/XSs1zAU1/outBjIYKpDkKL3DirruqJkuhRxh0NRno05kOr9YMDp6HD+G3kbOYlyoF7OPhn6ncapZxQ8QNmn4zwZ13GsteExw2mycWbB4UhuSJhmF6fEA+LoJujTG3I6FhHpat4iSyRTi6Ge8D0KH9WYC0FzGn7kKfG4tGNt9F2uJ6mzYRGZIWvDk+kwOQQYpYli5MDhQZIkY8Pj5NqEwareCQiN/Om4gLvX4mLrzsHLM6OHlEB7BuEXDc/ULsTvTMwg7KRtD3q+WcS+Gbg1pSVW5GaZNNZ7pMoRN7sx/f4I0EHyvbEveVT3GlEVf6sEcimryCWlLsVl3P3wehvYeQEVBcvXJ0eDwWkhp7ALuKLJQ05B6Yw0zUXltif9RBubAuk2jGCL91ppwRAPCcLWUVRm1yhmgDBZGDRSInHOcTOsIrlqCKFpUGk0/zrEQtAHew50VPwXwCn0nhBeaaJ+oQGhl6LOZf9irFlog4KbqKGJxWC0d+EIsciuH3IBeodC8jPKVsm5x0b0mAnNsCKX42FtGDGaiMi2VmkXkGdL6Cogh5F/OczJKiifmSDu17ng1MAqWzvRVaGqxmR353EIuu+F+5n78tFw71x7yHvH2uin0z2ZHg+T1SpXO0N2eZGk52cVrwnhkyEUqzQyZNasz4YHydhh8elYVlAHMVn/Mzfj7Jp6JpKcbRXC61z4NFjWl/HxOGFY0l7vRINj5Gos3hpWMZhJoDri+8bdv4Qq3vMLsa9nyuBPaWrMvhM3u/EcPBxdrBzjK38/jvH1vxU+OzUa74sFcDQyZgRGy+B6R5LRkCkssJsG17u1BsaK/hU3A+MnYmhiBO+nFY6xD2TX3EjcGFf1GTSwEVyciWO41wTplhF4YbPk3zYGWTgnpN1e2YsGuqEFd5MYz8yHIqncRCGM9+0sKKtHgZIxfQQoPZHGhDDf2XIY253N+LyMdlgkkowlxGmwueeG8M1QQ6KF045h/TmPyIxpp3X2bw/UJMbSOTkZn5FDqIoOUkrzT8O3ougsVsoo/MN18/p3bTQ++/p0/HsENmZjIby+/IEjMrH/6uC3jrXRd+L7nle/DB3vJ6QE78dJld/QGDGy6TXG+fSIb/TBIiJbV+Mvm0dMRSRuGqMleOLats8P95pGlL52bxMuBwU8FBdOb7QDxZeNWvS2quZV38e0Bca8rSXtrQMKXZpn9B6QM1haOd3/O/dcHKObX1Q2jQvDuUPu3w7JOnrs1/4WYLHqQTVxbE/0BUyiy/fDPRIxMzMbE4M7avA2oIO64WNIxRArHggsJvzZR4tt95aR4ANNgtOYt+V9REQqT8TNcP9u6EMKCdFr8BZnJ6LxMU3fU9MxtEGDtq1srKyKpiHNXYt9NH6gShYUxOBvr7aDIR3DaxQA2VUlNF/A6Rinjso3I01C+pnQh0kkZytNxMx1E2aNRuZO7GtlJnx2H6dXw8iLiOToaBdCv5kAb78W53xDTxDUna4+PsjNz5BMbsj6FBGpqbPWREg0DyeS99j/eaxVcyZ5UiGHPp05OyF0/El451tq7VZGlu8G42KqN2SrTCOG+9jVlf7fy/NhsTXXYWjh/VpcN30VwiiIzdW7gzDK7QYogrFYvIqf+BLoVtej/J2V+ycqQNFMb5RG/95uNGjcQIwHPAMjwnDEdjEsMlYfb76Bvui/7332jf5r5Atfq8V7TD8WjFdzLS7A8flIC1B6LHiZ96FLS4z6qbnQx114ZekVLCrWsSjyibH1rW2oi42FDbUHOcfEhr+jxgU5B+qoEorqdRrkz4P/BZ7n3HQIbeTnImvrTDF+9qZ6ezxZ8lTCwjuLHXMMCJG18UrUXQAB0nsyXjet98Bny8K9ezvBK2cMOlGnojkMj3BFE2GQ8afi/e5WQn9JkMewlcFte+h3Bo92UwnXCoBL1meRB8NatGee3Yq/1YCYjtMiuPRj8bfcfWhXqKrZ2jXMc9iIFKpvJ14N97D/Dsbs433tKcSWVdN5VNlm9XHsnkeI+W78/crjCPtqXi9XOzH631pz0vfK/aQiCjZQzp1BeCYP2l5FpFTA+thOCHEPPog65BLdA4ghkaRBo0KUmLHnpkJImEmwIZmYvRmNwOY5pWkAoqAKNsnTZ2JYaXEkGFLmKvbag4x/I+C1Lz8dF8vaXliMDIPk0sPDZTk92bQhxN0EqZb1YW4sGiYaYmMH3ccJB4wPMno7jufuEzp2RK0iLGWJUOKsGW7ovRA2yfx741glCs2g5Zoegp5hBbd5dk0Y1N1WfB52cuH1yQq5iQrkiYkw/xhiqPhoHYt6ykuPx3nKDWRmJL6+r14oY9tkUN1QzWA6JKRk6Nc6SGx+SMJURGRcT6cJMXR8z5LkPSJXUFi3oCI9iVM5wqcMjfW/P4/rY7wsBMswi0QuvZhMximxeAMO3MU4BlvPKXae6B0WxpngEZFEyM91SrZx4nleAcKKCWL9szk/GOo66na8jX5P+tA/Z6X4gG6VoJP6xmYkNDPVpjEkn7iYbBFSJIXNKgm7B1SAEnbXR4MQmrWD2LLJFZIh4FScoAZNZEn9GCCANeQP7vSU+RKLcXUrhjwspEIPi7kKO1UsV+J3mIhlvNh0Z+lZ1hGq2WkXtd9AaCBB7D8f+pp6BnKNiCe7q/HZnNHv8Xi8Bb1cK+evgSPn/GT0TLsfCDz7jJdvgcmUieuEgpq2HnjgzPjl8LkbK5GC28Iv9zbBi3MxPs+FckTcXG+G7yXCAdgsbJPk+z0YvNvrMT4/MRq+R3ppNqfPlhW5TI5mdMMkHYhn4RLCl+OFwfAIT6ITegLYaMWBy0LZyqiPK+04d7cQ0+fp1eZ994B+94ntmGCn08ZaGm009A5hH69hUzpw9OoN/tkjTJncOlqU1Z1HIRipP4j2081k8ktHA9m8/ZD3jrfRd9I3sGbQpl9EOOPC8PJ1izWa/uyDraIPnRWF9NT7MFEUCBFOmMBa26TBkbKwjgmsMfnZd4ExEf2an1ANYOK74TFTjs3CBN3e8ISl0S8w1rr82lz/79JZRXhQEWwJR/x3RU/ZwkYFoFTugIffruHQ72VsQCMfCMnNbIMYlNhoiA0Jk+AMgiE0hEYTIbjr69EQP30qhPZWqhANoboZvGMTd2HeZAzyl7axkQa6CONomPs6Qhs3bkYKZIaVzs0SOAgOAAAT6UlEQVRsJ37zwfuyhCTlNxPeNebk1svhfs++a7n/WhWoo4Yac1JwGJWGiEh7IrzexkmDxUQsdLQ+MNHL06M9p63s8PVlRp0OS+8Awfae6S4QJknOIV1fPHWTHLGjsNRmY3juKEGhkRnc8EnG53W4CncwrhcwN07rDwPu60D/wBNAW2lb1t93RMVZ/9vBbx1voy8SH6AWZa1+AIgaTBQqTNkEbfeQHIUhNQQH4+HXXj8fP2tJQDI57iFWCix0WmN2rdn4GqtULQdBwjd6OsMga3uoTB0bjZ6hoSlm52NMPw0Ss7RSVhCVNHclbjb2Oo1c6clo0JaBFrHFwkXDEFhacxQkbDs7E8Mc1gi9ZGHdOk8+St9A2K1thiLxtNEBPpo6qF+9HZLOUxPR26S4zPW7ceOzIqfxs3EMmYwbhiQipHNfDbWR9j3Y2G9riQ29O+ioTBaAHEPoTXBKq14JBuX2azG2cfnJuAFsKjLK78a+0iE2ZkhPhwXvJxTYdOx2G3HTuLEWY+Z9um8kznuL8fuvvxKkJBcur/dfIwKK4AQLt3Upsp6AXg2Gmhjisg3o+qsRkJCgPmbIhcV7di049V5DVAn6CdI0W+E3mVKpiJfg9BkslDwu7Xgb/Z6TlA7kyK0w0FXAsVJzSHIi2WbGvNWLt0cI3ojysPL9NrD3xZua5Z/BYj0FQjcu4nLoX24DRTXzg7TBux0kZzH5thVF4sFnTqKq8xMxjPHkzGri/kSSsdiermKqFNHLNcjjeH54Bel+Li5iMz6rO/EIP3s6GnULAZQz8IJhRbjJWiNvfUMQptCFnwHWenkzjtdjp4LxOFWOBnW1FuGEZuzPjsWxGgHX7nQxGiSbB9TzrSIMYeMxmo0bAaGqk1MhgV3txO+s1uMYMQxiUMxZwEs5LmPaB8I82eZLoLLWmHUN+YONGhBrKgZSh3C6kPXRjBscFoYUuX4Mckmpzb1y7OMdFSxyQ4yoiMjjTy6JiMgE5tkelMYWRgc3Rj6DYbUSPP0mIJ16Skwh90TZQuomOBnccDMVhMD0hF6HNEBxOX6nfkpzfQjpJCDTCR5+M/onidy/cNt7Uo0yxZiBumCy7bYKgCew4Ehw3W2HMAUha1nwfbSfCguMoQvy4fTAsGhUt50RHPlYbarJn3OPR33XbcRlLXZOgeVL0xH3fb8K4XNdjMsQ+u5BD3T8jXCt2g/ExVpAVaadNnZLw41MFcfxqlZwZhADpmLXaD7c904TCXCMtxlVv4XwDgkeGTdVPvnmLSCB5uIifv2Oerekfma+RY1X/d/EkE/73XEMWuB1sRhv/nocg8Zp6uGGTawKLDYRORUVoiHapD0Z78V0BERE7t4MeaaEjiqraNU4kZdm9qnoHb/4wkV5sI1djkZ/dylujNQ6sEY6grzKCTah29BFmOKbKwsDfeT6IFFcTsNdHWDzHSC4r+0FrztBQTCO+gBAjgtX9MT1+TinW2M4SashzUHNdPcpPC+TiQQfliujoIpMoLpWSQdCT7yhYj2s7OYGkNYNInHfYO/t4tRt2hH+O1FE5Uhb2vdRO27IjslY69ImDKFOZsoG5j4fDUpDES0pPKTTgEEOS5YlaBJGUaKv8XHGJMm1svhkMOD0VIgTLiiJGMMRPOIzyVf958H4dd8HHiFoye4p1NohvNP6ZvSIS98bQj2rtyImewqC1s+eWur//ZU/f4eIiIz84Fr/tTYMliVKE0VUiCfPKrw0vxC9bMIVT03FhOf2iKpC8YgPL7RPWkeY5178n3NXw4a6fj0e8TNfj/fdhtqUqEZt8xIQTtAvtjACNWPJC58xgi5QTufR18qNaIhFGTMPimebhGF6ajj5Olk6bTMqsZr6TJyzFp65DcqHLjmHbE4gNFGYQY3HDWy4yjLbpZ4EuZC0zePkx7lh6JzaKmL+cJSoMV1ZVcGjZ1iFjn7r2DZYe0JOH5W1LJxB9TycriLkEA0BNIVntzsPPiitl0hwbmED6KkSGp261DRyAgTyKL2KP4A2+yjbsTb6qXRPRsaDh9I3rjCCprojEulgRWLMzV2LD68KzvTsmhIrYS266TjTTAFq/lxMbE4g7so4t8XB69ehXQo90dQHw8K4vxeNEL9vgtfDcg4Pts2/FLw5BwrjHmh9+4ydYKNMxCetLJ78LNjMRhCqkfcGg8JqVm5cVpGbgN/BEzdkVBFG6hSplUmNoPF5wimJKjKWSqp45U/FzcTGrvkMaBpwEqDBMu+VkpLccO3ZlCAzSY9ailosBM6iJsjXHAFfyjGfvo9rwWOtqURmD+GCzZ34uywA6yoX0XIzJtPnkdsxyHL5DHidQBZvIZMO8lwZBLRfWEdI8HowhG0wz1LS0RKi5al4/Zv34ilrejrM0xr3OhhEtx3Ho3RfQyrzOClzD9ShSYB3yLI8Ecb+nafvxfvCD+RTg2upDvWwdQIlNIeRRoHdHjZ/S3YnnAACBoCA2tFcVneIctdRt2Nt9L13fWNvuHEiHXhUbmeggfvnWpmKtWo5ARHpC1FnQRiVB/1u+Vb4rcIlHBOx2TBuaypX5FVqzsTPGl798mxMqDJ+ubKtAuP4Pj395e14E+bBUD+WEoS2MLqIN5Orva+jit1ubSNuVncQ+67uhn7fWIO3xgmsf6fA2MkE97aiKWZmo0fP+gK7bxGRzFfCJlt7Mm4KttmLiFS0spM5AfL83NuYGXif/lUHlM5tPfqnN8Gxfw6oC/X2qJmQIjOlhqvOfjaO1fWfRWXSJD6rlc9NfL95CnFfZZOkJF97Ajq/ywwZhr8r5+NnVztxA5hTGvGdRvRch+W5xpCf4AaQhqPQNOjhEI58kQgzZkjyHRdiUvnle+FEyrlJ24tUmjQUB09Dn0FBU0oJ/5hwbYLDvqfOHnNTDVyAgu12v1vIea3DGTMuIaL6WIRnzgmBEs0hKn0iIvmvhrFpDwc4HWk73ka/5/rFFw2lEayPxMXaOWDAixthBuUgdtCFHexoZWcNYg5MHuW/PxhoMv8lJhVivH0yLhYWIaNf1olyrhyTjGuNONH2ioM0tfx96tk2lL524qX4PhPbk6+Ef9ffE/uSX8cEVWcst42kGBbQG6ux1iGzGgasBAoCngqsYtan8AwSULnwLzHfTPRemImnqDvPKQfNqzHE0EFobvSFsAHRcKRwitv6vvBZesyjNxDiipE/aZguLQxtcwc5DhMrfzkazy7EYwqb4XfX3w1tYp7MsKFmdBPqIMacRi7D1LeyO6jihUOQ24VxU25jjyIoosusKGwNsNn1bJxnBrEtHVSzgDh7Sm+BmzjHy8ALdZKknYohk6IaTSrbldaHFynVZ8JvFDfitVqjQM01BkNfuR0U9j0evn+vEp0jEr3RzzYDbg5N6ECcJzkVaKqPAwEFR6eR1XkIKmwWajZ7+F0jdRtCt3zU7dCNvnPuwyLyGxKcsX/svf/EgZ9N+T6Mz8rLSU3LSe8oDqEvd+GApQG+MU8/A/rcV1470/+7r0PKxCF2fEevX73n7CK4d2DArRL4T+5e7r9GniATYSHnSvpzFABBMZraz0YMyUv5bnzfKixZJ5BCdKewNqi+1FsHXw2lDy+ERcyQzsYKyNdeVKgp0h9phGU7JfWqZhAzZVgrEb5RWCtCE8yhtD4Q8gOdm/H6TKYVVS6xjgW8+HsxF/HqL4Gnx3RpcTLMgJqgY4IuRQ5S/NMiJpWzCJEBueVRdtxZ0HxU/HqC7jitcNdqAUI3wKDv5KIBL6yFX0mIy0N/wOhKqE7GMERdE7FkxkxoRFN282yYy4kiKoQxrOiRdRW39uOkNBI0eQpU26QzgDNmc2b3cryvmW/E+1p/l9KBg1+vgSpZ25iWl+L1c6hEHlYIVh6PY9SA0HxHN5s0DD3zGjZenMfUHM4jjNhQR4Jz47i0QzX6zrm0iPyPIvIjIrIkIl9yzn3Ge//y0M6lu/2EoHnUjPturAJXjgdlEyg9HGUZBUAQ+shChrGj3DuEgZH/HewPUdh5P25Gk3MRwWGIGC6Q3Da8UJ209Ko2n8VEwqTpmAgzjND+ZRon9QZRhGJanSKRgI6CFDxWE39spygmVPl+7bQlteL326hstQ23uQ+9X4SlJkGoZqyl5N6h0d/fDr/hqEoFiu0+GR8e8usfjzFm30ZsWxcmdW1p0MxCU/CaOgCdsvI2kTTs4nCCL/Py0kCQsIioHyNGoVd9Pc7pHriEmpNWLYq+ol8W2mrmkGOBGEleT3fNGYSycAIhb1HK5hFrSzBGm82w+VJLoUXOHxN04dSkcuk0/W/tP3QbyExZ0KhoA3Uw7XGMgb1MB/A1VP9irreeCCG5RKU95szMXAhFsmaAp1s3BErKUA9tk224fuwEsvleEbnmvb8hIuKc+20R+YiIDDX6Pe/6CUGDNrIKkZWxI7dAeqVZdupuUnDFdC+5qNpDCjq6e8OrSVNgP7SKvdEZoDmQ0Kl+JRQGdc6BjxysjimFfJHH5CCYVz8Ei6dGnHHLxNspeA34mx3RUzBYWaBgbNxEIgKqnWL+YPCo2oFHzFisPZseRSjIafKFWOwz8v6wsputuEEQHmpHcIaqGoDI1rTYhyEIJvuYi+jr+MKDa7MsvzF4jwn8tTpzLRjPBJoDG25KvcBEiAGJ934VOJ59l9+HJnDpfrgGN3luwiYB6ABd9Pjdth6SsniGhesQFjrPgkIFQmDcEtQFqljnX4nroH0Jk8o2DWDke43hp0+vt3hQotbe52sc736xGR5bC1W0ZJRl4Zq1LOC0ZsBZqMaTroVd6emz0JHEeobYyhcGUU9H3Q7b6C+KyF38/5KIPM8POOc+LiIf1/+tfO2v/Pprh9S3v2ibEZGNN/3U0bfvlH6KfOf09aSf39520s9vXzt/0BuHbfSHnXUSbq33/pMi8snD6c5bb865L3vvnzvqfrxZ+07pp8h3Tl9P+vntbSf9PJx22KnlJRE5i/8/IyLLB3z2pJ20k3bSTtq3uR220f+SiFxxzl10zuVE5GdE5DOH3IeTdtJO2kn7rm2HGt7x3necc78gIr8vAbL5Ke/9Nw+zD29D+04JRX2n9FPkO6evJ/389raTfh5Cc94PR4qctJN20k7aSft3rx2/crGTdtJO2kk7aW9bOzH6J+2knbST9l3UToz+Q5pz7sPOudecc9ecc7845P3/yjn3snPuG865P3TOncd7Xefc1/S/tzVZ/Qj9/Jhzbh39+U/x3kedc2/ofx894n7+D+jj6865Hbx3mOP5KefcmnPupQPed865v6/38Q3n3Lvx3mGO55v18z/W/n3DOfd559w78d4t59yLOp5fPuJ+ftA5t4vn+/fw3kPnzCH3879BH1/SOTml7x3aeL7l5r0/+W/IfxISzddF5JKI5ETk6yLy1AOf+ZCIlPTvvyEiv4P3Kseonx8TkX8w5LtTInJD/53UvyePqp8PfP4/l5DoP9Tx1Gv9oIi8W0ReOuD9HxeR35NQd/I+Efnzwx7PR+zn++36IvJj1k/9/1siMnNMxvODIvKv3+qcebv7+cBn/wMR+aOjGM+3+t+Jp39w61NGeO9bImKUEf3mvf9j773VbH9BQt3BYbc37edD2l8Wkc9677e899si8lkR+fAx6efPisin36a+PLR57/9ERLYe8pGPiMhv+dC+ICITzrkFOdzxfNN+eu8/r/0QObr5+SjjeVB7K3P7W27fYj+PbH6+1XZi9A9uwygjFg/4rIjIz0vw/qwVnHNfds59wTn3H74dHdT2qP38a3rM/6fOOSuQ+1bv8a20R76Whskuisgf4eXDGs9HaQfdy2GO57faHpyfXkT+wDn3glKfHHX7Pufc151zv+ece1pfO5bj6ZwrSdjM/xlePm7jeWA71nz6R9zelDKi/0Hn/hMReU5Efggvn/PeLzvnLonIHznnXvTeXz+ifv4rEfm0977pnPvPROQ3ReTfe8Tvfrvat3KtnxGRf+q9Jwn7YY3no7SD7uUwx/ORm3PuQxKM/gfw8vfreM6JyGedc6+qp3sU7Ssict57X3HO/biI/N8ickWO6XhKCO38qfeep4LjNJ4PbSee/sHtkSgjnHN/SUT+WxH5Ce99n3LPe7+s/94Qkc+JyLNH1U/v/Sb69r+IyHse9buH2U+0n5EHjs6HOJ6P0g66l2NHM+Kc+x4R+cci8hHv/aa9jvFcE5F/ISGUciTNe7/nva/o378rIlnn3Iwcw/HU9rD5eeTj+abtqJMKx/U/CaegGxLCDJZEevqBzzwrIdF05YHXJ0Ukr3/PiMgb8jYloB6xnwv4+6+KyBf07ykRuan9ndS/p46qn/q5qxKSYu4oxhPXvCAHJx7/fUkmcr942OP5iP08JyLXROT9D7xeFpFR/P15EfnwEfbzlD1vCcbyjo7tI82Zw+qnvj8uIe5fPsrxfCv/nYR3Dmj+AMoI59yviMiXvfefEZH/TkRGROT/ckER4473/idE5EkR+Z+dcz0Jp6lP+AOEYg6pn/+Fc+4nRKQjYcJ+TL+75Zz7VQmcSCIiv+KTR9bD7qdISJD9ttcVpO3QxlNExDn3aQmIkhnn3JKI/LKIZPU+/pGI/K4EBM81EamJyM/pe4c2no/Yz78nItMi8j/p/Oz4wA45LyL/Ql/LiMj/6b3/f4+wnz8pIn/DOdcRkbqI/Iw+/0OlbXmEfooEp+kPvPfQ8jrc8Xyr7YSG4aSdtJN20r6L2klM/6SdtJN20r6L2onRP2kn7aSdtO+idmL0T9pJO2kn7buonRj9k3bSTtpJ+y5qJ0b/pJ20k3bSvovaidE/aSftpJ2076J2YvRP2kk7aSftu6j9/6cZCiX9IjLCAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "recon = reconstruct(hidden1_b, ftr_phase)\n",
    "import sounddevice as sd\n",
    "sd.play(recon,16000)\n",
    "Pxx, freqs, bins, im = plt.specgram(recon, NFFT=512, Fs=16000, noverlap=256)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "recon = reconstruct(ftr_raw, ftr_phase)\n",
    "import sounddevice as sd\n",
    "sd.play(recon,16000)\n",
    "from scipy import signal\n",
    "# frequencies, times, spectrogram = signal.spectrogram(recon, 16000)\n",
    "# plt.pcolormesh(times, frequencies, spectrogram)\n",
    "# plt.imshow(spectrogram)\n",
    "# plt.ylabel('Frequency [Hz]')\n",
    "# plt.xlabel('Time [sec]')\n",
    "# plt.pcolormesh\n",
    "# plt.show()\n",
    "Pxx, freqs, bins, im = plt.specgram(recon, NFFT=512, Fs=16000, noverlap=256)\n",
    "# The `specgram` method returns 4 objects. They are:\n",
    "# - Pxx: the periodogram\n",
    "# - freqs: the frequency vector\n",
    "# - bins: the centers of the time bins\n",
    "# - im: the matplotlib.image.AxesImage instance representing the data in the plot\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 97,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(771,)"
      ]
     },
     "execution_count": 97,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import h5py \n",
    "import tensorflow as tf\n",
    "hh = h5py.File('ftr_refrmd_10h.hdf5', 'r')\n",
    "d=hh['ftr_refrmd_10h'][0]\n",
    "len_data=d.shape\n",
    "hh.close()\n",
    "len_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 99,
   "metadata": {},
   "outputs": [],
   "source": [
    "def reconstruct(wave,angle):\n",
    "    recon = np.sqrt(np.power(10, wave))\n",
    "    recon1 = recon*np.cos(angle)+recon*np.sin(angle)*1j\n",
    "    recon = librosa.core.istft((recon1.T), hop_length=256, win_length=512, window='hann')\n",
    "    return recon"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 92,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<PrefetchDataset shapes: ((?, 257), (?,)), types: (tf.float32, tf.float32)>"
      ]
     },
     "execution_count": 92,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python (myenv)",
   "language": "python",
   "name": "myenv"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.4"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}