3D_conv_TOPO.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 146,
   "metadata": {},
   "outputs": [],
   "source": [
    "import scipy.io as sio\n",
    "import numpy as np\n",
    "from sklearn.model_selection import train_test_split\n",
    "from sklearn import preprocessing\n",
    "\n",
    "# make sure you don't hog all the video memory\n",
    "import tensorflow as tf\n",
    "config = tf.ConfigProto()\n",
    "config.gpu_options.allow_growth = True\n",
    "sess = tf.Session(config=config)\n",
    "from keras import backend as K\n",
    "K.set_session(sess)\n",
    "###################################\n",
    "\n",
    "from keras.layers import Input, Dense, Conv1D, Conv3D, MaxPooling1D, MaxPooling3D, SeparableConv1D, Dropout, BatchNormalization, Flatten, ELU, Activation\n",
    "from keras.models import Model\n",
    "from keras.models import load_model\n",
    "from keras.callbacks import TensorBoard, EarlyStopping\n",
    "from keras import optimizers\n",
    "from keras.initializers import Orthogonal as orth\n",
    "\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# file = 'EEG_big.mat'\n",
    "# file = 'EEG_big4CNN.mat'\n",
    "# file = 'EEG_big4CNN_1subj.mat'\n",
    "# file = 'KOS_100Hz_noICA->0.5-38Hz,Env=1.mat'\n",
    "# file = 'Merged123->0.5-38Hz,Env=1.mat'\n",
    "\n",
    "# file = 'KOS_100Hz_ICA->0.5-38Hz,Env=1.mat')\n",
    "# file = 'Merged123->1-48Hz,Env=1.mat'\n",
    "\n",
    "# file = 'BCI_TD.mat'\n",
    "\n",
    "# file = 'BCI_0101T_TD.mat'\n",
    "# file = 'BCI_0202T_TD.mat'\n",
    "# file = 'BCI_0303T_TD.mat'\n",
    "# file = 'BCI_0403T_TD.mat'\n",
    "# file = 'BCI_0503T_TD.mat'\n",
    "# file = 'BCI_0603T_TD.mat'\n",
    "# file = 'BCI_0703T_TD.mat'\n",
    "# file = 'BCI_0801T_TD.mat'\n",
    "\n",
    "# file = 'BCI_FD.mat'\n",
    "\n",
    "# file = 'ks.mat'\n",
    "# file = 'ks_FD.mat'\n",
    "# file = 'ks_2TD.mat'\n",
    "# file = 'ksenia2_long_chunked.mat'\n",
    "\n",
    "# file = 'BCI_FD_fullspectrum.mat'\n",
    "# file = 'BCI_FD_lowspectrum.mat'\n",
    "# file = 'KOS_100Hz_ICA->1-38Hz,Env=1_FD.mat'\n",
    "# file = 'Merged123->1-100Hz,Env=1_FD'\n",
    "# file = 'Merged123->1-47Hz,Env=1_FD.mat'\n",
    "# file = 'Merged123->15-47Hz,Env=1_TD.mat'\n",
    "# file = 'Merged123->15-47Hz,Env=1_TD5_64_.mat'\n",
    "# file = 'Merged123->75-134Hz,Env=1_TD.mat'\n",
    "# file = 'Merged123->2-30Hz,Env=1_TD5_64_.mat'\n",
    "\n",
    "# file = 'Merged456->2-30Hz,Env=1_TD1_94_.mat'                # 0.71\n",
    "# file = 'Merged456->2-30Hz,Env=1_TD101_192_.mat'             # 0.60\n",
    "# file = 'Merged456->2-30Hz,Env=1_TD197_289_.mat'             # 0.85\n",
    "\n",
    "# file = 'DAS_1_ica.mat, DS2=200Hz, FIR=1-60Hz, centnorm=1, Env=1, ICA, 1-124.mat' # 0.53 \n",
    "# file = 'Merged456_197-298_protocol_1.set, DS2=250Hz, FIR=1-47Hz, centnorm=0, Env=0, ICA, 1-95.mat'\n",
    "# file = 'Merged456_197-298_protocol_1.set, DS2=250Hz, FIR=1-47Hz, centnorm=0, Env=0, NOICA, 1-95.mat'\n",
    "\n",
    "\n",
    "\n",
    "# file = 'DAS_CH.mat, DS2=250Hz, FIR=1-30Hz, centnorm=1, Env=1, TD, 1-124.mat' # 0.65\n",
    "# file = 'DAS_CH.mat, DS2=250Hz, FIR=1-38Hz, centnorm=1, Env=1, TD, 1-124.mat'\n",
    "# file = 'DAS_CH.mat, DS2=250Hz, FIR=1-47Hz, centnorm=1, Env=1, TD, 1-124.mat'   # 0.60\n",
    "\n",
    "# file = 'Merged123_1_64_ICA(-eyes)AUDpreproc.mat, DS2=64Hz, FIR=1-9Hz, centnorm=1, Env=1, TD, 1-90.mat'\n",
    "# file = 'Merged123_1_64_ICA(-eyes)AUDpreproc.mat, DS2=250Hz, FIR=2-30Hz, centnorm=1, Env=1, TD, 1-90.mat' # 0.61\n",
    "# file = 'Merged123_1_64_ICA(-eyes)AUDpreproc.mat, DS2=250Hz, FIR=2-30Hz, centnorm=1, step=1, win=2, TD, 1-90.mat' #0.61\n",
    "\n",
    "# file = 'Merged456-197-289_ICA(-eyes)+AUDpreproc.mat, DS2=250Hz, FIR=2-30Hz, centnorm=1, step=2, win=2, TD, 1-93.mat' # 0.76\n",
    "# file = 'Merged456-197-289_ICA(-eyes)+AUDpreproc.mat, DS2=64Hz, FIR=2-30Hz, centnorm=1, step=2, win=2, TD, 1-93.mat'\n",
    "\n",
    "# file = 'Merged456-197-289_ICA(-eyes)+AUDpreproc.mat, DS2=64Hz, FIR=1-30Hz, centnorm=1, step=2, win=4, TD, 1-93.mat' #0.7\n",
    "# file = 'Merged456-197-289_ICA(-eyes)+AUDpreproc.mat, DS2=64Hz, FIR=1-30Hz, centnorm=1, step=2, win=2, TD, 1-93.mat' #0.60\n",
    "# file = 'Merged456-197-289_ICA(-eyes)+AUDpreproc.mat, DS2=64Hz, FIR=2-30Hz, centnorm=1, step=2, win=2, TD, 1-93.mat' #0.75(unstable)\n",
    "\n",
    "file = 'Merged456-197-289_ICA(ANDeyes)+AUDpreproc.mat, DS2=64Hz, FIR=2-30Hz, centnorm=1, step=2, win=2, TD, 1-93.mat' #0.86\n",
    "# file = 'Merged456-197-289_ICA(ANDeyes)+AUDpreproc.mat, DS2=250Hz, FIR=2-30Hz, centnorm=1, step=2, win=2, TD, 1-93.mat' #0.89\n",
    "\n",
    "# file = 'Merged456-1-94_ICA(-2,3ICs)+AUDpreproc.mat, DS2=64Hz, FIR=2-30Hz, centnorm=1, step=2, win=2, TD, 1-95.mat' #0.65\n",
    "# file = 'EEG_ICA(-123_ICs)+proc_AUD_101_192_Merged456.mat, DS2=64Hz, FIR=2-30Hz, centnorm=1, step=2, win=2, TD, 1-92.mat' #0.59\n",
    "\n",
    "# file = 'NeoRec_2018-05-08_21-47-10+ICA(-IC1)preproc_aud.mat, DS2=64Hz, FIR=1-40Hz, centnorm=1, Env=1, TD, 1-123.mat' # 0.63\n",
    "# file = 'NeoRec_2018-05-08_21-47-10+ICA.mat, DS2=64Hz, FIR=1-40Hz, centnorm=1, Env=1, TD, 1-123.mat'\n",
    "# file = 'NeoRec_2018-05-08_16-52-10_-eyes.mat, DS2=64Hz, FIR=1-40Hz, centnorm=1, Env=1, TD, 1-124.mat' \n",
    "# file = 'NeoRec_2018-05-08_16-52-10_-eyes.mat, DS2=128Hz, FIR=1-40Hz, centnorm=1, Env=1, TD, 1-124.mat'\n",
    "# file = 'NeoRec_2018-05-08_16-52-10.mat, DS2=64Hz, FIR=1-40Hz, centnorm=1, Env=1, TD, 1-124.mat'\n",
    "\n",
    "SHUFFLE = False\n",
    "TEST_TRAIN = 0.25\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Original data shape: (450, 128, 60)\n",
      "Original labels shape: (450, 1)\n",
      "Original data type: float32\n",
      "Normalized data type: float16\n",
      "test input shape (113, 128, 60) Nomralized MEAN: 0.00010717 min -4.77 max 4.902\n",
      "train input shape (337, 128, 60) Nomralized MEAN: 6.21e-05 min -10.18 max 8.86\n",
      "test labels shape (113, 2) Nomralized MEAN: 0.5 min 0.0 max 1.0\n",
      "train labels shape (337, 2) Nomralized MEAN: 0.5 min 0.0 max 1.0\n"
     ]
    }
   ],
   "source": [
    "# get the Dataset:\n",
    "path = '/home/amplifier/home/DATASETS/' + file\n",
    "mat_contents = sio.loadmat(path)\n",
    "X = mat_contents['X']\n",
    "Y = mat_contents['Z']\n",
    "\n",
    "if X.shape[1]<X.shape[2]:\n",
    "    X = np.transpose(X,[0,2,1])\n",
    "\n",
    "if Y.shape[1] > Y.shape[0]:\n",
    "    Y = Y.T\n",
    "    \n",
    "print('Original data shape:', X.shape)\n",
    "print('Original labels shape:', Y.shape)\n",
    "\n",
    "# verify that the model REALLY finds a mapping between the input and the labels. If we get\n",
    "# our accuracy by chance, then we should get the same accuracy on a permuted dataset:\n",
    "# Y = np.random.permutation(Y)\n",
    "\n",
    "\n",
    "# winsize = mat_contents['winsize']\n",
    "# stepsize = mat_contents['stepsize']\n",
    "# trial_len = mat_contents['trial_len']\n",
    "# low_cutoff = mat_contents['low_cutoff']\n",
    "# high_cutoff = mat_contents['high_cutoff']\n",
    "# source = mat_contents['filein']\n",
    "\n",
    "x_train, x_test, y_train, y_test = train_test_split(X, Y, test_size=TEST_TRAIN, shuffle=SHUFFLE)\n",
    "print('Original data type:', x_train.dtype)\n",
    "\n",
    "# convert to float64 for numerical stability:\n",
    "x_train = x_train.astype('float64')\n",
    "y_train = y_train.astype('float64')\n",
    "x_test = x_test.astype('float64')\n",
    "y_test = y_test.astype('float64')\n",
    "\n",
    "# one hot encode the labels:\n",
    "onehot_encoder = preprocessing.OneHotEncoder(sparse=False)\n",
    "y_train = onehot_encoder.fit_transform(y_train)\n",
    "y_test = onehot_encoder.fit_transform(y_test)\n",
    "\n",
    "# convert to float16 to save space:\n",
    "x_train = x_train.astype('float16')\n",
    "y_train = y_train.astype('float16')\n",
    "x_test = x_test.astype('float16')\n",
    "y_test = y_test.astype('float16')\n",
    "print('Normalized data type:', x_train.dtype)\n",
    "\n",
    "leng = X.shape[1] # if you work in the FD, this is the height of the sample time-frequency image, othewise EEG channels\n",
    "chan = X.shape[2] # if you work in the FD, this is the width of the sample time-frequency image, othewise time samples of EEG signal\n",
    "\n",
    "if len(X.shape)==3:\n",
    "    streams = 1 # this is EEG channels if you work with frequency domain, in the TD streams = 1\n",
    "if len(X.shape)==4:\n",
    "    streams = X.shape[3] # this is EEG channels if you work with frequency domain, in the TD streams = 1\n",
    "\n",
    "\n",
    "print('test input shape', x_test.shape, \"Nomralized MEAN:\", np.mean(x_test), \"min\", np.min(x_test),\"max\", np.max(x_test))\n",
    "print('train input shape', x_train.shape, \"Nomralized MEAN:\", np.mean(x_train), \"min\", np.min(x_train),\"max\", np.max(x_train))\n",
    "\n",
    "print('test labels shape', y_test.shape, \"Nomralized MEAN:\", np.mean(y_test), \"min\", np.min(y_test),\"max\", np.max(y_test))\n",
    "print('train labels shape', y_train.shape, \"Nomralized MEAN:\", np.mean(y_train), \"min\", np.min(y_train),\"max\", np.max(y_train))\n",
    "\n",
    "# print('Window length', winsize)\n",
    "# print('Step size:', stepsize)\n",
    "# print('Trial length:', trial_len)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 91,
   "metadata": {},
   "outputs": [],
   "source": [
    "path = '/home/amplifier/home/DATASETS/'\n",
    "# file = 'cm.mat'\n",
    "file = 'cm_alt.mat'\n",
    "\n",
    "mat_contents = sio.loadmat(path+file)\n",
    "cm = mat_contents['cm']\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 169,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(337, 6, 7, 128, 1)\n",
      "(113, 6, 7, 128, 1)\n"
     ]
    }
   ],
   "source": [
    "x_train_rs = np.zeros([x_train.shape[0],6,7,x_train.shape[1]])\n",
    "tst = np.zeros([6,7])\n",
    "for k in range(x_train.shape[0]):\n",
    "    for i in range(2,8):\n",
    "        for j in range(2,9):\n",
    "            if (cm[i,j] != 0) & (cm[i,j]<=x_train.shape[2]):\n",
    "                tst[i-2,j-2] = cm[i,j]\n",
    "#                 print('RIGHT', cm[i,j])\n",
    "                x_train_rs[k, i-2,j-2,:] = x_train[k,:,cm[i,j]-1]\n",
    "#                 print('Trial', k,':', 'Moving EEG chan', cm[i,j], '(', cm[i,j]-1, 'x_train in matrix)', 'to', i-1, 'th', j-2, 'th matrix x_train_rs')\n",
    "#             else:\n",
    "#                 print('NO', cm[i,j])\n",
    "\n",
    "\n",
    "x_train_rs = x_train_rs[:,:,:,:,None]\n",
    "print(x_train_rs.shape)\n",
    "\n",
    "x_test_rs = np.zeros([x_test.shape[0], 6,7,x_test.shape[1]])\n",
    "tst = np.zeros([6,7])\n",
    "for k in range(x_test.shape[0]):\n",
    "    for i in range(2,8):\n",
    "        for j in range(2,9):\n",
    "            if (cm[i,j] != 0) & (cm[i,j]<=x_test.shape[2]):\n",
    "                tst[i-2,j-2] = cm[i,j]\n",
    "#                 print('RIGHT', cm[i,j])\n",
    "                x_test_rs[k, i-2,j-2,:] = x_test[k,:,cm[i,j]-1]\n",
    "#                 print('Trial', k,':', 'Moving EEG chan', cm[i,j], '(', cm[i,j]-1, 'x_train in matrix)', 'to', i-1, 'th', j-2, 'th matrix x_train_rs')\n",
    "#             else:\n",
    "#                 print('NO', cm[i,j])\n",
    "\n",
    "\n",
    "\n",
    "x_test_rs = x_test_rs[:,:,:,:,None]\n",
    "\n",
    "print(x_test_rs.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 170,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "Text(0.5,1,'17')"
      ]
     },
     "execution_count": 170,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline\n",
    "\n",
    "t = np.linspace(-100, 500, 128)\n",
    "fig, axs = plt.subplots(1,2,figsize=(15,5))\n",
    "plt.subplot(121)\n",
    "tmp = np.mean(x_train_rs, axis=0)\n",
    "plt.plot(t, tmp[2,3,:])\n",
    "plt.title(tst[2,3].astype('int8'))\n",
    "\n",
    "plt.subplot(122)\n",
    "plt.plot(t, np.mean(x_train[:,:,16], axis=0))\n",
    "plt.title('17')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 173,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0646cc369dec4a6eb428ca62fc2c06b7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "interactive(children=(IntSlider(value=63, description='f', max=127), Output()), _dom_classes=('widget-interact…"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "from ipywidgets import interact, interactive, fixed, interact_manual\n",
    "import ipywidgets as widgets\n",
    "\n",
    "def pltsin(f):\n",
    "    t = np.linspace(-100, 500, 128)\n",
    "    fig, axs = plt.subplots(1,2,figsize=(15,5))\n",
    "    plt.subplot(121)\n",
    "    plt.imshow(tmp[:,:,f,-1], vmin=-0.2, vmax=0.2)\n",
    "    plt.subplot(122)\n",
    "    plt.plot(t, tmp[2,3,:,-1])\n",
    "    plt.axvline(x=t[f])\n",
    "    plt.axhline(y=0)\n",
    "    plt.show\n",
    "    \n",
    "x = interact(pltsin, f=(0,127,1))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 174,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(337, 6, 7, 128, 1)\n"
     ]
    }
   ],
   "source": [
    "print(x_train_rs.shape)\n",
    "height, width, depth = x_train_rs.shape[1],x_train_rs.shape[2],x_train_rs.shape[3]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": 219,
   "metadata": {},
   "outputs": [],
   "source": [
    "def activation():\n",
    "    return ELU(alpha=1.0) # Activation('relu') #\n",
    "\n",
    "def conv_net(height, width, depth):\n",
    "    input_img = Input(shape=(height, width, depth, 1))  # adapt this if using `channels_first` image\n",
    "    \n",
    "    block1 = Conv3D(50, (3,3,50), padding='same', activation=None)(input_img)\n",
    "    block1 = activation()(block1)\n",
    "    block1 = MaxPooling3D(pool_size=(2,2,2), strides=None, padding='same')(block1)\n",
    "    block1 = BatchNormalization()(block1)\n",
    "    block1 = Dropout(0.1)(block1)\n",
    "    \n",
    "    block2 = Conv3D(100, (2,2,20), padding='same', activation=None)(block1)\n",
    "    block2 = activation()(block2)\n",
    "    block2 = MaxPooling3D(pool_size=(2,2,2), strides=None, padding='same')(block2)\n",
    "    block2 = BatchNormalization()(block2)\n",
    "    block2 = Dropout(0.2)(block2)\n",
    "    \n",
    "    block6 = Flatten()(block2)\n",
    "    block6 = Dense(20, activation='sigmoid')(block6)\n",
    "    block6 = BatchNormalization()(block6)\n",
    "    block6 = Dropout(0.5)(block6)\n",
    "\n",
    "    block7 = Dense(2, activation='softmax')(block6)\n",
    "\n",
    "    return Model(input_img, block7)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 220,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "input_31 (InputLayer)        (None, 6, 7, 128, 1)      0         \n",
      "_________________________________________________________________\n",
      "conv3d_21 (Conv3D)           (None, 6, 7, 128, 50)     22550     \n",
      "_________________________________________________________________\n",
      "elu_21 (ELU)                 (None, 6, 7, 128, 50)     0         \n",
      "_________________________________________________________________\n",
      "max_pooling3d_20 (MaxPooling (None, 3, 4, 64, 50)      0         \n",
      "_________________________________________________________________\n",
      "batch_normalization_32 (Batc (None, 3, 4, 64, 50)      200       \n",
      "_________________________________________________________________\n",
      "dropout_34 (Dropout)         (None, 3, 4, 64, 50)      0         \n",
      "_________________________________________________________________\n",
      "conv3d_22 (Conv3D)           (None, 3, 4, 64, 100)     400100    \n",
      "_________________________________________________________________\n",
      "elu_22 (ELU)                 (None, 3, 4, 64, 100)     0         \n",
      "_________________________________________________________________\n",
      "max_pooling3d_21 (MaxPooling (None, 2, 2, 32, 100)     0         \n",
      "_________________________________________________________________\n",
      "batch_normalization_33 (Batc (None, 2, 2, 32, 100)     400       \n",
      "_________________________________________________________________\n",
      "dropout_35 (Dropout)         (None, 2, 2, 32, 100)     0         \n",
      "_________________________________________________________________\n",
      "flatten_13 (Flatten)         (None, 12800)             0         \n",
      "_________________________________________________________________\n",
      "dense_25 (Dense)             (None, 20)                256020    \n",
      "_________________________________________________________________\n",
      "batch_normalization_34 (Batc (None, 20)                80        \n",
      "_________________________________________________________________\n",
      "dropout_36 (Dropout)         (None, 20)                0         \n",
      "_________________________________________________________________\n",
      "dense_26 (Dense)             (None, 2)                 42        \n",
      "=================================================================\n",
      "Total params: 679,392\n",
      "Trainable params: 679,052\n",
      "Non-trainable params: 340\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "# define input to the model:\n",
    "x = Input(shape=(height, width, depth, 1))\n",
    "\n",
    "# make the model:\n",
    "model = conv_net(height, width, depth)\n",
    "\n",
    "# compile the model:\n",
    "# sgd = optimizers.SGD(lr=0.01, decay=1e-4, momentum=0.95, nesterov=True)\n",
    "# model.compile(loss='binary_crossentropy', optimizer=sgd, metrics=['accuracy'])\n",
    "# model.compile(optimizer='rmsprop', loss=correlation_coefficient_loss)\n",
    "model.compile(loss='binary_crossentropy', optimizer=optimizers.Adam(0.00001), metrics=[\"accuracy\"])\n",
    "\n",
    "\n",
    "# SEE INSIDE THE MODEL:\n",
    "# conv_net().summary() #or\n",
    "model.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "init_lr = 0.000025\n",
    "lr_drop_by = 0.995\n",
    "drop_every = 5\n",
    "BATCH_SIZE = 10\n",
    "\n",
    "from keras.callbacks import TensorBoard, EarlyStopping, LearningRateScheduler, Callback, ModelCheckpoint\n",
    "\n",
    "early_stopping = EarlyStopping(monitor='val_loss', patience=100, mode='min')\n",
    "\n",
    "K.set_value(model.optimizer.lr, init_lr)\n",
    "\n",
    "def step_decay(epoch, init_lr=init_lr, lr_drop=lr_drop_by, drop_every=drop_every):\n",
    "    if epoch%drop_every==0:\n",
    "        lrate = init_lr * (lr_drop_by ** np.floor((1+epoch)/drop_every))\n",
    "    else:\n",
    "        lrate = K.get_value(model.optimizer.lr)\n",
    "    print('CHECK', epoch%drop_every)\n",
    "    return lrate\n",
    "\n",
    "class LossHistory(Callback):\n",
    "    def on_train_begin(self, logs={}):\n",
    "        self.losses = []\n",
    "        self.lr = []\n",
    "        self.batch_loss = []\n",
    "        self.val_acc = []\n",
    " \n",
    "    def on_epoch_end(self, epoch, batch, logs={}):\n",
    "        lrate = step_decay(epoch)\n",
    "        txt = K.get_value(model.optimizer.lr)\n",
    "        self.lr.append(txt)\n",
    "#         self.lr.append(lrate)\n",
    "#         txt = K.eval(self.model.optimizer.lr)\n",
    "        print('OPTIMIZERS lrate AT EPOCH END = ', txt, '\\n\\n')\n",
    "    \n",
    "    def on_epoch_begin(self, epoch, batch, logs={}):\n",
    "        txt = K.get_value(model.optimizer.lr)\n",
    "        print('OPTIMIZERS lrate AT EPOCH START = ', txt)\n",
    "\n",
    "    def on_batch_end(self, batch, logs={}):\n",
    "        self.batch_loss.append(logs.get('loss'))\n",
    "\n",
    "lr_decay = LearningRateScheduler(schedule=step_decay)\n",
    "loss_history = LossHistory()\n",
    "\n",
    "checkpointer = ModelCheckpoint(filepath='/home/amplifier/home/NEW_DL/weights/3D_conv_TOPO_wts.h5',\n",
    "                               verbose=1,\n",
    "                               monitor='val_acc',\n",
    "                               save_best_only=True)\n",
    "\n",
    "\n",
    "train_history = model.fit(x_train_rs, y_train,\n",
    "                epochs=1000,\n",
    "                batch_size=BATCH_SIZE,\n",
    "                verbose=2,\n",
    "                shuffle=True,\n",
    "                validation_data=(x_test_rs, y_test),\n",
    "                callbacks=[checkpointer, early_stopping, lr_decay, loss_history])\n",
    "\n",
    "# save the entire model and the weights after the training:\n",
    "# model.save('/home/amplifier/home/NEW_DL/models/EEGnet_CNN_Dense_Binary_Classifier_TD_1D.h5')\n",
    "# model.save_weights('/home/amplifier/home/NEW_DL/weights/EEGnet_CNN_Dense_Binary_Classifier_TD_1D_wts.h5')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 222,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(337, 2)\n",
      "(113, 2)\n"
     ]
    }
   ],
   "source": [
    "model.load_weights('/home/amplifier/home/NEW_DL/weights/3D_conv_TOPO_wts.h5')\n",
    "# let's predict a TRAINING sammple:\n",
    "decoded_train = model.predict(x_train_rs)\n",
    "decoded_test = model.predict(x_test_rs)\n",
    "print(decoded_train.shape)\n",
    "print(decoded_test.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 223,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABDAAAAFgCAYAAABNIolGAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJzs3Xl4VNX5wPHvSxLWsCbsSwKCshOQxRUsLriCuyCgoJVaW7eqVYttXUrr0l+1i9ZqpWqNBXepa1WwiGWRJQiyiUAgBIGEnRiW5Pz+eM84k8lMNrLn/TzPPDNz77n3nplkzr3zzjnvEeccxhhjjDHGGGOMMdVZvaqugDHGGGOMMcYYY0xxLIBhjDHGGGOMMcaYas8CGMYYY4wxxhhjjKn2LIBhjDHGGGOMMcaYas8CGMYYY4wxxhhjjKn2LIBhjDHGGGOMMcaYas8CGKbGEZFkEXEiElvVdQklIu+LyLVVXQ9jTPVWm9sKEblfRF6q6nqEEpEuInJARGKqui41kYg8LyK/qep6GGPKRkQ+FZEfVnU9QonIeBH5T1XXo6YSkU0iclZV16OqWADDlFhd/7AUxzl3nnPuhaquhzEmsurShlWntqI6XtiWN+fcZudcvHMur6rrYoypWr7N2y0iDaq6LnWZcy7VOXdOVdfD1EwWwDDGGGPKSXXqGVad6mJKz/5+xpQvEUkGTgccMLqSj22f51rA/o7VgwUwTLkQkRtEZL2I7BKRWSLSwS8XEXlcRHaIyF4R+VJE+vp154vIKhHZLyJbReTOKPuOEZHfi0iWiGwALghZd4WILAkrf4eIvOUfPy8iT4rIu/44C0XkuJCyfxSRLSKyT0SWiMjpIevuF5FXReQlv+0KETleRO71r2eLiJwTUr7AL5n+PVntt10lIoP88rv9690vImtF5MxjfPuNMcdIRC4UkTQR2SMi/xOR/iHr7hGRb0I+y5eErJskIp/7dm4XcL9fNs+3W7tFZKOInBeyzfdtRQnKdhWRuf7YH/v2LOIQDRE5Q0QyfBvzLfAPEWkpIu+IyE6//3dEpJMvPw29mP+L6BCLv/jlPUXkI9+erxWRK4t437qKyH99/T4CEkPWvSsiN4eV/1JELvaPnYjcKCJf+7o9KSLi1x0nIrNFJNu3/aki0iJkP5tE5C6/v4Mi8pyItBUdnhN4r1r6sgWGHYpIKxH5h4hk+uMGzheJ/v3Z41/7ZyIS8TpJRE4RkS9Ez2tfiMgpfvlYEVkcVvZ2EZnlHzfwf+vNIrJdRJ4WkUbR/n5Rjn2d6Lllt4h8KCJJIeuciNwiIhv8+/ZY4DWISD0RuU9E0kXPYS+KSPOQbU8T/d/fI3p+mxRy2JYS4TwqKuI53phq6BpgAfA8UGAYn4g0EpH/85+PvaLtcuCzGfGzIYWv+yaJyLyQ505EfiIiXwNf+2VFXXfGiMgvJHi+WSIinX3b+H9h9f23iNwW6UWKyNkissa/jr8AgXa1gW/b+oWUbSMi34lI65A26A7/md4mIpNDyl4gIst83beIyP0h6wLt7GS/brdo+z7Etwt7fF2ivVd9JHje2S4iv/DLh4rIYn/M7SLyh2h/XIn+XeRpEfl9WNm3ReRn/nEHEXld9Dy5UURuCSl3v4i8JvpdYB8wKcJxS9Ku/0K0Td4kIuNDtm0u2hbv9P9790nIeUeifJfwUvx7u1dEZopIQ79Nic9lNZZzzm52K9EN2AScFWH5SCALGAQ0AP4MzPXrRgFLgBZoA9oLaO/XbQNO949bAoOiHPdGYA3QGWgFzEGj57H+eLuAXiHllwGX+cfP+/VDfflUYEZI2QlAgl93B/At0NCvux/I9a8hFngR2AhMBeKAG4CNIfv6FPihf3wFsBUY4l93dyAJOAHYAnTw5ZKB46r6b2s3u9WFWxFt2CBgBzAMiEEvbDcBDfz6K4AOaND/KuBgSDs2CTgK3OzbiUZ+2RHfRsQAPwYyAfHbhLYVxZWdD/weqA+cBuwDXory+s7wdXnEt42NfPt2GdAYaAq8CrwVss33dfHPm/g2arJ/PYPQ9r1PlGPOB/7gjzcc2B+oH3AlsDCk7AAgG6jvnzvgHfT80AXYCZzr13UHzvb7bQ3MBZ4I+1suANoCHf3fbykw0G8zG/i1L5vsjxXrn78LzETPO3HACL/8d8DTflkcGtyRCK+5FbAbmOjfo3H+eYJ/n/cDPULKfwGM9Y+fAGb5fTQF/g38LtrfL8KxLwbWo+fSWOA+4H8h6x16jmzl39N1BP/XrvPbdgPigTeAf/p1XXy9x/nXngCk+HXPE+U8ShHneLvZrbrd/P//TcCJaLvbNmTdk2h72BFti0/xn8OiPhufUrD9nATMC3nugI/857GRX1bUdeddwAr0WlHQNjPBf/YygXq+XCKQE1r/kGMmoueJy319b/ftSqAdeAp4JKT8rcC//eNAG/Sg3/Z8f5yWIev7oefC/sB24GK/Ltm/3qeBhsA56DX0W0Abgu30iPD3Cm0Lt/n3o6F/Psyvmw9M9I/jgZOi/G2L+i4yHD2vBc6rLYHvCJ7XlwC/Qs+z3YANwChf9n70f+ViXzZSu1ySdj1wnhyBXkOc4Ne/CLztt0tG2+zr/bqI3yX8uk3AIv8aWgGrgRv9uhKdy2ryrcorYLeacyP6xf9zwKMhz+P9hz3ZNyjrgJPwDW9Iuc3Aj4BmxRx3duBD6Z+fQ8GL0b8C0/zjPuiFZOCLx/PA30O2PR9YU8SxdgMD/OP7gY9C1l0EHABi/POmvh4t/PNPCZ4gPgRujbD/7mgDfhYQV9V/U7vZrS7dimjD/go8FLZsLf5CK0L5NGCMfzwJ2By2fhKwPuR5Y99WtPPPQ9uKqGXRC+ejQOOQ9S9RdADjMP5iOEqZFGB3yPPv6+KfXwV8FrbN3/DBgLDlgfo1CVn2MsEARiDA3MM//z3wVEhZB5wW8vwV4J4o9b4YWBb2txwf8vx14K8hz2/GB2oICWAA7YF8/AV52DEeRC8kuxfzfzQRWBS2bD4wKeRv9Cv/uAf65acxegF6kJCgNXAyPhBewr/f+/iLW/+8HvoFIynkPT03ZP1NwCf+8SfATSHrTkDP1bHAvcCbUY75PFHOoxRxjreb3arTDQ0AHwES/fM1wO3+cT30C+2ACNsV9dkIbz8nUTiAMbKYeoVed67Fn1silFsNnO0f/xR4L0q5a4AFIc8FyCB4zhmGfpkPBEMWA1f6x2f49yE2ZPsdRA8aPAE87h8n+9fbMWR9NnBVyPPXgdvC3ys0OLQsyjHmAg8E/m5FvI9FfRcR9DvHcL/uBmB2yPsRfg6/F/iHf3w/PhAS5bgladfDz5OvAL9EA2WHgN4h634EfOofR/wu4ddtAiaEPH8UeNo/LtG5rCbfald3ElNVOgDpgSfOuQNoo9XROTcb+Asa2d4uIs+ISDNf9DL0QihdtAvyyUXsf0vI8/Sw9S8AV4uIoBeWrzjnDoWs/zbkcQ7aqAHfDzdZ7btf7QGaE9IFGo0uB3wHZLlgIrjv/H08hXUGvglf6JxbD9yGNog7RGRGoIubMabKJAF3+O6We3xb0BltexCRayQ4vGQP0JeC7cSWwrsMtjvOuRz/MFJbUVTZDsCukGXRjhVqp3MuN/BERBqLyN9819R96MVgC4k+I0cSMCzsvRiPBlTCdUCDIQdDloWeCw6hF2oTfPfVccA/w/YRsX323ZpniA6324cGBRLDtg1vn8OfR2ubdznndkdY9xj6C+1/RIdg3BOhDISd87x09BdG0CDOOP/4ajSQkoP2JGkMLAl5bz/wywMK/P0iSAL+GLL9LvQCumNImfDzZeAcE17vdDR40ZYo56wQEf9OxZzjjalOrgX+45zL8s9fJjiMJBH95T/SZ6C4z0ZxCrTZxVx3FnWsF9DeG/j78LY0oMA1s9NvtKHPF6JfuEeISE/0h7VZIdtnO+eOhjwPbZeHicgcP9xhL9pDurza5Wiv+3rgeGCN6HC9C6OUK+q7iANmULBdTvWPk4AOYee8X6DtYkBR592StOuRzpMd0PeuPoXb5UB7XqZ2mZKfy2osC2CY8pCJNgAAiEgTtMvbVgDn3J+ccyeivSOOR7vI4Zz7wjk3Bu1a9hZ6oRvJNvRDHNAldKVzbgH6q9XpaKMUrVEvQHTc4d1oN+eWzrkWwF78WMFjtAU4LtIK59zLzrnT0PfMod2FjTFVZwvai6tFyK2xc+5fovkFnkV/8Urw7cRKCrYTroLqtQ1oJSKNQ5Z1jlY4Sl3uQH9pH+aca4Z2pYVg/cPLbwH+G/ZexDvnfhylfi19mx/QJazMC2gA5Ewgxzk3v5j6B/zO162/r/cEyq9tbiUh+TQCnHP7nXN3OOe6oT3ufiaRcxQVOOd5XfDnPOA/QKKIpKAXzC/75VnoBXyfkPe2uXMu9IK+uP+lLcCPwv4+jZxz/wspE36+zIxS70APmu0Ucc4qTrRzvDHVhc9HcCX6pf1b0RwztwMDRGQA+tnMJfJnoKjPxkH0y2tApEDv95/pElx3FnWsl4Axvr690OvmSApcM/sf98LPG4FgyETgtWKCpqFeRoMdnZ1zzdFhChV9zfy1c24c+l3hEeC1sHNOQJHfRYB/AZf7c/owtDdI4Ngbw9rUps6580OrUUTdS9KuRzpPZvptj1C4XQ7UuUztcinOZTWWBTBMacWJSMOQWyzaoE0WkRTRaal+i4573iSavGeYiMShDX0ukCci9UXngG7unDuCjteLNsXdK8AtItJJNClbpEjii+ivQEedc/MirI+kKXrxthOIFZFfAeX1y9HfgTtF5ERR3UUkSUROEJGR/n3KRRs9m9rPmMoTqQ17FrjRt1UiIk1Ek5U1RXNCOLSdQDShWaUkKXTOpaPde+/3bebJ6MVIaTRF25k9ItIK+HXY+u3omN+Ad4DjRWSiiMT52xAR6VVE/R7w9TstvH4+YJEP/B8lDC6H1PuAr3dHyulLsXNuGzoM4ynRBKdxIjIcvk/k2t1f8AfOSZHa5/fQ9+hqEYkVkauA3uh7h//18jX0V7BW6Bh4nHP56P/a4yLSxh+zo4iMKsVLeBq4V0T6+O2bi8gVYWXu8q+tMzq+faZf/i/gdtHEq/HouXqmr28qcJaIXOlfU4IPwBQp2jm+FK/HmMpwMfp/2RsdRpeCBgE+A67xn83pwB9EEzrGiMjJ/lqtqM9GGnCpaE+37mhvgaIUd935d+AhEenhz0X9RSQBwDmXgebT+SfwunPuOyJ7F+gjIpf689stFA6s/BO4BA1ivFhMncPrv8s5lysiQ9EfDcvDO0A7EblNNCFmUxEZBiAiE0Sktf8b7fHlI7UxUb+LADjnlqHv+9+BD51zgX0tAvaJJk9u5P/2fUVkSEkqXop2PXCePB24EHjV9+h+BZjmX3MS8DM0WAVRvksUV6dSnMtqLAtgmNJ6D70YDtzud859go7leh2N/B4HjPXlm6Ef7N1ot6hsdBw0aOR3k2j34BsJdo0L9yw6Dmw5mqTtjQhl/ol+qSjNBfKH6IXsOl+3XIrvnl0izrlXgWlog7ofjZS3QseEP4xGXb9FI8q/KI9jGmNKJFIbthgdE/sXtK1aj8807pxbhX75no9+2e8HfF6J9R2PjqfNBn6Dfhk9VOQWBT2BJvPMQpNefhC2/o/or1K7ReRPzrn9aJ6hsegvRN8STCoZydXor1m70OBIpIvhF9H3LeLsKVE8gCZj24tekEdq98tqIvqr1xp0fHcgk38P4GM0cDIfzdfxafjGzrls9AL0DvTv8nPgwpCu6aBt/1noRWpod+y70f+vBf7c9zHaQ6ZEnHNvon+PGX77lcB5YcXeRpPSpaHv3XN++XT0HDkXTUidi+YKwTm3GR3SeQf6t0xDEwgWp6hzvDHVxbVoToPNzrlvAze0zR/vv+jfiSbQ/AL9DDyC5oko6rPxONoDeDvaqyGVohV33fkH9Avtf9Avns+h7XfAC2hbGvVa17dDV6DXmtlou/Z5WJkM9HraoUGckroJeFBE9qNJL6P1nC4Vf945Gw2Af4vO2PIDv/pc4CsROYCer8ZG6jFSzHeRgH+h7fLLIdvl+eOmoO1iFho4aE7JFdeuf4u2kZno/8iNzrk1ft3NaPB3AzDP1226r1u07xLFKdG5rCYLZGM1pkYT7R64A53J5Ouqro8xxlQEEZmJJlAM70lRbYnINcAUP3TOVCARcWjS1PVVXRdjTPnyvcVeApL9L//Hsq/pQKZz7r5yqZyJSETOQBNbd6rqutQmsVVdAWPKyY+BLyx4YYypTXw31l3oL0PnAGPQX9ZqBNH8HTehU/cZY4wpAz9M61Z0RqBjDV4kA5ei004bU+PYEBJT44nIJrRRv6OKq2KMMeWtHTpV3wHgT8CP/Vjeas+PAd6Jdq9+uZjixhhjIvA5iPag00A/cYz7eggdevaYc25jOVTPmEpnQ0iMMcYYY4wxxhhT7VkPDGOMMcYYY4wxxlR7NS4HRmJioktOTq7qahhjaoklS5ZkOedaV3U96hpry40x5cna8spn7bgxpjyVtB2vcQGM5ORkFi9eXNXVMMbUEiKSXtV1qIusLTfGlCdryyuftePGmPJU0nbchpAYY4wxxhhjjDGm2rMAhjHGGGOMMcYYY6o9C2AYY4wxxhhjjDGm2qtxOTCMqQ6OHDlCRkYGubm5VV0VU0INGzakU6dOxMXFVXVVjDHGGGOMMWVgAQxjyiAjI4OmTZuSnJyMiFR1dUwxnHNkZ2eTkZFB165dq7o6xhhjjDHGmDKwISTGlEFubi4JCQkWvKghRISEhATrMWOMMcYYY0wNZgEMY8rIghc1i/29jDHGGGOMqdksgGGMMcYYY0wtJiLTRWSHiKyMsl5E5E8isl5EvhSRQZVdR2OMKQkLYBhTQ8XHx1fYvp944glefPFFAM444wwWL15cpv3cf//9/P73vy9x+T179vDUU08VWy5anVasWMGkSZNKU8VqRUTOFZG1/gLyngjrG4jITL9+oYgkh6y71y9fKyKjituniHT1+/ja77O+X/4zEVnlL2A/EZGkkG3yRCTN32YVty9jjDHVxvPAuUWsPw/o4W9TgL9WQp2MMabULIBhjCng6NGjTJ8+nauvvrrSj13SAEY0/fr1IyMjg82bN5djrSqHiMQAT6IXkb2BcSLSO6zY9cBu51x34HHgEb9tb2As0Ae9QH1KRGKK2ecjwOPOuR7Abr9vgGXAYOdcf+A14NGQ43/nnEvxt9Ehy6Pt65ilpkJyMtSrB4mJehOB2NjI98nJuo0xxpgg59xcYFcRRcYALzq1AGghIu0rp3bGGFNyFRbAEJGGIrJIRJaLyFci8kCEMlF/TawozrmKPoQxlco5x1133UXfvn3p168fM2fOBGDbtm0MHz6clJQU+vbty2effUZeXh6TJk36vuzjjz9eaH+zZ89m0KBBxMYGJyl66aWXOOWUU+jbty+LFi0iPz+fHj16sHPnTgDy8/Pp3r07WVlZhfa3fPlyRo4cSY8ePXj22WcBOHDgAGeeeSaDBg2iX79+vP322wDcc889fPPNN6SkpHDXXXcB8Oijj9KvXz8GDBjAPfcEOyW8+uqrDB06lOOPP57PPvvs++UXXXQRM2bMONa3tSoMBdY75zY45w4DM9ALylBjgBf849eAM0WTe4wBZjjnDjnnNgLr/f4i7tNvM9LvA7/PiwGcc3Occzl++QKgU1GVLmpfxyo1FaZMgfR0cA6ys/UGkJcX+T49HSZOLDrIYcEOY4wppCOwJeR5hl9WgIhMEZHFIrI4cA1gjDkG+fmQn1fVtahRKnIa1UPASOfcARGJA+aJyPs+qhvw/a+JIjIW/RXvqoqq0Ofrsxj/94W8/ZNTGdC5RUUdxtQxD/z7K1Zl7ivXffbu0IxfX9SnRGXfeOMN0tLSWL58OVlZWQwZMoThw4fz8ssvM2rUKKZOnUpeXh45OTmkpaWxdetWVq7UIbB79uwptL/PP/+cE088scCygwcP8r///Y+5c+dy3XXXsXLlSiZMmEBqaiq33XYbH3/8MQMGDCAxMbHQ/r788ksWLFjAwYMHGThwIBdccAFt2rThzTffpFmzZmRlZXHSSScxevRoHn74YVauXElaWhoA77//Pm+99RYLFy6kcePG7NoV/PHo6NGjLFq0iPfee48HHniAjz/+GIDBgwfz8MMP8/Of/7xkb3b1EenicVi0Ms65oyKyF0jwyxeEbRu48Iy0zwRgj3PuaITyoa4H3g953lBEFgNHgYedc2+VYl+lNnUq5OQUXy5cIE4dLcgRGuyYMkUfjx9f9noaY0wtECnTdaFf/ZxzzwDPAAwePNh+FTTmWP3nPvj2S5j0TlXXpMaosB4YvgvaAf80zt/CG7povyZWiDlrdgCwaGNRPeiMqVnmzZvHuHHjiImJoW3btowYMYIvvviCIUOG8I9//IP777+fFStW0LRpU7p168aGDRu4+eab+eCDD2jWrFmh/W3bto3WrVsXWDZu3DgAhg8fzr59+9izZw/XXXfd93kypk+fzuTJkyPWb8yYMTRq1IjExER+8IMfsGjRIpxz/OIXv6B///6cddZZbN26le3btxfa9uOPP2by5Mk0btwYgFatWn2/7tJLLwXgxBNPZNOmTd8vb9OmDZmZmaV4B6uNklw8RitTXsuDBxKZAAwGHgtZ3MU5Nxi4GnhCRI4rYb0D+yzVL3eVMRIoJ0cDJcYYU8dlAJ1DnncCauTJ1JgKtWM1ZJQtN1xEmctg65Lgry/V1aH9kPYv7TFSxSqyB0ZgTPcSoDvwpHNuYViRaL8mZoXtZwqaUIguXbpUZJWNKbWS9pSoKNGGRQ0fPpy5c+fy7rvvMnHiRO666y6uueYali9fzocffsiTTz7JK6+8wvTp0wts16hRI3JzcwssC48rigidO3embdu2zJ49m4ULF5IapS9+pG1TU1PZuXMnS5YsIS4ujuTk5ELHDLy2aDHNBg0aABATE8PRo0e/X56bm0ujRo0iblPNleTiMVAmQ0RigebomOaito20PAsd3xzre04UOJaInAVMBUY45w4FljvnMv39BhH5FBgIvF7UvkKV9pe7Ll20l0RFq4EpU4wxprzNAn4qIjPQnnp7nXPbqrhOxlQ/7/8cDmbBTfPLZ397M+BIDuRkQ5PCPZmrjYVPw+zfQP0m0Ht08eUrUIUm8XTO5TnnUtAL2qEi0jesSIm7qznnBjvnBof/MmxMXTd8+HBmzpxJXl4eO3fuZO7cuQwdOpT09HTatGnDDTfcwPXXX8/SpUvJysoiPz+fyy67jIceeoilS5cW2l+vXr1Yv359gWWBvBrz5s2jefPmNG/eHIAf/vCHTJgwgSuvvJKYmJiI9Xv77bfJzc0lOzubTz/9lCFDhrB3717atGlDXFwcc+bMId1/S23atCn79+//fttzzjmH6dOnk+PHEYQOIYlm3bp19O0b3tTUCF8APfyMHvXRpJyzwsrMAq71jy8HZjuNYM0Cxvq8Ql3RLPKLou3TbzPH7wO/z7cBRGQg8DdgtHNuR+DAItJSRBr4x4nAqcCqovZ1rKZNA9/5pkI5Fz1PRiBxaL16ljPDGFNzici/gPnACSKSISLXi8iNInKjL/IesAHNofQscFMVVdWYyvfRr2HB0yUru30V7C+n2F5+Huz3v/nsqYRfbErjg3th3hPB5yvf0Pv//alwb5H5T+l7WEkqtAdGgHNuj/+17lwgdP7paL8mGmNK6JJLLmH+/PkMGDAAEeHRRx+lXbt2vPDCCzz22GPExcURHx/Piy++yNatW5k8eTL5vvvX7373u0L7O++885g4cWKBZS1btuSUU05h3759BXpsjB49msmTJ0cdPgIwdOhQLrjgAjZv3swvf/lLOnTowPjx47nooosYPHgwKSkp9OzZE4CEhAROPfVU+vbty3nnncdjjz1GWloagwcPpn79+px//vn89re/LfL9mDNnDhdccEGJ37/qwvdC+ynwIRADTHfOfSUiDwKLnXOzgOeAf4rIerStHOu3/UpEXgFWofkpfuKcywOItE9/yLuBGSLyG3Tmkef88seAeOBV3/tls59xpBfwNxHJR4PfDzvnVhWzr2MSyEsxdar2kgiMIMrOhpgYzWURfi9Stl6Y0fJkBJKGguXMMMbUXM65ccWsd8BPKqk6xlQf+fnwxd+hUSsY9iO9kIjmwE7I8QMFjh6G2GOcNf7ADsj3vYj3bIGOBXPQsW+b9sqIiSv5Pg8dgO3+63Z8W2jVNbhu/7ewe5M+btMLGjaPvI8dq2HBUxDbCAZOhIM7YMcqaNcfMr6AzQsg6WQtm3cEPvu9vh9n/lp/8algUlGzcohIa+CID140Av4DPOKceyekzE+Afs65G30Sz0udc1cWtd/Bgwe7xYvLNu7oN++s4u/zNjL1/F7cMLxbmfZhDMDq1avp1atXVVejwlxyySU8+uij9OjRo8hyixcv5vbbby8wC0hVOnToECNGjGDevHkFZlEJiPR3E5ElPq+DqUTH0pYXJTVVAx7p6dGDHIH7skpKgpC0K8aYasDa8spXUe24MZUqaz38xQcOfroEErtHL7txLrxwkT6+fRU0P8ac5RmL4e9n6uOzH4JTbwmuy8+Hx7pBvyvg/Mcibx/J2z+BZS/pY6kHE9+CbiNg1wZ4+nQ47FNUtugCN34ODQvnw+Otm2DFa5B3CM64V3uKfPZ7uHkpPDsSupwE4/6lZdd/DC9dpo9vXgoJx5XuPQhR0na8IkMk7YE5IvIl2o35I+fcOyLyoIgEBs48ByT4XxN/BtwTZV/GmEr08MMPs21b0d3jHn74YS677LKIvTiqyubNm3n44YcjBi9M3TB+vAYXnIOjR6PfH0u6aMuZYYwxxtQS29KCj7+ZXXDdwSz49BE46Ltj7lgdsm5HwbLOwf/+rENMosnP12EZ+33i+r0ZwXV7wi4uDu6E73bD0he150dJbZoHyafDhDegVTcNaHy3G978MUgMjJsBFz+tx/7w3sLb78uEL1+BQdfA8efComdgxSu6z1ZdYegNsPY92LlWy698UwMlUPC9rEAVOQvJl865gc65/s65vs65B/3yX/mu0Djncp1zVzjnujvnhjrnNlRUfYwxJXfCCSdzZrisAAAgAElEQVQwfPjwIsvcc889pKenc9ppp1VSrYrXo0cPzjjjjKquhqkBjiUftOWSNsYYY2qJbWkQUx+adykYwNjwX/jrqfDpb2GJHz69IyQ4cSAsgLFjlU6JOnMCHD4Y+Vg718DHv4Y0n1ArEMBo3rlwAGOfX3c0F754tmSvZf92HSJy/CjofiZc8jfYtxX+NgK2LIDzH4UTzoOUcXDa7dpTY+37Bfex8GlweXDyT+CUWzS56O5N0Fdn/2PoFGjQHP59Kxz5Dlb/G/pcqu9hZg0PYBhjjDHVVVkThDZurNsaY4wx5hjt3gSPJMODifBQG0h7ufLrsG05tO0DPc6GTZ9pLoev3oIXx0CDptCyK6z3gY0dqzXYAIUDGCtf154Iu76Bj++PfKzsr/V+5xq937cV4ppobonwAMberXrfsissehYO5xT/Wrb4CT87n6T3nQbDaT/TBKE9L4T+VwXLjrgH2vaDf9+mQ0RAX/vi56HXaO1tkXSK5uWoF6vLQHNynPcIbJ4PM66GQ3thwFho01vfy0pgAQxjjDF1zvjx8Mwzms8CNC9GpPuEBL0FluXkwLXX2owlxhhjzDFb854Obxj2I52ec/0nlXt85/RLd/sBcNxIzQ+xepb2Luh4Ivzov9DnYshYBLn7YMca6DpCtz2wveB+Vr6h64b9WIddRHotWT6AERiKsjdD82i0TNIARmhuykDvjFHT4LtdwV4bRdmyEGIaQPv+wWUj7obRf4Yxfyk4fja2vg4HOfBtMHiye6MGJE44T5+LwJin4IrnoXGr4LYDxmpA5JvZ0KgldDtD38Nty8uWSb2ULIBhjDGmTipJvoysLPjjH7XnRbSZSkJnLMnO1u0CM5ZYEMMYY4yJ4ptPIKGHfknvNKRgjonKsCcdcvfql++up2uOiLdugqOHdPhF/SYa2Mg/Cl/O1C/3HVKgQTPNURGQuUy//Pe9DM76NSSeAP8ap7ObhH6hz16v9zvXaq+HfVuhWUdNqHnkIOSETMa5b6vOAnLC+dB5GPz30WAujmi2LISOgyC2QXBZbH3NZ9GoZeHyiT0K1isQYEkISeLfpif0uqjgdiJw0R8hvh30u1JnSWk/AHL3FO5JUgHqVACj4uNBxhhjapupU7XnRWnl5Oi2xhhjjAlzJBc2fa4BAtBpPbPW6bScJbX6HXjxYr29dn1witBINn4G799dMKAQGPLQfoBOKdppiM68cfYDwdlIOg/TYR4LnvL17A1NWhccQrLydagXB70uhLhGMOld6Doc3r1Dk2gGBAIEeYdg10YdJtLcBzBAAyoBgd4ZInDBH7Snyru3a/03L9SknKHJPY/kag6KzkNL/v4FAhWBegWGuBQ1E0tAk0S4eQmM+q0+b5+i95UwjKROBTCMqU3i4+OrugrG1AnHMuuIzVhijDHGRLBlARz9LiSA0Rvyj0D2N/rcuWBuhmjmPwmZSzVp5tf/0WlCV75RuNzBLHhtsiaozFoXXL5tufa6aNNHn59yMwz9EQy5IVgmtgEkn6bTkIIGWuLbBAMY+fnw1ZvQ/axgL4f41nD1KzD4eh36kbNLX0/219BhkD92mg5DadYpmFcjtPfC3gztnQHQri/84Bew6m149Vr4x3mw/GUd6hIIyGQu0/cvkP+iJJokauAmELjIWg9N2uiykmgQDzF+5r+2vfW9tABG+TqGWfOMMcbUUTZjiTHGGFPOvpmtvRaS/Wx2bXrp/U4/jOR/f4ZHusKK1yJvf/SwBi8GToQffgQ3zoPWJ2igYtbNwZlAnIN3btceDIHjBmxbrseNa6jPe12oM3XUC/uKHAiyxLfTXBDxbYLTqG5bpsM9+lxScJt69XRICcCWRRpEyd0LPS8AxNfDQfNOIT0wQgIY+7YGAxsAp94KnYZqEKPvpXDGL2Dtu7D8X/4YgQSepeiBIaK9MEJ7YCT2KHqbaOIaQeueFsAwxhTPOcddd91F37596devHzNnzgRg27ZtDB8+nJSUFPr27ctnn31GXl4ekyZN+r7s448/XsW1N6b6sxlLjDHGmBI6nAPLZ2rPhKJ8Mxu6nKS/4gMkHq+zeATyYHz1BhzaB69fr8MwwntjfPulTjEa+MLeMgkmv6+zbiz9JzxzBsx9DN67SxNzjrwPEroHAxjO6ZCL9gOKf02hw1xAeykEemBs+1Lvu0To+dBhoM7gsWVhsJdD+wHQMhnWf6zPm3eERi10atJAACPvCOz/VtcF1IuBcTPgmrfh0mdh+J2QdKoOi/nvY7DyNWh1nPaqKI3EHsFeL9nr9T0qqw4pkPEFzP293nZtLPu+ihBbIXs1pi55/x74dkX57rNdPzjv4RIVfeONN0hLS2P58uVkZWUxZMgQhg8fzssvv8yoUaOYOnUqeXl55OTkkJaWxtatW1m5ciUAe/bsKd96G1MLjR+v91OnanLOmBhN3Bl+H5itZNcu7XkxbVpwW2OMMaZOWP6y5n5onAA9zopc5sAOvXY+81fBZXEN9Qv4jlWarDIzTWfQOPodfP5HOP7cgskkNy/Q+87Dgsti4jSJZrcRmiNi9m90efez4JRbYF8mLHtJk3RmpkFOVsHto0nsob0fup+pz+PbaMLKo4d0StS4JgV7SwTUb6wBiy0LNWgBGiBo01t7T4AOIQHthREIYOzfBrjgEJKAJgk64wfocI2Ln4Lp58Ic/zpPuaX41xIuobv24tizBXKyy94DAzTQk/YyzH5In3cYqNOxljMLYBhTw82bN49x48YRExND27ZtGTFiBF988QVDhgzhuuuu48iRI1x88cWkpKTQrVs3NmzYwM0338wFF1zAOeecU9XVN6ZGGD++5MGI1FQNdkyYoFOuRgpyZGcHlyUlWbDDGGNMFTu0H2ZOhPMe0aEYZbXZD2X46o3oAYxv5uh9oGdDQJteGsDYMAdw0OMcDQAse0lzW4QGMLYshBZJ0LRd4f13OwNu/0pnDwENbIjo8RY9o8GPRc9ozop+lxf/mkR0mEpAfBu9P7hT69umZ+FhJwGdh8Hi6dCuv05x2qKLvs5AACPQy6JFF53JBIJTqDbvWHh/oVom+9fpe6fE1i/+tYQLBCzWfaD3CccQwOh3OfQeE8zLUa9iQg0WwDDmWJWwp0RFcVHmWx4+fDhz587l3XffZeLEidx1111cc801LF++nA8//JAnn3ySV155henTp1dyjY2pvVJTdfrUwKwlkaZaDQgsC0y5ChbEMMYYU0W2LdfAwdr3ji2AscX3jFj9Dlz4eMEpPQNWvQ1N20O7sOEbbXrBmne0Dg1b6JCEejHQ+2LtJXD4oE5t6pwGMLqOiF6PevWgXtgX+uTT9Ev1omdgzbs6DKN+k9K/xiY+gHFguw55OX5U9LKdh+kMJqvehlbd9PUEhqI0bBE8fosusPG/+tr2btVlkXp1hKsXo7eyCgQs1r6v98fSAwM0WFTBLAeGMTXc8OHDmTlzJnl5eezcuZO5c+cydOhQ0tPTadOmDTfccAPXX389S5cuJSsri/z8fC677DIeeughli5dWtXVN6ZWsSlXjTHG1EiBfAU71pR9H/u26TCI40bCob2w/pPCZb7bA+s/gj6XFu610KYXuHz9st/tjOAX876XwpGcYC+BPekaPChNwkqABk01oLDmHYipD0OnlPYVqkAPjB1rtBdGm97RywaGqOzPDE5PGijfvFOwXMtkOHxA3799vgdG+BCSitCqGyCwca4mVW2RVPHHPEYWwDCmhrvkkkvo378/AwYMYOTIkTz66KO0a9eOTz/9lJSUFAYOHMjrr7/OrbfeytatWznjjDNISUlh0qRJ/O53v6vq6htTq9iUq8YYY2qkwPCFHauKL+sczH9Kp90MFZgJY8Td0KgVrHy98LZr34O8wxqUCBf4Yp9/tODwki4na4+NwBSpgWEqkRJnFue4H+h9yrhgIKK0Attt/K/eB3pURNKsfXCWkUBvh4Tu2hMkNIDR42y9Xz1Le2A0bB5McFqR4hpq/fKPaL6KmOo/QKP617AS5B7JIy6mHjH1bKJVU3McOHAAABHhscce47HHHiuw/tprr+Xaa68ttJ31ujCm4nTpokNCyrqtMcYYUyUCPTB2rtWcCkUNS/h2BXx4ryawHP2n4PItiyC2IXQYpLkQvnxFZyWpHzKV18rX9QtzxxML77dVN+0ZkXc4GGgArUufS+CLv+tUpFsWQv2mRfd8iKbPpTp85NRbS79tQGAIyYZAAKOYenQepj0rAsMzYutDytUF34OE46B9ir4/8e2CyT0rQ2IP7dVyLPkvKpH1wAB6/vIDbv6XfakzxhhzbGzKVWOMMdXamz+G/z5aeHmgB0beoWAw4z/3wSuFfwzjK98T4ps5wYSNoPkvOgzSL+h9L4UjB+HhzvBgoiYIzfpat+l7mSbGDBcTp9OpJvQI9loI6HOpBjYe6QqLn4NOJ5Yt90PCcTDlUz90ooziGkKDZnDgW81jEd+26PKBYSSJxweXjf4znDipYLm+l0HmMp2KtHklBjACgYvEY5hCtRLVqR4YkVMdqvdWfFtp9TDGGFM7FTflqs1CYowxpkqt/1i/xI/4ecHluzbqtJeZy3QYSatusCwVvtsFGUs0YAAasFj5uvaU2LsZstfrL/hHvtNEoKfcrOWSToNRv4WDWTrDyZLnYd2H4PI0GBHNBf8HRAhudBoM5z0K+/13tt5jjvWdODbxbeDQPu19ESkYEyrlag22dBhUdLk+l8BHv9TpXYubgaQ8BQIXNaQHRp0KYBhjjDEVrTRTrhpjjDGVJu+oJp2MbVhw+Xe7IXcPHH8eZKbpzBotOmvwAuB/f4IrX9DHW5focIgRd8N/H4FvZmsAI3OZ5q4I9DaoVw9O/knwGAMnwGvXaV6Hdv2i1zFaXgsRGPajsr3uitCkjQZvisp/EVC/CQy+rvhyLTpD55O0J0tlJPAM6DgYJAY6FhNgqSbq1BASy3BhjDGmOklNheRkvS6LjS14n5iot9Blycm6jTHGGFNqB3cCTmfEyM8LLg8MGWnbG1omwc7VwRlEUiZoYslAmZVvaO+Lk27SXhrfzNblmz7X+05RZgbpkAI//QKu+7D4Hgs1QXxrvS9JAKM0AslNK3MISYcUuCcd2vapvGMegzoVwDDGGGOqi9RUmDIlmPQzL6/gfXa23kKXpafrNhbEMMaYOsY57UFxLA744Rf5R+HAjuDyQP6Lll11SMSO1Zqrol0/GHmf/jo/73HYlwlfvQndz4JGLXSmkI2fwa4N2kuj63BokhD9+PViIK7Rsb2G6iKQ96IsiUSL0u8KHWLTdUT57rc4DZpW7vGOgQUwjDHGmCowdSrk5JR+u5wc3dYYY0wd8p/74Lmzj20f+7cHH+/NCD4O9K5omaw9CrK+1pk+jhup04D2vwqWvgB/6KW9N/pepuWPG6mJOl8Yoz06LgqZkaS2a9YRpF7598Bo3Aqu+Ie+7yYiy4FhjDHGVIHNm6tmW2OMMTXQ5gWQuRR2p+swj7I4EDJpwb4MYIg+3r1Rczo0iNceBS5Pb8eN1PXnPASdh4LLh7jGmmwSIPl07Z2xdzNc9Edo1bXML6/GGTxZ8300blXVNalzrAeGMXVEfHx81HWbNm2ib9++lVgbY0yXLsWXqYhtjTHG1DDOQfbX+njDnLLvp0APjK3Bx7s2BYMPgR4FsY00oSTol/QTr9Uv7QOuCk5f2rAZ9Dwfeo2GQRGmW63NGjaHpJOruhZ1kgUwjDHGmCowbRo0blz67Ro31m2NMcbUETnZkLtXHweSZpbFgW+hUSuIawL7QgIYuzdq/guAhO7aqyL5NIhrGHk/oa56Ca58sXYk5jQ1gg0hMeYYPbLoEdbsWlOu++zZqid3D727yDJ33303SUlJ3HTTTQDcf//9iAhz585l9+7dHDlyhN/85jeMGVO6ebJzc3P58Y9/zOLFi4mNjeUPf/gDP/jBD/jqq6+YPHkyhw8fJj8/n9dff50OHTpw5ZVXkpGRQV5eHr/85S+56qqryvy6jalLAlOtTp2qyTljYjRZZ+A+wedBy84OLktK0uCFTdNqjDF1SJbvfdGsE2z4VPNNBHpBFGfPFp2eE2D/t9C0nSbxDOTAOJKryTlbddPnsQ3gvEeg/YCS18+CF6YSWQDDmBpq7Nix3Hbbbd8HMF555RU++OADbr/9dpo1a0ZWVhYnnXQSo0ePRkpxYnnyyScBWLFiBWvWrOGcc85h3bp1PP3009x6662MHz+ew4cPk5eXx3vvvUeHDh149913Adi7d2/5v1BjarHx4y0YYYwxphiB4SNDrodPHoDMZdBpcPHbbZwLL1wEP5qrAYn93wZnzwgEMPakA65g/oqhN5Rr9Y0pT3UqgOGqugKmViqup0RFGThwIDt27CAzM5OdO3fSsmVL2rdvz+23387cuXOpV68eW7duZfv27bRr167E+503bx4333wzAD179iQpKYl169Zx8sknM23aNDIyMrj00kvp0aMH/fr148477+Tuu+/mwgsv5PTTT6+ol2uMMcYYUzdlfQ0xDWDgRPjkQVj/ScEAxtIXoUUSdAubenPzAr3fulQDGAe2Q+ueUK8e7Fil63aFTKFqTA1QJ3NgWC8nU1tcfvnlvPbaa8ycOZOxY8eSmprKzp07WbJkCWlpabRt25bc3NxS7dO5yKG+q6++mlmzZtGoUSNGjRrF7NmzOf7441myZAn9+vXj3nvv5cEHHyyPl2WMMcYYYwKy1+sQj/jW0GFgwTwY+7fDrJvhxdHw0a/g6OHgum3L9X7HasjP1wBG07bQvDMc2KFlM5fqdKCJPSr3NRlTRnUygGFMbTF27FhmzJjBa6+9xuWXX87evXtp06YNcXFxzJkzh/T09FLvc/jw4aSmpgKwbt06Nm/ezAknnMCGDRvo1q0bt9xyC6NHj+bLL78kMzOTxo0bM2HCBO68806WLl1a3i+xThGRc0VkrYisF5F7IqxvICIz/fqFIpIcsu5ev3ytiIwqbp8i0tXv42u/z/p++c9EZJWIfCkin4hIkl+eIiLzReQrv+6qkH09LyIbRSTN31Iq5h2q21JTITlZg/CxsXqfmKi30GWB++Rk3cYYY0wNl/U1JHbXx8eNhIwv4Lvd+nzDp3rfYxR8/kd4+6bgdoEAxs7V8N0uzX0R3w6adQQc7M/UYEiHQdCoRWW9GmOOSZ0MYET5gdmYGqdPnz7s37+fjh070r59e8aPH8/ixYsZPHgwqamp9OzZs9T7vOmmm8jLy6Nfv35cddVVPP/88zRo0ICZM2fSt29fUlJSWLNmDddccw0rVqxg6NChpKSkMG3aNO67774KeJV1g4jEAE8C5wG9gXEi0jus2PXAbudcd+Bx4BG/bW9gLNAHOBd4SkRiitnnI8DjzrkewG6/b4BlwGDnXH/gNeBRvzwHuMY5FzjGEyISerVzl3Muxd/SyuEtMSFSU2HKFE32CZrQEzTBZ3Z2wWWB+/R03caCGMYYU4PlHdFZQhJ8D4me54PLgzWaf4xvZkPjBBg3A06cDKv/rYk5D2bD3i06o8iO1Zr/AnwPjI76ePtXsHWJBkWMqSHqVA4MGzliaqMVK1Z8/zgxMZH58+dHLHfgwIGo+0hOTmblypUANGzYkOeff75QmXvvvZd77723wLJRo0YxatSoQmVNmQwF1jvnNgCIyAxgDLAqpMwY4H7/+DXgL6IZWscAM5xzh4CNIrLe749I+xSR1cBI4Gpf5gW/378650InmF8ATABwzq0LLHTOZYrIDqA1sOfYX7opztSpkJNT+u1ycnRbSxRqjDE11O507TkRGOLRYRC0TIaVb0DKeA1gdPuB5rU44XxY8g/YPJ/vs//1OBvWfaDBCtAeGI39NFdpL4PLh+5nVvarMqbM6mQPDGOMqYY6AltCnmf4ZRHLOOeOAnuBhCK2jbY8Adjj9xHtWKC9Mt4PXygiQ4H6wDchi6f5oSWPi0iD6C/TlMXmzVWzrTHGmGO0fRV8cK9OfVoWgRlIEvwQEhHoc6kOHdnwKRzcEexBkXwq1IvToEZg+Eh/P+IzMNQktAfGug+gQTPoeGLZ6mZMFbAAhjF1yIoVK0hJSSlwGzZsWFVXy6hIncTCB7xFK1Ney4MHEpkADAYeC1veHvgnMNk5l+8X3wv0BIYArYCIU/OIyBQRWSwii3fu3BmpiImiS5eq2dYYY8wxWvU2LHhKh3GURVZYAAOg72U6jOT9n+vz436g9/WbQJeT4Js5kJmmM5MknaLrNvgOlvHttFyjltqzo+twiIkrW92MqQJ1agiJMeXJOYfUsClt+vXrR1pa3UxPEG12lWokA+gc8rwTkBmlTIaIxALNgV3FbBtpeRbQQkRifS+MAscSkbOAqcAIPywlsLwZ8C5wn3NuQWC5c26bf3hIRP4B3BnpBTrnngGeARg8eHC1/4NUJ9OmaT6L0g4jadxYtzXGGFNFDvjcE1sWQLu+pd8++2sd8tG4VXBZ2z6QeDxkrYPWvaBZh+C640bCJw9ogs6kUyG+rQYr9m/T3hb1G2u5Zp00EWgg+GFMDWE9MIwpg4YNG5KdnV0TvhQbNHiRnZ1Nw4YNq7oqRfkC6OFnB6mPJuWcFVZmFnCtf3w5MNvpP+EsYKyfpaQr0ANYFG2ffps5fh/4fb4NICIDgb8Bo51zOwIH9tu/CbzonHs1tFK+VwY+H8fFwMpjfjdMAePHwzPPQFKSPo+J0fuEBL2FLgvcJyXpNpb/whhjqtD+7Xq/ZVH0MjvXwvqPI6/LWh9M4Bkgor0woHACzsDznGxoP0DLtvH5u+PbBssFhpFYAk9Tw1gPDGPKoFOnTmRkZGDd4GuOhg0b0qlTp6quRlTOuaMi8lPgQyAGmO6c+0pEHgQWO+dmAc8B//RJOnehAQl8uVfQhJ9HgZ845/IAIu3TH/JuYIaI/AadeeQ5v/wxIB541fcw2uycGw1cCQwHEkRkki87yc84kioirdGhKWnAjeX/Dpnx4y0YYYwxNc73PTAWFl7nHCx+Dj6cCnmH4Y61EN8muD5zGWQsgqE/Krxt/6tg0bPQ55KCy9v1h8aJkJMF7f2s5q17Qvrn0LRdsFyXkyF3L7Tqdmyvz5hKZgEMY8ogLi6Orl27VnU1TC3jnHsPeC9s2a9CHucCV0TZdhpQaLBApH365RsIzlQSuvysKPt/CXgpyjr7+cYYY4yJZP92kHqwe5M+bup7QeTsgn/fotOedhqqgYpVb8PQG3T9kVx480Zo0hpG3FV4v626ws+/Kby8Xj0dFrLiVWjfX5e16aX3oT0wTrtNb8bUMBU2hEREOovIHBFZLSJficitEcqcISJ7RSTN334VaV/lxTr7G2OMMcYYYypFfr7OEpJ8mj4P9MLYvACePh3Wvg9nPwTXfajDPFa+Htx29kOwcw2M+YvmsCiNU2+Fsx4I9uYIDCEJ7YFhTA1VkT0wjgJ3OOeWikhTYImIfOScWxVW7jPn3IUVWI9CaljeRWOMMcYYY0xNk5OtM330OAc2L9QARkJ3eGG0Jt68/j/BKUz7XgqzfwN7M2B3Osx/EgZfB90jdowsWrt+egto2xtiGthwEVMrVFgAw2el3+Yf7xeR1UBHdIy2McYYY4wxxtRegfwXzTtDx0GwaR5snAsNmsL1H0F862DZPj6AsewlSEuFlsnaO6M8NGoJP1mgM48YU8NVyiwkIpIMDAQiZK/hZBFZLiLvi0ifKNtPEZHFIrK4PJIm2sQRxhhjarPUVEhO1qHQiYl6E4HY2Mj3ycm6jTHGmHIUmIGkaTvoPAy2pcG3X8JFTxQMXgAkHKdJNz/9HezZApc8DQ3iy68urbpBbP3y258xVaTCAxgiEg+8DtzmnNsXtnopkOScGwD8GXgr0j6cc8845wY75wa3bt06UpEivfNlJpl7vsNGjhhjjKntUlNhyhRIT9eAfXa23gDy8iLfp6frNhbEMMaYchTogRHfVgMYAAOuhl4XRS4fmBr11Fuhy0kVXz9jaqAKnYVEROLQ4EWqc+6N8PWhAQ3n3Hsi8pSIJDrnssqrDs45fvryMjo0b8j5/dqX126NMcaYamnqVMjJKf12OTm6rU3Vaowx5WS/D2A0bQfNO8F5j0LK1dHLD54MsQ3gxEmVUj1jaqIKC2CIiADPAaudc3+IUqYdsN0550RkKNojJLsi6pO5N7cidmuMMcZUK5s3V822xhhjwhzYDg2bQ1wjfT7sR0WXb9C0+DLG1HEV2QPjVGAisEJE0vyyXwBdAJxzTwOXAz8WkaPAd8BY5yxDhTHGGFNWXbrokJCybmuMMaac7N8G8TZ1qTHlqSJnIZkHRaedcM79BfhLRdXBGGOMqWumTdN8FqUdRtK4sW5rjKl9RORc4I9ADPB359zDYeuTgOlAa2AXMME5l1HpFa1t9m+Hpm2ruhbG1CqVMguJMcYYYyrH+PHwzDOQlKQzjCQk6A0gJibyfVKSbmP5L4ypfUQkBngSOA/oDYwTkd5hxX4PvOic6w88CPyucmtZSx341npgGFPOKjSJpzHGGGMq3/jxFowwxnxvKLDeObcBQERmAGOAVSFlegO3+8dziDIzoCkF56wHhjEVoE71wLDkGsYYY4wxpo7pCGwJeZ7hl4VaDvg5PLkEaCoiCeE7EpEpIrJYRBbv3LmzQipba+TugbxD1gPDmHJWpwIYAVJkZg5jjDHGGGNqjUhXvuG/690JjBCRZcAIYCtwtNBGzj3jnBvsnBvcunXr8q9pTeccvHUTLH1Re1+ATqFqjCk3NoTEGGOMMcaY2isD6BzyvBOQGVrAOZcJXAogIvHAZc65vZVWw9oi/XNIS4V1H8KYJ3VZvA0hMaY81foeGDYpqzHGGGOMqcO+AHqISFcRqQ+MBWaFFhCRRBEJfC+4F52RxJTW53+C2EaQkwWfP6HLrAeGMeWq1gcwIrGghjHGGGOMqQucc0eBnwIfAquBV5xzX4nIgyIy2hc7Ay7j7pAAACAASURBVFgrIuuAtoBNqlxaO9bA1x/CabdDh0Gweb4utx4YxpSrOhXAsNQXxhhjTEGpqZCcDPXqQWKi3kQgNjbyfXKybmOMqTmcc+855453zh3nnJvml/3KOTfLP37NOdfDl/mhc+5Q1da4Bvrfn7X3xZAfwik367K4xtCgadXWy5haxnJgGGOMMXVUaipMmQI5Ofo8Ozu4Li8v8n16um4DNlWrMaaO27wQXrpMZxvJOwxDboAmCdBrNLRIgnoxNnuAMeXMAhjGGGNMHTV1ajB4URo5ObqtBTCMMXXaxv/C4f1w6m0QUx+G+uhuTCxc8Q/ItTyoxpQ3C2AYY4wxddTmzVWzrTHG1Ao7VkHLZDj7gcLrOp5Y6dUxpi6oUzkwjDHGGBPUpUvVbGuMMbXCjjXQpndV18KYOsUCGMYYY0wdNW0aNG5c+u0aN9ZtjTGmTsg7Wngaw6OHIftraNOraupkTB1VpwIYNnuqMcYYEzR+PDzzDCQlaZ65hAS9AcTERL5PStJtLP+FMaZOOPId/N8JsOyfBZdnr4f8o9DaAhjGVKY6mQPDkgEbY4wxavx4C0YYY0xU21dBThas+xAGXRNcvmOV3lsPDGMqVa3vgWG9LowxxhhjjDFlsi1N77csLDiMZMdqkBhI7FE19TKmjqr1AQxjjDHGGGOMKZNty/X+4E7YtSG4fMdqSOgOsQ2qpl7G1FEWwDDGGGOMMcaYSLYth6Yd9PGWRcHlO1ZBm55VUydj6rA6GcAITyJsjDHGGGOMMQUcPayBin6XQYPmsGWBLj+cA7s32RSqxlSBOhXAsNydxpjqTETOFZG1IrJeRO6JsL6BiMz06xeKSHLIunv98rUiMqq4fYpIV7+Pr/0+6/vlPxORVSLypYh8IiJJIdtc68t/LSLXhiw/UURW+GP8ScRSJRtjjKkFdq6BvMPQYSB0HhLsgZG1FnCWwNOYKlCnAhgBq7btY+93RwBw1h3DGFMNiEgM8CRwHtAbGCci4T/tXA/sds51Bx4HHvHb9gbGAn2Ac4GnRCSmmH0+AjzunOsB7Pb7BlgGDHbO9QdeAx71x2gF/BoYBgwFfi0iLf02fwWmAD387dxyeVNMtZSaCsnJOqNXbGzR98nJWt4YY2qkQP6L9inQeZjmvfhuj96D9cAwpgrUyQDGm8u2Mu6ZBVVdDWOMCTUUWO+c2+CcOwzMAMaElRkDvOAfvwac6Xs7jAFmOOcOOec2Auv9/iLu028z0u8Dv8+LAZxzc5xzOX75AqCTfzwK+Mg5t8s5txv4CDhXRNoDzZxz851GhF8M7MvUPqmpMGUKpKfr87y8ou/T07W8BTGMMTXStuVQvym07KoBDBys/xiWz4CYBrrcGFOp6mQAA7QXhjHGVCMdgS0hzzP8sohl3P+zd+dhcpVV4se/p6t6TSfdSXdnISHpbCwBZAuLg4JBBVQE/A0gGlRUJjqDjusIGAcFJzMoKKOjjkRRlGmEyBogJGyBAIEsBMhCErJ1d7rTSXrfq5eq8/vj3uq1ek1tXXU+z1PPrXvrvu99uzup7nvqvOdV7QTqgbxB2g50PA+oc/sY6FrgZGU8O8T4prvPBxs3ACKyREQ2i8jmysrKUKeYOLd0KbS0DH1eTy0tTjtjjBlzKt6FaR+AlBSYfrazbOpj/wTFr8ElPwWPN9YjNCbpJFUAwyaLGGPiWKi6EX3ftgY6J1zHuy8kcj2wELhrlNfuf1B1uaouVNWFBQUFoU4xca60NLrtjDEmZgJ+OLzNmT4CkJ4Ns/4BcmbAV9bAeV+L7fiMSVIJHzYcqsZFac0IP0oyxpjIKAOO77E/Azg0wDllIuIFcoCaIdqGOl4F5IqI183C6HUtEfkYsBS4SFXbelz7I336etk9PqPP8b7jNgli5szu6SMjbWeMMWNC6QZ49RfQ0QKdrTDt9O7XFv8dUlIt88KYGEqqDIy+iquaueiul2M9DGOMAdgEzHdXB0nDKcq5ss85K4Hg6h9XAy+5dSdWAte5q5TMximkuXGgPt02a90+cPt8EkBEzgTuBa5Q1aM9rr0GuEREJrrFOy8B1qhqBdAoIue7tTW+GOzLJJ5lyyAra2RtsrKcdsYYE/daamDFF+HQFuhohdkXwtxF3a+nZlrwwpgYS+r/gYcbfLEegjHGAE5NCxH5Bk6gwAP8SVV3iMgdwGZVXQncBzwgIntxMi+uc9vuEJEVwHtAJ3CTqvoBQvXpXvJm4CER+Q+clUfuc4/fBWQDf3dXQy1V1StUtUZEfooTFAG4Q1Vr3Of/DNwPZOLUzAjWzTAJZvFiZ7t0qZOJ4fE4BTsH2s6a5QQvgu2MMSaurfo3aKmCf3qpd+aFMSZuJHUAwxhj4omqrgJW9Tl2W4/nPuCaAdouA/p9zh2qT/f4fpxVSvoe/9gg4/sT8KcQxzcDpw7UziSWxYstIGGMSUA7noDtj8CiH1nwwpg4ltRTSIYoj2GMMcYYY4xJBlv+ApPmwoe+E+uRGGMGkVQBjGe2Voy67fbyesrrWsM4GmOMMcYYY0xcaK6E/BOsxoUxcS6pAhjHUvPi8v95jQvufCmMozHGGGOMMcbEheZqGJcX61EYY4aQVAEMY4wxxhhjjOlF1SnemWUBDGPinQUwjDHGGGOMMcmrvQn87ZCVH+uRGGOGkPABDKvTaYwxxhhjjBlQc5WztQwMY+JewgcwBnO0cfQ1MYwxxhgzuKIiKCwEEfB6Q2/z851HSopzblFRrEdtjEk6LTXOdpxlYBgT7yIWwBCR40VkrYjsFJEdIvKtEOeIiPxaRPaKyFYROStS4wnlWw+9E83LGWOMMUmjqAiWLIGSEmff7w+9ra52HqrOuUuWWBDDGBNlLcEMDAtgGBPvIpmB0Ql8T1VPBs4HbhKRBX3O+QQw330sAf43guMxxhhjTJQsXQotLSNv19LitDXGmKjpmkIyKbbjMMYMKWIBDFWtUNUt7vNGYCcwvc9pVwJ/VcebQK6ITIvUmIwxxhgTHaWlsWlrjDH97HsJ1v/PwK+3VDtbm0JiTNyLSg0MESkEzgQ29HlpOnCwx34Z/YMciMgSEdksIpsrKysjNUxjjDHGhMnMmbFpa4wx/bzzILx4B3S2h369pQo8aZCWHd1xGWNGLOIBDBHJBh4Fvq2qDX1fDtGk38IhqrpcVReq6sKCgoJIDDPu1DQP8AZrjDHGjAHLlkFW1sjbZWU5bY0xJmx89c4yqZW7Qr/eXO3Uv5BQtybGmHgS0QCGiKTiBC+KVPWxEKeUAcf32J8BHIrkmMaCVdsqOOunz7O5uCbWQzHGGGNGZfFiWL4cZs1y9j2e0Nu8POch4py7fLnT1hhjwsZX72wr3g39eks1jLMlVI0ZCyK5CokA9wE7VfWXA5y2EviiuxrJ+UC9qlZEakxjxYb9zjy87eX1MR6JMcYYM3qLF0NxsbPCSGdn6G1VlfMIBJxzLXhhjAm7IQMYVZBlAQxjxgJvBPu+APgCsE1EguuV/hCYCaCqvwdWAZ8E9gItwJcjOB5jjDHGGGNMsukKYLwT+vWWasidFb3xGGNGLWIBDFV9jdA1Lnqeo8BNkRqDc41I9m6MMcYYY4yJaz63DN/h7eDvBI8X6stgwnRn/lpztWVgGDNGRGUVEmOMMcYYY4yJOn8HdDRD3jzobIXqPVC2Ge45FXY946xM0lZvS6gaM0ZYAMMYY4wxxhiTmILZF7MvdLYV78Jr9wAKB9+EVrdovmVgGDMmWADDGGOMMcYYk5h8dc52+kLwZsL2x5zMC3CCGc1VznMLYBgzJlgAwxhjjDHGGJOYggU8s/Jg6mmwZw14UuHET7kBjErndZtCYsyYYAEMY4wxxhhjTGIKBjAycmDa6c7z06+D+R93Xjv0tnPMMjCMGRMsgDFChbc8wwvvHYn1MIwxxpiEUlQEhYXOggBeb+htfr7zSElxzi0qivWojTFxr2cAo/AC8KTDB7/ZHczY95KzzbIMDGPGAgtgjMIz2ypiPQRjjDEmYRQVwZIlUFLi7Pv9obfV1c5D1Tl3yRILYhhj+ggEYM1SqNzt7PcMYCy4Cr7/PhScAJMXQIoXSt8EBDInxmzIxpjhswCGMcYYY2Jq6VJoaRl5u5YWp60xxnRpPARv/AbeW+ns9wxgiEBmrrOfmgEFJ0Ogwznm8cZmvMaYEUn4AIai4e9Tw9+nMcYYk6xKS2PT1hiTgBoPu1s3Y9pXD+KBtHH9zw1OI7HpI8aMGQkfwDDGGGNMfJs5MzZtjUkWInKZiOwWkb0ickuI12eKyFoReVtEtorIJ2MxzrAIBjCa3Jp1vnrImOBkX/R13BnO1gp4GjNmWADDGGOMMTG1bBlkZY28XVaW09YYMzAR8QC/BT4BLAA+JyIL+pz2I2CFqp4JXAf8LrqjDKOmYAaGu/XVO9NHQglmYNgSqsaMGRbAGAUJFcE1xhhjzKgsXgzLl8OsWc6+xxN6m5fnPEScc5cvd9oaYwZ1LrBXVferajvwEHBln3MUmOA+zwEORXF84dXoZl70ysAYIIAx5RSQFMiaFJ2xGWOOmQUwRiFaNTCs0oYxyWUYKb7pIvKw+/oGESns8dqt7vHdInLpUH2KyGy3jz1un2nu8QtFZIuIdIrI1T3OXyQi7/R4+ETkKve1+0XkQI/XzojMd8gkssWLobjYWWGkszP0tqrKeQQCzrkWvDBmWKYDB3vsl7nHevoJcL2IlAGrgG9GZ2gR0NRjConq4AGMtHFw2Z1w9g1RG54x5thYAMNEVE1zOzcVbaHR1xHroRgT14aZ4vtVoFZV5wH3AD9z2y7ASfk9BbgM+J2IeIbo82fAPao6H6h1+wYoBW4AHux5YVVdq6pnqOoZwMVAC/Bcj1P+Lfi6qr5zbN8NY4wxYRQqdbjv52SfA+5X1RnAJ4EHRKTffYKILBGRzSKyubKyMgJDDYNgBoa/HVprBw9gAJz3NZh+dnTGZow5ZhbAiGOJMFHld2v38sy2Ch7aeHDok41JbsNJ8b0S+Iv7/BHgo+LMabsSeEhV21T1ALDX7S9kn26bi90+cPu8CkBVi1V1KxAYZKxXA8+q6igWvjTGGBNlZcDxPfZn0H+KyFeBFQCq+gaQAfQrDKGqy1V1oaouLCgoiNBwj1EwAwOcOhhDBTCMMWOKBTCMMSY+DCfFt+scVe0E6oG8QdoOdDwPqHP7GOhag7kO+FufY8vcyvX3iEh6qEZj4pM7Y4xJPJuA+e7UwTSc9/CVfc4pBT4KICIn4wQwxuYbdeMRmDTXed4UDGDkxnZMxpiwSfgAxrPbDg99kjHGxN5wUnwHOidcx4ckItOA04A1PQ7fCpwEnANMAm4O1XZMfHJnjDEJxg1WfwPnfXsnzmojO0TkDhG5wj3te8A/ici7OAHqGzRaRd/CKeCH5qPdq4vUl0NHs2VgGJNAEj6A8e2HbSp2NPg6/Nz57C5a2/2xHooxY9VwUny7zhERL06l+JpB2g50vArIdfsY6FoDuRZ4XFW7CtuoaoU62oA/40xdMcYYE2Yi8qiIfCpUfYrBqOoqVT1BVeeq6jL32G2qutJ9/p6qXqCqp7u1jJ4bvMc41VwFGugOYFS972wtgGFMwkj4AEYkjL1wdOTdv76Y37+yjz+8uj/WQzFmrBpOiu9K4Evu86uBl9xPyFYC17mrlMwG5gMbB+rTbbPW7QO3zyeHOc7P0Wf6iJuVgVtb4ypg+zD7MsYYMzL/C3we2CMid4rISbEeUFwJ1r+YNAfSsi2AYUwCsgCGCYv2TqfeX4d/sLp/xpiBDDPF9z4gT0T2At8FbnHb7sApvvYesBq4SVX9A/Xp9nUz8F23rzy3b0TkHHcZvWuAe0UkeD7usq3HA6/0GX6RiGwDtuEUffuP8HxXjDHG9KSqL6jqYuAsoBh4XkTWi8iXRSQ1tqOLA8EVSMZPhewpULnb2bcAhjEJwzv0KaavaK0OYpkexiQXVV0FrOpz7LYez304gYVQbZcBy4bTp3t8PyGmeqjqJpwpJaGuUUyIYp+qenGo840xxoSfiOQB1wNfAN4GioAP4WTTfSR2I4sDwQyM8VNh/DQoXe/sWwDDmIRhGRgmoiwIY4wxJtyKiqCwEETA6+29zc93Hj2PFRY6bYwZ60TkMeBVIAv4tKpeoaoPq+o3gezYji4OBDMwsqfA+ClOPQywAIYxCcQyMAbg6/AjAuleT7/XonVTHq1MD2OMMWasKCqCJUugpcXZ9/t7b6uru88NHispcdoALF4cnXEaEyG/UdWXQr2gqgujPZi403QYMieCNx2yp3YftwCGMQnDMjAGcNK/r+Zjv+w7zduMlAVhjDHGhNPSpd3Bi5FoaXHaGjPGnSwiucEdEZkoIv8SywHFlcbD3YGL8VO6j1sAw5iEYQGMQRysaQ153G7KjTHGmNgoLY1NW2PixD+pal1wR1VrgX+K4XjiS9OR7sBFMJAhKc6KJMaYhGABjFGI1BQSX4c/Qj1HnlqxC2OMMVEwc2Zs2hoTJ1LcJasBEBEPkBbD8cSXxiP9MzDSJzjFcIwxCcECGEP47op3aGrrjPh1tpfXc9K/r+a5HYcjfq1Isl8PxhhjImnZMsjKGnm7rCynrTFj3BpghYh8VEQuBv6Gs3y2UXVqYPTNwLDpI8YkFCviOYTHtpQzb3Lk087ePuhkA77yfiXelLEbBrBEDGOMMZEULMK5dKlTnNPjcYp1Brd5ec7r1dXdx2bNcoIXVsDTJICbga8B/4zzudFzwB9jOqJ40VoL/vb+GRgWwDAmoVgAYxgkRnkFYykYYJl5xnRz03sXA3NU9Q4RmQlMVdWNMR6aMQlh8WILRpjkpKoB4H/dh+mp0c1i7gpc5IIn3QIYxiQYm0JiImosBWGMCaPfAR8EPufuNwK/jd1wjDHGJAIRmS8ij4jIeyKyP/iI9bjiQtMRZ9tVvFNgwjTIzB24jTFmzLEMjGHQGN2GJ1JSg2VoxJ+NB2pYOGsiKWN4ylIcO09VzxKRt8GpEi8iVmTNGGPMsfoz8GPgHmAR8GUS60/G0Ws66myzeyyfevl/w7j82IzHGBMRw8rAEJFvicgEcdwnIltE5JJIDy5e7Kpo7LVvK26MnH3P4stre6q49t43uHedfWgTIR1uZXgFEJECIBDbIRljjEkAmar6IiCqWqKqPwEujvGY4kNXBsbk7mNzF8HU02IzHmNMRAx3CslXVLUBuAQowIn23hmxUcUZf8DuvocyUIDCPhKIT4fqWwHYV9kU45EkrF8DjwOTRWQZ8BrwX7EdkjHGmATgE5EUYI+IfENEPgNMHqpRUmg+Ct4MSB8f65EYYyJouAGM4H3oJ4E/q+q7JPG9qQi0dfp58p1yNFypBQmSopC0/yiM6UFVi4Af4AQtKoCrVHVFbEdljDEmAXwbyAL+FTgbuB74UkxHFCuqvf9+bqqEcZNt3rIxCW64NTDeEpHngNnArSIyniRKhw5VA+PuNbv5w6sHmJCZyqITLfBtjOkmIg+o6heAXSGOGWOMMSPmTk28VlX/DWjCyYhOXo9/HQKdcPV9zn7Tkd7TR4wxCWm4GRhfBW4BzlHVFiCVId40ReRPInJURLYP8PpHRKReRN5xH7eNaOQxpAqHG9oAaGjtCE+nFi02MZAgiT/x6JSeO+4fnWfHaCzGGGMSgKr6gbPdpbrNke1Q/lb3fnOlBTCMSQLDzcD4IPCOqjaLyPXAWcCvhmhzP/Ab4K+DnPOqql4+zDFEXHldK9NzM2M9jC5j8d5yLI45GdlfPpEhIrcCPwQyRaSB7m91O7A8ZgMzxhiTKN4GnhSRvwPNwYOq+ljshhQjzZXQWud8GiPirEIyY2GsR2WMibDhZmD8L9AiIqfjzOsuYfDABKq6Dqg5tuFF1xX/81rI4xLl272xGFgfaMgW0DDJRFX/S1XHA3ep6gRVHe8+8lT11liPz5hkVFQEhYXO7ymv19nm5zuPnseC28JCp40xcWoSUI2z8sin3UfcfBgYNYEANFeBv83ZBvzQUuXUwDDGJLThZmB0qqqKyJXAr1T1PhEJR8GgD4rIu8Ah4PuquiPUSSKyBFgCMHPmzDBcNrTq5vaQx0PVwOgpEFDOuOM5fvjJk7nu3GMfX9gKg8aRMRiTMWbUVPVWEZkIzAcyehxfF7tRGZN8iopgyRJoaXH2/X5nW13dfU7wWHBbUuK0AVi8ODrjNGa4VDW5614E+epA3f+09QcBBQ3YFBJjksBwAxiNbmr0F4APu/O5U4/x2luAWaraJCKfBJ7A+WO/H1Vdjpt+vXDhwri7u+8IBGjwdXLbkztGH8AIEbToe8//w8e3ke5N4cefPqXfucaY+CEiNwLfAmYA7wDnA2/gfGJmjImSpUu7gxcj0dLitLUAhok3IvJnQiS4qupXYjCc2Gmu7H7eUA4e97bEAhjGJLzhTiH5LNAGfEVVDwPTgbuO5cKq2qCqTe7zVUCqiOQfS5+JYLBMhQc3lPLn14ujNpaRGCppJAGTSowZzLeAc4ASVV0EnAlUDt7EGBNupaWxaWtMBD0NPOM+XgQm4KxIklx6BjDqy536F2BTSIxJAsMKYLhBiyIgR0QuB3yqOmgNjKGIyNRgFWUROdcdS/XgrSLv9b1VsR7CmNY3/mIzR0yS8qmqD0BE0lV1F3BijMdkTNI5llmnEZyxasyoqeqjPR5FwLXAqbEeV9T1ysAo6963DAxjEt6wAhgici2wEbgG541yg4hcPUSbv+GkTJ8oImUi8lUR+bqIfN095Wpgu1sD49fAdRoHxR8W/3HDiNuEc9Sx/w4YEx2v7ani9qdClr1JBGUikoszNe55EXkSp9aPMSaKli2DrKyRt8vKctoaMwbMB5Iv3NbsfuCYnuNmYBxx9i2AYUzCG24NjKXAOap6FEBECoAXgEcGaqCqnxusQ1X9Dc4yq4njWNINkqzKZV1LO42+To6fNIq/LE3YDFWgNpKuv88JFiZiTRdV/Yz79CcishbIAVbHcEjGJKVgDYulS53inB6PU6wzL885Xl3dfSy4nTXLCV5Y/QsTj0Skkd41MA4DN8doOLHTXAkITD0N6stgwnHgzYS07FiPzBgTYcMNYKQEgxeuaoZfP2PMi/YyqkGJnIxx4c/X0uDrpPjOT8V6KElpLC7VO1aISAqwVVVPBVDVV2I8JGOS2uLFFowwicNdqts0V0LWJMidCQdecfazJyfdB4LGJKPhBiFWi8gaEblBRG7AKRy0KnLDii+hPqWO5NtjIt1cDhSEafB1RnUcxkSLqgaAd0Uk+VJ6jTHGRJSIfEZEcnrs54rIVbEcU0w0V8K4AsiZDo0V0HDIpo8YkySGlYGhqv8mIv8IXIBz775cVR+P6MiSTY/iF3FQCmTUBhp5AsVkjBmOacAOEdkINAcPquoVsRuSMcaYBPDjnn+Dq2qdiPwYp+ZS8miuhqx8mDAdNACHt8GsC2I9KmNMFAx3Cgmq+ijwaATHEjc6/YEhz4lUiKHnjX607/l9HX5e3VPFxxdMifKVTcyM3VhZvLs91gMwxhiTkEJlTw/77/mE0VwJU06BnBnOvq/OMjCMSRKDvuGFKBTU9RKgqjohIqMKowJqAahk4rDbzFv6bK/9vjUwEvWe746n3+PBDaU8edMFnH587qj6sEQLY6zuhTHGmIjZLCK/BH6L8yfpN4G3YjukGAhOIZkwvfuYBTCMSQqD1sBQ1fGqOiHEY/xYCF7Q1sTr6f/Kl71rjqmbWK7UEE2l1S0ANPg6wt73GJ4Vk9gs4hRWItIoIg0hHo0i0jCM9peJyG4R2Ssit4R4PV1EHnZf3yAihT1eu9U9vltELh2qTxGZ7faxx+0zzT1+oYhsEZHOvstli4hfRN5xHyuH6ssYY0zYfRNoBx4GVgCtwE0xHVG0dbY7GRfjCrozMMDZN8YkvMReSSQ9m02Bk/hYyrEFpkPdfEfqvi/RbvTt/tgkk2MJ+oqIB+cTtU8AC4DPiciCPqd9FahV1XnAPcDP3LYLgOuAU4DLgN+JiGeIPn8G3KOq84Fat2+AUuAG4MEQw2xV1TPcR896HgP1ZYwxJoxUtVlVb1HVhe7jh6raPHTLBNJS7WzH5UPGBEh3f71m2xRoY5JBYgcwgBcCZ3FCSjmz5PCo+xhuAcpjulmPoyqXiRZEMWaMOBfYq6r7VbUdeAi4ss85VwJ/cZ8/AnxUnGWLrgQeUtU2VT0A7HX7C9mn2+Zitw/cPq8CUNViVd0KDF0MCBisL2OMMeElIs+LSG6P/YkicmypxmNNc6WzDWZcBKeR2BQSY5JCwgcwng+cDXBMWRiBYf0Z3+39I42U17WOrFGIqEG04whxFEMx0WLBqngyHTjYY7/MPRbyHFXtBOqBvEHaDnQ8D6hz+xjoWqFkiMhmEXmzx7J9w+5LRJa47TdXVlYO43LGRFdRERQWOr8PvV5nm5/vPHoeC24LC502xkRRvqrWBXdUtRZIrjv3vgGMHAtgGJNMEj6AUaaT2Rk4no97toy6j9U7emdvVDW2hTwveC94yT3ruODOl0Z1LRGQBIok2P1xfEqcf2EJJdSPpe9/oYHOCdfxocxU1YXA54H/FpG5I+lLVZcH054LCmyusokvRUWwZAmUlDj7fr+zra52Hj2PBbclJU4bC2KYKAqIyMzgjlsLKbn+3GqucrZ9MzDGWQDDmGSQ8AEMgBcCZ7NQdpNLY1j6e2N/dVj6SSYJFJMxJlLKgON77M8ADg10joh4gRygZpC2Ax2vAnLdPga6Vj+qesjd7gdeBs4cbV/GxJulS6GlZeTtWlqctsZEyVLgNRF5QEQeAF4Bbo3xmKKrKwMj39nOvwROuAzSs2M3JmNM1CRHAMN/Fl4JsCjlnYheJ1z36GpFKIxJRpuA+e6KHmk4RTlX9jlnJfAl9/nVwEvqvGGsBK5zVymZDcwHNg7Up9tmrdsHbp9PDjY4d551uvs8H7gAeG80fRkTj0pLY9PWmJFQeJx2JAAAIABJREFU1dXAQmA3zkok38NZiSR5NFdCSipk5Dj7J18On384tmMyxkRNUgQwtuocjmguHz2GaSR9rXw38h8wWtKCMcnDrSHxDWANsBNYoao7ROQOEQmu+HEfkCcie4HvAre4bXfgLKf3HrAauElV/QP16fZ1M/Bdt688t29E5BwRKQOuAe4VkeD5JwObReRdnIDFnar63mB9GTOWzJw59DmRaGvMSIjIjcCLOIGL7wEPAD+J5ZiirrnKmT5i6b3GJCXv0KeMfUoK7wTmcaKUxXooJgmV17Xy7LYKbvzwnFgPxcQ5VV0FrOpz7LYez304gYVQbZcBy4bTp3t8P84qJX2Pb8KZBtL3+HrgtAGuHbIvY8aSZcucehYjnUaSleW0NSZKvgWcA7ypqotE5CTg9hiPKbqaK7unjxhjkk5SZGAAlGs+06WKSNQ5OtYZH2+V1PLvT+4IS1/hEIkhxMPXFStfvX8T//HMTg6NdGUaY4wxUbN4MSxfDrNmOfsej7PNy3MePY8Ft7NmOW0WL47uWE1S87nBbEQkXVV3ASfGeEyR1+GDVT+AgxvdAIYVgjYmWSVFBgY4AYwsaWMijdQyISLXGG0m211rdoV3ID08+U45J04dz0lTI/M1D8WS+6DR56wuGUjmKI4xxowBixdbMMLEvTIRyQWeAJ4XkVqSoXByxbuw8V7Y9EdI8cIpVw3dxhiTkJImgFGmTqrZdKmiVmNzMz8cPYMg4bjd/dZDTuHS4js/FYbeeqtsbGPSuDQ8KQOHKeyWPb7Zz8cYY4wZO1T1M+7Tn4jIWpzVqFbHcEjR0dbgbKedDoe2wPhpsR2PMSZmkiaAUd4jgLFdw1eLoLXd3/U8XB+wyxgoSlTT3M45y15gyYVz+OEnTx7y/DHwJUVcPCVg2M/DGGOMGdtU9ZVYjyFqfPXO9jO/h6ajMHlBbMdjjImZpKqBATBDqsLa7y2PbTvmPmQMTrSobWkH4IX3jgCg9lm+McYYY4yJhGAGRvoEmP1hGJcX2/EkOVVl6ePb2FJaG+uhmCSUNAGMWsbTouluIc/IGO2n2n1v/jWePqofKfto3xhjjDHGhFMwAyMjJ7bjMADUtXRQtKGUtbuOxnooJgklzRQSkB4rkYwNwVDABXe+xFmzJsZ0LAPpF2oZy8EXY4wxxpgEJCKXAb8CPMAfVfXOPq/fAyxyd7OAyaqaG91RDsLX4BTvTM2M9UgMUNXUBnQXqjcmmpIogEHUAxjtnQE6AwGy0gb/Ng81haS8rpXyOFuCs++Ix+I0GGOMMcaYRCciHuC3wMeBMmCTiKxU1feC56jqd3qc/03gzKgPdDBtDc70Ecv0jQuVjU4Ao8HXEeORmGSUNFNIwFmJJBIBjIHqP1z129dZcNuasF8v0sb0FBZjjDHGGNPTucBeVd2vqu3AQ8CVg5z/OeBvURnZcPkaICN+VxFMNpVuBkaTZWCYGEiqAEa5FjBRmsjCF5H++2YhvFfhFBwaK/PDIrn6SbLGRHwd/n7ZMx3+AI0WsTbGGGNMdEwHDvbYL3OP9SMis4DZwEsDvL5ERDaLyObKysqwD3RAvnqrfxFHghkYNoXExEKSBTC6l1KNhIEyMb58/6ZR9hddlnkRfg9tLO137J//7y1O+8lzMRhNf/YzN8aY0SkqgsJCSEmB/HznIQJeb+htYaHTxpgYCPUJ1UB/AFwHPKKq/lAvqupyVV2oqgsLCgrCNsAhBaeQmLhQ1eSsRtjYZh/ImehLqgBGWVcAI4oR41GIZCbEsVBVtpfXj6ptnH5JEdcZ6P/3wQs7Y5+Rk6w/D2OMCYeiIliyBEpKnAzD6mrnAeD3h96WlDhtLIhhYqAMOL7H/gzg0ADnXke8TR8BdwqJZWDEC8vAMLGUVAGMYAbGjDBnYDS3hQxS99LpD4T1mpEwVOBk5buHuPx/XuOZrRVdx4Kf4A+UfWKMMcYkmqVLoaVl5O1aWpy2xkTZJmC+iMwWkTScIMXKvieJyInAROCNKI9vaJaBEVeqrAaGiaGkCmAcJZd29YR9Conf/ZR9sJU4zvzp8wO+1jduEK9p/XuPNgGwr7Jp4GCHfbRvjDEmwZX2nx0YlbbGjIaqdgLfANYAO4EVqrpDRO4QkSt6nPo54CGNxz9EffVWxDOOWAaGiaWkWkZVSaFC88IewOgYRnbFYP/BB/o1Ea+hANX4DbKYkbGfojHGjNzMmc6UkNG2NSbaVHUVsKrPsdv67P8kmmMatkAA2hptCkkcCWZgtPsD+Dr8ZKR6Yjwik0ySKgMDnGkk4Q5ghKpzMFqDZXFEy0BfTexHFjnlda3sOdIY62EYY4wZA5Ytg6yskbfLynLaGmNGoL0RUJtCEicCAaW6uZ0JGc7n4E1tloVhossCGGEQOMZshJ6zLmJZS2K4AQpFB55CMkYzMy648yU+fs+6iF4jHr81iRyUMsaYSFm8GJYvh1mznN/heXnOA8DjCb2dNctps3hx9MdrzJjma3C2NoUkLtS2tOMPKLMLsgGbRmKiL+kCGMU6lalSywSaw9ZnokynGPKrGKS+RTxkjpjw+69VO1mz43Csh2GMMXFn8WIoLnay26uqnIcqdHaG3hYXW/DCmFFpcwMYloERFyrd6SNz8scB0OizpVRNdCVdAONtnQfAGSl7w9bn/euLAWjtGHo1kpEYC2GRsTBG083X4aehxy+aoX5+967bz9ceeCuygzLGGGOMGYiv3tmOogZGTXO7TXEIs6rGdqA7gGErkZhoS7oAxruBufhVODtlT9j6LK9tDVtfMPRyppHQ3NZJR+fwlnpV7T/1wJZRHVo8LNDysV++wgd+8pxlzBhjjDFmbDiGKSSf/8Ob/PSp98I8oORW2eQDYHaBE8BosACGibKkWoUEoJlMdutMzpL3w9ZnIty6n/LjNUOe0/OWd8CvOR7u0uNUPMw0KgtzsM0YY4wxJqK6ppCMLAOjsrGNXYcbKRifHoFBJa9gBsZsm0JiYiRiGRgi8icROSoi2wd4XUTk1yKyV0S2ishZkRpLX28F5nNGyj5SGF7GwVDi4cY0mpLsyzXGGGOMMbHSNYVkZBkYb5XUAlDfajfY4VTZ1Ea6N4XjcjIBK+Jpoi+SU0juBy4b5PVPAPPdxxLgfyM4ll7eCpzAeGnlBCmL1iUH1TNpQZDYFwUdxuUtz8IYY4wxxkRcMIAxwiKeb5XUABbACLeqxjYKxqeTbcuomhiJWABDVdcBNYOcciXwV3W8CeSKyLRIjaenLTofgLNTwjeN5FgMFK+wIEFisTohxhhjjDEj1NYAnnRIzRhRs80JmoHR2u7nzf3VMbt+ZVMb+dnppHpSyEhNCfsUkgNVzZRUh2+1yHi1vbyeinqb2j0asSziOR042GO/zD3Wj4gsEZHNIrK5srLymC9cqpOp1AmcFaZCnh3+7qkoNxVtOaa+4vkmN1R5i1gnixhjjDHGmATmaxjx9BFfh5/t5fV4UoSG1g4CgcT5g/Xxt8u5bvmbHGnwxeT6lW4GBsD4jNSwTyG55dGt3Pzo1rD2GW/qWtq59t43+OVz8fFh+lgTywBGqASDkO8uqrpcVReq6sKCgoKwXHpL4ISwFfJcv687CvrMtoqRj2aAVIu4fatVHbhWp0U0BmQrf4TX6gOr2V4VssSOMcYYYxJFW8OIp49sK6+nw6+cPWsiAYXm9sSZ5lDZ2AZAWW1LTK5f5WZgAIzP8NIY5ikkRxvbOFiT2JkJD7xRQku7n8MxCkKNdbEMYJQBx/fYnwEcitbFtwTmMzvlCHOlPFqXHLZYLKM6HD1vwIcbp7B4hgkqri+mqb0p5Gu1vtoR9VXWWMYtr97C7W/cHvuaMcYYM4SiIigshJQUyM93HiLg9YbeFhY6bYwxODUwMka2AsnmYufviotPmgwk1jSS4NdyqC76N7/+gFLT3N6dgZHuDXsGRk1zO4cbfPgTKGumJ1+Hn/vXFwPdwSgzMrEMYKwEvuiuRnI+UK+qI09fGKXnAgtp1EyeSvsRN3hWI2FakWQ4Nh6owdfhj1j/T289ROEtz1DT3B6xawwoToMv8eBYpwepKr/Y+FtuffXWMI0oegGm8qZyrn7qav75hX8moL3/r/11x1+58OELefbAs8Pu70/b/4Rf/eyq2cXWqsROMzTGjG1FRbBkCZSUOO+51dXOA8DvD70tKXHaWBDDGEY1heStkhrm5I+jMM9Z6jMxAxjRz1Kobm4joFCQnQYEp5CE73vb6Q/Q4OvAH1CONiZmdsKjW8qobm5n3uRsqposgDEakVxG9W/AG8CJIlImIl8Vka+LyNfdU1YB+4G9wB+Af4nUWEI5oNO4pO3nbAicxE9S/8r1nhcifs2DNS3sq2zi2nvf4PandoQ8JxzTDP78ejEA+ytDf9p9rJThxynCHc+oa2mn8JZnuP/1A+HtONqkg2+++E1eLXt12E0eeLOE+95+kueKn+8XBBjx5aMcZ7pr0120+9t5p/Idntz7ZNdx74R3uWvzXXjFy92b76alo3c6ZEAD/Y4dbj7ME3uf4PI5lzMudRwrdq+IytcQDSJymYjsdpeXviXE6+ki8rD7+gYRKezx2q3u8d0iculQfYrIbLePPW6fae7xC0Vki4h0isjVPc4/Q0TeEJEd7tLXn+3x2v0ickBE3nEfZ4T/u2PM2LR0KbSMItO7pcVpa0zSG+EUElXlrZJazp41kZzMVCCxAhgNbsCgoj7yN/hvl9by2p6qrv2qRufD0e4aGF6awpiBUd/a0fXhWt8ME1Xl/94s4egYnnbhDyh/WLefD8zI4ZOnTqW6uZ1Of/Q+RE8UkVyF5HOqOk1VU1V1hqrep6q/V9Xfu6+rqt6kqnNV9TRV3RypsQykgjy+3PEDtgcK+UfP8G8kR+vDP19LXYvzprPrcOOw2mwuHnghF1+HPyJvyANlCoQs4jnMrIJwpfkH38we2nRwiDPjW+rE9bxc9jJ/3vHnYbdZv+8IKWlHaQ+0cagparOtjtnr5a/zYumLfPPMb3Lm5DO55617ONJ8hNSJ68mYtoKzJp/FvR+/l6MtR/njtj/2anv7G7fz6Sc+TUeg+9/5X3b8BVXlG2d+g8vnXM7qA6up89VF+8sKOxHxAL/FWWJ6AfA5EVnQ57SvArWqOg+4B/iZ23YBcB1wCs7y1b8TEc8Qff4MuEdV5wO1bt8ApcANwIN9rt0CfFFVg9f4bxHJ7fH6v6nqGe7jnWP4VhiTUEpLY9PWmIQxwgyMIw1t1LZ08IEZOUzIdJb6bEigAEY0MzDufm43P3piW9d+8JpTJjgrwmSHeQpJbUt39njfFToO1fv40RPbufPZXWG7XrTtrGiguLqFL36wkILx6ahCTUsMMubHuFhOIYkq8TQhnlAZCcLT/vM5I2UfM+RoFEYy9I18z5v9a+99Y8Dzrvrt65x++3NhGdVIqI4wUySljT+V3shT+56K3KAiKBBQfvPSHurC9AYjnmbS89eS7kln0+FNVDQNb+ZUKxVIipNfvL9+f1jGMhzirR91wKTD38GdG+9k1oRZfOmUL7H0vKU0tDdw6aOXkjF1JX7f8fxq0a84d9q5fGrOp/jLjr9Q2uD8xb6ubB2P7XmMoy1HefPQm4BTK+OR9x/h8rmXMz17OteccA3tgXae3PfkYMMYK84F9qrqflVtBx7CWW66pyuBv7jPHwE+Kk7RnCuBh1S1TVUP4GS2nTtQn26bi90+cPu8CkBVi1V1K/SeV6eq76vqHvf5IeAoEI6qysYktJkzY9PWmIThq4eM3KHPcx1yb3xnTMzqysBoaE2cIp7BYMyhKCzBeajOx8HaVto7nT8J9lc591Jz8rOB8E8hqW3p7qtvgCa4v/LdQzGZPhMOwQ+mPzg3r6sQqtXBGLnkCGCk+Mgq/A1Zs38dMojxdOB8AD6VsiFqQ2pt9/ODR96lvrVjwGDAUCGC4WZxhEvP8YyknkPqhHdp8lexoSJ6399wen1fFXc/9z5LnwjPihdp+S9CSjt3X3Q3AKsOrBrw3P31++kMOL90m7U762R/XXgCGMP5JZA5vYgbn7txVNNWXjz4IsUNxXzv7O+R5knjxEkn8u2zvs2i4xfRUryE1pKvkZPuFOb6zlnfIdWTyhee/QJritdw+xu3My93HuNTx7O6eDUAj+99HJ/fx5cWfAmAEyedyBkFZ7Bi94pjnlYTB4aztHTXOaraCdQDeYO0Heh4HlDn9jHQtQYkIucCacC+HoeXuVNL7hGR9AHahXVJbGPGgmXLICtr5O2yspy2xiQ1fyd0NI9oCknw75ppuRkJOYUk+LVURLiIp6pyqK4Vf0AprWkGYN/RZvKz08jJcr6v4zO8NLf7w1Zws2f9vr5TSII/186A8qfXxuZU8s0ltUzLyWB6bmbXNJyqJsvAGKmkCGCkT34GSa1HPC1kHLeCPh8sUqaTeScwl8s9A2c7hEswuWLX4UZWbC7j9pU72FzSPU3kWAs9xqPU3E0A7KvbN8SZ8anDnZvWMoplolr9DYi3oWu/rLGM1Ilv0lF3Dh85/iOcXnA6T+9/OuQUm22V27jqiat49P1HAWjWMjTgYZw3h3314ftebi+vH/A1Sa3Bk1XKwcaDrD+0vtdrVa1VrC9f369GRU9/3/13pmdP56LjL+o6dsOpN3DPonvwt86hZ1hsyrgpPPCJB5iUMYnvv/J9qlqr+OkFP2XRzEWsLV2Lr9PHit0rWDhlIfMmzutqd91J11HaWNpvfGPQcJaWHuiccB0fkohMAx4AvqzaFTW6FTgJOAeYBNwcqm34l8Q2Jv4tXgzLl8OsWc5UzLw85wHg8YTezprltFm8OPrjNSYutLdAa61T/wJGNIUkeGM/LSeT7HQvnhRJqABGMAOjurk9oosC1DS30+ZmXuw96gQw9lc1dWVfgBPAAGgK01KqwWzn7HRviAwM5+f68QVT+NvG0jH5Mw3WZgEsA+MYJHwAwzNuF2kTN9FefRFtRz6NN/t90vJe7nfeU/7zOS2lmFlyOKrje+ztcnwdoT85jnQoo70z0FUIaCQ6/AF+vDJ0EdK+6jpL8GQeJD0lm331+8b0p+R9fx776vbx8sGXBz5flScP30HmzD8AiiqsLl6NSID2qkUAXD7ncvbW7eX92vf7tb93670oyoulLwLQTCmB9ilMySgM6xSSprZObn1sGy0h1kj3jneyTsanjufh3Q8DsLtmN1c8cQWLViziay98jT9s+0PIfg/UH2Dj4Y1cfcLVpMjw3mrmT5zP3z71N2487UZ+eO4POTX/VC4rvIzGjkbu2nQX5U3lfPakz/Zqc8msS8jLyOPBnX1LNow5w1lauuscEfECOUDNIG0HOl4F5Lp9DHStfkRkAvAM8CNVfTN4XFUr3LpGbcCfcaauGGNcixdDcTEEAlBV5TxUobMz9La42IIXJsk9+wO490JocT/kG0EGRnldK9npXiZkeBERJmR4I3aze6iulX9/Ynuvv6F+u3Zvr8KXfakqdz67i7dLR7aEPDh/gze3+5mV56R1RbKQZ88MiODUkf2VzcwpGNd1PBjACNc0kppmp58F0yb0+9oO1bWSk5nKdz52As3tfh7c0L9IUGu7n+88/A5fuX9Tr8fDmyJXUKi+tYNvPLiFr9y/iRv/spmtZaHrspXXtVJR72OhG8DozsCIfQCjvrWDpY9vC+t0oEhK6ABGfVs9GdMew++bQnvVx+ioO5eOhg+QVvAipPT+x7LK70wj+XRKZLMw3i4dutig9KiWOdogRkp6BTtqNg16zvV/3MAHfjLyGhqrtlfw8u7hpX/v972EBjycnfMZWjtbqWiO2kq5YTPQFJ+7N9/N917+Hm3+0G88r5a/ypG2PXjSK0nJKAfgpdKX8LfOQDuduZyXFl6KV7z96oO8V/0er5S9wqSMSWw6somm9iaa9CAB3zQmZ8xkf93+sBVG/e3avfxtYyl/faOk32upE7bhb53OdSddx7qydeyp3cP3XvkeTe1NfO/s73F6wemsPrA65Fgeef8RvOLlqnlXDXr9vk0zvBl866xvdQUqzj/ufHLSc1jx/gryM/P56PEf7T1GTyrXnHgNr5W/1lU/Y4zaBMx3VwdJwynKubLPOSuBL7nPrwZeUuebvxK4zl2lZDYwH9g4UJ9um7VuH7h9DlpIxG3/OPBXVf17n9emuVvBqaURnvlWxhhjklP5W1BXCm+5xc4zcobdtKK+lWk5GV1/T0/ITI1YAGPNjsM88GYJD210ZmvuOFTPXWt287dBbphfeb+S37+yjwfe7P9311CC2RcnT3UCOhURrAXRs8bG/spm6lraqW5u7xPAcKaShKuQZ11LO+neFOZOHteviGfw57rguAmcctwE1u/rHyR6u7SWx98up6S6mcrGNiob29hcXMP960f+vR6uN/ZV8/TWCspqW1i7+yhPvRv686Bg/YuFhZMAGJfuJTPVExcZGK+8X0nRhlJe31sd66EMS0IHMMaljqOj9oP4Kq4F9QJCR+25iPjxZBb3OreCPN4MnMw1nlcQIpclsG7P4Df+4VhGFSB96hP8cfedg56zcZAVTkIJxlUCPb49g91D+zp9lLS/SmfjqUzPPAWAvbV7R3TNeNXS0cLGio20B9p59+i7/V5XVX7/7u/J9uSh6iF1wrtUtR5lW9U2OhtP6TpvYsZEPjzjwzy9/+leK23c++69jE8bz+3/cDudgU6e2v8UHdTj902jIH0mTR1NVLYO/G+pI9DBmuI1PL7n8SG/luDPsO/P8lDTITyZB+lsPI2rT7gaVeWG1TdQ1ljGXRfdxQ2n3sA/zv9HyprKeK/mvV5tfZ0+ntz3JBfPvJj8zPwhxzCY1JRUPjbzYwD8v/n/j1RPar9zrjnhGjzi4aHdDx3TtWLJrUfxDWANsBNYoao7ROQOEbnCPe0+IE9E9gLfBW5x2+4AVgDvAauBm1TVP1Cfbl83A991+8pz+0ZEzhGRMuAa4F4RCZ5/LXAhcEOI5VKLRGQbsA3IB/4j7N8gY4wxycHfAVV7nOcb3SzPEUwhOVTnY1puZtd+TmbqqDKOh2N/pTO14r7XDtDhD7B8nZMhO1hgIXjOWyUjz8AIBmJOmjYecD7Vj5TgFI4TpmSzv7KJfe7X2nMKSXZ6eKeQ1DS3MzErjWk5mVQ19Z4iU17nY7r7c503Obvre9/Tvirn2P/deB5PffNDPPXND3HpKVPDVog/lGB2ymP/cgEzJ2VxaICsmC0ltWSleThp6viuYwXj0+MiA2N/pZthUxVqwYv4k9ABDG+Kl/bqRQR83bXp/K2z0IAH77j+N9K/SfsAvvRaLkzZ1u+1aDnaeOypYJ004Mkspa69CmTkb9hh+mCf9YfW06HNdNQvJC/VyWLfW5cYAYwNFRtoDzhvhhsPbwTgaIOPpY9vo8MfYP2h9Wyr2sa5Ez+Lv+kEvBO2sv7wWgA6Gk/t1ddn5n2Gal81r5Y5S/nurtnNSwdf4vqTr+dD0z/EhLQJ/Hm78wlEoM3JwICBa4o8sfcJLn3kUr7/yve5bf1tlDWW9TkjwB/2fp/0yQMXDwV4vuR5Z7wNp3Fc9nFcOONCGtob+Nez/pWzp5wNwMUzL8YrXtYUrwHg/dr3+dFrP+Lyxy+nvq2ea0+8dsjv5XBcc+I1zJ84n2tPCN3f5KzJfHzWx3lizxOD1uSId6q6SlVPcJeXXuYeu01VV7rPfap6jarOU9VzVXV/j7bL3HYnquqzg/XpHt/v9jHP7bPNPb7JXfp6nKrmucumoqr/5y6LfYb2WS5VVS92l8M+VVWvV9Wx8RvQGGNM/KneB4EOmPMRCGa5jmAKSUV9K9NzM7r2cyKYgbG/qomM1BTK61pZvm4/T2+tQGTgqR3byupZv6+amZOyKKluGfGn710BDPcmOJJTSCrqfaR5U1hYOIl9lc1dN7lzJ/evgRGuqQe1Le1MHJfGcW6g4nCPr6+ivpVp7s91bkE2h+pbaW3vXQNkf2UTWWkepk7o/vlPGpfWqzhouO072syUCelkp3uZlpMxYPBqc0ktZxyfi9fTffudn50WFxkYwWBQqKBQPEroAEZImoq/dRaecX1v/vy8c9zb/Ht+Add7no/J0ADW7DhyzH00e7Yi4kQhJHXoKStBKemHyZrzSyp9fW94HT2ntgzH6+Wv4yUDf/Ns0j3ZTM6azC9fXsd1yyNfLBWg3R+5N6tXyl5hXOo4Tp50clcA40dPbKdoQylrdx1l+dblTB03lQXjL6aj4XRSUuv5+76/MCdnDtreu3jhh2d8mPzMfB7f8ziqys83/Zyc9BwWn7wYb4qXD03/UNfUG7+vO4ARqg7GW0fe4sfrf0yq5tN66B8BuoILv3/F+TfvnbCNkubtpOZuAGkfcJrSmuI1+Funox1Otbmbz7mZW869hRtOuaHrnJz0HM4/7nyeK36OI81HWPLcEl4qfYnTC07nZx/+GedODU8phFPyTuGxKx5jyrgpA57z+ZM/z6KZi2jpHLsBDGOMMSbpHXWzOi++DcZPc54PMwPD1+GnqqmdaTndGRiRnEKy72gznzx1GnMLxnHXmt0I8I9nzeBIg49Of/+M7nvX7WN8upefXuV8mPVWyciyoRvcqRoF49PJz07rN80inA7VtXJcTgZzC7Kpb+1gc3EtqR7h+Ind39twTyGpbelg0rhUjstxAhDBaSwt7Z3UtXR0/VznFIxDFQ5U9b7h3lfZzOz8cb3uWXKz0mjrDPQLdoRLz8Kmx+Vm9ls9BZwMlZ0VDV31L4LiJQNjnxucCm7jXfIFMAB/81xS0ivA0/2P3pN1APH42Jnh5ZTUrcyQ2C3xN9raBnUt7VTUt9LkfQdV5z9uSurAb4wpGeWkpHfP00rN3Ygn/SjPH3qQ5rZOLv7Fy2wZRoGhX2/5NWW+d/p9Da8fep3JqacCTnR2Xu48NPUIb+4f2Zv1aFQ0VfCRhz/CLzb/ouvY9qrtx7RSRfdUC2Vd2Tr70uYzAAAgAElEQVT+4bh/4ILpF7CtchstHS0EV5BqaK9my9EtfPbEz+KRVDobT0YDqdS11fDRmR/t1683xcsVc6/g1fJXeWj3Q2w8vJGbzripa3nRi2Y4K3ikMQkCWWR7JzI+bXy/pVTr2+q5ed3NzMiewaTGm+isP4fZ2Qu6liC989ldgJ/0/OfJ9IxHPG14x4cuxlrRVOFOd+nOFjl+wvEsPnlxv4KclxZeSnlTOTesvoGWzhb++om/8ouP/IJPzvnksIJe4SpWe8bkM1j2oWXHPGXFGGOMMTF0dCdICkw5BS74FqRmwbjhrVwV/MR+Wk73J/ATMlK7akeEU3NbJ4cbfMydnM3XLpwLwBWnH8fZsyYSUDjS55P1Q3WtrNpWwefPn8n5cyaR5k1hc/HIppEEAzETMlI5LjeT8ggupXqorpXjcjO7al68uOsIMydl9cogmNCVgTG8AMaDG0r5r2d3Dvh6bXM7uVndGRjBYEBwG5xCEgwY9J3ysL+yiTkF2b2OTRrnBFlqjnEayc2PbOWRt3p/yKuqvQqbHpeTwdHG7uDV01sPsejul7nkl68QUDjbrX8RlJ+d3pWBsWF/NR/9xct85K61IR/3PN+/4H9PSx/fNqrlZVW1KxC0v7L5mGrsbdhfzT/9dXPYltUdSMIHME6f0b/oT2fLPEQUb1b3DaB3/E5UnW/HC+My+bznxaiNcSAjrYZx3n++yAfvfJaWlJ1dN54pqc4b496jjb2ilM+XPE9W4e/InPlnN1MhgHfCNlRT2Fz1Amv37WJ/ZSPfef6nrNzXt45gN7+3jD9s+wPvND7S63hpYynlTeVMTftA17F5ufNIST9C32Vsj9W2svp+1Z6LdhbR2NHI/Tvu5y87/sLjex7nC6u+wLfXfpvWzhFGq/v8IHbW7KSytZKLZlzEuVPPpVM72XJ0S9fr7zc4zy847gLngKbT2XQyQMgABjjTSPzq5z83/CfzcudxzQnXdL12wfQL8IiHbHGm4YgIc3Lm9MrAUFVue/02qn3V/Pyin+PB+cV9dv4idtXs4kC984bmzdlCSnoVnzn+OwTac0nN2UIowZVP+k53CWXR8Yvwpngpaypj2YeW9Vri1BhjjDFmUAE/PHszVLnTjI++B5PmQmoGnPd1+O57kD5+8D5cwU/sp/etgdHaGbbi50HBv6vn5I/jqjOn8/WL5vLdS07oCp70XQZ0Z0UDAYVLT5lKutfD6TNyeGuEK5EEAxg5mamDTlcIh4p6H9NyMpnrBguqmtr7BQeyRxjAeGxLGQ9uKB3wZ1Hb0s6krDSmut/D4NcXzDQJfm9n5zsBg55THnwdfsrrWpmTP65nl+RmpTl9H8M0El+HnxVvHeTOZ3f1qstR09xOfWtH1/fluNzMXsGrZ7cdprqpjXNmT+KLH5zF+XN6BzAKxqdT29JBhz/AyncPcajOx+nH5/Z7pKQIRYN837aW1VG0oZRfPLd7xNlGhxt8tLT7mZM/jvrWjmOabvPoljKef+9IRDODIAkCGBee0D9iG2idgfrT8GQFp5Eo3uyd+JtOwO+byiPjCviM5zUiv5BpeLV1BvBmv49KJx215+GVVMQNYHzsl+tYdPfLADy17ym+/8r30Y4cUryNrC5ejSfrACneRtqOfgKRFNaUPUT6lKep8T5P0c6iAa/ZnvU6AEfadyGe7ihoMNNhaurpXcfm5c5DUjqR1BoktYrUiW+wsWIjDe0Nx/R1f/o3r3H9fRu69pvam3h0z6NcMusSLpl1CXdvvpvb1t9GYU4hrZ2tXbUmBrPlyBZuXnczGyo29HuzWFe2DkH40PQPccbkM0hNSWVjxUaC/152NWxmUsYkTpx0Yleb9qqLuXbel1mQtyDk9QpzCjlr8lkA3HLuLXhTvF2v5aTncMu5tzAj5ZPgXmVu7txeAYyHdz/MSwdf4jtnfYdT8rqLhJ6VfxGCOMu3eutIL3gBf+vxnDzhg3TUn4Vn3F78Kf1/eT5f8ryTMdM+9CceOek5fOOMb7D0vKV8fNbHhzzfGGOMMaZLfRls+D2s/7Wzf3QnTHY++EEEMicO3LaP4Cf1fYt4tvsD+DrC+wHavh41IdK8KdzyiZOYMTGrK3jSN4ARLO4YfP2sWRPZXl7f64Z4KMFMkgmZqUzLyeRQXWvYAzMAnf4ARxp8TM/NYPrETNK8zi3j3D4BjMxUD54UGXYNjP1VzTT6Oqlq6n+T7A8oda0dTMxKJSPVQ964tK6AVPB7GczMyEzzMD03s6suB0BxdTOqvWt0gFMDA5zgyGiVVLeg6ix5+vjb5b2+HoC5bgbGtD4/+32VTSwsnMSvrjuTO648lXSvp1e/waVUq5vaeaukloWFE/nVdWf2e3zlgtlUNbVRWhN6mvS96/aTkZoy4PKygwkGgT6+YEqvr2k0NruFaSNZmwWSIIAROovBg791dlcdjJT0I6Sk1dDZdDKdjadxIKMTj7eeaUR+qsNghvt29NyOw3zq185NuXf8e6ToOPwts8nLmEJKWu+voTPQyR1v3MEZBWfQvP9b+Nsm88B7D+Cd8C4aSKOj9jzOzb+EVypWkjZpPV7N4f2a93tlLXS9Uab4aM/YzGn5pwGKd3z3ShTry9czI3sG2Z6pXcfm5jrpdd7s98kqvJeMqU/y1ee+yoUPXchNL97E8yXPE9DuXy7/8eZ/cONzN47gO+Z4fO/jNHU0ccMpN/CfH/5PLiu8jC8t+BIPX/4wkzIm8VzJ0EvHPrjrQVYdWMWNz93IL7d/oys4o6q8UPICp+WfRl5mHpneTE4vOJ0Nh4MBlAC769/iyJGZrN3VPQ0p0DaV60/82qBTKn5w7g9Yet5Szpt2Xr/XrjvpOiamnNS1f9Kkk6jx1fDrLb9md81u7t58NxdMv4DrF1zvjNP915Obls/ZU87m77v/zrg5v0JSfPiOXI6I0FF/NiJKW/rGXteqaq3i7aP/n70zD4+qOv/458yeyb4RQshCkH0TgooioAW1atWK1FKpStWittVa21/dq63S2trWrlpxqUtxtyrUHRFRVDYRZIeEJJCQkH2dfe7vjzt3MpPMJJNkkpBwPs8zzyT3nnPuuXcmkznf+77fd1u3xIhrp1zL4vGLI24vkUgkkvCsXAl5eerazWDo/DkvT20vkQxaHE3q855VYG+EukMwLPQNn67Q7tgHppAkxqgpBNH2wSiqakEnIDfVGrRdW8S2X8SV19sw6gXpceqidWZuCi6Pwo4jDREfs9HmwmzQYTGqC/gWp8fvixFNKpsceBX1XPQ6wahUdYEeWEIV1KjgeIshogiMuhan/+5+KK+FRpsLRYFkn+AQ6CdRXm9HCPyRGdpcAhfbRVVtETGBJPsiMHoTWaDNNyXWxOPri/D6UiQKj/lELC0CIyD6xutVUzPazyeQNN97oaiqmX2VTczMTQnZbmaeKuKFSjkqrWnlna+PcvUZecwZk8ZTGw7hcEcuimki0AJNwOihD0Zti9P/GrQX76LNkBcwwuFuGY3eXIUupgRDnJqL5W6egLtxCgj40BrDdN2BqB/3kwMdaxa3J3CRG4mo+rOXvmJXeSOgoI/dT6xnMqAnzZLpTyHRONx0GLvHzmVjLwPFjKtmDntr92JM2oq7aQIoJuZnLsagM+JqmMZw1xLcipvdNZo4oaAYagAFY+I20Dm549Q7iNMPwxC3GwG4PC42VWxidtbsoPlrAoY5YzVCOGktvoF/LfgXSyctZW/tXm5dd6u/2kZ5czmv7n+VjUc3Bhy7a9xeNyv3rGTGsBlMSZ+CWW/moXkP8YtTfoFJb2JBzgLWH1nfaRqJoihsrtjMeXnncc+sezjcegBT2hoU4JOyT9hXt0+9fj5OzTyVPTV7cFCLznyUVk8D7pYxPP1ZccTzBtWosjMRIPCtsHDMQhaOWcjjXz/O9976HrHGWB6Y/UAHfwqA80edT5WtCq8zhZZDN+G15arjuVJxt+Zhj/kchNtfJndt6VoUlLDpLtGmL+4cSCQSyWBl5UpYtgxKStTfPZ7On0tK1PZSxJAMWpy+BZOtDjatAMULw8Z33icM5Q12UmNNWIxtd7r7SsAorGpmZLK1w131OLOBBIuhwyLuaL2NjAQLOp36havAZ+i4pRtGng02Fwm+89EqcpTV2fwL6mhR3k4I0oSL0ekdF+NxZkNEZVQD/SpCVbvQPCq0iInMRIs/FaG83saweDPGAP+N0elxFB5r9n+P1MSE9iJLslW9XvWtPX/9tUX97d8cT1F1Cx/sqfSdUwsmg84fGZIZ4N1RVm/D4fZ2SLsJRIvAeH93JYrSJlS0Z+yweOItBn+EQyBPfFqEXie4ZvYorp87mqomB69/WYbXq0T0HbuwqoVYk57p2UmYDDp/udzuElgWOJSRaTQ5cQWMxml4XYlYcx/DmPIpHttIFHcCXucwvI5hvBMby8m60GUqj1eEsQ6doYUYryoUpFsy/SkkGprx4+hEtY2r8WRSLCkI4cHdqPpVpJpH8PtTX8ZevpgYbz4AO6p2AGBM2kzzsPuxjvobppT16FwjmZw2mVzLqehjD+Dy2viq6ita3a2cMeKMoGPHGmPxupIAga1sCR5bHrOzZnNLwS28f9n7zBs5j8e/fpxaey3P7X4OgcCsN/Pa/te6PHeduYIb19zIvJfmUdZcxpUTrwzZ7ry88/xpJFsqtnD56svZW7s3qM3B+oPU2muZPWI2l4+7nDPSL8CYvAkHlTzy1SNkxWVx0eiL/O2/lf8tLAYLh/g3hjjVYMfTMqbLOfcGs97Mr8/4Nb854zekxqTyuzN/F2ReKQJijxaOWchfz/4rrSU3oriCDS6d1Wfj1ddgSv3YLzatKVlDTnwOY5PH9uk5SCQSiaQjd90Frd0sptTaqvaTSAYlWgQGtKWR9DACo7y+rdSmRkKMmpYbKGAca7Rz2m/XdPBQC6SwqplTl69hZ1noCIlA88b2hKpGUV5vZ0RAdZSUWBP56bFsK428YmCDzeUXZLRUlAv+9gn5d77d7dSBztAEDO0YJ/nSMjTzzEASLEaONbWd66cHqjntt2s41hh8/oELY00QOFzbymm/XcPu8ka/R4XmWTEiKYYjdTbsLo/fjyOQ/PRYWpwejvn8JoqqW8hMtGA1GYLaJcYYEaJ7ERibDtUy+8G1/j5FVS0MT7CwcEYW2Skx/OvjQp+BZzOjUmPR+0QpTbw62mDzR4eEe48A/mic93dVoBMwLTspZDudTjAjJ7lD1Zpmh5uXtxzm2ydnkZFgYfZJqUzMTOD2/35N/p1vM+P+D6hs7FxMKKxqZlR6LAa9jrxUa48jMLaU1GLUC2JNehmB0VsClbpAFHciLUU/xd00GZ2hBXdTm2+As34m22LMWKz7/NuEoZ7jxRNjx5HQH3R6y2EALJ5RAKRZhqMztIBoc0EubFBFmVGJahsUI9dMvgavKwl3S5tnQ7xRFRoMJDAybiQ7qnaoIasJ2xCeJMCLzlSHqWUOQghyLacidB6+an6Vuz+9mxhDTMgSmo5j52MvuwJPS/DiWK/Tc+vMW7G77fx+0+957cBrXJB/Aefmnstbh96i1RX4bc5DVdxfeLvobf8Wc/q7bD+2nXNyz+GhuQ+FjR4oyCggxZLCYzse44Y1N7Cndg/P7X4uqI1WFlVL5fjmyCtB0VOo+xu7anZx/dTrMeqM/vbZ8dn8vODnNLALU+pHJOhyUNyR1yuPlMDkk9+9s4cV6wu5dMylfLDoA87IOiNsP4POwDdyvgFK2we6FuXjaRmHyT4dU+pH1LvK+bLySzZXbGZB7oJul82VSCQSSe8p7eEapKf9JJIBRxMwsgrA3gB6E6Tk92ioow22IJEA2iIwAiuRfFZYQ2Wjg7+sCV/ZYWtxHceaHDyy7mCHfW3pAaHvrgdGD2iUN9gY0U5cyUuN7dZiL1DAmDoyiXsvmsjPFowl1qRnV3nkqShd0d5L5KrT8/jX9wv86R2BnDUunc8Kayj2Ldj/smY/lY0OPi+qCWpXVNWCUS8YMyzOv7j/eH8VlY0O1u6tpM4XIZHiEzDOmZhBq9PDG9vKQl477dpr6R1qBZKOYoFBryPBYuyWB8ZnhdWU1dvY6DuHwmpVrDLodfxwTj7bSuvZUlIXUsTSxCtNCOhMwNBSSMob7EzITCDObAjbdmZuMvsrm2kIiCT5qrQeu8vLt6aNANTv93/+7jR+tmAsN33jJBpsLv69objTcy2qansf56fFhYyOiYStxXVMGpFIdopVmnj2loxES/idXiv2su/RcujHOGvm+De76k4n3mXkf2nNGHBiTPqCuDEPYkzaGH6sfmR/ZbAy1uKra6yPOYziNWBWsgBIs6i1s3WmtiiMooYiMmMzsRrb8vWumngVLQdv8y9w269bpw2bxvaq7bS4a9FbizG2zKL10C00H/wFRpu6yM8wj8frjuXr5jcQQvDkuU8SZ+r4oe5unBZUmjOQ/MR8Fo1dxNuH3sbmtrF00lIuG3sZLa4W3it+z99OH1uI03iAp3c9rW7Q2dHHHWDhmIXcd8Z9fHPUN8MuvvU6PefknsP+uv3kJ+ZzTu45fFDyAejaRJ5NRzeRFZfFiDj1wyDRlIqz9kwcooKsuCy+NfpbHca9fNzlJDIZoXcyzDAl5LGjhaIoPPZxEb99e2/4Nt0Q26zNC0HR80blXVz97tWkxqSyaMyiaExVIpFIJN0kJ6d/+0kkA44mYMy4Wn1OGwt6Y/j2nVBeb/eH82uESiHR0ja2lNR1uKutUehLeXh3Z4V/ca5R0WjH5vIwelhnERhtiziPV6Gy0R5kLgrqHfiqduVWOyNQwNDrBD+YPYqfLhhDRoIlqikyRxtsJFgM/gV1eryZb04eHrLt0jPyMOp0PP5JEVtLav1pDu39GoqqmslLjWVsRrx/ca+lHWwpqfNHYCT7yp6eMTqVSSMSWPFJkVrStZ0wpV17rfRnUVVLB5NRjZRYk18giQRtEb+lpM4faaGN/Z2CbJKtRv6x9iClta0djqm99kVVLcRbDP4oi1DEmPT+azwzt3OzWi3l6MuAyjVbSmoRAqbntEVujB+ewE8XjOHn547j/MmZrNxYEtZk1e7yUN5g85/D6GGxlNa24vJ0z/DW4fawo6yBmbnJIaOPos2QFzC6vocs8NqzgQDFSzGSVT2ZPWYT6cNfwDz8TRRFhyntIxDRN8qJFrqYw3jtWQjUXLw0i/pBI4xtH8xF9UXkJwWr2upiP/yVmpo2lSpbFTuaVqvlZ+3TAYHiSvOnKuiEHmf1Nxhjnc/LF73MlPQpvrG7dw43TruRWGMsc0fOZUzyGGYMm0FeQh6vHWhLIzEmbAfUcqb76/ar3hvCwzl5kZlOXjP5Gq6bch1PnfcUV028CpvbhiFeTZHxeD1srtzcwUjTWTOPGCWHWwtuDYq+0BBCkM+1uJvHkG2a02F/NOiriAidNxHHsQtxeltYNnUZq769iuyE7D45Vih++tJX/XYsiUQiOd5Zvhys1q7bBWK1qv0kkkGJ5oEx/kKIy4DMaZ23D0Oj3UWzwx1k4AlhBIziOgpyk0myGnns4yJCUVTVQkaCGYNOxxOfFnXYB6FTKkBdxNa1urD5bjJWNztweZQO4kpavImaFmfEHhaN9jYBI5CEGGNUBYxQQlA4hvlSK17deoQ/vLuPJKuRgtzkDn4NRb4ohvx0dZHscHv8QtKXJXXUaAKGLwJDCMH180ZTVNWC3eXtIP4MT7BgNekpqmqhqtlBk8Md1jAzyWrsVhlVza9jS0mdOrbd7Y+kiDHpuer0PD7eX4Xbq4SIwFCjbwqrmslPj+vy+7vmg1GQF9rAU+PknCT0OhHkNbG1pI5xGfEkWEILfsvm5tNkd/PipsMh9x+qViu3aOeQnxaH26uErXYSjp1ljTjdXmbmJTMiyeKvHtNXDHkBo6dJH+UNZzPe4aQ5eQ+KPZNTyqegMzZgTNwS1fmFomemhh70lnI8tpH+LelaBIbfB8PLoYZD5Cd2HpZ37TNbeGFTWyzqtHT1H8mOxv/hsQ9H584I2c9VN5szk35Egqnn6ROpMam8/K2X+e2ZvwXUD6/vjP0O26u2s+noJpweB4b4nVickzEIA6sLV2NI+BqvK9FXDaVrRsSN4KczfkqcKY5p6dPIS8jDmLQVgH11+2hyNnVMf/FaGOf5FefmnRt2XBNJ2A5fS4I+9G2w3ppVdqe/iEC6C8RVfyo/zHmBm6bfFBSd0x+8teNovx5PIpFIjmeWLIEVKyBX9VtGr+/8OTdXbb9kSf/OUyKJGloEhiUJrlsD50WmxjncnqASpO1LbWrEW4IFjCa7i32VTcwZk8aVs3L5YE9lyKoYhVXNnJydxMIZWbyy5QjbD9dz8FgTB481salYXXiHMrWENvPLo+3LgLYTV9LjzHi8SsTpDQ2tLhIsHdMMEmOMQSkyXq9CYVUzB481hU1RcXm8Yc03y+ttEQsYANfNycfh9rLxUC1XzcrlzJPS2FfR6L/z7/Z4KalpIT89jvz0WLyKKiIdrrUxITOBRrubzcW1mAw6rKY2U9QLJg9nZLI6j6x2KSRCCEalxbKrvIENB1Uvk3CGmSlWU9A1bugkGkOL5tAJ2FXWwO7yxg5jX3V6LmZfadn2x8xMVMWrPUcbGd1JBRKNtDhVsCnoIgLDajIwMTPBL/p4vArbSuvDGn+C6qkxKz+FpzYcwului6qwOT0cPNbEF74UGb+A4Xv+rLDG/14/eKyJ+i7en1oUU0FuCpmJMdQHiHd9wZAXMHpKmZLOTdUOxjXH8IsKHU85V5NkS8SUtg7o2yiMj/ZVoTNV8kTxdZhSPga0N5zH//OW4uBwN525EqFz4bG33TmPNyaheI1+AUMY67F77H4Dz85486ty/89jk8di1pvx4vYbffYlOQk5JJoT/b9fPu5ysuKyeHDzg2yt3oDQO4h1nMWZI89kVeEqDLEHcDdODlmBoyuEEHz7pG9jsBajM5ezoWwDQJCAEXnkQ/cEip4KGmV9ZIyjRe5IJBKJZGBZsgSKi9VKZG5358/FxVK8kAxyHE1gtILeAEk5ENP5Qk7j1pe386OVX/p/P1wbWsDQ6wTxZoNfwNhWWq9WfMhN4eoz8jDpdTy+PjjCwuXxUlrTSn56HNfNycfp8XLJPzew4M/rWfDn9fztwwMkWY3+u+ftGRFQjQLaSqq2N6JM8/Wvbu5awPB6FZoc7ogiMJ7acIj5f/qYBX9ezxkPrmVfRVOHPn98bx/nPbw+SAQCaHW6Kalp8Rt4RsJJw+I4Z2IGFqOOq87IY2ZeMl4Fv0Hp4TobLo9CflqsP2rllS1qVMAN89Qbq58erCbZagz63q15TgDkpHQUA8ZlxLPxUC0/e0mNzh6TEVrASI41+SMwtpXWMf3+9/m8sCZk24pGO61OD/PGpuP2KryxrQwILs+aGmfmu6dkY9CJkBEYAHWtrk79LzRGJlsZmRwT0fWemZfMttJ6Glpd7K1opNnhDlt6VeP6eaM52mBn9fa2td11z25mwZ/X8+vVuzHqVSEIYPSwOPQ6wT1v7PS/1xf8eT3nPLy+U0Hi4/1V5KVaSY83+8+/L6MwwjuFnPAIHLZRvODYjlGoZp7za+N5LesIxpTPcNXO7bMjl9a2YhnxEY3uKswZ76CP24fiicUQtw938xgE01n0r8+D+uhj1A8Bjy0bfJ/7Qgi8rmSESRU7dOZjAB1SSLrCqDcyMXUi245tw9U4FULcoA+3Fo90jf5FUQ2LV3zBWzefyaQRiUH7LAYLP5/5c25ddytPNv8RrzsOs3ssF4/OZ93hdQgduJt67jtx0eiLeHjrX4nN/xt/26YanKZb0wPOITrmrb1NASmpUcO5/vDuvi5ads8DQ1YxlUgkEolEMmA4miCEb1pXFFe3cPBYM063F5NBx7bSOgw6wcTMjpHACTFGGn3RAFtK6tAJNSQ/zmxgUcFIXtlyhFvPHcuweHXxdbi2VU0PSIvlpGFxvHL96X4RQmNUWmzY73aaX0N5uwiM9otUzR+hqsnBuOHxnZ5vk92NouAvoxpIYoyBRnvbDdaSmlbizAZ+ce5Y7lu9m51lDR3G31vRRFm9jTe2lbH41Lbo4Ve2HKHF6eGSk0d0Op/2PLRoKhWNdtLizJycnYROqNd67tj0AEPLOP+i/p2dFZgNOs6fnMn9cbupbnb600cCuXJWLpOzEpg4ouPrescFEzh7/DBATcVoLxBpJFuN/jKtO4404FXgkXUHOX10aoe2WnrQooJsPtpX5Z9n+9fuzgsmsKhgZIf0jUCvjs5KqAaO0+qM7Ob4dwqy+feGYv6zsYR4XyROV5EbZ41NZ1xGPCvWF7FwRhZfHa5nw8EarpyVy6mjUshKjvFXbkmwGHn5+llBHhZHG2z89u29vPrlEa6cldth/N3ljWw4WMP/nacWg/C/9+ttYT1JesvQFzB6sTjb5j2Jc/Rf8op7LqN0FZxvr+alpvFYMt4GrwlX/ayIxtGZyzEmbQHhQfFYcNbMA68VcGNK+wiPfSSe5gn+9sJQhyFhBzMSL2bDHgOWjNUoXjNeRzqG+N00ujqWfNLFHMbrtqK4glU4xZnij8DQmdSaxV2lkITi0pMuxdaawmZXWqcL3p4u0d/fpc7t88Iav4Cx/XA9r2w9zP2XTGZBzgJOHX4qmyo24W48HSH0zBs5mwRTAvUt4LH13L1smHUYtiNXoTNVc8+FEynIKOjxWJ3RWyHE24P+kaSSaGKHLDoikUgkEomk33E2g7nzxXsoGmwuHG4vu8obmJ6jei5MGpFAjKljRGlgisXWklrGD2+r+HDdnHye31TKM58V83/njQfaFrGjfeVDZ3bhT9CejEQzQrQJF+X1dqwmvb+kq0ZbBEbXRp6aABMqAiPRF4GhKApCCOpanQyLN7NkVi4PvLXH7+kQiJbesuKTIi6fmY1OJ3B7vDzxaREzcpK6fc5JVpO/BGq8xci44Ql86R41uQYAACAASURBVPNr8F/P9FjiLUaGxZs51uTg1FEpmAw6CnKTeW9XJSkhqpzodIKCMFEG6fFmLprWtdCSHGvC7vJic3r8YsonB6rZXd7YQRjR9hfkJnPSsDgOHmtm/PB4dLrgL8oWo56pIzuWPQ2MAIpkAa9G8YQ3+gxk4ogE5oxJ4+nPipmZm0xGgtmfYhMOIQTL5ubz81e2s25/FS9vPky8xcBt548PWfWkIDeFggCdQlEU3vq6gic+KeKKU3P8JWM1Hv+kCKtJz/dPUztp53+0D408ZQpJJ7ziOYvlriu4030de7w5TBCHsZVdgbtpPJbMNzBnvIk+bjfow5ebEfpmYnKewpi0CUP8LkypnxA76h/orYXE5DyJOf1DYkY+hyF+p7+PKeVTANZvmYC74RSa999Ly8E7sJV9DyEUttev7XAcveWwz4y07U314d5jeF3JfgFDbz5GWkxaUHpGl/jWzJeOuZQFaTdF3i8EkaqLGotXfMF/vijF5vIghOCOU+8g05qNq15N7zDpTdx3xn04Ki6ht29lT/MEXLVzuGrSVUxKmxS0r6/MM7urR/RkGt2JxJBIJBKJRCLpdxxNYO7+ndoGvyBRh8vjZfvhemaEuRutLfDdHi9ftfMNGJUWyzcnDee5z0v8nhDagn90GJPOrjAb9KTFmf2LuKMNNjITLR2+U2opKJFUItHON3QEhhGPV/FXJqxrdZIca8Ko15GTYg1ZGrO83k5WUgxFVS2s2aPeSHxnZwWHa21cP6/rdPOumJmbzLbSOtweL0XVzaTGtgkcWhSGFj2gPYeKwIgG2rh1rU4KfaVPY016Vqwv7NC2sKqFWJOejASzvzJIdyIJMhIsCKF+b89Njb6v3PVzR1PV5OCdnRUU5CZHtE65aNoIMhMt/O7tPby7q4IrZ+V2WrI1ECEEN8zNp6Smlfd2VQTtK6u3sXp7OYtPySHRqr4vtfPvyxQSKWB0QhVJPO75Fi4M7FFySRA2RtKA7cj3cdVPx5i8EWv2s8SN/gM6c1mIERQsI15B6Oy0Fv+ElgN301p8AwgX1tzH0cccxl5+GV7bSCxZL2BM/hy99SDG5M24G6ahuDVVTw/oUFxpuFvz+KrufQJDS4S+CZ35mN/Ac5uvvM6j6wrxulIQejv6mEPozMdCRl80himtE22e3xi+SH1nf3vaQv+k5JP4y+wX8Toy/fvOyT0Hd/OkMD0Hhu4aaEZKX6V69NV8JRKJRCKRSLrE0Qzm7hnAe7wKTb6UiS3Fdewqb8Th9ob1A0iMMVJeb2flxlJanJ4OYffL5ubTaHfzos/EvvBYC6mxJv+irCeMSLQEpZCEMsWMNxswGXQRRWBoAka4CIzANrUtLpJ9c89Pj+1gUqpVbFkyK4eRyTH8fe1B3t1ZwSPrCslPi+WcCaEN+7tDQW4yLU4P//mihG2l9UF+EFpqxUy/gKG+bloJ1WgTKGAUVTUzbWQSi0/NYfWOoxypC664UVjVzKh0NT1IE8Qi8bLQMBl0pMepkREWY/T95WaflOpPkwoXmRJqTtfMHsX+ymaMOh1Lz8jr1jHPnTScvFQrj31cGBRR/tSnh1CAa85sG087/3DmsdFAChgRsserpihMECWAAfvR79K87z5aS5aheC3EZD+DMDQE9TEmb8AQtw/HsQvxOtSSpl57Dq3FN+GsnUVryfW4Gk6h9fAP8NozsQx/E2vuEwidE2cYjw1XfQE1zjJ0MZoYoGDJ/C8oelyNJ6vHCFjouhtm4HWkEZP9DDpzRUgBY+/RjsY+GmX1NlrCOBRDx7SIv354gDv++3WYtmGHCcnxktLQ3WkcT1EP4cSJwK3H03wjIVwta4lEIpFIJIOQHnhgNPvEC+HzWdDM7cNVZBiRFENZvY17V+1CrxOcOip44Tc9J5lTR6Xw1KeHcPkiBrqzaA1Fbmosu8sbsbs8lDfYg7wRNIQQpMeZuxWBEdLEU6u04quuUd/a5ieRnx5HcU0rnoAFgra4zEmxcv3cfL4ua+CG/2xlz9FGbpg3ukO6RE84LT8FvU5w3+rd7K1oCvK4O3lkElaT3i8kTc5KICXWRF5q7655OLTUlLI6G+UNdvLTYrnmzFEI4MlPDwW1Lapq8RuNnp6filEvQqaKdMb4zASmdbNPpAgh+Mk3TkIIVcyIlMWnZpMSa+I7M0cyLMHSdYcA9DrBdXPy2X6kgS+K1L+1hlYXL24q5VtTMxmZHBxpkpkU08EzJpoMfQ+MKLFPycarCCaIUj5gprpRMeFpzcd2eCnW3EeJyX6G1pLrwWtGZz6Kedg7uJsm4KoL9spQ3Ak4Kr/dtsEbQ2vxjejMVaoIohiDogwCcTdNxSj+hzFpMw5bDobErRji92CvvBDFmd6hveKJpbX0Wqx5j6IzNnbbwLOs3sYVT2zkzR/P7rRdoNjwwqZSfrewo6lmJAvl/jKUPHismZe3HOaO88f3WYpIV3T3VI8XQWcg+d+Ocn7y/Db+d9OZTM7qRiqURCKRSCSS4xNnU7c9MLTF/PTsJL4sref1bWWMTI4hI8zC7Pbzx7OoQI1UTrQaQ5o93jAvn2ue3sL/dpRTVNXCgl5GIXz3lGxWbS/n5S2HqWpyhC1LmhZvpipKERhaVHVti5pCAmr1DKfbS1mdjRxfSoOW2pKZGMMFkzOZlZ+Ky6Ng1AtOGhYd48XMxBjW//Jsv6gSOO6igpGcMzHDn1JiNuj56BdnERvCvyQaaNEoX/qqouSnx5GVFMPF00bw0ubD3DJ/LIlWo09ssvGddPW9kp1i5Ys75of05uiMf31/Bro+/OJ+wZRMNt+1gLS4yLwzQPUl+fDWecRGmDrSnkUFI3n4g/2sWF/I6aNT+c/GElqcHq6f2zHdKCvJErLyTbSQERgR0oqFEmUYE3QlHfZ5HZnYypagM1dgzX4SoW/CMuJFFK8V+9HLiOwevh6vYzielnF4WjsRGbxmJifNxZS0hdiTHsQyfBXullG4asMLDIo7GVvptbhb8jk98/QI5hLM9sPqH3uov8OuFv9d/e3e9boardHfa/MfPL2JFeuLOFIXeXjTzrIGtvrMiAYLx5NoFA3W768CYFd5QxctJRKJRCKRDAp64IGhLea/4atAsau80Z+OEAqTQcfEEWoli3DlKs8aO4wxw+L465oD1LQ4GT2sd9EAZ4xOZXJWAg9/sB+AzKTQ4kqkERiNnUVgBKSQ2JweHG6vPwJDMyItDDDy1FJbRiRZ0OkEYzLimTgigTEZ8VG9sZeVFOO/7iZD27JTpxN+gUUjMcaIQd83S1PtWFtL1OgB7bX94dx8Wp0e/rNRXd8dqm5BUYI9L1LjzN2+JlaToU/SRwLpjnihkRxrCnoduoPFqOfqM/L4aF8VXx9p4N8bipkzJi1kdZjMxBjK6+1Rq+TYniEvYEQzPH6PkssEEdrHwdMyDvuRJehiyogd/Uf0lkrs5YtQPNEvH3N+5o3YKy7BY8vF68jAfvQ7dPVSep0Z2EqXkZeY1+vjB15T7Y3pdHs7tHtvV0XQ9lDpDCtD+GI43V4OHmv7kO2Lt77XN61wn0d2lwe3J/ic6lpdXPboZ12O3VeeEtKrQiKRSCTtWbkS8vLU/2cGQ+jntDT1odOpbVeuHOhZSyQBOLpfhUQTMApyU4g3R1ZOsit0OrVaQ7GvbH1+Dw08NdTqD6Op80UghEohAUiPN1Hd7OxyvAabC71OYA1TZUVro5ULTfH5SeSnqYv1woDv1uX1NvQ64S8bO9RJ8l2f7UcaEAJ/qsqEzATmjU3n3xuKsbs8frPT3qYPDVWunJVLjFHPdc9uprrZwQ1hzF4zEy3YXB7/32m0GfICRjTZ480hT1dJLDamikJO1+0K2u9unoTtyJUgPDhrzsTTMq5P5nHHa/tx1Z2OvewKWot/3KF0al9w5ZMbO92/40jwHfGNRTVc/9xWnvuiLWIlomgAFO5dtZMFf/6YVp+T8kAw/p53WfrvzT3qG+4822/trirZEzGuO6LHg+/s5al2eYDHK4MpakQikUj6ipUrYdkyKPH9q/V4Qj/X1KgPRVHbLlsmRQzJcYLbCR4HmDoXMBRFYdX2cv+NMW1hlBxrZHo7I8jecMnJWWQkqHe2o7GIvWDycH+ZyxGdRGDUtjjweBXsLg+vbzsS5Feh0WBzkRhjDBkNoJmNNtpc1LWoAoaWnpESayIxxkhRdVslkqP1doYnWDqUxByqGPQ6EiwGnG4vWUnB5prXz8unutnBXa/v5JWthwG1Mo2kI8mxJr57SjaVjQ4mZyVwxujQHhxalFNZHxl5SgGjG+xR1Pq2NxpW8Yrp17xgWs4dhpUYaDO59DSPp/nAPTiOXThQ0+wUu6tnosAnB6q71V5TmyNlX2VbntRGnznMQPPpQfWcexNJN1D+Gj3lb2sPDPQUJBKJRBIhd90Fra1dt2tPa6vaVyIZcJy+qIAuIjC2ltRx8wvbWLtXLfcZ6AfxzUnDOWlYHOOGdy+KIxQmg46fzh9LToqV7JTel8A06HXcsmAsWUkxnXpgeBXVt+LNr8r42UvbWbU9uLqhoih8WVrvF0PaE2cyIIR6Xer8ERiqgCGEID89lqKq4BSSzMQTI/pCw+8J0q4k6un5qZw2KoXXvjzCun1VTMtOwmqSNpHhuG7OKJKtRn46f2zYdU6m771+rLHr1KieIF+dbqBVIvmJ4U2+8uazyzOK6w1vMUGUcpXrdvxODt7u5yT1F7vKG3scYhfqbr6mD288VBNmT+f9NToTSPoqf6o/iObcZQpJG4NMF5JIJJI+oTR8dfI+7SuRRA1Ho/rchQeGllpc6VsQaWaViTFGrjgthytOy4nalKI93qKCkVw2IyvsYi/d52VQ1eRgS7HqtfbYx0V8++S2Pp8cqGbP0UYeWjQ15Bg6nSDBYqTR5qLWF4GRHFACNj8tjk8OVPl/L6+3c3J231TJOF5JtpooqWn1p9RoCCF4cdksHL7oHlMf+XAMFUYmW9n2q3M7bTN5RAK7f3NenwlB8hXqBmWkcdibzibvOL7vvJO73NfyF/dC5uq/ZqToXoTCwNHzBXVnfhEuT2TpIV22Ueh/R88oE6nQcLzIMoNYH5JIJJITmpxerLF601ciiRoOX1RAF2VUtfSHal+1jgabC6NeENPHRonRorOI3LR4VcCobnawtaSOOLOBvRVNrA+4uffY+kIyEsxccnJW2HESY4w02FzU+6KgNRNPUE0rjzU5aLK78HoVKhrsYU1FhypaRMroEFVWhBBYjHosRn1USsie6Bj0uj6NYpECRrcQnOf8PZc7f0UzaljZOs/JAEwSxQM4r97R3bvZgQteAXy091ioUXszpQ7U+P5hDYY779E0ju0rBsN1lEgkEknnLF8O1h5EuVutal/JiYMQ4ptCiH1CiINCiNvDtLlcCLFbCLFLCPF8v0zM4Ush7iKFRDOg1Kp1NNhcJFhC+0EMNrQIjP2VTRRVt7Bsbj4ZCWYe+7gQUKvgbThYww9mj+q0goQmYGgRGIHVSjRD0qKqFmpanDg93rCmokOVJF9EymjpbzHoGfICRrTvLrdiIXBxvlfJxqMIJuqKo3ugfqQ310gBblz5ZYftv169q8O2nqZAfLC7koIH1vBZYXXUXs/upHZEHFGhdK+9RCKRSCS9YckSWLECclWLLvT60M+pqepDCLXtihVqX8mJgRBCD/wTOB+YCHxPCDGxXZsxwB3AbEVRJgG39MvkIvTACBWBEaqc6GBEi8B4d2cFAKePTuXaM0fxWWENP31xG794ZTtxZkOXaS2agFHX6uxQknS0z5C0sKqZ8nqthOqJJWCkWEN7YEgGH0NewOhr7JgpVEb4IzD0eHjE+BdOE3sGdmJhCLVuv/yxz6N+nKMN9o7HjqgKScfYjc3Fqqnn1+0qnbg8Xsbf805PpxgxfRVRcTymbhyPc5JIJBJJeJYsgeJi9fPb7Q79XF2tPrxeta0UL044TgUOKopSpCiKE3gRuKRdmx8C/1QUpQ5AUZRQ4bXRx++BEV7AcLq9lNaqbrVaBEajzUXCEBEwYk16Yox6tpbWYdQLpmQl8r1Tc5g2MpEtxXU02d38dP4YEiydn29CjMEnYLj86RIauamxJFuNvP11BUcbVAHjRDPxnDM2nfMmZfirzEgGL31q4imE+CbwV0APPKEoyoPt9i8FHgI0q91/KIryRF/OqS/YpeQxS6cKFtPFAS7Qb8KJgY2uCQM8s8GBw915ZRR/ZEM7ZaPR5sLu8vbomNEOOfR6Fb4sVY2XAgWPwOP0RBworm7hUE0LZ48b1q20j6EuRAz185NIJBKJJIpkAYcDfj8CnNauzVgAIcQG1O/t9ymK8m77gYQQy4BlADnRMFKJwAOjtLYVj1fBbNBR3aymRzTYXEEeD4MZIQRp8SYO19qYnJXo92J48ydndmucxBgjjXY3dS1Of7qEhsmg46rT8/jrhwcY5lvAZ51gERjzxqYzb2z6QE9DEgX6LAIjknA1Hy8pinKy7zHoxAuAXd48MkUtKTRyln47ALN1uzh+bBqjS7TPatzdHf4/9jnRTiH51/rCDqVjhRC9rkJy1h/X8YN/b+7VGN1hCKSSDmq6ylEWQpiFEC/59m8UQuQF7LvDt32fEOK8rsYUQozyjXHAN6bJt32uEOJLIYRbCLGo3fGv9rU/IIS4OmB7gRDia98x/iaGQlKyRCKRDB1CfSa3/4JiAMYAZwHfA54QQnQoU6EoygpFUWYqijIzPT0Ki8EIPDAKfeU/p+ckUdXkQFEUGodQCgm0+WDM7GGlQICEAA+MlBDizlWn52I26HhxUykWo66DyCGRDBb6MoUkknC1IcEuJQ+ASbpiztZ9hVvRkS4aGCuODOzEQtCb5XR/rElCrfdf3Nx200CbQovTTYuj88iNaPH+7oqw+3aWNaAoCgcq22pra4JHV+JFX5p9DvXl41A8vwhF32uBOkVRTgIeBn7v6zsRWAxMAr4JPCKE0Hcx5u+BhxVFGQPU+cYGKAWWAkEGbkKIFOBe1Lt2pwL3CiG0b1qPot6RG+N7fLNXF0MikUgk0eQIkB3w+0igPESbNxVFcSmKcgjYh/p53rc4u47AKKpS/S9OHZWK0+Ol0e5WTTxj+jSQvF9J8wkYBbkpPR4jMcaI0+2lotFOUggBIzXOzOUzs/EqMCIxZkgYoEpOTPpSwAgVrhaq9s9lQogdQohXhRDZIfYf9+z2qu5ZZ+u+YpKuhJWe+QCcqds5kNOKOo02V9eNwqDVVu4tmiZwuNbG3Ic+6vE43fnQfvbzkpDb3/n6KN/6+6e88VVZyP3RZkc7D5DOaO4ncWegCNSG7K4hc66RiL6XAM/4fn4VmO+LdrgEeFFRFIfvi+dB33ghx/T1+YZvDHxjfhtAUZRiRVF2AO3/aM8DPlAUpdaXI/0B8E0hRCaQoCjK54qq2j2rjSWRSCSS44LNwBhf5J0JVfBe1a7NG8DZAEKINNSUkqI+n5mjCcVo5Ybnt7PwkQ0sfGQDb+04GtSkqKqZ9Hiz34iyqslOo909pCIwNCPPgt5EYPg8MmpbnKTEhr42180ZhU6ceAaekqFFXwoYkYSrrQbyFEWZCqyh7Yt58EBCLBNCbBFCbKmqqurWJPojiaOBOI4oaXxPvxaA5z3zKfIO54whJmD89cMDQJuBEkTuRfDQe/siavf7d/dS6FPau0ve7W+xr6KpR317ghbSGBh9AT2PrDh4rAmXJzpCz/bD9RG37W2ay0Cydm8l4+95l20+/5FBTiSir7+NoihuoAFI7aRvuO2pQL1vjHDHinR+Wb6fO5s30LvPcolEIpH0DN9n/U+A94A9wMuKouwSQvxGCHGxr9l7QI0QYjfwEfB/iqLU9PnkHE14DLG8u6sCm8vL/spmVm0PvjFUVN1CflqsP0qhuFr1xBhKAsal07O4+RsnkR7fc4PJwOuRHBvaHyQ3NZY7L5jAki4qmkgkxzN9KWB0Ga6mKEqNoijaavhxoCDUQFHPt+sDdntziRFOypUU9inZfOadxCzdHgy4u+7cj1Q02Fmzu3Kgp9EjIvGiePvro1226S/aR3nsLAuOoAjUDcrrbSz483oe+N/u/pjaoC/1ql3at3ao6T2fFdZwoLL/xKs+IhLRN1ybaG3vjF6PNRg+yyUSiWQooijK24qijFUUZbSiKMt9236lKMoq38+Koii3KooyUVGUKYqivNgvE3M04TSokRX3XzKJqSMT/UadGkVVzeSnx/kX99oNpKEkYJySl8Kt547r1RhBAkYnBqfXzcnn/CmZvTqWRDKQ9KWA0WW4mi/0WONiVFV4ULLLmwfAOs/JgOBT7xTihJ2pou+j77rDTS9s47pnt7DFV5r0RGJrSS15t7/FoeqWiKMO7nz964jahZMDAkWMW176KmjfwWNtkRu1Leo/683Fx38kwZZi9TruOBJ5lEcgH+07xlVPbYrKXF77Ur3x/9B7+zjn4fW0OI4vwbCbRJqjnA0ghDAAiUBtJ33Dba8GknxjhDtWpPM74vu5s3lLJJLjgJUrIS9PFYENhuDntDT1EbgtL0/tI5H0Gc5m7EJNZ8hMiiE93hwU6Vvb4qSu1cXo9LYIDE3A6Kqs6IlGpAKGRDLY6TMBI8JwtZuFELuEENuBm1GN46KKtnzMTbX6t41Ki432Ydih5AOw1jsdgM+9E/Eqgrn6HVE/VjRYf6B6oKfQp3i8Ch6vwvXPbeGRdQcB+O+Xakjipwc6D12/7pm2qh/PbywN2y6cj4YW3VDT7OhUKNFqmg8GXB4vXq96Lh/uVUvDf9LD99APn9nC+v29Sx8Id1mj5bUyQESSo7wK0Kp/LALW+nwnVgGLfVVKRqEar20KN6avz0e+MfCN+WYX83sPOFcIkewz7zwXeE9RlKNAkxBils9b46oIxpJIJP3MypWwbBmU+GydPJ7g55oa9RG4raRE7SNFDEmf4WiiGSs6ARnxZtLjzFQ3twkYRT6xIj89lqQYIwad8KcaD6UIjGgQLGDIayMZuvRlBEYk4Wp3KIoySVGUaYqinK0oyt5oz+Hb07NYfEo2r/9otn/bxMyEaB+Gdd5pLHX+kjXeGYDqi7FZGceN+lVcof+Q462kausgvFPdHbPk0377IQUPfMB7uyr5w7v7OvTvzMRzzZ5j3ZqXAvx3W1nA7+prvau8sfN+/fCWiOSaRTKNMXe9w9X/VqMm+jv55LnPi/nJ81/281H7nwhF3yeBVCHEQeBW4HZf313Ay8Bu4F3gx4qieMKN6RvrNuBW31ipvrERQpwihDgCfAd4TAixy3eMWuB+VFFkM/Ab3zaAG4EnUM1DC4F3on6BJBJJr7jrLmjtgW7e2qr2lUh6y5VPbuSG57YGb3Q00+Q1k5FgwaDXkRZvptXp8UdUahVI8tPi0OkEqXEmfwRrghQwggi8HilhPDAkkqHA0Kk/FAaLUc+Dl00N3tgHKzAFHeu8Jwdtu8F5C38xPsJvjU8yVRRyu/uHfXPwHvDEp4cGegp9grYYD1TvNfzlTemdceWLm0rJSLT0uL+GJnRUNtp5devxUXK3s8vS04iL9nS3atc9b6rr7X9c0fMxBguKorwNvN1u268CfrajCguh+i4Hlkcypm97EWqVkvbbNxOcEhK47yngqRDbtwCTQ/WRSCTHB6XhAwr7tK9EoqEoUNlkD97oaKTOk+6vipHuSxOpbnYQazZwpK4VISAr2bc/3szOMvUGkYzACCbB0rasC2fiKZEMBfo0AuNEp44Elrp+yaPui1hsWMfV+vcBGCdKec10L9PFgQGeYe/pacWNgUBb9GqpED3l9v9+zQ/+vbnrhhFy9VObePqz4qBtRVXN7I+CKWVfRnnYXR4+7mUqSE8Jd16DuaKKRCKR9CU5vSg60Ju+EolGcqyJ+lZX8EZnMzUuE5m+G0NaOVHNB6O8wU5GvAWjXl2yaD4YAIkyTSIIg15HnFkVMZKkuCMZwpyYAkaEa5zLZ4a8CdnNQ+n4vXsxazzTudOwkkt1n/C8aTkFugPcZHi91+OfaPTmzntg185SSLpL+zWzJ0KBRFFUEaCy0d5h3zf+9DHnPrw+GtOLaB4akVwWrc3f1x7k6qc2sbei81SZDv2PkygkiUQiOZFYvhys1q7btcdqVftKJL0l2Wr0m5ZrKI4mKp3GkBEYAEcbbGQmtUW9avuFgDjTkA8k7zaJMUYSLAYM+hNziSc5MZDv7k749slZURpJ8H+u66klgYdNj+LCwKueuZyl206uqIjSMYY+QoiIIgrCRQVookVvU0jaxgu9PVgQCL9Yr7e5GH/Pu9S1vxsRRSIRJALDOXtyWRpt/e+nEvba9+80JBKJZNCwZAmsWAG5uerven3wc2qq+gjclpur9lmypH/nKhmaJFtNNNhcuD0+w223A+Fx0uixMMIfgaGmPvgjMOrtjEiM8Y+hRWgkWIzodPKGSHsSYowyfUQy5DkxBYwB+LyrI4EfO29mnWcai5138wfXd/Gg4/v6Nf0/mQHit29H3aO1I4rC1V2U6AxcpO8sa6Cs3ta7Q3aybO5MKAksExZNmuxdCyKBAkB3RYv+iKD4xSvb+d3bwVWV825/i3d3qoKfzBSRSCSS7rNkCRQXq5+hbnfwc3W1+gjcVlwsxQtJ9NCMJRtsvu8pDtWMs5kYMn0RGKmxZnQCqpqdKIpCeb2NESEiMKT/RWjS4kwMizd33VAiGcScmALGAPGlMpalrtsoVjI5RjLvek/hcv06YuiYQiDpOwLXvjf850tmP7h2wOYSDcrrbTz2cSGKorDnaCNT7nvfv2/JExv5oqgm4rG6k0LSl7y69QiPrS/qsL3R3nm0hxQ2JBKJRCI5PknyeVbUtfrSSJyq11cLFrJ8AoZeJ0iJNVHV5KCu1YXD7SUzRASGFDBC8+uLJ/G7hVMGehoSSZ9yQgkY9100kW+MH0Z2spoE+odFUztt39droWfc55IoWlmo/7SPj9R35fDn8wAAIABJREFU2JyegZ5CEJ29Zv1VuSLi40Rptf3DZ7fwu3f2Ulrbyp6jHf0o/rH2YMRj9WRK9a1OGrqTBiMjPiUSiUQiOeHQIjD8qbMOVcBoUqx+E09QjTqrmx2U+yJkNX8MkBEYXZGfHsdJw+IHehoSSZ9yQrnfLJ09iqWzR+HyeCnITeaciRlUNNj58wf7B2Q+W5RxbPWO4ZeGF/nIczLlpA3IPHpDV3fEo4mg66onWm3w9mw4WO2vPhKtShXhUikiHj5Kikqzr1a6VxmY8qLLfDXdix+8sP8PLpFIJBKJZFCQbFUFDL+Rpy+FxKmP8YsboJZKrWoKFDACUkh8HhkJMSfUEkYikQRwQkVgaBj1Os6ZmAHAzfPHDOBMBLe6bkSHwl9N/yCdev5kfJS1pltJpnuVHU4Unvq0uNP97+wMbYq65ImNPPN5SR/MiA5hHxELGFEQUmqao+uj0R8CSF8eYjCV9ZVIJBKJ5ERCM5es8wsYagSGJS45yPQ8vdMIDFXMkBEYEsmJywkpYBxPlCjDuct1Dafo9vOp+WYu1n1Grqjk5hAlVjOoJZbeGU4OZl7YVMrh2tZejxMtn4TI/CL6VhEoeGCN/5q0ONy9NgaN5NrIDBCJRCKRSCTdJcXaPoVEvVkXG58U1C7NF4FxtMGOyaAjNSA6IyHGQLLVGFSZRCKRnFjI+KtO6C9DwFXe2Ux1FzFBlPIb95Vcqf+A7+vX8KznXGqUBG4zvMg8/XZGimrWe6ZwleuO/pnYccaBMOkh3UVB6VNh4fVtZW3H6uRN9MZX5VE5ni8zhm/9vZ+8VHpx7Y7UteJwe6MyBS0lKAgZgCGRSCQSyXFJjEmP2aBrM/FsrQUgNjkjqF16nBmH28v+yiYyEy1B39mEELx185yglBOJRHJiISMwgP87b1yHbZfNGNmvc3jAfSVLXHexT8nhL+5FODDyoPFx/me6k8v169juzedDz3TO1O0knToAEmjhWv1bxNP7qIQTCUWJng8GdFwzOz1tC/T/7Tgatl9pFKJJok1fp5Cc+fuPojKOosDfu2FOKpFIJJLusXIl5OWp/xcMhuDntDT1EbgtL0/tI5F0RkqsyZ9C4mlRq6QlpqQHtUnz+VxsP9IQMtJiRFIMFqO+j2cqkUiOV6SAAYxMVj8cAx2QB5JqEnnUfTGn6fZiEB6+67yHH7tu4Xfu76ETCufrNwHwI8Mq7jGu5GXTb8igdoBnPXiIlnQRyVr/q8P1UTpa/xDNqCNFUXhpc6nfZDTafLz/WJ+MK5FIJCc6K1fCsmVQ4rOO8niCn2tq1EfgtpIStY8UMSSdkWQ1+SMwbA3HaFCsDE8Orpqh+VzUtjjJTDo+vptLJJLjB5lCEkBBbnKnd8z7k8c9F1JHPO94TqGOBAAOKiPZ683mQv1GXvPM5Qr9h+zwjiJfHOV1869Y7LyHUiWji5ElnfHUp4e63ccZhZSI/qY/fCy2lNRx22tf80VRLQ9/9+R+OKLMIJFIJJJocNdd0NqDIMHWVrXvkiXRn5NkaJASa/RXIXE0VtOkxAeZdEJbBAZAVpL0upB0H5fLxZEjR7Db7QM9FUkILBYLI0eOxGjsmRmvFDA6YSArGjgx8rxnfoftb3lO42eG17jJ8AYJopV7nUtxYOQl0/380vASP3HdPACzHVwoCpQ3hP5A+83/dnd7vKc/K+7ljLqmye7qcV+Xx8uGg9Xd6vP4+iJsLk+H7fe8sZPnvoismkuLL/Li9W1lXQoYW0tqKchNAeDrIw3dmqtEIpFIoktp6cD0lQx9kq0mjtar5p2e5mrqiCerXZRFepzZ/3OmNOuU9IAjR44QHx9PXl5enxvqS7qHoijU1NRw5MgRRo0a1aMxZApJAIFv8GnZiQM4k/C85Z2FTijcYFjNFu9Ytilj2K3k8aznHC7QbWS0KOt6kBOcgRSm9lU09ajflPve7/ExNx6qZckTG7vVZ/nbe/jzB/s7bO9MvKhpdvDtf26gvN7Gc58X8/bXkUczXfbo5/6fP9xb2a25BtJfxrsSiUQylMnJGZi+kqFPstVErS+FRGmtpZ44slOsHdrodep38hEyhUTSA+x2O6mpqVK8OA4RQpCamtqr6BgpYKCmjgBcPrPNuPPKWbmI47BgZJEygt3eXAAed1/g3/6k+wLsmPix4c2g9ufpNrHG9AsSiE4Fj6FAf5ZRbc9VT3UuJFQ2Dt5Qt4IH1vDV4Xqe/PQQ97y5i5e3HOlW/7J6GzPu/4C/rDnQ4zk8uk4ae0okEklvWb4crNau27XHalX7SiThSI410WBz4fEqGB11OEzJmA3Bhpw6nfCXTm2fXiKRRIoUL45fevvaSAEDGJlspfjBC5kzps0FWQgxoHfqO+NJ9/ms80zjA+9M/7ZaEljpmc8lug3kigoALDi41/gsJ+nK+Y7+44GariSAykZHp/trWpx8uKfnEQi9xRsFdaenQzy/scSfFxsJoaq4PPN5ZOktEolEIgnPkiWwYgXkqvdL0OuDn1NT1UfgttxctY/0v5B0RorViKJAg81FjLsBYU0J2S7Nl0ZyvBjsSySS4wcpYLTjB7PzeP660wZ6Gp3ymncuS1234W338q1wX4gbA8sNT2LBwTX6dxghailXUvi+fg2CcGaTCgb6plLE8Ui0SqhuK+2bCiNbS+r6ZNxIaLIPjvfBJweqqW6OXOyQSCQSSfdYsgSKi1VR2u0Ofq6uVh+B24qLpXgh6ZpkX2RFbX0DMdgxxaWFbJcebybeYiDe0jOTP4lEMnSRAkY77r1oEmecFPrD9HinimTucl/D6brd/Mf0O240rOY9z0wedH2PUbpK5ui+Dmofg50r9B/ynuk29ph/wAvGB7hO/xZx9MB6HBhGHSNF56UtdXjRhRVSBhfv7Kzok3EHY1WTQLTyaN2lOylbb3XDX0MikUgkEsnxQbJVFTAKSw8DEJs8LGS7OWPSOGeCrKwnOXGIi4vr92Nu27YNIQTvvfde2Db33Xcff/zjH/txVl0jq5AMMV71zKNVMfMX4z8RwIPu71GmpFGlPMeV+g9Y750GQLao5Bnj78nXVbDTm8eznnM5Q7eLu40rudGwir+6F7LSswAP+s4P6COVBlaZ78alGJjrfBgljDb2tPH3eNGx1PVL+qeYZ0c+7WZFjv5moBOX6rqRxhGK17d1NJJ9ecthfvnqjm6P9dB7e3s1F4lEIpFIJMcPmoBxsLiE84DktNAixXVz8vtxVpKhzK9X72J3eWNUx5w4IoF7L5oU1TEHghdeeIEzzzyTF154gfPOO2+gpxMxMgKjEwZrRYO3vbNY7LyHH7l+yiElEydGXvR8g/m6bSzTr+ZC3Rf813QvyaKZK5238y3ncu53X8n5zgf5luMB9nmz+Y3xGX5leLbD2CNFFd/RrwtKRxF4edj4CMNFHdm6KmaKtuoVybR9YMRi4wzdLs7Sb2ex/qO+vQid8EVR7YAd+3ihM++c6fd/EPXjPfnJoR71++dHhVGeiUQikUgkkoEiOVZNCTlcpt7sSM/IGsjpSCR9wm233cYjjzzi//2+++7jT3/6E83NzcyfP58ZM2YwZcoU3nzzzU5GaaOzfs8++yxTp05l2rRpXHnllQBUVlZy6aWXMm3aNKZNm8Znn33WYUxFUXj11Vd5+umnef/994Oqgixfvpxx48axYMEC9u3b59/++OOPc8oppzBt2jQuu+wyWlvVqP2lS5dy4403cvbZZ5Ofn8/HH3/MNddcw4QJE1i6dGm3rl0kyAiMbvC7hVO4479taRj/+n4BN/xn6wDOKDxfKmODbuU/5z6H83SbudP4AgCHvelc7bqNImVEUL+dSj5XuO7iV8pzXGN4l4+8J7POOx0dXq7Uf8AvDS8SKxwk08QKz0UA3KhfzVz919zvWsLPDa/ybf0GNrvHs0C3lRXGP3O58x62KOOZqduPQXipUJK507CSdZ5pVJDab9dksNAfwtkN//my7w8ikUgkEolEEkCKzwOjqa4SjJCYEjqFRCKJFgMRKbF48WJuueUWfvSjHwHw8ssv8+6772KxWHj99ddJSEigurqaWbNmcfHFF3dZlSNcv927d7N8+XI2bNhAWloatbXqTdqbb76ZefPm8frrr+PxeGhu7liNcsOGDYwaNYrRo0dz1lln8fbbb7Nw4UK2bt3Kiy++yLZt23C73cyYMYOCggIAFi5cyA9/+EMA7r77bp588kluuukmAOrq6li7di2rVq3ioosuYsOGDTzxxBOccsopfPXVV5x88slRu74yAqMbLD4l2++KDHD2+PQObfJSe1B3rB84RjLnOh9ipv1Rljjv4ELn8g7iRRuC37sXs8ebzUPGFXxf/wHvmG7n18Zn2Owdz4ee6fzC8DKTRDHX6t/il8aXWO2ZxZOeC3jfW8AF+o3EYOcOw/PohMJF+s8BOE23B5ei52rnbRjwcr/x6f46fckAs6+yaaCnIJFIJJI+YuVKyMtTo/sMBvU5LU19BG7TnvPy1D6SE48Yox6TQUcy6vcCETs4feckks6YPn06x44do7y8nO3bt5OcnExOTg6KonDnnXcydepUFixYQFlZGZWVXVcfDNdv7dq1LFq0iLQ09e8oJUWt6rN27VpuvPFGAPR6PYmJiR3GfOGFF1i8eDGgCi4vvKDe5P7kk0+49NJLsVqtJCQkcPHFF/v77Ny5kzlz5jBlyhRWrlzJrl27/PsuuugihBBMmTKFjIwMpkyZgk6nY9KkSRQXF/fsQoZBRmB0AyEEW+5eQN7tb4VtkxpnprimZyaY/UE1iVR7p3TZzoGJW1w/ZpXpbh4w/ps93hx+4ryJ/3lnkUQz75pv53nTAySKVt7ynMrPXTcCgjc8s/m2/jP+bvw7o3VHKVdSOFe/lXvdS5ml2812ZTT7lBz+6b6E/zO+zGh3GYVKFiZc3GZ4kWc951CiDO/7CyE57jheyxZLJBKJJDwrV8KyZeCLJMbjUZ9ratraaNu055IStQ/IyiUnGkIIUqwmklp8d4RjQpdRlUgGO4sWLeLVV1+loqLCLxSsXLmSqqoqtm7ditFoJC8vLyh1Ixzh+imK0mX0Rig8Hg+vvfYaq1atYvny5SiKQk1NDU1NPmExzJhLly7ljTfeYNq0aTz99NOsW7fOv89sVm/y63Q6/8/a7253dKscygiMTuhqORWqakKseehoQvuUHK503sH3nHdxvvN3/M97OiCoJ55bXTdixcF/3PO5yXUzTtScxk+9U6hR4lmg38ZG73j+5LqcTFHLLN0epohDbPSOB+Alz9m4FD2X69cBsEi/nmsN73BvCN+NE42nNvTML2KwM1g9ZyQSieRE5q672sSL7tDaqvaVnHgkWY2kiCac+lgwmAZ6OhJJn7B48WJefPFFXn31VRYtWgRAQ0MDw4YNw2g08tFHH1FSUhLRWOH6zZ8/n5dffpkan2KspZDMnz+fRx99FFDFisbGYBPTNWvWMG3aNA4fPkxxcTElJSVcdtllvPHGG8ydO5fXX38dm81GU1MTq1ev9vdramoiMzMTl8vFygEMo5MCRgSMy4jnjR/PDrnvh3NGBf3emQa25tZ5UZxV/7BJmcDn3km0P7PPvJOZ7niMu93X4g14G7kxsNpzOgC/dV3BGu8M3IqO2w3PYxQevvBOBNRIkDXeGVym/wQLDm7Qr8KuGPmG/itm6XYDcLX+vf9v787Do6rOB45/39mysAbCJiCLoiKyiCCgiIgbogUsKCgqWJVqrdJirShWcOHn1mKr1AUXREWw4IaiUhegVgUJioKAQAUliBoB2UkyM+/vj3sTJmESsjCTmeT9PM88zJx77j3vnJk5zJychXG+mVG3XW3CNtI5dI9lddNBNnK654uqDiMmrP/CGGOSz3ffVc25Jnk1qBWgvuwmmFK/qkMxJmY6dOjArl27aN68Oc2aNQNgxIgRZGVl0a1bN2bMmMFxxx1XpmuVdF6HDh0YP348p59+Op07d2bs2LEA/OMf/2DBggV07NiRk046qchUD3Cmj1x44YVF0oYMGcKLL75I165dGTZsGF26dGHIkCGcdtpphXnuvvtuevTowdlnn13m2GOh+gwXiIEmdZ3hL2cd35guLQ80sl6PEAorIjD+/ONp2SCdO14v+sY4t0MT5n91YE7TEfVSObrxwfv79j46M+G39SzJbqKv9zE5OJR3wifzhR4NwJJwe071fkVQPSwLH1OYb1aoH+d5l/J3/6Mc6cnh+rwbuc0/g3G+F1kU7sIY3ysANJJfuDn/t4UdJT09q5juv58wwoJwF54KDnAWLY2hI/iZyYHHeCJ4AQvCJ5b7/DT201S2s0GbVSqOO/zP0Um+oWfuFHZw6P2ivYQOuRVufXbRRn7AS4gsjV9j1IRtnOZdwZxQH0BsBIYxxiShI490poRU9FxT82TUCtCAXbb+han2VqxYUeRxZmYmn3zySdS80RbaLMt5I0eOZOTIkUXSmjRpUuoOJ88+++xBaQMHDixc72L8+PGMjzJE7rrrritcW6Ok67Vu3ZqVK1eWWlZl2QiMUhzXtC7zbuzN2LOPLTXfRSe1BGDs2Qd+RHc9MqNInvdv6ltCGXUqF2QC2kntwpEWAPPD3QD4Utuyl9TC9A/DHdmsDenvXcrqcEveCp/MQ8GhdPF8wxjfK/wreDp/yx/KEO+HPOR/lNayhfbyLVP9f+M7bcy/QqfT3bOG6YH7aSlOZ1Fd9jDSO58rvPMZ6l3ECO97XO2dRzvJLjXmq7xv0UNWRz3WmO28GJhET89qJvmfJpXcUq5U9Fe4lxCXet9nUcpY3g/8iUGe/5YaR2lSyeVEWUea5DGsDNvQ3uabwcLAWGqxrzCtDnuLxPgn30ssT/0tr6ZMYE7KXbSXCn4LrYBb/S/yV/8TnO75EoDHF9mWqSLSX0S+FpH1IjIuyvEUEXnJPb5ERFpHHLvVTf9aRM491DVFpI17jXXuNQOllSEiI0RkecQtLCJd3GML3TIKjtmy8sbUEJMmQXoF1i9PT3fONTVPRrqfDNlFoI7tRGeMKT/rwDiEDkfUw+uJPjGkIDUt4GXjfedz45ntSrxOWiD6X8F93rK/BOklXCPR/TvUjbAKn0R0agCE8TA75EyreSw4CMXDK6HT+CDUhWeC/bkleA2PhH7NX/Mv4leeT1iYchOvBe5gN2lckTeOCcErGZx7N2GEKf5HaE4OLwXu4k7/dO7yT+ev/ieY5H+G2/0zeDkwgQ4SfW2JUz0r+Iv/BZ4L3Ecfd3pGY7ZzoedDfud9nZmBe8iUHdyTP4IjZBtXe9+Kep2enlUsS7mWu3zTqM1eOsgG3gyM5//8T/OtNmapHstk/2MM9By8F3M0HsJc6X2bTHYAcJJnLQEJ8YvW4grfu3gJlXhuH88XjPbNo6Unh8u97wLQTrJZmnIdI7zvF15/mHchS8LHcV3eGHLVx0XeRWWKrbKasI3zPUsAuMU3C4kyTaimEREv8E/gPOB44BIROb5YtquA7ap6NPAQcL977vHAcKAD0B94VES8h7jm/cBDqtoO2O5eu8QyVHWGqnZR1S7A5cBGVV0eEduIguOq+tNhqhZjTIIbMQKmToVWrZzHXverSsOGzi0yreDfVq2cc2wBzxpm01LY+F+Gdz+StrVy8doIDGOKWLFiBV26dCly69GjR1WHlXBsCkkFaClj3cu7EKzfW7YTPhrXj7Mnx+fH5eH2Aw0ZmjeBtdrioGNPBs9ns2byZrgn4HRq/Cb/z0XyTAldyJxQH873LqGHZzUPBIexBedb0WYa8ef80TwR+DsfpPyJfLxckXcLK8NtSJf95KmfWrKfFwL/x3OB+7gi71bWaMvCaRVCmFt8s8jWTHZoLZ70T+bD8An09XyBT5wf1Vu0AVfl3cwSbc9JnrVc55vLS6G+5HBglE1nWc9T/r+yl1Qu875Hf+9SGrCTrdTlt3l/ZH64G2nkMi3wIA/5/8n2/Np8GO5Uar39yvMxE/zP00J+5u7g5fTyrCKoHu7Mv4KHAo9xtmcZ74RPPui8uuzmAf9U1oWb85PW52rfW0wPncM9/mdIlXwu877HjNCZnOxZQyPZwYT8kbwd7sEF4U8Y5P2Ie4OXkh/RNIzzvUhT2caE/FFlmrZSFlf4/o2HMA/mX8zN/n/xK89i5oZPqdC1hDACRdZiSVInA+tV9RsAEZkFDAJWReQZBEx0788BpoizVPQgYJaq5gIbRGS9ez2iXVNEVgP9gEvdPNPd6z5WUhlatOG7BJh5GJ6zMaYaGDHCOiNMGbz9Z/D6OeGqf0N4F6TbDiTGROrYsSPLly8/dMYaLum/8SeqyG/61/U9qsR8JY3uKK55/bRKRlS1PtNjoq6ZsYc0Zof6HvLH5w805OnQAEbn38T6Yh0h88Mn82RwADuoxSV5t/OfcGe2UZdsbcxPZLBBmzEi7zbCeJiXchv/S72c/6bcyADPYs7zfEonzwYm5w9lRN5trNXmnORZx1OhAZybex/H7Z9Gr9wpLNH2ANwXvAQ/QaYFHqS/51NayE+M8r7D9MD9bNW6XJA7iaF5E/heG/BqqDdn5z7A/HB3QNhHKr/Ju5n12pyH/I/SmO0AnOJZyYWeD2kn2YULlnoJFa4BcqH3QwLk08uzii+1La+HT2VTuBFjfK8wzvcik/2PcoV3Pm1kCxd4PmF64AEaspM/5l/H5OBQMmUnzwYeoIdnDUvCx9He8x0dZCMDPEvYpwEWhLsAMDt0Og1lF/08nxXW7VDvIq71vclg78fMS7mNbrKmYm+ACKnkcqn3A/4d7sajoYGsCrfiJt+/8FO2LZYasoPIT9i9vqdYlnItY7wvU4+S5w8mgebApojH2W5a1DyqGgR2AA1LObek9IbAL+41ipdVUhmRhnFwB8Y0d/rIX6Qie3oZY4yp3o4+E7KzYM/PkLsT0m0KiTGm/GwERhyMKWVqib8cU0jsF0HJJgUv497gpSV2hGzUZlyYdxf9PJ+RwW7O9H7Go4GH2acB1oRb8lq4N2E8DM67G0EJlvDR+Fab8qf867jZ9xKPB/5emL4y3Jpr8//IT2Twk2YwOO+eqOfvJZXf5Y/hjcDtPBJ4hK1alwHeTwuPZ2smv80by3HyHW09PzAjeCYjfO8z0PsxneQbpobOJ4yHp0IDuNM/nbayhV+oxa+9B9bW2KwNGZd/DSu1LQAfh47nFO8qssLHMDpvLJ+mXM9w7wL6e5eyINyFfe66JB+GO/Gj1meo9z/MD59MO8nmbt80Pg4dzwPB4Tzsf4Q5KXexOtySd8MnsUfTEJTdpLFDa7FZM1mvzm/gHp7VNJHtvBI6rci6JwAXeReRIbt5Ongeiof7g8OZHrifq71v8VjIWTxICKMIke/6uuxmnG8Wl/o+4N78S3gi9CuOlB+5yLuIzZrJH/0vc41vHmzrDg3aRq3/BBftI158uFdJeUpKj/aBKC3/IeMQkR7AXlVdGXF8hKpuFpE6wMs4U0wO2hNZREYDowGOtNX7jDGmZjmqH/znQVj1mvM4LaP0/MYYE4V1YFRCaX9kjDyS6j+wdsULV/XgsqeXFD72lXEEhjm0Q43iyNZGPBdy1jacEhrMaO88RvveZFJwROG5h9qxA2Bu+BTezOvJ2Z5lHCE/syDchY3l2F3kf9qc2/OvZHLgcXLVzwP5w3gv3JVOnm8Y65vN7MCd7CaNleHW3BEcxRnez7nV3YbW2dIWpofO4Y1QL7ZTG8VDG9nCqZ6VbNCmfBzugEbUxYPBYdwnT3Jb/lXsoDbvhLtzifcDfBLmrdCBeXUhvLwaOo2rvfN4xP8wvTyr2E0aY/J/Tw71OT/vXoZ6FzHAu4Qbfa+V+PzCKnjE+b17tfctxuZf5+4So4zyzme8bwZZ4WPIUmdx3EXhzrwd6s4Y38u8He7Obk3nucB97CPAlXk3s5Pa9PJ8xcP+KTRgJxvDTbjR9wqvhnoz2vsmQbwMyZtIA9nF+d7F3JjRpsTYElw20DLicQvg+xLyZIuID6gHbDvEudHSfwbqi4jPHWURmb+kMgoMp9joC1Xd7P67S0RexJm+clAHhqpOBaYCdOvWzfadMcaYmqRFdwjUhhVznMc2AsMYUwHWgVEBfq+H3GDYXQujaAdEp+b1WPh1DqcencmoU1pzRa9WRY73bpfJxvvOZ8HXP7F9T17hFJIrT21NesDLlh37aVgrwJMfHlhw8oJOzo9jZ8HPAws3/ubUNjzzUfSFKU3pQnh5LDSQx0K/oiJjW8J43KkhFfNKuA978tJYoy35VpsCsDbUkkWhzkwNTOZEz3rG5V9NCC9zQn240fcaeeolq3AbWmEbdQuvt0GbsSEUvRPlc23HuXkPFD6eHTqdQd6P2a9+Pii2JeysUF9Ged/hJM9assLH8mhwIDk4WwjvIp1pofOYFjqPFPLwuKMkarOfDNlFS/mJYyQbP0GWhNvjkxAP+KcyJ3AnmzWT3aTS3rOJd0MncVP+tUTW+4T8UZyacjMP+p+gHntoKTl4CfFC4F7mhPrwF98LbNSmjMq/hV2k8W7gZu71P0Vvz0peDvUhhwxyNIOvg0dyY/LOXlgKtBORNsBmnI6CS4vlmQuMBD4BhgIfqKqKyFzgRRGZDBwBtAM+xankg67pnrPAvcYs95qvl1YGgIh4gIuAPgUBuZ0c9VX1ZxHxAxcA7x2+ajHGGFMteP3Qpg987S6GbmtgGFMuOTk5HHHEEUyZMoXf/va3UfM8++yzZGVlMWXKlDhHFz8x7cAQkf7APwAv8JSq3lfseArOX+lOArYCw1R1YyxjOhxeu/5U3l31Y9QdRMacdQzndGjKCc3rcULzeiVe44xjnV0GVZWA18M5HZoWWQ8jsgPjgaHOYo9103zs2JfPdX2PYlCXIziuaV3rwKi0qvuxG60DJIf6DM+7nfbyHcv1aMDpcLjR9xqfazv2k1Lpcj8Od2BjuMlB29qCM9WmQ+4zhxyJkkug8P5+Uvhzg5ToAAAXl0lEQVRZ67FOW/ABXQ9kUuifex+jvPM5yvM9TdnOpPxLeSo0oMgIEYCfyODe4CXc63+avZrCqLxbSJf9PO5/iLv80/kwdALX549hJ7UAeCo0gOt9cwmp8ETogkrWSGJQ1aCI/B6Yj9NmPqOqX4nIXUCWqs4Fngaedxfp3IbTIYGb7184C34GgetVNQQQ7ZpukbcAs0TkHuBz99qUVIarD5BdsCioKwWY73ZeeHE6L548bBVjjDGm+jiqX0QHho3AMKY8Zs+eTc+ePZk5c2aJHRg1Qcw6MCK27zsbZ0jyUhGZq6qRK+oXbtcnIsNxtusbFquYDpf2zerSvlndqMe8Him146I4EeG8jgf/5Xzald25ctpSANIDzsv0/G968NryzYw5s13h9JVOLerxZfaOg86/f0hHjqifxuVPf3rQsZLc0O9oHvlgfZnzm9jIJVDYeQGwSZvwYP7FrNJWpZxVdmE8DMy7mzz8UY+XZRpNWe0mnSmhCyllx9dCs0JnkMFuPgkfz+faDhRG5o+jq6xlauiCIuuS/DM4mMHej/g0fFzhCJYCqlrq9K5EpqpvAW8VS7sj4v5+nBEQ0c6dBEwqyzXd9G84sFNJZHppZSwEehZL24PTCW2MMYc0YwaMHw/ffutsqxoKHdhudevWA2mtWsGkSba7SbVzVL8D99NsBIaJg7fHwQ8rDu81m3aE8+4rNcvgwYPZtGkT+/fvZ8yYMYwePRqAd955h9tuu41QKERmZibvv/8+u3fv5oYbbiArKwsRYcKECQwZMuSga86cOZO//e1vXHrppWzevJnmzZ2156ZNm8a9995Ls2bNOOaYY0hJcf7g+cYbb3DPPfeQl5dHw4YNmTFjBk2aNGHixIls2LCBLVu2sHbtWiZPnszixYt5++23ad68OW+88QZ+f/TfCYkglruQFG4JqKp5OMOUBxXLMwhn+z5wtus701avdxSM0IjUOrMWfzjrmCI/zub+vjer7+pP1yPrc9+vO1I7xcc5xzdhWPcj6e1OY3nnD6cV5h/R40iW3X4Wa+7uz9MjuxWmvzf2dG4659jCxw1qOX9h73dcY8add1yZYr6h39FR0wd0bBo13ZTdP0ODWVBsukdl7KT2YRnNcTgpHh4NDXI6L1yLw8fzaGjwQYuq7iWVc3Pv58/5B/c+l7LLsTHGmCo0YwaMHu10XoDTUQFOx8XWrUXTvv3WyTtjRvzjNDHUoC3Udxdxtikkphp75plnWLZsGVlZWTz88MNs3bqVnJwcrrnmGl5++WW++OILZs+eDcDdd99NvXr1WLFiBV9++SX9+vU76HqbNm3ihx9+4OSTT+biiy/mpZdeAmDLli1MmDCBjz76iHfffZdVqw6MFejduzeLFy/m888/Z/jw4TzwwIEp5f/73/+YN28er7/+OpdddhlnnHEGK1asIC0tjXnz5sW4dionllNIom3f16OkPO7w6YLt+n6OYVxJ4+0xp5HiO3QfU1rAyyu/OxWA4ScfWNlfRJg40Fn08dPxZ3LTv75g/PntC0d0nNm+CW/8vjf10/20bOBscTr72l60zaxFw9pFf9wOPakFv+zNp16an7vfXMXIU1pzUitn9ejZWZu4ec6XXNOnLX846xj25Yc496H/MLx7S25wd2DJ2riNm+d8yYaf9wBQL81P76MzmbdiC9NGdad5Rhqffbudca+s4O5BHXj2442owjc/76Fzi3p84Y4yCXg95IXCB9XBmDPb0SazFn946cDeyff+uiMNawWYseQ7cnblsmrLzsJjtVN87M51dpBs1TCdb7fuBWD8gPZMemv1QdfPrJ3Cz7tzD/laVFb/Dk1556sfYl5OrI3s1Yrpn3wb0zKibctrjDEmcY0fD3v3lj3/3r3OOTYKoxoRgWP6w8qXwZ9W1dGYmuAQIyVi5eGHH+bVV18FnM6HdevWkZOTQ58+fWjTxllwvkEDpxPvvffeY9asWYXnZmQcvEPPrFmzuPjiiwEYPnw4V111FWPHjmXJkiX07duXRo0aATBs2DDWrl0LQHZ2NsOGDWPLli3k5eUVlgtw3nnn4ff76dixI6FQiP79+wPQsWNHNm7ceJhr4/CKZQdGZbYELJqphm69V9I0lYpoXCeV568q3n8EHVsUne7SvXX03vDM2ilkup0aD19SdCTARd1aclG3Axsd1E7x8dG4oj2H3Vo34IObTiesFFnr458ReY5pUqewA+byXq2BA9MB9ueH8HmkcN0RVUUVPMV2cRl8YvODYj+nw4ERIKpKKKxR1y8pcE0fZwvOcFgPun5kTMFQmD15IeqllT7ESlXJ2ZVL47qp7NjndALlBkOEw07n0968ICk+b2G9hMNKSLXIFrv78kKkBbyoKtnb9xV2OEXKD4XxeeSg6RO5wRCrvt9Jm8xa1E8PFMa0Pz9MwOcp/BBGPtfIaRjRpmSEw0qeW14wrEV22ilw56ATAPhlbx6pfm9hnpxduTSsFUAE8kJhdu4LkpHuJ6wQVmXHvnwa10lBRMgLOp1V2/bk0ahOCrtzg9RJ8bErN8jOffk0rpvC/rwwO/fnR60TY4wxieO77+JzjklwZ06AHtdWdRTGxMzChQt57733+OSTT0hPT6dv377s37+/xGnOZZn+PHPmTH788UdmuMPSvv/+e9atWweUvDPmDTfcwNixYxk4cCALFy5k4sSJhccKppl4PB78fn/hNTweD8FgsNzPOZ5iOYWkPFsCFqxkX3y7PsDZek9Vu6lqt4LeJZN8RKRI50VZzwFnK9rITgcRidq5UJbrldZ5Eamk6xfE5PN6Dtl5UZC/cV1nscyC/Ck+L2kB5wd9esBXpF48HinSeQEU5hWREn+o+72eqA1Yis/LiUdmFHZeFFwnLeB0mng8B9dl5HWiXdPjkcLXJFrnRaT66YEieRrVScHjdrSk+Lw0qpOCz+sh4HOu1aRuamGZAZ+T3rReKl6PUC/Nj8f9t2WDdFJ8XupFjCAyxhiTuCryN6ga9HermiOlNjQ8qqqjMCZmduzYQUZGBunp6axZs4bFixcD0KtXLxYtWsSGDc4mDNu2OT97zznnnCK7hmzfvr3I9b7++mv27NnD5s2b2bhxIxs3buTWW29l1qxZ9OjRg4ULF7J161by8/MLp6UUxFGwTsb06dOpLmLZgVG4JaCIBHBWsp9bLE/Bdn1QbLs+Y4wxxhhTfUyaBOnl6G9OT3fOMcaYZNK/f3+CwSCdOnXiL3/5Cz17OuufN2rUiKlTp/LrX/+azp07M2yYs3fF7bffzvbt2znhhBPo3LkzCxYsKHK9mTNncuGFFxZJGzJkCDNnzqRZs2ZMnDiRXr16cdZZZ9G164EdASdOnMhFF13EaaedRmZmZoyfdfxILPsLRGQA8HcObN83KXJLQBFJBZ4HTsTdrq/Y9nwH6datm2ZlZcUsZmNMzSIiy1S126FzmsPJ2nJjaqZY7UJibXn8WTtuEtXq1atp3759VYdhShHtNSprOx7LNTAqtSWgMcYYY4ypXkaMsEU5q4KI9Af+gfNHxadU9b5ix0cBDwKb3aQpqvpUXIM0xpgyiGkHhjHGGGOMMabqiIgXZ930s3HWn1sqInNVdVWxrC+p6u/jHqAxxpRDLNfAMMYYY4wxxlStk4H1qvqNquYBs4BBVRyTMTFlyyomrsq+NtaBYYwxxhhjTPXVHNgU8TjbTStuiIh8KSJzRKRllOOIyGgRyRKRrJycnFjEakylpaamsnXrVuvESECqytatW0lNTa3wNWwKiTHGGGOMMdVXtH3hi/+yewOYqaq5InItMB3od9BJqlOBqeAs4nm4AzXmcGjRogXZ2dlYJ1tiSk1NpUWLFhU+3zowjDHGGGOMqb6ygcgRFS2A7yMzqOrWiIdPAvfHIS5jYsLv99OmTZuqDsPEiE0hMcYYY4wxpvpaCrQTkTYiEgCGA3MjM4hIs4iHA4HVcYzPGGPKzEZgGGOMMcYYU02palBEfg/Mx9lG9RlV/UpE7gKyVHUucKOIDASCwDZgVJUFbIwxpbAODGOMMcYYY6oxVX0LeKtY2h0R928Fbo13XMYYU16SbKuzikgO8G05T8sEfo5BOMlSfiLEUNXlJ0IMNb38RIghWvmtVLVRVQRTk1WgLa/q905FJFvMyRYvJF/MyRYvJE/M1pbHWZJ+Jy+vZIsXLOZ4SLZ4ITliLlM7nnQdGBUhIlmq2q2mlp8IMVR1+YkQQ00vPxFiqOryTcUl42uXbDEnW7yQfDEnW7yQnDGbxJVs76dkixcs5nhItnghOWMuiS3iaYwxxhhjjDHGmIRnHRjGGGOMMcYYY4xJeDWlA2NqDS8fqj6Gqi4fqj6Gml4+VH0MVV2+qbhkfO2SLeZkixeSL+ZkixeSM2aTuJLt/ZRs8YLFHA/JFi8kZ8xR1Yg1MIwxxhhjjDHGGJPcasoIDGOMMcYYY4wxxiQx68AwxhhjjDHGGGNMwqvWHRgi0l9EvhaR9SIyLsZlbRSRFSKyXESy3LQGIvKuiKxz/81w00VEHnbj+lJEulagvGdE5CcRWRmRVu7yRGSkm3+diIw8DDFMFJHNbj0sF5EBEcdudWP4WkTOjUiv0OskIi1FZIGIrBaRr0RkTDzroZTy41kHqSLyqYh84cZwp5veRkSWuM/nJREJuOkp7uP17vHWh4qtguU/KyIbIuqgi5seq/eiV0Q+F5E34/n8TXxU9PMRL+VtixJJWT87iUBE6ovIHBFZ49Z1r0SvYxH5o/ueWCkiM902M6HqWA7T9wljSpPo7Tgkb1ueTO04JF9bbu14AlLVankDvMD/gLZAAPgCOD6G5W0EMoulPQCMc++PA+537w8A3gYE6AksqUB5fYCuwMqKlgc0AL5x/81w72dUMoaJwJ+i5D3efQ1SgDbua+OtzOsENAO6uvfrAGvdcuJSD6WUH886EKC2e98PLHGf27+A4W7648B17v3fAY+794cDL5UWWyXKfxYYGiV/rN6LY4EXgTfdx3F5/naL/a0yn484xliutiiRbmX97CTCDZgOXO3eDwD1E7mOgebABiAtom5HJVodcxi+T9jNbqXdkqEdd+NMyrY8mdpxN6akacutHU/MW3UegXEysF5Vv1HVPGAWMCjOMQzC+ZDi/js4Iv05dSwG6otIs/JcWFX/A2yrZHnnAu+q6jZV3Q68C/SvZAwlGQTMUtVcVd0ArMd5jSr8OqnqFlX9zL2/C1iN09DEpR5KKT+edaCqutt96HdvCvQD5pRQBwV1Mwc4U0SklNgqWn5JDvt7UURaAOcDT7mPhTg9fxMXidCWl6oCbVFCKOdnp0qJSF2cL2hPA6hqnqr+QoLXMeAD0kTEB6QDW0iwOj5M3yeMKU3Ct+OQnG15MrXjkLRtubXjCaY6d2A0BzZFPM6m9B+XlaXAv0VkmYiMdtOaqOoWcBpFoHGMYytvebGK4/fukKRnIoaAxTQGcaYCnIgzAiDu9VCsfIhjHbhDB5cDP+H88P8f8IuqBqNcr7As9/gOoGFlYihevqoW1MEktw4eEpGU4uUXK6cydfB34M9A2H3ckDg+fxNzSfXalLEtShTl+exUtbZADjDNHSr9lIjUIoHrWFU3A38FvsP5wrsDWEbi1nGkeH9/MdVb0r1vkqgtT6Z2HJKsLbd2PDFV5w4MiZIWyz1jT1XVrsB5wPUi0qeUvPGOraTyYhHHY8BRQBecD/rfYh2DiNQGXgb+oKo7S8saixiilB/XOlDVkKp2AVrg/JWjfSnXO+wxFC9fRE4AbgWOA7rjTAu5JRbli8gFwE+quiwyuZRrxfOzYA6PpHltytEWVbkKfHaqmg9neOxjqnoisAdnSGzCcjuvB+FMSzsCqIXzHaG4RKnjskjk94hJXEn1vkmWtjwJ23FIsrbc2vHEVJ07MLKBlhGPWwDfx6owVf3e/fcn4FWcH5I/FgzJcf/9Kcaxlbe8wx6Hqv7o/qANA09yYBh+TGIQET/OfzIzVPUVNzlu9RCt/HjXQQF3CN5CnPls9d2hbsWvV1iWe7wezpCzSscQUX5/dximqmouMI3Y1cGpwEAR2YgzJLUfzl8j4v78TcwkxWtTzrYoEZT3s1PVsoHsiBFec3C+BCdyHZ8FbFDVHFXNB14BTiFx6zhSvL+/mOotad43SdaWJ1s7DsnXlls7noCqcwfGUqCdu0psAGfBvrmxKEhEaolInYL7wDnASre8gt0URgKvu/fnAle4q8D2BHYUDPGppPKWNx84R0Qy3B7Gc9y0Cis2h+pCnHooiGG4OLtAtAHaAZ9SidfJnef3NLBaVSdHHIpLPZRUfpzroJGI1Hfvp+E0tKuBBcDQEuqgoG6GAh+oqpYSW0XKXxPRYArOnLvIOjhsr4Gq3qqqLVS1NU69faCqI+L1/E1cxK0tr6gKtEVVrgKfnSqlqj8Am0TkWDfpTGAVCVzHOEOOe4pIuvseKYg5Ieu4mHh/fzHVW8K345B8bXmyteOQlG25teOJSBNgJdFY3XBWWV2LsybA+BiW0xZnReUvgK8KysKZh/Y+sM79t4GbLsA/3bhWAN0qUOZMnOkJ+Tg9aVdVpDzgNzgLFq4HrjwMMTzvlvElzgekWUT+8W4MXwPnVfZ1AnrjDHn6Elju3gbEqx5KKT+eddAJ+NwtayVwR8R78lP3+cwGUtz0VPfxevd420PFVsHyP3DrYCXwAgd2KonJe9E9vy8HVuCOy/O3W3xuFf18xDG+crVFiXYry2cnEW440/Ky3Hp+DWfHooSuY+BOYI3bFj6Ps9NRQtUxh+n7hN3sVtot0dtxN8akbcuTpR1340uqttza8cS7iftEjDHGGGOMMcYYYxJWdZ5CYowxxhhjjDHGmGrCOjCMMcYYY4wxxhiT8KwDwxhjjDHGGGOMMQnPOjCMMcYYY4wxxhiT8KwDwxhjjDHGGGOMMQnPOjBMwhOR1iKyshz5R4nIEWXIM6Xy0RljjKlKItJXRN6s6jiMMcZUjLXjpjysA8NUR6OAUjswjDHGGGOMMcYkF+vAMMnCJyLTReRLEZkjIukicoeILBWRlSIyVRxDgW7ADBFZLiJpItJdRD4WkS9E5FMRqeNe8wgReUdE1onIA1X43IwxptoTkcvcNni5iDwhIl4R2S0ifxORz0TkfRFp5ObtIiKL3Tb/VRHJcNOPFpH33Pb8MxE5yr18bff/hjUiMkNEpMqeqDHGVFPWjptEYB0YJlkcC0xV1U7ATuB3wBRV7a6qJwBpwAWqOgfIAkaoahcgBLwEjFHVzsBZwD73ml2AYUBHYJiItIzrMzLGmBpCRNrjtLenRrTNI4BawGeq2hVYBExwT3kOuMVt81dEpM8A/um256cAW9z0E4E/AMcDbYFTY/6kjDGmBrF23CQK68AwyWKTqn7k3n8B6A2cISJLRGQF0A/oEOW8Y4EtqroUQFV3qmrQPfa+qu5Q1f3AKqBVbJ+CMcbUWGcCJwFLRWS5+7gtEMbpZAa3bReRekB9VV3kpk8H+rij55qr6qsAqrpfVfe6eT5V1WxVDQPLgdbxeFLGGFODWDtuEoKvqgMwpow0yuNHgW6quklEJgKpUc6TKOcWyI24H8I+D8YYEysCTFfVW4skivylWL6S2uuCa5TE2nNjjIkta8dNQrARGCZZHCkivdz7lwD/de//LCK1gaEReXcBBetcrMFZ66I7gIjUERFrEI0xJr7eB4aKSGMAEWkgIq1wvocUtN+XAv9V1R3AdhE5zU2/HFikqjuBbBEZ7F4jRUTS4/osjDGm5rJ23CQE+yFnksVqYKSIPAGsAx4DMnDm1G0ElkbkfRZ4XET2Ab1w5us9IiJpOOtfnBW/sI0xxqjqKhG5Hfi3iHiAfOB6YA/QQUSWATtw2muAkTjteDrwDXClm3458ISI3OVe46I4Pg1jjKmxrB03iUJUSxvlY4wxxhgTGyKyW1VrV3UcxhhjKsbacRNvNoXEGGOMMcYYY4wxCc9GYBhjjDHGGGOMMSbh2QgMY4wxxhhjjDHGJDzrwDDGGGOMMcYYY0zCsw4MY4wxxhhjjDHGJDzrwDDGGGOMMcYYY0zCsw4MY4wxxhhjjDHGJLz/BxfAeOTAAu+KAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 1080x360 with 3 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "f, axs = plt.subplots(1,3,figsize=(15,5))\n",
    "plt.subplot(131)\n",
    "plt.plot(loss_history.batch_loss)\n",
    "tic = np.linspace(0, len(loss_history.batch_loss), len(train_history.history['val_loss']))\n",
    "plt.plot(tic, train_history.history['loss'])\n",
    "plt.plot(tic, train_history.history['val_loss'])\n",
    "plt.title('Loss dynamics')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('batch')\n",
    "plt.legend(['loss (by batch)', 'loss', 'val_loss'])\n",
    "\n",
    "plt.subplot(132)\n",
    "plt.plot(loss_history.lr,'bo')\n",
    "plt.title('Learning rate dynamics over epochs')\n",
    "plt.ylabel('lrate')\n",
    "plt.xlabel('epoch')\n",
    "\n",
    "asa = []\n",
    "optimizer_name = model.optimizer.__class__.__name__\n",
    "plt.subplot(133)\n",
    "plt.plot(train_history.history['val_acc'])\n",
    "plt.plot(train_history.history['acc'])\n",
    "asa.append('val_acc ' + optimizer_name)\n",
    "asa.append('acc ' + optimizer_name)\n",
    "plt.title('Accuracy dynamics over epochs')\n",
    "plt.ylabel('accuracy')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(asa)\n",
    "\n",
    "plt.tight_layout()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.6.3"
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}