diff --git "a/GAN_project.ipynb" "b/GAN_project.ipynb"
deleted file mode 100644--- "a/GAN_project.ipynb"
+++ /dev/null
@@ -1,7123 +0,0 @@
-{
- "cells": [
-  {
-   "cell_type": "code",
-   "execution_count": 11,
-   "id": "87a9239b",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "import numpy as np\n",
-    "import pandas as pd\n",
-    "import matplotlib.pyplot as plt\n",
-    "import seaborn as sns\n",
-    "import os\n",
-    "import pickle\n",
-    "import time\n",
-    "import random"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 10,
-   "id": "78f1e9ff",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "import PIL\n",
-    "from PIL import Image\n",
-    "import keras.backend as K\n",
-    "import tensorflow as tf\n",
-    "from tensorflow import keras\n",
-    "from keras.optimizers import Adam\n",
-    "from keras.models import Sequential\n",
-    "from keras import layers,Model,Input\n",
-    "from keras.layers import Lambda,Reshape,UpSampling2D,ReLU,add,ZeroPadding2D\n",
-    "from keras.layers import Activation,BatchNormalization,Concatenate\n",
-    "from keras.layers import Dense,Conv2D,Flatten,Dropout,LeakyReLU\n",
-    "from keras.preprocessing.image import ImageDataGenerator"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "730eee1e",
-   "metadata": {},
-   "source": [
-    "### Oxford 102 Flower dataset"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 86,
-   "id": "1a730a2d",
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Found 8189 files belonging to 1 classes.\n",
-      "Using 6552 files for training.\n"
-     ]
-    }
-   ],
-   "source": [
-    "flower_train=keras.utils.image_dataset_from_directory(directory=\"D:\\\\unikaksha\\\\GAN_project\",\n",
-    "                                             labels=\"inferred\",\n",
-    "                                             validation_split=0.2,\n",
-    "                                             subset=\"training\",\n",
-    "                                             seed=1337,\n",
-    "                                             label_mode=\"int\",\n",
-    "                                             batch_size=32,\n",
-    "                                             image_size=(256,256))"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 87,
-   "id": "1b9d6559",
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Found 8189 files belonging to 1 classes.\n",
-      "Using 1637 files for validation.\n"
-     ]
-    }
-   ],
-   "source": [
-    "flower_test=keras.utils.image_dataset_from_directory(directory=\"D:\\\\unikaksha\\\\GAN_project\",\n",
-    "                                             labels=\"inferred\",\n",
-    "                                             validation_split=0.2,\n",
-    "                                             subset=\"validation\",\n",
-    "                                             seed=1337,\n",
-    "                                             label_mode=\"int\",\n",
-    "                                             batch_size=32,\n",
-    "                                             image_size=(256,256))"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "9e8be9bc",
-   "metadata": {},
-   "outputs": [],
-   "source": []
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 88,
-   "id": "3dd1f991",
-   "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<Figure size 640x480 with 1 Axes>"
-      ]
-     },
-     "metadata": {},
-     "output_type": "display_data"
-    }
-   ],
-   "source": [
-    "for image,label in flower_train.take(1):\n",
-    "    plt.imshow(image[1].numpy().astype(\"uint8\"))\n",
-    "    plt.show()"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 89,
-   "id": "a0af030b",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "data_augmentation = keras.Sequential(\n",
-    "  [\n",
-    "    layers.RandomFlip(\"horizontal\", input_shape=(256, 256, 3)),\n",
-    "    layers.RandomRotation(0.3),\n",
-    "    layers.RandomZoom(0.3),\n",
-    "  ]\n",
-    ")"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 90,
-   "id": "c6ad6622",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "flower_train_gen = flower_train.map(lambda x, y: (data_augmentation(x, training=True), y))"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 91,
-   "id": "90038e56",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "normalization_layer = tf.keras.layers.Rescaling(scale=1./127.5,offset=-1)\n",
-    "f_data= flower_train_gen.map(lambda x, y: (normalization_layer(x), y))"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "2666230b",
-   "metadata": {},
-   "outputs": [],
-   "source": []
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 101,
-   "id": "2534270a",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "#Discriminator model\n",
-    "def discriminator(in_shape=(256,256,3)):\n",
-    "    model=Sequential()\n",
-    "    model.add(Conv2D(64,(3,3),padding=\"same\",input_shape=in_shape))\n",
-    "    model.add(LeakyReLU(alpha=.2))\n",
-    "    model.add(Conv2D(64,(3,3),strides=(2,2),padding=\"same\"))\n",
-    "    model.add(LeakyReLU(alpha=.2))\n",
-    "    model.add(Conv2D(64,(3,3),strides=(2,2),padding=\"same\"))\n",
-    "    model.add(LeakyReLU(alpha=.2))\n",
-    "    model.add(Conv2D(64,(3,3),strides=(2,2),padding=\"same\"))\n",
-    "    model.add(LeakyReLU(alpha=.2))\n",
-    "    model.add(Flatten())\n",
-    "    model.add(Dropout(0.4))\n",
-    "    model.add(Dense(1,activation=\"sigmoid\"))\n",
-    "    opt=Adam(lr=0.0002,beta_1=0.5)\n",
-    "    model.compile(loss=\"binary_crossentropy\",optimizer=opt,metrics=[\"accuracy\"])\n",
-    "    return model"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 102,
-   "id": "8570c642",
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model: \"sequential_14\"\n",
-      "_________________________________________________________________\n",
-      " Layer (type)                Output Shape              Param #   \n",
-      "=================================================================\n",
-      " conv2d_28 (Conv2D)          (None, 256, 256, 64)      1792      \n",
-      "                                                                 \n",
-      " leaky_re_lu_28 (LeakyReLU)  (None, 256, 256, 64)      0         \n",
-      "                                                                 \n",
-      " conv2d_29 (Conv2D)          (None, 128, 128, 64)      36928     \n",
-      "                                                                 \n",
-      " leaky_re_lu_29 (LeakyReLU)  (None, 128, 128, 64)      0         \n",
-      "                                                                 \n",
-      " conv2d_30 (Conv2D)          (None, 64, 64, 64)        36928     \n",
-      "                                                                 \n",
-      " leaky_re_lu_30 (LeakyReLU)  (None, 64, 64, 64)        0         \n",
-      "                                                                 \n",
-      " conv2d_31 (Conv2D)          (None, 32, 32, 64)        36928     \n",
-      "                                                                 \n",
-      " leaky_re_lu_31 (LeakyReLU)  (None, 32, 32, 64)        0         \n",
-      "                                                                 \n",
-      " flatten_6 (Flatten)         (None, 65536)             0         \n",
-      "                                                                 \n",
-      " dropout_6 (Dropout)         (None, 65536)             0         \n",
-      "                                                                 \n",
-      " dense_6 (Dense)             (None, 1)                 65537     \n",
-      "                                                                 \n",
-      "=================================================================\n",
-      "Total params: 178,113\n",
-      "Trainable params: 178,113\n",
-      "Non-trainable params: 0\n",
-      "_________________________________________________________________\n"
-     ]
-    }
-   ],
-   "source": [
-    "model=discriminator()\n",
-    "model.summary()"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 103,
-   "id": "9683e80f",
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Epoch 1/20\n",
-      "205/205 [==============================] - 3869s 19s/step - loss: 0.0096 - accuracy: 0.9956 - val_loss: 0.0000e+00 - val_accuracy: 1.0000\n",
-      "Epoch 2/20\n",
-      " 22/205 [==>...........................] - ETA: 45:35 - loss: 1.2188e-05 - accuracy: 1.0000"
-     ]
-    },
-    {
-     "ename": "KeyboardInterrupt",
-     "evalue": "",
-     "output_type": "error",
-     "traceback": [
-      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
-      "\u001b[1;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
-      "Cell \u001b[1;32mIn[103], line 1\u001b[0m\n\u001b[1;32m----> 1\u001b[0m history\u001b[38;5;241m=\u001b[39m\u001b[43mmodel\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mfit\u001b[49m\u001b[43m(\u001b[49m\u001b[43mf_data\u001b[49m\u001b[43m,\u001b[49m\u001b[43mepochs\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;241;43m20\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43mvalidation_data\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mflower_test\u001b[49m\u001b[43m)\u001b[49m\n",
-      "File \u001b[1;32m~\\anaconda3\\lib\\site-packages\\keras\\utils\\traceback_utils.py:65\u001b[0m, in \u001b[0;36mfilter_traceback.<locals>.error_handler\u001b[1;34m(*args, **kwargs)\u001b[0m\n\u001b[0;32m     63\u001b[0m filtered_tb \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mNone\u001b[39;00m\n\u001b[0;32m     64\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[1;32m---> 65\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m fn(\u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs)\n\u001b[0;32m     66\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m \u001b[38;5;167;01mException\u001b[39;00m \u001b[38;5;28;01mas\u001b[39;00m e:\n\u001b[0;32m     67\u001b[0m     filtered_tb \u001b[38;5;241m=\u001b[39m _process_traceback_frames(e\u001b[38;5;241m.\u001b[39m__traceback__)\n",
-      "File \u001b[1;32m~\\anaconda3\\lib\\site-packages\\keras\\engine\\training.py:1685\u001b[0m, in \u001b[0;36mModel.fit\u001b[1;34m(self, x, y, batch_size, epochs, verbose, callbacks, validation_split, validation_data, shuffle, class_weight, sample_weight, initial_epoch, steps_per_epoch, validation_steps, validation_batch_size, validation_freq, max_queue_size, workers, use_multiprocessing)\u001b[0m\n\u001b[0;32m   1677\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m tf\u001b[38;5;241m.\u001b[39mprofiler\u001b[38;5;241m.\u001b[39mexperimental\u001b[38;5;241m.\u001b[39mTrace(\n\u001b[0;32m   1678\u001b[0m     \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mtrain\u001b[39m\u001b[38;5;124m\"\u001b[39m,\n\u001b[0;32m   1679\u001b[0m     epoch_num\u001b[38;5;241m=\u001b[39mepoch,\n\u001b[1;32m   (...)\u001b[0m\n\u001b[0;32m   1682\u001b[0m     _r\u001b[38;5;241m=\u001b[39m\u001b[38;5;241m1\u001b[39m,\n\u001b[0;32m   1683\u001b[0m ):\n\u001b[0;32m   1684\u001b[0m     callbacks\u001b[38;5;241m.\u001b[39mon_train_batch_begin(step)\n\u001b[1;32m-> 1685\u001b[0m     tmp_logs \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mtrain_function\u001b[49m\u001b[43m(\u001b[49m\u001b[43miterator\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m   1686\u001b[0m     \u001b[38;5;28;01mif\u001b[39;00m data_handler\u001b[38;5;241m.\u001b[39mshould_sync:\n\u001b[0;32m   1687\u001b[0m         context\u001b[38;5;241m.\u001b[39masync_wait()\n",
-      "File \u001b[1;32m~\\anaconda3\\lib\\site-packages\\tensorflow\\python\\util\\traceback_utils.py:150\u001b[0m, in \u001b[0;36mfilter_traceback.<locals>.error_handler\u001b[1;34m(*args, **kwargs)\u001b[0m\n\u001b[0;32m    148\u001b[0m filtered_tb \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mNone\u001b[39;00m\n\u001b[0;32m    149\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[1;32m--> 150\u001b[0m   \u001b[38;5;28;01mreturn\u001b[39;00m fn(\u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwargs)\n\u001b[0;32m    151\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m \u001b[38;5;167;01mException\u001b[39;00m \u001b[38;5;28;01mas\u001b[39;00m e:\n\u001b[0;32m    152\u001b[0m   filtered_tb \u001b[38;5;241m=\u001b[39m _process_traceback_frames(e\u001b[38;5;241m.\u001b[39m__traceback__)\n",
-      "File \u001b[1;32m~\\anaconda3\\lib\\site-packages\\tensorflow\\python\\eager\\polymorphic_function\\polymorphic_function.py:894\u001b[0m, in \u001b[0;36mFunction.__call__\u001b[1;34m(self, *args, **kwds)\u001b[0m\n\u001b[0;32m    891\u001b[0m compiler \u001b[38;5;241m=\u001b[39m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mxla\u001b[39m\u001b[38;5;124m\"\u001b[39m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_jit_compile \u001b[38;5;28;01melse\u001b[39;00m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mnonXla\u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[0;32m    893\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m OptionalXlaContext(\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_jit_compile):\n\u001b[1;32m--> 894\u001b[0m   result \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_call(\u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwds)\n\u001b[0;32m    896\u001b[0m new_tracing_count \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39mexperimental_get_tracing_count()\n\u001b[0;32m    897\u001b[0m without_tracing \u001b[38;5;241m=\u001b[39m (tracing_count \u001b[38;5;241m==\u001b[39m new_tracing_count)\n",
-      "File \u001b[1;32m~\\anaconda3\\lib\\site-packages\\tensorflow\\python\\eager\\polymorphic_function\\polymorphic_function.py:926\u001b[0m, in \u001b[0;36mFunction._call\u001b[1;34m(self, *args, **kwds)\u001b[0m\n\u001b[0;32m    923\u001b[0m   \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_lock\u001b[38;5;241m.\u001b[39mrelease()\n\u001b[0;32m    924\u001b[0m   \u001b[38;5;66;03m# In this case we have created variables on the first call, so we run the\u001b[39;00m\n\u001b[0;32m    925\u001b[0m   \u001b[38;5;66;03m# defunned version which is guaranteed to never create variables.\u001b[39;00m\n\u001b[1;32m--> 926\u001b[0m   \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_no_variable_creation_fn(\u001b[38;5;241m*\u001b[39margs, \u001b[38;5;241m*\u001b[39m\u001b[38;5;241m*\u001b[39mkwds)  \u001b[38;5;66;03m# pylint: disable=not-callable\u001b[39;00m\n\u001b[0;32m    927\u001b[0m \u001b[38;5;28;01melif\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_variable_creation_fn \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[0;32m    928\u001b[0m   \u001b[38;5;66;03m# Release the lock early so that multiple threads can perform the call\u001b[39;00m\n\u001b[0;32m    929\u001b[0m   \u001b[38;5;66;03m# in parallel.\u001b[39;00m\n\u001b[0;32m    930\u001b[0m   \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_lock\u001b[38;5;241m.\u001b[39mrelease()\n",
-      "File \u001b[1;32m~\\anaconda3\\lib\\site-packages\\tensorflow\\python\\eager\\polymorphic_function\\tracing_compiler.py:143\u001b[0m, in \u001b[0;36mTracingCompiler.__call__\u001b[1;34m(self, *args, **kwargs)\u001b[0m\n\u001b[0;32m    140\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_lock:\n\u001b[0;32m    141\u001b[0m   (concrete_function,\n\u001b[0;32m    142\u001b[0m    filtered_flat_args) \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_maybe_define_function(args, kwargs)\n\u001b[1;32m--> 143\u001b[0m \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[43mconcrete_function\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_call_flat\u001b[49m\u001b[43m(\u001b[49m\n\u001b[0;32m    144\u001b[0m \u001b[43m    \u001b[49m\u001b[43mfiltered_flat_args\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcaptured_inputs\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mconcrete_function\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mcaptured_inputs\u001b[49m\u001b[43m)\u001b[49m\n",
-      "File \u001b[1;32m~\\anaconda3\\lib\\site-packages\\tensorflow\\python\\eager\\polymorphic_function\\monomorphic_function.py:1757\u001b[0m, in \u001b[0;36mConcreteFunction._call_flat\u001b[1;34m(self, args, captured_inputs, cancellation_manager)\u001b[0m\n\u001b[0;32m   1753\u001b[0m possible_gradient_type \u001b[38;5;241m=\u001b[39m gradients_util\u001b[38;5;241m.\u001b[39mPossibleTapeGradientTypes(args)\n\u001b[0;32m   1754\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m (possible_gradient_type \u001b[38;5;241m==\u001b[39m gradients_util\u001b[38;5;241m.\u001b[39mPOSSIBLE_GRADIENT_TYPES_NONE\n\u001b[0;32m   1755\u001b[0m     \u001b[38;5;129;01mand\u001b[39;00m executing_eagerly):\n\u001b[0;32m   1756\u001b[0m   \u001b[38;5;66;03m# No tape is watching; skip to running the function.\u001b[39;00m\n\u001b[1;32m-> 1757\u001b[0m   \u001b[38;5;28;01mreturn\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_build_call_outputs(\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_inference_function\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mcall\u001b[49m\u001b[43m(\u001b[49m\n\u001b[0;32m   1758\u001b[0m \u001b[43m      \u001b[49m\u001b[43mctx\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mcancellation_manager\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mcancellation_manager\u001b[49m\u001b[43m)\u001b[49m)\n\u001b[0;32m   1759\u001b[0m forward_backward \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_select_forward_and_backward_functions(\n\u001b[0;32m   1760\u001b[0m     args,\n\u001b[0;32m   1761\u001b[0m     possible_gradient_type,\n\u001b[0;32m   1762\u001b[0m     executing_eagerly)\n\u001b[0;32m   1763\u001b[0m forward_function, args_with_tangents \u001b[38;5;241m=\u001b[39m forward_backward\u001b[38;5;241m.\u001b[39mforward()\n",
-      "File \u001b[1;32m~\\anaconda3\\lib\\site-packages\\tensorflow\\python\\eager\\polymorphic_function\\monomorphic_function.py:381\u001b[0m, in \u001b[0;36m_EagerDefinedFunction.call\u001b[1;34m(self, ctx, args, cancellation_manager)\u001b[0m\n\u001b[0;32m    379\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m _InterpolateFunctionError(\u001b[38;5;28mself\u001b[39m):\n\u001b[0;32m    380\u001b[0m   \u001b[38;5;28;01mif\u001b[39;00m cancellation_manager \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[1;32m--> 381\u001b[0m     outputs \u001b[38;5;241m=\u001b[39m \u001b[43mexecute\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mexecute\u001b[49m\u001b[43m(\u001b[49m\n\u001b[0;32m    382\u001b[0m \u001b[43m        \u001b[49m\u001b[38;5;28;43mstr\u001b[39;49m\u001b[43m(\u001b[49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43msignature\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mname\u001b[49m\u001b[43m)\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    383\u001b[0m \u001b[43m        \u001b[49m\u001b[43mnum_outputs\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_num_outputs\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    384\u001b[0m \u001b[43m        \u001b[49m\u001b[43minputs\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43margs\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    385\u001b[0m \u001b[43m        \u001b[49m\u001b[43mattrs\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mattrs\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m    386\u001b[0m \u001b[43m        \u001b[49m\u001b[43mctx\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43mctx\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m    387\u001b[0m   \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[0;32m    388\u001b[0m     outputs \u001b[38;5;241m=\u001b[39m execute\u001b[38;5;241m.\u001b[39mexecute_with_cancellation(\n\u001b[0;32m    389\u001b[0m         \u001b[38;5;28mstr\u001b[39m(\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39msignature\u001b[38;5;241m.\u001b[39mname),\n\u001b[0;32m    390\u001b[0m         num_outputs\u001b[38;5;241m=\u001b[39m\u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_num_outputs,\n\u001b[1;32m   (...)\u001b[0m\n\u001b[0;32m    393\u001b[0m         ctx\u001b[38;5;241m=\u001b[39mctx,\n\u001b[0;32m    394\u001b[0m         cancellation_manager\u001b[38;5;241m=\u001b[39mcancellation_manager)\n",
-      "File \u001b[1;32m~\\anaconda3\\lib\\site-packages\\tensorflow\\python\\eager\\execute.py:52\u001b[0m, in \u001b[0;36mquick_execute\u001b[1;34m(op_name, num_outputs, inputs, attrs, ctx, name)\u001b[0m\n\u001b[0;32m     50\u001b[0m \u001b[38;5;28;01mtry\u001b[39;00m:\n\u001b[0;32m     51\u001b[0m   ctx\u001b[38;5;241m.\u001b[39mensure_initialized()\n\u001b[1;32m---> 52\u001b[0m   tensors \u001b[38;5;241m=\u001b[39m \u001b[43mpywrap_tfe\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mTFE_Py_Execute\u001b[49m\u001b[43m(\u001b[49m\u001b[43mctx\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_handle\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mdevice_name\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mop_name\u001b[49m\u001b[43m,\u001b[49m\n\u001b[0;32m     53\u001b[0m \u001b[43m                                      \u001b[49m\u001b[43minputs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mattrs\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mnum_outputs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[0;32m     54\u001b[0m \u001b[38;5;28;01mexcept\u001b[39;00m core\u001b[38;5;241m.\u001b[39m_NotOkStatusException \u001b[38;5;28;01mas\u001b[39;00m e:\n\u001b[0;32m     55\u001b[0m   \u001b[38;5;28;01mif\u001b[39;00m name \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n",
-      "\u001b[1;31mKeyboardInterrupt\u001b[0m: "
-     ]
-    }
-   ],
-   "source": [
-    "history=model.fit(f_data,epochs=20,validation_data=flower_test)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "8996963d",
-   "metadata": {},
-   "outputs": [],
-   "source": []
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "47daa149",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "\n",
-    "    "
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "d25326e6",
-   "metadata": {},
-   "outputs": [],
-   "source": []
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "dcb622da",
-   "metadata": {},
-   "outputs": [],
-   "source": []
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "2d36e7e7",
-   "metadata": {},
-   "outputs": [],
-   "source": []
-  },
-  {
-   "cell_type": "markdown",
-   "id": "ebc3020e",
-   "metadata": {},
-   "source": [
-    "### CUB-200-2011 (Caltech-UCSD Birds-200-2011) dataset\n"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 12,
-   "id": "a7bc6928",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "#conditonal augmentation (Text embedding-word2vecembadding)\n",
-    "def conditioning_augmentation(x):\n",
-    "    mean=x[:,:128]\n",
-    "    log_sigma=x[:,128:]\n",
-    "    stddev=tf.math.exp(log_sigma)\n",
-    "    epsilon=K.random_normal(shape=K.constant((mean.shape[1],),dtype=\"int32\"))\n",
-    "    c=mean+stddev*epsilon\n",
-    "    return c\n",
-    "\n",
-    "def build_ca_network():\n",
-    "    \"\"\"Builds the conditioning augmentation network.\"\"\"\n",
-    "    input_layer1 = Input(shape=(1024,)) #size of the vocabulary in the text data\n",
-    "    mls = Dense(256)(input_layer1)\n",
-    "    mls = LeakyReLU(alpha=0.2)(mls)\n",
-    "    ca = Lambda(conditioning_augmentation)(mls)\n",
-    "    return Model(inputs=[input_layer1], outputs=[ca]) "
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "9c4b10b1",
-   "metadata": {},
-   "source": [
-    "### Stage 1 Generator Network"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 13,
-   "id": "5d97acf0",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "def UpSamplingBlock(x, num_kernels):\n",
-    "    \"\"\"An Upsample block with Upsampling2D, Conv2D, BatchNormalization and a ReLU activation.\n",
-    "    Args:\n",
-    "    x: The preceding layer as input.\n",
-    "    num_kernels: Number of kernels for the Conv2D layer.\n",
-    "    Returns:\n",
-    "    x: The final activation layer after the Upsampling block.\n",
-    "        \"\"\"\n",
-    "    x = UpSampling2D(size=(2,2))(x)\n",
-    "    x = Conv2D(num_kernels, kernel_size=(3,3), padding='same', strides=1, use_bias=False,\n",
-    "            kernel_initializer='he_uniform')(x)\n",
-    "    x = BatchNormalization(gamma_initializer='ones', beta_initializer='zeros')(x) #prevent from mode collapse\n",
-    "    x = ReLU()(x)\n",
-    "    return x\n",
-    "\n",
-    "def build_stage1_generator():\n",
-    "\n",
-    "    input_layer1 = Input(shape=(1024,))\n",
-    "    ca = Dense(256)(input_layer1)\n",
-    "    ca = LeakyReLU(alpha=0.2)(ca)\n",
-    "\n",
-    "    # Obtain the conditioned text\n",
-    "    c = Lambda(conditioning_augmentation)(ca)\n",
-    "\n",
-    "    input_layer2 = Input(shape=(100,))\n",
-    "    concat = Concatenate(axis=1)([c, input_layer2]) \n",
-    "\n",
-    "    x = Dense(16384, use_bias=False)(concat) \n",
-    "    x = ReLU()(x)\n",
-    "    x = Reshape((4, 4, 1024), input_shape=(16384,))(x)\n",
-    "\n",
-    "    x = UpSamplingBlock(x, 512) \n",
-    "    x = UpSamplingBlock(x, 256)\n",
-    "    x = UpSamplingBlock(x, 128)\n",
-    "    x = UpSamplingBlock(x, 64)   # upsampled our image to 64*64*3 \n",
-    "\n",
-    "    x = Conv2D(3, kernel_size=3, padding='same', strides=1, use_bias=False,\n",
-    "            kernel_initializer='he_uniform')(x)\n",
-    "    x = Activation('tanh')(x)\n",
-    "\n",
-    "    stage1_gen = Model(inputs=[input_layer1, input_layer2], outputs=[x, ca]) \n",
-    "    return stage1_gen"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 14,
-   "id": "129b4b59",
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model: \"model\"\n",
-      "__________________________________________________________________________________________________\n",
-      " Layer (type)                   Output Shape         Param #     Connected to                     \n",
-      "==================================================================================================\n",
-      " input_4 (InputLayer)           [(None, 1024)]       0           []                               \n",
-      "                                                                                                  \n",
-      " dense_1 (Dense)                (None, 256)          262400      ['input_4[0][0]']                \n",
-      "                                                                                                  \n",
-      " leaky_re_lu_2 (LeakyReLU)      (None, 256)          0           ['dense_1[0][0]']                \n",
-      "                                                                                                  \n",
-      " lambda_1 (Lambda)              (None, 128)          0           ['leaky_re_lu_2[0][0]']          \n",
-      "                                                                                                  \n",
-      " input_5 (InputLayer)           [(None, 100)]        0           []                               \n",
-      "                                                                                                  \n",
-      " concatenate (Concatenate)      (None, 228)          0           ['lambda_1[0][0]',               \n",
-      "                                                                  'input_5[0][0]']                \n",
-      "                                                                                                  \n",
-      " dense_2 (Dense)                (None, 16384)        3735552     ['concatenate[0][0]']            \n",
-      "                                                                                                  \n",
-      " re_lu (ReLU)                   (None, 16384)        0           ['dense_2[0][0]']                \n",
-      "                                                                                                  \n",
-      " reshape (Reshape)              (None, 4, 4, 1024)   0           ['re_lu[0][0]']                  \n",
-      "                                                                                                  \n",
-      " up_sampling2d (UpSampling2D)   (None, 8, 8, 1024)   0           ['reshape[0][0]']                \n",
-      "                                                                                                  \n",
-      " conv2d_1 (Conv2D)              (None, 8, 8, 512)    4718592     ['up_sampling2d[0][0]']          \n",
-      "                                                                                                  \n",
-      " batch_normalization (BatchNorm  (None, 8, 8, 512)   2048        ['conv2d_1[0][0]']               \n",
-      " alization)                                                                                       \n",
-      "                                                                                                  \n",
-      " re_lu_1 (ReLU)                 (None, 8, 8, 512)    0           ['batch_normalization[0][0]']    \n",
-      "                                                                                                  \n",
-      " up_sampling2d_1 (UpSampling2D)  (None, 16, 16, 512)  0          ['re_lu_1[0][0]']                \n",
-      "                                                                                                  \n",
-      " conv2d_2 (Conv2D)              (None, 16, 16, 256)  1179648     ['up_sampling2d_1[0][0]']        \n",
-      "                                                                                                  \n",
-      " batch_normalization_1 (BatchNo  (None, 16, 16, 256)  1024       ['conv2d_2[0][0]']               \n",
-      " rmalization)                                                                                     \n",
-      "                                                                                                  \n",
-      " re_lu_2 (ReLU)                 (None, 16, 16, 256)  0           ['batch_normalization_1[0][0]']  \n",
-      "                                                                                                  \n",
-      " up_sampling2d_2 (UpSampling2D)  (None, 32, 32, 256)  0          ['re_lu_2[0][0]']                \n",
-      "                                                                                                  \n",
-      " conv2d_3 (Conv2D)              (None, 32, 32, 128)  294912      ['up_sampling2d_2[0][0]']        \n",
-      "                                                                                                  \n",
-      " batch_normalization_2 (BatchNo  (None, 32, 32, 128)  512        ['conv2d_3[0][0]']               \n",
-      " rmalization)                                                                                     \n",
-      "                                                                                                  \n",
-      " re_lu_3 (ReLU)                 (None, 32, 32, 128)  0           ['batch_normalization_2[0][0]']  \n",
-      "                                                                                                  \n",
-      " up_sampling2d_3 (UpSampling2D)  (None, 64, 64, 128)  0          ['re_lu_3[0][0]']                \n",
-      "                                                                                                  \n",
-      " conv2d_4 (Conv2D)              (None, 64, 64, 64)   73728       ['up_sampling2d_3[0][0]']        \n",
-      "                                                                                                  \n",
-      " batch_normalization_3 (BatchNo  (None, 64, 64, 64)  256         ['conv2d_4[0][0]']               \n",
-      " rmalization)                                                                                     \n",
-      "                                                                                                  \n",
-      " re_lu_4 (ReLU)                 (None, 64, 64, 64)   0           ['batch_normalization_3[0][0]']  \n",
-      "                                                                                                  \n",
-      " conv2d_5 (Conv2D)              (None, 64, 64, 3)    1728        ['re_lu_4[0][0]']                \n",
-      "                                                                                                  \n",
-      " activation (Activation)        (None, 64, 64, 3)    0           ['conv2d_5[0][0]']               \n",
-      "                                                                                                  \n",
-      "==================================================================================================\n",
-      "Total params: 10,270,400\n",
-      "Trainable params: 10,268,480\n",
-      "Non-trainable params: 1,920\n",
-      "__________________________________________________________________________________________________\n"
-     ]
-    }
-   ],
-   "source": [
-    "generator = build_stage1_generator()\n",
-    "generator.summary()"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "20881269",
-   "metadata": {},
-   "source": [
-    "### Stage 1 Discriminator Network"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 15,
-   "id": "f596e5b9",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "def ConvBlock(x, num_kernels, kernel_size=(4,4), strides=2, activation=True):\n",
-    "    \"\"\"A ConvBlock with a Conv2D, BatchNormalization and LeakyReLU activation.\n",
-    "\n",
-    "    Args:\n",
-    "        x: The preceding layer as input.\n",
-    "        num_kernels: Number of kernels for the Conv2D layer.\n",
-    "\n",
-    "    Returns:\n",
-    "        x: The final activation layer after the ConvBlock block.\"\"\"\n",
-    "    x = Conv2D(num_kernels, kernel_size=kernel_size, padding='same', strides=strides, use_bias=False,\n",
-    "        kernel_initializer='he_uniform')(x)\n",
-    "    x = BatchNormalization(gamma_initializer='ones', beta_initializer='zeros')(x)\n",
-    "\n",
-    "    if activation:\n",
-    "        x = LeakyReLU(alpha=0.2)(x)\n",
-    "    return x\n",
-    "\n",
-    "def build_embedding_compressor():\n",
-    "    \"\"\"Build embedding compressor model\n",
-    "    \"\"\"\n",
-    "    input_layer1 = Input(shape=(1024,)) \n",
-    "    x = Dense(128)(input_layer1)\n",
-    "    x = ReLU()(x)\n",
-    "\n",
-    "    model = Model(inputs=[input_layer1], outputs=[x])\n",
-    "    return model\n",
-    "\n",
-    "# the discriminator is fed with two inputs, the feature from Generator and the text embedding\n",
-    "def build_stage1_discriminator():\n",
-    "    \"\"\"Builds the Stage 1 Discriminator that uses the 64x64 resolution images from the generator\n",
-    "    and the compressed and spatially replicated embedding.\n",
-    "\n",
-    "    Returns:\n",
-    "        Stage 1 Discriminator Model for StackGAN.\n",
-    "    \"\"\"\n",
-    "    input_layer1 = Input(shape=(64, 64, 3))  \n",
-    "\n",
-    "    x = Conv2D(64, kernel_size=(4,4), strides=2, padding='same', use_bias=False,\n",
-    "        kernel_initializer='he_uniform')(input_layer1)\n",
-    "    x = LeakyReLU(alpha=0.2)(x)\n",
-    "\n",
-    "    x = ConvBlock(x, 128)\n",
-    "    x = ConvBlock(x, 256)\n",
-    "    x = ConvBlock(x, 512)\n",
-    "\n",
-    "    # Obtain the compressed and spatially replicated text embedding\n",
-    "    input_layer2 = Input(shape=(4, 4, 128)) #2nd input to discriminator, text embedding\n",
-    "    concat = concatenate([x, input_layer2])\n",
-    "\n",
-    "    x1 = Conv2D(512, kernel_size=(1,1), padding='same', strides=1, use_bias=False,\n",
-    "                kernel_initializer='he_uniform')(concat)\n",
-    "    x1 = BatchNormalization(gamma_initializer='ones', beta_initializer='zeros')(x)\n",
-    "    x1 = LeakyReLU(alpha=0.2)(x)\n",
-    "\n",
-    "    # Flatten and add a FC layer to predict.\n",
-    "    x1 = Flatten()(x1)\n",
-    "    x1 = Dense(1)(x1)\n",
-    "    x1 = Activation('sigmoid')(x1)\n",
-    "    stage1_dis = Model(inputs=[input_layer1, input_layer2], outputs=[x1])  \n",
-    "    return stage1_dis"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 16,
-   "id": "bd11a773",
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model: \"model_1\"\n",
-      "__________________________________________________________________________________________________\n",
-      " Layer (type)                   Output Shape         Param #     Connected to                     \n",
-      "==================================================================================================\n",
-      " input_6 (InputLayer)           [(None, 64, 64, 3)]  0           []                               \n",
-      "                                                                                                  \n",
-      " conv2d_6 (Conv2D)              (None, 32, 32, 64)   3072        ['input_6[0][0]']                \n",
-      "                                                                                                  \n",
-      " leaky_re_lu_3 (LeakyReLU)      (None, 32, 32, 64)   0           ['conv2d_6[0][0]']               \n",
-      "                                                                                                  \n",
-      " conv2d_7 (Conv2D)              (None, 16, 16, 128)  131072      ['leaky_re_lu_3[0][0]']          \n",
-      "                                                                                                  \n",
-      " batch_normalization_4 (BatchNo  (None, 16, 16, 128)  512        ['conv2d_7[0][0]']               \n",
-      " rmalization)                                                                                     \n",
-      "                                                                                                  \n",
-      " leaky_re_lu_4 (LeakyReLU)      (None, 16, 16, 128)  0           ['batch_normalization_4[0][0]']  \n",
-      "                                                                                                  \n",
-      " conv2d_8 (Conv2D)              (None, 8, 8, 256)    524288      ['leaky_re_lu_4[0][0]']          \n",
-      "                                                                                                  \n",
-      " batch_normalization_5 (BatchNo  (None, 8, 8, 256)   1024        ['conv2d_8[0][0]']               \n",
-      " rmalization)                                                                                     \n",
-      "                                                                                                  \n",
-      " leaky_re_lu_5 (LeakyReLU)      (None, 8, 8, 256)    0           ['batch_normalization_5[0][0]']  \n",
-      "                                                                                                  \n",
-      " conv2d_9 (Conv2D)              (None, 4, 4, 512)    2097152     ['leaky_re_lu_5[0][0]']          \n",
-      "                                                                                                  \n",
-      " batch_normalization_6 (BatchNo  (None, 4, 4, 512)   2048        ['conv2d_9[0][0]']               \n",
-      " rmalization)                                                                                     \n",
-      "                                                                                                  \n",
-      " leaky_re_lu_6 (LeakyReLU)      (None, 4, 4, 512)    0           ['batch_normalization_6[0][0]']  \n",
-      "                                                                                                  \n",
-      " leaky_re_lu_7 (LeakyReLU)      (None, 4, 4, 512)    0           ['leaky_re_lu_6[0][0]']          \n",
-      "                                                                                                  \n",
-      " flatten (Flatten)              (None, 8192)         0           ['leaky_re_lu_7[0][0]']          \n",
-      "                                                                                                  \n",
-      " dense_3 (Dense)                (None, 1)            8193        ['flatten[0][0]']                \n",
-      "                                                                                                  \n",
-      " input_7 (InputLayer)           [(None, 4, 4, 128)]  0           []                               \n",
-      "                                                                                                  \n",
-      " activation_1 (Activation)      (None, 1)            0           ['dense_3[0][0]']                \n",
-      "                                                                                                  \n",
-      "==================================================================================================\n",
-      "Total params: 2,767,361\n",
-      "Trainable params: 2,765,569\n",
-      "Non-trainable params: 1,792\n",
-      "__________________________________________________________________________________________________\n"
-     ]
-    }
-   ],
-   "source": [
-    "discriminator = build_stage1_discriminator()\n",
-    "discriminator.summary()"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "b59c5a57",
-   "metadata": {},
-   "source": [
-    "### Stage 1 Adversarial Model (Building a GAN)\n",
-    "Generator and discriminator are stacked together. Output of the former is the input of the latter"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 17,
-   "id": "d35ac36f",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "# Building GAN with Generator and Discriminator\n",
-    "\n",
-    "def build_adversarial(generator_model, discriminator_model):\n",
-    "    \"\"\"Stage 1 Adversarial model.\n",
-    "\n",
-    "    Args:\n",
-    "        generator_model: Stage 1 Generator Model\n",
-    "        discriminator_model: Stage 1 Discriminator Model\n",
-    "\n",
-    "    Returns:\n",
-    "        Adversarial Model.\"\"\"\n",
-    "    input_layer1 = Input(shape=(1024,))  \n",
-    "    input_layer2 = Input(shape=(100,)) \n",
-    "    input_layer3 = Input(shape=(4, 4, 128)) \n",
-    "\n",
-    "    x, ca = generator_model([input_layer1, input_layer2]) #text,noise\n",
-    "\n",
-    "    discriminator_model.trainable = False \n",
-    "\n",
-    "    probabilities = discriminator_model([x, input_layer3]) \n",
-    "    adversarial_model = Model(inputs=[input_layer1, input_layer2, input_layer3], outputs=[probabilities, ca])\n",
-    "    return adversarial_model"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 18,
-   "id": "40a96d77",
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model: \"model_2\"\n",
-      "__________________________________________________________________________________________________\n",
-      " Layer (type)                   Output Shape         Param #     Connected to                     \n",
-      "==================================================================================================\n",
-      " input_8 (InputLayer)           [(None, 1024)]       0           []                               \n",
-      "                                                                                                  \n",
-      " input_9 (InputLayer)           [(None, 100)]        0           []                               \n",
-      "                                                                                                  \n",
-      " model (Functional)             [(None, 64, 64, 3),  10270400    ['input_8[0][0]',                \n",
-      "                                 (None, 256)]                     'input_9[0][0]']                \n",
-      "                                                                                                  \n",
-      " input_10 (InputLayer)          [(None, 4, 4, 128)]  0           []                               \n",
-      "                                                                                                  \n",
-      " model_1 (Functional)           (None, 1)            2767361     ['model[0][0]',                  \n",
-      "                                                                  'input_10[0][0]']               \n",
-      "                                                                                                  \n",
-      "==================================================================================================\n",
-      "Total params: 13,037,761\n",
-      "Trainable params: 10,268,480\n",
-      "Non-trainable params: 2,769,281\n",
-      "__________________________________________________________________________________________________\n"
-     ]
-    }
-   ],
-   "source": [
-    "ganstage1 = build_adversarial(generator, discriminator)\n",
-    "ganstage1.summary()"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "9bd8cfe6",
-   "metadata": {},
-   "source": [
-    "### Train Utilities"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 19,
-   "id": "ccddd25a",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "def checkpoint_prefix():\n",
-    "    checkpoint_dir = './training_checkpoints'\n",
-    "    checkpoint_prefix = os.path.join(checkpoint_dir, 'ckpt')\n",
-    "\n",
-    "    return checkpoint_prefix\n",
-    "\n",
-    "def adversarial_loss(y_true, y_pred):\n",
-    "    mean = y_pred[:, :128]\n",
-    "    ls = y_pred[:, 128:]\n",
-    "    loss = -ls + 0.5 * (-1 + tf.math.exp(2.0 * ls) + tf.math.square(mean))\n",
-    "    loss = K.mean(loss)\n",
-    "    return loss\n",
-    "\n",
-    "def normalize(input_image, real_image):\n",
-    "    input_image = (input_image / 127.5) - 1\n",
-    "    real_image = (real_image / 127.5) - 1\n",
-    "\n",
-    "    return input_image, real_image\n",
-    "\n",
-    "def load_class_ids_filenames(class_id_path, filename_path):\n",
-    "    with open(class_id_path, 'rb') as file:\n",
-    "        class_id = pickle.load(file, encoding='latin1')\n",
-    "\n",
-    "    with open(filename_path, 'rb') as file:\n",
-    "        filename = pickle.load(file, encoding='latin1')\n",
-    "\n",
-    "    return class_id, filename\n",
-    "\n",
-    "def load_text_embeddings(text_embeddings):\n",
-    "    with open(text_embeddings, 'rb') as file:\n",
-    "        embeds = pickle.load(file, encoding='latin1')\n",
-    "        embeds = np.array(embeds)\n",
-    "\n",
-    "    return embeds\n",
-    "\n",
-    "def load_bbox(data_path):\n",
-    "    bbox_path = data_path + '/bounding_boxes.txt'\n",
-    "    image_path = data_path + '/images.txt'\n",
-    "    bbox_df = pd.read_csv(bbox_path, delim_whitespace=True, header=None).astype(int)\n",
-    "    filename_df = pd.read_csv(image_path, delim_whitespace=True, header=None)\n",
-    "\n",
-    "    filenames = filename_df[1].tolist()\n",
-    "    bbox_dict = {i[:-4]:[] for i in filenames[:2]}\n",
-    "\n",
-    "    for i in range(0, len(filenames)):\n",
-    "        bbox = bbox_df.iloc[i][1:].tolist()\n",
-    "        dict_key = filenames[i][:-4]\n",
-    "        bbox_dict[dict_key] = bbox\n",
-    "\n",
-    "    return bbox_dict\n",
-    "\n",
-    "def load_images(image_path, bounding_box, size):\n",
-    "    \"\"\"Crops the image to the bounding box and then resizes it.\n",
-    "    \"\"\"\n",
-    "    image = Image.open(image_path).convert('RGB')\n",
-    "    w, h = image.size\n",
-    "    if bounding_box is not None:\n",
-    "        r = int(np.maximum(bounding_box[2], bounding_box[3]) * 0.75)\n",
-    "        c_x = int((bounding_box[0] + bounding_box[2]) / 2)\n",
-    "        c_y = int((bounding_box[1] + bounding_box[3]) / 2)\n",
-    "        y1 = np.maximum(0, c_y - r)\n",
-    "        y2 = np.minimum(h, c_y + r)\n",
-    "        x1 = np.maximum(0, c_x - r)\n",
-    "        x2 = np.minimum(w, c_x + r)\n",
-    "        image = image.crop([x1, y1, x2, y2])\n",
-    "\n",
-    "    image = image.resize(size, PIL.Image.BILINEAR)\n",
-    "    return image\n",
-    "\n",
-    "def load_data(filename_path, class_id_path, dataset_path, embeddings_path, size):\n",
-    "    \"\"\"Loads the Dataset.\n",
-    "    \"\"\"           \n",
-    "    data_dir = \"D:\\\\unikaksha\\\\CUB_200_2011\\\\CUB_200_2011\\\\birds\\\\birds\"\n",
-    "    train_dir = data_dir + \"\\\\train\"\n",
-    "    test_dir = data_dir + \"\\\\test\"\n",
-    "    embeddings_path_train = train_dir + \"\\\\char-CNN-RNN-embeddings.pickle\"\n",
-    "    embeddings_path_test = test_dir + \"\\\\char-CNN-RNN-embeddings.pickle\"\n",
-    "    filename_path_train = train_dir + \"\\\\filenames.pickle\"\n",
-    "    filename_path_test = test_dir + \"\\\\filenames.pickle\"\n",
-    "    class_id_path_train = train_dir + \"\\\\class_info.pickle\"\n",
-    "    class_id_path_test = test_dir + \"\\\\class_info.pickle\"\n",
-    "    dataset_path = \"D:\\\\unikaksha\\\\CUB_200_2011\\\\CUB_200_2011\"\n",
-    "    class_id, filenames = load_class_ids_filenames(class_id_path, filename_path)\n",
-    "    embeddings = load_text_embeddings(embeddings_path)\n",
-    "    bbox_dict = load_bbox(dataset_path)\n",
-    "\n",
-    "    x, y, embeds = [], [], []\n",
-    "\n",
-    "    for i, filename in enumerate(filenames):\n",
-    "        bbox = bbox_dict[filename]\n",
-    "\n",
-    "        try:\n",
-    "            image_path = f'{dataset_path}\\\\images\\\\{filename}.jpg'\n",
-    "            image = load_images(image_path, bbox, size)\n",
-    "            e = embeddings[i, :, :]\n",
-    "            embed_index = np.random.randint(0, e.shape[0] - 1)\n",
-    "            embed = e[embed_index, :]\n",
-    "            x.append(np.array(image))\n",
-    "            y.append(class_id[i])\n",
-    "            embeds.append(embed)\n",
-    "\n",
-    "        except Exception as e:\n",
-    "            print(f'{e}')\n",
-    "\n",
-    "    x = np.array(x)\n",
-    "    y = np.array(y)\n",
-    "    embeds = np.array(embeds)\n",
-    "    \n",
-    "    return x, y, embeds\n",
-    "\n",
-    "def save_image(file, save_path):\n",
-    "    \"\"\"Saves the image at the specified file path.\n",
-    "    \"\"\"\n",
-    "    image = plt.figure()\n",
-    "    ax = image.add_subplot(1,1,1)\n",
-    "    ax.imshow(file)\n",
-    "    ax.axis(\"off\")\n",
-    "    plt.savefig(save_path)"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 20,
-   "id": "8bb3b1bc",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "############################################################\n",
-    "# StackGAN class\n",
-    "############################################################\n",
-    "\n",
-    "class StackGanStage1(object):\n",
-    "    \"\"\"StackGAN Stage 1 class.\"\"\"\n",
-    "    data_dir = \"D:\\\\unikaksha\\\\CUB_200_2011\\\\CUB_200_2011\\\\birds\\\\birds\"\n",
-    "    train_dir = data_dir+\"\\\\train\"\n",
-    "    test_dir = data_dir+\"\\\\test\"\n",
-    "    embeddings_path_train = train_dir + \"\\\\char-CNN-RNN-embeddings.pickle\"\n",
-    "    embeddings_path_test = test_dir +\"\\\\char-CNN-RNN-embeddings.pickle\"\n",
-    "    filename_path_train = train_dir+\"\\\\filenames.pickle\"\n",
-    "    filename_path_test = test_dir+\"\\\\filenames.pickle\"\n",
-    "    class_id_path_train = train_dir+\"\\\\class_info.pickle\"\n",
-    "    class_id_path_test = test_dir+\"\\\\class_info.pickle\"\n",
-    "    dataset_path = \"D:\\\\unikaksha\\\\CUB_200_2011\\\\CUB_200_2011\"\n",
-    "    def __init__(self, epochs=500, z_dim=100, batch_size=64, enable_function=True, stage1_generator_lr=0.0002, stage1_discriminator_lr=0.0002):\n",
-    "        self.epochs = epochs\n",
-    "        self.z_dim = z_dim\n",
-    "        self.enable_function = enable_function\n",
-    "        self.stage1_generator_lr = stage1_generator_lr\n",
-    "        self.stage1_discriminator_lr = stage1_discriminator_lr\n",
-    "        self.image_size = 64\n",
-    "        self.conditioning_dim = 128\n",
-    "        self.batch_size = batch_size\n",
-    "        \n",
-    "        self.stage1_generator_optimizer = Adam(lr=stage1_generator_lr, beta_1=0.5, beta_2=0.999)\n",
-    "        self.stage1_discriminator_optimizer = Adam(lr=stage1_discriminator_lr, beta_1=0.5, beta_2=0.999)\n",
-    "        self.stage1_generator = build_stage1_generator()\n",
-    "        self.stage1_generator.compile(loss='mse', optimizer=self.stage1_generator_optimizer)\n",
-    "\n",
-    "        self.stage1_discriminator = build_stage1_discriminator()\n",
-    "        self.stage1_discriminator.compile(loss='binary_crossentropy', optimizer=self.stage1_discriminator_optimizer)\n",
-    "\n",
-    "        self.ca_network = build_ca_network()\n",
-    "        self.ca_network.compile(loss='binary_crossentropy', optimizer='Adam')\n",
-    "\n",
-    "        self.embedding_compressor = build_embedding_compressor()\n",
-    "        self.embedding_compressor.compile(loss='binary_crossentropy', optimizer='Adam')\n",
-    "\n",
-    "        self.stage1_adversarial = build_adversarial(self.stage1_generator, self.stage1_discriminator)\n",
-    "        self.stage1_adversarial.compile(loss=['binary_crossentropy', adversarial_loss], loss_weights=[1, 2.0], optimizer=self.stage1_generator_optimizer)\n",
-    "\n",
-    "        self.checkpoint1 = tf.train.Checkpoint(\n",
-    "            generator_optimizer=self.stage1_generator_optimizer,\n",
-    "            discriminator_optimizer=self.stage1_discriminator_optimizer,\n",
-    "            generator=self.stage1_generator,\n",
-    "            discriminator=self.stage1_discriminator)\n",
-    "    def visualize_stage1(self):\n",
-    "        \"\"\"Running Tensorboard visualizations.\"\"\"\n",
-    "        tb = TensorBoard(log_dir=\"logs/\".format(time.time()))\n",
-    "        tb.set_model(self.stage1_generator)\n",
-    "        tb.set_model(self.stage1_discriminator)\n",
-    "        tb.set_model(self.ca_network)\n",
-    "        tb.set_model(self.embedding_compressor)\n",
-    "        \n",
-    "    def train_stage1(self):\n",
-    "        \"\"\"Trains the stage1 StackGAN.\"\"\"\n",
-    "        x_train, y_train, train_embeds = load_data(filename_path = \"D:\\\\unikaksha\\\\CUB_200_2011\\\\CUB_200_2011\\\\birds\\\\birds\\\\train\\\\filenames.pickle\", \n",
-    "        class_id_path=\"D:\\\\unikaksha\\\\CUB_200_2011\\\\CUB_200_2011\\\\birds\\\\birds\\\\train\\\\class_info.pickle\",\n",
-    "        dataset_path=\"D:\\\\unikaksha\\\\CUB_200_2011\\\\CUB_200_2011\", \n",
-    "        embeddings_path=\"D:\\\\unikaksha\\\\CUB_200_2011\\\\CUB_200_2011\\\\birds\\\\birds\\\\train\\\\char-CNN-RNN-embeddings.pickle\", size=(64, 64))\n",
-    "\n",
-    "        x_test, y_test, test_embeds = load_data(filename_path=\"D:\\\\unikaksha\\\\CUB_200_2011\\\\CUB_200_2011\\\\birds\\\\birds\\\\test\\\\filenames.pickle\", \n",
-    "        class_id_path=\"D:\\\\unikaksha\\\\CUB_200_2011\\\\CUB_200_2011\\\\birds\\\\birds\\\\test\\\\class_info.pickle\", \n",
-    "        dataset_path=\"D:\\\\unikaksha\\\\CUB_200_2011\\\\CUB_200_2011\", \n",
-    "        embeddings_path=\"D:\\\\unikaksha\\\\CUB_200_2011\\\\CUB_200_2011\\\\birds\\\\birds\\\\test\\\\char-CNN-RNN-embeddings.pickle\", size=(64, 64))\n",
-    "\n",
-    "        real = np.ones((self.batch_size, 1), dtype='float') * 0.9\n",
-    "        fake = np.zeros((self.batch_size, 1), dtype='float') * 0.1\n",
-    "\n",
-    "        for epoch in range(self.epochs):\n",
-    "            print(f'Epoch: {epoch}')\n",
-    "\n",
-    "            gen_loss = []\n",
-    "            dis_loss = []\n",
-    "\n",
-    "            num_batches = int(x_train.shape[0] / self.batch_size)\n",
-    "\n",
-    "            for i in range(num_batches):\n",
-    "\n",
-    "                latent_space = np.random.normal(0, 1, size=(self.batch_size, self.z_dim))\n",
-    "                embedding_text = train_embeds[i * self.batch_size:(i + 1) * self.batch_size]\n",
-    "                compressed_embedding = self.embedding_compressor.predict_on_batch(embedding_text)\n",
-    "                compressed_embedding = np.reshape(compressed_embedding, (-1, 1, 1, 128))\n",
-    "                compressed_embedding = np.tile(compressed_embedding, (1, 4, 4, 1))\n",
-    "                image_batch = x_train[i * self.batch_size:(i+1) * self.batch_size]\n",
-    "                image_batch = (image_batch - 127.5) / 127.5\n",
-    "\n",
-    "                gen_images, _ = self.stage1_generator.predict([embedding_text, latent_space])\n",
-    "\n",
-    "                discriminator_loss = self.stage1_discriminator.train_on_batch([image_batch, compressed_embedding], \n",
-    "                    np.reshape(real, (self.batch_size, 1)))\n",
-    "\n",
-    "                discriminator_loss_gen = self.stage1_discriminator.train_on_batch([gen_images, compressed_embedding],\n",
-    "                    np.reshape(fake, (self.batch_size, 1)))\n",
-    "\n",
-    "                discriminator_loss_wrong = self.stage1_discriminator.train_on_batch([gen_images[: self.batch_size-1], compressed_embedding[1:]], \n",
-    "                    np.reshape(fake[1:], (self.batch_size-1, 1)))\n",
-    "\n",
-    "                # Discriminator loss\n",
-    "                d_loss = 0.5 * np.add(discriminator_loss, 0.5 * np.add(discriminator_loss_gen, discriminator_loss_wrong))\n",
-    "                dis_loss.append(d_loss)\n",
-    "\n",
-    "                print(f'Discriminator Loss: {d_loss}')\n",
-    "\n",
-    "                # Generator loss\n",
-    "                g_loss = self.stage1_adversarial.train_on_batch([embedding_text, latent_space, compressed_embedding],\n",
-    "                        [K.ones((self.batch_size, 1)) * 0.9, K.ones((self.batch_size, 256)) * 0.9])\n",
-    "\n",
-    "                print(f'Generator Loss: {g_loss}')\n",
-    "                gen_loss.append(g_loss)\n",
-    "                \n",
-    "                if epoch % 5 == 0:\n",
-    "                    latent_space = np.random.normal(0, 1, size=(self.batch_size, self.z_dim))\n",
-    "                    embedding_batch = test_embeds[0 : self.batch_size]\n",
-    "                    gen_images, _ = self.stage1_generator.predict_on_batch([embedding_batch, latent_space])\n",
-    "\n",
-    "                    for i, image in enumerate(gen_images[:10]):\n",
-    "                           save_image(image, f'D:\\\\unikaksha\\\\GAN_project\\\\test\\\\gen_1_{epoch}_{i}')\n",
-    "\n",
-    "                if epoch % 25 == 0:\n",
-    "                    self.stage1_generator.save_weights('D:\\\\unikaksha\\\\GAN_project\\\\weights\\\\stage1_gen.h5')\n",
-    "                    self.stage1_discriminator.save_weights(\"D:\\\\unikaksha\\\\GAN_project\\\\weights\\\\stage1_disc.h5\")\n",
-    "                    self.ca_network.save_weights('D:\\\\unikaksha\\\\GAN_project\\\\weights\\\\stage1_ca.h5')\n",
-    "                    self.embedding_compressor.save_weights('D:\\\\unikaksha\\\\GAN_project\\\\weights\\\\stage1_embco.h5')\n",
-    "                    self.stage1_adversarial.save_weights('D:\\\\unikaksha\\\\GAN_project\\\\weights\\\\stage1_adv.h5')      \n",
-    "\n",
-    "        self.stage1_generator.save_weights('D:\\\\unikaksha\\\\GAN_project\\\\weights\\\\stage1_gen.h5')\n",
-    "        self.stage1_discriminator.save_weights(\"D:\\\\unikaksha\\\\GAN_project\\\\weights\\\\stage1_disc.h5\")"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "606f27d1",
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "C:\\Users\\hp\\anaconda3\\lib\\site-packages\\keras\\optimizers\\legacy\\adam.py:117: UserWarning: The `lr` argument is deprecated, use `learning_rate` instead.\n",
-      "  super().__init__(name, **kwargs)\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Epoch: 0\n",
-      "2/2 [==============================] - 11s 4s/step\n",
-      "Discriminator Loss: 0.9802191257476807\n",
-      "Generator Loss: [0.6961640119552612, 0.6650141477584839, 0.015574917197227478]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.48184134624898434\n",
-      "Generator Loss: [0.5850249528884888, 0.5450800061225891, 0.01997246965765953]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.5333389556035399\n",
-      "Generator Loss: [0.8463906645774841, 0.8030149936676025, 0.0216878242790699]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "C:\\Users\\hp\\AppData\\Local\\Temp\\ipykernel_2676\\16722043.py:114: RuntimeWarning: More than 20 figures have been opened. Figures created through the pyplot interface (`matplotlib.pyplot.figure`) are retained until explicitly closed and may consume too much memory. (To control this warning, see the rcParam `figure.max_open_warning`). Consider using `matplotlib.pyplot.close()`.\n",
-      "  image = plt.figure()\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.47618325776420534\n",
-      "Generator Loss: [1.3540817499160767, 1.3058273792266846, 0.02412717416882515]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.576083542779088\n",
-      "Generator Loss: [3.0371224880218506, 3.000938892364502, 0.01809178665280342]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.4288549982011318\n",
-      "Generator Loss: [1.8396023511886597, 1.797579288482666, 0.021011507138609886]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.5725578819001385\n",
-      "Generator Loss: [1.3126457929611206, 1.280745267868042, 0.015950273722410202]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.8466892631258816\n",
-      "Generator Loss: [0.7987891435623169, 0.7408242225646973, 0.02898247353732586]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.3347012852318585\n",
-      "Generator Loss: [0.4714033603668213, 0.3795074224472046, 0.04594796895980835]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.23857683420646936\n",
-      "Generator Loss: [0.509975790977478, 0.4569609761238098, 0.026507411152124405]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.21621597849298269\n",
-      "Generator Loss: [0.39061087369918823, 0.3531549572944641, 0.01872795820236206]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.24724604620132595\n",
-      "Generator Loss: [0.4967734217643738, 0.4569697678089142, 0.019901830703020096]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.22956613145652227\n",
-      "Generator Loss: [1.522817850112915, 1.4894189834594727, 0.016699453815817833]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.3451330265670549\n",
-      "Generator Loss: [4.71159553527832, 4.686409950256348, 0.01259271427989006]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.31048156949691474\n",
-      "Generator Loss: [3.0208652019500732, 2.974703788757324, 0.023080747574567795]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 1.286377330245159\n",
-      "Generator Loss: [8.078065872192383, 8.033745765686035, 0.022160058841109276]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 1.2558328942977823\n",
-      "Generator Loss: [4.335520267486572, 4.294609069824219, 0.020455578342080116]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.25108036841265857\n",
-      "Generator Loss: [1.5224553346633911, 1.490272879600525, 0.0160912424325943]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.5868438822358257\n",
-      "Generator Loss: [7.595462799072266, 7.56856107711792, 0.013450761325657368]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.3842968652024865\n",
-      "Generator Loss: [3.696214437484741, 3.6685736179351807, 0.013820387423038483]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 1.0542606416472609\n",
-      "Generator Loss: [9.764802932739258, 9.73652458190918, 0.014139214530587196]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.3704390770290047\n",
-      "Generator Loss: [9.32541561126709, 9.293708801269531, 0.015853524208068848]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.3744450802914798\n",
-      "Generator Loss: [0.5127507448196411, 0.48108744621276855, 0.015831634402275085]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.20261134163592942\n",
-      "Generator Loss: [7.319120407104492, 7.2795610427856445, 0.019779693335294724]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.22154655796475708\n",
-      "Generator Loss: [0.5853765606880188, 0.5413299202919006, 0.022023331373929977]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.24848091672174633\n",
-      "Generator Loss: [0.4261373281478882, 0.39824241399765015, 0.01394745521247387]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.21633487555664033\n",
-      "Generator Loss: [0.5965468287467957, 0.5641502737998962, 0.016198277473449707]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.3055696014780551\n",
-      "Generator Loss: [0.5238827466964722, 0.46521663665771484, 0.02933306246995926]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.23254890751559287\n",
-      "Generator Loss: [0.4697083532810211, 0.41412609815597534, 0.027791127562522888]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.22449417394091142\n",
-      "Generator Loss: [0.5463337898254395, 0.508608341217041, 0.01886272057890892]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.2007127658380341\n",
-      "Generator Loss: [0.5559038519859314, 0.5166953802108765, 0.019604243338108063]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.22016859483119333\n",
-      "Generator Loss: [0.4250626862049103, 0.3736666738986969, 0.025698000565171242]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.19318957749055699\n",
-      "Generator Loss: [0.4450433850288391, 0.4040531516075134, 0.020495112985372543]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.20447796490043402\n",
-      "Generator Loss: [0.41742050647735596, 0.3691825270652771, 0.02411898598074913]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.19861718312313315\n",
-      "Generator Loss: [0.5030301809310913, 0.4620949327945709, 0.02046763338148594]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1929033644264564\n",
-      "Generator Loss: [0.5071523189544678, 0.4739231467247009, 0.01661457121372223]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.19967280932178255\n",
-      "Generator Loss: [0.4714427888393402, 0.43816399574279785, 0.016639400273561478]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.2031445961620193\n",
-      "Generator Loss: [0.3809151351451874, 0.34519755840301514, 0.01785879395902157]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1938752414425835\n",
-      "Generator Loss: [0.4101487100124359, 0.37675684690475464, 0.016695931553840637]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17759674845729023\n",
-      "Generator Loss: [0.37680017948150635, 0.3431282639503479, 0.01683596521615982]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1953955094795674\n",
-      "Generator Loss: [0.4451413154602051, 0.4083906412124634, 0.018375344574451447]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1866152752772905\n",
-      "Generator Loss: [0.7822967171669006, 0.7482183575630188, 0.017039190977811813]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.19993027742020786\n",
-      "Generator Loss: [0.4690878391265869, 0.41948825120925903, 0.02479979768395424]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.18905645408085547\n",
-      "Generator Loss: [0.5142366886138916, 0.4719697833061218, 0.02113344520330429]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.23101993306772783\n",
-      "Generator Loss: [0.427396297454834, 0.3792785704135895, 0.024058863520622253]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.20659372577210888\n",
-      "Generator Loss: [0.38844016194343567, 0.3470289707183838, 0.02070559188723564]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1901704624906415\n",
-      "Generator Loss: [0.38711920380592346, 0.34456369280815125, 0.02127775177359581]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.18783968497882597\n",
-      "Generator Loss: [0.3948618173599243, 0.3499179482460022, 0.02247193269431591]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1915252854814753\n",
-      "Generator Loss: [0.6549585461616516, 0.6057220101356506, 0.024618258699774742]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.18054950918303803\n",
-      "Generator Loss: [0.41941961646080017, 0.3768080174922943, 0.02130579948425293]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.18988633051048964\n",
-      "Generator Loss: [0.44272011518478394, 0.4087149500846863, 0.017002590000629425]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 11s 6s/step\n",
-      "Discriminator Loss: 0.18048399151302874\n",
-      "Generator Loss: [0.37595367431640625, 0.34316951036453247, 0.016392074525356293]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17496303527150303\n",
-      "Generator Loss: [0.40165096521377563, 0.3578139841556549, 0.021918490529060364]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17897942382842302\n",
-      "Generator Loss: [0.39302635192871094, 0.3575895130634308, 0.017718425020575523]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 14s 7s/step\n",
-      "Discriminator Loss: 0.17537035774148535\n",
-      "Generator Loss: [0.38119563460350037, 0.3450235426425934, 0.01808605156838894]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1780482066096738\n",
-      "Generator Loss: [0.38575857877731323, 0.3476724624633789, 0.019043058156967163]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.182727460982278\n",
-      "Generator Loss: [0.45607152581214905, 0.41492295265197754, 0.020574286580085754]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.18447332979849307\n",
-      "Generator Loss: [0.38758498430252075, 0.35072416067123413, 0.01843041181564331]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17860571423079818\n",
-      "Generator Loss: [0.37579774856567383, 0.34342193603515625, 0.01618790067732334]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1858925880806055\n",
-      "Generator Loss: [0.37908297777175903, 0.3442193567752838, 0.01743181422352791]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.18772859932505526\n",
-      "Generator Loss: [0.38528314232826233, 0.3587524890899658, 0.013265324756503105]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1808448277588468\n",
-      "Generator Loss: [0.43304237723350525, 0.4028497338294983, 0.015096323564648628]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17957381592714228\n",
-      "Generator Loss: [0.40007010102272034, 0.3626365661621094, 0.01871676743030548]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.18053699872689322\n",
-      "Generator Loss: [0.43529215455055237, 0.4006400406360626, 0.017326055094599724]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1845625478963484\n",
-      "Generator Loss: [0.5139749646186829, 0.47341838479042053, 0.02027830295264721]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17584427335532382\n",
-      "Generator Loss: [0.4796595275402069, 0.45155781507492065, 0.014050857163965702]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.19239969493355602\n",
-      "Generator Loss: [0.5465446710586548, 0.5206610560417175, 0.012941796332597733]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.18761410604929551\n",
-      "Generator Loss: [0.404424786567688, 0.36992087960243225, 0.01725194975733757]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17783832902932772\n",
-      "Generator Loss: [0.3823916018009186, 0.3543917238712311, 0.013999941758811474]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17266782574733952\n",
-      "Generator Loss: [0.37809306383132935, 0.34391820430755615, 0.017087422311306]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17270548808664898\n",
-      "Generator Loss: [0.6664740443229675, 0.6267739534378052, 0.01985004171729088]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1800534851208795\n",
-      "Generator Loss: [0.37835797667503357, 0.3481042981147766, 0.015126839280128479]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17871860484592617\n",
-      "Generator Loss: [0.4661772549152374, 0.43449288606643677, 0.015842188149690628]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.18629638472339138\n",
-      "Generator Loss: [0.5019238591194153, 0.4734560251235962, 0.014233909547328949]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.180536136685987\n",
-      "Generator Loss: [0.38149377703666687, 0.35071855783462524, 0.01538760308176279]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17764301746501587\n",
-      "Generator Loss: [0.40358954668045044, 0.3741263151168823, 0.014731617644429207]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17080012267979328\n",
-      "Generator Loss: [0.37310791015625, 0.3462904989719391, 0.013408699072897434]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17130548738350626\n",
-      "Generator Loss: [0.47454968094825745, 0.44810569286346436, 0.013221989385783672]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.18367330575711094\n",
-      "Generator Loss: [0.6813294887542725, 0.6486695408821106, 0.016329985111951828]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1897322643053485\n",
-      "Generator Loss: [0.38306909799575806, 0.3530800938606262, 0.01499450858682394]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1736138232954545\n",
-      "Generator Loss: [0.469137966632843, 0.4357489347457886, 0.01669451594352722]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1746481669106288\n",
-      "Generator Loss: [0.410297691822052, 0.38105642795562744, 0.014620626345276833]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.1852193692830042\n",
-      "Generator Loss: [0.5768088698387146, 0.5471887588500977, 0.014810062013566494]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17311640443222132\n",
-      "Generator Loss: [0.5264151692390442, 0.4969061613082886, 0.014754515141248703]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16888699869741686\n",
-      "Generator Loss: [0.6626132726669312, 0.642403244972229, 0.01010502502322197]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16971732130332384\n",
-      "Generator Loss: [0.392791748046875, 0.36890941858291626, 0.01194116659462452]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1667094981239643\n",
-      "Generator Loss: [0.4054604768753052, 0.3775497376918793, 0.013955371454358101]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16862681882048491\n",
-      "Generator Loss: [0.40447133779525757, 0.374853253364563, 0.014809049665927887]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17563658353174105\n",
-      "Generator Loss: [0.5058242082595825, 0.4704281687736511, 0.017698023468255997]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17933404936411534\n",
-      "Generator Loss: [0.3794536888599396, 0.34335780143737793, 0.018047945573925972]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1693830450058158\n",
-      "Generator Loss: [0.4312605559825897, 0.40319862961769104, 0.014030968770384789]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17556392493861495\n",
-      "Generator Loss: [0.44084155559539795, 0.39919477701187134, 0.020823394879698753]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17348072211098042\n",
-      "Generator Loss: [0.7974593639373779, 0.7602275609970093, 0.01861589401960373]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1765500687979511\n",
-      "Generator Loss: [0.4340978264808655, 0.39812541007995605, 0.01798621378839016]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16941886499989778\n",
-      "Generator Loss: [0.4198617935180664, 0.3936614990234375, 0.013100153766572475]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1696752424177248\n",
-      "Generator Loss: [0.3989488482475281, 0.36275243759155273, 0.018098212778568268]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17254859328386374\n",
-      "Generator Loss: [0.6144299507141113, 0.5775778293609619, 0.018426064401865005]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1864816747547593\n",
-      "Generator Loss: [0.3965364396572113, 0.3642270565032959, 0.016154695302248]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1734401921639801\n",
-      "Generator Loss: [0.5290278792381287, 0.4967009723186493, 0.01616346649825573]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.172534832097881\n",
-      "Generator Loss: [0.42191481590270996, 0.39195775985717773, 0.014978524297475815]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.168708567682188\n",
-      "Generator Loss: [0.9239866137504578, 0.9011085629463196, 0.011439025402069092]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17294310401484836\n",
-      "Generator Loss: [0.5620108842849731, 0.5333402752876282, 0.014335304498672485]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17393866236670874\n",
-      "Generator Loss: [0.5168393850326538, 0.4787072539329529, 0.01906607486307621]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17645742278546095\n",
-      "Generator Loss: [0.37883812189102173, 0.3528320789337158, 0.013003027997910976]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17291862977435812\n",
-      "Generator Loss: [0.3965036869049072, 0.3713742196559906, 0.012564735487103462]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17384443146147532\n",
-      "Generator Loss: [0.38646000623703003, 0.36036360263824463, 0.01304820366203785]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17164584412239492\n",
-      "Generator Loss: [0.3994850516319275, 0.3680734634399414, 0.01570579782128334]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17196237533789827\n",
-      "Generator Loss: [0.3751278519630432, 0.3494475483894348, 0.0128401517868042]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16888764752729912\n",
-      "Generator Loss: [0.5003425478935242, 0.47001177072525024, 0.015165378339588642]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1720182311546523\n",
-      "Generator Loss: [0.37885040044784546, 0.3501875400543213, 0.014331430196762085]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17193801811663434\n",
-      "Generator Loss: [0.38392725586891174, 0.35279881954193115, 0.015564213506877422]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17428883849061094\n",
-      "Generator Loss: [0.4552914500236511, 0.41936779022216797, 0.017961828038096428]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17142450831306633\n",
-      "Generator Loss: [0.3909251093864441, 0.35883039236068726, 0.016047364100813866]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16852326712978538\n",
-      "Generator Loss: [0.4052236080169678, 0.37001973390579224, 0.01760193705558777]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1672885622137983\n",
-      "Generator Loss: [0.3810323178768158, 0.35226720571517944, 0.014382561668753624]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.18373181621427648\n",
-      "Generator Loss: [0.6394188404083252, 0.610511302947998, 0.01445375382900238]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.19671613787068054\n",
-      "Generator Loss: [0.37208110094070435, 0.34650278091430664, 0.012789154425263405]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1778201662018546\n",
-      "Generator Loss: [0.3983660042285919, 0.37093353271484375, 0.013716235756874084]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17354960551165277\n",
-      "Generator Loss: [0.4217820465564728, 0.39832568168640137, 0.011728182435035706]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17178144787612837\n",
-      "Generator Loss: [0.3721829950809479, 0.34306466579437256, 0.014559170231223106]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17107136059712502\n",
-      "Generator Loss: [0.3714357614517212, 0.34324324131011963, 0.014096265658736229]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16985933146861498\n",
-      "Generator Loss: [0.4401867389678955, 0.41247180104255676, 0.013857468962669373]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17393362350412644\n",
-      "Generator Loss: [0.5702003240585327, 0.5471206903457642, 0.011539814993739128]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16943995378096588\n",
-      "Generator Loss: [0.36107105016708374, 0.3407486379146576, 0.010161205194890499]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16973521593172336\n",
-      "Generator Loss: [0.38987433910369873, 0.36227506399154663, 0.013799636624753475]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16701886043301784\n",
-      "Generator Loss: [0.4122741222381592, 0.3849170207977295, 0.01367854792624712]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1716277181985788\n",
-      "Generator Loss: [0.5448815226554871, 0.5217555165290833, 0.011562997475266457]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1687299137483933\n",
-      "Generator Loss: [0.3949546217918396, 0.35768526792526245, 0.018634673207998276]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1687844840780599\n",
-      "Generator Loss: [0.4184192717075348, 0.3867832124233246, 0.015818025916814804]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16914064924640115\n",
-      "Generator Loss: [0.3805859088897705, 0.3574395179748535, 0.01157319638878107]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17298609676072374\n",
-      "Generator Loss: [0.441110223531723, 0.40844085812568665, 0.01633468084037304]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17360270822246093\n",
-      "Generator Loss: [0.36695006489753723, 0.3402103781700134, 0.013369848020374775]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16781856885063462\n",
-      "Generator Loss: [0.36394211649894714, 0.3446100950241089, 0.00966600887477398]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16979453315434512\n",
-      "Generator Loss: [0.4049004018306732, 0.3830277621746063, 0.010936319828033447]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17385048950382043\n",
-      "Generator Loss: [0.3585967719554901, 0.33945441246032715, 0.009571176022291183]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1694723754299048\n",
-      "Generator Loss: [0.42190027236938477, 0.4019508957862854, 0.009974684566259384]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.169932454758964\n",
-      "Generator Loss: [0.4543371796607971, 0.434335857629776, 0.010000664740800858]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16698016336158616\n",
-      "Generator Loss: [0.3634415566921234, 0.3408564627170563, 0.011292543262243271]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Epoch: 1\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17379584628361044\n",
-      "Generator Loss: [0.3577522039413452, 0.3399486541748047, 0.008901769295334816]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16611875274247723\n",
-      "Generator Loss: [0.4083532989025116, 0.3820931315422058, 0.013130085542798042]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16981480410322547\n",
-      "Generator Loss: [0.45322564244270325, 0.4259665310382843, 0.013629553839564323]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17780617461539805\n",
-      "Generator Loss: [0.41947945952415466, 0.3858738839626312, 0.016802793368697166]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1676520689507015\n",
-      "Generator Loss: [0.4091261923313141, 0.38311684131622314, 0.013004681095480919]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17036597337573767\n",
-      "Generator Loss: [0.43128111958503723, 0.4036552309989929, 0.013812937773764133]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1734127685194835\n",
-      "Generator Loss: [0.3618379533290863, 0.3409835994243622, 0.01042717881500721]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16824281342996983\n",
-      "Generator Loss: [0.4871261715888977, 0.45516437292099, 0.015980897471308708]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16999924747506157\n",
-      "Generator Loss: [0.41793525218963623, 0.37026557326316833, 0.023834839463233948]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17055944178719074\n",
-      "Generator Loss: [0.532861590385437, 0.5018796920776367, 0.015490951016545296]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17455832532141358\n",
-      "Generator Loss: [0.36303550004959106, 0.34091925621032715, 0.011058129370212555]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17146092635812238\n",
-      "Generator Loss: [0.3867221474647522, 0.3636346459388733, 0.011543749831616879]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16798607182863634\n",
-      "Generator Loss: [0.3572956621646881, 0.3360900282859802, 0.01060281042009592]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17161258905252907\n",
-      "Generator Loss: [0.3749561011791229, 0.35880303382873535, 0.008076537400484085]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17501034046290442\n",
-      "Generator Loss: [0.38381049036979675, 0.3527490198612213, 0.015530729666352272]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16860331935458817\n",
-      "Generator Loss: [0.5785412192344666, 0.5471409559249878, 0.015700142830610275]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16870543877303135\n",
-      "Generator Loss: [0.38838228583335876, 0.35851335525512695, 0.014934462495148182]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17488257051445544\n",
-      "Generator Loss: [0.4516022205352783, 0.42904067039489746, 0.011280778795480728]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17732026291196235\n",
-      "Generator Loss: [0.3619823157787323, 0.3441754877567291, 0.008903414011001587]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16745108766099293\n",
-      "Generator Loss: [0.45562463998794556, 0.4370197355747223, 0.009302457794547081]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16957885617011925\n",
-      "Generator Loss: [0.4287045896053314, 0.4118785262107849, 0.008413037285208702]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17324092709168326\n",
-      "Generator Loss: [0.39898210763931274, 0.3806677460670471, 0.009157184511423111]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1689320922087063\n",
-      "Generator Loss: [0.44481930136680603, 0.4176158010959625, 0.013601753860712051]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.16674932969544898\n",
-      "Generator Loss: [0.35118207335472107, 0.33533406257629395, 0.00792401097714901]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16819992047385313\n",
-      "Generator Loss: [0.3576042652130127, 0.3392186164855957, 0.009192829951643944]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16882829683527234\n",
-      "Generator Loss: [0.36622482538223267, 0.33876460790634155, 0.013730110600590706]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16585830420081038\n",
-      "Generator Loss: [0.36496609449386597, 0.3400167226791382, 0.012474680319428444]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1666502299503918\n",
-      "Generator Loss: [0.38084983825683594, 0.35977745056152344, 0.010536201298236847]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16603614990890492\n",
-      "Generator Loss: [0.3559083938598633, 0.33472976088523865, 0.010589315555989742]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16626825813000323\n",
-      "Generator Loss: [0.3608625531196594, 0.3350242078304291, 0.012919175438582897]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16647073911008192\n",
-      "Generator Loss: [0.3690946102142334, 0.34920641779899597, 0.009944096207618713]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1692058196349535\n",
-      "Generator Loss: [0.3625371754169464, 0.3365596830844879, 0.012988751754164696]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16997536386770662\n",
-      "Generator Loss: [0.3732745945453644, 0.3506496548652649, 0.011312464252114296]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16726714104879647\n",
-      "Generator Loss: [0.36841392517089844, 0.35006529092788696, 0.009174312464892864]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16600715326785576\n",
-      "Generator Loss: [0.3570045828819275, 0.3393431007862091, 0.00883074477314949]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1694513866386842\n",
-      "Generator Loss: [0.3895840644836426, 0.3698331117630005, 0.009875471703708172]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17276171086996328\n",
-      "Generator Loss: [0.5453651547431946, 0.5277018547058105, 0.008831647224724293]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16741786483908072\n",
-      "Generator Loss: [0.3843879699707031, 0.3657613694667816, 0.009313306771218777]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16834776167524979\n",
-      "Generator Loss: [0.36800047755241394, 0.34856563806533813, 0.009717423468828201]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16623327980050817\n",
-      "Generator Loss: [0.4009181261062622, 0.38308653235435486, 0.008915800601243973]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16928789918893017\n",
-      "Generator Loss: [0.3879125118255615, 0.36424922943115234, 0.011831637471914291]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17127257051470224\n",
-      "Generator Loss: [0.36610957980155945, 0.34529200196266174, 0.010408787988126278]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1690117270045448\n",
-      "Generator Loss: [0.4056251645088196, 0.38258737325668335, 0.011518890038132668]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16697075522279192\n",
-      "Generator Loss: [0.36058303713798523, 0.3400340676307678, 0.010274489410221577]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16551596325371065\n",
-      "Generator Loss: [0.3658486306667328, 0.34474658966064453, 0.010551026090979576]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1682825191601296\n",
-      "Generator Loss: [0.40609288215637207, 0.38417357206344604, 0.010959654115140438]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1743990409959224\n",
-      "Generator Loss: [0.36469343304634094, 0.3427026867866516, 0.010995374992489815]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16745789357446483\n",
-      "Generator Loss: [0.8354568481445312, 0.8140883445739746, 0.010684258304536343]\n",
-      "2/2 [==============================] - 9s 5s/step\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Discriminator Loss: 0.16982984548667446\n",
-      "Generator Loss: [0.3964448869228363, 0.3794506788253784, 0.008497098460793495]\n",
-      "2/2 [==============================] - 12s 6s/step\n",
-      "Discriminator Loss: 0.16938759066397324\n",
-      "Generator Loss: [0.6697938442230225, 0.6531023979187012, 0.008345713838934898]\n",
-      "2/2 [==============================] - 17s 8s/step\n",
-      "Discriminator Loss: 0.16977370387758128\n",
-      "Generator Loss: [2.859462022781372, 2.837353229522705, 0.01105437520891428]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17099883726041298\n",
-      "Generator Loss: [3.4141578674316406, 3.3959004878997803, 0.009128682315349579]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.2780838608744102\n",
-      "Generator Loss: [57.5842399597168, 57.56536865234375, 0.009435132145881653]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.7669053851685135\n",
-      "Generator Loss: [7.911134719848633, 7.892341613769531, 0.00939661543816328]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 1.3117623925209125\n",
-      "Generator Loss: [98.03166198730469, 98.00970458984375, 0.010979888029396534]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 2.4056636691094675\n",
-      "Generator Loss: [117.64867401123047, 117.62506866455078, 0.011801732704043388]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 2.4066649801018727\n",
-      "Generator Loss: [16.728708267211914, 16.70532989501953, 0.011689531616866589]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.5045298673212528\n",
-      "Generator Loss: [3.912365198135376, 3.8843350410461426, 0.014015069231390953]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 1.932625098338633\n",
-      "Generator Loss: [70.59149932861328, 70.57035827636719, 0.010569126345217228]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.8763693570507485\n",
-      "Generator Loss: [1.36116361618042, 1.3378283977508545, 0.011667625047266483]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.6450652023563634\n",
-      "Generator Loss: [3.319716691970825, 3.2867541313171387, 0.01648123189806938]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.38918108369398396\n",
-      "Generator Loss: [2.310241937637329, 2.275181293487549, 0.017530322074890137]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.2996680945507251\n",
-      "Generator Loss: [1.600480556488037, 1.5529499053955078, 0.023765303194522858]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.24070060866506537\n",
-      "Generator Loss: [9.476018905639648, 9.444986343383789, 0.01551638450473547]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.23392572251032107\n",
-      "Generator Loss: [1.9115113019943237, 1.8803954124450684, 0.015557916834950447]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.291397491004318\n",
-      "Generator Loss: [1.2948921918869019, 1.256650686264038, 0.01912074163556099]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.2415946024702862\n",
-      "Generator Loss: [1.5137697458267212, 1.4859565496444702, 0.013906599953770638]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.21148138132411987\n",
-      "Generator Loss: [1.355860710144043, 1.321687936782837, 0.017086386680603027]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.21975957024551462\n",
-      "Generator Loss: [2.590587854385376, 2.5471811294555664, 0.021703362464904785]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.19568662412348203\n",
-      "Generator Loss: [1.1067765951156616, 1.0722326040267944, 0.01727200858294964]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.19506385484419297\n",
-      "Generator Loss: [0.7975639700889587, 0.7632203102111816, 0.017171820625662804]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.2020061547809746\n",
-      "Generator Loss: [0.6563003063201904, 0.6226193308830261, 0.016840478405356407]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.19999883970012888\n",
-      "Generator Loss: [0.7539732456207275, 0.7181577682495117, 0.017907749861478806]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.19174541547545232\n",
-      "Generator Loss: [0.5636287927627563, 0.5262734293937683, 0.018677694723010063]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.22542133372917306\n",
-      "Generator Loss: [3.391948938369751, 3.3553080558776855, 0.01832042634487152]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.19922963762655854\n",
-      "Generator Loss: [1.0114272832870483, 0.9722586870193481, 0.01958431862294674]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.37087219534487303\n",
-      "Generator Loss: [18.71353530883789, 18.668882369995117, 0.022326193749904633]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.47285292346964525\n",
-      "Generator Loss: [0.829033613204956, 0.7850112915039062, 0.022011147812008858]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.22397881108918227\n",
-      "Generator Loss: [0.5942806005477905, 0.5497615337371826, 0.02225951850414276]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.21813420730177313\n",
-      "Generator Loss: [0.6996952295303345, 0.6561023592948914, 0.02179642766714096]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.20843163965037093\n",
-      "Generator Loss: [0.8573278784751892, 0.8159025311470032, 0.020712662488222122]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.18834336395957507\n",
-      "Generator Loss: [1.027643084526062, 0.9788804054260254, 0.024381328374147415]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1923854746928555\n",
-      "Generator Loss: [0.7623516321182251, 0.7282545566558838, 0.017048530280590057]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.18355153198353946\n",
-      "Generator Loss: [1.1679872274398804, 1.1298621892929077, 0.01906249113380909]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.20846794269164093\n",
-      "Generator Loss: [0.8024383187294006, 0.7683694958686829, 0.017034415155649185]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1928528374992311\n",
-      "Generator Loss: [0.6784680485725403, 0.6346690654754639, 0.021899502724409103]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1941780221532099\n",
-      "Generator Loss: [0.5794422030448914, 0.5382218956947327, 0.02061016485095024]\n",
-      "2/2 [==============================] - 11s 6s/step\n",
-      "Discriminator Loss: 0.21935425116680562\n",
-      "Generator Loss: [0.904646635055542, 0.8602195382118225, 0.022213557735085487]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1911653063216363\n",
-      "Generator Loss: [0.9832044839859009, 0.9462462663650513, 0.018479112535715103]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.2102399493851408\n",
-      "Generator Loss: [0.500117838382721, 0.46038636565208435, 0.019865740090608597]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.19488915757392533\n",
-      "Generator Loss: [0.6013811826705933, 0.5658959746360779, 0.017742611467838287]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1934375031851232\n",
-      "Generator Loss: [0.6782049536705017, 0.639863908290863, 0.01917050965130329]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.19082667381735519\n",
-      "Generator Loss: [0.44502174854278564, 0.4156520962715149, 0.014684827998280525]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17873958055861294\n",
-      "Generator Loss: [0.5339272618293762, 0.4944712817668915, 0.019728001207113266]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1849749393222737\n",
-      "Generator Loss: [0.4486980438232422, 0.4077763259410858, 0.020460866391658783]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1957180324716319\n",
-      "Generator Loss: [0.46563097834587097, 0.4288291335105896, 0.018400920554995537]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.18859698448795825\n",
-      "Generator Loss: [1.5433698892593384, 1.506518840789795, 0.018425529822707176]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17955090016766917\n",
-      "Generator Loss: [0.929273784160614, 0.8845124244689941, 0.02238067053258419]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.18189341621473432\n",
-      "Generator Loss: [0.5790578722953796, 0.5413727760314941, 0.018842538818717003]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17209526765509509\n",
-      "Generator Loss: [0.75950688123703, 0.7222777009010315, 0.018614578992128372]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.20075093463310623\n",
-      "Generator Loss: [0.6780570149421692, 0.6358927488327026, 0.02108212560415268]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.20808955523534678\n",
-      "Generator Loss: [0.5804647207260132, 0.5453739166259766, 0.017545390874147415]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1911735334142577\n",
-      "Generator Loss: [0.42684778571128845, 0.3924323320388794, 0.017207732424139977]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.19011279923142865\n",
-      "Generator Loss: [0.5841661691665649, 0.5475425124168396, 0.018311813473701477]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17914236274373252\n",
-      "Generator Loss: [0.5538751482963562, 0.508226215839386, 0.022824455052614212]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17397230066126212\n",
-      "Generator Loss: [0.9562854766845703, 0.9217334985733032, 0.0172759797424078]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.19711022335104644\n",
-      "Generator Loss: [0.49292683601379395, 0.45854172110557556, 0.017192551866173744]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.21020682185189798\n",
-      "Generator Loss: [0.7248889803886414, 0.6872766017913818, 0.01880619116127491]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.19582593758650546\n",
-      "Generator Loss: [0.5334466695785522, 0.48851144313812256, 0.022467613220214844]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.19626506276472355\n",
-      "Generator Loss: [0.47206273674964905, 0.4310867190361023, 0.020488012582063675]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.18287671820507967\n",
-      "Generator Loss: [0.4406943917274475, 0.39746785163879395, 0.021613268181681633]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.18299193761959032\n",
-      "Generator Loss: [0.5694998502731323, 0.5162140727043152, 0.026642892509698868]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17867000850128534\n",
-      "Generator Loss: [0.3978949189186096, 0.35563111305236816, 0.021131902933120728]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.19034564765752293\n",
-      "Generator Loss: [0.4693150520324707, 0.43242377042770386, 0.018445642665028572]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17805558450345416\n",
-      "Generator Loss: [0.401444673538208, 0.36052632331848145, 0.020459182560443878]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17494859670841834\n",
-      "Generator Loss: [0.39916127920150757, 0.36085131764411926, 0.019154977053403854]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17373917216900736\n",
-      "Generator Loss: [0.3969933092594147, 0.3601321280002594, 0.018430590629577637]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17829574846655305\n",
-      "Generator Loss: [0.3922646939754486, 0.3397202491760254, 0.02627221867442131]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.18334583344403654\n",
-      "Generator Loss: [0.3907313048839569, 0.34553292393684387, 0.02259918488562107]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1827138103544712\n",
-      "Generator Loss: [0.3840177059173584, 0.34055662155151367, 0.021730545908212662]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1866806780162733\n",
-      "Generator Loss: [0.37248528003692627, 0.3365044593811035, 0.017990415915846825]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.18573935534368502\n",
-      "Generator Loss: [0.4653454124927521, 0.43093734979629517, 0.017204027622938156]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17367476874642307\n",
-      "Generator Loss: [0.38152605295181274, 0.3407599925994873, 0.02038303017616272]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1732445124653168\n",
-      "Generator Loss: [0.37900397181510925, 0.33181267976760864, 0.023595646023750305]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17894858323415974\n",
-      "Generator Loss: [0.37332186102867126, 0.333858847618103, 0.019731512293219566]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17362131422851235\n",
-      "Generator Loss: [0.4131591022014618, 0.37411391735076904, 0.019522588700056076]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1760389762930572\n",
-      "Generator Loss: [0.38602253794670105, 0.34828853607177734, 0.018866997212171555]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1742742203787202\n",
-      "Generator Loss: [0.3729938566684723, 0.3387024700641632, 0.017145689576864243]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17380959934962448\n",
-      "Generator Loss: [0.37625303864479065, 0.33999109268188477, 0.018130967393517494]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17362700976082124\n",
-      "Generator Loss: [0.3813452422618866, 0.34442049264907837, 0.018462369218468666]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16966157831484452\n",
-      "Generator Loss: [0.3872207999229431, 0.35360291600227356, 0.01680893450975418]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17119372548768297\n",
-      "Generator Loss: [0.3900514841079712, 0.34923720359802246, 0.020407143980264664]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16921486234059557\n",
-      "Generator Loss: [0.5765373706817627, 0.53990638256073, 0.018315482884645462]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16737899294821545\n",
-      "Generator Loss: [0.43262147903442383, 0.3989427983760834, 0.016839340329170227]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16755579988239333\n",
-      "Generator Loss: [0.4554312527179718, 0.4190875291824341, 0.01817186549305916]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1797102899872698\n",
-      "Generator Loss: [0.3817053437232971, 0.3395741879940033, 0.021065576002001762]\n",
-      "Epoch: 2\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17401169703225605\n",
-      "Generator Loss: [0.3594488799571991, 0.33190321922302246, 0.013772835023701191]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17020215615048073\n",
-      "Generator Loss: [0.3816537857055664, 0.3484307527542114, 0.016611512750387192]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17633649520576\n",
-      "Generator Loss: [0.49695950746536255, 0.4622911810874939, 0.017334168776869774]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17027720558689907\n",
-      "Generator Loss: [0.387942373752594, 0.34857267141342163, 0.01968485116958618]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17372779792640358\n",
-      "Generator Loss: [0.388113796710968, 0.35283851623535156, 0.01763763278722763]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17165554815437645\n",
-      "Generator Loss: [0.3754979372024536, 0.33334729075431824, 0.02107531763613224]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16941950950422324\n",
-      "Generator Loss: [0.6505778431892395, 0.6192938089370728, 0.015642013400793076]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17471573024522513\n",
-      "Generator Loss: [0.5650069117546082, 0.5236581563949585, 0.02067437767982483]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17937259227619506\n",
-      "Generator Loss: [0.3953009843826294, 0.3429007828235626, 0.026200104504823685]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1763491338933818\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Generator Loss: [0.3986036777496338, 0.35978782176971436, 0.01940792053937912]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17681792378425598\n",
-      "Generator Loss: [0.37744131684303284, 0.34528040885925293, 0.016080450266599655]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17242084023018833\n",
-      "Generator Loss: [0.4094739556312561, 0.37326478958129883, 0.018104588612914085]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1713061386544723\n",
-      "Generator Loss: [0.38580119609832764, 0.3499787449836731, 0.017911231145262718]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17677571822423488\n",
-      "Generator Loss: [0.4177228510379791, 0.38673728704452515, 0.015492783859372139]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17192254995461553\n",
-      "Generator Loss: [0.45283180475234985, 0.4092029631137848, 0.02181442454457283]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1678741603027447\n",
-      "Generator Loss: [0.49263930320739746, 0.450339674949646, 0.021149812266230583]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1677050193575269\n",
-      "Generator Loss: [0.5579888224601746, 0.5233901739120483, 0.01729932799935341]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1690221366006881\n",
-      "Generator Loss: [0.3650253713130951, 0.33437302708625793, 0.015326175838708878]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16804369918827433\n",
-      "Generator Loss: [0.3849768340587616, 0.3567400574684143, 0.014118393883109093]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17360063971136697\n",
-      "Generator Loss: [0.3705819845199585, 0.3440471887588501, 0.013267393223941326]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17779486469225958\n",
-      "Generator Loss: [0.4751960337162018, 0.44869810342788696, 0.013248967006802559]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16928386507788673\n",
-      "Generator Loss: [1.1672372817993164, 1.1389933824539185, 0.014121958054602146]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16811944329128892\n",
-      "Generator Loss: [0.8764702677726746, 0.8488088250160217, 0.013830726966261864]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16715763500906178\n",
-      "Generator Loss: [0.4922735095024109, 0.4616849422454834, 0.015294278040528297]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17516483599320054\n",
-      "Generator Loss: [0.46365341544151306, 0.4276120066642761, 0.01802069880068302]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17063728647190146\n",
-      "Generator Loss: [0.44965311884880066, 0.4254313111305237, 0.012110909447073936]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17863933666376397\n",
-      "Generator Loss: [0.3799423575401306, 0.34472042322158813, 0.017610974609851837]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17993895220570266\n",
-      "Generator Loss: [0.4828231930732727, 0.44001466035842896, 0.021404264494776726]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17043651072890498\n",
-      "Generator Loss: [0.487692266702652, 0.44356775283813477, 0.022062256932258606]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.17111132841091603\n",
-      "Generator Loss: [0.40869224071502686, 0.3719571828842163, 0.018367523327469826]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.20011804762179963\n",
-      "Generator Loss: [2.0043623447418213, 1.9672002792358398, 0.018581055104732513]\n",
-      "2/2 [==============================] - 11s 6s/step\n",
-      "Discriminator Loss: 0.19770437170518562\n",
-      "Generator Loss: [0.5968405604362488, 0.5580974817276001, 0.019371546804904938]\n",
-      "2/2 [==============================] - 12s 6s/step\n",
-      "Discriminator Loss: 0.21312925321399234\n",
-      "Generator Loss: [1.3237717151641846, 1.2873269319534302, 0.018222380429506302]\n",
-      "2/2 [==============================] - 11s 6s/step\n",
-      "Discriminator Loss: 0.21279065348790027\n",
-      "Generator Loss: [3.36556339263916, 3.332369327545166, 0.01659701019525528]\n",
-      "2/2 [==============================] - 13s 6s/step\n",
-      "Discriminator Loss: 0.19690746476408094\n",
-      "Generator Loss: [0.9506070017814636, 0.9186819791793823, 0.0159625057131052]\n",
-      "2/2 [==============================] - 16s 7s/step\n",
-      "Discriminator Loss: 0.19260884018149227\n",
-      "Generator Loss: [1.7744770050048828, 1.746741771697998, 0.013867629691958427]\n",
-      "2/2 [==============================] - 11s 6s/step\n",
-      "Discriminator Loss: 0.1828139489516616\n",
-      "Generator Loss: [0.3788832724094391, 0.3509536385536194, 0.013964821584522724]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.2012077613035217\n",
-      "Generator Loss: [1.946556568145752, 1.9163496494293213, 0.015103446319699287]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.1937195472419262\n",
-      "Generator Loss: [1.3121068477630615, 1.282567024230957, 0.014769895002245903]\n",
-      "2/2 [==============================] - 12s 6s/step\n",
-      "Discriminator Loss: 0.18411833909340203\n",
-      "Generator Loss: [0.8744515180587769, 0.8482756614685059, 0.013087923638522625]\n",
-      "2/2 [==============================] - 12s 6s/step\n",
-      "Discriminator Loss: 0.2075507635017857\n",
-      "Generator Loss: [1.0493234395980835, 1.0180541276931763, 0.015634682029485703]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.20629937457852066\n",
-      "Generator Loss: [0.36480218172073364, 0.33375412225723267, 0.015524027869105339]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.20591497054556385\n",
-      "Generator Loss: [1.136017084121704, 1.096076488494873, 0.019970301538705826]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.2013473198749125\n",
-      "Generator Loss: [0.36319854855537415, 0.3298940658569336, 0.016652243211865425]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1907709245933802\n",
-      "Generator Loss: [0.37893542647361755, 0.3388539254665375, 0.02004075050354004]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17890530452132225\n",
-      "Generator Loss: [0.6149786114692688, 0.5822461247444153, 0.016366230323910713]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17737999447854236\n",
-      "Generator Loss: [0.45276376605033875, 0.42193281650543213, 0.015415480360388756]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17314861994236708\n",
-      "Generator Loss: [0.3772425055503845, 0.34835171699523926, 0.014445386826992035]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1882550767622888\n",
-      "Generator Loss: [0.5169978141784668, 0.47872766852378845, 0.019135072827339172]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.18400583427865058\n",
-      "Generator Loss: [0.5959926247596741, 0.5634856224060059, 0.01625349000096321]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17267936078133062\n",
-      "Generator Loss: [0.4199233949184418, 0.3889540731906891, 0.015484659001231194]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17074050293012988\n",
-      "Generator Loss: [0.37474215030670166, 0.3476373255252838, 0.013552404940128326]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.170090016676113\n",
-      "Generator Loss: [0.36929380893707275, 0.33120882511138916, 0.019042493775486946]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17443176126107574\n",
-      "Generator Loss: [0.3864007592201233, 0.3555888235569, 0.01540597528219223]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16932807319972198\n",
-      "Generator Loss: [0.36731278896331787, 0.3335162401199341, 0.01689828187227249]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17492744699120522\n",
-      "Generator Loss: [0.4135022461414337, 0.3665851950645447, 0.02345851995050907]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17587493814062327\n",
-      "Generator Loss: [0.38513973355293274, 0.33594179153442383, 0.024598972871899605]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.17289568050182424\n",
-      "Generator Loss: [0.38021543622016907, 0.33613622188568115, 0.02203960157930851]\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 12s 5s/step\n",
-      "Discriminator Loss: 0.17608480382477865\n",
-      "Generator Loss: [0.4568089544773102, 0.4205300509929657, 0.018139449879527092]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17570317807258107\n",
-      "Generator Loss: [0.3700219392776489, 0.3337629735469818, 0.018129482865333557]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17384685357683338\n",
-      "Generator Loss: [0.3604457378387451, 0.33480075001716614, 0.012822495773434639]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17448986181989312\n",
-      "Generator Loss: [0.3538672924041748, 0.33077478408813477, 0.011546261608600616]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17496684133948293\n",
-      "Generator Loss: [0.4231283962726593, 0.3902329206466675, 0.01644774153828621]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17173248740436975\n",
-      "Generator Loss: [0.36346757411956787, 0.3280644416809082, 0.017701566219329834]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16963951611251105\n",
-      "Generator Loss: [0.3912806212902069, 0.34945380687713623, 0.02091340720653534]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16938156182004604\n",
-      "Generator Loss: [0.3649074137210846, 0.33208662271499634, 0.01641039550304413]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17060795660654549\n",
-      "Generator Loss: [0.36955857276916504, 0.33921873569488525, 0.015169911086559296]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17048708847141825\n",
-      "Generator Loss: [0.36269038915634155, 0.3301836848258972, 0.016253352165222168]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16852921432291623\n",
-      "Generator Loss: [0.43569329380989075, 0.4117002785205841, 0.011996502056717873]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1677049634163268\n",
-      "Generator Loss: [0.4148292541503906, 0.384807825088501, 0.0150107117369771]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17243369750212878\n",
-      "Generator Loss: [0.37063783407211304, 0.33244195580482483, 0.019097935408353806]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16799236908263993\n",
-      "Generator Loss: [0.616144597530365, 0.588196873664856, 0.013973861932754517]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16677343528863275\n",
-      "Generator Loss: [0.4090512990951538, 0.3813449740409851, 0.013853166252374649]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17378613041364588\n",
-      "Generator Loss: [0.3572929799556732, 0.3301089107990265, 0.013592037372291088]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.171857919296599\n",
-      "Generator Loss: [0.3698786199092865, 0.3413102626800537, 0.014284173026680946]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16928718129202025\n",
-      "Generator Loss: [0.393309623003006, 0.3632696270942688, 0.015019997954368591]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16944985050940886\n",
-      "Generator Loss: [0.3758240342140198, 0.34627437591552734, 0.014774821698665619]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1679202547747991\n",
-      "Generator Loss: [0.45586174726486206, 0.42454367876052856, 0.01565902680158615]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1709235417220043\n",
-      "Generator Loss: [0.370307981967926, 0.3349982500076294, 0.017654860392212868]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1712113279500045\n",
-      "Generator Loss: [0.3917612135410309, 0.35793739557266235, 0.016911910846829414]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17546016390406294\n",
-      "Generator Loss: [0.366447776556015, 0.3328520655632019, 0.016797853633761406]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16931902820942923\n",
-      "Generator Loss: [0.36674201488494873, 0.3333273231983185, 0.016707338392734528]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17162590453517623\n",
-      "Generator Loss: [0.6583423614501953, 0.6261202692985535, 0.016111060976982117]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16615458956221119\n",
-      "Generator Loss: [0.4395010471343994, 0.40292173624038696, 0.018289651721715927]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1707400798331946\n",
-      "Generator Loss: [0.36301538348197937, 0.3357871472835541, 0.013614112511277199]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16916455372120254\n",
-      "Generator Loss: [0.35933223366737366, 0.3295104503631592, 0.014910896308720112]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16903906303923577\n",
-      "Generator Loss: [0.3879064619541168, 0.3603755235671997, 0.013765474781394005]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16872290277387947\n",
-      "Generator Loss: [0.42781153321266174, 0.39410972595214844, 0.016850905492901802]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17039909148297738\n",
-      "Generator Loss: [0.40220969915390015, 0.36968693137168884, 0.01626138761639595]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1712462271680124\n",
-      "Generator Loss: [0.3649415373802185, 0.3293185234069824, 0.01781151443719864]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16951370309107006\n",
-      "Generator Loss: [0.3812105655670166, 0.351068913936615, 0.015070827677845955]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16981495784420986\n",
-      "Generator Loss: [0.4546812176704407, 0.4224095046520233, 0.01613585092127323]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17320181749528274\n",
-      "Generator Loss: [0.4324555993080139, 0.40285536646842957, 0.014800122007727623]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17488716484513134\n",
-      "Generator Loss: [0.5285366177558899, 0.4982609450817108, 0.015137845650315285]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1687689737882465\n",
-      "Generator Loss: [0.7743476629257202, 0.7504233717918396, 0.01196213997900486]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16762044066854287\n",
-      "Generator Loss: [0.39241838455200195, 0.36062461137771606, 0.015896886587142944]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17522127623669803\n",
-      "Generator Loss: [0.47995924949645996, 0.44723474979400635, 0.016362249851226807]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17959866518503986\n",
-      "Generator Loss: [0.41698193550109863, 0.3875569701194763, 0.014712480828166008]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17317762642051093\n",
-      "Generator Loss: [0.41892701387405396, 0.38891711831092834, 0.01500494871288538]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17305987066356465\n",
-      "Generator Loss: [0.47074615955352783, 0.4352695345878601, 0.01773831993341446]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16771239886293188\n",
-      "Generator Loss: [0.3729965388774872, 0.3426097333431244, 0.015193401835858822]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16853168056695722\n",
-      "Generator Loss: [0.3650684356689453, 0.33520442247390747, 0.014932000078260899]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17474146596214268\n",
-      "Generator Loss: [0.3689756691455841, 0.33732056617736816, 0.01582755520939827]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.18525007914286107\n",
-      "Generator Loss: [0.4522797763347626, 0.42608386278152466, 0.013097955845296383]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.18514586449600756\n",
-      "Generator Loss: [0.5230544209480286, 0.4962800443172455, 0.013387191109359264]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17243194801994832\n",
-      "Generator Loss: [0.7475481629371643, 0.7200684547424316, 0.013739850372076035]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16911055691889487\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Generator Loss: [0.3695768117904663, 0.33503156900405884, 0.017272626981139183]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16928945627296343\n",
-      "Generator Loss: [0.4728425145149231, 0.44455772638320923, 0.014142397791147232]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16930212898296304\n",
-      "Generator Loss: [0.3599783480167389, 0.33253535628318787, 0.013721494935452938]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17089922665036283\n",
-      "Generator Loss: [0.39549434185028076, 0.3671080470085144, 0.014193142764270306]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16935424000257626\n",
-      "Generator Loss: [0.37844914197921753, 0.34465736150741577, 0.016895893961191177]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17427526228129864\n",
-      "Generator Loss: [0.3638364374637604, 0.3327910900115967, 0.015522679314017296]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16967234975891188\n",
-      "Generator Loss: [0.37338075041770935, 0.3408825397491455, 0.01624910905957222]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1684767285551061\n",
-      "Generator Loss: [0.4283868670463562, 0.388826847076416, 0.019780009984970093]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17265766870696098\n",
-      "Generator Loss: [0.39780279994010925, 0.3658710718154907, 0.015965864062309265]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.18220909201772884\n",
-      "Generator Loss: [0.36475956439971924, 0.335957795381546, 0.014400891028344631]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.18934649720904417\n",
-      "Generator Loss: [0.41896161437034607, 0.3865914046764374, 0.016185101121664047]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.1729845181107521\n",
-      "Generator Loss: [0.3856550455093384, 0.3563336133956909, 0.014660709537565708]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.16665091700997436\n",
-      "Generator Loss: [0.39443716406822205, 0.36699408292770386, 0.013721542432904243]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.16870928400021512\n",
-      "Generator Loss: [0.385134756565094, 0.3454698324203491, 0.019832462072372437]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17253159446408972\n",
-      "Generator Loss: [0.4223320484161377, 0.3889424800872803, 0.01669478602707386]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1696138363913633\n",
-      "Generator Loss: [0.46218469738960266, 0.430520236492157, 0.015832236036658287]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.17606147041078657\n",
-      "Generator Loss: [0.36321738362312317, 0.3355810344219208, 0.013818180188536644]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16890860415878706\n",
-      "Generator Loss: [0.38953182101249695, 0.36258000135421753, 0.013475916348397732]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.1705631630611606\n",
-      "Generator Loss: [0.3843799829483032, 0.3528372645378113, 0.015771351754665375]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16596704146650154\n",
-      "Generator Loss: [0.36664456129074097, 0.3316943049430847, 0.017475131899118423]\n",
-      "2/2 [==============================] - 10s 4s/step\n",
-      "Discriminator Loss: 0.1734373689396307\n",
-      "Generator Loss: [0.3598364591598511, 0.32885026931762695, 0.015493100509047508]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1679192708688788\n",
-      "Generator Loss: [0.412447452545166, 0.37802278995513916, 0.017212331295013428]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16857122254441492\n",
-      "Generator Loss: [0.3988707363605499, 0.36673304438591003, 0.016068845987319946]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16715452482458204\n",
-      "Generator Loss: [0.41287338733673096, 0.3837941586971283, 0.014539618045091629]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16851527956896462\n",
-      "Generator Loss: [0.45918309688568115, 0.4274182915687561, 0.015882402658462524]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17116099188569933\n",
-      "Generator Loss: [0.4278263449668884, 0.3968328833580017, 0.015496738255023956]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16645437563420273\n",
-      "Generator Loss: [0.3695369362831116, 0.34160447120666504, 0.013966228812932968]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17026940983487293\n",
-      "Generator Loss: [0.3806774914264679, 0.3470960557460785, 0.01679072342813015]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16891886584926397\n",
-      "Generator Loss: [0.437604695558548, 0.4072597622871399, 0.015172465704381466]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.17197988415136933\n",
-      "Generator Loss: [0.8755908012390137, 0.8477689027786255, 0.013910945504903793]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16553981264587492\n",
-      "Generator Loss: [0.385776549577713, 0.35627099871635437, 0.014752769842743874]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.18461283156648278\n",
-      "Generator Loss: [0.6167357563972473, 0.5833362340927124, 0.01669975370168686]\n",
-      "Epoch: 3\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.18786190560786054\n",
-      "Generator Loss: [0.40035948157310486, 0.37711235880851746, 0.011623559519648552]\n",
-      "2/2 [==============================] - 12s 6s/step\n",
-      "Discriminator Loss: 0.18272004136815667\n",
-      "Generator Loss: [1.1274621486663818, 1.0992467403411865, 0.01410769671201706]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.17238521241233684\n",
-      "Generator Loss: [1.0568554401397705, 1.0273264646530151, 0.014764475636184216]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17248140857554972\n",
-      "Generator Loss: [1.2871053218841553, 1.2538206577301025, 0.01664230413734913]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1691504044138128\n",
-      "Generator Loss: [0.4300333559513092, 0.4007378816604614, 0.014647742733359337]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1742007378488779\n",
-      "Generator Loss: [0.39807480573654175, 0.364327609539032, 0.01687360554933548]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17339405103120953\n",
-      "Generator Loss: [1.3116061687469482, 1.2861865758895874, 0.012709783390164375]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16906518844189122\n",
-      "Generator Loss: [0.45205244421958923, 0.4183502495288849, 0.016851093620061874]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1873390493565239\n",
-      "Generator Loss: [0.3874526023864746, 0.34461528062820435, 0.02141866460442543]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.18829722065129317\n",
-      "Generator Loss: [0.37572169303894043, 0.34317493438720703, 0.016273383051156998]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17947205118252896\n",
-      "Generator Loss: [0.3750980496406555, 0.3494420647621155, 0.012827984988689423]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17065593486768194\n",
-      "Generator Loss: [0.3777887225151062, 0.349997878074646, 0.013895421288907528]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16795213551085908\n",
-      "Generator Loss: [0.48876845836639404, 0.4596906006336212, 0.014538928866386414]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17764274420915172\n",
-      "Generator Loss: [0.36791539192199707, 0.3424377143383026, 0.012738832272589207]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17390999267809093\n",
-      "Generator Loss: [0.5064816474914551, 0.471156507730484, 0.017662562429904938]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16915999434422702\n",
-      "Generator Loss: [0.5535266399383545, 0.5199370384216309, 0.016794808208942413]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17003289924468845\n",
-      "Generator Loss: [0.4140845835208893, 0.38588953018188477, 0.01409752108156681]\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17494340310804546\n",
-      "Generator Loss: [0.9887570738792419, 0.9641479253768921, 0.012304561212658882]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16864456358598545\n",
-      "Generator Loss: [1.259760856628418, 1.2372093200683594, 0.011275769211351871]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1721153449616395\n",
-      "Generator Loss: [1.0383540391921997, 1.0166010856628418, 0.010876456275582314]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1798724218388088\n",
-      "Generator Loss: [1.026537537574768, 1.0049618482589722, 0.01078781858086586]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.18098223581910133\n",
-      "Generator Loss: [1.6254639625549316, 1.602280616760254, 0.01159169152379036]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17187901411671191\n",
-      "Generator Loss: [1.9116343259811401, 1.8881337642669678, 0.011750277131795883]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16785936272935942\n",
-      "Generator Loss: [1.119442105293274, 1.0923925638198853, 0.013524760492146015]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17427325650351122\n",
-      "Generator Loss: [1.0369290113449097, 1.0057811737060547, 0.015573892742395401]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1765032477560453\n",
-      "Generator Loss: [1.2746074199676514, 1.2532415390014648, 0.01068294420838356]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1773133582319133\n",
-      "Generator Loss: [0.9008806943893433, 0.8703116178512573, 0.015284530818462372]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17350599882774986\n",
-      "Generator Loss: [0.553489625453949, 0.5188716650009155, 0.017308976501226425]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1742647408391349\n",
-      "Generator Loss: [0.8560408353805542, 0.8184407949447632, 0.0188000351190567]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17124856074224226\n",
-      "Generator Loss: [0.4899701774120331, 0.45712441205978394, 0.01642288826406002]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16939445029129274\n",
-      "Generator Loss: [0.44835203886032104, 0.41534918546676636, 0.016501424834132195]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1710231167380698\n",
-      "Generator Loss: [0.36528483033180237, 0.33169418573379517, 0.01679532788693905]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16747618139197584\n",
-      "Generator Loss: [0.3959253132343292, 0.36428338289260864, 0.01582096889615059]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.17152298231667373\n",
-      "Generator Loss: [0.36590102314949036, 0.3351217210292816, 0.015389649197459221]\n",
-      "2/2 [==============================] - 17s 8s/step\n",
-      "Discriminator Loss: 0.171014094416023\n",
-      "Generator Loss: [0.36260536313056946, 0.333930104970932, 0.014337626285851002]\n",
-      "2/2 [==============================] - 13s 6s/step\n",
-      "Discriminator Loss: 0.1699954115319997\n",
-      "Generator Loss: [0.376579225063324, 0.35220736265182495, 0.01218593493103981]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.16831745652962127\n",
-      "Generator Loss: [0.39868417382240295, 0.3743106424808502, 0.012186768464744091]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.17254666108055972\n",
-      "Generator Loss: [0.35811346769332886, 0.3317609131336212, 0.01317628100514412]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.17576726434344891\n",
-      "Generator Loss: [0.35636886954307556, 0.3310058116912842, 0.012681528925895691]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.16766753987030825\n",
-      "Generator Loss: [0.615674614906311, 0.5932604670524597, 0.011207086965441704]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.16542478395422222\n",
-      "Generator Loss: [0.37208935618400574, 0.34535080194473267, 0.013369271531701088]\n",
-      "2/2 [==============================] - 14s 7s/step\n",
-      "Discriminator Loss: 0.16832562105264515\n",
-      "Generator Loss: [0.3566618263721466, 0.32984572649002075, 0.013408043421804905]\n",
-      "2/2 [==============================] - 14s 7s/step\n",
-      "Discriminator Loss: 0.16889795352471992\n",
-      "Generator Loss: [0.4174891710281372, 0.3834676146507263, 0.017010778188705444]\n",
-      "2/2 [==============================] - 14s 7s/step\n",
-      "Discriminator Loss: 0.1685707927099429\n",
-      "Generator Loss: [0.48286598920822144, 0.4541281461715698, 0.014368928968906403]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.17318387221894227\n",
-      "Generator Loss: [0.3672904968261719, 0.3333343267440796, 0.01697809249162674]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.1764324599353131\n",
-      "Generator Loss: [0.3796052634716034, 0.35096970200538635, 0.014317776076495647]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.16601770190754905\n",
-      "Generator Loss: [0.4525253474712372, 0.4256459176540375, 0.013439709320664406]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.16983347805216908\n",
-      "Generator Loss: [0.36368244886398315, 0.3389965891838074, 0.012342924252152443]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.16745390184223652\n",
-      "Generator Loss: [0.49696776270866394, 0.4649021029472351, 0.016032833606004715]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.1674857825273648\n",
-      "Generator Loss: [0.4553428292274475, 0.4275928735733032, 0.013874977827072144]\n",
-      "2/2 [==============================] - 14s 7s/step\n",
-      "Discriminator Loss: 0.16854040458565578\n",
-      "Generator Loss: [0.47163450717926025, 0.44506001472473145, 0.013287248089909554]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.17177693860139698\n",
-      "Generator Loss: [1.4163120985031128, 1.3923044204711914, 0.012003825977444649]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.16678603186301189\n",
-      "Generator Loss: [0.5191377401351929, 0.48639580607414246, 0.016370976343750954]\n",
-      "2/2 [==============================] - 14s 7s/step\n",
-      "Discriminator Loss: 0.16851148917339742\n",
-      "Generator Loss: [0.7513982057571411, 0.7245684266090393, 0.013414897955954075]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.1671780844917521\n",
-      "Generator Loss: [0.4843246340751648, 0.4546959400177002, 0.014814352616667747]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.16868130664806813\n",
-      "Generator Loss: [0.5577261447906494, 0.5187819004058838, 0.019472114741802216]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.17316252720775083\n",
-      "Generator Loss: [0.4807078540325165, 0.43994367122650146, 0.020382091403007507]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.17090155693585984\n",
-      "Generator Loss: [0.5332371592521667, 0.49608978629112244, 0.01857367902994156]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.17022196488687769\n",
-      "Generator Loss: [0.3977132737636566, 0.3658830523490906, 0.015915105119347572]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.17181193485157564\n",
-      "Generator Loss: [0.6883403062820435, 0.6567626595497131, 0.01578882522881031]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.16988847620086744\n",
-      "Generator Loss: [0.5016655921936035, 0.4781309962272644, 0.011767309159040451]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.1700495561817661\n",
-      "Generator Loss: [0.5215060710906982, 0.5000134706497192, 0.010746300220489502]\n",
-      "2/2 [==============================] - 14s 7s/step\n",
-      "Discriminator Loss: 0.1682173597801011\n",
-      "Generator Loss: [0.6369858980178833, 0.6077045202255249, 0.014640675857663155]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.17556356912245974\n",
-      "Generator Loss: [0.45972925424575806, 0.42892515659332275, 0.015402055345475674]\n",
-      "2/2 [==============================] - 14s 7s/step\n",
-      "Discriminator Loss: 0.17406872677383944\n",
-      "Generator Loss: [0.5918841361999512, 0.5546398162841797, 0.018622150644659996]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.16834616410778835\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Generator Loss: [0.4298703372478485, 0.40060096979141235, 0.014634690247476101]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.1731892686802894\n",
-      "Generator Loss: [0.4100964367389679, 0.38273897767066956, 0.013678734190762043]\n",
-      "2/2 [==============================] - 13s 6s/step\n",
-      "Discriminator Loss: 0.1688562433992047\n",
-      "Generator Loss: [0.3893231451511383, 0.36033597588539124, 0.014493582770228386]\n",
-      "2/2 [==============================] - 14s 6s/step\n",
-      "Discriminator Loss: 0.16702495381468907\n",
-      "Generator Loss: [0.5866925716400146, 0.5639981627464294, 0.01134721003472805]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16708332776033785\n",
-      "Generator Loss: [0.46769869327545166, 0.43964219093322754, 0.01402825303375721]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16866646718699485\n",
-      "Generator Loss: [0.43773967027664185, 0.4030134677886963, 0.017363108694553375]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17234401986934245\n",
-      "Generator Loss: [0.6119978427886963, 0.585772693157196, 0.013112567365169525]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16565213586727623\n",
-      "Generator Loss: [0.6033291220664978, 0.5777413845062256, 0.012793859466910362]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.18124656620784663\n",
-      "Generator Loss: [0.43062248826026917, 0.40589678287506104, 0.012362859211862087]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.17577537054603454\n",
-      "Generator Loss: [0.36576008796691895, 0.3395760953426361, 0.013091997243463993]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17059937847807305\n",
-      "Generator Loss: [0.4481413960456848, 0.42122882604599, 0.013456277549266815]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16753960397909395\n",
-      "Generator Loss: [0.37483203411102295, 0.34761202335357666, 0.013610003516077995]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.1683892916280456\n",
-      "Generator Loss: [0.47175318002700806, 0.4434732496738434, 0.01413995772600174]\n",
-      "2/2 [==============================] - 11s 6s/step\n",
-      "Discriminator Loss: 0.17184889347481658\n",
-      "Generator Loss: [0.3762768507003784, 0.34483587741851807, 0.015720482915639877]\n",
-      "2/2 [==============================] - 11s 6s/step\n",
-      "Discriminator Loss: 0.17118270747596398\n",
-      "Generator Loss: [0.37441641092300415, 0.34456950426101685, 0.014923447743058205]\n",
-      "2/2 [==============================] - 11s 6s/step\n",
-      "Discriminator Loss: 0.17513395394780673\n",
-      "Generator Loss: [0.3644357919692993, 0.3348451852798462, 0.014795304276049137]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.17076136620016769\n",
-      "Generator Loss: [0.36804628372192383, 0.3384634852409363, 0.014791400171816349]\n",
-      "2/2 [==============================] - 16s 7s/step\n",
-      "Discriminator Loss: 0.17195745458593592\n",
-      "Generator Loss: [0.467438668012619, 0.4388991892337799, 0.014269737526774406]\n",
-      "2/2 [==============================] - 13s 5s/step\n",
-      "Discriminator Loss: 0.1648971393060492\n",
-      "Generator Loss: [0.38459673523902893, 0.3524269461631775, 0.016084890812635422]\n",
-      "2/2 [==============================] - 12s 6s/step\n",
-      "Discriminator Loss: 0.1669784159748815\n",
-      "Generator Loss: [0.3533138036727905, 0.32847511768341064, 0.012419350445270538]\n",
-      "2/2 [==============================] - 11s 6s/step\n",
-      "Discriminator Loss: 0.17067766608670354\n",
-      "Generator Loss: [0.3967564105987549, 0.36987653374671936, 0.01343994028866291]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16588246267929208\n",
-      "Generator Loss: [0.39424824714660645, 0.3692878484725952, 0.012480195611715317]\n",
-      "2/2 [==============================] - 22s 9s/step\n",
-      "Discriminator Loss: 0.16645969259843696\n",
-      "Generator Loss: [0.37197351455688477, 0.3419869840145111, 0.014993272721767426]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.17064848163681745\n",
-      "Generator Loss: [0.35879912972450256, 0.3295177221298218, 0.014640703797340393]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.17256452462424932\n",
-      "Generator Loss: [0.5017028450965881, 0.4692632555961609, 0.01621979847550392]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16599899638822535\n",
-      "Generator Loss: [0.3816835582256317, 0.353555291891098, 0.014064129441976547]\n",
-      "2/2 [==============================] - 11s 6s/step\n",
-      "Discriminator Loss: 0.1673942123343295\n",
-      "Generator Loss: [0.364248663187027, 0.3350412845611572, 0.014603693038225174]\n",
-      "2/2 [==============================] - 12s 6s/step\n",
-      "Discriminator Loss: 0.1726404040819034\n",
-      "Generator Loss: [0.3650239408016205, 0.338690847158432, 0.013166546821594238]\n",
-      "2/2 [==============================] - 12s 6s/step\n",
-      "Discriminator Loss: 0.18243438463832717\n",
-      "Generator Loss: [0.41308820247650146, 0.38510584831237793, 0.013991173356771469]\n",
-      "2/2 [==============================] - 12s 6s/step\n",
-      "Discriminator Loss: 0.1647409971756133\n",
-      "Generator Loss: [0.37262943387031555, 0.34921616315841675, 0.011706641875207424]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.16650242352625355\n",
-      "Generator Loss: [0.3600670099258423, 0.3310520052909851, 0.014507504180073738]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.16642505804338725\n",
-      "Generator Loss: [0.4155694544315338, 0.38605988025665283, 0.014754780568182468]\n",
-      "2/2 [==============================] - 15s 5s/step\n",
-      "Discriminator Loss: 0.16991884306480642\n",
-      "Generator Loss: [0.36539018154144287, 0.33845558762550354, 0.013467303477227688]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16869395493995398\n",
-      "Generator Loss: [0.36097878217697144, 0.3338797390460968, 0.013549525290727615]\n",
-      "2/2 [==============================] - 15s 8s/step\n",
-      "Discriminator Loss: 0.16754743065393995\n",
-      "Generator Loss: [0.3700011670589447, 0.3382492661476135, 0.015875954180955887]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.16786337227676995\n",
-      "Generator Loss: [0.3803131878376007, 0.3524041175842285, 0.013954534195363522]\n",
-      "2/2 [==============================] - 10s 6s/step\n",
-      "Discriminator Loss: 0.16587493389670271\n",
-      "Generator Loss: [0.4442394971847534, 0.4165399968624115, 0.013849755749106407]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.1738493260127143\n",
-      "Generator Loss: [0.406830757856369, 0.37787529826164246, 0.014477735385298729]\n",
-      "2/2 [==============================] - 13s 6s/step\n",
-      "Discriminator Loss: 0.1687406963246758\n",
-      "Generator Loss: [0.3662053942680359, 0.342445433139801, 0.01187998615205288]\n",
-      "2/2 [==============================] - 15s 7s/step\n",
-      "Discriminator Loss: 0.17411155870649964\n",
-      "Generator Loss: [0.39604663848876953, 0.3712683916091919, 0.012389129027724266]\n",
-      "2/2 [==============================] - 15s 8s/step\n",
-      "Discriminator Loss: 0.16640230431767122\n",
-      "Generator Loss: [0.5716103315353394, 0.546708345413208, 0.012450982816517353]\n",
-      "2/2 [==============================] - 19s 10s/step\n",
-      "Discriminator Loss: 0.16533053667808417\n",
-      "Generator Loss: [0.5150612592697144, 0.4845603108406067, 0.01525047980248928]\n",
-      "2/2 [==============================] - 15s 7s/step\n",
-      "Discriminator Loss: 0.1652998610807117\n",
-      "Generator Loss: [0.4924704134464264, 0.46607333421707153, 0.013198541477322578]\n",
-      "2/2 [==============================] - 13s 8s/step\n",
-      "Discriminator Loss: 0.168816727511512\n",
-      "Generator Loss: [0.3559180796146393, 0.3302146792411804, 0.012851701118052006]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.16583217139123008\n",
-      "Generator Loss: [0.3596840500831604, 0.33405232429504395, 0.012815858237445354]\n",
-      "2/2 [==============================] - 12s 6s/step\n",
-      "Discriminator Loss: 0.1671649508862174\n",
-      "Generator Loss: [0.36001959443092346, 0.33011263608932495, 0.014953484758734703]\n",
-      "2/2 [==============================] - 14s 6s/step\n",
-      "Discriminator Loss: 0.16897286873063422\n",
-      "Generator Loss: [0.35874882340431213, 0.33057236671447754, 0.014088224619626999]\n",
-      "2/2 [==============================] - 11s 6s/step\n",
-      "Discriminator Loss: 0.16953081748215482\n",
-      "Generator Loss: [0.35611122846603394, 0.32748958468437195, 0.014310818165540695]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.16686296668740397\n",
-      "Generator Loss: [0.41733771562576294, 0.3835369348526001, 0.016900386661291122]\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 11s 6s/step\n",
-      "Discriminator Loss: 0.16641188046196476\n",
-      "Generator Loss: [0.3862459361553192, 0.3580699563026428, 0.014087987132370472]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.1824552961770678\n",
-      "Generator Loss: [0.4775696098804474, 0.45149070024490356, 0.013039452955126762]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.2016742865089327\n",
-      "Generator Loss: [0.37903594970703125, 0.3498556911945343, 0.014590123668313026]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.17399453035523038\n",
-      "Generator Loss: [0.414009153842926, 0.3878224194049835, 0.013093373738229275]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.1659562169425044\n",
-      "Generator Loss: [0.4454188048839569, 0.42162811756134033, 0.011895349249243736]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16620586908902624\n",
-      "Generator Loss: [0.39494770765304565, 0.36139971017837524, 0.016774002462625504]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16843102540588006\n",
-      "Generator Loss: [0.36677688360214233, 0.33793172240257263, 0.01442258246243]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.1664798087294912\n",
-      "Generator Loss: [0.4278429448604584, 0.4009026288986206, 0.013470158912241459]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.1725843984168023\n",
-      "Generator Loss: [0.35711050033569336, 0.3325357735157013, 0.012287361547350883]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16698741454456467\n",
-      "Generator Loss: [0.41720420122146606, 0.3933308720588684, 0.011936662718653679]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.170916668845166\n",
-      "Generator Loss: [0.4422612488269806, 0.4146150052547455, 0.013823124580085278]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.1652540029026568\n",
-      "Generator Loss: [0.41581404209136963, 0.38635069131851196, 0.014731679111719131]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.17078217075322755\n",
-      "Generator Loss: [0.3594152629375458, 0.33213597536087036, 0.01363963820040226]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16562949633225799\n",
-      "Generator Loss: [0.3609738349914551, 0.33017393946647644, 0.015399951487779617]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.1659727299338556\n",
-      "Generator Loss: [0.41476762294769287, 0.38532835245132446, 0.014719638973474503]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16557323735105456\n",
-      "Generator Loss: [0.3925667405128479, 0.3651074171066284, 0.013729657977819443]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.1667963459112798\n",
-      "Generator Loss: [0.37758809328079224, 0.348335325717926, 0.014626391232013702]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.1682746937731281\n",
-      "Generator Loss: [0.4498112201690674, 0.42085251212120056, 0.014479357749223709]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.16555526966112666\n",
-      "Generator Loss: [0.5974611639976501, 0.5716837644577026, 0.01288868673145771]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16686578941880725\n",
-      "Generator Loss: [0.41954803466796875, 0.3887883722782135, 0.015379825606942177]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16715402479167096\n",
-      "Generator Loss: [0.3729404807090759, 0.344959557056427, 0.013990465551614761]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.1672285598469898\n",
-      "Generator Loss: [1.1116423606872559, 1.085374355316162, 0.013133999891579151]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.16654381752596237\n",
-      "Generator Loss: [0.6341134309768677, 0.60593581199646, 0.014088815078139305]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.17511807626578957\n",
-      "Generator Loss: [0.3665095865726471, 0.3354770243167877, 0.015516286715865135]\n",
-      "Epoch: 4\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.18495365080889314\n",
-      "Generator Loss: [0.5454630851745605, 0.5219777822494507, 0.011742657981812954]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.16948716733895708\n",
-      "Generator Loss: [0.4033985137939453, 0.37507760524749756, 0.014160452410578728]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.16674012338626198\n",
-      "Generator Loss: [0.38142403960227966, 0.35396572947502136, 0.013729157857596874]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16651613284193445\n",
-      "Generator Loss: [0.4186111390590668, 0.38679608702659607, 0.01590752601623535]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.16563226768630557\n",
-      "Generator Loss: [0.3907448351383209, 0.36185479164123535, 0.014445027336478233]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16921527151134796\n",
-      "Generator Loss: [0.36227911710739136, 0.33043792843818665, 0.015920599922537804]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.17759753542486578\n",
-      "Generator Loss: [0.6777408123016357, 0.6519893407821655, 0.012875721789896488]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16594576268835226\n",
-      "Generator Loss: [1.3278249502182007, 1.2949550151824951, 0.016434945166110992]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.17587268678471446\n",
-      "Generator Loss: [0.4607061445713043, 0.42138123512268066, 0.01966244913637638]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.17521330273666536\n",
-      "Generator Loss: [0.7175789475440979, 0.6873071789741516, 0.01513589732348919]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16780833921802696\n",
-      "Generator Loss: [0.3743285536766052, 0.3474777936935425, 0.013425384648144245]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16736666491487995\n",
-      "Generator Loss: [0.41476038098335266, 0.3852865993976593, 0.014736896380782127]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.1658373635436874\n",
-      "Generator Loss: [0.4739883840084076, 0.4430987536907196, 0.015444815158843994]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.1702727860174491\n",
-      "Generator Loss: [0.35724127292633057, 0.33084940910339355, 0.013195924460887909]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16843747186794644\n",
-      "Generator Loss: [0.38547298312187195, 0.3496078550815582, 0.017932569608092308]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16539388859382598\n",
-      "Generator Loss: [0.40697598457336426, 0.3721731901168823, 0.017401400953531265]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16484116462197562\n",
-      "Generator Loss: [0.3596210777759552, 0.3300659656524658, 0.014777558855712414]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16480439850420225\n",
-      "Generator Loss: [0.36247745156288147, 0.3361097574234009, 0.01318384800106287]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16510303209724952\n",
-      "Generator Loss: [0.36044371128082275, 0.3358667194843292, 0.012288501486182213]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16601856920169666\n",
-      "Generator Loss: [0.3532679080963135, 0.3297356367111206, 0.011766135692596436]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16611196594749345\n",
-      "Generator Loss: [0.37229466438293457, 0.3487715721130371, 0.011761538684368134]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.17371926737541799\n",
-      "Generator Loss: [0.41519978642463684, 0.39063408970832825, 0.012282848358154297]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16547881541737297\n",
-      "Generator Loss: [0.4949989318847656, 0.4693142771720886, 0.012842325493693352]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.1658819711446995\n",
-      "Generator Loss: [0.37695813179016113, 0.34936484694480896, 0.013796648010611534]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.17241608526092023\n",
-      "Generator Loss: [0.39414674043655396, 0.36217600107192993, 0.015985362231731415]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.17376710086136882\n",
-      "Generator Loss: [0.3860653042793274, 0.3636394739151001, 0.011212920770049095]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16595079828766757\n",
-      "Generator Loss: [0.3630428910255432, 0.33259937167167664, 0.015221762470901012]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16608282589004375\n",
-      "Generator Loss: [0.3613685667514801, 0.3288751244544983, 0.016246724873781204]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16497243375852122\n",
-      "Generator Loss: [0.4118324816226959, 0.375696063041687, 0.018068211153149605]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16620764674917154\n",
-      "Generator Loss: [0.4026806354522705, 0.3709120750427246, 0.01588427647948265]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.1646093401068356\n",
-      "Generator Loss: [0.38710829615592957, 0.35491523146629333, 0.016096536070108414]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.1645543326594634\n",
-      "2/2 [==============================] - 15s 7s/step\n",
-      "Discriminator Loss: 0.1652485320373671\n",
-      "Generator Loss: [0.3597436547279358, 0.33007991313934326, 0.014831874519586563]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16592970842430077\n",
-      "Generator Loss: [0.3761609196662903, 0.34610089659690857, 0.01503001805394888]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16894916370802093\n",
-      "Generator Loss: [0.39055702090263367, 0.3627520799636841, 0.01390247605741024]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16552049390156753\n",
-      "Generator Loss: [0.40415650606155396, 0.38054755330085754, 0.011804482899606228]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1645762436928635\n",
-      "Generator Loss: [0.44042688608169556, 0.41656309366226196, 0.01193189062178135]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17051066536077997\n",
-      "Generator Loss: [0.3594999313354492, 0.3342670202255249, 0.012616457417607307]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1770795360316697\n",
-      "Generator Loss: [0.3531494736671448, 0.32859498262405396, 0.012277249246835709]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16656027637145598\n",
-      "Generator Loss: [0.5139154195785522, 0.49198099970817566, 0.010967211797833443]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16501183296350064\n",
-      "Generator Loss: [0.39065611362457275, 0.3648329973220825, 0.012911558151245117]\n",
-      "2/2 [==============================] - 12s 6s/step\n",
-      "Discriminator Loss: 0.1649114851388731\n",
-      "Generator Loss: [0.3824257552623749, 0.3562537133693695, 0.013086024671792984]\n",
-      "2/2 [==============================] - 11s 6s/step\n",
-      "Discriminator Loss: 0.16725199019128922\n",
-      "Generator Loss: [0.3645407557487488, 0.33107879757881165, 0.01673097163438797]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.16602606590822688\n",
-      "Generator Loss: [0.40152984857559204, 0.3728751242160797, 0.014327360317111015]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.16824608722527046\n",
-      "Generator Loss: [0.3658478558063507, 0.33270931243896484, 0.016569267958402634]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.17013296311779413\n",
-      "Generator Loss: [0.361298531293869, 0.3325863480567932, 0.014356089755892754]\n",
-      "2/2 [==============================] - 11s 6s/step\n",
-      "Discriminator Loss: 0.16437834838870913\n",
-      "Generator Loss: [0.3847960829734802, 0.3574129343032837, 0.013691581785678864]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.1641515490337042\n",
-      "Generator Loss: [0.35567909479141235, 0.33017420768737793, 0.012752437964081764]\n",
-      "2/2 [==============================] - 13s 5s/step\n",
-      "Discriminator Loss: 0.16499838728486793\n",
-      "Generator Loss: [0.42556899785995483, 0.39370977878570557, 0.015929603949189186]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.16569404141046107\n",
-      "Generator Loss: [0.3779250681400299, 0.35018372535705566, 0.013870670460164547]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16531831777683692\n",
-      "Generator Loss: [0.4082338809967041, 0.3818623125553131, 0.013185782358050346]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16464122801698977\n",
-      "Generator Loss: [0.3531458377838135, 0.32821589708328247, 0.012464972212910652]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16701537222252227\n",
-      "Generator Loss: [0.372311532497406, 0.33943498134613037, 0.016438277438282967]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16530036216136068\n",
-      "Generator Loss: [0.3867120146751404, 0.35972654819488525, 0.013492734171450138]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16380948287405772\n",
-      "Generator Loss: [0.4402015507221222, 0.41038790345191956, 0.014906824566423893]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.1647536688665241\n",
-      "Generator Loss: [0.442035436630249, 0.4053571820259094, 0.0183391273021698]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16473805832356447\n",
-      "Generator Loss: [0.3665708005428314, 0.32896584272384644, 0.018802473321557045]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16529932451521745\n",
-      "Generator Loss: [0.41724133491516113, 0.38227516412734985, 0.01748308166861534]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1661824104085099\n",
-      "Generator Loss: [0.3798931837081909, 0.34941384196281433, 0.01523967832326889]\n",
-      "2/2 [==============================] - 13s 7s/step\n",
-      "Discriminator Loss: 0.16685689303267282\n",
-      "Generator Loss: [0.3663750886917114, 0.3363268971443176, 0.015024088323116302]\n",
-      "2/2 [==============================] - 15s 7s/step\n",
-      "Discriminator Loss: 0.16618017565633636\n",
-      "Generator Loss: [0.4120415151119232, 0.3887244462966919, 0.011658533476293087]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16539428813848644\n",
-      "Generator Loss: [0.3664032816886902, 0.34440189599990845, 0.011000688187777996]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16523216321365908\n",
-      "Generator Loss: [0.35872432589530945, 0.3303970396518707, 0.014163648709654808]\n",
-      "2/2 [==============================] - 18s 10s/step\n",
-      "Discriminator Loss: 0.16547990402432333\n",
-      "Generator Loss: [0.37896764278411865, 0.35017773509025574, 0.01439494825899601]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16542009510158096\n",
-      "Generator Loss: [0.403539776802063, 0.3689565062522888, 0.017291642725467682]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.16409106991704903\n",
-      "Generator Loss: [0.38525158166885376, 0.3578510284423828, 0.013700276613235474]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.1652927259929129\n",
-      "Generator Loss: [0.3533809185028076, 0.3279705047607422, 0.012705208733677864]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.1646286701397912\n",
-      "Generator Loss: [0.4363657832145691, 0.40975141525268555, 0.013307178393006325]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.1644370008725673\n",
-      "Generator Loss: [0.35855749249458313, 0.3367784917354584, 0.010889505967497826]\n",
-      "2/2 [==============================] - 12s 6s/step\n",
-      "Discriminator Loss: 0.16482949937926605\n",
-      "Generator Loss: [0.35752642154693604, 0.33097028732299805, 0.013278059661388397]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16472860183785087\n",
-      "Generator Loss: [0.378016859292984, 0.3456293046474457, 0.016193775460124016]\n",
-      "2/2 [==============================] - 11s 6s/step\n",
-      "Discriminator Loss: 0.16406741703031003\n",
-      "Generator Loss: [0.38693177700042725, 0.36198222637176514, 0.012474767863750458]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.16481908179775928\n",
-      "Generator Loss: [0.3819853961467743, 0.3578326106071472, 0.012076396495103836]\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16734366286254954\n",
-      "Generator Loss: [0.3543660640716553, 0.3310539126396179, 0.011656073853373528]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16583442562114215\n",
-      "Generator Loss: [0.35238903760910034, 0.32776278257369995, 0.012313131242990494]\n",
-      "2/2 [==============================] - 12s 6s/step\n",
-      "Discriminator Loss: 0.16494313743896782\n",
-      "Generator Loss: [0.3850288987159729, 0.36006975173950195, 0.012479577213525772]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1651929185318295\n",
-      "Generator Loss: [0.35898736119270325, 0.33371156454086304, 0.012637892737984657]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1660370599493035\n",
-      "Generator Loss: [0.39746376872062683, 0.3715343475341797, 0.01296471618115902]\n",
-      "2/2 [==============================] - 12s 7s/step\n",
-      "Discriminator Loss: 0.1665205224271631\n",
-      "Generator Loss: [0.3597390949726105, 0.33069074153900146, 0.014524180442094803]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16797946870792657\n",
-      "Generator Loss: [0.36554428935050964, 0.3385855257511139, 0.0134793845936656]\n",
-      "2/2 [==============================] - 11s 6s/step\n",
-      "Discriminator Loss: 0.17149963347765151\n",
-      "Generator Loss: [0.36248356103897095, 0.33561262488365173, 0.013435464352369308]\n",
-      "2/2 [==============================] - 11s 5s/step\n",
-      "Discriminator Loss: 0.1694750538590597\n",
-      "Generator Loss: [0.35412007570266724, 0.3273569345474243, 0.013381567783653736]\n",
-      "2/2 [==============================] - 11s 6s/step\n",
-      "Discriminator Loss: 0.1672392253385624\n",
-      "Generator Loss: [0.4858878254890442, 0.45988431572914124, 0.01300174742937088]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16482586595520843\n",
-      "Generator Loss: [0.36562126874923706, 0.3365517854690552, 0.014534739777445793]\n",
-      "2/2 [==============================] - 11s 6s/step\n",
-      "Discriminator Loss: 0.16389035219617654\n",
-      "Generator Loss: [0.35599929094314575, 0.33324646949768066, 0.011376403272151947]\n",
-      "2/2 [==============================] - 14s 8s/step\n",
-      "Discriminator Loss: 0.164706398601993\n",
-      "Generator Loss: [0.4740229845046997, 0.44929009675979614, 0.012366442009806633]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16573649778729305\n",
-      "Generator Loss: [0.36327365040779114, 0.3401365876197815, 0.011568528600037098]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16539491014555097\n",
-      "Generator Loss: [0.46163466572761536, 0.43417081236839294, 0.013731930404901505]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16918120803893544\n",
-      "Generator Loss: [0.3662724196910858, 0.33950120210647583, 0.013385608792304993]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.17096562514052493\n",
-      "Generator Loss: [0.388098806142807, 0.3573337197303772, 0.015382548794150352]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16457880682719406\n",
-      "Generator Loss: [0.36573848128318787, 0.3394012749195099, 0.01316859945654869]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16529286278455402\n",
-      "Generator Loss: [0.35687294602394104, 0.33119434118270874, 0.01283930242061615]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16941530914482428\n",
-      "Generator Loss: [0.3984394967556, 0.3752051591873169, 0.01161716878414154]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17431159590705647\n",
-      "Generator Loss: [0.3621082305908203, 0.33655187487602234, 0.012778185307979584]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16609659857931547\n",
-      "Generator Loss: [0.3523296117782593, 0.33036303520202637, 0.010983280837535858]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16730464535066858\n",
-      "Generator Loss: [0.5041572451591492, 0.4776385426521301, 0.013259361498057842]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16571567277424037\n",
-      "Generator Loss: [0.36046677827835083, 0.3333168625831604, 0.013574963435530663]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16675247470266186\n",
-      "Generator Loss: [0.556318998336792, 0.5312560796737671, 0.012531468644738197]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16801766501157545\n",
-      "Generator Loss: [0.3871045410633087, 0.36187052726745605, 0.012617007829248905]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16525917865510564\n",
-      "Generator Loss: [0.35745322704315186, 0.32831668853759766, 0.014568273909389973]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16540788042766508\n",
-      "Generator Loss: [0.39707842469215393, 0.3712608218193054, 0.012908805161714554]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16523899751337012\n",
-      "Generator Loss: [0.35652396082878113, 0.33019697666168213, 0.0131634920835495]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16628245887113735\n",
-      "Generator Loss: [0.36151519417762756, 0.334019273519516, 0.013747953809797764]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1666059730341658\n",
-      "Generator Loss: [0.3747931122779846, 0.35237202048301697, 0.011210551485419273]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16907517229992663\n",
-      "Generator Loss: [0.3554101288318634, 0.33171749114990234, 0.011846315115690231]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16466736714937724\n",
-      "Generator Loss: [0.3628542423248291, 0.3392466604709625, 0.011803792789578438]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1641086536401417\n",
-      "Generator Loss: [0.4342121481895447, 0.4058682918548584, 0.01417192816734314]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16507738933432847\n",
-      "Generator Loss: [0.4810871183872223, 0.4562544822692871, 0.01241631805896759]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1688435175747145\n",
-      "Generator Loss: [0.3535479009151459, 0.3295983076095581, 0.011974794790148735]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16552174182288582\n",
-      "Generator Loss: [0.39130279421806335, 0.3678939938545227, 0.011704404838383198]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16529024037299678\n",
-      "Generator Loss: [0.3678763508796692, 0.34083831310272217, 0.01351902261376381]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16799726868703146\n",
-      "Generator Loss: [0.3585699796676636, 0.3322397470474243, 0.013165116310119629]\n",
-      "2/2 [==============================] - 10s 5s/step\n",
-      "Discriminator Loss: 0.16937637147202622\n",
-      "Generator Loss: [0.3542743921279907, 0.32826030254364014, 0.013007043860852718]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16562818441161653\n",
-      "Generator Loss: [0.3633078336715698, 0.33335447311401367, 0.014976685866713524]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16469637505724677\n",
-      "Generator Loss: [0.35419923067092896, 0.32877302169799805, 0.012713097035884857]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16813287176410086\n",
-      "Generator Loss: [0.3562330901622772, 0.33213934302330017, 0.012046867981553078]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1749414632904518\n",
-      "Generator Loss: [0.36751484870910645, 0.3409704566001892, 0.013272199779748917]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1660745574044995\n",
-      "Generator Loss: [0.3690163493156433, 0.34513187408447266, 0.01194223202764988]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16480249437699968\n",
-      "Generator Loss: [0.4174412190914154, 0.3959193825721741, 0.010760918259620667]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.16456690061932022\n",
-      "Generator Loss: [0.3866296410560608, 0.3569265902042389, 0.014851527288556099]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16484076900815126\n",
-      "Generator Loss: [0.35310351848602295, 0.327371746301651, 0.012865884229540825]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16523501432675403\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Generator Loss: [0.37614843249320984, 0.3522852659225464, 0.011931585147976875]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16777721928519895\n",
-      "Generator Loss: [0.3536750376224518, 0.3312028646469116, 0.011236089281737804]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1650715385985677\n",
-      "Generator Loss: [0.3644690215587616, 0.3426232635974884, 0.010922877117991447]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1671553061269151\n",
-      "Generator Loss: [0.3685956299304962, 0.3438794016838074, 0.012358112260699272]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16404423794301692\n",
-      "Generator Loss: [0.3853423595428467, 0.35992664098739624, 0.01270785927772522]\n",
-      "2/2 [==============================] - 9s 5s/step\n",
-      "Discriminator Loss: 0.1657894749660045\n",
-      "Generator Loss: [0.3563934862613678, 0.3318830132484436, 0.01225524116307497]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1648138800082961\n",
-      "Generator Loss: [0.36057156324386597, 0.3329862356185913, 0.013792656362056732]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16505658159439918\n",
-      "Generator Loss: [0.35950830578804016, 0.33221685886383057, 0.01364572811871767]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16410290688145324\n",
-      "Generator Loss: [0.3927435278892517, 0.36752814054489136, 0.012607688084244728]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16648591050761752\n",
-      "Generator Loss: [0.35606956481933594, 0.3291208744049072, 0.013474341481924057]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16861653997329995\n",
-      "Generator Loss: [0.47455430030822754, 0.4481063485145569, 0.013223975896835327]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16382326596794883\n",
-      "Generator Loss: [0.41020286083221436, 0.3865926265716553, 0.011805113404989243]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16600209915122832\n",
-      "Generator Loss: [0.3584953546524048, 0.33099234104156494, 0.013751505874097347]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16701577373896725\n",
-      "Generator Loss: [0.7769206762313843, 0.7521202564239502, 0.012400197796523571]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16422421470633708\n",
-      "Generator Loss: [0.39647865295410156, 0.3724963665008545, 0.011991141363978386]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.1645187284811982\n",
-      "Generator Loss: [0.4146750867366791, 0.38943201303482056, 0.012621533125638962]\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16806387425458524\n",
-      "Generator Loss: [0.37194937467575073, 0.34467586874961853, 0.013636760413646698]\n",
-      "Epoch: 5\n",
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.17272324481018586\n",
-      "Generator Loss: [0.524732232093811, 0.503046989440918, 0.010842624120414257]\n"
-     ]
-    },
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n",
-      "Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "2/2 [==============================] - 9s 4s/step\n",
-      "Discriminator Loss: 0.16846342326607555\n"
-     ]
-    }
-   ],
-   "source": [
-    "stage1 = StackGanStage1()\n",
-    "stage1.train_stage1()"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "383b4942",
-   "metadata": {},
-   "source": [
-    "## Check test folder for gernerated images from Stage1 Generator\n",
-    "### Let's Implement Stage 2 Generator"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 21,
-   "id": "f51c16c3",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "############################################################\n",
-    "# Stage 2 Generator Network\n",
-    "############################################################\n",
-    "\n",
-    "def concat_along_dims(inputs):\n",
-    "    \"\"\"Joins the conditioned text with the encoded image along the dimensions.\n",
-    "\n",
-    "    Args:\n",
-    "        inputs: consisting of conditioned text and encoded images as [c,x].\n",
-    "\n",
-    "    Returns:\n",
-    "        Joint block along the dimensions.\n",
-    "    \"\"\"\n",
-    "    c = inputs[0]\n",
-    "    x = inputs[1]\n",
-    "\n",
-    "    c = K.expand_dims(c, axis=1)\n",
-    "    c = K.expand_dims(c, axis=1)\n",
-    "    c = K.tile(c, [1, 16, 16, 1])\n",
-    "    return K.concatenate([c, x], axis = 3)\n",
-    "\n",
-    "def residual_block(input):\n",
-    "    \"\"\"Residual block with plain identity connections.\n",
-    "\n",
-    "    Args:\n",
-    "        inputs: input layer or an encoded layer\n",
-    "\n",
-    "    Returns:\n",
-    "            Layer with computed identity mapping.\n",
-    "    \"\"\"\n",
-    "    x = Conv2D(512, kernel_size=(3,3), padding='same', use_bias=False,\n",
-    "                kernel_initializer='he_uniform')(input)\n",
-    "    x = BatchNormalization(gamma_initializer='ones', beta_initializer='zeros')(x)\n",
-    "    x = ReLU()(x)\n",
-    "    x = Conv2D(512, kernel_size=(3,3), padding='same', use_bias=False,\n",
-    "            kernel_initializer='he_uniform')(x)\n",
-    "    x = BatchNormalization(gamma_initializer='ones', beta_initializer='zeros')(x)\n",
-    "    \n",
-    "    x = add([x, input])\n",
-    "    x = ReLU()(x)\n",
-    "\n",
-    "    return x\n",
-    "def build_stage2_generator():\n",
-    "    \"\"\"Build the Stage 2 Generator Network using the conditioning text and images from stage 1.\n",
-    "\n",
-    "    Returns:\n",
-    "        Stage 2 Generator Model for StackGAN.\n",
-    "    \"\"\"\n",
-    "    input_layer1 = Input(shape=(1024,))\n",
-    "    input_images = Input(shape=(64, 64, 3))\n",
-    "\n",
-    "    # Conditioning Augmentation\n",
-    "    ca = Dense(256)(input_layer1)\n",
-    "    mls = LeakyReLU(alpha=0.2)(ca)\n",
-    "    c = Lambda(conditioning_augmentation)(mls)\n",
-    "\n",
-    "    # Downsampling block\n",
-    "    x = ZeroPadding2D(padding=(1,1))(input_images)\n",
-    "    x = Conv2D(128, kernel_size=(3,3), strides=1, use_bias=False,\n",
-    "                kernel_initializer='he_uniform')(x)\n",
-    "    x = ReLU()(x)\n",
-    "\n",
-    "    x = ZeroPadding2D(padding=(1,1))(x)\n",
-    "    x = Conv2D(256, kernel_size=(4,4), strides=2, use_bias=False,\n",
-    "                kernel_initializer='he_uniform')(x)\n",
-    "    x = BatchNormalization(gamma_initializer='ones', beta_initializer='zeros')(x)\n",
-    "    x = ReLU()(x)\n",
-    "\n",
-    "    x = ZeroPadding2D(padding=(1,1))(x)\n",
-    "    x = Conv2D(512, kernel_size=(4,4), strides=2, use_bias=False,\n",
-    "                kernel_initializer='he_uniform')(x)\n",
-    "    x = BatchNormalization(gamma_initializer='ones', beta_initializer='zeros')(x)\n",
-    "    x = ReLU()(x)\n",
-    "\n",
-    "    # Concatenate text conditioning block with the encoded image\n",
-    "    concat = concat_along_dims([c, x])\n",
-    "    # Residual Blocks\n",
-    "    x = ZeroPadding2D(padding=(1,1))(concat)\n",
-    "    x = Conv2D(512, kernel_size=(3,3), use_bias=False, kernel_initializer='he_uniform')(x)\n",
-    "    x = BatchNormalization(gamma_initializer='ones', beta_initializer='zeros')(x)\n",
-    "    x = ReLU()(x)\n",
-    "\n",
-    "    x = residual_block(x)\n",
-    "    x = residual_block(x)\n",
-    "    x = residual_block(x)\n",
-    "    x = residual_block(x)\n",
-    "\n",
-    "    # Upsampling Blocks\n",
-    "    x = UpSamplingBlock(x, 512)\n",
-    "    x = UpSamplingBlock(x, 256)\n",
-    "    x = UpSamplingBlock(x, 128)\n",
-    "    x = UpSamplingBlock(x, 64)\n",
-    "\n",
-    "    x = Conv2D(3, kernel_size=(3,3), padding='same', use_bias=False, kernel_initializer='he_uniform')(x)\n",
-    "    x = Activation('tanh')(x)\n",
-    "    \n",
-    "    stage2_gen = Model(inputs=[input_layer1, input_images], outputs=[x, mls])\n",
-    "    return stage2_gen"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 22,
-   "id": "ca66f7ae",
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model: \"model_3\"\n",
-      "__________________________________________________________________________________________________\n",
-      " Layer (type)                   Output Shape         Param #     Connected to                     \n",
-      "==================================================================================================\n",
-      " input_12 (InputLayer)          [(None, 64, 64, 3)]  0           []                               \n",
-      "                                                                                                  \n",
-      " zero_padding2d (ZeroPadding2D)  (None, 66, 66, 3)   0           ['input_12[0][0]']               \n",
-      "                                                                                                  \n",
-      " conv2d_11 (Conv2D)             (None, 64, 64, 128)  3456        ['zero_padding2d[0][0]']         \n",
-      "                                                                                                  \n",
-      " re_lu_5 (ReLU)                 (None, 64, 64, 128)  0           ['conv2d_11[0][0]']              \n",
-      "                                                                                                  \n",
-      " zero_padding2d_1 (ZeroPadding2  (None, 66, 66, 128)  0          ['re_lu_5[0][0]']                \n",
-      " D)                                                                                               \n",
-      "                                                                                                  \n",
-      " input_11 (InputLayer)          [(None, 1024)]       0           []                               \n",
-      "                                                                                                  \n",
-      " conv2d_12 (Conv2D)             (None, 32, 32, 256)  524288      ['zero_padding2d_1[0][0]']       \n",
-      "                                                                                                  \n",
-      " dense_4 (Dense)                (None, 256)          262400      ['input_11[0][0]']               \n",
-      "                                                                                                  \n",
-      " batch_normalization_8 (BatchNo  (None, 32, 32, 256)  1024       ['conv2d_12[0][0]']              \n",
-      " rmalization)                                                                                     \n",
-      "                                                                                                  \n",
-      " leaky_re_lu_8 (LeakyReLU)      (None, 256)          0           ['dense_4[0][0]']                \n",
-      "                                                                                                  \n",
-      " re_lu_6 (ReLU)                 (None, 32, 32, 256)  0           ['batch_normalization_8[0][0]']  \n",
-      "                                                                                                  \n",
-      " lambda_2 (Lambda)              (None, 128)          0           ['leaky_re_lu_8[0][0]']          \n",
-      "                                                                                                  \n",
-      " zero_padding2d_2 (ZeroPadding2  (None, 34, 34, 256)  0          ['re_lu_6[0][0]']                \n",
-      " D)                                                                                               \n",
-      "                                                                                                  \n",
-      " tf.expand_dims (TFOpLambda)    (None, 1, 128)       0           ['lambda_2[0][0]']               \n",
-      "                                                                                                  \n",
-      " conv2d_13 (Conv2D)             (None, 16, 16, 512)  2097152     ['zero_padding2d_2[0][0]']       \n",
-      "                                                                                                  \n",
-      " tf.expand_dims_1 (TFOpLambda)  (None, 1, 1, 128)    0           ['tf.expand_dims[0][0]']         \n",
-      "                                                                                                  \n",
-      " batch_normalization_9 (BatchNo  (None, 16, 16, 512)  2048       ['conv2d_13[0][0]']              \n",
-      " rmalization)                                                                                     \n",
-      "                                                                                                  \n",
-      " tf.tile (TFOpLambda)           (None, 16, 16, 128)  0           ['tf.expand_dims_1[0][0]']       \n",
-      "                                                                                                  \n",
-      " re_lu_7 (ReLU)                 (None, 16, 16, 512)  0           ['batch_normalization_9[0][0]']  \n",
-      "                                                                                                  \n",
-      " tf.concat (TFOpLambda)         (None, 16, 16, 640)  0           ['tf.tile[0][0]',                \n",
-      "                                                                  're_lu_7[0][0]']                \n",
-      "                                                                                                  \n",
-      " zero_padding2d_3 (ZeroPadding2  (None, 18, 18, 640)  0          ['tf.concat[0][0]']              \n",
-      " D)                                                                                               \n",
-      "                                                                                                  \n",
-      " conv2d_14 (Conv2D)             (None, 16, 16, 512)  2949120     ['zero_padding2d_3[0][0]']       \n",
-      "                                                                                                  \n",
-      " batch_normalization_10 (BatchN  (None, 16, 16, 512)  2048       ['conv2d_14[0][0]']              \n",
-      " ormalization)                                                                                    \n",
-      "                                                                                                  \n",
-      " re_lu_8 (ReLU)                 (None, 16, 16, 512)  0           ['batch_normalization_10[0][0]'] \n",
-      "                                                                                                  \n",
-      " conv2d_15 (Conv2D)             (None, 16, 16, 512)  2359296     ['re_lu_8[0][0]']                \n",
-      "                                                                                                  \n",
-      " batch_normalization_11 (BatchN  (None, 16, 16, 512)  2048       ['conv2d_15[0][0]']              \n",
-      " ormalization)                                                                                    \n",
-      "                                                                                                  \n",
-      " re_lu_9 (ReLU)                 (None, 16, 16, 512)  0           ['batch_normalization_11[0][0]'] \n",
-      "                                                                                                  \n",
-      " conv2d_16 (Conv2D)             (None, 16, 16, 512)  2359296     ['re_lu_9[0][0]']                \n",
-      "                                                                                                  \n",
-      " batch_normalization_12 (BatchN  (None, 16, 16, 512)  2048       ['conv2d_16[0][0]']              \n",
-      " ormalization)                                                                                    \n",
-      "                                                                                                  \n",
-      " add (Add)                      (None, 16, 16, 512)  0           ['batch_normalization_12[0][0]', \n",
-      "                                                                  're_lu_8[0][0]']                \n",
-      "                                                                                                  \n",
-      " re_lu_10 (ReLU)                (None, 16, 16, 512)  0           ['add[0][0]']                    \n",
-      "                                                                                                  \n",
-      " conv2d_17 (Conv2D)             (None, 16, 16, 512)  2359296     ['re_lu_10[0][0]']               \n",
-      "                                                                                                  \n",
-      " batch_normalization_13 (BatchN  (None, 16, 16, 512)  2048       ['conv2d_17[0][0]']              \n",
-      " ormalization)                                                                                    \n",
-      "                                                                                                  \n",
-      " re_lu_11 (ReLU)                (None, 16, 16, 512)  0           ['batch_normalization_13[0][0]'] \n",
-      "                                                                                                  \n",
-      " conv2d_18 (Conv2D)             (None, 16, 16, 512)  2359296     ['re_lu_11[0][0]']               \n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "                                                                                                  \n",
-      " batch_normalization_14 (BatchN  (None, 16, 16, 512)  2048       ['conv2d_18[0][0]']              \n",
-      " ormalization)                                                                                    \n",
-      "                                                                                                  \n",
-      " add_1 (Add)                    (None, 16, 16, 512)  0           ['batch_normalization_14[0][0]', \n",
-      "                                                                  're_lu_10[0][0]']               \n",
-      "                                                                                                  \n",
-      " re_lu_12 (ReLU)                (None, 16, 16, 512)  0           ['add_1[0][0]']                  \n",
-      "                                                                                                  \n",
-      " conv2d_19 (Conv2D)             (None, 16, 16, 512)  2359296     ['re_lu_12[0][0]']               \n",
-      "                                                                                                  \n",
-      " batch_normalization_15 (BatchN  (None, 16, 16, 512)  2048       ['conv2d_19[0][0]']              \n",
-      " ormalization)                                                                                    \n",
-      "                                                                                                  \n",
-      " re_lu_13 (ReLU)                (None, 16, 16, 512)  0           ['batch_normalization_15[0][0]'] \n",
-      "                                                                                                  \n",
-      " conv2d_20 (Conv2D)             (None, 16, 16, 512)  2359296     ['re_lu_13[0][0]']               \n",
-      "                                                                                                  \n",
-      " batch_normalization_16 (BatchN  (None, 16, 16, 512)  2048       ['conv2d_20[0][0]']              \n",
-      " ormalization)                                                                                    \n",
-      "                                                                                                  \n",
-      " add_2 (Add)                    (None, 16, 16, 512)  0           ['batch_normalization_16[0][0]', \n",
-      "                                                                  're_lu_12[0][0]']               \n",
-      "                                                                                                  \n",
-      " re_lu_14 (ReLU)                (None, 16, 16, 512)  0           ['add_2[0][0]']                  \n",
-      "                                                                                                  \n",
-      " conv2d_21 (Conv2D)             (None, 16, 16, 512)  2359296     ['re_lu_14[0][0]']               \n",
-      "                                                                                                  \n",
-      " batch_normalization_17 (BatchN  (None, 16, 16, 512)  2048       ['conv2d_21[0][0]']              \n",
-      " ormalization)                                                                                    \n",
-      "                                                                                                  \n",
-      " re_lu_15 (ReLU)                (None, 16, 16, 512)  0           ['batch_normalization_17[0][0]'] \n",
-      "                                                                                                  \n",
-      " conv2d_22 (Conv2D)             (None, 16, 16, 512)  2359296     ['re_lu_15[0][0]']               \n",
-      "                                                                                                  \n",
-      " batch_normalization_18 (BatchN  (None, 16, 16, 512)  2048       ['conv2d_22[0][0]']              \n",
-      " ormalization)                                                                                    \n",
-      "                                                                                                  \n",
-      " add_3 (Add)                    (None, 16, 16, 512)  0           ['batch_normalization_18[0][0]', \n",
-      "                                                                  're_lu_14[0][0]']               \n",
-      "                                                                                                  \n",
-      " re_lu_16 (ReLU)                (None, 16, 16, 512)  0           ['add_3[0][0]']                  \n",
-      "                                                                                                  \n",
-      " up_sampling2d_4 (UpSampling2D)  (None, 32, 32, 512)  0          ['re_lu_16[0][0]']               \n",
-      "                                                                                                  \n",
-      " conv2d_23 (Conv2D)             (None, 32, 32, 512)  2359296     ['up_sampling2d_4[0][0]']        \n",
-      "                                                                                                  \n",
-      " batch_normalization_19 (BatchN  (None, 32, 32, 512)  2048       ['conv2d_23[0][0]']              \n",
-      " ormalization)                                                                                    \n",
-      "                                                                                                  \n",
-      " re_lu_17 (ReLU)                (None, 32, 32, 512)  0           ['batch_normalization_19[0][0]'] \n",
-      "                                                                                                  \n",
-      " up_sampling2d_5 (UpSampling2D)  (None, 64, 64, 512)  0          ['re_lu_17[0][0]']               \n",
-      "                                                                                                  \n",
-      " conv2d_24 (Conv2D)             (None, 64, 64, 256)  1179648     ['up_sampling2d_5[0][0]']        \n",
-      "                                                                                                  \n",
-      " batch_normalization_20 (BatchN  (None, 64, 64, 256)  1024       ['conv2d_24[0][0]']              \n",
-      " ormalization)                                                                                    \n",
-      "                                                                                                  \n",
-      " re_lu_18 (ReLU)                (None, 64, 64, 256)  0           ['batch_normalization_20[0][0]'] \n",
-      "                                                                                                  \n",
-      " up_sampling2d_6 (UpSampling2D)  (None, 128, 128, 25  0          ['re_lu_18[0][0]']               \n",
-      "                                6)                                                                \n",
-      "                                                                                                  \n",
-      " conv2d_25 (Conv2D)             (None, 128, 128, 12  294912      ['up_sampling2d_6[0][0]']        \n",
-      "                                8)                                                                \n",
-      "                                                                                                  \n",
-      " batch_normalization_21 (BatchN  (None, 128, 128, 12  512        ['conv2d_25[0][0]']              \n",
-      " ormalization)                  8)                                                                \n",
-      "                                                                                                  \n",
-      " re_lu_19 (ReLU)                (None, 128, 128, 12  0           ['batch_normalization_21[0][0]'] \n",
-      "                                8)                                                                \n",
-      "                                                                                                  \n",
-      " up_sampling2d_7 (UpSampling2D)  (None, 256, 256, 12  0          ['re_lu_19[0][0]']               \n",
-      "                                8)                                                                \n",
-      "                                                                                                  \n",
-      " conv2d_26 (Conv2D)             (None, 256, 256, 64  73728       ['up_sampling2d_7[0][0]']        \n",
-      "                                )                                                                 \n",
-      "                                                                                                  \n",
-      " batch_normalization_22 (BatchN  (None, 256, 256, 64  256        ['conv2d_26[0][0]']              \n",
-      " ormalization)                  )                                                                 \n",
-      "                                                                                                  \n",
-      " re_lu_20 (ReLU)                (None, 256, 256, 64  0           ['batch_normalization_22[0][0]'] \n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "                                )                                                                 \n",
-      "                                                                                                  \n",
-      " conv2d_27 (Conv2D)             (None, 256, 256, 3)  1728        ['re_lu_20[0][0]']               \n",
-      "                                                                                                  \n",
-      " activation_2 (Activation)      (None, 256, 256, 3)  0           ['conv2d_27[0][0]']              \n",
-      "                                                                                                  \n",
-      "==================================================================================================\n",
-      "Total params: 28,645,440\n",
-      "Trainable params: 28,632,768\n",
-      "Non-trainable params: 12,672\n",
-      "__________________________________________________________________________________________________\n"
-     ]
-    }
-   ],
-   "source": [
-    "generator_stage2 = build_stage2_generator()\n",
-    "generator_stage2.summary()"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 23,
-   "id": "40e89d49",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "############################################################\n",
-    "# Stage 2 Discriminator Network\n",
-    "############################################################\n",
-    "\n",
-    "def build_stage2_discriminator():\n",
-    "    \"\"\"Builds the Stage 2 Discriminator that uses the 256x256 resolution images from the generator\n",
-    "    and the compressed and spatially replicated embeddings.\n",
-    "\n",
-    "    Returns:\n",
-    "        Stage 2 Discriminator Model for StackGAN.\n",
-    "    \"\"\"\n",
-    "    input_layer1 = Input(shape=(256, 256, 3))\n",
-    "\n",
-    "    x = Conv2D(64, kernel_size=(4,4), padding='same', strides=2, use_bias=False,\n",
-    "                kernel_initializer='he_uniform')(input_layer1)\n",
-    "    x = LeakyReLU(alpha=0.2)(x)\n",
-    "\n",
-    "    x = ConvBlock(x, 128)\n",
-    "    x = ConvBlock(x, 256)\n",
-    "    x = ConvBlock(x, 512)\n",
-    "    x = ConvBlock(x, 1024)\n",
-    "    x = ConvBlock(x, 2048)\n",
-    "    x = ConvBlock(x, 1024, (1,1), 1)\n",
-    "    x = ConvBlock(x, 512, (1,1), 1, False)\n",
-    "\n",
-    "    x1 = ConvBlock(x, 128, (1,1), 1)\n",
-    "    x1 = ConvBlock(x1, 128, (3,3), 1)\n",
-    "    x1 = ConvBlock(x1, 512, (3,3), 1, False)\n",
-    "\n",
-    "    x2 = add([x, x1])\n",
-    "    x2 = LeakyReLU(alpha=0.2)(x2)\n",
-    "    # Concatenate compressed and spatially replicated embedding\n",
-    "    input_layer2 = Input(shape=(4, 4, 128))\n",
-    "    concat = concatenate([x2, input_layer2])\n",
-    "\n",
-    "    x3 = Conv2D(512, kernel_size=(1,1), strides=1, padding='same', kernel_initializer='he_uniform')(concat)\n",
-    "    x3 = BatchNormalization(gamma_initializer='ones', beta_initializer='zeros')(x3)\n",
-    "    x3 = LeakyReLU(alpha=0.2)(x3)\n",
-    "\n",
-    "    # Flatten and add a FC layer\n",
-    "    x3 = Flatten()(x3)\n",
-    "    x3 = Dense(1)(x3)\n",
-    "    x3 = Activation('sigmoid')(x3)\n",
-    "\n",
-    "    stage2_dis = Model(inputs=[input_layer1, input_layer2], outputs=[x3])\n",
-    "    return stage2_dis"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 24,
-   "id": "5ec36c5a",
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model: \"model_4\"\n",
-      "__________________________________________________________________________________________________\n",
-      " Layer (type)                   Output Shape         Param #     Connected to                     \n",
-      "==================================================================================================\n",
-      " input_13 (InputLayer)          [(None, 256, 256, 3  0           []                               \n",
-      "                                )]                                                                \n",
-      "                                                                                                  \n",
-      " conv2d_28 (Conv2D)             (None, 128, 128, 64  3072        ['input_13[0][0]']               \n",
-      "                                )                                                                 \n",
-      "                                                                                                  \n",
-      " leaky_re_lu_9 (LeakyReLU)      (None, 128, 128, 64  0           ['conv2d_28[0][0]']              \n",
-      "                                )                                                                 \n",
-      "                                                                                                  \n",
-      " conv2d_29 (Conv2D)             (None, 64, 64, 128)  131072      ['leaky_re_lu_9[0][0]']          \n",
-      "                                                                                                  \n",
-      " batch_normalization_23 (BatchN  (None, 64, 64, 128)  512        ['conv2d_29[0][0]']              \n",
-      " ormalization)                                                                                    \n",
-      "                                                                                                  \n",
-      " leaky_re_lu_10 (LeakyReLU)     (None, 64, 64, 128)  0           ['batch_normalization_23[0][0]'] \n",
-      "                                                                                                  \n",
-      " conv2d_30 (Conv2D)             (None, 32, 32, 256)  524288      ['leaky_re_lu_10[0][0]']         \n",
-      "                                                                                                  \n",
-      " batch_normalization_24 (BatchN  (None, 32, 32, 256)  1024       ['conv2d_30[0][0]']              \n",
-      " ormalization)                                                                                    \n",
-      "                                                                                                  \n",
-      " leaky_re_lu_11 (LeakyReLU)     (None, 32, 32, 256)  0           ['batch_normalization_24[0][0]'] \n",
-      "                                                                                                  \n",
-      " conv2d_31 (Conv2D)             (None, 16, 16, 512)  2097152     ['leaky_re_lu_11[0][0]']         \n",
-      "                                                                                                  \n",
-      " batch_normalization_25 (BatchN  (None, 16, 16, 512)  2048       ['conv2d_31[0][0]']              \n",
-      " ormalization)                                                                                    \n",
-      "                                                                                                  \n",
-      " leaky_re_lu_12 (LeakyReLU)     (None, 16, 16, 512)  0           ['batch_normalization_25[0][0]'] \n",
-      "                                                                                                  \n",
-      " conv2d_32 (Conv2D)             (None, 8, 8, 1024)   8388608     ['leaky_re_lu_12[0][0]']         \n",
-      "                                                                                                  \n",
-      " batch_normalization_26 (BatchN  (None, 8, 8, 1024)  4096        ['conv2d_32[0][0]']              \n",
-      " ormalization)                                                                                    \n",
-      "                                                                                                  \n",
-      " leaky_re_lu_13 (LeakyReLU)     (None, 8, 8, 1024)   0           ['batch_normalization_26[0][0]'] \n",
-      "                                                                                                  \n",
-      " conv2d_33 (Conv2D)             (None, 4, 4, 2048)   33554432    ['leaky_re_lu_13[0][0]']         \n",
-      "                                                                                                  \n",
-      " batch_normalization_27 (BatchN  (None, 4, 4, 2048)  8192        ['conv2d_33[0][0]']              \n",
-      " ormalization)                                                                                    \n",
-      "                                                                                                  \n",
-      " leaky_re_lu_14 (LeakyReLU)     (None, 4, 4, 2048)   0           ['batch_normalization_27[0][0]'] \n",
-      "                                                                                                  \n",
-      " conv2d_34 (Conv2D)             (None, 4, 4, 1024)   2097152     ['leaky_re_lu_14[0][0]']         \n",
-      "                                                                                                  \n",
-      " batch_normalization_28 (BatchN  (None, 4, 4, 1024)  4096        ['conv2d_34[0][0]']              \n",
-      " ormalization)                                                                                    \n",
-      "                                                                                                  \n",
-      " leaky_re_lu_15 (LeakyReLU)     (None, 4, 4, 1024)   0           ['batch_normalization_28[0][0]'] \n",
-      "                                                                                                  \n",
-      " conv2d_35 (Conv2D)             (None, 4, 4, 512)    524288      ['leaky_re_lu_15[0][0]']         \n",
-      "                                                                                                  \n",
-      " batch_normalization_29 (BatchN  (None, 4, 4, 512)   2048        ['conv2d_35[0][0]']              \n",
-      " ormalization)                                                                                    \n",
-      "                                                                                                  \n",
-      " conv2d_36 (Conv2D)             (None, 4, 4, 128)    65536       ['batch_normalization_29[0][0]'] \n",
-      "                                                                                                  \n",
-      " batch_normalization_30 (BatchN  (None, 4, 4, 128)   512         ['conv2d_36[0][0]']              \n",
-      " ormalization)                                                                                    \n",
-      "                                                                                                  \n",
-      " leaky_re_lu_16 (LeakyReLU)     (None, 4, 4, 128)    0           ['batch_normalization_30[0][0]'] \n",
-      "                                                                                                  \n",
-      " conv2d_37 (Conv2D)             (None, 4, 4, 128)    147456      ['leaky_re_lu_16[0][0]']         \n",
-      "                                                                                                  \n",
-      " batch_normalization_31 (BatchN  (None, 4, 4, 128)   512         ['conv2d_37[0][0]']              \n",
-      " ormalization)                                                                                    \n",
-      "                                                                                                  \n",
-      " leaky_re_lu_17 (LeakyReLU)     (None, 4, 4, 128)    0           ['batch_normalization_31[0][0]'] \n",
-      "                                                                                                  \n",
-      " conv2d_38 (Conv2D)             (None, 4, 4, 512)    589824      ['leaky_re_lu_17[0][0]']         \n",
-      "                                                                                                  \n",
-      " batch_normalization_32 (BatchN  (None, 4, 4, 512)   2048        ['conv2d_38[0][0]']              \n",
-      " ormalization)                                                                                    \n",
-      "                                                                                                  \n",
-      " add_4 (Add)                    (None, 4, 4, 512)    0           ['batch_normalization_29[0][0]', \n",
-      "                                                                  'batch_normalization_32[0][0]'] \n",
-      "                                                                                                  \n",
-      " leaky_re_lu_18 (LeakyReLU)     (None, 4, 4, 512)    0           ['add_4[0][0]']                  \n",
-      "                                                                                                  \n"
-     ]
-    },
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      " input_14 (InputLayer)          [(None, 4, 4, 128)]  0           []                               \n",
-      "                                                                                                  \n",
-      " concatenate_2 (Concatenate)    (None, 4, 4, 640)    0           ['leaky_re_lu_18[0][0]',         \n",
-      "                                                                  'input_14[0][0]']               \n",
-      "                                                                                                  \n",
-      " conv2d_39 (Conv2D)             (None, 4, 4, 512)    328192      ['concatenate_2[0][0]']          \n",
-      "                                                                                                  \n",
-      " batch_normalization_33 (BatchN  (None, 4, 4, 512)   2048        ['conv2d_39[0][0]']              \n",
-      " ormalization)                                                                                    \n",
-      "                                                                                                  \n",
-      " leaky_re_lu_19 (LeakyReLU)     (None, 4, 4, 512)    0           ['batch_normalization_33[0][0]'] \n",
-      "                                                                                                  \n",
-      " flatten_1 (Flatten)            (None, 8192)         0           ['leaky_re_lu_19[0][0]']         \n",
-      "                                                                                                  \n",
-      " dense_5 (Dense)                (None, 1)            8193        ['flatten_1[0][0]']              \n",
-      "                                                                                                  \n",
-      " activation_3 (Activation)      (None, 1)            0           ['dense_5[0][0]']                \n",
-      "                                                                                                  \n",
-      "==================================================================================================\n",
-      "Total params: 48,486,401\n",
-      "Trainable params: 48,472,833\n",
-      "Non-trainable params: 13,568\n",
-      "__________________________________________________________________________________________________\n"
-     ]
-    }
-   ],
-   "source": [
-    "discriminator_stage2 = build_stage2_discriminator()\n",
-    "discriminator_stage2.summary()"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 25,
-   "id": "f95b2b41",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "############################################################\n",
-    "# Stage 2 Adversarial Model\n",
-    "############################################################\n",
-    "\n",
-    "def stage2_adversarial_network(stage2_disc, stage2_gen, stage1_gen):\n",
-    "    \"\"\"Stage 2 Adversarial Network.\n",
-    "\n",
-    "    Args:\n",
-    "        stage2_disc: Stage 2 Discriminator Model.\n",
-    "        stage2_gen: Stage 2 Generator Model.\n",
-    "        stage1_gen: Stage 1 Generator Model.\n",
-    "\n",
-    "    Returns:\n",
-    "        Stage 2 Adversarial network.\n",
-    "    \"\"\"\n",
-    "    conditioned_embedding = Input(shape=(1024, ))\n",
-    "    latent_space = Input(shape=(100, ))\n",
-    "    compressed_replicated = Input(shape=(4, 4, 128))\n",
-    "    \n",
-    "    #the discriminator is trained separately and stage1_gen already trained, and this is the reason why we freeze its layers by setting the property trainable=false\n",
-    "    input_images, ca = stage1_gen([conditioned_embedding, latent_space])\n",
-    "    stage2_disc.trainable = False\n",
-    "    stage1_gen.trainable = False\n",
-    "\n",
-    "    images, ca2 = stage2_gen([conditioned_embedding, input_images])\n",
-    "    probability = stage2_disc([images, compressed_replicated])\n",
-    "\n",
-    "    return Model(inputs=[conditioned_embedding, latent_space, compressed_replicated],\n",
-    "        outputs=[probability, ca2])"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 26,
-   "id": "2ba0c590",
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Model: \"model_5\"\n",
-      "__________________________________________________________________________________________________\n",
-      " Layer (type)                   Output Shape         Param #     Connected to                     \n",
-      "==================================================================================================\n",
-      " input_15 (InputLayer)          [(None, 1024)]       0           []                               \n",
-      "                                                                                                  \n",
-      " input_16 (InputLayer)          [(None, 100)]        0           []                               \n",
-      "                                                                                                  \n",
-      " model (Functional)             [(None, 64, 64, 3),  10270400    ['input_15[0][0]',               \n",
-      "                                 (None, 256)]                     'input_16[0][0]']               \n",
-      "                                                                                                  \n",
-      " model_3 (Functional)           [(None, 256, 256, 3  28645440    ['input_15[0][0]',               \n",
-      "                                ),                                'model[1][0]']                  \n",
-      "                                 (None, 256)]                                                     \n",
-      "                                                                                                  \n",
-      " input_17 (InputLayer)          [(None, 4, 4, 128)]  0           []                               \n",
-      "                                                                                                  \n",
-      " model_4 (Functional)           (None, 1)            48486401    ['model_3[0][0]',                \n",
-      "                                                                  'input_17[0][0]']               \n",
-      "                                                                                                  \n",
-      "==================================================================================================\n",
-      "Total params: 87,402,241\n",
-      "Trainable params: 28,632,768\n",
-      "Non-trainable params: 58,769,473\n",
-      "__________________________________________________________________________________________________\n"
-     ]
-    }
-   ],
-   "source": [
-    "adversarial_stage2 = stage2_adversarial_network(discriminator_stage2, generator_stage2, generator)\n",
-    "adversarial_stage2.summary()"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 27,
-   "id": "c7cf4596",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "class StackGanStage2(object):\n",
-    "    \"\"\"StackGAN Stage 2 class.\n",
-    "\n",
-    "    Args:\n",
-    "        epochs: Number of epochs\n",
-    "        z_dim: Latent space dimensions\n",
-    "        batch_size: Batch Size\n",
-    "        enable_function: If True, training function is decorated with tf.function\n",
-    "        stage2_generator_lr: Learning rate for stage 2 generator\n",
-    "        stage2_discriminator_lr: Learning rate for stage 2 discriminator\n",
-    "    \"\"\"\n",
-    "    def __init__(self, epochs=500, z_dim=100, batch_size=64, enable_function=True, stage2_generator_lr=0.0002, stage2_discriminator_lr=0.0002):\n",
-    "        self.epochs = epochs\n",
-    "        self.z_dim = z_dim\n",
-    "        self.enable_function = enable_function\n",
-    "        self.stage1_generator_lr = stage2_generator_lr\n",
-    "        self.stage1_discriminator_lr = stage2_discriminator_lr\n",
-    "        self.low_image_size = 64\n",
-    "        self.high_image_size = 256\n",
-    "        self.conditioning_dim = 128\n",
-    "        self.batch_size = batch_size\n",
-    "        self.stage2_generator_optimizer = Adam(lr=stage2_generator_lr, beta_1=0.5, beta_2=0.999)\n",
-    "        self.stage2_discriminator_optimizer = Adam(lr=stage2_discriminator_lr, beta_1=0.5, beta_2=0.999)\n",
-    "        self.stage1_generator = build_stage1_generator()\n",
-    "        self.stage1_generator.compile(loss='binary_crossentropy', optimizer=self.stage2_generator_optimizer)\n",
-    "        self.stage1_generator.load_weights('weights/stage1_gen.h5')\n",
-    "        self.stage2_generator = build_stage2_generator()\n",
-    "        self.stage2_generator.compile(loss='binary_crossentropy', optimizer=self.stage2_generator_optimizer)\n",
-    "\n",
-    "        self.stage2_discriminator = build_stage2_discriminator()\n",
-    "        self.stage2_discriminator.compile(loss='binary_crossentropy', optimizer=self.stage2_discriminator_optimizer)\n",
-    "\n",
-    "        self.ca_network = build_ca_network()\n",
-    "        self.ca_network.compile(loss='binary_crossentropy', optimizer='Adam')\n",
-    "\n",
-    "        self.embedding_compressor = build_embedding_compressor()\n",
-    "        self.embedding_compressor.compile(loss='binary_crossentropy', optimizer='Adam')\n",
-    "\n",
-    "        self.stage2_adversarial = stage2_adversarial_network(self.stage2_discriminator, self.stage2_generator, self.stage1_generator)\n",
-    "        self.stage2_adversarial.compile(loss=['binary_crossentropy', adversarial_loss], loss_weights=[1, 2.0], optimizer=self.stage2_generator_optimizer)\t\n",
-    "\n",
-    "        self.checkpoint2 = tf.train.Checkpoint(\n",
-    "            generator_optimizer=self.stage2_generator_optimizer,\n",
-    "            discriminator_optimizer=self.stage2_discriminator_optimizer,\n",
-    "            generator=self.stage2_generator,\n",
-    "            discriminator=self.stage2_discriminator,\n",
-    "            generator1=self.stage1_generator)\n",
-    "\n",
-    "    def visualize_stage2(self):\n",
-    "        \"\"\"Running Tensorboard visualizations.\n",
-    "        \"\"\"\n",
-    "        tb = TensorBoard(log_dir=\"logs/\".format(time.time()))\n",
-    "        tb.set_model(self.stage2_generator)\n",
-    "        tb.set_model(self.stage2_discriminator)\n",
-    "\n",
-    "    def train_stage2(self):\n",
-    "        \"\"\"Trains Stage 2 StackGAN.\n",
-    "        \"\"\"\n",
-    "        x_high_train, y_high_train, high_train_embeds = load_data(filename_path=filename_path_train, class_id_path=class_id_path_train,\n",
-    "      dataset_path=dataset_path, embeddings_path=embeddings_path_train, size=(256, 256))\n",
-    "\n",
-    "        x_high_test, y_high_test, high_test_embeds = load_data(filename_path=filename_path_test, class_id_path=class_id_path_test, \n",
-    "      dataset_path=dataset_path, embeddings_path=embeddings_path_test, size=(256, 256))\n",
-    "\n",
-    "        x_low_train, y_low_train, low_train_embeds = load_data(filename_path=filename_path_train, class_id_path=class_id_path_train,\n",
-    "      dataset_path=dataset_path, embeddings_path=embeddings_path_train, size=(64, 64))\n",
-    "\n",
-    "        x_low_test, y_low_test, low_test_embeds = load_data(filename_path=filename_path_test, class_id_path=class_id_path_test, \n",
-    "      dataset_path=dataset_path, embeddings_path=embeddings_path_test, size=(64, 64))\n",
-    "\n",
-    "        real = np.ones((self.batch_size, 1), dtype='float') * 0.9\n",
-    "        fake = np.zeros((self.batch_size, 1), dtype='float') * 0.1\n",
-    "\n",
-    "        for epoch in range(self.epochs):\n",
-    "            print(f'Epoch: {epoch}')\n",
-    "\n",
-    "            gen_loss = []\n",
-    "            disc_loss = []\n",
-    "\n",
-    "            num_batches = int(x_high_train.shape[0] / self.batch_size)\n",
-    "\n",
-    "            for i in range(num_batches):\n",
-    "\n",
-    "                latent_space = np.random.normal(0, 1, size=(self.batch_size, self.z_dim))\n",
-    "                embedding_text = high_train_embeds[i * self.batch_size:(i + 1) * self.batch_size]\n",
-    "                compressed_embedding = self.embedding_compressor.predict_on_batch(embedding_text)\n",
-    "                compressed_embedding = np.reshape(compressed_embedding, (-1, 1, 1, self.conditioning_dim))\n",
-    "                compressed_embedding = np.tile(compressed_embedding, (1, 4, 4, 1))\n",
-    "\n",
-    "                image_batch = x_high_train[i * self.batch_size:(i+1) * self.batch_size]\n",
-    "                image_batch = (image_batch - 127.5) / 127.5\n",
-    "                low_res_fakes, _ = self.stage1_generator.predict([embedding_text, latent_space], verbose=3)\n",
-    "                high_res_fakes, _ = self.stage2_generator.predict([embedding_text, low_res_fakes], verbose=3)\n",
-    "\n",
-    "                discriminator_loss = self.stage2_discriminator.train_on_batch([image_batch, compressed_embedding],\n",
-    "                    np.reshape(real, (self.batch_size, 1)))\n",
-    "\n",
-    "                discriminator_loss_gen = self.stage2_discriminator.train_on_batch([high_res_fakes, compressed_embedding],\n",
-    "                    np.reshape(fake, (self.batch_size, 1)))\n",
-    "\n",
-    "                discriminator_loss_fake = self.stage2_discriminator.train_on_batch([image_batch[:(self.batch_size-1)], compressed_embedding[1:]],\n",
-    "                    np.reshape(fake[1:], (self.batch_size - 1, 1)))\n",
-    "\n",
-    "                d_loss = 0.5 * np.add(discriminator_loss, 0.5 * np.add(discriminator_loss_gen, discriminator_loss_fake))\n",
-    "                disc_loss.append(d_loss)\n",
-    "\n",
-    "                print(f'Discriminator Loss: {d_loss}')\n",
-    "\n",
-    "                g_loss = self.stage2_adversarial.train_on_batch([embedding_text, latent_space, compressed_embedding],\n",
-    "                    [K.ones((self.batch_size, 1)) * 0.9, K.ones((self.batch_size, 256)) * 0.9])\n",
-    "                gen_loss.append(g_loss)\n",
-    "\n",
-    "                print(f'Generator Loss: {g_loss}')\n",
-    "\n",
-    "                if epoch % 5 == 0:\n",
-    "                    latent_space = np.random.normal(0, 1, size=(self.batch_size, self.z_dim))\n",
-    "                    embedding_batch = high_test_embeds[0 : self.batch_size]\n",
-    "\n",
-    "                    low_fake_images, _ = self.stage1_generator.predict([embedding_batch, latent_space], verbose=3)\n",
-    "                    high_fake_images, _ = self.stage2_generator.predict([embedding_batch, low_fake_images], verbose=3)\n",
-    "\n",
-    "                    for i, image in enumerate(high_fake_images[:10]):\n",
-    "                        save_image(image, f'results_stage2/gen_{epoch}_{i}.png')\n",
-    "                if epoch % 10 == 0:\n",
-    "                    self.stage2_generator.save_weights('weights/stage2_gen.h5')\n",
-    "                    self.stage2_discriminator.save_weights(\"weights/stage2_disc.h5\")\n",
-    "                    self.ca_network.save_weights('weights/stage2_ca.h5')\n",
-    "                    self.embedding_compressor.save_weights('weights/stage2_embco.h5')\n",
-    "                    self.stage2_adversarial.save_weights('weights/stage2_adv.h5')\n",
-    "\n",
-    "        self.stage2_generator.save_weights('weights/stage2_gen.h5')\n",
-    "        self.stage2_discriminator.save_weights(\"weights/stage2_disc.h5\")"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "359dd5d8",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "stage2 = StackGanStage2()\n",
-    "stage2.train_stage2()"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "dba73ffa",
-   "metadata": {},
-   "outputs": [],
-   "source": []
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "82181ecb",
-   "metadata": {},
-   "outputs": [],
-   "source": []
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 10,
-   "id": "0ea29319",
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Found 11788 files belonging to 3 classes.\n",
-      "Using 9431 files for training.\n"
-     ]
-    }
-   ],
-   "source": [
-    "train=keras.utils.image_dataset_from_directory(directory=\"D:\\\\unikaksha\\\\CUB_200_2011\\\\CUB_200_2011\",\n",
-    "                                             labels=\"inferred\",\n",
-    "                                             validation_split=0.2,\n",
-    "                                             subset=\"training\",\n",
-    "                                             seed=1337,\n",
-    "                                             label_mode=\"int\",\n",
-    "                                             batch_size=32,\n",
-    "                                             image_size=(256,256))"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 11,
-   "id": "e9e92c9e",
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stdout",
-     "output_type": "stream",
-     "text": [
-      "Found 11788 files belonging to 3 classes.\n",
-      "Using 2357 files for validation.\n"
-     ]
-    }
-   ],
-   "source": [
-    "cub_test=keras.utils.image_dataset_from_directory(directory=\"D:\\\\unikaksha\\\\CUB_200_2011\\\\CUB_200_2011\",\n",
-    "                                             labels=\"inferred\",\n",
-    "                                             validation_split=0.2,\n",
-    "                                             subset=\"validation\",\n",
-    "                                             seed=1337,\n",
-    "                                             label_mode=\"int\",\n",
-    "                                             batch_size=32,\n",
-    "                                             image_size=(256,256))"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 12,
-   "id": "633699db",
-   "metadata": {},
-   "outputs": [
-    {
-     "data": {
-      "image/png": "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\n",
-      "text/plain": [
-       "<Figure size 640x480 with 1 Axes>"
-      ]
-     },
-     "metadata": {},
-     "output_type": "display_data"
-    },
-    {
-     "data": {
-      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAa8AAAGiCAYAAABQ9UnfAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy88F64QAAAACXBIWXMAAA9hAAAPYQGoP6dpAAEAAElEQVR4nOz9z69tS3bXiX5GRMy51j4/7nU6s5zpBEwZyi5R4AYYPWNKFnSw5CfxMHSQkBBNkFuWRQPLHRtZtqCBaJRAooVpIPEP8Bqmg5DcQzSeLIT8eAijgnRiOzPvvefsteaMGOM1RkTMmHOtfW6mL66so9xx7zp7rfkjZvya4zt+h5iZ8Vyey3N5Ls/lubxHJXy7G/BcnstzeS7P5bl8q+UZvJ7Lc3kuz+W5vHflGbyey3N5Ls/lubx35Rm8nstzeS7P5bm8d+UZvJ7Lc3kuz+W5vHflGbyey3N5Ls/lubx35Rm8nstzeS7P5bm8d+UZvJ7Lc3kuz+W5vHflGbyey3N5Ls/lubx35Rm8nstzeS7P5bm8d+XbCl7/6B/9I77/+7+f8/nMD//wD/Nv/s2/+XY257k8l+fyXJ7Le1K+beD1L/7Fv+Cnf/qn+bmf+zn+3b/7d/zYj/0YP/ETP8Fv/uZvfrua9Fyey3N5Ls/lPSny7UrM+yM/8iP8qT/1p/jH//gf92N/7I/9MX7yJ3+SX/7lX/52NOm5PJfn8lyey3tS0rfjocuy8G//7b/l7/ydv7M7/uM//uP82q/92s311+uV6/Xaf6sqv/u7v8vnP/95ROT3vb3P5bk8l+fyXP7HFjPj448/5stf/jIhfOtKwG8LeP32b/82pRS++MUv7o5/8Ytf5Ctf+crN9b/8y7/ML/zCL/xf1bzn8lyey3N5Lv8Xlf/yX/4Lf/AP/sFv+b5vC3i1cpSazOyuJPWzP/uz/MzP/Ez//Y1vfIPv+77v4//xw3+KNE2AEAzAar3+jwGGYPXcRDvWn9j/NbbnCtv94++7Mp6NdWy/P620fpoZihHMaP9hYGpQj1Hb158voEPbto9sz5fW7vrP2BkRRIQA/VMb5X9Ce5Zg9bsg/XwIrXKIeF0iYV9/+682Tuoz23lvnxDpF2Ctg7KNj4ggIRBCRCTUz74+keBtrte2voR2DWFrv+zHf/vrD5Yg2/iI9Odszw5+LozHtz5oXQOyW1vqjx0Wn7b5MPU1YObHzDBTtB43rWsBMNOhEl8BYtv3YXkg+MvtVxu5rJRSsKJYLpScUS1+TK1fWbJS1CjFWEuhqLelmNbmevvMwCdMdvPVitX5DHX029qhz0k9ZKDF+MJ3v+B7v+cF/9sPfsirz70ippcoH/C7bwrf+PiRjz6+8vHHj5hlUoQPX594eYokK4THC+eipABTCqwpYvMM84nPfeEFq8KbJfOffvNrvHlTMIMX5wkmQbOyrisPYnwYhVfJ+2TF35mTGacgSBLyaeIxRTKQVTE1phg4TYkXHyTEDC3G248XvvbmysfXwseP2mfHJKBasDVTcua6LswSOU8TX/rCh3zuu06kKbCUwpu3j5RcEDNeponzHDnNiXBOPObC43Xhd/7r1/nkk0cu18xlKZgJp5R4+WLiww9PmBTWvPC7X/2INx8/crksXDNknzxmEb74uQ/43Aev+PznP2Q6T74KzJAEcRIkQFkKZS2oFjQvXK6JOE2cXp14/eqEKZSixADXXLgsmd/+rd9l+eQtFOXhlCBG1CAXJcYAZlyuK//Hv/h/8/r1a34v5dsCXl/4wheIMd5IWV/96ldvpDGA0+nE6XS6OZ5SIqZEA69GMCT4y1FJRqfnAX+hGt0I9UQjCa04fT1A1fB7d2aoY/ewO3WOx0bwCg0+zWgmSFOt360Da69DNsAJwAY1lRoY98Grg0MlvGzg1QC6gwGNNm1EPrTnBDqYBxqhD32YNiBr4LWBwAheVPBqwNLnRiCMzw0BKkjEGA/g5QASQ6h13IJPH6UnwKvX1Y6LbGAoEEIFLTbwkhA28AqxD/MteDlz0iZfKk5oZUxCZV7MjFIXkZqiWgBDFUc68+OVHUPYvndA021NdKak3hejkHPGglKAIIYWCOLtpa69KIG1GATDRBAtqAXEFLV63EJl1GRbe0+AV19jbU2E7R7ntYwgcJ5OnE8nXrw88erlTJzOSHiBJl//RY3LktEipCC8Oj/w6iEyoUiAOa9ECcQ48fLDmXg+k84nXr18IGO8XDKXS+aTNyuC8OrhjCRhzYXHy0oCXgbhRQArSlkNKzCVzGkOTHMivD7z6mHCYsAIaFEiwiSBNDtwrauyLpGwGEkzcyXs3llhBVZRiglqAY0JixMWEhImpilyOoNEYVkzy5IRSYQ0kaaJeI6UspJFYRJWg6sqVy2U7O/ehxZ4cYrMUyBI4INl4WMyb6Tw9rqy4kCaQuJkBbteuHwUKflMSok0T5xPE9MpIlG4hpVFVrQEQgqVRkQmTaQ4IUkwC4QAkxhnzagVPo6BfFkIKmjQyqxOnE+RNCXml3p37Xyz5dsCXvM888M//MP86q/+Kn/5L//lfvxXf/VX+Ut/6S998xU1YtRpg7/KofOegyRme/608atPDptZI6tOxK3JQP5ijm0QF5V6G+5Wd1P9Rtz6WfG26yCBbm3cS35itklE/WA/ebjHnLC3o7aBnkhthxxBcHxY63etwUZphS4xCk1CtD7o1uprknEH7WFURlAdxmpofZ1Z6b92betizVN6cwVClWodlJ6S8jdBcDvn13JnsQzy1TteQEH66NwyNAId3qwzLEiTuKrUYrd33pZt9Wm/zipoCmLSJSZtAGqGqVb8q79Nq4RV6nMbg1a/V0lRCNu6sf0Y2DBPo7ZjY6jqcYHSxrw+RdqbLEaISkowTcI8CRojMQopNSk5wHmCLJgJxQLzaSadEmmOCJFghQS8Ok0EVTDhYQ4OXgJBjXX1NazmQBuTYMHQqNic4DwxPZyYHyZMhGKQTQnF3yfBpZSQhHQOzGvgqoI9FvJavO9BsOzS9Aj8hpHVWLKScuAhJc6nyaV7QFTQAFn82gIUvA1FhCJCNiWrkat0HYsyRUgBCoYGJaXCSVceVVCJTDFyEkXWhTffKJRP3pLmE6fzmQ/KmekciUkouZDXTFHFivLmbcYskqKi0UgpEmMiBoEIxczbngKESFFnMMAQVUJITHMixfjkO/PNlG+b2vBnfuZn+Ot//a/zp//0n+ZHf/RH+Sf/5J/wm7/5m/ytv/W3vvlK2ouzo3jCEUA24cP677368G7VG11EmkJvq4+NqDVpwd9XGw7cqbcSorGuri6zRv7HKqTRsn1/kC45dmJgNhAB2Ah6lewGcrnv37HnT5engK3hkwww1tWMbVw6iN1/zAhNstG+fX9GItmJpWG3LTuUA+jdnH0CzO5UIfdGz44Nln5sW0PtsG4TL4ZtC+DmkXswH4H8TmkMQ32uUse+Vb1hTwWp9tcO4NVUg5UtGVTjYx2IOSAOzMqTLb8ztG1VIlsjpVfunxAgRpgn4XyKGIEUhCkJLpALIU5IFCpVR2J09bE4I6hFETVOKZBj8CWEORMnRhRhreNRxKX+EAKESjPmCFNEUiTFWF9zQ4ggiovMrs4OMTCfA/MSmVYlhNLHWsRBvy2EUFXgJsZalGUtTFE4T4EUA6QIqtjqa9PEmRI1B2o1QfG/xcxVvFUVLKUQihBNCaUwo4goITh4F4SIMlEoRbkuK28tEOfCaTUkwmmNxMnX8ZoLWZWclbePK0UFQbEZTvPEPClTCkj0/nRVd33lp6oNs+hvRAwRSZ8Nfr5t4PVX/+pf5Xd+53f4u3/37/Lf/tt/40/8iT/Bv/yX/5I//If/8Dddh9jwRlSC8RT52RFGNn7XDndt56Xy8Rv3eK+0l3j3kF7ZkWO+BSET2WDRxssdAo6yRCOEXUU6EOU9kN6HpnbdTa/uCDathkaXt6ufHmeDgxDixKlLX73uo5RFxbsKDUJXOXqb5UlwaRy8oYQ2YgdJgD5en17f0PLazlCBZlAv3vZ6Y2o6eG+8jC/PPeOCDLavT2lHA60NuA43mrAHtSbTG8126pKFA5PQ1s7IxjVJbAMsjG2mKuG1UYWxW0syHBuYpSeGeTsuu/YEfNwUIwY4TYKeIum7HkiTkCKkCq4hCqfphKwTZIVFQQUtwKo8lguookU5iXCxwDUr67I6oTUoGfJSCOIqzBgD0ySkFGFKpBQIKUKI1cbgg9BtvwlMDQlCEng1JRadKQgfvS0sxchFKagT9gBEIUpEkqAiPF6vRFF0TUymnF8lTjEynYWF3G2xKmAWsTJRNJJVyAprBitKJnNZLpQ1UkwQKZTLBbmuTNklRStKtqb+dXX1ssAbC+hV4W3mzZqZp8CUBOcLClmN6+JAthTlusLn19e8PE+8OE+8eJiI0Rf9ulzRNftiCXCaXJLUogQLRK3Mxmco31aHjZ/6qZ/ip37qp37P9w/a81qOBGuw57Tz/U5jg41BIqkXO9fphLBK+f3e23bsf5htNiWtBOselIxA2njbIy1rBLS1abyv9bf1o7VRbLhnaKA7DjRriW12pcM4joASDH9hdm1qFHkPSP3+qpLr0tNwObZJDs0G1dodhnb0PjQbFK72akzKUMsAQp2X5576VsaO7fq7Xd1JrkE17vUR2oFdqyu0GlyNtoHyVmNgUP/uVGutPRug+gmvqQwXtXsdk3U4PljXDMaR6FKYKWoFtYJRfO2rIqaIFefum/CldZ0NINaZAxuYsYZ5Yt02uvVM67Jo42fbCPT7tv76VaNitdoaxUjAhKHBePFiZnoIhGDkTy4sGYIJU0gQi0tAVrBrpvo7oTFga8aKSxhZjWVV3l4zsTpnoO6AEcUoAhakghjEU6KpSZfrSjEnmWaFeZ4hgoXIuhYoBVNnDGYCD1Pkw1eJIIE1O4AtcnW1pAimm6NRMeOyFILBJQXOD66GiyliSXERVJCQCLEQkyKxYGSKuTMOCivG47JyeftImAMlGIsIFiIqCQUyhWxGzhkxWE3I+N/FlNUybz5+QxIH4ySGRJf81iJc87WqKIVVC6+miQ/mifx6Ik2hqg8D0RQTpYgRSyQCRQwrC3lR9PJNcG7vKN9W8PrsxTpBMtkfv8cbezlQ4hHvZPgjA48rAzEa3t+nOHcZX+RB5bLh3x2JoGkTnpTxGEQr66AkDH0/dHtTQI7wuN1juz7I7s/usQO92l3WiNA9aO7nxmZtKLYH7f1z7lXV+hxsA2vM7RNPzILfddOOoTVmjES3qXa6p0FrXVXvHZmjbY6HcbwB+tHitWdPev9H6Xz42jULdeDNZNes7ZZNgmqSFjSpfZy8ep0phrr6sklTVdJx3xDjho2yQZ3dbhgGfjd3+wHYN6F2QPoCbEyjDBqBBnetHw5irr2rxF6q04q6t6Jk9U9RQhVnJfhKWYuia8GKq9aKKasq61JZDRNUlRQEjU095440ET9nppCNSaTbb7RAczTK2aAYoq7qFIQUAuc5Vc9NZc0OHVnU7Vd9TvyPmlX1n/XxDiEQkzssEQISAikpcXI7YIiNKXRmNIhPRlYlq9/P5G0oVdXo2lUlq6DeLVYTVtyhZDFD10IwdyiKYsTkTkqmwqUUilr1MjRsych1YQonTnNkmhLpnIjR1a85VzumQQrmEmiBkvdest9qee/Bq7Ly/deOr7Xt9W2G5UEeOSLeUIf00yPBHm+tvDKj9Naf36SN0VbAyI37r0647uCViGxM7kjD2GhdqH6KreaNsI0KJDkODE260CrNdMIe6oswUCIdwEraw2snpUlEO4ole3Vub0aVbxo4UIUqa3UNc1Wf5VqZ6kJeCVtgG9cOPPX8JmU3Vd/W790QVGmi2R+7d2S90BpQ9UNapY6h/uYiX8EYXEIdAyxGiTbghvYOmI1wVbtMm9TWHx/mUAFmq9FbFuosNpDZHD5ozzBIto2nKV3l5QqsUvtlXdpqNXWnkdYgcwnc59XJepOSTIQogSaN95V4YPLEQMV2/biZb4RI6JoBd9N3L9xoRrJMLKEOkoIU1JTlIkzXTMgZyZk0n5AUsVOiWGF9o1yvBbJLHKWuqWVVrNKICKQkaIiUEMjBCWvIhWV1qQ31BRBTqN6ABTEn6Mujt1NEmRJgQpLIy9OJEKrElwsSlOtVYfE2qGbEXJVGDGgQ1roAokAMgs2xvpyRUJ+tBM4PkXmObh8zOIXAaYI0wyqwSiCGRHz1giWurGnF4pV8EdZVuWZYFbIJVwssCAv+N5eCVucPTDlNgRgDgcCSqTatFkqxYsGYJWMPZ3jAnVbmSJBIWXL1mFUSRhBnmpac+SzlvQYvkw1gRlJzR5mIsXcJvzGJHK4/HhjNV42IjtfeZTYHtUgr2322v6F/lW5wb8C4g7xBZBn5lhvb2ChRDA/pBLWC0Eg8Pq1I/zQV2u1d30w9raLGVLT2j5hn0MehYZ40zqAbxHwc916SY7tGceDpYlbrcBHgm3DdrSh/uGzAscMiEgdEa23cGACqirUzMp37CSDqCF4lieaGv3seDnI7ADNq/Jdgah4zhlUX/BoYolUtqIaqobgJQmuIxqgdbLZhB7b23CrvmYO7x/vVtg9j2JwMpE5iH7LDOElwycLvjYhFRCOWV7QUtGRKBL1oH2RJ0deFqaNP1UxKDEzzTDi/IC8LGlayrWgumIhL70HI64pm7++UAhpmiIn0YiJGd8bKmsk5U9RQFeJaJTuBZS21/0bJhfkcmE5CmCIBSAJnDaQQaREP5xR587jy9nHlzaVgktx2J1KlG7dhXTKEDOcoTGlCYoQQiHMiEUkh8LnXZ95+fEGXlXwKfPfLmRenyItz5OUszCkyJ3c2iQ/GuhYev/ExNs/okilvM49X5VqUC3AlkBGX0Ar+MX/3LktBKEgUSnGgDmLuEAIEBV2v6BQgB6ZwZo4RBLJkSi6YQJJISIIWQ/mOlrxGz8LxpR6I0VBGUjYa3o0RTGT4d3jSAWtGCa7FiY13bqoX2d1XqdTORX0DokEBZzeKm/0L3yWKkefeTm/WmgHcW5+tgaTcAfAR/IcHHYClH5I9gLQ29RqMSpwPdjWrtpLaQh2uGB1EdoDZgGskkO1aw+OS+lhtI9LnReDI2DxVdqA4gCVD23weB7nzZsLoGsidVU2kSzq3zx2qaQ4efZJ0d0/vSwNfmu3N6pqqgGUVwKqasLnHd2/Cemc/V2t3E9d2fvQ+3M0lbdWMi8QlWl93bdh8HK3PhVXu0zpT1q73+L4IlhtPUastNGFUqsNEsHF6at9qbJyEGocXAxbcy80EZhWuslBwiWASYQqQojtqxODjqGvuqjw1D8KluLT89rFQsj/rfI7MDxMxRWIM1aZoSIBobmdWMXQOmEZMlaXawVTZHIjE1XpLVtKqpKhMMfnoBo9njCJEM16dZ149TOTLRH5IfPAy8eKUePmQmAPEEDxGcUoYHnzucYk+JkSliJJFyGabL0qdZ18/Hoeq1ZYXJAyERVylK7h9DCOqVnVjU5gbqxnLkonmgBijr6H1O1tteMtjN4I3cocj1IzEfLx3JMhHft1p7N6Z4PbOjaD1l723QXbgt2/pECRdq7slryMZb6XH7Q8tGCH1jvv4IDk2UiO1jRzq6LfY3oXcNgSmc/nvlFRaOzaq31yrW5+l1qnNpiNDdhSkOpZsoGUytn4ESCqo7Soe5lO4nYh3t33swX4O2uyyt9UM4td+jY0r6y7OHVbTcGzDqGFFj+3e6pam8uvg1SSu0rNj0Bwx2kdaw61xAX2e2hx1sGs2stYKETbHljpWdQykSZIbgtcg8m1EBBsyUNAZGhHPqhLEA22JQggVmGqbPHBciLgU4Cppo5RCKC46iAghRcKUkLKSQiKqUCgsSdzLXY1T6B7xTFP0qEArlOqsUagAthYUYS3K1z++sF4zVpTv+Z6XNchYmGLoklqIVcLFwUtSgApgj1fPijHaZQEyxnUtBPGA7DQlWlxYCEIMQiLx6jzx4cOMLRm7znzwcuLlOfHqYYbSninEGCjZ58W02tIqWDanv4x0iduZgwpfpmiVvX2Iq/2LgKiQrDCJMAdhAiYzkhqiAsVQMa5FeXtZocApKrMnReK6fAeD1y61Eceg473UMx59+qyX4/EjyBmupTgSm6fuuy23pKtz542QVDuB3W3xHq72ZFV2qaPGL8JAu+807lPVZcfT9yjwVtnw3fa/+81VWujg496QYrKphIeAryO3L4e/TtRsePzxjqdqOX73olWiCTKsm8EAaWKeteXOvBx72odCmlr43cfuFmm2PN1XbMPKdM+Lvm4U9zLEClpWt2GoVulEq91l47INZ1h6IirdAEuagwdsa7NO7bau6nwaEGratmH9bfbK/YiLbN6GIu6aLnPCbCaKYUGYJ1BdKVpYimBxcpBeM1IKlldKXplFMVvRckVC4uEUmOKJck4g7qQwLZmpLNhF4GrM0Z0JEkbIpRlX6xi6/W0poKW4g0MxPl4Kl+uVsmY+py+BqvEw9dRcQdw+lSpAmHAWwakHXFaDUGDx+KmcDY0GMfJ2zZh6jFY0I50m0nkizA6sQZSHSfjcq4lZzpyl8DAF5uRppMpVWbNRsmFFWa8r+bJQHq/o6usgVR97K6AWu9pY1TAtNEOp1jmJKfD6PDGdgs9/hnBZeUB5GeB1CiQg5MzlkwuyJhaBr18z3/hk9WBt83RbKUR3TPkM5b0GryNw7EmP9Bdpu+KW2B356bsANKqGOlHYrtjbORrxlJ1DltQKNvXg9qTAWLdsFws3zg8bGT+6hMvu7CAn3hmb1qZBIpLt2FPFaJe25zTKtbVsEDxu4KBNRTg8wqvYVHCg2/chn1cf3gEL393a2xZYc9CQ/bVWn7WTZ2SY2+MqudNHGdbC8bV0AUToziK71bSvfXdvW3tmPT1Xl4ZsU+F1G1V1/vA4aKOm0vB0U+aSl5VBiqJKZjTJrn4649SksT1E79o6BKTu5koNiWO05HCfjWttA7QmFYTgDgIyTeSc0RKJAiEqIjUPR0wVx12tZVrZn6kGKZsh68okkFKAlDATiioxCHKeXJ2qmajmoQM1i0Sz64bo7TDzzBFmIB6T7DFna3A1oiprzqxZEAtb5o0gG7hXxiKKcJoCL15MHrhgmZJdvWsKoRglQEZZMPI1uPYhxTq3LhUFEU5TxM4RW0+IOrOVs5FXZV2VXJQ5Tbi3okCAKN6JkxRSZcBGSVtEmGKE6JRpza4ynGLgxZx49XLyOViNxZS5wGSFCQiq2Jp5fHtFi3IV+GTNvMlGzs4QXqqkXPQ7WPLal+ZBNxLBjQPdOMNDmOeB2gyX1VrvEMhOT101stdENQJj269uuzjKTK1+q23fFtCd7jU86//Idhi6jvm2LU2NM7h/fTPUvxPOWv3u9wiPsHHUTeIxNkeEu4+okmVvRjVat8Hc182GF4Na7J7HzQZ6I9i+q4zq1XdwgndQZqO9A4gf22R7oNpD1lZla3u/rY+n9Pq3kAXpgKKqBNlUgAxxWhvAVUmr2bwGFWBLDN3UVza2rqmRhna2MR0bvq1ZGQ8Pc39gYbqttY5Z6yttDbR8iOJOHCFSmkcjDmxBAYluR8LVf2qCiWfCcHHZsJwJMRJCJKYJNXfvFsG9EcuKrQFbsoNddWIJ0duYYkBrPj9Fu1otFjjPAcsRsRpDlgvrIoQYicVBiqnmxOxMgKvBpxR5OHseyVKMy8WdQkTdLuSqPGU1yEsmxYDlUtWmPqdiMCeBKWLnieWSMfXkyuuqDoiqFYRdfRqTq/siYGrMwVNIRaCoaxkkiCcjrjbF6zW7U0sUzinxwdmDkUtRPrmunBYlZYj42ioKbx9X1mJcg/A2Fy7FY8iLCmuldLl8J3sbPnF8oymdHWV8gezwd3fPnWP3r5G75zoIdQ57I4sjYbjHxbbyLi9Cap17zt0XWetuGDjgXUsH48y7+vn0Qb/PX8fx4IZCw9cnizWCPtoRuT8H23N9PGPjHGS40rbx3tvobkFu37bbhu5Ao14jx1kYjVzHGTz83DEYQ2lJc7v7/NCljcfYgGvH/JgDWLNLlGbLsubXV8+poppRzV1NaFq2jPXmqZP0CFojzDaQw7rX4XhF+60Ma3Vg0vawNnwzPNtEO2r+CeZSYskZLgu2KlpVaJmFUh0cigrEjKqR18xqAeKMpEBJiRB81qzgKjBRovnOBNEgETy3XkyUEFCJSExIisQ5erBt9Jmap0DWmsLJhKxGKUp5mHg7T7x5nCircflkJeWCpIkpGBYFsQlCzSOZDSseeDylAAkPXl6LB15XySsvhTQHshkrhVWLZxWZI+viyZWDZdLV8zKm4KmvPE66xooBYQpMIXm6qWnidBJSeUm4RmxZuF4eeauZkIVQIldzZ44YAx+8nD37u8Dbx4Xrda1gGXg9p263mq8J3hSiAqauXlXhkyy8WTNX4KLKtUhfZy3xtz6Zi/SbK+89eKnVl2aj0P0F37QRcpeANEoxxuY0IHDGcPTi2uoCX4ytbGq/elw2AtLS9TYVTPeCNOnGdY+laoljt7ZJrdszXAiIq0IaLMImrR1dzTeJTqolqXpPUbNCc3QhaDLIOGLS667I0AmTimzqv6N0U1Pl7Aj5jqhX7nuEdzlu7LFd7tJ0JeYCgwhSL2yzUmPFjshs1ncaGMG1O3nceeYwKH0d7ITVcbCrFtId1lu2Cj9f2jfbj3VT96vtW9B+B4a0YZgT+iZRdQBoqj8d7FLWt9OpYpgPl1Ywq+unFB08Ebfeq1LtXPT62nPMjuuFDqqIdABr68g9SD0GSZEet9fWWFvHNzZadTuN4vYYMCRWDWihZ8lg9ToyBeapEt7oTgiqkF1way76lhcsTN4qC+4IcZ4QnRFZkCSEyZ0kYv2E6NusqHlM1Fwlt7IWwkl4SIHzHPjkkytWMvkCOapnX6/qRqSOcwEjulAY3R4YzYOAg7hqTs3IWliayQljDcaUjVIgqpLVcxeW64qaezzmJXv+wazu0BSjO3ZEz/Y+x0SyhLw2ZIroMiEx8MpWyCAlkEMgRM+r+OGLE9OckBB4e5l483ilrEqyCCZY8VRapxSwUwSNHohtcDVYNLNqZEVqXJ11t3tL0qXBz1Lea/Dalf073V+QXgbCKbvLD7r8Dlzb+WOqoZshl+HBh+s6kPY625PvwWk9Js2zZ3T9lt3ltxaISm2OSV6H28Lu+ptWHmqTbnrb9Xkc56Hrx7KzufVKbRvcoVjt3zYyMkzGOFsbR9KvG54nQ43H0T1m1mhxam2Ox+HoMWwD0LUMG9LAsF3zDvWr3XzbO2ZsPTmA8U5a2RifdnDn3GHb+U1d2D6ba3xXATaJy9wwX27w0GhS3da6dlFrRm1zv68+f1hsJnsv1Zsy5g5jeyPGOeogWsGx5ZdE2t5n6oeCIKfkdrIpQpXoVLXuTdc4roKpVHYsEkIgTRE5TwQrzgRFj2FC6+Mq8Q8GBJd0inm6qpjq3nLBWNZMLK7zMwsO3E1aFpeorBja1HEaOnh5cuDG/PjYl0rpBWXFpb22PY7huRp1zb7XWi6s10xeVr9GImGG7gVqzkSkGLDTDAIlRaYgnHRxd/8MltwVP8XABw8T82kipsBp8mTI6+JhAi3IuEliOiV3+rmsVVp0pq2Y9dR6WtdeC9PxvMff6eAl28s3+n2N1OQe+AyklT2xuL1mX/xhoXHttlel3GvfRndu23FsG9WJYvAo7ne6qm3DgCa3bFFUm1TW8hbarmVHNd0wCoMK56ZJT/647ZfTGTkcP5xHujPE1mqpYyl3AHHnGL8R0Pu8ST0lO5vbLi9iG8ND81oKm/67S3pHyN+P5yAfHVq9H66RMWjNl6Ohqa3nYaz6+hznVHBqV6BngDdFFFeTqef5cxf55qzhgJBLqXE91lU52AaKLXtIy3ChtuVQbf1pkqHBlpf64BnaGQQ5zOkw7kI3ayHR+yRBIAgWKmCViGAES0xBCZNLoi09U0gTpFSdKwTMvQ8x35+sOyRopuQVNQEiYRLiNDEl32/KVZLKel1gSUhInM+enFZckKK0yF01RAunBNOUfO+xx0dkXZGCg1jTi1iVaIv5BpgRpPggRJRJjClEVlFUjFJViEUL2QpvoxBVmbTwYG1jGleX5rKSl5Xrx4/kxytCIE4nJEZUC2sA0cBkeH/nhE5C0JnpxQPn+YJcM9NSOCXPPxhFeDjNzOeJOEUephOnOLPkzJozrItn78/KaZqwEClzYhEhXwsLhVKZzzb2jfFL4ja2KQbks2kN32/wko3qcCQdW9rWjaTDlip0zIkouztG0Ki/ZbvfXbkrI1cv23HCInXTwIEr35XdU26e1drTpa4Kyvv2ADZ6n3msS1OAju4q7Rk72mHtWKXgfRjbf52PHu6p147EvT1iwBHZPWwD1T27IL151lRyDEBuBhJAa0zXuMgH4a09auuCbX1tANqB1P/dAqPBrCAS9xA0YuLo2DKM6A0kj6g/jM04FCrSt4nYKrOuZmv/9rbLAAoE1MpuJJtWEMSliab+q2Dle3RZt3WVnNFcKFp2oGS0rEdec2uJmcc2WaCrED0WReg5w2iReD3fBhD3JsE7o9ffgIHrC0H6NiYhRGKMWIyUVSnXjF6vUArx5C7b54fQIwY8iXH7XYh1mZoamovnHVRnTFTdRpiLEUoiRCFVvX/bYfrjdaXYBEQeLjPzudrAJk8JVbJvVjkrTFNgmgLf9SqhJ/FsF9+4OniqEcRVjsEjFEiVLojjJ7PCgwZezpFSDGpwsw+1Z56PcyTMCZkdoM0UCr67dS7oklmvC/njT8AEm1bkQbHkHoOLQYiZmCLnVxPz+USaA8mM1yFxWlaWaybU3JCiSswrXBVbI1ESpxhIcUJfTORrRK+Z8rhyrsn2CfDJnFjeel0PWZjE8yamrNg1g8EcA69OiVMb9M9Q3mvwGld/xw+5c1rG10Y68d5z0098l/3vIxC053fC1djY3V23k3TIPsiOeLWKWrsdIQ+dg54jcGiHE45KTCqh3kukT7R97Nsgldy77951N72947lxD+zunr93w71rb/qxnT1WczPXNxUMHnBsoLd9l57pYiTzm+fcKI1tHdxU2F7PbTzX9rsH6B7PmTMzTX23Z8i2/u48CxsRVPdI3PZ7GoDLBmAa+tYVkU2rMO7r0ps02oGHbz0F1FHPcOhtt1M226/QAp57FogYKFZ8M8TrCrkgcSKkwDSlLkmSPTmv4X2NiS2PY/W+01Ltv+YS1rpmAuZOCVGw6lq+rIXH1ZPagpKLMVtiypGTRpbsXnxlKWCBEIRp8sDm4l7zaFjcVtdVn7I5U0XPTUgUSMakgpbAKQlzMCxYz4MJ7qCUUqyfKq0UaDbsNokNqK0YugqCgx1TJBdFJBBico/D6GpKCYF5mmpYQvRg5yVja82okcW9v+Lm+RkinuhXQa8+1tMUCXMgA0sIlBShBJcizUjL5iQ0h8jrc+I8RapB8/dc3m/w4vYllsrNHy56B7UcAGA04D9B7ccXU7DBbm+9Pe3Cg1P+oZ6BaB0e927CPACjHGnK1peWSGjwPh+81fYc8CDz3PT1XrvuH68t2ejtTV09O4eMEu/YrcGhZayzt5kb1Gn9kZY7cKcu3QOqdUmQG3Dd22fkjr1mQ/NO7wfJg8pRN25jFLJu8Hg41qUuuZ1Dq0xLT8fU57GtsjbHLg158owabKy+AaPWbTpK0ZotomoUjSqt1ecNDijW6gCa08cIWsMqvwO125xt6nz6aI2jYLIxkv5vqOAVCdG3ATZTyppZLgus2R0q5sRpmiuwKctloWTPladiBGI3+YkquZR6HijqgcaXFYoSU6TEiF4K12vmshQ+Mep+VwW5rqQ1MU2Rl+fJM7JnJS+FYsH3E5sjcwgUMbJF1ilQSsCKu43nxuQQiCkSpojM1ZBmATHlYVKWyVMr5eISCzhonJIT+9McIJnnVjRfex5S4KAI7rmY10ekBMLDTDhNXC1jGghSEJnAhPmkxBSYT4kpTZzOxnK9sj4urCjL20zS4IxAdKk4BE/3lRKQhRyKO7ScJs4Pc0tPQjpnXpqnn8qqvHnMBJRSHLy+8PLEw1wlyM9Q3nvwAupbc4eoPcHhP1nFp5wbM73v7xng4+aB98j4dtyg77jRajkSXme3bOjnRorbtzCIQXqU0oYntiwko3fiXcnk5u53lZsW70t7YL/abojeOyLCnnik7QBodJHf2bo+ta3jc4/QMvZBtmdJA5CDdNEQp1f5dH82T0fZzfmGEvujYzOO5/aOGG3DSUOqW7mqkM299FrOv9KNXfs29FRlzT1f25rfazfaLuLvnrMqFXBkDtp9GyvoefgGxbUBqp7tPQo6B/dYrFuihBDABCWTzVi0IDEg84S27BgxEJIRWbHgKbKaN6VkRdcrawhcY2Q1WErkasLjqhSJ/n4VMFXiYlyyEkNEVcnrSi7uKXiaEtOpvo8WiPOEqXQbmkrNlx8i8SGSHibinDAKoh409mIGWz0bfMk1Jgr/fUqeoPcheAb/Fl5R1HyLlCkxn2c+jomFQs6ZtC5o8MDrLIFsirKyvP2ES5k4TYlzSjy8emA+T5xfzpyScU3GY8h8cslYEFQC0zyRUkQRci7YmilrZl0WVnyXZIuRU4jYaWJOEYkn0iyoGG+vK3OE5eoOHx++8ByM8h0NXgNo2fitv5cbsdmDhQxYUDn2kYTLgbDJ9su58oO0J3tCsm0qaGMTt+u7dHZLLH3riO2eMbj6eO1W9WZ16P097KNSmfiBZmxc+0a69wRmLwXdL0+qF+tz+rz0odzadevaf9u+Y0/vtGD7NIcBp5Y0++Qxpceun8OD9wT2DpgeEXE3xPdAcfvq42+7K3prjl3rqsHbsmXG2PB79EfcORtSfQvY8M6q+9cgTN0sd7NhNR05qnahbN9tHIgmQVbGooc4DLf2Zw5D1ngDqRCqpWBWyKsHD8cYiKeJNLndR2s2fKsZ3hugxRjdPb6u3UkACyArmoGsvnVNE83Apb2wcVeSvd5qgqJ5aq6rYqkNprdXqa7tq3qwi4KKZ/MI4qyABleFBonIHAnRM2a4jcsDkyO4Xamm9ipq7naOEGpbG7skzcMlRG+guIqVlJBUHLBTwKMUPJZvrdKchkASIagya8Gufl04uX0sSSCliCRfay01VAix2mcdHG1dkbJQLvgeYOCeGKoE841EpxBdahPh1Ysz15DJS2FKgSkKbcuq32t5v8Hr8Hrbnb+bUmcAGLGD/aJ+q7qsDV4G14f+Lg+EYvf8PbmR4dm3RaqGaQ9yWw3bv92suXMZ356x0RgZ7tyOOeHQ3XVNgeOKjA0A6UcPnW7Pv7FhbZIO/enDmI3ANhIqAzoD0Eb7hoLfPOd+kTttc2eWUEGM/oRBcVXv6X0/gnA3VknjBjaQa1c2xqATbO/X1pJ7/WI/F30dVm/AOl9mSkC2jU66h8b42/bAVRtjxuaMUa81pAJXAzbp66rjewV8915kN2fjwhr5i631dQw7w7axVbtOt/dRzT0KBURc7evZi8xzMZbi6sJlRVQJITCdPb9fnKLbnYp/RHw7kZACU4o1JZOrIKcQKngJGlytZqGuBQEJNd4oRUoRcoB49c0cKdZ9VAyjFKkbPvo9qXohZi2eNb36YGY8rVSIVdUW3XszEpBqtwr4TsaaFc1O8OcQPP5LqtejekZ6qbtgFytMTEhIyGRISu6YIzWLRkrESQnF7VCWAiWAroV1MRaFEhKzrEw1jb8sggRzld8sBBWSuMRasm/cKdCz3wRRLGdYV8LqQeNL8b3SUkqeHxFBNCAnz+iR5onXLx9IYeHCQqy739w4MH2L5b0Gr/YSdRoFO6ZwBIKNgdwIrPYzY43tVZSbOqBOzK4Vo6NFq6a9uBt7a4dzDRga8TnyIGHsx5GwAkewGsF2tIhswFSDnq1xuLsGs4EA33R556U7L7/jfa190q+920c5EHdaStNWv9A3uaxlnIsWFjSsko5H+2K7RdQBjrCb7XHbinGNHYWSfUe2FrXMJ8dXtvex27acK7Wu5K0rd5C6do/sKTIGhxDx9a1sWdtb29uGmP354/zjBL0LJYdu7dp9/HEYV6vPD/0ZY037i/0aw8RjmKysLDnw9nolYqQID+fINHt81pIX1jX3/clOc3JHCPHdj/15oYW6OZMvvi0KKSKvTwTUbTkpcUozJ1MeihLyGy5mLKruqResZnOPYEoSYUqBDx8i8yRE2cIPMN8uJcW6NcqU3CkE821SxO2QS87ky0J+XMjX7EHPBMSq57DhUlBI5Enc9V0iFmckCTFNxFcXlm9kyuJAExDPGjILLx8mSBGNgcd8QUtmuSpvViG9EOaz8JCE07IiJXG9fMI1zmgADYUohYxRcubNx0KaJl9TmlkvbyjLBcqVcl0oS+D6mAgS20IjTA+YwunVmenlzGmafNfmnLnm1e1dz1uiNM6oSSjja7GR9APzSM8P12s5AsB4hD33+U0Q+D3TKjfHtqv2ZEFgcGWvx5p9xMarALQ3zpBudB/VmntZcKu3k5IqLew6JXvg3S6kD3g/tUOdfe036rnaorbv1u6xIhzZgq2mPUA1rn9kEpr6czd3Xae5zebdSMA+Bq1B99qyzeF4xmR/3eHsndkdVIJSARGpzZUahFrvrpPuktOQ0qnmJtQOZi5OHdf9+Hd0tNg2xmxjGHb7aW07KdvW0F6dbT0Y3jWrDd7tSj3M3pY+aj+O7Xn7/NMOpCG4M0Db+BCkJt/1UN7Qx8+IUdzmFT37QymljpPHZqkaZtUhZPDcE6nOCDESLRCLkudIyjUprprHoAUISViLv1/JMhGpgFPHsTJUoqH/DmI1G38NYF5dmjY1ypo90HhdKatvTBmaqrWqGlMKyGkizpM7sdQEuyKRmCKIVI/K7GMSfBxS8j3Msrg9cZ5jtX0FlBrTtprnXywrC1feqoMdk6CTj1lRWB8vyGWt86Ksjyu2FFgVyVXWWg2zFXdKDESNrJfFGYWTS8MpeX+ujxld8R0OPkN5r8Gr0xvYKJrd4kzjvnfEartpT2+aIV5uyc7N88eqhrKzbRy5bXtXvdt9jYTusUFu7j/W1Mn5sbMyfL9pMD2Rabu3Y8oTz2m3Sf/bgHO48oiJsr/3tsZju2wHlHeY++FZe2miS8+y1QPbNiebALw5HthAfDfZ+/jETUrb9p5qbRhG7cAY7fq8w/TtnhuoMdt/2n/DMbXd4ur/Wl901vvZr5Jx1qQ/23c9rkDSEyRv62Jc7zf+hi507PojN04dg31PbiG2HfN5c0CZz7Fvz2GIOylYAXNVlgUf9xBCVRcKWQt5NdbVbVAmxQHUBCUgEWJyYuqvlHmyX4MYhFJjuibwlEwVM4jCY52HiBJJm5t/GOxmJYyjitSg8bZJZptDzQWt8XeaFaJLghYqsMbguzI/zKQ5OVjVfmKGhIhZ3ZV4LRXQPRVaCg5yhvdpnidKNFaNSHB2QvPq0ipgxbiuUGKCORIJHlJQ4PLWt0wBt2WVx4zkQlg9Lqxtl601z2IMbmNclxVbErFMSIquao2eqFfV1ZmfpbzX4DUmPNq5Jg90Y+PS+6F6j+0ILbsX78hfb9eMFNl/7nj92xs6oaj32e4sjfw3j8NeL80uth3vxNK2e22sm/HPbd6+sf+tn6Mq7F4vdsb2gfLck0T6tcLNEI7zMJDF4ex9WGoqLrfP1RS5Ij3jdb2ot20P2k/M4015FyxuANqYJSf0zWLIO+69lfgbA9JU1jcmaxs2ALTSpSyAHpylw6fbwvy+vkR3Th2DHaoly+xq19qC1pAj6B7U3KNr0HhlI9cdDuvQN6bCuJXIxmEb+VDnpSIxnjnPiVIyOa+s1wW7Zt8vK3qC2yAeWxWrejBTuFxX3r5ZeXy7el3JCX4IAS2eBDekyDRPhODbmWBAlQROKRGnwmyKSmaK1UaDMYl7N2oxz4EYIyFFCBNx8lRS6xXfWsUUckaX7MHHdZPGHo9XMqYFEd/mRartTU4TafZ4NqZAenH2pMHRpSnMBTnLUC4eLFwuCyGv1YlFMJtwt0vhlCbiKfIQIw8xMOWFc16wdUHXlSBGMrBsLGVlWRMhRy5Fua7GR59Esk4eViGFkDNBC0mNZL4tTVFIZkzBmKJyDhnWglS7YQyFpIWJggRhyYXH5Ts4q/xIoO5JCuMGeUcC04nKKGXUg51hvVPpRrIZqNLgmi0eCNm4Xw8u3ervXGp/yQdCMDbROis7ANEBjhrgdpWed6BXs/PyGkC3I8mY7bCb8GnjucetA+xJG/dxoG5dVI4Q5Rz61q/eNnXFkvQHj9zIJhlYHTMHNatpnw7g0ft3OHRzFHpW3Xv405tn42bBdSwqBZGwu/0ob7XRGZ87qhrvMRjszjo4tS1NtGaQbx+ae3wHuQ1h7fDpz+vAtYFXlxQqJjbpstXrfNOmphxfDaGth23Nj+ZTX8rGbqBHm1uzITagbe2tvz2BcOF6XXzrElWiwHRybl4CiLpbdxbhcVl4XFYe10ycIqL4Osme5iIE78ecku8KLNWtvRhWqHaw6OMbtSY1BrXim1WKM5sP80w4z8icyFYTVSugBc2luoIXbC0VuKw6N2/qX3BgjRMwJcKUmB4mV5emqs48VSeNEFAJlOKZUtY1U65XyvWCXRdM25YowjWuSIyYBKIGQhJsCpxenQhrIK2CvF3JqkjdaaAYXHPhrSp5iVyLcM3w8dUzwDuFKFCqR6EGIr5eijmYzAqTGieB10vh5TUzXdY6l8Vjnqfq+fidvJNyE0EE9obsSi06MTY56NThHqEd1SN7snIEiHs8/QgOfv3ITbZmyaG6obmfXgYdonQwqlLHnQpG9/ku8sjWz/7w3n4vu2xMcuznvt2776OHXv/3zkjdqVCOHMONLYk+pq3uAznc/tr9ZzRHi700eee6XWWtXXVtDLY1P71PSdW74Q2m+XZ6e2+ZD/r1FXAq6jQ1YUOhtqPxlsfwqFKkf9+Ga78Cu4axO8JswN+6KSaE6irfMtDvLKgyMIUc+KNKlN0xRWpi2oFBkaaa3TM0tqvLtuGoBF5rkPW6FLe1FN88kyDV7gRopkggB+G6riyleHaHJF361FLjIRVyMda1YPgWIKUClxmumkvR7WxqsHrmigbisTY8xeiJgE8JzSAF3+izet+hiloN2tKWjmuzPlrjWJuaMEUHrCkSJgfQtheZqyVr7Jm4+ruUQllXdPXsIy6pQymBNavb44JTtFB3dZ7OCaIh0QhlpiwrCHUPMU+bdVmVi8E1C9cSeKtK3UCtrz8MggWC+a4ABSMaTGYk4FQMWRW5Zh4ua31v3LkmTeJgP32Hu8pLXQ7NpdyZn5F4MpzhcPzAEXMg3HcfudU1fmt8Z3cOEdmpAaFj7buBquHtEfXGemgv+j2YFdq4bOf8ur61yV1cGInKvpW7Ww6gsqkd3ym87I/dwc59T2RTO8HmNSiHq4Quee18M97VFtk/8TggzXHheL8Mn/Y7VGI7rrbAoK5tjMIxoO1uw2pb2kaRVfzZ2bd0k7hUSz1WDuLV6JW4qRy3Ge2DSXMq6IyHQahb8zSeWA+S1riijiMJviXMJtFvY9xc8HcZ+8exDezGQU3d4ERhXVeWZeV6WWG1RkdJGlyVZcpyvXgQbYysufiWPafEw+sTZTU0q++b5TvCgSjXJZOLZ71wQd78MwlxSkQixRLlcUVXdUmh1MDi4ptBxhSI54RdfVsSvRbsusDSwKv0QLu2BZInHQ4+TjWfIzTwilhNUizVMUO0hjCIEGrSZBMHrzWvNWA6ozhmXqukmQjEmKoThW/18iIGVFKdojM5GZo9uW++FlbLLLnw8UV5LIFFjTUKIWlViBjFo/S7jdczuhTvTxHEfIPU9bKwACkJr3MkTRAm4/wwE2dD7zGo30J5r8FLpe5VpSOJlk7sej64O+XTQET6NQO33hHFuv2p/hyef6x9jKUazspANoe9rfo1MtRSv+99SEbrw54ENwLUzoZW066CEapHaWJr+1Gy2XHpTQpp3lHbRZ0gjqqjsfMNRNsQjrf6g+ueRF1tum9HGNRMrS9mQrjLyFWGohHxPm+jlM1OMtgO1TCG0W7ptKffOaoAO7QeJxuj73Rt7hG2O6e4iqp5FNb8TVq2zSOVBlq3Xoempd/fJJVNcttYq+6SL1IdU0JXgRpWs8c2yWfoRJcI991r+XnH9b0FFxwZxhEAFYiDytw8m7mZz30DYVVKKU6c10zJBSmCKIQpEueJdApIMvK61msVycqMbyz5Yopk8UBhnUIfe9VCVHcuiCl6PJ06wESLxEmQ6K7715BYr5mF1aNxq57s8rhic2KOHiwseUFyJuQCOfueZOrBAr51SwWk0FSFEyJSEwuDpOQehRooq+8ZphghKCFNyARh9hRNEiIxnQjTCTmdSAolCGqBUCLx5QPzaWY6TRTc21EwysXjrCQI8eUDdp5Yl5X1zSPLm0cyHphcRFlVuRalWHQ1YXDPyuxeMD0x+BiOYZULsaJ8/XFlyc2Ge+LhHHgQ39H5FAN66oEvv6fyXoMXu9dk/D5mKq/kq6tMxvvHTAaDnWdgnY+EExi1d72W7b4BLhsRtx2ZrKq4TlKcK2Nf6d5eR+3Pp8Ft/TtIBMe2t1HaEZjB4WEHCrIRvPEJx+9tFDrQ3pw8iENtPIbrhBE8ZDD0t/bu23EDeHefWEGjAwhYdbAZM4t420dgvl8237zGtEhv47ixpTMndYR3B2stsg1EJdO3M9slLnbSl5+6PcdT11DPtccPTAdt3VZQ9eZt1+6k/zZu7I8dpf7Nnno/Rde2SjdwG2vpI1NtebkUd3svvpsypWpWLHhQcwiERF87mhVdtb57ASnquQBxJsR3BzaKBmKxuo1KQMSw7IHJVHAPQSAoMflGkCQPKPaAb/PdhR8TEoPHkWkhWHEtUGVCtAZjEwSpuw/TACxFt5ELqAkSfKdnMUGzS9WrFoIocfbgYVLq9C7ESJxn4vmMSHTHFAKUyPTizHSemU+Tu8avGcsZ0wzigdZxStiUyBKwRSly9ewgElxlKp7VQ0udpYpQxZrGy991F8TcZZ9h/q51W5opwIuzIMFzRE7iWUY+416U7zd47V+cdswqWNj9axq9rr93ZppGiI4kzLYXf/zW6x1wY/fC71pQX+gjOO0bv5XQiMX+pAzfj6A7cv6t/niHHI9Eu69AGY9JxRrZtekGtEcppdUtA3wPBrMjAduDN9vkHFWDw99wGNve7/o5Bt86DtpBBti3F5rTx/18KBukSe9fO2O17Zt7/d6RYZz5TWTZALDLHcPFm+PMJpVYQ5L60bZJY+vjsW+dSxtsZIPIJJU7PjIZI2g1QGzHt3EYlYFbP9uc7izGcmfDwQ6cw1zptpZCC1cW3/JjLYVSMloyWrcZwTxbBupb2sSabknUYF1ZHzMSCiEV8iM9JZbUtEdtF98UPeBZksc+qWoFDcWS5070nIRKiEZIiRIKGU/tJG+u7p+hxvQQCVoIwT0hfTcv9S1B6iaZ6SEisgVthyn0mMftqGAmlFVZc+FxKQQJzC+EU4gwqa9Vc5Xl9OLBpaii6BQoBE7qoHY6T5zOEybG8nghX67YY6nMSyCmCVJglYgtSg4RDcWzmSRjKuZqw1UpmjHCthsOEIJ6nFj04G8tWrVg/sk1PIBL4fwWRIyH2XubDExv37dvpbzX4NVfeAGr26+D5wLrgNSIxaaKp51qH72tsr+i4YYn/pQBr+pFp8X7F31Eznt4NVKCpwjuqH4pbBJmlzNle9YOpHatH2o+ssbtcMefO0R974a478BI0Q5P3Y4NA7Aj9xsBbenpRmfCHRjIE1AzTGyTPftYvmPq7tcm/bhVpx9XeN08tfIBLUbuKNWN7d7YI4GaZLW5T1eO3aOtaA4QXZU4jkHHlg02DNnu3GjIIP01NSEeajDUY+OF41DY9ufo7B7uHq9HKjPRRl/vMgeDdqDrnt2uoxLdbdwczqwoEoqDhQmaYXl7xSxjJRGKE1uycilOONWUfMluo1LDQsLKTEieuom0LZcE5KLouqBFyZxQ9awduRTUPBaMECBGLCauS/bfCJLOxOiBzJaNLMrVjLcKZxEepsT5YUZmD6K24o4tbsfCVYW6jXUxY83uYanF1YESIvHkY6ylUJYrMSXiwwtSDIRzxMSdLoRISp4LUlBkSogWrktwT78CtoDkjBYjRk/uC8IUIy8ETpeV+Vr4nY9XlrVubGrQlMNmQpx9A9CQIhfNeMCXr8EkEOv85xJYirBU70oJvkHnZynvN3i1smNz96qNztDXv3c93drtw4t6dK/v13wadkEHsDtM8VYG4O3SRhcDt9q2rh1r21R/T+dQPLatsd5H2W2rs9e1A7WN4G5xuIOd7GALOw7UETC7nWgX2LYPKdjfO/xo+Fg92brX5TuNv6PccCgHLuEO3e7jYbtb9rKhHK4dj477M8G2tKpgSHM66oHIdTF2VR6barCpB6ECXXWXH3022v2tRb4cB3CoTgItQLt5j/U6mzT47qEa+rkfr03WHRWiR6vuYUBa86pKLYRAJJBUKAHfUyp6UHJHTFGPMVoNWwsU3y4kmK/NtulmNN/bzPKKrQE137o11bSHwcyT1ahvxBisZcUYQhDEbZUhDM5BdUFZzdwfmgQSVrJ4RvjVnMiq1GDnGKqNyLqmwCXBSPU92RgXBLNAbV5nYkrNar9eV8rV5y62HI0irgJV8fRX6i77wapsJwmrbvG6ZGJwh4uI8DBPTCLolJAE8ZSIl8wl+8TlqjLt/0ngxXnyrPNTIkbf58yyO7VMIq4alC0biklAYvWivKc2+BbKew1e7h3Ufw3/wvgKtd9+fnh9BoeAvWyyH9SjPWpXhY3XVW8lmjS3kbk7t76jYwdRo7VCxoOjw8YRwAbkad2RY6/2rdmCgUccGFre/wxPGUBLhjvGW+QwdHce3XtwtDvVdT9S+dqPpwF7U8fJ4fix9ffKLSluUt54RX+6DHXKNg5ix2fZfkxafwemqmc5b84Kg15wA7EGTqMXou6uY7jO14xUk2YdgTETBHQAG5/ZnX2G9dZbsHtf7sCRNClu6+fGDd0Wd62vIFod8Fq6JiRCqdugRCFOAVvNHSuSZ8pAlJILurpLuhQhiccRtTFyTNRqD3N/UAkJi3gIQnUCoXjGiNht1er5CsXcRhTca88JsWe6D/7DxyNELBglrqxBWMVdwrMYKp7XsCcECK4pCghIIITkkr0YqrnGQrodDKx6KDp4revK9bJwebtQ3m57ek3qYxWkJnSu3qliipRSM4gkirkUta4rKbgqMwXh1cOEzsnBcIbzUjhfVh5XxQSuS+nrW8RzPX746sTpNDOdEmmC5XIlX2FalTm0xD3GLM3xKRJjIsY0MDa/t/J+gxfQXiNtB8wj7seXpamgBsrCyDtvtY2oxM4UZIe/x1bsz3xzk3IDmnIgB/WCWzLsdzUm1AYQaUb00SFC6y1bSPJGzI41f1Pg2qp54mInFnvIfle9DZsMe0e7tta1cElgcD5olY3A0q4LQ3PvQ96GLrID5a0oMlhwNovdlnp2fGK/35qMbEM/DpKvbTYsZctZuMV1bVLA6G14s87M2FLcD70TVymG0NSGNVFsJSy3Zat7jFm8X/YMznbnOA+1bdLA8sAAmdueQnD1oqGIeDCxmBGDIFPk5avEVVbKqjCJ74sVgZJZkGrjFOaTOzWEIFgUkOiB07qitY2xSl1WA691zZScKQpxSkjyjyfvddtNMHX37iBoFM7zTDpNxNMMc3KgKwYhIWcjRN+eJKSIivAmKydxZxKTUMMKIuIoXD0hPZ+ghOAAEpzgp9NEmCPZtGYQeeSr//0bXD9eiCJ87nMPfCBnpimSxFMTp+jSoIgiFISafT4EzIxivr9WSoFpCpweXjj/VPz9fVmUD14UTAIvzxfPiFE9YwVPQfX69ZlzdQz54PXE4yeJ/PbC/BYi2aVAA6IxocRVSSEyTxPpaMT+Fst7DV5eDsQoDFwi0gNWR8eHnZv6TQ3Ou484d7x+/74eyeIAR7Jd3jnRRlzHe9p19+ayP6c9XTux3xcbiEY4PLu20RuwcfvSGrkjt/cePz7mLsBstL9Kw7JB5Qan3oau6m4gUYGr7bt17Fu3IRmu3qkAc+BFdk3ZVzES35rV4Nip4/dBepKBkenPbOMr48jv45v2nvvbSmlzss1jJfmN8g/qwRbzNQYjDxf2NdfPHCSwbYPOUD+u7jEGO5o1IN7W6g0A3hmtEcgbQNZQ1gG4NxBvN40A1lzFgwRCT+o72u2EIJGQJlBYQyGbkouxZCWpe/5lnImTIKRT8q1FQvAUS2KYNQ/WmvOvSktaJV5TV+1qV6k6U1GqA4KYJ7hNUbAU0CkyzYlUt2mxFNGc3TlEImkWSIpFJYaAhchSPH1UiK46E9MKFoqJ515s8xdqfkJJgpBq0PI2A2bGYoUVRU1YcyYvq6s7Q6mZL6LHes2REjw7fkvUZSKkFIkSSVNkPkXSDFqMshpkT+NEgu96cQITLovv5dXmPiC8PCcezonTeWYugZNmVi3Mea3qWqWYcK0qzJJX1stCNNml9Pu9lPcbvA4cd0OC8eXacgZaJ0jt3rsAVusQ2wJObXfBUHf9IjvO9L6MsWvT05ftpB8biPtNPQ31agNH69hObtgtEKF6tzwtCe3GlG08kBvQGomQVU6+O5DY8OgwQFjl5MdBeEqIG8diD/ibm8o477s+vEMyfOqZe+ZDdlcJ+/b0sRwFnf7ddoT7eL7hU1f1jfYsBseN46fePdq0Nuak9quuxRZy0IZjkyil/i83MWq19t0oyeFo+zsmBTgM4G7MWlMDdpMWq7vrEzwbRIs768ycgAVEoksl0dVWaoauvjNviVKDgenAnE4e7BtjqC7oLjUHEUrd6XWbB6P0sRc0gEmV1NV8vzBzm00MDODliXLTlEjz5O7uPUOHkNJENENCoZiPWC6BVXBbUPQ+9ez3eIY0fy2sb1iZxEBqTsMotZ8+giqKxZpBRNU9MkUpdef1QMQkMVIyVzO7anFKEyFEd1+fJ9LsW6uYFmytW6yEyOuHGSFynTPXde0TKGY8nCIPc+Q8R6JCWibWU2a6BmJxTtV38fbA+pIz63Vxp6fvaPACNucIf6mj3O6NdeemTuxvOfAbEjVID+3s4eVsd98BhQ2I9hfbnWueorUjOb114fgUO44c672VS+49550NYo9x/f4moewSAVonFL2fckhcNQLxEwNoNPvm/Rn71NII952rb1SP/cEyoJYMddAzcfiVYfh+aHqVRBux3whnXxkdnLQn4tW6R9d2rvW1jdqmPvQAVN9ZuG1epdvIjGpQCcN4bv+6IX0Yg83r41PLU1PWjw3MxZgEeqeTp8ZgISQiSRIWJkwSxoSZeJqlsjqFV+V6LZSsSHSJYk6udgwEZHJvQgkBs+Aqt+iEVK6r5zHUgtnUXdURgWjd6cGoGSyuKwF3iEinwDQJURIpClgiSt2C0nJ38oDqho/HZl0wSjHKojxePSD5HIUpxfp+qG+N0qYsVQ++ENwVXXxHYotCNiXFwDwFXr+c0MkIBucAMXggstW8h5YjtkamomSKe00uRpCJGBPTae4bQwYESsCy72K9XNfuPPPidGKaYCmFx+u1mWWxokwJohi6rL5rdRCfiyiE7GEEWc2zn6hn0Ze3F7cvhuPK+dbKew1ezS1YpMWH3L5IN3BU36UtQ4ZT3D0kyI5Q9IC8Xc0Omht3udk0Ro57LxQc+dcRevZA2upsoVIbpwq7rTeG6rt6U3DJx7a+bwAycuz1uTfUZ6xXdhJFl8Bkf60TJ3YgIFTuWcYesxunbQSsd7znAhzcyrX1S9rcNUy5A0hNPXQ83J4hMjRbtvmrEpu1ftW6d6CBq2Wb5LSpXX2VHJ17ZOsW2+g3yWiTakYw33w1tP84BiL3Oa1c925X2qFNYac23ADZuf09WI1eh/T6b+1do/OG2fDu2LjO6tjaOE/096Ru3dxBrK0Jvy3hAQmBvFr1mKvbcKAkUabggOJrqyqig+AbTlYVKR5HFULdGDJ540rOrIuRqp1NJKE22mdqPGZL0VXnVtWYxN3CY4ysSx3Ha8ZicfucWM2LWL0ILTKjrEW56Mq6Sk1fJ9hM3/LabVTVjTxJdwxxL46aALqmpnFnEXewsOCpqoIWwGpyYU/NFCWypsSpCCa+K3JZzDcViL59CcHTNGkpEGLNUlJc4iT18IAUBUmucs1FnREwKmgLYoEpRmRywCrpDVo9O0MQ5jiTYoKUSCnUDTo/W3mvwQvuSAvsCfQdPnt/r2zXd4Jqdrz07oOOUVz7p9886Ubgae/zrf3qqdJYs43IOFD5j+600uD2XvufqnZozzZuxwqa6u9w/0YT2d12bxi7d5v09E97em99XtqDupRTPwOZh2G2b+d6AxXnUZ7g9CoIbWrQDYgafB7Bt38fOJWWOWXs+3hdZ0RueJjGym4njmDVcWVw1OigP2geRKS2aWQqhBZI3aXegwpv7NsGahsYHXrDZmUbKjk4Q0mvQ3YMR2/X+NTKUMgwyaZGXotnrqCQqgovBpiCINXpYhcuVAFMpEpe0L+740e1+bTsGqHah4t0sEZqXNqGrQOwOkNgwCI1pslKM6VB3RQzRHfCEFWS4SmaMHI2hEJgdZd+keotGHr8mSRcKlHZXvWBKxBxDdM5BDRUe13JaI5goKtxvWaCKGmtDEZ0CHbVZHHvSjFKc2MPAaJStNR9vByITWvbYmMiBZNtP7Wy4nWpeB5FPP5U58k328Q9JqfphMYIIRKjJzW+7yz0zZf3GrxGGtkcV0YDervK2PKwWbO+Dy/wzvplt0GoI63xOze37rExul3Q/c/669yxxFsXjuII489DkLXI7op7z+537qSygdDdw9ljJbb/atjWzv6M1vuhwkqV3RVilD7uqzRbGEEY6u0AIWxEUMwN2ahvMT6AK/U5pboIgJB2Y1U2tOjHD+jXkGQHaoNrjOnmRGAjcFsN07ztnfZatmFrayHegIG3Rek7cnXblna9QheM2K60LYSggfLOUeNp4Gqlx3PV79okrFGtN/Rk5B8AMkNG99Yno6f8MfEAbM/AbhSRajtqBLv9V0EjCBYEESNbQdeVZTWu17eIZWLtdwBmESRBjv5eJ3xrDsFqnttIrECQcSALGojBPQlTdMlp7Krl5ryB5ykUEDVSdM+9ENxVXvo0KKV41gkkcApT3zcsppk4zajBwoouPn4x+FjkNcOilBSZJiEl4TxHpjn6nmCTry4tkHOd0xYgjaefiqacF+VyuXpS3cvCEl2qKypcrlfc1WPlUk5ef83osVa1abj6bs4iQkgJmd0rsgDlaqRUXE06T6QUO3BKcik+xppxY/U2xnkixoloJ9ayoCqUa0bSxHSeMHHPyjkZU/KtXT5Lee/By/OI7QHmhleWkfOTKqZXvz0ZoosGerwn/Vsx4Lj1e3/NZWhL/acMHGdqL+pQbwe4Wqc2woTT8GBN8ujo1xn0MfWT1kaL3fDEexAdG3zTg60uqW3STrjaKDUwayM9qOca6LX2y9ave2WDndofczWTink8krWM24M0a9DjZXqpg95UgFW6CLbVv2kX9wT8Vu14h6Mw22cauXfL0/zEZlfZ1V9Bqu3NNbjAN6nLE/NaB5cunTWnjA7j/kRffxWMmxTWPQwP/TZ2En+LB7vhbczG6d2+tbne84E3I7hJDUYIrgrsoFqv8GlpDao9E09GTKhxXJqJQlV1wbJ4oC/BiWjsjVdKXglhIsbAKcUtWXCBlCKSrEoNudp4bOA6lFI8ILcUt9XMp0RMASKUuiWLrVBKBdwYSbMn840xEOMJJLgdMnoqqxCF+TSRLpmcC8u6smYh58A8Rc7pwYOyJRCjb30iUkG90S+BuBaCKaEUyvKIPL7BLgv5MbOkCQ0RlchlMYK5rUsFAidEAuk8kfF0W9e3F/LjiplgIZFenZE5IVOs+RqDb3disGZnHkopRJMaL2hINkTxtqdEPCVMjMvbE5xXIJGmmdPLGRPfnVm04M4538mSV13v1gg8d+lIt29VxoeGTk1FsbMDDddzOM7Na2w7OjZKwaMUtf3Yc/jHb1tm5nrsSEePYnbt9G4bD7m9bwfFHaMHJNyhzWEEd+O0u7P2brO/yG40titu+9tUV9b7Md65AeRTdW6ANPAbjMmyxme++x1pJPQw5p0gN9CgE2w/v6k3e+vaMB5b3O7r9ivdQHHwJByBa5Ok2j3WnzF6FDaV4vYs6c/Zsv4LPXHwzRqqbu6boW03Ltt1Q1+GtXYjuO7uHmzCd87ueYHDtW7IRmLw3YeVaj8NUDeTtFInJbKBrNUQAPMtS1KzrdXxbkkEYvAUTKZKyQNwSnUyyL6HmIVY2+LJd00LWvAUUsXHPkhde9XdP0hzgGlz5330kAAPyi64DU2L747R5l0AUes0SOqznR+p3p/mThmsC+QVyormjEpEie6B2NaMKrquaAxoCIRze28MyxnNq6efQrEpeTjAlJDkY+0Jd5WSvb5cwcukQIFyufq+XpJqVn6qCjJCTJAc3GOsQdhmDMv/M5X3GrwaUREOem/gSFXG7A2d2Er73ewiUuPEhjpuvu8dDO6/nO2ZA9GzmouwAU7Y37t//59wZjg8be+gP9QQhstsJPB3WmrHHzL0dIuIukOfdrUa1YaFbNRskBaPSraR4Dudkn7dbsuUQ3HQkjvft5FqoDIeH4Sz7fnNpnaUFZ54dgO44yhuxPuWITmmWWq5C6nehK4qbEDkklb3HGzXd6mr7ee12cCaS7wN/dgAqzoAdN1Ynas7OeVaG3a9aBLbIH354xpQbJqEphrtwD3wQE8BXbO9NUakUr6qHvOs6+mUKKaeB2/3GqjHYJkxRZdyQiXKqgUt0QOLawYMFNY1E3Cpi0DdckXJOdftSrwjqr5RZc4KUTiZNNggK5Ri5FW5rAoCoQjrasQOYmDRk0A1r1BVw7J6rr8ghBQo2Ryo6tYvpgXLhlqghODbi4Sq9u0cmOdsLKWgy5WgC8FWBFdxSwzYHJgsEhbzbV/WC+CbYfJyJvoKrJoOrR6B6umi5olAIE6pz2nJuWpK1Xdvrgl3VTP5cfF9z9LM+Zw8ldfkA2nB4+2Isa+TUte5qDMAn6W81+Al0tKP+O+NTz1ct7NQDOwx24t3fDfeXW6dA5rabntj3dgadgFP7LIyH5/XggePXO7dcoeKdtAZgOv+5cK9r7eijDWE6LbEcACLdu0mpRwrfdeRbRzrrgt7qWWY0PF43wdKtXKlPmauwtrP77EFxwwPHdXqw7s3ZR+Pob99TO5PytOMTHuUDexmtzi5JKYtR+GtCrHt1XVb1SCVDWxsU4X6DrwNuNiYKUZQbPXYsfI2OsPobUO2l5qeLiPobVDXFMYby9IYIM/nJyABiYkwn7F4JjwEppgI+RFdLuj6sTsIBIEIdnVV83SasFMihlg1MnW+gxPfJWdvc6geFqkyls3xALguK7kUshoiypozBGMy8Sz3puQgrIDmgqyFr/+ucJoD8xx58WDEk9OEvGaWJftWKrkQitusCLBKYZom5jkRJ8FwQHYpqoJWSDBR1Yfmm06uBV3VcxsuK5oLUwjEFxNyPhMeZs6XhLy9EN9m5ProqrqSWSa3aaEwIzycEjYVisIjhZAzcvHBqAn8SXPqNkCWzGW5YutKuCysbxcWBFJgCsJ0PRNOE2FVEhBiz13sb3tIaMg98fBnKe83eO2khO01O2Y2GMtR0Gh4I5V43ZNx/HuTSvaEYuR4j5z3/smDS7E8QRSaHUAGwoQNmFJT4LDZX3aOGAP2yK4H7E8ydKEa0re23P47lrHn3X4owFHqGq7e2lJHsl4Tjlcemuw11jdopK21K22jAOm3DgS8VrhlRxkobr1275W3B+S9S85eGt2nj9r3V47fjgxGxy47AJQNoLF5DzJc38QtOwDLsXQ7VwP2YT31RlWmZGu+q7m2R9otqA0lDO6D97Kc3LQJ2OQXcWmuu2qO4w5Y9cOuxC5IApkI5zOyuD0mPC7E2cHLs7Krx0VNExrdBVFzoeTiKaSQKoA5kyDZkKk+N0AprssyrbZGpAK/J/gN2cNxmmNJ22GgjdWaPUJZixFDJNWBsTJI0kWrQ5OvIQswR2FqruOCS0JZa27DYYxq0t/G4Ggp5FLtcsXXRUqTO008zC5h5ky8OoNQrPiOyW8vlJqOKpox+6ZoDibFVYm2LOSw7cBMlOqMAjEKmGIlo8sCy+qS5SpcP3mLmhLz7KAXwMQDuUnN8cm9PEX13VzPN1Heb/Dq75Yb59tr33MZIrfbndTjYXzhjBpDsVduHXCh1cCmpLlD3vq7XhtX1TrSAPIgGTThZtevXoM1TWY3qDcpo3Fv4/U343NHQrgF5H0FnUg3R4YDcd+5hjM4PIzY0Dq147YH1UetZBcjJHU8pVmutvEce9GuP8ovQlP+yviIA9i27lRAbCAkhwHoc7M9pWH/Xf/JxlCMThFjm7s9pl9agakFFlfPwg5c2zKxdrFt0NztXHZ4Tp0LqaqmUGNpxmwbbRDaDtj7Mq68bZwbQ9XmYscGDEtoBL5771Gfg3tPbcuoqVV7oHWEcELSmfDiFRa97dPDI+lh8kzu5pkwkrhnnERzVddaKFNGTjWhbhSKNrtLIWmsfRJyVrTUTyOwNdOGlkKhJhCO4o5ERTs9EfGNGK/FVY0pBFQDIdZntfmuqs2IZ7KXAFPyoN4peQ5CM6Os6kmKxfcrMxPPtRjdHqDqiXWzQa7gJWacYmKaZ+Jp9p2Pl5WYXPpec2ExZXkbUElIiKQQmCdncgrwtiilrJSrUmwiJkcgmTyNFEGwObBcKsiWlVg3CVWMy8efkLWQ1kx8mEkRQvS8jJrEGQyr6dkE5Ds+t+HArR9fLNv9teHckQuvL3gDtuHYk7FBvU4Zf9JeXLlpxVgvt7aQ9hF3qW4h1+XI+druiXtC0LySjCecFI7y2BOcdQemwxPuqMx2vzowD6SrDvU+690AobK/v2XWjmPAQu3PRuTGvIlHcLvjPXoocvPl00sHssoxNgeIvrHiU5UNw7vLT2jWs2hg5vFCNfHulj2DbtsC6+rEevtO9dc65PaRA9Mh+4wmrVnHOqTXMdwfqjNB0/QZLZFTZxzeVTbAbyrC/Zrb5kIxMmYRrHRvw+Xt6sBCoSyCsCKqROAUQKLHHsXTqSc4zmsmX1f0smJWmPNMqnkITSZKLqwls1yvoE78r5crOXuGkxcvXhDmiZDcVuMu8W4zklSz0hcjSSKe3C8hpsB68byAbx8z09V8f6/JPfBijEhK8PiWoB6vlkRJ4raySTyheFH37ms7HksEXames9FtddEoSbmKsainpcKMaVVi8dABCZN/SIgFzFZyUd48KiaeGWSaI8UiMUWIETXPXZhVWfPMfJ6IYWZKgfMckSRMJ0PXmaVlpV8UyW7DfXzzBsmZcFk4yUtef3BmfohMLxIaAsmUcApcPqpbzeh3sNoQqSAw5F5ygGiv+1ZuAopl4AMPQtNGWp11v+/SeQSt4aiNz9gAU8GzWA/XVu1Ava5Rqn3grdWKwjHHFFQVnHWuuHH4N7T00MYOck9DYa1Gh44MAzWGGDTZpztA1LYM9/THy73xcUlYuzquoW+713pAsxN+dcliEK1utJXW3IvbEwYxrK8b20CIClD36htLn7QNNseVtYcS6w6cY9yWjrsbdweMClz1uP9X6BFgn4oS1fEhbADWYtSaqnmTZhgAq7NUvc2jPc2nQmq2E7Y5YHPXb0xe/6+KhSKha5ODVDWv4baTrjM24vYm0gJpsxrLWvja1z7i7UdX1sfCBx+84sWrwOlknFncbT0Hd0TQBpGKLcU96UqhXCDXZLthjr6/gBgF31rEsmLZCKru8JEi8ymRHhLSnCqWsI1HASs1v2LwjPcpBeZTwErGSmFdnOWMKTAHOAlIFAiBte7sbMUZQVUPlyjmLvtmVrc+caZDTJAomA9clbgFI3C1yFtLZBOsuBt9WTyGDPWc9SSpEmRAqjPQUrIHfWtklpmohk3KqnXnZHXHlKCCmu8L5mn3BCUynyaCGQkH27CshCUj2Z0wdFlJb1bKPFGiMs3KdPKAcMRYPUocy+unLOp3l/cevPpfE5pOvb2io4SxOcTT3U733PlGghq524PaZuPpEsxOvNg3Z6yJ+jxjH8TaaWurw6gOEdYr6XBlo6P6oEIa+kQj8q0N283sGtwId6txR6iflligCV+2e+Z4QwewKqU113Xr98hGRCX2SoXRjjIoABtYdFTZer5lw5CGp+2GDSNlu03qQQuygUHPgt/qGjrKfrxG0LIG1u0wdVfkPoqDdCF0sPJNBXXbjBDzfZda5vh23MD9U29tY++yeW2xXWHnzNSlaOMwvdKPb6rIoe0D12HDwcYzNy1CGGob/9qQ5tJooOgN62tE2kaJFbzM46yWtyvf+N2P+MZ//xqXb3zC8uFLPv89LwgfeNxQvmYg+C4kcar2T8XW4qCkSrlCSMWBSJUoRkAd5HLxa1clGu5ZkCKn80Q6u7v4shREA1qMXLWZpe7UXMTTRcUQmJKQk1GisRTFqofihJt7mtPCSg2GruCkZpS6wWRnBiqD69oG9xi14M5eWkMGVAMXTbyx0qWvuBb0momXwhRdRanJM9qrO+djeJCyqRGLco6BaO5/uKqwmrAgFAtMVhMb45KiVEZmmiaiBKYQeVRFHyOFhagZNUVzIb9dWU+ZGCPznJmnSIiC1mBvNf0OB69d2axbe36YDlxHmnzknWWo5aAQHK7ZXs/Q1Fx36vZKjkBg/SW3/s/Wxj1YNgCSLY3Sp3DfNzi8f/SnlqM9R+4cG+Sqw4PafleNwG+ceq+9Y8E+HmtTag2z1Iio0XO67Rtm9ZkOgKMatj8n9Kp6sHXwpm7ofm9h9J5uJ7Uaz70p2u0hradOmBvWHtSmXZJiB1jN/7sFKneVobbEsU2VuElL+9iufcND3Wgy9DyGh0kfEi06F72t6S5HSWUGatVj5pp7S7CtkXE1WGM46l+jy+/bNUf1t0AUIYaESKCshU8+Wvjkq7/F248/4rI+kq5vePH4mtP0Ejt/QF5dMlAC80mx4LtWFW17oKlnkFjwpLtzpFBhWCGsGa1Sw6uXD+gpYfPM6ZSqJ2MgTMK1us6XBS7XTM5K1sKyKpHAGuBFgomIJmE9ZRKerT2d3ZMwiEEuWBCyuMv4JIFggpqyLnmQroW6dTTmuZSqw0ZgKYU3a+CjS+R3HwMfXYRlDWhW+CRTZCVOK69eT8Q0wXzGXrzganBRyAJZPK7rUgofXTNRAUtcguAyWcCCBzwXCVyLwZIJdQNNib4dTDyfiXPk7cePZHnLVK4ELRQrvFk+onwjc11OWMme4muO7hlqmayZJX8Hu8rvpI7tvdy9YOMLN74s+2uaZFZVHvW9c4VKlSR2AlE9Vu8Ko3v78ND+okKXdEI9oU/5wMtAMhtF5Cjp3cJwc1LZ2jeAx3GNHCnQIIXs4OpgNzmWUTgJzasN7Q3t2zPVZzThJdSGam/X6DzQuNK4t71sAzIMxOCdyV4CaHkGm5XvCMTv5gXaKnpX8tDjqrL6nDt7hdm2BnzP+Pp3yK7RVYajZDU4aageJa86xqPtbQiSRQZ1YFPjjS0d7WV26EtdbDv2xEYoH9ajNA/coYYOXFu1ZtQUTsP4NqOeVKancoIpJk6nwMuXxusPEvN0puRIOk/MaUaKsDxefSv5mJimyDy5fUrVPJgYdwhSc9VrYyqseO4+W1ZYM6kUIsY0J+w8oafJxztr1YnhOQepOQOzO2U471E9Cc2ASEzCHDzxbZRIionplAixqvFUPcC6bqtiIUASCB6nZlrHWNw+ZhWwJCZabqpscCnKm5z5qGTeqLLWmKuX2TivrnKVGN25JAbWUsgKqwbyUiD6mCzF0OzOHqbGMrlkZDEQQ0TjhIbESvANKtU1O1Pw89PsuzCXYizXDG8LkhUpBiWTL49cS+Ftlbams9v/xAcVjZ8Nft5r8AJuqJAd/rZLOpgcJKHxdRrv2l12w8U2z7h3NOeJc70t9wBvIAKtsbKjCntyuwNHjlLSePZ45TAWrfqd5PIEsI6NszYKG9Jbb/+WE9GxsxHasarxgbWVT4zZNi4ydKeNZpXO6kjZME4duA4SsLAB6Qb2snVxN3QHiDtIal3iutP2rv6zo9Q1qAgPoHTz+yBtjQ4abfoayDcPwq5OrSC3eTIyLoFdl48d3PEMT5Rbj9CRVWh3369oy614tFD7ViDTFHl4MF69mjlHRXNCUyKFBCrkGtsk4nnyYnIprJjbegotX6OCuKdhjIFcqjpxXZFcPHEuntHdUkJT9A0ZiyefZXK1HSG4hFScWPfxbOMdnGmY8H22gnisWUwBd0ZxKdpDGMQz17dNwkKVTgd1cOwZ8t0bsKmsi7la8lKUt1l5LL7liKmxFAcuNXpy4DAl8rpg14xeFQ2LB2SbULSQdXtvcmxMj6udVSJFIqsJpqEyngG3HPo+azElYkqEyb0RBUNMCWqYrhQ1lreRZQ6YzUwiECISEpKmd6yuTy/vNXiN6oqdF1s3pm/Httdo73l1tH3daPq2Wg8qwkYY7l9721gZE04cOzL+2XrX2dtGMff9+WZKN7Y/1bD7GPzuMlL+evMo3O3sH+bSmAximJPwWylv+777U7vfjCehB3jJ0PaGMz0uptuNxrVQf9fm3x8Sq93rsni1pdRciQ35ugOJQ/jINB1jqpqrRgMnGAKS+x5eLoVY38+rqg1rnNAWc3UAhu7u7yrD/ka07BdDW3paKageFFvOvO6i3nuxzcsm/R+8CyujcjPCA5B6nN2erfLbRo6zqXT9WBDhnBLzywn97s9x+foblk8eWRUIEW0OCHHyRLun5OmbTNAQWEQ9Y0RZSRhpjpxOJ9I8OXOQM2VZsTX7WogRbT6UEuo+XhlVCJpgjuQC1zWz1pyHAeGUEvOUmE8J5kCKCREhzdE9ABGktgX1VRACxClUl/+ITYGamoM+LBEHtjA4GtWxLkVZs3BZI2+ucLm6vSsUIRej1LU7peA5GWOg5AwXRa+GSvE90BCyJrdviaeOomZ9RyKigaUIsQiPJTGFSBTfCTpKcEeSRaHGmxHgKsXVKdk4ayBrxkpmMeONwJyNc5jgxUSYJtLDw9038Jst7zV4DbZgwF+SYZeieqxxkjK8bI3wVA6VnRBS72tSw3Z8p0QaqGankc2Da/DKG3f96I6BdyWMrUKB0dlu60kF5UG46NfvGl7b4wKK9e047paO9/cvuHe08YfSLuiDN3icDddaa0dFG6392+quSWFl6JschJwD/myMSHtQq/DWTR8a41AZgo5628w3G9LmWXokt5tdqNtQ28T36d4H4XqzqqOO+z+DeXofq89s2dx3Krx7UlJdaLtWSe1TqNnMQ/Ds++3NOKqLx0Vu7OrfYhc3NrAHyva2HNbebnE0jmOvnLXDedtxHPXlCZXpCLhjghbMVkoRYhLSQ6QweULYGAkpMj1MzKeZmBJmiVy0bisfelYRw4Ep1DQPqoYVJ66xBLJ6tonFQK6FEAsSqg01+jYm61JYF+WajbeXwmWpGdlRztOZVMFrquowMSEUY2VjSHziAmFKBDVCMaLSvUG1e724qlxiA2N8S5gikHyX5zUrl6XweM1c10LOBSvFE+2uC+s6Ybm5CxYPuK6u/hbxgGMEJHBKJ2KKaIyUGMkhYTWfIQGyuYQXV1BxG96cIhoDWbzefC2s2cgEdIromqq/fyZWho11JT9ePU9iisyTM1qn6bh/x7dW3mvwAoYXcHsz7XDy0ySL9kLeAJjsa+zEuv0amcexrkPFfY+lXZ3bxUd1X4MogZ1r/Vi/O+3dIdhjJw4dku2BO7XHbe1PFRn+7J89OhCMIDXe1hiJkeaPXqCHVt5pVgOPFgW3yW/NY7B7FXbGRIYqtoFx4ryBxD4cYkTK9mjbVJIySDhj0+hDcsAdG0CpSlhPfPaXH9mpDbTbhoZNymlbjVC54v3KP4zwbk1sX0aGYicY3YzGFno+jtZO4SGNIWoMXYX9fuFWu7S1JPX9aOEDakzThD6oq9cUlzBj8Litea77ZoUq0Xr/Q3T3cgmh7ozsCUsdvMwNR54QA8VVbWHN2JKRGGF2acgQ8mJcV+WSjSVvKjqrQfVBXB0puJpWi/Vg5y0GpqpzQyQk93RUqJyxVefbrc0SoqvtrKooS0EsoKl6Ohbz7BpDXCDqORrzulLWlXLNZDMPlC7FmZDge4aJufoypch8PqEpUlJgsUAxoflRtES6S3HAlegxZCoe1KwKuUA2QSUS5oTl2p+Sh40C3LOzLCt2XZFrJqToAW2fobz/4LUr9WU5uEbdI/7Y9g5tr5IN3PSeq++kqr2E1gW5PaJ07nUgmoNnVSO/u/bYnjB1O1Jvy8ZCHwnJ/e+b6GLjCTv8rd83D/pbMuXv3n6gNua5pa8JdAO/bIRtn+3eaC7wfWxtGyc3sO8ff2xJOy3iWi+wHjvUnWbqTWNA+CYNNAlr8yU1xmDsm5mDXsfWhrD93Aak9aci+bZ2AD0OeFMPjnYfq9nQqwQ0SG79EXfWMx24QlfNbvPZULRpHtr7McCzbZ6HzeFliEf2HHvDaLTjbbR8fof10Xze68R3wBoXzziwXTKpruB1fMSMaEJ8eCCdZ7Kp7xtVM7mTYD7NvuGkGWV1X3IxIUaPvZI5Eqs6UB2l0KWgS6EshaKO2MEMWRaCuAdxPJ3rbsYBvRqXq/K4KEuBtXj7erajNjjZdyBeV09eGwRiDKTQtqTx9EqibhMSU49Fs6qRMPNdnWkbWTpzmrNn1PDYuUCp6sFc9+Qq6tubWFGWZeFyiTy+ufD4cUJndyJRze7NOQXSPKE1yPvFywemD1/CnCgp8LgY16xcVk+rhRgqxpILQgZgSoEcQ43JFAqRIp5BY36RyWGipJVSHhEVQvElUcqKLbA+RjTOpJTI39H7eTVuzaCpx3zatYKTB/N5qbAheyI7EjIYCfieULWPZ+SuyLfjjDfqYmOttn+5b3iNbsvwC6QF1+5KlTSstW/L6zc+fetnrfemnkNr7wlfT13LATgrQWpgsEdI/x0qKy3DCPfg4MqpWiWVXt39ljQgatc1aTQgnqUbGUcQqFJfH4NSz4TdtWZy+8hOjeoz1RApEJPr+VU6cCLNMUW7Sre05/ZnaO1jdYc3D5DVcfm0xLvBaiDPOLaA2GbC62vJ1WMhNtf40Ns+eimO3/v5vlQbcBwy1Q8s1l3Gbxyuejx0z9C95DxYKqmvKUF8Rkp9vP+puyQT67QJIU7EORBViUVZKEgsnn+wSmcBSJN71uUls769YG9XyIYQmU5z39ZDzO05FgILsJoDyWQQsqExYCkQsiHJg+BVCiEJySJTgCW7I0cKigTfTmS9GNmMTz5ZWK6KSGKahXkyomifo2ji2TWyIou78ZvUVEmpbs5aCVoxl7BWhZILKXowt1quWfPV12NlAEqxunVJZrlcWN5AXKID6JQ4pwlJE5cMxEBEeHj5wMsPz8RzhCny5mq8eZuJb1auiid2logRPHj5mikFXqye1iqlwFoZoxiE+YMH8kMhXxOLnnzSrwVWV7PmvLK+UTQbnE6UJx3DvrnyXoNXL7u3zH/ck0IkNM79yFvvA5iBMfXdoE7ZTNaj6mUjnGM5uiRIB9nxrk72xz4MrO5RxTNCxLtyFx7vferCm5F4Yj3tx+xexcdR2EitDG2x9ozx966WHYmuB2V3oJkIWqyQtE3dZPOwa4xK5f03SU/uS3ntvk2Iaja9/YBsDMnmhr+BgWxSV/ccsy5hbDN+GLkmgDAATV82o2fbtgLGfbqaKNwkrZvnDCrJcaQ7VA3nZABNN3nZ4a5t8sb3x++U4fc2t+MIynCiQumuPQ3opIrXodp+pM5nU6dq8QS2oTKoQSohLxnNBY/FgzBFQnI1oAjuDJECOkeURMA8o0WqDgviqjOMnlVdgnszpgCnksCUKOrbj+DSmObC41q4roUYBEmR2OrIzQkHVFuGeatMnA2LqmoyRsZGZIty7iripnlo9Mx/xwBJlKgFWa9gEasxYkEiKQZODxOLGkGBFIgpMsVIiIESlFUgYVzNHYVEDGlBxQaXxb0Apikwq2DFmY5gwoNMDvQYcvYgcl21LmNPSpwK2HJxBjS+KxTl08t7DV7fBG1mA5HhXFfTDQgFVRrbiPQoLWzv8AZj7QJXtXSqUaWDVufYihGjxjd4eOl3xLIRJavPGe7cY8Re9pHhfXhqXLpzhDFSqy3Vz/76RlT2dez+DGdluMiGcW7OKNKv297Rp3n8G9OP7dvw1J2jEnDLXclhAOl93j9zAz+qNNPnfDg3CuL9ATuwsGrhaCAGNI4Zo9SMCWZuk9mQ0Lpw30C0K/CahNOCkdsaGkHehsG6W2z/bVBtd4AZ5nCTj7ex6u9Ie/7gjHF86o6JERkyanhKp86EYERxd/IWNVcab0JNp5R9d+UYChqkquLwtFNaqoMC7i4+BcJcVXFBIAk2B3g41bQX5llApoBFd8dHHDCLGe7V6JLFCakA6mq/mAQVWGoqq0spXFWJkpkkMIk7kLEWT56bDc0rbeNmCQbBPKVXEFqmcR9KgUiVVD2voLvNR5fi6jUScOePEJjdMZKZTMxKsAgaYZoIJKYoPLyYakosrQy6r6OIMJkxYSR8M04HbkiVaVAzHouyambKkXOJRFOcfRA+KIkpgkxGPJ14m66sMfu2NLXeYMaSFYuG2Xeww8Yd034/0yFi+3r3OhteeGfeN+5XBv3djmDXF7BB3Z53vNe2pwnzSG3scPVITOTw+x1d2ojE/dN3WwX3ifh4X/sYmzNG6L2ON22+eaYN98iW4cHCABS7v21OcILfOtekmmrnkt1D7jR61+Nx9O4AmI3YtrEEHSQ6mI1V3DyolprpAaN5GnYprHK2/dhhsDYHlLAxPqPkFbaNJrfMJhWkd7akAe1lO75LyCsOIdZ3qaxAZEfV4cZw7HorOACMw96Yiqde0XplY5R6KJNZDWZW1KrzQfYsIzFFSqmem3igfy6gS2ZSRdaVaAr56q21iHKmqMdLRQLMM4SEkZCzt0HFVakaPCg4SPQ50sp4RA88TwIpJVpgOViPJ1vEsEmIyZinyOkcmGMFZjN3WHjM5JK9szGQ5og0qTAEt3nVwZvm5M4ggFaVkaKkSyGlSIquEnRVKESMV1PidQq8DsZL84zvokJ+DJgVdJ2ASGi21ZKxa/HNN1EPBVgyRTOlrCzFM5ZMMRKrY0wx45IXuBhJhIcp+biI8YbMi9mYQ2EKwpwcgFczgvnbPonzECEr62dLsPH+g1fPpdcIy0Es6K+byB3C3OxSW+69I5Pf75AKLxXsdi75B9rTX9y7EPMO6JHOmG90srXhwFQ/Ddvbma0d7caRWPnvp1oIdOmxwdMIXkdwGnsV7jWwihDbliq2o/nbLdYJZwewcazv9ncPyU0luLlpjBkl28w1cBwB23bPGpkJGWb0sLIGPN2rq7ddj62PvWdor3Fcahu4NUnr0Ltt3tu/zilTwSuEzdbV6h+iuXpLdxk/RzVhd66wvv5uuKRRz9uPDzqKhjxDS/fANfbgtsSmzm8j6mIPFGO9ZN+9GGGeJ2KqW5GESDhN3n4t6LpgJbsjRFk9i4YI6/WMFiOUiThHSAkNsasJTarzQdvyXqBUkHf3cgPbMuI3e6dh7sVXd7+WGJjEmIB5Ek4zTMGIWqrdk5rKs2XNkBpI7H/7djZVLRin6EHOwZ0ifKPK4oAWQrWjCWmaiFPkdBIeTsIpCUkgavH0dWaUIizryhoTj3LyLV+CuDRXSo1DLCyLx7Gpui3al6qwFCVIcWlYArmmMFtM0Zw5hcA5CessrEUIpoSSieZZ81V8k802tzGIMxlq95bDN13ea/BqSOHE3cZ3cnvpGwAc3pxdhovhZd0RkDs404ClOUy0dFJHCcAP+b97Wn5rCPfrqxz3xAt/vOeeP4Zhu8S8N0vjjlT17i1fZAPUjf/f2mbQFa1Dv9uzNzmnqe+kp23qFQytNfbg6wT9ANnjYI6ICpXgtFitMaP6Vt9GiK0DZGvfLt+ebV9MRjvQ3pt0wFg2Q9oIXEMAcv80CaxtcTJGE7YhGdS6jIyX9F2AfbPJLdpxF5Dc2j78246PktxOHS7DJDajb7MlHezEldT2a/ZtH9c/ddqk245HBq2+wnsWxHVUWDbyomSayiwRYiGKYBFknlyKXzzWyYqnRrZS0DVT1FgvV0oyJEGSk3sQiks+yuSIEoLnEaxtU12rmtm2LP3m3qBVAHGbzeo5+wwjpLpPl8ApwZyMSXz/qu4uFCvTIXRpyzfTlD4Qnt8CUt1CRVMkW0CtIMUzb4TKtCQJzPPEHIUXIfIiKrMoUTxzCJqhKLooyyVxlcijKPEUq2QX8Zya7gCTs1KKUpp0jmf4L9m6fTXE6HyFKqVkT0Y9RSYixQJZAxEjZWckonm2/jywf74NzDEi91sv7zV4HblUZ15sdzKyvSyNqw4i/SVqF4f+a/R8G/nt/XN33OVdUejIv987fuea6gE3vsx7cn6Lufu2HfMZ3LuGI1W/ad1TJ8dny/BNxmNDJc1eeICkOw2q9Zu3LXQbz9he35xvxwqY7bZeuVtvb9PQgYZvO/S5V8U4mptUs5OxbADkmj3DMypseQsx65k0tH930LJhUu2g527qwD7WDUS6XXLQPjQw7ipBw6x0deDmik/rPJ68qwGoth50gLstDZBugaut6LK/8ubWXkWg57+UYYudggN7LsU3QwRPaFuWCvzt/tArMoSMcDW4mifO1WKEN2+xGZiDJ5AIEWJyqYxUJa9ICKm/yqpSQQsiFVyoNknfWKu2d8WKEoKrxBrwRMmYFSfw2SjVsSSeZ05TophLUasKsdRQjwihLQQ1pKg7YBCJAgWXznyDRyNG48Vp4vMfnnh5nnhxnplt5ZQX0nLh8nj1fbaWhaxXlhJ4tMhjyJx44RtNTg8euByESCSsdadnxddvqXFkhW6XhQtFxB1RzLiUjOjEg8yE9ICJsap62MLqeQ5FAjxMfQfqYqvv0PzZPOXfmXn091R+/ud/vqN0+3zpS1/q582Mn//5n+fLX/4yDw8P/Pk//+f59V//9d/Ts54knhVV+ut9xxjvx+n0YDMe71j5Lm1IXbfuOHqb3aPTxhoboYd9u6h1t09L4DoyuyPX336H4RoXvLdPN+EfmN+2fcKO/AwPkztjsl0nu1u8vffpexsro8YD9YZUIt91hNKcxV1FMUzUrt+t8w0amrpNRunhKF20yw7efP3yI4szzK7ZkAtxRLUqKd20a/jVbUtD3V3S0u3Zg7qwf8b/TPpnG5jQZ3fnVdhsXC12aFgD7HpvfXz3o+XtuVm37KWfseYW9zg+qTuPiHQKMvoa3h3svlbqgXB48YYVbyG4J9qUkDkRzxPpNLkKLBfykrFLwRZ1aSRNhNMJOZ3gdMLmM2E+k04nYkyYwKrGm6y8WZTH1bhaoISJUrOnE12lGFKCOEF1LQ/zRJwm0jwznWfCNBPSREgzMZ2J0wlJJ8/TlyYsJkwSmeDbixQhq2e+l5SYzp7OKswJrRncta8BMDXKmsnXhXxdKNeaxipngiopwpwC5znx6jTzXS8f+O4PX/KFz7/mw8+/5uXnXhNfvySnmVWCB2CrMaHMUki2Es0dLIIEUoxuQ5siMfpc+Capw2tRs8KoFVb1fdDyupDXBS2Fdc081kwgvmYhqwc7FwNJifnlC04fvGL+rtekVx8Qzi+x6XRcMd9S+X2RvP74H//j/Kt/9a/67xg3r5K///f/Pv/gH/wD/uk//af84A/+IL/4i7/IX/gLf4H/8B/+A69fv/7WHvQU/R2/jG5/IyHbPAFumG/hQJeE26wEDRxGTrmLYnq40Hb33spi49Exg8Q9wt4aNPa0gfFAxlq3bxjoOx09njmoiBiOt/W8XdGkkH1bxm/HuLWxfduRVscWH7fP3XsQb33w6Ztxwk6Kkv3P/RTBll3jOB6HFvXfxjCYR5HuDvDtxalBEqp5DRuw7JbkwDhUld7I3FA3mZQwagcaALrqh93xoZlU78SgaNna3Jiztt53q6ExeTZcK67gHkHnuG52pc+j9fv3HFtjw+hOMRIC1iRsNXyHgYCuq2/YWHQLFBPP+m5p6lvCyHwCiYRSiCmyiJDxhLahKEHMg5glVJFncJhACDZ1bI1VbbhjeOrYxhTQ0vJllM7QaKmxfChqzv4Fj1ompoiZ29ay+L5uMHjymmElo2uoS0lqzkFDRD2eagqc5sCLU+LVw4nXL8+8fv0ANqHXiGJcP5qxcEUITFpIYsziqsxIDczWBmKCBLeD9RhGmgaoSefSw1PG3b7BWItwXX2rmClS7XSNwQkwRdLDmXCeCXNArxlWxex6u16+hfL7Al4ppZ201YqZ8Q//4T/k537u5/grf+WvAPArv/IrfPGLX+Sf//N/zt/8m3/zW3rOOznjHXHfLmlc7MbJjwR/H3XSvt2CSCNm7VcjBiMM3Rr+Xfd9f/evo4u+4+4N8nT+tbV1D6pbvTdODnt86Nc/rR66348RYhvBkQ63sOOu2acR2iodwXYAkTo/JpAxWlaOpNScc2Mw8vbXXy//NzRbF8Vzx8k7xkTGWja0by7E27lAC+INsLfX1C1gDFcxOQg0g3dzyGhBym7g96bKMBbbCIUAPRt97VyXcqLnvou1J4aRKbB6PkC1QkinnlyVdcWKu3uHWUghYSVzfXxEApRS1Wtm3XlE28aDzZi+kxjpkp/i9tUgdxgy2Y3oLSdA1WLUTyegFcxCDJhMWEjky8fD3jkRKCAKyZrTnnc3JVQjIc7EV4osK6EUJGWwiFlkzYUgDl6IMCff3TeFQEqxOk7ANLWlJ6SaPb0DfKhSiCpMkbwqumodj4KYoiEilrCS0FBopjVSJE0JzCgESlgQ9fUSghAb+KmSL4ZIhpSRKoHKFDmnyKtTJL+IhM+d+fDVA69fnPnw5ZmghSVFPgbKwyvKYthFCddPiJNwToE1BHdIKQVbM7auYAET9/L0naHFx1SrY6TAVF/wVYSrBVar6kFgzZm3WvjG1yd4dULO0TP0zzME91R8ePWK+dUD6YMz1+tCvmSYHvks5fcFvH7jN36DL3/5y5xOJ37kR36EX/qlX+KP/JE/wn/6T/+Jr3zlK/z4j/94v/Z0OvHn/tyf49d+7dd+D+DlL74M/25Syb6EJkk9YevpZeAmd8ROtrQ5TrTp6se9rHGQWgYxQzpHbHXrxnZu7/Sw9W17tr/XnZptZ2UALR1v2YjxdvEwPge2v3O+d4ajA1OrqjvJtAb4fXq8XeslgY6ysu/k9vVGADJG5Wy71NNBWX3W5kzQ/1MQMUrjGs1oW04IdR+1zvg3ccM2ATqwn4um+muG9nuUuK6Dvh9XXZeqzaZVwaFnh2c7bhsH25QDO0gfpJS2/aWKb0ohJkSLXIM7KgSFy9tP3OhelMf16ttwBOH86gUfvnwFCNnU95qKE+Hs3m5gqBU+/vhr5MXlBkpX9NY9qqqi2Eq1N7U5qiNiG7D7z22fOZdNthmNg80uVAS0OiZSs04YhXzJmAoyJIaVKTKdom+2GEJ3OkAgqPHw6gHyBLkgsjJpoGhgXsVjkCZjjrhnXgzV7TxWBwrPadvetyCDzasmVA4ECImcV/eUnHwrFqnZZKIFD3FQQXiBhiuIUiK+2WP1NE1u1KqppDxeS9SlHC0eWiGaSeHksWxiTAlenhP28oy8WnmYYQ5KqkHZDj6B06uXXIuylEy2ggTzcIF4hjRBjCjCmn0Neu5iYZ4iL4AVIy3CkkGj1DRXngfyzTVzWYGrx2ypKcXg8njhFJVkE+dpIr54cMeREEinmel04sXDC+I0scTMZblDqL+F8j8cvH7kR36Ef/bP/hk/+IM/yG/91m/xi7/4i/zZP/tn+fVf/3W+8pWvAPDFL35xd88Xv/hF/vN//s9P1nm9XrleNxHzo48+Ag5SCmyc9AhQ7WW6ox4bYWwDqUos7VYZZjcXV1ImW/U9w8NI/CrAbO287aPVN19kJFwHD7wjwj0hNd1kpD9+byKb3QGsw+8uf1Tvs81bcGtTVyMdnnXMztAlrps+7Mj1oT3Hu/fzsJtqG9omjQHYEGGfBqxVZDd93j1EuGV49mL3BryjTauC3vh7O74xXjv14r53bCnMRsG5ic01q4QGxDIlZ9brwnV94xslGsRZmKYz03zig899jpcPZ0/casYcEloKmjOp7owboiAUHt8+slwX8qWlvaI70Izvybh2bvgh2f/tfWmMGJu99uj04cvTvfty1vpeVIlA3E18nidCihiCqqd/CqJMASwkLAtW3Pg7G5hK3fjRCMmYozEHd3xI1QGi7XmzLQt35NrsgYO6F+tBvJ4iMfTA88aruZB9AlWEgoXi7jHVthljRCLddChmW/5DvBJnu9Q/5tLmFAPnKbDOgSkowXweg4BVEJumiXw+IetDbYgzODHMME/IPCEp1vAAaVPqwCfwgokYYVoFDe7hCb5PmvWsG0ZRI6iDtuaMLp5AOJ7OhHnGuckAKfZdvmNKhCRI+r/ZZpQ/8RM/0b//0A/9ED/6oz/KH/2jf5Rf+ZVf4c/8mT8D3BLMdwXHAvzyL/8yv/ALv3BzfHN7aAtrPHd77T3QGknaeP4GuDqobBRz957Ww53YGJtKr7HTDHn6Gg262eywkoEBCw4Q1inF2MZ2/LaPsseGfn873nu1vbit/zI+vXLMvapDhSMGDbYh2cjtMBZSuXTd19G+VTQY48y2+6kR+/4wbYTF2pOOnMXwdQBt2/VvrwIVxro6V0J3dtiN/dY6B6iae083taGOwckdtJyL34NjUwbSCfo2fs0rUYhJIcxIJeTx+oZPPvmYr33t64SHQkgT0+nMl/7gl/j8F76HDz/8bv6n7/0yy7LwySef8NWv/BYxKx997Wv8zm99lQeb+ODlAx988Jrvfv1d/M7v/BYfffwRH63mKjsxQoguSQrdy2+//vaM1zg6m+OUj3Hz9u3/SD/pa1CNQnUOKNVDOBSPFYonRBIxnYipea8pGhRRiGpE8T2utAhkV22exFXRpRgihThlYlrdcy+0tEyCiZCktBH3ubbmYhXrUUUpTNFzVKoJE5vKeFvHrtdctSC2QI2dAgfh00mqyzyEFYTsfZCmtvV29eBlHCgSMIvxYoJoGbKyXtTnyAzUwau8eHCp6XzGssfBTSTi6YRME/E8d1UpoWXBAEmRNEeWMpFzZnlcakyWE7UUXUILoWbzL+KSXymQPQFveiWkh9ljyXDdYzFlXYvvJC3iIQufofy+u8q/fPmSH/qhH+I3fuM3+Mmf/EkAvvKVr/C93/u9/ZqvfvWrN9LYWH72Z3+Wn/mZn+m/P/roI/7QH/pDu2vuAtM9on2nbITtUy+9KZtn/uaZ2F5kP7EHJ9nynW/aNJFBJbkHC+lXf2vtum3oVskTzPHdcpSLtvpkd16GH3tw2x2sh+xWTVgx7XYeau0NOCqF7Cq/sF1sVFPSoRKrz0QgWLPDOcBIiBuAS8uN2IBscJTvwNXAd+vf9rXu3tucMpo9iU0Ku3HmGJ67H+fGvVQJxzyy0OpHl0fWdSG/fct//++/TZGF+WXgJ/5f/09efvAFTi8+5PUHn+M8v2CaTszTmbIs5PXK5Y88cr1e+D//6//Jf/z//X/5yn/9b/z3b3yDr3/jI17FRM5XTtPEH/if/wBf/9pvsy4rQYXLcq1ZQQ5tZGRIjZtEYhWfQgjde7VPa3U9D9WT0gxy8Yzvy9UQEukUmGb3sHRtmrIWQaZEiMYUIM7mzh1qXCzXbUkMckYmd9NeipEzqBWMt1i5QgmYBvf9iE6QPUa5tcWQUkc9QkhTde5wJ4pYGR6z2LtdaqZ6rEplM+4Vqa4etBAIVYr0NW9Y1LrJY+WvYqjmBarDhjtw5JzR64peV+xyZS2FFYEwEWJyMIoeqHw+zaQ5ER4yqiuqhTVnwjQT00Q6z3Utu1t8FCFKBTALPCQjT4klQFl9D5kownmCReHVC2OaI3pdYVl5UQpzMKagRJR5OhHmRBGhWKAgLEthBda1oHlkXr/18vsOXtfrlX//7/89P/ZjP8b3f//386UvfYlf/dVf5U/+yT8JwLIs/Ot//a/5e3/v7z1Zx+l04nR6wq2yv/Hbq3+f2G5fd9zRkVO3W0mM8X29JS/7Y4N4NEqDnZw2zgvYXBs2FaUM10rfZuRW+hjtULftGfn1/fmDRrUDoxyO3dQ6sNmjV2PXFt7Y/9gw60Yg2q6V4cItd2CrU/rY90SkNxPrtTg/3BQ7Q+Oqis8qMB7H0qX+8bvcfK+d7tLTHpCtfm3A1IKSt+waG2i1KmzPkezGXer/G3ghuLedgmlhWR4plzcsj4+8+eQblCh84Yvfw/f9z1/mj3z/H+Pl689zOr8ipcnBpjsheobzeD5jOfPB6w/40pf/AMtayJcLer3w5rd/i9df+F5efPhdvPjwu3h4+Yo3n3zER1//BlIyQsHKaAsepaht5cnunPRz0vpWHWOQfbyOOQa5x16uqaDUmYggsTMn2ZSEumQSC5GChAJqJFWK1HmViCRPuJutEELEiqAFSlWb6oon+U3B95mq0pSqb/6oy4KgxGjMDw+u7kuRODmgBqQ6FIU+AptFzzBxm1NziPD1ImChho2ZS44IwYRA8HCbtkhCXWfmCYDLkimXlfXxQlmLj1lYOZ0fPDMHba8sn6eYAinMNc/igkgkRN86xvC4ts7Q2sZEi+BxeCmR1ZMh+w7SgUQgzsJqhTWCijKvypTE8xvXuL2AeAC4ufNUVvM4t7qly2cp/8PB62//7b/NX/yLf5Hv+77v46tf/Sq/+Iu/yEcffcTf+Bt/AxHhp3/6p/mlX/olfuAHfoAf+IEf4Jd+6Zd48eIFf+2v/bXfw9MOnN8gUggM+xodbrkhgN9cGT3qvGzEdpQK2qlGdBqxVrP+ah/jYgZs2JmPxuPH695VRlXNEXjv1XEEsLtHN6q0/3m3ou3B3amwzUuTVnaq4luOvqXlG9vV6hqBzlCQyCj9hirxdI/G+ryNYRi9Im9Lt0/5JG7Lpv6zs0ZVQGvbdFhL/1SBqwUpG9a1L+3ePegfgKuCLjh4qyllWXj8xtcobz/h7dtP+J2Pv8aXfvB/4X/9E/8bP/a//+/8T1/4Ac7nl8xp4rq85ZOPP+F6vbCWC3m94NuIJC5v3nKaT/yh7/vDiEQe33zC22/8Lv/xP/5/+EP/yw/xvX/4f+XVd383n/v4e/jt3/5vXNb/wHVd4AqlZHeyOIKvdDkZBkeH1r++79rICDG45dT61OiZHfI1E0NyO8qcsFA2VSGZGJSQVkRXMEXFCNJcrQIWIiITZp7frxB9w8VVWa91J+C3GbtmwhSIc0CidfC6LivL27eYZmJSXn/wum6EGZCHRAhV1Ti7GpfmVm/Jl4UqkUwQJQYjJR8bg+44KQYSfC0EEZfAwv+fvH9pkmXZ8vuw33L3iMjMqtrP87rn3EcDDaDRBIziwyiAMg1kppkGGmgi41fgB+AYH4dDjSQzjmQmyQgjRIoQQDQeRKP7Ps85e++zdz0yMyL8pcFyj/DIqn3u7b6USYc3ttWurMxID3cPj/Vfa/la/xVXXWnhp1Q6pjAF5rNnuj8TfCAmCMYi4uhQa9aEtFhupre4vlMexUkVgjI7hV9U124t9yN17RtVsjtnSL7csJjpnebEDZ0FE5lsZiYySKDrtPo1RjdecwKxpTqACEmyAtfCiPLXP/4nB69f/vKX/Gf/2X/G27dv+fTTT/mH//Af8o//8T/mZz/7GQD/xX/xX3A+n/nP//P/nPfv3/MP/sE/4L/6r/6rv3qO13I0QPLUJ60mX98XUd9w9bl+TwTiJdbVB21L6NPCQV4ErmS9We2RLs5uj9x0Mpdz7cX5vwtwtb1a2n2ip099Vt97HJS+fncV3duxfWwWN+BDE5WYi2Yml4ZIbvYyL/v1+LX+rfGbrRYvzYlSzgoXM/mUglCtrgZjVKKa9rwVVVvKp5wjrWW/7HVVlyJ5k/ibC83Aus6anjSuOD97bt+95XR/x/nhjutXL3j+0y/5Wz/6D/hP/uP/JZ9+8iP2w2fcvf+GoxQvU4pM08TsZx7OZ3wIiBE6ZwnzA9M8M3rPtc0crjqu5JrfvH7FL/78X/L+zXf83f/4f8Owe8mXn97w+tlXfP3rX/Dh9jvef3jLw+0H5RJMj10/dfsKSvoA6hZHWH6r8EyLd2F9BtWFGsiMOXN/mng4zbjbiZtXgh0crncKjjZgTKnYlzsiEMTioyNES0yaMhCjJSTDOZak4ZiZU2RKniiR2CciE85mOivsDMw+MHvPacyc79XK8eOJq8OJ/a7n6mrg+qbDDYLrha7vlWfQWqx0JDMXqyYt919T00pulTIBl7WwrpcESM74mLRGV8xKhWUyGEMKaDkVn8AXF2LKJGOJIWA6QbAacRsjISRC1LtgOotJME+BGBNSFIMMylUYPM4IztiF0X7Zn05qLeWQMNaVZGlH3u/oc2ZKUXO7DRgrZGvwKWNDVALhziFGuSQFkPh7svLy/wXw+i//y//yez8XEf7RP/pH/KN/9I/+p7too9m3AQiLbiyNSFiA7AlUQ7+fjSyLKjdnr/Kq2Z8Q1Tg3+rNQNOpq/ZRvLVq2/tu4/hakbQfVCuktiKgm/7HjCZuoQcYlZqoRGG3jsuzv1P6vYNYmBW/16uKyK+OQ5j4s/a/GVdXQN9GV6z5YliradZyqqVY7abs7uJTnyOXcmsBrtoEOS4XnAkzSWmX1/3Kt7Xw3sFp9j+vUsORz1e9XJblxE25uQs7N3VEFSnJerlKXcH20cy68fdMZP54IKbB78Zw//nt/n08+/ZRXL1/y6uWndBbm4zfEFFQwhEi8v+P24T2znzHDAW+UWy90HWEa8ePEfB4Ru9NS8zHy7PknHO/OBH8inz7g5x7ISIz0MrAfnhFuDDAQ5iNhPhPG8wLKglmAq/UgVCWlNbG0mCK07nNdJqKWT9bKvikkRAIMZ3Z5z4BhP6TlHmbpQAZiNswYxmTw0RCCIWZHTIaYDVMyhCxa6JFE7HpwGbcTJE5aOLJYHcYluj6x6yeSOHJ/YnaO2RnlWExOAxNSxs2ZoZvp+4TrHK7P6kasm7glOCezpkuo5WUoBFRIvecFKEKhlYqhsNp3OpOpPBlWhM5YkoQy97rHSrZYcc1UZ3wIpAlsVMUqhUCKpbSMqXMIpKwM+xQ3fVbwjSV6NZMJKWFjRHxAJoMEjw1RAzmdKSmGhmhF13pM5BBLLri6V53NZGtxXatq/tWPHzS34XJkni590sjmJwX9pey+FKTN58Ljv+spq7uqcYGUL1xaP7l5r01X1v2Y5lNZZWH9bN2coWltHcDWEbk2fjk1tc/av6cmbtvEZgzLuJq+lJPaz+vYpX7Qsvbnx27TpbmG3Xz7eRvkUXZWGsBfNNjLwpUXismq6MhjC6vOvmxCNRZQW8ZVrvs4BL7Maq4/7T5X24/yn7Apu7PeihoEkgk+kPyEGU+EMCHO8OrHX/F3/v6/z6evP+Xgeva9gfDAfHpPNJb5YcTfn5i++Zq3H74hZM+rT78k7vfgOoLrSDkyn0bm44jbFUJW77m5ek6clEzVnm/xudAXRZCQ6WRgNzhi6phcj7EWP08bZUVv/HYtGlnvRU0AXtaIrOuFhiopJfAxMc9B5/c0kY1Tct1Uk9aVYDflAZ8tE4ZzhMkL3oOPlpQMMQkBSyhzm2MiGId1hs45jNEkX5MihIRFNIE6R7Lrkf2Jue/JMRJFmJ0hRY+JSj4728SuT/RDpIPCQWgxdrnVYLIqIqWkSk4GI31Z73ldMynrXptX4M7VnWcUvLLqIPTWEo0hJa28nHOA7DSnddG5FHDSnAlB1CKPmhcXYikIUAmCy9qnkAXHwpifQl5ckAm16Moq1STnGJRJ3gjiDOIMyQmEpPu0QehsyZEzukdorWiR0N/j+GGDV4sgv+V4wsZYm1lUjYuTP/Z301ZuflpB/7Fznw7YX3t3gZnfe+TNq0eB4r9lWuTR3LUGYCu8K8H4U0yYC0BdNPLU6Or5db7sxWftBFZNdTsHjyCt1FXSJFagMH7X66yjuCx9os54U+M5FvivILMGFDZ5Pq0/rKo5TVCGugYrTRBLVGFtZxn4xYgWTbyuwaxRc06E79695fa7t7w+ZNgPfPblj/k//O//j+ztHj8eef/NXxAGGDroOwinmf/mH/+3/Hf/9J9z++4NL5/teXUz8Lc//0sOn/0Yt7/Gup5T1DpOOWWO9/cYOohCfnjLTz//imf753yWPeP8wDzPjKeJd2Pmdo68GRPn0PPs5oab5y8xXc/48J44j5UDYzkS20erBgQ0mgGbP5dJ0XdCStSiHdPkQR6IaWZ3GBjCgWgtPmbivGOKljEZjuee8zlwHj3H80jysYTfl0TjGEneY3c7hqHnarfH7DqcsYhAwCMmlZyqgat+x3DjOTx7gT+dMCScFebwgB9HTqeRt9+dOewc+53jOmR2Q6TrHP2u09y0Ss3CTAyB4JUZxZqEMR3inAaQZFXAYgEk3Uc1S6XkZCHaTOiy7mUF3Uv3IZOj8iL6ccYXwK8KZcoZorLBm87iAH9W8mAh4Yyh2/V0zuFMh7GJ5IMytKSAIBhjy35gJvmZeR6RcSZmTWofDh297XC9I4mQYiCHiPeJHAzGBrAzlfnr9ytF+UMHr+UogqTR7KAFDDaC+PKc2kRu5NIGF2WVcN8nlOVSODX4sAJS3j6pedtWa2CtfciL0bDp86Nj7WdahII8GncFnFz7ktHN5gvJurEbZbWVWqtpdYlWoSTbOb+Ys1b4L9V5WxdrvbJs4XhpLy//LSjVgmeu4WqFGqrapjV4YAM9rSWWVwuAy8+Wdoq7uOnDwp6RNXpqAbJUgUt/r7yD+t1WQRGj3BnLbpkYjMk8nCdimLEpcE49f/w3/pSvfvY3cWK4u39DnI8Yc+L27Xf0naXrHP+X/+t/z7/+81/wy6+/5TiOXN8+8OnNwMurH/Gqe0d3OpL7PfchImJwRZLYZLBYPn3xCYcUGD58zXj3azABS8LGxHUcCMEQkmB8It33nMaBbt9DfyCIIc5HwC6FQuu8Z9ZnDkBKBV6qFSbrfNVCner9TPigQhYXlCw3Oo7Bwlkw3hCz5TRlptlznhLHh3vO5zPn8czD6V45EUNx3ZWw8BwCxg30/cB+f6Df7+g6p4Cz7+n7nq5zuN1af8u6jOkFyQFDRLqD0kDRM85wTJHzOTGHxPV1ZLeDmGDoO5zTiPeUIilE0hxJDx4xETEe2/eaLFz5Ga3DiAanSKGuymXxJYSUjQYlVeaWUl1agsUErZ9VPQbOaRmTCozKcSn0e0NsKkJa0VphrtP8wZQyyWoNscFZ3YN3plwL3WOLXqmmctRaaFmQkLTAbEiFWiUT40QwfhmP7dyT+6V/leOHDV6NQFqfi+avpzwybN9bj3xx1hPC7qKN1uKq5252v54yfeSiza0aup627Dt9ZABFaC95SY/63/YSFcAFGCtjVMvc8ZFuPJpDNhbMxVWW/ac1fFpl2JN35/GxWDqyqSGVaRimlkm55OlorCyUQkoj7T8O9Zc3SffQqunYzpsKAlmsqaaJxi9Yiw6unIarZXZpYWxuZXWh5e378zySgoZpSzfw6esf8fnrLwg+cD7dkeOJgZnpeIs3hmQ6/tt//q/59v0996eRUwzc+5HgPR/GxHB3Rz+MsAscQ8LZjqHrkRQhgZWO6xef4N59QO6+Y3r3b7l6aTCdxYtlSAf2DBzcwOl8wk+W2fe47hOsMRpSPVP2Rcs+SnXLZr0/y7ouk3Gpx2mEnkappURxb2l5FBPBZUvAMUZDGnVhhAD354lxDJzPM8e7E+P5yHk88XD6QJy9uusqXVfKECPG9nTdwLDbs9sf6HqlMLq6uWZ/ONDvdgxWOQURi9gO06XCauJ1jTmlnjL9zFxY1vM5I0b3inJWWi1TntAU1RUYfcCfJg2jtw6XwHSdgqTVnLhlD8I5jYSoz12ZG2IqJMWavpCKOzDmQqVbIj6tcxo4ErWGlohG/4kTfEmql5xXxVZYiIrValQaLY2kNMSyz6ZUnXp9iYGEMsnnLiImYXN5SsuemRb6BFJHjb79fY4fNnjVY/XxPBJV3yO6njxWkGoi3lqE+us0nqsQL6DUCOXL5hrZ/D1S/nfsRn58wuP5WTkWi+7bfPnCIlpkcGMe5tW5twZr1HIp2wFUcF/q0j4BglsLqPwU3zvSJH+3Y7xgiP+d1kBubkozwFoccnFDLqer9WSaPao1fyuuwTgXYKW/P6ZhSsFJPdeWexBixAeP8+8J4cQM/I0vfoyzwvl0h49qjdmUyXHGOcvtaeLr21uOs8V211zdHDgd75jnifs58u0xkf2D1n7KlikYZNdhh4H54QNWBGMycvsL7v+Hf0n4zTd0ceSr/+0V5uWBN+cXPDycye5AHF5zfgg6Aoncfvia3bBTbZ0tGXJ7mx7/1cxZ8+kSZRfRyr0R5hncoSfYgdkMPHhHjEIIM8f7M7e3t8znkWk8c7q/ZzydGM9njqd7rXhcObMqk7BkOtdhXUfXDXS2w/UD3bDj2fVzdocrdocDN59+yvXNjYaaW0fuMzZZXOrAn4gkbBYO3TXGReYUGE8nwkOkP0Wu9x4bE+w0L0pyIs2eMJ053t8Tcg9mYP/cMBiHcxqMYZ3gxCKmJxtXrJpEKFZWmgLpfCZPI/iAiWW9igHXaXFMAWeEbudIPpBC0L1LChnx3jKJEEMmew0OMSZga6HMnNS1J4beGqRwQM5WwEBMWg9NItg5kaaRWSYQsF1msA5rLWaoGZgF3GbUjft7Bhz+oMFLLv7YWCrl+Jjh8mRr1WNULYbqOapy+vIrLSixWl2/9dqNYXhZLmTpzRNWVv3Opu1GAC9N5ScE9pODb3vWhnDXRtY9uG2Cr2xagDUEenv9+v9j27goZGUMDXAvTRRgIJUSIM0AhIu4jNbabQC3qvJGms+K1dYmn23mKTetrUEdC44uf5f+1eix6irc7HOl1YJ7dKz2+hr0mBFjCPPIOJ34j/7T/5Rff/2GD+/v+aMf/wTXWU4PD/i375nv3jMfPzDdf8Pnn7zgmCL3pxOn0y3H4xk/ewgZUiTEzLvvbvn0M2VRzxm++3DP4QDO7iBlht2eK3G8/yf/nPz1G+x8ZveF49kf/T2GH/+Ufv83uPnmHZO94ti94P/0izfEEAtjhrqfYsiA3U7rYtw/9RBJsc7XynB6y1Kp1AtZHOJ6DavvD2S3J5iB05x4OD5wPo083N9xevee8/Geh/sPPDx8YJoD8xwg+2Wv0ojVelzWKkt7Ycf3wWO9x8wTdjwRYmDnR3bTmTkmzseRYb9jf7Wj21uSGLIBjCViIBksCWetMmNcZeJ5Zpw9/sOZ5/ue3grJan5ZnDxhnJi8x0dDtg6LoTOiwRJG3Xe1AkYqLv1UFENJAZMmUpwQPNYkeoRsBds5up0qEtZo1eJ+12FjYPaR8ewLPZfBJOiMwTlNiE4xkaMQvEae5qBW1RITWZhB6vOTE4zdBDWgxuueXUTNrNiprmAkQaeKiE3gI/iUmHz4yLPxux0/aPAiP+UUelpYXOrml0DD5nMWYbI6qB6bMcJ2L6q9TsGBBoRap1Yu2CDl5RrB9vR16veaFnJu2nwM5JvhFfSp4N7aBVKvDwtYbYAns2H7kWVQT1tNjztzOdvr1fPmjMdgvroCm29Le0fyMtEizTcWy7UBntZavJgePbVcq71+jUasAJTzcs9rKGhlNK/MGqvVVcG39qNet450vffVoswZxAjD4cBwc8WPf/q3wF7RDW/Z7fccT2fOpzN3390y3X1gOn7gfPeG612HuRp4dr3HukyKM/M0ohJWaaVijFDqYmXgPM4IZ8b+xD5HOE+kMHL+9Rvc6YztEsNNB8NLZHhNf3jJ4XDG4khS9quauUoplnHIel+WzwtINdrgEvsirOVQygylXH/AdI4uCeJA+p0Wj0wQxhO37285HR+4u73l+PYt54c7Hh4+cB6PhKhuu84JxlnEap4RRpQktuswXVdongwS1c0cyUwpgJ81gdd0hX3dk3NkZw50ztYwPaIYkhgtxopgJCO2A5NIEsmx+BpKTmkMGkHqJ6/UTlFj+FKJ8pMSSWhonteU1ujndtqrOxvlc6S4A43V4pLOZJxRCidJBuMiMYApe0055IWqyzgI9akLBYRiLHu3FL9fhlwUoELmPPeONKkFRg6NB0LX+kLLILq2rTEEtBJzTO1T+Fc/ftjgRfGblht7CSStJbUKyJWypmr3wGaP5TLgo4WwR8BwKcMfCe76X16f52odlA4v7C/NOe2+0jqu/LSrrV6rtYwMi4Btx7SIVmn7soLGpT3SYO/aHxrAo2GtWEiFn1ANGtfu2u5jEJRmehaTrFpB28lYvruJs8wKcCmrtrgZ1AWA1S8sikBz+rJ2yBf3osxgYtnbqoEadT6XaspLh0r/NtdeZ3iNNjQEI3z21U/42R//Lb768m9g8oFBek73H/j6F7/kzW++5he//DkxTOQwQ5j5ox9/ys++fM2f/OQr/s//9f+bh/M96RiWaxhj2F1Z3ZcSIQhM80z2kcFHXlztCbdvufvuA+O3d+x2wu66Z//6hruz4/x2Qj78Av/wG+7GyNcnsETqHkiWoEJKBMl2EYhrmkQBbVF38hK5Kkn3X5pitZlScbuslf3VHrcTYrbMw56IxY+BD2++5vbdOx7u7nn/7j3v3/2G6XzCT2cwYKzBOId0B8wwYLuOod9hSyVk2w/sO33f9h0GKTlNCaxjAuYQ8KcHAol+HpnGkWdR2O135MMAWJLpCJ3mUtXSbTkq24VYLb/iug7bGUyXmI6RefKMp4kp6n6e1hcrJMvOYAtVVI4q5AOJZFb2d0RzzbLRasy5FJWsBFUWDUV3RuitBlv0xhByJiVDOI3kEIkxah0zk8EpuKSQSCHix8IPmTNiDV4MtvCCGmtwxpJ3wnw1MMdA9DN4j6GUlnGQrSMaC0mtPLHQdY5kIAaNtPx9jh80eH3M4nlSvD+SxttjkZesgnmxGIrl8sjq+ki/Lo2expO1ptjmx0DU4uJiCbR9pzQm8vQYnzq+Z9wtSK7z1xDgNifWIawVth7j+LKP1EBb8+Fydhtr17bXdniNaKxQueyUrRi4IO2a93XZt8fDr/D+sQy3yxUkF78pAFVD43/LtvNHPlwUqGI5YrRM/N/507/HF59/yeeffsH9u7d89+YNb37zNV//2T/jl7/5De/vbrk7H7m56el3HYMbmINnPiXMPPC/+1//Q/7r/T/ln/3Zv+V863n++sDrF3t+8ukVWIu3AzvXs7Md/TTTj9/x9hcj5v6MHSeuftTz2Z/+mE9+/Ckvv/qSPFwxP4yM53uGm8ThuueLmz3/vnvOh9t77u+PvL83zAFyynSpatxcuHZZAjgEFZQigthKyrvyAqaikTvruNn1BCw+O97NcHd7z/3tHW9+/nPef/s1p/t77h/u8PNZa5eJ4PYHXD/ghoHDs2d0wx7X9/Rdj3Md1jps17MbdgW8ejrTaQ5TuWm1GkDyuo8VRi3PkhLs9juuxgNXzw9gtfgiyZCmieh9oXvSABbJif3+wP4KdrsZ/+DwAc5HzxzR/VIJ+DmQfIIuIU4DPBJo4Iqsemg2BtMP2P2BsL8hJ0M2ATBaMdIkUphIsyvkvurWFFFS3X5wkDviDHlUADNZLaLOCDF7/OyZ7x+IsTxrXQc+Y7oIYaA/DJjesu8dvMwcRTQcPqbi8rTYQaNYI0LE0GPLP6FzqlB18Q/a8toK3vW9/MT79XhKOD1u4fL1AkB5sQEWYfnRW7AaPhe/i4Av2mmW1TNXP93GIKwieDU6VkvjY9feCG65OLmWHVmEeAs328i9ZcaaBrfAt1phm2tcahfSzN1yynqlXPu1uC+LxZVRYjyKC+/R2OsLdZGtaocs/W+nYPlELgFsvcn5QklYrK/qLqwvm2jDzQ2/GOV22RUrLK+wXC2Q5zfP2fU7kg/c3n3H269/xdc//wu+/fpXnI4PQOTVswMvnl8xDB2dUxfOeTpzd/uW17uOP/7x5+Qc+eXPv+XHP/6cl88ODHlmf3VN13dI57BG6CQx4DlPZ3qTGZ7tef2nX/Lib/6Y69cvsbsXRLHkOBPGI/ubHYPbY7oDf/zjA7c3ez7c3rF7d8939zPnMSqno7D8FPtL74qppUXUPSbkknxr1nWwGKnKxmCtI6HgME0jp7s7Hr57x4d3b3m4fc94PipwxYQYg+s6+v2BbtjT73bsrm4Ydsqs3/UFuKzFuI6+H3Bdr+Blu4XyTQNw1KIOxpO9Ev7GFJimEzlHhMQwdNjBYjuNQow+gkQgVn4u5fNLQFKA1iDBpO7CLORSYTvOnug90UOyFrHrE6DKUnlKRMus0Dlkv4OkYf8CmKHTKtREckATs7MhzKK1s6zB2lI1OmdC0DyslLKWjsmQ5kCaZ/2JuZAka6K0jQAW6Ts6Z+nEKi1W32GHjjAPS2Si3XWFJaXcT+vK3mbhvLQG+4fMsLHuHzSuI4EnIw/z4z9k+9/2RCklbKh/VgF0KZ1aJoZVkLcC89I4ztIURpGthdXaVPLxF2xOqi6uRkB+DIjqZy13YTtbG8LZFiSEJZS7wsNlmy2MZfLKjVY/bwFVwFRrqva/9lfW3jxuXbZXl6bXObHmo2WyVOs2b9rLS0uLpNzMez1nnYJ1Htr9r7xYCWUvLKX6am3o0rqWdg7WfDAj0DnL1e4KEzP379/z9u03/PLf/Rv+4s/+OXfv3+CGnqvDnh999pKXN8+0IKMVkkSO0wNvvv13XPXX/Hs/+4Kf/Pgz/sn+n/Gnf+vvcD3s+MX/8K95/vwTzOAY4xlrhcHBwcBJEt3LA9efvuCn/+l/wsurlwxmRzprQAbxSJwTNl9h5EDXHfjy8084vbzi9v6Km53w518/8OZ25C7EZcFUy6syrxujBTTtQiGlmroxyvq+Wbzl/SSOlIWQMw93D9y9fcvtm2/48OZrzsc7YvBoAUaL63r6w4Hd9TN2uz3Dfs/u5kaJirsB1/faB2MxzmG6Gm3YY61b11HSvKUcE1E8wXhSDMR5ZvaBlGZymtn3O/b2in43YF0idkH3/nIsCo0hGcM8RuZO63fFmAkx41PCYLWmXQ6EecZPHd5mgnWYWv4mm23AmAjZGegtcn3Q9R08IknrZ1mrblfvCQI5lLnvdP/QDg76DiNCDp4QvDL4RxCfSdNMPE9kPxcmDpjw+nnImgYzdFoaBqsVufset+t1jVgtGuquBkKSwo6V1B0qqtEYI0XR+EMGr4pdmc2ezuX+SpXvT4h+Nic9/cdv6cRvbXkTkf2o9fL1R5D4vSbdRRc+1vamj40c3YDFxTWf6Iw0fXxyLlfz5NL2Wi7Ycrin8iXT/LXeyGotbVUANYQu7djWcluv9bvkj6jFJBvBIOX9zIbs4iL0Xc9OKS9AViMOt3be9ljZOVqFS5lBjAFrDUPn2FtL8mduv3vDn/+zf8r/+Of/I3/xzTfkmHh52DHsDOwypzRhreX1Jy949eIVicQ8n3n3F7/garfn9bDnf/XHf8x494HzPHF4BlfPHa7vGebE4eYF+1efcri54u/8+IFnzw48e/GMF9d/Qni4J4xn7Pmst8d7rkZhvv0L3qcjb/IDf/qzr7C7HmeEq/e3fC57zGB5d6dlP6RMoDMOU/KNnHO6L1OmyIjQWcEZhxhKBJ9R64gOiZZ7Dw/jxN3xzNe//AXvfvEX3L39hvu7dzqfxmLoGYaBbr+nv3nG4cVLDrsDh2GPu7mmH/a4rsc5UyxoIYvFOac1sMQuNF1SrBzJpUozAWfVasE5fAxQSGzvH+7IFjAWtxsQNIk3nhIy62qIWN59OOInIU6Gh7sT4zSRcqI3grUln8tPzGer7B3K4YTYYmnlYruKwRiD6xzsB2IG2zlyDDgCXT8oKGUYjyeCj8wp4aeEGwZc39EnQXl39fkLMRLmoFbjFJEQkBjKwlcaq5gMPpzBeswcSRhCyGB1n41k6HcDaTZ0Vug6w+5qIKKRrTEEuqKshBjJUZWX3w+6fuDgVdQ6qguIvJaY2EaWtSUs1lemMQUuw+GlITdUQCxaf93jkOZLT6BBjX6r12zFeqvMt5r4ph+ZS6V9czwZgf3IimzbzY3V1Hag9q/p7+a9S+tnO/QW0JZrbbr0BOBcoihQoyEr8Gha1+O9PY0ALBZeCZKofLl1D3Gxfep/7UQ2NdKaZVN6k1cAbvLx2r6ulE+5ANtlYMYySJ482mjJMkZjDbv9nlevPwGB83ji/fu3/Pwv/4LT8YHd4LjeD7x+dcOzZwdePL/B4dj3HQ4YHx5Iolx0k7NI9IQxcPYj0SfEGp69fI33iXE+cxpnDi8+Zdf3mMOOw4tPcSYSnXC+uyWfTjCdMdMdZgzk8Ux4+IAzAesCN53n/i//nDgcGN2OXx0z77PhLqolWC0sEd3YN4U7z0qpX1UtsjpNUoBLdD6stdikG/9TSBzPE3d399x995aH+1vO43lRUo0xWNvjDnu6/YFhf2B/uGa/O7Db7bG7a1zf46zVUh2pejxqRYlSvqZW2M7rel0CtUypvmfUeshZQ8hDikyzx44zV7u+cARqLlQyGkGYBe5nr6H/Aab7E36cSVFTQJwxmE5D0OuaiFG5JiULmES2un9UAhL1mcoJSQlnBBGHs2pN5pyJITDPXnkxYykYaYIGfEwTzgo5x5L3FQvPYhmt0YrRqXAmZgNWLLkUAU2zJ4weMYZTZ9g5LUVkTYfbFRe4VeWgAqm4DmdWlpkcy3P4OyrnHzt+0OC1iFyRxfR6akJ+ZyOmOTFvhF5Z2I31sEjtR3L48ZVk86oAaQNaW4skP/pW+/WPg9b22FhHxWrKPO72U99sh/KRSHPacW5fNUrDgo86f4+xtQHzZjC5XvDRjn/5lVdQzJI3wLX0oXVXtuNZrrnmV1Xl5LENXQGrfZ03PzQUN+3cpuVGtX3J7aRQ7UvnHPvDFa8/+ZSUE8fjA2/fveHXv/k1KUw8vxr44tPnvHrxjJvrK569vIaYcQgmJe4/3OqegusIfc/oJ+Zp5P74nmG4ph9uOLx4xf1395xHz90UefHJj9gNA7bv2D27gdMt0/GO+w9vsfOMCRNMH+juT6TTkfnDO4bdgDs4nhnL7W++4eiueHA3/DxfcbSeyZSIuWIhmAJeYnSfRqsE5yVnsFYVkCVxWCtdG+sIScPRZx84nUbu7+64f/+O0/GBaZ60zRLk4voedzjQ7a/o9wd2u2t2hwPDbo/ZHRQMjSZhp+WGV8ui1FsTs+gfNQdwga9iNQtomHhKZNHqyz5EpmnmoJIfMQZrLBSwSGSOQcurzOcA5xMyzZgQoVNLxTkLzmqNNFRxi7EkxOeoxRzrqq0IlxLZBxwa8NJ1DjGixMEpMc+zVitW80xZOIIgc9YyTVmJmElprYBhjabpZSF4W7wQomVTgpZoySESJ68eEpsxw4B0WpyzG1ypkCBlj67sdxrR/LCke4e5RKHk3y/Y8IcNXu2RG1667zFYfodDaKTx7/oN7cP3vPdUn9bbu/79fdD3fdfmoq0VqLYWZAtg5uLvtr9bUf7xPtTWF/kubT/ycoJcfOev8roK+eXdXGI2W3/sGl//NO6LrEnLC6Do9y/qKq5XXdytQkv3tPQhZxaV40mt4lEvHo0MYLc/cPPsOS9fvebD+1v+8i//kn/+z/8FfQevbg68vt7xk7/5I3b7a/rdnuurZ5yPH/De433AWyBNyHji5TAQgyfgePHJj7H7K4ztiWfPNEZCcuxunnPzxRe8fPmaT16+wjrh9OZrxm9/zdtf/ivccdak3TjSj0dsnDE7w7n3+BQYT5Zf2h9za654MAcerAXpQJxSTtmoeUfG4MRokngFJlZ+SAHEVmyvc5Mx4hDjSMZwd//AuzdvePPrX/HhzTeEEBAxWAy263H9wO76hv75S/b7K64Pz7je3zDs93T7HbkbEEkkEikm4uxJKZBjiQos/3JSrs2cwTpbAkYEazvEWkTKvpyo2026BMqqxRw8fvLqjrOO3bBnTImUAkyZKcN5jtyPJ7rTiSGM7JPHW7jawa6zyKHXvLOSNhC9WoMrG78+ZQr+CWIgjSOA5rH1qiTklPAp4tG8rQgYAtZYkhFG7yFEBTPvcaKcltkJdE61/xw14dx4JCV1yYZICpkwRsCTfGR6GEmjZzjsGK727K56LY7pE+PoF+DqnQJrrRgwe91LDeEPOlS+2RkpErS6fPSErRa+OYoCtibhfgy0Ps6OtyZjPi0wFzHeaO7rR0/3qeEAXq6xvpYnxd+jlmrO0eJ2u6C6YmX7rvJ8oeWt+PBkw3q04e0rg3sDdVLJYKS59mrR1T2PNuG4gkiN+FpchhkWvkE+couWWXl6fi73P5NUvTqtuWB1MDS3VCgWmRa6XPe4EpVGSotQrnO2VZ9Wjb2+pYnQKwlySlHLypMZz2fOD7d8ePc1b3/zl3w2OIbeQme5unrBcNhjnSOGGWcd1nXsri32/oHz8cjxdOL+w3dM54noIzfPnpHOXgsVeiG4Hd3VDc8//wmvPv2c58+ec3V9w3l6UCGc4UN2GJOxRrAS6YYetx/odntuvXAfMrdz4t7uCNaRjENsv0lCtSXZVsSUeVQgqCXvqwUqgBEtk6E7MOrKC0mYA8qi8XDieHfH8f4DIehejBirRTWHvUYV7q8ZDtcMuwPDbqd97nqMdXgiKXhS9MRxJs4TKcyk6aych1UHqX56EQ2ndw7rHK7X9qzrsJ3+Vs5ASza62lLKzNNMv+s0ks9lrOuUuNapBRoxeIxaLz4hIbDvDTFoInBnpVQcFrWKQiYlQYjKMpIsZIPkQI76E2Mg6WKmjxnbCckojRPWFUb4SGeNEjf3rhQnDcQ5kGbP0HcKelIsQIGcDclayBmTM3bfYbMlhaSuyeK/iALee6x3dEGfEWUFgRD84hGJWYuERjRQxUePQYjh/88qKf//4tjYLlWtu1T32bp1Ls0cFZSyOTs/OqH5zkdMqU19pu85FrIBVhxbvpm32Jab35fbRVXcLm8/0WWQwjfYXuDjjHtbAGsAhsthb2L2FuhY9gs389HOpvZX7afFIfJEv+t92YLlU+dtUa1VXvRLC5hWTbYd4wK0q8W5Fq2kKDrlzOoupPzOmyt+5Hhq/7DumWWM1Tvj54njwwfOD++ZTx84HF7SDx39fsdut6frB8QIcRqx1mGcxfQ9YZwYAT977j/cch49IWZyEtIhkmxH9Jnrz1+yf/GKF59/ybPnLznsNYT84XTLNE2M08wRh5FcuPUyXe/ohx6ev+Ld+5n3Z8930ZOMhtvrTwcmapi0aBLrSskElSBWXYVpUawWpg2gRhtmtEjwGDLnOXI6nzmfjkynk+ZxLazrFtsPuN0eNxwYhj39sKPre2zV9IEUPH4eCfPEfDwRxjNxHonnB/w4aTh4rQIpSkTrun6pjNzvrugPJU9st0cGMJUivihpKUXC7Ol6hyYbOw06cSUsXyeBgFFACgnjI3EOWnAype1+VkqkOZagRaPKAblYnBowIoVNPkV9OGKImGhL9Kqy0mMiSKnNJlrLS8iEEDVq0ieMS4UuqrDAiDKCUNyySMYODhHlhgwmk4MqbzbDFCIxKFN+ziXlQYScgtaBw5BsIiUhknXvL0VA1koLf83jBw1eGzddEZj1wbgUJrk5Xy7e/93g5vHxfd97yoV4+cdvu/YFvn7k3O+3DFfL73uuIyxJkE93ahXkq9zfQOoGuC6jKxf4Wyye9Vgzs35bN7d7cdrx+ndB+zKGWjqibqU0nXhkYK/Q/NtWRdnnqvyFue4VrGfnj3x/+8qUtnSenHP0rqezDpMT3/7qL2C856evbvBx5MuXn/LlV19w6FWLzQjZQud2gBDPnvF4xI9n8DPTORGyZcqZf/3zr/nkZ3+b3fNnRDvz0z/9X/D6Rz/h5tUXyDxjcmY6zTy8fc+bX/+ad7/+JXQ91hrN49m9on/1it31c25efMrx//XfMU139F1JbHWFcR00Yq8AlCkM6IKGxZslFSGViarWqiwKRqXdCjlz9Im7cea7u4nb9+95uL1lPB6VdZ1cKjH22P0V3eGa3bPn7PZXDP2A6wd8DoQpwvlMnEbG4wPT+cj9+/ec7z4wTyem8QE/eQ1KQKmKpIKXdXQFwPZXz9g/u9FAkGfPOFw9o98N9Ps9UqwwAabRMuwHjFhsb3GxJ+dE8HuCGQnGko0hSKfh7z7iJ2GeA12I7GMCiSRJpBjxZ0/0WemcEEzsscnRmwgR5ULsHGNUEDrfnzmfz2ANyVh1JQarFbXnyGQmYghAZg6ROWSMcfhc2C8kYaZALU8jpbCkE7C9YRgGAEYXYApIAofBp4mcI/M40o+uBHwYrIU5KitIVx/RrHhqrbKw5D/kJOVNTa4iPVIjRfJG2j3hUpJV216iA/OlEFtO3rwnSylF/UlNyyZDNtqXhQFlY81sOrd8mEufPvLxbwWhxWTLLJNTR74Vs2YZ96NrbGwg/UoFF1nGUKP8SpmLaq2UT0vxh03HWzaR2LQFjauwaX39uLbaZpfxhDaSlXNPNk7Ni8HlMkWrHVRfr71dQbkmhopUwKrv17/z2o0ih3N7uUWx+MicA2KE/e5A3w/EqGHKOUWSCD/66gteffaS65sDxjqy6JoS13M8n5UtPSbuPtzy/v0HvvvultPxzIcpcDcH5jHywvU8++RTfvSzP+LTz3/K4eoFvenINhGmkfPDA3/2z/57/PigxK5WSYbEWuz+wPDsBdk4vvn2G5KfccYgux0SVbiJ1Rh3DcgTKpmsKUCwJaYpEySawFurg0utOU8mRmEMieM5cXc8Mz7cEcOsgRBdR8wKXq7fsxsO7PfXDPsr3DAg1hBSwI8TcZqI08jxw3tOD/dM5xOnhzv8dCZGTwhBy5fUO7K4NYqVOBqMWNzdLcN7teoONzc8f/kJ++sb9s+e0V8dcF2Hcw5ve+UKTKL0U4PWD7PjhNsNdCQGItP5SverUmA2mYCyuvtzILliRZGYvcf7TPCC8x7rDC4YnM3aP6sgSfDMc+R8nCGnJZQ+27KbJ4Zx9CSftKwJiThqFKJYp3XDOouxRi3AUpxSaiCHlAc2KqB11pJd1r9jxmZPmCPeR+L1DtdrYAxG2TQEo0a1LcFJWDqvBWLy71mN8ocNXnwsr6h+mh+Jr8tz1tKuuZXOm+/ki688OnLRHJ/6sEXW9vzvMTXa533rbnp6DOri2l6uifT/KHP9I6CSLczV9hbQb97M26+p4qAT0QQJrkpDG3y3BNxtpnuFvvao0V9tQnoZdZE3W/NqBVTZ9FfaCb9YFHm90Pp2040FrOrZeQWwaulVRv7NnJT+ybbr25kpgogMfhpJfsKQGIaeV69ecHNzxW4YNGE0a2TcHAPTPGml2xC5u3vg9u7I7cOJ8TTzEDKnCKbbc3j+ileffcHnX/2M/fVznBs0Ui1mpmni/v4Db998S98Z9rsBsRrabrqO7qA0S7OP3H74QEqp7Gt1iMmLCw9McQ+yKF8VwGBdm2vx0HaSG3s1Q8yGOWSmOTCeR+bxrKTCxqzPpSkJyf1A15fkY+vIJEKYGY8PzKcH5uMD9+/fcT4emadzacuX3KW6b7mGxC8gBssammclOXZdzzSeSSExjWe8n7lJL0m7gTzscGYghEAISaMtrQadmMLm4bqAix2+20E3QhzItkQk5oz3gZQyomQihJQJKTFFSCHggkNiJIqCihhlkDfWkiUyzjNxjlgj7DqH2fdI1Oi+eZxhIQ/WopUZo3vOUsL7nSN7DyRyKpRdpYJzLiVqpPB5rvcrYXKGHEkJog9Yq2kRxtiNJtd6XLRKhOowv8/xgwYvfRBWAGv1f2CjwX9c/FfJ0mjJRdq2WvUl0OQl1faylaIdk7Hfs/+1aOtNjzcA0byUfHHxRsg2InVjHW3YvduhlpOfkKerptV8Zi4BDqib7VD3MT7C9Cj1OrII/6rINcbh0pkqSkBKlJQsgSbLjl35YpZ6r2Wxyszy+WOwrXPSjk/Pb7/Q3s/WDqugl0tBQ7brI7NQfG1noSYvlYppF/iZS69jSszTSD7fMx1vcSbz+tUzfvzpK66urum6gTke8T4x+8Dtwx1kjUgLp4nffPOO2/sH7s8j93eevLuiP1zx4vUn/NGf/Hv8zT/5u7z65Atl846JOI34aeLD+/d8/atfcn9/y4sXL3D9XoFhPzDsdhxevcJ2Pf50x7u3b8kGBVpjSqKrKQBlCj/hxe039e7XgRukul1LZNy6JIWEIUbLOEVO55Hjwz3T6UQIXj0ZSYHeuI5h2NPt9nS7HW7Qfa44nZkeHnj47g0PH95z/PCeh9N7ovekGMtj3eTo0aQ8oFRIi5KyCJXAPI9wFu7vPvBwd8fh6prnL14g848Zbq7pr68xpuN0PmO6HpcOao2KxViH6zpSivQhEnZ7Up5BPGJncqeBDNM8YzuLyRo2HzP4lBlzJMwzg9PikV1OGJNwCH3X0/WJ2SfGeWK8D1iE2FtuJGvM/ew53x6VYR8QEgkhO0cUhzJeWJzrNLcrBlLUWmC2PFkxZObi8k0hYAJIykhMxcIGkUzwswa0GEtnLLEk76eUtUZbhuih7nV+v1L+248fPHjVI9U8ncX8L0EB+cnT1yOzuja2Zggb0XUhEB+Bxgp9mnAJq2B8otXcvG4F5hM4+Vc4ZNvgdhjr54t6uTF9tgMqRyI3AMIaxHCBtAsYPO6Rfia1va39tL3w9r1cSkXIE7Agj16sVh6smnOr+S3XXtzFzdw/vlV6Rnm/tb5qxeRHQ6hNb6wuWd6vuWUbt2hWN83D7R0//1f/gnffvuXF8wM/+tFXHHa95mvdPfBwf8ccVBuPaSLFhPeJ8xjwOZGMAztwdIkvvviML3/yR/zJv/8f8jf/zt/lxatPQRx+HEkxYkPk/vYDv/7lz/m3//LPMNIhtiMZy26/Z391zX5/wO0OBB8Is1cVwTpMiUwzeQ3GaH82gS5SACxDzqJaujTrX5olmIWYDDE5zlPkeJ45ns5EP5NSRIm3jBaN7HaY/RW232Ntj8kQz2fm0z3nD+/48O1vON/fcj4eCdE3EaEWTbPQh35jRSyxA42GR1Wm6juBcT4RkmecTkzzmWcvX/Ps5WsMjt70WNOxf36D5JJjhSUkIRaWDOl6bNpjCThrcIPDDkpThRWyEYwzZGuIJuOzWskyB4wIvstYIpKjWnrnCT97UgafZlJKymd4ispIH6LmV/lADglrQBJgHdH2zP0IKWNzZrjaKcmwdIRx0vuVspITh6Ss+z5AzEqwawRxA7agiDGmJqmBVeBLWddsTUNIpSSMNcLvll7y8eOHDV6PJEcRDAto5CdcZt9jjW0E+6W1s/14TWKWjSWThUYMr/3Y2C+ytVMavH3Ur6dvb17katsHmmuvlo40fd/aE5dtbuT3IxTNVQKXVlcUr3uGjzTvpcF1RrS3qxW2dncV8vq7WlVVu0jL91u4z7TzzObTzVQsU9QwpHABpln35i5wedHQF0aPJlCjtdo2Y2/cheuY2hGs4w3TyN3dLb/6za+Q6DGuZ391g3GOME9MPhBiYp4DPgRCDhgMKatVZ6xG+yUCu85xc3PDq08+4auvfsLV9Y3y9mUhp0SYJs6nI9+9e8t3797y4cN7rvYD1vXYrteIvWGH6wes7TjdPTBOI7o3pVGEWZRpYusaXYHa1H3bAm51Fuuvjda9vF0sr2SZQmaeA9M0EXNcnjWtVeWW6sfW9SUvKpOmEX86MT08MB7vmaeRGL3es5ow2zwruXaw9H9Jdq9/13V3odTkHIkR5jlz/3C3zMn+cMN+d03YH4iTL9t6am2oYVeuXVg4xDhyzXoTDXIRVyMpFbxyYelIueShxUgiEZNHUiD4mTBFkg+QS+FHMhID4kuEccqIlH0sEibrPnzKkTjOzJ2FlLAkdiXU33QWor5P1CClnCIxZaaYkAiu7lV2Sv9Vx4BQgpoiuZRQQVaOoyUf9ylf+l/x+IGD13os4CAUYZe3+ywUx8XHwD7LJtF5bbd+f0WWnJvSJpvzn4KdJ1T6fAESS//Wcxfj7eLZX5towKm+l4uwqCjbWgEX11z0yoscqNzuIQlIlmasT119jUhpZdIytNzSaVHGthXgjR2j3yldSIt8a4BLGldcDZN8Gj82OnMF+8eMXnkJZc40874oHIv00X9pfV2vupzZmPCboJDLfhVBUwd7Pt7z3Xdv+PlvfslXr19hu4Hh+ga6TGTGx6SRajEyz4ExRHaFjsh1ka7fwdkTcuLlfsfrZ894/cmnfP75j+j6QfcwilCbz0feffsbfvWLv+TNt1/z8HDPi+c3dMOOftizP1wxlJBzZx0PDw8cj0dlyDDViVwKEtZIwQrKNVij3u7NOrq4L+0aLKI1ZoNPjsknxmkFL63SbdUN5xyu69n3NSzeApEwnpiPd5zvbxlPR4KfFyuw7sMtnS24JI0ANWZVSkQsQtkTW5IS2/oLiZQjx9NDYbrwXF1dc7i6YX+4IpxHTO/0zJSLEK+ajAGxgCUGIUUV9rYzmJJzhQjSJXIwhV1DmelTTMTgiWEkh5nZT4Sgbj0RoTcGmyI2BsyUlj2lWZJGvlswuRSgTBF/GpXEd/ZIDDy73mF7oXMdqbckr8+ewakFFhNzBBIkq+5G2/fKqmINRhKSIikpI78UwmUxssgVQRCrwFfTGf66x/+MwEuPSyhZtOjfNk+L5H0MYFVR2AhDVLBXAb3mOUtZ749dXU/1F9Z8q7abqwXFqrp/T3ub0+sVmj2semiv6kP4ux2LPlCukGB1DTWodDn3+YnXtQ/CI0re5a8i/td4zrahpRtVe7uwuhqBtO5L2s21yHkVuBShtUQd5iKe86bTy17XZiQNgMk6tu2cZ9YrCa2nxLmO6+fP+c0v/h339x/46idfcvv2PSEJz5+9An+rSlVMnB5m3rz9wDh5husbBnHEEHn48MB88kiM7Dvhj/7Wn/CTv/OnfPmzP2a/uwJjiClyOh958/Wvefvt1/z8L/6cf/Vnf4YxcPP8GfubG4a9gsFuv9ecqX6gHwameWacJx2bCCJ2uWMlwLDM+zqu5VlYHr5l4sEoS4QG96Tl84QQsuCT5TiOnM4T8+lMSGjBRbFYsdhuhx327Psdw9BjnCHGEX+6Z7y/5eFWA0u01loqQLRd62veWaWEqkqcgowIS+5XLnlpqhQm1mw0dUWezw/4eeRwfc2wu6LrB56/fkXn9iCotRRjKeyoHoaYRUPUzx6ZA+ls6LqeQ6elWcQaXJfpAnQ2Ln0wIricECIiEZMDhISJ0AuauhCTVoEOHtNp2RGbKUlkBklRn+WQ8OczPgb81BFmz37oiWHPPg2Y8vCl8mh0nS1bMF4ZU4wptcU6+qGjGzpSmvDniTBFQsgaGWnAug7TWd23JGkVa4Q0e36f4wcNXq1Wm9laF+2+K2w1vydx7GP4UIGrtL1oZ/Vn0eJqH6owe+RveHx9YXloHhlnrT++gYltW41qmxthsTTZIG5ef/022FrHKIslIpu2y94GLK6hbf7W1pG3frXmXm0HK9TItHV4Uhqv87xYYa0ikdP2SnU+a5rAo5vfuIwzVJZ65WLLG01wrYysF21LobQKziaedemGLNK8ug/r2slNj62xXF3dsL+5xpjEzaFnPgd2V1ccbg5Mdydcr/lGs4+EDMkIxiTm8UgKqViCQTXZ/sAnX/yUV59+wbPnL0CUXDWEyDieufvwng/ffcf7Dx+YpzO73U5rXh2u6IdBXUalhpYtSbipWCCVnHZ1DxcbbBl2u4aLli3V+9E8jHXqGmWxKigpCyHC7APBe42mLNbKYnUZ/UnWqAYvmZwj03hkms6E4BfTbmF8YI0k3Li2pV64ruFmfRTFyNoaKJTX9WjUhVmVp0zidHrgdDpyOJ2UdSINC1chJT9Q15BR9ygwhkTvAy5kpvPEsN9jrQZBLJOStaqylOtbAec0QMZkxxQ8IaWSnpPJolbh5AOSXblnFkwF4lQiRZXdRAteajRhmJVhPsyubA+WgBYoCkHCGoOxlq5zDLse11usM0rIbB3RRyRk/KTKniUX8FRiZGspwJgJv6v2/JHjBw1e63Hpk1ifE9l88ETk3GJWPAVppZWNsr0K9iqUNkK6+W7eCDken1VBZrGQtgH3l/BVe/lI978As40F185Hg2UV6C9h8BGwXYwx1UaLm20B7c31MmscbNVon7JYqvNNK3stGWIFOC730Jbx5fz4bhVBtO5xNlZRox/I5YA/erRJ0Xm5bgWvVfY2ErEIzUe6SHtOo0WICLthz/XzZxx2js5f8e67e66e3bC72jM9aGSddR2T9yQxiLNYG5nHMzkKOTvIHqzB2AMvPvkRz1685nB1o0zzMRK853Q+cXd7y+2HD9zd3UJOOGfo+15dhf2Asw5TytFLAa86XmMq9Y8sUZt1DZlm3bXYtKyvorhUr3yNgtV7LGUtlZL3OeN91NDxsIKXmBJybp2WoHdSmOgzxMA0nZnnkRiD3l7R/CJjiqVYwGDdA14rry1KSwO0YqoiWsBvseTUZWpM4Wwsa2M8nxhPR8bzCT/PdDFCSdzOJcRcCYA1cTdgIGZG77FzYj7NhOuAdU77F7UysRa+FGqSn3O632Szsrm7OWJjQmLGkEiSiCTGEHTcxtJXzkIihojYjMngjBCryzyhbB8haZBHbNa70XWUU9IipsXa2u2HpcClEbQ8jbNgEz6j+2VkCFrF2ViH7Y0GpCTt9+9z/MDBq2jZ9dcikNdspUW3vpT6bRvt76V2Do2F00JPmwS7vPkk1dKqaObysMtFoSjWh/gimXmVjturPbYhWIv+bbv6COwugvIfja2Gmqt5r/2OVDdfucIF9q7a7BoMosJJmbq1qF4RCOW7cTlL26tkrdWlmaQkMpfvfBxnpPlph1Tuerun0QDfwushq/Wg93AFvBVadUxSQ6zbtVBDflfdYx3rxQyvrmUp49ccore//jlX+z3RwvH+xE9/9ke8ev0pIRref3vLw8Md9w/3fPteAyv21sI0YQLMMXEMM+/niB0OPH/+gtMcmOdICoCxhDByOh55+5uvubu75/7hgfsP7zlcXXFz84ybqyturm8K/VSPGzowQiISspT9kETX7xCjSoZIsYSVvA7KjFYnoC1vCFo3SsvYby1zU9a/iIZwQ0dMjqns63nv8dEj0mnwgnUw7JBhwPQdTiCnwu/3cGI6nvDTTEYK+Bq9dll/ktUqWXCsvBCU6DbDUhYFlD2+HnZxlboFcM1KxQgIMet+4nR/y3Q80V/tQXp8zsQCfClrAIfB0ItjzobzHCF5/HnET2fEZmIwnE+BaQp4HxHJWOnAWMQFnFicEYJEpBPwQgweO3msj3QZjsU9mZLuLfXG4sTSu4yNAWsSc8h4FLz7vsPajiyWkNUTUmmfUpoJUaMXjXR0naHvHf2+XxSRnNSNGlPCZyVwDkHJmLsZTN8peEmPmDrPv1WD/N7jBw1e+sAoUzW02v/qiqiyLedVgLVHi2cNHDTFLYswq/V+2vOWe7BCyfKINic+DRUXOPWUWbVUgazmk1x886Kd5vurqGiYHcqEbMbZXHK1NB4vq9y63vJ6Uguvl1DbuvTqGDfRn08Op9Xd23PlI695NMHVwpRi8Ty6jNTxFE28dekWU1IjDGuDaz5QVYagBS59Vx9uljaXDf4LE9KoSUBKkW/ffMuL58+waCiysZbpdOLtr7/h7bv33D0cOY1nrp4/Z2cNhsQ4RuzVFftu4Ga/o395h08Cdsf1s2v63YCxhhSiRqTNM36aGM8nzVkSYbcrfInDsCT7dl2HtcoMb0spk+oldSVEvo5ep7GJOGzmocqlhdeysVxr6AI5k5q2EoaUDSGJhufHSCQp0ayxJcqwx3YDthtwrkPQvLV5HDWRmWppseRymZYE2TTqW66WYFYzqwbllL5uqMXElGVW65SxKfxcZUTwM/N4Yj4fif4Zxmkpk0QmpkxMEYnKOpJESFaIoowaIUViisQUSGLw0WtkqQ8Ya4lGiBZirjlqBZwQkmQlufURE9W156xDuo5UiJXVWkuY4BcFfL8TbCml4waH23e4XUc39IiBlBM5JIL3RK/ghRUkJXUHkomoRZlCwkdNto5euRfrbIc0E/yACQ6XMjEn5W/8g7a8NosncyGHmnM+anahLoG1LZEnGqltSAtqT1wnt6DxlJRuvimrpXh5LJj7EQDctJrzYvmsgLJaCMvIW+BdUX65YgVh/VguelwQQdbdtw1m5JXt4rEdVAXa+tlTe26ZLZBWYHzU3tP4TXVAbomwpGjW69jac7fuPBq6sdLDYmnlhXqcRpFgEXTtmsnVnVonnepebkFd+xNz4ng+cnV1AFHhZvuOeZz48PYd96ezWlLZcvP8NR2asxNsx/7ZJ+yurri6uebw8p7TGBg9HG6u6fc7jLOkGIm+hFT7mWk8E7yn6zr2+70ysHdKQOu6Dtt1OGu1+rGxYGq2nFI9JVYlZr03q9p2CW25matl7orwXfcOdWmlrNGGIUIIWowxlnmSGiJve0zXY3ulZBI0IGL2laNQFvBKWRPKjZTwbWoovO4faU29EgmYeeQSXvf26rpWJbkmYy+1LJu1E6Nnnib8NBKDx6WO6l7PKH+iqdcRyEat0iRqmWn+YNKgjhj1pzCvJyvEpGHzKSsghqzgERKEqGBgY/FiWAuuw/QD3X6HyxmbApIMJgkYGAaH6XvoOmzvcIPF9Q7bdRijSceYTI6R5L1GPFqjxLxR12LOqJI0RyafFLxCCWwpRMCaXlAIfKPm7OWkytrvc/ywwYtm8TQPSBWipVoA8LTVVcOZW6tiaWwDYI3guUSvRrq2lEzbHjYSt/neohQ+javfeyi4pQtw1t9Ctbe2ILTJknpKGD9xjXYcNQCkPtyXRyoPpdl863Fbbe5ZW614CZlvghza/ZRtq9Le9OZt7WeiaNf1g7xq3WUCtr0q62YNlKkMJY018dReG8WiKH3ZeJPavl5En8YQyClzuLrhy69+zDye+cWff+DLTz4njEfevH0DzvHy88/o989wvSPLhHGGL69f89mPfkLfD+QYOZ/vmOeID5nh8JKrm5f0uyvm0wk/T0zjyHg+cX9/R4yRTz75lJevXrLfH9jtDxinAqsbdvTO6d6XMWSnEWzG1NXUzMuT92R9J+dV0dMMYA0WIDdzWl4m0MTaJIwhE+ZACmVti1X+Pecw/YAddtjdnn63Z55HQoyc53PJeXO6nwKamFu4LpctyRyWNUyswNk4h1NeWQ2WYCBKG/r0WFuCJQovXyouQRL44BnHE+PpAT+NmL4jWeWpqDldi+WUNGAHqxZnQoMorBGsAyvKYJF9JBkhpkRIiRANU0qEHPE5c54ipzlwDBl8wqXETjLGdLhhoLu64vDyBpsCxs+k+ViClQz7qx37mz3ZddA5XGdxnaVzylWZOgNBSHiin0hzYs6RyTpMUoaVEDUgKITI8RyIKZJJ6lYcBDEQgvJ1hjng50BUrYP4hwxemimur1MFnFRlWsOGdyFtLzX5y9RXila2rmppLCt9fNvQ+Vb7fHyF0osql+uzIc01v8e6agXmJiNd5OnzL46l5tmF5XHJhpGfCO1ft+FW4GIxCS+/X2ynxe1Sxl2B4nIuH33/sVHVjnCL87n5lmrPyxfLnla1btb5Tw2+N4Sxy3dkeb32ZU1Ozs2DJs152ujKMLqEXUuBYlMtvG1iQEoJYyyvX7/UqmJu4PmnX+D2V8w+MOeeF6+fY4cruuGaV5//mK6z2M5h+z37m2daHTjOdFfPi0YuGNGgizB5xvGspU7GE+8/vEeMYdjt2O93XF1ds9/v2e0P9EN1G/Y4Z4rLUMAJrrNLIvAy/42SlpeIwua2lPtd3a75Yr2Uu7a0lyhMDDEzx4QPnpSSlp83BbjKjziHSndLzBkfA6OfiYslswb+LGpccREb7NpHKWkRUq0DEFN6lTJI2StrXfVGENthrME5DTFKKZIqnxJCyIlxPjNPI3YYkGGgvWiUYmGlBFmIKTOHwPF44vp+QHLSEjTjiMwB6yNYQ7YaETjNCWzESiIkDbmfkuU+CilZuizMoozvV9Yy9B391Q5JESZDHvbkpPl63f4K9+wKGTQ8X5KS6poYlLk+JRyZXSdIJ0iC+Txzvn8gzJ55nkkRgo/MPnKeAxSGEGvA7jplETGdurVJhHkG54qS9/uRG/6gwQt45Ma75O6usmlT+HDbwgXUrCJpEchV+ObNc7tcoHU7bvwIubUOLo5Go2/KTj3ZqycNs1xhtDZSraInzT8ezclT2nOu42kGJK0w1+tuKtk3yNu6/haMai5QBd/ljDwF+WufpXajhDM//q6sHW+IeRsLsy6COkcLiK35PRtj+8JtvKWxydsAw7xGra2Dyau+kqXoQasgXFyRYtV6yiDGsru6AePItge3x/UH3d/pB/ZXNwyD7vWYfke32ymvXOqLxVhGHBN+9vh5wvuZaRoZp5F5ntViEEffDwzDwDDsGHY7XNfRdQ7nnIY9F0sDK0osa80KWpublNf3m7mryqO+lcrjU6GqKhx5AbBEcYXlrO7CFMk5l4KPpT9GQVSsXZk+srpaQwjaqhFyMstrvQlblWh5vttnTVYe0KrvrNvMUsgxZKkjZqzVsh6iBSNjBGS1oGY/E4InxoDN/RpynrO6Mwt4LVZbDEznwHgctARJyjDNGB+xUROdSbmkPeh+U7ZlnxBLlIjHEDAEnXFcFhzCQYyCftZ9vdgP5KBywg69rqNBLVbmGqZVFL2yto0I1gpW65jgvS+uS4rFGfFzYJwjxqnrMUcLWdk3jDWajI0mOluryq39PZl5f9Dgpetsuyhr+OpG4F58abPXtPHLb79R86ZyFXg8IWifktIocLW49VEB3RgltdMVXFZRt4527erKbVG/uwjHZoAXYndz/QoyDQyu1snmm9KesennIpAq9VR9+KG47cpZNRmO2s+89HVxI0q9n7LMiyzgqe/V6s9se3QxpXkJ45aLm7DCWp0rGiu67sc042IxIVitrfq6fistIAUGjYKuY9VzU67h5jrpuVhzMcYlUGN3uGaeRsT27J+/Yk7CgArrlJUeKBuLFcPgStFFOSDRL+2lHBijZ5rPjOeR+/t7jg8PCgZGhW5f2DR2hyuurq4YWoZ2ZxaNODmUHd2axnLPVG/AMk3NKtloY2WFbtZgEd4VwFIW5f5LGmHqY1JCVwTEqSJiVsBwRvfkBC2rkmLCz6HsxynrRyoKiZFSGZgqE8z6bLSeFRqFTlCOQaX1UG5NgZJti7EdtnN0ncMCIc4KbNYSoiEbjSL1syf4gPRRATkkjdxj3e+JUaP5og8cx5EHJ8g0I4c9Mmn+VxeFEAek5Fn5oFWuxRq1yFwkd4nsHLMJkARPRhKYIOwi5KTh6TIY4uGK6imwhx3dbsD1DmOFTCRltSRTBHLJI8ya12adQUzQ4BKvwSNK3Fvyw6aE6Zyu92DJoYdOsK7DmVwC7LICtBGy+/1qovygwWs5cnErlfWYmudscyyC+nsOkY1XbE38VSFbRdpTbVzGeagQ3Z75Mbvo6WNrHX1vv5vPdb9Hrb4n+9mc1773fe0vfX5C+65vm2V+8pq8Td7sKzy+RssW//gMtZrX7/7W+7f57qI+oyJMyhXX+7heuoX2FVgXQb1YaY+7mUsiKZhCzWDQzMyqKenJKSckqtDp+wER4f50i0iiL6U9vvnVr4gpYpzj1evPiUnwc2B+OGJvnmM6Q8iFMy9Egg/k4maLKTLNE+fTifP5xLmwsvtxgpyLW9AioowOfTeUmlh7bD8oZ2DfYYvlFU3CDQPGdVo7rN6PrIJDra5FtVjnJq9z3671lGIBLharq1JDLZV2Y1EgRJNrxWq5jq7r6K2lk/Jjna7xyr239Ed5ANseFT6PRVmj7XG5vyknqi0oIiSjcsAaELFL7pvpOqxzuM7RW4OJgsSS7GENIk4BKYTyOxHmoMz23iuQRd0bmqNHQsDOylx/ugWZZuzosaWeg0NBRXJhCpGkoC6O7Cx2l+gRnj0TnOuZvGeclOXCzYnuOPP6PNPvnVpPh52mEeSMOGXpTSmQkmgIvrMYi0Z75kiIichIwhalwDInTTOxMbOzStDcWegcxFyALXSEGHExY8XgBsjGAr0mWousBOZ/zeMHDV61Ym75S381EYFbsX9pc1yKQNl82jZRLZ3molsA+h40SrQPUn3A2hbWCLnWwln/b/vanLcRzKsVVAV8/dZjEZIXq2a9SrVSHg9njWVpQr5zc24bpVItmPLfeo11DIkKasqOsAGNzCI8lutf9gd5zPfYolpez1tnotZrziArJ19rhy2zswnyyMt4Vib5Cmbt9czGWl0i16qgzHUjdu2TMcpmkQRi2fTOEhjHiZS15IV+TQVV8J55HslASInOaXbUNGpkW3WVnk8PTNPEPE08nB6Y5gkfgiaZZr1fmmPmCTHqtBkNyhBbSGOtJtc6U6reWiGFbaTt43VVPpN67ws85ctzZdUyi0KgwZzV9VmnSnSPyZgyV07LixiLq0nC0ihoue1X08+FV7AA3GJmC+S03uvlaECvWG0r5VjpkzVLGRG1yCwxQY6A2AXg11IrWj8spaQ1tmJxGyadU1POCcHjvTBNjt4WthPjFpg3OSGOQo+YMb1hZzqytSQczs6cR6OsKJESEJYIIWCDRoxaI9jBaXqRKYS/ZfKSVVAxzuKswcwz4jRgJttEtmC6WCJP1SpNwhJ92RtlzciSdT6KFVGLZ2IsWSySksqTvPEd/ZWPHzR48Wjp5RVsFv9PeSAas2jRnC/MgUXDLhkoVcg+5ilcd8aWdi/xULaPrEGWC6u41urLqwdNOy7br7N58wnMXeFu/U4rtusDpOc1ILqijJa4YJ3Nqvmu5z0x06WA1cocXedDsFKFRW4HssxZXmZ1HViuRgqrdmxLiPLqslqB+kJlKZdvwUjWaxffPSYrhU6ZSCkovIDlosE397q4B1ONlGOFwupzrWwMRpZRr9dEWdhhXQM5awi3dT2DG8g5Ms8RExJz0PRsE4QYM5ge0+3x0ZOP98zjCdyuaOaZ8/GeMUQ619F3Pfe3H0q5FM+Hu/fM48jkPd5HTHU/Jng4nun6HVc3UftntAilglcpUCgW62yhOcoaoScgppI1X4QqSZnzlBaBq+q1hktvjkXCq78wJ9F9KslrmLqhlGFxGNHijs5ZurrvZTRfrioTNYRcryfkJGuye/s8rU/M2pfyd41yzaUDNZmZXAJLRDn9xDi6bo9FOQankEi+MKkvQSPafiQRiSXPK2nJkFjy+nKVBpmYAnOA49RDD50TnBWcZARl2nDOYbqMdBk3wM1uYJeEYcic+on7kyWK5mANTjkS5xRwQbBYnGTczmq5kgqgKUE2RDLGCc4oa4btHcZbbD8wFwXDxYTz6hZPWXPMJGuJnEHJNYhG11jKQs4GyYIzHWI1O6ymBfgU+H2OHzR4qSbV/FFfFubvp2MlKrrJ9q3mT2kW3uWxvdyl5tkK2RXUNmAp60vTyNfLsO/tw7VaR7WvVUBvxXrTpYvxXLSI0JYDWa+5jOjSB9o0qIpxAYCiJTcBx+t5zZhq/1enUdN3WedjwVTZzm691+ueVwvHW527uOnZFtKUIuDUsltSCQq+1r6takre/rQWfl7hD0wJvqhKzgJti66xQGW5jrVWhUYsDAlYck6EmHj+8nUJN4bJT8j5SM4Rd31FTh6RQJbAhxBIKTGOI67vmBEeMqQ4L1c92B7bF5aLl6+K9ZVJKXA63tF3lqurPVZELRrbFV7DAuix1PGqIzLrbJvlmckLZdgTS4+F+oYCVNXayoUuCdF9IJElrF2pnWrpkLLfVVyHxjmk0xw0SuJ0FCHVyr25sqE0Skizlts+5hqlI5WIl+aurmdnKHtfUqIee6QfsPsD1giOjJsnpmkmxrSVO1KTlDXQwYSkicgxKIF3WZMASRyBjlMWUjQMxrITU407DbLsM8ZlkET0al51xvDiMDDsOoZpoNv3nO96Oivsekt2MOdECvVeJZxBlZlyD3JKCrpByLPgOg3vx3XQByRlrDEMrsNkQ0zgYyROM0yJPCckRJxYLAq4ptzyoJU1kaizOUdPSpHzNPP7HD9o8IIVvJ6I3t4cVfC3Un7V4y9PrPteTwCJNO9UMMrNQ7HiTblGaydQaGpYtH61VliF9bKvUq2XNW+odTCuE7DCbBu+n5vP1mFvv60BkU+M8WLctSd6qqy1j4TNfDYtb1822NIGR7Tg3E7vZUuSpc1WuJyA5oq5Ab5mzhdlpEmUbm97jTikzkfd3yr7INVKgGJpt7C0ulM3WQBlr6ueqW7Scl7O9F2HdT0ZzYNBNMF2d7hSKqGkFXSD64hJ91kowQ7BT4xlw/98PrHLB0A0Kq2oBymlQuaqRK9XV9clmTcqHVKOhOC5v7/jfDoixmCtwwy2FJnUvWNbouvaZ2VlVs9Fscqb+V6wavmdm5iXfPFTEopzvVO6SJb5ErX0MPq3GK3kjJS2CnjmEnFZv7/Qxm0Um3L3heIxyEtu5nbNbVW6jFb0FmPVjWs6xPZIN5R9IwU3mwV8KGVMCmiVuVmjDdVlmFOTNF0iGWNxq2UMMRmCkjfSFe3VGBCjazKmiD8rw7y1Pfurnv3QY0RIPmJDwBgWKidi1v2rOZAdBdgcxlmdLmfIovROaQ7EKIRC/FzZ452z2CQ40bXmY+ZEJoWgT0qIGKNRib2zC4FwzuB9QJlKRME7K3fj73P8wMErb4UQW/G76MdVK2yFS9X2Fml/sY8FKnw2FsgqHKs5dAmal22IbAHssQ64drpm/Zemaf+vl8wXgrltIhdAvAzEaP9a2mmSdtfHOz9WAgrYyPIfOlcL2MraOdYHf3vtbeDIshtWpnPdNpNWvqxQI+33tq4qqfO2mYnyaesSLE0te8SZBVw2vc11t6aSF6kYknrvmz0xrVGUln5udKPS1jbCcr1m13W4rmcKCkIiBuc6hn4HotbR7fk7zSECur4vtZwSwc8kH/DBcz6fNaEYTWRd9yI1OCLHCDmzPxww06yBFzmSciDlzPHhgeP9Hc519P2uhKXrujVYbNdhnV2iRS/v7gLudbzru83fUFMK6usteBVn3+L2rSu+oGi9j0qFrpZZ041lX6tqRtU1Iatbv6iJ5Surq7tuv22PFb7Ssnak1NrSQpLG9tD1ZU+osKfHVPpiC2tI5TVcFaDECmD6nOu4xBqScUTj8GKZsyVmg63cjJLVfWi0jRwzx+OEST19l9kNO/p9j/QQdoE8O016doa+s8wpEEJkPo5EJwydwaIcjsZp6ZSQFLxijBo8U8L5MQbTK6+lKdG0GfAxMwelHosAMS1q4tB1dK6kWWRNPF9WR8l5/MOmhypHdX8tC3pR75f4oVa+LscF7n3vFTYQVqTU91p7H2m8bmgvut3qA3wslHMlsa2f5wU/NZ/JLC1Vh9XGjFle1f0nNmUINkbTYyPoURvkC1dsXv7bAEwjplhqnsECDM1oygft64vmN528PPIT728F5hKi3ohEUnVTyQImdQ8wkdY+5kQbZfhEF/U9swJobgV8jTIsYxdRQlp112hknbGWfhhIMTFNnkO/04gv6fH7M1c311w9e4bb75kfPFPQUOXz+Z6cNPBjms5M08x5Gun7gdlPxBi4Gnoykbo53neO7JxaZVaw1tB1jg/397hhYNj1pGmP4DCdpd9dMVzd4PZ3cHsLJjVT3sx7A16xgJRUyzVDSuoilMLfmFqwqy/L/cpRhWZlolivJ8u/RVHKjTcAdXCKZA1EqIweabW+agDGRh8tFlsWswYS0dprq4okmKVkjOt6EA2myMX1RkIfWOuWgI3WakcUqGNOxKysE8YI0ndYOZBMRxDLCUGwBClWptEIUZshhExKMz4Evrudyb6ncx4fDc9TUuupVN6WWNIrcuL8MDEeZ+4+HHFW2HWW59c7Xr+CvR3oXa8Jx7MSI58eRi1z4gR6qyBthC4JXe+UaR7hPI4wenI3k7uAWId0Dtv39H2HdMKcJsZRUwZyjrheg1Fy/D4B+tuP/1mAV2vkr4J6DZ+9lLebnC1pHsO6lyFsv7VsRG8bWr7XClhpbIPm/VVP1bY2j4eswQt1/8Vs2sprZ6Vqkds+1r/bGD8uPm1F/bod1YBPY8auXa8ad+2fbCew7htcfDE3/y1ar2yMEaolVZWBpYHNOSzA+/jIzf/bYykmWfbj6j2t+05m0z4ln6yVpGkRWY8F2dJNlihCMeunta9L5AFFc9d8mWHYMc+e7CN2GCBqGYpxmtkd9iRRQbs/XLE7XDHs9oBW0805Y53DOi1GmclMJQk5hpE5B3JWMiLXZXa7AyKGOSS6bgAgxoQrdEBd1zHsdgzDAFm5AnvTayh9v+P1Z1/g54n3X3/drKm6d1L+KkJzBZsK+DX0ve5vpaVQJFACO4r7rrhENaF3peyt7kN1ZepFcqpBHmrB6P6bgpe6Asv+jVF332Larw9Oeam5YpS2K4ehUB3GQl0FLOBoQRxQ9otSImdfrOKo7kBbZ0nHIkl/FmW6MsynZpxWXXhZLESLTxlJmT5m9lnwISOScEZ5H72PHMdEmLzmTJkzgQzWMOfE8TRRg2JSjvgp4KfI2SeMhzGAmEC/D0RkSbaevYJXPM50O4eTjs72xMppKUYtxaKIdRiSsWTrMPseRC25OsnLs5ZrDllCAmSjJVh+n+MHDV55+x8rSGwFdQsylzKwClAend+s8ir4ZAWX9tuLbiYXV38U9MBG8Aur27PtQctovVQPYDXSFgBuBrOOt3lCNzQCTyP2cvZFe9VSra5Ifc9sPlsFeXNNoC2wufJP5FWw5/Vr9dSFgLwZ0Bp2veq+G7iWx8rJ2kjLaLGFdVnGUzf2N1de+ilF6FVmjwVIlxMrOFUAa/rwCMDWvQ3XdczeKxO60yiyhG7k5xTRsk/CMOzo+gFjOxXqJapA+QYtyWRS1DD4FD0pzljUsjPWMPSO/X6PEcvxNLI7HBARvA/0u16Tfp3TYpR9r/sTUftji4Xx4sVrjve3ZGMgR+qELcpYXTzLjaSxNoqgLoDUskwIuURJFNd8TrCQ0+oFWqt9gZXqbkyZnLQ/1fqSMt/RlL4kUcsltevxUu0r7cp6jcd6UtOLJsqRzJJoHmNQMM1NFG3pp2b8Vtez9qC6D03drijuO4NBkim1zRJzzISUMUFdklYKg7tPjHPGzwmRAA+TWr1O8AjncV721iY/a+HSmInl1riUcDYxnD0xQfAJoqhr0UeYgtaOGwzWdWQt8MWidifIKWMLk0e2BrfrSNkse371fi9PYbn3MSaNBP1D5ja8FCZVkD2liT+tuV98U7bflYszPt7o9tO8hFVruxuWB9aHEWTZ92mBDBoAq/9lMEkIDWDmJZnjsv0G8RbhwqP9wbY/l+NrQWS1vpr3NkC6IvpTI11BXjbvXv5x6f59am/vqRv5GMDK+M16T2ugzML6vrnWeh1TLlHpqBBZIhQrU/ly4tPhrEXdZAEuU2qq131GsY6rmz0pCXOYOBwO5JzZDYMKumy0CGU/kDEEHwmzx8+hlElTt2NOahf0fUeKZxgnbm4OmvrgHJ999gqDCkJrDc9ev8IYw3Qe6fuemNUyurl+VuiOBJ+LMA6BGCIvXr3mdLzH9TvSfKbaJCVqfpsqcvE6FeCqT2rlMKzuWFJUhajwDOUci5tr3ROr4GQAI7oXmVIkCkvINWgu0WLbFMZrqfu6hmXxV13CZAWDGn0nRsO9F0+HyGa9GmMR5xBnMZ3Vgox9RwjajxRnsi3Rt52skZkhqOWVM0jCipZBqXO01ojTQAd9XMscxUwIlnnyRCsYIwSvIfkhZOZomNG1MB1H7uZZa/HlzDyFZf/qOEd6q1GJrtOcNEMiJk+SzGAndiWvvipJV/2O3jikH+h2O2wpRpmisnzkGIhzgHPAxYS1WmYnJC15mQvzic0GcXGZdzVW9U7Nf8iWl8qIqo1BUWkujtQIma3IAlaQWYRvG9Onkk+jmcqnG1qZC4sDuWh9fXDrYyCLdpqwyPI0rYwFLbg17i/qA9EibLsjxhpUUS2bOq7lb1E3xqINsRXAiwuxEemyjqPm9qxJviUYvbgQljPXqAFordECgrluwrf5c1Vxr/eh6ddT4Frv/dMARwmUWY9I1dCT7knImqvU+FCX+1/brgZFKuAnpnRSKJZ2s59W6Ijq2KobCvR7uRgbCcP+cIW1HSkqpU7OCdd1iCl1q/qBNI6qZWfPeLwvgQpCDHOZn8Q83nPd7+gHy3W3Y99HfNKAAZON7jsYg+0OmDBhXcfNs2eqTadETpGh0z2VXNYHsHDhdWbgcPWMT778Ed/9u79kDlokkmyQGgGYU1k6Zd3kWNx/9caVF+WcjJCSXxUBWZceQJdhBiYDSKkbL44lYlRES3xoZInWlJK67kQTgTFlXdYgFl2/Fpb1vzjzjaXSRtVrWKr7sARqFKZ9m9UbkgzKRGItgiMx4LMshMtkQZIW2vQpIzFhSxZFCZ4ki9E90LInmKNgjXBloS+ExN3QkY0qIBLA+5LUHjMpFsVIhDlnxhB0fzFnQgjK+J6UgiurYYWr9w3hPoI/BnpJ9CRcTjgROmtKtQFD74Rh6Ih0xJA4H0/Mk1c6qPNEejjTxcAuG66cxafEnDIn7/FHwAnSZXyEEIUQVX50VjYFP/86xw8cvGRxGTylp0sVbu1GC1BF4UZWL6dkaov6vEn7leVa2+s0zcr2k4Xk8hHw6RcWOK2a/qMxLMjDCo/bvm3qJjXRVYswWWXzRjteALlpYy0LsjWlWohu/zfN+7lerJ2Qdrx16sz2rbW1NTJ0CahojsXK045Se7/ISCp2bL+3uAdlVQpqG+v+ZF5aq+7CxQQrHV/VmvXv9ftVESntNpbqooyUN1TAacJpEq29lHIkuw5jOozTqrMxZULw5BhI3mOMKxvyE1m0Aq8JZ4wFZyN2MDhJKqiyhZx1v99ZetchfsJQaKK6HnKCFHC2ABdSK4WwuO5E6PsdLz/5lPd/+YsScKHFBsGu7F4FqVJRzNRlxrJ/lUjLnp0yTkRybOd0e7/qg1HXqFrNeVXuRJYpb5W+nNt7gro7U1VO8+L2X+7NsgbS5s5SVCtdBqLJ0M16qZVTjDFYLDZ3xJTIIW7uvSmFFykJycpIceGczKsVbQX2ThhKkUjbO3Up5oxNWXOl0npvTAVt0ZpeMamrMmSWH1U4NcDCIKVApzLgn72WVwlE+hzprAHcGiXZKp1ohKCfAvM0MZ9HGGdMziXisqRXSCb6gI+ZHAySNWcs5nIrRBCxWPO04vm7Hj9o8GodZkuuyIItrTonbCXmdtKeIHBavtZq+Cr7V40+XwhS2k3/Vuivzw710TBmDbBeXYdrWw0OqRWRVQS0/IGpvRBcjPdSpDfwntd+tbyD2+Py3bzFosZagQuDd+n3oj5swbTOQy0TUq2uXGyhzV5VO4KLkeWLDy96XgVVbs+5OHfFnHU8CzPGqtFgKNr5Ighl+V0tovbasAaHVEGYi8CN80yaJxJaYNCIAkmK0JkMKZLmWfnvphMpRq7cQIgTMXnm8YSViI0TQzqSxxMyWHqnfH8mlfDqNJKyWgbOZt2PsPoc9H2vQjMqs0YqlD8WrXQbYqQTzScadgM//ukf8ef/7J+SzjqepDYJphDULUTDOetqKOAVs9Z3yqSlYGJKiRRDIRNgIRWoqzpK1mjaUqwxSiJJBJMQSdpfV1EMMIIt7tFMppahyYuCt6aArDJjVTDUa2CW5ydRFVpZPI6L1S3qOPUxFCXEYG1HJ0DKxAoRkhGJpCSQAwodatFXKq7argZwJToLfW8R1+OGPeI6snOqUMSECQkk4SmUUiL4rHPmxBEKe0zMGWeVt1CpqArQ1bpsRblKyRBjLPtuWddxyYGZYmD0gW7y9MczGCHExDzNzCclfT4+nODkySbT98poImU8PgTOMRCNweYB163z2InDma4w///1jx80eFX2+Kqpwwpg0giOj7FlrEKuhE4XMtut4C/ntC+ekOsLNpYHqGpXGbM8Y5WtoBXlVeSZ2n9g/WZpizqmvBHEtRuaXLjp5cYy+tj4lxMffZyffPvSiGoQ+fFnzR/r/lF54+LkOqXy6Mvbc9bZYtEGt7ej3r8VlFfMqn+VSKkCJPaRFbh2wEjhyyh6kCALmWiFfGPs8o16he34130UW1yCu2GH957ZB0LK7LqODMwh4ccSICIwn86krK7haZgYz/ek4Nk5Rzx+i8HjerBdjzNlbWXLrnOItXRpxsm17meRtDCiGPw8I9fKWYi1CGkRWqodJ1KOSvOVEr3r+PT1j9jf3DCdz8RjKKHvRm9uLkBUBGm1vLRGl1/2poKPy55KjTo0Rh+eOkdtpKEplYera80m3ZcxKP+elkcpYdep5BgVRUSfu7SaSOVIBY2UWaW5c0thV73hi2qZcgnS0A+TlOjYpO47hMJzqDwk1kair8FCGbKCFqX0idiulAlRq7UWpDWdwXSCGwz7w8Dheo91PVjHeQrKpjEHrMCM4MUgXpXggAaRzHG1nJFSwgQhlDVUi6au6R8ZK5HeZG6M4coJzhlc58DA6D3pKEy+fD8l4uxJ55E0TqTzmTBlJgtjFubJwqDA3B+E86wVA1LWACJjhM4YrDHYzl0+KX/l4wcNXnpsIwxbaSbAQr25uMXYCs/yZ+Md0vcWwVwXYQuGj9FrYZzQi1Ev1kYFSr3O5so0AQ+yQO1HDIWasqgPzdJ32TAFbOGkMU/yZb+fwOLWcGxNwlyBo9ipSwcvQELYzE/z1UdzsVxPaEBdAegSPJe9o03vH8eOLZ2QJ/pQ3Xnlx1A0xaqdL/1Zx7do3ssV13nQrkpzncu71n62HsY6xmkillIoJz/plazDmMLMHSOn460+9FbwzMTpRAqeeRa6OGIkkpLTjXZqPlXGOAVlU+8VKshSUJdlDL4kRtu1aGhdgiVIotagkqwaeed6Dtc3nO7umB9O+tkSIq8gUSMFa82qFEvZj6yuxhibaMWkuZOpJWddog4zUvZmF4dm6d9y70Q0/6nkQJWELp3r1u1b/cX1Pm600CIX6v1e1n5VgqRsncmijOq5pvBU6muMCmZbxpZiqg/CAhKUOaGAs+576Rq05eJGpFAwOfa901IlriPnrJUEktAlg00Wl8FHSr0zVUo0GCQr32CxpGsKiAJuCZSpMi5rbbBeYG+Fm05wvcX2jtkardw8ec5jKPcjYWPEzhN5npGoe2wBYTaGkAVrHKa3DL2ht7ovl3FYp3W9RLTYqXXdmu/31zx+0OCVRX2suboLlrvCVlA30XC52ZN5Shw+tiR0l1Uu3gW2+0N5DQqvp9ZIoho5uHDEtQ+K6ANK1fBqf1mFYZMptfS3Gns66owt85GzhnEsD4xeZDPexZor362gJxfIvvy1QZL2j7xOrZhlj6DeiU30ZBWkRfuTBqBkM9ZHusXSa9nsp33skGWfpN3nFBEoeVbV71/dRiuAsa6bqqGDJr1eDGoN4tleewG/9R3qrk5K6sYxznK+HfHec3N9xd3tHWIsV9c3mMESg2c8n7j98JbdoWcYBqYpIX6C4Lk/zXz+LOKs4JUtVoltkyFGcGZlOsilz9YIY4r4oGUu5nlGpMd0dr3JFSySgkeKQecpC5INL1695nx3x8Ob70qidwGvFGkTuhNadDD4Uqolx7LHxbJnWSM3k60s/7ks3qhRiClR2PV0TyWXfK4yvRYpTOWOy31PynqqK32Zf2mS4YWluHa7p6PAqa5gpHIaKrcEIoXlXgWwKdGH2MLFmJXPT2xoLlIXS8nrEgUwyRknhk4MDoGQMcnQYRmM5dA5ut5hB6fFHlMiJkOPpQcCwhQCprd4Y8nZ4LyUPcdy6TrftfAlukdnFoaSREdmZ+DaCS8GoRssdtdzh+UhJM5z5DQGTIq4nDnkxFWeIQZsUFCLyTJHQ8Bh3YDbdRy6jnlImAA+GLo+61aJMRjRWnR/0KzyHxN0LZQtrrbGAHnqeCpYYj1WCtvLR2RtrwLMotaUByU3X5cFnCq/YQto6l54DKlt0G4LrbWUW6aE39IARwkTTmQkx6KNyvpMLZp2a1XkR0C3wmarHOSSQ8MChNv+NvJQQJlAVsraJ+f6wtR6pFLkOtLK3v64DZ3iGkhRNegypJKfo9WETXEbCmYJeS/STJqMsOWeNYEaRWExRkEytbkqufldLNKUq+tKVFj7wPH9t5jk6STijyO3332Lc5a9eckUd/hxZL4/Mh2PhPGBuTN8+vzAdLpjnifGlDnFawbncF2nmrEH04Fkw85oocSzOJjOmBRJ5obTPKmrKztSDJB7jHGkMLFYXuX+LIpGndeU+PyLn/Dw/gO/TP+mrJsMOZbcNLUsYvTEmMu+VlwohlLOlWpE10diXTmiDBea3WswxYjKBhKaI2VMpnPKdm7EPH74NwohSyBXslnvfa5ei6rU1We33PdlXa7raomaLZaeE4Ntfzqn7j+rxMpiTCmaqft1dc3HMj86jYkqJzprlC8Qi0mQSlFHP3rSHMk2kp1VUBRBjBYSxSXckHnWZXzfE6zlQGI3WEY/c54nzlNimiMxaHHJun6Dn8v6VwXAOWEwhl1nuNrtGHY9djdAZwjnwPk0c3ucSN7ToSVZehtxJDoDs9EgnXMSRmvphgF72GH3Azch0ofM+aw1n41oyobtesQYwvwHTsxbtcsFWhZhnFfvXf1s862njwUcHqHiY4ukjfJbrZmL7zYuwUaa0z59rVFTI63a3Zx6+uqavAS4vFhhKKYU37c0QmmZlUdjXsMqWASvvl5tokVjbWdg6a80LVfgaKdtnZ1Umq9W3nYk2/tymYitb67f2Ixkc14LbbIwmAhFi16Aq1qAq5CiCr5FkK1WYgUyStmOJW/sCfdHq2RUbSH6GVLE7Q19b/A+8u7bb0nTiMkdJpwQm5lO99x+eI/JsLeZgwFHz5hz2cIRfASbBJtMKfIYICd6N2ivs+6hdSgh6uwDKQQN6HCmqeq8cu8teVUCtSJ5Ku8ZyVxd3bDbH8AKKWZIyvCRUliYMWJQGqBqWVX34+qu2rp6haq8lRQKU2LJl0AZrbzbGXAul+KQei3JSgG1Sf+oC1XqOmjXo7BQvVUrrgIa1R3eWHHt+isBDyr41UWngR0KbCmuz33Lp6l5Z+V3rmFWat2mpNeu+WkhJmYfGWfdD5XOkIPmTEmWkkYh0Gu4/mEwhK4jihBJWCcMs9DNGUSZVqJP61ZJ81AKQmeV47DvtEqy6Symc9je0Q8WlzLWa9qF3hMDfUkHyapEGqupH0kMQbSIZ3aWbtczhAA+EEIizSUYKyspsFghpD/oUHkWAVorLwhQq8S0xtb3W1YseR6rdbE8P481PC4Uv1zv52WQwJNXKv254Eu8PH/RCmnysjKV5aId3DbwNheew3b3LC2ftRdtgWOJ4i8vqvCVZsEvoNDOy/Ja0wJWa6W95NpOrj1svrsBuwIal8C1uPiaS69HPbmFGBVgC/O9mA1wiaxABpQoUqkTuuYMoQ+tlL4pAFZrXl29pLQoSqqf5GZ8dX4zOUwQZwazZ+h7zjFy++ZrXL+n64QujeQI4+mWD++/5dOXL3jeZW56Qy6ErCGp5ZhE6ymRLTkHQkqkEOjtoFZGhnkOdGIwRjX6HAJGtNCgKwUnde8pFbLY1SIRo2CiYfF65/eHa4b9Ads5wjiRUyTnSEyBGu4dSq5RjT5UJY8CYLIua9HcwAog1a2bjSEbDYxR9yA4A72B3mZsoVaLdV8sJY18zOpebFY6NaBnWUvC4pEgNWCXV0VMZPVigM5jXaVWaoFIUwoypoW2alM3T2RZ7zVQSYl606pcp7VIpcmZbLTI6BwD4+yZvIfZkIwCY86lKGffYyxgDJaOZG1ZCxHnMpOHbk5abytl/ByVPqrMgSlr2BphcJbdYOk7LYGCs0hnMb2l33V0IWFdwFgNbe+M4PYdRJAUkSgYG8niwFiigWCFZAXbO7qiZMw2MeWk3hoxdKWgZ3DfL5N/2/GDBq/NkVfxtciM8kZL6XSJQ61ob9vaQsk2KOSRni0VMKW4AdeN5rRpQTaSWsr7K09cCS8RsxCK1gdYshbWMzktfOeyXGVtb+lSyZgnCpXzbWuRlPNy+73i6mrGKs3POrnrfJisUZpWWK6TpbEEUQgwm3bScpOWc1ZlezOvtXOLy/FyoE1fMywBCO19qhaXEVuKOJZrZ1k2tc2iNmsEnf5lWQp1NmSI1VJd9luLZh1TxrRRAbbouNlyfdURzAlJiU9f7Hk4J04PI998+4b/4O//lBcvHLsdvHnzjuO7bzm//5Y//tsveHZzQzfs+S4OvP/FW44PR7767JovPv+cvt8xz4aTnwlemQ8MD3Qd2B5Ob48Mn3yBcwMEr4zyVi0u47Qkb0gRsvIdpqhFLZ3RUGZjLUaU3HU+negO11w9e8nzL37Mr2//JSmodp/CXJRHDbNfwIr1JtRco81qM5ZsnJ4bIhmLHzNzCFoXK+m9Um5YYd8ZstOIwZTBxISJ6lasqmB9phYlKktRKps1W9ZbBuWRzGWfr0R5LopbKzUKqz7Gajl7IyQCJC3a6fqByc+E5EsgkNPyLZTnIqP5V6oJkEtJkVyKU0Yn+OKGTz4yPIzsQmbwmQmwpqOzDvqObuhwzhGM1YhEIIQJa2b2MXEVM7vecDcIeyeaH2Y1MhERiAYnhuve8eracegNQyfQdeA6rOvou45dnzjsEtf7Hb2ohXZ9bZHJkOZAmiJdnHHOwTBA32v4fkz4aVJlKQRc1LywHIVoIV+jFqytGx9/veMHDV4bb00xZbduMZW26qpKi1ZJ1bJFNmdfgtbafrOI86rJtd+t0XIiLFFXS5OsuVzru4sNsmqIta3mqa+gUQNDBNZIp+V9/aR602nAo+1jYZJbzq9j2kLa+uhSXCOmef8y6q86ZtYRyfLhFlDWo/GqrNPaWEBbcGqh9ZFB8zRwC1SGhconWMtoVJISaX63Vbb1vspykWpxmWWtrCEA7fqoGnWSVGiJLEIsgizy7v0tn70+8PrFa/b9nn/95/8jv/nmDTc3e16+eMmzmwNk+PVv/h2Ekb/51Quev3yOdHtOyfKbX71hOh8ZHHz15WdYNzAnwzkqaOq+SiClrtAVRe4/vGd3uEa6ns46QtCigs57og+l/wlBLad6Y+qmv9JQaQ2wlJQ/cb/b8fkXX/KLP/sXxBiQEohQwStR3YTrzVm969IoKOsdM5LI4UyIE/PxTJpGCEFBNUGIQRkjQkBJcaVU5S1kVcVtCFy41osiI6t6uhJfK+vK5oGubuMlVL6sgGqh14dxOUeDXVJMdE4ZOJRGStR6LEpOSGph6fOaVE5Vl2r5aYO1UkrMc0CMVm32Jbw858LMHnUha501LTmSSgawydBhud4JJIPgMMGqW9AWazpoIsGh77m6tgxO6CykLAQr+LJtYEXoreUwdDhrGXrL/sqReyFOntnOYAXnOrpBi2Faa5CcCKeREHSdpdGTZ0/KhpwtMURMZ1eF8K95/KDBq4LJLCJr+QABAABJREFUpV311JTU0GZ9LZtzqsX2ROvb9i5wcfu+UPeK1EXQ2BWZ9TOgRiZqI1JCgllArG6UL/8tD387Pq1AW69dNca1fVnHvXQxbwH/0Ui1rRqltUYZ59rVCwuMZbyLZluEf0bYJEBvbkrNg1uvvJ5ZQNlsDC80yGXdodiSDj/xcokkLJGFVFchRYJlxNQIyLZvK/hW4Kob+ev+nizzrnsu6xdyCcbRPpdNfAPjdGK3f8WLl6+Y3t/z9bdvePvdO7769Bk3NzcMQ894mri7vePldcdXn7+mP1xzTh0PU+DNm/eQAlc3B16+eMFDEHyAWQMYV+ypSkBUtvlpmuj9hNvtiMGDJOZ5IgSPVT0fK6rY1XFWFgwo4fclTyvGwDD0vHr9yVIws7rrFxdh87PO4qr4LKpPnTJRRgbSRAoZP92T/UiOiWxqIm3CF8JYZ+KitVd7qpZfqUGLNUCoLIBNPld9tUTWVpNfIBtRUDKy9jWvbmakrtMSOKQRJaQKjiVsnjLiWIA8pkhMNVClmaHiWq3Pu1me/0QMgeA1fSI6nV+TItFHoolKjWWd9illUojkoG1ZhJ1zsDcYcViveWjWqRsy+ghZ2HUd+ytHZ/UepDkSjMoZW6KCO6NFK13n6HvLbt+rlWgtPqMh+l3HMOwYhh5nBXImTjN+ngk+kCbd+1KflCXFQEruo1yrv+vxgwav1mJYLRAeY1pFmk0y18dnbguFv9tRH8rajypkizigLthVRJcFW3T05YJSHkREH5y8xrrlKtjLo2EWt2ELsxe9/istELn4rdeKrNNXXZna1wqQJTx/oViqBTFlmYn6yeXV8nLGMoOLY+/REMqEtqOV7RmbW1strpqYXBfH2uvS83baGsRcAG9tsADUehnTZKcnUqmXljAmIAjOOg7Djusffc6rmz1mPvF/+3/8Y/7dr79l2A38h3//79JfXfFwPPGrb97w7Grgk89e8+Krn3EbD3x3N/Hu/Zlv3r7j7/3p3+DLLz4hIZzHyJzUPRVCxrkd+91QSsbvwDievbhBTGaeR+Ro8eNIzJnTOPLsxWtc3wGZoesRo25EbEdIiew9zvuS/pDJzEQPfdfzyevPS/i78nPmEi2oVnPegPnyIjd3qy73YsF0JtO5ERdn8O/Ip7Py8XUdWEeIkSkkzmNgbyK9qFUbcyBlrQxtSmFDdQAnHUs51vWUl7xP7YdFTNHQqk96UZpWLg7d53KIdWrdBq8uMafBC0lg8kFzrIwydcwxYlLE5kCYA8mH4hqlKFYZTQ5XlyUJOqO8giaBixHxkWwiBCHjSUR8jJhhIA495spju4GchTDNzGcPOWAksrvZ4XrHvheedQPWqvUlvdPgnZixGHY7h5BJKXKKEyGDD4l+DhATvQjXe1eIiB1XXQfOMttOg4VsZNd3XO0Hrq96MoGYAufzzHQ8kyYPUyDPATEO2RnyNJOd/cNm2FiOCla5hbP1MFLdBKvY0t+XNtj2WIXrFgOW79cN5ycjO1gsKqEme+YlMdbUqK5cLbZVMm85QRotWOpDxZKIKYVSScOM82LVLUED9XqU2kRZ54nSpxbkL621dsSr6DHLnlh97iGvz39rQDUAt4LuOrNpAaLVbbqC/brZ3oJcarV2YPEI1SjEqm2X5FJj1qjCyuOwnLPM8prgugRqNH7FVQaXfZtC10VhgFgANWtyKFldWVk6Ptze8258wz/4j35KT+T2/ZF//m9+wWdfveQnX33Gq09e8Js3H/ju7Xd8ePuGP/rjn3J1dUNIHWOY+c2vf8PXX7/l1YvnPLs+4Kzl7Xd3BHrAaomMFJhiJKRIJwkfZ1zfsxuGUsk2MY5nTucTKQvGRU7He4a4o+sc0RgtDCiJrh9KhVtNlNYaU+CDB8kYO7AbrsB2YDR6MeS0Wm3oTcnSTOuiODSrulFCrIGbg8P1iVfPLQ/vE3nW/KCVcKOApFFlzXiPNCanLF6F6hwva2jxWqzrLJf7rN0SJRSs97GsmaqEUaL86o9BIGk166HvESfgNGBBl0bC+4DMvkRfzsRxInuvQRpG1CJv7FPtdsZm6HKiS5HOlzntelXuSg2w+cEjkyf1HTZH0pBIIkx+ZJ4DpIiVxO4AXW8ZnCOaVMDL4vYDvkaETsrYoeVXNGcwx4TMudw7nYvrve6xGSOkGJGclYOxdySEoXcMnbLsJxEkqjLrfSCOM2acIWTEFJfp+UAE5r+SefD4+IGDV25+byP96lGXoaGEGS/WQi57E3l57/uus3W3bZ1dG8ugiuPqwpAGvIoFuAjpTPPQNZF9F9darZP6YQWu5qzcQM6jvraPbn17peHZWDQZctPwsh2wtNKYtVJ7tto/0pzyaG/rok/LnMoKUlXO1ZD6zZeXCzT3cPmoAM3iBpLFDSOyGE3LHCxzkddLpM11mgvXNbOAHouVsR13ISUqikPX95wfHjge73h1M/Dh7R2//NUbjuPIJ6+e8+Xnn5Aw3N/ecby/J+fIi5cvcf2eEIW7D7e8f/ee2/cf+OKPPmffdzhjuJ8mkin8eYiWMEkZHyNaK9EiXQTTq8sveFIOum+VBYJnnkasNThnNBcoq8IT4xoFm1OkZhNqnlLE2KxcfsNA9p4cIov/ZzXJL5cWFXzW9+v8aRL/0BtM53h23TF04H3Gh7jmyCELt2CWvDA+SLN/vahH7TJnDWnSZbGsYnQvS+9VroSGAtX80r3PGqWqhSit0SjNyg5hjEGcFmPM9ZkUIYegNdb8SPJzqZOWl6i/DZl2XWa50mBFdXxGqyViWCszp+hLYFBGfFeCXkTz60IhAZZcEpSVscPYkl9lLZ1zWKP33Ht1B6ekLPU+aKoDMWkCdWd1jVhNE8gZQlDAk5xxThPjnTVlv0+jEhMJyvlJNy2RmJEEYpQaP5Xoyt/n+IGDV30I1CqhANQSaVTQQwWjOnhysWxW0V1NhIuF3bgh8+X1yk99Vtc8n0aYZdaHJrM4+FqrYg3DXZ/0lqVd22oeyMzygKgBlVmcIjluBO4KZuv4WtfcKvTXOK1cgesCSFbgkkVArw0U92chMdWHvjzYrONZ26B5tQIxLLpuHYC67Mpel54gyjvXTHdNm6vWlCmFAjMKXpWCpx1LK1zX/uT1GpdRce31K3jWryxMHCXcO6ey8iwvn78kTUfG+5kvnl/x//wnf8b//b/5F3zyyYE//dt/gy8//5xf/+o93/zFL0h4Xn3+it3+OdlZxnniz//1z/nmm7dM08jrq47rzjFYizUdk/fE5MlJGKdATBEkM+yvcF1PP+yYwgRB97NSygy7KyyGEAJ+OtP3HbDD+1nLy2Pp+h0ikJwjhoHUuWZ913B6uHn2gjsfmMOxndXN0ahALCke7dQXhU4Edr1lsIZPX13xzc098zxxPM30V+vUW2dK+bricku5PFulXZFCXV2f5fr0rFZOe2/r3taSz1bW0JIdJjTgZRW03aCkt0bACqbrcP0OnMOnCFGwzpKnE3Eamc8PhOkEaAXstTfVakTbESnsIhkbIxBIQLCahNwVGqwctXJzlgypJwaIxmjATqHWMjmToiocOUa1elJhpvcJK+qFSsBUqjKfJ89x9IQpkENibweG3tL3DmcFScpNOY2TFhOwBtutQVEpl8hJCxZDbxxjcXmFnHEpgyRsSoifF/fr73P8oMGreqXTkmOxVdUzqiFoMK0eW2G8Pki/LQ9s02hpSJpnYrv5GBdLyQJtbanFeipWX43ArUKPEtWWlr2uNf+ofei31o9K0kXry42ldjmuxXKpoHY57gbAyyuzvJGpewmLWyblds/7EnJVsMvjq1xcbjM79YOqVSfWaKw2GLFGlYmoMDGyJriqwKn9qqkLzfxtinheCjbZTBMkzcZcGiiaqIiWuZcatCGajyMgFv7Nv/of+PKzG/7X//A/4J/89/+WX3/zjq43/L0/+RlpCrz9zTecP7yhE0+33/HJ69d0hwN3x5E3b+74t7/5mr1L/PjzK7764iXHhyPvbo9M0pFcTyQyTWdsb3HZYbNwP53JD5BjoN91ysyeYDydyQG6fmDY7zUHyEfmyWOtXROVU9QQeTSp2ltXxt+VAo7Kjfj5l1+R5pnvjudmX0wXs6mMKtIqVywLpH0GDdCJxWbV9D/75CU/+iqQ+MDx+IGcZsgBERW6ysGRSUSt41WfhcpMY6Ddr7oMClgeh9biWQJ59E0r0mzRNXAj1ers6XY93eEK2/XgOhBblLiEM4ZkhWwzGM2FyyQNuvj/kPdvsZZkaX4f9vvWWhGx9z63zJNZlVXV1dWX6Z6eIac5FCmSEAmZI1EkRUESYNmQDfpBD3oQIEMAIQmCBb2MXihYMEQB5KsACpIFPlk0JEPiBbIl0pRkznBmqJnmTE93V9c1s7Lydm5774hYa31++NaKiH0ya2Z6ipZd7iicynP2juuKtb7r//t/vpuu67B6qyCOphND6mFNKBUjtM2jPadHUW9HTYqzdnQuYc4cBLIrYyQWpkxKjpGoGVeMHCeQcqbfj1xve7bDyPVu4Go3GhIxC/lEcV6Mo3DlGXvLLwYFibmkPZQ0VtBJZNAVvnWGQGwCoW1JbSQGowgTZ4XQvlGcy/ic+Dzbj1zi/N/9d/8d/9w/98/x1ltvISL81b/6Vw++V1V+/ud/nrfeeov1es3P/dzP8Wu/9msH+/R9z7/2r/1r3L9/n6OjI/75f/6f58MPP/xcD7Kco3rrd1tIty3BQ8E1K7SFRjr498BULx7X7avY75NuKZ/PkAWdDjzw4PT29VkcUxSFltDjS9dl0qKyOLcurjX93H6G5ZNPtFSzaLHDrB5qOloPnnS6jVlxLStq5oHSxePbxzWUOhsO0y4Hyk4XQuTQ8JhRYIf/TghDmVFjh77UfDVZPM1yTKZzl4tO56qCrp6xNESsscl6NVEhj1vaAKfHR3z//Udc3WxpW8/x8Rlp7NnfXJLGLaujFSdnx5ydntD3A5eXlzx78pRxv+V45bl/d4N4xz5ldkO0Nhgl/6aaaVrjwXPeMybLtQgQx2S5C1XiOEyfO5FC0BsZx7EUFhs3ofHvAWWfmKzmCvFUBCpi3ZebtjOd9JK3unwXi2k2/buY+8yGpXeeo/WGO3dPOT3dsF55yBHRhMeUiudWxEI5CJnXPPPkdYtMzR8nQNHi3qYw8jLGLSzmzG1t63A+EJoOF9pCDLw4Z1HWPgR80+CbykihJQxbWfft/h2CFyGUDs0SgoUgQylALmFB9Q71zrggvSktkaL8akG3K4TM3s053vJOY4z040C/39Pv94z73rpz9wNxPxB3PQwJSXmqawwFQNK2ntA4Y2Yp3aKnFjhjLGCRvZ1rjJYX84U+q22Q1lq74D3i7Ty+KZ7b59h+5KNvbm742Z/9Wf7SX/pLr/z+3//3/33+g//gP+Av/aW/xN/9u3+XN954gz/5J/8kV1dX0z5/7s/9Of7z//w/56/8lb/C3/7bf5vr62v+2X/2nzXW6R9hq0K64MkQ3JxwVxZIvaUqqego6horntGM2ztUOXk+siZaF5W9mYl3s+yXD66WF4Sl0y4wnVemH6lfzHVHygF1z7KVwaFCuq1w8+KTXP6ux1QyGg7GAZkBF9NPGT1jvqn4qNqdqWxV9y3uQtDSI6jezazSiuw7UIIc7Ac1m68Uhuz6bSFnRYsyKa0qLNZ+SPvkam3XsomgCAujfDE35vd5KKmKcjICRGrfLqmfVxRjfa4qEFUhZo5XwqoxJfAP3v2Qy+0Nq1WDb48Z+x3D9jnkkbP7dzh/8Bp379zlxfMLPvn4IR9/9D6NDtw/aXnr/jHbmNgr9AhJxISaeEBYr6xwVYMnpQHnoWkb9vuenApDxjjineC9WLFsTAzjyH7f0w99WXtSuPdM2uespNEY6FUcKRdFI3ByfEbbdNOQTeOimalT9sL7PQhTzLaS5VEKQ4l3nvXqiHv373J+74zT0xXkAUlGChucdRw3fJw7sKbqmjVvbuYM9dR5bTVQ5U1ygH7UErYrH9cV7NTC4fWeXQlbOx9oivJyLuBwC+ookz2h7WjXK8Kqm4pxNSVmZvnSEkbtedrg8V2DrFq0bcltS247tOuQJqCNJwVHaj05eNR7xLnClmK+rhct79jhnUxANdQAN/uhZ3uzZXezZb/dMfY9aejJfY/uB0JMtEAX7PjgHU1wrJpA03hCI7i2JI+x0onYj4y7nv12y74osDRGEMF3wZ5/tYKmQZsAIRSvzLpEf57tRw4b/pk/82f4M3/mz7zyO1XlP/wP/0P+nX/n3+Ff+Bf+BQD+4//4P+bBgwf8Z//Zf8a/8q/8K1xcXPAf/Uf/Ef/Jf/Kf8E/9U/8UAP/pf/qf8uUvf5m/+Tf/Jn/6T//p3/G9HDggUxuEV2wLk/sz9lj4C/U8r9rzFg7v4JKmWOYo3Yzdm83Cw/uY+pFNFmC9riHW8rzWF+edwSNVWJiyriwaCwtyeasTsvA2I8fhjvUuDv3Swx2Wx09PWQ66jd6sRc5VuNyGwd/essqkY1Tm5z4cg/kefG2JIdYSfWpNzwzYmN4NhydZok3ruz/Iu03W9HxRkfIcUizhGMu1LXcibeD5s2f88Pu/yT/zT/4vaBr46ONP+crbb/Hex494/OKGLl2x6jyyOibmzFtvvoXzLc+eXfP02QseP37Kp4+e8s2v3ueNt15nfXbOvgehofEOkYbdbkRzJriGvOsLY7rDZUfuMzvd8b13f5O3v/xVjk9O8G0wCiDJjPtrxiES2hXdekO3OqJbCeLawj4+B1orkazznhQzoglphNOTM9quIxWFMa2Y6Zfl4ij71OZVOn9ajRYQnDOlcP+1I+Lg2F3suf7eY7rgaBtjcM++hiStrb0F5PJsWNblI8JEXV2UBECikhXMObjp3U53Vd95NYuL0aKgNSTojKVimnNNA0gpOA40wUMAlSPG62tUI+O+t5xPMYbGnBhzonNKI9B1Ld57aFZQqKjEN+CCtRFxDpcaC906ZVQlj6PdeE6mvIKn7briLQVc40koQ2lLs+0HQn0iEbwas/xalEYsb+68oU1jMtaW2GckKV5LaLMJ1uomjcZYTzIAhmZCyrikSBB8aPAtSJstl6ZKj7BKgvf+pYzGj7p9Pr/t1vbuu+/y6NEj/tSf+lPTZ13X8cf/+B/n7/ydvwPAL/7iLzKO48E+b731Fj/zMz8z7fOjbctwxfzJ8q85dCQLL0FuKZPZ25kX3yzUYIGOOzQiD641Lww5OIct7OLNTXF6ihN1y6uqPogsQofF09LFOaZjcinUXHhjksvP9Bkl7KiTZ7fAGy6s16UVffh8r8oXlNtkipWqTl7bEub+ypcj83Mt733yd8tHNbwzCZVJORVrtzIb1L9Ln6pJ4S2myNzAdHFXByFBnXKU88/y+5JjKw8vlOJSMRaEYYi0TeCrX3mTEKx+ZoiwvbnGaeZ41XJ2dmSejHqOT+/Sdh39MPDB+w+53u1QVY66htfu3WG1OSK5zpr6OY96R9Q0ARJiysaIkRVSYuW9NSF0maPNxizufW9s5ECOA8P+hjhsGfotu5trUrRQT9ZEKh2Ccy61W8zjZoaPklKiW60ITTN5pS8nNucJVecSU5j45TlUp4aIY73ecHJ6wt3zO5yenrA52tCuOyONDQHng4FypERaFvNabDBsrUihfyrvW0qoS3MqbVwW77fcvizu2T53xSBa8GLWdT6ZfzIZUNXjdKVnVbva0KxW+NAgIuZ9pbnTQC6eZ4qx8C0KhIDrOsKqo101hLYhtC2hbXFtC20gB2cdr2NaHMsUInXOGkBWlKGII6vQj0o/JPohMgxWexZyYi2ZlWQ6l2mlcEamzDAaO30cEjma6y3l3OLD9JyhaxDnDCKfMykaaESLLM3OkUQYFVOIETR/Pu31DxWw8ejRIwAePHhw8PmDBw947733pn3atuXu3bsv7VOPv731fU/f99Pfl5eXQJ1gh9uhNq4S60AfTd9IqbE6sBpFFotpGXQsR9V9yolryG8OxhWbrSwqcKikSSBLdVHKDWlJYpk1z6zgKsRgurEDfTddn2kCWGGm1rj+IvQ4tQJWRWoVJsYIN6HAqkdXY+X1mktW01t+WAWG1PGYnu2Vzl+1aHX50fSPTK9K0ULlNKMj5/8mX0lqaLCECn0JCLmiuKrChjmfUd/P8j6FSRFUKPah9ywzurIKJwpeNWerCarNDr1nv7tms2r4ia98DScjQ4SsnufPnxNEuHN6zMmdE7a7PSllTk/v4QPc3Gz54Q8+RI82BCfcOz3i/O4ZsjpikJaYI+qt5irGEedbNGeGGGlXncV2Y+So9QSv4DPnd++wj8q+77lz5xQ0E4vyohNjrhgjR8dnhQIqGqVPytZY0TlquxAwgZjU2M+btiU0Ld4HlmUXKrfe8cLok8VcXs6luT7QwB5tt+bkJHLn3jl37+6Q9QntZkNoGgK1CNiXRBbzGqlzPANSYOBFmE+haq1NMNVYKqYV8LKxKgtlZCHqyoxZjNnFupxnqgUrxUHwls9qVmti3zNsrw39p0Khxydh9FHjMBpLhm/IIRDazoqgg2PMHi/BGj2GVFuCEVVxOeNE8VN4HCpS15WIgHeWr4wJ9mMmJKXyrzYpEnKik0xwkF0Gp0hOpGjKy7uERvtbkxrit3QQd77BhfJZ8CSBmBU/ZGLKpEL2XHO0EWGICg7S56TY+P8I2vA2cq+2n/6ttt9qn3/v3/v3+Hf/3X/35WOYJ06trYHbQvPWAQdf1omYJ+VWOdrmhL1ysBZ/m21KSR1cqkrJaQUBs1Vbl80EtJgKkMu3eYbLTx6DlGLhCeXlLL82KacpDX6Q0FbNk3C2oEiela/cupdFvde8qCuqcanUlg+r0363vvhtvps3U1FuAl3MocayOGtvrgkoUR9IZl46/a1fmiyU1i2NNd/f0puoOZGDe7EcoBZKIFJiuLlAt4lnsuP3f/vbXG/3bLef8PDTC372p7/Mt77+Fh998px1Ezg9bpEmcHH5jCfPnvPk+SVht+P+acebX30d/IqoDWPy3Owy3WmDb4KF31LhM9SR0K0gJnzqOV11XKfI5U2PS0LXbex7J4UsdSQg7Hc7khshRFQTaIIcScPeci1NwMuarKXGS42nMefMmCKdQNt2bDbH9NfXE+DDhqrMoimSMHPE2P9fVmBzqZijcYHV+piT+69z722Ptmv86ojWN0ZiGw315kthr5/WRVlb2bxnL3maC4qFGWtvsRoSnRCqdTYWr1oW794VuLwx4YNkLGw2RvBqhlNarDOxY7z3+K5lfXpmjPv93nj+UMgZEUMgxpzYxUA3GqJQgrOCaB+QEib13hOCpwkNaRByFGsI2QS8CC5AiomokMdM0yjeKy5RoPCJ633ixU3CpWjNPjVx6jJrlFaVtQhZjO4qjT3j3hspgF+RkpJiYtiNxgovZqjE4HFOCv1Uw5AjYzJKqNrHTBVrHCqOrDCiBx0AfrfbP9Sw4RtvvAHwkgf1+PHjyRt74403GIaB58+ff+Y+t7d/+9/+t7m4uJh+Pvjgg/nL2fCdl8JtwbUEN9wy/M17WOY2dOmwTdewf28JuRqCKiEzQ+ZR2h3MYTmFBbuFnVAP/rOrLUOCtQW6W8Qqa7LXkD7WD6gSoS6BIQe08ItHniAoVcjkhTUsugixKVqjQVVhMA9hBXS4GuZTnSEiwoHMf2nT5Ts4fE9CyTDoYsjrbotxX4bwKqJMMIFRY/muLLB5GOy/hbpdno1qg9dkfg0lQVnQzGCFZZhxGjRVvCSCEzZHLW+8fUZ7dMq+j3z66FPeeO0OJ2dnJoRFCE2DhEAm8ejRUx4/vmA7Kqcrx/npivN7d2g357hmjYinXa9pQ4fH0W974mhec9uuuXNyzNnRik3rcd7uP0e4utnRrlYcHR8zDoP1/crJ6nW2O0iZddcxjvadPVrGuYyTOZaQcylezodzyTlPCE0JxVW0IosQeNl3Ucx9OBeYjK+pC3NOBjJRRxs6jk7PWB0d06w6QtuaErGbsh8t3Zzrz3Tu4hknXSivW1EWYA7Na+HOZJrAc7qheIeCtR+hhB9TIqdoiqzEyEQWx4uj8Q3dakO3OWZ1dILvWlzwSIm41FhCzDAk+4lZGEvRcIy1gDgRC72UhUvNiHEhFGBECz5YaDAmdvuR3W5kvxvZDpGbMXM9Zi6GxPM+8aJPXAyZi0G5HJTrMRPVzFnLtZmRqTkTk5IQkjhTRv2A7vbobjf1b0tZEV+MR6kNOHNBr2oJX5oxmjSTSg7u82z/UJXX1772Nd544w3+xt/4G9NnwzDw3/63/y1/9I/+UQD+4B/8gzRNc7DPw4cP+dVf/dVpn9tb13Wcnp4e/Bxsn+Ec6eJn/mVeRPNHJUxQnf9qZU/5m1dt5WQ6K5/DG9GJnfvgh1tMh1qVX1UETDubklicdz7JYQ5sobyWFztUrIcL/EC/6+xDHeYjqmCeD5pyAtO4LCzeeuhkcS+eZRrjxWUP3tv8vVsIDaiKhwnFNgEpFjkIU2QLxXkb3aHzuQ4cxsWzzp7YLYSiLPkjl0Kt5josCrRqHI7EatXw2hv3UHFcXW15+umnvPngnM3REVkCbdsS2hWu6Yg58fT5Fc8vbhizcnbUcnq8YrXe4LoTkBZVwTctwXvjqk+ZOIzklGnbluPVik3b0jWh1DQaoe2YoenWrFYrhr4nRYPFj0NkHAYDfHhfhEwyb6AMlkg1huw91QaTk21XQ1O+0mEtfuqwS/XeDybMYr3UAT005bQwsHsxwd92K8v5hGaut8sZ1VpDtVBKCwVmzDYzLF0P3mKdA7dSA3L7l2X+c4FUrmusjJ1ma4w5raEyn4L4Ulu3oTs6IXQdPgQDbeicp7LcV/2pyksZR1NeMWWGmIuisChxVjHofgiEtkG8JyEMMbPvI7t9ZLcf2Q6J3ZjYRVNg12PmKirXI1yNcBWVqwi9Ckk84o1ZQwpU0cimBSsFUWQcoO/R/R4dLdxp5RQ156bVKqYicJ135UdIqv9QlNePHDa8vr7me9/73vT3u+++yy//8i9zfn7OO++8w5/7c3+OP//n/zzf/OY3+eY3v8mf//N/ns1mw5/9s38WgLOzM/7lf/lf5t/4N/4N7t27x/n5Of/mv/lv8u1vf3tCH/5ON1EmbjtXBOFnWfyv+nxZuOz08Lvln4ce2UKZ1FldleLkfMxhkqWVp3Wf8nJr2K845+ZeM1/mcKnNCsJyDCUvVKxF1XkfFS2JahPEmuvke8XD1RUvh6pl8nz08BC3+H2pYHIZQ1kefGsM3cERr9oOFc7So9Zi/npkbnEibs57TZ7W/HDLMFY1FCbBwvxeD/E1/qVxWM6rA+WLWdpBYNO2fOXBHb7767/ESWp587Vv80u/9Pf47m/8Gu+/+5v8L//X/wyXV3sun1/z4Js/QXt0Qka5evGIpzeZmyFz3GXeevs+Iax4dp3RI8/NMND3PR6lbYXQCHfuHvHxex8TnPDmg7uEJKAtIkc8HXfWsLJteOP1r3F09y5o5vLmAmJv5LB9ouk6Uhq5urnmzXsPUDUWepwjlYR6yImYrE4opUwIhXxXMykNaLbuzWiaFP1kYy0Hqky9lyO5BSkrlF5T9mnKmZyBHPH1xYtHXYNrM04z3FxDHNEUyzopsPxKA1bme6rKbDJcy/mcTKi/qrwOIwZlsihoCVE0lDIM7wiNL7yRimZr9KlOSsSklBuorYnQrOg2gsOKwYfdFnZbhu2l3avaXSRgVOsGE8l4LwQvaHAFjJJISS0/lkZWTmnbhk4amrbFtYqmgX6XGGLGjyOyi+xy4nIbud5F9hFyGRRB6LPSIVw7x5gDp77hpFvRHTeoBBBfGiibqkgeQtqhw0CKkYGB3HbkVcfYWf+3VgQaQ2VWpYxiRkkGjSUfmf5nVl6/8Au/wD/xT/wT09//+r/+rwPwL/1L/xJ/+S//Zf6tf+vfYrfb8a/+q/8qz58/54/8kT/CX//rf52Tk5PpmL/wF/4CIQT+xX/xX2S32/En/sSf4C//5b9sMNHf1VaElN7+hBmGPk3eeRNuLbIDT6BO6/L7wiLTSZotV6lO+1W0XXHfJkWk1QItnkmtP6lKavIYqqU4eU9S/l7eaFVgt1V2hYaUyxewuiGt5MAhYTpaCnhl8SwyC4Gci3I9UFXTBcziWg7iZyioSd+LnWspZKaxq4cuhJkux9rJdI6FYzSP31JS1euqHsDul3dUw4QT6GTx3NM9C0wFYpPmrrk1EE3sbnZ851c/5vd/+2foVmt+4zu/yoe/+RvE7RX37p3x6MklOVro7b2HjwnuBd45fMiM2bE5OeLtN07pk+fZ5Q3Pn+3YyTPazTHdes3pyrNebWi8MO4uuX/kWTWO14+Em6tLnl3c8Pj5Je3917i4fMaw3fH7vvJ1nHMMfc+4662FvBj34hhHvEC36kgpTiAXF1rGMRLGZM0ry/x2lVlFZy9Ns70T5xyV7WhaB5TSgenFzNvSEJDFJ9O7LB5NjAMpVaNOkLbFB09KwfrApIikiEsJKcJVhIlxwubO/M58ATdpaWk/t7pZTiALHrvi1S8baApm2oSCLNXO40oROMIUwpc8e2uarawihBbWsDk9wzlvwBFx5JTJKeJymGoYx6TscsI5pfUZT1tq4RNDtMaOaRxZeSX4apwqQx8Zx0xKMCYr4VDn2A0ju11kHIpSFWtHkorsSZgn5DM49ax9y/G6nSI61LYtcaTREZ97NPdITOhYEKoC/d7RBaOT6pqGrA7je64efEITjDGRx8TYD7cX5I+0/cjK6+d+7ucO4tm3NxHh53/+5/n5n//5z9xntVrxF//iX+Qv/sW/+KNe/qVtWewIOvH+Vb9lGQ6YBPAkgLQUMi/un6WQncNFsjjXlB1bXIuFvzVpGan3UWthdPKuFk9ADfEZN54enGL57+RcLZ+uICZFa7x+ccdT+M6sSJm17mQl17V9W65PinQxDvUZXrXNpFELcbQ4cR2n+q0Vey9wXksAyNJQmOTK/J6WIcNlGJHyLpdCdDkXfvtt8d6q1JXF1adJMD+NMQIqMY08e/Gcb3zrHZIqP/zhB2wvX+CdcnrnlBeXWyMnVchty3DzHBQ2pxvGMRHE061WXG8Tl5d7Xry45Plu4M69e5zLOZvTE9aNFaT2jaPZBNaN47jzvPf4hovrG652ezY3W+Jo7Vi81AaG0SiHCrlnxrojI0LTtuScSBHLETXGqJFrs8TyuHVca2mG9fkqua06TzgsMZhf/sIIXIz2pDpKaGmhciYlqVlRdSDJFIrz1qFYoaBJ5rxwyX2qc68wRKsSLu9tEXJG53m2zGfKQnFVQ69yZToRY4xwis+Q1GH43TzlhFEbI4sQgA8t7WpDGkeGvrE1WFwTiRlSRpMRX/WAK+GktiljIkocIkM/EseREaUNA1kri/vIMGbGlOmjmnJ0cNNH9mOyzs3F6zR5kqeGnJrhJsMmw6jW6FI1GWQ+poJmHXE54jUxedwKaEJzJMZEU1GZ3urjUFAHtYIhFSsnp2TFzJ9j+0JzG0IRza8w9JeKq4rF20Ja5nV3MNkny/7W/tPJpiUGNW49/V2VVdkvL66d6/e3FZjOxbzLD3XKQC8Fcf12UofTAJinVkM3RZFILhpKF//e9pOWQmpWHpNf5oSJgX9S1OX4haJ5aZhufaqv+O32+No3hT1BFswjYmEe773VrRQuw6m8ZmFQ1JepL523PNVBOLG+68ULX04AhOnNTEOtxeuw3xuPtcZYNeSYub654ocPf8j5ClanZ8TQ8eiTS7aXW8iJbx53vP/uD7m82rE5O6fRSPaOTyL0+5E87ojDDU8/eUarA/fayDvfuo9rFbyyfv0OQxMJInTdio8fP2SflXbd8vDXv8ODt7/Ma199m5tnTwnHa1IaCRKIsbdEekoIpgSabk1Oych5FQgWTsyaCqP5jIodJ2XCFErMaYEyRI3Lc/pZhBFrKbJUOPfs9RR1ML0De58KavBsxOEI9r49iLdAuxVmv/q95dkOXCxmM/LqZ64WhdR7OtRV82mLgVTJv2sd4MSoIRYDGaMxtBuvYCrCOhQGD4O9d92GOAyE5mY2KFOCPhvTBY7eO/aUejYc3vILRM3c9AN9H4lDQmJkiMr6emCz2pU5L2TxpjDETv1sFxkLqALnZ0agbGwoqUzlq5hZDYnTfSRlCFhRekxqtWTjAENPQcYTPKxdQBqh94VDVksuLiWLkogjuFKMnkFLMXmOkbjb8Xm2L7Ty0sUvC8Pp0PFZ/PsjnXsx8V91fJ3neuuTKTQ43WG9uQzqyIWqKWchSOE2w+ppckErzUS4UMzlSQhM15TKhGBfpPqN2n65hr3EwoiCs7CEyJzHmjT6K/2uxeaKIgFVmZ70cMDKaZZWbDn9QTB4iveVS7uFyhY/H7T8eFHPteQuXMKcDybDNGLT3ZenmgXX5HliaDoRBT9BdkpoeB7f6Xfm8Q/e8fqdu7z7G79OjAM//dO/l7i/Rvuet197wGt31lzdJD55tufxw09AHW3bcb0duLzZM8bIm68d8+LRJ7y42PHp9ciX3rzHveOW8/U5XSP8zLe+wTe+9g7r+8fsYmBIQkzKyf2G3c0N3334gg8fPefuvXPe+fKXSDdXnL9+n9N753z8vd9k1Z8VmDImQNUYEVyzJoswDj2b9YZhGIlDxK16hjjSpogAKY1lFhuDSbXdUoqW7/N2zglpuwj31vewfBez6ZOn8LpnXrRKRfRZy3hyAn84RRVH8g1CwOFtIpcyETNmax7L0I8HHh01FLiYX/V/1cuuSlAcIqZEnLObsN8L40XGagOdFDiqGTVRk13DASREba4GcezEY00wPV5qs1S7nzRmsiRSk8jOg1NSGVfNRvo8JqWP2RTlmNiPFv5tg+3nnKdpW+6cHIODqIkhmQLSStlTF5+ad2dISk8isR8iF9d7djct6xaCiHV1jgPkERFI6hFp8W3DyfpkqvnTsAIVhqTEGK3OzUPo5nkzhXpyYej4HNsXWnlVSb2Yd7dl2OQavJzvYMprvPLMkzU57yP1RBMCsCy4W3L8cKVVMGxBZVH/Lu1bsFj5VDCti9Da7OhQF159NEuK6+I2Zh9qsoMPHLfpRLNwqR/VcOL09WcrM6n2dA3RLL+d/pivs7TA56TG4fevMi8WeL6pnuu2AjsM6yy8r1soE7n9RLL4ub3p5DsffFiHaI5Q2c03QLtqabLjZNXQ783TPdocsTo65uLqBdfPL8g50rYrVusWJ57z81O8g3e+dI/dp58gGnE+8+DBHR7cOeL8KNCsGt548z537t1lK4n9tmcYoVkdsRuFZ9cjHzx6xtX1ls3JKQnP8Z1zmm5lPZqGnjj0uNBYfqMgUh3gvXHykRN5HIykdYzkGElxJMU4h+6QAyOjTplakvBZ26yOXvaB58gAh3OC5e7OeAG9t5yWs/xOLh7G5HZPNuKsJkV1SZY5GSE1MnGY51qs9YUVrMs5NidDS8GyO5iDtfi5No2c5ZBORARTmNtVlOr8d51bqpWlZzaw6nW1eDVZS0mBWq0fqlb4qw7nM63CZhURZ8g+zcaYQS5gkpJiyEvBVYRFVksnGlBHFjlmY1zJIRB9QMXTeCEcbWjbFtqGKA1pSKQhMoxWSqABXCj5w1xKDHT21j/P9gVXXvPcqwsqT3PstkX+8nY7fPQ72nRefPUCy2MPCzSFKQRWtMwU7tM8EYACE+LIiS3OZevymhPLWhgyKGXVWhKh00IsCmyp8BSm8NvC8DrQuNNY1dxb/VOm5zkI39UrSlFfh7piusS8DyzzU0svpj5lvnWc6HyMKwt9RhUulVe9u6JMq0IvJ3up1OFVf9T7yaCuHp+na5aRLENmfxvZqiP2O+4/uI93mTbt2KkgvmW9aRB/zM3lQx5/8D6rTcvJyREnxyccrzbc/9Y7nJ52fOMrX+bDX/8uN53w5tmGf+T3fYPzu3dYdx33H7/L2fnryNERV0+f8uyTJ+RBef3tDR8/3fLRR8/57vc+Yru95nq35fl1z53X30KcZ3t1Y2I6WVYjFQYNCkN7CI3xQsbMOG4Zh5ExqymuYSCOfaGKMsJeV8mJdXJEC/JuslCoUHk3jf3LucZlubLJS0EdzIzsbjKm1HkjgvUdrmnBe+tnVQEArnpReQoHKlIooux91nupntjSpq1h8GUsYVp15d1XtTYpXXvw4oFZ4fLk0Gier1HJpcuYSIUzO8dE9uycPYNzqBdw3thlxC8QtcXwlaqnXQHLRCz6YqCKMZaOAaokUXb7Pc4Z7tGlhEsGsnHFUE4qZBXUS4V0Td3HpTSS1BJF8SGQHcbl6Rz7kK01Shtojk/o1g1NFxiysL3eMYyJYUxW7JwzrgHvSpeCmEilJvDzqq8vvPKaNr3178F3rxDUy11/RCtgZg74newHcEizAyasI9anB2BqdyJY7x5MaEgB3lZrsxYyHzyAFO9Glx/WbJsUpZSnZVqt5tkBUWaxwuLcumDdsL9NHS/CiNVarc6bTE9c6UxveV2/5YAd7GNGac1HzCEWcQslxkKxwiRcpxxiubdbDt/8jHLwSymqFBMSB/OiWsn2e+s9QZT3P/qAo7NT7pyuuXNnxV4sRxFc4IcffMT17gV37nr+6f/tn6VrNwTX0LYbNnfWQOLi0SPuPnide28+4Cd++if56jd+AlEY+h3ef4koDc9vdnz0nd+kSQPiHB9+NPJLv/ybPH56wcX1lnWj5GHP/uo57s47jMM14/7a2BkKXM/7IqxwZAJjGhnTwJhHcszEMvY5jZAixEQcI2MaSzFyIJRpHHOeisFNALvZM62jOUUuQFwB/FTI9HKmFfsOLyUVK1Y7pYKKxzUN2qygXSHduqRwDaXmSosNNOOzL3WMc22X6a/DOeeYva9F6GFxM/NcrHuqJjT7UtNVvB/nyIUiSVDEG5Guk0iuBYflfKmESLWCq2Ses+o8GgLaOqQ18mHxQnC+dD72tE1jCj7be6xe2lRHamYvCQz9mJTr7WAhTQSXjRGlEZM3e1WGDBmPw+HF0TgIreN45Tkpysi31h05k2nwls8KiaG36/quIXRr/KohrD2kyNg7RifkIRIBjdb20s5jxe6Wr3SM/G7R5bZ94ZVXlT3mssOtuTp7SmXnw6/1MITAUuHMAn46/rOuvzjfUiBahHEhPSvknbqoZPYy6l1ozS2VRTHtz8SNuMzpqVRPq3ph5TIHKmeWLLPvtxyP5QLm8JnrI1WnaVJZ1aOahdRSQc2e1u2zU0KjYqinl0a0nK9CmYuAnFux3w4dzsPLwppb6NNXvK9ZqVYPWaSE1SbBlg9DhPXYrKhzNN7Recd18Kxb6EKmHxOhaUj7kZtnl7z55pu8dveMvP8mX/7GzyA0aBKyZI5PNuQ4cPHkkne+8ZO0refNd97Gr4+JYySNCuGY3dUVVy9eoNsXuPWGKI6PHj3n2eUl292OlJSuFY43LWd3j9EUSSVcaDVbhujSXDlQxOrAUMY0sut3oA7nguWvKmuF5qnPVx1N8z6XIZ9qHL2sAuoRn/UeXmVQ1pU6GSbO4ZoWbTqkWVkLEh9xocE3nXkv5RgzzlxRnMu3vTA0C1ipygyp0OSFNVNn88zDOB8vlAhf9coWkZuXWvPMK3pax4bgKEi9srBExBRw6XFFsG4FUtrXtN4RvIFTsuQSYSwqs7yb2cNVnAitE3PwyrU7gbWDViBoJojSOzFqrcYTgqP1gdAKR13DujWF2TQOsw1GUJnQyhoakwBthwSjfdIspEHRaFrVZUMo5igMqvjgpvCoeE/2Fn78PNsXXnl9lvd0oISWXyxWWQUWTDmn25J04QkcciwcnndSSQcLcvFnVTBOpqJqA1RU674ANTRPN+WKJpLi/Silo/C0npQa7pqFMJP1OYuQeaAOn8DOr+JmFo+qrIuFeHurZ6zh/+UXtQ6o7nDADfkZDlf1NOs5pxdQrXGZhdjEGD/lGqZXczDWNRwxh41vzYNbmnlCD2oVz27yYKtho6qlNqe8Ew9tCKwbY8s4XjvWjXK9T4TVinE3sn1xwe//A7+P06MT1s0aHnyNYZ8Z+8ieG6RdIboHv+IrP/V76LrA8cmGqz4xKAzSknPL9cU1l59+go83SHPKkBo+ePgp19s9Q4qA46j1nB0fcffeHfaXO9JgPZVUPDFG8xiK5M1qqOcWg/fv+h3edawaAyHMgjUT40jl+cyiuJrPqeuiTpsy9UU/g7KnII0mL76M7+S1LHctysEVZeBCUV7tCucba7MRGnzXWXitCO8Cg4Lp/VUjaxbs001i9GvV6Kk6TA/WQCmjkQos0QkiPymQrBNgyPna3WBhVDGXkKDZwpmVySSX8zqDbElpzihN8eCw9iOt9wTvrP1LFrzopEDr2Bti0nrdBQer4GlCoQRQZaNw5GElEJLSAr2a8G87a2TatQ0+mPLatIGubWkawYkSh3EOv4tAaKxcoekslIuQozLuI2lIMGZcNDJfVaxgug34xuM6DyGQQy5dun/32xdeeU3ayMm0KD5zt5c+K8feUlwL+wvUrBm41UJkcenlrwcLcfl7vZSbY+SqruS9Mp6wqLrIBcxRm0iWK1Ret1lSF6VWF14VCeVuFSqCr7IPTr6kSXgq9ctSd9RdltbzZw/ivFmUYkZP2W3PObJXmhOvMMtzXZxu/nE131WEw20NWs+tlUGfxbDpvM9LTyBFaZV98tShSqZIr3MOzQkpZKuv3bvHydGKHHve+8H3+Oqbf4B75/d4ernl/Pw+3Ztf5hu/5/dy+vZX8OEY0opHHzzi4tkT+v2We29/iZtxz7Dfsxsy7f3XoWsYyDz8/q+TyfjG8clv/Drx5gWMWzb3X+fDC3j09JJHHz8mxhGP0nTKl77xDsdnZ2wvdsT9njHBqA2jGpxacsY31rrC41CvrEJDaDua1Rq0xQF5HMjBW3gHiON40Hcp5TjPFwnGSj/0U362ILrnkr2D932wqiiZnIndfPpesG673hGcQ7IH1+Bcg4g3r6vrYLO2fFPNb3lbH6aj8uJ6bpqNJh5qoS4lBDlr3OUtV8UKgjhjYRFnRmVlkvGLdSKFecN7j45pYnS3YzA5osqYkykwsuUenbdQ8sqzWhmCz4WW7FaIb/C+pQmBWCy9NjhaL2SndHOTZsYEPng2bcOdTcdq3Vr4NEWOE6XlibJSx75QQV1LYHV+TLtesVp3gNKI0ImzZrLJeqVpIWo1+0fwbQBxZCcM2QgZU8xcv+iJ24G0HfDjaKHDrAxEvK4ITugk4Lpg76n7MUYb1txGydkWu7l8V3c6MLeZJNis52ThZNQFJNUxAZgg7Isl9vK9lOMm4+5AYJcjxeCu1VR1JURoZKO1CsxMVKEWHlcrV26FQ6rQnb2H6fPF89QwJEXJ+YMzvGrBAoVpHjlcxHNYbs6Smf6Xgkup+alXnZ2JiWAiQK27Vm9tue+yP1IlCXYcNJucjY9qFmqxRYQlZ3V2t661MHJkUcdkSeoK7i4FpFKUmPMEgeCE8zvH3FxccHN9wVe/+VXabkVOyqprzD4PK1YndxlHZXdzSb99yru/9qtGqUTm6vK5jUNOxN2WbnNEv9ty/fRTPv3wXSRHgge9eUbLiAbh4xc9Hz685snzGzuPKutVw/07Rxyd3iG0LTkO3OyuS27LckY+NKYgQsBadRjFlJa3FVxDP8QS9nLESs1URqjmW62coMzQEuqKKbLf7akciNXLrrWHywrIBb7pJQNifgMUzsIy30QKMMl48FyOM9uDLo22KTBnvlfOlkNGrRZsQXo5w+TzDFQoDCsCxgbhZn+tjhRIAWcs1q4P03yvqrjKEzMqy32VZ54o2uo/zpSnK2MagqdrgpE2d62R7zpnoBVM765WnrwPhNRYBYvYeIxJCE3Lqms4WnecnDQWuhuFplfWqnQ41mIhRY8D37LZdLRHHavjNSknK6QvoBiL9OTSx8tGIXjLrYGtwzo3YuFdjDGSYsSREUk2fhliSpAcISs+BLqVYx3/Z6aH+v+57RZqZdEZpXw/eyUHXgfcWkW3lFKVi+U7XZx39oZuHzB9eyAgK6qpzto6masFemiVmjdW249XeEaVzdNVZDYi67Hzoxy4TkutNJ3tEK13eHxVTEsNM6kbVQs1VgWms66a5ZdOwmD29nRxn1Xg3Faeszd1gLYqvIXiKDDjW3crr3jvi2dBDq9z+/tJaSklLGYPtgwfumIWOQcnmxXPH37E1bNnvPOtd2iahpyUxtl5Eg78mpuLa3Y3W64vLnn8wQ9wjcMFx2430LRl/Ac4vX+PGAc+ff+H3Dx5iM8jrWBUPK1jFOXhsx2fPHnG5eWWrMZ7t1m1vHZ+Yt6TCGnsGcYBU1IBgjMggRi7Q1a7/7ZxxGEsQ+dIabDO0GL9uuYxqOMyo8PMzrB3kFJkGHrm8nU9mFPz+5DF/zO1HGRaFQsj6aBjgpY2P7m08cB+rwpubkFSlURZXIXOQVDUFejO1FizlJCogU4mltDqwswLijlGX03RUopSQoA2R+v8Lt/rvGZtbi2Pv7UcHeaWlX2CszBh13rcqkGdJ6pDG29qMMO6C8iqockZddb2BSAmCE1L1zZsuobTdTAZMgIp0WWl1UznTZmLeGgajlYN7bql27SM2cJ+eSjApTwrL+tUbQZO7SYBNQytpRRDS8uZBcFzeac5JUieMWeCczSt0K26z1iVv7Pti6+86jZZ1LO/81nbge5SDhbSb3fMS3/o4fn0FTsdKFdkcT2Z/o/UYmQtFqct0qW+XQrgWRHcCtNVMITKXCgpSzW5VLMy/VuFUl3HC6zi4qL2sGJY/oVKUuQl5JBdqd7dhEqkeFV1j6UCFag9uSzsUnJczL9XzrnZu6v8jmU05Vb9Cr/FuwMDYExKawGHL3keANGMK95cTiMtcP3ikmefPuVn/9DP0pHIQ8/1iyua9Rn77TXPrx7y+Hu/xnZ/TR/3HOXIk2eXXPY7S46vSx+mbeTDX9+TYs/u2WNa37MKnnVo2fiGx33k0WXPD37wmF0/EEuE4cGdI+6fn3L/tXN22y1jHBnjSHtyQn+1Y7i+pj09IpWC2UYCeMGLQaetliczxkjUjBdv0PSSh1A1cEDKiZQsvFzLF1z5PqXEGCPNK978b7maZF5vSzRr7YoQY2IclWEP/ZAhiTXHDIE8JvI4WglACWFqKcKuJ8+acLmE3JcdlzlcB9UIVSlISJnn88Ej6NL6M64/TRnnwjQnU2GXTylNLVOM5sMfGrJS16XMHIwKOoJTofGOTRdojwLZBYYcUF+Mihxp1h0jSlo58l5pncfajAASaLywaj3nm9aQlSkyjAkZMz5lQvCsJBB8Q1PChU3X0jQBn5QhZkY1FnunFuKMw0BTjEjrNmkw+Og8uAAu4EKDayMyBjQGxt5NuVIV6GNEBocOykoF7wOr9rOnyO9k+/8L5VWFVrUEFp8C80LxLLym+aBZ5y13XpylfqH68qfzQbfU1cJzWyo3l2fFMwluEROisrhv0ZJgLy5WFR4wh91q48XFzZuXMPs5MHtx1eOZ6GGKpXz4NMWnmqF2t75j8qpme3Sp/t28nxZlfXCKGYelaFEK5ZW40nK9tk+YFJmfGDX8ouZrGnVdeqRTJd10nfpu6u8zCmwmU0Wr8JRZqFEVurGHBye0Xvjogx+yWgtvf+U1Qo5cXFySc2Zz95wXTy4Zhqfsxh7fdORt4ubFJat793H9Nc02052dMtxcAZmT8zP662dIjpwctbRZCYw4vaFv7vPBBw/53oePyWNv4URR1q3n9XvHHJ9uGLJ1z46a2A89aetpuxWr9RHj0FMqe8hZaTdrJCeGmx5pWpLCdrunWa0Q50kqdKFBURPCMZr3izVf9FUJOCkdl6N5RmYrTcbT9I51Oeo6hcY52G8h29W8pzQm+n1ktx2J+5awwsKEXZ2XlVevvnurjVKwrLHYPk7NyMqTsSNmeFHCfoWCfioj0RqSnmEntZliXU5+OpXivS/1b5BjpJbQZDXFmrMjJZvzzrmZ78+VNZgxzsnSDy0lU36o5cO8CNJY+FfVwpmrdcsYMnEN9NB4WxcJY4sXhQYlFONDRRhLRFMRCIHQriycvG7xBdihKZNiIic1GqlooWsnggsOL4WQ2Bv9VFSIaSQPJXzqPL4NtNrhXGa/3xFHISqMZEacNVDtE+Fiz2rVclvy/KjbF195LVBlWv/UWXhP+qBYVRXtV+T4QsTf2m4J788KS73ymIWVNf2tS+FZT10Vz8L9q48jVr+iJd91+zL54Phbt4EyRw7rtXVWejpfb0IEHoj52/c4b1VpVMDZS89e2QQQai1YvY6homaPblIuInZMWeQHdVyLHzcpvln5TJct/5NqSc93PIWpls9VH3VSVEvPS6parseYQm+CZ90EdttrVquWtt2wu76mH0ZUhI7AzdUFfb8naY80pkDImT5GNCVErXGhU2aBlkecJroQ6CJGfKuRp1cDn77Y8uz5NS4nnCSa4Dg/WXG06WiCZ4iJ7K1tSVbr89U2a3zbMPT7gtL0k2Ggam3anTN+QM2ZtW8KjNloi1Qx4tSUDOIMJmAdgLMxTqn03TJ29GpIiMyjXA2jWTkdGnmH9uIcbrfcSWIcRtKgON9AM0IXyl4l26WzkTWFsaspVHNT0/nrFUs+bJHHzlU7FaTJYQn9vOoribcIJd+0DL3XpygMEuXYnPPMTiJSFKYZYKrMIdLy/pZ9+WpOeQ7DO3zbMHqje8IrjZ+7PMcIJCP5rQhJwEpSXIHFhIBvWyisGOJs0eRkzSNT+SHpAvG5aPLqheDMGMpjYkwZcdmUoReaxiPasAue5ByRRASiYkqxjwVJLTTNjzHacAGCLR6XTgIMzKp2i8W0dLpg9j6WZ3wJIb5QYtNCqMe9UpvN9UGUS1eSXjsq14/NCptyK7PHlahUMs4UV0HQacqHiMcifA/YRKRew84nk39RyWUXWnR65gIQyTrlmJBbHhPyKslvi0tvfS7WCXc+3o6d/y7vqVjxUqSBK/UtzrkJllyV1tQteTrFnEpfDsjyfiYRdEsB270Lsx2jC+U117C5ch0EgijHq5bzsw03+0vunt6lCy3f+85v8NpXv0S73vDpxxdcXb0wbrnzt/iVX/ol1uuGO3dOePTeh4j2oJEXDz/g9Tcf0HUd189fcLxZEYj4fkebBxLQ55b//pe/z+XVFZTQ3jrA2VHDN7/6gAHPMGZ28Qp1jd2jD6SY6PuBpNa0su1WtG3LZr1mO/Zm2QdrVyHO432gabuJ8QFxhXQ3EuNI03XmXaiaYPQg6sgpoprKWNb5Vsb7lmFzwKpR5qB5Zkx9+CrjhRG7qhH+psgwJGgafG5pwpEpTSmgmtpZnJqTU6YOy4VEmJQr3M9kgWLzXNVKSib5kKY1knOacpx12lc4xrRUq6E1WWeKYs0xrU7OUYE/c/Dc2NqdeCbG9cLaHpNQOXcUIVkvEWI5n3MGme86885zNqomJ/ba1HtShDQm0nawztdljWUXyB7UKbld0azXuK7FdYGEs/GOiTFmxjEyjiMuQ4PJgja0SOm95RpPK87Ct8Oe/XWPSCCEls3pCoJHpIEQiM4zuEyflX1WhhQZ+sj2BsaoHG9WfJ7tC628PnuzKXCwpPTw27rNikxf/oxX5X5m72ppo02W16QQZ4WybHR5mwZpErDTPvaNMguAqjbV18LE209xuNW2D4tbuGVNlt8EZhFe+TBe3l7KS7GwQl+6m/lq0zOKTsv3IKxbdhARvHgkBLP0xFqRUxSZd4vFTzGqpY5ffdezr+Rg4ieZx7OKzipISyI5p2qkL+bArL6cExovtC7TBsdqveHOa6/z8PvvcfnkKW//xBtc7BJPnn7Ki8cf8843vsX11Q3/w9/629y7f0YIDTe7xNHJmocfPeXy6oJvfOMr1mRxHDndRDYBGAfG4RJtA+99cs2vvveCF89vSEkRdXSqvP36Xc7vHhFWa7b7wQhXM/gS3hFx1s8pJkQSoQ3Wft0JvvEwipUZqeVXVk3HZnOE+KYoMk/XdYgT8wDKWKhCihEfAtW72G13pHFc1DGBoWRvV3rNY/7SvJrG2s2zT/Ns8KmiwaHBaIwsZOlxrsLdS/EuViM5zQfNJXSXcSmhPhVP3lNj77WpZpHwpuTEoc5bnoxsXivO8oRibexr2D2h5DhaQ0psvZnOVCTVSaoLI8zOJTnO3qIsxkGYOBMRTEmrKUSkwQUrjG+DILTk7Bl1P3mQ3juGnMz7TnYNX/LE2rQgwe6zW0HX4roG37VojJa7HFLp3hyJOeIjRAHJgmtLuFOcgT3J1nEgRcbrHVkF5xraxiOt9QZLCDsV9ioMCINao81BoY+JfL1lu+tfOS9+p9sXXnktQ0JQPJGFqLJpvqz0rx7QLNIOznfrXEsvZxaWi+tODthvtUCZPJ6lUDevR+f9pjyO3Dp4cW/VCq5e18FOMt0nRYFN97Y0Mhcnn8N/s8Jdeqe3n0MX/y4Pu20QqCzHdr4vAaO6cbNSrDUxrize6mlV769GR2U6+/zu6juZEKVF6x96ZLK4R53f2W2Qxi2jA1GC82zalpVPdF1LaFsuX1yx392Q8kB2Lc8+fcj19TVdawWZu92O7eVT3vzSA/p+5OrZBffO7+C8wZSPT0/JfY+kxGrtCOOI6EgI8NGnl3z85IbnV1tj8sbG5XTVcvfOEcenRwxJ2RUuQvVG3VQ9cydijOfB473ii7sxjGN5Tle8tIYQ7Kdy9YUQCKGZ6xp18m2t661Xy82ibHc3xGjK69DTncdaZIkIrGHrstfSWwamUpCitFDLA8nE7FDXsk0eKTWPVXEcrsFDQ6UqEqnFSlJlg1anyYzCGuGYohM6KxihGKtSyge0sI/Yl1oTf8jBPVvOt6hnWRp8NUxuP96V4uSlN6ezcWbKTXAeBKu9i2L1h6Cl8DlBjmiy+ipCMIXoA5QOFuo8eIf4oihRNCdStDYoqhmkhDFjkZzBFpZqZszW6XqM0YqQx2hsYqIM+wEyRKwFy5AzQ1ZGFSJieTkgJaUfIgaF/N1vX2jl9Vk23cKApuY0ZqX1KtHMwg2agRnTfJyUxUKBTR7Wq7bD8y/zZUvshBaSKCPFXOSjEGrPIHNX6ilLiC2b9XtbbVaBIErJTzCFY4TiAS41WNFcygx8sp9bflq5p6mFymI7GIZybKV9qoaDWaDlhhylRYfVUVU4vOV/vAnfWuNV67tq3rEKlPr/2xqTsu/tdyP1/RdRIDC/wllIVDaJ+SClaRx3jtd0Xjk62uBCy3f+p1/i7AhOzjueXCrvffd75LjlD/+T/yiPnlxy+fwZ67Cn2Rzz5OmHvP+93+T8H/ujHB2vaXzk9O7r3HzyXSRuWa8ekG+e0WhkvWr5q7/2PS73kewcOGt00wTPG2+ccn7/mNCu+PTRNRc3O8Q7NqdHdOsVaRjZD3saH+i6jmazRnVvCzxnrm+u8WGF9wEcrNsVTdMhwQpOfQg0TWuKkCKoJ9lvAA5NFgpLqlxcvmDo9wdhXKnHoROA42CsF3SRysLoWOoaBYPEm3UvUZGYcWkmrBY1EIoWgIcic2l5mVeVQHdaF1kRmWYkCAbpFsvh1R8brkzlV5qUjc4GcIWDp5TK0qjntPxifdAannS4qYnlZB6K5ZRCoRoLoSEU5eWcGXQ6kXOWNeVAg5uKwDOOlCOipXxgHGEYrdi8OqfBE5pmsgCzk4nQ2Dggi7IbR1IJdYqzMbDojUCWKT86jCNj8dZySsiQyWNm1MTuZk8eMiOR3RDZxcQ+ZZK6orxqmYIyjgWc8jm2L7TyqrLywMKuk5gSp84VFHB43G97YpgswduYNUPv5Wm9vXyw1URMt7a4oGN5wzPSzi0WvVfMNanoJVfyaHXlOwoM99VeUpoGx/72i5zY0vJz5W6m51scU5vsLc+99E2mEakLfNLsc96oel+uCJcpE5ejjWNBe9XAvQUnZHp5c7+uWYpIfQiq9Wxdgx0z2fEEGplufqldvd0jZgAkIBWGf4efZo8T4bXXzmmcMIwR33R8+uyK/uMn6PCcs7e/werolL/yf/6/8K2ffIs3v/QtmqMv8Z2/8V/S+Mzv+Znfxw8/+JhGlZ/65td59vHHvPbGXc7uvMX+8Qecn27YtBs6Rp5L5NHFDe8/vObZPoEKKxc4C47N2YbjO0e8/eW3ePbsgsvHT3hxs6NZr+m6Fat2xfWzF9Y9OLS07ZqmbWmdZ+wVDQLe4ZIaaaszQdluNiUMWGrYsk7Qd+e8oeqoRoMJ2ZRzASrA808f0+92Zr0XoTcBMqb1U6Euk5qawBKVxSOLgK+6wsJkSRMpjaQ4kLaZ1B4R1dHlQNRs7Pe5hvX8QumB10wWU8iIoKG1WZVd8fhd8SCzrbNSlFwNXUmJ4D21PSYozikVja/FOJJoaEYpKGEPNOJAvHEQKljvtAxpBMm4IFO+DoXOOSTYMU0I+GBovsrMYcZVBp+AiKgjuBZEiAn2MTHc7MnjiCfh0oiOCfaRnDxZG7ITupPOFKEooga0SeNgpRL7kTQkpDfPzd5PIm+3pBDIbUDWnj5G+jFxddUz7AaICZ8yaegL0z9cPhNSE4he6MfMEIV9ErZjsmfHACkpqzUWXTTF/d1sX2jlBRxoj9shRPmM/eaYxUJoF+F7gF7jUFC/hPqbTicvfSoHeywEwQF8eN5rGfqQgr6bkY4lVSy16M+8gtonqJ5kqSgPr1AwegIsyG1dtazmPWZFdmscbuu32wpzyUlwODqHyt8W5KwVJ8TWdMPCAQEv8/eHuIuaDrdrT6ZCUaDL29D51ym3VX+mOyztHxzmBTZNw/nJKTlGnl8/YbVec3X1gqefPuYrX/oyVzd7Pnr0nK6F49MzCB3f+Qe/SZDIyckJq9PXufr1X+T+6RmvPXiTjz9+j7ZxrLqOzeqS4+AIAjFGbobMs+uRD59cWxsPb/m/83snrE5WtJuO633i2eWOm5st2TlC0+JDKJZ/ab0eGlabIwSI40Bw1UMB7xvABIZ1oi65FdUJKAMlFCbOAPZlwGp+qDac1KzsbrbE0cI+OscAl69nWi8GoqmG5ex9zGUUs/ZZ2mgyATHq58Ujq0XTulgzxdvWnMr0srDYPNWyKVldAKpqaE91RgaXuaZSOxNonTRGWzYtNy2KzzwUK7kw8MgUIlVKraYZJFrg/RYKl7kvmLp5ci/dVUqoT8o9ZuMLtKRuNjLcCGlUkiZCjOgY0d6YSLLz5NAYO4oTrFBbrXNyzgi5oBzreinPEJOFIKPRQ+WhRdW6Zg8xsd1Z14GQTUknLCQ4jsa1mdWX2roCSCOXEoOSPw2O1jmCfj7l9fmO/v/ytkT1TVO8/lNn/KvcrNsu0+21p7P8O1RLVcXoUjZOW21rIrXSX2vYgDk0UQRl7V1UALGLIOAr1EPp8WMoJV/CEH6e/FUgCNQk0fSIpecQUuqlmD2cGYZuWCrR+Wdava9yL+dhLt6iTkXLsydXMwRlvA4IBut1SxhImBd0+fELlOHUbv2ld5EXI1UJZXUeB+TWeGiBI+v0wzQabhKm3jlWbcPd41OOuxVxGBiGyNXlC549+Zg3vvw1Pn1ywa/96nd44/UzNien7Eb4hV/8JU6PWu7dOyc3J/TXL/BOuXvvNU6OGrrgaES4f+JZO8WlRB8TL7aRTy8HHj67JqjgvYe24fzN1zg5v0OzWvP42TXPXtxwte1xviE0LS4Ecs44H/ChxYeWzeYIRBjGYQIYqDhcaEHMxfGh8F2WV2L8hSXUWzruVttJ1ZoK5pRMeKZEjon9dkccYxH0uU7u4l3dgv6UV7LMj5nDXUe9vB2pSqTUCpZXWMP3E5R8ak1SyDTqCiplCJazcgbuWMyNnIuwLsAGnSexXchJqQur96JTHaA1wZxXaEU2armnlKL1TNNMFksIZKSwVBhTRQVrOAS/7A5eQoKooNlANWaI6xQ9N8WciWMkDyN5jJCttismYYyZOCTyENHBwofaj+R+IMdk0HYglRqtZGj4CUBl5SkWy9Cc8WSII3kYSH22A9QooPohsusT+8FCgkk82bviEdtz4DyU+q8gjpV3rINj3TjWrWfTBdbdjzFU/sDLKn+4VwjaH2W75ZO9tM3YvBk6e6gjy0KcEEWHZ6qw+dnfqT7ErWtajKL86hYHz9rETp1nz3FxbIXJzgBzOfxZ5HZmQTOrgvrb7fGYvc3bN1xPe3jkEo857V8Ye1xp+yCVPcNX5nhTXpXfcMH6ON2F2bq/FTfaoRFglrtSczYmxBae1wQwMRaKOwHGcURCw2v3X+f/9l/8V3zt62/yx/7YH+av/82/wycfvofGLb/vT/xJfvFX3+Xhw8fo1XPe+iM/xXZQ/tp/9Tf49tfeZu0bHr73Pn/gH/sTxCffJ10+xJ+e8vTiBdshkvwRv/wb3+XxsxckP+K6E9quZb1pCUcnfPr8GU+ePWd3tcN5T3d0zObkmNA0NMFAFnsGQruiWx+z2480oaFrGzT1hKYDPMOQjfTVudIUsBQgO0dz3DAOA2OMrI+ODrgMx/2+FIorIlYDFmOk329JMTLXCVbrvXoQr7IOi1dT14BW06F8VwEP3lttUXC4xpRQRkjZQCk1oTNZ92CeYUroOJJjgqAzV29pB4M4Ywup1yt8WfNtlwOqwpKMV1eUebHYvFlpLmeiRsjmZeSciTkz5mzPU4l63aQTS+lFoVoqp7M5bzHJmAq7yJBsXXgtYAkj6o4p0+9HJAuaMrHvSUQSiWFUdMiEmGlUcSmiQ8+IoquGxq0JjVFapBxRlIYC3ChznuDxIZG8Q+jRfY/GSO4H3MbRNo7N8Zp9L4y9GTLaOsRZs8lOAk0IuOAZGiWFgTBE1l3DyarBO2OUHMZE58Ln9py+0MrrYKvWE1BnRkX8HArgBRjhYIG90sEov+hLf2eWAnxG1tU9q+fvZjleIgJLvOOcZJ7vZelx6eH91sPcoVKTUjBTleCBB1cVWElsV1aBeqoD5VIA5zXvtHym27U7t7WtlGNufbi46cUxMiOtrI5rcpAmVNdt56meoPZHm7xgXV6MRQ509rZApjBhtbjrz3ReTG55oBVhs3LkNHJ9vePDDz7mzp0NIo5Hj6/5wW/+Jicbz9m9u/zw46c8+ugh+5srvvrO60Q5Yr+/wm+f89bb36YVIQ0jOlzTBhMkzy6veXIDL24yjz95n6vtHuc9p0dnxN465AYvPPr4Mde7HeMQ8V2gaztWXcd6vSa0rSHUxNOsA943lgPyDaqRmJKNVUw23s4BkVyEtyvIQu+9Jd+rxzoNjIWQJoYyJ4UTV4lxKGCFmaFFD6Zt9cLmczG/Fvt9ET6Xkmus5R01DOzETewNIt7mrgNx2VpyiIXUsmC5L3sy6z+VHVI4EHOuQB1dTKoC1ChADlOC2WxBZ+EuUYuNzE1Va7jSni2lNBXFxxSNlDYZyMaLGNLTL1c7BSShpmyLQqwG39KgrYaFecRlHYtYZ+tk6M+kyWSBtxq9LG5S8iKmKMxzzqQhoTKa/kUJHrzXicsyZ7X55D0eGPe2mJwXnFfrHO4d3UppV8GMylEI6wbXeCR4cmnf4r0jtoBzdH0kj4njzlnXbQ/DPhHEmTf3ObYvuPKaJVCVkQeO/S2jf8mjVnbnpZ3mw29dSQ8+smUy7zcLSmbrUqo6qKeslmdxz8vO5htVcTz7OAcZPKmFlVUILDJUhQ6Hg/PLfCPTDcwPWAXJgbFcrqivdKsOdpz/fdVu08eLL8u+UoSTE1008JND5SUcvIuZNYGDt6AHI7g0MIrA0MNPK8nschzms5ggboGVF1ZdQx57ri8uePTxI+7dPyPj+ODDT3n66Sfc/+Y73D2/yw/ef8TF82esPHz5y2+y7ZX9fuCsTdy7f5/GKcPuCh2uTEiGhifPL3iyDzy9GHj3h58wxkhoGlbdmqv+siiNzNNPn1lXXi8064b1as2qW9GtVjgfTKDjaX21660INsdkbAY481RECXVOKsY6r9nCsz4cKKI5b2w1cOIqvZibWCDGcZy82NnUOJwi08Vk+b7s88NghEBB6FUjqRp1DoPwm6IqoURnZMM1Z1f3rsXAMwigzOUSKpzm+0QJtVi4RUObwijlK2qh7oXdNPNfUgAnKRndlEjhgLSf4OfwtxZWEwtZzyHEqry0zEupCnUio3ZFaTvyBLt1aKlfS9k4FCn74s0DTMWgDYU7UcVCnmlMBpBRpXECwboLVMh/zooXQ/vihAGm+/DevEgJntZBs/ITKKxZtYSuwXcNBEcoazo1BolvnIMmsemEEAygsvcJr1jY+XNsX3DlNW8Lg7H+7xU7lX8/Q+BOwvglxXWo426f7vD3Yukvcjy6VLDlw7S82CtOfvuTKqCrZzJX7jusMdEERp+FusxCW+bg+eF9Sz3/LOhnZfAKocTLH88LvGrDCbN38MxVo1u40E9WdSU3XbTvmjy0Ay9yukOzKD9rrIpEKu/N7qAu0Il6Z5GXqe9KknL/bsOdkw3Hx+c8efgJF48/5unTj/jWt/843/kH3+N/+B//Hm/dO+KNt9/i5O49fvFv/l/5yS+d87WvfImvfutn+Rv/xV9j1Wb+2M/9AYJ3HK/XHJ/fIe8/5NHjkYefDPzqbzzk6T6z60fGYeD07JiUlItnW443DUM/8OjymvXdO3SbDd1qxapbcXq8oe1KI0DXIs5yXYoxUqSUiSkh4vFBiKV1iqbEkAa8b/Au0DSBxje0oaVtOpRU5opCiiWUZEwTHopQtfcxDgM3221BGS7fz623UEFHtwwLmx636xMpysNUFql4QBgIpSpqVZmUltQ8qBShLra/oDS5Lag9Z2zmrwrdC0yo02xKTnzJkdV712zCXzOpIFMrbdaIhzjaOTRbHmkcjQE/+JmxBKY1a0wkljvMKc+1Yqp4hOA9ofGE1rxivC/tWYpiQxDxZElkSSV6ESw8GRJJnBVtO0eozCjeoykz7AfSHmJhpu+6QGg9KWZSNLi7SxSgTkKHZGvEeWg8rmlwTWAjQruKREwxh03HetOxPu7wvjoNyqCZ3LWMOHSInGw8bRcIXcuu2ZNSZrf9sfa8lkuj2ovzgllaTbPiKqGn2XA7PN9iDS7Ftyw+m7/Xg2U53csC7bbk4Zg9Gywxnee798wiOk93b1dw6ouaKp4VTBViyEz9VO/dK2ahlXuRslBNIviD565WoQkDpvO76W5YKOLlSNf/z1m1pSpzCGJdJVEppQNivYt85UwrQIEZWVjPoNSQkFnvTIpRC+np/CLqi6xIxvq5eVhTq5Ncva4ywGWELSFvUPB3zk842gTarqPzju+9+wOGOPL7fvb38Pd+6df45OEjOq98+2d+kj4p777/MXfXnm/85Nd54803+OF7j3jt3orze6e88dVvcHrnK9A/o795yM0Ofvjhp7z/wSPa42PS5WN0iDx4402evXhqnIReGRPQeFarI1anJ6yPjlmtNqxWa46PVoTg2PeRZrUBZ+E2B3inND7T51yY+D2MvTV09EJMiSa0xkEXPH5CGM6DVj2V+mlKyVgaZDZD9vs9z548YWo9X12a+oeUkV14WFrJjqcJN/tKAiAVeFPmWxZ7faomEAv/nhaATVUKB+saMKJqh7QNIGV56VRGoXkhH9QAINOcrpD/rMUYBEqz2ASlOWNkiBHxHj86kiQM7heNZqlSKIlDfWHpl0zO5mFI8Xi1sNKnlJGYSKRS6wniBQm+1OOZMpKkxYuyUGouTCE57yfFa8wX5kl7JzSbzpS+N29MRIgA0aD/UEOq9sw5ZcZhhBTRcSRtBwOdeIfrI3QR58VCk9VrFEF8wDeBtvW0jSNlJWZFdz15MCXosrHUdMHRNQFtjcoqN5/lRfzOti+88vrsrXo1cDDDC8z8s5yzzz6dTFZj3SqQoF5tvt6hYpyVoBwotkkoL2/xwBOa/38YEV8o04VVOT1qsfQmRTyv1/nSi6jJHE6Vg/1feohpLA6/e8mKXl62jldtP1sRkqVfl6tFyks0oXDw+yQUdXnm5SgcfjQ/tv02IQvLTy4Ct+4XgmfVtpwedbSr0rl2GLi8ukKC5+jklI8++hX67TWnxx1vvPUWP/jgMRfPn/Ol1+/y5pfe5OzuOe+/+4gHD8659/o9upO7hLYljsoY93zw8SWPP33O1fUNp/deowkBTUoI3nIlObIKgUEyvmlo1mtWR0es1xu61Zr1ZkOzavHi8Hk0stoCZvFS5rSU7NECjFO9E++sWNWFplBwvSL3isx5GDDB7w5DdGM/cHV5WRQQizO89EYWL0OxIuYFOlSrRzIjUQ90kSqSM9K4qbDVPKU5N3aQvCyd1FWKt1C8KpHac6sqq8WVyt9mIMp0roouFBT1tazCgCI5mdekhUmCOCJxNMRdLl3XpWVqvVw8t6zlu0UvMgqbvE7jMxtyE8NMzRMs8naaWSCBK6LZqLLEe7yY5+Uby2ElNePS6kczNQpr51ugTIexIAxH0jDa/XgPfURG6yoQRc1bS4mUzDCpMSBf7fFsObYUIymm8myFgMBBGzxJlCH8WNd5HXILSvFBEiU3pMUNtl3LXsvFXb9YKqBb22KR2j7VI4KKg3MwhQan85jkXpC/LhatMnkVIqXoWUElFy+nYrCU2v3WFlmd4LP+nYQOZT6Xm21Kx+ZcOq3a95ZZ8/U59DNqJaQu71mcHOqxBT5RmFD1syC0omRX7tnOZ9JKfCCH1uiIFtB9P0H+7S06mGmKYBI0yzinya/y+WTmzwNeO1FHwx5P7cyjMgkXh3B2dsKdkzVNkzg+OUUTPPrwY1arDbuYePe9x1xevODuScvXvvw6Zw/eIv/gQ/L1c/7Jn/vDvPaVrwPCaXiXn/mDf4CwOeFqL/S7jwjaE90Rf+2//i+5c7bm7M4JOSn333iT/W7HR++9RyYTWk9YtcQxElYb1id3ODu9S9t2NE3D8dEGxSzfo9MjdtseVSWEUBpJJjKZ1gdiTqRs6EIjmYVuvaHrWhPsmZJrzIiOZV4Wn3dhPIErrByCsgcNDP3A1cWLkss1AZsW66QKUlubde6YUp1fTTGVxLwNCnJUUcvxiaFJfU6IC4Z+jIkQLK8rUuAZauFjQa1dC4XQV8FnQ/XZdQrwQcCpSX3NafreTevBip9Vs8G81Rl6ORhUXJLiEqWUpHALxoiLmVGMCNg7TxRrBGmREo/1RUkFn1DWiIBoQgozPwlQB9nWhBSwTM5iAA0UkcRIJMWRlMyjUay0o/WZ2AgBR+cczVGLXwUL9akp1piUfrRSAS0E4BrU2qCkhIwjjCM5jvT7vckh7xlDQ2oa3AgxZvrtQD8k9mPidJ2JTWZsEo7EMEb6PtJv9+z3PWnMhOwYxpbQKG2KdK213xn1x5hhg4OQxez9VKuxCt9XRNhvn+Qzvlp+pzNzBcw1W8zq7wDSUVFJnwF+mCiPpg8oyWY5sJxtwc2POS185oDkDBRZnHM6hylBM9bsmCXcf0ZIvGw911uvz/gqb1JlcamDb5lQkUAhfnWEEPCliV8pxZn43ZboQpGDq1DRFy/f6eHFD96YzrU5BUhGFLNAA+axrE43vPXaXe6fHrO//oSjO/fZ7SIffvwLPHjzPg8ffcov/uIvcO/I81M/9XV+z8/8FN//zv/El968y09/6yt8+Vu/l+uLF6CZb/+xfxy3OWMfM/v9DeiOJx9/zMN3P+D1N84Z+pGrq4Ev/8QDPn74kN1uz9033jCY+hjZDQN3XnuD1eaY9dERx8fHxjfYeLquRdWh2ep6rGnkPLMVAefxXvAEsirDsEcNbFiADyV35YphkJU0jngfpuJfY5A3hRC8w1VKpWL9x3Fk6PclKrsw2BxTIfJtsoCXt3nBzjN28eMD2Y2MKHkcIKzKfTBPDu8LoKGYk3lxL9nySmhhBCmelaEPmc4zkQEc3FYpeBbLTyVxjAUkITnifSS4kUzApUxMA/3Yc73dIlFpfMNmfUzerK1Jpse816w4TcSSS5u6QtdJLbmExK1WLQ6OXBRexoAaUZP10BqjcTGKx3lDYLrO47XDq4Fz2rbBNw0+BLyKcR16JbnEGC3XpqUuK0ZlUKXJ5jppEhzBmGdSJo+RdL1DfU/Kym5QtjGxHUaeXV4QU0eMLeuVIR3HlOnHgTEZ6jUm2A0toRE2ydCxQYLd0+fYvtjKC5ks8ckHuwV+0Om733o5HRxzC8EociBGp/Me3kq1/JfXWrhHy+NuKYSKttPqVb1CsdVjzaGbhfdtStK6u91rtaTn800Ip6rzKypq8XzFabT7XLDHT09VBsaiFjIN0qRY6v3Wz6R2QvaFgNTNfbtehTRcqOj5jl490aVY+DUaOynzEg6qIJ4aLVIt1DSYwnzt3l1aD2N/g3ONIfGAdtXgSUjck/trvvIT3+DB6/dZrzc03nH33n3OX3+djAffIIDrjri4uGHXD2y3PXl/yfOnl7y4uCSNivMB17TWdiInRs1odiQ1gd2uGjYnd1hvNqxW5imFYCS7tUmiFi+8hsZc8WrF1Z5LNmxOjVRYyxj5EObpWMe8fJBLAaopnjpzZs+2FtXHmKa+T654u0uD7bczEOu7mj/TaX5O2FkR1Fm+KDtnhLPlOZcw1KkjsciEDlwaONNc1Io6ljlPt/z+s+68AjZ0/j4Xjy2nkZQcfb8j9jtiv+Xm4gKGTOs7xrv30bMT87ocVKQkSiHQLVGihRzI5cYrcCbnRM6Fp9Ex58pGK0LWZHRT1rzVwnvaWP2ZlNKISsGkAN6e36OM0XLAsSiaOCZizPiMKa9cvNWKRBwjUQR1jqjQRxiTEjMM40g/CG1v8ZKsBr+3QnG1MgIy/ZDomkxKM8/pyx0IfrTtC668bk/CQ12xFHkHXsOroIMwQYU/6zqOuSJ9eX5Z7nXLmCu2qO0vr17gta398jub1PNnmQnugbJ0aha2a7mun7RAgT9PYAidYudL72sS7GLipX53u7brIMtRvcFb4dLKqbt4CGtr4r1ZgT4YUONAeZVnKAJVFtepVnz1omu49PAtyYQ+nBSYplKXVB+2wMDV7kdFwDu++ZUv8+jjH/DwyVPefPA1tjc3pJT50le+xEff/XXSzXMenHX8oT/0s/hmzfZqz9tvv83JG1+mPTrj0w8/4OjufXzb8unTLQ/f/z67IRLDmvjpI/a7F2Qizx5fce/L9zm+d8rjTz+lz5HRwcWzS5x3rNcb7p7f4/TuPdarFV27IgTLiXnvGGMqeRCPU2t1L2rzIBEtF+YcMcYSSrWQohAQMSRbHC18KkEO3m1KkYKQprI6IELSbDyUpcZqHHtijFOR8JLqrBp6h2/l9so8/F2nQnNZfG1tSXIIaNPAuIhuFBi8edR1zRQJkEoSKOcJgWjKMS0U21xXSVWkk/EmRSEuZruWeq2ag1LQOJJFiTpydXVJ3t2Qbq64+vRT0j7ThhX3z18n379vYYVQwgs5FkoQMxSkgFFcibsn1RLJtnmaKotIUiSYEZJVLazXD2jKNNnjGo93jtAGchqRnHFkA3hQCHUF1FUzxpC34xDZ9ZExWj8x7SNNFiRjgBnnkZxwqgyD8SBac0lHH4URKxLPam1U+j7j8RSsTQGemKEdkyFrG+8ZN+20gHX8cQ4bOpi72X2Gdb7UZIs9fkd+mMxicymPlyc/tCRvq9K6LVSFlCkkUIlxD2924T/c0oKmZMpilkNFKsw5IkGmBLY9xoxGzCz6VwnFytJpcdb7v40j4RWKd+btmBGHE5lxCQOKE5rGimGNhqi0PXEF6D/R5Bj0umphcTK90yrq6sjM1modXROBTpfZk1p8Wb1DuzcXIfnE3Xvn3HvtPh++/z6nZ0ecn5+yfbbH+RZ8YHV8xtGq5Se+9iV+/x/8GU6/+lM8+fQpT54+5Bvf/iOMKfHixY6bHrgZSRfXfPD9XyHfjGQ8uV3x6OPvgjiasOLswRHqhYvLLY/e/4jm7A7ZG03T+esPOD074+6dc05OTowY1guhoAT7MRHCyhZ8VuJYemupkuKIb0y4xcKGXllcmuCpNX/BNwiFFols5LGl/HZpAPsQ8CHggiv1UdVYcjx//pTrmyubi3mmLhJkQnDeniRS+DdvldaVFOairrF6VWpceepbZH2MI6AuGNW1Okg6125p9b6w3FEuwXEJpSjbY9KY0gdv2YJjOZOKxzfNMhDxJsBd4eETkDwQr3piiuRhz6OHHxO316Tdlm1/g46B1fqUcbeDDF4CzgeyC6VA2PKumi1EOKFhSzhTkynhlJNlpwsvlpb1EcS6E4sTpFQAWN1vVazmZRoBsCM5syQ1qZHnaqbf92z7nl0fudqNjNsBH5UuYRB/zIMjql2DjMsJ1EKofRayNBbCRmhFWQVh03qapplIrrURIx72npgTu35EVFh5D86AOzc3ez7P9sVWXlWwHaIXWHoG1eI+yEfx6kDi4R4vfzaJUp0F9hLAMe+tsxs0nW2u0QKZvK0pzLEIiQBz9PGW0mXa9+VtVl1LtoLiUVX3igVwX8FA95ZSnyiSFohIVVlgIfSVSmy6WanXhyXxqPO1S3L5vXZLFhZQ+ToOS8+tXniJErtlIFQFLzOLhikuQ4RVRT1t5dqb1Yo7x8d8+MPfoAv3OG5OWHUNrm3JOFzOnNx/HR8CR+d3efz4CdvtDtd0fPLpp+xurhj2O5rQ8OLJI1I0WPouRfpxYL/bMY4WZovA6WbDro/sx5718bExa6PcuXuPszt3OT45YXNkoUJf8oGak+VLRMp7EMQp3uvSNikeyHJGuyn8dNtcqyFbG7qiRaa56uaXAKWxQa2pUy4vn7Pb3tSBp4YSDkLXOhs+c93hbS+sTpcZYTdZPdM6duAC6gOVyT1rxi34CWvX4oOLV/UjdRwKUH6am8sFVSZbDa1Nt7aICAjW2DKNVlgfB+Luht3VC7ZPP0XjAGnESyQJViM1xy7ma1C6F6ixWaSkpqQqN2I2FF9OOinl6Yd5PYUQyE0y2Hyo95dhUOKYcALZe1QMUqhSiNQypJQZhkgcs/Xbipk4ZjQqbSpGeqFti5gBKdlV/w2vmUaEsZqH6vCaCaKE4AiNQ3LpIOGEUOD4IRg92RhHbnZ7pDEQ0X438Hm2L7jysu1A0Muhsqpy+GA63fYqyh9yO9Gs8zn08CNqEpgSekgLQSHz6jm4QVez5fX3paNVjpub98mBIJg+WwiKg4RzFXCTJWv/m+5CmGh76vFV2FecX/W+rIlgPX6RJ1g8Uol2HDxnFbDVihaxPJer/IXewoc13zWPx+IeD7RX+bi8z6p6ZxNh7tBcRWh9rlRYtw/z8QJOaJvApms5bhueXTzn7qrlvG05OtqQ2xVjUoiRkwdvQ9OSQ8e7v/L/pO1WHJ/e5YfvfpebixfoOPLOV77K1YvHaI7cvXOP7cUN+2HHi+sdOXtSHxn6LefHrzHsnrPd7njtzbe4ePYCVXj9jbc4Ob/Der1ive5o2zDBpPt9wvuASGDfR3xwhfHe5njOGTeFaSsnn42hEwWNLCy6yUio8HxDaxflJaUeT5g0vvduug5ELl48Y3tzY/V7C4aY5XbYeKEabYtwvGI5NKXUj7nFO2RhacisvHBoUis4xs2eS/W2dHq700yo0Q2cM3CJs/5ZOkVoXFlixo94oMvKvUlhqMhpIGvCkwn7G8br5+yefkJ/8QLnhab1dKtAFIdvOCAooLyPjFFIqWK1UKU4mFyJfRMpZmI0pZBLKDPXMoUSYvdtQyKjhQcUAU127jREW2NtgKq8sEaQMStpzPR9ZoxKShCTtVaRaMAMBQspyyxvxBvi0udSfeqMTUSwteVyIqA0XmhC8Y6LPPZk1CtNUEYZiFm53ilZIk5g2P8YM2zUnnHAgSZ6VV7KLXa57VW97MfogdA73GMGE2R5xS6HLlKx4vxCHi9zW0XgoC/dw9JzqqefPBqtC9SuMalNsWf3r3ieqrQ9C0rbovyKr4Lm2drUosSg5IoWiqoCIWqYst5kLlLIwdTWQxo/hQxdQcM576YcSxVuIpYaOBjDl9ym+XleEpxlXy0wcfvzcGQrw/87X32b7YsLfun/9R4/+a1vsh62bC+fkXxHuv7ESGlDgzQnPH74Eb/2y/89u6snJA0kVkgIHB0fsV6v6S+fGATdH9HjoF2h2jO82DPGS45OTzg6v8cPvvsDjl+7y90Hr/Hp82e0R3c4Ojnl7N497ty5Q9c0BO8IG0+/H+i3e9p2ZeMvjtU6lGfKELSEwZTsFO99keOZrmkt35ci4zjSNC3O+aLoCvJ0iHhn8yBVQltV1DnC0RGaIzk5GtfhEMa+5/Lqmn53beg/rS9ZJiOrLjBF57rx2wQKJedYZbtDCpG21UJpsfBrc9ImOJwGsnoE4/VLzlvPNkBK8Yc5agYwkAxO01RYHZwnYqQd8SB8X/JczGUmTH+BqjP05NiTdxHn4KxtOAk7hnbL0WYgXkE4OaK7c0LYeLbPBzS1+IJOnIp5Uy3TUMaYGQt9V9a5Ds/ak6RSQ2UphSxCkgp2MvLbxgWcxxRfsvyeZkVjmvPKocW5QE1MJDJDsuvuR6WPMKQiG8WRMULhYYj40dahNB5pWvP044AUhKOi+KDsVZCsBKd4UVNgTbAcXaGrdziCazhatWiCOCTGMdNf95CU3f7H2fMqLQugLJalnKsOyWcAMMrhk+NjP7oIr01nZQpnVItuuT9FocBs7S0sTTdZdWWfZdRCqzcxu0MzgzzToqrgdkFm7+kgiSBMAl1mmTFn62TasxYOzKFHe7ak1ZNRfJaFk2fCSG89/3S/hf/NyWxlG0GsEXT6qSB5AcyAmm2hFpHWFzEFdnJVktZTaL7koravKnG7RXsGLUWhk1GupdFhV1BZcHZ0h4unL/jw6RO+sn/NhDXQOMc47Nnu9ry42XF9+R2uL5/D/oKkQAnbrI/O6LqWpjEGg9M79xhT5rvf/R7f+OrXaQg8eu9dXnvzAVfXO9774UMevPGAmzRw9eKCo5O7HJ3e5ej4hJP1ytpPqIUIjYyhso8Yv131KnNVylMfCwFxE2EuKKHx5JgNIi9hmqcIeOexVu4Jr76AWYSccmE+L4oDQyvWa/T9wKePP0GjITXVCeLC/B6m0IZWe8ju2SnkOVS3XGv2ipUk0yoyZVHWia8UYhogO1I2YmCfM15ziSRYLZcXsf5TZMRlupxYqWMljoAwiDB4Ydu29KUHWdM2JK2w7tG8bUrBtCg+jjgcmiNHY8+Rh7unHSebjqSBVRe42TSEO0d0r58jJx2rcEXqoW0cQRRXiXFJ5V9dGGQZkUxFkbrCVq9iSEMNDZXyRpyx0wevBDX2FBHriuwx5ZVjsmaWwRGCEQrn4s3paEXHhnI1KitF5hCyM8RrX1CdokJoA03rcC000ZOGaK1V0ohRVdn9NiguZzTNxdKueJomZxTnlbZpCCJkH5F9RTH+GAM2qrCedAtLUQ1TvK96MFqdjaJYFsLysDjz1Vdb2PCFNb4syuphLb+bFuys0KRcE6wSXRcCYo7oHWbR6pnsuWYhMCuXEudYDEDVOfXrCqRYNs28/ZypKghdUj0xUTvNYcTlkOislKfxKyFCZygoL4WGaKG86gRfel2URbO0QlSrAstV1U0Q/ymUWxFkCrVZonlglF5PNQPocN4TVgGPY4iJq36P7HtrrtwEBGXo92yvL3jx6VMef/gDUhzwjUN9i2862m7F5viUJojVrbWB0HSMaeDq4pqj9TH5eMAHZX1yxPPLHc+fXfCVn/gKL54+4WrX89qXTjk6PeV4c8S6C9QqPDB0GSolXOiYuPvyrMDzFN42QyHnIhAFnBe0cNQ558tMyhN/ZFYh52yAhOINaMoGpS7nnN5PCSsPw8CLZ8/KeR3qZ1dr9oHn+Z5YTMdlyHCaU/aOsixsMK2TAJxoaa5dQndq7BGaMTSdTuRohVTdlXCLGSpdTpyIcurAS2IILb3ztALXN+bfrdYtMWd2w8gwjKRxLHMpQ8CeIkOKPW2/5aiBs6MNx74lBQhBuLNpaE+PWN+7A2cruqSMWyvC9aKFgNcKlGuVZb3PSXZQ+tXV3ikl96eluzhS15M9m1PzCutideU9BO/wjcMHoWnsfWjO5Dj3/0oxEbOWdVFDo5YXS5jyqgbkShy+CTQrR5MDzkfSMBL7TGsQEVTAq1o36dIgbJaD3kiOUdBM2wSLzDhlHNV6pv04FylXipuaeqrb7B2VRZ4X/smh4zDL/4Mzv0KFlWtVRIHc2tuVE07yV2Tysm7d9HT1A+CFsFBg8/3NV9CD/Q5uc6EcdXqg+R6rl2PMBYcK7BDMUmH0NVRoYTaXdVJOWavCmc9Tx9HyWkZDZBBvTwjBrH4/13hJ7YlUFN1sWSyUsMNyKzrfsd66HmpkqapWWxIXEGMlmYVqdjyZHevVXV578A6//Pf+LuKV3/v1r+AItF1DaD3PLh/z8Mkn3Fy+IN08wQdFQ4eGjo6WZn3CenNGEMf6+IT15ojVas0v/71fIMbEn/jTf5of/oNfY9/v+fq3/xH+9t/6W5xtjvn2T3+d7/3gXU7u3OWdL3+V87t3OT09YtW1KBnfdSCOqAWC7hzBOVLSEtKzsN4SbBHTcuFn83J9oN/3kNM8vpqRDN5DUivedarGuafZhHaORpLsHSlGaFvEOzQmeo1s91tudtelPomD+TJPptsrYt7DfLl08K2UT+XWvrCA7OeIT0rKHqceJZKw1va+aWlXR3TdMY3r8B04l3AucycI971w3niOwhq9f07abLjMiasXN8Sc8W0g5czl1Q2fxJGHlzsrmhU4kRVNYf7Yjleshj3HCKcK3Y1xAI79Dcen55zeO+f8/ms0J57LOLK/GXGnK7pGcYwQbfwp+d8scUFXVowMKePhXQm3+9LHrDZJLY0kFYg2J3LJ9alYIbZvPU0QmsbTrRpTDjkS00hKsRQlW74NKX28HKRBiS6SNOEL+746OBOh8UbI264d466H3pFI6KgEhY2UDHTKxH6k30UrzHaOxguuKX3KsgFJSB4XPD17RkmMPZ9r+4IrL0taJp3IagxiysGyOtiWeu6gB9Hi/wf7Ty6OFvLO6YtyxIzkq8bjwZlkAUmQqlRv3aDMobDbhZRKXeD1jPnwIaagy1TOVI5d5NOq6zY9pc7hy+qpVaNvvsMSsqu3XbXrofY0LIpZiS4Yl5pzDu/CLXCGm3IZk6cqh6NejYs5X1WQboXNoYbAquc1KVu13E0qzPHmULg5BJwz7eaI06MTXj864r/53vf4ymtnfPUb79Aetahk+v2eJ0+fk/c7XIpc9yPtagPquc6Ok9NzXFjhQktoA77riAr/4Dv/gLvn9wkh8MPvfRe/aon9jh9+/we89eabkJUX2x2bkzPu3D3n/Pyce3dPLbTkIYSGfYwgjuBLMXF5o7lapmJ1WgZEMVZzaxZp41m9sgo40pLPQfPEXJJjQjVOXlAmkjVZ8jy48t68oRjF5lJKievtDduba/NKXAVDHUxn0Hn+LtNcNUR9MJvLP47igdfzzI6XrWNNuBTJvUPVo4ZdM4XsPeum4+T8Hl6ExgktGecVcZnTcc9dyZx52DQdcn6HfLymDUL77Io4RoOT55Fm3THExOV1X5SX4+xkxaox72m3g+PdnkAipozbJoao9CkgzQbftPjgaMicHrdsVg3jcYf4DHlAcWZMpIIkrPO1wPodRtfmSy1kaAJN0xCCKS/FGcuGZmLK5CGRY2HYKKCaarWGXDy7Ejqv3SRc8LgMRGWImTElokIUYR+zeWPGClwlHT4qG/Wob/BdzbcKKQwQoxXBq6ESrV+YAUIEXxC9FFJoh5dglGwaIUHXBHyXSeuWz7N9oZXXzEa+cJ1e0gzMobTFcbBAIM5OzYGQn1qJLI7Tg3PIVOw5LdrpPpZu1BL1txT+3LJYF17F4j6n006wSg4/Z170h6ddKIdy6PL35VbDeK/S+nrrL62qsZ6zeqPem0XnC8pQKvnuDD0+KEheGg+z22jjqXroFS6U1XQnkyCuNV25rF27saKCEVVW3YbGN6T9lu32BvKG49a8rrHfM/Q9/c2W2O9J40AWj/MdqobaarsNPnSIb2jWHWNMjMOe/W7L3fNz2qbl0yefIN6x3e8Z+j3nZ3foh5H9fuD07h3O7pxxdnbCpmtBR6TQMR2E1VQxeLcj51jGyZRU7buVcy1vqGNnudCcEs4bi2A+GDtT4Cmn8q4tf5aLIVRLGapnV6+RnWN7c8Nut3spcvHK5qnL7VXk14cBhPJT80zl/SrFW1Qkj+QBlGC8n85BsnBq17QcnZ5Zga4T1mKeBD6z6XdsiKzEuB793RM4WhPWAZUVcRgN6JUHFM92Fzm9joa8c57T046jVaBx0PdHbG52+LEn6h5NyuASY9ci3TESVtaWRBLrzYqM5dZyYZPPKkielU2uNWo5TQAiC/Eav6ehc0vjRpGpfC5n0GQFxWm0LsaugDXqKwhAEiHFYIjEkhu1kKQBckZV9mNizEoUYUgWqUhSio7Lq9olZciQslB53NR7kvM4MWx1QArriPXuStHyW0bKL3N3dFfCvuVZvLPeYCG8DC37UbYvvPJyZdCX/Z2WQrp6FTVMfJD3WQrPuqgmM7DWxFRwdo1R1/8vcj2f5eYd3OyP/HQvH7Zc/NWCXejtesRhlckcMlSdUZeTjnUVFGJHVE9tfsJFWLAoRNMzWsJRrnhdxpzhCoNG9bpcyWktFZer7VfKTR7eP9zysQ7uYfqpoRdMEKcSSsnF/RRXWjeYrcvd07vE3Z6//92/x0+8dYcHr5/j1mtONw1Pr6/YX9+wdolHL57TjyObO2c8vgH1DSfHG8R1dOsNR0dH+OMTfv3v/wpPHj3kH/3Df4hnL67odzu+9M7X+ev/9X+NI/KT33ib7733BLxndfcO73z9q5zfOeFo3ZGGPf1+tPcROtqmNSXvrDDYBeMuHG9uCjuJPXNOM6N313b27CkRvPWtiinRCtYYsYRqsmYDX6gaQawY/HkYI+IE3xrPnPcNwQdC15iVH0ekC7x49oyby0saOSxFeGmqHnw0w2oOZ1D9rh624GwpNVBeQVNpNRL35H0iSUv2bcnLWbHuatMh915juHNGvH/OUbsCgaiRZj+ySSMhj6Q0EDYr2k3Hyeka508Z+gGXMqSRtr2Da89wJ28Qs3kfR8cdx2tPVzp43ry4It1s2V6+IG0HIpkxKL49JqxPcG6DNMpqtTEW97FjnzKaBwuzx8FCs1K4CVNEc8KrsWHUhsveSflxBaYvU+7WOAYT9CNpP5JjpgkO/EI+CaRgdVw4yM6TnUdaD42SkmPEcT0M9GOaUNlZIbtArC1bVJEhcrOLHIXIphVGYEBI4sEHnBpFmCv3mTMM0WoQvVO8CkGM4T5Lpk/GMi8x0wRLI3Tt51M/X2jl5cqCR9WUU3GXDR5n/0q2xLC1CneThJTqVRV01aHyqptMtSGVq2sS81Lduc/QXNXDuu1iTF/Pny3DME4cqVLcwAxWEFv21cqeQjDILATkQBzMoZiSoa0eoiwuWr2oKn+0KBs3Pyq5aLNZjdh1S1bdCFIXrdqNncD6SLnKseZeVmJy4PIWAU0ZawrXHrOwnAShzvU6KVuxZ85W6FkBG0600NcY8eib9094//0L/vu//yv8b/70H+a11+9xdHLMxdMXbF88ZX/5gqfPnxHWDWxWXIzOhF7bcXK0QdsTcJ7tPnL15GPasOKNt96mTwqauLq45O//yndwLtM0LZ883ZHGnpPT+9z/0pd5+603yWlPP2xJY0RdAzj2+0i76nDe0wRvz5Myw7At9VVmRAXv6fOelCJdt7J+XJQwa56VtKY0wcedd2iK5nGl0vVNFR0SQc2Q8wl8mFtwKIo4g95fXl7S73bEYSzsLJb79EgJb8uUb156gUv/2DHD0ad3WIMPzhQprmrngppr1FpneAhOiVpIa3Mk5RFRD1452Rwb4CAlQhMQb+skDSN+tFYl/c2NtZlR6LZC745I3coMtDggbsVRs+HBaSJjymuzPqb10DglBOXOa3uGfc/V8wtyj6EoG6XBs+5aVl1DewxeBlQjLgUkCSS7b5fjwri2XBSx5BolmxwDq/mKyWrwRpNVSa0eKyclR4XBarLImTSaPZI1M8bIqErrHas2IR40BHJo8M7Cc95baD+pcRPasjIEa14YJIISgathwG8VfCaJee5xzLSFPio4Z0axlC7OmtGYjMYrKOoLN6Sv7PklIpDN2/Y/zi1RppAKNc4PJlSxoqPSVK7qj+qFzAtt7vWz9CoWV5j+FVl+JPPvJktfOlIrEu+3U1wHVuysGBcq6MAHq5d/6axa7kFK7YwcAjFk8f+l9Ts9n9bn14PzH+QFZQ7JIWJhnEKhI7L8kbJLVVhufgdTvqs+3cJwN5duNhbKT3mc2SPTul8hK81amk0y7Z80o2JWbNcE9rsbYr9j1cK983OO1hs0Ja4vLtjdXDP0O6Jmgm9Q1xKTY7Xe0HUrmm6NNhYq7Pc911dXtF2g69YMY+T6ZsvV1RXXF885OVsh4rje7tgcH3N2dsbd0zPzgsojK4IL1h04ZyX4xhBzWtgXSt+oykQiIiVkaEZNEwJTjykobUDMYo5xwVM4hajmkg6jI7KQk5TxQrC8pLd7cN4QjNfX16RoeTKZ39LszR9Oofmzadelx7WYRNT8cFWYizlc5rEXMQSdsyJZm1LGxo5mcoqFx9CVnmNh6k5AkMmk02Ykp8Fq/wZI0kCBobsCImqd56gzISyuoV0d48HY2ZtMux4IfU/0LXksOZ1OCAqtczTe4VdA2kLscT4hY57CtXW8VIuhkRIaIy6n0lWhDFmhjsopIUlRySQtCNGyHCTrtBa0GGxRM8NoeTV1gkuKC6Aho41CZ2vU+C9LfKUcWyMggs0BKyAXA3Oo0sfETT+ClJBntK7P2TlycMWgN/O3hpt1jOA8yXtDjjoLFWbnLDxJKVD3P8ZhQxYhqXxQ94Sh1bSs4oWQXDI4VLRT9Vpg9sAOFt2BdH3FPXC4Kxwyeiw9r1qUebt+c9ptdphs/zkrfiu8VmddVSazoquHzH22ZsFvDurCNtaJj6BY8mXZlw/snFavMcP+Mau5CVOifwoTVsb4Os5TFXlVaOW7hbl+23dVLFZSPYoJTar13k3IW44rld+lfG/PmjTjJdM2Dffv3uX9996n317x+7/1Nsd3ziGN7C9e8OzJY4bdlphGms0RvXSM2iBtw+b41AhyfYdrOvb9NRc3N+x3N3RH92k3J9xcXvDo02dcv3jO6Rq6tmWIiZubK9759rd57d59zs9Oub66MASma8je07QtIp6UYLVegyr7/TXDyKSUmlJ0KuLY7baogg+W1O/73jyznBjHkSr5+9I9OdRxr50LS72hjV/CS7CaIkpUInhC2xTiYo9K5MXT56W4WcrrctM8rJROt6HykGcA0+L/t7dpOd0yYhzFU/eW2G8aZchuquVzokBi6HeggvgA4mi6bKAHh7VESZGURswbNY8mi+DXnZHZBk+IBjBCCi41NLimJYfjaZ56n5Dc47oB2paUGuse3AVUCj1SUaq6b1Hd4sIWpAesEthVIEROpFTg63HAF2NEypqjtkRJGcZEFiPCNURtXaNS1ryBk6IqQ8r0vYUSs4AMI23jyI3Rk2kbkK4liKNxhr0VtRowY9QozDcCTgVPBVdZ9+WbcbQ+ZtkKrtsmkLwnNrUkAFBL3aTR1m+MQusF7xraNpjBVSiw1AhVZ1KC3+X2hVZek8Bl9mZqXgcsSVjReeahzUS4mdkudItz2fYZocCXrq4v/br8lnJfL/tWi79lroPJy+M+8x7KQi83fHC+hbtWobhVIxrbs5batipWKqfbIZP8UqmYV6kTKkzF8oDiZ3ShOBMAtRjZibF8TMnoZdiwXO0VZkDhIiwaCKMtmsAXmqcRsXxXJmkmJmOPJ831RVUU5pg4uXPMH/2j/zj/p//jX+D8uOV/9U//MU5Pj3n6+BFPnj+haRu2g2cbEyoduxTAt9w9O2fTHSMhWBJ8TFxfX/Ls2SfcOX0N1czN5Qvef/8R/dUFPg0437Dd3dCt13zt6+/w2muvcXJ8TNMGmsHqy7wLrDYb9kNvDSOblmHsSSmxH0a6bkOKkXEYLKS2eK2hdEFWzaVmy2qwDIloijwEPyM6c8Y7exMxZ8Y8FOAAkEbEB1wI1mOt7XDtCgHSmBlu9uyvrgrisYQcD4N/ixk6q6n5i9v7HExfU1Rq86QRwYu3Pm+NN/or5yA3rFqhHwVSUb7Z4Odjj+VQMA92v70uyssh4w50NKh4HIn7PSmOoEJ3dIpvW3zjiTfX5KEnjQNowoXG0IPH9/DdGte0iAghJ4JmQrBiZ0LAdxt82yLeHiQPA9FZz7+bfW+FwOVZs3Mkte+GcbD3PfasashQzKe0arAaQdVyjM50UyLWx0ylsLUXTydVYl/zpgbJhOwL5dSIaxpEBfUBB3TBEYODMRNwhNAUTk2rc3NkXIKu8zTBIc48LgOHKH7VWJfmxtEhoAZAiftMHyMxCRqEoWtpGgAr4xCvjJKsBEQOJe7vZvviK68q/GVhmQBmKs4EtbM7slRUVaAu0YA67Xew8Kq3oK8e8FcppXrcbFyWuMhCwWr1hmT2AF31fFiE6cpNLaItRXnN9WRSw4ULT3EJUNH6HCWiugwGVcaf+iyF0Hoe13ItKiqt5rPcAqkmixxX8Yrt4FptX8d+UkMHY1xb0Nffa15rgsWrLe5ckvtJzfo03tjFGxAQ8dw5s1qqH/7gPVqfuXO24e79+9xsbxj6PZqVPkZDXiHkKAVwYqzZ2Xl8sIZ+H3zv+2RV7t97nXG3Z78diTFx8+wx7crjVh395SWrs2OOT464c3ZK03jACEnbtp0UsAsBhgHFwpqx1OFYLZxDCyuJKW1QMQPAl/jSOI6F328eG3tHrsg3V4qMF6S1NWSVzTOa2ouAPWN5f6Bsd1uurq+Y6iTLm1raaAftfRZBAH3pg1ueV3EcJiPtwAU7mNg2X7zgssOr5S8tGqDkcSxsJJFlUFwAiTsopL2qiTT0xosoDjf2hajWsb+5Iu93pO2WPPa4piF0K7rsCJuIth1N53EMhhiMPTVHlApzvZNCnJzVPJ9kYdqDpzZoY4V5LhjwWawROZAZy9CHveJytoLuUFwpf4Cajs/FaLU1Ut+3eXwko99qRFg1ntwGiErTNDRtw3rd4dwclpWYWXWBpvUgiRSt4zIOmsYRgsMXflJJamOdRii8iUkdY4oM0dOmZOjcmBmjnVsw4+rzbF9o5VUnvPEem9AW56ymgiKAF3wRkyOyUFYH1E5LiPUt1gr76FBAHmSObntB9ZrThCz/y+mWl7eAkZYP3ay55vte3Gtd2K/06Cb9KpOSPAg4yvxvWVLzHlUZFgGx0FpMyE4/Q3m991PeoApeX2rvZPqv3JRUhayHA1XCfBRB/FLRdEnw2l61xUnxvEpOxwqZF2MlgrjA/fvnOIX/8e/8D9y/0/H2l+6zPj3lk0cfMex2eBH2w8CYLbcwjtC1Ad8Exqw0COI8LnS8/8F7PHjjLb72la/x/e/8EvsR9v1I/+JT1l/+EuJbrp4+5v7JG5zdvcPpyQneKSmNqGaOj44Zh4GUa/mFjY5zroT9oGk6ZjZ+XwbHlLNMbBPWEsU+z1NHZBzF46q0XMKY4uTJarXQK6mqhGmONSGU3FImoVxdX/Li8nK6z9vuU/UEBEoso3rGtc5s4QEvpr9OTz3PNTvWobrkGjSjTiVbLs4Zm0nMxShU0HHAwP5izTxztOeLGfK+UEepMeOXEgMXHDEOJZfk2G+viVdXpIsXjDdXuLYlrI9QCTTjSF6tCanBuYjLkTTuyTKSfSQbAQchN7gmkMaRNAykfkRinkLz0/vEAEykjCRDgM6h9cV8oBLuLvPzdX1YramNlyNLKp6aHVeNuwPFWXOj2ZR36xybtrHwoApt29B2bVFepvCNpT6yWjW0rUcllHBnJntoW09wQnBSwrmK1DBtzOQkDOro44iLHj844jjSD5H9EJFk7WD8jzPDBhzYbGQVEvXFmzVg5e262HcpVGcP7PY5q8I4KDyuXsFtLfWq+ykK6KXXs+itNR8ze1oHik0OuQnc4r5nf0znG57/WQRIq3ov1De3BNGBaheoZLwOmVoyWJO8wlHoK9GuowkWdqrhmuD8rLwWlv10n/XetT5vAe5rBRzMHpeW1zeFXgpSyRLUxhSgJQZf37AXJvSabxo26zOeP3vGX/t//N/5P/zv/3e889YDLh4/ZntzzW5n9UveWQHlgHB6dpdRHSoNR2enZPE8fvSY999/j5/+vT/LMI784Ic/ZJSGF8/tPG9+6yd59PADNEfe+eY3OLt7Trda43zDOPR03Yau3RCzWliqKCDzxOwZnAtUTsjtdmsesXgTHMVbSmMkDqONQ0rF8zL6oXG0+q7QNnjnEbW8ihNHjj05jpZHiRmHsW3gfCGvBecb0jgw7nckJ9xcX7HbXgN6WM+ls3KaZ2t9t24B6Jnft5R5PJmRtyyuJRKxmi0xiRWmq9A2jk4CY/LkVOiucqYbeyIDSRzOtYwSTYHlkSGZZ0pWC5uqXSXEhPY7crT73l29YLh4zvj0U9Luhuw82rTs+z3r4zNWRxvk7jG5wbpqj3uiNkS/JjbHyOYU33b40JLzQNrtSP2ePI42753D46ecobHARLJGBGNjD1nwUYw3MlntVOsd0gbLeSn0ydYhDpyO5IrwzVgYVXPhqgxItnB6ch4XHE1jjDdTmM+V+rUuMHQN7cpyUu2qwdGQUzSWlSCEJlgYunHEOBqQpIHGG5o7pcQ4RsIw0gwjTbZSjdEpOxI3w8DgYNTMMIz0Y+Rm30NUjtvA6sc95zX/XoEGZuUrhyShNeFbnYmXz8C0yIAFz97kkE0Ld17AeusscsuxqGG52wuZ6d4mL2O61gHYfVLEs+KtCmGWAof7H45PPZubRMPs5XkWvHMLa3kZvagFxrUsQSpnoZvbm/hF+LDWK1lh46GXdQCsrGNA7Q2kk0dlIC2l/pdz6blbQRrl34k9HoPze2fj2bYNP/1Tv4f3v/8DHn78IQ8e3OXk5BSHcPX8Kfv9Fs2ZEDxPrxLJWa6j2ZxYOUUBAXz80QeM48j5vXP6fs/1zZYXl5cMNy9AlVW34tnTJ2y6lvX6hLO7Z6w2G0Jh9U46mABrGtSkMZohDiN9bzVATcmrqCrjOM6emVj9WhzHIjjmHGDOarROKMHZGBpoxvp4aU6QS0FzigbNnmaIueZZM961tE1LLt9nFS4uro2Bos7cya6ZZ9mBAqvrqyouWc66MmcXoe/DgFr18pnmgLU9sbCawf5l8vCcN5CJpIimESFN4A7NZvnnsWcKU7tiwrlaVlD6KSiQM95lfICxc+TRFRh3pL+5slytjsQVuCgkycjYM45bouwYQ0/qe0K3IXQr80ajKa1cFAtO8TVbqFaTpzFCpejyNZoiE+u8ZLXOzd46SgvLXNj/m70/ibVtS+96wd83xphzrrXLU91ziyhuhMP2w8ZOMtNkiiQblAIsARJYmFYKeGQqW0gI0wEJCRoI6XVouEELgYAG9JBIpEQgJfJLkZkkprIxdjiKeyNu3PJUu1xrzjmKbHzfGHOufc4N2xF6T9yMN6/O3XuvNcsxx/jK//f/rD7SaTbZ+0LXGVCqE7LH6PO1caQ4h3SB0Hl8ryAVRMi9x4vQe8+w6QmdpxtUyae5liYUrT0TrCO35s0RrUfLWd28mJPV5WVFaDqhF0eHI8ZE3Gu0YZ4iY4zspghRO1SXg7Ycv/XtM628Drdy8PvSjG+xGH9z2+ICHTDSrxVY23Px2+4K/rWSaPpJ7tyFnf+g9bhZow0iXmMs9f8rR6sqtIbIOzhXvexy440waqWYm3J9hR6vSKgFsr2gCRel5Zfmkg1p6OoJTJDcvaf6S6HB4S1cuHhepdV81QLWxqKR7O+kEBcNEVv20Al9P/ClL3yZ//cv/D/5znvf5kd++Ats+oE0R26vLkizQqczsIvA0NNtTug2J3jnSaWw2428ePaMru9449HnuXz6nMvLSy4vL8j7G4bQE7znxYtLvvDGI+4/uMfR6TE+DDhREIG0FjMqoHL1FnNkMuXlfaeCypRXzkXbUJRilu3EPM2az3Im5gtaC0i2ujyDI3uldNKYVjTPNJKzFicXC4VX6i3tDTaQcsSJYuaurm7Ic2rKKy2TV5P+h1P6pe3AoFzZZS183giglzm4zIFsVFdVean73boqOB2Dgii7eZmNFd0r6jRp/ZTWe2nNIQUbu1W0o2TIEe/AdQ636WAKlKj5qnncEbwQXCGNA3N2OMlIHJn2E3MJTDIxjzPdZqLbHlGCxzfDVhu7CqUxrJes87WkiOSEw5CGTkCMysxCDR7wXiheGeRjycSiiFoxo1PEUXwmdx4fwFkPsIqqTAVFUnYe3we6zre6qpQ0v9kXR7/tNX/VKc3YnBVuX0xxOYz42Dur5zJD2OZCslIMbYGixm7nhK4IY0za9HLOxJiYU2IfE5Kq9/a/KK+2VU+hLDJx/e0qhFU/+S7n+g3+XkXaDr4/WNjNg3n5SgUo7uV7bN+tPq2Q/rpLOdh7rTwXc7jUxVqLi1iFJFfKcn2hg7GxHFet9ndeLbeFr1BtYvXADOVlsGOsSZ4489wO1Hv1/lZKa/0P9b4AXfBmBS4eVyHFZPRRqFBGxzGKcHJ0ysnpOTcXt7zz7Y8pRfg//59+hifvvM+ziwt2NzvOj7e898lT3v3oGdv7b7M9PWVzdMzm5D77KfL86TP+/b/7t/zvfup/j3TCJy8+Znc9cvXsGdfPP+KHfvQn+eBbv87li0/44n/3Q7z12hucHp9SusCUVOiELrDxp8y58PT5M463GwOW6MJVzkGPC8obl0s1mhPejK95PzHuR2KMDJttm9Q5F4a+JydtdOmdaLhngv3uBiVeEOZxUq+rFPChzbeSE51X2L0LPfvpGvGQs2O3mzQvk9e5xzopFm303cpLD4tHynf5bu11LTHinFMLl+YpKSdgFmTYajuW4qBEZN4ru4vXnIxY2FANgAC+09yhR70Va/0iFFLScfFdx3B6RpygjDMpRiYyLkVcnJnmmSIdoOSzabdnmgv7dA3bPUwjLs4wHEGwoq1iRp09UzHllVImTRMSEy6bchCxOigUwJAi2h5H9EMcITvL26mi8Gh40YkjDF4NA1EEYZwSo1N0qes9DB2bkw1dp8CSeZyVQ9M5gu/pNr12ZHYYqbN6vORIwVOckOdoBpiGbXOJpJyJOTMZxLdGaBpALWXmKbGLmTkZmARtgUOB3QQ5/S/K62Cr2R0wcE8FYchaOMvB782lse1AOZQa3qqelXlBNeEKhxrojqJa79MUm5i302TBolidFf1V4a7PcZjreomeh3UFjp4nG5GtKOCtqguctaps4RtXA64GeJHqZbF4Wl7j5957gveGsnIrCiinuQWbwCLGXyaVmf6QdaFuVQG10N8qJbcYIKUR0dYiyXZA0edwxgUIjs997gvcv/eQX/iF/wc5jTy8f0pXPHG8IZeJzfGWy8sr9rPDDWfkbkB8D3iuri55cXnD7e0tb37+TaY0QYYgHU8++jrBFx6/+ZjvvPNVimTuv/6Yx689Imx6phKJVxF/dGLzJUHxBK8hHRHwJkBImX4YGitBNqUVU1IWfjIpzoz7UcM3PmgXXgMeaNlH9UqEMAxq4edEF5w2ooza3LBWt+ZStM1GUZDLxinb/1wmfOjZ7UdudyPO2SEoamw9p6uiWVuBB/O7GRJlmeirnWsk48C4EzPAjRjWOa+eVEI7YTMbuwREGVWAZq0TUmNAUXBJIHvRAuU5G2+gEtdqyDmS48w0Z1xJ5DITi6O4Ht87um0iMxJlxnce33dIPxDFI9JRSmKMhXmMTFNkN1s+zQfEd/RhUCaaogAURSXq3E8FFfQxMqcZnyOBbDRqTpuQGsjGg1oxKWvY3Qmt75cIJdX1EMkpI0GNRtcLbDqKN/TjFCF4JDiCoQNFhDTXJioqa6Rkajwwz5GU4tJ+R2k9lYm+5aW1Yajzjq7v6YZCEk9xiVQKrkYHZnARXIYcUwNberTIuU8RZev93rf/P1Fe5U4OSmrEqvky62F6lVfV1loNpB3GANtB63BdC0O8wrPijsI5AB2y6Ls1zVQ9b93zoMfYnQPXqlaV0nKDB0XAVfmW6gG92i5uam9tQdU+XCv4+6vCh0tR8hJuXDgNadc9uOvmiEl7hrthwwPPrKHlFoW3tkgcDlzg9OSMo6Nj/uuv/ipnp1veeuMh027PNO/IeSaEnoubkTEKvj/W+ibfUYpwdXXN7c0tMSfu3b/PFCftgTRN5DhC6AidZ3dzwcmD+5zev8/x0TFaJK8y1Rd1dw0voIvcCjQVOaesGt7a20+GmFSGEHu3OZPSTEqpsfJXNnmA4J3mtUq2NjTBhJAZSDlbDVRZxgm08NvG3zkVcomMdx3zHNnvdm3eLGG+lTffELgr5bRSYi8bVSwTv73wetxS9rF4664BgzBl66z2iDwbFLveiyo7BRU5CrmGXlZ5MmWQKFbQW1JiTlGbJ0ohFfUmRByuG5Ckc7UbemVA6TuyC8oRmCEWz5y0KHi28KwqkgrL13VGgVySonnFNxb5ZPx+LicoxXgrV/9EayRLVuVV/KqBJfXdliV3FhOuBOrK9p0yWLiq/LyNqa1lZ2tTLOFez6WyTOdYDclrNFPBWoVi4UEtPQhODZUQHF2nNGdJLG/nsxmsBR8dPlsz2KIGtMdqywwt/P1sn2nlJXKwtHg5RFEFvzRet1c5qofqwI6qC/XOAaoUNc9wJ2v9aXfZ7u67Ocly8H2pq3r5UgNqTd6riDx0GpsCups4bwqkovvuXrntpPvZwBZ1oRakoVgAs4E4xBacJd6b51Zrw5baFSkVLpNe9sKqF7U4VCpwqCFDDblkK8IszVxfD5rj+OiInODy4ppf/erX+b/+93+C3/ZDn+P9d77Bbn9LzgmZZ77z7JIynDGcnXN+eo5znmmOfPThEzYnWzZHW4btEZ98+CG3Vxfsrp7xxufe4PL5C95/933OH97jwePXOL93DylO27k7z/H5hnGKxClTQqAQ2XjHsNkw5REpHike74PV4hRL5Ot4hC6QxpE5zsQ0W0NK9VpjimrZWpipJC25CN4f1E3FWRsPFqOXqtBzEEhK1uu7Hu+CJf6FlGCOkWme2gsQRJkhDAK/vLM7SuiVU78cTN125NoRsym69JozgIXXouniBF8isff0seDnjCSlHSoI2Q+4QQVz7ga85ctEMLYINedcUVg2SVGqY9RQnhMh1vsvkH2P3ziGvmczbLSA23ti0N8LieIGYvFkEhJEab1CQIJ2l665zUKixFnXZ4ApJ/W69iNpnPBR+615Q8WGvqcLgWD3TcrkWTsw5xC1j1ftIJkTadLGkMwzbo64ocP1PdI7vCuE7IkpW87PqRfri3XAruAaFFCRk9kNouAg48IMncN70dwbME2JFBOkhO805CzO0296nE9E75GYEK/KahMyU4lK2gv4ueCyIiw7M1ym/AMMlTdg4cGyagKbxet62d9Yn6SuJbNfDtwlVt7Vcs07B3J40LIdqsq1alr7b/W45e6FRUGtCXLX8PoKsliQjGJOWf17fWV9kINwjSkxPaR6TaZ4jHDXVYtt7WGZJyCO1itqEVR3vTpb0KVUXWrPd+g3rnCFVI6+XKqyMkqdXOu6lrEBUdZqoLjEyXHP17/26zx7fsP5G+ecHg8MXrieJo6Oz3n64oJff+c7xEFzXP2wYb+fiSUyp8TxcUcYBpDA9dUtH3zrm3Te8frrr/ONX/1VkMz2qOd4GOidIwhsj44Zp5GUMrfjSN8PFk7VkGrKhdvrGyQEIOJK1vyN6zF/UUOHKWotzK0yyXd9T9f1TNNEnCOlQB86tdBJFLTGy3tPSbPWb+WkYR8L7dTZoEoyNVh8PwyIc3ShJwwbPvjwI8b9iMNTWmFGLZWoU2o9t9fMH3eNxpd5MyuPS17v2bwAO8rqi+qZq6nVO6H3hT7APhXjrPT4fgPdoAZWCMg84XzG90lJYVsOLSHFmdcBLigwQdGs6hLo50HRshSCD0jocaHD+UFb/biEPzrCT0cQOzywOTqlGzaE0DXgRQFynBU4gl4zzlG993EizSM5R4rBxL23QvhOm4EW0RByniMo14UW0JfaAbmG4LWYeNK+JUx4Bra44DkKATobxeSYp0qEa+tK1SupZIiLB5yzhVtFtHGsvYeMdimY55kyzbjoKCHjOqMV67TeM40TFfHde+G4d/QuM3Qdbsy4OONibsCW6ftzvD7byqu2A3mV8bdWXMuHy3Jbezoa1pP2h6zO8CrFcXjOO+cvhwv5VZtwmIOTg7tdaUy565GtIMnrJ6+HvGQIr5SJ3LnvughKXdW0EFWDxzeltSixdV+uNU9ktTpbuLQ2Omt3UVgQay+/tVIOhWArRF4rLkNg5jVKqUiLOp2f3+fb3/4a7733IT/+Y19m03eMu5Hb/Q4fO25uI9czHG03+H5DFzqyOPa7HftxYnt0DK5jP808+eQjzQmQub29IeaRYTNwdHpMPwxsNgPD0AMVtp8oWfBH1kYnJTZHx8Q5Mo0jne8sNJgQggnPZApZkXLRoNaudqC2t1zfhXeasSw5t3khUihRa31qz6g2jg0EY7PLBXwIhH5QQYmWJEzjpPkT6qEr12ixOg7fcz1nvdTq3cnydbuXNo/bElmtrzpv7X6Thc1IxvaA5YLFvG5xhG4giQlzESRlxCd8GOxc1jU6z6qOLaroXSVAyqZW9ZriXAttNYOt1iyatyddh+sHAyY5/LBRL9Y85PUj13C3GEVTmmbSOFJSROH6Oo4iRojs1HvLonWdjcDXe5KINVktjaAXmxcaYs6UcSbMA14cISitU6lMYOowga2dbPLHOdFjDTEkWXN0zqlR2Poe1jmaI2WetcdXCvhcyY0dPkCJFYGWcQhDR/tOJMFUoGgxvMNysN/H9tlWXmicGLTCf9mWxVaTyGui3BYR/BRltuy53l6hiOTudythvt5N7noki7J5+Q4qpVVuwqmi/lciY3W9vNzLXSP4VY8Baq3W0KJJFKV70sWknkMVmPq7rzRQTanVEMRKidXHd/W5Dx5yEXAmvBqjYgEN6coCk685ghSJMVk32TshKNBWH97R+Y633voSv/A//iLvfedd/vx//39henLJ06fPeXZ5ye3lyC47wtF9wnBM32/p+oGC4/b6Qy6ur3n0oz/OuNsxvnjB17/6K/z2H/0K+/0tX/var/Lw4RGn5/c4ObuPRzi9f4/Tk2PG2xvmeSTlgkcBLYXMftzxxptf5ObmhpvdLZ14ouU6+s6RciTmTJyURDaa8uqHQVkP+o7b61sohc57CB5nQj3OM10XwKE1WnHUnFezF9QwybkqNM1xha6nGwb67ZYijpgL+5sdyQZ3HWXQ0GxeaZwFX3gAVJLDCacKsGZg88HUdHbInWBGmwOlYKz6GhJjmgz2bwwSKJeEuI6u6zHwmym7iCs9NXeGZKREZH9LidrSw2eVGUUK0Ytdz1hHnLN7yBRXWZjMAxGnCrLr8Jst5IwTj98cEaxov3qqDq3z0h6DapSkcSTud8y7G81XSWohW8QAKs6Rq8IomDHkFaDh1EuKpqhcVo5SHzScmTPM+8gwJjrxdAFyJ8RUiAnirICPFDwFR5GEBK1Bq80xc1KF442GTAFYmufMKZLzrOUJ80SaABcpndY6Ou+QIKQQKDnptJFC3zv64kgIyc3KbBI9ErWOMHx/uuszrrzMeshWY9CEW1mH03Qrd36qO7wYluvt7jGftomZqq/KU7/qfL+VvRzLjVVy3JWj1RBDB1wa1XHLK2/n7vlXnpKzPEP9XbnKVDE5y3U5t0YVagzcIiR6b8pd1GTc3YetOMp1oHCtZGvHZArW2iSTcmSeZ2KMzTLM7bC1l6pAqdPTI85OT/i//9/+JbvdDW9/6S02MfLs8pLrq2tC6Ph4vKK4ntOTY7bHp4jv2M2Rjz/+iHsPXuP1z32Jjz98j48//IBpGvmx/+5Hee9b7zKPt2w7z/bojKHv6aVwfHqOc45x3DNPM8GpB3a9v6a/7ej7ns3xMdeXL0gI25MzgvPKAJMSl5fXKLJO8wrjpDVZfd9pvoWi9V3zuBSHe4eUbN5awqWCz4Bk8jRqOYVXGh5puV1nghVEHP1mq+z4CY6Ojxmvb3nx4kK9NymwMgDrq3Tluyicu3PrYHt5UdzNda69sXV+NBrNU5kyxKiQ6uxx6i5oDqrrWk8zYiR1G4q3XmbiFHKeJrLPrW6u7yIuKjnxlIuSObtCKgnSpDDxrJlkcdpYVXxQtnmH8iweZ0XOZYcMm8rstMRERfAh4ErQXGKKzPPIPN4y7m9xxcJmYqUKXvvg4RxZvHqDxWoCk2a5kyvKDTgn5jnjEwQcQy+43uMrwl4wSq3ExqMdCnJimnTcrLdmI2vA6ZyhFvzngoTQCAcyysY/T5EyZWTKMM9IgiKJOBc4OVbDNwSGodewZE6MczIyXo0ixH5kDig9V5r0mb9P7fWZVl6UYvmU0voAqVGvq61yETYwR7NKdZMarmhhi8NluBb/DRW4CnH8ZrZX9vN66eTr6+lFPs3ve4WubXuU1YnqU9bC0nUYqJ5b+yCtwBeVXNcbwtBJU2C1V5KrSo+KJqzgjJXTKYeCSS//sltoDtfK26oe1xqBpz8plQt/NQiKnsGFDgkDX3/nG7z5xj0+/9ZDdrdKpRRz4tnlNXQbQn9EGI4Ytsfc7keubm/Ynp2TxXF9c8OTTz7S0F3OvHj+lBgnnHdshyP6boPDk2JqXk4uyrdX0qxgik6ZvGNMeByzn8EFRII2GLRia/VuAwWYplHrkszKF1EmhjhPasU6e+CcrP4pUUokzoqwDc6YG0RhyMlyLfqemgvcBLHretwwsN+NjOOo9U7ZJG/zssxMqAP+qVNd7vw8NBhrGP3TTqNz5M5MLygq0v7lWBuOqidY86fFyKAVN+UpPlCcFnuv0a90WeutciYUh0MLhRXWrUhAKeolgXWeELE6sWBQeK+CXjrCsNWeWlmQrrecoCHxbJ6LaiaNLcSZFGdinIhpasXJWh8ZrCuDIic1Z6YlHylrXVQmK61UysSUmbMabiKeEqQR2VRSG7UCdf459NljtvYnovMvoGE7/JLvcsVQjOIoQde1RRNJc6bMiRIzkkoLr1KKdcNUdg5XW34LStZrTWpDFyhFEbDJewTrIPB99kT5bCsvFpFdcT7rOq8G7a1RuVV0rsHJgZWof0lpyOr/B3B8O/6uDrurrGpY96X9ltNST6udau2Kqwhba3Ypd8+zMGkvey9ootWttzFxtYC4hkRa6M/i0+vcVm19slZezhBiNdluSsute6nYOMq6ERdm2a9eyaK4lt9bGGMFIc/rFgrFIlhi9W8OCp55dnzrvW/xO3775/lf/9gPcXm1J6P9jj58esnJ4y8wbM8Im2O6fst0fcvF1RVf+dEf48nHH/P044948snHnJ+dIxTee/cd7t07YbvdcLQ5Yug3iGg/IozwtuCYE6T9HgSG81PtsTRHShZSl1AoNszTiNbSCV3X4X1HSpmr/Z7Oh0a3RSnWEmVUL0BEFXdSZo6cle5pjpEsKjBSqs0lVXk5MQ/FhTYnne8s59UxbI746PlHTONelUv1uNZrouqyuxPpwLpbzb4DJ//Q4y/FPLssq5lY1+wSO9AopBoqKG0jKRVittDgina2eKEGExEhh9AAB341fymCT8mQbg5XoqLfUsK5YBEbHVvntE7JFRDfIb4D32m9lKDroXMa2cgFQt9YX0qc2hA4ESQEKj44ppnZlFdfEaPWq622FNK2QpXFwmvKT6ESSimWMnPKzEat75yDzqvOK2L1VRoJKblAKpoHzFojl33W82Ulm/JiOS6/GPhpThpWDFkVoaVQ05RJUyJP2iSzwe8F1vH8mputDTHrc3nv8V2AriP5oEaA63DhB1p5ycGvNWfyqr209tcUw4GCOdBoiwt2Z4/vfv214F5dc/336poqFGpPqyoUVPioMl6S4oaJeuUla2hxbf+uw3M2zQ2PYZ5SY8gQi2ujnuvK86pchaaVzONaGAHqs7l27EI99fKN2l81NAg0bQWt8FGLkaN5LpEYo3liNG3XXo/dQQHOh553v/Z1vvHuB/Sbnm3XEXLh4vI5H3z0MVf7EU4fEdyWoTvi6PQ+//XXvsbjx4/53/6On+I//Zf/zHfe/QY3lxf8yI/8CL/6K/+Fq6sLzu6dcXx6xPboiOOTEwgw9FuOtsdsj3v2+5n91RW7/Q0nm4FN6OlLYJfVC1PLsiPnzBRv6QcrdsUpTdXugjkqm7zzCpn2PjDtrpmmPdO0x7kOyVqvNDhl6c5xQuKEN3LefSmI5W9KFlyx0C+erhuIJYJz+GGL+KB5kJsbZmsTopB4BZCsc1RYrliBDPaupKkOXlpBNvmW5WezoXmO0ryStm+LBAil1BxUQRKte/IkTqmfSkKIDZRSaZ+1BUhp4W5KpvcdxXlqsbAkB8XjUQWHRLwv9LLR6+RE8UWBBPrQSKdoxn57rApCClIiTpRSoojgpFMlkzIk81walVWgOPV0pnHPPO5J097yZZXw1uOskBgglTqzHbEIqRjfIqLk0TkzJ+WhdIZSqt2IS8zMpUDMcKv3EmNmnjMlO3LKhK7XHmG2loeuEBAkRUKM5GlCSiGLg6MBrItzjFG5OMcZnyEEIfhAN2w03FkEEojPuKAAl9E7IKilOet7jRn22VGkR7qBYfj+1M9nW3mZz9uUE2vSUFUGa6CG/lJVxl1XjAUg9wov6bsHCe/6a4fKq4XtqAt3vfiXM1dv7uBsppRVcK/cMVn8xfU9NAxGWZQMZol6LAxo+Sxf225IsUJkaaFCMcUlUj+nKalXPK4t2uWJm72+MgbU29K6obxCx1XrtRbiarK6NDqlysVowSwz+IuiASPsb3ZcXFzyQ195g2EzsJtmSopcXu+5mRIn9484u/8A1/Xc3F5zcnzM1eUF/+X5E55++B0G7+jPT/n4o/cIoXB+fsL5+SnbvqfrugZcCSEQOkeKReupUmKzOabrAs4LRaJ2oQ0d3WajZNAidJ3yuMU0k1Ih52TchEIfvIVUZoTC7vZG4e6r+VCyFtcqoatZzzk1peKUSoaUk1m/CtVPKSHBW52SV/b4UthdX2thbX1b9WVZa5ka4a01ZodGYg0bLdNuzbqxrB9TVrZ6CqtcSzE1KarA1rO4mLAudV6U2p+qtMJsdQcKWTRUl6n1E6Y9JVtxskcx6YmSzYjre/AeX8DLoPMsRig9kpPmvbCclygxrhdR8JE4lCZ3tTYwbJ0Pem/2dbW5ci6UmCBF61+lxt7SeVyvkxEL0+rgJtFmlKu4BWDlCzaeKUFxxRRMYYqRPYm904L+FAsxZiTqPfYZJASiITeT5dQkRYgTLhm83wXyHDV/FZO1JKIZs9J1+L6n3wz4vjMaK+zZqvwoLeyfsijHYUqMhprsKEvrp+9x+2wrr4NtCee1BVNlvSmzui1+zVotSTvLgWdW4//tezu3eUrfVbVVTVLa1FvCMaxyURwqu1f9vvhh9QmWHWqL8MW1WT/DCvYu6zCgWaqmtZvyqiCOVUX+mi1jfc6XVacp6Ds+2GKvr9gzKM3jKgY1T/YzG1t8teSbZS/Lm3BOCN4x7Sd2u5H9NPGlL77BMHTc7EfmaeZ2ikxRONmecHR6xjjPXF+8IHQdz559wgfvv4fLE/fOz+k3R7zz7a+zPdqy2Ww4PTrWmipr+RJcTzBln2Yt2Mw5c3x8RufASaKUCecGvNd6sRIz4iF4Ic6q8Oa0KA1xnuA88xyVnbtExv0tNenfPJ6SiVGtflcUYFBM+dScY7Hx9NK1ovKc9W8fehWuzpHnyP523+b1GthU7k7l38Bqq9dty2tVOF5prFaz0P5exwbMg17ZZPXYjDUbZQH1tNqtovkwHIvnpQMF1oCyzufiHM5ZrzFRw0Jx3eo5+WQNQnOgWI+uYopbTNPWqEW2dVZM6YpORH027yEuI1LMKyxZ29OUqChBpDQkby38F+caBL5GJhYm+ZUEkKXXWcaqCaIafvOcGKeZUhLOZVxS5RRTxs8Z5YlTJvlclGMwp0wqCZ8ikiIuRr1OnCnzrDm3GK00w+RC0VyfHwa67YD0oRH5qmyrC1WNjYKQiuWCLeyZSmECwm/gEvxG23fj13zl9gu/8Av8sT/2x3jrrbcQEf7pP/2nB9//2T/7ZxtyqP77Xb/rdx3sM44jf+Ev/AUePXrE8fExf/yP/3Hee++97+tBat8tXUiFFnc/lP4HW3XRl10WG+fTjluAEC8P/PqQwyDMHWVYlZ6wKrjVF+/sfjR1Ww7ua63E2ulWP9u1iz6XE48YSawPDgmHdDQ1z9W5QHCOII7grKHk2utaCaFyd3xXd7Hcq6Li1mWp1YpWa9RAGDYeuRSmqN1WlUkjNcXVTlyrTE1oHA0Dr91/yK987Zu8/+QpftPzv/mJH2PoOj768GPeffcjwnDMo7c+zxe/9ENcPr/l4w+e8cnHz/i3//b/w3vf/hZD5zm/f87V7QUffPwejx495N69e5wcnxJCx7bfsN1sGY6OONpukSLsr0fGOFkORhkZuqHHdz2ZAKFDOk1GX++eM47XlBy5vbkgxZngHEM/0IUNTjqmaYQ0Evc3XDz9mBgnkIIPnnneKzKrRFIaSXFW5nSwEGvU8a3ku6I9vcQpq0HXD7gw4LoNR+fn3N7uuL25VQ/ImP+LoJBzT6MD485MW62KgzlePfHqTen3ru2h8xgNZZoHXs+jfOmLoVkFdQFlRc9CTBouK3VliBpUrTxIY3aUnIk5knLUsoNxR4ojKStoR8OMSmgrRvvEMNANgb53DL4QRKmLnCS8LwgR8kRJIyVNChcvkVSSRg0oZMst4r3mxmRRLNpqraL1RuI8Uop2Mu59YOh6LYT2HeI6MkI2SH5bOVWb2zN3oi1HEG1gGwncjDOXuz3XtyPPrkaeXs08uU48vc082yUu9pnbCLf7mdubkXmcSTG1rgqxWLiygM8zbh4p0460vyXv9zBNuFQI4gleC7eH4y3D6Zb+dMNwekR3NCCdvvc4JabdRB73lHFs5Q6Ss1E1allAohD/56aHurm54Xf8jt/Bn/tzf46f+ZmfeeU+f+SP/BH+3t/7e+3vvu8Pvv+Lf/Ev8s/+2T/jH//jf8zDhw/5uZ/7Of7oH/2j/OIv/mKD9v5mNpGl3HVxCkTbA5CbMC1Nq628rlL/Wg5+mRHjZQG92r1eDkEr4FeuSbOgalizhVXamQ7B/LX2Sko5uEC19PQcq1Dg+t5qZ9tqhWesKWMFWljCWdbtTcS8q+WfF7fwGZp0EhYv7m5fMWmfia20pU5FvaTFpK4hkVyLjUsm50jM0ShpooUL88pPM4FWTy8W69emRkhyvP/xMyZXePDojOlmr2wTrvBiTpw/POX0/JwkA9/8+ld58eI5Y5wIktkMA/2m5/nFBX3fcXbvmPPTUzbHx/RdRxeUADYXmMiM80wQrSeb09RQg0p4WnXHhu3JKeK0Y28Imq/Z72/UujdrW8sAtLaGkpj2O2KcyHHWd1iKLng0lFooBCcqDI1lpPaqEkEtYxcIvlejwN5jCYHO4PHjODHH2PgRDyZwFZIHsQBalKD9Lauf1SO2P+5CnZqnvCL3PSjMN+9Gnf+awarGpAdiC/kjojB1XGMmyCUvaOKijBGSE2WMzHnf5n+MszLEl0xwjtIFivfgOlUQ80yJO9J4owXEKYHvKJJIziNpwnWd5hWDKEGyIfJCV2yeg1X5Q2V8saaOMc7KkhJnck5aaxc6+n6gC72iQJ00ctzqTzgbr0xSmi7rV9d5lW/OQXZoYTOwT5EpZlPWDleEnJVz00thyuBS1saRnVOaLdcpe4gIYnMvl0yeJ8oYyM4deMVFLE/edcjQ4zeD/p4iUjLTflyAVzEqHZpz+KI9ytRw87hs6+v73H7Lyuunf/qn+emf/unvus8wDLzxxhuv/O7i4oK/+3f/Lv/wH/5D/uAf/IMA/KN/9I/4whe+wL/6V/+KP/yH//Bv7YaqUlprMHNf1/mlutBe5YQt3gosQftatCltgtZGka9ik38VJJ56TD33St8chDJffej6LMudmgUrHCrLdhoRxJWFj9Bp/Vb1WhZ0oYYBvBiR6VqpNeW1WNgt0CqrMSzlQLHVIVl7aYWCK3JQy1VbnSQjNq2EoAttzxJCWft0GuqB0PWUDJcvrri42XPv0Rmfe/Mx035mP42M80T0gWF7hO96nj1/wdMnH3F5dUHBcXxyrOwCWduoD9tjTk9OOTk5tmaSHd4V8hyVyDVG5jQZJZYWs1a2fTC0JA4XPL7Tmq+YIiF4ZdiYZyh9e455nrV+LWshbYwTKWrIx4uFC00gFKPzEe9puaWC5mRYpmtDemULjxlSTqyx5jhO1ujR5k2pK+IwxFsjGHVqK8r1DrVYW2vVNLQ18lJ/JrHxsePbulnNreWX5XomxBu9mJinJspAofOsNAWq672YcaREysWeZU4zEmdFG4og2416Xz4pQCTO5GlPHM3wycrUkQ26nsVBDIqMC44pJooL6gYZOnFZJM5Igis7TDSYvJIsY0jIYEhDXympGhmxei8OaWwbgnIFKiWG0DnXCIedF3zWFEC2tVa9vlQ0zFlKIYoQUU7jmDIuCL4A1i6mlEwSp95fzmQSTDPZu/bONHyrnlMx5n+8p3hpJNQxaReDUtS48GjuywHBObrgGDqHz57e6su+n+1/kpzXv/7X/5rHjx9z7949fs/v+T38zb/5N3n8+DEAv/iLv8g8z/yhP/SH2v5vvfUWP/ETP8G/+Tf/5pXKaxy1JqVul5eXQBWqhnRzNMTSgacghwjEtmAO/J6V/rjj9bQFUi/4inBhPV0713dJRIoVVqqF3G5j5b3oVtqzLQla+fRLr5SO/augDK9N8kzyWmF35S1cmKwFazpnJ62PUD22JvTubK3z83psxK2exeiEDcJbi2yVZVsh3zFqO5C8KkheAo7Vo1tqd+7fe8CTj57yy//+vzADX/nyl/i9/4ef4urJcz768BmX15fcf/2LHJ3c5/Z6z3/4N/+a6GZCr2wJx8fHXFxdcfnJJT/223879+7d5+joSFkr0Lqp/XjbemF1zjEDY5zZzxPHR0cMQ0/otFWJIto93XDEOCW8E/rQKxnspH2NRDLisuWvZiBDFbQ5IlbV40S0aHRKlqaoRkFaxtQJwfU6v0vCSQBE92FD6JX+qvQDydCN+/2IM2OjKhyFievLrpRj1cKmCsyV8XdgKb6ikeDd5XEYWX55vdXcVv1ksRudFSM7RAJONOQURIuKM8rjqCwWqhx7lMlhkkKaowF/MlOaSLtbyjwhkuk2x1rr5jp715EUR/LtlTW4zIBXpWV329taKs4zFwdhwA+ZrdsS+kDwqrhEDPKeEzGrwTOOe+I0UuKsAB50/fng6YeO0HU47+m9bx0TwB24u5UWzINerxSt7w0dSYSA4zZE5i4iKRMpi3wBIkJ0nuQCuSofr/+Shzk7ZoScBJdEOyOMEyVoiI/iiVmJjXMRppwJOTOUYryNkXmcbZ6v3rADcQUXCoP3SFDPjk4RrOP435jy+umf/mn+1J/6U7z99tt885vf5K/9tb/G7//9v59f/MVfZBgGPvzwQ/q+5/79+wfHvf7663z44YevPOff+lt/i7/xN/7GS58vwAua9ZVFY+kZlC8sZ8sBVP938SAOfOK7a3Ex+1fXWyvGO5RTa6/P/qhKUxelrUwTEo2XsSm+ZbK1dhSmsdZez/peqzhYFI0uslCJdc07qMALXKV/qcAMbwp+SSmtQ4V6uXKos0yjLwoTcwhrkA/LB5rKFSuiLdWzSkZHo8W8mudK2u6kVC/GDBGzytUWVWGSJRPHyOXlDe8+ecLZvWMenB9zvtny8fwhezwpnHDSn/Lk/fe5uLwkMjG4jiLClAsffvIxZ+f3+NG3fpzPfe5tVURBhVCMxqodPLc3l5av2AOFzupy8F4tzVgobkFYOR8MB6H5wnFMpLlACWyOj0kxMU97bWXh1Tqd9jsKSd9B64lWDazK61fngobPOhdwFnrVnldOa5Z8h+824IK2xvCBeVIvz2GdmEuFWJRmcEm19EuBdPiq2xw7WBrWuoNPs+UWAJG2BbH1WG2bOn9dnd/LfQQvVl8kdH0wYj4gF+aSkTxD0pBXMaUtWGjcQwmZURxzjqSYFISRhVy0oNnNajS44IigyM9YyCXYDM6NE7DYek1ZCYOlOPZJwBfC5AlHESHgeysnEA3HYcbZHCPTNFmuUgEbRWj8hSrgFanYezHxYCi9rHRNSTydsyaXTpBNhzeSZem9dilOQufVGWwgGb8Yoh1KPFAclt/UPF1yGqVJzjFJYJe9MbcUNq6o5+8gFmEyVv5IZpgzfkqE25m5ZOZxZtqPWsDvtI40pkTvEgRHDiBDIADHnfYGlFLo3H9jyutP/+k/3X7/iZ/4CX7n7/ydvP322/zzf/7P+ZN/8k9+6nEVnfSq7a/8lb/CX/pLf6n9fXl5yRe+8AXgZZ2jU9ms1ZXXcDcmX/d+6SSvWo13dvnUM5V6B/bZSglJzUvdOUtTWqbADnSm1POtsnGyHFtWZ5HKhiEGCXaHnlX1yg5YMu6EB3GLEKnCSVZeWL3+q0douaNSYzktlFg7IuemuGqLk1xqDqyGl4TKptE040p5+9BxfXnDxdU1t/PEl17/ImdnJ5QCzy9eIGFge3bC7nbk4uIFNzfXdF2vQhzUUu16Tk7PePjoNbbbI7q+11COqyTAaXX/el/eFmUNoGWNjSgrho010EKvuQjznMlF2TScc6QyK4lu1k7QOUXj1VsNbzUI7PmdLKiqZV5Vj9xZaEkLkp3vEfHGreeXMS1aB6V+zto1gjYB7kQn1sN/d0nIap9y54sDY9IMuVeBm5A6t5a8c0W2ajhMkYJLQxbrKFCMIDZFa0PirQuC+pK+0/YgriiE3Ns4iFelggvqsZnXpiS9nuw6+0x9/trBW8SYLgq4rI0lpVI35Rq+tH+iRovSO2U1zuaZFLMVUKskysUaVKaCL0vRvzMkSk7ZFJwBVZzTshAHLnsFPpBBFDUYBDondLZoa8mPsuOAR/NNwSvtm6udvJ1yQGo+0hOLp+SMT4mQNHCL5RO1aFoNnDlnppgIU2RKiWmcmfbWzcDpe4yltBrFOGsRfUGBO72BxXL+b7xI+c033+Ttt9/m13/91wF44403mKaJ58+fH3hfH3/8Mb/7d//uV55jGAaGYXjld3UhVR3uWRbEIWPFKxbniv6iLqAmd+9c4/CD5Tj9KW0/ASU5rYqhneCu9bpWZlWBVWFPs0jX51/7N6UtepUCDZhRJ6y15KiFyIgahqrMXIO/Y0LZr+61TanVDYssfy/hoGKKpnqYVdKldreNZNdQhSlpB9icsqKemkLLi3VeSWAbcZyGUbw4tv2Wb/76e3z08ROSE37ix3+E1157xPU08c673+YrP/aT3H/0Bv/pF/8Dt/sLKIXTk3PeefcdYooMRxu+8pUf57XHr3H//n2cd/Tdhr4fEJnZ729JadQi3pwNlNHhBOZZezL5LmjUTwQXtorcdFpQOvQBCsQY2U8J7z3d0Kl8S1EbKuasoI/W8j2AGEenxngpFLxoDOEg/F009OpDABxJlJLH+YDvBuakDSrFe2KM6s2YYo2pNGVQgRqHBRsrg8teeqmxvE+3WJa5UqXmiiu6eXbr+W5KzosiW+t1lVPTE5worVmE4ir/kVCiKPw6zUxlpAQHXhAJZNFi7955tseneN8Tg7L09zlRciTOk/VOE4p3SreVHFrM7Kj0U8Vloz3S/mB1+RXDp4sLZqystHupT6fvQ2sBI+NuzxyXkHgSTyzCHGGcMhIK4gohWPdx0LBeUmWR8ApWEc35dRKQXJdZIpIIktk6IXqHk0xCQ6zewpPBPHbnHF3nV/+0TU8Rh5NAkqCtTxJ0EYI5k9o+RcOyFBhTRuaICOynxH6aGaeoT+9QsBBaPC+lMLqkjB+iCGgfHD4Ime43mFTfffufXHk9ffqUb3/727z55psA/NRP/RRd1/Ev/+W/5Gd/9mcB+OCDD/jlX/5l/of/4X/4LZ275WOKCmWKCRRTAl5EWZEt1pzLondsSb5iTa4/ebUneHADn+KpfdqmFoha7isniop0WCP8l1tqrtnqnqXlgZQazTcSV+cXlgzXZhRKUNpyXGp5lYomqeNSXa0CQq2PkZVQK+2HrlcTBvUZqoJtnlTRfJYBNOIUD9qcVACHHi7atLKYwGqiD0TUixyK8I1vvcuTyyuOtlt++4/9JJfPn/NffuXXePyFH+bpi2ve//hXuB5fKA3UOPPr3/g6/WbDg3v3efTwNR7cP+fs9JTj4xM2G811eQfznHBeCN5D12v7eWMpv7m91Hi9gzTtGYZjun4AHCFstK9WgXGcW0nAcHwCRlT64unHzPsdJSXOT8+YJ+V60x5WltPzCpNWw8KrtV0LgKSOvWbH41QAZVj3/QbxniwZ123I4kilsp7oPy18fnmu1jhF9W9oI27HVuFdmapMWNdC4lduq7B5qt5ezd2xTLcarq5cmepD6pdF9RIxpZbwp++QKMrmgNZzleyMRktPWBA2/YbgeuYhk5yxZpSsIWBj2RdXCCWTYyZPSUN1YkQgOZPGSdvXV8GP6dCU8C7Qdxv8dlByXScqnO0ZSoY0z8zTxDSOzFFRjKVkYhHGBLspc7uPFDdr/VjvkK4QvCMEpx6l1dEYHzFeoO+FHAs5ZiRnfEyEmNgAqQ9E9Dm8QBeCdijwQaNQTqMOQ+cJwdN36nVOpSBdInQBmSMpOsaUmXMCScQwUHkKxeb5nDLMkTFq4fGE5t4Qp6Rd4tmlxJQy0cZWrCzHS6FHQ+/fz/ZbVl7X19d87Wtfa39/85vf5D/+x//IgwcPePDgAX/9r/91fuZnfoY333yTd955h7/6V/8qjx494k/8iT8BwPn5OX/+z/95fu7nfo6HDx/y4MED/vJf/sv85E/+ZEMf/qa3u/ELMVeXqggMqVSMQX6xL1cnuTuC0k62btZbZUcVzMvu8tIpKgdiqYvx4Nz1sFV40wRCXdDSlnjdeYEviIUYWNdgWc6lhV1WSazGlmEnrwqvXqtawRX1deAJHgyPrO6rrL5cCTFZlFhVXLTcTFIYckUVmie2AAWc6egVPN+EmgB9FyBnnn3yCbv9niLC0dER037k8vKKZy8uOLr3kGkc2e1uSKkwjrfMc8R3gXvn55yd3ePegwecnZ9zcnrGyckp3ge8V+mcx4WgNudYEflKWeQMNWkoMC2uTkzTRAjBFPyCkKMUvHfM+5H97Q2311cIGn6cswoyHT6vgspCvmGF8GpF8EUsjKWlBFp8S3uvubGihOZJNyNCbIxXnlZriWNbXk01J1U50XKvCz7DoE7SVpmes6IR70zbOrdLI+rUuXOI9ZCmdOpJXFEgWjNsbE73QffTVvfWmLKoEK8TupYjuOAIvhAq51/twB1nK3BWpnklvlVIeHFO/+VMcp0W6EILsymVVNb2MtZF2dUC3Wr0CWiIM5FjJM1zQ9WJQEKbS04FplRwc9YWJZRGkh1Qg9QXzdVlUBSfMSXXXl95ninThMyRkLKyVli+23voO8/Qe4beyoFF8F2g8866Q5T2zp0TvHeUYB4uik4sxRFxVtNmYcdaCyqOWnTtg1+o5bzec56VGPt2TCQBJ5nOFbqgXuXCf/S9bb9l5fXv/t2/4/f9vt/X/q65qD/zZ/4Mf+fv/B1+6Zd+iX/wD/4BL1684M033+T3/b7fxz/5J/+E09PTdszf/tt/mxACP/uzP8tut+MP/IE/wN//+3//t1TjtWzrkIQqgAakEOV6y1IT3yvgBqzc/Xqm5n6wBPb0m9VBr7qF9qPqssO9lqR0RUPW0uhXWa+rJicmx1aKS10oXaDQ2DJKbWVSlZc4qG1NTPBoDtCUnN1lVYbLdcqBEDpkG6EpWR0+LdQ1idaOKW1X8xpKVuReY0WnIc2qKiwG93VQw/AKEUZzaEPfM97u+Na77zDFRDf0nJwe8+Sjj3n65CkX19eE7Slx2pOmiRSFF8+fk0vh/v17PH78mLOzc87u3ef8/kPOz+9xdn5OjAlKIqVZLWR7ghQn48orJCI+2PsST0S7HzNFpjip1wWEblCIOiA5IUkBGZfPn7K7uWRzfEQ39Exx0hCVANI1AtPggzYRpDQqJFMfFlpV9KMztgycFrVmh1FC9aoIZFkV2V4Nzhk9kRpz1c/KsuxHWaFb70QAFltGLIS5kPm65qPXqbqUcpSChbjLMp/v2nsWHi72nCF7JFnuJ5XWa84HLc7NEXwuzMbeK1nzeThvcHpBghkDYNEXvd+SstFKaTv6LFic3GmBeVAG9ByiKp+U8V3tIq491QStd/JWI1UolrfMCjbB+njFmTzP+s6wrt/OE3FMCGMulFnJd7MZO945+qKvV8fNqcFXDUJJlKiKMU578jjBnPGp0IlX78cHfBCGzrPpPZuNBwNASX0WQcfOIjHeiRKPBKEEyNmRsihaVRzZ0J++8+odek8XPFPGmkqqoR6MAEHBMMqEvxsjyYBnnSScV57Dpdr8e9t+y8rr9/7e3/vp4QLgX/yLf/EbnmOz2fDzP//z/PzP//xv9fIHW8DjcURqYau2hVgvODWGTC2V0hanfn+opO6CRpZvD03KV6my9Wdrh1CN9+VMtR5FZIHDVlReKXdOVM8hS9LbDCsVrLXXk1OPoLLCV6h8VWzB1YLH0gAAYhZxNRjvPtt3m1aNn/DO2DQOtOptWT4rxdhId5UCimZE5LLQA3m7cPEWWs3mNXrH1gWmOfL19z/iZsocDR3bIPz7f//vOTo+5uHDB1zf3DBPM/tx4utf/zZvvvWY1157yNtvf4khDBwfHXHv3jkn9+5rPVcXSCUyjnumcccctcjSuUA/HEOJJBmJCbrQU3kaPJ5pmthNe2ROPL+9ot8e8dpbX+TRo9dx3jFNe77+S/+B2+sLxv0NXd8po8QccV1HCQp+rvDnynVXyOZ2REqadK1VoIKY4eOU8LSUgHQbpN+AhGqKNG8/saY1K81pVltGQEqLSC4NGDTEJ67WJxrQo6hCqimtShBQo87erpOoa0gVXN1nrbAaCN08yDZnjN+yJM+cCik7ckpW0wSdeKUqBPyobT1SyZQ8U1JW2ifpleGkGmSlepCuMfeXoh6BVHoyyQQxhvfQKYjIOUoIuKT9wGp0o8vmZdei5GVR2DNlUp6IcWKeRsZpR06JgBBCrwz1wVv3ZCGLoijzrAhGXCH0QoczHkPRe4mZNCsHIRbBmPczZT82st9eXAOXdE7ovWPwnk0wlvoiSs2UonU/0EhDsjZE5uRr2FkyUdTLlb5XY8AU02YI9J2GHYs4pgxTVsO4C4Zmjpno1GPbp8IYlRkmOI0i7L5PmDx8xrkNq3JyNTRR0ESvLEJ0FTFjIe9cBezk03GIL8HE9ZDD79f3Yp8e/n9VBtrOJQd/tj1ekRdw9VyiQkmtvcr6Xq0osWLkSutkyEPnrD2EnruGGFmdf+0n1mddhMsicGoYsHlT1XZf926xX2oRsra3r/25SmtlXi30UpYatub3igqbYgWWwQcev/6YZ+++xwfvf8huVot+6AL3jo+I80xMiS5l5nnm+XNFGD58pN7WwwcP2G6OONlu6LtAyZHtZsAFx5wiMSXmeWKaRx2bloMT7eCclKul25wAjpKFOCd8sHze7pau6xmcY7Pd0ofOwraeaRqZo3LEhRrSdSoAmqcilShZqmNrbCRa/FlzXpojdQubeNggXY/0W7K102i4RDvekcnrYqtSx3lp7onIqhfeEpqqczebRbWw2ZQ2bxqPaFlMvIUgsb5Vm+GWOFvBHFbWpTSC3mS6O5qwbSm/gnqiRVF7LmtOMFs4Q43CbLyYpbl4y4qi5W4XujMtoq1kxkvovTJ+WJ8vZ+Ukzr+0cswqbojanBMx1kaqkZwitVC35qE1+mFNXq0hJQgJo8RKSixQ6ju1hl3FOeZYKNG8r1Q0v1+K1sKZEbt+H7kUYjK0YM5MFK3ZyoVpr2z9NdqSSmk0VZFMElWczjlctxhY2lldI1nBKbLQicmHrOdY/Bt9joRQS2H2SXuE5bKqy/gets+28jLF5GThEWxTa624oIU2VqqiWaHrFMDap1xbrXc9LLWJXva7WnilulFy97xy5whdVMVCOkvPsGW/Ki4qC8YhE0b9jAPqpxYeXCsrWd9SZRC588RyOIawRKYPFFcbsLV4KKaUyoIgTMlqu3TiNuesTub6zCvS05Y7yMoicP/8nK8++89858OPmCKEbWAz9BwNAzcmrKZpYr/fc3l1xX6344d++Cs8eviQs9Nz+q7jaDvgvZoC/dBpQauxH0zzxGzhndpHrKDWuRYgd/SbIyiOnKCkkWDxzZQTvXNsgnZB7oy81xWI06jNLbEGiTWMa9a/vvvajFCa0imGkigqfVShUkx4KLuB6ze4boBuq7PzjnfT5tqd/KxQkFwVzOHPplCqrVVfzhqpu/p+fYnmtdmUWCuopnxE101tO1JXUf3ZCN2KCvIkNZ9k17JeVNVIVYVajCTbdjEjq833OjnNMjhUYKbCpBhNk7ItZgulZWfsFmI5tUp/xt1x1lVSipLdRst1KaltUvnjarNRK2sRZZSvPb0oqtpjQeu2Qg256rov3lGSfp9SoSQFojSa0dLsAGXZx1OZ+WPM1hOtMKbElAtzzNawNaoS8o6ci8LhRfNy2fqLOedwtfeYebSWAlOZk9VwKDhrHJvvMNGroV0opAJTEVWo5o19r9tnWnnZLEbQkFOVxYsvVUy5GeTXgALNQ2ohi7sr8vASmHV3V/GswRiaO1r/vRyuP/W/GrJbn6suh3xw5HKiqoycA9+FtgCxZxO3JF2ds7PX0CKHSusly3e11cLpV95/U1x3hipXhVPafs3rMkRhSmnVGRnWp6oCa5UVASwn57WI9IOvf4N33v+A7zx/zojwI2++wfHRwNXtNY9ee53r6xuePHnKt97/iPv37/PWW2/xlS9/xZ430vtM8Inj4y1n9+/jAsxTJM4z+9tb5nEiTjPTuCdZK4hCNGjxQN8fqeVbMs5ZJ9wkzC7DdsPJ2QnHp8cMwNAJc4zs99eMz54iJbE93rDptYOxC1pPFCzc63xNplueo0QFWWjKm0RWLQ546QlhwG+P8NsTigTrd1aVz8oD+g22ZcRLe/eFRWDWk9R86creO1gFuZTD5SO2Lqo+tTmlhA/lcMrZRMxmnRcJFNdpnZrLiFd4uPOCeBStmiIxAVkoJWIJG7tFZSzHrnsIlqrPIxTnCXlRUDhR5n0jK3ZFUaaCghEaXOpljbUyTkvrRTfvZ8b9jmm/Q1JW8usQ6IeeIfQaosQTvKfvOkKn/dS8aDHyDNrMsWhYtYiG1zWX1xGNod25QOd6fFH4Q4yJ5FBQUSrMcwYiUxJiFuZc2M+RcU5MU+TyelS+Re8Y+qBh7eIoPpCzkIq+l+A969EsZtSIeIqDVJLWNFJIKDq05rWdc/R9r6HIbKkGb/Viv5mJ+l22z7bykmVxVGuvNEWzoAxXDpBKdSkWDlkOPFhUr5qkNPl88LlaeHe0wKsPP5jsd89VRXgT4OYBOUHbOpjyapajxdvXKMNqFbVW5tS2Jqvr1v/VZ7KHlzYGi0fW6rRWD9HamJhnuyYO0bBBPujNlS2eXhWWnsuuVJoZYclpWe6pJLb9lpIzv/KrX+Xp5TX7rMivs+Mjzu+d0W83vHhxwYvnL3jx4oIHjx7yuTff4uHDh1rfliY6J9w76ZW38PiYfnPEFE2hxsju9poYR2Kc2N1eG+zaQALiSSky7ne4LlBzeXGetC1JAel7pnGm8zscmfn2mjhPcHvNGYnZJ0pQRg0fNPziauGpwRkLyfKDiRyVLoqUSGlSvSWOfjii35zSdRvoe4rTHFdxWA0Ni6tRXp5gNbwnyGph3IkAoEXclfC2hp/gcI5rDsnQktjvL82fej/GF2/zpE5Fc4RsTliYGCsIbrGObDkcy5OWzG6cGOfCPEMKQY0f521prebQaq2pjWuCs64TQxVW187JwgWq7WbEvN/S5nxbA2asNXFu46T53aQe9zRajzatuQxe4ekhBAv966HeKbkzeK3NFGO/WBmDJbPqxKC1fckHovRaVULElaQM/PZaolHFq+NcmFMmpsKUssLbY2IfEzklfCqkrKH4Ks+i0zxZwbHx2vXYW91WsbeaRbQ7c4GclCXeidWMOm0npPJGwWMxZS1VcCgH5/cE0Fu2z7byajGnuiDKKgSo4ITFWajIuoqoMm9CdGKLLEHAYotSVrN1bXHW5Hfj5GwBk5WWOLxJ6hnrAl7DuQpL/VdFc7XwXQsxrLyq6nmt/q0/cyyKbFFKyybtJqCRDa8X/cqCX2Mu6/qteatqYdW9aoPJtbelP0sbiUbpuMiC5a5ayNJCoTkz7ia+/cHHXO8nsji2m56T4622KBk2XF6+x9XVNeM088b9+zx8+JD79+6zu7khOOiD52jTs90OdL2yeMdxIs6JNM9M445kLdrned9AOwHf8j8JbZdRDK6eZoUIlZJxXUcaI/M0Me93yDiR44Qf99wLnr1kZsmHYd0iC3t/qSEnVVwx7jU8llSwONfhfE+3OSYYL19xBksq9V0uM3NZCXfn4/Jd+8red1legc36l/zu1RyS9u9AsbV5VI1Ie7ZS2vHtbuTleZWLlSRkzYVkQ6ZqgbtAUeTaOM6Mc2ZOjiSRIr6t33aHzdCyNW+5FqqCd9Lymm1lWuRCnHk4As6c3tTOuVpJJjeqh6nPoIjQPM+keaasatS8VwXmDXZvQkC9OxFcCNp92yuKkQaKyi3snpPmlGJRyqZs9VTewqe1eSUFpVsz5VJRfzErO0ZMhTkXYrF+X6YUKwuVIGSn+Uas/EOsU0EFueSiObJYbL1n9fy8Wdy+MnlIBaYIkjLERAV7VVaa73X7bCsv2w4X6lITBYJRpekEFHAGbyo1wWizb310re5sFmi7ihz8uc6pvewBLwhHXbB6rGM16esiLqkdobUTNMi7wt+91WEswl2qu+SWOPxSw2WTrd1FtUtf7adXBfSymnv1w1XFZlk6SrGFU0zYxJX3VeHndlABSHYdV+/Pwrqiz+CdY7sdePbuh3z0nY95/8UVc4HtduBzb7zGg0f3KTiePHnBdz74kOAcZ6envPWWel2bvuOj99/h82+8wf2zc042p2w2R5SuZ4yFeU7sbnfcXF0xjzt2u2vlHDToXCnCfh9BNM/U+Z5pN1on5BmsqNg5R9gMkIVxnvnw3Xc5lcCRd5wExw/fO2Ocb7nNkfeDzbmcEBwxJxDoJBDjSIoj87QjTXskF1wWxHcMm2M223P6k4dk77SIOVk4HPOK3aKBqidNDectVoq9rWp8SVNcCn3XF63gJ08xOLm8qhZHQF0+QyCaUXjX27HZsRwkyz2s90tRm3viZkoZKcmrAphGYpq0rgthjJHb3Z5xLkQKSTpEPKHD3onH4evCWimIquQdxdd7SAcxdCmKJUVEPSNjI0mO1mJpae+i51NB7sjFmOSzUlfleSTNWuQcXKALymYROi2HQNSnnFNkjjM+ebrNEZvNEX3f03W9enBpZop74i6SJjW24m7HNM9EdYtwWfDF4RFikycRn5zRYTkgNbLrDMwUkmh9YCpFw3klEgyliXlOxRbpGCM+C5IdQbRNUMyZEmd208h+KoyTco6GIJROSxZ8p6z5HmXx9zHBqHV2WTLpB1p5rRenLbIENXnQXNwa/1YQkoawHCpDq0FVy1bWxuoBUSya2JWmjIyRG5v45uUt+m6R+FXI63JfgRSQZtHVEF8ptHotDQcGQwzaEqxxQOfwK+WpKEO7dbfUj2WpAsbW88rjW99llqps1ypMRSRW6HpgaaPs3sWomxqxbrZ9raW6DVTzKgHwNUykwhJRy8yLPYd3SPK89+wF3/zoI2Z7h5tNzxufe42Y4cXFC77+7rs4L5zfP+XN1x/zIz/829hdPefm4gmnfWC72dAPPRKicsWlzD7tubx4xn63Yz/tmKZ9q5+CyDzNlFLoukGT7ymRvXIwVrAKlGZMeO9h6AgpM6TIUYDjvuf8aIsf3sbPM0wzZ/GCyxi5oXDlOx2zkhnjjTJt5JmURsge3w0Mw5ZhOMdvjnHDllx7dZVMcVr/1PpYYd52VkvbXlJ7u2qUFfPuV/D41sZemSQWqq+MkM3Ak7YWqnOj+bFsHlYlAGgXwlfvGkA8LaZgtUa1bUgsiSlFpjnjpglvDPcpCimq8lIgjZ5sijP7aWZOiooM/SlSPIVu0UOSaQiGgs5fD4IKeMleyWglIFYw31alhePFe5yDHACXKFEfLOVsjDTqqbQ5LQVJOhDJQnNKrJuR4AldoOsHwnBEsbChFyFNkclNOBybEPGD1Y8NPb4HmR3xdtI82jQy7Xbsbm7BQEo5F2PhsLdUCYWl4JWGBERzg8VqQcUJOWZKtJ29rt8MTDlprVhQlKMoNQklJebRipedIjzd7HDRMUa0kWyOzGRK8S3fJXOmyyqrNDCqCqeAMcnwfW2fbeV118xbuwg1fih3PzLYK+r6LmGTpSZGF14VuOsQ2it9k5VqaAbu4X6yHL8csw7JmNdBDfdVxOCdf81MrJ7W8u/wfmjPLnfGpVDrd2RRtC89lJgHYs/eFHhZDfUyKq2AMq9+X4cwWGx0KK1+pVkWKwCK9xr7f/7sBS+urrm0Bnf9ENhsB462W66vb7i4uGS/u+X07IR75+ec3zvn9PiY+eYCcuHk+IRhsyH0XUM7pZzZ72fmaWSeR+Z5T4xjI8jV4tK0PJ89fFaG0YN3sA7jKhpNm0f2XcfQCb2P9P2GPm0IsZBuE0PMDElzOPukCz7OWqfjgRA2lG4g9FuGzRH9cIp0g7auv+sC11dh1nYpi5e9nmftqIMPV96YzTvN29S9q3Y7vKq+rkUhVu11F3Vb11OpHRGKeT517trBpWDs6wk3R1zWfVOCnJSLcJqj1bkV5nlmjLMWuIrDZWU9QRyHUQVT8i08i0UinObHcsFT9L1ZA9S6NgDNfdkweF/I2Rg4WhlBfc7VVQvGCGJ92ihacxmUTzBY/y5Zwc0ree80zcz7iXRi1Y/GcUhUkt/Z9hnHif1+wpmWUiVR7yUv4VJBi+SpIDJHrqh7LySDzh8ioqWx3uSyhA/rw5WclduwFrq7jBSlIcu5Kv9FXqRcSClbiLy0lILY/q0J2PexfcaV1+GaBFVK7TNbpGUV2HcCFPOC1rC36i3bSdXFrstwEQBtYRbNnVXIbjFlUFY3dNcJW6uQ9lOWBe+8GGxa/4lo7LgqNVfzJK8ah7LUrNXi0N9oqzyq63OW9f+rN2nnr/mLqqR0uScqJFctQp2kReVHyx+WYqFBln8iQmX+V/u8aIt0H/h3//XXePLsOdEG9PjsmNOzUzoZ+OZ33uXm+preOb74uc9z7/yce2fndCURnKcbjrl/cszJ2RnDdqA4FRpxnrm6vCClaEWkOw3VxahtW3JUoWghV7dSrKHrmiAuBUUKikNytFbxkUxi2J4zBEHyC/rtKYPf0suWN062nI3X3Iy39Fc7PtrtuZonpjLTiTD0R5ycPGLuVWGFrkcbepUGdHEcWFvmrQulqJVdHEgRDXGJ0gvdNVwO5uXq26UDQo0PZBOw9WLrmrxiNFLrDNna21/m1+GEM6FcPQXzHvb7meIdIok0T6rAjN1/TlkLkE3JjXEmZcA5Qkx4I08+6CtnqLb6+E48GIWWs15znQsQtaVMztHmrCozR0cxOqjSeVwrmZBmszVMi62NjNZMlZwgRo2KBIeXjqHv6PtAsHnoLfeZ0I7DsRRCuaE7O8FtByXSLVajlQv7cWK/2zPe7Njf7llCuYLWoqFzwLv2ciVGrYcTwUugeChKR9+UywH204xVRQrr/NZ+akaxUDTnlouyVYqgytU5i2hVai4blaTIwyK5sZtkU2jzqNEN5h/gOi94yR596bMqKossAAvs+2ChqpgPIvNAaQShWjGxKKW7iWbgJQaBT7vP9V9VyYoJ78prJiLKpL3iINSH0Is4OUxO630UtYjNej783h1ce60+vttWQ6bAkkZsea1q6ZcWwig26WshssbR18q8vYnDCzkdBIfQdT23Vzd8/OKCTz55xm6eKQKdg9cfPeL4aMvXvvY19lfX9N5x78E93nz9Ea8/foOH9x9y9eF3mFLG9wP3Hr3G9mxLCEoA+vxqx35UL+v29opxf8O0vyZOe8hRNW2GftjgQ6+tNdxS49L3A9XW8T4szyEOjs6QkiFFnuSOq/2Ok3iFuBHZHBP6E/xwzEk54ki23D8eeOviIy6vn/Hh0484TRds/H2Ozr/AJ8eP2KXMfppMaHgDuYzWcFVvtYYAc874pHeTKIgLeImYu9JQtZ33ivpciaxiAievOtq2OYkqJ4cZMfUos+IW/+PVynD5pqz2X5Sd2M4N0m05xpxEX0VR5oeYtdeb1k/NatEXIceZkvS9OZRzrwJH4jxRvbwiiUxA8C3/40SUhks09C9ZyDFatwVvXI4GU08FXFCG9DizEBzUyauLI+ZISkYHNc90AnSefnAM/UDwvXYnWLckMaVbcuJmvMVfXFBKYbvZkkTJlLEmrXFWHs1pmlRRlUxxnpQNfJILoQu0cHGptWlC59B0gYG/MKNRIweGWhYjOahREEu7VOMpWmPWXFJ7qxkxdGYxAEwNGxbcVCjON8VXUw8pKrCDnMnTD3Sd1ypktjhQK8/LhGyh5YfWO1WEn5iVJqwPVk2nHtoCeyiHq1MF+GIMHyjBpujWx7hFF7VCY7dGD2LW/upENYSzduVa6CWvgz2AxcHNy5FSNHRyd7xWT3Agidr9Lg+1wIWtfYkpNte8LPveQhHZJrP+vb75JURRn7XeT5bCEAIvbm75zvsfMcWIlEIvwmbT40shTRM3t1eUNHN6cs5bb73FptfW4kMXeLG7QvotXdcRc8IHj+865jGy30+M1hgwzSNp3msH3TQpawMF53qGzZbQb0B868Ml3hHCUN+2ocHq8FXYc0Gy0hWNOVC80JWJFKG4ic2wxXeOIEKQnodnDzkeNgzOwaXm/sLFx9zvtgxu4LYL3CbNoay9m/bOLP+qfHIKOfaSmS0PIsiS5yprj0vMys6sfSmdG5XzsprvshQAr+ZeKTV8SovOr2/PtT8WlF+p87JOZ6ntRfX+UgZEjZ5SmVjKYgAVU7s6Nc0DSEb2XLIaEXUu1dCvoUicZApBn8U5yqrYuNZy1cVfrN9cAePsW1g4xFWdYCZAndMWKlQKtAlSbAjCPqBdC7wxVHixMoxFeZELOUf2NzdI0aanBE+ME+PtrdaMTXvmeSbXQm0M3VjHJxckZTOAF/mAmFFTMq7Ys9hAOTOU9X24xiFZO61XnhEbckMYLjIhFS14X3vTxYwOu0EzVwpr8mYR9QC/T7zGZ1t5vSyH66KqJr/+LqDE7Cz5q6on1Howx6YcrFKEJcxYw2eHWMb6YpvaOLwrWT69a5GKOFVkbuV1UQW6VB1W8+htIeoEsDxDy7xDe7pSla1b3dtSGV+35UmXSXUwpmUBvGD5rMo+UZVXje/nGm4oFspONXYuNlmleZLra9euq1WE9l64urriW++9TyITRFnYz463pP1I3O+Zxz2dwNn5CV/+8peI8xWOhOTION7QH23pNh27ccc9dw8XAvE2Mo4T0zQSTXHFeU+cd5Q0AeDE0fU926MTumFLzK6xxSPgpGOh5cLCKYKywrsGfohE5lIY8yPS+AlT2pOmmft9pPfaFddFOLv3mHuu8PDeOU++4xmvnpE+fpf7IXB08ho3Rw+ZclK8ttXbULEI4sjK+Y0rhc4HnNOi+8t9sWaFdVLnFtrS9XH4vpcXLwvkvHoXzdaQGo1rn63tKGTljYjloUzhUJVbtV3yyh6z81rEmWLKK1UX92DRqOIRYz8pJashkmZKjgiaFyzFlE9dGyj0XvKML+oNIBru0nVIM6JKUtRdA6DgKcYxiFhR89rrrCHKosonxclQhrOi7VxgGDyd79BO10G7CMuq7CULxWVFwN5cM+33TONI6DtyTkzTjtvrK8bdjmkaKWgtVTVIK8F1QWvBKmmBdYVU8Ja9tIzm4dz6nVo4VGvRFoq5BYhpHr4Jzkor1kobUCBbPWfOBTEidCVW0/B6qVyYIrigIdmDdgbfw/aZVl6HnWEXFeGLqoxKEloRgs3LYkEA6uDbJLZCyHo6gYMkaFqt+pYlqkHwtpxL+/9LSWTqotX/yYoF3hldTp1Umt7SsxzwL5pQqIi/6iCJcOdqdQm6xfJeKY9PGdC2te7GsCitrCg3rXGyvJflXROaaF+s5XpDcnjuZtTX/BzKDTgM/Oov/Qc++vApriS2HjZ9T9d1uCGwu72i5MzgPF/6ytu89tpjeu95fPYQcY6bq+dkcZzeO+f04UOurq6IMZPyyNXNFXEemcYdtzcXxHFHnvequErCuZ7QbTi995huOMZ3A15CM4Q0rBbuvNFaHF5xZ6oktMu6Uv8k/4gLMjdkLnZXnPo9Jy5xnyvmuCN0W47v36d/+FPMux23T5/x9J1fghcfc7TZ8oVHX+FqOOPWbbjaj0TREl5fMluXcTHhxolNfsJm8AybHu9PiaEnO8d+vz/IiTsLF5Vc2nwRxHrhLfms3Iw/zFPRv9fMGW1by5+777iAlSDbbFx7BIvBVnOmat3ng1OvDb9qeipPIJBnJarNkYxHrIKzGZW5kHKixFnZ891tayRZzJjSZpe+LSDt/ReJEa2j8oOSR+fD1VW1skYjImkcmXe3jLsb4rSnC1qc7EOPM1ovV3ns1rW5UgFUqgDnnImX0QhsFS6/v70lxkgp0SiZ7N4pFVitxqWly3GukRe02k+rM/QrAywIyjBiSjxUHkOp5AaLcVE1liL0i9pURZVTlWneetKJMwLe9iqdIi9NztX78v9zs8r/t7Qd5lMW1SS1ZqUu0Cr8VzG/lQ2qi9oQM7V2pvYCo3BHRd5dvcv6FVZ5Ilkco7aP0GLdLVxoS65aJfXOlsDM6r7buatieBlZhoUiXmb9WLzHqpiXsOZKwZQVwrCFbqyZZA0TglmbuSHBEjW0YMWNK4VevZd6rxUI4bC26kVbXIy3t5Q4s+k8D++faR1MzoxJQ5b9EHj02iMev/Ym985O2XROW6XsJ3a7ieOz+wTfIwVOT89IMTPHmXEc2e9vmcadsmPMk1rsSeP3XT/Qb07x/RZcR2FpaHowf+wDfVf1e+3vJaDtSYpb0fp4YtEOy7f+iERkzBEpI8dxx0YiQwLnNvR9T3hwj8yPMs4T+xQptxec5sRmOKLfnHA1R+YEQQonnceRKSkyCPTzni7dcC9N7EJgDB0pnDLPWn8korVK6gmX9hyrmVtniT0Li/dks7Luv/Y+BNGyPZvsjmVuOtEQYxXQNKNrneNVj6txH1ZHkObYNGNvcdlqiZZWEeecGnBd5/Zi9GWchonTxJRmXDDPC+0yXmroLAT1nkWZIWLWuRyJuEodRc0hFpCM5GKF1MlAQDPzPBHjjHfBPE5pudI2sJW30oaqsChv7dgAM0At1rbIR6lMKkCppTFleV/F5mMpqzTEyviVVduiRmKwGu+1eFmMcJ3L1RauIeEilo5ox6//WxnpxYAc1VGodauCeb/f+/aZVl6HWkQO/m6DB7X2tMloC7lqCLD+bnNyEbm1LqVOr8Mr1V5U7YSr22nvvmkue6mCVZzTUHZLUbGsd28KTJVxW8GrJ1/s0fZDlonT9qseUhU+rzR2lnPo9daKKzfKp1wRWaUqKpbf12eywZbV4int92WiV+tACviUyXEmeOH0eMNbjx/y9OKCF7c75pjog+P4eMvrb7zO/fOHnBwNHA1C8B27NDLu9pw/fks9npg4vXePi8sbLWrd7xn3N8zTaL2+JiVNLdZYcNjSb08QP6hgK3L4xm3gDv3WRdjrYrZvi4VCS2nhpkRhdFvmnNjnhKeQ0p7MzGbS+ed9x3B6RDn6IW7GkbK7Ib33LcJ0xUYm/OaIhGd0wpALJ73gSiaKdswNaSKMN5wWKLGQfGB48ECRlJgBUednVQR3nq8aOPXrgjRDUESW5tZFc5TVE1vPq0qAXRXbekZmk4ALOe5S01gNKrdaqNqWpVoLtDmzCFcFENW8q7QbWeZeQTkH0zSS9zeW/BYoDi8KcMhOcP2gYT2vsPaUlNV+yhPd5hgfOmVtL+tJX1lXNFSZovXwylGBCyxro+n3gxBJaXRYqU0pPX9reNqoqUr9SueVrfcDuWfh2waBb2tM2vtY296HcqI+VjU41nN8HUaWdqyikfPqWqsT1vuqNsdK3vn6+6uF0W96+2wrr9UA19khQPb1RSlDgAHabIIcVGe0F+rQQt21zGqxZFkusSCVM5UXYyk+toQ+GkpS1FINS6qV4Yy5WmxKNyqn9QVYTUIRizKURfuSKUt+ejFK1+HBskxoTewvllZTMJoZQdCkLpLbOUoj0a2NEPUf0KxEbGyKwrYOFHxlDdcGoMv7EjCvxBr0dZ5QCt1+T3808IXX7vHg3n3Oz9/g6tf+K/vrawLwpbe/xIOHDzg5uUecbuhOAm+9+SYfPpmY4h4IFCeM855IZHP+gMuba66vL9mNN+x3V+R5T572MO+AhHhhe/aQzfFD+t6apRpqaol4CA5HIapF22CoxRBb1TOxd7Wqg/IrA6YUIeKIAu+Vc5ycsC0jX9g94f58ydAfEeQ18hTpj8558PrbbF9/myff/BVefPRt5l3k0ZtfZnN0Rrd7zs51uM2W7dk5V9MVXZzYpNe4PnuT6TvvcPvsEx5+YausICQQp+1RSsFnAW+ZxmwWdFaKKpU3NklSUig06mUrn2ZBcrKW7ya00mweXc0fG2gka/8zFc4FKVFzQahnG4IjBE/XefAayssYw7pB5Tu3No4cYl5+coGUIymOlHlHThtwQXNLXW3voc84u8rWnsj7yYAETntYmdfluo7W6r5guVwhuYFwNBKGDZujLZSkx0d9p6UoL2WZIzJOyDSBgPdW/5syk1fGlOAK3mUbAzOirZxE55JbGdk1veFsrQkixuFY8wqYQVHQWsSq6P2iGPQc+rwavKuypOj8LO2COk8sn165U0u7jwOda4u5IL6WoENtRaXvy7Xr5BK1v6AZE9HC1qtV9j1tn23lBU2BrV3ig6+qImgm4vKd6omqZFTgVwW0bri5dJ69Ez5s1uqBDQPVMGJRXGsm+OZ9yKHtIatzHOTUqhVZlehKaf9GpMDmSynyLGP9fur9VdJd4xqrea5ctJUJRoqaKjx+yYMVpNVx1cXVnrop+5VlWO+5hitsv5IzEgL+aMuP/LYfb+P0zQ++zTTPnJ+ecfboAQ8ePuT4+ISh33L/1LE9Gri8jdzs9qRScP3AFGc6r6ztzy8uGfd74qxFyWRN7Oc0ARkXAqHv2Wwf4v22Diq1RKCFdktNPPuWg1xGVdoYw2IwHLySagyUaiao1ZEpjKXno/KAcX/F0bTndPwAmQvsd7AfSfTE1JP8KSUKuxcvmHc7Qons9hd0TnDbLc+fPaP3cNQHPvjwE8YZ5Oic/cUFR97RbQYuxoQ2mC94l/SZzHqXUoxNAzWyjHVi8Yyt87e97FofWMz1dvU4aIWnrearqZCyGEai1/NOW39sBnC915CtBJScRfM9aZ41dykLqatGP5VHMKWZGCctyrXvte/W8l586MihJxfth6VoRCtiLtZiVMx4cwZMiklzZpIQFxSME9b5c42N5Byt9YkiDUucEQrBCZ0TOl9zTNI8JwpaX0axjjeyLBFbF0Jp3opzzmq00uq90ByrFr2BFmW6AxFdKTFa5MleRlvXeotLjnwRgapd1aDNC7NQFVHr69T3nnNzJurayKakfVHDL6YfZOVVFmOjCfPV17ISNMsw1vxXVUB1jVb6Jl5tD5Q67+yYUs9cQ0x1SreLA7QE5SK09bi7RKLV86qeYNUBd13rT9dVhxO1iVUTTJRinmUNa7Udlolt2ijn2qG2EoMeQpfrNVb6nbsjf/B7faS1UquhxJK1hUPfcX/7GvM0s7vdc3l7hQ+es/NzHr72kOOTU+2XFQKnp1uc81zejgrL9QHvzUMoQC6MN7fMc4XGT9ptNyvU2DvBdz3dcEzXHSPOW56gzqfCanmvXtLLimsJV63drPa/O8NRXU/9PeG5KkeUPDOmkZxGhpQUgZYcxR0xR0hsKAVFos0T3nv2L54RBHw85urpE7rOEY8Gntzs6Iaerj9id3nF+YNTNn3PVQzGQF4IxIbokwzkRC5e378UVQJ5RYdVFEJeG2NSNPmeLZTlCSzTqaAxjgpEyCY0tZSgdT4ohRCCtgrpHL4PiA8KkEiFFBM+wlRXY1FwgjQooJEZW3eAks0DKrQicwHEFbxX5YULZCVO0nco2thVcOQsiJXWaA4rWelfpKTlX40waD5NKFk/z3m2nxFnCl0Rh7Wuyi0zwhZDMcO3Eggvme5FxohdS9u3cCAbms0otPKIg3DcgRFZz1pa5Wdp83T5vqXXq9G8khNNAdUP2w2sHst+USTiIhebwbyS160A/nvcPtPKa41jqJOVO7/DIjRbzy87qCoMWESVIMSVt3J3W/kXFoKzeLVZRG41cWqbD7fyRg6F3DITX1JSrO/vjsZaycg6uQ/GparTogso52zOm4ZdqnAtq/3BkFm5KPdZXtV1NTh8vcJ6IS4TvBqWy/0uHqcSvd55FIEgjlQKl/Oe49Rzc7nj8uKSR2f3CdsjQj/QdYHj4zP6viO4maOjU673Mx88e8ZbD99QKy4nnE/s55np5pZEYB4NEj/tyHFHSTMQ6YZjhqP7bLYPwHXLuDX9dNdgeDmX+L1vVfIYvx+FF/41Lsk8LXte6y/pxeFS5PbiQ66uZ/YjbI43xKxMIK5kbj58nxL3XB8NPH/yEZIzwXk+OPkiDx6/xdnmhOcfvMvn73WcHB1z2T1mdlrTFpLNkPrCYlJFYHx4YrWDLRxuOR2l9SvL+275VEjF+rUBFWHozBBo6D87psZCtr4w9LDpIQwd0g3Qb8mxkOaZeVJFWPNXIWn7DUEVaU6RNI/EcUcfo/IQSqI4zWXp2vCEfkAQ0vFMwWsesKDIw/XKM6UtqLeUneDFa72gF8jJWNEhiUYtSlKI/DzuiNNInmcbu9yUWBEF9VA5SxtXqafCyhu1Up0lpkkka/foKp9KjeC0VIPDuaU9S2tW26ab2P3UPPOhbFlE0JLNqjnfxWta3RiL3L27HmrEQhZ9RTFluawtcw4KfJ+667OtvA7cEHsbzeKClWtgxZfVcCw2gLK8xBoadC2MqGcrtU12O5N5WEWMYsU2t1qoNrkQMY4wnSEiGMKwHrJy951NcipM/q6cr0qx8BIkq+1ZhcQSRs01HFAwdN3i9WnYsIYDV6CMpNZ1rZxvFlkdv2XAVyNyGBv3Vh/TjIA6qVu7ej0+A+RMHGeCEwieo/MTSjnWHI0Thu0pJSdCGHj7y1/ixYuRy9tEmsD1vYZT48w87dmPE3PMbDaOKe6I4zVxf21t0h3SHXF0+hgfjkE2CBGlYbIxORjPlakhy+hKqfmeO9tLltHyyup7ayFWEWovKYq2WJmk42k6gXlHGZ+x+87HzHRkCUzxGTkXwmbD/c9/kftv/h/xQ4DjwOW/+n8Rn37CdHvBnmdMxyck1+OnG66uXnDrMu/f7viPX/0qVzfXhqRc2oBqPlQh4853JnSF3gcIS32eKxjFkiXenV+YLayuSKhwbC3yNhpCZWKwfJIgDN5zOghnUngzzkgCHwTxXQsru6gGicuCBMGFhCuVSkwLuHNOpGlPmnatS7UL3iqM7FVYwXm/PUZcsE7Z+rmw+AfZjANKIqNk0Ri7iq+5v7KANEDIaSZOI/O4b73YLB2luT4LsxnIUMP2dYHnupIEjGYJk01htaZrGxJ9D1Ux6TtLYnGlGhmw6wbnWfeVqEKxIn3bdG3LeJGcFYjlCkuovADm3bkVRdmrl8DKuka7J9fcMKbIgBWU/nvbPtvKq6zDNi/ZAatXtwxphTY3k4MVjLd6LLLWD2V1vkVgr63PA0Elyz6rX9u9NGuGtUWvE2ot19YHt6OE5egDzbY8Yw0FSBOUbajabwvjgibSSylKwVPSQvVS1DTSHFcFb6gdRane2mphrCezLEK/FVgvoh9ZBmS5q2wJY8AHLQ4V65QbQk9wQt93dMMRV7trxjmz3R5rQzsBKYnb/Y6YtDi6elwpjlC066z3gTBsCN0x4vqDMG5pv63HtxzepP1dZ0J5aQHfMU9lPfvuLvjD+VpzKHs6YCaXwFhEQ1xSkDmBV+UiRyd0J/fxm468Kez9hskNIOph5N0tkecEL+xubskpc7275cPvfJvnl1dWmVptH7H6nKBgBd/hitbsdM4jjXJIhVmz7EUZ9b33Bnu3Ym0wmiWHC50aJ040sW85RIcweMej40A68TzYaJNOQkYSpCzkiMLVs6jwL0JlM1G8QiGJCvWUFOXnUjSlovNfQQqWK3ZadA61LEashq2+oELOCgKRbGg/U17BqypJObZ3W3JpLW4UaagQeYXtqwLK0Lg59ZUfWqTrvHpj27i7rg+m36J41vpBi5DLyphdfj0MGa4OlUOvqy7IZv6awl3knP0hsDQrXW6rnvjgOrJ82aRx3ecAJ/C9bZ9p5dVg7MUmmrkGrTpcFnd68cerJaxvr9k364lEhc8b9VKL7S+V5lWgtxNXhpqa5OYuq6B9ZxPG+4r0W03IWrsha4qdRQWvAQOrszYTSkEp+UBO1jobpMbVK9S9KiVjzLAmcc0ttX1yya2kYFHj63dQfx5aeQ3c0jyu1SQWLI6/LMROKgmsnjGVrDU34nGu4+x8y/Zoy+1t4pNnF/TdwOffeqNZGj54nj5/weboiK7vuL56znR7SU4zXegIYUu/OWJzfA6y1XGSGSlDu6c2Dw5khimspoykjevipR2OyEqtHyqudtLluasHUEEz2XlcOEE2R4THQhkn0hRJaYM/OaacnrLv74E7w8VMurri43li8j3h5DW2XUCuLplePOX081/k5vkN++kFu+JIT5+SLm9IqWjtjjhVWKIUWDhVMC6rN+ZFqH2dckb7jMmdLt5GeFuVOgIxRRBw3us7tUmdkvIKehSJ+cXHp8yvnfD68WtI5wmuI0skzk57fE2Rec7Ggl7IRvGEy0gWEA3wpxhJ04jvBrTpli5GBzivnkR2BReUuzAjIK0yzOad9uby5vOEOgmcwxVR9CHqOdU8FaU0Qud5HJnniXmOiIdYKrVcsRYxFfSiM2RB7y0wMGfrpq19ywvnlh/Tg7KXVc59OUOdiesUxNqw10NKQyoWVxWdenFtrkqVDyvZZHuugPEHn5tNu57aaoBWWda0ac118wNe51UWT2PtTdy1lguFJAVfSrOIqkBZ7734BvrT27E1T1ra+atgznZ9p4TOZilVxo4lQLZCFq6tI6EJwMI6IMBCivmq8BRViK797qZVQay4k6ajbXy0RcG6+LgyxEuquDA0LGbKLa9OXRfUop4P72C1Tlafabij1DzZKvSxiH1dFL5+UBRyvtke0w8DznvO7t1HxPPtb33M2ekpznuudrfcv3ef/f6Gm+tLun6gUJimPbvbS0pSctsQejZH9/HdBhgoRBvBroWPX70Jrxz+9R5yONeWgVjGa/1+Dvc2ge8VRNxZXy3fOxVkmzeJux1lHCk4JhcYs+Pq2x8wzt8hBOH4JPD62/e4eDrw/JMJGZ+x3V3QEzk/PeKTb32T66sbus//kJLJOSOFNlaJ1jJEm6nhcNrKA1GaJwMcOI/5MfrMuRRjEVfF5716aHNKdMG3salsHk6EGAaFyYtmTV3fQ+dN+amhFGMizsmIaGGaiyrbrD9T7RtXx6/AHDNxSoQ+IbZv5fcrOUPRdjUaHfZa+uBqHVbB5UJKM94Y570ZKkVEZ4rlqVzyykhRAF+IaSJmbVey3+/YG31Z8lronIp25O6i4HzBp2JwfB33gx55LDnCavylUpiLs7VTKCUZ+MqKrO3+63laZK4UZRQR/enxrX+gq2hjWdWf2WJ0bkljLLNZ/9+QkXZ+SpVCsuS110acycJq1OLKsl8t4v9Bpoc63BbVAiY/SoUAL69hDVOoPuzCjFGadSJlgfdWz2uBkFaUISvnqL5EWENX19sB6nAdEqjhQzvXga1TVte4Y+WroqnPV1jnwtZhaV2kZtVVrytnK0AuLYnerK3qlbVxWU3Cu09V1lMcKilwHZu1cbB63LZAbfQPzwHqEYSA7zecnZ1SJLCfIte7Pfcf3cd5p+AF55jnkevr5whiyMKJkvYAON/TDyf40Cuq0P6roZKymgqv2kp9iPYKXuV/rh+srH63M9zJo8nKRPW19QRmncsiDHAKPsiSmOZEzBMpKYx7jpl+0zFsj9luOm7DSEkzKe5xZWYThPunAy/yTBdHvnz/hP8QPBciZFNitYSjAQnsPv2S3m/NTwXNY9Y7L6vp4J0j9MpQkfa5WTRi08MYiyjcIXyt6+ZwEFs0QHtFqZerzC3F0IBKU6QNOaGQGxPFks/FGFxSI5CutkS5e10xWHott5EFVCLZPJ9mHNf3oz2/Ssra9XhOxJiYo6IUxzlRXKS4CF7h4bmAkttLW/9mq60AYwukvebn1ezU/SqLXJubq+Oa7FhP5iofagSEBTBRozHrdXkgZprRutTN5WI9uur0d65FADPVy7LZXIVaE2yyusid+/wets+08mogAvtd8zxWkNs8Ed1hHT5chL191qiS7TubNVq0Wb2UskTUymrimPVaFVZd9pWzcFFoNKH0suJSIS5I87hoX1fva3nRpSksoKKVbCAq+K+VGDYrVY87YM+wxoiHybGVsj94ztWqOVg0VeHb4S9p7EXpI9KS2cWUlxKM6h2mapVLQfyA+I4wbHj85mM+eXbNxc2e/Zzo+w0hOK3forDfX3N58QnDcI807UjzDZL3uHDSkIVSASRS6a1o15VWcFsNj9VWx0Da4wCZpXZvUc9a63TXG9atdoxuxpGwzI8Kpy7q9ej9FXIsFeDG9c2OaZpIMWnBMB7v9b0MXaCThIy3xDjiXOFo0/Ha2cAHLpNd5n/1+Ix/1XV8Io7ZVWFmnJqiXu8asSYsXpOCkZa53Uw3e5fB2tyLc0z7iZK02N05pcYSZ2HIOieh1XzpwlVYT8sm1hBzQaHxRUewmh25CPO8QONLidrluM4nAW/XK0axRMk0D0ZPh71KRIp6/XapLKo1JWsR8SRL8YS+R9corbRdSWz/5jmSRPBdNLzhTAkF7wvaSaeOuyox67+JN4UIWvhf2zAVrPBA0OLjFQdqRiCvIzZFDZEa1q1KsCpnkaa4iqClC41fcGWhHSg/9W4TSg8Hyk5fE2Jm/xvwazHc61rCYd3ca2jZmZIt7Xm/1+0zrbxq6w0cCvEWZdnWVhDLO2ivttTCuRpmXGCAqzwAAMbfSURBVMAaBxI4r/I/TZG9wvICE4hCqESY0BBZzZpdWVqy3JDdn533FVbI4f3X7Y5J1D47/LXWjiCVEcM8rZib8qoow/VVDkEIVS1Wv+DVltLaUKu+1h192wyv+lw6Hp5oiKt1+Ftc4LU3v8Bm29H3Hbf7wrff+4AYMz/yIz/CJ08+4mg78Prrj/j6V3+Fy8unjOOeOH1IzhFI+O6E7ckDun6rgsfeRebueNo9Vx1UXu2Gibzy41dsDdT88lXMCJrnSGWo9x6tq7Jj4qzKaY6RyxcX3Nzcst/tubm5be/Ke0cXelIKTOPIg8/dY3OTkGHPZRFOZOY6eNw4MW+OGG9H8jd/hfPOcXZ8zLPbG1UqpoAEDeVUhJ6CESz/1YwyY+C0heXNkBPvOD45Ybq9ZZq0yeAstW+YCmkvaqnHvG+Ygw5PlxK9NZPEmCRS0hxXzLVTQbGmjMqvGVMmpsSUckPhpZQ4MrTsHBP9oCqy9p5qRpz1wKrr0IkW/+pad1Cpz9A8mXb+FZMpUMRpiDeDl8I0oZRfcSLuR6ZR2+7EPJPJDHMkxkTXD3gfCSEQ+tyUeahUVaL3Mq/WTWjMFUVDkAglJRrQpGoNb+i9UpS9p57P+AQxb3Td7yxRDaSMS8XGYSGaFimtvRFFi6hTTsrCU7QAWXWc4PJsz6DKKRVVls4YgyWL5cOqcrRozvfndAGfceWlExTI1iagQHW6q00tNuC1JqtJ57zCl1VvBBaFtdrM+KyxD31ZzrVJUGRhS675rsbf5iqP06KJhGLWkLz0+frplqvf/bzoQrMJXA72x/isSpvUd/txretzDo9/hQKTBV13MCD1OFsQr97WHub68xqOXIEezBsJXUc/DHTDhmHT45zwyZNnCsvuPDc313RDhwue/W7HxYsnjOO1GRxJQzK+Z7M5w4cNIt2B1qlkXoc3tHiPte9Qsx7bcTXQuB6EVzxvG8vDrbC0Rt+PUzOAfPB4a+GRcyKOswqKFLm+vuL2dse4Hxmn0eaTIM4D2j9qd7Pj+nlht49MXkuQr+PMJ/vI1U0hnTwkzo4Pnl3y+sOH3PYjL769U+UiBnvHUbt3q5IyK7nO27Lka6uRUsmkSYXdzY0yUhRrx1KXkCiLhSJZIcdka0c9YEQULVpHyNZhyplkSiqlpH/b/K31ZHUe55y1weEUFeRRlKy3lERli1hW+iILDvPQy7M2UJ3QmlHWoG7CoObeQdZOwUpRNTPFiZTV8IipMMeoxklISEzmRTmyi83r7bNbyYlK4abCXhqy0npwOYuW5NXMqwqsUsrZBJXq4a+jPGZ9paJsirWlSbH8mDOEsVGvUkVltmvOOamx0qxUXa++1MhBXc9ahOGKtsYRAfISmVrSOKJNKb+P7TOtvFqIq+akqgIQ6yVFTYLSBLkdyNrbqvmutYe1FkBqpdWJIE15Ca7Foj9NeTXhLNXK5dALWUz+5WIvKatDBVGV36J47HOpj1QXrD1TLTgu2Vg0OFBeqwsfOKBtFdv31Tssy+7tvg5VrAmJelMHCqzQiD8LBx6nAKEf2J6c0m+2hC6QUuTJk2ccHR0hIlxevOD8tfs4gevLC25uXpDTjPeCSML5gdANDMMp4jrU7FsGvuWb7oYHD0d79YCq0JrfWRahpx7LkmdZWs+86ryKmJtjYj9Otp/gfMCHGUohzjPjbk9KkZwTNzc37Pd7xnEixYjzHvBU8t+cEvvbHVfPMze3E7PTsMx1TEiKXFwn0tE9chQ+eO8THn35Hrt+5Nfe/6A9qaAIvGJ1Wq6Fd1Zjthbo0CDvNd+6u70lOAVeCAp4aMLT5o1HWSYqWCE4h++0Y3WbRjaBqpJPMRGThgSzIZAa+8taycVEnGbibOHDnFthcFvNIgeLpRoiYv/LLTJSSKnOW123zrRCMk/HRoVE1g7K88w0m/IqhZISKSf1EnNBzGNCMinW3Juduyqv4JrycgguY40hNTQttU6Q3ERBs33v2GL1HVWB07qv2xCsc4fFWE9cTblU2QVGXVXIKWvOy6I5SxpAFZtzfukhZpvH4XK9R7cM58pEn+MPcidls7KqUKwM1M2TWDlRtf5DQQhldXw9VzlcQaCLuSkhs5IsbJIkGNx1UT/NchIWGDjLAgEtgq73eICMLOVgAtZXXHMM7ZFb0fSSm1rXlrUt13xAIa7qX0quyWA7tlSLqN6HjaUt7wqXrWNQuf/WOKnVga/cXpblh4q4jlnKhaOjezx+/UucPTzm+SfPePbkGU+fPeX1118ndJ5vfvNrvN6/xvXlC9759V+GMhFcISCEbkO3PSV0Wzy18Ni845X3AEuZ9Et1kncfqw1LFfYmFBZX7RVK63As6tSa58huP3Jxed2mmqK/MinN2jF3N6phIoUYNexUeeJK0hxSMeLplGG+nck+spsjzhWSjxD37Oc97zx5jpxu6Y4HvhWF3+YLjwfP+ckxV+NoS0dwEqgNNUVkyTMJ5FBzsainaKGlUK1tWxOpKDquWBgqBM+wPeL06JihG9j2G7rje1ZOUkAyr93vOT8BF3a6jkox70UVkyIOoxXP63yPKasyS4k5JSWMzoVxNxK2e4Zpomy36h3VzpfOafgKfYa7hSiIvQcAceRWjOtIZHxeweWTofjIzClrG5RpzzjumGOklErMvW6+qnwbGYfEYsqyMIn1wxKQlPDiFPxSPN5Z6E/c0rOriNJhWX4vBLEJ7FYyxcLRLqx6cy1haQ0JVu8VYlzSBo7YDO/Kx1jQd1pZdhBDfLZjCt4VuuAQl9p6Uq9qkU9tacgie36wPa8DD2olPEpZCZI7ouTA+lKFsjgf5uGsvabGKKAWqhjiQIzWaK3uGkP8ikVj3dIBaoX7iiSz3fdyrjrl28+DEN8dwUhFCy6V64sjaeGVtCJHzcs5StE0ePUoqqXcrr32ulbSvOm5tZu2crDWm6wFe0NzianFhJfO8iuFL7z9Q2yP7+GD4to++vgJH334MW+88Ro3N1eE4HnjjcfEccfu5gXj/hIfsMXqGTan+HCEc73WJBUagOUAJPPyXS7Pe6CAlvkhi3S483zcOeaVZybmzDxH9vuJy8vrxuvWBW0iklOtFVJ6IXFmIZsnASowKUJKMM8JJBMT3OaRUgoheFwujOLZ4/nG8yfcl/uUnHmaHN/48Am3xbHdHnEbjZyXCtqoprGCTrSrcmUxX56DVS61GYJFZ1E/DDx87RH3T845Pjri+OyMYdjQdwNDPxCDJ3igJL79zje53V/APHG2GfBOIe3Jz8SkDPAxZaaoLU8wj0sRiJkY1cuqXtk0TwzzRJ7nFlL0KMWThkWLloDZjNf0YUZrQVeruGCNYTV6sQ0dFG2sKCmzr8oVBWukOSot1DSRppkU40oemaKUqghYwm4mAxpsLxWyr0AiUYCHx8AXSlZcsnqZuTSztR3unK1ky69XGinnPZ0PC19jEeWOzKV5hmbdreqxIHRLtKmQTNmp55dy7c5tMi/QvLI6Z1Myx+IV26K8fpA9L14O1bQq9XJXzqyt/eXnQR1VDQFVT8tqWGqTOifeXma15lfKy8IqlbtMF/nqQrK6z3Z/NRmu2zrMIfWWKxNpUxSlfd+U1IHQXOq4mvIqVXUtw3Cg6NrpVSgdhPJkUVxLKGkxGNajWvVTPe7gOVje0eF42xfecf/ha4R+i4hnd3PLxcUll9fXvP3lz3N1cUVKjtfffI0XTz9hf3NJTiMhDNqTK/SEbovzPUr7k5bry0oltxqCAze33ZwcfLUgqJYi5cPtLtBmre7Xn6oHrDVB19fKeiEC/RDwkik5keaowk9QSx+Wd0nBiRbS5lyIUUNRcy6McVZvSIQUE2N2zHg+ubkG4/W7zA55cU3yHSF0akSZYD80VFhZ8SuU7mpmKgu9Jf3NuOqGge3xMfcfPeTR+SOOj485PjnBdR1d6Oi6npFIcJCT5ojm3S24iTl2BKe0THOOxCQNdj7NivgVCxNmVHmllIhlIYyekzaC1H5tEe99I8Ot/zK0/SvqVCq9khXIt3Vlb7D3gYxYB+GMGCRfw4OmvGbtD5ditEJstxjADcla67p0Dla5vlCDVkWm9aU517nn7NUoiCSm6gEJkhYDuBlYVLShGiLOOS0rKap4dOxMcVk4sA5QUzVScI05RWVkNqOhEnXXuQnuQJnWMgctGUiLoLmzFgoaJv9+ts+88qpb9QIaocSrrOyyFi40YZRM4TUL1FgDnHNI8Di8CUSolfm5lFYkmFd5nybMmhdHCwWXlZVeY9srp+zg9yYrXqG01t+32jNTKHVirBPaB6ex+1g7DKsztGuUNj7rCpRluxvO/PRN7vxcX8trWKYL9Jtj3HDE0ckxwQn/3//xP3Nzfc12u2HbH3Etl0AmhJ6P3vsmN9fP6ILQBU8/HNFvjvFusxgYQSh4NEdVE+O03mzlYATubN/lq1d+Lcube0m52VvNpTDNkd1+z4tnL4gxIU44Pt0wBGd8cVmFWVGYvE4h49LDfkdDZTOJLIUxJWKMzPPMOE7sbmfN2WTPpcDVxxfkBLvUc5uScmj2VWLWu9UEPiKrtiWQSYSyhEoXwJMQy0IxFULgjTfe5Oz8HqHreHF1ydMXz0nzZOe3fFpJSEnEOPHON9/h849PGR4csdtbqxqnsPQxOnb7ievdjv1+1MJmrOccOj4pRmakRdvHaUe47bm9vmY4OtLQWNeBd81ryEUpnjLG4yggWB64cT3a/3K2gm2tNwQoU4IYySSmEsnjTBpH5v1eOyinSE6FPqi304VA6Drz5HTIDfbAEuOgkQNU1o5F2WqRNKLgsJhRT9TmXM4Fb6FFciaEamAHCzl6BZZZNUwuME6R/X42MMyyzvUtLSZX50G8hU6dsuvHooXiJmjMWE1At4A2zNDT/GPtAn2wIJqc+oEOG1a2DKEW3inKZak1KgeDZTlSFg2yfO5wC5tACC3H432gorGq9Z4xSL5JsVaBspLlzZiXRcRXHFcVfoU2bxY1UJZcVtU4hbIyYFZekD1zMf772kAy5dRYqUth4ShrSr0NBAfFx3XywYGX4cSgudUiL3fycCtLfTXyFiqswlC/9SqjzUhw5Bw5Gra8/vrnGfrAtLvl4uaGr37jG3z5Kz/EgwcP+PWvfpXHb77GMHQ8/fg77KdLILLptpzdewPEU7IhQc0ryK2ehGbRNvXfchF21wfurtgCPHyXiwuZAYNbv+StrQ2AQzXmqveXNU8y7SecE9JGwG8MyGCVULUEpJ7LvGtn6EDQcE/MhTnO7PYz4zQxTnvSfjbG08xVKohEveuUmTBLeAKdjXVcXJsCsJArZ6vpqR2Uc3taC1UXDVcdb7fsdzfM80gpatmXksFYWxClInJVWpbMphuIM1zfRF5cj2xCRlxklsQ+OsZ5ZjeOzIbMdCLEktRDyoUYZ5JbMsMlw7jfcX15webomBQzfuhxXbBQlnVNQFBurBpFMaBCHRspuv4zePH0fSKIKqDiIYkWROdpZhr3TMaqUUrRSjURhcYHjwuBru9bar2AQtJVaFCbEAsLeXUR0dYnB6FcXStSFZUZn9k0kmixGhvX0YWA7wNdp0z4zitbS0aBJCUpofEcC3NRwEgNBecWzqe1QRKU51IbhTpiTtoJ2eakFWYi4rT2MiaSaHg7Z3UMignKGn6uM/v75OX9bCuvAwF651/zsprQsf1k+XPxlJYEJUbt0j63nj80tVPtJm1TsCzo5Tf9aSu+ybB11mhRIoXSFFzzzMoqlHBw9/Wp1Y1fFzJmO6hOOl+bAq6Fc71GpbICAzJIe6qX79kE9EoWH+jRA8HPy3/Y4lCFuIwfpthdGBiGI46OT+hC4PnFBR998KEya4dA3wXiPBFCwAk8e/IhKU2E4Dk6OiZ0W23stwbfsDYQqpt5qEzKSvHInYPWex76U4War3vZx1qdoqCgi9XRmjy3e2rhXFbMD85AA+Y5m7laORTX8G5EGSaSeV17g9JP0548znqOAtpc0d5tVmVUUkbSyiM372MZOssJUZFtFbnqDt9lGxbdd9wr/D6nqLQ/pRhoqCgfoBNEYXwABAmMU+R2V7i+gRiSKi8yYxLmGBkn5XWstZIxR1zS/FfMqbF/VOUl44Tsbjm+vSWWQsgJNxsZb9F6ML28wxEsTK9AEynJ8tMFH3otqnaJbvJsfAcOYsnEFJnjTJoUYTjPE3GeTe7o/XgfLHKj+e+1IVIlQx2/8tI8WtZ8qY5u/XstD0rdpwq6bPl2JUt2zita0U6fSiEajF9zXZlYitbgmTFbuX0dauxkC9dWyXeX41RFlzSPcVklqgI1TFuXVamT7eAZvp/tM628VP5VTbRMitI8jDZySzivhvIAGrwdZce2IkTnF+WlHXQrfFiv4Kp1X2gKbOmmfMf2lkUbudU9uvr5wWQuJh9XcNQ7z1sfoV2nLFZ6KdnwravC47u3xaH1I26ltOqY3llg9cfLpsLLm6yesVrMYOzo6MSuITwnmePzhxyfPaDfbOj7gY8+esJ/+k+/zJtf+DxQuL664NFr9+n7wO72kvfe/TWOtz1H5/e49+gxu716K14WK3wZp9Lu5xWqqv3+iibcTde9qnj8ZZEj69e+ut4ibHwNQxv8uVRvIJny8g7vNVeayt1wyqrg3aSL1hMlpnHi+uaKadwzj6N6BeUONnI9/+0ELdLpVi9YhGbxg0Hha0Bt8QTM7tCQHInb3c2qfjDR5YAUDdGtoxFJSlM4TgIpwjw7Oj/TBw3Nz2VgSlrPlXKCtOShpxRxOSNZczW+C1ozBuQEUymMAt3miH4a6YaN1gda65OcYmOpcC4QURBItFISLTQQvOvx3hG8I84TZ1lwoWOfM/vdTpGF+53xGe4Yx9GWnNj71VYx6qJYqLBgXtMCdtBtYcPQGjlTdrkCH3Qu1Hq3YgerCDCjV6BYQHIdOaqTL+fMNEV248w4R8ZoSElyi6jQ5otyWk4uQtSIi2ehhyolEZuS031LMm7IrOHxjLLyV6j94bpZr8TvT3t9ppXXSq2bIjFh0QTIChlnh9R2DgtNk8JTxRsoo9ar2ILR9u26CdJ6bmlSYAkXrkENTaGibo4Ds54tJCiHifCmf6tVs7JuGl8Yy3fVRatItIW+ypRkE6R3PQRZXbdOpXWAqzTXXtaCqko6dRuo0Nd1C4fCYXnAelomypJ3QnucIUrz87nPfZnjs1N85/nWu9/mgw8+4up2x+9++yt85zvv8OzZx/zwD3+Fi2cfc3P9nJPtEacnnXpdm3PK7ZVZzPrSc3XqYNFALCGKV3tNTVPpvddnfLVzpfPijsu2HsWDcKL9VLtI24z0Xc+N7E1AKztZEEFch+SorA/Wjv7AkLFQdoEWftzvd8y728UDkNyWhcdZnkgIbT5rDqLm9DlYI0IMrhUa+5SZXX0GXSvrx87GQj/JfjWPMkm0s3IqyRjV9Xzan03D80kmchJiFPK8NxSuo9ApEwxFqZuKvRdn/mHNLwt4HxpP4JQS4hz+dkdOhWHY0PUDWTzBBetqLBSj1PISKE7HJ2KsEcVquooQvMHWPdwcX+JCR0S4ub1lnkbitOf68pJxvyfFjBRnxRn6frqup+97vPdM42SoX5TVvtA8Un1mlQ3J6RoKXlGeXe7oDZc8TzPTfiYaHL/mtQ2IqopRbtnmTAmOaHkwcmKaR27Hkd00stvtGGdDaQqNVQVbo9gaTbHQp0zoPMF7dtOeOUYkaedzMUVLVnCUiLZ/mXNiitoUtua8RKztTlsN9T3+ACuv6srqy1sSnW1YFinW8l2Hyss8r+CN0kkn9WLhqvJaqKHsXJgyqlhszGKq3t7Kdzmon5J6ltVLW+R/U4T188VuX0IENZTQ2pWY4srti+UyKw4R6iMtV385TFjDee0263NKe8wDYbq2qhfZXRPRSzatKsVKNVRQJo2jk3tsthuc80zTzLvvvsM0jzx67RFXly9IKRK6gHee/e01+9srOu/ZHp3SdRviPlLzenZzB4pf73vlCb50r+vXcldTrZTc+sByuG+5Y02uR+Hgb/NClRJKGd0LNFRZsfMqsrWQZVFcy8/lFnLR3ItCxqMaMeuHXxlK3Pl89aN912pxalyoggcW+9CADfrMxZKpmgup9UoanUhSPbZc00gLZ6cpr4IWEWeBPC8jWGTpd64EsMXaoAhLcMMEote8lUMZIEQE7wIUYewHQuhJ3hHwBOfogtN2Is5ZbZsnwaoAl0ZcHbzHOUfwwrjbaXsXcezHmRRnchwZb2+I42RjD7UZU6zvMyvdVU4VXl69l6q8DMJvK8ZlGtt8BYaJKxTxzPPMPEctkdA3oMXcZqk4BNyo3IZdILhgXZ5n5mlkP09Ms3pe85z0OFnq3UAffpkHAiWTomP2nn1UAIazcJPOB5Y5YNR8s9XfKaTfUimClkhQG13qAOWDcOpvfftsKy+hReurk5Gx8JwtDDPyVWmZ8qrhG9AkqAtqM4nzNvl10auFIauQIE05SEt4oxeuBRxrSdlCLFocuKgSqBCc9WfNwGY5lcrjYmihJTeWzFJc2puU1b2p4rpLknvgHSw3qb+55ZO1zL4r08udnwKtpmT9xcJLbnYAC4VWkUI3bHn8ubfpOk+KE9dXt/zar/4arz1+zFe+8mW+9e43OTrdcnp+RsmJ/e0l4+6aTRc4PnkABG6vrqnecClFw3E2Pst91z+kqZlPc6ja2Bx4YMu4yWIV8FJZwd2xvuN9gVnV1hRRxBn6rSxhPlHBKuZx1X9rFg8w1gMLqykbRzZBYddchWqa5wKLtXvnXTXjRliVdpv8smeuoerKCVpKxmUNP+ayWnNAITXUZH0hWhxfViTEeo0ohh+p8GsE70oL1WM5N2VUX70XpLW7l6JUTU5UUU3jjA8B7wIxOAIOL47NEMgeCxt2YF5QSsUMBxuNkrUTtPP4YJEZgSxCLk6VVZrZX10yj6PyqpoyRoSpaJ1TmnW/lJIqsVKgxDaeldUE0aLlGhpOXiwVoeUD+MA0zUzzTJymZkzGVAxJqMhnTQdmcIIPfVNccdxb3VxinKN6b0Zxv4QNi7EGGQCtFHJUoz47RzRvWHK2ljkAmeQLsThC9vjomaw+L8ea57bItBHd1E7UklWpfz/bZ1p5JXtp1ZvIKANzXeYOlnCg1Pbk6zYQ1kvImJiXokIOXJWK3KtItkrm66S2BsesGLWtaGJSpcPCu3iorO56zSIri9gtcwqqraXFm42bsCqzO6e6OyXkpf9Xr0zu5EPWW1nG4c5WGy2vz792TpYPdTF2UmtGVNHce/CQ8/sPefToNT756H2uL2+4eHEF4hS8sT0jhMTZ2Qnboy3f+PovEadruj6wPT4lyxZSwTGSqL2j0Jty6yddGxeforQ+NTZ4+GQ1dPoSzyOfNn4v7yPe4YKj64O2RS+2gPNypTVJ6ssKzApe7biStJamormaxVNM4d7Rqy0awOF8EWr4SAl3VTerQG3Hr95rq29cbofsdU1p+I1Gy+aKoYJXt9f+Vz07e3etzKNUBaNVu5WCrXV/aIcvT5KLKS/xEOdWn5lKJogqr7H3/P/a+9dYW6+rPhj/jfmstfbt7HN8jo/t4xM7jhNCKThv3reBAlEKKRdDpJTSVIK2X4iEUGnjSFGCqlI+kE8EIQGVQkulVyjcG74kBQkUav65QBTl/+YNaXOB+m/IzY59bB/7XPfZe6/1PHP8P8xxm/N59rnYJOmO97T3WWs9l3kZc8zxG3PMMcfETBWtDn1f1pRW/eD2eSLQjMTUWJz0SYL9cWZAYjESZyyXuzZTzUmVXMJeHnD12h6W+z0GhGOImLHK7ioPuNKXw9aDrkvg2Qw868Brc3CaYdX3WC6X6C0uJjAQmdkwM2Pos5wqvQTP5xL3cQXuJe6jOPjwoB0xIFM50oRT4b3EjJQzeh6QqDie9F0KI4kxk60FDAbPV0hUgH6WOgusnCBHpaRyMnqZ6TF6DECfwf2AfqlbKZ5fOtTgpXtgQL5nyohMZfG0LNiSmG0cvJS5Idpu2djni9WmidloKxqSrksV85esGqWg1ApDuqDQd+GeOxVoBR8iqkUuADvCRLXXuAHZPb9qgWpavMwPs0mfpm0BuBLcIcVzq9d93DxX/jGTWQsMVMwhCl5WOgFpvsDpO85g+8RtGJZLEBOu7lzFlx77Eu6//34wMx778uexWF8vIYiGHsv9q8hDj/W1dRw/fgqr5aoc5zIC5Fg//0pQ7zu/WbWLSi/U61zaj6gzM9qWAkndtckpR5FY2n4qjhqpk2C8VNaW8lBCHZVjRGAx6Q4ERGY5YFHmKcw+DnRWpHMYDh6CzMXMDZv0IzCB8Z3Bgb5TNUPuisA2PtSZanTxDx8MFM17gK2ThnN7zautTMeyrDMX125lLlYFEtaRpZ0CeMxFnA7QvDQcmjsmrPqhAJIomlk2FedVX2YaJC793IFTLv7EDOhG5mHINjspm6Z7W35QcyAADH3Zz7eSU6hncmQCM+RYEapmxDqrVTpzn4BhAPdd8d7suuIh2A/I/WB6tcXM126RWeMS2cNIDQNIzctDligb+vwg45cKAEr4N53ZdwCIk9FBZZQ6tzABWKGEy6IMThqJA7KXjkCcQJTBXZbZl5yKLe15IemQg5cMUhA8eKp0BmnYINdiOzkyQMOmqGtrXAeDzA5KaqEkJh/wZL8AUCpmwlJDy4VAAKu4bXNlH5KkTGiNLPnZGoQDV3yf1JYAiJs2DLxKniLGKWjlKrusbLL3tOJUDbAGIo1MYb3EkMwpoI+llLDY3MTJ03dic2MLz51/BsyM3d09PPX0U3jV//UP8OQTj+HxL38Rr/h73wwCo9+/hqHfA2iGNFtga+s4zp+/hH7IoK4z55RxaV5/o5TL2fhS+C5qCQXdeApEAt3KWz7zUZeGMBdyMqUSgLXruuJ9iDJTycMA1jBIsh5m/BhT4AmdxpAILPcwVQCAPWO1NEGvtVLHJFT1NQDksl/HNTYHLa0G51z2pyloqfmPylqUKk5q/jZrQVgv1nUxa5oIUwa7cgB34G9J4qOsANjMrjPKBuxyr/BMQpfKdc6D7H0axENOzshKhDwABDmsFcXa0We3aejeN0odUvKtKgDAw4C9Yc/WOIfgAFaCY0t9SaPhFOuMmYcpleC+XcIwJGQxMWtUDx2vNuEGZNYn+6tYQkGJQ1eHjNz3FgfSZs6k7ZDlEXHEAJdtAWU/a9nsrzKteBGy8X/KxdoFGgCpZ9l+w2WCIG0rJsLiMKOBlz2u4vNLhxu8wnledrwDEToqrtOdzMjUizDJbExNWBZ0F5CjyMnWzAAfSD5cwsBhh6+gRImAFG3X7ivIlkyjkNAUQa38FU06ASGStlYqCET5zPajBhgCeYy/CTlcWk3yvXXiqF/wMtm+UbxDZMCpbclgzEWJ2NjYwt//P78DHQGXLl3A4195HMvlEpkZ9738FfjKl7+A/f1rOHVyGydP3o4LT30JVy48jVW/wB133I7NjU1cvXINYEgsNxFwOgBVQB6QvI+4udAAUQty1prY9nCZ/H40K1bf1SyUCLOZxMhMpQV9P2AlG3rNQCPCztwCc2+gAcBPCgDEjpu9G7T/1C0b6u4c93OJAkRSlqxJsNCHbHom2y8cOeV/sjh8JbiCj73Km07Ik0Q4ZtN5BDRbQCZdtw40MGVQAbYQPQG2zlaeT6CumLD0vMTMEtuTGQM6dH22Ayv7vkSMUCVWI+t0lEpUdABDx8i53rpiip4Af9aeFj4oumU2k+hytZK+L2eBGdukrgI9JcYggYS7TOgHATMZlw7yLOtQJaVEpr/osS0qLjK4zMQkxqk6j6i80/cpq0MVyfreqpi0M1ncVlPsqfhRZwy2jJI1uHNKAvxZZpuEPhVLwJCpHNw5DFgdPFRvKh1q8NKBZ2sDSUPvk61rGZMBFhG+DDV33bZI/mpG0azhJiGWgWYiitv9WdMmJpUPUbtELUIA+H6IwvBuItRNx7ovSDcSGnx4JQotoAzoMBgOXkecMoTmxiqXlgeNPOq2FlCYDI4t2UCkDHBZK1AnivXt49i+7RTWuxm6tRn6YYXMwN/+7eextX0c97/8Ffhfn/5LLNYWOHbiBPZ2LmJ/fw+rDMznc2xsbGGxWC8HK8phiVZbUpjwUFbklS/JXL4deB0nwjXtF7czjoEsZsGFF/SGevg1RuwCshQchqjUN0tw2YEdvFg72forzLhL5SQyuHixsWvTemKz9ojXC/6czNLV/KTVZ6kpCaOrsqWmOVMNRCAyk/AqCnCwl02x+JB8311pFwGlvEQGisZgUKcjMoWijKcApIDxQhLw0eC0RMBAch4W5CDQoWxPAHT9OJvipyHbOqhDF8DQNYHKmB7M4aG98kBiVQpRNmorzXJG0mgbBCSNVaasEs3UVECESYEj2WPaH6UPZJxnMkedPqOYjFFmQVlnmbnMJPVIFK0Hicag2zAAXcNjIBdHkhxMzNHMrKGvwGzAlcwCVNqRh0IA6sp6br9aFQeW/CIOzKvi2YGLRIFS5hVNwISQED7sOYprMgWoyH6UwaIdwVao8Fn5HvnPjt1w+7dql6oxRvdyFRb63YSEaq6yUTG6yJM9rLOc0DaDXBiq2dgnF1YqEFuZbKndH8YByEOymCOKAwoGCJ9UGHzz+G3Yvv10OXocxVvy6s5VXL56FbO1NcznMyz3drGxuYGtY8ext3OpRFhgwrHNdcxnixJnTc9G0ipGqQHvxyBSIiS5cJQ7tcJxAEWkDAUpDo9G/OfRO3U1i+laQvYIv2Y5Z03XGeIOQOtSK9d5I8vBonpUiFdASmwtBspTCHwUkm/p0GdhIKYc6uAZGgYyV/mk71NLd39lrOHVnpQUXhhbCmwUoVad3MSbZLbq+CcCl8sGWgy5HEsvNFSaqM1BY+kocOrY81JawpW3dRYIoLj4N6NF15JIADqDAM4SxLvmkdEAQjAjZ1/jUz5Qu6NtZWAAFs1EwSsD0l5kXSP0sjLK0osCGgALN+fjgnSKa+OMbf8WQFTiQWbhPdtvywAjg/oESl0JYCxbPV5IOtTgZR3aEVJXvAkBiK25gFSSMC0e4cDPLbLoAgA02oa7Prt0qpaXdCwwzKtKNzLLGIGNfrjW5fKlFjTm/q75B6XStO0m4rWa+lYaYw9s13SAZWUq8VBr5bIxXMApZVo9CtwF4MTLAZw1JVMOihDJEl5oe3MT997/TTh19h5c+MKjWO52+MqTT+L/+cT/F9/2wP+F1XIff/HhP8OZU6dx111ncerOO/G5//E32N0f0M3muOfuM9i91mN3d4UOCb0QWxfeTQngwTTUMhw7uUPezil8Io+DSPGhCXB0AaobOhGE9YHqQNnESh1m3QxrsznmsxK+ZznoGkBZVJ91nWnQiSYUBOgsXBbl+wHqnBANlkUiSvwXZTnL1zcb2zxIGbWcOSJtDCF/gmVgpMUA5qpPsubbvmMYbK7ybDSv6KTVh3gdsoNbPQrsQZQjNMnoC/WgA4sjEpnC0xevkWoGxAAoFyeSRIwhZZmxFLAoZ9tPNBoKpTqDJAMij2gflVN3mNRtPim0T8OHKU+RKkyBSJyoCnSbxfWXWI7MUcAdFBjK+xReUvADyJpVKRcEjwVlNIcJOFtPBcyT2zYOiVI2FIOiZM5IK5HX0NOtyz7AF5IONXh1uhdDbO1m95Z1rRL3STyuEoG6EibGXI6FacT51YZ/K4QSqUIq2olqIwJaHanjKKzDIw8gsLEPwwBOmh3X0eD1EMIokspnNHtWyjnAGmFDfcYFbUeCoggBi97RTiXkj6AuwCzmNVHFTWAHNZ581lDalNDNZzh+z72gzFheuABaW8Pf/vX/D489/jhAHTaPHcPetSI8T5y+HT33OHfucayGDutrCRtrC4AW5bj1YbBoJQAQz0obYSmZSJFnfEYQUSwag/RjGoIi8oVP6bdJB4uaGGJW6ZBmHZJsguV+KJtkBYx0YZyorEcUqwLU80F4Y7DNzSV5613QKI85rSIw28Gp5SGfVQn/eRIzIkjWcfReAKlIMw6rjsQV/4K7Aow6OxIhqD4hdVuczszBr9TQ09vPpGbkIGQ1IHaYcZiA5gAqWl/q5LTmDMpCR5JQTbL9ZQrAIhdJGGIfM43OE4HCQE7uGWirCVxmfJly8YTmsoaokJQhIaMGOTJGwLpU29WSWEftm7Lpm8VUqCOAfa1fVWTpFPVELtULYGxJFR0yTSVGytdN3GVWG2XEixq8iiOAmwnJhISZCyiZSmOmRBTwUhncWLOts03zjBzGDjRF0wyaZQCw8p6YD1QRrmQCh6+MONOKwGXFSxU0UGvRN71alTzNZGsLgXVDEjCqfsU8qBL00aI0kZm9Er/oIKFuhuOn70DXJax2d7G7u4cnnngSF567iDvvuhu7u7tY7u9jfWMD61ubWK32cfXqFWQG1hcLbG2sY+izgXqwfXpnWJ28zuMN2vGbqijhIc1bNAyKwlBjQbXYZQoJVXkc6KGojkJyxlLqVEhxJVzjVgttLmndENY5TKAq99WkiZVlb5ibzLVpxmeaH/ur7MK2lliSH8c1P1QOO5Wwl0wiyFF1k+0VVSNZFEXfxAEHxyklI3Zn+KImtnjXf6uQLabPMkNLYPW0NDCeAi9G5CJzeDmgUgT31CuzH5cpei2QyhRl44ygGNial9BE8/X+jltfYN9s4zh0bRzVewBEdkS50RK2XnLg8A+rR6rwVRVEAQSNgO/vPP90yMGrkxAxMn3VhVo1D3Yk4WjINyDqfi6Q+3OEhf4qjmAFa3KdSELJFDQqvJDDA9WHmBWVqzN0XYvZDSARtCqPMsCEEwdAMSGHhu9UjicYkMc1OVfFnNELyASoVjkMEaIRQAXo6zFcr63oOmJxJCBwmuHMS+7B6tJl7Fy4gCe+/BX87d9+AdTN8E/+8ffj//PfPwBKwNl778HGxgb2967h2tWroDxge+skjh/bksMbe2mbiIsKX/UHWd+WvuJQR4KL7EA7aywcwGLiqiD5qv2d5JUgtSZwS96AbtFIs4RuXgAMckqwm3Fcey59pk46hOKyXdYKXOFRQaVNqQWgV8mD8eo6iz6TzWtNmZcn2jFeWzPRyFVBMJ5owI4wgFnWoLMGWITjQgTlqpwGBA11S1m2TsUsunyGGj6tPZq3bVzzdSMAZQ+a7sfK9Q5CDZJfnFk0S68niedrTi70UxgLSiMNYGDAGPLJDDGW6IDlylhA8Kp7X/u/xQRZeD/yhMoXcWwtG8ZR6BdFVtzP57N3d/RiGBGcr1QesddJHTzQEcASBciip2RZ64ttf/7pUIMXdbKeFWZeZTG8BOxESqBOBIEEwXQh5w4MbB1GPjiACU1LNldS7DBhnmh68UmzZmS/9egOs/qyxIPOMPAqcpS9kJGWWTYeWwBeuEHI5ceUamqPT95ry/A0poNmMYrmIcI2dR3Ovux+3POy+7F66mk8e20PTz93EX/5P/8H1jYW2NjcxJNPPI5Ttx3HiZO34f5XfhO+9DePYvfaDigzTm6vYdbNMMiGzRJlAlitBqT5zLFG2woF6whkXv/k6Nt45F2POJEZpn7ffCJmpAR0XQkRtTafYzGTUGSyH6eEedJ+VRAD/NhdBHwpR3kM4LK+ouF7rKZkAtfqIJYIVk8yQzDJnNn2adnzSpJJOeMqnilHFXnKi1GwKfBzIgmDNdL7tIEGOObEELCoSFtXVHRsVzDcjt9mtpW5ODWQmMV5yJJPVgKI8M8+Y6lmvQoEXABMN16p1x5zAeggG8iqIciUs8VuzLK2ndVUrBaWnuWYE11hK7OuwZQeDuPAQUFlEklZTsLrrc5O9IUmggGRlqMWHlaqiplwUCEpa3JlqwI8ZN8Nyr+ZdKjBS8/LUXByl3ldX1ChBqgJkU17BnRoVx2ls5VKC9e9RPqTg+kQtTYS85Mf2rG2SVS/qy1e94Ro50sZbhKS7HQWptUw2SLgxcEUNK6NSnqYYPSL1XzE10L8WSfU2JNKq6s2cwKQZh22jh3DyVO3Y++5i1iuVri6s4Onzj+Ll973Uqyvr+OxL30Rm+tzrK0twDljtdzH0C9ByDh+/AQSJazkrCTvHhKaUuye0L7wrAi5QrVmPS7WferlkImDQvsMIyoOkRaVNRMOiQm6eb6s1wJwr9Jw6nWudGfAjUBuhvE1KAFu8TwrM+Zadlv9o0nYeM5+oOIXBXwO3Q+YqdyytO8+g6zzaL+6UjamfRSxzfNaRfJN4GVsqtCGjB/yCmukjTi2pK7lQ0yUNqtQYZtD2/VdtupEGKgUTaCsDSuJdfNV6AwbTqygwlVfMUToV05TChYlmkcOskQVh7iOF012hYddcR9xOitIN90Q60vePqeBfCH/qTFWS7HaaTx+bfTj1lO68SP/+yYNt0Nd8v0zeuifai667hU2+ikA6BTa9u4T+2ZNAbFgnYBxiXxXMDJ8hNyvBKuyGEzoZGbZ2zOUYwOGXLntOnAFocFcFMKKyV0QlZW88m+u2NPbbCBHFK6Eu4bzWv/rcxcb0EVOL1rkfL6O9Y1NbB7bAtY2sex77Fy9imcvXsaZM2dx+tTtePSRR5BmZZZ85eIFDH0PzgMSZdx22ykkSlgu901WlHV+19uiw4F9soB4Lr0cPTETyAaUz4WVStH4FtrI8SktOigU+lt5zQSYrsHC9QUmibEpjkbi0q1u7wZewdSivKC8xYCtiZZQPLmqjwkLDletelGRI+FbGQnMcK9FkZsi/OM6ieUaFIc4NNTLr5g7SxghnY3re76vbAxK0BrEOgtNKntGABMX3jBzYQyjVkA9V0K+XQNTujHYZ8F5KO7cuv7czLoqwJBA2SQzaZtt2bFFuVkWcDMi2JcMWGWDnN+Vcwmo2/e9jY+4v0/rkQREs5TF2p+se7tyADxvt0lDvccyruGsYkAMiBt8UCtqbLI+gPR5/HPZe9Bou7V0yGdecfalu7tha18+ICJgtZpyEbapUlNlkKsANJdsDh6k4g6u6wXqiQco+pUjy1k2nqKYeXS3PzAAQ18ELZMFLo2Cp9JX1GusUpylkmCoz6S1KbSxaq/J0xq6ZFXGM5eyWsas6B+uFdrKoKWEB171f+L4ieO4/NxzePyJr+B//s/P4Mlz5/C9/+h1+PIXv4yhX+Jbv/WVeNn9r8D+cg+PPfkl9MsltjfXcWJ7HTs717Dqe8s9WuiJctWv5XIu/e8xA6yVkRaQvop6cNTjzUuveo8tAoutgUVEHfnh17otS9TXAgqEbtZhNpuhm5WtDn0uEcK5H8B9D05JNF0Sq4uYBBPJ2kWIXSf3C59pE0pvMCiYaEq9oynQAk7rGpreSGF1MMjq8V5u5yRugDo+puOmjCWZ6SC5owe5+TmCtI5ZVzd0dm+DxXswD6WCuax/62xWtwrUzlLe85aH2S/Z86/aITP4XGVhNWtn9tFECa7DdXuFCOChODYFF3LjQIl2DwAZEj5syOUQTQHgBFSBvyMQu9Kj9xnM0fKkFJaqgMxztOZguTcBNyUIQVmzU78mLTUGNFAlNwMl+kl54AWlQw1eHekBknHvFqzD4wS5Xg3QjhBvQf0VVciSgZvz7GphbAv3RoRq4wVc81RttkzzJYqCaEK6+bFsBCW0fDFytbVE1V/tHSlPGANqG6XOImzd1TyFmaXTjACLKOJtggl9tYqoGbP8X97v5gusbW3h2PY2Uupw7doenj1/HhcuPIe93V2cvftuPH3uSfT9EqfuOIvd3V3s7V8DDxmJeyxmG9hY28D+/iDx0BgRgKIXIbMuijtlYo1dONUbfzl8AmGliNwZ5CDK+yfVdyjcibcagU9UooZ3EuOwdI1r3zln2X4YlZXSGYkUWCRjmQmYKZW9SOt/yYsErYyaOjNAy3pcXag2JleQH54xCVWTxDftm+g3mV1mZAVQKwt9IFtUjEB+bSRYlR/LGTOiSZWx186c3QEpIlOAIgI0RqBLDDgQqBLZSAWVFVUtpW9dJ9V9Z/oIS7QT9igjgEfloLoOZd1IKVlvoYkSaBTKKtBIFS7NoxIaAZ5s9mWPyJfgbFK7s7E3vekl28agfCIptx15i+lQg5cF3k2dDE4KwilomdAOLQuhHriWzTQ4UipR6GwLwJVWJGCgDNtOURQOxWQxQASTTusz+6CaKNwYyBDVamRjjY2xS/ibOJUvPFIL9lYm26xL0QgB1CAL30GI6ozQqlRppVpNwmJ9E8dvvxNr62tYLle4enUH5596CnvXroEAbG1ugoiROsKJU7fj4rMXsNzfleNrgPmsw3y+wLWdqxIVoKGLmZa0ejUg6wwgWW/pHpXG1u8ZFDJQ2LNeuX6rfGPjq0jFaaQLBNJ8lD+obKDvuuIpa3trJE4nZzmplpzX9L8uFbNn0r6Ps30RELZ0Y30nz2hVlVf0Va7rKZV1eoNHj1BkKtSesEpXBaRyGGhYe7HDE5V07HJ+XAUDaJY2RbMbYh76MWRzEtA+M7713L2qUTjHRdQaw43/UfERV+31JI4fwdU+kpqEFg44FPKvwcfy1+tBJhiP2LpYoEOVqLqsTmH1eGbrBDVEB+4J5buwMjq0szy9pr8CL0PXY8eVvOV0qMGLAnBZZHi9qbyYC1hx0vBpQZshmFI+FbQ2ZOWkDlOSiDlEqBfRBaAGMPq+L/ZzOYPJcMHyHjOdfyXToMs4lvC5IqzsKJjao6QWslHIGSOzgbcrRQ4CqiFGsAaCcVIYMiUvFQTcceddeNU/fB1Wu5dx/twT+NtH/waf/tT/wLc88CpsnziB3/3d38Irv/mbcM9L78dtJ07imaeexP7eNWDo8ZKXnAZxwpUrO4HZp/sjdoe1yvEqNhVAMGGIgG9DYE28Ir8VFAJAcq0q+EZznshBqefKRDeblWPi53PMiIo52dZEBhStv6yJabRzPdl3QIm9pwdZMoDUEZglonf2jbIlvp2alSNzhXWgKfCaqP2IUPo9/DbhjMBLnUSf4QzKg1tZIS7oEXwsI83P12Ro4oGgD2BQZVEYs8pRzP4eqDhmFcfJAULggBN/1YLjMy7PIsv6ZQJstgSUvjaKBjBjZGBoqmNhwTTILktUep1Jq8IWAT1IDtdiyl1jz/o5pWOuxkWITAKRUSLXINky4AdKmjXHiDPG0jhWJxS8W02HG7zEVFjCkyjJyQRY3Grn4BO0cLP9AQjDI27vHV2lACSWZ8nXziQK527pLvgsjhmIzGv/UrV+VCVri9fTBbYZY6JCZsCi5kJtRClZS2IgSSw908oRCvOdUbnSqoqwZgJy8jYQE47feRqbx7aw3N3Bs+fP4/zT53HpuYvYOrYFMKNfLbF9bAO33347NjY28cRXHsew7DEjoFsAs26Boc/ohx6ca6a3PS7VmpxroPpFx2acCByYmD2wKLRrfYT5pvcWkhqpPZm3f2G4S3GZecmZXrMOs1mHoS8HSg6sMeikp6i41hM6rK8vkLoZ+r7H/rVdU1oopXK4IOmape+dijNkCzOm1bZFJmOWKN3q1Mw2leeMDgJUZTuDu20nDQUFQCNykLpU6GxNiWsL/OOZXqPE10RmFCBu6qm8Udw0MkBu7fBWqmCOzBLvTPdyjAiiYy9kYO9Viqk0JNfVNEeGiCt62/gmWn1UcSY3uzpI5LrOhpCeo/UxKSC1jZN/UqroPhJ4LHUIr+nMsNJHSJVrHzPWloOUhZtMhxy8CpOmeihV3VVJnmAWiAF5g8iCbg6s5xz+rwGXdpwA19Rf9FriIcPOLJgQrFXe4RkzCJBBc3nOTJ8BVAPKqUbr/KHAFYhjj7cXKnZHlQWTD6KkwFYE1W2334GNY1tY7l7Dxeeew8ULF3DlylWcuv12rFZLLC8ucdttx7G1tYUudXj24jNA32OWCGtrswLhrB5PNR1K59bCycCZRaAcMBh8yMQM62sjbqiAiw54X4T1LQxChpu7NUAv0IN1Y60oQQpeiRJmXQlO3M1mWK56XF7MSoBfAw3IGqnsUTJWKDNGq/cIGKjpaMJImFWt97xUOdK9YyqkIq+4JyrbcTtxfbECMNSh0PSZHCtsgAHLV82KVb8kjesp+eq7ylej2WZ4n32cxzLtq2JtZeXh+kGt4WjtiWvHlEpZmjKkKXD5soCpuyzgJspnnP1V8iy0r/Xvi2q0X+T6Mz4RzYYiC/x7uBcUTe0HUplUYeB1VcsbpkPtKl98dOwHAKFNAjjFa1Rr7OH9ksf0Qv3kcgBUmwoDh901Vqf25TjuFfKqFzu8DqLwPQz3GmxFKOl9FdxyvEP5zfKXAVLn+HqwtH+xga24pamLo0QmFDXMFukZShtreNkrvhV33vESrHZ3cP7cEzj31Dk8d/kSvvO7vxvPnH8Gn/h//x+87GUvAxjYvXYNDKDv9zGfd7jj9GmsdnsMyyEI7rJhNzrzeXsaELIG0GQUhFFbocN5ApRcKl2fFu1710m254aB2WyGmcy60qwAdgx/RSlh1nXoEmEx77CxPsdtJ7Zw8vgmTm5v4LatDazPEmZJTL9l85gIazm6HgkJnZwk3sG4iQFkNo2/zLw7+Us3bk4AHOqSxWhsoAu+3xJmHTEeTMETVJUydn6v1tCuS1T5R0yg0dW/o4QZSmjmuCmWzIwKNJK0HvBTrBPWW9u3p1/QMlFTR4ctK49qnfLoz7iTy9aDaSJkAIO56tejvO6X9jrbFqLmCQ77DKNSPiFjAG3HuG5xNV64DQlAx4x0gDn2ZtOhnnkZUxIMnFyGCcyQH8qmdloG2UJ5eZxgx3jre9ZJamarVTGWQcN2RIHu3yp7MsrJpb15kOlOd7VUciarbLXnwTiCfH2Akix86yOy1pFCU8nrBkjUEctQZgeiIZV9NLKqZeYsuKbM2kL2uolZSVuvA5ATYWNrCy+5/2UYVis898zT+Nu/eQQXz19C182xffIk/t9PfhJr63N827d9C17+8lfiy489jqtXr4IyY2Ozw9piBvRzEK9AyMFxQnunnNbKQr9kJq5S52iS8PXAoAUSecxgztJ271I1yyQiWwOEkcuB0RUM/ZsSsGqeEroH2pZZGhdvQ3HYmM06F4pBey1KmJyWMEtYW5tjfd5hfUbYObaOtRlhBqADoyfnUZJ2JGUQRlF4mO08ONekIQ4ppbGs9G43pCqpZIal3oW2PUXsaGU7SA6UIvOOZJY1LqupCMgsGrqxbqiLmNlUQ1VFxvtOeFTGCVMpo0tJtH5CyoPLXtT7rLSOzm7KNxR7u3oaDW+plUBn6D5xacGsfLewUXYlOFuERTy9XzwUmzqE51ii5GvuhdYDiOLWmTFIVLXTGbcer2Lt1CcjjeIXdcTR+no5WccdUPguWI4yABDflH5yvXSoZ15AUASri+VDO9/27rTaQaCesTRH4IpdzPVv9vx1tpV5sM2mnGuX+FAtA4TIEsrIBiA2/rXTddqvbFrhXKVjRSGrmq+W5n9R4Nf0I0K1kdRLJWuzru9Rl7DYWMftp+9Cv9zH1cuXceXiRVy+fBVdN8OJ207gK195HN1shjvuuAOARsPukfMKG+sLLBazck5X0z+GD6ORM/FzpPh6I6bHiNI0ZhGAI8iSqXfH3+Czq6Z2I6MkoWym7zp03cz1DgMurjb7JmbME2N9kbC5lrC5SJjPyswM2dd1bVIO+KCoDmStBaXVSs2j8TmKe6y8JQzYzMhmVGZedToy+7u+/2diNFU0C4oIOSgYtxJCnZ1Hi+OH14nk4XK/ceSqCq9/xs1gXP0XOrtiShrRpskx/C6fVP0O8ifQy3hvgv/YXvMHuGlPuRbuo6a1tShGAhn91cpyDU3lWhVUASEvvg5ckpQ7kjq3ng41eEVhnKCRFNhOMo2dTMTiuVXMLIMGV01UDqcT81vbcWWx1/+0vLLjX/7jQY7s6MvpoEMP5AyyjZJeY8tb0IdIvCBT8qCz0iJSUw6KBq0mkIRiHNLTnAsDqfbKMvCVSQnazdni36kJQhlK3qlMA9H9lmVAuZda2Q0KdOsLbBzbxh2nzuDapSvYv3IFiwycv/AsZos57rn7LJ568hxm8wWOnbgNX/ji57FcLUEd0PfXcNuxE9hc28D+crf0jdU3eb3JBxyjLMKb4KqGXK6E2thLjUwbrGBNZTZcGbFUuaGO1IMm/1CXIOhYNqAqYIIA6hK62QyLxQIk/WsCgQg5UXHgGHrQaoV53sfWGmN7M2FrwVhfn2M2n2PIcOEsQFa8a6VPU0JOCUNKyFScbMpCe/mzvWOJZCZPiDMsrtamcskbkIg2xWTYdap0FcbOWY6k58JbxZcum3Wk/G7/avmrOFJONhZ+N21Mg+USMhIylRN6iTrMqTMrAxOArqu4wI2GUWqk8L0GXzfgEUAdXBWpWYF9mAjAeF1LtRnxRWsnhfHGLsUs0IGCJLmPjQGYga2qsloJgq5vDgQ/HDLwKQdKeANUIKX2bsDfZD4GxQzChiJTdPHrwgdxmxJeWLol8HrXu96F7/iO78D29jbuvPNO/OiP/igeeeSR6hlmxjvf+U6cPXsWGxsbeP3rX4/Pfe5z1TP7+/t461vfitOnT2Nraws/8iM/gscff/yWK2+DYgJyCjO4FgGoEnp9tC94x02/ZhPazIP8ZeR+wCAmwmG1KuFbxC3ePPIQw1C56YSyrOkkd/NXjdKO4hrXrHxzmVtWN1QjFWEZyytmB/km7S/nnaEaXEIxGUwu+I1cHi8ITISBCIwBL33pK/GSl7wcuztX8aUv/C888fiXsbM74OWveDkuXbyIj/7FX+DvP/D3cPeZO7C5mGNvd4VhuY85gFO3ncCqX2F/WeIXtgE74xiIunOt70YwYoAHEA9jF/Dr7YUAMLbZq3ZaX2P2YX3g4HPpgmaODEIx6XYy65rNZhKDk+WU6AFALhFf8oBhfx+rnatYPnceq8sX0e9dRTfLOLY+w8aiKyfmDhwCoAIaBiZ3hIGAIRXQGlItAJmKGzeBkKjUp+tmFqGmVNi/ZwE2SmS83KciILliFgY4I/c9eDWAVwOwGmCmXnlGeSxFWitLkowLwIRlkvqwKnQyyyrjWk+S6IBZQk4kQA0bh1Z6IwOUFoYZ7KqfPUH18xMDtM4v/I2k+YFP2xBzeoSkeOWe8W2+tZ1EZ+6+Wduf8zEUC5zmaFeQx62gg19rHvR1fM/jhcHXLYHXRz7yEbzlLW/Bxz/+cTz88MPo+x4PPvggdnZ27Jlf+qVfwq/8yq/g137t1/CJT3wCZ86cwQ/+4A/iypUr9szb3vY2vP/978d73/tefPSjH8XVq1fxxje+sRysdkupmdhXGrRcMOEm/1WgxOEP1T1lixgfrdjM5S9nsMYl1Lh0tkHS61Zp4CTDlZKt64TbMnCjFsv2hJl0gnlER7qbRUbqjgzICDx+PY7harmrYSqS923ZDAzqCOubx3DbyVPY3NzCpcuXAC5u7leuXcXa5hYGzrh27SrOnDkD5oxLly/JWOqRaMDWxgZ44LJXJHgu1auAUfBXzWoUkSAERiaTXD/WvHFgirw0UnpUNWApUjmOnJjaikbwERVvw7L2lSzkUVlyUIFTnDiGVY/V/j72r17B8toOhv09pI4xnxFmiRoztTgOocxKnP0akSFsxcJXGhdUZ1IUlSl9J64rCv/ZGjK5OKy19bLOpse+GEXUGSCsA6t50E1/E39KW3lWz+dLahokHyfRdMjWV3VflCq2SkrTWWy9ippjap68vmoEVRFsVmhIFNJ0HjVVXTIovZ0e+ri11hf7qrxsS0/Mn6bbUpt1AT0xw2Qua22ACPdVC0ZDhyfyvfV0Sw4bH/jAB6rf73nPe3DnnXfik5/8JL7ne74HzIz/+B//I37u534Ob3rTmwAAv/Vbv4W77roLv//7v49//a//NS5duoTf+I3fwO/8zu/gB37gBwAAv/u7v4t7770Xf/Znf4Yf+qEfuoUaxQ5QCsnu9kbU2ZakYP+uw6hwYEuyDmawbPwUgucCZHkYwIOA19CXzccyUAl+HEEC2SbFUnwn+NPGfKgXNUld9iOoTaoasn6mU3dSD0xl6qiLibglLSW23wWtnb4LMS8UFAdIN3ozUtfh+O134eTJ25BShy8+/TQ21jcxm13Fc1cv4TTdB+o6zBYJd5+5G489/iVcuHAB2xunkHmJRDMc29jAlcs7yEPculAkWaxZiViYTftTskVFhVJ90x0tvN3aTr+mJSCUXydG6Cci1JVQM6duQBYBBQqCSZUPBQ1GQhJzW4lxmFKyOIWZxdQsYNSvVlj2+9i9vIv97TmQNkCzDvNUzGlZFCndIK8UAwDkEI4IALjeM6TrXF3q5MggKvveBomyYN5tFPgjKFF1d4BRTLxGQ0bwsoXTgLkOhySL/iAUN/csLzDL/mKnoY3owPQWoEA8LssrZELS+jp2C1smFXAV/kvGHGEIusAnr9OBC6NGFKrzmFQP6wscvo24MigKgJIwjOU242YcGLqh0ccCAKqSUZkO2QGqHUvWLm6ys1fZr00t0D3P9ILWvC5dugQAOHXqFADgC1/4As6dO4cHH3zQnllbW8P3fu/34mMf+xgA4JOf/CRWq1X1zNmzZ/HAAw/YM23a39/H5cuXq7+S1EDWQlWtSFXfG23H1m+MoO52W8CqmAh5KEFT82qFvFyBV8tiFhl6iNSpGDmO8bKJOg5AsvrYTEo3dFYMdQN9jmBmEcsPZW2sE6CxZ0YvOwPqFm8Of5O8Jde6RNja3MTLH3g1emacf/pxfOXRv8ST5y8A68fwLd/2bfjIBz+AvauX8A/+j1dj50rxPNza2gb6HZw6cQzHt49h5/IVEFMJ8WWDpwaXSqGYeCK3w8nsSE0GJjTcnNvGpLyZZLNcCoAXTGvTRIsivnRrlxJmXcJisZAYhwlDTxh6tkgbzCVUFOYzzDePYba2jjSbIwPYX/XY7/vihZl7gAeMOk3kVgIKkKkAEiImEOZImM1n6BaL8jefo5sV86F5E8aTyTVfGSokmn00WDnIuKCiQH8VdCyxPivC6LjRLygnQtiJ6Ck5aFHCLHWYpw6z2dyi9OswxsAl6nEOK6lad5QyLMAB62d5P666RlW4xoHnI4AncprKZlSY/9YZp+pSkcrjl65Xj5utq1bTgbeZtx0ARdoRqKxcpQ03Wc3rpOcNXsyMt7/97Xjd616HBx54AABw7tw5AMBdd91VPXvXXXfZvXPnzmGxWODkyZMHPtOmd73rXThx4oT93XvvvXZPtQSmOHx8wJCNtPKnZhk7awZ6O8ys7FiBEsmZJYJ3Hgb7KzOtAeAMyrVZSmCqDIxUzEJVFBCXfvZpukkwi8Q2RhODByC2m5U9uQxU+cVep9TYnElvoGXT8L0FAGJQN0O3WMPxrRO4dPECLl28iO2t41iueuzs7ODK5SvY2FjD9okT2L7tFK5cvYy+H4qGjxVmqcMszZDFdVvLjgNANeVYsbru4flWDpiaTQgkMJgLMOYKcjv8DjATVhqu2t8q6vEoL12Uj3ptl0rIp7LGVFzdzewnelCi8sx8Nsd8bQ1IM2ROWPbA/rLHqh8MFuqjNthmcL6/TIDeCO5mNo320c1mMhPsghJAoyb6ESnss0UBsKIskf/piwZM5GUHmrmyR67EaT3FlMmps3fL2JLx1XXlaCStN0NMlWUc694nN7274uGchYoBGTWNyB+txf6NhHBjzhzRJWYhl6ZjrYqks3vOJ1VlgsnaWmLrXsU8b2UbO8c+8jo6MVyeqjlSZWb1B2MLe6fyfQpLMDmPRtwtp+cNXg899BA+/elP47/+1/86utdqoTcTheB6z/zsz/4sLl26ZH+PPfaYlwXo8K00A8k1AFhVmt2rLsm5N3bUeg6u78Pge7eGHnkYbIDolForVAYXqllXGIsIRv54sTCVeA1Wz8MHXwyeqyZQy7J9JuSTIHU5oBt0luboe3B/pfkcs/UNbCzWcenZ53D5wiUcP347GMDOzg7OP/0UTp++HSdOnsR8fQNXd3bQ96tSj1RCMiU7sgQWqLSeITbejnqVWoNn3f5qIIV2GHeQg5gn9oyaZLpEs46ms4cW9m19xbTMenBrf5bgvKkcjWJ7oYqjT84FfDoFr/kCs8UC1M2RMcOqB/aXA5a9HFnZSAFzOGCPVK/AxQDiUSREKC77AlxdNwtWAPJZbFT+LPCtlKVmdeu3IACDQFTJ75uWBcB0M7NwRCXgzXyRwkZsmJmQumTgRnq6hIxlNb1632ghIV9pmU8qD+D7KMuBpn3jR1sIaQRAAJk6ez1HEEC7JBasLIG37Huo3KguUX0LUjEOLZ3dan+hrZ+8GYEr1k2z4iCHeVy/yN8537rlI6bnBV5vfetb8Ud/9Ef40Ic+hHvuuceunzlzBgBGM6inn37aZmNnzpzBcrnEhQsXDnymTWtrazh+/Hj1V5LOoMr3StD5WGluR/E4IXrYD6EbhgFDv8Kw6sESKYNl5lXiFw5Q77PS5wlIvunU8jXbQ+BywgGRBNjeVX2vo3pTtT8lwFRkP1KL0ZJHiTQQYvg1baaJCB3NQ34NjJN33IHTd5/FpYs72Hv2Mew+9wTOXyPc/dL7MZsv8Ln/+Wl8yzd9E9YWM3zpsS+CunXZRrDEmbMvQQKw2tuHbqo0MliYIcIBQ0PqMaabOgC0qObKpRLnIFCOAHnAI5rP2JH4Onk2fCkppbLeNZ/P0M0JqWNk7tEPvWxpICzSDJtrCxzbXMPm1gbWNrYxWzsGYIblirFa5essuUiPcoZ6MJoSI9I6E6EnN7/NZjPMZ3MBsC5mBhNCAliZh3JagtE4gHuge4n9EKngsyqbaU03QJjbQUq/p074pBPvwiQmxY5AuunbAFey01m4hW4hcPK6sIKo5IPO0LSaoOiesvJbUcYFDSFc8iyatgXnkzgS6SDOatfw67YdwGIVMePb0+5QesFBLGbnGwlUPZlIcW3LLjFak6E++jWdeTEzHnroIbzvfe/DBz/4Qdx///3V/fvvvx9nzpzBww8/bNeWyyU+8pGP4LWvfS0A4DWveQ3m83n1zJNPPonPfvaz9swt1CgQplwx5QY6MMg0hcT2lmuKLBuMkc3UkiWYbjnlWGdgQ5nq2vJYRmKS/SdyrhhpOJriepyoE21QvlMK0Qk0iI/Mhqg4R+iajIGvalthIOjsw5nRBUarrRHKYFMhY1ereHTFIJBI95BB2h1mFzIa02KBk6fP4LZTd+DLj/4P0Gwd862TuHrtMh555H/h8uWLePk3vQyLzW3MuoSOl9LOAR16O8wPBImmLe0VTbD4dGbrpWxUUWHI4a4EQ2LfFlEJkUA4sn1wZA9Wem+Mvi18QXqdfV1QbloHkdKc1VlDZwcGFUJRss5hlMvUkc92KKEf5DiqzOhQ9nLdtn0Md5w8ic3jt2G+tgZKCfuZsBQe7XIOc1jdA6f9rQxPYCYLBWTOR6xWAz8Pr+yfKkphl8mjkCkFiGEHfsoszN27G+Eq+VBFU6AcvOhjs+xzk++5RJ9Psk+ydI8KxcKDLLMuTqkCozB9QkTITAKgOturlEgZWF0yp5X6v4SUZkhpXj7RIXE5vFbNslqognEJnab/udLEAgxhzlmAMOn1EjGl7EwTQRMU7tZsbn1DKObgZsdcPUmO0CWnCQSzMmGAKoDK3cotBAc8NwlLvVpwCjNCDbRMuvwiJyfo+mt1APDzSLfkbfiWt7wFv//7v48//MM/xPb2ts2wTpw4gY2NDRAR3va2t+EXfuEX8MpXvhKvfOUr8Qu/8AvY3NzEv/pX/8qe/cmf/Em84x3vwO23345Tp07hZ37mZ/CqV73KvA9vOnH8wuMBxGwahPv8lFM/NcQQA348NhfCGqHVJV4GOme1+soGT/lXI5OzMAwT2eZZBxkOWrszood28gap2YW0OZPqW2y/iNOo9sG/G1lEiJuwpri3KgAiOxMXJpQ8OsJiYxObx05gbbGO557+EtY3bwMttrB/6Vl85SuPY2Ntjntfem8ZOLnsMRvygFkizDsg90OlaFgkDwZ0DcqPeCAbuB5omKxfAfemq8wcKpBMu1OgoYqObXDmwCLhk8L7Xi1vAEyx8FpPOXBw9Um2Mbh4HSIlsXSV+ykRFmtzHNvaxPGtTWxsbYApY5mXGDKwGspR8LZaSipgtI0qhBrmEfOsOVpoZwiT5CCIOsjp39aqMLsCwrpxUAMoED70E3vx9n5iJbuvz2pMvZK/eEdy7WmoLv62PmP70pp1TC5AontBEwhevehFqkMmjEm5ntAhpVmZqRFAg1pbirILFeLGELp+RtCNyWUtkAygFNidN5VlHXRChxmNFURQven05Mlryt3O1GU8Sz+qQsFaT8ve+gPwEiMITiX2QWfyS6Rs8I9K4d/nn24JvH79138dAPD617++uv6e97wHb37zmwEA/+7f/Tvs7u7i3/7bf4sLFy7gO7/zO/Hf//t/x/b2tj3/q7/6q5jNZvixH/sx7O7u4vu///vxm7/5m2UQ31ISpM9OeCYqdm4xtSmTEoCMwQjLLLqGLSAKcEmYIt3LBdWwxH3X7cTCqEHhIw4nTchj2qIqmrTX/uCWUWTZmMYoVonXIDQnF6WNOdksmXZOldWJi3C1GW15qaMOZ+55KXI/4Mr585inPVzZ2cVymGG+2MD+7g5uO3EnXvbKv4fPP/JpMM2RZhvY2zmPM3ecxvbWFnYuX0FHM5ToIXU7EtwmP24j14SWFhldgwnHQxs1zikH0M/Qx5KqOxkU4L1NxR2/OinswGdjzlrPlIqn33xeTHUZxfWdUsJibQ2bGxvY3NjA1uYGbju2jWt9j90hAblDv8xYrQYMhBDjUt3nnX+0xYX/Aa5QgMEDY+h7oJuBMtAz0HPGkAcQ1Jwc1tWUHzJbGBvXGSMQli/m3OGalQn7DOfvSLngGoBGkpt5PNKTiTCjhMwZK9a103pdM+YdaaL/qDcwCOBOdwYmYFack5IAT79aIfcr5FVRyMr6YnnXFM6mzqkS+qEtQMOVcZ23VjoqnUr14EADBFq2bR6VRmT5xLIVxA464fggedXqcje6z4GvXki6JfC6mU1lRIR3vvOdeOc733ngM+vr63j3u9+Nd7/73bdS/FSNYF1qA6IMCdUSygIwMIDKTn4ZTFk0jTK1zsWtVubZOvMqh/XBhH3FQtQZYGpcZhY7/CDvJdFwqgjNPmLqZgAYdT9Ni0OdQTBg7r0J0272EZgyyvpZLDNqnwj56loek6xMEYHmHV7y0lfg6ccew+XnzmO+cRtm1xKeu3AR/+OvPouX3nsWJ07ehucuXcBsvi40HrC1vgDljNX+Mqz/s0RK8CGr7SKmcPRUgkP/BDGqiz7gY6DmCVyq7qjcNW6qQN8oN1FevNSAYxwqE48nlC0C8/kM89kCs27urUhAmiUsNuaYrXVIc8JsnpAGAmVgtRqwHHr0suCdpc6lHQ0PBZazoNBQ4VWkYJ8zMAxI1MnYKKbaMl688kWBCwqEjgvqKiHsAngsL3iKo810hcJnQcGysW3mLLFuMIO4OGcUg4mMWSAccokqXykq1BsmyG0DMYoMKbO0hMVsDfPFBjrZj7fc30O/v4dVLns7lVWyvR/X4cc8ZtelWbU4EHnFGZNMo0+RjJVq9iZlt88e+EtmkjLGqX2q0UMa97aQ1/TV6t4kbLT53Xo61FHlwQqoxZxkZiee0AKI5RDQMhAou26Wcw4eSjCAqz6lwAJSRcBqp5c/csCIOhADahtkXN/rktp/bQy4cI66mMISBebXAVleC2W1X7W68sO8wdhqKk+XWexsPsdiYwPzxQZyv4/ltSugtXX0/YDl/h6uXH4Ot5/8FqxtbmBn5zJAM4DLdoLtrU0kIgyr3iPCU6gDAN2LautSQciphm4mpwjAadzXXL2t7dGbDD+wkezj4F4B1ExsWrvIdJIFhykFY4Ls1W918y6RNmZlrxdITikAKBFmixm6RYduXhrJmZGHAf3QY8jcbi00vtMx4LdqBQ+sbVIh6oBQQY+9wvY5JXGYMnxj+3R7dXP0lCeyP+wzAlNwbBxyaIKMXLnHWcLEifm/OgMs0KAyfAblxmej5OZDLhu2u26G2bzsf0tdknXxAatlJ1tlYO8qkCSCzy2mCMJ6gxAVN6M8kZj3hW9HzBU5PsgcJ7aUPTX3lLqGiqlCM5kojJ5xqAw0o2vczInrB/261XS4wQuAD7ayQMginNTczhJ5gQMQETMweFxEZpnBqGsU4AOmSeYskZKpccqGRWsqgTCLmUsGrCx8MxKqI4IrezLMiSMohRVwJdQMJ5fDlaDpUT14al0wrt80ezoAYdIEUJa4c4TN7W0cO30Sly5fQYcBG3PC48/uYrlcYrXax+3HN3H77XdglZd45oknsLZ+B7jvkYYlzrz0fly+dBG7165hPp+Dkre7mC5ln5PUsGwYlRiNirTBi6XQkmBeW5EQcR3C0bkmEbevRBU49nmktx71QZ51JBkXQJp4dTIlObNr1nVYzDvMuw4dJTFVZyAx5msd5htzdOtz9Oix7Pewv7+HfrmUINMaCYOCIqfFs9WDtYKAO10qraGxNYEyf1BCMQaHQsQZBdsjAighqg3Zug4ZsMdxxOx145CXedM2Y8JmXXaQq+91SpyBLKEKhlWpapKZSyxTZIMxQAM2ZV2sBCYuY6Er1gxKmKc55os1dIsFaJaQiTDwAOxfQ14hqAtaFqAzURfqXK3ZkZgbHaRQvx86ztamTNErsqW87mOi4ttAy7FqFbw9YV8whLoCsnVGOshM0DGbRjZy8rqQLLuUmoaG15rWpCJ0K+lQg1fOZdE6JQKyCGEumiAAWftQwciyJwvSqYBRj3XPQ+lsNcY0og/a3dofTLqHq+6UEnTXBy6LcCYe9JHQkWz5WmxoZVQpfHSkQygr5qDx5qZS9EhrM3E+CtKdZEMjz5FnPU6evgNnX/oKfOF//c8S8mpxDLP5NfzV5z6HzD3+j1e/GucvPg3OjK2141jlfawtgI3ZOi5cuIihXxXArxKrduFtbdo37oO2zm3ioNlO6X7lfW0vVVJF37vV1KDhgY8poKiXadnntba2wNraAvP5DHkYsFr2WO4viymsI2Bthtwl7FPGPvpyXtyQZcYBdMrGkzWIIlSKVxIlyGnOxduxePxF70E4QzY6PKMy5uqEwa2Joc01p49rNqW369jJQPFAFEUlExuP68kIRVBGPU5nHiTni7XwQqFO6ubi1wbSPZFUwlXN5kiLBdKsQ8dA1y8x6+boaX9SwR1TSkvPaN0UAry5o5LSUMd3Nc5rgsXSD+b2qOzSDdkUYIsgHjF/8rmJsgoo1088n1F1o3SowUu9ojLrwjoqcwmL7R4o7Ep2vhbrWnO5x/6iRoNvk3VKZM3KW82V0XiBSNxqIYCWx2KXYibmoaSutHrrBt0/kvIOiqEUyd7gzu5FLyGNrA0AiTqsHVvHfG29qAHLXSyxwLInXLpyBZkHzOYd5hsb2Nu5As5ARzNg2Md8fYH1zQWW15Y2yDMYnbqOV9UVjTLMGGwNLKC000olpWrTbUtFCBM3A3ys6Zp3m2sLB49XeL+3aXqjfdPf7Jo3ESHNyv6q2bxDN5+Bh4y+H7C/u4/d3T3sb65juTYH8QzLYcB+32N/uUI/ZAwy86kBIdLA5Y47DLGtIlqEjaSnKCdT7EqthSrU6ANoBCrfQFfAFMfrhgdXIymOvSA4Yw/atxR4dkLCjqrB7vI9qkrIXI5IMxYrIE92/ppthKYU+rpG+9ZpyxtYt8FoOsrBnyKihnbwylXtax8qz0S9TAExrpnX0kWuGD2n1UO7LkNQeTC6YcR6R+CM9Tk4/5tLhxu8hK05ZwvgWTRC5QjGIIyYuLjKmiswXAlhLifPMrPLwzGbG4Bpx05Gcw/rCoBhkUW40EeGRuH3NilT1660Bw1Od8iY8Euk5lm0A0QYmUv9hlB/FSuJCLedvgvdbIbdq5exSITLez0uXV3i8XNfwdb2BjY2N7AvHp6gsnO+65dYW6xjY3sTu1fPQ3fEcR7AqXgbkgT99ZaRR4HQNnNotQWfi2LIB1Ikoo/v6Oqt9HehjJAHt1Q0+gXwnJB9pNgpTBUBrFp1CGPVQzPNMFvMMV/MMZvPsJ97LJcDdq5ew+ULl7E57zBLhH5tgd29FXb3Vri6u4f93Jcz6RDb6mAi1CwkpGLKTsaTLBEuAEKJbM+pA1MqIbvgxjAKNKASlTnMuBpQJnLlL2jfpT/I2u0VjV9EwWNXYWwEBp6UWsm4hTylzwOUdX+eDeRJBKn8aySfcgI1yT6rYiZlmfUl2a8ZYz0mUnlCVgfl59QWrbqWVLwS7BGDlUaqPLbVn0yR4Q/WIOwR69jA1xDFPGj/XkcdpMFRn1zm6NhNUncSk2g0LtlWl6rWL2bw4gHgHqAkS0lCSqFuZiCnQQYcBQNiIVpnxK3FVjSxFVaIQFVSvGtJTe3Gkc4FhMK8Q0oBQEuN3Z1eZoVigyeQGxmo5t3Ael6DFpnkotaV44oGsQ0+oIlly4w55hgoY9WtcN/9r8LVZ57Aub99BEteYJVm2Ms9vvToo/jef/yPsLm1gc9/6Qs4cew2dCkBuceJrTl4fx8XnnpOwJtlE3cSpUEpC9Pc1GszK12FboO1yV0nVHiUvTwu/GIvmRANRCHNNGxIrZUPpyFFE4+pyc2zpN6kimAHJSo+7XChC5R4fLP5GrrZokSXp4S91YCLl6/h6cfPI+8tsXPpCrbWF7h0eQ/PXtzB089eRL8/IDFh0aWwxaAtUYR+9CzSXk4lOPAiJcxmc/SLOYYuga/2RVgbD0b1iYLJjmvTUGBG5TLf0ajP6Z4id8JXsa19dyAFuShHHbO839m+ZAbKdEnbG/qhBCOOBk6qSgYDxLnMPhlmussAODOWe7vY2t0FKGG50WG56tGvVsCwKnvymKFw3oW8NeCwOeGr4pDYHURYa+ZApVFnOJWNxMqBdQAuJ7eCWxhNsamtOuY3oGMO8LVJIzYCqvvzrMsmOkkoIczUutFJDbI5fzWb1ONIbKMpPI90uMELKgj0uIwyjFmCTyoDaooMkANjMDTmH4Vrxl/QaTag8olqLTwKvLABUCRxGWCsmr57ROojmofljcCgel2EtLrnR9isddeaVX11iOLF8h6Fn6IZxo2o87U5NrbWwX2JWk4dYRjmuHDhOVx47jncffddYDBWfY/1xUYJ1QMGMGA+X8OQWTbShv6yRun+l6gICIubRGoE8gicaQTqersajJVjh3dcDf7Sp9US3JielmW45dWStybNh03dhOAa47DrOiTZ5zgMGUswLl3dRZolrIYel+cz7OwscenqLnb2luhzDdbWhxCQiSp/1UrhHyrhoYauHNxIKaFLhAGuaLWOPTq9JACUy+pwHfInm/NNjWiw90aUIAdIQk2fdnhZf5CApzk8cHiIRRkNyuMon4k+UXkbiMkABu6x3N8rpxETMKz2S6gzzsWZqHEMGeNHzbQc/ry13kctjWLMDCAGFYj5SDnt+CAvM9ZLOSFVD6OpewuU9SyLzYM6qvnyhpXvSui4kOsoKjeZDjd4cQnyWJC/IEWZ1RStsTAii3bGgUVQwkDpd5TB22oIIHggWwrLuoRayab4TuipKECoZtcUXvXXg6uu3tMxDn3J61YDmAvMkTtyeKdN7mLL4VrRmxdrW9g+dRqrvZ2ytrW2wLCb8Oz5p/Hcc+fxspe/DEMesLu3h82t42ISWSFRxix1RZjxUKgmZQzQYElcYslV9QzfAzrU8qBpeHg5mYCtHTqiQlKTYXyl9IGW61qjKQ5AEZgm4II5UgRIS38afbKVoxE2ulmHriuCf8iMZR5w+do+kAj7fQ/qCHt7PXau7ePa/gq9nP8VZbbWxLdjqemOkHKol4aJSgW8hiQR2pFsoy1B5wBKUt0sUK4lFm9Eed7c07OvWyXZOBzlVwRZpUfwT6yecbNh6K4qhJNQXceHzAZVAWvP8qq0jarLyXeHEmQ9rTw35Iy9/T3MWKLF9CvkfokBGYPIltY5YQxXsZ+8YDMjNtUpvwu929C1EQ8OAuMKj8b4CB0hcQ+oeZJGpUPeNSWTvL/Ln2rgbsat/QVKBhzKVhla9/jzS4cavIYhY6Ch2PQ5ajViXhLGIjAoMZLunWCWuGkyABKLQEpImElHwR1zTCg68afkp/aqriNB6mK9nYEkrJAx1O/hwJ8VUNLUAzeVaqMCJW1PdrOSlJVEc93aPoF7X/6tuPL4F7Hqe+T5Ji4/cwk7Vy+hX17FnXffhSeePIflqsfm1jb65R42Fgmnjp3A/rU9F2gElHmvGkJ1SJohqQIbH5hB5dBBReGhli6jFgdBAZ9xhAwPeNMfUaUgQqkJ8dDHOttiHBSTQ1vl5RPYjj2ZpQ5d6spWC1mvefbaHvaZsbEcgNRhf3+F3b0lruwt0XMxz+j6bV1mLanUoSK6oycmEDoQzaCHLxJnEywE2dCuNCO1TrCFlkqZitMIEPpP2ygglyjsu6L6XljRiWAWx1ulALZPVlMKBomrY+QlpbTeMEVGhyVRCPybAjiKIM+M/eUulv0+aHWt7MPrV8h9j9Qc69FgRHVdWdfNruE+4aasaNGTs92L7g9NfKf6QmOHmHhhNE2zq8YbaMdXDYb6rnZR5IBe3vyahof63y3lISOnguwDD8VcSJB9CcV00qFo8ZTZTQxg0RClK3QBX/ItA901FE8yqBiog9tI0oET8Yo83ArJIjaLQIiaoYWW0ee8RBtMVgWEarC/U4kvZaTAjypDUrLlWWtBqY+sNqWEte1tLDaPIeWENEvA0GG1n/HXf/1ZHFtfx8mTL8X5C8+i62ZYTzNkZMxoKAyVZfbLAzIGzNFVGr1pxzJbKLPjwPhxkMJnElH0KfMfNOZViQF084PTZxK2TK7XIHPQRu+p30x8wwFp3SHtSXocSEJx0pGz4jIyljxgZ3eJ5Sqjm8+xv7/C/nKFVZ+dPmxLaXUhMgZIXOfICi8zLg0qbdH2YgxPNbvZTB4SCBcWvJmR0ROKu352x5Ek448I6FBmXkTqROUaZifOH+YiHmYvujvJhCW5d2YtTWWtTyLUcq6dN7JmSG2f+qZ4dbBJ5EeuUCLLCwQ/kNZin5ZAswoFWiUtxuQG1fzpsxKubxBgp1ZHdiPYOPXnJ4CF2/KcTuQP2Est4LaJ2+vk72jOdlJ1Uxnd15XjckEjQkNPTJR+8+lQg9eQi9nQZ16+oK+rWwQSF9wSBd44TJw0KmBQIEix+1sSk/wfjhgJjNUML/doBJlCyWDkpG7z9cM6WLXcGLPPBgmK51iM1qOPRI6eWnexS1qWqIVRYDCAzRO3YX1zE+h7oOswDAOu7ezjqSefwO1//5W4/fTt+MqzT2F74wRmXYclr7DWETpi5H5ws638q/Z1c6u2FKzmVpeJpExPsE3MhHqgRet7PUCboRf2W+nzVD3d2uMFBHWdpZld2M86o+lGhGHLKLOh0rVF3Op5cUDGKmdwv8JqNaAbGMvlCstl7xFhtPBg5tTMVeBrQZE1mAiUBWxMkciyPsy2zuLbeJMHEqYyIyv7g2WbSnSQoLLfkojQyZqUmb5C1IgUaJ8DGVthWGRky8vxu4/2yHE622mkcKinBBOQmZfWG4ks1JrSTw+p1finAIc1a/KCqrLYm9wmdtp6hTnwkEsedZIwC07NugbODnAGnWiZkZtrUXmOG9BHJGtnVF47k7OWJ008x141VZujH8HzTYcavHLOFt+taHDKRG4pHoRREpVd5DqUiuDOMva7AmYS9Z0yBbBg1LHXa1CrwEqwMZl8VNgKruii8hGKBkzVxjDUsy4DLm9zLC96GsV7k5HRAq/UZegRHiVH5owejJe87JuwtphjtXMJQ17HM8+ex5e/+GWc2FrD+vo60myBfrXCMO+L22Ze4bbj68CQsbe7LBpAAjrSo0R0ba3eLq3adLS12xgN9CxAzdaQMlg5jFeNBTImFMu7NBq+Eym8V2PsCHVhAiuswbTKy3WLEuFpfJYH9MtrWK1WcqYXsF/0LPDOrh0sadXkMuNAzjLbcdq2G3ejIgFAtiQMAFZgHlAiY7DNMMxNgIByDIpurNaZIiFlqX9K0PidSfrCtgIgoYOHbJJY1xKNpAwUAwnhA1tXC3SqD1wsfeHhq8Rli0LQXt8vID/D+BUk15idSUGsUi5Ki0vAg8KRXS5R9vWENJbyTDGLYltl0SjV1yi49Kso0NZPcJvlO8ljZo45oNjAvzXOsoBym1/9jn5TikjoBcurYPAI/qp3CcDcIOygyt5cOtzgZcdJi2ARs2HxOozal54RBT8sNIp9TjUClIuIEr/wuzBkcL4oTxaAK2ZLHmmK+q0D2TEKSX4DPouqgFFnRimyiICqSO/JsC1w99cqxdmcvSKaEzOyDO7ZfI6NzQ0kHpB7YJV77Oyu8ORT5/DFL38er/ymV6DPS5w//ySObx6XtbOM9cToV6uipSaGhZbiZtyYaVf3HQXCRw4PXaA1FZgr/5Ip8uZ6XIRJpQsXkzFUFJDlaQONlEW4qsa07hqTrOFx5+/w+IV6zxeNrnVdktOSO6wtFthYW6AjoB8GYAAyCYywoDjEhZ0h/RaNp/XaYQHEQheHPD1HiQEewEMPLPcBcRUf+hU0vJKFxBJ358QyU6NSr8KOBB1+WrqdIh6OKulYIuZL/cxsyCUSfonuIaqiAY4DF4MNLHy24c4A1vIwXtu+IyavJ6GAbgQZUwCkp5Jst2E5VYLCDjQuR8aUYktJGuGnvDxytRjNuKbqCG0ePPZ6K+LjthYRTOHFMM4V2FM99qegT70zK09plqgrqX1W6+GArc12ixegDjRV+DXp4Ew0atetpkMNXo708pvYNgzaNS4EpcxQL964FqyiriakmhQLKF0vmK5CFxjebcS1pogISjCu8z1cbErOVFHm0ho6nJ0TK4Fs1Z4ovDZBBiKY7ZrRzWbYvu0UWCI4ZAaefe4Crly+hNVyF5vHtnDhygXs7e/jxPYx9LlH4oz1WSohi3QBP4a6oUjrSLUaTIocIFen5f3qjChtpzQskqHQhcXUQirjnOaTYF8jTkt+u2vYynV9K93zenzitFdaqO7ZdR4mamNjHSkl9H0PDH6elkaAN94s+0AwsO/dcx6EMj6QCi+W04yFEDlDD3IB52IaTuXK0PfgPJTrnMuBj4C7psN8bquW+SiqTecmNhVbqHZ4UicL75gDnKGklToz0yDb9qoSgRkaxMYjRUzlhGai4vUuSk94lmRdjpzO7fCrc5K1Tw58KZ/dxAxeNyV739X1NO6K46Jai2UbZ9A9h9rGydZL3qz77tjXOfW94MSl7XXhqXIOiPAVJ122zNjMmMscg8wv4YWkQw1eAILmVTptgJ5EqgFoXGJlEpMelEH1yQ4anYMT214XPfPUWNX4TrtSdaTySzMuZbMFldWoHeprx9BDNsj2nk12Y7NQ7QNhaj2LoOdYJYpMW7Zn+2J3M5gVuBIj5x7z9QXO3PtyLHd3hPc6/NVf/zX2r17C6ZPb2NnbxZAB6uZAN0Na7WGeMo6tr2Nnd1kiNCT4PiGbrercyQWQSE/pCwUbhzmdDYLFjdueE4BkUQAykMXhQc1JSShWuj7BhjQ5rdx8GgVKnL0ctHYQV8Xi6lDsDSXvwVsYNLdulrC2voZj28dwYu8klvv7WK16IA/FVMi6xiIChiGLvLlEOAfZ7yyRaksgaplBcTHb9RDzbJbzuoiQiIHVsuz7Ysaw3AP3S/Ag3rBU8i5W4GztLvtzGcQZCbmKhm/yKpf1LJs9OcLaLEwFcs9F8VR5R1Rc05mTz7xsxigfFVh7Fw5Cn8JxUmciFOqIpxu5MwuIQDljRiUYtfIZRI/KKEsKaqYklNmiyhF1n0lcZmRD4J7WC7qwPMv0knQIlPrxIE5m7MsJlZCXXMTsoBYMvVeOsXEZo4xIqgCZeqHjR64ofbUqys82Y4oST9GJRnPLkqcvZ7iZvijBpi9b+JEXOXgBrrQ5MzX3ELVDlV/SkWp+MAYCOhVw1tmSR0Nr02wm1fVmBmE35Jg76XxfB2oMh6GwARzMMo0QjGapwOOudUk4m4oa+moCZCc/hgEzPRQx9wAGLPczLu+s8JUv/Q1O334Cd529C0898wyOnziJE9ubQO6xmBV67e/3IqjbJteapi3xhBpGSLWIAxKdPC6MVzZ2MQFp3xv0uCzz2amXbkpNW0XmsmF4lOFkIlB8IAiaUZEH5gCoCXixWGBjYwPbx48hDz2WyxX6fkDOvXv/ZRc6OWeJqM7IeUAvUpY4Y4Ae71NMj+rVycOAQfg8Z0bPgyla1HXlDK/cY4GEfUoYekbOKJuXZwlri4S1RSdrpMAqJfRdwpDFdM9uLk8pGWDolgww0A8wYTno7iFmzMDo5Ei9rLMBUXyymD1LveHKFkIkD4aYOsmEuTldGRPIdWlzypDTDQRsEwl9hF/FXKBnmBX1KSMLX3YABgwWBaMASZElGjHHQCLwue4urfhEZEXS9SzFDPLtF5oDBwUiKlwKoWp5AKwJ5YDSqkaefCdjq4zXvBrhs03xHlUPat4+MNRqcGCw8VtI3xDgBQShJvSqQAcoA0k1P6CxnCkTy1t6rINNnf2d2FWRdbxT2UxfEb+q756Z/0bcYQN/CVp+GVRUvxTeDbSgsKenui4aIlEAQBkwIKxtHMNifQP9/jWAGTs7O/jK489gMU+YzZIFc02pQ6IOedjDbFZOsR2yQwKpJhHrKmav2tsoAJhuNA/vcXg30ia2LG7yjB54bPTT/FG/a8pfoF2Ns16GCJPirCp5BmWiebqqoc8+D0qErpthsZhjY2MDw6p4Fw7DgJwHMA8ShqgILp1FpZyL+Y8ZvYAAsXj/icmwH/Tk8KKccC7vZmb0emwIM0pMwx55SFhxjzkx+q5HPzC61KHrOiwWHdbWFrYRfJhl9MOAPg8yweMA4mQOG0oKBjD07mI0sBx6I6DXC3ANuR5XWa4zM/JQ1uEgszgdzxlllm7nkkmxGmtPAYohHpBEwTohNGCSkFfKBDIHYXcCSSJDtFllliHrxo53JU5qMHXXHKGyR2hkypbMd8Kyh/Gc/5oYW3KdCj0pUk+BnOu528jMrdaMKWSyHmtUNkbgaxavUn2KgwnTszXv7hBm6oWkbwjwikY3muoDIlmkRuja2MlF3HG4N9IMrJ/DkqRqSEHbr1/R2QGjGhNotBQtQBm96VSv1xSIop59h0Gp5U8lDp+Moi0fO3Ea65vr2L96EYv1E3juuQv4X5/7DM7ccQqYdbi8cw1bW5uYdTPwwMCwwmJjjo4Y+3srqVtkV1O3DOEpydpJUy/d+5VJzIVcz0QLcJrLg9WeSSnt4OVrbB5L0BxgrEARb1xlF76LNmprNEF/jp1nAEleVlUOwjpgVEz0gzCbdVgsFjh27Fg5tLMvgJVzcYsvsy8qnxDwGmRNihkrORqEGODcm7QZBg+2SZyBVdmvNIDFTAfZk5elrBVWHWNvlrBa9VgOWQ7L7DCfz7G+sV5mp0SgTOizgJcCF0NMfKKeBPAClfU0TT0rjYvJceCMnIGBWQ6KFcUiO2APAwAeyniS2R5YnFpymZEOQ7Y9Ux4+qlCaAcwoWddHR5R43h97BF0DQUC2AagrMels2P+yzNzT4EsWPhxUcy3bDpJE9Ckm1chz6rEZRlHkTx/gYc1OTKHqoUgCM2Zi1xgZ2uxaUpmip2M0Mmm07ATGJaOolh5mdVGJbGVV/DWyTN1aOtTgpYMEcFOQ8aX8mcEs6Z6EskEzqUqjlFSPbtHk1GOvqzlnqhJWrv68fpeQPUVQzyoyjXHKI60No+Ql1VfjL13zCfpl0epa16EkrJ0z1jfWsFjMsbp0EU8+t8Qz557C7uVn8NKXfDMuXr2KCxcv4b77XwEaGOAVTt12DHv717Acei+fMQLfUoEiqDwM8XR7tP+sP4WY2YahFQJEz7DQzkavFJkU1p4w7qOoUVaNASpNvNZdXUjUaDzumwNbS0DXzbC2Rui6DmvzhZngBtkcm2VGk22zrHgMCvD0OZvmzIOCFwN5FeQmgwW8gFxiI+pMIw/IwwpDv8JyY1GOul/16PvBwlfN53N0axKNQ/qlAOtQBL4G4OOMFRT06ygKnH1zdWZxuuDShoHd9AipPnMJPpCzAJzsOtZ9V06njDwwhmHAahiwJzPOIo87E5IavzERIXUi2OVEz4Fkzdhc+chMldEsGgGryyUKP+cyu8mUMDDAeZClBwUx75IkNFFPSlNmCeUw3JZ1qt8qG2DgFnabKje6ciUPZoT1RgDEGqI85huUXda7CuKulKFrxi95NRX4psZXTCwusCNZdIvpUIOXJ+9tjUxervpE1xQ9/RX9zClqUPpfmL4HZlIt3oVl9DaDa0wcSw66DLkWH2cBalasZpGhDAQNxsW8i2d3gKh5v9KimtkDa11SwvrGBojKYFgsNvHYI4/iyqWLuOPUSVy9dg19zlhfXy8U5iWIGGuLDeztl3WJNOUnbldK++qoJYTK5qftkGvqrSUjSQaVO1Oohul7+jg4c2j+8l0HWJhNKV2b1auJFNCMQz/IAPdzi3gyJ511T5lxjUZUeKrrOqSF72lSJ4Uy0wBYZh1qIgSzOXRAyqEha2PBvB+kSJnRIWcQspkNKTPAA/LQY+hX2F/MsNqflz18Q4/UlbBVi/kcmHXQtWLKOvMr5ku2AcRYhl5IGjhCZjSmXGTyLQAs7uixrZDN0gJcOZe1LehMLKsjiry7GjDkjGXO2BgGo2FC8u8pmfu+nZpcsBy98o9aXMR5ZBgyhkw2+1PTZAZjGOSEiFxmwj2K48kQZHK251GeIzfbK+9kbYcOhwAIbN8grKyDSPlNwV6lgivlJh2qDNVpxPNusLGWZ/FWVOhGCFt/t/FZl+CPEESRev7pUIMXA8FkVsBGw/KDXAAWeFEtgm2Gpfs+Iu1NQdGQMyiLtaokUVWavsO+OVJfr6FHGEpEd9CCHD9lFsGROSrWDYmA4IRhWcMFv6+gUXiHRgMjg5C6GTaPnyxUyhnzxQYef+xL6Ai45+478YWnnsZ8fQ3Hjm0hMyFxBlHGrEumocb1K4tEIXXSLd5uRQqQ0fB1q7kVGc2yobR+3I+4cUrZgn3TGzq4Td+MQESRTv4WtFmWt+asA5TcHBwLDYzidZtIoUgFL5vZiOasAIasexVFFMhuX3VwsBiHOXBlntXgNRQXeF0bg+RLXNbXhmGFhYBXXq0wDKtyenDqsJjNwV2SaDYMGjTosjtsAAAyYcV+qHwaFGwB7oK2n5NtAyBOAswSbDgX7TyT5C2zm1zQBJxzia5TWokhM9KyR8+MJZe1OG0fIRmwpk5MhkQlnKOuv2SgzyVKPBPQsfgpMjDkAcNQPDEzCgplBgaU2Z6CF/UZPRdz7BIK7vVWh8zikahjMSVZr2Npj3sXM7sibXzLDKKZrV3XmGQanih7rgQ5iJhYs1xhJTifE2Hk9IFw37IMCpn9VKUOgHrNEEJdrCydDT7/dMjBy2cfIbaA3HX1g7gwZEfecWYa1E5nB70SNLc4ncZ5WJUvaiGLcBdABYpBWYdVSDmmul5aFRC0ybmRklETIoUEQj0jgZyvU7iWJBJFpOJitsA99/09nH/i83juufO4dG2FDit0szn2GdhfLrF5fBvHT5xEXi6xtZhhQYzzTz+DQS3tA9y7T+qly7UEso2rbSI746rV3tTMq81TmunCMJp+0fLCjFfXdcTUwdYvbCSsvLdAYYE81kR/xEbE/mgb5Z9RAz340SSaKJuXnCoDpaqFR1Ux0zPfdH9gJwJyAFvwXEYGd2t2xhRlBs98vShLAC9zUhC3+62NNfTLJYahx9D3ZcacEmbdrOwZ04oPBbhUhbNgAUgA68yHxbTptDRxnJN59yUmQMyPQJhVIYvZTpWYDORBQFvDU4mJeX+JPg/Y5wG5zxaP0MALAAn4AmKySqlst8goJ1gTIxOX38ImfR6qGRdkFtZzBoZsYekGHopDSQZWSMjoLYgChjIONOCy0SKlsv+OGeiLi71uQWCjp0TvB2Q/VoL4kIpCV57L4ZQBzhkdJQnzxaZksWwp0WqoORpKQ+VJRr3PTflUZFUXeF9HqimdcZNpMKfruFNXen33haRDDV4uxKOBMNxTW1wAEpbBnSXOoa6hlE4I0pfLgPdF5yDg5B0tRpkG/nYQu6Gadp3NVyCWWlKCzvjiu5NNd2lc6qs/RdgpSLswZJCuOXER3ImKNWh9bYbZfA37yx5f/uLfYHt7G6sh45kLz+LEiePYWN8EIaGjHnrWrMZLKOUmy9foIoDJtqBIjfmMIhkrPVBdiln6zBeFs4MB13tyDCzCwPKZGMXiQmJ4VH0ODj9O/9i30x3R5ujvMHmbAFTrga4th5wY4u6sOowIzSj0RHgrzXQGAjW76Xd1BpD3rV6MSkv2CpTTgrvZXExsZX9gohK4lm1RpohOO0SIS0xb3dPURfrDwcuVgNIfvq6pFge2mba1gRW8WLYB+D42BY7MjLw2x5AHrA1lrY7F0YWiFCYqrvuk6y0aTV9nPqKoah0YthYHARSI0B8kFqTWtxfbIHOJudqjzP4yMwShpBuyASMrwDCAIQfw0n5ntWoKndT7UrbZSGeWD31fzJggA3iCrttBzMxCtxx5RpVCaWN2nh3Yl0ZKZBwfG1mZ3CRW8GVUViGEvWbOAy906nXIwQuG7gTtTMCAK5LTXNGkE4Q5XdPW+651uNusDFfvz0og1eiodWC/XfdYeTvWz9xjW6BUfGo62bzuvL3VGwJqTNS4SFRiuWhBYv9HLidSr/oBzz37NM7ccRb9/j4uX76Ee+55KdYWawADs47Lup1pbRxAI6QwhdG2hhU3r0Ako9x1vnYzoOdboTRc55sAkSAtqe5lBwoDNbaX3PGqrhxb7bwdbq49SMu4fhrF7GMv24ScaewcBHsASP2PNcKJAgYFQaguLyUvK0PGgbcgIaWu3sRNcTar40l7tNzJKBueI3hlqmfPXaBSNLsXBwZ/TuMgko5RoVOPDDvBWc2Gwid97jEMGYu+x6pflRiQWleGlUzJvf10A5m66isNNWNmmKAvWkK2rQaDhoySCq5s6ZXBA2MF2VeHXDxzhfQF0FSJIFM8qC8xRU1CZAVP2eit5GeGWGLBIOMNQvE61fXAJK7rquQMAmo569YCAS+LDQv05Pyg5Wt/aO8wwrKH1K+IATePmrLSKA4Mw7nnOVrqdLjBq/LycqSovVhafcCTjAHxDFJDYrDFEow5D9QRlCnDz9qXpy05Zl7PGKMObjORGEexffA6HFBpOSFXw2MCZimBuhlyz3jmi5/HpZ097O8vsb62hqu7u9jre6RZh/Vjx9ExkIYVjm2uYbUs2q3asj3uXqi9ToaUaVU8KuNbA2owcOAKCoE9VX4faHbQvJUn6rFT6xAH0s/BP5Zcf9eZgl8voBE8sawKtUVgHCE9ZM3+jNeFw/UIWCrwS0QYBSR9rnyRs9ps1pb9OY7Psnu6SY3LrErN5trGuq41tHtys26z/1Ce15CzKdBYA+Nm1LENAef9GUIbs0TAkTwGKhu28zBgtVrZjNNmDPK767pSVpdsTVzXsxSYeAizO4lbUd4vgKQBknUtEkzoswMhDxkDAUwZGbqVAECm4hkpM6pyGnV5ZzYMZg5UHhgYGDRKCcvMC75dYuCuMkX2kNmhmBBtnQ6yZld8bEpVwGUrhQBwRtkvqHSArDPqbE3HpW12Vh4wQC68piZWaD8LiXRWm3XscKHj7gT/3Gw63OBlEkIEIwHgDiEaIKITPckgiRJAzRoamiZDZhYoqyFJBHRY6vfSVchSgpqwvGYBwnTmQS7gys4LNRnGUVq+ZKsmIeWhuV//tEmOmRsVGB0WikBSoaLtZpy47SQ2trYxDIRnnvw8Ll2+iO3jJ3Hu2QtI8zluv/00ciZ03KPjHkQLMEt4oRDo1KEmMK5qaWEGFAHbNLHYGotLaAbPYCKBTfbAsDW2KIxVC6wJoyJ4rCgU2a2UjmgXCM1FJSkbQdPEAxPJcC6A6S0mm3UdoDpxoIXDv79svEVq3vFnysZVshBl2o/eiwQk0ewjoxklYzOtNxH3xBFqKiXz5ENzlz37+I6gFsH7X/s8paDZl6rKjDFhltxRhdQ8r9JXgwUTmSck4LOLSsFhIJPusUNRE1iBRSSEgL56FepMV6PPM9zVv6yr+8xHwQuZQcPgYM/i/g8E4ITQXlaymJExk3oXyg9aAXbvTNWxDJgVvPS7enayrLkB4iDj7dH1NDtrICj05ZSYLJ6hqMBUB7gCospanVHnIePypQt4vulwgxcV4FDNrA0m6YMarr4FWWKLmfq4CkQWzx+LEF+uVeBEriXZrIhh6wy1uCJ/SNdgAl7FgxgrYU4RhlxkGGZKfeNEwzfoBuAiFyZmhkHRiObzBdbW1rHc2cfetavY390FNjawzAM2ujVsbW0DXNbG5p0E31VvbITDF+tCAyg3Yo5CvQLtbTR4SyIxYIpAQCybxTDMtMqh1EIf+aX0Duba8tvrGJ1szCwa9KMWJG8lGWiGvh5lE8pH9dR0hszNBYMStitTiRG7q1Y9Sr7CXGisBrF+sp58MDnI8h89NAHmujZo65utPhfrUK0TBv6iEsSp2kposZa80Tp/C8YC8/I1j1IDSwcdZg9ZNuicRvZ4FQBQ+lAxK6KYIdWEqSwdPRcLTzG4txUnAy8WeWH5MsMhBBjQhdlQmPWwA1cFXmDx2oxAK96TnM3b0e7r7FP3uQHmZGMKP+uMNcvePb3v49fW9cKsDAz0q3Ca/PNIhxq8EpUwRSXpBLl8t3EiLDElBjS6scaCM7dQEXqD6JEJsgAKoByVTvacAZegic+/KGiGHMeD3o1QVKeogRJMK66FTvXo+GUTZOVKCW3jcy/Layj29uX+PpAYAwZcvXQVXddhfX0Dmxvb6HevYmNzga2NdVy8eEkOLSzSvJyYyhgIIJtvefTsQoIyh61bX/VEI4hz8wn49uRaGHLmAEBU4R2rtCDNqzgk+H3Px3WGVu2giYdg+3Ti75BdlV/cHN2WHXUVrWsRkCaxcOupBX2ufk8/7bxbrUvQxGtywy4Z0Ohaml4n+JO6fkUtKe2tKLxdUWD42pBydVs5r03t/sSIdhePWFsUpazXFeDYaeGBASgED+jM1NiJHYM6bzEHhrKVpDgLsTboLMQVNuN0NQHabmj3MK1sdQBYQiUr5WziZVl59A5z5hCwQsiqmAwd7LJst1CzIaCmwzLO1fvTvUB1TVUOCGYHcsFy61ffXA6sXszgFeWMMaEyZ7kI5uxqczMA43KZCkU9tyezaGNyUGXF+wJm9dyrXtTUyzk8gdH3AwRJIzUoke3Iz3BR1EYSqWNNaP4VkQLtCF3qQGmOfpnx5OcfKfnM57j43HO49757sbGxhcTAiY05KAF7q5UNgFJ+ECKVzC/rjhpU1Bb/bdbDXtcw2LTv1CO0blFsC1U/2f4t16e6m6H9PYINEXcCfuGdaWufik8Vfd0IyNqntX9imn76+aXnE6egLd97BlbpSHuXy67htTQeU30qjwgyJmb9hXipYV9Wh1ZWmsa8rkMFrhmC4F64PKqL13rcDv+sIrdrK+VmZjeb29AIVVAAy8jQ0xdcTSnfZpiVmuksh0oB7rVXVOpitxTOTZ0AHWzdXxUGRjAdKsdLvmXmJc/J7MjPpmTz9mQu6286kcxMyOI9ojMqXc+LszUFX90CUICWsFyuDu6zm0iHGrwAZxiTS6bBOjsUoCIBI/gLrVapJj8KGoUYF9Wkoamsg7mN2u8z3NCOqh7N0n2o+FgPtXpD2+ONtWJCPpa3aK2sbW7KjTEb02yB+XyBlDrs7e2ghziwzGdYW1vHYjYH5QGztQ4DD+hXva9hSaw9EwAc6UNVfxwUXaKCfq4Hd/sQt29EDzilmhbqSBWSPl0LqQZzQ4VQd0t8KPRZDGlUmX/jG8pUrVt8m8irH3PwkElTUOFMzM1NpWv7bDCU+1si9aaULHsvZqh32kLDp/0ihzqnEYd79XAEh3pzPZuuSrIKhv1+1D4XNBxqlKW2tZUXb9v6+BygAXqdZZK1KY3yqH8ZdSSf4thAyBo3VflF2g8BGVbwMnrayboFTAW8ADQWhoRo9iN5z8GLSmBj9Q4NCqqCoVYrixlTQTH3ah4VpYDFOcPkZwAu1r1uZR1xf3+JF5IONXjVwXN9yLuZJjABADvMS98P70WmzRlIScwonG3G0B6s3ooNH4QEFQzK4My1q304r7yqDwkjqwu2AhGgGNYsonsTDN8qECB3X8nCsOoNuFhsYG19AykRVtRj99qAVQa2tk9gNl8Uc9ywAnXryMsV+v2lDJnyfjYvgCg4pQIZxXfG2uWV9ZUyru/LP1FwoXqPiqAigCjSH9Ao3hWI67qkDEy4ulHVqACGFko1we3RcL+GiaobHL/lVxBEBzxZp5B9iZKQUZdn2oMv55AKk1i2zgCcH5WXXICL0wEYQK7nLpVZoqmv8VsN5HUzmpFJalTWkcHV/WKOa83JcdVUpSkZi3luYRMGedVN6MK7UEt1txvlN418EUtHhVBxPdm25wRIKtzlZmx3K/f6U8icJXJPEiclEpljypD+o7xMAuaBfpTMNQqZkux5VItA1RIAZb90WWqhqh6MBD1hQats+ydBYkIiFFOlghuBB3UhKSJNdwWUiUBBuRhaq6y79WCGBfN+vulQg1eddNRTGJCeriMumvuq5QJyRJDsO8lI1AnTuC9hgEnLIZoUai1ah9xY+NWCr34ayWd4NqDkJR0yra1f3zYhRWXK32VgNptjtljDyTvOYO/KJVy7eglL7nBpbwc0W+DM2bOYgTAnYG19jr1Vj9zrURtk6zEGpJCI2yjhb0B+PEJNg7GAUFpHOkx9dziOJjuq1wqu59UnAOVKgYJuAMuGhAe6tYPgQa8mCpJntL5R0E4/F36bqgvEFdSDy2m/S57VLKrmQq0ah5qRaN1giJPClFnQpf9o64nW2VgzWCZEH/B4leViO96svCjwqXb7AVwvqIsPKmM4kaANkVQTos7H1YLYgrqouAygymilRMXctPETiZt/GeJI4Yjl1k7hxajYGRrHGnIY83RA+YSg/DoaazlWO4oqD8EPzwwHw4LkkNICdAnATPmKghQMypLZsRiYz/YnaXOz6VCDl5/FpOaLhpOCXuSir01kfTjKX6fSLO4grPtRoudbEJoMmwG0ZehALcJbh5l+N2nibxhzeh1H51XFNlTtMT20tNuIUFx41zc2cPy223H76Tvw5Jf+BhcvPofdvX2g6zCbLzDrOlAeQEjoZh3y7m6JaB7LZm+VjgMTMwxwXPBma5ALsRHN/VecH0FMDyTRUDQSAEmeulZJXoTlRiHHQpYGMKoZhboblLenjm1pa0sMP2wxlEmAb3BvKzYFNKGt9XUO/8YSfGOpA12TB2t/iAVApyFNPXU2q72p0T1aTFWB462IQOYcmUQzj+V4ycIRHDJnuKOUXg3Dqp7T+4NM0OUe4aXwJPt481iB7Hk4p9b+FCLYra5UYpbWFgAtT7dwVHA1KUeqO+oeHNaDyPpHLwhF2ZVWVUhMZnAIFhCByEryrSbVWMhAnhgHFIImW4gnuVDLNN9sXLWVGRrYOI4pVVaVRzRwOlPZ2vBC0uEGL8DMIPVYYxs1NdmnFbBqTcZxyPfGUFkczmCJjO6gaYPeHy3DqMqSYtb23epFLuxaXY9AYcmrXPc9O62wUwFMDn6h0MTAkIC1rS0cv/0OnDp5O778t3+NC5cu4NruNcw3trG2sW6BfUv0jQ48rMBDNgFQiXz5ooPPPSOL9lUOL1Tb/BhYq54I2n6MrlDyR7DoqQmxISorQFFzw6nPwRM1vCQ0I7WO1B6CB3gVGtWtXS3IxLWYAjqN2Aj5BIoonlRP6raAEdUCL+qF1qQt5kH233oslUttLW3cxzHF0r11saN8NLSgq2+YnYIVmMNg1adVwll/+zi3drPdLs8Irte8ZW9XdQmvacMN/rIUppMXBTMjfQN4Vr5K6UohqQG4rrh7EqZAQ99TGmZ/OvPSu6T7TzHOO8X2q5nR5VAO+fv70mdmao9Z+np+NJO2a7BepI3ipuXuY00oMTNfSDrU4KVx1wh1iBl3y86mf7iRJxI0dnGddBGy/CgDLDHAuWwg7giVluHmEs8vLkZLhUULKTbpjuIsJkVR4Ot5jHDHU1ZhhihQqX5Sys9cnum6DidPn8TasWPYHwY8+cRjuHDhWezs7mDAgDtvP4219S3kvsfm5iaQB1y7tgviSrSiFgfR+ATT2jP5dptOgMzAt7X7BE009oUtN6gQ8d3YTicDfR2cdfu9VyjUlOx+JXMR14iun6IyYb8tnwhYnlgW9ctzUwGJ2zIc7EfmLw4gAKWP/y7llTUKcNw6IjxBDgZOkfKNjb6VmhJEZxlROkuhprlU1QDef5H0knU9Nr2OJshZx5n2t7QhuHHbunKTR02Ng5IrTuFXpRCbQjORW6QbAc169NTTASDsqrQvHE2hdS9Km5jguNgdDjYJONjHrrN6k/aBU9zKQaTfOH8PcQArv1q5jmukX6N0uMFLPwvXgInNRVW1k0ZcyXtlNpOM+anKb5R0us5FcwfE+YFkjUcHGDk0jgZOMzpLHUQTZcDcKig+6mBn7QSDuN41pYzKoipqwFTTfURQlsVSxvpiDVubW3jybx/F1SuXMOQe29vH0XUJRIx5lzDrOgycyzEaUDBVb6FQrP7Uwa3mtgkcIWlELSbqdmheofbyF4R2UpBuxZXKIJZo9YwSdURCf7H2NTlwBQWk/vdmhB6jsp+49A7V8h8cAJqhpklt18EgptlE4UKVsG3eJQRNvbpcVx/wyBUCTiqohXQ1lesFF8TL3gyfVcR9X9Hrj0FVe1Te1p6A9mJVYdP9uX5Mv2ksT1VU6lMDDkqRobVOB781pqO/SNNPVZcKXdJYsYU757j1gmBRAUjfpREAloxs59q4Fka0ybvwWaNfjf5K2ofOb+O1SIQ8vhYgdqjBS3XL8q3WQqc1pPbiwQSOXmvaVVGrKpquDpJKrIDb36SCIJhSSASpCdRYrbFPXD0T43LMSSPw2Oig+foCcBT2i9kcG/MFLp5/Gnu7O2Awtja3SyRxAItZCa1adtfnOsbc9egZ21X91j9vV1wzCUji9+ST4/WRN1pouK5rUMynfSY+C0w27KBmRaA58J1p4Y5W0OjTHPlkstQbV6wqpubZYBdEIHz1c+rX5IPc/m6Vh/Cebk5WVlcFIeThAZddNRnjd6MCClDoqPexzpNVtNcq85oi80TVobdo8n500JpKZQ2qVMS2BEwLHqjK5D00dvMit5OHMkwiVB/xvUphGrUBNxaQgY0rDqU6vxuoW9e9+3eRDjV4VeaGMJaiuWkyNf0+RebWJAQV5CgurUkjPiizmw0j9n34JuPBFjBtYMsjldYpkQ+pYh2Ry+Vlks3XcZe+ZphNE9YSy0EoMybM14pr/NDv4dlzT2DV72E+n2Fr4xjSfI75rMOx9TVcvXoNq9USBomyBlVvzizlWTu1ktIxClhEsq5IEWCl2eJO651SDzyVkbWWHrX76w+i8b0UZgpjOFa+qXBqSpAp+E2tO0y8FGDkFtKNIM1bX8uesWJzoxQ16lEZLIqEzAw8KPE09SMbaO10u4Z5zcE99zqELo1lTwnmRjGZopAepjOlMsSVuuul+ombPYEqArKXSPKvUzZQV5SzQpXwfjhstqwZuxJ5sBpZz2gnU1TaJgGwfrROpSWtzvnVh6npdKjBC3DxpQp6caxwQeix92o3c/veLIDXIX/8etHIotCVcENQeHCmTLZhkStepRGTCpCFCAC+ITQMsUqDLW0p4apywzk6sEXzs3sZs/kCW+sbOLZ9AqvdXeyePw/qGOuzTWA2QwfCPCXMZzPMZh2YezD35tGkjhhlb5eHbdK9J6PhJDMVxwgfMJR9oqXPqjuwKx0aOYCqGVeCLKgHhKm1UR3AORRZi4+KqCCAXbmwdqHml6lUXIVNK8HUUNagzjEdtGm7TmNg0CsKCO383EUfV7/smWZNrG6L1z6p1sCitFQqhzpP1DaHSv/QqzY9Cn0fwIcE/CcFYNz9zG3bapiNVIiwejC0+g31LWjroGZUg1p5QCkxdsKKOUTa6yyKwXyQgwLbVpPyT7K3CarAuewYOT82mrDLDTe7253I59cBugP1fhnQcXIw3X+ASedbVKRuNh1y8KKJb6g7CLDQSrFTNU3tZvF7IW/TMmKPjfQlLzMwjv7Wurh4oSCDiyOpx4PTd8I6SWSCsHDvDZOBHdZxdOTP5jNsHtvC5uYWrl24gJ2LF9DNErrZAjRbw2I+w/raGrrZDMOQ4ZtjuZICU231aSSHS2FAUaAVBy0/ruBbE2IedZsPErr2hRCWsKLQdrf3SiEJA6zO0aRwI5y8epUJser4US4HCwKpEMeM4QrMZAoNHhfL1mUjD7eYr0xzzDwm01sLTFPBpAJMBA2ugDB611ak94oBYbbsFRIo4LpvOf5DqsCE+0Hoqvk5CnXnqlGuY8oWFK2eCTmPno3Akdqbro1VeTk8Rw9MWGXLK9zgtfkNy29pUTBHVjJO+ehAQCJEuae5Hshr1xlsUfTE69MvTScGT1H4ltLhBi/rj+ssr7aEnXrMTF+up9S6lY8uwywT7BwYutFBVEuJlyi8q4e0B4XNg/m2q3kNVBODOIXJY/QRE92Qi6s7CFjf2MDtt5/Gxm0nsPvcs7h29TI2NtaQZlvo5utYbC1w4uTt6PsBTzz5BPJQjluPUaDrxfgRCcv1aubq3M22RjIGI5txmfAMg7UqRbc+S4SAiFxKCANHfzPF9bAYLLmRAM4qqlHUi9Icr+gaWACK8ZqYilGurnF91cuukLVNZFRh1jlBBBDU9LDPQPNYaLUPofqoXgm7f2T/jwNXXY5+i44a9Rony34id1KaysFL1RvadQr2dit2JeLWiqAUkf5Sfqodhuo+aFUVF/It91Z8OSFTWCsW22RjOsyaxQ2d1C96QkzpfraRThA9oxTUKhE0DU7WjWlM04P4s6rBFDsF+taMpDKhkWYvDLcAfAOAl7mxGr4ooVrtJwoh7b3A5GWbuOBUCl1ed78OEKa4VTRGbgCIfJsmU4g35k8U8LGntOaMEsvedbpS7wlzg9Vd3ZfZjJVlEzCLgAUW62tY9QOefOoZ3HPsGBIGrK8tMDt2VzHJdR22T96Gq3v72Lm2i2Xfy7YDgp5Vpi1oTVKEoIEGQnVJ4qcZccs3PTm2fWF6sJQSbHcUi68n6cGhHsd+pECHDE1TZ1Tj1Bbya/3B8lBvUD96hatCSsCwLopKwOrjQoCqNtccVYuWKLnYKx3rNLHPK+Zj5Uh9R+0avQl7jigepBrLmQJDblqlKZvPW1wfdaVtXOf6spdtQCR9GY+zmVrPmqQqo2l8C1q1M1MLUG3/qF4zbW+RqgrveE0CyIRvFgmEy2Gakyq4AYVTVD2c82iPrzRWN0KP6u/yxzb3H+BMBBCSxTsEpsKaOFA2pUwxZlvKBODfajrc4AUoJ6HyNgvjF/AVGhtMQRMoD8o/mpVIwoO6FVAGdSDTwW99QuPhEQoLlayvU+hs25c0NVi4zstaZDMlBmQf2dbWMQCEZb8CMCAlxnzeYXFsq4j/1GE2m2F/eRV7+3tgDRt9UPsjjeNFIUK7wdd4PKrKB2QL8lI5TFkPEtqjQRCAi6sBHNYlpwZOBUxNXnY7CKEJFbwFsDazyoOVGkHVrL9O3WohBMBITjh/y5WJPN2hwzNNCGYraR4FHrChxSPdYJy/4Y1o3gc+GSkWCepAYlQip+BUeCjPp4ZrGvFaHEdjSOfqOR9L6rZOXh15ZwokfIxW61NRvvAUDScAkTwTZ89QlzDcle+M4qGJxUhQLtRtPIAc7PW15yPhJ+vfpMm13Xq7yAtJhxu8EsApcJMKhKDuGcBwc34PCCnuf4IL29Jpk+qDX2eJeUjV9j3z9tErbfwzzy52PbU3taaVwKzXvOBtDQK2fHL4ZJw4dQp93+PSpYvoOsZ8Rlhb6zDbPoa1+TqIEnb2drF77Rp2r+0CyBLP0TNt9VWvq2vtuhGZwgivtwFo1AKSXyE/K6d+vjV7VJqiEkDoWZtMEHs1/BuBtX7OPbUCsIxAjcITzeCkJn+9NArpzLYOO2rbDRa3vR/GawYR0CoRLUBOTZBfNzcr1NQ5Rj5uBVXZ5KxXHarsPooXoZ7YTACGaiCogG15QXh9QsQyQg8yRv0dn65POIiKz4RAZcCdeuqcvIjrrkR6CYoakiehPpWubp1IDmJ4KAKCR+x0czrgyxOFb5JX2xsReHgKnrUSzT6x+GCVX6x1AVEL3RWe5eo9f2WaXsIrcRvH80yHG7xCsuMCMAUU8tWcIfy/8nvq1K2JS6TT7aC1srJhGFikQrZwUmrln2XnbxVW1MBMIkInNZcoRmrxUoCjmLg4JQwg9APjmaeexmIxx9b6AmuLLazdtYHjJ+7C1f0rABL29pb4/Be/iN39VTn1tc17SqBGJpaUJEq2Rj4pqZj4ZCJoORtpAw2rwaBtMlrV112Ix3JCtGzNbwRI7FMDG+y6Sigm3rBPx7IaMdUEeDWJ2m/BTB1np9MymGCOFOocQeUvs/OwApAK1yz1t6RylDEJjAY9QSopvZLSSl5LXHh6CDxoJrdm9X6S3+2JKdq5wRKVYK3p1JJoqgRufrUq4sEiswUwpckthDFSvoIuKOVwnUeF13sfo9LkoDAZ4WPUdhaWKXyRq+0MdQV9ZAdFK5Q3RWxTlyQGV1m3dCeR0pZQObMktNQXnqHxGLvVdLjBi2Hn27TdpELUSVevS9lz+p5oza5N2o2GY4I2r4K3MQO13TXFdGNXW9ea2nuWH5HYyZuGWnWlbowS400ijuxe28Xe/h66WcJtl65gfbGG2WKBxfwELj33HC5duoT95bIcPR6BtmLg6CpNXjTHSBHBbm/AweP469JnLfjF6ocGCtXddX1SY2zeqfT0IBF49EzbwqjWjJNbpvXdFOjFoX/b+khbgnnwhpFdpiS2Pt8ScELY1K1o2zp+XonPss8wHuVOZrekiRcLjXWdNAoxDrmP+qN1mQN8/5zURc/hM1KM8hYy1d08KvWgdscbUUGqTueOddfyWhlhHX8gU04kqj6n3tKYguNnVCuprtS5iwysbL2SSUW769W4KmMMhPZvU8/ya2Q3gVNuClhvLX1DgJeO4Upwht8knNnG6PMuSfb85AMjuDN9tWi0mYEUtEb2bjXHNcvmZphbdLDreJ9ZluS6oroNg1FAjhJSIuzt7WGVB6w4Y+OZZ3DixAkc3z6GtY1juHLtCTx74TkMw+A8zs5kdcgeDp+R1lNw45cJ42d8f099l6WvTIZFHbECNoKZXbgWXtPyfEq6TyEoJruncuce3R+/EOvpfUXV59QLxcjFmJztoqmufG+MTN4u/c4uMCzKSzUN1j7QftOZKCO6JfksS94JM+Z22MT+YrRu5RPtN9rEfCjeGs3c9aYpVVRmh9aMmBeH9o3ed46pga4VxkpBrZMrOzUFonIz1kAiFNnKNqN5lxp+jy0/YBWRlQ6qTE4zdl3L5prO2CoZ6CPL4SoCeZ1/yadRsifTixi8EiXZEJxNu/eo4PIMAou1fFSBQxlm0dVzSnMjqDYjb8hDA2epD0Zvcfv+5NXm6VG/slXZTUUTtVN5LGuBzIyOCEwJmRlPfOGLeGoxw3xtjhPHTuDScxewc20neD9L/MLxeJysrRHbP7y6wvgGqjdIHP7sSgSZhjRM8Ggi1fEColiAAKr8ER0Km8EdgWU8Nw+/RNKMwkVVkFm/o+7jzyfZvp5qgcd5YRKTYw1kfQQ8hDsuWKxmcUYIjaKB4vQSQMOfYGO3g1LbYneib6+jIVu95uX9eZBiMpFf+KWHpx6wSHbD1C4qmNgeKQBa72zttE03omDVtW9Vt2iyr2lVcytPUCRCYSiHaTKHaT6VJ424BG+Xg5Vdrwkyyu+FQdON06EGL5sWKxOwm5Zq4euaPAc1Lptw9EfLeVOxo2DffRalHaesKgGduBYyVZDamI/WMQ7B5Pc1cfwnPO5t1gX4BmlScZpnwBwviCSyA2fk1QrLoceFZY/l/rLMHL3WIqRypQ0qqMdaFyCXhWaiSqBH4cLseUWx2bbWlIGGXrZQbBfJgdL/kQIYTGKTL4UbcJX/S11bpxHffKyap5u/XEhov3qdx0Igwq9DVrtqYgLN4ED5Ur8TwIPnpM4cCka6gaDVkHXjL7iWrVUUFO3pIoDjuln8M/4iEYrs/B4FaayG9aPWxYeKPemnA7D1LyH77DbB1+ioiGSLOWFdKUJVaZh83ForGdDo/aRjSMrjUG/dZhI5wie+hIrZ2KHDKenvSWfBaliNIQQzIGPyZIAqj3DRMnJZQORnn1DyNfZ6QzgDzdEj8aw5y/U6SFMrCrVU5EBTq7dWneunawetUca3nA41eAGoBy3VLEUtgzUdpAqTmQLkavk+ftiBx3fCWN+JkqMDbsQLjZYS15aq6B9T+ooFkI2gSuP7+lNNRQ1zFO/KDM4Zw8BY7u/LQG5ypMhXHiinaYJ8D8DlFaiEqtFIP0VYttYzK5Pr0tunRgAPmDOORtb3+kjH6L4b/QlYP7f7mrSNFEozWSCeYaE1qDlhVF1IhaoSPNfxCDbzHMXIEsowgVMZiAcHMsIjNbEtZ65aVKkYobUOgFVTCLJg7+PLwDf0d3X2UyWvAp2qrGsn7ilTdaRD6YtUgaJ+b0E19n2dKLzctDPmlai+UymdLSeORiZqJp+W1k5z6flqtsShPr5WLHpHHQAhllQtAkZ5GLdsyEc1HsatUnk3IRVD/q6QtGMh+vLU/frC0OvwgxcKqfRwPQck2MF0JfxcFDb+HlXfeGz6Qt2hOXSjCTTIUGcGBCDKuGXxzCrPlqMa1KTQQTcTkhc/OS5a4V++TR0R7y+pATPLXEfNqkN5tZxNhiRxAseluJuEzxrsEDsOjuI2y+mKrT3Vs6/MfkRLBAwtknUPDUjo02ho0lZw9oHEBAtebLU3TQChUwLodKFfOYAqgTgFjdhLZftGVX9oicwEihpxeEgpYKLOwhlorjczeIWXQq3GJTjDqImqakW1vSKPr7GXFJPPsQKoCJUzvH+jmB1n5M9FoyWad0birlI+GipR7eZNBHQeMWCin/qA6sojjvStIdPKSmROWN6X6nxUrrYOfebUwrIaSwjR8xGYiqQ+begMzVAjyDSdwzoakyNOQKBoIdBZaAlFOricsV6s1/p8BLXU4GZGGw3wU5RrR2Srhkx00fNM3xDgBQiZozmJ1V3bxUQ9e3CX0fIYQ9dMKs0Lwg7qTacDJ2g1qg1leTlRMdfpAHStVIEsjHAG2Oww9QyM9B0zIbQqjLCbDkydXdgMjW2csvyTwTY7MeEs5tYcBXF2s5Sf3HywF541z3qjHQph+d+aEXokjkWjx3i3DwcyRGD1QVwcVfTocrOQkOicQZDpYC5ygOw6SHmlBi0zGVcS6wBX5/BSkpnwFO0mAcDI4/1Z8HZiNg3d6FFTyvLVtawwa698AOXhqPRVstTyLMoZc9jMHGpcvAw12stEi9RtOuoWHASmTSPY8hsjnbtXF7OoeIZGIT6lEFT5yAD3kDweks1owBWPOerUDgvc9nhT5VY9cZERo1a0d51A+m1omqMSjcmhNMo4hKvGA3IvW52D8xorzzd1IWpaTvBQbteXBXWD28S4OeXt+umQg1fQj4O3Uu2tqgAmHWjjx0HI5NmBSZ09aPJZ99KCBuO2oqPwvl7YFpMv10UHBNDxh11YtFZ/zbsAlDJeBC02RgJ0TSPGM2wBxMlaBFURHjSqd71/BY2ACWXVTatKqwYTReCiSrhYf1TgWWuGGhmfq2vhvXET2kxR1sHIsqaoWACTTDTdpa3WOvmQZEm2HaECroq4MRPnjphHeZSrLoDwdLTTTYoTu6h7liKjKzGEtub0EdcMD2ii9QcFz9LQP+1YiWA6iQoHJapppLwo7ddbtUnV+7Y9obq862+5sI9jK9D8gCpWCkbTvvgSYdxeg3jyGRdTbMHI6DtRg6pFVp8ypgS4VHGKa9mT9Kam+3jU8XGbyEE1upV0uMGrkR0goFObdBynRjQK4zsyH8stH/ye2DQQZyISW3h9NIjlphJVNBrRD/0ByzfqWCVzToUJ9MykdqeMNzy2YZwyFTOf/ilYAWUwxggaFpFDGd5QJZZ9wJqXamAqyEOHBHaf4NSYiw80L5HCHxobvjyra48iSypBgHBDstRBWWeUQUju3ECwwRrHorPPBM0lS68Llb1JFbBpuyY1myZND+t6H7D3Z7ukU4E+vM+1h6MZsFYo2maRX+MsCos+Fze7llmQgaEV3WgH1pcuDKOHTsm2rnNLQ7UC1DleP13PbdyU3RHMUM1X8JPXyxpQbc6FRRbVMq9TIQ5kSqJw5rCC2K5h27/J35WOsGUMiv3hIaxLeSymVqXzGK5jS1Q/Y3KXsCQ3c5A7ZpGpsqgBMX6p4FTl4wtIhxu8ANRCYGLQy+0OBPPdkoEV92ek8PpYqVONc+wzFkGS4WaxpL84+aI61GNMf4XOSzqKvCyqV66b1KyHUD1I1WQIyPlX+h0FuLLcd+BSmVjDiL6l9Q1xEGQ4qYBhECWLrKHDRM07/nzMvdYP9RBBaD0qIBRQR9h/FjMEUAU7jo4zDCTy89dy1WPJBqC1/waKgeWPug8dHCeePWDGXXPPWD04OFxUuwIFqNFO+7IEaCbgwDy8aB8LucLYoAs279ShprKaE9WXhdm2a7QtlYaF/vWV4FZZq2nnwEbQc9Gm1bvrNrZNkm1cSvDyldJjQIlXyjtk/7bP1r2vigRKRBsii4zv96f5L7eXlXezgtrEu5RM8YrmWJMek0WVWms7gyXfxwpYJMLNgFBE7Buz5M2kQw1e1ZqEaH3KbNEUYppT+7Ys2Lr9tvSMdv/IGjbi2Kn9KCKyWavkQpO1zpWroZQrhVkRJkDHvRxNnrB6+2eY58G96QAzE4pwCevLqHISM6Mz6bQmSfYfKh3Caen7hQ5aDja6BG7W/NyF26QLXESrZxZGmmqVD+yhhlzc+HY0wMHTQBQFi9Kq1ezrNrriMp1bMFWZbJP6islw3CKnSQQuhMjkxskReOyGtlF43xS5lo5teQRVSFx81U9N0UBzVrOijqvKkxIjyiJCalBr6v7WyP8NdUcmy6lKtqnuxombN3jRpufc3qn+ja72xuKWnM8B6SLzpJ2uhy+FoPZUJIRZWiw/lDWVHzhQ21tQWZivM7XUfh2JJ/vdjLXnmQ41eJlwJQ9raeKCGmKrZqd2YYq6BayDdWGcQgdOm38mpXkpSZQM9UAsym+yDueQJ4c8K4aZ6OBGtmE0+6rMNF4GhacKLrlQcx1wIps4smJ9jX5U4hmCSl3Yn/WhWYNXKEXa7GLT9vJESSME8iMmouBqclTAoagsxDYVHVIdL+z02iDoWoj1ca+UimbSuiIUiUcEO81ZHTaCEKlw+WbTBDocpMGObATBGUXpSlYv8k3JVTGqJjkv2lOEClhtwwFBYt7V20+mGmqTtIOHUoUnUWjG3nDwp4lM4gzR+2zERzpeq9KdClG5siHh1ZEspEW2XBAglevuq/MyaK/y176w9o/AKxqAp+Iv+kwySQYE8mDl9kQDpii8W84LrNUa+111aejpml3QkrlyKHlh2DXZ4kOTyhILGUGN0RPcRZUzMmeswv2EiXh7QNUZrUD3q1HzbAWpS31mcyy39YY2y3EuyoQdzCW2rakyOwOcuVhvcp2ZR2yuwawFqraNOWfbsGwAGjm9Anr/OsmDalYyYKhXCZoGST51fah5KkseuaFam9sIkAlg3ZQOWbuhA7M4uE36vjwUlYPUvKf1HbfIn1Ilpo3WEX+PZn9tnamYkuLlWJdxJJBQFa27dhO1PRDboZ6OmnO93Tx0NTiwbEvLek+jPDHRD5Gm9VW9g8ALByeyMeStme746+czqkb1Tmwljf6ttCMZvzeW28rt0eu1lEMoZ23Bfo1rEYvLkKUDfbYSRQKcoW61RNKn8g1o3WpWjZAYCY2MsR/lraXDPfOKrrYApsxHJPZkQ312xlK5HLdRTOleo4Vj6K5/qq+YQNQNmgJ0EpopkW/TZDnIsBZ3KrBQqzATC7iTphvFGfMaVOHXCBqvpovX4MrdavNF8xwLIYCQSaF2wjDIzsaqidu6VqSTjKhJrbf6GQeAt98GcPBoKI6d7ISp3g6AnOR3K01MGWkEU9jRyYiu2/W6iNUnKkSsRXl9ojOPTzhjLynzXl9w1HtxKeSlIOZZUHhPlZ3aRO69pj1nwMUZZogmHTPhZfY+9f6O9PRUrT1WpjE2ymhGdYQVgJkkKoYL8Yp/qrEYL9BoguY1q5U9cO2/66M3vkNVHYQrqif0W/U+A9HSwn5gk13TcRMaLbIheD2rXXByGlLok9jpxFWemm/JLHZRDtoIBTrXVHKa1q2tGj2iRez2F5ION3iBANZjOJouYWErgngUaTgoqjtSd4/LPx4n0bi/Lq/6qXtNhI0VrCQ/kvzFcI0BAxJ3iGq/CTB9jQDyIT8uXwSZmiSs/kFr1j2btV7IBrgZ3MwG47Afr+LZH0e2JbDslWSCHCaSPAd1WQqJZZBqOclvlOERMMUKZ0ADcDUqRbgOWRdwoan1MuipdAGlSOOsEXI2ODL8i5QI9Gm8LP2reuYF3ZgYFQW1waPZkYgYCr1BKDvMjVGUUs5L5ZeakoWzgpu37+NVIQvYfJhVSCUTpOUiezuNZ3LAU9lnptYP2U4QPfiYnDMcFrWcCC6xHwjE8haJ+V0tKlScNVwpVZiLxn7NxWlRg0N7ylakPtuzVjeq6WwYRG3t4ybemLf3oyo6iXUJIzqsUHibZCs+KiC1MWIVaWVFgEDW5RIYbVrQLXkxipNYiy3SPlWClGOopmZVB55uu2/V0Ja+MMPfoQYvIppwACSJHlEINnLSaDq7Zubp6fy0q61n4OspzXW4eElhQKjLbbWUdkD20KyDAmbXrE3xUZXGBTTNW0s1W52VVc9PZHu9JJxMREhpem+ZCXxj5AaQQ4MZ5bEpg07UXSvHQgXQquZNv0pbq4gfYCRxj/dglc1gE5qZdhvyV7CqIolI/5c4erXonA7e+/xSrGnssfGcd3qXj+8Vq3lAM57qe5tFJYjtiVwzCG/U46wJd2FdowpOVAR0nHr/sQ4eBXFolBxqIOdG9AzAH37Wt2nEAzUnTb3oNdD7NTcfXK86JweqYoU54G0jI1XgUq7VI5iad9psjA8OqF0N/hTJUvPMxHv6/aAKtGrWC023BH3vete78B3f8R3Y3t7GnXfeiR/90R/FI488Uj3z5je/ubLjExG+67u+q3pmf38fb33rW3H69GlsbW3hR37kR/D444/fcuUb8WdCsNwbs7aZqORPjz6nINArUVOtPaggKy9XLt+xYw3npJsYsq9KL0ldq3Up71KuMgotDWtPKoSVZSrvLWbEE3N1g7In/91cHY9tZmTJi8FlFkleZ3WRjzzeikyvdql1hCgGIas2qN6WUYELnybWuJ6DubZaJ42EVCN1cMeuQInt3ni1kzGRvbwy1epW0WGf/VtddRgHaB6ZBadVAv9PnjHGJLCNOXk7IczM/f1KRSOlf6ivPJlo2sSmZaq3rurz43mP1zq4vYGZglccjLe9ZgH0iYv5N4poZcMbpdAVOk50H5962kqT0RAhJBYa1mpDmTe0xrjymZQWDf9RrHwcuKBiDRBHl8zJ6aNjjhigsOpUM1Wobf37IHpcL0UwrFtcr6/dKK9WukWpo9FaXki6JfD6yEc+gre85S34+Mc/jocffhh93+PBBx/Ezs5O9dwP//AP48knn7S/P/mTP6nuv+1tb8P73/9+vPe978VHP/pRXL16FW984xsxDLe+gDdWqqSH5MMGqgFRRX5juEYaHtgxFYgJ8/ABLygTeOeVUaNeiPEo7NiRXmcELqnz5zh6wwAZYZWXMBLytfCtQWj6Oc1HwauuVguG41ocpLu18+HwVlAGPAqDA45mXSsrAeX1twKXKTdspuZoEqmrN4JzE2Rjhx79HVvSPENtO8d9b3nEvpjqVIdzlDDsERAbwVbxUAV9VW3GX70SlWchoXJWUQWwkseWD4Xc9KTwQA9logY8PfOmbhx/TnFNzdP+q3ZVCj2OoAaGbK4vWE0EoG3vAaIjuCVqL7jxP9aYpjHUHogDPVYm5tGY/m2gNuNm9F55l7Qjxy2unsdEWfUDtYubWkJeGGR5uiWz4Qc+8IHq93ve8x7ceeed+OQnP4nv+Z7vsetra2s4c+bMZB6XLl3Cb/zGb+B3fud38AM/8AMAgN/93d/Fvffeiz/7sz/DD/3QD910faa0WY+dV1QstberqS4Mb/s3SDQfrDiACWu2uH79wrcyywB6lEC2HQiZByR0ZgqIA32CL638ah3koKrBeT0xkMOetAKeeVRIztmAABD6Gpe63kRUzDhl5hpH8AFsSfUtFx3w2IvEVlYUP3XTZHbEjCKwS718Rlwyqzz3JqpiB+URoAFzixhzXXqyGUHAGiMlyGxnWg+snH3UhIumjhPTG9OgR6Od2gvVrWz1lJUtcrrERXlz60bZJMsVMxRPN5rqiAiCjd5XP1Uh5wFJeKcTnhces1UqMet6MOyY45Qwjdp9BAWFiOlR1ahwE7l6i2LV/WOinTzOt16UYNtwXM8iNVKlWIV8A4nXwBdiR22oqlitS4RnoiNTUAqKck3GJwf5RKrZMec8Lie0OPqcVfz8d5Re0IrZpUuXAACnTp2qrn/4wx/GnXfeiW/+5m/GT/3UT+Hpp5+2e5/85CexWq3w4IMP2rWzZ8/igQcewMc+9rHJcvb393H58uXqT5ObJ2GanDGq8FRkSiZCFqnMVNyMB/iulNxoQ8AUqx+sQdSMStXFgWWJOqxFMdw0FweLLm76epqPhtqE6SmDMaA4oObM4Oz5intCU1nPt9RJAY8rc2McmolksZzcjKCCxupl621OlGhkMakXCEiKtA3da4HV0mjqYW3vWNjUoqMI7srUFWbCoSMnSiPoNgZzfuAM5sFop84MRI0TfWwz+cAf7yW8iVSZjbyv6hqTNd7Ne6j6zusSAVouwvu6XNOcc8w6PJMsI12STwp0mUHidVse4Yo0dYaB5pWJrDVf1S1ueWaSlPWwLDKienjMZ9UMI6yhjjhWTKLKXxUbUf2oAggTNf5IMVefuUxtNXFalP5ocaTaKtEQxJYuQ90qngAb0DGXJU+TEWg5brJyX9X0vMGLmfH2t78dr3vd6/DAAw/Y9Te84Q34vd/7PXzwgx/EL//yL+MTn/gEvu/7vg/7+/sAgHPnzmGxWODkyZNVfnfddRfOnTs3Wda73vUunDhxwv7uvffecqNZDK+EKVphVWsSDPh6gL3f5lm1GLXkae+FPPQqNU8oSID9k3xgjMRyYOZi5kIFGhUjNuLd/iZty16HUbsicDXoQEGQTRn6JstiF5X1O9raKa1NSmT/a++38g460DCuu9VJhYV8txpMCG41g8ViJDMvVteaKhpO0LRN17ll5VllWqcPF+gs9THeQGiD0KBeI9N61f1Ru6FzLUTNPBhzcadv5wkEmtZ9quqg/dk6lhQhKnq7Z1Pfdh6q+0kVhzrVCtcUgDG74I7bGH0Ut1yK0a92nE0/YQ2MxJgsswI3rsdKPban5VMNOq0kmk4tCGmZ1DRKwTtWcbrkG6dbe/r66Xl7Gz700EP49Kc/jY9+9KPV9R//8R+37w888AC+/du/Hffddx/++I//GG9605sOzG+8l8rTz/7sz+Ltb3+7/b58+TLuvffeShvSE2xITtAqih7BTsa14QaA2LSImh3IBJyzi1bQXw9fJwSRljHKwRCMUY4fcS207JdK4WmN22aLvBwzcZYfMbbalZvZ04ELpLExE/jsAkKFXpltJF1kT+TLLTVlqu+R9YsrvWPOTchyE45THOI8wKP6Vu+zKCxERjkO90z4Ki+2DgvRDhLRUxQRckLU9ZNZduQqkuwqU2RdW6kDy3M1D0fFQzlf85reTBqUgolqGgBrf8gzNFJiQhXZ4A9x/Ub7Oo06i1DZItkKsfPnSv1CmQz4achGGaufKSBgi0Axhs5mHIPL7Ii03d4zDtVjGAJg+9DcvUfM57JZZMTUVf9KC7TpbWKAuFYShtBCUDSpOl/pBRZlpeLNqdQiiDJjgL8QohJxS4rRKix4lzfr8dYqA3bR8sS0PLqF9LzA661vfSv+6I/+CH/+53+Oe+6557rP3n333bjvvvvw6KOPAgDOnDmD5XKJCxcuVLOvp59+Gq997Wsn81hbW8Pa2tp1y2Eq/nfE8fh6F96muUGAEgwOnm/KMIkBi1kogV45aH4+QF1M2hVS8Q4XgAD0ULsOZauOvpOz7HlKchAjiSs36e6UMLCUWSvR1CHGcNR21A4hsFleVlmBJNsJBnkHtpQe3dXDsHGzCsFMhAnq4FsLCKMzvM6gsKVU7B8DabST8kYmb28M6zXWD3wekCV2ZAaQGpCISccmaR9LmUld0ORI9VRtbG7AWMDMh2xplwmP6ulaeaGwTsFACdLeCpFJ8BeXfs6Iru5kZqJkDWRpm+70MiAO5mH9y8g2+ym8IPXnVrfn0BxCWSOUa4nBuZjD/YDIwUSf9rd6lPoomlIqPMBwtfaZGGRHuiawLRR5LnIugLwjPOSDxjiIwr8AkDnLXsBa6CaIYgF2vKXwAOmuSeWSesvGJP20prmzvOJzpUwKdeRpgNN7ytDM1ndWvap0k0ggLksmEVi8ZrBWMGePUBT6qz5yZdw6h7FKswjj6e8OuIBbNBsyMx566CG8733vwwc/+EHcf//9N3zn2WefxWOPPYa7774bAPCa17wG8/kcDz/8sD3z5JNP4rOf/eyB4HVQKmQNI7K54xweB8x17OI2IJquoepXc7/thJCDmZi4eksxtayTKMDqxfBQm39bWX2SWk2u1hp9PYt9WaeZVTiZ6nLLtzg0OBAuDJURMZ2NoXlIIeWjNoxU79Vkb+pIzbVxHzC1prADyBoUm9jVYaw1QB4ysftR0vqAvxmzzVSq3c3JAVPJHL9rbYIXl4nqdv139BlMf3GoAFVXRlq0/et9Kd8N3F3FC3qCzHQ8Z5rIND5vyFbfnSJa83J9b7ovAt8EnrAVJpoe3YHC4V0arwWNaOVj29Z3m3E/lj1R4rQLA15I/HdcdA1sMcexOIncEv8NraAJmoR1Xi9LazUixN9ZuqWZ11ve8hb8/u//Pv7wD/8Q29vbtkZ14sQJbGxs4OrVq3jnO9+Jf/7P/znuvvtufPGLX8R/+A//AadPn8Y/+2f/zJ79yZ/8SbzjHe/A7bffjlOnTuFnfuZn8KpXvcq8D282tbwRdV8blBBBFoGk0ThrSRWEeuSjFjSYi4ecRNmI99wg5ZVTU0jFVCwL/UjIXPTHRFQWtiPuVtxCYjIoNxSQovCKVddzuxiMnNkGaKkT2fu1ywjsHRtw5K72VlZg5LhPKAoRCvuMTFslXdp3goRIkKh7NizJRwRShIsBFEI/MlCO5Ahgp1YW1tkFkZvfiOwjEECrIO87zUhmG8ZXFAqwC6i+q0nHZPMIiMeDvPQy2TfAvbhMi7VGe+9p3+nsipVsGuOR7ak6s1iXKiLIOJGOFwEuDu+WGkvZPG6r0yUqOKjoZscUVYpCW6M4QLj57eXEpTz2yo/bxW79IGrjc/gzU/Mr76FyBFNV4xgBhusXKLzlOXjkEx2r4UIto9QOrJpNha2aX7J9mmyzdbJGqQIT5WEtCwK1TIZo8rrxaLooeXFN74OWiG4l3RJ4/fqv/zoA4PWvf311/T3veQ/e/OY3o+s6fOYzn8Fv//Zv4+LFi7j77rvxj//xP8Yf/MEfYHt7257/1V/9VcxmM/zYj/0Ydnd38f3f//34zd/8TXRdd0uVLzHNE3IT+AYBrMLDbj6JF6vvU4QP6kSLlm0kgfCgAhgBk7MiPVwxgzGwxjWTDY6UAnNfT3x4ng6Y8ZuXGMVa+6nfMo+fimDrXp1tm2JVpgUesweQjfBE4fkJO9qBqQDJATEbRwAoFaCxINXN0ZbHTdch9r+6ik2UW70R9tqFKVQrmOrvUzzLCDHZEVtfc2Sz+hXkemvmVeKMak9kfTRVS9JSg7teICmcH8v9ON9qRGL41oBfbCn5u/VqNYerZGSZdurwUjSPceheVxjGVJFWcMAOMNTNPCoF4/U2/6zpBLRRaMbwHCEyok77XJMBAJBuiYgyjQzzDqLQKKNKTwiAy+GBBpjqk9mbay8g3RJ43ajAjY0N/Omf/ukN81lfX8e73/1uvPvd776V4qcqJAxECAtdJSlDwZndR27UEeVS0H/MzDfBd/5DZ1MEDWwKHpurypMc8hDGC1XJXNYwEhEys61/QfINIq+ufmhTZcxg9sgh0bsxrH9weEe1dK4zx2gmILSOAQLq9rTtnNJYg9087DWrjlq4TjLnh1C1dtBaHzZehk1TUDtg+ACz9ZIDcChyzQ3wSistsiKsu0qNU9gfNpVV6S6npfFSkAfJZpB1k2oFRdd5OSgMLZ+T8TGIzRt+3CW1MCuv8KhYo5LuoRbZXI3HRruLQJFUwWNf/3GnnPCG0ih0WWHNccc4RSKI1m2Jke8pulG52KhM2xFkiwyYik06nld7HrFnx/WBjDnFmyjbHKQVTxrwaBsodYy3FIeKUwi1LzXvh/s2vg5WEFoG+rsALU2HMrahEmC56l23GfVXOCJbBFk782rdUdWFo9zLzrYcgM3UlAGQAZvU64oITMlKtcFEZck657BXg7uSBRUBVv4GECV06HyWA9lXVemW3soSZFe0xbKJrAhLzsjM6HkoQZlyMRsm+GnKjPKMnu815HomEy3WSU5Jzl0WcxwwS0CisgCdq7qJYIbz7iBOFXH2kRRkUfa6TEWXbEGJFV0oLETTACSJDSfaZQKVGWy0v8aZjl7vILMvNWXq/c7KqUxZKe5Xk4/UQU+RppTK96SuLNpPEp/PzIpFMnVUD0HlRhOyOQekCtfNESEXTz3pc1ciGEAvX+VZzqbQaH7VdioZBcS55JUH4yezbug9ZNCgY2oQUxRBHTYcoLThXWPihUUOSBYoOdAUhE5eH8htKx3IPBPVhKubzql5f0oIZ10ftKkWOWhFujtsIAtYmV2IlFpCM/J3CjmzjNk43uHLCxxqZkBkKpf/jiG1ALcSmKKgilzYwRf4lSF7WkNTi2TT0d3ZdWMDimXqd3GZIWB8crIHC7B3GotCNfMK4KXbp54voBH/XULh1yg9/vjjvtfrKB2lo3SUjtKhTY899tgNvdan0qEEr5wzHnnkEXzrt34rHnvsMRw/fvzrXaX/7ZLuhTuiz3Q6os+N0xGNrp+O6HP9dCP6MDOuXLmCs2fPIqVbj5dxKM2GKSW85CUvAQAcP378iHGuk47oc/10RJ8bpyMaXT8d0ef66Xr0OXHixPPO94WdBnaUjtJROkpH6Sh9HdIReB2lo3SUjtJROnTp0ILX2toafv7nf/6GYaNerOmIPtdPR/S5cTqi0fXTEX2un77a9DmUDhtH6SgdpaN0lF7c6dDOvI7SUTpKR+kovXjTEXgdpaN0lI7SUTp06Qi8jtJROkpH6SgdunQEXkfpKB2lo3SUDl06tOD1n//zf8b999+P9fV1vOY1r8Ff/MVffL2r9DVP73znO2FHp8vfmTNn7D4z453vfCfOnj2LjY0NvP71r8fnPve5r2ONv/rpz//8z/FP/sk/wdmzZ0FE+G//7b9V92+GJvv7+3jrW9+K06dPY2trCz/yIz+Cxx9//GvYiq9euhF93vzmN4946ru+67uqZ76R6fOud70L3/Ed34Ht7W3ceeed+NEf/VE88sgj1TMvZh66Gfp8rXjoUILXH/zBH+Btb3sbfu7nfg6f+tSn8I/+0T/CG97wBnz5y1/+elfta56+7du+DU8++aT9feYzn7F7v/RLv4Rf+ZVfwa/92q/hE5/4BM6cOYMf/MEfxJUrV76ONf7qpp2dHbz61a/Gr/3ar03evxmavO1tb8P73/9+vPe978VHP/pRXL16FW984xsxDMPXqhlftXQj+gDAD//wD1c89Sd/8ifV/W9k+nzkIx/BW97yFnz84x/Hww8/jL7v8eCDD2JnZ8eeeTHz0M3QB/ga8RAfwvQP/+E/5J/+6Z+urn3Lt3wL//t//++/TjX6+qSf//mf51e/+tWT93LOfObMGf7FX/xFu7a3t8cnTpzg//Jf/svXqIZf3wSA3//+99vvm6HJxYsXeT6f83vf+1575itf+QqnlPgDH/jA16zuX4vU0oeZ+Sd+4if4n/7Tf3rgOy8m+jAzP/300wyAP/KRjzDzEQ+1qaUP89eOhw7dzGu5XOKTn/wkHnzwwer6gw8+iI997GNfp1p9/dKjjz6Ks2fP4v7778e/+Bf/Ap///OcBAF/4whdw7ty5ik5ra2v43u/93hclnYCbo8knP/lJrFar6pmzZ8/igQceeNHQ7cMf/jDuvPNOfPM3fzN+6qd+Ck8//bTde7HR59KlSwCAU6dOATjioTa19NH0teChQwde58+fxzAMuOuuu6rrd911F86dO/d1qtXXJ33nd34nfvu3fxt/+qd/iv/7//6/ce7cObz2ta/Fs88+a7Q4opOnm6HJuXPnsFgscPLkyQOf+UZOb3jDG/B7v/d7+OAHP4hf/uVfxic+8Ql83/d9n5299GKiDzPj7W9/O173utfhgQceAHDEQzFN0Qf42vHQoYwqD4RD/SQx8+jaN3p6wxveYN9f9apX4bu/+7vxile8Ar/1W79lC6RHdBqn50OTFwvdfvzHf9y+P/DAA/j2b/923HffffjjP/5jvOlNbzrwvW9E+jz00EP49Kc/jY9+9KOje0c8dDB9vlY8dOhmXqdPn0bXdSOEfvrpp0fa0IstbW1t4VWvehUeffRR8zo8opOnm6HJmTNnsFwuceHChQOfeTGlu+++G/fddx8effRRAC8e+rz1rW/FH/3RH+FDH/pQdVDiEQ+VdBB9ptJXi4cOHXgtFgu85jWvwcMPP1xdf/jhh/Ha177261Sr/z3S/v4+/vqv/xp333037r//fpw5c6ai03K5xEc+8pEXLZ1uhiavec1rMJ/Pq2eefPJJfPazn31R0u3ZZ5/FY489hrvvvhvANz59mBkPPfQQ3ve+9+GDH/wg7r///ur+i52HbkSfqfRV46Gbdu343yi9973v5fl8zr/xG7/Bf/VXf8Vve9vbeGtri7/4xS9+vav2NU3veMc7+MMf/jB//vOf549//OP8xje+kbe3t40Ov/iLv8gnTpzg973vffyZz3yG/+W//Jd899138+XLl7/ONf/qpStXrvCnPvUp/tSnPsUA+Fd+5Vf4U5/6FH/pS19i5pujyU//9E/zPffcw3/2Z3/Gf/mXf8nf933fx69+9au57/uvV7P+ztL16HPlyhV+xzvewR/72Mf4C1/4An/oQx/i7/7u7+aXvOQlLxr6/Jt/82/4xIkT/OEPf5iffPJJ+7t27Zo982LmoRvR52vJQ4cSvJiZ/9N/+k9833338WKx4H/wD/5B5ar5Ykk//uM/znfffTfP53M+e/Ysv+lNb+LPfe5zdj/nzD//8z/PZ86c4bW1Nf6e7/ke/sxnPvN1rPFXP33oQx9iAKO/n/iJn2Dmm6PJ7u4uP/TQQ3zq1Cne2NjgN77xjfzlL3/569Cav/t0Pfpcu3aNH3zwQb7jjjt4Pp/zS1/6Uv6Jn/iJUdu/kekzRRsA/J73vMeeeTHz0I3o87XkoaMjUY7SUTpKR+koHbp06Na8jtJROkpH6SgdpSPwOkpH6SgdpaN06NIReB2lo3SUjtJROnTpCLyO0lE6SkfpKB26dAReR+koHaWjdJQOXToCr6N0lI7SUTpKhy4dgddROkpH6SgdpUOXjsDrKB2lo3SUjtKhS0fgdZSO0lE6Skfp0KUj8DpKR+koHaWjdOjSEXgdpaN0lI7SUTp06Qi8jtJROkpH6SgduvT/B2eZKzrAzUn4AAAAAElFTkSuQmCC\n",
-      "text/plain": [
-       "<Figure size 640x480 with 1 Axes>"
-      ]
-     },
-     "metadata": {},
-     "output_type": "display_data"
-    }
-   ],
-   "source": [
-    "for image,label in train.take(2):\n",
-    "    plt.imshow(image[31].numpy().astype(\"uint8\"))\n",
-    "    plt.show()"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 13,
-   "id": "b39ab5ee",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "from tensorflow.keras import layers \n",
-    "data_augmentation = keras.Sequential(\n",
-    "  [\n",
-    "    layers.RandomFlip(\"horizontal\", input_shape=(256, 256, 3)),\n",
-    "    layers.RandomRotation(0.3),\n",
-    "    layers.RandomZoom(0.3),\n",
-    "  ]\n",
-    ")"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": 14,
-   "id": "a5138b88",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "cub_train_gen = train.map(lambda x, y: (data_augmentation(x, training=True), y))"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "5f9e1933",
-   "metadata": {},
-   "outputs": [],
-   "source": []
-  }
- ],
- "metadata": {
-  "kernelspec": {
-   "display_name": "Python 3 (ipykernel)",
-   "language": "python",
-   "name": "python3"
-  },
-  "language_info": {
-   "codemirror_mode": {
-    "name": "ipython",
-    "version": 3
-   },
-   "file_extension": ".py",
-   "mimetype": "text/x-python",
-   "name": "python",
-   "nbconvert_exporter": "python",
-   "pygments_lexer": "ipython3",
-   "version": "3.10.9"
-  }
- },
- "nbformat": 4,
- "nbformat_minor": 5
-}