vit_ecg_model.ipynb

{"nbformat":4,"nbformat_minor":0,"metadata":{"colab":{"provenance":[{"file_id":"18BsCxOzoZcNUipgayPGNHevKr83zZYOa","timestamp":1697045481919}],"gpuType":"A100","cell_execution_strategy":"setup"},"kernelspec":{"name":"python3","display_name":"Python 3"},"language_info":{"name":"python"},"accelerator":"GPU"},"cells":[{"cell_type":"markdown","source":["Vit ECG Image Classification Model"],"metadata":{"id":"7CJJUmxZx2dy"}},{"cell_type":"code","execution_count":1,"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"a-hv13auxqqg","executionInfo":{"status":"ok","timestamp":1697229824168,"user_tz":240,"elapsed":27855,"user":{"displayName":"Mahsa Bakhtiari","userId":"12793844954925855500"}},"outputId":"3c0a6b6e-7314-49f4-b4e7-e67776099fd1"},"outputs":[{"output_type":"stream","name":"stdout","text":["Mounted at /content/drive\n"]}],"source":["import tensorflow as tf\n","import matplotlib.pyplot as plt\n","from tensorflow.keras.models import Model\n","from tensorflow.keras.callbacks import ModelCheckpoint\n","from tensorflow.keras.applications.resnet50 import ResNet50\n","\n","from google.colab import drive\n","\n","drive.mount(\"/content/drive\")"]},{"cell_type":"code","source":["# Preproccesing the dataset\n","train_ds, val_ds = tf.keras.utils.image_dataset_from_directory(\n","    \"/content/drive/MyDrive/ML /ecg_image_classification/ECG_Images\",\n","    labels=\"inferred\",\n","    label_mode=\"int\",\n","    class_names=[\"ECG Images of Myocardial Infarction Patients (240x12=2880)\",\n","                 \"ECG Images of Patient that have abnormal heartbeat (233x12=2796)\",\n","                 \"ECG Images of Patient that have History of MI (172x12=2064)\",\n","                 \"Normal Person ECG Images (284x12=3408)\"],\n","    color_mode=\"rgb\",\n","    batch_size=32,\n","    image_size=(182, 256),\n","    shuffle=True,\n","    seed=2,\n","    validation_split=0.1,\n","    subset=\"both\",\n","    interpolation=\"area\",\n","    crop_to_aspect_ratio=True,\n"," )"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"KTnE5R0EztoA","executionInfo":{"status":"ok","timestamp":1697229850342,"user_tz":240,"elapsed":6950,"user":{"displayName":"Mahsa Bakhtiari","userId":"12793844954925855500"}},"outputId":"1b656ad1-f763-4b33-a865-41fba32f0458"},"execution_count":2,"outputs":[{"output_type":"stream","name":"stdout","text":["Found 928 files belonging to 4 classes.\n","Using 836 files for training.\n","Using 92 files for validation.\n"]}]},{"cell_type":"code","source":["def normalize(image, labels):\n","  \"\"\"Normalize training/test set\"\"\"\n","  return (tf.cast(image, tf.float32)/255.0, labels)\n","norm_train_ds = train_ds.map(normalize)\n","norm_val_ds = val_ds.map(normalize)"],"metadata":{"id":"V5tTG9d75qeO","executionInfo":{"status":"ok","timestamp":1697229854366,"user_tz":240,"elapsed":366,"user":{"displayName":"Mahsa Bakhtiari","userId":"12793844954925855500"}}},"execution_count":3,"outputs":[]},{"cell_type":"code","source":["# Verify the normalization\n","i_2 = norm_train_ds.as_numpy_iterator()\n","image, label = i_2.next()\n","print(f\"Image Shape: {image.shape}, Label Shape: {label.shape}\")"],"metadata":{"id":"2PfFZZb-63l4","colab":{"base_uri":"https://localhost:8080/"},"executionInfo":{"status":"ok","timestamp":1697229879804,"user_tz":240,"elapsed":22604,"user":{"displayName":"Mahsa Bakhtiari","userId":"12793844954925855500"}},"outputId":"79146a9d-9cc5-42c7-a8d2-3899629a955c"},"execution_count":4,"outputs":[{"output_type":"stream","name":"stdout","text":["Image Shape: (32, 182, 256, 3), Label Shape: (32,)\n"]}]},{"cell_type":"code","source":["# Show the image to see if the resize algorithem keep the crucial information or not\n","plt.imshow(i_2.next()[0][0], interpolation=\"nearest\")"],"metadata":{"id":"6my0hQ0z7mIl","colab":{"base_uri":"https://localhost:8080/","height":453},"executionInfo":{"status":"ok","timestamp":1697229888602,"user_tz":240,"elapsed":809,"user":{"displayName":"Mahsa Bakhtiari","userId":"12793844954925855500"}},"outputId":"a5692ed5-24c1-46ab-8d7c-e7c0fd151e78"},"execution_count":5,"outputs":[{"output_type":"stream","name":"stderr","text":["WARNING:matplotlib.image:Clipping input data to the valid range for imshow with RGB data ([0..1] for floats or [0..255] for integers).\n"]},{"output_type":"execute_result","data":{"text/plain":["<matplotlib.image.AxesImage at 0x7883f8afe110>"]},"metadata":{},"execution_count":5},{"output_type":"display_data","data":{"text/plain":["<Figure size 640x480 with 1 Axes>"],"image/png":"iVBORw0KGgoAAAANSUhEUgAAAioAAAGQCAYAAACAiUs9AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAEAAElEQVR4nOx9d7xkRbH/95wzMzdsuptzICw5SlgyqARJSjCBKGDAAKjgU0R9Kjx/4lOfT5Lx+cAAigqC+hQTJpSMgCAgrCy7sDncvZvuzJzw+6PrW326Zy6w113Zu3TxWebOzDnd1dXVPaeqq74VFUVRIFCgQIECBQoUaAuk+MVmIFCgQIECBQoUaCAKDyqBAgUKFChQoC2WwoNKoECBAgUKFGiLpfCgEihQoECBAgXaYik8qAQKFChQoECBtlgKDyqBAgUKFChQoC2WwoNKoECBAgUKFGiLpfCgEihQoECBAgXaYik8qAQKFChQoECBtlgKDyqBAgUKFChQoC2WXtQHlauvvhqzZs1CZ2cn5syZg7vvvvvFZCdQoECBAgUKtIXRi/agcsMNN+DCCy/EJz7xCdx///3Yc889ccwxx2Dp0qUvFkuBAgUKFChQoC2MoherKOGcOXOw33774aqrrgIA5HmO6dOn4/zzz8eHP/zh57w3z3MsXLgQI0aMQBRF/wp2AwUKFChQoED/JBVFgTVr1mDKlCmI4xfmK6lsZp7aUqPRwH333YeLL75YP4vjGEceeSTuuOOOluvr9Trq9bq+f/bZZ7HLLrv8S3gNFChQoECBAm1aWrBgAaZNm/aCrn1RHlSWL1+OLMswceJE5/OJEyfisccea7n+sssuwyWXXNLy+YIFCzBy5MjNxmegQIECBQoUaNNRX18fpk+fjhEjRrzge16UB5WNpYsvvhgXXnihvudAR44cGR5UAgUKFChQoCFGGxO28aI8qIwbNw5JkmDJkiXO50uWLMGkSZNaru/o6EBHR0fL50VR4EUKsQkUKFCgQIECbSQN5jf7Rcn6qdVq2GefffCb3/xGP8vzHL/5zW9w4IEHvhgsBQoUKFCgQIG2QHrRjn4uvPBCnHnmmdh3332x//7744tf/CLWrVuHs88++wW3EUVRyPoJFOhfRLSE/vKXvwAAxo4dC8AExwPAzJkzAQAPPPCARvPPnj0bALB69WoAUC8qj2x33HHHtn389a9/BQD09vYCALbffnsAwOjRo53vyUO1WgUAzJs3D11dXQCAl73sZQCAxx9/HACwatUqAMA+++zj8DJv3jwAUG/ulClTnHGmaQoAGlPHQP6w9wQKtPE0mHXzoj2ovOENb8CyZcvw8Y9/HIsXL8Zee+2FW2+9tSXANlCgQIECBQr00qUXNZj2vPPOw3nnnfdishAoUKAXSPR2/Pa3vwUAvPvd7wZgsvgA4POf/zwA4Pzzz0d/fz8AYO7cuQCA+++/HwBw5plnArBejjzPAUA9MOzjgQceAAAcdthhAIBPf/rTAIAvfOELDg/vfe97AVjPy5///GfstNNOAICbb77ZaeOQQw5x2jjhhBOce2+88UYAwEc+8hEAwP777w8AuPzyyx3eAwUK9K+lUOsnUKBAgQIFCrTF0pBITx6IQtZPoED/euIZ81133QXAxocsXrwYANDd3Y3u7m4AUABHxn00m00AwK233goA6v1gm3xNkgSAKbUBWO8Hv2e8yS9+8QsAwMEHHwwAWLNmje4JtVoNALDtttsCgMauLFy40BnPgw8+CADYZpttAADjx493+mJ7/rF02HsCBdp4GjJZP4ECBQoUKFCgQC+EhrRHJWT9BAr0ryNaQuvWrQMA9PT0AABWrFgBANh9990BGE8Lv9trr70AANdccw0A4IADDgBgM2kqFXcLYh9r164FYD0p9KwwY4exLfSWPPHEEwDMnnD88ccDsNk5t99+u9MnediwYQMAm5n09NNPAzAlOgCbYcTrSGHPCRRo8DSY9RM8KoECBQoUKFCgLZZetOrJ/wz19fVh1KhRWL16dYDQDxToX0TcKoiJQq8GY0EYA7JkyRKt48FYFXoz5s+fDwCYMWMGALQgTrOPvr4+ADZWha+McWF7JLZTr9e1b3pr6J1hm8RLoaeE4+js7NQ2yn3yPb1EwaMSKNDgaTC/38GjEihQoECBAgXaYmlIx6gMJusnyzJFnGTmAEtN77zzzgP2A1jsh+222w6AtQ5JtCj/8Y9/ADA1jQBrPS5btkzROzcX0Tq88847HR5mzZqllu+mJmZNMPuDmRzExnguPokKyliDrZ2oS0899RQAm2nyYljp9DBsrFdy1KhRbT/nnD7zzDPqjdhtt90AWK8LEWZJA61f8sQ1yjXLGBXGj/hEDw5g0WunT58OwHpWqJfLli0DYNcnP6cek7iun4/nF0K8l+uTsmQ8zWDbe+CBB1S2K1euBGDHFTxAgbYkGsz6GdIPKoMJpo3jGH//+9/1/jKNGTMGgLvZAVBXMjdLPojQVcz2+BDCTZogV3zAGTt2LGbNmrVR/G4scbNdunQpAPsD8fDDD2P9+vUAgAULFgCwPxr8weLDwgv98aTC/e1vfwMA7LvvvgCAW265BYD9gXj00UfV3c7UT7rwGcD42GOPAQCOO+64F9T3YIj8cl74A3TPPfcAsLI6+eSTAdj53RzEBzTOCedg2LBhAIA99tgDwHM/7A2WsiwDYIHaqN9+X5QX03mZCsy55H38YSSc/aOPPqrzyzYJAMdA1hc6vzx2+cMf/gDAPqA8+uijACxMPvv7+9//rgYHdYow/QSfY1vUcz4M8UiLDyp88OIR0pNPPgnAzg3XGPcN6v8LIbblH2GRB+oe52DRokUA7MMWweg4hqIoWuaHa2vOnDkAbNAzU8XDA0ygF4NCMG2gQIECBQoUaKuiIe1RGczRT57nGnhHa4aWEy0jWn/8nH3wvqlTpwKwliktKVqWtGjWrFkDwB4tVavVfxlIFK1IBgwuW7ZM+aKVyld6Vuhp2VivDy3rCRMmOH3SO5BlmcqKcOW0UpnqyntJm1NOHCePEWi9khd6OfxAz01BHBetCh4f8hiCPPnXb0piEUEeO/jAZj75FhDXBz2P1COul9GjR6tXkbL2vVMvdFy8jl4Mej84hmeeeQaAXWP9/f3aF48g6X3hNQSmY9v0bvlFFumloWeQngiu63IA7wsdE2XEtUEPET0m9OawTV5H7x91k94RrtVKpaLHRyy2yHmijAYKXA4U6F9JAfAtUKBAgQIFCrRV0ZBOT+7t7R0S6cm+Ff1i04vJz5YmixebaGH/8Y9/BADsueeeAGwq7FAnFhek94YxFZuLyttZ0LFAgbY86uvrQ09PT0hPDhQoUKBAgQJtHTSkY1QGk/WTpil++MMfArDnu4yKf6HEc3Jaic+XmUEely9frunCG8s3M42YgcG4Er9Q2kBUFAV+8pOfALCZNxz3xmaWbCzvS5cu1SwHnsEzJoUplDxPZ2YVx0deSfRAMB6IvDO+aFPy/a8g8n/44Yc7n2+JvA6GCKHv0+Ya39Yit0CBtlYKWT+BAgUKFChQoK2KhrRHZTBZP0mSqFVOHInrr78eALDrrrsCsKXpGWnPbABmrjDynudrbIdZAYy6Jy7JI488AsB4D5iNs7GZNeXsHcBifhAThHKgx+ZPf/oTAOvBePvb366ZAvQI8cmWmTjMlmA2FHEo2Paxxx4LoBW06/koiiL1qDCTgt4oZoWwT77nuA488ECHZ2YukGdiRTAzg+Pl2HbffXfN0hpK4VhDidfB0NY+vkCBArWnkPUTKFCgQIECBdqqaEh7VAYbo0LsAr4S04MeEXolGB9BzwqxIWjV08tB7wY9L/S4MH6E+BzleJKN5Zt4CuSR3h16JuhJ4FiIzlnGm2A8CMfN7/g58SF8iHSOh16PjeU9iiLMnj3b+YyeIcaqMPaGWSHPPvus0xezYChryoN4E/SaMOaD8vH5CBQoUKBALx6FGJVAgQIFChQo0FZFLzkclbVr1+Lee+9t+x09B8woIZJloECBAgUKFMglZvVtDO5TwFEJFChQoECBAm1VNKRjVAZDzz77LK6++up/aZ/PV0vlhV4TKFCgQIECbSn03ve+FwBw6KGHbtZ+gkclUKBAgQIFCrTF0kvOozJ79mycccYZAGyGyeTJkwFYdNPBejcY28KsGmYJMVvmrrvuUlwQZqUQiZVVc/2+/Qhp/3tmHvHVR2gtX8+/Q/ZLoECBAgX6Z4mYZJubgkclUKBAgQIFCrTF0pD2qAwGRyVJEvU+zJ8/H4BFQyWWx7777gvAIsx2d3cDsJgdxBnp6+tzPmcEMzFCTjnlFAAW+2P+/Pn6HXFAGC390EMPAbBYJkRiJT4KkViJwPr4448DaK1GO2LECAAWAbbc3yte8QqnzUCBAgUKFGiwtLF14oCAoxIoUKBAgQIF2spoSHtUBlPrJ89zRZalV4M1bujFoMeF74lQyydBeivoDVm7di0A66kYPny40ycRYEeNGqVeF37Ga1evXu28J0/lsQLWe0PPCVF0yTN5IUou71u3bp3+HTKLAgUKFCjQi0Gh1k+gQIECBQoUaKuilxwybdkLM9BZGWNTWG2YdXOeLyOHEdA8t2uXPeTf64t/IJ4Gum9j3oesn0CBAgUK9GJSQKYNFChQoECBAm1VNKRjVAZDLyRTaNKkSc5ruzbaEXFTnuv6gXBRno+e776NfR8oUKBAL4Tonb3tttsA2Grwxx9/PADjQaY3+bHHHgMA7LTTTm3bYtwfsab89z7RO+3H1w201wbaOil4VAIFChQoUKBAWyy95DwqgQIFChTohRO9sTNmzABgMxTvueceAAbjidcwo/Kpp55y7u3q6gIAzJs3D4D1uPT39wMAnnnmGQDA7rvvDgD4/e9/DwB43/veB8DG/xHrirwEemlQ8KgEChQoUKBAgbZYCh6VQIECBQr0vMS4EGI4LV26FIBBy6anZMKECQAsphQ9LCRiUpWxpQDrWWEfRNzmdZWK+akaDBJqoKFPm3zWL7vsMuy3334YMWIEJkyYgJNOOknh3klHHHGEBrXy37ve9a5NzUqgQIECBQoUaIjTJsdRedWrXoU3vvGN2G+//ZCmKT7ykY/g4Ycfxt/+9jdFgj3iiCOwww474NJLL9X7uru7X3BONXFUNiYPO1CgQIECDZ62RMitkM049Ggwv9+b/Ojn1ltvdd5fe+21mDBhAu677z4cdthh+nl3d/eA6b8vlAYDoR8oUKBAgV44MZD1u9/9LgBg4sSJAKD7eZ7nmrp85JFHArCFWH/1q18BsGnIkydPBgDsv//+AAZOO+b9pDRNAdgjIFLY/4cebZEQ+owQHzNmjPP5ddddh3HjxmG33XbDxRdfjPXr1w/YRr1eR19fn/MvUKBAgQIFCrT102YNps3zHO9///tx8MEHY7fddtPPTz/9dMycORNTpkzBQw89hIsuugiPP/44brrpprbtXHbZZbjkkks2J6uBAgUKFKgN8XiFga3PPvssAOBPf/oTAFNIlR4Qphmz/AiLpNJ7vmbNGgA2/ZgBuTRkGVx78MEHOzywPXpzAr20aLM+qJx77rl4+OGHcfvttzufn3POOfr37rvvjsmTJ+OVr3wl5s6di+22266lnYsvvhgXXnihvu/r68P06dM3H+OBAgUKFChQoC2CNtuDynnnnYef/vSn+MMf/oBp06Y957Vz5swBADz55JNtH1Q6Ojpa0twCBQoUKNDmJ8YU7LrrrgCAnp4eANbLMW7cOE2U2HbbbZ1XAr0RoI0eF7aZZRkAYL/99gNgQwXY9tixY53rAr00aZM/qBRFgfPPPx8/+tGP8Lvf/U7z4Z+LHnjgAQA20CpQoECBAgUKFAjYDOnJ73nPe3D99dfjlltuwY477qifjxo1Cl1dXZg7dy6uv/56HHfccRg7diweeughXHDBBZg2bZqeWz4fMb2pt7c3pCcHChQo0L+A/J+KwaQG+23Qg8LYlOdr85/hYVPwH+ifp76+PvT09GxUevImf1AZaPKvueYanHXWWViwYAHOOOMMPPzww1i3bh2mT5+Ok08+GR/72Mc2GkclPKgEChQo0L+GwoNKoE1Bg3lQ2SxHP89F06dPf8Gek+cjotoGChQoUKDNQ4wP+epXvwrAwuAfddRRAEw20JIlSwDY2BTiphBKn5lCzNrhMT8LGzLrh/eTmDVE6Pw777wTgIXYHzt2LOr1OgCLscJ4Rt7Dh6BrrrkGAPDud7/b+f5nP/sZAGhmKn9TCJmxww47OJ8H+udoMHIMhRMCBQoUKFCgQFssDemihAGZNlCgQIH+NURPBT0YhJ2YMWOGZvPccMMNAIBDDz0UgPXG0JNC5FnS8uXLAQCHHHIIAOCOO+4AAGzYsMHpkzgtLHrYbDYBAI888oha6D76Lb08/I3wjxn4OTOW/vKXvwCwx1E+lkv4rdk0tEUi0wYKFChQoECBAg2WhrRHZbCUwzzRRQhnjoECBQr0nCQei4PE6zFyxAgA0FIm3d3d+NujjwIALvzABwBYxNnhw4cDsHgqJNrUhx1+BABg5apeAMCOO+0EwMab9Pf3AwAyqTfEHbu7uxsAMKqnRz0ijGfplL56Ro8GAMyaNQsAMHfuXABAKl6eqsSozJYYlErFeHsefPAhAMAI8cAEP8rzUyRSKvT9pv1tDR6VQIECBQoUKNAWS5s8PflfQf9MenJabyDvNaiH8UvToRQoUKBAm4xYXRmwmTRDkfhTmLOS8xAey7+aisjoQFwYT0osmWFR3OpZGUx6cpiJQIECBQoUKNAWSy85l0LRuwrpOnOWGdU6zWdyBhvBPBXyiRp8Ooz85zl58paHRT5F6rdR7lwHmLPPqOS7ohVSxObDikyFbcptowB5bP9sWfB6MhXZ00IAiKLcjrOQNjhO3qJ907KQD/LCucDPgy805oc8R85r0ebEMmftjogylmsj9yy68MbLYbmjs3Kz34t8ixxRJPInP76ItHHvvfbizkEh8lC9oDxEnk77LfPpyoZdUOdUtu2nUcdlWeVcuhwjKsmicM+PraGon6AtDSifSHhm/RWxpkTOKCI7PwPJpqUzzpdcJzraYqFrwwM5gvPS326GCTyLOW7RY7bt6yDXTftzd9W1PFILUiUQubpDAfgt6bxShlyDOfVCxqV7VVK+DIhKe4DuFS2rTu4RfnVdeN/rJ7F0UbjXeWMwW44w4tXkiaLn+YnhnOheyhu9tUfWdT9rlWNZ951PCs4nlS92rmdX3JMj0bm8rErmE+993LoR6Z4nfGdum1GsA5GX2OHB2xb0j5aZLDLddlRnWnhx2/DXXqE35O4F5MlduqUhRmUOnYsyyf7qHjdG7qE+/3MxK8GjEihQoECBAgXaYukl51GJAHRMMHn2WcWgJlrrLzWv8kG1YjwuWTO1NwPIc/O+o9Pk+BdNWoHikYkyXijXmaj0Zr1hGRHLI66ZJ85KYV4zmg707khftU4TyZ43ModnbS43uAKdNXNdQ3AG1OpBE7Wq8CvmC5/29bla2qgKT2nmelRo7XZIVH2z0ZD7aXGJnMQarkp/jTQtGTniQRKrRjOw1OJOy8NHrdNE9zdEdv5zub6vikdK5qpAJmNK0dlpMg/SRrPctCXyJCiaqciuiGh509oXK18smNwzuSifmuhFWm9oG2oRi9VC2aQN6pxrzdY6DG5Elprv1cun8pG+OkS3dC5KzUiTxK5IM+mLZmtuZFSrGRk3pS+OQy0v6avKcUlfkXhU8kLWQ4dpJ8sKa51qG+Z9rcusqYauBcpH9JpezNjMhcqYlrfobGe3aafJ+eZ1UabrsKNjuPRVR5kKWXs1VmQXcaQ5vQHkheu3W8ZN/XZXXyZ635l0okkZq1dS1lSHWTNFSg+BqzuZyLBTZEickEivM+8rVfIsXgGu4QjIM+mLe4WuBcup6Zt6wIr0vILzLzo4wF7iO7Mq1QryVNaMjIPz1tVlMoQajYZ7k3qtzUu1w6zfvEGe6WGiLlpvHVDO3CytO5nPW265BQBw7HGvknsyudbIrFIxa4v6TopEllx7+j15jbmvy7hrnchkT1GRxPS6mreJvM9UP02b8+c9DQCYveMupq+6Kx/fo1KR/S1VnlL1YM+dO8+0tf0Obdvg+u3w1m/hrT2oF1PmInd/H/h7VuvsVP2E561Nly2VcboeseBRCRQoUKBAgQJttfSS86gUUSkWw3vIs2f6bb8uNcLvGfNQeJ+71z0Xued+9h6NHsj9RuwJo/OpeiTan7vHeeuzre9ZsLnwjHvwrhggQcyehfMsn1Ye36NV1m1bsvEztMwwUF0I6wZz2rPnzt4BK55rOtrH3lAesc7zc1sFlEM5pomq0VrfYqB4B1qKzz3ugdtt0wPZydzP1XMSu7rXSt6NbN/jRUMAym1ZlwoZbs+j915jlrz1oHFQjAVpc6bPmDEbL+K1oRayfNCyEOQ+LzZrIPlEni46feXtr21Z1c9jePpeDN172l4sffhBBsqce2/L+lXr/4Xx5tyr8W+uF9onNrlsqUGPfWbJIgDAy/bYy+PJ9czl/mSWOZDxPv73xwEAxx17rHOpyn6APcN699rb73YO7H7Ruju7ellabABszaO777wLALDDTrs69/t+iLiNbrFZ8vN///d/AIAL3rej3Nv+t4FttIiubQxKGyqs987nl16ZVolsGgoelUCBAgUKFCjQFksvOY9KW2qxVrZk2gRPrAM9Uiv9kxLwD1jb0XNYRhvTxQvwJ2xEa4PurO11ZUeSWhp+hphv/TxPF88/qo04Ex5ADzZ69tuKr8Wl8lxdtrx/Pj/YwDO2Md+055HW74CJRS8qPT9TG6mmA97ZOgfP1bJvxbf7tLWnvtUG3XbBApMtYj0qz33/cxFjSwbymAyoiy3XGZr3tIkn2WbWNKfd6AXx2V5mfnzb897drplB6qfl+QWu98KV2/NcPDimnoeCRyVQoECBAgUKtMXSkPaoRFH0nGf0be8pbIR17B3OpbQCUnPeVmf+vXekzWyJZl2+YNA03LNdpOZ9I1lvvs9zMI2ebeRNyX6RM13yxLiXrCnX1Tc4vHhJJCgk26Ap1zGriOf0WZqhiCXamxZj7lqQzCDIJOo7kiygIiuEx1z6gI6nzAR5LiQDIKtby9RaWOaPZlNkm8TOeDKeD8t4GvV1Iq7MGT8pJYokVTkjFk4sYsjQqK93xy33Mo6lIQNihgFyN1si4zRLBkcsNUGYeePjMDT7hdcsV5nlPp4IlYkZCdIWcTOaVBS/D8pYMjoayTp3bBrTFCHTLJaK8mMakRd532hsEFmJDjLGQ3jOpZ1Udcti1JghmNe02VA5xCU+TF+in/3rXF7gsFSaZ8mK88MrpM+0f70zBs1Mi4FIJqwuukM9tfOUyniolG6wCjNKCsmwy2X98oyeyR9ce5m01yhsfIeGalBW9X4OQG5y956Ce0pDdIuZhsJTxvltMguMWXA2DozrriFtWtkV7riajCdwvlYdZTtFQ9YN58SLi6G+V6Kq7nXUON7DtQdvvnUuMsl20b1GdEgxXET35PvYq74c5TZTslE3r4sXLjT3MOspYgYg8X5UEbQN0wf3dU6wWTe/++1vAQAz3ny6M7Y8KlQfdT+WW8mTVoumXsjrMxKrwv2NVaYTzifIk8iYc8a+i0Kzedb0rXHGFxWuR0mz/ercZLkPUi/Iu/nrb0/8DQAwc7rxII0Yxmw+mePGBs0oo07kyfPE75V+pzf2NxsIHpVAgQIFChQo0BZMQ9qjUhQFBlWqiA90iYssGnlt8ckv1noF7omkomYqqqqL5AgvejxOklIGhhuJnSSciqz8cSnq20UL1KZ5HbyxVISXciYC+9LGy6OB4qAUnnz08tzLjqmwAVpaIhci8ZZQRa315noICsqWyI2aauAiGlI+FvWVXiB6oIS3xL0vjhIdALNxCncaUUXV/SCJ3feUPQ0tIpuSd8o19yyFJFLMFZ0Hz7LkuBWzRUUWaRvlPvl5VK2473UqZKyI1DtHLwzb8lE92TKtP7sK2JZr/XPeFS3US4qJoggF++L8EQeDayVpbyNZRFeXV4uqKfqj+kLmyUPRcujOdWrP5GPvc3fcutpr7ppS5F3rkjNdZ4nltTIAqicn1su00nFlLtNcvzq/OT1zMlzuFyUcC3/d6fxStxS8WkE+5Dq3b21H9xL3Z0KHLzznUckzSq9LEbttJO7a08Qrj9dI9YKyl7eaBOjuF4iBirfuqFuxrBFFKFIwbOqgzGfMeXT3MWaWPTV/HgBgfcN4xYZ12krQUYkP8yJtqPOycMY1UPafv/Z4lbcT2XVTFMp37+pe4cW9i+Mg6rkqpa4VjxfRsaVLl5m34onZfY/dWniJKq5nq01JH4fKv9OD+c0OHpVAgQIFChQo0BZLQ9qjMhgqIqBaE9RP1qGgwSmP7ak8JdeqBv0ybfqYAPIELiiRecrHZ4uOWb5OkT899EHAxjuwknNG64Dn4eJhSWrSV4NogW5YPeNjalK/iEildmw5YkHipSWU+ciqsXuumqdOE/oEXqm5yLQ6FvDc0tyfVM11aaNZSnaRp/yqxJB4mRYx66ZIXzVB6mz0SxyJF4Ae03IXOfLsNC5ZmjXlt+ncq8/1Mt+Vas0ZF9FkaTvECT0vrAXi1e2Q11qnoK9u6NcPE63VIrIR1GNFiSTyqvTQSd3Lms73lDGVkbrlo6/mEbQGSCKysXpMz4Igr1I+daJsehaPXM95TxnjpXNh2qmK/PK00PNsv3ZVlcirOp/CC9eMxEfFFbblK6HIU/QcRBvOuF4sCjJRexsb+p0mVHeEX7opiMTsY0JUBFU27W86DVhcDsj4O9FMXcRO6nPCvppuvJeGS+heYda56irjKtgO13BBdGSLcUPPSVXaSJvqApRr5L1awxVHLrFGmHA9cL6b2kf5Bnrx4qSiCLyx7imR8ELdcvcj8vTMs4vN97oVVUoc2H0trsj+px6bkoUujHR0GJ2IVPYuxapbpq1sAOTeWoe7v/nejYTrqtapa0GHFbvorjZG0PS1avUauV7m29vfSBQHQz9iomY3iBKc2rWjcS30CLneDsrQzkVdvzHjknkXD9SKlSsAAGPHjpXr+Jgg6Mi1LuXDzoNb42lTU/CoBAoUKFCgQIG2WHrJeVTKWSN+TMrzknd94ZvSah3y+3b9F84rr9Lzc7V+GHvRnpWBeffbGeAylOIBnDvR4v1ouSJy+9Cr/FCe0n1+1k/htWGRdQfilU0997hfSM6/xTTxr2w/roF4aplvnXfbjmZCaezNQFwVztcW7fiFyYfk8K6Ntde5gZqylpqh3NcHfw7aNNQiY68Nnxe/Df8oP/LWic9LWT4DzbxmOUTu/LboIHsYAAykRe9LMVovdP0qebLM/b3keXEprHdnQBm1vLrjaNG5Afqwe5W2AEBimArno5Zx26Y9vRDP0qKFi/hB2/t9KutX4Y17zZq1Hp+8ydugVPVEL1rk157a8RZ58+UPl7f09/c/Z9v+vl0MsIaL0nerele5fPvr15u/51t7/cxQG2g/LOu5z+9zjm7wFDwqgQIFChQoUKAtll6SHpVCnxAlJ12+S7SMiDxhM/ddC3a4MSjEetD25HPFWeDTdalej/Yp1yaFfe4137vVgxXDRV55ts12Ci/eQOsxeLVyoqLQuA2FT9E+c70GgMX0sAAxbh8SvJJ71aY1NAfWwiRPhV/0RLMYzDWa4+C5MfzquYXK3vm4ZFG7uCRREVn5U4acJ7j3arR87NYXidRi5mU8OHYtMGuZp9qOnm9zPDo+3ku5uNgWmoHlzXPsDqWkyxRECd/As5QKr0KxzpPXRhG5582xyjJzv/f0peyAKHwFzsm/i22hnLjhPvbsWy01r/6MxmYQt0RiGYrIjsPzeDWa/dJE5LTB66nvnCtVC427cPcNivHeu+8DAMw56EC7fn2rXdceG7VYHIYX08cXL78cAHD+u84RnkRuuduOYtiUeLLrWPAuUsEkkeyNWpXxLTKsiPFc0rZOq8878WXMBY///e8AgJ1n72Qv436ja8CPJSmcPlvWFOfRW99+DA/FqnPlyMS03btqldNn7G09hbeHWs8BBdA+Jif25IKi5PGI3D3UKjb3Cncv1Z9e4v/AnYQBazvp/pZrW2vFg2QvdnUjVt7ctWf3Lca4wOubc0Kd5hrOkEVufI8fr7Wp6SX3oIKoHCxpgqrs1sEFaSjXwCV31fD6BgP9WLqd93tK2myY4CXzg8lFZS5pMmI1o1JB2rI/8gCQSQAUU4T9IxS22y99RV4xwziy/EaahucCd3F8BAqzIGRw+qoLsFehwEEk3VWMWMoBk5qWSOWnHLwfTc9NmTPYNPV+JDXFVnjmxwpOx9ccTY9fBdMrOF4JoiYrmTvfCgem6dlc6B6Ilez4Kd27mR1vpNmW8jBE4C5uVPyxl+i5fuqo+5xm5cYHvX5F3xOeyEze8gCl8xW7usVAXAVP44+n95CQKU/UC/fhodEsAwT6a8Zcu5464d5qr+O66Cc4F3nQpwYAQF30PNNAdo7NFjSsM4hWxvX73/8RADBm7BgAwN577yXfw6EU7oNYwaBxne7CeXvb738HANhn333LT84yPvPaFNmoHcJO+Tsm+vGPuU+a73Orv4Cdi1QUneuhfNzKezaIbB64/y8AgNWrVwMAjnj54cK+tClp1Rqg6hUAbHDfol4Ij1/7+v8AAD73mc+Y61MrjNwzmBi46a89Pyhex5upBsu43Pm1act2DmgENZoM2BWZc7+L+BArfac0gry9Rni2e6iRD5MOmhLYXpWA6byIFKhPH3aJV0nAQqa+S1+cP0Ib1OvUYwJ0tn9AzxmjTTkWecuRszXq3L2AbaRypOM/iOsDmz7gyt7EvVS3XLv/5zb6WcZDtjfPk0o4+gkUKFCgQIECbbH0kvOoFCil38YCbOShCtFaja05Z0iuJ8xzRR4jcwW3IhFHWdrpEOChNEPBZ0OxsPT4gYBmGgiVu68CulSk/NxLQZPH3kqFac4u7DWKtARwRdc3zTmmOhqeEl6XuQBXdE8TbCqjFUseND3ZfF6RdLc0tSarL6NIQbQI/CTXipeDc8XyBmqhEHpeLU6RT+y60s24ZXzqOXLBtyoE7uK41CpmaqjHvEKu08NiZQwASa2q7VAWififaZ3EBMDKfRA6N0Xcw3mDPSIyH1DGNtWSPBXaGcefqafEtSQrMu6m50mw4xXdSiStn7IHvUKu3mQo1Fq14HLShqTpZg336Ec9ZXIf9dgve5DnZu0xFVrd06qLqXp8OA+ppFUuW7IcADBy1CiHX6Zz0nJWnmn9Vuz6NWTe0xP1rMChx0lS8rqJDMU6jWvUQeGt8ADQcqap06vrAoTxWCNWELbYkVcelWQs/PZJKuyKlaukLeq9u2Z8G5hHIHY98HO9wuEtiSvIePSmDhBpg/uRwrq7KdHq9YsIVic8KSKctMelq0cpFW1BdYYAbnJNJSZIpOe15r6tak63juG5mhAozrSzbJkBQCN8P+9PKhVW+rCOLVfEul/R08L08ky8MpSPOtB1MihI/hZxjy4dFUknK5cvd2RTRK4MuRcyxZneGXqa7DKnzEX3YsrT3XMrcQVIPJcYvcw+oOUmouBRCRQoUKBAgQJtsfSS86ggKj3FKuS9oXVrTdG1b193HQCgIk+/hx1yKABghx13BGChtJOEwWlu3EDhHR4mYvUWeV6Cjpencj6lw/WEqCEtFhpBu7w6gtbSFAuNPBf6YMun6sgW9GKxOS+A00KFRw6PNkiyIn0Ya7bICIBFXugVEctF+ovyvOTZMS8ERGq11mhpinxoBWTe039EuUkftIIY80OIcsTqdWgWLr+2MdMHrZ2c4xIetIiZBy1uYzpogZvvVT7Nhv3OdV7Z+aTXzbNWE/Fe0DJVDwUtrZJVZ9rhmXi5P1rf9E7Qw2LPuQGrn5kGJfqeMo5L2iGYm+L9i5wq9DgWgBfvUHj6madu/IxfloFejKzkjTNyku9lvpOMRfqo8LGe86u3SQt0uvrMNnQf0EUVOa/qGfQWH2M4mHJarVY1HsoGn8ia4lrw9gol4WHVyl4AwPp1pljd8BEj5L7cbYdeoJTzXiAncKHsS0zTtZ4PrmtCx3vw9XDHF6uOSoFBidVZtmy5tFeVZhIbK6elAnz9bIloBQBsWL/e4dEv7+DPAT2G5ZIS9MZUBMhNb/VKn9CLqZ4Rr3BmEnN/E9A98QI1NKYx0vGasSUqG7t3SpMqGwatuONPxdPGtafFOOV7rjnuxbHu6wyAjlWPCa4YeyVctKyIFFfkngIWneS65bxX3HVeZ0wLx5DbNUw+bMA6vVUhRiVQoECBAgUK9BKjl5xHpUD5TNZ9oq4LVPTOO+0CABg5aiQAYLmcAdKjYklTWZzW7B9yVek4z/uqlM3B9y0uBpf8I8DIHQM8i6NdRW2/6ZYwCP3cfLNayoiPGD7cZVYvdMffwnpR+lAHyqd/L3NmIF79M1ztzPWKDchDO/K8GK25dfQwyFUD5eDp+GmhWp5s2rQ/fq8PbUwsKJFPpikpnoy915b2oggRC8MVvp4W3j3uOOB7FrxU0ucVbVHqq6UNlxeffX++BwZXczMXWluyxJiVvr4+AMCzzy7kN23bGGg9PN86L7cVDaSAA4iexIJw9nv/hgHeA1Y23iLxZZt7GUn+PjYQ2VgsfmL13JdZC2jaAE2vXm3mpKOj1v4Cv912PNqKnACAtevW8mL3Zp/84furyutruXiShs0Y3tqs6nv7Qo/6vbwuXbLEbWOA7YHFZlvWcGHnc+LEiXKPr598Ty+Vu1cOtKcsXmzKGvSM6nHacdbeAIClm4uCRyVQoECBAgUKtMXSS86jEhUlHBV5iNdnQ7G8mEfOgBAt4sQCcRJHkjYFI6Dp4pHYEu+838S+5FkpRkXOqH/ww5sBAK876RR+QW5MHzyjr5exWNBi9RZyxrlhvfF+NOR8sbvLFKLK0xyIvUJSmfs0rPzLuX+938jp/Pe8BwBwzTXfMDw1iE/RvmBcyswFGTcKy+6KFcYqGT1mjMO/D7HPmINmY70zPvUUybnpWjnjrgsvo4ePdK7LixxN4dc/J2efTbk3VlQ9D0cjorh4HaPgaZnKbcRbaMjneWZjEeQiBfASLIfCw6wpdN6lUfLseTtS3tfYIB97gFJ5KXuJ2Vu5ew2zWJoChJb7xTc144Q4KsSHkblgTJaum6Ztl6LmuOWeZn29fODhvshfWeHOf+y1Q3mRF83EKc1FRnwYkQ37GjbMFCkkBglxYQqVPVmR8TJuoF/WL7OCFK6iKeNOZWx1zVaz45Z7Gg3nfeFlGFHGkyeONzdyr+Eao7zIu18xFDa2Km2IZ0HwU5IOwYvKmFEkspXrLTib+z08GWeSqbh82VLTT2r2pCSqIPLGxeymlPrpgeZxUTH+YSS9tVqEUy6jbhWCbRK5tnVUWF3pXWXietb0mnFnHo5Kyowr1aVc2wCgBR4VjkXiA8eONnvVhg1mLKlcEMWRXVMcnQwz87L3OH8LFyww18kewt+GLPOyuqS9VO6rKF6QBXyzcVnSt+5vLk+qxw39wL2PeFKyh4wb3SOdszAms2GpB3XkqRfvxzAWbB4KHpVAgQIFChQo0BZLLzmPSpkYocwcd0aJRxL9TEu72mlcL0QNJe6EYgAQ0kT/YAdevEglshaiWBI3/vBGAMDRRx4JAOgZaSwLnu1rFg8tEQbs0wqSR++qRJnfdPOPAABr1hnL9Ywz3mRuSxI9m9bzf316F+uGOf9ywTIp9008imqV6pI78vDjB1gGXeUDa7VceOGFAICrrroSADC8Z6QznkxklpAXnqczG4RdSdu3/e42AMDy3pUAgLPPeItph5kORdXC9CeuzOjF6JAsiVxBTjgcNxOByJaZ4pO4Z9rEX6AFg7hQzw90XIYylaEjOlR9zBcP8ZGyrkbCM20YYvmoZyHSDCDFPaFxx2yPCvEkhJda7LSh2RAF9cKbd2a0EMOmzEts+TDjIjaFq8caJ0IZy6eci4J4IXJbJXYtVGtgUz6FZuPlOm53XTPjSmMuyIt6KSW7oSp6rPgzbmbHwkUm1iWJrb5Tx+iNSiqx0wZNZcqUpNlrlCn3ooRZHyI/XQ/kWZotgFjGzb5S2b9WrOiTTmVc9Aipa0FemGlHrCLibIgcFi94xvAg+lKV9vKoUN3KuEcW1E9vj1RXk+gFs3kUsiMqf13KFnQ9iprBlQNVEUbfWjPO2RJLGNWIJ5I6bel8ViJtA7C6pd5J2VO7RxhPHOe/qjhJmeqOJj2ph9iNMeO8su8xPWNEPsQ44X2uRyVWPCV37eVxpFlXup/pWnG9VzWuPb8NjWdzwWpyFYi7PohLk+cZ4ordiVHiWD1Im5iCRyVQoECBAgUKtMXSkPaoRFFkraIXSEUEVGuCjhmrK8R51SdnwfBIGzyPE5RUFvnq7AQg8R+wnhOtndAUpMNu80Se1pu2LoO00dVpzo8XLDDolqN221X64pm0ua7S3W0+bbhF6kipnJevW89zdw6YWAgNJNIXzRWLAyJtSBEznmkvWmgsxp7Ro6UPc11np4l70VgfMkF0UWm32ml4zhpNlem4SZMBAM8uNufcO40ea9iUJmLWqaCMRXbNuotPQat/2WLj9ekaJWfc4mlINM6ggY6a+S6VGIyn580DAMycOVNkZdqqSuYB67IQs0StfPmrSowTWwRD+iocnhv9dTW1iFWgfdWM7jSpW3JdU9roEN3i2bXGR7CYnUxGrXu4yEdimITXIoq07lFF9J3jT8iL6GcH+WWtnpInDABSsUhrNTOfaZ2xDXJmLXEFHHeWtp6fZ6Kf1S7hlzV/wOEzBsm8r1YktorxMHLd3LlPAABm77KLMEecmVzbYRxHp8imscH0RQTPrilGB2PFl+H88ezejQepcFwybpuZYfRh2szpIp8ONPXsXjgmXkaXmc+iyT3CjdFJJQ5g3XqJAxIvgSIxix4koqMR17UWioTGDdQkLm3EyFEAgLV1xnfQw8b5rUpfkL64fmXf6pK1JzrKeIndd91DpCRYH0kVOVGZc64702dnp+HhsUcfBQBst912Tp8NImqDnmNpUw114sQQeZpean5fKL5NtWbGXRMZxV6GGN9HREcmmjP3ZOGF+1tDvl+9qhdlYt+1jm6ksi+pTniZUZWKq1vkZccdZwMAOmRNEeWYpKFZdFJ2yG+R7k22llX/esYCcjXRw2colf24NsyMK6U+yBXcK4mqnDWpH/x9JMaXnDB0d+tvos0ulTY83CNS+Xd6Y3+zgeBRCRQoUKBAgQJtwbTJPSqf/OQncckllzif7bjjjnjssccAGBTHD3zgA/je976Her2OY445Bl/60pc0F3xjqCiK9nn1L7gB98ku8l99lEQPXdWClPDVPYfOS/ECA9GuuxrLcPEik7u+2267OW22jq99W7xu7VoT+V6puKiTZRvZnoN61onHL89Fp06dCgCoi/XQKRbmQLy04JMg0jZHi3dm4aJFAIAdd9q5bUs2g6o9LgG7GDZcrL7MtVBtTEfR8gR/zz33AABmzpwlfbkZVK2aMJCOudfZytBRm2uYWeV+rt+6R/DwkXxbxt+iW61z4fflX2NLy7t60dqOew498PeWV23LCzFo4cXDumiZCw8T5N577gUA7LArPY/O1+bP3OfXvE4YN975XvElWsErHF6iAdb5yhUmLqomFnp53jkvuSc724L7nl6OYZL90rvGxFsMHznC4aVVflbO9hLzWa9k/Uzy9tZW3bFct+uL8qEHlZ5mZ069JeLP409u+TEA4P0So8Ybli41ntUO8QJFKhc3E8u21xobodXfvfH4q5dtKDZJy5ri9e7cjBB0YH9s/q9DuVfdv1r6cr8feG3yDxf9unwdEXNZHdvnwfLrfz7AOpZXIhr7+7irFy9sL2j3+WB+szeLR2XXXXfFokWL9N/tt9+u311wwQX4yU9+gh/84Af4/e9/j4ULF+KUU07ZHGwEChQoUKBAgYY4bZYYlUqlgkmTJrV8vnr1anzjG9/A9ddfj1e84hUAgGuuuQY777wz7rzzThxwwAFt26vX66iXzvCIMDkYip2jc8YYeBepdcuo6MK5nqiTq/vMk+ziZ0x8ybbbbwugVFgyafccKBgsUndm4qQppo0li+T71LlaswDkvUbRM1aBMRDyxD1C0HTr/Wud6xBF8G0RRvXzvFxr/Mj3S5cYa4dekFW95ox/1Gix8qQqqSY0eOiM7TwRu0hswRJBZtS4FmtCyLhdrBL1SPDQVt739poYlViRLV0cCx93AQBiPQE2fT/44IMAgO3l3HjeU3MBWEtlzn77AQBqcgauFLnnsVHCuUlbv6cXy/Pu+LqlWC4Rz88Zu+HWztFsItuS005URCVFJEm8i5dp41+lVbC1tgl7dXnS6tItnkfYOC22rtlL1G83PkCrBinKKAfKuTKvrFTMejPLFps4KtaC2W6bbTSrqcwNAKxYwSw2Vg2mB4y6JXOgnhV6FDyPkVy3XjCLxo4dK90UQClmxFzrr184fxVaodu8blhvPKLrxTPamqE30HrIEcWurEYLsmiLT1ZrGblYLDpOL3uRclqyzOxRo3oEmTWmHqSAN9+R18Y8iQvrkzXF9U9eOrSWj6sXvr5bo97i57C205q1pm16TDS7J3fbsPsU45rovaX8yAOzhTz8JWfduDJXz4kOn54h1vJhtpr70+v7y7T+TuLypK+F9RTXxMNlw2O8uUjcGmwqY4dz28DYcaLPujexfhp5TEv8eJwPpVo/TzzxBKZMmYJtt90Wb3rTmzB//nwAwH333Ydms4kjJRUXAHbaaSfMmDEDd9xxx4DtXXbZZRg1apT+mz59+uZgO1CgQIECBQq0hdEm96jMmTMH1157LXbccUcsWrQIl1xyCQ499FA8/PDDWLx4MWq1Gnp6epx7Jk6cqPUF2tHFF1+s+BuA8aj8Mw8rPNuLa4w0N8TMAWb1MHslFYRWoqMqkqG8f/KJJwEA20gWScbHP3lCbZRQZSOJLF+3xlhMHZJhs36Nsc4KptZ4Z/aMD2E15UjPLvVgtdylVsDMS4iQmpGQeeekhden8FiXaPKX7bk3AFvzaMrUKdKOG3ehlqmYQ3XKOS+wQfhfL/guREFlZoqinHrOGM5VoQikLi4BEW7X1l0ETAvgW6DByHz5bpHEx/zxj380w5BKvo888ggAYKTgJvz5z+bheca0aQCAadNmipyoAB4Oo/Cu3r80t2iftM7FY8AMHEWTVNwU0RnCY2bu/FoZUy/qTjvWaipUlio7ry1mC7CqKivxEnGZ2RPUC8YoIGUfLrpqU8dU2PF4VGe2D3kRyj1DjH1xzdFJMnbsOADAj350k2FF0EJ7Ro0CAMyaMUMXQV2qGnPex02QGBXJxEmJpCu8fO/67wIA9n6Z0XfW9rL649qgy5aZujxVwY5oNuqIpK/b//wnANYD9No3vMGMQ7J1CkWwLRxeRvWYcTBLhK/zn/oHAOAHP/oRAOCd73i3uX5UqbqypxNrNpg9prvTzaBaL2jOX/v61wEABx1wIABg/wPmlFmya0/uozyJ8Kt7SzPFbb/6NQBggYz3tNPeKLIzvIwRr9PtfzJy2Ulk+/TTTwMA5uxvvJbcc9n25y//bwDAjtvvBACYPsOswb333N1cXxS67pYsMb8himrdJNqrGdA3v3mteS/ejLe85QxIZ5DGnHFzE2lQj4iaLHv0ir7l+NKXrwYAnPduMx+MmaM+E8uIfaxYbrx69HZRPqymfMklnwQADYfYYw+JxWpKO7qGc3Bhdkp8jyLQyvi/+d3vAAAW/MOg4V4gv6HDulhl2o0DssFZwrLuG1zn5m2j2bT7t3qQ6H3dPPk5m/xB5dhjj9W/99hjD8yZMwczZ87E97//fXR1dT3HnQNTR0eH/qAHChQoUKBAgV46tNlxVHp6erDDDjvgySefxFFHHYVGo4He3l7Hq7JkyZK2MS2bg4rInndFguTIJ1M+DRJpsskcfz4kieXJGh/NJrNgTC58XJWaGmBcgXldIefpnR2dGCFR/JGgmFIOTWJTVKQPxo2kxG4Qi1ws7v5+4rDI03RunvoXC1rmmJGj5XoZY960eAJaf4JnufI2c+s3EGVw3HhjxT766F8BAHvuvRevEF75lC9YEOItqoh8u7q7Nc6hZ4zha4XEvxRylk1MC40bYB0SsWpyQRMlngKtvv5+c10H40eIo0IDJK+rpcl57V1t4nfWrjWW5fx5xqq78+67AADveue7AAD33ncfAGC01L44/TSD8tsl+DBx4nocaIn95KabAQC//M2v8eY3GavtF7/6OQBgh+1MHMzpp7/Z8JQQD8ONYqCVs15wNa75xv8CAKbMMN4sxs38+Cc/AwCMHWXkeuobXm/bVZwcw+DfHjeZdz/7meHl7nvuBgBc8cXLAQBXXW2sw+GSebK/9HHoIYcAAGqCBdIQj0u1g/gKZg5u+K7xSOy7z37qnXriCYN78tazzjZ80qvhxc/QofLo3wzexi9++QsAwArx4l1yyaUAgN/98Q+mnXFGJ489+lUAgHv/YuYqShKLA8O4IFnXG9YZ70tHp+GbtVu+8lVTw+rVJ5wIAPjt734LAGg2jPz22HtPAKX4Gc1Eq4g8JOsijnDzLf8HABgxwshw7z3NvZ/61P8DAFz8oQ8bHjoEV0UGvkxiNnbcwWTBUS8eeeRhAMDtf/gdAODcc88FAPzX574AAHjDG4znYqedd0KcMbCDCMmyh8jaWbfB6OcXvngFAOD973s/AOAm8dIMlz1j112MtyNh5pLsVat7zbrp7mLWn6Ef3vgjjJQaW4ccdgQA4L8vNzp13rnvdeRRkXFxv6cBuoFeDFlTP7/VyPG1r38dAGD5M0Y+v/qF0QvisXTWaqjVZG+QPWb6DONtJ6r3WvFWE7l3xx12AABccaXR9zeddjoAYIx4Z+m15iayqs+MW7OGJObjxhtvxFlvPgsAcMv/mXXICvMnHH+CIyOIDMdI5lkmHmB6JP4snqbXnvpaAMBT882e9P0bvg8A+MQnPgkA6JI1iCxHvWH4GjdhgulCZJuKx7Ox3qzLd7/LeHu++Z1vAQAOPcis5933MBmmjI/KpGbdurUS36jIzfzdEw85YsQV13HAuK0WL/Mmos2Oo7J27VrMnTsXkydPxj777INqtYrf/OY3+v3jjz+O+fPn48ADD9zcrAQKFChQoECBhhhtco/Kv/3bv+HEE0/EzJkzsXDhQnziE59AkiQ47bTTMGrUKLztbW/DhRdeiDFjxmDkyJE4//zzceCBBw6Y8bM5iMi0qTzN8mzvpz81uf47CbbJaKma+VPBANhtd/MEOkLOISfIk+zLX24ymBgnsGSZefr//g03mPbFWs6zDO9/7/sAmAc4wD4FNxqMY3FjTzKxDheJp+T6b18HAFgsFthHPvIRAMCYHuOpmSbxFM2GnMOLVyiJbI2TQiyvu+41HoQf3XwzAODYY48BABx2+BEAgAXPmNoeRx1jrNaVK3uN/MRz1BBvBs+Efyx1hh588CEAQLdYWRdccAEaDaKZmuEpUqMXi0Nfwv2SifPTn/3CkdcnPvbvAKzV39EtKJwjTF8rJbNj7Ngx2tqqPsP39d+5HgCwRtp67FHjYTjjTcZTMkLiHO68604AwDX/ew0AYPXqVQCATwo+0D77GE/D+nXGczRr1iwAFhtm3/32BQBsN3s2vv61rwIALv3UJwEAq1aYtj5+yScAALsJHggxXRY8PQ8A8JrXngwA+M/Pfw4AcN57zgMA/PgnNwMAvv71/wEA/NtFFwMA5j9l7rvssk8b+YwYoR6k/efsDwC4527jQfnQhz4IAPie6OeNN5t4j4svMtY+vX4/+5mxEr/85S8DAJqsUrvS4Ie8693vBAB8TcZ42GGHAwCuvPoqnHCCsSjfIx6Az332PwEAp0qsxnDxBP5IdObNZ54JAPj5r24FALzlzW8xPPzkJ0YOn/ssAGDiBIMJ8rkv/BcAWyG3U+KKvvzlL9tMIvGudch651k+PS6/+93vAQAnveYkAMA222wDANhutrHWb7zhBwCA7/7AyInZFTOmzwAALF1q1mCHeJaiOFEP0MknmzbpGXzPeUYOl4kciGg6bryxsNPMvJ85y2QO0gt21113O3KMhYePSyzDV79iZH/zj2/GGIlvicVLMWLYcOHTeC9/e5sxEN/+9rcDAEaKvp911llmvDeZumM33WSs+OmzjDw2SCzTgqdNUsT22xv5MLZn9eo+vO5U4/ngOn7bW98KAPjvK4xn5c8CUfGyl+0DwHhZAWis4bBu10vz2KN/AwCccILZk2bP3B4AME14+u4N3wNgdPG9IttKpea0QU/qjT8y43r1sUYnJ0w1yMR7SizSe88za2unnY0365STzdqjPnCfXyhYV9OmGp77mw1MEpTjM043e8hXvvoVAMBHPvZRAMDJJ50EANh/v/0d+VAHK5KB9teHjbf63PeYsey5114AgDkvM3vJu883FexnSYzOB//t37Cy1+wlWole2v7Zz826fblk1jIr7X3iQTtfxjtFeD/80EOERxOjtEqyWateuAW9ntVqTX9f/ikcs42gTf6g8swzz+C0007DihUrMH78eBxyyCG48847MV4W5H//938jjmOceuqpDuBboECBAgUKFCiQT5v8QeV73/vec37f2dmJq6++GlfLefi/moqoFVeAXo0zzzTxBESJrFWNBTb7gu31XgBad+Opp8w54hN/N1k/99xlPBSTp5rz1zPFKhw9zjyRP/7oY/jq174GAFgj56avfs2rAQC9vb0AgCuvugoAsGGt+b7SaSyo2TuYqPdz3vYOAMAwOfP9xv+a8/VOeTLv7DaehdnbG57vFSt6l912wi9vNpH5cx8zOCE77mosiP8n5+df+YqxnB980DzdDx9p2uqWKP+FzxqLYq3EoNx0o7HEV0qV5cMOPRgAcLJYqHOfNtHmV111lXpSdhWvVEfNPK3/4lZjQS8Qj8DqtUYOM8VL8dGPGstkbZ+Rx3/9l7Gkd5htYj0YN7G7eCZ++MMfAgBGiodp3lP/wLBu8/cJx5sYhGu+aTwlxHQZP97MD7OiiJI8rMtYfSOkRsZn/9NYwwueNd4tIpL+7rcmpoHxNLvvbmqhPP2Pp9CQNseKd25sj7Fu/vMznwEA/GOuyeaYv2CBjNtYce87z3jePvxh4zGZKFZdn8TVHHPU0U67I8Qi3UXk0Gg2tGLxz39pZHzYoYcJnwBgvVIrxUNCBE6uj+OOOw4AcOyrTF+s3UQvxxVXmArYH/uY8erRA3n55VfigDnGOusWy/mjFxtvzV1/eQAA8HfxZr36BKP/1D2uGVqxmXgYJ0psy/YzjTfDhx09cH/T3+hRw7Vm06yZxvJnrAXjvL4vsTQLFxuPyPHHGh5sfSLz+tpTTwIAvO5Np5lP65Lp4GFFXHWVifl44oknVKd8YkzGpZ+8VMZFXA3pUbKAHn9irjOwilajJT6JecdKv++WeKoChdYPY0wdPbzfus5kf6wRz+KkCeNKLVoeTj35NQCAU159PABgWa9Zc+NkXhnbRk/x0iXGezRt2vQWJNXRUhX9Ex//OADgvz5rPIN/fdjE3Gwn+9M4iTXi/vCrX/4SAHDc8cejHU2dYlCy332O8eat6u3Fl79i9tSdxSOyaKHxbDIGad0G40EdO360M+5uiRP6osRodcqe9J//adbmBR/4AICS91c84zf8wHhoXv/a15aQpU2r7+R8yF7w9W+YzKrxY43+rpYq741+s4cyDm63XSS7x8M2mjzZeD2+da2JL1ksXturrroKx4vXcpisMdbqmicxdyfKfqdeeuH2Pz9j9jEiDF9++RcBANtsY+Zk9nbmNVZgFk05FN5cxKR/BYVaP4ECBQoUKFCgLZaGdPXkwVBU2Cfk3H1Q1CdPxoWkMOfGtqoqr5Oqs4LDMma0Oeu94ML3AwAqfBLViq7mddttZ+Gd7zDnwxd+wOS0M8/+lJPMuei6tQZ1t1vO05mvXh3OSsQumuQ75Lz5W9/6JgCgfwOxT0y2wTe/bT6/+847cMrrTET5Ka8+yTTB8cs5+jvfZtp66G/G6vnxj01sTi5VlV/5iiNMX9caj8Qe4jl4o7RbaDVd87r9dua8/b3nnYtLJWtjrGTQHCQW97wF5tz78IONN4ZR7ZyLatVYccPFq/HRjxgPw3e+Y6zEVat7AQAzxNI+6CAT60Qr47Unn4ROrwrsbmL1Hn2UAR5sShYL+14v2SFauVfO4pkMME1wZGixvp7jJwaEyHPihAmYL+f6xCSg/ZHkZulNnWws7SlePZZ99jHn58xEy8Qz8/BDxtt1xCFGXqxczD45p5U40vk4/jgDGcC4CFbZHTnMyGjVqhXSh2ReeEizWsNIrP6Xv/wIAMDeL9sLANDNWjeyoIZ1d2MY40GkLy6Jww45CADQ7Hcrb3/0Qxc5/GeagWY+6BLvFivc0nvA6yj7HbafrQu7JhVnm1I9majO69cbC7tTqokTw4SZdiUmpB1i88jac8GjMWac8TgsWbwIk8eb+WTsjOLAyPus6VZ7tkIWGRPZNnczEbnuE8keYeyZ4vDAVtom8i5xcbrEc7BuzWqn70jm29Y6YsaZGf+kicarpTgiucv7Eol9mT59KjLi/nAPZWVxYnJI22edabLdviuYNY9KDbgDDzR6MXeu8Si98uUvN/exOrby5nqzRg4fhte/9lQAwDvfbbwZ++xr4jp+8+tfAQBec9wJIh/RqZRrSjInxQNH+Z17rokH+Y//+A8AQH8/sU4azuv4cWO1Db/+Gz0s73jH2wAAX/+q8fo89FezfreV/eruO40X/q1nv1XGy2rq0qxMbyJ6wf3z7DPPxLveY7J5GCN5t8QzHXHYoaYt3RtMI0lm9lJmRLKvc99l5PbpT18GABghmajTpcbbYkF/niCxLnmW2vm2sMbmu3zzeFiCRyVQoECBAgUKtMXSS86jAgCZeCmSLsnGIfwA63LIa8q8cdaIYAOCozBhvHnCHC+1cLQ2iFxWSJZNk5ZLXOhZ7GLJGLB1QsxT73DxnPC5NE6IiSBtyOMw4wxoDHV3m3iDdevWyn3m87PfepaO2y+O69Y9gtYw2mabWQCAO+++02lrD0GD3HPvPdwbvbKz7IY8x0mEZcvNeGviIenqNFbMTjvKeagwlxOqQrAKaL34gIfENlm/yLRLC3v33Yy3RKMNihzNzG2D4yFmS0Vk2i0eFeK/qHXE4i9EphTcEFvjRvSmIgitjX4OAhOJD5S4usEsD5b8sMC0LraNTlnhjnuiWO5pWmdX5lUsmgiRtXLEsuK4aZ6Ml1iFlJg1KhfGQwjPBdF0BT+lat6PHdtTZk09L3EUI9bxEsNG9FgssSRm9oNCeEobhSMfIvGy3tQ68aCpRcc6Lbw/srrTkHlQ+BPhZZVkcY1LzNrT+kp8kXVcCLZPWlq/gF17yroo1vLlS7HPni8zX7G+isR1ZAXrx4h1yz50+IKnI33lxEKR8dFLkhZ+fSnLCuvr0DMWSxu2OLTggCgmk9xLvebnEfE4zHwnpT4A6NwulczDfQ/YR8uaZVqbzKyhtEkPgfl8mHhG3/5240H4qeAA9Yt1P2yk1BGi2S9MqhdIa4HZfXD6DGP506NNNOBemecpUwRJO3I9Y1qbi/Mp4yQy9etfZzKZ6DlatNhkQW6/rYkjS/Om7pla9ZlrjpWAxCP0NtmHP/kfxrNcE28eY7g4Xt3m1PVqXjinlPO4caNxpaD23nGnqQa/SOrF7bOPydKkXqtXjmuPMFrUB/ng9DcbPJnrrjNxptE0I7cf/+QWAMC73vEOaadu1wyJ2EIVt8bTpqLgUQkUKFCgQIECbbH0kvSoEEclFyuP5+cx68TIk2ZV0E4LL0aF5+TM2y/0SZ1P1zyYFYu901jqaTNVq3wfwRMYPspYEERm1Sq5NLGl70Ss/VzOuOFZObMF2+DxJyVrQN0GtDwyJBLlrSV+PBTUXCo6j5S8fGKTEOmS1m5NouObqXtYH9PaE6s+lrPxLM2wrSCyTptucAASkV1Gi0SY0KqhMu6KjDttuKi5s3c0WVDPikeF+BJFKpZJYeegViUfzNoQtRdvFb0Q22xjZEjcCHrIEvVyuFHvFhNBSORaFW9Rs5li5512kqZi55pY4jqyJpGHxQKTxqijNuPAfDFhsollGT6yW65z5VOUPBUcF/FzNNZKWJkkmCS77LSLKxePmJFS6TDrQa1khTROnX62m719qYK2XMo10+nPp7tmaKglMme77GyyIcYKvMFf7jXWY0V0kHEYtLhj2Jo3NcFqYa0e0pgxxpNUFas2EQTTzLfaiUEkWTS69iJ33qm769f1axaaVu1mjIrWFXO9cyTuKVwXxOwYM5YYGealknAtig5SYSIbQ1StkV8XmbhSdT2FukcUtOZzMmN4aZkrjtdcT8yYjs5uFNIW1x3j3qpdNbJnXmPyb96f+GrJ+vq6ieEgllGs7j/Jjkqom66rMYqsDI6TWJTrrzdYU31r6F2mu828xtx7uH9pV6Yvrr29BMtk5SqTqfP1r5sMnT13M57larVT14JfLZs6oV0Lj/9xqYl7+YEgztZTt15WQbRvaYBeb+7dGvtU2Iwg3svYyg7JAFSME+mjIjV+yLPquVy3/bZm/yM2140/NDhCXC+MP6l0dtk9QCjydGdT05B+UImiyML8vkAqolLZ+oFKyvNHk67Uwm4G5nv+2vDHh8opP/yR556VdqIi00YuvMCkvlUE/pnXeJlg0IXKNsiClxZGOORdJMDVlmovXSfjZQBvRnelV5yKP6oEGYtYNt0rUe6XP6f81EUcER6+wAc+8EGRAV23ifPe/mBRhtKG/ni66Zw7zDYw2NsKQJbKD+6PTRGV5lH5pS9cfd8AgL32NgGsTf6wefqgPBNa3AuIVNc6N9VGiosuukj6dH80CMaWK7/uQ2MUUcacG/P6/vPPBwB0yPGLL592pA8gmftQwGOk06Q0AAsl+sSHDjt+t4y88izu60996lMaDE1ZaZl7fz51bfEhQMYlc3aAFMyjrB+8xzyo2HMsjomU2/UrfcUS/UpDokseuAhgR92xA3EDQe33rq7Ce1erdijUgX+NXb88GnDXCvWWDxn/+IdJW99uWwLAaUMyJik6ynIdpYsie7ZpXlisUEAlrbA8a2cAnq1uyb7BBxkxrkaMGKlHHNRjXWn8pba7grwT3WFbPCLTH1HvaIzFHL1gVcM1jzhYXsP0MXLkKGdU3PfsuPwfVe5NoqMRH3DNfesEUmL6NAP4lsQJMm/dFdqH+8qjP+41tm6te3wMXS9s0N0v4e0XZcp0Drhf++tT5NPMnO8jb+11yrwukfT9nQXyoKxfke6NXCTu8alP5d/pjf3NBsLRT6BAgQIFChRoC6Yh7VEpimKjIXydGKBiYCvUIT4tqtXrWpQ2rc+1EvMWa8+2YS3IyP284FNxUf4U/jMovTc5n9DFmu3qlFTcer9zY9GS2FduXZ7SvYzJww47WIbnjTvKnPe+QdY6I3nJw+Edl9hQTOcOvw3KVIMRxcBg4TFr9bnWgsug5cd8LMcGanm4beh4vAFaa8/13liubTt+12rlqKxy7x7Xw+DrWqccIeRe+qevi4jKfFMJOF7xCPGIxAsWz/15beExcz63XiAbMFwUrjIVGnTotdEyN97Mq9fD1feohSdrYfpB40oaNUlrvv286dpz43xLtrPbTlO8ABuadbt+I/fKyLnDflN4fU+Uoz16kAhzP9A2V5TmXYuhqlfSvC5eZo6RJrEgpJ6eeDqka9MVoNVvpiWb9irqHcu1Ld/Ta9sged444XGMFCvNW9bFQGTngG0U3vyuF6gG6BGta5dH3jzavcPXKfO+b00feyy14srG7g2+p8RL15ZPWcjUkrv2NJC/ZW8qcSEXL1y4UL7z9xJ3v/P3SDv97twxCWSEHDMXUZkXzyvj9thCjgdsELD7waMSKFCgQIECBdpiaUh7VAZFBdCUp9hIA9wM5Z5loQGKtJJo3WlZbAkoaooFJueQTP+j5cL+8jxHnBMgSJ5mJYCLuE0RAZ/IrnyfShtF6j9huxZmk+fQDBRjGmReIFcQKlrUco0XRMg+C7ahZqHIpc4+XF5JNKZpLURZZq2XiLe68RBJ4cqMjbIN5YkyFi9A7sUNFKlnqeaFFkRkG5QDg+foncgUCcm15qx+iKwZg5Jn7nWFOwdRntu0Si8QV4Nkc/eVxHREDeT2vFX0BjRKugVY/UFR8j5QZ0TmkeSAxxKzQv1MM+o5z9PdeACmN+pc5K41lRLULC1svJYXi9Cgpy9zvTEkvqeXIvZg7RWUsUnwutThJY5jXVsqG8/jyXFRZjZYlOBihXMfgQJVxjlTjc1dSxYbD8PUqTNUxrSoobITHdQlJ3NBK5XB1DKeZ+ebkgpHCpgXdVLHT/1gfEaRqT5zr4DM+4Qe461YL6UvInVbuGBlqlvyB+egaJJX0/cEgb1vNErrKHUtbG6m1C3d77wAfIKsMTB3tMA1WCA7Ga+nk443SGRHLzKDPllEkrzk9CgShFDXluuNTJn6nnG+zWtdYtfGSrJBs5m2JD/Y3ZmxJYxjdGPUGIw6QspYMEg29vZJm+bs7lnlPWpsj+Hn4b8+Ynjw9taB1p56uejMZzyN6OZjTz4BADhZSitQjmmzoUCk2gfTxb3yEpuKgkclUKBAgQIFCrTF0pD2qAw260cfewXYy6aU0eMgT8fypE2wKoKRMfVOz+ckYIIAS4wviDzLM45jFLGk6xECXJ7yEz55e9k6NtZb3ieMfmeWCy3WuvOe6a+kOEqtY0QzT8SjxKQQgspp9qJ8T/AuGbdaCxU39VC9PBp9Lt9UqrBmm1g1Mn7tQ7MBRPaEu6bVXmEWjKQ38oxU2skZZ1GVTA+mmkcNm27JFEpCvkt8S0SIfFqaBHwTXhK1auiZYM4gAbVEPmKJaJprXNHpjBljkdJrZd7HMgd5zPkU7w7B01Q+tCRFbjRoqKPMsFI9ApDRoyDXVqVsu1rxosd5e93S9MWsLjxJxonMO9Qb5Hk1olgtLM1SyekBkHEk1B33HJ0pwPq5yoWCpCdNxsQ1TG9YkdpkLcqYoGECQrZOSiSMGjnC3MsMLE2plnnPXA9LrIiBXFvm86qAsS1bvgyouOuOa6Zl7VEH2WMhc8X1LPIbMWqkjI/AZ7TURZc1JiIpee2kL1kznK9uKSmg46kYfbD5JsKlet6kL01rNi8dkva9fsMK014UAzIeesC4l+hEcY148qnInlkXDwNT5iPVRbmw4a4XJ+srp4y5XwvAmfRdMPNIvK2K76frXIjrhQUjufZk/1+zTjxS6qXOLVBfwuxNrm/ZKznx+ltjPli71qROj5Q1ybRz/Y3x5kL3JsqvSNXT0y0AdSy7Qhnr75P+XsH53s8gyhvMzDNjWbPaxOR0dAxzxp0XQMS9hG1yPx4gPTlk/QQKFChQoECBtloa0h6VwWT9AEBNAH08o1QfOfnkSIAkQu6XsKgB2DLZOfGvbCqHaUdMEAJLNUvAU7y2QqsHhMpvj1GhAGAEnfKKP9FIrgk4VaNOPAK2ViggFx/b/UJ5vJqWo8YiwKUO9tHf/nueyxKkqJk2beIU+aTscvfp2ub2i+ykYBhBu/wMIxZp01IDTXrDbJsqf8YieLxwvFWxChuN9tlgSdUFncoz13qgpUagu3SDC4pU7iuR+CiNPfLiPSodHmgXrcPIlXZF5NOUuCELa1AoFklcIZw5LSa3L5Vxw/ViqTUrc1GrunIkz6rLMqdplmshR83WktcOAahr9NPDQhIrPyEoFwGuPEwieeUabjYYZ2TXi14j817vd2Xzj6cMRskBBx0ofYpVzLN9z+CjfNIB9GeiFJRMkg4rf3+eRHdytbh1k3DaqooeEwZevSQ0qBXPnpD7mf2+oIdHZNeQWB2uHXEgKNaLF5sCb+1V/LXnLfQOmYMkjpGlrheavLANXSnu17pH9vWtcb9Q81+8JDImKmU5m4ptdBAEUKz6pgdKxnnmOm4y9sbbwTpk/TYa7l7Su6rXXC/eu0qtpsU1dc/nvkQMH2I4eXsFeZsgJTboDfJCVHRsCvjW4HwXNktHXvr6JCuptBYAdVYiEcC3rN/HSxIZ02tNHJUlpnTAyJEjnes6ah1o1L1YS6+UgE8h6ydQoECBAgUKtNXSkPaoDDZGZUCcBduy/N999T0PkR9X8QJa9dtA4UHdl/hs24Y+RPN6Pj3zmdN79nQaZCyO9513ZwsPyjQtsfZosYrP4LUfFeWuGP/god22MCzvPLnYq2Ln1XLvt1e2TgYijsdH4nRv9NEjfWIchY1VSjGQVrS0oJYzocgZi+PGQ+kcMo5I46vcySqiGLF4zgjnHvnzpW35uufJ3NNRy3J7fYrKdpXG77AMgY9l4ZO/9sqjsuMsvPl25KzhMe11Z7zgiYyR7I3n30LccbbOqOFlw/r+F96G757Q8RhZr19r4iHGjTUZNgXdtrqmvLlAOc6Lemys3meeMRlEO+6ys2mrYDyF06T9wAszaNV6L8bN+cabv5Y58DYJeVm5cqUZ77jxXt/t92CnvciNX2uIt2LBggVtx6HvtTF+w72J7ble3H7JciLybVT6Tp1sHvp3i26J7J9aYDxm+x90iDs+12mPuGBsi1dSoM3KsfrMRtgKvTvidY5cr33LmpOPGS80SrB8oLEusR23rjWnqRYKMSqBAgUKFChQoK2WhrRHZbDU8lTuWSlaEJC4Gd4DoD5NemdtLQ/PhfuH8wTuB1sUrhnT8szZ0pcXR+Bd1/YB1yui13Kr9/mAz72eXPz4Exuj4/JS/rLw2vCrhlvQ0FYPictK+/GWeXu+a3yE0lbZ84WYKPRUDMTUwF6dlngA7/sWIXq82svcw/6opYOi5drIG7gtquiP233vF9JrHUvkfd+KgmzZas8v3+beemj1pLEZ25d7XWT13GuDa2WtFKtjTAOv9yEgLM9uO77nidRM01IbrgcAA/LrttklRQ3Xrl3T9jqrHgOvC1/fqd/Du7vb9qnvB4pZadMHADz11Dz9/vn2Cu2jpRPzx5IlBoumS4s6+pe58TOuervzTQ/AmDU9AAa2xgdav7qvy/fDBeukU4o02jkonve3wfcM8npm6IwWjJuBnQzPt84trZWsJOVFRdV+/bb05H2/dOnStp8D1kVuHYOt87IpKXhUAgUKFChQoEBbLA1pj8pgsn6iwmZIxDV5Eufjmp4fm9dGY725x38qlldF6vRQGfX8UMyrZn298msL8rpR3qnk8BPzwdZhkNdGv7YBAHHunmlyCE3hmZVyFc8ChWY1KE6Ax4utCsxS4szl91AVG3Y85hUOEWclr2/Qz6whJd81Pdl746EIG5Sdh20Ta+Vr815RgmXcrHhdRJHOo4Um4LwJ/oIiUjLDxvVuZdY0dXhTjBfPRknrRBHNrIzUxRMJD8xEcbO4GHOSNjdIG64lbs/CzV/1+jpHDhboNLKy9lFw6TkUvhsyT9TjyOOVfaeNDY58fK9YyjPsNBswDqrZ7/ELZ/jaWDM1+h63BErJ/Mp60LGVLW5po+HzK0ytWdMLABgtlrciVUfuuHg9x61zoUie5rUq2SRplqKgbnnuGUXSJQIxs6LkupgxW4ngbIh1zMwyG3aQiXzqzn2RYdDhl0LdIDInnIjiRVEPqPZeHS62QxlTX4YNM94AZkMVWdOiPMPVd5vpSF10sxGJyDp1isl+mTTJxKhkuZuxo+jg/CCyvEe6V5hrhg8zno9EKoznioJseEhknefefEL13I17qkk79IIQ+wrNft07fEcRX3Pvj0LiRKhjIwUnRysfx0SHlnZkbHnKSbLYXNQdtjFs2DD5jjJ2PUNce5qJR151P5DrcmYkTdS+zPeyN9U3lHRCiDhWSXvfR8j6CRQoUKBAgQJttTSkPSqDyfoBYOsvJDxHplXDuhJEQUyc+3gdnyYrPOuPXVRCeyQo7UheepHl6kHIFHnVtEE8kNx72iwU5ZQYKGId0KGgT9yZ01eu1ZnlpYS7wgjywgsMKTzLKbc3OLwwp9+vT6Nn2sSIqJXG7RnG1vKgDOlRkW9oSbKvgpgHcp+fmUAEU9bcYHXpNEUiCJzqxSLqJa1edZnIuL3jVo239/BFbGYDL5DrK0RCzdWrphYxUXATypiv8rn0UWUbauXL/V79HiLxshaIxZeIrCVEHeF3rHUjliGr4OaJpw+cG9Et4ixQlegd4Kti48Sx9TZp56Kf1HMfw0fGRRwhrtHcW9/0lPF7jS+iN6AodI1wLaQqQ9MWEUahmC0ukmlLHIDuA252H+dijNSn+fvceRYhmfNN7yXbYMZN4uoScTaSyHgpxk+cIONXYcv9wjNRkUuYGcqutxaoz+PGi7dCZSzZLX5wDj0QVaIBu+LoGW2yS5YtX6ZS0X2Ieso9kl4NWvFc55w24fXIo48GAFQ7uFadru3c6HoiSnBRQlY245k0dSoAoHOEiS3RvUdF6MqHAXL0WsSxuyZJk+hhKHmNrC6QX+plXu7Sjkfm8+hjzHhHj2aMCvc/d+/lvp0r9I14ZJEjknmaNnUaAOCII45w2rAljNx9m551G4vDteeiYRMfSOc2K60r1QlXRlnqY7TAGZ//9wul4FEJFChQoECBAm2xNKQ9KoOJUSkiiyibMv+eD5aKrmceX2sVQRhtuoiVfI6OvZoJhXo7WOvH3FCpmnaaWb2ljcRDptU4AmmTXp6KoKZmcg5rkQA5LnNdtWbOURv9LqZDhAKJoH0SklfPWr02aOXQKWPjRsiLGU/OKsr6gCxeD7EqK+LJaGbNcji8GbdYa5me8XI8jCcwrzXpqy4xC3a8Mgd6NiqWuniqElo2captEGGT9TkUf0Gs24ogr3JcjFngXFUoP1o1fhVS4anSIWfZ2Qaol4kZQ0TBVdlIrIqHL5KI7jHmgZ6WmJMn7ejYRD5qBcYRYrHeqWOMxaAB7cu4kbkVeZViyqekx23aoXyiZo6sBe9BrqF+ZjKfatW53h2uLa2OTFZYO0jmKs6Jksv4oRSckZr0VaQSayGsjBpprNgZM2YY+bDekjVj5cXwVOW4iULqY6AIT8O6uhERQ8eb90TGozWoPLyfPDLjpA6y0nHsWdqJVu4WZFrWWInsvLGNpnghJ42fDADo6ekxfSb0+lQdHtQlGNGrJwjUEg9j/Y7mr9GC2xEnVWT0DNMrzfEIoqziisCVD71VxxxznLnPQ6AupJ0a9YFevnJ8lKefO++4C8pUsD4WUaGlzlQz87BpFA1bdLThIhp/4N8ucNqpVjvRkLWgfVE/E9dbRU8D95RDDtgfANAhcSX1DdzfWD9MxCOvrA3VzMtr2MwX8VP2edk+Rkb0Kuu4uPZEj6Uv1XPVUXr7DA8fuejDZfGol7BS69TYIzsPXnybRyFGJVCgQIECBQq01dKQ9qgMJkalbAzFLVgkbpaDRpZ7XhDrYaCVayhXK8qNdbBPpJFitMReGy08yXs/e8Dig7j32ZI5Lk+a3t7mjNBvy4/NiL2Ym5QxLRG9PR6vOib3+TcujTH32/DwJXL/8Z1teJa5vo3d+2Nth0/6lhc7bvdaemcKz4NSutG5n53HnldLx4aSfDxrzUcg5sdJ4euWa3FbXnXSzCu7pCVXiv4hoKzqqaePA4FARF42T6Yi9ORDq8+PcYiiEp6IG5vgyKY8Do33cmVt0W+9eed8y/25t2ZddkQ20sY5b38bAKBW4/bnr0lZvy1rj7xQ783r6DE9AIARo0aUxoXnacMdh8a3yeUzZ81whzNAOzoHKKwee2vhnee8A4DNgmGMSqzz6e1Xsbd+2ae0x9ov9EzHRVGK22jfRuTpp16vmUjentKyD1Bu7hqOigKZrsP2+qn3+vuS/kVeXJ20yaDm/czp0932S134fRbu9Ng9cID9zeLDUA4+k658zFxFJe5t29Y7LXd6+5KP5eMjbpOVSVKHyF/DBfKWeWzZ5zwKMSqBAgUKFChQoK2WhrRHZdA4Kk03z5yWRKIxGYJHwDormt3C/HvJdmhKHIRkYGRyJpxoPRLzOauu5nmOSDBKmHER672F15fwRuunIm2kfJJ1sVsg7abEayCmBS2QLEeRSN8SO5MpXor0oRkVMk5Gh2uVXDl/lrPbNHMzcfgkT9iVPCa2SQaLQSLXsPgmcSS8LB72ybnKyAvPm2mBa5iNOzd6sJzlWu2XsswEP6UQGaXEMGBiFTOMPJ6ID6MZRrl7ps+Uh7xZF15Sm/lFfBzOh8QxZYoDQoFIX2nkfG8xW+B83kwajnyiUtyRxcfJ3WuIWSNZDipjjdjnebO0I3EVWYtuFd73qfRTqD5SNszyier9Li/SI+OhNEtCUw2YsSNrjvISXlLGsMhcJFGs8TwcV65tmGunTZ0s45f5p3OqcOcxkzbTCsdNi5QZVua67bfZBgAwetQYlY1NtRK9bBJjhl1Qn705kjiiaTOmiZxkryIuUpI7nysOD3LV65ixFcwMo6eFlde5joWHVlRo2bdEtmyXczl1ipHfq44+xlxXwlHRrLuc2ERaWt5eC7v3Kn4I9YRZUmSG3lzGZqlXxPKuOEgN7mtuFXB6DBTfKSIWiVcBmPMtGD2Z7iXStULg2Hgo9s2NiPuYeoaZvcasN/lCdVT2pibnzo9d4p4se5PNFs1VX1PdIyl7tyo010xaZ9/c9+GMD17fnFM/BjNvNpCm7rwy7mmgX+MQoxIoUKBAgQIF2mppSHtUBoWjEtknQ0Zm6xlsRCwDeToUK0bxVOSxNk6IkyEfM8slktohzHBQbAxpN45RSIZBotVk+RQr+AhVFya3IhkbxBGIFfmP9xfO/SktS2ao8Aw7KZRffsYsDaIlKuqh3iPfE3ZGWkg5Hs2CYbMiVxoZfHBOKmqN8Sya1goxXaKKWw23UrjWgGY7EANGQRGYRRE5PGstlDiyOAIyjzGzU+SVyKKaWRNTdu5Zv6LnFpQPeZL3MvAmvWBJtaWNgm2QJ+JhcP4ZF6I6KrxQBemhIRaMMFPRdiymD609Ylsw84QKH4kLkai/1DXbhrmObgB6ypiBoJk2Xm2sOIoBT/8S6SvzeLFhBNIGPQeEuKD3StYksSAsngrXJDP4clToGSO2EDFXIPPO+AkbAOK8ZRZYUvh6wfuYyWI+HzZiBACga9jIljpQ6p3kMJm9w8rDGssgHlFZFxd98MMlzsz6BcpZFqIvMeesQCJ7A/1iScJ5dHFgIjpIRLc0e4vZH5wKYrvIfFOXR44073eTasyIYkDWr+qn4hi52SB2zxAZy/zRdld9V9tc2iPWkRYHY8ZSgYriX7Ev2derXlybOqvpBfAytGQy8sJbexr+JPKQyuZ5niv/ha9DIMaQvE9YBVnWK7cU2Stqutd6bg56MZgtKeumKDIbQ1O4soTGkkDukbXnZav63jx66+mB4j5pr8uUF91L4MrGxhLC+z7EqAQKFChQoECBtlIa0h6VQeGowFZNTWn92HBp87lYAZ0dgheSup4Gns9VuwRFMaVJ4j4NE3+lY5ipWmpiVfjU7lp7SUTPCW2v3O1L8u3zJs+C2ZNpj7U/OrsMGiPPPvnsmmYpKoL6qOemPFfmGW7uYrbAQw/lmXXXCMNLs+4GcaiXg/gTUgm12bTYFlrRVrEcKrzZoZTjEdk1dDxsp3Duo3dIa2LouBvo6jJt8Dyc59y00nM5064JfoLW+ohcTwHnn3gDuQXekL6kHamFYmr+8F5aIy52Cc+XC3FbEduhprpFFlweqJM1kU9WbzrfFyX+qtR3Im96SJRd3Wa+G8QLgm3DXCd63E09duMrVC+kxkqWFdq31gcR3JSO7hFeG3DaonwqNcHNyHwZm/sUC4kxW4oanOs8d3ZTd1hvhuNx+0DBtediGDFmh/OZN70aOcJzU/BnOjuGqyxIun5rrjcr92KOmsozddCND2MtMK5hRZdOLWYG962atJE16SmQcXs4qfTK2BUjfXE9SG0bxvroGlbsFnrianZtFYytMvd0DR/m9VFCbQawZnUvAGCVoAXPmjZLemIsHl1rYuXHLs8FCo2h4Xz/z9f/FwBw1plnuW0QH0bQnClzEnnuEBnbr+lS4XUi565OjT+0aS+MlTNvq7IvZeIpXNW7AgBw3z33AgCOPeHVhhetiUTvH9uR/UJ+izKpU1QgUt265Se3AABOOeW1Thv2d8jbK7SWkbtXWq8XPUZubTCuvc6uYbqX+Htj1oJWzstCjEqgQIECBQoUaCulIe1RGUyMShGhJcfdf8BjRslALduKpX7ZZdtHuR2XAflOLoq93P7I88r4lXnhfl3K03e9Qy1p+7CyaqmTQ554Nimfa8Q5P2dtEMYw+A2xvciNATBj8uMBmAUz0HjcM2k/C0DHyXgST07lMfk6YisMswvGHMi4NP7F7VNjFBQswWFRvUP6WjSsocU2VDYDjdtFKs4YcTCAfOzYXOu5iCLEEWXjxg/wmmQAGRe+7nA85Uq9pQvtXNnmWq7RIBu3cIy/Qlr02Ps+AWNWqINebZGiJJu4/bhiG/nhjpPzzfXr8+DxrO1pDEjUgo8Sw5UxPE8giW2ojFX36P1weWmB0SnK+ss23KwVcll4vPh7jO5rqru+Beyu4QhxS18t4yZ563fpUuNhePSJJwAAs6bOcq8fYC+1co7smhEZrurt9a4hK/SkRu2/h79+jW79/vY/AgCOOOQg6aeko94eQM9oErm/EYzvWLdmreFx+Qq5z5ePq4SJIh2z2A/XcKHjmT9/fukO+D9HpX2JbbiZVTZOzPy1ZOkSAMCokaMA2BpeiepFrHtJ629ne99HiFEJFChQoECBAm21NKQ9KoOJUXHudw3j0pPlANfr956l5renr76lUkJ/9W/2Px+Il8K/zreKovbtxC1GnG3Se7T2x5GXrJfyFz4Pz8myZv3wGonV8c0a30TxSeXGOXAttwFYdXtomXfe274NfTuAJWbfujwV5Qtdh1LrvD8f4x7PA1mF5dsir41W/XTvbeFJ30ZtP3/edkr3+Pabf4+/tnwaaE4KX35R6zWta6Z9Y+3WjNO3kM3gkdfSjQO14fPSolsD8OKjW6s3WK3mUoyAXtt+vtiGttkij6Lt577s/TVcRK0yabfunM+1URd/YyC9GEg+5TYG6tSX2UB67q9f0r333AMAOOiA/QAA1cTNdnTa8HTC32v4ft2G9c79/vxaFl0FKn9Pj2GdNddUz4vy25J+ttcLn4tHHnkEADBj+gwAwDaCE1Rewz6fz7XfAtjyYlRmzZqlRzLlf+eeey4A4Igjjmj57l3vetemZiNQoECBAgUKtBXQJveo3HPPPRZNE8DDDz+Mo446Cq973ev0s3e84x249NJL9X23RGtvLA221g+ryGr+uGeV8amw3lgv18G9QD5JFQFSMjqIU+Cd/TelnTzLFf1Vo6UzMwWMtPfrCRH9sqhvcFiIiTfh4ag06vKkTiTXUo2QVCrxMpMgUsROxhpIRD2RW4l1wuvk+6b0ociOHvHTtN9E8kdFYcdLTwrREbXmj7DtTWedfRGB1EN6zfmsLXNKTJDc2jCoN4zsFOVUMooglU+JUUPMiyJzx6VeH43J4eS4SI60xFQ+mUWmVY8PX1ntmFHyuTuulNH9HlIxR8X4oX7ON/vJrA5qNhdlJ+PKc9cb098U3dIKr9A2yuNqyHxqhW+vJhCr7BZpYfXcC/xobFgn/LoZZSTKJxP5JKrnLmZLRvRQZtmU5Sxt9HP9esjSClFBpF3N3qKMzStlDNEfZiBxbArtUZTmvaUmkaG0KfOduWvPxmCY+IEmdTV1M6vUAylzyCygko2rXTf61zt8Fr4Lgftd5lri1tKWuRJeCkUR5lwwBkZuaDYRZdwDeY3skaw4rftdv3NvmpMHIveKnDz5+dWmtZ5RYT9tSGX5BU8/ZdhixXWi3rLekOi7ZqZ5cUNNZsXIvrCqdyUAoH+drDXJZCqaGzR7M/JQcNkX21DMIdGhVZLtVJc1pfWXPK8YM24yZnuW0MKJb7V0+TL5jhlBzAB0J7YpfSleFpxhq770rTK89Ul2Z5FNcXhKG/2KUq0qpXhGmydGZZM/qIwfP955/5nPfAbbbbcdDj/8cP2su7u7VPDoxSBR2KRWflsCKnNhnW2AkAuHzPsSAicxeE+xq9xg0yRJEBVuanDuAX9Z96MofuZu5coTQco46QpzLffHHrhXXtggSV6T2CBAw7CbMhrFLiCS/2ASifz8wL+8BLZmPo9KC88FGVPQKY6bPyJewJcFVxNXMRe+RVRyr+OPSemhmeBhCr4Us28ufjifkyilVAHxZA40uJibkvcDHlftDwkvYc6jYjNJcGjsBsvpD1Tsu5nN50nq/yASlY9BpvZhUI/JCJLFoOCMm6w0VXGDMdl4Lg+4OldaQkAeCrnJEsQwznWAOu+in7bYoBcErfMlP4oUOnkqCGxGPeI64Hxz3ZR1Q/SWqfBswwvAjRO3bT40xwpPz/1CdEwr6aFlbDouLfHgPtwmsskoUJs+uHpB06qbfLDhwnB5sUd/sQZ/2gfMxHlvAd+E3ypB99w5KLx9i8BoVi/4QM/rYtUp/thp2QLvSIB7BvtatGiRjM/bk9m3vDJtP9I0ZRZEtDqjx0jyXaUicAwEn9N9W+bC0x0tS6B6I0GkDLaX9iqgHtmAVhtYzC5piLlrb/VaE0xbpK5uJb6MVY7t9QKwxp+CJCrongXDK/Oivy38bfDMhEz2lJXyEDUzmWWu51GXGoutgG9sy8IJbFrarMG0jUYD3/nOd/DWt77VeYq67rrrMG7cOOy22264+OKLsX79+udoxZzB9fX1Of8CBQoUKFCgQFs/bdZg2ptvvhm9vb0466yz9LPTTz8dM2fOxJQpU/DQQw/hoosuwuOPP46bbrppwHYuu+wyXHLJJS2fDwrwLQKqNQEH86x4LcYmT4fVKoGTWBCLjZiXpEpQLlo7tCLcI5NajQBpzVKAk1hWNVpOLsiOpg6KVZd0mOOxvO4WI6Q1TM9KTa5r9tfdsRUZKuSXpw3e0y8vpYdIAa7ULWN4rLIPBnEpCa9spybgVf2lNF35Nq66MM72VM2TXccIZzw2vVF6EUuEIEUZrcXSnHZUpQ0ZT0t6qlgpVQF6+tlPfwoAOPrIV6LcGCHm1bOWuV6NQiDXOzrNuBsb+kuueRfgr1IRwDceWdEikXF1CCCcFlBUd5gUGMvNfcNqo01fLQF1QBwZGcfilbCycd3Rq9cYQ6FnVI/ThkWYNzzXZD6bXl9a9kDSGPOsaJENG6t2CmiczmfhtKEAiGK90gP1l7/c7zRXEXC+vMEin9TlTI90a51m3hsb+l12CQ0g/NLLaQEBXTd9VUAUU4LqcUgsQkn5VLtsET6dCFkTspeg6Xod/UBfBR3UIxM7LiMXejFlTh2ArciRTdZgEcqi/LXOZ4udynIecn2lJnNFMDIVID2Tdu1lCsgo61e8O9WO4U7XkediiXLT1nI5boC/J0fSjnhabGo1XwrdlwiK1pTCl7pdE/qeXp8KwdM4V+64ap3STsOdy8gr1lmrdSlQm26zPF6KBZgvoVfSjGPdBikiKjd0yHpo9Lt7qbYj3hyWMaCcC6Sq5zbI3zvDo/cRlA+BKOvuVQQfTejdcT3tGhhc0i8CTNrNgvOOtvTPBtNu1geVb3zjGzj22GMxZcoU/eycc87Rv3fffXdMnjwZr3zlKzF37lxst912bdu5+OKLceGFF+r7vr4+TJ8+ffMxHihQoECBAgXaImizPag8/fTT+PWvf/2cnhIAmDNnDgDgySefHPBBpaOjQ2GNyzTYYFoSA7Tsc6gE6sknFS8YzVaBdz0RuYLfuFDNfvtFYUM8tYAW4wDgWVoe3LUPwlagPQ9sx2/fhlbageTwPSq+h0DapPXiWWY6bjbA2A2iXutI8rIQ5LvIGY+FYnZDvNTe9pHhFKyK6Y2e/AgTX+SlFNZcPyuP33orzMs13/omAODIo14hH/t9erE4ypmrLzly3HbbbQCAl+29BwBgdM8YlCnXs2j3zFp1qHB55PWX/Mf/AwB89nOfd74vB+/aYcXOuJOyNQrgU5/6FADgc5/9nPO5xhB7c1N48kvlfUdkr/fXiB97Y2XnrTFKQXXPvH7pa18FABxVinVz5VPmqb3uFN7aqHrfa0xZ5OpQURpXuR3tJ7efs40ocq/1C8j5+qxri2EHDPTnOtYgcrT0yVcbW+LpjnoEqJecX09/I/c6ks+r3UdoadtxF/6a4qW5y5MGNItXj8f/LXqjQbPunBYlnvy+liwxgGWZN1++3vvj0rgaX34M4fPkB0St+xJc7wOBPTNvblrk4//WeHqjPKK0LqTNhgS2Fr5Oadve75LHi93fnCGU9v/W36TWeea17b0lWyzg2zXXXIMJEybg+OOPf87rHnjgAQDA5MmTNxcrgQIFChQoUKAhSpvFo5LnOa655hqceeaZpXN9YO7cubj++utx3HHHYezYsXjooYdwwQUX4LDDDsMee+yx0f0MJkYlKoBUM0FcS5plwG3assSDaBoqgzvkXJyZNkwLk8/jwnoxgFKsR14gUkj01OlTR6FR3mLlybl5hXEyzGhgoAmtILHIyVMk7TBKvsgKW2RQ8io19VWDVqQtXkdrLvOe5pneyz5AVmiZyNB4xptlNkWQMtF4AGcYSoz7oeygBRTdzAb2FWvap41MN7y78jfDFRnJKzOD8qo5m31mwQK9V240L1pinme3bjyRXscsmTzFX//6IACgJpknhxx8qOmL4/HnQCiX9FNI2joV5NkFCwEAS5aalEQWRdNMF8ons9Ze4clGU7pFHk899RQAoL9fCgdW3RIIWmrem3dNqaZcJWvgjjvuwJrVqwAAR0qcDwsBZpKmy3HR8tS4GVqkzAqTYT37zDPSpSvjInPHH0eRFvxT3dE0UyOz2/94OwDg5JNPKn9tS0JIHIkWKdQCgdwH3Oy4+U8b3rbddltEOl/qdpQ2vfVeeGf8ct29998HANh+ljna7uyUWAcuF6bfM9ND9SZHTn1m6qjo2J133QEAOGh/4722MSuiK9b0ZlPmdu4DmlFovlgn3o9h3UxfLUrjdnWObcQ27c0ZNmXKvVPXr7CkWYC6OdO2trzT65RmzJzjPiT7MfWXMWmMn9L9q3DeZ6m0I3KqrzPpzJFABuTVzI6t8PZjIQt8lrjjkgwlwv4zjkjlFLuxHpruzVx4wlhk9rdvyWLjQVKvnAJEcBjSR5Prnt+7a06z95uUj/t7V5TWHjMB1SsTM96vvbdki4xR+fWvf4358+fjrW99q/N5rVbDr3/9a3zxi1/EunXrMH36dJx66qn42Mc+tjnYCBQoUKBAgQINcdosDypHH31026em6dOn4/e///0m62cwMSqIgIL4ETUX86C3z1iBX/rKVwAAnZJ5cfChhwAA5uxvIJSThGBj8jTJHPqKm7/Oom8Zrb0k1ghsZjcUMc9gZRyKF2I+r0lJ8lTkmUgGB4GACEYWe7gNsfBC6z+pFHrGmvAzxXJwi/HRNo/oDZMH8ar0pR6pSqX8Nfh0XfHibkw7PHN1vQ+5Yj5ETlvEBCB2CTE+NBJd3in4Fj0oIj89067EGkNBMKJEsj3If1Wi3HmOPIXHkAkxP9g3MVxkDIyO59TJ2Jq0mioVBUJau87gJ3D+6O2IK8RkkIwEaYvjVpwQciB6M3z4cK8dVy8iFCprOw/EUZHGRB6jRo0EAKxauRIAMHnqZG9ctOJFM4j5IMFIlYorv76+Pjz19NPOtUnF9eJEHgy54gNp7IV87hUWTGqSHUdLlngixKGJihbcm1j6XrXawBrMf9Z4QGxxUlk7bEuCGarSUqpZEYyHcufkC5d/AQBw5X9fYWVD55vco+BwIgZa1BqLIN/fc/fdAIAO8e7ttOOOMgZZe7yeAFsaE5GjIuOkTixbthwA8Ns/mKJ6Bx98sOGB3rzYXVOxBxBGj1KiGXrm/ZVf/jIA4KMXXWSuKwrVLa67SPjPPfBE7inMJKl0CB4H1xKxauR6jVnL3WKVhRbrK1BhDJaHk1IhPk5CNxbj4sQzpmtL9DhmPAn3HPP90pVGjlHN8GZ1MFO90ywd5crblzhu4iZV3L4qVe7FLo5K7LfDdRPnOl99a/pEdNK28mComrh6QV782CVi9ChMENeD8m5jeWI9KYmc72zhVpe22BiVQIECBQoUKFCgf5ZeckUJC1g8jFw9CYzZMK8HHWgsjxEjDA5DnZgkcj3jKxK1hv3cf57tytNyZxkTQiwHnsHTYiTiqh/vIBZJpVMwKlLX6qEnhufTxEpRTAieOxaZ4lxwnDz/Z1u5WlDCS+qeYTL+pdYtfTS8yHWeI4vFRmspTVPtg3EdtEBsGXOGmksbgjtQ7SSeCM9uadUyZkWsIcFZ4Nm9hZRvolrrlO8oDHUpSBtmXLRIiJ/Qt3o1AKCnx3gctHxB7M2VxlVIOzJXRROYMH4CAGDd2nXOPZWY2B1ebIZ43ypitenRt3S1YIEbq6EeGsZAKOZNobqTVCy+ieGW596m8dE9owEAy1eY0vNTpk5FmRgvQ0wMq1tiYUk/tS6jF8tWLEdf3xrpK3ZlQ4yPpiqfjM+NzaLXkl6eTsHIoDVMb6AtzUBK1atY7RBvpOjS4oXmLH+N8JboPIqsFeWYZ/Fce8RLIs9EhZUYgYXPmk8rFfWeWhnLuOnh9Mxdvs0Ezn7xYhODRMRi9QLo2nS9wIpKGsU6n5UOWvzmmuVLl8t7uZV6rL4n8qQASwCAak2wO0QOifxcPPTgQ+ZyaaeSVDUGSTMHKbuuLuGP698tQrhG1phmJ0b0uLjxFhoHx33Sph7p3lerel469TZx3PQoE4OG8ylrUeagSo+JG8qhr9w342pikWMVidnNiFOvtXYl65My5W8R16/KR+ZVYlKSGn9DqO+5zj312WajeSi53AfktyFvuvsAYy/pcVm3zuxVa9cSiFX2SZFPpdaFTPdZ1WC5xs/a4mX/XIxK8KgEChQoUKBAgbZYGtIelcHGqETeOZr/gMc2/Xoahb3AvLIdTS938/D1DJBn22lm+9IDcrdWBEq1aRyelGcvulybc3kqFGW3RIzMVgwTtziV7QxtP7eoi8ZySdHwr5D7xTpqEwGu+AIxYzXczvQOrx6R1kLx22HtJMboeMXLTBtufaDCGxmRd2lBTJhovCBNr0+NbNd4INfUUi+RnCOnjSbWrDGxKcO7O6UNjyevxotmwagl5mcWmdcJEwyP1NFMUYRL/NJjSJ3IXJ3gOCZMmggAWO+Vni8FijjtoPDngpeZ9rq6ujBzxkznOx0X4waaLKrneUS8vpqCHjpmjMGfUe8mYxp0vbj4HEApbid1M1J6e40VX+j6fe49hHEmaUrsCPf7DUSRTWKbhuaRzgFjcFowiAwPKbN69Dr5Wi+LnOvbdUfd6e3tBWDjH/xRDmTXal8sMOfVHtXkEnp94lg9xPDmMxogboJ/rBaPSkU8TgNZ2yo/3u4ouuiMF/fktxT5e2TLXitxYjGRy93vV61aBcBmO8VJYotiajyY5xnS4qleXxrD5K5fn3dtR+OJWgvBTpL165OVOeUjP/dNZtzRoyK8CK91Qa7t7V3lcyPtJFoXroUGWEohRiVQoECBAgUKtNXSkPaoDDZGJSr9Xf7LBxm013nncXrWGznf+9YC/PclXnv7jCXBp9fpU6cNyG+ZbH69/7nh5cYf3AgAeM1rThy4HW8cAz0FP/bYowCAHXbcWS6jRT1Arry8tjz9FsCateYcdZXEQWyzzbZoNxJfhlHLF+zDfPPzW38BADj21Sc615W9WraNAaw14bgpmAbjxo4DAGzYsMG5LfIUx+oFvUL0WNkYljFjx8p3mX+zQ/at6ynKPK8Xs7wSWpEtMQ+WWcYvFfrqevp403LJDlm1cpXzvcYN6LjcM3/fM0XasGED1q5Z4/ShNV681ddSnkSJ35vXyVNMJlL/epkT7wbreShaFjhjCnjN9Gmy1nRc/N49u7eN+y5Gd72vYpxFEQ0oE29YrVfJB6xD9Yzgxuy4ww5tr49cB4YZhFdHZ51U6mWleo7TdyA9r33r6RbjUSIn32+ggUkfLWvPvK9JHMzoYd1tvx8I2de9lGvB9TZ57D/nJ3Kj/N/Fahnd0wMAWCvydK8mX+yaa8blhd/3ildm6ZKlcp3Lkr+1RhpP5/3WFIUOs1IdyJPEteftS24XJe+NeaX3Uvv0fu/MCvF9ZM+t9yFGJVCgQIECBQq01dKQ9qgMttYPz73VctKnYKIEmjNnRf3keRxRMIlk2mSsgltvRIH+5AA5bUj11tzWZ/j7Y48AAJ78x9MAgDNOPw1ykWUUQCZ9ouFVD9ZKzdKX8PDLX/0SAPCaE441X/AYPsuAlPyyC++8nOfiFXPBd677DgDgQx/4IACgexirqRJd1IvhEO9IU87y84adm/nz/gEAuOeeewEAb3mzKVTJ6qI2Y4rjNm03KTvOgQ2fBwB87/s3AACOPV7Gm7lnxkWWo9Egv9IJZSqvrKqcN42Mp0wcDwBYscxkiUybYs6AiY9CfIbICxBgVk1atxlORNxcJx6GtEnhcz6ZzST8Co+pApfmzvdrxCJjNkn/BhOhr6ELtFYyaGQ+sxxiRRqWSyT+ZZ999zU8rl7jyodxEBn1WBA6NW6Gei9InSLHJIowcvgwEYrIVtZMs1/iYHI3lkjrMHk1bLQSrtzfK1lAuawH6onis0R2XHVFwTXvV1N29EpqjRRXdoxzYvxBJPrD9Rxb9w0AYOLESTL+hkVo1T1AZCW6ZYFk3b2CshwxcpiwzLgaF8E485CnSwLUCr1pg/pn2ly/TmROdGBmTsFFPYaPk6TjdtFSe1cZr2hD5FuNqhabhVkv3PvqjEXiuF3kXqJCs3Kv7rFeraRGZvpKFPvEenCon0sWm/lduVyynBg/Igi8TDQjXooir7Ip6igLGcs8DxvWhTJpVeW4DsZr+ZWLU2bvVRjnaNpeI3FDdVkrDVlTilnj/ZxxDUceWmyU5+ptVeIa0HgYdz4jWQ5cv75XpOkhrhO5m7+LlE/WqGvcl6pf7HpffAoxKoECBQoUKFCgrZaGtEdlMDEqciOAUkQ6LUdGVhN3gmd8ROZLXNwBPcvnE7piAMj3xGOgVRwDFPlKOdduNIzFwSwY/8QvqnhZMWycEfnMNJGPH33sMelLITDNe8RqxcSxH00i0e5as8i8rlxleOwVVNXhI8RKZlwJ5aFMSzsVLwMgirFsubGE61ojQvqKiW7rnYdW+N61uIgKyvPWZ5951r2P2TRqJZfOuRU+xc2GqYrVs3SZOTfuGjlCxr1OxCOom6ynpPPpxm7QY0Ge80qCdRuMGTNizCjhj/NJi9qN+9DquP738rpmnfF6jBCPRX/dWGRdgleBUuxKJJakVjSNXJ1QfZC219MD4ek5a6VojItmsDBDCc4YKtUaNoilqEijau15bdhezIvWejF9EhV0G4nVmPv44wCAhsR2ffP66wAAbz3rbGUism5E4cG8rBPvzKgemQvtWqx75VF4Jp6QrnOPV5Hbcok3yIuiBYn497/7HQDg+BOOd9pQhFINHBGrVry9MTGPiPkh2DVnyTgv+/RnAABTphnMm6goo4WatoiHsY7ZXDLOWBwNC541a6dnpJHHyJEGL8iuNSK4ijyE54ZmqBAzpFB01N/+9rcAgBNPfHWJk1JcDNcesUi4hrif8dVDoIZXG6mQ6+Lcordy3U2YaDygRMdl5+ppKLRzpw+t9UTUbPHerFjVK9eZKyuKAp2VXJluWy2IsvrbYq6bNlmwiohAqzL393u3dhlKGat+/Iqua6+ifJJwP+ae2z5WJfKBYzgXWsfHytn/LVTE8QGcJSFGJVCgQIECBQq01dKQ9qgMJkaliCziYq41XMTSkOe2RN5XxarRHHfiUohXY41UEX30YZMdQ2TTfffZ27QnT/IVQeNMG019Kl++1Jzzrl1rLJFY0RLJEa07OU/uIBqsX/HUtLeyV1BFWacmcr0dMTLErCeh1ZO9OIHIrRPUKXLS7BHhkeiieZ3n8e7ZaKPffN4xwiCZRmmO9YLMulKsE8ok9x6uOV5a8ZUOxsXwrJu8mxvHjjcZOhZ9lefpMrYoVWTa1JtHvv5Zqst2DxN8BPHysFaK2i1EFyU6ZuZ7PeCMDfUGRg2XOjoSH0GrNCFPDfesN4+ZDSHznbGqLL1T5rpZ22wDwJ67dxHhl9682HoWOO/UYy4Zxk41xavXKV6ZyEMszcUiqxFVlmf0xOrReiVGPus3rC/h/oisxNqrdJr5zPp9DB45o2f8D70DIq8J40zc0I9vNFlt3/6eiU2a/4/5AIDPXPZZAMCHPvgBrbFVFX4bojuscdUlsgezJUSon7nsUwCAadNM/NSbTz/d8OzNFWeL7a7v3yDNVTUe4PrvGv64lyxd0QsAmDh+vPQpLWmohZHxuLEGHyfK6aUzr3+8404AwEUXXQwAuOWWHwMAzn3Pe4SnROM6EkHxXSHxEDttP1uGSQ+YmZPrr/suAGDWzFkAgNNOe4PDU1zz1rno4MQJxmMRx2Z/qFQ6cNNNt8hnZr6eftrMy0677AoAePKJJw1Py4xndekSE//14CMPAwD2ftnLTN8eMm0hAVVEni60cjk9GfYzzvdIqV2lnrWIKLGyf8vaKppejAedOrLXsuL1uJ4xwotcJ+u/Uuts0Qn1hNMLJbLWauly3Uypjs29gmtQszrVUSNrWOrScf9HZGMFNcOOC9vPZvP2Je6l8H5j6PVZtsLM0cSJU+Q6ev/4W9SFnHyot7Y9/hcpxKgEChQoUKBAgbZaGtIelX+WFHZArXr3nM1/8vNP1ni2v2zZMgBAvzzh7rOf8ajQ6r/37nsAAGNHj8H0mTMBWEu5u7vb6ZM88Xz57nuMJfWyffcHAKTrzdPwnXffBQA47nhz9v2sVISdJhgRJYNDxljgvvv/AgDYZtosAPasfu4/TEbOrJnTnHEzEn3RokUAgClTJFOHXqEVZtx/e/RvAIDDDj0IAPC5L3weALDbHnsBAE4+8TVYsdJ4fHaYbWINrEXEc0/zdq3EYHRLnaANkrFw15+N1+PQww8DAKxYYbwd/f2SmeBhW6hxERUa/0Jr5fY/3W7GP8JYXiOkEvHDfzXW3TjBEZj/jLEKWeHa6oPrQbHkxnzk9iN0DxPvkvISOe99lEi9DhyH29PYccaTtHyJmYOeYSauxmJkRDre0kfyKvEQjBciKqpkSfi6Q88gx6W2E+fOA4OI41i9kVqLSnlxZafjcwBBrHeT108U9M1RUn+LVdg/dcl/AAA+81kTs7F6TS/GjBot/Io1Km30CQ5Gt1jMbHuFeM5mb789AGC5xMUwzmR093CnHZ74b5D4oKlSG6kojYuoxmed9VYAwBevugIA8JY3nQEAmMx7NCRB5tumbznf/11iz4545SsAAP+QNXvzzTcDAF590kkgWT5NG13eHjN/wQIAwB577gkAuO222wAAB8v6nSHr3FZmNkTMp0wRju33TfnsbW85EwDwuS8YD9drxCsxY7rxINz2OxPDss8++wAAlkuGzgLBjdE9VuVodHTpInPdk3PNuA874giVj7+WtAn1Tsgfko1Ij2BDUI47u8TDxpgtZu5EvN9p1gKTI8Zdd5p9eLXE6Oyxxx4AgMkTJzvXwtN3EvuKlVevL3hjs0E7aEpWTpr63g0ZrszJX+69DwBQl5W7/z4my69adWOayOvECS7ujt91jEj59Odrc1HwqAQKFChQoECBtlga0h6VwWT9xCXoC2ZptJTL8FNvaClKPjnrNmj8gFz11a98BQBw++1/AACMHW0suw0bjFUwZcoUnH/euwHYTJNR3aPcruSp/7JPfxoAcPjhhwIArr78CrnPeBpGDjdW0rq1JvZhwbynAQCzZs2ScTLjwXD3zW9fp/Vw7vjjnwEA8+fPAwBMn26suyliBZz8xtcBACZPMu+ZiUBbuneVseI/+5n/BADsN2c/AMCl//5JANZr8vTTT8vYSjiizISSaHZaEl//n68DABY+a86ueaZLTwxjMvbaazcAwGOPGhyacaN6ZJimoVzm6N77jRXx41tuQSFxDrS05wmmy+jRxqOyr1h3jzxuPEMdUmW0V5BaVy0zPIwZLzU11Oh1MW8sXCizf0rotn6tDyXipPi1QBiB7+KsbBBP20SpyrxoyWIAwPbbbivXUScjjUlYv67usNnZ4VbyHS/emYXiOYs9Xhg/UWjbbtaHj67ZWetQ5OXW9cl5d/toqacltFy8lTNnzDBtd5k4ojGTzfgnTTS8H/2KlwMArrz8anz8o/8ufbt8Des0Z/T0PjLD6Mf/Z+IrTjzuOADWy/nZz30OAHCmeEUe+avRuVe+8nAZg7mfcScRCjz79FMAgFkS51KrGd37tw9cCAD4/Of+CwAUZ+boVx0FANh2m1kAgMVLzRzQA0H5MZ4oEZmf9OqTAAA3/sjE7Fx5xeV4xaFHAAD2nGMs5jWrewEAw7vNuJl59ctfGDTnt5z5FgDAcUcbHr541dUAgL13M16BJ5426+RNr3+j6btq9KlHvFqU67PPPINp4oVJJP3wogtkvLJv/fWvfwUAbLeDiZeZLHEue+5u+pog2EWU6eKlZh/42te+BgCYPt14otf2GS/Xaom/mTR5MvYVD3bfGvNZrLuNeAqEzy9edZWRh8SNEbl3G1k742UeDzzQeJYYc0dd1WRJkeOvfvlLrKEnZU+zL1337W+Z8U0ye+pxJ7zKyEziZtaLl65TYm5I9EhdcuklAIBddtgJAPCIeKs/+OGLAABdEl9TS+IWj6nWIBOd+dKXvgQA2GsPw9swiRm88vIvAABGiJf3+BNPAABMmeRWTbfrm5l49Fpndp/x1thAjpWQ9RMoUKBAgQIF2mppSHtUBlU9GUBd4hogkdQ8y/vZ//0fAGCf/Y2HYFiXsax+IfEgO+1oPAUjRpkn0W6xVP7+xN8BADdIJsJIQZdctGghAGD48B4AwM9//nPc/4CJE1klFmdVrLcVclb7+S8Yi+sd7zwHADB1sjkv3H13c578zAKDfTB7trECPvtf5umYuBLbzDLn7KyEe/OPTXbAuPHjcdyrDHorcTWeeNLwvc22xlq5/IuXAwBuuvEmAECXWH1EmWTMyhVXXAkA+OhHP2rGJ0/q/a86BoDN/rjzboNC+6ObbkJDvDL94qVoyBz86U4TezJ9luHhrDPfhjKtl8wqxv98TuTTKRHsh0jMSl2i5h992MSZ/PrXv1EeK7GxQu570Mh+5WoT1X777SZW5b57DZ8777E7AGBCj7Gs/vG0sdZHjDZer4YgmfIs+777TOzRzTcZeW23/XYAgNVS12iPXfdUJNUx43oAAEsEq2XCeEEzzYiGbF7/JPEz8+Ybb9TBBx0MAKhVjQXWJfE0k6caC/YPkg1ywH4mhokVsRcvXoT/ueZa07fULmqKjqwV3Tv44AOFB/HWiKzrcoYfx+7h9P1/Np64X/7iVwCA977vvQCAqqwjYgKtX7cO4yVLZ6V4xLhmctbqEav0rrvM2vrtb02cxDbbGhnuvbfxCnD+GXs1d4GRy/s/bNCSydyx4g2Ztc12+PL/Gu/c8qXGG/O+8wyfy4QXVlWuC7ot52ikxGzRFfPBD5o+Pv4JY+We9WYTf3HNN78JANiD+iLerf7+fvz4/35qrj3rLAC2VhMrGh97rFmDPT3GK/H9G38IAHjr2ca7MUIq81IfbvuN0eODDjsEgEVHziX24/jjTIza0/Pm4S7Zp1aKZyFTVFfxhMq8R4LloZlyMhfvPfc8AMCXvvZVAMAb32g8KZdfafaFY44z3oGxE8YLL2a+b775ZpwtsmG8RC7W+Jlnms+/3GfWxLx58wAAHZKZNFZrYUnclIzrmyJjeqKqibm+Lu3PnTsXAPCnP/8J06YZz++SpWZtMdiiIfN73wPGu7rzLrsAAI468mjDo3hE/+cb3wBgPQ7f+rbpm1lQixaafXyFVBNeKa9/+ctf8N7zjcyYEfPG0wzC+BrxfF4lXo2LPvRvAErZneLp7heP67XXXgMAOOccs+/Tu71qldHNiy822V4TBSPmPe88BzVZd52dRjaMi3r0EeP5myXxkAcdeICRi+g9cWC4pq691oz3vPeYsaSyV8+XzC1F06Zbo7+/hLAtL8SBGeDnOGT9BAoUKFCgQIG2WoqKQUG7vrjU19eHUaNGobe3V9EUXyilS5agkCfQgvn08l0sZ5Ks05DwHFExLOSpUSyJX//OWIH1umvdxDzTF0uGuAZZM8O35QzzwDlzAAB33GOs1PWSzfPGNxgrhmearF0TSaZCkbmH+evXmyf33/z21wCADdJOl0SyM1Pj+GOPRiRWuYZWEIhRa3wYr8UDjxhcmDEjjIX5hGAg3P+X+wEAb3i9sTS21QrIJNZEkrNNifX48x13oCm1Ie6/13ghdt/d4Cs8u9CcRb/ljLdIG0THNdcnNcHdEEvrsb8bZNJHJUalKZ6Zl+1nvGBf+6o50/7Pz3xG26lKG7Qcb/jB9wEAp5xyMgDgrw8+BAD42c9/DgB47amvBQAsX2ks8jvvMHO0n3gtnpo7z7Qncjtb0EIV2VHiEr557XfQLbp2wquNxX/V5eacnBWpO2Re6d05WbI3Zm5nZHuXeEz+8EcT9/Qyidh//cmvAQB8S7AwFgvKaE10bWTPSJxxxpsBAJ0y7xlr1Ugcwf+75FIAwL77G4trt92Nh+C6b3/bjEMsr1xQUafNMHETRx9lLNKvfNXEZI0dY6xi4lcsW7IUr3+diXP6/g+Nx+BtZ5s4j54xJk7oFxIn8fRTJqbj7eeY7+c9ZTwmN/zQxF50iPze8y4T2/WNrxtvybnnn294Y10aoQiZrtc1G4xFfe01xlqdJNbowmdN1sueexsv5eRJZlw7iMdUo9a89ZtL/NgvfnErAOMpBIADDjJ6cdZZb8f/CH9vf7vxDLJGVSxZbH4R7T7xcnzjG+a+bWfvCMBWGW+KF+/stxrPRLXKPYkpKdJcBOSifz/52c/MZ9IHvZc9I4fLOE2cyMxZs0VmJBaBkpgM8Vr+9WETX3LtN/4XALD33nsBAF57qlk/1333BpwtcTysSZTnhu9ap+nzvyTe57gTTwQA/PD7Zg02RD677WriKCqCPzJVEHf322cvaZdZUS6aeP+G9bjiShMHw5i6+QtNDNIH3me8MV/6kvn+gvddYGQo85kSB0kmI5d9/XeCJvzXv5lMqxWCPHzIEcardb/UK7vgwg+ik78hzATiXipzMV884N8XXB3GCW470+jcttuZOXhG1u/JrzlJOGJtH9EfwbTqXWViYq688gocepDh51vfMev18suN5+trXzd74IXvN+NlXbC4U9COWaNOFGTe0/MAAN/4XzO/HeIR306y4I4+ysQwjRlj1nel1qGgzz6OSv1vJqZm2F57mT6JyFvyifT19aGnpwerV69+wb/fwaMSKFCgQIECBdpiaUh7VDbmiYzUXLwYscReNOWMVssoiEeATotajZ4QOXel6cEqul2CcMukGK2my4qngugp586NesPm9st54v98y5yPxpHh5e1ieVqrTirSdtOzwLbdacuEiZtv+gkAi3h7pkT2R0WGmIiLMhCLlgiRg9QdqgqypVgS3/ym8QJtJ5k3p55mvD5NwSGwJplYYsJbUkLk5TXXf9fUZlnbZ/h7+znvMvcoJoNrSVS7jEXW9KqwEiPiB9//ntxnvjnlpFPMfURjzRvo7Bph+QDwfbHmXvf61wOAoua+9/0mloExOF3dRl6plFN98CHjxdlBrN4RI1x8DWb2VMSb9c3//aZ6M177emN90pv11Dxj1TPzYOedjYeFSJY18SSwWu6ttxor/s+C2/DJj5kz6w6RD2OUylqhCKMii7TJulLm8w9+4AMAgLeINUyPCu/bIGf8nWJhkSfOBWOXWF33m+KJ6erqwmniGVwucSHflpiDiujgAfsbL8R+UrmZ6XjExFgimVaMg7r6SuOJ+vFNPwIAnHaW0WutZFvYtccMKVrzdYlvomxu+J7xQm2QSs5nn/12871CuUhciXiSqpIdkQpCq9Y8Kuihux4AcOSRx+BesbZfdYyJ1+LmkEjGET2iuVrzwj6zKKSu1A9ERxnDcMYZRp5E1S7gersKQCt117q47lIZr1kj/f1Gz9/8ZoPlAkGWtRl5Mm7hrSoZVnUPNfWmm4y3ixmJf/zTHTjp1a+RgYhHRbwTnRKf99nLjIfzgguMlc94qDdITMeb32K8f+tkLXLfIk+xhFNynTPOLoKNi/nIR8ya6Be9PfgAk73zyqOOAABMGCuZRVXGdLjVo1mxurPL6Oj/Xmv0eZ7EwzAWZtYME/tx9HEntOyBkTq6ZA8UD1Bdan6dKB7TY48+EgAwSlBvzzr7bKcdxVVShG4XmbyZphq3whixww818Xp77klPoeGXulXpFk9SnVl70qXIkrp7ww9/AADYV9bm2jVmj3rr2WaOql3daGobXDSCSk6PivBAj0qkKPCD+/0e0sG0g0lPLqLyUY98piBTXAT0p/KuyL2eAEMKNFQ41xWKNuYe08SISjDMkuo3ssf0qbrjIQR5pABabgae8paxVDfh8VkFTX11pWxavw8FPDIXjJ9ggjBXSEDXMbL5Dny/91oei3xWlc2xlnS4/HF8OgdyJ2G/PQA03pULhvwyCXCryUbOxZeXuWBbXpss6LdaCkXyuIy8dMoR1v78USWkNGXK+c4ENpvfF4UGVPsy23FH87BD6Gz98WNhPO/6GdNNeu6vfnMbHIrc6yjoIoI+69pZcB8GEz4UsVSCN5/d8qCey4+tarnKh2Ta65aNMEakMiT0/QUMvOWPvmzwtkf++Jt3DFDl0VCFhUEVTKBweCjKvLvLsaRT3PzN5wwiJricDzrnby32B93VD+4Dzz67EDOmmgBFW1zSlWnk6bWWPJDO+CDCchz9As8f63XkxdXhdm2xj8WSdj558kSXhwHWv7e16J5D4rj/cu8DAICpU6aW2qJM3KZ9HSQI3Vve9CYAdm5qHV7ZDu9+f04Bq8cEohsmZRoIq8DAblUdtuGhDhYEGSyMrr1s771NOxIAPk9A9njkEhVtZOgBuHEOmGCwoxy78UQkqXLPcNshq3EhP/Te4UclTrDzTiaFeaUX5EtgTv1tzAieyL74W+Aa1ttvb46baZj3CpT+pEkCAJfbWVRuvAmKBvg9DunJgQIFChQoUKCtloa0R2VQ6cmFdRUqQI58RU8KXzNxldtjFrH65Mk9b0rBKbpf1ZIRa14sbrrc8yxTk4ht8Km/IApy5pYz5xERxPLWh2HhUTlTnsz96wXem58XeYooFXexVjtnG7TmxfWbyAXywL3oWeMVILBVTrkor0Jiqqo8xAVrivfR3DPXdI0YJuOhO12+jlxeMhl3nrpyoSXDlNJV6/pkbAQls3JmG1pEUK4h/7Qg1khgI49GNDCQesEifwSt8ywDuvXZX6N/AyaLNaLHbPSAyXgYbOl7VKh71KFxEoS6ZLEBeKMOJzo2F5TJ9CmyYnn2zIXa5vEbPSE5j1FiV8YqB847ofYpF/m+W9L5H3/0cdVbdYWLTmQ8ospc1zHTO+nMZAox01fVQyFzSD3PPV6iqFC9TpVfkb2WQjBtjRLvDmVtnRwMhpejYNHjPHN1WPcN4WHposV6PJqX1p0wY9436Tkiv2yE4yLP5nqWDKCMI6YnU1AM9EeBTGQSiex4hs1rG7oPsW+5Xq3i3OE9Uhn7BSQlWFmOl6fOmKpryeoE91B3nnNvz3jNq18NAPiqBICOGSsFAOkZhrs+WDiv0KM+qxPnSsD1XXeaAPTM03sGMkexu/ZsO9xzzHzvKuB0UwUQ8eOfugSADRrPmg07vwN541hEUeTxeQny/9FNJsi8bwP3N2mH+5/cTq9fpL8PVh5nnWG8UT8SCAo9FsvcI61M9D9piOcwdfcKeikJAPfJT34SAPCVLxkAwD0k8YF6kTeaLWuGLo+BfCUhPTlQoECBAgUKtNXSkPaoDCZGJYL1JDA8wtbHk7M8+YIAZ3Hint2yhDeh9DXMQs0+uV6AlQjrjqR0PixmOwuE7bLDznIPORXrXeIj1IJI3PNxLeGdmDPQUZJCtkaC1WLhtajUkBESOSEP7hltUggAlBaXM8wQcnyEANllHE/VPY8uRH5xh/TJ/kpatmixOS/vFuCyXKyeROMnpE0ZT1o03TYit08GvBIunGPTduIaUgbk8btK5FxblfiHMRON9V7rdCHmC3YuupbJXPiAaJXEne9pM6ahJmfTTPWllyKlDDkumd4KdYZl04XHkWPcoDOe5Wd+O0JRAcQV6rF7Da3SCVI6YaQA2kk8d8lMMq9JLvpOk7RCq5G6aG6cPMUE7/3tkUcRVVzZVBPqlqwZT8/11JtrUmRN4KyCa7BWcXmhLpdsrljmkymiccX1iHLe+7lGKuVvS21JuQr1XqinyRCP7If3GF1+asE8zDlgf+FXxiXBsWyDffvr18pYPAmiW2vXSYyK6FFGC53t8T0iVAVOgTKmfo6dYLwUWihSeCtUdpxPWb+yZ2Qsd1D1YlRkbpcsMtAC4ycdqvNt2yAUgnifqFNceyJt8tQx3MSJdUqAus63KKX1/rqlQeLCpjRvO9t4s+574H65l+5LFbozLl0PBZuUtQdPz+WVgIYskJkhs+vOK2hIb4XOu1xH4MZc5mCEeMzoQSwY0CzNEkhNgQOpq7nVc/4e/UOCfnkPNbqWeGum4s5nIYNgavUICXBdsNh40idNNzEvFV17KSLu/dqK+5vnU4hRCRQoUKBAgQJttTSkPSqDhdBnBDYtZVoUBGrjmV5FIvBzTTsWC1XOjyu0vD0Qp0jbkTiCrnKas1gzPNOVtidNniKN0JqXs0o5J2ZqWaZn8m7GQTMz56qxPLETSErLyDcLVDr98Qi/PMuUlMJEwanMdZoNI3JLxMvDVGltB4xRccedNzOVcUX4myixG7E14817ei1Sb9wNxp4wQ4Pn6eb6/g1ME5TIdhlTmjVQrXVJmwzOYaqcWAjS13veda75WsGlyBOtGgKmyf2agMI5YEq1kc9RRx0DteKZrSUyrXS56YbUnVS8eFXOlVzfIdkBp7zGnOnnIvuKlHFImywSZi063qvzJePnfL/3XJOJM6ZnrMhK2qCrQF6aBO/q6Hb6jhi7IvcxHXLb7bfT9FnOO/WzKhlWeWMA3SHYooz39Dea9NVE5iyTMVHPVQdLa4/AbJwHLapJ3RKd6RRrll4P+LFoTUIDUH6cK/EayHUjhxkLdHGxAsMkXTwp3Pg1zremJXuZN83cyKcmKf104rC9PDU3VKSgpCZg5NYPlMr61b6k7y6ZtyVLFsktzHZSN6XKztxH+AUXnoEe5ymTTWbTM/MNSFlnx3BdU3b9yl4yjGnF7EldEOb/0tea1QZif+fZkqavmXP0pNBDpe5rkFLJSmP6fCL3ct65Z6huyf5GPeH4U4FnIOAZ9YRuu741xrtn13CXXb+eh4ze+ISexMzta/UKE9/D0iHqcWIathf/xfTkXPfcDIXwP3myWXcPPWSAK+3+ZqgpvzW69pruD1akMXgyWpHxIw+bVOPhw4zHlQB5SVcn0KQeU5Elvid11zUpxKgEChQoUKBAgbZaGtIelcHiqBDjIfe/Y7S/mJwx4Zrp3fAeBPl97mE75F7t66R0nY+j8qxAJ/OpWC2PyI3I5lMoLcsc7rjZ7Mjh5um3KueSGj8S5TYWRSyErGj/9GufX421QIAjWjkcdwbvfska4N0JvR+F7ZvFFw846ECH88Jrg7OTRDwX9UBB5I9pUwwUdUr8EYuwoq+JN48k26f5/uWHvwJAyQL3eInV8+bJTx0QUuxNgh6MfFyLg/FMnAOehxe2noHTh2ZkyUSedMprzHW0bmJXPoVN4VB+YsZcea6/kZJ5lcQED3Pb0IB+epb8M3xmKAjvs7aZBQCYMX1mi6zVDia/kesJ0T5VluaPN51+ujM+yxNjefwMrFzXjM5DQVA0c80qKYw5cnSP9Nl+XDrvXqaJ2sY6RHN/b+9qjQspvD1AsVfoaWlZv5QxY1VML53i1VBd1HgyN26oiGxXqu/atumTJQRaoD+UCc67u//ZPtzMlqVLTImJSqVq97XClR09ID6uik1XNN8vFnDJ0WPGOF9z3ImuB9e2LiKU9JxeN8nCWrrE4duGErbf1zk3uqZyN0Np4cJFcr9cFyWu/A3D0heBzvxYDkNPP2Pi/vbez2AzRcpL5LwSR0XnQtdNobJhkUFm3dkYM3pMpI3Y3XMs/pfoueJMmU87xGs5TkpkKPholCBtBWlqO05SiFEJFChQoECBAm21NKQ9KoOJUYlLD3NRG58KYJ8s/c95lqc4GnAtbn0yJSpq3MpbTIhpeardWwqjTZsxzfle8TC8NqxN42Kg8LLtdjCFpEaMkLNuWvJR5DsjSq25FodFRTSfv/vd75T7+BRtz0ld5mgVM46ElOvT+0gpb08URIXIJiyEerP8Ntynf74fPtJYEbvuupN8nDpXl7NByvyY74R/NbW8z/X8lZaGx1PkWvlWubx2ULIoI38+6UnxcDV8GXsehcJvR11nNmugoHWvFhTbtJkizni0Mdda8vFF7Lz7lpgIMsoBz1tnMYs4Xu97z6K0X/DFk4/HS1QaRRT743LnpRIzTsz1INh5l8s1i8/7WNeovJfXzq5OJAqVK+tTs3Lk48j93u45rufs/vtN5goLwvnIwxrjQS9IkWsb9PBGhSsjQqEroq7OBdy2It9+ZR+uvIYPMx65Yd0dilWTcU14bdj5ae+FXSkFQCdNHOd8bheyz5PsG4Xtk591SAwKy2hExUDjyp3vueYKr72eHhODxOKbDhqt5/nWtdUC3cpxG69GRWDviRNkPTKMIyvKl9v9oLB7k49urF7Hlt8QT7/9/U3H4ranGTxc33F5Dbtt2IOC9r/HIUYlUKBAgQIFCrTV0pD2qAyKCltfJYZkt/jnq/LkmErxNZSslvJ7ZosUDdfDEtmDdgC2iFuR5yAsLGNQ3vbWs813zOZxH2r1qZgF0dTQ1tgG18Ka0DMaADB+lPFcKBprUWgtDKJj0qIo5AlcYUGIQCi8HnbIIXJdLryUEWett0CxA4iQ2CDPucrmveee68gGRA3NXdnTPqHs9Dq4qMJTJbbnjVJgEKlrqQK2cJ3yS2heopwSQVgLJrkWs/WgwP3Es4r5dbPRr+1ofRnVCdNJyjiYzLWCaL0oKnLmW73s2nze7G8476MyT5QlZZJxvl0lS2U9MIMmilxeeGEuiJ2tMibPgmSaeXygFJMhtWu08qdQ7hlZig/D9eJ50hRtuAUV2jak644IrbG5dtSYHgAWo4jj96PWyHPacNuBF5dA3ntXrrQIo4lrAzJzRpFpvXgYogfnwjOL3Y0aPtLliXOYunsRCrvGMkWxljblmkQ8DLrfEYHWdRwqBlGqyLTuHHCd11POd2bb0jg90xb3WtXfzLXEiXZ6yiknm/EKpg/RkG08g5v1V/bIcS9tiH5OmGzqRDH5paUAJNFyyYvnWWlKO5HWgjI0bdo04UTuqzftnuJ5inLCAmVUEters9ceewCA1oaijGP1xgqvkavv0P0g1wmbOWMGAGBvqU2kMtYBy14hxTmRurE5doOT/V1iUXbfbXcAQL/UOurskNidZh05M8G41xO7Jd48vo/gUQkUKFCgQIECbbE0pD0qg8360b8Td/hqMdA68GMTiPKaubgMUaVabkafrvm0bY9ZYxR6Pk48FMmTJwqmtiLn/5n79BsRVdKLry4Eh0GRKgupQkorKWuqrBSttuAZJKPlzbhi9Sx5Fljuen0Qtx935FswSaJ90JuhaKnSB1FU1XL2amho/RI9tOV1jPmRcctcsLYG8oYdN2XPcSUuIqOqUsJx0aISq4YxDVF7+USsT0K9SKo2g0JxX6RllTEzKyL3Oo47Ia6C9KFn2LHTt2bBxNZMUqsUHJec2bMGCvVUh+NlVqjl3XA+L7SWjAyGeqH1WGLFUdHz7ZzeSa4lN3aDFV2zlNkiiTt+z1LXO2WubPJBZrFFdB4kS0OQl+kZeXbRPNN3wiwtN94gVs8ax8Uzelc/SHFcASRTQvXWq8EUqZ678U5cv3w/erTxjMZVFxPDrmEXLTmPkxKmSokfAKuW9QIAxkyUjBp6Y2pSwdxOtHnPtaOxCi72Ca9evWqFfBrZeWCMVObXEaKiiuzo5ZGvX/WqY83XRAfnFErdsZxeYE0plLlEUdorDB14wMFwifuxqxd2/5f3yrP/vaEvfO4zwgLnrtCspiJi5hHbkjpR3JcSZgGZ70896TUyDlmT3K4UPtvdB3PVd857BtaHmijZXMce/SrhhaOiDopHWdGBS/C20rrpSjKshNdP/8f/ky5ZLZ6V3iNEFeq5sCUDSD0dJIWsn0CBAgUKFCjQVksb/aDyhz/8ASeeeCKmTJmCKIpw8803O98XRYGPf/zjmDx5Mrq6unDkkUfiiSeecK5ZuXIl3vSmN2HkyJHo6enB2972Nqxdu3ajmWfWz8b8A4BarYZarYYojoyVFMXyD+ZfHANxjEpH1SBBxpH5F7n/kloNSa1mzqOTWD8vYvkXmafbWmcNtc5a2zbiSgVxpYKokhivQeRfY3iq1KomUjyJzD/vOvZVqdVQqdVQRBGKKEIu/xBHSCoVJJUKokps/vnjl3FG1Yr5J/+xjxzm2btS60Cl1lHqm//kaV7aIS/GrWP+8ZqkVkVSqyJOEuPh8cctVOvoQK2jQ2Xq9xnFMaI4RqVaNYjDbeRS66iZ2jgDzCPnu1qroVqrtcie/+JKIv/MnLX2Zf5VOzpQ7eho+12UmH+VjhoqHTWrmzIOXlep1lCp1vTzKDL/OAbKsdpRQ7U0Nr0uinR+dR6EF+1ThubLp9yG0WegiO2861pq0T2jo1Fi+eW1hfxH2bT0QfnIOkhqFSS1itUpXgfzj/pD+ZTb4cBqnR2odXZoG3ZNGM/M1GlTMXXaVERJYv55MmY7Op8D6E+tWkWtWsXChQu1D64pvk+qVSRVkU1pr/DHn3TUkHTUsGrlSqxauRLTZ8zA9Bkz7J5TSZCIjKJK4vIkDHNfKsSDu65/A9b1b8D69eYfaglQs+PV9atrTMZdq6JaK68pMybuA/3NOvqbdWct+HsB9Zwyt/MofgzZO+2+FZt/3EPluqRaQVKtIK7Exvuq445LfXWg0tFh29B/hqc4iREnsd1LS22U939dU94cVWoVVGoVu67aXGPXudEpf1+y+lwAkV0PHK+/tvS3SORTXjf+uP3foVb+3fVr9xZzCX8fuL9xDVpeAMTmN0bvFVnoGpP/nu+3emNpox9U1q1bhz333BNXX3112+8/+9nP4oorrsBXvvIV3HXXXRg2bBiOOeYY9Pf36zVvetOb8Mgjj+BXv/oVfvrTn+IPf/gDzjnnnI1mPlCgQIECBQq0ddNGx6gce+yxOPbYY9t+VxQFvvjFL+JjH/sYXvMacwb3rW99CxMnTsTNN9+MN77xjXj00Udx66234p577sG++xpUviuvvBLHHXccPv/5z2PKlCkt7dbrddQl0wQA+vr6tL/BxKh4Qc42Z90mpDvfa8Q2r9KAfTlPZeAB74u8djTJprBntox34Hk/b/XqjfjZEfZ94XTFw3ueN/OcWZ9eMxvVwgybwhtXLO8T7dqTg54fM97EPRtWnrxzysgZA2VG1FpvHP4clGTnXkcSXgaai6ywGQeFx58mRA00Lm8u1BBweaSIi5w1YHjuXJaPOx6brGGzNszXMn+FLE2iY/p64etk4bZfRECUUzbRc7ah5MncDjd236v8cue9833uNsL6QRr/4vPig5XoXMH5Ptb3bMfDnUDRMi7Ot6qzZDexBpL9wo1BKvy116KD5o8JEiOQZ5n25a81+95rS9vmvJtXrcdVuLK2lcpb17AWeWZtG8mcmjzB8Fcjym3O+CGPhxa9cPWG140fOx4AsK7PVJ+O8pKMdd25cWoqF21L9KFov36jloUi7Ufux24bz70v+SrGfUk1J+c+wPXbdO/313CJLz8WJyr8GC1Xt3T/J4veemFzcebKR+Nx8lz71mRFX48Z18VxMSLI39+83z/dS1r0X38pbBuFN08+NBlb3pJiVJ566iksXrwYRx55pH42atQozJkzB3fccQcA4I477kBPT48+pADAkUceiTiOcdddd7Vt97LLLsOoUaP03/Tp0zcl24ECBQoUKFCgLZQ2adbP4sWLAdgoZNLEiRP1u8WLF2PChAkuE5UKxowZo9f4dPHFF+PCCy/U9319fZg+ffqgz7torWRwUUw180CeSFkzgq8aNa1VY+V6vipGCi3N3LkuzzLrpJEnUubZkxe2oRgn0rfiiLAtxTSAfC5jkutSZi4xCyi3WAdQGAG2xSdreiNcHBWtIipy4XjSFsuFEf9yveAsZFlWQtA0X9F7k0W+lUJcFU/GqZs9YdE1Ob7YGTfNyyxPdRy5ysabV+KpVF3Z2dpFtLykb2ZJca5oNRG3ojQHhYdmTJMpybzxKaKsWJqebhUeyijvq4hc2CflXACKuZNJho2OFxSRp7/E+lCE0va8WLkJLwU/z/U64rqohUiciCb13K0araigfJ+641THmi834SVXD2Kh3hX/Wur5o48/BgA4+NBD3XFplpSbcZGk7dvhzsP7h48YbtedWpjMpJG2JbPC1tkSGfN7uf+Z+fPl/sLhTTNyqGvcY6LCVi2njKWtOnFgvLnIEt/8FRnKfbHINk1Z+8q81KS6cDer8Ra53QtyV5b+/PnzToyTuGZ+itiOeqBE3xNFUY5KnEpD/r7kZ1pxHVMlW+bbXbfcQ8lr4e+P8jZJM7s+hRv9PdLEKa4J2e+JXs39KXVl3IKqzC3Zy5bMskyz03QN6Hr0fhuov0133yutKnkr+xsrsItu+bg7eZYhTb02mEnY6vI24yr9Tg/mN3tIpCd3dHSgQ8pcBwoUKFCgQIFeOrRJH1QmTZoEAFiyZIlWA+b7vfbaS69ZunSpc1+apli5cqXe/0JpMDEqUWGr4yZd8nQuD3h8kkwqtM5NXEySeBgfkjufE3eElohCHsgTqeSrN1MJJE6ACq29Kp8q3Vx2G7Ii3xNfRc5LpZiwbUfuViwI4bkqgSYWKqVicTPk6ZfgmVrqRnAmaCEzpZ/YAMTAyGQ8SeLLXksDSzsin0oZ9YVWOhFm5RZae7xQcBIa6Qbpi/d7Z7sFvV9m3JXYlUuU1NAQfuOKuasisq/Qoowjtw3P0lRPGo0lWtp6DC08iV6k0l+5HeIssK1UcDPiqhuLkVDG/D5uL+PI09GI8BT0LMBkF5m2jEWdVGhhShtiGTelL37vx+IQyyRLuR5Exqo/bEfmOy7rjLzIfOaZ24YvH2LQ0MNo14P5vCqWd542HPkoIGYUqYzqGfWUrJhrp0wxHt/hUm2WKLpJ7M17h+E5La1fwK49joBVZlf19qLCNhTvQ3ByiGWhbXBcYu2LjDmfS5ca7/K4iaOla3oS2L477hhAJBglmaw7UXed36kTJgtLwn/ieiftnlN1eCFMlEYoyB7UZF2tomnXnayluGIMS3rpuITixF2/Cdce95TY10HXe6kVzDnuArqRZc0Nbftg59wLOa5YQ3VEHjV3H9A9R39jXKyiNK8j5poqB4cBuqFxr6hUXA8wfxuUF+k79mLNdG/y1l4URToRXHf8/Um830RiEnFP0fmHNy56UATBmTK2OFSy9rIGksSNgWTsWWY3aoe2qBiVbbbZBpMmTcJvfvMb/ayvrw933XUXDjzwQADAgQceiN7eXtx33316zW233YY8zzFnzpxNyU6gQIECBQoUaIjTRntU1q5diyeffFLfP/XUU3jggQcwZswYzJgxA+9///vxqU99CrNnz8Y222yDf//3f8eUKVNw0kknAQB23nlnvOpVr8I73vEOfOUrX0Gz2cR5552HN77xjW0zfp6LBhujUhVLsoiIailPszyrlHPTRJ56NZJdnzzFopbaGYz5UL4Kr50avR2ZtlEwRkUROokWK20wLkLOtCsd5Tas9UPeiSqbSFVOIjmqdLIciVhteeSetUY6LnpAaP260eAcd02O4bLUtUA1c4VxGeQlzZUTIvEq8qiWGWW0fOG0UdH6Eozh4PXuuWoiPFNesXpaUlQ7a8Iv4yD41E90SzkHF6s/82rZ0IrjGXCkpgbH5I4tEf3KmlY+WqGX9UOoE03GKtDq4bi9OVAjRMYn8xt30dvB2AiLOkkLKalw/K4HibE5rDKbcm7UrUXdaghPJmuECJb0RDC+okZLPM9a9JMehaTD5UWtdD0up2uJw3ZdbYwfo47qmDT5oFCdqFa8uAci08r8jxlvKvWqN9JDRWXmStLJvrjQtUSyXGauGz16NCK6tsi36ExCPdY6LJ6MZe0RoTqW14gouDLuCj00fmYhYo1fqFD/tMKwa49qPAVcdFFdezKf1EGroyIfGWNFvCZxXNP1So8eY22SbqI3Ux9kPGpVy3W6fj1rm94urczNL9RNpvtSVZB2U0+3tC/uS4L2m7OviN4PMwdcD17ikq7hWO6LOxKkLbKBtEWvtHjEuY45DGmDupU2Xd0qvBinRLx2GouGXPngPKW5q1s2Fo2/Vx3Shru/Rd4+p+jRsubsmOhJriLXPZRtuHGZPv3LY1TuvfdevPzlL9f3DHI988wzce211+JDH/oQ1q1bh3POOQe9vb045JBDcOutt6Kzs1Pvue6663Deeefhla98JeI4xqmnnoorrrhio5kPFChQoECBAm3dtNEPKkccccRznjFFUYRLL70Ul1566YDXjBkzBtdff/3Gdt1Cg8VR0ZovkWv1aYi1nlmL50CzO+SFdVlYpyVmnAHjQuSpMqWlbqLji8weYUbe02scuR4T+6QqVm5ieEkyxtV4ljaPRvkEn7IiLClXSzrRWBQ3OlxZUrkwkptWrRlvrJ6HhnOfeiAYFR/TY9EsyU7aiF2rVFlg1orIrpKYOIKmVBNWVmj9xu5cZJpdRGspQ+LX5PFkT0o0joKyc62FWM9f6f1ys34KwaegByNP6y1t8FrKRudRM4ukLo0GnbB+kM1qMWNhTJNUAGe1ZbYTx4gy0S2eLUeMqWJfrn6qQeYRvV70PGitK+WcMVqsqRPZLKXIte5iRzblOIKsfLlaolod2Q0TKvHiZWYgUwua3gla6czWYe2Snp4e06boolYmj5kuRau4y2nHx/Ko1Uw/XR2ddgB+RoX0EeeMNSmcNnh1Epu29pRKuHZvEotbAhES8UzkjDsryYbeSvY1UWJTJkxyszFL0pe3rjcypo56d3V1GnnUOlhPLLZ1krg+QY+x8KLeW9EH7VrqbYmOZhG9X+5emhSuZ6a87dMbzTioyN+XZGI1fEj3pdS9jmtP9to89/c3N5MyiWvI4ffF/Yj7OsdLjzDXBevquB7wFhwVTz6MUTJg32yD9cAoG29P4Z5Lr5XsbzauRtYv1wHlo7XqqKuRtsM9wPv5aYNzBY+nLSBGJVCgQIECBQoUaFPSkEhPHogGG6Nin1p9a9b93s/Isc+Bvo3hWrkWRdG1rgqLWKKfRd57C8Xp8eRbaHCfwOGNxefRPAzrQX77a/RT3yrkbV4fHjZI6ULv+9ZYDbW0/TYLf1wDzEHhtaPftzsj9cfrXhv5MoVrSfjP/8Rl8DEDoshvpyj97VuSPi9eL5E3Hq/Cc0s7kTfuomTct/RFr9wA33vei8jTLZW1enfgUd6ytmxb7cej0z+Qnvu66Oty2WJr+cu99sQTTwAAzJg+re24oufRQZ8mTjS4UNttt21Jp1wO1Ds5wPq0ZL4fOWqEc107GZuGy+2031MOO9zgxdQ6vfSdNjrj8OzvE3Jfz+iRAICJkybIt5ltawD9LDxPYGkhO6++/Cwz/vqw+hO1zI+3L7X01X5PsXpOT2mrbpnrS/ubJ8OoZS9s7zkeaF9v3d/gfF/myUrOX7/tZRVpH/7acnlpkac/GUW5no+7R7bqqTTxT8aoBI9KoECBAgUKFGiLpSHtURkMRYWNYm+xafTIjt/Lub+HWULEypwIp0T+Y1Q5myOqpJyZFllqn589pElbXsE9o2U0eCzYLxaZ1mHZoocSDZZZRbSe8wyRZC1EinaZC98FmTKvCevMpO73iqZo4gsswqGLAcAMpCJm3+WMffmrSUwXfyDMQCKarJFd7p+JaliR+YTYCZqZw9oyeV6Sv/RJRMqcqJDEVSA+jpvGZRNvGPXO67w4jJwZOaVMLM8oobUTUzaMsSistQIAWepmGvm1fKhzScPwruiTJQRgxg0UUi9EdY4yluh/omJmTZ63R+4rda5Zd3kiwjNfGSeT57aeFI1azku932mDZEuGcK3J/WxHrvP1IpV1UXYw5jnXDvty9XjS+LHSmMw/0T+9TDKLptt02vHtUOIRnfnmMzQzwq/ZExNZmain/jl94a7fd7/rnW6frN8Scw7dMeUlmUQpsVtM38OHdzsyYuZI5KUrxsoz9xyRcerqBTGPzjj9NOWRfVlQbu4VjJ3KnfGQ8ibxofz1wDg5aYfxIcTbKbfBDJtUsDx0r3D3Jd4V6XrwkJqJbMvyWpm3y3LtcU6TGAXn1fNCcl6YcaXygasXqqOpu2dqX+SdwL+p3S9y/X0R/eQe4HtvOK6Gu6eoN6TFccLPPW8efy+ade2TlHjZm5uagkclUKBAgQIFCrTF0pD2qAwm60duBGCj4/kUyAyb1MJpAih5AYg/oQ4IWvNSbbPw/Ry0huXJPqpoXIB//JloFgynxHxR1TNfZnB4j+5C9ilarB7yxE8jwEakD1A91cvmiGJXPfRss2B0vK8+pr2EZ50RI9aLUqYEhUeLW2TvZRoRL4XZPbGXucN5TxQBUsbGcUe0NCO9KSJ0IzFc5LVS6RTuJeqd41Ke/TNv9pU4PNn4C2ZHoQT34J/NiwyZ7QBXLloVmrWaeJvWBIIzJpv9ZOWjZUVoWbItZtboMGORg2TksC+5zr5nxpabyaCkOpq0xK+odkqnrPALz9rV2i+ameHFzTDTTnmWDA1mLiBHpNYr+XXHU9BbxbWneEmUrXfuXlq/QDm2h3ESzHqq6DzzEq4FWtjM9vJjEgpPPaZOnyGXUafczEPywvUUReW/22fWRR7OkWYHebF09EzA0wsFXRUmp02ZLp/H6mW0sSQEwhGd4Ku3Z1gMJzjf22wo4T13JzFyxsZ92e+Dn4v3sfD60mw4rj25i/tWhfPrxm5YXmKbfeZ5QogOW+hak3XK8eh0lvaKUjuKBqtbjzu2vMiRaHyM99tA0r1S4cmlDXe+I6IK69rjXuuvYXqgYs1sLGmw06dP/2zWz5B+UBlMMG0RWaC2zHNXxVpATiZeQaXcHwcqTUVBePygUhd4h9elRbMUVEjFI1AQy3izE1kU3Bx0M23vjtQHFIK11V3FiZDZBZxR87zAP7p2ebShwF7sSr7nuOteah6PrWQsvC6t23Fr4avE3UTtKFwXqfYlKXV2P3AfzCqcU0J2q/wiy4d+55Y957grIrtmwz3qUohxQs9HdB37gZOcg3bpyXaE5WuakgKp6ZgRH8DkB02PLex4nHY8+ThBxqIqFcKzE2o7cjdg6oxCcbfsPbI+mN6Y0zfOsRG0i/D2RUm3uPFKG+S3v+58Dz0Scx8mW2KM5ZU8Zw13ToFcZWRlU3fGlbNURpVgasJbOUiyRBVJP87qTYcH3+BJKlUULC7n6SfHjab7uV1bko5PvajzyIRXE9acxhXTsy14o91T+GPGMy09VzMvahR4cOcsncH1q8CPrl7YYn/WYMkUsMw9VuAc2PHGXhsiOwVJ5HyW9BjlH0/vybfUNvsqvH0pgjufvI66Q80htIGuX0Ii2AMdc31pf2vquiMr7u+RlgjQQqfuHPjr1z8i1nYkXT/1U4th962Uv1O5u77VKOK46+5xnI4r8dYej76o54XlOWvweI3j4ZFk+4eQEEwbKFCgQIECBdpqaUh7VAZz9FOu8WYtYR4T8Gl/wA7lDzet1W+HlnXra8lyVkvbD6pz3W36LFnYNhxeWo4nXB55QV7EFiLZ4xee9deSnqwGGa0lP1WOL27xRjs1RYlx17JQS5vWrhfI2RJz2DI+t28/NdzI3m2rNXWW1qBYZJ5stW31rPjtCMUeWJXzfftgM3u8QL5dYChbYr50lOUy5/Dk5Pt6bvjCu9jKjC+u5Wg5dV3fGlyuJ2GEnme7pZH7BdI8i7sofdOOdJZb2nHlY+csLg/IbWOglOfC1XcL28/jRzif6xEKrePcrgcLkujeY+HJ2ZUbwWg9CC6P/t5kvYHuSMxf7jzB89LZu7xjmpb5j7z3/j7hzoXRo9y5p/D2Un1tgf7neOlBG2APjTyeIuvtLVSnXN2y00v99Hd2f3/jnKTO9/Yuz64vb2t67OZ5EuF6kFrmwGWhtK9Tx9gejwypD2VPj9tGS+qwjIdBxWlLurW/fuXzlvRsq1/+vuQf8foUAN8CBQoUKFCgQFstDWmPyqAA34rSeSGDiDToCM4fmiJauB4UBqMxzUvTNf0AOabzyhl5UaRqSakVkLOAGANZ3fQ8TVtkXIB6Thio6/HM9L/CTZsrUGjAqQayet4JhT3XglOuZaFWPdP/CvfMXlvTs3EWyEu1Dy0uJ/EvhTpCyFNcuseWQS8KjsuLm+FrRsvaLdpYFKmmH1tYd/In42DKa87xuqnQNP4KDdj15KMXSsqhzoENTtYATE3fjL1r5Jw4p4wZm+Jbd651w4KBVj6leCrpi6nq7MtaecIveWC6pg8EBnfeVUc9vWFqfZHnykcpoEO64L1u4Db1QkvNq5OC64VrUNrJ3Tl0rF+5R2NyCpdf8p9LGi8nR4MF1crl+qWMGQ/iQgREmrbfRCExRzYoknuFxDJQTz2PUJQzPZl9NNtexwWj4QOOXNz9CIW7L5FyLVYofavngZuFwC4QQr7w9MKPti7sNXlp3Zlb3dgF7hkaUZRRxu598HSUe3bEAHbG05T40TRd3Ze4f7tenKKlL3czyVRHM+dz3Uu5RvPEel+osPodvVYcKdt0f2xUt+DF1inPcnfurZsS6JqmY/u/Lapb3r5UasOMy/U4M8aF+1mL/mSpzqPtg56+9r/HIUYlUKBAgQIFCrTV0pD2qAy2KKFaLV4aGy1ttqil6FkGPHKzRFJaR7TINA3Muy+117E8Nh/AmQlII79gCi3TNck3LRPNCHXPgHPJGUwaBJ/jebodm5Yadw0Fe47O9DWxXprylExvBR+EmxxP5qU/yrgzWlWNcno20/M4Oo8XPXsncJm8l74s9pJf7IvDc+MJaKEWiHXcavSIYZExXZGWBUHocteas8W7zEuqaaycPJc3zRLLY/XGacE38Bpa2rT6ZL6YfsuMGlp/7CJyLa4my8yrhU2eSnkn9F5RRl6yRybzlOk8exaPzEWqxqO0oxaXsEAws8yMVNgQ9ijj3GlDu/BSnQnKaHkRHuS12aD85GuVnzm9B4C0wYk2L4sWLgEArF7XBwDYaaedhTl6ORjL4/KiWX0qas05N++pg81c168Oh2nyng6qx5RrhXMi8qH3k6+aQpq7eqNeA0TI6H0R2a1etRoA8ODDDwEADj/kIAClLDZpw+YoyvqlF4C8pFxUnFNv7aWRFuRUkEjVGRdULc/cfYtNV6jnOfcJ6VshFOQ+3sAipAVKaTuQNrgXaPCJyggAEk2GcuNECPinfXhzlNPrIXPYbGa2z6h9TE6kWV7Sl+Qba7wIwUczV8bWWSX7uu4Xsn8UqQ47ZfHAzO3Lj5nMmVHlZV6p91P3WFe3ynsKAKRpoacNdutzY4l8CjEqgQIFChQoUKCtloa0R2WwRQmrzNmPPdOS599i9hA/QSGn9QhXnoprBFXj+avnBRFzr1qz2AgW+t31uigOSO5aczwnrDEHvtUdIu0JDkOH5NR7aOgFMlQS1yNky3izL7FABeMkoQciLpzrq1UDspUWEvuh0eDy9E+gqBrxS0rxCNJZotH8vFe+pudBhlklDkPh4gwoWBcNEWJ4iNVQCM9RmqHaKfIXPhIFuhILUayBmoyr2XCFp7gjXgaWDRBy31M+zbxuQbc0LobzTUwS9yyabXcQG0HjK1zrkLpW65C58HA3yoYmY2oYi6KgcJzPGjE+6DLx5kSuq3QQf0L6Uuwf8z1xN4C8BJLF4UtfXSIbYl1446IciGWhuEE0pOVtrSbtNF1MF0Sp9t3RYYD8GnWjp3PnPQUAWLBgAQBg9133MLcq4JUbV0OvFnUwpUXtrb0bvn8DAOD0087QOACrp8RHccHiCnURynXy/vf/v70vj7OrqPL/3nvfe71k6U6nk3QnEAi7QAghbJFFGFASlwFZFMURkWETF0ARcUV0REEdR8V1VJyfCoKKIgozKJtIQFkispMQCGTfOp2lu997997fH3W+p27V6yaLQBbr+MGX927dqlNVp6rrnDrne/58FwDgtYdON+Pg4Q0RR4WftLTmERCJeTZpNfyu7jWWo8eeeAIAcNQRh0tbEgWSuJYhtsF1z37X3O7inj/fAwA4cvoR5vUkUf+uTMDEJAAO5UrZeZfAlvxeZtQfMVty1z9MITkFhK1BF4/s2CXS1sJFiwAA3d1dpoguCuLDSF2Ud29OuA/U664fBi3OLF+uNKkcq5xrdBLln5Yx+kXS0u2uX7seDFE6aHkpE+uIc5FlWkdfXx8AoKW51eFFu81+yZqpc/02RAe5c5F7OFGQv3vlpgqi3MVRUWuzJ9ek4KMSKFCgQIECBdpuaZu2qGyuj4pCiSeeX4PGk4tWn7h3nRZuQU7moh1kiXvvHvkhKYquGFv/6RJRJHknOzhSa8LYdkKkMymXavnk3YtISVztMM5iVaU1+ZgH/qmGAmKrJ/TYd0/ACtUcu5gOqnnQv4Kou2ls+SQyrfQj9bAcCA1O9FOtw9PEFC6c2BY0E0h5at5ZqV6Al5c6yD/rVn75u3s3bbUlF8siVfwMONRYT2F+9brX7Rd51ORzfF6i1cK7w5am49itx1KuRpy4RLlmWVov3HGxfj/e/TMjibxxo4aZ0W+mMM6ZXVTSdp0Me/x6FhW4/VL/IT+IQr+7/hJArFEOyo+nwS1fvkz6wYgF1w9GUXJzT6YYBahas6F77p0FAHjXv50OeBEScerLpVhvEk+epe077zIWlalTpwIARo4YYQrkukAc3uiblgMqb9yXlq9YDgBYvXq1Mx7q70H5J7P0r9H55jy7vin/d9v/AQBed/iR8loMZNxbxHIUZ04damnw5n3BCy8CAF5Yaqwghx50iKlH/SNy7zv9RSyxRlqMr7/hBgDARR+6wJTlfHK5kgdvzXCvib299oGHHwQAHHDAfvJ+pJ+N8keLCf/GyAPxH1m2bDEA4H6RmZPf/g4ZH+5NHB7KoLdXpXbtccl88z//CwBwyUc/Zl6FSxrFk7iyaH3vMqe/NVrcvPnXvylxpLJmVxbnG4NS8FEJFChQoECBAm23tE1bVDbXR0XJu2vu6TGax/d/+N8AgCa5uz/itYcBAPafdgCKL/jofDYngnepq2/lBfOF+46PAqinXU9pz/w2Gzvl8kDK8gbLh3+yJQuRx5ufT8/vd8MM6OPBeHXHTpNv+eOiXwfv71CIvo1z8RJ1eYiLQ7fFUoPP1VB9Kz63fkzuDbJa63y/kMh/7kb7WLcCr1xxXBh15sua2+3Gu2rf98afYA+XwoddMXJO/yfXOmXb8O6/vXwiQAOT7vs+L8V145eVZ/0D/QCs70FBx5MPX9DdteTfx5OeEB8Q5HnDurNrynwOvX5dC9SyZcbqM2LEcIeH2Cs/mHbqIyY7PmJF/oeQJb8eX56XLl3qVpTnDWM15K7sPV++3Fh9lrFOf//zol+8qXW+8Cftr1+XYry89B7q73O0IO23r4kSK5eKKLGeTESDftW2+tauBwD0r+tzfrf7Ibzf+dVdN8hz3ZftfLjv+oLagOLs53qTz9tuuw0AsNeeewIAdt55Z6+cnW/dt6LBx5IUfFQCBQoUKFCgQNstbdMWlc3xUYlyG9WhpzReI8pJfOqU/QEA7e3tAIDenh5TjJoYkUwVuNFDmVRERPN7rbpe+Y3r7smYGA1Epk08f5A6UW8H+qQNQ7EFPzD1iE/Hvfcaj/yD9p8qBel1ngGSDTTlnaQezoUn9dUwd5Tf+4GxLL33Pe8RZgThcABsVN4nK+YfVeE5qVmrhs3oacrSv4Ppz30tlXgRtZqMHf1BMlfTWLzE3Pl27zDBtFV3NZSsnmGg6movkUQoRILoqPJAX4vUjbzSTCKMmmHK9sybb4nAqPcTd8NGv2Rah3CuAWeZUxfHNK26/fV1EEa75KX1Tn/jAutE80xkmUcKZyr8phzjfinPnCCuFYQZetOoz+GR42nrkSiwLFMrC7UnZsWt962T/jpN2H7JnMTiYKNp7jnhxBHp73PathaJSKN3dN5lBpevMDgqvb2rpQ43ak1JvhIvB1KPWv+8tdc2cqQpVh1ATPweWrO0DtMW116UWn7NC6bcipXLpLgsMkVTFbkQXJI05x4mvgKZfVYX7Zb7WVNzs/CSOZ9JRMuD2/E617WsPYvpY3ZMWj/q9X6pp6yyxTFKKRMD3hgrwrbhm340jETimvStGjXxcUp07cnw5TZaqyb9/tvs2YY/lWfZtxjNqVBVmdYBADnHb4B9MQXXrVkLABjoM/1Fk13DmbeXknHub5G3VyxaYsbuhUVm36rK3wbKRaL4OFIPEXmrlBu79hghqgZO7Q9xnigHIjP97l7jG/4ZubV0sVkn48eONeV33ME8Jy8D/YowrdYRL6u7T8FHJVCgQIECBQq03dI2bVHZXB8VvkMvbz0NEj+kRPRYOYELDgfzTBDbgMECMePybZiE+V3OgbFGE0BH/Jbf3QwAmDJ5MgBgp512cnikJlFK5D6UHtg8uSvAgPlYscyc1H/965sAAAdPO1De4+ux8leOXH6tJUi+y3jcfcedAICz3/vvpr8V3kfSC7zBeQEAUEk85iLg1v+7BQAwfuKOAIDJr9nH1KnBMPSeN3Ukio7oeugTwZNawje+/W0AwBX/8QVTnumFZA6SCEg8V/TcixAqtxiNsyZaAjEv/HEpwZWLKHFxeCxmSsHj34t+KXlRAVBvf2pggmGgdbj9WSia2PiucVKP17dCFEiieVGoAbpjmXhZj0sxMS9cjYdRELnXP8LCEk+Ir/3Xf34dF154obwsdTPqQ7E7PMcWzrtMIHmi5ahScddgppEZ7hzkyHV92rwjifTP1L3LpF2c8dBoJ0/TK0euHOtIe+JNeYmiWNc3h0ijc3T+S/yHaVOjolx5V0sj+5lwzUbCmzvmeQkoaxSiKbNosZGVZsHqIH6QNZi548S1V1aZSorFNAKtJHglJWLdRCjk4KHVwvAXe2PHcrRi0N9jjVgt0GBhFewOL7uyRq7khTplbAbEQqp4KWKNsXVwclzZscYBd52sWt3jtFkWPKEoj21Emcc29wq1cujfFv7J5d8I8i77gD6W9cB9z/sblecZShw7wcvRPdSLrErIr65B6afvJ6hC67alfxd1H4gQCw6U9bGE+65HwUclUKBAgQIFCrTd0jZtUdkcyiOLCpoyd08DXoZogaI5VOWErrl+xFeFmrjmAtHTpdwdis9C0moQA+sDNT15zpnzLACgu9v4VuwyaTcAQI33/5o207RVGmbaypnjxNN6F8vd57p1cq8ceZgpWRWJIHVGml/GzW2UZeY+OSq1AACWLjX3x6k1vRhe5M5bEVzBx8yFISd5KZdW65g/f4H5t/C13777m7L0JZE6mOsnl/vlcvMw01a/29aqnpUAgNU9a8x7Ynmiz4LmLcmraCpLHcz9ojld5FPuXr/wxSsBAB+9+GIAQBMtaRpVIJqFtKUowmphEATIFjN+ixcuVA/6U089GcWOlioyhgMuCi6nn6iqmg1bXvza178JAPjS5y932tLxKeDrEEclEYsHoyFo4bv/foPlMGm3XQEA4zrHOnXoPbncXStiZ9XLWCw8rhTN7qmn51hTkIfMyTVTpw+GijERXWWMBRX1U5+6DABw+eWmv/SjKsv41YlETB8f1PXflWYzNtU++kmYupubWqQNInDSZ4fzybUnMjjMRN6kA5KdV1pat8742/RLX8qVikXK1bER3zOZJ0KLau4jKU1ZGt3RCQBYvNDgiuy5+x5Sj/gwiHUkErVYM7gDmrm5IgillKlxXV3SFp0ZuJDd/U8zsmdEuTb1pCJb9FXJ1CdP9pikrP5b2oT0u9JS8dpgm6bgmjVGZlb3rpHfiStF6y39LuQ9avVSWx7Z9UvZItFKpdYIYg8JQmtaZVZod1xKlFGZy/Wyp6ohgMPX1KzrTi1jXltJWXy0iCDM9SBrpiTrvNZPxz95TuwX8WUqEYG6RsfIXOd+/vz5bFw+XYwe+lByr6grCi73Shc1t79ffHH0JoCyJn/3mpvtHsC9QpG3B0em/UcpWFQCBQoUKFCgQFst/dNZVIrh3k5cePEX63LulFPt34vPH6oeP0MqIoAZeVuazQmZqJH2TVfTstfmQ93rmQJr1xqNRO/LvS5lWaT5dSzRkuTdVcrTtevWyu/0FvesPQ1oAbw/rXu/A81iXbG5W3iKt3fNxToyz4rj9/fFBQukvx665GBz4M0j/HkTXuY9a6xcPXInPbZZLAzUzOgP0HC855wRydf8OufpOaoZ27HxOuRpabnnsW8fmI+nn3pSvnqYHy4riKK8gFXiLnPy98ijj5rvcsc/bsxYpw5bJ+txeaJ2yF+XLjZWvbVr1zUwZHFihuDbGx/6TTz9zNMAgBotJz42RgNmRFTAidCfDF/iBzGMmreHF1O4RQegYKINvKr9wYt8yPO8YYXZvWLQqgrr3YxL51hjUal7kWfab6+C4r6ReWPBbMALF5m1opFIdPPwuFXuG/Yc8f2QyMNFCxcIj2LBKZds5Ay8tQKf3MW4avUqAEBra4v2o1jA8uT5xxXmn3Ju8XGk/9q9wX2Q/F5yX+NcaF4mb9DVeuf87K0NnS937XEMLZ6OzwXfc/cUv5kcke5La2Tvb4Ag0q/uvuTvKfaPhPlYuXKV0wf/j5FZw4MLorVsvrwULCqBAgUKFChQoK2WtmmLyubiqGj+ENSdZxnvaIlDQHRNG2hv6pD7OPXx8E6/9qBKCwwz4GaI5Z3+Ku8B4dSl98SukaLgD+FaJFiOUT8dnaMcHnMbfoBU3lGdJPI0ZQ/7oXtCtzxn215/IjcXkJ/hl5mesyjD8lXG32XUmHbzm46Vi7iaqce9+W6zB3N8TPlVK019XZ2jnd85R5lasWy/MwVf8O5m5XuT3KcvX2ZwBMaO65T3pGr/zrsI5gB7n865XLxsMVato8XMhRi2subOJ/FD2F+Wq0s6bGKV2LmQcowaKlo/YnceEHHeSs7vy3tWeDxSs4TDs7WUebIoz1evMX3de9/XWAcZElMUMUN35K0VJfpByHiUBs8n5K8X1WQjNMgfv4+QvDk5M/3SeiUTTOvmsJHDpFvcBzgX7hhz/+Cdfpandk/hlqFj52It5R7fich/b2+P05ZmFycejbdeims4ZhZoWb/LZE8Y19HhtM35mvPMMwCACROMn1xzK6ND3PHLCRwjc9YtPi8oyoPOpxtBBC8Dr+4Z9COS+W1qoi8L5YN+Fq7Fybd+RMh13a1cbeS4VXyTrM9g7ryT6t7pZcP29jeWpxWkX3BUWsUHKEOm0XqKauztpdpfaWu55PphXqIh12/kzm8Gdw6ATNfUsGGUV7htyvcodv9e2fXrcFxY/+aX5cuXSH3M04VCPd769n10PAo4KoECBQoUKFCg7Za2aYvKZlEEpOKxHEUV95F4R2c1oiTK3Sc9u+s8LUt8/XJzl/f32X8330UznThxIgBgv30NRkom0QLIMr0fTplNkyiPNYkQ8k7mirwpntqqHHin0lWrjff8bhI9lNYEMVE00NU9q/G/t94KADj55FNMW8RV4Cle8BcG1ovmIFEeeY0YH6KJVw2vfWuMLwKRL/0MwFm/9VBvF/TOFctWSr/MI45xlrlajWICyFwxgoqn8fWSM6N7B4PLQg0cdasfmPqBLOb4y29SJpN+3XnnXQCAkcPbAADPzp0HAOhZ1QMAmLzP3gCAESPMc0Wd9PwpcnFeqQrPvStX4TUStZFJBAFxTzJBImbEgmqeMif1qvv7utXGv2K4RJBRTNSDX1GCrVxQkmJqqey3jEcuEQSxouFyXcDpF+U9HWBElpT3sqwuk8izcaM7VWaUF5H3x55+DACw+yQTaRQTi8fLol0XK0dZ8DiWLlgIAJj3rJmbJx97CgBw0403AgAOOtDgBh1x5GHapo6NrOsaI+iIN0Ljh0QzXHWlifrae++9AACnnvI2U07q0bEVOVq72vgGjB5trHr1Wl39uO77y18AAE89ZXxszjjjPdKmp4FTqxWehkmUGyjvsi+k0ubHP/5RAMAlHzGRaaNHGWtJVvRRGTAdG9ZiopVKxH2quxaEG2/8FQBraTrvrLOkLuFJ+k3Mor41xrLQznVAWa6lWLbEIOo+O2cuAGDaISYvGqNDnn7S+FYtXWAsCsxPM/fZ5wAAk3YVbJsa5ZfWP641VefN7wU/i5wZ5SUislmiW3JGYHnWnpj+TiktKJxX80FUaPrTtbea8VkiFqq2kab/0UAVUF+ivFiFtkUUbGjElPkcNsLsh5RRtSCqXwycf2SMUGLEVZ6iJrmrKoKnontg7q497kuMMIw8ZGlrlTN7T7Ps8wPMR1RjOdkHqimymmeVoa+hhy31clGwqAQKFChQoECBtlr6p7Oo5CggdjILJlEh5SSpSLNyvC0Tf0SwHYivQJTNec8/D8BqKG944wwAwOT9pgKwmBB5mqohRLUSOd2XBXmT991f+uKXzDvywv5T9wcAvOlNbwYAzTFRFl6JYEt0wVj6xuiYH/3gh5j5xpkAgM994QoAwHnnnQcA6O42vijMBbRIMpqO6RoPAOiVk3Vbu+H53nvvAwD84le/NMMhmvWZ73kvAKBJYv67pN72tjY0tRpNkQijiigq2t5VX/wiAKBH/ASOOOIIAMAM4Tkuu9CPy1YYy8yalcaSRJRJomfO+vOfAQA33XQT9p9q5uHUU08FAMwV3IHf/9//AgDGCn7I008ZLf2o1x0FAFgpFpWRo9oBAA/85SEAwN8fMdEyx59wPABgdKfxZVGsC6JyViooydwTyZFa3EN/+xsA4Jc3/AIAsMuuxsJw2jtPAwDc/UeDv3LM648FACxeZuZkguTdiKW/rDdTnAYZjiiyVhr6/ZSo9YiGJFrewhdNZFIi+BLwLEWKrlk2/ZrzzBwAwFe++lUAwCc/+UkAQI9gYWRxDIj8UTG8+ptfAwDsurux+H37O98FAFxwwQUAgEmTdnTGh21zbOvC0047TQIA/McXDBLxwQcfAgD4491/AgAc8S+vK6CjCt6N3Kevk2ikAcGsoAw+/PhsAMAb3/QmAMAz4rvxy9/8BgBwyslvk/IyPKVI+musXHtKltkkKSGXefnLXx8EYDPQcs0ceuihpqyiABtKBQNlmGC2cP1yL6Ll6OyzzgYAfOfb3wEAvFPkZZfddgPEMhRR1mTMiBtFdNtZ95q1ccwxxwAAnnvO7F/f/eGPAABnnnG6Mz5xxcjNilVmbbaNMn5wJcj+mZTwc5Fj8j9hJzOfO+y4EwBg4qSdAQB33W1ykXHPmSdrcQdm6BW0W1pBEvqV0HWF+FAFVFnivgyI9WblKrM3xGUiTLtmiigmkjAjbCBtiIVQ5Ib+J51ju2Q8XOTyuFRCVkQdL1BDhA1NAtL2+t51wmPZ6ZfvF8bQM1paIbIX5xFScVdrER+VmH/XPPtDTKs8kYiJOEurs4x1KtaetnZjpSNKNj8h9cSlMgqQ2U5b9RD1EyhQoECBAgX6Z6Nt2qKyObl+8shGEtQ1WkeeeR7bannh/SPLy2d/n7E0/PJXxpLy3jPPAAA89sQTAICv/OdXTDlBrtyhezzefcYZ0pipo6nFaLFrBeXyyi8ZS8rZZ5n8OuPGjgEA/M+1PwMAfOKTnwAA9Ala4nul3Ipl5o64Y4w5/VML/t53jeb6iUsvQWubuRedNm0aAOAjH/kIAKC9QzQk8Reg38fuu+8qvBnN8YUXXwAAPDT7YQDAl6/6MgBgfZ/h5epv/BcAoFfwKpaL1ePLX7oSddFWaFEhfz/92U8AAIe/7nUAgOmHHgwAuP0PxqJAbX2U3Au/7uijzJhK1BStN8xb86tfGytPTbLVfu4/PosnHzd+Al/4krHa3P7HPwIA1q1dK2NjxnS8RD+sWGl8jQ482Pg93PaHPwAAli4xVo2T3mZ8fL4oVqDddzNWgibp2xn/fiYAoF6vazQD5/spsdr89a8PAAAu/9znAACPPGr8nE4XbfZt73wnAOCXvzT9GSVa7A7CY+7dQ990828BAI89ZnxA1vSuwTmifTcJZg/Hvn2UkQP6L7UMa3V45NxcffXVAIBeQQ+dKpaplTKvX/jCfwAArvzylfK6qW+fffZR/n7/e5Pjaa89ja/Om044AQDwL6LNf+pTnwIATBhvNOxjXv9608/xxnLUPX58kTXNlbKH+P6cLr4f3/7WtwAAP/rRj3HAAUa+d9zR+Ip10AIgGvbqVca3jGrvfbOMtePfzzY+GoceZqwes2WO/v0883tXp8mvdPa55wAAlorPQusw4xMRRzHmzjM+GjvvbCwJrz/2XwAAt915BwDg//3EyHuzzMnMmcZiOEz8Ktb3MeMz9xwzjs9KvSedchIA4ANiifrmN78BANh1110VH+aEkw0KclktK6zRzM/DDxpr3rnvfz8AYOoBxp/kmTlmnXzmM58BYC3B75NyPZJ1ukX8pEi9vb1oF6vjSSedCAD49nfNfOyyk1kbc597znzfxfii7LfffgCA3e7fTfpJHw8yaz6ffdb0+7e//p1pa73ZJ2mpqlWreOepZq30SJb7CRN2KIygreu5eYaHu/98LwBgsfg9HSF7zwqxJB933HEArCVqtch/0S8GMBaMH11zDQBguVg8d9xhB2eMDhTfqXFjzD6+VvacUe1mDRL3qiqW9WfmGGvlk4KXxHqnHWTqWS6+QEcedaQiKo8UH6PMi15audKs059fe50ZK7FO7TTB7O8vLngRAHD88cYyPF6sXLGXK0jrZWRi0qxZne0o03dmcGTakOsnUKBAgQIFCrTd0jZtUdkcHBW3AvORSTTA//zoGgDA62eYE3VbezsA4Mc3GW11771N9MfIEeYedt48E4Hwbcng2yUZbd/85rcAAB595BEAwHCJeFm1YiWu+PznAQCvnT4dALB8ldHeqTnzzr5tuHj/y/3rGWKJWbzQeM13jjKRBrf+n4nkqYhlgd70f/3rXwEA//qvhpemStlm2xBL0Ve+/BUZBjMQa9YYTfMh8cGoxO459p4/GT+ACy76MACgJv40LaINnn/++0xbZaPZVeW+8stf+hKmHmi0XOZNuuOO2wEAe+/1GgDANGZ7llP70aLlvO4Yo2FDvMyv/rbR1OjZ3yaaHPu9cLHxtzhPrAlpNoD9pxmNcZREZ3QJAiuxDL7wH2ZOXnekabNzdKf0z1hl5j9n7tHPPNNYryLBIeFcPSs+DfQH+oXc1w8MDKCv32jIq0UbvfbaawEAn/u88bFIpV+T9zbZpH/6P0bjrghS50/+538AALfd9n8AgKn7G020Z2UPAODK/7wMADBd5OljF18i/U7xObHWjBT5oxb0kYsvMmMoUV20BK0WjfR73/8+AKvtT5q4MwDgD3eZOXtSIjhOfcfbAQCXXyYauNxl/+amm/H448aquErqfOfbjRWK1soWydVzwQc/BADoHmcsgV/4kvGfomXhyCOPBACsXWP8Xx573FiMjjz6aABAq0ScffiCCwEAN/ziF7j3XqMxr19vLGG0LDHygv1d2WPkfViL0X6ZFyuSNbf/lCkAgE990lh9usVX4Wtf+08AwE4TjdVk1HDBZwFwy80mK/q/i6WTETYzREu//uc3ALAowN8TX50LL/igjI/5WC7a/bNznwUA7CbWTWrLo9raAQDnv+98AMYq1N9nrA2zZxuLyYBEecQiSyvEItghVlrF2ZB9cA+JvDlBfK+mTDVr9qv/aXyRmPF8R7G4Ep/kF9dfj5Pfaiw9Zckr9YHzDF833mwsISOHmz1z/gtGi6fFgb4q3Be449wlkXiPPW72ovd9wFh1uN8vlwizec89hxt/bfx36Cs0RqwXtKQsFWvztWJZuOQTxoIaif/Hj64xvjmvPey1AIDvyH7+PhnbFvHdUsgjkaOf/OQnmCKWoSmTjaz8XSyja8Ry8sMf/sC0+VGzLoeLrNQH3OiYOWJJ+V/xmzvxRDOezWLV+ul1Zt+YuKMZ+z/ddTd2FUvuOs8Kt0ai0b7+NWPhvkT2hBGj2827MrZHyh773e8af6cz32sswQtkHxvb6Y4jijcQ6oviWlR8nxXSq46jcvfdd+Mtb3kLxo8fjyiK8Otf/1qf1Wo1XHLJJZg8eTKGDRuG8ePH493vfjcWLlzo1LHzzjvrtQ3/oxk9UKBAgQIFChSIFOWbeLy55ZZb8Oc//xnTpk3DiSeeiBtvvBEnyL3z6tWrcfLJJ+Oss87ClClTsGrVKnzoQx9CmqZ44IEHtI6dd94ZZ555Js6SuH3AxPITYW9D1Nvbi7a2NvT09Ki2uLFUX7IEdcmKDEZMiGqREMNB7tk00sTL7aDRHWVmlXTvkxNFeDW/x03ifZ5lCuhHpNkbbzKRBWsFB+X0042PAu/xiOlSaiZyI13UhRc6tIvH/y9+fZP5Lh7cM2aaCKSRrc2IK2WnDkZY+PgQsWhFf7nPWGW6uo2fwN3iN/Ju8cVJvXh8nnqpuUdSD/JYtZFvfdv4PbQMNxrVe959utO2Jh+VOmLpN8fBA9fEDWKJ4t13m+AT7CbaRpbWkRD1UubnF9cbi8cpcpf/4vPGYtLaZvxgZj88G4C1HLzvHKORM6OpYgX4Hu5iHXn0afNeEpfx6BPGAlAVbfeE408AAIwcJVFfjNZh/ylbMle+X9Qvrr8eADB5ssHoeexpo4mdKGtQ75HzSGWM8/nQAyYS5WaJZjnxJKO1PSW4MSuXGs3zuJlG+6f2plnAm5g1WiZTPjlXjI674YZfqiXpbYJFUiICq9SReVlWiZtTF9yJSPzIaFF48UWjiX/ve98DAFwnGibxhzh+pThS+evtN75T373aaIw7TzIyQsRZav9HHXUUAKBdrBTqq8N+MwOw8DxPMp//9Gc/BQDsvvvupq9vfwd+JBr0mWe+V8ZO9pKK6U/KtSZr7/7775fPv0pdxveGmZnp03D6u95heBbfFtajcBtZrvP0y18bn7nRYhl8cbFRFDvajK/OQeLDM6bbWIhoUYmJxCv9jGWunnjscQDA9df+HAAwZX/jqzT9UGOB+P3//g5nvPvdMnTu2JXEWvWfYpU57Z2mH9f91MwffVcY5UfLcF2stW99q7HupMJbnDAnlPCcAQ+K9fhzEgl22OHGCvfmt5gIydtuMVads84167hFrDvcx4msnQruynMSxfmt7xnLImX1pBON/w2twUe+7gjlm2NIxF3Oz6IlxgJ+g6zbx54wY3mc+GKN38FY5R580KzND5x/nlMf11bCfVB4vuX3t+I3vzWW/uHSn/POORcA8IsbzDzR4lthhKDIMWWPfwf61pt18q2rjSVpvVi3Jk40/jZHHnk4AGDnHYzPV7lSQa3uorrTr6X2pLGkDtt/f/mdliNrE+nt7UV7eztWr1690X+/N/nqZ+bMmWoS9qmtrU3T2pO++c1v4uCDD8b8+fMVCA0wB5MuQjFvISopPLf5TuGgYyaFrhgKZwrIpiMvlBj+qSFb8pXo9XwvTuxvsjgWiEPTeDF9M9w414VZ8epiiBlc3iWkjsQ/gB2Ez87qeqjhgQyJa1DjppcRxV6Eb71sms1tw71hiJ3+kmcCTGmyNsQa29kxxly/1GjqpeOyhM6yP3R4Zt0M+VUIeh6OhMcFLxhH370lnDfWUNyyDdkWQCNeffFz4i47AwBWiJMoQwUZjh3JgTTxDg10RiVTebN5b7/9zCEiykt4Ug4tVenHyJEj5Jn0K3ahwvkHWh26FfnJUPcEc2h8TsI699zN/JEsiUDksT08MXyUB7RpcgV24EG8ZjNlH3vcOPg2iVPtJHEE5eDnDNcUVhQwTWWPcsTwz1hDJZmMMtJ0ClJH5Moe0xIkmbsljZKD57MCXz5Swl+hBxMKK+vJkQi/oypm/j7+sY+ZV6TQDdebjZyHgU6Gl+vhSXgpe8kcZR/YQw4ml112GQB7aHp23jzd49QR33PILxF0TMJMDz/cANTtJdcqvBq4WZyjY4VIEPBFf/x0LUPX8ymnmis5ys5/fu1rpg6RkS4eUKRuBSxk3bm79vZ9jbme3fdy099f/cocdFeuMNfWHaNGaX+ZdoR7A/c5iuW4cebK64ILzJXf7XeZ6+QVK8y11HL5PPM97xGe5FDpJ8gTx/84Bg58rXF+vvBCc/03Rw7e991vnKQ/eNFFUpaHO26Ebj8hIeO772Hm9yNyxU3Ygj/ddTcA4JSTjIKz9z57Qak4D7B76Q6yXi+88AIAwA/lmomJO58U5/cL5QpUU6jIN8It8AoxkTl708w3YuedTag+Heh53X/ppR+T/nL/dfcQAntSJlvFleEccRL/wx/uAAD0CIDp7AcfNuOyi7l+zCL7t1P/ZkS6OeCVoFfcmXb16tWIogjt4u9B+uIXv4jRo0dj6tSpuOqqqzQqZDAaGBhAb2+v81+gQIECBQoUaPunV9SZtr+/H5dccgne8Y53OCaeD37wgzjggAPQ0dGBe++9F5deeikWLVqErwqAlE9XXHEFPvvZzzb8vrnhyQRXSxVCXjQLJm0TxapSNk5otE7Q0pDLoYpmWMIia3JCSQKlJvMWc6WVVmtI9WqDVxnmh74+SZRGU4niOdelDkmEJTDmqXZbTK28MhCtYG3NnNgt3HuGRJwnI3mZVwP+9VFM6Hype7mE67aNMKbjcpPhhQ5wFt3ITTiWNJnxqw9U1arRVJHf+gTGPWaCPGp3vE4wz5sEvprwzwWXLABAzwrjpNosieRaRrRLPcJRPoCmsuG3TujsiJY09wqnLPwOEFpeLUPSzZggZgIlzTwA1EhlDpokXLXWN2CBnAiaBIJGNQtPMg5MQkaHT7lGqcs4ULYYnvuiWJCmHXRIkQVNOAhE6iyYiLxTGdCEZyIbhAYn8F2uMih1igxWBLSPc6EJIIXHSqvhube3VxO/qcO6tFUWi0hNQvb16gLkTcZHLInsz/wFxkmaFsUSZYtzqtdwdes022raqvYNSF3uOu1nagtanooJ3wCFR+f1BWHMufZ4hTRcwBvnPvscJu5Aa5TMO9evWJYIR84wXHI9Sqx3lRaG/kYOr3qtLOMSi9WD11x5VJinFnds4lJFyxTbjiLKM2dB5kDmqtTizbcUo2zSeXxM13ikUeLUlUkSzdJwXllTixeLsczFUeLA/rVvfh0AsJOElEOt1MIT/1TxSlw4TiMoENkbZhrAvgX/LVc2tBjxU8x5sThyE+DN7n9yhSn7+vgdmKzRjOfPfm4gInbehVcgrSrHNlGpfNCSVJa9VuqmQ/73BbBvmFzHkXRvku+JiGTcbOZK94ukjn32NteET9FqyzUlFkW9XJV0FuXh3I+FZ1rtpSSDR972dnNde8klxgn3nQKVQHeC8rBW5cOGRKtPAwajfzQ8+RU7qNRqNbztbW9DnucaFUO6SExxgImnr1QqOOecc3DFFVcoLkaRLr30Uued3t5e7Cj354ECBQoUKFCg7ZdekYMKDynPP/88br/99g06zBxyyCGo1+t47rnnNMSsSE1NTYMeYDYnPDnKB/+3qdD7HOLd3L8v9d73jB1uHd5nV5cJz2MSOpubTVVk7w2fd2oqbqnhTP2tfbLv+2z548DvYyXM7w+3mzvLPcRBdUM8+A0Mdn6mz0mklhT33aHmxq9rjYAxdXYba0Cs3riiNeZRw7xG3qdqFvLuosXGAa6LYZzqODM4Tw3MF3hfK+B3JTpwDyVzDeTV5fFcE5+NUe3UyCgvxYKD1+FHFHaLr1i/hGM38qSC7/3s+jjws31Uu9W06MejOfg4L25VqtUPITvUdseNHTsoaw5PQ65Ptx+jJLS9oXzk8hL5a89re7hYiRYsWoK9xGcIQ/THJ18GffluaW6RAjLG+RC85XljZKh85/zWBoyDMx3b1WihxQfvp9/fMTIHixcZK9fEjg71QRlqO9a6fBGSF9bKOu4aN27Q/uk4ef6Epqgr2EySul6S9g05B2zDs5D564a804G0RYD13KfaIe+7P19iIRNre5v+bXTXr50Tjxcd50jrHiaWwzniWO/zTetV5HWvYV/2lsFcAdubQJDJoiD4e6lXR0PV/2B48st+UOEh5ZlnnsEdd9yhmUVfimbPno04jjHW34QCBQoUKFCgQP/UtMkHlbVr1yo4DWBAz2bPno2Ojg50d3fj5JNPxkMPPYSbb74ZaZpisWinHR0dqFQqmDVrFu6//34cffTRGDFiBGbNmoULL7wQ73rXuxQm/JWkKC8c+nLXR0NTdHtwzqmoILHcE+fqmc07bXkvcj22bfZwqQc5Ymkzk3fpPR3HnArPb0TvqHPnd8t77nxfscp4zSeqJZjf63luwxDlHxpuy35pvw01iebA/nSOHiXfebfLtPGud3nmgf+kaFCssXr9Wqe/GhLqJedKdcypBjI01pQY2W60ibXreoWXulNPluc6/uzvvvvtLW2SG3PfShjwJUuXAAAm77OPvAenfKTe9EyPLo8j+jyIn02cY806A740MhnmjE2uc8D+uHLBpck2+EleFguPjLjJLBqV8mCVOQ0cJ6Pmw/N7aW7h2KdOv5RnGVv9jFx5UR8tAOvXCuy4hBvnfh3e2lGF0gOSYh8eePhhAMCbBYyR9WiftLzV1vx55xoZqApAVsYQDfqLWf4LX5Fy7OGtF6/eNevXoKNztLyTO3UwakP9g+jD49XBsaaPz4CAUWpfhDeuYXdv4vxw/jg28t1LRqd+BZSHyJWL+lAmOPm+4EXjozL9iNdqG7oHePvWEa87UtrgngGHt/kvPg8AmLTLTlKPNxe6Dug3Rt+7XGUpk/7XasYyOCAhzuo14VmvVM5zd05YD3SPNm3Of2G+MwwZcl13ao3RdeglGfSi3p6ePw8AcOTr32Ce6l7McfTHnPsBn2daZpSkQKlm9lmxw1lEPzf375dPkWdxfF784No62s3PIotm/+c6ZkvufL3ctMkHlQceeABHCyokYP1NTj/9dFx22WW46SaD47G/xFGT7rjjDhx11FFoamrCddddh8suuwwDAwOYNGkSLrzwQscHJVCgQIECBQoUCNiMg8pRRx31kndMG7p/OuCAA3DfffdtarOD0uZE/SAC6hKtEtFvwLtgo9d7qpEJPC3y1CsnVfErYbQHtSG98yQgFuvJMj3VRh5mA2H8c03JTXVOLASShC+y6qvLu/w+TKIhygpyRl7qQJ0e6HyVmoSctAkQJh77mYBprVnV47RRZ7RERi1Z6qOGzVBzuROPcysXCwTgaQdJjEXfA72bpVWKmrgkF2QEFbUFplxvE6/59aIlpxLpEqv1IEXqjR2xR2jFIIx9JJpHj+CpMMkbJKqFUxbTv8ZLwCXBXqgSujxNMWY0cWwYSULZERmkFkSQQSlWo2x5Wh/bWLfe9LfCSLTUHZ8IkaZtp+ZMmfOtcoTObxsu9+UqIG60RFqtufVQlqmJynPU6goyxTHmUNX7pUxD8jKpU9eS4b2VkWrS5uGvPVx4kXVRdy0qiKETVZd5Z5p7RussXWai2HYSvBjKObx1ziiavCqD7AMkWgcTQ9UUTYK5E2lUn9RRY2SY/KwWFZf9TNbWUkmV0SYJ5ygnlEWVc8UqyjXyLx2gBs2qTRkmk7SySO1fWNFoMJkzWXu6nt3AFuRiWWwqlwvrmENDXswcHLD/lEHb5No78/QzAADjBNOG+55GYsnkKdhi0WLF7ohMHHqoSSfRTx8V1kVcLJ0Lz3pBsDaph99HCL7Q9EMPleokmqo6gMhPzkcrMtcrQasyrkHT9usON0Bxe+w6yTyX/Z+DS9sXI0rThDwVLKfy7z13M/gmTLuiskLisozt3yHDqivHTHIo2z+mypyxj1yyWTW1a4YyQVC5If4eh6SEgQIFChQoUKDtlv7pkhLmKJzoBJOARG0+lfu32MMGyFV7FW0u5n2pnCbZRsT7SKknZj1x4U7PPBsQrQ91eTshtoFolmVqitQkBj+N8k68XLFYFqa4vFfKkBNx10OUtVYZuT8WTSkW34Ue8cgfJ+nTof40HqoqrRiCGULcEeS5jnlZPPJ7xUeFOBlsE76fD+eACI3ssKqL5peWZqNNxETo1bvexPLrhTFYo5T0Q5F4DaT0OEmYpgiWRaRd01jxZ4vKyXCKOFLsmn5JHJYTuyfy6xRmRAMjL+r/I22MHmOiInrXsz6mKXC1owwAStQ6qaZxrN01Qz+y3Y4QSxMxTLRE5vKu2qOrkWbiGLW+OoBlAsfPuijPaEg/4Fp9LKaR4XVYm7HyMElhLJgw8ORCWYrq1o8DROt1I8vGjzdyvH6d1Ck4IxaW3rUoqQx6y4bEuVq3br1i8dApS/1YvLUXq8+BoXpGpGbzvE+wfCZIqoVcfZG4NkXOMxViRUXlPFnrhnl3rKDC6vrn+lSi3wHngHuHu/bY5EJJa5BHCXLupXlNfzPddn2tSHbrMeUOFksZMZgitU5Hbnlv3Zi2aCk0zybtshs7bn5XnyruMW6/rDGE5YTnhLwY+tznLje/p2zPIm5bXuQzc32KOOa5mEzf/CaD+ZKUabXNnfcVZ4f7Adw5y/O6DkG7WKHaOzpdXviPkrtf5bpvKxqN1Mlipty3vmkSwGo6F/WLTOy681aDn1VEeXm1kxIGChQoUKBAgQK9WrRNW1Q2iyKgLMkI6dutp3XeK4pWU5JyqZ8QTw6ipWY5eRNtknf2Uh8TrbE9RXiEzX2xfKmJ0pm0066DsptIHcrLgHcnynJyFE8kiohp1ZWXPNLoDt73Zql7l6le39S85fclSw1yaVm0OPaHSIgkcsQ8JMyVUh+oqdY9Rk7/1JCHuq+k/agsac79tpRnmYsVvau8B6J5Fvit0YfCO9BzvitS7inJ7WExDWQOmG+IfkWZe9/OZJQV0fpr/QNaB5VY5uooyVww51HhBtfUIWNdp++S1NMqKJnM4VTh+KQuBkoURSp/zJtENEm3JWDturVOXazEYliItU6eKxqn56rBPnV2dmLVypUoFopZh4xNtd9DpuUYenmY6HdBVN128dlgPZxT6y8DJKJ/cR4Y/cF+cU3RIsp8Uerv4omk7hcD/o7h9r+e1hHFrqapYyN1cK/Iff8mWb/E22FOICbpi1UGxVKR++jIjXWor4VQfx/H3Mtd5vXK33Oywr5l3qdvihnftrY2TRpKawz3t7LwwvlWOxrXhaw9Wm/pB+XXQ7+wTCMM7RzQ6tggnx6/bIt7pCY69cpx7VXprKFGK/fvQ7lcadiXdK9Qazv9At1op8jbK6p97nogWZ5lv5A1HOW2Le7LitKsFguuPWnLGx+fZ25Ssa49Wlpkrjin5bLuW34k1Wa4n2wUBYtKoECBAgUKFGirpX86i0qOgj6kXuzUHF3nBT0cMvqBjz3kSq3HR5H1ggIi515VNAaq2kN5S3vavxckUMDwMJ8294JbXzTYv30vfu9Bk+S8WLHcWH1KfjZZRSQlT57zh98A7BiNYDRDFHlFoiGq8MZHvlbF4mCtHW47Tr+9MbFe727VtFYQNbJxZnxePC0vsj8///xzAIAdupkp3NPG2V8vgqRBRr0fmPnXY8HOwSAT7o81uz8gGhazCPtjrRgtXmOetCs1NzdbzU/9PIjh4L8zOE/6u3wnZgdryL1u6ppE5PptDNIf1tEiURINchq5bcP92Rb0sI3mzn22cUyGqCP35dNjkblhxnWOcXnxr/cHm2evyDKxiLaIlVV91Rqq4L8Gnyu/aeINRYgKY+ZalAp2D7etITTvyBv73G/cn9O8sHbgPmvkmzxFTrmGgfDaUGTXl9gJ8qFMCb5Muaw0ltPv3AcHl/cogkUYHgK9RPMuDYFx0jBO/t89nxdnDXt/XxQ/ZtP9TzaGgkUlUKBAgQIFCrTV0jZtUdkcHJUot/feeUIvZlNH4iHzpfRgV6TVyPmueAPES+GdYOaeYLOM6JKp/kgE0r5+46vRMbqNpeWTd9nyjbgiXlIN1XWlzT6JpnlRNFD12M5yixtSHIwCn3p6l/vzNrEsJHLf2lKRe2TiihCfQ94na+qroxEAmTayYrmJBhk/UXBUFO+ELJEH81lPmTXZRUKkxWSkRIWsXkyN28MQyAsZiBVh1GuL+BOpuZtdL3PCO28CCGQSFaAe/L51hPg7xLpAqho0813likDLCANPE6EsSRRIZgcGANA63GA69EokVloc4+IAZdZL31ppBo+86F2zxuXRwwNSfAmdCx9VVT64rvIc4yTCRJE0OYZ1b00J0Qcj4/h46LdEfFUZJWKrF3kU51b+6hmxKQZHr125wlgK08zliZEnitQsay+DP57mHxPGjwcAjBg5Ykh8FMVkUfRbd+1FXrmaZtX2rLeMsCIyrzMHMsb0h5Dv/RJZOG6EzMkQPCp2h/bbky2hMWPMXCxasqjAi4vQSosR14IfzWf7K/MtGB6czzh3ZauukZhEnC50gfuarhkP/Zbyq5PhrxmxEUjbur/B9+GT6rhG81qDpdDugfTZYKSNKzt8bsfHrQfeXpxlfqROcY90ZcHPD6RZzv359KwfFlF8MNkqrIe0ZmWI/eY/NmAp8/+9sRQsKoECBQoUKFCgrZa2aYvK5uCoANYDO05598a7PCECUdath7UUNI/lQJgSd6SgSQIW64D4DdQikeX2XpGe1KKdtzZX2GixKZunRD26qfUJT7y7lLY6Rdupq5XAak+pZwGCao7SbbGcWJRUU2dVMvWSF9sfN/6+7l3A5uoZbvMMLVy8EACwH1MspERela902SEvHmJv7mkq9HVZ+ehypz6ikOZRpBoxsQr8e2JqTploscN5l6+5itxorrzuaWK8y6bcCAZGlGZokciJUUSolTrUskCUyIxWDNEYFYnU1cTaRxrLW/c44/NClGV9v3CXry5XnC+ivmomXvNzs1iOxggOQ+4jkcaedlxnPTLP+tz8PnmfydhNUF8LMMjm1RrRMd11a+dLxlwiLoiFMnEHwfCRtaXRUN74pHGkC5TzAM/iSTyUikSk0FoZ+ZZQrkFvrrgubLBErOOkqKDcAxLXUuQjFPv+bzWxWi1aYCyEnWLVzHXPcesp+q5QXmscO+lPW5tZI2vF2ko0p1zRc80H12/k9ZsopJT/csXMyYQdjCUpSlOdB80PxD3VmydrvZU2RDNPpPHIwx/xo0loDS9mVVZ8q5on5/DHOHL7zTnQfZsWA8qDa0nS/E38Xqs1zIPyzcEiv+wX8XU8GeU+HXsmCe1bwxrOdI9UK2Tm7qVqCREzT73qrr2C8dW07c3ZUD5aaVpTpHBSXbF5Brd9BByVQIECBQoUKNB2S9u0RWVzKI9snLiiqnGy7QAAbYBJREFUegoRkTKTk3a5ZKJeNBeOnm7Nibsk8euZnFhVU9HkE3JKLhvsizRLtWwuORwefvBhAMCM17/R40m0N9bRRBwG8uLdL4q2m0Sm3NrVkk2YmmqWI2KmZkFe1GiO2K2DqJnUPNeulwgT0SQTif3Pcjc/BceP1SYyfvVaXcdupeQNIi4A0SSppHIMeZ8aV4YJ/y4GiFpxZC5GS3QExy/W+9sBRX9kfzUbMlEwaSkTVNsPf/jDwpvwpFe1PNfTV8nV2GhpKcn41LMa3jDTzGvXWBc1MpZ5yjw0UGu1EywMtXLRemPa+MiHLzRt6RjTz8LyFFNjFBRJIiZT6aG/zFnnnuOU8y+aqYGWBPWYPh2Ub2pgsSQeGtPZiUwsSFoVo7Oamh1+I0+vy1NaGokwbD7ec8Z7TSlps0Jecn98UtVyE5ahJUgq61ltfHJGd4yW8ZCxJrYJM+LS2kd0XVozvLWXihy9MP9FZERapr8Arbcl1iF7T+JptVKuTCRjsZS1DKP805IgODOxrMFi/iZZC2Xhl5bA4SPaAQAl0XpT5s0pE+WVH96+xXWeehY26WO9SgTuRNdUpD4Vsh+VjE9V3Vu/pNhDsM1y13qjmZ8zWr8py9BPWltjGTvuS9Zy6u5LUYmWtNStS6zW3AfqzKoNdt+tJ04qBb8Pt1/WsulGzKi8yxgnlRbhxd1L1cohfyeikjs+SOtqKST6NejHQtnjnOi+1CSvuushpm9T5q69XC0v5Fn2gbiiuYesAZd7gecj+DJRsKgEChQoUKBAgbZa+qezqAAW9TLVmH3Gk8vJnGiAgs2RMtpD/UaIpsicGvQz4H0qT6KRU09Wz2zuFjkjjpI76LFjxwkP7tmR8em0AjEjro8Twr4Q4bG5pYUVyEek/fazHbMWRXBkTgg59TP3DTwUydg7PfvRIiwX1VNto1lO/+O7TB4djrW9D5X+khciVtZ4hyttyQvDh5v79zVr1jp9sHgDkR1/WmloIdFwpcRpa9r+B0gBygO1Iub2kbthRfZVc5Cpp4BovMukXaQOygib5Nh4ETZClmfXkkIaI9YAlVHNZkptKtJ/s0yuPjquReDAaQcCGBq5V6PiyJNmm3bHh5hAeZ4jopwqFoUrn1Gt7jzX3C7MwqprxrR1gPg0qYWCqKmenwGQ6RoqiWaYUhOWjtUGiPEi8qxWvSHWA8ePvkkejgq/dnd16Vhov8GxkTo8/7DG9Wt4eeaZZ5zn9K9INI9P4jwHcvXN4RjTEpLYWsz/07IKF0/F+oO4PEc2ZbuUN+/1i+9aHCc2DMeTHa0DrqVEsU9id53Dk1G1CnDvKuCnkDKuu5IrMyRaMxSpWfMPuczS5yMWvKio6vmiaa4gaS8p2XWnlfCTbYkvCo1StKxEbr8V/Vxlj/5U7lyoj04U6d8A3QMU/dZ1UlEZbBgf199HUYKJTJu7+1txDWfp4GPzSlk+gkUlUKBAgQIFCrTV0jZtUdkcHJW4oJhaCwmjXxhy4mpK9mdqAazE1YZ5d53rCTQulDInUt4HUq0vMzdP+0ipK/PeYbZVDcGQJ+79Kn0yuieYaJAVK5e5fYgSZDyt62+uNmv5Ff4lGmLCROPdHxOPInLv6m24PS0OHl5JZL/F8s6Ika3OOFBpo7WKc+HWYRujR/7oDhMFM3yEZE/2rF7IE2s58e+LeWcd0b9D/CI0G6vbL/t+5DxXC0XiaTSRedv0Iy5WYXU5yhbvg1V79bBKVDviC249FhCUGk5WyJIs/Ot3VVPlm9sv33pjefB44mDn1PILGqeqQO4aiYhx4YUU5H7+GTXLaQGX9dzjRS1JsdXG+SqzJzNiTr6P8vxooqI1CrB3/h5P6odB2ZXPUiWxbTO6g3VoVmhm0XWxXRQ3RHhg9M9Ikevcy3xr9xy+lhc4dX1t1q4z/mrMjxTR9yJ2/YBs3e4+ZrdY9s3U2zq8RfvPfmfqU5UU3rCV2LWn4Y9SzrMIqkx6Mqo82giXSLF0pA7PZJp7+rgPIG5z+HAOPBwRWNkqvpcX6tI9QMYm1onh3uHKlmaBVp48y6JXX6R7U2HtZd4Y6+bnYdpwX1L/NT9CkFZt1wqYe/Nu5SC1+7Av7142aVLAUQkUKFCgQIECbbe0TVtUNhdHpcp48ty1Suj5We7waoqJwPboHS3IlwPmQVan1sy7e6moznLyfpppZAHr2HXnSQCAimasJOqrnFSrLiKtBab1kAzF4383qa9L8FQUFbee2pO0XpOzDv4gz1WhEB53NZmd6ZtQlWy09NGw987CSpU+AEUMAcPp8f/6rwCA4S3GopLW3Ey8ikEjWBBRf+S0nXs+LB3towAA+0/e19Qng02rUFofQM3Pm0GtheiZxNuI3Dlgx1K+p/4l9PXJnHKUm4GI41NtzCekkJoiM8TNydw5qIlGatFQXQ98vdOWxMD6fsEakjGrt3zPVbYot8JvTb5XXU1Sc4UIzzVB7LXInPIpclCriTadZpZfKnmUnfX90h0PqVQ+icUSZ651ykYekJcBKe9ljI0jQPpVFQyTqE40YEYWub4btQHhiVZKaomUucRt2/ZNvku5pcuWKnqvgnwS/6fKSCJ57KHDMkKrLv3q6jKWUWZuzmQ96L7g4dDkyJEy8qtfNGThgfvYeMnlxKzBUcmziEbueqDvB6NhrJyb76uWr5DyAxadm+tY1mGtf72Mg4cgLHzX+iUajP3SDNZcs0SdlZ81p5e1CtD3qibrrkEmyDdlixbT1F1b9P+rcW16SOUqF/QriwEokjL3TplHhkExsopYTrm79qp98ph7qWdpzBjlSYRbka8sz3VQav2eBVEzc0s/ZF0PRGyMvLhrVH3YZO+xhimZW/qdDQyorFlDP7N6Y1AKOCqBAgUKFChQoO2WtmmLyuZQHhW83YmTop7WkhOBp3aeLPVd8Y5WXBV6bovHs1pm5HRMjUUwMaKkcJKUE/T7PvQhAEBJ7iBV26dK1iK8iKd65CrxSuS5Raa0qbnVKRdXUgXQ4Ok3ydwTte8/k8HcaV9+2edNedHu6B2fwPWej6n9N/M+VbSASH31cfiRR5m6idTItiN3DqJEfFgEsyH2MjdTG+gcazTPURJFpIi/EqkVx82IJPrDS12hRNwJlQOiLKp/hcw3MTs4B8S8oce7XvlLPXGl4FtD65N7d625S3gXTcwTRiBkdgxNAdOvCmWUsgVP6wMQN6mtCzIY0rSMUUItSHih9udpPAxu0/eFF1o9aGHSe/ckKZhKDA8lqmccW0UYZd2p85hjTX8SW06eU17YtnJbB5rcyJg88bIFizyMkvxBCfGQcmL7cC44bj7eklQvnRwmaMFdY7oRJ0361FTu+n/RR6Wk2qp8lkQ+iANEuReholzQnyBh5Eoh6U1EAWR/ZYxGtI6QOsx4lIjIS78HoZimVFl7nM+MUTBSbvz4iQCAsROIFlxWV4yY6y6h/wrf9XyQhJpaBQeI+0CJPHHtyRyqeYfMFuaU2zX354j7kruX2jGknLrzmtD30EMLtj4axNWhL0ei8+mr/CojtHTGpp8RfVgYvSWymMD1N9E9WX1g+IOpJ87rDFbUOjSllWfWSJpFJhU1lr5Xbr9i9S9x54q/ChQQEMV2P/b2iigefJ7/UQoWlUCBAgUKFCjQVkvbtEVlc6J+AKAsGkW9wS5hiHfwTUS2VMwHQ3WiYwrKZl6jeuSaO1K5h2xpNRpKbaBq26DmT9wI3g97YfmZ11ZWdfPr8AzOXBFNgp9SHRhweMmy1GYD9jKVar+IbMm8Q6KtlYgLIm20NBtealXbnyLxLrgiJ/lq1aJS6rUxLQaR94C+Jak7dgPSH89GoP4XJWIEKEKr1JNlaBILT606ODomrTsV6Tf7pR791iwCAEhEM828XCD83iz+NwP9/cqp5tHg2Mh8ViXja+T3u0lQJNkfjhvlhLIlyKXVfjpCkdVIte2SaD9EgyWxjlbWMcB+u1aMVOSiWeaz5pVLVUYFgbmeFzBm4JRpbml2+VVsDkNcF/TNYNu2HvO9tUV4rrp5eBDZMW4WbX2gn34xYgGRXE5jx5pswoob0dCW8Nwk6KqKM+PKKrGQWlpaC6Ydb2xaxHpbp4+JKzu6fmUMd/ByG6UqN7QOEuvCIoRybJqazdhwr+gU1ObuLtPfOhFNG9RUyij3EpmrAXedc7xK8pkkJZ0nu0eKzLSYsa720w/MfGheNNkzicibDbFGI7VyMBrQavI6drJX1Ln3eft7rrJl1kPNWw+6fpvNHDBnko3YEkuKrKumprKOjZ121x+O6OWcG7+tlhaz51T73P2a68LyXPF4jrTOFsonZcEzGdMfprmZ88m91PMQoxGwTFwWz4eH66G5ye6RblP2HY/+0aifbfqgsjnOtHlUGFzvr579c/TSA9kw0P5fTyE6fA7GYeR96uWI90cx9mPpbHymWy7y2opcporJrqIhBkAXohra3IPMUAIWebzZVAKF596BxLblmQ79OjYg00PzVAjjHEpE9NCk9lWvDvn0DglD8oLGerx9R696/KJq0W6IX/baYD1DyWiBx9hvxDuAJF5/hupdFLvz6fPir5eo4f9t6OdQdSjPQ/Ck40MHwaGYzRv7rXVI3Se+9a0AgFGjRvncOOXpbKsyOsSc8JB80kknNnCc+GPjVWXHx12/H7n4Yue9Iee70En+Ebf9NZ/HHHOMlPGcSD15sOvX5dEnOmu+4+2nNpTjuttQHf4f9sbnvJ5wrwhzXyCw4X3Jys4GePIA3RqeDyHvg1EMd51bnoaoY6j9gH0YpHjUMH+D1xF5dfh/c4aaC3/duGvYK8vroyGGJjjTBgoUKFCgQIG2W9qmLSqbRXnxROeZUmy2OvO0AVSKHzTLuTDCdHj0taY88+op1GV5GTw8zWoQ/nPPaqEhtIPXA0SWT56JPU058uCrI08F4fPM64+1GtDs6QHiFZK4acIsJgRjCCleeuwiXyPnuDExXtkD6yvUl+WD12GHltdpdHx0+6VSwRBE1bx8axBDzxvn2xqUXNlRAwsnOuVzgle5VVlQJjjl7FyR96gQAhw7ZXzAPzs+/N3lOVIxp2Mgn4umxYRxhXWlc6654eVR5vLbIDtMsmeDqj1m3HosuFlhVmmq1jJuHa2tkl7CD/EfKrZSZdArz9ek3in779fYpvbP65dnOPPX74QJxjlcQ2Qz3TDkg+NYlDFep7gpA6BrzHzWuc4b1i/cNn0Z9cpNlbQGiHNbRrvH+XT5bFh7dXdc/D1UjVne2suLc8UweYWb8Prly2+DzLj15B4shW9DUIf1PBtkD2AZtuU62/oGB7ZFP1e+5acUadgvskgtIln20uu34e+Zxwv3FJ13D6TR/1uT5wXAN23rpa1R/ygFi0qgQIECBQoUaKulfzqLSgTrLKlBqA2XfOa0WKv1yzuu+sPiBEZiYjkCg+k9u1zY1QlClmVQ5GyeWiX8lGnQVWNWBUpA4wSkjqdeRtL5GkqtJqA+TKBWCONjem+1HHgAVnTsJeBZSmAjdVSMpN8ENHMdAn2NozrQp0yqVj4EmJie6nkwl7aq0h/VvHSM3ft4AoopKFfBr4Djb9OWy9CoUUbepZ9a3QUCpBucvSeWX7xQcdVYq5bnTK1trkzUKBNsiw6O0i+CT6lTqjfRVH4GZIz5vk0LEOk8pOJUyfnKFFJeuk3Z8pxJNWEg573a7/Cc5a5WVScoWZoqHz7VBggA5mmznk8V117sTSd5qQ8Q3MuTi9g62LMOWkQsSr2MS70AnlUkz/JYr3lrj6BtnoW1Xu0rzFNU/LAgamyr7q4pdaanA6eAdKX0dZF1TO2ZwHLO3pS5ewXBx3w3Nw2JJ9ChL1vyyfFL/QSSnm9EXs20P6m1CTh15Km7Yem65d5SdfekBpdM8qZOqbTmFcZ4wAUk5HrXsFvKDnlSdDp2RPqgIJQuz9ZWJuNa69e2fHm3+7drKcx07GR/k72iLmsq8cZYnYYJ3qhrL0MsMlFTOSYvnllaiOB7arUFh4VWWmmDSUf9tSdf0+qAdXLnM4bEx6+M7SNYVAIFChQoUKBAWy1t0xaVzQ1PZnidTWPOJzyJm89S7IOMyUdOKGY55xFIyV44m9r0/lHayTPFGsoVelmqkPA8e0fPO2upUngtuQpLwZdDYMEF/MfeR0b6nKfe2AO2sv1yQckS1UQM09SWCEJlVRE49UWEqpbQvDzN9G5Vb0HVgiCaES0sng9LIm3lic+rqy0mijglmgu1A9SGTOvuR+LEAnSUab/du/zMc733wcio0ZBno3qzEXc+VSZiV3PKVYN2gbJsf10rF+c78+7dc/vPAp+Z853w/dpW2bsoF6JsKRgZtSfPN4BRJ0kSuz4jhTpiBVn0/UfoV8IxTAZ7rDxHZQJl0V/A3q9zHkoC/OVbkNRyROBCJgrk2tOCMl6y9hL6djDPHFtU4LtSwWwDp04oyKK85G0tHEMCgUUl18Kk61kqLss4qkYPaLoFXZ8lV97pN8Gw1phAYSzg9VtltBQ7j/0Q3CiKFFRPo7LEQhzrfuQB/OXuutCIpdgbP7jmKxumbC1W3JdK3M8pwESv1zrgjI/vL1RXXw8XjFFFi2ZsWoHjEmLPd6Qh5AZu/yBgi7QwlQRsUgHx1Bgi77GtyF17yFMdU4aJ0zpf+Ktg2CVMgYyPppKwu7G8L2OaJF415JmAdyUk8Db02P3b6VNIShgoUKBAgQIF2m5pm7aobC6OCnEPUp6Y5ZlNsW1OfIlgBvOOVhOsyamxTIuBJiXkadL1K2A99bRq6+BpmLDlkQ+hnHl1iOY84CZz01OzfCtLW9XUBUgDChpSGhW5tHfY1CgIgKVRI2RaeBHwoUyBlRxW1DpCbb+W1hrK8HRfV490CL9utActTexPgyc+NTGxaul9eiExV6lEILda4U1LPOGzX6kC2YlmKd9Uy5df6Nui/icykKWKAI2lBcA39aSXMjI2VfUjUFORaUs0rXpOnj1/IM8SV6UfDjXwKFL/jljGxiaLlLED6xCZGfCA7sBuueNTy+jb4FozFNyqnlst1gfLYx11z/+L1gC1yshciPXD+pG5Y1wf8ADfkCn2UJnz0Mekgxwb0f5LBE8T34WM4GkyV7r2BHwvdcfHrlXhuVRBve4Blska4vqlr0HqWyO1nIwP2/L8JPy0B/p+gW/2SxNX6gbn7m8aDaa1+P12eWmA/iisvUzb0hAbALBrT+WTPni501ZMgD/KqOcYYa2irj9FHsFaSmQtpJm793FPoR8QLemcbw1YIc8VAlVyvt31QhktlSrWP8TzrdNP2Tzog6IWGMqxv35VtrhvuONjQQczu34TrkuuFXcv4Wcs/UoVbNHlhQiAHOus7lleRH6SUgX1dPCxadhc+XPAUQkUKFCgQIECba+0TVtUNocKsAPIBXJYdRu9c+dz0X48XAK+F0Wi9Sl0sXcC9e6r0zwr3OvzZEwNgb/X3TrUZ0W++nfW9vLeqSeXO2ImTsxzizebe74F/j249blwo1tsAjTxr2H0i0YqudqUauR52jgm6nvhjTm0cee7hUh374npZ2CvU+vO4zSzYNuZzo+rOSp+ApOZydw0YBp4YBEcn4Z791zSHWS1Bl8iizlCnmQsGzA/RPvROWDUgGhHnlai9846pwXtU3ESROYy3r27Y2yjX1yrXdFaYd7jGJN31zqQ5VlDegZoGUbgWOj3Yr9zVSmlvx4mjfYv99aLWv/yBiwejQIB+aXVg/4T5Ntde1Eh4Z957o8PG7JylXlaOsdOIwwVq8O10iL3+1l3eaf/E4t7EPRAVtiXXFlQa5y+w/XprhUVcw9fpji2ha6pfCWIdL/RQBp/38q99ettB0RytXLhWk50vj0LXJQXrE7cl7x1acfI1csbZMmXE5Wt3HmdyVwNP24dBbuLvCtrj+PDXITZ4OtXZUnngPWJrEo99TxF7NeRu7JT9NsynDV7z2nlFc45d0xK68lPlBXknPuSrls4db7cFCwqgQIFChQoUKCtlv7pLCrI7d1jVHajfqzCZE6FdQXWcK0WjGqhLwMxPnzkUnp215iQK8tVc1KNWPwcUtVyXE90Ypnw7hapX861vNTZN/UWZzM5crnv18Owd4pnv/lJHhV/hG0wqRW1ZM+pQbFBHF5crU6xS2LP5ECrBcdO2lL0WPoyMEqIuAqxO55kJspy1Ol7QewOpmunb4LgJtCPgGOXeHKRaQp7RnGxJY6PqZc4FnGe69gpFgk1bCZx4/2xpzmmTHbm3fkn2k3ze7UoW4V6IuRIa4xIYLSXaykgb5Qt9epXlGQ3uiPz5p28qwgK/kKeZla+FSVWZEcSBEaeRYjENaNasYeSSkpljKF4DsJLHNm1oGNox6RQVBM+InWjohRPou7KoFqFfCRf4aEeVXVvyK0JxXxwjOnP5qEl8z3lOaUW7FkQuR/UfeTiTKOZuAdwDUWeiUmtPIzy8NYv9zM7Pq41xE+gGSdFWXAjrLQ/ap1yrZn0G1P3L12/LiQzI5UiFowKfoCpK1uKVs0mGcym0V51t186xtxDKaNkgQIjfeL+WKtbucvd/hH/KJa/MbouPRRnTfDJOiN3zlgPk61GxX1Ax1DqUEwqzhOHw/S31u+tX7hEedf1760b/n3LqgON2XO5WW5GRM/GULCoBAoUKFCgQIG2WtqmLSqbg6OSR41e7blnAVm4eIGpXw7aY8eYNOkVSXseCZiJnv1Fw9bMkTztE8dAfjeRLox5F0uKMFPS2H5q4HLCbmFac+E5cSNPVC8lBgY1Gukbh6dUSbVsPEQd7Jd2PK64/OsdqHj7J27kAU/i9j16k5caU0Co5in33JrDQ+pImpziSZnYLfLp+cHw7jopMTqAOUYqln8+ixg5ITXSy53dT8g/NQyqosIjrRcJl49EEygWiK3HRi8Yfkpw/UPUw16iONQCxogKP+GS9KuUuBon5YJzk0cAu6GaFjEbeNeeEO2U3vzsjzuvbEpv4YlDROuggIOkxBeK44KVTaIbis5haJQdtXLR6sEx9bGMiI3h96kYmRC7/Yo8eVfMHWb6LXEdyBzwTr7k+vBEitPhReix/ixBVPb6VXJ9sBiBlfh1lFxLAtcvo0S0HnnDzhVrSNTaQBnQda6Lz7RR5lpTrB5GqnjrV3nm2pPqGB0j5bK8KBO0OlYc/nTsGN0odVeaywXOiuvX9fnIY29/i5vsUxmKLHLr8NeMJtRhRI7fL/Xtke+6oXtrmFGRWaYYRFa63T3fouPKeMSFNQK73+vepLVwnmnNEJ6JfZOniIjJopFR3CvcfhFPSLMl69y4Pnu61BTThmMuvMTWpysusQ6n1xrp6VPAUQkUKFCgQIECbbe0yQeVu+++G295y1swfvx4RFGEX//6187z97znPWrp4H8zZsxwyqxcuRKnnXYaRo4cifb2dpx55plYu3btJjNPHJVN+Q8AypUKypUKeM2X5ea/VatWYdWqVbjpt7/FTb/9LW6744+47Y4/4q8PPYi/PvSglq+nGepphqRcQlIuIYojRHHUUF+aZ0jzDKVyGaVyGZmcXTMAaZYjzXLESWL+i+PCidicWvPcnNqzLEO5XEa5XEYemRN0Dps9EzA+Lmmeo1KpoFKpNNaT50hKCZJSgiiJESWxPiPfaZYhzTLlpVavo1av2+fSRrmpgnJTRV/kczvOpt5SqYxSqWyfF/iJS4nRLqMIiCIdlyzPkeW58uLPVeN/ZhziUglxqdRQT5ZnKJUrKJUryPm/wphkea51lCpllCrlhjYQmf84LklSQpKUGsaP815pakKlqcmrw/STY0SZ8NvKshxZlut8R/I/5TmS/ygXHB+PF/YpR4SkXEZSLhfKmP9lWYosS+18UmZUtsx/HKdSpYLSIHORZinSLEW9VkO9VsM999wzdL+amlBuahqEXxmnOALiSPvfIINpijRNUWlqRqWpubD2+F8u6yxvnAddU+Y/ymcUx4jiWOfPXw+UHz6hjHEuuS5K5bKtg8+E31K5hFK5VJCDwdce5cKfC863v4ZtObs+k0oZSaWMPIrMf/5YR9TCzfza8Re5kPnWPYdzw7ZYr/yXlEqDyITpT6XShEqlybaQR+Y/b80kpTKSUtmOrccr56hhry2OnSeflAP2hyJdShKUkqRRRtlvWVPFsXVkVGiwdYcoBqJY+U1KJSSFfclvi/v1UOtB9ybuF4VyJrrO/m3g31sdY93nzH8Nf/fkP0Q5EOX6fpzEiJN46DVcaWqQHX6nvPs02N/hTaFNPqisW7cOU6ZMwdVXXz1kmRkzZmDRokX637XXXus8P+200/DYY4/htttuw80334y7774bZ5999iYzHyhQoECBAgXavmmTfVRmzpyJmTNnvmSZpqYmdHV1DfrsiSeewK233oq//vWvOPDAAwEA3/jGN/DGN74RX/7ylzF+/PiGdwYGBjBQQEHt7e3dVLaV8qjRMdm7icQ+e+8DAGhrbwMArF+7btC6/KyTfn3+PX1UaGNAvL1bklbzQ8nlQeviD16suyIzymNG2lx/0w0AgOOPP97hxdwrux75DUiTUlsm59fLP3sZAOCzl3/e58rlNXd/12ihQv2rZc4qoi0OGzZM3vXHzo3AsKEVg/M8++9/AwAcMG2aW09uz+AbuhLlWF79TXP4PkcOzXpX7Ve0gfp4t93f34/164ylsLNz1OAdaHxb/p8+O8okAOBr/2V4vPD89znlBq1piDGD+MMsX7oMANC9wzDvOTtuPYDc7y5P/FyydCkA4Lc3/xaHHXaY1ysPxderw18zXFuXXfYZAMDnP/8f0ieRLZ1eb4ByGwAWeTl+elavlu6Z/nS2jJHng4+h1rMBAUqU58KP9BtQV4uo+LPNiqvFzZP77rsPADDtgAOcfkXexuDNjFOGCKx9smeuWrUKANA1rtPU6Am23SM4XuK74O05lIvb7vgjAOANx7wePun6HWQPKPaHtLbH7AvDx4xyfvezyGsQ1WDVSX/XyB5Dud5xxx2kjiF48teHbtfu/rZy5UoAwOjRbU7xl1rKQ63L62+4EQDwL0ea9bHDTiOGYMarp2HZ5BqdteBF41M5duw4ry7va+zXLZ9cOjI+K1esAAB0jBrlceZmlXcrI4bPpltLNoZeER+VO++8E2PHjsWee+6J8847Dyuk4wAwa9YstLe36yEFAI499ljEcYz7779/0PquuOIKtLW16X877rjjK8F2oECBAgUKFGgro5c96mfGjBk48cQTMWnSJMydOxcf//jHMXPmTMyaNQtJkmDx4sUYO3asy0SphI6ODixevHjQOi+99FJcdNFF+r23txc77rjjZmdPbjjz6aHVtZCo5jHUnRpP6EM21Pg+rRbX/+J6AMCRRx4JANh9193M8yEMCL4W5NPSJUsAAPfJYe+EE453npuD7hD8+hqG0FNPPTXog6E0Mn3uqdoRgNtvN1rYDuMnAAAOPviQQdvWOci9OfAaIRbOj3/8YwDA/tRAXSMWkA1yJ+ozLJ7599xzDwDgjPe8BwBQaXKXh0XAZPTA4NWyvdkPP4xn5z0LAPi3006VbnpmiMaXvbbc57PunQXAWlRsHg6vusi24Rm8tOjPbjAy+La3vw0AMGGssYL6Vo/IXw/eGLP0smVGk21uam7sF2kDsqPoqRKxM3euGb/Ii5Kwa2oQa6anIbINWisGBgxOxoknn+jWMMTaGgptk5rll666CgDwyUs/ObT1zt8LGsxd5vsPfvADAMCee+wOABg2bLjLmtc3pUK7fPb3Rx4BADz73DwAwNtPOdGpo3EOfOuUO1f0obv2OnOV//p/Obahm9Yiyioyv4D5oIXw298AAFx8yUcBAE1etKLdY4c0A2uY3eyHHwIAPPPU0wCA008/3elOwz7eOKjy4crNV7/yFQDAF77wObce+r0MQhbN1h3kRx/9OwBg+kHT3HJDCc4Q+2CcR3hhifl7eeOvbwIAXPiBD5mykTuPUcOa8cndcz972WUAgP/6r/9y2lSWkDdYpdTyM0QT/2jUz8t+UDn11FP135MnT8Z+++2HXXfdFXfeeSeOOeaYzaqzqakJTU1NGy4YKFCgQIECBdqu6BXHUdlll13Q2dmJOXPm4JhjjkFXVxeWyl02qV6vY+XKlUP6tQxFm+NBHOUWaTKpuHeW1OLotazZdplvpE4UUEGslEyocc3FOOAdMZH8atX18t3mIVmx0mifL85/AQAwaeJE8653C1sTNMWSoCUq/oanqCyUetTjuia8yok+zTLktapTh03MAed7XjJtLpYTe8bMxYLsWqvKe8RCcVlBXZAQswHer8dYssTMeUkQDDkmjFzw76SJvJoPGP8gYtyQx6Vifeul3wH7SyRHWgHSHNVqH6TjKDJMH4Ranbl9DN/VuilfLrdKk6ZgXfJsEAtD0X9Fm6CfUK1qeH1+/rPoXW38A4gkqnlHCFFAtGMPLbZOjATpd1+f4WnVcnONqll6OT7UAguAxXWpOxEcBaKaErump2c5AGDZcmONG9cxSvpDouwIjxxHRY8V64d8p49Aa2srQMRgot8SFbR/nfO7b80gmm6fIO4uEhnsF9TRspSrDxheFAVZNdhYEXYHKGPybPly09+WZsmjJOi2vtGOddW5lqTf7G8sP/fLmvzbQ38DYBBCI8I+cw8gOI+0pb4pRAGldiljvGr1CukXEUs91FEZV84JEZqjHJrBuCqys2D+fADQ/ZbzxPxJCbMiu8vCZgWvSb/le78gm/YsM/WlMi5xXEZUd/dArt8aZUZRfWV4ZKEvWrAQALCmpwcA0NTebopTQZcxT3Mz1rG3X8S5zcXWs8qMXU+vWXP1lAjRsucwX1LNXWuaPowIrlXKplnnf3/0UfNcZE3HJ4nUYhR5+3Gd2bBzd69Y32v6uXDBiwCAiTvv6NSZENtHxUSwWwT1WffcDFi+zPSTPptEf468HG5ce3k/efUt5FK11P3UU0+a31MXmZ1ykA70IScyMkvInpIN4aOy1WdPfvHFF7FixQp0d3cDAKZPn46enh48+OCDWub2229HlmU45JBDXml2AgUKFChQoEDbEG2yRWXt2rWYM2eOfp83bx5mz56Njo4OdHR04LOf/SxOOukkdHV1Ye7cufjoRz+K3XbbDccddxwA4DWveQ1mzJiBs846C9/5zndQq9Xw/ve/H6eeeuqgET+vBCleCTVj7/dEEB5pGYlYTjVp83sivgq5oMLy1Mc7wdhDiIwI0gBgxHDj7U1fC+aZsRlQhUW6Q2gdcJ7TAlFnxktqLkRn5J1mhAJ6JN+VfwiApT3pmpdKFd4XS9tl9tt8T4U5q4HnwnPk8RyhtaXFtC18Kaoryyoyrfws/W5A0ZWvKyWSYeyYsc4L9vqT1i0gjrWDThuci3LiIvTGPjKjZymL4DEpRMMFM8HW0xSdYyWyRMZK86toG9ICc76w3zLGGS1Qos1072B8fJJKWVgQpE9aKGJrqUg4JryDZj4OGQ7m+MkGzGdSZvZo6Q/Fg9fsmpBFtFmF8jUfq3qNdatcLimcrcoxNX+Zd2aL1elSmGDz0VIxVo/2tjapR9rUO3+uTeZdsvfxRPXVjLxiYaB1sr5G8s+wnCL0Sh2K9uvKsWa8FRaq682cDB8xQsrFQO6tO/IrbTBXE8ouMi2FsiTwoIuWG4vr7hJ5qHm1YpYTVG3VjnOVT+5fA7K3dHaOln66Gcvhzaf2O+ZYy/qVcViXGuvIiJGGJ8ogEAMlrg277op15J5OTBZozRwQaw33ZBLRnpFSnjie1ieC89wrWFwdo01/uadSoJOMa05zUMt38sreiHVAPptl76L1l+MTIbIve4DESRa5P3hyq1ZdRQf2ssB79VBWc90vgdViTR7TOUbqclGvlZcK5cK0lRKB12uNaMnt7aPkRe6bglwrpaM4RizzrX8xuH9tMKpx82iTLSoPPPAApk6diqlTpwIALrroIkydOhWf/vSnkSQJHnnkEfzrv/4r9thjD5x55pmYNm0a/vSnPzk+Jj/96U+x11574ZhjjsEb3/hGHH744fje97738vUqUKBAgQIFCrRd0CZbVI466qiXvGP63//93w3W0dHRgZ/97Geb2vTLQnkERX5MmfPDdwZnfhLRGJidsmB6MOWYK4j3zaqxu34HSclEQTCTKAD095l/09+BmkeqV7l6cSp1mINeVmUsu+uZzlj63XfbXTrhaQ3IrGWE/iDeiZrWnFyej2rvkLqoaUt/Ks3Ci+2PIcnzIe+T57RaxyoJUR812tRZKsszLxrCZgs2H+Wy0Waqct/M/i5cZHwXOjtEW6RFquZm3c1RR1naos+R7wRPDbO12eCJzHvW3O3vs+9rpD+mXKx5LDz/AyVz7i81GZ7T1I6lzaciZWRsanIPzPlWq448z0XjXLjQ+JF0dJjxo2WtLNpelRppQaFRXA2xlGgGVmmlScZs6bLlMhCuhVEd+aWxRMaxKMemgPgwiB9NkpRQ0L+kX5HT7zxjHeoZYd4VGY1Ve3WtdoouLWNcF2sQ5zBHXeWvImUG+ox/Q10sKiOHD5dxYd4UWlR47871a76pHA8w7bj4SS0x41aWfaJcrqBWc+/1df0S+Vf8I3yLEvs1cmS74S2jrwLXMXPEcA3TMlXAtshpGTP80qLSPEywmkgaruVv/8ya6/a7Jr4sS5ea/ka0AkQyfqWSZlrWzL2yl5TKZq/QDPLssDQyUixmS2V/2GGHHd1xkT1F88x4Gb3NIHCsxQLS1CpNJE4dXIPMQ5Rmst+xHva72fC8ZIHZU0eOHCn9dddyudyMqr8WaPmjNY4+giJ7O4ovIvfesuylXL+R+je6Vg7KaqZ7WIYlEunpk+ZTYrckO3jcKvs29wr92+D+TVm+0szFunXGx4uRZwDlqxn1qpvN2Zedl5tCrp9AgQIFChQo0FZL/3TZk90Tn+vwkasGKZ9wP/VlnjzlHvmeu/8MAFi00Hiw77X3HgCAI15r0AddDk2bAxIxsEBO7b6av2qVQUNcvngRAGDPyZM9Zsir+WHdOhNNMVruo+Fp6H//+yP40f/7fwCA95/3AQAmIst0n5EDpvSARNCMbBspLLnWG0ZD3HHHnQCANil3wAH7Fbtgxy8Cho809/grViyXZy5/VUbtyMm8EnuXvg39NffRo8VCA0XTtBYkABjo70dT83CnLR+zRC0MovWWPP8BanHz5z8PABgpd/Sqaam1xLWwxEmsvjTqFyPa3/wXTJTWww8Z7IfDDz8CADC8dZjDIonz2z6qzR0Pr5zOVVx8mnmfhsaN6xK+3dbUqiOfke9fIJ9zBCNml4lGC+5bbywq3eO7tRQ1x29+xyDqrlm7BgAwQrS0c845BwBQrrjWTVoD2gUd0yK50ppl5OTppw1mBpGOx3ePK7TtEueLVg7PVQkvLjSRGB2j2wEArWJxavCjoUVFoqVcPA7zjFFKC55/DgAwaa89AVgNWXmM3DnqEHn29x4fPdRakLzFViD6M6xfv979Xf9h/vXEkybKY+JE4yM4vLmlsTIAy1YslXIE3NScx+prt1zKdI3p9LnxWwcAtI+kD5L8ELF/shdJhGZ/n1nvbWJxcidPojUZObVitdMWx+iBvz4AAPjjn/4EAHjTjDcCACbvu6/DE6smIm1F9gWN6syt1YMWtFJFfO90ojzcH/l9zRCo6rTOXnvddQCA52WvOf4txxsep+zndTvHunVmLY0RP7gC+AwAYNFC87fjvlkGP+iEE98qdWROORItjm1i5eKcDhvm7humT678RYMJYIFC9uRAgQIFChQo0HZL27RFZXNxVDKNN2dIjReBwAiK1D152vtXc3Ivi+b99NMGwZWooY8/ZhAhD9hviinXZDUU1tEhp9Ymiawhvsa8554DAFzz4x8BAPbczVg9fiu+P2ed8e8ArEY+eT9jaaH3f6zRL6Y+piX4+8N/wxe/cAUA4Ic/vAaAzWl0gKC6tjQbcXhRYvyZO2KF4E90jjZa0ve++x0AwAixkjz59OMAgLnPmnGoyB3whB13AgAcOG0aholfQIURNqnrD/OZT30aANDWacaluWzquOCCDwMA+gU3o6XJaLn9/X3Cm7lPVfRUwTr5/g9+CABYuGAB2tuNlem8c88DYKM/WlvMne0awV0Y3dnhjB2VhLvvudeMpSCbMgfU6E4zHiee8FYpL30Si1MpThosJMtkLP/nx/8DADji8MMBAF/5ypcBACefeBIA4FnRqF5/jEH/pOWtpcIlK8yJjP7lL38BAPz+lt8DAE552ynYe889pCyjtdy10icIrar0aUSC+eFH11xj2hQ/mOmHvRYAsGqF0TT/74//BwBYK9E+lRbjG7DHLrupBvnz642GuM9rjEXhmDeY/DAPPmC02//3EzMOZ55xhmla1uaSxcZaQWylnlU9AIBxYnH405/uAgD89S+mHkbsjO8ej+kHHyr8GB+LbrEcrVljNNCEkVGigT/6xBMAgFt/fwsAoFo3Wny7yPf73v8hZ5y4XywVjJf999vf8JBmWuZrX/8vw69g0/zopz8BAFz60UsBACNaZU+Iube4PHEcGMnzosz/pwQ1tLPTrM3x4w3sw2GvPQzLZMzeLJozo38U64PsC4/f/d53AQAt4ifxo//+GwDg4o+YNdfaZngfLutkhViQu6XNLBeMl2qEz3/OoLaOFmsU/Z/O++CHnH5w7Lhtr+1dI7xxHEz/lywxbf3n175qCor/yYGHmLl9y5vfDMBY6DpGmr2lVvV8L6SNP//ZWLz/fJdBnr7g4osBAD//qUHYXbxokbxlXpg5w+SyWyJYTa0i17Fng/vGf31dMUeYy+uij1wIABhWEURh+lyJBYJBJdy3iM1yzTVmv588xezJJ55oLClXf/NbAIDlgrv1yGzzt+X8952LCiOkiO8jY7xUEKK/8fWvAwDeKLn5Lr/8MgDApR8z/W+SyDq+xzW2Ky3t9IvyrFxRVrd4X3oL8dI2j60eRyVQoECBAgUKFGhzaZu2qGwWRUBV7j1VwZS73KrcNxIlsC53dPRwrhEdUI6Rc+fOAwA8LbklPv6JTwAA1q83WsLnPm+yDi9f2QMAGNU+Cp/8hNGokqRUqAmY96yp6zc33ggA+MTHTLlYtKF7RQP96EdNToy9J5t71TvuvAMAsGyp0aZ23FF8XOSe9rZbbzO8ffxjiptw1plnAQC+JxrV3yQnyPKlRrNYIX4Vr51utP314v19059/AwDYddddAQBHig8Oj7vX/I/RCmKJJnjkEYPoOLy1Ff0SeZGIJtm72ozR1d/+NgDgfecZa0f3eIOLskiiei756CUAgAmixRGhlXfXO4nVpi7RPp/59GcBAKecZCwTe71rD/SsNv4dn/y4mZ8BibQhtsuxx/wLAKBLct2kdTOGD0vukEcfNeNzwYUXmI6KMvH9H5m8LJyTXSYZXv7tXacrT82CgtonY/i1r30NAPCJTxoLUkUiiSbvayxjX/+m0YJmvulNAIArv/QlAMBy8e05bLoBReR9/GWfucyMj+CrfEzk5ic//QnmCd7RZLHsMYJGEaBFC2JUDFFvv/OD7wMAjvkXk/KiWaIgviNzRfDGD77v/QCAdevMvfsa6eOse2Zh7jNzTZuyxRx5xFEAgJpEzkzex9y5P/gXM8Zf/OKVAICSrItFYh046uijAVgE4ln3Gu24JJFM7z/HyA2VtFv/+L+4/hfGirNiuZHjE99qctzQ2riqx1iE5sqau/22PwAAPvD+90tdZmz/9reHAQAfvtBoybQSnHHmewEAq1cZnkaOMFaz/v4B/PjH1wAA3jzT+D/svZexaq2TSIsrrzT9HCZWil12Nzm+DhNLwcgRxo+GEVTPzjXj+L3vG/iGb33LzEFZrJZ//7vJHTP/xRfQL74oXxXrXJNYJZlbjfvXV79qrBQHHjIdAHDEdLOOq1VjifnKV01um0z8aYYNM+uEa2/nnScBsODD37j6a/j3M42lt7vbtPXEk8bK+rOf/BQAMEasj5Mm7QwA6BIZGjWqHQCwWOa7Zyfz/Ic/MnvJpz/9KemvWUf/98fbAQDnnXsuAGD0mE4cJBbh1bLPdkvS2kcfewwA8MLzxjr54Y98BIDFcHrnaacBAG7+/e8AAL1iGfzed82+OFyswPvJ2qzL+N0i1u1x48bhnLNMpvXVgjj71a8YS9qIFiMTjLg68UQjgxMl6qdfrJk33WTy9IzvMuNB6xytm+8SHv8qvmxHy3r48MUXY4dus47HSRTWujXGj+frXzM8fPITnwQANMlaeY1EMX7iE2bv4br/qORZWrHMrH9iF7G/XLMakTRQU6RtNSEShGYz/E82hqJ8c+wwW5h6e3vR1taGnp4e6yC3kVRfsgRoNcKTSshXXQb9V7/8JQDgMDHHU1Bv/b0xp894o9l8Rgw3QrhEoKRHy9VCmSGIEp7c12c2jlLJLPT+9evxk58YE/DrjzUm/YceMZthtd/88Tzt1HcAABIBkWMIHa+P0pob+ryix2zGz8ofpVl/MVc9+7xmbwD2UDFxp/EaEkyE5Uwh8GnaE+h7ORw99NBsAEB3l3Gy+72Mw/kfMM649VrNeT8CYdNp5jUb3fd/8N84+OCDAQDPPPOMeSZw/Ecc/ToAwDgBLWJoN6/f2G+GNz4824wXHduefdrUN+0QUz+dK2e84VipJkWzONPaEF7DH6GnecVx+GFmwx4hMnX7HeYP2PvOMw6fmYIvSSi57NTLxUzdJ9dR/++nZnM+dPqh2kZrq5GNvfYwm8Xuu+8pY+iOOXmryJUgx7h3jannllvNpvrW441p+GfX3gAAeI8kYCsu5v+RhI0MjeyRP/YVOTx1CVjeauGxRQ6RR73uKABQAEZuEevkD+Hdd5lrlxkz3mB4F6fK5mazrr7z7e+iu9u8O326+WM4qq3dtC3XMTUJ68w8szWvOIbJGqNT4chh5hrmrjvMwfws+UNFB0CbtK1uQ5gTcxj4+fUm+eKOO+wAwIBUAvaw+653vcvwQIArXk8QUp1hynJV/H05yK2QsO7d9jSHjbeecBKuu9ZcJ5z+7n+TsZEwXZVjM5/94nj8p3uMY+fTkgD04OnmwLJ+rTlcPyPr+sz3msNRkxwa1am2sIZTufZ89jkzZrwO5mFmihxY//a32QCAE+WaUcEWJfyUB/FyxfD80EMGRZwOwvfdZ/aYU9/2dgDAHX+6C28/+RSpQyD+Rdl78AHTFq8uX5Ar6znPmgPYW2aaA3lVlIfFoqCccqqpr20Er05dMEJe3yZxjIcfNnvCzpPMAeqOu+4EAKwVuf63d8meKkpBSWAlaimhDODw/KiM1xVXGiXhlFMML289/gQANnHkOeeeq1c31DiffdY4mFfkimep/I343U03mzF7p6mLTvQDIlz/9s53OfVY53vzwYM513Ka13Uj/9bV5ip+38lmfnko3IdOwtxLZd0/J/I/bozZc79x9TcBWAV+190NxAUVFUIiUNYqzU02DN9LzFt9UpTTKYYXBdssXN709vaivb0dq1ev3ui/3+HqJ1CgQIECBQq01dI2ffWzOeHJeWRhrRlWWhIY7lPfdjIAG3RHoLBTxcrB026WEbzHmBgzJrnSEFlTsFW0qLKYAZvKZZz/vvNNG3Kaf3C20VZqolkTxpzHfAtbLKA97K98Hy1w0aPazcl0nmgsTMB2xJFH2L56MOQ2GRsd+QTIiM6/Xtg2rwCsw647juRWobhFqzj738/SEn+S0MDh4qBGCHx19FOAMHHkIuy/OOFOm2bSo/MIP+8Zo3EuFvCjffbeW4aHjoSphV9XOH5D7ZIA7UPi8Pfkk8apcvEiE2Y+YoQbfpx4kOO8SqMZm5rL5Z83joX1gSp+eaOx0q1cbq4bJh1nroeiIcdQWiC8u5RrkzBOykPPSmNJm7jTRKeeQgU4/d3vMe8IfD+1sUWLjXbbIVDZv77pt4Zf0SgnTJhQ7KVaexg6+ya5ltJknaLZkeem5mYskavI0aqN1Z1+U7YS5dvUseMEY4khwNV8MdvTglBhWgZC9NN6l6qnq/JNYMdTTj7ZqeN50Sg5LiWCBapXMVk0/6BlleBy7zvvfU59//0Dcy3z9DPPYJ999nH6aVMmyNqStkojTD9miqPjceJkvF4sTffNMg7czQyZ1z2LyQwlxYJCJkSasmNvWQNplVdYxkn2L/cZZ/+TTznZ4dGuPTe4gGuOlhiWfEESoBKOYfLkyQWIA3GWF8voEUceCcCCZvrm+wGxLP3xDnOlQ/C89hFt0qSbfDHjOBKsD8C0A8ye0CTXaSpRTOinwH5ct9LvlBMtH1L3gQcdBAD4+bXmCvGXvzJreLGsm112Nmu4VEo0tQX7tdtuxrrGvZVXPQdMmSptm/KzH3rY7Ufsrl/1SpBPf93EiHS+JoilcOlyY7059NBDnf6oc7HUQcsimfzwRRfId1P+jrvvBgAFlLvrzjsBAG89/l+FqahgheNPXkCKRyE8OVCgQIECBQq03dI2bVHZnPBkh9wDZwFsyH3c2LBXgYL8bEKbcqpkaFx7W+egVTRCEtPK4T/IC0+BNZKgq1msGkwbPzif/mneHRiGdW60P5BXfwR7Aqffw8C69f5bg/LiVznUnCyTkLyDDzxoAyUHGWNq95q0zfxuLW98b3Aucq9GG+YLLJdQXk02Gb+0NtHQAqvymKafBR1bB6tJRc1LiEaLSS4WFobID4jMNIpxPsTvfMxxMiWGDx+uPgQbrMPPgEie+VisVASKGz5iuFPAX8LIC1XpQ7dtzler+Kqpdugl5dyg0NHXSXx/1qxdi24mySzwM2i/Guoaog19TmutfN2EbY/WVYZEE1ZhkEYc5nL/dzVIGP12sYTv7jdtqnWipCx4gQqaRFJ+tj5F3nNazMjDxijf3nRxvyJYoIbVbmAu7NYcOw/6xXdjkVhaOyTkPEfjPA69dtz9mc7SpebWQUo38jiofOTuJ5OpEk5gKPL/fMVw54LzS8A3OlG/9B849m/wMiE8OVCgQIECBQq03dI2bVHZHIpye18aeVoeD3qZ3LlXVbWQ0yKv5AWMqVozXvCxeG4zgoFKAu/2o6o5kaZZiljRzOUeWO4oWxgxlDKpmSGNapBwNp7MY/JCbYLlRPNqFq0pFUtKXk+B2I200GRe/JD7Vqakp8f5vDnGQ79VwNaq1fVSngB4phzHk74OqZRDHul9ML3kGYYKJgZjgj+6HEgdbCsjbxnH2BRcuMRoOWPGGU02SczzTCIgsnqGmoy/JhH0lHh6sHeOMf4+cyUklOF7qfgPpRx7Wl7Yf1q55Ht1QLTfzKYXIJAVo3wQ9Ut32C9Ix80/agQl8+aKbTwzx0Q7HXpYt/xOzyoLd807+rxecupmgBHHw44L26D1TeYsY7h+n8ODRovpc4lgqqfoE8h/1q0RNP3rhBfPwiKTwbDHlEkKRRQfedSEu3bIeNYlwipVuZC5iCONZuun/EmZLKZ8mvkeWCvzKwkvC1kq3boHRN51HxDeRVa7Jdx72eKF2EuSgmYeX6mkzNAm6q68c37JS1pzI1JYPotFrgVsrYjqn8keMCBjVhIWOgWgkNYNrsEYbh2cT5WHqjvG1pdFxlF4HDasBai7ewvnoObtFUrSZipAjkx8WJK1RZmjXFMe1DpY7LfwOyCb62oBLtP9XfpbV0upyDFlRlize07uPK/WzdwxImnffY0fUq3Wr3zqPMiYp7nrY+fvFbT6VFLKvSR09ZLNci9NCI2hWWszO0/ySctPntacOjKZp1o1d+vw/v6lUk/XOAMmuOBF44tUbnLnJK31a51Kieuz8nJTsKgEChQoUKBAgbZa2qYtKpsT9QPAatSJ60lNGOsS76wJS89ycprXxGJUjxJGaLj1aOSJNJgkiT0aijrWLTD1fX0CQldyz45lSbWud9OaAM+pxkYu8ATO6CHhLUliLUsP+sg1nCDh3bWUY0TJI4+a2Pj9p5i4/IQHc+HVWlSknpJ3Bx7Fen8aybNe8TnQQIPYbbtEqH2mOy95iQK95GxNErmV6HvsflnfoQ8K+8t5IuR/JFYAUqck+yLPCTVOzq/XT0aRqNUnKWlE2eIlxiOf39VKFzNKgv0kEKA7qLloLG0SqTR//nwAhQgd9rtwF06rlVrQvNTzxAXhWLZIJJZG1Ei/SgIopXf52k8zkknsR81EivehEWICPqZWGPLCOmmVqrhOOV0C9HfvfSZFwF4CkAZ4azOxsph70OIQTKJcLGdMNkcoeI2Cg8tT7CWnjNXfgkJrPiolG6nHhG6UDdVq4byCWNanrknKQ7/5rIoFKpaUCYl8MtGgruGCUbREOdaF6PajR4AcKYN54u419jV3LTauZ/OxeNkS+Zrb8edeJ1glJM4Tx4Etl1uEZ7EYqbywPooDrb+kwnwn3Fuk8h1l38oYIVUijL1H9IeJuT/LmHL9ipyP7RgtPJj6mAATeW7lT0jXjKwttaqX3DVF0MU+sZRw3ROHRucidvcLrmHksd1v9G+A7B3Kk/w9E/wY3TMTV95VBlPXR4UpNiaIX6HKTZ41/J3SqK+4YZSd5/6/N5aCRSVQoECBAgUKtNXSNm1R2ZyonzyyCLIpu69VEFXR/NBcMRqm9SswH2ndaGYV8aLPavbe0LQh96vyXpOgahKN01TiasxMfGWnRO6ka3J3K6f4rOrG7VOjqksiteXLjLbYJgnViGuQpTUkouWqv4gmK+P9uLTV1IzBqKXFRFwkTa3CS9V5Tm0Aoh2VpE9pta6qEdEPYz15eyIoPGUpx874JNQEVVY1cJn3zlHG6rFy9Vp5X7QKqSdNB9DSKnVU3btba50RfgU/Zs06U9cIQcUkdk0mY1mKeWfr3rvzXr5JYLNr/QNqKejqmlDsnmLVME0D5bgu2lxLszyvuwiQI8VHY51E6KiFxppezEccIZdLeWrjdY3EoR+BoN4Kgqdavbw5IeJpRSIU6pRjWtZEblqGGbnp6+/TyAP2n/f/leEjdGyK/bKIxOajVJb3M8+6I/1ltEQqc2pRllP152iWdVcVXyP6P3AOVgoEfhIbq2WdiTLF14Hrt0IclX7Xf0x9nKQPfb19aq1S1yL6ZEmEEWq0yrn+XXUZw4qgKK9YYeDMaaHh+JXEwhrlngxG1h+N8lev0rfElC1LdBd9EeLEtXrQwYL+bpVWSfY34K1zUcGZWK+9vVP9Y7juuB+1jGg33wfoF8g6uOeY+X3+RZMItZsWwtzV9lUmuW8U9v2M8y37MeteK5DytqD5SGR/Sz0/C/pRlWTt1epE3jbjt3iJidBslueVSouOjXJDCymtjWJZUn8nopYLyjPHmMR+quVJ9oOk2UWqBurqe6PWWplfzrf6nki5CuVigMkbRQ7075aMj6yx/j7Tt7i9LH0gMu1wTStj50H8wBRa36UQ9RMoUKBAgQIF2m5pm7aobA45t2Me4II96HkezENeqXnhI/5B0UPvG4x40mbSPt8Tu5Ea7AHO07HisU2LxKD1qOO851wyaI0Wo0TvFqOhSvr1RIV/m7JLl5rT/46SOCzyu9NQh/vpF2dOnFJrs9POS92C5latG/R5j/gw7LH7ri9RS4GiIT4LfK4XDcrny5exyOdJXTZcuehdbawgjYi0L8GXN5oagSDWmQnjxerjPW+gQfpZpCRJNKrBx5hpmFGvrgi+jBlatHjRS7ZZ/L2xCNeh0QjHyRopiWU130B/8oYOcy7cvrW0thR8FgaXrQ3pksSJ6VndY3gd6+GyNA7Y0DVHLp/DWocNwc3gHW+okU36izYasopB6nJrpTWXSTd3kzwzDfgpQ+xREQAfIIX8jRCrciP5+7K/5gbfW+kHSKRi56k/D0OsOX5bI9ae7lGjh+BxiHUyyDzT6jpW/F6G7Ncm0spVZh+cOnX/QThzZUulctPdTzaKgkUlUKBAgQIFCrTV0j+fRaXgVkBsC57efY90q8XSl4Oe6a5PCu9lc97x6iHYPbnnAGIilMpd5YDc1beKthPlxEMhk7T2eCdYLSd3mHryNrysWS8aN+8Oo8hqgPrh8m8hCnKnzfUSDbPjBEFBzXm3K5/CLDMaJxpFYcePbTVLBs8VgpZpMyuxrHiPe/1WS0PEe3f5nR77pch5nhU0zkb9yMWoiJjbRCZ+/kKT0+MgGJRbxccBI2tYn1sP810wGiZHjiVLDPbCqFEjnX7Bmy8IRkEcu3e+OfPoFPBRAJuKnb5N1SrvnW3UgEXIpZ8H+XX7MVa0dqulurgZCqbLeafsUX7Ub8B8NDc3o6yWBfEL0ulg3ZQhVwXz550yVBcZ3WmHiYPygsI9u6aj99ZvHnHtmd/9MbP9cnmBv/bg+jaw7efmz7Pr0kNktdY71uXKvaKDCq/0i2MurNwbF76XqZ9BioS+NZQZ+XxhgYkQa66Ij5oGi7gYIOQoabAEu/vdDjuYKJA7b79Dfq9rv1M/KtHDpkFUd743N5k/QWPGGmTukcPFz0TblPHx5tTOmR07rkdi2LhoIYU6dB7dObBrjPPrjuN6yUtU7Fvm7aEWaNeNwlM59dY/fZB0n2NuI91UWJx/g9i3TKPQdtzBWKdXiRXTyif3SI4l+8t9QMZNo1XhEH3XymX6vkSF+rnPuFY5RfV9mSlYVAIFChQoUKBAWy1t0xaVzcJRiYAB8Xpmdl/fZ0EjMOjt3nA/aH6g53OuKJs8Vbvaj0Y4ZLnV8qSqdeK7MKqNWWalJR5vxeubbTFiZyj/kkzQQUdKxAOjiyLkFk1Q66AG5kb/RKnrwb3gBeORT0989XRXlcW1bqRE3VTky1QtB6MkC7COQ+r2M4/d7zp2mqWUvEr3RZtfuMD4MDDSRecqs3X42qhagqTu4cPM3fMaQbYc2yH5l8TbXbEwiB6Zu9oRrUEaJZHmaJJ77YpgjaTEg9C5cK1c7JiNtHBlkkrg0pUrpJyVrSLlEQpomET5lKKeT9VCyYK71257mJ9lDH0/gbTf9/QXlinnEpFQr9exQvyacldpRXV9v9tflWNpk4OsSqD5fclC+qjI2uvn2vOi4GL7RdevBsaYdxlxtmTZUilHq4+3LlJ//ZJld3y6u41vwKLFy5CluVOWQ6hRElzfqoHTokKeJYqJaMh112rHOeUa4/t5BG0716gs03qLWFJGCx6IWtQ82VNEbZGljOvca4vjyMzAWZpbc4yuT5kn1kE0VG/tMCKF2cC1Ho3ikv7mngVuEL/Aar9pi/5ba9f0Stuyn3HCOBeKzO3KXn3AWmsAoFmihFZLTidGv6Badyz0gB07nV/iwmg58/xZyQq+1z77mV8ZLaMWVbd/lAfOWZTnOhT0c8z075HLEzFodE/xMWkYaSly0SaReY8/8nfT/3e+SwratZf5SMOK0RRwVAIFChQoUKBA/2S0TVtUNgtHBUDCU1/idl/v/4gzEDOTJ7UWwS6QiJpI3vfAZvWulLH0mr8iyrQ0sQrWEqGVdSREJqUWLxoXT+Za0kULpaZeFe2nJjHwiBmXX9M7TUXMzYiKKpVk7t3mqE6jga3qNZpEmcilikTKu3+e8uUoTyTXmOiMxdO0PBOfklwRa3n/SQ1Z0H6JaClVxDnnzLTZv970s5lZSInkqKpZpMiUNkdLUqxCVUlmbqW1p6lF8GSI5OgpEfayn889dE3kaB9tLGWrqDF670DRi3l/bogRBvDy8ETSv2UrVjk8K44Bl/QgSkseE7mS1izXF2W05Evy14X6ixAdmFoj8zSp75b1L2ob1eG0SS0uJrpl6ql9sWcxk3GqiE8TMwCXBQuI85x7mW7zqG4xLDgv6nxl+F8qWEPjx02Q7po2UrVyuHhIOsY2JTBHxvleSsqAoIDSOpmJJSD2ULBj4t1IuZqUq8ie8+QzTwMATjrlZOGF/TPlk4rIB619BV8s7kuRLJrRnWZe1w8wvxTRX10EWvXRkTHV9Vvy5UGsIeoLEmmbrCOTtpPE14Vd/6hYvveuM/4fI9pHQRoX1ri2Ive77EHmGfdZw0O7ZDduazOWAY496+D6zDxEct0zIsqW+X20oIenmmeMmC4A/b9yD6049/Ooce2Jv8haibRjjrLIQ//V+aYFhfVzP8/r1mdSUI5HdZg1x/Vr/bwEmZe/q++k6xfDNdosmD9cN2O8SNI4Kes+7mcmT31rDfsTcFQCBQoUKFCgQNsrbdMWlc2lsmQsrvsRB/yU0ytRMH0EQ5bT3AfUgxUlMnJKluTEW0R4tN4N5nQ5XjAs7D25nP7ZltTRgEzrfTYJNsTiRQucB3EeIdGMxfShcTVq1XJE0yJeQEXGq0xUUPle85Bp1UiQu/2uVmvaL+Jh7C0ZSC0aKnmgVm64occ5EW2t07z5B3NmPPXsHHkeOc+jKFK+a5mLLEqihkUrBnPBsC7yqLl96NvizbdaAQTbpTYwoP1qH2mifqjla26PjGi51BghvFD2iI5paLRYaEaMGCHjIxmtmQmWnYrsWJFvWvF8Ywu1O/uuywu1JrZFOWYUBV0dOM79/f1Y3dMjVam5zpQRS0C13/WLIiWxq/2NGmW0Y/oJDBP00bJk8q57yLRA1DAPVfExoVWO/dC1QatdnX4R8jvc8nUPoZUV8F5+wYIFBRwJN5KC85nR/yfLnHKxt1esET8LzjOrZb4qRZxWvxGL9sL1mVY53+ZJc0VQbQuWr2LdDTnPuH49VFkSI9qSJFF/Nl+OrXx6Y+dp1fTpYBSMxdPJnX5nnl9YsR/ab2bz1vmSNRe76zivuZE4rEdltEbfJfPe4iUmtxH3i3K5jFrm74G0fMj32LU+MXJwtFgcm4kW7uHisHf+fkFk2hiR7kNE4E0iN7+UXcecT5FjD1Fb9znpV5K4Vh8fPqtULinKt/07NPh6frkoWFQCBQoUKFCgQFst/dNZVPLCXa7FCzHf9TToO143nP6pxRjKvGghGzs/+HvFf/NkOiReih/N4d0J2gfmo1000LW9a9w2i/Vkbr9t3dQo3ey4ayWe/qX6U+TJ4spY9YD9YYTJaw8/zO2Hz0whh8lgbZGoBaxbx1w/3t1wQXsYauwsfoD5B737hzE/ize/ZMrnqQFVNs9RFauTxXpIvSJ+3Q4rDbLEDMd9fX2D12O5aQiMaGhLPl9cYKxvI0V79++dFR/Ir4cy5cloV1cXXpj/glMXQxH8OgbpsPdd1gktS5lr9cj917K8gLnj8SulqCmPIX6MPy4acQevHpc35UH6P3z48MZ+ZW5Zu/YaZcWUE+sALQiUXx1jb58oRCD5+xLfYF6hJPH8IIZYU5E3xl6QHEYK4qvNnRU1TkTDPA3RX/lkNnArm64fRe6PeaGvft2aJ0ny6Ng6PSczT84zrx5/jMd1E0fKyrLHbuM4NPwNMQWXi89ak1i5GjCuGg1HDTxRzjOxIJXLvm/KBurg3xg/spDrZMCLlnT67b5Dv66h0L7/UQoWlUCBAgUKFCjQVkvbtEVlc3BUotze82UFbUR+AGC9/2uR8ZLPvTs9fq9JtlLedbK+2Duh1mpyR57V9HSbSsTEk088DgB4+ylvN2VS+lHIaZnx9TVil7gnV9UGUuIumHLLli9zyudpikhyQqSK9eBpTnKPGpdNXXW596+xf/J+rdbvtEnSLMzCc55IucxaEdavNyi3qjFKfyPvNK9jrG2JRq1j7GZRbZG7Xh0/+T1NU9R1/D3kUfkHUUDVmV/voCtOf4jlEHvZk1UEOVU1uZeu1/DY448BAPafvK9pi34iSVXqpKXAtU7U67HyX+wP2+TddpXjw4itguZGNNeYkQKcL2qr0q9hkpmb/hBp6uGK8M7fH0f5nf4h9OWKAIxqb3PKEqsmGuhz3s091ZNRDik1NfHpWC2YGKwnFVRZXcuMuIqsFXLAkx36c6xZYyxm9Fmw+4GMdWrXDABEVU+OvbU3qq0dgFlzDWtC+GIWdGJw6Lrk/T+xLoTnmkTojBol46g8uWuzaDUj1kZV5olZzJ+QPeY1e7/G8CC8RGXxB/JUb44xpJ40dX27hou8rJW1nGV1K3/SL8ptfci9QliUsad/U+eYDq9NRj+6/mDFNZzrfIs1Sua1RVBuuYfynUTqZlZwNaDVOX6RMw6Un0UvGstjTcYPSax12H6J7xDlscTs9SJL4kfCSMpRHcZ3jdmzYy+KLdX9W37mespSla0Rgv+0dt16aYu+N+xXKnxD3y02wr9bmWa+TmUcLC6SvGh4rA0grXv9VraHsNIFHJVAgQIFChQo0PZK27RFZXNwVAB72k08jA4b8u9GxdhoD3nK9+TkmiRudIjVcmgdEE28FFuEQtFyR42WmH/RQFmXHocTyfBKjAdhJWbkiRQrCa5AWbSkzjFjTDliwSQWk0WvqmMP06DkWikYgdHfb7SiUoX4DBwXN7KB6mFccSM94sTm4W0SXIxRgpfAk3iibvJkwR3jUsnllQoWcQiqj/9N2nIxLkpxGanMo843+yuVJTJ2cU7NrCptEpPE9WinzNEqpPqB4JNwfJJKrJYFRq9opJjyJLKmETqUIXmumopEJMg99DrJv4TcnQsXhVXGTH0U2Agxaoxs0frWLnOSlFytjhYZXQ+sJ2KEDvFyKF8JurrGOmUTmqsKMmH5tJRxgTB4QmSOa7YiUW3UCktl8mp1LpsfSvpdkf5yfsVa0d09XvrnrrmYGEVlWs48ec/ctTei3WjFtVrNjrFGXMRuHYrgmTh1xIq7QRRVwQaR+eaSzfy5LFhUKMfUfLlXMI9Ox6h2AHaPyIirBJIrH+TFX3uUUe5ZeZba/F7EENJ9S+SYuElsSv5RaRGrpVgtWiULeqw8DOEPRhnOgTyh+cHsFS3Nps6u8V3Stuebk3HfduVc5Zl4KdLPsePMXrpylfErKYsMAhlKMpa+e1NJ2qRFhN+JwUIrECMqWQ+8enQD0FxgjFCKLT6UyFrr8GHSL3dPSRTzxF17dkQl2keEcVhrxXmu+6AiM6d2jxBSo7LnB0UKOCqBAgUKFChQoO2WtmmLyuZQHgEVyfGT8QTNh0ToE+2s4vkokNTLupl3oD6OiuupXxbU1PpA1eKeSJGSaB7jxprTvzp7s4DcbZYEJZU4KiSbGsdoCTtOnAQAWLde7qmp6WQ5EumP5vrxcFR4z6yol9Kv3feQHDAyUuWK6TdxVPR8LOiiZD0pG57rVYtYwwyu4wTtMIkEydOPmBJ030qT0RKIiaCahlRIX43hw4g3wRM9C6ZoEn6JueGfzhXtsyzWHvE5IFmdXzQYanuav0TqkfkutZh6atWq8tXZ2enUWRI0VM0XpffF5nlFogEUw0fGp3uCsQJYHBV3LhymaTESLUfzxvCBh+g5YqQb9WNdtxhVwHHkvMv9ucheSeTm4IMORq9ETkU5tTmWkTr6ic1BPzE3N0os1gH6Zuy2627Od64HYoVYn49UNbZKyc6D4dP8PkaQWonZofhCXBfe+q20uPLDcbG2O1Oua1xXQf7kmaypkshpXnN9zLh+GVlVkfFZ07tW2qDGLmuTmWzV4mjxc7gHlAVzhr5XlH/iIpHfGGV91/RbHSEMzxW7bxX729VlZHDSTjsbnpKKReumTEi/rcy4/iB+hna1/EZE5qZJTSwJisTdEIIFCnpZ9vXXvMb4g+08yeyFFg1YPrmve/hYJO4D9M1obTbj1tHRLk3T2tWC1JMJtpXp2qN/mPjJSMFJwtuw4WKN0/xMcMrRdykhRgxz/kTQeaJVeb3k0Yo85G2OcZn9GnDXnkbseAjdLcyFx74pvk4LwHXHeWDmbs+f8+WiYFEJFChQoECBAm21tE1bVDYre3KB8kE8yJ3vdKDwU6FaBgb99LPO2nv4qJCzw/xjxowZ0ibrgPM9KkRxGGJuCfeejxrK3vsYxNc9dtvd7VNu6yA/2RDnVG1bvh966KGFOgrjNtTYWxOL8s5+n3PuuQCAVsEoaTx/u5PAObCItW6xiTtOBAC88Y1vglPAGR7Os3c3a1VJpz/Udqy25/kNcfwa+u/6HyCP8LojX2f4nCjIww36uCtjdrapUdIiJr9LgaOkXiujnuxFKPhDuf3TIZL+/cu/HO11Y3B/Ah9xGf5zoREjRmBY6zCnqB1zyu9g8wRFNdZ8LGIxmDB+vMOz+lGwb4WO64gU5sE8MZ+nnPI2AMBYwVHRbvjrWR0PXD8Yf+1NEFTp6dOn25wvDbgwns+Kl3Y398odeMgh8rq7Fgez57CLtgr2w5Q5+OCDAQDd3WM9loaQ34YFwn67LX/wQxfIDxGQ0ZfC2XDsfPv7G9uQ8nvssZdU5fp92fUw+BqOouIX09YBB0yTZyzrW04GXzORt37VwiT1MEu2IvpGse2X8uMibTfu5+b7xy6+xCnv86L+QJ7fkLNupNABUw8AUIgo8/cCeHPhzbvPKz/HdI5xfi6u4Ya/nX5dHv2jUT9RvomeLXfffTeuuuoqPPjgg1i0aBFuvPFGnHDCCRtk4sorr8TFF18MANh5553xvKS5Jl1xxRX42Mc+tlE89Pb2oq2tDT09PRgp0OQbS7UlixGJCTRLvAmjgypNXOLIR5M5KVfzM69+3HBNnS4JtSyLw1jRPE9ToJpy6VznhQKiJry00lzpQxdLfXVeEZk/EGm1zy2R5igJ7Dgd8ophwwAACQWNxFmMkOu5ljfFWuSQUZOwTWuOlcXIKySpp163Vz/cxzQxos17LyXkj4uYksutYr734MvtBm4+E3FG0/FR03gNzXJNQNAwWwf/Yeqgk9yKZUsBACPlaiAe5I8C0Gjm1NQBIhe1/vX2mZfEq1Qx3xWK2juQVuRakaGknDO2odeKIltVlS1e28TKKJ0f6xJSSKdvyjET/zHs0v/jwLYSudJKCQQlvzO1fROfp6kN3eYAyFooDRPZ6e+T/ugAOeNDB/Y0rRd/1nqaWA95Tu3a4xmg3OpC6JN4+GniGDMEVq/GNrx+i8RxSCrNSOsDTh25XJ+yDg3d98HHuKcISNnTTzwKwB6aeaLn9ZpeFRYBDXUPECWgztB3F75dr4tK3l5Nh2SuvRZXLizR4VvkqxxruGqeMV2DJAsdbnihfDb80RMHzn4JhW6q8DpK3YcLLVrnYzjOmVwz3CtkX+IhgP7ZdCYdYl9nQr2mJu4XEo4vsvXE408CAPYVZbDUVG6QCV5hkadI7vZSvRqTgxivFWV/G2CKEPZJPnnFl1QIoU+YilzXTkWch2t17zpGO2Y+yuKoXOf4kGcF15PvIhe33vK/AIDXv+H1Uo+sh9YW1KsuGBz/7lcfM3AMrfvv74xH8fKmt7cX7e3tWL169Ub//d7kq59169ZhypQpuPrqqwd9vmjRIue/H/7wh4iiCCeddJJT7vLLL3fKfeADH9hUVgIFChQoUKBA2zlt8tXPzJkzMXPmzCGfM0kc6Te/+Q2OPvpo7LLLLs7vI0aMaCj7alBeKiN78UUAQFZm912zHJ2o7HOX6DhWTwiYxNBa19xLDS0dpB5q43VR/xKIo6tvEq4JwFV58LAvUiqAQVnSTC4b6kkZUjaE1StX4DOG5XptSH/6Sn5/XHtlRvCiUhmNZArVRNP0QwcjMSHXM3PqT0sVDEY8/VOLrJdckzHrSfMB9Mfsz+D9Zr9S6ddw+T1bR3Ay0xpBtpIh5oKAWzXRfuKiYkNQJen3UPOp4HIlz9F7iLYGky02p1Y7b4xpe8hEtjI+94dH+GeSNx+CXYGypJ1+hmsPMsz+GPt1qIUgJVDWUOMjWmODDFqedWwqg5chGFk/Q2mHkIsNjTGpLlbQUtw09PotvXRbBGqsl8z63Ukc8LNFJhEe5TzzrrwcfsWqUZd1l+iV5uD7Uuzvf/Tfl0R76aDrt3HtFefKX799G1y/hB8Qp1OvniwSq4do89EgckGZyHyZ8HDoOU+D70sblovXdBgwuvoL5u/HYHJBi3Gd4fMNaQuk3AZkVHmquXMVOc82sH7Jk9RRH2xfKvBMq3wse9Mx+042vy80ySeHHGfY9btJ1zObQK+oM+2SJUvwu9/9DmeeeWbDsy9+8YsYPXo0pk6diquuusqi3w1CAwMD6O3tdf4LFChQoECBAm3/9Io60/74xz/GiBEjcOKJJzq/f/CDH8QBBxyAjo4O3Hvvvbj00kuxaNEifPWrXx20niuuuAKf/exnXxaeKqNHA6NHvyx1BQoUKFCgQIFeWdpkZ1rn5ShqcKYt0l577YXXv/71+MY3vvGS9fzwhz/EOeecg7Vr16JJYreLNDAwoM5GgHHG2XHHHTfLmTZQoECBAgUKtGVoc5xpXzGLyp/+9Cc89dRT+PnPf77Bsocccgjq9Tqee+457Lnnng3Pm5qaBj3AbE54cvFcpuFrG3lWGzItulePz9NLtbkh/jdUbkM8FcsNxdem8rQheql6GqCwN9CvlxrLzXl/Y2hj29iUNjdU54b4/UfnJFCgQC79Azr6JrdhQ5rddZwNAZC2KX+bNjkx7xbeS7aqpIQ/+MEPMG3aNEyZMmWDZWfPno04jhtwDQIFChQoUKBA/9y0yRaVtWvXYs6cOfp93rx5mD17Njo6OjBxogHf6u3txQ033ICvfOUrDe/PmjUL999/P44++miMGDECs2bNwoUXXoh3vetdmrhtY2lzkhKuXbsWv/zlLwEAp512GgDgd7/7HYACZoHQvvsaKOZbb70VADBGEv2tXLkSADB37lwA0Iim3Xff3fm+fr3B0bjvvvsAmJMknz3wwAMAbLK68QJotdtuuzlttggmR3d3NwAbkUQI9SVLTFQAI6jIG+uhae2xxx7Dww8/DAA46KCDAABr1qwBAAyTtO28XiO8/YsSHVWpGM99JsLr6zPRMHvvvTcACw//6KMG+2H+/PkAgMmTjdf48uXLFeCN706YMMHhnxElHA+WJ89XXXUVAOCCCy4AAD0A33nnnQCAd77znQAsXsMvfvELAAZin/NGCHHOC+dr7VoDV875LAtc9RFHHAEA+NWvfgUAOProowHY6IAHH3wQALBKkpURkp3tzZs3D8899xwAoEMiBt773vcCAG655RYA0DXDNjlvN998MwA0+HcFChTo5aEXXngBgF3HJO5N3Me4t9ISQDymOI4b9l+WIWAh9zXuuT795je/AWAtK9wPli9fDsDuG3zOm4Vly5ZppBT3pYULFwIAFixYAAB46qmnANi/AccddxwAm3ZkS9HmWLI2+aDywAMP6MAAwEUXXQQAOP3003HNNdcAAK677jrkeY53vOMdDe83NTXhuuuuw2WXXYaBgQFMmjQJF154odYTKFCgQIECBQpE2uSDylFHHbXBE9HZZ5+Ns88+e9BnBxxwgFoYtgSVy2XV1pcuNQiktFrwkydQnlBZnpoyNWr609DKQWsBT9U8kff3G0yBESNG6ImYJ2eeznk65kmcGvZeexloaZ76ae1YvNjEtlObpyWBp+VFixYBsFp+S0uLWk6oBfCdZ555BgCw0047AQD+9re/AbDWDVogWDf7SysQT//knTg7PNknSaIWIFpxaK3ZddddAQCrJYkdx5waCOH7X/va1wKwY8654nhwTjiO5D2KIh3rHXfc0eGLFiRaucgTy3FsOf+sm21Ty6G1i3LDvuy88846v+SfdRxzzDEAjKVrsHcPPPBABAoU6JUj7sujJQqU+xz3kvb2dgDWEksL8UMPPQTA7IPcU7m+uZ657vk3gXsm93US9xL+zv2R1mp+963dPT09DZaRFStWALD7Fvdz/q3w296WKCQlDBQoUKBAgQJttfQPhSdvKfpHcv0AQ3tak2gZIPGEvbERJppQbSNSXtvEd/Ggv2/ovQ15UBd52dgon02tm8T+DvX7YHUPNVZ+Hf77/ifvgvldc80U+uC3tSH+/Xkfapw2Jton9tAufT5Jfh1DjUOgQIH+MeLa29g/gT4yc5ZlDXuDvxcMtdcMxcPGRmAWeSZf/l4yVB1bek95VXL9BAoUKFCgQIECvVr0iiLTvtK0OTgqQOPJeENUGiqvyMvczitJrwYvQ7WxMW1vLH8bmm/fUrMpbfm/b+68bwxtDJ+BAgV65egf3RMHe9/fnzbUxsu5L29Nf29eijbnb/Y2eVChSSvk/AkUKFCgQIG2HeLf7U3xOtkmDyr0hGZ0RqBAgQIFChRo26E1a9ZoVOqGaJt0ps2yDE899RT23ntvvPDCCyHfzxYi5lwKc7DlKMzBlqcwB1uewhxsWdqU8c/zHGvWrMH48eM3+up7m7SoxHGs6IEjR44MgrmFKczBlqcwB1uewhxseQpzsGVpY8d/Yy0ppODJFyhQoECBAgXaaikcVAIFChQoUKBAWy1tsweVpqYmfOYzn1FI+kCvPoU52PIU5mDLU5iDLU9hDrYsvdLjv0060wYKFChQoECB/jlom7WoBAoUKFCgQIG2fwoHlUCBAgUKFCjQVkvhoBIoUKBAgQIF2mopHFQCBQoUKFCgQFsthYNKoECBAgUKFGirpW32oHL11Vdj5513RnNzMw455BD85S9/2dIsbZd02WWXaZZq/rfXXnvp8/7+fpx//vkYPXo0hg8fjpNOOglLlizZghxv+3T33XfjLW95C8aPH48oivDrX//aeZ7nOT796U+ju7sbLS0tOPbYY/HMM884ZVauXInTTjsNI0eORHt7O84880ysXbv2VezFtk0bmoP3vOc9DetixowZTpkwB5tPV1xxBQ466CCMGDECY8eOxQknnICnnnrKKbMxe8/8+fPxpje9Ca2trRg7diwuvvhi1Ov1V7Mr2yxtzBwcddRRDevg3HPPdcq8HHOwTR5Ufv7zn+Oiiy7CZz7zGTz00EOYMmUKjjvuOCxdunRLs7Zd0j777INFixbpf/fcc48+u/DCC/Hb3/4WN9xwA+666y4sXLgQJ5544hbkdtundevWYcqUKbj66qsHfX7llVfi61//Or7zne/g/vvvx7Bhw3Dcccehv79fy5x22ml47LHHcNttt+Hmm2/G3XffjbPPPvvV6sI2TxuaAwCYMWOGsy6uvfZa53mYg82nu+66C+effz7uu+8+3HbbbajVanjDG96AdevWaZkN7T1pmuJNb3oTqtUq7r33Xvz4xz/GNddcg09/+tNbokvbHG3MHADAWWed5ayDK6+8Up+9bHOQb4N08MEH5+eff75+T9M0Hz9+fH7FFVdsQa62T/rMZz6TT5kyZdBnPT09eblczm+44Qb97YknnsgB5LNmzXqVONy+CUB+44036vcsy/Kurq78qquu0t96enrypqam/Nprr83zPM8ff/zxHED+17/+VcvccssteRRF+YIFC1413rcX8ucgz/P89NNPz48//vgh3wlz8PLS0qVLcwD5XXfdlef5xu09v//97/M4jvPFixdrmW9/+9v5yJEj84GBgVe3A9sB+XOQ53n+ute9Lv/Qhz405Dsv1xxscxaVarWKBx98EMcee6z+Fscxjj32WMyaNWsLcrb90jPPPIPx48djl112wWmnnYb58+cDAB588EHUajVnLvbaay9MnDgxzMUrRPPmzcPixYudMW9ra8MhhxyiYz5r1iy0t7fjwAMP1DLHHnss4jjG/fff/6rzvL3SnXfeibFjx2LPPffEeeedhxUrVuizMAcvL61evRoA0NHRAWDj9p5Zs2Zh8uTJGDdunJY57rjj0Nvbi8cee+xV5H77IH8OSD/96U/R2dmJfffdF5deeinWr1+vz16uOdjmsicvX74caZo6HQeAcePG4cknn9xCXG2/dMghh+Caa67BnnvuiUWLFuGzn/0sjjjiCDz66KNYvHgxKpUK2tvbnXfGjRuHxYsXbxmGt3PiuA4m/3y2ePFijB071nleKpXQ0dER5uVlohkzZuDEE0/EpEmTMHfuXHz84x/HzJkzMWvWLCRJEubgZaQsy3DBBRfgsMMOw7777gsAG7X3LF68eNB1wmeBNp4GmwMAeOc734mddtoJ48ePxyOPPIJLLrkETz31FH71q18BePnmYJs7qAR6dWnmzJn67/322w+HHHIIdtppJ1x//fVoaWnZgpwFCrTl6NRTT9V/T548Gfvttx923XVX3HnnnTjmmGO2IGfbH51//vl49NFHHd+4QK8uDTUHRZ+ryZMno7u7G8cccwzmzp2LXXfd9WVrf5u7+uns7ESSJA3e3UuWLEFXV9cW4uqfh9rb27HHHntgzpw56OrqQrVaRU9Pj1MmzMUrRxzXl5L/rq6uBsfyer2OlStXhnl5hWiXXXZBZ2cn5syZAyDMwctF73//+3HzzTfjjjvuwA477KC/b8ze09XVNeg64bNAG0dDzcFgdMghhwCAsw5ejjnY5g4qlUoF06ZNwx//+Ef9Lcsy/PGPf8T06dO3IGf/HLR27VrMnTsX3d3dmDZtGsrlsjMXTz31FObPnx/m4hWiSZMmoauryxnz3t5e3H///Trm06dPR09PDx588EEtc/vttyPLMt1IAr289OKLL2LFihXo7u4GEObgH6U8z/H+978fN954I26//XZMmjTJeb4xe8/06dPx97//3Tkw3nbbbRg5ciT23nvvV6cj2zBtaA4Go9mzZwOAsw5eljnYDOffLU7XXXdd3tTUlF9zzTX5448/np999tl5e3u741kc6OWhD3/4w/mdd96Zz5s3L//zn/+cH3vssXlnZ2e+dOnSPM/z/Nxzz80nTpyY33777fkDDzyQT58+PZ8+ffoW5nrbpjVr1uQPP/xw/vDDD+cA8q9+9av5ww8/nD///PN5nuf5F7/4xby9vT3/zW9+kz/yyCP58ccfn0+aNCnv6+vTOmbMmJFPnTo1v//++/N77rkn33333fN3vOMdW6pL2xy91BysWbMm/8hHPpLPmjUrnzdvXv6HP/whP+CAA/Ldd9897+/v1zrCHGw+nXfeeXlbW1t+55135osWLdL/1q9fr2U2tPfU6/V83333zd/whjfks2fPzm+99dZ8zJgx+aWXXrolurTN0YbmYM6cOfnll1+eP/DAA/m8efPy3/zmN/kuu+ySH3nkkVrHyzUH2+RBJc/z/Bvf+EY+ceLEvFKp5AcffHB+3333bWmWtkt6+9vfnnd3d+eVSiWfMGFC/va3vz2fM2eOPu/r68vf97735aNGjcpbW1vzt771rfmiRYu2IMfbPt1xxx05gIb/Tj/99DzPTYjypz71qXzcuHF5U1NTfswxx+RPPfWUU8eKFSvyd7zjHfnw4cPzkSNH5meccUa+Zs2aLdCbbZNeag7Wr1+fv+ENb8jHjBmTl8vlfKeddsrPOuusBkUpzMHm02BjDyD/0Y9+pGU2Zu957rnn8pkzZ+YtLS15Z2dn/uEPfziv1Wqvcm+2TdrQHMyfPz8/8sgj846OjrypqSnfbbfd8osvvjhfvXq1U8/LMQeRMBQoUKBAgQIFCrTV0TbnoxIoUKBAgQIF+uehcFAJFChQoECBAm21FA4qgQIFChQoUKCtlsJBJVCgQIECBQq01VI4qAQKFChQoECBtloKB5VAgQIFChQo0FZL4aASKFCgQIECBdpqKRxUAgUKFChQoEBbLYWDSqBAgQIFChRoq6VwUAkUKFCgQIECbbUUDiqBAgUKFChQoK2W/j94TgkWfHI+ggAAAABJRU5ErkJggg==\n"},"metadata":{}}]},{"cell_type":"code","source":["def vit_block(x):\n","  \"\"\" Define the transformer layer blocks for vit model\"\"\"\n","  prenorm = x\n","  x = tf.keras.layers.LayerNormalization()(x) # Add pre self attention layer norm\n","  x = tf.keras.layers.MultiHeadAttention(num_heads=32, key_dim=16, value_dim=16)(x, x) # Make a self attention layer(with 4 heads)\n","  x = prenorm + x # Residual(Use it so model predictes only the changes insead of the whole information)\n","  prenorm2 = x # Store the output before normalization\n","  x = tf.keras.layers.LayerNormalization()(x) # Add pre mlp layer norm\n","  x = mlp_block(x) + prenorm2 # Residual\n","  return x\n","\n","def mlp_block(x):\n","  \"\"\"define multi layer perceptron(dense*relu*dense)\"\"\"\n","  x = tf.keras.layers.Dense(units=2048, activation=\"relu\")(x)\n","  x = tf.keras.layers.Dense(units=512, activation=None)(x)\n","  return x\n"],"metadata":{"id":"NRuexaLb2JxL","executionInfo":{"status":"ok","timestamp":1697229891095,"user_tz":240,"elapsed":5,"user":{"displayName":"Mahsa Bakhtiari","userId":"12793844954925855500"}}},"execution_count":6,"outputs":[]},{"cell_type":"code","source":["# Define the model's artitecture\n","input = tf.keras.Input(shape=(182, 256, 3)) # Define the input neorons\n","x = tf.keras.layers.Conv2D(filters=512, kernel_size=(7, 8), strides=(7, 8))(input) # Patchify\n","x = tf.keras.layers.Reshape((26 * 32, 512))(x) # Flattening the inputs\n","# Add Positional information to the input because self attention is permutation invariant.\n","pos_emb = tf.keras.layers.Embedding(input_dim=26 * 32, output_dim=512)\n","pos_indices = tf.range(tf.shape(x)[1])\n","x = x + pos_emb(pos_indices)\n","\n","for i in range(8): # Add vit blocks to the model\n","  x = vit_block(x)\n","\n","x = tf.keras.layers.LayerNormalization()(x) # Final layer norm\n","# The magnitude of the residuals increase with the network depth, we need to do\n","# a finaal layer normalization to avoid really large activations here.\n","\n","x = tf.keras.layers.GlobalAveragePooling1D()(x) # Reduction of spatial dimentions\n","\n","x = tf.keras.layers.Dense(units=4, activation=None)(x) # final classification layer\n","\n","model = tf.keras.Model(inputs=input, outputs=x) # Make the model\n","model.summary()"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"xVLeZDRb_oHx","executionInfo":{"status":"ok","timestamp":1697231081322,"user_tz":240,"elapsed":1380,"user":{"displayName":"Mahsa Bakhtiari","userId":"12793844954925855500"}},"outputId":"49109e9a-e327-4ad3-df76-0d5aed4d0dde"},"execution_count":15,"outputs":[{"output_type":"stream","name":"stdout","text":["Model: \"model_2\"\n","__________________________________________________________________________________________________\n"," Layer (type)                Output Shape                 Param #   Connected to                  \n","==================================================================================================\n"," input_3 (InputLayer)        [(None, 182, 256, 3)]        0         []                            \n","                                                                                                  \n"," conv2d_2 (Conv2D)           (None, 26, 32, 512)          86528     ['input_3[0][0]']             \n","                                                                                                  \n"," reshape_2 (Reshape)         (None, 832, 512)             0         ['conv2d_2[0][0]']            \n","                                                                                                  \n"," tf.compat.v1.shape_2 (TFOp  (3,)                         0         ['reshape_2[0][0]']           \n"," Lambda)                                                                                          \n","                                                                                                  \n"," tf.__operators__.getitem_2  ()                           0         ['tf.compat.v1.shape_2[0][0]']\n","  (SlicingOpLambda)                                                                               \n","                                                                                                  \n"," tf.range_2 (TFOpLambda)     (832,)                       0         ['tf.__operators__.getitem_2[0\n","                                                                    ][0]']                        \n","                                                                                                  \n"," embedding_2 (Embedding)     (832, 512)                   425984    ['tf.range_2[0][0]']          \n","                                                                                                  \n"," tf.__operators__.add_34 (T  (None, 832, 512)             0         ['reshape_2[0][0]',           \n"," FOpLambda)                                                          'embedding_2[0][0]']         \n","                                                                                                  \n"," layer_normalization_34 (La  (None, 832, 512)             1024      ['tf.__operators__.add_34[0][0\n"," yerNormalization)                                                  ]']                           \n","                                                                                                  \n"," multi_head_attention_16 (M  (None, 832, 512)             1050624   ['layer_normalization_34[0][0]\n"," ultiHeadAttention)                                                 ',                            \n","                                                                     'layer_normalization_34[0][0]\n","                                                                    ']                            \n","                                                                                                  \n"," tf.__operators__.add_35 (T  (None, 832, 512)             0         ['tf.__operators__.add_34[0][0\n"," FOpLambda)                                                         ]',                           \n","                                                                     'multi_head_attention_16[0][0\n","                                                                    ]']                           \n","                                                                                                  \n"," layer_normalization_35 (La  (None, 832, 512)             1024      ['tf.__operators__.add_35[0][0\n"," yerNormalization)                                                  ]']                           \n","                                                                                                  \n"," dense_34 (Dense)            (None, 832, 2048)            1050624   ['layer_normalization_35[0][0]\n","                                                                    ']                            \n","                                                                                                  \n"," dense_35 (Dense)            (None, 832, 512)             1049088   ['dense_34[0][0]']            \n","                                                                                                  \n"," tf.__operators__.add_36 (T  (None, 832, 512)             0         ['dense_35[0][0]',            \n"," FOpLambda)                                                          'tf.__operators__.add_35[0][0\n","                                                                    ]']                           \n","                                                                                                  \n"," layer_normalization_36 (La  (None, 832, 512)             1024      ['tf.__operators__.add_36[0][0\n"," yerNormalization)                                                  ]']                           \n","                                                                                                  \n"," multi_head_attention_17 (M  (None, 832, 512)             1050624   ['layer_normalization_36[0][0]\n"," ultiHeadAttention)                                                 ',                            \n","                                                                     'layer_normalization_36[0][0]\n","                                                                    ']                            \n","                                                                                                  \n"," tf.__operators__.add_37 (T  (None, 832, 512)             0         ['tf.__operators__.add_36[0][0\n"," FOpLambda)                                                         ]',                           \n","                                                                     'multi_head_attention_17[0][0\n","                                                                    ]']                           \n","                                                                                                  \n"," layer_normalization_37 (La  (None, 832, 512)             1024      ['tf.__operators__.add_37[0][0\n"," yerNormalization)                                                  ]']                           \n","                                                                                                  \n"," dense_36 (Dense)            (None, 832, 2048)            1050624   ['layer_normalization_37[0][0]\n","                                                                    ']                            \n","                                                                                                  \n"," dense_37 (Dense)            (None, 832, 512)             1049088   ['dense_36[0][0]']            \n","                                                                                                  \n"," tf.__operators__.add_38 (T  (None, 832, 512)             0         ['dense_37[0][0]',            \n"," FOpLambda)                                                          'tf.__operators__.add_37[0][0\n","                                                                    ]']                           \n","                                                                                                  \n"," layer_normalization_38 (La  (None, 832, 512)             1024      ['tf.__operators__.add_38[0][0\n"," yerNormalization)                                                  ]']                           \n","                                                                                                  \n"," multi_head_attention_18 (M  (None, 832, 512)             1050624   ['layer_normalization_38[0][0]\n"," ultiHeadAttention)                                                 ',                            \n","                                                                     'layer_normalization_38[0][0]\n","                                                                    ']                            \n","                                                                                                  \n"," tf.__operators__.add_39 (T  (None, 832, 512)             0         ['tf.__operators__.add_38[0][0\n"," FOpLambda)                                                         ]',                           \n","                                                                     'multi_head_attention_18[0][0\n","                                                                    ]']                           \n","                                                                                                  \n"," layer_normalization_39 (La  (None, 832, 512)             1024      ['tf.__operators__.add_39[0][0\n"," yerNormalization)                                                  ]']                           \n","                                                                                                  \n"," dense_38 (Dense)            (None, 832, 2048)            1050624   ['layer_normalization_39[0][0]\n","                                                                    ']                            \n","                                                                                                  \n"," dense_39 (Dense)            (None, 832, 512)             1049088   ['dense_38[0][0]']            \n","                                                                                                  \n"," tf.__operators__.add_40 (T  (None, 832, 512)             0         ['dense_39[0][0]',            \n"," FOpLambda)                                                          'tf.__operators__.add_39[0][0\n","                                                                    ]']                           \n","                                                                                                  \n"," layer_normalization_40 (La  (None, 832, 512)             1024      ['tf.__operators__.add_40[0][0\n"," yerNormalization)                                                  ]']                           \n","                                                                                                  \n"," multi_head_attention_19 (M  (None, 832, 512)             1050624   ['layer_normalization_40[0][0]\n"," ultiHeadAttention)                                                 ',                            \n","                                                                     'layer_normalization_40[0][0]\n","                                                                    ']                            \n","                                                                                                  \n"," tf.__operators__.add_41 (T  (None, 832, 512)             0         ['tf.__operators__.add_40[0][0\n"," FOpLambda)                                                         ]',                           \n","                                                                     'multi_head_attention_19[0][0\n","                                                                    ]']                           \n","                                                                                                  \n"," layer_normalization_41 (La  (None, 832, 512)             1024      ['tf.__operators__.add_41[0][0\n"," yerNormalization)                                                  ]']                           \n","                                                                                                  \n"," dense_40 (Dense)            (None, 832, 2048)            1050624   ['layer_normalization_41[0][0]\n","                                                                    ']                            \n","                                                                                                  \n"," dense_41 (Dense)            (None, 832, 512)             1049088   ['dense_40[0][0]']            \n","                                                                                                  \n"," tf.__operators__.add_42 (T  (None, 832, 512)             0         ['dense_41[0][0]',            \n"," FOpLambda)                                                          'tf.__operators__.add_41[0][0\n","                                                                    ]']                           \n","                                                                                                  \n"," layer_normalization_42 (La  (None, 832, 512)             1024      ['tf.__operators__.add_42[0][0\n"," yerNormalization)                                                  ]']                           \n","                                                                                                  \n"," multi_head_attention_20 (M  (None, 832, 512)             1050624   ['layer_normalization_42[0][0]\n"," ultiHeadAttention)                                                 ',                            \n","                                                                     'layer_normalization_42[0][0]\n","                                                                    ']                            \n","                                                                                                  \n"," tf.__operators__.add_43 (T  (None, 832, 512)             0         ['tf.__operators__.add_42[0][0\n"," FOpLambda)                                                         ]',                           \n","                                                                     'multi_head_attention_20[0][0\n","                                                                    ]']                           \n","                                                                                                  \n"," layer_normalization_43 (La  (None, 832, 512)             1024      ['tf.__operators__.add_43[0][0\n"," yerNormalization)                                                  ]']                           \n","                                                                                                  \n"," dense_42 (Dense)            (None, 832, 2048)            1050624   ['layer_normalization_43[0][0]\n","                                                                    ']                            \n","                                                                                                  \n"," dense_43 (Dense)            (None, 832, 512)             1049088   ['dense_42[0][0]']            \n","                                                                                                  \n"," tf.__operators__.add_44 (T  (None, 832, 512)             0         ['dense_43[0][0]',            \n"," FOpLambda)                                                          'tf.__operators__.add_43[0][0\n","                                                                    ]']                           \n","                                                                                                  \n"," layer_normalization_44 (La  (None, 832, 512)             1024      ['tf.__operators__.add_44[0][0\n"," yerNormalization)                                                  ]']                           \n","                                                                                                  \n"," multi_head_attention_21 (M  (None, 832, 512)             1050624   ['layer_normalization_44[0][0]\n"," ultiHeadAttention)                                                 ',                            \n","                                                                     'layer_normalization_44[0][0]\n","                                                                    ']                            \n","                                                                                                  \n"," tf.__operators__.add_45 (T  (None, 832, 512)             0         ['tf.__operators__.add_44[0][0\n"," FOpLambda)                                                         ]',                           \n","                                                                     'multi_head_attention_21[0][0\n","                                                                    ]']                           \n","                                                                                                  \n"," layer_normalization_45 (La  (None, 832, 512)             1024      ['tf.__operators__.add_45[0][0\n"," yerNormalization)                                                  ]']                           \n","                                                                                                  \n"," dense_44 (Dense)            (None, 832, 2048)            1050624   ['layer_normalization_45[0][0]\n","                                                                    ']                            \n","                                                                                                  \n"," dense_45 (Dense)            (None, 832, 512)             1049088   ['dense_44[0][0]']            \n","                                                                                                  \n"," tf.__operators__.add_46 (T  (None, 832, 512)             0         ['dense_45[0][0]',            \n"," FOpLambda)                                                          'tf.__operators__.add_45[0][0\n","                                                                    ]']                           \n","                                                                                                  \n"," layer_normalization_46 (La  (None, 832, 512)             1024      ['tf.__operators__.add_46[0][0\n"," yerNormalization)                                                  ]']                           \n","                                                                                                  \n"," multi_head_attention_22 (M  (None, 832, 512)             1050624   ['layer_normalization_46[0][0]\n"," ultiHeadAttention)                                                 ',                            \n","                                                                     'layer_normalization_46[0][0]\n","                                                                    ']                            \n","                                                                                                  \n"," tf.__operators__.add_47 (T  (None, 832, 512)             0         ['tf.__operators__.add_46[0][0\n"," FOpLambda)                                                         ]',                           \n","                                                                     'multi_head_attention_22[0][0\n","                                                                    ]']                           \n","                                                                                                  \n"," layer_normalization_47 (La  (None, 832, 512)             1024      ['tf.__operators__.add_47[0][0\n"," yerNormalization)                                                  ]']                           \n","                                                                                                  \n"," dense_46 (Dense)            (None, 832, 2048)            1050624   ['layer_normalization_47[0][0]\n","                                                                    ']                            \n","                                                                                                  \n"," dense_47 (Dense)            (None, 832, 512)             1049088   ['dense_46[0][0]']            \n","                                                                                                  \n"," tf.__operators__.add_48 (T  (None, 832, 512)             0         ['dense_47[0][0]',            \n"," FOpLambda)                                                          'tf.__operators__.add_47[0][0\n","                                                                    ]']                           \n","                                                                                                  \n"," layer_normalization_48 (La  (None, 832, 512)             1024      ['tf.__operators__.add_48[0][0\n"," yerNormalization)                                                  ]']                           \n","                                                                                                  \n"," multi_head_attention_23 (M  (None, 832, 512)             1050624   ['layer_normalization_48[0][0]\n"," ultiHeadAttention)                                                 ',                            \n","                                                                     'layer_normalization_48[0][0]\n","                                                                    ']                            \n","                                                                                                  \n"," tf.__operators__.add_49 (T  (None, 832, 512)             0         ['tf.__operators__.add_48[0][0\n"," FOpLambda)                                                         ]',                           \n","                                                                     'multi_head_attention_23[0][0\n","                                                                    ]']                           \n","                                                                                                  \n"," layer_normalization_49 (La  (None, 832, 512)             1024      ['tf.__operators__.add_49[0][0\n"," yerNormalization)                                                  ]']                           \n","                                                                                                  \n"," dense_48 (Dense)            (None, 832, 2048)            1050624   ['layer_normalization_49[0][0]\n","                                                                    ']                            \n","                                                                                                  \n"," dense_49 (Dense)            (None, 832, 512)             1049088   ['dense_48[0][0]']            \n","                                                                                                  \n"," tf.__operators__.add_50 (T  (None, 832, 512)             0         ['dense_49[0][0]',            \n"," FOpLambda)                                                          'tf.__operators__.add_49[0][0\n","                                                                    ]']                           \n","                                                                                                  \n"," layer_normalization_50 (La  (None, 832, 512)             1024      ['tf.__operators__.add_50[0][0\n"," yerNormalization)                                                  ]']                           \n","                                                                                                  \n"," global_average_pooling1d_2  (None, 512)                  0         ['layer_normalization_50[0][0]\n","  (GlobalAveragePooling1D)                                          ']                            \n","                                                                                                  \n"," dense_50 (Dense)            (None, 4)                    2052      ['global_average_pooling1d_2[0\n","                                                                    ][0]']                        \n","                                                                                                  \n","==================================================================================================\n","Total params: 25734660 (98.17 MB)\n","Trainable params: 25734660 (98.17 MB)\n","Non-trainable params: 0 (0.00 Byte)\n","__________________________________________________________________________________________________\n"]}]},{"cell_type":"code","source":["# Compile the model\n","model.compile(optimizer=tf.keras.optimizers.Adam(learning_rate=1e-4, beta_2=.99, weight_decay=1e0),\n","              loss=tf.keras.losses.SparseCategoricalCrossentropy(from_logits=True),\n","              metrics = [tf.keras.metrics.SparseCategoricalAccuracy()]\n","              )"],"metadata":{"id":"IP1rhBG6DrNR","executionInfo":{"status":"ok","timestamp":1697231086247,"user_tz":240,"elapsed":339,"user":{"displayName":"Mahsa Bakhtiari","userId":"12793844954925855500"}}},"execution_count":16,"outputs":[]},{"cell_type":"code","source":["checkpoint_callback = ModelCheckpoint(\n","    filepath='/content/drive/MyDrive/random bootcamp material/ecg_model_checkpoint/heart_model_checkpoint.h5',  # Specify the file to save the model\n","    save_best_only=False,            # Save only the best model\n","    monitor='val_loss',              # Monitor the validation loss\n","    mode='min',                      # Mode can be 'min' or 'max' depending on the monitored metric\n","    save_weights_only=False,         # Save the entire model, not just weights\n","    save_freq=5                      # Save every 5 epochs\n",")\n","history = model.fit(norm_train_ds, epochs=50, callbacks=checkpoint_callback, validation_data=norm_val_ds) # Train the model"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"Yj5oleK7GjXv","outputId":"fe3ef5c6-d3ef-462d-aefb-7646c5f80fa9","executionInfo":{"status":"ok","timestamp":1697232203408,"user_tz":240,"elapsed":1113415,"user":{"displayName":"Mahsa Bakhtiari","userId":"12793844954925855500"}}},"execution_count":17,"outputs":[{"output_type":"stream","name":"stdout","text":["Epoch 1/50\n"," 4/27 [===>..........................] - ETA: 11s - loss: 6.6364 - sparse_categorical_accuracy: 0.2812"]},{"output_type":"stream","name":"stderr","text":["/usr/local/lib/python3.10/dist-packages/keras/src/engine/training.py:3000: UserWarning: You are saving your model as an HDF5 file via `model.save()`. This file format is considered legacy. We recommend using instead the native Keras format, e.g. `model.save('my_model.keras')`.\n","  saving_api.save_model(\n"]},{"output_type":"stream","name":"stdout","text":["27/27 [==============================] - 50s 916ms/step - loss: 2.5116 - sparse_categorical_accuracy: 0.2895 - val_loss: 1.3927 - val_sparse_categorical_accuracy: 0.2283\n","Epoch 2/50\n","27/27 [==============================] - 21s 737ms/step - loss: 1.4246 - sparse_categorical_accuracy: 0.2656 - val_loss: 1.3810 - val_sparse_categorical_accuracy: 0.2283\n","Epoch 3/50\n","27/27 [==============================] - 22s 738ms/step - loss: 1.4500 - sparse_categorical_accuracy: 0.2787 - val_loss: 1.4535 - val_sparse_categorical_accuracy: 0.2826\n","Epoch 4/50\n","27/27 [==============================] - 21s 728ms/step - loss: 1.4891 - sparse_categorical_accuracy: 0.2249 - val_loss: 1.4483 - val_sparse_categorical_accuracy: 0.3043\n","Epoch 5/50\n","27/27 [==============================] - 23s 831ms/step - loss: 1.4794 - sparse_categorical_accuracy: 0.2883 - val_loss: 1.4032 - val_sparse_categorical_accuracy: 0.2283\n","Epoch 6/50\n","27/27 [==============================] - 21s 735ms/step - loss: 1.4245 - sparse_categorical_accuracy: 0.2799 - val_loss: 1.4223 - val_sparse_categorical_accuracy: 0.2283\n","Epoch 7/50\n","27/27 [==============================] - 22s 773ms/step - loss: 1.4205 - sparse_categorical_accuracy: 0.2787 - val_loss: 1.4916 - val_sparse_categorical_accuracy: 0.1848\n","Epoch 8/50\n","27/27 [==============================] - 22s 733ms/step - loss: 1.4141 - sparse_categorical_accuracy: 0.2847 - val_loss: 1.3995 - val_sparse_categorical_accuracy: 0.2283\n","Epoch 9/50\n","27/27 [==============================] - 21s 735ms/step - loss: 1.4062 - sparse_categorical_accuracy: 0.2560 - val_loss: 1.4018 - val_sparse_categorical_accuracy: 0.3043\n","Epoch 10/50\n","27/27 [==============================] - 23s 806ms/step - loss: 1.4461 - sparse_categorical_accuracy: 0.2632 - val_loss: 1.5885 - val_sparse_categorical_accuracy: 0.2283\n","Epoch 11/50\n","27/27 [==============================] - 21s 730ms/step - loss: 1.5493 - sparse_categorical_accuracy: 0.2464 - val_loss: 1.3978 - val_sparse_categorical_accuracy: 0.2826\n","Epoch 12/50\n","27/27 [==============================] - 21s 738ms/step - loss: 1.4378 - sparse_categorical_accuracy: 0.2644 - val_loss: 1.3956 - val_sparse_categorical_accuracy: 0.3043\n","Epoch 13/50\n","27/27 [==============================] - 23s 777ms/step - loss: 1.4326 - sparse_categorical_accuracy: 0.2751 - val_loss: 1.4066 - val_sparse_categorical_accuracy: 0.2283\n","Epoch 14/50\n","27/27 [==============================] - 21s 731ms/step - loss: 1.4325 - sparse_categorical_accuracy: 0.2823 - val_loss: 1.7094 - val_sparse_categorical_accuracy: 0.3043\n","Epoch 15/50\n","27/27 [==============================] - 23s 827ms/step - loss: 1.4633 - sparse_categorical_accuracy: 0.2990 - val_loss: 1.3854 - val_sparse_categorical_accuracy: 0.2826\n","Epoch 16/50\n","27/27 [==============================] - 21s 732ms/step - loss: 1.4144 - sparse_categorical_accuracy: 0.2679 - val_loss: 1.4606 - val_sparse_categorical_accuracy: 0.2283\n","Epoch 17/50\n","27/27 [==============================] - 22s 783ms/step - loss: 1.4060 - sparse_categorical_accuracy: 0.2488 - val_loss: 1.3896 - val_sparse_categorical_accuracy: 0.3043\n","Epoch 18/50\n","27/27 [==============================] - 22s 737ms/step - loss: 1.4113 - sparse_categorical_accuracy: 0.2907 - val_loss: 1.3709 - val_sparse_categorical_accuracy: 0.3043\n","Epoch 19/50\n","27/27 [==============================] - 21s 735ms/step - loss: 1.3913 - sparse_categorical_accuracy: 0.2859 - val_loss: 1.5028 - val_sparse_categorical_accuracy: 0.3043\n","Epoch 20/50\n","27/27 [==============================] - 22s 799ms/step - loss: 1.4313 - sparse_categorical_accuracy: 0.2691 - val_loss: 1.4917 - val_sparse_categorical_accuracy: 0.2283\n","Epoch 21/50\n","27/27 [==============================] - 21s 741ms/step - loss: 1.4302 - sparse_categorical_accuracy: 0.2811 - val_loss: 1.3803 - val_sparse_categorical_accuracy: 0.2826\n","Epoch 22/50\n","27/27 [==============================] - 21s 742ms/step - loss: 1.4460 - sparse_categorical_accuracy: 0.2691 - val_loss: 1.3873 - val_sparse_categorical_accuracy: 0.3043\n","Epoch 23/50\n","27/27 [==============================] - 23s 764ms/step - loss: 1.4177 - sparse_categorical_accuracy: 0.2644 - val_loss: 1.3890 - val_sparse_categorical_accuracy: 0.3043\n","Epoch 24/50\n","27/27 [==============================] - 22s 773ms/step - loss: 1.3847 - sparse_categorical_accuracy: 0.3050 - val_loss: 1.2529 - val_sparse_categorical_accuracy: 0.4348\n","Epoch 25/50\n","27/27 [==============================] - 22s 795ms/step - loss: 1.2503 - sparse_categorical_accuracy: 0.4522 - val_loss: 1.2923 - val_sparse_categorical_accuracy: 0.4022\n","Epoch 26/50\n","27/27 [==============================] - 21s 737ms/step - loss: 1.2257 - sparse_categorical_accuracy: 0.4629 - val_loss: 1.3924 - val_sparse_categorical_accuracy: 0.3587\n","Epoch 27/50\n","27/27 [==============================] - 22s 766ms/step - loss: 1.2321 - sparse_categorical_accuracy: 0.4593 - val_loss: 1.2629 - val_sparse_categorical_accuracy: 0.4348\n","Epoch 28/50\n","27/27 [==============================] - 22s 737ms/step - loss: 1.1838 - sparse_categorical_accuracy: 0.4629 - val_loss: 1.1869 - val_sparse_categorical_accuracy: 0.4348\n","Epoch 29/50\n","27/27 [==============================] - 21s 738ms/step - loss: 1.1353 - sparse_categorical_accuracy: 0.4785 - val_loss: 1.1716 - val_sparse_categorical_accuracy: 0.4457\n","Epoch 30/50\n","27/27 [==============================] - 23s 807ms/step - loss: 1.0495 - sparse_categorical_accuracy: 0.5203 - val_loss: 1.1119 - val_sparse_categorical_accuracy: 0.5326\n","Epoch 31/50\n","27/27 [==============================] - 21s 735ms/step - loss: 0.9367 - sparse_categorical_accuracy: 0.5574 - val_loss: 0.9276 - val_sparse_categorical_accuracy: 0.5761\n","Epoch 32/50\n","27/27 [==============================] - 21s 741ms/step - loss: 0.7751 - sparse_categorical_accuracy: 0.6711 - val_loss: 0.8584 - val_sparse_categorical_accuracy: 0.6304\n","Epoch 33/50\n","27/27 [==============================] - 23s 740ms/step - loss: 0.7704 - sparse_categorical_accuracy: 0.6579 - val_loss: 0.8159 - val_sparse_categorical_accuracy: 0.5978\n","Epoch 34/50\n","27/27 [==============================] - 22s 778ms/step - loss: 0.6714 - sparse_categorical_accuracy: 0.7237 - val_loss: 1.1576 - val_sparse_categorical_accuracy: 0.5761\n","Epoch 35/50\n","27/27 [==============================] - 23s 806ms/step - loss: 0.7780 - sparse_categorical_accuracy: 0.6758 - val_loss: 0.7382 - val_sparse_categorical_accuracy: 0.6522\n","Epoch 36/50\n","27/27 [==============================] - 21s 737ms/step - loss: 0.6606 - sparse_categorical_accuracy: 0.7273 - val_loss: 0.6749 - val_sparse_categorical_accuracy: 0.7065\n","Epoch 37/50\n","27/27 [==============================] - 21s 752ms/step - loss: 0.5455 - sparse_categorical_accuracy: 0.7727 - val_loss: 0.6080 - val_sparse_categorical_accuracy: 0.7283\n","Epoch 38/50\n","27/27 [==============================] - 22s 733ms/step - loss: 0.4042 - sparse_categorical_accuracy: 0.8409 - val_loss: 0.7190 - val_sparse_categorical_accuracy: 0.7283\n","Epoch 39/50\n","27/27 [==============================] - 21s 737ms/step - loss: 0.3640 - sparse_categorical_accuracy: 0.8565 - val_loss: 0.4996 - val_sparse_categorical_accuracy: 0.7935\n","Epoch 40/50\n","27/27 [==============================] - 23s 820ms/step - loss: 0.2203 - sparse_categorical_accuracy: 0.9211 - val_loss: 0.5179 - val_sparse_categorical_accuracy: 0.8043\n","Epoch 41/50\n","27/27 [==============================] - 21s 740ms/step - loss: 0.1910 - sparse_categorical_accuracy: 0.9318 - val_loss: 0.4431 - val_sparse_categorical_accuracy: 0.8370\n","Epoch 42/50\n","27/27 [==============================] - 21s 743ms/step - loss: 0.1565 - sparse_categorical_accuracy: 0.9390 - val_loss: 0.4099 - val_sparse_categorical_accuracy: 0.8696\n","Epoch 43/50\n","27/27 [==============================] - 22s 735ms/step - loss: 0.1730 - sparse_categorical_accuracy: 0.9342 - val_loss: 0.5161 - val_sparse_categorical_accuracy: 0.8370\n","Epoch 44/50\n","27/27 [==============================] - 21s 754ms/step - loss: 0.1425 - sparse_categorical_accuracy: 0.9629 - val_loss: 0.3652 - val_sparse_categorical_accuracy: 0.8913\n","Epoch 45/50\n","27/27 [==============================] - 22s 796ms/step - loss: 0.0855 - sparse_categorical_accuracy: 0.9689 - val_loss: 0.8477 - val_sparse_categorical_accuracy: 0.8370\n","Epoch 46/50\n","27/27 [==============================] - 22s 768ms/step - loss: 0.1113 - sparse_categorical_accuracy: 0.9653 - val_loss: 0.2614 - val_sparse_categorical_accuracy: 0.9457\n","Epoch 47/50\n","27/27 [==============================] - 21s 746ms/step - loss: 0.0212 - sparse_categorical_accuracy: 0.9940 - val_loss: 0.3691 - val_sparse_categorical_accuracy: 0.9130\n","Epoch 48/50\n","27/27 [==============================] - 22s 733ms/step - loss: 0.0056 - sparse_categorical_accuracy: 1.0000 - val_loss: 0.3835 - val_sparse_categorical_accuracy: 0.9239\n","Epoch 49/50\n","27/27 [==============================] - 22s 767ms/step - loss: 0.0020 - sparse_categorical_accuracy: 1.0000 - val_loss: 0.4495 - val_sparse_categorical_accuracy: 0.9239\n","Epoch 50/50\n","27/27 [==============================] - 22s 797ms/step - loss: 2.5455e-04 - sparse_categorical_accuracy: 1.0000 - val_loss: 0.4722 - val_sparse_categorical_accuracy: 0.9348\n"]}]},{"cell_type":"code","source":["# Evaluate the model on the test set\n","model.evaluate(norm_val_ds, verbose=2)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"ESNDDVaUjvVM","executionInfo":{"status":"ok","timestamp":1697232214076,"user_tz":240,"elapsed":1443,"user":{"displayName":"Mahsa Bakhtiari","userId":"12793844954925855500"}},"outputId":"c6c00cd4-d5a8-45f6-f6a6-e61777d23ff2"},"execution_count":18,"outputs":[{"output_type":"stream","name":"stdout","text":["3/3 - 1s - loss: 0.4722 - sparse_categorical_accuracy: 0.9348 - 785ms/epoch - 262ms/step\n"]},{"output_type":"execute_result","data":{"text/plain":["[0.47217148542404175, 0.9347826242446899]"]},"metadata":{},"execution_count":18}]},{"cell_type":"code","source":["# Evaluate the model on train set\n","model_loss, model_accuracy = model.evaluate(norm_train_ds, verbose= 2)"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"oQJG3Sm5k2gJ","executionInfo":{"status":"ok","timestamp":1697232224698,"user_tz":240,"elapsed":6099,"user":{"displayName":"Mahsa Bakhtiari","userId":"12793844954925855500"}},"outputId":"0126e7dd-16e9-4c49-ec94-a3f382cb9fa0"},"execution_count":19,"outputs":[{"output_type":"stream","name":"stdout","text":["27/27 - 6s - loss: 1.8157e-04 - sparse_categorical_accuracy: 1.0000 - 6s/epoch - 235ms/step\n"]}]},{"cell_type":"code","source":["# Plot training history (loss and accuracy)\n","import matplotlib.pyplot as plt\n","\n","plt.figure(figsize=(10, 4))\n","\n","# Plot training & validation loss values\n","plt.subplot(121)\n","plt.plot(history.history['loss'], label='Train')\n","plt.plot(history.history['val_loss'], label='Validation')\n","plt.title('Model Loss')\n","plt.ylabel('Loss')\n","plt.xlabel('Epoch')\n","plt.legend()\n","\n","# Plot training & validation accuracy values\n","plt.subplot(122)\n","plt.plot(history.history['sparse_categorical_accuracy'], label='Train')\n","plt.plot(history.history['val_sparse_categorical_accuracy'], label='Validation')\n","plt.title('Model Accuracy')\n","plt.ylabel('Accuracy')\n","plt.xlabel('Epoch')\n","plt.legend()\n","\n","plt.tight_layout()\n","plt.show()"],"metadata":{"id":"M0Tu8S7XtSNI","colab":{"base_uri":"https://localhost:8080/","height":407},"executionInfo":{"status":"ok","timestamp":1697232283552,"user_tz":240,"elapsed":479,"user":{"displayName":"Mahsa Bakhtiari","userId":"12793844954925855500"}},"outputId":"801b2123-424a-4ed7-d7b1-7e7801c306f0"},"execution_count":21,"outputs":[{"output_type":"display_data","data":{"text/plain":["<Figure size 1000x400 with 2 Axes>"],"image/png":"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\n"},"metadata":{}}]},{"cell_type":"code","source":["# Export the model\n","model.export(\"/content/drive/MyDrive/ML /ecg_image_classification/vit_ecg_model.hdf5\")\n"],"metadata":{"colab":{"base_uri":"https://localhost:8080/"},"id":"8oxDdnxjKwTI","executionInfo":{"status":"ok","timestamp":1697071060945,"user_tz":240,"elapsed":4571,"user":{"displayName":"Mahsa Bakhtiari","userId":"12793844954925855500"}},"outputId":"0a195c95-a711-454d-ae1b-fac55dec5bc8"},"execution_count":null,"outputs":[{"output_type":"stream","name":"stdout","text":["Saved artifact at '/content/drive/MyDrive/ML /ecg_image_classification/vit_ecg_model.hdf5'. The following endpoints are available:\n","\n","* Endpoint 'serve'\n","  Args:\n","    args_0: float32 Tensor, shape=(None, 182, 256, 3)\n","  Returns:\n","    float32 Tensor, shape=(None, 4)\n"]}]}]}