my-own-facematch-liveness.ipynb

{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.7.10","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"# This Python 3 environment comes with many helpful analytics libraries installed\n# It is defined by the kaggle/python Docker image: https://github.com/kaggle/docker-python\n# For example, here's several helpful packages to load\n\nimport numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\nimport os\nimport tensorflow as tf\nimport cv2 as cv \nfrom kaggle_datasets import KaggleDatasets\n\n# Input data files are available in the read-only \"../input/\" directory\n# For example, running this (by clicking run or pressing Shift+Enter) will list all files under the input directory\n\n# You can write up to 20GB to the current directory (/kaggle/working/) that gets preserved as output when you create a version using \"Save & Run All\" \n# You can also write temporary files to /kaggle/temp/, but they won't be saved outside of the current session","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2022-11-25T20:22:34.647285Z","iopub.execute_input":"2022-11-25T20:22:34.647596Z","iopub.status.idle":"2022-11-25T20:22:41.630462Z","shell.execute_reply.started":"2022-11-25T20:22:34.647562Z","shell.execute_reply":"2022-11-25T20:22:41.629471Z"},"trusted":true},"execution_count":2,"outputs":[{"name":"stderr","text":"2022-11-25 20:22:35.544268: W tensorflow/stream_executor/platform/default/dso_loader.cc:60] Could not load dynamic library 'libcudart.so.11.0'; dlerror: libcudart.so.11.0: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /opt/conda/lib\n2022-11-25 20:22:35.544413: I tensorflow/stream_executor/cuda/cudart_stub.cc:29] Ignore above cudart dlerror if you do not have a GPU set up on your machine.\n","output_type":"stream"}]},{"cell_type":"code","source":"# Detect hardware, return appropriate distribution strategy\ntry:\n    tpu = tf.distribute.cluster_resolver.TPUClusterResolver()  # TPU detection. No parameters necessary if TPU_NAME environment variable is set. On Kaggle this is always the case.\n    print('Running on TPU ', tpu.master())\nexcept ValueError:\n    tpu = None\n\nif tpu:\n    tf.config.experimental_connect_to_cluster(tpu)\n    tf.tpu.experimental.initialize_tpu_system(tpu)\n    strategy = tf.distribute.experimental.TPUStrategy(tpu)\nelse:\n    strategy = tf.distribute.get_strategy() # default distribution strategy in Tensorflow. Works on CPU and single GPU.\n\nprint(\"REPLICAS: \", strategy.num_replicas_in_sync)","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:41.632278Z","iopub.execute_input":"2022-11-25T20:22:41.632618Z","iopub.status.idle":"2022-11-25T20:22:48.186271Z","shell.execute_reply.started":"2022-11-25T20:22:41.632586Z","shell.execute_reply":"2022-11-25T20:22:48.185312Z"},"trusted":true},"execution_count":3,"outputs":[{"name":"stdout","text":"Running on TPU  grpc://10.0.0.2:8470\n","output_type":"stream"},{"name":"stderr","text":"2022-11-25 20:22:41.641166: I tensorflow/compiler/jit/xla_cpu_device.cc:41] Not creating XLA devices, tf_xla_enable_xla_devices not set\n2022-11-25 20:22:41.644297: W tensorflow/stream_executor/platform/default/dso_loader.cc:60] Could not load dynamic library 'libcuda.so.1'; dlerror: libcuda.so.1: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /opt/conda/lib\n2022-11-25 20:22:41.644336: W tensorflow/stream_executor/cuda/cuda_driver.cc:326] failed call to cuInit: UNKNOWN ERROR (303)\n2022-11-25 20:22:41.644363: I tensorflow/stream_executor/cuda/cuda_diagnostics.cc:156] kernel driver does not appear to be running on this host (e085986736f2): /proc/driver/nvidia/version does not exist\n2022-11-25 20:22:41.647930: I tensorflow/core/platform/cpu_feature_guard.cc:142] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN) to use the following CPU instructions in performance-critical operations:  AVX2 FMA\nTo enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.\n2022-11-25 20:22:41.649640: I tensorflow/compiler/jit/xla_gpu_device.cc:99] Not creating XLA devices, tf_xla_enable_xla_devices not set\n2022-11-25 20:22:41.655916: I tensorflow/compiler/jit/xla_gpu_device.cc:99] Not creating XLA devices, tf_xla_enable_xla_devices not set\n2022-11-25 20:22:41.688824: I tensorflow/core/distributed_runtime/rpc/grpc_channel.cc:301] Initialize GrpcChannelCache for job worker -> {0 -> 10.0.0.2:8470}\n2022-11-25 20:22:41.688887: I tensorflow/core/distributed_runtime/rpc/grpc_channel.cc:301] Initialize GrpcChannelCache for job localhost -> {0 -> localhost:30021}\n2022-11-25 20:22:41.708930: I tensorflow/core/distributed_runtime/rpc/grpc_channel.cc:301] Initialize GrpcChannelCache for job worker -> {0 -> 10.0.0.2:8470}\n2022-11-25 20:22:41.708985: I tensorflow/core/distributed_runtime/rpc/grpc_channel.cc:301] Initialize GrpcChannelCache for job localhost -> {0 -> localhost:30021}\n2022-11-25 20:22:41.710373: I tensorflow/core/distributed_runtime/rpc/grpc_server_lib.cc:411] Started server with target: grpc://localhost:30021\n","output_type":"stream"},{"name":"stdout","text":"REPLICAS:  8\n","output_type":"stream"}]},{"cell_type":"code","source":"GCS_DS_PATH = KaggleDatasets().get_gcs_path()","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:48.187378Z","iopub.execute_input":"2022-11-25T20:22:48.187700Z","iopub.status.idle":"2022-11-25T20:22:48.551366Z","shell.execute_reply.started":"2022-11-25T20:22:48.187659Z","shell.execute_reply":"2022-11-25T20:22:48.550428Z"},"trusted":true},"execution_count":4,"outputs":[]},{"cell_type":"code","source":"IMAGE_SIZE = [192, 192] # at this size, a GPU will run out of memory. Use the TPU\nEPOCHS = 5\nBATCH_SIZE = 16 * strategy.num_replicas_in_sync\n\nNUM_TRAINING_IMAGES = 12753\nNUM_TEST_IMAGES = 7382\nSTEPS_PER_EPOCH = NUM_TRAINING_IMAGES // BATCH_SIZE\nAUTO = tf.data.experimental.AUTOTUNE","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:48.552354Z","iopub.execute_input":"2022-11-25T20:22:48.552595Z","iopub.status.idle":"2022-11-25T20:22:48.558729Z","shell.execute_reply.started":"2022-11-25T20:22:48.552569Z","shell.execute_reply":"2022-11-25T20:22:48.557640Z"},"trusted":true},"execution_count":5,"outputs":[]},{"cell_type":"code","source":"import matplotlib.pyplot as plt\nimport matplotlib.image as mpimg\n\nwith open('../input/selfies-id-images-dataset/11_sets_Hispanics/0001cc1e1e--61af86f374501f072edf6d2c_age_22_name_Juliana/Selfie_9.jpg', \"rb\") as local_file: # <= change here\n    img = local_file.read()\nimg = tf.io.decode_jpeg(img, channels=3)\nimgplot = plt.imshow(img)","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:48.561952Z","iopub.execute_input":"2022-11-25T20:22:48.562321Z","iopub.status.idle":"2022-11-25T20:22:50.051203Z","shell.execute_reply.started":"2022-11-25T20:22:48.562279Z","shell.execute_reply":"2022-11-25T20:22:50.050507Z"},"trusted":true},"execution_count":6,"outputs":[{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 1 Axes>","image/png":"iVBORw0KGgoAAAANSUhEUgAAANIAAAD8CAYAAAACP/oEAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAAsTAAALEwEAmpwYAADg6UlEQVR4nOz9WaxtWZaeh32zWc1uzjn33ogbERnZV8dSVZFVJsssGhBgGhYlWm4IwbBE6kGyIViGIcIvfjDdADJkC+CDGxiwIZi2CUsPEq0HCy7btClKFm1CEKEiKZdYxeqzMqsiIiNue9q9VzPnHH4Yc6619j773puVrErfEGJG3jzn7GY1c80xxxj/GOMfRkT4Ynwxvhj/YMP+//sCvhhfjP8kjC8E6Yvxxfh9GF8I0hfji/H7ML4QpC/GF+P3YXwhSF+ML8bvw/hCkL4YX4zfh/EDFyRjzJ82xvyaMeY3jTF/4Qd9/i/GF+MPYpgfZBzJGOOAXwf+FPAR8AvAnxORv/8Du4gvxhfjD2D8oDXSHwd+U0S+JSID8FeAP/MDvoYvxhfj9334H/D5vgz87uLvj4CfW37AGPPPA/88gPf2jz14tDpxGAOACKQojEMiRjl411UGgyHl1wUh/y9/+3BYa7HOkmIixogIGGMQkelfOScGyO+Xgy0/U4ZMXzi4bMzRFZTP6fGmA5JkeU/zd/QjZvqZj0BKgkiaTmnM/O3ymeO7N6bM5eH1G2Omf8uvSZJpLqfPWbu4keXBuTfKlUi+1uU1HP5uEETv2+RnkWQ6wP3vmsXUlflIB/f9qnstxzm+f8nPZHG8ZyLy+P5d/eAF6Y1DRP4S8JcA3nv/TP6Jf+qPAPmGEBKQouHqReDpx3u6u0AIwpCEGCPWwLqpWLX1kWAUgQBjZkWcRB/WxdmWuqnp+44Uha4bqJuaZy9fMg4jYECEkISYUr5Yi3V6vGEYCCFg86KKSc+LCFh9eN47rHXTdVhr9e8khBCISb9vjCGGyDgGjLH3Flr5jHMO51y+H2EcR/q+n+63DGvt9Ld+1kzHmuchISJYY3HO0bYt1lq9BwBriDGSUiKlpIKQkl6Pc1mYDHVVT9d49Fync6YUGYeBcRymxar3oYs9poT1HmssRsB6R103pBjBQAwRSQEwpHybKSWMAUkRYwzjOE7X7pzDe5/PnRjGAUmSBU2vrcwpCCnJNDflmkSEq6ur77xq3f6gBelj4KuLv7+SXzs5nKlp3Joh7QEhxUQCPv1o5NkngTgEkggb79k6x0uzZ11XrJoK8g4mcjhhy21SRHAWGmshBoa7kaHvGcZAP4zcXidC1k4gJEkYHC5PekqS/w3Tw3BeF7aJEe/cdC7jzLSgQwiUB7Zab/B1zfMnn+oClYQxlpQEa9107fP1l0Vj8ucF7/3BQl1+vvzt3OGxykIvr1lrsWYWuBBGvK8wxhDGkQRISmDzzm6zUAqkGJCg5+5TxHlPU7fTcY0x06JOeUOLcZyupQho+WyIEUKgrmskJWy0OGup6waMIYY9ImplpJj0tRhVG6dIjHESzqqqMLZsTOM0Z5OFgS6JKAkrh1r494If/KAF6ReAHzXGfBMVoD8L/NOv+rA1lkfNV3jafYeQejCWyyeRJx+PpKiKP2UtUVWJB+0KawyTGhdmQTL64Iup4IzBYbCAk0Ta77Ei1CJEiVgiURLeCGIMQ1STyQAha5uyU8UYqaqKcRwJXaCqqulhlUXunVMNmRIpWcYxEGOi7zuqtsX7mhCElKKapCcEqPxeBGS5KJY7+vIzy531UCskXBZ0EYEEkZjfV43s/YgxTjeAlBBEF5s1JFAtlIQQ5mtNMRJjAIGqrrHY+ZryRhGC3vvSfDy4ViClSN/tJxN6HEfqpma9XoMkUooqcIJumggxBlLWxsaYae7DEEgp5Q1sMafGZFM4C5U93IREBHdk8r1q/EAFSUSCMebPA38NcMBfFpFfft13rFS07pydPCdE4cmne1JU4YgxYi2IF1yjtyJJTamymL14vPfEGHDZBJIkeASPwQONFbBCTIIFWu8YHNwNI2OApBfPkARnVZOEpNqxCGoxJUIM00NUE0c1VwyJ7YN3uLm6JMYBMEhK7O5uGfqOulnRrFZcvngK1mKtPdA8cNpuL4tjKTjHvt38PZ3TctyiCXQe7XQOY7Lmk4QFQkjzLm2Keax/28W1Lhdc1+/z5gVJLCmFrDGEpTN1bLaWTUES2fyK2cQMdPtIGEd9nlkoRVWT3keIy7WmG9tifsqzKpqvaMfldYhKrv6LugGn9JYJEoCI/FXgr35vnzaEFGmrFbejIYbE2KdpkkBYtw3rTYMRM2mestu1basaIEbSOCIxYkQQImsDW5ftY3RBjEm1W4hqiknt6EwkJZBksp8reGsYUQETZmEyBirns+9i83WGrLUSl8+eM4wDwzDMPoRAGAPWjaQYcdYfaKMylibYsVNchME5Ny2Y02P+bhGmSeAyiKCCpNpbEsjC1SkOv7qYZvKFnHNYa4gxa0jKQh6QFHHW3dsQlj/1d52MAuBUxpNMQvI8lTnu+55hGKgqNd/1mIYUZ+283BTKd6d/LMGdGcQpFow+5Nnsiym9EqBajrcObFgOYyBGwbDixWeR1Vk25YLupOt1w2rdAExqfBhHrIWhH6iczdBewho1STbO4ZPBkWi9UQeWDDqYeZGRBAd4A9FC7S1hjDqpDkavjm4IQoRsk2enOS/2YRwnxzXGxO3t7Xx8ObzPcRjyAtZjHAMGZdEvhWxpWgL3dteYzcT86elnOU/5TjELi1YpYIwIWfscIV0JrNMPzNdjDj5jRH2qhArf0kc71gTWWjCJlALG+AOEsmg8UF+q+DkhBDVNM4J3CnVczomIoqDCrKUR1XqTQJeb5dC/fJMQwVsuSIqUgTcVl88Cu52uvyEONHXFelVj0YXXtCsQwTqrO7wxpBCpvGe1WnG2XrO7fIEPHZVTH0kNPR2qmQRvYcy7rX4iYQ04K/j8QENKVNZQ9qqYdZXzFaCoUsooosiMRJXzHP/U3+M9n+bYD1qabyfh9vx3QdvMgfDMxy7aqCzi5feOz6XXkrDWH5h0xccy2dQD8N5N2lEvKJ/LGmxWbc454jhO9zCjZSYv8jTNWRlFAMtxy7yc2kiOzd+lGaw/9V+ZObuYTz3Z/B5Ghd6aN4vSWy1I1jqqqmYcOs7Pz/jt3/mUEFVINus2Q7Wepm7wzmPI5q2AN5ZH51u27QrjPetVhdldg6hvkiTh7exsl6myBrwzxORIErEYktHXbdb+1hictSpmMi/YGOZFr6BCebjpYLEW4Tg20w7QpBPvLz93/J1TC08/Z/I5BWsV/CifKf7HqWuaNWLROAFr/cFOrxqkLFiLyBzfm7SnNTgcCQVcjDFYY4giGDsLs2pud/D943s60HhHfuByPo6RSxEhymyiLe/hvqAt5jn/X5J0cMxT461OWjXGsD27IMTEew8f4J1nHALr1ZrKV1hjqapabfD88MdxZNW2vPPgAe8+eMCjsy11P9I9uyTs91iMahhjcSVOgUHxPBUJZ7Odbh3ezfu6zf+cc6rRjNFYRzGlYmQMgTHG7D+92qleaoNXoWslZnP8vVPvT/GdE5pq+ZnifM/mUiBmgAQgpEhCst/D9NkYE+M4EOOISEQkTudbntvkOJSxZrr78n6QxBgDISWNDTUNrvI477HeT7Eou/CLjv+V15fHPfa5lvO0nJPiE53aiE6Zhbo29CdvABzeao0kInjrWbUb9t3Il99/zHft7IdU3lE5jyXHCUKgchUPtmd85f332L28or+74fbqmpQiosY+VkQFJwsR6KJ3xpCMQUKChGq7SiBGJCaqJIxpgY4JyMIFSUI24bKfcbR7Ho9jlK3c8/F3lq/VdZ3h40OEamnmTH7HiXEcSxKKgCWdnzzi5EfIwWI9yBZYmE+KhhmMcVjnSKKIW4pq2okknAjGWozRzQijwiYmC3YGLsZxhOzPGLi30Fm8tjRTl/M2CVvRRNaSxoDABPsvn8PymAdzajLM/wbg7q0XJFJi1a4Rc433nrPNhn4ccNbSVDWGoglGjPFcnG3ZtCtC33Pz4op+GNgPA94q2oY1+IxeqZOrDqfNO5XDaKxEr2BCeSyWJNlGz/9fzEj1kHTRubw4rbUko6CCcwbBzulKr1gUy7+PneWyQIZhmP5+lbDp34m8n77iPSFGvU5nLWIUyImTZjG6D4joPVrNAFEtOqNcxewpJqLOqWp0Y41qsRCIRqF6Z70imjk2Va4rSQFl4hQkLZlIwukN5pVrZvkz+zxFP54CcO5pI7N8+qf90ePxVgtSQSKNMazWa+rmlrr2DGHEWg0UhhQZcvygdo7Nao1D2F1eset79sPIEAKrylHnZyypPCFNiYmoP6D+kAqIIwfr8r5YzDSLBiOtsQSV82wYZmE0Fu89NiOGxcE1xhIU35u0BsyOM8wCNJt8hyZh+cwpwTv+TBH0co7yenHuJz9GEiSNd1nrNBicd/oS3IYZHseUbIXD+Mps2iXSdMmSkbU05daNYSBYgxsDzmuKk/N+Ai+WaOUpoVnO1al5WAImx6jbseY6jn0VgTV22iqXs8rrxlstSJADbZJom5qmwNliqCtPxND1ETGCdx7vK5yviP2O3e0NXYiMad5VjDHoU1YhQcDnNBwRISGIGCQpIheTCorG6GxeJHos73z2wyNY3bWtMzhXUVWVmjUpZiccrIlYazhGpJbo3PGYdn7uO8bL1/T7ZcfN3tmM5B6gXAowLEzK/NkYEwVY0FQnP/tU80okxoRNanIvF9eBL0N2KYqTfrTjl8yNNCSCCVinQlXOrec0BXWYNFf5/tKcLOdeXsMUcD16/9S1zgeehWjSVDlpViRlR+nV460WpLJDxJQ0L23VEseEFUX0xiESRk3PQQyVdUgY6Xc917uRIWiaj8lQtTMGK1G1jbGQodkkmq+li9bivCUmiN6j4RIL1uHEkKwHY0koDL4ah3x9qhWtsdkpD/RdP6WyiAg+a4IYQ87Rm6Pzr/OjTr1/aNadNvN0l/eLnViyJkmLoKU9WCQT2pa1qPc+BybjLIwGhhRyWGAZRJgOAilnr2fNaqcFOsPv5XyaEWJnEMaU2JFMC+F48R9rkzIny/s//v34c+X9cl33vp99OoCmbtjT3TtWGW+1ICEZthSQKHhfqQGVM5j7fpg/moS28qRxoB8H+qCZwuMYSQbq2tE6B0F35BR1p62cZ3BC9DWjdYzWE1zNiCNZp86p6KLwgCQmJ9sloZbVPVMC9KHEpAsqxMA4Rrp+R9/tubu7ZegHJCX6Qc2hV2klPRaUWNCxMM12vr6mUPJy4SkMr5/TLABYLObsSxUFoPdQSjGyZrNzpvkcs0mqsWMROrK5bWYhSNl0RDeo/KQWj3fhCyZBw7cQssl9PJ/l5ynhWGrnw7m7nzK1jEGV75ZLKyb8HNIIWAvd/tVCBG+7IAGSoj6QGNW3cSBBGIdAGEess1PSaB80MW4/RoaYtEQBqw9pMnVy6QJCMI6+veCucfQ4ojHqdKekSt1YnDMabzImq303WyrZK06S05BSytomErNpICbhrAOXcE3F9uKCdyTngY0jdzd37Hc3DH1P13X3hElIaNDTwJHmKVpG/2kgk+NvpzjB2MArTMhEyWYwUy7dvPAl24nGmCkNSb+nwqJaJifcLoCHUz7eMUI2gwIpgxQKYsQc4yoLfz7mYaB6iXwuz3GMwC3Pebxp3QdjZIKZND/wcx6QFci5byHHO3IOQRLCGIjF1LAaTY8xcjcKwxgZY4So8GneM0nGIs4xths6UzHYisHMAcoCL3tfqyNstY6lwMSgC8yaXAdTnHUBMaX+SB9EiIkxRmJJaE3jIsUlYqzDVzV1u+JhekSMI91+z+72hpvrG4ZhyCk+ZYcETFm0s2mWUlmQ9xGpsvsvNdFylLUkCJaiaZi+M5uVuhnJkZ8wo56HgeilrzI9yyOTbKmNJg1xFBNaHuM4nlaOeWyOHQMVSwEr0FG58aVQlQ3gOEFYLYECl8+lH8fj7RYk0by6cRxzPEConGcX9gwMMxybEslooC/2gTBqtmWIIaexAL5mv37AkKA3PsOtFme1ZqWua6qmoakbnPcan8oLZQ7uRdU+qezAMqFRUWJGvfS6K4FmYdsHGbPGSpO/MY5jLgjUgjzvKzbrDQ/fecy+23NzdcntzRXjODAVh77Wn5rnbTGLJ02k/Ff+X0b4pl38EOUDWCq7JTq2PG7RGiWr/PhZHv+9jAEda6qi/ZaaZ/n+7Mu9+hzluqbvG52kEpQt97YMAC8FmVx2Y8xicl8x3mpBSimx7/YZAUua41b5CYAoWdYhhHnHk6CpPdmp9t7RrtbY7Tk3piaalFG/irZdsV5vaJuWVdvSNBU+C5AxczZzEhWAEIP+nsEPzafLPohoXlYSweTFaIt5lwQnFckmxE5GA7HV90MIjGGk7zuGbk8/DNjK07YrHjx6h9ubG65evmAY+jljB33w3vuDUoF5FAdaN2BrD82XEg+aomAWkIL+zX7SZCIZFlp5IciLUxYhXNY5HQvBUnCONdhyFBBivtbTIMOp7x6bkxgzpQd55zSEUc6bP5+yCeuW1cSYnBf2esQOPg+CtO8UhjSGMUR8Tp8fYqDNGgMKUKRmoLEgxlCvV2wfPMLXtZp2CZqqYb1ac7a94Hy7YbPZaJ6eY7L/57yzlNNQ1D+iIF4mEW0ubbcGSW6CT1NSjYPMWc7J5ETMZMAuYy5CRKiqCu8dTd0QVhuGYaDvA91+x253w7mv2J5fcHt9w8sXTxkzyFIe+NIk0dfmxaDCdKhlQJFD/Vx2yMtPbM5EKNkK2bSSpDt5gYmZIkw5Jy0vNiP3QLwk9zMUVMA1rnYsDK/ycco4EE6Kqp6/e/yziJW1Nl/eHBMjf7UUfE6a1y7n65hl4/54qwWpH0c+/uwpVeVp6ppuGEAMTVOzHwai6EMrO9wYtCTaWsPq4pxqtcHkxNS68qzXG863Fzx8cMFms2HVNBq/SIaYwiQoxlj1T/KWnkSIIRLCSEyJEGL229LsIwDHfoiIwqc+gxsp6QJ3ouiXEY3JJCMEVGtUlcc7R1Ul6rqmbVfs9nfc7XZsLyzNasXt1UuuXr6cTKhDZ/rUTM7aaV6gS0Gb4XGRBHF26CfzyZjJX5vOxZFvYk12SI/y48xit9NfKMvZ4CYz61VxHz39q9G45bQfCNARMHNcgCiQNw2ZBM4sCirL905dz/F4uwWpH/n13/4IawyVc1SN52y7pmobzO0dIQRC5QkxYgXMEGjbFc7X+GatEyBC26x49OARj999zMX5GW3b4CuF0FNKiFG4tUT0+2FP1/f0Q2Dfdey7PWEc6YeOYRyJIU1gRwEbiuljTckSUJi+8pWSb9QqUM57nFViFof6AHpeNc9SMiQSVQXONnivhYKVb+m6PXfccPbgEU274sWzp3Rdl6tOi47QcWoB6IKZPpE/B4rulUWsf5exLKk/PM6J+Nac/nc4DjTRfJ3Lff4YGCivnfq9HOcArMip+cf+VvnsMuMBSp6hTJZyCRSX38sGoCGCz7kgiQh3d3uqyhG9Z9/vub6+4eLiAmsN+32HrxSOjjFgvaPdXMypHsaw3Zzx/nsf8P7jdznfbql8TpZc7M7D0LPfd9zu7rjb3XK7u+Vud8d+t2cYRsYQGMKY2Wsk+2hapyIiStIhMhWdmazFCkmKtXaKxTR1RV3VNG1LXTVUVY2t/FRcV/ydlMDYXHrQrjRrI3/mdndDiJGH7z7m5uqK2+vrnJlQtMphqpHOZVkgLD4zm3wzfH7f1HrdjnwoIIevn0LX5vEqAOT+MV8XOwJyriAHKGEZc0rUoTAv0boZgNA5nGNbCWPdFJR93XirBamYFuOYiCHQNDUxRp4/f6HZDCT6flBTbrXh4sEjUnaQrbM8OH/AVz/8kA8eP2azWkNOqIwxMA49+67j6uaay6tLrq6vubq+ZbffMQaN8ZTFuNqs8VVF3w0Lmqf5wfZ9z2q1Yr1es9vt6PseZK7psdZS1zV3t3fcLGIidVUp6LFuqZqW1XqLr2aBEYkqrHmR1HUFrMmBFvbjyHq9xhjD9eXVgfZYmkvzgl6Yd4jGYqf7mM3C1wnO0mQ6lde23PHL38fa4Ph4p8YpAdXnsfBpmLXNcQ5euYZTmqh8bi6cyZMzoSlqWWSutTI7r5wTeMsFyWQzyXvP+dmWs/M1Nze3XL68IoSY5cKwXm84O7vAWp+5CyzvPHzE17/8ZR6/8w6b1UpNN2PY9x0315e8uHzJsxfPeXl9zdX1jaYbhZQzvHWhWaMAxPXlFc55qqrm7OwMYwxDP6jJF5Sh5vr6Gucc2+2WpmnY7XaklBiGAYmJQYTNZsPd3d2USxZD0O+9dBhnqeqa7dmGdrVitdrgfZsf5PwYnTPqR+X4Vt/3pJSom4Yhc9qdivDDkqehxICWi/vVGuJVz6b8fJMwLIXiWDud1lanzzedSxImAzjTvSyOdxwXOiXorzq/ASRqTmUZzjmseb1WeqsFSSFQ5YC7vLyk6/Y4r456CEqe2DQNDx48oqqrKUn0wYVqonffeUezwa2lHweu7+548uwJL54/47Pnz7m+viHEpHGjpmazrbBWVb3GeHpCGKcKztubW64ur/Des91s2KzX3N7dTQ9uHEdevHhB27acn59zeXlJGEcQMhI3sFqvpkznruty/Mgjg97j86cKzddtQ7s9x1u9326/p+v27Hd7xnGg73vGoc9l7WnSMsDBolku+EMNNWeHL2acY8DkTc/nVYLwKq32vQjNq753bKYpzG7yWjhdk7S8xjeZoceCfqzdXzfeakECaJqGpq5YtQ3vPX7Eg4sz/tbf/sXMQgpnZxdUVaVl3gbOz7Z89Usf8u7DRzRNm32KOz59+kT/PXnK9fV1ZiRNU6bz0PfcXF9rmbr3VG1N0zSsVi19P1JVNXc3N+x3Cjzc3t4y5oDvdrtlt9sholC2tZZPPvmEpq4x1jL2SiAZ+56u23F2ds7d3S2317c5PqYZGzHEe8HMNFFY5b/TzF9gQP09WRgeC7j3VabUMUQ8v1diS8XfOvSzyvlPQdWnxjEgcSzgS7Pv1EI/XtyF4KYgq4W19nXa5vgaX2UyLoVpee43CVAZb7cgGXhwccaD83PqyvGl999lHIccoDY0TcN2u80PN1I3K778/vs8fucRdVPT9Tsu+z1Pnz3jo0++y9NnLxXlkoRzFXWdy61DUHbQpqHvuqw9em5JippVFRbYbrfUTUOMkd1ux+3tLSlFXrx4zoPzC7brDXe7Hc9fPJ8K8KyxylSaM8RFIk+fPFWTfPFwRWaWGyOnkbHlgjhe4NZoMNUeLdZTjvbhTxWgQo08k4/MrD/H3zkFU58SqFNCe7zjl/E6v6z4Os65SfscX8exQBXteywIS8E9vsZDbr/7OXyvG2+1IFljePToEePQc3Wz467ruLq6ygiV8OjRI81IzmbeowcPeeedd6nblkTk5dUzPv7kEz59+pzLlzeIoLl0ld62sxbrHN45+qFnHCMPzi8IITIMHcPYk2IiDAPXfU8IQXnVxsBut9P3k+7e15fXKIwsRxpF7sVPFHY/DDgaYyYhKuN4Fz0uPyjj1I577FwvX1uCD6VAcl4wc+HfcUn20t84vvZjof1eEbdX7fgHqBpzsHx5niW4cqzRjDFsNhtNx4r365mW87A813LTWSbLvmm81YIkInznd39n4lXQsojAGCIYy3q9mXbypva8++gh21VLSgNXVy/57qef8fF3P+Plyyv6PtDkAGzX9+x3u0wqWCr3wBrHzbUCC9bbaZdXnoTIy5eXvLy8zBkLuYAuFQYie28R50ueAQORSfCWGgny7nrCGV+O40V5vJjLZ8rCOPX95WdOmS7LRXbM1fAmU/HUcY6v/U2mYPnuMsu8vHaM/J0CLcq/d999l9VqxbNnzxjicHCM4/tc3u+UYW7tZPnA/U3leLz1gtR1nSanmsI/oPj+er3G5zw7gO12w4OLNWIiLy9f8p2PPuJ3P/ou19d3hKA7VwgjIQZFx7ZnjMOAclGPGaqO3N7d5fw5fd0aS9PU1E1DiKOm/xSPZBKi+5oAOFjQy8rVY6d++s7RwnmTWXFs1x+//jowoHzuFLnk8vdjosoyjqHvpaCUey7FfKBJvsem6XKelubVKUKT4wyOU9ey/Hu/33Nzc4MxM+HmK004a/O1aulH6a6hLFH53x8k2GCM+TZwg8a0g4j8rDHmEfB/Ar4BfBv4J0XkpdG7+F8B/ziwA/7rIvJ333SO4lQno2XcxujuWNf1VHlqDGw3a9qmYre75qNPPuajjz/j6uoWa63m03mHoPzQ3dBzfXWlGiIG9vvdglo4Z3gDiMYafvLLP8of+yM/wb/zN/8Drq6uKbwGi3lgzhQw92z3IkSqAaDAtoffvze39147pYEW7y5+3vdBXieUp7TH8e8HGoWca2dKwLloN00i1k0rMo59vm8t09fT3EfHlucqWvLYXFzO6+vmq3yn6zqMMZydnTGGERGmMnKFs81BoHXZ9uXYh9O0uz/4ONJ/TkSeLf7+C8C/KyJ/0WiP2L8A/PeA/wLwo/nfzwH/CkdNxk4NfchaqxJDr2kdeWKHQUu6vVcuhyfPnvLk6VO+89F3efHiaprMEELOsFYi9yIkhTnopCOeFAUUhO989DF/4md+ks26mZIvOfETTtWzzFkFBW063uCXO+7xTvyqRV4W3Pz+/fSf42MsTadXLeIp3GlzuNIoTO69o6orLTNxnqqqqCpP5R2+3Kcx2rwAuOs6rq9vePbs2SRAp/yq42s55fstNcmbNGz5KSL4qiKJxtiO5/XU/B7Ptchstr8pLPAHYdr9GeBP5t//VeBvoIL0Z4B/TfRq/5Yx5oEx5ksi8t1XHUgo2dY256ZpPAef6PqO+HLUUnKJfPfTzwCmliEinF7UQELz43JCyD0n/jht/+rqlm/97scYmQv5ymfLDs1ksh3GH+ZRTIfTUK0thIr55eOHvjzusu6nABjaU6nc74GVeLhb25nU8jj7oG0bzs/P2J5tpo4Pm/V6gYA5uv2ey6tLuq6n6/eavDuEKc/QGO1vFFOk8hVVVdP3w8nN4/j+XoUsLsGQYxP0WBgOvqs72MHxDze4+VnfA1KMhhHEKJJp/4A1kgD/ttGkrf+taLe99xfC8Snwfv79VNvLLwOvFKQpTpKTKkOK2iFBhJub23nHWKjk40V0/6CLnUWYOtXpW/d5Ecil6i9eXinh/kKIlgJXzp2SOfmwyx0dL3C9XtUmajKmSahU+5G12Kt26nmxOTf7kYW4pGlqVqsWYyxVVeFyt4xSWNh1Hfv9ftL8u92e/b4DJM+N5jGGMeCcB8vkMzjnqH2NqRpsMZeM3sPzZxrwPtVx8Hiey70sU32OP/+qOX/dT0D9WH9C42VBOQXxizri+plcFvLaNcU/uCD9wyLysTHmPeCvG2N+dfmmiIgxb+AxOhpm0UPWWsM4DJBj9gYLcrgzLb4HHHKtnbSrl+r/6DjH6JdqFbWtxzBSmILumRf5MyK5QK4oLTktnFrGnT8AGKMaRe85a4tJ+5hJ0NQHmf9573M7zZlfoe81gyIlmez+ruuzeRsPknWLaVvupbTudM5RN5oHqMdu1FTyPmu+mP2khT+YEpFD7obiH77JtDzW4Mv3T5lbx+N1gnRK6JZxKYzBO6c1aflr1liEmPst5ZX3Bwk2iMjH+ecTY8y/hXYt/6yYbMaYLwFP8se/p7aXsugh672TmRsuJ73Lqx3poqJPL97TqNqp942qh6kK1liDs4dm4LHpNv2eJciAZg6bGVY9uAZTyrvz583craGqtLufwR043kczRYyR/X6Y+CrK/RfmnmEcptYn5djG2qklp/J75w4bmb/OGJNja35q4+lcZlF1Lt8X0zFL7KuYRop2av7h3d0dstjN3+TbnDK7yrGWz3SpmU5qofK30cpXIzOjrHMOX1U0TU3btFgLVVVrFr5RegJJEYzWoMWUcNYegEunxvctSMaYDWBF5Cb//o8C/xLw88A/C/zF/PP/kr/y88CfN8b8FRRkuHqdf5RPwuTk3WPImcerdq1jm/jYwV36Rkf3dmDCIbCqW/pupJSOv2lRlMv33lE3K9qmUQ69lObYU74GBU1mBh7nVa2llBiDNh0WUYYiazwhjPO85DFnjEv+22JMLhlJiSS5o8YUYFRo23sFDtbrlcZODko/7ORnmIN7y35hCS7nn0r9NVsApVthAVumuT0axybdqU1yKUSnzMIlgDH5f6ZsWsxC5LVItKkbVquG9XrF+fkZ66amzp06YgyEGBmGUc1YLN5V/M2/+e+ffObwD6aR3gf+rXyjHvjXReT/aYz5BeDfNMb8c8B3gH8yf/6votD3b6Lw93/jTSdQiFW37cm04745Zu1s1r3JJFgK0fLv488W00tNE+WKMFYOBOz1i0PNrw8+eI9Hjx7x4sULLZsXmTSIkqEcVrmCYEenhX5ZQ7ipnkkXeeMa1Zal/FvkIGA4U1dZ1Y2WqQvH0gQu/01cdMcaumw0Mufy6XWWeFgOLk+9Cw/nRptJx8kPXMZvpmf8CrPsdZvhciy19T1zsQhVNoXrup7SyrablvWqZbXecLZtaeuaVV2TkoZHkmiZiSVzXpjXi8r3LUgi8i3gp0+8/hz4z594XYB/4fd+nkQJlB0jWvNkH5zn1LWe/Pk6oZsfXv5nzMFiOrb3759XW29uNxuePHnCMGqXbhcDwahgpDibfEuztGSHC4Kz6sQLpd2MmoTWGZShvJCbzJnRBYSQAloV5GIxtFxE2ZZEmHn/TphZ9xe+ZOHIC5hMT7aIZekCV81qzH1r4NScfS/Ccmoca6Uy/0j2M3OrmdKiM6XAOAwM+fWhclSuYoyJpm54tNqyaloaX6sms5Y4fI4JIvO0LhxWm3fg408eVTae0DKvPc8rtNL8upCSmTpNLM9z6uEXAQwh8PEnn5JS5OHDh+x2O8ZR01XiIv9rmc81+wBkU0mdDGdngkTJaUZZp6LK6NWpNxPykTV3jEtHuvieSk2GMVMzsnkOiqnmWK3UBBz6bqqFAlHLewGKgJqkp2i7XvVsTgnX0lx7FcB0OO/lOGp+SnbSpo0KONuu+OGvfIkf+upX+fKH7/Ho0bucnZ3RtjVtXbHdnrHabGg2G1xzBgLh+XP+2//Df/H0/PKWC9IcWCm1M+XBlodjmOutchsTeb1AnTbjTqM9s12uBoxRo/sgIr787mQyZlOm63q6rqeuK25ubqjbhqqp8/o9TI4UkanvUUqBGLMmFKb3TF7khcV0ykFkSaJfFpMWBJf3SzOnqdPCJBhrNpsNm82WuqkYxiELfJh8GxUkFeRxHHEZvXPOIzKqZiotb4xDW2W6k/N5at5OAUHLzxx/93XH0l8VDSkmZZm3qq5Yr1u+/sFjfu4P/zg//M1v8PiDR1xcnOF8g8Hi64a6qXGrFVQbjKmRmIj2NJ10GW+3IJnDWMlkqlB27KWJpa8DzM2E7zeiKuNV5t1s3x+ibMYWwvj73723YwqISTkUYRmGkQePHuCraooNISApTrluIQRiCKQUlJWoXK6I8k8bixhhHMacUaDpUoaMThmDJCGqyszXX5p/qSZzzqlmM4YhCW3T0DTazFrLQhQsiHJMvAgwX+vxHGrZhZ0C0yWzvWja4yrVN5lqy7mFOQv7FKhwaDlMV1aOgjWer33tazx4cKEC5Qyb83OCrbi82RGtpQsGYztEwFY+x8Q8oQ+kMbK/vuGzJ8943XirBclMdv6h3V52mjJE5jyu7M5MD7JM9HLCX/UgT/lSM3RtJjKT8t6rEL9SF5QkZsF3GhhcgCExZbhaEuMwMo6jtta0DkRbzKSYiOOYzyX4ulYoGm0mvXTqnah55ZzkWBHaHFpgHEfqplGhzNXFMY4MI2AMIROrCEyoHZwucTjWovN922yKKtdfQciWDZSXP5fP89j0K+M4BviqTfH+NSZE5sZnH7z3mA+/9D51s8o9hz0v7kb24xX++S3Vx1dY5xScsJaq8iQJhCEyDJEwDOxvbl97zrdakICjned0v1Brq+nzalndJ+g43hVPnaeMY99Hy8+LNpp5tA/Nv0PY+3AcazlBchPnkJNlS2xnHDUjPYaAhKhonNUgobc2a6ICP88EMeVaBW2/OQwDkiIihV0UyMHdMYxq7QGkiDXV5AeZDBsjMmmz5RzNPpydzr28/7Qwv5eb1lIAXunLLeb/FIp3imzl+NqmZ0OCzJn36ZNn2KrBuzuscThvc2NoizUO7+0iA7z4rJqA66yj8rNV9KrxuREk3emLtsk/5xYTs7Y6+t7x72UstVP5e3nO+Y/yUzDOTgDYq/ypMuZSAlsOqtkPMfs9KSNnGUhQTTESxp6QIXFrLM6oJlQ2Vg9iiEkDhRzdQ1nYfd/rufNCcM6pUEXJtMqJtqpZ1c3k7xnjNEvCOc1SWBz7VWaZsRazKI0vaGOKicT9jun35vYVc7d8b/mZ49eX7y0D5fq4lLwGEeqm5fHj93Guwk8m4qHpr9aBPpMwpozLqDWg/uXn2UfKo0yWy4FEMaL1+qWBa96drbWZq/n0pC+PVcZyoZxEjMi7kcQpS6CkA5XjndopS1LpwSLP5leKafKJSvR+HAfGsSeGMRcOghhhCEKV2ZHUz/G54DBO0fZlDGkM44LQfz73dL2Ad56zi3POz85yND9St9rnKZEUaBhUa5VshmM+b5uPG8pGlJISdVqLeMm5ee7EJlbAi0OL4fj5nHoux8/o1O+HFoK6AUki2+0ZvqqorF+YrHPgGzK18uQ2mEl2BG3b87rxdguS0SCkyNKUy/kpYmYbyiyi4sX5PrKvTz2skugpomT4x7uvSA5ZSnZ2rTup4Y93zgNTp+x4i9LywglRVxUxRvp+N2U3lOVuTUmBUfLDkr9WBCrEOJlnNqrJp+hezKzBgpFEDJm3YHHPq6bhwdk5j997D+cs3TjStA3WOGKKPHv+DEkR5z11raxHQsljFJyxiOQuhKJaLoQwbTAazDxEN+dnMc/V8b/j53VsMVRVddDR/VjIlucpWsp76PZ3OUWompN7DRjKBjRnzS+udpE4vKBufsV4uwUJ0FhE2aEsh7y4GQIv3btLpP1oV5uOZAoCtXgIRnKQrpiMRw+k/DSHaODJK12crwhgsdOBKSdtHEfN30rKDdH1HTGUwCVztzprCCkQk5lMQJNBlSjkcpGSAKtZ10McEDE4lxhFBd+LZmxHSbgs0CZD2V0XSdawcWu+/OFXSAj7vVJ+Vb5icg+NxXib51sQmbVwyKlLKmyZ9SgKdVXRHzVYnp7ckV+5nL9XgQvHwlW+U8axditrZHd3xzDsASFZh4slZjZbDCXDw7AIaOsZprXzuvFWC9LSF9EJKtnXhze8/LwcBWePj7g0K/RYTK/ds7PLhIr6Y1OKjXDvgZfPHwrv8iGbiQUn5vL4cdTeSIiZzCBndOOYqn8pD1UomQSShUaKCSiwH0acMxRzRcRMcaSYIilrchHD3X7Ht77zbWJMDOPIatXyjW98nYuzMx5cPOLRgwfEFNnvuuxHCFWl7XHqpiYMvWbDZxNVTTRHSJHaeQVKaq/OurP3nsUpK+E4ID0/n1kTlPKP41KKU+DE8ljX1y958tFv0a62+LrGWo/m3xmcy3Fq4yhEMNZqapW1MuXa/YElrf4gx4FWofgt+ld53xgmIvVT97x8QAcQa9Qmzc5YjDNTK5YD6NeA9q6di/rua7sMa8dDJ/uYxaf4Q9qDNWYCyjnLu870Xwo8hAkmF7GEqD5hMTWK4Kv/IlP8pvgzMYXJ1JUUQBQqj0NPPxRCEBj7gV/9lV/jyWdPuDg/x3lPSImz7Tmr1YqUhKZpqOuGtm0JdcV+v580TdmESmmFZoorjFwyx5cCsPQjl+bwoW+k/6yd+RaKaXpsZi0L/46PBcLQDXzr279D05S6KRUMa8BXGtb2E0+Dpm9Z7/DW4hAq71+XMw18TgRpOQSdYEoO3OT8a6/PYh4dC9PrdkURg/NOW6ykSFo8LFERAslIzuLgxxro2L8qcLJhNulCGLOZUeDjgh6NeF9NJJEhFjJ9ma5xmaQKy+TUEtEvl1eur/iTM8J5DI6AJYlyqH/22VOePHma29UY2qbh4aNHPHz0Du8/fo8PPnifx++9w+XVDZ9++ildt58ScAF88nqfQAhxIsu8nwd4OHfHoILuPXa6VhUkJkE6ZRKe/luPHZNwu+uIQUES68xcbmIV6bV5M3XG5m7thsZ5WmPBmTcKylstSJOBdGQKyDEUOU1sWRyGQ4I4o4CEYYoxLUeBbbXuyOOyU58kYcb8sEua0gmnuIyifcpPhbDnxTIMA3e3N0hOpixCBIHChxesmhLLc2k/6vvZASXrXX/me8/+/HyLs/ZOhUH1gDvicFHGmKaug/v9nv673+Wzzz7ld7/zHb71rd/kZ/7oz/DwnXcVKAkjsmpVkNIc7DVGTVHXuayRzIHWOHyWx2bZvEkUhqj8ZPHeT5rtuOzitHDNVbDDMGgFLwkX5s2nxrDxjgeuYo2hMoYafb1ORrPmx/Q5h7+PkLbDtw5NgvmBHD4stXdlVmCLHbB8Mom20lwyyxhrMNFMh0sS8YsygFM+1XIcxFuM0UXZdZoCtAAKRA4RKMFo9wljcBlXEgshpKnzt8Uo7561GCMHFavlvuVAmHRarLGZ1izNG85CkIo2W35NRJAo3N3dsNvd8dlnn3F2cc7XvvY1vvzhl5nyHkXo+x5jjHbUqDzDOOKvbjEms87aOUh76ActN0g5+txcymHMacLLV837knU2hEByudO6g42xfOBq3m8qLpoGExIkMM5BDKQQIGqSa4rjAW30qfF2CxKnJ+vYFDj+ufzMvHMVc/D0iDGSjPZ9XQ6zXFaZTObYhAMOhObgtfzZ/e52WjyH6TWWkgIlIrmyocD7lBPqcY2WoHufG0yTwMxUvoXvYeKTE5PzY3WTwFgq60jYnBSbJjPZWqea+GhzmmF8KGDPi2fPuXx5ye3NLT/5kz+lPaqSTPxxMUW6fqBt14vyhUM/pjybJchQEm+Xr2uGxKHWX2ql5VzfszREyK1zFRSpa5yBbYh8ycIDGTAh0HdKHtqsthhjGcMNJiXImSGWOZPkVeNzIUiv2vW/1+8fIzz3Yg8y8zeUfqcHGgUwOavc5ODN8aKAw91yIjq0aIVqSll7lHhOuac0aQK1//MlodENby0mRZw1eKvl32VTKBC7MQbrbC6vEAwWKyYnuhqi1jjgrVPHWYR9Zh8t4QWf6268lakfUyHKXM5ZMd1ijPz6b/w6V1dX/LGf/WOcbc85v7hg6PucWWHo9vtJkJbzX57LwTNg9oHmeUXpvaTMm6Wu/QEsPifS3t9Ey8MzxiAxYiWxDsIGg8cwioBJWFEtmmLEpAETBkw2V61eDPHzniIE95MXy2vL3ex1Y2l7T+jPEVCQqwwmoYIFGjTtlsUXmd9fOtLHtLZloSuilUvMDZARQHWgXV6sM/o43ZNVdMlbC5Lw1iJWzU/J1bUiQlTW/ZxSZA92e+sstdF2NeM4gmjW9+QDZt9PW2xamtrrfVrL3W6v9yrAQotMm1tMPPnsM/69f/f/xVe/9nV++mf+SK7radntd4RxpK79gXAsn8Ex/8WxOWyto2lWxCiIaBZIm32yMSfzFv+QkoHCfI3WWQqXeUoBO4acEQOjgEP/pZQgBGTYk8IIsaRuZRKaQ7vk5HjLBUl31cQxG01Z5DOqtRyntNi00+ed/3hmiiU1fWfxQFMOeipwUN6eHVaz0FAzdC4TYYqaPFpFOhfx2QkyXpo15Zw2P0SXA4M2L3j9pmXM3xMSRnI/XBG8USiXbE75ymdhKUWBinqena3Z7/tpiVS1p2kqqpwdjTGkGBhGQ4pCSGnyH2UqQddFuNvt+PVf/zVub2/4E3/i57i4eMCZO2PoB+qmmdiMyr0VXu9jAZKplmzWTIVXQhvIOb72la/wW7/1rawZD9FZk03h5eYG2VIQsNpnmsEII4JHQQUAI4I3lpDUn1Z/MqOER77ZqfFWC1LZsW3uzmazCVNiE6BO+5IDYTkOH9Ihd7TWz+TP6Yv6EMsDWJoHmIlNpjygqRxcYNmL9cBUWTj/RZuV3VJNHg8IQaDkfFlrqH2tDzBrKmMMVS6UK1rTG4NxnpDCXK+DPvC6ria/xFs1BZ21hOQZx5Gmqlitapq6out7YhLatmZdV1ycbdmuVtzu9ly3dzhnGYeAjTCWqt5TpnYSPv30U/7G3/h/85/6oz/DB+9/ie12Q9s2U4/dQ9h91kyHloY+EX3dskwHWq9afu5nf5rnz5+z7zomeP8APLkfV1IKMclaVcGnQC4FNQurI8ZcRqIbl3OaxGucMii9brzVglSGyXCwcxbrlBa3rmowJvdrvd9t4D5DUCac434i5BK9W/o7xWcx2ANn00zp9uVhHfXgKeYK8042jhxc14yOySTIhVPAOw+SiFmIMGDc/MAFJeYwQOU80cRMbqKkkHVmypFs7hXfbN3UmLZWKNlaHp+fc323ow8Dq6ZmVVdsmpoH65o0Dryz3fL8+gapZu0RE4Slxl/OX0pcXV3x27/5W3zjK1+lvdiw3ay4rNS8K5rl2MQ7Nu+Kibbk+rNWOdy/8c0fYbv9D3Hu2fTMFcyU/J1j1FZjhISAWIP1JqdLCVLmFjNtYuvNGTGMhGHU5+E9Bmjq5rVr9K0XpKVWMdbQtg0/+eM/yX/lH/tHePfdc/6v//Zf52//R3+flAKfffYZYcFFUMwlVe3mngAd/zx8mNMVTOZXic+Uza84uy5nSJ/y2courOZdzkQolMH5/SRCJOkuaLImYjYRrc/8eHnhaoNmzXYwxlB5T8rH1q52WmCniawRUGSv8o6L7YrbfQcCF+uad89bnl/f4Z3HW8fKe9rKcb6q6DrPsGroRktnDMOomeImMgl5SoWRNOfZGcPV1TUkYd02rFdr6rrGOsvZ2RmXl5fTfB9oo8Xvx7Glqqq0OLFueOfxl9lszw5QuyJ85fkd+mSqWcZRaZVNJpAha/sgQnIKe6/Pt7z/lS9x8/KK6yfPKIjpm7QRvPWCNE+2zXRUznmcNXzw3rt882uPef+dB7Rti7WGp0+f4f1seqVYyBJhGaA9RopO+VjFVJiBCrnnVxV7fwleLK+9kCUWE9BapxFzX+VschX2cQhg9JpjFHydU1UwE8cCzJounxzvLJVTxK32Xr/rHW3tiQkkdxz02YRsvMUm4bxtGEOgdo4PHm5592ytSaxBGDIgIUm42DTUlWXfB267nrt+oB8C/TgyhEhMece3hUNO+/3e3t3yW9/6Fu88/qOsmpa2XeGd5+HDh1xdXS2eqV1slIDVBGQJ8UCYyhyHEKiblg8++JBf+ZVfuZefd2wmlnO0dauNqvP81pmAU0RIJcRgDVVbsdluIRp2z7UP1pJE83XjLRekeefS1CA1p+72O01PGUY1V4zkHkhzzKGYA8p37ei6PUOKR4v9dOypnLdw04WgmkDJPg4FsTDusNB0OvEzv3YIgRSDfkYUSWubFtDM6RAi5PSm5YNT82/epWtfaawnaQlFU1ka52i852zdcrbSQr2QhMpZhjFyN4zEZKisZd1U1JUjxETtKnx2MR6dbVlVFdbC7a6jDwG/XVNXjiEEbnY9Hz29JEnCWzMJ15i1vzWWMUY0SV03j7/zi/8R7zx+hHeOdSbi35xtpwU8bUpy8EPnMJd9lPsu8zGEgZQSDy4eZMDITCbCsQCVNaBWQ8IaRwJCEkISpHJaHFlyEbOl4ZqG7UPP849rhr7HWKMdHj/PXc2X02uNw+WirK7r2O1uM8Q6LLKCtfAu5fjM2fmWhw8fAPDxx58w9GGxcy0TKA/jTGVUVUXXHV5KeV5LTXQqiVLjMNpGBoQUFRWyxi6OVco/1LRzucFwyR9c7tiVtay8o3KO2jc4Z6ic43zdUnvLw+2ad8/WbNaNnj/7Ll0feHZ5x9Vuz2bVUjeeYYhESaybmtY5KgytNbRNxaP1in03cNP3YKEfBsZx5Gxd4zO5/03XUXtHiEIUYQyRcR8m5C3GyO3NHf/B3/oFfvoP/zRn2w2Vc2y2Z9lPSjlrSxkY7wUvDkCiWUiGfuD25or1ZrVEuxfJyqfLZ1IGl6IIgcSYElEcYi2jTF4qcQyEcVC65smcM5Sg9+vGWy5IOiYbGCUWKQ5r5Sv6IWVnc6boUjOwoq7bAzOi+BxFqxz6TPfP9+jRI25vr4FZc83YEJOzba0npXB4LudYrzdcX11CjBqnQh+o817jGaIaaxKyRWpMOVdhLmqrivN1Q+Mtta843zScr1vayrGuKx6sWt652FJVjlVd09Q166bGO8fdPvDZy5d88vJKg7xbxxgjtXdsmxZnDbUzNJWjdo62avFOMyli27CqKi7Wa+76ntuu5/10xl03cNMNXN3tuU6dLrQEURIlQ+7p06d8+zvf4Zvf/KG82NOkRUpgu5jctphnWUMsY02bzYqrqyv6vufm9o7N9gLrPGbK6+NA+O5ZHeR4kwAYxiT0MdFWjmhgRAky97ue/e0t6+3ZXKQYIylqxfHrxudCkKAsdJl2D+Ws9vjKT7u+NQbj3VTK0O077m7vAMmFcw5r48ndjmw6zkgRvPPOQz755KM5edLk0mlmP0xKbIg5em+tY9Wu+PCD9/nlX/p7jFHLGbQ6U0vAh6FXghLRNH09Z7a1pg54uZeRCM5A4w1na+WoXlWWd9YVDzcb2qpiVVc8WrWs65rKO1yttFLeOqqt52z9AV9+9wG73Z6Xdx1BhMZ5Gu80Tuag8Z7a6d/Vus15ihBWLeNF4q4fuOv23PQ9XR/45PKWMYSpDOWGHhMjJmV8VBLPnz3hx/7Qj+G8YxxzTCiMapGZcrsl/cmCkVJ1ps/CKMm9tZZxDNzcXHN+do6vPMOQTbhsmp1eN6U4M4c78k4YRTM+jDUEDL0kdkPP1fOXxDHSDyOxVAYP6hO+brz1gjSjOrrQC5mHzd0SnANrPSFo8O/hwwd895NP88QPU//YEnNQm/vo2FPGwkxYYq3l/PwiR8ezj+aKAM7fLyaldp4A5z3eVVjnqNuayvts4uUUmxiIe30oVVVR1zXDMMywODlh1WlcqPWG1ldUzkJKrCvHe+cbNk3NxarhrGloqoq2cay9Y13ldize4apKu1rk+7zYrIjnZ7w/jvTDiLNo1nbuaGizsMcI2zYjhRiGYHHOsK8t+8bx8s5xaTrGbUPtDLf7ntpZmsrRp8AwCl03EiSx73Z0/R7nLKEftGq2GxR1vEdqOZuzinRaUmJCRlNKXF6+5MMPt1TOT99ZxqjKWGbip/J++TzKtDSKJ6GcgAGhT4Fnz15y/eIaiVETiDEESQfFBKfGGwXJGPOXgf8S8EREfiq/9ojfY59YY8w/C/yP8mH/pyLyr77p3JDV8mT2FKRMTSNjLJXzOGeIQVGs9aqlrhUVC0EpfwvNElMk/HD7KsJx7LA2bZuj8rmMIb++BBb0zzlj2ZaAXxjpup4hjIwhPxQRCBFnoG3aqSxgysQwepzaWdrKUVtLZS3vXmw4a2pSjDxYt3zw4AwHPNquOWsb6qqi8Z5t09LUHl+pIDtfYa3TYkW0vLytarwbWFcNZFKPMI5IVWllqwjGVgRR/84Zr+xJIrTOkXzIJmXFRdtwud9zeddzvl7RhcDLfcfT670KyRgY+4GPPvpdnHP0fadc5hmCTuaokYEpQdPDYKpC2Bqcvbq85Jvf+Obko+rTPK2ODkChuiKGISODGVzIS8Fbw6qq2BTCzJAmy6NwsL8pSeh70Uj/R+B/Dfxri9d+T31is+D9i8DPotf/d4wxPy8iL990cmuZukCYUuxlLNZoaoyzNXVVIzHinULGdd3kXKw8UxPy9aazqWYqAESKEZd9mZTkiJTRLlhy5iOU91MUXrx4wRiV2kqyM6zlzX6Kg4wlUptH5S2Vg21bsakr1nXFqva892DNeVtTO8dFU7OqKhpvabzLyJ2jyhkfPhfUeV9NaT1jGKeF4YzNPY508frKQ1JuvRijCkEpOMyxrSAa9PVYVl7NwrayVN5ytlrxeAzcjoH2+o4YhWGMWvpB4qNvf4d6tWIYB61QtRNKkE31OD1sSXHSLrpZard55xxjCNzd3bJer9is19ze3EwmNouN8FWaaRTl3XN5Y/bWsHGeM6dz6DDEEEkhlTwUbPb43pTP+UZBEpH/jzHmG0cv/xngT+bf/1Xe0Cc2f/avi8gLAGPMXwf+NPBvvOn8CuvkFA2jrUqstRgx4Cwmd63zqw031W2Orzh2u/3BUUxGzBIxm4nLiSmR+gPEgTGG6XfJBnYpDJyBBccpMsQwDjx5ckUIYzZVFDionNUAapp34zwnGGOoneOsrXn/fMOjbcOm8WzrlqZynLcNrfMqRJWnqSy109iUo8Ax+p93mQfPzLGYCQGcNGECktY8GiHla4ghYHIbk5Qg5vSllMGA1lUkoDIRuwLbjVRWYzFnTc35quXqrqOrLCKOcRgIMTJe9FTezwxJ+bmoP1Xm+lDDWDubz5JGhn6PNZbNeoMxTvMZTZj81WWK6RJMcs4RSzm5LfE3clBWTe4eZbd1Jjd6y345UlKRXj2+Xx/p99on9lWv3xtm0fqyOOb59Zw9oESGVdNoXX2l9TXrVs2Yymeiw0Ugc16sCi+/2hSYkZ8CU2sOm1FYJ6flzLGiYhLOu6ERSCHQ93u6bjc/YGO1FGJRUrDMIzMmE0Fay8WqZd1UvHu24aytab2j9Z7We7yx1N7ReDVrvXVYEsb4nEpVtKrOn8uk++VcheuhwPPOGr0vY6GESrxHcv8lawSMxSeLYGmoEIOagdbiRe9rjAlHojKwbWoebDeMMXHLQN+PpEz60qw3iwWeg96UcEQ8WPzFjAdomob9vmPf9fnvdlramtqbkYscWrLGTM+lmN7OeYzVEIJ3johhJ0IjQpO/WNUVtauzposq4EbT0143/oHBBpHfe5/YNxxvan1Z17XoLkrePTXvyVltGuVtNZFGKs5vsM5n59Lh3BGd7xQBP3nm6XPFHHDWHCSqlrSdyVeCycTL106SRBwH+n5PjJoaqT6a1raIyIEATZC3Be8M1giVg8rljnrGsKkb2spgxVK5IkQGZwQk4n2tTZGzVigBT5Pz2zAlJ02d5iLMFgGpMEaZWNWcU/ahRMKkXOqAEkXajO4JAk5NXpsMtdf0p9omNrWjaz0Pxpp93xBizGlbKnzez6SRZeNSIpLc3W+RzV/mJ4REVVeAsN/3CPDg4sHBc50EUIrvswg05Xkv/2G1Jc6AocISjOoxZ5SmK8SwiPU55XDwrxeV71eQfq99Yj9mNgXL63/jeznRFEOath/NAK9yMmExBMZcn4PAOGiLSn908zMCOEPYh47tIdhQsilAA4+lGK9oo2JmLj8f45jRwmXwV2m2gEmICnjivKfyltopg5BHoe7zVU3rPZW1OCMYMRrvyWZJEfKS6KoFfg6bVCva3PRFRLQdWSa3FGsgGiTESbuazE5rrSNNJJXLiVNfVc8zI51pHLAYKmsYraUyGShxaoZ2m8AQAv2g2RvD0E9mMBR/UgseNcSktVYsigkVfR1pmgZB2O133N3d0q7b6RhLwZuBi4yAWi08Kf2SphIWNKNejHpBId+k8444xHzP2obUOout69eu0zcnEZ0eP4/2h4X7fWL/GaPjTzD3if1rwD9qjHlojHmI9pv9a9/LiZaL3VqLd546xxWcm+M+c5eFlHeUGRSYewplkCCXJCwDoLNAkc0DzRywJleYZ5ADmDTSUohKGUNKgWHsJ5Nu2gREtWqJeRlrMN5SecNZo6BCBTzYrPRaUsoaaiE02bx1xkxp/vPeYKeSgOKsz9WtaurYhS9XrFuNU5YK4YQwkzmWRVoCpd5V85wxm9vOWnzmGa+9o7KakHvR1mzbmtorAcrN1SXPPv3uPce9ADZLRLScv4Ayde4rtd/v6foBX1UHxzj2Nct6UfMwWwwJSiWB7mLqZydjSRkAEaDdrKmamqqtabZbNg8e8PDxu69dp98L/P1voNrkXWPMRyj69hf5PfSJFZEXxpj/CfAL+XP/UgEevofzT79PE2MdrvIKGmAzyYVMWkBrlLQYrxSVzejNzI12Gokxk72tLKNzvX5B3sr5luZgjIEwdjnxtORwLbkP0nx8q/+csdTWcbFuqZzu5uum4mxVZVNHz5kS+MpRO0/jHJBUAvL9O+MOhBqyEEU9QMoEKdNGIEnNt5wkakoGt53nabnTT5uCHHJOCEz0xJpNYmjEs24quhDZDz2r2unfQ6AfIrfXVzTr7b0S9nLO5VhqpKJdu75jHPqDjofLNXLymZq5glmyyWudQ7zF2Lw+vKNuKi7ef4+z84dcX19hK8dqu6E5P6PebE4v0Dy+F9Tuz73ird9Tn1gR+cvAX37T+U58Dzj0RTQQ67KXL1hncJbcXRugdNNWwSqJq8ZAjMdlEuU8xTTQ91IiBwzz4sq8cs67XI80O/UAkkLWijEXp1lMTiQzplSymulBgvpEmYaatnY82rb0fZicYdU8UJVqTWPA5BLy3J2wEB4uNapEyaiHQEhQadBEFoiYXqLCVsXvEwz2iICkHFfL2ufGyiEVzaUC5XItlBdNmD1rG+1F1Cgd864LDEGthSoGhbqzv7gMwi59WRCc04YBxSTs+57b21tub6+PPvtqiLpsnM55Lh6d8/WvfYOHF2dsVyu22y3rdcv5ZsN2vWF7dka7XpMQbOVxlc/xuNMAVRlveWbDAkWbdlxo6jkgWxzrQv4x28FknyXn5VXVgX2+PEeJH4mUn2hfVrMoi05CEjXT7ASBF3NNKCT3x88yb+STj+StxnEshtpZNm2Dt0bjQd7xoG2wBrZNw6ryrOoqZzmYSRNaM2vN4jxDLjQ0JdWp8A1IjnPmz+T5jJlqqnCCq116aBotfQ5ny0aUiMZMeruyniRRm6qp/UpdlaZlBqh56h3eGdZNzb4fFHRoV8qWtBhlwU/3MoUl5iz7YRj47LNP+e53P9X5XQhdCVGQS1DKPaRc3eyc42vf/Ab/tT/3T7HdbnOj64V5n62QclXLmKG9t24Ox1suSPNkLJlonMnmiOg/NS3UmT5w+oRJeywbcjl/SOw+z5E67GUhptxR3bmKVLoGYnKTqtKZL2nQ0jlIiZTmAN68EEuirD0QNmssq9pzsc3JqNZy3ta0lc9Ou89FgKUJVq6qzT6Oagyr5Qs2z1e+Z1U2pSCxdLdYEk1mLZzfiycIKO/7MnYyVycksJTwZ8G11lCJBQ9j1HNv1w0XmxVNSMQkB72XyvmO42plaLqW4atfeY+PP/odxmHk088+4263mzSmlI1A5jk4FeGoKs8YhKZW+uXJ3J+/nIXpMP6kb7wemP5+wYYf2DhG04SEyWkm4zCSSqmCaDPmys5FdsvvLQWntIuf0aNCvDELUTHnbGb5SVJMMpkYVBUIUEjaWZd7rh9O6WR2YLOtPi/wMSYyOzC1szgDtdXOfOoP+UzDle+pZF1Ipt41JSdvwmpR1pzFAhPJ3A9CysSHJikFVTXt3GqGuknrJ+yRtigLUym/8vEktyyY1qFMgIhBtYq3jtZZHp6tebBZs2qb7GulKafRe59ZjFw24cvGoSetK8+P/+gP89677zKOIy9evMQ7vzA9FXWcIW4t6Fz6YZrL6Oi67oBLfF4jMwo7W0Fmasz2uW7rUhb1tCvk5EFjc6NhY3OsJgsCykj6Kuez+ECl4G92nMsuPFdrimixYMwNvcYo2pOIJZ+DZKafbMI5bXYcFucvqKLRi5nUnzGWIURu9z2Ptw2rjedi1eCtpfWOTe2prNOUIaum4MQ6ajIJTKbfwtqJ9WapUbR0QH0lyR30xBpELCmO04bhRYhikWTAiQp9Shirft0SBbTOTbZPaTigJuKCD5BcymI05Wlde6yz7Ebhet9zu+uAkgI0gz9LDbNE8qy1WL/i4aPHfPt3P+bq+pLHj99X9tphwBST3hXQTyatNAl8nvvb2xv6viOGQERZqGKIpBQVMAqBGCIhd+oYRu26MYTPtSCpb1N+N1bNLQoKZW2u0jQqZMYw5AZcy3HI3DkfN+XY00xUPzuthb10igXBTBK5/MdhfVPtK8DQDf20mEU0j23V1gxdP6F4zlSEKPRDwBuF21d1xTqbdnpMO5k3is4ZvMk+otOsc+dzhsMC7tWSEUWkpJC0WIMtO7donpspXOMimuVMZm2yltpYolHELmY0UjBZo3uqpJ3bY8oxKWuwSTWRSUpu6a2lrStMSNz0OwxQ1w7jctwqyeHGZGe0UhFLhZC6fcf2bA3Z7/2Rb3yDoeu5u7vT2NfUXzhnS0yNFsimLYQ4cnl5w9/9u/8xFxcX9P3AOMbcuCCQJGZ0NdOb5UwJy+feR4LtdkvXdcUtAAFvcnNho+XHug3pQ8Yu+X4O4z3HglKEqSRmFLKU4lNdXl8TUk6oXApa8bWsJqeU75gc5/JVRVXXuYw8YIzw8OKc9997n9/8jd/Sh5aSInNZABAlK2ysoXJqZpUesoWz27lcAOgMYizGVlhX6TGcAWvxvs4Vnl4znZ3NFFParsSIFt8ZHClEMDb7NxocRcDZSsuwTcIWmjByvlqSSeNY5/DItOuDmrxV5qgQD13IQWuTqL1j3Xpe7lTb6CbDZN7ppqWLuAhB0VaffvopMaOoMUaaumK7XlG4PFIopSrz8yCbeaAbkBEhjCO/+Ru/xfkiM8LauZ3L9LfVxF4DiLXI57mwT0Som4Z3Hj3ks6efTa97r6n9BkOIxffO5h5Mi/Mw4l2CoTqKz1SEbK4Hkmkju7x8OTUFK48lxoRzCW+WLECZNtg4nEcj5LVq0hgC2/WKdl2z3a6p6orUBSTBrusVto8tLkfaLQVI0P48ywCo7tKqpQxgi4bUtAOc9zjfgDOYTIFVjlfqsWJKitLFwrK0YOHR1Z3J5PV+TEparEdCcghrJBCzT5WSxVrBiuCtaG5eFGIUIkJl9TnFpKZgW3m8tcSlWwcHCzilknc3b2p/9xf/46wRmRhrt9uWIjilfKUcqwxjFNQsQXlrHSEcdhBMueTdyrzhphTVbC4+2P18j4PxloMN6mT/V//Lf4pVUxcDmKatMMbQD2ojW5ubcDEjY8ths0+VkpJ1lDYfy0ZZy52pPJzdbsc4qg9W1GEp8BMrFLG0xmFy7Y+rqkz5pF+pneedR4/42le+yj/+X/xT/MiPfHNCv2JS2ztJLgEp/s10Pl18IcbcwU/jP7th5G6M7FNkCJFRhDEJUQxBIGXehJgiIUWGEBhjYIz6d6Er1vlaoGaCAiJ5AQuZAqwIr9NAuPee2nnNaPAen9OUtLudgiNNpRtBlU27Om98qvQ1AVZ901OpWqpBlW5M5+Xm5jp3N1TTbtXWtKt1BhZyP90js3sikjFCVfnp9WKyL/8tgYn7geJXBe/n8VZrJGPAJuHx43P1Lwblf161K0VSoi5GEWEYwzTJxQxbCkgIZWfRoZNZzIvZ5JMCoU7XkNGb7IyCyYmVahUp+J3NFy8Y0S7kJC2YM8YwjAPvvvuQb3ztq/zx//Qf5dd/7TdJeVeNovwB5aySgZMkWr5gRBMsrXMMUXjR9QwJrA3UQ6KpEs2orSn9EKhcT+0bmkrLS6xRONjm/lAWyZzfTOlPikdkwsTlTj2118wl1whiLTbHtipvMTEiY8YNg0GkEL4oKKT0Y5ZtU9GNgbtafVy70DbzPNvFvKsG9d5NVcTe51QfBFd5Vqt2CpQurY/ycxKK6Xg6ybFoJHO/DmDSRm/I9j4eb7UggQZaX7640vLkLBybdTuZNGnKVCgTNps7kmHOw/QZbWlS2poslfJs+tkJ3ZsfCBmVKrt1IomdNJWjpB85Kmt5+OARNzd3XN9ca7S8aXn8+F1+/Cd+nLpuGIYBY9TPGqIwxNwGzBqM5JhPRt2SQD9GXuzuuNp1+KrGWEdIwqpdE6K20fRG651Wtedis2HVNDy6OGfdVJPGUxNRNx1fFkuOKyEq2FNnwDxDiWz6ojzYYkSRQu8wY0CiEG1ikDHXFkEqJmUGOJyztHWNt4GUDFVbUxo6H2ojM825MUpws16t6IdeszicQ6L6W48enE3moN7GUoscZkkocKGJzCbD2kaEYOY8xpJlUVDbpel/P5B/ON56QRrHgf/zz/81bvf77DQbHj04x+adfszdwEvqmyQ0xmDU5j+kUTqExWdkZyEs+WDGgK/8XMGaHWGbTQ9JgrFahFw6lOuDt1TWZ1MU7u5uadoGZ7RkeXe3B9HetRo4VTg7oYsYg6IOxjBG4WbfMcTEmKAPEWM1a8Bby9XdnstdD8bQDQq6rHyFoDC1JOFstear7z/iw3cfcLZaqR9mNQ41xpTp2gSbid1iUg1kADGGMUJKVrM6cguaISV2o/ZA2nfK27frenY5RlM7w6oqQkqOf6GtU2BqJjAzP81jmZak5pnn4cWK5y9fToy2IpqGVJ+d412FoT84xoy0zs81pIi3jnEYc8M3pQRb1pgJIDFO6+SYXvl1460WJINh1+25++6OMaTc3LeiqRvlc7ZJfRhTfELdzXxVsV6v6fv+YMda7nZqEthJ5R/vjBhom5qQBUn9zrmyc3r+C4vRGHBYLTysKvoQcdaybVeMYeTy8oqh08rdGKPW+8ScNT5pN72uJHDbjby47XQmnPojZ6s1bVMrRmktn13esB9GhjEyxpGuD9mBVy1w0w08v77msxcX/NCX3uXhdsuqrjVL2xtIOchs1IkPUTsDapqe0MdIEujGQBd049oNPc9v7njy8oa7bqCtPA/WLTFFbvZ7Ykw03vL4rKXKtT+CmqpjjDRtg/dVRk9lSkBdChGojRFSoGk3uat8pnTO5tfD7Za6rtnv/eTzzSa9lksYdHONIVK3tQrVODDGUYltRDeUJEpNLABJlKcCDoK3rxtvtSBhYBjHOZva60Tt+05NGZ/RHZSnAciOvk5mnWtIDoOyMMca7gdsp1wvY3lwtiH0fT5GNuvs/L2UZMoGKMdMknCVxn3O1iuee8+qrbm5ueHJJ5/hDGy2a+52O/J2nc1FzeROWLqgCZ5dCLRNzWa1YtWs8N7mJs2Cw/Jg3VI5x2Xf0Y+Brh/ZdyOlH23M4EqP45MXVwjwwx9AWq0JtWfT1jgnlH6yMQegxxAZo7AfR273HVe7jk+vbrjpem73A12mq+pH9VkNwqcvr9iuNE/QGhj6yH6InK09m0Z7Lhkx3PWR7faCIZfxK0o3+zfHGQcxJqqmoQA9WGWa3e/3/NBXv8bZZsXNzS02Oe3ibkp2yeHCT7kaAJTdNsaAsXMpRhItZMQoYENKEzr6vYy3WpAURFDwYI6uR15evuT5yytWtaXvO3b7Peum1WpWwDvL6Ayr1UqTTlNiv1dNUOIUxdcp51mCE/o5w2a95qq+zRrK5j5EstBeMqXCkIN2vrJZiCs1k9BiP0lCv+9oq5oHD8549vTZQpPpdxOah9YPujms61qZkdbNxH3nTcWYIjEljAjbxtM2ZwwxEMZAirAbenbdSEiRu25kjIkR4enlNR7DN94X6sHh7Blt7SmFgUmEIUT2fU83Bq73PU8u7/j06pq7YeR2NxBiYoyBmFtdVpl031k1tc/WDZtWG5mNIfLZ5R1t7XnnfIsgjCHw+IMP+O7TZ1OPozLfS8Bget4x5YTj+ZmJCLc3d7SZpL+YiWAo7T6XjeTKBllIHmOMpHHEeF88AlikRBnReBuSkAwk3YclDsdbLUigzqaxUNcVbdPQtg1//9d+nSdPn7JatazWa77+lYa68jx+9yHb7VqLzLxXog1ruN3t+b//P/497u5mWFMfVKEi1LEUJjURlTvPAFLU/0L4jJnz59RBy60rBR5s19x2mt2wXrV0dx13d3dU3lP5aoJrMTMJvmAYcopT5byWU3hLlEgY9zijbKZJJJfbaz9Z55WTPJIYhpFhrOn7kX0f2a8C+6BaChF2Xcfl7Y5NW3O+WTEGcFbvO0TVNLth5HrX8eJ2z6fX11x3e9VyGaavfINYgzfaCGDb1mxXNd6eayZD5YkpgDGs2opPnr3EmFt81TDGyHa9whtLNCaz5N4HG8ozCiHOlGgpTYHbu66jaVrOtpvFdw/Bi/K6iOQAs5nqxVLMXRLNzGZbaLecn5sY6Crk853ZcH6+5R/7U/8w3jn1j7zPiZoluRQev/+O+jxyqtpVf+9HNYfKd3QozHwvGr7weSrnMgyao/lLk87lGEhSTeJcJh4MMrGjeqsUWXVVMTjl0O73O9abda6nUsywrZQA0ptSh2QzoiYkROmhRFNcNEPBTvwCVVVRN/WCYRaMTZBLBxpvOUs1SVImOjSEGEjJ0fWDCqdTSDrERDcO9ONIFyJX3Z4QA7W1RCPUmxpvNYa0qStWrcaSXEkwlVL3YxDjCZI4X7UMD8643XdIJqJUv7ZojENOjWW6kC76oGlJTknvnVPh2nUdbdtysd3AIh1oCaFPT9qYRe8r3TDSgjcDmICfYpqWAtDJ33qDjfdWC5L3nocPHlCym5eqWqViRmcMZn4dDjTPoVpewqJQojeLt3UYsL7SfLXs+wwpTEHEQjxZPlygZZtJK6vKse+GnAnhsZIYXz4jPPsuZ+EWY/TaGl9x3iqa5pydmiIXPyyOYdopNfky0oegrD0GJcJvW1xdE2NgHOMCENH0p8promol6ru13tJ4T6lpMkZTfMaUGGNi14/s+j6DBo7KaAJq7SxtW7NpG7ar5iCDPsVEiEIYEyFGohFSUJi9qTxRWvoxYlA/bw68HqJ2ywyHfugzBZjFu2oSQID9vsdZy8X5uaIJHIFFyycuMsUWY4pUyU5gDZCTfxVa9/m6StvLeXP9HGukyYBd3MNS48j0f+Xjmt6zDMSlpOCETKjOoU802+WHOV6lZqfyDjGaOqQ0tofgxHSunBmtHeIszld411Ngo353S/fZjsfriq+er/BO2XGcMaxaR1XlEhA7P7MSbU8pESSy65W0HqOdIyqvHeVSEGIcEInEMeBEckaGXr9SkOnNGzTWVHuD8+B9LsVIgshIjJEhBvb9oEiZKLWBNdBWnov1hnWbwwsp5cROM2WCi0SiRMZMoqLEmo6mUv4462etZaKmF53yUWMplESUYMVXhFFZmVIK3N3uCTGxPV8zt8A8bNuzNPPmuJBmbEgm/CwCU3gwlrwX3yv0DW+7ILEQihJdT2nq6F0ykkPmTBvGkXHUvj3DMDCMmr3QdT1dN+Qjni5YO3R0AVO6WpRGwrO9fMoO10/MRYR91xNipuKyak48eXFLLVvaumZde27SgLOJTaO8dbWvtJO2NVqzmBM2Y4rcdntudh0JQ0yG5y/ueH6356rriQnq9ZowBjaV5/H5msfblm1VMYZEXYk2WBblcbBUeCvU3lFXmqUwhqDJn0A/Bvp+IPa56RiC8zb7UYEoGg8KvTIm9WNg1w2MIeIqT9XW2aeSTI2mGeuV1yDwOA650DBMc1i4NYYhtzFNhZAFhqHDV47+rsN7Ry9C1+8YhoGHFw8xxnEsROW5TM/HGGIcNeFX0JIYUa1kS5oSLMy52U8rqOrrxlstSF235+/90i8rS2cI2tcnRlIKE8GJqm0y6jQTjBRcuUyOTAmf+ollj9nye0GG9AgGh88Mr6Cmgy1YNTCbjyklzQ42ZCSrIoyqvbzTNix1XTMOe7715AXnmzXvbFcgsGkaNrW2U6m8xXvN7k5JSL0QU6APgW4I3OxHnt52vNwPjLbC+Irn1x0xCV999D43t8/Yj4bBC51zbG9v+dL5Coth1db4yjH2naY6iQZ167rCVw7pwIURa8maLWnk1MyUZMMYuLq7IpobxjhSeUdjlUpMRBAjvLy9Y7y65uxceyKRmVRtAVWqin23V3QtReam1nKQSGyd0zkUNTerTMO8Wq242+3Y7ffs+4GL7fnBwi/jlBaJMWIzE1LZSyVvFCV4O7G6Lp5t4Ux83Xi7Bakf+O3vfHw/WAr3tMrSVFtO4sxBd/i94+OVxbIMtMZC6atf5BCUmP8VO9rkBYHA2baluxxAYN/1VN7y6J0znnz2gmfXd1ysGvohsGlqaqfkGpXXjg7WWIKJSOUYxxGH+iBX+57dMLBZbzh//0P6ELm+uiKmkReXN2zO32G4vWKzXvETP/UzfPrt3+Tl3QvO31nRVp7tuuE2p9pKTJnbrnAygBUhxZGhHxAJ1I3NJRHCdTfyG7/zKZ+8uOLh4/foozZedkZ4cL7FxYF3tis2qxpJEbndcbHdYG3OCLFqujaV424c8b7GGoglPDA9N433FM2fUmK331FVDcYY2lULYuiGwG13y/nZZvInl5wPp8Y4RnxVgrJpISRLwSsm/uFaktcrpLdbkJDZT3iV7QvzJLxqEkXS1D3PGHckhCVVaDkygmQMipAb5kZTy/ytI6FF0/Sds2zWDZ88GzGIktkbw6Nti3RrntzsOCNDwW2tBXBO65AslraqiMkBA2bVsm0atqsVbdXw5PqWaCyr1mNW7/Dk00+Qq0vC3Uv6cUdb1Xz53Q94d3uBf/9Drp9CP/SctRXvXJxBHDVbIwlhGIhDr1ooRoiJGLRWqqodj87OaLzjuy9uuJOGvVj6kPj05TXrh49J5hYR4d2v/QguBT799m/QP7viwabh0fmKpqpYrWplS3Ja5tBWFbdDj3MOjyWZNHVJL/7pMiiOKOHJarUihIjN1GMxRPa7Ox59cKHJrGY2rcsa0bWhzwy0ONEYg7FuCvIv0Tn9OaN7paDyOBv81HirBUko/MvzzZ38XLnJaSfLf2Z/Zc63k+lf0S9L+PX4eOqYMmXunEynX06ynVOU+j4pjbBx9N2e3X4gbVdsGw9mw2fXOyxqzlmrQWQMxBTwrtYYkqlJlcLtTa+wcTeOJIE1PRs/sPqHfoxf/Z3f5Wa/xwAfPn6XH/vKe7i7z1jtn7NZezxQOU9dVdTOYadWNwaJQhoDcYzZfwKDUHnHg23DOCZ2wfDhV79K8g23t7d0Y0+6u6Zyhrpe8fDBA5rKUo07vvvpJ1zvO9rKEdYrrQXKxYopBWrvDszq5SgZDuUZlE8Mw8jF+QoRoWmbHNIYuLre88M/tKGpKu4Wvu/hBrc057WSVzcLDYloFXHhiJjLNpaX9yYhgrdckEBTNzi4j0kEODC1IDvGqF1P4WfQ7gvG5CSTXOCnZuD9881R8AKbzw9HU+850Ehlxo3ReFJJKNUyCZfriTQWIrbFuh3eCXVufzkZErkadtVW+NzQrHRoD1HNloeyAQNDTIQE0t3xwXrDV/7YTyHGkLqO2jn8/jnOCu1KQZSmXmONcHd9g0lCU1W5dkiQEMC7iS5Mif4NKYjm1e073m0dm7jjwQfvsLt8j4+ev6QPe9arlh/6kR/j3e2acPOSu7jnrHE8XG1552zNpvW0ztF6i/da3Fc7x9CPrFeRZaHlbHnMaUNFIMZhyOlehlIbFcbA7fWOqqomOoLpOS3WiVobZX3kJF2JuUl2YcPNhYhRN93cWy77i3Ou5uvGWy9IQG5rKBpPUoN4mgBd63aSrRJXAtUodV1P8QIWk3xfsRz6WCVfqzjAUjgLkiBu8eUjZMhYS9PkbgaivZFC0MRau14x7jyr2rJtE229Z9XU+VyFRKQQgpSWMQYvhqpOuKqiqj23+479oLVRNo6Y3S2+rglJSTz6MGq7TGepa6+1SdYyjKOaVM5iJGJEhd25Cmeh8iOVraidZxhHdrues+2Gx49qiAbrav6zP/WH+Ozqmqu7HVVVs91UyM1nXF9f4SWxrj3eWc5XDQ9WmnFS136KxTir0PYwDMSUN5gDU3mOLZXnEUKg8pWm/lglfokp8vzyBV23y8T8UHxhpQIwsynBHEgHzQQfw0jXddR1PZmSdV1P7VSbRuNkde7D+/lvxiwyaYaS9r4cBpDMNTA3Ep05v43RPkcpzZpsEsLy6XtmXfGR9LWUH0xWSNNXizPMQUGaIaUwVWcmSVp0CKzOtqyGO9I48PS6Q0SzHwQFyPSylkCJo/I1xuj7q1ZYr1q2w0g3RLqhZ+iV2TWNAZPm9pXOOdq2ZrtesW4ccRwIFHJLZS71rsLXNZWrEUlUrqKtWmpf0zjtOmiyFogxcHNzwxgTGwPtttXJCHdand6ApJpVrQ28tm3Fuq1pvEOsXr/L5emV91PcKeVrWs7hnHGiczyOo2YemJwS5RwxRD57/pxf/KVfZRjHOY6+RPDugVQaJ9IwyXggwHVd86UvfYnNZkMIQenBMmWB8+6NWun7bX35Pwb+m8DT/LH/gYj81fzefx/459A2bP8dEflr+fU/jbbFdMD/XkT+4pvOneeCpU9jptfV10lZgOZKyZSDfPq5lBJDP0zfOeZyOGWrq8kxc35P0e2SZrS4himLIKNfPmuAfhiyWSc45+mzpJytN4TB0VY3SmiSUcXJgjWS/boKZ5Ug0ppMYuIMq8ayFbTjuEAMaCl50IbTKSNx1kHT1DSVQ4L2qPV5zhCn2Rbe44yWhIeYtDTcOe275JS0JYSRddNiqoqmhXC3J4Q+p0W5iYF1XTcYI6SkDZ7b2tNWXjksknb8A42peZvTnLwH0Ty4TGeTAQHmeSf33c2vGwPrVUvf9VxeXrHdbnhwcc7T55ev9GExudZogrd1U4xxzn5xmSa60LQVP6lcyz+wIHG69SXA/1JE/mfLF4wxPwH8WeAngQ+Bf8cY82P57f8N8KfQJmO/YLT15d9/08l1hy+ZCrNpN6nxvDgKp3eBQkFNgiiRYRhPCkwZy8lffk75CgopI2WlTw+ldKDQg2ispPaezarhfNtyuVMAoKorMA5rKzZtTTTCqqlYtTWFZlhEpupUweYGYtlvcV575VYO4z1kLrskQt9HVmKIYcx2vjrUzmkmOikSxJKiw3rU7DEGjMVkJLIktCKCc5pjOFOAASRWTUPtK7xz1H1NPw6ETF8logrlbNXgnfapnQgerVb+kgvpnNX2kpvzhhATpIi1mkcHxT8qCauZtTblMvm8WdaNlsfsd3d8/Utf4v13H/Nbv/27RFisifwM9ZFNfnOhJpAkpHiIvMaYsrewDLDbe8c8Nb7f1pevGn8G+Csi0gO/bYz5TeCP5/d+U0S+hd7sX8mf/Z4EqaqU3qkkMU6LfYI8y86Ro9A2t7VPEa3qPDTdTqWOzJ/JJkVKhDHkoC8TW2mxwSUl5ZVbaDVjNNVm0644P9vAZy9IKTEOI2McGROs6hoh8mi95uW6ozCz6sNlzjDP/p7ScFl8bTUbvaoxVltaJhGci9l3y93Tc2KnU358xmHQ6ltyfNWW5gN5b0iapzZxJwAlh9Hl8nTNgjC4uuJ8UzOMiX4MjClpwqoFLTVV+q4puIoWzfUxFMWAUQJbtu2K/b5DrFJ82aw1SkvOJcV0SkIY1VwehpGmron5+SSEszMFYZY+bvmZLNpRJG++4xBx3k3mo24ClpJG4XPZTVl75d8fJPz9540x/wzwt4H/rmhj5S8Df2vxmWWLy+PWlz936qBm0fqy1OAkdPdvqhq0SSFjGPNEz9ppcZTZLMv/nYoxHed36Wv5J0xMm5L5xUsmxXJiS/yBfBzvtP1mCJGmdvjK0njt5zTEyC4IawzbtuFiu+b6bp9TacgMO0Yh2UWqigGs8/iqpa5qjHMTyuWtU+49gdLR0Iou1jgG4qBmXYijCp7JnzEWI4mYRkLwkxaHQkySlb8BSRHvtGOGt5ZVLYzBTQCCzYIZJSLEqRZpGEfAMIyRZDR/wLuauvJAmlqbluTWFCNzq9A53ar4Sb6q6fqeuq5AhH7oGfqOiwdnLNujHYZD1ILRciNlkvKmUtM4Ki2a9UrZvFwPc9D91S7Acny/dFz/CvDDwM8A3wX+59/nce4NEflLIvKzIvKz2kJlrg2ZIFFradsVm82WpllNoMJ8kDTtQMaowz8z4sjx+WazcYK603Se0mEvSspO7SGiJDAteucc63XLl957l4cXZ9kx9jRtRVM1GGOVRmvQBNOLzQrrleNtFCGkNJkT00O1GkC0xqswuUxl7JyWaVhHa5UnvHGKuFWuUiYjQTeAJKRs46SUFrxummQ6+YTZ5PGuRJPKf7pQKutofMW6bti2DTWGs6Zh09Rs6ppt3bCua1Z1TVvpv9KiRo/gOF9vWLetklQ6rcGae1UZYE4WZQpZCH3f4Z1nv+/Q3DrNfLm9u+NsvdHaKA43uYLilC6GMLPCMqGpWttl7CIckn1S52aO+OPuj8fj+9JIIjKxNRpj/nfA/y3/+arWl7zm9VeOok3IkPaSzLH4KApZrgEIYWQYcqn1xB9tST4dBk45BXcbBAt5URljWLcNvvKZLD5ico2/KcAGTF2yRQTvDKumoqnVtKybFovVTC5rsM7T0VCPe8RYGquIVS4CmALAZZdc7pDCvDM657BiSEZNqJS1QoGPRbSyVNdRaXVfhCroxhTAmQpxkgEal2NLMzm/pKSIoChdsjPqu2mNVsWqbjFOZ64kqMZkCLaEDhLWlmRS3d2dt1ivz9E7p1nis4t2YGprVrcKfd91VFXF3e5uuseuH3jx8jlnmw3WQsQcPN/pd2awQCQp0OGU61CBGZOz5HVuU27U4F7BUHRqfF+CZHL/2PznPwH8Uv7954F/3Rjzv0DBhh8F/kOdRX7UGPNNVID+LPBPv/lEZC6zw2K9xXVkssdRedaqmvV6nfuVFgJIGMbZZIP7k1J2Q2uKeQXGaGpP4VKDwzyuKfvBzAVpGjnXIr+r2z0hZa5spwibqSqirXj58jO2dUUcdtrmJBWyfrIZFkkpAwGiYd6CWBZ03NoqO/RjTn9aONgpIWInjaQbgWS+B0PpBIHItBMXIayco/Vqnu6HhJcSwZNpYS3J753VGF4qGjVaMAM2JiUXybE17d4BVdNS1w1jVIg59bJICj1E3YyZEdNhGFmtVlxdX+Fzp/MYIy+vdnz9Kw/zpjgHWDO/0+FyMoYYNH9yGIesmZNuxt5zfn7G48fvKrLoPV2va2i73dKuXt9D9vttffknjTE/gwr7t4H/Vl5cv2yM+TdRECEA/4Lkem5jzJ9H+8Y64C+LyC+/+dzaywhUOy1t5+N/BaUr6Sfee6pazRstJNNjHsOYk4kmiZSWAPvinAB594Ki2WRaTMt0FiXLX2Gd5+Jqp+aB0SzoKEK93hBXW1a11gp95/lLUlIqrhiFMWljaTXysu2eHag0BsRXpGmDAWM81s6pVJOfEc0c70IYhsAYwtTbyefrr1yds6+ZtF3jPK0bucvayFg3ccq5/E9JMwulmMOKgMkd/YLgbcLagHV2aqlpTKRerVhvtrx4+ZJms8Lu3QJxXTj3xYg2imqOIbD1Gj8qwdEYIy8vr/nJH/tRnLWTGTiPYi4eWiDkn1qyYdisV1hr+dKX3ufHfuxHlRNeEi9eXvLw4SPqusK+wQn6fltf/h9e8/l/GfiXT7z+V9Ees9/zMMbQNNUEP6c0x4cOC6901zTZfCjJi+pjZBt4AWPeD8BOf03oEhjELAsEC+XxjNAtG1KpGSUYgZvrW273A5LL30NSQfOV18VcN+z6G778/jucf/cZd/tOUcl8rii6cKpMhq+IZWaQTQmxy517jrtI1h5GIKWST6atbvbdSEnIdLmMwJhcgFjXyt9nlfSxqSvqvsMZ/S65vF1jWYUUP89TmVeU+0CMwcQIuTWnFMZIoylUSaBtG3a7js35RdZu92uJpknNP8I44r3TjhHFBEuJq8vLw6piBCkEYDnptTzniZY4qY5NKdF1g26IzjGGkRBGvPfsdnvW6xVtq91F3gQ2vNWZDcZoS8iS4iNo+0lk1hzWam+e/I08b5lVxiQQMxELLiHvcvxDaFOmRZmSIFG5EkQEEkiUSUgnkCKpzW3Q9dI2Nftu5Ppup4tQcvxD1GR7+OACscLL33pC9exFTuaEkOmXQ0wknwho6beXREUx8UoIYMTkVCIFC+brlxRzYyxtVxNToh+Um6+p6ikcICiTq1+1tOsNpuuQvqOyQiOJZr+nrixDyHGmvOMXrbV8JrIQZMoyLhsQzBXGGPb9yGa1Iiah8TWGNNEDzwHTRYl3fq5jGHKD6giiZmMIkcvLK0pLnwIa6WZJ9ovntTRpo6TmcgiBbhi42+/YbreEMHJ9c8UwBnxVsdluFxvr68dbLUhQOBdycBQzoSgGcjWppXKH8HUo6B6ZYCRXvOr79+NIrxppQbIvhrmTg5jpGR+XJBtRvuoiQDFrMmNKG0WLXZ8jzZZPn73gbt8RUlS+g6Rl3k3MrTqLEx8TyeQYWrCAx9o5yFzSbUhCyohcicNI1m7rVZMj+laFyYLxqkFs7fGx0s0DqJLgfIPzPTbp96H4Ok5Bl+U9Y5BsUs4l5xnRFGUKCmIYElztOjAVIajZZ63FiSCZOEW1g4UwNyAzJjPA5ue1u72dSt2vb28Qoxn3y+erFsW8eU7dRnJjAcmAlKTAMIw8eHjBlz78kEfvvEvdrFCf+XuQoDzeakEymLkdo4iSvOcIt0Wj9/r+vFMCuJSm3Lppd1rA1mWcMvPKzmXNzPVWhK00vSq75nIHLe0wNVs9ZtYaNS0SEDJaljIw0Vw8wqY9EeG2HwmS6MdIU0UG76izIIQYSc5rUHIIk4YNIUz+Xsi5fOTvxNy9oiTNWqdVtzGluSYnCWPUfL1q3xHGqIvVeMjVtxiPMNINuthY61MhN6WWrMFlQu00ABuDpvQkyWlMIgxZq93t92zOVwjkLhYFiY2HCFm2FqaULpUirLHc3N7kzyXu9jv6YZh8v+V35hIZm4+vG1yYfEVNgRpzV77JbTDq/8W8bNSgmMviT423WpBgdoBnSJQp4j5rFQUblnUumk6iv7sMv74JwlxqluLoTlS4lGZaWUBgEqqDa7OGu13H7d0dVbvWhZYFO2bTI6WIXbX464rzzZpPX94SSYSU0/kzkhdCUL7qGLXZWG5HYnMGd8rXG0LOSicuWsBkd12Un2BMkS5Eun7EOEvlHV1K9DEy5iK3GDRRdj8ErvuePiRFG2Pitut5EEfG6GcofkGqWEr/Q9DyBI1dqfCOSc297fYMd31JZdV/qqsK7702FGDu/Vt6Ky8ejLIU5U0rhJCzEzSWNIQRX80Ntie/VQSmlpyzxhvHUa2cqJTX+/2Ort/naw95DhNjGLm6uuWT737CRx999Nq181YLkjGz6VR2K0WNShynZIbrKOaMLl5d+CEEhqFnht8ONdEh3LpE7DR4l1LhP5tjMssxmQyiyNi+V5rf612HDyXNRcA6oqTczwlsu2JsNqSbS0opQUgqTGPUFBxvHZX4CXU0GWZm6AnWYoybqkhFQok/5qvX3bt08t6HwLPrPd958oKrfceDzYof/cpjHvaBblBARtD2MVd3Oz59ecvTyyvWjQZfu1F9rSqDH9ZqOlSJ10mCMYZMjqJCHFOkHyN9UG13cfEO8vEVoN3wqnpua2qdUf+3PJOlH5sfkbKu1jhrWbUt11c3dN3A3e2dNgVbPs9sChqnpn1c5NVJBn9S0rqkq+trnjx5yi/+0i/zK7/xLUJM7HZ7rq9vuXx5yTDMLFSvGm+1IBW0pAiStWaO65TYT96hxhDY7ztur29yP1WDydRPfdfrWmahcY4ABz3kQmOZ3BRLyrnmmMRyx9O0m4JQCLddz0fPXnC962HXaxxrUNMhjVFT0pw2ke7XF1xf3ykalh3ymISQYBgTlUvUIWJNFiQ0JQhriM5iiDNKmedLu+6ZTLurgVXvHbdXN4gkHp2t6ceRF1e3fMvCN770mIhlHDXmlYCXux2fvniJMYa6igxjT0qRkHJDagHC7AdZRXymOp8YFTTpx5EhBsbcZXx7cYH1SnPcVD4XN5ryqDWmtXi2BXFTg8Iw7DsqX3G32ysDr1HWoZu7OyrnWTIGTL9KIkSIacQazYBJUSmhNcsjcH1zy6/+xm/zK7/x24o6GmXZrZzHGTN1+3vdeKsFaUaIMjVWTnef/Z0cw0kQhoHKWd65uCCGETctPE/sA88mNtD7GujYV4JsyuXcOpWlONNLIYefzUCEkIgpEKKymKZkCCIMMWqLlFwHE4La6WOC5tFj6ptrGDWAO4qaZtFZxhgZwpi7MCg6mVKceLennqcZnhbL1IXcWQeifWofPzrn2dU1+yHw3oMNX3nvnH4MvLzteHp1Rx8FI4aIECTShUBd6cYRwoipjDZoToExOMRmNDPPQ2n4nUTzEceommg/DgTRal5rK9bbM5q6QkSph9dVw8XmjCf981loKACGPgU1+SyRSD+OVHXNbr+jbbVpQoiRq+sb7fLuHM54jSeSLQkU6XPOK0hUQiDOTv5vTBGTM83JMH+M2gamoJOfc/ibXP2YBcnPDJhFW2nDLkPja+WFHkcswpmv8GgDsL3rcGUCl8J0YnKWQqUO8/xeKVHQr85Jlfo96MNIiIHdXrjbdRjjVSMOSrzS9R0mw67lHFXlNeds0OtR8y4SoiVEyxgT3qnPQdCd3ZqUA4Ta59Vbi3UWX1dKHwbaDAywNnG+rvmJH/oyz17eEGNkvVoRMFycDzy7vuPqriNE7aJhjFAZC94x9gPbdcP5qqFy2gViCCM4nwPA85xJTkrV9jKR/aipWiFBSgbXrFi1azbrFX0/5ORew9c+fMyzy5eEmLs/LK2FbKKWBd+PA6tVizHC2XbDZ58+JcbI9e1t1m7Z/C85kiZ3GczazU7PSwPXIZufoNpNWwWpdjUoh1+T/bjPtSABlK7V1swMPfq6/lyvWowd2e07drc7nEQu6pozMay8x2F4YuuD1PgiLEtilGO/yWCmdpr6OJlKKgqzTOnwJiIY0U4O13d7RZbueryLSBS6fUcIkX3f4Ww+ry0kIHYigjTZV1KYXUgkYoiMZsSktABdgqKY2Vdp6ppaPMYISIXxpdQjIkk7RzTO8d7Dc4YQqCpPEkNTVygBj7AfRiViQf2+3lhWtuZi1WjHPyuMccBawyAlry7mrHjtiC15bseoiOAQI0Nu+rxeb6mbhrqq2e3vlOcvjvzUj/4Q3/roCc8vL6feruU5zfB33rRCpK4qxqD+oMldNPq+Y9U2i+RXWIJCc2C3QPP6PSXTl6kzvWByYV/KvYRV4xXi/teNz4EgLVrb54yDQqKv7xuqPtLte1wMXNQ1W19Rqe2hwEOMREpX88O4z3Ei63L0wzAVnEk61CQHsHcWpqEfeXp5hzOOruvZbtaAcHu7o87NtWxVIzFMZkaIEdesiVcvdIdPBWrX6tpgEyYEUlTwYbruLHi+8njx2m82lvvQLuQpJaUzzr5NioLzTitv8z2e1RXjpia01QRohBjYd9CPvRbkJWV7jcEySJ/z6xwxKLOtiEw1T9Y6+uIfhexXJcE3SqfV1jX98JK6rokh8tM/8eP86m9/xN/55dvJuihCVFC4cs8pJU0CTioMJZxwd7vXfrnWzjmWisAwBXZl3ixL2UtKQX3p/DyldGdEMNZrxwyJdNL9J4CzIY9CMjir2Lm83IhoUmPtqXOMhFxpOSVl5jy6JbjAUgMd+0gijOOgGQK6wqcyjvyJg/iSMYYQIjc3OzUfjGG1ahGEIYzUoiQsvlJi/hgjttIYkOTNIiVBrGQ4X4V3DBGcBjMkHsavnHU47xhDr5F643A2YI0Sn8QoB9S/+bKRDJC0dUUMkdSMGl+yWk/VhxFvBHKJxTiO9APEpBnuldfq3ZRk6paues0gJjKkRD8mhqjd1pMIm+05YVRt2Pc9VVVxfXPHqm3543/kH+JXv/Vtuq6nz8+6xAbLs5hM6Py+gWkjK5kJcFRZGxOSwyBLdDcfdA7cLp4jY2mAZkhOM1JE5gyJV423WpA0TiPqD5Sg3GI+CrhTOs0hpXPEZLxPjYWX+VLHmujUKH7ZJHgiuRVIycQ+KmlOCWNyMDRrLhEtr06ptGWJDEFRL2OVltcmSE2rgpLTa2J+eFESTmyupdJ2mCXoqhepxwwp4oNei7NKNumsVR4E0dQOYww45c+z+fxEwQOt8+BNBgG0aVgS3Qz6biAmoR8CY6av8s6yaWqdg6RBWH08hoTGjYYxMQqEqJtZXWtrmbqq6YeBuqr53SdPuLq54Ye/+iFffv9dXl7fTJbG8vEcLHTdFrJZbkEiXd/z6OHDRTwvr4GcJlQaqenBDFAorHMHc5GpV3BKmc7ZRwaGicTmTajd91vY9wMZKSbubvf0dx1hGO8F3ICMHpkJAi/1NHMajzrJIuUh5MZSy5iROcWmmev6c+mzZPTHTO8e7ZSLXauYIWnZ2Dfb5N2CiCWlhPWOer3GNqtsuzPtnnoMUfhWCkZ2mPYyRkXxCjPOGCP9EBhGjUmV6wQ1u6Zjl5SeGDGScCgL6rpptXDQanf0zbqlqjyCJUSNce2HoPTJ45BLRQrXnkz3GUWIUqqba5yriOOAQbjb7bHO8dnzS37p175F0zT8oa9/lcrlBGXdQllufuX5KF+gPm+X6+mvrm+mZzWlGS0YVOcsh0SSSExjti7m+VY0NUyFnEvrw5g/oMK+H9SIKXFzs6OvDG3np5qjqmmoG81ads6SfN7djc01PUI0UQUsBoYwZkdTj2utxcrhIoNjTbUMDmr2RMr0xRR7PT8o/d1Px5i0WE6fEQRMRt5Q9M9nxtHKe7qxw1Q1JnTa/jJGxNnMFJSwYrJZVUzANMWwkDQ5+wYDgzaAbtFycjGl2jeXTmSKKZPjcnVdqcBnVIucOVI5pz2EDMSk6GHGu9DTaJBZg8Ta9dzkOS6xsCgq7HVm5hn6AWst+31H027ou4G/92u/wR/58W/yza9+yLppuJ4Cq2kqDVk+n5hTiWKM+MrT7YWb2zu6XFIP2tmipEgtR9FCIpqB4Zyn7wba9YaqqtQPEkXvvK9IoiSan3vULsbI0+cvcMbR1hWrxlNXnqpSVeudZ7Va0VQeW9UMMXHb94h3tLkU2xeQQiN9cxBxof4nM2AxipBNGcNSoFQdxzl6MtnSc7A25l0tJI32h6h8BgrPahJqilpDRbsm7W/U8RWtR3LYLHzaoKty5ICrRXIxYMimlfOOGAMxCqu6mkxDn/P0wJC6QBhiNg9F24nWNU1d63yEwPXdHTf9QB97cGBzFkmiFCiU+YMhxEzWonVTJel1TFo6XzIVmnajPA6jXl/I8bm+7/j7v/FtfvFXfosf/+Gvcr7d8N2nTyfNcww2zEhp1khWSUwKl3nZ1EocaonQlmei2krnpGnUX0tlo6mqCaW12aRzFqU0+1z7SEnY9wMGJVLv9o6qcpxvW4VBiXS7AUkRQmBtLY1vlHTE5PaQBn3IC84zydS1PnfzLoHE6X3U5i3E++XB2tLPisPEVX2AkZQsIczgQ4xCCJFSZxDGTAgiEWuYkk8xBqqGmEs/omgmeJWzrKMIQ0w04jO/uGZwpyTaybwfSQZeXHVU3vNg09LVAYMhZr67cYhaEZqUKccbTbVqvPI8WGM0DpaEPgaGNGqg1Vv1rSrNyi4Z6VUpi0eoKqis1baaSRhEmP+zuGatGiOqhhjHga7bIyK8vLrm7/zSr/OTP/p1PvzgHX7tt78zBeLL8yhmVdd1jMOAc167dDjN+xvGkVLTVDYzfTaHbECSaaQLtF3XNV3faS6kmClEoqEO3XBwhi4tuzOeHm+3IMEUWxjTQJ+dvu2mwXmvaTekHIlG0+9RB1KDahqMi5l9cUo2laUJRm5BmnsSZc2F5Bqg7AMIWdhY8CjIEfVT9mmsySn8+WQhBC1/H0ds75DkMAhtM6NTUcg75YC1Qogj0RnqyiOpABWaceCcy8HhbGaizn0XhJe3Oz59ek3jMucAFifKK+5yfydrEpWNVN6wZ5wC3ENSAQ4o60/xcQZJJKPZAJLN1YfbJncz1963zhoi0IXImASM5u6Rzci+HyCJAkMidEOHNbDf7/mt3/mIjz59wde+9N4ic3v2hWbB0GfgfcUwDNS1xnfGEBE0FepAi8FMV03Ok1igtk3b0vUdxEiyhq7r1DsTg5ghWz2aOf8Gy+7tFiSyQ14mRPLi9pkDO4Sk8RBJiDiSMQRJpKNa/YLuSSoOq7Z3iVETXMvuJwWgyFpHzZBSmSoTG5FZPOjlLhhGLZWwJYfMLtJd0ix04xin1o7aeR0SBmnWDLsBSy4/SJrVXTrKjZnmylRmqm2svHZEXzWGxw/OMFievdhx87Kn9p7aWFZVgzcKNDhTNpaEGKHPuXHKuaDpTMo8aqico48a1L0bB26HPc4ZHp43mY3HYS1UlWafdDEpo2wmr0skcDV9N6gmsIYxpFzno7v8OPZcXd/y7//dv8fP/dSPsV017LrdJBAxm8PAlBvXNA3juGO9ycm+IXLXaR5eWNRj6fNRa6Rw1y2RW2uUrDLGiPXK1052GUoirXVKff25BhsElE7JWhy524PRLuWr3KGBPFkhhCnju/hDBfbmCO5eZmyXTIkp4BbT5BONubmxVskWAkWTHfuja1VPFuVmSFjjslabfYoySvqKolIlOJhIxmGwNKjDPkQNGDbOY5xjGAN7q/RdoTCqWstmpRnRde763jw+Y79eYQZY+5azdk1lHI1XqqymUXrhcRi53nfs+j1jFmzvK00REq3HeTnseb6748n+mnVwbLaed85bmsrr3IrkjuPCIBBF/SV9AhZbtYp4RjUxDbBqapxR4Ohu19H3Pb/x27/Lf+YP/xjvPXrAi6srRubM/4m80pgpm2TqMZuFbbfbafB96VPpFOtzyOleNgfkvfcTgf4wjpyt11hvc2DXKTqY8xlLWfvrxlstSMXxKzEbfQgNRmC320+E6gBVXUM/6G7EISK3BAaKEJWJUa2kWeWKziwrTNWom6/nsIxiSQu2PM8yH6/ch0j5OUfbC2Su/VRLvU1FkEG7Noim+pvchkWiEsAXHr0QI96o+VFbP7XMFCNYr+GDPg1UowdfcdZsaNuW1XqNrzxxHFmt9ozjJrO25j6u/UAcR/ZxYBdNDsJatnXFdl2xbqppsYEKfR8SQxANO1ijeWzicKIl4d5luEI0u6Fdr7nb7bm960gpcHt3y83tjq9/+D7f+ugTYhJWK+2JNGREbgIbvMv+zlzEF4bcsUJ2B/N+CGMfLq+6rnHe0e07Hr33mEJaCRbrwXqHSWBSmioHXjXeakGaalFEoHDKGYghIbGk7Oju6Z1lBRjvSVFITnIkAtq2vheMnbrCZQRNwbCZo07fy6XizDEMHUq5leKM0BURMllLihTYe6EVczmBBmTNZLMXYnfjHHsMLuk5LFo2n1JSXrnsQCvi5qZdVzLqqE660UwDC+KEbhjo94HWVQwp8DBc8E5dUQHeGFzTEuiRQclPUlSz83bYcTP2XPY7LocdgUhVaWvOwolh0EDsMEa6rEHJjKUBQzIeiREbA9LtJ/g6JeHmbse+G7LppNRi17uOr3/5S8D/dzLtjDnsBKEbl973Mk7U9SPbVcvL66s55pR93SmsIUKhY9aYUeC9x+/x8ccf8/zpMx6++w7Pnr/I4Q7AaAVAzCxOrxtvtSCBwq82Q9gCdN1AU1c0rkLDHxGSMACm8kjhPJOENRof2W43VJVfPBxzoJnKxHvPQTZDygmXqVAWZ+ABsl+FIkHABIwr4budtA1o2g2i2s9X+tOKYeg7hn6fQQoFJOIYscZQGagFxiTsx4HGVTgyDB/nwGO2EBWGFi2vaGsV633WqGFI3MWB/jZw23Xc7PcaeHXKahRj0HYrGTbvh4HL/Z7n/Y6r2NHJiHFQ144657SlXLaektBHoU8goqBFQhiTJRjBSkLCCF7je1e3t3TDQBTLzd0d52dbfuLHfkjRTWP46ofv47I5pb6aRcRPfpIxpa4sl1jklKCu77jYPMRaR8wpQra4Aws0VrJJmJLl6dNn/PAP/zA3N7e8ePqMm8srjZ1Zp/9MsToO442nxlstSFVd8eWvfokUAy9fXjP06gf13UjwYTGhFiNCYyxdSHknj5p9naBzboJeTyWqlr9DKNwP2TYn82BnkAOYEhyrSltKjqlkipLNtaSMQzZnrBtDValGvLu7nUgJh7HPibAaVyql1q6yXIvQIrTWEBKElHAErPOUjtwhCFVlKE25giRMMmjmqmomVxmSUd7wNCZiSNzFnuHukravqTP3dgjK0e28Rwzshp6rcc+dBP5/7b1rrG1Zdtf3G3Outfbe53HPvee+6tn16Fe5bffLTwwhCIIx/tIgJYhECuahKB+wEj4gxUCkkAARRAFEJIJkIksGEYwTQFjIwRiwQRax2213t/tFdVdXdT1u3fe95733XmvNOfNhjDnX2vs+quwqd52O7iydOvfsx3rMNcccY/zHGP8RJOIM+p7WSpmcEqUkvouJZYz0QbNJEBX+W4fH+HrK2XrK5saEpqqYL5aIdzz79Pv40ksvE2PiO174EH/oD/wert28zYWdHaZTE/BWS1KWy+UoW2FAOZ0FpqtK64+Wbces2mBrZ4v923vkTO+c91iyQVKEoP2h5vM5r772Gs889yyvv/4aB/sHxlwbim81nPfh41QLUlVVnLuwS7dYcPfO4YC6iBgFlqJQml8XOepaUow0togVZQvs9V1p5zHmdchj0Ew5cDe0vkwpm2VDgHBoy+gQGXLf9LUhmJtLQLz3bG5tsGx7nNOcrr7rrYTck0K0rnCOynvCMnAcWibUuMrRi8OTcChkLSiHQZDM26fX3YZAlYyDzgkb04ZFF2h7BWmiUyFfho5l6HFBaIyquA+J3uDflsCJi+DUF3Hi2JhoZkmIxiqUsJhT1LKKqNB4TJFre8ccdolnLlzk7PY2F3bPMl/OOTw5IYaeK29eYf9gn6efeoLv/8THuHTuHBd2zjKZ1Ny5c8cWfiL0vYUONHdRfSaz9NcIYJZdS+0qLj/5GCcnx3SLThNyY7S8RilOkmaFaGOBw4MDvv7SSzR1zaRpmFvNmMbquGetPHCtvsO1/ts6xFalbioD4YmaN5Z+I9poOFNvHYWE6xN9WNKGQE9kadnR62OchTCOX2g4djXbO0b1WTIkn/uNjo+jPouifCI6tbvnznE8n/Pcs89w9dpVQujzzVHXWoxIjAU00JwxLdleevCuIvURklfzqO802Oy8BUMrYrRMeImmvXShOe/Y8A7X9cxToE+QnLV4MeFfJEXBekkEMRIUEsYAgfjERlNTVwq5tyFY8mqiS4mFZTdEtLTiaNHxua9fIcTI1Ru3mTSNMpeScF4bm01mE373930X3/uJT/DU45fY3Njg6vVbnN3Z4qsvvaqL2WFtoIYQQ9u2RXBymlDuHxv7QL9ccvnCZa7vXufkjnZcxwLkVEOA15m/3XdaWhJC4DhZqINkjQscztXWKfBbXCNpCn/Lsu1omoambpjP5yvZz3U1QdwG3jWW5awGX50iddfRx572+LpOhl8VHuAeRzb/1vDRQK6ShSyRSVYUcBBz8nMVZ+6cAbBctjz14Q/ya5/7PLdu3+bypQsKikwmVN4zXyxIouysMShkfnR8SEjKIb5YtnhJJO81DoUH70geLSk3LjjnXQkVIA4lBRoK4qbWQ7aLsFz2LNFEuIjGjfoumGOezKyFiRcLRlZULie6KstRjPrdLmg1b0yJJJEQEl96/SbztuejL3yI7/3oR3jy8Ys89fhjzGYNTTUlimfZGbNqShzP51y7eYe+j7zx+jU+96WvAUmFL2rMza0wywKi+XAxRmvSrOGAo/kJcvUWVdReUqFVV8A5Z+yzps4s1AWrLFBuFHciJmLqcVQEGcIkDxpvh/v7abRb32XUavnxlNLfEpFd4B8Bz6L8338kpXRXVBf+LeCHgRPgj6eUft2O9SPAf2+H/ssppZ982LlzkHO57IpJNa5JyombOjfGp5CUUFAiQOR4uUfXd6taY81HGiN1eXIzt1leYJo7N4AN2axQ3gjzk8wKyLGHZdvS9YFvf+EFvvIfXmR7a8b53bP2/cjJfM58saSzAjk9lRbwvfD4OaabE5ZWpevE43otTRIg4ohixXy2wypnnpVsG3qW52viPZMEk6piXvW0faDtI8G0euPFsiagzrTEzoCNlOm1NNM8jBKAYlKYHoRXb+7x8tWbVJXnxu27HJ3MdYMAzp45Q+UrnKtZ9lpNfO36NX7h3/8K+3uH/MC3fwe//Pn/wN7BERsbG6R+TvBQW4Ox5XI5bGYpMWkmdF03CFLfs7845qybUPeOxte0slzZHHOzuCA9KeViUUFhiVGlNNk+1+ZpWCn/w8bb0Ug92kjs10VkG/g1Efl54I8D/zql9FdF5MeAHwP+O+APol0oPog2E/s7wPeZ4P0PwHejl/lrou0v7z7oxDEk5gstlRbv8aJxFsHRBeumFzv69kCRMiBGE5okdP2ywNohjPilFY4p5xkDDuU1IBgpfxp9Ju9eIUbL4q4sQDgcTyxNKQTHl77yIt/z8e/ke77r49y+c4vXr1zViHpUjTOdNJw7u8Nsqkyo06ahbRc0tMrFgND1CScaw6qdQACRHu8SgvU8dRRmUIeCHBre1dSgEm+rtD1nF5Qmqw81JKtjqq2UX4xqOEa6PrDoO5I1tA7WHaM0R0OB+sN5x4uv34Sk17G3f8AvfPqzfOaLL3LuzBbf+aEP8NEXnmd7+wwvv36Vl159ndffvMrENzyxcZa7n71KeP2Q9z3zATa7Y1567cuaReKE0K3m0aUUaSY1y+ViZWOd9z0SE1NxzO1xjAshyzOOCZwVPTrNx7xfzFG1ZlQ2prfwld4Oif5VtJkYKaVDEfkK2oXvU8DvsY/9JPCLqCB9Cvh7Sa/sl0XkrIg8bp/9+ZTSHbvInwd+CPiHDzp31/fcvHW7/J0LuzQUY4iMlUjIKCazorFGUHGeHAHjmBlNGKsQZ0qJw6MjQh/K3znPznDtIkjZLChgQ0qEXnvJzE+Ef//pX+eJxy5w8eIuTTOhqSsOj0+KAB4cHnNweIRYSlBKiad3N7m0MaHxnraPtOaXSEpMa4/zFV0w1tkgJVtZAJE0CJEJUp0RP2BaqYkaQipzhwWJxWt8KyTl1svk/VrNM8xPRuxSctSTCbPts3zie7doKkcfFvSx58a1I67fus3ewRF394+4fWuf82d3ePHV17l5d4/HdnZ44ezjyM05169coe3mHFy9xs1wwOJkQbIiSDUdUdoyIMXEZDLlbr9XNjIRoe179volU1+xMZlxdHK88tzz7xJAZ1yHtmryw2D2O4tHPWz8pnwk0V6ynwB+Bbichh5J11DTD1TI1ttcPvmQ19fPUVpfKjIzEgL9gDqirBZ9rWcZAEUL5d14jK6NzndfjQRYC3nLDUtDUC5RrLiSsjI+9hDjCXSypI89L792ha+/+ka5Zm8+SIza0MuhJmqOC3Whx6MayDeOk1ad/PxARTRu5kNEJODjgGaWRsrWdbyyfkcKs7uiuXrLsRMsydaVSaZLinj1XnPOUjcIW0rQp1QKB6fndvnkD/wOZSvtOm7d/Aa3b99kftJy9cqBZp93Pf/v/hfYbBqO5ksq57m8EK5efYm4iLR94JiWazdvcbM/VJ+ri4W5Kc9b3nzqpil9asfQ9mFo2aai8YJ3FcGQyqGDhliwtYgOYykZa6YxL0fgXUoREpEt4B8DfyaldDBWgSmlJDmg8Q5HSunHgR8HmE5naTrbNDjSdhOro8/VkDkgl5Mbs78kAk2uy4m5K8RowY80y4OU9np+VQmADtd6D5yehTLngxHA5SAtij5677XxsiQmjTCbeBqL1uf7rM0mr50uXtdUHM5b2j5o1yTRbnigKUYxaEfyfD8iom0yqxrvoXJCbYFGbI68c1ojlSDHZ/I9EAeGU3tVTWfR8vE+WoYIicnmNrOtHSJCH06YbZ7lfAhc2dhTqFy0AvUE2BOF7mupuNFDH+e4qGGGAzoOw5xF1xJSVOZWYwwq8y/Csl0OXUAsvJADpz3W8EyEylW09Igk2/w0fSmvn5KuBffdSPP5APq2u+f98XhbgiQiNSpE/yCl9E/s5etinfvMdLthrz+o/eUVBlMwv/6LDzuv947nnn2Gxy5fYGNjwtb2Bmd3zrAxm4G1OTmzvc3x0QFX3niDr379G9q9z2nXhA++//18/ktfYv/omFs3bjNfLIadTWTlAa1P6v1AiUGQNFo3FqR1H2pcrxRSKigaSZNBXUYXK6i9Bn4zAWbllJs7xYR3CtfWlQA1hydLFl2HSKVmZA1OOpwXfO0NeqeYm5Vzejzv8a4eYmY5Y8OQyKGdp/43+HwCKASc6AnRyspH5JlPPv9hPvHJ7+Ezn/kMr7xxk5nXDWPSaMOuxXI5oGKo79umjm+EwKFrmJmjf5g69sKcZW8NsG0+4+jZ5OtXAdXnWNWVlqg4IThhkRLJAvBOzOfLccK0qnVW0odYNQGLposR93AO/beF2gnaWOwrKaW/MXrrZ4AfAf6q/f5no9d/VER+CgUb9k3Yfg74n0XknH3uB4E/97BzXzh/jh/7sz/K4089T0pwZneXC7tnCaHnzu07XHn9VSR5+rjkwx94k63Zv6dfdMRefaOPf9d385nPf579vUNr+1juqfyUjIWREI0FqQhQ/hvIyiylxM72GY5Pju85dh4rWk1Uk0hMJGux4iS3VLTvJEZaTx/upPJ4EWrncAIHJ0uWbY+mUCQFJfoe11ZMKkdPpHKx+BY550/9IbJsECN0fdQs55R1jijdcTFlrGFY0lKPLkaNJeXyEnFMzp6n37rIXue4tXeXWb2gjj09xj4bNDZDgZdVa7TtnIVbMhGN8SxSz6Lvh8YFBlUTE1FSKXeJUZmNCh2a90jotUAzRUKyDhdroY7yfMjPMz/7geasILhr62G9bH19vB2N9DuB/xL4goh8zl7786gA/bSI/CngVeCP2Hs/i0LfL6Hw95+wC7ojIn8J+FX73P+UgYcHjZP5gv/rn/4LLl68SNNMeerp9/H7/5P/mMXJAbdv3eLw8IDFYs4v/OK/Y3//AEmBFCOVF7a2Znz+P3yBk0WLUnlpZnceZedJq87oeOTo+braz8ISYmDvYN+c9fvfw8p3U9JsBAnKHhQdXRuYeNEufCgdswIpqXRnqKy3a11ptwvvhP2jBXMLCwxnWCoQk2ocAe8CfaVl6n0fNCwgqi0AbThcYbVe2f/SgGxEm59FKxRNCePzNoKQpKQwMQb+5T/5af7lL/4St968wuHBHZz0OIHloiWh7K0pJnoZelWJWQTLGOkc9KGzBmmrG9sw3xRXRgEeh/fKZOu9V767Yr6v+s3jZ5ZNPHvC5Rm5UcBfnBuAlZi//3DP5e2gdr/Eg92I33efzyfgTz/gWD8B/MRbnTOPru24evUN9vbusLW1zWKxz/U3X6b2jtt399nf38fXFS+9/DJ929O2LSkpzfHTT1+iaZWxp6q0hWRVDb1WRSCmESQ+0jw5P6uYeyJlYnM6/UqwljR6ODrWNVze6TQ7whAwp4t19YtqyjW1sx1VY1WV1zaVta81Y9oJd48WFmey2FLS4sSMyDlx1FWDpEhKvbUAFZwzFlKREbGHpv+EFFgurSdS12u/o2hFhTHQ99m/yLn1iVdfeomXrv0a2E6fcxVTsrIV3CDApskzlUqORTmtblQFuGYV3Bs0F6pKiW92ds5QNzVXXr9KSomqrukXSw5PjkvNVn4WVWWJy6x2Ky9WhGnvmB+lQfkZiX3YONWZDTFG9vbvcnx8xMHBXe7evUHoOqbTGTEm9g72WC6XLBZah9R1HSStLN3YnLBcLjg+OtYWlFAyiou6jsGQPT1fDtrBsHOUOEUanPh17TUWovH7efLH3A7ZNMSSVEMItKEidoGYWvouUDnH2Scv2AejwdKKwNWVpv5U4nHiuXN0wtw2DD2p/i96CjfddNrgo0OprR0uKWIoyYK3mUQT9X/attcmAMn6NMXAsu9ZtJ0RugAmHCTNKazrWjWcd2xubhH6nsPDQzOTrMpZjOEprXdhHypgx/M2NrXETFPQfrLtcklVVZzf3eFydZGrV67rXHvHyXJhWjOV2q0BLfUrkP9AZe3KQ8/mryCaLZbemrfuVAtSH3r2b9/WGElV00yaYUezsuNsL4fSelEh0y9+4cvaq6cPJgx+xZF05tiLDA8128G5S6D3vji6xT9iNYC7DlAA95gnK04tq2lJJ8uWk2VLBi9mk4pnHr/AVlOVPlDZrvdeO6HXtZJA1tZd/M6RY26R/xg1346pQ6TiaLlg2XdMJg3TxpDM6EhiFbaVlO8or0RPFzXvTJzTToLW9ynTE+s1edMyWiD3vmd32dg+y7U33+T48IjenklmPVVIPhXfcjxf498rGmJtrvIQhOVyiffaNPnixS3EONXVmtCGz+rqjDY20zpZM2KaxlnXiWQhCISCWN4PeLrfONWClJJG/xMQ0rzYvDH0oweySoOVx/jm9X1NbHFDsOSeBe6d14xns/3HNF3OJjrXwORFm8dq0utw/dlUXAcg7uWSFmaThheePMf7ds9obMlcr9yBD5TJqBJPNZHS7tE5x96x53i+ICzaYo5tz2YKV1cap1m2gc2NKU1T46MzIv9eG0En9cvavmfZtUrw2HcctS2HiyWLZYskLZkXI5bJtDCx7/jqS1/B+XqgWV6bg8FHGYKb40VeTOq17+WQgQZPg2lQnbuu0zStixfOU1dKsXX3zh1q60yS51XMwcpUBWpgaC2ZNmhQ013MVM1In2omIyz4VmYRCiFyeDwHhiDoqlWVGLva6ybVsHhHoPVIuLIQ3ZP5IBrMrKoKCQM1lDqlct+d863G+g48nNNsfu945sIWH7h4Bu8cXQ+qkExbGjedNw50ze62rglO+7E2lefwZMG87WhDZNFHtrqejUlN7SqOlz3LLrAxa2jqiSKIon2YOkM1u76n7VvaEDhpOw4XC04WLTFq82Srj9Vy8hjpU2JrqpwRCwsvqJl479ysO/4Pmh/cyFgWRzasnFQlPy6R2N7e5IMf+ADPPfM+3v/887z+xhXIDKsObRwtmXVVrEWQMy4GpSmonaOqrYldXWubnMrjXIWvKuu5pGQwn/30Lz/w+Z5qQQJDee5roeZ0HN0bM/E69zxAw3oZIM1B9u41z/Jx8991rdRP+e/iAJt9PRbM+41102AMu4J2/N6Y1HzbUxd48twG2xsTTdK1WxZrPB1CsA4UFEH3vsI7NcG8X5qAVcyXS04WLUfzlmUfOG4rZlaU14XI8XJJU82txNupNkqR3vLpliGwaDuOlwuWVnvkRHvxOrGgpmCB4cTmpObJ82d449YBfcgMtK5kWYigHHQpm3hDA7k8J5WvqKoG8aP3JQtABeKUccnqtmYbUy5fvMSzTz3JbDrju77ro1y8tEtn9APaeKzW2Fxl/mVdDZ3KHeC0GUBd1XjnNRZn5Sca97JNy/li7j9onGpByja4/WWv3bvoM+qU/QwdVpyV8mLW11TQBllKzpXOfKvmV2A+n5fds65r2q5dBRZGWeNjs244X4nzlZFjOVm7PnZ2g+95/+M8sbvF0UlLXVV0i56qLCZzgo2gP1jpQNaStVQ2UWNtCr7yzJfaWuXgZMGx14TYpuvwTmNSla+onDYeaGOw0ojeNFqv2Q9o7MfZYpI1syBEbXp2cWfKma0pJ50n4LQpnMvUZJobOCak907wlV1vpUyv+X2R3LVDtUtdaz5dZYWP4jxV3TCbTJlNGppqwlNPPM7GTCuRc/5dXetGoY0Bauqqpqqcmse+wlW1zkFdUTnsvIrwOtNYTvRevsXbugjOZaxf/17f3bPAgFWzMnZkzeZ1uWEVdixb8Gb/aiOpkX2ONicGBR6m0ylPPHaJazducnR0TAkUJgs0joULzN4e3YPXnTxH9r0TNhrPsxfP8O1PnufC5pQELJa90o+ZAlWm2MEsRTAeCe3EUNVVif2kepQF4JXvzjtoO09rPtC8iyz7DgTrWNFpcNLmJBhKl7Mjshmru7oVHpoD3xu4o9oZJpWjdnBu1tAGmGzMqLa2aJopzaShbhptpFxpB7y6qmnqGt9U1HVtJRaDFqi8U748r0KchUmvR7VGpiw+Pl5wsjjkwx98P7u752hqCxH4oYufM+FwTqxYbzhX5kJXTTakUImoKR3TvT74+jjlgjRGdGAQpCxA5VOsYJdp9c+8UzuXg2+WAeyGPjveCFa8mSSIdm+IMTCbTPgd3/VR7uzd4Stfe4WTkyVEhdtzcNCJt4pO00gup6YEnBcaM1lq57i8OeXJMxW7Gw1b04maXH0AToqQOGtE5mQgf8nzEKNmwDszYZwxKaW6yFXR5ksfmERNS+pibjpmvOQRxMXBd5SBqCWJElPmLol1NWRNp6Rd+RLYgtQapj4FZjXM6oSkBTM34/ylXbbPnmcymdE0E6raUxl/XF3XIGqCidMM9bqqjBrNU3mPq7IW8yBDu5l8k33Xczy/ynK5YHvzCS5fusx0Oi1Ny1ao0kSrqPPzz4msJDea34zw6UR6EdxbCBGcckHy3rGxuVFSOgabmtKAuPK+eFBO39BdjMGWz3ayOEdV6U5WidNd23au8uCcZ1JXTCaeqq64fUcJMc6e2eKxy+f5zm9/gStXrnJ0eMwrr1/BVzXeV2UR1LVXaLrW73vnVMuEjjQ/YkrPualnc1pTVxVNba1M2hbDY0coF0UocrNhQBd6VFRNavUbJpOJlUgPZp42cVYO7z4lfMwtK0fUYkNgCzLwonJpDE4a/PXO4V1lRX6qtZCEE21OXXnPsmvxAk1dAYlwdJebLx3QXnyMy898kDMXdplOJzR1raabV74KX9XGWWH3IVoB2zS1VSJ76qq2Z6/UyMtlS9urRm2aCe2yKylBVVWtJRwPGQswig0iqyiWhTjU2JAS5B4m6sHjVAtS09R88LlnjbxCOwR4Ue3hKkW6au/Nvq7wxpKZe87qA/BllxPv8JVQ+dpKqC2Z09qlJFvIlVeEJ6XEzVs3uXbtBhcvXsB5z+bmJme2tuhDx8c//h00tcHJ3lCeKtNIaSfyxdExV19+mcOrr1H5hlkzY9ooe09TVXhx9DGyQHd29QtyR3LTdDJ6+IZAaQaAWP1Sr35D5ZWApGRhYBneib5PuBA0/ScNRXk6pCSxSszaPmtxnQtvAWSt1tbMBoXDHd5FAyIULau8o8o86ySWd27wxv4d7rz+MruPv49zl59gsrGhhYTO01R9oQXOZSniHF1rrVYEWtcqaOAruhhol51SL6MNrUNUsKRtlyyXtfVP8qPyiUEzraO7QwYJI2her73PZu63siCd2dri9/1H349UlT4ccz69CYbLRPGW4ezMyXROfQkRFUZNjcm2sBgnt8UnhEHj4VYmEuCpp5/mjTdeI0a4eP4iOzs7hVQ9ZzzkB29HIsaeg9s3ufq1r7L/5mukbsmZqmLabDJrcke8pPlhJMQ4AkCRMXNEDPlSrgbvsokzaBy9UAaTB6icY2LUxV6U+CT4SF9FfK/tMCPJmoLp7cYYSV75xvX0idriZt47aufQbPuhDaSmSNn1imj7HPO38gZfWcA4j+7gNjeO7nL3zVfYfeJZzuxeZLa9RZpONT0ppEIq45yWmlR1oq48Xeq1PMJpmXyyGJ+WRjjmy5au7+lDx3I5LwhhnqJk+Ue5mK8Ez6G4ChmIGkb2uVcLQe83TrUgbcxmfOyj3z5Cr7OjmCVh5OaPo9iiXAMiYrX2+ikneUIHXyIPESUTyc60c47NzU3SXmK2scHXX/oGTzz2pCFBo12t7Otqau1dv8b1l17k8MZ1fOrZajzVZEbtHU2l5pw4KZkYuVzBiaOpK3xVYyXAKqCD1I/2xCEwnO9XU5jUf8xJMDQ6X10f8VGRwD4mQlS0TU087b4UAJeEaOUdlZdyXc4kox+Xa48WnWpMrDQeXNIlW4kYYieWo6jXmtpjbr3yZW6+OuHiE+9n97GnmGxtUM8aUqqIKagg2XMayii0kC8agJTNsrryIK4Q6MPQYS8DMHqdmutTsiVE7kFVV8dwj5mq7UHjVAuSOMdkMh3+Xlv461kE5T1kxZ8oQc00mMQ5kKslBvpGzN5WmTTtG5vikOGgxYPOWGh0YbWLBXtXr3H95a+x3L9L4xJbjWNaz2jMt8k/WUhFhHGrxWxGIVC5CtLASCQi1go2mySazuSNsyoDCyBoiii4FHExqukqjjZY1rn5SCFq/9feNEFCjBEoGnqGxY9cma2ILcbxmjJNnrOnFbhxRVMJ4MlZGK5w8IWU6ELHnStfxYtjl/cRFz3t1NFMGqbTCVjPopgG9E3suJFoRXoY4uaVyjqNAAMTokSm9ddevCXjZG03HZBe7vP7W9i0g6JwV3Y/4L5CNA5+Piw/ypBrO57oA0NtfxUqnVAtNR8K+lJCCeEr/cz8+Jhbr73K7VdehvaEaeM5u9Go8Iiy8lTF3Bx3ZGfocSQalEUcTd1o25Ja+/5kv8RXPuMQKjjODUIGq8cVLc/wftS3ySVc5Wg7JeF3XnCVK5zWLqj5UiXdeUvWRZlsvcacYyACVeVWQH49d1sgc+fESC1BcuxMkkLSZj42Xhlj9668yMZ0k+2zl1jsLbh66xVCnHP2wkV2Ll5kc+uMtTn1JSBbivQkC5LQGdm+9pkdgSnY57LfZqBKxMznNZ9p/PfbzVw53YJkSIsua7GXHpxBkH+vJ0GOP3PvxIxy8gDJKQVktS+l1iUB7bJjcRK4c+0Nrr34ZdJyzkZTszmrtfGWV/TPO+2I543ayongMtomiShDfUy+1lndcNfopySmEs8pPtEDnum6kEqdlHnINUZmOXCei/lTHktMNUFLCVIUJDqSG/wPRDV5FwIS88LT0vXeoGhBOSiaxjHOthZAWXh8CTNkEnsR0eP5BKnn+iu/weSFH2Bjawd38UmuvPplXvyVX8JVFdPtHbbOnWNr5xybO2fZ2DrDZGOTqqloplNiVGBmuWgH8y5BsoB8rkJSOrfhWSdbE6vd+FbX19tNBTvdgqTurfoD5ITTASjI6n5YLHH03nrS6vDv8e88hvjUKhuQVocmaj/j8PYxErXX6htf+gJN6phNJkxqYVp7DQRa8LAgXpm1EzFlFwf6ZLsub5/bmFTcPdEbdN5g80z0YuZTVk1JzCSVVSRKRLQRmqfU+YioT1aJmmp90sXmvVDbDq7lEWIkinnjslzEpGXafYbnNYBExVBZGpLGnSorPPSSARmb/5ggROvCYdo1aU2YqvoTrnz913nuO34Xs9kW73vuo1Su4uabX2exf5vF/i1upajNpH2tMamqZjKdUk0mcLzgeHGGo8vn2JhMcZMR9CGQnM6hZK1DVrirtU/j9fCbGadbkGwHzwZecvmGrfN23VBParq2Yz6fF7OjaRrq3BaxVSJ2uNccHE+eQrpowVlSwsF+2XKwv8f1q2/iOqFfRmaTGTQNWzsXae9e1d0u5SrSIcu5+Bn5fGYbRrBYTCzkliCFr2FW14QUaaoGqoQQC5yfEzZ1t9dF4kSFlJEwqdkX6YLVyuVUJmNgrcTTx0TtvVbmpgH0iGnovpCiwd2dgjfZr/RGJpL9s2RqU5M8NaNCLYdcop6sc7rRpjk1p8XappC0yK8/vsW1b3yJpz/03Uxnm7zv/R/FETjZewNnzq0DvHgqSZCWMG+Jx4ktEcKtA77+725w7cJlLr//g1x8+nmmm5tFayaJZT501kdKvrgO95p1b5XVAKddkNDgmhNtP9hY7lQIvT54gdB13Ll9m/39Q/q+p6pqZrOazc0Zk+kE59RPabuuLIQMNoS+p122LE5OODk64uTwkKODfY6PDjk+2Kc/mdP1andvbl/giUtPWr6ZY2PnIsc336A12LcIhNeWmiClk52kUM4ao3b1ztkJplIsS8CzOZmwN5+zNZ0M5qxBvANQYc2YE8TkiOKo3JALppkV4JJqGUmyglQRgpIi2t8paU6ZMvMYdVmycohgG5loyYpLFGRP7N7xev1N5dmoPYP3du+IMZYAlohulCloyYirKpZ3Xuf49hPsPvY8tat4+vmPcvOVBRKWuKQJvxJt3ZtGyT5ubjPT3XqTV66/wRtbv86F5z/E5fd/iO1zu9ptwyyBUc1t4a1LSEkNKvP4/wcfybmK85efg+i4/ORl6kYvt1suOTg8ZP/uHov5MSed8Nobb3Ln1k2lv+96Hn/yGd7/keeIfcf29rZmN7cdJ4eH3L5xg/2bNzneu8Xi6JC27VTzJTOXQGNNQFM5UnSE5RHdYk6YbuCdMNs8S8IbjKwLZ952JBFmzUTRoRhstzXfxLSQFh/GUfRdd2rvYGM64ahd0ofItPGQ3ODAm2kbhVXmzzSgfupIAwjOGeGkmX/qe2lKkEupfFayVPlUzLBIIvmhAbXGjXR+Cr+BadEoqjk3m4qJ16wGjY8phYqMoOOxmT1kYhvjka+ovGf/6lc5d+Ep6mYDt3We84+9QH/3VUg9odefAgzl6EZMSOdIYqayd7QnB1z5/Kd58yu/wbmnnufJD32Ec48/Rj2ZULjLcn17TIhEBM860r0OVN5vnGpBOjg85F/83M9zcfcSHwuf5H3PP4kAd2/d4eBwn7t39/jFf/tvef2N19ieTdjamCHeEyeeq7duc+bGFo2H/Zs3uP7qK+xfv8nx0RFEJan3DiqBjUpIubeoBVYzo0B2VfsQ6ZbzYppU9QQ/meBZkNLQ8mVulbuNmTmuLHrTSAw0XnlRCY7K2eITx+akYd51bEw1Mzthu7cZ9iKuwNwZBIHBLMl+iWA58SLgBhRvYGQd0YjlGJFEcIrS9TEYimmZEOaYD9/XZNaldfk7O51RiRKpJJvBIUSroyBjqKZ1IiRnvl0OuMeW/VuvcuHpb6fCceb80xwu70A3x7slwTn6viPQG2Bjm4WBJpLQ6/DaHSS0C25+/SvcfPXrbF96nKe+7Tt57NlnmUwmYFkk2VzOVbIDwJUeol+HcaoF6fjkhE//6qfZPb/L1176CpcvP0HlHLduXy83/MZrr3N0dMzx0QnL9jqOyFO7O7jQ8YWbr5K6ObFdUqMT40X9q0o0J04s4JhhUF0omE+v0xmBLkZCt9AUJPFE8dTNlCZqw7MUMykIdN2CuWjqihTHW5RvwQ3Z1uumgyD4FNmaTrhxcEQIMK0rEtZryR5o4dLDE6IofF2pRDlzJELIQmPWXISYpAhPzoYX5/XaxQzepB9OaJzKXLsV/yanVGUwoet7vBMmTaUMYVFKrKvkmILFhCwjwnktSDTylSQ570D9q6Pbr3Pu8vupJpv4eoNm6wLt4RW8cZ17KqNTZvgZK+mkUUEPRNFNp+sW3LnyDe5cfYNXLz/Bc9/5cR579nmkno7KJCxZlxHy+DbGqRakEAJ37txiPj/i2pXX+exnP8OkqUESXdcjeNq2pZHERBJbTtiohOroDrUH6YUKcJXH66asvYUM1dPqUot5uEGQRCiCqiXmjjZEfIqa52dmTe0rbUnptPdqTOoDhaBgRdctRovOmo5ZFnVlaPa6MDlxNN5zZjbhpFsynajP0fY9VWNd9kQ3BH3YQ1wkp5+LDNqplKibU5EyUOELHxLJeettmwBPlEDqOr3P5AxljCX5t/K+ZEeHEFi2S85ubFA5T8x9paAIWl7lAiUnT0n7NVnVyZDJ5kygXVhwdOt1dp/8NiIwPfME/fE1kg9IcEgI5GZyw/wJ4ixzxfwo7512z0hQO215E1Jg79obfO7mVS48+Qzv/9j3cPGpp6nrZhAiBrl8WEwyj1MuSD137txUAkARIzKxFPuUOLfRcGnimDhRjmwrCXaSqFA+OFIy4VGykCqnz5PMlPCFcimbPSWfDbFSA8sYz5F0PN5VBm97Gu9JlacPmhmdu/TlzAXlmLA+t07TWzTr2x62nTPHO5zAZtPw5t27bM0apnVNUyB1LXQbfzbFpKSYLlFhrt5Y04kUjRYM0q+zqopRszdk2JF7IxLxIZP6AwhNrT6M+lKq5bqgFF2NeCQlyx4MSIqWmmPXieDFF43mK8uOt+Cqaq7BNHWSOLrzGmcf/wC4imqyRT3boV/cguQ1BNAnxEcybwdAsLCAy89OFMTpkiYY+RhNmBShvf3mqxzdvc1TL3wHH/zYdzPb3NTsDtw9m9zDxqkWJO8d53a2hqIvM4smBLYlslXrgmgkaxTbNcTMB81xoXZDVD4vWi2ryHb6oDUQilZyoj5KtCyDqlIKLI1HAClR+4rG0vadRPooVjWqQt11Qt8H1VghaNNhlLMP01Ka/aAmYGULzSXYnEy4tX/IE+fOauazZJ43MT4HCrTrHBBzfEmvbT04LeLIRSdZI0tS4sacda0weSR4j5dQihErp5qoqlTj9Wa2tX3PZtNQOyHFwJiMRiwep89CK1Z95azozuszECOSsTSlXOoAkLoT5ns32Dj/lMbytp4kLO9Q1aLsqzGSckO4EaCCgyCRFDL5jD6/mBLRaw5ftBqtXoTYL7jy5c9yeO0KH/iu7+HiU89C1RRw51se/p42DR945ilq62weFwuW+/s04tisrE4GNQeyyZVZakQoD9IXFk1Fnsxnp/LKn5070jlBnU9bdM45LaBL6nh7V9mPo0uRlEJJ/48xIsaRV7IFDPpz4gje4YL+JKvpSTFT+lrXDDMXvQs479isJ8yXLXcOj9nd3iT67N8k5RewJNF1Z1iNo/Xgs5TSEhENcBcHO/kVAhjVvtDSg/lnzgt1pVo9aiSVkALHy5btyURhdIP1gZJUmjkPtF7Lar4s40OU5pU+BlLqC6w9wPxweOd1Ns49SUyJZuM87f4mxBOkclRRN7oQFRZ33utGkhLeJYIPJZVqIJLNsyPWNA3amAg4TvZv8eVf+jc89cLHePajH6NpZjj8SsD7QeNUC9JsOuWjH/4I3iVuvvEN7t68zqaPTGtt21EK/UgFStW1M3ie3kotMlS7jlxVMlCrqO8DkOFddYJ1d5VSgOacR0hGZzxKiTH7v3Yq0CElpBF66XFBES6kxiU0rpXUFxt3RtC2i1LMk+3JjJvHR1TOc25zpjCxE4gOKusuMcr60PkQzeRGqbwQwa+lVjnThjpnA6d23jyEqOUiTmNUzjuaxlE7Txs0gNuFyGbTsFE3mv6EQtiYj5lLP7xzRYhKPyPnLYXHslecDP5OGrRsd3yTfnlMNd0iAtXGefqjk0KIkhqtLk4GkORSD0mrKGVKQ5rykG4lhKAlN10M2li6a3nti59leXLEh7/vB5jMNiG5t9RK76T15V8E/ivgpn30z6eUfta+8+eAP4WWcfw3KaWfs9d/CG2L6YH/I6X0Vx927rquuXThPN948UscXHmVjUbUuRfjQXBi+P9A/jgsCPMnjHQjT15GYjI6N3Yos/mTX02j152HutlUIcrOp5kwmR98bKsLlh6UtL+RghEOhd0d06YeuPNKTGcI3OaxOamBTW4fHjOpK+uppIhaCLrNql848OrpYkwrKBmiGdPazS8Z+miBYLe6GZRSg5T7OKkWqUTL0LWTvON4uWRSVWxMJ2pGF1MuE3CqyZwL7DLXgojg6oaUbOMIOp/BBSRmE1uP0QDzgyuc2/gIXQhMNp8gzq8hElYqXjXQmk1L9Ruzni71YgZ8OM2aVQQUCH2i7wJ1H2hj4qSPXPv6iyQSH/6+38lksvmWAPg7aX0J8DdTSv/r+MMi8hHgjwLfDjwB/CsR+ZC9/beB3482GftV0daXX37QiZ1zXH3t61x/5UWmlWNWVdTeW7GbBSBl9fOkBAYzVwUQGIELMiwau16j7E1D4Z8I4FYIIsHhau0coYTzUVtjpsHmjqOgTl7QpZyg9uq7WVcFGFpr2pMvx9bD5JQcoWkqcI69ozk7G1MNcMpQkQpSuklk8klxkqvWipmXRH/EZeoyvb3M4JPnQ69dsyNqL6Q601PZwjQhbXzNuS0rFXGCL1waVhZvc+vECvWS3rt4T3K+TK3DUaWK1CWSSzhypXOlcaqDq/DYh9V39Vu45iy0d60IMhb0LgZHqQ9LQwwww/bjZFpn6yWmQKiEqorUXaAJWp01D5EbL3+NajLhA5/8/lLq/qDxTlpfPmh8CviplNISeEVEXgK+1957KaX0sj2wn7LPPlCQunbJla9+kVnt2agrplVddhUxpA1Wk1MzA433WZikZPeu7LqsZk0PAK0urGjxBGQwFxU9zGk+HSl0kCYWN1kVojErqzduCOfyg7VrsUZoqe+JoR+QwpyDprdHSDCbTLh+94CDkyWb52fUVda0aYh5mBD01jsoz032f7I5mzMhsHsWGZm7I0DDedXoZY4tdhRT4s7xkknl2Whq7eEkgseXa3H1iCZslD2Pc+C8Ja8mUsgVt0o2Q1SUTXwiiaK0sT2hPdmn2jxHjIlqdgn6Q9XOIShRS9Ju7/qwoibukh+JK881C5KYI135WhFIHwk+UPUBcQHf9hy2Hde+9iK7lx7nwtPPPGTJv7PWl78T7YP0x4DPoFrrLipkY0rKcYvL9daX33efc5TWlzubMybesVlXTKy8PEfsYcg8gLw41WRTXmzVRkqwOCxgZPj8PVxleTc2nyuRK8CNLMR2WEmJ0C20EjVGAmFFmFNKRlaSOSKMpMV7M9zSqMo2QVUp4sVqIu09vZWqmteu3aRPsFnX2r0QK7fIgKWMshUMWKH4jXZMlzehHC8bAIIcd8pFyJkWOeSk3mgd8ZxwZnNCXakPlc29vHlI5clZu5pXmEbCbtebQw7iqKQiEDTdyImdX83R2juO715hd3tXu/1Nz9KfNDgXyEyaKY39ZMtLCWZRmPaT0bPXe3QMLW5iyWdEPKDtbQ67nm98+fPsXHp8famujHfS+vLvAH/JntBfAv468Cff7vEeNNKo9eVTF86mjbpi2igXmrisHgyWM3NGbIetvTO+toGvLO/CuVVi9nuyQK6gMfaQ84POwqdmW4ZmdbJj0EbFMcbCfUD2j0T9L19J6d2qdLiQTZ+8iADEecbt6dfmo3yuqiqevHSRq7f32dpQrZQ1cy4PiCWHT+ufVKVaqUQI1lZmdK4V019NzBCj8WKrT9N1AReMQLLr2TtesrPRsDObWcxMqYxjUMQvKV6Ad2KaplBSDGZmjNbaxRSXaPjCC+DTiGYM6sYxP7xK6F9A+1zN8JMdfNhf8Q3HbKgxRqIEixmu0XLdZ4hzkALJJbwPGufq1WQ92rvDrauv3/d7efyWW1+mlK6P3v+7wD+3Px/U+pKHvH7f4Zwwm9SWWuOKmaBOsylo24lqr4I0ziQYmzPJhCir92jJm/ee0w/lDXl3RifaWWwhSiQYV3YMCXEKEmR/q87cbd4bPZcvO6z6NxjEbloppkKLG62lZ9G6I9PMe8+FcxVtH9g7WnC5bvCNMe2YrwDKppNipG1bnBNNDwpBa4ZSohexcgRXtHCy7uV2UpxzTMwvcPS0TitqF8uWPkYubWyXTHTnFKnz5qLmEETMPtpIW8aIQtQMN6lVrQxomiQjf0wm2wnCkuXxXWZnLqsfOL1EOj7UgLREFTzn7aR5Q6pH/x5g9fVuITElsH5SzikljYsqzFV0hD5y/ZWvPWyp/tZbX4r1j7U//zDwRfv3zwD/p4j8DRRs+CDwaZu2D4rIc6gA/VHgv3jYuZ1YENS7ooEEtKBMKJBv7Sl+kThl4sxO9rgtSKl9FSl0w0WYCjxqtFMxFRMgkWjqhtpPyvdD6EqvHXMfjDdPE4u9E7wfkMBsymRzJ5caKMljRro83nCAbAKutIEEqpR48uIuX33zKhsbU3YnE0MS9VGqFu4JoSdFCH3EeYutFJAj0CWII05Au7EyU85pUWFKicbVpOCh7ThqA7tnttjamCo7k8XvQoglwK1CZL9TpgfTDIYQo2Vou/xIh/acmM+WHUMTqkSiroX53Sts7Ch4LM0ZOKlx0oP4IWcvb5wpFUR0bC6P/cDy6K1BQclsAXCpVGe7BPu3bz1sqb6j1pf/uYh8XB873wD+a3vgXxKRn0ZBhB740ymloM9JfhT4ORT+/omU0pcedmK1n4eodUGYxpnMIlSVlPiOG6Ez9ztecbxH7xefZNT+UbMjVOuJJFw9RZymvkIkBS1vztC48r8NcSvVhBrPESgClP0Zy0IquWfYvY2bB2vgNMe2hlH5mmcuXeTN23tsb25obCbHy0SITrPDxUW8rwixh3xNIuW+sFtNKZteJr4G45fGbCHiHex3HYeLjmcub1u3dI8z7ZJ5GLIJlwv71A2UAnA45ywDHsS0k7LbhnI9ZvfqMzJNXtee5clNUjcH1xDxUJ2BeFdJP32JBpZwQgoBMfSQEFdAp5T7SOkCMP8vmhkPCpNnCrdE6JYPW6rvqPXlzz7kO38F+Cv3ef1nH/a9+w1BzSGX/Q4rJhObOC+UhZtz3MZjTLMLqzZyQdAybD6aRLW7pZiJvmq0ewOZuqpnNqmYTOoi7N6b451pdlFgIkWtSnWmKfN1eO8NNVsV/vG/8/0kGfdi8pzf2eFk0XLt1m2ee/qpskix+/GVJwVPCI7aTJxs0kQUws41dsOUrPZdLeUeRs98NJ+zuz1jczot+YkDYqhC4SyuFyUqrXEyEMINgED2U1XzdOT403jH0GPmYhHNvat9z/zwNtOdJ/QYk7Ok+Z6ljkWSSEkL0u8URAWRAegpwVmxXcQKBjNpZkg5qIuBEQPv4IOGe+i7p2Ao+WNFZcVaJefLaT+gulYgoqor6xWbmU79sAOLrGiq8lr5nQVMGECKnBqiVE+umpJ3ShGHIzKbNMbKWlHXSlvsvDPye2fxq7EZl9uxrHYPzL/XBSgXAaJXNiCCThliH794HhHHwdGJaqKU6adWhTVnE3hvjKkZ6RT7ccN5xyZP7oAXY+LgZMlJG7iws61wt8t9kBhZB5ndCEtHMp+VEWLHeDMzrb9mQYgtcKu4Aiw7w1ec7F+l8HHU2wSaYf6SlOvO5vtYqHK4LguSAEPYyeic7SeiJDV5nsYuwv3GqU4RUtOn0tCDmTk5ebW0BxkJjoIR95p12TRM2QE2MzGbAJi5la2KZOUSCVT9J8HX08G+jhqrqGrl+s5WRd5xfTbJkKLLV7IOzL8r6OHIVB1rAzEHX7ymvOT3cyrNpK65uHuB6zdvsjGbMZs0ZK0SjS+v8tppQxBiDCsCO2RBDAKkpfzBIGRHikLbB24dnnBmY0ZT1Qz0v7p71271XuIoYxxJBWhRUEVTl7TUYhQPTKmgjQCp+F2Dz+ochPk+MbSq9ZJDqk1iuGubw/2ptezwKxuMAiw5hmV1WAx+abLFklLGFh8+TrdGkiFI6Eb/zrvseKeVvICzpmEQIHWAdcFk8sLx8YBSnSmj1yujOgZt4JXEleOQEpOmpqlHmo9cq4SVQK+m4aekbVNC6gkpEJWJW6Pr1hg6a4GQBge40C2XEm/KNW9vzjh7dofbd/borEq3pColDRM4CwZX4lcESaf4Xp9ShV1h4y5E7hzPmdSei2c2R2Zg1njajaNosny80XyqxtPS8Ey6b9OjXyn/H2+AyeJ4yfjGTdBpmR9cw1mqlJucIQQtjRibpavryDamcnwpWikWxtloDQZieQbq2g8b28PG6RYkhgftcnmBG7q5qaOvAINbWXBmQrnBAb8fAJFVdnm/CFZ2xFPZbZ2fWNOtwSdoGj8kxdqCyUCB2tyWRCmqjUJmbY1quuSdTx/SQNFVhCmGgiyqz+UKlBxsoTlJ7GxtUk9q9o+OtN9sjMonDprrN844uM983CtIyhOBwHHbcePggAtntpnU1Yhma2QOigPR/rs4R/KuIG6Z00+fiSuLedwUYCydeQNwaP1U1pY6J4GqgvnedTt3IvpNuuBK7VcumS+aPWu99ZUeFa3LweKQojLNxoFFCciZW285Tr1pp3Nt3Sa8QqbjJFRtN5LRtfvb+uuv5YAgqGOM+Reri8l2MBMO74e2IiQFObSXUM6KG8zENDpvsiBnkgGuF2e0X5Ehg32tSZqQ7f4hjhSIuMoZG8Kwf09qz5nNGa9dv4kXCgORiEHPUU3ecebECvy78m+d+YTQhp69+ZyzW5tsz5oS28mfjikV1iFGpqIDgugZHSBWgZvsBA41CXHgCjeeaoCckRFXAIFBmLzzLE7uEJYniNOmZpWb4cKB8nXkkIYi7mrSpnHyruFKMZVOjXlEI4oZKMkyB9/gqz5onHKNNFRTZse1ss5tzhhNx5poJWZ0n3gBjHbSYr5g/179XtY63kFVCeK0dim3NGFkeqWk5Rg+H0ZWfaKkBlxmQjYtY3UyqoyK4w6MEln1lRACXdA2JlFZPu7RJhNfcenMNtdu3+V4roQsub1LSRka+2mMnP5VRa3nTInbhyd0oefxc9u6adh3vLPSCpuTch2j6xl81Qw+DOAODGjrMNeaoZ3LLigbUxx82KQxnYqO+f41qkpZnNJkd2BoClrflbV91u46l3q8nCyczeDiEzHMkT6HEXr5Fiv1dGskASxyLlZ/5DOaBkMkegRxj7VSKSkvDj/kvdyVh6W7yXqegwpIJlDTIrRkcGgybun8Da1YrdAu4VIYVMuCLSZK1GTYbDplx9piLBmeLt+NAyWwE4XdCyIlluSZP++ErY0ZZ7uOw0XLZNJY2ACNpaALcSWBVQwVK4tH03RSSuobHR3z2LkdJlV9j1lMAp+M466EcFzJxs+cd1o5rCZpGJ07Yb2bjN+BpLwSKT8zNA6km1wCPClFAlBXcLJ3ja1Lz2kMqjlDd6B9mnLwuzw/t+LUDaZzCKN0pWGzCTHTj4k1qU73eG/3G6dakEAXobM6lgxTF2BhBBbAmt2eMlQMw14/wJlhJSNmpD1WNJgidAmP8w1itMHmpmptv2hsyyFIpR3t/KioLB9/BSVjOG8mUsk7eDSEbSXuZUIndvEF6bIxRi7Pn9lmkRJ7J8fsTKZgGQRYwmzeiEy2jQBSYz1YZvui77l6d59LZ3fY3dku2mQoQMyanHIsYdiMEkPSaybgT6bBnW2OFtyh9spulAnti6Mfhw1INcoAQVeVJyz26RbHeD+jjwmqLVK/r6F+cZaGpHZcrkDW9KAhxzJG1dnjxN1sGuQywYLqvsU6PdWCJCIl9cMxSoO3xXe/z6/+O3sS6T7vmy0lVmZh8DKmxnMmdMjlDIbgiQgptDhrAuNG+5UKhl5XlFgc6wcJuzriIxadNDSB1g9TaqgyKesYaMnHG2uZyWSC73tuzRfEvmd3e5uy0kcClPPhRgly5nxHbuwfsux7nj6zrfC5fd4b30GOaYXMB2fm7WiLKPOh9ymk5EoeHGA0xyos5RLSGFxafb7ZTAOUggw42bvB9sXnaNtIPd2lP9xDREMXEiE5inVQTLaU27kY1B1G8Ha5l9FzN6FK3+o+UlVVK7a0E1n5G8agz7oDbdxttjPJaNfRliP6uxIlSMnlE8LYr1DbOzdbTkRIPbUXa015L+qFnbO4YlmDyOr15d2ccUl73iRELCisxYa5recYLFhfdFmgqqri8fPniQkOTuaaw1dVuLpGqgoxCqyxX4MTosDeYsmi73nqwnllLsoBaq9JhKKJhErn5dzI1B42lNW5GPylfK0l66PY3AN4N95k1oPq+b0QAlUlHN+9Si5AjPUmPblS1oQmrm6gOXZUTNm0dq35M3lLSKlYhvIWonLKNVIudZZy007M7C0Tm0Ci3fCgfbI/VEyk/E5KhmAJIsNE62uZQopSph2tq53zFU58KWeujFYrjXbZ/FCKr4NqvvKs0vAQsyBUPocSc0GrtpuMBTHUAOZYwMcLcnW+hsW2uTED4M1btzl34Ty0fSlDUa0ajPl/+P6867i6t8elcztsb2+uVBOPBXhsdqaE8S+ob3GPWZq1YVq9vuxL5mC0F9G5RVOtYkzg7PrCWNelAu7ExR798gTvJoSQwE3p4hFVQk3aXCVs51WT0eqqDB3MPuj4PUYbaBHKt7DtTrUgIVh9kcKQXlzZrctHZCjoGg/nUrHNxwswL5zxAy3+lNnw0Qs9aO6dYGBHVXaxGJRZNCdgFqKNkk9GeSDlOpPgyEFVV5Ix9TZHO3n5Gcyfcer/g1JVhqI8GwkmTcOlc+e5c/cA+si5HQ2o6sKQAkLEEFm0PbcPj9jd3ub8zo6SmLB2D8bIox07IkheeJj5Gcvnx4ilJvAOeXfF5MsaGQgxgFNoXGm9dXvJaFo+VrKGAFpblRR0uPg8y3aJr7foT8xPisk6WFipiN13hrYhpzgZyMAgNBl6j0Ygg8gKf/l95/+h756CkaFRPwq2juHpvDCywOSxEhthJDAMm/BYmFby8oDKjp9EwPyY8k0jV1EoeNzWMmugIYDsxSFJg6lZZLK/l69zrGXGD71ojzXgYiUelv2bkRYYQ8ob0wmbdcPdg0P2jk4oWTij47V95M1be4QAl8/vWlrRUIaimQmuBDudpW6kqBonQ+2ydj+a5a6afyDeXIXfV822zO9gOZPiyPtGtKRSnX5NpK284+Tum1byIaTmrPXI1cyQEMwH6gOpNxAjY+kFrRza2gTTUjkDPPdVWF9P9xunWiMJOeVGCqS5atOOS7EH023sYAIrZc3AijkzoDmxqP9sjohkts5MZEiJCTW5fHnkn0RbWFm4JWWTcbj2AQWy4OxoLxM3+u6K1l01r8YmpGoiveYxN10+p/eOzdmUpy5f4PreAQfHcy6dO6O9kWLg6GTBjbsHzOqGi7tnqKvaSjkSIfT0MWl6lM22FJVuZtmov+5wzYP2F8favdjViaj5RbYalKymwEM5yThCkI5xLhxRtSNAv9ijWx5TVxO61JCkIsWe3p67M4EWr9W2Q+DcEEKSdh4kv2fmdA76MZKmh4xTLUgw2NlD7FTKw8jqXk2OlBsnoO+uRu+zw5vtXcmAQDarbFEoswz4kEoPbWRA7LTxXCZ0VELCzEfgnXbXzgt7KPUeNKLdxIoDnn+i5Cb0qr1W58DeWQMW9F4pJP7ZPNSNZNgoZpMpj5113Do44MqNO0QR9o9PWCyXXNza4NKFXSaNBlhz+fa9WlA3mOw7ACssTnq7I1N7sKRXhcllqH+14C63wEmSM7W1VWVMCcnpPClpyVHZ7wKHd69z5tLzxAWkapt+eRtXTEt7wsa5keeFFE37RAt0j7SOmeVv4RatjFMtSBlRctmGIwtSsp17yKLOm1yShEQgisUzU3n4klJZiE4yP5tONuKIFox0UQguDCn2oi0tTZmQQq+LTTR+FAml4lYPbiXPMuyeQ6azK75GRuQGpMrbYsyaLS/kISsaKJWb+t6qqZed53wY/XF4n2jqmou7uxweH/PiG2+yub3B7taMSztnqZu6aLEh6p+FcuxLrpo5KxtEshqeLD4WVlgFH+y3QdTj+/IypBtFdCNzeKokuBQ0KTdqVji2mXrvOdm7xrnLzynferVNf3ITn9S8jmRwavCtBxh8yG3sk6UNRbMk0sqlfWuXUcDY9s9FZJjdvG7mybBrKkhNRpVUew0PPRNiZDuisJ9l8wkGbgW0qC8f1wkKm2cIOUWcZA63oR4oWQvN3MRLSpwqC9MABw9QtiOYAbWKlOV7sPu1QOIKT3YWpqgFbuO5yUBFVTskCdubm1w8e4a7J3Nm04ZmNqOqa7NihiTb4rPlNB0Zsi9gVYiyWQ1ifIDjbOv7mOUpx5fyPeaeSnawlKtu9Zl7m+M+6AYVrWjQhaTB2eUJVd3QuhkpeYh99trImEPeKPKcpERhRxqusyynlVKPFUan+4zTDTYIpVTCyVCch+3m46Fzb0K36nkAaq6NQQcY7ZDOIW6IWXjLNMjOqLiKFccqhlISPlINQC5h14TM3EtmKCocCgYHJlIpWecmzsM9rQEIwy2NkilHJl2GxMdaKkbtbpdNsazlN5opBycLJs3McttU+zjvEV+N2ICy4FOuZQXIyf5ozpqOVpBnlFY6t+O436rBpECEpVmNOoOUtjUY/4Xxh5fUJhmet0uB+d4N3fDqCVEq85U15y4QCSkQYk9vbVMHLT4EbXW+LMeumHcZgAo8bJxyjbQadxhWru409zjf2SFN490wlR0nI1GSESRASq2RDAmOIzhaBFxVl3PqE1Ry+ZgU3csPNe+oeeEVTYOULO9cXp6zoTOY4pzTTIHBMBqueQ18WA/GjqFxRdVkIHdxWkMlmG9jvmOfAs4LXRwgBF++r2Zu1gwp3xOrqVTjf+v1mR9i/ueQOJwXoVkIMsTwypMu9zdoaOcSkiISHSHzgRuLk5gJm0jUlXCyf5Mzl5/HVQ1dvUlcHNuz12fmHeXfzjIptBrWTDoDGmLQ1K+iqfR/a9bPveNUC5KawQP7zXhnznZ7jmXEmLRe376bUu6MkOz9SBQtffBmf3vLHBg/Uik2gF1BUg5yNZq1zifFFirtfkCyAGIakiA15UfQ0oBUWFslBwkRyOysTpDKfKNovtJo8xgT7K9rUrvTFfs9ZuHB7HqB3igHQkqEPnL7+Ji7Ycn5y7scdh1X9u9y8cwOG3Vd/NGx2VnWE6uLfwW2RzMOcFKmbjCpc1Kxmr922VaCkU0mMyFRyDvnwWlqloENyTIeqFTAe/WVnMD85DahW+qm6DMjbxaegZSFRNks+xitGlgIDCCWJhdHc7rzUvgWFqQHjbFaHmcsJ0PuQql6zIJkiA8YtKDAg/nSClqQC9ByTMRy3KLW3GQzIqVAij25y59t2roe4kjYE8DQ6lKF3pfs5lLbYxdQTDxWNc7q5jE49RStIGQOgxx7iubfZKF2ojtv2wWu3LnLjcURbewQNJ3oYO8usfHs9lN2ZrPSkS+N6nWi1W/lzeJBcRWRbBJDjsnovQ1c36txJG/PcbA0BtBG31e7IWcZm1nrvAqcafaUeuaHd5idfYx5NSPLvZiPlxsUCK6UXPRJ6cr0Ypx1x7BYUzbtzE8L6VscbABGAMGwyMa/o7kkOEcIumhCQZciIeSkw8HcE0OGBAUCsqDA4EfYJ3HVRJWTJXU6qy2Ko+sSVynt7diVisY54D25gLBkIBTNtZoypLdxbxuRsvtbCUYcXe8YtcsoZ7l+WwgxJW4dHXGrPcZVQhU8XehZtC07OzssBY5rR9W2nJnNzA9JKz5XSvcX7nXNpKtPQYHC78Dgow4a1eI1unvonJqvlrWgEz8ogzgqdnEQvWrYFCO1g/ndN9nafQrXnKFLDXVaWivOIZCcmxQowckYLEnE2GsrnLwp20YaGQEhDxinXpDG6FX++34+g3m6utMYd0G0nSj2tpsmrdwkKQtRCLHY8uNzwVAlGZPDV9Ni1STriJ7jUeUaRterPlrWQkKGXgv0DqQMfbO6IMcoXj7eyn1qZFozJbL5YpWdWTtmQdLFaM0oY+DGyYEu1pgDoJ4gmiG3XPbcCid03rM1m1KJFk+u8yCs+2NjRGvd+hkDQsPGN/5EKmlQmHYS005C1FhQ9mXV3lP/NqGblhdcHK5jcXKbmDqqyYxFNaXqlxYG0ZY5GcmNpbTcElSLj0uZx6z5IoGYHg40wLeQIK2/NoZ7sSrHEJT0b4xmKX3vkIKfDYiqCrjkVHNFKUWCMAYrLKjp65Gaj6ZNBgRqBVYXV9qeYOba+nWn8e69dl+qbSjXMX6vmExO71tpwQScmrP5/gp3HIqKxQRtCPQkpnVDJLBse7yr2JhNERIn8wXzZc++8+xOZ5ydzvBr1ybOQ6Xa3YkrvsbwPLLWj/dsBgPayiDk+uYInGEEaLhynPVNSvP38jPLpfiQ+gX9yT7TZoujagPCIaCxqGBEJrn9Th+DaR4zH0WfV0yDmaz7ViidLR42TrcgyeqOPEaLYJRlnfIERetjqo5vP+o/FFOk63pihNkEmroiFu6HHnEVIYycejEU0Hl86Y0jppEGpCtrAOeUnQiXiwHNPDBI3FlJQL7edc2TkjrUWYutZC7kex6ZRMiQIkSB3XXkIGNGCJ0AztFMJsxmEyQlqrpjMplwcHRM10e6XnPSFqnjaL7g3HS2soAHTVtRyu1tWxIyzZmzBGMZLfAcqhgESa9/FeTJmiH7i07UGlD4gUIrjAg47R276kc6Ei3H+zc599hZqnobljc03JAUcMn+YxcCXfabk8UlnZj/WmaYXObuEium9P3G6Rakse3E6g6dF1c2PUJQM05praDrOhUkg5RDiCy6SN/pblo3fUG1cB6JI5hTDOGKuRZNilCEbmGOr1etM2qnmcQsL5eTVSNOtFSCtZL4se9RoFw3lFKUGVjzBzV3LOU3i0ivm7wDtZiaU1Xl2J5tUU1rNjcanGhwM4TEyWIBBGIUur4tx78fqCAiyjGB3azt1r7KJfiWvyOKtq08R1WResWSRRArpTDTVDLqZ7HAkmWB2l0kHK4812EetSz9YP86u499EFdv0gbPzEeCRByVaiVRH6gPwUw8cCGWxtArxk/RTEnDAQ8ZbxmQFZGpiHxaRD4vIl8Skf/RXn9ORH5FRF4SkX8kIo29PrG/X7L3nx0d68/Z6y+KyB94q3MPjkkucZAiPGNtlNW/8pKlQs3U9z19H1h0PSddYNEFjpYLFl1P10Xavif0YeV4RZgAEQvGuooMA/VdaxD4aEdkFGyUYXEoA2yNM86DbNasMsC6QWNZFjlwz2ds/ooWXN3Ny/wOxxwJrpDYqD1PnL+kUHj+ScJ0OmFzNmXaeCoPk7pmYzJb2+1X43mFidZlejRXNGLOGND4XI1zuldHcfaTkcuxWTyYtd45m4t7CW0yGioM5xeR0jxBBLrFPn03p5lsEmRi/nMWWsukQGW9C5G2V5N32JtGFsLYhXinggQsgd+bUvoY8HHgh0Tk+4G/hra+/ABwF+0Zi/2+a6//TfscstoS84eA/11Exmb4fUdKQ9pNXNsdB21kfYoi9F0WoJ4uRtqQWHQ9y67nZLnkpO1YdB3LbqmfC0rM2Me+LILQGzecaEHfEAdJdO0cDEJf2bHWiEsyG84KU06uVC1tId09gvKgv8f8FGUWRpYe91nwmP0votfR9EvOnz1D10du3t1jsVgynU7Z2tpiOp1x7swWj+1sszmblmyD8bXkkgr9rZpXvLblxFEyNfTz1mbFuO70Wtzouu4Fkco9M2grZa6tSmPnqq4Q74qpXObL+A0lLmmXB0ZlPbFYVAaQtBN7zvRoQ6QzzZRNyjK1ZpmIVQ4P2Y33H28pSEnHkf1Z208Cfi/wf9vrPwn8Ifv3p+xv7P3fJ3qFn8JaYqaUXgHGLTEfOPLNjdM6Bk7qVUqlYELR9R3LEJi3PSdtx8G842CxZN71HLc9yz7oTmQ/sY/6EwdiRk0rAZwWrGXzIvYLfS0N5x/7cuPFPGZHvbeWKguXM4KS4ft+5E+VBzX6TP6cE6cpSOJGkH0+d2aJ1aflRZh6x05InN3eoqlr2q7lZK4dwieTCWnZ8eTOWWZNdU91bLkvxbBVu7qR/2SghxThE1smFmTIx1jb2AWU0cfM3fW5wgnOa6m9r/wgQG51vsffOTnaw1cOmWwpOjpKKepzuQwKbnS9cX2nCH3PCqxoql8Uvn3oOn1buXYi4kVbutwAfh74OrCXUurtI+P2lk9iLS7t/X3g/Pj1+3zngWNscmVhSSuawP6OidD31rY+0QUtWDs86ThatBzOWw4XHSfLnmUXaPvIsuvpOjX/Qoz0ISiaQyJG0Ug9Q8cIhNLeI1kMIldT3mOymcmSTFukkRDknW6saYa2i4OZk9u86Ov6GZ9NQKel75oNoQt38Bn1NRmZTw6tNK5TYnK8ZMdPWRy3HB6cMD+aw7zj+YsX2N3eGqiabe3k5FXVQJpiI3j1E9NYmCjm6xAT8kMeIwx0XaN7yloov67pSNksFWsjqvcN1n2EUVdGnzcSzeubH93R3Mp6gy7mtKS8qWgGTLINswu9bdCmHaP2n2WsKRNvWSH7tsCGpMlSHxeRs8A/BV54O9/7rQwZ9ZB9bPeMolNJc64Qt2JC6W8LrhXtpH5S1wW6PtGGwLLrOFqoAPUhcX4z0IdIHxJd6M0Mg8phbUmclalDqiYlaAqJGHqcH5FiZIFOaXVXyrtp8VMs9oKZdeVVhe/FkLVsXiKutK4RtIJTMtqVDLBABQSDiwezCVgJMdslIXiBiRcuziacqS/SxkDjaiY7lbbpHAq/1syvhEL7GktTBTSkD0lWUfnvHOcaASN5wylgwpo/onG1weF3eDRsaqaj+GLKa59ZbcYccYircE65xZcnh6S+R6ShT8rvkBKKcCcxIYp0wSpjQ4Q6IuJzFJhsXCZ1sLhnMtfGbwq1SynticgvAL8DOCsilWmdcRvL3PryDdE2cjvAbR7eEnN8jtJD9tve93gWFyPqW0OTkrGJGtwdU6LrNdO3mGgpsuwihycty14nbmG+UR+hDQnn1fuWoJopmyYhJhPeLKiBFOaD4Ji2ycQZCgD6Agas7mFSfrI2UwWkpseYsF2HAr/FGYeyg2dSvlQOO8qYyLGvNTMwpmjdPBSijinReA+24AdBHDRMtOxOEQti5niNraysVZJRV6k/KSUFZ/1ZDUmsQ1k6IviqKtx+XlR4gsWqkgxZHlKu00Bx55UgxQoyccqP18We2C+pqoZOLONEpGyIIT8PoWy8KjgjqwLlXJGgGmkdvVwfbwe1u2iaCBGZAb8f+ArwC8B/ah/7EeCf2b9/xv7G3v83Sa/iZ4A/KorqPcfQEvPhw8xT3flW0zQsHKj/tiTVPhpdrZmCYos+mvknwKLrWbTmJ4VAHxScaLuevo/EoEHMLgBGAiIihNAT+g7E6mdGdE8ZNSwm5xrqo2tUxvJUsoxDsuy+0pNJNYz3Yg5/hbcSgoxqDTRZ9/pTyAAlZ82STcRcep1bz+TmaMWUHPli+TyY9TjErAYeByH7SyMhXPPn7tGMI02kG5DDGTmMvmEZH9msGvmdIub/GbDh7VpSPq8TYupou4Wagt4KFo2RqbCniplr5gMVcMP82txgenzuh423o5EeB37SEDYH/HRK6Z+LyJeBnxKRvwx8Fu0zi/3++yLyEnAHRepID2mJ+cCR/QU0ejCmvgq2kGOMhtLp4hdbRaqSxSqMR76Cc7Stwp5tH6g8SNfr5OVjlwcdqaumLK6+b5UhK5s0jMySsUPOkMG9cjvm+MaVXZ0SKFTrzpOZcgapy3VF+nlvAIUi8mnYZGx6NPSlizxrdJFBO5TrlJxpodc2FPKZFsmmFmIho8GEFsSySqByFZFRGcpovr33yCi4nF8fD9140LKJfP6YkLXVOcxvMsFRjRTj2LcEJLJcHjDb2iK5GbE/zouJJDmpGWsHk0pKlXZRT0Xr5vVSAvsPGW+n9eVvAJ+4z+svcx/ULaW0AP6zBxzrvi0x387ICE4yEyN3EogxGnPMKHiaEiGJJasaYOEcuEQbevYXLee7yLLtqazozgVNfMxsP5gZ4KtGd3iBvltA0nOFGPGj6P16sLWYMNl5LujREFAtqJ1+U9HHqK8l8uf0s855tFFYmZDigmjWuyCiTDiSL5g0nDclnITiX9nFAqPFubYpAIXaKhKLc19ieUCJ/quTZy5RUkEfQcorAsxwDYrkmeAZaaEKdfbP8v1i4Ya8gWX3S0U9jecYoV8uYVvAz4jBAX1BD0OyYHRKiIPa51jeAGDkYG1c2wQeNE53ZkPxU0dl5CkVRqE+asv7zDCzGljV1Pc2Yk27FIgNMbB/0nG4WFI3jiYIdRK6PuCcEL2lvwQsc6HR3Tb1dO1c7WwRolOt6Ep6vdh16O4Yyb2VhhiK3pPFLFYET2MgkOj7UPgCkubJlFKLTCOVTTZzhjRALJqRLppUQIqaVRFD9plQLRWHxe0snSf7AElGybajRyDWMHlsonmnXZe60KupC3nTJ/vn2TqonC/0xvmex881d58YUolgKFiUkjuo2eRx1QdMem1IJNiGVomnX7Z6/JzxbXmY+iyhi5FF6KgcNLUhkTKG4KFKjmChDt5CmOStnKj3cojIIfDie30do3EBeHif+G/eeHQt9x+/ndfyTErp4v3eON0aCV5MKX33e30ReYjIZ07L9Ty6lvuP9+pa3lZA9tF4NB6Nh49HgvRoPBrvwjjtgvTj7/UFrI3TdD2PruX+4z25llMNNjwaj8a3yjjtGunReDS+JcYjQXo0Ho13YZxaQRKRHxKtpH1JRH7sm3TOb4jIF0TkcyLyGXttV0R+XkS+Zr/P2esiIv+bXd9viMgn3+G5f0JEbojIF0ev/abPLSI/Yp//moj8yP3O9Q6u5y+KyBWbn8+JyA+P3rtv9fO78RxF5GkR+QUR+bJolfZ/a6+/Z/NzzxjX9pyWHzQx5evA80ADfB74yDfhvN8ALqy99r8AP2b//jHgr9m/fxj4f9BQ/PcDv/IOz/27gU8CX/ytnhvYBV623+fs3+fexev5i8Cfvc9nP2LPaAI8Z8/Ov1vPEc33/KT9exv4qp3zPZuf9Z/TqpG+F3gppfRySqkFfgqtsH0vxqcYKn5/ktVK4L+XdPwyWlby+G/1JCmlf4cm+b6Tc/8B4OdTSndSSnfRIswfehev50HjU9y/+vldeY4ppasppV+3fx+i1QdP8h7Oz/o4rYL0W6qmfRdGAv6liPyaaIEhwOWU0lX79zXg8jfxGn+z5/5mXNOPmrn0E9mU+mZejyiZzieAX+EUzc9pFaT3avyulNIngT8I/GkR+d3jN5PaB+9JvOC9PPdo/B3g/SgJzlXgr38zTy4iW8A/Bv5MSulg/N57PT+nVZDeVjXtuz1SSlfs9w20pP57gevZZLPfN76J1/ibPfdv6zWllK6nlELSCsu/y1BG89t+PSJSo0L0D1JK/8RePjXzc1oF6VeBD4py5zVoceDP/HaeUEQ2RWQ7/xv4QeCLrFb8rlcC/zFDiL4f2B+ZGe/W+M2e++eAHxSRc2Z2/aC99q6MNR/wD6Pzk6/nftXP78pzFBFBC0a/klL6G6O3Ts/8vBuIxW/HD4q8fBVFff7CN+F8z6Oo0ueBL+VzogxI/xr4GvCvgF17XYC/bdf3BeC73+H5/yFqLnWo7f6nfivnBv4k6uy/BPyJd/l6/r6d7zdssT4++vxfsOt5EfiD7+ZzBH4Xarb9BvA5+/nh93J+1n8epQg9Go/GuzBOq2n3aDwa31LjkSA9Go/GuzAeCdKj8Wi8C+ORID0aj8a7MB4J0qPxaLwL45EgPRqPxrswHgnSo/FovAvj/wPm3XHc0n6jFQAAAABJRU5ErkJggg==\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"code","source":"images = []\nfor dirname, _, filenames in os.walk('/kaggle/input'):\n    for filename in filenames:\n        images.append(os.path.join(dirname, filename))\nimages = pd.DataFrame(images,columns=['filenames'])","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:50.052335Z","iopub.execute_input":"2022-11-25T20:22:50.052963Z","iopub.status.idle":"2022-11-25T20:22:50.315577Z","shell.execute_reply.started":"2022-11-25T20:22:50.052931Z","shell.execute_reply":"2022-11-25T20:22:50.314536Z"},"trusted":true},"execution_count":7,"outputs":[]},{"cell_type":"code","source":"images['given_names'] = images.filenames.apply(lambda x: x[x.find('name')+5:x.find('/',x.find('name')+5)])","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:50.316968Z","iopub.execute_input":"2022-11-25T20:22:50.318168Z","iopub.status.idle":"2022-11-25T20:22:50.331773Z","shell.execute_reply.started":"2022-11-25T20:22:50.318119Z","shell.execute_reply":"2022-11-25T20:22:50.330813Z"},"trusted":true},"execution_count":8,"outputs":[]},{"cell_type":"code","source":"images[\"image_type\"] = images.filenames.apply(lambda x: x[x.find('/',x.find('name')+5)+1:x.find('_',x.find('name')+5)])","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:50.333097Z","iopub.execute_input":"2022-11-25T20:22:50.333665Z","iopub.status.idle":"2022-11-25T20:22:50.346342Z","shell.execute_reply.started":"2022-11-25T20:22:50.333601Z","shell.execute_reply":"2022-11-25T20:22:50.345164Z"},"trusted":true},"execution_count":9,"outputs":[]},{"cell_type":"code","source":"images = images[2:].reset_index(drop=True)","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:50.347578Z","iopub.execute_input":"2022-11-25T20:22:50.347813Z","iopub.status.idle":"2022-11-25T20:22:50.358619Z","shell.execute_reply.started":"2022-11-25T20:22:50.347787Z","shell.execute_reply":"2022-11-25T20:22:50.357648Z"},"trusted":true},"execution_count":10,"outputs":[]},{"cell_type":"code","source":"images","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:50.359941Z","iopub.execute_input":"2022-11-25T20:22:50.360306Z","iopub.status.idle":"2022-11-25T20:22:50.384510Z","shell.execute_reply.started":"2022-11-25T20:22:50.360278Z","shell.execute_reply":"2022-11-25T20:22:50.383864Z"},"trusted":true},"execution_count":11,"outputs":[{"execution_count":11,"output_type":"execute_result","data":{"text/plain":"                                             filenames given_names image_type\n0    /kaggle/input/selfies-id-images-dataset/11_set...    Matheus      Selfie\n1    /kaggle/input/selfies-id-images-dataset/11_set...    Matheus          ID\n2    /kaggle/input/selfies-id-images-dataset/11_set...    Matheus      Selfie\n3    /kaggle/input/selfies-id-images-dataset/11_set...    Matheus      Selfie\n4    /kaggle/input/selfies-id-images-dataset/11_set...    Matheus          ID\n..                                                 ...         ...        ...\n428  /kaggle/input/selfies-id-images-dataset/18_set...  Anastasia      Selfie\n429  /kaggle/input/selfies-id-images-dataset/18_set...  Anastasia      Selfie\n430  /kaggle/input/selfies-id-images-dataset/18_set...  Anastasia      Selfie\n431  /kaggle/input/selfies-id-images-dataset/18_set...  Anastasia      Selfie\n432  /kaggle/input/selfies-id-images-dataset/18_set...  Anastasia      Selfie\n\n[433 rows x 3 columns]","text/html":"<div>\n<style scoped>\n    .dataframe tbody tr th:only-of-type {\n        vertical-align: middle;\n    }\n\n    .dataframe tbody tr th {\n        vertical-align: top;\n    }\n\n    .dataframe thead th {\n        text-align: right;\n    }\n</style>\n<table border=\"1\" class=\"dataframe\">\n  <thead>\n    <tr style=\"text-align: right;\">\n      <th></th>\n      <th>filenames</th>\n      <th>given_names</th>\n      <th>image_type</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>/kaggle/input/selfies-id-images-dataset/11_set...</td>\n      <td>Matheus</td>\n      <td>Selfie</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>/kaggle/input/selfies-id-images-dataset/11_set...</td>\n      <td>Matheus</td>\n      <td>ID</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>/kaggle/input/selfies-id-images-dataset/11_set...</td>\n      <td>Matheus</td>\n      <td>Selfie</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>/kaggle/input/selfies-id-images-dataset/11_set...</td>\n      <td>Matheus</td>\n      <td>Selfie</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>/kaggle/input/selfies-id-images-dataset/11_set...</td>\n      <td>Matheus</td>\n      <td>ID</td>\n    </tr>\n    <tr>\n      <th>...</th>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n    </tr>\n    <tr>\n      <th>428</th>\n      <td>/kaggle/input/selfies-id-images-dataset/18_set...</td>\n      <td>Anastasia</td>\n      <td>Selfie</td>\n    </tr>\n    <tr>\n      <th>429</th>\n      <td>/kaggle/input/selfies-id-images-dataset/18_set...</td>\n      <td>Anastasia</td>\n      <td>Selfie</td>\n    </tr>\n    <tr>\n      <th>430</th>\n      <td>/kaggle/input/selfies-id-images-dataset/18_set...</td>\n      <td>Anastasia</td>\n      <td>Selfie</td>\n    </tr>\n    <tr>\n      <th>431</th>\n      <td>/kaggle/input/selfies-id-images-dataset/18_set...</td>\n      <td>Anastasia</td>\n      <td>Selfie</td>\n    </tr>\n    <tr>\n      <th>432</th>\n      <td>/kaggle/input/selfies-id-images-dataset/18_set...</td>\n      <td>Anastasia</td>\n      <td>Selfie</td>\n    </tr>\n  </tbody>\n</table>\n<p>433 rows × 3 columns</p>\n</div>"},"metadata":{}}]},{"cell_type":"code","source":"def build_facematch_dataset(images):\n    X1_dataset = []\n    X2_dataset = []\n    y1_dataset = []\n    y2_dataset = []\n    for i in range(len(images)-2):\n        X1_dataset.append(images.filenames[i])\n        X2_dataset.append(images.filenames[i+1])\n        if images.given_names[i] == images.given_names[i+1]:\n            y1_dataset.append(1)\n        else:\n            y1_dataset.append(0)\n        if images.image_type[i+1] == 'ID':\n            y2_dataset.append(0)\n        else:\n            y2_dataset.append(1)\n    for i in range(len(images)):\n        first_sample = np.random.randint(0,len(images))\n        second_sample = np.random.randint(0,len(images))\n        X1_dataset.append(images.filenames[first_sample])\n        X2_dataset.append(images.filenames[second_sample])\n        if images.given_names[first_sample] == images.given_names[second_sample]:\n            y1_dataset.append(1)\n        else:\n            y1_dataset.append(0)\n        if images.image_type[second_sample] == 'ID':\n            y2_dataset.append(0)\n        else:\n            y2_dataset.append(1)\n        \n    return tf.data.Dataset.from_tensor_slices(({\"input_1\":X1_dataset,\"input_2\":X2_dataset},\n                    {\"facematch\":y1_dataset,\"liveness\":y2_dataset}))","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:50.385731Z","iopub.execute_input":"2022-11-25T20:22:50.385981Z","iopub.status.idle":"2022-11-25T20:22:50.396622Z","shell.execute_reply.started":"2022-11-25T20:22:50.385953Z","shell.execute_reply":"2022-11-25T20:22:50.395545Z"},"trusted":true},"execution_count":12,"outputs":[]},{"cell_type":"code","source":"dataset = build_facematch_dataset(images)","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:50.398244Z","iopub.execute_input":"2022-11-25T20:22:50.398733Z","iopub.status.idle":"2022-11-25T20:22:50.479376Z","shell.execute_reply.started":"2022-11-25T20:22:50.398691Z","shell.execute_reply":"2022-11-25T20:22:50.478391Z"},"trusted":true},"execution_count":13,"outputs":[]},{"cell_type":"code","source":"face_cascade = cv.CascadeClassifier(cv.data.haarcascades + 'haarcascade_frontalface_default.xml')","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:50.480535Z","iopub.execute_input":"2022-11-25T20:22:50.480769Z","iopub.status.idle":"2022-11-25T20:22:50.517179Z","shell.execute_reply.started":"2022-11-25T20:22:50.480744Z","shell.execute_reply":"2022-11-25T20:22:50.516294Z"},"trusted":true},"execution_count":14,"outputs":[]},{"cell_type":"code","source":"def get_face_boxes(image):\n    height, width = image.shape[0], image.shape[1]\n    global face_cascade\n    face = face_cascade.detectMultiScale(image.numpy(), 1.3, 5)\n    for (x,y,w,h) in face:\n        return x/width, y/height, (x+w)/width, (y+h)/height","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:50.520565Z","iopub.execute_input":"2022-11-25T20:22:50.520820Z","iopub.status.idle":"2022-11-25T20:22:50.526701Z","shell.execute_reply.started":"2022-11-25T20:22:50.520791Z","shell.execute_reply":"2022-11-25T20:22:50.525774Z"},"trusted":true},"execution_count":15,"outputs":[]},{"cell_type":"code","source":"def process_path(img):\n    with open(img, \"rb\") as local_file: # <= change here\n        image = local_file.read()\n    image = tf.io.decode_jpeg(image, channels=1)\n    \n    try:\n        x1, y1, x2, y2 = get_face_boxes(image)\n        image = tf.image.crop_and_resize(tf.expand_dims(image, axis=0),\n                                          tf.constant([[y1, x1, y2, x2]], dtype=tf.float32),\n                                          tf.constant([0]), [192, 192, 1])\n        image = tf.squeeze(image, axis=0)\n    except:\n        image = tf.image.resize(image, [192, 192, 1])\n    \n    image = image/255\n    return image","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:50.527841Z","iopub.execute_input":"2022-11-25T20:22:50.528124Z","iopub.status.idle":"2022-11-25T20:22:50.539754Z","shell.execute_reply.started":"2022-11-25T20:22:50.528073Z","shell.execute_reply":"2022-11-25T20:22:50.538597Z"},"trusted":true},"execution_count":16,"outputs":[]},{"cell_type":"code","source":"def parse_func(img):\n    return tf.py_function(process_path,[img],tf.float32)","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:50.541086Z","iopub.execute_input":"2022-11-25T20:22:50.541329Z","iopub.status.idle":"2022-11-25T20:22:50.554888Z","shell.execute_reply.started":"2022-11-25T20:22:50.541302Z","shell.execute_reply":"2022-11-25T20:22:50.554247Z"},"trusted":true},"execution_count":17,"outputs":[]},{"cell_type":"code","source":"dataset = dataset.map(lambda images, labels: \n                      ({\"input_1\":parse_func(images[\"input_1\"]),\n                        \"input_2\":parse_func(images[\"input_2\"])},\n                       labels))","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:50.557560Z","iopub.execute_input":"2022-11-25T20:22:50.558088Z","iopub.status.idle":"2022-11-25T20:22:50.632956Z","shell.execute_reply.started":"2022-11-25T20:22:50.558039Z","shell.execute_reply":"2022-11-25T20:22:50.632262Z"},"trusted":true},"execution_count":18,"outputs":[]},{"cell_type":"code","source":"@tf.function\ndef rot90_augment(images, labels):\n    img1 = tf.image.rot90(images[\"input_1\"],k=1)\n    img2 = tf.image.rot90(images[\"input_2\"],k=1)\n    return {\"input_1\":img1,\"input_2\":img2},labels","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:50.634157Z","iopub.execute_input":"2022-11-25T20:22:50.634970Z","iopub.status.idle":"2022-11-25T20:22:50.641717Z","shell.execute_reply.started":"2022-11-25T20:22:50.634926Z","shell.execute_reply":"2022-11-25T20:22:50.640875Z"},"trusted":true},"execution_count":19,"outputs":[]},{"cell_type":"code","source":"@tf.function\ndef rot180_augment(images, labels):\n    img1 = tf.image.rot90(images[\"input_1\"],k=2)\n    img2 = tf.image.rot90(images[\"input_2\"],k=2)\n    return {\"input_1\":img1,\"input_2\":img2},labels","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:50.643269Z","iopub.execute_input":"2022-11-25T20:22:50.643632Z","iopub.status.idle":"2022-11-25T20:22:50.654410Z","shell.execute_reply.started":"2022-11-25T20:22:50.643601Z","shell.execute_reply":"2022-11-25T20:22:50.653338Z"},"trusted":true},"execution_count":20,"outputs":[]},{"cell_type":"code","source":"@tf.function\ndef rot270_augment(images, labels):\n    img1 = tf.image.rot90(images[\"input_1\"],k=3)\n    img2 = tf.image.rot90(images[\"input_2\"],k=3)\n    return {\"input_1\":img1,\"input_2\":img2},labels","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:50.655425Z","iopub.execute_input":"2022-11-25T20:22:50.655943Z","iopub.status.idle":"2022-11-25T20:22:50.667877Z","shell.execute_reply.started":"2022-11-25T20:22:50.655911Z","shell.execute_reply":"2022-11-25T20:22:50.666659Z"},"trusted":true},"execution_count":21,"outputs":[]},{"cell_type":"code","source":"@tf.function\ndef left_right_augment(images, labels):\n    img1 = tf.image.flip_left_right(images[\"input_1\"])\n    img2 = tf.image.flip_left_right(images[\"input_2\"])\n    return {\"input_1\":img1,\"input_2\":img2},labels","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:50.669340Z","iopub.execute_input":"2022-11-25T20:22:50.669866Z","iopub.status.idle":"2022-11-25T20:22:50.679455Z","shell.execute_reply.started":"2022-11-25T20:22:50.669829Z","shell.execute_reply":"2022-11-25T20:22:50.678398Z"},"trusted":true},"execution_count":22,"outputs":[]},{"cell_type":"code","source":"@tf.function\ndef up_down_augment(images, labels):\n    img1 = tf.image.flip_left_right(images[\"input_1\"])\n    img2 = tf.image.flip_left_right(images[\"input_2\"])\n    return {\"input_1\":img1,\"input_2\":img2},labels","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:50.680902Z","iopub.execute_input":"2022-11-25T20:22:50.681269Z","iopub.status.idle":"2022-11-25T20:22:50.692879Z","shell.execute_reply.started":"2022-11-25T20:22:50.681229Z","shell.execute_reply":"2022-11-25T20:22:50.691808Z"},"trusted":true},"execution_count":23,"outputs":[]},{"cell_type":"code","source":"rot90_dataset = dataset.map(rot90_augment)\nrot180_dataset = dataset.map(rot180_augment)\nrot270_dataset = dataset.map(rot270_augment)\nleft_right_dataset = dataset.map(left_right_augment)\nup_down_dataset = dataset.map(up_down_augment)","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:50.694168Z","iopub.execute_input":"2022-11-25T20:22:50.694458Z","iopub.status.idle":"2022-11-25T20:22:51.894285Z","shell.execute_reply.started":"2022-11-25T20:22:50.694428Z","shell.execute_reply":"2022-11-25T20:22:51.893377Z"},"trusted":true},"execution_count":24,"outputs":[]},{"cell_type":"code","source":"dataset = dataset.concatenate(rot90_dataset)\ndataset = dataset.concatenate(rot180_dataset)\ndataset = dataset.concatenate(rot270_dataset)\ndataset = dataset.concatenate(left_right_dataset)\ndataset = dataset.concatenate(up_down_dataset)","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:51.895392Z","iopub.execute_input":"2022-11-25T20:22:51.895619Z","iopub.status.idle":"2022-11-25T20:22:51.906254Z","shell.execute_reply.started":"2022-11-25T20:22:51.895586Z","shell.execute_reply":"2022-11-25T20:22:51.905071Z"},"trusted":true},"execution_count":25,"outputs":[]},{"cell_type":"code","source":"dataset.__len__()","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:51.907454Z","iopub.execute_input":"2022-11-25T20:22:51.907810Z","iopub.status.idle":"2022-11-25T20:22:51.925863Z","shell.execute_reply.started":"2022-11-25T20:22:51.907702Z","shell.execute_reply":"2022-11-25T20:22:51.924835Z"},"trusted":true},"execution_count":26,"outputs":[{"execution_count":26,"output_type":"execute_result","data":{"text/plain":"<tf.Tensor: shape=(), dtype=int64, numpy=5184>"},"metadata":{}}]},{"cell_type":"code","source":"def get_dataset_partitions_tf(ds, ds_size, train_split=0.8, val_split=0.1, test_split=0.1, shuffle=True, shuffle_size=100):\n    assert (train_split + test_split + val_split) == 1\n    \n    if shuffle:\n        # Specify seed to always have the same split distribution between runs\n        ds = ds.shuffle(shuffle_size)\n    \n    train_size = int(train_split * ds_size)\n    val_size = int(val_split * ds_size)\n    \n    train_ds = ds.take(train_size)    \n    val_ds = ds.skip(train_size).take(val_size)\n    test_ds = ds.skip(train_size).skip(val_size)\n    \n    return train_ds, val_ds, test_ds","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:51.927259Z","iopub.execute_input":"2022-11-25T20:22:51.927581Z","iopub.status.idle":"2022-11-25T20:22:51.934616Z","shell.execute_reply.started":"2022-11-25T20:22:51.927551Z","shell.execute_reply":"2022-11-25T20:22:51.933568Z"},"trusted":true},"execution_count":27,"outputs":[]},{"cell_type":"code","source":"train, val, test = get_dataset_partitions_tf(dataset,dataset.__len__().numpy())","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:51.935825Z","iopub.execute_input":"2022-11-25T20:22:51.936073Z","iopub.status.idle":"2022-11-25T20:22:51.953808Z","shell.execute_reply.started":"2022-11-25T20:22:51.936038Z","shell.execute_reply":"2022-11-25T20:22:51.952712Z"},"trusted":true},"execution_count":28,"outputs":[]},{"cell_type":"code","source":"STEPS_PER_EPOCH = train.__len__()//BATCH_SIZE","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:51.955413Z","iopub.execute_input":"2022-11-25T20:22:51.955703Z","iopub.status.idle":"2022-11-25T20:22:51.964650Z","shell.execute_reply.started":"2022-11-25T20:22:51.955673Z","shell.execute_reply":"2022-11-25T20:22:51.963493Z"},"trusted":true},"execution_count":29,"outputs":[]},{"cell_type":"code","source":"train = train.shuffle(100).batch(BATCH_SIZE).prefetch(AUTO)\nval = val.shuffle(100).batch(BATCH_SIZE).prefetch(AUTO)\ntest = test.shuffle(100).batch(BATCH_SIZE).prefetch(AUTO)","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:51.966248Z","iopub.execute_input":"2022-11-25T20:22:51.966654Z","iopub.status.idle":"2022-11-25T20:22:51.982778Z","shell.execute_reply.started":"2022-11-25T20:22:51.966614Z","shell.execute_reply":"2022-11-25T20:22:51.981557Z"},"trusted":true},"execution_count":30,"outputs":[]},{"cell_type":"code","source":"from tensorflow import keras","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:51.983898Z","iopub.execute_input":"2022-11-25T20:22:51.984148Z","iopub.status.idle":"2022-11-25T20:22:51.988316Z","shell.execute_reply.started":"2022-11-25T20:22:51.984114Z","shell.execute_reply":"2022-11-25T20:22:51.987373Z"},"trusted":true},"execution_count":31,"outputs":[]},{"cell_type":"code","source":"def input_layers(image_size):\n    model = keras.models.Sequential([\n        keras.layers.Input(shape=[*image_size, 1]),\n        keras.layers.Conv2D(32, 3, activation='relu', padding='same'),\n        keras.layers.MaxPooling2D(3),\n        keras.layers.Conv2D(96, 3, activation='relu', padding='same'),\n        keras.layers.Conv2D(96, 3, activation='relu', padding='same'),\n        keras.layers.MaxPooling2D(3),\n        keras.layers.Flatten()\n    ])\n    \n    return model","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:51.989427Z","iopub.execute_input":"2022-11-25T20:22:51.989924Z","iopub.status.idle":"2022-11-25T20:22:52.001246Z","shell.execute_reply.started":"2022-11-25T20:22:51.989892Z","shell.execute_reply":"2022-11-25T20:22:52.000316Z"},"trusted":true},"execution_count":32,"outputs":[]},{"cell_type":"code","source":"def build_model():\n    base_model_A = input_layers(IMAGE_SIZE)\n    base_model_B = input_layers(IMAGE_SIZE)\n\n    concat = keras.layers.Concatenate()([base_model_A.output,base_model_B.output])\n    z = keras.layers.Dense(20,activation='relu')(concat)\n    bn = keras.layers.BatchNormalization()(z)\n    z = keras.layers.Dense(20,activation='relu')(bn)\n    output_A = keras.layers.Dense(1,activation='sigmoid',name='facematch')(z)\n    output_B = keras.layers.Dense(1,activation='sigmoid',name='liveness')(z)\n\n    model = keras.Model(inputs=[base_model_A.input,base_model_B.input],outputs=[output_A,output_B])\n    \n    return model","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:52.002443Z","iopub.execute_input":"2022-11-25T20:22:52.002675Z","iopub.status.idle":"2022-11-25T20:22:52.013572Z","shell.execute_reply.started":"2022-11-25T20:22:52.002650Z","shell.execute_reply":"2022-11-25T20:22:52.012914Z"},"trusted":true},"execution_count":33,"outputs":[]},{"cell_type":"code","source":"with strategy.scope():\n    model = build_model()\n    \n    model.layers[0]._name = 'input_1'\n    model.layers[1]._name = 'input_2'\n    \n    optimizer = keras.optimizers.Nadam()\n    model.compile(loss={'facematch':\"binary_crossentropy\",'liveness': \"binary_crossentropy\"},\n              optimizer=optimizer,metrics={'facematch': 'binary_accuracy', 'liveness': 'binary_accuracy'})","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:52.015093Z","iopub.execute_input":"2022-11-25T20:22:52.015377Z","iopub.status.idle":"2022-11-25T20:22:53.011932Z","shell.execute_reply.started":"2022-11-25T20:22:52.015348Z","shell.execute_reply":"2022-11-25T20:22:53.011184Z"},"trusted":true},"execution_count":34,"outputs":[]},{"cell_type":"code","source":"model.summary()","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:53.015046Z","iopub.execute_input":"2022-11-25T20:22:53.015474Z","iopub.status.idle":"2022-11-25T20:22:53.031821Z","shell.execute_reply.started":"2022-11-25T20:22:53.015435Z","shell.execute_reply":"2022-11-25T20:22:53.029552Z"},"trusted":true},"execution_count":35,"outputs":[{"name":"stdout","text":"Model: \"model\"\n__________________________________________________________________________________________________\nLayer (type)                    Output Shape         Param #     Connected to                     \n==================================================================================================\ninput_1 (InputLayer)            [(None, 192, 192, 1) 0                                            \n__________________________________________________________________________________________________\ninput_2 (InputLayer)            [(None, 192, 192, 1) 0                                            \n__________________________________________________________________________________________________\nconv2d (Conv2D)                 (None, 192, 192, 32) 320         input_1[0][0]                    \n__________________________________________________________________________________________________\nconv2d_3 (Conv2D)               (None, 192, 192, 32) 320         input_2[0][0]                    \n__________________________________________________________________________________________________\nmax_pooling2d (MaxPooling2D)    (None, 64, 64, 32)   0           conv2d[0][0]                     \n__________________________________________________________________________________________________\nmax_pooling2d_2 (MaxPooling2D)  (None, 64, 64, 32)   0           conv2d_3[0][0]                   \n__________________________________________________________________________________________________\nconv2d_1 (Conv2D)               (None, 64, 64, 96)   27744       max_pooling2d[0][0]              \n__________________________________________________________________________________________________\nconv2d_4 (Conv2D)               (None, 64, 64, 96)   27744       max_pooling2d_2[0][0]            \n__________________________________________________________________________________________________\nconv2d_2 (Conv2D)               (None, 64, 64, 96)   83040       conv2d_1[0][0]                   \n__________________________________________________________________________________________________\nconv2d_5 (Conv2D)               (None, 64, 64, 96)   83040       conv2d_4[0][0]                   \n__________________________________________________________________________________________________\nmax_pooling2d_1 (MaxPooling2D)  (None, 21, 21, 96)   0           conv2d_2[0][0]                   \n__________________________________________________________________________________________________\nmax_pooling2d_3 (MaxPooling2D)  (None, 21, 21, 96)   0           conv2d_5[0][0]                   \n__________________________________________________________________________________________________\nflatten (Flatten)               (None, 42336)        0           max_pooling2d_1[0][0]            \n__________________________________________________________________________________________________\nflatten_1 (Flatten)             (None, 42336)        0           max_pooling2d_3[0][0]            \n__________________________________________________________________________________________________\nconcatenate (Concatenate)       (None, 84672)        0           flatten[0][0]                    \n                                                                 flatten_1[0][0]                  \n__________________________________________________________________________________________________\ndense (Dense)                   (None, 20)           1693460     concatenate[0][0]                \n__________________________________________________________________________________________________\nbatch_normalization (BatchNorma (None, 20)           80          dense[0][0]                      \n__________________________________________________________________________________________________\ndense_1 (Dense)                 (None, 20)           420         batch_normalization[0][0]        \n__________________________________________________________________________________________________\nfacematch (Dense)               (None, 1)            21          dense_1[0][0]                    \n__________________________________________________________________________________________________\nliveness (Dense)                (None, 1)            21          dense_1[0][0]                    \n==================================================================================================\nTotal params: 1,916,210\nTrainable params: 1,916,170\nNon-trainable params: 40\n__________________________________________________________________________________________________\n","output_type":"stream"}]},{"cell_type":"code","source":"train.__len__()","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:53.033232Z","iopub.execute_input":"2022-11-25T20:22:53.033504Z","iopub.status.idle":"2022-11-25T20:22:53.044159Z","shell.execute_reply.started":"2022-11-25T20:22:53.033476Z","shell.execute_reply":"2022-11-25T20:22:53.043119Z"},"trusted":true},"execution_count":36,"outputs":[{"execution_count":36,"output_type":"execute_result","data":{"text/plain":"<tf.Tensor: shape=(), dtype=int64, numpy=33>"},"metadata":{}}]},{"cell_type":"code","source":"history = model.fit(train,epochs=EPOCHS,steps_per_epoch=STEPS_PER_EPOCH,validation_data=(val))","metadata":{"_kg_hide-output":true,"execution":{"iopub.status.busy":"2022-11-25T20:22:53.045625Z","iopub.execute_input":"2022-11-25T20:22:53.045941Z","iopub.status.idle":"2022-11-25T20:22:54.512176Z","shell.execute_reply.started":"2022-11-25T20:22:53.045908Z","shell.execute_reply":"2022-11-25T20:22:54.510732Z"},"trusted":true},"execution_count":37,"outputs":[{"name":"stdout","text":"Epoch 1/5\n","output_type":"stream"},{"traceback":["\u001b[0;31m---------------------------------------------------------------------------\u001b[0m","\u001b[0;31mTypeError\u001b[0m                                 Traceback (most recent call last)","\u001b[0;32m/tmp/ipykernel_21/2489583751.py\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[0;32m----> 1\u001b[0;31m \u001b[0mhistory\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mmodel\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtrain\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mepochs\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mEPOCHS\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0msteps_per_epoch\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mSTEPS_PER_EPOCH\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0mvalidation_data\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mval\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m","\u001b[0;32m/opt/conda/lib/python3.7/site-packages/tensorflow/python/keras/engine/training.py\u001b[0m in \u001b[0;36mfit\u001b[0;34m(self, x, y, batch_size, epochs, verbose, callbacks, validation_split, validation_data, shuffle, class_weight, sample_weight, initial_epoch, steps_per_epoch, validation_steps, validation_batch_size, validation_freq, max_queue_size, workers, use_multiprocessing)\u001b[0m\n\u001b[1;32m   1098\u001b[0m                 _r=1):\n\u001b[1;32m   1099\u001b[0m               \u001b[0mcallbacks\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mon_train_batch_begin\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mstep\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1100\u001b[0;31m               \u001b[0mtmp_logs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtrain_function\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0miterator\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   1101\u001b[0m               \u001b[0;32mif\u001b[0m \u001b[0mdata_handler\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mshould_sync\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1102\u001b[0m                 \u001b[0mcontext\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0masync_wait\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/opt/conda/lib/python3.7/site-packages/tensorflow/python/eager/def_function.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, *args, **kwds)\u001b[0m\n\u001b[1;32m    826\u001b[0m     \u001b[0mtracing_count\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mexperimental_get_tracing_count\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    827\u001b[0m     \u001b[0;32mwith\u001b[0m \u001b[0mtrace\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mTrace\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_name\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0mtm\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 828\u001b[0;31m       \u001b[0mresult\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwds\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    829\u001b[0m       \u001b[0mcompiler\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m\"xla\"\u001b[0m \u001b[0;32mif\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_experimental_compile\u001b[0m \u001b[0;32melse\u001b[0m \u001b[0;34m\"nonXla\"\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    830\u001b[0m       \u001b[0mnew_tracing_count\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mexperimental_get_tracing_count\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/opt/conda/lib/python3.7/site-packages/tensorflow/python/eager/def_function.py\u001b[0m in \u001b[0;36m_call\u001b[0;34m(self, *args, **kwds)\u001b[0m\n\u001b[1;32m    869\u001b[0m       \u001b[0;31m# This is the first call of __call__, so we have to initialize.\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    870\u001b[0m       \u001b[0minitializers\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 871\u001b[0;31m       \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_initialize\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mkwds\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0madd_initializers_to\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0minitializers\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    872\u001b[0m     \u001b[0;32mfinally\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    873\u001b[0m       \u001b[0;31m# At this point we know that the initialization is complete (or less\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/opt/conda/lib/python3.7/site-packages/tensorflow/python/eager/def_function.py\u001b[0m in \u001b[0;36m_initialize\u001b[0;34m(self, args, kwds, add_initializers_to)\u001b[0m\n\u001b[1;32m    724\u001b[0m     self._concrete_stateful_fn = (\n\u001b[1;32m    725\u001b[0m         self._stateful_fn._get_concrete_function_internal_garbage_collected(  # pylint: disable=protected-access\n\u001b[0;32m--> 726\u001b[0;31m             *args, **kwds))\n\u001b[0m\u001b[1;32m    727\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    728\u001b[0m     \u001b[0;32mdef\u001b[0m \u001b[0minvalid_creator_scope\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0munused_args\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0munused_kwds\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/opt/conda/lib/python3.7/site-packages/tensorflow/python/eager/function.py\u001b[0m in \u001b[0;36m_get_concrete_function_internal_garbage_collected\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m   2967\u001b[0m       \u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mkwargs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2968\u001b[0m     \u001b[0;32mwith\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_lock\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2969\u001b[0;31m       \u001b[0mgraph_function\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0m_\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_maybe_define_function\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   2970\u001b[0m     \u001b[0;32mreturn\u001b[0m \u001b[0mgraph_function\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2971\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/opt/conda/lib/python3.7/site-packages/tensorflow/python/eager/function.py\u001b[0m in \u001b[0;36m_maybe_define_function\u001b[0;34m(self, args, kwargs)\u001b[0m\n\u001b[1;32m   3359\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   3360\u001b[0m           \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_function_cache\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mmissed\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0madd\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mcall_context_key\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 3361\u001b[0;31m           \u001b[0mgraph_function\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_create_graph_function\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mkwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   3362\u001b[0m           \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_function_cache\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mprimary\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mcache_key\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mgraph_function\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   3363\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/opt/conda/lib/python3.7/site-packages/tensorflow/python/eager/function.py\u001b[0m in \u001b[0;36m_create_graph_function\u001b[0;34m(self, args, kwargs, override_flat_arg_shapes)\u001b[0m\n\u001b[1;32m   3204\u001b[0m             \u001b[0marg_names\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0marg_names\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   3205\u001b[0m             \u001b[0moverride_flat_arg_shapes\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0moverride_flat_arg_shapes\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 3206\u001b[0;31m             capture_by_value=self._capture_by_value),\n\u001b[0m\u001b[1;32m   3207\u001b[0m         \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_function_attributes\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   3208\u001b[0m         \u001b[0mfunction_spec\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfunction_spec\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/opt/conda/lib/python3.7/site-packages/tensorflow/python/framework/func_graph.py\u001b[0m in \u001b[0;36mfunc_graph_from_py_func\u001b[0;34m(name, python_func, args, kwargs, signature, func_graph, autograph, autograph_options, add_control_dependencies, arg_names, op_return_value, collections, capture_by_value, override_flat_arg_shapes)\u001b[0m\n\u001b[1;32m    988\u001b[0m         \u001b[0m_\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0moriginal_func\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtf_decorator\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0munwrap\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mpython_func\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    989\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 990\u001b[0;31m       \u001b[0mfunc_outputs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mpython_func\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0mfunc_args\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mfunc_kwargs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    991\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    992\u001b[0m       \u001b[0;31m# invariant: `func_outputs` contains only Tensors, CompositeTensors,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/opt/conda/lib/python3.7/site-packages/tensorflow/python/eager/def_function.py\u001b[0m in \u001b[0;36mwrapped_fn\u001b[0;34m(*args, **kwds)\u001b[0m\n\u001b[1;32m    632\u001b[0m             \u001b[0mxla_context\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mExit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    633\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 634\u001b[0;31m           \u001b[0mout\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mweak_wrapped_fn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m__wrapped__\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m**\u001b[0m\u001b[0mkwds\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    635\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mout\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    636\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;32m/opt/conda/lib/python3.7/site-packages/tensorflow/python/framework/func_graph.py\u001b[0m in \u001b[0;36mwrapper\u001b[0;34m(*args, **kwargs)\u001b[0m\n\u001b[1;32m    975\u001b[0m           \u001b[0;32mexcept\u001b[0m \u001b[0mException\u001b[0m \u001b[0;32mas\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m:\u001b[0m  \u001b[0;31m# pylint:disable=broad-except\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    976\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mhasattr\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0me\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;34m\"ag_error_metadata\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 977\u001b[0;31m               \u001b[0;32mraise\u001b[0m \u001b[0me\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mag_error_metadata\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mto_exception\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0me\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    978\u001b[0m             \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    979\u001b[0m               \u001b[0;32mraise\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n","\u001b[0;31mTypeError\u001b[0m: in user code:\n\n    /opt/conda/lib/python3.7/site-packages/tensorflow/python/keras/engine/training.py:805 train_function  *\n        return step_function(self, iterator)\n    /opt/conda/lib/python3.7/site-packages/tensorflow/python/keras/engine/training.py:795 step_function  **\n        outputs = model.distribute_strategy.run(run_step, args=(data,))\n    /opt/conda/lib/python3.7/site-packages/tensorflow/python/distribute/tpu_strategy.py:540 run\n        return self.extended.tpu_run(fn, args, kwargs, options)\n    /opt/conda/lib/python3.7/site-packages/tensorflow/python/distribute/tpu_strategy.py:1296 tpu_run\n        return func(args, kwargs)\n    /opt/conda/lib/python3.7/site-packages/tensorflow/python/distribute/tpu_strategy.py:1345 tpu_function\n        maximum_shape = tensor_shape.TensorShape([None] * rank)\n\n    TypeError: can't multiply sequence by non-int of type 'NoneType'\n"],"ename":"TypeError","evalue":"in user code:\n\n    /opt/conda/lib/python3.7/site-packages/tensorflow/python/keras/engine/training.py:805 train_function  *\n        return step_function(self, iterator)\n    /opt/conda/lib/python3.7/site-packages/tensorflow/python/keras/engine/training.py:795 step_function  **\n        outputs = model.distribute_strategy.run(run_step, args=(data,))\n    /opt/conda/lib/python3.7/site-packages/tensorflow/python/distribute/tpu_strategy.py:540 run\n        return self.extended.tpu_run(fn, args, kwargs, options)\n    /opt/conda/lib/python3.7/site-packages/tensorflow/python/distribute/tpu_strategy.py:1296 tpu_run\n        return func(args, kwargs)\n    /opt/conda/lib/python3.7/site-packages/tensorflow/python/distribute/tpu_strategy.py:1345 tpu_function\n        maximum_shape = tensor_shape.TensorShape([None] * rank)\n\n    TypeError: can't multiply sequence by non-int of type 'NoneType'\n","output_type":"error"}]},{"cell_type":"code","source":"model.evaluate(test)","metadata":{"execution":{"iopub.status.busy":"2022-11-25T20:22:54.513084Z","iopub.status.idle":"2022-11-25T20:22:54.513450Z","shell.execute_reply.started":"2022-11-25T20:22:54.513259Z","shell.execute_reply":"2022-11-25T20:22:54.513285Z"},"trusted":true},"execution_count":null,"outputs":[]}]}