Bangla Text Classification With CNN-Bi-LSTM Hybrid Network.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)\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\nimport os\nfor dirname, _, filenames in os.walk('/kaggle/input'):\n    for filename in filenames:\n        print(os.path.join(dirname, filename))\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":{"_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","papermill":{"duration":0.145271,"end_time":"2021-10-19T15:49:54.256874","exception":false,"start_time":"2021-10-19T15:49:54.111603","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T04:45:03.032056Z","iopub.execute_input":"2021-10-20T04:45:03.032402Z","iopub.status.idle":"2021-10-20T04:45:03.046677Z","shell.execute_reply.started":"2021-10-20T04:45:03.032355Z","shell.execute_reply":"2021-10-20T04:45:03.045789Z"},"trusted":true},"execution_count":19,"outputs":[{"name":"stdout","text":"/kaggle/input/banglamct7-bangla-multiclass-text-dataset-7-tags/BanglaMCT7/train.csv\n/kaggle/input/banglamct7-bangla-multiclass-text-dataset-7-tags/BanglaMCT7/test.csv\n/kaggle/input/bangla-stopwords/stopwords_bangla.xlsx\n","output_type":"stream"}]},{"cell_type":"markdown","source":"# **Updating/Instaling Libraries**","metadata":{"papermill":{"duration":0.029067,"end_time":"2021-10-19T15:49:54.318085","exception":false,"start_time":"2021-10-19T15:49:54.289018","status":"completed"},"tags":[]}},{"cell_type":"code","source":"!pip install keras\n!pip install --upgrade tensorflow\n!pip install --upgrade tensorflow-gpu\n!pip install xlrd==1.2.0\n!pip install openpyxl","metadata":{"papermill":{"duration":15.963129,"end_time":"2021-10-19T15:50:10.310307","exception":false,"start_time":"2021-10-19T15:49:54.347178","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T04:45:03.049330Z","iopub.execute_input":"2021-10-20T04:45:03.049751Z","iopub.status.idle":"2021-10-20T04:45:41.748331Z","shell.execute_reply.started":"2021-10-20T04:45:03.049711Z","shell.execute_reply":"2021-10-20T04:45:41.747467Z"},"trusted":true},"execution_count":20,"outputs":[{"name":"stdout","text":"Requirement already satisfied: keras in /opt/conda/lib/python3.7/site-packages (2.6.0)\n\u001b[33mWARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv\u001b[0m\nRequirement already satisfied: tensorflow in /opt/conda/lib/python3.7/site-packages (2.6.0)\nRequirement already satisfied: flatbuffers~=1.12.0 in /opt/conda/lib/python3.7/site-packages (from tensorflow) (1.12)\nRequirement already satisfied: grpcio<2.0,>=1.37.0 in /opt/conda/lib/python3.7/site-packages (from tensorflow) (1.38.1)\nRequirement already satisfied: absl-py~=0.10 in /opt/conda/lib/python3.7/site-packages (from tensorflow) (0.14.0)\nRequirement already satisfied: wrapt~=1.12.1 in /opt/conda/lib/python3.7/site-packages (from tensorflow) (1.12.1)\nRequirement already satisfied: six~=1.15.0 in /opt/conda/lib/python3.7/site-packages (from tensorflow) (1.15.0)\nRequirement already satisfied: clang~=5.0 in /opt/conda/lib/python3.7/site-packages (from tensorflow) (5.0)\nRequirement already satisfied: numpy~=1.19.2 in /opt/conda/lib/python3.7/site-packages (from tensorflow) (1.19.5)\nRequirement already satisfied: protobuf>=3.9.2 in /opt/conda/lib/python3.7/site-packages (from tensorflow) (3.18.1)\nRequirement already satisfied: keras~=2.6 in /opt/conda/lib/python3.7/site-packages (from tensorflow) (2.6.0)\nRequirement already satisfied: keras-preprocessing~=1.1.2 in /opt/conda/lib/python3.7/site-packages (from tensorflow) (1.1.2)\nRequirement already satisfied: google-pasta~=0.2 in /opt/conda/lib/python3.7/site-packages (from tensorflow) (0.2.0)\nRequirement already satisfied: tensorboard~=2.6 in /opt/conda/lib/python3.7/site-packages (from tensorflow) (2.6.0)\nRequirement already satisfied: h5py~=3.1.0 in /opt/conda/lib/python3.7/site-packages (from tensorflow) (3.1.0)\nRequirement already satisfied: opt-einsum~=3.3.0 in /opt/conda/lib/python3.7/site-packages (from tensorflow) (3.3.0)\nRequirement already satisfied: termcolor~=1.1.0 in /opt/conda/lib/python3.7/site-packages (from tensorflow) (1.1.0)\nRequirement already satisfied: astunparse~=1.6.3 in /opt/conda/lib/python3.7/site-packages (from tensorflow) (1.6.3)\nRequirement already satisfied: typing-extensions~=3.7.4 in /opt/conda/lib/python3.7/site-packages (from tensorflow) (3.7.4.3)\nRequirement already satisfied: gast==0.4.0 in /opt/conda/lib/python3.7/site-packages (from tensorflow) (0.4.0)\nRequirement already satisfied: wheel~=0.35 in /opt/conda/lib/python3.7/site-packages (from tensorflow) (0.37.0)\nRequirement already satisfied: tensorflow-estimator~=2.6 in /opt/conda/lib/python3.7/site-packages (from tensorflow) (2.6.0)\nRequirement already satisfied: cached-property in /opt/conda/lib/python3.7/site-packages (from h5py~=3.1.0->tensorflow) (1.5.2)\nRequirement already satisfied: markdown>=2.6.8 in /opt/conda/lib/python3.7/site-packages (from tensorboard~=2.6->tensorflow) (3.3.4)\nRequirement already satisfied: google-auth<2,>=1.6.3 in /opt/conda/lib/python3.7/site-packages (from tensorboard~=2.6->tensorflow) (1.35.0)\nRequirement already satisfied: google-auth-oauthlib<0.5,>=0.4.1 in /opt/conda/lib/python3.7/site-packages (from tensorboard~=2.6->tensorflow) (0.4.6)\nRequirement already satisfied: tensorboard-plugin-wit>=1.6.0 in /opt/conda/lib/python3.7/site-packages (from tensorboard~=2.6->tensorflow) (1.8.0)\nRequirement already satisfied: tensorboard-data-server<0.7.0,>=0.6.0 in /opt/conda/lib/python3.7/site-packages (from tensorboard~=2.6->tensorflow) (0.6.1)\nRequirement already satisfied: requests<3,>=2.21.0 in /opt/conda/lib/python3.7/site-packages (from tensorboard~=2.6->tensorflow) (2.25.1)\nRequirement already satisfied: setuptools>=41.0.0 in /opt/conda/lib/python3.7/site-packages (from tensorboard~=2.6->tensorflow) (58.0.4)\nRequirement already satisfied: werkzeug>=0.11.15 in /opt/conda/lib/python3.7/site-packages (from tensorboard~=2.6->tensorflow) (2.0.1)\nRequirement already satisfied: pyasn1-modules>=0.2.1 in /opt/conda/lib/python3.7/site-packages (from google-auth<2,>=1.6.3->tensorboard~=2.6->tensorflow) (0.2.7)\nRequirement already satisfied: cachetools<5.0,>=2.0.0 in /opt/conda/lib/python3.7/site-packages (from google-auth<2,>=1.6.3->tensorboard~=2.6->tensorflow) (4.2.2)\nRequirement already satisfied: rsa<5,>=3.1.4 in /opt/conda/lib/python3.7/site-packages (from google-auth<2,>=1.6.3->tensorboard~=2.6->tensorflow) (4.7.2)\nRequirement already satisfied: requests-oauthlib>=0.7.0 in /opt/conda/lib/python3.7/site-packages (from google-auth-oauthlib<0.5,>=0.4.1->tensorboard~=2.6->tensorflow) (1.3.0)\nRequirement already satisfied: importlib-metadata in /opt/conda/lib/python3.7/site-packages (from markdown>=2.6.8->tensorboard~=2.6->tensorflow) (4.8.1)\nRequirement already satisfied: pyasn1<0.5.0,>=0.4.6 in /opt/conda/lib/python3.7/site-packages (from pyasn1-modules>=0.2.1->google-auth<2,>=1.6.3->tensorboard~=2.6->tensorflow) (0.4.8)\nRequirement already satisfied: urllib3<1.27,>=1.21.1 in /opt/conda/lib/python3.7/site-packages (from requests<3,>=2.21.0->tensorboard~=2.6->tensorflow) (1.26.6)\nRequirement already satisfied: certifi>=2017.4.17 in /opt/conda/lib/python3.7/site-packages (from requests<3,>=2.21.0->tensorboard~=2.6->tensorflow) (2021.10.8)\nRequirement already satisfied: chardet<5,>=3.0.2 in /opt/conda/lib/python3.7/site-packages (from requests<3,>=2.21.0->tensorboard~=2.6->tensorflow) (4.0.0)\nRequirement already satisfied: idna<3,>=2.5 in /opt/conda/lib/python3.7/site-packages (from requests<3,>=2.21.0->tensorboard~=2.6->tensorflow) (2.10)\nRequirement already satisfied: oauthlib>=3.0.0 in /opt/conda/lib/python3.7/site-packages (from requests-oauthlib>=0.7.0->google-auth-oauthlib<0.5,>=0.4.1->tensorboard~=2.6->tensorflow) (3.1.1)\nRequirement already satisfied: zipp>=0.5 in /opt/conda/lib/python3.7/site-packages (from importlib-metadata->markdown>=2.6.8->tensorboard~=2.6->tensorflow) (3.5.0)\n\u001b[33mWARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv\u001b[0m\nRequirement already satisfied: tensorflow-gpu in /opt/conda/lib/python3.7/site-packages (2.6.0)\nRequirement already satisfied: h5py~=3.1.0 in /opt/conda/lib/python3.7/site-packages (from tensorflow-gpu) (3.1.0)\nRequirement already satisfied: astunparse~=1.6.3 in /opt/conda/lib/python3.7/site-packages (from tensorflow-gpu) (1.6.3)\nRequirement already satisfied: six~=1.15.0 in /opt/conda/lib/python3.7/site-packages (from tensorflow-gpu) (1.15.0)\nRequirement already satisfied: flatbuffers~=1.12.0 in /opt/conda/lib/python3.7/site-packages (from tensorflow-gpu) (1.12)\nRequirement already satisfied: opt-einsum~=3.3.0 in /opt/conda/lib/python3.7/site-packages (from tensorflow-gpu) (3.3.0)\nRequirement already satisfied: grpcio<2.0,>=1.37.0 in /opt/conda/lib/python3.7/site-packages (from tensorflow-gpu) (1.38.1)\nRequirement already satisfied: absl-py~=0.10 in /opt/conda/lib/python3.7/site-packages (from tensorflow-gpu) (0.14.0)\nRequirement already satisfied: tensorboard~=2.6 in /opt/conda/lib/python3.7/site-packages (from tensorflow-gpu) (2.6.0)\nRequirement already satisfied: numpy~=1.19.2 in /opt/conda/lib/python3.7/site-packages (from tensorflow-gpu) (1.19.5)\nRequirement already satisfied: clang~=5.0 in /opt/conda/lib/python3.7/site-packages (from tensorflow-gpu) (5.0)\nRequirement already satisfied: typing-extensions~=3.7.4 in /opt/conda/lib/python3.7/site-packages (from tensorflow-gpu) (3.7.4.3)\nRequirement already satisfied: termcolor~=1.1.0 in /opt/conda/lib/python3.7/site-packages (from tensorflow-gpu) (1.1.0)\nRequirement already satisfied: gast==0.4.0 in /opt/conda/lib/python3.7/site-packages (from tensorflow-gpu) (0.4.0)\nRequirement already satisfied: google-pasta~=0.2 in /opt/conda/lib/python3.7/site-packages (from tensorflow-gpu) (0.2.0)\nRequirement already satisfied: wrapt~=1.12.1 in /opt/conda/lib/python3.7/site-packages (from tensorflow-gpu) (1.12.1)\nRequirement already satisfied: protobuf>=3.9.2 in /opt/conda/lib/python3.7/site-packages (from tensorflow-gpu) (3.18.1)\nRequirement already satisfied: tensorflow-estimator~=2.6 in /opt/conda/lib/python3.7/site-packages (from tensorflow-gpu) (2.6.0)\nRequirement already satisfied: keras-preprocessing~=1.1.2 in /opt/conda/lib/python3.7/site-packages (from tensorflow-gpu) (1.1.2)\nRequirement already satisfied: wheel~=0.35 in /opt/conda/lib/python3.7/site-packages (from tensorflow-gpu) (0.37.0)\nRequirement already satisfied: keras~=2.6 in /opt/conda/lib/python3.7/site-packages (from tensorflow-gpu) (2.6.0)\nRequirement already satisfied: cached-property in /opt/conda/lib/python3.7/site-packages (from h5py~=3.1.0->tensorflow-gpu) (1.5.2)\nRequirement already satisfied: setuptools>=41.0.0 in /opt/conda/lib/python3.7/site-packages (from tensorboard~=2.6->tensorflow-gpu) (58.0.4)\nRequirement already satisfied: tensorboard-plugin-wit>=1.6.0 in /opt/conda/lib/python3.7/site-packages (from tensorboard~=2.6->tensorflow-gpu) (1.8.0)\nRequirement already satisfied: google-auth<2,>=1.6.3 in /opt/conda/lib/python3.7/site-packages (from tensorboard~=2.6->tensorflow-gpu) (1.35.0)\nRequirement already satisfied: werkzeug>=0.11.15 in /opt/conda/lib/python3.7/site-packages (from tensorboard~=2.6->tensorflow-gpu) (2.0.1)\nRequirement already satisfied: requests<3,>=2.21.0 in /opt/conda/lib/python3.7/site-packages (from tensorboard~=2.6->tensorflow-gpu) (2.25.1)\nRequirement already satisfied: google-auth-oauthlib<0.5,>=0.4.1 in /opt/conda/lib/python3.7/site-packages (from tensorboard~=2.6->tensorflow-gpu) (0.4.6)\nRequirement already satisfied: tensorboard-data-server<0.7.0,>=0.6.0 in /opt/conda/lib/python3.7/site-packages (from tensorboard~=2.6->tensorflow-gpu) (0.6.1)\nRequirement already satisfied: markdown>=2.6.8 in /opt/conda/lib/python3.7/site-packages (from tensorboard~=2.6->tensorflow-gpu) (3.3.4)\nRequirement already satisfied: cachetools<5.0,>=2.0.0 in /opt/conda/lib/python3.7/site-packages (from google-auth<2,>=1.6.3->tensorboard~=2.6->tensorflow-gpu) (4.2.2)\nRequirement already satisfied: pyasn1-modules>=0.2.1 in /opt/conda/lib/python3.7/site-packages (from google-auth<2,>=1.6.3->tensorboard~=2.6->tensorflow-gpu) (0.2.7)\nRequirement already satisfied: rsa<5,>=3.1.4 in /opt/conda/lib/python3.7/site-packages (from google-auth<2,>=1.6.3->tensorboard~=2.6->tensorflow-gpu) (4.7.2)\nRequirement already satisfied: requests-oauthlib>=0.7.0 in /opt/conda/lib/python3.7/site-packages (from google-auth-oauthlib<0.5,>=0.4.1->tensorboard~=2.6->tensorflow-gpu) (1.3.0)\nRequirement already satisfied: importlib-metadata in /opt/conda/lib/python3.7/site-packages (from markdown>=2.6.8->tensorboard~=2.6->tensorflow-gpu) (4.8.1)\nRequirement already satisfied: pyasn1<0.5.0,>=0.4.6 in /opt/conda/lib/python3.7/site-packages (from pyasn1-modules>=0.2.1->google-auth<2,>=1.6.3->tensorboard~=2.6->tensorflow-gpu) (0.4.8)\nRequirement already satisfied: chardet<5,>=3.0.2 in /opt/conda/lib/python3.7/site-packages (from requests<3,>=2.21.0->tensorboard~=2.6->tensorflow-gpu) (4.0.0)\nRequirement already satisfied: certifi>=2017.4.17 in /opt/conda/lib/python3.7/site-packages (from requests<3,>=2.21.0->tensorboard~=2.6->tensorflow-gpu) (2021.10.8)\nRequirement already satisfied: idna<3,>=2.5 in /opt/conda/lib/python3.7/site-packages (from requests<3,>=2.21.0->tensorboard~=2.6->tensorflow-gpu) (2.10)\nRequirement already satisfied: urllib3<1.27,>=1.21.1 in /opt/conda/lib/python3.7/site-packages (from requests<3,>=2.21.0->tensorboard~=2.6->tensorflow-gpu) (1.26.6)\nRequirement already satisfied: oauthlib>=3.0.0 in /opt/conda/lib/python3.7/site-packages (from requests-oauthlib>=0.7.0->google-auth-oauthlib<0.5,>=0.4.1->tensorboard~=2.6->tensorflow-gpu) (3.1.1)\nRequirement already satisfied: zipp>=0.5 in /opt/conda/lib/python3.7/site-packages (from importlib-metadata->markdown>=2.6.8->tensorboard~=2.6->tensorflow-gpu) (3.5.0)\n\u001b[33mWARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv\u001b[0m\nRequirement already satisfied: xlrd==1.2.0 in /opt/conda/lib/python3.7/site-packages (1.2.0)\n\u001b[33mWARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv\u001b[0m\nRequirement already satisfied: openpyxl in /opt/conda/lib/python3.7/site-packages (3.0.9)\nRequirement already satisfied: et-xmlfile in /opt/conda/lib/python3.7/site-packages (from openpyxl) (1.1.0)\n\u001b[33mWARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv\u001b[0m\n","output_type":"stream"}]},{"cell_type":"code","source":"import pandas as pd\nfrom pandas import read_excel\nimport numpy as np\nimport re\nfrom re import sub\nimport multiprocessing\nfrom unidecode import unidecode\nimport os\nfrom time import time \n\n\nfrom sklearn import metrics\nimport tensorflow as tf\nfrom tensorflow import keras\nfrom tensorflow.keras import layers\nfrom tensorflow.python.keras.layers import LSTM,Dense,Dropout,Activation,Embedding,Flatten,Bidirectional,MaxPooling2D, Conv1D, MaxPooling1D\nfrom tensorflow.python.keras.models import Sequential\n\nfrom keras.layers import MaxPool1D\n\nfrom tensorflow.keras.optimizers import SGD,Adam\nfrom tensorflow.keras import regularizers\n\nfrom tensorflow.keras.preprocessing.text import Tokenizer\nfrom tensorflow.keras.preprocessing.sequence import pad_sequences\nfrom tensorflow.keras.utils import to_categorical\n\n\nimport h5py\nimport csv\nimport seaborn as sns\nimport matplotlib.pyplot as plt\nfrom sklearn.model_selection import train_test_split\nfrom sklearn.model_selection import StratifiedKFold\nfrom sklearn.metrics import precision_recall_fscore_support\nfrom sklearn.metrics import accuracy_score\nfrom sklearn.metrics import classification_report \n\n\nfrom nltk.corpus import stopwords\nimport nltk\nimport re\nfrom nltk.stem import WordNetLemmatizer\nimport string\nfrom nltk.tokenize import word_tokenize\nfrom nltk.util import ngrams","metadata":{"papermill":{"duration":5.795641,"end_time":"2021-10-19T15:50:16.139673","exception":false,"start_time":"2021-10-19T15:50:10.344032","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T04:45:41.751147Z","iopub.execute_input":"2021-10-20T04:45:41.751815Z","iopub.status.idle":"2021-10-20T04:45:41.763234Z","shell.execute_reply.started":"2021-10-20T04:45:41.751774Z","shell.execute_reply":"2021-10-20T04:45:41.762458Z"},"trusted":true},"execution_count":21,"outputs":[]},{"cell_type":"markdown","source":"# **Loading Dataset**","metadata":{"papermill":{"duration":0.033638,"end_time":"2021-10-19T15:50:16.206566","exception":false,"start_time":"2021-10-19T15:50:16.172928","status":"completed"},"tags":[]}},{"cell_type":"code","source":"df_train=pd.read_csv(\"/kaggle/input/banglamct7-bangla-multiclass-text-dataset-7-tags/BanglaMCT7/train.csv\")\ndf_test=pd.read_csv(\"/kaggle/input/banglamct7-bangla-multiclass-text-dataset-7-tags/BanglaMCT7/test.csv\")\n","metadata":{"papermill":{"duration":32.6727,"end_time":"2021-10-19T15:50:48.912114","exception":false,"start_time":"2021-10-19T15:50:16.239414","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T04:45:41.764409Z","iopub.execute_input":"2021-10-20T04:45:41.764788Z","iopub.status.idle":"2021-10-20T04:46:03.449476Z","shell.execute_reply.started":"2021-10-20T04:45:41.764761Z","shell.execute_reply":"2021-10-20T04:46:03.448700Z"},"trusted":true},"execution_count":22,"outputs":[]},{"cell_type":"code","source":"\ndf_train['count'] = df_train['cleanText'].str.split().str.len()\ndf_test['count'] = df_test['cleanText'].str.split().str.len()\ndisplay(df_train)\ndf_train.head()\ndf_train.describe()\ndf_train.info()\nprint(df_train.category.unique())","metadata":{"papermill":{"duration":7.950669,"end_time":"2021-10-19T15:50:56.896111","exception":false,"start_time":"2021-10-19T15:50:48.945442","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T04:46:03.451912Z","iopub.execute_input":"2021-10-20T04:46:03.452192Z","iopub.status.idle":"2021-10-20T04:46:12.339453Z","shell.execute_reply.started":"2021-10-20T04:46:03.452156Z","shell.execute_reply":"2021-10-20T04:46:12.338656Z"},"trusted":true},"execution_count":23,"outputs":[{"output_type":"display_data","data":{"text/plain":"            category                                               text  \\\n0            economy   বাংলাদেশে সফররত বিশ্ব বাণিজ্য সংস্থা (ডব্লিউট...   \n1             sports  ইউরোপিয়ান ফুটবলে আজ ঝড় বয়ে যাচ্ছে। একদিকে সংবা...   \n2      international  বিজেপির নেতৃত্বে ভারতের উত্তর প্রদেশ থেকে শুরু...   \n3              state  নারায়ণগঞ্জের সোনারগাঁ উপজেলার একটি সরকারি প্র...   \n4             sports  ‘ছাই থেকে ছাই, ধুলো থেকে ধুলো...’ আধ্যাত্মিক ক...   \n...              ...                                                ...   \n98878      education    সহকারী অধ্যাপক  টিচার্স ট্রেনিং কলেজ (মহিলা)...   \n98879  entertainment  অক্টোবরের ১১ তারিখে ৭৫ বছর পূর্ণ করবেন অমিতাভ ...   \n98880     technology  কাজী নাঈম আল রশিদ। সফটওয়্যার প্রকৌশলী। কাজ করে...   \n98881     technology  ইয়াহুকে মার্কিন টেলিকম প্রতিষ্ঠান ভেরাইজন কিনে...   \n98882      education  অধ্যায়-১১প্রিয় শিক্ষার্থী, আজ ফিন্যান্স ও ব্যা...   \n\n                                               cleanText  count  \n0        বাংলাদেশে    সফররত    বিশ্ব    বাণিজ্য    সং...    293  \n1        ইউরোপিয়ান    ফুটবলে    বয়ে    একদিকে    সংবা...    206  \n2        বিজেপির    নেতৃত্বে    ভারতের    উত্তর    প্...    167  \n3        নারায়ণগঞ্জের    সোনারগাঁ    উপজেলার    সরকা...    181  \n4        ছাই    ছাই    ধুলো    ধুলো    আধ্যাত্মিক    ...    148  \n...                                                  ...    ...  \n98878    সহকারী    অধ্যাপক    টিচার্স    ট্রেনিং    ক...    389  \n98879    অক্টোবরের    তারিখে    বছর    পূর্ণ    অমিতা...    147  \n98880    কাজী    নাঈম    রশিদ    সফটওয়্যার    প্রকৌশল...    673  \n98881    ইয়াহুকে    মার্কিন    টেলিকম    প্রতিষ্ঠান  ...    161  \n98882    অধ্যায়    প্রিয়    শিক্ষার্থী    ফিন্যান্স  ...    204  \n\n[98883 rows x 4 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>category</th>\n      <th>text</th>\n      <th>cleanText</th>\n      <th>count</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>economy</td>\n      <td>বাংলাদেশে সফররত বিশ্ব বাণিজ্য সংস্থা (ডব্লিউট...</td>\n      <td>বাংলাদেশে    সফররত    বিশ্ব    বাণিজ্য    সং...</td>\n      <td>293</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>sports</td>\n      <td>ইউরোপিয়ান ফুটবলে আজ ঝড় বয়ে যাচ্ছে। একদিকে সংবা...</td>\n      <td>ইউরোপিয়ান    ফুটবলে    বয়ে    একদিকে    সংবা...</td>\n      <td>206</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>international</td>\n      <td>বিজেপির নেতৃত্বে ভারতের উত্তর প্রদেশ থেকে শুরু...</td>\n      <td>বিজেপির    নেতৃত্বে    ভারতের    উত্তর    প্...</td>\n      <td>167</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>state</td>\n      <td>নারায়ণগঞ্জের সোনারগাঁ উপজেলার একটি সরকারি প্র...</td>\n      <td>নারায়ণগঞ্জের    সোনারগাঁ    উপজেলার    সরকা...</td>\n      <td>181</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>sports</td>\n      <td>‘ছাই থেকে ছাই, ধুলো থেকে ধুলো...’ আধ্যাত্মিক ক...</td>\n      <td>ছাই    ছাই    ধুলো    ধুলো    আধ্যাত্মিক    ...</td>\n      <td>148</td>\n    </tr>\n    <tr>\n      <th>...</th>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n    </tr>\n    <tr>\n      <th>98878</th>\n      <td>education</td>\n      <td>সহকারী অধ্যাপক  টিচার্স ট্রেনিং কলেজ (মহিলা)...</td>\n      <td>সহকারী    অধ্যাপক    টিচার্স    ট্রেনিং    ক...</td>\n      <td>389</td>\n    </tr>\n    <tr>\n      <th>98879</th>\n      <td>entertainment</td>\n      <td>অক্টোবরের ১১ তারিখে ৭৫ বছর পূর্ণ করবেন অমিতাভ ...</td>\n      <td>অক্টোবরের    তারিখে    বছর    পূর্ণ    অমিতা...</td>\n      <td>147</td>\n    </tr>\n    <tr>\n      <th>98880</th>\n      <td>technology</td>\n      <td>কাজী নাঈম আল রশিদ। সফটওয়্যার প্রকৌশলী। কাজ করে...</td>\n      <td>কাজী    নাঈম    রশিদ    সফটওয়্যার    প্রকৌশল...</td>\n      <td>673</td>\n    </tr>\n    <tr>\n      <th>98881</th>\n      <td>technology</td>\n      <td>ইয়াহুকে মার্কিন টেলিকম প্রতিষ্ঠান ভেরাইজন কিনে...</td>\n      <td>ইয়াহুকে    মার্কিন    টেলিকম    প্রতিষ্ঠান  ...</td>\n      <td>161</td>\n    </tr>\n    <tr>\n      <th>98882</th>\n      <td>education</td>\n      <td>অধ্যায়-১১প্রিয় শিক্ষার্থী, আজ ফিন্যান্স ও ব্যা...</td>\n      <td>অধ্যায়    প্রিয়    শিক্ষার্থী    ফিন্যান্স  ...</td>\n      <td>204</td>\n    </tr>\n  </tbody>\n</table>\n<p>98883 rows × 4 columns</p>\n</div>"},"metadata":{}},{"name":"stdout","text":"<class 'pandas.core.frame.DataFrame'>\nRangeIndex: 98883 entries, 0 to 98882\nData columns (total 4 columns):\n #   Column     Non-Null Count  Dtype \n---  ------     --------------  ----- \n 0   category   98883 non-null  object\n 1   text       98883 non-null  object\n 2   cleanText  98883 non-null  object\n 3   count      98883 non-null  int64 \ndtypes: int64(1), object(3)\nmemory usage: 3.0+ MB\n['economy' 'sports' 'international' 'state' 'technology' 'entertainment'\n 'education']\n","output_type":"stream"}]},{"cell_type":"code","source":"display(df_test)\ndf_test.head()\ndf_test.describe()\ndf_test.info()","metadata":{"papermill":{"duration":0.105641,"end_time":"2021-10-19T15:50:57.036829","exception":false,"start_time":"2021-10-19T15:50:56.931188","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T04:46:12.340860Z","iopub.execute_input":"2021-10-20T04:46:12.341170Z","iopub.status.idle":"2021-10-20T04:46:12.915252Z","shell.execute_reply.started":"2021-10-20T04:46:12.341132Z","shell.execute_reply":"2021-10-20T04:46:12.914374Z"},"trusted":true},"execution_count":24,"outputs":[{"output_type":"display_data","data":{"text/plain":"            category                                               text  \\\n0          education  অধ্যায়-২প্রিয় পরীক্ষার্থী, আজ তথ্য ও যোগাযোগ প...   \n1             sports  স্পিন, স্পিন, স্পিন। লর্ডসে ইয়াসির শাহ্‌র লেগ ...   \n2      international  সিরিয়ার উত্তর-পশ্চিমাঞ্চলে গতকাল রোববার বিমান...   \n3      international  দক্ষিণ আফ্রিকার কেপটাউন বিশ্ববিদ্যালয়ের (ইউসিট...   \n4         technology  সহজে ডিজিটাল আর্থিক সেবা দিতে অ্যাপ ও ডিজিটাল ...   \n...              ...                                                ...   \n98879         sports  ১৯৫৯ সালে ইউরোপিয়ান ক্লাব কাপে জাতীয় দল পাঠান...   \n98880  entertainment  গানের দল সহজিয়ার প্রথম অ্যালবাম ‘রঙমিস্ত্রী’র ...   \n98881  international  ফেসবুকের ৮১ হাজার ব্যবহারকারীর তথ্য চুরি করে স...   \n98882        economy   পেট্রোল পাম্পে তেল বা সিএনজির বিল বিকাশের মাধ...   \n98883        economy  বছর শেষে ঢাকা স্টক এক্সচেঞ্জের (ডিএসই) দুটি সূ...   \n\n                                               cleanText  count  \n0        অধ্যায়    প্রিয়    পরীক্ষার্থী    তথ্য    যো...    260  \n1        স্পিন    স্পিন    স্পিন    লর্ডসে    ইয়াসির ...    131  \n2        সিরিয়ার    উত্তর    পশ্চিমাঞ্চলে    গতকাল  ...     75  \n3        দক্ষিণ    আফ্রিকার    কেপটাউন    বিশ্ববিদ্যা...     49  \n4        সহজে    ডিজিটাল    আর্থিক    সেবা    অ্যাপ  ...    172  \n...                                                  ...    ...  \n98879    সালে    ইউরোপিয়ান    ক্লাব    কাপে    জাতীয়...     14  \n98880    গানের    সহজিয়ার    প্রথম    অ্যালবাম    রঙম...    163  \n98881    ফেসবুকের    হাজার    ব্যবহারকারীর    তথ্য   ...    159  \n98882    পেট্রোল    পাম্পে    তেল    সিএনজির    বিল  ...     14  \n98883    বছর    শেষে    ঢাকা    স্টক    এক্সচেঞ্জের  ...    257  \n\n[98884 rows x 4 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>category</th>\n      <th>text</th>\n      <th>cleanText</th>\n      <th>count</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>education</td>\n      <td>অধ্যায়-২প্রিয় পরীক্ষার্থী, আজ তথ্য ও যোগাযোগ প...</td>\n      <td>অধ্যায়    প্রিয়    পরীক্ষার্থী    তথ্য    যো...</td>\n      <td>260</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>sports</td>\n      <td>স্পিন, স্পিন, স্পিন। লর্ডসে ইয়াসির শাহ্‌র লেগ ...</td>\n      <td>স্পিন    স্পিন    স্পিন    লর্ডসে    ইয়াসির ...</td>\n      <td>131</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>international</td>\n      <td>সিরিয়ার উত্তর-পশ্চিমাঞ্চলে গতকাল রোববার বিমান...</td>\n      <td>সিরিয়ার    উত্তর    পশ্চিমাঞ্চলে    গতকাল  ...</td>\n      <td>75</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>international</td>\n      <td>দক্ষিণ আফ্রিকার কেপটাউন বিশ্ববিদ্যালয়ের (ইউসিট...</td>\n      <td>দক্ষিণ    আফ্রিকার    কেপটাউন    বিশ্ববিদ্যা...</td>\n      <td>49</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>technology</td>\n      <td>সহজে ডিজিটাল আর্থিক সেবা দিতে অ্যাপ ও ডিজিটাল ...</td>\n      <td>সহজে    ডিজিটাল    আর্থিক    সেবা    অ্যাপ  ...</td>\n      <td>172</td>\n    </tr>\n    <tr>\n      <th>...</th>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n    </tr>\n    <tr>\n      <th>98879</th>\n      <td>sports</td>\n      <td>১৯৫৯ সালে ইউরোপিয়ান ক্লাব কাপে জাতীয় দল পাঠান...</td>\n      <td>সালে    ইউরোপিয়ান    ক্লাব    কাপে    জাতীয়...</td>\n      <td>14</td>\n    </tr>\n    <tr>\n      <th>98880</th>\n      <td>entertainment</td>\n      <td>গানের দল সহজিয়ার প্রথম অ্যালবাম ‘রঙমিস্ত্রী’র ...</td>\n      <td>গানের    সহজিয়ার    প্রথম    অ্যালবাম    রঙম...</td>\n      <td>163</td>\n    </tr>\n    <tr>\n      <th>98881</th>\n      <td>international</td>\n      <td>ফেসবুকের ৮১ হাজার ব্যবহারকারীর তথ্য চুরি করে স...</td>\n      <td>ফেসবুকের    হাজার    ব্যবহারকারীর    তথ্য   ...</td>\n      <td>159</td>\n    </tr>\n    <tr>\n      <th>98882</th>\n      <td>economy</td>\n      <td>পেট্রোল পাম্পে তেল বা সিএনজির বিল বিকাশের মাধ...</td>\n      <td>পেট্রোল    পাম্পে    তেল    সিএনজির    বিল  ...</td>\n      <td>14</td>\n    </tr>\n    <tr>\n      <th>98883</th>\n      <td>economy</td>\n      <td>বছর শেষে ঢাকা স্টক এক্সচেঞ্জের (ডিএসই) দুটি সূ...</td>\n      <td>বছর    শেষে    ঢাকা    স্টক    এক্সচেঞ্জের  ...</td>\n      <td>257</td>\n    </tr>\n  </tbody>\n</table>\n<p>98884 rows × 4 columns</p>\n</div>"},"metadata":{}},{"name":"stdout","text":"<class 'pandas.core.frame.DataFrame'>\nRangeIndex: 98884 entries, 0 to 98883\nData columns (total 4 columns):\n #   Column     Non-Null Count  Dtype \n---  ------     --------------  ----- \n 0   category   98884 non-null  object\n 1   text       98884 non-null  object\n 2   cleanText  98884 non-null  object\n 3   count      98884 non-null  int64 \ndtypes: int64(1), object(3)\nmemory usage: 3.0+ MB\n","output_type":"stream"}]},{"cell_type":"code","source":"def stopwordRemoval(text):    \n    x=str(text)\n    l=x.split()\n\n    stm=[elem for elem in l if elem not in stop]\n    \n    out=' '.join(stm)\n    \n    return str(out)","metadata":{"papermill":{"duration":0.042051,"end_time":"2021-10-19T15:50:57.113675","exception":false,"start_time":"2021-10-19T15:50:57.071624","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T04:46:12.916709Z","iopub.execute_input":"2021-10-20T04:46:12.917503Z","iopub.status.idle":"2021-10-20T04:46:12.923244Z","shell.execute_reply.started":"2021-10-20T04:46:12.917461Z","shell.execute_reply":"2021-10-20T04:46:12.922406Z"},"trusted":true},"execution_count":25,"outputs":[]},{"cell_type":"markdown","source":"# **Stopwords Removal**","metadata":{"papermill":{"duration":0.034106,"end_time":"2021-10-19T15:50:57.181905","exception":false,"start_time":"2021-10-19T15:50:57.147799","status":"completed"},"tags":[]}},{"cell_type":"code","source":"data1 =pd.read_excel('/kaggle/input/bangla-stopwords/stopwords_bangla.xlsx')\ndisplay(data1)\nstop = data1['words'].tolist()","metadata":{"papermill":{"duration":0.306621,"end_time":"2021-10-19T15:50:57.522956","exception":false,"start_time":"2021-10-19T15:50:57.216335","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T04:46:12.924580Z","iopub.execute_input":"2021-10-20T04:46:12.925234Z","iopub.status.idle":"2021-10-20T04:46:13.004560Z","shell.execute_reply.started":"2021-10-20T04:46:12.925194Z","shell.execute_reply":"2021-10-20T04:46:13.003746Z"},"trusted":true},"execution_count":26,"outputs":[{"output_type":"display_data","data":{"text/plain":"     Unnamed: 0   words\n0             0      অই\n1             1  অগত্যা\n2             2  অত: পর\n3             3    অতএব\n4             4     অথচ\n..          ...     ...\n727         727    হয়েই\n728         728  হয়েছিল\n729         729   হয়েছে\n730         730  হয়েছেন\n731         731     অংশ\n\n[732 rows x 2 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>Unnamed: 0</th>\n      <th>words</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>0</td>\n      <td>অই</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>1</td>\n      <td>অগত্যা</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>2</td>\n      <td>অত: পর</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>3</td>\n      <td>অতএব</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>4</td>\n      <td>অথচ</td>\n    </tr>\n    <tr>\n      <th>...</th>\n      <td>...</td>\n      <td>...</td>\n    </tr>\n    <tr>\n      <th>727</th>\n      <td>727</td>\n      <td>হয়েই</td>\n    </tr>\n    <tr>\n      <th>728</th>\n      <td>728</td>\n      <td>হয়েছিল</td>\n    </tr>\n    <tr>\n      <th>729</th>\n      <td>729</td>\n      <td>হয়েছে</td>\n    </tr>\n    <tr>\n      <th>730</th>\n      <td>730</td>\n      <td>হয়েছেন</td>\n    </tr>\n    <tr>\n      <th>731</th>\n      <td>731</td>\n      <td>অংশ</td>\n    </tr>\n  </tbody>\n</table>\n<p>732 rows × 2 columns</p>\n</div>"},"metadata":{}}]},{"cell_type":"code","source":"#df_train['cleanText'] = df_train.cleanText.apply(lambda x: stopwordRemoval(str(x)))\n","metadata":{"papermill":{"duration":177.769835,"end_time":"2021-10-19T15:53:55.328548","exception":false,"start_time":"2021-10-19T15:50:57.558713","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T04:46:13.005869Z","iopub.execute_input":"2021-10-20T04:46:13.006201Z","iopub.status.idle":"2021-10-20T04:46:13.009982Z","shell.execute_reply.started":"2021-10-20T04:46:13.006164Z","shell.execute_reply":"2021-10-20T04:46:13.009199Z"},"trusted":true},"execution_count":27,"outputs":[]},{"cell_type":"code","source":"#df_test['cleanText'] = df_test.cleanText.apply(lambda x: stopwordRemoval(str(x)))","metadata":{"papermill":{"duration":176.489362,"end_time":"2021-10-19T15:56:51.853171","exception":false,"start_time":"2021-10-19T15:53:55.363809","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T04:46:13.011558Z","iopub.execute_input":"2021-10-20T04:46:13.011950Z","iopub.status.idle":"2021-10-20T04:46:13.019955Z","shell.execute_reply.started":"2021-10-20T04:46:13.011793Z","shell.execute_reply":"2021-10-20T04:46:13.019266Z"},"trusted":true},"execution_count":28,"outputs":[]},{"cell_type":"markdown","source":"# **Stemming**","metadata":{"papermill":{"duration":0.035718,"end_time":"2021-10-19T15:56:51.925536","exception":false,"start_time":"2021-10-19T15:56:51.889818","status":"completed"},"tags":[]}},{"cell_type":"code","source":"#make sure to turn on internet on your kernel\n#importing stemmer\n!pip install bangla-stemmer\nfrom bangla_stemmer.stemmer import stemmer\n## stemmer function\ndef stem_text (x):\n  stmr = stemmer.BanglaStemmer()\n  words=x.split(' ')\n  stm = stmr.stem(words)\n  words=(' ').join(stm)\n  return words","metadata":{"papermill":{"duration":7.061834,"end_time":"2021-10-19T15:56:59.023501","exception":false,"start_time":"2021-10-19T15:56:51.961667","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T04:46:13.022133Z","iopub.execute_input":"2021-10-20T04:46:13.022342Z","iopub.status.idle":"2021-10-20T04:46:20.345812Z","shell.execute_reply.started":"2021-10-20T04:46:13.022314Z","shell.execute_reply":"2021-10-20T04:46:20.344939Z"},"trusted":true},"execution_count":29,"outputs":[{"name":"stdout","text":"Requirement already satisfied: bangla-stemmer in /opt/conda/lib/python3.7/site-packages (1.0)\n\u001b[33mWARNING: Running pip as the 'root' user can result in broken permissions and conflicting behaviour with the system package manager. It is recommended to use a virtual environment instead: https://pip.pypa.io/warnings/venv\u001b[0m\n","output_type":"stream"}]},{"cell_type":"code","source":"#df_train['cleanText']=df_train['cleanText'].apply(stem_text)\n#df_test['cleanText']=df_test['cleanText'].apply(stem_text)","metadata":{"papermill":{"duration":0.043403,"end_time":"2021-10-19T15:56:59.103700","exception":false,"start_time":"2021-10-19T15:56:59.060297","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T04:46:20.349503Z","iopub.execute_input":"2021-10-20T04:46:20.349754Z","iopub.status.idle":"2021-10-20T04:46:20.355025Z","shell.execute_reply.started":"2021-10-20T04:46:20.349725Z","shell.execute_reply":"2021-10-20T04:46:20.354226Z"},"trusted":true},"execution_count":30,"outputs":[]},{"cell_type":"markdown","source":"# **Changing Labels to Numbers**","metadata":{"papermill":{"duration":0.03582,"end_time":"2021-10-19T15:56:59.175690","exception":false,"start_time":"2021-10-19T15:56:59.139870","status":"completed"},"tags":[]}},{"cell_type":"code","source":"#['economy' 'sports' 'international' 'state' 'technology' 'entertainment' 'education']\n\ndf_train[\"category\"].replace({\"economy\": \"0\", \"sports\": \"1\",\"international\": \"2\",\"state\": \"3\",\"technology\": \"4\",\"entertainment\": \"5\",\"education\": \"6\"}, inplace=True)\ndf_test[\"category\"].replace({\"economy\": \"0\", \"sports\": \"1\",\"international\": \"2\",\"state\": \"3\",\"technology\": \"4\",\"entertainment\": \"5\",\"education\": \"6\"}, inplace=True)\ndisplay(df_train)\ndisplay(df_test)","metadata":{"papermill":{"duration":0.14663,"end_time":"2021-10-19T15:56:59.358840","exception":false,"start_time":"2021-10-19T15:56:59.212210","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T04:46:20.360024Z","iopub.execute_input":"2021-10-20T04:46:20.360946Z","iopub.status.idle":"2021-10-20T04:46:20.478937Z","shell.execute_reply.started":"2021-10-20T04:46:20.360899Z","shell.execute_reply":"2021-10-20T04:46:20.478143Z"},"trusted":true},"execution_count":31,"outputs":[{"output_type":"display_data","data":{"text/plain":"      category                                               text  \\\n0            0   বাংলাদেশে সফররত বিশ্ব বাণিজ্য সংস্থা (ডব্লিউট...   \n1            1  ইউরোপিয়ান ফুটবলে আজ ঝড় বয়ে যাচ্ছে। একদিকে সংবা...   \n2            2  বিজেপির নেতৃত্বে ভারতের উত্তর প্রদেশ থেকে শুরু...   \n3            3  নারায়ণগঞ্জের সোনারগাঁ উপজেলার একটি সরকারি প্র...   \n4            1  ‘ছাই থেকে ছাই, ধুলো থেকে ধুলো...’ আধ্যাত্মিক ক...   \n...        ...                                                ...   \n98878        6    সহকারী অধ্যাপক  টিচার্স ট্রেনিং কলেজ (মহিলা)...   \n98879        5  অক্টোবরের ১১ তারিখে ৭৫ বছর পূর্ণ করবেন অমিতাভ ...   \n98880        4  কাজী নাঈম আল রশিদ। সফটওয়্যার প্রকৌশলী। কাজ করে...   \n98881        4  ইয়াহুকে মার্কিন টেলিকম প্রতিষ্ঠান ভেরাইজন কিনে...   \n98882        6  অধ্যায়-১১প্রিয় শিক্ষার্থী, আজ ফিন্যান্স ও ব্যা...   \n\n                                               cleanText  count  \n0        বাংলাদেশে    সফররত    বিশ্ব    বাণিজ্য    সং...    293  \n1        ইউরোপিয়ান    ফুটবলে    বয়ে    একদিকে    সংবা...    206  \n2        বিজেপির    নেতৃত্বে    ভারতের    উত্তর    প্...    167  \n3        নারায়ণগঞ্জের    সোনারগাঁ    উপজেলার    সরকা...    181  \n4        ছাই    ছাই    ধুলো    ধুলো    আধ্যাত্মিক    ...    148  \n...                                                  ...    ...  \n98878    সহকারী    অধ্যাপক    টিচার্স    ট্রেনিং    ক...    389  \n98879    অক্টোবরের    তারিখে    বছর    পূর্ণ    অমিতা...    147  \n98880    কাজী    নাঈম    রশিদ    সফটওয়্যার    প্রকৌশল...    673  \n98881    ইয়াহুকে    মার্কিন    টেলিকম    প্রতিষ্ঠান  ...    161  \n98882    অধ্যায়    প্রিয়    শিক্ষার্থী    ফিন্যান্স  ...    204  \n\n[98883 rows x 4 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>category</th>\n      <th>text</th>\n      <th>cleanText</th>\n      <th>count</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>0</td>\n      <td>বাংলাদেশে সফররত বিশ্ব বাণিজ্য সংস্থা (ডব্লিউট...</td>\n      <td>বাংলাদেশে    সফররত    বিশ্ব    বাণিজ্য    সং...</td>\n      <td>293</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>1</td>\n      <td>ইউরোপিয়ান ফুটবলে আজ ঝড় বয়ে যাচ্ছে। একদিকে সংবা...</td>\n      <td>ইউরোপিয়ান    ফুটবলে    বয়ে    একদিকে    সংবা...</td>\n      <td>206</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>2</td>\n      <td>বিজেপির নেতৃত্বে ভারতের উত্তর প্রদেশ থেকে শুরু...</td>\n      <td>বিজেপির    নেতৃত্বে    ভারতের    উত্তর    প্...</td>\n      <td>167</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>3</td>\n      <td>নারায়ণগঞ্জের সোনারগাঁ উপজেলার একটি সরকারি প্র...</td>\n      <td>নারায়ণগঞ্জের    সোনারগাঁ    উপজেলার    সরকা...</td>\n      <td>181</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>1</td>\n      <td>‘ছাই থেকে ছাই, ধুলো থেকে ধুলো...’ আধ্যাত্মিক ক...</td>\n      <td>ছাই    ছাই    ধুলো    ধুলো    আধ্যাত্মিক    ...</td>\n      <td>148</td>\n    </tr>\n    <tr>\n      <th>...</th>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n    </tr>\n    <tr>\n      <th>98878</th>\n      <td>6</td>\n      <td>সহকারী অধ্যাপক  টিচার্স ট্রেনিং কলেজ (মহিলা)...</td>\n      <td>সহকারী    অধ্যাপক    টিচার্স    ট্রেনিং    ক...</td>\n      <td>389</td>\n    </tr>\n    <tr>\n      <th>98879</th>\n      <td>5</td>\n      <td>অক্টোবরের ১১ তারিখে ৭৫ বছর পূর্ণ করবেন অমিতাভ ...</td>\n      <td>অক্টোবরের    তারিখে    বছর    পূর্ণ    অমিতা...</td>\n      <td>147</td>\n    </tr>\n    <tr>\n      <th>98880</th>\n      <td>4</td>\n      <td>কাজী নাঈম আল রশিদ। সফটওয়্যার প্রকৌশলী। কাজ করে...</td>\n      <td>কাজী    নাঈম    রশিদ    সফটওয়্যার    প্রকৌশল...</td>\n      <td>673</td>\n    </tr>\n    <tr>\n      <th>98881</th>\n      <td>4</td>\n      <td>ইয়াহুকে মার্কিন টেলিকম প্রতিষ্ঠান ভেরাইজন কিনে...</td>\n      <td>ইয়াহুকে    মার্কিন    টেলিকম    প্রতিষ্ঠান  ...</td>\n      <td>161</td>\n    </tr>\n    <tr>\n      <th>98882</th>\n      <td>6</td>\n      <td>অধ্যায়-১১প্রিয় শিক্ষার্থী, আজ ফিন্যান্স ও ব্যা...</td>\n      <td>অধ্যায়    প্রিয়    শিক্ষার্থী    ফিন্যান্স  ...</td>\n      <td>204</td>\n    </tr>\n  </tbody>\n</table>\n<p>98883 rows × 4 columns</p>\n</div>"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"      category                                               text  \\\n0            6  অধ্যায়-২প্রিয় পরীক্ষার্থী, আজ তথ্য ও যোগাযোগ প...   \n1            1  স্পিন, স্পিন, স্পিন। লর্ডসে ইয়াসির শাহ্‌র লেগ ...   \n2            2  সিরিয়ার উত্তর-পশ্চিমাঞ্চলে গতকাল রোববার বিমান...   \n3            2  দক্ষিণ আফ্রিকার কেপটাউন বিশ্ববিদ্যালয়ের (ইউসিট...   \n4            4  সহজে ডিজিটাল আর্থিক সেবা দিতে অ্যাপ ও ডিজিটাল ...   \n...        ...                                                ...   \n98879        1  ১৯৫৯ সালে ইউরোপিয়ান ক্লাব কাপে জাতীয় দল পাঠান...   \n98880        5  গানের দল সহজিয়ার প্রথম অ্যালবাম ‘রঙমিস্ত্রী’র ...   \n98881        2  ফেসবুকের ৮১ হাজার ব্যবহারকারীর তথ্য চুরি করে স...   \n98882        0   পেট্রোল পাম্পে তেল বা সিএনজির বিল বিকাশের মাধ...   \n98883        0  বছর শেষে ঢাকা স্টক এক্সচেঞ্জের (ডিএসই) দুটি সূ...   \n\n                                               cleanText  count  \n0        অধ্যায়    প্রিয়    পরীক্ষার্থী    তথ্য    যো...    260  \n1        স্পিন    স্পিন    স্পিন    লর্ডসে    ইয়াসির ...    131  \n2        সিরিয়ার    উত্তর    পশ্চিমাঞ্চলে    গতকাল  ...     75  \n3        দক্ষিণ    আফ্রিকার    কেপটাউন    বিশ্ববিদ্যা...     49  \n4        সহজে    ডিজিটাল    আর্থিক    সেবা    অ্যাপ  ...    172  \n...                                                  ...    ...  \n98879    সালে    ইউরোপিয়ান    ক্লাব    কাপে    জাতীয়...     14  \n98880    গানের    সহজিয়ার    প্রথম    অ্যালবাম    রঙম...    163  \n98881    ফেসবুকের    হাজার    ব্যবহারকারীর    তথ্য   ...    159  \n98882    পেট্রোল    পাম্পে    তেল    সিএনজির    বিল  ...     14  \n98883    বছর    শেষে    ঢাকা    স্টক    এক্সচেঞ্জের  ...    257  \n\n[98884 rows x 4 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>category</th>\n      <th>text</th>\n      <th>cleanText</th>\n      <th>count</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>6</td>\n      <td>অধ্যায়-২প্রিয় পরীক্ষার্থী, আজ তথ্য ও যোগাযোগ প...</td>\n      <td>অধ্যায়    প্রিয়    পরীক্ষার্থী    তথ্য    যো...</td>\n      <td>260</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>1</td>\n      <td>স্পিন, স্পিন, স্পিন। লর্ডসে ইয়াসির শাহ্‌র লেগ ...</td>\n      <td>স্পিন    স্পিন    স্পিন    লর্ডসে    ইয়াসির ...</td>\n      <td>131</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>2</td>\n      <td>সিরিয়ার উত্তর-পশ্চিমাঞ্চলে গতকাল রোববার বিমান...</td>\n      <td>সিরিয়ার    উত্তর    পশ্চিমাঞ্চলে    গতকাল  ...</td>\n      <td>75</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>2</td>\n      <td>দক্ষিণ আফ্রিকার কেপটাউন বিশ্ববিদ্যালয়ের (ইউসিট...</td>\n      <td>দক্ষিণ    আফ্রিকার    কেপটাউন    বিশ্ববিদ্যা...</td>\n      <td>49</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>4</td>\n      <td>সহজে ডিজিটাল আর্থিক সেবা দিতে অ্যাপ ও ডিজিটাল ...</td>\n      <td>সহজে    ডিজিটাল    আর্থিক    সেবা    অ্যাপ  ...</td>\n      <td>172</td>\n    </tr>\n    <tr>\n      <th>...</th>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n    </tr>\n    <tr>\n      <th>98879</th>\n      <td>1</td>\n      <td>১৯৫৯ সালে ইউরোপিয়ান ক্লাব কাপে জাতীয় দল পাঠান...</td>\n      <td>সালে    ইউরোপিয়ান    ক্লাব    কাপে    জাতীয়...</td>\n      <td>14</td>\n    </tr>\n    <tr>\n      <th>98880</th>\n      <td>5</td>\n      <td>গানের দল সহজিয়ার প্রথম অ্যালবাম ‘রঙমিস্ত্রী’র ...</td>\n      <td>গানের    সহজিয়ার    প্রথম    অ্যালবাম    রঙম...</td>\n      <td>163</td>\n    </tr>\n    <tr>\n      <th>98881</th>\n      <td>2</td>\n      <td>ফেসবুকের ৮১ হাজার ব্যবহারকারীর তথ্য চুরি করে স...</td>\n      <td>ফেসবুকের    হাজার    ব্যবহারকারীর    তথ্য   ...</td>\n      <td>159</td>\n    </tr>\n    <tr>\n      <th>98882</th>\n      <td>0</td>\n      <td>পেট্রোল পাম্পে তেল বা সিএনজির বিল বিকাশের মাধ...</td>\n      <td>পেট্রোল    পাম্পে    তেল    সিএনজির    বিল  ...</td>\n      <td>14</td>\n    </tr>\n    <tr>\n      <th>98883</th>\n      <td>0</td>\n      <td>বছর শেষে ঢাকা স্টক এক্সচেঞ্জের (ডিএসই) দুটি সূ...</td>\n      <td>বছর    শেষে    ঢাকা    স্টক    এক্সচেঞ্জের  ...</td>\n      <td>257</td>\n    </tr>\n  </tbody>\n</table>\n<p>98884 rows × 4 columns</p>\n</div>"},"metadata":{}}]},{"cell_type":"markdown","source":"# **Removing Null values**","metadata":{"papermill":{"duration":0.036676,"end_time":"2021-10-19T15:56:59.432428","exception":false,"start_time":"2021-10-19T15:56:59.395752","status":"completed"},"tags":[]}},{"cell_type":"code","source":"#display non-null values\ndf_train = df_train.dropna()\ndf_test = df_test.dropna()","metadata":{"papermill":{"duration":0.420253,"end_time":"2021-10-19T15:56:59.889527","exception":false,"start_time":"2021-10-19T15:56:59.469274","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T04:46:20.480404Z","iopub.execute_input":"2021-10-20T04:46:20.480846Z","iopub.status.idle":"2021-10-20T04:46:20.973317Z","shell.execute_reply.started":"2021-10-20T04:46:20.480803Z","shell.execute_reply":"2021-10-20T04:46:20.972444Z"},"trusted":true},"execution_count":32,"outputs":[]},{"cell_type":"markdown","source":"# **Removing Small Texts**","metadata":{"papermill":{"duration":0.036856,"end_time":"2021-10-19T15:56:59.964075","exception":false,"start_time":"2021-10-19T15:56:59.927219","status":"completed"},"tags":[]}},{"cell_type":"code","source":"#counting text length\ndf_train['count'] = df_train['cleanText'].str.split().str.len()\ndf_test['count'] = df_test['cleanText'].str.split().str.len()\n# Remove the text with words less than 5\ndf_train= df_train.loc[df_train['count']>5]\ndf_test= df_test.loc[df_test['count']>5]","metadata":{"papermill":{"duration":6.793066,"end_time":"2021-10-19T15:57:06.795371","exception":false,"start_time":"2021-10-19T15:57:00.002305","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T04:46:20.974686Z","iopub.execute_input":"2021-10-20T04:46:20.975122Z","iopub.status.idle":"2021-10-20T04:46:28.547191Z","shell.execute_reply.started":"2021-10-20T04:46:20.975083Z","shell.execute_reply":"2021-10-20T04:46:28.546394Z"},"trusted":true},"execution_count":33,"outputs":[]},{"cell_type":"markdown","source":"# **Removing Duplicate Rows**","metadata":{}},{"cell_type":"code","source":"df_train.drop_duplicates()\ndf_test.drop_duplicates()","metadata":{"execution":{"iopub.status.busy":"2021-10-20T04:46:28.548553Z","iopub.execute_input":"2021-10-20T04:46:28.548835Z","iopub.status.idle":"2021-10-20T04:46:35.997438Z","shell.execute_reply.started":"2021-10-20T04:46:28.548800Z","shell.execute_reply":"2021-10-20T04:46:35.996738Z"},"trusted":true},"execution_count":34,"outputs":[{"execution_count":34,"output_type":"execute_result","data":{"text/plain":"      category                                               text  \\\n0            6  অধ্যায়-২প্রিয় পরীক্ষার্থী, আজ তথ্য ও যোগাযোগ প...   \n1            1  স্পিন, স্পিন, স্পিন। লর্ডসে ইয়াসির শাহ্‌র লেগ ...   \n2            2  সিরিয়ার উত্তর-পশ্চিমাঞ্চলে গতকাল রোববার বিমান...   \n3            2  দক্ষিণ আফ্রিকার কেপটাউন বিশ্ববিদ্যালয়ের (ইউসিট...   \n4            4  সহজে ডিজিটাল আর্থিক সেবা দিতে অ্যাপ ও ডিজিটাল ...   \n...        ...                                                ...   \n98879        1  ১৯৫৯ সালে ইউরোপিয়ান ক্লাব কাপে জাতীয় দল পাঠান...   \n98880        5  গানের দল সহজিয়ার প্রথম অ্যালবাম ‘রঙমিস্ত্রী’র ...   \n98881        2  ফেসবুকের ৮১ হাজার ব্যবহারকারীর তথ্য চুরি করে স...   \n98882        0   পেট্রোল পাম্পে তেল বা সিএনজির বিল বিকাশের মাধ...   \n98883        0  বছর শেষে ঢাকা স্টক এক্সচেঞ্জের (ডিএসই) দুটি সূ...   \n\n                                               cleanText  count  \n0        অধ্যায়    প্রিয়    পরীক্ষার্থী    তথ্য    যো...    260  \n1        স্পিন    স্পিন    স্পিন    লর্ডসে    ইয়াসির ...    131  \n2        সিরিয়ার    উত্তর    পশ্চিমাঞ্চলে    গতকাল  ...     75  \n3        দক্ষিণ    আফ্রিকার    কেপটাউন    বিশ্ববিদ্যা...     49  \n4        সহজে    ডিজিটাল    আর্থিক    সেবা    অ্যাপ  ...    172  \n...                                                  ...    ...  \n98879    সালে    ইউরোপিয়ান    ক্লাব    কাপে    জাতীয়...     14  \n98880    গানের    সহজিয়ার    প্রথম    অ্যালবাম    রঙম...    163  \n98881    ফেসবুকের    হাজার    ব্যবহারকারীর    তথ্য   ...    159  \n98882    পেট্রোল    পাম্পে    তেল    সিএনজির    বিল  ...     14  \n98883    বছর    শেষে    ঢাকা    স্টক    এক্সচেঞ্জের  ...    257  \n\n[98164 rows x 4 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>category</th>\n      <th>text</th>\n      <th>cleanText</th>\n      <th>count</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>6</td>\n      <td>অধ্যায়-২প্রিয় পরীক্ষার্থী, আজ তথ্য ও যোগাযোগ প...</td>\n      <td>অধ্যায়    প্রিয়    পরীক্ষার্থী    তথ্য    যো...</td>\n      <td>260</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>1</td>\n      <td>স্পিন, স্পিন, স্পিন। লর্ডসে ইয়াসির শাহ্‌র লেগ ...</td>\n      <td>স্পিন    স্পিন    স্পিন    লর্ডসে    ইয়াসির ...</td>\n      <td>131</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>2</td>\n      <td>সিরিয়ার উত্তর-পশ্চিমাঞ্চলে গতকাল রোববার বিমান...</td>\n      <td>সিরিয়ার    উত্তর    পশ্চিমাঞ্চলে    গতকাল  ...</td>\n      <td>75</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>2</td>\n      <td>দক্ষিণ আফ্রিকার কেপটাউন বিশ্ববিদ্যালয়ের (ইউসিট...</td>\n      <td>দক্ষিণ    আফ্রিকার    কেপটাউন    বিশ্ববিদ্যা...</td>\n      <td>49</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>4</td>\n      <td>সহজে ডিজিটাল আর্থিক সেবা দিতে অ্যাপ ও ডিজিটাল ...</td>\n      <td>সহজে    ডিজিটাল    আর্থিক    সেবা    অ্যাপ  ...</td>\n      <td>172</td>\n    </tr>\n    <tr>\n      <th>...</th>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n    </tr>\n    <tr>\n      <th>98879</th>\n      <td>1</td>\n      <td>১৯৫৯ সালে ইউরোপিয়ান ক্লাব কাপে জাতীয় দল পাঠান...</td>\n      <td>সালে    ইউরোপিয়ান    ক্লাব    কাপে    জাতীয়...</td>\n      <td>14</td>\n    </tr>\n    <tr>\n      <th>98880</th>\n      <td>5</td>\n      <td>গানের দল সহজিয়ার প্রথম অ্যালবাম ‘রঙমিস্ত্রী’র ...</td>\n      <td>গানের    সহজিয়ার    প্রথম    অ্যালবাম    রঙম...</td>\n      <td>163</td>\n    </tr>\n    <tr>\n      <th>98881</th>\n      <td>2</td>\n      <td>ফেসবুকের ৮১ হাজার ব্যবহারকারীর তথ্য চুরি করে স...</td>\n      <td>ফেসবুকের    হাজার    ব্যবহারকারীর    তথ্য   ...</td>\n      <td>159</td>\n    </tr>\n    <tr>\n      <th>98882</th>\n      <td>0</td>\n      <td>পেট্রোল পাম্পে তেল বা সিএনজির বিল বিকাশের মাধ...</td>\n      <td>পেট্রোল    পাম্পে    তেল    সিএনজির    বিল  ...</td>\n      <td>14</td>\n    </tr>\n    <tr>\n      <th>98883</th>\n      <td>0</td>\n      <td>বছর শেষে ঢাকা স্টক এক্সচেঞ্জের (ডিএসই) দুটি সূ...</td>\n      <td>বছর    শেষে    ঢাকা    স্টক    এক্সচেঞ্জের  ...</td>\n      <td>257</td>\n    </tr>\n  </tbody>\n</table>\n<p>98164 rows × 4 columns</p>\n</div>"},"metadata":{}}]},{"cell_type":"markdown","source":"# **Shuffle Dataset**","metadata":{"papermill":{"duration":0.064527,"end_time":"2021-10-19T15:57:06.938220","exception":false,"start_time":"2021-10-19T15:57:06.873693","status":"completed"},"tags":[]}},{"cell_type":"code","source":"df_train = df_train.sample(frac=1).reset_index(drop=True)\ndisplay(df_train)\ndisplay(df_test)","metadata":{"papermill":{"duration":0.158248,"end_time":"2021-10-19T15:57:07.159106","exception":false,"start_time":"2021-10-19T15:57:07.000858","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T04:46:35.998638Z","iopub.execute_input":"2021-10-20T04:46:36.000043Z","iopub.status.idle":"2021-10-20T04:46:36.058682Z","shell.execute_reply.started":"2021-10-20T04:46:36.000002Z","shell.execute_reply":"2021-10-20T04:46:36.056940Z"},"trusted":true},"execution_count":35,"outputs":[{"output_type":"display_data","data":{"text/plain":"      category                                               text  \\\n0            5  বিয়ের ব্যাপারটি প্রকাশ্যে আসার পর এবার ছিল জনপ...   \n1            3  ক্ষমতাসীন আওয়ামী লীগের ভ্রাতৃপ্রতিম সংগঠন বাংল...   \n2            5  বয়সভিত্তিক নাচ প্রতিযোগিতার আয়োজন করেছিল নৃত...   \n3            1  পুরো দেশ অস্থির। হবে না-ই বা কেন। দুটি ম্যাচ ...   \n4            4   নকিয়া ৮ বাজারে আসতে না আসতেই ইতিমধ্যে গুঞ্জন ...   \n...        ...                                                ...   \n98620        1  হোয়াইট হার্ট লেনে পরশু টটেনহামের গ্যালারিতে কি...   \n98621        1  ‘শৃঙ্খলা-ভঙ্গের’ অভিযোগে বোলার আল আমিন হোসেনের...   \n98622        2  বন্ধুরাষ্ট্রের প্রতি সামান্য অবিশ্বাস দেখানোও ...   \n98623        1  পেনাল্টির আবেদন করেছিলেন। রেফারি বাঁশিও বাজিয়ে...   \n98624        0  ২০০৮ সালের অর্থনৈতিক সংকট অনেকটাই কাটিয়ে উঠেছি...   \n\n                                               cleanText  count  \n0        বিয়ের    ব্যাপারটি    প্রকাশ্যে    আসার    জ...     99  \n1        ক্ষমতাসীন    আওয়ামী    লীগের    ভ্রাতৃপ্রতিম...    150  \n2        বয়সভিত্তিক    নাচ    প্রতিযোগিতার    আয়োজন...     60  \n3        পুরো    দেশ    অস্থির    ম্যাচ    দক্ষিণ   ...    151  \n4        নকিয়া    বাজারে    আসতে    আসতেই    ইতিমধ্যে...     96  \n...                                                  ...    ...  \n98620    হোয়াইট    হার্ট    লেনে    পরশু    টটেনহামের...    442  \n98621    শৃঙ্খলা    ভঙ্গের    অভিযোগে    বোলার    আমি...    186  \n98622    বন্ধুরাষ্ট্রের    সামান্য    অবিশ্বাস    দেখ...     51  \n98623    পেনাল্টির    আবেদন    রেফারি    বাঁশিও    বা...    107  \n98624    সালের    অর্থনৈতিক    সংকট    অনেকটাই    কাট...    344  \n\n[98625 rows x 4 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>category</th>\n      <th>text</th>\n      <th>cleanText</th>\n      <th>count</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>5</td>\n      <td>বিয়ের ব্যাপারটি প্রকাশ্যে আসার পর এবার ছিল জনপ...</td>\n      <td>বিয়ের    ব্যাপারটি    প্রকাশ্যে    আসার    জ...</td>\n      <td>99</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>3</td>\n      <td>ক্ষমতাসীন আওয়ামী লীগের ভ্রাতৃপ্রতিম সংগঠন বাংল...</td>\n      <td>ক্ষমতাসীন    আওয়ামী    লীগের    ভ্রাতৃপ্রতিম...</td>\n      <td>150</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>5</td>\n      <td>বয়সভিত্তিক নাচ প্রতিযোগিতার আয়োজন করেছিল নৃত...</td>\n      <td>বয়সভিত্তিক    নাচ    প্রতিযোগিতার    আয়োজন...</td>\n      <td>60</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>1</td>\n      <td>পুরো দেশ অস্থির। হবে না-ই বা কেন। দুটি ম্যাচ ...</td>\n      <td>পুরো    দেশ    অস্থির    ম্যাচ    দক্ষিণ   ...</td>\n      <td>151</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>4</td>\n      <td>নকিয়া ৮ বাজারে আসতে না আসতেই ইতিমধ্যে গুঞ্জন ...</td>\n      <td>নকিয়া    বাজারে    আসতে    আসতেই    ইতিমধ্যে...</td>\n      <td>96</td>\n    </tr>\n    <tr>\n      <th>...</th>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n    </tr>\n    <tr>\n      <th>98620</th>\n      <td>1</td>\n      <td>হোয়াইট হার্ট লেনে পরশু টটেনহামের গ্যালারিতে কি...</td>\n      <td>হোয়াইট    হার্ট    লেনে    পরশু    টটেনহামের...</td>\n      <td>442</td>\n    </tr>\n    <tr>\n      <th>98621</th>\n      <td>1</td>\n      <td>‘শৃঙ্খলা-ভঙ্গের’ অভিযোগে বোলার আল আমিন হোসেনের...</td>\n      <td>শৃঙ্খলা    ভঙ্গের    অভিযোগে    বোলার    আমি...</td>\n      <td>186</td>\n    </tr>\n    <tr>\n      <th>98622</th>\n      <td>2</td>\n      <td>বন্ধুরাষ্ট্রের প্রতি সামান্য অবিশ্বাস দেখানোও ...</td>\n      <td>বন্ধুরাষ্ট্রের    সামান্য    অবিশ্বাস    দেখ...</td>\n      <td>51</td>\n    </tr>\n    <tr>\n      <th>98623</th>\n      <td>1</td>\n      <td>পেনাল্টির আবেদন করেছিলেন। রেফারি বাঁশিও বাজিয়ে...</td>\n      <td>পেনাল্টির    আবেদন    রেফারি    বাঁশিও    বা...</td>\n      <td>107</td>\n    </tr>\n    <tr>\n      <th>98624</th>\n      <td>0</td>\n      <td>২০০৮ সালের অর্থনৈতিক সংকট অনেকটাই কাটিয়ে উঠেছি...</td>\n      <td>সালের    অর্থনৈতিক    সংকট    অনেকটাই    কাট...</td>\n      <td>344</td>\n    </tr>\n  </tbody>\n</table>\n<p>98625 rows × 4 columns</p>\n</div>"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"      category                                               text  \\\n0            6  অধ্যায়-২প্রিয় পরীক্ষার্থী, আজ তথ্য ও যোগাযোগ প...   \n1            1  স্পিন, স্পিন, স্পিন। লর্ডসে ইয়াসির শাহ্‌র লেগ ...   \n2            2  সিরিয়ার উত্তর-পশ্চিমাঞ্চলে গতকাল রোববার বিমান...   \n3            2  দক্ষিণ আফ্রিকার কেপটাউন বিশ্ববিদ্যালয়ের (ইউসিট...   \n4            4  সহজে ডিজিটাল আর্থিক সেবা দিতে অ্যাপ ও ডিজিটাল ...   \n...        ...                                                ...   \n98879        1  ১৯৫৯ সালে ইউরোপিয়ান ক্লাব কাপে জাতীয় দল পাঠান...   \n98880        5  গানের দল সহজিয়ার প্রথম অ্যালবাম ‘রঙমিস্ত্রী’র ...   \n98881        2  ফেসবুকের ৮১ হাজার ব্যবহারকারীর তথ্য চুরি করে স...   \n98882        0   পেট্রোল পাম্পে তেল বা সিএনজির বিল বিকাশের মাধ...   \n98883        0  বছর শেষে ঢাকা স্টক এক্সচেঞ্জের (ডিএসই) দুটি সূ...   \n\n                                               cleanText  count  \n0        অধ্যায়    প্রিয়    পরীক্ষার্থী    তথ্য    যো...    260  \n1        স্পিন    স্পিন    স্পিন    লর্ডসে    ইয়াসির ...    131  \n2        সিরিয়ার    উত্তর    পশ্চিমাঞ্চলে    গতকাল  ...     75  \n3        দক্ষিণ    আফ্রিকার    কেপটাউন    বিশ্ববিদ্যা...     49  \n4        সহজে    ডিজিটাল    আর্থিক    সেবা    অ্যাপ  ...    172  \n...                                                  ...    ...  \n98879    সালে    ইউরোপিয়ান    ক্লাব    কাপে    জাতীয়...     14  \n98880    গানের    সহজিয়ার    প্রথম    অ্যালবাম    রঙম...    163  \n98881    ফেসবুকের    হাজার    ব্যবহারকারীর    তথ্য   ...    159  \n98882    পেট্রোল    পাম্পে    তেল    সিএনজির    বিল  ...     14  \n98883    বছর    শেষে    ঢাকা    স্টক    এক্সচেঞ্জের  ...    257  \n\n[98602 rows x 4 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>category</th>\n      <th>text</th>\n      <th>cleanText</th>\n      <th>count</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>0</th>\n      <td>6</td>\n      <td>অধ্যায়-২প্রিয় পরীক্ষার্থী, আজ তথ্য ও যোগাযোগ প...</td>\n      <td>অধ্যায়    প্রিয়    পরীক্ষার্থী    তথ্য    যো...</td>\n      <td>260</td>\n    </tr>\n    <tr>\n      <th>1</th>\n      <td>1</td>\n      <td>স্পিন, স্পিন, স্পিন। লর্ডসে ইয়াসির শাহ্‌র লেগ ...</td>\n      <td>স্পিন    স্পিন    স্পিন    লর্ডসে    ইয়াসির ...</td>\n      <td>131</td>\n    </tr>\n    <tr>\n      <th>2</th>\n      <td>2</td>\n      <td>সিরিয়ার উত্তর-পশ্চিমাঞ্চলে গতকাল রোববার বিমান...</td>\n      <td>সিরিয়ার    উত্তর    পশ্চিমাঞ্চলে    গতকাল  ...</td>\n      <td>75</td>\n    </tr>\n    <tr>\n      <th>3</th>\n      <td>2</td>\n      <td>দক্ষিণ আফ্রিকার কেপটাউন বিশ্ববিদ্যালয়ের (ইউসিট...</td>\n      <td>দক্ষিণ    আফ্রিকার    কেপটাউন    বিশ্ববিদ্যা...</td>\n      <td>49</td>\n    </tr>\n    <tr>\n      <th>4</th>\n      <td>4</td>\n      <td>সহজে ডিজিটাল আর্থিক সেবা দিতে অ্যাপ ও ডিজিটাল ...</td>\n      <td>সহজে    ডিজিটাল    আর্থিক    সেবা    অ্যাপ  ...</td>\n      <td>172</td>\n    </tr>\n    <tr>\n      <th>...</th>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n      <td>...</td>\n    </tr>\n    <tr>\n      <th>98879</th>\n      <td>1</td>\n      <td>১৯৫৯ সালে ইউরোপিয়ান ক্লাব কাপে জাতীয় দল পাঠান...</td>\n      <td>সালে    ইউরোপিয়ান    ক্লাব    কাপে    জাতীয়...</td>\n      <td>14</td>\n    </tr>\n    <tr>\n      <th>98880</th>\n      <td>5</td>\n      <td>গানের দল সহজিয়ার প্রথম অ্যালবাম ‘রঙমিস্ত্রী’র ...</td>\n      <td>গানের    সহজিয়ার    প্রথম    অ্যালবাম    রঙম...</td>\n      <td>163</td>\n    </tr>\n    <tr>\n      <th>98881</th>\n      <td>2</td>\n      <td>ফেসবুকের ৮১ হাজার ব্যবহারকারীর তথ্য চুরি করে স...</td>\n      <td>ফেসবুকের    হাজার    ব্যবহারকারীর    তথ্য   ...</td>\n      <td>159</td>\n    </tr>\n    <tr>\n      <th>98882</th>\n      <td>0</td>\n      <td>পেট্রোল পাম্পে তেল বা সিএনজির বিল বিকাশের মাধ...</td>\n      <td>পেট্রোল    পাম্পে    তেল    সিএনজির    বিল  ...</td>\n      <td>14</td>\n    </tr>\n    <tr>\n      <th>98883</th>\n      <td>0</td>\n      <td>বছর শেষে ঢাকা স্টক এক্সচেঞ্জের (ডিএসই) দুটি সূ...</td>\n      <td>বছর    শেষে    ঢাকা    স্টক    এক্সচেঞ্জের  ...</td>\n      <td>257</td>\n    </tr>\n  </tbody>\n</table>\n<p>98602 rows × 4 columns</p>\n</div>"},"metadata":{}}]},{"cell_type":"code","source":"train_sentences=df_train['cleanText'].values\ntrain_labels=df_train['category'].values","metadata":{"papermill":{"duration":0.071228,"end_time":"2021-10-19T15:57:07.296059","exception":false,"start_time":"2021-10-19T15:57:07.224831","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T04:46:36.060222Z","iopub.execute_input":"2021-10-20T04:46:36.060438Z","iopub.status.idle":"2021-10-20T04:46:36.064372Z","shell.execute_reply.started":"2021-10-20T04:46:36.060408Z","shell.execute_reply":"2021-10-20T04:46:36.063660Z"},"trusted":true},"execution_count":36,"outputs":[]},{"cell_type":"code","source":"test_sentences=df_test['cleanText'].values\ntest_labels=df_test['category'].values","metadata":{"papermill":{"duration":0.071788,"end_time":"2021-10-19T15:57:07.431006","exception":false,"start_time":"2021-10-19T15:57:07.359218","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T04:46:36.065897Z","iopub.execute_input":"2021-10-20T04:46:36.066387Z","iopub.status.idle":"2021-10-20T04:46:36.075757Z","shell.execute_reply.started":"2021-10-20T04:46:36.066347Z","shell.execute_reply":"2021-10-20T04:46:36.075007Z"},"trusted":true},"execution_count":37,"outputs":[]},{"cell_type":"code","source":"print(\"Training Set Length: \"+str(len(df_train)))\nprint(\"Testing Set Length: \"+str(len(df_test)))\nprint(\"train_labels shape: \"+str(train_labels.shape))\nprint(\"test_labels shape: \"+str(test_labels.shape))","metadata":{"papermill":{"duration":0.075957,"end_time":"2021-10-19T15:57:07.570254","exception":false,"start_time":"2021-10-19T15:57:07.494297","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T04:46:36.077231Z","iopub.execute_input":"2021-10-20T04:46:36.077547Z","iopub.status.idle":"2021-10-20T04:46:36.594430Z","shell.execute_reply.started":"2021-10-20T04:46:36.077491Z","shell.execute_reply":"2021-10-20T04:46:36.593561Z"},"trusted":true},"execution_count":38,"outputs":[{"name":"stdout","text":"Training Set Length: 98625\nTesting Set Length: 98602\ntrain_labels shape: (98625,)\ntest_labels shape: (98602,)\n","output_type":"stream"}]},{"cell_type":"markdown","source":"# **Train/Test/Validation Set**\n**This dataset contains two csv files. One is Train set and another is Test set.IN EDA,we have seen that train and test set is equal sized and distribution of class is identical to both of the test and train set.\nI have used full train set for training and used 20% of the data from testset for validation set and 80% for testing\nThe validation set and the Test set is partitioned into stratified fashion to keep the distribution of class intact.**","metadata":{"papermill":{"duration":0.064416,"end_time":"2021-10-19T15:57:07.698168","exception":false,"start_time":"2021-10-19T15:57:07.633752","status":"completed"},"tags":[]}},{"cell_type":"code","source":"test_sentences, validation_sentences,test_labels, validation_labels = train_test_split(test_sentences, test_labels,stratify=test_labels,test_size=0.2)","metadata":{"papermill":{"duration":0.203136,"end_time":"2021-10-19T15:57:07.962753","exception":false,"start_time":"2021-10-19T15:57:07.759617","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T04:46:36.605700Z","iopub.execute_input":"2021-10-20T04:46:36.605940Z","iopub.status.idle":"2021-10-20T04:46:36.778256Z","shell.execute_reply.started":"2021-10-20T04:46:36.605910Z","shell.execute_reply":"2021-10-20T04:46:36.777493Z"},"trusted":true},"execution_count":39,"outputs":[]},{"cell_type":"code","source":"train_labels=keras.utils.to_categorical(train_labels)\ntest_labels=keras.utils.to_categorical(test_labels)\nvalidation_labels=keras.utils.to_categorical(validation_labels)\n\n\nprint(\"Training Set Length: \"+str(len(df_train)))\nprint(\"Testing Set Length: \"+str(len(df_test)))\nprint(\"training_sentences shape: \"+str(train_sentences.shape))\nprint(\"testing_sentences shape: \"+str(test_sentences.shape))\nprint(\"validation_sentences shape: \"+str(validation_sentences.shape))\nprint(\"train_labels shape: \"+str(train_labels.shape))\nprint(\"test_labels shape: \"+str(test_labels.shape))\nprint(\"validation_labels shape: \"+str(validation_labels.shape))","metadata":{"papermill":{"duration":0.530604,"end_time":"2021-10-19T15:57:08.532767","exception":false,"start_time":"2021-10-19T15:57:08.002163","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T04:46:36.779654Z","iopub.execute_input":"2021-10-20T04:46:36.779948Z","iopub.status.idle":"2021-10-20T04:46:36.813174Z","shell.execute_reply.started":"2021-10-20T04:46:36.779909Z","shell.execute_reply":"2021-10-20T04:46:36.812400Z"},"trusted":true},"execution_count":40,"outputs":[{"name":"stdout","text":"Training Set Length: 98625\nTesting Set Length: 98602\ntraining_sentences shape: (98625,)\ntesting_sentences shape: (78881,)\nvalidation_sentences shape: (19721,)\ntrain_labels shape: (98625, 7)\ntest_labels shape: (78881, 7)\nvalidation_labels shape: (19721, 7)\n","output_type":"stream"}]},{"cell_type":"code","source":"print(train_sentences[1])\nprint(train_labels[0])","metadata":{"papermill":{"duration":0.048235,"end_time":"2021-10-19T15:57:08.621791","exception":false,"start_time":"2021-10-19T15:57:08.573556","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T04:46:36.814571Z","iopub.execute_input":"2021-10-20T04:46:36.815012Z","iopub.status.idle":"2021-10-20T04:46:36.820225Z","shell.execute_reply.started":"2021-10-20T04:46:36.814972Z","shell.execute_reply":"2021-10-20T04:46:36.819480Z"},"trusted":true},"execution_count":41,"outputs":[{"name":"stdout","text":"  ক্ষমতাসীন    আওয়ামী    লীগের    ভ্রাতৃপ্রতিম    সংগঠন    বাংলাদেশ    ছাত্রলীগের    দুই    কর্মীকে    সিলেটের    জালালাবাদে    কুপিয়ে    জখমের    বাংলাদেশ    ইসলামী    ছাত্রশিবিরকে    দায়ী    ছাত্রলীগের    সভাপতি    সাইফুর    রহমান    ছাত্রশিবিরের    ওপর    পাল্টা    আক্রমণের    ঘোষণা    হামলার    প্রতিবাদে    বুধবার    ঢাকা    বিশ্ববিদ্যালয়ের    ছাত্র    শিক্ষক    কেন্দ্রের    টিএসসি    রাজু    ভাস্কর্যের    সামনে    সমাবেশে    সাইফুর    রহমান    ঘোষণা    ডিফেনসিভে    থাকলে    চলবে    অ্যাটাকিংয়ে    স্ট্রাইকারের    ভূমিকায়    স্লোগান    মিছিল    সমাবেশ    শিবিরকে    প্রতিহত    অ্যাটাকিংয়ে    ছাত্রলীগের    কর্মীদের    আপনারা    খুঁজুন    শিবির    কোথায়    প্রতিহত    করুন    রাজশাহী    বিশ্ববিদ্যালয়    লক্ষ্মীপুরে    শিবিরকে    প্রতিহত    শুরু    উল্লেখ    ছাত্রলীগের    সভাপতি    শোকের    মাসে    শোককে    শক্তিতে    পরিণত    জামায়াত    শিবিরকে    শক্ত    হাতে    দমন    জেলা    উপজেলা    পাড়া    মহল্লা    ইউনিয়নের    কোথাও    জামায়াত    শিবির    থাকতে    পারবে    খুঁজে    খুঁজে    দমন    দেশ    বিতাড়িত    সাধারণ    সম্পাদক    জাকির    হোসাইন    হামলাকারীদের    ঘণ্টার    গ্রেপ্তার    ছাত্রলীগের    নেতা    কর্মীরা    জামায়াত    শিবিরকে    সমুচিত    জবাব    দেবে    জবাব    কীভাবে    দেবে    ছাত্রলীগের    নেতা    কর্মীরা    জানেন    সমাবেশের    বিশ্ববিদ্যালয়ের    মধুর    ক্যানটিন    বিক্ষোভ    মিছিল    সংগঠনের    নেতা    কর্মীরা    ওয়েবসাইটে    ছাত্রশিবিরের    কেন্দ্রীয়    সভাপতি    ইয়াছিন    আরাফাত    সেক্রেটারি    জেনারেল    মোবারক    হোসাইন    বিবৃতিতে    দাবি    বিগত    দিন    এলাকার    শিক্ষাপ্রতিষ্ঠানে    ছাত্রলীগ    নেতা    কর্মীদের    জ্যেষ্ঠতা    কোন্দলের    জের    হামলা  \n[0. 0. 0. 0. 0. 1. 0.]\n","output_type":"stream"}]},{"cell_type":"markdown","source":"# **Tokenization**","metadata":{"papermill":{"duration":0.039494,"end_time":"2021-10-19T15:57:08.700854","exception":false,"start_time":"2021-10-19T15:57:08.661360","status":"completed"},"tags":[]}},{"cell_type":"code","source":"vocab_size = 100000\nembedding_dim = 150\nmax_length = 600\ntrunc_type='post'\noov_tok = \"<OOV>\"\n\nprint(train_sentences.shape)\nprint(train_labels.shape)\n\n\ntokenizer = Tokenizer(num_words = vocab_size, oov_token=oov_tok)\ntokenizer.fit_on_texts(train_sentences)\nword_index = tokenizer.word_index\n\n#464246 words in raw dataset\n#463651 words in raw dataset after stopwords removal\n#***** after stemming \n\nprint(len(word_index))\nprint(\"Word index length:\"+str(len(tokenizer.word_index)))\nsequences = tokenizer.texts_to_sequences(train_sentences)\npadded = pad_sequences(sequences,maxlen=max_length, truncating=trunc_type)\n\n\ntest_sequences = tokenizer.texts_to_sequences(test_sentences)\ntesting_padded = pad_sequences(test_sequences,maxlen=max_length)\n\nvalidation_sequences = tokenizer.texts_to_sequences(validation_sentences)\nvalidation_padded = pad_sequences(validation_sequences,maxlen=max_length)","metadata":{"papermill":{"duration":78.018667,"end_time":"2021-10-19T15:58:26.759234","exception":false,"start_time":"2021-10-19T15:57:08.740567","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T04:46:36.821639Z","iopub.execute_input":"2021-10-20T04:46:36.822112Z","iopub.status.idle":"2021-10-20T04:48:24.147078Z","shell.execute_reply.started":"2021-10-20T04:46:36.822072Z","shell.execute_reply":"2021-10-20T04:48:24.146294Z"},"trusted":true},"execution_count":42,"outputs":[{"name":"stdout","text":"(98625,)\n(98625, 7)\n464230\nWord index length:464230\n","output_type":"stream"}]},{"cell_type":"markdown","source":"# **After Tokenization Data visualization**","metadata":{"papermill":{"duration":0.04016,"end_time":"2021-10-19T15:58:26.840497","exception":false,"start_time":"2021-10-19T15:58:26.800337","status":"completed"},"tags":[]}},{"cell_type":"code","source":"print(\"Sentence :--> \\n\")\nprint(train_sentences[2]+\"\\n\")\nprint(\"Sentence Tokenized and Converted into Sequence :--> \\n\")\nprint(str(sequences[2])+\"\\n\")\nprint(\"After Padding the Sequence with padding length 100 :--> \\n\")\nprint(padded[2])","metadata":{"papermill":{"duration":0.052562,"end_time":"2021-10-19T15:58:26.932593","exception":false,"start_time":"2021-10-19T15:58:26.880031","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T04:48:24.148377Z","iopub.execute_input":"2021-10-20T04:48:24.148649Z","iopub.status.idle":"2021-10-20T04:48:24.158823Z","shell.execute_reply.started":"2021-10-20T04:48:24.148608Z","shell.execute_reply":"2021-10-20T04:48:24.157991Z"},"trusted":true},"execution_count":43,"outputs":[{"name":"stdout","text":"Sentence :--> \n\n  বয়সভিত্তিক    নাচ    প্রতিযোগিতার    আয়োজন    করেছিল    নৃত্যশিল্পী    সংস্থা    শিল্পকলা    একাডেমী    গতকাল    বৃহস্পতিবার    সন্ধ্যায়    শিল্পকলা    একাডেমীর    সংগীত    নৃত্যকলাকেন্দ্র    মিলনায়তনে    অনুষ্ঠিত    প্রতিযোগিতার    সমাপনী    অনুষ্ঠান    সারা    দেশের    নাচের    দলকে    পুরস্কার    পাশাপাশি    প্রতিটি    সেরা    ক্যাটাগরিতে    পুরস্কার    মোট    প্রতিযোগীকে    শিল্পকলা    একাডেমীর    মহাপরিচালক    লিয়াকত    আলীর    সভাপতিত্বে    সমাপনী    আয়োজনে    প্রধান    অতিথি    সংস্কৃতিমন্ত্রী    আসাদুজ্জামান    নূর    অনুষ্ঠানে    স্বাগত    বক্তব্য    নৃত্যশিল্পী    সংস্থার    সভাপতি    মিনু    শুভেচ্ছা    বক্তব্য    সংস্থার    সাধারণ    সম্পাদক    মাহফুজুর    রহমান  \n\nSentence Tokenized and Converted into Sequence :--> \n\n[8624, 2704, 9440, 1725, 514, 7067, 254, 1417, 6180, 16, 125, 835, 1417, 4808, 727, 1, 1225, 111, 9440, 2350, 340, 301, 19, 4418, 782, 218, 88, 315, 151, 4266, 218, 2351, 1, 1417, 4808, 1720, 5715, 2200, 1036, 2350, 6617, 27, 397, 6875, 2749, 1276, 70, 1518, 249, 7067, 761, 60, 15656, 1914, 249, 761, 62, 105, 8373, 49]\n\nAfter Padding the Sequence with padding length 100 :--> \n\n[    0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n     0     0     0     0     0     0     0     0     0     0     0     0\n  8624  2704  9440  1725   514  7067   254  1417  6180    16   125   835\n  1417  4808   727     1  1225   111  9440  2350   340   301    19  4418\n   782   218    88   315   151  4266   218  2351     1  1417  4808  1720\n  5715  2200  1036  2350  6617    27   397  6875  2749  1276    70  1518\n   249  7067   761    60 15656  1914   249   761    62   105  8373    49]\n","output_type":"stream"}]},{"cell_type":"code","source":"print(\"Padded shape(training): \"+str(padded.shape))\nprint(\"Padded shape(testing): \"+str(testing_padded.shape))","metadata":{"papermill":{"duration":0.048564,"end_time":"2021-10-19T15:58:27.023184","exception":false,"start_time":"2021-10-19T15:58:26.974620","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T04:48:24.160537Z","iopub.execute_input":"2021-10-20T04:48:24.160810Z","iopub.status.idle":"2021-10-20T04:48:24.171223Z","shell.execute_reply.started":"2021-10-20T04:48:24.160775Z","shell.execute_reply":"2021-10-20T04:48:24.170558Z"},"trusted":true},"execution_count":44,"outputs":[{"name":"stdout","text":"Padded shape(training): (98625, 600)\nPadded shape(testing): (78881, 600)\n","output_type":"stream"}]},{"cell_type":"markdown","source":"# **Model Creation**","metadata":{"papermill":{"duration":0.040586,"end_time":"2021-10-19T15:58:27.105314","exception":false,"start_time":"2021-10-19T15:58:27.064728","status":"completed"},"tags":[]}},{"cell_type":"code","source":"from tensorflow.python.keras.optimizer_v2.adam import Adam\nmodel= Sequential()\nmodel.add(Embedding(vocab_size, embedding_dim, input_length=max_length))\nmodel.add(Conv1D(150, kernel_size=3, activation = \"relu\"))\nmodel.add(MaxPool1D(pool_size=16))\nmodel.add(Bidirectional(LSTM(16, return_sequences=True)))\n#l2 regularizer\nmodel.add(Dense(10,kernel_regularizer=regularizers.l2(0.01),activation=\"relu\"))\nmodel.add(Flatten())\nmodel.add(Dense(7, activation='softmax'))\n#sgd= SGD(lr=0.0001,decay=1e-6,momentum=0.9,nesterov=True)\nadam=Adam(learning_rate=0.00005,beta_1=0.9,beta_2=0.999,epsilon=1e-07,amsgrad=False)\nmodel.summary()\nmodel.compile(loss='categorical_crossentropy',optimizer=adam,metrics=['accuracy'])","metadata":{"execution":{"iopub.status.busy":"2021-10-20T04:48:24.172833Z","iopub.execute_input":"2021-10-20T04:48:24.173237Z","iopub.status.idle":"2021-10-20T04:48:24.889408Z","shell.execute_reply.started":"2021-10-20T04:48:24.173202Z","shell.execute_reply":"2021-10-20T04:48:24.888702Z"},"trusted":true},"execution_count":45,"outputs":[{"name":"stderr","text":"2021-10-20 04:48:24.233563: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:937] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero\n2021-10-20 04:48:24.301065: W tensorflow/stream_executor/platform/default/dso_loader.cc:64] Could not load dynamic library 'libcusolver.so.11'; dlerror: libcusolver.so.11: cannot open shared object file: No such file or directory; LD_LIBRARY_PATH: /usr/local/cuda/lib64:/usr/local/cuda/lib:/usr/local/lib/x86_64-linux-gnu:/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/nvidia/lib:/usr/local/nvidia/lib64:/usr/local/cuda/compat\n2021-10-20 04:48:24.315627: W tensorflow/core/common_runtime/gpu/gpu_device.cc:1835] Cannot dlopen some GPU libraries. Please make sure the missing libraries mentioned above are installed properly if you would like to use GPU. Follow the guide at https://www.tensorflow.org/install/gpu for how to download and setup the required libraries for your platform.\nSkipping registering GPU devices...\n2021-10-20 04:48:24.317145: 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 AVX512F FMA\nTo enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.\n","output_type":"stream"},{"name":"stdout","text":"Model: \"sequential\"\n_________________________________________________________________\nLayer (type)                 Output Shape              Param #   \n=================================================================\nembedding (Embedding)        (None, 600, 150)          15000000  \n_________________________________________________________________\nconv1d (Conv1D)              (None, 598, 150)          67650     \n_________________________________________________________________\nmodule_wrapper (ModuleWrappe (None, 37, 150)           0         \n_________________________________________________________________\nbidirectional (Bidirectional (None, 37, 32)            21376     \n_________________________________________________________________\ndense (Dense)                (None, 37, 10)            330       \n_________________________________________________________________\nflatten (Flatten)            (None, 370)               0         \n_________________________________________________________________\ndense_1 (Dense)              (None, 7)                 2597      \n=================================================================\nTotal params: 15,091,953\nTrainable params: 15,091,953\nNon-trainable params: 0\n_________________________________________________________________\n","output_type":"stream"}]},{"cell_type":"code","source":"history=model.fit(padded,train_labels,epochs=10,batch_size=256,validation_data=( validation_padded,validation_labels),use_multiprocessing=True, workers=8)","metadata":{"papermill":{"duration":1063.631189,"end_time":"2021-10-19T16:16:13.647580","exception":false,"start_time":"2021-10-19T15:58:30.016391","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T04:48:24.890820Z","iopub.execute_input":"2021-10-20T04:48:24.891090Z","iopub.status.idle":"2021-10-20T06:31:52.264099Z","shell.execute_reply.started":"2021-10-20T04:48:24.891054Z","shell.execute_reply":"2021-10-20T06:31:52.263060Z"},"trusted":true},"execution_count":46,"outputs":[{"name":"stderr","text":"2021-10-20 04:48:24.894564: W tensorflow/core/framework/cpu_allocator_impl.cc:80] Allocation of 236700000 exceeds 10% of free system memory.\n2021-10-20 04:48:25.115806: I tensorflow/compiler/mlir/mlir_graph_optimization_pass.cc:185] None of the MLIR Optimization Passes are enabled (registered 2)\n","output_type":"stream"},{"name":"stdout","text":"Epoch 1/10\n386/386 [==============================] - 640s 2s/step - loss: 2.0560 - accuracy: 0.2692 - val_loss: 1.8943 - val_accuracy: 0.3688\nEpoch 2/10\n386/386 [==============================] - 624s 2s/step - loss: 1.1932 - accuracy: 0.6874 - val_loss: 0.6801 - val_accuracy: 0.8602\nEpoch 3/10\n386/386 [==============================] - 625s 2s/step - loss: 0.5373 - accuracy: 0.8933 - val_loss: 0.5194 - val_accuracy: 0.8894\nEpoch 4/10\n386/386 [==============================] - 627s 2s/step - loss: 0.4010 - accuracy: 0.9266 - val_loss: 0.4740 - val_accuracy: 0.8991\nEpoch 5/10\n386/386 [==============================] - 615s 2s/step - loss: 0.3274 - accuracy: 0.9461 - val_loss: 0.4593 - val_accuracy: 0.9040\nEpoch 6/10\n386/386 [==============================] - 609s 2s/step - loss: 0.2764 - accuracy: 0.9589 - val_loss: 0.4461 - val_accuracy: 0.9064\nEpoch 7/10\n386/386 [==============================] - 602s 2s/step - loss: 0.2364 - accuracy: 0.9685 - val_loss: 0.4551 - val_accuracy: 0.9055\nEpoch 8/10\n386/386 [==============================] - 615s 2s/step - loss: 0.2051 - accuracy: 0.9753 - val_loss: 0.4519 - val_accuracy: 0.9071\nEpoch 9/10\n386/386 [==============================] - 613s 2s/step - loss: 0.1798 - accuracy: 0.9807 - val_loss: 0.4756 - val_accuracy: 0.9026\nEpoch 10/10\n386/386 [==============================] - 615s 2s/step - loss: 0.1588 - accuracy: 0.9845 - val_loss: 0.4557 - val_accuracy: 0.9059\n","output_type":"stream"}]},{"cell_type":"markdown","source":"# **Learning Curve**","metadata":{"papermill":{"duration":1.030315,"end_time":"2021-10-19T16:16:15.704991","exception":false,"start_time":"2021-10-19T16:16:14.674676","status":"completed"},"tags":[]}},{"cell_type":"code","source":"print(history.history.keys())\nloss = history.history['loss']\nval_loss = history.history['val_loss']\nplt.plot(loss)\nplt.plot(val_loss)\nplt.title('model loss')\nplt.ylabel('loss')\nplt.xlabel('epoch')\nplt.legend(['loss', 'val_loss'])\nplt.show()\n\naccuracy = history.history['accuracy']\nval_accuracy= history.history['val_accuracy']\nplt.plot(accuracy)\nplt.plot(val_accuracy)\nplt.title('model accuracy')\nplt.ylabel('accuracy')\nplt.xlabel('epoch')\nplt.legend(['accuracy', 'val_accuracy'])\nplt.show()\n","metadata":{"papermill":{"duration":1.500424,"end_time":"2021-10-19T16:16:18.227470","exception":false,"start_time":"2021-10-19T16:16:16.727046","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T06:31:52.266207Z","iopub.execute_input":"2021-10-20T06:31:52.266436Z","iopub.status.idle":"2021-10-20T06:31:52.705380Z","shell.execute_reply.started":"2021-10-20T06:31:52.266406Z","shell.execute_reply":"2021-10-20T06:31:52.704722Z"},"trusted":true},"execution_count":47,"outputs":[{"name":"stdout","text":"dict_keys(['loss', 'accuracy', 'val_loss', 'val_accuracy'])\n","output_type":"stream"},{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 1 Axes>","image/png":"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\n"},"metadata":{"needs_background":"light"}},{"output_type":"display_data","data":{"text/plain":"<Figure size 432x288 with 1 Axes>","image/png":"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\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"markdown","source":"# **Evaluation**","metadata":{"papermill":{"duration":1.06426,"end_time":"2021-10-19T16:16:20.320319","exception":false,"start_time":"2021-10-19T16:16:19.256059","status":"completed"},"tags":[]}},{"cell_type":"code","source":"#accuracy calculation\nloss_and_metrics = model.evaluate(padded,train_labels,batch_size=256)\nprint(\"The train accuracy is: \"+str(loss_and_metrics[1]))\nloss_and_metrics = model.evaluate(testing_padded,test_labels,batch_size=256)\nprint(\"The test accuracy is: \"+str(loss_and_metrics[1]))","metadata":{"papermill":{"duration":18.385085,"end_time":"2021-10-19T16:16:39.974096","exception":false,"start_time":"2021-10-19T16:16:21.589011","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T06:31:52.706754Z","iopub.execute_input":"2021-10-20T06:31:52.707014Z","iopub.status.idle":"2021-10-20T06:35:52.519229Z","shell.execute_reply.started":"2021-10-20T06:31:52.706979Z","shell.execute_reply":"2021-10-20T06:35:52.518441Z"},"trusted":true},"execution_count":48,"outputs":[{"name":"stderr","text":"2021-10-20 06:31:52.710116: W tensorflow/core/framework/cpu_allocator_impl.cc:80] Allocation of 236700000 exceeds 10% of free system memory.\n","output_type":"stream"},{"name":"stdout","text":"386/386 [==============================] - 134s 347ms/step - loss: 0.1471 - accuracy: 0.9865\nThe train accuracy is: 0.9865247011184692\n","output_type":"stream"},{"name":"stderr","text":"2021-10-20 06:34:06.734696: W tensorflow/core/framework/cpu_allocator_impl.cc:80] Allocation of 189314400 exceeds 10% of free system memory.\n","output_type":"stream"},{"name":"stdout","text":"309/309 [==============================] - 106s 342ms/step - loss: 0.4548 - accuracy: 0.9061\nThe test accuracy is: 0.9061371088027954\n","output_type":"stream"}]},{"cell_type":"code","source":"predictions = model.predict(testing_padded)","metadata":{"papermill":{"duration":12.073365,"end_time":"2021-10-19T16:16:53.145067","exception":false,"start_time":"2021-10-19T16:16:41.071702","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T06:35:52.520369Z","iopub.execute_input":"2021-10-20T06:35:52.521067Z","iopub.status.idle":"2021-10-20T06:38:15.025175Z","shell.execute_reply.started":"2021-10-20T06:35:52.521026Z","shell.execute_reply":"2021-10-20T06:38:15.024295Z"},"trusted":true},"execution_count":49,"outputs":[{"name":"stderr","text":"2021-10-20 06:35:52.523306: W tensorflow/core/framework/cpu_allocator_impl.cc:80] Allocation of 189314400 exceeds 10% of free system memory.\n","output_type":"stream"}]},{"cell_type":"code","source":"predictions = np.argmax(predictions, axis=-1) ","metadata":{"papermill":{"duration":1.350268,"end_time":"2021-10-19T16:16:55.773345","exception":false,"start_time":"2021-10-19T16:16:54.423077","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T06:38:15.027101Z","iopub.execute_input":"2021-10-20T06:38:15.027359Z","iopub.status.idle":"2021-10-20T06:38:15.033791Z","shell.execute_reply.started":"2021-10-20T06:38:15.027324Z","shell.execute_reply":"2021-10-20T06:38:15.033046Z"},"trusted":true},"execution_count":50,"outputs":[]},{"cell_type":"code","source":"test_labels=np.argmax(test_labels, axis=-1)","metadata":{"papermill":{"duration":1.144988,"end_time":"2021-10-19T16:16:58.020714","exception":false,"start_time":"2021-10-19T16:16:56.875726","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T06:38:15.035180Z","iopub.execute_input":"2021-10-20T06:38:15.035481Z","iopub.status.idle":"2021-10-20T06:38:15.043775Z","shell.execute_reply.started":"2021-10-20T06:38:15.035435Z","shell.execute_reply":"2021-10-20T06:38:15.042937Z"},"trusted":true},"execution_count":51,"outputs":[]},{"cell_type":"code","source":"precision, recall, fscore, _ = precision_recall_fscore_support(test_labels,predictions, average='weighted')\nprint(\"Precision :\"+str(precision))\nprint(\"Recall :\"+str(recall))\nprint(\"fscore :\"+str(fscore))","metadata":{"papermill":{"duration":1.143419,"end_time":"2021-10-19T16:17:00.268536","exception":false,"start_time":"2021-10-19T16:16:59.125117","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T06:38:15.045600Z","iopub.execute_input":"2021-10-20T06:38:15.046544Z","iopub.status.idle":"2021-10-20T06:38:15.084198Z","shell.execute_reply.started":"2021-10-20T06:38:15.046517Z","shell.execute_reply":"2021-10-20T06:38:15.083540Z"},"trusted":true},"execution_count":52,"outputs":[{"name":"stdout","text":"Precision :0.9067235401713108\nRecall :0.9061370925824976\nfscore :0.9060858303839899\n","output_type":"stream"}]},{"cell_type":"markdown","source":"# **Confusion Matrix & Classification Report**","metadata":{"papermill":{"duration":1.110767,"end_time":"2021-10-19T16:17:02.489299","exception":false,"start_time":"2021-10-19T16:17:01.378532","status":"completed"},"tags":[]}},{"cell_type":"code","source":"#['economy' 'sports' 'international' 'state' 'technology' 'entertainment' 'education']\n\ncm = metrics.confusion_matrix(test_labels,predictions)\n# Transform to df for easier plotting\ncm_df = pd.DataFrame(cm,index = ['economy' ,'sports' ,'international', 'state','technology', 'entertainment', 'education'], \n                     columns = ['economy' ,'sports' ,'international', 'state','technology', 'entertainment', 'education'])\n\nplt.figure(figsize=(9,9))\nsns.heatmap(cm_df, annot=True, fmt=\".3f\", linewidths=.5, square = True, cmap = 'Blues_r');\nplt.ylabel('Actual label');\nplt.xlabel('Predicted label');\nall_sample_title = 'Accuracy Score CNN-BiLSTM Hybrid: {0}'.format(loss_and_metrics[1])\nplt.title(all_sample_title, size = 15);\n\nreport = pd.DataFrame(classification_report(y_true = test_labels, y_pred = predictions, output_dict=True)).transpose()\nreport = report.rename(index={'0': 'economy','1':'sports','2':'international','3':'state','4':'technology','5':'entertainment','6':'education'})\nreport[['precision','recall','f1-score']]=report[['precision','recall','f1-score']].apply(lambda x: round(x*100,2))\ndisplay(report)","metadata":{"papermill":{"duration":2.161845,"end_time":"2021-10-19T16:17:05.757205","exception":false,"start_time":"2021-10-19T16:17:03.595360","status":"completed"},"tags":[],"execution":{"iopub.status.busy":"2021-10-20T06:38:15.085436Z","iopub.execute_input":"2021-10-20T06:38:15.085697Z","iopub.status.idle":"2021-10-20T06:38:16.211630Z","shell.execute_reply.started":"2021-10-20T06:38:15.085661Z","shell.execute_reply":"2021-10-20T06:38:16.210926Z"},"trusted":true},"execution_count":53,"outputs":[{"output_type":"display_data","data":{"text/plain":"               precision  recall  f1-score       support\neconomy            88.27   89.33     88.80  11584.000000\nsports             97.84   96.33     97.08  12693.000000\ninternational      93.20   90.44     91.80  12250.000000\nstate              89.73   84.26     86.91  12146.000000\ntechnology         83.24   84.01     83.62   7948.000000\nentertainment      91.78   94.75     93.24  12194.000000\neducation          86.98   92.97     89.87  10066.000000\naccuracy           90.61   90.61     90.61      0.906137\nmacro avg          90.15   90.30     90.19  78881.000000\nweighted avg       90.67   90.61     90.61  78881.000000","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>precision</th>\n      <th>recall</th>\n      <th>f1-score</th>\n      <th>support</th>\n    </tr>\n  </thead>\n  <tbody>\n    <tr>\n      <th>economy</th>\n      <td>88.27</td>\n      <td>89.33</td>\n      <td>88.80</td>\n      <td>11584.000000</td>\n    </tr>\n    <tr>\n      <th>sports</th>\n      <td>97.84</td>\n      <td>96.33</td>\n      <td>97.08</td>\n      <td>12693.000000</td>\n    </tr>\n    <tr>\n      <th>international</th>\n      <td>93.20</td>\n      <td>90.44</td>\n      <td>91.80</td>\n      <td>12250.000000</td>\n    </tr>\n    <tr>\n      <th>state</th>\n      <td>89.73</td>\n      <td>84.26</td>\n      <td>86.91</td>\n      <td>12146.000000</td>\n    </tr>\n    <tr>\n      <th>technology</th>\n      <td>83.24</td>\n      <td>84.01</td>\n      <td>83.62</td>\n      <td>7948.000000</td>\n    </tr>\n    <tr>\n      <th>entertainment</th>\n      <td>91.78</td>\n      <td>94.75</td>\n      <td>93.24</td>\n      <td>12194.000000</td>\n    </tr>\n    <tr>\n      <th>education</th>\n      <td>86.98</td>\n      <td>92.97</td>\n      <td>89.87</td>\n      <td>10066.000000</td>\n    </tr>\n    <tr>\n      <th>accuracy</th>\n      <td>90.61</td>\n      <td>90.61</td>\n      <td>90.61</td>\n      <td>0.906137</td>\n    </tr>\n    <tr>\n      <th>macro avg</th>\n      <td>90.15</td>\n      <td>90.30</td>\n      <td>90.19</td>\n      <td>78881.000000</td>\n    </tr>\n    <tr>\n      <th>weighted avg</th>\n      <td>90.67</td>\n      <td>90.61</td>\n      <td>90.61</td>\n      <td>78881.000000</td>\n    </tr>\n  </tbody>\n</table>\n</div>"},"metadata":{}},{"output_type":"display_data","data":{"text/plain":"<Figure size 648x648 with 2 Axes>","image/png":"iVBORw0KGgoAAAANSUhEUgAAAlUAAAIoCAYAAABAuh6tAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjQuMywgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/MnkTPAAAACXBIWXMAAAsTAAALEwEAmpwYAADVvUlEQVR4nOzdd3wUxfvA8c+ThBJKEkoChK4gVaQTQDpSlCYCohQBvz9s2AFRlCJdUJqKoiBIF6QX6UVRegepSg2QENKA0JL5/bGb5JJcIIGcEH3evO5FbnZ2d+a23Owzs3tijEEppZRSSt0ftwddAKWUUkqpfwNtVCmllFJKpQFtVCmllFJKpQFtVCmllFJKpQFtVCmllFJKpQGPB10ApZRSSqUv12/j0kcHZPZAXLl8V9FIlVJKKaVUGtBIlVJKKaVSRR9x6ZxGqpRSSiml0oBGqpRSSimVKsa1Q6ogfQ6p0kiVUkoppVRa0EaVUkoppVLHuPiVAiIyWUSCROSAQ9pIETksIvtEZIGI+DhM+1BEjovIERFp7JDexE47LiJ9HNKLishWO32OiGS8W5m0UaWUUkqp9GgK0CRR2mqgrDGmHHAU+BBAREoD7YEy9jxfi4i7iLgDXwFNgdLAC3ZegBHAaGNMMSAUePluBdJGlVJKKaVS5SEIVGGM2QRcTpS2yhhz2367BShg/90SmG2MuWGM+Rs4DlS1X8eNMX8ZY24Cs4GWIiJAfWCePf9UoNXdyqSNKqWUUkr9G3UDVth/5wfOOEw7a6cll54LCHNooMWm35He/aeUUkqpVHH1c6pEpDvQ3SFpojFmYirm7wvcBmakddnuRBtVSimllHqo2A2oFDeiHIlIF6AZ0MCYuObfOaCgQ7YCdhrJpIcAPiLiYUerHPMnS7v/lFJKKZUqxsX/7pWINAF6Ay2MMdccJi0G2otIJhEpChQHtgHbgeL2nX4ZsQazL7YbY+uBNvb8LwGL7rZ+bVQppZRSKt0RkVnAH0AJETkrIi8DXwLZgdUiskdEvgEwxhwEfgIOAb8Abxhjou0oVA9gJfAn8JOdF+AD4D0ROY41xmrSXctk9Ad8lFJKKZUKkddjXNp4yJ7ZLV0+Ul0jVUoppZRSaUAHqiullFIqVbSPyzmNVCmllFJKpQGNVCmllFIqVXQ4tnMaqVJKKaWUSgMaqVJKKaVUqtzPs6T+zbRRpZRSSqlU0e4/57T7TymllFIqDWijSimllFIqDWijSimllFIqDeiYKqWUUkqlio6pck4jVUoppZRSaUAbVWlIRP4WESMixR50WR42IlJWRBaKyHkRibI/q9kiUvZBly0lRKS8iMwRkQsiclNEAkVkhohUccizwd7+HzqZ/5KIDHB4P8DOu9JJ3nkisuEu5alrzx/7uiUif4nIZyKSNVFeIyI9HN5PEZEdd1n+kyKyWkSCReSqiByz5yvgZN1OX4nqeSyZ9Ryzpw9wNt0h3wYRmZfMtJMiMupO86dmeSmc/67rtPd5IyJ172H5mUTkcxEJsj//ZSJSJIXzfWHvp1Ei8quIVHaSL7+ILBCRSHvf/FJEsjjJV1hEZonIZRG5JiJ7RaSJw3RfERknItvs4+JkMuX6RkQOi8gVEQkVkU0i0jBRnil32J9ecMj3vIjMt88lRkS6JLPOu9ZRRLxEZIy9Pa+JyJ8i8o6ISKJ8T4nIZhEJF5GL9nJLOFnn/9n79HUR2SkiDRJNbyjWeeSUvb4DItJDRNwT5Uvuc7iRKF8ZEVllL+uSiEwQkWwp/FxLJvO5uYnIDjtPM2d5wHqkgiv/pVfa/ZdGRKQ6UMR++wIw6MGV5uEiViNzC7AN6AGEAsWBtkA54MCDK93diUhrYDawCXgXOAfkBzoAq4AciWZ5V0TGGmOupWDxjUSkijFm+z0WrwPwF5ABqAgMAXyA7g55qgN/p3SBIvIksAFYCLwMRAGlgReBwsAue5mxagKjgNbAeSeLvA4UFZHKxpi4xpxYDdIi9vT05lkgxIXLHwe0wdrfgoEBwGoRedwYc6fPaxzQHvgAOAW8BawRkSeMMacARCQDsBK4aef1Ab6w/+8YuyARKQj8AewFugJXgfKAp8P68gPPA1uBPYBfMuXyBL4EjgAZsfarFSJSyxizxc4zCPgm0XyvYe13qx3S2mDtN0uB/zlbWUrrCEwBagMfAceBenY+AUbby6oELMM6HgYC2YF+WJ9rGWNMhJ3vBbv8A4DfsD6zpfbxHXuO6w5kAT4GzgBPAp8DRYH3HcrleHzFWgJsdqijN7AOOIq1DXIBnwH5gFaJ5j1sl8fRSSfrAOszLZDMNHU3xhh9pcEL62R2BavxcOhBl8ehXO5AxgdchiFYX0CZnEyTf2D9nvcxrz8QCUx1VlagmcPfG4DfsRoJ7yTKdwkY4PB+gP2Z7AMWJso7D9hwl3LVxfpN07KJ0j8Dgu4y7xRgxx2mzwD2J1Nfp5+BXZYiTqYNsOu+ChiVaNrnWF98CT6bZMq0AZiXzLSTiZedgu2a7PLSal8CytqfS91UrqMAcBvo7JCWH6uB8L+7zBcNvOyQlgnrIuBLh7QX7HxFHdLaATFAcYe02cCvgNsd1unm8Pco4GQK6+gOnAbG3SXfQWCFs3UC2ezPt4uT+e5aR6zGTTTwZqJ55wNbHd4PBy4AHg5p5ex1N3VIOwJMdiynfRxNd0jL7aSsQ7EuXJKcHx3yVLHX97xD2odABODjkNbczlfZIW0KdzjeE60nB1Yj/mV7Oc2Syxscecu48pWaY+Zhemn3XxqwQ7ftgMXAZKCUiDzhJF9tEVlvh8DDxeqCqOAwPTbUfskO5+4TkRftabFdLmUTLTNBN4Yd6t0hIq1E5CDWF3w1EcknIpPF6iKKEpGjIjJYRDImWp6nWF1Ip0TkhljddMPsaZ/Z8ycOjXcRK/Tvm8xH5AOEGWNuJJ5g7CPZYVnPitWVECUiISKyXEQKO0yvLyJb7fD6RRH52jHc7fA5NRaRxSJyBesKGREpJFaXY2xXxkpnIfxE/od1Zf1+4rLa5V+aKCkQ+AHoKSKZ7rJsg9XgbCEij98lb0pFYkWt4kii7r8U8MFqmDmr773G5WcD7WL3Hfv/dnZ6mhGRp0UkRkSKJkovaqe3TJTeXayunyixutjyO0wrYn92HUTkRxEJw4oWOO3+E5HXReSMWN11S7AiBveikf3//NgEY8w5rOhH0zvM9zjWF3lcVMc+5jYBzzjkawpsN8Y4Ri8XYjXamth18caKPH5tjIlJboV3mnYnxphoIAzr2HJKRMphRUhn3cM671pHrIadGxCeaN4wrEhVrAzANWPM7UR5iM0nIo8AjwE/JSrnXBy2mTHmkpOy7gYyAznvUJ8XsCKFSxzSymM1lsIc0lZjnVcct3dqDMKKhq29x/n/87RRlTbqAXmwviDmAbewDoI4Yo2rWGtPewkrXPsr1hUoIuKHFWqvAvTEuuKYBBS8h/IUwYpYDMM6oP8GcgOXgfewTiojscLB4x3KKMAirJD7V8DTQH97XrAajEWBOonW1xVYYowJTqY8u4BHRGSsiJROrtAi0gnri+QE1hduV6zQtq89vQzwC1Zk4zm7bC9ifeaJTcLqtmgBTBKRnFhfSiWAV+3lZ8UK4Xs6mT9WHawTl7OTYXJGYO0PicPtzswFjgF9U7F8R+4i4mE3hmsCrwAL7nFZsXYB9UTkE/vLIi3Mx/pMnrTf18LarvOTnSMpseua4JUoz0qshu1LidK7AEFY3TixqgNvYh0TL2NFHxY6We8orMZqW6yogrOCtcQ6ZpZiNUb2Yx0vifNNkWTGHTkoCZw1xlxJlP6nPS05me3/byZKvwkUdtjPS2J1B8UxxtzEOu5il18RqzFhxBpLdEtEzorIh4kvqlJKLB4ikktE3sUaApDkM3LQHuuicOE9rO6udTTGRGI1gnqLNWYyu1hjiNphbctY0wF/EflARHKI1S36hb382MZH7OeWYJ1Y2yznHS44wdoPw7D2zyQcLkAWmYRDCjKTdFvfxorGlUqUXlpEIuwL5d9EJPE5PLYR2w3r++eujItf6ZWOqUobL2AdFL8YY26KyCqgvYh86HBlPwzrS76xQ9ovDst4F/AGKhljYsel3OvVQi6goTFmj0PaWRwOFhHZjHXlM1lE3rRPOI2Ap4CWxpjFDvP+CGCMOWzP1xWr+yT2Cq0WVuMlOVPtZb8FvCUil4HlwFhjj7ERETesMPsCY4xjg9SxHJ9gjRNpYV/pYi9rjohUN8b84ZB3rjHmE4f6DsJqRJU3xlx2+AxOYp1IHE+ijvJjXUmmmDHmpIjMwDpZf5/oCjdx3hixIoGTRKSfMeZoataFNY7F0W/A26lcRmIjscZJfQp8KiLnsbbDF/dQPgCMMWEi8gvWF+Wv9v+/GGPCU/Ed3RrrouRO64kWkSnASyIy0Bhj7C+ll7C6YRy3hR9Q3RhzGkBETgG/iUgTY4zjsbnFGPPGXcrW167Pa/b7lfYXaeIxP9FYX3x3koP4SIijUJKO33N03P6/CvERNbHfiz1vVAqXn9f+/1tgAtYYoHrAYKzIztd3qYMzzxMfdbqK1ZW17S75lxt7zFIqpfQz7IzV3R17jBvgQ2PM1NgMxpjddmNrLtY5CqzGU2OH6HvsMhOvM9RhepKLTvsi8zVgROw5zYlaWOehxFHd48CLIpLBGBN7XFTCisA5Rr12Y415O4R1IfM+1vi8JxN9/uOxuomPSwpuilDOaaTqPonVfdYaqzEQe9UwG2tAb3U7T1agGjD1Dt0n9bFOys4G+qbWuUQNqtirxHdE5JCIRGF9Oc3AGnNRyKEMlxM1qBKbBDwn8V1uXYCLJGwgJmCMuW2MeR54AqthtBPryusPEYkNU5fAGr/0wx3WXRXrc3Y8+fyM9SX1ZKK8yxK9b4gVGo9wiHBE2mVJcndU4ircZbozw7D2gQ4pyDsda3xJkrsGAe4SmWmP9aVZDatxnxtYaDdS74n9JdYAqIEVmTmB1TjYJSIV73W5WMdFG7G6RduQ+q6/dVh1TfxKfMxMxvrs69rv69nvE+9bu2IbVADGmM1Y0YKqifIl3pcSsLdJRawor6MkUThjzMvGGJfcHWyM2Y/VdTNKRCrZjbqhWN1SYEUwUiq2pbvCGNPHGLPeGNMP6wLJ6X6aAiuxtldTrGjqbEnmzkgRqQY8QqKuPxcYjXXsdMWKSn8MDBCRlx3KUgaYibU9GwItsRpLy0XE615XLCI5sM5f+0gmAmp7wV5f4juFv8NqJI0Xkbx2Ob/GarjHbWtjzFhjzARjzEZjzDysY/sc1uD82LK0xzoHD05xBTRU5ZQ2qu5fU6wxKMtFxEdEfLCiODeI7wLMgXWSulODKdddpqfGRSdp72B1YyzAOilUBWKvvmO7DVJShp+wDtjY8TEvAT/eKRoTyxizzxgz2BjTCOsAPk/8QZzL/v9O689HorrZDawQko5HSPwZ5Ma68r2V6FWPO3exniO+0ZlixpgjWN2Sfe7WwLE/u8+AjuIwfsxB4jI7OmiM2WGM2WaMmY3ViKuP1XV7z4zlD2NMX2NMLayGZwxWo/heLcYaXDwEK2q45M7Zkwi165rgRaIuEGPMX1jHYGz3a1dgmzHmYKLlOetuCSLpWChnx5Oj3FjRgcTLc9qdkwKhWFHrxHIQH/lIThfgGrDDXn9zYCzWfhN7t2JKlh/7//pEedYBBe6lMWGMid1+vxhjOmENd/g0meztsS567tigvYO71tHu7noN6GaMmWKM2WSMGQqMwWqYxh63g4BjdoN4rX3R+QzWUIjYSGTs55V4nTkSTcded2asRngmrMh74m682HweWEMdfk6cxxhzGOtuwhewzpv7sO6w3oM1sN4puwtxOdaFQOydkiOxhi242d9hsds3q4hkT25ZKiltVN2/2IbTXKwDJxTrVtlMQFuxBrGHYn0h3WngashdpsfeRp14YKez7gBn7fy2WHc79TXGrDLWLfxXU1kGjDFXsSIMXbC+vAtx5+hScss5ifWZxY5FiD3h32n950l0y7b9+ebCGi+WYBWJ3l/G+lJ3Fum4U9fOBqCyWGOyUmsIVuOxTQryTsb6EvzAybTE5b2TP+3/E4+puC925HM1dx7Tc7dlXMUac/Qu1hi8xPtfWvoeK6KaHyuS7GwfdXb7vx9JG/Z3u26+hBUdSLy85B4vcDeHgYKS6HljOBknlJgx5rgxpgLwqJ3/caxz0S6HLqLDJNqOdsT9EYflx+5HiftmY9/f0wD1RHbb60y4Aqsx0w7rrtioe1x2SuoYO32Pk3L5EH+hVzJxHmNMKNZQhEcd1ue4TBzeX3Ycb2qfs2ZiDcJvaoy5U6O9AVY0ymnEzhgzGWusYjmsSH8PIPYRNnfiGA/KinXn6BfEf4fttafNJpnhD0afU+WUNqrug33Sa461w9dL9HoPa2evb395bAU632GQ51qgsYjkSWb6Wfv/uC9Le8BkSr/kPLGiZ44Sd02txRpUmewD32yTsPr5B2CNN7njiV6sQfjOFCc+CnAEKyqUeICxo63As5LwQXmtscYG/naXMq8FyhAf2XF8HbnDfJOwrvKdPujRofsyCWPMPqxozEck/XJKnPeGvY5uJGpYOonM3Ens3aFn7pIvWc62l73fPsrdozZ3MwHrM0n8PKK0Nh8rgjUb6zznrKuxoojERSHFGujvh3W1n2J2pHE3VgTYUevULMfBKvv/Zx3K5o91zK1IYZn+svfrXFgNlEkOk1cAVRJFRVtgNb5+sec/ifU4g/qJFt0AOGGSDqJPFXt/Su75abWxGgj30/V31zpiNYrAjtg4qIR1wXnJIV8FxwwikgvrhqCTEBcdPYp18Rqbx81+n3ibfY11s1CLu5x7ID4KtSG5DMaY68aY/XbjrCPW/v5TcvnFumHhGayhD2A9Cijx91dssOAjUjaEQdl0oPr9aYn1rJOxxpitjhPsQdB9sXbO1UAfYA3WA+8mYh201bHuLFuK1bffGfhVRIZgfSmWArIaYz4zxpwV6ynYg0TkGtaB8xFJIzTJWY01SHwr1hiZDlhXNInzrARmisinWHeB5QNqG2Neic1kjNkq1uMansS62+xuPhHrERMzsa6As2J94TTHHjxvrAHbvYEZ9iDvWVhXUvWBWXZjYjDWl9dCEZmAdXU1AlhpEg5Sd+YLrBPOOhEZj9WAy4M1juI3Y0xyV4KBYj2xeZaIFMCKKMU+/LM91hfAnaJYQ7AagynxLdY2rQFsTOE85ezxbW5YV+GfYI3PutuXbw4RcRZBWw58b38h/Iy1r+TA6kJ7AocvjXthjNnAHb4g0oox5rq9H72Btf+EOckWDCwTkf5YXeAjsCI6yY4PvIOhwHx7v1yAtV81SZxJRCYBde40rso+1icBY+zGR+zDP09hjb+LXVY/oJ8xxsMh7S2sqO85rIuWD7HuRHRsVM3DOjfNF5FPsLqsRgMzjTGOT77/BPhZREZiNfTqAp2wzlOOdYrdjx4Dsji832iMCRaRWlgXmQuw9s1cWBdPAVjngMTaYzVoVjuZFju4uzTxwxYqi/XolGBjTOxxk5I67rBfk+3P8m+sc9o7WOf02HDJN1jnnClY56WsWBHlm1jjUmMNAKaLdXfnZruOxbHuUI4t+0dYXXbDgBgRCXCY/5BxGJRvjz1sBUwxTh4jYXfB9sV6ZMZtrMbQ+8D/mfibcbyxosPTsQa258aKFPtjH8v2RcGGRMsuYv+5P/F3W6x7frjKv515CB6WlV5fWFfcR+8w/Wusu0Ey2e/rYB0A1+z09Vh3o8XmLwzMwQq/XsMKwbZ3mF4Ma+e/ihXZaUmihxiSzIPesMay/IDVCLuM1T0S+9DGsg75PLEiJmexIlt/A0OcLG+wXUavFHxOAfa6j9nzXMJ6SGZ7J3lbY11BXcf6clgGFHaY3gCrkXIdq7vsayCbw/S6ievkMC12IPxFu24nsU42ZVJQhwpYV38XsSJXgfa8FR3yJNgWDumr7DINcEgbAFxykvcjO++Gu5Qntp6xrxh7m80CHk2U1wA9Eu0jyQ0PLYLVGJhtb/vrWOMz1gKNkinLXR/+eZe6uOThn1iDig3WnbBOl4f1eI3TWHfFrQAKOuQpYs+f5AGIztaJ1fVyFmsfX451x6vB4eGf9md/MgX7WyasC4FgrON9OQ4PsnT4bE2itA/s7XYD68JsBJDFyfILYD2q4ArWcfZVMvk6Yl0I3cT6Un7VSZ7k9qW6Dp/jPOLPKWexvuirO1mWh13nb+7w2QxIZn0bUltHrLscv8dqsF6z6/ohiR6YjBXt2471sM0ge3uUd1K2/7M/pxtYF6UNnOx3d/y8HPK2stMDkvkcsmKdWy5j7b/bgVaJ8mTGitqescsUjhWpc7rMlOz7sa8L4TeNK193O0Ye1pfYH6BSqSIi24AjxhpwqtRDR0Q+w/oyfMTc40MqlVLOXYi45dLGQ16vDPf0PLQHTbv/VKqI9eOs9bn7AG+lHgixnpIf+/yfgdqgUkr9U7RRpVJrO1bX5Yfm3n8EWClX+hbr2UOLsX6TUymV1rSTyyltVKlUMcaky5Cs+u8wxtR90GVQ6t/OaKvKKX2kglJKKaVUGtBI1cNHm/9KKaXuh8t7FPQeN+e0UfUQOnrh2t0zpWOP5c2CZ4UeD7oYLhe1+0s8K771oIvhUlG7xuFZrdeDLobLRW0diWfNvg+6GC4VtXnIv/64jNr9JZ5Ve949YzoXtc3ps4rVP0AbVUoppZRKFQ1UOadjqpRSSiml0oBGqpRSSimVKjqmyjmNVCmllFJKpQGNVCmllFIqlTRU5YxGqpRSSiml0oBGqpRSSimVKjqmyjmNVCmllFJKpQGNVCmllFIqVTRQ5ZxGqpRSSiml0oBGqpRSSimVKjqmyjmNVCmllFJKpQGNVCmllFIqVYyOqnJKI1VKKaWUUmlAI1VKKaWUSh0NVDmlkSqllFJKqTSgkSqllFJKpYoGqpzTRpVSSimlUkUfqeCcdv8ppZRSSqUBjVQppZRSKlX0kQrOaaRKKaWUUioNaKRKKaWUUqmjgSqnNFKllFJKKZUGNFL1EBs7fADb/9iEd46cfDVlHgCREeF8NuADLl4IJE9efz4Y+BnZsnux5bf1zJg0AXET3N3d+V+PXpQpVyFuWdeuXuH1l54j4Ml6vPpOHwA2rlnB3OmTERFy5vblvb6D8fbJkaAMxhgmjvuMnVs3kylTZt7+cCDFHisFwNpfFjPnx+8BeL7z/2jQpAUAx48cYsyw/ty8eYNK1WrS/a3eiMh9fRbf9O9A09plCb4cSeW2QwHo+8rTdGtdg+DQKwD0/3IxK387lGTep2qUYlSvNri7uTFl4e+M+mE1AIX9czFteFdyemdl95+n6fbxj9y6HU3GDB5MGtSJCqUKcTn8Kh0/mMzp85fvq/zJ1+tFmtYqY9Wr3XAAcnhlYdrwLhT2z8mpwMt0/OAHwiKjeKyIHxMHdKB8yYIM+GopY6atc7rMwv45mTasCzl9srL7zzN0+3iaQ706UqFUQS6HXaVjnylx9erZ9Sm6tAogOjqG90f+zJo/DqdZHTNl9GDNN6+RMaMHHu5uLFi3n8HfreKHgS9QsVQBbt2OYceh0/QY9jO3o2Pwye7Jtx+3o2j+XNy4eYtXBv/Eob8uJq1nvhxMG9yRnN5Z2H34LN0GzLbr6c6k/u2pULIAl8Ov0fHj6Zw+H2rV86V6dGleleiYGN7/fBFrth5Ns3oCHJ7Xk8hrN4iOMdyOjuHJl79m6BtNeLpmSW7eiubvc5fpPvRnwq9cp3KpAnz5QSsABBgyeR2LNyXdfwvny8G0gc9b9Txyjm6fzouv5ydtqFAiv1XPfrM5fSHMqmen2nRpVtmq5+ilrNl2PE3rCf/eY7KAnzffD3gBv5zZMRgmL9jCV3N+o+//NaJby2oEh9l1+3oFK38/TOXSBfnyozYAiAhDvlvF4g0Hkiy3sH/OhPtr/1nx23HAC/H7a99pDvtrfbq0iN1fF7JmS9rur/dDA1XOaaTqIdagaXMGjPwqQdq8GT9QrlJVJs5cTLlKVZk34wcAnqhYjXGT5zBu0hze+mAA40d+mmC+6ZO+pky5inHvo2/f5rvxIxkyZiLjf/iJIo8UZ9mCOUnKsHPrbwSePc23MxbxRs+PmfCFdfKMjAhn1pSJfP7NNL74djqzpkzkSmQEAF9/MZQevT7h2xmLCDx7mp1bN9/3ZzFtyRZavvFVkvTx09cT0H44Ae2HOz15u7kJY/q0o2WPr6nw3GDaNqlEyUfyAjDk7ZaMn7Gesi0HEhoZRZdnqwPQpVV1QiOjKNtyIONnrGfI2y3vu/zJ12srLXtMSJDWs2tDNmw7yuOtBrNh21F6dn0KgNDwa7z/2c+Mmbb2jssc8lZLxs/YQNmWgwiNuEaXVrH1CiA04hplWw5i/IwNDHnbagSXLJqXto0rUrHNMFr0mMDYPu1wc7u/RrCjGzdv0+SNb6nWcTTVOo6mUUAJqpYtxOyVu3mi3Ugqv/g5npky0LVlNQB6d6nP3qOBVO34BS8PnM2o95x//kN6PMP42Zso22aEtf1aVLXq2aKqtf3ajGD87E0MeeNpu55+tH2qPBVfGEWLt79nbO/WaVrPWE3enERAly958uWvAVi7/TiVOo2j6kvjOXbmEr061QHg4F8Xqfny1wR0+ZKW709lfO+WuLsnPSUPea0x4+dspuzzXxAaeZ0uzSpZ9WxWmdDI65R9/gvGz9nMkNcbW/Us4kvbBuWo2HEsLd6bytieLVxSz3/rMXk7OoY+Y5dQsf1I6nQbzytta1KyaB6rbrM2EdBxNAEdR7Pyd+vC4+CJC9R8aSwBHUfT8q3vGN+njfPt2OMZxs/aRNnnhlt1axm7v1az6vbccMbP2sSQHs8AULJoHto2Kk/F9iNdur+qtPWvaFSJSEcR2SYie0TkWxFxF5EmIrJLRPaKyFo7X04RWSgi+0Rki4iUs9MHiMhkEdkgIn+JyFsOy35PRA7Yr3fstCIiclhEpojIURGZISINRWSziBwTkaoi4mb/7WvP4yYix2Pfp0TZJyqRPbt3grStmzfQoElzABo0ac6W39YD4JklS1w06EZUFEL8wXf8yCHCQkOoUKV6XJrBYIzhxvUojDFcu3aFnLmSFm3Lbxup37gZIkLJMuW4eiWSyyHB7Nr2O+UrB5Ddy5ts2b0oXzmAnVs3czkkmGvXrlKyTDlEhPqNm7Hltw0prXKyNu86weXwa6mer0rZIpw4c4mT50K4dTuauSt30axuOQDqVHmM+Wt2AzBjyVaa130CgGZ1yzFjyVYA5q/ZTd2qJe67/MlxVq9mdR5n+tJtAExfuo3mdR8HIDj0CjsPnebW7Zg7LrNOleLMX7sHgBlLt9G8njV/s7qPM8Ne7vy1e6hb5bG49Lkrd3Hz1m1OBV7mxNlgqpQtnGZ1BLgadROADB7ueHi4YYyJ+1IC2HHwDPn9rH29ZNE8bNxpRVaOngqmcL6c+OXMlrSelYsxf91+q57LdtK8ThmrPrXLMGPZTque6/ZTt0rxuPS5q/dw81Y0p86HcuLsJaqULpSm9XRm7bbjREdb22zbwTPk9/MCIOrGrbj0TBk9kn3uT51KjzB/w0EAZizfRfPapQFoVqsUM5bvAmD+hoPUrfRoXPrctfsc6nmZKqUKpHm9/q3H5IWQSPYcOQfAlWs3OPz3Rfx9vZLNn2A7ZsqASWZDWvvrPgBmLNtB8zplAWhWpwwzlu0AYP66fQn311X2/hp4mRNnQ6hSxvX7a0oZ49pXepXuG1UiUgp4HqhpjCkPRAMdge+A54wxTwBt7ewDgd3GmHLAR8CPDosqCTQGqgL9RSSDiFQCugLVgADg/0Qktk+tGPC5PV9J4EXgSaAn8JExJgaYDnSw8zcE9hpjgu+nvmGhIXGNnxw5cxMWGhI37Y9N63i107MM7PMWb3/QH4CYmBgmff0F3V57L8FyPDwy8Pp7H9Gjazteat2IMyf/4qlnWiVZX8ilIHL75Y17n8s3DyHBQYRcCsbXL09cem5fP0IuBRMSHERuXz+H9DyEXAq6nyrf0avta7Ntzod8078DPtk9k0z39/Pm7MXQuPfnLoaS39ebXD5ZCY+MijsZnrsYir/9pe7v583ZC9Y80dExRFyJIpdPVpfVITG/XNm5cMmK+l24FIFfruwpnjeXT1bCrzjWKwx/X7tevt6ctbuHrHpdJ5dPVvIn+YzC8Pf1SZvK2NzchC3T3uX0L/1Zt+0Y2w+eiZvm4e7GC00rsnrLEQD2HwukZV3rC6dy6YIUyusT1+CKq6d3loTbLyhRPYMS1dM7C/l9vTl7MTy+nkHh+Psl/2V5L4wxLBndlc2TXqdbiypJpnd+phIr/4jvwqlSugA7p7/Fjh/f5K2Ri+Lqk6CeV67H1zM4Iu4L3t/Xi7NB4fH1vHqHet6hUZDW/k3HZKF8OShfIj/bD54G4NW2Ndk24z2++bhdgrpVKVOInbN7smPm+7w14mfn2zHyDsflxbCEdYvbjmFxy3Dcx9XDK903qoAGQCVgu4jssd+/BWwyxvwNYIyJ7Xx/Ephmp60DcolI7NlmmTHmhjHmEhAE5LHzLzDGXDXGXAHmA7Xs/H8bY/bbjaeDwFpjXaLsB4rYeSYDne2/uwE/OKuAiHQXkR0ismPixIkprrgVmYqPSFWvXZ9vpi2g75AvmD7Z6npYvvAnKld7ktwODSCA27dvsXzRPMZ+P4up81dR5NHHmDdjcorX/TD4bu6vlG4+gGrth3PhUgTD32v9oIvkEun5qi1WTIwhoNNoijUfTOUyBSn9SPz+OLZ3azbv+ZvNe/4GYNSP6/HO7smWae/yWrua7D0aSHR0+vgQGrz2HTW6fUWr96fySutq1HyiSNy03p3rEh0dw+xVe+PSth86S6WO43jyfxPo1akOmTKm72Gu/6ZjMqtnRmYNf4leXywi8uoNvvv5d0q3Hka1jqO5EBLB8Lebx+XdfvA0ldqP4skuY+n1Uv10vx1Twrj4X3r1b2hUCTDVGFPefpUABtzDcm44/B3N3QfxO+aPcXgfEzuvMeYMcFFE6mNFwFY4W5AxZqIxprIxpnL37t3vuFKfHLm4HGIFuy6HBOOTI2eSPGWfqMSFwHOEh4Vy+OA+li6Yw8vPP83kCaNZt3IpU74dy1/HrKvlfPkLIiI8We8p/jywN8mycuX241LQhbj3IcEXyeXrR67cvgQHxQ8evhQcRK7cvuTy9eNScJBD+kVy5fbDFYIuRxITY3VjTp6/mcpOuqwCg8IpkCd+8H3+PDk4FxxOSNhVvLN7xo19yJ8nB4H2VX9gUDgF8lrzuLu74ZXNk5Cwqy6pgzNBIZHkzW219fPm9iL4cmSK5w0Ju4p3Nsd6+RAYbNcrOJwCeX2A2HplJiTsKueSfEY+BAaHpU1lEgm/cp2NO0/QqHpJAD56+Sl8c2Sl95glcXkir97glUE/EdBpNC8PmE1un6z8HRiSYDkh4dcSbj+/RPX0S1TP8GucCw6nQJ74K/38ft4EBkWkaf0C7QhjcNhVFm86RJXSVrdbx6cr8HTNEnQZ+JPT+Y6cCuZK1A3KPJLw4ick/Bre2TLH19PXi8DgCLueERSwIznu7m54Zb1DPYPTtp7J+bcckx7ubswa8RJzVu5ikT3oPOjylfi6LdxKZSddcUdOBnEl6iZlHs2bID3J/pr4uMzjk7BucdvRJ24Zjvu4enj9GxpVa4E2IuIH1rgpYB9QW0SKOqQB/IrdHScidYFLxpg7nW1+BVqJSBYRyQo8a6elxvdY3YBzjTHRqZw3iao167D2F+sLaO0vS6hWsy4AgWdPx/XlHz/6J7du3cTL24eenwzlh7krmDRnOd1ee5f6jZvR5ZW3yeXry5mTfxEeZgXx9uzYQsHCRZOsr1rNOqxbuRRjDIcP7iNL1mzkzOVLxao12L39D65ERnAlMoLd2/+gYtUa5MzlS5YsWTl8cB/GGNatXErAk3Xut9pOxTY8AFrWf4JDJ84nybPj4CmKFfKlsH8uMni407ZxRZZtsMY1bNpxlNYNrd7cDs2rsdROX7ZxPx2aW4OmWzeswMbt/+wdN8s2HaBjM2sQa8dmVVm6cX+q5t+04xitG5QHoEOzqizdYM2/bOMBOtjLbd2gPBu3H7PT99O2cUUyZvCgsH9OihX0ZfuBU2lUG8jtkxXvbJkByJzJgwZVi3PkZBBdWlTlqYDH6PzJjATjULyzZSaDhzsAXVtW5bc9fxN59UaS5W7aeZzW9a3xYh2eqcTSTda4o2W/HqLDM9Zg7tb1H2fjDmt81rJNh2j7VHkyZnCncL4cFCuYm+2HTqdZPbNkzkC2LBnj/m5YtRgH/7rIU9WK896LtWnzwTSibtyKy184X464L9lCeXwoUdiXU+dDkyx3066/aF3XGi/W4emKLP31T6s+v/1Jh6etm09a1y3Dxp1/2emHadugXHw9C+Ri+59n06yed/JvOSa/+aQdR/6+yLiZm+LS8jp0w7esWzauboX9c8Zvx7w5rO0YmPTORGt/tcaOdXimMks32vvrpoN0eKYyAK3rl4vfX389SNtG9v7qn9PaXw+m3f5634yLX+mUJDeoLj0RkeeBD7EaibeAN4AcwFA7LcgY85TduJoMPAJcA7obY/aJyADgijFmlL28A0AzY8xJEXkPq+sO4HtjzBgRKQIsNcaUtfNPsd/PczItAxACVDXGpOQ+dXP0gjX4c+TAPuzfs5OI8DB8cubkxa6vEvBkPUYM+IDgi+fxy5uPDwZ8RnYvb+bN/IF1K5fi4eFBxoyZ6PrauwkeqQCwZsVijh85FPdIhRWL5rJ43iw8PDzwzZOPdz4ciJe3DysWzQWgacu2GGP4Zsxwdm373XqkQp8BFC9pneBXL1vIXLvLsF3Hl2n4tHVHzrHDBxkzvD83b1iPVHjl7Q8SPFLhsbxZ8KzQIwUfRbypw7pQq1JxcvtkI+hyBIO+WU7tSsUpV6IAxhhOnb/Mm4NnceFSBPl8vfm634s8+6Z1V13jJ0szsmcb3N2EqYu28NmklQAUyW/dvp3DKyt7j5yha98fuXnrNpkyejB5cGeeKFGQ0IirdOrzAyfPhdypeE5F7f4Sz4pv3THP1KEvUatSMbtekQz6ZjlLNuxj+oiuFMybg9PnQ+n4wQ+ERlwjT67sbJ7ei+xZMxNjYrh67SYV2gwl8up1Fox7hdc/ncX5SxFWvYZ1IYd3FvYePkvXj6fF12tQJ54oWYDQ8Gt0+nBKXL16v9yIl1oEcDs6ml6j5rPq9z9TVsdd4/Cs1uuOecoWy8d3/Z7H3c0NNzfh57V7GTZpDZGbh3P6QhiR16wG06IN+xk2aQ3Vyhbmu/7PY4zhz78u8uqQuYRFRgGwYHQ3Xh8yz6qnf06mDe5ADq8s7D16jq79Z3HzVrRVzwHteeKx/IRGXKPTxzM4aX/J9e5Sn5eaV7XqOXoxq/44krJ6bh2JZ82+d8xTxD8Hc4ZaQyg9PNyYs2ofn/24gQNz3iNTBndCIqw6bDt4hrdGLuKFxuXp2ak2t27HEBNjGPbDOpbYDaYFozrz+vAFnL8USRH/HEwb2J4cXp7sPRpI10/nxtfzkzY88Zg/oRFRdOo/m5OBVqOsd+e6vNSsIrejY+g1djmrUnArftTmIak6LtPtMVm15x3z1HiiCGu/68H+Y4HE2N+P/b9eQbtGFSj3mL9dt1DeHDaPCyGRvNC0Ij1fqs+t29HWdpy0miV2g2nB6Jd5fcjc+P11SMf4/bXfzPjtOPCF+P217/T4/bVrA15qXsXajl8sZlUKH3UStW2Uy28TPHrhmksbD4/lzZIub3X8VzSqHmYiUhkYbYypddfMlrhG1b/VvTSq0qOUNKrSu5Q0qv4NUtKoSu9S26hKj1LSqPo3+CcaVUdc3KgqkU4bVf/+0XQPkIj0AV4j/g5ApZRSSv1LaaPKhYwxw4HhD7ocSimlVFrSTi7n/g0D1ZVSSimlHjiNVCmllFIqVdLzs6RcSSNVSimllFJpQCNVSimllEodDVQ5pZEqpZRSSqk0oJEqpZRSSqWKBqqc00iVUkoppVLFGNe+UkJEJotIkP0rKLFpOUVktYgcs//PYaeLiIwTkeMisk9EKjrM85Kd/5iIvOSQXklE9tvzjBPHnwZJhjaqlFJKKZUeTQGaJErrA6w1xhTH+m3gPnZ6U6C4/eoOTIC43wbuD1QDqgL9Yxtidp7/c5gv8bqS0EaVUkoppVLFuPhfispgzCYg8a9XtwSm2n9PBVo5pP9oLFsAHxHJBzQGVhtjLhtjQoHVQBN7mpcxZouxfs/vR4dlJUsbVUoppZT6t8hjjDlv/30ByGP/nR8445DvrJ12p/SzTtLvSBtVSimllEod49qXiHQXkR0Or+6pLqIVYfpHx9Tr3X9KKaWUeqgYYyYCE+9h1osiks8Yc97uwguy088BBR3yFbDTzgF1E6VvsNMLOMl/RxqpUkoppVSquDhQdT8WA7F38L0ELHJI72zfBRgAhNvdhCuBRiKSwx6g3ghYaU+LEJEA+66/zg7LSpZGqpRSSimV7ojILKwoU24ROYt1F99w4CcReRk4BbSzsy8HngaOA9eArgDGmMsiMgjYbuf71BgTO/j9daw7DD2BFfbrjrRRpZRSSqlUSemzpFxbBvNCMpMaOMlrgDeSWc5kYLKT9B1A2dSUSbv/lFJKKaXSgEaqlFJKKZUqKX2W1H+NRqqUUkoppdKARqqUUkoplToaqHJKI1VKKaWUUmlAI1VKKaWUShUNVDmnkSqllFJKqTQg5mF42IRypBtEKaXU/RBXr2D3qUiXfldVKJzd5XVwBe3+ewh5VujxoIvgUlG7v+T67QddCtfL7PHf2JaeFd960MVwuahd4/719YzaNQ7Pqj0fdDFcKmrbqH/9MQnWcakeDG1UKaWUUipV9DlVzumYKqWUUkqpNKCRKqWUUkqljgaqnNJGlVJKKaVSRdtUzmn3n1JKKaVUGtBIlVJKKaVSRZ/G5JxGqpRSSiml0oBGqpRSSimVKvpIBec0UqWUUkoplQY0UqWUUkqp1NFAlVMaqVJKKaWUSgMaqVJKKaVUqmigyjmNVCmllFJKpQGNVCmllFIqVfQ5Vc5ppEoppZRSKg1opEoppZRSqaLPqXJOI1VKKaWUUmlAI1VKKaWUSh0NVDmlkSqllFJKqTSgkSqllFJKpYoGqpzTSJVSSimlVBrQSJVSSimlUiVGH1TllDaq0piI1AVuGmN+/6fW+U3/DjStXZbgy5FUbjsUgKHvtOLp2mW5eSuav89eonv/6YRfiUoy71M1SjGqVxvc3dyYsvB3Rv2wGoDC/rmYNrwrOb2zsvvP03T7+Edu3Y4mYwYPJg3qRIVShbgcfpWOH0zm9PnL91z2fh9/yKaNG8iZMxfzFy0F4ItRI9i4YT0ZMmSgQMFCfDp4GF5eXvzx+2bGjv6cW7dukSFDBt59vxfVAqoTFRVFr/fe5syZ07i5uVOnbj3eea8nACOHD2X7tq0ARF2/TujlEH7bsiNJOQ4dPMAnfT/kxvXrPFm7Dh982BcRITwsjN493yXw3Dn88+dn5Odj8PL2xhjDiGFD+G3TRjJ7ZmbQkOGUKl3mnj8HSN/bMTXe7FCXLq2qY4zh4PHzdB8wg2/6vUjF0gW5dTuaHQdP02PIbG7fjkkyb4dmVenzv0YADP9+FTOWbgOgQqmCTBzQAc/MGVj52yHeH/kzADm8sjBteBcK++fkVOBlOn7wA2GRST8/V/DO5smEfi9Q+tF8GAyvDpxJ1PVbjO/7PJkyenA7OoZ3hv3EjoOn0009C/h58/2AF/DLmR2DYfKCLXw15zfKFfdnfJ/nyJTJrteI+ew4dIbHCvsysd/zlC9RgAETVjBmxkanyy3sn5NpgzuS0zsLuw+fpVv/WfZ+6s6kAS9QoWQBLodfo2PfaZw+HwpAz5fq06VFVaJjYnj/84Ws2XI0zesLzo/L1g0r0PfVpylZNA+1Oo1i16Gk2xDS13Gp0oZ2/6UhEfEA6gI1/sn1TluyhZZvfJUgbe2Ww1RqO5Sqzw/j2KkgenVrlGQ+NzdhTJ92tOzxNRWeG0zbJpUo+UheAIa83ZLxM9ZTtuVAQiOj6PJsdQC6tKpOaGQUZVsOZPyM9Qx5u+V9lb1lq9ZM+Pb7BGkB1Wvy88KlzFuwhMKFizDpu28B8MmRg3FfTeDnhUsYNHQ4fT/sHTdP5y7dWLT0F36at4A9u3fx26/WybtXn4/4af4ifpq/iBc6dKR+w6eclmPwpwPoP3AQS1as4vSpk2z+bRMAk7+fSNVq1VmyYhVVq1Vn0vcTAfjt102cPnWSJStW0W/AIAZ/OuC+PgdI39sxpfx9vXm9fR1qdhxF5XbDcXdzo23jisxesYMnWg+hcrvheGbKQNdWSQ+hHF5Z6Nu9CbU7f0GtTp/Tt3sTfLJ7AjDuw3a8MXg2ZVsO4tFCvjSqUQqAnl0bsmHbUR5vNZgN247Ss6vz7e8Ko3q1ZtXvf1L+uSFUfX4Eh/+6yJC3WzLk2xUEvPAZgyYsd/q5P8z1vB0dQ5+xS6jYfiR1uo3nlbY1KVk0D0PefIYh368moONoBn27kiFvNgMgNCKK90ctYsyMDXdc7pAezzB+1ibKPjfc2k9bVgWgS4tq1n763HDGz9rEkB7PAFCyaB7aNipPxfYjafH294zt3Ro3N3FJnZ0dlwdPBNL+/e/4bdeJZOdLT8flvTAufqVX/+lGlYhkFZFlIrJXRA6IyPMiclJEPhOR/SKyTUSK2XmLiMg6EdknImtFpJCdPkVEvhGRrcBPwKvAuyKyR0RqiUhbe9l7RWSTK+qxedcJLodfS5C2dsthoqOtK/1t+/8mfx6fJPNVKVuEE2cucfJcCLduRzN35S6a1S0HQJ0qjzF/zW4AZizZSvO6TwDQrG45ZiyxIj/z1+ymbtUS91X2SpWr4OXtnSCtRs0n8fCwgqjlnihP0MULAJQqVRo/vzwAFCtWnBvXb3Dz5k08PT2pWi0AgAwZM1KqdGkuXriYZF2/LF9G06ebJUkPDg7i6tUrlHuiPCJC8xatWLd2LQDr16+lRatWALRo1Yr169ZY6evW0rxFK0SEck+UJzIyguDgoPv6LNLzdkwND3c3PDNlwN3dDU/PDJwPjmDl5kNx03ccPEX+PN5J5nuqeknWbj1CaMQ1wiKjWLv1CI1qlCJvbi+yZ83Mtv0nAZi5dBvN61n1b1bncabbUZ7pS7fRvO7jrq8g4JUtM09WLMaUhX8AcOt2NOFXojAYvLJlBsA7W2bOB4cnmfdhrueFkEj2HDkHwJVrNzj890X8fb0wgFfWTPH1umTVKzj0Cjv/PMMtJ1FHR3UqF2P+un0AzFi2g+Z1ygLQrE4ZZiyzIsvz1+2jbpXiVnrtMsxdtYebt6I5FXiZE2dDqFKmUJrXF5wfl0f+vsixU3c+3tPbcZlaxrj2lV79pxtVQBMg0BjzhDGmLPCLnR5ujHkc+BIYY6eNB6YaY8oBM4BxDsspANQwxrQGvgFGG2PKG2N+BfoBjY0xTwAtXF4jJzq3rJ7gSyuWv583Zy+Gxr0/dzGU/L7e5PLJSnhkVNyX+bmLofj7ecfPc8GaJzo6hogrUeTyyeqysi+c/zM1a9VOkr5m1UpKlS5NxowZE6RHRESwccN6qgVUT5AeGHiOc2fPxjW+HAVdvEiePHnj3ufJm5egIKtRdjkkBF9fPwBy5/blckiINU/QRfLkdZgnT16CLiZtyKWl9LwdYwUGhzNm2jqOLh/I36sGExF5nbVbDsdN9/Bw44Wnq7D69z+TzOvv5xNXZoBzF8Pw9/PB39ebc0Fh8elBYXH19MuVnQuXIgC4cCkCv1zZXVSzhIr45+JS6BUmDujAHzN78/UnL5Alc0Z6jZrP0Ldbcmz5QIa924p+Xy5JMm96qWehfDkoXyI/2w+eptcXixj6VjOOLfmYYW81p99XK1K8nFzeWRLtp2H4+9r7qa83Zy+GAQ77qXcW8jukg/1Z+CZtiD9I6em4VGnnv96o2g88JSIjRKSWMSb2snGWw/+x387VgZn239OAJx2WM9cYE53MOjYDU0Tk/wB3ZxlEpLuI7BCRHRMnTrzXujjV++XGREfHMHv59jRd7j/hu28n4O7hzjPNErZFjx8/xpjRo/ik/6cJ0m/fvk2fXu/xYodOFChYMMG0X5Yvo2Gjxri7O90EKSIiIK7pYrib9LwdHflk96RZ3ccp1WwgjzT+mKyeGWn/dOW46WP7tGPz7hNs3v2XS9b/T10Be7i7Ub5kAb6b9xvVX/yMa1E36Nm1Id3bPEnvzxdQ/On+9P58ARP6veiS9bu6nlk9MzJr+Ev0+mIRkVdv0P256vQevZjizQfTe8xiJnzc1rUFUA+ccfG/9Oo/3agyxhwFKmI1rgaLSL/YSY7ZUrCoq3dYx6vAx0BBYKeI5HKSZ6IxprIxpnL37t1TXP676di8Gk/XLkuXvlOcTg8MCqdAnhxx7/PnycG54HBCwq7ind0Td3e3uPTAoPD4efJa87i7u+GVzZOQsGSrf88WLZjPpo0bGDZilNWYsV28cIF33+rB4KEjKFgoYbj/0wGfUKhwETp27pJkeb+sWE7Tp59xui6/PHm4aHcxxq4jtpsxZ65ccd16wcFB5MyZ05rHLw8XLzjMc/ECfnny3Ftl7yI9b8fE6lcrwclzIVwKu8Lt2zEsXLeXgHJFAfioexN8c2Sj9+cLnM4bGBQWV2aA/Hl8CAwKIzA4nPx+PvHpfj5x9QwKiSRvbi8A8ub2IvhypItqltC5oDDOBYWx/cApABas3UP5kgXp0KwqC9ftBeDn1bupXKZwknkf9np6uLsxa8RLzFm5i0UbDgDQ4ZnKLFy/H4Cf1+ylcumUd8WFhF9LtJ/6EGh3iwYGh1PA7vKO20/Dr3HOIR3sz8JJV+qDlJ6OS5V2/tONKhHxB64ZY6YDI7EaWADPO/z/h/3370B7++8OwK/JLDYSiIu9i8ijxpitxph+QDBW48rlnqpRive6NKTNO98Sdf2W0zw7Dp6iWCFfCvvnIoOHO20bV2TZBmtcw6YdR2ndsAIAHZpXY6mdvmzjfjo0rwZYd8Bs3J72d9xs/nUTUyZ/z9gvJ+Dp6RmXHhERQY/XuvP2u+9ToWKlBPN8OXY0VyKv0LvPR0mW9/dfJ4iMiOCJ8hWcrs/X14+sWbOxb+8ejDEsWbyQevUbAFC3Xn0WL1wIwOKFC6lXLz59yeKFGGPYt3cP2bJlj+smTEvpeTs6c+ZCKFUfL4Jn5gwA1Kv6GEf+vkiXVtV5qnopOn80FZNMmGX1H4dpGFASn+ye+GT3pGFASVb/cZgLlyKIvHqdqo8XAeDFZlVZusH6gl+26QAdm1mDnjs2q8rSjftdX0ngYkgkZy+GUbywtU/UrVqCw39f4PylcGpVKmanPcbxM8FJ5n3Y6/nNJ+048vdFxs2MHyJ6PjiCWhUftepVpRjHz1xK1TI37TxO6/rWeKMOz1Rm6caDACzbdJAOz1iRzNb1y7Fxx3Er/deDtG1UnowZ3Cnsn5NiBXOz3cldlA9Sejou70WMce0rvZLkTmD/BSLSGKsxFQPcAl4D5gFzgKbADeAFY8xxESkM/ADkxmocdTXGnBaRKcBSY8w8e5mP2cuIAd4E3gWKAwKsBd4xd/7QjWeFHqmqx9RhXahVqTi5fbIRdDmCQd8sp1fXRmTK6EFIuHWVs23/Sd4aMpt8vt583e9Fnn1zAgCNnyzNyJ5tcHcTpi7awmeTVgJQJL91y28Or6zsPXKGrn1/5Oat22TK6MHkwZ15okRBQiOu0qnPD5w8F5Kq8kbt/pLrt62/P+j5Hju2byMsLJScuXLx2htvMvm7idy8dRMfbx8AHn/iCT7p/ykTv/maSd9PpHCh+Kv7Cd9N5vatWzRqUIeijzxCxgzWGKv2L3akdRurC2LCV+O5ceNG3GMWYrVr3ZKf5i8C4OCB/dYjFW5cp+aTtfmw7yeICGFhofR67x0unD9PPn9/Rn4+Bm8fH4wxDBv8KZs3/0rmzJ58OngoZcomHByc2QNSsy3T23YEa1t6VnwrVfN8/GpT2jxVkdvR0ew9co7XPp1FyOaRnD4fSuS16wAsWrePYd/9QsVSBflfmyd5fZDVI9+5ZQC9u1l3to2YtIppi61BvRVLFWTiwA54ZsrIqt8P8e6IeQDk9M7C9BFdKZg3B6fPh9Lxgx8IjbjmpFR3qeeucamuZ7nH8vN1vxfImMGdk2dD6D5gBqUfzcfIXs/h4e7GjRu3eHv4XHb/eeahqGfUrnF4Vu15xzw1nijC2u96sP9YYNxzivp/vYLIq9cZ+V4rPDzcuHHjNm9/9jO7D58jT67sbJ7yNtmzZibGGK5eu0GF9iOJvHqDBaNf5vUhczl/KYIi/jmZNqQjObyysPfoObr2m8nNW9HWfjrwBZ54LD+hEdfo1Hc6JwOtRwz07tqAl5pX4XZ0DL2+WMyqPw7fqehWHbeNStUxCc6Py9Dwq3zxQVty58hGWGQU+46co8UbXz1Mx6XLxylsOHLZpY2HuiVyPpixFvfpP92ockZETgKVjTGpu9RKO6luVKU3jo2qf7PUNqrSo3tpVKVH99KoSm9S0qhK7+6lUZUe/RONqvVHQlzaeKhXIle6bFT9p7v/lFJKKaXSij5RPRFjTJEHXQallFLqYaadXM5ppEoppZRSKg1opEoppZRSqZKenyXlShqpUkoppZRKAxqpUkoppVSqpOdnSbmSRqqUUkoppdKARqqUUkoplSo6pso5jVQppZRSSqUBjVQppZRSKlX0OVXOaaRKKaWUUioNaKRKKaWUUqmikSrnNFKllFJKKZUGNFKllFJKqVSJ0bv/nNJGlVJKKaVSRbv/nNPuP6WUUkqpNKCRKqWUUkqlij780zmNVCmllFJKpQGNVCmllFIqVXRMlXMaqVJKKaWUSgMaqVJKKaVUqugjFZzTSJVSSimlVBrQSJVSSimlUkXHVDmnkSqllFJKqTQgRpubDxvdIEoppe6HuHoFC/ZdcOl31bPl8rq8Dq6g3X8PIc9Kbz/oIrhU1M6xeFbo8aCL4XJRu7/kXNjNB10Ml8rvkxHPGh896GK4XNTvQ/8bx2W1Xg+6GC4VtXUknpXffdDFcLmoHaMfdBH+s7RRpZRSSqlU0V4u53RMlVJKKaVUGtBIlVJKKaVSJeZBF+AhpZEqpZRSSqk0oJEqpZRSSqWKjqlyTiNVSimllEp3RORdETkoIgdEZJaIZBaRoiKyVUSOi8gcEclo581kvz9uTy/isJwP7fQjItL4fsqkjSqllFJKpYpx8etuRCQ/8BZQ2RhTFnAH2gMjgNHGmGJAKPCyPcvLQKidPtrOh4iUtucrAzQBvhYR93v5TEAbVUoppZRKnzwATxHxALIA54H6wDx7+lSglf13S/s99vQGIiJ2+mxjzA1jzN/AcaDqvRZIG1VKKaWUShVjjEtfKVj/OWAUcBqrMRUO7ATCjDG37Wxngfz23/mBM/a8t+38uRzTncyTatqoUkoppdRDRUS6i8gOh1f3RNNzYEWZigL+QFas7rsHSu/+U0oppVSquPo5VcaYicDEO2RpCPxtjAkGEJH5QE3AR0Q87GhUAeCcnf8cUBA4a3cXegMhDumxHOdJNY1UKaWUUipVHnT3H1a3X4CIZLHHRjUADgHrgTZ2npeARfbfi+332NPXGWtFi4H29t2BRYHiwLZ7/Vw0UqWUUkqpdMUYs1VE5gG7gNvAbqzI1jJgtogMttMm2bNMAqaJyHHgMtYdfxhjDorIT1gNstvAG8aY6HstlzaqlFJKKZUqD8OzP40x/YH+iZL/wsnde8aY60DbZJYzBBiSFmXS7j+llFJKqTSgkSqllFJKpcpDEKh6KGmkSimllFIqDWikSimllFKpEvMwDKp6CGmkSimllFIqDWikSimllFKponEq5zRSpZRSSimVBjRSpZRSSqlUSeFTz/9zHlijSkR+N8bUuEued4CJxphrLi5LecDfGLPcft8CKG2MGZ7G67lijMl2v8v5pt8LNK1VhuDLV6j8vFXEHF5ZmDasC4X9c3Iq8DId+/xAWGQU73aqz/NNKwHg4e5OyaJ5KNiwL6ERCT/Swv45mTbsJXJ6Z2X3n2fo9sl0bt2OJmMGdyZ92pEKpQpyOfwqHftM5fT5ywD07NqQLi0DiI6O4f1R81nzx+H7rVryde7fgaa1yxJ8OZLKbYcCUO6x/Izv255MmTJwOzqGd4bOYcfBU0nm7dC8Gn3+1xiA4d+vZMaSrQBUKFWQiQM74ZkpAys3H+T9z+YB9mc5olv8Z9l7EmGRUfdV/s8GfcKWzZvwyZGTybMWALBh7UqmfjeB0yf/4usfZlGiVJm4/DOnfM/yJfNxc3Pnzff7UCWgJqdP/c2gvr3i8pw/d5Yu3d+gzQudOHH0CKNHfEpU1DXy5MtP34HDyZot6a627Y/f+PKLEcTERPN0i9a8+NL/rGUFnmXQx72JCA/jsZKl+XDAMDJkyMDNmzcZPvAjjh4+hJe3D/0GjySv/739gLubm7B58hsEBkfwXK8f49I/f7cZnZ+phG/DgQD8r1VVXnnO2q+uRt3kjRELOXwyKMnynqpWnFHvNMPd3Y0pS7YzatomAArny8G0T9uT0zsLuw+fo9unc+P35U/aUqFkfi6HX6PjJ7M4fSHsnurijLPjst9rT9OszuPExMQQHHqF7v1ncP5SBF7ZMjN5UCcK5s2Bh7sbY6atZ5q9XzqqULIAEwd2sPfRQ7w/cj6Q/PEO8Hmv1jSuWZpr12/RfcAM9hw+m2Z1zJTRgzXfvEbGjB54uLuxYN1+Bn+3ilfb1KBH+1o8WjA3BRr1JyTcOr+0b1yB9zrVQwSuXLvBW5/NZ/+x80mWWzhfDqYN7mhvs7N0GzA7fpv1b0+FkgWsbfbxdE6fDwWg50v16NK8KtExMbz/+SLWbD2aZvX8pl97mj5ZmuDQK1R+/jMA+nZvTLdWAQSHXgWg/9fLWLn5TyqXKcSXH7UDQASGTFzJ4g37k9bRPyfThna26vjnWbr1mxFfx4EdqFDKruOHU+Pr2KUBXVpWIzrG8P7I+azZciTN6qhc44F1/92tQWV7B8iSmuWKiPs9FKc88HTsG2PM4rRuUKWlaUu20fLNbxKk9ezSkA3bj/L4s4PZsP0oPbs0BGD0tHUEvDiSgBdH0u/LJfy663iSBhXAkLdaMH7GBsq2GkxoRBRdWgUA0KVVdUIjoijbajDjZ2xgyFvNAShZNA9tG1WkYtthtHjzG8b2aYubm7iwzlto+cZXCcv8TiuGTFxBQPvhDJqwlCHvtEoyXw6vLPTt3pTanUZRq+NI+nZvik92TwDGffQ8bwyaSdmWA3m0kC+NapYGoGfXp9iw7QiPt/yUDduO0LNro/suf+NmLRk+ZkKCtKKPFGfgiNGUq1ApQfrJv06wbvUKJs9ayIixExjz2WCio6MpVLgo302fx3fT5/HN1DlkypyZJ+s2AGDU0P783xvvMGnmAmrVacCc6T8kKUN0dDRjRw5h+Jiv+WH2ItatWsHJv04AMPHL0bRp34npPy8ne3Yvli+2vrxXLJ5P9uxeTP95OW3ad2LiV6Pv+TPo0a4GR04GJ0irWDJ/3PaINWfVXqp0GkdAly/5YsYmRrz1NIm5uQljerag5ftTqPDiGNo2fIKSRfwAGPJ6E8bP2UzZdp8TGhlFl+aVAejSvDKhkVGUbfc54+dsZsjrafuD9s6Oy9E/rqVq+xEEvDiSFb8e5MP/s9b5SttaHP7rAtVe+IzG3ccz/N2WZPBIeuoa92E73hg0m7KtBvNoQV8a1SgFJH+8N65ZmkcL+lK21WB6DJ7NuA+dPkD6nt24eZsmb3xLtY6jqdZxNI0CSlC1bCH+2HeSp9+cyKnAywnynwy8TKPXJlClwxcMm7yGr/q0cbrcIT2eYfzsTZRtM8LaZi2sB2J3aVHV2mZtRjB+9iaGvGHtCyWL+tH2qfJUfGEULd7+nrG9W6fp+cfalkl/y3f8zI0EdBhFQIdRrNz8JwAHj5+nZucvCOgwipZvTmT8R21xd0/61TrkzeaMn7mRss8OterYsppVx5YBVh2fHcr4mRsZ8qbjObYCFduNoMWb3zK2TxuXnmNTK8bFr/TqgTWqROSK/X9dEdkgIvNE5LCIzBDLW4A/sF5E1tt5G4nIHyKyS0Tmikg2O/2kiIwQkV1AW/v9QDvffhEpaeeras+/W0R+F5ESIpIR+BR4XkT2iMjzItJFRL605ykiIutEZJ+IrBWRQnb6FBEZZy/nLxFpY6dns/PFrrtlWn92m3ef4HJ4woZRszplmb7U+g3I6Uu30bzu40nma9ekEj+t3OV0mXWqFGf+2r0AzHCYv1mdssywlzt/7V7qVn3MSq/7OHNX7eLmrWhOBV7mxJlgqpQpnDYVdGLzrqR1Nga8smYGwDubJ+eDw5PM91SNUqzdcpjQiGuERUaxdsthGtUsTd7cXmTPmplt+08CMHPpNprXLWfXrRzT7ajB9CVbaV6v3H2X/4kKlfHy8k6QVrjoIxQqXDRJ3t83raf+U03JmDEj+fwLkL9AIQ4fSnjlu2v7VvwLFCRvPn8Azp4+RbkKVuOhUrXq/Lp+TZLlHj60n/wFCuGfvyAZMmSg/lNN+X3Teowx7N6xjTr1nwKg0TMt2LxxHQCbN62n0TMtAKhT/yl2bd96T2H//L5eNKlRkh+WbI9Lc3MThr7RlL5f/ZIgb+S1G3F/Z/XM6HR9VUoX4MTZEE4GhnLrdjRz1+yjWS2rwVGn0iPMX38AgBkrdtG8tpXerFYpZqyw9v/56w9Qt/Kjqa7HnTg7LiOvxtcli2dGjD2812DIZu+7WbNkIjTiGrejE36V5M3tRfZsmdl2wIq+zly2PcFx6ex4b1anLDOXWZ/xtgOn8M7mSd7cXmlaz6tRNwHI4OGOh4cbxhj2Hg2Mi6442rL/VFwEbduB0+T3806SB6BO5WLMX2ft4zOW7aR5HStq26x2GWYs2wnA/HX7qVuleFz63NV7rPPP+VBOnL1EldKF0qyOm3f/xeWIqynKG3XjFtH2tsuUySPZn2+pU6VYCs+xxePS567a7XCOvUSVMmlXR+UaD8tA9QpYUanSwCNATWPMOCAQqGeMqSciuYGPgYbGmIrADuA9h2WEGGMqGmNm2+8v2fkmAD3ttMNALWNMBaAfMNQYc9P+e44xprwxZk6iso0HphpjygEzgHEO0/IBTwLNgNjI1nXgWXvd9YDP7V/Qdim/XNm5cCkCgAuXIvDLlT3BdM/MGXiqekkW2ge1o1w+WQmPjIo7MZwLCsPf1wcAf18fzl60TpbR0TFEXLlOLp+s5Pf15qxD18m5i+H4J3PCdJVeo+Yx9J1WHFsxiGHvPku/8YuS5HEsP8TXzd/Ph3NBYfHpF8Pw9/MB7v5Zulpw8EV88+SJe+/rl4dLQQm7v9avXkH9Rk3j3hd+5FE2b7IaQhvXriQo6EKS5V4KCsIvT96497n98hAcfJGI8DCyZc+Ou4eHvb68XAq21ncpOAg/P2sedw8PsmbLRkR4WKrrNPKdZvT9agUxMfHfOK+1qc6y3/7kQkhkkvyvtA7g4Nz3GfJ6E94fvTTJdH9fb85ejG9EnwsOJ7+vF7m8sxB+5brDvhyBv693knmio2OIuHqdXN6pCoTfkwGvP8OxZQNo36QygyYsB+CbOb9Ssmge/lr5KTvm9KHnqPlJGo/+vt6cuxgWX8cU7KP+fj6cdZwnKDyu/mnFzU3YMu1dTv/Sn3XbjrH94JkUzdelRVVWOhkikMs7i5Pzj8M2s4/TuPOPdxbr/OO4/YPC8fdL28ajM6+2q8W2Wb34pl/7BBHWKmUKsXPOB+yY3Zu3hs2Nq0usXN6Jz7Hx50t/P++4bRZfx6zkd0i35onf/g8DY1z7Sq8elkbVNmPMWWNMDLAHKOIkTwBWo2uziOwBXgIcQyOJG0Pz7f93OizPG5grIgeA0UAZ7q46MNP+expWIyrWQmNMjDHmEBD7LSjAUBHZB6wB8jtM+8ck3imfqVWWP/b+7bTrL73q3rYWvT+fT/Gmn9B71M9M6N/BJet52A7wW7du8fuvG6hTP75bsvfHn7Jo3hxe6dyOa9eukcEjw4MrYCJNa5QgKPQKu48ExqXly52d1vXK8vW8P5zO8+38LZRp+zkff72SPl3q/VNFdYkBXy+j+DMDmP3LDl59vjYAT1Uvyb4j53ikcT+qvfAZo3u3IXvWTPe8jn9yH42JMQR0Gk2x5oOpXKYgpR+5++mtdqVHeal5FT7+cvk/UELX+G7eZkq3Gky1F0dx4VIEw9+N74TYfvA0lZ4fwZOdv6BX1wZkyqj3gP1XPSyNqhsOf0fjfAC9AKvtaFJ5Y0xpY8zLDtMTx2pjl+m4vEHAemNMWaA5kDkNyx0bjeoA+AKVjDHlgYt3W4+IdBeRHSKyY+LEpP34KREUEhkX5s+b24vgywmv/ts2rsjcZLr+QsKu4p3dM24cQH4/HwKDwwAIDA6jQJ4cALi7u+GVLTMhYVc5FxxOgbw+ccvIn8ebwKCk3W+u1KFZNRau3QPAz6t3U9lJ96Nj+SG+boFBYeR3uOrLn8eHQPuK+G6fpav5+uYh+OLFuPfBQRfJ7ecX937b779SvEQpcubKHZdWqMgjjBw/kW9//In6jZqSr0DBJMvN7edH0MX4CNaloIv4+ubBy9uHK5GRRN++ba/vArl9rfXl9vWLi3pF377N1StX8PL2SVV9qpcrTLMnS3H45178+Gl76lZ6hJ3T3+GRArk4+NP7HP65F1kyZ+DAT+8nmfenNftoXrt0kvTA4HAK5ImPwOT39eZccAQh4dfwzpbZYV/2ItDuFnacx93dDa+smeMGVP8T5qzYSav6TwDQqUU1Fq2zosZ/nb3EycAQShRJ2DgJDA4nfx6fuPcp2UcDg8Io4DiPn3dc/dNa+JXrbNx5gkbVS94xX9li+ZjwUVva9prCZScXdSHh15ycfxy2mX2cxp1/wq9Z5x/H7e/nTWBQRBrVzLmgy1eIiTEYY5i84A8qO+mKO3IyiCvXblLm0XwJ0kPCE59j48+XgUHhcdssvo5XOeeQbs0Tv/0fBjHGuPSVXj0sjarkRAKxfS9bgJoiUgxARLKKyGOpXJ43cM7+u0sy60nsd6C9/XcH4NcUrCPIGHNLROqRMJrmlDFmojGmsjGmcvfu3e+W3allmw7QsZk1uLNjs6os3XggbppXtsw8WfFRlji5IyXWph3HaN3AOuF3cJh/2cYDdLCX27rBE2zcfiwuvW2jimTM4E5h/5wUK+jLdid33rnS+eBwalWyxh/UrfoYx08HJ8mz+vc/aVi9JD7ZPfHJ7knD6iVZ/fufXLgUQeTV61R9vAgALzarytKN+wBYtnE/HZtbg0g7Nq/G0g37/pkK2arXrsu61Su4efMm5wPPcu7MKUqWjh8jt25Vwq4/gNDLIQDExMQwffJEWjzbLslyS5Yqy7kzpzgfeJZbt26xbvUKqteui4hQvlIVNq5bDcCqZYupWduKDtWoVZdVyxYDsHHdaipUrkpqe7P7fbOKYq1GUPK5kXTuN5sNO//Cv8kgijYfRsnnRlLyuZFcu36Lsu0+B+DRArni5m1aowTHz1xKsswdf56jWIHcFM6Xgwwe7rRtWI5lv1kDhzft+ovW9coC0KFpRZb+aqUv+/UwHZpWBKB1vbJs3PlXqupxLx4t6Bv3d7M6ZTl60mosn7kQGjc+0S9ndh4r7Mff50ISzHvhUgSRV65Ttax1CnnxmSrxx2Uyx/uyTQd48ZkqAFQtW5iIK9fjugnTQm6frHhns64RM2fyoEHV4hxxcmdmrIJ5fJg9vDMvD5jldDvG2rTzOK3rW/t4h2cqsXTTQQCW/XqIDs9YN3K0rv84G3cct9I3HaLtU+Wt80++HBQrmJvth06nSR2TkzdXfPdiy3rlOHTCuouxsH/OuMZSobw5KFHEL8mAfYBNO447P8duSnyOja3jQdo2qpDoHOvaOqr797DHKCcCv4hIoD2uqgswS0Ri4+QfA6m5j/YzYKqIfAwsc0hfD/SxuxWHJZrnTeAHEekFBANd77KOGcASEdmPNe4rzZ8zMHVIZ2pVLkZun2wcXz6QQd+uYNSUNUwf3pWXWgZw+vxlOvaZEpe/Rb1yrN1yhGvXbyZYzoKxr/D6oFmcvxRB33FLmDb0Jfq//gx7j5xlykKrW2bKoi1MHtSRAws/JjT8Gp0+mgrAn39d4OfVu9k97yNu347mnRHzEoyXSfM6D+tCrUrFrTr/MohB3yznjUEzGdmrDR4ebty4cZseg2cBULF0If7X5kle/3QmoRHXGPbdL/w2vTcAQyf+EtcF+vawn5g4sCOemTKwavMhVv52CIBRP6xm+ohuvNSquvVZ9p583+Uf9HFv9u7aTnhYGO2aNaBL9zfI7uXN+FFDCQ8L5aN3X+fRx0ry2bhvKfpIMeo2bEzX9i1xd/fgrV59cXe37gyLirrGzm1/8O6H/RIsf92qFSyaZw0nfLJeA5o0bwVYY6JGDenP8DETcPfw4M2eH/HBW68SHRNN0+bPUvSRYgB07/Eugz7uzeRvx1PssZI0bdEagKdbtGbogA/p+NzTZPfy5pPBn933Z3E3r7WpTr3Kj3LrdjRhkdf5v8HWoy7y5c7O131a82zPqURHx/DuF4tZMror7u7C1KU7+fNv68u979e/MO3T9vTv/hR7jwYyZckOAKYs3cHkfm058NP7hEZco1O/2cmW4V44Oy6b1CxN8cJ+xBjD6fOXeWvoTwAM/24lEwd2YPucDxCEvuOWEBJmBdu3zOxFwIsjAXh7+FwmDuiAZ2Z7H91s76PJHO+//HaIxjVLc3DRJ1y7fpNXBsxMWtD7kDe3F9/1ex53Nzfc3ISf1+5lxeY/eb1dTd7rVJc8ObOzfcZ7/PL7YV4fOo8PX25ITu8sjOlt7U+3o6N5sos1LHXB6G68PmSedf75cjnTBneg/ytN2Hv0HFMWWwO3pyzexuQB7Tkw7wNrm308A4A//77Iz2v2snt2L25HR/POyAVpev6ZOqQTtSoVI7dPVo4v68+gib9Qu1Ixyj3mjzFw6vxl3hwyF4Aa5R+h50sNuHU7mhhjeHv4PELCrW25YOz/8fqgOVYdxy9l2tBO9H+tKXuPnGPKoi1WHRdtZfKnHTiw4COrjh9Ns+r41wV+XrOH3XP7WI+M+cy159jUSsfBJJcSfYDXQ8d4Vnr7QZfBpaJ2jsWzQo8HXQyXi9r9JefCbt49YzqW3ycjnjU+etDFcLmo34fynzguq/W6e8Z0LGrrSDwrv/ugi+FyUTtGu/zmqIlbTrm08dA9oPDD8/yIVHjYI1VKKaWUesik53FPrqSNKqWUUkqlykPUE/lQedgHqiullFJKpQsaqVJKKaVUqmjvn3MaqVJKKaWUSgMaqVJKKaVUqsSgoSpnNFKllFJKKZUGNFKllFJKqVTRMVXOaaRKKaWUUioNaKRKKaWUUqmiz6lyTiNVSimllFJpQCNVSimllEoV/Zka5zRSpZRSSimVBjRSpZRSSqlU0UCVcxqpUkoppZRKAxqpUkoppVSq6N1/zmmkSimllFIqDWikSimllFKpYnRQlVMaqVJKKaWUSgMaqVJKKaVUquiYKuc0UqWUUkoplQY0UqWUUkqpVNFIlXPaqFJKKaVUqhi0VeWMdv8ppZRSSqUB0dsiHzq6QZRSSt0PcfUKBq057tLvqk8aFnN5HVxBu/8eQp71hzzoIrhU1Lq+eFZ860EXw+Wido3DM+CDB10Ml4raMoLD56896GK4XMl8WfCs/O6DLoZLRe0Y/d+o47/8mATruFQPRrKNKhGJJD5qEttiNPbfxhjj5eKyKaWUUuohpJ1cziXbqDLGZP8nC6KUUkoplZ6lqPtPRJ4EihtjfhCR3EB2Y8zfri2aUkoppR5GMRqqcuqud/+JSH/gA+BDOykjMN2VhVJKKaWUSm9SEql6FqgA7AIwxgSKiHYNKqWUUv9R+vBP51LynKqbxnruggEQkayuLZJSSimlVPqTkkjVTyLyLeAjIv8HdAO+c22xlFJKKfWw0iFVzt21UWWMGSUiTwERwGNAP2PMapeXTCmllFIqHUnpwz/3A55YXYD7XVccpZRSSj3s9O4/51Jy99//gG1Aa6ANsEVEurm6YEoppZRS6UlKIlW9gArGmBAAEckF/A5MdmXBlFJKKfVw0kCVcym5+y8EiHR4H2mnKaWUUkop251+++89+8/jwFYRWYQ1pqolsO8fKJtSSimlHkIxD7oAD6k7df/FPuDzhP2Ktch1xVFKKaWUSp/u9IPKA//JgiillFIqfdC7/5y760B1EfEFegNlgMyx6caY+i4sl1JKKaVUupKSgeozgMNAUWAgcBLY7sIyKaWUUuohZoxrX+lVSh6pkMsYM0lE3jbGbAQ2iog2qpRSSqn/KP1BZedS0qi6Zf9/XkSeAQKBnK4rklJKKaVU+pOSRtVgEfEG3gfGA17Auy4tlVJKKaUeWiY999G5UEp+UHmp/Wc4UM+1xVFKKaWUSp/u9PDP8VgP+3TKGPOWS0qklFJKqYeajqly7k6Rqh3/WCmUUkoppdK5Oz38c+o/WZCHmYi8A0w0xlxLi3z3q3jBnEz7pHXc+6L5fBg0ZSPVShegeMFcAPhky0TYlRsEdP8+Ll9BPy92/fAKQ6ZuYsxPW5Mst3Beb6Z98iw5vTzZffQC3YYt4tbtGDJmcGdSnxZUeCwvlyOi6PjpAk5fDAeg5ws16PL0E0THGN4fv4o1O/5yWb0PL+1P5NUbRMfEcDs6hic7jgLgtedr80q7WkTHxPDLbwfpO3ZxknmfqlGKUT1b4+7uxpQFfzBqyhqrzv45mTasCzl9srL7zzN0+3gat25HkzGDB5MGdaRCqYJcDrtKxz5TOH3+cprXKVNGD9ZMeJWMGd3xcHdnwbr9DP5+ddz0z99rQedmlfGt3y8u7bkG5ej7v4YYA/uPBdKl/+wky61QIj8TP2mLZ6YMrPzjCO9/YX0mObw8mTa4A4Xz5eDU+VA69p1BWGRU3LoaVy/BtRu36D7oJ/YcCUxVXcaNGMCOPzbh7ZOT8VPmARAZEc7IgR8QdCEQv7z+9B7wGdmye7Fh9XLmz5oCxpA5SxZee/cjihYrwc0bN/jo7Ze5desm0dHR1KjTkBe7vpZgPRPHjWDt8kXM+eV3p+WYN2MSq5ctws3djf97szcVq9YAYNfWzXz35UhiomN46plWtOnQDYCL588x8tM+RIaH82iJUrz70WAyZMiQqroDfNOvPU2fLE1w6BUqP/8ZAOUe82f8h23JlDEDt6NjeGfEPHYcPI1Pdk++7deeogVyc+PmLV75dDaHTlxIsszC/jmZNrQzOb2zsPvPs3TrN8PeP92ZNLADFUoV4HL4NTp+OJXT50MB6NmlAV1aVrOOyZHzWbPlSKrrkpo6ThvameKF/QDwye5JWGQUAR1GUblMIb78qB0AIjBk4koWb9j/0NcxuWNyzTevki1LJgD8cmRjx6EztPvgx7j5KpUqwIbvXqfzJ7NYsD5pPR/EMelKGqlyLiXPqVLwDpAlDfPdl2NnLhPQ/XsCun9PjVcnce3GLRb/doROgxbEpS/cdJhFvx5OMN+I1xqyatuJZJYKQ7rXZ/y8bZTtNIHQyOt0ebo8AF2alic08jplO01g/LxtDOluPfe1ZOHctK1fmordJtLig1mMfacJbm7isnoDNHllPAEvfBbXoKpduTjN6j5O1fYjqNR2GGN+XJdkHjc3YcwHbWn55jdUeG4obZtUomTRvFad32rJ+BkbKNtyEKER1+jSqrpV51YBhEZco2zLQYyfsYEhb7dwSX1u3LxNkx4TqdZpLNU6jaFR9ceoWqYQABVL5scnu2eC/I8WzEXPznWp330ClV78gl5jljhd7rjez/LGsPmUbTuSRwvmplH1EgD07FyXDduP83jbkWzYfpyenesC0Lh6CR4tmJuybUfSY9h8xvV+NtV1adCkOf0/+ypB2s8zf6Bcxap8M2Mx5SpW5eeZPwCQJ58/Q8d+z7gf5vJ85//jq88HA5AhY0YGfTGRsZN+Ysz3s9m17XeOHIz/qdFjhw9yNTKS5Jw+eYJf163kyynzGPDZV3w7ZhjR0dFER0fz7djh9B/xJV9O/Zlf1/3C6ZPWsTD127G0aNOBb2cuJlu27KxZviDVdQeYtmQbLd+cmCBtyFstGPLdSgI6jGLQtysY8lZzAHp3bcjeo4FUfWEkL/ebyaj3nX/eQ95szviZGyn77FBCI6Po0rIaAF1aBhAaGUXZZ4cyfuZGhrxpLbdk0Ty0bVSBiu1G0OLNbxnbp02aHpPO6tjpox8J6DCKgA6jWLhuL4vWW9vr4PHz1Oz8BQEdRtHyzYmM/6gt7u5Jv3Ietjomd0w2fPUbAjqPJaDzWLYeOMXCDQfi5nFzEwa/0ZQ1244lu9wHcUyqf542qhIRkawiskxE9orIARHpD/gD60VkvZ1ngojsEJGDIjLQTnvLSb5GIvKHiOwSkbkiki2ty1uvYhH+Dgzl9MWIBOnP1S3NT+sOxr1vXvMxTl4I49DJ4GSXVadCEeZv/BOAGav20bzmYwA0q1mcGausE+X8jX9St2IRK73GY8xdd4ibt6I5dSGcE+cuU6Wkf1pW7666t3mSUT+s5uat2wAEh15JkqdK2cKcOBvMyXMh3LodzdyVu2hW93EA6lQpzvy1ewCYsXQbzetZ6c3qPs6MpdsAmL92D3WrPOayOlyNuglABg93PDzcMRjc3IShbz5D3y+XJ8jbrWVVvv35j7gr2eDQq0mWlzdXdrJnzcS2g6cBmLl8J81rl7HqVasM05fvBGC6Y3rtMsy007cdPI13Nk/y5sqeZNl3UuaJSmTL7p0gbevmDdRvYn0Z1m/SnC2/rQegVNnyZMvuBUCJ0uUICb4IgIjgmcW6Lom+fZvo27etMAcQHR3NlG/G8NKrbydbhm2bN1CrfmMyZMxInnz5yZu/IMcOH+DY4QPkzV+QvP4FyJAhA7XqN2bb5g0YY9i3azs16zR0KOOGVNU71ubdf3E5IuH2MMbgldX6IQrvbJk5H2xFeEs+kpeN260v4KOngijsnxO/nElPD3WqFGP+2r2AvX/a+22zOmUd9s+91K1aPC597qrd1jEZeJkTZy5RxW6kpwVndXT0XMPy/LRyFwBRN24RHW397G6mTB7JPtDxYasjOD8mY2XPkok6lR5lycb48+vrbWuycP0Bp+cfeHDHpCsZY1z6Sq+0UZVUEyDQGPOEMaYsMAbr2Vz1jDGxdz/2NcZUBsoBdUSknDFmnGM+EckNfAw0NMZUxBqj9l5aF7ZtvTL8tO5QgrSa5QpyMfQqJ85ZofKsmTPwfvvqDJn6a7LLyeXlSfiV60TbMd1zwRH457YOYP/c2TkbZDXaomMMEVdvkMvLk/y+2TkbHN+YOxccGTePKxgDS756nc0zetGttdWlU6ywLzUrPsqmqe+x6ru3qFQ66cnV39eHsxfC4ssZFEZ+P29y+WQl/EpU3In/3MUw/H297Xm84+aJjo4h4sp1cvlkdUm93NyELT++zekVn7Bu2zG2HzzDa21qsOzXQ1wISRiVKV7Ql+KFcrNu4mts/P4NngpI2tjz9/XinP3lbdU3HH9fqwHjlzNb3DIvhETGfZH7+3pxNsj5PPcj/HIIOXP5ApAjZ27CL4ckybN62UIqVq0Z9z46Opp3Xn6ezq0aUL5yACVKW1+yyxfMoWrNOnHLcyYkOJjcvnnj3uf29SMkOIiQ4CBy++aJS8/lm4eQ4GAiw8PImi077h4ecemXg4Pur9IOen2+gKFvt+DY0n4Me7sF/b5cBsD+o+doWb8cAJXLFKJQ3hzk9/NJMG8u76yERzrsn0Hh+PvZ+6efN2cvhgEO+6d3VvI7pFvzhOGfaLmuUrPCI1y8fIUTZy7FpVUpU4idcz5gx+zevDVsblxdYj2sdXR2TMZqXqcMG3acIPLaDaucvl60qFOGifO3JLu8h+mYVK6ld/8ltR/4XERGAEuNMb+KJAkttxOR7lifXz6gNLAvUZ4AO32zPX9G4A9nK7SX1R3g22+/TXFBM3i48UyN4vT7fn3CwtUvw1yHKNXHXWozft42rl6/lXgR6UqDbmMIDA7HN0c2lk54gyMnL+Lh7kZOryzUfukLKpcpxPQRXSnVPH39FnhMjCGg81i8s2VmzojO1CxflNYNHqfR6xOT5HV3d6NYgdw0eu1b8vt5s+abV6ncYTThV67f07r/yStCEYmLOsXat3s7a5YvZNj4yXFp7u7ujJk0hyuRkQz75D1O/XWcbF5ebN6wmiFjvvvHypsWurepSe8vFrJw3T6ea1ieCZ+055k3JjBq6lpGvf8sW2b05OCJ8+w9co7omJi7L/Ah1q5xRebaUapY2w+eptLzIyhRxI/vB77Iyt//5MbN2w+ohCmX+Jgs/UgeDv1lRVPbNSrPlMXb4vKOfKc5H3+1Is2OpfQSpUnfe6vr6N1/iRhjjopIReBprAefrnWcLiJFgZ5AFWNMqIhMweGHph2zAquNMS+kYJ0TgdhvUPP27CEpKmvjqsXYc+wCQQ5dQO5uQssnS1Dz1fgvqSol/Xm2dkmGvFIf72yZiYkxXL8ZzTcL4zdxSEQU3tky4+4mRMcY8vt6EXjJunoKvBRJAT8vzl2KxN1N8MqaiZCIKM4FR1LA4copv2/2uHlcIdC+0gsOvcLi9fuoUqYw54LCWbjOas/uOHiamBhDbp9sXAq74jBfGAXy+sSX08+Hc0HhhIRdxTubJ+7ubkRHx5A/j0/cOgKDwymQ14dzQWG4u7vhlS0zIWHJd3ukhfAr19m48wR1Kj3KIwVyc3BeLwCyZM7Agbm9KNt2JOeCwtl+8Ay3o2M4dT6UY6cvUaxgbnb+edahvhHk943vhsvv502gHVEMunyFvLmycyEkkry5ssd1HwYGR1DAz/k898M7Zy4uhwSTM5cvl0OC8c4R/2MMJ08c5auRn9JvxJd4efskmTdb9uw8XqEyu7b9TsHCRTl/7gyvdrDGtt24cZ1XXmzBtzMT3pSQy9eXS8HxA74vBQeRy9fP/vtiXHpI8EVy+fqS3duHq1ciib59G3cPD0KCL5LTzp8WOjSrwvujrDFaP6/Zw9cfPw9A5NUbvPJp/A0Ghxd/wt/nEkbxQsKv4p3dYf/08ybQjlwEBoVTII+1H8ftn+FXOWenx8rv50NgUFia1Sc57u5utKxXjpqdPnc6/cjJIK5cu0mZR/Ox68/4qM/DXsfYY7JRQAkO/XWRXN5ZqFy6AM87DFCvWKoAPw62TvO5vLPSuHpJbkdHs2RTfA/Cw3RMKtdKtvvPGDP1Tq9/spD/JBHxB64ZY6YDI4GKQCQQ26/lBVwFwkUkD9DUYXbHfFuAmiJSzF5uVhFJ04E57eonHDcFUL9SUY6eCeGcQ+Om4TvTKPniV5R88Su+/HkbI2duTtCgirVpzyla1ykFQIdG5Vi62Rrzsez3Y3RoZHVVtK5Tio27T1rpfxylbf3SZMzgTuG83hTLn5Pth11zd0qWzBnj7rzJkjkjDQNKcvDEeZas30edytZYi2KFfMmYwT1Bgwqsxlaxgr4U9s9JBg932jauyLKN1t05m3Yco3WD8ladm1VlqX130rKNB+jQrKpV5wbl48a/pLXcPlnxzma1yTNn8qBB1eLsPnyWos8MpuSzIyj57AiuXb9F2bYjAViy6SC1Kz4CQC7vLBQvlJu/zyW8K/FCSCSRV2/EDXh/8elKLN1k7SfLfj1Ex6crAdDx6Uos/TU+/UU7vWqZQkRcuZ6k6/FeVK1Rh3W/WIPp1/2yhGo16wIQfPE8wz7pyTsfDSJ/wcJx+cPDLnPFHoh+48Z19u7YSoFCRahcvRZTF6zhuznL+W7OcjJlypykQWWtry6/rlvJrZs3uXj+HOfPnqZ4ybIUL1GG82dPc/H8OW7dusWv61ZStUZdRITHK1Rm88Y1ScqYFs4HR1Cr0qMA1K1SnONnrDGN3tkyk8HDHYCurQL4bfcJIq/eSDL/ph3Had3gCcDePzdag6OXbXLcP59g4/bjdvpB2jaqYB2T/jkpVtCX7fY4HleqX/Uxjp68yDmH7qrC/jnjBqYXypuDEkX8OBWY9A7ah62Ozo7JI6esLuFn6z/Oit8OJ4i2lWo9Iu5YXbB+P++MXJigQQUP1zGZVh6GMVUi4iMi80TksIj8KSLVRSSniKwWkWP2/znsvCIi40TkuIjss4Mnsct5yc5/TEReup/P5a5PVBcRX+ADrK6suIiMMab+/az4IfY4MFJEYrB+9/A1oDrwi4gE2uOldgOHgTPAZod5JybK1wWYJSKZ7OkfA0fTopBZMmegfqWi9Bi9IkF623qlk4yxupMFw57n9VHLOB9yhb4T1zHtk2fp360Oe49fZMqKPQBMWb6HyR+15MC01wiNvE6nQdaV958nL/Hzhj/Z/cMr1u3i41YS46L7bP1yZWfO5/8DwMPdjTm/7GT173+SwcOdbwe8yI6f+nDzVjT/6z8dgHy5vfi63ws8+9a3REfH8O6IeSz56nXc3dyYungLf/5lRTP6jlvMtGFd6P/GM+w9fJYpC61xEVMW/sHkQZ04sOgTQsOv0enDKS6pV97c2fnuk3a4u7vhJsLPa/exYvPhZPOv3nKUhtUeY9es94iOjuGj8cu5HGE9wWPLj28T0HksAG+PXMDET9rhmSkDq/44wso/rFvOR/24gelDOvBSiyqcvmDdvg3wy++HaVyjBAfn9eba9Zu8Mnhuqusy6tM+HNizk4jwMLq1acwLXV/luRe7MnLgB6xZvhDfPPnoPcC6DX/21IlERoTx7ehhALi5u/PFxJmEhlxizLB+xMTEYGJiqFnvKarUqH3H9W7dvIHjRw7RodvrFCr6KDXrNqJHl+dwc3fnlXf64O5uNV66v/0BA3q9TkxMDA2atqRQUaux89IrbzPq0z7MmPQ1jxQvwVNPt0p13QGmDulErUrFyO2TlePL+jNo4i+8MXgOI3s+i4e7Gzdu3qbHkJ8A6w627wa8iAH+PHGBVwfFR60WjP0/Xh80h/OXIug7finThnai/2tN2XvkHFMW2fvnoq1M/rQDBxZ8RGjENTp9NA2AP/+6wM9r9rB7bh/rmPxsXpoek87qOHXRVto2qsBPq3YnyFuj/CP0fKkBt25HE2MMbw+fR0j41Ye+jnc6Jts+9QSjftyQ4mU96GPyP2As8Isxpo2IZMS6+/4jYK0xZriI9AH6YLVhmgLF7Vc1YAJQTURyAv2BylhDnnaKyGJjTOi9FEju1iIUkVXAHKwur1eBl4BgY8wH97JCdVfGs37Kuv/Sq6h1ffGs+G8dkhcvatc4PAP+3YdJ1JYRHD7v0seyPRRK5suCZ+V/90+eRu0Y/d+o47/8mASI2jLCtc+2AV77+ZBLB39NeK70Hetg/ybxHuAR49CQEZEjQF1jzHkRyQdsMMaUEJFv7b9nOeaLfRljXrHTE+RLrZTc/ZfLGDMJuGWM2WiM6Qb8W6NUSimllHrARKS7/eii2Ff3RFmKAsHADyKyW0S+F5GsQB5jzHk7zwUg9rbf/Fi9S7HO2mnJpd+Tu3b/YXWBAZwXkWewHhuQ8w75lVJKKfUvFuPiuxQT3cDljAfWmOc3jTFbRWQsVlef4zKMiPyjt1OmJFI12A6zvY/VBfg98O+OESullFLqYXYWOGuMif3NtXlYjayLdrcf9v+xD547BxR0mL+AnZZc+j25a6PKGLPUGBNujDlgjKlnjKlkjEl6241SSiml/hOMce3r7us3F4AzIlLCTmoAHAIWY439xv5/kf33YqCzfRdgABBudxOuBBqJSA77TsFGdto9Scndfz/g5CGg9tgqpZRSSqkH4U1ghn3n319AV6xg0U8i8jJwCmhn512O9fzJ48A1Oy/GmMsiMgjYbuf71BiT9NkfKZSSMVVLHf7ODDyLNa5KKaWUUv9BD8OT340xe7AehZBYAyd5DfBGMsuZDEx2Ni217tqoMsb87PheRGYBv6XFypVSSiml/i1SEqlKrDiQdr/joJRSSql05SEIVD2UUjKmKpKEY6ouYD2dVCmllFL/Qa5+pEJ6lZLuv+x3y6OUUkop9V9310cqiMjalKQppZRS6r/BuPiVXiUbqRKRzFg/TpjbfnZD7O/weHEfj3BXSimllPo3ulP33yvAO4A/sJP4RlUE8KVri6WUUkqph9XD8EiFh1GyjSpjzFhgrIi8aYwZ/w+WSSmllFIq3UnJb//FiIhP7Bv7Ue6vu65ISimllHqYxRjXvtKrlDSq/s8YExb7xhgTCvyfy0qklFJKKZUOpeThn+4iIvYj3hERdyCja4ullFJKqYeVjqlyLiWNql+AOSLyrf3+FTtNKaWUUkrZUtKo+gDoDrxmv18NfOeyEimllFLqoaaBKufuOqbKGBNjjPnGGNPGGNMGOATo3YBKKaWUUg5S9IPKIlIBeAFoB/wNzHdloZRSSin18NIxVc7d6Ynqj2E1pF4ALgFzADHG1PuHyqaUUkoplW7cKVJ1GPgVaGaMOQ4gIu/+I6VSSiml1EMrPT9LypXuNKaqNXAeWC8i34lIA+J/qkYppZRSSjmQu/WLikhWoCVWN2B94EdggTFmleuL95+k7X+llFL3w+UBkPZTd7v0u2r2SxXSZRDnrgPVjTFXgZnATBHJAbTFesyCNqpcxLN6nwddBJeK+mM4nhV6POhiuFzU7i/xrPjWgy6GS0XtGvevryNY9ey38tiDLoZLfdq4OJ6V3n7QxXCpqJ1j8XzykwddDJeL+m3Qgy7Cf1ZKfqYmjjEm1Bgz0RjTwFUFUkoppdTDzbj4lV6lqlGllFJKKaWcS9FzqpRSSimlYsXoc6qc0kaVUkoppVJF21TOafefUkoppVQa0EiVUkoppVJFf6bGOY1UKaWUUkqlAY1UKaWUUipVNFDlnEaqlFJKKaXSgEaqlFJKKZUq+kgF5zRSpZRSSimVBjRSpZRSSqlU0UCVcxqpUkoppZRKAxqpUkoppVSq6HOqnNNIlVJKKaVUGtBIlVJKKaVSJUYDVU5ppEoppZRSKg1opEoppZRSqWLQUJUzGqlSSimllEoDGqlSSimlVKrozX/OaaRKKaWUUioNaKRKKaWUUqmiz6ly7qFtVImID/CiMebre5h3CrDUGDMvDcqxAehpjNlxv8tKK5kyerBmwitkzOCBh7sbC9bvZ/D3a6hT6VGGvfk0GT3c2X3kHK8O/Zno6Ji4+SqVKsCGia/Rud8sFqw/kGS5FUrkZ+InbfHM5MHK34/w/uglAOTw8mTaoBcpnC8Hp86H0vHjmYRFRgHw+bvNaVyjBNeu36L7oLnsORrokjoXyOPD94M645crO8bA5J8389WsDUwb3pXiRfIA4JPdk7DIKALaD08y/1M1SjGqVxvc3dyYsvB3Rv2wGoDC/rmYNrwrOb2zsvvP03T7+Edu3Y4mYwYPJg3qRIVShbgcfpWOH0zm9PnLLqmbozc71KVLq+oYYzh4/DzdB8ygevlHGPp2S9zchKvXbvB/A2bw15lLSebt2fUpurQKIDo6hvdH/syaPw7H171na9zd3Ziy4A9GTVlj1z0n04Z1IadPVnb/eYZuH0/j1u1ol9aveGE/pg3vEve+aP7cDPpmOTOWbmPa8C4U9s/JqcDLdPzgh7h9zFGHZlXp879GAAz/fhUzlm4DoEKpgkwc0AHPzBlY+dsh3h/5MwA5vLKkaLn36ua1K2yfNY7w86dBoOqLb5O7aCmOblzC8V+XIW5u+JepzBMtu3Fy+3qOrJsfN29Y4Eka9RpLttx5WTf2g7j0a2EhFK5cl4rPdU+yvkOrfuLvLasRNzcqPNedfKUqAXD+0E52z5+IiYnhkeqNKPVUWwCuhFzgjymfcfNqJDkKFqNap/dw98hwz/X9pt8LNK1VhuDLV6j8fMLj7O2O9Rj+bisKNPiIkLCrcemVShdiww/v0PmjqSxYuzfJMiuULMDEgR3wzJSBlZsP8f5I6zPK4ZWFacMctl2f+G33ea/WNK5Z2jrvDJjBnsNn77lOyTk89z0ir90kOiaG29ExPPm/b+Lr2r4Gw3s0pcAzwwgJv4ZX1kxM7teGgnl88HB3Y8ys35i2fHfSupbwZ+JHra1z7B9HeX/scquu2T2Z9mk7CufNwakLoXTsN4ewyOtWXd9+msbVH7PqOnQ+e46eT/O6qrTzMHf/+QCvP+hCPIxu3LxNkx7fUa3zWKp1HkujgMcIeLwQ33/Sls6fzKJyxzGcvhBGx6crxs3j5iYMfr0pa7YdS3a543q34o1hP1O27SgeLZibRgGPAdCzU1027DjO4+1GsWHHcXp2qgNA4+oleLRgbsq2HUWP4fMZ17uVy+p8OzqGPl/Mp+JzQ6jTeRSvPF+bko/kpVOfHwhoP5yA9sNZuHYPi9btSTKvm5swpk87Wvb4mgrPDaZtk0qUfCQvAEPebsn4Gesp23IgoZFRdHm2OgBdWlUnNDKKsi0HMn7Geoa83dJldYvl7+vN6+3rULPjKCq3G467mxttG1dk3Ift6PrxjwS88BlzftlJn5cbJ5m3ZNG8tG1ckYpthtGixwTG9mmHm5tYdf+gLS3f/IYKzw216l7UrvtbLRk/YwNlWw4iNOIaXVpVd3kdj50KIuCFzwh44TNqdBjJtes3Wbx+Lz27NmTDtqM83mowG7YdpWfXp5LMm8MrC327N6F25y+o1elz+nZvgk92TwDGfdiONwbPpmzLQTxayJdGNUoBpGi592P3/InkLVWJpz/+hsYfjMcrT0EuHt1H4P4tNP5gPE0/+poS9VsDUKRKPRp/MJ7GH4ynWqf3yZozDzkKPEKGzFni0ht/MJ6sOXwp8ESNJOsKP3+a07s20eTDr6n92kB2/jSBmJhoYmKi2Tl3ArVfHUiTj77m1M6NViMP2LdoCiXqtuSZft+RMUtW/v5j9X3Vd9qSbbR885sk6QXy+NAgoESSCw83N2HwW81Zs+VIsssc92E73hg0m7KtBvNoQYdt16UhG7Yf5fFnB7Nh+1F6dmkIQOOapXm0oC9lWw2mx+DZjPuw7X3V6U6avDWZgK5fJ2hQFfDzokGVYpy+EBaX9krrahw+GUy1Ll/R+M1JDO/RhAwe7knr+n5z3vhsIWXbj+HRgrloFFDcqmvHWmzY+RePvzCGDTv/omfH2lZdA4rzaMFclG0/hh4jFzGuZ3OX1TW1YoxrX+nVw9yoGg48KiJ7RGSkiPQSke0isk9EBsZmEpHOdtpeEZnmMH9tEfldRP4SkTZ23roiskFE5onIYRGZISJiT2sgIrtFZL+ITBaRTIkLJCIv2NMPiMgIh/SXReSoiGwTke9E5EsRyS4if4tIBjuPl+P7+3U16iYAGTzc8fBwJzracPNWNMftCMa6bcdoVbdsXP7X29Zg4Yb9BIdedbq8vLmykz1rJrYdPAPAzBW7aF6nDADNapVm+vJdAExfvovmte302qWZucJK33bwDN7ZPMmbK3taVC+JC5ci4q5Gr1y7weG/L+Dv65Mgz3NPVeSnX3YmmbdK2SKcOHOJk+dCuHU7mrkrd9GsbjkA6lR5jPlrrCvKGUu20rzuE1bd6pZjxpKtAMxfs5u6VUu4pF6Jebi74ZkpA+7ubnh6ZuB8cATGGLyyZgbAK5sn5y+FJ5mvWd3HmbtyFzdv3eZU4GVOnA2mStnCVClbmBNngxPV/XEA6lQpzvy1ewCYsXQbzes9/o/UMVa9qiX4++wlTp8PpVmdx5luR52mL91G87pJy/JU9ZKs3XqE0IhrhEVGsXbrERrVKEXe3F5kz5qZbftPAjBz6Taa17O2b0qWe69uRl0l+PhBHqluRc7cPTKQMUs2Tvy2nJJPtcU9g3WoZ87uk2Te0zs3UqhS7STpkUHnuH4lHN9HyySZdm7/FgpVrI17hgxky5WX7L75uHzqKJdPHSW7bz6y5c6Lu0cGClWszbn9WzDGcPHYPgqUfxKAIlUbcG7/H/dV5827T3A5/FqS9M/ee5a+Yxcn6RJ6/fnaLFy7l+DQSKfLy5vbi+zZMrPtwCkAZi7bHreNmtUp63TbNatTlpnLtgOw7cAp67yT2+u+6pUan735NH0nrEpQV2MgW5aMAGT1zEhoRBS3HXoJAPLmymafY63z2Mxf9tC8ltWAbFarFNNXWOeh6St2J0if+cseALYdPGufY7O5tH4pZYxx6Su9epgbVX2AE8aY8sBqoDhQFSgPVBKR2iJSBvgYqG+MeQJ422H+fMCTQDOsBlqsCsA7QGngEaCmiGQGpgDPG2Mex+oWfc2xMCLiD4wA6ttlqCIirez0T4AAoCZQEsAYEwlsAJ6xF9EemG+MuXXvH0k8Nzdhy9S3OL38Y9ZtO8b2Q2fwcHejYsn8ADxb73EK5PEBwN/XixZ1yjBx/tZkl+fv6/X/7d13eBRV98Dx70looSShJGCoAkoRlA6CCCqKShXFQhHw/b2oqFhRbAhSBMUC+ipiAykWLICiFFFQEKQrXRSlJEBCSEiAIJCc3x8zCRuyCIFdlmzO53n2yc7dmdlzd7K7d8+9c4fY+ONf1rHx+4mJcj6ooksVZ3ei86G4OzGV6FLFs7bZuSf5+DYJx7fxp0oXlKJejQosX/d3VlmLBtXYsy+VP7cn5Fg/JjqCnXuSjse5J4nyURGUjizG/tS0rC7S2D1JxERHHN9mt7NNenoGKQfSKB1ZzI+1griE/bw26Xt+/2YIf80dRkrqYeYv3US/oR/x5dh7+OPb5+nWrjGjP/gux7blc9QxmZioSGKiItnp8Ys6Nj6Z8tFu3Q941j2ZmKgIv9bvRF3bNuDTOU4jOLp0CXbvTQGcBnS0l8Z5THRk1jEBN+boSGKiIoiNTz5eHp+cdRxPZ79n6mDiHgoXD2fZlNeYM6o/y6aO5dg/h0lNiGXvn+uZ9/IjfD9mIInbfs+x7fZVP1GpQc5G1faVP1KpQUvc33rZpO1PpGjJqKzlsMgypCUnkpacSFjk8fKikWVI25/IkYMpFAorRkhoaFb5of2Jvqh6Nu1b1SEuYT9rt2Tv+o+JiqDjVZcy/rPFJ902JiqCWM/PEPeYwsmPXUx0ZPbPnfj9fvnfVYWvXunF4vfu4a6OjQBof0VN4vamsPaP3dnWHff5UmpWjmLr9MdZMfF+HhvzTY6GQUyZcGITUjziTiHGbQxGlyzG7sQDTl0TDxBdsljWNjtP/Fw+hw1Ik3vnc6PK03XubTWwCqfhchFOA2eaqu4FUFXP3PN0Vc1Q1Q1AWY/yZaq6U1UzgDVAFaAG8JeqZn76TQRO/MRrDCxQ1QRVPQZMcddpAixU1X1ug2maxzbvAn3c+32AD7xVTkT6isgKEVkxfvz403pBMjKUZr3GUr3TCzSqXZHaVcty56CPePHB9vz03n2kHvon6wvzpYfa88z/vvVZ6z+QPyKKhRXio9H/x4DRn5N68HBW+a3XN2La7PNm2NsZiSwRRvvWdanVfghV2z5DsbBC3H5jIx7ofhU39R9H9RsGMWnmUkY9clOgQz1rBQuE0u7KOnwxb43Xx/31P+bL/WpGOkk7/6T6FTfS9omxFChcmI3fTSMjI51/DqXS5pGXuaxzH5Z8MCrbey/x780UKFSYyJgqOfa5fdWPVGrQyndB+llYkYI8fte1PD/umxyPvfTYTTwzNmf26kyd68+da/q9Q/P/vEXnRydxd5emtLisMo/feSXPvzs/x7rXNr2I37bspmrnF2na501efbg9JYrm6Ow4bXkhT6Pq31tedd4OVD+BAC+o6tvZCkUe+Jdt/jlhe2/l6fjxNVDVxSJSRURaA6GqmnN0uLPeeCCzNaUPfjDwtJ9j/4HDLFy1leuaXcxrU3+izb3OS3RNk4u4qFIZABrUrMCHQ7sBUDqiKG0vr8Gx9Ay++nFD1n7iElIoH33811756Aji3F9V8fsOUK50CXYnplKudAkSkg5kbeNkw5zUffmo49v4Q4ECIXw0+r988u0KZnx/fMBraGgIna6+jBbdXvS6XVz8fiqULZm1XL5sSWIT9pOYfJCIEmGEhoaQnp5B+bIliXN/FcbF76dCuZLExicTGhpCePGwbINv/eHqpjX4OzaRvcnO6zv9+1+5/LKq1L2oPMvd7pHP5q5mxhv35tg2NkcdI4lLSAagQrnI4+XRkcTGu3Uv7ln3SOIScnYr+kvbFrVZs2kn8fucDGh8YirlyoSze28K5cqEk7AvZ3dRXHwyLRtddLwuZSP5acUW4hL2U97NboBTx8zjeDr7PVNhkWUIiyxD6SpO13DFei3YOO8zikaUocKlzRERSleuASL8cyCFIiWc99f2VT9SqWHOhlNS7FYyMtIpVam69+eLKM2hpOOZ2LTkvYRFlnbvHy8/lLyXsIjSFCoWzpG0g2SkpxMSGsqh5L0UjSjts/oDVK1QhsoxpVn20eOA89ovmTKAlne+TINalfjwhV4AlI4sTtsWtZ3PnQVrs7aPS9hPeTejDu7/rZt1PNmxi4tPzsrCO88Z4Zf/3bi9zvMlJB9k5o8baFnvQipfUJJlE+5znjcqnCXv30vL/75Nzxvr8/LknwDYGruPv3clUaNyGVZsjPXYXwrlPTL55aPDiXMzcfFJBylXuji7Ew9QrnTxrGEacXtTqHDi5/Je/33GmrN3PmeqUoHMXP0c4C4RKQ4gIuVFJBr4HugqIqXd8lJn+FybgSoikvlp1hNYeMI6y4BWIlJGREKBO9x1lrvlJUWkAHDzCdt9CEzlJFmqM1EmshgRxZ0xNkUKF+CaxtXZvC2BKDdlXKhgKI/2bMU7XzrdfbVufpGaXUZRs8sovvxhHQ+Nnp6tQQVOt17qwX9ocklFALrd0ICv3XVmLdqQNei9x40N+Pont/ynDXS7wSlvcklFUg4ezuom9Idxz3Vn81+7GTv5+2zlVzetwe9/78nWBeRpxfptVK8UReWY0hQsEErXtg2YteA3AH5c8Ttd2tQHoHuHpnztls9auJbuHZoC0KVNfRYuz9mF42s7difRpG4Vwoo4Y3GuanIxm7buJrx4EapXcrp3rm5ag81/7c6x7ayFa+natgGFChagckwpqleMYvm6baxYv53qFaOoHFPqeN0XrnXrvoUu19QDnLPqvvb4svO3W68/3vUHMOvHdfRo3wSAHu2b8PXCnLHMW7KJNs1qElkijMgSYbRpVpN5Szaxe28KqQcP06RuFQC6edTldPZ7psLCS1I0sgwpe5wxMns2/0p4uUqUv7QZ8Vuc/6PU+Fgy0o9RuLjzZaoZGexYffKuv8peGluZytdtyvZVP5J+9CgHEneTmhBHqcoXU6rSxaQmxHEgcTfpx46yfdWPlK/bFBEh+qK67FyzCIC/l80npm4zn9UfYP0fu6h87TPU7PA8NTs8T2x8Mpd3f4k9ianU6vh8VvmX89fw0Mhp2RpU4HTrpR44TJM6lQHo1q4xXy90fnvmPHbHy7u1awxAkzqVSTlwOKub0FeKFilI8bBCWffbNK7Oyk07qdxhFDW7vkLNrq8Qm5DC5Xe9xZ59B9ixZz+tG1UFnK68iyuV4a+4pGz73J14wP2MreDU9fp6fP3TRqdOizbR4wbnc6jHDfWzlXe7vp5T10sqOHV1uwkDLUPVr7e86rzNVKlqoogsFpF1wLc4DZMl7liDA0APVV0vIsOBhSKSjtM92PsMnuuwiPQBprkNo+XAuBPW2SUiA4EfcDJfs1R1BoCIjMBpdO0DNgGeP5umAMOAj3Ib18mUK12CdwbdSmiIECLC59+v5dvFmxhx/w3c0KIWISK88+VSFq7885T7WjqxP816jQXgwZemM/6ZroQVLsjcpZuZs8Q5Y2f0hwuZPLwbvTo0ZvtuZ0oFgNk/b6Zt85qsnzaAQ/8c5e5h0076PGereb2qdG/flLW/x7L0YyeT99wbM5mzaANd2zbMMUD9gqgI3hzUjZseeIv09AweHvUpX715H6EhwsQZS9m41WmYPD1mBpNG9uG5fu35dfMOJkx3BvJOmP4z7w+7k3UzniMp5SA9B/qsTXxSy9dt48v5a1gy5XGOpafz6+ZY3vviZ2Ljk/nopf+QoUpyyiHuHuK8/u2urEOD2pUYOu4bNm7dzefzVrP6s6c4lp7OQyOnkZGhgPLwqM/46n/9CA0JYeJMj7qPncmkF3rz3H3t+HXTTiZMX+r3OgIULVKIq5vW5P7hn2SVjf5gHpNH9aFX52Zs35VEjyec17tBrYr83y1X0G/oRySlHOKFd+ewaPJjAIx4ZzZJKc6g6Qdf+NQ9Lb8Qc3/ewJzFG/51v77S4JZ7WPrhaDLSj1G8dDmadH+I0EKFWT51DN++0I+Q0II07fFw1hiphD/XERYZRfEy5XLsa8fqn7jynsHZymLX/sK+7Vuo264HERdUplL9lnw74l5CQkNp2PVeQkJCs+JY+OYgZ0qFZtcScYHTSLmsYx+WTBjF2lmTiaxQlarNrjur+k4cfictG1WnTGRx/vhmCEPf/paJM3L/f7N06gCadXsJgAdHTsuaDmPuYo9jN+E7Jo/sQ69Ozdi+ax89Bk4AYPaiDbRtUZv1M57l0OEj3D146lnVyZvoUsX5ZIST3S8QGsIn835j3i9/nHT9kRMWMP7pLiyfeD8i8PRbc0l0B/Qv/aAfzfo4MwM9+PJXjH+6i/sZ+ztzljpnY4+e/COTn7+NXu0asn1PMj2edd4bs5f8TtvLL2b9Jw9z6PBR7h7xhfcAzHlD8vIo+/OFiBRX1QNug+xL4H1V/dJ97Bagk6r2PM3dadjlp9/9lxelLRlJWP37Ax2G36WtfoOwBv0DHYZfpa0aG/R1BKeeg+acfDqSYPB824sIa/jgqVfMw9JWjiHsimcDHYbfpS0amvNMBx+7euwSvzYevu9/ud/r4A/nbaYqjxksIm2AIsBcYDqAiLwO3ADcGLjQjDHGGHMuWKPKB1T1sZOU/9tAemOMMSZPsl4u787ngerGGGOMMXmGZaqMMcYYkyuWqPLOMlXGGGOMMT5gmSpjjDHG5IqNqfLOMlXGGGOMMT5gmSpjjDHG5IolqryzTJUxxhhjjA9YpsoYY4wxuWJjqryzTJUxxhhjjA9YpsoYY4wxuWKZKu8sU2WMMcYY4wOWqTLGGGNMrliiyjtrVBljjDEmV6z7zzvr/jPGGGOM8QHLVBljjDEmVyxR5Z1lqowxxhhjfMAyVcYYY4zJFRtT5Z1lqowxxhhjfMAaVcYYY4zJFVX/3k6XiISKyGoR+dpdvlBEfhGRP0TkExEp5JYXdpf/cB+v4rGPJ93yzSLS9mxeF2tUGWOMMSavehDY6LE8CnhVVasDScB/3PL/AElu+avueohIbeB24BLgeuBNEQk902CsUWWMMcaYXFFVv95Oh4hUANoB77rLAlwNfOauMhHo7N7v5C7jPn6Nu34n4GNV/UdV/wL+AJqc6etijSpjjDHGnFdEpK+IrPC49fWy2mvA40CGu1waSFbVY+7yTqC8e788sAPAfXy/u35WuZdtcs3O/jsPpS0ZGegQ/C5t9RuBDuGcSFs1NtAh+F1+qCPA820vCnQIfpe2ckygQ/C7tEVDAx1CUPD3yX+qOh4Yf7LHRaQ9EK+qK0WktX+jOX3WqDoPhdW/P9Ah+FXa6jcIa/hgoMPwu7SVYwhr9HCgw/CrtBWv5p9jGeT1TFs5hsSDx069Yh5WulgBwhr0D3QYfpdPfui0ADqKyI1AESAcGANEikgBNxtVAYh1148FKgI7RaQAEAEkepRn8twm16z7zxhjjDG5EugxVar6pKpWUNUqOAPNv1fV7sAPwC3uar2AGe79me4y7uPfq/NEM4Hb3bMDLwQuApad6etimSpjjDHGBIsngI9FZBiwGnjPLX8PmCQifwD7cBpiqOp6EfkU2AAcA+5T1fQzfXJrVBljjDEmV86nCdVVdQGwwL2/FS9n76nqYaDrSbYfDgz3RSzW/WeMMcYY4wOWqTLGGGNMrti1/7yzTJUxxhhjjA9YpsoYY4wxuWKJKu8sU2WMMcYY4wOWqTLGGGNMrtiYKu8sU2WMMcYY4wOWqTLGGGNMrliiyjtrVBljjDEmV6z7zzvr/jPGGGOM8QHLVBljjDEmVyxT5Z1lqowxxhhjfMAyVcYYY4zJFUtUeWeZKmOMMcYYH7BMlTHGGGNyxcZUeWeZKmOMMcYYH7BMlTHGGGNyxRJV3lmmyhhjjDHGByxTZYwxxphcyciwVJU3lqkyxhhjjPGB8yZTJSKdgd9VdUMut2sNHFHVn0+xXkegtqqOPNMYz5SIRALdVPVNfz7PRZWjmTTqrqzlC8uXZuhbs3hj6oJs6738+C20bXEJhw4foe9zk1izaScA3Ts0ZeD/tQVg5LtzmPLVLwDUr1WR8UN6Ela4IHMWr+fRFz/zZzWyjBt0Bze0vISEfQdodJtz2AbdeyPtW9UlIyODhKQD9H1uCrv2phBevAjvD+1JxXIlKRAawmuTfmCSG7+n+jUrMH5Id7cuG3j0pS8AKBlelEkv9KZyTCm2xe2jx8APSE5NA+DlAV1o26I2hw4fpe/gKVmvl2/qeDs3XFGbhKQDNLrtRQAmjbiTiypHAxBZIozk1DSadR8NQJ3qF/DGU7dSolgRMjSDK+58lX+OHMu2T6cud1L5glJs27WPHgMnHq/LYzfRtkUtty4fsWaze+zbNWbgf64FYOR785gya7nP6ujUM+exzPRgj6sY+XBnKlzzFInJB7PKG9auxIIPHuLOpyby5fxfc+zz/DuWOevYpU09nu57PTUvLEvLO19h1cYdADS6pBJvPH0bACLC8PGzmfnDbzn2WTmmFJNe6EWpiGKs3riDu56dzNFj6RQqGMp7z/egfq2K7Nt/kB4DJ7J91z4AHuvTht6dmpGensGjo7/guyWbcl2X4YOfYfFPCylZqhRTps0A4Pt5c3jv7f/x919beXfSx9SqXQeAXXGx3HFzBypXrgLAJXUv4/GnnwPgvv/2JnFvAoULFwbg1TffoVSp0lnP88P8uTw94GHem/xJ1v48LV38E6+NHkl6ejodbrqZO/v8F4C42J0MevIx9icnU7PWJQwa9gIFCxbiyJEjDH32STZtXE9EZCRDR77MBTHlc13/E0UUD+OtQXdQu9oFKMo9Q6bS6erLuLFlHY4cO8ZfO/bSd/BU9h9Iy7Httc1rMfqxLoSGhjDhyyWMnvAdkHlse1Mq0j22z0xyj20B3hvqHtvkg/QYOCHr2J5PbEyVd+dTpqozUDs3G4hIAaA10PxU66rqzEA0qFyRQD9/P8mWbfE0u30kzW4fSfNuozh0+Cgzf8j+ZdT2itpUqxRFnU5DuH/YR4x96nbA+SJ6uu8NXNlzNC17vMTTfW8gskQYAGOfuo37hk6lTqchVKsUxXUtcnWYztikr5bR6YFx2cpe/XA+TW4fRbNuL/HtT+t58r/XA3B315Zs2rqbpne8SNu+rzPy4U4ULBCaY59jn7yV+4Z+TJ3Ow6hWMYrrmtcC4LHebViw/Hfq3jSMBct/57HebQBo26I21SpGUafzMO4f9jFjn+zqhzqOz1bW86kPadZ9NM26j2b6978yw/2yDQ0N4f2hPXjghWk0vG0Ube/+H0ePpefY52O9r2HBsi3U7TKCBcu28Fjva9y61HLqctMI7h/+KWOfvAVwj/1/23Jl79do2etVnv5v26xj79t6jstRXqFsJNc0q5HjSyMkRBjWvwPfLd180n2en8cyex3X/7GL2we8z6JVf2Yv/3MXLXq+TLNuL9HpgXG8/tSthIbm/Dge3r8jr09ZQJ3Ow0hKSaN352YA9O58OUkpadTpPIzXpyxgeP8OANS8sCxdr2tAg64v0PGBcYwZ2JWQEMl1XW7s0JlX33g7W1nVatUZMXoM9Ro0yrF++QoVmfjxF0z8+IusBlWm54aPynrMs0F18OBBPp06mUvqXOo1hvT0dEaPGs7Lr49j6ucz+W72N/y19Q8A3hz7Crd1v5NpM2dTIjycr6Y7Deqvpn9OifBwps2czW3d7+TNMa/kuu7ejB7Qhbk/b6TezcNpctsoNm3dw/ylm2l46ws0uW0UW7YnMOCua3NsFxIivPZEVzo9MI76N4+g6/UNqXlhOQCG9+/kHNtOQ0lKOUTvzpcD0LtzM5JSDlGn01Dn2D7Y0Sd1MOeGXxtVItJDRJaJyBoReVtEQkXkgIgMF5FfRWSpiJQVkeZAR+Ald91q7m22iKwUkZ9EpKa7zwkiMk5EfgE+Be4BHna3aykiHUTkFxFZLSLfiUhZd7veIvKGxz7GisjPIrJVRG5xy1uLyEIRmeGWjxSR7m4d1opINXe9KBH5XESWu7cWbvlgEXlfRBa42/d3X4qRQDU3xpf8+ZpnuqpJDf7amcD2XUnZytu3upSpXy8DYNnav4koEUa5MuFc27wW85duIinlEMmpacxfuonrWtSmXJlwShQrwrK1fwMw9etldGjt/UPQ1xav/pN9+w9lK0s9+E/W/aJhhVCcn0uKUrxYEQCKFS1MUsohjqVnZNu2XJlwShQvwrJ12wCYOms5HVrXBaB9qzpMdl+XyV8vy1Y+1c3aLFu3jYjizuvluzpuZV/KwZM+fnObenw6ZxUAbZrVYN2WONZuiQNg3/5DXsc1OHVZ7tYlex2nfuNRlxJhlCsdzrWX12D+ss3Hj/2yzVzXvKbP6ujUM+exBHjxkZt4eszMHHPe9LvtSqbP/5WEpFSv+zs/j2XOOm7+ew9btsXnWDft8FHS3f/PwoUKnPRXf6vGF/GFm6WbckJdprh1/GL+r7RucrFT3rou0+au4sjRdLbF7ePPHQk0vqRyrutSv2EjwiMispVVqVqNylUuzPW+TuadN8fSo/d/KORmsU60Yd1aKlSoSPkKFSlYsBBt2t7ITwt+QFVZufwXrrrmOgBuaN+JH3+YD8BPC77nhvadALjqmutYsXzpWc+nFF68CFc0qM6E6UsAOHosnf0HnM/IzGO4bO3flI+OzLFt4zqV+XNnAn/HJnL0WDrT5qyivXsMnWO7BnCP7VXusW1d1+PYrqF144vPKn5/UVW/3vIqvzWqRKQWcBvQQlXrAelAd6AYsFRVLwN+BP7rdt3NBAaoaj1V/RMYDzygqg2BxwDPrrMKQHNV7QKMA151t/sJWAQ0U9X6wMfA4ycJ8QLgCqA9TqMn02U4DbVaQE/gYlVtArwLPOCuM8Z9zsbAze5jmWoCbYEmwHMiUhAYCPzpxjjg9F7Bs9O1bUM+nb0yR3lMdCQ7dx9vaMXuSSYmOpKYqEh27vEoj08mJiqSmOhIYuOTc6wfSIP7tWPLrMHcfn0jhr71DQDjPvmJmheWZeuc51nxyUAeG/1FjjdmTFQEsXuSs5Y96xJdugS796YAsHtvCtGlSzjbREey03Ob+P3ERGX/svGXFvWrsmffAf7csReAiypFocDM1+/m58mP8sidV3vdLrpUCXYnunVJTCG6lFuXqAh27k7OWs+pf4RTfuLrcg7q2L5VHeIS9mc1EjPFREXQ8apLGf/Z4pNum9eOpTeN61Rm5acDWfHJQPq/8GnWF3Sm0pHF2J+allWe+Z4Esr1f09MzSDlwmNKRxSif4xjvJyba/3XcFRtLrztupt//9WLNquyfO8MHP0Ov27vwwTtvZb0nN2/cQPye3bRo2eqk+0xI2EPZchdkLUdFlyUhfg/7k5MpXrwEBQo4o1eiy5YlISHe3SaesuWcTFCBAgUoVrwE+5OTz6puVWJKszfpAOMHd2fJ1Md589k7KFqkULZ17uzUjDk/5xy5EhMVmf14xCdTPjrCObYHPI6tx3vO833qeWxN3uDPTNU1QENguYiscZerAkeAr911VgJVTtxQRIrjdOlNc7d9G6cRlGmaqubs93BUAOaIyFpgAHDJSdabrqoZ7hiush7ly1V1l6r+A/wJzHXL13rE2gZ4w41tJhDuxgwwS1X/UdW9QPwJ+/ZKRPqKyAoRWTF+/PhTrX5KBQuE0q5VXb6Yt/qs93U+GvzmLC5qN5iPZ6/gntuuBODay2vy2+ZYqrYdRNM7XuTVx2+hRDHvv4BPx/nwQ+nWtg2Y5mapAAqEhtD8sgvp88xkrvnPWDq2rkvrxhedcj/n46++sCIFefyua3l+3Dc5HnvpsZt4ZmzO7NWZOg+rD8DyddtoeOtIruj5MgN6t6FwofNmiGuulC4TxZfffMfEjz6n/yOPM/jpxzl44AAAg4ePYvKn03nzvUmsWb2K2bNmkpGRwdhXXuSBR072e/f8UiA0hHo1K/DOZ4u4vNuLHEr7h8f6tMl6/PH/XEf6sXQ+/mZFAKM891T9e8ur/NmoEmCim52pp6o1VHUwcFSPf1qm432wfAiQ7LFtPVWt5fH4yftL4HXgDVWtC9wNFDnJev943JeTlGd4LGd4xBqCkw3LjK28qh7wsv3J6peNqo5X1Uaq2qhv376nWv2U2l5RmzWbdhC/L2fXSVx8MhXKlcxaLl82krj4ZOISkqlQ1qM8OpK4hGTi4pOzpbUz1z8ffPLtSjpffRkAPTs2Zcb3TjfJ1p17+TsukRpVsrdn4xL2U75sZNayZ13iE1OzuoLKlQknwX3t4uKTqeC5TXQEcQn7/VSj40JDQ+h01aV85tEwjo3fz6LVW0ncf5C0f44ye/EG6teskGPb+H2plCvt1qV0OAlJzr9mXMJ+KpSLzFrPqf9+p/zE18XPdaxaoQyVY0qz7KPH2fTVIMpHR7JkygDKli5Bg1qV+PCFXmz6ahA3XVOP1wZ2zer2ypSXjuWpbP57DwfS/uGSahdkK09MPkhEibCssVaZ70kg2/s1NDSE8OJFSEw+SGyOYxxBXLx/61ioUCEiIp3nrFn7EspXqMj27X8DTnYJoFixYlx3/Y1sWLeWQwcPsvXPLdz33950aXct69f+yhMP3c/GDeuy7Tcqqix7du/KWk6I30NUdFkiIiM5cCCVY8ecEzTi9+whKira3SaaPbt3A3Ds2DEOHkjNiu1MxcYnExufzHK3q/nL+WuoV7MiAD06NOHGlpfQ+5kPvW4bl5Cc/XhERxIbv985tsU9jq3He87zfep5bE3e4M9G1XzgFhGJBhCRUiLyb537qUAJAFVNAf4Ska7utiIil51qO1cEEOve73UW8f+buRzvCkRE6p1i/RNj9Ktbr2/ktesPYNbCtXRr3wSAJnWrkHIgjd17U5j380baXF6TyBJhRJYIo83lNZn380Z2700h9eBhmtStAkC39k34emHOs5TOlWoVo7Lut29Vh9//3gPAjt1JWeNKokuV4OLK0fwVm5ht2917U0g9cJgmdZx/w27tGvP1QueDfNaP6+jhvi492jfJVt6tXWMAmtSpTMqBw1ldS/50dZOL+f3vPcR6fCHOW7KJS6pfQFjhgoSGhtCyQXU2bt2TY9tZC9fRo31jty4edVy4nm43etYljd2JKcxbspk2TWscP/ZNazBvyckHiPvC+j92UfnaZ6jZ4Xlqdnie2PhkLu/+EnsSU6nV8fms8i/nr+GhkdP4asHabNvnpWPpTeWYUllfqJXKlaRGlbJs83KG148rttDlGuejr7tnXRauo7tbxy7XXMbC5Vuyyrte14BCBUOpHFOK6hWjWL5+m1/rkpS0j/R0p+MgducOdmzfRvnyFTh27BjJSU4X5bGjR1n800KqVr+I4iVK8O33i/li1jy+mDWPS+pexqjX3shx9l+tS+qwc8d24mJ3cvToEb6b8w1XtLoKEaFBoyb8MN/pRPj26xm0bO10hbdsdRXffu2crfjD/Lk0bNwUkdwP1Pe0JzGVnXuSs87Ibd2kBpv+2s21zWvxSK823PLQO6QdPup12xXrt1O9YhSVY0pRsEAoXds2YNZC53/ZObb1APfYuv/j2Y9tvaxje76xMVXe+S3frKobROQZYK6IhABHgfv+ZZOPgXfcwd234Iy/esvdR0H38ZznVcNXwGci0gmnoTMYp9swCfge8N3IyuP6A/8Tkd9wXsMfccZheaWqiSKyWETWAd/6c1xV0SKFuLppTe4f9lFW2f/dcgUA7362iNmL1tP2iktYP/M5Dh0+yt2DJwOQlHKIF96ZzaLJTkp+xPjZJKU4g24ffOFTxg/pQVjhgsxdvIE5i3I168UZmzj8Tlo2qk6ZyOL88c0Qhr79Lde3qM1FlaPJUGX7rn30H/EpACPfmcP4Id1Z/skTCMLTY7/K+nW3dOoAmnVzzg94cOQ0xg/uTlgRty6LnbqMnvAdk0f2oVenZmzftY8eAycAMHvRBtq2qM36Gc9y6PAR7h481cd17EnLhtUpE1mMP2Y9x9Dxs5k44xe6XlefT+dm775NTk1j7JQFLPrwERRlzuKNzHbjf/OZ23j3859ZtXEHoyfOZ/ILvejVqSnbdyXR48mJTl0Wb6Bti1qsn/60U5chHwPusX9vLos+fBiAEe/OzTr2vqtnzmM5ccbSXO/n/D6WOeuYlHKIVwbcTJmSxflizN389vtOOt4/jub1qvJY7zYcPZZOhioPjpyW9f/65Zi76Tf0I3btTeHpsV8xaUQvnuvXjl8378waLD1hxlLeH9qDddOfIWn/IXo+5RzjjVt38/m81az+7CmOHUvnoVGfndEkjYOefIzVK5eTnJxMp+uv5v/uuY/w8AheeXEEyUn7eKx/Py66uAavvfkOa1at4N233qBAgQJISAiPPzWI8IhI0tIO8fB9fTl27BgZGek0ano5HW+65V+fNyEhnpHPD+Ll18dRoEABHnniaR6+ry/pGRm073gTVatVB6Bf/0cY9ORjjP/fWC6uWYsOnW8GoH3nm3n+2YF07Xg94RERPP/C6FzX3ZtHRn3GB8PvpFDBUP7emUjfwVNYNPkxChcswNdvOSd2L1v7N/1HfMoFZcJ5c9Ad3NT/bdLTM3h41Gd89b9+hIaEMHHmUjZudTJpT4+dyaQXevPcfe34ddNOJkx33g8Tpi/h/aE9WTfjWefYPjnBJ3Uw54bk5RZhkNKw+vcHOga/Slv9BmENHwx0GH6XtnIMYY0eDnQYfpW24tX8cyyDvJ5pK8eQePDYqVfMw0oXK0BYg/6nXjGPS1s19uzSc6eh2qPf+rXx8OfLN/i9Dv6QN0dGGmOMMSZgLCHj3fk0+acxxhhjTJ5lmSpjjDHG5IplqryzTJUxxhhjjA9YpsoYY4wxuWOJKq8sU2WMMcYY4wOWqTLGGGNMrtiYKu8sU2WMMcYY4wOWqTLGGGNMrlimyjvLVBljjDHG+IBlqowxxhiTK5ap8s4yVcYYY4wxPmCZKmOMMcbkimWqvLNMlTHGGGOMD1imyhhjjDG5Y4kqryxTZYwxxhjjA5apMsYYY0yu2Jgq7yxTZYwxxhjjA5apMsYYY0yuWKbKO8tUGWOMMcb4gGWqjDHGGJMrlqnyzhpVxhhjjMkda1N5Zd1/xhhjjDE+YJkqY4wxxuSKdf95J/bCnHfsgBhjjDkb4u8nuKDv5379rto1/ma/18EfLFN1Hgpr0D/QIfhV2qqxhNW/P9Bh+F3a6jfyx7Fs+GCgw/C7tJVjCGs6INBh+FXaLy8F/bFMWzmGZVv3BzoMv2tSNcLvz2EJGe9sTJUxxhhjjA9YpsoYY4wxuWKZKu8sU2WMMcYY4wOWqTLGGGNMrlimyjvLVBljjDHG+IBlqowxxhiTO5ao8soyVcYYY4wxPmCZKmOMMcbkio2p8s4yVcYYY4wxPmCZKmOMMcbkimWqvLNMlTHGGGOMD1imyhhjjDG5Ypkq7yxTZYwxxhjjA5apMsYYY0zuWKLKK8tUGWOMMcb4gGWqjDHGGJMrNqbKO8tUGWOMMSZXVNWvt1MRkYoi8oOIbBCR9SLyoFteSkTmicgW929Jt1xEZKyI/CEiv4lIA4999XLX3yIivc7mdbFGlTHGGGPymmPAo6paG2gG3CcitYGBwHxVvQiY7y4D3ABc5N76Am+B0wgDngOaAk2A5zIbYmfCGlXGGGOMyZVAZ6pUdZeqrnLvpwIbgfJAJ2Ciu9pEoLN7vxPwoTqWApEicgHQFpinqvtUNQmYB1x/pq+LNaqMMcYYc14Rkb4issLj1vdf1q0C1Ad+Acqq6i73od1AWfd+eWCHx2Y73bKTlZ8RG6hujDHGmFzx90B1VR0PjD/VeiJSHPgceEhVU0TEcx8qIud0RL1lqowxxhiT54hIQZwG1RRV/cIt3uN26+H+jXfLY4GKHptXcMtOVn5GgjJTJSK9gUaqer8P99kZ+F1VN7jLzwM/qup3vnqOM/VA99b07nw5qsr6P3bRd/AU+tx0Ofd3a021ilFUuPpJEpMPet22e/smDPy/6wAY+e5cpny9DID6tSoyfnB3wooUZM6iDTz60ucAlAwvyqSRvakcU4ptcfvo8cQHJKem+b2O457rzg1X1iFhXyqNuo7IKr/39lbcfWtL0jOU2T+t4+kxM3Jse23zWowecAuhISFMmP4zoz+YB0DlmNJMGtmHUhHFWL1xO3c98yFHj6VTqGAB3hvak/q1KrFv/0F6PPE+23ft83sdwfuxfP2p22jZsDr7Dzivc9/npvDb7znf8+fzsRw36A5uaHkJCfsO0Oi2kcef/wWP5x/oPH/7VnUYdG87MjIyOJaeweMvf8nPa7bm2Gf9mhUYP6Q7YYULMmfxBh596Yt/3S/AywO60LZFbQ4dPkrfwVNYs2mnz+pYuFABvht3L4UKFaBAaAhffr+WYe/M5Z5bmnP/7S2pVrEMFa57jsT9hwBof+UlDOrblgxVjqWn8/irM/n517+91LM845+9zannz5t49JUZbj3DmDSsB5VjSrItLokeT08+Xs9HOtG2eU2nnkM/Yc3mM/6OyMbbcezSph5P972emheWpeWdr7Bq4/GelMf6tKF3p2akp2fw6Ogv+G7Jphz7rBxTikkv9HLfhzu469nJ7vswlPee70H9WhWd9+HAiVnvw9PZb27Nmf4xP8yeDqq0vr4z1990B599OI5VS35EQoTwiFL0fXQQJUtHsfG3lbw65DGiysUA0Kj5VdzU/f8A+PbLqSycPQNEqFilOv995FkKFSqc7bmOHjnC2y8P5q8tmygeHsH9Tw4nqqyzr5mfTGDhnJmEhITQ895HubTh5QD8tmIJk8a9TEZGBq2v70SHW8/qJLWzF+AZFcRJSb0HbFTVVzwemgn0Aka6f2d4lN8vIh/jDErfr6q7RGQOMMJjcPp1wJNnGpdlqk5fZ6B25oKqDjofGlQxURH0u70VLXqMptGtIwkNCaFr2wYsWfMXN97zP7bFJZ5025LhRXm67/VceecrtOz5Mk/3vZ7IEmEAjH3yVu4b9jF1Og2lWqUormteC3A+zBYs+526nYexYNnvPNbn2nNSz0lfLaXTff/LVnZlo4to37ouTW4bScNbhvPah/NzbBcSIrw28FY63f8m9W8eRtfrG1KzajkAhj/Yiden/ECdTkNISk2j903Oh1fvzpeTlJpGnU5DeH3KDwx/sJP/K8jJjyXAU6/NoNkdL9Lsjhe9NqjO92M56atldHpgXLayx3q3YcHy36l70zAWLP+dx3q3AeCHZb/T5PZRNOv2EvcM+Yg3n73d6z7HPnkr9w39mDqdh1Gtoke9TrLfti1qU61iFHU6D+P+YR8z9smuPq3jP0eOcf19b9O0x6s07fEq1zWrQZM6lVjy29/c+MB4tsVlb5j/sHwLTXq8QrOer3LPsGm8+ZT3eMY+3oX7XviMOreMolrFMlx3eQ2nnndezYIVf1D3lhdZsOIPHrvzKqeezWtSrWIZ6twyivtHfsbYx7v4rI7ejuP6P3Zx+4D3WbTqz2zlNS8sS9frGtCg6wt0fGAcYwZ2JSREONHw/h15fcoC6nQeRlJKGr07NwPc92FKGnU6D+P1KQsY3r9DrvabGzv+/pMfZk9nyGsTGP7mFNYsW8SeuB20u7kHI96ayvD/TaFe0yuYPvXdrG1q1KnH8P9NYfj/pmQ1qPbtjWfujE94fuxERo77mIyMdJYunJfj+RbOnUmx4iV4+f0vuL7zHXzy/hsAxG7bytKFcxk57mMGDBvDxDdeJCM9nYz0dCb+70UGDB3DqLc/YcmCOcRuy/lDI59pAfQErhaRNe7tRpzG1LUisgVo4y4DfANsBf4A3gH6AajqPmAosNy9Pe+WnZE82agSkR4issx9Ed8WkVAR6SMiv4vIMpwXO3PdCSJyi8fyAY/7T4jIWhH5VURGumX/FZHlbtnnIlJURJoDHYGX3Oes5rlfEblGRFa7+3pfRAq75X+LyBARWeU+VtMfr0eB0BDCChckNDSEsLCC7EpI4dfNO0+ZXbn28prM/2UzSSmHSE5NY/4vm7mueS3KlQmnRLEiLFv7NwBTv15Gh6suBaB9q7pMdjMgk79eRofWdf1RpRwWr/qTfe4v/Ex9u7Zk9AfzOHL0GAAJSQdybNe4ThX+3LGXv2MTOXosnWlzVtG+tVOXVo0v5ovvVgMw5atf6ND6MgDat76UKV/9AsAX362mdZMafqvXibwdy9Nxvh/LxatzHr/2rep4ff6DaUey1ikWVsjr2I1yZcIpUbwIy9ZtA2DqrOVZ259sv+1b1WHqrOUALFu3jYjiYZQrE+7LambFXrBAKAUKhKCq/Pp7HNt3JZ10XYBiRU5Sz9IlnOO3bjsAU79dSYdWdZz6XFmbybNWADB51go6tLrELb+Eqd+uBGDZuu1ElChCudIlfFI/b8dx89972LItPse67VvXZdrcVRw5ms62uH38uSOBxpdUzrFeq8YX8cX8XwGYcsLxysy2fjH/V1o3uThX+82NuB1/Ua3GJRQuUoTQ0ALUrNuA5Yt/IKxY8ax1/jmcBpy68ZaRns6RI/+Qnn6MI/8cpmSpMjnWWbVkIVe0aQdAk5ZXs37NclSVlUt/pFmr6yhYqBDR5cpTNqYCf/6+nj9/X0/ZmApEX1CeAgUL0qzVdaxc+uNZ1flsnQdn/y1SVVHVS1W1nnv7RlUTVfUaVb1IVdtkNpDcs/7uU9VqqlpXVVd47Ot9Va3u3j44m9clzzWqRKQWcBvQQlXrAelAD2AITmPqCjwySv+ynxtwTrFsqqqXAS+6D32hqo3dso3Af1T1Z5zU4QD3wP3psZ8iwATgNlWti9Oleq/HU+1V1QY4c2I8dsYVP4m4hP28Nul7fv9mCH/NHUZK6mHmLz29VHhMdCQ7dx//sI/dk0xMdCQxURHExicfL49PJiY6AoDo0iXYvdf5ot+9N4VoH31Yn4nqlaNpUb8aP374GHPffZCGtSvlWCcmOoKdezzrmET5qAhKRxZjf2oa6ekZWeWZdYyJjsh6XdLTM0g5kEbpyGJ+r8+/HcvB97Vj2SdP8OKjN1GoYM5e+7x4LP/t+TtedSlrPn+KL8b05Z4hH+XYNiYqgtg9yVnLmfX9t/3GREey03Ob+P3EREX4tE4hIcLSSQ+zffZzfL9sC8vX7/jX9Tu2qsOaTwbwxSt3cc+waTked47f/hNidhqC0aVKsDsxFYDdialEl3LrGRXu93qejvJREezc7RHHnv1Z/3uZcrwP45OJiYoEICYqMuu967wPD1M6sthp7Te3KlSuxu/r15Caksw/hw/z6/LF7EvYA8C0CW/yYM/2/PzDbG7ueXfWNn9sXMtT/brx0rMPsnOb85VQqkw0N97cg4fu7MgD3W4krGhx6jZsluP59iUmULqMc1JaaGgBihYtzoGU/SQlJlA6qmzWeiXLRJO0N4GkvQmU8igvVSaapMSEs6qz8Y8816gCrgEaAstFZI27/DCwQFUTVPUI8Mlp7KcN8IGqHoKsFCBAHRH5SUTWAt2BS06xnxrAX6r6u7s8EbjS4/HMwXMrgSreduB56uj48ac82SGbyBJhtG9dl1rth1C17TMUCyvE7Tc2ytU+zkYgr1RQIDSEUhHFuPLO0Tz16nQmv3hX4ILxgZMdy0FvfMVlXYZzRY+XKRlelEfd7ixfC/RVJzyff+YPv1Hv5hHc+uh7DLr3Rp/t198yMpRmPV+leodhNLqkIrWrlv3X9WcuXEe9217i1scnMOjutmf13P4+GyuYla90Ie263smLT/fnpWf7U6nqxYSEhALQtXc/xkz6muZXXc+8r5yGb5VqNXh14kxGvDmVazvcymvPPw7AwdQUVi5dyCsfTGfslG/45580Fn//bcDq5U+BzlSdr/Jio0qAiR7pvhrA4H9Z/xhuPUUkBCh0iv1PAO53s05DgCJnGe8/7t90TnJigKqOV9VGqtqob9+TTsXh1dVNa/B3bCJ7kw9w7FgG07//lWaXXnha28bFJ1Oh3PGJY8uXjSQuPpm4hP2Ud3/1A5SPjiTO/bUcn5ia1WVSrkw4CftScxWvL8XuSWb6/DUArFi/jYwMpUzJ4tnWiYvfT4WynnUsSWzCfhKTDxJRIozQ0JCs8sw6xsXvz3pdQkNDCC8edtKB/r50smOZmXU5cvQYH878hUZ1cmbk8uKxPJ3nX7z6Ty4sXzpHpjAuYT/ly0ZmLWfW99/2GxefTAXPbaIjiEs4ngXypf0HDrNw5Z9cd/np9fgvXvMXF5YvRemIotnKneN3PAvjxOz8P8TvS83q1itXukRW93dcQso5q+e/iU3YT4VyHnGUjcj638uU430YHUlcQjIAcQnJWe9d531YhMTkg6e13zPRum0nhr7+Ic+8NJ5iJcIpVyH7+6z5VdezfPH3AIQVK06RMOdY1WvSgvRjx0jdn8y6NcuIKhtDeGRJChQoQOPmV7Flw285nqtU6SgS9zqZsPT0Yxw6dIDi4RGULB1FopshA0jaG0/JMlGULBOVlTkDZ+xWydJRZ11n43t5sVE1H7hFRKIha4r51UArESntnmLpOeLzb5zMFjjjogq69+cBfUSkqMd+AEoAu9z9dPfYT6r72Ik2A1VEpLq73BNYeObVy50du5NoUrcKYUWcal3V5GI2/7XnFFs55i3ZRJtmNYksEUZkiTDaNKvJvCWb2L03hdSDh2lStwoA3do34esFawGY9eM6erRvAkCP9k34euFa31fqNH214DdaNXbGWVSvFE2hggXYe8K4qhXrt1G9UhSVY0pTsEAoXds2YNYC50PuxxW/06VNfQC6d2jK1275rIVr6d6hKQBd2tRn4fLfORdOdiw9x/10vOpSNvyxK8e2efFY5nz+dQBUrXB8DEq9mhUoXKhAjkbt7r0ppB44TJM6zliabu0aZ21/sv3O+nEd3do1BqBJncqkHDic1WD1hTKRxYgo7vwGK1K4ANc0uYjNf+cca5SpaoXSWffr1ShP4YIFss4MzLQ7MdU5fm5DutsNDfn6x/VOfX7aQI92Tla6R7tGfP3jBrd8Pd1ucD7ymtSp5NQz8dz/+Jm1cB1dr2tAoYKhVI4pRfWKUSxfvy3Hej+u2EKXa5zxjN09j9fCdXR3j2OXay5j4fItudpvbu1Pdjor9sbvZsXiH7i8dVt2x27PenzVkoXEVKgCQPK+vVnZlD83r0c1g+LhEZSOKsefm9bxz+HDzhm8a5YTU7FKjueq3+xKFn03C4BlP31P7csaISI0aNaSpQvncvTIEeJ3x7I7bgfVLr6EqhfXZnfcDuJ3x3Ls6FGWLpxLg2Ytz7rOZ8MyVd7luSkVVHWDiDwDzHUzT0eB+3CyVUuAZGCNxybvADNE5FdgNnDQ3c9sEakHrBCRIzhnBjwFPIszK2uC+zezIfUx8I6I9AeyBr6r6mER6QNME5ECOGcPZD89xo+Wr9vGl/PXsGTK4xxLT+fXzbG898XP9Lv9Sh7p1YaypUuw/JOBzF60gX5DP6JBrYr83y1X0G/oRySlHOKFd+ewaLIz1GvEO7NJSnE+1B984VP3dPVCzP15A3MWOx/Yoz+Yx+RRfejVuRnbdyXR44mzGtN32ia+0JuWDS+iTGRx/pg9lKHjvmHi9CW8Pbg7K6Y9xZGj6fzfoEkAXBAVwZuDunHTA2+Rnp7Bw6M+5as37yM0RJg4Yykbt+4G4OkxM5g0sg/P9WvPr5t3MGH6EgAmTP+Z94fdyboZz5GUcpCeA89NHU92LGe8cQ9lIosjAr/9HssDw53e7bx0LCcOv5OWjao7x++bIQx9+1tGT/iOySP70KtTM7bv2kePgRMAuOmay+jWrjFHj6Vz+J+j9HxyYtZ+lk4dQLNuLzn1Gjkta6qIuYs96nWS/c5etIG2LWqzfsazHDp8hLsHT/VpHcuVCeedQbcRGhJCSIjw+fxf+XbxRvrd2oJHerambKkSLJ/yCLN/3kS/EZ9x01V16XZjQ44ey3Dq+czk4/Wc9DDNer7q1PPFLxk/yJlSYe6STcz52RlnN3riD0we0YNeHRuzfVcyPZ52/v9nL95E2+a1WP/5QKeeQz/1WR29HceklEO8MuBmypQszhdj7ua333fS8f5xbNy6m8/nrWb1Z09x7Fg6D436jIwM58vyyzF302/oR+zam8LTY79i0ohePNevHb9u3nn8fThjKe8P7cG66c+QtP8QPZ9y/g/+bb9nY+ywJziQkkJogVB69RtAseIlePe1YezauY0QCaF0dDn6POBcRm75ou+ZP+tzQkJDKVSoCP0GDkdEqF6zDo2vuIZnH+hJSGgoVarV4KobbgLg8w/f5sKLa9Gg2ZW0atuRcS89x6N3daF4iXDuGzgccMZ2NW3ZhoF330ZIaCi9+j1OSKjTDXnnvQN46Zn+ZKRncOV1HahQudpZ19n4nuTlFmGQ0rAG/QMdg1+lrRpLWH2fTSF23kpb/Qb54lg2fDDQYfhd2soxhDUdEOgw/Crtl5eC/limrRzDsq3nviv0XGtSNeLs5pg4DcW6fuDXxsPBaX38Xgd/yIvdf8YYY4wx55081/1njDHGmACzTi6vLFNljDHGGOMDlqkyxhhjTK7YeGzvLFNljDHGGOMDlqkyxhhjTK5Ypso7y1QZY4wxxviAZaqMMcYYkzuWqfLKGlXGGGOMyR3NCHQE5yXr/jPGGGOM8QHLVBljjDEmd6z7zyvLVBljjDHG+IBlqowxxhiTOzamyivLVBljjDHG+IBlqowxxhiTOzamyivLVBljjDHG+IBlqowxxhiTOzamyivLVBljjDHG+IBlqowxxhiTO5ap8soyVcYYY4wxPmCZKmOMMcbkjp3955VlqowxxhhjfMAyVcYYY4zJHRtT5ZVlqowxxhhjfEDU+kXzPRHpq6rjAx2HP+WHOkL+qKfVMXjkh3oGax3Drn/Fr42HtNmPiD/37y+WqTIAfQMdwDmQH+oI+aOeVsfgkR/qmR/qaFw2psoYY4wxuWNjqryyTJUxxhhjjA9YpsoABF1/vxf5oY6QP+ppdQwe+aGewVlHG4/tlQ1UN8YYY0yuhF07yr8D1ec9YQPVjTHGGGPyK+v+M8YYY0zuWC+XV5apMsYYY4zxActU5UMiUlpVEwMdx7kkIiWBiqr6W6BjMcYbEQlV1fRAx2HMabEpFbyyTFX+tFREponIjSKSJwcDng4RWSAi4SJSClgFvCMirwQ6LpN7IlJURJ4VkXfc5YtEpH2g4/KxLSLykojUDnQg/iIio06nLK8TkYtF5B0RmSsi32feAh2X8T9rVOVPF+Oc5tsT54N8hIhcHOCY/CFCVVOALsCHqtoUaBPgmHxGRFJFJMXLLVVEUgIdn499APwDXO4uxwLDAheOX1wG/A68KyJLRaSviIQHOigfu9ZL2Q3nPAr/m4bzQ+4ZYIDHLXio+veWR1n3Xz6kzjwa84B5InIVMBnoJyK/AgNVdUlAA/SdAiJyAXAr8HSgg/E1VS0R6BjOoWqqepuI3AGgqoeCLcuqqqnAOzgZ1VbAVOBVEfkMGKqqfwQ0wLMgIvcC/YCqIuLZBV8CWByYqPzqmKq+FeggzLlnjap8SERKAz1wMlV7gAeAmUA9nF9YFwYsON8aAswBFqnqchGpCmwJcEx+IyLRQJHMZVXdHsBwfO2IiIQBCiAi1XAyV0FDREKBdkAfoArwMjAFaAl8g5NhzqumAt8CLwADPcpTVXVfYELyq69EpB/wJR7/p0FVVxtT5ZU1qvKnJcAkoLOq7vQoXyEi4wIUkz/sUtVLMxdUdWswjqkSkY44X8AxQDxQGdgIXBLIuHxsMDAbqCgiU4AWOI2PYLIF+AF4SVV/9ij/TESuDFBMPqGq+4H9wB1u47EszvdPcREpHmQ/AAB6uX89u/wUqBqAWMw5ZDOq50MiIpoPDryIrFLVBqcqy+vcbturge9Utb7bpdtDVf8T4NB8ys2wNgMEWKqqewMckk+5jYsDgY7Dn0TkfpwG8h4gM9Whnj9+TN4Q1up5/86ovnBQnuzet0xV/tRQRJ7GyWgUwPmSCpoPNhG5HGgORInIIx4PhQOhgYnKr46qaqKIhIhIiKr+ICKvBTooXxKR+ap6DTDLS1mwGOFlmNh+YIWqzghAPP7wEFAj2Kd0EZGCwL1AZoZxAfC2qh4NWFDmnLBGVf40BSctvZbjvxaDSSGgOM7/t+dg7hTgloBE5F/JIlIc+BGYIiLxwMEAx+QTIlIEKAqUcecay2x1hAPlAxaYfxQBauKMawS4GfgLuExErlLVhwIVmA/twGkoBru3gILAm+5yT7fs/wIWka/ZmCqvrFGVPyWo6sxAB+EvqrpQRBYBl6rqkEDHcw50Ag4DDwPdgQjg+YBG5Dt342Q3YoCVHG9UpQBvBCgmf7kUaJE5AaiIvAX8BFyB8wMoGGwFFojILLIP4A62sY6NVfUyj+Xv3W56E+SsUZU/PSci7wLzyf7B9kXgQvItVU0XkZhAx3EuqKpnVmpiwALxA1UdA4wRkQdU9fVAx+NnJXEyrJmZnGJAKfd/OVjOdNzu3gq5t2CVLiLVVPVPAPfM4+CaLd8yVV5Zoyp/6oPTzVAQj8GiQNA0qlxrRGQmTndKVsMjmBqPACLSBRgFRONkcjLHyAXNxJGq+rqI1AFqk33aiA8DF5XPvYjzP7sA5xheiTPOqhjwXSAD85XMzLGIFFXVQ4GOx48GAD+IyFacY1mZ4Dtb1XhhZ//lQyKyWVVrBDoOfxORD7wUq6redc6D8SMR+QPooKobAx2Lv4jIc0BrnEbVNzizcC9S1aAaI+dOVtvEXVyuqnGBjMfX3JNI3gOKq2olEbkMuFtV+wU4NJ8TkcJA5ufsZlUNlmwjAGFXPOvfs/8WDbWz/0ye8bOI1FbVDYEOxJ9UNb/8MtwTzA0q1y04l3FZrap9RKQszpUAgk1jnMk+wckiB1WjCngNaIsz2TCq+mten4PLk4hcrarfu9ljT9VFJOiy5CYna1TlT81wuhn+whlTFVRTKmQSkQrA6zgTRYIz6PfBEyY8DQYrROQTYDpBOkYOSFPVDBE55l4PLx6oGOigfElERuI0qqa4Rf1F5HJVfSqAYfmcqu44YeqIYBpr1Ar4Hujg5bHgGmJhY6q8skZV/nR9oAM4Rz7AuTxGV3e5h1vm7aKueVk4cAi4zqMsuD7AnYZjJM618VYCB3CuDBBMbgTqqTrfViIyEVgNBFOjaoeINAfUncvpQZzZ/4OCqj7n3n1eVf/yfExEguXyX+Zf2JiqfMody5DZzfCTqgbd6b4iskZV652qzOQtIlIFCFfV3061bl7iXmi4deb14USkFLAgmDLIIlIGGAO0wcmQz8XJHgfVZKAnuZrDSlVtGKiYfC2s+VP+HVP18wgbU2XyBhF5EPgvxzMZk0VkfBCesp4oIj2Aj9zlO4Cg+vCG/NHN6Tl7uqr+fWJZkHgBWC0iP3D87L+B/75J3uJeWqh7oOPwFxGpiXPNzYgTxlWF43HWalCw7j+vrFGVP/0HaJo5v5GIjMLpSgm2RtVdOHV61V1eTHCe1hy03Zz5aUZ1Vf3InU6hsVv0hKruDmBIPud2gT0AVMHj+0dVOwYqJh+rAbQHIsk+rioV54esCXLWqMqfhOyDQ9M5/mUVNFR1GxAsH9b/JkpVPaePmCAiDwUqGB/zNqO64nxJBcWPABE58QLfmRnGGBGJUdVV5zomP5qOM6XCVwThJbLcazTOcE8wCLYxf9nZ0CGvrFGVP30A/CIiX7rLnXE+6IKKO4vxGJyzHRUnG/ewqm4NaGC+F7TdnB4zqg8CXlPVFBF5FmhA8AxUf/lfHlPg6nMVyDlwWFXHBjqIc2C1iNyH0xXoOVltUM2RZ3KyRlU+pKqviMhCjo/B6aOqqwMZk59MBf4H3OQu347T8GgasIj8w7ObU4GfCb5uzltU9XkRuQKnkTEa5wK1ef5YqupVgY7hHBrjTuQ6l+zTfwRTNg5gErAJZ06u53HGkQXNWY6Ajak6CWtU5V9rgF24/wMiUklVtwc0It8rqqqTPJYni8iAgEXjJ/mkmzOzu7od8I6qzhKRYYEMyNfcKQbuxRmgDrAAeFtVjwYsKN+rC/TEaRh7XiIrmLJxANVVtauIdFLViSIyFecEEhPkrFGVD4nIA8BzwB6Oj6dSIGhO3XZ9KyIDgY9x6ncb8I17qjqZp67nVSLyuKq+KCKv49QvG1XtH4Cw/CVWRN7GGXw/yr0ESEiAY/K1t3Cux/mmu9zTLfu/gEXke12Bqqp6JNCB+FlmQzjZvWblbpxrcwYPG1PllTWq8qcHgRrBNjeMF7e6f+/meKNDcLoBFagaiKB8KLM7YUVAozg3bsWZtHa0qia718gLtqxjY1W9zGP5exEJtvnj1uGcGRcf4Dj8bbx7tuozOJfkKQ4MCmxI5lywRlX+tAPYH+ggzoEngNknDG4eGizjN1T1K/fuIVWd5vmYiHT1skmepaqH8JghXlV34XRfB5N0Eammqn9C1okWwXQJF3AaVJtEZDnZx1QFVfe1qr7r3v2RvP/jzTsbU+WVNaryp63AAhGZRfYPtlcCF5JfPKOqnwbj4OYTPAlMO40yc34bAPwgIltxMqqVCb4TDp479Sp5n4iMAF5U1WR3uSTwqKo+E9DAjN9Zoyp/2u7eCrm3YBXUg5tF5Aac68WVFxHP09TDgWOBicqcKVWdLyIX4UwgCbBZVf/5t23yGlVdGOgYzpEbPC+ErapJInIjTndgcLAxVV5ZoyofUtUhACJS3F0+ENiI/CbYBzfH4Yyn6ogzMWamVODhgERkzlZDjs82Xk9EUNUPAxuS77iXbhmFM2hb3JuqanhAA/O9UBEpnNkoFpEwoHCAYzLngDWq8iH3bJRJQCl3eS9wp6quD2hgvhfUg5vdi2D/KiJTg+y0+3xJRCYB1XCmO8nMsioQNI0q4EWgg6oG15xNOU0B5otI5pUO+gATAxiP79mYKq+sUZU/jQceUdUfAESkNfAO0DyAMflcPhncDFBFRF4AapN99ubgHCAbvBoBtVWDul9lTz5oUKGqo0TkNyDzgt9DVXVOIGMy54Y1qvKnYpkNKgBVXSAixQIZkDkrH+AMAH4VuArnV3EwdXPmF+uAcgRnwz/TChH5BOcagJ4nyXxx0i3yKFX9Fvg20HH4TVC3/c+cNaryp63uFAOZs433wDkj0ORNYe4gZ3FnVx8sIiuxeXHymjLABhFZRvBONxAOHAKu8yhTPDLKwUBEUjk+N14hnEldDwbh2DFzAmtU5U93AUNwPsgU5/IJdqHPvOsfEQkBtojI/UAszmSDJm8ZHOgA/E1Vg22KCK9UtUTmfRERoBPOhd2Dh42p8kqCu/vemOAnIo1xZlePBIYCEThz5CwNZFzGZMpnl1TySkRWq2r9QMfhK2H17/dr4yFt9Rviz/37i2Wq8iERmQd0PWFiuo9VtW1AAzNnRFWXu3cPEHyTReYbQT7dQH66pFLmscwUgnMSwuEAheMXebXR42+WqcqHvP1iCrZfUfmJiFyMM1VEZTx+KKnq1QELyuSaiPxB/phuIOh5TKUAzkS8f+NMQBzs1zzM9yxTlT9liEglVd0OICKV8ZKSN3nGNGAczrQYwXatuPwk6KcbEJEonGtynjj9R1D9AMgvY8dMTtaoyp+eBhaJyEKcLoaWQN/AhmTOwjFVfSvQQZgz49FVlB+mG5gCfIJz6ah7gF5AQkAj8qGTjRnLlB/GjuV31v2XD7lno/QEHsA5C/A3oJyqLgtoYOaMiMhgIB74kuxfxvsCFZM5fSd0FZ1IVTVozswVkZWq2lBEflPVS92y5araONCx+YKI9HLvtsDJxn3iLncFNqjqPQEJzJwz1qjKh0TkLSADuFpVa7kD1ecGywdbfiMif3kpVptR3ZxvRGSpqjYTkTnAWJzrV36mqtUCHJpPichS4ApVPeYuFwR+UtXgmlbB5GCzLudPTVX1PtyzUVQ1CWeCOpPHuPNTDVTVC0+4WYMqjxGRiSIS6bFcUkTeD2BI/jBMRCKAR4HHgHeBhwIakX+UxJnoNFNxt8wEORtTlT8dFZFQ3L5/d/CozeSWB6lqhogM4Hg3g8m7Ls2c5gScHzsiEmxn5Cap6n5gP84llRCRFoENyS9GAqtEZAHOuNUryQeTuxrLVOVXY3HG30SLyHBgETAisCGZs/CdiDwmIhVFpFTmLdBBmVwLcbviAXCPYbD98H39NMvyugk4l4m6FPgcaMXxubpMEAu2N6w5Dao6xb023DU4v6I6B/up3EHuNvfvfR5lClgXYN7yMrBERKa5y12B4QGMx2dE5HKgORAlIo94PBQOhAYmKr96Eyf7H6aqM93G8ueAjVsNctaoyqdUdROwKdBxmLOnqhcGOgZz9lT1QxFZAWTO2dRFVTcEMiYfKoQzrqgAUMKjPAW4JSAR+VdTVW0gIqshqyvXxq3mA9aoMiaPE5GiwCNAJVXtKyIXATVU9esAh2ZyrxRwUFU/EJEoEblQVb2d3ZmnqOpCEVmEM25sSKDjOQds3Go+ZWOqjMn7PgCO4HSvAMQCwwIXjjkTIvIczmzjT7pFBYHJgYvIt1Q1HYgJdBzniI1bzacsU2VM3ldNVW8TkTsAVPWQO8GryVtuAuoDqwBUNU5ESvz7JnnOGhGZiXNppYOZhUE2a7yNW83HrFFlTN53RETCON7VUA2PmdVNnnFEVVVEMo9jsUAH5AdFgESOjxsD5/82qBpVYONW8ytrVBmT9w0GZgMVRWQKziUy7IKuec+nIvI2ECki/wXuwrlIdtCwCw2bYGeNKmPyOFWd63Y1NMPpanhQVfcGOCyTe1HAZzhnxNXAmeeoTUAj8jERuRh4CyirqnVE5FKgo6raGEATFOzaf8bkcSIyX1WvOVWZOb+JyCpVbXBCWdaFh4OBiCwEBgBvq2p9t2ydqtYJbGTG+IZlqozJo0SkCFAUKONOLpg5OD0cKB+wwEyuiMi9QD+gqoj85vFQCWBxYKLym6KquuyE8yiOBSoYY3zNGlXG5F1341yMNgZYyfFGVQrwRoBiMrk3FfgWeAEY6FGeqqr7AhOS3+x1T6TIHIx/C7ArsCEZ4zvW/WdMHiciD6hqMF4/zQQZEakKjMeZUy0J+AvorqrbAhqYMT5ijSpjgoCINAeq4JF9VtUPAxaQMV5kzhDvThcRoqqpwTJrvDFgjSpj8jwRmQRUA9YA6W6xqmr/gAVljBcnGYy/UlUbBiomY3zJxlQZk/c1Amqr/UIy5ykRqQlcAkSISBePh8JxJgQ1JihYo8qYvG8dUA4b8GvOXzWA9kAk0MGjPBX4byACMsYfrPvPmDxORH4A6gHL8Lg8jap2DFRMxngjIper6pJAx2GMv1ijypg8TkRaeStX1YXnOhZj/o2IROFkpqqQ/aSKuwIVkzG+ZI0qY4wx54SI/Az8hDOvWuZJFajq5wELyhgfskaVMXmUiCxS1StEJBV3MsXMh3DO/gsPUGjGeCUia1S1XqDjMMZfrFFljDHmnBCRYcDPqvpNoGMxxh+sUWWMMeaccLOqRYEjwFEsq2qCjE2pYIwx5lyJALoDF6rq8yJSCbggwDEZ4zOWqTLGGHNOiMhbQAZwtarWEpGSwFxVbRzg0IzxCctUGWOMOVeaqmoDEVkNoKpJIlIo0EEZ4yshgQ7AGGNMvnFUREJxz1Z1563KCGxIxviONaqMMcacK2OBL4FoERkOLAJGBDYkY3zHxlQZY4w5Z9yLK1+Dc+bffFXdGOCQjPEZa1QZY4wxxviAdf8ZY4wxxviANaqMMcYYY3zAGlXGmH8lIukiskZE1onINBEpehb7miAit7j33xWR2v+ybmsRaX4Gz/G3iJQ53fIT1jmQy+caLCKP5TZGY0xwskaVMeZU0lS1nqrWwbm8yD2eD4rIGc13p6r/p6ob/mWV1kCuG1XGGBMo1qgyxuTGT0B1N4v0k4jMBDaISKiIvCQiy0XkNxG5G0Acb4jIZhH5DojO3JGILBCRRu7960VklYj8KiLzRaQKTuPtYTdL1lJEokTkc/c5lotIC3fb0iIyV0TWi8i7OGeV/SsRmS4iK91t+p7w2Ktu+Xx3HiVEpJqIzHa3+ck9g80YY7KxGdWNMafFzUjdAMx2ixoAdVT1L7dhsl9VG4tIYWCxiMwF6gM1gNpAWWAD8P4J+40C3gGudPdVSlX3icg44ICqjnbXmwq8qqqL3GvGzQFqAc8Bi9xrybUD/nMa1bnLfY4wYLmIfK6qiUAxYIWqPiwig9x93w+MB+5R1S0i0hR4E7j6DF5GY0wQs0aVMeZUwkRkjXv/J+A9nG65Zar6l1t+HXBp5ngpnAvnXgRcCXykqulAnIh872X/zYAfM/elqvtOEkcboLZIViIqXESKu8/Rxd12logknUad+ovITe79im6siTize3/ilk8GvnCfozkwzeO5C5/Gcxhj8hlrVBljTiVNVet5FriNi4OeRcADqjrnhPVu9GEcIUAzVT3sJZbTJiKtcRpol6vqIRFZABQ5yerqPm/yia+BMcacyMZUGWN8YQ5wr4gUBBCRi0WkGPAjcJs75uoC4Cov2y4FrhSRC91tS7nlqUAJj/XmAg9kLohIPffuj0A3t+wGoOQpYo0AktwGVU2cTFmmECAz29YNp1sxBfhLRLq6zyEictkpnsMYkw9Zo8oY4wvv4oyXWiUi64C3cTLhXwJb3Mc+BJacuKGqJgB9cbrafuV499tXwE2ZA9WB/kAjdyD8Bo6fhTgEp1G2HqcbcPspYp0NFBCRjcBInEZdpoNAE7cOVwPPu+Xdgf+48a0HOp3Ga2KMyWfsMjXGGGOMMT5gmSpjjDHGGB+wRpUxxhhjjA9Yo8oYY4wxxgesUWWMMcYY4wPWqDLGGGOM8QFrVBljjDHG+IA1qowxxhhjfOD/AdCqK4ZDWlfgAAAAAElFTkSuQmCC\n"},"metadata":{"needs_background":"light"}}]},{"cell_type":"code","source":"","metadata":{"papermill":{"duration":1.130234,"end_time":"2021-10-19T16:17:07.996667","exception":false,"start_time":"2021-10-19T16:17:06.866433","status":"completed"},"tags":[]},"execution_count":null,"outputs":[]}]}