geoffrey-hinton-kg-embeddings-pykeen.ipynb

{"cells":[{"cell_type":"code","execution_count":1,"metadata":{"_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","execution":{"iopub.execute_input":"2024-10-12T21:22:31.483067Z","iopub.status.busy":"2024-10-12T21:22:31.482689Z","iopub.status.idle":"2024-10-12T21:22:31.842968Z","shell.execute_reply":"2024-10-12T21:22:31.842078Z","shell.execute_reply.started":"2024-10-12T21:22:31.483026Z"},"trusted":true},"outputs":[{"name":"stdout","output_type":"stream","text":["/kaggle/input/genaik2025-aylin/results_clean_v2.txt\n","/kaggle/input/genaik2025-aylin/summary_clean_v2.txt\n","/kaggle/input/genaik2025-aylin/cypher_query_v2.txt\n","/kaggle/input/genaik2025-aylin/summary_clean.txt\n","/kaggle/input/genaik2025-aylin/results_clean.txt\n","/kaggle/input/genaik2025-aylin/cypher_query.txt\n","/kaggle/input/genaik2025-aylin/neo4j_query_table_data_2024-10-12.csv\n","/kaggle/input/genaik2025-aylin/Geoffrey_Hinton_KG.csv\n"]}],"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","\n","import numpy as np # linear algebra\n","import 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","\n","import os\n","for 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"]},{"cell_type":"markdown","metadata":{},"source":["# Sources:\n","- https://pykeen.readthedocs.io/en/stable/reference/models.html\n","- https://wasit7.medium.com/tutorial-knowledge-graph-embedding-with-pykeen-22d3b7847cea\n","- https://medium.com/@shreyash.pandey.katni/comparison-of-knowledge-graph-embeddings-kge-models-2bcd52ca3f76"]},{"cell_type":"markdown","metadata":{},"source":["# Installing Libraries and Modules"]},{"cell_type":"code","execution_count":2,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T21:22:35.361533Z","iopub.status.busy":"2024-10-12T21:22:35.360747Z","iopub.status.idle":"2024-10-12T21:23:15.742893Z","shell.execute_reply":"2024-10-12T21:23:15.741886Z","shell.execute_reply.started":"2024-10-12T21:22:35.361493Z"},"trusted":true},"outputs":[{"name":"stdout","output_type":"stream","text":["\u001b[31mERROR: pip's dependency resolver does not currently take into account all the packages that are installed. This behaviour is the source of the following dependency conflicts.\n","beatrix-jupyterlab 2024.66.154055 requires jupyterlab~=3.6.0, but you have jupyterlab 4.2.5 which is incompatible.\n","ydata-profiling 4.10.0 requires scipy<1.14,>=1.4.1, but you have scipy 1.14.1 which is incompatible.\u001b[0m\u001b[31m\n","\u001b[0m"]}],"source":["!pip install --upgrade --quiet pykeen\n","!pip install --upgrade --quiet matplotlib seaborn"]},{"cell_type":"markdown","metadata":{},"source":["# Importing Libraries and Modules"]},{"cell_type":"code","execution_count":20,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T21:46:21.133511Z","iopub.status.busy":"2024-10-12T21:46:21.132591Z","iopub.status.idle":"2024-10-12T21:46:21.139121Z","shell.execute_reply":"2024-10-12T21:46:21.138043Z","shell.execute_reply.started":"2024-10-12T21:46:21.133456Z"},"trusted":true},"outputs":[],"source":["from pykeen.triples import TriplesFactory\n","from pykeen.pipeline import pipeline\n","from pykeen.evaluation import RankBasedEvaluator\n","from typing import List\n","from sklearn.manifold import TSNE\n","from sklearn.decomposition import PCA\n","import matplotlib.pyplot as plt\n","import pykeen.nn\n","import torch"]},{"cell_type":"markdown","metadata":{},"source":["# Preparing Data"]},{"cell_type":"code","execution_count":4,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T21:23:26.493302Z","iopub.status.busy":"2024-10-12T21:23:26.492737Z","iopub.status.idle":"2024-10-12T21:23:26.520282Z","shell.execute_reply":"2024-10-12T21:23:26.519380Z","shell.execute_reply.started":"2024-10-12T21:23:26.493261Z"},"trusted":true},"outputs":[{"data":{"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>head</th>\n","      <th>head_type</th>\n","      <th>relation</th>\n","      <th>tail</th>\n","      <th>tail_type</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>0</td>\n","      <td>neural networks</td>\n","      <td>invention</td>\n","      <td>developed</td>\n","      <td>computer vision</td>\n","      <td>invention</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>1</td>\n","      <td>neural networks</td>\n","      <td>invention</td>\n","      <td>developed</td>\n","      <td>natural language processing</td>\n","      <td>invention</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>2</td>\n","      <td>John J. Hopfield</td>\n","      <td>person</td>\n","      <td>bornIn</td>\n","      <td>Chicago</td>\n","      <td>birthPlace</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>3</td>\n","      <td>John J. Hopfield</td>\n","      <td>person</td>\n","      <td>worksAt</td>\n","      <td>California Institute of Technology (Caltech)</td>\n","      <td>institution</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>4</td>\n","      <td>John J. Hopfield</td>\n","      <td>person</td>\n","      <td>worksAt</td>\n","      <td>Princeton University</td>\n","      <td>institution</td>\n","    </tr>\n","  </tbody>\n","</table>\n","</div>"],"text/plain":["   Unnamed: 0              head  head_type   relation  \\\n","0           0   neural networks  invention  developed   \n","1           1   neural networks  invention  developed   \n","2           2  John J. Hopfield     person     bornIn   \n","3           3  John J. Hopfield     person    worksAt   \n","4           4  John J. Hopfield     person    worksAt   \n","\n","                                           tail    tail_type  \n","0                               computer vision    invention  \n","1                   natural language processing    invention  \n","2                                       Chicago   birthPlace  \n","3  California Institute of Technology (Caltech)  institution  \n","4                          Princeton University  institution  "]},"execution_count":4,"metadata":{},"output_type":"execute_result"}],"source":["kg_df = pd.read_csv(\"/kaggle/input/genaik2025-aylin/Geoffrey_Hinton_KG.csv\")\n","kg_df.head()"]},{"cell_type":"code","execution_count":5,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T21:23:30.136827Z","iopub.status.busy":"2024-10-12T21:23:30.136405Z","iopub.status.idle":"2024-10-12T21:23:30.185201Z","shell.execute_reply":"2024-10-12T21:23:30.184335Z","shell.execute_reply.started":"2024-10-12T21:23:30.136787Z"},"trusted":true},"outputs":[{"data":{"text/plain":["array([['AI applications', 'characteristics',\n","        'large language models like ChatGPT'],\n","       ['AI applications', 'characteristics', 'self-driving cars'],\n","       ['Geoffrey Hinton', 'advocates',\n","        'ethical implications of advanced technologies'],\n","       ['Geoffrey Hinton', 'advocates', 'responsible AI practices'],\n","       ['Geoffrey Hinton', 'bornAt', 'December 6, 1947'],\n","       ['Geoffrey Hinton', 'bornIn', 'London'],\n","       ['Geoffrey Hinton', 'coAuthored', 'David Rumelhart'],\n","       ['Geoffrey Hinton', 'coAuthored', 'Ronald J. Williams'],\n","       ['Geoffrey Hinton', 'coFounded', 'Vector Institute'],\n","       ['Geoffrey Hinton', 'collaboratedWith', 'Alex Krizhevsky'],\n","       ['Geoffrey Hinton', 'collaboratedWith', 'Brendan Frey'],\n","       ['Geoffrey Hinton', 'collaboratedWith', 'John J. Hopfield'],\n","       ['Geoffrey Hinton', 'collaboratedWith', 'Richard Zemel'],\n","       ['Geoffrey Hinton', 'collaboratedWith', 'Ruslan Salakhutdinov'],\n","       ['Geoffrey Hinton', 'collaboratedWith', 'Vector Institute'],\n","       ['Geoffrey Hinton', 'collaboratedWith', 'Yann LeCun'],\n","       ['Geoffrey Hinton', 'collaboratedWith', 'Yoshua Bengio'],\n","       ['Geoffrey Hinton', 'developed', 'AI applications'],\n","       ['Geoffrey Hinton', 'developed', 'Helmholtz machines'],\n","       ['Geoffrey Hinton', 'developed', 'distributed representations'],\n","       ['Geoffrey Hinton', 'developed', 'mixtures of experts'],\n","       ['Geoffrey Hinton', 'developed', 'neural networks'],\n","       ['Geoffrey Hinton', 'developed', 'the product of experts'],\n","       ['Geoffrey Hinton', 'developed', 'time delay neural networks'],\n","       ['Geoffrey Hinton', 'earnedDegree', 'Ph.D'],\n","       ['Geoffrey Hinton', 'earnedDegree', 'experimental psychology'],\n","       ['Geoffrey Hinton', 'earnedDegreeFrom', 'University of Cambridge'],\n","       ['Geoffrey Hinton', 'hasAward', 'IEEE Frank Rosenblatt Award'],\n","       ['Geoffrey Hinton', 'hasAward', 'Nobel Prize in Physics'],\n","       ['Geoffrey Hinton', 'hasAward', 'Princess of Asturias Award'],\n","       ['Geoffrey Hinton', 'hasAward', 'Rumelhart Prize'],\n","       ['Geoffrey Hinton', 'hasAward', 'Turing Award'],\n","       ['Geoffrey Hinton', 'hasNationality', 'British-Canadian'],\n","       ['Geoffrey Hinton', 'hasRelation', 'George Boole'],\n","       ['Geoffrey Hinton', 'hasRelation', 'Mary Everest Boole'],\n","       ['Geoffrey Hinton', 'hasRole', 'advisory role'],\n","       ['Geoffrey Hinton', 'hasRole', 'advisory roles'],\n","       ['Geoffrey Hinton', 'hasRole',\n","        'cognitive psychologist and computer scientist'],\n","       ['Geoffrey Hinton', 'hasRole',\n","        'pursuit of knowledge and innovation'],\n","       ['Geoffrey Hinton', 'hasRole', 'research'],\n","       ['Geoffrey Hinton', 'hasStudent', 'Ilya Sutskever'],\n","       ['Geoffrey Hinton', 'hasStudent', 'Yann LeCun'],\n","       ['Geoffrey Hinton', 'invented', 'Boltzmann machine'],\n","       ['Geoffrey Hinton', 'workedWith', 'Google Brain'],\n","       ['Geoffrey Hinton', 'workedWith', 'Stanford University'],\n","       ['Geoffrey Hinton', 'workedWith',\n","        'field of artificial intelligence'],\n","       ['Geoffrey Hinton', 'worksAt', 'Carnegie Mellon University'],\n","       ['Geoffrey Hinton', 'worksAt', 'Google Brain'],\n","       ['Geoffrey Hinton', 'worksAt', 'University College London'],\n","       ['Geoffrey Hinton', 'worksAt',\n","        'University of California, San Diego'],\n","       ['Geoffrey Hinton', 'worksAt', 'University of Toronto'],\n","       ['Geoffrey Hinton', 'worksAt', 'Vector Institute'],\n","       ['George Boole', 'developed', 'algebra of logic'],\n","       ['Howard Everest Hinton', 'is', 'father'],\n","       ['Howard Everest Hinton', 'was', 'entomologist'],\n","       ['Ilya Sutskever', 'bornAt', 'December 8, 1986'],\n","       ['Ilya Sutskever', 'bornIn', 'Nizhny Novgorod'],\n","       ['Ilya Sutskever', 'coFounded', 'OpenAI'],\n","       ['Ilya Sutskever', 'coInventing', 'AlexNet'],\n","       ['Ilya Sutskever', 'developed', 'GPT-3'],\n","       ['Ilya Sutskever', 'developed', 'GPT-4'],\n","       ['Ilya Sutskever', 'earnedDegree', 'Ph.D'],\n","       ['Ilya Sutskever', 'hasNationality', 'Canadian-Israeli-Russian'],\n","       ['Ilya Sutskever', 'hasRole', 'Chief Scientist'],\n","       ['Ilya Sutskever', 'worksAt', 'OpenAI'],\n","       ['John J. Hopfield', 'bornAt', 'July 15, 1933'],\n","       ['John J. Hopfield', 'bornIn', 'Chicago'],\n","       ['John J. Hopfield', 'collaboratedWith', 'Geoffrey Hinton'],\n","       ['John J. Hopfield', 'developed', 'Hopfield network'],\n","       ['John J. Hopfield', 'earnedDegree', 'Ph.D'],\n","       ['John J. Hopfield', 'earnedDegree', 'bachelor’s degree'],\n","       ['John J. Hopfield', 'earnedDegreeFrom', 'Swarthmore College'],\n","       ['John J. Hopfield', 'hasAward', 'Nobel Prize in Physics'],\n","       ['John J. Hopfield', 'hasNationality', 'American'],\n","       ['John J. Hopfield', 'worksAt', 'AT&T Bell Laboratories'],\n","       ['John J. Hopfield', 'worksAt',\n","        'California Institute of Technology (Caltech)'],\n","       ['John J. Hopfield', 'worksAt', 'Princeton University'],\n","       ['Ronald J. Williams', 'invented', 'backpropagation algorithm'],\n","       ['University of Edinburgh', 'earnedDegreeFrom', 'AI'],\n","       ['Yann LeCun', 'bornAt', 'July 8, 1960'],\n","       ['Yann LeCun', 'bornIn', 'Soisy-sous-Montmorency'],\n","       ['Yann LeCun', 'coAuthored', 'Geoffrey Hinton'],\n","       ['Yann LeCun', 'coAuthored', 'Yoshua Bengio'],\n","       ['Yann LeCun', 'developed', 'DjVu image compression technology'],\n","       ['Yann LeCun', 'hasAward', 'Turing Award'],\n","       ['Yann LeCun', 'hasNationality', 'French-American'],\n","       ['Yann LeCun', 'hasRole', 'computer scientist'],\n","       ['Yann LeCun', 'hasRole', 'key figure in deep learning'],\n","       ['Yann LeCun', 'workedWith',\n","        'convolutional neural networks (CNNs)'],\n","       ['Yann LeCun', 'worksAt',\n","        'Courant Institute of Mathematical Sciences at New York University'],\n","       ['Yann LeCun', 'worksAt', 'Meta (formerly Facebook)'],\n","       ['neural networks', 'developed', 'computer vision'],\n","       ['neural networks', 'developed', 'natural language processing']],\n","      dtype='<U65')"]},"execution_count":5,"metadata":{},"output_type":"execute_result"}],"source":["triples_factory = TriplesFactory.from_labeled_triples(\n","    triples=kg_df[['head', 'relation', 'tail']].values,\n",")\n","\n","# Display the first few triples\n","triples_factory.triples"]},{"cell_type":"code","execution_count":6,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T21:23:33.707344Z","iopub.status.busy":"2024-10-12T21:23:33.706701Z","iopub.status.idle":"2024-10-12T21:23:33.726874Z","shell.execute_reply":"2024-10-12T21:23:33.726003Z","shell.execute_reply.started":"2024-10-12T21:23:33.707304Z"},"trusted":true},"outputs":[{"name":"stdout","output_type":"stream","text":[" 1:AI applications 3:characteristics 74:large language models like ChatGPT\n"," 1:AI applications 3:characteristics 81:self-driving cars\n","22:Geoffrey Hinton 0:advocates 69:ethical implications of advanced technologies\n","22:Geoffrey Hinton 0:advocates 80:responsible AI practices\n","22:Geoffrey Hinton 1:bornAt  16:December 6, 1947\n","22:Geoffrey Hinton 2:bornIn  33:London \n","22:Geoffrey Hinton 4:coAuthored 15:David Rumelhart\n","22:Geoffrey Hinton 4:coAuthored 43:Ronald J. Williams\n","22:Geoffrey Hinton 5:coFounded 55:Vector Institute\n","22:Geoffrey Hinton 7:collaboratedWith  3:Alex Krizhevsky\n","22:Geoffrey Hinton 7:collaboratedWith  7:Brendan Frey\n","22:Geoffrey Hinton 7:collaboratedWith 30:John J. Hopfield\n","22:Geoffrey Hinton 7:collaboratedWith 42:Richard Zemel\n","22:Geoffrey Hinton 7:collaboratedWith 45:Ruslan Salakhutdinov\n","22:Geoffrey Hinton 7:collaboratedWith 55:Vector Institute\n","22:Geoffrey Hinton 7:collaboratedWith 56:Yann LeCun\n","22:Geoffrey Hinton 7:collaboratedWith 57:Yoshua Bengio\n","22:Geoffrey Hinton 8:developed  1:AI applications\n","22:Geoffrey Hinton 8:developed 25:Helmholtz machines\n","22:Geoffrey Hinton 8:developed 67:distributed representations\n","22:Geoffrey Hinton 8:developed 75:mixtures of experts\n","22:Geoffrey Hinton 8:developed 77:neural networks\n","22:Geoffrey Hinton 8:developed 82:the product of experts\n","22:Geoffrey Hinton 8:developed 83:time delay neural networks\n","22:Geoffrey Hinton 9:earnedDegree 39:Ph.D   \n","22:Geoffrey Hinton 9:earnedDegree 70:experimental psychology\n","22:Geoffrey Hinton10:earnedDegreeFrom 52:University of Cambridge\n","22:Geoffrey Hinton11:hasAward 28:IEEE Frank Rosenblatt Award\n","22:Geoffrey Hinton11:hasAward 37:Nobel Prize in Physics\n","22:Geoffrey Hinton11:hasAward 40:Princess of Asturias Award\n","22:Geoffrey Hinton11:hasAward 44:Rumelhart Prize\n","22:Geoffrey Hinton11:hasAward 49:Turing Award\n","22:Geoffrey Hinton12:hasNationality  8:British-Canadian\n","22:Geoffrey Hinton13:hasRelation 23:George Boole\n","22:Geoffrey Hinton13:hasRelation 34:Mary Everest Boole\n","22:Geoffrey Hinton14:hasRole 58:advisory role\n","22:Geoffrey Hinton14:hasRole 59:advisory roles\n","22:Geoffrey Hinton14:hasRole 63:cognitive psychologist and computer scientist\n","22:Geoffrey Hinton14:hasRole 78:pursuit of knowledge and innovation\n","22:Geoffrey Hinton14:hasRole 79:research\n","22:Geoffrey Hinton15:hasStudent 29:Ilya Sutskever\n","22:Geoffrey Hinton15:hasStudent 56:Yann LeCun\n","22:Geoffrey Hinton16:invented  6:Boltzmann machine\n","22:Geoffrey Hinton19:workedWith 24:Google Brain\n","22:Geoffrey Hinton19:workedWith 47:Stanford University\n","22:Geoffrey Hinton19:workedWith 72:field of artificial intelligence\n","22:Geoffrey Hinton20:worksAt 11:Carnegie Mellon University\n","22:Geoffrey Hinton20:worksAt 24:Google Brain\n","22:Geoffrey Hinton20:worksAt 50:University College London\n","22:Geoffrey Hinton20:worksAt 51:University of California, San Diego\n","22:Geoffrey Hinton20:worksAt 54:University of Toronto\n","22:Geoffrey Hinton20:worksAt 55:Vector Institute\n","23:George Boole 8:developed 60:algebra of logic\n","27:Howard Everest Hinton17:is      71:father \n","27:Howard Everest Hinton18:was     68:entomologist\n","29:Ilya Sutskever 1:bornAt  17:December 8, 1986\n","29:Ilya Sutskever 2:bornIn  36:Nizhny Novgorod\n","29:Ilya Sutskever 5:coFounded 38:OpenAI \n","29:Ilya Sutskever 6:coInventing  4:AlexNet\n","29:Ilya Sutskever 8:developed 20:GPT-3  \n","29:Ilya Sutskever 8:developed 21:GPT-4  \n","29:Ilya Sutskever 9:earnedDegree 39:Ph.D   \n","29:Ilya Sutskever12:hasNationality 10:Canadian-Israeli-Russian\n","29:Ilya Sutskever14:hasRole 13:Chief Scientist\n","29:Ilya Sutskever20:worksAt 38:OpenAI \n","30:John J. Hopfield 1:bornAt  31:July 15, 1933\n","30:John J. Hopfield 2:bornIn  12:Chicago\n","30:John J. Hopfield 7:collaboratedWith 22:Geoffrey Hinton\n","30:John J. Hopfield 8:developed 26:Hopfield network\n","30:John J. Hopfield 9:earnedDegree 39:Ph.D   \n","30:John J. Hopfield 9:earnedDegree 61:bachelor’s degree\n","30:John J. Hopfield10:earnedDegreeFrom 48:Swarthmore College\n","30:John J. Hopfield11:hasAward 37:Nobel Prize in Physics\n","30:John J. Hopfield12:hasNationality  5:American\n","30:John J. Hopfield20:worksAt  2:AT&T Bell Laboratories\n","30:John J. Hopfield20:worksAt  9:California Institute of Technology (Caltech)\n","30:John J. Hopfield20:worksAt 41:Princeton University\n","43:Ronald J. Williams16:invented 62:backpropagation algorithm\n","53:University of Edinburgh10:earnedDegreeFrom  0:AI     \n","56:Yann LeCun 1:bornAt  32:July 8, 1960\n","56:Yann LeCun 2:bornIn  46:Soisy-sous-Montmorency\n","56:Yann LeCun 4:coAuthored 22:Geoffrey Hinton\n","56:Yann LeCun 4:coAuthored 57:Yoshua Bengio\n","56:Yann LeCun 8:developed 18:DjVu image compression technology\n","56:Yann LeCun11:hasAward 49:Turing Award\n","56:Yann LeCun12:hasNationality 19:French-American\n","56:Yann LeCun14:hasRole 64:computer scientist\n","56:Yann LeCun14:hasRole 73:key figure in deep learning\n","56:Yann LeCun19:workedWith 66:convolutional neural networks (CNNs)\n","56:Yann LeCun20:worksAt 14:Courant Institute of Mathematical Sciences at New York University\n","56:Yann LeCun20:worksAt 35:Meta (formerly Facebook)\n","77:neural networks 8:developed 65:computer vision\n","77:neural networks 8:developed 76:natural language processing\n"]}],"source":["training = triples_factory\n","validation = triples_factory\n","testing = triples_factory\n","\n","d = training\n","id_to_entity = {v: k for k, v in d.entity_to_id.items()}\n","id_to_relation = {v: k for k, v in d.relation_to_id.items()}\n","\n","for i in d.map_triples(d.triples):\n","    s,p,o =int(i[0]), int(i[1]), int(i[2])\n","    h,r,t = id_to_entity[s], id_to_relation[p], id_to_entity[o]\n","    print(f\"{s:2d}:{h:7}{p:2d}:{r:7} {o:2d}:{t:7}\")"]},{"cell_type":"markdown","metadata":{},"source":["# Models"]},{"cell_type":"markdown","metadata":{},"source":["## TransE"]},{"cell_type":"code","execution_count":8,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T21:27:13.680147Z","iopub.status.busy":"2024-10-12T21:27:13.679762Z","iopub.status.idle":"2024-10-12T21:29:33.935580Z","shell.execute_reply":"2024-10-12T21:29:33.934639Z","shell.execute_reply.started":"2024-10-12T21:27:13.680109Z"},"trusted":true},"outputs":[{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"aea1e912283c4d2b97e7c03b421c6d54","version_major":2,"version_minor":0},"text/plain":["Training epochs on cuda:0:   0%|          | 0/1000 [00:00<?, ?epoch/s]"]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"1cbb31e57f6a4a488891326a0e6547fc","version_major":2,"version_minor":0},"text/plain":["Evaluating on cuda:0:   0%|          | 0.00/93.0 [00:00<?, ?triple/s]"]},"metadata":{},"output_type":"display_data"},{"data":{"text/plain":["<Axes: title={'center': 'Losses Plot'}, xlabel='Epoch', ylabel='softplus Loss'>"]},"execution_count":8,"metadata":{},"output_type":"execute_result"},{"data":{"image/png":"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","text/plain":["<Figure size 640x480 with 1 Axes>"]},"metadata":{},"output_type":"display_data"}],"source":["result_TransE = pipeline(\n","    model='TransE',\n","    loss=\"softplus\",\n","    training=training,\n","    testing=testing,\n","    validation=validation,\n","    model_kwargs=dict(embedding_dim=3, random_seed=42),  # Increase the embedding dimension\n","    optimizer_kwargs=dict(lr=0.1),  # Adjust the learning rate\n","    training_kwargs=dict(num_epochs=1000, use_tqdm_batch=False),  # Increase the number of epochs\n",")\n","\n","# The trained model is stored in the pipeline result\n","model_TransE = result_TransE.model\n","\n","#plot loss\n","result_TransE.plot_losses()"]},{"cell_type":"markdown","metadata":{},"source":["### Visualize TransE embeddings"]},{"cell_type":"code","execution_count":11,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T21:30:46.872663Z","iopub.status.busy":"2024-10-12T21:30:46.872232Z","iopub.status.idle":"2024-10-12T21:30:46.878053Z","shell.execute_reply":"2024-10-12T21:30:46.877057Z","shell.execute_reply.started":"2024-10-12T21:30:46.872624Z"},"trusted":true},"outputs":[],"source":["entity_representation_modules: List['pykeen.nn.Representation'] = model_TransE.entity_representations\n","relation_representation_modules: List['pykeen.nn.Representation'] = model_TransE.relation_representations"]},{"cell_type":"code","execution_count":17,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T21:44:42.313418Z","iopub.status.busy":"2024-10-12T21:44:42.312611Z","iopub.status.idle":"2024-10-12T21:44:42.318521Z","shell.execute_reply":"2024-10-12T21:44:42.317426Z","shell.execute_reply.started":"2024-10-12T21:44:42.313378Z"},"trusted":true},"outputs":[{"name":"stdout","output_type":"stream","text":["Number of relation embeddings: 21\n"]}],"source":["print(f\"Number of relation embeddings: {relation_embeddings.shape[0]}\")"]},{"cell_type":"code","execution_count":18,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T21:45:23.129839Z","iopub.status.busy":"2024-10-12T21:45:23.129423Z","iopub.status.idle":"2024-10-12T21:45:23.736552Z","shell.execute_reply":"2024-10-12T21:45:23.735556Z","shell.execute_reply.started":"2024-10-12T21:45:23.129800Z"},"trusted":true},"outputs":[],"source":["# Extract entity and relation embeddings\n","entity_embeddings = entity_representation_modules[0](indices=None).cpu().detach().numpy()  # All entity embeddings\n","relation_embeddings = relation_representation_modules[0](indices=None).cpu().detach().numpy()  # All relation embeddings\n","\n","# Reduce dimensionality of entity embeddings\n","tsne = TSNE(n_components=2, perplexity=5, random_state=42)\n","entity_embeddings_2d = tsne.fit_transform(entity_embeddings)\n","\n","# Optionally reduce dimensionality of relation embeddings\n","relation_embeddings_2d = tsne.fit_transform(relation_embeddings)"]},{"cell_type":"code","execution_count":26,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T22:01:02.029324Z","iopub.status.busy":"2024-10-12T22:01:02.028469Z","iopub.status.idle":"2024-10-12T22:01:02.786215Z","shell.execute_reply":"2024-10-12T22:01:02.785228Z","shell.execute_reply.started":"2024-10-12T22:01:02.029271Z"},"trusted":true},"outputs":[{"data":{"image/png":"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","text/plain":["<Figure size 500x500 with 1 Axes>"]},"metadata":{},"output_type":"display_data"},{"data":{"image/png":"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","text/plain":["<Figure size 500x500 with 1 Axes>"]},"metadata":{},"output_type":"display_data"}],"source":["# Plot entity embeddings\n","plt.figure(figsize=(5, 5))\n","plt.scatter(entity_embeddings_2d[:, 0], entity_embeddings_2d[:, 1], alpha=0.7)\n","plt.title('Entity Embeddings (2D t-SNE) TransE')\n","plt.xlabel('Dimension 1')\n","plt.ylabel('Dimension 2')\n","plt.show()\n","\n","# Plot relation embeddings\n","plt.figure(figsize=(5, 5))\n","plt.scatter(relation_embeddings_2d[:, 0], relation_embeddings_2d[:, 1], alpha=0.7, color='r')\n","plt.title('Relation Embeddings (2D t-SNE) TransE')\n","plt.xlabel('Dimension 1')\n","plt.ylabel('Dimension 2')\n","plt.show()"]},{"cell_type":"code","execution_count":22,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T21:48:04.307940Z","iopub.status.busy":"2024-10-12T21:48:04.307300Z","iopub.status.idle":"2024-10-12T21:48:04.314616Z","shell.execute_reply":"2024-10-12T21:48:04.313563Z","shell.execute_reply.started":"2024-10-12T21:48:04.307896Z"},"trusted":true},"outputs":[],"source":["pca = PCA(n_components=2)\n","entity_embeddings_2d = pca.fit_transform(entity_embeddings)"]},{"cell_type":"code","execution_count":25,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T22:00:47.057937Z","iopub.status.busy":"2024-10-12T22:00:47.057523Z","iopub.status.idle":"2024-10-12T22:00:47.389998Z","shell.execute_reply":"2024-10-12T22:00:47.389086Z","shell.execute_reply.started":"2024-10-12T22:00:47.057898Z"},"trusted":true},"outputs":[{"data":{"image/png":"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","text/plain":["<Figure size 500x500 with 1 Axes>"]},"metadata":{},"output_type":"display_data"}],"source":["# Plot entity embeddings\n","plt.figure(figsize=(5, 5))\n","plt.scatter(entity_embeddings_2d[:, 0], entity_embeddings_2d[:, 1], alpha=0.7, c='blue')\n","plt.title('Entity Embeddings (2D PCA) TransE')\n","plt.xlabel('PCA Dimension 1')\n","plt.ylabel('PCA Dimension 2')\n","plt.grid(True)\n","plt.show()"]},{"cell_type":"markdown","metadata":{},"source":["### Computing some evaluation metrics"]},{"cell_type":"code","execution_count":13,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T21:32:43.252542Z","iopub.status.busy":"2024-10-12T21:32:43.251820Z","iopub.status.idle":"2024-10-12T21:32:43.338667Z","shell.execute_reply":"2024-10-12T21:32:43.337690Z","shell.execute_reply.started":"2024-10-12T21:32:43.252502Z"},"trusted":true},"outputs":[{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"365f4ae4baaa4f919e32de075f1ee853","version_major":2,"version_minor":0},"text/plain":["Evaluating on cuda:0:   0%|          | 0.00/93.0 [00:00<?, ?triple/s]"]},"metadata":{},"output_type":"display_data"},{"name":"stdout","output_type":"stream","text":["Hits@1: 0.3387096774193548\n","Hits@3: 0.6827956989247311\n","Hits@5: 0.9086021505376344\n","Hits@10: 0.989247311827957\n","Mean Reciprocal Rank: 0.5523144602775574\n"]}],"source":["# The result object also contains the evaluation metrics result.metric_results.to_df()\n","\n","# Create an evaluator\n","evaluator = RankBasedEvaluator()\n","\n","# Evaluate the model\n","metrics = evaluator.evaluate(result_TransE.model,\n","                             testing.mapped_triples,\n","                             additional_filter_triples=[training.mapped_triples, validation.mapped_triples])\n","\n","# Print the metrics\n","print(f\"Hits@1: {metrics.get_metric('hits@1')}\")\n","print(f\"Hits@3: {metrics.get_metric('hits@3')}\")\n","print(f\"Hits@5: {metrics.get_metric('hits@5')}\")\n","print(f\"Hits@10: {metrics.get_metric('hits@10')}\")\n","print(f\"Mean Reciprocal Rank: {metrics.get_metric('mean_reciprocal_rank')}\")"]},{"cell_type":"markdown","metadata":{},"source":["## RotatE"]},{"cell_type":"code","execution_count":24,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T21:55:23.538125Z","iopub.status.busy":"2024-10-12T21:55:23.537168Z","iopub.status.idle":"2024-10-12T21:57:55.913576Z","shell.execute_reply":"2024-10-12T21:57:55.912688Z","shell.execute_reply.started":"2024-10-12T21:55:23.538082Z"},"trusted":true},"outputs":[{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"55e913995fc84a1eaeff5e69a501eed3","version_major":2,"version_minor":0},"text/plain":["Training epochs on cuda:0:   0%|          | 0/1000 [00:00<?, ?epoch/s]"]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"1887478f0c35443fbebe1022cce8e8d0","version_major":2,"version_minor":0},"text/plain":["Evaluating on cuda:0:   0%|          | 0.00/93.0 [00:00<?, ?triple/s]"]},"metadata":{},"output_type":"display_data"},{"data":{"text/plain":["<Axes: title={'center': 'Losses Plot'}, xlabel='Epoch', ylabel='softplus Loss'>"]},"execution_count":24,"metadata":{},"output_type":"execute_result"},{"data":{"image/png":"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","text/plain":["<Figure size 640x480 with 1 Axes>"]},"metadata":{},"output_type":"display_data"}],"source":["result_RotatE = pipeline(\n","    model='RotatE',\n","    loss=\"softplus\",\n","    training=training,\n","    testing=testing,\n","    validation=validation,\n","    model_kwargs=dict(embedding_dim=3, random_seed=42),  # Increase the embedding dimension\n","    optimizer_kwargs=dict(lr=0.1),  # Adjust the learning rate\n","    training_kwargs=dict(num_epochs=1000, use_tqdm_batch=False),  # Increase the number of epochs\n",")\n","\n","# The trained model is stored in the pipeline result\n","model_RotatE = result_RotatE.model\n","\n","#plot loss\n","result_RotatE.plot_losses()"]},{"cell_type":"code","execution_count":27,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T22:01:12.583077Z","iopub.status.busy":"2024-10-12T22:01:12.582225Z","iopub.status.idle":"2024-10-12T22:01:12.587506Z","shell.execute_reply":"2024-10-12T22:01:12.586548Z","shell.execute_reply.started":"2024-10-12T22:01:12.583037Z"},"trusted":true},"outputs":[],"source":["entity_representation_modules_RotatE: List['pykeen.nn.Representation'] = model_RotatE.entity_representations\n","relation_representation_modules_RotatE: List['pykeen.nn.Representation'] = model_RotatE.relation_representations"]},{"cell_type":"markdown","metadata":{},"source":["PCA and T-SNE could not be plotted for the embeddings of RotatE because of the complexity of embeddings."]},{"cell_type":"code","execution_count":28,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T22:01:18.094910Z","iopub.status.busy":"2024-10-12T22:01:18.094066Z","iopub.status.idle":"2024-10-12T22:01:18.100373Z","shell.execute_reply":"2024-10-12T22:01:18.099409Z","shell.execute_reply.started":"2024-10-12T22:01:18.094870Z"},"trusted":true},"outputs":[],"source":["# Extract entity and relation embeddings\n","entity_embeddings_RotatE = entity_representation_modules_RotatE[0](indices=None).cpu().detach().numpy()  \n","relation_embeddings_RotatE = relation_representation_modules_RotatE[0](indices=None).cpu().detach().numpy()  "]},{"cell_type":"code","execution_count":29,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T22:01:20.198211Z","iopub.status.busy":"2024-10-12T22:01:20.197547Z","iopub.status.idle":"2024-10-12T22:01:20.202845Z","shell.execute_reply":"2024-10-12T22:01:20.201880Z","shell.execute_reply.started":"2024-10-12T22:01:20.198144Z"},"trusted":true},"outputs":[{"name":"stdout","output_type":"stream","text":["Number of relation embeddings: 21\n"]}],"source":["print(f\"Number of relation embeddings: {relation_embeddings_RotatE.shape[0]}\")"]},{"cell_type":"code","execution_count":32,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T22:06:09.921692Z","iopub.status.busy":"2024-10-12T22:06:09.921310Z","iopub.status.idle":"2024-10-12T22:06:09.936359Z","shell.execute_reply":"2024-10-12T22:06:09.935135Z","shell.execute_reply.started":"2024-10-12T22:06:09.921656Z"},"trusted":true},"outputs":[{"data":{"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>Side</th>\n","      <th>Rank_type</th>\n","      <th>Metric</th>\n","      <th>Value</th>\n","    </tr>\n","  </thead>\n","  <tbody>\n","    <tr>\n","      <th>0</th>\n","      <td>head</td>\n","      <td>optimistic</td>\n","      <td>adjusted_geometric_mean_rank_index</td>\n","      <td>0.988112</td>\n","    </tr>\n","    <tr>\n","      <th>1</th>\n","      <td>tail</td>\n","      <td>optimistic</td>\n","      <td>adjusted_geometric_mean_rank_index</td>\n","      <td>0.991331</td>\n","    </tr>\n","    <tr>\n","      <th>2</th>\n","      <td>both</td>\n","      <td>optimistic</td>\n","      <td>adjusted_geometric_mean_rank_index</td>\n","      <td>0.989712</td>\n","    </tr>\n","    <tr>\n","      <th>3</th>\n","      <td>head</td>\n","      <td>realistic</td>\n","      <td>adjusted_geometric_mean_rank_index</td>\n","      <td>0.988112</td>\n","    </tr>\n","    <tr>\n","      <th>4</th>\n","      <td>tail</td>\n","      <td>realistic</td>\n","      <td>adjusted_geometric_mean_rank_index</td>\n","      <td>0.991331</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>220</th>\n","      <td>tail</td>\n","      <td>realistic</td>\n","      <td>adjusted_hits_at_k</td>\n","      <td>1.000000</td>\n","    </tr>\n","    <tr>\n","      <th>221</th>\n","      <td>both</td>\n","      <td>realistic</td>\n","      <td>adjusted_hits_at_k</td>\n","      <td>0.993885</td>\n","    </tr>\n","    <tr>\n","      <th>222</th>\n","      <td>head</td>\n","      <td>pessimistic</td>\n","      <td>adjusted_hits_at_k</td>\n","      <td>0.987792</td>\n","    </tr>\n","    <tr>\n","      <th>223</th>\n","      <td>tail</td>\n","      <td>pessimistic</td>\n","      <td>adjusted_hits_at_k</td>\n","      <td>1.000000</td>\n","    </tr>\n","    <tr>\n","      <th>224</th>\n","      <td>both</td>\n","      <td>pessimistic</td>\n","      <td>adjusted_hits_at_k</td>\n","      <td>0.993885</td>\n","    </tr>\n","  </tbody>\n","</table>\n","<p>225 rows × 4 columns</p>\n","</div>"],"text/plain":["     Side    Rank_type                              Metric     Value\n","0    head   optimistic  adjusted_geometric_mean_rank_index  0.988112\n","1    tail   optimistic  adjusted_geometric_mean_rank_index  0.991331\n","2    both   optimistic  adjusted_geometric_mean_rank_index  0.989712\n","3    head    realistic  adjusted_geometric_mean_rank_index  0.988112\n","4    tail    realistic  adjusted_geometric_mean_rank_index  0.991331\n","..    ...          ...                                 ...       ...\n","220  tail    realistic                  adjusted_hits_at_k  1.000000\n","221  both    realistic                  adjusted_hits_at_k  0.993885\n","222  head  pessimistic                  adjusted_hits_at_k  0.987792\n","223  tail  pessimistic                  adjusted_hits_at_k  1.000000\n","224  both  pessimistic                  adjusted_hits_at_k  0.993885\n","\n","[225 rows x 4 columns]"]},"execution_count":32,"metadata":{},"output_type":"execute_result"}],"source":["result_RotatE.metric_results.to_df()"]},{"cell_type":"code","execution_count":33,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T22:06:28.332868Z","iopub.status.busy":"2024-10-12T22:06:28.332497Z","iopub.status.idle":"2024-10-12T22:06:28.419100Z","shell.execute_reply":"2024-10-12T22:06:28.418187Z","shell.execute_reply.started":"2024-10-12T22:06:28.332831Z"},"trusted":true},"outputs":[{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"0f55b6ca433e42249d5199bfe6907789","version_major":2,"version_minor":0},"text/plain":["Evaluating on cuda:0:   0%|          | 0.00/93.0 [00:00<?, ?triple/s]"]},"metadata":{},"output_type":"display_data"},{"name":"stdout","output_type":"stream","text":["Hits@1: 0.7419354838709677\n","Hits@3: 0.9139784946236559\n","Hits@5: 0.978494623655914\n","Hits@10: 0.9946236559139785\n","Mean Reciprocal Rank: 0.8407471179962158\n"]}],"source":["evaluator = RankBasedEvaluator()\n","\n","metrics = evaluator.evaluate(result_RotatE.model,\n","                             testing.mapped_triples,\n","                             additional_filter_triples=[training.mapped_triples, validation.mapped_triples])\n","\n","print(f\"Hits@1: {metrics.get_metric('hits@1')}\")\n","print(f\"Hits@3: {metrics.get_metric('hits@3')}\")\n","print(f\"Hits@5: {metrics.get_metric('hits@5')}\")\n","print(f\"Hits@10: {metrics.get_metric('hits@10')}\")\n","print(f\"Mean Reciprocal Rank: {metrics.get_metric('mean_reciprocal_rank')}\")"]},{"cell_type":"markdown","metadata":{},"source":["## SE"]},{"cell_type":"code","execution_count":null,"metadata":{"trusted":true},"outputs":[],"source":["result_SE = pipeline(\n","    model='SE',\n","    loss=\"softplus\",\n","    training=training,\n","    testing=testing,\n","    validation=validation,\n","    model_kwargs=dict(embedding_dim=3, random_seed=42),  # Increase the embedding dimension\n","    optimizer_kwargs=dict(lr=0.1),  # Adjust the learning rate\n","    training_kwargs=dict(num_epochs=1000, use_tqdm_batch=False),  # Increase the number of epochs\n",")"]},{"cell_type":"code","execution_count":40,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T22:18:52.556952Z","iopub.status.busy":"2024-10-12T22:18:52.556212Z","iopub.status.idle":"2024-10-12T22:18:52.970582Z","shell.execute_reply":"2024-10-12T22:18:52.969650Z","shell.execute_reply.started":"2024-10-12T22:18:52.556909Z"},"trusted":true},"outputs":[{"data":{"text/plain":["<Axes: title={'center': 'Losses Plot'}, xlabel='Epoch', ylabel='softplus Loss'>"]},"execution_count":40,"metadata":{},"output_type":"execute_result"},{"data":{"image/png":"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","text/plain":["<Figure size 640x480 with 1 Axes>"]},"metadata":{},"output_type":"display_data"}],"source":["# The trained model is stored in the pipeline result\n","model_SE = result_SE.model\n","\n","#plot loss\n","result_SE.plot_losses()"]},{"cell_type":"code","execution_count":41,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T22:19:02.603552Z","iopub.status.busy":"2024-10-12T22:19:02.602799Z","iopub.status.idle":"2024-10-12T22:19:02.608271Z","shell.execute_reply":"2024-10-12T22:19:02.607237Z","shell.execute_reply.started":"2024-10-12T22:19:02.603508Z"},"trusted":true},"outputs":[],"source":["entity_representation_modules_SE: List['pykeen.nn.Representation'] = model_SE.entity_representations\n","relation_representation_modules_SE: List['pykeen.nn.Representation'] = model_SE.relation_representations"]},{"cell_type":"code","execution_count":42,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T22:19:04.790122Z","iopub.status.busy":"2024-10-12T22:19:04.789221Z","iopub.status.idle":"2024-10-12T22:19:04.795751Z","shell.execute_reply":"2024-10-12T22:19:04.794717Z","shell.execute_reply.started":"2024-10-12T22:19:04.790081Z"},"trusted":true},"outputs":[],"source":["entity_embeddings_SE = entity_representation_modules_SE[0](indices=None).cpu().detach().numpy()  \n","relation_embeddings_SE = relation_representation_modules_SE[0](indices=None).cpu().detach().numpy()  "]},{"cell_type":"code","execution_count":43,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T22:19:07.658371Z","iopub.status.busy":"2024-10-12T22:19:07.657975Z","iopub.status.idle":"2024-10-12T22:19:07.663741Z","shell.execute_reply":"2024-10-12T22:19:07.662666Z","shell.execute_reply.started":"2024-10-12T22:19:07.658332Z"},"trusted":true},"outputs":[{"name":"stdout","output_type":"stream","text":["Number of relation embeddings: 21\n"]}],"source":["print(f\"Number of relation embeddings: {relation_embeddings_SE.shape[0]}\")"]},{"cell_type":"code","execution_count":44,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T22:19:11.058785Z","iopub.status.busy":"2024-10-12T22:19:11.058389Z","iopub.status.idle":"2024-10-12T22:19:11.147837Z","shell.execute_reply":"2024-10-12T22:19:11.146966Z","shell.execute_reply.started":"2024-10-12T22:19:11.058747Z"},"trusted":true},"outputs":[{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"a0becc08d6ce465eab1ec779ab988f38","version_major":2,"version_minor":0},"text/plain":["Evaluating on cuda:0:   0%|          | 0.00/93.0 [00:00<?, ?triple/s]"]},"metadata":{},"output_type":"display_data"},{"name":"stdout","output_type":"stream","text":["Hits@1: 0.6720430107526881\n","Hits@3: 0.9731182795698925\n","Hits@5: 1.0\n","Hits@10: 1.0\n","Mean Reciprocal Rank: 0.8257169127464294\n"]}],"source":["evaluator = RankBasedEvaluator()\n","\n","metrics = evaluator.evaluate(result_SE.model,\n","                             testing.mapped_triples,\n","                             additional_filter_triples=[training.mapped_triples, validation.mapped_triples])\n","\n","print(f\"Hits@1: {metrics.get_metric('hits@1')}\")\n","print(f\"Hits@3: {metrics.get_metric('hits@3')}\")\n","print(f\"Hits@5: {metrics.get_metric('hits@5')}\")\n","print(f\"Hits@10: {metrics.get_metric('hits@10')}\")\n","print(f\"Mean Reciprocal Rank: {metrics.get_metric('mean_reciprocal_rank')}\")"]},{"cell_type":"markdown","metadata":{},"source":["## DistMult"]},{"cell_type":"code","execution_count":45,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T22:19:28.186402Z","iopub.status.busy":"2024-10-12T22:19:28.185572Z","iopub.status.idle":"2024-10-12T22:22:13.918504Z","shell.execute_reply":"2024-10-12T22:22:13.917750Z","shell.execute_reply.started":"2024-10-12T22:19:28.186362Z"},"trusted":true},"outputs":[{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"db20cb5c4f79475382b9f8320ae07fe5","version_major":2,"version_minor":0},"text/plain":["Training epochs on cuda:0:   0%|          | 0/1000 [00:00<?, ?epoch/s]"]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"80c1458a17784508823b3d88d6869f0c","version_major":2,"version_minor":0},"text/plain":["Evaluating on cuda:0:   0%|          | 0.00/93.0 [00:00<?, ?triple/s]"]},"metadata":{},"output_type":"display_data"},{"data":{"text/plain":["<Axes: title={'center': 'Losses Plot'}, xlabel='Epoch', ylabel='softplus Loss'>"]},"execution_count":45,"metadata":{},"output_type":"execute_result"},{"data":{"image/png":"iVBORw0KGgoAAAANSUhEUgAAAkAAAAHHCAYAAABXx+fLAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuNSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/xnp5ZAAAACXBIWXMAAA9hAAAPYQGoP6dpAACK5ElEQVR4nO3dd3wUZf4H8M/upkESSgq9SA0lJAQQNIAgInA0BYQTRAQLNrDrqT/PEz0Py6FiOUVBRAQbCBYQFRRUBKX33jsphCSkZ+f3R7Kbmdmpm92dkP28Xy9fkt3ZmWef3Z35zvdpNkEQBBAREREFEbvVBSAiIiIKNAZAREREFHQYABEREVHQYQBEREREQYcBEBEREQUdBkBEREQUdBgAERERUdBhAERERERBhwEQERERBR0GQEREfvLVV18hISEBJ0+etLooRCTDAIiIKsV1kd+xY4fVRQmYt956CwkJCe7/kpOTMXjwYLz++uvIzc31yTG+/fZbfPTRRz7ZFxF5CrG6AEREl6vnnnsONWvWRF5eHtauXYv33nsPf/75Jz799FPYbLZK7fu7777DgQMHMHHiRN8UlogkGAAREXlp4MCBiImJAQCMHTsWU6dOxY8//oitW7ciJSXF4tIRkRY2gRFRQOzevRt33nknunTpgpSUFNx2223YunWrZJvi4mK8/fbbGDBgADp16oQePXpg7NixWLt2rXubtLQ0PPXUU7jmmmuQmJiIXr164d577/XoZ7NmzRqMGzcOnTt3RkpKCiZPnowDBw5ItjG6L6OuuuoqANB9/YIFCzBkyBD3MadNm4bs7Gz387feeitWr16NU6dOuZvZ+vXr51WZiEgZM0BE5HcHDhzALbfcgsjISNx5550ICQnB559/jltvvRWffPIJkpOTAQBvv/02Zs2ahdGjRyMpKQm5ubnYuXMndu3ahZ49ewIApk6dioMHD2L8+PFo3LgxMjMzsXbtWpw5cwZNmjQBACxduhRPPvkkevXqhcceewz5+fn49NNPMW7cOCxZssS9nZF9mXH8+HEAQJ06dVS3eeutt/D2228jNTUVY8eOxZEjR/Dpp59ix44d+PTTTxEaGop77rkHOTk5OHv2LJ566ikAQGRkpOnyEJEGgYioEhYvXiy0bdtW2L59u+o29913n9CxY0fh+PHj7sfOnTsnpKSkCLfccov7seHDhwuTJ09W3c/FixeFtm3bCrNnz1bdJjc3V+jWrZvwzDPPSB5PS0sTunbt6n7cyL7UvPnmm0Lbtm2Fw4cPCxkZGcKJEyeEzz77TEhMTBRSU1OFvLw8QRAq6ubEiROCIAhCRkaG0LFjR+H2228XSktL3fv75JNPhLZt2wqLFi1yPzZ58mTh2muvNV02IjKGTWBE5FelpaVYu3Yt+vfvj6ZNm7ofr1evHoYOHYpNmza5R07VqlULBw4cwNGjRxX3FRERgdDQUPz111+4ePGi4jZ//PEHsrOzMWTIEGRmZrr/s9vtSE5Oxp9//ml4X3oGDRqEq6++Gtdddx2effZZNG/eHLNmzUKNGjVUy1ZcXIwJEybAbq84/Y4ePRpRUVFYs2aNV+UgIvPYBEZEfpWZmYn8/Hy0aNHC47lWrVrB6XTizJkzaNOmDR544AHcd999GDhwINq2bYtevXrhhhtuQLt27QAAYWFheOyxx/Dyyy+jZ8+eSE5ORt++fXHjjTciPj4eANzB02233aZYnqioKMP70vPWW28hKioKISEhaNCgAZo1a6a5/enTpwEALVu2lDweFhaGpk2b4tSpU4aOS0SVxwCIiKqMK6+8Ej/99BNWrVqFtWvXYtGiRZg3bx6mTZuG0aNHAwAmTpyIfv36YeXKlfj9998xc+ZMvP/++5g3bx46dOgAQRAAAK+88opiIONwONz/1tuXnm7durlHgRHR5YVNYETkVzExMahRowaOHDni8dzhw4dht9vRsGFD92N16tTBqFGj8Nprr2H16tVISEjAW2+9JXlds2bNcPvtt+PDDz/Ed999h+LiYnz44YcA4G5mi42NRWpqqsd/PXr0MLwvX2vUqJH7fYsVFRXh5MmTaNy4sfuxys4jRETaGAARkV85HA707NkTq1atkgwPT09Px3fffYeuXbu6m6UuXLggeW1kZCSaNWuGoqIiAEB+fj4KCwsl2zRr1gyRkZHubXr37o2oqCjMmjULxcXFHuXJzMw0vC9fS01NRWhoKObPn+/OVAHAokWLkJOTgz59+rgfq1GjBnJycvxSDiJiExgR+cjixYvx22+/eTw+YcIEPPTQQ/jjjz8wbtw4jBs3Dg6HA59//jmKiorw+OOPu7cdMmQIunfvjo4dO6JOnTrYsWMHfvjhB4wfPx5AWf+eiRMnYtCgQWjdujUcDgdWrlyJ9PR0DBkyBEBZH5/nnnsOTzzxBEaOHInBgwcjJiYGp0+fxpo1a9ClSxc8++yzhvblazExMbj77rvx9ttv484770S/fv1w5MgRLFy4EJ06dcLw4cPd23bs2BHLly/H9OnT0alTJ9SsWZNzARH5EAMgIvKJTz/9VPHxkSNHok2bNliwYAFmzJiBWbNmQRAEJCUl4dVXX3XPAQSUTQD4888/Y+3atSgqKkKjRo3w0EMP4Y477gAANGjQAEOGDMG6devwzTffwOFwoGXLlnjjjTcwcOBA936GDRuGevXq4f3338ecOXNQVFSE+vXro1u3bhg5cqSpffna1KlTERMTg08++QTTp09H7dq1MWbMGDzyyCMIDQ11bzdu3Djs2bMHX331FT766CM0btyYARCRD9kEcR6WiIiIKAiwDxAREREFHQZAREREFHQYABEREVHQYQBEREREQYcBEBEREQUdBkBEREQUdBgAERERUdBhAERERERBhzNBa8jIyIGvp4m02YDY2Gi/7JsqsJ4Dg/UcGKznwGFdB4a/6tm1XyMYAGkQBPjtB+DPfVMF1nNgsJ4Dg/UcOKzrwLCyntkERkREREGHARAREREFHQZAREREFHQYABEREVHQYQBEREREQYcBEBEREQUdBkBEREQUdBgAERERUdBhAERERERBhwEQERERBR0GQERERBR0GAARERFR0GEAFGAFxaVWF4GIiCjoMQAKoHVHM9HnrT8wf/0xq4tCREQU1BgABdC+c7kodQrYfiLL6qIQEREFNQZAAWS32QAAgsXlICIiCnYMgAKoPP6BU2AIREREZCUGQAFkc2WAGP8QERFZigFQANnLM0ACIyAiIiJLMQAKIFcGyMn4h4iIyFIMgAKoPAHEPkBEREQWYwAUQO4mMGuLQUREFPQYAAVQRSdohkBERERWYgAUQO4mMKelxSAiIgp6DIACqKIJjBkgIiIiKzEACiCOAiMiIqoaGAAFkKsJjH2AiIiIrMUAKIDsnAmaiIioSmAAFEBcC4yIiKhqYAAUQBUBkLXlICIiCnYMgALI3QRmcTmIiIiCHQOgALJxMVQiIqIqgQFQANngGgbPAIiIiMhKDIACyD0RIuMfIiIiSzEACqCKiRAZAREREVmJAVAAudcCY/xDRERkKQZAAWR3TwVtaTGIiIiCHgOgAGITGBERUdXAACiAKprAGAARERFZiQFQANk4ESIREVGVwAAogOxcCoOIiKhKYAAUQJwJmoiIqGpgABRA7iYwxj9ERESWYgAUQBVNYIyAiIiIrGR5ALRgwQL069cPnTp1wujRo7F9+3bN7T/66CMMHDgQSUlJ6NOnD/7zn/+gsLCwUvsMFE6ESEREVDVYGgAtX74c06dPx/33348lS5agXbt2uOOOO5CRkaG4/bfffosZM2ZgypQpWL58OV588UUsX74cr732mtf7DKSKJjBGQERERFayNACaO3cuxowZg1GjRqF169aYNm0aIiIisHjxYsXtt2zZgi5dumDYsGFo0qQJevXqhaFDh0oyPGb3GUhcDJWIiKhqCLHqwEVFRdi1axfuvvtu92N2ux2pqanYsmWL4mtSUlLwzTffYPv27UhKSsKJEyewZs0a3HDDDV7vU4tr1Jav2EUzQft63yTlql/Ws3+xngOD9Rw4rOvA8Fc9m9mfZQHQhQsXUFpaitjYWMnjsbGxOHz4sOJrhg0bhgsXLmDcuHEQBAElJSW4+eabcc8993i9Ty2xsdGmX6OlTnYRgLKJEH29b1LGeg4M1nNgsJ4Dh3UdGFbWs2UBkDf+/PNPzJo1C//617+QlJSE48eP48UXX8Q777yD+++/3+fHy8jI8WlzVU52PoCyDJCv901SNlvZD4v17F+s58BgPQcO6zow/FXPrv0aYVkAVLduXTgcDo/OyRkZGYiLi1N8zcyZMzF8+HCMHj0aAJCQkIC8vDw8++yzuPfee73apxZB8HF/HUG6X/64/I/1HBis58BgPQcO6zowrKxnyzpBh4WFoWPHjli3bp37MafTiXXr1iElJUXxNQUFBbDbpUV2OBwAykZWebPPQOJM0ERERFWDpU1gkyZNwj/+8Q8kJiYiKSkJ8+bNQ35+PkaOHAkAeOKJJ1C/fn08+uijAIBrr70Wc+fORYcOHdxNYDNnzsS1117rDoT09mmlik7QFheEiIgoyFkaAA0ePBiZmZl48803kZaWhvbt22P27Nnu5qozZ85IMj733nsvbDYb3njjDZw7dw4xMTG49tpr8fDDDxvep5VsnAmaiIioSrAJbI9RlZ7u285Zu87mYOKCLWhcpwa+vvNKti/7kc0GxMVF+/wzJCnWc2CwngOHdR0Y/qpn136NsHwpjGBiZx8gIiKiKoEBUABxLTAiIqKqgQFQALnXAgMjICIiIisxAAogu7sTtLXlICIiCnYMgALIBq4GT0REVBUwAAqgiokQrS0HERFRsGMAFEDi1eCJiIjIOgyAAoijwIiIiKoGBkABxLXAiIiIqgYGQAHkagJj/ENERGQtBkABxLXAiIiIqgYGQAHkbgKzthhERERBjwFQAHEUGBERUdXAACiAOAqMiIioamAAFEA2toERERFVCQyAAsjOTtBERERVAgOgAKpoAmMAREREZCUGQAHkagJj+ENERGQtBkABZBcthsrZoImIiKzDACiAbO5GMGaBiIiIrMQAKIBsFfEPl8MgIiKyEAOgALKLIiA2gREREVmHAVAAiTNAnAyRiIjIOgyAAkjSBGZdMYiIiIIeA6AAYhMYERFR1cAAKIBECSA2gREREVmIAVAA2cQZIDaCERERWYYBUADZOQyeiIioSmAAFEDiDFAp28CIiIgswwAogJgBIiIiqhoYAAWQ3VaxGEYJIyAiIiLLMAAKMHt5GsjJJjAiIiLLMAAKsJDyAIh9gIiIiKzDACjAXP2AStkERkREZBkGQAHmcDWBMf4hIiKyDAOgAHPY2ARGRERkNQZAAWZnAERERGQ5BkABVtEExgCIiIjIKgyAAsxRXuPMABEREVmHAVCAufsAMQNERERkGQZAAWbnPEBERESWYwAUYK5O0Ix/iIiIrMMAKMA4EzQREZH1GAAFmJ19gIiIiCzHACjA7OU1zsVQiYiIrMMAKMBCOBEiERGR5RgABZh7FBjjHyIiIsswAAowzgRNRERkPQZAAcbFUImIiKzHACjAyhNAHAVGRERkIQZAAebgPEBERESWYwAUYK4mMA6DJyIisg4DoABzcBQYERGR5RgABRgXQyUiIrIeA6AAq1gMlQEQERGRVRgABRgXQyUiIrIeA6AA4zB4IiIi6zEACjBXHyCn0+KCEBERBTEGQAHmWgy1hE1gRERElmEAFGB2rgVGRERkOQZAAcbFUImIiKzHACjAHK5O0GwCIyIisgwDoAALsZdVOfsAERERWYcBUIBFhJZVeV5RqcUlISIiCl4hVhcAABYsWIA5c+YgLS0N7dq1wz//+U8kJSUpbnvrrbfir7/+8ni8T58+eP/99wEATz75JJYsWSJ5vlevXpgzZ47vC29SZJgDAAMgIiIiK1keAC1fvhzTp0/HtGnTkJycjHnz5uGOO+7AihUrEBsb67H9W2+9heLiYvffWVlZuOGGGzBo0CDJdr1798b06dPdf4eFhfnvTZhQM6ysyi8xACIiIrKM5U1gc+fOxZgxYzBq1Ci0bt0a06ZNQ0REBBYvXqy4fZ06dRAfH+/+b+3atYiIiPAIgMLCwiTb1a5dOxBvR1fN8gxQfjEDICIiIqtYGgAVFRVh165dSE1NdT9mt9uRmpqKLVu2GNrH4sWLMWTIENSsWVPy+F9//YWrr74aAwcOxL/+9S9cuHDBp2X3lqsJjBkgIiIi61jaBHbhwgWUlpZ6NHXFxsbi8OHDuq/fvn079u/fjxdffFHyeO/evXH99dejSZMmOHHiBF577TXcdddd+Pzzz+FwOAyXr3zSZp+qKeoD5I/9UxlX3bKO/Yv1HBis58BhXQeGv+rZzP4s7wNUGYsWLULbtm09OkwPGTLE/e+EhAQkJCSgf//+7qyQUbGx0T4rq0vDvBIAQEGpE3Fxvt8/SfnjMyRPrOfAYD0HDus6MKysZ0sDoLp168LhcCAjI0PyeEZGBuLi4jRfm5eXh2XLluGBBx7QPU7Tpk1Rt25dHDt2zFQAlJGRA19P2FySXwgAyC0oQXp6jm93Tm42W9kPyx+fIVVgPQcG6zlwWNeB4a96du3XCEsDoLCwMHTs2BHr1q1D//79AQBOpxPr1q3D+PHjNV+7YsUKFBUVYfjw4brHOXv2LLKyshAfH2+qfIIAn/8AaoRUdILmj8v//PEZkifWc2CwngOHdR0YVtaz5U1gkyZNwj/+8Q8kJiYiKSkJ8+bNQ35+PkaOHAkAeOKJJ1C/fn08+uijktctWrQI/fv3R926dSWPX7p0CW+//TYGDhyIuLg4nDhxAq+++iqaN2+O3r17B+x9qQlxcDV4IiIiq1keAA0ePBiZmZl48803kZaWhvbt22P27NnuJrAzZ87AbpcOVjt8+DA2bdqEDz/80GN/DocD+/fvx9KlS5GTk4N69eqhZ8+eePDBB6vEXECO8h5apU4BgiDAxp52REREAWcTBCb51KSn+74N+GJBMfq/sw4AsP7h3u7V4cm3bDYgLi7aL58hVWA9BwbrOXBY14Hhr3p27dcIyydCDDYhooCHK8ITERFZgwFQgDlETV5O3l4QERFZggFQgNlFGSB2hCYiIrIGA6AAc7AJjIiIyHIMgALMIerzzCYwIiIiazAACjCbzebOAjEDREREZA0GQBZwdYRmHyAiIiJrMACygDsDxCYwIiIiSzAAsoBrLiCn0+KCEBERBSnTAVBBQQHy8/Pdf586dQofffQRfv/9d58WrDqzsw8QERGRpUwHQPfddx+WLl0KAMjOzsaYMWMwd+5c3HfffVi4cKGvy1ctuTJAJWwCIyIisoTpAGjXrl3o1q0bAOCHH35AbGwsfvnlF7z88suYP3++zwtYHTncTWAMgIiIiKzgVRNYZGQkAOD333/HgAEDYLfb0blzZ5w+fdrnBayOQtgJmoiIyFKmA6BmzZph5cqVOHPmDH7//Xf07NkTAJCRkYGoqCifF7A6Yh8gIiIia5kOgO6//3688sor6NevH5KTk5GSkgIAWLt2Ldq3b+/zAlZHIQyAiIiILBVi9gWDBg1C165dkZaWhnbt2rkfv/rqq9G/f3+fFq664jxARERE1jIdAAFAfHw84uPjAQC5ublYv349WrRogVatWvm0cNUVl8IgIiKylukmsAcffBCffPIJgLIO0aNGjcJDDz2E4cOH44cffvB5Aasjh72s2hkAERERWcN0ALRx40b3MPiffvoJgiBgw4YN+L//+z+8++67Pi9gdVQxCszighAREQUp0wFQTk4OateuDQD47bffMGDAANSoUQN9+/bFsWPHfF7A6ohNYERERNYyHQA1bNgQW7ZsQV5eHn777Tf3MPjs7GyEhYX5vIDVEQMgIiIia5nuBD1hwgQ8/vjjqFmzJho1aoQePXoAADZs2IC2bdv6vIDVkXsmaI4CIyIisoTpAOiWW25BUlISzp49i9TUVNjLO/Q2bdoUDz30kK/LVy1xHiAiIiJreTUMvlOnTujUqRMEQYAgCLDZbOjbt6+Pi1Z9uTJAJQyAiIiILGG6DxAALF26FMOGDUNSUhKSkpIwbNgw9wrxpI99gIiIiKxlOgM0d+5czJw5E7fccou7yWvTpk147rnnkJWVhYkTJ/q4iNVPCPsAERERWcp0ADR//nw899xzuPHGG92PXXfddWjTpg3eeustBkAGMANERERkLdNNYGlpae4FUMVSUlKQlpbmk0JVdxV9gCwuCBERUZAyHQA1b94c33//vcfjy5cvxxVXXOGLMlV7rqUw2ARGRERkDdNNYFOnTsXDDz+MDRs2oEuXLgCAzZs3Y/369XjjjTd8Xb5qicPgiYiIrGU6AzRw4EB88cUXqFu3LlatWoVVq1ahbt26+PLLL3H99df7o4zVjt3GAIiIiMhKXs0DlJiYiP/+97+SxzIyMvDee+/hnnvu8UnBqrOKxVAZABEREVnBq3mAlKSlpWHmzJm+2l215nAwA0RERGQlnwVAZBz7ABEREVmLAZAF3H2A2ARGRERkCQZAFqjIAFlcECIioiBluBP09OnTNZ/PzMysdGGCBfsAERERWctwALR7927dbbp161apwgQLB5vAiIiILGU4AJo/f74/yxFU2AmaiIjIWuwDZAEuhUFERGQtBkAWCHG4FkNlAERERGQFBkAW4FIYRERE1mIAZAFXHyA2gREREVmDAZAFHOwETUREZCnTAdCvv/6KjRs3uv9esGABbrjhBjz66KO4ePGiTwtXXTEAIiIispbpAOjVV1/FpUuXAAD79u3DSy+9hD59+uDkyZN46aWXfF7A6sgdADH+ISIisoTheYBcTp48iVatWgEAfvzxR1x77bV45JFHsGvXLkyePNnnBayOOA8QERGRtUxngEJDQ1FQUAAA+OOPP9CzZ08AQO3atZGbm+vb0lVTdgZAREREljKdAerSpQumT5+OLl26YMeOHXjjjTcAAEePHkWDBg18Xb5qiRkgIiIia5nOAD377LMICQnBDz/8gH/961+oX78+gLLO0b179/Z5Aaujij5ADICIiIisYDoD1KhRI8yaNcvj8aefftonBQoGIeVLYTADREREZA3TAdDp06c1n2/UqJHXhQkWjvK8GwMgIiIia5gOgPr16wdb+VIOSvbs2VOpAgUDLoZKRERkLdMB0NKlSyV/FxcXY8+ePZg7dy4efvhhX5WrWnN1guZiqERERNYwHQC1a9fO47FOnTqhXr16mDNnDgYMGOCTglVnHAZPRERkLZ+tBdaiRQvs2LHDV7ur1lwZoP1pl3Aw/ZLFpSEiIgo+pgOg3NxcyX85OTk4dOgQ3njjDTRv3twfZax2XMPgAeDxr3dZWBIiIqLgZLoJrFu3bh6doAVBQMOGDfHaa6/5rGDVWYloEbCiEqeFJSEiIgpOpgOgjz/+WPK33W5H3bp10bx5c4SEmN5dUCooLnX/OzLcs87yi0tRUiogOkK5Preduoh95y9hdOeGmiPyiIiISJnpiKV79+7+KEdQ6ZMQjwbR4TibU4hChQzQ9f9bh8ISJ359oCdqhDo8nr/zs20AgMZ1ItCzRYzfy0tERFTdGAqAVq1aZXiH1113ndeFCRahDjtmjOiIWz7eLAmACkucOCcKik5cyEfbelGq+zmWmccAiIiIyAuGAqD777/f0M5sNhsnQjQoPKSs/3lhSUVz2D1fbMPOMznuv9m6RURE5B+GAqC9e/f6uxxBJ6I8ABJ3ghYHP0REROQ/PpsHiMxxZYCKSgXMXndMcZsSp4Cz2QVYfzQzkEUjIiKq9rwatrVu3Tp89NFHOHToEACgVatWuO2225CamurTwlVn4SEVnZtn/XEMPVt69uUpKnFi2Ad/AQDeuakTujevC0G0fhiXEiMiIvKO6QzQggULcOeddyIyMhITJkzAhAkTEBUVhcmTJ2PBggVeFWLBggXo168fOnXqhNGjR2P79u2q2956661ISEjw+G/y5MnubQRBwMyZM9GrVy8kJSVh4sSJOHr0qFdl85ewEGnV/+fHAx7biDtIbzuVDUC6fAbjHyIiIu+YzgDNmjULTz31FMaPHy95vEuXLnjvvfdwyy23mNrf8uXLMX36dEybNg3JycmYN28e7rjjDqxYsQKxsbEe27/11lsoLi52/52VlYUbbrgBgwYNcj/2wQcfYP78+XjppZfQpEkTzJw5E3fccQeWL1+O8PBwk+/YP0Ls0h7Oe8/nemwzbcU+97/LF5BHUak4A8QQiIiIyBumM0A5OTno3bu3x+M9e/ZEbq7nRVzP3LlzMWbMGIwaNQqtW7fGtGnTEBERgcWLFytuX6dOHcTHx7v/W7t2LSIiItwBkCAI+Pjjj3Hvvfeif//+aNeuHV555RWcP38eK1euNF0+K53PLXL/214+JKy4lDNHExERVZbpAKhfv3746aefPB5ftWoV+vbta2pfRUVF2LVrl6TvkN1uR2pqKrZs2WJoH4sXL8aQIUNQs2ZNAMDJkyeRlpYm2Wd0dDSSk5MN77MqcigEQFxNnoiIyDumm8BatWqF9957D3/99Rc6d+4MANi2bRs2b96MSZMmSZbKmDBhgua+Lly4gNLSUo+mrtjYWBw+fFi3LNu3b8f+/fvx4osvuh9LS0tz70O+z/T0dN19ivljHh7XPs3u22G3wWYrGxnmUuwUOFeQCm/rmcxhPQcG6zlwWNeB4a96NrM/0wHQokWLUKtWLRw8eBAHDx50Px4dHY1FixaJCmHTDYAqa9GiRWjbti2SkpL8sv/Y2Gi/7NebfUdHRyAuLhrZQsWnGxIWgrg4/5WxOvDnZ0gVWM+BwXoOHNZ1YFhZz6YDoJ9//tlnB69bty4cDgcyMjIkj2dkZCAuLk7ztXl5eVi2bBkeeOAByePx8fHufdSrV0+yz3bt2pkqX0ZGjs+HmttsZR94RkYOkhvXco/u0pOfV4j09BycT7/kfuxiTgHS0zl5ohJxPbOvuP+wngOD9Rw4rOvA8Fc9u/ZrhOk+QG+//Tby8/M9Hi8oKMDbb79tal9hYWHo2LEj1q1b537M6XRi3bp1SElJ0XztihUrUFRUhOHDh0seb9KkCeLj4yX7zM3NxbZt23T3KScI/vnPte8Xh7RHzxYxeGtUom5Z7DYbBAEoEvUBKioV/FbG6vCfPz9D/sd6Zj1X3/9Y15d3PRtlOgB65513kJeX5/F4fn4+3nnnHbO7w6RJk/DFF19gyZIlOHToEJ577jnk5+dj5MiRAIAnnngCM2bM8HjdokWL0L9/f9StW1fyuKvp7d1338WqVauwb98+PPHEE6hXrx769+9vunz+VD86HG+MTMRVV+gvaOoob/kqFg2DL+KIMCIiIq+YbgITBAE2hV5Ge/fuRe3atU0XYPDgwcjMzMSbb76JtLQ0tG/fHrNnz3Y3gZ05cwZ2uzROO3z4MDZt2oQPP/xQcZ933XUX8vPz8eyzzyI7Oxtdu3bF7Nmzq8wcQEpC7DZJB2c51zPiUWBWDIl3CoJ7SD4REdHlynAAdOWVV8Jms8Fms2HgwIGSIKi0tBR5eXm4+eabvSrE+PHjPSZWdJk/f77HYy1btsS+ffsUti5js9nw4IMP4sEHH/SqPFaY0L0pPlx/HClNaqNTw2h8vOGk5HnXkHdx0CNeSPWLLaex+1wO/jmgLRx2/wQo3+48ixm/HMLrIxKR0sR8sEtERJcv1+S7SkmQy5HhAOjpp5+GIAh4+umnMXXqVERHV3QyCg0NRePGjU33saEK96Q2R4/mddCsbk0s3X7G4/nsghKs2p8GcZIoK79iRuxXfy4bkTcgIR6pLTyb1M7lFGLvuRxc0yrW6y/v8z/sBwA8/vUurLyf674REQULpyBg0sKtqBFqx7ujk3SvI8WlTmRcKkKDWhEBKqF5hgKgESNG4KOPPkLt2rWxZMkSjBo1CpGRkf4uW1Cx2Wzo0qQOAKBmmMPj+Vl/eK4Yfy6nEIB0biC1yRGHvf8nBAD/HtwOA9vXU9yGiIhIyemLBdh9tmzUcWGJExGhntcpsTs+3Yo953Lx4djO6NSoViCKaJqhTtCHDh1yj/zauHEjCgsL/VqoYBepEAApOZFVgE0nspBdUJEJUgqegIo+ROuPXahs8YiIiDTtOVe2NNby3ecsLok6Qxmg9u3b46mnnkLXrl0hCAJmz57tXnpCbsqUKT4tYDCqGWa8b/pLKw/g5eEd3H/rDQE0MULQr/sgIqLqLyzE9GDzgDF0pZ0+fTreeust/PLLL7DZbPjtt9/gcHhmGmw2GwMgHzCaAQKArPwSSV+gYqf2yDAzK8jnFpYgKtz0QEEiIqrGzCxDGeq4zAOgli1b4vXXXwcAtGvXDh999JHHWlvkO0YCoCub1cGG41nIyi929wUCpPMEKTEa/8zfcAJv/noE0/6WgMEd6nu1DyIiqn6cJi4CoX4alewLpkOzvXv3MvjxM7V+PGJ1a4S6/73xeJb73wXFpZi2Yh8WbT2t+DqjX9s3fz0CAJi2Qn26AaLqYt/5XHy1/YypDClRsDLzM7nsm8Dkjh8/jnnz5uHQoUMAgNatW2PChAlo1qyZTwsXrMQBkN2mnG6sEepwT57457Es9+PLdp/DH0cu4Ltd59A/IR61I0IkwxXNnuCVji2wFxBVM+PnbwYARIeH4PqEeItLQ1S1mckAhVSnDNBvv/2GwYMHY/v27UhISEBCQgK2bduGIUOGYO3atf4oY9CJrRmGqHAHHHYb+rZWXhQ2xGFDeHlkfV7UBLa3vOc9AFz/v3V4edVByet8cYPr7T4OpOUit7Ck8gUg8pMDabn6GxH5SV5RKV5eeQCbTmRZXRRNuoNtRBuEXe59gMRmzJiBiRMn4rHHHpM8/t///hf//e9/0bNnT58VLlhFhDrwxcRuCLHb8MWW0/j5QNnjDaLDcbY82Alz2BER6sClolJJPiYzr1iyr8XbzuDJ/m3cf5vpvOZLW05exOTPtyGmZih+uPdqawpBRFSFffjncSzadgaLtp3Bhkevsbo4EuKgx6nTCiCemy7UUY0yQIcOHcJNN93k8fioUaNw8OBBhVeQN+KjwlG3ZphkWYvEhqLZtx02RHjVtmpNBLTmYAYAzwCNiIjKnMzK9+n+jmTk+SybJL5y6N1IF4qWaarKo8BMlywmJgZ79uzxeHzPnj3sHO0H4gAouXHF+luhDjsiQs1/sazqveOv9cmIiEjZmI824p4vtuNoZl6l9+UURT16fUmLSi+PAMh0E9jo0aPx7LPP4sSJE+jSpQsAYPPmzfjggw8wceJEX5cv6DlEHZgT6kW5/12WATI2X5A4GreqD1BIFU6DEhFVBdtPZ/tlv0cy8nBFjPLkxUaJm730MkDihbqr8qAZ0wHQ/fffj6ioKHz44Yd47bXXAAD16tXDlClTMGHCBJ8XMNjlFZe6/920TsWicnabzXAG6EJekfvfZnrvq/HmCx1STVYPJiJSs/F4FpyCgO7N65p+7bZTF5GWW6S/Ybm1RzJxJCMPsZGhuLZ1nO7aXJUlDnr0MkBaN925hSWIDHNUiRXlTQdANpsNEydOxMSJE5GbWzZiIioqSudV5K2cgopRU3VEc/8UlzoNZ4DEQZRV2ARGRNVZQXEp7v1yOwBg9dRURJpY0ggANprsq/PQVzvd/x6emIV/DkxQ3dYXZ19x0KOXAVq46ZTi41tPXsRdn2/DjZ0a4JmBbX1Qqsox3ThXUFDgXhg1KioKFy9exEcffYTff//d54Uj4KJoodMQUVtqUangHgav5y/RPEG+GAXmVRMYAyAiqsbEk8bmF+nfdBYUl+Jg2iV3YKGzipGmFXvOaz5vNtlyIa/II8sjvnZotSTsL59U1EW86QfrjgEAlu44a65AfmI6ALrvvvuwdOlSAEB2djZGjx6NuXPn4r777sPChQt9Xb6gd7FAed6c4lLjv5YZvxxy/1v+xT2ZlY8SE/sCvOtIzQwQXQ74LSUtB9MuSeZdE1u5P73iDwMRx+2fbsXYjzfht8OZAHzTPcEX1h3NxIB312PaD/slj4sDIq2inpPVT1V5X0pMB0C7du1Ct27dAAA//PAD4uLi8Msvv+Dll1/G/PnzfV7AYHdL18YAgP5tpRMiFpU4UbdmqNJLDPv1UAZGzNmAh5bs1N+4ksQBkC+WGygsceKbHWeRlqt8MiJzBEHA4YxLKLVqoqgqIrjfPWk5fbEAYz/ehCHv/6m7rZFA+kDaJQDA97vPAahcoFBUKmD32ZxKlqjMB38cBwAs23VO8rjRDFCJ7BxSlX9TXjWBRUZGAgB+//13DBgwAHa7HZ07d8bp08rrT5H3rroiBssm98C/h7SXPF5U6sT4bk3gGlwVLVq1/anr20CN+Hv7+eaydlrxUhr+Ig6A5D8Qb7z/xzG88ON+TPhkS6X3RcCnm0/h7x9twgs/cO03IiX7zqvPEi4PCMzc5Lm21FnHWtdtC7bgz2MXFJ977Otd+O/P3s/T9+3Os3hBlBHSense5/cqHAGZDoCaNWuGlStX4syZM/j999/dMz9nZGSwM7Sf1IsOdwcQA9uVrVN0c5fGaFKnBr6b3AOrp6ZKskEjOjVQ3Zd4BFclmpxNE/cB0lux3og/jpSljdMvGR81EWx+2HMe93+5HVkGJp+cva78rm+3dl8CufM5hXhm6Q4cSr/kVRmJqgPxsG/AXDDjCiacssDhx73n8euhDFPl+PWg+vafbzktCczyVPspeRb++R/246DoN641E7S8e4Y4OKwCA78kTAdA999/P1555RX069cPycnJSElJAQCsXbsW7du313k1VdYLg9vhlympaBNfFmzGRYUjMiwEt/coW4h2aMf6msMLzQxlVOPN6xySAKjyoVdVnl69qnhm+V78dTwL75d3PPSH577fh0/WH8e4jzf77RhEVYHWGUd+U2emOasiAyR9zf8t24tHl+5yN0vPWX8M7/x2pOw1BvavtE1ReTk//usE+ry1Fj/tSzNcTjGtJL5WE1hV6w5kehj8oEGD0LVrV6SlpaFdu3bux6+++mr079/fp4UjTzabDVHhnh/b4A71kNgwGo3r1NB8vfjLGcjuHuKThy8CII4qM+5Skf4CtN7embnuCqtL3yF+q0iN1o1lkeycVuoUcC6nEOdzCpFdWIJ3fz+Kfw1q675xFXOPAlP5CZU4BZQ6Bby3tuxGZnTnRprNce79KjyWV1SC8JAwvFUeSD2/Yh+uaxuH/edz0SouEqEOu6EWK60ATC0A+mlfGnZp9lMKPNMBEADEx8cjPj5e8lhSUpJPCkTesdlsaG5gpk/xiC+tL7EN6k233lzqxK8p9sHFMqQKT69e1fhzKnqjUzEQXe7E8Y8gCJKAyLPZBxgq6yz9xDe7seSO7qr7lzeBifctzqAXlTrxyNJduuVV2t2lolLUFV0mBACfbjqFN9YcRt/WsXj1ho6GsjSaGSBZNkwQBOw5l4Onv/NcQstqPHsFGfHvVOtLrJVg8SaNKQ62fNEHKJQZIMP8WVdGJ+O8XFSPPBb5m/zcWSRvAlM4uWbla/fFU2s2c2WAXOwamSh5kCYn7/fjFAR8svEkAGC1Rv8hOe0MkDQYFISypTiqIgZA1dTsm5MVHxe3M2t9ie0+vmiKzwfydLE3qvICe1WNP+vKmwV5iaqivKJSPP71LvewdDnxGVEerHg0gSmcW9VOt+5O0GpNYKVOSVbF6KlZaX/yAKjUKXgMJDFyE2AqA4SqOxcQz17VVHLj2ogM87w7d91JCIKAHWfU22MdGncZXjWBiX4AZideVMJO0MYZCYC8rU0GQFRdfLr5JFYfzMA/lytPBSHOvMj7vGmNfNJ6DFDvBO3et1NAifjGVXErhf0q7O9SsTwDZHBnBvbtUqyQAarMLNf+xLNXNab0HS11CigscaL7a79pvlZz5mYvonnxD80XTWAhdvNf3XM5hVhzMMMnEzFeTvwZLIZXsyYwCh7ncwqx7mim+3yQla8+WOCZZXskE8aOn78Z2aJliuTD4JUu+GrBRsVSGGp9gATJTaPRAQdKAZX60HfP8mgx2weIGSAKOKVV20udAlbt1x/6qNXO7M1X2ddNYCFeXNSHvf8nHvt6l9dDPy9XYewEbVh1ziuWOAXsOJ3tk1GY1cGQ9//EA4t34tdDZXOKaX32P+yVnjOOXcjHgvK+M4DnTZ1yE5jymTO/xIkf9pzXXPZIvD+jWRulw+UZGBFqhFZAIz+/swmMLKH0nStxOvGv7/Vn+/V1v1nxj98XQ6a96djrOur6o8qzpVbG0u1ncPvCLcjMqxoTM4qHovozAxRRzQKg6uzd34/i9k+34sWfDlhdlCrFtQq72akgCkRZH/lFX7kJTOX4x7PwzPK9qpMelpQKkqyKWqbI83ie2xUUmwt+F246qbgfpWDty62n8fCSncgplD4noPKzXPsLz17VmNJ37nyusQu01pwX3hD/7HxxM1CZjr3+uBt58acD2HEmB+/+ftTn+/aG+C4/UMPgfdG3i/zn4w0nAHiu8RTs9M50RpqE5Fk1pZs8b5vei53SDJBaXyE5pTjJyDJE4t2/vvowshWaBh9dugubT2ZJHntl1UH8fjgTS7ZLV3oXBMFw0BZoDICCTGGJ8kVK/uPUOik4BeCjP4+bWv5A8OIHrEWc1TAb0PjzbiS30Dcp5soS90kIVB+gApXvlpKTWflVblI0gMPgg5HrXk+t2V/v2r37bA7eLp9Y0OWzzZ7rYnobAxTLM0Ba87eJ3oNSwGWk+VP+KvmwdpdZa43NMJ9bWBLQZZfMYABUjZm54zAbFLzz+1HcPG+T4e2lS3CYO5YSRyXWFvPn3UhVSfWKU/JGsnneZvzELZFmmjZHzNmAiQu24GRWvlfHDQYFxaWsnwBS+wXofa9vW7AFRzOln9NKhX6W3p4aikud0hn8DUYTStmed34/aqgPqOT4Ku8/w+A6jB+sO45z2YWmjhkoDICqsZQmtQ1v6+/mC0EwdgdjlLhZRz4CQ48/O+S5gquC4lJMXbwDX2zxvBMMBHGmz2zq/Wx2gaHRInLe1OvhKjZBWlXqBD32400YMWcDdp3Jtroo1ZpN51O3ugNvsWwiRK0Muuu3Pmf9MQye9afiNk9+a25GZrXz67EL+YZbAVYfTDd1zEBhAFSNTftbO/2NyhlpG64M8e61Tij5xaV4c81h7DitfdKXTw1vhi+yNIIgKJ4YXCenpTvOYv3RC3j154OVP5gXiiRLnuhvL74EDPvgLwz/QPnkKSfedRVt5r9sncwqAAD8GGSjFv3lYNolDJm1Hku3n5E87kp+qmVBfdFkXxm5BSU4dqHiRkHr/Pn5ltN4c81h97ph3pDfMGll2H/Ye97QPqvqyEMGQNVYbGSY4W3lczfoTanu8XqnoNn/xSnJAKnvZ+6fxzF/40nc/ulW92MX84vxwg/7sPXkRXGh3P80++PyxSi051bsQ+83f8fZ7ALFfRcUm8+g+JI4OPPm7aoNyZUTfzcCOb/SuqOZeHjJTpzPqZqpdV+qoiOI/erPoxc8OtmasfpAOl5aeUCS2Z7xy0Gczy3yGAXnOtWptQL/Vj5MXk4vc6Tkp31ppkeKPrN8L6at2O/+W+/8NV80PN8X5BMbitUMNTYPmC/mfvMHBkAEQPtLbuQC+uQ3u9H/nT9w/IJynwXxSVzrQqm0Zsybvx7GNzvP4a7Pt1XsQ/S82eyVL1Lay3efh1MAvtwqvZt07dvqpTqkJ0k/NvmJdh3Ic9wDi3fi98OZeHmVNRm2QLK6CSbQsvKKMWXxDtz9+Xavb1Ye/2Y3Fm87g69EI5LUOjm7M0Aq+/rn8r1elUHJ09/twcB31+Olld5PReDvTKt891rBS02F1QaUaF1frMQAiABoBxFG7uzXHMpAqQDMXqecehWfxLUulEozPCsFRZILr1PAscw8w/1WzFxQzucUYtmuc6pZJvkICdd7C/HhREqrD6TjsaW7dBdTVCoH4N8MgiTMsuBCnZbr2wxQcIUaVVNmfkWGpLLfqXM5FRna6IgQla3Kfqs+nvlD0+JtZ/Q3UhHogFgrw16DGSCqyhIbRgPQ7xCt9QXV+u7K+8HsOac8tNloBsjoDM/iXew8k4Ob5m7E6LkbDL3WzF3luI834bkV+zDvrxOG9uX625dDzx//ZjfWHMrAe2uP6m57MP0S1h7JlIx0M3Lv5e3JXzDYtOkvRkavXe4ZFF802V5OxPcUlX3n4hu7qDDlAEgvA1TV+Pv7IP+5aAVARkvi7z6m3mIAVM29dmNHPHV9Gzw7sK3mdvIvud58Ei5vrDks+VstC2O0D5DSDM9Km4vL9NyKspmtjU7yaOa36OoLozZ7tPyH7Qo8xJksX2VGLuTpZ4DGztuEh77aib3nc31+fCVGO7db5Wx2AQa+ux7/+/2I/saomhfBqler/mX0XGGEOFiICq8IgMR9g9yfucm7gAWbTuL+L7dXpnheCfTvrEjjDthVj3rnmKoaxDMAqubq1gzDyKSGqKWa/i2j9QXVGgXx5dbTko6oaj8WcXil9QMWZ4C+3lGWJlbavDI/J29OIGqZKXm9ufYt3t5Xdz9mzs/iSQb9eb6s7MVKEAS89ethrDmovAxAZc1edxxZ+cWY+6dyBu9yYOTzy8orxsJNJ6vMUiyVIf5O6V1YC4pLMWXRdny6+ZTi859vOY1nlu2BIAiIFPVXyRJ18v/z2AXkFpZ4dTH863iWF6+qHH+3JsnXkNSaIsXVclBFW7h0MQAKEpqru8MzKBBvrXcCFs85k5VfjKHv/4ltpy5KtjE6D5A4c/LvH9U7ClbmLsibuxGHSvQhD25K3Bmgiu3VZt82y8yok3xRJs7bmjJST+KPwZsJJlfuT8fHG07isa93mX4toJ+xsSIrtetsDr7ZcdZnmTcj7+GpZXvw+urDeGypd/VYlYgvpnpfqaU7zuLPY1l47ZdDAMpuyL7bJV2K4Ye9aTiXUyi5sItHae45l4tHl+6S3GAcTb+EY5lVa44ql0BnU7QyQK5JEr2ZR64qDI1nABQk1C7gLpoZIJM/uHM5hXhoyU4AFSdvp8GTmlKgZuboe87lYMqi7Zj31wk8+sU2xaH53vzu1DJAHk1grk7QolFgvvqh6/WrFl9w80UneG8vxEYyV5IMkBehlr+HsVtxip24YAte+HE//jym3GxqlpGPb2N5JmLHGf3lRapiM5+YmUlTxd/zjcez8Mqqg5Ih42Li67j8nLb5pPSGre9/V2PUhxu9mhDU35Zs974DtTe05llzndu8mStp2gr9Rbn9TbtdhKoNvVFJrovdxfxivLf2KNIviUdiaO9b6SRVWOLEcyv2Ydmuc+jQIBpdRJ2wf96fjogQOwa0q+fxOsU+QLKUuM1mUw2iJi7YAqcA/HksCwAQbhPwcN9WuuXVozQ6DfCcP0mtLnxBrwlM3JFdGgB5d7wSpxPhOvdIgsHA1l90mwUDlAFyfS/FDmfk4aorYiq9b19nsfSywVYTBydm3vq9Gv1xbDabZL9K31WlDvXnfTzK0Bd+P6w8L5GvyOtcK7vjOv95k5Vascf6CT6ZAQoSdp2T3lfldxVvrDmMRbIhmnrRvdIIMqdQser07rM52C/qlLv2SCb+b9leZOUXe9xh6Z2cXcdSy2rIf4dnZBMVlm3jTQCklgFySqaDd50IxOXTSiG7HEjLxX1fbsdOjWUP9OpGfKeWJwqAjLxfpZO/kaGrkgxQ+XvPLSzBNzvOqg7b92WnbP0mMJ8dStW2Uxfxt1l/YqXKjM2VXdOrsuHz1zvOYNLCLe6/q3oApNaxvtQpVOq7I26iVbpgK9XKpSqyuLGvqX0DnILgsTyN1nnANQ1IVR3lpYcBUJBQmwTMZfnu8zihsraL3klHcZIr2WuULsL/+GY3+ry1FgfSKoIjpROT+KUFJaVKuzfFqz5AKheNtNwiyaKwSk1+RpbqePCrndhwPAuTFm5V3UbvsiXONF0qrHwfIGNNYBX/dn0m//5xP174cT+eUOnXE6hT5cX8YvxywP9rED28ZBcyLhXhqe+U11i6Zf5mjJizQTO41VLZgPHfPx7ATlHTmC/nqPIHSSfo8v+XlDpx87yNuG/RDq/2WeoUJDdySiucK50is6tpAKT2Ffhu5zmPx9QWQwVEnaAZANHlbn9armImQC8RoHSHIH9EaReudvf5G066L7ZKF13xY/nFTtX9GaX3W52z/hgmLdwiaUZSu2jIJ2l07VuSARIFJtkFxfjPT/s9OomnGRjCrzfnjTjQyhGduL1uAjMQuInfp+sCs2p/WdCx5ZTyBT9Q58p7v9yOAgPNj5UNMPQCXNfs6Cv3eReM+bq+9G6GrCZtAiv7977zuTiamY+Nx7O8+rycgnRBUaXzjNIgg+z86hkAqdkqOy8B2n0YXcERAyC67B3LVFvGQvvLXVji2VFQ/nvQ2sX3e85j+Ad/orDEqfhDEv8AXaM3KjUKTOO1TkHAe2uPYeeZHCzdUTGaRC0Akme/KprARNuIyj9zzWEs2X4Wd362TfI6I5ckvRt3cQZIfEz5sFajzGeADB7Hp31a1CvlQJqxlar9deqWv01vV0fx9TxOZhNA6ZeKcEJliRs1rt/nzjPZ+Mc3u1WbAJXem/i3XVQq4LtdZyV9EsU3XEbfSolsRXWloFUpLjS6Jt7lRu2nrXTDoLYaPFB2nnlm2R7MKB+FZ5bVI8HYCZrcsvKLFU8oen2Aikr0T9B6W6TlFmHnmWzJRdfV7CR+resHeuqiZ98eo9SGa1/IK8L4+ZvdfxeLfvhqTWDyTtCuk6x4RJQ4MDmYrjy01mG3eQQcO05n4/XVFRNN6mWAxCcTsx1JlfZsJAASB1dGbwLF21WFZIRS/RSVOBEWYixiUQtQ5I962/dGqV5/LF+FW2kggR6z5fjbe+sBAD/cexViauovsHzqYj5u+2QLxqQ0wgfrjgMo+z7+98aOku1e++UQfj6QjgW3dkHtGqHux8XfuwUbT3rM8VNUWvHZGJkJHCgLqsT1WGzgnAWUNaNWBX9rXw/f7zG28roRau9eaRFnrSawfedzJc2rZuVaHGAyA0RumXlFkgn0XPTnATIwVNTgXay4bb7UKaC4VJoVKiguxcbjWaozMxuhFtC9u/aoZDZp8YlYLQMk39PZ8mHd0gyQKPWucsejtP/7F23HDlG/Eb2pDMR3anprgT33/V7c++V2zUyaPLhTIh0FZuwzrmrJckl5bDbM/fM4es78vVKrkSvR+/zUyKs1t7AE/7dsL/5v2V6vhml7G4gdVgne5d79/SguFpS4gx+gbK3AFbIL+KebT+FcTqF7AIaL+Hv3xxHPEU/ejKosNZABUvr+XqgiAVCtiBBc0yrW78dRqlutLI1WdsgIb4bP+xIDIHL7Ya/yKBa99l0j/SyMzBQqCJ7HulRUKrnDLih2YtG20/o7E+1TTu3Cvv20tM+KOBgLMdF+8dexC7IZkrX7HgDKFyVXfycXvetnocqJSqkJbNnu89h4PAt7zuWq7lupo6ic0fmdJOWRTGtQ8Xh2QTG+3HoaF0zMZuyTDJJ0oTr87/ejAICXVhpbad7oKdz7DJD0COKgx8hn5FEOLyvNaICr9j3QW1W9oLgUb/16BFtE/VD0ppUw+k6cTqBEtC+lrIZSVWZcqhozazvsNtNNl96Qn3MA/y5kanXfITaBkS6976iRAMhoc4rHzMqlTsnxC0pKK/2DVEvpyverVOa953LwxRbtAGzvuVzUqVmR0hfvRq0elDJAIbJmMb3zn9rdmGd/LFHHZY3PxdBnJtrXnnM56Ny4lup2ruYK8V7F1+Jnl+/D2iOZ+PlAOt4dnaR7bF9Re5d69Z1dUIwHFu+UfG+0+uv4KgCq7JBjvSkxxCTBquHXmCvP/vO5eOKb3QgPsXtkiZR+6gfSLuHhJTsxsUdTwwFwiSBImr6LFX4rSsGWkcEJgRDqsAdk+oIChWy+ZifoSp6LS5wCjK0n7x8MgEiX3p2fkTSosRFFnif3UkF6/IJip6Fh5VoyLhXhmx1nMbxTA8nj8tOL+MftLG+Ou/WTLbr7dwoC5JM3VuxTuexKJ7fIMIekE6bS6J0ZvxzC2ewCvDK8AwpV+jXIH5UsNaAVABmaB6ji36+vPowaodLT2ScbT+JoZh62nLyIT27tghqhDtXv09ry5o6NAV5fSa0K9C6u8zec9GgyljY9Snfsq9FX4qA1t7DUnbEySq05d+Gmk1h9IB2vj0xEZPnK6V51cjfZyLlyv/roOKXv57++34tLRaV4dvk+PHBNC0PHcBppAlN43Q4vpy7wtRC7Tff70yK2pseoVLOUmsDEg0HklAImM0otXkSMTWBBZGC7ePe/H7imBQa1r4focP0YOCu/GK+vVu/lb6RN3lgGyPOi+8bqQ5J+OV9sPe1V/x/5yfuFH/fj+IV8FJc6sXjbacVRKuIybziRhdQ3fjd0rFJZh0tx7ZhpAhMv3ggoX5A/23wKqw9mYPe5XNXgSv7eJRPCafUBMpkBAuCRHZu55jC+3nEWxy/kY9X+tPLX6O7Ww1fbz2DryYsex/NNC5hygfTWXlPKfGoFlL7qBC3+XN5fdwyLt5lbGkGtGK+vPowtp7Iln6GZxW7/PHoBN8/bWKlOsXJKwfIlL/o9lToFyb6UfiverGUXKKEO/SYw+c2HN5Q6QWs5k125mbLZB4gC5sn+bdz/7t6sLl4Y3A7ROqvEA8B/fjqAhZuUV1sGgCMZ+sONDaXtBc8fhPzuUN5Pxyilw6dfKsT8DSfx0sqDGDFng8fz4qyVmTurnIJS1QyQWlZF6a48Uhacal1AS52CamZMfo5R658kpzTB5bZTFzF+/mZsKZ/DSb6FVhldu9M759UIlZ6WVh9Ix/SfDuCuz7fhxtl/IbsgMB1TvUnYyL+/4vo1EwBprYclzrgeV5m6QoteJkE8/5V0+QjtD27K4h04lJ7nHgjgC75KEJQKAjLyKr43Sk03vij3y8PaAwCuT4hHxwbRkuce7NMSNb0MUkLs+k1gSusemmWkO4NRbeMjdbcp9aIPmy8xAAoiEaIfX4nRqxEqJnJTs2y3/vBMI01gJYLgtzZ3pZN3cakgWVdHfl3wtq9FVn6RJODSm4ANMJgBgg0FxaX440gmikqckoukDer9eeQPiy/Srq+B4jB4hYvEnZ9tw77zuZj8+bbyfcubeRSLAKCiM7bevERRosAvr6gUj3+z2/336exCScbDF61KktKIdqi3a8UpI2TTD4jrUGU9XQ8FxaUY89FGyX7ExIGuN/Nh6V1I1dZ3syJBopeVMToM/rGvd7mDdkC5CUze/8gb/drG469HeuM/Q9vjvTHSfmw3d2mMUKNfAkg/p7IMkPHA1VtmM0BajDT5Wr2EBgOgICLOMjSoFQEgcEOSjXzRn1+xD/tEa4b5ktLhp/90QNLGL2/y8PbHmVtYqrhERNk+VfoAKZwsXP0wXOw24Jlle/HgVzvx8JKduCiapdZm07oYypvARP/W+AYYawKT/q2ZARKk/1cTVf6+d5zOxntrj2oe09eDwMR/qF1cP/jjGF5ZdVAx+JIHoeIsmtE+QK5Zj108MkCVDYB0yiHeo2RZCh+dLIpKnLjz062GtvVVE0luofTC7s+RTa7vTbhoHqm/ta+n2vfqvl5XKD4u7p4QYrfpfm739FTejxlG1i00ysjX3Ug/Q39iJ+ggs2hSN+QVlyI2Un9CM18ycjHNzPNf04ZSPw+9yRS9DYBKZIs2ioMMtROvUuAQEep5f7LmUAYA4K/jWRg0a737cRvUAwvx8hw2m02aAdJ4i8ZmgpZuk66RwRM8/qEsKrws83W7ykXS15MnqgURSocRBAHvrzsGAOjZwnOld3HdCpB+3kabwCJkzSRaGSBvvqN69ScN2JWbcivjQPolbDPYlK03TNrbr0JlB1Io+e8NHSR/22w2PD84AVn5JRjbpbHq6yb1aKbYkT063OFeUDjEYYddI13x8/2pyDAxfUQgGAn4KzOjvy8wAxRkmsfURPv6FW3Tvvz+yZtsxKxKdbqO6s2dpLfTtJc4nZK+MWoXFDGli6P8jtGjKUv2gNrJRBDK3sst8zfjn8v3Sl6ndYF5b+1R3ROU/GnNfhSCsSYwvTmXflJZdd3XbDbPle3FTbRKF1F5nxlx06/8E87KL1bsfK/3uYr7AHlzAdF/iXK/H19drMzM/aJ3TG/L9MNe382qDABdm9ZGn9ZxHo//rX19zeBHS5QsA6QWUHw5sRuiI0ICusit1rnexUhx2ARGlvLl1+/hvi1Vn7M61WnoPCn7wXoz4yxQdtev1YlVidLJK0wWCGjux2aDWrwmQMDGE1k4kHYJK/acNzQxI1DW9+vBr3bim53qw2DNXICMNoHpnTcla3z5IB2k9RZe/PEAXvhxPx5dugsbj2dhyPt/VhxaYXtJf69S6bxW8rq6/n/rMGLOBpyXBY0eAZDsGEWS6RnUy67GKQhYczBDsr6XOFCT9AGS9V/75UA6rpzxK8bP34yC4lKvAhAzkzfqfVe8vX6ezPJ+KR0lESG+n81G3ARWM9Sh2gR2RWxNAIFd5Pa5QQkejyXUi5L8baR/ltUTITIACnK+Smt3algL8VHhqs8rjSgKhPM5hTiYfsmrE7W3nQpLnLJh8AYOrRQAySes03oPNmisSSXrjCu9SGt/LuuPXsALP+xXrQsztSrI/g8o102g1wcTZ6TEo2DsNhtWlg/d3346Gws3nZS8TmmxVXGmsUQQpHNJqVTWbvlcQrIN5Z+rdMkT89/r/WmX8NjXuzDyw4qRj1MW76g4nrgssuzlE+Ud0vedz0XvN9fi9oVbTR/fzM2QXobA6guoi3zkohq1oOD1ER3RrG4NyWO1IiomU+3evA5CdDpQByoB9NT1bdC3jWe265FrpTfAhjJAnAeIqoMQh02zj4NVX/Q953Ixdt4m06tZA2WTLnpD3geo1Cnglo834coZv6q+Rlx3Z7PL705lVaZ3rle7GApQn8/FdYHWu1tTu9CYuf66tpVc0BV2UNafydiOD6dfwpqD6hPpiR3LzMOxTM/pDMSHEk/3IK8SeYmU1okSx5OlTkES+GfmFSkGqfKmNPnn6OtRYEo2nVBefkIvW7jrbA6eW7EPvxww9hmo7UeN3igwq+eRcQk3OLxd7SalV8tYLL79SsljHRpEYfLVzfHvwe1QKyIUETqL8/qzCezfg9u5/11HZeoU+aANIxkg9gGiKqVx7bLRYc8MaKOzpVSozigFs19zIxM0mjHRwJ1qlqwTtrcZoGLZ8h0H0y5hvyxbEC47mYkDoGEf/IV1RzM9Tg5aF/qyUWDKzwmCIJ392cDM1J77UH7czAnMdfKXxD9KG9pshgPmiwUleOzr3dgmWj8KUA7Ybpq7ETfN3egxc7nakby5nEgmmXRKM0AfrDuON3894vEaecd4M32A/J0BMTKFw7Jd59yZIbP71KO3pebEk36KB+7teQWa1InAlc3quB/TC07cr1UZ8aWkQa0I3JXaHAPb1wMAhOs0s5lZ4sSsAaJJdOWd9F3kN8BG6p99gMhS8u/fe2OS8OsDPXFDp4ZoY2AiKxe9DJBZdWoEfoCi/I7e20nBSmSzziotUlq3Rqjkb3nw+NnmUx4n//MaI6ycgvrFQBDU1/9yrYum98kZnWNIi6sWnJKyKG9rtsnUFWBuP52NzLwizRNrniywNdoMbGSzElk9y5sYP9l4Uv4SjyDUczJF6fbiDJDS5/LXsQvufy9QOJ4ZknXsfJTF9eVFTysAjwh14Kor6vrkOMmNKta4a1c/Ckvu6I5JPZpWHMtgH6BRyY1wS9cm7r8XTeqmuu11baXNTPKbJjl/9gESZ3Nia4aVH0+6jccNsKE+QJwIkaqQGqEO95TqZlaNDjUwU6kZoSZWX/cXbycFk1/0FKfdl5245R2ua0WEmutfIwiqFwOnIF0HqdSLDFCJyr7N9CFT2lapCcMG8/O0CELZGmJ3fLoVwz/4S/N9yb+mqhkg2fffSLZLnjExcrGXL87r2QdIur0kAFLY/TPLylZdP5qZhzfWHNY9vpzaJJ7rj2YqbG3etxqd6s3S+prYbTafNAsl1IvCf4a2d/9du/zmRdzkozRlhZpWcTXd/66pMppqVHJDj4BGLwDy9yiwx65thQlXNkVC/bLOzvLfh/zwRmrED7MRmGL9VYYsJT9/iH+QZm4o7HbfZoCM3iU+2b+1z44pV5k+QPM3VNx5K13MxQ+tO5rpMS9K7YgQUx1snIJGExhkM/qK3pbRu3pfZICUXqO0XwHGZg6XW1e+RlxhiVPz+yO/sKhVs/zbbOStyqcYMDLfjEcGSKcJTBwsK2X9XNtfVOijZIQkWyj692+HKx8Anc0u8Ml+XLSawOw241mR+zWapgRBkJzbapU3z4uHghttAgOkgYyZoMXKDBAA/L1LY0wVLT4rP5r8/G+kSS7oM0ALFixAv3790KlTJ4wePRrbt2/X3D47OxvTpk1Dr169kJiYiIEDB2LNmjXu59966y0kJCRI/hs0aJC/38ZlS35XLs68mElVlzoFn7W535DYwHCnZTNZKrO87QN0JrsQOaJ1edQWXiwoLkWJU8Drqz3v0mtFhJhrXnIK6p2gBVmzk8Fh8JL9+yAD5Lqwi0ddKQVATqfgkRXRJ0gCdq0ASl5k9QyQ/HX6ZXIK8gDI8zXydeXM9gHKEy0G6otOpFqZTl8vEOrryU61OkE7TNyUqWVigLLvh/izd/VPjI0Mcy9t0ai876QR4qDHzE2jfgBkeFeqHuyjPpWJ3vHk78VIcazuA2TpTNDLly/H9OnTMW3aNCQnJ2PevHm44447sGLFCsTGxnpsX1RUhEmTJiE2NhYzZ85E/fr1cfr0adSqVUuyXZs2bTB37lz33w6H7+doqI7e/3uy5O+cAuOL6zlld0mVMa5bY3y766yhBIg/b3p89dNUyiTlF5fixjkbVPs6RYWHmDq+E4J6HyBIm8CUOkHr1aMvMkCuk534c1UMgGTDx40QBOkJV2vFcI+gQW0maFvZ4ihmSiJpAhMEFCv0Ixvz0UbMuLGj+2/PPkDaxZMsVqpQdrPl/df3+1SP5+tROr6adsNFq0O13WYzfFOmt5J6nRqh7sCnVvlvNio8BHPHpeBifrGkQ7Qeh2hKZ3MBkHYZfXH+NbOPsiawivqXZ6CMZKSsnsbA0gBo7ty5GDNmDEaNGgUAmDZtGlavXo3Fixdj8uTJHtsvXrwYFy9exGeffYbQ0LJ22CZNmnhs53A4EB8f7/E4aUtpUlvyd26RNAC6O7U5Zv1xTPG1JU7fBUChdnvZj8vAydLoYohWUsokFZQ4UVBShIxLRYqzqso7LusRBM8J8yTP6XSC1vPBumOSzpsuZi6QruY2SQCk1C/I6d1SBeI9ZWk0/8iPqFoFOsPRlYjvaLWawB5dusv9b3EAdOpivsd6ePLD5hVpd4KumG5Av7yfbj6Fn2VD2MUv8/XsFfleNiur0fr62jVGRsrpzWwc4rBj+d09YLNJZ2SWT/5nhHhBVLUMdus4zwEoes1sPgmATJxPPTtBS/82squgDYCKioqwa9cu3H333e7H7HY7UlNTsWXLFsXX/Pzzz+jcuTOef/55rFq1CjExMRg6dCjuuusuSZbn2LFj6NWrF8LDw9G5c2c8+uijaNSokeky+uPa6tpnVblui0+S8jLJFxC8K7U5vt9zXnF1eKdT8FknvFCHfFlSdQGc/d1rek1pStkKs3feTkFQDZjkfYDkEyEa+S4u330ey3dLlw+w2cxlG0qE8mOJjqeWATLbN+C/vxyS/H1RM3spbS5T+7JtOSXtl2Xkvcqb9wz1AXJWlOfG2Rs8nncKguS8If4+KV0/XNsbGUk3U6GT9JdbT2NIx3pIbFjL5xmbe7/U7uJgltbvxGG3GW5i0WoCaxFbEzYbUMPA8g9GiAOgUIdN8l385NYUbDiehRHJDT1+l2Ehyl9U13ZKwUuruJo4lO4595UatbEnSucIeYbHow+QwdXgfX0tNLM/ywKgCxcuoLS01KOpKzY2FocPK49cOHHiBNavX49hw4bh/fffx/HjxzFt2jSUlJRgypQpAICkpCRMnz4dLVq0QFpaGt555x3ccsst+PbbbxEVZS5aj42N1t/IS/7ctxmNY2rgwqmyu+W4OO0yxcVFw6HyC7E57Ij30XuqFx9d1oHOwMmrVnQN3W2sVuTFXc6hrAKc0lpXSya6Vg2ERXjOTAwA4eEhqBFZMUv39FUH3f8OCQtBXFw07ForLaqIi4uGzUQEGhoeiri4aFwUKl6jVDM2hw01K/m5lmo0e9epG4W46Ir6KAoxdhoMMTDUOTJKVG6HHWE19Bcddn0GasGG3W53ny9iY6Mhzm0pBQA2mw1xcdGIMHHhk5u4YCuOvjQE0bn+W6DYF8I05gsLDXHAZnA0aYN45XPXzVc2xeMDExCrMcu9WbHZFdNZ1IuvJeks3CsuGr06Kt+s18lR/izE5+27r2mJWb9WXD/N/q7r1PL83bVvWEvx2iAPcOrJ6jBC4bN55Pq2+HV/GkIcNqw/nIlSp2DptfCyWg1eEATExsbihRdegMPhQGJiIs6dO4c5c+a4A6A+ffq4t2/Xrh2Sk5Nx7bXX4vvvv8fo0aNNHS8jI8eni4UCZdFpbGy0X/btjecHJeCVVQcxqUdTpKfnaG6bnp6DUpU72oKiElzMUr4Am5WddclwBig317dr+vhDrom+VC7Ltp8xtX1WVh4uXVKeJyg/vxjZ2RVZu0OiSRmzLxUiPT0HghejMdLTc1BiYq6k7NyyY6WlVxxfPikhABQVlSI9I9fjcTOOn1f/Lmdk5sJWWFFXmdnGvkNFxfqfY2ZWRdBx8VIhMhUWO5Wbu/YoJqQ0Us0wFJeUIiMjx33eyBat+q00is/pFJCenoN0hVmvzUhPz0HGBd/8pv3lZIZG+QQBGTnSzzY+KkyyoK1Lgcp55LE+LSAUFCG9wHcrrV8SHSsz0/j3PPui8ucpPm/f1b0J6tcIwfM/7AdQ9t0xIz9PetMVGebAK0Pb6V4bACBL9l0pUch8j+pYD+OSG+Dxr8smzixxCj6/FrqusUZYNgqsbt26cDgcyMjIkDyekZGBuDjPdUYAID4+HldccYWkuatly5ZIS0tDUZHyF7RWrVq44oorcPz4cdNlLOuH4fv//Llvs/81rVMDb43qhC5N6iiWU14fakqdglf9cepFed4hh5i4a6kqTYlavB1NZkapoN4c4IR6X46yhVu9O6bZfkolpQL+99sR3Dxvk/sxpSawUkF59JQZmXkak0Y6Bcn33GiCzshbFb+fgmKnYoCn5KWVBxQ7TLuO6zr2nrM5ksVTlYokoOz9GT22GkGwvo+Gnt8OqQ+ptwE4fVEa2LwxIlFx25phyrkAf5xzE+KjUC8qDJ0a1jL1OrU+PlrnbbMfn91mk/Rrevr6NmhQK0KxPPLiyDsuKBXXXl5eV3Od/Lfoy2usofdrfFPfCgsLQ8eOHbFu3Tr3Y06nE+vWrUNKSoria7p06YLjx4/DKbpbPXr0KOLj4xEWppxqvnTpEk6cOMFO0X5W6hS86o8zIqmhx2MhdpvhC6vx3kLWEQdA3tRRw1r66XenU1C9WAmC+ggxdwdcLyNJc6PAnPjwzxOSx5Q6QTudlZ91OLdQPVsjP6TRE6aRzcSfQX5xqeIs4Ep+O5Sp2l/FFdj+uj8N4+dv0R1K7tqN0WOr+WbHWew4XXbnH1MzVGfrqkleV63jIxVvumqprG/lD2Ehdnx9Vw/MHpts6nXt60ejm2y0WVKjWh7btatfEcCY7UsYYrfhvTFJhrbVmwhR/Pzozo1wW/emCClvknS1TFo9DN7SeYAmTZqEL774AkuWLMGhQ4fw3HPPIT8/HyNHjgQAPPHEE5gxY4Z7+7FjxyIrKwsvvvgijhw5gtWrV2PWrFm45ZZb3Nu8/PLL+Ouvv3Dy5Els3rwZU6ZMgd1ux9ChQwP+/oKJ0he5QbT+hXt8tyZ4dXgHyWMhdpvhC2sgOkHLp6QXEw9nViN+L95kyYzUheZEiBrZocqegMxmgOTUMkBGZ6hWozWJpbwujF4kjLxXaQao1HAWpmaYQzXocz26fIe5ZtHK1uELP+7Hu2uPAvD/LMP+cn2C9MZXrWOur9ce1BNit5meuNBht+Hd0UmSZTnevqmTx3Zt4qPw7ugkLL79StMZIDNzJ43pLO2rJP+OiPfzxHWtMaV3C4/nrJ4I0dI+QIMHD0ZmZibefPNNpKWloX379pg9e7a7CezMmTOSTlwNGzbEnDlzMH36dAwfPhz169fHhAkTcNddd7m3OXv2LB555BFkZWUhJiYGXbt2xRdffIGYmJiAv7/qSO3HoXQhCzEwCUdYiB1928RhXNfG7pW4HXab5asEiw1sVw+r9isvRNq9WR10bBCNXWf128iBsqGiZhvEjFx4zSyFIWZ2vh35Mc0EUEpD7pVe7xS8mQhRSmsdt3M5hXhg8U7c1LkRxqQYHx1q5Csp/gwKSpyGh/Nn5Rfju93nNPdpNlCTL69SGSFVYGkabzx9fRuczMrHnnMVfW2UvvK+nMXe38RlVZu/yJUpMjuRpd0mzalr3bBNuqoZ3l9XNi3KVc3resz8rBU0u5rArM4AWd4Jevz48Rg/frzic/Pnz/d4LCUlBV988YXq/l5//XWflS3Y/Wdoezz93R7JY2pzUcTUDPW4izJySnHdBXVqWAtAWQBks2lngEYkNcCS7WXrCQUiTgrVCeTM3MnZZZOHGWHkHLFw8ylcEaM+ckptHyWVuAN7edVBzQVa5ZRmHFadCbqS2Qutflfvrj2KI5l5ePXngxiT0sinTWAlsiYwoxmgEqegOCQdAE5mFWDcx5tQYrCcRaUC/u+7PR6L+1bG5ZgBsttsiAoPwaD29aQBUBXv16THzKrvZm8kHSYyUyF2G36692r8uO88BrarZ+o86ArifD3TuFmXZ1hPAXF9Qjz6tpZOUxChcMfRtWlt/N+AtogIdWDx7Ve6HzfT3GPm/No6rqKNOxA/n1CNTtk2m0117gwl3txpGjmJbTl5UfVi69RoAnNlgLy5vC3eZq5JRikoUboYlTiFSvcB0moCk2e9jB7JyLVE/BnkFxvPAOnZf/4SDqcZH5H14740bDie5ZNjA/o3AVWSSpEv9wAoxMR51XQTmE06L5HekerUDMWYlMbuBWLFtA7tOg9anQFiAESmKK1H896YZDSpU5Z9aFa3Igth5pRpZpIx8clYCEAIpNWUZ2bBxbLtzV9IjN7EqTVnafYBCuByzEqzACuV6nxuEV5edaBSxyrQGP4rv5gb78ekv5149FqpU8CxTGNr2lV1ZibTqypcn7L8RkwtAGoRU1Px8arGzE2U2QyQ3R6Y2fVdTWBWB6MMgEhTw1rSRf7MrHqs9Dv66vYr0Squ7EQjvhB1b1YXfVvH4q6rm+nuN0yUcglEE5g4/S9vZrJBmpLWayrwpiXB6ElM/W5KvQ/QsQv5+Gbn2YCciMxMB1DZYfCu5rYJVzb1eE6e0fNd+OPZ8Xjb6WyVLUnMH6Ow1K7jaguofjiuM+aO6+weWWXmXBdIZgIgs+dHzz5A5l5v9OANyke2xhsYKONPlvcBoqptcmpzZOYVYVD7egD0VyQWs8GGGTd2lKx7FOKw4dmBCZj1x1FMvaZi5WGH3YZXb9AfUQV43sHPvjkZd362zXC5zBJ3AJUPu7fZbJK7iMgwh+YyDN41gRnbTq25xSmozwOUmVeMF8onTRNrVCscp7ONz0RdOyJEZ/kJ4ICJJpzKcmWbaoZ5fl/XH7sg+Vtr4VQxI5+Dr5q8gk3dGqHI9mLCUC2ubKv8F6cW7EeFhyCxYS28NKw9Fmw9g+Htq+bUKWZOIaYzQDbp/isV/2g8N7ZLY3RpWhu9OzbCBROTQfpa1QxxqcqICg/Bv4e0R6+WZX2BIgwsB+BmA65pFYtnBrRxPxRqt6FDg2jMHNlJccE/I0JlGaDkxrW92o/h44nOCPGyOURsNmkGSGtNIcC7JjDDGSCNYdRmMzz1ZZk/JeIFJB/r18rU/v3N1QRmZFLNyZ9tNbTP3QZG+lW283aw8uc8Q+3rS5dA0vst1IsOx4sjOqGVl+cnfzPTId1sYtcGm8+awLROWyEOOxIb1rJ89B0zQGRKRKj0gtKtqXrw4fpqiy9CZmZ5ViNpAgtQH6BXh3fAvvO5qFUjFH+JOpfaIF2EUG1Yqos3v/e4yDBcKtLvS6KWfRA0hsGrCTVR0IhQu+KEbFZyLeSr13m3uNRZ6eY2MV/uqzpoEVMTR1SW5OjcuBa2li84GxOpv2aaWa6bjeTGtfHajR3RtLyf4uX+CfmzD5BH7HM5TLVfCcwAkSljUhoDAHq3jMEbIxLxX42JAN0paNFvyMjcQHrE+wjIMHh72VxFd/e8wuMHY7PZJEGNXgbIG+O6Nja0nVYnaLOjLUINDG1zZTtWPtJHN/Czit7dsq+XKWEGSEptcMMD17RAxwYVQXNdhVFElSX+5Hu3isUVsZdHJ2c9ZrLIvVqam/9OaUX56owBEJnSrG4NrJ6aihk3dkTPljGIVFlDB1C+efDFfCLSUWBl1Nb48QVxwKWUHhbfkZ03sYK7URGhDgxPrK+7ndqcPloTIaoxMuzZFVSFOuym5iYJJL0J/JRGppk1Krmh+ztQmYklqyO1pq1aESGSGwejGSDxb+2fA9pqbltdr+XyLLyWJ/u30d9IRF5n/uoDVFUwACLTIsNCDLUTK23hiwBIcpdS/ivrafJOxwxxmRUX+BOVJ0tn8rmUJub7K9lsxoamqmaAoD7yRY2RDJArpgp12KvsRHmByAC1jK0JV7xYXTtBN62j3ydMSWxN5cAmOiJU8p2uH2VsNFCH+hWrfNe5TNcnq6zJVzdHs7o18MA1LXS3jTK5xIc3fRQvZwyAyG/0siXeEl/UAtEHSDxsWjkAqvi3Vh+QmJqheOK61qaPb4PNUGpavQ+QuTW7AHMT34U4zK9rFCh6AdBFH8yWbLdVrJ9U2RXYq6rXvcywxkaqZIDCpRmg+GhjGSDxIFS9U0lV/U5WVlxUOBbffiVuVZjiobI8V3j3ntlzjhUYAJHfuH484t+B0REG4gkV5cRBVCB+Y5IiK5a/4rH+Ggun3t3zCtSKMH/XaoOxztNqkxoKEFSHwavRmv1aaVurR3Oo0ctk6a2sboSAiottVW4Cq0w/m0gvFwvt0qSO4uO1IkIkzaYxKpkiOfH37AqdiQuNxj9av9lgE4hJEKsSBkDkN5X5Lb05Sv2OUzySTOtyk9SkNuaMTfa+EOXEy38o/WDE7/OBPi0Vtijjbf9vm81Y5kxtAVGnYH7NHTMZoFCHzev35m96GaALecbXMlNzqbBE1Aeo6maAFt7W1evXmll+wWVSj6bocUVdxeciwx2S7EIdg8GZ+HfQpE4NvDs6CbEq/Yfkc3Yp+eneq/Gfoe0NHbs6ur/XFWguutn0yABV4nd9GSSAGABR1dS4tsEMkMY+Qh32Ss8R9Mn4LpLJH5XS6uKThtZoKNfdVUuTo1HKRpoZaAJTa37RWApDjZE+QC4Ou63qdoLWKdfxCwWVPkZOYan786nKfYDiIsNQ08vRet5k+Aa3V+643yY+EvWjIyQd8yMNjp6U/w66NauDViq/JyMX7zo1Q4Mu6+HSrG4NTOzRDF9M6uZ+zLMuqnfdMAAiv/HXiUVyMta4sPuiY26CbBI1vfOB1iFd/XjuuEp/uQ/5IY1UpdpQdwGC6QnRzGSAjARo3ZvVMVcAH9F7H0cV5qj5e0ojj0WAteSKMkDbdZa+6NggWvN5f/t2cnevXudNAKQUFI9IaoBPbu2CELtN0lyoNcN8m/hIXNc2DnPHdVYsh9pXr4rG5AEXFS4NLnu1jMHdqc3xv9FJAKSnNPm2lXEZJIAYAJH/uE5ACfWitDc0SZyc0PqR+WLOITmlk6o40NNKu7ta7vSGZnvu39j8HGrNL1qrwasxW0Yt47o2xotD2+MhleZBf16n9CbeVJqk77F+rT3WwNOSW1jibgLUW1YjzOK2Qm/6oAHeBRNKr0lqVMsdLIunbdC6WWlUKwIvDeuAxIa1FLdT+80x/ikzZ2xnDOtYkY2rUyMUd17dHPXL1+Gy2Wx4qn9rTOndwiPzrrSUjFGXQxMYZ4Imn7u9R1N8+OcJ9wWvdXwk3h2d5LGMhLfE2QatH5nDy1mnr4ipgaMqq3grZTp0+0i7ylP+pJlZlsv2aayJSW0EmiAIqstkqPHVsPY5YysWl7ylWxOk5RZhwaaTkm3CQ+woKG++qxFqr9TcPHabdPp/vQxQlkonaDPJyxs6NfDI/AxIiEdhiRNncwrRsFY4Vh/MKC/P5XnP6c33QSlbM7hDxYVYnLHUyhaLg3fFTJTaS4O0aUuuZWwknh2UgG93nVPdZmRyI8nfD1zTAocz8tCtaR1TxxL//ioTPAUKAyDyuXt7tcCkHs0knYe7+akJRHxZf3FIO2w+eRGLt52p1D6vviIGRzNPKfZLUDqnGr02uIIns5kpG7zvQA0AG09cNP0aXzUfyMutdP2PCHW4A6Awh/kA6KbkhlhU/pmH2G2SQFDvwl2okjUzMyNu9+Z1PQLUVnGRuL28qfObHWcv+wDImz5e8jq8slkdyQ2EPCjv3zYOK/enm9onoBH/GCtm0DFyK+TtEPvXRyTiiW9246n+bdDjiro4kHYJI5IaerWvQLg8f41U5UX4cWkE8YlNnAEa0K6epJlFaeTTC4Pb6e5/cmpzPNW/NT5TGDWjmG4XN4FpnHVdFxGzd9Nlw+ADezr31fHkF06lO31x/w9vmt7EwbW8yUtvf2ozZJvtvyb/TMWZCnGRzPStqkq8+T7oBU3FspnL/zO0veJoMPEnZCYQ09u0qk7dcDlLbRGD1VN7YkjH+oiLDMPH47swACLyJemJS3oBE99hK/V7GdS+Hj4en6K5/4gQO0YmN0IDhX4gehl4rT5Armuf2QVh5SvOB8rU3i28ep3k4i+rD6Xrf56o34w3TS3iYEWeXfM24DBbDPn2kjoQlU8pA6TVAfhyplf18jmTbDYbJnbXzjwo7VM1ONMJ2i7TWLTS/P2+q+qs8Eqq5y+PqqWB7eJxXds4xInm/ZDHOOILj9rdfbzOtPtawYZSZkC8udY51/Vasxdl+YKrgWC32TChe1M83s/8zNWhKtkPQLn+xPUhPnmOSGpg6HjiQ8hPvmaDTfc+TVa4/CIsfrk46FG6OLSOi5T8/eXEbhjSoZ6p47sMVnjdrL8n4Zu7vBv9VRl6WSOlTvtjuzbG7Julc3eJf+NKv13VUWAqx532twREhjnwxkj/rR9YFU3p3QINa4VjcuoVVhelymAfIKqyProlBR/9eRxHMvJwS7cmklTqtW3isP5oJga2U79QqHWQ1psUUOu0rfScOOuj9VpX/4XLowms7P81TCy86BLqqOjULA94lPpwiINWcd3ckNgAS7af1TzWfb2ukAQrnk1RxsstZvZl8uYUcTl6tojBtW3ikJ5bhBYKc9bUjw7HrrM57r9r1whBTY1FhrUoBf1qszH7m14Tk9K0DXabDcmNa6Njg2h3nYiXu5nYoymOZubh+oR43eOr/WQGd6iPQe3rVdulMtTc1r0pbtPJsAUbBkBUZXVsEI1Xb+io+NzLw9qj1Clo9vFQG/qtNxpNq/+H4jndaCdo1zB4L5rAzHTK9QVXHXjTl0uc0ZGXW+ltiD8mcROWkQ7Dk3o0w++HM0T7l2dibB4jw4wwmwGSv0/xxb9mmAOvDO8AAFgoGwHnKqPaa8XaN6yFPWe05xkyMw9j58a1sPWU9v4qQy/A0BqZOHNkIvr/bx0A6fcjMizE45yg3gla63ccXMEPKWMTGF2WbDabbgdXtQDIZrNhXNfGXh/X2+ftXmeAAj/TsutobeMjNbf7740dPB4Tvz95dShd3MWfkjg4DDOYvpHOwyTlFASvRgMZH9lX/n+7Z+BllPxt2m3Kl+7vH+ytO5NzEy9XbfcHvTqUd4IWqy3qDK0Xu4o///fGJIme0HkhBT0GQFRtaU19Y2SdIOXXeZL0AdJ4rbsJzHQfoMDPauu6pjSPqYl3buqkuE23prXRt7XnQpLiAMgjA6SwH/Gq0eL+XaEhxt60Vt04nd7NSG40gHHtW6sTtNL2Wo+F2G2qzTd6xbqhUwPc3qNqNHPoNoEZnZtKZzPxUbqK5q2JMri8BgUvBkBUbQkasySqJRcW336l5j4VL2DiPkAa5/yKJjD9i6t45e6yAMj3EZDWrMTi43VvXhdDOnqu6xSlskK4ODNnJAMEAM8PTsCo5Ibon1ARUIUbzACJR5rJd9+gVrhX8+EZrW/X8eQXc7XPWC+ABrQDB71ihdhtuLdXC+2NAkSvDl0LpdaKUP4e9WlVthzJuG7msrX/HNgWCfWi8KDGwsREAPsAUTWmNgoMUM8K6C3KqLwUhv5+AXEGSPvCHh5ix+PXtcbT3+0p2yf8EwCFOuwoKlVeukF+tLtTm2OZbCZZtbqS9AEyMA+QIAB/a18ff2tfH9/vqThGmMHh4ZJdiv74ZUoqQh3ygfjGmJ3cUt6cY+bz8hxBphEA6bybqrSwp14d3nZlUzSIDkf35nUVn395eAdkXCpCvWjtUZvytzw8sQGGJxobQUjBjRkgqra0Or6qJT/0OhvrXdi0nnVdnMTZAeW1jeTNajavRzNp0cpEyS+kSluqZYBCRf145K/TuyiKL/BG+wCJPxPx/l0Bmj+bwFzHk89po9oEpvCY/DunVUd6xYqp6d1aX/6gV+9hIXYMS2zgXpNKzmG36QY/AEwv10DkwgwQVVtaw93VTs66/U4UnhZfLLXXAis/hixDojQc2Cbbpz8yQFr7VOqYK6eWAQrRyAApBZjidy9+NsRuw8fjUzDhky2q5QSkdS79LGwe+zTKaHW7F/aUDb9S7wOk/5hW4KBVrPrR4QFfaiPUYfMI/gJtVOdGiAx3oHPj2paWgy4/zABRtaW1ArpapiciRLsJTOlVNp3nXSqWwtCeGM9mk/4wbTbpdt2b1cFj17bSLKcRWn1NtCb2c1HPAHkGIWp/A+p9tex2G9rXj1Yto4te1s6/fYDKtpMvRGtmwVvFBXZVjq8VHDU1OQJMXu3XtvHs0K7nK50+c4EQYrdhaMcGaFKnhv7GRCIMgKjaKtUIgNSm4tEbuaK0R6N9gJSGwas1Q4mHVdtgkyyXMK5bE/Q3MBGcHq236jl/j0IGSCUAEk/iZ2QxVOlxxOUzFkRI6l/heW+yZ1rBs3TfZf+Xz2pcU7UvmWdZTC1wrlEWo32m1LjmKjKjQa0I9G4ZU6njElmFARBVW1pNYF43KVUi2+8KKsRBllrAJV9eQ9y0Ybf5ZiFHM6ONFDNAsov8w31bokVMTdzds7loP/p9iSrbgGK0CdIMrQ70Yq73J2/G1OtML2bmszQTtBrx3pgkXBFTwz1/zv9GK095oMXsJJNEVQUDIKq22mo0n3h7nVTKDBhuLlH4tRmZL8Zmk97d2202nyw4qHXhNTJ6S54BGte1Cb6Y1E0yl4/8ECU6MxV712FZ+/VGPp9nBrRBeIgdd6eWBW9K/bK0ji3PANVQmbBQuQ+QcjOo4R2UC5EFyUZ0bVoHX0660j1/zpXN6hp7oYjRYJGoqmEnaKp25o9PwZIdZ/HU0I5AYZHiNt4OF9ZrAtOidCFWHgVmkzQd2WCTzNnjqwyQVmBgqA+QSpZDHCPKh20rNUtKtzdPaybosuf199GufjRWT+3p/jyMXtTV+gCpZYCUm+gMHUp3W4ckaLapL4bnYyUBOg6RrzEDRNVOu/rRePr6NojXGELrbfhg5lwvv1gZDYAECB4X9TCHNAOkFbzIV9M2Wj6t5xRHgan0AZKM6pK9TCmwMNrfRo24rIodihVeI1+U1CHLqhkNgLo1q6P4uNpipkofm5nmWK0tQ2RBsi/8PaWR6kzgLnqLCxNVVQyAKCh531dEYci67LKU1KgWgLJlCcSULkpqC6PK+wDJm8C0MkB6Ey2K92P0OaVN1TNA6hdEvcCidg3zSWl5c6ERd/RoJvlbXp96TWBDOtbHfb2uwD+ua634fESo8mdQ2U7aWpuKAzhfTZvwWL/WqhMVuhidr4moquE3l8gEpeui/Frz+oiOeHFIOzzSVzpUXSkuUBuNJp/PJkzeCVp0zHCN0T8P91VfDkAzADKwuKfaMPiYmhV9gOR9YfQCoG5N62Bsl8Z4ZkAbze3E9Dr/Kh1xYPt6eOr6imPIAyC9cjapHYFJPZqp1oFq3Sr2UVLYTCXXozUTtPg9GAl/vM3b1I4IwQ2dGmD++BQAwOPXtUazujU8An6iqo59gCgo+bQPkOzvWhGhGNCunsd2SlkFxYAK0iYupSYwcflDHTYUliiXd2yXxmgZWxNTF+/0PI5mfxJ5mTypDfUOC7Fj1f1XK2aqlAILSR8gmw2PmJzjSG8YvFoTWzPRvDHyJIbWFAqAbyemlAebWgKdAVJSLzoczwxo6/67Wd0aWHz7ldh26iK+3nHWb8cl8jVmgCgoKV0eRiU31H+hwoVR61rz7uikipcqhE9q/Sc0m8BkF0y1ZrSy16o3l2ldIj2bwCr+7t6sDv57Q0fNi2ytiFDF7IhSh1mlejFDuhSGQp8qld1rrVqvt1K51mfeOi5S/XWif7erF4V/D26neBJW27/4/ckzLuL348f4x9AIRNdCpmLsKURVDQMgCkryC8TSO6/EEyr9OcSUm8DULwjdmtXBDZ0aoFfLGLSIqenxvFp2wrMTtHrzhjzAqR8Vpvm8kXJrjQIb1L4e+rT2vMAZoZcB8obNpvxvrWMC0iVJ5Bd1vQyQVgjgahrSM//WLhjYvp5iBshI/PJ/10ubCcWf8+P9yr7LE65saqgsZqh9n8RVNn1Ye58fl8jX2ARGBKBxbWPT6BtpApMTNxfIqV2c5XPbiDNA8tfIm6viosLx7ugkRIY7yp9XLqHWnbw8qeSrVcb9MWeMXsnUjijOnMmDEL1O0PL6uPOqZpi9/jge7ttSsxO6UlnNTGCoNeeR+PMc0rE+rm5RF3Vr+H5xVLXyimvMn01wRL7CDBBVe83qlgU3qS0qRrN4e4JWGuFUmXN9qQD8Z2h7/D2lkeRx+ezG4j5A8ouzUiDTrVkd9zpaaoFOn9axaBVXEzcqdF7VygBVJoRRzABVYn9ySp+rWtAl3lR+UW9fP0rzOPKjTE5tjm/v6o5xXZtovu6a8sxZxwYVk3Sa+f7Igx7xfuSfc0zNMJ8FrgAwrmtjAMD9vVsoPi9uyvTVMHwif2IGiKq9/41OwvLd5zCiU0UfH2/Pz+L459u7uldqX0BZE9j1CfG4PiEen2857X5ccqcP6cWtxCmddVhv2Ltak0V4iB2f3dYNALBU1nnVYy0wzSMY17aeev8YNZ0a1sKOM9mGtlW63qsNy5cEQLI6uqFTQxSWOPH66sOGjmOz2dCglv5ipDE1w/DrAz0lAa3agrhGzL45GVe/8TsA85NjavVVAoCaofKlTlrh7tQrVDu/yzuzE1V1zABRtVc/OhyTejRDnZoVzQG+WArMdcGrzMlePTsh7vMjHfUlzwDpXfeMXBg/HNsZwxPri46vXp7KGNyhPv5xXWssnNDF/ZjWvEEA8NZNie65lfQoDRNXa80SbysPQkLsNoxJaax+nErUR41Qh+QzGZGk1Plerd+W9O8QSSBl7HT+8fgUjOvaGFOvUc7k/G90J7SJj8Q7CuuCqS/ySnT5YQBEQcq7C5hSBqMyoYFSAGSzyZc1kD4v768kvpje1+sKj/0ZCYA6NaqFAQkVQ/e9WVjTCLvNhps6N0Kb+IomJr0msMiwELytMRux3kSIqvvXyAABnn2rxHzZxBMVHoLJqc31NwRg1/i2GV0frn39aDzct5XqHEZXNquLhRO6IrGhsaDThSti0OWGTWAUlLy9vreMjcT7f09GvGikVWViBdXshMLIpgW3dkFmXpG7T9PwxPpYf/QChic2cDfVTJLNcAwYzwyIj2lmbprKMnLhrBHqwA2JDfD1Tu15ZswUWxzkKS6hofHB+rqJ59ZuTZBfVOoeXdelaW18vOGEwnHV9+GL9eEqwxcL9BIFEgMgCkqVOVWnNKkt3VclLoZqw+DtsiYwAGhbT9ox958DE+AUBCzffU7zGEZXKhAfUyv74esJXYzuTm9oOmDuc20eUwOdGkYjOiLEdPDg60t9RKgDD/SpmLU79Yq6eGNkIh76SjqBpdGJEK2Q1LgWUlvURbO6ntM9EFVFDIAoKPnyBr4yu1Ibbq03t42L3rpggPHmLOkxq97dvJFFUxvXqYEtp4x1mLbbbJgztrPh4/duGYPfDmeWv9bwy7xis9nQs0UMosIdyC0srXhc49vmmvbAKnabDTNHai+cSlSVMACioCS+jDyg0hnU8L4q1QSmfFHX6gMkF6my8rh7XwYnQtQaFu5XBjuPGJlDaOo1LSAIAoZ0rK+7LWA80Hvnpk64slkddH/tN9crDb2usspGi4kCIIXDTundAptPZmGgwvIrRKSOARAFJ9GV5FY/zJZrlJE5arTu+gGgZ4sY9G0di3Yqc9cYvVSLO9ga7DbkE4abwJzKjzeIDgdQNqy/bo1QPPe3dr4pmIh8/bVAxYev3tARjy3dhYevbal63Nu6N8Vt3a37DhNdrhgAUVDy5fWrMrPeqiU15IuhanHYbXj1ho7qG6iUT/6oVU1gRkcPqfUBCguxY83UnrDblMvtsJVNOOmNtvGROJGVj8SG0ZLHA9XdJqlRLfxw71XuTul6wTARGccAiIKS1ZeRRrXCcTq70KNDtYt8MdTKULtYewZA4k7QAQyADG6ntnAsoD0/jcNuQ6mXEdD8W7ugpFSQLEUCBDZAFB/rchxo1bNFDNYeycTNGvMqEVmBARAFJas7Qb/392Qs2X5GdbI96VIYlSts7YhQ1KkRiqz8Ys3txEfRutBWdvV2bxkZBabEYZemgMzUpt1mQ1iIwhB5r0pSeTd3aYx/fb8PV11R16ISmPffGzvifE4hGtXWnymbKJAYAFFQ8uUdvDf7algrAvf1Uu98LV8KozIcdhuW390D9325HVtFI6TkxRYf0+o5ZZQYGQWmRN5E6YuP3qrqGdyhHtrVj0KzOsYW760KQuw2Bj9UJXEmaApKvu0D5MOdlfN1h9tQhx1vjeqEj8Z1Dtgx9cSUL00iXtBTi7crycvnx/HFW6sRas2Qc5vNhpaxkbrrvxGRPv6KKChVwWlu8I/rWgMAXhza3i9BVUSoAx01ljcI9DD4D27ujL+nNMLLwzsY2t7bjswe2axKvLcpvVugd8sY9GsT5/U+iKhqYBMYBSVfjqbxVXPaTZ0b4cZODRDisOOiTn8d35CWWzIMXuM9ta9vLGOjp1ndGnisX2vD22t1gtYiD4BcS4l4o2y4OYecE1UHDIAoKFndCVqNq2lDHIAEapFJvbXAvrmrO85lF3osyREo3jaBibNZ/drEYUrvyk18SUTVAwMgokryR2uReJ/+in/k5ZYEQArvqWGtCDSsZV1nVnknaMMTPIrejNHmNiKq/tgHiIKSTzNAfoiAxM02QoBSQOLDVGZyR3+Rz3b8oGjxUC1WLxJKRFUTM0AUlHzaB8hne1LeZ6Bm3REfJ6BrgRnUp3UcvpvcA3GRYThnYl6ZqvheiMh6zABRUPLlJbF/QjwAoGtT5VmdvWEPQBuYRx2IM0BV9MxQPzocDpPzylTFOY2IyHrMAFFQiokM9dm+4iLD8NsDPT2WS6gMewD6AMmJZ3iuik1g3roipgYOpl+yuhhEVMUwAKKg1K1pHUzq0RStYiN9sr8IH0+MJ+64G6ilJ8RHqU4B0OPXtUaow44RSQ2tLgoRVSEMgCgo2Ww2zaUorCYOPyJC/DPrsDzGEY8yd1Sf+AcxNcPw/OB2VheDiKoYBkBEVZDNZsPDfVsit7AkcOsoiYaBKc0DRERUnVje1XHBggXo168fOnXqhNGjR2P79u2a22dnZ2PatGno1asXEhMTMXDgQKxZs6ZS+ySqisZ1bYLJqVf4bf9xkWGSv8VNYAx/iKi6szQAWr58OaZPn477778fS5YsQbt27XDHHXcgIyNDcfuioiJMmjQJp06dwsyZM7FixQq88MILqF+/vtf7JAo2Lw9rjwlXNsU1rWIlj4vnAfLH3EZERFWJpQHQ3LlzMWbMGIwaNQqtW7fGtGnTEBERgcWLFytuv3jxYly8eBHvvPMOunbtiiZNmqB79+5o166d1/skCjb92sZj6jUtPIIctnoRUTCxrA9QUVERdu3ahbvvvtv9mN1uR2pqKrZs2aL4mp9//hmdO3fG888/j1WrViEmJgZDhw7FXXfdBYfD4dU+tfhziQPeYPsX69m89g2i0aVJbTSqHWG43ljPgcF6DhzWdWD4q57N7M+yAOjChQsoLS1FbKw0DR8bG4vDhw8rvubEiRNYv349hg0bhvfffx/Hjx/HtGnTUFJSgilTpni1Ty2xsb5Z9TrQ+6YKrGdzvprSy6vXsZ4Dg/UcOKzrwLCyni+rUWCCICA2NhYvvPACHA4HEhMTce7cOcyZMwdTpkzx+fEyMnJ8vhK3zVb2gftj31SB9RwYrOfAYD0HDus6MPxVz679GmFZAFS3bl04HA6PzskZGRmIi4tTfE18fDxCQkLgcFTMi9KyZUukpaWhqKjIq31qEQT47Qfgz31TBdZzYLCeA4P1HDis68Cwsp4t6wQdFhaGjh07Yt26de7HnE4n1q1bh5SUFMXXdOnSBcePH4fT6XQ/dvToUcTHxyMsLMyrfRIREVHwsXQU2KRJk/DFF19gyZIlOHToEJ577jnk5+dj5MiRAIAnnngCM2bMcG8/duxYZGVl4cUXX8SRI0ewevVqzJo1C7fccovhfRIRERFZ2gdo8ODByMzMxJtvvom0tDS0b98es2fPdjdXnTlzBnbRstQNGzbEnDlzMH36dAwfPhz169fHhAkTcNdddxneJxEREZFNENjKqSY93T+doOPiov2yb6rAeg4M1nNgsJ4Dh3UdGP6qZ9d+jbB8KQwiIiKiQGMAREREREGHARAREREFHQZAREREFHQYABEREVHQYQBEREREQYcBEBEREQUdBkBEREQUdC6r1eADzWbz3z79sW+qwHoODNZzYLCeA4d1HRj+qmcz++NM0ERERBR02ARGREREQYcBEBEREQUdBkBEREQUdBgAERERUdBhAERERERBhwEQERERBR0GQERERBR0GAARERFR0GEAREREREGHARAREREFHQZAAbRgwQL069cPnTp1wujRo7F9+3ari3RZmTVrFkaNGoWUlBRcffXVuO+++3D48GHJNoWFhZg2bRp69OiBlJQUTJ06Fenp6ZJtTp8+jcmTJyM5ORlXX301Xn75ZZSUlATyrVxW3n//fSQkJODFF190P8Z69o1z587hscceQ48ePZCUlIRhw4Zhx44d7ucFQcDMmTPRq1cvJCUlYeLEiTh69KhkH1lZWXj00UfRpUsXdOvWDU8//TQuXboU4HdSdZWWluKNN95Av379kJSUhP79++Odd96BeBUo1rN3NmzYgHvuuQe9evVCQkICVq5cKXneV/W6d+9ejBs3Dp06dUKfPn3wwQcf+OYNCBQQy5YtEzp27CgsWrRIOHDggPDMM88I3bp1E9LT060u2mXj9ttvFxYvXizs379f2LNnj3DXXXcJffv2FS5duuTe5tlnnxX69Okj/PHHH8KOHTuEMWPGCH//+9/dz5eUlAhDhw4VJk6cKOzevVtYvXq10KNHD2HGjBlWvKUqb9u2bcK1114rDBs2TPj3v//tfpz1XHlZWVnCtddeKzz55JPCtm3bhOPHjwu//fabcOzYMfc2s2bNErp27Sr89NNPwp49e4R77rlH6Nevn1BQUODe5o477hCGDx8ubN26VdiwYYNw/fXXC4888ogVb6lKevfdd4Xu3bsLv/zyi3DixAnh+++/Fzp37izMmzfPvQ3r2TurV68WXnvtNeHHH38U2rZtK/z000+S531Rrzk5OUJqaqrw6KOPCvv37xe+++47ISkpSfjss88qXX4GQAFy0003CdOmTXP/XVpaKvTq1UuYNWuWhaW6vGVkZAht27YV/vrrL0EQBCE7O1vo2LGj8P3337u3OXjwoNC2bVthy5YtgiCU/WDbtWsnpKWlubdZuHCh0KVLF6GwsDCg5a/qcnNzhQEDBghr164Vxo8f7w6AWM++8eqrrwpjx45Vfd7pdAo9e/YUZs+e7X4sOztbSExMFL777jtBECrqffv27e5t1qxZIyQkJAhnz571X+EvI5MnTxaeeuopyWNTpkwRHn30UUEQWM++Ig+AfFWvCxYsEK688krJeePVV18VBg4cWOkyswksAIqKirBr1y6kpqa6H7Pb7UhNTcWWLVssLNnlLScnBwBQu3ZtAMDOnTtRXFwsqedWrVqhUaNG2Lp1KwBg69ataNu2LeLi4tzb9OrVC7m5uTh48GDgCn8ZeP7559GnTx9JfQKsZ1/5+eefkZiYiAceeABXX301brzxRnzxxRfu50+ePIm0tDRJPUdHRyM5Odl93tiyZQtq1aqFTp06ubdJTU2F3W5nE3u5lJQUrF+/HkeOHAFQ1pyyadMmXHPNNQBYz/7iq3rdunUrunXrhrCwMPc2vXr1wpEjR3Dx4sVKlTGkUq8mQy5cuIDS0lLExsZKHo+NjfXow0LGOJ1O/Oc//0GXLl3Qtm1bAEB6ejpCQ0NRq1YtybaxsbFIS0tzbyO+KANw/+3ahoBly5Zh9+7dWLRokcdzrGffOHHiBD799FNMmjQJ99xzD3bs2IF///vfCA0NxYgRI9z1pHTecPW3Sk9PR0xMjOT5kJAQ1K5dm/VcbvLkycjNzcXf/vY3OBwOlJaW4uGHH8bw4cMBgPXsJ76q1/T0dDRp0kSyjetckp6e7r4B9gYDILosTZs2DQcOHMDChQutLkq1c+bMGbz44ov48MMPER4ebnVxqi1BEJCYmIhHHnkEANChQwccOHAAn332GUaMGGFx6aqP77//Ht9++y1mzJiB1q1bY8+ePZg+fTrq1avHeg5ybAILgLp168LhcCAjI0PyeEZGhsddMul7/vnnsXr1asybNw8NGjRwPx4XF4fi4mJkZ2dLts/IyEB8fLx7G/loJdffrm2C3a5du5CRkYGRI0eiQ4cO6NChA/766y/Mnz8fHTp0YD37SHx8PFq1aiV5rGXLljh9+rT7eQCa5424uDhkZmZKni8pKcHFixdZz+VeeeUVTJ48GUOGDEFCQgJuvPFG3HbbbZg1axYA1rO/+Kpetc4llb1+MgAKgLCwMHTs2BHr1q1zP+Z0OrFu3TqkpKRYWLLLiyAIeP755/HTTz9h3rx5aNq0qeT5xMREhIaGSur58OHDOH36NDp37gwA6Ny5M/bv3y/5Uf7xxx+IiopC69atA/I+qrqrrroK3377LZYuXer+LzExEcOGDXP/m/VceV26dHH3S3E5evQoGjduDABo0qQJ4uPjJfWcm5uLbdu2uc8bKSkpyM7Oxs6dO93brF+/Hk6nE0lJSQF4F1VfQUEBbDab5DGHw+EeBs969g9f1Wvnzp2xceNGFBcXu7f5448/0KJFi0o1fwHgMPhAWbZsmZCYmCh89dVXwsGDB4V//vOfQrdu3SSjZEjbv/71L6Fr167Cn3/+KZw/f979X35+vnubZ599Vujbt6+wbt06YceOHcLf//53xeHZt99+u7Bnzx7h119/Fa666ioOz9YhHgUmCKxnX9i2bZvQoUMH4d133xWOHj0qfPPNN0JycrLw9ddfu7eZNWuW0K1bN2HlypXC3r17hXvvvVdxGPGNN94obNu2Tdi4caMwYMCAoB+eLfaPf/xD6N27t3sY/I8//ij06NFDeOWVV9zbsJ69k5ubK+zevVvYvXu30LZtW2Hu3LnC7t27hVOnTgmC4Jt6zc7OFlJTU4XHH39c2L9/v7Bs2TIhOTmZw+AvN/Pnzxf69u0rdOzYUbjpppuErVu3Wl2ky0rbtm0V/1u8eLF7m4KCAuG5554TrrzySiE5OVm4//77hfPnz0v2c/LkSeHOO+8UkpKShB49eggvvfSSUFxcHOi3c1mRB0CsZ9/4+eefhaFDhwqJiYnCoEGDhM8//1zyvNPpFN544w0hNTVVSExMFG677Tbh8OHDkm0uXLggPPLII0Lnzp2FLl26CE8++aSQm5sbyLdRpeXk5Aj//ve/hb59+wqdOnUSrrvuOuG1116TDKtmPXtn/fr1iufkf/zjH4Ig+K5e9+zZI4wdO1ZITEwUevfu7bPpY2yCIJoOk4iIiCgIsA8QERERBR0GQERERBR0GAARERFR0GEAREREREGHARAREREFHQZAREREFHQYABEREVHQYQBERGRAQkICVq5caXUxiMhHuBo8EVV5Tz75JJYsWeLxeK9evTBnzhwLSkRElzsGQER0WejduzemT58ueSwsLMyi0hDR5Y5NYER0WQgLC0N8fLzkP9dq0AkJCVi4cCHuvPNOJCUl4brrrsOKFSskr9+3bx8mTJiApKQk9OjRA//85z9x6dIlyTaLFi3CkCFDkJiYiF69euH555+XPH/hwgXcf//9SE5OxoABA7Bq1Sr/vmki8hsGQERULcycORMDBw7E119/jWHDhuGRRx7BoUOHAAB5eXm44447ULt2bSxatAhvvPEG/vjjD7zwwgvu1y9cuBDPP/88xowZg2+//Rb/+9//0KxZM8kx3n77bfztb3/DN998g2uuuQaPPfYYsrKyAvk2ichHGAAR0WVh9erVSElJkfz33nvvuZ8fNGgQRo8ejRYtWuChhx5CYmIi5s+fDwD47rvvUFRUhJdffhlt27bF1VdfjWeffRZff/010tPTAQDvvvsuJk2ahNtuuw0tWrRAUlISJk6cKCnDiBEjMHToUDRv3hyPPPII8vLysH379oDVARH5DvsAEdFloUePHnjuueckj7mawAAgJSVF8lznzp2xZ88eAMChQ4eQkJCAmjVrup/v0qULnE4njhw5ApvNhvPnz+Pqq6/WLENCQoL73zVr1kRUVBQyMzO9fUtEZCEGQER0WahRowaaN2/ul32Hh4cb2i40NFTyt81mg9Pp9EeRiMjP2ARGRNXC1q1bJX9v27YNrVq1AgC0atUK+/btQ15envv5zZs3w263o0WLFoiKikLjxo2xbt26QBaZiCzEAIiILgtFRUVIS0uT/CduflqxYgUWLVqEI0eO4M0338T27dsxfvx4AMCwYcMQFhaGJ598Evv378f69evxwgsv4IYbbkBcXBwAYOrUqZg7dy4+/vhjHD16FLt27XL3ISKi6odNYER0Wfjtt9/Qq1cvyWMtWrRwD3efOnUqli9fjmnTpiE+Ph4zZsxA69atAZQ1n82ZMwcvvvgibrrpJtSoUQMDBgzAk08+6d7XiBEjUFhYiI8++givvPIK6tSpg0GDBgXuDRJRQNkEQRCsLgQRUWUkJCTgnXfeQf/+/a0uChFdJtgERkREREGHARAREREFHTaBERERUdBhBoiIiIiCDgMgIiIiCjoMgIiIiCjoMAAiIiKioMMAiIiIiIIOAyAiIiIKOgyAiIiIKOgwACIiIqKgwwCIiIiIgs7/AwNisSBI+NtXAAAAAElFTkSuQmCC","text/plain":["<Figure size 640x480 with 1 Axes>"]},"metadata":{},"output_type":"display_data"}],"source":["result_DistMult = pipeline(\n","    model='DistMult',\n","    loss=\"softplus\",\n","    training=training,\n","    testing=testing,\n","    validation=validation,\n","    model_kwargs=dict(embedding_dim=3, random_seed=42),  # Increase the embedding dimension\n","    optimizer_kwargs=dict(lr=0.1),  # Adjust the learning rate\n","    training_kwargs=dict(num_epochs=1000, use_tqdm_batch=False),  # Increase the number of epochs\n",")\n","\n","# The trained model is stored in the pipeline result\n","model_DistMult = result_DistMult.model\n","\n","#plot loss\n","result_DistMult.plot_losses()"]},{"cell_type":"code","execution_count":46,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T22:22:24.697068Z","iopub.status.busy":"2024-10-12T22:22:24.696025Z","iopub.status.idle":"2024-10-12T22:22:24.701555Z","shell.execute_reply":"2024-10-12T22:22:24.700663Z","shell.execute_reply.started":"2024-10-12T22:22:24.697024Z"},"trusted":true},"outputs":[],"source":["entity_representation_modules_DistMult: List['pykeen.nn.Representation'] = model_DistMult.entity_representations\n","relation_representation_modules_DistMult: List['pykeen.nn.Representation'] = model_DistMult.relation_representations"]},{"cell_type":"code","execution_count":47,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T22:22:27.547531Z","iopub.status.busy":"2024-10-12T22:22:27.546807Z","iopub.status.idle":"2024-10-12T22:22:27.553236Z","shell.execute_reply":"2024-10-12T22:22:27.552245Z","shell.execute_reply.started":"2024-10-12T22:22:27.547491Z"},"trusted":true},"outputs":[],"source":["entity_embeddings_DistMult = entity_representation_modules_DistMult[0](indices=None).cpu().detach().numpy()  \n","relation_embeddings_DistMult = relation_representation_modules_DistMult[0](indices=None).cpu().detach().numpy()  "]},{"cell_type":"code","execution_count":48,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T22:22:29.048161Z","iopub.status.busy":"2024-10-12T22:22:29.047790Z","iopub.status.idle":"2024-10-12T22:22:29.053115Z","shell.execute_reply":"2024-10-12T22:22:29.052198Z","shell.execute_reply.started":"2024-10-12T22:22:29.048124Z"},"trusted":true},"outputs":[{"name":"stdout","output_type":"stream","text":["Number of relation embeddings: 21\n"]}],"source":["print(f\"Number of relation embeddings: {relation_embeddings_DistMult.shape[0]}\")"]},{"cell_type":"code","execution_count":49,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T22:22:32.278067Z","iopub.status.busy":"2024-10-12T22:22:32.277686Z","iopub.status.idle":"2024-10-12T22:22:32.359104Z","shell.execute_reply":"2024-10-12T22:22:32.358253Z","shell.execute_reply.started":"2024-10-12T22:22:32.278029Z"},"trusted":true},"outputs":[{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"673d8ce550314bb8a307388aacf465c9","version_major":2,"version_minor":0},"text/plain":["Evaluating on cuda:0:   0%|          | 0.00/93.0 [00:00<?, ?triple/s]"]},"metadata":{},"output_type":"display_data"},{"name":"stdout","output_type":"stream","text":["Hits@1: 0.03225806451612903\n","Hits@3: 0.478494623655914\n","Hits@5: 0.5\n","Hits@10: 0.5752688172043011\n","Mean Reciprocal Rank: 0.3207639157772064\n"]}],"source":["evaluator = RankBasedEvaluator()\n","\n","metrics = evaluator.evaluate(result_DistMult.model,\n","                             testing.mapped_triples,\n","                             additional_filter_triples=[training.mapped_triples, validation.mapped_triples])\n","\n","print(f\"Hits@1: {metrics.get_metric('hits@1')}\")\n","print(f\"Hits@3: {metrics.get_metric('hits@3')}\")\n","print(f\"Hits@5: {metrics.get_metric('hits@5')}\")\n","print(f\"Hits@10: {metrics.get_metric('hits@10')}\")\n","print(f\"Mean Reciprocal Rank: {metrics.get_metric('mean_reciprocal_rank')}\")"]},{"cell_type":"markdown","metadata":{},"source":["## CompleX"]},{"cell_type":"code","execution_count":50,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T22:23:40.779833Z","iopub.status.busy":"2024-10-12T22:23:40.778897Z","iopub.status.idle":"2024-10-12T22:26:31.474422Z","shell.execute_reply":"2024-10-12T22:26:31.473531Z","shell.execute_reply.started":"2024-10-12T22:23:40.779788Z"},"trusted":true},"outputs":[{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"0576cee1070c440dbcadfafa1b417b5a","version_major":2,"version_minor":0},"text/plain":["Training epochs on cuda:0:   0%|          | 0/1000 [00:00<?, ?epoch/s]"]},"metadata":{},"output_type":"display_data"},{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"198a4812fd554f23a71e5560ba20c87b","version_major":2,"version_minor":0},"text/plain":["Evaluating on cuda:0:   0%|          | 0.00/93.0 [00:00<?, ?triple/s]"]},"metadata":{},"output_type":"display_data"},{"data":{"text/plain":["<Axes: title={'center': 'Losses Plot'}, xlabel='Epoch', ylabel='softplus Loss'>"]},"execution_count":50,"metadata":{},"output_type":"execute_result"},{"data":{"image/png":"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","text/plain":["<Figure size 640x480 with 1 Axes>"]},"metadata":{},"output_type":"display_data"}],"source":["result_ComplEx = pipeline(\n","    model='ComplEx',\n","    loss=\"softplus\",\n","    training=training,\n","    testing=testing,\n","    validation=validation,\n","    model_kwargs=dict(embedding_dim=3, random_seed=42),  # Increase the embedding dimension\n","    optimizer_kwargs=dict(lr=0.1),  # Adjust the learning rate\n","    training_kwargs=dict(num_epochs=1000, use_tqdm_batch=False),  # Increase the number of epochs\n",")\n","\n","# The trained model is stored in the pipeline result\n","model_ComplEx = result_ComplEx.model\n","\n","#plot loss\n","result_ComplEx.plot_losses()"]},{"cell_type":"code","execution_count":51,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T22:26:43.595299Z","iopub.status.busy":"2024-10-12T22:26:43.594929Z","iopub.status.idle":"2024-10-12T22:26:43.681353Z","shell.execute_reply":"2024-10-12T22:26:43.680430Z","shell.execute_reply.started":"2024-10-12T22:26:43.595264Z"},"trusted":true},"outputs":[{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"19b5397bdf7e45258d3c5f52dd491a25","version_major":2,"version_minor":0},"text/plain":["Evaluating on cuda:0:   0%|          | 0.00/93.0 [00:00<?, ?triple/s]"]},"metadata":{},"output_type":"display_data"},{"name":"stdout","output_type":"stream","text":["Hits@1: 0.7956989247311828\n","Hits@3: 0.978494623655914\n","Hits@5: 1.0\n","Hits@10: 1.0\n","Mean Reciprocal Rank: 0.8862903714179993\n"]}],"source":["evaluator = RankBasedEvaluator()\n","\n","metrics = evaluator.evaluate(result_ComplEx.model,\n","                             testing.mapped_triples,\n","                             additional_filter_triples=[training.mapped_triples, validation.mapped_triples])\n","\n","print(f\"Hits@1: {metrics.get_metric('hits@1')}\")\n","print(f\"Hits@3: {metrics.get_metric('hits@3')}\")\n","print(f\"Hits@5: {metrics.get_metric('hits@5')}\")\n","print(f\"Hits@10: {metrics.get_metric('hits@10')}\")\n","print(f\"Mean Reciprocal Rank: {metrics.get_metric('mean_reciprocal_rank')}\")"]},{"cell_type":"markdown","metadata":{},"source":["## RGCN"]},{"cell_type":"code","execution_count":null,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T22:51:33.202415Z","iopub.status.busy":"2024-10-12T22:51:33.201990Z","iopub.status.idle":"2024-10-12T22:55:11.165915Z","shell.execute_reply":"2024-10-12T22:55:11.164574Z","shell.execute_reply.started":"2024-10-12T22:51:33.202373Z"},"trusted":true},"outputs":[],"source":["result_RGCN = pipeline(\n","    model='RGCN',\n","    loss=\"softplus\",\n","    training=training,\n","    testing=testing,\n","    validation=validation,\n","    model_kwargs=dict(embedding_dim=3, random_seed=42),  # Increase the embedding dimension\n","    optimizer_kwargs=dict(lr=0.1),  # Adjust the learning rate\n","    training_kwargs=dict(num_epochs=1000, use_tqdm_batch=False),  # Increase the number of epochs\n",")\n","\n"]},{"cell_type":"code","execution_count":61,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T22:57:43.862253Z","iopub.status.busy":"2024-10-12T22:57:43.861535Z","iopub.status.idle":"2024-10-12T22:57:44.269131Z","shell.execute_reply":"2024-10-12T22:57:44.268233Z","shell.execute_reply.started":"2024-10-12T22:57:43.862201Z"},"trusted":true},"outputs":[{"data":{"text/plain":["<Axes: title={'center': 'Losses Plot'}, xlabel='Epoch', ylabel='softplus Loss'>"]},"execution_count":61,"metadata":{},"output_type":"execute_result"},{"data":{"image/png":"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","text/plain":["<Figure size 640x480 with 1 Axes>"]},"metadata":{},"output_type":"display_data"}],"source":["# The trained model is stored in the pipeline result\n","model_RGCN = result_RGCN.model\n","\n","#plot loss\n","result_RGCN.plot_losses()"]},{"cell_type":"code","execution_count":62,"metadata":{"execution":{"iopub.execute_input":"2024-10-12T22:58:05.009545Z","iopub.status.busy":"2024-10-12T22:58:05.008818Z","iopub.status.idle":"2024-10-12T22:58:05.088930Z","shell.execute_reply":"2024-10-12T22:58:05.088054Z","shell.execute_reply.started":"2024-10-12T22:58:05.009504Z"},"trusted":true},"outputs":[{"data":{"application/vnd.jupyter.widget-view+json":{"model_id":"ff21e1631d0046d590fb603908152d3b","version_major":2,"version_minor":0},"text/plain":["Evaluating on cuda:0:   0%|          | 0.00/93.0 [00:00<?, ?triple/s]"]},"metadata":{},"output_type":"display_data"},{"name":"stdout","output_type":"stream","text":["Hits@1: 0.016129032258064516\n","Hits@3: 0.05913978494623656\n","Hits@5: 0.11290322580645161\n","Hits@10: 0.3172043010752688\n","Mean Reciprocal Rank: 0.09552431851625443\n"]}],"source":["evaluator = RankBasedEvaluator()\n","\n","metrics = evaluator.evaluate(result_RGCN.model,\n","                             testing.mapped_triples,\n","                             additional_filter_triples=[training.mapped_triples, validation.mapped_triples])\n","\n","print(f\"Hits@1: {metrics.get_metric('hits@1')}\")\n","print(f\"Hits@3: {metrics.get_metric('hits@3')}\")\n","print(f\"Hits@5: {metrics.get_metric('hits@5')}\")\n","print(f\"Hits@10: {metrics.get_metric('hits@10')}\")\n","print(f\"Mean Reciprocal Rank: {metrics.get_metric('mean_reciprocal_rank')}\")"]}],"metadata":{"kaggle":{"accelerator":"nvidiaTeslaT4","dataSources":[{"datasetId":5861421,"sourceId":9611111,"sourceType":"datasetVersion"}],"dockerImageVersionId":30787,"isGpuEnabled":true,"isInternetEnabled":true,"language":"python","sourceType":"notebook"},"kernelspec":{"display_name":"Python 3","language":"python","name":"python3"},"language_info":{"codemirror_mode":{"name":"ipython","version":3},"file_extension":".py","mimetype":"text/x-python","name":"python","nbconvert_exporter":"python","pygments_lexer":"ipython3","version":"3.10.14"}},"nbformat":4,"nbformat_minor":4}