model_explanation_cyclicity.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "id": "1a91b813",
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    },
    "ExecuteTime": {
     "end_time": "2024-04-16T01:12:49.733396Z",
     "start_time": "2024-04-16T01:12:49.719814Z"
    }
   },
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2"
   ],
   "outputs": [],
   "execution_count": 1
  },
  {
   "cell_type": "code",
   "id": "c7f4188d",
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    },
    "ExecuteTime": {
     "end_time": "2024-04-16T01:12:57.975447Z",
     "start_time": "2024-04-16T01:12:50.579597Z"
    }
   },
   "source": [
    "from gnninterpreter import *"
   ],
   "outputs": [],
   "execution_count": 2
  },
  {
   "cell_type": "code",
   "id": "2ab76fb3",
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    },
    "ExecuteTime": {
     "end_time": "2024-04-16T01:12:57.996736Z",
     "start_time": "2024-04-16T01:12:57.976601Z"
    }
   },
   "source": [
    "import torch"
   ],
   "outputs": [],
   "execution_count": 3
  },
  {
   "cell_type": "markdown",
   "id": "395a07f1",
   "metadata": {},
   "source": [
    "# Cyclicity"
   ]
  },
  {
   "cell_type": "code",
   "id": "52747602",
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    },
    "ExecuteTime": {
     "end_time": "2024-04-16T01:13:00.771656Z",
     "start_time": "2024-04-16T01:12:57.997499Z"
    }
   },
   "source": [
    "dataset = CyclicityDataset(seed=12345)"
   ],
   "outputs": [],
   "execution_count": 4
  },
  {
   "cell_type": "code",
   "source": [
    "dataset.show(7)"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2024-04-16T01:13:00.906453Z",
     "start_time": "2024-04-16T01:13:00.772721Z"
    }
   },
   "id": "d1ef7ef233c222df",
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "data: Data(x=[96, 1], edge_index=[2, 192], edge_attr=[192, 2], y=[1], G=Graph with 96 nodes and 96 edges)\n",
      "class: red_cyclic\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 600x600 with 1 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "execution_count": 5
  },
  {
   "cell_type": "code",
   "id": "4c1617a4-666d-43de-8f0e-c5301b74375f",
   "metadata": {
    "collapsed": false,
    "pycharm": {
     "name": "#%%\n"
    },
    "ExecuteTime": {
     "end_time": "2024-04-16T01:16:43.738243Z",
     "start_time": "2024-04-16T01:16:43.680631Z"
    }
   },
   "source": [
    "model = NNConvClassifier(node_features=len(dataset.NODE_CLS),\n",
    "                         edge_features=len(dataset.EDGE_CLS),\n",
    "                         num_classes=len(dataset.GRAPH_CLS),\n",
    "                         hidden_channels=32)\n",
    "model.load_state_dict(torch.load('ckpts/cyclicity.pt'))"
   ],
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<All keys matched successfully>"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 6
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-04-16T01:16:55.671948Z",
     "start_time": "2024-04-16T01:16:45.355469Z"
    }
   },
   "cell_type": "code",
   "source": [
    "dataset_list_gt = dataset.split_by_class()\n",
    "dataset_list_pred = dataset.split_by_pred(model)"
   ],
   "id": "3ce389a1b5d24cab",
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/homebrew/Caskroom/mambaforge/base/envs/gnninterpreter/lib/python3.10/site-packages/torch_geometric/data/in_memory_dataset.py:300: UserWarning: It is not recommended to directly access the internal storage format `data` of an 'InMemoryDataset'. The data of the dataset is already cached, so any modifications to `data` will not be reflected when accessing its elements. Clearing the cache now by removing all elements in `dataset._data_list`. If you are absolutely certain what you are doing, access the internal storage via `InMemoryDataset._data` instead to suppress this warning. Alternatively, you can access stacked individual attributes of every graph via `dataset.{attr_name}`.\n",
      "  warnings.warn(msg)\n",
      "/opt/homebrew/Caskroom/mambaforge/base/envs/gnninterpreter/lib/python3.10/site-packages/torch_geometric/data/in_memory_dataset.py:300: UserWarning: It is not recommended to directly access the internal storage format `data` of an 'InMemoryDataset'. If you are absolutely certain what you are doing, access the internal storage via `InMemoryDataset._data` instead to suppress this warning. Alternatively, you can access stacked individual attributes of every graph via `dataset.{attr_name}`.\n",
      "  warnings.warn(msg)\n",
      "/opt/homebrew/Caskroom/mambaforge/base/envs/gnninterpreter/lib/python3.10/site-packages/torch_geometric/deprecation.py:26: UserWarning: 'data.DataLoader' is deprecated, use 'loader.DataLoader' instead\n",
      "  warnings.warn(out)\n"
     ]
    }
   ],
   "execution_count": 7
  },
  {
   "cell_type": "code",
   "source": [
    "evaluation = dataset.model_evaluate(model)\n",
    "evaluation"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2024-04-16T01:17:25.264432Z",
     "start_time": "2024-04-16T01:16:55.672948Z"
    }
   },
   "id": "b563ee40b6912eef",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'acc': 0.9720199704170227,\n",
       " 'cm': array([[2874,    0,   54],\n",
       "        [   4, 2647,  126],\n",
       "        [  69,   66, 5561]]),\n",
       " 'f1': {'red_cyclic': 0.9783830046653748,\n",
       "  'green_cyclic': 0.9642987251281738,\n",
       "  'acyclic': 0.9724578261375427}}"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 8
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-04-16T01:17:49.838450Z",
     "start_time": "2024-04-16T01:17:49.711514Z"
    }
   },
   "cell_type": "code",
   "source": "draw_matrix(evaluation['cm'], dataset.GRAPH_CLS.values(), fmt='d')",
   "id": "43b9aa358576e104",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 2 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "execution_count": 9
  },
  {
   "cell_type": "code",
   "id": "d4a27bdf-0586-4652-9989-53b8257bb2e9",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-04-16T01:18:03.066434Z",
     "start_time": "2024-04-16T01:17:52.590536Z"
    }
   },
   "source": "mean_embeds = [d.model_transform(model, key=\"embeds\").mean(dim=0) for d in dataset_list_gt]",
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/homebrew/Caskroom/mambaforge/base/envs/gnninterpreter/lib/python3.10/site-packages/torch_geometric/deprecation.py:26: UserWarning: 'data.DataLoader' is deprecated, use 'loader.DataLoader' instead\n",
      "  warnings.warn(out)\n"
     ]
    }
   ],
   "execution_count": 10
  },
  {
   "cell_type": "code",
   "id": "2c966c81e3166b9e",
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2024-04-16T01:18:30.407909Z",
     "start_time": "2024-04-16T01:18:30.383679Z"
    }
   },
   "source": [
    "trainer = {}\n",
    "sampler = {}"
   ],
   "outputs": [],
   "execution_count": 11
  },
  {
   "cell_type": "markdown",
   "source": [
    "# Red Cyclic"
   ],
   "metadata": {
    "collapsed": false
   },
   "id": "702a0c6a5a0480c5"
  },
  {
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-04-16T01:24:24.954307Z",
     "start_time": "2024-04-16T01:24:24.928225Z"
    }
   },
   "cell_type": "code",
   "source": [
    "cls_idx = 0\n",
    "trainer[cls_idx] = Trainer(\n",
    "    sampler=(s := GraphSampler(\n",
    "        max_nodes=20,\n",
    "        num_edge_cls=len(dataset.EDGE_CLS),\n",
    "        temperature=0.15,\n",
    "        learn_node_feat=False,\n",
    "        learn_edge_feat=True\n",
    "    )),\n",
    "    discriminator=model,\n",
    "    criterion=WeightedCriterion([\n",
    "        dict(key=\"logits\", criterion=ClassScoreCriterion(class_idx=cls_idx, mode='maximize'), weight=1),\n",
    "        dict(key=\"embeds\", criterion=EmbeddingCriterion(target_embedding=mean_embeds[cls_idx]), weight=1),\n",
    "        dict(key=\"logits\", criterion=MeanPenalty(), weight=1),\n",
    "        dict(key=\"omega\", criterion=NormPenalty(order=1), weight=2),\n",
    "        dict(key=\"omega\", criterion=NormPenalty(order=2), weight=2),\n",
    "        # dict(key=\"xi\", criterion=NormPenalty(order=1), weight=0),\n",
    "        # dict(key=\"xi\", criterion=NormPenalty(order=2), weight=0),\n",
    "        dict(key=\"eta\", criterion=NormPenalty(order=1), weight=0),\n",
    "        dict(key=\"eta\", criterion=NormPenalty(order=2), weight=0),\n",
    "        dict(key=\"theta_pairs\", criterion=KLDivergencePenalty(binary=True), weight=5),\n",
    "    ]),\n",
    "    optimizer=(o := torch.optim.SGD(s.parameters(), lr=1)),\n",
    "    scheduler=torch.optim.lr_scheduler.ExponentialLR(o, gamma=1),\n",
    "    dataset=dataset,\n",
    "    budget_penalty=BudgetPenalty(budget=10, order=2, beta=1),\n",
    ")"
   ],
   "id": "d87f5622bedd957d",
   "outputs": [],
   "execution_count": 12
  },
  {
   "cell_type": "code",
   "source": [
    "cls_idx = 0\n",
    "trainer[cls_idx].train(\n",
    "    iterations=2000,\n",
    "    target_probs={cls_idx: (0.9, 1.0)},\n",
    "    target_size=20,\n",
    "    w_budget_init=1,\n",
    "    w_budget_inc=1.1,\n",
    "    w_budget_dec=0.95,\n",
    "    k_samples=16\n",
    ")"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2024-04-16T01:24:36.084946Z",
     "start_time": "2024-04-16T01:24:34.522485Z"
    }
   },
   "id": "97b92f8fee7143da",
   "outputs": [
    {
     "data": {
      "text/plain": [
       "  0%|          | 0/2000 [00:00<?, ?it/s]"
      ],
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "65dd7f1c5c7e4732b11df2fe241f5174"
      }
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 13
  },
  {
   "cell_type": "code",
   "source": [
    "trainer[0].evaluate(threshold=0.5, show=True)"
   ],
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2024-04-16T01:24:46.228235Z",
     "start_time": "2024-04-16T01:24:46.139831Z"
    }
   },
   "id": "a7c082da89a24be6",
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "n=17 m=19\n",
      "logits=[101.8824234008789, -38.539520263671875, 16.946470260620117]\n",
      "probs=[1.0, 0.0, 1.2965373911318937e-37]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 600x600 with 1 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<networkx.classes.graph.Graph at 0x359ea3d60>"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 14
  },
  {
   "cell_type": "markdown",
   "id": "383c9c7a23eb249f",
   "metadata": {
    "collapsed": false
   },
   "source": [
    "# Green Cyclic"
   ]
  },
  {
   "cell_type": "code",
   "id": "20c1d36b425af72d",
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2024-04-16T01:47:07.580080Z",
     "start_time": "2024-04-16T01:47:07.490988Z"
    }
   },
   "source": [
    "cls_idx = 1\n",
    "trainer[cls_idx] = Trainer(\n",
    "    sampler=(s := GraphSampler(\n",
    "        max_nodes=20,\n",
    "        num_edge_cls=len(dataset.EDGE_CLS),\n",
    "        temperature=0.15,\n",
    "        learn_node_feat=False,\n",
    "        learn_edge_feat=True\n",
    "    )),\n",
    "    discriminator=model,\n",
    "    criterion=WeightedCriterion([\n",
    "        dict(key=\"logits\", criterion=ClassScoreCriterion(class_idx=cls_idx, mode='maximize'), weight=1),\n",
    "        dict(key=\"embeds\", criterion=EmbeddingCriterion(target_embedding=mean_embeds[cls_idx]), weight=1),\n",
    "        dict(key=\"logits\", criterion=MeanPenalty(), weight=1),\n",
    "        dict(key=\"omega\", criterion=NormPenalty(order=1), weight=2),\n",
    "        dict(key=\"omega\", criterion=NormPenalty(order=2), weight=2),\n",
    "        # dict(key=\"xi\", criterion=NormPenalty(order=1), weight=0),\n",
    "        # dict(key=\"xi\", criterion=NormPenalty(order=2), weight=0),\n",
    "        dict(key=\"eta\", criterion=NormPenalty(order=1), weight=0),\n",
    "        dict(key=\"eta\", criterion=NormPenalty(order=2), weight=0),\n",
    "        dict(key=\"theta_pairs\", criterion=KLDivergencePenalty(binary=True), weight=5),\n",
    "    ]),\n",
    "    optimizer=(o := torch.optim.SGD(s.parameters(), lr=1)),\n",
    "    scheduler=torch.optim.lr_scheduler.ExponentialLR(o, gamma=1),\n",
    "    dataset=dataset,\n",
    "    budget_penalty=BudgetPenalty(budget=10, order=2, beta=1),\n",
    ")"
   ],
   "outputs": [],
   "execution_count": 15
  },
  {
   "cell_type": "code",
   "id": "4c0bd9796d611c52",
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2024-04-16T01:47:27.014246Z",
     "start_time": "2024-04-16T01:47:19.569632Z"
    }
   },
   "source": [
    "cls_idx = 1\n",
    "trainer[cls_idx].train(\n",
    "    iterations=2000,\n",
    "    target_probs={cls_idx: (0.9, 1.0)},\n",
    "    target_size=20,\n",
    "    w_budget_init=1,\n",
    "    w_budget_inc=1.1,\n",
    "    w_budget_dec=0.95,\n",
    "    k_samples=16\n",
    ")"
   ],
   "outputs": [
    {
     "data": {
      "text/plain": [
       "  0%|          | 0/2000 [00:00<?, ?it/s]"
      ],
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "9c2f98684c684a0a84d84c911b54fe8c"
      }
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 16
  },
  {
   "cell_type": "code",
   "id": "7b268b4019492234",
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2024-04-16T01:47:28.806824Z",
     "start_time": "2024-04-16T01:47:28.740882Z"
    }
   },
   "source": [
    "trainer[1].evaluate(threshold=0.5, show=True)"
   ],
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "n=4 m=4\n",
      "logits=[9.884783744812012, 41.78472137451172, 13.449748992919922]\n",
      "probs=[1.3996953485031875e-14, 1.0, 4.946270172959122e-13]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 600x600 with 1 Axes>"
      ],
      "image/png": "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"
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<networkx.classes.graph.Graph at 0x359fab9a0>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 17
  },
  {
   "cell_type": "markdown",
   "id": "4f4d3ddd-9bb6-456c-9a68-5a3edd019e6e",
   "metadata": {},
   "source": [
    "# Acyclic"
   ]
  },
  {
   "cell_type": "code",
   "id": "a71e2ec8-f880-401d-9b89-0f7235d02c3f",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-04-16T01:47:51.213703Z",
     "start_time": "2024-04-16T01:47:51.188731Z"
    }
   },
   "source": [
    "cls_idx = 2\n",
    "trainer[cls_idx] = Trainer(\n",
    "    sampler=(s := GraphSampler(\n",
    "        max_nodes=20,\n",
    "        num_edge_cls=len(dataset.EDGE_CLS),\n",
    "        temperature=0.15,\n",
    "        learn_node_feat=False,\n",
    "        learn_edge_feat=True\n",
    "    )),\n",
    "    discriminator=model,\n",
    "    criterion=WeightedCriterion([\n",
    "        dict(key=\"logits\", criterion=ClassScoreCriterion(class_idx=cls_idx, mode='maximize'), weight=1),\n",
    "        dict(key=\"embeds\", criterion=EmbeddingCriterion(target_embedding=mean_embeds[cls_idx]), weight=1),\n",
    "        dict(key=\"logits\", criterion=MeanPenalty(), weight=1),\n",
    "        dict(key=\"omega\", criterion=NormPenalty(order=1), weight=2),\n",
    "        dict(key=\"omega\", criterion=NormPenalty(order=2), weight=2),\n",
    "        # dict(key=\"xi\", criterion=NormPenalty(order=1), weight=0),\n",
    "        # dict(key=\"xi\", criterion=NormPenalty(order=2), weight=0),\n",
    "        dict(key=\"eta\", criterion=NormPenalty(order=1), weight=0),\n",
    "        dict(key=\"eta\", criterion=NormPenalty(order=2), weight=0),\n",
    "        dict(key=\"theta_pairs\", criterion=KLDivergencePenalty(binary=True), weight=5),\n",
    "    ]),\n",
    "    optimizer=(o := torch.optim.SGD(s.parameters(), lr=1)),\n",
    "    scheduler=torch.optim.lr_scheduler.ExponentialLR(o, gamma=1),\n",
    "    dataset=dataset,\n",
    "    budget_penalty=BudgetPenalty(budget=10, order=2, beta=1),\n",
    ")"
   ],
   "outputs": [],
   "execution_count": 18
  },
  {
   "cell_type": "code",
   "id": "dcc3278f-a073-42e2-b446-0115025e391f",
   "metadata": {
    "ExecuteTime": {
     "end_time": "2024-04-16T01:48:03.083387Z",
     "start_time": "2024-04-16T01:48:01.217187Z"
    }
   },
   "source": [
    "cls_idx = 2\n",
    "trainer[cls_idx].train(\n",
    "    iterations=2000,\n",
    "    target_probs={cls_idx: (0.9, 1.0)},\n",
    "    target_size=20,\n",
    "    w_budget_init=1,\n",
    "    w_budget_inc=1.1,\n",
    "    w_budget_dec=0.95,\n",
    "    k_samples=16\n",
    ")"
   ],
   "outputs": [
    {
     "data": {
      "text/plain": [
       "  0%|          | 0/2000 [00:00<?, ?it/s]"
      ],
      "application/vnd.jupyter.widget-view+json": {
       "version_major": 2,
       "version_minor": 0,
       "model_id": "97428971157941c18a0924e2934d3174"
      }
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 19
  },
  {
   "cell_type": "code",
   "id": "10896fc3a8e2ea3d",
   "metadata": {
    "collapsed": false,
    "ExecuteTime": {
     "end_time": "2024-04-16T01:48:05.437584Z",
     "start_time": "2024-04-16T01:48:05.363070Z"
    }
   },
   "source": [
    "trainer[2].evaluate(threshold=0.5, show=True)"
   ],
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "n=16 m=19\n",
      "logits=[-4.710405349731445, -8.151561737060547, 4.258912086486816]\n",
      "probs=[0.00012723830877803266, 4.0751410779193975e-06, 0.999868631362915]\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 600x600 with 1 Axes>"
      ],
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAeQAAAHiCAYAAAA597/kAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguNCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8fJSN1AAAACXBIWXMAAA9hAAAPYQGoP6dpAABLNElEQVR4nO3dd3hT1f8H8HfSdFNKywZZZc+yZTjK3jJl7ykgijIdqPxQEEQQkd2yLaggG2QjCIKMsoeMAmUUKLSleyW/P67wLZB7c5MmNzfJ+/U8Ppaek+SjIu+ec8/QGAwGA4iIiMiutPYugIiIiBjIREREqsBAJiIiUgEGMhERkQowkImIiFSAgUxERKQCDGQiIiIVYCATERGpAAOZiIhIBRjIREREKsBAJiIiUgEGMhERkQowkImIiFSAgUxERKQCDGQiIiIVYCATERGpAAOZiIhIBRjIREREKsBAJiIiUgEGMhERkQowkImIiFSAgUxERKQCDGQiIiIVYCATERGpAAOZiIhIBRjIREREKsBAJiIiUgEGMhERkQowkImIiFSAgUxERKQCDGQiIiIVYCATERGpAAOZiIhIBRjIREREKsBAJiIiUgEGMhERkQowkImIiFSAgUxERKQCDGQiIiIVYCATERGpAAOZiIhIBRjIREREKsBAJiIiUgEGMhERkQowkImIiFSAgUz/YzDYuwIiIpfFQHZRWfoszDwyE1HxUf99Iwto2BD48kvgyRP7FkdE5II0BgOHRa4mKj4KfTb0wZ+3/kRIyRDs6bMHbuvWA926CR1y5QJGjgQ+/hgoUMC+xRIRuQgGsotZd3Edhm4ZitjU2Offm9Z4KiaOXAOcO/diZ29v4L33gLFjgSJFFK6UiMi1MJBdRGJ6Ikb/MRphEWGvtOnghiOLs1DnnsiLPT2BQYOACROA4sVtWygRkYtiILuAE/dOoOf6nrj65KponzKPgYhFQK50iTfS6YB+/YBPPgFKl7Z+oURELoyLupyY3qDHjMMzUD+svmQYA8C1vMDolibeMDMTCAsDypcH+vYFLl+2XrFERC6OI2QndffpXfTd2Bf7IvfJ6l8yFvj5d6BBlBkfotEA774LfP45ULWqZYUSEREAjpCd0oZLG1BtYTXZYdyrai+cbrsFDco1Me+DDAbg11+BatWADh2AkyfNL5aIiABwhOxUktKT8PHOj7H41GJZ/f08/LCgzQL0qtbrf9/8+29gyhRgxw7LimjVCpg0Cahf37LXExG5KAayk4i4H4Ee63vgyuMrsvrXf60+VndajaCAIOMdTp4Evv4a2LjRsoIaNxaC+e23haltIiKSxEB2cHqDHrP+noVP936KDH2Gyf5ajRafv/k5Jr09CTqtzvQHnD0LTJ0qTE1b8lvljTeEZ8zNmzOYiYgkMJAd2L2Ee+i3sR/23Ngjq39x/+L4udPPeKP4G+Z/2OXLQjCHhwvHbJqrbl0hmNu2ZTATERnBQHZQm69sxsBNA/E45bGs/t2rdMeCNguQxytPzj74+nXg22+BFSuADNMj8lcEBwvB3KkToOWaQiKiZxjIDiY5Ixljdo7BwpMLZfXP5ZEL81rPQ59qfaCx5sj09m1gxgwgNBRISzP/9RUrAp99JpyfrZMxdU5E5OQYyA7kdPRp9FzfE5diLsnq/3rR1/Fzp59ROtCGp2rduwfMnAksXAikpJj/+jJlgE8/BXr3BtzdrV8fEZGDYCA7AL1BjzlH52Di3olIz5I621KggQafvvkpvnz7S7i7KRRyDx8Cs2cDP/0EJCaa//oSJYCJE4EBA4Szs4mIXAwDWeXuJ9xH/039sev6Lln9i+UuhtWdVuOtEm/ZuDIRT54Ac+YIf8XHm//6okWB8eOBwYMBHx/r10dEpFIMZBXb+u9WDNg0ADHJMbL6d63cFQvbLESAd4CNK5MhPh6YNw+YNQt4LG/h2QsKFBCufRw+XLifmYjIyTGQVSglIwXjdo/DvOPzZPX3dffF3FZz0b96f+su3LKGxERg0SLgu++ABw/Mf31gIPDRR8CoUYC/v/XrIyJSCQayypx9cBY91/fEhUcXZPWvXaQ2wjuFo2zesjauLIdSUoQV2dOnA3fvmv96f3/ggw+ADz8E8ua1fn1ERHbGQFYJg8GAuf/Mxfjd45GWZXobkQYaTHxjIiaHTFZu4ZY1pKUJe5inTQNu3jT/9blyASNGAGPGCNPaREROgoGsAg8SH2DApgHYcU3ehQ5F/YpidafVCCkZYtvCbCkjQzj1a+pU4N9/zX+9tzcwbJjwnLloUevXR0SkMAaynW2/uh0DNg3Aw6SHsvp3rtgZi9stRqB3oI0rU0hWlnBO9jffABfkTdO/wMMDGDQImDBB2DpFROSgGMh2kpqZigm7J+DHf36U1d/H3Qc/tvwRA2sMVN/CLWvQ64Wbpb7+GoiIMP/1Oh3Qty/wySfCYSNERA6GgWwH5x+eR8/1PXHu4TlZ/WsVroXwzuEol7ecjStTAYMB2L5duJP52DHzX6/VAj17Cqd/Vaxo/fqIiGyEgawgg8GAecfnYeyusbIXbo1rMA5TGk+Bh5uHAhWqiMEA7NkjBPOhQ+a/XqMBunQRLrKoVs369RERWRkDWSEPkx5i0OZB2PrvVln9i/gVwaqOq9C4VGMbV+YADh4UgnmPvGsmX9G+vRDMtWtbty4iIitiICtg57Wd6LexHx4kyTsYo2OFjljSbgny+nC/7QuOHhWeMW/bZtnrW7YEJk0CGjSwbl1ERFbAQLahtMw0TNwzET8c+0FWf2+dN35o+QOG1BzinAu3rOXUKSGYN2yw7PWNGgnBHBIiTG0TEakAA9lGLj66iJ7re+LMgzOy+tcoVAPhncNRIV8FG1fmRM6fF7ZL/fKL8MzZXA0bClPZLVowmInI7hjIVmYwGLDwxEJ8vOtjpGamynrN2Ppj8XXjr+Gp47WDFrlyRTj5a/VqYV+zuWrXFkbM7doxmInIbhjIVhSTHINBmwdh85XNsvoXzlUYKzqsQLPSzWxcmYu4cQP49ltg+XLhJDBzVasmjJg7dxa2TxERKYiBbCW7r+9G3419EZ0YLav/O+XfQdg7Ycjnk8/GlbmgqChgxgxgyRLh7GxzVawo7GPu3l04cISISAEM5BxKy0zDZ/s+w/d/fy+rv5fOC7NbzMawWsO4cMvW7t8Hvv8eWLAASE42//Vlyggnf/XuLRzRSURkQwzkHLgccxk91vfA6ejTsvoHFwxGeOdwVMpfybaF0YsePQJmzwZ++glISDD/9cWLAxMnAgMGAF5e1q+PiAgMZIsYDAYsPrkYH+38CCmZKbJe81G9jzCtyTQu3LKn2Fjgxx+BH34A4uLMf32RIsC4ccDQoYCPj7WrIyIXx0A20+Pkxxi8ZTA2Xt4oq39B34JY0WEFWpRpYdvCSL6nT4H584Xp7JgY819foIBwH/Pw4YCfn/XrIyKXxEA2w94be9F3Y1/cS7gnq3+bsm2wtP1SFPAtYOPKyCJJScCiRcB33wHR8hbjvSAwEBg9Ghg1CsiTx9rVEZGLYSDLkJ6Vjs/3fY6ZR2bCANP/urx0XpjZbCZG1BnBhVuOICUFWLoUmD5dWKFtru++A8aOtX5dRORSGMgmXIm5gl6/98LJ+ydl9a9SoArWdF6DKgWq2Lgysrr0dGDFCuGQkchIea8JDARu3uTUNRHlGE8/EGEwGBB6KhQ1F9eUHcYf1P0Ax4ccZxg7Kg8PYMgQ4N9/hWAuJ+P+6Y8+gj6XL+48vYPIWJkhTkRkBEfIRjxJeYKhW4Zi/aX1svoX8C2A5e2Xo1XZVjaujBSVlQWsWydcZHH+/CvNSd46vP1VCZxLj0J6Vjo6VOiADd0svPCCiFweR8gv2R+5H9UWVJMdxq3KtMLZ984yjJ2RmxvQrRtw5gzw++9AjRovNM+qk4mTKdeRnpUOALgRe8MeVRKRk2Ag/ycjKwOf7PkETVY2wd2Euyb7e7p5Yk7LOdjWcxsK5iqoQIVkN1ot0LEjcPIkkjb8ir9fAxLdgTmvv9gtMjYSnHAiIktxyhrA1cdX0fP3njhx74Ss/pXzV8aazmtQtWBVG1dGapR3eiD878UiMvDVtkfjHvF8ciKyiEuPkA0GA5ZGLEWNRTVkh/H7dd7H8SHHGcYuLCiwtNEwBsCFXURkMZcN5NiUWHRb1w2DNg9CUkaSyf75fPJhS48tmNt6LrzdvRWokNSqVJ5Som18jkxElnLJu+UO3jqI3r/3RtRTeYdAtCjdAss7LEehXIVsXBk5gqCAINE2BjIRWcqlRsgZWRn4fN/nCFkeIiuMPdw8MLvFbGzvtZ1hTM9JjZAj4zhlTUSWcZkR8vUn19Hr9144dveYrP4V81VEeOdwVC9U3baFkcPhCJmIbMHpA9lgMGDlmZV4f8f7SExPlPWa4bWHY2bzmfBx5xV79CqpQOYImYgs5fSBPGbXGMw+OltW37zeebG0/VK8U/4dG1dFjqy4f3FoNVroDfpX2m7F3UKmPhM6rdP/r0VEVub0z5DfKf8ONDB941LToKY4O/wsw5hMcndzR7HcxYy2ZRmyEBVvwY1RROTynD6QQ0qGYELDCcIvDEDlB0DdO/9rd9e6Y2azmdjZeyeK+BWxT5HkcDhtTUTW5vSBjBs3MOV6cfyxLRDRM4HzC4AZu4Wm8nnL49jgYxjTYAy0Guf/V0HWw73IRGRtzvugKzoaqF8fuHkTOgAtsjXVjwLerzQA37afC18PX3tVSA5McoTM07qIyALOG8gFCwKpqUabPPTAXP/uAMOYLFQqQGKEHMcRMhGZz3nnaTUaoHFj8fa9e5WrhZwO9yITkbU5byADDGSyGU5ZE5G1OXcgN2ki3nbqFBAbq1wt5FTy++QXPTjmUfIjJKQlKFwRETk65w7kkiWBIJGRjMEAHDigZDXkRDQaDbc+EZFVOXcgA9LT1vv2KVcHOR3JSyY4bU1EZnL+QJaatuZzZMoBLuwiImty/kBu1Ei87dIl4N495Wohp8IpayKyJucP5IIFgSpVxNv371euFnIqPK2LiKzJ+QMZ4LQ12QSnrInImlwjkE3tRzYYlKuFnEbJPCVF2yLjImHg7ysiMoNrBPLbbwNakX/U27eBGxzNkPl8PXxR0Leg0bbUzFREJ0YrXBEROTLXCGR/f6BOHfF2TluThbiwi4isxTUCGeAxmmQTkpdM8DkyEZnBdQJZamHX/v2AXq9cLeQ0gvJwYRcRWYfrBHKDBoCnp/G2R4+A8+eVrYecAqesichaXCeQvb2FUBbDYzTJApyyJiJrcZ1ABrgfmayOe5GJyFpcK5ClFnb9+SeQmalcLeQUivoVhbvW3Wjb3ad3kZaZpnBFROSoXCuQ69QB/PyMtyUkACdOKFsPOTw3rRtK5ClhtM0AA27F31K4IiJyVK4VyDqdcEiIGE5bkwUkF3bxGkYiksm1AhngfmSyOl4yQUTW4HqBLLWw68gRICVFuVrIKXBhFxFZg+sFcpUqQP78xtvS0oRQJjKD1AiZe5GJSC7XC2StFmjUSLyd+5HJTBwhE5E1uF4gA9yPTFbF07qIyBpcM5ClFnYdPw7ExytXCzm8AO8A+Hv6G22LS41DbEqswhURkSNyzUAuXRooXtx4m14PHDyobD3k8DhtTUQ5pbN3AXah0QjT1suWGW/fuxdo107ZmqwoKT0JO67tgLfOG97u3vDSeYl+7a51h0ajsXfJDq9UQClEREcYbYuMi0StIrUUroiIHI1rBjIgTFtLBbIjMhgAjQb3E+/j3d/elfUSrUYLb91/Qe3u/cLXRoNcpK+cr7O/n1bjXJMzvIaRiHLKtQNZzPnzwIMHQMGCytWTUytXAjt3AosXIyVD/l5qvUGPpIwkJGUkAQpuwXbXupsf6jn4YcBb5w0PNw+bzQbwtC4iyinXDeQiRYCKFYFLl4y3798PdO+ubE2WOn0aGDYMSE0Fzp6FYf6X9q7IpAx9BjLSMvA07alin6mBBl46L3jpvHBi6AnJEDWX5DWMcRwhE5FpzjVvaC5nOEYzNhbo3FkIYwA4fx6V2vRHB5GfM1yZAQakZKYgNjX2+Q1NUfFRSM9Kz/F7c1EXEeWUawey1H5kRzggRK8H+vYFbrz4B74uIQkbfgGm7QbcsuxUm8p56byw6MQiVJ5fGdMOTcvx+5XwLwENjE+H34q7hSw9/0MQkTTXDuSQEGHFtTE3bgA3bypZjfmmTQO2bhVtnngYeIu3/xnV9beueG/be0hIT8DXh77GmegzOXo/T50niuYuarQtQ5+Buwl3c/T+ROT8XDuQAwKAmjXF29U8St61C5g0SbLLlnblkbtVezQv3RxvFn8TdYrUQZUCVVA6oDSK+hVFoHcgvHXeChWsLgduHXj+daY+EwM2DUBGVkaO3pPT1kSUE667qOuZJk2AkyeNt+3dCwwcqGw9cty6BfTsKWxzEvP222j3+x6005n+T2wwGJCelY6UzBSkZqYiJSPF8q8zU5CSIe/rDH3OAtCaIqIjMP3wdHz+1ucWv0epPKVw8JbxQ2UiYyMRUjLE4vcmIufHQG7cGJgxw3jbvn3P9/aqRmoq0KUL8PixeJ/ChYG1awEZYQwAGo0GnjpPeOo8rVSkPFn6rOcBrcQPAkkZSUjNTBWt5//+/D90qNABVQpUseifhyNkIsoJBvIbbwDu7kCGkdFadLSwLapSJeXrEvPhh8CJE+LtOh3w229AoULK1WQhN60bfD184evhq9hnfrrnU0w7bHwRV4Y+A/039sfRwUeh05r/vwYvmSCinHDtZ8gA4OsL1K8v3q6m7U/LlwOLF0v3+f57oGFDRcpxRF+GfIlK+cV/wDp5/yRmHplp0XtL3YvMETIRmcJABhxjP3JEBDB8uHSfHj2AUaOUqcdBeeo8saz9MsmjO7888CUuPrpo9ntzypqIcoKBDEjvRz5wAMiy8x7Slw//MKZyZWH0rKbn3SpVt2hdjK0/VrQ9PSsdAzcNNHvvcKFcheCl8zLa9iDpAZIzks16PyJyLQxkAKhbV5i6NiY+Hjh1Stl6stPrgd69gUiJZ5B+fsDvvwO5cilXl4Ob3GgyyuctL9p+7O4xzD4626z31Gg0ktPWPNOaiKQwkAHAwwN4803xdnvuR/7mG2D7duk+y5cD5copUo6z8NJ5YVn7ZaKnawHA5/s+x5WYK2a9Lxd2EZGlGMjPSE1b2+s58h9/AF+auChi/HigUydl6nEy9YvVx0f1PhJtT8tKw8DN5k1dc2EXEVmKgfyM1MKuv/4C0tKUqwUQju3s1Uv68I9GjYQRNFlsSuMpKBtYVrT9SNQRzP1nruz348IuIrIUA/mZ6tWBwEDjbSkpwNGjytXy7PCPJ0/E+xQpAqxZI/vwDzLOx90HS9svlZy6/nTvp7j25Jqs95O6hpFT1kQkhYH8jFYrjDjFKDltPWqU+HGegHCQybp1QMGCytXkxN4o/gY+eP0D0faUzBQM3DQQeoPe5HtxhExElmIgZ6eG/chLlwKhodJ9Zs2SPsyEzPZN428kw/TQ7UOY9888k+9j6hmyQeoRBBG5NAZydlILu/75B0hIsO3nnzoFjBgh3adnT2DkSNvW4YJ8PXwR9k6YZJ+JeyeaHOX6efohn08+o23JGcl4lPzI4hqJyLkxkLMrVw4oavxOW2RmAocO2e6znzwRDv+QWjxWpQoP/7ChkJIhGFlH/Ied5IxkDNo8yOTUNaeticgSDOTsNBr7TFs/O/zj5k3xPrlzC4d/iB1gQlbxbdNvUTJPSdH2AzcPYNGJRZLvIbkXmYeDEJEIBvLLpKatbXVAyJQpwI4d0n1WrADKim/PIevI5ZHL5NT1+D3jcTPupmg79yITkSUYyC+TGiGfPg3ExFj383bsACZPlu4zcSLQoYN1P5dENS7VGMNqDRNtT0xPxJAtQ0QXaHHKmogswUB+WbFi0iPRAwes91mRkaYP/2jcWBhBk6JmNJuBYrmLibbvubEHoaeMr4aXPM+ae5GJSAQD2RgljtFMSREO/4iNFe9TtCgP/7CT3J65EfqO9PazMbvG4Hb87Ve+zxEyEVmCgWyMEgu73n9f+hapZ4d/FChgnc8jszUv3RyDagwSbU9IT8DQLUNfmbou5l8Mbho3o6+JehqFjKwMq9ZJRM6BgWyM1IldV68CUVE5e//QUOEAECk//ADUq5ezz6Ec+77593gt92ui7Tuv78Sy08te+J5Oq0Nx/+JG++sNeqOjaiIiBrIx+fIJZ1uLyclq6xMnhNGxlN69geHDLf8Mshp/L38sbrtYss/HOz/G3ad3X/gep62JyFwMZDG2mLZ+/Fh4bix1+Ee1asCiRTz8Q0ValW2F/tX7i7bHp8Vj2NZhL0xdc2EXEZmLgSzG1H5kc88kzsoSVlTfuiXex98fWL8e8PEx773J5mY1n4XCuQqLtm+7ug2rzq56/muOkInIXAxkMW++Kb66+e5d4N9/zXu///s/YOdO6T4rVwJlypj3vqSIAO8ALGorfULXh398iPsJ9wGYOK2LI2QiMoKBLMbPD6hbV7zdnOfI27YJgSzl00+Bd96R/56kuHbl26F3td6i7XGpcXhv23swGAyS9yJzhExExjCQpVhjP/KNG8IiLSlNm5oObFKFOS3noKCv+D3Um69sxprzazhlTURmYyBLkQrk/fuFSyGkpKQINzjFxYn3KVYMCA8H3IzvWyV1CfQOxMK2CyX7jNoxCplZmfDz8DPa/iTlCeJT421RHhE5MAaylHr1AG9v421PngBnzoi/1mAQ7jY+fVq8j4eHcPhH/vw5KpOU1aFCB3Sv0l20/UnKE4zcMVJy2prPkYnoZQxkKZ6ewBtviLdLTVsvWQIsXy79/nPmSD+nJtWa22ou8vuI/yD1+6Xf4enmKdrOaWsiehkD2RRL9iMfPw6MGiX9vn37AsPEbxQidcvnkw/z28yX7HPu4TnRNt6LTEQvYyCbIvUc+dAhID39xe/FxAjPjV/+fnbBwcCCBTz8w8F1qdQFXSp1EW1PzUwVbeMImYhexkA2pWZN4cAOY5KSgH/++d+vs7KAnj2lz7rm4R9OZV7recjrndfs1/EZMhG9jIFsipsbEBIi3p592vqrr4Ddu6Xfb/VqoHRpa1RGKlDAtwB+av2T2a/jCJmIXsZAlsPUMZoAsGUL8PXX0u/z+edA27bWq4tUoVvlbuhYoaNZr4mMi4TeYGLbHBG5FI3h5ctc6VUXLwKVKxtvc3cHTp4UjtqMl9hb2rw5sH079xs7qejEaFSeXxlPUp7Ifs2dj+6gaO6iNqyKiBwJR8hyVKwIFCpkvC0jA+jQQTqMixcHfv6ZYezECuUqhDkt55j1Gk5bE1F2DGQ5NBrp7U83JP5gfXb4R7581q+LVKVX1V5oV66d7P5c2EVE2TGQ5ZIKZClz5wJ16li3FlIljUaDhW0XIo9XHln9OUImouwYyHJJLewS078/MGSI1Ush9SriVwQ/tPhBVl8GMhFlx0CWq2RJIEj8Bp9XVK8OzJ/Pwz9cUN/gvmhVppXJfmcfnMWDxAfgukoiArjK2jxDhgChoab75ckjrLw2J8DJqdx5egeV51fG07SnJvv6uPsgKCAIQQFBKJWn1POvn/3a213kghMicioMZHOsXQv06GG639atQJs2tq+HVC3sVBgGbxmc4/cpnKvwCyGd/a9CuQpBq+FEF5EzYCCbIzLS5Kg3tG1RvL18P8rmLatQUaRKyckwfDQaLZOXYFcZ232Ml87rlVF19tG1r4ev7T6ciKyKgWyOWbOAMWOe/zLVS4c73pkoEyv8+o/SQJtegJenD75v/j2G1RoGDZ8hu56zZ4Hu3YFLl3DbH6gwEkjxAJpdA5bscEPbrlk4X1CZUgr4FvhfSOd5MbCL+BWBm5Z744nUgoEsV1KSMDp++PCVplQ3IMEDqDAKeJLtzohWZVoh7J0wFPYrrGChZDcGg7CQb8wYIC3t+bdXBgt/73MG0AC4kB+oM0QIaXvycPNAyTwljYZ1qYBSyO2Z274FEj1jMACpqUBcHBAbK/w9+9cvfy9/fmDhQntWbBEGslwzZwLjxok29+oIhAe/+v283nmxsO1CyWv6yAk8fgwMGgRs2iSr+5S3gC8s3NqulHw++UQXm72W+zXotDp7l0iOJDNTONFQTqAaa5e60vZlQUHA9evW/2ewMQayHElJQKlSwKNHRpvPFARqDAMMEmtr+lTrg7mt5sLfS+QqR3Jcf/4J9OoF3L0rq/up5tXwU59yuJgahRuxN/Ao2fjvKzXTaXUo4V9CdLGZ3MNRyIEYDEBiouWBmpioXK0BAcAT+efKqwUDWY4ZM4AJE0SbO3YDNlY0/TbFchfDig4r0KhUIysWR3aTmQn83/8Jt3zJ+d8oVy5hGq1Xrxe+nZCWgMi4SNyIvfHKX5FxkUjPMmNkoBIBXgGiYV0sdzG4u7nbu0TXlJZmeaDGxwt3vjsCrVa4Z0DrWDsQGMimJCYKo+OYGKPN+uBq+OCbNzDvxHzZb/lRvY8wtclUeOm8rFUlKe3WLaBnT+DIEXn969QB1qwx+y5svUGPewn3EBmbLbDj/hfY0YnRFhRvX24aNxT3L45SAaVeeXYdFBCEQO9Ah18MaTAYoDfooTfokWXIQpY+6/nXeoMeWfos0a+NvUbs9YWjk1DxXrq8QI2LE57Duoq4OMDfsWYkGcimfPst8Mkn4u0bNwLt22PntZ0YsGkA7ifel/W2lfNXxqqOq1CjcA3r1EnKWbdOOCQmLk5e//HjgSlThItGrCwpPQk3426+OLLOFtipmY73B3Buz9zPw9lb540ifkWkQ8wKIWftwFTqrutl/1ZC//CLinyWw7l5EyhRwt5VmIWBLCUhQTgyU+xZRM2awIkTz4/HfJz8GCO2j8CvF36V9fbuWndMDpmM8Q3Hc/uJI0hOBj76CFi8WF7/ggWBlSuFu7DtwGAwIDoxWjSs7yXcs0td5ijiV8Qh6rSXH+9UxajQc/YuQ50iIoQjjB0Il0lKmTtXemHAV1+9cFZ1Xp+8WNt5LdqXb48R20YgPk3ijmQAGfoMfLrvU2y9uhUrO6xE6UDzpjNJQdn2FsvSsiWwYgVQoIBt65Kg0WhQ2K8wCvsVRsPiDV9pT8lIeWV0nf1ZdlJGkh2qfpGfh5+9S1C1BG8X+UFeqxWOJA4IEP4u9nX275V1vMOZOEIW8/SpMDqOjTXeXrs28M8/opdHRMVHYcCmAdgbuVfWx/m6+2J2i9kYXHOwwz8/cyoie4tFubsLjzlGj3a4BSXZGQwGPEp+ZHSh2Y3YG7jz9A4MsP0fHXWK1MHxe8dt/jmOakJKLXw7/aS9y5AnVy7TISrWniuXS1zUwxGymB9/FA9j4JXR8cuK+RfDrj67MPfYXEzcO9Hks7ykjCQM3ToUm//djNB2oSiYS6GjnEjc48fAwIHA5s3y+pcpI5x3XquWbetSgEajQQHfAijgWwD1Xqv3SntaZhpuxd8yGtbXY68jMd06W1x4OIk0RUfIHh6WB6q/P6Bj3JjCEbIx8fHC6Fhs0U7dusDRo7J/Yrv46CL6bOiDU/dPyeqfzycfFrddjI4VO8qrl6zvwAGgd2/Ze4vRr5/wiMOPU6wGgwGPUx6Ljq6jnkbJXvTUqUIn/H75dxtX7Lh6eb+O1ROOyeus0fwvLM0N1Dx5AC8vlxil2hN/ZDFmzhzpFbQmRscvq5S/Ev4e9Dem/DkFU/+aavIPo5jkGHT6tRP6V++POS3ncJSgpMxMYPJk4JtvcrS32JVpNBrk88mHfD75ULdo3Vfa07PSERUfZXSx2fUn119Ye5HHO4+ClVtOq9HCTeMm/F3r9sLXz9rEvpb7GmP9KviUA/qUkxeofn4O/RjFFXCE/LK4OGF0HC+yIOv114G//7b4J8W/o/5G3419ce3JNVn9S/iXwMqOK/FWibcs+jwyg0J7i0labEosbsTewO342wj0DsSlmEs2CThrhSKvvyRrYSC/7KuvhBGSmD/+AFq0yNFHJKYnYuyusVh0cpGs/hpoMLbBWExpNAWeOs8cfTaJ+O03YW+x2A9iL7Ph3mIick0M5OxiY4XR8dOnxtvr1wcOH7bac5Rt/27DoM2D8CDpgaz+VQtUxepOq1GtYDWrfD5B2Fs8ejSwZIm8/nbeW0xEzotzLdnNni0exoAwcrbiooY25drg/Ijz6Fyxs6z+5x6eQ+3FtTHj8Axk6R3kTFk1O3tW2L4mN4xbthRewzAmIhvgCPmZJ0+E0XFCgvH2Bg2Av/6yySpDg8GA1WdX4/0d7+NpmsQPBNm8WfxNrOiwAqUCSlm9HqfnonuLiUjd+KfLM7NmiYcxYPXRcXYajQZ9gvvg7HtnEVIyRNZrDt0+hGoLq2FZxDLwZyozPH4MdOgAvP++vDAuU0ZYxPfxxwxjIrIpjpAB4Q/pkiXF7+t84w3g4EFF9uDpDXr8cPQHfLL3E9nX7rUv3x6L2y1GAV/7HdPoELi3mIhUjD/yA8D330tfnm3D0fHLtBotPq7/MU4OPYnggsGyXrPpyiZUXVAVW65ssXF1DiozE5g0CWjcWF4Y+/kBq1cDy5czjIlIMRwhx8QI9x2LBfJbbwkjKzucUJOWmYavDnyFGUdmyD7ZaHCNwZjVYhb8PBkkALi3mIgcBkfIKhodv8xT54lpTafhYP+DCAoIkvWa0IhQBC8MxuHbh21cnQP47TcgOFh+GE+YICzcYxgTkR249gj50SNhdJwkcs1cSAiwf7+iJYlJSEvAxzs/RmhEqKz+Wo0W4xuMx+RGk+Hh5mKHV1iyt3jVKqBZM5uWRUQkxbUDecIEYMYM8fY//xSmrFVk85XNGLJlCB4mPZTVv3qh6ljVcRWqFKhi48pUwgHvLSYiAlw5kB8+FEbHycnG2xs3BvbKu8tYaQ+THmLolqHYdGWTrP4ebh6Y1mQaRtcb7bzn7lqyt3j6dODDD7mdiYhUwXUDedw4YOZM8fZDh4TtTiplMBiw/PRyfPjHh0hIl9g/nU1IyRAsb78cJfKUsHF1CjP33uKyZYV7i2vWtG1dRERmcM1AfvBAGB2npBhvb9oU2L1b2ZosFBkbiX4b++HQ7UOy+uf2zI25reaiT7U+0DjD3aaW7C3+6Sfh2kQiIhVxzbm6GTPEwxiQvu1JZUoFlML+fvsxvel0uGvdTfZ/mvYU/Tb2w7u/vYuY5BgFKrSRnOwtZhgTkQq53gg5OhoIChIP5ObNgZ07la3JSs5En0GfDX1w7uE5Wf0L5SqEsHfC0LpsaxtXZmU3bwK9enFvMRE5FdcbIU+f7jSj45cFFwrG8SHHMa7BOGhgejo6OjEabcLb4L2t7yExXWIvtpr89htQvTr3FhOR03GtEfL9+8LoODXVeHvLlsCOHcrWZCMHbx1E3w19cSv+lqz+ZQLLYFXHVaj3Wj0bV2ahpCRhb3GovH3Y3FtMRI7GtUbI334rHsaAQ4+OX/ZWibdwdvhZ9K/eX1b/a0+uoeHShpi0bxIysjJsW5y5zpwR7i2WG8bP7i1mGBORA3GdEfLdu8K0pdge1datgW3blK1JIRsubcDQrUNlL+KqWbgmVnVchUr5K9m4MhMMBmDePGDsWO4tJiKn5zqBPGqUsN1FzD//CIt/nFR0YjSGbBmCrf9uldXf080T05tOx6jXR9nnMJGYGGFv8RaZN1hxbzEROTjXCOQ7d4TRcbrI/cJt2gBb5QWVIzMYDAiLCMPoP0YjKUPk/O6XNCnVBMvaL0Mx/2I2ri6b/fuFvcX37snrz73FROQEXGNeb9o08TAGgK++UqwUe9JoNBhcczDOvHcGDYo1kPWavZF7UXVBVYSfC4fNf3Z7tre4SRN5Ycy9xUTkRJx/hBwVBZQpIx7I7drJP3LRiWTpszDj8Ax8ceALZOozZb2ma+WuWNBmAQK9A61f0M2bwr3Ff/8trz/3FhORk3H+EfLUqRwdG+GmdcMnb36Cfwb/I3vx1q8XfkXVBVWx85qVD055trdYbhhzbzEROSHnHiHfuiUs9skQ2cbTvj2wcaOiJalRamYqPtv7GWYdnSX7NSPrjMSMZjPg4+5j+Qebu7e4UCFg5UpuZyIip+TcgTxsGLB4sXh7RIQwMiMAwP7I/ei3sR+inkbJ6l8ubzms6rgKdYvWNf/DzpwR7i2+fFle/1athGfFvLeYiJyU805Z37wJLF0q3t6xI8P4JY1KNcLZ4WfRp1ofWf3/ffwvGoQ1wFcHvpJ/mIjBIKyIfv11eWHs7g7Mni2sgmcYE5ETc94R8pAh0lOhZ84A1aopV4+DWXdxHYZtHYYnKU9k9a9TpA5WdVyF8vnKi3fi3mIiIlHOGciRkUC5csI2GmM6dwbWrVO2Jgd0P+E+Bm0ehB3X5J3v7a3zxoxmMzCyzshX71o2d29x//7A3LnczkRELsM5A3nQIOnp6rNngapVlavHgRkMBiw6uQhjdo1BckayrNc0L90cS99ZiqK5iwo/FE2eDHzzjTBdbYqfH7BwobAFiojIhThfIN+4IYyOs7KMt7/7LvDrr8rW5ASuPr6KPhv64NjdY7L6B3gFYEHd/0O3z8Llb2eqW1fYWxwUlINKiYgck/MF8sCBwLJlxts0GuDcOaByZWVrchKZ+kxMOzQN/3fw/2QfJtLjHDBvGxAgcckWAGFv8ZQpwiIuIiIX5FyBfO0aUKGC+Oi4WzdhkRDlyIl7J9BnQx9cjpG3ZWnqHuCTv0QaubeYiAiAs217+vpr8TDWaIAvvlC2HidVu0htnBp6Ch/U/cBk3xr3gTFiM9atWgmr3RnGRERONEK+elUYHev1xtt79ADCw5WtyQXsubEH/Tf2x92Eu6+0eei1iFigR6VHLzW4uwMzZgAffMB7i4mI/uM8fxpOmSIexlotR8c20jSoKc4NP4ceVXq80qZ302JDvTxIc8v2zbJlgaNHhSMzGcZERM85xwj5yhWgUiXxQO7VS7imj2zm2J1jqBdWz2hbucfAks3AW436c28xEZEInb0LsApTo+NJk5Stx8Vk6jMxfNtw0fZ/8wIxn44Ghs5WrigiIgfj+IF8+bKwd1VMz55AeYnjHCnHFhxfgIjoCNH21mVbo2MP+TdJERG5Isefsu7ZUzyQtVrg0iXhoBCyifsJ91FhXgU8TXtqtN1L54ULIy4gKICHfRARSXHsVTUXL0rvK+7dm2FsY2N2jRENYwD49I1PGcZERDI49gi5e3fgl1+Mt7m5CdPZZcooW5ML2XtjL5quairaXjawLM4NPwdPnaeCVREROSbHHSFfuCB9JnWfPgxjG0rLTMPI7SMl+8xrPY9hTEQkk+MG8uTJ4rcHubkBn3+ubD0uZuaRmbjy+Ipoe7fK3dCsNE/gIiKSyzGnrM+dA6pVE28fOBAIC1OuHhcTGRuJSvMrITXT+I0Rfh5+uPz+ZRTxK6JwZUREjssxR8iTJ4u36XQcHduQwWDAqB2jRMMYAKY0msIwJiIyk+MF8pkzwPr14u39+wOlSilWjqvZdGUTtl3dJtoeXDAYI+tKP1smIqJXOd6UdadOwIYNxtt0OuGSiZIlFS3JVSSlJ6HivIqIehol2ufIwCOoX6y+glURETkHxxohnz4tHsaA8OyYYWwzUw5OkQzjITWHMIyJiCzkWCPkDh2ATZue//JqILC6GvDFn4Cbzh24dg0oXtx+9TmxCw8voPqi6sjUZxptz+udF1fev4K8PnkVroyIyDk4zlnWmZmAtzeg0QAGAxI8gA7dgYsFgGOvAeF+vRDIMLYJg8GAEdtHiIYxAMxoNoNhTESUA44zZa3TCWdWnzsH/btd0K+jEMYAsLMMUKfMPpx7cM6+NTqp1WdX4+Ctg6LtDYs1RP/q/ZUriIjICTlOID9TuTKmvh+MDRVf/PaNhNuoF1YPv16QOL2LzBabEosxu8aItrtp3DC/zXxoNY73W4mISE0c7k/Rbf9uwxf7vzDalpyRjG7rumHinonI0mcpXJlz+mzfZ3iU/Ei0/cPXP0S1ghKHtBARkSwOtagrJSMFQT8GITox2mTfFqVbILxzOAK9AxWozDkdv3scr4e+DgOM/xYp6lcUl0Zegp+nn8KVERE5H4caIXu7e2Nrj60olruYyb47r+9EnSV1cPbBWQUqcz5Z+iwM3zZcNIwB4IeWPzCMiYisxKECGQBqFamFk0NPIqRkiMm+N2JvoH5YfT5XtsDCEwtx8v5J0fYWpVugc8XOClZEROTcHGrKOruMrAyM2z0Oc47NkdV/fIPxmNpkKty0bjauzPFFJ0ajwk8VEJ8Wb7Td080T50ecR5lAXm9JRGQtDjdCfsbdzR0/tPwBKzushJfOy2T/GUdmoHV4azxJeaJAdY5t3O5xomEMAJ+88QnDmIjIyhx2hJzdyXsn0fGXjpLHOj4TFBCEDd02cGWwiP2R+9F4ZWPR9tIBpXF+xHlZPwQREZF8DjtCzo7Pla0jPSsdI7aPkOwzr/U8hjERkQ04RSADQH7f/NjVexc+fP1Dk32f7VeesHsC9ytnM+vvWbgcc1m0/d1K76JFmRYKVkRE5DqcYsr6ZavOrMLQrUORmplqsm/z0s2xpvMal9+vfDPuJirNq4SUzBSj7bk8cuHyyMsomruowpUREbkGpxkhZ9cnuA8ODzyM4v6mL5vYdX0X9ysD+PCPD0XDGAAmh0xmGBMR2ZBTjpCfeZT0CF3XdcWBmwdM9vVx98HSd5aiW5Vuti9MZTZf2Yz2a9uLtlcrWA0nh56ETus4l4MRETkapxwhP5PfNz9299mN0a+PNtk3OSMZ3dd3x/jd413quXJSehI+2PGBZJ/5reczjImIbMypAxkAdFodZrecLXu/8ndHvkOrn1u5zH7lbw59g1vxt0TbB1YfiIbFGypYERGRa3LqKeuXnbp/Ch1/6Yjb8bdN9i2VpxQ2dt/o1PuVLz26hOCFwcjQZxhtD/QOxJX3ryCfTz6FKyMicj1OP0LOrmbhmjgx5ISs/cqRcZGoH1Yfv5z/xfaF2YHBYMDI7SNFwxgApjedzjAmIlKISwUywOfKz4SfC8f+m/tF2+u9Vg8DawxUsCIiItfmUlPWLzNnv3KzoGZY22WtU+xXjkuNQ4WfKuBB0gOj7VqNFieHnkT1QtWVLYyIyIW53Ag5O3P2K+++sRu1F9d2iv3Kk/ZNEg1jAPig7gcMYyIihbn0CPmZR0mP0G1dN8kp3Gd83H0Q9k4YulfprkBl1nfy3knUDa0LvUFvtL1wrsK4/P5l5PbMrXBlRESuzaVHyM/k982PXX124aN6H5nsm5yRjB7re2D87vHI1GcqUJ31ZOmzMHzbcNEwBoDZLWYzjImI7ICB/B+dVodZLWZhVcdVZu1Xfpz8WIHqrGPJqSU4fu+4aHuzoGboWrmrghUREdEznLI2wtz9yhu6bUBwoWAFKrPcg8QHqDCvAuJS44y2e7h54NzwcyiXt5yyhREREQCOkI16tl+5UclGJvs+26+89vxaBSqz3Pg940XDGAAmNJzAMCYisiOOkCVk6jMxfvd4zD46W1b/cQ3GYWqTqao79/nPm38iZEWIaHtQQBDODz8Pb3dvxWoiIqIXMZBlWH12NYZsGSJrv3LToKZY23kt8vrkVaAy0zKyMlB9UXVcfHRRtM/2ntvRqmwrBasiIqKXccpaht7Vesver7znxh7UWVIHZ6LPKFCZabOPzpYM404VOzGMiYhUgCNkM5izX9lb542l7Zfadb/y7fjbqDivIpIzko22+7r74tLISyjmX0zhyoiI6GUcIZvBnP3KKZkp6LG+B8btGme3/cqj/xgtGsYA8FXIVwxjIiKV4AjZQj+f/RmDtwxW7XPlbf9uQ9s1bUXbK+evjIhhEXB3c1esJiIiEsdAzoGI+xHo+EtH3Iq/ZbJvyTwlsbHbRkX2KydnJKPy/Mq4GXdTtM/B/gfxZok3bV4LERHJwynrHKhRuAZODD2BxqUam+x7M+6mYvuVpx2aJhnG/av3ZxgTEakMR8hWYO5+5bH1x2Ja02k22a98JeYKqi6oigx9htH2AK8AXHn/CvL75rf6ZxMRkeU4QraCZ+dgr+64WtY52DP/nmmTc7ANBgNGbh8pGsYAMK3JNIYxEZEKcYRsZfZ8rrz2/Fr0WN9DtL1u0br4e9Df0Gr4cxgRkdowkG0gJjkG3dZ1w77IfSb7Wmu/cnxqPCrMq4DoxGij7VqNFseHHEfNwjVz9DlERGQbHCrZQD6ffNjZe6ei+5W/2P+FaBgDwMg6IxnGREQqxhGyjZmzX7lJqSb4pcsvZu9XjrgfgdpLakNv0BttL5SrEC6PvAx/L3+z3peIiJTDEbKN9arWC0cGHkEJ/xIm++6N3IvaS2rjdPRp2e+vN+gxfNtw0TAGgFnNZzGMiYhUjoGsAHP3KzcIa4A159bIeu/QU6E4dveYaHvjUo3tep42ERHJwylrBWXqMzFh9wTMOjpLVv8x9cfg26bfiu5XfpT0COV/Ko/Y1Fij7e5ad5wdfhYV8lWwuGYiIlIGR8gK0ml1+L7F9/i508/w1nmb7P/939+j5eqWovuVJ+yZIBrGADC+4XiGMRGRg+AI2U7M3a+8odsGVC9U/fn3/rr9F95cJn78Zck8JXFhxAX4uPtYo1wiIrIxjpDtJCfPlTOyMjB823DJ18xtNZdhTETkQDhCtjNLnisX8CmACXsniPbpUKEDNnTbYK0SiYhIAQxklQg/F47BmwcjJTPFZF+tRiu6zcnH3QcXR1xEiTymt1kREZF6cMpaJXpW7YnDAw/L2q8stef4i7e+YBgTETkgjpBVxpxzsF9WKX8lRAyLgIebhw0qIyIiW+IIWWWenYP9cb2P5b/ovx+p5reezzAmInJQHCGrmNRzZfdMoN2/wKBTwLVA4MTEPljZcaUdqiQiImtgIKtcxP0IdFwYglvapwCASg+BsUeAXpfd4ZGaAQCI9wJSb0eiYP6SdqyUiIhywviZjKQaNfZewImZT9GrEzDpT+CNqGctGc/7+KcC/rsOA71K2qNEIiKyAo6Q1ezUKaBhQyA1FZla4GogUDFGpG/jxsDevYqWR0RE1sNAVquHD4HatYGoKNN9n7l+HQgKsl1NRERkM1xlrUYZGUDXruaFMQAsW2abeoiIyOYYyGo0dizw55/mv275ciAry+rlEBGR7TGQ1WbFCuDHHy177Z07wK5d1q2HiIgUwUBWk+PHgWHDpPs0bw7UqyfeHhZm3ZqIiEgRXNSlFg8eCIu47twR7xMUJIT2hg3A4MHG+7i7A3fvAvnz26ZOIiKyCY6Q1SA9HXj3Xekw9vUFNm4EAgOFBV++vsb7ZWQAq1bZpEwiIrIdBrIafPwxcOiQdJ9ly4CqVYWv/fyAbt3E+4aFAZz4ICJyKAxke1u6FJg3T7rPJ58II+jsBg4U73/xInDsWM5rIyIixTCQ7enYMWD4cOk+rVoBU6a8+v0GDYDy5cVfx8VdREQOhYFsL9HRQKdOwvNjMWXKAD//DLi5vdqm0QCDBom/du1aIDEx53USEZEiGMj2kJ4OdOkC3Lsn3idXLmERV0CAeJ++fQGdyP0giYnAb7/lqEwiIlIOA9kePvwQOHxYus+KFUDlytJ9ChYE2rYVb+e0NRGRw2AgK23JEmDhQuk+n38uTGfLITVtffgwcPmy/NqIiMhuGMhK+vtvYORI6T5t2gCTJ8t/z5YtgcKFxduXLpX/XkREZDcMZKXcuwd07iwc3CGmXDlg9WpAa8Z/Fp0O6N9fvH3lSunPJCIiVWAgKyEtTQjj+/fF+/j5CYu48uQx//2l9iQ/eABs327+exIRkaIYyEoYNQo4elS6z6pVQMWKlr1/mTLA22+Lt3NxFxGR6jGQbW3RImEhl5QvvwTat8/Z50gt7tq+XXp0TkREdsdAtqXDh4XRsZR33gG++CLnn9W5M5A7t/G2rCxhGxUREakWA9lW7t41vYirQgVhqtqcRVxifHyAHj3E25cu5YUTREQqxkC2hdRUYR/xgwfifXLnFhZxiY1qLSE1bX31qukbpYiIyG4YyNZmMAh7jf/5R7yPRiOcUS11OYQlatf+3xWNxnBxFxGRajGQrW3BAtOHcUyeLH3kpaVMXTjx229AfLz1P5eIiHJMYzDwwaLVHDwINGkCZGaK9+nQAVi/3jrPjY15/BgoUkT8FqmFC4Fhw2zz2UREZDGOkK0lKgp4913pMK5USTg5y1ZhDAB58wqhL4bT1kREqsRAtoaUFGER18OH4n38/YVFXH5+tq9Hatr6+HHg3Dnb10BERGZhIOeUwQAMHw6cOCHeR6MBwsOBsmWVqalpU6B4cfF2jpKJiFSHgZxTP/1k+tCNr78GWrdWph5AmBIfMEC8fdUq4XxtIiJSDS7qyokDB4TRaFaWeJ8uXYBffxVGyUq6dQsoVUr8MJBffgG6dlW2JiIiEsURsqVu3xYWcUmFcZUqwLJlyocxAJQoIfywIIb3JBMRqQoD2RIpKUDHjkBMjHifPHmERVy5cilV1aukFnft2iX8UEFERKrAQDaXwQAMHQqcOiXeR6sF1q4FSpdWri5j2rcHAgKMtxkMwPLlipZDRETiGMjmmjMHWL1aus/UqUCLFsrUI8XLC+jdW7x92TJAr1euHiIiEsVFXebYtw9o3lz6uXHXrsLo2B7PjY05cwaoXl28ffdu6WfNRESkCI6Q5bp5UwhbqTCuVk1YLKWWMAaA4GCgVi3xdu5JJiJSBQayHMnJwiKux4/F+wQEABs2AL6+ytUll9Tirg0bgCdPlKuFiIiMYiCbYjAAgwcDp0+L99FqhX29QUGKlWWWHj2E58nGpKUJV0ESEZFdMZBNmTULWLNGus/06UCzZsrUY4k8eYQDSsSEhYkfIEJERIrgoi4pu3cDLVtKr0Tu0UMYYarpubExBw4AjRqJt584If2smYiIbIojZDE3bgDdu0uHcfXqQGio+sMYAN5+W3pfNBd3ERHZFQPZmKQk4U5hqcVOefMKC6J8fBQrK0c0GmDgQPH28HDhBDIiIrILBvLLDAYhuKTuDHZzExZxlSypWFlW0a+fsADNmPh4YP16ZeshIqLnGMgv++474XYmU32aNFGmHmsqWhRo1Uq8ndPWRER2w0Vd2e3cKQSW1L+S3r2BlSsd47mxMb//DnTuLN5+7Zr9z+AmInJBHCE/c+2asIhLKoxr1gQWL3bcMAaAtm2B/PnF23ktIxGRXTCQASAxUVjEFRcn3id/fmERl7e3UlXZhocH0LevePvy5dLHgxIRkU0wkA0GYMAA4MIF8T5ubsJz5eLFlavLlqSO0rx3T5i6JyIiRTGQv/0WWLdOus+sWUBIiCLlKKJiRaB+ffF2Lu4iIlKcay/q2r5deKYq9a+gXz/h3mBHfm5sTFiYcEa3MTodcPcuUKCAsjUREbkw1x0hX70K9OwpHca1awMLFzpfGAPCVZJiN1NlZgKrVilbDxGRi3PNQE5IEBZxxceL9ylQQNgiJHZLkqPz8wO6dRNv54UTRESKcr1A1uuFaeiLF8X76HTCc+VixZSryx6kFnddugQcPapcLURELs71AnnqVGH7kpQffgDefFORcuyqfn2gQgXxdi7uIiJSjGsF8tatwBdfSPcZOBAYMUKZeuxNo5EeJf/yi7BHm4iIbM51VlmnpgJBQcD9++J96tYF/vzTeZ8bG/PgAfDaa8JCLmPCwqRviSIiIqtwnRGyl5ewzalEieffSvAANj6bsS1Y0LkXcYkpWFDY+iWG09ZERIpwnRHyfwyPHuGfIa0Qqj+JtVWARE/g4kIdKq47ADRsaO/y7GPrVqBdO/H2ixeFw0SIiMhmXGaE/CTlCeYem4vgdU1Qr8ZJhNYSwhgAwj4Ocd0wBoCWLYHChcXbeeEEEZHNuUwgf3/ke3zwxwc49/DcK20rcBppmWl2qEoldDqgf3/x9pUrgYwMxcohInJFLhPIA2uIL0yKSY7BpiubFKxGhaQWbj18KExrExGRzbhMIJcOLI0mpZqItoeeClWwGhUqUwZ4+23xdk5bExHZlMsEMgAMqTlEtG33jd2IjI1UsBoVktqTvH27cDUjERHZhEsFcocKHZDXO69oe1iEi2/x6dwZyJ3beJteD6xYoWw9REQuxKUC2VPnib7BfUXbl51ehky9yAEZrsDHR7gBS8zSpbxwgojIRlwqkAFgcE2RO4AB3Eu4hx1XdyhYjQpJTVtfuwYcPKhcLURELsTlArlS/kpoUKyBaPuSU0sUrEaFatUCqlUTb+fJXURENuFygQxIL+7adnUb7j69q2A1KmPqwol166TvkSYiIou4ZCC/W+ld5PY0vnhJb9Bj+enlyhakNr16AR4exttSUoA1a5Sth4jIBbhkIPt6+KJnFfHFS2ERYdAb9ApWpDJ58wI1arz6fQ8PoGtXIDhY+ZqIiJycSwYyAAypJT5tHRkXiX2R+xSsRmX27wdOnnzxey1bAnfuCHck169vn7qIiJyYywZyzcI1UaOQkVHgf1x2cdf580DHjq/ej/zHH8Ann/BMayIiG3HZQAakF3dtuLQBMckxClajAnfvAq1aiS/aCgsTDg/hXmQiIqtz6UDuWbUnvHXeRtsy9BlYeWalwhXZ0dOnQOvWwrS0lHfeEVZiExGRVbl0IPt7+aNr5a6i7UtOLYHBFUaD6enCyPfsWel+kyYBg8UPViEiIsu5dCAD0tPWl2Mu40jUEQWrsQODARgyBNizR7pfv37A5MnK1ERE5IJcPpAbFGuAivkqirY7/eKuL74AVpqYmm/WDFiyhFPVREQ25PKBrNFoJM+3/vXCr4hLjVOuICUtWQJ8/bV0n+Bg4XQud3dlaiIiclEuH8gA0De4L9y1xgMnJTMFa8454clU27cDw4dL9ylWTOgndiUjERFZDQMZQD6ffOhYsaNou9NNW584Abz7LpCVJd7H3x/YsQMoUkS5uoiIXBgD+T9Si7sioiNw8t5J0XaHEhkJtGkDJCeL9/HwADZuBCpXVqwsIiJXx0D+T+NSjVEqTynR9tBToQpWYyOPHwsHfzx8KN1v+XIgJESJioiI6D8M5P9oNVoMqiF+7eDP535GUnqSghVZWUoK0L49cOWKdL/p04EePZSpiYiInmMgZzOgxgC4adyMtiWkJ+DXC78qXJGV6PVAnz7A4cPS/UaOBMaNU6YmIiJ6AQM5myJ+RdCmXBvR9tAIB522HjMGWL9euk/79sCcOdxrTERkJwzklwyuIb4n+UjUEVx4eEHBaqxg9mzghx+k+7z+OhAeDrgZnx0gIiLbYyC/pFXZVijiJ77VJywiTMFqcmjdOmF0LKVMGWDLFsDHR5maiIjIKAbyS3RaHQZUHyDavvLMSqRlpilYkYX++gvo3Vv6qsR8+YS9xvnzK1cXEREZxUA2Qmq19eOUx9hweYOC1Vjg8mXhmsQ0iR8cvL2BrVuFETIREdkdA9mIUgGl0CyomWi7qvckR0cLe41jY8X7aLXA2rXCs2MiIlIFBrIIqQsn9kbuxfUn1xWsRqbEROEUrps3pfvNnSuMoImISDUYyCLal2+PfD75RNtVt7grMxPo2hU4dUq634QJwIgRytRERESyMZBFeOo80S+4n2j7stPLkKnPVLAiCQaDcHPTjh3S/Xr0AKZOVaYmIiIyCwNZgtTirujEaGz7d5uC1Uj45hsg1MRz7ZAQYNky4fkxERGpDv90llAxf0W8UfwN0XZVXMu4YgUwaZJ0n8qVgQ0bAE9PZWoiIiKzMZBNkLqWcce1Hbjz9I6C1bxk925gsPjiMwDCfcbbtwN58ihSEhERWYaBbEKXSl3g7+lvtE1v0GNZxDKFK/rPmTNA587CYi4xfn5CGBcvrlxdRERkEQayCT7uPuhVtZdoe1hEGPQGvYIVAbh9G2jdGkhIEO+j0wkXSgQHK1cXERFZjIEsg9Se5Fvxt7Dnxh7liomLE8L43j3pfqGhQDPxw02IiEhdGMgy1ChcA7UK1xJtV2xxV1oa0LEjcMHEjVNTpgD9xLdsERGR+jCQZZJa3LXp8iY8THpo2wL0emDAAODAAel+gwcDn31m21qIiMjqGMgy9ajaAz7uxq8ozNBnYOWZlbYt4NNPgTVrpPu0bg0sWABoNLathYiIrI6BLFNuz9zoVrmbaHvoqVAYpK46zIl584Dp06X71KoF/PKLsJiLiIgcDgPZDFLT1lceX8Fft/+y/odu2gR88IF0n5IlhasUc+Wy/ucTEZEiGMhmqPdaPVTKX0m03eqLu44dE86f1ktsqwoIEM6wLlTIup9NRESKYiCbQaPRSI6Sf7v4G2JTJO4hNse1a0DbtkBKingfT09g82agQgXrfCYREdkNA9lMfar1gYebh9G21MxUhJ8Lz/mHPHoEtGwJxMSI99FogNWrgTfEz9omIiLHwUA2U16fvOhUsZNo+5JTS3K2uCs5GWjXDrh+XbrfrFlAly6Wfw4REakKA9kCUtPWZx6cwYl7Jyx746ws4ZnxsWPS/UaPFv4iIiKnwUC2QEjJEAQFBIm2h54ycTexMQaDsJp682bpfl26AN9/b/77ExGRqjGQLaDVaDG4hvj51uHnw5GYnmjem373HTB/vnSfhg2BVasALf+zERE5G/7JbqH+1fvDTeNmtC0xPRG/nP9F/puFhwMTJkj3KV9e2JPs5WVGlURE5CgYyBYq7FcYbcu1FW0PjZA5bb1/P9C/v3SfggWFvcZ588ovkIiIHAoDOQekFncdvXMU5x+el36D8+eF25syMsT7+PoKp3CVKmVhlURE5AgYyDnQskxLFPUrKtq+5KTEyV137wqXQcTHi/dxcwN+/RWoXTsHVRIRkSNgIOeAm9YNA2sMFG1fdXYVUjNTX214+lQI46go6Q9YuFDoR0RETo+BnEMDawyEBsavO4xNjcXvl35/8Zvp6UDnzsDZs9JvPGmScLcxERG5BAZyDpXMUxLNSjcTbX9hT7LBAAwZAuzZI/2m/foBkydbqUIiInIEDGQrkFrctf/mflx7ck34xZdfAitXSr9Zs2bA4sXCWdVEROQyNIYcHbxMAJCelY7XZr2GR8mPjLZPaDgB30aWBoYOlX6j4GDg4EEgd24bVElERGrGEbIVeLh5oF9wP9H2qDWLYBg+XPpNihUDtm1jGBMRuSgGspUMrml8AVbNe8CiVXHQZGWJv9jfXzj4o6j4FioiInJuDGQrKZ+vPN4q8dYL3ysZC2z7Gcglce4HPDyAjRuBypVtWh8REakbA9mKsl84EZgM7FgNFEoy8aLly4GQEFuWRUREDoCBbEVdKnVBHq88AIAfdwAVHpt4wfTpwv3HRETk8hjIVuTt7o3eVXsDAMY0B04Ulug8YgQwbpwyhRERkeoxkK3s2eKuB35ASH9gZ3kjVzS2bw/8+CP3GhMR0XMMZCsLLhSMOkXqoN5r9TCnSyganooRTud65vXXhfuP3YzfpUxERK6JB4PYQEJaAvw8/f73DYMBmDpVWMB15AiQP7/daiMiInViICspKUm435iIiOglDGQiIiIV4DNkIiIiFWAgExERqQADmYiISAUYyERERCrAQCYiIlIBBjIREZEKMJCJiIhUgIFMRESkAgxkIiIiFWAgExERqQADmYiISAUYyERERCrAQCYiIlIBBjIREZEKMJCJiIhUgIFMRESkAgxkIiIiFWAgExERqQADmYiISAUYyERERCrAQCYiIlIBBjIREZEKMJCJiIhUgIFMRESkAgxkIiIiFWAgExERqQADmYiISAUYyERERCrAQCYiIlIBBjIREZEKMJCJiIhUgIFMRESkAgxkIiIiFWAgExERqQADmYiISAUYyERERCrAQCYiIlIBBjIREZEKMJCJiIhUgIFMRESkAgxkIiIiFWAgExERqQADmYiISAUYyERERCrAQCYiIlIBBjIREZEK/D/YnR25+OP8owAAAABJRU5ErkJggg=="
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<networkx.classes.graph.Graph at 0x359ea27d0>"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "execution_count": 20
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5f2ee2c4-fcdf-4a7e-be20-6a329b4909a4",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.10.12"
  },
  "widgets": {
   "application/vnd.jupyter.widget-state+json": {
    "state": {
     "07b43483463e4ae3b9a7b160d8938299": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_name": "DescriptionStyleModel",
       "_view_name": "ErrorWidgetView",
       "description_width": "",
       "error": {},
       "msg": "Failed to load model class 'DescriptionStyleModel' from module '@jupyter-widgets/controls'"
      }
     },
     "0b87e5fec1ff41aeb4c9e624a222f89b": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_name": "LayoutModel",
       "_view_name": "ErrorWidgetView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "error": {},
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "msg": "Failed to load model class 'LayoutModel' from module '@jupyter-widgets/base'",
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "0cf8ded72d3147d681343573a3c5c3ad": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_name": "LayoutModel",
       "_view_name": "ErrorWidgetView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "error": {},
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "msg": "Failed to load model class 'LayoutModel' from module '@jupyter-widgets/base'",
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "0ef53961cc214c23a5cc520142ed3530": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_name": "DescriptionStyleModel",
       "_view_name": "ErrorWidgetView",
       "description_width": "",
       "error": {},
       "msg": "Failed to load model class 'DescriptionStyleModel' from module '@jupyter-widgets/controls'"
      }
     },
     "10730754cf7a43eca9820d10763e9a41": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "HTMLModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_name": "HTMLModel",
       "_view_name": "ErrorWidgetView",
       "description": "",
       "description_tooltip": null,
       "error": {},
       "layout": "IPY_MODEL_1958cff3714b451e95a8fbc5e76541b6",
       "msg": "Failed to load model class 'HTMLModel' from module '@jupyter-widgets/controls'",
       "placeholder": "​",
       "style": "IPY_MODEL_0ef53961cc214c23a5cc520142ed3530",
       "value": " 11531/11531 [00:33&lt;00:00, 255.26it/s]"
      }
     },
     "14391e150296461ca407c9ff78c5ae7e": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_name": "DescriptionStyleModel",
       "_view_name": "ErrorWidgetView",
       "description_width": "",
       "error": {},
       "msg": "Failed to load model class 'DescriptionStyleModel' from module '@jupyter-widgets/controls'"
      }
     },
     "1958cff3714b451e95a8fbc5e76541b6": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_name": "LayoutModel",
       "_view_name": "ErrorWidgetView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "error": {},
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "msg": "Failed to load model class 'LayoutModel' from module '@jupyter-widgets/base'",
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "218cfca13d80402ab877a02fc12211cb": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_name": "LayoutModel",
       "_view_name": "ErrorWidgetView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "error": {},
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "msg": "Failed to load model class 'LayoutModel' from module '@jupyter-widgets/base'",
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "26ccacccb90843b7ac727e9595ab1596": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "HBoxModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_name": "HBoxModel",
       "_view_name": "ErrorWidgetView",
       "box_style": "",
       "children": [
        "IPY_MODEL_34fa3b9e3ba74ee399312caeceb99f69",
        "IPY_MODEL_54daa2fea4ec45d0a23452074efd359e",
        "IPY_MODEL_10730754cf7a43eca9820d10763e9a41"
       ],
       "error": {},
       "layout": "IPY_MODEL_6d28c2f4d48043aa90f29bb764583bf9",
       "msg": "Failed to load model class 'HBoxModel' from module '@jupyter-widgets/controls'"
      }
     },
     "34fa3b9e3ba74ee399312caeceb99f69": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "HTMLModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_name": "HTMLModel",
       "_view_name": "ErrorWidgetView",
       "description": "",
       "description_tooltip": null,
       "error": {},
       "layout": "IPY_MODEL_e9446ec84ebf43ada6872814f2acb0bb",
       "msg": "Failed to load model class 'HTMLModel' from module '@jupyter-widgets/controls'",
       "placeholder": "​",
       "style": "IPY_MODEL_690ee1b6a3cf43adb5164f200cad4fb4",
       "value": "100%"
      }
     },
     "45e44448b1ea450db792b21878b84f01": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_name": "ProgressStyleModel",
       "_view_name": "ErrorWidgetView",
       "bar_color": null,
       "description_width": "",
       "error": {},
       "msg": "Failed to load model class 'ProgressStyleModel' from module '@jupyter-widgets/controls'"
      }
     },
     "54daa2fea4ec45d0a23452074efd359e": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "FloatProgressModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_name": "FloatProgressModel",
       "_view_name": "ErrorWidgetView",
       "bar_style": "success",
       "description": "",
       "description_tooltip": null,
       "error": {},
       "layout": "IPY_MODEL_218cfca13d80402ab877a02fc12211cb",
       "max": 11531,
       "min": 0,
       "msg": "Failed to load model class 'FloatProgressModel' from module '@jupyter-widgets/controls'",
       "orientation": "horizontal",
       "style": "IPY_MODEL_9706bfc895af4431a44ab22509435387",
       "value": 11531
      }
     },
     "690ee1b6a3cf43adb5164f200cad4fb4": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "DescriptionStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_name": "DescriptionStyleModel",
       "_view_name": "ErrorWidgetView",
       "description_width": "",
       "error": {},
       "msg": "Failed to load model class 'DescriptionStyleModel' from module '@jupyter-widgets/controls'"
      }
     },
     "6d28c2f4d48043aa90f29bb764583bf9": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_name": "LayoutModel",
       "_view_name": "ErrorWidgetView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "error": {},
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "msg": "Failed to load model class 'LayoutModel' from module '@jupyter-widgets/base'",
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "6e12fbb8b0f349e09c32c14601a1eddc": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "HTMLModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_name": "HTMLModel",
       "_view_name": "ErrorWidgetView",
       "description": "",
       "description_tooltip": null,
       "error": {},
       "layout": "IPY_MODEL_0cf8ded72d3147d681343573a3c5c3ad",
       "msg": "Failed to load model class 'HTMLModel' from module '@jupyter-widgets/controls'",
       "placeholder": "​",
       "style": "IPY_MODEL_07b43483463e4ae3b9a7b160d8938299",
       "value": "Loading graphs: 100%"
      }
     },
     "7821741103e94b03923bba81dbc4c2a9": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "HTMLModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_name": "HTMLModel",
       "_view_name": "ErrorWidgetView",
       "description": "",
       "description_tooltip": null,
       "error": {},
       "layout": "IPY_MODEL_0b87e5fec1ff41aeb4c9e624a222f89b",
       "msg": "Failed to load model class 'HTMLModel' from module '@jupyter-widgets/controls'",
       "placeholder": "​",
       "style": "IPY_MODEL_14391e150296461ca407c9ff78c5ae7e",
       "value": " 11534/11534 [00:56&lt;00:00, 247.95it/s]"
      }
     },
     "9706bfc895af4431a44ab22509435387": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "ProgressStyleModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_name": "ProgressStyleModel",
       "_view_name": "ErrorWidgetView",
       "bar_color": null,
       "description_width": "",
       "error": {},
       "msg": "Failed to load model class 'ProgressStyleModel' from module '@jupyter-widgets/controls'"
      }
     },
     "9b5d28e41af94931b674dcd1ddf14562": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_name": "LayoutModel",
       "_view_name": "ErrorWidgetView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "error": {},
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "msg": "Failed to load model class 'LayoutModel' from module '@jupyter-widgets/base'",
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "c5606e09c1de4100be70fc1ef5fb367d": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_name": "LayoutModel",
       "_view_name": "ErrorWidgetView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "error": {},
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "msg": "Failed to load model class 'LayoutModel' from module '@jupyter-widgets/base'",
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "ce6b2459d7fb430d9b0a8820031189de": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "FloatProgressModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_name": "FloatProgressModel",
       "_view_name": "ErrorWidgetView",
       "bar_style": "success",
       "description": "",
       "description_tooltip": null,
       "error": {},
       "layout": "IPY_MODEL_c5606e09c1de4100be70fc1ef5fb367d",
       "max": 11534,
       "min": 0,
       "msg": "Failed to load model class 'FloatProgressModel' from module '@jupyter-widgets/controls'",
       "orientation": "horizontal",
       "style": "IPY_MODEL_45e44448b1ea450db792b21878b84f01",
       "value": 11534
      }
     },
     "e9446ec84ebf43ada6872814f2acb0bb": {
      "model_module": "@jupyter-widgets/base",
      "model_module_version": "2.0.0",
      "model_name": "LayoutModel",
      "state": {
       "_model_name": "LayoutModel",
       "_view_name": "ErrorWidgetView",
       "align_content": null,
       "align_items": null,
       "align_self": null,
       "border": null,
       "bottom": null,
       "display": null,
       "error": {},
       "flex": null,
       "flex_flow": null,
       "grid_area": null,
       "grid_auto_columns": null,
       "grid_auto_flow": null,
       "grid_auto_rows": null,
       "grid_column": null,
       "grid_gap": null,
       "grid_row": null,
       "grid_template_areas": null,
       "grid_template_columns": null,
       "grid_template_rows": null,
       "height": null,
       "justify_content": null,
       "justify_items": null,
       "left": null,
       "margin": null,
       "max_height": null,
       "max_width": null,
       "min_height": null,
       "min_width": null,
       "msg": "Failed to load model class 'LayoutModel' from module '@jupyter-widgets/base'",
       "object_fit": null,
       "object_position": null,
       "order": null,
       "overflow": null,
       "overflow_x": null,
       "overflow_y": null,
       "padding": null,
       "right": null,
       "top": null,
       "visibility": null,
       "width": null
      }
     },
     "f64c7011415b4ea3a3043c3a81009193": {
      "model_module": "@jupyter-widgets/controls",
      "model_module_version": "2.0.0",
      "model_name": "HBoxModel",
      "state": {
       "_model_module": "@jupyter-widgets/controls",
       "_model_name": "HBoxModel",
       "_view_name": "ErrorWidgetView",
       "box_style": "",
       "children": [
        "IPY_MODEL_6e12fbb8b0f349e09c32c14601a1eddc",
        "IPY_MODEL_ce6b2459d7fb430d9b0a8820031189de",
        "IPY_MODEL_7821741103e94b03923bba81dbc4c2a9"
       ],
       "error": {},
       "layout": "IPY_MODEL_9b5d28e41af94931b674dcd1ddf14562",
       "msg": "Failed to load model class 'HBoxModel' from module '@jupyter-widgets/controls'"
      }
     }
    },
    "version_major": 2,
    "version_minor": 0
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}