{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Solving the Flexible Job-Shop Scheduling Problem (FJSP)\n",
    "\n",
    "The following notebook explains the FJSP and explains the solution construction process using an encoder-decoder architecture based on a Heterogeneous Graph Neural Network (HetGNN)\n",
    "\n",
    "<a href=\"https://colab.research.google.com/github/ai4co/rl4co/blob/main/examples/other/2-scheduling.ipynb\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"></a>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: torch_geometric in /Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages (2.5.0)\n",
      "Requirement already satisfied: tqdm in /Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages (from torch_geometric) (4.66.1)\n",
      "Requirement already satisfied: numpy in /Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages (from torch_geometric) (1.26.3)\n",
      "Requirement already satisfied: scipy in /Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages (from torch_geometric) (1.11.4)\n",
      "Requirement already satisfied: fsspec in /Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages (from torch_geometric) (2023.12.2)\n",
      "Requirement already satisfied: jinja2 in /Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages (from torch_geometric) (3.1.3)\n",
      "Requirement already satisfied: aiohttp in /Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages (from torch_geometric) (3.9.1)\n",
      "Requirement already satisfied: requests in /Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages (from torch_geometric) (2.31.0)\n",
      "Requirement already satisfied: pyparsing in /Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages (from torch_geometric) (3.1.1)\n",
      "Requirement already satisfied: scikit-learn in /Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages (from torch_geometric) (1.4.1.post1)\n",
      "Requirement already satisfied: psutil>=5.8.0 in /Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages (from torch_geometric) (5.9.7)\n",
      "Requirement already satisfied: attrs>=17.3.0 in /Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages (from aiohttp->torch_geometric) (23.2.0)\n",
      "Requirement already satisfied: multidict<7.0,>=4.5 in /Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages (from aiohttp->torch_geometric) (6.0.4)\n",
      "Requirement already satisfied: yarl<2.0,>=1.0 in /Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages (from aiohttp->torch_geometric) (1.9.4)\n",
      "Requirement already satisfied: frozenlist>=1.1.1 in /Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages (from aiohttp->torch_geometric) (1.4.1)\n",
      "Requirement already satisfied: aiosignal>=1.1.2 in /Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages (from aiohttp->torch_geometric) (1.3.1)\n",
      "Requirement already satisfied: async-timeout<5.0,>=4.0 in /Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages (from aiohttp->torch_geometric) (4.0.3)\n",
      "Requirement already satisfied: MarkupSafe>=2.0 in /Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages (from jinja2->torch_geometric) (2.1.3)\n",
      "Requirement already satisfied: charset-normalizer<4,>=2 in /Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages (from requests->torch_geometric) (3.3.2)\n",
      "Requirement already satisfied: idna<4,>=2.5 in /Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages (from requests->torch_geometric) (3.6)\n",
      "Requirement already satisfied: urllib3<3,>=1.21.1 in /Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages (from requests->torch_geometric) (1.26.18)\n",
      "Requirement already satisfied: certifi>=2017.4.17 in /Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages (from requests->torch_geometric) (2023.11.17)\n",
      "Requirement already satisfied: joblib>=1.2.0 in /Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages (from scikit-learn->torch_geometric) (1.3.2)\n",
      "Requirement already satisfied: threadpoolctl>=2.0.0 in /Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages (from scikit-learn->torch_geometric) (3.3.0)\n"
     ]
    }
   ],
   "source": [
    "! pip install torch_geometric"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "\n",
    "import torch\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "from IPython.display import display, clear_output\n",
    "import time\n",
    "import networkx as nx\n",
    "import matplotlib.pyplot as plt\n",
    "from rl4co.envs import FJSPEnv\n",
    "from rl4co.models.zoo.l2d import L2DModel\n",
    "from rl4co.models.zoo.l2d.policy import L2DPolicy\n",
    "from rl4co.models.zoo.l2d.decoder import L2DDecoder\n",
    "from rl4co.models.nn.graph.hgnn import HetGNNEncoder\n",
    "from rl4co.utils.trainer import RL4COTrainer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "generator_params = {\n",
    "  \"num_jobs\": 5,  # the total number of jobs\n",
    "  \"num_machines\": 5,  # the total number of machines that can process operations\n",
    "  \"min_ops_per_job\": 1,  # minimum number of operatios per job\n",
    "  \"max_ops_per_job\": 2,  # maximum number of operations per job\n",
    "  \"min_processing_time\": 1,  # the minimum time required for a machine to process an operation\n",
    "  \"max_processing_time\": 20,  # the maximum time required for a machine to process an operation\n",
    "  \"min_eligible_ma_per_op\": 1,  # the minimum number of machines capable to process an operation\n",
    "  \"max_eligible_ma_per_op\": 2,  # the maximum number of machines capable to process an operation\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "env = FJSPEnv(generator_params=generator_params)\n",
    "td = env.reset(batch_size=[1])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Visualize the Problem\n",
    "\n",
    "Below we visualize the generated instance of the FJSP. Blue nodes correspond to machines, red nodes to operations and yellow nodes to jobs. A machine may process an operation if there exists an edge between the two. \n",
    "\n",
    "The thickness of the connection between a machine and an operation node specifies the processing time the respective machine needs to process the operation (thicker line := longer processing).\n",
    "\n",
    "Each operation belongs to exactly one job, where an edge between a job and an operation node indicates that the respective operation belongs to the job. The number above an operation-job edge specifies the precedence-order in which the operations of a job need to be processed. A job is done when all operations belonging to it are scheduled. The instance is solved when all jobs are fully scheduled.\n",
    "\n",
    "Also note that some operation nodes are not connected. These operation nodes are padded, so that all instances in a batch have the same number of operations (where we determine the maximum number of operations as num_jobs * max_ops_per_job). "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "jupyter": {
     "source_hidden": true
    }
   },
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAxoAAAH4CAYAAADNU5vyAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjguMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8g+/7EAAAACXBIWXMAAA9hAAAPYQGoP6dpAADk4klEQVR4nOzdeVxU97k/8M/AsO+yCSr7DuKGirLNqLjHnSxNbYwm19g2drltb5vYNr03/po2t02bpI0mJNGbxiYFUWPiBjoDgqCiuLDvi4LsyDrDLOf3B5lTRxZhFmZ73q8XL4U553u+6HA4z3d5Hg7DMAwIIYQQQgghRIPMdN0BQgghhBBCiPGhQIMQQgghhBCicRRoEEIIIYQQQjSOAg1CCCGEEEKIxlGgQQghhBBCCNE4CjQIIYQQQgghGkeBBiGEEEIIIUTjKNAghBBCCCGEaBwFGoQQQgghhBCNo0CDEGJ0jhw5Ag6Hg/r6er3rB4/HA4/Hm/a+6Oq6U9Ha2oodO3bA1dUVHA4Hf/nLX6bcxq5du2Bvb6/5zhFCCJkyCjQIIXpv06ZNsLW1RV9f37jHPP/887C0tERnZ+c09ky/lJaW4o033tB5gKWqn/zkJzh//jx+9atf4bPPPsPatWvHPG5wcBBvvPEGhELh9HbwEW+88QY4HM6YH4cOHWKP43A4+OEPf6h0bnt7O370ox8hLCwMNjY28PDwwJIlS/Bf//Vf6O/vZ4/btWuXUruOjo6YN28e/vSnP0EsFk/b90oIIari6roDhBDyJM8//zxOnz6NEydO4Hvf+96o1wcHB3Hq1CmsXbsWrq6u2LlzJ5599llYWVnpoLcTu3DhgtbaLi0txe9+9zvweDz4+flN23U15dKlS9i8eTN+9rOfTXjc4OAgfve73wGAzmdpPvjgg1EzKEuXLh33+K6uLsTExKC3txe7d+9GWFgYOjs7cefOHXzwwQfYt2+fUntWVlZITU0FAPT09OD48eP42c9+huvXr+OLL77QzjdFCCEaQoEGIUTvbdq0CQ4ODjh27NiYgcapU6cwMDCA559/HgBgbm4Oc3Pz6e7mpFhaWprUdaeira0Nzs7Ouu7GlOzYsQNubm6TPv7jjz9GY2Mj8vLysHz5cqXXent7R/0/cblcfPe732U///73v4+lS5fiyy+/xJ///Gd4e3ur9w0QQogW0dIpQojes7GxwbZt23Dx4kW0tbWNev3YsWNwcHDApk2bAIy9N6KwsBBr1qyBm5sbbGxs4O/vj927d7OvC4VCcDicUctx6uvrweFwcOTIEfZrd+7cwa5duxAQEABra2vMnDkTu3fvntSyrcf3Svj5+Y27BEfRl4aGBnz/+99HaGgobGxs4OrqipSUFKXv78iRI0hJSQEA8Pn8UW2MtUejra0Ne/bsgaenJ6ytrTFv3jwcPXp0zO//f//3f/Hhhx8iMDAQVlZWWLx4Ma5fv/7E7xcAamtrkZKSghkzZsDW1haxsbH45ptvlPrO4XDAMAz+9re/sX0fS319Pdzd3QEAv/vd79hj33jjDaXj7t+/jy1btsDe3h7u7u742c9+BplMpnSMXC7HX/7yF0RGRsLa2hqenp7Yu3cvuru7J/V9qaKmpgbm5uaIjY0d9ZqjoyOsra0nPN/MzIz9fzTUJXKEENNBMxqEEIPw/PPP4+jRo/jXv/6ltOa9q6sL58+fx3PPPQcbG5sxz21ra8Pq1avh7u6OX/7yl3B2dkZ9fT0yMjJU6ktmZiZqa2vx4osvYubMmSgpKcGHH36IkpISFBQUjPuQPJa//OUvSuvyAeCdd97BrVu34OrqCgC4fv06rly5gmeffRazZ89GfX09PvjgA/B4PJSWlsLW1haJiYnYv38/3n33Xbz22msIDw8HAPbPxw0NDYHH46G6uho//OEP4e/vj7S0NOzatQs9PT340Y9+pHT8sWPH0NfXh71794LD4eCPf/wjtm3bhtraWlhYWIz7/bW2tmL58uUYHBzE/v374erqiqNHj2LTpk1IT0/H1q1bkZiYiM8++ww7d+5EcnLymLNWCu7u7uwSo61bt2Lbtm0AgOjoaPYYmUyGNWvWYOnSpfjf//1fZGVl4U9/+hMCAwOxb98+9ri9e/fiyJEjePHFF7F//37U1dXh/fffR1FREfLy8ib8vhS6urqUPjc3N4eLi8u4x/v6+kImk+Gzzz7DCy+88MT2x1JTUwMA7PuDmBaZTAaJRKLrbhANsbS0hJmZ8Y77U6BBCDEIK1asgJeXF44dO6YUaKSlpUEikbDLpsZy5coVdHd348KFC4iJiWG//uabb6rUl+9///v4z//8T6WvxcbG4rnnnkNubi4SEhIm3daWLVuUPk9LS8PNmzfx3//935g7dy4AYMOGDdixY4fScU899RSWLVuG48ePY+fOnQgICEBCQgLeffddJCcnP3HvwocffoiysjL84x//YP/tXnnlFSQlJeHAgQPYvXs3HBwc2OMbGxtRVVXFPkSHhoZi8+bNOH/+PDZu3Djudd566y20trbi8uXLiI+PBwC8/PLLiI6Oxk9/+lNs3rwZAQEBCAgIwM6dOxESEqK0VOhxdnZ22LFjB/bt24fo6OgxjxWJRHjmmWfw61//mv2+Fi5ciI8//pgNNHJzc5GamorPP/8c3/nOd9hz+Xw+1q5di7S0NKWvjyc0NFTpc19f3wlnGnbv3o133nkHu3btwltvvQUej4fExESsX78eTk5OY57T0dEBAHj48CH+9a9/4eTJk4iOjh51bWLcGIbBgwcP0NPTo1Y7AwMcNDZaYniYA0tLBj4+w7CzYzTTSTJlZmZm8Pf3N4jlraqgQIMQYhDMzc3x7LPP4p133kF9fT272fnYsWPw9PTEypUrxz1Xse7/66+/xrx58yY1Uj2RR2dORCIR+vv72aUwN2/enFKg8ajS0lLs3r0bmzdvxoEDB8a8nkQiQW9vL4KCguDs7IybN29i586dU77WmTNnMHPmTDz33HPs1ywsLLB//34899xzyM7OVgognnnmGaWResX3WFtb+8TrLFmyhA0yAMDe3h7/8R//gV/96lcoLS1FVFTUlPv/JK+88orS5wkJCfjss8/Yz9PS0uDk5ITk5GT2QR4AFi1aBHt7ewgEgkkFGsePH4ejoyP7+Xizagqenp64ffs2/vu//xsnTpzAoUOHcOjQIVhaWuLAgQM4cOCA0ozYwMAAu1RMYfny5UrfCzENiiDDw8MDtra2U5o5LSvj4KOPzHH+vBnq6jhgmH+fy+Ew8PdnsGaNHC+/LEN4OAUd00Uul6O5uRktLS3w8fGZ0v+poaBAgxBiMJ5//nm88847OHbsGF577TXcu3cPly9fxv79+yfc/J2UlITt27fjd7/7Hd555x3weDxs2bIF3/nOd1TKTNXV1YXf/e53+OKLL0btGXn48OGU2wNGNgJv27YNs2bNwv/93/8p/cIZGhrC73//e3z66ae4f/8+GObfDwKqXq+hoQHBwcGjpuwVS60aGhqUvu7j46P0uSLoeNJ+hoaGhjGzMD16HU0HGtbW1qMezl1cXJT6WlVVhYcPH8LDw2PMNsbaCzSWxMTEKW0GBwAvLy988MEH+Pvf/46qqiqcP38ef/jDH/Cb3/wGXl5eeOmll5S+l9OnTwMYyUDl7++P2bNnT+l6xPDJZDI2yJjKkrm6OmDvXiAzE+ByAal09DEMw0FtLQcffWSGDz7gIjkZOHwY8PfX4DdAxuXu7o7m5mZIpVK1B8H0EQUahBCDsWjRIoSFheGf//wnXnvtNfzzn/8EwzATLpsCRmoZpKeno6CgAKdPn8b58+exe/du/OlPf0JBQQHs7e3HHUl6fAMxADz99NO4cuUKfv7zn2P+/Pmwt7eHXC7H2rVrIZfLVfredu3ahebmZly7dk1phBwAXn31VXz66af48Y9/jGXLlsHJyQkcDgfPPvusytebqvECuUeDHn0xmYxjcrkcHh4e+Pzzz8d8/fFARRs4HA5CQkIQEhKCDRs2IDg4GJ9//rlSoGFubo5Vq1ZpvS9Evyn2ZNja2k76nNRU4NVX/x1cjBVkPErxukAAREQA770HPPJWJFqiWDIlk8ko0CCEEF17/vnn8etf/xp37tzBsWPHEBwcjMWLF0/q3NjYWMTGxuLgwYM4duwYnn/+eXzxxRd46aWX2BH6x9c/Pz6y393djYsXL+J3v/sdfvOb37Bfr6qqUvl7euutt3Dy5ElkZGQgLCxs1Ovp6el44YUX8Kc//Yn9mkgkGtXXqUy7+/r64s6dO5DL5UqzGuXl5ezrmuDr64uKiopRX1fnOppYXhAYGIisrCzExcU9cbnTdAgICICLiwtaWlp03RWixyb73j94EHhk9eWUSKUjHy+/DLS2Aq+/rlo7ZHKMcbnUo4x3mzshxCgpZi9+85vf4NatW0+czQBGgoPHR97nz58PAGyFZV9fX5ibmyMnJ0fpuL///e9KnytGyx9v7y9/+cukv4dHZWVl4cCBA3j99ddHbQx/9JqPX++9994bNdtiZ2cHYHSwNJb169fjwYMH+PLLL9mvSaVSvPfee7C3t0dSUtLUvpEJrnPt2jXk5+ezXxsYGMCHH34IPz8/RERETLlNxaiuOptin376achkMvzP//zPqNekUqnaG27Hc/XqVQwMDIz6+rVr19DZ2UkbvInaUlNVDzIed+AA8PHHmmmLmCaa0SCEGBR/f38sX74cp06dAoBJBRpHjx7F3//+d2zduhWBgYHo6+vDRx99BEdHR6xfvx4A4OTkhJSUFLz33nvgcDgIDAzE119/PWqtvqOjIxITE/HHP/4REokEs2bNwoULF1BXV6fS9/Pcc8/B3d0dwcHB+Mc//qH0WnJyMjw9PbFx40Z89tlncHJyQkREBPLz85GVlTVqrfb8+fNhbm6OP/zhD3j48CGsrKywYsWKMfch/Md//AcOHz6MXbt24caNG/Dz80N6ejry8vLwl7/8RSnjlDp++ctf4p///CfWrVuH/fv3Y8aMGTh69Cjq6upw/PhxldI62tjYICIiAl9++SVCQkIwY8YMREVFTWmvR1JSEvbu3Yvf//73uHXrFlavXg0LCwtUVVUhLS0Nf/3rX0dl+tKEzz77DJ9//jm2bt2KRYsWwdLSEmVlZfjkk09gbW2N1157TePXJKajrm5kuZQm/fCHwIoVhrFn44033sDJkydx69atcY/h8XiYP3++yoNDZGoo0CCEGJznn38eV65cwZIlSxAUFPTE45OSknDt2jV88cUXaG1thZOTE5YsWYLPP/8c/o/89nzvvfcgkUhw6NAhWFlZ4emnn8bbb7896gH22LFjePXVV/G3v/0NDMNg9erVOHv2rEpVmhUZj8aqqSAQCODp6Ym//vWvMDc3x+effw6RSIS4uDhkZWVhzZo1SsfPnDkThw4dwu9//3vs2bMHMpkMAoFgzEDDxsYGQqEQv/zlL3H06FH09vYiNDQUn376KXbt2jXl72M8np6euHLlCv7rv/4L7733HkQiEaKjo3H69Gls2LBB5XZTU1Px6quv4ic/+QmGh4fx29/+dsqbyg8dOoRFixbh8OHDeO2118DlcuHn54fvfve7iIuLU7lvCopZqEf3jOzduxe2tra4ePEiTp06hd7eXri7u2P16tX41a9+hQULFqh9XWK69u598l6MqZJKR9q9cEG183ft2oWjR49i7969OHTokNJrP/jBD/D3v/8dL7zwglJRVG3KyMgwyr0Q+orD6ONOPkIIIcTA9fb2wsnJCQcOHBhziRYhkyUSiVBXVwd/f/9xq8eXlgKRkdrrQ2kpME79zwnt2rULly5dQm9vL1paWtg9USKRCF5eXnB0dASfz9dIoDGZGQ19M5n/W0NGezQIIYQQLbh+/ToAqLQPhZCpOnRoJIWtNnC5wAcfqH7+woULMWfOHGRkZLBfy8jIgI+Pj9Is3rlz5xAfHw9nZ2e4urpi48aNqKmpUWrr3r17eO655zBjxgzY2dkhJiYGV69eVTrms88+g5+fH5ycnPDss8+ir6+PfY3H4+HHP/4x+7mfnx/+3//7f2yRUh8fH3z44YdK7TU1NeHpp5+Gs7MzZsyYgc2bN09YmJP8GwUahBBCiAbduXMHf/7zn/HCCy/A1dVVrSVihEzWmTOaXzalIJUCZ8+q18bu3bvx6aefsp9/8sknePHFF5WOGRgYwE9/+lMUFhbi4sWLMDMzw9atW9k03v39/UhKSsL9+/fx1Vdf4fbt2/jFL36hlOa7pqYGJ0+exNdff42vv/4a2dnZeOuttybs25/+9CfExMSgqKgI3//+97Fv3z42W55EIsGaNWvg4OCAy5cvIy8vD/b29li7di2Gh4fV+0cxAbRHgxBCCNGgjIwMvPXWW4iJicE777wzqi4KIZrW1wfU1mr3GjU1QH8/YG+v2vnf/e538atf/YpNGZ6Xl4cvvvgCQqGQPWb79u1K53zyySdwd3dHaWkpoqKicOzYMbS3t+P69euYMWMGAIzapyeXy3HkyBE2ocXOnTtx8eJFHDx4cNy+rV+/Ht///vcBAP/1X/+Fd955BwKBAKGhofjyyy8hl8uRmprKpqL99NNP4ezsDKFQiNWrV6v2D2IiKNAghBBCNOiNN97AG2+8oetuEBNSUwNoe8ctwwDV1cC3mcGnzN3dHRs2bMCRI0fAMAw2bNgANzc3pWOqqqrwm9/8BlevXkVHRwc7U9HY2IioqCjcunULCxYsYIOMsfj5+SllzfPy8hqVPfBx0dHR7N85HA5mzpzJnnP79m1UV1ePysQnEolGLesio1GgQQghhBBiwL4tB6T319m9ezd++MMfAgD+9re/jXr9qaeegq+vLz766CN4e3tDLpcjKiqKXaI0meKaj2eU4nA4SkurpnpOf38/Fi1ahM8//3zUee7u7k/sj6mjQIMQQgghxIBZWRnGdRT7Gjgczqj03J2dnaioqMBHH32EhIQEAEBubq7SMdHR0UhNTUVXV9eEsxqatHDhQnz55Zfw8PCgZZAqoM3ghBBCCCEGLCgI+Hb7gNZwOCPXUYe5uTnKyspQWlqqVF8GAFxcXODq6ooPP/wQ1dXVuHTpEn76058qHfPcc89h5syZ2LJlC/Ly8lBbW4vjx48jPz9fvY5N4Pnnn4ebmxs2b96My5cvo66uDkKhEPv378e9e/e0dl1jQYEGIYQQQogBs7cHAgK0e43AQNU3gj/K0dFxzJkBMzMzfPHFF7hx4waioqLwk5/8BG+//bbSMZaWlrhw4QI8PDywfv16zJ07F2+99daooEWTbG1tkZOTAx8fH2zbtg3h4eHYs2cPRCIRzXBMAhXsI4QQQgjRY5Mp6rZ//0itC22kuOVygX37gHff1Xzbpo4K9hFCCCGEEL32yivaraOxb5922ibGjQINQgghhBADFxEBJCdrvjo4lzvSbni4ZtslpoECDUIIIYQQI3D4sHYCjcOHNdsmMR0UaBBCCCGEGAF/f+C99yY+xg79mIdbWIKrmIdbsEP/hMe///5Iu4SogupoEEIIIYQYiZdeAlpbgQMH/v21cJTiFRzCepxBAGphhn/nAZKDg1oE4AzW4xBeQRki2NcOHgT27JnO3hNjQzMahBBCCCFG5PXXgY8+AkIt63ABq1GKSOzDBwhCjVKQAQBmYBCEGuzDByhFJC5gNUIt65CaCrz2mo6+AWI0aEaDEELI5PT3A9XVgFg8UiI4KEgzifUJIRr3ElKxm/Mq5BwpwAAWmDglleJ1PkeAUk4EzJj3ALw0DT0lxowCDUIIIeMrLQUOHQLOnAFqa4FHSy9xOCNVwtavH8mtGRExfjuEkOlz8CBw4ADMMPWlK1xGCoilwMsvj6zBev11bfSQmAhaOkUIIWS0ujpg9WogMnKkClhNjXKQAYx8XlMz8npk5MjxdXW66S8hZERqqvIGDXUcOAB8/LFm2iImiQINQgghylJTR2YnBIKRz59UBUzxukAwcl5qqnb7RwgZW10d8Oqrmm3zhz802QGEXbt2YcuWLbruhkGjQIMQQsi/HTw4smRCJJp6mWGpdOS8l18eaYcQMr327tV8eXCpdKRdNTQ1NWH37t3w9vaGpaUlfH198aMf/QidnZ0a6qR66uvrweFwcOvWLaWv//Wvf8WRI0d00idjQYEGIYSQEbTkghDDVVoKZGZqJ9DIzATKylQ6vba2FjExMaiqqsI///lPVFdX49ChQ7h48SKWLVuGrq4uzfb3EcPDw2qd7+TkBGdnZ810xkRRoEEIIYSWXBBi6A4d0nxZcAUud2Qvlgp+8IMfwNLSEhcuXEBSUhJ8fHywbt06ZGVl4f79+3j9283mfn5++J//+R8899xzsLOzw6xZs/C3v/1Nqa2enh689NJLcHd3h6OjI1asWIHbt2+zr7/xxhuYP38+UlNT4e/vD2trawDAuXPnEB8fD2dnZ7i6umLjxo2oqalhz/P/tiLhggULwOFwwOPxAIxeOiUWi7F//354eHjA2toa8fHxuH79Ovu6UCgEh8PBxYsXERMTA1tbWyxfvhwVFRUq/dsZAwo0CCGE6O2SC0LIJJ05o/mfYQWpFDh7dsqndXV14fz58/j+978PGxsbpddmzpyJ559/Hl9++SWYbxNNvP3225g3bx6Kiorwy1/+Ej/60Y+QmZnJnpOSkoK2tjacPXsWN27cwMKFC7Fy5UqlWZHq6mocP34cGRkZ7FKogYEB/PSnP0VhYSEuXrwIMzMzbN26FXK5HABw7do1AEBWVhZaWlqQkZEx5vfzi1/8AsePH8fRo0dx8+ZNBAUFYc2aNaNmZV5//XX86U9/QmFhIbhcLnbv3j3lfzujwRBCCDFtJSUMM5JDSjsfpaW6/g4JMWhDQ0NMaWkpMzQ0NPYBvb0Mw+Fo9+eYw2GYvr4p9bugoIABwJw4cWLM1//85z8zAJjW1lbG19eXWbt2rdLrzzzzDLNu3TqGYRjm8uXLjKOjIyMSiZSOCQwMZA4fPswwDMP89re/ZSwsLJi2trYJ+9Xe3s4AYO7evcswDMPU1dUxAJiioiKl41544QVm8+bNDMMwTH9/P2NhYcF8/vnn7OvDw8OMt7c388c//pFhGIYRCAQMACYrK4s95ptvvmEAjPt/98T/WwNHMxqEEGLq9HTJBSFkksZKP61pDDNSsFOlUyfXt2XLlo36vOzbvSG3b99Gf38/XF1dYW9vz37U1dUpLYPy9fWFu7u7UjtVVVV47rnnEBAQAEdHR/j5+QEAGhsbJ/091NTUQCKRIC4ujv2ahYUFlixZwvZRITo6mv27l5cXAKCtrW3S1zImVLCPEEJMnR4uuSCETIFYrJfXCQoKAofDQVlZGbZu3Trq9bKyMri4uIwKDMbS398PLy8vCIXCUa89umHbzs5u1OtPPfUUfH198dFHH8Hb2xtyuRxRUVFqbxYfj4WFBft3DocDAOwyLVNDMxqEEGLK+vpGKn5rU00N0N+v3WsQYsqsrPTyOq6urkhOTsbf//53DA0NKb324MEDfP7553jmmWfYh/GCggKlYwoKChAeHg4AWLhwIR48eAAul4ugoCClDzc3t3H70NnZiYqKChw4cAArV65EeHg4uru7lY6xtLQEAMhksnHbCQwMhKWlJfLy8tivSSQSXL9+HREREZP41zBNFGgQQogp0/MlF4SQSQgKAr59WNcaDmfkOlP0/vvvQywWY82aNcjJyUFTUxPOnTuH5ORkzJo1CwcfqbmTl5eHP/7xj6isrMTf/vY3pKWl4Uc/+hEAYNWqVVi2bBm2bNmCCxcuoL6+HleuXMHrr7+OwsLCca/v4uICV1dXfPjhh6iursalS5fw05/+VOkYDw8P2NjY4Ny5c2htbcXDhw9HtWNnZ4d9+/bh5z//Oc6dO4fS0lK8/PLLGBwcxJ49e6b872IqKNAghBBTpqdLLgghU2BvDwQEaPcagYEj15mi4OBgFBYWIiAgAE8//TQCAwPxH//xH+Dz+cjPz8eMGTPYY//zP/8ThYWFWLBgAd588038+c9/xpo1awCMLEE6c+YMEhMT8eKLLyIkJATPPvssGhoa4OnpOe71zczM8MUXX+DGjRuIiorCT37yE7z99ttKx3C5XLz77rs4fPgwvL29sXnz5jHbeuutt7B9+3bs3LkTCxcuRHV1Nc6fPw8XF5cp/7uYCg4z2R06hBBCjM+tW8CCBdq/TlERMH++9q9DiBESiUSoq6tTqg0xyv79I4kXtLHfissF9u0D3n1X821/y8/PDz/+8Y/x4x//WGvX0EeT+r81YDSjQQghpkyPl1wQQqbglVe0m9Rh3z7ttE2MGgUahBBiyvR4yQUhZAoiIoDkZM2nquZyR9r9dlM2IVNB6W0JIcTUrV+v3SUX69Zpvl1CyGiHD48EHJr8WeZyR9rVsvr6eq1fg0w/mtEghBBTR0suCDEO/v7Ae+9NeIjUGugLBHrDR/6UPmlbwPvvj7RLiApoRoMQQkydYsmFQKD5kVA+n5ZcEDKdXnoJaG0FDhxgvzTgCzRvAjqXAiIvKA8zywHrFsD1KuD9FWDX8MhrBw8ClLqVqIGyThFCCAHq6kYCDpFIc21aWwOlpTQaSoiaVMpMlJqKoTd/gMofDqM7BoAUEw8vf/u6SyEQ8r4lbH79dwoypgFlnSKEEGL8JrHkYspoyQUhOtO8Hrh+lIPuBd9mlXvSGpZvX+9ewMH1oxw0r6NxaKI+CjQIIYSMeOkl4M03NdMWLbkgRGcaGg6isvJlyBkxYD7FgMGcgZwRo7LyZTQ0HHzy8YRMgAINQggh//b668BHH40se5pqmkwud+S81FTgtde00z9CyISam1NRV3fgyQdOQl3dAbS0fKyRtohpokCDEEKIspdeGtlbweePfP6kgEPxOp8/ch7NZBCiE0NDdaiuflWjbVZV/RBDQ3Uqn79r1y5s2bJlUscKhUJwOBz09PSofD2iXyjrFCGEkNH8/YELF0YCh0OHgLNngZoa4NH8IRzOSDG+detGUthSdilCdKqyci/kcs2mqpbLpais3It58y6odP5f//pXUN4h00WBBiGEkPFFRADvvjvy9/5+oLoaEIsBKysgKIgqfhOiJwYGStHdnamFlqXo7s7EwEAZ7OymPpjg5OSkhT4RQ0FLpwghhEyOvT0wfz6wdOnInxRkEKI3mpsPQXvjx1w0N3+g0pmPLp0Si8XYv38/PDw8YG1tjfj4eFy/fn3UOXl5eYiOjoa1tTViY2NRXFzMvtbQ0ICnnnoKLi4usLOzQ2RkJM6cOaNS34j2UaBBCCGEEGLgOjvPYKQYhjZI0dl5Vu1WfvGLX+D48eM4evQobt68iaCgIKxZswZdXV1Kx/385z/Hn/70J1y/fh3u7u546qmnIJFIAAA/+MEPIBaLkZOTg7t37+IPf/gD7GnQQ29RoEEIIYQQYsCk0j6IRLVavYZIVAOptF/l8wcGBvDBBx/g7bffxrp16xAREYGPPvoINjY2+Phj5cxWv/3tb5GcnIy5c+fi6NGjaG1txYkTJwAAjY2NiIuLw9y5cxEQEICNGzciMTFRre+NaA8FGoQQQgghBmxoqAaAtjdcMxgaqlb57JqaGkgkEsTFxbFfs7CwwJIlS1BWVqZ07LJly9i/z5gxA6Ghoewx+/fvx5tvvom4uDj89re/xZ07d1TuE9E+CjQIIYQQQgwYw4iN6joTeemll1BbW4udO3fi7t27iImJwXvvvafrbpFxUKBBCCGEEGLAOBwrvb9OYGAgLC0tkZeXx35NIpHg+vXriIiIUDq2oKCA/Xt3dzcqKysR/kj67Dlz5uCVV15BRkYG/vM//xMfffSRyv0i2kXpbQkhhBBCDJiNTRAADrS7fIrz7XVUY2dnh3379uHnP/85ZsyYAR8fH/zxj3/E4OAg9jxW5PO///u/4erqCk9PT7z++utwc3NjM1f9+Mc/xrp16xASEoLu7m4IBAKlIIToFwo0CCGEEEIMGJdrD2vrAIhENVq7hrV1ILhc9bI7vfXWW5DL5di5cyf6+voQExOD8+fPw8XFZdRxP/rRj1BVVYX58+fj9OnTsLS0BADIZDL84Ac/wL179+Do6Ii1a9finXfeUatfRHs4DJVrJIQQQgjRWyKRCHV1dfD394e1tfWYx1RV7cf9+x9AOyluuZg1ax+Cg9+d8pnPPfcczM3N8Y9//EML/TJ8k/m/NWS0R4MQQgghxMB5e78CbdbR8PbeN7UzpFKUlpYiPz8fkZGRWuoX0XcUaBBCCCGEGDg7uwi4uCRD86viuXBxSYad3dT2QRQXFyMmJgaRkZF45ZVXNNwnYihojwYhhBBCiBEICTmM69cjIJdrbmbDzIyLkJDDUz5v/vz5GBwc1Fg/iGGiGQ1CCCGEECNgY+OPoCDN1pQIDn4fNjb+Gm2TmA4KNAghhBBCDMBk8vd4e78Ef/83NXI9f/+D8PLa8+QDicqMPScTBRqEEEIIIXrMwsICACa9FMnX93WEhHwEMzNrTH2VPBdmZtYIDU2Fr+9rUzyXTNXw8DAAwNzcXMc90Q5Kb0sIIYQQoudaWlrQ09MDDw8P2NragsPhPPEckagedXU/xMOHFzEScEy0d2PkdSenlfD3fx/W1n6a6TgZl1wuR3NzMywsLODj4zOp/1NDQ4EGIYQQQoieYxgGDx48QE9Pz5TPlUqrMTj4JYaHL0Mma4JyBXEOzM3nwNIyAba2z4LLDdRUl8kkmJmZwd/fny1IaGwo0CCEEEIIMRAymQwSiUSN8/shEtVALhfDzMwK1taBMDdXr+I3UZ2lpSXMzIx3JwMFGoQQQianvx+orgbEYsDKCggKAuzpAYUQQsjYqI4GIYSQ8ZWWAocOAWfOALW1wKNjUxwOEBAArF8PvPIKEBGhu34SQgjROzSjQQghZLS6OmDvXiAzE+ByAekEm0gVrycnA4cPA/6Uc58QQgiltyWEEPK41NSR2QmBYOTziYKMR18XCEbOS03Vbv8IIYQYBAo0CCGE/NvBg8DLLwMi0ZMDjMdJpSPnvfzySDuEEEJMGi2dIoQQMiI1dSRI0GR7e6iqMCGEmCoKNAghhIzsyYiIGJmR0BRr65HN5LRngxBCTBItnSKEEDKy8XuqS6WeRCodaZcQQohJohkNQggxdaWlQGSkdtsPD9de+4QQQvQSzWgQQoipO3RoJEWtNnC5wAcfaKdtQggheo1mNAghxNQFBQE1Ndptv6pKe+0TQgjRSxRoEEKIKevrA5yclCt+axqHA/T2Avb22rsGIYQQvUNLpwghxJTV1Gg3yABG2q+u1u41CCGE6B0KNAghxJSJxcZ1HUIIIXqDAg1CCDFlVlbGdR1CCCF6g/ZoEEKIKevvBxwdaY8GIYQQjaMZDUIIMWX29kBAgHavERhIQQYhhJggCjQIIcTUrV+v3Toa69Zpp21CCCF6jZZOEUKIqaPK4IQQQrSAZjQIIcTURUQAycman9XgckfapSCDEEJMEs1oEEIIAerqRgIOkUhzbVpbj8xm+Ptrrk1CCCEGg2Y0CCGEjAQD772n2Tbff5+CDEIIMWEUaBBCCBnx0kvAm29qpq2DB4E9ezTTFiGEEINES6cIIYQoS00FXn0VkEpHPiaLyx35eP99CjIIIYTQjAYhhJDHvPTSyN4KPn/k8ydtEle8zuePnEdBBiGEENCMBiGEkImUlgKHDgFnzwI1NcoVxDmckWJ869YB+/ZRdilCCCFKKNAghBAyOf39QHU1IBYDVlZAUBBV/CaEEDIuCjQIIYQQQgghGkd7NAghhBBCCCEaR4EGIYQQQgghROMo0CCEEEIIIYRoHAUahBBCCCGEEI2jQIMQQgghhBCicRRoEEIIIYQQQjSOAg1CCCGEEEKIxlGgQQghhBBCCNE4CjQIIYQQQgghGkeBBiGEEEIIIUTjKNAghBBCCCGEaBwFGoQQQgghhBCNo0CDEEIIIYQQonEUaBBCCCGEEEI0jgINQgghhBBCiMZRoEEIIYQQQgjROAo0CCGEEEIIIRrH1XUHCCGEGIj+fqC6GhCLASsrICgIsLfXda8IIYToKQo0CCGEjK+0FDh0CDhzBqitBRjm369xOEBAALB+PfDKK0BEhO76SQghRO9wGObR3xqEEEIIgLo6YO9eIDMT4HIBqXT8YxWvJycDhw8D/v7T109CCCF6i/ZoEEIIUZaaOjI7IRCMfD5RkPHo6wLByHmpqdrtHyGEEINAgQYhhJB/O3gQePllQCR6coDxOKl05LyXXx5phxBCiEmjpVOEEEJGpKaOBAmabG/PHs21RwghxKBQoEEIIWRkT0ZExMiMhKZYW49sJqc9G4QQYpJo6RQhhJCRjd9TXSr1JFLpSLuEEEJMEs1oEEKIqSstBSIjtdt+eLj22ieEEKKXaEaDEEJM3aFDIylqtYHLBT74QDttE0II0Ws0o0EIIaYuKAioqdFu+1VV2mufEEKIXqJAgxBCTFlfH+DkpFzxW9M4HKC3F7C31941CCGE6B1aOkUIIaaspka7QQYw0n51tXavQQghRO9QoEEIIaZMLDau6xBCCNEbFGgQQoiJkkgkyL1+fXouZmU1PdchhBCiN7SUZoQQQoi+YhgG169fx4kTJzDQ2oo4ABxtXpDDGdkQTgghxKRQoEEIISakuroaaWlpqK+vH/mChQXaHR3h0durvYsGBtJGcEIIMUEUaBBCiAlob29HRkYGbt68Oeq14jlzkFRaCnMtbAqXcTi4HxUFt8FB2Nraarx9Qggh+osCDUIIMWKDg4M4c+YMLl26BJlMNuYxORERWFFSopXrmzMMjlhbo+0Xv8DSpUvB5/Mxe/ZsrVyLEEKIfqE6GoQQYoRkMhlycnJw+vRpDAwMPPH4/d98g7DmZs3OanC5AJ+P3vR05ObmIjs7Gz09PQgKCgKfz8f8+fPB1VZFckIIITpHgQYhhBgRhmFw584dHD9+HK2trZM+z7W3F2+kpcFCJtPcxnBra6C0FPD3BzAS/Ny+fRsCgQCVlZVwdHREYmIiEhIS4OzsrKmrEkII0RMUaBBCiJFoampCWloaKioqVDo/rrwc38vJ0VyHUlOBPXvGfKm5uRkCgQBXr16FRCLBwoULwePxEBQUBA5HqzmwCCGETBMKNAghxMD19PTg1KlTyM/Phzq3dGdnZ/ygpwc+H36ofqcOHgRee+2Jhw0NDSE/Px9CoRCtra2YPXs2eDwelixZAiuqvUEIIQaNAg1CCDFQYrEYFy5cwIULFzA8PKxyO5aWllizZg2Sk5NHHu5TU4FXXwWk0pGPyeJyRz7ef3/cmYzxMAyDsrIyCIVC3LlzB9bW1oiLi0NSUhI8PDym+B0RQgjRBxRoEEKIgWEYBvn5+Th16hR6enpUbofD4WD58uXYtGnT6D0SdXXA3r1AZuZI8DBRwKF4PTkZOHyY3ZOhqo6ODuTk5CA3NxcDAwOIiooCj8dDVFQULasihBADQoEGIYQYkIqKCqSlpaGpqUmtdsLCwrBjxw7MmTNn4gNLS4FDh4CzZ4GaGuDRXxkczkgxvnXrgH37gPBwtfr0OIlEguvXr0MgEKCxsRFubm7g8XhYvnw57OzsNHotQgghmkeBBiGEGIDW1lYcP34ct2/fVqsdT09P7NixA3Pnzp367EB/P1BdDYjFgJUVEBQ0LRW/GYZBXV0dhEIhbty4AQ6HgyVLloDP5z85UCKEEKIzFGgQQogeGxgYwOnTp5GdnQ25XK5yO3Z2dti0aRMSEhJgbm6uwR5Or97eXuTm5iInJwfd3d0IDAwEn8/HggULqCYHIYToGQo0CCFED0mlUggEApw5cwaDg4Mqt8PlcrFixQqsW7cOtra2GuyhbsnlcrYmR0VFBdXkIIQQPUSBBiGE6BGGYVBUVITjx4+jo6NDrbYWLVqEbdu2wc3NTUO900/Nzc3Izs5Gfn4+JBIJFixYAD6fTzU5CCFExyjQIIQQPVFfX4+0tDRUV1er1Y6fnx+efvppBAYGaqhnhmFoaAgFBQUQCARobW3FrFmzwOfzqSYHIYToCAUahBCiY11dXThx4gSuXbumVjszZszAtm3bEBMTY9Ij+QzDoLy8HAKBgK3JsXz5cvB4PKrJQQgh04gCDUII0RGRSIRz584hKysLEolE5Xasra2xbt06rFy5EhYWFhrsoeHr7OxETk4OLl++jIGBAURGRrI1OczMzHTdPUIIMWoUaBBCyDSTy+XIy8vDqVOn0NfXp3I7HA4HCQkJ2LRpExwcHDTYQ+MjkUhQWFgIgUCAhoYGuLm5ISkpCXFxcVSTgxBCtIQCDUIImUYlJSVIT09Hc3OzWu1ERkZix44d8Pb21lDPTEd9fT0EAgEKCwvZmhw8Hg8+Pj667hohhBgVCjQIIWQaNDc3Iz09HSUlJWq14+3tjR07diAyMlJDPTNdfX19yM3NRXZ2Nrq7uxEQEAA+n4+FCxdSTQ5CCNEACjQIIUSLent7cfr0aVy+fBnq3G4dHBywefNmxMXF0d4CDVPU5BAKhSgvL4ejoyPi4+ORmJgIFxcXXXePEEIMFgUahBCiBRKJBFlZWTh37hxEIpHK7VhYWGDVqlVYu3YtrK2tNdhDMpaWlhYIhUK2Jsf8+fPB5/MRHBxs0pm8CCFEFRRoEEKIBjEMg8LCQmRkZKCrq0uttpYuXYotW7ZgxowZGuodmSyRSMTW5Hjw4AG8vb3B5/OxdOlSqslBCCGTRIEGIYRoSE1NDf71r3+hvr5erXaCgoKQkpICPz8/jfSLqI5hGFRUVEAgEOD27duwsrJCXFwckpKS4OnpqevuEUKIXqNAgxBC1NTR0YGMjAzcuHFDrXbc3Nywfft2LFiwgJbp6KGuri62Jkd/fz8iIiLA5/OpJgchhIyDAg1CCFHR4OAgzp49i0uXLkEqlarcjo2NDTZu3Agej0fZjgyARCLBjRs3IBAIUF9fD1dXVyQlJSE+Pp5qchBCyCMo0CCEkCmSyWTIycnB6dOnMTAwoHI7ZmZm4PF42LhxIz2gGqjHa3IsXrwYPB4Pvr6+uu4aIYToHAUahBAySQzD4O7du0hPT0dra6tabc2bNw/bt2+ndf5Goq+vD3l5ecjOzkZXVxcCAgLA4/GwaNEimqUihJgsCjQIIWQSmpqakJ6ejvLycrXamTNnDlJSUhAaGqqhnhF9IpfLcefOHQiFQpSVlcHBwQEJCQlUk4MQYpIo0CCEkAn09PTg1KlTyM/PV6vgnrOzMzZv3ozY2FjD3Tjc3w9UVwNiMWBlBQQFAfb2uu6V3mppaUF2djby8/MxPDyM+fPng8fjISQkhDb7E0JMAgUahBAyBrFYjMzMTJw/fx7Dw8Mqt2NpaYk1a9YgOTnZMOsvlJYChw4BZ84AtbXAo78yOBwgIABYvx545RUgIkJ3/dRjipocQqEQLS0t8PLyAp/PR2xsrGG+JwghZJIo0CCEkEcwDIOCggKcPHkSPT09KrfD4XCwbNkybN68Gc7Ozhrr37SpqwP27gUyMwEuF5goq5bi9eRk4PBhwN9/+vppQBQ1OYRCIW7dugUrKyssW7YMfD6f9uoQQowSBRqEEPKtiooKpKWloampSa12QkNDkZKSgjlz5mioZ9MsNRV49dWR4GEqaXu53JGP994DXnpJe/0zAoqaHLm5uejr60N4eDj4fD7mzp1ruEvrCCHkMRRoEEJMXmtrK44fP47bt2+r1Y6npyd27NiBuXPnGu4a/IMHgQMH1G/nzTeB119Xvx0jJ5VK2ZocdXV1bE2OuLg42NP+F0KIgaNAgxBisgYGBvD1119DKBRCLper3I6dnR2eeuopJCYmwtzcXIM9nGapqcDLL2u2vT17NNeekWtoaIBQKMS1a9cAAIsXLwafz6eaHIQQg0WBBiHE5EilUgiFQnzzzTcYHBxUuR1zc3OsXLkS69atg62trQZ7qAN1dSObuUUizbVpbT2ymZz2bExJf38/8vLyIBQK0dXVBX9/f/D5fKrJQQgxOBRoEEJMBsMwKCoqQkZGBtrb29Vqa+HChdi2bRvc3d011DsdW70aEAimtifjSbhcgM8HLlzQXJsmRC6X4+7duxAIBGxNjvj4eCQlJVFNDkKIQaBAgxBiEurr65GWlobq6mq12vHz80NKSgqCgoI01DM9UFoKREZqt/3wcO21bwIePHgAoVCI/Px8iMVitiZHaGio4e4HIoQYPQo0CCFGrbu7GydOnMDVq1fVamfGjBnYtm0bYmJijO/Bbv9+4IMPNDubocDlAvv2Ae++q/m2TZBIJMLVq1chEAjYmhw8Hg+xsbGwtrbWdfcIIUQJBRqEEKMkEolw/vx5ZGZmQiKRqNyOlZUV1q1bh1WrVsHCwkKDPdQjQUFATY1226+q0l77JohhGFRVVUEgEODWrVuwtLTEsmXLwOPxMHPmTF13jxBCAFCgQQgxMnK5HHl5efjqq6/Q29urcjscDgcJCQl46qmn4OjoqMEe6pm+PsDJSbnit6ZxOEBvL0DpWrWiu7sbOTk5uHz5Mvr6+hAWFgY+n4/o6GiqyUEI0SkKNAghRqO0tBTp6em4f/++Wu1ERkZix44d8Pb21lDP9NitW8CCBdq/TlERMH++9q9jwhQ1OYRCIWprazFjxgy2JoeDg4Ouu0cIMUEUaBBCDF5LSwvS09NRXFysVjve3t7YsWMHIrW5MVrfXL0KxMZq/zoFBcDSpdq/DgEwuiZHTEwM+Hw+/Pz8dNsxQohJoUCDEGKw+vr68NVXXyE3N1etgnsODg7YvHkz4uLiTG+pCc1oGLWBgQHk5uYiOzsbnZ2d8PPzY2tyGO2eI0KI3qBAgxBicCQSCS5evIizZ89CpEaBOS6Xi+TkZKxdu9Z0M/b09wOOjrRHw8jJ5XIUFxdDIBCgtLQU9vb2SEhIQGJiImbMmKHr7hFCjBQFGoQQg8EwDAoLC3HixAl0dnaq1daSJUuwZcsWuLq6aqh3BoyyTpmU1tZWZGdnIy8vD2KxGPPmzQOfz6eaHIQQjaNAgxBiEGpra5GWloba2lq12gkMDMTTTz9Na9UfpcU6GjIOB/Xr18Pl//6PRs71jFgsZmtyNDc3Y+bMmeDxeFi2bJnpzvARQjSKAg1CiF7r6OjAiRMnUFhYqFY7bm5u2L59OxYsWECjto/TcmXw//fd76LRzo6qWeupx2tyWFhYsDU5vLy8dN09QogBo0CDEKKXBgcHcfbsWVy6dAlSNUbabWxssGHDBvD5fHC5XA320HjIZDJ0L1kC19u3wZHJNNcwlwvw+RCfPq00ck7VrPVXd3c3Ll++jJycHLYmB4/Hw7x580wvUQIhRG0UaBBC9IpMJsPly5dx+vRp9Pf3q9yOmZkZkpKSsHHjRtjTJuRxtba2Ijc3FysDAuC8fDmgxub6UaytR2ZL/P0BjIycV1dXQyAQoKioiEbO9ZhUKsXNmzchEAhQW1sLFxcXJCUlIT4+nmpyEEImjQINQoheYBgGd+/exfHjx/HgwQO12oqOjsb27dsxc+ZMDfXO+DAMgytXrkAikSAxMXFktDo1FXj5Zc1dJDUV2LNnzJd6enrYata9vb00cq7HGhsb2ZocDMNQTQ5CyKRRoEEI0bl79+4hLS0N5eXlarUzZ84c7NixA2FhYRrqmXHq7e1FVlYWFi9ejDlz5ii/ePAgcOCA+hc5eBB47bUnHiaVSlFUVASBQICamhoaOddjAwMDyMvLQ3Z2Njo6OuDr6ws+n4+YmBiqyUEIGRMFGoQQnXn48CFOnTqFK1euQJ1bkZOTE7Zs2YLY2FgaDX+CO3fu4P79+1i1atX4D4epqcCrr45koZrK/hgud+Tj/ffHncmYSFNTEwQCAY2c6zlFTQ6hUIiSkhLY2dkhPj4eSUlJlC6aEKKEAg1CyLQbHh5GZmYmzp8/D7FYrHI7lpaWWL16NVavXg0rKysN9tD4iMViXLhwAUFBQQgPD3/yCXV1wN69QGbmSPAwUcCheD05GTh8mN2ToarHR879/PzA4/Fo5FwPKWpyXLlyBSKRCNHR0eDz+QgLC6PMYoQQCjQIIdOHYRhcvXoVJ06cQE9Pj8rtcDgcLFu2DJs3b4azs7PG+mesamtrcefOHSQnJ8POzm5qJ5eWAocOAWfPjhT1e/RXBocDBAYC69YB+/YBkwlgpkAul6OkpAQCgYAdOVdUs6aRc/0iFotx7do1CAQC3L9/H56enuDz+VSTgxATR4EGIWRaVFZWIi0tDY2NjWq1ExoaipSUlNF7C8goMpkMAoEAjo6OWLx4sfojzP39QHU1IBYDVlYjFb+nKaNXW1sbW82aRs7111iZxWJjY8Hj8eDt7a3r7hFCphkFGoQQrWpra8Px48dx69Yttdrx9PTE9u3bER0dTQ+Wk9DW1obLly+Dx+MZ1ej/4yPnVM1af/X09LA1OXp7exEaGgo+n0+ZxQgxIRRoEEK0YmBgAN988w0EAgHkcrnK7djZ2eGpp55CYmIizM3NNdhD48QwDPLz8yEWi5GUlGS0D3SKkXOhUIibN29STQ49NlZmscTERCQkJFBmMUKMHAUahBCNkkqlEAqF+OabbzA4OKhyO+bm5lixYgXWr18PW1tbDfbQePX19SEzMxMxMTHw8fHRdXemDY2cG47HM4stWrQIPB4P/v7+NFNJiBGiQIMQohEMw+DWrVvIyMhAW1ubWm0tXLgQ27Ztg7u7u4Z6Z/zu3r2LpqYmrFq1CpaWlrrujk7QyLnhGBgYwJUrVyAUCtmaHDweD4sXLzaozGI5OTl4++23cePGDbS0tODEiRPYsmWLrrtFiN6gQIMQoraGhgakpaWhqqpKrXb8/PyQkpKCoKAgDfXM+InFYmRmZsLf3x+RkZG67o7eaGpqglAoxNWrV9mRc0VNDho51x+KzGJCoRDFxcUGV5Pj7NmzyMvLw6JFi7Bt2zYKNAh5DAUahBCVdXd348SJE7h69apa7bi4uGDr1q1YsmQJPQROQX19PYqKipCcnAz7acr+ZGiMZeTcFCgyi125cgVDQ0OIjo4Gj8dDeHi4QdwXOBwOBRqEPIYCDULIlIlEIpw/fx6ZmZmQSCQqt2NlZYV169ZNXKWajCKXyyEQCGBnZ4elS5caxEOYrjEMw9bkMMSRc1OiyCwmFApx7949eHp6spnFbGxsdN29cVGgQchoFGgQQiZNLpfjypUrOHXqFHp7e1Vuh8PhID4+Hps2bYKjo6MGe2j82tvbkZOTg6SkJLi5uem6OwZprJFzqsmhf8bKLLZ06VLw+Xy9rMlBgQYho1GgQQiZlLKyMqSlpeH+/ftqtRMREYGUlBS9fFDQZ4qq6oODg+DxeJRNSQMMdeTcFD2eWSwkJITNLKYvaa8p0CBkNAo0CCETamlpQXp6OoqLi9Vqx9vbGzt27KANyyro7+9HZmYmFi5cCF9fX113x+gwDIOamhoIBAJ25JyqWesnqVSKW7duQSAQoLq6Gs7OzkhKSkJ8fLzOZ0cp0CBkNAo0CCFj6uvrw+nTp3H58mW1Cu45ODhg06ZNiI+Pp1F4FZSUlKC+vh6rVq2ClZWVrrtj9B4+fIjLly8jOztbb0fOyYh79+5BKBSioKAAcrmczSymq5ocFGgQMhoFGoQQJRKJBJcuXcKZM2cgEolUbofL5SI5ORlr166FtbW1BntoGoaHh5GZmQlfX19ERUXpujsmRzFyLhQKUVVVpVcj50TZ4OAgrly5AoFAgI6ODvj4+IDP509LZrH+/n5UV1cDABYsWIA///nP4PP5mDFjhkkVzSRkPBRoEEIAjCwfuXHjBjIyMtDZ2alWW4sXL8bWrVspm4+KGhsbcePGDaxatYoKzemBx0fOY2JiqJq1Hno0s1hJSQlsbW0RFxen1cQJQqEQfD5/1NdfeOEFHDlyRCvXJMSQUKBBCEFtbS3S0tJQW1urVjuBgYFISUmBv7+/hnpmWuRyObKzs2FtbY3Y2Fh6iNUzipFzoVCI9vb2aR05J1PT3t6O7Oxs5OXlYWhoCHPnzgWPx0NERAT9XBEyjSjQIMSEdXR04MSJEygsLFSrHTc3N2zfvh0LFiygX+Iq6ujoQHZ2NhISEuDh4aHr7pAJKEbOFdWsp2PknKhmeHgY165dg0AgwL179+Dh4cFmFrO1tdV19wgxehRoEGKChoaGcPbsWVy8eBFSqVTldmxsbLB+/XqsWLECXC5Xgz00HQzD4Nq1a+jr68OKFStow7yBGWvknM/nG0w1a1PBMAxqa2shEAhw48YNcLlcNrPYrFmzdN09QowWBRqEmBC5XI6cnBycPn0a/f39KrdjZmaGpKQkbNy4Efb29hrsoWkZGBhAZmYm5s2bR8vNDNzw8DCuX7+OS5cu0ci5nlNkFsvJycHDhw8RHBwMPp+P+fPnU2YxQjSMAg1CTADDMCguLsbx48fR0tKiVlvR0dHYvn07Zs6cqaHemaaysjLU1NQgOTmZ0tYaERo5NxwymQxFRUVKmcUSExORkJBAmcUI0RAKNAgxcvfu3UN6ejrKysrUamf27NlISUlBWFiYhnpmmiQSCTIzMzF79mxER0frujtT098PVFcDYjFgZQUEBQE0ozWux0fOQ0JCwOPxaORcDykyi129ehUymQwLFy4En89HQEAALYEjRA0UaBBipB4+fIhTp07hypUrUOfH3MnJCVu2bEFsbCztH1BTU1MTrl+/jlWrVhnOiGlpKXDoEHDmDFBbCzz6XuJwgIAAYP164JVXgIgI3fVTj9HIueEYHBxEfn4+hEIh2traMGfOHDazmKWlpa67R4jBoUCDECOjKPR2/vx5iMVilduxsLDAmjVrsHr1alraoybF3hhLS0ssW7bMMEZI6+qAvXuBzEyAywUmShqgeD05GTh8GKD9JuOikXPDwDAMSktLIRAIUFxcDBsbG8THx1NmMUKmiAINQowEwzC4evUqTpw4gZ6eHpXb4XA4iI2NxZYtW+Ds7Kyx/pmqzs5OCIVCxMfHw9PTU9fdmZzUVODVV0eCh6lkJeNyRz7eew946SXt9c8I0Mi54VCkns7NzcXQ0BCioqLA5/OpJgchk0CBBiFGoLKyEunp6WhoaFCrnZCQEKSkpMDHx0dDPTNdDMOgsLAQPT09WLFiheGsyT94EDhwQP123nwTeP119dsxcoqRc6FQiLt379LIuR5TZBYTCARoamqCh4cHkpKSsHz5csosRsg4KNAgxIC1tbUhIyMDRUVFarXj4eGBHTt2IDo6mkboNGBwcBAXLlzA3LlzERgYqOvuTF5qKvDyy5ptb88ezbVn5Gjk3DA8mlns5s2bMDc3x9KlS8Hj8TB79mxdd48QvUKBBiEGaGBgAN988w2EQiFkMpnK7djZ2WHjxo1ITEykgnsaUl5ejqqqKiQnJ8Pa2lrX3Zm8urqRzdwikebatLYe2UxOezamhEbODUdvby+bWaynpwfBwcHg8XhYsGCB4cxiEqJFFGgQYkCkUimys7Px9ddfY3BwUOV2zM3NwefzsWHDBnpw0RCJRIKsrCx4e3tj3rx5uu7O1K1eDQgEU9uT8SRcLsDnAxcuaK5NE6IYORcKhbhx4waNnOsxmUyGW7duQSgUorKyEk5OTmxmMScnJ113jxCdoUCDEAPAMAxu376N48ePo62tTa22FixYgG3btsHDw0NDvSP379/H1atXsXLlSsN8qCgtBSIjtdt+eLj22jcBNHJuOO7fvw+hUIiCggJIpVIsWrQIPB4PgYGBtASOmBwKNAjRcw0NDUhLS0NVVZVa7fj6+iIlJQXBwcEa6hlhGAY5OTkwNzdHXFyc4T5E7N8PfPCBZmczFLhcYN8+4N13Nd+2CZLJZLh9+zYEAgEqKyvh7OyMhIQEGjnXQ49nFps9ezb4fD6WLFlCmcWIyaBAgxA91d3djZMnT6KgoECtdlxcXLB161YsWbLEcB+E9VB3dzcuXbqE5cuXw8vLS9fdUYlMJkNnZyecYmJg1dSkvQsFBQFqBspktLFGzqkmh/4ZK7NYXFwckpKS4O7uruvuEaJVFGgQomfEYjHOnz+PCxcuQCKRqNyOlZUV1q5di1WrVtHomYbduHEDnZ2dWLFihd5vopdIJOjo6EBbWxva29vR3t6OtrY2tLW1oaurCxYiEf565Ai0+ljK4QC9vYC9vTavYrJo5NxwKDKL5eXlYXBwEJGRkeDz+YiMjKTgkBglCjQI0RNyuRz5+fk4efIkent7VW6Hw+EgPj4emzZtgqOjowZ7SAYHB5GZmYmIiAi9WoImFouVgohHg4menh5MdJuf3dGBX2dkaL+TRUXA/Pnav44JYxgGZWVlEAgENHKu5yQSCa5fv45Lly6hqakJ7u7u4PF4lFmMGB0KNAjRA2VlZUhPT8e9e/fUaic8PBwpKSmYNWuWhnpGFCorK1FeXo7Vq1frJG3t0NCQUhDx6J8PHz5UuV2/tjb86uRJzXV0PAUFwNKl2r8OAUAj54aCYRjU1dVBKBSisLAQZmZmWLp0Kfh8PmUWI0aBAg1CdKilpQXHjx/H3bt31WrHy8sLO3bsoIcILZBKpbh48SI8PDywYMECrV2HYRgMDAyMWuKk+LO/v18r16UZDeOmGDkXCARobGykkXM91tvbi9zcXGRnZ6OnpwdBQUHg8/mYP3++3i/RJGQ8FGgQogN9fX34+uuvkZOTA7lcrnI7Dg4O2LRpE+Lj42FmZqbBHhIAaG5uRn5+PlauXAlnZ2e122MYBn19feyypseDiaGhIfU7PUWDbW04evIktPnuYTgcdNfXY4aPjxavQiYy1sh5bGws1eTQQ49nFnN0dGRrcmjiPkTIdKJAg5BpJJFIIBAI8M0330CkRgVmLpeLVatWYe3atbCxsdFgDwkw8lCWm5sLAIiPj5/SLBHDMOjp6RkzmGhvb4dYLNZWt6dEJpOhrq4Ozc3NOFdTA181Eg88SZujI3797LPw8vJCZGQkIiMjERwcDAsLC61dk4xvvJFzqsmhf5qbm9nMYhKJBAsXLgSPx0NQUBDNXhODQIEGIdOAYRjcuHEDGRkZ6OzsVKutxYsXY+vWrXB1ddVQ78ijenp6cPHiRSxbtgze3t5jHiOXy9HV1TXungmpNupRaFB/fz/Ky8sxNDQEKysrvPbgAbY+eACtLM7gciF5+WXcfeklFBcXo7S0FN3d3bCwsEBoaCgiIiIQFRUFDw8PenCaZnK5HLdu3WJHzp2cnNiaHDRyrl+GhobYzGKtra2YPXs2eDwelixZAisrK113j5BxUaBBiJbV1tYiLS0NtbW1arUTEBCAlJQUBAQEaKhn5HFFRUVoa2vDypUrAQCdnZ1KGZwUf+/o6FBrydt0Mjc3h5ubG9zd3eHm5oaGhgYUFRXBysoKVlZWMDMzg1d3N95IS9NeJx6pDM4wDFpaWlBSUoKSkhJUVVVBKpXC1dUVkZGRiIqKQmhoqE423JuysUbO+Xw+VbPWMwzDoLy8HAKBAHfu3IGNjQ2WL1+OpKQkeHh46Lp7hIxCgQYhWtLZ2YkTJ07g+vXrarXj6uqKbdu2YdGiRfQLX8MkEgna29tx7949nDt3Dra2tuByuWhvb0dnZ+eEaWH1iYWFBdzd3dkPDw8PeHh4wN3dHTNmzICZmRk6Ojrw6aeforq6esw29n/zDcKam2Guye+ZywX4fODChXEPEYvFqKqqQnFxMUpKStDW1gYzMzMEBQWxy6xmz55N7/1pQiPnhqOzsxPZ2dnIzc3FwMAAoqKiwOPxEBUVRT8vRG9QoEGIhg0NDeHs2bO4ePGiWktorK2tsWHDBvD5fFrLrgaRSMTuj3h8dqK7uxsPHz5Ed3c3Zs+erdeZXaysrNgg4vFgwtnZedwHC4ZhUFBQgC+++GLCfUGuvb14Iy0NFjKZ5or3WVuPzGb4+0/6lPb2dpSUlKC0tBTl5eUQi8VwdHRkg47w8HDYU+E/raORc8PxeGYxNzc3NrOYnZ2drrtHTBwFGoRoiFwux+XLl/HVV1+plYrUzMwMiYmJ2LhxIxwcHDTYQ+M1ODg4ZrG69vb2cYsfyuVy3L9/H9bW1npTzMzGxkYpgHj07w4ODlMepRwYGMA//vEP3Lx5c1LHx5WX43s5Oap0fWypqcCePSqfLpVKUVNTwy6zunfvHjgcDnx9fREVFYXIyEj4+flRxjUtG2vknGpy6J9HM4vduHEDHA4HS5YsAZ/Px5w5c3TdPWKiKNAgRE0Mw6CkpATp6eloaWlRq63o6Ghs27YNXl5eGuqdcWAYBv39/eMGEwMDA1Nqb3BwEK2trfD29p725SAODg5KS5weDShsbW019uBWWlqKI0eOTLmY37qbN7GlsBAMoN7MxsGDwGuvqdPCKD09PSgtLWVnPAYHB2Fra4vw8HB2xoM2MWuPYuRcKBSioaGBRs71mCKzWE5ODrq7uxEYGMhmFtPnmVtifCjQIEQN9+/fR1paGsrKytRqZ/bs2dixYwfCv90wa4oYhsHDhw/HLFbX1tamVjrgR6/x4MEDAMDMmTO1Nhrr5OSkFEQ8+qe20xFLJBJkZGTg0qVLKp0fExOD7w0Pw+pnPwOk0pGPyeJyRz7ef1+tmYzJkMvlqK+vZ4OOuro6MAyDWbNmsUFHUFAQPVRpAcMwqK+vh0AgoJFzPSeXy9maHBUVFVSTg0w7CjQIUUFvby9OnTqFvLw8tTYMOzo6YsuWLVi2bJlJLP9gGAbd3d3jpoUdHh7W2rWHh4dx//59eHh4qD36yuFw4OLiMiqIUHzoatNsU1MTPv74Y5Vm1qytrfGd73wHS5YsGQnA6uqAvXuBzMyR4GGigEPxenIycPjwlPZkaMrAwADKyspQUlKC4uJi9Pb2wtLSEmFhYWzgoS9L5IxJX18fW5ODRs71W3NzM7Kzs5Gfnw+JRIIFCxaAz+dTTQ6iVRRoEDIFw8PDyMrKwrlz59QqvGZhYYE1a9Zg9erVRpfJRS6Xo7Ozc8xgoqOjQyc1Jjo7OzE4OIhZs2ZNOqAzMzODq6vrmLMSbm5uerVBXy6XIzMzE6dOnYJMJpvy+cHBwXjxxRfHrs1SWgocOgScPQvU1ACP/srgcIDAQGDdOmDfPjaFra4xDIP79++zezuqq6shk8ng7u7OBh2hoaFG97OnSzRybjiGhoZQUFAAgUCA1tZWzJo1C3w+nzKLEa2gQIOQSWAYBteuXcOJEyfQ3d2tVluxsbHYsmULXFxcNNS76SeVStHR0aFU8Vrxd32qMSGTyXDv3j04OTmN+bCjqDHx+AZsd3d3uLq6GkSV5M7OThw5cgSVlZVTPtfc3BybN29GcnLy5AKw/n6guhoQiwErKyAoCDCADFAikQgVFRVs4NHR0QEul6uUQtfb25tGdTXk8ZFzqmatnx7PLGZtbY3ly5eDx+NRZjGiMRRoEPIEVVVVSEtLQ0NDg1rthISEYMeOHfD19dVQz7RreHiYDSYen53o6urS+xoTvb296Orqgr+/P7y8vJSCCUVA4eLiYrBL1hTB77Fjx1Tav+Ll5YU9e/aY3Jp6hmHYFLrFxcWoqKiARCKBs7OzUgpdW1tbXXfV4IlEIuTn59PIuQHo7OxETk4OLl++jIGBAURGRrI1OQz1Hkn0AwUahIyjra0NGRkZKCoqUqsdDw8P7NixA9HR0Xo3mqeoMTFWMNHT06Pr7k2alZWV0rKmxsZGzJkzBytXrpywxoShGhwcxOeff47CwkKVzufz+di+fbteLf/SFYlEgurqana2o7m5GRwOB/7+/mylch8fH3rYUgONnBsOiUSCwsJCCAQCNrNYUlIS4uLiKLMYUQkFGoQ8ZnBwEN988w0EAoFK690VbG1tsXHjRiQlJel0U+TAwMCYG6/b2trQ19ens35Nla2t7Zj1JTw8PGBvbw8Oh4PW1lbk5uaCx+ONvd/ACJSXl+PTTz9VKRB0cnLCCy+8gMjISM13zEh0d3ezQUdZWRmGhoZgZ2eHiIgIdsbD0dFR1900WMY2cv773/8eGRkZKC8vZ4sa/uEPf0BoaKiuu6Y2RWaxwsJCNrMYj8eDj4+PrrtGDAgFGoR8SyqVIjs7G998882U6zI8ytzcHHw+H+vXr5+WESBFjYmx6ku0tbVhcHBQ633QFAcHh1HLmxSfT/RvyTAMrly5AolEgsTERIN8YHkSiUSCkydPIisrS6XzFyxYgJ07d9Ko5BTI5XLU1dWhuLgYJSUl7PLJOXPmsEFHQEAAZVdSgUQiwY0bN3Dp0iWDHjlfu3Ytnn32WSxevBhSqRSvvfYaiouLUVpaalDfx0QezywWEBAAPp+PhQsX0nufPBEFGsTkMQyD27dv4/jx42hra1OrrQULFmDbtm0aXw6gqDExXjChTgas6ebs7DxmsTp3d3dYW1tPub3e3l5kZWVh8eLFRrvf4N69e/j444/R3Nw85XOtrKzw7LPPYtmyZUa3hGy69fX1sSl0S0pK0NfXB2tra4SGhrKVyo11Jk2bxho5N9SaHO3t7fDw8EB2djYSExN13R2NUmQWEwqFKC8vh6OjIxISEpCYmEiZxci4KNAgJq2xsRFpaWkqZex5lK+vL1JSUhAcHKxyG3K5XKnGxOMBhUQiUauP04XD4WDGjBljBhNubm6wtLTU2LXu3LmD+/fvY9WqVUa534BhGGRlZeHkyZMqpQUODAzE7t274ebmpoXemTaGYdDU1MQGHTU1NZDL5fD09GSDjpCQEKN8X2qLMdTkqK6uRnBwMO7evYuoqChdd0drWlpaIBQK2cxi8+fPB5/PR3BwMA1oECUUaBCT1N3djZMnT+Lq1atqZU9ycXHB1q1b/13k7AlkMhk6OzvHrHzd0dGh1p6Q6WRmZgY3NzelJU6KP93c3LT+UCAWi5GZmYnAwECjrabe3d2NTz/9FBUVFVM+18zMDJs2bcKaNWuMchmZPhoaGkJ5eTkbeHR1dYHL5SIkJIRdZqXNavTGRC6X486dOxAIBAY1ci6Xy7Fp0yb09PQgNzdX192ZFiKRiK3J8eDBA3h7e4PP52Pp0qWUWYwAoECDmBixWIwLFy7g/Pnzas0QWFlZYe3atVi1atWoEXqJRIKOjo4xg4muri69qTHxJFwul60x8XgwMWPGDJ3VmKitrcWdO3eQnJxsNGugH3f9+nUcO3ZMpf01np6e2LNnj8GkUTZGDMPgwYMHKC0tRUlJCSoqKiCVSjFjxgw26AgLC4ONjY2uu6r3DGnkfN++fTh79ixyc3Mxe/ZsXXdnWjEMg4qKCggEAty+fRtWVlaIi4tDUlISPD09dd09okMUaBCTIJfLkZ+fj1OnTuHhw4cqt8PhcBAXF4e1a9dCLBYrBROKv3d3d+t9jQkFCwuLMYvVubu7612NCblcjkuXLsHR0RGLFy/Wu4cMTRgcHMQ///lPXLt2TaXzeTwetm/frtHlaUR9w8PDqKqqYmt3tLa2wszMDIGBgYiMjERERAR8fHyM8j2tKfo+cv7DH/4Qp06dQk5ODvz9/XXdHZ3q6upiM4v19/cjIiICfD7fYDOLEfVQoEGMXnl5OdLS0nDv3r0pnSeVSiESiTA0NIShoSG4uLggICAAw8PDagUr083a2nrMWQl3d3c4OTkZxMNNW1sbLl++bNRpaysrK/HJJ5+oVHne0dERL7zwglGvCTcmnZ2dSil0xWIxHBwc2BS6ERERcHBw0HU39ZJi5FwoFOLWrVs6HzlnGAavvvoqTpw4AaFQqNY+PWOjyCwmEAhQX18PV1dXJCUlIT4+3mhno8loFGgQo/XgwQMcP34cd+7cGfN1hmEglUoxNDSkFFAo/q5YWmVra4uAgAC4uLjo7UO5nZ3dmPUl3N3d2RoThohhGOTn50MsFiMpKckoR8OkUilOnTqFzMxMlWbC5s2bh507d9KDqYGSSqWora1FaWkpiouL0dTUBA6HAx8fH6UUusb43leXPoycf//738exY8dw6tQppdoZTk5OtDTuEfX19RAKhbh+/To4HA4WL14MHo9HSzxNAAUaxOj09/fj9OnTyMnJgUwmg0QiUQoiHg0qJsrkY2FhAT8/P73ZwOno6DgqmFB8GOPoUF9fH7KysrBo0SKjLRDV3NyMjz/+eMqzbcDIPqFnnnkGy5cv14v3J9GM3t5eNugoLS3FwMAAbGxsEB4ezgYeLi4uuu6mXtHlyPl4P3uffvopdu3apdVrG6K+vj7k5eUhOzsbXV1dCAgIAI/Hw6JFiwwmsxiZGgo0iMFjGAY9PT1oaWnBuXPncOnSJTx8+JANKKaaycnMzAyzZs3CnDlzpv3G5+LiMmaxOlVrTBiq4uJiNDY2jrnZ3hgwDINLly4hIyNDpbS1AQEBePHFFzVer4XoF7lcjsbGRnaZVW1tLRiGgZeXFyIjIxEVFYWgoCBKofuIsUbO+Xy+0Q5WGCpFZjGhUIiysjI4ODiwmcUokDYuFGgQgyCXy9HV1TVmsbr29na0tLSgrq4OIpFIreu4u7vD399faw/1ihoTYxWrc3d3N/kHBrFYjKysLPj5+SEyMlLX3dGKnp4eHDlyBGVlZVM+18zMDBs2bMD69etpKY0JGhwcVCoY2NPTAwsLC4SGhrKzHR4eHjTDBRo5NyQtLS3Izs5Gfn4+hoeHMX/+fPB4PISEhNB72QhQoEH0hlQqHbPGRHt7+7g1Jvr6+lBTU4Pe3l61ru3o6IiAgAA4Ojqq1Q7w7xoTjwYQioDC1dWVfsmNo76+HkVFRUhOToa9vb2uu6MVN27cwD/+8Q+V0tZ6eHhg9+7dJp/RhoxgGAYtLS0oLi5GSUkJqqurIZVK4ebmxgYdoaGhJjUTOha5XI67d+9CIBDQyLmeU2QWEwqFaGlpgZeXF/h8PmJjY/UisxhRDQUaZFopakw8Xl+ivb0dnZ2dk94MKxKJUF9fj7a2NrX6Y21tDX9/f7i5uU1p5ITL5Y5ZrE5RY4JGmydPLpdDIBDA3t5+0oUPDc3Q0BC++OILFBQUqHR+QkICUlJS6JctGZdYLEZlZSU729HW1gZzc3MEBgaylcpnzZpllD9fkzXWyLm+1uQwZWNlFlu+fDl4PB7V5DBAFGgQjXu8vsSjwURPT49aNSZkMhkaGxtx//59tQrfcblczJkzB7NmzRo3KLC0tBwzLayHhwecnZ3pF5MGtLe3IycnB0lJSXBzc9N1d7SiqqoKn376KTo7O6d8roODA773ve8hOjpaCz0jxqytrU2pYKBYLIajoyOioqIQERGBiIgIo0wiMRmPj5x7e3uDx+PRyLkeUmQWy83NRV9fH8LDw8Hn8zF37lwa0DMQFGgQlQwODo5ZrK6trU3tZUxjUVTabWhowPDwsMrtcDgceHl5wdfXFxYWFrC2toanp+eYsxOOjo4UTGgJwzC4evUqBgYGwOfzjfIXhlQqxenTp3H+/HmVguvo6Gjs3LlTI8v5iGmTSqWorq5mZzvu378PDofD7oWKjIyEn5+fUf4cToRhGFRWVkIgENDIuZ6TSqVsZrG6ujo2s1hcXJzRLrU1FhRoTEF/P1BdDYjFgJUVEBQEGOv7m2EYDAwMjNoroQgmBgYGpq0v3d3dqKmpUWldOzCSptbGxgZBQUFYvXo1wsLC2GDCzs6Ogolp1t/fj8zMTCxYsAB+fn667o5WtLS04OOPP0ZTU9OUz7W0tERKSgoSEhL0771pSjdBI9bT08POdpSWlmJwcBC2trZKBQOdnZ113c1pRSPnhqOhoQFCoRDXrl0DADazmLZrckil/RgaqgbDiMHhWMHGJghcLt3/noQCjScoLQUOHQLOnAFqa4FH/7U4HCAgAFi/HnjlFSAiQnf9VAXDMOjt7R03mFA3g5O6BgYGUFtbO6lKyZaWlrCxsYGNjQ2sra2V/vTz88OOHTsQHh4+Db0mEykpKUF9fT1WrVpllEsUGIaBUCjE8ePH2YKPU+Hn54fdu3fr12iqMd8ECeRyOerr69nZjvr6ejAMg9mzZ7NBR1BQkMkksaCRc8PR39+PvLw8CIVCdHV1wd/fH3w+X6OZxQYGStHcfAidnWcgEtUCePSRmQNr6wC4uq6Ht/crsLOj+99YKNAYR10dsHcvkJkJcLnARKnuFa8nJwOHDwP6lBSGYRh0d3ePWuKk+FOdZUjaMjw8jIaGBjx48EBpyYmVlZVSAPHo383NzUe14+joiC1btmDZsmU0IqVjw8PDyMrKgo+PD6KionTdHa14+PAhjh49ipKSkimfy+Fw2LS1Y72XdcJYboJkSgYGBtjZjpKSEvT29sLKyopNoRsVFWW0+6ke9/jI+ZIlS6iatR4aK7NYfHw8kpKSVM4sNjRUh8rKvejuzgTABTBRvaOR111ckhESchg2NnT/exQFGmNITQVefXXk9+ZUamlxuSMf770HvPSS9vr3OEWNibHqS7S3t6tUEEwXFDMszc3N7P6JRwOKyQYLFhYWWL16NdasWWOUo+aGprGxEYWFhUhOToaDg4Ouu6MVRUVF+Oyzz1RaUuju7o7du3cjICBACz1TkaHdBIlWMAyDe/fusZXKq6urIZfL4eHhwe7tCAkJMfr77HSMnBPNaG1thVAoxJUrVyAWi9maHKGhoZNeitrcnIrq6lchl0sxcYDxOC7MzLgICnoP3t50/1OgQOMxBw8CBw6o386bbwKvv65+OwqKGhNjBRMdHR1qZWCaTmZmZqNqS7i5uaGpqQlCoRAPHz5Uq/3Y2Fhs2bKF8qPrAblcjuzsbFhbWyM2Nlb/9htogEgkwpdffokrV66odH58fDyefvpp/XpQ09ebINE5kUiEiooKtnZHZ2cnuFwugoODERERgaioKHh5eRnlzzqgnZFzoh0ikQhXr16FQCBga3IoMotNVFumoeEg6urUv//5+78JX1+6/wEUaChJTQVeflmz7e3ZM/njh4eH2RoTjy916urqUist7HSysLBgg4nHq18/XmOiqqoK6enpqK+vV+uawcHBSElJoSltPdHZ2QmhUIiEhAR4eHjoujtaUVNTg08++QQdHR1TPtfe3h47d+7E/PnzNd8xdej6JkgMBsMwaGtrY5dYVVRUQCKRwNnZmV1iFRYWBltbW113VSvGGjnn8/lUzVrPMAyDqqoqNrOYpaUlli1bBh6Ph5kzZyod29ycispKzd3/QkNT4eVF9z8KNL5VVzeyj1GT+5+trUf2UT66XFkkEo1KB6v4e09Pj+YurmVWVlZj1pdwd3efVI2J9vZ2ZGRk4ObNm2r1w8PDA9u3b8e8efPo5q4HGIbB9evX0dvbixUrVhjl3hiZTIavv/4aZ8+eVSn4j4qKwgsvvKB/aWun6yZIjJJEIkF1dTU729HS0gIzMzP4+/uzy6x8fX2N7j6tGDkXCoVobm6e9Mg5mX7d3d3IycnB5cuX2cxiPB4P0dHREIsbcP16BORyzd3/zMyssXhxqcnv2aBA41urVwMCwdSWIz+JuTmDuXM7sH//12ww0dfXp7kLaJmNjY1SAPHonw4ODir9whgcHMQ333wDgUAAmUymct9sbW2xceNGJCUl0RpZPTEwMIDMzEzMmzcP/kb6YNna2oqPP/4YDQ0NUz7XwsICO3bsQFJSkn4+bGnjJsjlAnw+cOGC5tokBqGrq4tNn1taWgqRSAR7e3ulFLp6F2yrICcnB2+//TYKCwvx4MEDvPLKKwAw4cg50S1FZjGhUIja2lrMmDEDq1adh7n5HQCqP5eMxoWLCx/z5pn2/Y8CDYwMuEVGaq/9lJTfwsXlgfYuoAZ7e/txq1/b2tpq7IFIJpMhOzsbX3/9tVo1OMzMzMDn87FhwwaTrWqrj8rKylBTU4Pk5GT92m+gIQzDICcnB2lpaSqlrfX19cXu3bv194FD2zfB0lKA0kubLJlMhrq6OnaZlSJQnzNnDqKiohAZGYmAgAD9ybg2BWfPnkVeXh4WLVqEbdu24cSJE0hKSho1ck41OfRTQ0MD8vI+h7e39vZTLF5cCjs7073/UaABYP9+4IMPNDuQp8DhyBARkY24uC813/gkOTk5jRlMuLu7w8bGRqvXZhgGd+7cwfHjx9Ha2qpWWwsWLMC2bduMds2/IZJIJMjKysKsWbMQHR2t6+5oRW9vL/7v//4Pd+/enfK5HA4H69atw4YNG/R75k2bN0EuF9i3D3j3Xc23TQxSX1+fUsHAvr4+WFtbIywsjF1m5erqqutuThmHw8GJEyewZcsWAGOPnCtqchhrBj5DVFW1H/fv/x2anc1Q4GLWrH0IDjbd+x8FGhgpbltTo732HR3b8Oyzv9Za+xwOBy4uLmzw8Ggg4e7urrMR5sbGRqSlpaGyslKtdnx8fJCSkoKQkBAN9Yxowr1793Dt2jWsWrXKKJZAjOX27dv47LPPVFry6Obmht27dyMwMFALPdMwbd8Eg4KAqirttU8MFsMwaGxsZIOOmpoayOVyzJw5UymFroWFha67+kSPBxqPerwmR0xMDPh8Pvz8/Ka3k2SUgoIgiETau/9ZWwchNtZ0738mH2j09QFOTsrFbjWPwYsv/ggWFmKVWzAzM8OMGTPG3DPh5uamVzfhnp4enDx5EgUFBWplynJ2dsbWrVuxdOlS/VzTbqLkcjlycnJgYWGB5cuXG+X/jVgsRlpaGi5fvqzS+cuXL8czzzxjGJtBp+MmyOEAvb0AVVYmTzA0NITy8nKUlJSguLgY3d3dsLCwQEhICBt4eHp66uV9Z6JAQ2FgYAC5ubnIzs5GZ2cn/Pz82Joc+vR73FRIpX3IzXWCcsVvTeMgPr4XXK5p3v9MPtC4dQtYsED719m27X/g5nZvwmPMzc3h5uY2ZjDh6uqq9+tXxWIxLly4gAsXLqhVcdzKygpr167FqlWrYGlpqcEeEnV1dXVBIBAgPj4enp6euu6OVtTW1uLTTz9FW1vblM+1s7PDd7/7XSxcuFALPdOS6boJFhUB+pbOl+g1hmHw4MEDdm9HZWUlpFIpXF1d2U3l4eHhehPQTybQUJDL5SguLoZAIEBpaSns7e2RkJCAxMREzJgxQ/udJQCAvr5buHFD+/e/RYuK4OAwX+vX0Ud6vGh4eohVn2SYErl85J9aUWNirD0TLi4uBrlRTC6Xo6CgACdPnlSr4B6Hw8Hy5cuxefNmODk5abCHRBMKCwvR3d2NLVu26H3Qqwq5XI5vvvkGZ86cUakAZkREBF544QU4OztrvnPaNF03wem6DjEaHA4HXl5e8PLywqpVqzA8PIzKyko28Lh8+TLMzMwQGBjI1u6YPXu2Xs52PM7MzAzR0dGIjo5Ga2srsrOzIRAIcO7cOZWqWRPVMMz03Jem6zr6iGY0bk3PYN7Jkw1ISnKCk5OTUd04ysvLkZ6ejqamJrXaCQsLQ0pKCmbPnq2hnhFNGRwcRGZmJqKiogxjv4EK2tra8Mknn6Curm7K51pYWGD79u3g8XiG+bNNMxrEQHV0dKC0tBTFxcUoLy+HWCyGg4MDu8QqPDx8WjddT2VGYyxisZitZk01OaYHzWhon8kHGv39gKMjLU+eqtbWVqSnp+POnTtqtTNz5kzs2LEDUVFRhvmQZuQqKipQWVmJ5ORko/xFxzAMcnNz8a9//Uul5X5z5szBnj174OXlpYXeTZNpuAkyAJpKSjAnPJx+zolWSKVS1NbWsrMdTU1N4HA48PX1RUREBKKiouDv76/xVQP9/f2orq4GMJIZ8c9//jP4fD5mzJgBHx8fldp8vJq1hYUFW5PDoO81ekgq7UduriNoj4b2mHygAWg/4cqsWYNoarIxil+w/f39+Prrr5Gdna3S8hIFe3t7PPXUU0hISDDKZTiGTiqVIisrCzNnzsR8Ix2F7uvrw2effYbbt29P+VwOh4PVq1dj06ZN+p22drK0fBNsd3LCgWeegYODg1LBNkrxSbTl4cOHSil0BwYGYGNjg/DwcERFRSEiIgIuLi5qX0coFILP54/6+gsvvIAjR46o3X53dzcuX76MnJwc9PX1ISwsDDweD/PmzTPIpdb6iLJOaRcFGtBuCnlACheXL5CYmI6XX34Zq1atMsiCZlKpFJcuXcKZM2cwNDSkcjtcLhcrV67EunXrtF7Dg6jm/v37uHr1KlauXGm0e2Xu3r2Lo0ePqpS2dsaMGdi9ezeCg4O10DMd2b8fzN//Do5MC3nkuVzI9+5F9f797DIXxWizj48Pu8wlICCAHpyIVsjlcjQ2NqK4uBglJSWoq6sDwzDw9vZm339BQUF6nfVJKpXi5s2bEAqFqKmpgYuLC5KSkhAfH08Bu5pG6mh8AEAbD4FUR4MCDWi/KG509LMYHr4Nc3NzBAQEICUlBatXrzaIrD0Mw6CoqAjHjx9HR0eHWm3FxMRg69atcHNz01DviCYxDMNuroyLizOKGbjHicVipKenIycnR6XzY2Nj8eyzzxpVkCwWi3HzH//Aspde0t5FHqsM3tvbywYdj482Kx78NDHaTMhYBgYG2BS6JSUl6OnpgaWlJUJDQ9n3n7u7u97eAxsbG9maHAzDUE0ONQ0MlOL6de09BFJlcAo0AACrVwMCgWZnNTgcGby9y7Fhw7vo7+9HS0sL2traYG5uDn9/fyQlJWHFihWYO3euXo7k1dXVIS0tDTVqLqlQBFcBAQEa6hnRtO7ubly6dAnLly832jXA9fX1+OSTT1SqUG9ra4vnn38eMTExWuiZbjAMg6tXr6KzsxMJCQlw3LFD8zdBLhfg84ELF8Y9RDHarHjoq62tBcMw8PLyYjMJ6ftoMzFcDMOgubmZff9VVVVBJpPBzc0NUVFRiIyMRGhoqF6uRBgYGEBeXh6ys7PR0dFBNTnUcPv2anR3C6DZWQ0uXFz4mDdv/PufKaBA41t1dUBEBCASaapFBubmEqSkvAFHx072q1KpFA8ePEBLSws7wxEYGIikpCTExcXBXg92jHd2duLEiRO4fv26Wu24urpi27ZtWLRokd6ODBHg5s2b6OjowIoVK4xjv8Fj5HI5zp07h9OnT6u0rygsLAy7du0yqhH2iooKlJWVYcmSJfD29h75ouZvgoC19chshr//pE8ZHBxEWVmZ0mizhYWF0mizh4cH3VOIVojFYlRUVLAFAzs6OmBubo6goCD2/Tdr1iy9ev8panIIhUKUlJTA3t4e8fHxSExMhKurq667ZxCGhupw/XoE5HLN3f/MzKyxeHEpbGwmf/8zRhRoPCI1FXj5Zc2197e/ieHmdgoCgWDUAw7DMOju7kZzczPMzMzg7+8PBwcHLF68GHw+H76+vprryCSJRCKcO3cOmZmZkKoxqmltbY3169djxYoVNKqix4aGhnDhwgVEREQY136DR7S3t+OTTz5BbW3tlM/lcrnYunUrVq5cqVcPFepobW1Ffn4+goODETnWelFN3wRTU4E9e1Q+nWEYtLS0sGvrq6urIZVK4ebmxj70hYaGGmVGNKIf2tra2A3l5eXlGB4ehpOTk1IKXTs7O113k9XW1gahUIgrV65AJBIhOjoafD4fYWFhRnMf05bm5lRUVmru/hcamgovL9Xvf8aCAo3HHDwIHDigmXZee23k762trcjIyMCtW7fGPFYkEuHBgwcwNzfHzJkzYWFhAX9/f3YKVNujzHK5HLm5ufjqq69U2hyrYGZmhoSEBDz11FO0OU3PVVZWory8HMnJyUa130CBYRhcuXIFX375JcQqFIqbNWsW9uzZg1mzZmmhd9NvYGAA2dnZcHZ2Rmxs7MRLNbVxE9QQsVisVLBNsRQ1MDCQXeaib6PNxHhIpVJUV1ezsx3Nzc3gcDjw9/dnAw9fX1+9WAotFotx7do1CAQC3L9/HzNnzgSPx8OyZcsoMJ9AQ8NB1NWpf//z9z8IX1/N3v8MFQUaY0hNBV59dWSp8lQG9rnckY/33x97EK+yshJpaWlobGwc83yZTIbe3l7Y2trCwsICZmZmcHBwQHx8PJKSkrSydKOkpATp6elobm5Wq52oqCjs2LHDaNf3GwupVIqLFy/C3d0dCxcu1HV3tKK/vx+fffbZuIH9RDgcDpKTk7F582ajWEYmlUqRm5uL4eFhJCYmTv4BQ1s3QQ1ra2tjU5hWVFRALBbD0dGRTV8aERGhV6PNxLh0d3ez77+ysjIMDg7Czs5OKamBrjP3MQyD6upqCAQCFBUVUU2OSWhuTkV19auQy6WY2p4NLszMuAgOfp9mMh5BgcY46uqAvXuBzMyR35sT/a5VvJ6cDBw+PPFyZMUGzBMnTqCnp2fcY7hcLlxdXdHe3g6GYcDhcDB//nzweDyEhoaqPWLX3NyM9PR0lJSUqNXOrFmzkJKSgvBw082oYCiam5uRn5+PlStXwtnZWdfd0Yri4mIcPXoUvb29Uz7XxcUFL774IkJDQ7XQs+nFMAxu3bqFe/fuYfny5aqt09bWTVBLHh1tLikpwf3798HhcODn58c+9Pn5+enFaDMxPnK5HHV1dez7r6GhAQzDYPbs2ez7LzAwUKcDGD09PWxNjt7eXoSGhoLP51NNjjEMDdWhsnIvurszAXAxccAx8rqLSzJCQg6b/J6Mx1Gg8QSlpcChQ8DZsyP1rB791+JwgMBAYN06YN8+peyNTzQ8PIzMzEycO3duworEs2bNwuzZs1FRUcEGJl5eXuDxeIiNjZ3yFGhvby+++uor5ObmQp3/ekdHR2zevBnLly+nG5SeU6StBYCEhASjXFYyPDyM48ePQygUqnT+4sWL8Z3vfAe2traa7ZgO1NfXo6ioCAsWLNBMuktt3QS1rKenR6lg2+DgIGxtbZUKBhprwE10r7+/H2VlZWwK597eXlhZWSEsLIwNPHSV6l0qlaKoqAgCgYCtyZGYmIiEhARa9vyYgYFSNDcfQmfn2W+L+j363MSBtXUgXF3Xwdt7n0mnsJ0IBRpT0N8PVFcDYjFgZTVSTFfdJFE9PT346quvcOXKlQkf/GNiYhAYGIiioiJUVlYCGNl0rZgCnTlz5oTXkUgkyMrKwtmzZ1Vas65gYWGB5ORkrFmzhtZ5GoCenh5cvHgRy5Yt+3d2ISPT2NiIjz/+GA8ePJjyuTY2NvjOd76DJUuWaKFn06urqwu5ubmYPXs2FixYoJ2AUhs3wWkgl8tRX1/PjjbX19crjTZHREQgKCjIKJbLEf3DMAzu3bvHvv+qq6shl8vh4eHBpnAOCQmBpaXltPetqakJAoGArcmxaNEitiaHMQ5KqUMq7cfQUDUYRgwOxwo2NkHgcvX//qdrFGjoiaamJqSnp6O8vHzcYywsLLBq1SrMnTsXBQUFKCgoYGdDwsLCwOfzER0drTTDwDAMrl+/jhMnTqCrq0utPsbGxmLLli1GlebTmBUVFaGtrQ0rV640ygcouVyO8+fP46uvvlIpbW1ISAhefPFFzJgxQwu9mz4ikQg5OTmwtLREfHy8Uf5fa9rAwAA721FSUsKONitS6EZFRVFhUaI1IpFIqWBgZ2cnuFwumw0uMjISXl5e0/qgPzAwgCtXrkAoFKKjowO+vr7g8XhYvHgxZY8kaqFAQ48wDIPi4mKkp6dPODrr4OCAzZs3Y8GCBbh69SqEQiHa2toAADNmzGBrcrS2tiItLQ319fVq9Ss4OBgpKSk6SblLpk4kEuHChQsICwtDSEiIrrujFR0dHfj0009RXV095XPNzc2xZcsWJCcnG/SInVwuR0FBAXp6epCYmKgXNXgMkWK0WVGp/PHR5sjISISEhOhlwTZi+BiGQVtbG5vCubKyEhKJBC4uLuz7LywsbNqWdcrlcpSUlEAoFKK4uBh2dnZsQhqqyUFUQYGGHpLJZLh8+TK++uorDAwMjHuct7c3duzYgYiICJSWlkIoFOLu3bsYHBxkM1t5e3urvObS3d0d27dvx/z58w36gcyUKDbDJicnG8V+g8cxDIOCggJ88cUXEKlQWM7b2xt79uzB7NmztdC76aOoYBwbG/vEZZNkakQiESoqKtgHP30YbSamQyKRoKqqip3taGlpYWttKVI4+/j4TMv7r62tDdnZ2bhy5QqGhoaoJgdRCQUaemxwcBBnz57FpUuXJiygFxkZiR07dsDZ2RlffPEFvvzyS7S0tEAikQAYmQHx9vaGu7v7pDZu29raYuPGjUhKSqJlGAZCkbbW1dUVMTExuu6OVgwMDOAf//gHbt68qdL5K1euxNatWw16GUBzczOuXbuGsLAwhIWF6bo7Rk8x2qx46KuoqIBEIoGzszO7xGo6R5uJ6ens7FRKoSsSiWBvb8/uLYqIiICjo6NW+6CoySEUCnHv3j14enqyNTmMsQ4T0SwKNAxAR0cHTpw4gcLCwjFfZxgGDx48gEQiwcyZM2FpaQm5XI62tjY0Nzejv78fwMgeDy8vL8ycOXPMjdxmZmbg8/nYsGED5Z43IC0tLbhy5QpWrFhhtPtnSktLceTIETx8+HDK5zo7O2PXrl0GnYK5t7cXly9fhqurK5YsWUKZ3nREIpGgurqane14dLT50YJtNNpLtEEmk6G2tpYNfBUrF3x8fNj3X0BAAMzNzbVyfUVNDqFQiJs3b8LCwgKxsbHg8XhGm2yEqI8CDQNSU1ODtLQ01NXVARj5oe/q6kJdXR0GBwcBAFwuF7Nnz8bs2bNhZmYGhmHQ19eH5uZmdHR0QC6Xg8PhwNXVFd7e3nBycmJrdGzbtg2enp66/BbJFDAMg7y8PMhkMiQmJhrlw41EIkFGRgYuXbqk0vmLFi3C888/b7CBs0QiQW5uLmQyGRISEmifgJ7p6upiH/oeHW1+NIWutkebienq7e1FWVkZ+x7s7++HtbU1WzAwIiJCa/sqHq/JERISAj6fj/nz59NACFFCgYaBYRgGhYWFOHLkCG7evDlu0T9ra2v4+fnB3d2dfQAdHh7GgwcP0NLSwqa49fLywq5du/D0009TuloD8vDhQ2RlZWHp0qUGv99gPE1NTfj444/R0tIy5XOtra3x3HPPYenSpQYZgDEMgxs3bqClpQUJCQlU78EAyGSyUQXbAGDOnDns2nptjjYT08YwDBobG1FSUoLi4mLU1dVBLpfDy8uLne0IDg7W+NJRqVSKW7duQSAQoLq6mq3JER8fT0E2AUCBhsHp6enBqVOnkJubi/v376OpqWnC/RsODg4IDAxU+oGXy+UYHh6Gk5MTRCIROBwOrKys2JocXl5e0/GtEBXdvn0bzc3NWLVqlUHvNxiPXC5HZmYmTp06BZlMNuXzg4OD8eKLLxpshpSamhrcvn0bMTEx8PHx0XV3iIr6+vqUCgb29fXB2tpaqWCbob5Hif4bHBxUSqHb3d0NCwsLhISEsPuLPDw8NDoQc+/ePQiFQhQUFEAulyMmJgY8Hg/+/v4GOeBDNIMCDQMhFotx4cIFXLhwQamS+PDwMBoaGvDgwYMJC/65ubkhICAAjo6OWLt2LZKTk2FpaYnm5mYIBAJcvXqVneUICwsDj8fDvHnzaApUj4hEImRmZiI4ONhoNwJ3dnbiyJEjbFHKqTA3N8emTZuwevVqg3zfdnR0IC8vD35+foiOjqZfzEbk0dHm0tJS1NTUQC6XY+bMmUopdI1x4IDoHsMwaGlpYVM4V1VVQSqVwtXVVSmFrqZWNQwODuLKlSsQCATo6OiAj48P+Hw+1eQwURRo6DmGYZCfn49Tp06Nu0wKGMnIU1dXN25RPg6HA29vb+zcuRMpKSmjsqQMDQ0hPz8fQqEQra2tAAAXFxckJSUhPj5e5RS5RDNqampw9+5dJCcnG+x+g4kwDINr167h2LFjKqWt9fLywu7duw1yBmBwcBA5OTmws7PD8uXLaWmNCRgaGmJHm4uLi0eNNkdGRsLT05OCTaIVYrEYVVVVKC4uRmlpKVpbW2FmZoagoCD2/Td79my1338Mw6CkpAQCgQAlJSWwtbVFXFwckpKSqCCmCaFAQ49VVFQgLS0NTU1Nkz5HsTn80fobzs7OCAwMZB9Q7ezs8NRTTyExMXHUQw3DMCgrK4NQKMSdO3fAMAy4XC5iYmLA5/Ph5+enke+NTI5MJsOlS5fg4uKCRYsWGeWDx+DgID7//PNxs6o9CZ/Px/bt2w1upEwmk+HKlSvo7+9HUlISpUg1UYqsgYolLpWVlexos2JTeXh4OO2hI1rT0dHBvv/Ky8shFovh6OiIiIgIREVFITw8XO2CoI/X5Jg7dy74fD7Cw8ON8vca+TcKNPRQa2srjh8/jtu3b6t0vuIXV3t7O2bPng0XF5cxf5A9PT2xY8cOzJ07d8zXOzo6kJOTg9zcXDZw8fX1BZ/PR0xMjME92Bma1tZW5ObmgsfjGe1a7vLychw5cgTd3d1TPtfR0RG7du1CZGSkFnqmPQzD4O7du6irq8Py5cvh7u6u6y4RPTI8PIzKykr2wU8x2hwYGMiurdfEaDMhY5FKpaipqWHff/fu3QOHw4Gvry872+Hv76/y8tTh4WFcu3YNAoEA9+7dg4eHB3g8HpYvX041OYwUBRp6ZGBgAKdPn0Z2djbkcrnK7djZ2WHTpk1YvHgxsrKykJmZyRbvG0toaChSUlIwZ86cMV+XSCS4fv06BAIBm7fbzs4O8fHxSEpKMtqHYF1RLJcbHh5GYmKiQe43eBKJRIKTJ08iKytLpfMXLFiA7373u2qPsk23pqYmFBYWIioqCsHBwbruDjEAHR0d7Np6xWizg4MD+9AXHh5OS1uJ1vT09CgVDBwYGICtrS2bQjcyMlKlrHgMw6C2thYCgQA3btyAhYUFli5dCh6Ph1mzZmn+GyE6Q4GGHpBKpRAIBDhz5gxbD0MVXC4XK1aswLp165SWYXR1deHkyZO4evXquOdyOBwsW7YMmzdvHvemwTAM6urqIBQKUVhYCJlMBg6Hg+joaPD5fISFhdEom5p6e3uRlZWFxYsXjxv4Gbr79+/j448/xv3796d8rpWVFZ599lksW7bMoN5ripzzM2fORExMjEH1negPqVSqVLCtqamJHW1WLHNRZ7SZkInI5XI0NDSw77+6ujowDANvb282hXNQUBC4XO6U2n348CFbk+Phw4dsTY558+bRnjUjQIGGDjEMg6KiIhw/fhwdHR1qtRUTE4OtW7dOuMGqvr4eaWlpqK6uHvcYS0tLrFmzBsnJyRMWB+vt7UVubi5ycnLYZS+enp7g8/lYtmwZrSdWwZ07d3D//n2jTVvLMAwuXryIEydOTJiSeTyBgYHYvXu3QW0iFIvFuHz5MjgcDhITE43y/5XozsOHD5VS6A4MDMDGxgbh4eGIiopCREQEXFxcdN1NYqQGBgaUCgY+fPgQlpaWCA0NZWc7PDw8Jt2eTCZDUVERhEIhqqqq4OzsjMTERCQkJFBNDgNGgYaO1NfX41//+hdqamrUasff3x8pKSkIDAyc1PEMw+DWrVs4fvw42tvbxz3O2dkZmzdvfuLIsVwux+3btyEQCFBRUQFgZNQ5NjYWPB4P3t7eU/uGTJBYLEZmZiYCAwMRHh6u6+5oRXd3N44cOYLy8vIpn2tmZoannnoKa9euNZiRWkUWrfb2diQmJtIvSaJ1crkcjY2NKC4uHjXarHjoCwoKomCXaAXDMLh//z4bdFRXV0Mmk8Hd3Z19/4WGhk44gPkoRU2Oq1evQiaTYeHCheDz+QgICKAZYQNDgcY06+rqwokTJ3Dt2jW12nF1dcXWrVtVXoYhlUohFArxzTffTLhca86cOUhJSUFoaOgT22xubmaL9ShqcoSGhrJToIbykDid6urqcPv2baNNWwsA169fx7Fjx1RaFujp6Yk9e/bA19dXCz3TjsrKSpSWlmLJkiUUaBOdUYw2K2Y8enp6Ro02u7u700Mb0QqxWIyKigo2hXNHRwfMzc0RHBzMvv+8vb2f+P4bHBxkU++3tbVhzpw54PP5WLJkCQXNBoICjWkiEolw7tw5ZGVlTbgx+0msra2xfv16rFixQiM/ZAMDA/j6668hFAon3IA+b948bN++HZ6enk9sc2hoCAUFBRAIBEo1ORRToLRxcWT08dKlS3B0dMTixYuN8pf94OAgvvjiiwn3Bk0kKSkJ27dvn/QImK61trYiPz+f/UVKiL5gGAbNzc3saHNVVRVkMhnc3NzYtfVTGW0mZCoYhkF7e7tSCl2JRAJnZ2dERkYiIiIC4eHhEw62MQyD0tJSCAQCFBcXw8bGhk1IY0jLaU0RBRpT0N8PVFcDYjFgZQUEBQFPSnojl8uRl5eHU6dOoa+vT+VrczgcJCQkYNOmTVp5UJ9MSl0zMzPweDxs3LhxUqPvDMOgvLwcAoFAqSbHokWLwOPx4O/vb5QP2E/S1taGy5cvG/UNsrKyEp988olKaWsdHBzwwgsvYO7cuVromeYNDAwgOzsbTk5OWLZsmXHP3KlyEyR6Z7zR5kcLts2aNcsk789E+yQSCaqrq9nAo7m5GRwOB/7+/mwKZx8fn3HvpR0dHcjOzkZubi6GhoYQFRUFPp+PiIgIrb5npdJ+DA1Vg2HE4HCsYGMTBC6X7n9PQoHGE5SWAocOAWfOALW1wKP/WhwOEBAArF8PvPIKEBGhfG5JSQnS09PR3NysVh+ioqKwffv2aVmGUVlZiX/9618TFgm0tbXF+vXrwefzJ51dorOzEzk5Obh8+bJSTQ4ej4fFixebxBQowzAoKCiASCRCUlKSUT6QSqVSfPXVV7hw4QJUubXMmzcPO3fuNIhZL6lUiry8PIjFYiQmJhpvAgR1boLEILS1tbEbysvLyzE8PAwnJyelFLrGurST6F53dzebwrmsrAxDQ0Ows7NjC1ZGRkaOuc9teHiYTb3f1NTE1uRYtmyZxgqgDgyUorn5EDo7z0AkqgXw6O81DqytA+Dquh7e3q/Azo7uf2OhQGMcdXXA3r1AZibA5QITJclRvJ6cDBw+DFhZNSM9PR0lJSVq9cHb2xspKSmImOZf3ooH4pMnT6Knp2fc49zc3LB9+3YsWLBg0qMIEokEhYWFEAgEaGhoAGAaNTn6+vqQlZWFhQsXGtR+g6lobm7Gxx9/jHv37k35XCsrKzz99NOIi4vT+1FURUKFpqYmxMXFGe17Vq2boL//9PWTaJRUKmVHm4uLi0eNNkdGRsLX19coB0qI7snlctTV1aG4uBilpaVoaGgAwzCYM2cO+/4LCAhQGuR8vCYHl8tla3LMnj1bpX4MDdWhsnIvurszAXABTJQpceR1F5dkhIQcho0N3f8eRYHGGFJTgVdfHfm9OZUsnObmDMzMZFi27J8IC8tV+fqOjo7YtGkT4uLidHozV2RDOn/+PIaHh8c9LigoCCkpKfDz85tS+/X19RAIBCgsLIRUKmVrcvB4PISHh+v9A+dkFRcXo7GxEatWrYKlpaWuu6NxDMPg0qVLyMjIUCltbUBAAF588cUppUHUlfr6ehQVFWH+/PnwN+aHaVVvglzuyMd77wEvvaS9/pFpoxhtVhRsGxwchJ2dnVLBNicnJ113kxipvr4+pRS6fX19sLKyQlhYGJvC+dElyL29vWxNjp6eHgQHB4PH42HBggWTrsnR3JyK6upXIZdLMXGA8TguzMy4CAp6D97edP9ToEDjMQcPAgcOqNMCA4CDmJiTWLjw7JTOtLCwQHJyMtasWaNXyzB6enpw6tQp5OfnT7gcZunSpdiyZQtmzJgxpfb7+vqQm5uL7OxspZociilQGxsbtfqvK8PDw8jMzISfn5/Rbg7u6enBkSNHUFZWNuVzzczMsGHDBqxfv17vR0e7u7uRm5uLWbNmTWkGzyCpfxMc8eabwOuvq98O0RuK0WbFQ59itHn27Nls0BEYGDjlgm2ETAbDMGhqamLffzU1NZDL5fD09GTffyEhIbC0tIRMJsOtW7cgFApRWVkJZ2dnJCQkICEhYcLAuKHhIOrq1L//+fu/CV9fuv8BFGgoSU0FXn5Zc+0lJv4fwsLyJnWsqg/p06mpqQnp6ekT1kKwsLDAqlWrsHbt2ikHS4qaHEKhkL2GlZUVli5dCj6fb1CpQhsaGnDz5k0kJyfD3kg3y968eROfffaZSmlrPTw8sHv3br2fFRCJRMjJyYGlpSXi4+ON/wFK0zfB1FRgzx7NtUf0Sn9/P8rKythlLr29vexos+LBz1gTXhDdGxoaQnl5Obu/o6urC1wuFyEhIez7b+bMmUqp96VSKRYtWjRmTY7m5lRUVmru/hcamgovL7r/UaDxrbq6kX2MIpGmWmRgbi5BSsobcHTsHPeo4OBg7NixY8rLjnSFYRjcvXsX6enpbOrasTg4OGDz5s0qL/9qaWmBUChEfn4+W5MjJCSErckx2SnQ6SaXyyEUCmFra4ulS5ca5ci3SCTCP//5TxQUFKh0fkJCAlJSUvQ6laZcLkdBQQF6enqQmJhotMGiEs3fBAFr65HN5HoeUBL1MQyDe/fuKRVsk8vl8PDwYDMJKUabCdE0hmHQ2trKvv8qKyshkUjg4uLCpnD28fFhZzkUNTl4PB6WLFkCmew+rl+PgFyuufufmZk1Fi8uNfk9GxRofGv1akAgmNpy5CfhcGTw9i7Hhg3vjnrN3d0d27dvx/z58w3yYVQmkyEnJwenT59ms0iNRd0N7SKRiC3W8+DBAwAjVcuTkpIQHx+vVxWXFSn3EhMT4e7uruvuaEVVVRU+/fRTdHaOHzyPx8HBAd/73vcQHR2thZ5pTmlpKSorKxEbG4uZM2fqujvTRxs3QS4X4POBCxc01yYxCCKRCOXl5eyDX2dnJ7hcrlLBNi8vL4P8/feo3//+98jIyEB5eTlsbGywfPly/OEPf5hUkVuiPRKJBJWVlez778GDBzAzM0NAQAAiIiJgaWmJiooKtibHunWXYGlZAkCmwV5w4eLCx7x5pn3/o0ADIwNu2lxCn5LyW7i4jDwk29raYsOGDeDxeEaxDGNwcBBnz57FpUuXJtwIHBkZiR07dqi8/IlhGFRUVEAgEOD27dtgGAbm5ubsFKgua3IwDINr166hv78ffD5f7/cbqEIqleLrr7/GuXPnVEpbGx0djZ07d+pVYPi4lpYWXL16FWFhYQgLC9N1d6aXtm+CpaVAeLj22id6jWEYtLW1obi4eNRosyLoCAsL01hK0um0du1aPPvss1i8eDGkUilee+01dikZpQTWH52dnWwK57KyMohEIjg4OGDOnDlgmDrMn/9XrV178eJS2NmZ7v2PAg0A+/cDH3yg2YE8BQ5HhoiIbCQkpE2p2J2h6ejoQEZGBm7cuDHuMYqig0899ZRaD5xdXV1sTY7+/n4AgI+PD/h8/rTX5Ojv70dWVhbmz59vMMvfpqqlpQWffPIJGhsbp3yupaUlUlJSkJCQoLcjl319fcjJyYGrqyuWLFlilIHiE2nzJsjlAvv2Ae+Ontklpkkx2qzIZtXS0gIzMzP4+/srLXPR13vGRNrb2+Hh4cHObhP9I5PJUFtby6Zwdnf/AsHBpTAz08bjMBezZu1DcLDp3v8o0MBIcduaGu217+7eg7t3xfD09NTeRfRETU0N/vWvf6G+vn7cY6ytrbFu3TqsXLlSraBAIpHgxo0bEAgE7PXs7OwQFxc3LVW3S0pKUFdXh+TkZL3eb6AqhmGQnZ2N9PR0SCSSKZ/v5+eH3bt36+37XiKRIDc3FzKZDAkJCUb5fzhp2r4JBgUBVVXaa58YtM7OTqUUuiKRCPb29oiMjERERAQiIiL0ejb0UdXV1QgODsbdu3cRFRWl6+6QSbhyJQDDw3Vaa9/aOgixsaZ7/zP5QKOvD3ByUi52q2kcDtDbC5jCflJg5AG1sLAQGRkZ6OrqGvc4V1dXbN26FTExMWqPXI1Vk2Pu3Lng8XiIiIjQ6MiYRCJBZmYm5syZg7lz52qsXX3y8OFDHD16VKWikxwOh01bq4+b9hmGwY0bN9DS0oL4+Hi4uLjouku6RTdBokceHW0uKSlhZ1J9fHyUCrbp471FLpdj06ZN6OnpQW6u6rW0yPSRSvuQm+sE5YrfmsZBfHwvuFzTvP+ZfKBx6xawYIH2r1NUBMyfr/3r6BOJRIKLFy/i7NmzEE2Qycbf3x8pKSkIDAxU+5p9fX3Iy8tDdnY2G+R4eHiwNTnUXQPc2NiIwsJCJCcnw8HBQe3+6qOioiJ89tlnE27yH4+bmxv27NmDgIAALfRMfbW1tbhz5w4WLlwIHx8fXXdHP9BNkOix3t5epYJt/f39sLa2ZgsGRkREwNXVVdfdBADs27cPZ8+eRW5ursoVqcn06uu7hRs3tH//W7SoCA4O87V+HX1k8oHG1atAbKz2r/PMM+/Ay2vqa9yNgVgsRl1dHe7fvz/hRmJPT08EBQVppEAfwzBob2/HvXv32IDD3NwcXl5emD179pTTlSpSN3K5XMycOdMg1w4/iVQqRWVlJZqbm1U639vbGyEhIXqZ5GBoaAjNzc1wdHSEm5ubUf7//f/27jyorTPdE/9XQoDYEQJhsA0GsxnjeMU2GDCyjRPbiVfoSnfSiWN3Tzq32j2/mZqaP7r73pmq6dTUVNf0H7dT08kt7Did3m7Aa7zFxBarN7ybzRiDsY0wWhCbkITEOb8/iM5FIGEE2vV8qlzgcM7RC8GPznPe932euUrq7cV/+fd/d/rr9Jw4gajt2xEWFkY/fzInLMvi+fPn3Nr6rq4uMAyDhIQEbrYjPT3dpfv0zH75y1/izJkzqK2t9fj+QOQ/DA3dxN27zr8JXLPmBiIjNzj9dTyR3yca9DDPdeRyOSorK2dcjiMQCLBlyxbs2LHDYRVIent7UVNTg2vXrnE9OdLT0yGVSrFq1arXTsGr1WpUV1ejsLAQEonEIWPyNE+fPsWxY8egUqnsPjcsLAwffPABVnngL/jo6Chqa2sRGhqK/Px8j0yC3M5FQfB/7d+Pl7GxCAkJQVxcHCQSybSPkZGRlISQWRsdHbUooavRaBAYGMg1bMvJyYFEInHq7xTLsjhy5AhOnTqF6upqpKenO+21iOPRjIbz+X2iMTICREbS8mRXam5uRmVl5YxPzsPCwvDOO++gqKjIYWtx9Xo9bty4gerqavT29gKY6MlRWFiIoqKiaZsNzXtNBgcHIZVKPXJN8HyNj4/j3LlzuHjx4pzK1i5fvhwffvghoqKinDC6uRsfH8e1a9cwMjKCoqIin6z05jAuCIIsjwd5ayv6tFoolUooFAruo0aj4Y4LCgqaloCYPxeJRJSEEJtYlkVvby/XJfrJkycwmUwQi8UWJXSFQqFDX/ef/umf8Le//Q1nzpyx6J0RFRXlkNl54lwm0wjq6yNBezScx+8TDYAKrrgDwzBoaGjAmTNnMDw8bPO4+Ph4lJaWYsWKFQ67yTD35Kiursb9+/e5nhxr1qyBVCpFamoqRkdHUVVVhTfeeMNj9xvMV19fH44ePYru7m67zw0MDERpaSk2b97sUTd/LMuiqakJnZ2dyMvL89kZKIdzYxA0Go1QqVQWyYf5c7VazSXAAoHAIvEwf5RIJIiJifHPssTEJoPBgCdPnnA9Lfr6+sDn85GWlsYlHosWLZp3/LJ1/pdffomDBw/O69rENW7cSINe77z4R1WnKNGgEvJupNfrcenSJVRVVc3Y8C8rKwulpaVYvHixQ1/fWk8OoVCIhQsX4pNPPvHJDd8sy6K2thYVFRVzKlublJSEw4cPe1zX7JcvX6KxsRE5OTm0fMFeHhoETSYT1Go1lErltCREpVJhfHyiiy+fz0dsbKzVJEQsFtOSOQKVSsUtsWpra4PBYEBkZCSys7ORk5ODZcuW2b13j/iGJ09+hZ6ePwFwQvyjPhqUaADUFNcT9Pf34/Tp07h586bNY3g8HvLy8rBnzx5ER0c79PWNRiNu3ryJo0ePYnh4GGKxGKGhoSgoKHBJTw5XGRoawp///Gc8evTI7nN5PB7eeustvP322x514zY4OIi6ujrEx8c7pFSyX/LCIMgwDPr7+6clIOY/5iSax+MhJibG6p6Q2NhYBAUFOXRcxPOZTCY8ffqUSzxevnwJHo+H5ORkbrYjJSWFZsn8hFbbgsZG58U/6gxOiQYAYPt2QCZz7AM9gQCQSoHLlx13TV/37NkzfPPNN3g6wzKOoKAgvPnmmw5tlPfy5UvcunUL27ZtQ39/P6qrq9HY2Mj15MjJyYFUKnV4Tw5XevDgAb7++usZl6rZIhaLcejQIaSlpTlhZHMzNjaGuro6AEBhYSHdMM6XDwVBlmUxMDAwLQkxfzQXhQAm9mlNnQUxL9Fy9Hp+4pkGBgYsGgZqtVqEhoZyJXSXL1/u8IdbxLM8eLAdGo0Mjp3VEEAkkmLlSv++CaRE4wddXUB2NjBDuwe7CYUTD/Ko0p19WJbFvXv3cOLEiRmrIEVHR2Pv3r3YuHHjnG/+zd2vAwMDkZ+fb3GdkZERNDQ0oLq62qInx+bNm5Gfn++wqljOZjAYUFFRwd2U2ysvLw/vvvuux9x0sSyLW7duQalUWt3ET+bIT4Igy7IYHh62moQoFArodDru2MjISIt9IZMTEm/590/swzAMuru7udmOrq4usCyLxMRE5OTkYPny5UhLS/OoWV0yfzpdFxobs8Ewjot/fL4QubktCAnxnPjnDpRoTFJeDvz854693uHDjruevzGZTJDJZDh//rzFm/9USUlJKCsrQ0ZGhl3X7+/vh0wmQ0FBAeLj420exzAMHj16BJlMhtbWVgATsyobNmxAcXGxRzdm6uzsxJdffgmFQmH3uWFhYXj//fexZs0aJ4xsbtrb29HS0oLc3FwsXLjQ3cPxPX4eBFmWxejoqNVZEKVSaTEbGBYWZnUWRCKRIDw83GtnPoklrVZr0TBwcHAQQUFByMzM5GY7qOiEb5DLy9He7rj4l5lZjoQE74l/zkKJxhSffgr89reOuc6vfz3/65CJmYVz586hpqYGDMPYPG7lypU4cODAjEmD2e3bt6HRaLBlyxa7yta+evUK1dXVuH79OtftPD09HcXFxVi9erXHlMBlGAbnz5/HhQsXZvyZ2bJs2TIcPHjQY5YLKBQKXLt2jasYQzdxTkRB0CadTjetPK/54+DgIHecUCi0OgsSFxeHqKgo+v31UizLoqenh0s6Ojo6MD4+jri4OC7pyMzMdNiSXuJ63d2foqtr/vEvJeVTJCf7VvybK0o0rCgvB44cmViqbM9yZYFg4s9nn3nVQzyv0dfXhxMnTuDBgwc2j+Hz+SguLsbbb79ttXfC6OgoLl++jBUrVmDp0qVzHoter8fNmzchk8m4nhxRUVEoKipCYWGhW/tKKBQKHDt2DF1dXXafKxAIcODAAUilUo+4GdJqtaitrUVkZCTy8vJoc6arUBC0m8FgmFamd3KvEPNbbWBgoM0kRCQS0e+4FzEYDHj8+DHXqVylUiEgIADp6elc4pGYmOgRsZTMnlxejo6OI2AYE+zbsyEAny9AevpnNJMxCSUaNnR1AR9/DFRVTbxvzvRea/56SQnwxRcetRzZJz1+/BgVFRV48eKFzWNCQ0Oxa9cuFBcXc2tpHz9+jPb2dpSUlDhsvwHLsnjy5AlkMhnu378PhmHA5/Oxdu1aFBcXY+nSpS57k2FZFvX19fjmm28wNjZm9/mLFy/GoUOHkJiY6ITR2cdkMqGhoQF6vR5FRUXU+ModKAg6jLlXyNSlWAqFAmq1mpt1FAgEiI2Ntdq0UCwWe8yMKZmOZVkolUqLErpGoxHR0dFYvnw5srOzsWzZMmoe6iV0ui60t38MjaYKgAAzJxwTXxeJSpCR8YXf78mYihKN12hpAT7/HLh4caKf1eSfFo8HLF0K7NgxUSaeSti6DsMwuHHjBk6fPm2xZGGquLg47N69GyqVComJiVi1apXTxqTRaLieHOa13IsWLYJUKsX69eudWhVpeHgYX3/99YyzPbbweDxs374du3fvdvsGR5Zlcf/+fbx48QKbNm2CWCx263gIKAg62fj4ONcrZOrG9Km9QsRisc0yve7+t0ssGY1GdHR0cImHXC4Hj8dDSkoKli9fjpycHCQlJdEMlofTalsgl38OtfriD039Jt8y8yAULoVYvAOJiZ/4dQnbmVCiYYeREaCjAzAYgODgiWa31N/HvQwGAy5fvozLly9bfYqv1WqhUChQUFCADz74AEuWLHH6mEwmE+7cuYPq6mp0dnYCmJhhyc/PR3FxMeLi4hz6eo8ePcJXX301p7K1MTEx+Oijj+zeSO8M3d3duHv3LlauXOmz3di9HgVBl2IYBhqNxmoSYq1XiLXO6XFxcVT62QNoNBq0tLSgqakJra2t0Ol0CAsLQ3Z2NrfMiiroeTaTaQQ6XQdY1gAeLxghIWkQCCj+vQ4lGsQnDAwM4MyZM7h+/TpYlgXLsnj16hUAYMGCBdzypQ0bNmDfvn0QiUQuGVd3dzeqq6tx69YtrifH8uXLIZVK572p2WAwoLKyErW1tXM6f8OGDfjxj3/s9mVJGo0G9fX1WLhwIVavXk3rmQmZBZZlMTg4aHVPiEKhmNYrZOpSLPMeEU8pW+1PGIZBV1cXmpqa0NLSgu7ubrAsi8WLF3NJR2pqKs1SEZ9AiQbxKS9evMBXX32F2tpaxMfHW611HxgYiJKSErz55psue5M19+SoqamBWq0GMLGsq7i4eE49OZ49e4Zjx46hr6/P7rGEhobivffew7p16+w+15H0ej1qa2sRGBiIgoICBAYGunU8hPgKlmUxMjJidRZEoVBgdHSUOzYiImLaLIj5c9pP4BrDw8MWJXSHh4cRHByMrKws5OTkIDs7G7Gxse4eJiFzQokG8Sl37tyBUqlEXFwczpw5M+ONeGRkJHbv3o1Nmza5bJ0swzBoamqCTCZDS0sLgInEZ8OGDZBKpa/tycEwDC5duoRvv/12TmVrs7KycPDgQZfN6Fhj3l8zMDCAwsJCREREuG0shPgjrVZrsSF9chIyeQlmaGiozSQkIiKCZh8B1NbW4ve//z3u3LmD3t5enDp1Cnv37p3z9ViWxYsXL7ik4+nTp2AYBvHx8dxsR0ZGBi2HI16DEg3iE3Q6HS5fvozs7Gykp6cDmNhkWVtbi2+//RZardbmuQsXLkRpaSmys7NdNVwAE+V6a2pquOpKAJCWlgapVIpVq1ZNmzZXKpU4duwYt+/DHgKBAPv27cPWrVvdenPQ2tqKx48fY8OGDUhISHDbOAgh1un1+mkJiPnvAwMD3HHBwcFWE5C4uDhER0f7TRJy8eJFNDQ0YO3atdi/f/+8E42pdDod2trauP0d/f39EAgEyMjI4BKPycuDCfE0lGgQr9fe3o7W1lZs377d6n6D0dFRXLhwAVevXuUquFiTk5OD0tJSl98AGwwGrieHXC4HMDHbMrknx7Vr1/Dv//7vFuuuZ2vhwoU4fPiwWztp9/b24ubNm8jMzMQyqkxEiFcaGxvjeoVM3RvS398/rVeItX4hvtwrhMfjOTzRmIxlWfT19XGzHe3t7TAajRCJRMjJycHy5cuRlZXl9n13hExGiQbxWiaTCVeuXEFcXBzWrFnz2uOVSiVOnjyJu3fv2jyGz+ejoKAAu3fvdvmSHms9OcbHx6HX62EymRAZGWn3U6uSkhLs2bPHbfsfhoeHUVdXh5iYGKxfv95nbzAI8Xcmk2larxDzR5VKxS31DAgIQGxsrNUkxNt7hTg70ZjKaDSivb2dSzxevXoFPp+P1NRUbrYjKSmJZjuIW1GiQbySXC7H9evXsXXrVkRHR9t1bkdHByoqKvDs2TObxwiFQuzYsQNbt251y026RqPBX/7yF/z5z3/GyMgIACA8PBwJCQmQSCSvfTMWiUQ4ePAgsrKyXDHcaYxGI+rr62EymVBUVITg4GC3jIMQ4n7j4+Po7++fthTL/Lnph2aQfD4fMTEx05ZimXuFeHrBCFcnGlOp1Wo0NzejpaUFra2t0Ov1iIiI4EroZmdn05444nKUaBCvYu5+zbIsCgsL5/ykhmVZNDY24uTJk9BoNDaPE4vF2LdvH9atW+eyp0JjY2M4efIkZDIZGIaBSqWCXC7H0NAQgIn9FgsWLEBCQoLVKfLc3Fz85Cc/sbuSlSOwLMttiiwoKHDrpnNCiOdjGAYDAwM2y/RO7hUSHR09bRbEnIx4wsMMdycak42Pj6OzsxPNzc1oamrCixcvAADJyckWJXRplpk4GyUaxGsMDAzgypUryMvLQ2JiokOuaTQa8f333+PixYsz7n9ITU1FWVmZ0xvJPX/+HEePHuV6gEw2MjICuVwOhULBLUOIiYlBYmIiRCIRQkND8ZOf/ATr16936hht6ezsxIMHD7BmzRokJye7ZQyEEN/BsiyGhoZsJiHmIhoAEBUVZbVZoUQicdmeBU9KNKYaGhpCS0sLt8xKq9UiJCQEWVlZXOIRExPj7mESH0SJBvEK9+7dQ19fH7Zt2+aUJkZDQ0M4e/YsN1tiy7p167Bv3z6H1zRnGAbfffcdzp49+9qytUajEX19fZDL5dwb7cKFC/HJJ5/grbfecnnte5VKhYaGBiQnJ2PlypW0HpgQ4nTmXiHW9oQoFAqLSoPh4eFWZ0EkEgnCwsIcFrM8OdGYjGEYPH/+nEs6Ojs7wbIsEhISuKQjPT3d45eqEe9AiQbxaHq9HpcvX0ZmZiYyMzOd/npyuRyVlZVobm62eYxAIMCWLVuwY8cOhyxPUqvVOHbsGDo6Ouw6z9wZWCwWc13HAwMDsX79ekilUixevHjeY5uJTqdDbW0tQkJCkJ+fT11sCSEeY3R01OosiFKp5JahAkBISIjVWRCJRDKrXiEjIyNc7F69ejX+8Ic/QCqVIiYmBklJSU79Hh1ldHQUbW1taGpqQnNzMwYGBhAYGIjMzEwu8ZBIJPQQicwJJRrEY3V0dKC5uRklJSUu32/Q3NyMyspKrtysNeHh4XjnnXdQWFg4p0opLMvi5s2b+Pvf/26xBGC2EhMTcejQISxevJjryXHt2jXodDoAwNKlSyGVSrF69WqHJgEMw+DatWsYHh5GUVERdQ8mhHgVc68Qa00LJ+/ZCw4Otpj9mJyEmHuFVFdXQyqVTnuNDz/8EMePH3fhd+UYLMuit7eXm+148uQJTCYTYmNjkZ2djZycHGRmZkIoFLp7qMRLUKJBPI7JZMLVq1cRExODdevWuW0cDMOgvr4eZ8+eteiWO9WCBQtw4MABrFixYtZPfLRaLf7617/izp07cxrb1q1bsW/fvmlT2+aeHNXV1ejp6QFg2ZPD3gpdk7Esi6amJnR2diIvLw8SiWTO1yKEEE9kNBqnVcYyf1Sr1dzSWoFAYHNPSExMjM9ssjYYDBYldBUKBfh8PtLS0pCTk4Ps7GwsWrSIZjuITZRoEI/y6tUrNDQ0YMuWLR5TsUiv1+PSpUuoqqriyjBak5WVhbKyMixatGjG67W2tuL48eMWXXZnKzo6GgcPHnxt0zuWZdHR0QGZTIZ79+6BYRjw+XysXr0aUqkUaWlpdr0xvHz5Eo2NjcjJyeE6rxNCiD8xmUxQq9XTZkEUCoVFrxA+n4/Y2FirndPFYrFXLzNVKpVc0vH48WMYDAZERkZyS6yys7NplptYoESDeASWZdHQ0IDx8XEUFRV55NMRtVqN06dP49atWzaP4fF4yM/Px549exAVFWXxNaPRiFOnTuHKlStzev01a9bg/ffftzuIDwwMoK6uDrW1tdza5IULF0IqlWL9+vUzloUcHBxEXV0dJBIJcnNzPfL/CyGEuBvDMOjv77eahEzuFcLj8SAWi612To+Li/OqDdgmkwkdHR1cNauXL1+Cx+MhOTmZ61S+ZMkSn5ndIXNDiQZxu6GhIXz//fdYv379a2cDPEFXVxcqKirw9OlTm8cEBwfjzTffRElJCYKCgvDy5UscPXp0xj0ftgiFQvz4xz/Ghg0b5nWjbzKZcO/ePchkMm7s5o3cxcXFFkuhxsbGUFdXBwAoLCxEUFDQnF+XEEL8GcuyXK+QqQmIUqm0KK0uEomsJiBxcXEevy9iYGCASzpaWlowOjqK0NBQi4aB81m+S7wTJRrErR48eAC5XI5t27Z51ZMclmVx7949nDhxAiqVyuZxUVFRWLhwIdra2l5bttaatLQ0HDp0CGKxeD7DnebFixeQyWS4desW1xBr+fLl2Lx5M0ZHR6FWq1FYWDhtVoYQQojjsCyL4eFhq0mIQqHginsAE/vtpiYh5o/uaNA6E4Zh8OzZM26Z1bNnz8CyLBYuXMgts0pLS/PqZWRkdijRsMPICNDRARgMQHAwkJYGhIe7e1TeSa/Xo6qqCunp6cjKynL3cObMvHH9woULFm8IwMQmusePH2NgYADh4eFITU2d9dMcPp+PPXv2YPv27U6ddtZqtbh27Rqqq6vx9OlT9Pf3IzMzE++88w42bdpEa22JJQqChLgMy7IYHR21mYSMjIxwx4aFhdlMQsLDw92+7FWr1aK1tZXrVD40NITg4GCLErpxcXFuHePrmEwj0Ok6wLIG8HjBCAlJg0BA8e91KNF4jZYW4PPPgQsXgM5OYPJPi8cDUlOBnTuBX/wCyM523zi9SWdnJx4+fIiSkhKfuZEdGRnBuXPnUFNTA4ZhoFAo0NHRMW3zeGxsLFJSUmbsVLtgwQIcPnzYZTXYFQoFGhoawLIsXr58yfUQMffkKC4u9pp68MQJKAgS4pHMvUKmLsVSKBQYHBzkjhMKhVYTkLi4OERFRbk8CWFZFj09PVzS8fTpU4yPj0MikXBJR0ZGxoz7B11Fq22BXP451OoL0Os7AUy+ZeZBKEyFWLwTiYm/QFgYxT9rKNGwoasL+PhjoKoKEAiAGYoNcV8vKQG++AJISXHdOL3J+Pg4ZDIZoqKisG7dOrc/YXGGzs5O/PM//zPu3r1r8xgej4fExEQkJSVNWy4mlUqxf/9+l+yJ0Gq1qK2tRWRkJPLy8riZE4VCgZqaGjQ0NHCzNKmpqZBKpVizZg1NdfsLCoKEeC2DwWC1TK9CobDoFRIUFGSzTK9IJHLJ+7Rer8fjx4+5xEOtVkMgECAtLQ3Lly9HTk4OEhISXHrPoNN1ob39Y2g0VQAEAGaIfz98XSQqQUbGFwgJofg3GSUaVpSXA0eOTLxvzvTeOpVAMPHnj38EfvYz543PG/X19aG+vh7FxcUO32/gKdra2nD8+HFoNBoMDAygs7PTYmp7KoFAgKSkJCQmJnJla5cvX+70cZpMJjQ0NECv16OoqMjm7IrBYMCtW7dQXV2Nly9fAphYI1xQUICioiKPKT9MnICCICE+y2g0QqVSWU1CrPUKsda0UCwWO2VZL8uyUCgUFiV0jUYjoqOjudmOZcuWOXVPilxejo6OI2AYE2ZOMKYSgM8XIC3tj0hMpPhnRonGFJ9+Cvz2t/O/zu9+B/zmN/O/jrdjWRbXr1/H2NgYioqKfLLMnclkwunTp1FVVWXx31mWRV9fH549e4axsTGb56ekpOA3v/kN8vPznfrEhmVZPHjwAM+fP0d+fj5iY2NnfV5HRweqq6tx9+5drifHqlWrIJVKkZ6e7pOzU36LgiAhfstkMqG/v9/qnhCVSoXx8XEA/9ErxNpsSGxsrMNmvo1GIzo6OtDU1ITm5mb09vaCx+MhNTWVSzySk5Md9h7U3f0purrmH/9SUn6H5GSKfwAlGhbKy4Gf/9yx1zt82HHX8zbmsrW5ublYvHixu4fjFD09PTh69CjXhdua8fFxvHz5Ei9evLCoPBUQEIC0tDRIJBLweDykp6ejrKwMycnJDh9nd3c37t69i5UrVyI1NXXO17HWkyMxMRFSqRQbNmzwiDW1ZB4oCBJCbGAYBhqNZtosiHmJlrmCIY/HQ0xMzLQkRCKRIDY2dl5LgzUaDTfb0dLSAr1ej7CwMGRnZ3OdyiMjI+d0bbm8HO3tjot/mZnlSEig+EeJxg+6uib2Mer1jrumUDixj9Iflys/fPgQPT09Xle2drZYlsWVK1dw6tSpGbuFT2YwGPDs2TMoFApEREQgKyvLal30DRs2YN++fQ5ZmqTRaFBfX4/ExESsWbPGYU99zD05qqur0dHRAWBiw6G5J0d8fLxDXoe4EAVBQsgcsSyLwcFBq0mIQqGw6BUSHR1tNQmxt1fI+Pg4urq6uMSju7sbALB48WJutmPp0qUICAh47bV0ui40NmaDYRwX//h8IXJzW/x+zwYlGj/Yvh2Qyexbjvw6AgEglQKXLzvump7OYDCgqqoKqampyPbRCjQajQbHjx9HW1ub3efy+Xxs2LABKpUKT548sXlcYGAgSkpK8Oabb86pSZPBYEBtbS0EAgEKCgqcmuy9ePEC1dXVuHnzJvdEKzs7G1KpFDk5OT65XM4nURAkhDiBuVeItVkQhUKB0dFR7tiIiAirCUhcXNxrq1QODw9bNAwcHh6GUChEVlYWl3jY2iP64MF2aDQy2Lcn43UEEImkWLnSv+MfJRqYeODmzD24LS3AsmXOu76nePbsGe7fv+9TZWunun37Nv76179aBMbZio+Px+HDh5GcnAyWZfHw4UNUVlZCoVDYPCcyMhJ79uxBfn7+rG7YGYbBzZs3odFoUFhYiIiICLvHOVejo6O4du0aZDIZ18RQLBZj8+bNKCgo8NnfCZ9AQZAQ4iZardbqLIhSqcTw8DB3XGhoqNUkRCKRTOsVwrIsXrx4wc12PH36FAzDID4+Hjk5OVwJ3cDAQGi1LWhsdF78y81tQViY/8Y/SjQA/OpXwJ/+5NgHeWYCAfDJJ8C//qvjr+0pGIaBTCZDeHg41q9f75Mbg0dHR/GPf/wDN2/enNP5mzdvxoEDB6btYTCZTKitrcW5c+eg1Wptnr9w4UKUlZVh2Qw3a62trXj8+DE2bNiAhISEOY3TEViWRXNzM2QyGZqbm8GyLAIDA5Gbm4vi4mKn7EEh80RBkBDigfR6vc0kZGBggDtOKBRaVMeanIRERUVBr9ejra2NK6Gr0WgQGBiIjIwM5OTIwOd/C2DcCd+BAAsXfoL0dP+Nf5RoYKK57dOnzr3+DKtkvJpCoUBdXR02b9486ypG3qa9vR3Hjh2zqD0+WxEREfjwww+xYsWKGY8bHR3F+fPnIZPJuKoe1uTk5KC0tNQikejt7cXNmzeRmZk5YyLiDuaeHNeuXeNmgVJTU1FcXIy1a9dSTw5PQUGQEOJlDAYDVCrVtKVYSqUS/f39XJnewMDAaRvSeTweVCoVnj9/jrS0/4Hw8CGnjVMoTMPGjf4b//w+0RgeBqKiLJvdOhqPx6Krqx/hPtSpnmVZ3LlzBzqdDps2bfLJdfgmkwmXLl2CTCbDXP6ZLF++HD/60Y/sWr6kVCpx/vx5PHz40OYxfD4fGzduxKZNm/Do0SNER0dj3bp1Hv3/YGxsDHfv3kV9fT3kcjkAIDw8HBs3bkR+fj6io6PdO0B/NjyMmNRU8JwbBIGhIfhUECSEeCyTyWSRhEz+qFaruQqQQiGDvXvL4dyFGDwUFAxBIPDP+Of3icb9+8Dq1c5/nf37/xdiY186/4VcwGg04uXLl4iNjXXpHgBX0mq1ePz48YwN92wJCAjA0qVLER8fP+dlZIODg+js7LRYnzrZ6Ogo+Hw+MjMzsXjxYo9OMiYzbwqUy+VQKpVgWRY8Hg9isRiJiYmIioryyaV3nmyRSoV/PnnS+S907x6wapXzX4cQQmYwPj4+qVfIdQAOLOltw9q19xARscrpr+OJ/H7dwqSKa07FML7xo+7v78fIyAiSkpJmVTLO27AsC7lcjq6uLoueF7MVGRmJzMxMm922ZysqKgqrVq2CUqlEV1cXVxpQr9fDZDIhNDQUfD4f3d3d6Ovrw5IlSxAXF+fxN+k8Hg+RkZGIjIxEamoqent70dvbC5VKBZVKhdDQUCQmJkIikdCyKhcRzOH3fE5cFWwJIWQGAQEB3H6OoaER3L3r/NdkWf+Nf37/Tu6q/mJ8vhM2WbrQ+Pg4enp6EB4ejqSkJHcPxykMBgPa29vntBeDx+MhKSkJSUlJDrvZ5/F4kEgkEIvFePLkCbq6uhAUFDSt3K15k5tcLkdqauqcmxW5WlBQEJKTk7F48WKo1WrI5XIMDg6io6MDz549g0QiQWJiIkJDQ909VJ9mctFsWHt3NyKSkhAbG+uTvXUIId6Hx3PNTaCrXscT+f3SqZERIDKS9mjM5MWLF3j48CGKiop8dqnUgwcP8M0330Cn09l9bmxsLN577z2nVFPq7+/HzZs3sWjRIiQnJ+PSpUu4efPmjHtGVq1ahV27dtmsF+7J5HI56uvrcffuXYyNjQEAMjIyUFBQgOzsbK9ZIuZVRkYQk5Li1D0aLID//NFHMAQGgsfjITo6elplGPMTRuouTwhxFZNpBPX1kZiIUs5CezT8OtEAqOCKLQzDoLq6GqGhodiwYYPHL8uZC71ej7///e+4cePGnM4vLCxEWVmZw2+OdDodamtrERISgvz8fItlRD09PaioqEBra6vN8wUCAbZu3YodO3bMexmXO5h7ctTU1HB9Rsw9OTZt2oRwb83aPZWTgyCblobBxsZpmzLNH/WTupFHRUVZTUAkEolX/i4TQjzbjRtp0OudF/+o6hQlGlRC3gqVSoWamhoUFRUhLi7O3cNxiidPnuDLL7+EWq22+9yIiAj89Kc/xcqVKx06JoZhcP36dQwNDaGoqMhmkztzr4rKykr09vbavF54eDjeeecdFBUVeeVsAMuyaGlpgUwmQ1NTE1iWhUAgQG5uLqRSKfXkcBQ3BkGWZTEyMjIt+TB/Prm/THh4+LQkxDwzEhoa6pMPQwghzvXkya/Q0/MnOLYruBn10aBEA9QUdzKWZXHr1i2MjIxAKpV65c3p65hMJpw7dw6XLl2aU9naFStW4IMPPnDoXghz4tDR0YG8vDzEx8fP6jyGYVBfX4+zZ8/arFAFAAkJCThw4ABycnK89mbMnPzW19dzPTlSUlIglUqpJ8d8eXAQHB0dtZmEDA39R+370NBQi9mPyUlIRESE1/7eE0KcizqDOxclGj/Yvh2QyRz7QE8gAKRS4PJlx13TmbRaLaqqqrBq1SosWbLE3cNxit7eXhw7dgzPnz+3+9zAwED86Ec/QmFhoUNvWnp6etDY2Ijs7GxkZGTM6Rp6vR4XL17E999/D9MMv8TLli1DaWkpFi1aNNfhut3Y2BgaGxshk8nw4sULABMzTAUFBdi8eTNEIpGbR+ilvDAImrsGT01AlEqlRVGH4ODgaQmI+WN0dDQlIYT4uQcPtkOjkcGxsxoCiERSrFzpJTeBTkKJxg+6uoDsbGDSUuF5EwonHuSlpDjums7S0tKCzs5OlJSU+ORmTJZlUVNTg8rKShiNRrvPT05OxuHDh2c90zAbQ0NDqK2thUQiQW5urkNudtRqNU6dOoXGxkabx/B4POTn52PPnj2Iioqa92u6C8uy6OzshEwmw507d8AwDHg8HlatWoXi4mJkZmbSDaQ9fCwIGo1Gq0mIQqGw6BosEAiszoLExcUhJibGJ2d1CSGWdLouNDZmg2EcF//4fCFyc1sQEuIFN4FORInGJOXlwM8d2LelvBw4fNhx13MGo9GIqqoqLF68GCtWrHD3cJxicHAQX331FZqbm+0+l8fjYefOndi1a5fD+oaMjY2hrq4OwMRm8qCgIIdcd7LOzk5UVFSgs7PT5jHBwcF48803UVJS4pQxuNLg4CDq6upQV1eHgYEBABPLxYqLi7Fx48ZpJYGJDX4SBE0mE9RqtdUkRKVScT10AgICEBsba7Eky5yEiMViWq5HiA+Ry8vR3u64+JeZWY6EBM+Lf65GicYUn34K/Pa3jrnOr389/+s404sXL9DY2Iht27Z5Te8Fe927dw9ff/21xYbS2YqNjcWhQ4ewdOlSh4yFZVk0NjZCoVCgsLDQ6bMJLMvizp07OHXqFFQqlc3joqOjsW/fPp+oLDY+Po779+9DJpPhyQ+l3oRCIfLy8lBcXIwFCxa4eYRewJ+CoBUMw3Bdg6dWyFIqldzSRHNH+6kJiEQioV4hhHip7u5P0dU1//iXkvIpkpO9L/45AyUaVpSXA0eOTCxVtme5skAw8eezzzzyIR6HYRjU1NQgODgYeXl5Xn9zaY1er8c333yDhoaGOZ2/adMm/OhHP3LYk/AnT56gubkZ69atc/n+CJPJhKtXr+L8+fMWZUSnSk5ORmlp6Zz3iXianp4eVFdX48aNG1xPjqysLEilUrzxxhu0JGYmvh4E54hlWQwMDFhNQhQKBfd7BgAikchqEkK9QgjxbHJ5OTo6joBhTLBvz4YAfL4A6emf0UzGJJRo2NDVBXz8MVBVNfG+OdN7rfnrJSXAF1949p4MtVqN6upqFBQUOHS/gSd5+vQpjh07NuNTfFvCwsLw05/+FKtXr3bIWBQKBa5fv47U1FS3V3waHh7GuXPnUFtbyy0NsWbVqlU4cOAAJBKJC0fnPKOjo7h+/Tqqq6u5nhwxMTFcTw5fbUI5b74aBJ2EZVkMDQ1ZzH6YP+/r67NI8iMjIy0SkMlJSGhoqBu/C0IIMLFno739Y2g0VQAEmDnhmPi6SFSCjIwv/H5PxlSUaLxGSwvw+efAxYsT/awm/7R4PGDpUmDHjoky8Z5cwpZlWdy+fRuDg4OQSqUO22/gScbHx3H+/HlcuHBhTmVrly9fjg8//NAhS5pGR0dRU1ODiIgI5OXledTPu7e3FydPnsTDhw9tHsPn8yGVSrFr1y6bvTy8jbknR3V1NR49esT15Fi3bh2kUqnPVlqbN18Jgm7Esiy0Wq3VWRClUomRkRHu2LCwMKuzIBKJBGFhYT45A02Ip9JqWyCXfw61+uIPTf0m31vwIBQuhVi8A4mJn/h1CduZUKJhh5ERoKMDMBiA4OCJZrre0KDYXLb2jTfeQGpqqruH4xR9fX04duwYnj17Zve5gYGBKC0txebNm+f9Jm4ymXDt2jXodDoUFhZ69NPJtrY2VFRU4OXLlzaPCQ0Nxdtvv43Nmzf71MZXaz05lixZwvXkoPX1NnhrEPRw5l4h1vqFTO4VIhQKbSYhkZGRlIQQ4kQm0wh0ug6wrAE8XjBCQtIgEFD8ex1KNHxca2srOjo6sH37dp9cF8yyLOrq6lBRUWGxPnq2kpKScOjQISQkJMx7HA8fPkR3dzfy8/MRGxs7r+u5irkT+enTpy1uaKaSSCTYv38/Vq1a5VM3M9Z6coSHh6OwsBBFRUWIiYlx8wiJvzMYDNM2pJuTkMm9QoKCgqYlIObPRSKRT/27JYR4D0o0fJTRaMT333+PxMRErFy50t3DcYqhoSH8+c9/xqNHj+w+l8fj4a233sLbb7897yf1z58/x507d/DGG284rEKVqxkMBly+fBnffffdjH1G0tPTUVZWhuTkZBeOzvlYlkVXVxfXk2N8fBw8Hg8rV66EVCqlnhzEIxmNRqhUKqtd09Vq9bReIdb6hVCvEEKIM1Gi4YNevnyJW7du+XTZ2gcPHuDrr7/G8PCw3eeKxWJ89NFHSE9Pn9cYNBoN6uvrkZCQgLVr1/rEjahGo8Hp06dx48aNGY/buHEj9u7d65NduIeGhlBfX4+amhquJ8eCBQtQXFyMvLw86slBvILJZOLK9E5NQlQqFcbHxwFM7Mcy9wqZmoRQrxBCyHxRouFDzN2vAwMDkZ+f7xM3vlMZDAZUVFRwDe/slZeXh3fffXdeN4sGgwG1tbUQCAQoKCjwyfX83d3dqKysRHt7u81jAgMDUVJSgrfeessnl+WNj4/jwYMHkMlk3M/BXBK6uLh43svtCHEXc6+QqQmI+Y95VpPH4yEmJsbqnpDY2Fivb/TpCLW1tfj973+PO3fuoLe3F6dOncLevXvdPSxCPAYlGj6iv78fMpkMmzZt8tmmZF1dXTh27BhXotQeoaGh+OlPf4o1a9bM+fUZhsGtW7fQ39+PwsJCny+Lat53UllZOePPPDIyEnv37kVeXp7PLsGQy+WQyWS4efMmDAYDgImeHMXFxVi5cqXPft/E/5h7hVjrmq5UKrnff2Ci2efUWRDzEi1/mfm7ePEiGhoasHbtWuzfv58SDUKmoETDB9y+fRv9/f3YsmWLT05zMwyDCxcu4Pz58zP2f7Bl2bJlOHjwIKKjo+c8hra2NrS1tWH9+vVITEyc83W8kclkQk1NDc6dO8dVaLJm0aJFKC0txTIfLnGq0+lw/fp1yGQyLvkSiUTYvHkzCgoKfD75JP6NZVkMDw9bTUIUCgV0Oh13bGRkpMW+kMkJiSdX45sPHo9HiQYhU1Ci4cVGR0dRVVWFnJwcr92E/DoKhQLHjh1DV1eX3ecKBALs378fW7ZsmfMyslevXuHGjRvIyMhAdnb2nK7hK7RaLS5cuACZTMat77YmJycHpaWlPr20iGVZtLa2QiaTUU8OQjDxb2J0dNTqLIhSqbTYTxcWFmaza3p4eLjXLvulRIOQ6SjR8FKPHz9Ge3s7SkpKfHKKmmVZNDQ04JtvvrGYqp+tRYsW4fDhw3OefRgZGUFtbS2io6OxceNGWhoziUKhwMmTJ3Hv3j2bx/D5fBQVFeHtt9/2+af8KpUKtbW1qK+vh1arBQAkJydDKpVi3bp1PrmHhxB76XQ6q80KFQoFBgcHueOEQqHNJCQqKsqjkxBKNAiZjhINL2MymfD9998jPj4eq1evdvdwnGJ4eBhff/01Hjx4YPe5PB4P27dvx+7du+e0jMxkMqG+vh5jY2MoKiryySTOUZ48eYKKigp0d3fbPEYoFGLnzp3YsmWLz99wG41GrifH8+fPAUw8uS0oKMDmzZshFovdPEJCPJO5V4i1poUajYYr0xsYGDhtL4j5c5FI5PYHQpRoEDIdJRpepKenBzdu3MDWrVvntd/AkzU1NeGrr76asXmcLTExMfjoo4+QkZFh97ksy+LevXvo6elBQUGBT5ZtdQaWZXHr1i2cOnXKonnYVGKxGPv37/eZMsAzMffkqK6uxu3bt7meHG+88QakUimysrJ8/mdAiKOYe4VYmw1Rq9Xcvj2BQIDY2FirFbJiYmIQEBDg9LFSokHIdJRoeAFz92sej4eCggKfvEkZGxtDZWUlampq5nT+hg0b8O67785pk2FXVxfu37+P1atX09r6ORobG8P333+PS5cuzbjULTU1FWVlZUhNTXXh6NzHWk+O+Ph4SKVS6slByDyNj49DrVZPS0IUCsW0XiFisdhmmV5HFVGhRIOQ6SjR8HAajQZXr15Ffn6+z26u7e7uxtGjR9HX12f3uaGhofjJT36C3Nxcu89Vq9VoaGjA4sWLsWrVKp9M4FxtaGgIZ86cQUNDA2YKLevWrcP+/fv9ZjkRwzC4f/8+qqur8fjxYwATPTk2btyI4uJiv6tkRoizMQwDjUZjsSF98kb1qb1CJicgkz9/Xa+QkZERdHR0AABWr16NP/zhD5BKpYiJiUFSUpLTv09CPB0lGh7s7t27UCqV2Lp1q8+Wrb106RK+/fbbOZWtzcrKwsGDB+1e5qTX61FTUwOhUIhNmzb55M/W3V6+fInKykq0trbaPEYgEGDr1q3YsWMHQkJCXDg695LL5aiursaNGze42Z/MzExIpVLqyUGIC7Asi8HBQatd0xUKxbReIda6ppt7hVRXV0MqlU57jQ8//BDHjx934XdFiGeiRMMD6XQ6XL58GdnZ2UhPT3f3cJxCpVLh2LFjePr0qd3nCgQC7Nu3D1u3brVrFoJhGFy/fh1DQ0MoKipCWFiY3a9NZo9lWTQ3N6OyshK9vb02jwsPD8c777yDoqIiv7rJ1ul0uHHjBmQyGTebJxKJUFRU5BcNIQnxRCzLYmRkxGYSMrmXUEREhM0khN5fCJlAiYaHefLkCVpbW1FSUuKTT3lZlsX169fxj3/8Y05laxcuXIhDhw5h0aJFdp3X3NyMjo4ObNy4EfHx8Xa/Lpk7hmFQV1eHb7/91qKW/lQJCQkoLS3F8uXL/WoZG8uyaGtrg0wmw8OHD7meHGvXrkVxcTFSUlL86udBiCfTarXTEhDz3yfHt9DQUKt7QuLi4hAREUH/ponfoETDQ5hMJly5cgVxcXFYs2aNu4fjFCMjI/jLX/4yY/+FmWzbtg179+61q0xqT08PGhsbkZ2dPadqVMRxdDodLl68iCtXrsBkMtk8btmyZSgtLbU7mfQFarUaNTU103pyFBcXIzc31+dLBBPizfR6vdVZEKVSyRWDACb2Z1mbBZFIJB7fK4QQe1Gi4QF6e3tx7do1ny5b29zcjOPHj8+pbG10dDQ++ugjZGVlzfqcoaEh1NbWIi4uDuvXr6fA7UFUKhVOnTqF27dv2zyGx+Nh06ZN2LNnDyIjI104Os9gNBpx+/ZtyGQyrk8J9eQgxHuNjY1BpVJZTUL6+/steoVM3pA+OQnxhF4hhNiLEg03YlkW9fX1YFkWhYWFPnkzbDQaceLECchksjmdv27dOrz33nuzLltrNBpRW1vL/UyDg4Pn9LrE+To7O1FRUYHOzk6bxwQHB+Ott97Ctm3bXlv9xVc9e/YMMpkMt2/fhslk4npyFBcXY9myZT4ZNwjxJyaTiUtCJicgSqUSKpWKK5YSEBDA9QqZWiFLLBa7pFcIIfaiRMMOIyNARwdgMADBwUBaGhAePrdrDQwM4OrVq9i4caPPlrZ8/vw5jh07NuNGYFuEQiF+8pOfzHo2gmVZ3L59G319fSgoKPDZmSFfw7Is7ty5g5MnT0KtVts8TiQSYe/evdiwYYPf3lgPDw9zPTnMzRHj4+NRXFyMvLw81+zpcmQQJIS81vj4OPr7+6clIeZeIeZlqHw+HzExMdNmQeLi4hAbG0vLLh3AZBqBTtcBljWAxwtGSEgaBAKKf69DicZrtLQAn38OXLgAdHYCk39aPB6Qmgrs3An84hdAdvbsrnnv3j0oFAqfLlt7+fJlnD17lmuYZI+MjAwcPHhw1stDnjx5gqamJqxbtw6LFy+2+/WI+xmNRly9ehUXLlyAXq+3eVxycjLKysp8thrbbDAMgwcPHqC6uhptbW0AJmZ+NmzYAKlU6vgHF84IgoSQeTP3Cpm8IX3yHpHJvUJEIpHVPiFxcXE08z8DrbYFcvnnUKsvQK/vBDD5lpkHoTAVYvFOJCb+AmFhFP+soUTDhq4u4OOPgaoqQCAAZti7yn29pAT44gsgJcX6cXq9HlVVVcjIyEBmZqZzBu5marUaX375JZ48eWL3uQEBAdizZw9KSkpmtQ5VqVTi2rVrSElJwYoVK/z2SbcvGR4exrlz51BbWztjb5XVq1dj//79kEgkLhyd5+nt7UV1dTWuX7/OVXHLyMjgenLMaymFM4IgIcQlzL1Cpi7FMichkx/oREVF2ayQ5YvVL2dDp+tCe/vH0GiqAAgAzBD/fvi6SFSCjIwvEBJC8W8ySjSsKC8HjhyZeN+c6b11KoFg4s8f/wj87GeWX+vo6EBzczNKSkpmvd/Am7Asi5s3b+Lvf//7jE+kbUlISMDhw4dnNSMxOjqKmpoahIeHIz8/n9al+qDe3l6cOHECjx49snlMQEAAiouLsWvXLr+vWa/X63H9+nWLnhzR0dHYvHkzCgoK7N9Q74wgSAjxCOZeIdZmQZRKJVfxDpjoc2QtCZFIJAgNDfXJB3xyeTk6Oo6AYUyYOcGYSgA+X4C0tD8iMZHinxklGlN8+inw29/O/zq/+x3wm99MbPK6evUqYmJisG7duvlf2ANptVr89a9/xZ07d+Z0/tatW7Fv377XriEdHx9HQ0MDdDodCgsLfTJhI5ZaW1tRWVmJly9f2jwmNDQUb7/9NjZv3uyTSxHtwbIsHj9+DJlMhgcPHoBlWQQEBGDt2rWQSqWz68nh6CBICPEqo6OjNpOQyZUjQ0NDpy3DMich3torpLv7U3R1zT/+paT8DsnJFP8ASjQslJcDP/+54673f//vIJKTv4dUKkVMTIzjLuxBWltbcfz4cYsa4bMVHR2NgwcPYtmyZTMex7IsHj58iGfPnmHTpk2IjY2d42iJNzJ3dD99+vSM5ZElEgkOHDiAlStXeuUbnKP19/ejtrYWdXV1GBkZAQAkJSWhuLgY69evt57YOzoIlpcDhw877nqEELcy9wqx1rTQXKQCmNg3ZqtrenR0tEfGaLm8HO3tjot/mZnlSEig+EeJxg+6uib2Mc5h1Y8NLIKCGLS28pGa6nn/oObLaDTi1KlTuHLlypzOX7NmDd5///3XLnl5/vw57ty5gzfeeANLly6d02sR32AwGPDdd9/h8uXL3CZHazIyMlBaWork5GQXjs5zGY1G3LlzBzKZDM+ePQMw0ZNj06ZN2Lx5838k7o4PgoBQOLGZnPZsEOLzjEaj1SREoVBM6xViLtM7NQmJiYlxS68Qna4LjY3ZYBjHxT8+X4jc3Ba/37NBicYPtm8HZDL7liO/jkAASKXA5cuOu6YnePnyJY4ePQq5XG73uUKhED/+8Y9fW6Z0YGAAdXV1SEhIwNq1az3y6QdxD41Gg9OnT+PGjRszHrdx40bs3bsXIpHIRSPzfNZ6cqxYsQLFxcXI/i//BTwKgoQQJzCZTFCr1VaTEGu9QqxtTBeLxU5bHvvgwXZoNDLYtyfjdQQQiaRYudK/4x8lGph44LZ8uXOv/5rVQV6BYRh8//33OH369JzK1qalpeHQoUMzlq01GAyoq6tDQEAACgoKqPY3sam7uxsVFRUzVjgLDAzE9u3b8eabb1IJx0mGh4fR0NCA6upqaDQaJGg0+J8VFc57QV8JgoQQh2MYhusVYq1poblXCI/Hg1gstto1fT69QrTaFjQ2Ou8mMDe3BWFh/hv/KNEA8KtfAX/6k2Mf5JkJBMAnnwD/+q+Ov7Yr9ff348svv0R7e7vd5/L5fOzZswfbt2+3OSXKMAxu3boFtVqNoqIiREREzHfIxA+wLIsHDx7gxIkTUCgUNo+LjIzE3r17kZeX55ZpeU/FMAwePnwI9sgRvNHQgABnvB34ShAkhLgcy7IYGBiwmoQoFAqMjY0BmEhCoqOjrSYhr+sV8uTJr9DT8yc4djbDTICFCz9Berr/xj9KNDDR3PbpU+defw5tJTzGrVu38Le//Q06nc7ucxcsWIDDhw8jKSnJ5jFtbW1oa2vD+vXrfbZLOnEuk8mEmpoanDt3DqOjozaPW7RoEUpLS19bgMDvUBAkhHgZlmUxNDQ0rU+IUqlEX1+fRan9yMhIiwRkchLy8OEb0OudF/+EwjRs3Oi/8c/vE43hYSAqyrLZraPxeMDQEBDuZZ3qR0dH8be//Q2NjY1zOl8qlWL//v0ICgqy+vW+vj5cv34d6enpWO7MtWvEb2i1Wpw/fx4ymWzGhn8rVqzAgQMHkJCQ4MLReSgKgoQQH8OyLLRardVZEKVSyVXiEwjGUFp6HM7dBspDQcEQBAL/jH9+n2jcvw+sXu3817l3D1i1yvmv4yhtbW04fvy4Rbm62YqMjMTBgwdtJg8jIyOora1FdHQ0Nm7cSEtZiMMpFAqcPHkS9+7ds3kMn89HUVER3n77bf9eqkdBkBDiZ8y9Ql69qodO977TX2/t2nuIiFjl9NfxRP7d3QqAweBbrzNfJpMJp0+fRlVV1ZzOX716Nd5//32EW3lyaTKZUF9fj7GxMWzZsgVCoXC+wyXEKolEgl/84hdob29HRUUFnj9/Pu0YhmFQXV2NGzduYNeuXZBKpf5ZfICCICHEz4SGhiI5ORki0Svcvev812NZ/41/fp9ouKoQjTcUvOnp6cHRo0fR09Nj97nBwcF49913kZeXN60ULcuyuHfvHnp6erBp0yafbV5IPE9GRgZ+/etf4+bNmzh16pTVxpJ6vR4nTpxAdXU19u/f73/llF0UnO63tiIsJgZxcXGIioryr58xIcQj8XiuiX+ueh1P5PdLp0ZGgMhI/16ezLIsrly5glOnTnFl5OyxdOlSHDp0yGrH7mfPnuHevXtYvXo1lixZ4oDREjI3Y2Nj+P7773Hp0iUYZni6npqairKyMqSmprpwdG7kgiDIAvjPH30Eww8zRkFBQVbr5EskEohEIkpCCCEuYTKNoL4+EhNRylloj4ZfJxqAfxdc0Wg0OH78ONra2uw+l8/n45133sFbb701bZ+FWq1GQ0MDFi9ejFWrVtGNA/EYg4ODOHPmDK5du4aZwl9ubi727ds3Y98Xn+GCIGhsaYFKpZpWIUahUECtVnP/LwQCAVcVZmoSIhaLaU8XIcShbtxIo6pTTuT3S6cAYOdO5/bR2LHD8dd1hNu3b+Ovf/3rjOVAbYmPj8ehQ4emzVLo9XrU1NQgODgYO3fudFoXT0LmKioqCh988AG2bNmCyspKtLa2Wj2usbER9+7dw9atW7Fjxw6EhIS4eKQu5IIgGBgYiISEBKuVvkwmE9ewa3IC0tTUBJVKxTUI5fP5VrsGm5MQijeEEHuJxTud2kdDLPbQm0AXoRkN+F9ncJ1Oh7///e+4efPmnM4vKipCaWmpRQMchmFw/fp1DA0NoaioCGFhYY4aLiFOw7IsmpubUVFRgVevXtk8LiIiArt370ZBQYFvPlH34CDIMAw0Gs20JMQ8K2I0GgFMNOyK+WEPyNQkJDY21maZbUKIf6PO4M5FicYPtm8HZDLHPtATCACpFLh82XHXnK/29nZ8+eWX6O/vt/vciIgIfPjhh1ixYoXFf29ubkZHRwc2btyI+Ph4Rw2VEJcZHx9HXV0dvv32W66+ujUJCQkoLS3F8uXLfW85oBcGQXPXYGt18hUKhcVenOjo6GnNusx/pwp4hPi3Bw+2Q6ORwbGzGgKIRFKsXOlBN4FuQInGD7q6gOxsYFIjyXkTCice5KWkOO6ac2UymXD27Flcvnx5xnXptqxcuRI//elPLfoNyOVy3Lp1C8uWLUNmZqYjh0uIW+h0Oly8eBFXrlyZsTDCsmXLUFZWhoULF7pwdE7mY0GQZVkMDw9PmwUxf5y8ZDQiImLaLIg5IaHZWUJ8n07XhcbGbDCM4+Ifny9Ebm4LQkI84CbQjSjRmKS8HPj5zx17vcOHHXe9uZLL5Th27BhevHhh97lBQUH40Y9+hIKCAu4J7tDQEGpraxEXF4fc3FzfXEpC/JpKpcKpU6dw+/Ztm8fweDwUFBRg9+7diIyMdOHonMhXg6AVWq3WZhIyPDzMHRcaGmo1CZFIJAgPD/e9mS1C/JRcXo72dsfFv8zMciQkeGb8cyVKNKb49FPgt791zHV+/ev5X2c+WJaFTCbDiRMn5lS2NiUlBYcOHYJEIgEAGI1G1NXVgWEYFBYWWuzRIMQXdXZ2oqKiAp2dnTaPCQ4OxltvvYWSkhLfaPjnS0FwjnQ6nUXiMfnzyb1YhEKhzTK91CuEEO/T3f0purrmH/9SUj5FcrJ3xj9Ho0TDivJy4MiRiaXK9tyfCwQTfz77zP0P8QYGBnD8+HGbFXVmwufzsWvXLuzcuRN8Ph8sy+L27dvo6+tDQUEBoqOjHT9gQjwUy7K4c+cOTp48CbVabfM4kUiEffv2Yf369d5/g+kLQdBJDAaDRZneyR81Gg23NDUwMNDqLEhcXBxEIhHNBBPioeTycnR0HAHDmGDfng0B+HwB0tM/o5mMSSjRsKGrC/j4Y6CqauJ9c6b3WvPXS0qAL75w/56Mu3fv4i9/+Qu0Wq3d50okEhw6dAgpP3wTHR0dePToEdatW4fFixc7eqiEeA2j0YirV6/iwoUL0M+wjyE5ORllZWVIT0934eicwJuDoJsYjUao1WqrSYharQbDMAAmeoWIxWKrSYhYLEZAQICbvxNC/JtO14X29o+h0VRhohPETAnHxNdFohJkZHzh93sypqJE4zVaWoDPPwcuXpzoZzX5p8XjAUuXTvTJ+OQT95ew1ev1+Mc//oHr16/P6fzCwkKUlZUhODgYSqUS165dQ0pKClasWOH9T2gJcZDh4WF8++233DJCW1avXo39+/dzSw+9ljcFQQ82Pj4OtVptdUmWSqXilrfy+XwuCZnatDA2NtY3lucR4iW02hbI5Z9Drb74Q1O/ybfMPAiFSyEW70Bi4id+XcJ2JpRo2GFkBOjoAAwGIDh4opluuId0lO/o6MCxY8dmXNphS3h4OD744AOsXLkSo6OjqK2tRVhYGPLz8+nJGiE29Pb2orKyEk1NTTaPCQgIgFQqxa5duxAaGurC0TmJJwdBL2buFTJ1FsT8+eReISKRyOqekNjYWNo3R4gTmUwj0Ok6wLIG8HjBCAlJg0BA8e91KNHwciaTCefOncOlS5fmVLY2JycHH374IcLCwnDt2jWMjo6isLDQN26KCHGB1tZWVFRUoKenx+YxYWFh2LVrFzZv3kzdq4ldWJbF4OCgzSRk8jK+qKgoq0lIXFycb3e2J4R4LEo0vNirV69w9OhRPH/+3O5zAwMDubK1TU1N6OrqQn5+PuLi4pwwUkJ8G8MwuHbtGs6cOYOhoSGbx0kkEpSWluKNN96g5Yhk3liWxcjIyLSlWOa/T96nFxERYZF4TN6oHhoaSr+PhBCnoETDC7Esi5qaGlRWVnJT6vZITk7G4cOHMTY2htu3b2PFihVIS0tzwkgJ8S96vR7fffcdqqqqZvy3mZGRgbKyMiQlJblwdMTfmHuFWOsXMrVXyNRZEPPnERERlIQQQuaMEg0vMzQ0hK+++mrGdeG28Hg87NixAwUFBbh+/ToWLFiAtWvX0psIIQ6m0Whw+vRp3Lhxw+YxPB4PGzduxN69e6lkNHE5vV5vMwmZ3CskODjYZhISHR3t9+8f//t//2+cPHkSbW1tCAkJQX5+Pv7P//k/yMzMdPfQCPEIlGh4kXv37uHrr7+eU9na2NhYvPfee5DL5eDz+SgsLKTqJYQ4WXd3NyoqKvDkyRObxwQGBuLNN9/E9u3baTMv8QhjY2Ncr5Cpy7L6+/un9QqxtickJibGL3qFvPXWW3j33XeRm5sLk8mEX//612hqakJLSwvCwsLcPTxC3I4SDS+g1+vxzTffoKGhYU7n5+XlISUlBSMjIygsLERkZKSDR0gIsYVlWdy/fx8nTpyAUqm0eVxUVBT27NmDvLw8v7hBI97JZDJxvUKm9gtRqVRcyeeAgADExsZabVoYExPjs0URlEolJBIJampqUFRU5O7hEOJ2lGh4uM7OThw9ehQqlcruc8PCwlBYWAgej4f169cjMTHRCSMkhMyGyWRCTU0Nzp07h9HRUZvHLVq0CGVlZcjKynLh6AiZP4Zh0N/fbzUJUSqVXK8QHo8HsVhsNQnx9l4hHR0dSE9Px6NHj5CTk+Pu4RDidpRoeKjx8XGcP38eFy5cmFPZ2qSkJCxZsgSrVq3C8uXLnTBCQshcaLVanD9/HjKZbMaGf2+88QYOHDiABQsWuHB0hDgHwzAYGBiYthTL/HFsbAzARBISHR1tkYBMXpLlycsLGYbB7t27MTAwgPr6encPhxCPQImGB+rr68OxY8fw7NmzOZ2/ZMkSFBUV0RIMQjxYX18fTp48ifv379s8hs/no6ioCO+88w7CqTEe8VEsy2JoaGjaLIh5ZmRyr5DIyMhpsyDmZMTd/Z8++eQTXLx4EfX19Vi0aJFbx0KIp6BEw4OwLIu6ujpUVFRwT3dmi2EYMAyDbdu2Yd++fRAKhU4aJSHEkdrb21FRUTFjP5yQkBDs3LkTW7Zs8dm17YRYw7IstFqtzSRkcnGUsLAwq0mIRCJBWFiYUytk/fKXv8SZM2dQW1uLlJQUp70OId6GEg0PMTQ0hK+//hoPHz606zyWZdHf34/MzEz86le/Qnx8vJNGSAhxFpZlcfPmTZw6dcqitOhUsbGx2L9/P9asWeP3ZUUJAYDR0dFpe0HMScjk5pkhISFWl2JJJBJERkbO+d8Ty7I4cuQITp06herqaqSnpzvqWyPEJ1Ci4QEePnyIP//5zxYNlGZjeHgYBoMBv/zlLyGVSp00OkKIq4yNjaGqqgqXLl2acVZz6dKlKCsroyenhMzAYDBYnQVRKpXQaDTccUFBQVZnQSQSyWt7hfzTP/0T/va3v+HMmTMWvTOioqIQEhLi1O+PEG9AiYYbGQwGVFRUoK6uzu7zXr16hY0bN+K///f/7vZ1qYQQxxoYGMDZs2dx7dq1GYtB5ObmYt++fRCLxS4cHSHez2g0cr1CpiYharWa+3cnEAiszoKYy/QGBARYvf6XX36JgwcPuvA7IsQzUaLhJs+ePcPRo0ehUChmfc74+Dh6e3sRGhqK//bf/hs2bNjgxBESQtzt5cuXqKioQFtbm81jBAIBtm3bhh07dtDeLEIcwNwrZOpsiHlplrlaHJ/PR2xs7LQlWRKJBGKxmPZTEQJKNFyOYRhcuHAB58+fn7G05WQsy0KhUMBgMKC4uBgff/wxoqOjnTtQQohHYFkWTU1NqKysxKtXr2weFxERgd27d6OgoICqzRHiJOZeIVNnQcwfJ/cKiYmJmTYLYk5IvLlXCCH2oETDhRQKBY4dO4aurq5Zn6PRaDA4OIjExES8//772LJlC20CJcQPjY+Po66uDmfPnrWotDNVQkICysrKqH8OIS7GsiwGBgZsJiEGg4E71twrxFoSQjOTxJdQouECLMuioaEB33zzjUWgmYlWq4VSqUR0dDRycnJw+PBh6uxNCMHo6CguXryIq1evck9PrcnOzkZpaSkWLlzowtERQqxhWRbDw8M2y/TqdDru2MjIyGl7QswfaU8m8TaUaDjZ8PAwvv76azx48GBWx4+NjeHVq1cIDg5GfHw8tm/fjj179tBaT0KIBZVKhVOnTuH27ds2j+HxeCgoKMDu3bsRGRnpwtERQmaLZVmMjo5anQVRKBQYGRnhjg0LC5s2C2L+GB4eTiseiMehRMMOIyNARwdgMADBwUBaGjBTs96mpiZ89dVXFrW8bWEYBr29vQCABQsWIDY2FocOHUJGRoajhk8I8UFPnz5FRUXFjEsyg4ODsWPHDmzbtm1+a8PtDYKEkHnT6XRWN6YrFAoMDg5yxwmFQptJSFRUFCUh82QyjUCn6wDLGsDjBSMkJA0CAcW/16FE4zVaWoDPPwcuXAA6O4HJPy0eD0hNBXbuBH7xCyA7e+K/j42NobKyEjU1Na+9PsuyUKlUGB0dxYIFCxAcHIz169fjxz/+MU2REkJmhWVZ3L59GydPnkR/f7/N40QiEfbt24f169fP/qZjLkGQEOIS5l4h1poWajQarkxvUFCQReIx+XORSEQFJGzQalsgl38OtfoC9PpOAJNvmXkQClMhFu9EYuIvEBZG8c8aSjRs6OoCPv4YqKoCBAJghqXQ3NdLSoB/+ZceXL78Bfr6+l77GkNDQ1Cr1dyUZ0hICN577z3k5uY68DshhPgLo9GIK1eu4OLFi9Dr9TaPW7JkCUpLS2fuYjzXIPjFFwA1EiTE7cy9QqztCenv7+cqXwoEAosyvZM/isViv0xCdLoutLd/DI2mCoAAwAzx74evi0QlyMj4AiEhFP8mo0TDivJy4MiRiffNmd5bp+LzGfB4Jmza9A9kZTXYPE6v1+PVq1eIiIhATEwMeDweMjMz8dFHH0EkEjngOyCE+LPh4WGcPXsWdXV1Mzb8W7NmDfbv34+4uDjLL8w1CAoEE3/++EfgZz+b4+gJIc42Pj5utVeIQqGASqXC+Pg4gIleIWKx2GoSEhsb65P7R+XycnR0HAHDmDBzgjGVAHy+AGlpf0RiIsU/M0o0pvj0U+C3v53PFVgAPKxbdxpr1ly0+IrJZEJvby8EAgEWLFgAHo8HgUCAvXv3Ytu2bbR+khDiUHK5HCdOnEBTU5PNYwICAiCVSrFr166J5ZrzD4ITfvc74De/mf91CCEuxTAMNBrNtKVY5r8bjUYA/9ErxNaSrKCgIDd/J/br7v4UXV3zj38pKb9DcjLFP4ASDQvl5cDPf+646xUV/RlZWQ1gWRavXr2CyWRCQkIC9wQgMTERhw8fxqJFixz3ooQQMkVLSwsqKiogl8ttHhMWFoaf83hY9oc/OO6Fy8uBw4cddz1CiFuxLIvBwcFpsyDmz6f2Cpk6C2L+3BN7hcjl5Whvd9xNYGZmORISKP5RovGDrq6JfYwzLGu2E4uAACNKSv4/MMxTxMfHIyQkhPvqtm3bsHfvXuoOSghxCYZh0NDQgLNnz1qthCceGsL/rKhA4Pg4HDa3KhRObCanPRuE+DyWZTEyMmIzCRkdHeWOjYiIsJmEhIWFuXzsOl0XGhuzwTAOuwkEny9Ebm6L3+/ZoETjB9u3AzKZfcuRX4fHM0EiacaePf+P+2/R0dH46KOPkJWV5bgXIoSQWdLr9fjuu+9QVVXFLYEAgF+dP48suRwBjnxLEAgAqRS4fNlx1ySEeCVzI2JrZXqHh4e540JDQ63uCYmLi0NERIRTlpk/eLAdGo0M9u3JeB0BRCIpVq707/hHiQYmHrgtX+6865eV/Q+IRK+wbt06vPfee1S2lhDidv39/Th9+jRu3ryJBI0G/7Oiwnkv1tICLFvmvOsT4ia1tbX4/e9/jzt37qC3txenTp3C3r173T0sr6PX663OgiiVSgwMDHDHBQcHW50FkUgkc+4VotW2oLHReTeBubktCAvz3/jne+UC5uDzz19fvXGueLxxtLdvwRdfCO2rXU8IIU4UExODQ4cOYcuWLVC//z7GeTzHzmaYCQTAn/4E/Ou/Ov7ahLiZVqvFypUrcejQIezfv9/dw/FaQqEQixcvxuLFi6d9bWxsDCqValoS0tjYiP7+fq6yXmBgoMWG9MlJyEy9QuTyz/H6ErZzJYBc/iekp/tv/KMZDUw0t3361HnXT0kZR2dngPNegBBC5oFNSwPPmUEwLQ148sR51yfEA/B4PJrRcDGTycQlIVMbF6pUKq5XSEBAANcrZGqFrM7OvB+a8TmHUJiGjRv9N/75/YzG8PBEs1tnevYsACMjQDh1qieEeJrhYfCcHQSfPgUFQUKIo5nbBSxYsGDa18bHx9Hf3z9tKVZzczNUKhVMJhMEgjGUlnbCmYtN9PqnMJlGIBD4Z/zz+0Tj6VPA2XM6LAt0dACrVjn3dQghxG4uCoK9dXUwOnMzHCEeQKlU4vnz5+4eBpkkIiICERERSE1N5f4bwzAYGhqCSnUDPN5xJ4+AhU7XgYiIVU5+Hc/k94nGpJLPPvE6hBBiFxcFp+P/9m94JpG45LUIcZfKykrcvn3b3cMgsyQWK7B9u/Nfh2X99ybQ7xON4GDfeh1CCLGLi4LTwf/0n2hGg/i0f/u3f0NpaSnefPNNdw+FzJLB0IyentNOfx0ez39vAv0+0UhLA3g8564c4PEmXocQQjyOi4JgQmEh7dEgPi8uLg5JSUnuHgaZJZMpBj09PADOXD7KQ0iI/94EWq/15UfCw4FJy/acYulSen8lhHgoCoKEzNnIyAju37+P+/fvAwC6urpw//592qfhJQSCcAiFzo1/QuFSv90IDlCiAQDYuXOi1LszCATAjh3OuTYhhDgEBUFC5uT27dtYvXo1Vq9eDQD4r//1v2L16tX4l3/5FzePjMyWWLwTzlvgI4BY7N/xj/powPmdwakpLiHEo1EQJIT4KeoM7lw0owEgOxsoKXH8Az2BYOK69P5KCPFoFAQJIX4qLCwbIlEJHD+rIYBIVOLXSQZAMxqcrq6J91q93nHXFAonHuSlpDjumoQQ4hQUBAkhfkqn60JjYzYYxnHxj88XIje3BSEh/h3/aEbjBykpwB//6NhrfvYZvb8SQrwEBUFCiJ8KCUlBWppj4196+md+n2QAlGhY+NnPgN/9zjHX+vRT4PBhx1yLEEJcgoIgIcRPJSb+DCkpjol/KSmfIiGB4h9AS6esKi8HjhwBTKaJP7MlEEz8+ewzen8lhHgxCoKEED8ll5ejo+MIGMYEwI74BwH4fAHS0z+jJGMSSjRs6OoCPv4YqKqaeN+c6b3W/PWSEuCLL2ilACHEB1AQJIT4KZ2uC+3tH0OjqcLEJvGZEo6Jr4tEJcjI+IKWS01BicZrtLQAn38OXLwIPH1q2TyXx5voQ7VjB/DJJ1RYhRDigygIEkL8lFbbArn8c6jVF6HXP4VlB3EehMKlEIt3IDHxE7+vLmULJRp2GBkBOjoAgwEIDgbS0qjZLSHEj1AQJIT4KZNpBDpdB1jWAB4vGCEhaX7d8Xu2KNEghBBCCCGEOBxVnSKEEEIIIYQ4HCUahBBCCCGEEIejRIMQQgghhBDicJRoEEIIIYQQQhyOEg1CCCGEEEKIw1GiQQghhBBCCHE4SjQIIYQQQgghDkeJBiGEEEIIIcThKNEghBBCCCGEOBwlGoQQQgghhBCHo0SDEEIIIYQQ4nCUaBBCCCGEEEIcjhINQgghhBBCiMNRokEIIYQQQghxOEo0CCGEEEIIIQ5HiQYhhBBCCCHE4SjRIIQQQgghhDgcJRqEEEIIIYQQh6NEgxBCCCGEEOJwlGgQQgghhBBCHI4SDUIIIYQQQojDUaJBCCGEEEIIcThKNAghhBBCCCEOR4kGIYQQQgghxOEo0SCEEEIIIYQ4HCUahBBCCCGEEIejRIMQQgghhBDicJRoEEIIIYQQQhyOEg1CCCGEEEKIw1GiQQghhBBCCHE4SjQIIYQQQgghDkeJBiGEEEIIIcTh/n8xIBfNp6uZCwAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "# Create a bipartite graph from the adjacency matrix\n",
    "G = nx.Graph()\n",
    "proc_times = td[\"proc_times\"].squeeze(0)\n",
    "job_ops_adj = td[\"job_ops_adj\"].squeeze(0)\n",
    "order = td[\"ops_sequence_order\"].squeeze(0) + 1\n",
    "\n",
    "num_machines, num_operations = proc_times.shape\n",
    "num_jobs = job_ops_adj.size(0)\n",
    "\n",
    "jobs = [f\"j{i+1}\" for i in range(num_jobs)]\n",
    "machines = [f\"m{i+1}\" for i in range(num_machines)]\n",
    "operations = [f\"o{i+1}\" for i in range(num_operations)]\n",
    "\n",
    "# Add nodes from each set\n",
    "G.add_nodes_from(machines, bipartite=0)\n",
    "G.add_nodes_from(operations, bipartite=1)\n",
    "G.add_nodes_from(jobs, bipartite=2)\n",
    "\n",
    "# Add edges based on the adjacency matrix\n",
    "for i in range(num_machines):\n",
    "    for j in range(num_operations):\n",
    "        edge_weigth = proc_times[i][j]\n",
    "        if edge_weigth != 0:\n",
    "            G.add_edge(f\"m{i+1}\", f\"o{j+1}\", weight=edge_weigth)\n",
    "\n",
    "\n",
    "# Add edges based on the adjacency matrix\n",
    "for i in range(num_jobs):\n",
    "    for j in range(num_operations):\n",
    "        edge_weigth = job_ops_adj[i][j]\n",
    "        if edge_weigth != 0:\n",
    "            G.add_edge(f\"j{i+1}\", f\"o{j+1}\", weight=3, label=order[j])\n",
    "\n",
    "\n",
    "widths = [x / 3 for x in nx.get_edge_attributes(G, 'weight').values()]\n",
    "\n",
    "plt.figure(figsize=(10,6))\n",
    "# Plot the graph\n",
    "\n",
    "machines = [n for n, d in G.nodes(data=True) if d['bipartite'] == 0]\n",
    "operations = [n for n, d in G.nodes(data=True) if d['bipartite'] == 1]\n",
    "jobs = [n for n, d in G.nodes(data=True) if d['bipartite'] == 2]\n",
    "\n",
    "pos = {}\n",
    "pos.update((node, (1, index)) for index, node in enumerate(machines))\n",
    "pos.update((node, (2, index)) for index, node in enumerate(operations))\n",
    "pos.update((node, (3, index)) for index, node in enumerate(jobs))\n",
    "\n",
    "edge_labels = {(u, v): d['label'].item() for u, v, d in G.edges(data=True) if d.get(\"label\") is not None}\n",
    "nx.draw_networkx_edge_labels(G, {k: (v[0]+.12, v[1]) for k,v in pos.items()}, edge_labels=edge_labels, rotate=False)\n",
    "\n",
    "nx.draw_networkx_nodes(G, pos, nodelist=machines, node_color='b', label=\"Machine\")\n",
    "nx.draw_networkx_nodes(G, pos, nodelist=operations, node_color='r', label=\"Operation\")\n",
    "nx.draw_networkx_nodes(G, pos, nodelist=jobs, node_color='y', label=\"jobs\")\n",
    "nx.draw_networkx_edges(G, pos, width=widths, alpha=0.6)\n",
    "\n",
    "plt.title('Visualization of the FJSP')\n",
    "plt.legend(bbox_to_anchor=(.95, 1.05))\n",
    "plt.axis('off')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Build a Model to Solve the FJSP\n",
    "\n",
    "In the FJSP we typically encode Operations and Machines separately, since they pose different node types in a k-partite Graph. Therefore, the encoder for the FJSP returns two hidden representations, the first containing machine embeddings and the second containing operation embeddings:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Lets generate a more complex instance\n",
    "\n",
    "generator_params = {\n",
    "  \"num_jobs\": 10,  # the total number of jobs\n",
    "  \"num_machines\": 5,  # the total number of machines that can process operations\n",
    "  \"min_ops_per_job\": 4,  # minimum number of operatios per job\n",
    "  \"max_ops_per_job\": 6,  # maximum number of operations per job\n",
    "  \"min_processing_time\": 1,  # the minimum time required for a machine to process an operation\n",
    "  \"max_processing_time\": 20,  # the maximum time required for a machine to process an operation\n",
    "  \"min_eligible_ma_per_op\": 1,  # the minimum number of machines capable to process an operation\n",
    "  \"max_eligible_ma_per_op\": 5,  # the maximum number of machines capable to process an operation\n",
    "}\n",
    "\n",
    "env = FJSPEnv(generator_params=generator_params)\n",
    "td = env.reset(batch_size=[1])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([1, 60, 32])\n",
      "torch.Size([1, 5, 32])\n"
     ]
    }
   ],
   "source": [
    "encoder = HetGNNEncoder(embed_dim=32, num_layers=2)\n",
    "(ma_emb, op_emb), init = encoder(td)\n",
    "print(ma_emb.shape)\n",
    "print(op_emb.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "The decoder return logits over a composite action-space of size (1 + num_jobs * num_machines), where each entry corresponds to a machine-job combination plus one **waiting**-operation. The selected action specifies, which job is processed next by which machine. To be more precise, the next operation of the selected job is processed. This operation can be retrieved from __td[\"next_op\"]__"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "tensor([[ 0,  4, 10, 15, 21, 27, 33, 39, 45, 49]])"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# next operation per job\n",
    "td[\"next_op\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "torch.Size([1, 51])\n"
     ]
    }
   ],
   "source": [
    "decoder = L2DDecoder(env_name=env.name, embed_dim=32)\n",
    "logits, mask = decoder(td, (ma_emb, op_emb), num_starts=0)\n",
    "# (1 + num_jobs * num_machines)\n",
    "print(logits.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "def make_step(td):\n",
    "    logits, mask = decoder(td, (ma_emb, op_emb), num_starts=0)\n",
    "    action = logits.masked_fill(~mask, -torch.inf).argmax(1)\n",
    "    td[\"action\"] = action\n",
    "    td = env.step(td)[\"next\"]\n",
    "    return td"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Visualize solution construction\n",
    "\n",
    "Starting at $t=0$, the decoder uses the machine-operation embeddings of the encoder to decide which machine-**job**-combination to schedule next. Note, that due to the precedence relationship, the operations to be scheduled next are fixed per job. Therefore, it is sufficient to determine the next job to be scheduled, which significantly reduces the action space. \n",
    "\n",
    "After some operations have been scheduled, either all the machines are busy or all the jobs have been scheduled with their currently active operation. In this case, the environment transitions to a new time step $t$. The new $t$ will be equal to the first time step where a machine finishes an operation in the partial schedule. When an operation is finished, the machine that has processed it is immediately ready to process the next operation. Also, the next operation of the respective job can then be scheduled.\n",
    "\n",
    "The start time of an operation is always equal to the time step in which it is scheduled. The finish time of an operation is equal to its start time plus the processing time required by the machine on which it is being processed.\n",
    "\n",
    "The figure below visualises this process. "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "env.render(td, 0)\n",
    "# Update plot within a for loop\n",
    "while not td[\"done\"].all():\n",
    "    # Clear the previous output for the next iteration\n",
    "    clear_output(wait=True)\n",
    "\n",
    "    td = make_step(td)\n",
    "    env.render(td, 0)\n",
    "    # Display updated plot\n",
    "    display(plt.gcf())\n",
    "    \n",
    "    # Pause for a moment to see the changes\n",
    "    time.sleep(.4)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "if torch.cuda.is_available():\n",
    "    accelerator = \"gpu\"\n",
    "    batch_size = 256\n",
    "    train_data_size = 2_000\n",
    "    embed_dim = 128\n",
    "    num_encoder_layers = 4\n",
    "else:\n",
    "    accelerator = \"cpu\"\n",
    "    batch_size = 32\n",
    "    train_data_size = 1_000\n",
    "    embed_dim = 64\n",
    "    num_encoder_layers = 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages/lightning/pytorch/utilities/parsing.py:198: Attribute 'env' is an instance of `nn.Module` and is already saved during checkpointing. It is recommended to ignore them using `self.save_hyperparameters(ignore=['env'])`.\n",
      "/Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages/lightning/pytorch/utilities/parsing.py:198: Attribute 'policy' is an instance of `nn.Module` and is already saved during checkpointing. It is recommended to ignore them using `self.save_hyperparameters(ignore=['policy'])`.\n",
      "/Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages/lightning/pytorch/trainer/connectors/accelerator_connector.py:551: You passed `Trainer(accelerator='cpu', precision='16-mixed')` but AMP with fp16 is not supported on CPU. Using `precision='bf16-mixed'` instead.\n",
      "Using bfloat16 Automatic Mixed Precision (AMP)\n",
      "GPU available: False, used: False\n",
      "TPU available: False, using: 0 TPU cores\n",
      "IPU available: False, using: 0 IPUs\n",
      "HPU available: False, using: 0 HPUs\n",
      "/Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages/lightning/pytorch/trainer/connectors/logger_connector/logger_connector.py:67: Starting from v1.9.0, `tensorboardX` has been removed as a dependency of the `lightning.pytorch` package, due to potential conflicts with other packages in the ML ecosystem. For this reason, `logger=True` will use `CSVLogger` as the default logger, unless the `tensorboard` or `tensorboardX` packages are found. Please `pip install lightning[extra]` or one of them to enable TensorBoard support by default\n",
      "val_file not set. Generating dataset instead\n",
      "test_file not set. Generating dataset instead\n",
      "\n",
      "  | Name     | Type           | Params\n",
      "--------------------------------------------\n",
      "0 | env      | FJSPEnv        | 0     \n",
      "1 | policy   | L2DPolicy      | 81.2 K\n",
      "2 | baseline | WarmupBaseline | 81.2 K\n",
      "--------------------------------------------\n",
      "162 K     Trainable params\n",
      "0         Non-trainable params\n",
      "162 K     Total params\n",
      "0.649     Total estimated model params size (MB)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "812cd22f33dd469cb501b27936cc1105",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Sanity Checking: |          | 0/? [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages/lightning/pytorch/trainer/connectors/data_connector.py:441: The 'val_dataloader' does not have many workers which may be a bottleneck. Consider increasing the value of the `num_workers` argument` to `num_workers=7` in the `DataLoader` to improve performance.\n",
      "/Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages/lightning/pytorch/trainer/connectors/data_connector.py:441: The 'train_dataloader' does not have many workers which may be a bottleneck. Consider increasing the value of the `num_workers` argument` to `num_workers=7` in the `DataLoader` to improve performance.\n",
      "/Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages/lightning/pytorch/loops/fit_loop.py:293: The number of training batches (32) is smaller than the logging interval Trainer(log_every_n_steps=50). Set a lower value for log_every_n_steps if you want to see logs for the training epoch.\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "fcda6b7e273945b3880e38863d6c53ae",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Training: |          | 0/? [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages/lightning/pytorch/trainer/call.py:54: Detected KeyboardInterrupt, attempting graceful shutdown...\n"
     ]
    }
   ],
   "source": [
    "# Policy: neural network, in this case with encoder-decoder architecture\n",
    "policy = L2DPolicy(embed_dim=embed_dim, num_encoder_layers=num_encoder_layers, env_name=\"fjsp\")\n",
    "\n",
    "# Model: default is AM with REINFORCE and greedy rollout baseline\n",
    "model = L2DModel(env,\n",
    "                 policy=policy, \n",
    "                 baseline=\"rollout\",\n",
    "                 batch_size=batch_size,\n",
    "                 train_data_size=train_data_size,\n",
    "                 val_data_size=1_000,\n",
    "                 optimizer_kwargs={\"lr\": 1e-4})\n",
    "\n",
    "trainer = RL4COTrainer(\n",
    "    max_epochs=3,\n",
    "    accelerator=accelerator,\n",
    "    devices=1,\n",
    "    logger=None,\n",
    ")\n",
    "\n",
    "trainer.fit(model)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Solving the Job-Shop Scheduling Problem (JSSP)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "import gc\n",
    "from rl4co.envs import JSSPEnv\n",
    "from rl4co.models.zoo.l2d.model import L2DPPOModel\n",
    "from rl4co.models.zoo.l2d.policy import L2DPolicy4PPO\n",
    "from torch.utils.data import DataLoader"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Lets generate a more complex instance\n",
    "\n",
    "generator_params = {\n",
    "  \"num_jobs\": 15,  # the total number of jobs\n",
    "  \"num_machines\": 15,  # the total number of machines that can process operations\n",
    "  \"min_processing_time\": 1,  # the minimum time required for a machine to process an operation\n",
    "  \"max_processing_time\": 99,  # the maximum time required for a machine to process an operation\n",
    "}\n",
    "\n",
    "env = JSSPEnv(\n",
    "    generator_params=generator_params, \n",
    "    _torchrl_mode=True, \n",
    "    stepwise_reward=True\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Train on synthetic data and test on Taillard benchmark"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Policy: neural network, in this case with encoder-decoder architecture\n",
    "policy = L2DPolicy4PPO(\n",
    "    embed_dim=embed_dim, \n",
    "    num_encoder_layers=num_encoder_layers, \n",
    "    env_name=\"jssp\",\n",
    "    het_emb=False\n",
    ")\n",
    "\n",
    "model = L2DPPOModel(\n",
    "    env=env,\n",
    "    policy=policy,\n",
    "    batch_size=batch_size,\n",
    "    train_data_size=train_data_size,\n",
    "    val_data_size=1_000,\n",
    "    optimizer_kwargs={\"lr\": 1e-4}\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Using bfloat16 Automatic Mixed Precision (AMP)\n",
      "GPU available: False, used: False\n",
      "TPU available: False, using: 0 TPU cores\n",
      "IPU available: False, using: 0 IPUs\n",
      "HPU available: False, using: 0 HPUs\n",
      "Overriding gradient_clip_val to None for 'automatic_optimization=False' models\n",
      "/Users/luttmann/opt/miniconda3/envs/rl4co/lib/python3.9/site-packages/lightning/pytorch/utilities/parsing.py:43: attribute 'policy' removed from hparams because it cannot be pickled\n",
      "val_file not set. Generating dataset instead\n",
      "test_file not set. Generating dataset instead\n",
      "\n",
      "  | Name       | Type          | Params\n",
      "---------------------------------------------\n",
      "0 | env        | JSSPEnv       | 0     \n",
      "1 | policy     | L2DPolicy4PPO | 25.5 K\n",
      "2 | policy_old | L2DPolicy4PPO | 25.5 K\n",
      "---------------------------------------------\n",
      "51.1 K    Trainable params\n",
      "0         Non-trainable params\n",
      "51.1 K    Total params\n",
      "0.204     Total estimated model params size (MB)\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "5b3c9b45d4fc4eb2a60a3bf80ac42a6f",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Sanity Checking: |          | 0/? [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "c0bdcd678f8140b7b5b6b123fa5084c8",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "Training: |          | 0/? [00:00<?, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "CHECKPOINT_PATH = \"last.ckpt\"\n",
    "device = \"cuda\" if torch.cuda.is_available() else \"cpu\"\n",
    "try:\n",
    "    model = L2DPPOModel.load_from_checkpoint(CHECKPOINT_PATH)\n",
    "except FileNotFoundError:\n",
    "\n",
    "    trainer = RL4COTrainer(\n",
    "        max_epochs=1,\n",
    "        accelerator=accelerator,\n",
    "        devices=1,\n",
    "        logger=None,\n",
    "    )\n",
    "\n",
    "    trainer.fit(model)\n",
    "finally:\n",
    "    model = model.to(device)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Folder does not exist. Creating folder and downloading taillard instances...\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "bash: line 11: wget: command not found\n",
      "bash: line 12: wget: command not found\n",
      "bash: line 13: wget: command not found\n",
      "bash: line 14: wget: command not found\n",
      "bash: line 15: wget: command not found\n"
     ]
    }
   ],
   "source": [
    "%%bash\n",
    "\n",
    "# Define the folder path\n",
    "DATA_PATH=\"./taillard\"\n",
    "\n",
    "# Check if the folder exists\n",
    "if [ -d \"$DATA_PATH\" ]; then\n",
    "    echo \"Folder already exists.\"\n",
    "else\n",
    "    echo \"Folder does not exist. Creating folder and downloading taillard instances...\"\n",
    "    mkdir -p \"$DATA_PATH\"\n",
    "    ! wget http://mistic.heig-vd.ch/taillard/problemes.dir/ordonnancement.dir/jobshop.dir/tai20_15.txt -O taillard/tai20_15.txt\n",
    "    ! wget http://mistic.heig-vd.ch/taillard/problemes.dir/ordonnancement.dir/jobshop.dir/tai15_15.txt -O taillard/tai15_15.txt\n",
    "    ! wget http://mistic.heig-vd.ch/taillard/problemes.dir/ordonnancement.dir/jobshop.dir/tai20_20.txt -O taillard/tai20_20.txt\n",
    "    ! wget http://mistic.heig-vd.ch/taillard/problemes.dir/ordonnancement.dir/jobshop.dir/tai30_15.txt -O taillard/tai30_15.txt\n",
    "    ! wget http://mistic.heig-vd.ch/taillard/problemes.dir/ordonnancement.dir/jobshop.dir/tai30_20.txt -O taillard/tai30_20.txt\n",
    "fi\n",
    "\n",
    "exit 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Provided file name ../../ai4co/rl4co/data/jssp/taillard/15j_15m not found. Make sure to provide a file in the right path first or unset test_file to generate data automatically instead\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Done evaluating instance type 15j_15m with reward -1408.0999755859375\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Provided file name ../../ai4co/rl4co/data/jssp/taillard/20j_15m not found. Make sure to provide a file in the right path first or unset test_file to generate data automatically instead\n",
      "Provided file name ../../ai4co/rl4co/data/jssp/taillard/20j_20m not found. Make sure to provide a file in the right path first or unset test_file to generate data automatically instead\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Done evaluating instance type 20j_15m with reward -1380.699951171875\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Provided file name ../../ai4co/rl4co/data/jssp/taillard/30j_15m not found. Make sure to provide a file in the right path first or unset test_file to generate data automatically instead\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Done evaluating instance type 20j_20m with reward -1349.9000244140625\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Provided file name ../../ai4co/rl4co/data/jssp/taillard/30j_20m not found. Make sure to provide a file in the right path first or unset test_file to generate data automatically instead\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Done evaluating instance type 30j_15m with reward -1374.0999755859375\n",
      "Done evaluating instance type 30j_20m with reward -1371.699951171875\n"
     ]
    }
   ],
   "source": [
    "# path to taillard instances\n",
    "FILE_PATH = \"./taillard/tai{instance_type}.txt\"\n",
    "\n",
    "results = {}\n",
    "instance_types = [\"15_15\", \"20_15\", \"20_20\", \"30_15\", \"30_20\"]\n",
    "\n",
    "for instance_type in instance_types:\n",
    "    \n",
    "    dataset = env.dataset(batch_size=[10], phase=\"test\", filename=FILE_PATH.format(instance_type=instance_type))\n",
    "    dl = DataLoader(dataset, batch_size=5, collate_fn=dataset.collate_fn)\n",
    "    rewards = []\n",
    "    \n",
    "    for batch in dl:\n",
    "        td = env.reset(batch).to(device)\n",
    "        # use policy.generate to avoid grad calculations which can lead to oom \n",
    "        out = model.policy.generate(td, env=env, phase=\"test\", decode_type=\"multistart_sampling\", num_starts=100, select_best=True)\n",
    "        rewards.append(out[\"reward\"])\n",
    "\n",
    "    reward = torch.cat(rewards, dim=0).mean().item()\n",
    "    results[instance_type] = reward\n",
    "\n",
    "    print(\"Done evaluating instance type %s with reward %s\" % (instance_type, reward))\n",
    "\n",
    "    # avoid ooms due to cache not being cleared \n",
    "    model.rb.empty()\n",
    "    gc.collect()\n",
    "    torch.cuda.empty_cache()"
   ]
  }
 ],
 "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": "3.9.18"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}