{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "51363966",
   "metadata": {},
   "source": [
    "# DragonNet: JAX vs TF Benchmark\n",
    "\n",
    "Compares the JAX (`flax.nnx`) and TF (Keras) DragonNet implementations on the\n",
    "IHDP semi-synthetic dataset using ATE, MAE, AUUC, and wall-clock training time.\n",
    "\n",
    "Reference: Shi, Blei & Veitch (2019) — https://arxiv.org/pdf/1906.02120.pdf"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "2613c834",
   "metadata": {},
   "source": [
    "## Setup\n",
    "\n",
    "**TF backend:**\n",
    "```\n",
    "pip install tensorflow\n",
    "```\n",
    "\n",
    "**JAX backend:**\n",
    "```\n",
    "pip install \"causalml[jax]\"\n",
    "# or: uv pip install jax flax optax orbax-checkpoint\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "d21524db",
   "metadata": {},
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "4de019de",
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import time\n",
    "import warnings\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import seaborn as sns\n",
    "from matplotlib import pyplot as plt\n",
    "from sklearn.metrics import mean_absolute_error\n",
    "\n",
    "from causalml.metrics import auuc_score, plot_gain\n",
    "\n",
    "# fmt: off\n",
    "%matplotlib inline\n",
    "warnings.filterwarnings(\"ignore\")\n",
    "plt.style.use(\"fivethirtyeight\")\n",
    "sns.set_palette(\"Paired\")\n",
    "plt.rcParams[\"figure.figsize\"] = (12, 8)\n",
    "# fmt: on\n",
    "\n",
    "try:\n",
    "    from causalml.inference.tf import DragonNet as TFDragonNet\n",
    "\n",
    "    HAS_TF = True\n",
    "except ImportError:\n",
    "    HAS_TF = False\n",
    "    print(\"[INFO] tensorflow not installed — TF DragonNet will be skipped.\")\n",
    "\n",
    "try:\n",
    "    from causalml.inference.jax import DragonNet as JAXDragonNet\n",
    "\n",
    "    HAS_JAX = True\n",
    "except ImportError:\n",
    "    HAS_JAX = False\n",
    "    print(\"[INFO] jax/flax not installed — JAX DragonNet will be skipped.\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "62cd229a",
   "metadata": {},
   "source": [
    "## IHDP Dataset\n",
    "\n",
    "Semi-synthetic dataset from Hill (2011), used in the original DragonNet paper.\n",
    "747 observations from the Infant Health and Development Program (IHDP) study.\n",
    "We use one realisation (`ihdp_npci_3.csv`) to reproduce the notebook benchmark."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "9633a572",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "n=747,  actual ATE = 4.0989\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>treatment</th>\n",
       "      <th>y_factual</th>\n",
       "      <th>y_cfactual</th>\n",
       "      <th>mu0</th>\n",
       "      <th>mu1</th>\n",
       "      <th>x1</th>\n",
       "      <th>x2</th>\n",
       "      <th>x3</th>\n",
       "      <th>x4</th>\n",
       "      <th>x5</th>\n",
       "      <th>...</th>\n",
       "      <th>x16</th>\n",
       "      <th>x17</th>\n",
       "      <th>x18</th>\n",
       "      <th>x19</th>\n",
       "      <th>x20</th>\n",
       "      <th>x21</th>\n",
       "      <th>x22</th>\n",
       "      <th>x23</th>\n",
       "      <th>x24</th>\n",
       "      <th>x25</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>1</td>\n",
       "      <td>5.931652</td>\n",
       "      <td>3.500591</td>\n",
       "      <td>2.253801</td>\n",
       "      <td>7.136441</td>\n",
       "      <td>-0.528603</td>\n",
       "      <td>-0.343455</td>\n",
       "      <td>1.128554</td>\n",
       "      <td>0.161703</td>\n",
       "      <td>-0.316603</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0</td>\n",
       "      <td>2.175966</td>\n",
       "      <td>5.952101</td>\n",
       "      <td>1.257592</td>\n",
       "      <td>6.553022</td>\n",
       "      <td>-1.736945</td>\n",
       "      <td>-1.802002</td>\n",
       "      <td>0.383828</td>\n",
       "      <td>2.244320</td>\n",
       "      <td>-0.629189</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0</td>\n",
       "      <td>2.180294</td>\n",
       "      <td>7.175734</td>\n",
       "      <td>2.384100</td>\n",
       "      <td>7.192645</td>\n",
       "      <td>-0.807451</td>\n",
       "      <td>-0.202946</td>\n",
       "      <td>-0.360898</td>\n",
       "      <td>-0.879606</td>\n",
       "      <td>0.808706</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0</td>\n",
       "      <td>3.587662</td>\n",
       "      <td>7.787537</td>\n",
       "      <td>4.009365</td>\n",
       "      <td>7.712456</td>\n",
       "      <td>0.390083</td>\n",
       "      <td>0.596582</td>\n",
       "      <td>-1.850350</td>\n",
       "      <td>-0.879606</td>\n",
       "      <td>-0.004017</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0</td>\n",
       "      <td>2.372618</td>\n",
       "      <td>5.461871</td>\n",
       "      <td>2.481631</td>\n",
       "      <td>7.232739</td>\n",
       "      <td>-1.045229</td>\n",
       "      <td>-0.602710</td>\n",
       "      <td>0.011465</td>\n",
       "      <td>0.161703</td>\n",
       "      <td>0.683672</td>\n",
       "      <td>...</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>5 rows × 30 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "   treatment  y_factual  y_cfactual       mu0       mu1        x1        x2  \\\n",
       "0          1   5.931652    3.500591  2.253801  7.136441 -0.528603 -0.343455   \n",
       "1          0   2.175966    5.952101  1.257592  6.553022 -1.736945 -1.802002   \n",
       "2          0   2.180294    7.175734  2.384100  7.192645 -0.807451 -0.202946   \n",
       "3          0   3.587662    7.787537  4.009365  7.712456  0.390083  0.596582   \n",
       "4          0   2.372618    5.461871  2.481631  7.232739 -1.045229 -0.602710   \n",
       "\n",
       "         x3        x4        x5  ...  x16  x17  x18  x19  x20  x21  x22  x23  \\\n",
       "0  1.128554  0.161703 -0.316603  ...    1    1    1    1    0    0    0    0   \n",
       "1  0.383828  2.244320 -0.629189  ...    1    1    1    1    0    0    0    0   \n",
       "2 -0.360898 -0.879606  0.808706  ...    1    0    1    1    0    0    0    0   \n",
       "3 -1.850350 -0.879606 -0.004017  ...    1    0    1    1    0    0    0    0   \n",
       "4  0.011465  0.161703  0.683672  ...    1    1    1    1    0    0    0    0   \n",
       "\n",
       "   x24  x25  \n",
       "0    0    0  \n",
       "1    0    0  \n",
       "2    0    0  \n",
       "3    0    0  \n",
       "4    0    0  \n",
       "\n",
       "[5 rows x 30 columns]"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "DATA_PATH = os.path.join(os.path.dirname(os.path.abspath(\"__file__\")), \"data\", \"ihdp_npci_3.csv\")\n",
    "\n",
    "df = pd.read_csv(DATA_PATH, header=None)\n",
    "cols = [\"treatment\", \"y_factual\", \"y_cfactual\", \"mu0\", \"mu1\"] + [\n",
    "    f\"x{i}\" for i in range(1, 26)\n",
    "]\n",
    "df.columns = cols\n",
    "\n",
    "X = df.loc[:, \"x1\":].values\n",
    "treatment = df[\"treatment\"].values\n",
    "y = df[\"y_factual\"].values\n",
    "tau = df.apply(\n",
    "    lambda d: d[\"y_factual\"] - d[\"y_cfactual\"]\n",
    "    if d[\"treatment\"] == 1\n",
    "    else d[\"y_cfactual\"] - d[\"y_factual\"],\n",
    "    axis=1,\n",
    ").values\n",
    "\n",
    "print(f\"n={len(y)},  actual ATE = {tau.mean():.4f}\")\n",
    "df.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "23546a75",
   "metadata": {},
   "source": [
    "## Train DragonNet"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "62e5bb15",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[TF DragonNet] fitting...\n",
      "\u001b[1m24/24\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 1ms/step \n",
      "[TF DragonNet] done — ATE=3.9952,  train time=4.42s\n",
      "[JAX DragonNet] fitting...\n",
      "[JAX DragonNet] done — ATE=3.9220,  train time=1.71s\n"
     ]
    }
   ],
   "source": [
    "SEED = 42\n",
    "HYPERPARAMS = dict(neurons_per_layer=200, targeted_reg=True, verbose=False)\n",
    "\n",
    "np.random.seed(SEED)\n",
    "\n",
    "results = {}  # label -> {\"ite\": array, \"train_time\": float}\n",
    "\n",
    "for label, cls, available in [\n",
    "    (\"TF DragonNet\", TFDragonNet if HAS_TF else None, HAS_TF),\n",
    "    (\"JAX DragonNet\", JAXDragonNet if HAS_JAX else None, HAS_JAX),\n",
    "]:\n",
    "    if not available:\n",
    "        print(f\"[{label}] skipped (backend not installed).\")\n",
    "        continue\n",
    "    print(f\"[{label}] fitting...\")\n",
    "    model = cls(**HYPERPARAMS)\n",
    "    t0 = time.perf_counter()\n",
    "    model.fit(X, treatment, y)\n",
    "    train_time = time.perf_counter() - t0\n",
    "    ite = model.predict_tau(X).ravel()\n",
    "    results[label] = {\"ite\": ite, \"train_time\": train_time}\n",
    "    print(f\"[{label}] done — ATE={ite.mean():.4f},  train time={train_time:.2f}s\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "4da1e8cb",
   "metadata": {},
   "source": [
    "## Results"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "b00cebee",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>ATE</th>\n",
       "      <th>Actual ATE</th>\n",
       "      <th>MAE</th>\n",
       "      <th>Train time (s)</th>\n",
       "      <th>AUUC</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>TF DragonNet</th>\n",
       "      <td>3.995167</td>\n",
       "      <td>4.098887</td>\n",
       "      <td>1.199390</td>\n",
       "      <td>4.416656</td>\n",
       "      <td>0.552354</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>JAX DragonNet</th>\n",
       "      <td>3.921982</td>\n",
       "      <td>4.098887</td>\n",
       "      <td>1.212568</td>\n",
       "      <td>1.710895</td>\n",
       "      <td>0.552620</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                    ATE  Actual ATE       MAE  Train time (s)      AUUC\n",
       "TF DragonNet   3.995167    4.098887  1.199390        4.416656  0.552354\n",
       "JAX DragonNet  3.921982    4.098887  1.212568        1.710895  0.552620"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Build df_preds following the same convention as dragonnet_example.ipynb:\n",
    "# model columns + reserved columns tau / w / y\n",
    "preds_data = {label: v[\"ite\"] for label, v in results.items()}\n",
    "preds_data.update({\"tau\": tau, \"w\": treatment, \"y\": y})\n",
    "df_preds = pd.DataFrame(preds_data)\n",
    "\n",
    "rows = []\n",
    "for label, v in results.items():\n",
    "    ite = v[\"ite\"]\n",
    "    rows.append(\n",
    "        {\n",
    "            \"ATE\": float(ite.mean()),\n",
    "            \"Actual ATE\": float(tau.mean()),\n",
    "            \"MAE\": float(mean_absolute_error(tau, ite)),\n",
    "            \"Train time (s)\": float(v[\"train_time\"]),\n",
    "        }\n",
    "    )\n",
    "\n",
    "df_result = pd.DataFrame(rows, index=list(results.keys()))\n",
    "df_result[\"AUUC\"] = auuc_score(\n",
    "    df_preds, outcome_col=\"y\", treatment_col=\"w\", treatment_effect_col=\"tau\"\n",
    ")\n",
    "\n",
    "df_result"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "a1b3834e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAwkAAAMKCAYAAADODwgiAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjExLjAsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvlcelbwAAAAlwSFlzAAAPYQAAD2EBqD+naQAA499JREFUeJzs3Qd0FNXbBvA3vQMhEHrvvSldELEggiAoKKKC2MCO+LdiL2ADQZoFFQsgIgiIgtKUXqR3Qm8hQArpZfOd5/JN3GxmNm3LbPL8ODkJmcnuZDI7e59bveLi4rKFiIiIiIjo/3lrXxARERERETEkEBERERFRHmxJICIiIiKiXBgSiIiIiIgoF4YEIiIiIiLKhSGBiIiIiIhyYUggIiIiIqJcGBKIiIiIiCgXhgQiIiIiIsqFIYFKlNdee00GDhwor776qpiB2Y6npOJ5JiIqeT7//HP1Hjp58mR3H4rpLV68WJ2rl19+2WGP6euwRyKH2rdvn/z9999y6NAhuXz5sqSnp0u5cuWkUaNG0qlTJ7n22mvFy8uLZ93Gtm3bZN26dZKUlGSKc2O243G0I0eOqN9vz549EhsbK1euXJGyZctKpUqVpE2bNtKlSxf1tbOV9PNcXJMmTZI1a9aor6dNmyaRkZF298f95oEHHlCf4dFHH5Wbb77Z7s9YLBZ56qmn5Ny5c+Lr6yufffaZVKxY0e7P4P6GN7Ts7Gxp3ry5vPnmm1JS3HvvvZKamlqsx3jhhRekffv2cuHCBRk5cmSBf65nz54yatSoQj2X0XN4e3tLSEiIhIWFqb9nixYtpHXr1lKnTp1CPT4Vj+3fp0GDBjJu3Di7P4PXPF77MGbMGFV28CTHjh2TsWPHSlpamvps75x0795d3X8Ks13Phg0b5KOPPlJfDx8+XPr06WP4mI58ndh7jfv7+6vyX5MmTdRru1mzZrr7XXfddfL444/LihUrpEePHnLTTTdJcTEkmMzy5cvl/fffl+3bt9vdr169evL000/L0KFD1cVJrvO///1PoqKipGXLlvL666+XylO/atUqGT9+vGzcuNHufj4+Purm/Oyzz6obGLkHQhzeOCAlJSXf/X/77Tf5/fffc/6fkZGRb0jAfahfv35y1113qf8/88wz8sMPPxjuj8d8+OGHZefOnRIQECBvv/22lCSrV68udmhFIUX7m2l/v4KoWrVqoZ+rsM/RqlUrGTZsmNx///3qdU7OZfv3wde33Xab3fvq+fPnc34Gfydn2LFjR85rFyEfYd9R8HgICPg9UeC2d04iIiIKvd2osK79TM+ePe0+piNfJwV9XLSa416MlhXbCjgECVTofPjhh6psguMvbvmQIcEksrKyVJcU1PIBWgmuv/56VcCqXr26BAUFqRf81q1b1Rs4CqlIxXjz/+CDD9x9+KXK5s2b1Y0xMzMz333feustVcOOF29JgN8ZN58pU6bkfK9z587SrVs3qV27tmpFQMvX8ePHVSET1+fKlSvVx8cffywjRoxwynGVtPPsbrNmzVKf8aaGe9M///yj/qb4G9uDmiu8GeLncZ+aM2eO3H333br7vvvuuyogAK6ppk2bSkmCgGR0j0BFEFq/4Ntvv1U1kEYFDFt4reVXI1qtWrUiHbPRc6AAk5CQoN53cP9Dqx3+dgj/+D2//PLLfK8Ncrw33nijUAVWZ7h48WLOMaBiwFHw3vHrr7+qr59//nkxo25Oep3YPi6CElpdca/AfRiVyaiQ+euvvyQ0NDTXz6IlAa246I3yyy+/yJ133lms35EhwSTQrIwLCNBMhbCgl8hRyELtFILB1KlTJS4uzg1HSwXVrl27EnWyXnzxxVzXKcICWlT0vPLKK/Lvv/+q8Lt+/XoVHpylpJ1ndzp58mRO1yTclyZMmKDe/L7//vsCja1B4R+16Hgc/Dze8GxrtvHmqXWDQGVIYbrSeAr8Xka++OKLnK9REVSYcFu5cmW58cYbi318xXmOo0ePqnsBCiuouLrjjjvkzz//lAoVKjj1uOiqMmXKqMIogubChQulf//+Je7UoKYc3RBReaDXimAGlZ30OtF7XLSmPPLII6rQjy5RBw4cUGVA9GywhntJ7969Zf78+SosFDcksJ+KCSAtawUvBIOlS5fabbJDrROa4X7++ec8KZLIWfBmpF2nGG/wxx9/GAYETdu2bdX1jCZSdknwDAgDGF8QGBio3pRQYwU//vijalXID/rkIjyiNTQ+Pl6efPLJXNtRuHnsscfUc4SHh6sKEY6v8ix169aVuXPnqnEXWt/x0aNHu/uwSg106atSpYr6Gl19CtKq7UlQ+blgwQL1tVFLZGl8nYSEhOTqOYL3ZD2DBw9Wn9HjQWuxLCq2JLgZkjKangF9x1DDhDfZgkAN1DXXXJPre3jj1foEDxo0KOdi0fPee++pCwgDa7SBOtbQ3HXmzBk1cA41gnhspGB8YHCilnZvvfXWXD+HG9aiRYtU7fHZs2dVX0B0Q+jbt69uYcCRx1wQiYmJalA4xn1ER0dLTEyMKhChthO/K34fDBQyGvyJ2gHYvXu3mknAtobn66+/zvk/Csd79+5VA47eeeednO+j4PzVV1/l3OTz62qBIKl1AcFNAmNS9Bw+fFgV3tHUiO43OB48Nwp6uGEVFf5GOPfadYoZJ4y6SOjBjRHn3dF/k/zOs9G1DPhboksMvo/fBcEHb0gouJZW+DsjDMDtt9+uaqUwgBndhvBaRvP2Lbfcku/joJ80AsaMGTNUVwS8JrT+9RhAeerUKfU1Wim0wk5hPfTQQ+oax2snv/EMuM/i+TGwHn9nvRYR3Fdwb8NgfOyH3x19fjFZBGrmSvN1oQf3cnQhxOvo9OnTamYVdBGxreCyHgiq3bvw+kbhCfcpfI3WJtswefDgQdUideLECXVPSE5OVn8T/L1xDTZs2LBAx4kaXBwbCmi4h2DA7z333CM1a9ZUr32tWwc+4z3VCK5/3C9Q8EKrKO4Z6DqC98COHTsa/pze74+fR/cTvIegZwC6FOMas/f81tD9GDXUGJeIri3ffPONej0UByYpwPW/du1a9f6OMUN4beK1jGPz8/PTHWh76dKlnO+h26Dt60Rv8G9+EBC0SRMGDBggpeF1UlBowcfrAEEK7/d6brjhBvV3wP0Rr7PitLQzJLiZ1mwEuNmgkFMYtgU1vMlr/QNtA4Qt9JXDvkYXK44NFyEGFeKmdt9996luAtZQ0MUNBAVYzGiC/nIo5OMGb+27775Tg23wGY/nrGPOzyeffKJmhNBuQLZQA4rwg2Y8vNCMBn8Czoltf9Dy5csXaNYdNJ9i8C9+d7xp5DeuBM2GW7ZskVq1aukW9nHDQOELTYwoEOl1AcFsJ+jDWpSBTLgW0CcSevXqpd5oC8uo1as4f5OCzm5kfS2jJhsDZpctW5Zrn59++kn9HVBI9rRZQBwF1zPeyACD7QDnAgVlvKbxOi9ISLDuL41CN2YmwWwbCIE4z4BAVpxuEnijRGsqCpLoh4trxAgCqFbrprWMaNDagesBXQKM4N6G1xAGBdJ/tNYmhHTcd1CZYXtvth4Iitce/mYIBNYD6K0HlWIfjHPSrkM9uJ5QsYSQaXRfQT/uJ554QubNm5dnGwZ2YuIFFIC1Y0MFlb0KEnR/wWPaQgDALG6oONEbC2L7++O+g+sNX1tDoEaBGJ9tC+R6MGkJ7o24L+O+heBTmIobawg/qDzRO+doPa5fv776bN3tR2+gLbqX2tIb/JsfbdIEhCd8lIbXSWHgmsd7PoIc3jdtK9Bwv8IMmLinodKwOONWGRLcTOv7C/nNHuIuuEliMCJqfVAQxWA6TO2Hmy8uQtSKY+AoaosRGPAiQH90BB7UEKP7AmonsC9utu6c5hA1mHiB4Vzj+FBTiJYb1GYhgOANDAPEccPFjQrdZTSoacIbCW6mqL1Bored3aggN3dADTkKvKiZxXOi9tuophxvAggIMGTIkDytMbhZ4PfRCvEo1KHvI26uCDIYvISBw2gJwb5aX/CiXqeO7g9dnL9JYWk1yhhUhtCElgWEJgzMRW0cal7w5ovtBZ0NoyTRWqsQAlFQ0+D1j9c0Cjgo9OQ3hapW24muRAiVuA+gphOBAVCLW9wJF/B3QsEFXaDQ0mFv0CTuQYBCFF4b1nAdaAEBAQitnrgG8QaM2mO8rlCIQu0q2R97gft8fvcR3P9RGYRCPu6DqHyxro3GeUcNP17neGzsg78HCkP4PgpYqBTA+w+uq9mzZ+s+F0KuVthE6xHunSjE4/rFPeW5554r0Ox0uO9r1w8qaPC4qNhBuMR1gQ9UUOC6QcWPvSmfN23apFqx8Lvh98fxoJUElW04D7hX4zkKMvYH3TdR6MTrAL8TKpK0VtLCQCsf3rtxb8R9F+cJ90VUqKClA70b8LpFTwC8/rVWb9wDcB7RsqK13uJvazs9Z0FbfKzLG1plZH6VhiX1dWIPyl74ewN6ChiVG3DucF/D2LCCTDphhCHBzdBFwt5MFmaAGx+aHfHCtR6AiJsc+trhJombHArOKFihOc16IB72Q9cF/Dz2w0CbotZ4FBdCDmp+jPrHoyCEoIOCAYKPdZ8/JH98aN3B0GpQnAIzbu4ICSjI483MtoZTo71BoTCLG7gtFHK0gIBjtp35BM+Dmnp8oBCIAVAFrQ3W7N+/P+fr/MYhuPJvUlgISxg0hppl6/mrcX2iSRZ95VFgwSwSpa2PNUIZap30pktEQMO5R00qCmXo5lAQqM3C9YgaX3T70K5j1JbiDa44UKuJ1yMKV+i6YRQSUJhbsmRJzt/ZujsnCv6o5AD8vVHoMmrtwvkxA1y7tt0crbl6amYUCrVZsNClxx5UEqE1Bl1stcoOdEWxblXAdYEKKaNuaLie8D6CQj7umyhwde3aNdc+aK3SAgKuXdS4W7egoqCP61nr6msE3Wa1+y9awnDto1ZYg/sx7qk4JgRKfEb3DiP4u6DVy7aiDK1qCBm4lvHawGw4BXmPxLnr0KGDCh8ICQ8++GC+a5NYQyUM3o8REBB80M0HAd768dHdEEEfhU3cH3H9aZUAtu9/CD/FneoaLZZaizDCnateNxqtAO7O14k9uLa1Vnd73dysuxihBbeoIYEDl93MesYXs9ZcomCAN0m9ube1Aile1AgLqDm0nakDLwzcGAE1P7ihuQtqbuwNoEUBUjtWvPk4c3EuFHy1v7nRfPIY36G96aDfbo0aNXJtRy2OVshByDCaGhE1TFpTcVFaEqxr+my7VNmGCdyIjT5QS+TOv4l2LestcIM3aq0rF8JbaYPaeNTiombKNozib46aRECXwcJ46aWXco2hQah1VHcuBGBAVzK0/uhBFzytEKoNINSg4KN1z7PXLQLN90UdO+Fo2tz3Rh96XT6cCa9dLfDlN4MZ7l+odbZtDUWB07o1Nr9zjVn+tHuhbbdB0KYSRxc09AfX62KJ2np0o7FHu1eiwI7uRNYBQYNArVXw4FisK1RsNW7cWDfA4ffXKiUwHsboWtajBQ78XGFb57A/XvP4e8ycOTNXQNDgfR/nEHbt2qW69zmT1toI9roQOvp1o32g9cTdrxM9uE8hxOF+qkFoM2LdooUK3KJiS4KbWa/IadtX3yxwsRn1BUfrB24wuIDR/caon511jQDe0I0ez1VQK41BWhgPgi4m+DtohQXM+6wV0NFHE/2xnQGFMbSyTJ8+Xd2cUKtp++aIY0RztHWByJo2jzSg/60R/I1QQEKoQP98vKEUdIB8Ya5TdGeyN283QoLRmAhX/E3QRG6vWx+a2TEw3fqNqrTQCv+oPdSrsEBhCN0LcG4wKYF1d6T8aga1gcqAAhBq0xwx2xXGP6HQhfCHoI2/ny0tgCMYou+4NetaV9QYo1YWgcDM8lsnwR2VTVrhGTWc9v626K5Z0POL1z/COroWIcyhgsl6Fh+0EAG6UlhDAUxbjBRdy4KDg3UfH/chtDLoreQLuAdpM8Og9dVeDT1q8LV7MY7ZaGwhwrfR/c/62sV7JFouCgK1yahwQmURukwihBdklV8EZ9xvtWvKXk8GtBigwI6CNrqw2Jvet7i09ztw5GQBBVlfBLBAaFEnRXHU68S21QP745rA+degBcheOcr63Fn/XGGZ+25YCljXumMgU3EXwXEG1H7Yu+hRy4IbuL1B19a1z7YDtlwJNeIoTFuvJmsPCtPOhII/QoLWrxrNzNa0pm4sUqY3Q4T2Zoi/Q34zGGgBDgV1dHOz11RZ1OsU1wAKktZQc4hwYoa/CY7P3nSb2vzVWgGkMKxn6XImR69qajswHd0LjN5k0SKAWil0sShISEBBD4M0rQelIySghtb2Wi8KvBFi5it0QUMtG7qPWBcKETi1gh5Csu3fHi1HKJzhmNCVBG/OKBAiTCAwOLIm01FcsU5CYWmvF9SU2gt/BZ3vHn3z0fpZkC5etvcD65r8/Lqr2DsePI5WSZHfvdW67zy6ShmxN4ud9dz5hb3/ICijFQOtApjpC60C+cF7gPa6tA3PtvC6QXcZFDbRmuBM1l3PHLk4ZkFfN87sPRBfwNcJzrNewR4/g1ZYdK3M73exDgnF+Z0YEtzM+k0ItW2Fnd3IFfLrG6nVjNjbz/oF4a45nfG8mD1CW+UVhQDU1qCvnvUAIPSdRv990JspyJFQ2MMbFQrRmFXHuuCE2nNtQCUKn3pN3VoNO27iqBXTaMdt/dl6CtLCNndaX6eoyTe6TnFTt715aU3/Zvib5Hcta7WcRblGrWfpciZHrmpqO2AZPv30U8MuaVqLEvpqo6tCfuMK0Mdfm70NM8og9OJvjcI8xsU4YpVlBG2EBIRXHJf1vOrW43msXx/WUPuK7isIShgYi24l+ADUyCKcI+jodcWgq62HmJ4U8lskyl5XRQ3Gj+Dvgdc5ri+MI0EtN1oB0S1He79BtwvUrtreD6wL2Pk9n73t1guV5vd74b6CcIrzYG+BU3vrGlm3sBT2/oOWVYRgvI4RljF7VH4BSXvvAG0wuN57h/Y1xkuAMxfFBOuWFusW7NL0Oulm0+qBLnh4b8WEEkYtY7asZ+Iq6IQqehgS3Ay1udq85KjNKu4MR4VZlMgsC7C46phRO6UVRjGADs2yeopSi1zcQg5CAt7wMF4DBWXrPuLaPvbeWFD7otc310hhb7621ykG2TmCWf8mRYEgbNuK4gyObkVA4dq62xoGd+cHb3b4XdHNwggCk7ayMAY+oqCNGkuEQLyBoT8t9inOGxigyR0tWyjgo2uRFhJwr9CmW9X20YM+12jFQjcDdNlAWMA1iVpWDDDE1JeYRQlhtySubFtc2sxrBalxL0hXIwwoRqEUFQWoJDGaScuom5D1bC9G0yprtPtrcR/Heh93dRtGaMJrEq9NtCwgMBf0b4FWBetJVOxxdsEdrebWXb5K4+uksgNaC63PnfU5LSyGBDfDYDm8WLUBqi+//HKR5rG3LqjgTRc3P705nY36/rmTq44Z01xqtUf2BvwYLVDiLFg2HYPocPNFIUcLCVqhHM28Rs3iKOAgYGC604kTJxb4OTF+pDBQyNJmZsAbEd6QinOdmv1vUtSwa7ZuIAWBgb0oWOA1iHEJ+RXkUGDGLEjY1ygkoLYRg81R2ENfbhS0AS0HuHawhgK6LaA1ArNXFYfWSoB+xBjYrk33p03Xai9k2wZhrQse7kOoWUULBVoaEMLRJQ59sR3ZBaIkQGWGdQ1oceD+rnV7w5S5RgEB75e2YxE01j9jPRZGD8Y6FGTgp9FzWT+PVoFlbwpUZ8J4NtxDMTEDus3l16ppPREJZisr6KxEBa3JLirryTnstcqU5tdJYUNCcdaaYEhwM9RuYYAKAgJuNGjmtldYsu3egHmnbQc4oSkLg2DtTbOFvp72ZmFwNVccs/aiQV89ewVczNJkj9Z1ylFdkVDoQJcGFL7RVIxuNeiiofVttVfAwY0dtZ94c0WQyK8Zs6hwk8F1ippZ/I0wRai2gm5xOOpvQsXvaoRpJFHjnx8UoBESMB4GM4HoBU40lWt9ahEQrAd9Yhtq7tFqhqlRMeiyuFMd4jWCGVjwmsTYAgQRrasRAiieozBQG4wxF/hAoQW11uiuh9pAR7WilQS4BrRWKBTOCzLFZEELNvYGYKMAbN133RqCKLr/oB82ZuKxtwCe9foveo+DyR0w5gGDka1nldE7Ho3e4HlXQWEfoRYhHa0J9iazQJdRvC7xvopKgqJWcFh3JXbEe6L1OgueUDnkjtdJQVgvaFucbp2cAtUE8Aak3RBRw6bNVW4Pul+gFk9vIRmt1hnrGxgN8kQNHmqFzcIVx6wVVFArZN0f0xoKFvlNgab1K3VkLYcWBPC7o5lYm5EFTd6YwcUItqHfLlph0Ezv7OtUCyFo8dJmxigOR/1NqGjQz1gb/F7QgjQqJbQpK63HMmjwPW1dAtyjbIMHwiBWz0aNJGpfR44cWewuDGg50AZSowUOoVm7PjGex2jBoYLAKucaZ49R8rRr57777supQUdraHFrma3DpNZH3hbCgb0FOdESpk1Jiu5KRpMmoFLOunZXr/CrFeZwLEYTKyA8IuxqFT6FXYPGkdCtBOvOaH8fvdWmrVs+tYoevN9oY4cKy3qchSPeE/Eeo83pb3QNlPbXSUFo5w6vh4JOGKCHIcEEUEuL9I8LB2+WGICEwYnWSVCDGgLMXIDaCqNFpbTl5dHXGE3+1gNWUWOAGyxuChgEYxauOGZtvAeCBvpHWy+aghcw+k9j4HB+XXG0Qbuo6S/O/MPWunfvnjMwEitgan3bUcCyN7gO2959992cwhlu+ka1Lyg44U2xMN2SrKFGFS0IeFPAGzUCCmqqUIDXKzyhhgqLGNm70Tvqb0JFg7+nBrMEFQTuU9oUiCiEWBfw0cqEAAmYg15bidUWZklChQigcGK0X2HgjVgr/GHgpvambLs2gjVM5YpuSrhG9dbwQNcXbcA8gpE7a4nNAOcIr3fMPISaZ0xiADjftgvwFQUqy7T+2nhPtB3jg+mJMdEBxp/YrhljDQut4TrFfQXTjmKciW3BDQHAOgDqQYFb616G+5P12B04ceKECqH4DLj27Q1OdgV009LeS/LrcoRWPdQy4z0Ws3rhPVavhQbnEa1omHDAdnYjrKistQJrlQPFpbVqYGyQWcZOmul1UpgxEOi+XJjpzm2xu5FJoNsIaj2wEiTmIccNEh/oZ4gQgeZvFKBwk9ReNBjcghumLQyuQ4ERtfKolcZnFLLwQsfNEW/qGISHvshmac5zxTGjwI1CKc4zHh8rk2JWCLyZaHPzozYSNZtagUMPBkaiWxj+Dtgfj6PN8oLPKOQXFmp18GaGAon1YnMF6UuNY0XTOmr60V0JH7hm8CaKWgTU0KPArtXUo991UWfIwcBTNL2jhhghCbX8+EBYwRsTarIwgA+rj1ovVIXmfzT720775qi/CRUeXlNaTSNqmgrTbxUFCtSsouYQK6yjoITX6iOPPKICPq47vEbs1Zih0IVuZOjygZYFPGZxFllD7THmDkdFgzYrGK4neyuEI1AgoOADIQD9yfGB6xTdpXC/Bfwffb1Ly3gE23na0cUMrddo8bNu6cXrfvz48Q6d+hcFUXS/xH0EBV50m0HNMq413HNwbaGiDLNwGY05QAjFCsS4HnEvQgDG99C9F++juLfgesd7irZIoN6UlNgH70u4D+O6wvTAeAxMnYvzgfcmLVzi/obncze0mqG2uiDHgkCDIIbfD4vwoYIOAQsFf3QBxfsKzhfeP7TwgFXUrWnTc+N9G91REbxxrrWxTai40pu+O7/XMsY+IbysW7dOVaKZ0d9ufJ3Yg3uXNhBda1UrKoYEE8GsJZhhA2/c6EaEWWTQTx8fGrzw8CJFQRU1uXq1Fihwoskd0w9icCEuVrxwAbUGqHlGlwEUuM3CFceM58ANH/NIoyCPG5A2sw5uiCg4oyYov9ldUKBCDTfevFBDYD1rQUGm+DOCGk90qdLedBAQ7a0Caw3jWBA08caJ7mo4Lq32wnqgGmqAi3vTwLoZGCCKNwUEBFyzaOGynRoP1yb6mmMKQ4RZvT7GjvqbkH3WLT1arR8K91r3gML22Ue4w+PgWsXrFW9+uHa11wJq0Nq2bWv3MfC3R0EOwRMFMMxshesqv2lqjaCQj+sMlSsFDdkIoKgtRlBBIQlv7taDVDGYG/ed559/Pk/hqCQzmqcdEKJwD8S5xmvberVkR0BrDa5NBD7cC9BqgA+tVhStyqjowL3OHhwf7qGoPMEUymj1xQfCAO6BuF6t31uNalsxyBSFQcy+hppy6+PBNYx7HArWhS0IOxNej3htFWRNA7wvYJA/XscIXyhc2v4cXgcI8Hj/0Avd+FugbIL3BNv3noK+h1nD+CgEQ7wWMV7TrCHhvBtfJ/agDIlKG1QuY2KU4vCKi4tjJ0sT1/ShNhaLTaHPOQpNSOiF6c+G9I/BvqgRQK2I9UqM6KuJ2mXcHLUZdayhiRY/h36i9lZjxBssjg+Pb7TwGgopWtMnamHw4YxjRrcB1D6jxs96gRu9c4taKYQRDCTCedX6LePxtX6sKBgYTR+G3wnNzLghogYB/8fN1PqGVtDj0aDWRKuxwU2mKN1sUHBDaxRqf7TZZfA7OnL1SmtoUcF5QEjA3wzhAP1K0ZRfmOlti/M3ye88F/RaxnnDGxMKwO5eFdyRUFjWugKgSxCuBbzGtMJOUQa9a9eqdq60+wD+5ihMFHRFZRRKtMIarvfizA6DN2xtPndAYbKg3T9w/Wlv+ng9I/Cje6Pe+iTFpd3HAOcqvxmlcJ5xvrUWZEdPg2v7HNbw98V7Du65eF3aW3lYD2qitUInAkB+a2tYQ0sBXo8o7KCl0nq9FgRS1Nqi8iG/ge9oTcDj4H6C+4p2L8SkD2jBBbRk5vc4aN1AN2Dc6xBmUZDN73VT0N8/v/dI678P7q35db3F72u9cjwK90azRVnDfRQ/i+COvzl+P7wmCzK1K1oR0dKPn9Wml0WrRFHWGEErDwaL4xjQ7U+vkJ3f66Iorxvrv1dDnWN31uvEka9xVBiiuxNa37WZ5YqKIYGIqIRDIR415XijRTBwxPS1RJ4OLV4zZsxQlTuo7HHXGgeUFyp1UHmByi4MDHfEbHqlwZb/n4ENlTTouoyxYcXBdwoiohIMXYq02aFQm8iAQHR10LG2Fg0KVQwI5oKWGm2mJkxvXJAF7UhUtzit+3JxAwIwJBARlVBowsYbrdb076qBc0TuhrFZmI0I45xsoWscBiyjayNCM2ZQI/MZMWKEmk0QrTx60y1TbhgfiG5z6NJmtCJ5YXHgMhFRCYOpDdEfGf3z0T8YMLUkZmchKg3QeoYJL9CVCDMSYRAzYPCyNtUyxs9gn9I0KN2TYKwOJnHBvaw4k4KUFhEREWq2KoxpKOyYCCMck0BEVMKg0KNNFYxm+3vuuUfNxlWc+bKJPAkCMtbAwGBg2wU6UfjEbD2YtQqzFxGRPoYEIqISBrM5oQ8vat8w4xhqU4lKI0wFidmz0HqAsICZ0dBX292LnhF5AoYEIiIiIiLKhQOXiYiIiIgoF4YEIiIiIiLKhSGBiIiIiIhyYUgoBVJTU+Xo0aPqM/FcmgGvSZ5Hs+E1yfNoJrweeS7NgCGhFM3wQDyXZsJrkufRbHhN8jyaCa9Hnkt3Y0ggIiIiIqJcGBKIiIiIiCgXhgQiIiIiIsqFIYGIiIiIiHJhSCAiIiIiolx8c/+XzMJisUhSUpJDpi3FY/n7+0t8fLxcuXLFIcdXWvFces55DAwMlJCQEPH2Zl0IERFRYTEkmLQAdenSJQkNDZUKFSqIl5dXsR8vPT1dFcpYYCr+34bn0vznMTs7WwVsvI4iIiJ43RMRERUSq9hMCC0ICAhBQUHFDghEpRFeN3j94HWE1xMREREVDkOCCaEGFF0liKh48DriSuNERESFx5BgUmxBIOLriIiIyF0YEoiIiIiIKBeGBCIiIiIiyoUhgYiIiIiIcuEUqGQaiYmJasrKgsL0sJgHPyEhQWJjYw33K1u2rJQrV87uY9k+ho+Pj4SFhUmZMmVK/fgQ63NTvXp1dW5snT9/XtLS0qRmzZql/nwRERGVBAwJZBq//vqrPP744wXe/6uvvpKBAwfKt99+K2PHjjXcb8yYMfLqq6/afSyjxwgODpbmzZtLnz595P7771ehobSxPjcTJkyQ4cOH59ln6NChsnXrVhUWijozV1xcnFpcrWrVquLn51fs4yYiIqKiY0gg08Cc9jVq1Mj1PazGi8IjWgJQs29bgLcWHh6uHkOvJaGgrB8D8+ujBn3z5s3qY8qUKTJz5kxp166dlFYffvih3HPPPU6Zovejjz6Szz77TNavXy9NmzZ1+OMTERFRwTEkkGn069dPfVibPHmyqsX+3//+J6NGjbL78y+99JI88sgjxToG28dASPnnn39U4Xj79u0yZMgQ+euvv6RevXpS2qBFZc+ePfLFF1/Ik08+6e7DISIiIifiwGUiO9B60bt3b1m+fLl07dpVdYeZNGlSrn0uX74sJ06ckMzMTPX/rKwsOXfunNrXVkpKiuqSk5GRUaDzjjEaaEnRxMTEqOcygsfHc+OzEdvjxWccU3p6ut1jefDBB6VatWqqyxHCU2HgOaKjo3P9Ltbw/Npj4mscHz7w+xIREZHrsSXBA60/k1So/bPxYckWL+8s8RLn6lwtREoi9JF/9913pXv37vL777/Lp59+mrPt7bfflq+//lq2bNmixkl89913qqvSyy+/rFpALl68KF9++aXMmTNHFXyzs7PF399funXrph6zUaNGeZ4PoeSVV16Rw4cPq/+3aNFCPSe65CxdulQ9pq/vfy9fdId6/fXXZdOmTWKxWNTg4o4dO8pbb72Vp3uUdrz//vuvGm+AY8PxogsRxha8//77umMCAgIC5Pnnn5dnnnlGdQtCq0t+jh8/Lm+++aYsW7ZMkpOT1fcaNGigzkvfvn1z9rv77rtlx44d6usBAwbkfB8B7ccff8z3eYiIiMixGBI8UFyapYg/ibhQcmH8gG0tOwrLmJHHEVq1aiUVK1ZUtduorUetuu0A6TVr1khQUJAaW6GNoVi0aJGMGzdOfY2BzyiM41jRbQldmFDAj4iIyHmcVatWqX7/aJHw9vaW8uXLy+7du+WOO+6QunXr5jku9OHv37+/agnASt14LLQWrFu3ThWy8fwdOnTI83MvvPCC/Pnnn2psB8Z8oJYfgQFfGw30RohAS8rUqVPl0UcfVcdm5MiRI3LTTTep3xXHhdmo0MKB4PPwww/Lxx9/nDMIunLlyurcYCalKlWq5AQgnG8iIiJyPXY3ohIDNeAoyFt/oLbekVCYBb0pV1GQnzdvnpw9e1Z9PXLkyJyCLmr0Dx06pL6/YsUKVYB/+umnVXein376KdfjoAUBAeGJJ56Q06dPq8I2CtZdunRRoUKvsI+AcN9998mxY8ckKipKfUaBHtOSovZfz86dO2XhwoVy5swZVeP/yy+/qBaEWbNmGf7+KLyjBQFdgz755BO752r06NHqPCE84Xjwe+C5EFpwTt55552cblFoZcHsUTB//nx1nvBh3WJDRERErsOQQCUGZiZCDb71h6NaETTaWALrrj7WhXvUnKPW3FqvXr1UYfiGG26QWrVqqa5D7du3zykAo1CvOXnypOzbt0/atm2rCtHaDE4oVM+YMSPPeg+nTp1ShenGjRvLxIkTc7bjMx4f39+1a5cKG7bw+Ndff33O8eL48HHhwgXDsQNw5513SrNmzVSrAwKRHrS2/P3336rLE1oL0J0Jx4DjrVOnjgo0CBB6oYeIiIjcj92NqMRwxOxG9qB2XytsR0ZG5tmu16UHnnvuuZzaeXQ1QphByMDYAdSso7Zfoz0+Cte20H0JMwytXbs253sodEPnzp3zLHKG/3fq1EkOHDig9rMNTAgitrBGAWD8gNECdAgV6I6ELlGY9QkDmW0dPXpUfd64caMKFEYwSJmIiIjMhyHBA5UL8C7iwGUvpw9cLsnQfx+rQmNcgF4B2nYdB0Afewy8RasGgkLr1q1zau7RYoDCvTWthSI1NVX3GGy/rw0w1gYF29K+rzcQ2d6CZRhcbc+tt96qWkMwSPupp57Ksx1jKQDjDIzWqcBzYDA0ERFRaZCcniXHLqfK8cspEp2YLg+2v1oxZ1YMCR6osDMIocYafdYxo45WeKPCQTcgrW//XXfdVeCfwyxEmP7zxhtvlDZt2uTatmDBgjz7169fX4UIBBJ0UcIgaA3GDaBrkd7+K1euVFOuWhfI0WUIg6DxN8d+job1KzBD0XvvvZdnW8OGDVXgQSvCb7/9lue6wzWJAGO9IJ7WEpLfVKxERERml5KRJfN2XpCNx2PlXHyaXExMl4SUDMlKuSIphzdKaMubZHCrSAkJMG9R3LxHRuTmGZJQiEU3ndWrV8v333+vWgVQ+MXMPAWFLj6oscdgXNS8YzpSzDyEAcPo028LswVhXAAGNw8cOFAN+kUXoIMHD6qpS20L0Oi6dMstt8gff/wht99+u7z44ouqzz8GCmNGJYwvuO222wy7DhXHddddp8Y04HezneUIYWXw4MHyww8/qN9j2LBhqgUG5wKBa8OGDWqQN8ZLaDD7EaDlBeEBrQz4zBmOiIjIE2RnZ8vZxHT5Yes5WbIrWuJTcq+JhIBwYc6rkh4dJVnJ8XLscitpXiVvLwSzYEggspkhCR96sEbClClTJDQ0tMDnDK03mOUI04Zqsx1pnn32Wd3+/B988IHcfPPNampT6zUDmjZtqhZ0w8xI1gOnsXbC3r171WxFGCdgDQOlMW7AWV577TUVatBiYmv8+PEqrKA1Ax+2bM8jWluw1sPnn3+uPoDrJBARkdmlZVrkwKUUmbvtnKw+GCNJaVl59slKSZDo2a9KxoWrY/biVn8jkydVkBnv6085bgYMCWRq6NNuveZAUfcp6PNYd30JCQlRU562bNlS1cZfc801OV23rKEWHT+rN+MRvPHGG2o7Wg8w6w8K7li9GAOHf/7551xrJEC9evVUoRrrCGANBXTVwVSumOoUYwFsa+3RWoH1GaZNm6ZaPTCtKmrlUcs/atSoPK0I9o5Xb5t2bvTCEX4HBBNtMLX1zE7Yf/HixarFANOeYjAzfhc8FgZUY5Yka02aNFEtNmhhQWsOzjNbEYiIyKzSsiyy9UySzN12VjZGXVJhQY9tQNDMnfaRNKkUqhYpNSOvuLi4kr3ClgdCQdKRhSOOSRCPO5d4HtvH37Ztm6ptR+FfbzyDJ3HlNeno15OZYCA7usQheGHmLOK5dDdekzyPpeGazM7OllNXMuSX3Rdk7qZTakCyEaOAoEHFISrailPR6SxsSSAyGQx07tGjh5rOFWsqwNatW1X3HdyYCjNwmoiIiBwnPi1L/j2fLHM2n1atB/bkFxBq1KwlS5YsMWVAAIYEIpNBlx2MMXjyySfzbMNibXfffbdbjouIiKg0ys7Olvg0i5xOzJBNJxNkzsYTciHhvzWOjAJC3LyxhgGhdu3aqkuudVdns2FIIDIZjIfAGIOvvvpKraWA2ZCqVaumpht94IEHOI0tERGRkyVlWORiSqZcTM6SS6mZajDy1mOX5Y9d5yTTYr+nfvPyIvu/elsSz0Z5bEDwmJCAueHnzJkj//77r1qRFs0y1157rZotpnHjxro/g+4ZmFEGP4tBmBh0innuMQ2jO/YnKgx0M5o4cSJPGhERkYtaC2JTs+RcUqZcSM6UC4npcuJikhzHR0ySnItLkXyygXSpGy73tQiT5x8eIkcP7vfogOARIQHzqWNGF1uoYZ09e7b66NmzZ65tS5culfvuu0+ysv4bSHL48GHV7wsfrVq1cun+RERERGTeYHA+KVMuJaXLusMXZc/peLl4xX53ImuBvt7ywk11pVs1X+nfv7/s2bPHMCCgnIhZCT2B6ZffRUG8Y8eOai745cuXq7ngcYIRHDA7CqaNwh9Zk5iYKE899ZT6ObQ0bNmyRc0rjwWdrly5ovp5u3J/IiIiIjKPLEu2HIlNk5Unk2TjuRTZdT5Zft56Wj74bb+s3n+hUAGhcaUQmTu8tdxcL7REBQSPaEno3LmzWk3WdroozLOOGntMa3X+/HmpUqWK2oYmHCzs1KtXr1yLYmFxpv3796sVXtFVCN2VXLE/EREREbkfKnHRlWjfpTRJycyW2KR0WXPgghprgOBQWMM6VJMnutUSPx9v9diYolwvJHhiQPCIlgSjOdQxuLNs2bJqJhgs9qT5559/1Od77703z/7aarTaPq7Yn4iIiIjcKykzW7acT5Ft0akSfSVdftl6Sj5aul82RV0qdECoGR4oM+5uLs/2qKMCAqA8+vbbb8sTTzyRa986dep4ZEDwiJYEPUlJSarmHtNE9uvXT62Mq4mKujqSXJtf3hpWztXGD7hqfyIiIiJyD8xEdNErVI5cyJTUjCz5+2CM/HMwRjKy9FdH1hMe5Cutq5eRtjXKqs/NKoeKj7dXnv20oACfffaZCgjogeKJAcGjQsL69etl2LBhaqGp2NhYtVIravM/+OCDXPvFx8erzxEREXkeQ/ueto8r9re3AqC91Wjx4SjaGAl8duTjlkY8l553HvH49l5vngzjsqw/E8+lu/Ga5Hk0k5gUixyMz5QUCZEtRy7Kir3RkpiWWaCf7Va3nHStW1ZaVw2TmuEBKgBoMtLTJMPOz77yyiuqt8uAAQOkQoUKpnoPKsyq076edOO5cOFCzv/xx0KLQlxcXK6WBK3AoddNSfuedaHE2fsbOXv2bK7ZkawhADnjTT8jw94lTTyXrueKaxI354SEBCnJoqOj3X0IJQbPJc+jmfB6LJos8ZIYrzKS4BUo+88mqLUNYgowGBmluM41AmVg4xCpUQZF5FSRpFQ5nVT4Y8AgZpQHMXbWLNA1vjBT9XtMSMAA5oMHD6pCBQYq//rrrzJlyhTZtm2bbNy4MScohIaG5tTmW4cHQKAA6zEMzt7fSNWqVQ234bERFBwFtbU4b35+frmSMPFcuosrr0nUmlSqVElKIlQmoBCB38+R94zSiOeS59FMeD0W3aVUixyIy1LjDhZsOyaHzl/J92d8vER6NYmQYddWUeMNyMNCAt4AtTd69O3C4mVpaWlqbAKmRr3jjjtyRpBj0TUECtuCOL6nzY6kcfb+RWnuwVSqRgO2i0Jr2UBhzJGPWxrxXHreecTjF6Z51RPh/ljSf0dX4bnkeTQTXo+FG3uw/1KanEjIlM1Rl+T3XeckPTP/nh29m1WUx6+rJdXL5X8PvXjxorz00ksyfvx4KV++vJR0Hl1i1Gr1L126lPO9Dh06qM9oabC1YMEC9RnrLrhqfyIiIiJyHiyI9s/pJNl+NlG+XB0lv/57Jt+A0L5WWZk9rLW837dRgQPC7bffLvPmzVOfL1++LCWd6UPCuHHj5Ntvv1VN6tpgR3THmTVrlkybNk39v23btrn6gAUFBcl3330nP/74o6qxRLcGLMb2999/q1aIbt26uWx/Krzp06dLtWrV5JtvvjHc5/vvv1f71KxZM9dYFWuHDh1SLT/4e+iN8UBLVPfu3dXsA1orkD1YwA/LqKM/Hz7j57BWB66Jd955R44cOSKlFc4N/h7NmzdXLWG20CcT20eMGOGW4yMiopIH5UIsirb+dJIs3xstk5YflGMx9gcQ1K8YLFPuaiqf391cmla+Wtlc0ICwb98+9X+shVAagoLpQwIKek8//bQ0atRIde9B4QzdebDqcUpKigwdOjRXSECXpJdfflkNCh41apTaHx8oxKHbwYcffqj6Qbtqfyo8hC4MSrc3eBshDftgQOrcuXN192nYsKE8+uijaoE721mw4N1331UreGNOY1xf+cH1hudMTk5WnzHL1okTJ2T16tUqJGIBvf/9739qBq7SRjs3p0+flqlTp+bZjjCN7divONDN8IYbbijWYxARkedLzbTI5nMpsu1cknz191FZuvOcZGQZr3cQEewnb9xaX34a3ka61itf4PFwF20CgqY0BAXThwRMI/X444+rGmPU/KKWEn30UCCbPHmy+rD15JNPyqRJk1QtMgommN2kWbNmMnv2bFXj6er9ybGwVsWmTZvUi7NcuXKqRcfImDFj1N9m4sSJKixoNmzYoOYwRqHzmWeeKdTzb9myRU6ePKkCAh4TTY9DhgxRNxyMkXn++eeltEJLGiYUcNZNEwGtuEGDiIg8G1ZNXns6WbacjJfJyw9J1IVEu/v3qBUoc+5vJne0qqy7voGRmJgY3YBg/Z5kpulNS93A5Xr16qkaX3wgJOAPgtmDMI2TPffff7/6QKjAvsHBwW7dnxznhx9+UJ/vu+8+CQ8PV93RMJjcukVJg1YdFFpvvPFGFTZXrVqlriO0AgUEBKgua/ldS7bwt8Z4GLQcYR5kBNibbrpJzYd89913q25SDz74oO6CeyUdWlLQyjdhwoScBWWIiIgcISs7Ww5eSpOj8eny94ELsnzPefn/nui6qpYJkBd61pRq3glSJrBwRd6YmBi1YK9RQEDXY6ykbG+2Sk9n+pBgDYU6fBRGWFiYqfZ3hPtm7SzU/tn4Z8kWL28vwT9n+u7+Vk59fHRbQfeiihUrSo8ePVSBHSEBrQl6IQFat26tWn9QcMUH1qg4duyYvP/++9KgQQOHHRuCCFoUMF7ml19+yRUSmjZtqgIvukm99957akYuDLj/559/VIsUmjOxDd8/evSo6meJY8Pjaa0UttCCMWPGDLU/zsc999wjjzzyiPq52267TbVqWEO3KIQitH4gKOEGd+edd8rDDz+cJyhpx4vWMQT03377TbWaYQzGa6+9ps6pHjwexpR8+eWXKohVqVIl3/N24MAB1TKHBRMxjXBkZKT06tVLRo8erVqK4Pjx49KlSxdVSYApkDG+QYPuTbiRExFRyZWcYZF/o1Pk3JV0+WnTyXynNh3ctoo83b2W+GRnyqlTCQ4NCPXq1VMrKZfkgOBxIYGu2nU2/zl/Syq0BJw5c0aNNfD19VXrZ6CLy88//6wKs0Yh8sUXX1QFXYxNwJiHrl27ymOPPebw4+vTp48KCWjZsJaYmKgG3KPZ0rrbkzZ+AauJr127NtfPYED2unXr1LicN998M9e2Tz/9VF5//fWc/6N7zxtvvKG6Ymld4Kx99dVXquuVNvhf+5mtW7eq8ICQZT0VKY4X4z3QOrJ58+ac769cuVJ1t0KBHmNxbOEx0EUQq6FjfM4nn3xi93z9+eef8sADD+Q6XgQF/M5//PGH2o7WGm1MA+B30L62PodERFQyXUzOlO0XUuXIhUSZvfGExCcbL8RZKcxfzVjUrmZZ9f/U1MK9R2hdjPbv31+qA4JHjEkgsqaNPxg0aJD6jBr2u+66SxUsly5daniyEB4wrzECAn4GY1mcsYgXuh4BWgZsIRygJnzOnDlqNiUM8sXNRhsQjxp6FMJR0EftOqbZRc09asq1hfoAP4euPOhKhZ9BINm9e7cqkC9cuDDP8547d07N64zf99lnn1WFftSOoBUCLRAojGtduKxhUDe60/3000/qeLFo4S233KK+h9YbI2jFwFgPtIygBcAIQgjCHgLA2LFj1XGhVQSzhOFviqCAoAGYSQq/N1omMMgcX2sfbEUgIiqZUCkUFZcmG88ly597z8vnq47YDQhd64bL3OFtcgJCYTEg5MaQQB4DBWW0BtSvX1/atWuX8/3Bgwerz3oFXdvuOdpNB7UAzqDVxlvX2FtvwzGiKw1CAcY1aEEFLQOYMQu1/ViLA4VsdDNCTQaCjfUUregDidpzzMqELjnalKwYB2Hb4gD4XTFT1PDhw1XrA2Z9Qg0Izhu6BsH8+fPz/BxCCM7ZzTffrI63cePGarA3WLeG6EGhH8eNrlVG8Hvgb4p9n3vuOXVcWJymZcuWqqsUnk/7O+E8aecL5xFfax9oUSIiopK3OBpaD7acSZKZa47Kst3nxWIw/gBjkZ/qXksm39VUwoP9nBYQlpTwMQi2+O5KHgMFWXRLQS2zNRQm0f8ftfAYb6D3AkYrA1ohMMAY6xmg8Nq7d28VOBwJNdugtxIjWhn0plpFYRrdlHbs2GH4uNYz+qClAVB4t6X3vcOHD6vPeA5bmE4U4zr01njA2AZ05bKGlgeMw9FbC8Ea1p/A+hToBobZo0JCQvLsg9YSQAuP1mKgBSx8aL8zvnZGqw8REZlTYvrV8QdbT8bJz5tPSlJaluG+FUP9ZdztjeSaIrYeaAGhb9++Oe9LRgGhSgHG2ZUkDAkeqGXVMNMOXHZFVyMMPsZAV2sYiKsNakaXGmsYIIyCKgruqAlHoRk3A9TEIzxY98UvrmXLlqnPegN7K1SooPszqC1HQGjfvr3qz48adRSqcVz4XT/++GPdn9Oblclerbre74nCt9HPGM3YhZ/Raymxha5QGMyNNUQwSFwvHGnjH+xBCwtbC4iISj68t5xJzJSd0cny265zsu5Q3q671jrWLifv9W0oESH+TgsI9evXV+/TpS0gAEOCByrsDEIoPKO7CdaXcGSB2JXQ3Wbbtm3qa3tzEiNI2IYEdMnBIOCZM2eqbjP4QNccDOZFtxZHDWDG2ACtrz6aLAtKq8V/9dVXc63WjZsl1nMwGveAvvsIFtbwPVtYfBDQ0oIafmsYB4CxAc6YrhVdptBagyBm2/oD2ngMBDesWm3EOiDg+kVoICKikiXDki17L6bK7vNJMnvDCTkTa7wmDroXPdqlpjzcuUah1j2whbIB3q8ZEPR5ZomRSh1tvAGm1bQetGr9ga48aCWwno0Hg24xABiz9OBDg7776MePAcD2BtfmB+EL/fPxOBiwiwCDgbS2hXd7tOk8MSsSajQQDrBQG1o6MIuQLRS8UZuPVgZMUarNZvT777+rlghb6GaEVgdMf4pQhFmWcNxr1qxRU6YWNtQUBoIPCvZ6YxNQc4PWCoxJwIBr/A4YY4BjPXXqlBqwjUXwrGGmI2zDFLZERFQyxKVmybrTSbL8wEWZ/OchuwGhcpkA+WpIC3msa00GBCdjSCDT0vqho+YY3YhQ2ESh1nrQqvXHiBEjcgUKzOqDxb3QcmDbZQf7o8sSCthYQ6Eg3Wc0WO0bAQNNj5jTH7X/eHz0oUfBF4XbwsBAZnRFQv99jANAv3/MaoTfGcFDrwYeYQnHPnLkSDVuAOMwsE4Cau+1QccarMOA84BggM9oWcCxI8wgIHXs2FENanYGrLeAtRO0cRHW8HfBgG2Mb0AgQpMuzgN+l06dOqlggelurWHKW5znNm3aSOXKlVXAQggkIiLPnb3on1OJMmfzaflh/QlJy7AY7t+zYYT8NLy1tK1RttjPi8lB7LUgLCmFYxBsMSSQ6WjdibQ1D/766y+Jjo5Wg45R4DWCQjIG1S5YsEAVJFH4x+w5CANYmdkWFmPDqs1Y0AyrJBcUpjFFAR3PERgYqAZODx06VPVZxLSfeoN07YmIiFCDslEAxuMBCvtYkA0FeD3o54/WCwQE3OywCjkCg9aSgBp3ay+88IJqRUD4QE09ghd+Fl2xcL7QFc1ZMP2qdWixNnDgQPX3RWDB3wizNqF1oVmzZmptC7Qy2D4WWj2wD64T/B24TgIRkedJyrDI5nMpsuHkFZm28ohsOGI8/iDA11tevaWefHxHYykbVLTZi2wrId966y3d92stIFSuXFlKO6+4uLiCV6GSS6DLCWqTHcXTxiSgsItuNFhPALXsGOCKgckIDUaFTQ0K7igAoxCJwjxuBPYK7dgXP4OCc1BQUL6PjWPBB44D++d3PBoUZnHu83sO/K1A+ztpvzt+Tm+gsraPdhxY/wEFa8wYhDUIjJ4Dv7e9Y7d3vHrbtPOOFho9+FvgeTG+AEHI6Jq0/l3yoz0nHs/ewGZHv57MBEEJ3a/QsqUFTOK5dCdekzyPBZna9EhcuhyPS5ddp+Nk/pZTkmqn9aBuhWD5oF8jaVAxxOHXJLr0YsyctkAnWvNR4ceAcBUHLpNpoMCHfvKLFi1SBUd0KQEUGgtacLQuuBoVWK2h4F2Q/bTHRlApSuAqaOuC7WMa/e4Yf4DB3Bjwi+5DWD0Z/frHjRunfieMW7D3HPkdu73j1duWX/gxminJVkH/zgV5TiIiMg+0ep9NypQDl9IkJdMif+45L6v3X7D7M7c3j5SXbq4nwf76lWTFhRZ8rAeEoICurgwIuTEkkCmgywm6C2nTYqIbCvqskz7MSPTGG2+oD637kAYLk6HGhIiIyAwS0rJk78U0iU3LkvRMi8zbfFL2nI433D/Q11tevrme9Gvp/HIAggK63WLmQLYg5MaQQKaAfuUICOiXjlmI0N+ejN16660qDKD14OTJk6qGBmMjHnroITW9KxERkbvhvelYfIYcvJwm6NuekJIh3607JqcvG89eVDciSD7s31jqF7F7UVEUZkbC0oQhgUwBg5LPnz/PftUFhIHKGHuAD7QiYOyFJ4w3ISKi0iEt0yI7Y1LlYsrVlu5zcSny7dpjEp98tceAnj7NKsort9R3WPcibZwfFQ1LFWQK6DLDgZdFP3cMCEREZBYXkzPlnzPJOQFh/9kEmb7yiGFA8PX2ktd61Zd3+jR0WEDArIjXX3+9rFq1yiGPVxoxJBARERFRsVmys1XXos3nUyQ9K1syMi2yZMcZmbX2mBqLoKdsoK9MH9xMBraunLM+kiMCAtYtwmKnw4YNkw0bNjjkcUsbdjciIiIiomJJybDI9gspEpd2NQyciU2WnzadlAsJaYY/UzM8UD67q5nUKu+42erQdRnr6Rw6dEj9H9OIjxkzRk2FjfF8VHBsSSAiIiKiIotJzpS1Z5JUQMiyZMuqfdEy9a/DdgPCtTXLyvf3t3JqQNBg6nK0KPz9998Oe67SgC0JRERERFSk2Yui4tLlUGy6+v/FK2ny0+aTcupSst2fu6NlJXnllnri5+Pt0ICALkaHDx/W3Y4pThs1auSw5ysNGBKIiIiIqFAysrJlZ0yKXEi+Ojh527HL8uu/ZyQjy3hGIX8fL3m2Rx25p10Vh40/KEhAqFOnjsyfP5/rLxUSQwIRERERFVhCepb8ez5FkjOzVfei33edlXWHLtr9mSaVQuTdvo2kXoVgh57p/AJCgwYNZPLkyRIZGenQ5y0NGBKIiIiIqEDOJmbI7phUycoWSUnPktkbT8jh81cM9/f2EhnRqYY82qWGQ7sXFSQgoHvRzz//LKmpqQ593tKCIYGIiIiI8u1etO9SqpxJzMwZf4CpTWOu2J+96N0+DaVltTIOP7sICH369JEjR47obm/cuLEsWrRILT566tQphz9/acCQQERERESGLiRnqtaDNDQfiKiWgx83nJDUjKvjEfQMalNZjT9w1OJo1s6dO6daEPILCOhixFaEomNIINOZN2+ezJgxI+f/GNwUFBSkBh7hpnDjjTeKWVy8eFHuvvtudVxPP/20uw+HiIjIYTIs2bL/Yqqc/v/WA8xmtOHIRfltx1mxXM0LuoOT3+zdQHo3c84YgMIEBCoehgQyHTQhbt26Nc/3Mb/xt99+K8OHD5cJEyaIGWCRFhxrixYt3H0oREREDl37AK0Hqf/fepCSnikLtp6W3afjDX+mYqi/TBjQRFpUDXPKX4IBwbUYEsi0nnjiCenXr5/6+vLly7J8+XL56quv5Ouvv5ahQ4dKu3bt3H2IREREJc7x+HTZd+m/sQZHLySq1ZPjUzIMf6ZZ5VCZMLCJVAoLcFtAWLx4sVpZmRyDIYFMCwufXHvttTn/v+WWW+TSpUuycOFC2bBhA0MCERGRA6E7ERZGwwJpkJllkb/2RsvfBy6IQe8ipVeTCqqLUaCf48cfaAEBg5SjoqJ0tzdp0kR1MWJAcCyGBA900003FfpFjw/07ddbvGTAgAEycuRIu48xbdo0+eWXX/J9rj///FOcqXr16rrfz8zMlN9++01Wrlwpx44dEx8fH1WrgBaHZs2a5do3KytLBY6OHTvKG2+8oVon/vrrLzW4CaHkySeflPDw8DzPsXHjRvnmm2/k+PHjUq1aNXnggQekXr16hseKZeGx/759+9RzYq7me+65J1fwsT2esWPHqvEY6FqFv1WvXr1U9ypvb285ffq0+jvs3r1bQkND1e/Wu3fvIp9LIiIijSU7W/ZcTJPTVzJyZi+as/GEnIlNsXuSnuxWS0Z0qu7QxdFsvf766wwIbsCQ4IG2bNni0MezLbTqwfRhjn7ewkKhGysm6h3zbbfdJps2bcr1vVWrVskXX3whX375pfTv3z/n+whMGEeAIIBCu3Ww+eeff+T3339XYQODpTWzZs1SA5PxsxqEpldeeUX3WHGcjz32mGRkZOR6bHSVQjCxHuSsHU+5cuXkzjvvVPtpcGyoOUEgwe+IgdKapUuXynfffaeaX4mIiIoKC6Jtv5CqZjFCWNhw+KIs231OTXtqpFyQr2o9uL5BhNNP/EcffaQqAG3LIU2bNpVff/2VLQhOwpBApvXZZ5/JTz/9pL6OjY2Vo0ePqgL1gw8+KB06dMgzgBg16126dFHdlPD/PXv2yKRJk+TFF19UYxtsazkQItA0+f7776sbDQZMjx8/Xvbv368WX7nvvvvUfqjBf/7559XX+B5actBSgb6P48aN020WxXgKBISBAweqgBIQECArVqxQgQUhoXPnznmCzurVq6Vy5cry8ccfS/369VUoevvtt+Xzzz9XoQUtER988IFERESo8RlTpkxRx86QQERERZWelS3bzidLbJpFtR7M33JKjl9MsvszHWuXk7dvayCRThp/YAtrHaDyDe+pWlBgQHA+hgQyrZMnT6oPaw899JCqUbCFcQroMoSa9zNnzqiQgECBbjoo/J84cUJq166d62d8fX1VF6W6devmfA/TrN58883qJqSFBAQGPN7jjz+uCu3p6eni7++vavaDg4Nl5syZeaZwTUlJUT+PpeA1eFx0l0KXIrRM2IYEPz8/WbJkSc5xdu/eXa5cuSJvvfWW2oZQgmPWtm3fvl3Wr1+vnsu61YOIiKggUjItsuVciiSkZcn6wxdl+R77rQd+Pl7yVPfaMvTaquLtxO5F+QWFpKQktiC4AEMCmX52I4w3wJLrH374oaqJb9u2rQwZMiRnP2xHd5w1a9YYPhYK27ZatWqVKyCANpUpZlPSYEwBYD0EW/iebUhACwZoIcMaWjsQErTHtD0e2yCDwViAsQdaQNCgFgUhAS0Xtr8HERGRPckZFtl0LllOxqbK/M0n5cSlZLv714kIknG3N5LGlULddmK1oIDKugoVKrjtOEoLhgQPVJAxBIUZuFyjRo18HwP7FPZ5HTm7UadOnaRbt26qOxG6D6HQjD78gBsGAgK642AcAArMISEh6nedPn262m49lkCj/bw11NiDxWLJ+R5uRtrNyVbZsmXzfE/bX+/xw8LCVFclvRUg9R5LOx69bVpoQEgiIiIqbEBYc/CiLN5xRjLttB7A4LZV5NketSXISbMXFYbeezE5B0OCByrsDEIo8GpdZND9pigw+1F+MyA5G2rZMeD3vffeU+MVXn31VfX9HTt2qM8YpNy6detcP/Paa68V+3mrVKmiPqN7j22g2rZtm+H+mzdvloYNG+ba9u+//6rZjDA7EhERkTsCwoYzSTJ/2xlZd/i/yTD0VCsbIG/0biDta+Wt9HI0tOCjcq2o5RRyPP4lyKOgpQC16hjMGxcXp75Xvnx59dl6ViCMIXjnnXfUegrF1bNnz5zAYd1NCOEEYxSMpqjF8yMUaDAzw7PPPptrHyIiIlcGhNUnr8iMNUfzDQh3t60iP49o65KAgLGEN954o4wePTpXSz65F0MCeRQ0Mz7yyCOSkJCguhIB1hhAFx709Uc/fXRJwvoFmCXommuucUhIuO6669QganR5wnoGGBfRo0ePnFYDazfccIMapIwB0/jZli1bquPAx969e9UxYmwCERGRKwPC0iNx8unyw3LwXN5xeprq5QLlqyEt5KWb60mwv49LAgJm6cMMhlhbiEHBPBgSyOOMGjVKLSaGkICwgEI4uh+hheHs2bOqII5Zh7AoWdeuXYv9fBjb8MMPP8gdd9yh/o/F1PCBIGA9e5G1b7/9Vh5++GEJDAxU4QLLyONxsHAdZjDibEREROQqSRkWmbM7Rib8cUjOx+cdE6e5p10VmfdgG7mmZt5xcM4KCFhJGQFBg6Dw3HPPsUXBBLzi4uLsj1Yhl4uJiXHowiCOGJPgSqiBx+JttWrVksjISN19sIZAfHy8GqysDRDGugTo0gMYvIyBvVjjALP/YNVlBAcNpjjFz+HnbWEbFlrDWgV6fxsEBExlilYEnNedO3eqv5ftzESAAcq4+WEcArZj4LIeo+PB74jfFYO4K1WqlGsbwkd0dLSakQlhxJO48pp09OvJTHB94bWCsTKedg2YDc8lz2NJvR4vp2bK5xvPyrzNpyTTol/k8/fxkrduayi3NnXdvVILCNr7ti30DkBYKC6+touOIcGESntIMDOeS887jwwJVBAsSDgGz6O5zuOJ+DT54K9jstbO+IPywX7y6cAm0rKa62YNQgUeuhgZBQRU7C1atEgtHlpcvCaLjrMbEREREZUgmPZ727kkeff3w3I0xnj15PoVgmXyXU2latnAEhkQqHgYEoiIiIhKiKzsbFm0/5JM+CtK4pMzDPfrWjdcPujXSEICfF0aENDFCN129TRv3lytpMyAYA4MCUREREQlQHqWRT5bf0Z+2HDScPwB3HtNVXnuhjri4513gVVnYUDwPAwJRERERB4uITVTXl56RP45ZDz+wM/HS964tb70aZ57IgxnY0DwTAwJRERERB7sRFyqPDN/v93xB5Fh/jJpYFNpUjnUpceGAdgYg2CvixHGIGgLo5J5MCQQEREReaiNJ+LlhV8PSJyd8QdtqpeRCQOaSHiwn0uPjQHBszEkEBEREXmgH7adk09WHLU7/mBwuyryv551xdeF4w+0gIBByidOnNDdzhYE82NIICIiIvIgGVkWeWf5UVm487zd8Qdje9WXfi1cO/6gIAEBi4BiFiN2MTI3hgQTz3Hs5eXa1E9UEl9HREQlbYDyEz/vk52nEwz3iQjxl8l3NpFmVcLE1RgQSg6GBBPC6opYITAoKMjdh0Lk0fA6Ks5qpUREZnI6LlUenbtHTsemGu7TpEqYfDawiVQI9Rd3WL58OVsQSghvdx8A5RUSEiKJiYmSkpLCmlCiIrYg4PWD1xFeT0REnm7H6QQZ8s0OuwGhV7NImTW0hdsCAowYMUJee+21PN9v2bIluxh5GLYkmJC3t7dabTApKUkuXjSe77igLBZLTo0qHpt4Lt3NFdckHhuvI17zROTpft8XI2N/OyQZWfpdKH19vOTR62rLwx2qmqKr8ujRo9Xnt956K1dACA8Pd/ORUWEwJJgUCjZhYWHqo7hQGEtISJBKlSqx6wXPpSnwmiQiKlir6Ix1J2XqPycN9ykT5Cev3tpAbm4QboqAYBsUFi5cyIDgoRgSiIiIiEwGrQZvLT8uS/dfMtynaniQvN67oXSs4foBygUNCqNGjWIFpYdiSCAiIiIykUtJGfL637Fy6LLxAmlNqpaR52+qJ20rB4uZcfIIz8WQQERERGQSe85dkWfm75OYROOA0LVhRXmocw1pXSnILV2MsP6Br6+vVKtWzeXPTa7DkEBERERkAkv2XJA3fz8s6QYDlLFo8u1tqkm/VpWlVWSg2wICFkrz8/OTxYsXMyiUYJzqhoiIiMiNsizZ8vHKo/LKkkOGASHAz1seuK6u3NaikgoI3m4MCFgw7ejRo9K3b185e/asy4+DXIMhgYiIiMhNktIy5bG5e2TWZuPCdvlQfxl5QwPpUKuctK0UJD5uCAjHjx+X2267TQUEDYICQgODQsnEkEBERETkBjGJ6TL0+12y+US84T71K4XK4z0bSKPIYLmmSpD4os+RGwICwsDp06fzbENQeOWVV1x+TOR8HJNARERE5GL7o5Pk6fn7JDohze4A5d6tqkij8gFSp5y/21oQjAICtG7dWiZMmODy4yLnY0ggIiIicpEMS7asOhov7y49KHHJ+jMYobXgjmuqS6/GEdIsMliCfN3T8SO/gNCmTRtZsGCBlCtXzuXHRs7HkEBERETkAheSM+X3g5dlxuqjaiyCnrBAX3moW21pE5Ymjcr7SiADArkJQwIRERGRE2VnZ8v+y2my8nCszFp7TNIyLbr7lQ/xl3f7NZbWFf3kjEHtvSuwBYGAIYGIiIjIiQFhz8U0+ePARZmz4YRkWvSnOK1WLlCmDW4utcIDJTU11W1/DwYE0jAkEBERETkpIOy4kCKzNp2RlfuiDfdrXDlUPh/cTMoG+bn178CAQNYYEoiIiIgczJKdLetPJcqkVUfl4Lkrhvt1rF1OJgxoIsH+PqYOCG3btpVffvmFg5RLEYYEIiIiIgcHhMUHYmXiiiNyOTHdcL9eTSrIO30aip+Pe5etOnbsmFo9mQGBrDEkEBERETlIliVbpm86I1+vPSkZWfoDlGFQm8ry4k31xMcNi6NZY0AgIwwJRERERA6QnmWRt/88Jot2nDPcB6HghRvrqpDg5YbF0QoTENq1ayfz589nF6NSiiGBiIiIqJiupGXKW8uPyfK9xgOUywf7yScDmkib6mVMcb6TkpIkOTlZdxsDArm3ExwRERGRh4tJypCXl0bZDQjNq4TJ3OGtTRMQoHnz5vLrr79KeHh4noDAQcrEkEBERERURMfj02Ts70fk74Mxhvv0b1lJvhnaQiLDAkx3nlu0aJErKGgBoWzZsu4+NHIzhgQiIiKioiySFpMi7yyLkg1HLunugxEHY26sK2/2buD2GYzsadmypSxcuFBuuukmBgTKwTEJRERERIWcwWjb+WSZuua4bD8Rq7sPxiQjHPRrUckjzm2rVq1k3rx57j4MMhGGBCIiIqICyrRky/rTSfL5P8dk96l43X18vETe69tIejWtyPNKHsu8bV9EREREJpJhyZbfDsXKe0sPGAYEX28v+fiOxqYJCFFRUbJo0SJ3HwZ5ILYkEBEREeUjPStbvth8VmatPyGpGfqLpPn5eMmnA5tKl7q5ZwtyZ0DAOgjR0dHyxRdfyIABA9x9SORBGBKIiIiI7EhKz5Kxv0fJiv0XDPcJ8PWWyXc2lQ61y5kmIPTp00fOnbu6sNvDDz+sFm+744473H1o5CEYEoiIiIgMnIpLlafn75eomCTDcxTi76MCQruaZU0ZECArK0seeugh9TWDAhUExyQQERER6dhwLE7u+WaH3YBQr0KwzB7W2tQBwToofP7552Kx6HeXIrLGlgQiIiIiG4v3XJA3lh5WsxkZ6d2sorzWq74E+fmY4vwdOXJEjUHQCwhw7bXXyty5c8Xbm3XElD+GBCIiIiIrMzeelkmrj4tRPMAMRs/fWFcGt6ms+vl7QkBo3769/Pzzz1KmTBmXHxt5JoYEIiIiov9fRfmDlcfkxy1nDc9HRKi/TLijsbSqZp7CNgICuhidP39edzsDApXYkBATEyNLly6VAwcOqK8jIyOle/fuavlw2yaz5ORkeeyxxwwfa/To0dK6des834+Li5Mff/xRdu/eLb6+vqpJbtCgQRIYGKj7OIXdn4iIiMwrI8siLy05JH/uv2i4T/NqZeTTAU2kQoifmAUDApXakPDBBx+oj8zMzFzfnzp1qnTq1Elmz54t5cr9N91Yenq63UVDhgwZkud7Bw8elH79+uVK4N99951Mnz5dFi9eLBEREcXan4iIiMwrOT1Lnvh5n2w7qb9AGtzQuKKM69tAAny8PSYgdOjQQebNm8cuRlQyQ8KxY8ekfPny0rt3b2nQoIH6es+ePfL111/Lhg0bZOzYsTJ58uQ8P9exY0cZOXJknu+3adMm1/8xwn/48OHqBYbmuKFDh0pGRoZadGTfvn3y9NNPy/fff1/k/YmIiMi8LidnyCNz9sjhC8YzGN3Zrqq8dENt8TVRQDh8+LAag8CAQKU2JLzwwgvy2WefiY9P7pkDbrnlFrn99ttlyZIluiGhatWqqrY/P3/++acq3Ddv3lx+++038fO72oQ4cOBAlcDx+EePHpW6desWaX8iIiIyp9NxqfLw7N1yNj5Nd7u3l8hD19WWkZ2qibdJBigXNCBgkHJYWJjLj41KDvNEYgO1a9fOExCgW7duEh4eLmlp+i/sgvrrr7/U50cffTSnwA/owoRWAli+fHmR9yciIiLz2X8+Ue79dqdhQPD39Zb/3dJARjEgUCll+pYEI5jiKyEhQW644Qbd7SdPnpSXX345z0Bn26nKMBga0HXIlvY9bZ+i7E9ERETmsul4nDw9f5+kZOgvKhYa4Cuv3tZQejUIN80Up1oLAsYgREdH625HV2uMQWALApXakIApysaMGaNmNnrllVd099m6dav60EyZMkW9eDDQGS0QmkuXLqnPlSpVyvMYlStXzrVPUfY3kpqaKq6CwdzWn4nn0t14TfI8mg2vydJzHv88eFleX3ZMsgwWSSsf6i9v9q4v7asGF7u3giPP44kTJ1QXowsXLuj+DCoqMSYSvRxcWcYwO0+4Jl2pMLNwemRIQAsBxgNMnDhRdzrTsmXLyoABA6Rly5YqSOzdu1fmzJkjGzduVNOjYrVBjXbR+Pv753kc7XvWF1Zh9zdy9uxZtTy6KxnVPBDPpbvwmuR5NBtekyX7PP52JFm+3nnFcHu18CB5rktFqZJ1SU6dyr/Cz5XnEZOkYLyjXkho1aqVmgkyNjZWfZD9c1la+fj4FGrMrEeFBMwshBaEmTNnyvjx42XYsGF59gkNDZUdO3bkai0AhAN0TVq2bJnqilSzZk31/eDgYPU5KSkp52sNvme9T1H2N4KB1a6C0IIXB1o/9MIN8Vy6Gq9Jnkez4TVZss8jeiB8tems3YDQoFKovN6rrjQI9zfteUSFJ8o+q1evzjVI+YcfflDlHyr4uaQSFBKQoDFYeMGCBfLRRx/JQw89pLsfFjazDQiA5IQxCeirhz59WkhAYX3Xrl1qqtWKFSvm+hl8D6pVq5bzvcLub8Qdi67hxcHF3nguzYTXJM+j2fCaLHnnEQFh4urj8s2mc4b7tK5ZTsbeWl/qh5vjmI3OI75Gt+l7771XVq5cqdaL+umnnzgGoQjnkkrA7EbaKsp33323CggTJkwwDAj5uXz5svpsnSS1dRNWrFiRZ39Md2q9T1H2JyIiIvewZGfLe8uj5JtNZwz3ua5RRXnztgamCwhGgoKCVMvBM888w4BApTskxMXFSf/+/WXVqlUyadIk3S5G1n755Re1joE19P2fMWOGKtgjRVqPY8BaC5i5AKslW89KtGbNGvVYaL678cYbi7w/ERERuR4GJo/97ZD8tF1/LQHo3aqqvNCzjtQuGyCeBEHhjTfeYAsCle7uRs8++6xs3rxZzRyEmnqttt4aZi7Spvv6448/5MEHH1TrK+ADUJjXFhzBmAbrqcGaNGki99xzj/z4449qmlQ03aH/2qZNm1S4wP4YCF3U/YmIiMi1MrIs8sKvB2XFIf3Bx5jVdMA1NWT4tVWlRpn/1jwiIg8KCVgLAVDIX7Roke4+6IKkwcrHR44ckX///VeOHz+e832MHxg9erSMHDkyz89/8sknqs8iZj3SBgNhCjE05eGjuPsTERGR6wLC0/P3y7qjsYarKN/dsaY8cE0VqRxiroBw8OBB1fPhzTffdPehEJk/JKBm/v7777e7j3XLwC233KI+zpw5o8JCSkqKGmzctGlTNahZD7ogTZs2TcaOHaumS8UUURhXoDcAuij7ExERkWsCwjO/GAcEH28vub9LLXmgXRUJD/QxXUDQ1kHAQrCY7p3InUwfEtCdpygww1BBZhmyhjBRmKlJC7s/EREROS8gjF5wQNZG6QcEPx8veahbXRnappKE+ptrSCa6RWPMo7YGwuLFi1Ul59dff80ZechtzPUqISIiIiqkTEu2jFl4QP4+cnUWQ1sBvt7yRM/6Mryd+QOC5q+//pJRo0ZJZmam246NSjfTtyQQERER2QsIzy88IKsPGwQEP2957qYGMrB5BfHFgASTBQR0MUL3Ij34flpammF3aSJn4lVHREREHjvN6QuLDspKg1mM0ILw/M1XA4I3pjTyoICAsY5YDyEkJMTlx0YE5mpzIyIiIipgQHhx8UH568BF3e3+vt7ynIcGhI4dO8rEiRMZEMitGBKIiIjI4wYpP7fwgCzfbxwQxtzcQO5qYb6AsH//frsBoXPnzqoFITg42OXHRmSNIYGIiIg8RmJ6pjz60z5ZZdDFyM8HYxDqmzYgYJCyUUDo0qWLzJs3jy0IZAoMCUREROQRTsWny8Nz9sm2E3GG05w+d3N9Gdyyose1ICAg/PTTTwwIZBocuExERESmn8FoR3SyvP/HYTkSnai7j6+Plzx7Y325u2VF8TJpQLh4Ub97FAMCmRFDAhEREZlWSoZF/j6VKFNXRsnxi0mGXYz+d3N9ucsDA0LXrl1l7ty5bEEg02FIICIiIlNKyrDI6uNXZNqqKDl1OdlwmtP3+zWWng3Ki9ns27dPjUFgQCBPxJBAREREpnMlPUv+Ppko01cbB4Rgfx+ZOqiptKleVjwtIFx33XUyZ84ctiCQaTEkEBERkanEp2XJ2lNJ8sWao3Lykn5AKBvkK1/c3VwaVQoVs4mPj5d+/foxIJBH4+xGREREZBqxqVmy9nSizPznqByL0R+DUCHUX74d2tKUAQHKli0rr7zyiu42tiCQp2BLAhEREZnCpZRM2XgmSb5dd9xwFqNKYf7y1ZAWUiM8SMxs2LBh6vMzzzyTKyBgkDIXSiNPwJYEIiIicruY5EzZeDZZfthwUg6eu6K7T4UQP/niHvMHBOugMGHCBPU1AwJ5GrYkEBERkVuduZIhO6JTZO7mk7L3TLzuPuHBfvL5PS2kVnnPCAia4cOHS2RkpPTo0YMtCORRGBKIiIjILbKzs+VofLrsu5gmP285JTtP6q+kXDbQVz6/u7nUqxAsnui2225z9yEQFRpDAhEREbklIOy/lCYHL6XK9+uNxyCE+vvI9LubS8PIEDEbi8Ui3t7suU0lE69sIiIicqksS7Zsv5Aqu6KT5fNVUYYBIcjPW6YObiZNK5tvFqM9e/ZIt27d5MiRI+4+FCKnYEggIiIil8mwZMuW8ymy61yiTFtxWM7GpejuF+jrLZ/d1UxaVStjyoCAdRDwuW/fvgwKVCIxJBAREZFLZGRly+ZzyfLv6QSZvvKIxCVnGK6kPOnOpnJNTfOtpIxggJWUL126pP5/7tw5BgUqkRgSiIiIyDUB4XyyrIu6LF+tiZKU9CzDaU5nDmkhHWqXM21AuHz5cq7va0Hh+PHjbjs2IkfjwGUiIiJyekDYdC5JFu+Klj92nTPcr3b5IJk6qJlUKxdour/I7t27VRcj24CgadSokVSqVMnlx0XkLAwJRERE5NSAsOFskny7/qRsPaZfwIbW1cLk0zubSrkgP48LCNdff73Mnj1bgoI8aw0HInsYEoiIiMhpAWH1yQSZtuqYRF3Qn8EIejQoL+NubySBfj4eFxCwSNqPP/7IgEAlDkMCEREROSUgLD0cK5+tjJILCWmG+w1uW0VeuLGu+Hh7me6vwIBApRlDAhERETlUela2/HogRqatOipJaZm6+yASPN+zjgy5pqp4eZkzIGCQcmxsrO52tiBQSceQQERERA6TKr7y/e5Y+eqfY5KZlW24SBq6F13fIMKUZ37Xrl2qi5FRQLjhhhvkhx9+YBcjKtEYEoiIiMgholMssvKCv8xabxwQKob6y+Q7m0oTE66iDAwIRFcxJBAREVGxZGdny+HYdFl6IE6+W3/cMCA0jAyRz+5qKpXCAkx5xhkQiP7DkEBERERFlmnJlp0XUuWfo7Hy3dpj6v96utYNlw/7N1arKZvRzp07pX///oZdjHr27Cnff/89uxhRqcGQQEREREWCQLDpXLJsORFvNyD0aR4pb/VuYMoZjAoaEDAGITDQfIu8ETkLQwIREREVmiU7W7ZHp8iWE3Hy3drjhgGhX4tIeaN3A/E24QxGYLFYZNSoUQwIRDa8bb9BRERElN8YhL0X02TJ7mj5FrMYeWhAAG9vb5k1a5ZUrVo1zza2IFBpxpBAREREhXLwUqp8tvqYLPz3jBjkA48ICJp69erJ4sWLpUqVKjnfY0Cg0o4hgYiIiArsQEyKvLzooGyMumS4T5+mER4TEKyDwpIlS1RQYEAg4pgEIiIiKqBtZ67I8wv2y6XEdMN9bqgdKK/cVNujAoJ1UPjjjz+kUqVKHKRMpR4HLhMREVG+/jx4UV5dckhSMyyG+zzcsarcXDXTIwOCplatWu4+BCJTYHcjIiIisuuHrWfl+YUHDAOCv6+3fNS/sTzUsap4mTQgGE1vSkT6GBKIiIjIcBajSWuOywd/HZVsgwHKESF+MmtoS7mpcQXTnsUdO3ZI27ZtZebMme4+FCKPwe5GRERElEdGlkXe+eOILNx9wfDs1KsYIjMGNZWKYQGmDghYKC0uLk5Gjx6tvvfggw+6+7CITI8hgYiIiHJJTs+S//16QP6JMu6i07FueZlwRyMJ9vMxdUDo16+fxMfH53yPQYGoYNjdiIiIiHJcTs6Qh2fvthsQ+rSqIpMHNPa4gGAdFJYtW+aW4yLyFGxJICIiIuV0XKqMnLtHTsam6p4Rby+RQe1ryFNda6jByma1fft21cVILyDAzTffLNdff73Lj4vIkzAkEBERkZyJS5Vh3++SGIM1EBAK7u1US+5rW0lC/Dw7IHz33XcSEGDecRREZsCQQEREVMpdSc2UJ3/eZxgQQgJ85YHr6sitDcMlIsjXYwPCLbfcIrNmzWJAICoA877SiYiIyOkyLdny/K8HJOpisu728iH+MrxbXbmmWqjULONv2r8IAwKRYzEkEBERleJ1EMb9GSUbjsXpbq8aHiTDrqsjdcMDpXFEgKkDAgYpJyQk6G5nCwJR4TEkEBERleKVlOdtP6+7rXLZQHn4+npSIdhPWlcKEm+TrqT877//qi5GDAhEjmXekUdERETkNKsPX5KPVhzT3RYWeHUMQliAr7SrHCR+mNbIAwNCr169OAaBqIgYEoiIiEqZ/ecT5cVFByVbZ5ufj5fc37WORIT4S7tKgaadyWjbtm35BoRvv/2Wg5SJiojdjYiIiEqRk7Ep8sS8vZKSYcmzDe0FgzvWktoVguXaysESHuhj2oBwxx13MCAQORFDAhERUSlxNj5VHvpxt1xMytDd3qtlFWlbs5xcWyVIyvgzIBCVZgwJREREpcCFK2ny8Ow9En1Ffy2Ea+uWl5ubRUqHqsGm7WIEWOcgvzEI/v7mnaqVyFOY9y5AREREDnEpKV0FhNNxqbrb61cKlXs71JTO1UJMHRDg448/Vl2NbN16660MCEQOxJYEIiKiEiw+JUMenbNHjl9O0d1eIyJYHu9RV7pUDxF/H3POYmTN19dXvvjiC/X1ggULcgICBimzBYHIcRgSiIiISqjEtEx5bO5eORyjv5py1XJB8syN9eS6mmGmneY0v6CQmprKgEDkBAwJREREJVByepY8MW+f7DufqLs9skyAPH1jfeles4xHBQTboGCxWNiCQOQEDAlEREQlMCA8/tNe2X5af4BvRKi/PHVjA+lRJ0yti+CpEBSIyDnMPTqJiIiICiVJdTHaI/8aBIRywX7y5I315aa6ZSTAx5zFgL1790pKiv4YCiJyDXPeHYiIiKjQrqReHYOw88wV3e1lgnzliRvrS6/65STQ15xFgC1btqiByEOGDGFQIHIjc94hiIiIqFASUjPl0bl7ZNdZ/YAQGuAro26oL7c1CJcgEweEAQMGqHUQVq1axaBA5EbmvEsQERFRoaY5fWT2btl7Tn+Qcmigr4zsWV/6NiovwX7mDghXrvwXchgUiNzHnHcKIiIiKnALwkOz98j+6CTd7WGBvvJEz/rSv3G4hPp7TkCwDgpTpkxxy3ERlWacFoCIiMhDZWRZ5LkF++XQBf2AUDbITx7vWV/6NAo37UrKmzdvloEDB+oGBOjTp4889dRTLj8uotKOIYGIiMgDZWdny7vLo2TziXjDWYwev/HqGARPDghff/21+Pn5ufzYiEo7hgQiIiIPNGvzGVmwM1p3W3iIvzzRs570bhBu2jEImzZtkjvvvJMBgcikGBKIiIg8zIpDF2XCquOG05w+dWN9ubVBOdPOYsSAQGR+5rx7EBERka7956/Ii4sOSbbONj8fb3mke13p7cEBoW/fvuxiRGQCbEkgIiLyEOcTUmXkT/skPdOSZ5uXiNzXuZYMbFbBtAulISBgDEJiYqJhQJg5cybHIBCZgDnvIkRERJRLchoWS9snsckZumfmttZV5cFrK5t6DIK9gHD77bczIBCZiDnvJERERJQjy2KRpxcckOOXknXPSoe65eX562tKmL+PKc/axo0b8w0IX331FVsQiEyEIYGIiMjkU52OW3FCNh+P091ev1KovNO7vpQL9DVtQMAYBAYEIs/CkEBERGRSWZZsmfVvtMzbdkZ3e8WwAPmwf2OJDPUXTwwI/fr1YwsCkUmZs9qBiIiolEvPypZlh2Nl2ppjujMZBfv7yPj+jaVueKCYVXR0tKSkpBgGhC+//JJdjIhMii0JREREJpOSYZE1JxNlyuqjkpKelWe7t5fIq7c2kHbVwsTMtJYCH5/cYyX69+/PgEBkcmxJICIiMpEr6Vmy6Wyy/LDxpJyLS9Xd56EuteS2JhXEEyAQwIgRIyQrK0v9/4svvmALApHJMSQQERGZRHKGRTafS5HVBy/K9hOxuvv0aBgho7pUF0+iBYUlS5bItGnTGBCIPABDAhERkQmkZ1lky/lkOXD+ivy2Q3+gct2IYHmvT0Px8sLSaeJxQUELC0RkfhyTQERE5GaZlmzZcj5FjsQky6y1x8WiM1I5NMBHJg5sogYsExE5G0MCERGRG1mys+Xf6BSJupgiM/8+KqkZeQcqw3t9G0mt8kFiNhs2bJAdO3a4+zCIyMHY3YiIiMiNC6XtikmVqEupKiAkpmbq7vdYlxrSvX55MZv169fLXXfdpcYY/Prrr9KoUSN3HxIROQhbEoiIiNzkwOU0OawCQpTEJqXr7tOzYYQ82rWmmM26detUQEhKSpK4uDg13enu3bvdfVhE5CAMCURERG5wNC5dDlxMlW/+OSYXEtJ09+lQq6yMu72ReJtsoDICwqBBg1RA0CAo4HsHDx5067ERUSnrboSbzoEDByQmJkYiIyOlc+fOUqGC/Tmit2/frmo1sIjLtddeKw0bNnTr/kRERFpA2B2TIt+tOy6nLyfrnpQWVcNk4sCm4u/rbfqAoImNjZUFCxbIjTfe6JZjI6JSFBJ++eUXGT9+fJ6aCfR/fPzxx+X111/PMxVcfHy8PPDAA7J69epc37/nnntk8uTJ4uvr69L9iYiItDEIh2PT5eDlNJm78aREXUjUPTH1KgTLZ3c1Nd1MRvYCAmCK0zFjxrj8uIjI8Uxfmv3zzz9VQGjcuLE0aNBAypcvL3v27JFt27bJxIkTJTQ0NM8N6bHHHlMFeOzbq1cvycjIkKVLl8rs2bNVK8Sbb77p0v2JiIgQEPZdSlOtCHM2npC9Z+J1T0rVsgEyfXAzKRfkZ8oxCMnJ+i0fAwcOlE8//VTOnTvn8mMjIsfziouL05mN2Tz++usvqVmzZp6uPFOnTpWXX35ZqlevrkKDdRegHj16SKVKlVRBvkqVKur76BZ08803qyXh9+/fLxERES7Z3wxSU1Pl1KlTUqNGDQkMDHT34Xg0nkueRzPh9eg55xLTnO6OSZUT8ekye+MJ2XcmQXe/CiF+8s3QllIj3FxTna5du1a1IBgFhDvvvFOmT58umZmZfL9xAL62HYfnsujM1dFRB/o16vX1HzlypISFhakxCtaw5DuMGjUqpwAPLVq0kMGDB0t6erosW7bMZfsTEVHplpWdLTsuXA0IP24wDghhAT4ybXBzjw0I7GpLVLKYPiQYwc0K3XyaNGmS6/va9Gvdu3fP8zPXX3+9+rxr1y6X7U9ERKU7IPx7PkVOJaTL9+uPy/6z+gEBYw+mDGomDSNDxEwYEIhKL9OPSTAyYcIE1YT07LPP5vr+2bNn1Wd0UbJVq1Yt9fn8+fMu298Ijt1V0Lph/Zl4Lt2N1yTPY2m4JtHFaG9slpxNypLv1x2XQ+ev6O4X7O8tn/ZvII0i/F363lCQMQhDhw6VlJQU3e0DBgxQYwPRxQgfwNe2Y/A8Og7PZW6F6U7pkSFh/vz58vHHH8uwYcPU4i3WtBus3kkICrrahGt9w3P2/kYQNjB+wZWio6Nd+nwlGc8lz6OZ8Ho037nEYL9or7ISmx1gPyD4esmrnctKeFasnDoVK2aByUGeeeYZw9Byyy23yPPPP284SJnXpGPwPDoOz6WoKfvr1q1bckPCnDlz1NSn6AOJoGArICAgJzkGBwfn2qbd7LR9XLG/kapVq4qr4Fjx4sBga39/f5c9b0nEc8nzaCa8Hs15LjGL0cF4iyQkZckvW04ZBgSMQZh0R0NpWtlcXYzQgoBWeqOAgBaESZMm6Y5B4DXpGDyPjsNzWXQeFRIwMOqll15SU7BNmzZNJSJbmIJ07969cvr0aSlXrlyubWfOnMnZx1X7G3HHLEN44+PsRjyXZsJrkuexpF2TCAj7L6XJ2WSLLNt9Xraf0G8dKBPoKzPubi5NK4eKmfzzzz92uxjh/RfvxXrvv9b42nYMnkfH4bkswQOX33vvPXnxxRfVgmX2blBNmzZVnzds2KBbOwLNmjVz2f5ERFR6YKG04wkZsv7wRVlz4ILuPmUDfeVzEwYEvK9hFqPiBgQiKhlMHxJQK4N+jx988IHcf//9MmXKFPH2Nj5s9JMEtDRcufJfEy9q/n/44Qf1s1jPwFX7ExFR6RAVlyZH4tJl96k4WbL9asuyrRB/H9WC0MRkAUGbfMOoKyzCAwMCUeli+u5GY8eOlS+++ELduBo1aqQK57YefPDBnObhbt26SYcOHWTTpk1qSlLUfGCqVBTg4+Pj5d5775Vq1arl/Kyz9yciopLvWHy6HLycLsdiEuWnTSfVwGVbvt5e8smAJqYMCID32cWLF0vfvn0lKioqV0Aw6uJLRCWX6VdcxjLvK1assLsPbmbWKxxjbAB+7sCBA7n2u+6669TA55CQ3IPEnL2/u3G1QZ5Ls+E1yfNYkq5JBASMQ4iOT5XpK49Iaob+zHXv920ovZvlP2bN3TD7nhYUChsQ+Np2DJ5Hx+G5LMEtCX369JEGDRrY3cf2ho6a/L///luWLl2qFj9DF6D27dtLz549dbsqOXt/IiIqmY7Fpcv+y2kSl5QuX/991DAgjO5R2yMCgnWLwueff65a89mCQFQ6mb4lgYqPKdpxeC55Hs2E16N7z+XRuHQ5cPlqCwICQnxKhu5+915TVZ7vWUe8vLykpOM1yfNoNrwmS3BLAhERkdloAeHExST5du0xSUnXb0G4uXEFGVNKAgIRlSzsG0NERFSEgHDgbIJ8tSbKMCBcU7OsvNunoXibKCCsXr1aJk6c6O7DICIPwJYEIiKiAjqZcDUgbDt+Wa2mbDHosNsoMkQmDGgi/r7epgoId999t+p+kZmZKWPGjHH3IRGRiTEkEBERFcCF5EzZezFN/j5wQX7fdc5wv3Y1ysjEgU3VqspmDAjwzjvvqC5Qzz33nLsPjYhMyjx3MCIiIpOKT8uS7dEpsmJftPy557zhfjc0jJBxtzeSABO1IKxatUruueeenICgefvtt9VnBgUi0sOQQEREZEdyhkW2nk+Rlfsv2A0IA1tXllduric+3l6mDwiaI0eOSHZ2NgdWE1EeDAlEREQGMrKyVUBYsf+C/GGni9EjnWvIqOtqmqqwvXLlShkyZIhhQEB4+Oyzz0x1zERkHgwJREREOrKys2VbdIr8ue+C/LbjrO45QvH6xZvqyt3tqprqHOYXELBt8uTJXCiNiAwxJBAREdlAF5xdF1Llj70XZNH2M4bn57Vb68uAVpVNdf4YEIjIEcwzsoqIiMgkAeHg5TRZsueCLNx22nC/V26pZ8qAYG8Mwr333ssWBCIqELYkEBERWYmKS5f5uy7I/C2nxGAZBPnfjXVlUJsqpgwIaWlpdgOCtzfrB4kofwwJRERE/+9kYpbM3hEjC7aeNgwIo3vUlnuvMd8YBAYEInIkhgQiIiIRifMKkoVbo2WJwSBleKp7LXmgQ3VTna8VK1aogchsQSAiR2JIICKiUu9sskXmHUyVZbuN10EY2bWmjOhUw1TnigGBiJyFIYGIiEq1M1fS5dN/Tsvq/RcM93m4cw15tItnBYShQ4fKpEmTOAaBiIqEIYGIiEqtc4np8u7yo7Lu8EXDfZ7oVkuFBDPZt28fAwIRORWnOCAiolLpVEK6jP3tiN2A8HzPOqYLCNC4cWMVEvSwBYGIHIEtCUREVOrWQdh/KVXGLY+SnSfjDFdSHturvgxsba51EDSYxvTjjz9WX3/99dc537/vvvvk008/ZRcjIio2hgQiIio1sizZsvVcskxYESX7zybo7uPtJfL2bQ2kT/NKYma2QYEBgYgciSGBiIhKhbRMi6w7nSifrTwqURcSdffx8faSD/s1kp6NKogn0IJCp06d5M4772QLAhE5DEMCERGVeAlpWfL3ySsyffVROXUpWXcfP28v+aBvPbnBQwKCdVAYNGiQuw+DiEoYhgQiIirR4lKzZOWxePl8zVE5F5equ0+Qn7e82KmsdK5TTsw2fsLLCyMkiIhci7MbERFRiW5BWBYVJ1NXRhkGhLAAH5k6sJE0q+gvZrJ8+XK57bbbJD4+3t2HQkSlEEMCERGVSInpFlm4/7JM/vOwxFzRX3CsfLCffH1vS2laOUTMFhAwlen69etl4MCBDApE5HIMCUREVOKkZFrkx50XZMqKw3IlNVN3n8plAuSboS2lQaS5AsKyZctUQEhPT1f/37p1qxqUzBYFInIlhgQiIipxsxjN2HhWZqyKktQMi+4+tcoHyrdDW0qt8kFitoCAqUy1gKDZsmWLCgpJSUluOzYiKl04cJmIiEqM9Kxs+fjvk/LT5tOSbbBPo8gQmTa4mUSEmGsMwh9//CH3339/noCgadasmQQFmSvUEFHJxZBAREQlQkaWRV7746j8vvu84T7X1iwrEwc2kdAAX48KCMOGDZNPPvmE6yAQkcuY6y5JRERUxIDwv8WHZeWBGMN9bmoUIe/1bST+vt4eFRCGDx+uFkzDeghERK7CkEBERB4ty5It41adsBsQBretIi/eVFe8TbbmAAICxiBkZGTobmdAICJ3YUggIiKPDgifbzorP289Y7jPqOtqyiOda5huUbL8AsKDDz4oH330EVsQiMgtGBKIiMgjWbKzZd6eGPly7XHd7d5eIi/dXF8GtaksZvP777+rLkYMCERkVgwJRETkkQFh2eF4mbQiSjKz8s5jhDaDt25rKH2bR4rZMCAQkSfgKCgiIvK4gLDuZKJ89OdhSUrL0t3nietre2RAGDFiBLsYEZEpMCQQEZHHyM7Olu3nU+SjP4/IxStpuvsMbF1ZRnSoJmazdOnSfAPChx9+yDEIRGQK7G5EREQe43Bsmnz8V5Qcv6i/8nDnuuHyys31TDdIGQHhgQceyLcFwWzHTUSlF1sSiIjII5y5ki7vLYuSvWfidbc3jAyRj/s3Fh+MWDaRuLg4eeyxxxgQiMijMCQQEZHpxaZkylt/RMn2E7G62yuF+cu0Qc0k2N9HzKZcuXLyzTffSGBgYJ5tDz30EFsQiMiUGBKIiMjUktKz5I1lUbIx6pLu9hB/H5k6qJlUCPUXs7rhhhvkxx9/lICAgFwBAWMQ2MWIiMyIIYGIiEwrIytb3vnrmKw2WE05wNdbBYT6FUPE7BAUZs+erYICAwIRmR0HLhMRkWmnOh2/6rgs3XVed7ufj5dMvrOptK5eRjwFgsLq1aulcePGbEEgIlNjSCAiIlNOdTp57WmZt/WM7nYfL5GP+jeWDrXLiadp0qSJuw+BiChf7G5ERESmM3XdaZm57oTuNswS+k6fhnJ9gwgx2yxGREQlBUMCERGZyqdrTsjna/UDArx8cz3p3cxcqykvWbJEWrVqJevWrXP3oRAROQRDAhERmaaL0YcrjsnMDacM93n6+toyqE0VMVtAGDZsmMTHx8tdd93FoEBEJQJDAhERmWKQ8tt/RMn3W/THIMDwjjXkwY7VxUwWL16sAkJmZqb6f3JysgwaNIhBgYg8HkMCERG5VUaWRV5ZfEjm79SfxQju71hdnrm+lpgtIAwfPjwnIGiSkpJUUIiKinLbsRERFRdnNyIiIrfJsmTLi4sOyl8H9RdK8xKRBzrXlGeuqyFmsmjRInnwwQfzBATNvffeK3Xr1nX5cREROQpDAhERuW0Mwvi/jhoGBG8vkQe61JInO1c31ZoC+QWERx55RMaPH2+qYyYiKiyGBCIicotvN5+Ruf+e093m6+0lD3StLY+0ryI+SAseEhAeffRRGTduHAMCEXk8hgQiInK53/fFyIRVx3W3+ft6y7CudeTeNpES6GueoXO//vqrjBgxggGBiEoFhgQiInKprSfjZexvh3S3+fp4yfBudaV/0wgpF+DjMQHhsccek/fff58tCERUYjAkEBGRy0RdTJZn5u+TjKzsPNvQqejuDrWkZ/1wqRrqZ6qAgC5GWVlZutsZEIioJDJPOy4REZVoF66kyeM/7ZUrafqF7dtaV5UbGkZIw3B/8ZSAMHLkSLYgEFGJxJBAREROl5SWKU/M2yfnEtJ0t1/XsKL0al5JWkUGmqbLTkECwnvvvWea4yUiciSGBCIicvpiac8tPCAHLyTpbm9Ro6zc3raqXFMpSM1qZAYMCERU2jEkEBGRU9dCeGPpYdlwLE53e+0KITK4fU25tnKwBPl5e0RAGDVqFFsQiKjEM8cdmYiISqTJf5+QJXtjdLdVDAuQ+7rUljaVgyU80BwzGSEYTJo0yW5AePfdd9nFiIhKPIYEIiJyijnbzspXG07rbgsL9FVTnTaPDJLqYeaZycjHx0d+/vlnadmyZZ5tjz/+OAMCEZUaDAlERORwKw5elHF/HjVeLO26OtKoYpA0Kh9gurMfHh6uuhy1aNEiV0B455132IJARKUG10kgIiKH2n46QV5afEjyroQggnHJQzvXlkaVQqVVZJBpC91aUOjXr59069aNAYGISp1ih4SoqCiZMWOGbNy4UWJiYiQtTX96O+jSpYt89913xX1KIiIyqWOXkuWpn/dJWqZFd/vAa2tIi2pl5JrKQeJnkpmMjJQvX15+//13CQkJMW2YISIyZUhYtWqV3HPPPZKamlqg/RMSEorzdEREZGIxiekyau5eSUjN1N3eq0UVaV+nvFxTJUiCfD2jt2toaKi7D4GIyLNCAqa1e+qpp1RAqFOnjowZM0aaNWsmZcqUMfyZoKCgoj4dERGZWGJaplpN+azBYmmd6kdI98YVpW3lICnj7/6ZjHbu3KkGJ7OFgIjIwSFh//79curUKQkICJAFCxZI7dq1i/pQRETk6YulLTBeLK159bLSp3U1NQahQpD7h8L98ssv8vDDD8ujjz7K2YqIiAwU+W4dGxurPjdu3JgBgYiolEKr8utLD8vG48aLpQ1qX1MaRwRINRNMdaoFBKyDMHXqVPU9rHtAREQOCgmRkZHqc0ZGRlEfgoiIPNykNSfkN4PF0iLLXF0srW54gNQr5y9mCAgPPfSQWCz/DapGUECXo1dffdWtx0ZEZDZFHjnWoEEDqVu3rhw5ckQuXLjg2KMiIiKPWCxt5kbjxdKGXVdXapULkGYVAtze918vIGimTJkif/75p1uOi4jIrIo1vcS4ceMkMzNTnnzySUlMTHTcURERkan9dcB4sbQAX2+1mnKt8EBpUylIvN0cEObPn28YEACTcNx0000uPy4iohLZ3ejkyZOyd+9e6dChgyxbtkzat28vN954o9SqVUu8vfWzR82aNWXgwIHFOV4iInKzhbui5a3fD+sulubj7SVD0cWoQrBcWzlIfL3dHxAwBsEoIDz99NPyxhtv2F3jh4ioNCpySDh69Ki8+eabOf8/e/aszJo1y+7PdO/enSGBiMiDByl/s+mMTFx93HAfLJbWpEqYXFs5WLUouNPPP/8sjzzySL4Bwd1doYiISlRIqFKlitx3332F+plGjRoV9emIiMiNLNnZ8snKY/LdlrOG+2CxtHa1w6Vd5SAJ9WdAICIqlSEBBf7Jkyc79miIiMiU6yBgmlOjWYygS4MK0q1xRWlVMVDKB/qaugXhmWeekddff50tCEREdrh/VRsiIjKt5PQseX7hAVl79OraOHp6Nq0kPZtVkqYVAqVKqHvXQpg3b55aJI0BgYioeBgSiIhIV1qmRZ6Zv082nYjX3Y6e/H3bVpNO9StIrTJ+Uqesv6kDwrPPPiuvvfYaWxCIiBwZEjBQWRuYjPURunbtmu9AZVv4ufvvv79QP0NERO7pYoQWBKOAgFmMBnWoKS1rlJPygT7SJCJA3Omnn36Sxx57jAGBiMjVIQFTnk6cODFnliJMZ6r9v6DwcwwJRETmlmXJlrG/HZY1Ry7rbvf39VYrKdevFCaBPl7SplKgW9dCYEAgInJjSEAowGAvrUXA+v8FhZ8jIiJzT3P6zrIj8vs+/UHKIQE+aiXl6uWDBUsgtK0cJAE+3qYNCKNHj5axY8eyixERkbNCAgr4mE/amu3/iYjIswPCRyuPyS87o3W3B/n7yIju9aRKuSD1/+YVAqVcgI+40+HDhxkQiIicwL0TWRMRkWlMX3tSvjdYBwFdjIZfVzcnIGCgcvUw985kBC+//LKMGTMmz/efe+45tiAQERUDQwIREcm3m07L9HWndM+Er4+XPNC1jtSICFb/DzfBQGUNVkt+5ZVXVCjQ4OtXX32VXYyIiMwwBWpCQoIcOnRI4uLiJDMzU3efChUqyDXXXOOopyQiIgf4euNpmbj6uOEsRvd2ri11I0PV/zFQuW2kewcq6wUFhAINAwIRkQlCwu7du9W806tXr1b9WfOb3ejXX38t7lMSEZED4J495Z+T8sV6/RYE5IDBHWpK4ypl1P+Dfb2kfZVgCfA1XyO0dVDA10RE5MaQ8O+//8ptt90mKSkpEhwcLJUrV1brKZQpU0YNdD548KDaFhgYKK1bt5amTZsW83CJiMgRLNnZ8uFfR+XHbecM9xl4bQ1pUaOc+jrUz1vaVwmSQBMGBA3DARGRSULCiy++qELAzTffLDNnzpStW7dK//79pU2bNqrF4MqVK/LSSy/J999/rxZfs24OLoqMjAzZv3+/ZGVlSZMmTVT4sIVtu3btMnwMhJeyZcvqbktLS5MTJ06Ir6+v1KpVS3x87M/aUdj9iYjMsg7CW38ckYW79Gcxgr5tqkm72uXV1+UCvOXaysHi5+OeGvrFixdLhw4dJDIy0i3PT0RUGhU5JMTExMjmzZvFz89PLaoWGnq1v6q1sLAw+fTTT2XHjh3y8ccfS58+fVSLQlHmwf7jjz/kr7/+UmMfYP369botEwgmPXr0MHysOXPmSK9evXJ9D/Nrf/DBBzJ16tScx69YsaKaGUNv8bfC7k9EZKaVlF9efEiWH7ioux0xoF/batKhfgX1/wpBPtK2UpD4YlEEN8A9e+TIkdKoUSMVFnCvJSIiE4eEY8eOqc/16tWTqlWr5mrqtR64jNr1e++9V7Uo/PLLL4UOCXisRx555OrB+vqqMJKYmJjvz5UrV05q166d5/t6rQhvvfVWzurRaGlIT0+X06dPy1NPPZVz/MXZn4jIDFIzsmTMwgPyT1Ss7nbkgDvb15Q2tcLV/yuH+EqryEDxcVMf/9mzZ8uoUaPU2IkDBw5I3759GRSIiFykyJ1LtUAQHn71zQSCgq7On63Vrmu0wjrGKxTleQYNGqS6M0VFRUm3bt0K9HM33HCDGkxt+9GpU6dc+6G70GeffSYBAQGycOFCNc5iz549qpUAMCgbXaqKuj8RkVlaEEYvMA4ImMVoSKfaOQGhRpiftDFJQNBoQQEt2UREZNKQUL16dfX57Nn/Ft7RWhTQyoCxAZoLFy7kjCkoLNTMf/755zJgwADDsQTFgbETaK146KGH5Prrr8/5/pAhQ6R3795y6dIlWblyZZH3JyJyNxS031kWJeuO6gcEjDW4v2ttaVb96j22UrCvNK8Q4LaBwHoBwTooLFiwwC3HRURUmhQ5JGAmo0qVKqmQoHX/QXDAB/7/zTffqO8lJSXJV199pb6uX7++uBLCyd69e+3WOmGwNdx66615tqHQD9u2bSvy/kRE7jZj3SnDQcoBft7yYLe60rBymZxZjNDFyF0B4ccffzQMCPC///1PHn74YZcfFxFRaVPkMQl4A7n99tvliy++ULU69913n/r+o48+qgbwjhkzRm27ePGi+sAA56FDh4qrLF26VI2B0GCQM2ZX0grympMnT6rPDRo0yPMY2vfQxaio+xtJTU0VV8GYCevPxHPpbrwmXXcel+y9KNPWXr1v2Qr291EBoVr5qysp+3qJtAj3lsz0NNFfEtO55s6dK88884xhQMBKyqNHj1Yzyzkar0meRzPh9chz6Sx6M4M6ZQrUxx9/XE1FWqNGjVzfQyEZrQdoFgasmzBhwgS1r6vgTaRmzZri7e2tWjv27dunugRh7AA+a9DSoc3EZAvHDdYDpQu7vxEck3WXLFeIjjae7pB4Lt2B16Rzz+OO6DR5f12c7jZ/X+9cAUGys6VSVqxcPOeeygTMXPT2228bBgS0Htx9991y6pT+wm+OwmuS59FMeD3yXDoSuvBjwh2XhAQMSH7wwQdzfQ+F8o8++kieffZZNaAXA3zbtWunW6h2BsyAhJaMESNGqBmOIC4uTj788EOZMmWKvPzyyzJw4EB1XNr+tjMyabTvafsUZX8j2vgNV9VI4EaD7mH+/v4ue96SiOeS59FTrsdDMcnyyeYDkpWtP4vRkE61/gsIItKwnI9UD6kk7prm1F5AQAvC888/79Rj4Gub59FMeD3yXJpBsUKCPdWqVVMfroYpUvGGYg1h4d1331VjBTZu3Cjbt2+Xjh075mwDjFvQWgI02lgG6xmcCru/I5p7HAWFCHc8b0nEc8nzaObr8XxCmoz+9Ygkp1t09+/Xtro0qvLf/QszGdUr756Byj/88IOqVDIKCC+88IKaQttV+NrmeTQTXo88lx45cNkTaVOxYsE1TcOGDdVnLPhmC9ObAhbxKer+RESulJ5pkdG/7JeYRP1uQz2aREr7ehE5/w8P8JFmbprJ6Pvvv5cnnnjCNAGBiIicEBLi4+Nlw4YNuh+7d+8WVzEaD4DB02vWrFFf16pVK+f72jSmWIfB+o0KA4vxBma9T1H2JyJypQ9XHJW95/Xvg1gD4abmlXP+H+jjJW0rBYq3mwLCk08+aRgQXnzxRQYEIiJP6m6ENQswRR1mMbrnnntyvo8uPP379zccKLFixYpCr7asOXz4cE7hH2EEMCham+ECtfshISHqa3Q1whgEzGKktRxgbMSMGTPk3LlzanyE1hoA2A/Ttq5bt06GDx+uPtAXcNKkSXL8+HHp3LmztGjRosj7ExG5ym97L8hP28/rbqsXGSoDrqme02Lg6y1ybZUgCcAXLvbdd9+pFertBQR8EBGRh4QEFIKxojAG3WJxs4LCLD6vv/66WoisKDAlHgrl1qwHTP/5559y7bXXqq8RFjCN3rJly/I8Tp06dXLWbNBgADOCz1133aVWUMaHBoMBtZWUi7o/EZErHIlJkrf+OKK7rWJYgAztXFt8fbxzBi5fUylIwvx9TBcQ0L0I3YyIiMiDQgIKwOhWgyZibXYgW926dcu1PgHeCNCC8Pfff8vRo0cLNfWS9foD2tSjerRWBPj444+lX79+smTJEtUCgeNFqOnevbsq2AcH/zebhwa1/2vXrpXp06errlFo+Wjfvr2MHDlSKlSoUOz9iYicKSk9S55bcEBSMyy6U53e27m2BFoFgtaRgVI+yGnzVhhCKzQDAhGRZyjUuwRaArRF1Ixgu+0UoOiGhOlHV61aVaSQMHHixALvi+dHIMBHYaCVYfz48U7bn4jIGVAR896fx+X45RTd7Xe0qy6Vyv4381GziACpHOLnlj8GKnwwHXZCQkKebWxBICIyF+/CdDXCfNwo5EdE/DczRkG0atXKcEYgIiIquqVRKfLX4VjdbR3rRUjrWv9NyVyvnL/UKuu+tVLQLXT+/Pl5po/G+jXsYkRE5KEhQVv1LzIyUnd7+fLlpWfPntK2bVvdbdoMQ0RE5Bi7zybKrF3/TelsrXr5YLmt9X+LNlYP9ZWG4e5fTNE2KCAg/O9//3P3YRERUVG7G2mDzPRWGoaWLVuqG78ejAuwfgwiIiqe6IQ0eXlplO6KykH+PmpFZW2gcvlAH2leMdAtayHYCwrr16+Xp59+2t2HQ0RExQkJWmvAmTNnpLBOnz6d6zGIiKjoLidnyKNz98iFxIw82xADBneoKeEhV1sN/H281EBld6yFkF9Q0GalIyIiD+5uhIG6mEXo7NmzatagwtAWMWvevHnhj5CIiHJcSc2UUXP3yLFL+gOVezStJI2q/Nfnv1XFQAl0w1oIRETk2Qr8zuHn56emN4VPPvmkwE+Ahcy0NQswZoGIiIomJSNLnvp5n+yP1p8Sun6lUOnZtNJ//y/nLxWDXTvV6axZs+T333936XMSEZHjFap66YknnlCfZ8+erVYYzg/WRbjvvvvEYrHILbfcIo0aNSr6kRIRlWIZWRYZs+CA/Hs67/ShULlsoNzTsZZ4Y6W0/x+H0MDFA5URELAOwv3338+gQERUmkJCly5dZMSIEeprrLx86623ypw5c9T0qGlpaSoMxMXFqdWRMWMFWh6OHTumpkz98MMPnfU7EBGVaFmWbHl58SFZe1R/qtPyof7yYLe6EhxwtdXA3/vqOARXDlT+9ttvVUCAjIwMBgUiIg9X6HbocePGqTcA1Bht2LBBfWjwhmQ7g1H16tXl+++/l5o1azrmiImIShHcU99ddkSWH9CfQrpMkJ881L2ehAX9t0Baq0jXjkNAQLCdpUgLCt9995306tXLZcdCRESOUeh3EYxNQFejuXPnSqdOncTb+7+HsA4IFSpUkNGjR8s///wjrVu3dtDhEhGVLjPWnZL5O6+uU2MrJMBHRnSvmzOTkTvGIegFBOugcODAAZcdCxEROU6R30kwxgAf6F60fft2iYmJUV2OypUrp1Zlbtq0qWnm5CYi8kS/7Dwv09ae1N0W4Octw7vVlcgygTnfwziE+i4ch/DNN9/IM888Y7gd3VLtbSciIvMqdnUTQkGPHj0cczRERKSsOXJZ3vnjiO7Z8PXxkge61pFq4cE53yvj7y1tK7luPYSCBAS0JhMRkWdy7dx4RESUr91nr8j/fj2gu5oyJi+6t1NtqVMxNOd7EQFe0q5KsPj+/8xGzvb111/Ls88+a7j99ddft7udiIjMjyvsEBGZyMnYFHly3l5JzbDobu/frro0rvrfYmllLMnSorwPAwIRETkUQwIRkUlcSkqXkXP3SmxKpu72ns0qybV1I3L+XyfMWyKzE1zWxWjmzJl2WwjeeOMNtiAQEZUQDAlERCaQnJ4lT87bJ6fjUnW3X1unfM5qyogELSoGSp0wH/W1qwKCvTEGCAgcpExEVHIwJBARuZklO1teWnRQ9p5P1N3eqEqY9GtXPWfGOCyUViPsv3URnI0BgYio9GFIICJys0lrTsjqI5d1t1UvHyxDOtUSn/8flIx1EKqEui4gfPXVV3ZbEN588022IBARlUAMCUREbrRod7R8vfG07raIUH811am/r4/6f8UgH2ngwnUQEBCee+45uwHBaCE1IiLybAwJRERu8u+peHnz9yOGqyljsbTQwKszVQf7ekmryCCXLVK5ePFiBgQiolKMIYGIyA0wQHn0L/sl05J3MQR0LRraubZEhAao/6OnUdtKQeLv47pV7LFIZufOnXW3vfXWW2xBICIq4RgSiIhcLDEtU576eZ/hVKd3XFNdalstloaZjMoEXO1y5CqhoaHy008/5QkKCAhPPfWUS4+FiIhcjyGBiMiFsizZ8uKigxJ1MVl3e/fGFaVd7fI5/69dxk+quXCgsr2gwIBARFR6XO3sSkRELvHhiqPyT1Ss7ramVcvIzS2q5Py/fKCPNI642uXIXbSg8Mcff8idd97p1mMhIiLXYUsCEZGL/LDljMzedk53W5VygTKoQ82c1ZODfL2kTWSgy1ZTzi8oMCAQEZUuDAlERC6w6tAl+XDFMd1tmMHo/i51JMDPJ2cmo45VgyXA1/m36OzsvAOniYiIGBKIiJxs77kr8uLig6JXHPf18ZL7utSWciFX1z8I9rsaEIJcEBA+//xzefzxxyUrK8vpz0VERJ6FYxKIiJzobHyqPPnzPknNsOhuH9S+ptSMCFFfh/h5S4cqQRLoooDwv//9L+f/kydPFh8f186gRERE5sWWBCIiJ7mSmilPzNsnl5IydLff2rKKtKhRTn0d6uctHV0UEGbMmJErIPz444/y5JNPskWBiIhysCWBiMgJMrIsMmbhAcOpTtvXjZDrGlVUX4f5e0v7KkES4OOagPDCCy/k+T6CAkyZMsVlqzoTEZF5MSQQETnB+8ujZOPxON1tDSuHye1tq6nCeJn/Dwj+bgwImmbNmjEgEBGRwpBARORgC3dFy/yd0brbKpcNlHs61RIfby+1DkK7ykHi5+38mvvp06fLiy++aLj93XffVYOYiYiIgCGBiMiBDl1IkveWR+luKxPkK8OuqyOBfj4SGewjbSKDVFhwNgYEIiIqLIYEIiIHSUzLVOMQ0jLzzmTk7+stD3StK2WD/aVaqK+0qOiahdLyCwjvvfeejBo1yunHQUREnoUhgYjIQYuSvfn7ETlxOUV3+13ta0jV8CCpVcZPmkYEuKTv/7Rp0+Sll14y3M6AQERERhgSiIgcYM6/52T5gYu62zo3qCDNq5eTBuH+Ur+cvykCwvvvvy8jR450+nEQEZFnYkggIiqmPeeuyEcrjuluqxERrNZDaFI+QOqUu7qqsrMxIBARUXFxMTUiomKIT8mQZ3/ZL5mW7Dzbgvx9ZEjHWtKiYpDLAsLUqVPZgkBERMXGlgQioiJKSsuUx37aKxeupOtuH9yhpnSsEebSgPDyyy8bbh83bpw89thjLjkWIiLybGxJICIq4kxGD8/ZI/vOJepuv75JpNzerILUZUAgIiIPxJYEIqIiBIRH5+yVvQYBoW7FEBnVtabUKxfgknO7e/duuy0I48ePl0cffdQlx0JERCUDWxKIiAodEPaowcp6ygT5yYu3NJBGEYEuO68tWrRQKybrYUAgIqKiYEggIipEQHhs7l7ZY9CCgBWV3+zTWNpXC3H5OX388cfzBAUGBCIiKip2NyIiKmhAmLNXdhu2IPjK2NsaS4+6ZVyyDoJRUMCibq+++qp88MEH8sgjj7jlOIiIyPMxJBAR5SM5PUtGzrUfEF68tZHcWK+s2wKC5oknnpBu3bpJy5Yt3XocRETk2RgSiIjsyLJkywuLDsius8ZjEJ67pYH0ql9OvN0cEDQMCEREVFwck0BEZABdd8b/dVT+PhJrGBCevbmB3NYgXHy8nR8QrlzRDypERESOxpBARGRg1uYzMvffc4YB4ckb60vvBuXEz8f5AWHy5MnSpUsXOXnypNOfi4iIiCGBiEjHnwcuyierjuuem5AAXxnVs77c1jBcAn29XRIQxo4dqwJCnz59GBSIiMjpGBKIiGzsOJ0gLy8+qHtefH28ZFjX2qoFIcTP+bfQSZMmqYCgYVAgIiJXYEggIrJyMjZFnpq/T9KzsvOcF3QqurtDLenbJELKBvi4JCC89tpreb6PoNC3b19JSkpy+jEQEVHpxJBARPT/LidnyKif9kp8SqbuOenduqr0b1FRqoT6Of2cffrpp7oBQfPkk09KSIjrF20jIqLSgSGBiEhEUjKy5Ml5e+VUbKru+ehUv4Lc2bqy1C/n75KA8Prrrxtu/+ijj+Shhx5y+nEQEVHpxXUSiKjUy7Rky/MLD8iec4m656Jp1TIypH11aRUZ5PTF0vILCB9//LGMGDHCqcdARETElgQiktK+FsLbfxyRf6L010KoXj5I7utSS66tGuz0tRAmTpzIgEBERKbAlgQiKtUm/X1SFu6K1t1WPsRfhl9XRzpXC5UgJ091ioDwxhtvGG7/5JNP5MEHH3TqMRAREWkYEoio1Pp68xmZueGU7raQAB8Z0a2udKkRJuUCnTuT0YQJE+TNN9803M6AQERErsaQQESljiU7W+buvCCTVh3T3Y4VlBEQbmsULuFuDgjYPnz4cKceAxERkS2GBCIqVbIs2fLr/svyyV9RYsm7FIJgXPKI6+rIkFYVnd7FCC0Eb731luF2BgQiInIXhgQiKlWzGC0/Eief/HlY0jMtuvvc17GmPNy+ivg6eZAyAwIREZkZQwIRlQpYQXnNiQSZ8NcRuZKqv1jagDZVZHS3Gk6f5jS/gIBBzMOGDXPqMRAREdnDkEBEJV5KpkXWn06SKauOyoWENN19bmxcQV67ua7TA8KlS5dk6tSphtsZEIiIyAy4TgIRlWhJGRZZfyZJvl1/Qo7FJOnuc23NsjKub0OnBwSIiIiQRYsWSYUKFfJsY0AgIiKzYEggohIrIS1LNpxNll+3n5WdJ+N096lfMVgmDmwifj6uux02bdo0T1DASsvsYkRERGbBkEBEJVJiukU2nUuWvw/GyJoDMbr7RIb6y9RBzSQ0wPU9L7WgEBkZqQLCAw884PJjICIiMsIxCURU4mRYsmVbdIrsOh0vi7af0d0nNMBHpg5uJpXCAsRdEBS2bNkiZcuWddsxEBER6WFLAhGVKNnZ2bIjOkV2n0mQ2RtOSLbOWgiY3vSTO5pIg4oh4m4MCEREZEYMCURUohy8nC6bTsTLN/8ck4wsnYQgIm/2biAdapdz6nHs3bvXqY9PRETkTAwJRFRinLmSISuOXJZv1yIg6C+W9mS3WtKneaRTj+ODDz6Qrl27ypw5c5z6PERERM7CMQlEVCLEpWbJ4v2XVAuC0WrKA1tXlhGdqjs9ILz33nvq65EjR6rPd999t1Ofk4iIyNHYkkBEHi810yLz916UmX8fNQwIvZtVlFdurufUtRDGjx+fExC08REICnPnznXacxIRETkDQwIRefxMRggIM1ZHSZpBQOjVpIK8fVtD8fF2bkB4//3383wfQWHUqFFy/Phxpz03ERGRo7G7ERF5rExLtiw7EidTVkZJWoZ+QLi5cQV5t28jNaORs4wbN0596EHLBVZSrl27ttOen4iIyNHYkkBEHinLki0bTifJlJVHJSktS3efng0j5L2+Dd0aECZNmiT33Xef056fiIjIGdiSQEQex5KdLdvOJ8uX607I2bgU3X16NCgv4/s1Ej8f59WFoHsRuhnpYUAgIiJPxpYEIvK4gLA9OlUW7oqWHSdidffpXDdcPuzfmAGBiIioiNiSQEQeA4OAd8WkyvrjcfL7zrO6+9SOCJKP3NyCMHnyZBk6dKjTnp+IiMjZ2JJARB4TEHZfTJO90ckye8Nxsegsphwa4COTBjaVkADn1X8wIBARUWnAlgQi8giHY9PlWGyafL/+uO5AZQxNfr9vI6lVPshtAeGzzz6Te++912nPT0RE5CoMCURketFJmXI4Nk0WbDstZ2P1ByqPuq6mdKtf3mmtGAgIWE1ZDwMCERGVNOxuRESmlpRhkZ0xKbLmwAXZbjBQ+YaGEfJQ5xpOeX4GBCIiKo3YkkBEpl4sbdv5FNlw5JIs231ed5+6EUHyzm0NxNvLyy0BYcqUKTJkyBCHPzcREZE7sSWBiMw7UDkmVTYdi1XdjPRgoPIEJw5UzsrKkl27duluY0AgIqKSjCGBiEzpeEKGrD0WJ7M3npBsnZmM0G7wXt9GUtuJA5V9fX3l22+/lV69euV+brYgEBFRCceQQESmczklU1YciZXv1h6TLL25TkXk5ZvrSXcnDVS2FhAQoILCLbfckhMQpk6dyi5GRERUonlESDh58qR8/vnncscdd0i1atWkUqVKcuDAAbs/s3r1aunXr5/UrVtXGjZsqN7QjboNuGJ/IiqY1EyLLDsSJzP/PiZpmRbdfUZ2rSmD2lZx2SlFUJg1a5b07t1bBYR77rnHZc9NRETkDqYfuIw+wS1btszzfYtFv/AA8+bNk0ceeUT1adYsXbpUVq5cKYsWLZL27du7dH8iKuDr3ZItq44nyPRVUZKUlqm7z91tq8ijXZwzk1F+QeGHH35QLQlEREQlnelbElAQr1Gjhjz88MMyf/58uemmm+zuHxcXJ2PGjFE/9+KLL8qhQ4dkz549MmzYMElNTZWnnnoqV8Bw9v5EVPDX+o7oZJm2+qjEJWfo7nNr04rywk113VZQZ0AgIqLSwvQhAQMHd+/eLR9++KH07NlT/Pz87O6Pmvz4+HjVNQmF+MjISKlevbpMmDBB2rRpo7opbdy40WX7E1HBByrPWHtSTlxM1t3epU45edsJU50inPz000+Snp7u0MclIiLyZKYPCYW1fv169Xnw4MF5agAHDRqkvl63bp3L9iei/F1MzpRvN5+VTVGXdLe3rBomH93RRPx8vB0eEN59913VffCBBx5gUCAiIiqpISEqKkp9btq0aZ5tzZs3z7WPK/YnIvuSMyzyy96Lsmj7Gd3tVcsGyOS7mkqwv4/DA8I777wjH330kfr/77//zqBARETkKQOXCyshIUF9Ll8+79SI4eHh6jO6C7lqfyMYv+AqWjcKdqfguTQL7VpMSUuXNeeS5Nt1x3WnOg3w9ZYP+tSTQK8sSU3NcmhAGDdunHz66ae5vo+gcP/996vZ1Pz9/cXs+NrmuTQbXpM8j2bDazK3wMBAKbUhwR5vb2/T7H/27Fk1c5MrRUdHu/T5SjKey+JDJNganSKfrzktian6Mxk90S5MAlMvyalTlxz3vNnZMm3aNPn66691ty9fvlyWLFki1157rXgKXo88l2bDa5Ln0Wx4TYr4+PioqftLbUgICwvLmYUoNDQ017bY2Nhc+7hifyNVq1YVV6ZovDiwvoQn1I6aGc+l487j3ugr8sO2i3ImNkV3n2HXVpZBnaqLIyEgvP/++4YBAUF/0qRJMmDAAPEEvB55Ls2G1yTPo9nwmiy6EhcS6tSpI1u3bpX9+/erWYes7du3T322TlHO3t8RzT2OgoDgjuctiXgui+dcskUWHUmT7SeuBmtb19ULl6dvqOfQmYwQEN5++20VAowCwvTp03MmIPAkvB55Ls2G1yTPo9nwmiy8EjdwuVOnTuoz1lSwpX1P28cV+xNRbheSM+XnPZfl951ndU9NzfAgeb9vI4cHhLfeeks++eSTEhcQiIiInKHEhYR+/fqpbkBz585V/Y4xQBiDjV977TW1fgFaArp27eqy/YnoP3FpWbJw70WZs+mEGpNgK8TfRybf2UTCAn0dHhCwlokeBgQiIqK8vOLi4vTeq00FC5dp6xNkZGSoFY3RbKStfrp06VJp165dzv7ob/zss8+qr7EPCgnawmyo7e/evXuux3f2/u6GIHPq1Cm1cjW7G/FcunOq00UHY2XSn4ckKS3voH28mjHV6XX18s4c5syAMGPGDLnrrrvEE/G1zXNpNrwmeR7NhtdkCW9JwKCTtLQ09YGAYPQ9zfDhw+Xbb7+VVq1aqZHcAQEB0rlzZ7Vasl4B3tn7E5V2aVkWWXk8QT5fHaUbEODZHrUdHhDefPPNEhsQiIiIpLQPXF6wYEGeIGANhXRb6BaEDxQUtBYHe5y9P1FplWnJlvWnk+Tz1UflUuLV9RFsDW5bWe5vX83hAWHixIm62xkQiIiISkBIKM60nYUtwDt7f6LSBIX17dEp8uU/x+XEpWTdfbrVD5cXbqznsNcSnvONN97Is1CadUDAYml33nmnQ56PiIioJPKI7kZE5JkOXE6THzaflj2n9VchbxwZLONvbyw+3gwIREREZsKQQEROcSIhXZYdvCyr91/Q3V4x2Ec+6ddAgv19HPacGKTMFgQiIqLiY0ggIoeLSc6UTacSZd6mk7pTnYYF+MirXcpJRIifQ5+3QYMGut2W0MXoiy++YBcjIiKikjQmgYg8R0J6lmw7lyxzNp2UxLTMPNt9vb3kw9vrSwVLnMOfe8iQIerz448/njM1sRYQBg4c6PDnIyIiKqnYkkBEDpOWaZGt51Pkr/0XJOpCou4+o2+oI22qhTntrCMofPbZZ6pFAQHhyy+/ZEAgIiIqJLYkEJFDZFmyZWt0iuw/d0X+2nted58eDcrLkHZV1PomznTvvfeqz0FBQTJgwACnPhcREVFJxJBARMVmyc6WHRdS5WxCuszZeFL+v6dPLlXKBMibvfXHDDgzKBAREVHhMSQQUbGkZ2XLv9EpciklU+ZtPikJKRl59vHxEhl3eyMpG+TYgcpERETkHByTQERFlpRhkQ1nk+VyapasPXRRDp67orvf491qSevqZYp9pjEYefz48XLgwIFiPxYREREZY0ggoiKJTc2SDWeSVVA4dP6K/LHrrO5+neqUk+EdqzskIIwdO1bef/996du3L4MCERGREzEkEFGhnU3MkE3nkiXdki0nLyXJ9+uOi0VnHEKFED95t09D8S7mOAQEhFdffVXNWgQxMTEMCkRERE7EkEBEhXIsPl0NUkYoiI5PlW/+OSYZWZY8+yEWvNe3kUSE+DskIEyZMiXX9xEUbr/9drYoEBEROQFDAhEV2PmkDNl/6er0pbFJ6TLz76OSkp6lu+9jXWtKh9rlnBIQNJcuXZLDhw8X6zmIiIgoL85uREQFXkl554VU9XViaqYKCHozGUG/FpHyaJcaxQ4Ir7zyikydOlV3u4+Pj3z11Veq2xERERE5FlsSiChf6VkW2XY+RbKyRdIysuSbf47KxStphgumvXZr8dZDKGhA6N+/P/96RERETsCWBCLKd6G0f6NTJSUzW9IzLTJr7TE5E5uiu+81Ncqo9RB8vYsXEF5++WWZNm2aYUCYOXOm9OvXj385IiIiJ2FLAhHZte9SmloHIeP/A8LRmCTd/RpFhsjEgU0l0M+nyGeUAYGIiMgc2JJARIZOJKTLyYSMqwFh3TGJupCou1/N8ECZNriZhAX6FisgvPTSSzJ9+nTd7WxBICIich22JBCRrkspmbLvYpqa3vS79cflSLR+QKgY6i/TBjcv1lSnDAhERETmwpBARHkkZ1hke3SqCgg/rD8uh89f0T1L4cF+Mn1wM6leLtBpAcHX15djEIiIiFyMIYGIcsm0ZMu26BRJzsiSH9afkIPnDAJCkK98cU9zqV8xxKkBAbMYcZAyERGRa3FMAhHlKrTvjkmV2JRM+XHDCTlwLkH37JQL8pXP72khDYoZEMaOHStffvml3RYErKpMRERErsWQQEQ5ouIwUDldflxvHBDKBvrKjLubS8PIogcEQAsBAwIREZE5sbsRESkXkjNlb0yqfLf2mGFACAvwUQGhcaXQYp+13r17S/Xq1fN8ny0IRERE7seQQESSmG6RzWcS5Zu1x+SwwSxGCAjoYtSkcvEDAlStWlV++eWXXEEBAeHrr79mFyMiIiI3Y0ggKuUyLNmy9nSifLHmmBw1WAchNMBHTXPa1EEBQVOzZk1ZsmSJCgpaQOjbt69Dn4OIiIgKj2MSiEoxDB5ef+qKfPbXETl1OVl3nzKBvmqa02ZVwpxyDLVr11ZB4cCBA9KrVy+nPAcREREVDkMCUSm2KzpFxv9xWM7EphhOczrdQWMQ8gsK+CAiIiJzYEggKqWikzLkoxVHDQNCRIiffH538ddB8PLyKsZREhERkTtwTAJRKV1Refr607LrVJzu9oqh/jJzSItiB4T//e9/Mm3atGIcKREREbkDWxKISpksS7Ys3HdJft1+Vnd7pTB/+WpIC6kRHlTsgPDFF1/kfG/kyJFFfjwiIiJyLbYkEJUyW88lyRd/H1NhQW+a05n3tnR4QHjppZdk+vTpRX5MIiIici2GBKJS5GR8ukxadUwuJ6Xrbn+3b0OpXi6wWAHh+eefzxUQNC+++KLMmDGjyI9NRERErsOQQFRKxKdlyRcbT8ue0/G62x9oX026148odkD48ssvdbf7+fmpdRGIiIjI/DgmgagUyMi6Og7ht53ndLe3rBomT3avVayAMGbMGPnqq68MA8KsWbPk1ltvLfJzEBERkeuwJYGohMvKzpa1p67I12uP645DKBvoKx/1byx+PkW7HTAgEBERlTwMCUQlmCU7W7aeS5Fpa45JrME4hPdvbySVygQU7fEtFrYgEBERlUDsbkRUQqGG/9/zKTJ5ZZQcPHdFd58RnapLl7rhRQ4IGINgr4vRd999J7169SrS4xMREZH7MCQQldCAsDMmVT5bfUx2GwxUblu9jIy6rlaxWhBmzpypu50BgYiIyLOxuxFRCQwIey9eDQjbT8Tq7hMe7Cfj+zUSX2+vQj8+AwIREVHJx5BAVMICwoFLafLZmhOy5ehl3X2C/X3ks7uaSmRYgMMDgr+/P7sYERERlQDsbkRUghyOTZOp/5yQDUcu6m4P9PWWKXc1leZVwooUEJ577jn5+uuvDQMCpjnlGAQiIiLPx5YEohLiREK6TF97Sv4+GKO73d/HSybd2VTa1ihb6MdmQCAiIipd2JJAVAJEJ2XIuOVHZb1BCwLGHkwc2FQ61C5XpMefM2eO3RYEzGJ0yy23FOmxiYiIyHzYkkDk4aIT0+XJefsMAwLGJn/Yv3GRpzqFwYMHqw9bDAhEREQlE1sSiDzYsUvJ8tjcfXI+IdUwILzXt5Hc0DCiWM/j4+MjU6dOVV/PnTtXfWZAICIiKrkYEog81NaTcfL0/P2SmJZluM8btzaQW5tWdMjzaUEBMygtXLiQXYyIiIhKMHY3IvJAi3ZHyyNz9hoGBD8fLxl/eyPp17KSQ58XQWHatGmyfPlyjkEgIiIqwdiSQORhlu2PkbG/HTbcXjbIVz67s6m0rFbGKc+PoNC6dWunPDYRERGZA0MCkQfZdSZBXllyyHB7tfAgmTaoqdQKDyrSNKdpaWkSFFT4nyUiIqKShd2NiDzE6bhUeernfZKRla27vXn1svLVkBZFDghPP/203HXXXZKUlOSAoyUiIiJPxpBA5AESUjPl8Z/2SmxKpu727o0jZVy/RlIlzL/IAQFrHaxdu1ZNdcqgQEREVLoxJBCZXEaWRcYs2C/HL6fobu/coII83q2W1ChmQNAwKBARERFDApGJYbrRd5dFyaYT8brbm1QtIw92qikNw4sWEJ566qlcAcE6KDzwwAPq+YmIiKj0YUggMrGvN52RBbuidbdVLRckD11XW9pUDhIvL68iBYTvv/9ed3tAQICMHDmy0I9LREREJQNDApFJ/Xngony6+rjutjJBfvJQ9zrSpUao+GJZZQcHhB9//FF69uxZpOMmIiIiz8cpUIlM6EhMkrxqMNWpv6+3DO9aR26oU1aCfL0LHRCefPJJ+eGHH3S3MyAQERERMCQQmUxiWqaM/mW/pGZa8mxD7597OtaSWxqGS3igj8MDwuzZs+WGG24o8rETERFRycDuRkQmgoHCY387JCdiU3W392ldVXo3qSDVwvwcGhACAwMZEIiIiCgHQwKRiXyz6YysPHRZd1uHehHSv1VlaVTe3+EBAWMQ2IJAREREGnY3IjKJTcfjZNIa/YHKNcoHy+Brq0vryMLNZFTQFoQePXoU+biJiIio5GFLApEJRCekyQu/HhCLzrIEIQG+cn/XWtKxWkihZjLKysqSJ554ggGBiIiICo0hgcjN0jMtMnrBfolNyTQYqFxTetQuKyF+3oUKCGhBQDciPWxBICIiInvY3YjIjTKyLPLS4oOy51yi7vZbmleRWxtHSMVgX4cGhDlz5sj1119f5OMmIiKiko0hgciNAeH5hQdk1WH9gcpNq5WR/2vvPsCjqvL/j39JI6GG3jsWQLogSLMu9rLqiiLYFQu6rhXLupa1ru5P3ZW1u+qqVCkiuiLVCqIoHaQjRRBCJ6T9n8/xf7MzkzshE5IwM3m/nocn4d4zYXK49875nvM951zUpYG1So9sovL8+fNt1KhRvucIEAAAQFGQbgQcBpnZuXbb2MVhA4RaVVLsiuObW6d6kU1Ulk6dOtmbb75pycnBy6QSIAAAgKIiSADK2P6sHPvjmEU2a8X2sDsqX96rufVqUiWiicqBzjzzzKBAQQHCiBEjSDECAABFQpAAlKF9WTl2y+hF9uWqDN/zFZMS7Mq+Le20I2talZRDuz29QKFatWouQOjXr98h/TwAAFB+MCcBKCN7D+TY0FEL7dt1O33Ppyb/FiCc0rqG1atcMremAoUff/zR0tPTS+TnAQCA8oEgASijEYQbRy6079f7BwhpKYl2Vd+W1q1JtYgnKh8MAQIAAIgU6UZAKcvJzbNhE5aGDRAqpSTaNf1aWbsGVaxDndQiTVTWMqc//fRTKbxbAAAAggSgVOXl5dmTU1aGXcVIuylfe0Ira1m7knWtl2aJCUULEG644QY76aSTbO7cuaXwrgEAQHnHSAJQiv49+2cb8d1G33NVU38LEBrWSLOu9dMsNSmhyAHCyJEjbefOnXb++ecTKAAAgBJHkACUksmLttjfp60On2J0QiurVz3VOtZJteoVEyMKEDwECgAAoDQQJAClYPaaDLv/w2W+55ISK9jg3i2sbrVUa52eYg2qBG96VtQAITBQGDhwoO3bt69E3jsAAABBAlDClm/ZY38au9iyc/MKnNOMgwHHNbNmtStb/cpJdkSNlEMKECQtLc1efvll9xUAAKAksAQqUIIy9mXZLaMW2a7MHN/zZ3duZO0aV7fKyRWKtJKRAoQhQ4bYqFGjfM8rMNBGaX379i2R9w8AACCMJAAlJDcvz+6buMw27Mz0Pd/3qDrW84japgWMOtdNs6SDrGSUnZ190ABBowsECAAAoKQRJAAl5JUv19nnK7f7nuvYNN36d2jgvm9Xq6JVO8hEZQUISjE6WIDQp0+fEnjnAAAAwQgSgBLw1artNnzWWt9zLepUtgu7NbGEChWsUZUka1w1+ZAChEqVKhEgAACAUsWcBOAQbdqZafdMWGoFpymbVUtLskt6NrOkxASrkpxg7WoXPg/BSzEaPXp02ABBcxAYQQAAAKWJIAE4BFk5uXbHuCWWsS+7wDlNObikZ3OrmppsiZqHUC+10HkIRQkQlGLUu3dv/s8AAECpIkgADsEzU1fZ/A27fM+d3qGhNa9d2X2vEYSqKYmFBgjXX3+9jRkzxvc8AQIAAChLzEkADmFH5ffmbvQ9d0zj6tbryNru+yZVkwudh0CAAAAAok3cjSTs2bPHLr300rDnH3jgATv22GMLHN+0aZO98cYbNn/+fEtKSrJu3brZFVdcYVWrVvX9OZGWR3xZtW2fPTR5ue+52lUr2gXdmri5B7XTEq1d7Yphfw4BAgAAiEZxFyRkZWXZjBkzwp7XqjGh5s2bZ+edd55lZGTkH5swYYK9+uqrNnnyZGvYsOEhlUd82ZeVaw9MXOG+hkpOrGADezaz1OREN1G5c700t6pRODt37rQFCxaETTHSCke9evUq0fcPAABQ7oIEjyZ33nHHHQWOt2/fvkBP7tVXX+0a/CeffLINHjzYBRovvviifffddzZ06NCgPPFIyyO+5OXl2fDvdtrq7f4bpp3XtbHVT0+ziokVrFuDNEs+yIZpNWvWdAHmOeecY0uXLs0/XrlyZTdJmQABAAAcDnEbJNStW9dOOOGEg5b7+OOPbcWKFS4FSY2yxMTfJpf279/fevbsaZ999plrvB111FHFKo/4MnLeL/blev8AoXvLWtaleU23ktGx9dMsLaloU37q1asXFCgQIAAAgMOt3E9cnjp1qquIa665Jr/BL5pbMHDgQPf9lClTil0e8eP79TvtuVnrfc81qpFmZ3X+Lc2sU900q36QHZXDBQpdunRhBAEAABx2cTuSoN7+m266ybZs2eJGFfr16+fmESQnB68y46V4dO3atcDP0GTkwDLFKY/48OueA3bnuCWWk1twy7S0lEQbeHxzS05MsLa1Klq9ysW7rRQoaCSqsM3WAAAAykLcBgk//PCD++N555137LnnnnPzBdQY82zfvt19VSARyju2bdu2YpcPZ//+/VZWDhw4EPQVkcnOzbM7P1hmW3YXrD815y8+rqnVqJxiTSonWP2KuWX6fxuruCapx2jDNUk9RhOuR+qytKSmppbvIKF+/fo2YMAA69ChgyUkJNjChQvtzTffdKvIXHvttS6tw6NJxxI6whB4zCtTnPLhbNiwwXJycqwsbd68uUz/vXjxzoJdNnf9Xt9zJ7WtZ0c1qGZVc/dZxZ07bN3OgmU02X3atGl2yimnMEoQgmuyZFCPJYe6pB6jCdcjdVmSlCbfsmXL8hskaG7A3Llz3eRPj9KMtIeBUo5mzpxpq1atshYtWuQvMym7d+/O/96jY1KlSpX8Y5GWD6csl0lVj4QeNBpBSUlJKbN/Nx5MXLjFxi31D66OqF/VBQkaQWhdrapVqFDNN0DQsrsTJ060rVu32p133kmgwDVZYri3qctowzVJPUYbrsniS4rHKCkwQPA0atTILVk6YsQIN1/BCxIaN27s0pJ0LDSFSMe8Mp5Iy5fEcE9JUYBwOP7dWPXZ0q322JQ1vufSKyW7NKOja1W0Vukpvg1/jSgNGTLEBQjy7LPPuo33hg0bRqDw/3FNlgzqseRQl9RjNOF6pC4Pp3K1upFSfCRwBECrycinn35aoLyWOw0sU5zyiE3frM6wuycsNZ95ypaYUMEuPb65HdeosrWuUTFsgKDUtnHjxgUdf+qpp+zxxx93+y0AAABEq7gLEjRB+auvvgo6pomkapzNmjXLjTJ06tQpKBVJ8xZeeukll6bkmTRpkpu7UL16dZdLXtzyiD0/bthlt45ZZFk5/g35czo3tNNbVbWm1fxTt8IFCB5tvLdu3boSfc8AAAAlKe7SjTTn4Oabb7YaNWpY8+bN89OAdu78bUbp/fffHzSS0KpVK7eD8iuvvGKnnnqqdezY0eWvabKzVz4wfSnS8ogtP23ZYzePXGj7snJ9z5/ctp6d1zzB6qYlhA0QtIfG+PHjfc9rvsqoUaOsadOmJfq+AQAASlLcBQmXXXaZ2xth+vTp9v333+cfV8Bw11132aWXXlrgNUr/0FyG119/Pf81asypvHqED7U8YsP6jP02ZMRC27E/2/d8z9a17ZoeDSx5x8ZiBwijR4+2Hj16lOj7BgAAKGlxFyT07dvX/dHIwcqVK23fvn1uJaFmzZqFfY0mkz7xxBN2zz332LJly1wA0KZNmwKrFxW3PKKf9kC4/v0FvnshSKdmNWxwzybWomqCrd/hHyBohClwed1ABAgAACCWxF2Q4KlWrVrQ3IOiSE9Pt+7du5daeUSnfVk5NnTUQjeS4KdNw2p2yXFNrHO9NKuQXTCIIEAAAADxJu4mLgOR0CpDD3/8ky3evMf3fMs6le2Sns2sc71KlpZU8HYhQAAAAPEobkcSgKJ4e84G+2jhFt9zjWqk2aDeLeyIGhWtXuWkYgUIY8aMseOOO47/DAAAEFMYSUC59fXqDPv7tFW+5+pUrWhX9GlpdSsn21G1KvoGCFdddVXYAEE7fxMgAACAWMVIAsolzT+4a9wS383SUpMTbXDvFlY9Lck61U2zxJDN0rKzs91OytobI1yAoFWMGEEAAACxiiAB5c7eAzk2dPQi36VOFQ4M6NHUaletaO1qp1qVlIQCIwjDhg1zS+z6IUAAAADxgCAB5W6i8h3jl9rKrXt9z/dv38COalDNmlVLtkZVkgqMIFx//fWFBghKMWLFKwAAEOsIElCuPDl1tX2xYpvvufZNqlu/o+tY21oVXZBQISTNSPthaM8NPwQIAAAgnjBxGeXG+/M22XtzfvY91yA91S7t0dR6NqpszaunFAgQRMceeeQRu/jii4OOEyAAAIB4Q5CAcuHzVRn2t09X+p6rlJJoQ09qZSc3r2o1UhML/TkKFG6//Xa75ppr8gOEsWPHkmIEAADiCulGiHvLtuy1eycutayc3ALnEiqY/enU1nb6EemW4DN64McbUVCAcMYZZ1i3bt1K4V0DAAAcPgQJiGub92TZ7eOW2I69Wb7nr+3dzC5qXyfin6tA4cEHHyyBdwgAABB9SDdC3NqVmWPDJiyztb+GWcmobR274fjGZf6+AAAAoh1BAuLS/uxce2Lqapu7Zrvv+XYNq9qjZ7QuMEH5wIEDboRg+3b/1wEAAJQHBAmIO9m5efbq7I026YeNvufrVatoL17YxlKSEgsECJdffrk999xzdt555xEoAACAcosgAXG3WdrHy7fbv79cY3k+59OSE234H9paeqUU3wBh8uTJ7u8//PADgQIAACi3CBIQVxZv3WfPT11pB7ILrmSkzKKnzzvKWtWuXGiA4PEChYyMjFJ/3wAAANGEIAFxY+u+bHt26mrbvGO/7/k/ndjC+rSqWSBAGDx4cIEAwbNq1Spbs2ZNqbxfAACAaEWQgLiwLzvXXv9mo81Ztc33/Lnt69mgbg19A4SPP/7Y9zXVqlWzDz74wDp27Fgq7xkAACBaESQg5uXk5tnk5dtt1LfrfM+3rlPJ7uvfKmglo8zMzCIFCF27di219w0AABCtCBIQ8+Zt3mcvz1jlOw8hNTnBnj2/jVVMSggKEDQHgQABAADAHzsuI6at23nAXv58jW3M8J+H8JfTWluzmmkFRhA++eSTsCMI48aNsy5dupTaewYAAIh2jCQgZmVk5th78zbb1yt+9T1/fsd6dnq7uvl/J0AAAAAoGoIExKSs3Dz7dEWGjZ7jPw+hVe1Kds8pLfP/ToAAAABQdAQJiMkN0+Zs2GOvzVxlmVkF5yFo/sEz5x9tqcm/7ahMgAAAABAZ5iQg5qzZmWV/n7Ii7DyEB/q3sha1KhU5QBg/frx17ty5VN8zAABALGEkATFlV2aOPfHfFbZs0y7f8+ccU9fObl+vSAFC9erVCRAAAAB8ECQgZuTk5dmzM9baV2EmKreslWbDftcq/+9DhgwpNEDQKkaMIAAAABREkICY8c7cTfbBdz/7nqtRKdn+cVE7q5Ty2zwEueqqqywt7X/Ln3oIEAAAAApHkICYMHNlhr0wbaXl+ZxLSUywf1zU1hqlpwYd79Onj40cOTIoUCDFCAAA4OAIEhD1Vv6614aNX2JZOQVDhApm9tjZR9oxDar6vjYwUPAChE6dOpXBuwYAAIhdrG6EqLZ7f5bdPHqR7c7M9j1/U79mdurRtQv9GV6gULVqVQIEAACAIiBIQNTKzcuz28YttZ+3+y91esYxde2aHo2L9LMUKAAAAKBoSDdC1Hp22mqbvTrD91yHxtXs4dNbW4UKSjgCAABASSJIQFSauOAXe3u2/0pGDaqn2vMXtLGcrAM2aNAgmz59epm/PwAAgHhGkICoM3/DLnto8nLfc5UrJtkzv29jaRVy7LLLLrOJEyfagAEDCBQAAABKEEECosrmXZl2y+hFvisZJSZUsLv7t7ZW1RNdgDBlyhR3fP/+/S5QmDFjxmF4xwAAAPGHIAFRY19Wjt06epFt25vle35A9yZ2aovKNnDgwPwAwaNA4eKLLyZQAAAAKAEECYialYwenLTcFm/e43u+z5G17equtW3QZZfZZ5995lsmNTXV0tPTS/mdAgAAxD+CBBx2eXl59sxnq+yTJVt9zx9Zv6rd0quhXXfFoLABgoIDbZTWsWPHUn63AAAA8Y8gAYfd61+vt3e+3eB7rk7VinZj70Y27MYrwwYINWrUIEAAAAAoQWymhsNqzLxN9vyMNb7n0lIS7YoeDe25O6+z6dOmhg0Qxo0bxwgCAABACSJIwGEzZelWe/STn3zPaSWjP3StbyMeudlmTZ9W6AhChw4dSvmdAgAAlC8ECTgsZq/JsHsmLLXcgiudmvZQ/n2nuvbJ03+0bz73X9aUAAEAAKD0ECSgzC3etNv+OGax714IctYxte2z/7vdvv9ypu95AgQAAIDSRZCAMrV62z67YeRC23Mgx/f8SUem28x/3Gk/fj3L9zwBAgAAQOkjSECZ2bQz04a8v8C2h9ksrXuzKjbnpWG24JvwAcKECROsffv2pfxOAQAAyjeWQEWZUGAwZMQC27gz0/d8+/pptvjN+8MGCDVr1iRAAAAAKCMECSh1ezKz7aaRC23Vr/vCbpaWM+d9W/DN52EDBK1ixAgCAABA2SDdCKUqMzvX/jh2sS3ctNv3fLPalWzg8c0tu/OttnHRXJv3/XdB5wkQAAAAyh4jCSg12bl5dvf4JTZ7zQ7f8/Wrp9rlvVtaSlKCdWlW18Z9MNa6dOmSf54AAQAA4PAgSECpyMvLs4cnL7dpy7f5nq9ZJcWu6tvS7apct1KitaiebOnp6TZ27G+BAnMQAAAADh/SjVAqXv96vY2f/4vvuaqpSXZ135ZWNS3ZaqQmWue6aVahgrZQs/xAYdOmTXb00UfzvwMAAHAYECSgxH2/fqf9c+Ya33MaObiqX0urWaWiVUtJsGPrp1liwm8BgkeBgv4AAADg8CDdCCVqx74sGzZhqfltppyYl22DezW3+tXTrHJygnVrkGbJIQECAAAADj+CBJToPIS/TP7Jfy+E7P2W9/HjNvXVxyw10ax7gzSrmMjlBwAAEI1IN0KJGfHdRpu67NcCx3Oz9lvO5Cdtw6I5tuqHb6xhlWQ78dlnqHkAAIAoRZCAErFk827729RVvgHC7gmP2fbl/9v/4K03XreUxAR7+umn8ycsAwAAIHqQ74FDtvdAjt09fqllhUxEUIDw65hHggIEz6uvvmqPPPIItQ8AABCFGEnAIXv8vyts9bZ9QcdyD+y3X0Y/bJlrf/R9Ta1atezCCy+k9gEAAKIQIwk4JO9+u8EmLPgl4gBh4sSJ1rZtW2ofAAAgChEkoNimLfvVnpqyMqIAoXbt2gQIAAAAUY50IxTLgo277O4JSy2vQIDwkGWunR82QJgwYQIjCAAAAFGOkQREbH3Gfhs6apFlZufmHyNAAAAAiB+MJCAiO/dn29BRC23b3qyIAgTNQWjTpg21DQAAEAMIElBkB7Jz7baxi23lr/uCA4RRf7HMdQt8X0OAAAAAEHtIN0KR5OXl2V8mL7dv1+4ocoBQp04dRhAAAABiEEECiuSlL9bZpIVbIgoQNEmZFCMAAIDYQ7oRDmr68l9t+OdrIx5BOProo6ldAACAGMRIAgq1Zts+u3fisqBje5d9SYAAAAAQxwgSENbeAzn2p7GLbc+BnKDjVY45ydL7DCxQvm7duowgAAAAxAGCBISdqPzQ5OX209a9vucH3XyHXXnLnUEBguYgkGIEAAAQ+5iTAF//+XaDfbx4q++5zs1qWP92daznqfda/cpJ9tprrxEgAAAAxBFGElDA3LU77Jmpq3xrpkF6qg3o3sSOrV/JEhMq2N13321fffUVIwgAAABxhCABQX7ZlWl3jFtiuXkFKyYtJdEG92puPRtXttSk/106tWrVohYBAADiCEEC8u3PyrFbxyy2bXuzLDdzr+Xs3p5/roKZXXxcUzuxRXWrXjGRWgMAAIhjBAnIn6j850nLbdGm3S5A0D4Im969x7J3b3PnT25X385qW9saVEmmxgAAAOIcQQLyd1T+ZMnW/AAhc/0iy972s21+d5i1qHzALu3W0Fqnp1BbAAAA5QBBAuy/S7a4HZUDAwSPAoVlr91p9XN3WIUKSjoCAABAvCNIKOcWbdxl901c5hsgeDasXmFDrr/usLw/AAAAlD2ChHK+ktGNoxbZ/r17wgYIUq9ePXv22WfL/P0BAADg8CBIKKf2Hci2ISMX2a/bdxw0QPjwww/tiCOOKPP3CAAAgMODIKGcrmR0x4Rltnz9Fvtl5INhA4T69esTIAAAAJRDBAnl0ItfrLeZC9f/FiD8vDhsgDBx4kRGEAAAAMohgoRy5stVGfby1MUECAAAAAgrKfwpFCY3N9dmzpxp8+fPt8TEROvWrZv7E8227smyu0Z9Z5tGMIIAAACA8AgSimHr1q02YMAA+/bbb4OOn3nmmfb6669bxYoVLdrk5OXZbaPn2/J/31doipEmKbdu3brM3x8AAACiB+lGxXDNNde4AKFhw4Z2ww03uL+np6fbpEmT7P7777do9Mb3223m838iQAAAAMBBMZIQoa+//tqmT59ujRs3dulGNWvWdMcVLPTr18/eeOMNu+OOO9zSodHiwx/X25sP32qZPy/xPV+/fgP78MOJjCAAAADAYSQhQpMnT3Zfb7zxxvwAQVq1amWXXHKJZWdn2yeffGLRYtHaX2zoVYPCBgh16ynFiAABAAAA/0OQEKEFCxa4r7179y5wrk+fPkFlosF/Zsy3zK3rfM/VrFPPPprEHAQAAAAEI90oQps2bXJfmzRpUuBc06ZN3dfNmzcf9Ofs37/fysIDF/Wy5LxX7fm7rrHc/bvzj1epWcc+HP+BS5sqq/cSDw4cOBD0FdTj4cT1SF1GG65J6jHacE0GS01NtaIiSIhQZmam++q3gpF3rCiN7g0bNlhOTo6VhYE9W1m9v79o9992o+Xs322p1WvbGy8Pt+TkZFu3zn+UAYUrSiCIg6MeSwb1WHKoS+oxmnA9UpclSUv2t2zZssjlCRIi5AUCChYqVarkG0AUJUrTykhlGUWfcqxZy3fftSG33Gb/eesN69L2yDL79+OJ6lIPbU1MT0lJOdxvJ2ZRj9RjtOGapB6jCdcjdRkNCBIipL0EFi5c6Hrga9SoEXRu7dq17mvdunVLdLinpPQ7vrut+HGOJSQwFeVQKUA4HP+H8YZ6pB6jDdck9RhNuB6py8OJ1mKEjjnmGPf1888/L3Bu1qxZ7mv79u0tWhEgAAAA4GAIEiJ0+umnu68vvviibdu2Lf/4ihUr7L333rOkpCTr379/pD8WAAAAiBqkG0WoR48ebtO0GTNmuGVQzz33XMvKyrLRo0fb7t277dprr42qjdQAAACASBEkFMOrr75qAwYMsLlz59rw4cPzj59xxhn26KOPFudHAgAAAFGDIKEY6tSpY59++qkbTZg/f77L8+/evbv7AwAAAMQ6goRiUmBw4oknuj8AAABAPGHiMgAAAIAgBAkAAAAAghAkAAAAAAhCkAAAAAAgCEECAAAAgCAECQAAAACCECQAAAAACEKQAAAAACAIQQIAAACAIAQJAAAAAIIQJAAAAAAIQpAAAAAAIAhBAgAAAIAgBAkAAAAAghAkAAAAAAhCkAAAAAAgCEECAAAAgCAECQAAAACCECQAAAAACEKQUE4kJiYe7rcQN6hL6jGacD1Sl9GGa5J6jDZck8VTISMjI6+YrwUAAAAQhxhJAAAAABCEIAEAAABAEIIEAAAAAEEIEgAAAAAEIUgAAAAAEIQgAQAAAECQpOC/It6sW7fOPvzwQ9u0aZPVrVvXzjzzTGvevLmVdxkZGfbpp5/a999/b7m5ufbggw9aWlpa2PJ79uxx9bh06VJLTU213r172/HHH19i5WNRXl6eff311zZ//nz7+eefrVq1atatWzfr06ePVahQIezrZsyY4V534MABa9OmjbsmC6v7SMvH6vU4depU++mnn2zfvn3uHj3ttNOsXr16YV+jspMnT7atW7dao0aN7KyzzrKGDRuWWPl48MILL7hr84gjjrCrr746bN1PnDjRVq5caVWqVLGTTjrJOnfuHPZnRlo+Fj3wwAOWlZXle+6CCy5w93monJwc90z97rvv3DO1U6dO1r9/f0tOTvb9OZGWj/Vn5eeff26zZ8929/fRRx9d6HPsxx9/tM8++8x27NjhngXnnHOO1axZM+zPj7R8rPnb3/7mnluFefjhhy0lJSXo2C+//OLuVbWDatSo4a4v1X04kZYvD9gnIY795z//sT/96U+WmZmZf0wP4KeeesquvPJKK4+2b99ugwYNco3O7Ozs/OOrV6+29PR039csXrzYLrroIlu/fn3Q8fPPP99eeeUVS0pKOqTysWj69Ol28803F/gdpUuXLvbuu+9a/fr1g47rOhw8eLB98sknQcdbtmxpY8eOLRC8Rlo+VunD7cUXX7T9+/cHHa9YsaI99thjvo1bNX7/8pe/uIaWp1KlSjZ8+HA799xzD7l8PPj4449twIAB7vt+/frZ+PHjC5T55ptvbODAgQUaINdcc41rmBxq+Vil4DTwcyPQ888/7+7L0MBJz7w5c+YEHe/YsaO7V2vVqnVI5WOZGpz6vP32228L1LGuydBG6H333Wf//Oc/g46pwfrWW2+5DphQkZaPRQrCV61aFfa8AqLly5cHbZj20Ucf2fXXX2+7du3KP5aQkGDDhg2zO++8s8DPiLR8eUGQEKd++OEH18OlRoEaAbrJFi5caKNHj3a9vGp4+fUGxbs1a9a4DyIFBKeccop9+eWXtmHDhrBBgnqve/To4XoN27dv7+pSH3Bvv/2267XRw0MP6eKWj1Uvv/yy3X333e53PeaYY6xBgwa2du1aGzVqlBtF6dmzp+u1DnTXXXe51+mBftlll1nVqlXtgw8+sEWLFrm60oiBHsrFLR+r1KOoXn41ZFu1auWCV12XX3zxhTuve/W4447LL68Rh9///vfud//DH/7gGhlqbE2aNMkFFnqtfk5xy8eD3bt3u2tTv+uUKVN8gwTdj127dnUNfpX93e9+Zxs3bnSdK3v37nWNfjX+i1s+lqkBq1Gma6+9tsC5E0880Y3oBdL9qZFTjVBdcsklrrH2/vvvu+ftqaee6p4Lh1I+Vu3cudM11PV7qT7PO+88V7e63xUM6do54YQT8svrc2Lo0KHuvlQw2qRJE5s2bZrNnDnTfT5p1CVwhCDS8rFK9aT7L9SCBQtch5Qa908++WT+cX0W6ZmpURtdTxrFV5Chsnq+6usZZ5xR7PLlCUFCnLrqqqvcQ0gNOUXCgT2KGkpWKoMeyuWNegnmzZvnGrHq0dcDWn8PFySMGDHCPYAUUKmnwRsK18NJH5Ya3lQPhnpli1M+Vul30FB548aNC4yinHzyya7RpKBUjQBvBOfII490da4PMKV/iHrPFax5D3vvQRxp+VimlLcOHToE9YIFBkk33HCDPf7440FBhQKIZ5991t3nnvvvv9/+8Y9/2OWXX27PPfdcscvHg3vuuccmTJhgI0eOdKl+fkGC9yxUSoGehV6KnK43pWuoUadr2DseaflYpobsscce6wLJg1FKpRpYderUcdeZ0lpFnSO9evVy6V4aeVQ6UXHKx7J7773XjRIqqFQHndLTPFu2bHGdeIEjrl6P+XvvvWenn356/nGNJo4ZM8Z9luszvbjl4811113n7nHdg3qGBt7///rXv9wIzt///vf8497ns0a71XlS3PLlSex3w6EA5Xcq11O58LfcckvQOV3w1atXd70N4YaT45l6o9WzU9SUH6UsyB//+MegXFn1nqvxpV7zWbNmFbt8rFKjPTRAEPUwdu/ePb8XzaN8WeU4K+XKa/CLrtFbb701qO6KUz6W6YM+NEAQBUPe6JRHvWlfffWVa8SpcR/o9ttvdz8nsF4iLR8PFHQpre/pp59293s43kiXRvcCG/Z9+/Z1nQgaYdSIbHHLlxdeOqCuL6/BL+p0USNOAq+xSMvHKj2/1AOua0XBeGCAIAqSAgMEBU9q8GuUNLDBL166S2C9RFo+3iioVEeAgoPAACHw9w5NE9JIqtJUNcKyefPmYpcvTwgS4pCGztRjrhuncuXKQefUk61ebgUIGvJE4ZTaIn6TjtXrJeo9LG75eKSeQAWigSksXr2oMRVJPRa1fDzSB2Dg7ytLlixxnQC6h0MDC+Uht23b1n2YeTnzkZaPdeqZVcfI2Wef7YLywugaU9CpXsJQXp1rxKq45WOdgvyXXnrJ/vznP7t5CKE59R7vPizqMy/S8rFK14KCdF0vehYqyNT8I6WlKXAPVdgzT2lzmqfh3c/FKR9vNIKgkWWlrgVSR5zSu5o1a1ZgYQYFbN5159VfpOXLm9ifQQnfGfripXqE8o5ruBMHr0ulBqlBFa4evfouTvl48/rrr7tUpEceeSRopQnvd/YbfdB8BuXLh9ZjJOXjgYa7lfamxpkaZMuWLbNLL73UjaZEcm9rtSmVq127dsTlY50mcGqiqFItDtbLq55I9RL6zWvxrjvvGRlp+Xig6yI0VUXz3HSPB6Zmer+z3zXm91kTaflYpXtZ1NOvhmxo6pZ6/9944w0XeB7smefVjVYx0vNB9R9p+Xjz73//283F0AT4QLp2tJrUwdo/Xv1FWr68IUiIQ94qKaHLgXl0YwWWQ3gacQlXj97DPTBtK9Ly8UQpQmpUaJ6AVj4K5P3OfnWj3hpdk6H1GEn5eDBu3Di36lZgI0I5xYGpLd7v7N3DB7vGIi0f642yJ554wk1gDExj8eM9+8LVS+gzMtLysU6jTpqvpXkB6vRQwKqlIZWXrbxtLSDgKaxuvOsrsF4iLR+rNC9LNIKwbds2t9KW5lmp11q94Dr+0EMP5c83KuyZV9i9XR4/b+bOnetGm9SBEtohxzOyZBEkxCFv7eVwD1rN4A8sh/D0oPUe9qG844H1GGn5eKFeMk2K1eRs9Y6FTt4s7MNfw+E6Htiwi7R8PBgyZIibAKsUBaUq/Pe//3WTbtUg81JcvHrx7uGD3duRlo9lmmOhybZa4vhgvIUDDlYvXrlIy8c6BQOhS3MqPVWTtjWfTT3UXh64d+341Y3fMy/S8rHK+x2UzqfVmrRqjueKK65wq2NpmVKlIGn+2sECpNC6ibR8PFG9SWiqUVHqJdwzkvaSP+YkxCGlY3hzE/yoJyOwHAqvSz08/IYa/eox0vLxQKsMaRKiAoR33nnHt4ewsGtSxzTcG1qPkZSPB1oe8cYbb3SjB5rwqIaFAgblhHu8nNmD3dvehMhIy8cqBVQaydJEZa1U4v3RnjCyYsUK93elKHg95Zo4qiVM/TYNC62XSMvHOr8NpFq3bm0XXnhhgTkDhd2r4Z6RkZSPVd7voHSVwADBW6xAS3ErH15zuA5WL3re6bgmP3uT8SMtH0/LG2vlRtWrPnNCaZEGpQQW9ZkXafnyhiAhDilHUWsjK6dUw5yhN5iG6vTw0KZUKJzySUVr8ofyjnllilM+HnLAb7rpJtfDqDW7ww19U4+R06o56uXS5EPPUUcd5XodtalXaBqBVtfRiidaK93LQY60fKzSjvLeiJbmdnh/FGyJNv3T37UsceA1qQa/9okIpSU4vTLFLR+PvJW2AleHK+zeDlePkZSPVe3atXP1FLg6WSDvuDfHxfudtZyn34pdmlugVfK8UdpIy8cLBQhamEX7a/jND9IzU2ldCr5CF2fRwgba+VqvU90Up3x5Q5AQp5QXrk1ANJQZSPmP6r3QPgnxsPNvafNWSNGKFIGbuegDTsvMKhgLXH0m0vKx7NFHH3UbwylF5s033wxa8jWUeny00pY2UFKD1aMgVuv3y1lnnVXs8rFKqwopfcOP0rY0KhXYq6o6Ud1oEm3gDr/qOdSOyvqqlX2KWz5WafUm7U4d+sdb0lC94Pp74DKw3r2q44GpBlrPXkuZ6jVa/am45WOVgiBN/g6lxpLWjZfAPQw0d0aNKHUSaNGCwB5YLUWrRmrgvRpp+VhVrVo1d+9pYqyXHhO45KY68bRYgIJ00aR4BRZqqOp5GhhMeJ/jgfUSafl44dWlNo8Lx7tXQ3eZV6eWRgP1GRw4lyHS8uUJm6nFKe34q3xm5SUqn1kPdS3hpYmRSgdR74N6GcsjNZa8JR+1Cooe4sqn93rBle6hJTy9/HdtDqaeGaVuaO16Nbj0kNfDWJMktfeEJ9Lyseq1115zOeDqhVEOuN86/6pT9dB49LsrSNX1p4aChsGVJqKcXfWae0t+Frd8LNKEUO0r0aJFC7fHhAIC9QDq+vF6tbTGemDurUYCNXKjTgAtz6fUEB1TQ1UNk9mzZwcNjUdaPh53WPfbTE0NfW1ypQnPanCpQacGga4x3ccK0gJXloq0fKzSJnvDhw938zv0eyr413Wq60S0e7dWOAqkvUuUyqV7VPeqngcatVFHiSbsahTnUMrHKt3HmnugESjde3oeKq1FHQMK0P/617+6kViP6kArmilQ0udI06ZN3b46CqaUIaAOk8BlzSMtH+uU5qYGu+oycFQwlD7f9VxVp5ICd923ahNppEoBqkYcA5eOjbR8eUKQEMemTJniGqS//vpr/jE1fvUADt18pTzxdqkMJ3CnYNEwpBppeuB79FAeOnRogZGa4pSP1VGEwJ5pP+ph9TYEEzWk1LOrACOQHsiayxC6BGek5WORdpbWjsoaZQrswRL1XKnBpt1TQym41YZ9Gnb3KLdWDVW/9ecjLV8eggRR41f3qr56NMKqnZW9TfsOpXwsUoeGdgpWIymQfk/13moFqdDJsAqg9FkTWscasVZAETqhO9LysUydGXr2B44sK/DS9aL7O5Q+nzUPKTBNSZ0IeuZp5OBQy8cybxd67WKt4Kgwc+bMcRPEvTkfoutKn1t+r420fHlBkBDnlFqkXgv1vmrinXq/4m0iU6SUpxz4wA41ePDgArtjqtdHG+Aoh1u952pYafOVcCItH2vUqxhucyWPhrrVsxVKjQ/VjXrXvB2aC8ubjbR8LNL9OW/ePJc7rwaENl9SCk24OR6iESqlsakXTKNWWrKysJVMIi0fDzQqo8aSelWVFudHIyxKpdF1pvteo1SFjaxEWj4W6fmlkSYFQwostZKYgnN9hhRGHSwapRIFZ/pTkuVjleYCqldac2c0/0ej/IVdMxrd1r2q61efG7rGCkvnjLR8LC+SoeeYGvNFCSQ1D0v1ovQ5dbqo/VNY2lCk5csDggQAAAAAQZi4DAAAACAIQQIAAACAIAQJAAAAAIIQJAAAAAAIQpAAAAAAIAhBAgAAAIAgBAkAAAAAghAkAAAAAAhCkAAAAAAgSFLwXwEAsejTTz+1CRMmBB1LSEiwypUrW8uWLe3kk0+2Fi1aWCz75ptv7J133rEOHTrYtddee8g/7/PPP7cRI0ZY586d7aqrriqTfxMAYgVBAgDEgQULFtjbb78d9nyFChVs0KBB9swzz1hycrLFop9++sn9jmeccUaJNNiXLVvmft6OHTvCBgkl/W8CQKwgSACAONK4cWO766673Pe5ubm2fv16++ijj2zRokX21ltvWcWKFe3pp58+3G8TABDlCBIAII6kp6fb4MGDg44NGzbMLr/8cvvwww/tjTfecH+vWbPmYXuPsaRHjx72/PPPW9OmTQ/3WwGAMkWQAABxLjEx0Y0uKEjIzs6277//3s1R8OjYZ599ZnPmzLHt27dbjRo17LjjjrOTTjrJvTbU1KlT7YMPPnBlLrvsMpfq9PHHH9vmzZtdkPK73/3OunXr5vteNAdAcwHOOeccO/XUUwuc37Ztmz344IOWlpZmTz31VES/57x582z27Nm2bt06y8zMdKMq+j3btWtXoKwCJdWD97qhQ4fmn6tSpYo9/vjj7vutW7e6etHPO+GEEwr8nEOtO6UzTZo0yY34aP5I79697ZRTTono9waA0kCQAADlgBrMnp07d+Z/v3jxYjfKoPz8UG3btnUpSq1btw46vnDhQpenrwbykiVL7J///Kfl5eXln1c60yWXXGIvvPCCJSUFf8x8/fXX7rWaTO0XJOzatcudr1atWpGDhKVLl9qAAQNs1apVBc79+c9/tksvvdT+7//+z1JSUoKCFQUksnbt2qD5HBpl8YKEwuYkHGrd/fzzz+53zMnJyT+v99m/f383WTpW544AiA8ECQBQDqhx6qldu7b7qp7/s88+2/WWqwdcPdvNmjWz1atXu4as5jGcddZZruffe00gzXXQpF+t/HP++ee73n+V1YjFe++9537mY489Vuq/2y+//GJr1qxxPf3HHHOMC4j27t1r3333nXuP7777rtWpU8ceeuih/NcoCNAIwMiRI61Tp05BE5c1b+NgDrXuJk+ebBkZGW5lJY126GcogJo4caJ98skn9q9//StodAMAyhpBAgDEuU2bNtm9997rvq9atap17drVff/Xv/7VNXLr16/vGsyNGjXKf416zZX2otc++eSTvpOdFSCceeaZrsfcS60ZMmSIvfrqq3bHHXe4hu51111nzZs3L9Xf7+ijj7Yff/wxaLQkML3noosusldeecVuv/12N0IhF198se3Zs8cFCZpvEDqP42AOte4UIFx//fUuWNFStXLTTTe5kQ/NgVBgQ5AA4HBiMzUAiCNKYVHjUn/U6FQPv4KC+fPnu/P33XefVapUyaW7jB071h278847gxq5ol7x2267LT81RyslhVIqkRrLobn3V199tUu30WtGjx5tpU2jBH4BgmhuQM+ePd3Iwrffflsi/15J1F29evXs4Ycfzg8QPN6IhlKo9O8AwOHCSAIAxBFNnvXbL0GNUk3WveKKK9zflUe/e/du9/25557r+7MUYGgEQnMYVqxYYUcccUTQeaX2+I0SaE8G5fAr5cabHFwWNEdg5syZLvVIowRe43zjxo35cw9KQknUXceOHX3Tmpo0aeLqT+9d/4YmggPA4UCQAABxuk+Ct1KPdlpu37590CRipcqIVtTxy5kXpdKoIauVfVQ+tKFb2LKgauwG/julSQ1xpe4oz78wmhRdEkqi7mrVquX7Oo3KaHRBk5kDJzQDQFkjSACAON8nwY+XIpSVleV6rUPTXkSNVC/lxW85TzWAw/HO+b2uMIGrJBXVrbfe6gKE1NRUO+2009ySp5oI7K1mpEnUX331lZWUkqg7AIh2BAkAUA41aNDAfT1w4IBbo99vVEDpOV5vtlc+kNJ6wtEqP36v80YzNEfAj95LpKMIEyZMcN9rEnLfvn0LlNFqQdFWdwAQ7Zi4DADlkPYp0DwF0epEfrQ7s2hirpc+FEh7JPhNBt6/f7+NGTPGfa9Nw/zSbLT/gJ8pU6ZEvPypGuNaftUvQNAqQrNmzfJ9rbcPgUYEyrruACDaESQAQDmljcBk+PDhBRrS06ZNc0uZSuAeAqG0rGjgvAOl2GjCrvYRUM7+hRdeGFTe24lZeyloH4NA48ePdxuzRbqykdJ99u3b514fSL38gwYNCto8LlDdunXzgx2/FYhKu+4AIJqRbgQA5ZSW6dRmYwsWLLBzzjnHevTo4ZbvVBqRNvbS/IAuXbqEXa9fPepaWlUbgqkXX0ur6nXr1q1z5x999FG3e3EgbXjWpk0btxKRdhbu1auXK6OGulZD8s4VVfXq1d2mZgoQ1HDv3r27W3FJS8FqZSXNjdCux34jFxrl0ORi7dSsZVK9FYc02dvbcbm06g4Aoh0jCQBQTilFRz362mxMNLn3/fffd1+1DOeAAQNs3Lhx+ROA/RrZL7/8sis7adIkGzVqlAsQNIKgDcSuvPLKAq/RJF5tFKZ9FJTmM336dLfngAKECy64IL8HPhLPPfec27xMZs+e7eYmfPHFF27jOKX9KAAIN8n7qaeecoGB9iXQ67R8rPY2KO26A4BoVyEjIyPypSQAAFFFPdpK39GqPupZj9SWLVtszpw5LodfP0NpQeGW93zhhRfsgQcesEsuucSl22gOgnrPtcOwXnv88ce7BnphlN6j+QwrV650DW6NRmgCsPYGUNCgxrUa2oG034Aa/yqnEYlQGo3QyIYa6VoK9thjj3UTpdVwX758udtUTisf+e0tod9ddaD5DQoatCNzUf7NSOtu4cKFNnfuXGvVqpUbRfGjQEUjEXoPfnspAEBZIEgAAEQkNEgAAMQf0o0AAAAABCFIAAAAABCEIAEAAABAEJZABQBE5KSTTnJLj2ryLQAgPjFxGQAAAEAQ0o0AAAAABCFIAAAAABCEIAEAAABAEIIEAAAAAEEIEgAAAAAEIUgAAAAAEIQgAQAAAEAQggQAAAAAQQgSAAAAAFig/wcPoD8vI5StOgAAAABJRU5ErkJggg==",
      "text/plain": [
       "<Figure size 800x800 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plot_gain(df_preds, outcome_col=\"y\", treatment_col=\"w\", treatment_effect_col=\"tau\")\n",
    "plt.title(\"Cumulative Gain — JAX vs TF DragonNet (IHDP)\")\n",
    "plt.tight_layout()\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5bfb1ee2",
   "metadata": {},
   "source": [
    "> **Note on training time:** JAX train time includes one-time JIT compilation\n",
    "> on the first mini-batch. TF time includes Keras graph building overhead.\n",
    "> Both are measured end-to-end from `fit()` call to return."
   ]
  }
 ],
 "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": "3.13.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}