{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "6c4d4faf-ab84-4a72-a80e-535b211747cd",
   "metadata": {
    "tags": []
   },
   "source": [
    "# H2O Sonar Model Validation Explainers\n",
    "\n",
    "This example demonstrates explainers that are based on [H2O Model Validation](https://github.com/h2oai/model-validation) (MV) tests:\n",
    "\n",
    "* Adversarial Similarity explainer\n",
    "* Backtesting explainer\n",
    "* Calibration Score explainer\n",
    "* Drift Detection explainer\n",
    "* Segment Performance explainer\n",
    "* Size Dependency explainer\n",
    "\n",
    "Please see H2O Model Validation [documentation](https://docs.h2o.ai/wave-apps/h2o-model-validation/) for MV tests features, purpose and use."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "b2ecd37e-c508-448a-8528-b53fe375a04c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABckAAAXSCAIAAABM2vDvAAAABmJLR0QA/wD/AP+gvaeTAAAgAElEQVR4nOzdd1QU19sH8LtLWTqCNFkQsABWjKKoBAFRY0HFqFE0CtgbGo2ixti7sSsRJAYbaIwGeG3Ya1QiohEQpFsQqdJEysK+f0wymd9scdlFinw/J+fk7sydW4aZ6+yzM3c4QqGQAAAAAAAAAACAXLgN3QAAAAAAAAAAgCYMsRUAAAAAAAAAAPkhtgIAAAAAAAAAID/EVgAAAAAAAAAA5IfYCgAANCXe3t4chnPnzjV0iwghJCMjg9mqzp07N3SLAD6txnkmfmrNs9cAACALxFZAjE9x6YDLEYBGKCAggCNi8eLFkvJv2rRJNP+GDRvqs80Anyux56NYJiYmDd1YAPh8YPABqBOIrdQTjFkA0FT8+uuvHz58EF1eXV0dGBj46eqNi4tjDoYuLi6fri4AABoGHwCQEYYLkAKxFQAA+B/v3r07ceKE6PKzZ8++fPmy/tsDAAAAANDIIbYCAABs/v7+Mi4EAGg+9u/fn8vw1VdfNXSLAACgsVBu6AZAc7F///7t27fTH3V1dRuwMQAgXUxMzIMHD3r37k0vef78+bVr1xqwSQDNjampaXx8vNhVXC5+G2sYWlpaWlpaDd0KgE8Lgw+AfHB6NAxTU9N3EiQlJTV06z4JLS0tAwYVFZWGbhEAsKmqqtJp1l0q/v7+QqGQSvN4PFlKe/78+bJlyxwdHU1MTHg8nq6ubocOHaZPn37z5k1WTmqu6y5dujAX3rp1i/lIs62trZS6Hjx44OnpaWZmxuPxTE1Nx48f//jxYyn54+PjFy9e7ODgYGRkpKqqqqur265duwkTJoSGhgoEAikb5ubm+vn52draqqurGxoaurq6Hj9+/CM74t8Nt2zZ4ubmZmZmpqGhoaqqamRk1KFDhwEDBnz33XcnTpzIy8uTpRxoPjgcTgsJdHR0qDyFhYUWFhb0acLlcm/cuMEs5NWrV7q6unQGdXX1uLg4eq3YaebPnj07fPjwVq1aqampWVhYTJ06NTExsbaNz8zMPHz48Hfffefq6mpra2tiYqKmpqampmZsbNynT58FCxY8fPhQ7IZim3Tz5s2xY8fy+XzpJ7gcldZ28JFlYn45hhdFek0wvEBdk2XwYartMS/2gL9w4cLw4cNNTU2VlJS++OILKuf48eNZOYVC4eHDhx0dHXV1dfX19fv16/fbb7/RJRcWFq5fv75z584aGhp6enqurq4nT54UbYB8A1RthwucmM2REOrFgQMHmLudz+d/dJN37961bt2a3oTD4Vy/fp2Z4eXLl8wBTk1NLTY2llrl5eXFrO7s2bNCofD//u//3N3dqS85rVu3njJlSkJCgtiqxW5Oef36dXBw8IIFC1xcXGxsbIyNjXk8Ho/HMzIy6t279/z58//666/alilp7Y0bN8aMGWNqaqqqqtqqVatx48bFxMRI2WOJiYlLly7t27evsbGxqqqqjo6Ora3ttGnTbty4IXuTzp8/7+7u3qpVKy6X261bNzpzTk7O5s2b+/fvz+fz1dXVVVRUDA0NbW1t3dzcFixYEBoampubK6VtAI0Ta2jy8PDQ1NSk0jweLycnh8pWWlpKjzZKSkqenp7MrdavX88q9v3799OmTeNwOJL+6RkyZEheXh6dn3UyimVjYyM289mzZ1esWCH6S5qKisrp06dFu1xSUvLtt99KaZuVldW9e/fE7q579+61bNlSbHdYP/F16tSJueHFixe1tbWld3Dfvn1y/x3h8yDHpYJQKLx9+zbz+LeysiopKaHXDhw4UMphxjqbwsLCJk2aJHpw8ni8o0ePsuqV/s/6rl27pB/whJAJEyaUlZVJLzYiImLu3Lmi24o9weWotFaDz0d7LffwokivMbyA4uQbfITyHvOi59GCBQuYS+zs7Kic48aNYy4/c+aMu7u7aC3Tp08XCoVxcXFt2rQRXTt79mxWA+QboGo1XODEbJ4QW6kn9XzBpMjVkujmzEuHurpakh5bqdVVhbD2X+QkNUnSyI7xET5XrKHJy8trxowZ9MfNmzeLZhs5cuTSpUuZW7FiK2VlZcyHiSSxtrYuKCigNlEkttKzZ09Jm+jo6GRnZzPbVlpa2r1794/WpaKicuXKFda+SktLE/t7HYVVLDO2kpOTI8tTkBhDQO6vNz/88ANzwzlz5ogtcOjQoawNWWdTu3btJB2fXC738uXLUraVI7ZCCJk4caL0JtnY2EjaVvQEl6PSOoytKDK8yN1rDC9QJ+QbfOQ+5lkHvJubG2uTrl27UjlZsRUrKytJtWzfvt3S0lLS2vDwcGYD5BugZB8ucGI2W5hvpVFzcnJatmzZpk2bqI/p6elLly6lbtQPCAi4cuUKnXPo0KHz5s2TVM6SJUtSUlJEl1dUVHh7e5uYmLDCNAoKDQ3lcDgy3icvlp+f3/Pnz0WXV1VVTZkyxcnJycjIiF744cMHNze3Bw8eSCnw4sWLffv2ffDggZ6enqQ8u3fvZk0nIRQKCSG5ubnjx48vKSmpdTcAmqC5c+cePHiQSgcEBPj5+XG53J9//pnOMG/evKtXr0opYeHChczz0cDAYObMmR06dMjLyzt06FBsbCy1PCkpaebMmadOnSKEDBo0qEWLFvn5+cxxg8/njxkzhv5obGwstjrqxl0dHR1ra+u0tLSCggJ6VXFx8ZEjR5YsWUIvmT9/fkxMDHPzXr16de/ePS8v78KFC2VlZdTCqqqqsWPHpqSkMO9SmTt3bnFxMXPbzp07d+nSJTMz8+7du6ximSIiIoqKiuiPAwcO/Prrrw0MDMrKytLT0588eXLr1q13795J2hyarczMTEm/GQQHB3t7e9Mf165de/Xq1b/++ov6eODAgbFjx1paWjIPfmNj4+DgYOk1UpcK3bp169Chw6tXr/7880/hv08C1tTUTJs2LSUlpVaP9Hbp0sXFxaVDhw56enrKyspZWVlXrlw5e/YsnSEkJOT777+n7/8XRV0MyHiCy1GpIoMPiyLDi9y9xvACn4KMg09dHfOis7nRIw9Leno6IYTP5+vr6yckJDAfOFq8eDGVaN26tZaWVmJiYk1NDb127969I0eOZJVW2wFK9uECJ2bz1cCxnWaDFQ+WIjg4mLlhVVVVr1696LUcDufGjRvp6enMqdSMjY1ZP92IDax269bN09Pzyy+/ZA2XrVu3rqyslLK56H0rXbp08fX1/fnnn0+cOPH777/v3bt3+PDhrOpYz+/U6r4Vio6Ojr29vb6+Pmv5tm3bmNvOnDmTudbAwGDFihXHjx/fvXs365HIsWPHfrRSpi5dugiFwqCgIObCgQMHHjhw4Pfffz9y5MiaNWs8PDyoeA1iz9AUid63IhQKv/zyS3pJRETErVu36I82NjY1NTVS7ltJS0tTVv4vas/n8zMzM+m1lZWVzAGNEPL06VN6LR12oTg7O4tts+iZ6+XlRd0rV1JS4uzszFzl7u5Ob5iSkqKkpMRcu3PnTnptcnIy6xvUDz/8QK8VnW+CvqlHKBRevXqVOVUN+d/7Vpj3FPD5/JqaGlaPqqurr1y5cv/+fdn+aPDZkvtSQSgUJicnMy8M2rRp069fP/ojh8O5cOGCaI2iZxM1sxLlwoULzNOZEHLixAlJ27L+WX/w4EFcXJzYbrIGkB9//FF6k2Q8wRWpVMbBR0qvFRleFOk1hheoE3IMPooc86IHfMuWLYOCgrKyskpKSp4+fXro0CEqJ+u+FULI+vXrqeP88uXLos2j2xASEsJcrqamVlVVRTdA7rFCKNtwgROz2UJspZ7U8wWTIldLopszL5jq6mrpo7EVGa8qFPkiJ+PIjvERPmNiYysnTpyglwwaNOibb76hP+7Zs0cocrIzYys7d+5krtqyZQurRlawct26dfQq+WIrrVq1qqiooNeyJvJkxjiYryojhNjb27NKZt6eQwjp0KEDvYp1/zBzMiaKr6+vpHo3bNhAL1dRUTl+/HhxcbHkvwk0X4pcKgiFwkOHDknKv2DBArE1ss4m0ZNi6tSpzAyTJk2StC3rn3UpHj16xNxwyJAhUpok+wmuSKWKx1YUGV4U6TWGF6gTcgw+ihzzrANeRUXl0aNHYhvGiq2wamFOTEkI6devH72qpqaGeZM7IeTVq1ey7ArpY4VQtuECJ2azhfcENQHt2rXbs2cP/TEtLe327dv0x/nz5w8ZMkR6Cfb29nPmzKE/DhkyhDWoXbhwQcbGODg4dOrUSewq5hcwQghrbKqVVq1aHTx4UF1dnRCipaW1Zs0a5lrqhkBKeHg484ZAX19fU1NT+qOKisr06dOZ24aHh0uqVEVF5fLly9OmTTMxMdHS0urSpcuUKVMIIRoaGnSenJyc0NBQ1vNBXC53wIABskwwAdAkjB492sTEhEpfuXLljz/+oNKamprMxxDEioqKYn5ctmwZ53+xTsknT54o2NrJkycz7xmxsLBgrmU+xcOa9n/YsGGsolj33yUkJJSWloptp+hceiNGjJDUQkdHRzpdVVX17bff6ujomJmZDRgwwNfXl5oJW9K2ALKbMmUK89Z0WpcuXbZu3SpLCaKbsw516a/fYikpKQkMDBw9enTHjh1btGihqqpKDQI9evRgZsvPz5dSiOwneB1WKgdFhhdRsvcawws0lDo85r/++mtZ5m0hhIwaNYr50dzcnFUOneZwOGZmZsy1rOHik44VODGbLcy30jRMmTLl4sWLp0+fZi2X8YJJ7NUS8weu2l4thYaGXr58OSEh4c2bN2VlZVVVVaLZFLlwkf2qQvSL3LJly6SULOWLnKSRXXR8JITw+XxbW9sOHTr06dNn4MCBhoaGUioFaFqooOT69esJIUKhkA5fTpo0ScpkrpScnJxa1aX45UX79u2ZH5nBUEII83FrVttYv3cRQvh8PpfLZW6Sm5tL3TbIGtBY13Nil9BcXFyGDh3KCmFnZmZmZmZeu3Zt//79KioqEydO3LVrV4sWLSQVAs2Qqakp6/1TNNZxTjt48ODt27eZhzqHwwkJCZHxvekfPbBlf2PopUuXJk6cKMuVAD0pg1iyn+B1WKkcFBleRMneawwv8CnIMvjU4THv6uoqY8NY822zTg3WJNCsZ3WZLfnUYwVOzGYL9600DFNT03cSTJgwQewmBw8eZN3bJvsFU91eLVlZWc2aNeuPP/5ISEgoKioSG1ghil24yP1l6aOkfJGTNLJT4yNrIT04Tpw4kc/n+/j4FBYW1qolAI3ZzJkzWU8OEkLEvr1LQZWVlQqWwBoYWY9/SyHlzWKihP87r16ttiWE/PHHH2vWrGE1lVZVVXX48GHmD24AhBAOh9NCAtZ3Blp8fDzr24JQKGTOfP/RGhXMQElNTR01apTo9xYej8fj8ViNF0qYtJIi+wleh5UqrrZDBEuthjUML1Dn5Bh8FDnmZZwumoh8I2C+TZUQwgrcSDrN62eswInZPCG20jDq+YKpEV4tSSf3l6WPkvJFTsrIjvERmhs+n8+aUd/Z2blz584f3ZB1mgQEBORKdenSJQWbKvv1HKttL168YGXIzMxk/QxO35JmYGDAXP7y5UvWtq9evZJSNY/HW716dVZW1sOHDwMCAhYuXOju7s58gJEQcuPGDdaNeAC1UlRU9O2331ZXV7OW//DDD3///bcsJXz0wJbyahumwMDADx8+0B87dux4/fr19+/fl5eXl5eXsx4lkE72E7wOK5WDIsOLqFp9TcXwAg2iDo950d9yPqn6GStwYjZPiK00DQpeMDXCqyXp5P6ypMgXOSkjO8ZHaIZYd6nIeNMKa/boe/fuGUjG4/G0tbXpzKwoquiIpyBW286fP8/KcO7cOebHDh060D+CdevWjblKdI4q5lsbJeFyufb29jNnzty5c+fZs2dfv369evVqZgbFZ5+B5mzWrFmiX28IIRUVFRMmTGD+8y2J6KPHrNNEysuSmVhH8s6dO11dXekfnD/Rca5IpYoPPooML3UCwwvUswY/5uWm4ABVq+ECJ2Zzg9hK06DgBVOTvlqSTpEvcrWF8RGaFVdXVzs7O+quNEtLS9YEcpJ4eHgwLzuOHTsWFhYmmu3NmzcbNmywsrJiPjzIuupKSUmR9MihfEaOHMlsW3R0NPOtRsnJyaxps5k3o7EeDIyJifnpp5/ojzdv3gwMDJRUb3R09K+//vr+/XvWcg6H07FjR+aSOg8nQZMmFAoLJWMdLUePHj158iT9cfjw4UuWLKE/Pnv2jPlRkujoaOarPS5fvnzkyBFmBtEnZMVinbnMp3Hz8/NZ/3TWFUUqVXzwUWR4UQSGF2goDXXMK07BAUqW4QInZrOFuWwbBnXBJGmttrY2c7QSvWCytbWlL+upC6b9+/dLqY66WqJfFdS0rpak8/Dw8PPzowemY8eOeXh4iH4JfPPmza+//rp79+6MjIzaRs2jo6OfPn06btw4TU1N5nKMj/DZkyNc2KZNm6lTpx48eJD6KBQKv/7666FDh7q6upqampaWlqampt67d+/evXusW4UJIXw+X1NTk74Wefv2bY8ePXr16qWmpkYI+eabb5ivn5dD27Ztvb29mdN4f//996Ghod27d8/Ly4uMjGTGqfX09BYuXEh/tLa2Zs1L5+fnd+zYsS5durx58+b27dui3aG9fv166tSpc+bMcXR07NGjR9u2bXV0dCorK589e0bvKAprSIFm7s2bN3p6epLWPnz40N7enkqnpaXNmzePXqWvr3/w4EE9Pb2LFy/GxcVRC/39/YcMGSL6Lg+WuXPnBgUFdezY8dWrV3fv3mU+3tu6devRo0fL0vKOHTvevHmT/ujr65ucnNy+ffvU1NQDBw5kZ2fLUkhtKVKp4oOPIsOLIjC8QENpqGNecQoOULIMFzgxmy3EVhpG/V8wNd2rJekU+SInI4yPALWya9euv//+m/mI3IULF2R50TuXy/3qq6/oVz4TQmJjY2NjY6m0ra2tgrEVQsiePXseP34cExNDL3n06JHoC+NVVFROnTrFelhy//793bp1Y76njNm8du3apaSkSKm6oqLi+vXr169fl5TBzs7OyclJ9r4AUAQCwcSJE0tKSugl/v7+1GvUjx071qtXL/p3kSlTpsTGxkqaO4wQ0r1795iYmCdPnojGVblc7i+//KKioiJLk2bMmBEQEED/s1tYWLhu3Tp6bdeuXZ8+fSpb52pBkUrrZPBRZHhREIYXaBANeMwrQsEBSvbhAidmM4Rngho1SRdMPB7v2LFjzEucKVOmSHljDvVq4SdPnoSGht65c4cZWKnt1RJzRm5qMJo0adKaNWuys7O7du1aq97VlV27djk4ODCXXLhwYcmSJRMnTpw5c+a2bdvu3r0rd2CFRo2PP/3006xZsyZMmODt7b1t2zbmzUcYHwEoGhoa169fZw0XYnXq1Ik1+GzYsEFXV/fTtU1TU/PmzZvUm9QlsbKyunXr1oABA0SXR0ZG6uvri27Sp0+fU6dOKdKwrl27hoeH1+G83dB8rFu37sGDB/THMWPGjB8/nkp369aNeUtpTk6Oj4+PlKLWrFlD3+LKpKqqevjw4YEDB8rYJDs7O39/f7HHs5ub26+//ipjObWiYKWKDz6KDC+fFIYX+EQa7TEvneIDVJ1cq+DE/CwhttKo1dUFU5O+WvooRb7I1RWMjwBMGhoagYGBycnJq1atcnFxadWqlZqamoqKir6+vp2d3fjx4/fu3ZuQkBAXF8d6i3yHDh1iYmJmzZplbW2trq7+Kdqmra197Nix2NjY77//vlevXgYGBioqKtra2m3bth0/fnxISEhSUlKfPn3EbtunT5/ExMTFixdbW1urqanp6ek5OjoGBATcuXNHyq2I7u7u0dHR+/btmzBhQvfu3S0sLKgHP3V0dDp06ODp6fnbb7/FxMRYWlp+iv7C5+3PP//ctGkT/dHIyOjAgQPMDMuWLWP+/HDhwgUpDxFzOBx/f/+LFy8OHz7cxMREVVXVzMzMx8fn77//njRpUq0aNmvWrHv37o0bN87ExERZWdnQ0NDFxeXQoUNXrlxhPWBbhxSptE4GH0WGF/lgeIGGVf/HfJ1QcID66HCBE7P5EkK9YF3rSPfw4UOhUHj37l3md3UjI6Pc3FxmmQKBgHW/xr59+6hVXl5ezOVnz54VCoWiV0sJCQliWyt2c1pUVJToYFRTU5OQkMDcqlOnTrKXKX0tc2IXQgifzxfb7NTUVOlf5GrVTVpVVdVHx0eBQCB2WwAAgEZLxn8HAQAA4KM4QsbjIfDZ8Pb2Zs5We/bsWXd39wZsDwAAADQ2uFoAAACoK3gmCAAAAAAAAABAfoitAAAAAAAAAADID7EVAAAAAAAAAAD5IbYCAAAAAAAAACA/xFYAAAAAmqPDhw8zX3CAiWwBAADkhvcEAQAAAAAAAADID/etAAAAAAAAAADID7EVAAAAAAAAAAD5IbYCAAAAAAAAACA/xFYAAAAAAAAAAOSH2AoAAAAAAAAAgPwQWwEAAAAAAAAAkB9iKwAAAAAAAAAA8kNsBQAAAAAAAABAfoitAAAAAAAAAADID7EVAAAAAAAAAAD5IbYCAAAAAAAAACC/Oo6tLFmyhMPhXLx4sW6LlZ29vf3x48cVL6dbt24nT55UvBw5PHnyREtLS3oeSd20tbUNDw+Xva7a5q9VYwAAAAAAAACag7qMrVRVVR05cqRly5ZBQUF1VaZAIKhVfj8/v169etVV7SCL2u7z2v5NAQAAAAAAABqzuoythIWFvX//Pigo6OzZs2/fvqWX29raLl++3NXVtWfPno6OjnFxcR9dvnLlSicnJxsbm7t372ZlZY0dO9bIyMjMzGzevHllZWWEkNevXxsaGp45c4baZMSIERMnTiSEbNu27a+//qLL2bFjh5OTk7q6uqOjY2Zm5pYtW4yNjQ0NDXfu3Enl2blzp42Njba2tqWl5aZNm6iFvr6+z5498/X1tbS09PLyIoQUFxfPnj3b3Nzc0NBw/PjxeXl5ot2XpTpCiNjuvHr16quvvmrRokXnzp3//PNPOrMs9bIkJCQ4ODjo6ur27duX3qVSyklPT3d2dtbQ0OjRo0dsbKyU3bJr164+ffrQGz548EBbW/v9+/fMfS62d6J/04/2AgAAAAAAAKCpqMvYSlBQ0JgxYzw8PPh8/uHDh5mrIiIiwsLCHj58OHPmzGHDhlVUVEhfHh4eHh4e/vz5cxcXl9GjR6uoqKSmpkZHR8fExCxatIgQYmZmduTIkenTp6enp+/cuTMhISEgIEC0SYcOHQoICMjNzdXV1XVxcSkpKUlLSztz5syyZctSU1OpciIjI4uLi8PCwvbu3UsFa/bt29exY8d9+/ZlZGQcOXKEEDJx4sTs7OzHjx+/fPlSV1fX29tb7B74aHWEELHdGTt2rJmZ2du3by9cuMDsiIz1MgUGBgYHB+fl5X311Vf0LpVSTmBg4J49e/Lz83v37j137lxqodjd8u233z569CgpKYnKc+TIkTFjxmhqajJrF9s70b/pR3sBAAAAAAAA0GQIhUKhUChHWvi/0tLSOBzOjRs3hELhqlWr2rZtW1NTQ62ysbHZvXs3nZPP50dGRkpfvn//fmphYmIiISQnJ4f6eO3aNTU1terqaurjkiVLOnTooKWl9ejRI2pJjx49jh07Rtfr7+9PpY8fP66rq0tv2Llz59OnT7O64OfnN3v2bCptZ2d34sQJZtfoNhQVFXE4nMLCQtbmslQntjvPnj3jcDgFBQXUwqNHj2pqakqvl9lNVhu2bNlCpQUCgZGRUWRkpJRymLs6NjZWXV1dtEzmbhkxYsQPP/wgFArLy8v19PRu3rzJbIyUPxazIhYFj0OkkUYaaaSRRhrpBkwzlwAAQLNVZ/etBAUFWVpaOjs7E0K8vb3T0tJu3LhBrzUxMWGmMzMzpS83NTWlEq9fv9bS0jI0NKQ+tmnTpry8nH6kZe7cuUlJSW5ubt27dxfbKrocTU1NExMTLpdLfywpKSGEhIeH9+vXz8LCwsLC4uDBg7m5uaKFZGRkcDicnj17WlpaWlpadu3aVVdXl25qraoT250nT560bNlST0+PWti+ffva1stkaWlJJZSUlMzNzTMzM6WXw+fzqYSWltaHDx+oyVAk7RZvb+9jx47V1NRERES0aNGiX79+zKql/7HonQMAAADw2aAuqTkcDofDoZYgjTTSSCP9eafFUqb+J2SE3mubJoQIBILDhw/n5+fT36sJIUFBQf3796fS9BMxAoHgxYsX9Pd5ScvpRpuZmZWWlubl5RkYGBBC0tPTeTwela6urvby8vrmm28uXboUERExcuRIKf0UKysr65tvvrl48WL//v05HM7y5cuTk5OpVXRYhBDSunVrDofz999/6+rq1rYKFrHdsbOzKywsrKio4PF4hBB6qhr56n3x4gWVqK6ufvXqFZ/Pr205UnaLu7v7jBkzrl+/fuTIES8vL9axJeWPRRh/UxZFjj2kkUYaaaSRRhrphk1Lv9QGAIBmom7uWzl37lxOTs79+/ef/OvgwYNhYWH5+flUhsDAwPj4eIFAsH79enV1dXrGDUnLaTY2Nr179164cGFpaWl2dvbKlSu9vLyowMeaNWuKi4uDg4MPHTo0ZcqUV69e1bbZ1L0kXbt25XA4b968CQ0NpVeZmJg8f/6cSrdt23bQoEHTpk3LysoihOTm5p4+fbq2dUnpTseOHb/44otdu3YRQioqKvbs2aNIvQEBAYmJiQKBYPPmzTwez8XFpbblSNktKioqEyZM2LZt25UrVyZPnixL75hRKgAAAIDPjJCBuQRppJFGGunPNS1W3XzvDQoKGj16dPfu3c3+5e3tbWxsfPToUSrD7Nmzp0yZoq+vf+nSpbNnz1I3aEhZznT69OmysjIrK6svvviia9eu1Dt3rl27tm/fvlOnTvF4PA8PjwkTJnh6etb25b7W1tarV692dHR0dHT09fUdMWIEvWrp0v4mzKkAACAASURBVKUhISF6enpjxowhhJw4ccLU1LRv377a2tq9e/dW5E03Yrvz+++/X7t2rWfPnu7u7sxmyFHv9OnTvby89PX1z507R+/SWpUjZbcQQnx8fK5cueLo6GhlZSVj7wAAAAAAAAA+YxzpoZc6YWtru2XLFg8PDxmXAwAAAAAAAAA0FXheAwAAAAAAAABAfoitAAAAAAAAAADIrz6eCQIAAAAAAAAA+FwpN3QDAKBZqMNXVCIiDAAAAAAAjQqeCQIAAAAAAAAAkB9iKwAAAAAAAAAA8kNsBQAAAAAAAABAfoitAAAAAAAAAADID7EVAAAAAAAAAAD51dN7gurwFSEA8Hno3bv3/fv3fXx8EhMTmQmhUFhcXPz48ePly5c/ePCgoZsJAAAAAADwEbhvBQAal+3bt8+fP9/c3DwyMtLQ0LChmwMAAAAAAPAR9XTfCiUzM7M+qwOAxoPP58uY89mzZ0ePHs3IyLh169aECRP27NnzSRsGAAAAAACgoHqNrQAAyCghIYEQYmFh0dANEQ/POQLIQigUNnQTag1nN0Cda4pDAQBAbSG2AgCNka2tLSHk5cuXDd0QAAAA+Mw1laCq7FGqptIjSRCPg6YIsRUAaFw6duw4adKklStXFhUVnThxoqGbI01NVPuGbgJAI8V1SG7oJigEl/UAdaKpf8MHAJAdYisA0LgsXry4tLT08ePHkydPzs7ObujmAAAAAAAAfARiKwDQMB48eED/nCWaAAAAAAAAaCrwDmYAAAAAAAAAAPkhtgIAAAAAAAAAID/EVgAAAAAAAAAA5IfYCgAAAAAAAACA/BBbAQAAAAAAAACQH2IrAAAAAAAAAADyQ2wFAAAAAAAAAEB+iK00pP79+/P5fD6fb2Fh0blz5+HDhwcFBVVVVclXWufOnfl8fmRkpIJ5ZGdra8sXZ/bs2XVS/icl466o2z0GAAAAAAAAnx/lhm4AEENDQxMTk1evXsXExMTExPzf//3fqVOn1NXVa1uOnZ1dUVFRixYtFMwju27dupWWlhJCnj17VlFRYWRkxOfzCSFWVlasnAKBQFm5SR5sdbvHSFPeFQrKzMwUXUgdMEKhsN6bAwAAAAAAUGdw30rDGzZsWGRkZGxs7ObNmwkhMTEx/v7+hJCCgoIff/yxb9++7dq1GzBgwKlTp+hNioqK1q1b5+Tk1KZNm27dulEb/v33348fPy4sLKTyBAYGOjo6WllZ2djYuLm5nT59WjRPfn7+8uXLqSpcXFz27t1bWVlJraLu19i7d++4ceNsbGz69+8fHR3NavnJkyfPnTt37tw5MzMzQsiIESOoj35+fvTmI0eOtLCwuHfvnpTuSFnFJDbbxYsX+Xx+p06dsrOzq6urR44cyefzV69ezerFmDFjbGxsBg8eLNoLQsj333/fq1evtm3bWlpaOjs7Hz58mF7F3GNS9omULojuCsnHAgAAAAAAADQ9iK00Flwud/Lkyc7OzoSQc+fOVVVVjRo1Kjg42NDQcM6cOUVFRQsXLgwJCSGEVFVVff3114GBgQUFBSNHjuzVq1dSUhKrtMePH69bt66srGz+/Pk+Pj5mZmbp6emsPJWVlR4eHkePHuVyuR4eHnl5eVu3bv3uu++YeXbt2mVtbd2+ffvnz58vWLCgtp3atm0bj8dzd3dXUVGR0h1Jq5gkZRsyZIiPj09hYeGiRYv8/f2jo6Pt7Ox+/PFH5rY///xz3759nZ2dY2NjJ06cmJeXxyo8ISGhT58+vr6+48aNS09PX7Fixd27dyV1SnSfyNIFeldoamrWdjcC1AnDQWlch+TwW6Wfequwm6Vch+R5P+XUsoENbMiCTK5D8sqA/IZtRmJGJdchmeuQnFdYLT0n808zyDdTpU9yfFplvbQR4B+JiYkcDofD4Yj+wyojAwMDDocTHh6uYJ5aGTx4MIfDYV0qAAAAKAixlcalTZs2hJCcnJyrV6+mpKSoqan5+/t7e3t///33hJD9+/cTQq5du0ZdzUREROzatevgwYNBQUGscqhHdVRVVc3NzQcNGhQQELBkyRJWnmvXrqWlpSkrK4eFhW3fvn3Pnj2EkIiIiJyc/74RzZ07d/369Rs3biSEZGRkFBcX16o748aNO3XqlL+/f2FhoaTuSOkpk5Rsq1at6tKly82bN7dt26atrX3gwAEVFRXmtr6+vosWLQoMDDQ3Ny8tLT179iyr8LCwsD179syfP3/16tWurq6EkNu3b0vqlOg+kaUL9K7o0aNHrfYhQF2x78Bz6KSmr6P0SbeqqSHL9udxOGTxt3q1b6MYVQIxj4yJXdicLZ2sV11DfvhZzu+3AA3F3t7ewcFBX19fwTx1S+6Z7+CzFxcXJxQK6Ws8Pp9/5MiRzMzMysrKvLy8qKioJjHnIBOrR4SQjIwMoVC4ZcuWwsJCoTgnT56kNxQKhYsXL6a3vXTpErVw+/btksoH+IwhttK4pKamEkKMjIyoySnKy8sdHBy6dOlCfV1/9epVRUXF69evCSH6+vrt2rWjthKdv6Nv375jx459+/btggULhg8f3qlTp4MHD7LyUFXo6+sbGhoSQmxsbKjlb968ofN06NCBEKKjo0N9fP/+fa264+TkxKxLbHekrBJtrdhsqqqqs2bNIoQIhULquRtWM1q3bk0I4XA41LNLrIk/ampqduzY0aNHDwsLi/bt21+9epUQkpubK6lTovtEli7QuwKgbum5pXIdkrcefddnyisNp5TB8zMzcwXzt+fquqZajEg/euG/eGh0QkVUfHlBcbXcW70rqVmyN8927AsNp5RWQ9LEfpO/eP998qsq5y/ULVv9E+LcEfLOenSGulOKrmuq3YSXVOHUrRm3Yj4QQqITyqmP5ZVC8u/tGJsOF3w5/RXPMeVmzAexC3PfVc/dltPu6wwt55RO415sDC6oqPwn4JL8qsptzmst55RuE18Gnytm3gYybUO25ch0zX4pal+mdBz3wv/0Pw9IdvF8celBGSFkY3AB1yG5i+cLQkheYfX87bntR2dQRR0+Jz64LOPOlN7g/rNfazmn9Jj8Miq+nFm4jG1w7aFhaqB8/s/3aZn4TgifVnJysqurq6amZvfu3aOiolhr8/LyfH1927Vrp6mpaWdnRz9j++7du8WLF9vY2Kirq5uYmCxfvpxaHh0dHRUVVVBQQAjZsWNH+/bt1dTUdHR0unbtevToUdE8ubm5c+bMadu2raamZseOHTdu3Ej/U0vd3rJp06YBAwbo6Oh07txZ7EO4nTt3vnTpEiFk48aNHA6nc+fOzG0dHR1VVVVv3rwpqSNS+gjN0G+//TZ58mTqmElNTbW1tXVzc2voRtWZhw8fRkVFRUVFlZeXE0Levn1LfUxOTmZm8/X1VVJSIoR06tRp0KBBDdNWgMahOc6p2TjV1NSEhIRQt0sMGzbM1NSUEKKhoREUFEQNWISQyspKJSUlavrPgoKCtLQ06j4X0elRuVzu7t27t2zZkpSUdO3ate3bt+/YsWPGjBnMPFQVBQUFeXl5BgYG9INFrVq1ovNQVXM4HPk6paamxqxLbHekrBJtrdhs796927BhAyGEx+OdPHly7Nix9vb2zG2fPXs2fPjwysrKlJQUuijaxYsX/f39tbS09u3bZ2pqum3btvv370uZXVV0n8jSBXpXAHwKqwLz+9ur81Q5l6PKuni+0NNW6milGhVfPmNTzsBeGq0MxA/1tdqqskroPPNVXGqlvo7S+EHaJe9rnol7AiXsRikhZICDBvUxKr58yd48k5ZKy731KyqEsakVqa//+/IvZWhZGZDv2kNjTH8tLXWu6EIVZY7TjNdJLyvbm6t4DtIOv/V+ZUB+fGpl6AaTyirh0O8yU19XtTVT6WbN+373/8RJn6ZUunTXaGumkpkjOPR/Rb4/5dpaqLr11Ph2iE7AmcIXbwW9Oqn16axmYqBcWSXsN/N1YkZl365q3w7R+fX/iqasz64SCKd76MqxM/V1lCQ1uEogHPZdZsrrKstWKrYWqgt2/Ndg2dvA5ZL+PdWPXyyJuP1+oWedzb0NwFJVVTV06NCUlBRLS0tbW9v58+cz11ZWVjo5OSUmJvbt23fSpEmHDh3y8fGpqqry8vLq169fXFycvr7++PHjS0pKnj17xio5Kipq8eLFJiYmP/zwQ3l5eWxsLPVPNlNFRcWXX36ZlJTUvn17T0/P8PDwH3/8MS4u7sSJE3SedevWzZgxo7S0NCoqysvLi/UlkBAyadKkAwcOvHjxolevXn369GFe8/z444+urq5jxoxRVVUV25Hp06dL6uP06dPrYP9Ck6Kmpta3b19CSP/+/WNjYwkhqqqqlpaW1FpDQ8O1a9d+9dVXJiYmL168CAkJ2b59OxUKzMvLa9my5YoVK/r379+rV6+XL1/OmDGjEc7HN3DgQCqRmJhoY2Pz22+/saYOIIQUFBS0bt169OjRp06dWrRoESEkPz+/ZcuW9d1WgMYBsZWGd/78+UePHr1+/frdu3eEkO7du8+dO1dZWblNmzZpaWl79+51dnYuKSmJj49XUlJyc3Nzc3Nr3759cnLyiBEjBg0aVF5eXllZ+csvvzDLjI6OXr58eZ8+fYyMjBISEgghosOcm5ubpaVlRkbG119/3atXL+pnHHd3d2Nj4zrv44ABAyR1R8oqGUv47rvvsrKyxo0b16dPn++++27OnDlXrlzR1f3vu0dAQMCLFy/S0tJyc3M1NTWHDx/OLFkgEFCJysrKp0+fxsTE1GHv5NpbnycqJihWrYJ3eKmQWAs9W2yZZ/DjgfxNhwvKK4RRp811NLm6/VMrKoVPUyolxVZqtdWFe+/jUis5HHI3yMzWUpVIeDAn5nkFIaRrOx71seR9DSGEp8KxbKVs3Vp1xRR9DbX//txcybdO+rjr/PIjeyyiF4bdLE16WamsxLl90NxYX8nD5f3wRW9OXinZ8Z1BbGpl6usqLpfcDDDjGyofv6gxec1buoTbB83UVDk1NeRDRc2bPMG5u++v/lXm1lNj6WS9m4/KXrwVDOylsX5WS6qKxIxKdR4nZH0rTTWOubHytA3ZW468kxRbkb4zy8prJDU4Lq0y5XUVl0tuBZqZGysHnyueuj6bKvP8n+9lb0OXdjxCSqKflYuuAqgrN2/eTElJ4XK5t2/fNjc3Dw4OnjJlCr32/PnziYmJ6urqoaGhmpqa5ubmU6dO3bx5s6GhYVxcHIfD+fPPP21tbYm4h25KSkoIITwez9LS0tra+scff9TQ0GDluXDhQlJSkrKy8p07d4yNjUeNGuXu7n7y5MmdO3fSIZKlS5euXbs2Ojq6Z8+eKSkphYWFrDf9LV269MaNGy9evBg4cCD1wwzNx8fn0KFDhJCwsDCxHZk+fbqkPiK20gyVl5dnZmaamZmFhIScO3cuKirq1q1b1E+VPB7v7t271tbWycnJJ06c8PDw2LBhQ+fOnT09PenNV61adfDgQS0tLQcHhyNHjrRv377hukLGjBlD/zDJDDh+VHBwsK+v78KFC2/evDlx4sRHjx7l5eV99dVXn6aZAI0dYisNLzc3t6CgQEdHp3v37iNGjPD29qamCwkLC9u9e/f169d3796tpaVlbW09btw4Qoiqqio1P8jVq1fDwsJ0dHTGjx/PKtPQ0FBfXz8iIqKoqEhTU9PJyWnFihWsPDweLzw8fPv27bdu3QoLC+Pz+VOmTJkzZ86n6CPVZindEbtKlhKCgoKuXr1qYWGxfv16TU3Nq1evnjt3buHChb/++iu9rZ+f39WrV5OSkjp16rR+/XrqGSiau7t7VFRUeHj4hg0bnJycXF1dIyMj66p3APWjRwc1QkjLFkqEkDZ8FYMWSoQQLXVuRWV1SVlNnWz14q2AEGLQQokKrBBCVJTFBMXyi6oJIS20/4mauNpreA3TCYks8V6bTQhRU+VsmN1y0YT/mYpFIG7OVrde7K9VzIUvs6nGcI31lQghndv+E8p5lS14nSMghOjrKPENlQkhnduq0pvX1JC1QfnHL5Zk5Qtq/u1fdoH4KWOpKj5UCK1G/jcReEZWVXmlUE1VTMel78zMXIkNfpUtIIS01FUyN1YmhHS34cnXhhZaXELIR2fABVDEq1evCCEtW7Y0NzcnhHTv3p259uXLl4SQDx8+0D/dE0IyMjKeP39OCDEwMKACK4QQ1rRohBBXV1cvL6+QkBAvLy9CiJqa2saNG6mfwVnlGxgYUD8CUY/zUK2ivw127dqVEEL/vlJaWsqKrUgxYMAA6R0pLy+Xsgo3qDZDY8eODQwM7Nq1a5cuXQghlZWV69at27hx49ChQ62trQUCgZOTU3Z2dlhY2Llz58aPH79o0aKsrCxq261bt65evdre3v7hw4ft2rVr0aIF/R7P+mdsbCzfb6tv3749ceKEl5dXcHAwj8fbtWvXpEmT6rx5AE0FYisN6fr161LWGhgYsH5Roenp6a1Zs2bNmjXMhXFxcXTa0tJS7JuMmXkIIYaGhlu3bhVbBas01hwlLKIzv7IqIlK7I2WVLNmYPxYFBgaKZrCysjpz5oyUFm7atGnTpk1iK2Vmk7JPpHRBdFcA1DkVZUL+fcRGbMhD8a1aGysTQvIKq5NeVlq3/ue+FdGtdLW4r7L/uV2FEMLlkOBVxj8vNXqWXnn+z/drDuavDSpYNEFPU537/kPNu+IaQsjTlAoiQp0npj30QnMjqjE1Oe+qjfSU4tP+KcHMWLnofQ0hpKC4+m1+tUlLpdjU/x5c+uNm6daj77Q1uMfWmpgbK68KyL8Z84G+EUqJyyGE0DEXqgpNde7pLa2U/r2/pqJKqCxhVl/pO1NKgwtLawgh74prikprdLW46W+qWFvJ2AYqHKarhWnU4BOipi179+5dUVGRrq4u6xWEVMBFU1PzzJkz9FOxFRUVZWVlhJC8vLykpCRra2tCSFVVFSu8wuVyDx8+fODAgWfPnp0/f3716tVr1qxhxVao8vPy8nJycoyMjOLj45mtosjyLDOVp6aGHUFWV1eX3hFlZWUpq6TUCJ+rBw8e2NnZWVlZOTg4TJw40d3dfcOGDadOnaJm+svLy8vOziaMS0Fzc3M6tvL06VNCSFFREfVRS0urAWMr/v7+8+bNo9IZGRmicxdKsXPnTi8vr6FDh2ZmZp46dQqxFWjO8C8BAAB83NC+mh0sVRMyKh2nvR7hpPmhQlhRJTyzlX3ncJe2vLjUyucvq4b0JYSQe08/zN2W69xd3cRAKS6lghBipK9ECOlhy7v9+MPKgPx7Tz8E/FFU68Y4arYzU0l5XeU88/WXdmoRt98TQsa6aZkaKBu2UPp31as+XdQjbv/39mhB9T9xlIpK4aPEigdx//P4jEUrZUJIyKXi0g81Dp3VRrtqWbdWTXpZuSm4YKCDRsn7midJFUpKZJijPK9Rl9JgA12lNnyVtMyq0Uuzvuqj8fPv/11bD3PUlL0NiRlV5J8ngwA+FRcXF+oZ2K+//nrw4MH+/v7MtcOGDbO2tk5KStq4ceOgQYOKi4ufPHmipKQUHh7eoUOHhISEvn37jhgx4sOHDxUVFX/88Qdz23v37s2ZM8fZ2blVq1bU1BVGRkas2ocOHdquXbuUlJR+/fp9+eWXERERhJCxY8eyplH7KOp7Y0hISGlpae/evSdMmMDKIKkjw4YNk7KqVm2Az4O3t/fp06fT09PT09Pv3bvn7u5OCNHX16fu8DIwMDAyMsrJyenUqROVn3ofBaW6upp8Fk86P3369Nq1a25ubvv27cM7tqCZww9cAADwcTxVzu2DZgs9W+jrcEMvl1yPLrOxYN/VTwhxd9IkhNx6VEZ9NG6pbNBC6eSVktWBBdcelg3opfHbplaEkP1LjHp2VEvNrLrx6MOaGbWe9E5NlXMnyHyGh25FpTD0UklLXaU1M1oeWWNCCFFR5pzfzXfprp6ZK3iSVLFpjgG1ibISZ6yb9pwxuspKZOn+vL/iygf3+Z8IxcIJeg6d1LILqvedKrwSVUb1d97YFm/yBBuDC4LPFVdUCb8ZoF3bpn60waoqnPO7TPt9of5XfHloZMlCxgNTtWrDzUdlhBD3L+UJ/QDISFVV9cKFC/369fvrr79CQkJY95XweLw7d+7MmzfvzZs3GzZsCA4OrqioGDduHLV84cKF+vr6oaGh169fp19NSDM2NjYwMDh58uSqVauuXbs2YMAA0dtv1dTU7t69O2PGjIqKitDQ0JYtW65du5Z+nZDsFi1a5ODgkJ2dvW/fvsuXL4tmkNQR6augGQoODs7JyXny5Mnly5cfPnxICMnKyoqNjb1w4UJKSoqysvLt27d/+eWXI0eOEEJ+//135rs4PydTpkwZOHAgK9jKNGbMmAf/Cg4Ors+2AdQnTv2ES6mbM6U/VwIAnzEpc9nWSiP5hYca02qiGnLmucapskpo5ZGRX1T9+pwVNedI/SssqaEnfAkKL5q5OUdfRyn3cht533jWBPwVX957yqu+XdXuBpk3dFv+wXVIJo3mhK0V6uxuii0HaISaygkl47T6CQkJtra2u3fvXrhwISFkxYoVAwYMsLGxadmyZUFBQVRU1MqVK6kbr4yNjdetWzdo0CBjY+OXL1+GhoZu27aNepkx9Z6gUaNGhYeHt2vXjnqblbm5OfOuFklk35My9iguLq5Tp06izwRt3bp12bJl1BLqPUF79uxhvieI2nDJkiXbt29nFhgZGfnVV1/t2LFj8eLFdDZmhkePHrFe6ClW4z9mAEQhtgIA9QGxlebj0P8VT9+YvcJHn3rhTv1btCs3Pq2yb1e13MLqw+eKy8qF62e1XOGj3yCNqR/fLM86fb309kGzL+3UG7ot/0BsBQBI0zmhZIlEtG7dOikpicfj+fr67t+/vx5aJarOYyuNVuM/ZgBE4ZkgAGgYpaWlQoarV69yudzAwMDCwsLS0tKYmBhtbTmfv4CG5eOukxZuOWeM+HcV1wM7a176m6qtR9+dvFzapS3v2FqTH7w/58AKIWT7AsMX/2fVeAIrAACfmb1798bFxfF4vMLCQtE3JAAAEMxlCwANxd3dXVlZ+dSpUwkJCatXr87Pzx84cOCMGTOmTp2akZHRr1+/pv6TS7PF5RLLVmKmYqk3XsN0vIbpNGAD6l9rE/xrDgDwCbVv376srOzOnTs//vgj/a4fAAAmXI0BQMO4efMmIaSysjI3N/fq1auEkNGjRxNCunfvnpKSsn79emoKfQAAAICGNWTIkIZuAgA0dngmCAAai/Pnz0dGRs6dO/fWrVvJyclmZmYN3SIAAAAAAICPQ2wFABqL8vLyIUOGtG7devXq1VZWVtRtLAAAAAAAAI0cngkCgMZi6NCh/fr1u3HjRklJCSEkLy+voVsEAAAAAADwcYitAEBjkZ2d7eLiMm/ePIFAEBIScurUqYZuEQAAAAAAwMdx6ufl4dT7PjIzM+uhLgBohPh8fp2UUz9D1kdRY1pNVPuGbghAI8V1SCaN5oStFersbootB2iEmsoJ1VReTSj7nmwqPZKk8R8zAKIw3woAAAAAAAAAgPzq9ZmguvrhGgAAAAAAAACgkcB9KwAAAAAAAAAA8qun+1bwyBxAM9fUn/sFAAAAAACQpJ7msgWAZq6uYiuNZMhCqAhAFo3khK0VnN0Ada7xDwVN5cTHXLYAjRmeCQIAAAAAAAAAkF+9zmULAPB5wM8pAJ8rnN0AAAAgBzwTBAAAAAAAAAAgPzwTBAAAAAAAAAAgP8RWAAAAAAAAAADkh9gKAAAAAAAAAID8EFsBAAAAAAAAAJAfYisAAAAAAAAAAPKr49iKi4sLh2HatGm1LaFbt24nT56s21ZR7O3tjx8//ilKFuvTdQQAAAAAAAAAGg/lOi/Rz89v5cqVVFpFRaXOyxclEAiUlT/eET8/v27dutVDewghAoGgfioCAAAAAAAAgIZV988EqaioaP2Lx+MVFxfPnj3b3Nzc0NBw/PjxeXl5VLadO3fa2Nhoa2tbWlpu2rSJWujr6/vs2TNfX19LS0svLy9CiImJyd27d6m10dHRWlpaVNrW1nblypVOTk42NjZ3796VVAvTtm3b/vrrL3rzHTt2ODk5qaurOzo6ZmZmbtmyxdjY2NDQcOfOnXSe5cuXu7q69uzZ09HRMS4ujlqelZU1duxYIyMjMzOzefPmlZWViTbpiy++YHVEbH+prdasWePq6tqpUyd7e/vY2FhqeWFh4Zw5cywsLHR0dHr06PH8+XNCiCzdBAAAAAAAAID69MnnW5k4cWJ2dvbjx49fvnypq6vr7e1NLTczM4uMjCwuLg4LC9u7d++ZM2cIIfv27evYseO+ffsyMjKOHDkiveTw8PDw8PDnz5+7uLhIqkWKQ4cOBQQE5Obm6urquri4lJSUpKWlnTlzZtmyZampqVSeiIiIsLCwhw8fzpw5c9iwYRUVFYSQ0aNHq6iopKamRkdHx8TELFq0SLRJsbGxrI6I7S/l/PnzYWFh8fHxY8eOnTt3LrXQ09PzxYsX9+7dKywsPHz4sLa2tpSdCQAAAAAAAAANRlinnJ2d1dXVW/7r999/53A4OTk51NqioiIOh1NYWMjays/Pb/bs2VTazs7uxIkT9CpjY+M7d+5Q6YcPH2pqalJpGxub/fv3U+m0tDRZaunRo8exY8fozf39/an08ePHdXV1q6urqY+dO3c+ffo0lWf37t305nw+PzIyMjExkRBC13Xt2jU1NTVqW2aTRDsiqb82NjZBQUFU+tmzZ+rq6kKhkAruZGZmMreSsZsAAAAAAAAAUJ/+m6aEw+FQoRY50kzTp09fsmQJlY6Pj+dwOD179qTX6urqZmZm6urqhoeH79y588WLF4SQ4uLiAQMG1DYqZGpqSiUyMjIk1SLL5pqamiYmJlwul/5YUlJCpU1MTOj8JiYmmZmZysrKWlpahoaG1MI2bdqUl5fn5eUZGRkxyxQlpb8Gu7fpTwAAIABJREFUBgZUQl1d/cOHDwKBICMjQ0NDg1WajN1U5O+INNJII4000kgjjTTSSCONNNJIS0+Lqvu5bLW1tc3MzKh0RUUFh8P5+++/Wd//s7Kyvvnmm4sXL/bv35/D4Sxfvjw5OZlaRcc4KFpaWuXl5VSaNb0I1TFCSOvWrcXWojj64SCBQPDixQs+n29mZlZaWpqXl0cFRNLT03k8Hh0coZvE6oiU/oplYWFRVlaWlZXVqlUreuGn6yYAAAAAAAAAyO2/7/9CoZAOwNQ2LUnbtm0HDRo0bdq0rKwsQkhubu7p06cJIdSNIV27duVwOG/evAkNDaU3MTExoeZtpdjb2589e5YQUl5evmfPnlrVorjAwMD4+HiBQLB+/Xp1dXUXFxcbG5vevXsvXLiwtLQ0Ozt75cqVXl5erHiQaEek9FdKj2bMmJGVlSUUCuPi4rKysmTspiJ/R6SRRhpppJFGGmmkkUYaaaSRRlp6WtQnn8v2xIkTpqamffv21dbW7t27N/XSH2tr69WrVzs6Ojo6Ovr6+o4YMYLOv3Tp0pCQED09vTFjxhBCtmzZ8vjxYzs7O3d3d2dn51rVorjZs2dPmTJFX1//0qVLZ8+e5fF4hJDTp0+XlZVZWVl98cUXXbt2pd8rxMLsiJT+SnLy5MlWrVr17NlTV1fXy8uLis58om4CAAAAAAAAgNw4UuIuzZytre2WLVs8PDwauiEAAAAAAAAA0Hh98vtWAAAAAAAAAAA+Y4itAAAAAAAAAADID88EAQAAAAAAAADID/etAAAAAAAAAADID7EVAAAAAAAAAAD5IbYCAAAAAAAAACA/5YZuAAA0FxwOp66KwkRRAAAAAADQeOC+FQAAAAAAAAAA+SG2AgAAAAAAAAAgP8RWAAAAAAAAAADkh9gKAAAAAAAAAID8EFsBAAAAAAAAAJAfYisAAAAAAAAAAPKrv3cw1+HrVwHg89C7d+/79+/7+PgkJiYyE0KhsLi4+PHjx8uXL3/w4EFDNxMAAAAAAEAa3LcCAI3O9u3b58+fb25uHhkZaWho2NDNAQAAAAAAkKb+7luh3Lhxo55rBIBGwtXVVcacz549O3r0aEZGxq1btyZMmLBnz55P2jD54F48gI8SCoUN3QQAAACA+lDfsRUAABklJCQQQiwsLBq6IQDQvCByClC3EGYFgOYAsRUAaKRsbW0JIS9fvmzohkhTE9W+oZsA0BhxHZIbugkAALXTVOKqMsaqmkp3JJHezUbeuybdeBAl40mH2AoANDodO3acNGnSypUri4qKTpw40dDNAYDmCL+0AygO3yEBoPlAbAUAGp3FixeXlpY+fvx48uTJ2dnZDd0cAAAAAAAAaRBbAYAG8+DBA/oXLdEEAAAAAABAk4B3MAMAAAAAAAAAyA+xFQAAAAAAAAAA+SG2AgAAAAAAAAAgP8RWAAAAAAAAAADkh9gKAAAAAAAAAID8EFsBAAAAAAAAAJBfU4qt+Pj4uLq67tmzh/p4/PhxV1fXgQMH3rx5s34aMHLkSFdX17t374quysvL27x585gxYwYOHDhy5MjZs2dHRESIbTYhZPz48a6urgcPHmSWsGrVKldXV1ZOenPK4MGDfXx8fv/9d6FQyGoAMxtl9erVcvfUz8/P1dX10KFDcpcgI9GdU5/qrZv1ScpRKkc2AAAAAAAA+Cjlhm6AnAIDA0+ePKmqqrp27drevXvXaluBQKCsXMcdX7t2bVxcnK6ubrdu3UpLS1++fBkTEzNy5EgZNy8sLLx//z6Vvnbt2uzZs1VVVZkZTExMLC0tc3NzU1NTf/75Z4FA4OnpKVqOnp6eiYkJlTY3N1egQ/L7FLv3UxAIBA3dhIZkY2NTWlqqo6NTVwXK8ne/ceOG6EJXV1dCiGi4EAAAAAAAoKloAt+BWYRC4e7duyMiItTV1Tdu3PjFF19Qy4uKio4cORIVFZWfn8/n88eOHTt48GBq1ciRI4uLi6dNm3b//v3nz59v3rx5/fr11JKYmJjExEQjI6PFixd36tRJejmSVFZWxsfHE0J27tzZpk0bQohAIHj79q3snbpy5YpAIDA1NS0pKSkpKblz546bmxszQ+/evRcsWEAIWbRo0ePHj2/fvi02tuLs7ExlY5LUo23btj169KioqKi6utrU1HTUqFEeHh6EEB8fn4yMDELI8ePHjx8/bmlpGRwcTH0B3r17t52d3fPnz2fNmkUIuXTpkqqqqujutbe3l2M3uru7v3//fsaMGXfu3ElNTe3atevSpUtDQ0MvXbqkpaU1derUQYMGUTmpGqdOnfro0aOkpCQ+n79gwQLqz1dYWHj48OG//vqroKDAxMRkwIAB48aNU1FRISKHAR1bYXZT0j5hbi56zJSUlBw/fvzevXs5OTmamppDhgyZPn26lD0vyx/ozp07q1at0tbWDg4O1tPTmz9/fnx8/OjRo+fNmyd9DzBJ6c7z58+Li4uLi4uld61Wp5W9vb30PzEAAAAAAMBnqSk9E0SJjIyMiIjQ1tbesWMHHVgRCATz588PCwvT09Pz9PQsLS3dunXruXPnmBseOnRIVVXV2dlZXV2dWnLkyBFLS0sLC4uMjIzNmzfLWI4oVVVVAwMDQsjGjRt/+eWXP//8s7y83MzMjJnn1q1bc/6Vn5/PKuHixYuEkP79+3/55Zf0R1G5ubm5ubmEEE1NTbEZmLXcunVLeo/S0tK6des2YcKEwYMHv379es+ePTExMYSQQYMGGRsbE0JsbW1Hjx5NRzQIIRwOR9JOYO5e+XYj5ddff9XS0lJRUYmOjvbx8Xnw4IGFhUVOTs5PP/3E2m8nT57s1q1bz549k5OT/fz8CgsLq6qqfH19IyIilJSU3Nzc3r17d+jQoS1btohtp6Ojo2g3Je0TmthjZsGCBadOnSoqKurfv3+XLl2oyJSMO0FSNicnp1GjRpWUlGzdujU0NDQ+Pt7GxoYKaUnZA6zCP9qdj3ZNjtMKmi3DQWlch+TwW6WNqurMXIH32myzYek8xxTDQWm9p7w6cKaIWtXF8wXXIdn3p1w6s9XIdK5D8nL/PGYJo5dmcR2SWTnpzan/tJxTuni+2HWiEDdgAQAAADRPTe++lYqKCkJIz549bW1t6YX3799/+fIlj8dbuXKlmpqakZHRtm3bQkND3d3d6TyDBw/28/NjFuXp6enj40PdhZGZmVlaWvr48eOPliPWmjVrduzYkZaWlpaWRghRVlb28vL69ttv6Qzv3r179+6d2G0TExPT09MJIa6urgUFBRcvXoyJicnOzqa++VPCw8PDw8OpNIfD+eabb8QWxayloKBA+p7Zu3evqqqqUCisqKjIz8+/f/9+dHR09+7dPT09Hz9+nJ2dbW9vP3XqVGb5UmIrzN17584d+XYjIWTMmDEzZ8785ZdfQkJCKisrDxw4oKmpOXTo0KqqqrS0tJYtW9I5J06c6OnpKRQKJ0yY8Pbt2xs3bhgYGLx+/VpJSWnv3r16enpOTk7Lly+/fv36nDlz6A2Z7fTz82N1U9I+oSsVPWaePHmSnp7O4XD279/funVr8u/TRrIck9KzzZ49Oy4u7uHDh9HR0RoaGqtWrWI9dCO6B0aNGsXM8NHuMMl3OoieVgByqxIIVZQlDjLyGb/i7Z9/fzBooeTSXb3ofU1iRuX1h2WzR+vKuHnuu+pzd99T6dBLxT8tMFBT/Z8WWrZS6dRG9XWO4O/kiu9351ZWCZdO1qvbLgDUofj4+GnTpj19+rSsrGzkyJH0pUWtDB48+NKlSytWrNiwYUOdt5DJwMAgPz8/LCyMvumyPtVbNwEA4PPQ9O5bMTMz43A4169f/+mnn+g5GnJycgghFRUV48eP9/Dw2LZtGyHk7du3lZWV9IY9evT4f/buPS7H+/8D+PvudJdOIpFODqWWHL5jihw6OBQ2mTLnpK3fbGpsxjwqhBCGhTFbYlKW44xkThE6kFkLqaQjldJBp7vT/fvjs927d1eXu3Tk9Xzsscd939fn+lzvz+e67tzX5/ocJLLq378/ifUBqaiokCafBpmamgYEBAQHB3t7e48cObKmpiYgICArK0uUwMHB4eo/xBtNiOjcuXNEpK+v369fv//973/q6upCoVCi60qvXr0sLCysrKxmzZp14MCBESNGNBiG+FHYbXZjJRIIBAcPHnRycrK1tbW3t2ezvTTW+iOhtra2/ofi1dvsaiQiY2NjIlJXVycibW1tdXV1OTk51ieivLxcPKW2tjYR8Xg8NsVMXl4eO666urqGhgYR9e3bl6VknX3qxylBKBS+sk7qXzO5ubnsoKxhhYhYC4iUlcCRTF5e/qOPPmKB2dra9u7dWyLg+jXQ1OJwF615XysAoZDcNubJmCebfpSelVeTX1Trse250Yw0lXEpQ+dmHDxbIkrJ+ptsPPhi9CeZfMuUiLsVok8mLMlWt348aHb6rfhKlpgjn8ZUVglvxVcQ0eXvdS7s0ok+oJd3oZ/vZ91fuaPI4fMl1TXC/rryGqoyhS/rTkVIdo2ZYqn82/befwTp2wzvQkT1EwB0KLt27YqOjjY0NPT393dzc2v7AKqrq6VPPHz4cHNz827durVePI1pUpzwtklISBAKhbt37yYiHR2dQ4cOZWdnV1VV5efnx8TELF68uL0DbBrx4jBpaWlCoZB1/S4qKhI25OjRo2xHoVC4fPly0b4XLlxgH27btq2x/NvYwIEDDx8+nJmZKRAInj17FhYWNm7cOOI8d1evXhUKhceOHRPP58aNG0KhMDg4uM0il6i6DnuxmZiYsJPORnI0Jjw8XCgUStla3aTE+fn5QqGwwVZ4jk2tofP1Wxk+fLiLi8vGjRvPnz9fUVHh6ekpJyenpaVFRIqKiuvWrZOR+bvBqLq6WlZWVrQjn8+XyIqlFO+LIU0+DQoPDx83bpy2tra2traZmRm7j3358uUriyMQCNgEnxkZGWxOE1GGzs7OothE8600VWMlunnzZkhISJcuXTw9PbW0tA4cOHDv3j1RWxVLKT69qKKiYmVlJSsR65sjQbx6m12N9E/DBCs499yoKSkpVlZW1dXV6enpRNSjR48ePXoQUXFxcVFRUdeuXdnYHLapwTglinn9+nWOOhHfpf41U1xcnJWVxQaCsVldpawEjmQlJSU//PADESkoKISFhU2aNEliRpX6NSC+VZriSFO0pn6t4C0nFJLbxtyAMyX/M+aHf6ejriIzdF5GYlrVqMGK8+zVDpwpXrQ+t7pG+InDvz1HvPcVWA/r4mijoqL092W2PuCFm4N6aXldzP3KhT45SSf6VFULx/5fFnc+9Skq8HR6yGXl1czzzp06WnmEmeK4/ykN0P/PTOHHr7y88/Dv5ptnBZINx6wFZ9YE1Wf5NQd+Kwn8rWT2RNX6B8rKq8nKqyEiNeXO98QC3h5WVlZsvHB8fLyHh8eaNWsmT56cn5/v4+Nz/vz5Z8+eGRoaLlu2bOHChUTk6up66dKl58+f19bW9uvXb8mSJZ9//jkRmZmZsTnmfH19fX19Bw4cmJCQwP7tiIiIGDdu3J07d9577z0iqqioUFRUZH1PfH19z507d/v27XPnzk2YMKGxg0q4c+dOQUEB64fbtWvX4uLizZs3nzp16s8//xw7duyBAwc2b9586NChrl27btiwYcGCBfRPV5cNGzZcunQpLi7OyMho165do0aNYhk+f/58zZo1Fy5cyMnJMTAwmDt37vLly9k/ZBJxitpWxIvZWJ2I737lypXY2Fh9ff39+/ez4xYWFvr6+v72228ZGRnq6uouLi6bNm2Ssgag4/vll18sLS3z8/MjIiLU1dVNTExsbW337t3b3nG1mNu3b6uqqhLRkCFDFBUVc3Jy2M/O5ORkMzMzlsbd3X3Hjh21tbUDBw4Un0ygI7Czszt9+jSfzy8rK7t9+7aCgsLYsWNTU1OvXbvGce4OHDhgZWX1/vvva2hosAeT/fv3t7S0ZJvaqyxtc7FNmzbN19fX1taWPT+GJul8bStEZGNjo6iouHbt2oiIiKqqqrVr11pYWOjq6mZlZQUFBQ0fPry8vDwlJUVGRqapSwg1Ox8/P7+dO3fq6upqaGikpKQQUffu3dm8ttyuXbtWVlbG4/FGjhzJfppUVVXdvn07Nzf37t27r98poLESTZo0iSWorq5OSkp68OCB+F6sH8TFixcrKipMTU1tbW0HDBgQHx8fEBCQkJBw5syZ5h20qaeDW2ho6LNnzzIzMwsLC5WUlKytrZWVlXV0dLKzsz08PAYNGnTz5k0isrKyEh9JxFFMUbNCg3XSGAsLCwMDg/T09M8//9zS0lIgEFRXV69bt07KSuBItnnz5ufPn9vb2w8ZMmTz5s3r1q376aef2D9vjdWAeM6i7kVNKo6UsTU1K3h7eO4tSEyrshyidHZ7b3UVmVMRpYlpVUp83pH12sqKPL2ech9vyN18qFC8TcRlqtpPXv/pzbdyvsZat+53HlaOWJiZklVd9LLualz5K/NpUOgm7U835cWnCP56LCAiBXmet2s3T5d/H4PnvqjNfdFAXzwiir1fmfC4iohmTlB5ll974LeSy7fL03NqDHr9+0/nnuNFe47/PdWRjAx9OQcDgqDjcnR0zM3NTUxMHDRokI2NjYWFRVVV1ZgxYxITE0eNGjV//vyAgAAXF5fq6upPPvkkPj7eysrK0NAwKysrICBgyZIl7Hf8/Pnz9+7dm56ePmLEiJEjR7IelAzHwGEvLy9ra2tHR0cVFRWOg76yCN7e3jY2Nnw+//fffzczM9PQ0DA1NY2Jifnkk08mTJggCmbr1q3Lli3T1NQ8fvy4nZ1dSkqKlpaWQCAYPXp0UlKSkZHR7NmzT58+7eXllZCQEBISUj/O8vLye/fuSRSzsToR7b5u3To3N7fS0tKYmBhnZ+fk5OSqqqqxY8cmJCR069Zt1qxZL1++fPDgwevUAHQoioqKrAXNxsbmr7/+IiIFBYU+ffqIEvTo0cPHx2fSpEm9evVKT08/cuTItm3b2PwG+fn53bt39/T0tLGxGTFiREZGhpub261bt9qpKI2aMGECe5GYmGhsbPzLL78sXbqUfcL6yL948UJfX3/GjBmhoaFffvklERUUFDT227uN8fn8wMBAPp//119/2dnZPX36lIjU1dWNjY25z93x48d3796tpqY2a9Ys1njh7OxMROnp6ZcvX26XsnAH3IJXmoaGxsCBA9lKIByMjIz2798/YsSIR48efffdd+KbAgICxo8f36NHD1lZ2dTU1N27d+/ZsychIYE9JPb09PT09Lx//76ZmVmDKYmoSYlF+vfvf+XKFQsLi4cPH7q7u9cvo6am5po1a+zt7bW1tVNSUnbs2HHw4EEp6r4JXrttpfF/RyX8/bj8v7d/IhENLc7KYdSoURs3bvTy8rp169aqVat8fX137dr1888/x8TEHD58uEuXLgYGBvb29k3Kk4jk5eWblw9bsSUzMzM9PV1NTc3S0nLRokXSPNJnY39GjBjh6+sr+tDNzS05Ofn8+fOv37bSWImsrKzi4+OvXLnyww8/DBs2bMSIETdu3BDt5eTklJSUlJKScvLkybKyMltb26VLl27ZsiUtLe2PP/5wcXHh7tfX7GpskkWLFkVFRaWnpxsaGrq7u7NxQLt27Tpw4MCdO3cuX76spaX14Ycfzpo1q7EcJIq5YsUKjjppjLy8vL+/f1BQUFRU1KVLl9jsMCR1JTSW7NixY1FRUdra2u7u7kpKStHR0REREX5+fuJd4xqsARHuUyyNtjmP8IZJTKsiIpf31dRVZIgoI7eGiCoEwr7TnojSpD2rrqwSiiYusR3RRSKTQUZ8IlJX+buHVGlFnTT5NMjCTPHeEf0nT6tj7lcGh788e6PMe1/BzPGqRnp//2j43LHrrq//7vPVd9qT9Jx/V2c/cKaEiEz6KAzqz3+nj1Czq2x+Ue3B34rXfPLvT0Y230oXRV7f3vLzJ6sN7PefTjEAHcqSJUuio6PZLf3OnTuJ6NSpU4mJiUpKSsHBwcrKynp6eq6urps2bfrkk08iIyMVFRXr6uoqKiqePn169uzZixcv2trarly58urVq+np6RMmTJDorS3q4Vifi4tLQEAAe81x0FcWYdmyZX5+fp6enhs3bqysrIyNjVVTU1NTUxMIBPHx8aK2lVWrVq1cuVIoFPbr1y8tLS00NHTJkiVhYWFJSUlycnKRkZE9e/acPn361KlTjx49un37dtGO4nHa2dlJFLOxOhGFt3LlSh8fH9ZzJyUlpaioKCIigvXruXnzJpsosLq6+uzZs82uAehQKisrs7OzdXV1jxw5cvbs2ZiYmGvXriUlJbGtfD7/xo0bAwYMSE5ODgkJcXBw2LBhg5mZmfhan6tXr96/f7+Kioq5ufmhQ4eMjIzaqSjk6OgoWvBRvM30lQIDA93d3ZctWxYRETF37ty4uLj8/HzRQ9z2ZWFhwZ6kent7s4YVIiouLo6NjSUijnNXUVFx9OhRNzc3Z2fnvXv38ni8+fPnE9HBgwc5+oC3Ko6LrUWuNE1NTRUVFfaCiHR1ddkAgufPn5eVlUkklpeXDwsLMzQ0TEtLS0xM9Pf3F986ePDgiIiIlJQUXV1dV1fX3bt3JyYmHj58ePHixQYGBrGxsVFRUc+ePWss5eXLl5uUWHRcX1/fM2fOPHr06N133w0PDzc0NBSfMEFBQSEyMtLExOTWrVuHDx92dXUNDAyUl5f/8ccfW+D0/KMz9VsJDAwUfzt8+PDw8HDRWz6f7+Hh0eCOv/76K8cnOjo6V8Vadrp27Sp9PiLz5s0Tn7mWI2wiOnr0qOj1jh076u+yf/9+jt2lPIpIYyVaunSpqOFZgo6Ozvfffy/+Sd++fcW7nM2YMUP0usFq4ajGxsIWX4DG0dHR0dGR+xC6uroSraREpKGh8dVXXzV4rPqZ1C8mR51wXDNqampsbSaJXaSshMaSOTk5iV6vWbNmzZo1EgkarAHxOKUsTot/HeBt5jJV7fD5l59uylNXkZlhraKnJUdEykoyxzdry/5z2yWoFsqJjY1T4ks2jrCU4k330uTToINnSxxtVfr2lu/bW37UIKWzN54Q0YuSWqJXPJCpEAh/ufSSiBLTqmTMk8UzXP1xd1FsUyyVRU0zAJ1ORkYGEVVUVIg/aU9LSysvL1+3bl1QUNCzZ8/q6urY51L2D2cTuksYP378Kw9aWVmpqKjInTm792M//fv168deqKioCAQC8YHYbMI1Ho/Xp0+ftLQ0dkT2f01NTTbtnWg4Q2ZmpuhOUjxOCXV1dWvXruWuk8GDB9M/c8YRUWlpKRs9oampKVqBQV5e/nVqADoaJyenH374YfDgwYMGDSKiqqqqdevWsSemkydPHjBgQE1NzZgxY3Jzc0+dOnX27NlZs2Z9+eWX7F6RiPz8/NasWTN8+PDbt28bGhp27dq1/rKPbaNnz54SM0JKKScnJyQkxNnZmfUQ2bFjB2uG6AhYwwoRJScn19/Kce6IKCAgwM3Nzdzc3MTEpFevXn369BEKhVLel7WSxgJukStt586dc+fOFb1lE1wQ0fz584OCgiQiYT346urqxo4dm5mZ6eLiIj5UasyYMZWVlTIyMkpKSr179546deqECRO++eYba2trAwODixcvenl5caS8fPmyn5+f9IlFx926dau3t7eCgkJaWpq2tvbMmTPFewNMmTLFxMSkoqJizpw5ZWVlmZmZAQEBq1atenvbVgAAoIN7f6yy1fAuC31y5njlHNusPcVSeYC+QlJG1cbAFxPMu7wsq7uXJJCVpSmWDS8k35hm57Nofe7nW/IG6Ctoacj+mSwgIm1NuUH9X92p8PiV0uLSOh6P3h+tzFpSKquEF6LL03NqLt8uH1+vrw1AZ6Snp0dEysrKJ06cEM2lJRAIzpw54+fnp6qqGhQUpKen5+3tHRERIXpay1KK2hdYDmVlZWxWgvj4+PoHYnPScx+Ue5I1hnVTZ4OPOLqs//nnnzNnzhQIBA8fPhQdkf0/Pz8/Ly9PS0uLzRpDRGyutPpxShTz5MmTHHUivov42Cg2z31+fn5SUtKAAQOIqLq6+nVqADqa6OjoIUOG9O3b19zcfO7cuVOnTt2wYUNoaGhycrLo7LM2uISEBLaLnp6e6I6XfV+Ki4vZWxUVlfZqW9mzZ8+SJUvY67S0NAMDA+n33b59u7Oz8+TJk7Ozs0NDQztO24qo9dPIyKj+6HiOc0dEsbGxCQkJZmZmzs7OrPn18uXLrLW0vTQWcItcaTt27Dh+/DgR2djYuLu7f/zxxwUFBUR0586d+pGwP2IFBQWZmZlEdPfuXdEmGRmZtWvXzps3T1tbW9SZscFmO+lTSpmYFbyqqiopKUlbW1u0zAjD3iopKYlm5CSiPn36sElFGzxoM+CPOAAAtKT59qrFpbUe257PXPXs1Bbt6/t1NwS8OB9V5hv4QrWLjGlfhYWT1JqaJ1+B17x81n/a/XJs+aOM6odpVd3UZKaNVV73afcuiq8ezRp4ppiI7Ecqn9727xJdwxdk3H0kCPytBG0r8GaYMmXKgAEDkpKSfH19J06cWFJScu/ePVlZWTazABEJBIK4uLjo6GjxvdhN15EjR0pLSy0sLObMmTNs2LDr1697eXndvHlz3759zTvolClTWqpc3377bWpqalJSUm5uroqKysyZM4lo8uTJhoaGKSkpY8eOHT16NOt96eTkVH8ZvgaLKfo132CdNGby5MnvvPPOw4cPR40a9cEHH1RUVAgEgpCQkNauAWgzCxcuPH78+JMnT548eXLr1q2pU6cSEVvcit12ampqamlp5eXliVYkEF9IlM2O117DTFpKfHz85cuXbW1td+3a1aHW2IqKisrJyenVq9f69etjY2NZQwObbyU2Npbj3DGBgYHffvvtggUL1NTUqF1nsWUaC7hFrrS4uLi4uDiveeQZAAAgAElEQVQi6tq1KxFduHBBfHcJbJOGhoa6unpxcbFobVYi+vDDD1euXPny5ct58+ZlZmauX7/eysqKtTizGER/SzlSNjUxY2pqSkTy8vKsIZtViwh7W1ZWNmPGDNGslHw+v8G+ls2GthXofDAaBTUAHdDz3/+dvXuJU9clTl1Fb/2X9yBqYOCM+C71PzHUla+L+Xc8sJaGrPT5iHi6/GfmWnF/hUg+lHvy678/Dq7s1aV67vz87zOQ+rsDdDp8Pj8yMnL9+vXnz5/fsGGDqqqqqampi4uLk5NTZGRkSEjIihUrxo8fzxbaEO315Zdf3r179969e7t27SopKZkzZ86ePXsWLVr04MGDq1ev+vj4NDYclfugLViu9evXnz179sGDB0OHDvX392fPNhUVFW/cuLF69erff/89ODhYX1/f3d19xYoVjWUiUcyAgACOOuEurK+v79mzZ4ODg9XV1dl8fK1dA9BmAgMDv//++6SkpLy8vCFDhhDRs2fP2FSjYWFhKSkphoaG169fv3HjxrRp04jo2LFjook/3iSLFi0aMGAAR5uj+HwuDx8+bJsLXiAQLFq06NSpU4MGDUpOTr537568vPzAgQMPHjwYGxvLce6Yw4cPb968mTW/FhUVnTp1qg1i5tBYwH/++WcbX2kRERGpqan9+vU7efJkeHi4aMU0Elvmlc/nDxs2THwFDNbrZ+7cuSoqKtHR0aJegfVTNjUxs2LFCiMjIyMjI21t7dLS0tDQUPGt586dY/0HPT09f//9dzU1taFDh9bW1p47d+71K0QEbSsAAAAAb4WgoCCJkfNaWlq7du2qn3LPnj0SSzCIGBoaStxBmZmZsbkhmS+++EL0Oj8/v34OjR1Ugvi+4t3XJWYTq38IIyMjttq0hJ49e/7www+vPBZTv5gcdSK+u6GhofjD4e7du2/fvn379u3i6aWsAeiY2Egu1kHDy8tr/PjxxsbG77zzzosXL3799Vdvb+/y8nIiqqysHD169Lp16yZOnDhnzpyMjIxdu3Zt2bKlnaNvHRkZGWwiocaIz+fSlsPfzp8/P3z48JUrV1pbW48YMeLFixfXr19ng184zh3z/Pnz33777cMPPySi4ODgFhw5IiXxK4074Da+0qqqqiZPnszWCWJ/4kRTQB47dmzMmDGzZ8/esmXLpUuXwsPDHRwc2Kbt27e/++67Q4cOdXd3V1NTc3V1bSxlUxMzq1evnjp1qpmZ2b179zw8PCTmwxIIBGPGjPH29ra3t/fy8mKrtrX4BDq81+2EJvU6Qdyauk4QALwZ2NLRnbE3LOuIKN6xAgBE2ATAnfGrTf98uztp8G85TU3NgoKCU6dO1f/ZDe2ic32bONYRF9HX109KSuLz+e7u7tzrZrYeKetTmuJ0ZNzF7OCle/3gO8KVBiJSfukaXS2vBVhYkFBIQiFVVtKVK6TbQP9qAAAAAACAjs/f3z8hIYHP5xcVFZ04caK9w4E3Fq60Tqo121aYDRto9myytKTNm1v9WAAAAADwtsrPzxcKhei0Aq3EyMiovLw8LCzMxsZGtAILQIvDldZJtf5ot8eP6dQpevCABg9u9WMBAAAAAAC0Ant7+/YOAd4KuNI6qdbvtyIjQ3p61Lcv/XcZJAAAAAAAAACAN0Dr91sJCCAiys8nH59WPxYAAAAAAAAAQNtq/bYVHx86cYKSk6nNl60CAAAAAAAAAGhtrd+2kpZGf/3V6kcBAAAAAAAAAGgPrdm2Eh1NHXvhcQAAAAAAAACA19T6c9kCAAAAAAAAALy5Wn9MkHSsra3bOwQAAAAAAAAAgCZDvxUAAAAAAAAAgObjCYXC18ughWZUec0wAADaFg/zSQG8yuv+xmgn7NvdSYMH6FA617eps/zLLmV9dpbiNIa7mB28dJ06eKhPyi9dRxkTBAAAANBx4LcvAAAASA9tKwAAzdFZnsIBAAAAAEBrQ9sKAAAAwL/QcgoAAABNhflWAAAAAAAAAACa77X7raBNBAAAAAAAAADeYliDGQAAAAAAAACg+dC2AgAAAAAAAADQfGhbAQAAAAAAAABoPrStAAAAAAAAAAA0H9pWAAAAAAAAAACar+F1gngttbKy1F53KWgA6PBa8A8L/mIAAAAAAEDHgX4rAAAAAAAAAADN13C/FaZtngy3fR8ZAAAAAAAAAICWgn4rAAAAAAAAAADNh7YVAAAAAAAAAIDm4xoT1JYwMggAAAAAAAAAOiP0WwGAdmNhYSEUChcuXCjxoq6urqio6OrVqxYWFu0dIwAAAAAAwCtI228lOjp65MiRampqOTk5SkpKX3zxhb+//9y5c4OCgq5fv758+fIHDx7o6emtWrVqwYIFLHFgYODChQvrb20w/6tXr7ZcoQCgI7K2tpYy5bZt2xISElavXh0eHm5kZPT8+fNWDQwAAAAAAOB1NG1MUElJycmTJ52cnIKDg9kn2dnZ9vb2w4YNO3369OHDh52dnfX19RUVFTm2WllZtWwZAOAN8+DBg59//jktLe3atWtz5sz57rvv2juiBmAkI4A02mbNwZaFbzdAy+qMfwcAAJqqaW0rlpaWBw8eVFRU5PP5ffv2JaJjx46Vl5d/++237733noWFRUhIyKFDh/7v//6PpW9wK9pWAEAaDx8+JCIDA4P2DgQAAADefJ2lXVXKtqrOUpzGcBezg5euUwcP9Un5pWta24qzs/Nnn31WWFg4f/78Y8eOEdHTp0/pn5sfFRWV7t27Z2dni9JzbwUA4GBiYkJEGRkZ7R0Il7oYo/YOAaCDkjFPbu8QXguetAO8PtxDAsDbo2lz2erp6Y0bNy4uLm7hwoXsE11dXSJKS0sjotLS0oKCAj09PVF67q0AAA0yNTWdP39+QEBAcXFxSEhIe4cDAAAAAADApcnrBK1YseLLL780NjZmb52cnJSVlb/66qsLFy4sXry4urra2dlZlJh7KwBAg5YvX75nz55nz57Z2dnl5ua2dzgAAAAAAABcmjYmiIgmTpw4ceJE0Vttbe3w8PCvvvpqxowZ+vr6hw8fHjt27I0bN4iIz+c3uLUlwweAziw6OlrUW7j+CwAAAAAAgE5B2rYVCwsLiYHHKSkp7MXo0aNjYmJEnz958uT7778nokGDBtXfCgAAAAAAAADwJmnymKBXCg0NvXr16tatW83MzFo8cwAAAAAAAACADoXX4DT4rE9+28yQz4519erVNjgWALQja2vrlsqqI6zfwf52YZ0ggMawdYI6wre1qdryVxDAm61zfZs6y6hkrMFMHb50nTp4qE/KL13L91sBAAAAAAAAAHh7oG0FAAAAAAAAAKD50LYCAAAAAAAAANB8HbdtxcXFxdraevz48VlZWeyTpUuXWltb79+/n3vHadOmWVtbs3WgX19+fv6mTZscHR0nTJgwbdq0xYsX//rrr61xoBbPkFWguDVr1jQvqxUrVlhbWwcEBLRIYBxYzN99911rH6gxbVbSNiPlRdXiFzMAAAAAAMDbg2sN5o4wy05tbe3+/fvXrVsn/S7GxsalpaVqamotEoCPj09CQoK6uvrQoUNLS0szMjLu3r07bdq0Fsm8DWhoaPTq1Yu91tPTa5cYampq5OSkXe27vXSKIFtPy35r6K2vTwAAAAAAeKt09JsfHo8XGRmZkJAgsaLzli1b4uLiiouLa2tre/fuPX36dAcHB7bp0aNHJSUlJSUlW7ZsOX/+vL29/YoVK4jozp07X3/9taqq6okTJ+Tl5YuLiw8dOhQTE1NQUKCjo+Pk5GRnZydx9Kqqqvv37xPR9u3b+/XrR0Q1NTU5OTniaXJycr7++uv4+PjevXsvX7584MCB3OFNmzatpKTE1dU1Li4uKSlJR0fniy++YHuJayy8jRs3Xrx40cHB4YsvviCimJiYb775plu3bseOHZORaaAX0rhx41jKV+bcWNguLi5paWlEFBQUFBQU1KdPn8DAQLbgy86dO4cMGfLo0aNPP/2UiC5cuKCgoCAq48cffxwVFfXo0aNNmzYZGRm9srbrmzp1allZmZubW2Rk5OPHjwcPHrxy5crg4OALFy6oqKi4urpOnDiRu0qLiooOHjwYGxv74sWLXr16jR8//qOPPpKXl28wyD179tQvaWOnUrT73bt3ExMTtbS0RGf/5cuXQUFBt27dysvLU1ZWtre3/+STT7hrXpoTFBkZuXr1alVV1cDAQA0NDQ8Pj/v378+YMWPJkiVSXlQcV6boWyN+lTZYusbCq1+fw4cPlzh6/eXA2IXUWZYPAAAAAAAAaFDDY4KErYDlfLURjcU3fvx4Ho+3b98+ic9TU1OHDh06Z84cOzu7rKys77777u7duxJppkyZQkTXr1+vqqoioosXLxLRxIkT5eXla2pqPDw8Tp06paGhMXv27NLSUj8/v7Nnz0rkoKCgoKmpSUS+vr4//fTTzZs3KysrdXV1xdPs27evtrZWTU0tLS1t06ZNUoZ39OjRoUOHvvfee8nJyStWrCgqKhLfyhEeuxO+dOkSK9S1a9dYLTXYsMISfPaPa9eucRe8wbAnTpzYs2dPIjIxMZkxYwZrzmC4ezYFBAQoKCiMGzdOXl5emtpuzIEDB1RUVOTl5e/cuePi4hIdHW1gYJCXl7d169aCggKOKq2urnZ3d//1119lZWVtbW0LCwsDAgI2b97cYJBKSkoNlpT7VB46dKhPnz4GBgais19TU/PFF1+EhoYWFxfb2NgMGjSItddIeclxJBszZsz06dNfvnzp5+cXHBx8//59Y2Nj1qrVWA3Ur0xpvjjcpeMuhXh9vvrUQieU/bxmoU+u7pQnfMuUHhNTLRZl7j1RzDb1mJgqY558+lppSx2rxTPsdAEAtIb79++PHDlSWVmZx+OJHvw0iZ2dHY/H8/LyavHYJGhqavJ4vNOnT7f2gRrUZsUEAIA3Q8edb4Xp37//hAkT7t+/zxoRRPz9/VetWjV//vzPP//c3NyciO7cuSOx78CBA/v06VNWVhYVFSUQCCIjI+mfBpeoqKiMjAw+n+/t7e3g4LBw4UIiCg4Orh/A2rVr+/Xrl5qaeuTIES8vr+nTpwcFBYkn+OCDD7Zv386mMsnOzi4tLZUmvLlz5zo7O69Zs6ZXr17l5eUSrUsc4ZmamhoZGZWWlkZGRtbV1d28eZOIOPqAFBYWPvzHixcvuAveYNizZ8/W19cnouHDhy9ZsmT27NmizLnbVuzs7LZv3+7l5VVSUiJlbTfI0dFxy5Yt7PdfVVXV3r17/f39WQNZamoqR5VGR0dnZWXJysr6+/t//fXXq1atIqIrV66It8iIghw4cGCDJeU+lbNnz3Z3d2c9g9jZj46OfvLkCY/H271798qVK318fHx8fEjqS4472eLFi42MjG7fvn3gwIEuXbqsXr1afNwN90XFSPPF4SjdK0shXp+NnlHozGZ55vwcVlJVI7R6V6m/rnxiWtWV2+XtHRQANMGuXbuio6MNDQ39/f3d3NzaPoDq6mopUw4fPtzc3Lxbt26tGk+DpA8S3kIJCQlCoXD37t1EpKOjc+jQoezs7Kqqqvz8/JiYmMWLF7d3gE0jXhwmLS1NKBSy55FFRUUNPjU/evQo21EoFC5fvly074ULF9iH27Ztayz/NjZw4MDDhw9nZmYKBIJnz56FhYWNGzeOOM/d1atXhULhsWPHxPO5ceOGUCiU/hbm9UlUXYe92ExMTNhJZ50SGhMeHi4UCjds2CBNnk1KnJ+fLxQKG3xawLGpNXT0thUicnV1VVBQ+PHHH2tqatgnQqHw4MGDTk5Otra29vb2UVFRRFRYWFh/X9aScvHixRs3blRUVJiamvbt25eI8vLyiEggEMyaNcvBwWHLli1ElJOTwzqDiDM1NQ0ICAgODvb29h45cmRNTU1AQIBoel0iGjx4MBF17dqVva2oqJAmPG1tbSLi8XhsMhQWjwh3eOziCAsLu3fvXklJyYABA1ihGuTg4CDqHDR9+nSOnKWvVQm1tbUNfj5s2DBpivNKxsbGRKSurk5E2tra6urqcnJyrFtEefm/N3X1q5QdV11dXUNDg4hEtfT8+fP6QTbolXXSv39/IlJWVmZvKyoqcnNz2UFZMw0RseYPKSuBO5m8vPxHH33EArO1te3du7f4vtwXlTTFkVC/dK8sBXd9QmdXWSW8FV9BRJe/17mwSyf6gF7ehX6+n3UXT5P2rGaSe3aXMSmDZqffiq9kH368IbfPtCfKY1MUR6eYfpS+5/i/napY3xDfwBc2i7PUrR8PX5Ah2ktcflGtx7bnRjPSVMalDJ2bcfDs3+PXFqzNkTFPdt/695c67FaZjHly78mptXX/2d11fa6MefLHG3LZ299jymXMkzUnpAqqhByxiZMxT5YxT752t4KI7jysZG8rq4QcsRHRt0cKB8xIUxqTom79eMicjJ/DShrMHKDNWFlZ/fDDD0QUHx/v4eERGxubn5/v7u5uaGiorKw8ZMiQgwcPihK7uroaGBh06dKFz+e/8847e/bsISIzM7MLFy4Qka+vL4/HY0O2eTwej8djj8Hu3LnD3lZW/v1dZt1PNm7caGlpqaCgEBERwXFQcXfu3ImJiXnx4gURde3alcfj+fn5WVhYKCkpTZo0KTs7293dXU1NTV9f/+effxbtxQ7n6+trbW2tpqY2bNiwW7dusU3Pnz//7LPP+vfvr6ysbGpq6uvrKxAIGgxSQUGhfjEbrBCJ3cePH6+mpmZmZiY6aGFh4fLly42NjZWUlHr16sWe9BCRlJUAHdwvv/yyYMECdmE/fvzYxMTE1ta2vYNqSbdv346JiYmJiWHf6JycHPY2OTlZlMbd3V1WVpaIBg4cKN7DvSOws7OLi4ubN2+ehobG7du3MzMzx44d6+TkRJzn7sCBA0T0/vvvs5sIIurfv7+lpaVoU7tom4tt2rRpCQkJrC8/NFVHn2+FiLS0tGbMmBESEvL06VP2yfXr10NCQrp06eLp6amlpXXgwIF79+4JG5qyYeLEifv374+JiSkuLqZ/mlpYnkSkqKi4bt060Wia6upq9ndBXHh4+Lhx47S1tbW1tc3MzNjt6MuXL0UJ2J2zeA8OacJLSUmxsrKqrq5OT08noh49ekgUmSM8W1vbffv2/fHHH2zeEGkmLpEmZ46wWUrxIigqKlZWVrJ6EO88Io7P50tTnFcSr2GO6VHrVymr1eLi4qKioq5du7KxOfTf2hYFyUiU9JWnkqUXP/ussMXFxVlZWWz4GJvVVcpK4E5WUlLCfhMrKCiEhYVNmjRJvHsI90UlTXEkNFY6jlJI1Ce8YRQVeDo95LLyauZ5504drTzCTHHc/5QG6CuIp1nhnz/2f0rd1WXvp1Yt9MlJOtGHiOJTqqze7dJfVz47rybgTLH71ucmBgq273UR7bUtqHDpbA3NrrLHr5Taf5GdfLKPlsa/X42qauHY/8tKTKsaNVhxnr3agTPFi9bnVtcIP3FQ/8yxa9D5l0culGz9QlNRgXfySikRzbVTk/3vg4OPHdQDz5acuFq6e4WWogLvyPkSIppvr8pX4L0yNnH1++pxxBZzv/Jr//xe3WVXLewmEAj/eix4nIUn4dDOHB0dc3NzExMTBw0aZGNj8+67744ZMyYxMXHUqFHz588PCAhwcXGprq5m04TFx8dbWVkZGhpmZWUFBAQsWbLExMRk/vz5e/fuTU9PHzFixMiRI1mzPsPdm9XLy8va2trR0VFBQYHjoNy8vb1tbGz4fP7vv/9uZmamoaFhamoaExPzySefTJgwQTyYrVu3Llu2TFNT8/jx43Z2dikpKerq6qNHj05KSjIyMpo9e/bp06e9vLwSEhJCQkLqB1leXn7v3j2JYjZYIeL3NuvWrXNzcystLY2JiXF2dk5OTq6qqho7dmxCQkK3bt1mzZr18uXLBw8eEFFVVVWzKwE6DkVFxVGjRhGRjY3NX3/9RUQKCgp9+vQRJejRo4ePj8+kSZN69eqVnp5+5MiRbdu2sRa9/Pz87t27e3p62tjYjBgxIiMjw83NTdQk13FMmDCBvUhMTDQ2Nv7ll1+WLl3KPpk+fToRvXjxQl9ff8aMGaGhoV9++SURFRQUdO/evbEM2xKfzw8MDOTz+X/99ZednR27l1RXVzc2NuY+d8ePH9+9e7eamtqsWbP27t1LRM7OzkSUnp5++fLldikLd8AteKVpaGgMHDiQ3WZyMDIy2r9//4gRIx49eiSxxmtAQMD48eN79OghKyubmpq6e/fuPXv2JCQksNsWT09PT0/P+/fvm5mZNZiSiJqUWKR///5XrlyxsLB4+PChu7t7/TJqamquWbPG3t5eW1s7JSVlx44dLd6o3QnaVoho7ty5YWFhrH2ExDpKVFdXJyUlsX+lGqSmpjZmzJgrV64kJCR06dKFTZxJRBYWFrq6ullZWUFBQcOHDy8vL09JSZGRkbGwsJDIwc/Pb+fOnbq6uhoaGikpKUTUvXt3Nq9tY6QJLzQ09NmzZ5mZmYWFhUpKSqLApAmPz+fb2dkdO3YsJiZGTk5u/PjxHMFI4MiZI2zWD+LixYus74+tre2AAQPi4+MDAgISEhLOnDnT7INKH/kr1a9SZWVlHR2d7OxsDw+PQYMGsfFTVlZWHH/uJUoq+pn4yitNxMLCwsDAID09/fPPP7e0tBQIBNXV1evWrZOyEriTbd68+fnz5/b29kOGDNm8efO6det++uknVVXVxmpAIjbpvzgcpWuDUwkdWegm7U835cWnCP56LCAiBXmet2s3T5d/e+x/+qG6//IeN/+sGOOWlZJVXfSyrquqzPX9uooKvLo6qhDUPc2vOXuj7FJsuXj7xTfO3VYu0BAKqf/0tLRn1aGXXi5x6iraeu5mWWJalRKfd2S9trIiT6+n3McbcjcfKvzEQd3CTPFdY/7dR4KTV0s/mqD66/UyInKeIrni1chBigP7KdxPrTp7o2zyKOVTEWVE5DpNnYheGZu4+hNbccT2sqyOiPjyvD7acgP0FTwXdeui2P6r78FbbsmSJdHR0eyWfufOnadOnUpMTFRSUgoODlZWVtbT03N1dd20aRO7w4+MjFRUVKyrq6uoqHj69OnZs2cvXry4efPmq1evpqenT5gwQaK3dmNTvzEuLi4BAQFExH1QbsuWLfPz8/P09Ny4cWNlZWVsbKyampqamppAIIiPjxdvW1m1atXKlSuFQmG/fv3S0tJCQ0N1dHSSkpLk5OQiIyN79uw5ffr0qVOnHj16dPv27aIdRUESkZ2dnUQxG6wQ8bYVNhb4zp077733XkpKSlFRUUREREJCAo/Hu3nzpomJCf0z2ujcuXPNrgToOCorK7Ozs3V1dY8cOXL27NmYmJhr164lJSWxrXw+/8aNGwMGDEhOTg4JCXFwcNiwYYOZmZn4+PrVq1fv379fRUXF3Nz80KFDRkZG7VQUcnR0FK1CIP5VeqXAwEB3d/dly5ZFRETMnTs3Li4uPz9/0qRJrRNm01hYWLDf9t7e3qKH9MXFxbGxsUTEce4qKiqOHj3q5ubm7Oy8d+9eHo83f/58Ijp48CDHU8lWxXGxtciVpqmpqaKiwl4Qka6uLnuk/fz587KyMonE8vLyYWFhhoaGaWlpiYmJ/v7+4lsHDx4cERGRkpKiq6vr6uq6e/fuxMTEw4cPL1682MDAIDY2Nioq6tmzZ42lvHz5cpMSi47r6+t75syZR48evfvuu+Hh4YaGhuK9+BUUFCIjI01MTG7dunX48GFXV9fAwEB5efkff/yxBU7PP157TBCPJ+V/QiIhkZW1dYP/cR9EWVl5wYIFordWVlbTpk2TlZX94YcfHj58OGLECI59p06dyl7Y2NiI5teUl5fftWvX9OnT8/PzDx8+fP78+aqqqvr3okTk6ur6zjvvFBUV3bt3T0ZGxtLScsuWLdwP56UJb9GiRfn5+enp6YaGhps3bxb1N5MyvGnTprHb/lGjRoluraXBkTNH2E5OTu+8805hYeHJkydv375NREuXLjUxMXn69Okff/zh4uLS7IO2oPpVqqCgsGvXrqlTp1ZXV1++fFlNTc3FxUXUF7dBEiVt0pXGyMvL+/v7Ozk5qampXbp06e7du2zpaykrgSPZsWPHoqKitLW13d3dJ02aZGVllZeX5+fnx1EDEpk3ozjShwdvCQszxXtH9B+f6hO8odfU0cpV1ULvfQXJmf92xxg9VImItLr93XBfWlFXV0c+PxboTX2iYJmsavX47I0yIsp98Z+xhH17yxMRj0d9tOWIKDOnRnxrRm4NEVUIhH2nPdGalMqG9qQ9q2ZDchY7diWiwN9KIuLKC4prh5nwzfr/pysNw1pSjpwv+fV6aWlFnYWZoll/BWliq69GbDtHbNbDuzhPUct+XrvQJ3eUa6bmhMfbg189yhKgLWVkZBBRRUVFnz59evTo4erqSkRpaWmVlZV1dXVr167V1dWVl5dXUVFh05azca/cRCO4JYgeBXEc9JWZs3s/9tO/X79+mpqaCgoK7H5AvE8x/TMQmMfjsee6GRkZ7Liampqsr7toAcrMzMz6QdYnTYWwceJsFDMRlZaWsp6kmpqarGGFiNjT4NepBOhQnJyc4uPjBw0atGrVqtOnT+fm5np6erJNkydPHjBgQE1NzZgxYz7++GPW8WHWrFniLRd+fn4eHh5LliwhIkNDQ9EkA22vZ8+e5v9g635KKScnJyQkxMLCgvUQ2bFjR+sF2VSsYYWIxEcwiXCcOyJizazm5uYmJibjxo3r06ePUCgMDAxsm8gb1FjALXKl7dy588mTJ0+ePNm6dSsRRUVFsbesd5IE1oOvrq5u7Nixc+bMEXVlYsaMGePs7Ozr6/vll1+eP3+eiCZMmODn55eYmEhEFy9eXLp0KbuFaTAli1b6xCJbt26dOXOmubn5s2fPVFVVZ86cKb51ypQpJiYmFRUVc+bM2bNnz9q1a8JcT7cAACAASURBVImI+8awGTpuvxWJa/fDDz/88MMPRW+XLl0qcRZFfv31V/G3//vf/xqc1LNr164eHh6vDGPevHnz5s175YF0dHTEj8IRHqOrqyvRe0oiQ+7wdHR0+vbtm5qayjEgqLEvf2M5y8jINBa2jo7O999/L/5J3759WQc5ZsaMGeJbJU4Bx0G5YxZfgMbR0dHR0ZHjEA1WqYaGxldffdXgsernQA2VtLE64Tj7ampqbGEmiV2krITGkjk5ObHRocyaNWvYDMoir7yoOE6xRG1wlK6x8BqsT3jzHDxb4mir0re3fN/e8qMGKZ298YSIXpTUEv3dd1RBnui/Y2dORpT6/Vyo2kXmsE8vvZ5yq/cVRNytkHjqE58smDleRVAlfJhWRUS6Pf/zb5OelhwRKSvJHN+sLRrsI6gWyskSEc2ZpLrC//mVO+UK8jwicp4q2WmFmW+vumpP/vmo8vziOvqnqUWa2BhlJZmyirrCkjoiik8RSBObDI8CV/f8fqXWgydV526Wrd1f4PPjiy/nSDZ6ArQj1vqvrKx84sQJ0ehOgUAgJyd38uRJPz8/VVXVoKAgPT09b2/viIgI9sCWpayr+3daI2Vl5bKyMjaBV3x8fIPHEj3f4jjoKwNmDRPs8RJ3l/U///xz5syZAoHg4cOH7KA6OjpElJ+fn5eXp6Wldf/+fZZSfP1H8UXuJIrJUSESu4iPjWKTr+Xn5yclJQ0YMICIqqur5eXlX6cSoEOJjo4eMmRI3759zc3N586dO3Xq1A0bNoSGhiYnJ4vOPmuDS0hIYLvo6emx5/D0z/dF1DdfRUWlwXUe28CePXvYjTcRpaWlGRgYSL/v9u3bnZ2dJ0+enJ2dHRoayrp4dASi1k8jI6P6/bU5zh0RxcbGJiQkmJmZOTs7s0aKy5cvs9bS9tJYwC1ype3YseP48eNEZGNj4+7u/vHHH7PVPxpc9YL9BSsoKGBt0+KrjsrIyKxdu3bevHna2tqizowNzt4ifUopE7OCV1VVJSUlaWtri+a+ZNhbJSUl0TQRRNSnTx8200WDB20G/AXvfG7cuPHXX389efJEX18fYzEAoC0tWp/7+Za8AfoKWhqyfyYLiEhbU25Qf66ufDW1f997CKqEcYmC6IQG/gH7NrgwNbs6KaMq90WtipLMzPH/6Y43xVJ5gL5CUkbVxsAXE8y7vCyru5ckkJWlKZbKRKTE5y2cqrYjpOj8rTIFed6cSQ135euuLjt9nMrRiy9v/lmh2kXmowkqUsbGDDPhX/+jwntfwa34in0ni6WJ7eafFZ9veT7uXaVemrIJKQIi0uom1SRTAG1mypQpAwYMSEpK8vX1nThxYklJyb1792RlZadMmSLqfiIQCOLi4qKjo0V7sZuuI0eOlJaWWlhYzJkzZ9iwYdevX/fy8rp58+a+ffuafdAWLNq3336bmpqalJSUm5uroqIyc+ZMdXV1Q0PDlJSUsWPHjh49mj0ScHJykpgYvrFiin7N168QDpMnT37nnXcePnw4atSoDz74oKKiQiAQnDx5sm0qAdrAwoULjx8/zp7w37p1i/WXZ4tbsdtOTU1NLS2tvLw80Rx54mtisPHa7TXMpKXEx8dfvnzZ1tZ2165dHWqNraioqJycnF69eq1fvz42NpY1NLD5VmJjYznOHRMYGPjtt98uWLBATU2N2nUWW6axgFvkSouLi4uLi6N/Vmi5cOGC+O4S2CYNDQ11dfXi4mLxZVU+/PDDlStXvnz5ct68eZmZmevXr7eysmItziwG0d9SjpRNTcyYmpoSkby8PGvIFu+TKHpbVlY2Y8YM0TwJfD6/sb6WzdMJ1gkCCWfOnDl+/Hj//v3XrFnDPW8cAEDLWv9pd/OBinmFtdf+qJCRoWljlcP9e3NPI+Jkq/qZo7qcLK3cnR+bUGk3Url+mnX/1/1pfs2DJ1VDB/DP7ezd879tEHwF3vX9ukucuj7Nr/ENfBF4tkRQLRRvf1ns2JX9LXx/tHI3tUbbLz52+LtLy6yJqipKMlLGxuz+Wus9U8XH2dVX4yrWuv07ZxNHbD27y2l2lT168eWaH15cvl0+fkSXXzY2YQQ7QBvg8/mRkZFLlix5+vTphg0bAgMDBQIBW5DOycnps88+k5OTW7FiRUxMjHg/2S+//NLc3Dw3N3fXrl2///47Ee3Zs+e99957/Pjx1atXfXx8mn3QFrR+/fqnT58+ePBg6NChYWFhPXv2VFRUvHHjhpubm0AgCA4O7t69u4+Pj/gCQxIkislRIRxYYZctW9atW7fg4OArV66w1Q/bphKgDQQGBubl5d27d+/3339nA+efPXvGphoNCwtLSUmRk5O7fv36Tz/9dOjQISI6duyYaOKPN8miRYsmTJggMbeoOEdHx+h/tNnIGoFAsGjRIoFAMGjQoOTk5Bs3bsTExGRnZ7O5JjjOHXP48OHq6urevXuzXh6nTp1qm7Ab01jAbX+lRUREpKamsh6OX3/99c6dO0WbRJ3v+Hz+sGHDxPsBsF4/c+fO9ff3nzNnDkfKpiZmVqxYERwcfOvWLW1t7dLS0tDQUPGt586dS0pKUlZW9vT0tLCwmDhx4ooVK5YsWdKybSu8120obaF7+4iGhu0AwJuNzdXSSR/XsJbNuph2m3buzdBjYmpBce3JLdoO41ReJ5+hczPiUwRnvu09dXSj7SPQxmTMk6lzfsHZt7szRg5EpKmpWVBQcOrUKQcHh/aOBTrZt0nKZ5YPHz40MTHZuXPnsmXLPD09x48fb2xs3L179xcvXsTExHh7e4vuz3v27Llu3bqJEyf27NkzIyMjODh4y5YtbAACW71l+vTpp0+fNjQ0ZONQ9PT0ODoLiEhZn1IWh63JUn9MkJ+f3zfffCNKxtYJ+u6770Sjy9mOX3/99bZt28QzDA8PnzRp0rfffrt8+XISW/NFJC4uTjRvLgfuYkr/gNnMzGzlypXW1tZaWlovXry4e/fuli1bIiIiuM8dc+LECTYrxffff//5559LecSWCl78SiMijoBb8EpbuHBhYGDgKy9FY2Pj/fv3v/vuu48fPz5w4ACbl6BHjx6FhYX+/v6zZ8+uqam5dOmSkpKSg4PDoUOHFi5caGhoGBQUNHToUD6ff+jQIVdX18ZSElGTErMyrlixYurUqSNGjEhMTPTw8IiMjJQovpaWlre3t729va6uLlu1LTAwUMqlgqT90qFtBQDaC9pW4PXbVk5fK735Z+X24EJjA4WEEAPOhUqgTaFtBdoF2lY6lM71bZLmjldfXz8pKYnP57u7u+/evbsNoqqvZdtWOqyWaltpF68ffEe40kBEyi8dfoQCAEAn9sPJ4p0hhUOM+Ed9e6FhBQAAWo+/v39CQgKfzy8qKjpx4kR7hwNvLFxpnVRrzmVrYUFRUUREAgHdukULFpAUndwAAODt8fz3fq+Zw/nvdFokEgB4M+Tn57d3CPDGMjIyKi8vj4yM9PLyEq3AAtDicKV1Uq2/TtCGDXT3Lh09Sps3UyOLGQMAAAAAAHRk9vb27R0CvBVwpXVSrd9/+vFjOnWKHjygwYNb/VgAAAAAAAAAAG2r9dtWZGRIT4/69qX/LjENAAAAAAAAAPAGaP0xQQEBRET5+eTj0+rHAgAAAAAAAABoW63ftuLjQydOUHIyVVa2+rEAAAAAAAAAANpW67etpKXRX3+1+lEAAAAAAAAAANpDa7atREcTj9eK+QMAAAAAAAAAtLfW77ciHWtr6/YOAQAAAAAAAACgyVp/nSAAAAAAAAAAgDdXR+m3IhQK2zsEAAAAAAAAAIAm471uo0ZLzaiCthUA6FR4mE8KQAqd8dkJvt0ALauz/B3oLN99KeuzsxSnMdzF7OCl69TBQ31SfukwJggAAAAAAAAAoPk6ypggAIDOpbM8hQOApsK3GwAAAJoKY4IAAAAAAAAAAJrvtfutoE0EAAAAAAAAAN5imG8FAAAAAAAAAKD50LYCAAAAAAAAANB8aFsBAAAAAAAAAGg+tK0AAAAAAAAAADQf1mAGgDbCa6llxbBCKgAAAAAAdCTotwIAAAAAAAAA0HxoWwEAAAAAAAAAaD60rQAAAAAAAAAANB/aVgAAAAAAAAAAmg9tKwAAAAAAAAAAzYe2FQAAAAAAAACA5mu7NZhbcPlVAHgzWFhYREVFubi4JCYmir8QCoUlJSV//PHHqlWroqOj2ztMAAAAAAAALui3AgAdzrZt2zw8PPT09MLDw3v06NHe4QAAAAAAAHBpu34rzNWrV9v4iADQQVhbW0uZ8sGDBz///HNaWtq1a9fmzJnz3XfftWpgzYO+eABSEgqF7R1C0+DbDdDiOt3fAQCApmrrthUAACk9fPiQiAwMDNo7EAAAAHjzdZZ2VSkbqjpLcRrDXcwOXrpXnqMOHj/UJ833Dm0rANBBmZiYEFFGRkZ7B8KlLsaovUMA6LhkzJPbO4Tmw2N2gBaBe0gAeEtgvhUA6HBMTU3nz58fEBBQXFwcEhLS3uEAAAAAAABwQb8VAOhwli9fXlpa+scffyxYsCA3N7e9wwEAAAAAAOCCthUAaDfR0dGirsL1XwAAAAAAAHQKGBMEAAAAAAAAANB8aFsBAAAAAAAAAGg+tK0AAAAAAAAAADQf2lYAAAAAAAAAAJoPbSsAAAAAAAAAAM2HthUAAAAAAAAAgObDGswtKT8//8cff4yLiysuLu7SpUvv3r3t7OymTZvW3nG1AI6iTZs2raSkZP369aNHj26RY7V4hi1Iytg6chEAAAAAAACgZaFtpSX5+PgkJCSoq6sPHTq0tLQ0IyPj7t27HbltpaamRk5Oqmug0xWtfRkbG5eWlqqpqbVUhtKfqY7s6tWr9T+0trYmIqFQ2ObhAAAAAAAAtIxOf7fWcVRVVd2/f5+Itm/f3q9fPyKqqanJyclhW9kN5M6dO4cMGfLo0aNPP/2UiC5cuKCgoDB16tSysjI3N7fIyMjHjx8PHjx45cqVwcHBFy5cUFFRcXV1nThxIstEypRbtmxhHUxqa2t79+49ffp0BwcHlgPrT/Hxxx9HRUU9evRIVVW1sLDQwcHhiy++IKKYmJhvvvmmW7dux44dk5GRkbJoTE5Oztdffx0fH9+7d+/ly5cPHDhQmkhcXV3j4uKSkpJ0dHS++OILtpe44uLiQ4cOxcTEFBQU6OjoODk52dnZEdHGjRsvXrz4ysgbyyEyMnL16tWqqqqBgYEaGhoeHh7379+fMWPGkiVLpI+No3SPHj0qKSkpKSkRr/O7d+8mJiZqaWmJ6qex0tU/U5s2bRo+fLiUlyIAAAAAAAC0Jcy30mIUFBQ0NTWJyNfX96effrp582ZlZaWurq54Gh6P19juBw4cUFFRkZeXv3PnjouLS3R0tIGBQV5e3tatWwsKCpqUMjU1dejQoXPmzLGzs8vKyvruu+/u3r0rnkNAQICCgsK4ceNmz55NRJcuXaqqqiKia9euEdH48eMlmiekKdq+fftqa2vV1NTS0tI2bdrEPnxlJEePHh06dOh7772XnJy8YsWKoqIi8a01NTUeHh6nTp3S0NCYPXt2aWmpn5/f2bNniYi1Yrwy8sZyGDNmzPTp01++fOnn5xccHHz//n1jY2PW4CVlbNKUTtyhQ4f69OljYGAgqh+O0tU/U0pKSo3lDNBx9JiYKmOefPpaaYc6dPbzmoU+ubpTnvAtU3pMTLVYlLn3RDHbNGh2uox5svvW56LEfac9kTFPXrUnXzyHGSufyZgnS6QU7c7+UxmXMmh2+o6QIvTBgg7u/v37I0eOVFZW5vF4okcCTWVnZ8fj8by8vFo2Ngmampo8Hu/06dOtehQObVNMAAB4M6BtpSWtXbu2X79+qampR44c8fLymj59elBQkHgCjrYVR0fHLVu2sF85VVVVe/fu9ff3l5eXr6mpSU1NbVJKf3//VatWzZ8///PPPzc3NyeiO3fuiOdgZ2e3fft2Ly8vJycnIyOj0tLSyMjIurq6mzdvsq3NKNoHH3ywffv2NWvWEFF2dnZpaak0kcydO9fZ2XnNmjW9evUqLy+XGDMSFRWVkZHB5/O9vb0dHBwWLlxIRMHBwURkamoqTeQcOSxevNjIyOj27dsHDhzo0qXL6tWrJQbdcMcmTenEzZ49293dnfWyYfXDEZuI6EzV7zUD8Eaqrmn5lolZnjk/h5VU1Qit3lXqryufmFZ15Xa59Ls/L6w9e6OMvQ6+UFJZJRlhH235KZbKRnoK91Orvtr5fMvhwhYLHaAV7Nq1Kzo62tDQ0N/f383Nre0DqK6uljLl8OHDzc3Nu3Xr1qrxNEj6IOEtlJCQIBQKd+/eTUQ6OjqHDh3Kzs6uqqrKz8+PiYlZvHhxewfYNOLFYdLS0oRC4ebNm4moqKhI2JCjR4+yHYVC4fLly0X7XrhwgX24bdu2xvJvYwMHDjx8+HBmZqZAIHj27FlYWNi4cePYpsZO39WrV4VC4bFjx8TzuXHjhlAolPit3qokqq7DXmwmJibspLOH8RzCw8OFQuGGDRukyVb6xPn5+UKhsLGnBdxbWxbaVlqSqalpQEBAcHCwt7f3yJEja2pqAgICsrKyJJLV1tbW39fY2JiI1NXViUhbW1tdXV1OTo71VigvL5c+pVAoPHjwoJOTk62trb29fVRUFBEVFv7nt/6wYcNEr9l1FhYWdu/evZKSkgEDBvTt27cZRRs8eDARde3alb2tqKiQJhJtbW0i4vF4vXr1IqK8vDzxreytQCCYNWuWg4PDli1biCgnJ4f1VZEmco4c5OXlP/roIyISCoW2tra9e/eW2Jc7NmlKJ65///5EpKysLKof7tIx4mcKoBMRCsltY56MebLpR+lZeTX5RbUe254bzUhTGZcydG7GwbMlopSsv8nGgy9Gf5LJt0yJuFsh+mTCkmx168eDZqffiq9kiTnyaUxllfBWfAURXf5e58IunegDenkX+vl+1l36shw+X1JdI+yvK6+hKlP4su5UhGTXmCmWyr9t7/1HkL7N8C5EVD8BQMdhZWX1ww8/EFF8fLyHh0dsbGx+fr67u7uhoaGysvKQIUMOHjwoSuzq6mpgYNClSxc+n//OO+/s2bOHfW5mZnbhwgUi8vX15fF4ZmZmRMTj8Xg8HutGeufOHfa2srKS/ul+snHjRktLSwUFhYiICI6Dirtz505MTMyLFy/Y265du/J4PD8/PwsLCyUlpUmTJmVnZ7u7u6upqenr6//8888sGTucr6+vtbW1mprasGHDbt26xTY9f/78s88+69+/v7Kysqmpqa+vr0AgEN9LPMgGi9lYnYh2Hz9+vJqampmZmeighYWFy5cvNzY2VlJS6tWr16pVq9jnUlYCdHC//PLLggUL2DXz+PFjExMTW1vb9g6qJd2+fTsmJiYmJoZ9nXNyctjb5ORkURp3d3dZWVkiGjhwoGgqgw7Czs4uLi5u3rx5Ghoat2/fzszMHDt2rJOTE9va2Ok7cOAAEb3//vsaGhosZf/+/S0tLUWb2kWbXWzTpk1LSEjo2bNna2T+ZkPbSksKDw+vqKjQ1ta2sbFZunQp+/Dly5dEpKioKHot0Q+FYZ0mWMcW7llLuVNev349JCSkvLzc09PT399/6NChVG+iUD6fL3pta2urqqr6xx9/hIaGUiOdVriLVj8q6SNJSUkhourq6vT0dCLq0aOH+FYtLS0iUlRU3LJly7Z/bNy4kf35liZyjhxKSkrY70sFBYWwsDA2oYz0sUlTOnFssJJ4/XCXjhE/U//P3p3H1Zz9fwB/35Z72zetWpFqkuVrK3vJUviOjDJCkvAdM8XwNYxJjGjIMijGMJMYKWMfv8ZOUVQUJiEVbpuU9v1Wur8/PubO/d7qdiUtvJ4PD4/63PM5533O53Nv93M+55wPQFfB59OiH/J+/aP0X+ac6z8baKtLj/5P9u7jJToa0t+4aRSXv56/Ie+XM6XCu/j+XCjHlnIeq6Qk/+av0obgIssebMse7IfPauetf0lEtXX8FvNpTI7N0teSIaI5vnk+PxX+caOyqoZvZsQWTnPiWvmw+VnMv9xC0b5vpgdn5njlabZKRBTyf0136GTn12fn1xORiiL+sELn5ezsbGFhQUR9+/ZdunTpwIEDR40atXv3bh0dnZUrVxYXF3t4ePzyyy9M4qSkJFtb29WrV8+bNy8tLc3Ly+vq1atE5ObmZmxsTERDhw5dunSpm5ubIH8x43PXrFkjJyfn7OzMZrPFFNoiX19fNTU1Dodz6dIlKyurP//809LSMisra+HChbm5uYJkW7dutbW1nThx4t27dx0cHPLz83k83siRI/fu3SstLe3q6pqfn79mzRpm0GjjIJWUlJqsZnNtwvDz87O0tLS0tHz48KG7uzsR1dbWjh49evv27QUFBTNnzhw5cuSjR4+Y7e/SCNBJyMnJDR8+nIjGjh07YcIEa2trLS2t7777TpBAS0vrp59+evr0aWVl5aNHj3x8fARf7Zh76d99992VK1fKysqSk5OZrDqb8ePH29jY2NjYMN+Hf//9d+ZXX19fJkFRUZGRkdH06dOJaPny5UQksppBB+JwOCEhIRwO58GDB2ZmZiNHjhw6dKi+vj7TFSvm8J04caKsrIzD4cycOZPJinlHZ2RkCL/l25P4k61tzzR1dfU+ffrIysqKSdO7d+/IyMjKysq7d+8y4/cFgoODMzIyqqqqeDze48ePv/rqK2Z7cnLyxIkTicjHx4fP5ycnJ7dVYkavXr2uXbtWVVWVmJjYXB01NTWDgoLS09MrKyv/+usvkT8B7w5r2balgICAnTt3GhgYqKurM1fm3bp1YxZ/NTMzS0pKCg4OTk5OPnv27PuLQTAopq6uLjU1lfkTLgaHw3FwcDh+/Hh8fLyMjMy4ceOaTCamau8SybFjx3Jzc7OysoqLi+Xl5ZkVfwVsbGwMDAyys7NDQ0MHDx5cVVWVnp4uJSVlY2MjYeRicti8efOrV68cHR379++/efNmPz+/X3/9VVlZWcLY3rad3yq2t80KoFPx2VuYwq0d0V8+4sfuqkpSp6MqUri18hzWkQ16inIsQx2ZBRvzNh8qXuikKtjFY4rKr2v+5/bIKjf17xd1S3hcM3ReVnp2XUl5Q2RiVYv5NOnYJr0vNuUnpfMePOUREVuW5eup4ePxzyyDvKLXeUVNDCckotsPa5Kf1hLRjPFKuQWvD/xf2dU7VRkv6411//nruedEyZ4Tb9ZjkpKi5bPU3669ANqRl5dXXFxcSkrK8OHDd+7cefr06ZSUFHl5+bCwMEVFRUNDQ09Pz02bNi1cuJCIoqOj5eTkGhoaqqurX7x4ERERcfnyZXt7+1WrVkVGRmZkZIwfP15ktLbIqmfCPDw8goODiUh8oS1atmxZQECAj4/PDz/8UFNTc/v2bRUVFRUVFR6Pl5SUxIw5JaLVq1evWrWKz+f37NmTy+UeO3ZMX18/NTVVRkYmOjpaR0dn2rRpU6ZMOXr06I8//ijYSxAkEQ0bNqxxNZtrE+bVVatWrV+/PiEhYciQIenp6SUlJVFRUcnJySwW6+bNm0yvFjPb6M8//3yXRoBOoqamJicnx8DA4MiRIxEREfHx8devX09NTWVe5XA4MTExZmZmaWlp4eHhTk5OGzdutLKyYpY7ZKxdu3b//v1KSkrW1taHDh3q3bt3B1WFnJ2dBY9NELwjJBESEuLt7b1s2bKoqKjZs2cnJiYWFBQwV8UdzsbGhhl+7uvr++LFC2ZjaWnp7du3Sezhq66uPnr06KJFi9zd3ffu3ctisZje1YMHD3bUoy3FRNtWZ5qmpqaSkhLzAxEZGBgw985fvXpVWVkpnFJWVvbcuXOmpqZcLjclJSUwMFD41X79+kVFRaWnpxsYGHh6eu7evTslJeXq1auHDx9evHixsbHx7du3Y2Njmd7wNknMlOvv73/27NknT54MHDjwwoULpqamItMO2Gx2dHS0hYXFrVu3Dh8+7OnpGRISIisr24b92uhbaUvMk2WysrIyMjJUVFRGjBgxf/58ptfw66+/3rJlC5fLvXfvnoeHx/ubc2hra5uUlHTt2rV9+/YNGjRo6NChMTEx4neZOnXqiRMn+Hz+8OHDhTsXhImp2rtEMn/+/NjY2IyMDFNTU29vb8G4O4asrGxQUNBvv/0WHx9/+PBhBQUFY2NjR0dHySNvLofjx4/Hxsbq6el5e3vLy8vHxcVFRUUFBAQIf0cUH1sr2lnC2N4qE4BOKIVbS0Qe/1ZRVZIiosy8eiKq5vF7TH0uSMPNraup5cux39zith+qIJJJ394cIlJVejOMq6K6QZJ8mmRjJXf/iNHzF3XxD2vCLpRHxFT6/lw4Y5xyb8M3N2S+clYL+ubNwLQeU59nvKwX7HvgbBkRWZiw+/bifGLC11STLih5ffD/Stct/GdWkYmebJ+ebAU5Vo/usm6TVPr0/J9BMQCdWWZmJhFVV1ebmJgINnK53JqaGjab/f3334eGhubm5jY0NDAv5eXlSZJtfX19442CWyBiCmUG+YrHXP4xX/179uzJ/KCkpMTj8YSH0zLThFkslomJCZfLzczMZG6KaGpqMgPdmTk+RJSVlSW4kmzuDhOjoaFBfJsw86OZWdtEVFFRwdzq19TUZDpWiIi5FfyOjQCdh4uLy759+/r169e3b18iqq2t9fPz8/f3J6JJkyaZmZnV19ePGjUqLy/v9OnTERERM2fOXL58uWCMVUBAwLp16wYPHnznzh1TU1M1NbXGD09oHzo6Oq2bA/Ly5cvw8HB3d3dmhMiOHTuEx7J1LKZjhYiEZzAJE3P4goODFy1aZG1tbWFhoaura2JiwufzQ0JC2i14yaNtqzNt586ds2fPFvzKrHhARG5ubiLrbNra2pqamjY0NIwePTorK8vDw0N4qtSoUaNqamqkpKTk5eW7d+8+ZcqU8ePHX716NSAgwM7OztjYhvdrxQAAIABJREFU+PLly4I1wtskMfPq1q1bfX192Ww2l8vV09ObMWOGyBX35MmTLSwsqqurZ82aVVlZmZWVFRwcvHr1avStdFJz5syZM2dOky/16NFj7969gl+ZgXMM4UfDODs7Ozs7C379448/hDORMOXXX38tmLYjQiRDhr6+fo8ePZ49e9bchCASWzXhPPX19YXXfBUTCcPAwGDXrl1iMlRTU1uyZElzu0sSeXM5CGZaEtG6deuYVXjfKjYJ21lM+4ipXZNHCqBL8Jiicvh8+Reb8lWVpKbbKRlqyxCRorzUic160n/f0ubV8WX+mf1G8hzRzhEmpfD0AknyadLBiDJne6Ue3WV7dJcd3lc+IuY5ERWVvSYSN9iViKp5/N+vlBNRCrdWyvqfr2UHI8rWLugmiG3yCEVB1wxA12JoaEhEioqKJ0+eFMxI5fF4MjIyp06dCggIUFZWDg0NNTQ09PX1jYqKEtywZRIL+heYTCorK5l1x5KSkhqXJXjgnZhCJYmZ6ZtgJh+JGbL+119/zZgxgxk3zhSqr69PRAUFBfn5+dra2oK5wMLPPRR5Kp9INcW3iSC98MQoIyMjptDU1FQzMzMiqqurk5WVfcdGgM4jLi6uf//+PXr0sLa2nj179pQpUzZu3Hjs2LG0tDTB0Wc64JhJDURkaGgouOJl3iylpW/mtyopKXVU38qePXu8vLyYn7lcLjMhTkI//viju7v7pEmTcnJyjh071nn6VgRdn717925yjLmYw3f79u3k5GQrKyt3d3em+/Xq1atMb2lHaS7atjrTduzYceLECSIaO3ast7f3ggULmOldjR/WwXyCFRYWZmVlEZHwk1KlpKS+//77OXPm6OnpCUYyNtdt17aJmYrX1tampqbq6ekxzSKM2SIvL8/lcgUbTUxM5OTkmOWE3h0+wT92MTExDx48eP78uZGRUdeajdJ1Iwf4sP17tKLtYIV561/OWvPy+Ga9ySMUzYzYqZm1P4QUjbdWKK9suJ/Kk5amySMU3yrbVuczf0PeV1vyzYzY2urSf6XxiEhPU6Zvr5YXMzpxraK0ooHFon+PVGSulWpq+RfjqjJe1l+9UzWu0VgbgC5n8uTJZmZmqamp/v7+EyZMKCsru3//vrS09OTJkwVjT3g8XmJiYlxcnPCOzHXXkSNHKioqbGxsZs2aNWjQoBs3bqxZs+bmzZs///xz6wptw6pt37792bNnqampeXl5SkpKM2bMUFVVNTU1TU9PHz169MiRI5l7GC4uLo0Xs2+umoJv8022SZMmTZr0ySefPH78ePjw4Z9++ml1dTWPxzt16lT7NAK0g3nz5p04ceL58+fPnz+/devWlClTiIh5uBVz2ampqamtrZ2fny945qPwsyCY4VQdNc2krSQlJV29etXe3j4oKKhTPWMrNjb25cuXurq6GzZsuH37NtPRoKqqam5uzkwLEnP4iCgkJGT79u1z585VUVGhDl3FltFctG11piUmJiYmJtLfDye5ePFi40eyCOesrq6uqqpaWloq/DiRzz77bNWqVeXl5XPmzMnKytqwYYOtra2gx5kJQ/BZ2oaJicjS0pKIZGVlmY5splmEMVsqKyunT58uWN6Bw+E0OdaydbDk3sfu7NmzJ06c6NWr17p168QsQdcJdd3IAT54bo7Ku/6rVVfPn7E691pC1Y39Bl4uai8K6v1DikIiynh1/Bnjmp5+KAaHzWpdPhu+6GbdRy6/+PX1e9VSUjR1tOKFwO4Kci1/aIScLSUix2GKZ7Z1P721++mt3c/v0h9ozqHmV7QF6Fo4HE50dLSXl9eLFy82btwYEhLC4/GYh+i5uLh8+eWXMjIyK1eujI+PFxkfunz5cmtr67y8vKCgoEuXLhHRnj17hgwZ8vTp08jIyPXr17eu0Da0YcOGFy9ePHr0aMCAAefOndPR0ZGTk4uJiVm0aBGPxwsLC+vWrdv69esFTxdqkkg1xbdJk5jKLlu2TENDIyws7Nq1a8zTHtunEaAdhISE5Ofn379//9KlS3fu3CGi3NzcBw8eENG5c+fS09NlZGRu3Ljx66+/Hjp0iIiOHz8uWPjjQzJ//vzx48cLnpzVmLOzc9zf2m1mDY/Hmz9/Po/H69u3b1paWkxMTHx8fE5Ozty5c5kEYg4fER0+fLiurq579+7MKI/Tp0+3T9jNaS7a9j/ToqKinj17xoxw/Oabb3bu3Cl4STD4jsPhDBo0SOT+NzPwZ/bs2YGBgbNmzWrDxES0cuXKsLCwW7du6enpVVRUME87Efbnn3+mpqYqKir6+PjY2NhMmDBh5cqVXl5ebdi3wmq3jlLm6ld4QgQAALNIcFe8Y8N8pjXEd9iycwCdHzOXqsu9wZl3d5cLGxiampqFhYWnT592cnLq6FiAqEu9oSS8V/f48WMLC4udO3cuW7bMx8dn3Lhx5ubm3bp1Kyoqio+P9/X1FVyc6+jo+Pn5TZgwQUdHJzMzMywsbMuWLczsg4KCgm7duk2bNu3MmTOmpqbMgiCGhobNjRQQJmFjSlid5OTkPn36NJ4TFBAQ8O233wqSpaSkmJub79q1SzAjntnxm2++2bZtm3CGFy5cmDhx4vbt21esWCFIJpwgMTFRsG6uGOKrKfmNVSsrq1WrVtnZ2WlraxcVFd29e3fLli1RUVFEJP7wEdHJkyc/++wzIvrpp59EnkrzLsFLGL/wmSY+2rY90+bNmxcSEiI+jbm5+f79+wcOHPj06dMDBw4wCyloaWkVFxcHBga6urrW19dfuXJFXl7eycnp0KFDzBN5TE1NQ0NDBwwYwOFwDh065Onp2SaJmTquXLlyypQpQ4cOTUlJWbJkSXR0NBOqcAtoa2v7+vo6OjoaGBiUl5c/evQoJCTk4MGDLR4Lkux9h74VAOhI6FsB+IChbwXaH/pWOpsu9IaS5HLXyMgoNTWVw+F4e3u/v2dTiNe2fSudVlv1rXSId+9b6QxnGgiT5H2HOUEAAAAAAAAtCAwMTE5O5nA4JSUlJ0+e7Ohw4IOFM62LQt8KAAAAAHwgCgoK+Hw+Bq3A+9C7d++qqqpz586NHTtW8AQWgDaHM62LwnOCAAAAAAAAWuDo6NjRIcBHAWdaF4VxKwAAAAAAAAAArYe+FQAAAAAAAACA1kPfCgAAAAAAAABA66FvBQAAAAAAAACg9dC3AgAAAAAAAADQeuhbAQAAAAAAAABoPfStAAAAAAAAAAC0nkw7l2dnZ9fOJQIAAAAAAAAAvD8YtwIAAAAAAAAA0HosPp/f0TEAAHQ9LBaro0MA6Bq63DcNvLsB2lyX+BzoKu99CRuzq1SnOeKr2clr1+Ix6uTxQ2OSvO8wbgUAAAAAAAAAoPXae70VAIAPQ5e4BQcArYB3NwAAALwtzAkCAAAAAAAAAGg9zAkCAAAAAAAAAGg99K0AAAAAAAAAALQe+lYAAAAAAAAAAFoPfSsAAAAAAAAAAK2H5wQBQDthsVhtmBvW4QYAAAAAgE4C41YAAAAAAAAAAFoPfSsAAAAAAAAAAK2HvhUAAAAAAAAAgNZD3woAAAAAAAAAQOuhbwUAAAAAAAAAoPXQtwIAAAAAAAAA0Hrt+gzmtn0CKwB8AGxsbGJjYz08PFJSUoR/4PP5ZWVl9+7dW716dVxcXEeHCQAAAAAA0CyMWwGAzmjbtm1LliwxNDS8cOGClpZWR4cDAAAAAADQrHYdt8KIjIxs/0IBoMPZ2dlJnvjRo0e//fYbl8u9fv36rFmzdu3a9f4Cax0MxAOQEJ/P7+gQAAAAAN6vDuhbAQCQ0OPHj4nI2Ni4owMBgI8Iek4B2lyX6GPtKu99CRuzq1RHjC5x2gAIoG8FADovCwsLIsrMzOzoQJrVEN+7o0MA6LykrNM6OgQAAACA9oC+FQDojCwtLd3c3Hx9fUtLS8PDwzs6HAD46OB+KUCb+ABGTwAASAJ9KwDQGa1YsaKiouLevXtz587Ny8vr6HAAAAAAAACahb4VAOhIcXFxgjtajX8AAAAAAADo/PAMZgAAAAAAAACA1kPfCgAAAAAAAABA66FvBQAAAAAAAACg9dC3AgAAAAAAAADQeuhbAQAAAAAAAABoPfStAAAAAAAAAAC0HvpWAAAAAAAAAABa7yPtW/Hw8LCzs7Ozs/v9998FG7/55htm4969e9uklKlTp9rZ2cXExLRJbm1FOKqVK1fa2dkFBwc3+WqrM+ko79LgHXuwJC+9c55UAAAAAAAAHzOZjg6gg506dcrFxUVKSorL5SYkJLQ6n/r6ehkZGZEt5ubmFRUVKioq7xxmWxIflYQxd86qvYuuUqM2j7Pxqfv+REZGNrndzs6OiPh8fvuEAQAAAAAA0LY+6r4VZWXl/Pz8Gzdu2NraHjt2jIhUVFTKysoECUpKSg4ePHj79u2ioiJdXd1x48Z9/vnnsrKyRDR16tSysrIFCxbExsY+efJk06ZNGzZsENny5MmTsrIyQYbHjh07e/Zsfn6+rKysrq7u559/PmHCBJGQmkxTWlp66NCh+Pj4wsJCfX19FxcXBwcHJr0gjLt376akpGhra69YsaJPnz5iihNE5eHhweVyiSg0NDQ0NNTExCQkJEQ45i1btpw/f97R0XHlypVElJCQ8M033ygrK588eVJ8Jj/88MPly5ednJyWLl1KRPHx8d9++62Ghsbx48elpP4ZKrVly5bExMTS0tLXr19379592rRpTk5OLdZLzF4CYgI4ceKE+GaR8Eg1d1Cio6PXrl2rrKwcEhKirq6+ZMmShw8fTp8+3cvLi6mUp6dnYmJiamqqvr7+0qVLmUqJEFNH4TjFtJIk54zgRB08eHAzbxHovPq6Zjx8VktEW5Zorpitzmx0WJJzKb6KiP47W33rEs02KUhrwrPC0tentug5jVFqkwzfnXBIjktzLsZV+XhobPiiW5MJJMmEiJrMp/29S2t3wiMFAAAA8PH4SOcEMRwdHWVkZI4fP15SUnLlyhUzMzNzc3PBq3V1dd7e3n/88Ye0tLS9vX1xcXFwcPDmzZuFcwgODmaz2WPGjJGXl29uC+Px48d79+6trq6eM2fOtGnTdHR0cnJyROJpMk19ff2SJUtOnz6trq7u6upaUVEREBAQEREhvOOhQ4dMTEyMjY25XO6mTZskLG7ChAk6OjpEZGFhMX369MbdB5MnTyaiGzdu1NbWEtHly5eZvZjeJTGZMB0BV65cYXa8fv06EY0bN064Y4WInj17NmDAgFmzZjk4OGRnZ+/atevu3bvi6yXJXmICePLkSeuOgkgaMQdl1KhR06ZNKy8vDwgICAsLe/jwobm5+RdffCHY9+jRowMGDBgyZEhaWtrKlStLSkqoEUnqKKaVJDlnmjtRocvZfazkdQMR0cNntUzHSuvU1YuOG2K2DP6EY91HTkNF+h1ibGMthiRJzJ2wXu/ow6sRvCcPHz4cNmyYoqIii8VqfHNCQg4ODiwWa82aNW0bmwhNTU0Wi3XmzJn3WooY7VNN6HKSk5P5fP7u3buJSF9f/9ChQzk5ObW1tQUFBfHx8YsXL+7oAN+OcHUYXC6Xz+czVz0lJSX8phw9elSwL5/PX7FihWD3ixcvMhu3bdvWXBEAH6SPetyKhoaGvb39xYsXN2/eXFdX5+LicunSJcGrcXFx2dnZ0tLSgYGB6urqo0aNWr169bVr17788stu3d7c2HRwcGDGdAg03sKoqqoiImYchIGBgZubG4fDkSRNdHR0ZmYmh8Px9fWVk5PT1tbesmVLWFjYlClTBDu6urp6eHg8efLkiy++yMnJqaiokKQ4V1fXe/fu5eXlDR482NPTs3HMffr0MTEx4XK5sbGxNjY20dHR9HeHi/hMLC0te/funZaWFh0dbWdnd/PmTaZlRPIPDAxks9l8Pp/H4xUWFsbGxiYkJAwcOFBMvZSUlFrcS0wARUVFrTsKImliY2PFHJTFixcnJyffuXMnISFBQUFh7dq1wpNuZs+e7erqyufzZ82a9fLly8jIyGnTpr1ty4gcApFWunfvXovnTHMnKnQtGirSmS/rT16rmDFOaUdYMRF1U5UuLH0tSPCq+PX3vxRejKt6WVhvrCs7y0F5xWx1DptFfw9z2Li427mblXce8SJ2dJ+15qXIlvFDFRIe8wpLXxeVvSai7UeK950qzcqvZ8uwTPRk/ztHbe4k0elpzaUpKHnt92vR+djK3IJ6UwP2165q86aoCIcRmVB9+2GNka7MvtU6w/vJiclKEJJg8I5/SJF/SFGfnuwH4cbCCTw35IVElM3/t8qva3SI6FJ8lcOSHA0V6Zw/ewjXq3E+/zLnhJ4v/8pZLegbLSI6d6tyyrIXut2ksyJ6Sv/vLYkFG/Ou3Kl6Vfz6dQP11Jf9ykX1K2c18fVqbhdhc79/KSaAJltGuEYSHiz4OAUFBcXFxfXr12/BggW9evVq/wDq6uqEb9KIMXjw4JKSEg0NjfcdUmOSBwkfud9//33EiBEFBQVRUVGqqqoWFhb29vZttXRjZ3Dnzh1lZWUi6t+/v5yc3MuXLzMyMogoLS1NOJm3t/eOHTtev37dp0+fxvdrAT4SH/W4FSKaMWMGEcXHx2tqatra2gq/lJ+fT0Sqqqrq6upE1KNHD2b7q1evBGkGDRokkmHjLYx//etfEydOLCgo2LRp01dfffXpp58eP35ckjRMGDweb+bMmU5OTlu2bCGily9fMiMyGMx3I0VFRebX6upqSYqTBNOTcvny5ZiYmOrqaktLS0E7iMfcCjt37tz9+/fLysrMzMxEduTz+QcPHnRxcbG3t3d0dIyNjSWi4uJi4TSN6yXJXmICaPVREEkj/qDIysp+/vnnTB3t7e27d+8uvK+enh4RsVgsXV1dQVZv2zLiW0mSc6a5ExW6Fo9/q7BlWTvDi/OLXx+5WD7IgjP4k3+6Anm1/FGLsveeLJWWItcJyvnFr31/LvTwyxPOwffnQjm2lPNYJSV5qea2MOIf1nwTWFBR3bB6noaXi5qxnszT7DqReJpLU1vHH/2f7N3HS3Q0pL9x0ygufz1/Q94vZ0oFO24ILrLswbbswX74rHbe+pcSFjfHUcVYV4aIhvaRW/q52hxH0b6DBU6qRHQysqKmlk9ER86XEZGbozLTuyQmny+d1YjoyMUyZsdT1yqIaLaDinSjv5lJ6bW2AxW+ddeYN1klLbPWe+urq3f+GT3UuF4t7sIQE8C7HAgAW1vbffv2EVFSUtKSJUtu375dUFDg7e1tamqqqKjYv3//gwcPChJ7enoaGxsrKChwOJxPPvlkz549zHYrK6uLFy8Skb+/P4vFsrKyIiIWi8VisZiBogkJCcyvNTU19Pfwkx9++GHEiBFsNjsqKkpMocISEhLi4+OZ+yJEpKamxmKxAgICbGxs5OXlJ06cmJOT4+3traKiYmRk9NtvvzHJmOL8/f3t7OxUVFQGDRp069Yt5qVXr159+eWXvXr1UlRUtLS09Pf35/F4wnsJB9lkNZtrE8Hu48aNU1FRsbKyEhRaXFy8YsUKc3NzeXl5XV3d1atXM9slbATozOTk5IYPH05EY8eOnTBhgrW1tZaW1nfffSdIoKWl9dNPPz19+rSysvLRo0c+Pj6CO3YFBQV8Pv+77767cuVKWVlZcnIyk1VnM378eBsbGxsbG6ZL5ffff2d+9fX1FaQpKioyMjKaPn06ES1fvpyICgsLOypggA70UY9bIaKePXsOHDjw7t27n332mciKntra2kRUWlpaUlKipqbGrCpCRFpaWoI0jUc0NN7CYLFY33777bJlyzIyMuLi4kJCQpjr5xbTfPvtt0QkJyfn5+cnmFNTV1cnLf3PwG9mO4vFeqviBDuKWUN0woQJ+/fvj4+PLy0tpUaDVsRkYm9v//PPP9+7d4+57dN40MqNGzfCw8MVFBR8fHy0tbUPHDhw//59kUwa10uSvcQE0OqjIJKGOTeaOyhlZWXMN1c2m33u3LmJEycKL6qSnp5ua2tbV1fH/IkSPp3eto7NtZL48BjNnajQteh0k3adoHzoz7L5fnm8Wv7Xruqh5/9ZMercrcrUzFoZadaN/YY6GtJOtpX/Xv7i6OXy7V9r6mm++bjzmPJmTIdA4y2M8soGIuLIskz0ZMyM2D7zNRTkWBKm+fNmZQq3Vp7DOrJBT1GOZagjs2Bj3uZDxQudVJkdV7mpf7+oW8LjmqHzstKz60rKGyQpbtVc9ajEqoyX9eOHKjS5TsqwvnJ9erIfPquNiKmcNFzxdFQlEXlOVZUkn4HmnLtPeKciKz4fr/zHjUoicp/cxLiPG/sN5Nishgaq5jW8KKiPiKm8crvKfohCc/VSU5YSvwvDxkquuQDe5UAAODs75+XlpaSk9O3bd+zYsQMHDhw1alRKSsrw4cPd3NyCg4M9PDzq6uoWLlxIRElJSba2tqamptnZ2cHBwV5eXsw9eTc3t71792ZkZAwdOnTYsGHMPQOG8B8jEWvWrLGzs3N2dmaz2WIKbZGvr+/YsWM5HM6lS5esrKzU1dUtLS3j4+MXLlw4fvx4QTBbt25dtmyZpqbmiRMnHBwc0tPTVVVVR44cmZqa2rt3b1dX1zNnzqxZsyY5OTk8PLxxkEpKSk1Ws7k2YXb38/NbtGhRRUVFfHy8u7t7WlpabW3t6NGjk5OTNTQ0Zs6cWV5e/ujRIyKqra19l0aATqKmpiYnJ8fAwODIkSMRERHx8fHXr19PTU1lXuVwODExMWZmZmlpaeHh4U5OThs3brSysnJ1dRXksHbt2v379yspKVlbWx86dKh3794dVBVydnYWLMAn/KaWREhIiLe397Jly6KiombPnp2YmFhQUDBx4sT3ECZAp/bOfSvN/xFt7M3VoZ1d45eimnmASDtYtWpVVlaWpaWlyHZra2t9ff2cnJwlS5b07duXmVdia2srmBD0Vh4+fLhjx44BAwZoaGg8e/aMiNTURAeBN5nGxsbGwMAgOzs7NDR08ODBVVVV6enpUlJSNjY271gcETFDJy5fvsyMSRF8ORBQUVEZNWrUtWvXkpOTFRQU7Jo6dk1mwuFwHBwcjh8/Hh8fLyMjM27cOJG9Xr9+M3C9rq4uNTWV+arRIsn3ajKAVh8FkTTiD8rmzZtfvXrl6OjYv3//zZs3+/n5/frrr8xwSiI6duxYbm5uVlZWcXGxvLx84yZtXctIHh58YJbNUjv0Z9m5W5X6WjIzxikJ961k5tUTkaaalI6GNBFZ9XrToZaVVy/oW7EfqiCSYeMtDLvBCu6TVY5cKJ+3Po+I5NisjYu7LZ+lLkkaJpJqHr/H1OeCxNzcOmZQBhH17c0hIlWlN91/FdUNkhQnCc+pqst3vDpyvqyunl9R3WBjJWfViy3Jjoud1Rb654X8X5mOhnRh6etBFpzGOzY00PpfCkPPl+cW1jc0vNmYV/TPnKzG9VJRlBK/S4sBvMuBkKTi8GHz8vKKi4tjLul37tx5+vTplJQUeXn5sLAwRUVFQ0NDT0/PTZs2MVf40dHRcnJyDQ0N1dXVL168iIiIuHz5sr29/apVqyIjIzMyMsaPH79x40bh/EUWVhPm4eERHBxMROILbdGyZcsCAgJ8fHx++OGHmpqa27dvq6ioqKio8Hi8pKQkwTXh6tWrV61axefze/bsyeVyjx07pq+vn5qaKiMjEx0draOjM23atClTphw9evTHH38U7CUIkoiGDRvWuJrNtQnz6qpVq9avX5+QkDBkyJD09PSSkpKoqKjk5GQWi3Xz5k0LCwsiqqurI6I///zzXRoBOg8XF5d9+/b169evb9++RFRbW+vn5+fv709EkyZNMjMzq6+vHzVqVF5e3unTpyMiImbOnLl8+fLc3Fxm94CAgHXr1g0ePPjOnTumpqZqampNrsTXDnR0dJglFFvh5cuX4eHh7u7uISEhHA5nx44dbm5ubRseQJfwsc8JIiJtbe1BgwY1XtGTzWYHBQVNmTKlrq7u6tWrKioqHh4egpGcb0tdXV1VVfXatWshISF3794dNGjQ999/L0kaWVnZoKCgadOmFRQUHD58+Pz587W1tU32cbxtcUTk4uLyySefFBcXnzp16s6dO01mJVikY+zYsU2ue9pcJlOnTmXuXw0fPlzQsyBga2s7depUaWnpffv2PX78eOjQoeJr1Iq9GgfQ6qMgkkbMQTl+/HhsbKyenp63t/fEiRNtbW3z8/MDAgIE+86fP7+goCAjI8PU1HTz5s3MjLN3bxkJw4MPTz9TDjPkwWuGmqzM/3R2G2rLEFFBSUN+8WsievjszdB3A51/etXlOaL94423MKRYFLJWpzSy151DRt8v6lZTy1//S5GEaZhIFOWlzu/SvxT05t/Z7d1l/h5Kxcy1Ee6rl6Q4IpKWYhGRoJOiMWYG0PnYqp9OlFJTg1aay2fWRGV1ZalrCVU/hpUQkfuUJgatnIqqCPituKyy4fB63Rv7DWwHyhOR8CCzxvVqcZcWA3iXAwEgIjMzk4iqq6tNTEy0tLSYddO4XG5NTU1DQ8P3339vYGAgKyurpKTELIiel5fXQo5ERFRfX994o+Aui5hCJcmcubWuqalJRD179tTU1GSz2UpKSkRUXl4uSMbMRGaxWCYmJkyhTLmamprMBSQzx4eIsrKyGgfZpBbbpF+/fkSkqvrmc6aiooIZo6qpqcl0rBARM5z2HRsBOo+4uLj+/fv37NnT1dU1IiKCzWZv3LiRGX5iZGRERAUFBcxJkpyczOxiaGgo2D0pKYmImOHhRMScyR1iz549rL8x5+1b+fHHH4lo0qRJOTk5zNNXAT5CH+mcoJCQkCa3MytTCKirq//3v/9tMuUff/zxVlv09fWZDx0xmkujpqa2ZMmSFsPQ19ePFBr+01xxIrv89NNPzb3K+Ne//hXZaFSR+EwE23v06PHs2bPGE4KISEpK6uuvv/76668lCVI4AAn3ajIAMUdBeN9rkK8kAAAgAElEQVQWjxQ1f1BcXFyEJxCtW7du3bp1wgkMDAx27dolpnTJW0ZMK0l4zsCHIXiNTmpmrU1fOZHtk0YomhrIpmfXjflP9sj+csy8Ehd7pe6arfnkv5VU/dWWV2MGyutqSien84hIW0P0kTTNpZk8QtHMiJ2aWftDSNF4a4Xyyob7qTxpaZo8QvFdiiMiYz0ZIjpysayiusHaSm7WRNFu3G6q0tPGKB29XH7zr2plBanPxzf9tbVxPvIc1rwpKjvCS87fqmTLshrnTET1r990ivBq+YkpvLjkli+KJN+luQDe5UAAiGCu8RQVFU+ePCmYN8rj8WRkZE6dOhUQEKCsrBwaGmpoaOjr6xsVFSWYoMokbhDqj1RUVKysrGRWB2MuF0UIbs+IKVSSmJm+CebeiZjlZv/6668ZM2bweLzHjx8zherr6xNRQUFBfn6+trb2w4cPmZQGBgaNg2yymuLbRJBeeGKU4Oo6NTXVzMyM/l4l9x0bATqPefPmnThx4vnz58+fP7916xZzS5JZgJnpttPU1NTW1s7PzxfMEM/OzhbszoxWFjP1u6tISkq6evWqvb19UFAQMzgL4COET3B4L2JiYh48ePD8+XMjI6MOmYrS4QEAtBsjXRkj3SY+zOXYrOhfDNftK7wUXxV2sdxIV9ZrhtpKt1ZODNHpJqOpJn30cnlxWYOyAmvcUIXNXpoSpuGwWTf2G2wMLjofW+kfUqSsIGXZgz1vorjH1khSHBEtm6V+N4V3P40XdKykrFKlyR6QBU4qRy+XE9HMCcoiC/SKz2exs9rOoyV8Pv17pGKTzzZ2sVeOuV8dfrF81e6CcUMUHIYpnrleIaZSb7tLkwG8y4EAEDF58mQzM7PU1FR/f/8JEyaUlZXdv39fWlp68uTJgrEnPB4vMTExLi5OeEdjY2MiOnLkSEVFhY2NzaxZswYNGnTjxo01a9bcvHnz559/bl2hbVi17du3P3v2LDU1NS8vT0lJacaMGaqqqqampunp6aNHjx45ciRzp8HFxUVkyXkx1RTMeGqyTZo0adKkTz755PHjx8OHD//000+rq6t5PN6pU6fapxGgHYSEhPz000+pqan5+fn9+/cnotzc3AcPHhDRuXPn0tPTTU1Nb9y4ERMTM3XqVCI6fvz4ixcvOjjo92P+/PlmZmZi3hfCS7o8fvzYw8OjvUIDaCfoW4H34uzZs4mJib169Vq9erWYle0+4AAA3ivmYcONnd+lL/yrjob0z6u1m0z56lLPFreIbLyyR79xAmGmBrLNpdFWlw5coUUkunizcP6mBrIN8f8s49dcViK7xB4wFJOAiMYOVhDOtsk0TeZjaiDbtxcnKZ3X5IQgIpKWot3faO/+pokWFlOv5nZpMqrGATTXyCIltniwAIiIw+FER0dv2LDh/PnzGzduVFZWtrS0ZC54XFxcoqOjw8PDV65cOW7cOAcHhzNnzgh2XL58+d27d+/fvx8UFFRWVjZr1qw9e/bMnz//0aNHkZGR69evb270pfhC29CGDRsiIiIePXo0YMCAwMBAZh5QTEzM2rVrL126FBYWZmRk5O3tvXLlSjGZiFQzODhYTJs0iamsv79/REREWFiYqqrq/Pnzqb0aAd4TZqgRMzpjzZo148aNMzc3/+STT4qKiv744w9fX9+qqioiqqmpGTlypJ+f34QJE2bNmpWZmRkUFCQyRv5DIph51xzhJV0wRAs+SKx3HYTWRletHbiWLQB0LGYtmC43IJbps2t80Q4fhjPXK27+VfNjWLG5MTs53Lj5BTo/2ADahJR1GnXZd3eXCxsYmpqahYWFp0+fdnJy6uhYgKhLvaEkuRtnZGSUmprK4XC8vb13797dDlE1JmFjfgA3F7vEaQMg0DW/rAEAALxP+06V7gwv7t+bc9Rft0P6NTo8AAAAEBEYGJicnMzhcEpKSk6ePNnR4QBA5/I+h2PZ2FBsLBERj0e3btHcuSS0dBMAAECnJTK76iMMAAAARPTu3buqqio6OnrNmjWC5ygDADDe/1S3jRvp7l06epQ2b6Y5c957cQAAAADwsSooKOjoEOCD5ejo2NEhAEDn9f7HGT99SqdP06NH1K/fey8LAAAAAAAAAKB9vf++FSkpMjSkHj0oK+u9lwUAAAAAAAAA0L7e/5yg4GAiooICWr/+vZcFAAAAAAAAANC+3n/fyvr1dPIkpaVRTc17LwsAAAAAAAAAoH29/74VLpcePHjvpQAAAAAAAAAAdIT32bcSF0cs1nvMHwAAAAAAAACgo73/tWwBAAAAAAAAAD5c739OkGTs7Ow6OgQAAAAAAAAAgLeGcSsAAAAAAAAAAK3H4vP575ZBG62o8o5hAAC0LxbWkwKQzLt+02h3zLu7y4UN0Dl1oTdUV/nLLmFjdpXqiNElThsAgc4yJwgAAACg8/gALksAAACg3aBvBQCgNXAvBQAAAAAAGJgTBAAAAAAAAADQeljLFgAAAAAAAACg9d55ThDGmwAAAAAAAADARwzjVgAAAAAAAAAAWg99KwAAAAAAAAAArYe+FQAAAAAAAACA1kPfCgAAAAAAAABA66FvBQAAAAAAAACg9Zp4ThCLxWrnIPh42BDAR6ANP1vwoQEAAAAAAJ0Hxq0AAAAAAAAAALReE+NWGO1zW7j9x8gAAAAAAAAAALQhjFsBAAAAAAAAAGg99K0AAAAAAAAAALRes3OC2hNmBgEAAAAAAABAF4VxKwAAAAAAAAAArSfRuJW4uLhhw4apqKi8fPlSXl5+6dKlgYGBs2fPDg0NJaIbN26sWLHi0aNHhoaGq1evnjt3LpM+JCRk3rx5jV9tnH9kZGQbVwsAOh87OzuRLTY2NrGxsR4eHikpKcI/8Pn8srKye/furV69Oi4urkOibREG3AG0CI9LBwAAgI/EW4xbKSsrO3XqVG1tbVhYmGBjTk6Oo6OjnJzcmTNnhg4d6u7uHhUVJeGrAABN2rZt25IlSwwNDS9cuKClpdXR4QDAx4UFAG2qo9/TAADt4S3WWxkxYsTBgwfl5OQ4HE6PHj2YjcePH6+qqtq+ffuQIUNsbGzCw8MPHTr0n//8R8yrtra2bV4NAPiQPHr06LfffuNyudevX581a9auXbs6OqJmNcT37ugQADojKeu0jg4BAODtdJVuIEmGBHaVuoghvpqdvIJdOnhoTMJxuG/Rt+Lu7v7ll18WFxe7ubkdP36c2fjixQsiMjY2JiIlJaVu3brl5OQIdhH/KgCAGI8fP6a/P0AAANoZJjQBvDtcQwLAx+Mt5gQZGhqOGTMmMTFx3rx5go0GBgZExOVyiaiioqKwsNDQ0FDCVwEAxLCwsCCizMzMjg4EAAAAAABAnLd7TtDKlSuXL19ubm4u2OLi4qKoqPjf//734sWLixcvrqurc3d3l/BVAIAmWVpaurm5BQcHl5aWhoeHd3Q4AAAAAAAA4rxd38qECRO2b98uvEVPT+/ChQu1tbXTp09PTEw8fPjw6NGj6+vriYjD4TT5aluGDwAfohUrVuzZsyc3N9fBwSEvL6+jwwEAAAAAABCH1Xg6MTMxstXTjJ8/f+7j4xMeHv7gwQMrK6sWimexCM9gBvg4NH4Gc6t1hnUQmI8vrGUL0CRmLdvO8FZthXf8IgQAAl3r3dRVVofBWrbU6SvYpYOHxiT8EHu7cSuSOHbsWGRk5NatW1vsWAEAAAAAAAAA6OraftzK2xWPcSsAHw2MWwH4eGDcCgBQV3s3dZXRBBi3Qp2+gl06eGisw8atAAAAAAAAAAB8PNC3AgAAAAAAAADQeuhbAQAAAAAAAABoPZnmXuhs08CmTJlSWVnZeLudnd3atWvfX7lcLnfr1q1Pnz7l8XgjRozYuHHj+yurRZmZme7u7kR05swZVVVV4Zc8PDy4XK7wltGjR69fv751Ba1cufLOnTtz5szx9PRsbbASYcJ2cnJaunTpey2oSe1WzfY0derUsrKyDRs2jBw58t2TAQAAAAAAQIua7VvpbCwsLKqqqojo6dOntbW1GhoaOjo6RGRgYCCcrL6+XkamLSt16tSpR48e9ezZc/Lkyfr6+pLv2OaRSEJdXV1XV5f52dDQsJ1LZ3RIxVuhvr6+o0PoSObm5hUVFSoqKm2VYVc57gAAAAAAAG2uiWuh97SU9zs+Emjbtm3MD3Pnzs3KyrKzs/Py8iKiqVOnHj58eMGCBbGxsU+ePNm0adPgwYO3bNmSmJhYWlr6+vXr7t27T5s2zcnJif6+V79gwYK7d++mpKRoa2uvWLGiT58+TM7Hjh07e/Zsfn6+rKysrq7u559/fu7cub/++ouInj17FhQU5O7ubm1tXVJScvDgwdu3bxcVFenq6o4bN+7zzz+XlZUVZC4cCTPcZtGiRdHR0U+fPu3Xr9+qVavCwsIuXryopKTk6ek5YcIEIiotLT106FB8fHxhYaG+vr6Li4uDgwMTVXZ29vbt2x8/fmxkZDR9+nTxrTRmzJjGA0Cay7y5VhIMgQkNDQ0NDTUxMQkJCWGe8LJz587+/fs/efLkiy++IKKLFy+y2ewmKy6mRs2RsK0Ex9HT0zMxMTE1NVVfX3/p0qXMcWzu6AgffSZOQd+KcDWbaxPxJ095eXloaOitW7fy8/MVFRUdHR0XLlwo/rC2eICio6PXrl2rrKwcEhKirq6+ZMmShw8fTp8+XXDaN9cCwsRU58mTJ2VlZWVlZeKrJqYKTR538Ye4ybc/c2p1lScIfORyXtX7/FR45XbVq5LXKopSvQxk3SerLJ6uSkRaE54Vlr4+tUXPaYxSm5TV5hl2uQAAAAAAoAv5QNZbCQ4OZrPZY8aMkZeXJ6Jnz54NGDBg1qxZDg4O2dnZu3btunv3riDxoUOHTExMjI2NuVzupk2bmI2PHz/eu3dvdXX1nDlzpk2bpqOjk5OTM2bMGCMjIyLq2bPn9OnTLS0t6+rqvL29//jjD2lpaXt7++Li4uDg4M2bNzcXCePAgQNKSkqysrIJCQkeHh5xcXHGxsb5+flbt24tLCysr69fsmTJ6dOn1dXVXV1dKyoqAgICIiIiiKi+vv7bb7+9f/++urq6kZFRYGCg+Ha4fv36l3+7fv06k0NzmTfXShMmTGDGBFlYWEyfPl3Qo0FiZ4oJV1xMoS0S31bCKY8ePTpgwIAhQ4akpaWtXLmypKSkxaMjHOeIESMaV1P8mUNNnTz19fVLly49duxYaWnp2LFj+/bty/RMSdgIzSUbNWrUtGnTysvLAwICwsLCHj58aG5uznRpiWkBkcxbrE6LVWuxCk2e8PChmunz8rdzZbX1fNuB8r0MZFO4tdfuVHV0UADwFh4+fDhs2DBFRUUWiyXobX9bDg4OLBZrzZo1bRtbY5qamiwW68yZM++7oCa1WzUBAODD8IH0rTg4OPz4449r1qxhbrYHBgauXr3azc3tq6++sra2JqKEhARBYldXV29vb2Z8R05OTkVFBRExE46YESvDhw9ft26dh4fHtGnTzMzMiKhPnz5eXl5Dhw6Ni4vLzs6WlpYODAz85ptvVq9eTUTXrl0TXPaLRMJwdnbesmUL8yWmtrZ27969gYGBsrKy9fX1z549i42NzczM5HA4vr6+Tk5O8+bNI6KwsDAiun//fk5ODovF2rVr15o1a5gxC2IUFxc//ltRURERicm8uVZydXVlepQGDx7s5eXl6uoqyF9M34pwxcUU2iLxbSWccvbs2e7u7uvWrdPV1a2qqoqMjGzx6AjHuXHjxsbVFH/mUFMnT1xc3PPnz1ks1u7du1etWrV+/XpmmRsJG0FMssWLF/fu3fvOnTsHDhxQUFBYu3atyKSbxi0gknmL1RFfNUmq0OQJDx+kmlr+raRqIrr6k/7FIP24A4b5F3v6f9lNOA03t36id47CqPS+rhm3kmoE2xdszDOZ+lxxdLrcyHTLzzP2nHjTD6g14ZmUdZp/SNHYxdmqdk8Hz80U3ktYQcnrJdte9Z7OVRqTPmB25sGIMiKa+/1LKes0762vmDTnblVKWad1n/TsdcP/7Ou5IU/KOm3Bxjzm10vxVVLWaZrjn/Fq+WJiEyFlnSZlnXb9bjURJTyuYX6tqeU3FxsRbT9SbDadKz8qXdXuaf9Zmb+dK3uL5gZ4P4KCguLi4kxNTQMDAxctWtT+AdTV1UmeePDgwdbW1hoaGu8vnua8VZzwsUlOTubz+bt37yYifX39Q4cO5eTk1NbWFhQUxMfHL168uKMDfDvC1WFwuVw+n7958+aSkhJ+U44ePSrYkc/nr1ixQrDvxYsXmY2COQeN829nffr0OXz4cFZWFo/Hy83NPXfu3JgxY0jssYuMjOTz+cePHxfOJyYmhs/nS3hF0yZEmq7TnmwWFhbMQdfU1BST7MKFC3w+X8LVS98qcUFBAZ/Pb/KGgZiX3ocPpG9l0KBBgp/5fP7BgwddXFzs7e0dHR1jY2OJqLi4WJCgV69eRKSoqMj8Wl1dTUT/+te/Jk6cWFBQsGnTpq+++urTTz8VeTsx8vPziUhVVVVdXZ2IevTowWx/9epV40gEzM3Nmb2ISE9PT1VVVUZGhrnPX1VVxeTJ4/Fmzpzp5OS0ZcsWInr58mVtbS2Traqqqra2NhExHT1iODk5Rf5t2rRpgoAbZ87j8cS3khivX79uvFG44mJq1GLm4ttKOKWenh4RsVgsZomZ/Pz8Fo8ONXOAGC2eOdTUyZOXl8cUyvTUEBHTAyJhI4hJJisr+/nnnzOB2dvbd+/eXSTgxi3wttURXzVJqiCmPeEDI8dm6WvJENEc3zyfnwr/uFFZVcM3M2ILp1kZWPC6gbqpSj98Vjtv/UvB9qT0WtuBCt+6a8ybrJKWWeu99dVVoQEv20KLxwxSmGijcPcJz3FpTn6x6CdMbR1/9H+ydx8v0dGQ/sZNo7j89fwNeb+cKf3SWY2IjlwsY/o4Tl2rIKLZDirS//uXbYGTKhGdjKxgkh05X0ZEbo7KHDarxdhENO5ebi62+Ic13wQWVFQ3rJ6n4eWiZqwn8zQbl2rQwWxtbfft20dESUlJS5YsuX37NhEVFBR4e3ubmpoqKir279//4MGDTGJPT09jY2MFBQUOh/PJJ5/s2bOH2W5lZXXx4kUi8vf3Z7FYVlZWRMRisVgsFjNgNiEhgfm1pqaG/h578sMPP4wYMYLNZkdFRYkpVERCQkJ8fDxzr0hNTY3FYgUEBNjY2MjLy0+cODEnJ8fb21tFRcXIyOi3335jdmGK8/f3t7OzU1FRGTRo0K1btwQZvnr16ssvv+zVq5eioqKlpaW/vz+PxxPeURAnm81uXM3m2kR493HjxqmoqFhZWQnKLS4uXrFihbm5uby8vK6uLnPjR8IWgM7v999/nzt3LnNuP3361MLCwt7evqODajN37tyJj4+Pj49n3s4vX75kfk1LSxNO5u3tLS0tTUR9+vQRHvDeGTg4OCQmJs6ZM0ddXf3OnTtZWVmjR492cXEhscfuwIEDRPTvf/+buaYgol69eo0YMULwUodon5Nt6tSpycnJzOh+eFsfyNqTHA5H8PONGzfCw8MVFBR8fHy0tbUPHDhw//594dUcpKSkqNEQDBaL9e233y5btiwjIyMuLi4kJIS5LhUpiOnjKC0tLSkpUVNTEzyaR0tLq3EkAszFNlNi4/U+mTzl5OT8/PyY2Iiorq5OWlqayba8vLyyslJRUTE3N/etmkVM5jdv3hTTSkxK4UaTk5OrqakpLy8nIpHxI40rLqZGLQYsvq2Epaen29ra1tXVZWRkEJGWlhbTXGKOjkicItVs8cyhpk4ewSmRnZ3NrKzMrOoqYSOISVZWVsZ8CWaz2efOnZs4caLI2JDGLSD8qiTVkaRq4qvQ5AkPH6pjm/S+2JSflM578JRHRGxZlq+nho/HP7eUv/hMNXCF1s2/qkctyk7Prispb1BTliKiG/sN5Nishgaq5jW8KKiPiKm8crvKfogCs9e37hqr5qrz+dRrGpebW3fsSrmXi5pwuX/erEzh1spzWEc26CnKsQx1ZBZszNt8qPjpaZOB5py7T3inIis+H6/8x41KInKfLLpC87C+cn16sh8+q42IqZw0XPF0VCUReU5986g18bGJkGp0P6K52Pat1iYijizLRE/GzIjtM19DQa5zPX0PPkLOzs55eXkpKSl9+/YdO3asjY1NbW3tqFGjUlJShg8f7ubmFhwc7OHhUVdXt3DhwqSkJFtbW1NT0+zs7ODgYC8vL+Z7vJub2969ezMyMoYOHTps2DCml58hZnDrmjVr7OzsnJ2dlZSUxBTaYhV8fX3Hjh3L4XAuXbpkZWWlrq5uaWkZHx+/cOHC8ePHC4LZunXrsmXLNDU1T5w44eDgkJ6erq2tzePxRo4cmZqa2rt3b1dX1zNnzqxZsyY5OTk8PLxxnFVVVffv3xepZnNtItjdz89v0aJFFRUV8fHx7u7uaWlptbW1o0ePTk5O1tDQmDlzZnl5+aNHj96lBaBTkZOTGz58OBGNHTv2wYMHRMRms01MTAQJtLS01q9fP3HiRF1d3YyMjCNHjmzbto3p0SsoKOjWrZuPj8/YsWOHDh2amZm5aNEi4a7AzmD8+PHMDykpKebm5r///vvXX38tkqaoqIhZFPLYsWPLly8nosLCwm7duonm1RE4HE5ISAiHw3nw4IGDg8OLFy+ISFVV1dzcXPyxO3HixO7du1VUVGbOnLl3714iYh7VmpGRcfXq1Q6pi/iA2/BMU1dX79OnD7NapRi9e/fev3//0KFDnzx5smvXLuGXgoODx40bp6WlJS0t/ezZs927d+/Zsyc5OZm5kPHx8fHx8Xn48KGVlVWTKYnorRIL9OrV69q1azY2No8fP/b29m5cR01NzXXr1jk6Ourp6aWnp+/YsaPN+7U/kL4VYYJRFXV1dampqY8ePZJkr4cPH+7YsWPAgAEaGhpM34GamlrjZNbW1vr6+jk5OUuWLOnbt+/NmzeJyNbW9l0+QWxsbAwMDLKzs0NDQwcPHlxVVZWeni4lJWVjYzNgwAA9Pb3c3FxfX9+hQ4e2Yspxc5lPnDiRSdBkKzHjIC5fvlxdXW1paWlvb29mZpaUlBQcHJycnHz27NlW1+ht4xfj2LFjubm5WVlZxcXF8vLydnZ2ioqKb3V0RKop+FL4VmeOjY2NsbFxRkbGV199NWLECB6PV1dX5+fnJ2EjiEm2efPmV69eOTo69u/ff/PmzX5+fr/++quysrKYFhDOuXVvBAlje9us4MNgYyV3/4jR8xd18Q9rwi6UR8RU+v5cOGOccm/DN3+ARw6QJyJtjTd/WSqqG9SUpRoaaP0vhaHny3ML6xv+nq2TV/TP4JQe3WWJiMUiEz0Zbm5d1kvRZ3hl5tUTUTWP32Pqc8FGbm5dTS1/sbPaQv+8kP8r09GQLix9PciCY9WLTY14TlVdvuPVkfNldfX8iuoGGys5JlmLsTWpXuj15mIb1lfOfbLKkQvl89bnEZEcm7Vxcbfls9TF5wzwXnl5ecXFxTGX9Dt37iSi06dPp6SkyMvLh4WFKSoqGhoaenp6btq0aeHChdHR0XJycg0NDdXV1S9evIiIiLh8+bK9vf2qVasiIyMzMjLGjx8vMlpbqnHv4988PDyCg4OZn8UU2mIVli1bFhAQ4OPj88MPP9TU1Ny+fVtFRUVFRYXH4yUlJQn6VlavXr1q1So+n9+zZ08ul3vs2DEvL69z586lpqbKyMhER0fr6OhMmzZtypQpR48e/fHHHwU7Csfp4OAgUs3m2kQQHjM1OCEhYciQIenp6SUlJVFRUcnJySwW6+bNmxYWFkRUV1cXERHR6haATqWmpiYnJ8fAwODIkSMRERHx8fHXr19PTU1lXuVwODExMWZmZmlpaeHh4U5OThs3brSyshKebr927dr9+/crKSlZW1sfOnSod+/eHVQVcnZ2FjyUQLjPtEUhISHe3t7Lli2LioqaPXt2YmJiQUGB4EKjY9nY2DDf9n19fZmOFSIqLS1lRu2JOXbV1dVHjx5dtGiRu7v73r17WSyWm5sbER08eLCjHr8g5mRrkzNNU1NTSUmJ+YGIDAwMmJvcr169qqysFEksKyt77tw5U1NTLpebkpIisiRov379oqKi0tPTDQwMPD09d+/enZKScvjw4cWLFxsbG9++fTs2NpYZMdBkyqtXr75VYkG5/v7+Z8+effLkycCBAy9cuGBqaio8qJ/NZkdHR1tYWNy6devw4cOenp4hISGysrK//PJLGxyev32AfSu2trZJSUnXrl3bt2/foEGDhg4dGhMT0+Je6urqqqqq165dKy8vV1BQGDRo0H/+85/GydhsdlBQ0IEDBxISEq5evaqtrf3ZZ5/NnDnzXQKWlZUNCgr67bff4uPjDx8+rKCgYGxs7OjoSEQyMjIBAQHbtm1LSUkpKyubMWNGUFBQm2QuvpVcXFxSU1PT09NPnTpVWVlpb2//9ddfb9myhcvl3rt3z8PDQ/ycSTE1akPz58+PjY3NyMgwNTX19vZmxuy91dERqebKlStbcebIysoGBgaGhobGxsZeuXJFUVFx0qRJJHEjNJfs+PHjsbGxenp63t7e8vLycXFxUVFRAQEBwl9km2wBgda9ESSJ7a0yESbS+yNMzN3OxvBQoY5yMKLM2V6pR3fZHt1lh/eVj4h5TkRFZa+J3vStsGWJGk2cORVVEfBbsbKC1OH1uoY6Mmt/Loy6Wy18DJPSeDPGKfFq+Y+5tURkoCP6h8lQW4aIFOWlTmzWE8z34dXxZaRp1kTllYGvriVUsWVZROQ+penHirs5Kq/eU3A+tqqgtIGEBq20GJuAorxUZXVDcVkDESWl81qMjS3LClmr89Mq7UfPa/+8Wfn9/sL1vxShbwU6m8zMTCKqrq4WvtPO5XKrqqr8/PxCQ0Nzc3Mb/u53ZObAtkjwGD5h48aNa7HQmpoaOTk58RGbyJsAACAASURBVJkz137MV/+ePXsyPygpKfF4PGZoLYOZFMxisUxMTLhcLlMi87+mpiYz1p2Z5kNEWVlZgitJ4ThFNDQ0fP/99+LbpF+/fvT3vGYiqqioYAaWampqMh0rRCQrK/suLQCdjYuLy759+/r169e3b18iqq2t9fPz8/f3J6JJkyaZmZnV19ePGjUqLy/v9OnTERERM2fOXL58uWAoekBAwLp16wYPHnznzh1TU1M1NbXGjyZoHzo6Oq2bBvLy5cvw8HB3d3dmhMiOHTuYbojOgOlYISKRSUwMMceOiIKDgxctWmRtbW1hYaGrq2tiYsLn80NCQtoteMkDbpMzbefOnbNnzxb8yiwmQERubm6hoaEikTAj+Boa/p+9+46L6lj/B/7sUpYOUkVAinSMepUIUUTsiv6ixhaxgRpzNRBjvgZjlNjAqyaaxBaTXIMmisbujb0SWwDFimAoglSBpUrbpZzfH2P27l1gWVaqft6vvPI6nDMz55nZ4u6zc+bUeXt7Z2RkBAQESF8qNWjQoKqqKj6fr6mp2a1bt3Hjxo0YMeLzzz8fMmSItbX1xYsXJWuEN1jy8uXLGzduVLyw5LxfffVVSEiIurp6Wlqaubn51KlTpb+xjh071tnZubKy0s/Pr7y8PCMjY/fu3cuXL+9guRWFvxG9/LDayJerSIXvzSy5pJY5efKkTAE+n//JJ5/Un7EmU9jCwkJ67U8LC4stW7bUr8JmIknv6dKly//93//Jb1xC+qYqkydPnjx5coPlDQwMPv744/rVicjKykp6ntV7773XYDE5L/XGGm9slIjIwsJi586d0ntsbW3ZjDhG+m7QDXZcTo8aC1vBsZKwtLSUmYFGjT86DTZSv5tyxkTOk0dPT4/dm0mmioKD0Fgx6UvSVq1atWrVKpkCDY6AdJwKdkdO1+R0ocEHBV5vc9flfrQpz7G7umkXlQdJIiIyN1Z9q0cT14XV1L58+xeJudgnoqg42dVqN0cUPc2qTkwX5xbW6mjypw7XlSkwdqC2Y3f1xHTx+vDCER5aL8rr7ieKVFRo7EBtVRXyH6f3zYHis7fK1dV4fqNk6zJG+ioTB+scvPji5oNKXS3+tBE6CsYm0c9ZcO1eZciuglsPK3cdK2kyNgMd/keb8gf31exqrBKXLCIiU8OmL4oEaGNWVlZEpK2tffToUcn1niKR6D//+c/GjRt1dXX37dtnZWUVEhISGRkpyWuzkpL8AmuhvLycLen18OHD+ieSvpdcYyeVfyEww6aps3S8nCnrDx48mDp1qkgkSkhIkJyR/V8oFObl5Zmamj5+/JgVZtfz1o9TppvHjh2TMybSVaR/LWBrsQmFwsTERLZqXnV19auMAHQ0UVFRvXv3trW19fDwmDFjxrhx40JDQw8dOpSUlCR59FkOLi4ujlWxsrKSfONlr5eSkpf/rOjo6LRXbmXHjh2Sm2akpaVZW1srXnfLli1z5szx9fXNyso6dOhQx8mtSLKfDg4O9Wdwy3nsiCgmJiYuLq5nz55z5sxh6dfLly+zbGl7aSzgFnmmffPNN0eOHCGioUOHBgUFzZ8/n90MpMH7YLA3sYKCgoyMDCKSvg8pn89fvXr1zJkzzc3NJZMZG0zbKV5SwcKs42KxODEx0dzcXLIUJsP+1NTUlKwaQUQ2NjZs4YsGT6oEvIkDAIBC1v3T6HJMxV/p1QlpYkM9/nhv7bX/NGpyGZEpw3Rv3K88cP7Fsu3C4W9rjX5H+8QfZdIF1n5odPpGeXyquI+j4Lv/MzGrl4MQqPOu/WgZurvw7J/lYeGFulp8V1t1/1Evp6gsnGzw7cFijqP/56VtqNdo/mL+BL2DF18Q0fsjdXU0+QrGJrH9M9N5obnxqeKrsZWrFxgt2ZIvPzYzI1VjA5WDF18UldbpavGG99faEChv8XyAdjF27FhHR8fExMSwsLCRI0eWlpbev39fRUWFrSxARCKRKDY2NioqSroW+9K1f//+srIyT09PPz+/fv36Xbt2beXKlTdv3ty1a5dyJx07dmxL9Wvz5s1Pnz5NTEzMzc3V0dGZOnUqEfn6+trb2ycnJ3t7e3t5ebFfCKZMmVJ/qfgGuyn5NN/gmDTG19fXxcUlISFhwIAB7777bmVlpUgkOnDgQGuPALQZf3//I0eOpKampqam3rp1a9y4cUTEbm7FvnYaGxubmprm5eVJVs3LzMyUVGdXcHf2CbkPHz68fPnysGHDtm3b1qHusfXnn38+f/68a9eu69ati4mJYYkGtt5KTEyMnMeOCQ8P37x58+zZs/X09KhdV7FlGgu4RZ5psbGxsbGx9PeyGOfPn5euLoMdYpd9lJSUSO4fQkTvvffesmXLXrx4MXPmzIyMjHXr1vn4+LCMM4tB8l4qp2RzCzOurq5EpKamxhLZbFgk2J/l5eWTJk2SrJwgEAganGupNORWAABAISsC/mflWmn5F+wk2/aWanXR/72OV4VP2z8z3f6ZaWPNOlipRe6ylNkp3SARmXZR2brUhMiE6rG3VHurh+BhsqixC4KYoe5a0lE1GZtMAD17qEeHW0n+XDztvwtyNRbbpR0WcuIB6AgEAsH169fXrVt39uzZ0NBQXV1dV1fXgICAKVOmXL9+/cCBA8HBwcOHDx89erT0im+ffvrp3bt379+/v23bttLSUj8/vx07dsydOzc+Pv7q1atr1qxpbMqk/JO2YL/WrVt36tSp+Pj4Pn36bN26lf22qaGhcePGjS+//PLChQsRERHdu3cPCgoKDg5urBGZbu7evVvOmMjvbFhY2KlTpyIiIvT19efOndsGIwBtJjw8fOfOnYmJiXl5eb179yainJwcttTomTNnkpOT7e3tr127duPGjfHjxxPR4cOHJQt/vE7mzp3r6OgoJ+covZ5LQkJC2zzhRSLR3Llzjx8//tZbbyUlJd2/f19NTc3NzW3Pnj0xMTFyHjvm119/3bBhA0u/FhcXHz9+vA1ilqOxgB88eNDGz7TIyMinT5/a2dkdO3bs3LlzH330keSQZP6dQCDo16+f9CqNbNbPjBkzdHR0oqKiJLMC65dsbmEmODjYwcHBwcHB3Ny8rKzs0KFD0kdPnz7N5g+uWLHiwoULenp6ffr0qa2tPX369KsPiARyK9D54GoUjAAAc+KPspsPqh6liJxt1H0HaLd3OAAd3b59+2SunDc1NW1wKbcdO3bI3IJBwt7eXuYbVM+ePdnakMzixYsl20KhsH4LjZ1UhnRd6enrMle81j+Fg4MDuyG0DDMzM3YDPvnnYup3U86YSFe3t7eX/nHYyMhoy5YtMheeKzgC0DGxK7nYBI2VK1cOHz7cycnJxcWlsLDw5MmTISEhFRUVRFRVVeXl5bV27dqRI0f6+fmlp6dv27Zt06ZN7Rx960hPT2cLCTVGej2Xtrz87ezZs+7u7suWLRsyZEj//v0LCwuvXbvGLn6R89gx+fn5v//+O1uQISIiogWvHFGQ9DNNfsBt/EwTi8W+vr7sPkHsLU6yTMHhw4cHDRo0ffr0TZs2Xbp06dy5cxMmTGCHtmzZ0rdv3z59+gQFBenp6c2bN6+xks0tzHz55Zfjxo3r2bPn/fv3P/74Y5n1sEQi0aBBg0JCQsaMGbNy5Up217YWX0CH96qT0JqzAqUciq+3AgCdlJy1bJulg0ydZRMR60+FAMWZjHxaUFJ7bJP5hME6yrUwZnHWpZiKt+wFe1aZ9bLHHcE7EL5HEnWYV2tzsVd3Jw3+DWdsbFxQUHD8+PH6H7uhXXSuV5MiK+t37949MTFRIBAEBQXJv7dD61FkPJt1l4COSX43O3gHXz34jvBMAwkF38QwbwUAANqHzHU3Sjj7Ha67AQCANrJ161Z/f3+BQFBcXHz06NH2DgdeW3imdVL8posAALSOsrIyTsqlS5f4fP4PP/xQXFxcVlZ29+5dXd2Gb/sCAABQn1Ao5DgOk1aglTg4OFRUVJw5c2bo0KGSO7AAtDg80zqp1py34ulJ7M7YIhHdukWzZ1Pjqw0DwBto3Lhxqqqqhw4dSkhIWLVqVUFBwYgRIxYsWDBv3ry0tDRvb+8OPuETAAAA3hxjxoxp7xDgjYBnWifV+vNWQkNp+nQaOJA2bGj1cwFApxIZGXnp0iWxWJyfn3/p0qV79+7p6OgQUd++fWtqatatW1daWtreMQIAAAAAADSh9XMrKSl0/DjFx1OvXq1+LgDo5E6fPs3u5fbHH38kJSVZWsremhcAAAAAAKCjaf3cCp9PVlZka0sZGa1+LgDo5KqqqsaMGdO9e/dVq1bZ2tpOmjSpvSMCAAAAAABoQuvfJ2j3biIioZDWrGn1cwFAJ+fr6+vt7X316tUXL14QkVAobO+IAAAAAAAAmtD6uZU1a+joUUpKoqqqVj8XAHRyubm5Pj4+gYGBNTU1+/fvP3ToUHtHBAAAAAAA0ITWz62kpdGjR61+FgDotLp27SrZjo2N9fT0bMdgAAAAAAAAmovHcdyrNdAyd0jFfVYBQEGv+q7VQtj9oeuiHdo7EICOiO+RRB3m1dpc7NXdSYMH6FA616uJ10Lfa1qbIuPZWfoih/xudvAOdurgoT4F38Rafy1bAAAAAAAAAIDXV0eZt0KdJJ8NAEprqSR9B/n5C785ADSpg7xam6tz/dIO0JF1rldTZ/mXHfNWqMN3sFMHD/Up+CbW+uutAAAAAHQ2+OwLAAAAikNuBQBAGZ3lVzgAAAAAAGhtyK0AAAAA/BcypwAAANBcyK0AAAAAAMCb7nXKq75OfWlQp+5gpw4e5MB9ggAAAAAAAAAAlId5KwDQRpCkBwAAAACA19Ir51bwZQkAAAAAAAAA3mC4JggAAAAAAAAAQHnIrQAAAAAAAAAAKA+5FQAAAAAAAAAA5SG3AgAAAAAAAACgvFbPrfj4+PCkzJ8/v7kt9OnT5+DBg60Rm7Oz84kTJ1qvfIPc3d337dv3io0AAAAAAAAAQAfRFvdgDg4ODgkJYdtqamptcMaamhpV1Q56e+ng4OA+ffooXr4j9wUAAAAAAAAA2uKaIDU1NZ2/CQSC0tLShQsXWllZmZiYvP/++0KhkBXbsmWLk5OTrq6ujY3N+vXr2c6goKD4+PigoCAbG5s5c+YQUdeuXW/cuMGO3rlzR0dHh207OzuHhIQMGjTIycnpxo0bjZ1FRkJCgoeHh76+/oABA+Li4thOOXVTU1MHDx6spaXVr1+/R48eyYn8m2++eeeddyQVo6KidHV1y8vLN23aFBMTw3bm5ORMmTLF1NTU0tIyMDCwoqKiwb68yuADAAAAAAAAQKtqh/VWZsyYkZube+/evfT0dH19fX9/f7bf0tLy3LlzpaWlx48f37p169GjR4lo27Ztrq6u27ZtS0tL27t3r/yWT5w4ceLEib/++svHx6exs8j44YcfwsPDhULhqFGjxo4dKxKJ5ETIyn/33XcFBQWenp4fffSRnMhnzpwZGxubmJjIyuzdu3fy5Mna2trSZ580aZKamlpKSsqdO3fu3r376aefNtgXRUcWAAAAAAAAANoe1zjpAs3dlhg8eLCmpqbR3w4fPszj8fLy8tjRkpISHo9XXFwsUys4OHjhwoVsu3fv3gcOHJAcMjMzu379Otu+ffu2trY223Zyctq+fTvbfvr0qSJncXJy2rBhA9uuqakxNTU9d+6cnLrSp3j06JGmpmb9QZOO/N133/3iiy84jquqqurSpUtkZCTHcf369fv11185jnvy5AkRSU50+fJlDQ2N2tpamRPV9yqPC7axjW1sYxvb2MY2tlt7W3oPAAC8CdpiIY8PPvjgs88+Y9uPHz/m8Xhvv/225Ki+vn5WVpa+vv6JEye2bNny7NkzIiotLR0+fHhzT9StWze2kZaW1thZZKrY2NiwDRUVFSsrq6ysLHV1dTl1LSws2E4dHZ3Kykq2GEpjkfv7+y9evHjdunUnT540MDDw9vaWPnVmZqaOjo6JiQn7087OrqqqSigUmpqaSvcFAAAAADoXjuOIiMfjYRvb2MY2tt+QbXm5FVZCuW1purq6lpaWbFskEvF4vAcPHsikOXJycqZOnXr27NmhQ4fyeLzly5cnJSWxQ3z+/1y4pKOjU1VVxbZlVlFhHSOi7t27N3iW+lhChIhqa2szMjIsLCwUr9tk5OPGjVuwYMGVK1f27t07Z84cSXiMpaVlWVmZUCg0NjYmotTUVIFAwLal+1Lfqzwu2MY2trGNbWxjG9vYbu1tOR/kAADgtdTW66306NFj5MiR8+fPz8nJIaL8/PwjR44Q0YsXL4ioV69ePB4vOzs7IiJCUqVr165//fWX5E93d/fff/+diKqqqr777rtmnaW+Xbt2PXnypKam5l//+pdAIPDx8VG8LiMncjU1NT8/v02bNl28eHH27NkyFZ2cnDw9PZcsWVJWVpabmxsSEjJnzhyZRBIAAAAAdDqcFOk92MY2trGN7dd1ux2+yR84cKBbt24DBgzQ1dX19PRk98FxdHRctWrVwIEDBw4cGBQU9O6770rKL1u2bP/+/V26dJk8eTIRbdiw4d69e7179x43btzgwYObdZb6Pvjggzlz5hgaGp46der3338XCASK12XkRE5EAQEBFy9eHDhwoK2tbf26R44cqaiosLW1/cc//tGrV68tW7bIGzgAAAAAAAAA6Hh4kiwLAAAAAAAAAAA0F65AAQAAAAAAAABQHnIrAAAAAAAAAADKa4t7MAMAUEvfNAHXMwIAAAAAQAeBeSsAAAAAAAAAAMpDbgUAAAAAAAAAQHnIrQAAAAAAAAAAKA+5FQAAAAAAAAAA5SG3AgAAAAAAAACgPORWAAAAAAAAAACUh9wKALQbT09PjuP8/f0lG5KddXV1xcXFV69e9fT0bO8wAQAAAAAA5FFty5PxeLy2PB0AdF5ff/11XFzcl19+ee7cOQcHh/z8/PaOqAF4TwNQBMdx7R1Cs+HVDdDiOv5bQWd54Xf8kQR4M2HeCgB0RPHx8b/88svcuXP19fX9/PzaOxwAAACAzoTXybX3+AE0W5vOW2FWrVrV9icFgHa3Zs2a5lZJSEggImtr61YIp8XURTu0dwgAHRTfI6m9Q3gl+HEYoEXgezIAvAkwbwUAOi5nZ2ciSk9Pb+9AAAAAAAAAGoXcCgB0RK6urrNmzdq9e3dJScmBAwfaOxwAAAAAAIBGtcM1QQAATVq6dGlZWdm9e/dmz56dm5vb3uEAAAAAAAA0CrkVAGg3UVFRkmuwJRvSOwEAAAAAADo+XBMEAAAAAAAAAKA85FYAAAAAAAAAAJSH3AoAAAAAAAAAgPKQWwEAAAAAAAAAUB5yKwAAAAAAAAAAykNuBQAAAAAAAABAecitAAAAAAAAAAAoD7kVAAAAAAAAAADlqbZ3ANBiFi1aZGpqyrarq6uLioru3bsXFRXFcZz00bq6uqqqqsLCwsePH8fExNTW1kpa+OSTTx4+fHjlyhVWOCYm5syZM+yQqampl5eXjY2NtrZ2ZWVlTk7OzZs309LSpE/KxMfHHzp0qLU7GxwcrKWldfDgwSdPnrx6MWgbq1atanD/mjVriIg9UaHzGrM463xUxYoAw3X/NGJ7TEY+LSipPbbJfMJgHTkVFSz2JE3sOu0ZEeWdtzM2UGnByJulg4QB0OJGjx59/vz5FStWhIaGsj3GxsYFBQXHjx+fMGGCnIoKFnvy5ImLiwsR5efnGxsbt2DkzdJBwmiMgoNZ/8GCthQXF+fm5rZjx47AwEAisrCwWL9+/fDhw01MTEpLS1NSUvbs2fP999+3d5gA0NYwb+V1U1xcnJiYWFBQYGpqOmrUqIEDB0ofLSsry83N5fF4lpaWo0aNCggIUFNTY4fMzc0NDAwazEHY29svWLCgV69eGhoaWVlZJSUl1tbWrq6u0s1m/q2goKBVO9gs2dnZmZmZlZWVLdWgigq+SgE0oLqmgdSYu4vAw03DUA+vGoCOrrq6uv5Od3d3Dw8PQ0PDto8HoBP57bffZs+era6uHhkZmZKS4uzsPGzYsPYOqhni4uI4jtu+fbtkT1paGsdxGzZsIKLi4mKuIQcPHpTU5Thu6dKlkurnz59nO7/++usG2wd4XSG38rpJTEyMiIjYtWtXamoqEbEfZyTi4+N/+OGHTZs2nT59mogsLS29vLzYIWdn55KSkuzsbJkGVVVVJ0yYoKqqmpubu23btp9//vmnn37asmXLgwcPpJv9998uX75cPyotLS1fX9+PP/54xYoVCxcu7NOnD9vv4uKyevXqZcuW6erq8vn8efPmrV69evTo0UQUHBy8evVqb29vf3//5cuXf/jhh1ZWVg12efz48UuWLFmxYkVISEhgYGD//v0lh7p162ZpaampqSlpcNCgQbNnz16+fPmiRYskDTYWnnStefPmhYSE2NjYNP0YAHQqJiOf8j2SwsILhy7M1B+S4j47/dbDKnZofmiuzfhUbe9kDa9k12nPdhwplqm1fk+h1wcZgoHJgoHJ56MqiCgsvJDvkfTW9GdEdCdBFP24qrC0tsnWGpOUUT10YabO4OR+s9OjH1fJHBUW1378db7DpDSdwcl9ZqTvOVXaZI8aqyLdoxGBWfpDUt6a/kxSS04YyjVY9KLus61C5ynPtAYlm495+sVOYZOtATTG2NiYx+OFhYUNGTJET0+vX79+t27dYofmzZtnbW2tpaUlEAhcXFx27NghU2v9+vUDBw5UV1dXV1c/f/48EYWFhfF4vJ49exLRnTt3oqOjCwsLm2ytMUlJSUOGDNHW1u7bt290dLT0IaFQGBQUZG9vr62t3bt37z179ijSoyZrsekDenp6PXv2lNR6lTAabLCoqGjp0qVOTk6amppdu3Zdvny5/KYUf8ik5efnL1q0qEePHtra2q6urmFhYSKRiIh69uxZ/8GC9qKhoTFgwAAiGjp06MiRIz08PExMTL744gt21MTEZOfOnSkpKeXl5fHx8StWrBAIBOyQUCjkOO6LL764dOlSaWlpXFwca6ejuX37dnR0dHR0dFVVFRE9f/6c/ZmUlCRdLCgoiP0A6ebmNnLkyPaJFaC9IbfyetLT09PT0yMi9s+wDI7jbt++nZKSQkSS6SfOzs4NTlqxtLTU0dEhoqtXr7548YLtrKqqysrKkpRxdXWd/zfp+SyMiorK3Llz+/fvX15efvPmTQ0NjQkTJvTr14+IEhISYmJiNDU1x48f7+XlZWVllZ2dffHiRUndAQMGpKWlpaSkmJubz5w5U1tbu36EZmZmaWlpN27cuH//vqGhoa+vr52dXWMjM3jw4Pz8/Pz8fFNT04kTJ8oPT2Lo0KE1NTXx8fFisbixlgE6ta/3FQ3upzXKU+vuX6Ixi7PyimqJ6GGy2Kev1udzDP3H6iWli4O+yr98u0K6VsiuAg11/uShOuO9ta27qhJRfzeNxdMMZo7Rq3+KJluTUV3Djf0kK/JupWkXVWdr9cWb86WPiqs57w8ztx8uNjNU+WyWYdGL2rnrcn86USKnR01WIaJ1uwtdbdVdbdUfPxX7r3kuPwzlGhRXc4M/zNi8v0hYXPv+SF2v3prxT8UKtgbQmK+++srHx2fUqFF3794dPXp0Xl4eET18+NDHx2f58uX+/v5JSUmBgYEyv3+sXLlSQ0Nj8uTJ48ePt7a2JqL+/fsvXrx41qxZ9U/RZGsyqqurfX19IyMjTU1NnZ2dP/74Y8khsVg8aNCg7du3m5mZBQcHFxUVBQQE/PTTT/J7pEittWvXurq6urq6Pn78eM6cOa8eRv0GxWKxt7f35s2bhULh+++/7+XlxT4eNNmUIg+ZhEgk8vLy+v7771VUVKZPn56Xl7dy5Up/f38imjVrVpMPFrQZyUfi/fv3r1+/fvz48VpaWomJiUQkEAhu3LixcOHC2traAwcOmJqahoaGyiTdvvzyy/j4+Pj4eDc3t71797ZLF+QbMWKEp6enp6fns2fPiOi3335jf4aEhEjKFBYWdu/efdKkSUT06aefElGHmsYO0Gaw3srrpn///pKJGxzHNfhLCFNQUNCjRw+WNzEwMDAzMzt37lz9YqwAyX2X1NHRkRSTbEg4OjoaGxtXV1cfPXpULBaXlJSwTEpsbCwRXbhwwcrKyt7evkePHiKR6MiRI9JLwNy4cePGjRs8Hm/x4sUGBgZubm4xMTEy7f/88881NTU8Hk9NTU1XV9fR0dHOzu7p06cNhnrz5s2rV69269ZtwYIFhoaGGhoatra2csJj7t+/f/Lkyca6D/Aa+HyO4bLZXTiOekxMS8upPnTpReAUg2s/Wmqo8+rqqFJUly2sOXWj/FJMxbC3tSS1Asbp/XulGdseszjr2fOaEf21JOutyGiyNRmRdyuTM6v5fPrjB0srM9XwU6Xz1uVKjp6+Wf4kTawp4O1fZ66twbMyU50fmrthb9EHE/Qb65GFiar8KkS0bFaX1QuM7iRU9ffPSM6sLn5RdzuhqrEwmoyhwQYj71bEpYh5PLrxk6WzjTr9fUWVIq0BNGb58uXLli3jOM7Ozi4tLe3QoUOBgYHXr1/X0NCoq6urrKzMzs4+derUxYsXpa9WCAgI2L17N9sePXr0s2fPRowY0dgSHk22JiMyMjI5OZnP51+7ds3Kyio8PHzu3Lns0OnTp588eaKpqRkREaGtrW1lZTVv3rx//etfH3zwgZweWVhYNFlr2bJla9asuXPnzttvv52cnFxcXHz79u1XCaN+g5GRkXFxcTwe7+bNm87OzkRUXV196tSpJptS5CGTHD1z5kxiYqKqqur169fNzMwmTpw4bty4gwcPbtmyZdmyZVevXpX/YEFbmjJlyg8//NCrV6+33nqLiMRi8dq1a8PCwnx9fR0dHWtqagYNGpSbm3v8+PFTp069//77n376aU5ODqu7cePGVatWubu73759297e3sDAoLi46UmdrWHy5Mnu7u5s29zcGtY4rwAAIABJREFUvFl1w8PDg4KClixZEhkZOWPGjNjYWKFQOGrUqFYIE6BDQ27ldVNcXJyXl1ddXV1cXPzgwQOZn0GkGRkZEVFZWRkRubi4VFZWsoS0DFaAlc/Pz69fgIikV72tT19fn4jU1NQ++eQTyU4DAwNVVdWampqamppbt25NmjSJx+M9evRIMveYKSoqIiKO44qLiw0MDFhT0ng8no+PT69evXR1dXk8HttZP78jkZubS1LTedTV1eWHx/5sLFMD8Nqw7aZGRDwe2ZirpuVUZzyvqaujNT8V7Dv7Iqegpq7uZbHcwlrpWsP6N5oZkaFIazIycmuIyEhfxcpMlYj6Ogmkj6bn1hBRpYizHZ8q2ZmWU10l5hrrUW0dNVZFQ/3lu8dbDgIi0td5uUZMWWWdnDDkxCCnwWfPa4jI2ECFJVaISE2Vp2BrAI2xtbUlIh6PZ2Njk5aWlp6eXldXt3r16n379uXk5NT9/apj/whKDB8+XMH2FWlNRkZGBhEZGRmxK3D79u0rOZSenk5ElZWV0lfapqWlVVVVaWhoNNYj9tOL/Fq9evWivz94EFFZWdkrhlG/QfZhydjYmCVWiEhNTU2RpmTU76D0UfansbGxmZkZEUku/MnIyGju915obVFRUb1797a1tfXw8JgxY8a4ceNCQ0MPHTrUvXt3IhIKheyVEhcXx8pbWVlJcisPHz4kopKSl1MUdXR02iu3YmZmxp5sSnj+/PmBAwfmzJkTHh4uEAi++eYbTKeCNxNyK6+bxMREOWkOhsfj9evXr0ePHkQUHx9PRM7OzomJiZJPS9IyMzPLysp0dHSGDBmSlZXFLgvS0NAwMjKSvixIDvYPhlgsPnTokOQUqqqqbFtLS2vEiBFEVFNT07dv3wcPHrCPQUzXrl0fP36sqqrKFvMvLZVdfcDFxcXLy0skEh07dqykpGTo0KHyl0RhJ5W+JY388JgGF/kDeJ08TBJNHa4jEnMJaWIisjRTPRZZtvGXIl0t/q9rulqZqX65qyDybqXM3Zw0Bf/9zq/C5xFRQ+8iRESKtCbD0lSViIpK60rK6vR1+KnZ//MytDJVJSJtTf6RDeYqf1/eKqrmVP9eObd+jyxMmqhCRGw/TyqVISeMJmNosMHuZqpEJCyuTUwXO3Z/OW9FTZWnSGsAjXnw4MHUqVNFIlFCQgIRWVlZHTt2bOPGjbq6uvv27bOysgoJCYmMjJS5Ixtbj4xhayU0+EmAiBRpTYalpSURFRUVlZSU6Ovrs2XgGJbm0NbWPnr0qGSReJFIpKr638+l9XtkYWHRZC22nyf1knvFMOo3KPnCnJiY6OjoSETV1dWKNCWjfgelj7I/hUJhXl6eqanp48ePpbsj/8GCNubv73/kyJHU1NTU1NRbt26NGzeOiAwNDdkHWmNjY1NT07y8PDc3N1Y+MzNTUpdlDDvCrRIltz0iorS0NHbdmeK2bNkyZ84cX1/frKysQ4cOIbcCbybkVt4srq6uVlZW+vr6WlpaRJSZmXnz5k0tLS0rK6s///yzwSo1NTUnT56cNm2amZlZUFDQ8+fPVVRUTExM7t+/L8mtuLq6duvWjW0LhcITJ05It5CUlFRQUGBkZDRo0KCUlBSBQNC1a1eO49jFqBMmTNDT07t3796zZ88mTJgwefLkXbt2Se7s884773Tp0sXIyEhHR0csFks+W0jw+S+/gqioqLCVa5s7JvLDA3hDbI4oeppVnZguzi2s1dHkTx2uezX25WIoIjEX+0QUFSe7lKwMa3NVItp/vrSsss6jp4bfKF3pozW1XLNaIyKfvpp2FmpPs6onLcsZ9Y7WzsP/81Pe2IHajt3VE9PF68MLR3hovSivu58oUlGhsQO1G+uRgQ5ffpXmhtFkDA3yHaDtYqOekCYeOD/z3UHalSJOVM0d3WiuXGsAzObNm58+fZqYmJibm6ujozN16tSrV6+yQyKRKDY2NioqSn4L7KvU/v37y8rKPD09/fz8pI9KJnIq2BoR+fj4sEt033vvvdGjR0uvfTt27FhHR8fExMSwsLCRI0eWlpbev39fRUVl7NixcnpkYGDQZK0WD6M+X19fFxeXhISEAQMGvPvuu5WVlSKR6MCBA81tqn4HZc5ib2+fnJzs7e3t5eXFLkyeMmUK+7gl/8GCNhYeHr5z587ExMS8vLzevXsTUU5OzqNHjx48eJCcnGxvb3/t2rUbN26MHz+eiA4fPlz/xhGvgYcPH16+fHnYsGHbtm3DT5LwxsJatm8WHR0dNt8vMzPz3Llz4eHhYrHYycmptraWLW3boKSkpB9//PHhw4dVVVUWFhb6+vrPnj1jE14kzVr+zdTUVKZ6TU3Nzz//HBMTo6ur6+3t/Y9//ENVVZVNjHznnXccHR2LiorOnj17//79x48f6+vrs397mKtXr+rq6pqYmDx//px9hpBpPD4+/vbt23V1dSNGjLCwsEhOTm7umMgJD+DNsfZDo2xhTXyquI+j4PS33cwMVaYM0100WV9VhZZtF8bEVY1+p4lv+Ev8uni4aeQW1m47VHwxWnaR2ua2RkTqarzT33Tz/odmzOOqiHMvlvh1kT4qUOdd+9EycIpBtrAmLLww/FSpqJqbOvy/CZ36PWqySnPDUK5BVmvJdANDPX7EhRdX7lQ4Wasp3RoAs27duuzs7Pj4+D59+pw5c8bMzGzKlCmLFi1SVVUNDg6Ojo5m9+CT49NPP/Xw8GD3BLxw4YLM0ea2RkTq6upnzpzx9vaOiYnZv38/W+GSEQgE169fDwwMzM7ODg0NDQ8PF4lE06ZNk98jRWq1eBj1sVpLliwxNDSMiIi4cuWKk5OTEk3V76D0UQ0NjRs3bixYsEAkEkVERBgZGa1Zs+aXX35hR+U/WNDa2LwhSQZh5cqV0dHRpqamgwcPrqurO3ny5KhRoyoqKqqqqry8vH788UeBQODn51dQULBq1arZs2e3a+ytaO7cuSNGjGjsDmKTJ0+O+lt4eHgbxwbQNnhtOQmNzahctWpVm50RFDF9+nTJbeo7juDgYC0trYMHDzZ49yJ4naxZs4Y6xoTY5mLvaXXRDu0dyCsxGfm0oKT22CbzCYMbXaioc3n9etR58T2SqDO/ujtF5MbGxgUFBcePH58wYUJ7x9IyXr8eyXjtO1hfZ3lBSV/81Zju3bsnJiYKBIKgoKDt27e3QVT1KTiSinSHiOLi4tzc3OpfE7Rx48bPP/9cUuzJkydOTk7fffed9AKFrO5nn3329ddfS7d57ty5UaNGbd68eenSpayM9NHY2FjJurlydPwnDIAMXBMElJ6eLrOCGgAAAAAASNu6dau/v79AICguLj569Gh7h9MyJCslSzS4dqFk8Wb5dRnpqW2NlQF4/eCaIKCbN29KLx8LAAAAAAAyHBwcKioqzpw5M3ToUMm9fgAAGFwTBADtDNcEAbyucE0QAFDneUEpeBFNu2vZa4I6rI7/hAGQgXkrAAAAAAAAAADKQ24FAAAAAAAAAEB5yK0AAAAAAAAAACgPuRUAAAAAAAAAAOUhtwIAAAAAAAAAoDzkVgAAAAAAAAAAlIfcCgAAAAAAAACA8pBbAQAAAAAAAABQHnIrAAAAAAAAAADKQ24FAAAAAAAAAEB5PI7j2u5kPF6bnQsAOpe2fC9qKXhPA1AEXt0AQJ3hraCzvPAVHMnO0p3GdPwnDIAMzFsBAAAAAAAAAFBem85bAQAAAAAAgNaGeSsAbQy5FQAAAAAAAAAA5eGaIAAAAAAAAAAA5SG3AgAAAAAAAACgPORWAAAAAAAAAACUh9wKAAAAAAAAAIDykFsBAAAAAAAAAFAecisAAAAAAAAAAMpDbgUAAAAAAAAAQHnIrQAAAAAAAAAAKA+5FQAAAAAAAAAA5SG3AgAAAAAAAACgPORWAAAAAAAAAACUh9wKAAAAAAAAAIDykFsBAAAAAAAAAFAecisAAAAAAAAAAMpDbgUAAAAAAAAAQHmq7R0AAEBnxePx2jsEgE6A47j2DgEAAACgdSG3AgAAAPA/kDkFaHGdJc2Kl3+n01meWvDaQ24FAOCV1EU7tHcIAB0U3yOpvUMAAAAAaAvIrQAAAAA0AL+FArQIzAQBgDcB1rIFAAAAAAAAAFAecisAAAAAAAAAAMpDbgUAAAAAAAAAQHnIrQAAAAAAAAAAKA+5FQAAAAAAAAAA5SG3AgAAAAAAAACgPORWAAAAAAAAAACUh9wKAAAAAAAAAIDykFsBAAAAAAAAAFAecisAAAAAAAAAAMpDbgUAAAAAAAAAQHnIrQAAtC6TkU/5Hkkn/ihTvMqTNDHfI4nvkSQsrlW8BTnFZBpUghK9eHXyx2HM4iy+R1LIroK2CbLNRqBVT9TGg/ZmMjY25vF4J06caFatJ0+e8Hg8Ho8nFAoVb6SxYjKtKUe5jryK+mFLxzB69Ggej7dy5cq2ibDNut+qJ2rjQQMAeJMhtwIA0LrcXQQebhqGeiqt3cKrn6iDa7KDr/0ItAYMWotzd3f38PAwNDRsg0Za5FwdmfwOvvbdbw0YtJYSFxfHcdz27duJyMLCYu/evVlZWWKxWCgURkdHL1y4sL0DJCJydnbmOI7jOGNjY/klz507x3FcaGioIs02q7BQKOQ4bsKECc06BNAZIbcCANC67iSIoh9XFZb+d+bF+j2FIwKz9IekvDX92a2HVaxYUkb10IWZOoOT+81Oj35c1VgL89bl8j2S5ofmskMXoiv4HknGI56KxJx0MfkNsskgf9ytJKI7CVXszyoxR0TzQ3NtxqdqeydreCW7Tnu240hxkx0selH32Vah85RnWoOSzcc8/WLnyx+c84tqP9qUZ/9ems7gZLdpz8LCC0Vijh169XF4a/qz81EVRBQWXsj3SHpr+jOZAsqdXYnuE5GwuPbjr/MdJqXpDE7uMyN9z6lSIjoeWcYemmxhTW0deX2QwfdIWrIlXzqGsPDCoQsz9YekuM9Ol8QgTU48SvSilQYNZNy5cyc6OrqwsJD+niOwfv364cOH6+np9ezZ89atW5KSSUlJQ4YM0dbW7tu3b3R0dIONzJ07l8fjzZs3j+2/cOECj8czMjISiUQy55LTGpsM8scff7Aq7M+qqioimjdvnrW1tZaWlkAgcHFx2bFjhyJ9LCoqWrp0qZOTk6amZteuXZcvX05E+fn5ixYt6tGjh7a2tqura1hYGAtS/jjICVu6gz179jx//jwRhYWF8Xi8nj17ynRfTgDyHwUlRkAoFAYFBdnb22tra/fu3XvPnj1s//Hjx9mjk52dXVtbO3DgQB6P98knn0hiCAsLGzJkiJ6eXr9+/aRjkNZYPHJ60ViV1hs0kPHbb7/Nnj1bXV09MjIyJSXF2dl52LBhrXGi8ePHx8XFmZmZtUbjAPCKkFsBAGhr63YXutqqu9qqP34q9l/znIiqa7ixn2RF3q007aLqbK2+eHN+Y3XnT9AnoqNXy1gqZP/ZUiKaNUZXoM6TLqZIgzxe/X30MFns01fr8zmG/mP1ktLFQV/lX75dIacv4mpu8IcZm/cXCYtr3x+p69VbM/6pmIhEYm7Qgszvj5ao8Gn6SN28otqQXQUBa3NbahxmjtGz7qpKRP3dNBZPM5g5Rk/6qHJnV6L7bAS8P8zcfrjYzFDls1mGRS9q567L/elEyUQfncApBoWltfPW5W7YW3jrYZW7i8bGoP/55fDrfUWD+2mN8tS6+5dozOKsvCLZK7aajKdZvWilQYMmrV271tXV1dXV9fHjx3PmzGE7q6urfX19IyMjTU1NnZ2dP/744wbrfvDBB0R09OhRlgrZt28fEc2aNUsgEEgXU6Q1XkOv+YcPH/r4+Cxfvtzf3z8pKSkwMPDy5cvyuyMWi729vTdv3iwUCt9//30vL6/4+HiRSOTl5fX999+rqKhMnz49Ly9v5cqV/v7+8sdBwUFgXba2tiai/v37L168eNasWTIFmgygwUdBiREQi8WDBg3avn27mZlZcHBwUVFRQEDATz/9REQTJ04MDAxkGbENGzbcunXL3d1906ZNkrpfffWVj4/PqFGj7t69O3r06Ly8vPrty4+nwV40VqX1Bg2kaWhoDBgwgIiGDh06cuRIDw8PExOTL774QlLAxMRk586dKSkp5eXl8fHxK1asYK9fNmvjiy++uHTpUmlpaVxcHGtHji5duri5uampqckp4+DgcPXq1fLy8rt373p4eMgc3b1797NnzyoqKkQiUUJCwkcffUREcXFxo0aNIqIVK1ZwHBcXF9dYyeYWlujRo8eVK1cqKipiY2Mb7KaxsfG2bduSk5PLy8sfPHgg8+4B0DlwAACgFPYuWhftIP8/I30VIjq2yVyy/eU8w7poh5g9VqyFwks9zm+zICI+n579x7Yu2mF3yMufpPLO28m0UBft4GanTkSH/mVe9oe9jiafiB5GWMsUk98g2772o6V0GBXX7euiHdj/a/50eBHZY5yXNhEtm92lfgyS/45tMiciHo/if7Nme0Q37euiHY5uNCciVRVezlm7umiH37d0Y2fJOm3bUuMwylOLiFYEGNYfauXOLqf7ckaAnUtTwEs9aZt33u7fK82IyLabWl20Q+V1+75OAjY+etr8pKM2MqH+6yPjumiH2igHG3M1Itq61ETmRE3G09xetMagyfmv837SeJXIjYyMiOj48eOS7S+//JLjuNu3b7Nmi4qKOI67cOECEfH5/PT0dI7jfv75Z3Y0Pz9fphE3NzciOnz4cHl5uY6ODhE9evRI5lzyW2Pb169flw6jsrKS4zj2/9ra2rKysnHjxhHRsmXL6ndE2vHjx4mIx+MlJCSwPWKx+NixY0Skqqr6/PlzjuNOnTrFzpKdnS1nHOSHLROD9Ne5+kflBCDnUZAzAo11n51IU1MzLS0tPz9/9+7dRGRra8uOVlVV9e3bl42Pnp5ecnKydKgbNmzgOK6urs7GxoaItm3bVv9E8uNpsBdyHsRWGrTm6lxvBaQY6WuCMjIyOI57+PDh+vXrx48fb2BgICkmEAj++usvjuMSExP//e9/s3zKgQMH6O/cSlVV1datW6OiojiOS0pKavBcxsbGNjY2NjY2S5cu5TjO09OT/amtrS1TUk1NLSkpieO41NTUiIiIkpIS1inJNUG3b9/eu3dvSEjIDz/8UFNTw3HcsGHDli1blpaWxnFcdHT0t99+u2zZssZKsmeX4oWlu3no0KF79+5xHFdaWmpqakpS1wSpq6uz95ObN2+uWrWKvSGwzLIi2vLpASCHqoJPWQAAaClvOQiISF/n5QoXZZV1Gbk1RGSkr2JlpkpE7At5Y+aN1//0m/z9Z0ura7iyyjrPnho9e6jLlFG8wRqpqRJ1dbTmp4J9Z1/kFNTU1b3cmVsob/nbZ89riMjYQMXZ5mUMaqo8IkrPZfv5ZoYqRNSzh0ASmLmxaouMgxzKnV1Pm9/c7kvOVSnibMenSnam5VRXiTkNdd7/zewyI+Q5x9H0kbo9LGV/abTtpkZEPB7ZmKum5VRnPK+RPqrIw9FSvSBlB81AFxNgm9arVy8i0tfXZ3+WlZUZGBhkZGQQkZGRkZWVFRGxL+QNmj9//pIlS/bt21ddXV1WVubp6cku7pCmeGs1Nf99mtXV1a1evXrfvn05OTl1fz9dcnNzG6n60rNnz4jI2NjY2dmZ7VFTU0tPT2c72dUKkggzMjLMzc0bGwfFw26SnAAaO7uBgYESI8BOVFlZyfIjTFpaWlVVlYaGhkAgWLp0qZ+fH8dx06dP79Gjh3RdW1tbIuLxeDY2NmlpaawpaU3GU78Xenp6SjyI0n1p7qAp0vKbZsqUKT/88EOvXr3eeustIhKLxWvXrg0LCyMiX19fR0fHmpqaQYMG5ebmHj9+/NSpU++///6nn37K6m7cuHHVqlXu7u63b9+2t7c3MDAoLpa9HPXbb7+dMWOG5M8///yTbcyaNYvNZZPw8fGxt7evq6vz9vbOyMgICAiQpCyZQYMGVVVV8fl8TU3Nbt26jRs3bsSIEZ9//vmQIUOsra0vXrwoWfm4wZKXL1/euHGj4oUl5/3qq69CQkLU1dXT0tLMzc2nTp3K0lLM2LFjnZ2dKysr/fz8ysvLMzIydu/evXz5cjYjDKCzwEciAIC2psIn+t9LcixNVYmoqLSupKyOiFKzq+VUZ1cAnf2zYueREiKaN16/fhn5DWpr8tlRInqYLJLsPxZZtvGXotLyul/XdL32o6VPX00ikv8bXnczVSISFtcmpovZnuoajoisTNn+Onady+OnL89iafbfnP4rjoMKn0dEkvSBNOXOrkT3JefS1uSf/c7iwraX//1nczdVFSooqV22TUhEGuq83f8prb9AycMkERGJxFxCmlgmPAXjaW4vWnzQQBEqKipU75IcS0tLIioqKiopKSGi1NTUBuvS31cAnT17lq2jMX/+/Ppl5LfGftwuKioioocPH0r2Hzt2bOPGjaWlpfv27bt+/bqPjw8RcU096bt3705EQqEwMTGR7amurmbJEaFQyK5zefz4sXRgjY2D4oMgqV7X4NOXqMkAGnwUlBgBdiJtbe1z585d/Nvvv/+uqqpKRAUFBcHBwUSkoaGxe/dumTVKHjx4QETsoglJU82Kp34v5FdpjUGD+qKionr37m1nZzd9+vRTp06pq6uHhoY6ODiQ1OuFJbzYFTQk9eizlyR7CRARm5sm45tvvpk4ceLEiRO3bdtGRPPnz2d/RkZGypRkzRYUFLAE2d27d6WP8vn81atXZ2ZmskQtm+XU4OotipdUsDDruFgsZu8bbFgk2J8y08FsbGw0NDQaPCNAx4R5KwAA7c+nr6adhdrTrOpJy3JGvaO187C8JVSN9FUmDtY5ePHFzQeVulr8aSMa+Bwmv8F+zoJr9ypDdhXceli561iJZH9N7cuP4yIxF/tEFBXX9GKlvgO0XWzUE9LEA+dnvjtIu1LEiaq5oxvNfQdq21uqJWdWD/4w06u3xslr5UQ0ZZhON2N5/+40axyszVWJaP/50rLKOo+eGn6jdP8blVJnV6L7RDR2oLZjd/XEdPH68MIRHlovyuvuJ4pUVGjsQO2AtbmZeTUB4/S8+2oGrM31C8m5t8+6i9REj80RRU+zqhPTxbmFtTqa/KnDdaVbVi4e+bVafNBAaT4+PnZ2dk+fPn3vvfdGjx4tZwlVIyOjiRMnHjx48ObNm7q6utOmTWtua/369bt27drKlStv3ry5a9cuyX7JHBaRSBQbGxsVFaVI5L6+vi4uLgkJCQMGDHj33XcrKytFIlFERIS9vX1ycrK3t7eXl9fJkyeJaMqUKd26dWuRQSAitnTI/v372eQdPz8/maiUCECJERg7dqyjo2NiYmJYWNjIkSNLS0vv37+voqIyduxYIvL398/MzAwICBg8eLC/v//06dPv37/fpUsXVnfz5s1Pnz5NTEzMzc3V0dGZOnXqq8cjv0prDBrU5+/vf+TIkdTU1NTU1Fu3brHkArsZE8txGBsbm5qa5uXlsUv8iCgzM5Nt1NbWUlMZvdjY2NjYWCJi84bOnz8vqS6D7e/SpYu+vn5JSQmbKiXx3nvvLVu27MWLFzNnzszIyFi3bp2Pjw/LnbEw+Hx+kyWbW5hxdXUlIjU1NUdHR5KaG8WwP8vLyydNmsQaJyKBQCA9zw6g48O8FQCA9qeuxjv9TTfvf2jGPK6KOPdiiV8X+eXnT3i5Cun7I3XZkivNanD7Z6Zvu2qkZFVfja1cvcBIsn/KMN1Fk/VVVWjZdmFMXNXod2Qv5K5PoM679qPlkukGhnr8iAsvrtypcLJWIyINdd71n6wWTNAXibmI8y+M9FVWLzDau7prC47DEr8uHm4auYW12w4VX4z+nxVelTu7Et2XjEDgFINsYU1YeGH4qVJRNTd1uO43B4pP3Si3s1D7bqnJnLF6U4bppD+vmbv2f9Z/XfuhUbawJj5V3MdRcPrbbuxinFeMR36tFh80UJq6uvqZM2e8vb1jYmL2798vuUCgQZJ1B95///0Gf9aW39qOHTvefvvtlJSUq1evrlmzRrJ/ypQpixYtUlVVDQ4Ojo6OHj16tCKRCwSC69evL1myxNDQMCIi4sqVK05OThoaGjdu3FiwYAHLsxgZGa1Zs+aXX35pwUH49NNPPTw8cnNzt23bxhZqkaZcAEqMAOt+YGBgdnZ2aGhoeHi4SCRiCa9vvvnm1KlTdnZ2W7dunTNnzpQpU9LT0wMCAiR1161bl52dHR8f36dPnzNnztSfBaBEPPKrtMagQX3h4eF5eXn379+/cOECW54mJyfn0aNHRHTmzJnk5GRVVdVr1679+9//3rt3LxEdPnw4Ozu7NSKJjIx8+vSpqqrqsWPHPvvss2+//Vb6KJtdRUQCgaBfv36enp6SQ+xavxkzZmzdutXPz09OyeYWZoKDgyMiIm7dumVubl5WVnbo0CHpo6dPn05MTNTW1l6xYoWnp+fIkSODg4MDAwORW4HOhdfkzE8AAGgQ+01GsmAngOJMRj4tKKk9tsl8wuAGvie/NvgeSaTANSYdEHt1d8bIoWMyNjYuKCg4fvz4hAkT2juWdtC5XlAKXgOVkJDg7Oz87bffLlmyZMWKFcOHD3dycjIyMiosLIyOjg4JCWG5FSIyMzNbu3btyJEjzczM0tPTIyIiNm3aVFVVJRQK2ay0EydO2Nvbs4VsraysGpuTQkT+/v7h4eHyyzg5Of344499+/ZNSUn5+eefv/vuOyIyMTERCoUqKipbt26dPn16TU3NpUuXNDU1J0yYsHfvXn9/f3t7+3379vXp00cgEOzdu3fevHmNlSSiZhVm3QwODh43blz//v2fPHny8ccfX79+XXKIjYCpqWlISMiYMWMsLS1fvHgRHx8fHh4uucG5fJ3lqQWvPeRWAACUhNwKKA25lQ6uc30VhI4PuRXqPC8oRXIr3bt3T0xMFAgEQUH+HrwrAAAgAElEQVRB0muyQrvoLE8teO3hmiAAAAAAAACFbN26NS4uTiAQFBcXHz16tL3DAYCOAgvUAQAAtLX8C3btHQIAtB2hUNjeIUCLcXBwqKiouH79+sqVK3Nycto7HADoKHBNEACAknBNEIB8uCYIAKizvaBwz+lOp7M8teC1h2uCAAAAAAAAAACUh9wKAAAAAAAAAIDykFsBAAAAAAAAAFAecisAAAAAAAAAAMpDbgUAAAAAAAAAQHnIrQAAAAAAAAAAKA+5FQAAAAAAAAAA5SG3AgAAAAAAAACgPORWAAAAAAAAAACUx+M4rr1jAADolHg8XnuHANAJdMZPGuzV3RkjB+iAOtcLCv+4dzqd5akFrz3V9g4AAAAAoCPCVywAAABQEHIrAABKwu8kAAAAAABAuCYIAAAAAAAAAOBVYC1bAAAAAAAAAADlIbcCAAAAAAAAAKA85FYAAAAAAAAAAJSH3AoAAAAAAAAAgPKQWwEAAAAAAAAAUB5yKwAAAAAAAAAAylNt7wAA4E3B4/FaqincPB4AAAAAADoOzFsBAAAAAAAAAFAecisAAAAAAAAAAMpDbgUAAAAAAAAAQHnIrQAAAAAAAAAAKA+5FQAAAAAAAAAA5SG3AgAAAAAAAACgPORWAKDdeHp6chzn7+8vs1FXV1dcXHz16lVPT8/2jhEAAAAAAKAJqm15Mh6P15anA4BO6uuvv46Li/vyyy/PnTvn4OCQn5/f3hE1AG9oAAriOK69Q2g2vMABWlzHfyvoLC98BUeys3RHjiZ72sH7qMgj1cG70HF0/DcQwrwVAOiA4uPjf/nll7lz5+rr6/v5+bV3OAAAAAAAAPK06bwVZtWqVW1/UgBod2vWrGlW+YSEBCKytrZunXBaRl20Q3uHANBx8T2S2juEV9IpfiUD6PjwyzwAvAkwbwUAOihnZ2ciSk9Pb+9AAAAAAAAA5EFuBQA6HFdX11mzZu3evbukpOTAgQPtHQ4AAAAAAIA87XBNEACAfEuXLi0rK7t3797s2bNzc3PbOxwAAAAAAAB5kFsBgHYTFRUluQa7/gYAAAAAAECngGuCAAAAAAAAAACUh9wKAAAAAAAAAIDykFsBAAAAAAAAAFAecisAAAAAAAAAAMpDbgUAAAAAAAAAQHnIrQAAAAAAAAAAKA+5FQAAAAAAAAAA5am2dwCvp0WLFpmamrLt6urqoqKie/fuRUVFcRwnOVRXV1dVVVVYWPj48eOYmJja2tr67Ziamr777rtmZmZqampPnjw5ePCg0iGNGTPG2trawMBARUWlqKjozp07MTExTQbPxMfHHzp0SImTzpw5097e/tq1a1euXFEm6A4mODhYS0vr4MGDT548efViAPI9SRO7TntGRHnn7YwNVExGPi0oqT22yXzCYJ12iadVA2iD3o1ZnHU+qmJFgOG6fxo166QKFpN5vFow8mbpIGFAZzR69Ojz58+vWLEiNDSU7TE2Ni4oKDh+/PiECRPkVFSw2JMnT1xcXIgoPz/f2Ni4BSNvlg4SRoMUHElq6MGCthQXF+fm5rZjx47AwEAisrCwWL9+/fDhw01MTEpLS1NSUvbs2fP999+3d5gA0NYwb6UVFRcXJyYmFhQUmJqajho1auDAgZJDZWVlubm5PB7P0tJy1KhRAQEBampq9Vvo37+/paVlYWHh2bNnY2NjFT+1iorsR+p+/fqpqaklJSXl5eWZmpr6+vq6u7vLaaGsrCzzbwUFBYqfuqXU70JnkZ2dnZmZWVlZ2VINdt6hkLGqEewo1xzt25F24e4i8HDTMNRrtydDuwfwKqprGnjOdOoeAbSg6urqBve7u7t7eHgYGhq2cTwAnctvv/02e/ZsdXX1yMjIlJQUZ2fnYcOGtXdQzRAXF8dx3Pbt2yV70tLSOI7bsGEDERUXFzf4SYz94svqchy3dOlSSfXz58+znV9//XVjp2h7bm5uv/76a0ZGhkgkysnJOXPmzODBg9khCwuLvXv3ZmVlicVioVAYHR29cOFCIrp69SrHcYcPH5Zu58aNGxzHRUREtGXwCo5zswiFQo7jmszkOjs7s7N0tGR0x4R5K60oMTHxzJkzRDRnzhxbW1sXF5cbN26wQ/Hx8WfOnOHxeO7u7mPHjrW0tPTy8rp69ap0dX9/fxsbGyIyMzMbM2ZMZGRkUlKStra2j4+Pvb29jo5OcXHxo0ePbt26VVNTw+ZKXL582dHRsVu3bhERESkpKdKt7d+/PzU1lYh4PN4///lPMzOzHj163Llzp7HgWYTSe7S0tNipdXV1CwsL//zzz/v377ND48ePt7Oz09LS4vP5RUVFMTExMTExkvkv3t7e3t7eeXl5O3fuXL16NRHt2bMnLS2tW7duCxYsIKLQ0NDGutDYSQcMGODu7q6np1dbW1tcXHzr1q0HDx7IdKGxui4uLtOmTausrNy5c2d5eXlAQICVlVVUVNS5c+dYDFeuXLGzszM3Ny8sLDxz5kxGRkb98Wmwy+xQt27dtLS0NDU16e85LJcvX7a1tbWwsCgpKfn9999Zg3LGs8lHEzoRNushdKHR1TuVMY+rundV/WG52YBeGg0WTsqo/nB9bkx8lZO1+sfTDKQP3UkQFZTUFpa+nODWZVhKSVndvz4yPhFZ9iBJ5P0Pzd0hZhv3Fu09XWqgy1/3T6PZvnpEJCyuXfvvwrN/lucIa+wt1T+ZbuA/Tq/JqDbvL/rhWElGXo26Ks/GXO3/ZhrM9tWTCSC/qHb1TwXnoyqeF9RYd1XzG627dEYXgTpPfsvzQ3Mv3a7IL6qtrSM7C7WPpuh/NPl/utnYAK77p9HlmIrYJyIHK7WtS01Zg3Jak4Rx5mb57XiRJLcSFl4YFl7oZqf+6IC1TI+UiE3O49Xk4DfYo8aqyH/I5IShXINFL+rWhxf+fr08/Xm1vg4/4P/prV9kLL81aEdsskNoaOilS5diY2MdHBy2bds2YMAAIpo3b96lS5fy8/Nra2vt7OwCAwM/+ugjSZWwsLDTp0/fvn379OnTS5Ysefz4MRGFhYWFhYW5ubnFxcXduXOnoKCgsLBQTlPyJSUlLViwICYmxsnJafHixdKH/j97dx4f0/n+j/+azCQzk31fyKaySSJaSxJkE4SiRS21i2j1pw3KV6kldlpKtSgtb4KKPahKitIEQRJbEJFmIatIMtnXmUwyvz9OTeeTTCZjkEVfz4c/jjP3fZ/rPmdmMuc6932OQCBYvXr1H3/8kZeXZ2dnN3/+/MDAwBZ71FwtaY/++uuv+Ph4a2vr3bt3M1VUC0NxgyUlJevXr//999+zsrL09PRmzJjxzTffKOiRkserkcLCwpUrV164cOH58+c2NjaTJ09euHAhl8t1dXVterBaPBbwhvB4PObw+fv7P3z4kIg0NDSYH/BEZGJisnr16iFDhpibm2dmZoaFhW3evFkoFBKRQCAwMjJatmyZv7+/u7t7VlbWrFmzbty40XZdke/WrVs6OjpE1KNHDx6P9/z588zMTCJKTU2VLTZnzpytW7fW19e7uLgEBAS0TazNGzp06JkzZ7hcblVV1a1btzQ0NHx8fJ48eXLlyhUiOnbsWP/+/QUCQXR0tJ6eHpMd27Vr1759+/z8/D744AMDA4OSkhIi6tq1K3OxfN++fW3Skde1n+Ve1IdXh3Erb5yurq6uri4RMd+ksiQSya1bt5jTZmdn50avJiUlCQQCIsrPz4+Njc3JyeFwOEFBQX369GloaHj48KGWlpa/v79sutHf318sFiclJYlEokatMYkVImKxWBwOh4gUj0Zxdnb+5AVnZ2c2mx0UFOTu7l5VVXX9+nUejzdq1KhevXoxhc3MzDIyMmJiYhISEgwNDYcNG/bOO+88ePCgtLSUiHJzc2NjYx88eCDb8ea2K9uF5jZqaWkZEBCgoaFx7dq1+Pj4srKyplfVFAT8+PHj+Ph4Pp8/cuRILy8vKyurZ8+e/fnnn9K6/fr1y8jISE9Pt7CwmDJlipaWVtM45Xa5uU75+voWFhYWFhaampqOHj1acXhyd0XzBwo6hrV7i527aDh30Xj0RBS4+rncMnViyfAvc6Pv1pgacJxsNOZtKWyx2RW/FOnrqHE1WBfjqrtPzIy4XuXcRSM7XzxrQ0GeQCyqk/h8lrPjRKmZIfurqYYlFfVBa/P3nClTHFXco9qvtgkqaxqWBBoGj9O3seCk5zS+pi0USbxn5ewKL2Or0cQAnYKS+pCfi2asyW+xvw/SRH49Nb+ebhg4XDc1SzTnu8LLt6qV2YGbD5X49tIc4ql592/h+/NyC0rqlWkt5OcinobaWH/tkT5aNuYcInJ34c37WH/K+3IyAi8bm+Lj1eLOb9qjFqvI3bEKwlCtQVGdxPez7C1hJYLS+gkBOl49+ElPREq2Bm3ou+++8/PzGzJkyN27d4cOHVpQUEBEDx488PPzW7JkSWBgYGpqanBw8OXLl6VVli9fzuPxxo4dq62tPXXqVBsbGyJyd3efN2/e1KlTG7WvuCm56urqhg0bFh0dbWpq6uTkNHfuXOlLIpHI29t7x44dZmZmixYtKikpmTFjxp49exT3qMVaa9ascXZ2dnZ2fvTo0fTp0189DLkNikQiHx+fLVu2CASCCRMmeHl5MX+pW+xRi8dLllAo9PLy2rVrF5vNnjhxYkFBwfLly5lkTYsHC1pTbW1tbm4uEYWFhW3YsGHkyJGampopKSlExOVyY2JiZs+eXV9ff+TIEVNT03Xr1u3fv1+2+ooVK5KSkpKSklxcXA4cONAmXVBs8ODBnp6enp6eTErl2LFjzH9DQkKkZYqLi62trceMGUNECxYsoJbOMloZl8sNDQ3lcrkPHz50cHDw8vJyd3fv3LnzwYMH6f9mxwICAjw8PExMTJYuXUpEJ0+eLC8v53K5EyZMYJpivgcyMzNb/AJ8E1rcz3v37s3MzKyurhYKhY8fP5bNgDOjVJYuXXr9+nWRSOTn5yd9icVi7dmzRyKRPH782NLSkojs7e2joqKqqqru3r3r4eEhuwkTE5OdO3emp6dXVVUlJSUtW7aMy+US0RdffCGRSJi7QNjZ2UkkkvLycjabzWKxCgsLJRKJt7e3bBiXLl0qLy9PTEyUm1nuuDBu5Q1yd3d3d3dnliUSSXOp6KKioq5du2prN57JHx8fb2lpaWxsnJ2dff78eSLq1q2bkZFRQ0NDaGhoZWVlcnLypEmTXF1dL1y4wFRJSEj47bffFITEZrM/+OADIyOj58+fX7t2TUFJbW1taUja2toODg7GxsZ1dXXh4eEikaisrIxJTDAzlfbt2ycWi1kslrq6uo6OjoODwzvvvHPp0iVbW1t9ff309PRG91tRkFuR7UK3bt3kbvT3338nIrFYXFpaWlRUdPXq1abjmRUHfPHiRSsrKzs7u65duwqFwpMnT8re7yYmJiYmJobFYs2bN09fX9/FxaXpvWnkdvnJkydyO3X9+vWoqChmnI6hoSGPx+vSpYuC8JruCujoFk81WDXL6PbjWvfA7LScutKKBn0dtav3ahZvFzAFhvXX8uzOS8upU1OjK79YWplxQs+Vz1ybr7jZ+RP1vw02Xr6raMP+4lqhJO6kla6Wmp5/ulAkeZAmqq5tSM4Q8bmssLUWWjyWlRnnk3X53x4o+XSUnoKoKqoaiIirzrK14DhYaywLMtTksRptN/JGVUqWiMNmXd1tZWbIHuVX9cGCZ0f/rNjypbGClono6m5LngaroYFqhA3PBOJzMVWX4qsH9tFscQd+Pd1w8TQDiYS6js7IyKs7fqkieJx+i63NGKH7v+VmzPL783Izn4sHu2tK77fSyMvGFn23RsHxirhepXjnN+1RZxOO4ipyd+ytx7XNhdFiDHIbjL5bnZguYrEoZo+lk60GvZhRpUxr0IaWLFmyePFiiUTyzjvvZGRkHD9+PDg4+Nq1azwer6Ghoaam5tmzZ+fOnfvzzz+lsxVmzJixd+9eZrlv375RUVGZmZmDBw+WewsPxU3JFR0dnZaWpqamdvXqVSsrq9DQ0KCgIOaliIiI5ORkPp9/+PBhLS0tKyurmTNnfvPNN59++qmCHnXu3FlxrcWLF69evfr27dt9+vRJS0srLS3V19d/lTDkNhgZGZmYmMhisa5fv+7k5EREdXV1586da7FHLR4v2QKRkZEpKSkcDufatWtmZmajR48eMWLE0aNHv//++8WLFys+WNDKxo0b98svv7i5uXXv3p2IRCLRmjVr1q9fP2zYMAcHB7FY7O3tnZ+ff/r06XPnzk2YMGHBggV5eXlM3Y0bN65cubJ37963bt2ys7PT19dnLky2vrFjx0pvF2BhYfFSdUNDQ+fMmTN//vzo6OjJkyffuXNHIBAMGTLkDYSpCk9PT3NzcyIKCQl59uwZs7KsrIz5ec9kxywtLcPCws6dOxcXF3flyhUmO1ZTU3P06NFZs2ZNnz59165dLBaLSWXu37+/Teant7if3dzcmG88S0vLmTNn7tixIzk5WTYNtG7duqioqJMnT1ZWVjJrmMTKzJkz7927N2TIkMLCQnV19cjISDs7u4yMjOTk5G3btkmrM+lCBweH1NTUI0eOjBo1at26da6urhMnTrx06RIReXp6MmOCiEhHR6dnz541NTXGxsZVVVWxsbHSdlasWLF7925tbW0PD48DBw7Y29u/6V3XapBbeYNKS0sLCgrq6upKS0vv37/f9KIEw8jIiIikb3EF9PT0iKi6upopLG2QGRdDRM2d2zM0NTXHjx9va2ubmZl55MiRpuNoZMXHx8vOCfL09CQidXX1L7/8UrpSX1+fw+HU19f7+fm5ubnp6OiwWP+cgzVNFTWlpiZn2JRsF5j+Nt1oTk5OQkKCm5sbMwZELBZfvnz55s2bsu00V5fD4YjFYrFYfOPGjTFjxrBYrIcPHzIDnqWYUX8SiYT5FcU0JYvFYr1Ul/Pz80lm4JKGhobi8JruCujouttziUhP+59be1TWNOjrqBWX18c9qmXWONlqdDblEJGRHtvKjENEPR25LTbbqxuPiIz02UT0Tmd15t6l2nw1oai+oroht1BMRDVCSZeRT6VVMvLqakUSBVEN6K05fbhu2PmKwNX5RMTTYK2bbbRgkoHsdrPyxURkrK9mZsgmIteu/4SanS9W0HJDA63eU3Toj4q8InFDwz9N5RfLuY13U106qRMRi0W2FpyMvLrs52JlWhvo3nLWhqFCbExnmztezC5SsPOb9qi+gZqrwtP453um6Y5VEIaCGBQ0mPmcObhsJrFCROoclpKtQRvq0qULEbFYLFtb24yMjKysrIaGhlWrVh06dCgvL6/hxdua+XvEGDRokJKNt9iUXMwEWCMjIysrKyLq2bOn9KWsrCwiqqmpkU6dIKKMjIza2loej9dcj5irIHJrMctubm704gcAEVVWVurr66sWhoIGmav3xsbGTGKFiNTV1ZXpkaymvWtUgFljbGxsZmZGRK6urtK9+rLnvfCmxcbG9ujRo0uXLh4eHpMnTx4xYsS6deuOHz9ubW1NRAKBgPmwSKduWVlZSXMrzLDusrJ/xgAyU/7boA9EZmZmzJtNBc+fPz9y5Mj06dOZ4SFbt25tV8OpmMQKNZnHJNVcdoyI9u7dO2vWLA8PDycnJ3Nzc1tbW4lEEhoa2mrBy2pxP3t7e9fW1qqpqfH5/E6dOo0YMWLw4MGyuZXQ0NCZM2fKVtmwYYOTk9P169eHDx/OvA+ZWxY0NDT4+PhkZ2fPmDFDOgFKQbrw77//zsnJsbS07NOnj6+vb3p6urW1ta+vL3MDymvXrsleCG8/KcXXDrmVN0h6v5XmsFisXr16de3alYiSkpJabJB5x2tqamppaVVVVZmYmDDry8vLmYXmbkdHRKamphMnTjQwMLh7925ERITcxxK1uGmRSHT8+HHp7yoOh9PQ0NCtWzcvLy+hUHjq1KmysjJ/f3/pDwsmpyvNPjAtaGhoMPcikfsNLtuF5jZaX19/5syZiIgIExMTe3v7AQMG+Pn5NcqtKAiYiDQ1NQcPHkxEYrG4Z8+e9+/fl72pirm5+aNHjzgcDnPTJunulVLQZbmYjcpmuBWH13RXQEfHViMiYv3fM9BRvtoNcf+m6i/GVRNRSXlDWWWDnrba02ctvwHUOf82y5wAy7Iy5RCRFl/t5LcW7BeZTGGdhPPi5q1yo1JjUegKs52LTZOeiiKuV63aXbR6T3Gj3ArTsqC0oaCk3tSA/ejJP3lDSzOOgpZPRVduPFiio6n262pzKzPOip+Lou/WKHnh50GqcPwgbaFI8jhDxGxImdb43H8jYKuxiEjmE/aqsVmacqj549Xizm/ao84mLVQheTtWQRgtxiC3QWsz5uDWp2SJHKz/GbeizmEp0xq0ofv3748fP54ZB05EVlZWp06d2rhxo46OzqFDh6ysrEJCQqKjo2X/EjF/i6WY+6Y3yPuQtNiUXMzY8pKSkrKyMj09PencZCY8ItLS0goPD5fer10oFDJzlpvrUefOnRXXYlay/u9Xz6uEIbdB6QlzSkqKg4MDEdXV1SnTI1lNe9eoALNGIBAwjyBgbrAi7Y6CgwWtLzAw8OTJk0+fPn369OmNGzdGjBhBRIaGhsxvS2NjY1NT04KCAhcXF6Z8Tk6OtC7zg7w93KRf+tgjIsrIyGDmnSnv+++/nz59+rBhw3Jzc48fP96ucivSRLC9vb3cE67msmOpqanx8fGJiYmurq7Tp09n0pqXL19mEqxtQsF+VlNTW7Vq1ZQpUywsLKQXsBudbTGjS2QxOeJ9+/ZJE3zMl09RURHzBr579660sOJ04eXLl6dPn+7j4+Pr6xsZGdmjRw9fX18mVd1ou+0npfjaIbfSNpydna2srPT09DQ1NYkoJyfn+vXrLdZKTU0tLi42NDScMWNGVlYW82F49OhRRUVFi3WDgoJ4PF5dXZ2uru7EiROJqKio6I8//lAy4NTU1KKiIiMjI29v7/T0dC6Xa25uLpFIUlJSpJ9eNpvdqVMn5q8+g/mcuLm5aWho5OTkPHz4MC8vz8bGxt/f38rKSvGDihRstLa2dvjw4RkZGZWVlcxXRlVVlfIBE9GoUaN0dXXv3buXmZk5atSosWPH/vzzz9In+/Tt29fAwMDIyEhbW1skEkl/0Egp6PKr78+XbQreGn49+e90Vn+SWzdmcd6Qvpo7T7zqn5nh/bUcrDVSskQbQosHe2hWVDUkpAjZbBreX84thKRuPKj5YlOhb0++uTE7MU1IRKaGjc+eh/XXsrNUT8up8/0sx6sH77erVUQ0bqB2J2NFf1PE9f/8dhSKJHeShbGJtcr3Zcvhkie5dSlZovziem2+2vhBOlF3ql+qNRsLDhGFXSivrGnwcOVNGqLzirEpPl4t7vymPdLXVlPheCkIQ7U3wLB+Wt1sNR5niPp/kvOht1aNUCKsk4RvtFCtNWg1W7ZsefLkSUpKSn5+vra29vjx46U3yBcKhXfu3JEdjy0XcyoVFhZWWVnp6ek5adIk6UvSAZVKNsXw8/NjZst+9NFHQ4cO/emnn6QvDR8+3MHBISUlZf369QEBAeXl5QkJCWw2e/jw4Qp6pK+v32Kt1x5GU8OGDevWrdvjx4/79ev34Ycf1tTUCIXCI0eOvFRTTXvXdCt2dnZpaWk+Pj5eXl7MHOFx48Z16tSJFB4saH2hoaE7d+5MSUkpKCjo0aMHEeXl5T18+PD+/ftpaWl2dnZXr16NiYkZOXIkEZ04cUI6LeVt8uDBg8uXLw8cOHD79u3t7ergzZs3nz9/bm5uvnbt2vj4eGbQkJ6enqOjIzMtqLnsGFM9NDR0y5Yt06ZNYyYKtNVdbBkK9vNHH320ePHiioqKKVOmZGdnr1271s/Pr1FquOljTENDQ6dOnfrLL7+UlZWFh4fTi9yfgYGBnp5eWVkZM8iOoThdyORWJk+ebGNjc+XKldLS0tmzZzN3jWx0e5r2k1J87XAv27ahra3NJAVycnLOnz8fGhqqzP1KxWLxvn377ty5w+FwunfvXl1dHRUVdfr0aWW2qKGhQUTq6up2LzS9SNLipuPj43V0dHx8fN577z0Oh8NkK5OSkm7dutXQ0DB48ODOnTunpaVJa928eTMnJ0dLS8vDw4MZnhMREZGbm2tgYNClS5fo6GjVNlpZWVldXe3q6jpgwIAuXbo8efKk0dPRFAfct29fBweHkpKSP/74IyEh4dGjR3p6eswfPEZUVJSOjo6Jicnz58+ZHy6NGlfQ5Vffn/CfpaHOitjayec9fvyj2sPnK+b/36EiKuBqsK7utgwep/9MIF4fWhx6rlxYJxk/SEdxLTMjjrE+++ifFSt/Kb58q3qQu+axDY3Hn/M0WNf2WM0apScUSQ5fqDDSY6+aZXRglbnilscN1Pl8rB6HTYt3COITa4f2fYlz8jWfGT0TiJOeit514Eb80MnMkP2yrc2fZODhwssvrt9+vPTPuMY3qVUhNsXHq8Wd37RHqh0vBWGo1iBTa/5EfUNdtcMXK/66Xe1oo65ya9Bq1q5d++zZs6SkpHfffTcyMtLMzGzcuHGff/45h8NZtGhRXFzc0KFDFbewYMECDw+P/Pz87du3X7x4Ufall22KoaGhERkZ6ePjEx8fHxYWxtx5kcHlcq9duxYcHPzs2bN169aFhoYKhcKPP/5YcY+UqfXaw2iKqTV//nxDQ8PDhw//9ddfjo6OL9tU0941KsDj8WJiYmbNmiUUCg8fPmxkZLR69Wrm1puk8GBBK2DGDUnPbJcvXx4XF2dqaurr69vQ0PDbb78NGTKkurq6trbWy8tr9+7dXC530qRJRUVFK1eunDZtWpvG/gYFBQUNHjxYNn3ZyNixY2NfaM1pNUKhMCgoSCgUdu/ePTU1NSYmJi4uLjc3V3osQkNDCwoKEhISLl68eOvWLXqRHWNe/fXXX+vq6jp16sQMr/ru2wUAACAASURBVFDyzOvNaW4/S0fJcbncXr16MfdzaNHZs2eDgoLYbPaRI0c+/PBDIoqOjn7y5AmHwzl16tRXX331ww8/SAtHRkampaVxOJyrV6/+73//Y+6+LE0XMoNTmITL1atXr1y5oq+vb2pqKhAImj7O9W3Fas2MEZM5W7lyZattEeClMA8/Pnr0aHJyclvH8h+yevVq6oDZa+YLTXZSD7wJzKOCT22yGOXb8l2cOoS3r0cKqHmkUgf8dNOLD3iHiJx5pu/p06dlnxvYob19PZL1dveuOR3lA9XoIr9c1tbWKSkpXC53zpw5O3bsaIWomlJyTyrTHSJKTEx0cXFpOido48aNX3/9tbRYcnKyo6Pjjz/+KHuvQKbuV199tXnzZtk2z58/P2TIkC1btixcuFBaTLbAnTt3WhzATkr0VMk+EpGrq+vixYsHDBhgampaXFx89+7dTZs2MRd6ly1bNmjQIEdHRyMjo+Li4ri4uJCQEGluhYjCw8M/+ugjItq5c6cyj59XPn4lu6DMfmaz2du2bZs4caJYLL506RKfzx81atSBAweY54sxz/wePXr0mTNnmLqya4KDg7dv3y4SiUaNGvXHH384Ojru3r27Z8+e6enp+/bt+/HHH4nIxMREIBCYmZmtWbMmICDAzMwsKyvr8OHDmzZtkt6jKikpqVu3bmlpafb29tra2iUlJRwO5/jx49JEs+xG7ezsmDvgWFlZyU6Ue5Wd2eaQWwH4F3IrbQK5FVDg7ctEvH09UgC5lVbw9p2rv309kvV29645HeUD1eJZ7rZt2wIDA3V0dEpLS52dnaW3pG1lrze30p69xtxKm3hduRWgjvAFQpgTBAAAAAAA0CJ7e/vq6urIyEh/f/+2SqwAQLuFcSsA0MYwbgXgbYVxKwBAHecD1VFGEGDcilQ77yPGrbxG7f8LhDBuBQAAAAAAAADgVSC3AgAAAAAAAACgOuRWAAAAAAAAAABUh9wKAAAAAAAAAIDqkFsBAAAAAAAAAFAdcisAAAAAAAAAAKpDbgUAAAAAAAAAQHXIrQAAAAAAAAAAqA65FQAAAAAAAAAA1SG3AgAAAAAAAACgOuRWAAAAAAAAAABUx5JIJK23MRar1bYFAB1La34XvRb4QgNQUof7dBM+4ABvQPv/KugoH3wl92RH6Y4CLfa0nfdRmSPVzrvQfrT/LxDCuBUAAAAAAAAAgFfRquNWAAAAAAAAAADeMhi3AgAAAAAAAACgOuRWAAAAAAAAAABUh9wKAAAAAAAAAIDqkFsBAAAAAAAAAFAdcisAAAAAAAAAAKpDbgUAAAAAAAAAQHXIrQAAAAAAAAAAqA65FQAAAAAAAAAA1SG3AgAAAAAAAACgOuRWAAAAAAAAAABUx2nrAADgv4LFYr2upiQSyetqCgAAAAAA4BVh3AoAAAAAAAAAgOqQWwEAAAAAAAAAUB1yKwAAAAAAAAAAqkNuBQAAAAAAAABAda16L9vXeCdLAAAAAAAAAID2AONWAAAAAAAAAABU1wbPYF65cmXrbxQA2tzq1asbrfH09Lx58+aMGTOSk5NlFyQSSXl5+b1795YsWRIbG9sm0QIAAAAAACipDXIrAACKbd68OTExccWKFefPn7e3ty8sLGzriOTAJEcAJUkkkrYO4eXg0w3w2nW47wEAgJeF3AoAtDtJSUkHDx7MyMi4cuXKpEmTfvzxx7aOCAAAAN5yHSWvqmSiqqN0RwHFPW3nHezQwUNTynzukFsBgHbq8ePHRGRjY9PWgSjSEGff1iEAtF9qHqltHYLqcJkd4LXAOSQA/EfgXrYA0E45OTkRUVZWVlsHAgAAAAAAoAhyKwDQ7jg7O0+dOnXv3r1lZWVHjhxp63AAAAAAAAAUwZwgAGh3Fi5cWFlZee/evWnTpuXn57d1OAAAAAAAAIogtwIAbSY2NlY6DbvpAgAAAAAAQIeAOUEAAAAAAAAAAKpDbgUAAAAAAAAAQHXIrQAAAAAAAAAAqA65FQAAAAAAAAAA1SG3AgAAAAAAAACgOjwnqL37/PPPTU1NZdckJSUdP35cbuFFixZpamoePXo0OTlZyfaNjY2Dg4OJaNOmTdXV1a8Y7UthuhYfHx8ZGcms+fLLL/X19WNiYi5duqRkI1OmTLGzs7t69epff/2lQmFl9phqtQAAAAAAAOA/ArmVjqGysrK0tJRZLioqalqAzWbX19c/e/aMx+PV1NS0bnQdmGp7DPsZAAAAAAAApJBb6RiSkpKkgzsYzNCJy5cvOzg4dOrU6fDhw+np6Z06ddLU1OTz+USkqanp5+dnZ2eno6NTXFx88+bNhIQEpq6RkdEHH3zQuXNngUAQFxcn2+zXX3/N4/EuXbrk5ORkbm6emZn522+/eXl59ejRo7a29q+//rp//z4RjRw58p133tHU1FRTUyspKYmPj4+Pj28UWJcuXTp37lxWVvb7779nZ2e/+k7o169f7969dXV16+vrS0tLb9y40b9/f2ZQj4+Pj4+PT0FBwc6dO5uLTToCSLaw7B5r2v79+/dbrEVEfD7f29vb0dFRT0+vtrY2ISFB+XE3AAAAAAAA0NHhfisdg7Oz8ycvODs7S9f7+/uLxeKkpCSRSCRbns1mBwUFubu7V1VVXb9+ncfjjRo1qlevXsxLkydPtrW1raqqEggE77//ftPNDRgwoLa2ViwWd+3a9fPPP7e3ty8sLNTT0/vwww91dHSIyMzMLCMjIyYmJiEhwdDQcNiwYe+8845sC76+voWFhYWFhaampqNHj1aya0zjcllaWgYEBGhoaFy7di0+Pr6srMzQ0PDBgwfMcJ7c3NzY2NgHDx4oiE1uYcXtt1iL2Z8zZszo16+fpqZmYmJiVlaWiYmJgv7+l61sBvOq5GW0bUfeDt0nZqp5pMr+G/d1ntySJgFP1DxSz1ypVL7x5AwR06agtP41xasspl9zviuUruky8qmaR+qSnwTKN/L+vFw1j9SQn+UMElSmsDJ7TLVaAK/i0aNHffv21dLSYrFYo0aNUqGFoUOHslis5cuXv/bYlGRsbMxisc6cOaO4WCvH2XRzSsYJAABvE+RWOgZtbW3LF7S1taXrExISDhw4EB4e3mhgiIODg7GxcV1dXXh4eHx8fHR0NBF5eXkRka2traGhoUQiCQ0NDQ8PP3/+fNPNxcbGHjp06NatW0TE4XD27NkTGhoqFovZbLaZmRkR7du37/Tp01evXr1w4UJaWhoRNcqtXL9+/Y8//mDG2hgaGvJ4PGW6xmazmZU2NjbShIuvry8RaWhoEJFYLC4tLf37779PnDgRFRUVExMjEAiIKD09/fz58zExMQpik1tYSm77LdYiInt7e1NTU4lEsnfv3jNnzhw/fry5u+EAtENmhmwPFx7zz8Fao2mBOrGkdzeuhwvPUJfd+uF1UKrtMexneNO2b98eGxtrZ2e3bdu2WbNmtX4AdXV1r9hC7969PTw8mIsfb8irB0mtEie0ocTERIlEsmPHDua/nTt3PnDgQG5urkgkYoaEz549u20jfCmNukNEGRkZEonk22+/JaLS0lK5V7mOHj0qrSuRSBYuXCitfuHCBWbl5s2bm9tEK3Nxcfn111+zs7OFQmFeXl5kZCRzfkHNH76oqCiJRHLixAnZdmJiYiQSyeHDh1st8o7yZnNycmIOurGxsYJi58+fl0gk69atU6bNlyosEAgkEoncywYKXnrtkFvpGOLj41e9IJ19Q0RPnjyRW15PT4+I1NXVv/zyy0WLFo0cOZKI9PX1ORyOrq4uEVVXV5eVlRFRXp6cK9XPnj1jyhBRSUlJdXV1fX09MzRGQ0ODxWL5+fktWLBgxYoVS5cudXBwICLZjA8R5efnE5FQKGT+y2QuWuya9J4yfD5fmnAxMDAgooyMjISEBF1d3dGjR3/yySeLFy/u27dv09aUiU0uJdtvSl9fn9lXTAqGiOrrW/tCPYDKxvrr3Nxnxfxb/7kRvRg9sWF/sden2dz+adF3a24/FsY9qi0uryciQWn93M2F9mMytH3T3p2ctf9cubSp1Ow6/9k52r5pvaZlxT2qld2KwcB0NY/UjQdL+gZla3qnDZ2bm1sonru5UG9Aus2HTw9G/tvIJ+vybUc+1fJJ43mlOX+c+dPJf74TpFENDs7VG5DefWLmjQf/ZxMq2xJW4jAmg++dpjcgvcekrIOR5d0nZl6IrSai9aHFah6p3SdmKohNbmHZPab8JhrVKqlo+GqbwGlcpqZ3msX7T5bufIlxNwBN+fn5/fLLL0T04MGDuXPnxsfHCwSCOXPm2NnZaWlp9ejRY//+/dLCM2fOtLGx0dTU5HK53bp1++mnn4jI1dX1woULRLR+/XoWi+Xq6kpELBaLxWJduXKFiG7fvs38t7b2n48nM3xjw4YN/fv319DQiI6OVrDRLVu22Nvb83g8XV1dNze3gwcPNu3F7du34+LiiouLZRsfNGiQrq6uq6vrjRs3motTwXabBim3+4ySkpKFCxc6Ojry+Xxzc/MlS5bI3ZxsnERUWFj4+eefd+3aVUtLy9nZef369dIfSM31AjqQY8eOTZs2jXnzpKenOzk5DRw4sK2Dem1u3boVFxcXFxfHfK6fP3/O/Dc1NVW22Jw5c5hLpC4uLgEBAW0TazOGDh16586dKVOmGBgY3Lp1Kzs728fHZ9y4ccyrzR2+ffv2EdEHH3zAnIkQUdeuXfv37y99qU20zptt5MiRiYmJzAV1eCnIrXRszV1dYfImIpHo0KFDB184cuRIQ0NDeXk5EfH5fGYsCZMaaITJDjAzL5pmCrp16+bl5cXlck+dOrVv376MjIymLTQ0NEhbUEFycrI04cIMqZVIJGfOnPnmm292794dFRXF4XD8/Pykm2CxWMrE1qhwo5fktq+4Fr3Y1ZqamkZGRswa6egbgPbv5F8VfYOymX8n//p3NkrIz0U8DbWx/tra/H//TIjqJD6f5ew4UWpmyP5qqmFJRX3Q2vw9Z8qIqE4sGf5lbvTdGlMDjpONxrwthU23teKXIn0dNa4G62JcdfeJmRHXq5y7aGTni2dtKMgTiJkyD9JEfj01v55uGDhcNzVLNOe7wsu3/n1+2dq9xc5dNJy7aDx6Igpc/VzJfuUVNZvujHtU+9U2QWVNw5JAw+Bx+jYWnPScuinv69qYc4jI3YU372P9Ke/rKoitucKqbUJ2V/t+lr0lrERQWj8hQMerBz/piYgAXsHYsWOdnJyIqHv37vPmzevZs6e3t/eOHTvMzMwWLVpUUlIyY8aMPXv2MIUfPHjg5+e3ZMmSwMDA1NTU4ODgy5cvT5061cbGhojc3d3nzZs3depUaePN/YlkLF++nMfjjR07VkNDo7mNxsXFLVy4sLKycunSpcHBwTY2NszI0xatWbPG2dnZ2dn50aNH06dPJ6KmcYpEIgWdbRSktra23O4TkUgk8vHx2bJli0AgmDBhgpeXV1JSkoLdwhAKhV5eXrt27WKz2RMnTiwoKFi+fHlgYKDiXkBHwePx+vXrR0T+/v4BAQEeHh4mJiZLly5lXjUxMdm5c2d6enpVVVVSUtKyZcu4XC7zEnMtfenSpZcuXSovL09MTGTaaW8GDx7s6enp6emZmZlJRMeOHWP+GxISIi1TXFxsbW09ZswYIlqwYAE18+SNNsHlckNDQ7lc7sOHDx0cHLy8vNzd3Tt37sxkbxUcvpMnT5aXl3O53AkTJjBNMZ/NzMxM5guh9bXam83AwMDFxUVdXV1xPPb29lFRUVVVVXfv3vXw8JB9ae/evZmZmdXV1UKh8PHjx1988QURJSYmDhkyhIiWLVsmkUgSExNftrDckrK6du36119/VVdX37lzR24fjY2Nt2/fnpaWVlVVdf/+/UZfxa8O97J9O6WmphYVFRkZGXl7e6enp3O5XHNzc4lEkpKSkpGRUVJSYmBg8PHHH6elpfXp0+dlG1dT++dci81md+rUydLS8nWHL4eVldXw4cMzMjIqKyuZNGpVVRURMUNd3NzcNDQ0cnJypNkcubE1Kvzw4cMW21dci4hSU1MLCwtNTExmzpz5999/q6urs9nsY8eOvbE9AfA65RfX5xf/k3p4XiSWrp8xQvd/yxtfr4i4XpWcIeJzWWFrLbR4LCszzifr8r89UPLpKL3ouzVpOXVqanTlF0srM07oufKZa/MbVZ8/Uf/bYOPlu4o27C+uFUriTlrpaqnp+acLRZIHaSILYw4RXd1tydNgNTRQjbDhmUB8LqbqUnz1wD6aTAuLpxqsmmV0+3Gte2B2Wk5daUWDvo78KwSy/ZJ19V7N4u3/DAAZ1l+rb3ceEXHVWbYWHAdrjWVBhpo8FhFF36nOfC4e7K659v8z+reuvNi++cJYbmGpiqoG5TchFXmjKjFdxGJRzB5LJ1sNIqoT4zZD8EqCg4NjY2OTk5P79ev3ww8/nD59Ojk5mc/nHz58WEtLy8rKaubMmd98882nn35KRNeuXePxeA0NDTU1Nc+ePTt37tyff/757bffRkVFZWZmDh48uNEgbekPA7lmzJixd+9eIlKw0YqKCiLicrm2trYODg7Lly/X1NRUpl+LFy9evXr17du3+/Tpk5aWVlpaunjx4kZxKu5soyCb6/7AgQMjIyMTExNZLNb169eZRFVdXZ26unpzu4URGRmZkpLC4XCuXbtmZmY2evToESNGHD169Pvvv7ewsGiuF3IvfUE7VFtbm5uba2lpGRYWdu7cubi4uCtXrqSkpBARl8uNiYlxcHBITU09cuTIqFGj1q1b5+rqOnHiRGn1FStW7N69W1tb28PD48CBA/b29m3VkbFjx/bu3ZtZlr4zlRQaGjpnzpz58+dHR0dPnjz5zp07AoGAOTFuc56enubm5kQUEhLCDMwnorKyMmYegILDV1NTc/To0VmzZk2fPn3Xrl0sFovJnO7fv7+tbvz3pt9sxsbGzGB/Zl6PpaUlh8MhosLCQulpkZS6unpkZKSdnV1GRkZycvK2bdtkX3Vzc4uOjk5LS7O0tJw5c+aOHTuSk5N//fXX2bNn29jYxMfH37x5Uzp5QvnCckvKprrWr19/9uzZv//+u2fPnufPn7ezsysoKJC+ytxY08nJ6caNG7/++uvMmTNDQ0PV1dUbpdpfBXIrbyexWLxv3z5fX187OzsfHx+RSFRYWHjv3j0iqq+vDwsLY54TxOfzb968Kfd2tgokJSXdunXL1dV18ODBT548SUtLY35hvFGVlZXV1dWurq58Pl8oFD558uTPP/8kops3b1pYWJibm3t4eHC53LNnzyqIrVFh2SxJc+0rrkVEYrE4NDTU29vbwcGhe/fuQqGQ2c8AHcIXY/W3fyXn7ssD3eWc1WTli4moRijpMvKpdGVGXl2tSJKdLyYiIz22lRmHiHo6cptW79WNR0RG+mwieqezurE+m4i0+WpCUX1FdQMRNTTQ6j1Fh/6oyCsSNzT8U0s2RdLdnktEetr/DA2rrGk2tyLbry4jn2Y+/ydtVFxeL52v5GSrsXSG4fThumHnKwJX5xMRT4O1brbRgkkGTRtsMbbmDOitqeQmZDEBG+uzmcQKEalzFI0LAHhZWVlZRFRTU2NraytdmZGRUVtbq6GhsWrVqkOHDuXl5TW8eLszU30VE4vFctcPGjSoxY0OGDBg+vTpYWFhzJVhHo+3fv165gK4Ym5ubvRiKjQRVVZWNs1KKNiu9H5w0iAbGhqa6z5z3d7Y2Fj606LF67rSrRsbGzNXbphJQ0SUnZ0tPYNVphfQbo0bN+6XX35xc3Pr3r07EYlEojVr1qxfv37YsGEODg5isdjb2zs/P//06dPnzp2bMGHCggULpGeVGzduXLlyZe/evW/dumVnZ6evry+dIN/KzMzMVJ4D8vz58yNHjkyfPp0ZIbJ169amA7jaCpNYIaJGk5ikmjt8RLR3795Zs2Z5eHgwj0+1tbVlbljZasErH+1rebP98MMPkydPlv735s2bzMLUqVMPHTrUKBLmcbQNDQ0+Pj7Z2dkzZsyQnSrl7e1dW1urpqbG5/M7deo0YsSIwYMHf/311wMGDLCxsfnzzz9lb/6tfGG5JWVzK999911ISIiGhkZGRoaFhcX48eNl7/IzfPhwJyenmpqaSZMmVVVVZWdn7927d8mSJcit/Ifs3LlT7vpNmzYpXllVVdXosc1SAoFA9ntB9jHMzG2rGLGxsbGxsXIbj4iIiIiIaDGw4uLiVatWyS1G8rr2ww8/NFe4uLj4wIEDctf/73//k12jILamhWWjldt+i7WIqLq6+sKFC8x0a4C3A58r50zeypRDRFp8tZPfWrBf5DSEdRIOmyxNOURUUt5QVtmgp6329Jmc6YrqHCIiZuqA3EzBqejKjQdLdDTVfl1tbmXGWfFzUfTdGtmLQ8xGFU4+aMEoX+2GuH8v1EgkFLrCbOdi06SnoojrVat2F63eU7xgkgFbjUVE0hyK4tiaFpalxlJ2E7KszThEJCitT8kSMfcYrhNLkF6B18jKyoqItLS0wsPDpVNZhUIhh8M5derUxo0bdXR0Dh06ZGVlFRISEh0dzVynZUo2yLxxtbS0qqqqSkpKiKjpo/QYfD6/xY2qqant379/165dSUlJERERK1euXLVqlTK5FaadRpOSGsWpYLtNg1TQfWtrayISCAQpKSnMDd2YcStNd4ssZusCgaCgoMDU1PTRo0fMetmhtXJ7AR1FbGxsjx49unTp4uHhMXny5BEjRqxbt+748ePSNwyTm5POgLCyspKe7jKfGmaCORFpa2u3VW7lp59+Cg4OZpYzMjKYmW7K+/7776dPnz5s2LDc3Nzjx4+3n9yKNC9sb2+flJTUtEBzhy81NTU+Pj4xMdHV1XX69OlMJvTy5ctMjrWtvNE329atW0+ePElE/v7+c+bM+eSTT5i5Xbdv324aCfPNVlRUxDxQ5e7du9KX1NTUVq1aNWXKFAsLC+moxuYyd8oXVqYk03GRSJSSkmJhYcHsFinmv3w+X/auEba2tjweT3qbsFeE3AoAAChleH8tB2uNlCzRhtDiwR6aFVUNCSlCNpuG99fy68l/p7P6k9y6MYvzhvTV3HlClZ+G4vp/8ihCkeROsjA28fX8nVPgxoOaLzYV+vbkmxuzE9OERGRqyCYiGwsOEYVdKK+safBw5U0aoqMgtqaFVduEbK1h/bS62Wo8zhD1/yTnQ2+tGqFEWCcJ3/hyg7QBFBg+fLiDg0NKSsr69esDAgLKy8sTEhLYbPbw4cOlw0+EQuGdO3dkr7Iwp1thYWGVlZWenp6TJk3q1avX1atXly9ffv369Z9//lnljV6/fv3zzz/39fW1sLBgxoeampqq3LtGcY4ZM6a57Tatq6D7w4YN69at2+PHj/v16/fhhx/W1NQIhcJTp0413S2yDQ4bNszOzi4tLc3Hx8fLy+u3334jonHjxnXq1EnlDkK7EhgYePLkyadPnz59+vTGjRsjRowgIkNDQ+ac09jY2NTUtKCgwMXFhSmfk5MjrSt7i8MO7cGDB5cvXx44cOD27dtfy8O2XpebN28+f/7c3Nx87dq18fHxTKJBT0/P0dGRmRbU3OFjqoeGhm7ZsmXatGnMw0Da8C62jDf6Zrtz586dO3foxe04L1y4IFu9EeYlAwMDPT29srKyLl26SF/66KOPFi9eXFFRMWXKlOzs7LVr1/r5+TG5YyYG2WmkyhdWUFLK2dmZiNTV1Zn0d6MH6TL/raqqGjNmjPSOolwut7lBlyrAvWwBAEApXA3W1d2WweP0nwnE60OLQ8+VC+sk4wfpEJGGOitiayef9/jxj2oPn6+Y39KcF7nGDdT5fKweh02LdwjiE2uH9tV63T1ozMyIY6zPPvpnxcpfii/fqh7krnlsgwURzZ9k4OHCyy+u33689M+4asWxNS2s2iZkMbt6/kR9Q121wxcr/rpd7WjT8tQDAOVxudxr164FBwc/e/Zs3bp1oaGhQqHw448/JqJx48Z9/vnnHA5n0aJFcXFxQ4cOldZasGCBh4dHfn7+9u3bL168SEQ//fRTnz590tPTo6KiVq9erfJGzczMjI2Njx49umLFisuXLw8aNOj48eMq965RnAq225SC7jPtzJ8/39DQ8PDhw3/99Zejo6Pc3SKLx+PFxMTMmjVLKBQePnzYyMho9erVcp+CBB1UaGhoQUFBQkLCxYsXb926RUR5eXkPHz6MjIxMS0vjcDhXr1793//+x4yPPnHihPSuH2+ZoKCgwYMHyz5Xq5GxY8fGvtBqM2uEQmFQUJBQKOzevXtqampMTExcXFxubu60adOYAs0dPubVX3/9ta6urlOnTswoj9OnT7dO2M1pP2+26OjoJ0+eMEMdv/rqK9nJB9IhgVwut1evXp6entKXmFE/kydP3rZtG5OGVr6wgpJSixYtOnz48I0bNywsLCorKxv9HYmIiEhJSdHS0lq2bJmnp2dAQMCiRYuCg4NfY26F1ZqJUiaxtHLlylbbIgC0f8zP8Q530Yb5QpOdYAIAjah5pFKH/XR3uLAB2qcO9IFSclrW48ePnZycfvjhh/nz5xPRsmXLBg0a5OjoaGRkVFxcHBcXFxISwpycm5mZrVmzJiAgwMzMLCsr6/Dhw5s2bWJmHwgEAiMjo9GjR585c8bOzo65G4iVlZWCkQJSSu5MJbuTmJjo4uLSdE7Qxo0bv/76a2mx5ORkR0fHH3/88csvv2xU96uvvtq8ebNsm+fPnx8yZMiWLVsWLlwoLSZb4M6dO9Jb5yqguKfKT6NzdXVdvHjxgAEDTE1Ni4uL7969u2nTpujoaFJ4+Bjh4eEfffQREe3cubPpg2nedPBt8mYLDAwMDQ1t8d3o6Oi4e/funj17pqen79u378cffyQiExOTkpKSbdu2TZw4USwWX7p0ic/njxo16sCBA4GBgXZ2docOHXr33Xe5XC6zhs1mhfUS6AAAIABJREFUK1l45syZzZWU9nHRokUjRoxwd3dPTk6eO3futWvXGnXf1NQ0JCTk/ffft7S0rKioSEpKCg0N3b9/vzLHQpnPHXIrANDGkFsBeFshtwIAHegDpczprrW1dUpKCpfLnTNnjuxtMlvT682ttGevK7fSJl49+PbwZgMpZT53mBMEAAAAAADQgm3btiUmJnK53NLS0vDw8LYOB95meLN1RMitAAAAAAAAtMDe3r66ujoyMtLf31/6+BWANwFvto4IzwkCAAAAAABowfvvv9/WIcB/Bd5sHRHGrQAAAAAAAAAAqA65FQAAAAAAAAAA1SG3AgAAAAAAAACgOuRWAAAAAAAAAABUh9wKAAAAAAAAAIDqkFsBAAAAAAAAAFAdSyKRtN7GWKxW2xYAdCyt+V30WjBfaA1x9m0dCED7peaRSh32093hwgZonzrQB6qjnKoouTM7SncUUNzTdt7BDh08NKXM5w7jVgAAAAAAAAAAVMdpzY11iIw1AAAAAAAAAIDyWnVOEADAWwODOQGU1OF+aeDTDfDadYjvgY7y2cecIEY772CHDh6awpwgAAAAAAAAAIA3q1XnBAEAvDU6xCU4AFABPt0AAADwsjAnCAAAAAAAAABAdZgTBAAAAAAAAACgOuRWAAAAAAAAAABUh9wKAAAAAAAAAIDqkFsBAAAAAAAAAFAdcisAAAAAAAAAAKp7I7kVPz8/loxPPvmkuZK9e/c+dOjQm4gBAAAAAAAAAKAVcN5Qu4sWLQoJCWGW1dXVFRR79913m64Xi8UczpuKDQAAAAAAAADgdXlTc4LU1dW1X+ByuUT0/fffOzo66ujo2NrabtiwgSm2adOm+Ph4ZtnJySkkJMTb29vR0TEmJsbJyWnVqlUDBgxwcXHp3bv3w4cPmWLl5eWzZ8+2srIyMTGZMGGCQCBg1m/dutXW1lZfX79Tp07r1q2TuwYAAAAAAAAA4PVqvbEhlpaW58+ft7W1TUhIeP/99x0dHceMGdOozJkzZ6Kjo42MjJj/RkRE/Pnnn/r6+hs3bvziiy+uXr1KRJMnT1ZXV793756WltaXX34ZGBh47ty5lJSUFStW3Lt3z87OrqysLD09vemaVuspAAAAAAAAAPyHSJqQXa/kcqMWfH19+Xy+0Qu3b99uVGDRokWzZ8+WSCS9evX69ddfmZWOjo47duyQlnF0dNyzZw+znJSUxOfzJRLJkydPWCxWQUEBs76srIzFYpWWlqanp/P5/OPHj5eXlzMvNV3TtJsq9xfLWMYylrGMZSxjGctYxjKWsYxlLP/Hl6Xe1JygTz/9NOGF7t27E9GZM2d8fHxsbGxsbGx2795dWFjYtFanTp1k/2tsbMws8Pn8mpoasVickZHBYrH69Olja2tra2vr5uamp6eXm5v7zjvvHDx4cOfOnebm5j4+PleuXGm65g31FAAAAAAAAAD+y+TMCZKmYVRYltLR0bG0tJT+Ny8vb/z48X/88Ye/vz+LxVqyZElqamrTWiwWS3G41tbWLBbr/v37enp6jV4aO3bs2LFja2pqfvjhh9GjRxcVFTVdI9u+bLbptfQdy1jGMpaxjGUsYxnLWMYylrGMZSz/d5al3tS4lUYqKiqIyM3NjcViPXv27PDhw6q107Vr14CAgE8++SQvL4+ICgsLT548SUTJyclRUVFCoZDP55uYmKirqzdd8xq7AwAAAAAAAADAaKV72To4OKxcubJ///4mJibm5uYffvghkxxRwZEjR1asWNGvXz+BQGBqavrBBx8wg1O+/vrr5ORkNpvt4OBw8uTJ2traRmtaHBQDAAAAAAAAAPCyWHJHswAAAAAAAAAAgDJa7xnMAPAf9xrHjiEpDAAAAAAA7Ucr3W8FAAAAAAAAAOCthNwKAAAAAAAAAIDqkFsBAAAAAAAAAFAdcisAAAAAAAAAAKpr1XvZ4inIAPDWwBcagDI64p2n8ekGeL06yvdAR/nst7g/O0pHFOgo7xkAWRi3AgAAAAAAAACgujZ4BvMHH3zQ+hsFgDb3+++/N1rj6el58+bNGTNmJCcnyy5IJJLy8vJ79+4tWbIkNja2TaJVUkOcfVuHANBOqXmktnUIrwRXTQFe3VswgAIAQEkYtwIA7c7mzZvnzp1rZWV1/vx5ExOTtg4HAAAAAABAEeRWAKDdSUpKOnjwYFBQkJ6e3qRJk9o6HAAAAAAAAEWQWwGAdurx48dEZGNj09aBAAAAAAAAKILcCgC0U05OTkSUlZXV1oEAAAAAAAAogtwKALQ7zs7OU6dO3bt3b1lZ2ZEjR9o6HAAAAAAAAEXa4DlBAACKLVy4sLKy8t69e9OmTcvPz2/rcAAAAAAAABRBbgUA2kxsbKz06YxNFwAAAAAAADoEzAkCAAAAAAAAAFAdcisAAAAAAAAAAKpDbgUAAAAAAAAAQHXIrQAAAAAAAAAAqA65FQAAAAAAAAAA1f1Hcys7duw4e/bs2bNnT58+HRYW9t13340cOZLD+fepSWFhYWfPnvX09GzDIIlo1apVZ8+enTJlystWZDr42WefvYmolKFk5EuWLGEOxKuH+oqHrG2PuJJbbydvS2hzj56I+s3M1vZNU/NIHf3VMwUlTQKeqHmknrlS2WqxvWnvz8tV80gN+blIQZk27LXspl8xjKY9ffuOJryK5ORkFovFYrEEAgERGRsbs1isM2fOtG1UQ4cOZbFYy5cvf9mKbR6/ypGr5lX62+b7ChITEyUSyY4dO5j/du7c+cCBA7m5uSKRSCAQxMXFzZ49u20jVFKjjhBRRkaGRCL59ttviai0tFQiz9GjR6V1JRLJwoULpdUvXLjArNy8eXNzmwB4i/2nn8FcWlpaVFRkamrq6Ojo6Ojo5eW1fPlyoVBIRKmpqdra2hUVFW86BjabXV9f/6a30j7p6em5u7szy76+vqGhoSKRqK2CabUj/ipee5D/5bdfh7bjeGlsYq2bHXfmSN2uluoKSvbuxi2taDDUZTd9qU4sUee8hY+7rhNLFPS6Q3tb+wWvRe/evUtLSw0NDd/0hurq6tTVFX3tqKbV4n8LYF+1N8eOHevfv79AIIiOjtbT03Nycho4cOCuXbvaOq5XdevWLR0dHSLq0aMHj8d7/vx5ZmYmEaWmpsoWmzNnztatW+vr611cXAICAtomVoD24T86boVx/fr1+fPnT5kyhfn6c3R0HDNmDPOSvb29g4MD84UyatSoX375JTw8/NixY9u3b/f392fKMIMIxo8fv379+mPHjm3dutXJyYl5ae7cuXv37j158uSpU6d27tw5fPhw6UaZWuPGjdu0adPp06e7d+/eXOEdO3b07NmTiMaPH3/27Fkm46urq/vZZ5/t3r37xIkT27ZtGzhwYIvdPHr06NmzZ8eMGbN58+aTJ0+uXr3ayMjos88+O3bs2L59+6TdUblH2traQUFBu3btOnny5MGDB6dNmyY38qYGDBjAZrOfP39eWVmpra0tOxxDupfWrl177NixHTt2KBOJ1IIFC2THwvTu3fvs2bMHDhxQU1Nr7mgqc8RlNXcg+vbte/bs2bCwMENDQzU1tU2bNp09e/bTTz9VvHsbaa6PskEq3kvNhdf07Sc3AGjPBszO+eV0GRE9SBPO21J461HtJ+vybUc+1fJJ43mlOX+c+dPJUmnh24+FcY9qi8vr6cWohw37i70+zeb2T4u+WyO7cnBwrt6A9O4TM288qGXqCkrr524utB+Toe2b9u7krP3nyqXNbgkrcRiTwfdO0xuQ3mNS1sHI8uZWNmIwMF3NI3XjwZK+Qdma3mlD5+bmFornbi7UG5Bu8+FTaZXmNt19YuaF2GoiWh9arOaR2n1iptx+yfaaiEoqGr7aJnAal6npnWbx/pOlOwVKRiu3IhEVltR/sanA7qMMbd80l48z14cWC0WSFg9cc52SuxW5PZXtl4IYFBxT6OhSU1MHDBigpaXVs2fPuLg42Zdu374dFxdXXFxMRFu2bLG3t+fxeLq6um5ubgcPHmTKMOMd1q9fP2DAAF1d3V69et24cYN5aebMmTY2Npqamlwut1u3bj/99JO0ZabWhg0b+vfvr6GhER0d3VxhV1fXCxcuENH69etZLJarqysRCQSCOXPm2NnZaWlp9ejRY//+/XK7Jhu/vr4+i8XauHGjp6cnn88fMmRIbm7unDlzdHV1ra2tpd1RrUclJSULFy50dHTk8/nm5uZLlixpLnJZyuyfQYMG6erqurq6KrNXpaZOncpisYKDg5n/RkZGslgsCwsL5sqH3EMpu68UHG5oHTwer1+/fkTk7+8fEBDg4eFhYmKydOlS5lUTE5OdO3emp6dXVVUlJSUtW7aMy+UyLwkEAolEsnTp0kuXLpWXlycmJjLttB+DBw/29PT09PRkUirHjh1j/hsSEiItU1xcbG1tzZxALViwgIiKihQNLAV4u/2ncysMiUTyxx9/3Lt3j4j69+/f6FUHB4egoCAej3fixImIiIiCggILCwvZAh999FFiYuLdu3e7du26evVqPT09IrK1tX348OGJEycuX77cqVOnzz77rEePHrK1pkyZIhKJrl+/Xltb21zhqKiogoICIkpJSfn999+joqI4HM633347fPjw0tLSU6dOaWtrz5s3b8iQIcp0c8qUKZWVlWKx+L333tuxY0fv3r2zs7ONjY2Dg4MNDAxU7hGHw/nmm29GjRqlq6t77dq1pKQka2vrppHLDYk54b969WpsbCwRDR48uFGBCRMmZGdnZ2dnW1tbz58/n1nZ4r4looiICCLy8/PT0NAgIuZv1ZUrV+zs7BQfTVLiiDO9bu5A3Lx5MyIiQkdHZ+7cuWPHjnVyckpLSwsNDVW8extRpo8K9lKL7xPZt19zzb4JH8jDvCR30KkCrRl2ezPGX9vJVoOIunflzvtY38OV/yBN5NdT8+vphoHDdVOzRHO+K7x8q7q56iE/F/E01Mb6a2vz//3+X7u32LmLhnMXjUdPRIGrnxORqE7i81nOjhOlZobsr6YallTUB63N33OmjIjiHtV+tU1QWdOwJNAweJy+jQUnPadO7srmYljxS5G+jhpXg3Uxrrr7xMyI61XOXTSy88WzNhTkCcQKNj3lfV0bcw4Rubvw5n2sP+V9XcX9Yjri+1n2lrASQWn9hAAdrx78pCciZaKVW5GIhCKJ96ycXeFlbDWaGKBTUFIf8nPRjDX5io9ac51qbisKeqpkDE2PKXR0dXV1w4YNi46ONjU1dXJymjt3rtxicXFxCxcurKysXLp0aXBwsI2NTVpammyB7777zs/Pb8iQIXfv3h06dCjz9/rBgwd+fn5LliwJDAxMTU0NDg6+fPmybK3ly5fzeLyxY8dqa2s3V3jq1Kk2NjZE5O7uPm/evKlTp4pEIm9v7x07dpiZmS1atKikpGTGjBl79uxRpr8hISH6+vpcLvfixYuurq4RERHOzs7Z2dmffvppXl6eaj0SiUQ+Pj5btmwRCAQTJkzw8vJKSkqSG3mjYFrcP2vWrHF2dnZ2dn706NH06dOVrEVEX3zxBRGFhYUxf5HDw8OJaPLkyWw2u8VDqczhhjettrY2NzeXiMLCwjZs2DBy5EhNTc2UlBQi4nK5MTExs2fPrq+vP3LkiKmp6bp16xqlF1esWJGUlJSUlOTi4nLgwIE26cLYsWNjX2j6o1cxZtT5/PnzTU1NJ0+efOfOndu3b7+hOAHav//0nCBZz549e++99xplGYiIz+cTUV1dXUFBQW5u7rFjx5hJQ1InTpwIDw9nsVh79uwxNTX18vKKiIj4+uuvRSIRi8XicrmGhoZ9+vR5991379+/L611+fLlbdu2McvNFQ4PD+/evbupqWlCQsKhQ4eIqG/fvpaWliKRaPPmzbW1tYWFhcwJPHOxRbHffvtt//79U6dOHTdunLq6+v/7f/+vurr62LFj6urqXbp0KSkpUa1HvXv3trGxkUgkixcvzsnJoRfTTBpF3pSDgwPzO+batWuGhoaDBg1yc3MzNTVlfhIxwsPDDx8+bGdn9/3331tYWGhpaVVVVbW4b4no77//Tk9P79q1a9++fa9du+bh4cHsc319fcVHU5kjTkR9+vRRcCD27dvn5OTUs2fP9957r7q6etOmTWKxWPHubdS+Mn1UsJfc3NwUv09k337Q4QSP049LrE3OEPVz421dYEJEfr34PA1WQwPVCBueCcTnYqouxVcP7KMpt/qMEbr/W27WaOXiqQarZhndflzrHpidllNXWtEQdac6OUPE57LC1lpo8VhWZpxP1uV/e6Dk01F6Ff8/e/cd19TVxgH8SYCEDbIVkCEgiopVFOoCK6JFq1aldVRBbX2r1Vp9tdS6B1at1bZatbaKWhXrHjhfB+JEcCECAjJFVtgzAZL3j6tpmpAQIjL09/3008/l3nPPfc65iUmee+655UIi4mqwbNuqO7XnLJpqpK3JunS3QnalvCbMHW+4dpbJ4m35a3YXVPFFEUes9XXYBh884wtE0UmCiiqhvEMHTm4Tdq8iLbtmcG/tVV8a19suIjp7qzzmmYDFoht/WDE5qeoa0bX7lfVGW+eOzPqEdIG6Git8h7W5kdoor/KP5r04+L/Sn74xaWsi9yP1zM3yOhtl2katzqMoaKmSMcieU0M9XE1p3cLCwpKSkthsdnh4uLW1dXBw8NSpU2WLMfeNcrlcW1tbJyenxYsXa2v/61+DhQsXBgYGikQie3v71NTUQ4cOzZo16/r165qamkKhsLKy8sWLF6Ghof/73/8kx8ZOmTJl586dzLK8woGBgVevXk1LSxs8ePDq1auJ6Pjx4/Hx8VpaWgcOHNDR0bG2tp42bdoPP/zADOdUbO7cuevWrVu0aNGaNWuqqqru3r2rr6+vr6/P5/Ojo6MlfwEq36LS0tKYmBgWi3Xz5k1mpGd1dTURyUYupd7+CQwMXLFiRVRUVK9evZKSkoqKigwNDevdi4g8PDx69Ohx//79Y8eOffrppydPniSigIAAZU6lkmXgTfPz8/v999+7devGDAcWCAQrV64MCgry9fV1cnKqqanp379/Tk7O8ePHQ0NDx40bN2/ePHF+cN26dcuWLXNzc4uMjHRwcDA0NCwqKlJ4tMZnbm5ubl7HB6gysrOzQ0JC/P39g4ODuVzupk2bZFOTAO8O5FZesrS0JCLJFAPj8ePHV65c8fT0/Oabb4hIIBDs27dPcv6wnJwcIhKJRDk5OWZmZqampiwWa/z48V5eXkZGRizWy+/rzK96sYcPHzILyhQWMzExISIOh/Pnn3+KV5qZmXE4nHpnKmHujSwpKWFiZhaqqqo0NDSYbIJqLTI1NWWqZRIrRKTk/B3MKJXnz5+npaU9f/68pKREX19/0KBBISEh4jKpqalEVFHx8gq8lpZWRUWFkt119uzZ2bNne3t7FxUV6enpJSUlpaWlZWRkKD6bpMQZp/pORHV19fHjx+fPn89iscLDw7Oz/3XFWLZ7pSJv0Euizl5SEB6zLH75wVtAKKQVf+TvO1ealV8jFL5cmVMg9204qHcdX7u7OnKJyED35UQeZZXC9JwaIqrki+xGpoiLpWZVVwlEA920/Yfp7z9fGrAih4g0OazVM4znjGsju3LeBOlsNaNnJ00iMjZUIyJ7Sw0TQzUi0tVi8wW1pRXCzDy5h9bkyM3X1NkuIkrLriEiE0M1JnNBRBrqrDqbIBVtnTsSEdMzJoZscyM1IurS4eXo7oycGgW5FXn9+TStus6jKKZMDLLnFLmV1i4jI4OIjI2Nra2tiYi591bWwIED/f399+/fz4ye0NTUDAoKYsbqM+zs7IiIxWLZ2tqmpqamp6cLhcLly5fv27cvKytL+OrfEebTSszb25tZUKawWHp6OhFVVlba2tqKV6amplZVVWlqaipur5ubG736wLW3t2cWdHV1+Xy+1LxjyreIubXBxMREfAutMnPHKNPkbt26EZF4LGpZWZm+vr6SHTVz5szPP/98165d5ubm+fn5PXv2ZG5KqvdUKlkG3rQ7d+64urra2dm5u7tPnDhx+PDhq1evPnToUPv27YmIx+Mx5z0mJoYpb21tLc6tREdHE1FxcTHzp66ubtPnVn777TfxXWmpqanMtU/lbdy40d/f39fXNzMz89ChQ8itwLsMuRVisVhDhw7t3r07EYlvkRUTiUQ///zz1q1b27dv7+bmNmHChPHjx0v+0razs7tx44aGhgbzXYfH4/Xp02fMmDGVlZXMoNOJEyd27dpV/COZIU6FKC7MfBiz2S+/EDPPAqiqqlq7dq04haGhoaFMOoMpw9xMITmMQlaDWsSEpK+vb2lpyQyJZMatSEUuhcPh9O/fn4isrKxOnTolXj9o0KCDBw+K7/hgKpG8AUSZvmVcu3ZtypQp3bp1YxrLjMKt92wqWUbxidDT02OuOAkEgsGDB1+5ciU+Pl5B90pFrnwb5fVSva+TZpwzGBrdsbCydXsL9bTZf62wsDZXX7o9P+x+pYK7prS4dbyW1NhERJKvMmszdSLS0WIfWdtW7dX7mF8tUlcjNouCl5pvDTSLTRGcuVm+fEf+ij8K5o5vI7tSXm5FQ/2fw8mmEhQcmojU2CwiEmeRFLeLiNqbqxMRr6g2IV3g1P7lwBB1NVa90da5o4Y6iwmPVyTMLaw1a6P2JPnluDYrc0Wfp/IaVVElqvMoCloqrk1xDLLnFFo7KysrIiosLCwuLjYwMEhJSamzGJvN3r1797Zt22JjY8+cObNs2bLly5dL/th+9OjRJ598wufz4+LiiMja2vrYsWPr1q3T09Pbt2+ftbX1kiVLwsLCpO6+FF+GUVxYTU2NXn0wMZUTkY6OztGjR5lNRMTn8yWfzCgPk/VgPv4UZ0CUb5H4t25CQoKTkxNJTM0rFbkkZfqH2V3yw1qZvRgTJkxYsGDBlStXmEsgzFcIUuJUKlkG3rSAgIAjR46kpKSkpKTcunVr+PDhRGRkZMTkQ01MTJhx2S4uLkx58fVI+vf381YqOjr68uXLgwYN2rx5MzMQDOCd9U7nVvr27evs7GxmZsZMDvr06VPmNldJzs7OM2bMiImJKSgoYPK44tQyY9SoURYWFu3atTM0NKyqqrpx44Z4flANDY0OHTp07NhRQQzi7EOdhfPy8ojI09NTU1Pz6dOnt27dyszMtLS09PPze/jwoZaWlr29vVAobNw7GxvUoqioqIyMDGtr6/Xr10dERHC5XGYGFqnIr127JnmIvn37amtri0SiyMhI5uNEQ0OjR48eZmZm3bp1k3fzC9XXXZIEAsHly5dHjhzZs2fPmpqa8PBwUuJsKlkmKipKwYn45ptvTExMLl26FBMT88033yxYsGDOnDllZWXyulflNsqjODx4y9TUvvxCxheI7sXz78Q0zhw6w/rqOLXnJKQL1gQXDHbXLi0XPkzgq6nRsL46Nx9VfrU+z7OHloWJWkwSn4jMjNRuRdexstEPTUQ2bdWJaP+FkrJKoXsXzQlD9BTX5ttHp5MtJy5V0Pfz5yP661TyRfxq0bwJhvVGW+eOR9e19e2r42ClkfS82vM/z/u5ap4MLyciv0G67eQPWlHQqOPr29V5FMUtVS0GaO28vLzs7e2Tk5NHjx49dOjQOidGJaJbt27NnDnT09Ozbdu2jx8/JiIzMzPJAj/99FNycnJCQkJOTo6uru4nn3winhaNz+ffu3ePmQRNHvHlmToLM5+b+/fvLysr8/DwGDNmjJOTU0JCQlBQkI+PT0lJycOHD9XU1Oqch15lyrfI19e3U6dOcXFxffr0GTFiRGVlJZ/PP3bsmGzkEyZMULLJ8ii/l5aWVkBAwKZNm86dO8fhcMSHrvdUKlkG3rTg4OCtW7cmJCTk5uYyE+RlZWU9fvz40aNHSUlJDg4O4eHhN27cGDlyJBEdPnz4xYsXzR1yI5s6daqTk5OCF/nYsWOZkWhEFBcXN2XKlKYKDaBJvdMjhA0NDZlxpE+fPv3zzz8XLlwoO7VnUVFRSUlJ//79J06c6Orq+ujRo3Xr1kkWYJ4IY21tnZycvGLFiqKiops3b549e7a2tjYgIKBjx473799XEIPiwidOnEhISDA0NBw+fPh7771XXV393XffnTlzxtjY+JNPPvH29tbQ0Lh+/Xoj9YcqLWJCOnnyZGlpqaenZ7du3ZjRK1KRSx2CGVp879691atXBwUFBQUFLV++/NmzZ1TXjLaSGtS3Z8+eZRI3d+/eZcYP13s2lSyj4ESMHDmyV69e2dnZO3bsuHLlys2bN01NTefMmaOge1+njXVqmtcJtBB+g/RmjjVQV6PALby7MVVD39dplGq5HFb4DqtZfoYveDVBwQXBoSX8atEn3npEZG6sbmKodvB/pct+L7gcWeHdW/vvNW3rXNnohyaiuRPauLto5hTUbj5U9L8IuVP2StU2d7yhkT77wMXSK1EVHW00lIm2zh2JSJPDuv6H9fRRBnyB6MCFUmMDteXTjfcst1CtUfKOorilqsUArR2Hwzl79uyAAQPu3r27f/9+eWMTzM3NTUxMDh48uHTp0suXL3t7ex86dEiywKpVq168eBEbG9u9e/ezZ8+am5v7+fnNnDlTXV3922+/jYiIGDp0qIIwFBeeN2+eu7t7Tk7O5s2bL168yOVyr1+/PmvWrBcvXqxevTo4OJjP53/66aev3xuqtYiJZ+7cuUZGRgcOHLhy5Yr4AoZU5Mo3WZ4G7TVz5kxmzMtHH30kfrhyvadSyTLQ6JhhSuIxGosXL46IiDAzM/P09BQKhSdPnhwyZEhFRUVVVVW/fv127NjB5XInTJiQn5+/bNmyyZMnN2vsb0R6evqlS5fE1xFlmZubu7+Ch1TCW4zVlIPQxB8bTXbEN2r//v16enpr1qxR8gpGy/f2tejXX3+1tbVdtWpVZGRkc8fyFnZvozh9+jS1ztGwzD9owgjH5g7VlBSZAAAgAElEQVQEoIViuydSa353t8bIFTMxMcnPzz9+/PioUaOaO5bG8Ta1yNXVNTo6+vTp08wdJW+N1vVuUnALtlj79u0TEhK4XO7s2bO3bNnSBFHJqrc/lWlIC9daXjMAkjB+GN5OHh4enTt3trGxef78Oe6FAQAAgJbpxIkTN27cePz4sbOzs6+vb3OHA4r8+uuvAQEBXC63qKhIdiYBAHjHIbcCbydmfuKUlJSff/4ZmW8AAABombZv337p0iVXV9c9e/bIewgAtBCOjo4VFRXXr19fvHix+Fk/AAAM3BMEAM0J9wQBvK1wTxAAtK53U2u5lQb3BAG0TMiOAwAAAAAAAACoDrkVAAAAAAAAAADVIbcCAAAAAAAAAKA65FYAAAAAAAAAAFSH3AoAAAAAAAAAgOqQWwEAAAAAAAAAUB1yKwAAAAAAAAAAqkNuBQAAAAAAAABAdcitAAAAAAAAAACoDrkVAAAAAAAAAADVsUQiUdMdjMVqsmMBQCvSlP8QNRb8gwagDLy7AaC1/DvQWt779fZna2mIAq3lNQMgCeNWAAAAAAAAAABU16TjVgAAAAAAAAAA3jIYtwIAAAAAAAAAoDrkVgAAAAAAAAAAVIfcCgAAAAAAAACA6pBbAQAAAAAAAABQHXIrAAAAAAAAAACqQ24FAAAAAAAAAEB1yK0AAAAAAAAAAKgOuRUAAAAAAAAAANUhtwIAAAAAAAAAoDrkVgAAAAAAAAAAVNf4uRUvLy8Wi3X48GHxmrNnz7JYLDc3N+Urefjwoa6uruIybm5u+/bte816nJ2dT5w4oXxgKlAmTgAAAAAAAABopd7IuBVnZ+edO3eK//zzzz87der0Jg7UKnz77be9e/du7igAAAAAAAAA4I14I7mVMWPGREVFpaenE1Fubu7Vq1fHjRsn3pqVleXn52dmZmZlZTVr1qyKigpmfUZGxpAhQwwNDbt06XLz5k1x+ZKSkhkzZlhbW5uamo4bN47H4yk+ekPriYuLc3d3NzAw6NOnT0xMDLPSwsLixo0bzHJUVJTk4Je0tDQfHx9DQ8Nu3bpt27ZNvEne+vXr19+9e5eInJ2dly9fPnDgQBcXFzc3t8ePHyveEQAAAAAAAABavjeSW9HU1Bw3blxwcDAR7d69e9SoUfr6+uKtY8aM0dDQePbsWVRU1P379+fNm8es9/Pzs7Kyys7OPnv27Pbt28XlJ06cmJOT8+DBg/T0dAMDg4CAAMVHb2g9v//+e3BwMI/HGzJkyLBhw/h8vuL6x44d6+DgkJOTc/78+V27dtW7XtKZM2eOHz/+5MkTPz+/r776SvkdAQAAAAAAAKCFEv2b5Eoll6Vq8PT0XLVq1YMHD2xsbGprazt27Hjjxo1Nmzb17NlTJBLFx8cTUW5uLlP48uXLmpqatbW18fHxLBaroKCAWb93714dHR2RSJScnMxiscTli4uLWSxWUVGRSCTq2bPnX3/9JXX0htbTsWPHtWvXMitramrMzMzOnz8vEonMzc2vX7/OrI+MjGQqEYlEcXFxbDa7pKSE+fPvv/9mNslbLxlnx44d//jjD2ZlbGyslpaW4h1f87xgGctYxjKWsYxlLGMZy1jGMpaxjOU3tCxJnd6M7t27GxsbL1u2jMVi9e3bNzIykln//PlzXV1dU1NT5k97e/uqqioej/f8+XNjY+M2bdow6x0dHZmF1NRUFovVq1cvcc0GBgaZmZkGBgZ1Hreh9RCRra0ts0ZNTc3a2ppZKU9mZqaRkZGenh7zZ/v27RWvl2JiYsIsaGlpVVZW1tTUKLkjAAAAAAAAALRM0rkVcRpGhWUpn3/++VdffbV+/XrJlVZWVmVlZTwej8kypKSkcLlcExMTS0vLoqIiPp/P5XKJKDs7mynfvn17Fov16NEjeckUKSrUk5aWxizU1tZmZGRYWloSka6ublVVFbNecoYXS0vLgoKC0tJSJhvCzCmjYL0yASuz4+ucFyxjGctYxjKWsYxlLGMZy1jGMpax3LjLkt7IfCuMyZMnX7x4cfr06ZIrO3bs6OHhMXfu3LKyspycnCVLlvj7+7PZbGdn5/fee2/Tpk1ExOfzf/nlF6Z8hw4dfHx8Pv/886ysLCLKy8s7cuSIgoOqUM/27dvj4+Nramp++OEHLpfr5eVFRG5ubqdPnyaiqqoqcSVM/O+9915gYCCfz3/x4sWPP/6oeH29VN4RAAAAAAAAAFqCN5hb0dHR8fb2lpzFlnHkyJGKigo7O7v33nuvW7duGzduZNYfPnz48uXLvXr1Gj58+IgRI8TlQ0JC2rVr16dPHz09PQ8PD/Hje+RpaD1ffPGFv7+/kZFRaGjo6dOnmQEva9euffDggaur6/Dhwz09PcWVsFisI0eOJCQkmJmZ+fj4TJw4kSkvb329VN4RAAAAAAAAAFoClrwBLaCMkJCQ9evXP3jwQMn1KlcIAAAAAAAAAC3TGxy38raKjIxMTk4mouTk5DVr1owbN07xepUrBAAAAAAAAICW7009J+gt9vz587FjxxYXFxsYGIwbN27u3LmK16tcIQAAAAAAAAC0fLgnCAAAAAAAAABAdbgnCAAAAAAAAABAdbgnCACaDovFaqyqMOYOAAAAAABaCIxbAQAAAAAAAABQHXIrAAAAAAAAAACqQ24FAAAAAAAAAEB1yK0AAAAAAAAAAKgOuRUAAAAAAAAAANUhtwIAAAAAAAAAoLomfQZzIz5+FQDeDh4eHrdv354yZUp8fLzkgkgkKikpefDgwcKFC+/cudPcYQIAAAAAAMiFcSsA0BJt2LDh66+/tra2Pn/+vKmpaXOHAwAAAAAAIFeTjlthWFhYNP1BAaAlyM7OVrJkbGzs3r17U1NTr127NmHChF9++eWNBqYajMUDUIZIJGruEAAAAADerGbIrQAAKCkuLo6IbGxsmjsQAHi3IHMK0OhaRZq1Vbz3le/JVtEcBVrFawZADLkVAGi5nJ2diSg9Pb25A1FEGOHY3CEAtFBs98TmDgEAAACgKSC3AgAtUefOnSdNmrRkyZLi4uKQkJDmDgcA3kW4ZArQKFr76AkAAGUgtwIALdH8+fPLysoePHgwefLknJyc5g4HAAAAAABALuRWAKA53blzR3w5S3YBAAAAAACg5cMzmAEAAAAAAAAAVIfcCgAAAAAAAACA6pBbAQAAAAAAAABQHXIrAAAAAAAAAACqQ24FAAAAAAAAAEB1yK0AAAAAAAAAAKgOuRUAAAAAAAAAANUht1KPsLCwrFeePn167ty5ESNGqFxbbGxsVlbW0KFDpdY7ODgwhzAyMmr0ykmiFRkZGbGxsaGhodOnT9fQ0HgTx2pQGSU9ffo0qy7bt29//crfNCX7oRG7CwAAAAAAAJqSenMH0DpkZGQkJiY6Ozt3795927ZtiYmJcXFxKtTz6NEjfX39oqKiRo9QGXl5ednZ2dbW1j179uzZs+eIESP8/PwqKytVqEqZhjRiYx8+fKirq0tELi4uXC43Nzf3+fPnRJSSkiJVUkNDo7q6+vWP2PQa/bXRMrvCwsJCdmV2djYRiUSiJg8HAAAAAACgEWDcilIuXbo0ceJEHx8fImKz2S4uLsx6IyOjoKCg27dvJycnX758+dNPP2XWf/nll7du3UpLS0tMTLxy5Yqfnx+z3tXVtUePHoaGhkRkb29/9OjR5OTk//3vfz169JA8HDMo4/3332d2Yf7kcrlEtHHjxqioqJSUlPT09OvXr0+ZMkX5VoSGhvr4+Li4uHz33XdE1LNnz1mzZiloBREZGBgsW7bs5s2bqamp0dHR33//vWxDlGmssbHx2rVr79y5k5ycHB4ePmfOHA6HwxRjxmvMmTPn0KFDiYmJYWFhvXr1kor8008/HTZs2LBhw5iUysmTJ5k/161bJ9791KlT6enpffr0UdAcBZsk1Vnsww8/zMrKiouLs7CwUFNTO3XqVFZW1sqVK6VacfTo0cTExIsXL8q2QvHpk+wuxX0irxWyXaHgxQDvIFOfZLZ74olrZVLr41MFbPdEtnsir6i20StX4HhYGds9cdaPuSoftFl8OCeT7Z64ZHt+cwfSgBMneXZ8ZmdqvJ/4JFnQJDECAAAAvCuQW1EWm83u3LkzEYlEopiYGCLS0NA4efLk1KlT8/LyfvvtN0NDw59//nnixIk9evRYtmyZjo7OL7/8smvXrufPn9vZ2UnVpqGhsW/fvj59+vB4vMTExNWrV8sesc7L+J07d75169avv/76999/29nZrVmzpn///g1qiFAo3LNnz7Vr14joo48+ktcKJsgTJ058+eWXRkZGJ06cuHv3rpOTk1RtyjSWw+GcOnXK39+/trb2+PHjJiYm33333S+//CJZZt68eQkJCQkJCR07dpTapIzAwEA+nx8aGioQCBQ0R94mSfKKnTt3bteuXYaGhps2bZo1a1avXr0ePXq0atUqyX1nzpx569atsLCwrl27HjhwwMTERKryBp2+Ovuk3laIu6K8vLyh3QitRXWNKmN83Dpx3V00jfTVGj0eFQiF9N0WHotF8z9r01h1ynaLah31dguc3KZWSN9v5TV3IPD2iI+PZ7FYLBaLx1PxdWViYsJisU6cOPGaZRpk6NChLBZr8eLFjVUhvFNiYmJEItGWLVuIyNLScs+ePZmZmQKBgMfjRUREzJgxo7kDbBjJ5jBSU1NFItHatWuJqKioSFSXgwcPMjuKRKL58+eL971w4QKzcsOGDfLqB3gr4Z4gpUyZMoUZYlBSUvLf//43Pj6eiLy9vR0cHKqqqmbOnFlRUfHixYuNGzd+/fXXCxYsICKBQJCRkZGcnPzzzz/L3nfTp08fOzs7oVA4atSoFy9ejBs3btOmTVJlhEKhbCQjR47k8/lsNltTU9Pc3Hzw4MEDBgy4fv16Q1v07NkzT09PMzMzea3Yv3//oEGDnJ2dRSLRRx99lJSURESyU7Qwt+oobuygQYPs7e1rampGjRqVl5d37ty5v/76a9SoUcuXL8/JyWHKbNmy5ccff3R1dT1//rydnZ2+vn5JSYnyzTl48OC8efOI6MMPP5TXHAUtlaxKQbEVK1b06tXLy8vL09OztLT0yy+/lLrpZvPmzVu2bGGxWBEREdbW1iNGjNi1a5dkgQadvjr7pN5WiLsC3iamPsn5xbWrZxifvVkeGcsP3dRucG9tXlHtyj8Lzt0uz+LVOFhxvhlvGDBcn4h+2l/4+7HijNwajjrLtq3Gfz8znOyrT0RRcfz84tqCkloiSsyo/s+anLuxVR1tOF9/aih5LLZ7IhFd3Wbl2UMrKq6qd0AGEVVcd9DksD5fnXMpsiKvsLZWSPaWGl/5GXw11lA2WnkBSDp3uzwxo9qrh5ZtWw3FeymIR7ZbJizOlu2ovMLa5X/kX7hTkZ1fY2OhMWGo3vyJbbgcFtMPX/6QE/GkysGKM2e84bRVOUSUe8HexFCtzsZ2HZ/GjPgICi4ICi5wsec8DrGRdyJktRn0rLhM+MNXJifCyh4l8ge8p7Vzifm6PYV7zpQY6rFXfWnMNFlxwPJOnJJhDOyp3c5E/czN8uTMantL1WfdAmhEbm5uRUVFiqecU6ZM46qurn6dmeng3fH333/37duXx+OFhYUZGBg4OzsPGjRo27ZtzR1Xo4mMjNTT0yMiV1dXTU3N7OzstLQ0IkpMTOzSpQtTZvbs2Zs2baqtrXVxcWFG+gO8gzBuRSkJCQnnzp2rqKjQ19f39fVlVlpZWRGRpqZmZGTkkydPNm7cSETW1tZRUVGHDh2ysLD49ddfQ0ND4+Li/vOf/0hV2K5dOyIqLCx88eIFET1+/FjeodXV/8l/sdns+fPnP3jwICMj49mzZ4MHDyYiU1NTFVrUoUMHIsrNzZXXCi6Xa21tTUQFBQVMYoWIZOfvuHnzZr2NZQ5RUFCQl5dHRExmStwJjNjYWCIqLS1l/tTR0WlQc8TpCQXNUbBJNto6iwkEAuaTkrl0lpqaKhVGeno6EYlEooyMDCKytLSU3NrQ01dnn9TbChUSbdBaLNmer8lhj/1AV1eLLagWDfjP8y2Hi8yN1BZMMiosrZ26KuePE8URT6oW/MorqxQuDDCa5Wdo01b92XPpt211jWjYN5lh9yvN2qg723Dm/JQneywWq44AopMEXj20v/M3Chimn5gumP1j3uXICqkyygRARMevlhGRt7u2knvVGY9st8iu4QtE/ac/33a0WI1N4330cgtrl2zPn7Iyh4gE1SLfbzKv3qtsa6Le3Yn735//1Q91NvazD/VtLNSJqLeL5pxPDT/7UF/eiZAbLtHS3/MN9dhcDutiREXX8WlnbpZ3tuNk5NRMX5ObxatRELCCE6d8GGw2fdBLSyikk+EY2gaqS0xMHDhwoI6OTo8ePSIiIqS28ni82bNnOzg46OjouLq67t69m1lfWFg4f/78jh07amlpWVhYLFy4kFkfFRUVERFRUFBARD/99JOjo6Ompqa+vn63bt327t0rWyYvL2/mzJkdOnTQ0dHp3LlzUFAQn89nijHDW9asWePt7a2vr9+lS5dbt27Jxt+lS5cLFy4QUVBQEIvFYn4fivft27cvh8MJCwuT1xAFbYR3jaamJnMX9gcffODj4+Pu7m5qaiq+id7U1HTr1q3Pnj0rLy+PjY1dtGiR+Asbj8cTiUTff//9pUuXSkpKYmJiWuzd3IMHD/bw8PDw8GBSKn///Tfz55IlS5gCBQUF7du3HzNmDBExV/jy85v/zlmAptcY41YUfOf9t5eDs7Oz69zatq5JLluImzdvfv/9956engcPHvz4449PnTp1/vz5zMxMIqqoqJg2bZp4jAmHw6murp4zZ05gYKCTk5O3t/eCBQvmz5//+++/S1bIpFQMDAyYwQhMFkOsoqJCW1ubmXqjU6dO4vW+vr6zZs0qKyubNWtWZmZmYGBgnz59WEr3P4PNZn/22Weenp5EFBoaKq8VtbW1zOQmRkZG9vb2ycnJVNf0qEKhsN7GMocwMjIyMTHh8XgdO3Zk1mdlZYnL1NbW0mvMZioeLKOgOQo2yUZbZ7E2bdownyJ8Pn/8+PGHDx+OjIyU3Ldz586nTp3icDiOjo706iyLNfT01dkn9bZCtcmJm0y2nLc/ETXolfxuTnw7Zbj+n4vNmeXjYWXxqQItLmv/qrY6mixrc/XPV+es3VP4+0IzIuJqsGzbqju15yyaaqStKd2xYfcrk55Xs9l07Xcra3P14NASZryGJHZdiffwHVaaHJZQSJV84QteTeiN8kt3Kwb10pYsU1ourDcAIrr/lE9E3Ry4Su5VZzyy3VJnRyWkC9TVWOE7rM2N1EZ5lX8078XB/5X+9I3J42eCZ8+r2WwK225laaq+75z25OX/vD7rbOwPX5mE3atIy64Z3Ft71ZfGJP9EfDHKQF7Ac8cbrp1lsnhb/prdBVV8UcQRa30dtsEHz/gCUXSSoKJKKC/gmGSBvBN35ma58mF0deASlUbFVsntUwCFqqurfX19k5KSbG1tnZ2dv/76a8mtAoGgf//+8fHxffr0mTRp0s6dO6dMmVJdXe3v7z9gwICYmBgjI6Nx48aVlpYylxAkRUREzJ8/38LC4vvvv6+qqnr8+LH46o4Yn8/v169fQkKCo6Pj+PHjT5w4sXjx4piYmJCQEHGZlStXTp8+vaysLCIiwt/fPzExUaqSSZMmbdu2LS0trXfv3u+//37btm3FmxYvXjxw4MCxY8dyOJw6G/LFF1/Ia+MXX3zRCP0LrUpVVVVmZqaVldX+/ftDQ0MjIiKuXbuWkJBARFwu98aNG05OTomJiSEhIaNGjVq9enWXLl3Gjx8v3n3p0qU7duzQ1dV1d3ffs2cP8wWyuYwdO9bNzY1ZlnxT1Cs4OHj27Nlz584NCwubOHHivXv3eDzekCFD3kyYAC0X7glqgGvXroWHhw8YMODbb7+9cOHCpUuXkpOT7e3tv/nmm7CwMD09PRcXF6FQWFxcvHbt2tu3b+fm5jKZEdk7kJnJX21sbHbu3Hn16lWpKWmjo6M9PDwCAwN79eo1efJk8XrxGBYOh+Pq6tqzZ88GxT98+HA3NzcrK6s2bdoQ0b1797Zs2VJTU1NnKy5dunT58uXExERHR8fQ0NALFy5oampyudypU6dK1unm5lZvYy9fvpySkmJnZ3fixImIiAjmMcOnT59W8DNbZfJOyqVLlxRsUrKGX375pW3btgcPHrx9+/Yvv/yybdu2QYMGFRf/c2V4xowZNjY2HTp0MDU1LS8vP3XqlGTNr3n66g1PhdqgdRnU+58sRnpODRFV8kV2I/95WlZqVvX7XTX9h+nvP18asCKHiDQ5rNUzjOdN+NecJhk5NURkbKBmba5ORD06/mvolqQaicSjUEgr/sjfd640K79GfMNiToH0LKoD3bTrDYCI8otrichQj92gvWrqmrNVsltk1zAdZWLINjdSI6IuHbjiTnieW0NERvpqlqbqRNSlA6ehjSX5J6JKINLk1J0u7NlJk4iMDdWIyN5Sw8RQjYh0tdh8QW1phTAzT27ACk5cg8Iw1GUT0etMXQzvuLCwsKSkJDabHR4ebm1tHRwcLPnd4MyZM/Hx8VpaWgcOHNDR0bG2tp42bdoPP/xgamoaExPDYrFu3rzp7OxMdQ2GZYZqcrlcW1tbJyenxYsXa2tLv8HPnj2bkJCgrq5+/fp1c3Pzjz/+ePjw4QcPHty4caP412BgYOCKFSuioqJ69eqVlJRUVFTEXK8SCwwMvHr1alpa2uDBg6XmvJsyZcrOnTuJ6Pjx43U25IsvvpDXRuRW3k1+fn6///57t27dunbtSkQCgWDlypVBQUG+vr5OTk41NTX9+/fPyck5fvx4aGjouHHj5s2bJ76+uG7dumXLlrm5uUVGRjo4OBgaGjbX40SJyNzc3NzcvP5yMrKzs0NCQvz9/YODg7lc7qZNmyZNmtTo4QG0fLgnqGF+/PFHIurUqdNHH30kEAhGjhy5a9cuc3PzuXPnjhs3jsvlnjx5ksfjFRQUjBo1asGCBf379w8PD58+fbpUPdXV1RMnTrxz58577703evRoqYEeCxcufPjwoa2tbd++fcWzQBHR6dOnd+/eXVtbu2TJkvfee+/q1asNCt7U1JSZjvfevXtLly79+OOPKyoq5LWCiJhNO3bsKCwsHD16dL9+/WQvHynTWD6fP3LkyH379nE4nNGjRxcWFv7444+zZ89uUPBKqrc5dW5Spobp06cPHjw4LS1t8eLFhw4dOn36tKWl5c8//yy577p16ywsLJycnJ48eTJx4kTmHiix1zx99TYQ3npa3H9+JFubqRORjhb73C+WFze//O/UT+04GqzgpebFVztE7mm/fLpxlUC04o8CqXqszNSJqLBEWFwmJKKUF//6eaOjxWa2ElF0El+8/lhY2bq9hSXlwr9WWITvsPLqoUVEsuOH2CyqNwAiMtBl06vhKor3khdPnd0ir6N4RcLcwloiepL8sgYrc3UmQ1FQUpudX0tEj5/98+gcBY1VY7OISJxwkXci1OVPGayhTvRqxKeGunTwCgJWcOIaFEZphZBenQIAFTD3vRobGzOjbqWedcjcHltZWWlra2tqajpt2jQiSk1Nffr0KRGZmJgwiRWqaxK3gQMH+vv7Z2Zm+vv7v//++8bGxsytr7L1m5iYMD8CxdM9MFExunXrRkQGBi/HbZWVNeARZt7e3oobUlVVpWCT8geCt8adO3dcXV3t7e3Hjx8fGhrK4XBWr17t6OjYvn17IuLxeMz0gsyjMIhIcrh6dHQ0EYkv1DHzGDaX3377jfUKc/uP8pi3qq+vb2Zm5qFDh95MgAAtHcat1MPLy0vyz6ioKMkxcjweb9GiRbJ7iZ9DLIVJbTCePXv28ccfi//8888/xcvx8fEffvih+M8//vhDvLxw4ULx/ckKKpci1Qop8lpBRIWFhcuWLVu2bJmCYynT2Ly8PGaKX8XFUlJSFA9B7Nevn4LdGQqao2CTMsV27NghXpZNIRFRcnKy5DmVjVDJ06e4T+SFp+AFAG+fYX11nNpzEtIFa4ILBrtrl5YLHybw1dTIUJf91fo8zx5aFiZqMUl8IjIzkv557dVDy95SIzmzekxg1pD3tbce/tclsp7O3PAHlUu259+Krtx+7J9hWTW1L/MofIHoXjz/TkzdPyFuRVfWGwARde3AjXkmeJpe/WGfevaSF4+SfPvqOFhpJD2v9vzP836umswkI36DdNuZqJsaqr3alPF+V62T4f/8+lLQWJu26kS0/0JJWaXQvYvmmIG6dZ6IYX0bNmmUMgGbGKjJO3HyXg91hhGfWk0v7wwCUAUz81dhYWFxcbGBgUFKSorkVuZ3o46OztGjR9XUXr6R+Xx+RUUFEfF4vISEBOaxg7KTxbLZ7N27d2/bti02NvbMmTPLli1bvny51ATtTP08Hi83N9fMzOzJkyeSUTGY4yq+1ZQpI/vcAC0tLcUNUVdXV7BJwRHhbRUQEHDkyJGUlJSUlJRbt24NHz6ciIyMjJh8n4mJiZmZWW5urouLC1Oeueme8Zo3xbcc0dHRly9fHjRo0ObNm2WHpAG8I/AZAADQ+nA5rPAdVqt3Fpy7XR4UXKCnze5sxwkYom9urG5iqHbwf6WFJUI9bZZ3b+21s6SfBc7RYJ3Z1O4/P+TefVKVX1w7d0Kbbzb+M8ZqywKzaatzYlMEV+9VLp9uPPfVJr9BejceVoZcKA3cwvPupT30fZ0T1+q4FKxMAEQ0vL9OyMXSa/cqvhlnqHgvefEoSZPDuv6H9bLf8y9GVBy4UNreQmPWJ4bfTmpDRBrqrDM/WzKP3XmYwF8z0+Sr9blEpK7GUtDYuRPa3I/nP0zkbz5UVFKuP2GIXp0nokFBKhmwghMn7/VQ5yHC7lUQ0fB+KmZ/ALy8vJiJ2EaPHj106NDffvtNcuuwYcOcnJwSEhfy8WUAACAASURBVBKCgoJ8fHxKSkoePnyopqZ24sSJTp06xcXF9enTZ8SIEZWVlXw+/9ixY5L73rp1a+bMmZ6enm3btmWm+TczM5M6uq+vr4ODQ1JS0oABA/r168eM3PTz85OcIF8ZNjY2RLR///6ysjIPD48JEyZIFZDXkGHDhinY1KAY4O0QHBy8devWhISE3NxcV1dXIsrKynr8+PGjR4+SkpIcHBzCw8Nv3LgxcuRIIjp8+LDUfHxvjalTpzo5Od25c0deAcn5XOLi4qSmRAB4CyC3AgDQ0uVdtJddadZG7df5pkTSj5q69JulbGGpSjracMK2/3ONV/Jpvl06cCKC/xmuPEdi05YFZlsWSP/Oka1cXgCSxgzUnW+ifu52Ba+o1sRQzcFKQ95eCuKR7ZY6O8rcSG37wrrDNjVUu7LtZT8wT9Ux0lcz0GWzWHIb62ClcXvXv2Yfl3ciZBVe7iBe/macIZNXko1cQcAKTpyCMCQrv/ukKul5dZ9umu/Jn2cHQDEOh3P27Nnp06ffvXs3Pz9/3rx5c+bMEW/lcrnXr19ftWrVuXPnVq9eraen17lz5ylTpjDrg4KCQkNDDxw4YGBgIDWDGxGZm5ubmJgcPHiwsLBQT0/P29t73bp1UmU0NTVv3LixdOnSixcvHjhwoH379rNnz/72228b2op58+bdv3//4cOHmzdvLikpkc2tyGuI4k3wjmDGKzEDNBYvXuzt7d2xY8dOnToVFBScPHlyyZIlzECtfv36rVy50sfHZ8KECenp6Zs3b16/fn0zh/7GpKenM7fLySM5nwsGecFbidUIg9Aa+JwaeVryc4IAoFE01gTGLWT0LDPgXBjRnLP6t147T5V8EZSzaIoR88CdZjFvU96TZEGfbpp5RbW7Q0sqqkSrvjReNMWoueJpAp8szDpypSx8h1U/V60mOBzbPZFazBu2QZh3d2uMHKAFakVvKGWeWti+ffuEhAQulzt79uwtW7Y0QVRSlO/Jhj5OtKVpFa8ZALE3PJWdhweJRCQSUVUVXblCErfCAgAQUVlZmUjCpUuX2Gz277//XlRUVFZWdv/+fT09veaOEd6IKcP1k0/Yzhwr91nFTcDViZvyonrd3sKDF8u6duD+tcLi+4C3ObFCRBvmmKadsmuaxAoAwNvn119/jYmJ4XK5RUVFR48ebe5wAKAFaZLhWKtX0/37dPAgrV1Ln33WFEcEgFZi+PDh6urqhw4diouLW7ZsWX5+/uDBg6dPnz5t2rTU1NQBAwa09ksuIA+bTbZtpZ8S0sT8h+n7D1N9bpTWqL0FhmEDAKjO0dGxoqLi+vXrixcvFj9KGQCAmii38uwZHT9OsbHUrVtTHA4AWo+wsDAiEggEeXl5ly5dIqIxY8YQUY8ePZKSklatWsVMoQ8AAADQ7CQf5QkAIOkN3xP08iBssrYmOzvKyGiKwwFAa3bmzJnz589/9dVX165dS0xMtMK9hAAAAAAA0LI1SW5l505KT6fqalqxoikOBwCtWVVV1Ycffti+fftly5bZ2dkxw1gAAAAAAABarCa5J2jFCjp6lBITqaqqKQ4HAK2Zr6/vgAEDrl69WlpaSkQ8Hq+5IwIAAAAAAFCkSXIrqan0+HFTHAgAWr+cnBwvL69Zs2bV1NTs37//0KFDzR0RAAAAAACAIqxGeGx4Iz3Fo62FRaPUAwAtVnZ2dqPU0wj/cDUG5hlGwgjH5g4EoIViuydSi3nDNgjz7m6NkQO0QK3oDdUqnk6ofE+2iuYo0CpeMwBiTTLfCgAAAAAAAADAWwq5FQAAAAAAAAAA1TXJfCvKaaybBQAAAAAAAAAAmgzGrQAAAAAAAAAAqK4FzWVLmKwI4G3XWHOqtZC5zVr7FHEATaOFvGEbpBVNvQnQ8rWiN1Sr+GTHXLYALVMLuicIAAAAoOVo7T9LAAAAoMkgtwIAoCJcTgEAAAAAAEJuBQAAAEAKMqcA76a37L3/ljUHoIVrjPlWAAAAAAAAAADeVXhOEAAAAAAAAACA6pBbAQAAAAAAAABQHXIrAAAAAAAAAACqQ24FAAAAAAAAAEB1yK0AAAAAAAAAAKgOz2AGgKbDYrEasTY85gwAAAAAAFoCjFsBAAAAAAAAAFAdcisAAAAAAAAAAKpDbgUAAAAAAAAAQHXIrQAAAAAAAAAAqA65FQAAAAAAAAAA1SG3AgAAAAAAAACgOuRWAKA5eXh4iESigIAAqQWhUFhUVHT16lUPD4/mjhEAAAAAAEAR9SY+HovFauIjAkBrtGHDhpiYmKVLl54/f97R0TEvL6+5I6ob/k0DUIZIJGruEBoM726ARtfy/yloLW/8lt+TAO8gjFsBgJYoNjZ27969U6dONTAwmDBhQnOHAwAAANDKsFq55u4/gIZp6nErAADKi4uLIyIbG5vmDqQewgjH5g4BoIViuyc2dwivBReHARoFficDwFsP41YAoOVydnYmovT09OYOBAAAAAAAQC7kVgCgJercufOkSZN27txZXFwcEhLS3OEAAAAAAADIhXuCAKAlmj9/fllZ2YMHDyZPnpyTk9Pc4QAAAAAAAMiF3AoANKc7d+6I78GWXQAAAAAAAGj5cE8QAAAAAAAAAIDqkFsBAAAAAAAAAFAdcisAAAAAAAAAAKpDbgUAAAAAAAAAQHXIrQAAAAAAAAAAqA65FQAAAAAAAAAA1SG3AgAAAAAAAACgOuRWAAAAAAAAAABUh9xK04mJiRG9UlhYePfu3U8++UTl2ng8nkgkGjVqlOwmZ2dn5igmJiaNXr9kK8rLy2NiYubOnctisaS2VldX83i827dvz507l8PhSNaQkpKyatUqceEtW7aIN7m4uPz1118ZGRl8Pj8rK+vs2bOenp5SB2UcOXJE5aYpT0Enq1AM6iVqoOaOt5Ux9UlmuyeeuFYmuyk+VcB2T2S7J/KKat9E/Y3lwzmZbPfEJdvzG3pQJYs1Sj+8ppYQA7yzhg4dymKxFi9eLF5jYmLCYrFOnDiheEcli8XHx7NYLBaLxePxGiFcVbWQMORRsjNlTxY0JakvsZaWlnv27MnMzBQIBDweLyIiYsaMGc0bIQA0PeRWmlpqauq5c+dKS0t79eoVEhLStWtX1eqJioqKiIgoKCho3PCUlJqaeubMmcTERBcXl40bN3777beSW3NycqKjo1ksloeHx8aNG69du6atrc1seu+992xtbev8xjB06NB79+599tlnbdq0iYyMzMjIGDBggJ+fn2S1Ea8kJCS80QY2SKOfCw0NjcaqCt5W1TUNTi25deK6u2ga6au9iXjeNHntbdWNAmhRqqurZVe6ubm5u7sbGRk1fTwArcjff/89efJkDocTFhb27NkzZ2fnQYMGNXdQDSB7vTM1NVUkEq1du5aIioqK6rzEdfDgQZK4Ajp//nzx7hcuXGBWbtiwoc76Ad5KyK00tTNnzvj6+vbs2ZOI2Gx29+7dmfUmJiabN29OSkoqLy9/9OhRQECAeJf//ve/iYmJVVVVJSUl0dHRkydPJpmvO46OjlevXi0vL79//767u7vkEZl/2pgBIG5ubsyfmpqaRLRz5860tLSKigo+nx8XF/fVV18p34rhw4d37979ypUrRDR69GjJrUeOHOnZs6epqenMmTOJyMPDIzAwkNk0atSo9PT0e/fuSVXI5XKDg4O5XO7jx4+dnJz69evXu3dvS0vLvXv3Slbr8cr3338vG5W8Pvz4449FIlF+fn67du3U1NRu3rwpEol+/vlnejXkZNGiRVevXi0pKbl3716fPn3qbLKCvpI8F0yF33///aVLl0pKSmJiYsQVKjjF4r1u3rwpEAi8vLzqOwPwzmFGXqzZXdDviwxu36Sw+5W8otqvN+Q5jknV9UzqPjF9d2iJuPBP+wudxqRq9U8yGPjMdUL63rMlRBQVx494UlVQ8nI0RGJG9Qcznut6JvWcnB7xpEryWMy4iWv3K4koKq6K+bNKICKiz1fn2I5M0RmQpNkvqfOnab8dKVI++KDggg9mPDcY+Mxtcvqt6JdHlFehbHu7jk+7cKeCiIKCC9juiV3Hp0k1SrXYFPSDgh5W0CJ5e4lbNHhWpsHAZ13Hp4l3URCDahUWlgoX/Mpz9kvT7p/U9sPk77fy6m0RtGrMSIegoKCBAwfq6+v37Nnz1q1bzKZp06bZ2Nhoa2tzudxOnTr99ttvUnutWbOmb9++HA6Hw+FcuHCBiIKCglgsVpcuXUjm4oGC2uRJTEwcOHCgjo5Ojx49IiIiJDfxeLzZs2c7ODjo6Oi4urru3r1bmRbVu9eaNWu8vb319fW7dOki3ut1wqizwsLCwvnz53fs2FFLS8vCwmLhwoWKq1LyfEnJy8ubOXNmhw4ddHR0OnfuHBQUxOfziahLly6yJwuai6amJvNl74MPPvDx8XF3dzc1NRV/UzU1Nd26deuzZ8/Ky8tjY2MXLVrE5XKZTQq+NLYokZGRzKXNqqoqIsrOzmb+TExMlCw2e/ZsNTU1InJxcfHx8WmeWAGaFXIrzYDNZnfr1o2IRCLRgwcPiIjD4Vy/fn3WrFk5OTnr169v06ZNcHDwF198QUTu7u4bNmzQ1dVds2bNli1b0tLSHBwcpCrU0NA4e/asl5dXbm5ufHz8r7/+KnvQOm+g6NatW1hY2A8//LB7925HR8ctW7Y0KMtuZWVlZWVFRCUldXxBFwqF27Ztu3jxIhGJh5+MGjXq5MmTsoU9PDwsLCyIaMmSJS9evGBWFhcX3717V1xm7Nixd14ZO3asVA0K+vD48eNbtmwxMjLatWvXd99916dPn6ioKMmxNgsWLAgLC7tw4UKPHj3Onz9vZmb2mn21dOnS2NjY2NhYFxeXPXv2KA5PbPXq1VVVVUeOHCkre4N3VUCrtmR7viaHPfYDXY46a8B/nm85XGRupLZgklFhae3UVTl/nCgmoognVQt+5ZVVChcGGM3yM7Rpq/7sufTl6Ooa0bBvMsPuV5q1UXe24cz5KU/2WK9u9fuX6CSBVw/t7/yNAobpJ6YLZv+YdzmyQsngN+wr9OypPcRD+/5T/odzMnMLa+utUNxeXS32Zx/q21ioE1FvF805nxp+9qH+68emoB8E1SJ5PaygRfXutWpnQWc7Tmc7zpNkQcCKbMUxKBOGbIWCapHnfzJ+2l/IK6od56PXz1UrNlmgZIugVfvxxx+9vLyGDBly//79oUOH5ubmElF0dLSXl9fChQsDAgISExNnzZp1+fJlyb0WL16sqak5duzYkSNH2tjYEFHv3r3nzJkzadIk2UPUW5uU6upqX1/fsLAwMzMzZ2fnr7/+WrxJIBD0799/y5Yt5ubm3377bWFh4ZQpU/744w/FLVJmr5UrV3bu3Llz585Pnjzx9/d//TBkKxQIBAMGDPjpp594PN64ceP69esXGxurTFX1ni9JfD6/X79+27ZtU1NTGz9+fG5u7uLFi5kLM5MmTar3ZEGTqaqqyszMJKL9+/evWbNm5MiR2trazAhrLpd748aNGTNm1NbWhoSEmJmZrV69WirpJvulsaUZPHgwc2kzLS2NiP7++2/mzyVLlojLFBQUtG/ffsyYMUQ0b948IsrPz5dXIcBbq6ETHLym5m5uc5KaNKSwsFCcIGAGVlRUVNjY2JiYmEydOlUkEiUnJxORt7e3SCRKTU2dPHmyh4eH+OYayTk+Bg8eLBKJamtrra2tiWjKlCnMIZj5Vpjlfv36kcy4Feb/bDZbR0fn9OnToldj/0i5+VYYtbW1Q4cOldwqOeRv8+bNIpGIueRla2srEok++OAD2cKffvopU1vnzp3r7TqRSDRr1iypMgr6kIi4XO69e/dEIpFQKCwuLu7QoYNkM5lhNSwWKyUlRbJyyU5Qsq+Y5RUrVkj2tqGhoeLwmL127typ4PXzLmj4vyjNj4lcGOH4pv8zNlAjoqkf6TN/Hl3Xloi0uKyUk3a5F+z/XGxORHbtNIQRjhc3WxKRjYX67mXmt3Zal11zkKzh2Pq2wgjHC5stiYjNprRTdsIIx51LzJmG5F6wF0Y4MsvhO6yEEY53d1szf1ZcdxBGODL/r7ntWBrWYXg/HSIKnNxGtv46g//hKxNhhGPtHUfbthpE9Ot8UwUVSrWX+W+IhzYRLZpiJFUzc1AVYlPQDwp6WEGLFOzF7LJ0mpFkrxZc6qD4XKhQ4bH1bYmIxaLYv22YaPk3HRRX9Ub/a+3v7uaOQinGxsZEtHbtWpFIJBQKbW1t6dXnb2VlpUgkqq2tLSsrGz58OBEFBgZK7sV8HjGGDBlCRIsWLZKq+fjx48yf8mqTKibGXF9hs9np6ekikWjXrl1Mr+bl5R07doyItLS0UlNT8/LymE9AOzs7xS1SZq+lS5eKRKLIyEjmWIWFha8ZhmyFx48fJyIWixUXF8eUFAgEiqtS8nxJdiZTobq6enZ2tkgkCg0NZQJ48eJFnSerhWstbyhSjtQ3Xg8Pj0ePHokr4fP5ixYtolffTqurq83NzYlo2LBhTIG2bduS/C+NygTwJpqTnZ0tvorJ5/NFEt91GfHx8aJXQ7+l9t2wYQOfz799+7aZmVlVVVVUVNT58+dFr31PUKOcU4Amo97Qlzi8ptjY2KdPn/r4+BgaGo4ePZqZk7V9+/b06vNYXNLW1lZTU/Pq1at79uyZOHEik8muqqpatGjRxo0bJetkUir5+fkZGRlEdP/+fXlHV1f/54yz2ezly5d/9tlnbdu2ZbNfjmBi/umvV2pq6pMnTyoqKlJSUvbu3fvkyRN5JTt27EhE2dnZRPTxxx8XFBSEh4fLFsvJyWEWHB0dY2Nj66zqt99+k02piCnow6qqKj6fv2HDhgMHDrBYrJCQkGfPnknum5KSQkQikSg1NdXW1papSlJD+yo6OpqIiotfXhDW1dVVHB7z56VLl+RVCMAY1PtldjU9p4aIKvkiu5Ep4q2pWdVVAtFAN23/Yfr7z5cGrMghIk0Oa/UM43kT2kjWk5FTQ0TGBmrW5upE1KMjV94RayQmVBUKacUf+fvOlWbl1wiFL1fmFCg75apdOw0iYrHItq16alZ1RnZNvRWK21sv1WJT0A8KeliTw5LXolohyduLWe7qyCUiA92XE8SUVQoVnwsFYcirMC27hohMDNWcbV/OI66hzlKyRdCq2dnZERGLxbK1tU1NTU1PTxcKhcuXL9+3b19WVpbw1RtD/IHL8Pb2VrJ+ZWqTwnwtMTY2Zr6o9OjRQ7wpPT2diCorK5nMAiM1NbWqqoq5mFFni2pra+vdixkabGBgwPxZVlb2mmHIVshcvTcxMXF2dmZWamhoKFOVJNnWSRVg1piYmDDfN8Q3/mRkZDC/zKHluHPnjqurq52dnbu7+8SJE4cPH7569epDhw4xX/94PB7zTomJiWHKW1tbZ2VlMcuyXxqLipS6p7XRmZubK/lDQFZ2dnZISIi/vz9zj/+mTZswnAreQcitNLWrV6/OmjXLx8fnwoUL48ePP3To0IkTJ5hP/fLy8jFjxjDfG4iIy+XW1NQIhcKAgIAZM2Z07tx52LBhK1asWL58uVRu5fnz50TUpk0bAwOD4uJi5tNarLy8XEdHp02bNvTq+wFj9OjRgYGBpaWln332WUZGxqpVq7y8vFh13gYg48yZMwrSHAw2mz19+vTBgwcTEZNCGjVqVGhoaE1NjWzh27dvZ2dnW1hYrFq16u7du8znjYGBQceOHSVvC1JAQR8SkbGx8fr164moqqpq2rRpe/fulby32dXV9dChQ8zd4+KqJDW0r5gARBKXCxSHx6isrFSmpfAu0+K+fNVZm6kTkY4W+8jatmqvbu7kV4vU1YjNouCl5lsDzWJTBGduli/fkb/ijwKp3IqVmToRFZYIi8uEBrrslBf/umlIR4tdXiksLBESUXQSX7z+WFjZur2Fetrsv1ZYWJurL92eH3a/UvnxiNGJ/E+8dfkCUVyqgIiszNXrrVDcXoYam0VE4tSJJNViU9APCnpYQYssTevZi1kp+Y+H4nNRbxiyFbY3VyciXlFtQrrAqT2HiKprRBrqLGVaBK3ao0ePPvnkE2ZSMCKytrY+duzYunXr9PT09u3bZ21tvWTJkrCwMNG/3xhaWlriZWauBGGd7zEiZWqTwtw4XFhYWFxcbGBgwFzJYDBpDh0dnaNHjzLHJSI+ny95EUi2RZaWlvXuxayX/Ix+zTBkKxT/YE5ISHByciKi6upqZaqSJNs6qQLMGh6Pl5uba2ZmJr6OxTRH8cmCJhYQEHDkyJGUlJSUlJRbt24xo7qMjIyYr38mJiZmZma5ubkuLi5MeearO0P2S2NzkbyKmZqaytx3pryNGzf6+/v7+vpmZmYeOnQIuRV4ByG30jwuXrx46dIlb2/vlStXnjx58syZM8zH86JFiy5evKivr9+9e/fa2tozZ8707dt369at165dy8rKYh4qJHtHblhYWHJysr29/bFjx86fPy81Je29e/cGDBiwevXqvn37fvnll+L14g97Lpfbs2dPDw+Pxmrd2LFj33//fRsbG2Zc6507d9atW2dsbNy3b99NmzbVuQufz586derx48e7du2amJj48OFDDQ0NFxeX3bt3i3MrY8eOdXNzY5bj4uKmTJkiWYOCPiSi3bt3W1lZBQcHX7t2bffu3SEhId27dy8sLGT2/e9//2tvb+/k5GRubl5WVnbo0CGp8F6/rxSHB9BQw/rqOLXnJKQL1gQXDHbXLi0XPkzgq6nRsL46Nx9VfrU+z7OHloWJWkwSn4jMjKR/PXv10LK31EjOrB4TmDXkfe2th/91iaynMzf8QeWS7fm3oiu3H/tnPo6a2pff/PgC0b14/p0Y6VlXFfvpQGFyZnVCuiCnoFZXi/2Jt97VexUNqtCmrToR7b9QUlYpdO+iOWGI3mvGpqAfFPSwghYZ6rLr3Uv5GJQMQ4pvH51Otpy4VEHfz5+P6K9TyRfxq0VH17VVoSpoXX766afk5OSEhIScnBxdXd1PPvnk6tWrzCY+n3/v3r07d+4oroH5KbV///6ysjIPD48JEyZIbhVfDFCyNiLy8vKyt7dPTk4ePXr00KFDJee+HTZsmJOTU0JCQlBQkI+PT0lJycOHD9XU1IYNG6agRYaGhvXu1ehhyPL19e3UqVNcXFyfPn1GjBhRWVnJ5/NDQkIaVJVs62SP4uDgkJSUNGDAgH79+jHT1fn5+bVr147qO1nQxIKDg7du3ZqQkJCbm+vq6kpEWVlZjx8/fvToUVJSkoODQ3h4+I0bN0aOHElEhw8fFs8t+DaJjo6+fPnyoEGDNm/eXOdzxwDeeo0xly2Lpfx/IiJ5/71rli5dSkRdu3b18/Pj8/nM/Gft2rVbvHjxlClTuFzu33//TUQ5OTnMTGkrV64cNGjQpUuXZD99BQKBr69veHh47969J06cKDWq5auvvoqMjOzQocPAgQOXLVsmXn/48OGtW7fW1NSsX7/e3d39/PnzjdU0c3Nz5nPlzp07c+fO9fT0LC8vHzFihEAgYO55rtO5c+fc3Nz27dtXVFTUu3dvGxub8PBwZsCLuFr3V2SfXa2gD+fOnTt8+PDk5P+zd99hUVxfH8DPLGVBEJCuUpUmYokNFEGMNWosURQ1CNZEAxp9LTFiQcVYYjcmmihWVNSgCSK2iI2AEVtooYmA1KVI36Xs+8eY/W0oK6zAgn4/T54ns7N37j33zg6yhzt3khYvXnz8+PHz588bGRn5+vqKjl27dm2nTp2sra2fPn06ZsyY2jOc332sJIQHIAWuInP3sIGHs0Y6r9LHN883sJBfIZw6vD0R6WnJa2vInb1RtP5Q3q2/SocPaHduS83Z44oKzJXdnRw/Un4YVe4XXLT0v7NaDqzQ7W+tlPiq4nZE2YYFWqL9zsPaL5qiLi9Hqw7wHkaWjx7YuO/kG7/QSudVRr8Q9LbgXtnTSU9TrrEVLp3Rwba7UlZe1X7/ghvh/1mnVrrYJIyDhBGW0KOGHNXwGBoYRg3sIUuna2iqcfyuF/3xqNTSWEG6qqBt2bRpU3p6enR0dO/evYOCgvT09JydnRctWiQvL79y5crw8HDR4mj1WbZsma2tbVZW1v79+2v/e93Y2ohIUVExKCjI0dHx4cOHp0+fZle4ZHG5XHaJ9/T09M2bN/v6+vL5/GnTpknuUUOOavIwamOPWrp0qaampp+f3x9//GFpadnYqmr3rkYBJSWl+/fvL1iwgM/n+/n5aWlpeXt7i56fKPlkQXNj5w2JMgheXl7h4eG6urpDhgyprq6+fPnyqFGjSktLy8vLBw8efPjwYS6XO2PGjNzc3PXr17NP/HwvzZkzZ8SIEfU9QUz8qRTiv4cDvDeYJpiB1rC7SN5eTZPUAq3V5cuXhXUtjitbPB5PS0tr0qRJly5dknUs0ComxDYWO0tctGAn1ElnZFLu66pft3ecOERV1rE0jfevR82HYxtPbfnqbhORa2tr5+bmBgQEtLZ/ZKX2/vVI3Pvdu/q0lQuqIXfHGxkZxcXFcblcT09PKdZnbRINHMkG3uwfGRnZvXv32vcEbdu27ZtvvhEVi42NtbS03Lt379dff13j2BUrVrDL1ooEBwePGjVq586dy5cvZ8uIvxsRESGajS5B6//AAIjDPUHQQh48eHD//n1ZRwEAAAAAIKV9+/a5u7tzudyCgoKLFy/KOpymIVopWUR8SWYR0eLNko9liU9tq68MwHsGuRVoIexSsgAAAAAAbZS5uXlpaem9e/e8vLxEz/oBACDcEwQArUdbnPmJe4IAJMM9QQBAbeeCauBNNDLXtPcEtVqt/wMDIK4p1rIFAAAAAAAAAPhQIbcCAAAAAAAAACC9Zs6t2NmRUEhCIZWX0x9/kIFB8zYHAAAAAAAAANCyWmTeyubNNH062dvT1q0t0RwAAAAAAAAAQEtpkdxKYiIFBFB0NPXs2RLNAQAAAAAAAAC0xkE44wAAIABJREFUlBbJrXA4ZGhIpqaUmtoSzQEAAAAAAAAAtBT5lmjkyBEiIh6PvL1bojkAAAAAAAAAgJbSIrkVb2+6eJHi46m8vCWaAwAAAAAAAABoKS2SW0lOpr//bomGAAAAAAAAAABaVjPnVsLCiGGatwkAAAAAAAAAANlhhELhO9fRNNkT5GAAPnBN8OOoxTFIHwM0AK5uAKC28KOgrVz4DRzJttKd+rT+DwyAuBZ5ThAAAAAAAAAAwHuqRdZbaRgkJgGgzcEPLoD3Fa5ugA8QLnwAkForyq0AAAAAAABAk0CqCKAl4Z4gAAAAAAAAAADpIbcCAAAAAAAAACA95FYAAAAAAAAAAKSH3AoAAAAAAAAAgPSaYi1bLJIEAAAAAAAAAB8qzFsBAAAAAAAAAJAecisAAAAAAAAAANJDbgUAAAAAAAAAQHrIrQAAAAAAAAAASA+5FQAAAAAAAAAA6SG3AgAAAAAAAAAgPeRWAAAAAAAAAACkh9wKAAAAAAAAAID0kFsBAAAAAAAAAJAecisAAAAAAAAAANKTl3UAAABtGMMwsg4BoA0QCoWyDqHRcHUDNLm2+KMAAKCBkFsBAAAAAAB4A6nVNgdpO2gNkFsBAHhX1eHmsg4BoJXi2MbLOoR3gt/XAZoEshUA8N7DeisAAAAAAAAAANJDbgUAAAAAAAAAQHrIrQAAAAAAAAAASA+5FQAAAAAAAAAA6SG3AgAAAAAAAAAgPeRWAAAAAAAAAACkh9wKAAAAAAAAAID0kFsBAAAAAAAAAJAecisAAAAAAAAAANJDbgUAAAAAAAAAQHrIrQAAAAAAAAAASA+5FQAAAAAAAAAA6SG3AgAArZHOyCSObfylO8U19scmCzi28RzbeF5BVZNXTkQ9pr/k2MZ77sgR7YlKEszakGn06QulwQmdxiSNXZp+53GZqKT4f87fZEgdErxPtLW1GYa5dOlSo46KjY1lGIZhGB6P18BKJJSpUZt0pOvIu5A8CKNHj2YYxsvLq2UibLHuN2tDLTxoH5TIyEihUHjgwAEi6ty58/Hjx1+9eiUQCHg8Xnh4+MKFC2UdIBGRlZWVUCgUCoXa2tqSSwYHBwuFws2bNzek2kYV5vF4QqFw4sSJjXoLoM2Rl3UAAADwnquoFCrIM409ql83bkFRtaaaXHOE1HDBf5ZOWpnOFwhVlDn9rZUEFcK7T8q6dFIY0keZLaCnKWfSUYHdtjBSlF2k0Ir069evoKBAU1OzuStpkoZas7d28L0fgeaAQWsO586ds7e35/F4ISEh6urqVlZWw4YN+/HHH5u8oQkTJvj4+AwbNiwrK6vJKweAd4F5KwAA0MTYWSFbjuUNnp/KtU8IeVzGK6ha/H2O+eRk1SEJvWemHAssFBXeeTrfYnKyskOC+tDEXjNSTgS9eetRDD88qjyvsIqI4lMrPl6Ypjokoe+slPCocvG22Akj7ESSRzHl7MtygZCI5m3OMpnwQsUxQWlwgvW0lz9cKGhsR/gC4ZxNmXyBsEdX7j8XjO8dNgj3NUwLNHUd015UZsrH7f88asj+57NIS6oBg/fNo0ePwsPD8/Ly6N85Alu2bBk+fLiampqNjU1oaKioZHx8/NChQ1VUVPr06RMeHl5nJXPmzGEYZu7cuez+69evMwyjpaXF5/PFG5JcGzsZ5M6dO2zN7Mvy8nIimjt3rrGxcbt27bhcbrdu3X744YeG9DE/P3/58uWWlpbKysr6+vqrV68mopycnEWLFnXt2lVFRcXa2trHx4fP57PlJYxDQwaBiGxsbK5du0ZEPj4+DMPY2NjUKCBd69KNAI/H8/T0NDMzU1FR6dWr17Fjx9j9AQEB7NlJT0+vqqqyt7dnGObrr78WxeDj4zN06FA1NbW+ffuKxyCuvngk9KK+Q946aBLGTfKggTglJaVBgwYR0ccffzxy5EhbW1sdHZ1vv/1WVEBHR+fgwYOJiYklJSXR0dFr1qzhcrn076yNb7/99ubNm4WFhZGRkWw9EnTo0KF79+4KCgoSypibm9++fbukpOTx48e2trbibx05cuTly5elpaV8Pj8mJuarr75i90dGRo4aNYqI1qxZIxQKIyMjm6qwSNeuXf/444/S0tKIiIg6u6mtrb1///6EhISSkpJnz565u7tLHgqA1ga5FQAAaBZrf8pVUuRM+VhVUZ5x/CLtwPkCPU25Fa6a+UVVczZl/XzpNRGFR5Wv2McrLqte7a7p4axh3FE+Ma2iRj0VlcKxX78KeVym20Heylhxyc6c2m0xdU2LeZ4gcOrT7hs3TfexavEpAs8dObf+Km1UF8IiyzNzq4ho45danbTfzPRUV+UM6K4kKnPhj6KBc1LZ/y78UcdNRgBEtHHjRmtra2tr66ioKDc3N3ZnRUXFmDFjQkJCdHV1raysFi9eXOex8+fPJ6KLFy+yqZBTp04RkaurK/vdTKQhtTF1XSrPnz93cnJavXq1u7t7fHy8h4fHrVu3JHdHIBA4Ojru3LmTx+O5uLgMHjw4Ojqaz+cPHjz4xx9/lJOTmz59enZ2tpeXV41vR7XHoYGDwHbZ2NiYiAYMGLBkyRJXV1fxd6VrXboREAgEDg4OBw4c0NPTW7lyZX5+/uzZs3/++WcimjRpkoeHB5sR27p1a2hoaL9+/bZv3y46dseOHU5OTqNGjXr8+PHo0aOzs7Nr1y85njp7Ud8hkgetIeNW36CBuPLy8levXhHR6dOnt2zZMmHChHbt2sXFxbHvcrnc+/fvL1y4sKqq6syZM7q6ups3bxbl44ho3bp10dHR0dHR3bt3P378eJ1NaGtrm5iYmJiYsLf2GBgYsC9VVFRqlFRQUAgKCnJycsrOzo6Njd23b5/4uz179gwJCfnuu++OHTtmbm5+4MCBYcOGEdHJkydfvnxJRA8fPty7d+/JkyebqrCIj48Pj8f7559/+vTpExwcrKurK/6uoqLivXv3PDw8srKytm/f3qFDB19fX/anH0BbgXuCAACgWcwep/aLlx4RBYQUxyYLlLnM6U0dVZQYQz35eZuzth7Pnz9Rvaikmoi4CoxJR3kLI8U1czTbKdX87hfyuCwhrYLDoTuHDAz15H0DC+duqjkRmlPXXwruHjZQUmSqq6mMX53Oqwy8X3LzYemw/u0a3oXM3Ep2w9yw3r8QZuVVZeVV1SgPUMOqVau8vb0fPXrUv3//hISEgoICDQ2NkJCQhIQEDodz9+5dQ0NDX1/fOXPm1D524MCB3bt3j4qKCgwMHDNmTEBAABHNmzevRrGG1Map61K5d++ekpJSdXV1WVlZenp6YGDgjRs3anwpqiEoKCgyMpJhmAcPHlhZWRFRRUVFYGBgXFycvLz8vXv39PT0Jk2aNG7cuLNnz+7atatjx471jcNff/3VkEFgj719+/bLly9HjBhRe5WHoKAgKVrX0NCQYgSuXLkSGxurrKzs5+enoqJiaGg4d+7c7777jv0e+P3334eGhl67du369etqampnz55VVPzf3YKrV69etWqVUCjs0qVLcnKyv7+/h4dHo85Inb2o7xDJgyZ53CQPGtTg7Ox86NChnj179ujRg4gEAsHGjRt9fHyIaMyYMRYWFpWVlQ4ODllZWQEBAYGBgS4uLsuWLWOP3bZt2/r16/v16/fXX3+ZmZlpaGgUFNScaLlnz56ZM2eKXv7555/shqurK5tvFXFycjIzM6uurnZ0dExNTZ09e/bRo0dF7zo4OJSXl3M4HGVl5U6dOo0bN27EiBG3bt3atm3b0KFDjY2Nb9y4IVqdp0kKi5resWPH2rVrFRUVk5OTO3bsOHXqVHapGtbYsWOtrKzKyspmzJhRUlKSmpp65MiR1atXs1lLgDYBuRUAAGgWwwa8yWKkZFUSURlfaDrhhejd5IyKcoFwaL92bmPVTgcXuXtnEZGSIrN5odayGR3E60nNqiQiLXU5Qz15Iupj+Z+/1YurFFvctrqavH/OPXW1KCO3srr6zU5REqSB9LTe/CsZn1phbVr3WipfTdHYv0KnUdXCB6hnz55EpK6uzr4sLi7W0NBITU0lIi0tLUNDQyLq06dPfYfPmzdv6dKlp06dqqioKC4utrOzY2/uENfw2ior/5cErK6u3rBhw6lTpzIyMqr/vVTeuo4D+ydrbW1tNrFCRAoKCikpKexOPT09IhJFmJqaKspu1B6HhoctmXSta2hoSDECbFtlZWUmJiaincnJyeXl5UpKSlwud/ny5TNmzBAKhdOnT+/atav4saampkTEMIyJiUlycjJblbi3xlO7F2pqalKcRPG+1Dlu9TWH3EqdwsLCevXqZWpqamtrO3PmzHHjxm3evNnf3z8+Pt7IyIiIeDwee1LYO2iIiP3ME9Hz58+J6PXr1+xLVVXV2rmV3bt3X7hwgYg+/vhjT0/PefPm5ebmEtGjR49qlGSrzc3NZU/i48ePRW9xOJwNGzZ8/vnnHTt2FGVa2VNfW5MXZjsuEAji4uI6duzIDosI+1JZWTk5OVm008TERElJiZ2yB9D64Z4gAABoFsrcNzNQDHXliUhFmXN1b+fr+9/899vOTvJyxGHId53e69td/zputGGBVrlA6P1zXo16DHTliSi/sPp1cTURvUj/z01DKsoc9l0iep7AF+3/NaR424n8wpLqk976dw8bOPVRJiKhsHFdGGijpK8lR0TrfsrN4L35Ovq6uPphFH7Pg8aRk5OjWrfkGBgYEFF+fj77nerFixd1Hkv/3gF09epVdh2N2pNW3lobe+NAfn4+/ftFjvXrr79u27atsLDw1KlT9+7dc3JyIiLh2y4V0XdF0V0PFRUV7Dc6Ho/H3ucSFRUlHlh949DwQRAdLkofiJOudelGgG1LRUUlODj4xr9+//13eXl5IsrNzV25ciURKSkpHTlypMYaJc+ePSMidkEKEvt23fB4avdC8iESBq0h41bnoEFt7u7uqqqqL168OHv2rGipEXbBYDbHoa2tzd4F0717d/bdtLQ0dqOqqore9qmLiIi4dOnSpUuX2FzJtWvX2JeiSkTYPR06dGAzYmw6j/XZZ5+tWrVKTU3t888/d3BwCAkJIbGTy4Yhyow0YWGWtbU1ESkoKFhYWJBY/o7FviwpKRk9evSIf3366afiuWCAVg7zVgAAoHmNtVexMFKMSxFs8c0bYduuqKT6aRxfTo7G2qs8eFb21facIX2U9bXlIhP4RKSrWfPBQE59lLt0Vkh6VTF5Vcaoge0Onv/PX/P6WnHvPilb+1Nu6POyn359LdpfWfXml1S+QBgRyw+LlCYbwlVkjnjpfbYq4+9EvsWUl70tuBWVwqgkgftYNdGSKxf+KHoU86bybqaKR9fW/Tc9gNqcnJy6dOmSlJT02WefjR49WsISqlpaWpMmTTp79uyDBw/at28/bdq0xtbWt2/fu3fvenl5PXjw4KeffhLtF31v4fP5ERERYWFhDYl8zJgx3bp1i4mJGTRo0Pjx48vKyvh8vp+fn5mZWUJCgqOj4+DBgy9fvkxEzs7OnTp1apJBICJ26ZDTp0+zk3dmzJghHpIUrUs3AmPHjrWwsIiLi/Px8Rk5cmRhYeHTp0/l5OTGjh1LRO7u7mlpabNnzx4yZIi7u/v06dOfPn3aocOb6Xg7d+5MSkqKi4vLyspSVVWdOnXqu8cj+RAJg0bvMG5Qg6+v78GDB+Pi4rKzs3v16kVEGRkZf//9NxEFBQUlJCSYmZndvXv3/v37EyZMIKLz58+np6c3RyQhISFJSUldunT59ddfg4ODxdeUZdN/RMTlcvv27WtnZyd+IDsfbebMmaqqqmFhYaJ83LsXZq1cudLc3Nzc3Lxjx47FxcX+/v7i7165ciUuLs7CwmLNmjXs/XS9e/euqqq6cuXKOw4IQIvBvBUAAGheXEXm7mEDD2eNdF6lj2+eb2Ahv0I4dXh7ItLTktfWkDt7o2j9obxbf5UOH9Du3JaONQ5XVGCu7O7k+JHyw6hyv+Cipf+9Y+jACt3+1kqJrypuR5RtWPC/x/Q4D2u/aIq6vBytOsB7GFk+emDN1f4a6JNBKn8dM/r8k/YaqpyHUeUvMyocP1KePExVVCArryo8qpz972+xiTMAb6WoqBgUFOTo6Pjw4cPTp0+LFl+ok2hNRxcXF1VV1doFJNf2ww8/9O/fPzEx8fbt297e3qL9zs7OixYtkpeXX7lyZXh4+OjRoxsSOZfLvXfv3tKlSzU1Nf38/P744w9LS0slJaX79+8vWLCAzbNoaWl5e3ufOHGiCQdh2bJltra2WVlZ+/fvv379uvhb0rUu3Qiw3ffw8EhPT9+8ebOvry+fz2cTXrt37w4MDOzSpcu+ffvc3NycnZ1TUlJmz54tOnbTpk3p6enR0dG9e/cOCgqqfYeFFPFIPkTCoNE7jBvQv5N6KioqiMjLyys8PFxXV3fIkCHV1dWXL18eNWpUaWkpEZWXlw8ePPjw4cNcLnfGjBm5ubnr16+fNWtWM0UlEAjGjBlz9+7dAQMGzJw5U7R0DhGdP3/+4MGDlZWV27dvt7W1DQ4OFj9w165d4eHhenp6np6eI0eObMLCrHXr1nXu3NnGxubp06djxoypcdsan89nl4ju1KmTl5fX7NmzuVzuuXPnmnRsAJoX89ZpnwAAUB92vmt1uLmsAwFopTi28dSAe0xaIfbqbouRQ+ukra2dm5sbEBAwceJEWcciA23rgmrIPVBGRkZxcXFcLtfT01N8TVaQibby0YL3G+atAAAAAAAANNS+ffsiIyO5XG5BQcHFixdlHQ4AtArIrQAAAAAAADSUubl5aWlpUFDQxx9/nJGRIetwAKBVwD1BAADSwz1BAJLhniAAoLZ2QeG5SG1OW/lowfsN81YAAAAAAAAAAKSH3AoAAAAAAAAAgPSQWwEAAAAAAAAAkB5yKwAAAAAAAAAA0kNuBQAAAAAAAABAesitAAAAAAAAAABID7kVAAAAAAAAAADpIbcCAAAAAAAAACA95FYAAAAAAAAAAKSH3AoAAAAAAAAAgPQYoVAo6xgAANoqhmFkHQJAG9AWf9nA1Q3Q5NrKjwJc/m1OW/lowfsN81YAAAAAAAAAAKQnL+sAAADaMPydBOB9hasbAAAAGg73BAEAAAAAAAAASA/3BAEAAAAAAAAASA+5FQAAAAAAAAAA6SG3AgAAAAAAAAAgPeRWAAAAAAAAAACkh9wKAAAAAAAAAID08AxmAGhRDMM0VVV4zBkAAAAAALQGmLcCAAAAAAAAACA95FYAAAAAAAAAAKSH3AoAAAAAAAAAgPSQWwEAAAAAAAAAkB5yKwAAAAAAAAAA0mvp5wQ14SNCAAAAAAAAAABkDvNWAEDG7OzshEKhu7t7jY3q6uqCgoLbt2/b2dnJOkYAAAAAAIB6tfS8FZa+vr5M2gUAmcvMzGxgye+//z4yMnLdunXBwcHm5uY5OTnNGhgAAAAAAIB0ZJNbAQB4q+jo6BMnTiQnJ9+5c2fGjBl79+6VdUR1wH2OAA0hFAplHQIAAABAM0JuBQBatZiYGCIyNjaWdSAA8GFB5hSgybWJNGtbufYbNZhtolON/Xi0iU5J0CYuB2gU5FYAoFWzsrIiopSUFFkHIkl1uLmsQwBopTi28bIOAQAAAKDZIbcCAK2UtbW1q6vr2rVrX79+febMGVmHAwAfIvxdEaBJtPUpBgAAb4XcCgC0UsuXLy8uLn7y5MmsWbOysrJkHQ4AAAAAAEDdkFsBABkLCwsT/Tmr9gYAAAAAAEArx5F1AAAAAAAAAAAAbRhyKwAAAAAAAAAA0kNuBQAAAAAAAABAesitAAAAAAAAAABID7kVAAAAAAAAAADpIbcCAAAAAAAAACC9Dyu3EhISkpGRkZGRkZqaGh0dHRgYuGDBAgUFhbceeObMmYyMjG+++aY5omrWyllmZmZsxzU1NWu8JRqThQsXinaePXuW3bl+/fqmaqiG6OjojIyM0aNH136LDWnLli2NarohNTe8ZAuclLdqYEca3l8AAAAAAABoDvKyDkAGcnJyMjMzDQ0N+/bt27dv3/Hjxzs7O5eVlck6LhmbO3fu4cOHq6qqLC0thwwZIutwpPfs2TM1NbWCggJZB9JCmry/CgoKFRUVTVVbbfr6+nXuz8zMJCKhUNh8TQMAAAAAADSHD2veCiswMHDkyJHdu3dnZyX07dvXw8ODfWvXrl2PHj168eJFSkrKvXv3Zs+eTUQhISFOTk5EtGTJkoyMjJCQEAmFiejLL78MDQ19+fJlfHz8H3/84ezszO7X1NT08fH5888/k5KSbt26NW3aNAmVi6uvIXbCwpIlS/z9/ePj40NCQvr37y86qkuXLhcvXkxKSrpx40afPn0kj0lBQUHnzp3Hjh1LRF988QUR5efnixfQ0tLaunVrWFhYUlLS3bt3lyxZoqio+NaG6uxyo0joY33j3KtXrz59+mhoaLCxXbhwgW3dxcWlzjk1hoaGZ8+eTU5OFlVe50mpr7m39veTTz7JyMiIiYnR19eXk5P77bffMjIyNm7cWKODFy9ejI+Pv379uvhJFKnvM1Cjv5I/EvWdDtFRv/32W0pKyqBBgxp7mgAAAAAAAD5kH2JuhVVdXX38+PE7d+4Q0aeffsrutLa2Dg0N3bdv37lz50xNTbds2eLg4HDhwoW0tDQievLkyc8//3zhwgUJhfv06bN+/XoVFZW9e/cePXo0LS3N1NSUiBQUFC5fvjxnzpycnJwffvhBQ0Njz549M2fOrK9ycXU2JHp32bJlcXFxcXFxlpaWe/fuZXcqKCicOnVq0KBBPB4vPj5+8+bNkkfj7NmzFRUVCxYs0NbWnjx58vPnz589eyZ6V1FR8bfffnNzc6uqqgoICNDW1v7mm2/YtiQ0VF+XG3um6uxjfeMsTkFB4fTp0/b29llZWVFRUd7e3nVWvm7dOjk5ufz8fFHltU9KA5urs79Xr149evSohobG7t27PTw8+vfv/+zZs02bNokfu2jRotDQ0JCQkB49evj5+Wlra9eoXPJn4K3DJSE80VGrVq3i8/mBgYElJSX1ngn4UOmMTOLYxl+6U1xjuzWoEU8Dw/tkySuObfzan3Lrq6fFNGG7re3UAABAmxYcHCwUCt/6PUKKwgDvpQ/xniBxiYmJQ4YM0dXVZV9OmDCBz+dzOBwlJSU9Pb0RI0Y4Ojr6+PjY29sbGBjcvXt369atomPrLHzv3j0iEggEqampSUlJe/bsYe82Gj58uJmZWXl5+aJFi0pLS9PT03ft2rV48WJbW9s6KxcnoSEiOnDgwI4dO3r16hUcHGxqaqqmplZYWDho0CBTU9Pq6uqJEyemp6e7uLjs3r1bwjhkZ2cHBARMnTp1z549ioqKhw4dEp+XMWzYsC5dulRWVk6cODEnJ+fq1asnT56cOHHihg0brKys6muovi6fPn26saepdh9VVVXrHGdxgwYNMjExqa6unjRpUmZm5uTJkw8cOFC78hMnTqxZs6Z///6//fYbW/mBAwdqnBRHR8e3Niehv97e3v3793dychoyZEhRUdGXX35Z46ab/fv3HzhwgGGY8PBwQ0PD8ePHHz16VLyA5M/AW4ersLDwrafj7Nmzy5Yta/hJgQ9Wv27cgqJqTTU5WQdSN6nDa+X9AoCmNXr06GvXrq1ZswZfBUE6kZGR3bt3/+GHHzw8PDp37rxly5bhw4fr6OgUFhYmJiYeO3bsxx9/lHWMQPTfM8XuSU5ONjY23rZt25dffqmurl77kHPnzrm4uLAHEtGKFSu+//579q1r166NHDmSiHbu3Ll8+fI664cP1oc7b4XVtWtXIsrOziYiDoezfPnyJ0+epKamJiYmjhgxgoh0dHTqPLC+wg8ePPD399fX19+3b19gYGBMTAx7i42BgQERKSkp/fXXX1FRUbt27SIiQ0NDLpcrOcK3RhUdHU1ERUVF7EsVFRUi6tSpExHl5+enp6cT0d9///3WoTh06BARDRs2LDMz8/fffxd/iw0+Ly8vJyeHiGJjY9n9nTp1ktCQ1F2urXYf6xtncWxsBQUF7EIeorBrCA8PJ6Lc3FxR5bXLNKQ5Cf0VCATsv68Mw1y6dCk5ObnGsSkpKUQkFApTU1OJqHPnzuLvNuqTWedwSQ6PLVZfpgZam/yi6hX7eFbOL9s5JHT8JOnbgzx2f05+1Vfbs80+S1YdktB92ksf3zy+QEhEHYYlcmzjt53IHzgntZ1DwujFr17lVC7+Pkd9aKLx+BcnggpFNbOzHnx88z5emKY+NLHfrJTQ5+W1A3gUww+PKs8rrGJfztucZTLhhYpjgtLgBOtpL3+4UCBe25ZjeSM8XqkPTewx/aV4bXX2gldQtfj7HPPJyapDEnrPTDkWWFi79bcSD6++GHpMf3ktrJSIfHzzOLbxPaa/rN0vCfHHp1YMW5TGBukbWMixjefYxvMKqt4aW33njoiSMypHeb5q55Ag3pCEAZFQFREJhbRgSzbHNt562su07Eoi2nk632JysrJDgvrQxF4zUsTPO0Br06xrfrVYE/BBOXfu3KxZsxQVFUNCQhITE62srIYNGybroN5VZGTkqFGjiGjNmjVCoTAyMnLSpElCoTA3N7dTp05ycnIPHjwQCoV79uyps7CMo2+Yv/76Kzw8PDw8vLy8nIgyMzPZl/Hx8eLFPD095eTkiKh79+5sYgWgTh9uboXD4cyaNYtdtDUwMJCIxowZ4+Hhoaqq6uHhMWHChNDQUCJiGIaIqqqqRNus+gpXV1cvWbLE3Nx81KhRO3bs4HK5bEbz1atXRFRaWjp9+vRp/5o1a1ZVVVXtysVJiIrFHl5jBVA206Gurq6mpkZEhoaGbx2Q6Oho9tv1kSNHavzOwQavqanJ3qtiaWnJ7s/IyJDQkIQuvzWYGmr3sb5xrj0IGhoa7Lykbt261Vm5QCCgWgNY46Q0pDkJ/e3QocPatWuJiM/nT58Wq0kBAAAgAElEQVQ+vfaKKtbW1kSkqKhobm4uilzkrZ+Btw6X5PDYAljOuU0QVAiHfJG683Q+r6DKZWT7wb2Uo5MERMQXCB0WpP148bUch6aPbJ+dX7X2p9zZG7NEB647lKvRnsNVZK6Hl/aY/vLKgxJrU8XUrMoFW7IzeJXiTXx/Kn9I33aj7No9/of/yZJX2flvuWCfJwic+rT7xk3TfaxafIrAc0fOrb9KRe9uOpJnbapobaoYlSRw986U0AtBhdDxi7QD5wv0NOVWuGrmF1XN2ZT186XX7z5otWP4/BM1Y315IhrQXWnJNI3PP1Fr+LGCCuGYr1/djijrqC3f24L7f3tyGhhGfeeOtXIfr6qatNTlxBuqb0AkVyUU0oItWb9cfv2RJffOTwYGuvLhUeUr9vGKy6pXu2t6OGsYd5RPTMMXS2hFtLW1GYbZsmWLvb09++2Ux+N5enqamZmpqKj06tXr2LFjNQr7+PgMHTpUTU2tb9++7D+LRJSTk7No0aKuXbuqqKhYW1v7+Pjw+fw6m1BUVLx27RoR+fj4MAxjY2PT4p2G94eSkhK7Vt3HH388cuRIW1tbHR2db7/9VlRAR0fn4MGDiYmJJSUl0dHRa9asEf1li8fjCYXCb7/99ubNm4WFhZGRka1n2buTJ0++fPmSiB4+fLh3796TJ08GBAQcOHBAU1Pz6NGj33zzzaBBgx49erRy5co6C8s4+oYZMWKEnZ2dnZ0dG/y5c+fYl+zv7ay8vDwjI6PJkycTETvFW/QXWYAaPsR7gsaNG9evXz8DA4MOHToQUUREBHuriLz8m9FQVFTs1atX3759RYewq29MnjxZRUXl8ePHv/76a32F+/Xrt3Xr1j///DM7O5v9Ms/j8Yjo5s2bSUlJXbp0+frrr0NCQtq3b9+9e/fq6uqbN2/Wrlw8WglRScCuumpsbHzkyJHbt2+LL30qwdKlS7t27RoREVFj/61bt168eGFqanrp0qXw8HD2cb+///57ZmZmbm5ufQ1J6HJDgpGsvnEWFxoayoYdEBAQERHBZtMbqMZJSU1NfWtzEvq7d+/ejh07nj179s8//9y7d++PP/44bNiw16//96Vx4cKFxsbGXbt21dHRKSkp+e2338Rrlu4z0PDwpKjtHbEzieokIWdU2wf4UKGg0JLIRAHD0P2fDaxMFImoolLI7o9LEcjLMXcPG+ppyk10Kvl0WfrZG0U7v36zds/S6RpbPbS9fszdciyvnC8Mv2CopsJR/ziRLxA+TxB01P7fvwXfuGmumtVBKKSuk5KTMyr8bxZ5OGtICOnuYQMlRaa6msr41em8ysD7JTcflg7r3459d5Vrhw0LtB7FlA9wT01IqygoqtZoz6mzF4H3S2KTBcpc5vSmjipKjKGe/LzNWVuP58+fWMdk3UapHcOqWR1CIkpfZlaOGNBu05dajTr2YXR5YloFh0MhPxl01pE/dbXdrA31fp7F1XfuWF9+pr5vuc6DZ2UOC9LYhm5HlNY3IJKrWvNjbmyywL6XcuCuTuqqHCIqKqkmIq4CY9JR3sJIcc0czXZKjbjQAFqGl5fX0KFDp0yZoqio6ODgEBsbO2jQIFdX1yNHjsyePbuiomL+/Pmiwjt27Fi6dKm2tvaFCxdGjx6dkJCgrq4+ePDguLg4c3Pz6dOnX7p0ycvLKzIy8syZM7WbKC0tffr06cuXLwcMGDBw4MCOHTvKosfwnigvL3/16pWBgcHp06cDAwPDw8Pv3LkTFxfHvsvlcu/fv29hYREfH3/mzJmJEydu3rzZxsZm+vTpohrWrVt3+PBhVVVVW1vb48ePs39pk7lt27YNHTrU2Nj4xo0bXl5e7M7ly5cPGjRo1KhRI0eOLCwsdHFxYf9IWWfhVmXKlCn9+vVjtxt1yfv6+np6ei5dujQkJGTmzJkRERE8Hq9RXyvgw/EhzlvR0dFhpwlERESsW7du0qRJpaWlRPT7778fO3asqqpq7dq1H3300e3bt0WHHDp06PHjxzo6OnPnzmWnutRXmMfj5eXlTZw4ccWKFQ4ODnfv3l2wYAERCQSCCRMmHD16VE9Pb+nSpS4uLlwu9/Lly3VWLk5CVBJUVFTMnDkzLCzso48++uyzz9j7fd7q1atXd+/erb2UKZ/PnzBhwqlTpxQVFT/77LP8/PwdO3Z4enpKbkhCl99dfeMsrqKi4vPPPw8NDe3YsWP37t23bNnC7q+srKxVX001TkpDmquvvwsWLBgxYsTLly+9vLz8/f1///33zp07s/MnRbZt26avr29hYREVFTVz5kz23isR6T4DDQxPiqpAhl5mVhKRtoYc+42aiBTkGSJKyWL3c/Q05YjIpuubP4ilZr35tPftpkREWhpyRNSls4K2hpyiAqOqzCGiotJq8SZMOykQEcOQSUd5IkrNlHS9VFeT98+5huNeKNrHt3dKDLxfQkRZef+b6tLDnEtE6qpvFjEpLquurxdsF8r4QtMJL3RHJc3bnEVEyRkV5YJ3zaDVGYPUx7K32GiqyXXWkScim66KDayqvnPHGtxbmYh0NeVFDUkYEMlVxSYLiGj2p2psYoWIhvZr5zZW7VVOlbt31qC5qdojEnf5/edJcACtwezZs2/duuXn55eXlxcbG6usrOzn5/fVV19t2LCBiL777jvxwqtXr16/fr2/v7+JiUlRUZG/v39QUFBcXJy8vPy9e/d++eWX48ePE9HZs2czMjJqN3Hp0iUrKysiGjFixJ49e1atWtWiXYX3jrOz8/Pnz3v06LF69epLly5lZWWtWbOGfWvMmDEWFhaVlZUODg7z5s1zc3MjIhcXF/Gv99u2bVu8eDG7WoeZmRn78MfWic/nsyuPMAxz5syZxMREWUfUUHp6erb/Ej3ttCEyMzPPnDljZ2fn6+vL5XIlL2EJH7immLfSqL8zs/+r66/WHfX1myAYidgH69anqqpq9erVq1evrv3Wixcv2OcTN6RwnU/nJSIejyf6OSu58gY2xGaIRJXUSMEmJiZOmjRJ9PKXX36ps/76xkQ8m05EOTk5K1asqLOkhIbq63KN4CWEJKGP9Y2z+CE8Ho+dwkdE7DNxCgoKRAuRSKi89kmprzlx9fX38OHDou3aSRkiSkpKEh/D2uHV9xmoUUzyR6K+8CScC2htjPTkiYhXUBWXIrAwejNhQUGeMdRl91dn51fpdpCLSnozDd5A780PeQV5on9/Wot/D6/teTx/6nBVvkAYkywQr6FOv4YUbzuR374d56S3vqGe/LqfckMel4lPJ5Lj/K9dyb1gu6CizLmwtaPcv2l/foVQXo6IKDO3qrisWkOVo63R6LVm64xBjsMQUfXb0iy1jzXUkyeivMKqzNwqfS25vxMF4uUlxFnfuWPfVVSo1VD9AyK5qtnj1E5eLfryu2x1Vc7koapExGHId53ewVW60S8EVx6UbDic6/1z3rIZHd7SeYCWNXz4cHaDXYOsrKzMxMRE9G5ycnJ5ebmSkhL7kn1aH8MwJiYmycnJKSkp7C2u2traenp6RCS6zSc1NVX0r6GoCYCmFRYW1qtXL1NTU1tb25kzZ44bN27z5s3+/v7x8fFGRkZExOPxsrKyiEi0EImhoaEo8ff8+XMiEs1oVlVVLSgokEE3amEvKw7nf3+M19LS2r59OxGVl5fPnTv3xIkTopvyahduVWqvZdvwY3ft2uXm5jZmzJhXr175+/u7uro2T4zQ5n2I9wTBh2P58uUWFhaPHj3S0tKaNm0aEf30008f4I0k8N4YM0ilm4liTLLAfl7aeAeVMr6QXyG8uK3jGHsVMwOFhLSKIV+kDe6ldPluCRE5D1PtpN3oH/I7/fKTXlXEpQiy8qpUlTlTh7eXULiy6s3VxBcII2L5YZF1rH3bwF74bdK3MFKMSxFs8c0bYduuqKT6aRxfTo7G2qsQ0cKtWZfvliyZprF7Wb2rODeKcUd5Ijp9rbC4rNrWRmnGKEndFOfUR/nfoU4d2EP58t3/PPBYQpz1nbv6Ghprr1LfgEiu6lNHFad+7dy9M2d4ZZ7f2nG8g0ro87KvtucM6aOsry0XmcAnIl1NPA4JWh1lZWV2g127TUVF5eLFi+z6kUTE5/NFd8gS0bNnz6ZOncrn82NiYthD2GXgeTxedna2rq5uVFQUW5Jdyr1GE0TE1lz91gwrQAO4u7tfuHDhxYsXL168CA0NHTduHBFpamoSEfuYAm1tbV1d3ezsbPa5M/Tv7eesOlfKaw3YVUhmzpypqqoaFhbm5+d37NgxAwMDX1/fO3fuHDt27MyZM717987Pz6+zsIyjbzrPnz+/devWsGHD9u/fj5WwQQLkVuB9FhkZOXz4cDs7u7KyspiYmCNHjgQEBMg6KADpcRWZu4cNtvjmBd4v8btepK7CmTNejYiUFJl7PxuuP5R7PbzU71qRkb6Cx1SNla7STEzY+IXWlfsl0S8EvS24e/9PR0/il3DnYe3vPy07c61o1QHe8P7tRg9UuXSnWEJ5Cb1gd24+knf1zxIf37z27TjWporuo96sMlstfHOgFD2q09IZHR7H8p/G8/f7FxSWqDU8t6Igz1zZ0/mLLVkPo8ufxvG3LNL+ans2EcnLMZLjrO/c1UfCgLy1KtdP2r8urlr8fc7U1RkB2zuaGylqa8idvVGUX1jdvh0zfEC7rR7aDewvQMsbO3ashYVFXFycj48Pu6bD06dP5eTkxOeT7ty5MykpKS4uLisrS1VVderUqerq6mZmZgkJCY6OjoMHD2ZvenV2dmYfGlgb+1fr06dPFxcX29nZzZgxo2V6B+8lX1/fgwcPxsXFZWdn9+rVi4gyMjLYp2cGBQUlJCSYmZndvXv3/v37EyZMIKLz58/XeHBB67Rr164+ffr07t3b09NTTU1NT09v3LhxSUlJixcvLi4uHjt2rLOzs6+v78SJE2sXfp9yK0Q0Z84cCwuLsLCw+gqIr+cSExPTwMUu4T3DNEGKtDH3BEnQAvcEAUBrIGEt20ZpDX/hYRffrQ5vFcvOvSOdkUm5r6t+3d5x4hBVWcdSU+exL8r51X+fNZZiJk6TY1fkZbd/vvT6i++yNdXkcq53YZjWFWcrwbGNp9ZxtTYWe3W3xcjbHG1t7dzc3ICAAPYbGhFlZ2dv2rTp6tWraWlp7du3t7a2nj17tru7u6jw9u3bAwMDHz58aGVltW/fPgcHByLKyspat27d9evXs7KyjIyMZsyYsXLlSvY2otpNJCQkfP7550+fPuXz+W5ubuKPIoLm0IYuqAYuqx8TE2NlZbVnz56lS5euWbNm+PDhlpaWWlpaeXl54eHha9euZXMrRKSnp7dx48aRI0fq6emlpKT4+flt376dfegvj8fT0tKaNGnSpUuXzMzM2Ef/Ghoais9qqU+jBrNRzwqQlcZ+PBrYqcjIyO7du9e+J2jbtm3ffPMNuyc2NtbS0nLv3r1ff/11jQNXrFjBLjEjEhwcPGrUqJ07d7KPDWWLiReIiIgQ5VkkaBOXAzQKcisA0NKQW2mdWm1u5e9Efq8ZKac26jd8dkmzWrY7JypJMKinUk5B1bHAwtJy4aYvtdbM1mxtcbYSyK1A06qdJYE2oQ1dUA35xm5kZBQXF8flcj09PdnnjbY85FbaRKckaBOXAzRKK11tCAA+KMXFxUIxN2/e5HA4hw4dKigoKC4ufvz4cfv2+LL64erRlVsdbt56Eha9LLgv0iu2ncg/e724R1fuSW/9b901qfXFCQAAzWHfvn2RkZFcLregoODixYuyDgcAWotmnrRsZ0d//klExOdTaCjNmkUNmOEGAB+acePGycvL+/v7x8TErF+/Pjc3d8SIEQsWLJg7d25ycrKjo2Nb/9NEm5BzvYusQ2gb3MaquY2VtFQKAAC8x8zNzUtLS+/du+fl5SX+kG8A+MC1yA3hmzfT48d09ixt3Uqff94SLQJAmxISEkJEAoEgJyfn5s2bRMQ+PLtPnz4JCQmbNm1il9AHAIAPHI/Hk3UI8KH75JNPZB0CALRGLXJPUGIiBQRQdDT17NkSzQFA23flypXg4OCvvvrqzp078fHx4g/RBAAAAAAAaFVaJLfC4ZChIZmaUmpqSzQHAG1feXn5J598YmRktH79elNTU3YaCwAAAAAAQCvUIvcEHTlCRMTjkbd3SzQHAG3fmDFjHB0db9++XVRURJgEDgAAAAAArViL5Fa8veniRYqPp/LylmgOANq+rKwsJycnDw+PysrK06dP+/v7yzoiAAAAAACAurVIbiU5mf7+uyUaAoC2TF9fX7QdERFhZ2cnw2AAAAAAAAAaqJlzK2FhhCenAgAAAAAAAMD7q0XmrTRMZmamrEMAAAAAAAAAAGicFnlOEAAAAAAAAADAe6oVzVsRCoWyDgEAWgKDWwUBAAAAAOA9wjRBRqOpviYhtwLwYWiq3EprSMgiTwTQEK3ham0s9upui5EDtEJt6IJqK/+yN2ow20SnGvvxaBOdkqBNXA7QKK1o3goAAABAq9LWf3cHAACAloHcCgCA9PA3BwAAAAAAwD1BAAAAAAAAAADSw3OCAAAAAAAAAACk1xT3BGG+CQAAAAAAAAB8qDBvBQAAAAAAAABAesitAAAAAAAAAABID7kVAAAAAAAAAADpIbcCAAAAAAAAACA95FYAAAAAAAAAAKSH3AoAAAAAAAAAgPSQWwEAAAAAAAAAkB5yKwAAAAAAAAAA0kNuBQAAAAAAAABAek2WW3FycmIY5vz586I9QUFBDMP069ev4ZU8ffpUVVVVcpl+/fqdOnVKyijrqqTOCsUjES/Qu3fvs2fPNqotKQ6p88Am6TgAAAAAAAAANK2mnLdiZWV15MgR0ctffvmlW7duTVh/M1m5cuWAAQPepUDLaCVhAAAAAAAAAIC4psytTJ48+dGjRykpKUSUnZ19+/ZtFxcX8QIZGRnOzs66uroGBgYeHh6lpaVElJqaOmrUKA0NDRsbmwcPHogKFxYWLly40NDQUEdHx8XFhcfjSWh69+7dAwcOFL0MCwtr3759SUnJrl27LC0t27dvb2JismXLljqP3b59+8OHDyVEIirg6ekZHR3t6elpYmLi5ua2b98+Ozs7UbHQ0FA1NbWSkhLxymscIqFfu3fvNjEx0dDQ6NSp0+bNm2sfKAqDiKysrDZs2DB06NDu3bv369fv77//Zve/fPly5MiRGhoaPXv2/PHHH986CQgAAAAAAAAA3lFT5laUlJRcXFx8fX2J6NixYxMnTlRTUxMvMHnyZAUFhcTExEePHj1+/HjZsmVE5OzsbGBgkJmZGRQU9NNPP4kKz5w5Mysr68mTJykpKerq6u7u7hKa/vzzzyMiIuLi4tiXx48fnzJlioqKioGBQXBwcGFhYUBAwL59+y5evCihkvoiEdm/f7+1tfX+/fuTk5OPHz/u6ur6/PnzqKgo9t2jR49OmzZNRUVFwiH19SsuLm7dunU3b94sKCiIiYkZM2ZM7QNruHLlSkBAQFRUlLOz81dffcXunDJlipmZWVZWVnBw8NGjRyV0FgAAAAAAAACahlAoFAqFUmwL/2vIkCGbNm168uSJsbFxVVWVpaXl/fv3d+/e3bdvX7ZAbGwsEWVnZ7Mvb926paSkFB0dzTBMXl4eu/PEiRMqKipCoTApKYlhGFHh169fMwxTUFAgFAr79u178uRJYS3jx4//9ttvhUJheXl5hw4dQkJCahRYuXLlwoUL2W3xStjt2NjYOiOpUbhXr15nzpwR1enq6rps2TKhUFhcXNy+ffs///yzdmDih9TXr8TERGVlZX9//8LCwjoPrBGGpaXlzz//zG5HR0crKysLhcKYmBgOhyOq4dy5c6IuiHvH845tbGMb29jGNraxjW3RtvgeAAD4YDXxc4J69+6tpaW1fv16hmHs7e3F30pLS1NVVdXR0WFfdunSpby8/OnTp1paWh06dGB3mpubsxvJyckMw/Tv39/ExMTExKRnz57q6uqvXr2S0LS7u/vJkyerq6svX76soaHh6OhIRJcuXXJ0dDQ2NjY2Nj58+HBOTk59h6elpdUZiWTz588/depURUXF+fPnDQ0NxW8RqlN9/erSpcuJEycOHjyor6/v6Oh4586dtzatra3NbigrK5eVlVVWVr569UpTU7N9+/bsfiMjo4Z0AQAAAACkxv5KzTAMwzDsHmxjG9vYxvb7vV0nefZ/QrHUe2O3a5g3b95XX321ffv2GvsNDAyKi4t5PB6bFHjx4gWXy+3Vq1dBQQGfz+dyuUSUmZnJFjYyMmIY5tmzZ+rq6hKiFzdu3LgFCxb88ccfx48fd3NzYxgmIyNj6tSpV69e/fjjjxmGWb16dXx8fH2Hd+7cuc5IauBw/pONcnBw0NTUDAwMPHr06Ny5c996iIR+TZkyZcqUKWVlZXv27Jk0aVJubm6Ntt6qc+fOeXl5RUVFbHqFXfimtnc519jGNraxjW1sYxvb2BbflvyrNgAAfCCaeN4KEc2aNev69esLFiyosd/S0tLOzm7p0qXFxcVZWVlr1651c3Oztrb+6KOPdu/eTUR8Pn/v3r1s4a5du44cOXLevHkZGRlElJOTc+HCBcntKigozJgxY/v27Tdu3Jg1axYRFRUVEVHPnj0ZhklPT/fz85NwuJWVVZ2R1KCvr//PP/+I75k/f/7GjRvDw8NdXV3fekh9/YqNjb19+zafz1dWVtbR0VFQUKizLcksLS0/+uijVatW8fn89PT0HTt2NPxYAAAAAJCCUIz4HmxjG9vYxvb7ul2nps+tqKioDB8+vMYqtqwLFy6Ulpaampp+9NFHPXv23LVrFxGdP3/+1q1b/fv3Hzdu3Pjx40WFz5w506lTp0GDBrVv397Ozu7+/ftvbXr27Nk3btywt7c3NTUlIgsLi/Xr19vb29vb23t6eopXXqf6IhG3atWq06dPd+jQYcqUKeweNze3mJiYTz/9VHS7k+RD6uxXWVnZN998o6urq6mp+csvv1y4cIFhmNptScYwzIULF+Li4nR1dUeOHDlz5kx2Dg4AAAAAAAAANB9GcuoF3qqystLAwMDX1/eTTz6RdSz/cebMme3btz958kTWgQAAAAAAAAC8z5p+3sqH5pdfflFXVx81apSsAyEi+uuvv5KSkogoKSlpy5YtLi4uso4IAAAAAAAA4D0nL+sA2rYOHTqoqKj4+fk1dt3ZZpKWljZlypTXr1+rq6u7uLgsXbpU1hEBAAAAAAAAvOdwTxAAAAAAAAAAgPRaxWwLAAAAAAAAAIA2CrkVAAAAAAAAAADpIbcCAAAAAAAAACA95FYAAAAAAAAAAKSH3AoAAAAAAAAAgPTe6RnMDMM0VRwNhKcaAQAAAAAAAECrgnkrAAAAAAAAAADSe6d5K6yWmUvS8nNkAKBZNeFFjRltAAAAAAAgQ5i3AgAAAAAAAAAgPeRWAAAAAAAAAACkh9wKAAAAAAAAAID0mmC9lZaEVVcAAAAAAAAAoFVplnkrwcHBDMM8ffq04YfweLwNGzZER0c3RzwAAAAAAAAAAM2ktcxb4fF43t7eNjY21tbWEorhaSAA7w1MQwMAAAAAgPdDM6638sMPP2hpaTk6OpaXlwsEgnnz5mlqaurr6/v4+BDRkydPunXrxuVyjYyMQkJChg8fTkTOzs5mZmbNFxIAtGZ2dnZCodDd3b3GRnV1dUFBwe3bt+3s7GQdY00MADSArK9Uach6zADeQ7K+rAEAmlEzzlthGMbb29vT0/P69evJyclHjx69fPnyP//8s2LFCkdHx+Dg4NevXz98+LCyslJTU/Ps2bMODg6HDh0aP35884UEAG3R999/HxkZuW7duuDgYHNz85ycHFlHBAAAAAAA8D/NmFuZP3++tra2p6dnfn5+ZGSkhobGp59+2rNnzxUrVkRGRk6ZMuXKlSt9+vQxMDDw9fXt1KkTEbETW5ovJABoi6Kjo0+cOJGcnHznzp0ZM2bs3btX1hHVVB1uLusQAFopjm28rEN4J7gZGaBJtK1JK20i2ob/dGoT3ZEAP4ehrWjGe4IUFBTYK1koFNrY2BQUFFy9evXXX38lou7du8vJyfn7+7948UIgEJw/f15JSYmIEhMTeTxe84UEAG1XTEwMERkbG8s6EAAAAAAAgP9oobVsv/jii6dPn86YMUNRUdHb29vR0fHkyZMrV67Mzc3t0qWLm5ubkZHR8OHD16xZc/v27eDg4JaJCgDaECsrKyJKSUmRdSAAAAAAAAD/0Sy5ldGjR4vmbok2jh49evToUVEZV1dXV1dX8aNu3LjRHMEAQFtnbW3t6uq6du3a169fnzlzRtbhAAAAAAAA/EdreQYzAEB9li9fXlxc/OTJk1mzZmVlZck6HAAAAAAAgP9AbgUAWouwsDDRcmu1NwAAAAAAAFqnZlzLFgAAAAAAAADgvYfcCgAAAAAAAACA9JBbAQAAAAAAAACQHnIrAAAAAAAAAADSa4LcCtMi3j3O1sPGxobtlIKCgra29sCBA3fv3i0QCCQcoq2tzTDMpUuXiCgqKmrgwIEqKioMw0ycOJGIRo8ezTCMl5dXUzXHklxtA9WuRLwvrVADw2vlvQAAAAAAAIAWg+cEyYyenl7nzp2Tk5PDwsLCwsL8/f1v3brVrl27GsUqKioUFBT69etXUFCgqalJRPv37w8LC+vZs+e8efO6du3atM21APG+tF1N3gv2RDdVbQAAAAAAANBi3mneirCRpDuqdg3vhylTpkREROTk5Bw8eJCIwsLCtm3bRv9OiNiyZYu9vb2iomJISAgRPXr0KDw8PC8vz8nJ6dChQ0T0/PnzxYsXP3z40MbG5tq1a0Tk4+PDMIyNjU2jmiMiHo/n6elpZmamoqLSq1evY8eOEVGd1dZZkpWfn798+XJLS0tlZWV9ff3Vq1fXV4moL0SUk5OzaNGirl27qqioWFtb+/j48Pl8tkLROAwfPpirmYEAACAASURBVFxNTc3GxiY0NLTOrtUZVUBAAMMwWlpa6enpVVVV9vb2DMN8/fXX4pX7+PgMHTpUTU2tb9++dVY+d+5cY2Pjdu3acbncbt26/fDDD6K3xHshIVQJI1bniQZoQj2mv+TYxnvuyBHtMZ3wgmMbv/oHHvvy6105H32eojk8UXVIQs8ZLw+cLxA//MGzsonL03VHJXHtEzqPfTFzbWb0i7rnu7ENcWzjvz+dL9o5evErdueKfTwi+nhhGsc2furqDPEDHRakcWzjZ67NlNyR2GQBWxWvoKrOAjojkzi28ZfuFLMvo5IEg+amqg5J4NjGT1qR/smSVxzb+LU/5UpupYFtsRpVZ8MrqdERAAAAAGgrsN6KjHE4nIULF44cOZKIzp8/L9rv5eWlpKQ0ZcoUVVVV8fJTpkyxsrIioh49eixZssTOzs7V1dXY2JiIBgwYsGTJEldX10Y1JxAIHBwcDhw4oKent3Llyvz8/NmzZ//888+1q62vJFuJo6Pjzp07eTyei4vL4MGDo6OjiUhybHw+f/DgwT/++KOcnNz06dOzs7O9vLzc3d3Fy2zcuNHa2tra2joqKsrNza12j+qLatKkSR4eHnl5eXPmzNm6dWtoaGi/fv22b98ufuyOHTucnJxGjRr1+PHj0aNHZ2dn16j8+fPnTk5Oq1evdnd3j4+P9/DwuHXrVn1jWztUCSMmUt+J/hA0YeJVth1puw4FvC7jC8cMUrHpyo1MFCz+PudQwGv2rXM3ipy+TPvtXom8HA3qqcQXCM9cL+rvlhL6vFxChQf8C6qqiYiikgTXw0vF35o9Xp2Ifr9fkl9Uze5JTKt48KyMiGZ/qvYuvaioFPbrxrXtrqSpJicKIyyy3MxAce//6cyfpP4ulbewGh0BgObTJDc+AwAAiOCeoFbBwsLi+vXrmZn/++Pt7Nmzjxw5Urukh4dHWFhYbGzsoEGD9uzZQ0SjR4++ffv2y5cvR4wYsXnz5sY2d+XKldjYWGVlZT8/PxUVFUNDw7lz53733XdJSUk1qg0ICKiz5Pz584OCgiIjIxmGefDgAZv6qaioIKJVq1b9P3v3HdfUuf8B/BtGEvYKILIEAREVvWoFF+JG1ForqGAVUGuvFqx4rf6sOKji1TpoHW1tr+LCgfteRMCF1VJw1YFIAZUhKBD2TBj5/XFsmkIIEdl+3i/+COc84/ucJJA85xkyYouIiEhOTlZSUrp586ahoeH06dOnTJly4sSJnTt3GhkZMWlWrVoVGBh49+7dDz74IDU1taioSFtbW7KQxuL/9NNPt2/fHhsbGxUVFR0drampeeLECTabLZl39erVq1atEolElpaWaWlpYWFhvr6+kglu3rzJ5XLr6uoqKyuzs7PDw8MvX748duxYqVe1YajXr19vLLYmn2iANhDxrfHoQSpEJBLRPz7JeJQqiI6r+Gy6VqVA5Lstr7aOxgxW/d/O7iocVkFJrdOil4kvhP/ckvPomLnU0nQ1FTNe15y5VjZznHrwsUIi0tNSzC9+M/rDbYy63zaFkvK6E9Gli2doEdHhiBIiMu+mNPYDKZMTUzKrP9ucczuxqpc5e+msv73r9Sc8zy+u3bRYL+LX8juJgvDg7nefCvKLawtKaolo9OKXN+5XEtGjVMEXO/4asxMUUhAUUtDHkv34eP34ZdTFL6r9+j8Fl34rf8WvsTJhL/PQ9p6i2c8j/clzYcMypSYmosLSus0hBf+7WZ7xulpLXcFnqubmJTyphUg2JK+wdsPP+VFxFa/za8y7KXu6aKyYo8NhsySvwPW7lbefVJl1U9q32nCYPVfO5x2gI2ubGbLMpxQAAIAWhHErHcIff/xBRN26dRMfGTduXNtUl5GRQUSVlZU9evTQ19dfsGABEaWlpVVV1b87LSNleno6EfF4PKZjhYjk+WDEFMjj8QwNDYlIPJspMzNTnMbe3p6ItLTe3HkuK6s/VF5GVBwOZ8WKFUQkEok8PDwark1jYWFBRCwWq0ePHuKixOrq6jZs2GBiYqKsrKyurh4eHk5EOTk5jTWnYajyXNtWfaIBiOj0tdKh8zOZn1f5f5vnwnSsEFGdiCoFdUTUy1yZiOISqpg+kVVeOiocFhHpaiouna1NRAnPhGmvpH8n8ZmqyVZmfXu8MLewNjSqdJAtZ3BvjvisCoc1e7wGER2+WEJEIhEdiSglIu8pmg0XK6+uEU1elhVzv9JAR8nWnC3ZRSK29sd8LlvBbYy6usrf/pHNGKNu24NNRP16cr6Ype08UMW8mxIRDenD/WKW9ieT6o+RkVGXsFrk9NnLPaeKDHUVv5yrW1haO39jzs/niz+ZpNmwzMYSC6tFoz7L3BFayC+qnT1BY0R/lcTnQiKSWoiYQCgauejlD2eKFRXIY4JGbmHt2h/zfb7+29+fjfsL7CzYdhbsJ8+F3oFNzKsC6LCkzpBtbEat7Cm98kw0Zmphs9nyzKcGkC0hIUEkEu3Zs4eIjI2NDx06lJWVJRQK+Xx+fHz84sWL2zvAtyPZHEZaWppIJNqyZQsRFRUVSR07fOLECXFekUjEfPZmREVFMQe3b98utXyALgbjVtpZXV3dTz/9dPnyZSJyc3MTH1dRUZG/EEVFRaaoZlRnampKRGpqamfOnGHKISKBQKCkpFSvWBkpzczMiIjP5ycnJ9vY2JDEfScZsTEF8vn83NxcAwODJ0+eMMdNTEzqNU3GRlEyosrPz1+5ciURcbnc/fv3z5s3b9iwYZJ5Hz58OHPmTIFA8PTpU3FRYmfPnt26dauGhsbRo0dNTU3Xrl0bExMjY/pJw1BlxCZO81ZPNEAz5BTU5hTIWjqkuka0aHNuSmZ1f2vOam9dIsotqGFOmRr89Vo1M3zTYZpbUNvDSErnqaGeoscEjUMXS+Z/nSMQipZ56By9VCKZYP6Hmj+dL45/UpWUJnxdUJv2qprFIu+pUubsxNyvTH1ZraBAN/aZmBoqhYSXLNhYv1vTZ4rmfwIMG+b1ddeOT6hKShMOs+cGL9cnoklfZKW/rhk/RHXjP/Xeqq6Lv5YnpQlVOKzQjUZqXJapodLCTTlbDhU+O9cj5l5FvTIbS6yvo5jwTMhi0a2fTZhOn+oaERGtmqfTsBCxiNjy5AyhkiLrl59MDXUVP3Iun7o8+8Tl0h3LeEa8N0/Kqrk6Gxbp3X1aNcQ7M/VldVFpnbYG7pdAZxUQEDB69Ghmhiwzo5YZojt37tz9+/f7+PhUV1eLR31u27bN39+fx+OdPn3axcUlNTXVwMCAmWicnJxsbW3t4eFx/vz5gICAhISE48ePN6yloqLiwYMH6enpQ4YMGTp0qHi0LECznTx5cvjw4Xw+PyYmRktLy9bWduzYsT/88EN7x9Vi7ty5o6GhQUT9+/fncrmvX79m7q2mpKRIJvPz8wsODq6tre3Tpw+zCgHA+6OF+lbk2yNZ1GTi92ndhNOnT//222/p6en5+flE5OjouGrVquYVxaxpEhoaWlZW5ujo6OnpKX91kydPtrGxSU5ODgoKmjBhQklJyYMHDxQVFSdPnlyv2BkzZjSW0tXVtXfv3k+fPh02bNiHH35YWVkpEAjOnj0rOzZXV1crK6vU1FQnJ6cRI0ZcuHCBiNzd3bt37y5/22XE7+3t/fLlSx8fn1GjRnl7e3t4eDx48EBHR0ecd8eOHc+fP09OTs7JyVFXV585c6ZkyTU1b75eCgSCe/fuxcXFyR9Vk7G9bVEAzfa5m/buL/WZxxbTXqS/rpE8m19c6/5/r2LuVzr9Q+X8tu4aqgpEpK/zpiswM7eG6Q4gooycN8NVDHQbXQ3E31P70MWSiNhyY32lmePU6/WtDOnD7duTnfBMeDii5BW/lojGfqDKjN2oJzOnhoj0tBRNDZWIaGAvTsM0Y4e0zDZnMurKyKkhokqByGLaC/HBtFfVVUIp/6oaS/xHejUR8bQVxVdSWanp/5hMaTxtBUNdRSLq25Mjjlbct9LPmkNEWupvno6ySvStQCcmOUNWxhxkJoHUKb3yTDSWrMXFxeWt5lMDyMDlcpkbeGPGjHn8+DERsdlsZlg0Eenr6wcGBk6cOLFbt27p6emhoaHbt29nBlXx+Xw9Pb01a9aMGTNmyJAhGRkZixYtamz3hvY1fvx45kFSUlKvXr1Onjwp3iNCrKCgwMzMbMaMGWFhYcuXLyei/Px8PT0pNzYAuiR8Dms3OTk5Dx8+JCJHR8fg4OAbN26oqak1r6jly5c7ODjk5OTs3r07Ojr6rarjcDg3b9709fXNzs7etGlTSEiIQCCYNWtWw2JlpGRO+fv76+rqHjt27Nq1a7169WoyNi6Xe+vWrUWLFgkEgmPHjunp6QUGBh4+fPit2t5YVMHBweHh4ZaWlrt27fLy8nJ3d8/IyPDx8ZHMu3Hjxuzs7MTExAEDBkRERDBTk8Tc3d2XLFmipKS0cuXK+Ph4FxeXtwpM9rUF6AiePBcO8c6MuV+5cJrW5T3G4m/mQ/upMMupbj1UWCkQEVFBSe3uk0VEZGfBljpohWFvxWEWT/GdqS21B4FZf+TwxdIz18qo8VVsTQyUiKiwpK64rI6IXmRLmYXETFaSh6ICi4gaG9gnoy5m2I6aisKl74yjd7/5+e+O7kqKUspsLLGlsTIR8YtqkzPe7LLEjFuRHRhTGr+oLrewloiePH8zr8HE8K+uKEUFInlvbQB0dJIzZJucUSt1Sq88E40xDxdaSVVVVVZWFhGFhoZu3rx52rRpqqqqycnJRMThcG7durV48eLa2trjx48bGBhs2rRJcuNIIlq3bl1iYmJiYmKfPn0OHTrULk0Qc3Nzi/vT2w7pCgkJEQqF/v7+BgYGc+bMuXfv3t27d1spToAOCHOC2kFCQoKMs3w+X/bBo0ePHj16VPKslZWVjFEVsqsjIgMDg927dzc83rDYxlISkZ6e3s6dO3fu3NlkIZJtMTQ0ZLaUbkgymZWVlYzJOI1F5e/vL34cFhbWMIG1tfWNGzdk1Lt3717JfZcbSyYjVBlXTOoT/V5pONVLxuQvGbBPULONXJRZVFqnymVl5lR/+K9sIrIxY3/3L30VDmvXCv15G15fu1vRc/oLW3P2o1RhQUktl8364f8MZJe5P8AwOUPo2E/6uqpzJ2mu3pufza8hIm0NhenO0rfHch6oYmms/DyresaqVxOHqn7/982h35a5kRIRhUaVlFXWOfTlek7UkLOuycPVbMzYyRnCzSEF4x1US8vrHiQLFBVp8nC1hmU2lvjcN91792A/TRMOX/jyw5FqlQKRoFp0ZquR7MBch6tZmSinvqwe9dnLEf25F34pJyL3serdefivDV2T5AzZJmfUSp3SK89EY8la5J9PDSAPd3f3ffv22dvb9+vXj4iEQuHXX38dFBTk6upqY2NTU1MzcuTInJycc+fOhYeHz549e/ny5a9evWLybt26df369YMHD75z546VlZW2tnZR0Tv943sXhoaG9W43yu/169fHjx/38vIKCQnhcDjBwcGyNzAF6GIwbgUA4D1VWl5HRBVVoqi4CuYn9lElc8pzosb1H0ymjFCrrqFbD6vYyjRrvMbtg2YjBzSxQpBZN6VxQ1TrrS8rpq+jOHWkmrgKLlt6bxpbmXUxuLvTP1RuP6k6Flnq76kjNZmc/D11HPpwcwpqd4cVXf77ztCy6+KwWb/8ZOLrrp3NrwkKKQgJLxFUi2aO05BaZmOJmeP+Htq6mgrHokuv3a1gFgyWHRiXzbr5s+mij7QEQtGxqFI9LcUNi/QObehGAO8BZkZteXl5UFBQXFxcdHT0N998s2fPHnHfyo4dOzw8PIYNGyY5pZeZaFxTU+Pk5LRw4UIvLy+SOdFYPGd56dKlx44da6vGQZcVFxfXv39/S0tLDw+P8PBwNpu9adMma2tr8aKEzH4I4lueksv8PXr0iIiKi4uZX9XVpd94aBt79+5l/YlZUeWtMPdZXV1ds7KypN7aBOjCcAcMAKBrarjZ8IsLFpK/Vv9mLSP7yAEqTfakNFYR49J3xg0Pnt4i1wDjXubsmB//utssuTVyXrRlvcT1jhwJ7HYk8K9uCCsT5d8OmFLjZNRloKO4a4U+kX69LFLLbCyxnpbijmX6O5Y1XYhkQwx1FX9cLX2UkGQyKxPlunhZzyNAp8PMqN24ceOlS5c2bdqkoaFhZ2cnOat348aN4eHhzJTeXbt2MffYmYnG69ati46OPnbsmJmZmZ+fH7OkvVTLly+/f//+gwcPdu/eXVJSInWtOgD5eXt7nz59+sWLFy9evIiNjZ0yZQoR6erqMrPSeDyegYFBbm5unz59mPQvX74U562traWuMg730aNHV69eHTt27O7du7HZObxv0LcC7yPMxwEAAOgIpP5HljGjlhqZ0ktyTzRmyJ5PDfC2QkJCvv/+++Tk5Nzc3P79+xPRq1evHj9+/PDhw9TUVCsrq19++eXWrVvTpk0jolOnTmVnZ7d3yK1l/vz5NjY2jb2/3NzcBg8ezDx++vRpvcUQATo19K0AAAAAAAC8HWbVHmZ0RkBAwLhx43r16tW7d++CgoILFy6sXbu2oqKCiEaMGPH1119PmDDB09MzIyNj9+7d33zzTTuH3poyMjKYtaWlklzPRTzRD6BrYLXM8DMZK1A6OtJvvxERCQQUG0vz5pHEELj6usRYOACQR/NWrm2ofcfQMq3ApAyAxig4pFB7v0+bh3l3d8bIuzAej5efn3/u3LmPPvqovWOBt9O53lDyfEQxMzNLTk7mcDh+fn579uxpg6jqkf9ittQnrvbSWV42AG21lu2mTeThQcOH05YtbVQjAAAAAHQhfD5fJBKhYwXa3a5duxISEjgcTlFR0ZkzZ9o7HADoENqqb+XZMzp3jhITyd6+jWoEgM6mrKxMJOHKlSsKCgr79u0rKioqKyu7f/++hoZG06UAAAAAtCZra+uKioqIiIgxY8aIt1IGgPdcW01yU1AgU1OysKBff22jGgGgs5kyZYqSklJYWNjTp0/Xr1+fn58/fvz4RYsWLViwIC0tzcnJqbMPagUAAIAuYNKkSe0dAgB0OG3Vt7J/PxERn0+BgW1UIwB0NjExMUQkFArz8vKuXLlCRDNmzCCigQMHpqambty4kdmkEAAAAAAAoENpqzlBgYFkb0+mpnT7dhvVCACd38WLFyMjIz///PMbN26kpKSYmJi0d0QAAAAAAAD1tVXfSloaPX5MVVVtVB0AdAlVVVWTJk0yMzNbv369hYUFM4wFAAAAAACgQ2n9OUFxcbJ2aAYAaJyrq6uTk9P169dLS0uJiM/nt3dEAAAAAAAA9bXVeisAAG8vJyfH2dnZ19e3pqYmNDQ0LCysvSMCAAAAAACojyUSiVqimJYZmYLxLQDwtlrmj1hzMVsX1cVbt2MMAB2ZgkMKtff7tHmYd3dnjBygA+pcb6hOsS+h/BezUzRHhs7ysgFoq/VWAAAAAAAAAAC6oo41Jwi9kgDvj85+FwUAAAAAAIDRseYEEfpWAN4bLdW30hHmBAGAbJ3x3gne3QAtrrP8KegUb3/MCQLoaDAnCAAAAAAAAACg+TrWnCAAgM4F91IAuiq8uwEAAEB+GLcCAAAAAAAAANB8LTRuBfd2AAAAAACg8+tiw9a6WHMAOqwWWssWAAAAAAAAAOC9hDlBAAAAAAAAAADNh74VAAAAAAAAAIDmQ98KAAAAAAAAAEDzoW8FAAAAAAAAAKD50LcCAAAAAAAAANB877QHM4vFaqk45CQSidql0jauEQAAAAAAAAA6C4xbAQAAAAAAAABovncat8Jom2Ed9YartEulANCCWvD9hcFlAAAAAADQjjBuBQAAAAAAAACg+dC3AgAAAAAAAADQfC0wJ6gttcskHcwMAgAAAAAAAIDGYNwKAAAAAAAAAEDztUrfSmRkJIvFevDggfxZ+Hz+hg0bEhMTZSeTZ8XKuLg4FouloKCgra09evTouLg48Sk7OzsWixUSEiJOdvDgwSYLFAFAK2jyrQcAAAAAANApdJRxK3w+PzAwsMm+FfmtWLFi165dmZmZLi4ueXl5RPT06dOnT5+yWKzTp0+3VC0AAAAAAAAA8J5rxb6VvXv36unpOTk5VVVVCYXChQsX6urqduvWLSgoiIh+//333r17czgcMzOzmJiYcePGEZG7u7uVlVWL1G5nZzdv3rwDBw4UFxcfO3aMiJgulZkzZ165cqW4uLhFagGAFuTo6CgSiby9ves9qKurKyoqun79uqOjY3vHCAAAAAAAUF8r9q2wWKzAwMCbN29GR0f/+OOPBw4cOHTo0IoVKwICAm7evHn69Oni4uLbt2+fO3fO3Nz8xIkTRLRv375bt261YAy9e/cmovT0dCI6c+ZM7969fXx8hELhf//73xasBQBa1fbt25cuXWpqahoZGamvr9/e4QAAAAAAAPxNK+4T9Omnn/J4PD8/v8LCwoSEBG1t7alTp9rb23/55ZcJCQlubm4XL14cOHCgiYlJSEhI9+7diYgZ2NKCMSQlJRGRmZnZs2fPHj586Ofn5+TkxOFwTp8+vXr16hasCABaT2Ji4uHDh9PS0m7cuOHp6fndd9+1d0RvYBMxAHl0xvWV8O4GaHGd8U8BdCVd4A873kQdXCuOW1FWVmZewSKRqG/fvkVFRZcuXTp79iwR9enTR1FRMSws7MWLF0Kh8NSpU1wul4iePXvG5/NbpPbExMQjR44sWLBAS0vLw8ODmRC0e/duVVVVgUAQFRVVWlraIhUBQNt4+vQpEZmbm7d3IAAAANCVsaCzae+XDABRq45bkfTZZ589ePDA09OTzWYHBgY6OTkdOXJk5cqV+fn5lpaWXl5eZmZm48aNW7NmzfXr1yMjI9+9xu3bt6urq//jH/84fPiwoaHhmTNnLC0tz5w5Q0SxsbGff/55eHj4u9cCAG3G1taWiDIyMto7kPrq4q3bOwSADkrBIaW9Q3gnuEMI0CLw1RcA3gesd/ncwPpzWErLxdN0Xe1SaRvUBfC+afhJy9HR8bfffvPx8UlKSpJ8sG3btsePH69du9bAwKBXr145OTn1MrbXm5RpAvpWABrD9K10xn+j+AAA0II63RsKnUGdjjyvri7wtHaiN9H7qY3GrQAANNuKFSvKysp+//33efPmNexYAQAAAAAAaF/oWwGAjiIuLk58S6HhAwAAAAAAgI6pFdeyBQAAAAAAAADo8tC3AgAAAAAAAADQfOhbAQAAAAAAAABovhboW2mXTcvbpVIi+vjjj5lTvr6+jV2QpKQkJg2fz3/3y9uCeDwei8U6f/58s5NlZWV5eXkZGxuz2Wwej+fg4PDDDz+8VeEy6mpeCQ3LcXFxYbFYAQEBzSinBb371QYAAAAAAIBOAeNW3kJeXl54eDjz+NixY1VVVW1Tb3V1dYsUMnjwYAcHB11d3WYXMmvWrMOHDwuFQmdn5549eyYlJV29epU51bzC3z0k5uK8ezntpTUib5EXDAAAAAAAAMjpnfpWRG+pebnqlfAu2ZtdKePIkSPV1dU9e/bU0dEpLCw8d+6c+FRKSsro0aPV1NQGDhwYHx8vPj5//nwWi7VgwQLm1+joaBaLpaenJxAIiIjP5/v5+VlZWampqfXv3//gwYNMMmYsw+bNm4cPH85ms2NiYohox44d1tbWXC5XU1PT3t7+8OHDRLRgwQJzc3NVVVUOh9O7d++9e/eKq25YyN27d+Pj4wsKCmRnbExVVVVsbCwRXbt2LTo6Oj4+Pi8vb/PmzcxZycKJSFtbm8Vibd261dHRUUVFZeLEiVlZWX5+fpqammZmZkzwDXNJaixCqRdHXE7fvn2joqKIKCgoiMVi9e3bd+7cuZLjjCIiIlgslpGRUW1tbb0aG3s6zp07xzxr2dnZtbW1w4cPZ7FYy5YtEwcTFBQ0evRoTU3NQYMGMZdI/ubUuwLi1o0bN05TU7Nv377iAhsLr7FrAgAAAAAAAG0D41beQkhICBHNnj17+vTpRHTgwAHmeHV1taura0xMjIGBga2t7dKlS8VZPv30UyI6c+YMM8jl6NGjRDR37lwOhyMUCkeOHLlnzx5DQ8OVK1cWFhb6+Pj8/PPP4rwBAQFcLtfNzU1dXT0+Pn7FihVlZWVfffWVr6+vubl5amoqET169MjZ2Xn16tXe3t4pKSm+vr7igSQNC5E83mTGhrhcrrGxMRHNmTPnq6++unDhQkVFhY2NjYwsa9eu1dbW5nA40dHRffv2vXjxop2dXWZm5qeffvrq1SvZ1cmOsLF2zZ0719zcnIiGDBnyxRdfzJ079/PPPyei0NBQ5ik4c+YM0wRFRUXJjDKejunTp/v6+hYUFMyfP3/Lli2xsbGDBw/+5ptvxHm3bdvm7Ow8ceLE+/fvu7i45Obmvm1z6vn666/t7Ozs7OyePHni5eUlO7wmrwkAAAAAAAC0KqX2DqDTuH37dkJCAhHNmjXr1atXBw4cuHr1anp6urm5eUxMTGpqqoKCwi+//GJqahoSEjJ//nwm19ChQ/v06fPkyZPw8HBXV1dmqMvChQuJ6OLFi0lJSSoqKseOHVNTUzM1NV2wYMG///1vpjuGiHx8fPbv3888vnLlChFxOJwePXrY2NgEBASoqqoS0c2bN7lcbl1dXWVlZXZ2dnh4+OXLl8eOHSsOW7IQSU1mlOrUqVOfffbZo0ePHj9+TERsNnvdunVr1qxpLL2/v//WrVvXrFmzefPmqqqq27dva2pqampqCgSCR48eGRkZyahLdoSNtWvVqlXXr19PT08fP378pk2bmIMDBw68f//+2bNnZ82adeHCBSLy9vaul1H207F9+/bY2NioqKjo6GhNTc0TJ06w2Wxx3tWrV69atUokEllaWqalpYWFhTVcjuetLviqVasCAwPv3r37wQcfpKamFhUVXb9+XfarRcY16bAkB4UxyxtJHgEAAAAAAOgsP/yBzgAAIABJREFUMG5FXsy3Vltb2379+o0ZM4bH44lEImYkS2ZmJhHp6emZmpoS0cCBAyUzMj0pR48evXDhQllZmaOjY9++fYkoIyODiCorK3v06KGvr8/MG0pLSxMv4zJu3DhxIaNHj/by8mKWkh06dKient7OnTvr6uo2bNhgYmKirKysrq7OrAWTk5MjWbtkIWLyZJTK0dHx4cOHz58/P378+JQpU4RCYUBAQEpKSmPpBw8eTEQ8Ho+ILC0teTwem81mRlWUlpbKqKjJCKW2qzFLliwhogMHDsTExOTn5w8aNIh5CiTJfjo4HM6KFSuISCQSeXh49OzZUzKvhYUFEbFYrB49eoiLeqvm1GNvb09EWlpazK9lZWVNvlre9poAAAAAAABAS0HfilwqKytPnjxJf+4BpKyszOwBdPDgQZFIZGJiQkSFhYXFxcVE9OLFC8m8zAygS5cuMUtsMF0tRMR0xKipqUVGRl7+0//+9z8lpTeDiVRUVMSFKCgoHDx4sKSk5O7du4GBgVVVVRs2bDh79uzWrVtLSkqOHj168+ZNZ2dnanDnX7IQMXkySnXw4MGysjILC4vZs2eLVwyRuloKQ1lZmf4cksA8llOTEUptF4OZ7FNXVyc+4unpqaOjc+3atR07dpC0QSvU1NORn5+/cuVKIuJyufv376+3qMrDhw+JSCAQPH36VFzUWzVHahMkd6pq8tUi+5pAJ9LPI13BIcVvW574iMW0FwoOKav3vtl3bNnOvH98kqE77pn6qFR7z/Q9p4oks//6sPKjFdkGE59zhqcaT34xZ+3rxBdCGRUpOKRsDy0UH3RZmsUc/HIXn4jGLH6p4JAyc/XfZvCNXPRSwSFlztrXzWtgUpqQqYJfVH/No/alP+G5gkPK+RtlLZKsZbM0owSpGSd9kaXgkLL2x/y3LQegi+kgWwpCZ5eQkCASifbs2UNExsbGhw4dysrKEgqFfD4/Pj5+8eLF7R0gEZGtrS2zgiRzs1OGyMhIkUgkHvfdUon5fL5IJProo4+acbZdSD6tjLS0NJFItGXLFiIqKiqSukDniRMnxHlFIhFzU5YRFRXFHNy+fXtjVUDXgDlBcjl9+nRxcTGLxZo6dSrzjbeqqioqKio9Pf3q1avOzs6WlpbPnz//+OOPXVxc6q0Lq6enN3369BMnTvz6668aGhqzZs1ijk+ePNnGxiY5OTkoKGjChAklJSUPHjxQVFScPHlywwBiY2OXLFkyatQoIyMjZj6OgYFBTU0Nc1YgENy7dy8uLk7O5jQ7o4+Pz5IlS2xsbAwMDJgOBSMjo379+smZXX7NjpCImPVWQkNDmVFCnp6eKioq3t7ewcHBly5dYrPZnp6eDXPJfjq8vb1fvnzp4+MzatQob29vDw+PBw8e6OjoMHl37Njx/Pnz5OTknJwcdXX1mTNntmBz5AkP3iv7zhWbd1N2HaaW+rL69pOqpdvzlJVYn03XIqKTl0s/Wfe6to666SkOs+c+ThUejy49f6Ps8h6TYfbcxgrcE1bk76GjqEBPnguj4yskT/l8qBVzv/J/t8oLS+t0NBSI6NnL6l8fVhKRz1TNVm6ovKprRMpKrKbTNVXI4N6cotI6XU3FplO/pWaU3CLBtGqjAFpJdXX1W92MaV4VrVo+vJ9Onjw5fPhwPp8fExOjpaVla2s7duzYH374ocUrmjZtWlBQ0NixY+UZcg4t7s6dOxoaGkTUv39/Lpf7+vXr9PR0Iqo3kN/Pzy84OLi2trZPnz4TJkxon1ihzWHcilyYZWsnTZp04cKF8+fPnz9/PjIykpn7c+DAATabHRER4eTkdPv27dDQ0OXLl9fLLl4UY/bs2eJ1Rjkczs2bN319fbOzszdt2hQSEiIQCMQ9L/UYGhryeLwTJ06sW7fu6tWr48aNCwsLc3d3X7JkiZKS0sqVK+Pj411cXORsTrMzbtq0ycHBITc398aNGwoKCtOmTYuKimJWfmlZzY6QiJYvX+7g4JCTk7N79+7o6Gjm4JIlS5hOsalTp0rd8FjG0xEcHBweHm5pablr1y4vLy93d/eMjAwfHx9x3o0bN2ZnZycmJg4YMCAiIsLQ0LAFm9NkePC+ifjWOOmU+dGvu/2239TeikNE0XEVRFQpEPluy6utozGDVZ+ds7j+g0nyGXM7C3alQPTPLY1+/NLVVMx4XXPmWhkRBR8rJCI9rb++h7uNUddUUxAIRSei30ziOxxRQkTm3ZTGfiDljb9wU06PaS/UnFK5I1LtZqXvPf1mTE1KZvWYxS/VR6UOmpcR/+SviWwLNuYoOKQs3PQmvOj4CgWHFN745wKhiIj4RbVLt+dZz0hTH5U6YE7GwfASJhkzFmPzwYIRn2ZyhqfG3K8koh2hhTYz0lRGpmqNftbfM4OJs7F4pBZy96kg/klVQUmt7LbIIDUGIpIsWWfsMwWHlK2HC4fOz1QdmeqyNCsrr2bp9jyt0c/MP3whNYucF7nJRvXzSI+KqyCioJACBYeUfh7p8za8lhwkFRFbruCQ0t31eW1dw2oBWkvDre6a3BdP6vZ8eXl5S5Ys6dmzp5qamp2dXVBQELMnY8Mq2Gx2vS0F27zR0NVwudxhw4YR0ZgxYyZMmODg4KCvr//VV18xZ/X19b///vtnz56Vl5cnJiauWbOGw+Ewp5hRG1999dWVK1dKSkoSEhKYcmTQ0dHp06eP7C5Ia2vr69evl5eX379/38HBQfLU/v3709PTKyoqmAHXzJ4PRJSQkDBx4kQiWrNmjUgkYlaZbJHEjJ49e167dq2iouLevXuNtZHH4+3evTs1NbW8vPzhw4dSR5q3u/Hjxzs6Ojo6OjJdKidPnmR+Xbt2rThNQUGBmZnZjBkziIj5YpifjxGj7wWMW5HL9evXGx68d++e+HGvXr1u3Lgh/lVyqyAiGjNmjNQJIAYGBrt37254nJlwJMnKykrqtjJ79+5tbPvkhoVIHnmrjGJr1qxpbOXaermKiv76uL9s2TJmx+KGKRt7rKio2FiEUsOTPGhlZdVwYIiVlVW/fv0ePXok4890Y0+Hv7+/v7+/+NewsLB6CaytrSWf/YYhyd+ceg2RfNk0Fl7DQqALOH2t9O7TN30Qr/L/9gV79KA3k7/qRFQpqCOiXubKRBSXUJVfXEtEq7x0VDgsItLVVFw6W/uf/85NeCZMe1Xdw0jK5zCfqZq7w4q+PV7oPEglNKp0kC2Hp63IfAMnIhUOa/Z4jZ/OFx++WLJ4hpZIREciSonIe4omS9pIkUepQueBqj1NlLNya/b/t9hvW56tOdvpHyqTl2WlvqzuYaRsa87+Ysdf050WfqQVEl5y5nrZnpUGXDYr9FIJEc2dpMFhs4TVIqfPXialCYfZcz+ZpHngv8XzN+ZU14g+/ejNOkRrf8wfPUjVbYy6uopC/JOqL3fxu+kprvbWFQhEj58Jnr2sbiweyV4hyULkaYvUHiVGYzFItW5f/pjBKhw2Kzq+op9Huo6Gop0FO/5J1aLNueOHqBrxZP1rbnajPpmk+eOZovTXNUP6cIf25XbjKY0aqHL0UmloVMm2L3hcNuvstTIimuOiqYjbLtDmAgICRo8e7ebmxmazR44cmZSUNGzYsLlz5+7fv9/Hx6e6ulpy7fZt27b5+/vzeLzTp0+7uLikpqZqaWmNGDEiOTnZ2traw8Pj/PnzAQEBCQkJx48fb1hFRUXFgwcP0tPThwwZMnToUNkr6wPIo6qqKisry8TEJDQ0NDw8PD4+/saNG8nJyUTE4XBu3bplY2OTkpJy/Pjxjz76aNOmTX379vXw8BBnX7du3U8//aSuru7g4HDo0CFra+uGVfB4POYGLTO1x8TEhJkYnpeXV15eLplSWVk5IiLCysoqLS0tKSlp165dkmft7e2ZXThMTEwWLFiwZ8+epKSkq1evHjlyZPHixebm5rdv3/7tt9+YDT1bJDFTb1BQ0H//+98//vhj4MCBkZGRVlZW9TbWZLPZN2/etLW1jY2NPXLkyIIFC0JCQpSVletti9lm3NzcmIUjieht/0qEhIT4+fn5+/vHxMTMmTPn3r17fD6f6Y2Crg19K9D1nT9//tatW48fP7a1tXV1dW3vcOANVoNv5w2PyKNL7i6UU1CbUyBrOZLqGtGizbkpmdX9rTmrvXWJKLfgzdQzU4O//rCbGb7pT8ktqJXat2Kop+gxQePQxZL5X+cIhKJlHjpHL5VIJpj/oeZP54vjn1QlpQlfF9Smvapmsch7qpbUqH75yYTLZtXVUaWgLptfE36r/MrtijoRpb6sVlCgG/tMTA2VQsJLFmx8M1BlaD9uH0v2k+fC8FvlrsPUzsWUE9GCaVpEdPHX8qQ0oQqHFbrRSI3LMjVUWrgpZ8uhQnHfis8Uzf8EvBkjduV2BRFxlFk9jJRszNhr5uuqclmNxSPZDSFZiDxtkdG3UlpeJzUGqfw9tLf48gJ+yN98sKBKIIo/baqppqA15plAKHqUKpTdt9LsRq2apxNzryL9dc34Iaob/6nHHBzYi3P/D8HZ62Wzxmtc+KWciLwmd5TZXvBeEW91d+7cuSb3xWu4PZ+xsXFycrKSktLNmzcNDQ2nT58+ZcqUEydO7Ny5U/ylSHI3PRcXl3pbCgK8I3d393379tnb2zOT5YVC4ddffx0UFOTq6mpjY1NTUzNy5MicnJxz586Fh4fPnj17+fLlTJcEEW3dunX9+vWDBw++c+eOlZWVtra25E1Kxrfffjtnzhzxr7/99hvzYO7cuUePHpVM6ezsbGVlVVdX5+TklJmZ6ePjwwzAZ4wcObKqqkpBQUFFRaV79+5TpkwZP3781atXt27dOnr0aHNz88uXL4uXImqRxMzZbdu2rV27ls1mp6WlGRkZzZw5s95qI5MnT7a1ta2srPT09CwvL8/MzNy/f//q1avbq2/F0NCw4VB0Ob1+/fr48eNeXl4hISEcDic4OHju3LktGx50TOhbga7vxx9/vHLlSv/+/Q8dOqSggBuy0Al87qa9+0t95rHFtBfpr2skz+YX17r/36uY+5VO/1A5v627hqoCEenrvJnLk5lbY9vjzR7hGTlvhk4Y6Da64oa/p/ahiyURseXG+kozx6nX61sZ0ofbtyc74ZnwcETJK34tEY39QNW8m5T/HXV1FPhz/tFLpa/ya8RrSecU1Gbm1BCRnpaiqaESEQ3sxZHMtWCa1vLgvNBLJdU1orLKOse+3L492USUkVNDRJUCkcW0v1YHT3tVXSV805U2dshfvQmjB6t6TdYMjSz1DswhIi6btWmx3rLZOlLjkaxdshB52tLYNWwshuWeOlITD+rNJSI9bUUisjRW5mkrEpG6ioJAWFtaIWtCjjyBNdYoqRa7aX8alBPyvxJDXcX84tpBthzm+gO0MfFWd5L74onPMvvicblv1o2S3J4vLS0tIyOjtraWiHg8HvNdSDzNJzMzU9y3gt30oFXFxcX179/fwsLCwcFhzpw5U6ZM2bRpU1hYmJmZGRHx+XxmeRRm+gwRmZqaivtWHj16RETMnhhEpK6u3rBvJTg4+PTp00Q0ZswYPz+/hQsXMtNM7t69Wy8lswFCfn4+s5Pp/fv3xacUFBQ2bNjwySefGBkZiT8SN9aD0LKJmYYLhcLk5GQjIyPmskhijqioqKSlpYkP9ujRg8vlSm6L2Wb27t3r6+vLPE5LS2PWc5Tfzp07vby8XF1ds7KywsLC0LfynkDfCnR9kZGRrVQyJuNA23vyXPjhv7JfZFcvnKa1d6W+eBnXof1UdDUVC0pqtx4qHNFfRYXDKiip3X2yiIjsLNhSB60w7K04Yz9QvXqnwnemttRFYb2naK74jn/4YmlJeR01vort2ZiyrYcLNVQVjgR2MzVUWvdjfsz9SpGITAyUiKiwpK64rE5LXeFF9t9mysydpLF6L//SbxX84jr6c9AK/Tn6Rk1F4fQWI/EUFUG1SOnPPiJm3hNDgUUh6wy/X2WQ+EJ48dfyDT/lB/5cYNZNWWo8kiQLkactMkiNobG+FWUlIiJmnNZbLcQrT2CNNYqIFBVYRCSxixp5TtRYuSvv2t0KtjKLiLymYNAKtA/xVnfiffHOnDnD7JpHRAKBQHJfvIcPH86cOVNyez5jY2Mi4vP5ubm5BgYGT548YVIyOznWq4KkbSkI8I68vb1Pnz794sWLFy9exMbGTpkyhYh0dXWZDg4ej2dgYJCbm9unTx8m/cuXL8V5mc5B2eNw7927xyxHoK2tTURRUVGSJUhijuvo6GhpaRUXFzN9kYyPP/541apVpaWln3zySWZm5saNG52dncWjhpkwxD0jLZiYiOzs7IhIWVnZxsaGiJjLIok5Ul5ePmPGDKZwIuJwOOIdITqXR48eXb16dezYsbt378b62e8P9K0AAHQmIxdlFpXWqXJZmTnVH/4rm4hszNjf/UtfhcPatUJ/3obX1+5W9Jz+wtac/ShVWFBSy2Wzfvg/A9ll7g8wTM4QOvaTvpfQ3Emaq/fmZ/NriEhbQ2G6s7rUZDW1bz4UCoSie0mCuIQ3d5mcB6pYGis/z6qeserVxKGq3/9902g9LcXpo9RPXC799WGlhqrCrPFvCp88XM3GjJ2cIdwcUjDeQbW0vO5BskBRkSYPV2tYdeyjys+/yRs1UKUbTzEhVUBEBrqKjcUjj2bklRqD/DW2XmCSzI2UiCg0qqSsss6hL9dzooYKh+U9RTP4eNGl2HK2MstzokaLxwzwVuTZF6/h9nxaWlpWVlapqalOTk4jRoy4cOECEbm7u3fv3l1qLQ23FGyb1kEXFhIS8v333ycnJ+fm5vbv35+IXr169fjx44cPH6amplpZWf3yyy+3bt2aNm0aEZ06dSo7O7uVIomJiXn+/LmlpeXZs2cjIyMl15QV91FyOJxBgwY5OjpKZmQWZ50zZ466unpcXJy48/HdExPRypUrra2tra2tjYyMysrKGi5fePHixeTkZBsbmzVr1kRHR2tqag4YMKC2tvbixYvveEHay/z5821sbGRsDyq5pMvTp08lN8qATgrzIwAAOhNmXY+KKlFUXAXzE/uokjnlOVHj+g8mU0aoVdfQrYdVbGWaNV7j9kGzkQNUZBZJZt2Uxg1RbbieK0NfR3HqSDVxFVy29GER7mM1lrhpKSnSqj382wlVLkPfZGErsy4Gd3f6h8rtJ1XHIkv9GwzlWPjRm7ESsydoiGPgsFm//GTi666dza8JCikICS8RVItmjpP+zd9QT4mnrXjicun6fQVX71SMG6J6crNRY/HIoxl5pcYgf42tF5gkf08dhz7cnILa3WFFl//cb3uxmzZzZ3HqCDXs1gztTp598Rpuz8flcm/durVo0SKBQHDs2DE9Pb3AwMDDhw83VovULQUB3hYzAIoZlRAQEBAfH29gYDBq1Ki6uroLFy5MnDixoqKiqqpqxIgRP/30E4fD8fT0zM/PX79+/bx581ovKqFQ6Orq+ssvvwwZMmTOnDk7d+4Unzp16tT3339fU1PzzTffODg41BvZvXPnzvj4eENDQz8/vwkTJrRgYiJat26dsbFx3759Hzx44Orq2nADaYFAMHLkyD179nTv3j0gIMDHx4fD4Zw8ebJFr02bysjIuHLlSllZWWMJDA0NHf7ELNMDnR2rLZeBZAaGdcmFJwHgbTVv5dqG2vFPCtOEungp6/kDdCID5mQ8ShX8d0f3KSPerrOmSQoOKdQ5/+/jE0vHxOPx8vPzz50799FHH7V3LPAWOt0bSp6PKGZmZsnJyRwOx8/Pr96yrND25Hl1tdQnz3bUid5E7yeMWwEAAHhPnb9R9uUu/uNnAtsebNdhLdyxAgDQVe3atSshIYHD4RQVFZ05c6a9wwGADgF9KwDQUZSVlYkkXLlyRUFBYd++fUVFRWVlZffv39fQwGIQAC1p39nib48X9rfmnAjqhl3UAADkZG1tXVFRERERMWbMGPF2PwDwnsOcIABoHw1HZjo7OyspKYWFhT19+nT9+vX5+fkGBgaRkZELFixIS0tzcnLauXNnSUlJvVyYEwTQYWFOEABQJ3xDdYHJI+8bzAmCjgD7BAFARxETE0NEQqEwLy/vypUrRDRjxgwiGjhwYGpq6saNG8V78gEAAAAAAHQcGAEMAB3XxYsXme0Db9y4kZKSYmJi0t4RAQAAAAAA1Ie+FQDouKqqqiZNmmRmZrZ+/XoLCwtmGAsAAAAAAECHgjlBANBxubq6Ojk5Xb9+vbS0lIj4fH57RwQAAAAAAFBfO/StdIFlhACgbeTk5Dg7O/v6+tbU1ISGhoaFhbV3RAAAAAAAAPW1wz5BAAAtCPsEAXRY2CcIAKgTvqHwnaXTwT5B0BG06bgVvBoAQKwL/IcDAAAAAAAgrGULAAAAAAAAAPAu0LcCAAAAAAAAANB86FsBAAAAAAAAAGi+Nl3LFgBArKXWW2n3tWwBQLbO+EkD726AFteJ/hTgL0Cng7VsoSPAuBUAAAAAAAAAgOZr032CAAC6Etw9AOiq8O4GAOhi8IcdWhv6VgAAAAAAAN7Al3AAaAastwIAAAAAAAAA0HxYbwUAAAAAAAAAoPnQtwIAAAAAAAAA0HzoWwEAAAAAAAAAaD70rQAAAAAAAAAANN877RPEYrFaKg45YeVdAAAAAAAAAOhQsAczALSPFuycRa8rAAAAAAC0oxboW2mbbzVtP0YGAAAAAAAAAKBJWG8FAAAAAAAAAKD50LcCAAAAAAAAANB8nWy9FcwMAgAAAAAAAIAOBeNWAAAAAAAAAACar1XGrURGRk6aNOn3338fMGCAnFn4fP6ePXtmzpxpZ2cnIxl2AwHoMjAMDQAAAAAAuoaOMm6Fz+cHBgYmJia2dyAA0G4cHR1FIpG3t3e9B3V1dUVFRdevX3d0dGzvGAEAAAAAAOprxb6VvXv36unpOTk5VVVVCYXChQsX6urqduvWLSgoiIh+//333r17czgcMzOzmJiYcePGEZG7u7uVlVXrhQQAndH27duXLl1qamoaGRmpr6/f3uHUxwIAObT3OxUAAACgFbXiWrYsFiswMNDPzy86OjotLe3AgQMXLlz4448/vvzySycnp8jIyOLi4tu3b9fU1Ojq6p44cWLkyJH79u378MMPWy8kAOiMEhMTDx8+nJaWduPGDU9Pz++++669IwKA9wK6hABaHCb4A0BX1Yp9K59++imPx/Pz8yssLExISNDW1p46daq9vf2XX36ZkJDg5uZ28eLFgQMHmpiYhISEdO/enYiYgS2tFxIAdF5Pnz4lInNz8/YORLq6eOv2DgGgg1JwSGnvEAAA3kKn6FdtwV6qjtzeFu+M68iNlQd6JzuyVuxbUVZWZl67IpGob9++//nPfy5dupSUlEREffr0UVRUDAsL43K5Dg4Op06dWrVqFRE9e/aMz+fzeLzWiwoAOilbW1siysjIaO9AAOD9gg+yAC2is3+nBQCQrY3Wsv3ss8+8vb09PT23bNkSGBjo5OT08OHDUaNGWVlZaWlpeXl5mZmZjRs3bs2aNZ988knbhAQAnYWdnd3cuXP3799fXFx8/Pjx9g4HAAAAAADgb1pl3IqLi4v4Jo/4wYEDBw4cOCBOM3fu3Llz50rmunz5cmsEAwCd3YoVK8rKyn7//fd58+bl5OS0dzgAAAAAAAB/04pzggAA3kpcXJx4wHDDBwAAAAAAAB1TG80JAgAAAAAAAADoktC3AgAAAAAAAADQfOhbAQAAAAAAAABoPvStAAAAAAAAAAA0Xwv0rbDaxLvH2faysrK8vLyMjY3ZbDaPx3NwcPjhhx+YUzwej8VinT9//t1r6du3L3OJlJWVeTze0KFDg4ODhULhu5f8LpKSkpio+Hx++0bSIuR8vlrwaQUAAAAAAIDOAuNWWtGsWbMOHz4sFAqdnZ179uyZlJR09epV5tTgwYMdHBx0dXVbqi5DQ0N7e3uRSBQXF7d8+fJRo0ZVVFS0VOEgpxZ/Womourq6BUsDAAAAAACAFvdOfSuit9S8XA1L6BSqqqpiY2OJ6Nq1a9HR0fHx8Xl5eZs3b2bO3r17Nz4+vqCg4MqVK/VG6PTt25dJw+fz/fz8rKys1NTU+vfvf/DgQRnVubm53bt3Ly8v7/vvvyeiuLi4rVu3yi6HGWQRFBQ0evRoTU3NQYMGMQHLrprJtXnz5nHjxmlqavbt21ecKyUlZfTo0WpqagMHDoyPj5cMr3kFFhYWrlixolevXioqKt26dVu9ejURzZ07l8Vi+fr6MmkiIiJYLJaRkVFtba08NZ47d47FYunp6WVnZ9fW1g4fPpzFYi1btkz2BalnwYIF5ubmqqqqHA6nd+/ee/furfe0Ntk0eS7I8OHD2Wx2TEyM1BgAAAAAAACgg8C4ldbC5XKNjY2JaM6cOV999dWFCxcqKipsbGzqJTMzM/viiy+++OILPz8/FRUVImJyCYXCkSNH7tmzx9DQcOXKlYWFhT4+Pj///LPsShUUFBYvXjxhwgQiOnXqlDzlbNu2zdnZeeLEiffv33dxccnNzZWn6q+//trOzs7Ozu7JkydeXl5EVF1d7erqGhMTY2BgYGtru3TpUnHi5hUoFAqdnJx27NjB5/Nnz549YsSIxMREIvr888+JKDQ0tKqqiojOnDnDXGRFRUV5apw+fbqvr29BQcH8+fO3bNkSGxs7ePDgb775RsYFaXidHz165OzsvHr1am9v75SUFF9fX/GIpIakNq3JCxIQEMDlct3c3NTV1RsrGQAAAAAAADoCpfYOoCs7derUZ5999ujRo8ePHxMRm81et27dmjVrJNPY2Nh8++23IpFo4cKFlZWVvXv3Dg0NJaKLFy8mJSWpqKgcO3ZMTU3N1NR0wYIF//73vz/99NMm67WxsYmOjn79+rU85axevXrVqlUikcjS0jItLS0sLMzY2LjJqletWhUYGHj37t0PPvggNTW1qKjozp2cAwHJAAAgAElEQVQ7qampCgoKv/zyi6mpaUhIyPz585nE8rSlYYExMTEJCQksFuvXX3+1tbWlP2fHODo6Dhw48P79+2fPnp01a9aFCxeIyNvbW/IKyK5x+/btsbGxUVFR0dHRmpqaJ06cYLPZ4rwNL4h4jIzYzZs3uVxuXV1dZWVldnZ2eHj45cuXx44dK/XpaNi069evN3lBfHx89u/f3+Rz3alJjkRj1lTqXGPTOpp+HulPngs/d9Pe/aU+c8Ri2ov01zWr5un8+3MeES3bmXfjfmX662phtcjSWHnRdC1fd21x9l8fVm47Uhj7uKq4rI6nreg8UGXNfF07C3ZjFRHRN0t5K+boMAddlmZFx1cQ0b/m6Gxbyhuz+GXM/Uq3Meph/zYSZxy56OWvDys9JmiEbuwmT4v0JzzPL649+43RR6NasocxKU1oNyudiHKjLHnaik2mbw0dIQYAAAAAaEEYt9KKHB0dHz58+Pz58+PHj0+ZMkUoFAYEBKSkpDRM6e/vf+DAAQsLi8uXL/N4PCLKyMggosrKyh49eujr6y9YsICI0tLSmMEasv3xxx9E1K1bN3nKsbCwICIWi9WjRw8mvTxV29vbE5GWlhbza1lZWWZmJhHp6emZmpoS0cCBA8WJm1dgeno6EfF4PKZjhYiUlZWZB0uWLCGiAwcOxMTE5OfnDxo0SDyRSp4aORzOihUriEgkEnl4ePTs2VMyb8MLUu/y1tXVbdiwwcTERFlZWV1dPTw8nIhycnIaezoaNk2eCzJu3LjGCgRonn3niisFItdhan17chKeCZduz9t3rpg5dfJyqfM/X/73ZrmSIg2z5wqEouPRpR94ZcQ+kvUHZ09YUW0dEdGT50KmY0XM50MtIvrfrfLC0jrmyLOX1b8+rCQin6macgY8uDfHoQ9XVxNdDwDQKlxcXFgsVkBAQHsHAgAAXQH6VlrRwYMHy8rKLCwsZs+eLV6SQ7wYh9i6deu+++47IyOjy5cvMxOCiIjpoVBTU4uMjLz8p//9739KSrKGGtXV1f3444+XL18mIjc3N3nKefjwIREJBIKnT58y6eWpmpmAI7l/k4mJCREVFhYWFxcT0YsXL8SnmlegmZkZEfH5/OTkZOaIeFVXT09PHR2da9eu7dixgxoMWmmyxvz8/JUrVxIRl8vdv39/vUVVGl6QeoWfPXt269atJSUlR48evXnzprOzM8kccNGwafJcEGaCGEALivjWOOmU+dGvu/2239TeikNE0XEVRFQpEPluy6utozGDVZ+ds7j+g0nyGXM7C3alQPTPLY12GupqKma8rjlzrYyIgo8VEpGe1l+dIG5j1DXVFARC0YnoUubI4YgSIjLvpjT2A9V6Re0ILbSZkaYyMlVr9LP+nhlMSiK6+1QQ/6SqoKT2yu0KBYcUyZ9+HulExC+qXbo9z3pGmvqo1AFzMg6GlzQWbUpm9ZjFL9VHpQ6alxH/5G8dRo0Voj/huYJDSlBIwZjFL7VGPxs876+eJhn1Mrk2HywY75ulNfpZP490ca5mxCC7wMLSui938W3d01VHphpNev7V93wimrfhtYJDit+2PCZNRGy5gkNKd9fntXWNXRuAjqgN1nHHUvHw7hISEkQi0Z49e4jI2Nj40KFDWVlZQqGQz+fHx8cvXry4vQNseX369Dly5EhmZqZAIHj16lVERMSoUaOo8eZfv35dJBIxKxWI3bp1SyQSHTt2rH3aIB/JJ5eRlpYmEom2bNlCREVFRVLXBj1x4gSTUSQSMbdyGVFRUczB7du3N1Y+dAHoW2lFPj4+BgYGAwYMmDBhwgcffEBERkZG/fr1k0xz7ty5jRs3EpG9vf3u3buXLVvGrEE7efJkGxub8vLyoKCguLi46Ojob775Zs+ePY31rZw+fXrQoEEGBgbMHzJHR8dVq1bJU86OHTs8PDyGDRuWk5Ojrq4+c+bMt62a4ezsbGlpWVNT8/HHH2/bto1ZHZbRvAJdXV179+4tEomGDRs2f/58Dw+PWbNmMadUVFS8vb1FItGlS5fYbLanp2e9vLJr9Pb2fvnypY+Pz48//lhTU+Ph4VFYWCjO2/CC1Cu8pqaGeSAQCO7duxcXFyejFVI174IAyOP0tdKh8zOZn1f5f1vgefSgNx12dSKqFNQRUS9zZSKKS6jKL64lolVeOiocFhHpaiouna1NRAnPhGmvpH/98JmqyVZmfXu8MLewNjSqdJAtZ3BvjvisCoc1e7wGER2+WEJEIhEdiSglIu8pmhLdjERE8U+qvtzFL6usW+2t6+uubW6k9Oxl/RrNuil9MUv7i1nafjO1mQiN9ZWE1SKnz17uOVVkqKv45VzdwtLa+Rtzfj5f3DDU6hrR5GVZMfcrDXSUbM3ZX+zIE59qspDtRwtHDVKd6Kh6/w/BpC+ycgtr5al34/4COwu2nQX7yXOhd+Drd4xBaoHCatGozzJ3hBbyi2pnT9AY0V8l8bmQiJa4aRNRaFRJlVBERGevlRHRHBdNRfy3h46t4TruTS76LnXt+by8vCVLlvTs2VNNTc3Ozi4oKEggEEitgs1mR0VFEVFQUJDkTgIAzXPy5Ml58+Yxr95nz57Z2to2Nlu883Jxcbl3794nn3yio6Nz586dzMxMJycnd3d3arz5Bw4cIKKpU6fq6LyZRNyzZ8/hw4eLT3VSd+7ciY+Pj4+PZ0adv379mvlVcoKCn58fc4e1T58+zGqY0OXh01Yr2rRpk4ODQ25u7o0bNxQUFKZNmxYVFaWq+rd7tllZWcyDqKio77777rvvvjty5AgRcTicmzdv+vr6Zmdnb9q0KSQkRCAQiDsXGsrJyWEGXDg6OgYHB9+4cUNNTU2ecjZu3JidnZ2YmDhgwICIiAhDQ8O3rZrBZrMjIiKcnJxu374dGhq6fPly8anmFcjk8vf319XVPXbs2LVr13r16iU+u2TJEmYkyNSpUxvueSyjxuDg4PDwcEtLy127dnl5ebm7u2dkZPj4+Mi4IPUKd3d3X7JkiZKS0sqVK+Pj411cXGQ3pLGmve0FAZBHTkFt/JMq5kdYLWU4VXWNaOGmnJTM6v7WnNXeukSUW/Cmu9DU4K/ePTPDN1PwcgtqGxZCRIZ6ih4TNOISquZ/nSMQipZ56NRLMP9DTSKKf1KVlCa88Xtl2qtqFou8p2rVS1ZaXkdEHGVWDyOlqU5qJ4KMAhfp1UtjY8YOXq6/01+/vLKuUiDq3YN99OtuF38tT0oTqnBYoRuNPnfTWv+pHhFtOVRIDcTcr0x9Wa2gQDf2mYRu7Ba8XF98qslC/s9Ld/1C3ZObjXoYKZdW1IVdKZWn3lVzdb77lz6z8E3qy+qi0rp3iUFqgRGx5QnPhCwW3frZ5MBaw1NbjE5tMSIix77cgb04RaV1Z6+X1dbRhV/KichrsrzzsADal3gddzab3eSi7w3XnhcIBCNGjPjhhx8UFRU9PDxyc3MDAgLqDW4VVzFt2jRzc3MiGjJkyBdffDF37ty2bCl0MVwud9iwYUQ0ZsyYCRMmODg46Ovrf/XVV8xZfX3977///tmzZ+Xl5YmJiWvWrOFw3tyN4PP5IpHoq6++unLlSklJSUJCAlNOB8ThcEJCQjgczuPHj21sbEaMGDFkyBBjY+PDhw/LaP7p06dLSko4HM7s2bOZcpiNHdLT02VsBNHxjR8/3tHR0dHRkVnE4OTJk8yva9euZRIUFBSYmZnNmDGDiJivRfn5+e0YMLQN3CdvRWvWrKm3cq0Yn88XP264VCrDwMBg9+7dTdaSkJAgO4HscqytrW/cuCF/FsnIraysJOfC9OrVS7Ioya2Cmlegnp7ezp07d+7c2TCXlZVVv379Hj161HBCkOwa/f39/f39xb+GhYXVSyD1gkgGqaiouHfvXvEkr8aSkcymyXlBujZWvQEM0o7IAyvgSmq4lq3k2fziWvf/exVzv9LpHyrnt3XXUFUgIn2dN3N5MnNrbHu8Wbw2I+fN4BED3UaXO/H31D50sSQittxYX2nmOPWjl/42JWdIH27fnuyEZ8LDESWv+LVENPYDVfNu9f/pjB6s6jVZMzSy1Dswh4i4bNamxXrLPev31BDR8uC8A/8rseiuHL3HmKetmJFTQ0SVApHFtL9mIKa9qq4Sirjs/2fvvuOaPNc+gF9JIGGDGBmKiAqICOppreAAceBAarUVFRyAtvpWoS3WcTzgQMS6bR21tsfiqEjRin6KirQKjlqhahFBPYDIlhH2TBh5/3hsmjJCRFbw9/3wB+R57plBcue6r+cfD6TMvDoi6q3N6a+vRERvDfk7vkZGJS/nsK8yEbFYZGKolPaiNjO3jtlcI7tdazMeEWlrvJy6iuqGtvVBRoXMPcvX4UjuMmWllyd/PFfno8C8oJ/L9HU5haX1b1vwrAY3k5MYoBuS5HEPCwtrNel7s8n4k5KSlJSUbt26pa+vP2fOHGdn55CQkH379hkaGjZqgoimT5+enp7u6Oi4bdu2zh8s9CQ1NTXZ2dlGRkanT58ODw+PiYm5ceMGs6udx+Pdvn3b3Nw8OTn5zJkzs2fP3rZtm5WVlaurq6T4pk2bvv32Ww0NDRsbmxMnTpiZmXXdUFpka2vLJHPcuHFjTk4Oc2NpaWlsbCwRtTT86urqkJCQ5cuXu7u7HzlyhMViMeuYx48fV4j3b3Pnzh01ahTzu+RlRB5BQUHe3t4+Pj7R0dELFy68f/++QCCYNm1ax3QTugusrYDiuXDhwu3btx89emRhYeHk5NTV3QFQDImpolmf5zzPqf3wPe3D6/pIPoqPsVbV1eIUldXvPFE8foSqKo9VVFZ/8McSIrIcyDUxVG6pwuGmvMnvqF37o8prno6kNmkezlprvhKcvFReVtlALWSxZbMoaJP+1+v1Hj8XXfqtcsu3hf7fFTVdW9n8beFXP5YY8pUiD/br10eJ/oqyUVdln9thKNnwIqwVKzVZCzLSUyKi4rKG0ooGbQ3285y/9xy1Wkl8snDeFA2hSPwkTURERvpKTOuy22Vul14qfJ0+NFuhsb4SEQlK6pMyRObGXCKqrRMz94LbNM11Bwqu36viKrOIyN0ZQSugMCR53KWTvkuOMknfVVRUmD+lc8+npaVlZGTU19cTEZ/PZ2JOJdt8MjMzJR+KkCoeOoiLi8vRo0eHDx/ObP8XiURbt24NDAx0cnIyNzevq6uzs7PLy8sLCwsLDw9fsGDB6tWrX7x4wZTduXPn5s2bR40a9ccff5iamuro6JSUlHTpaJrBLKwQUbPX5Whp+ER07Nix5cuX29jYWFhYGBgYmJiYiMXioKCgzux8m+nr6zeNYZdHbm7umTNn3N3dmWCf/fv3IzjuTYC1FVA833zzza+//jpixIgTJ06w2djXBiAXu+WZJeUNaiqszLzaWZ/nEJG5Mferz/uo8lgH1vRZsiX3+r2qwXOeWwzgxqeIisrqVbisI//Wk13nMT/9pAyRrbVKs0cXz9DacLgwR1BHRDqa7DkOzVxK+U589apdBRPeUjXgcxJShNRcpExYdEXAsSIiGm7KPRRaQkQGfKXPFuiYG3OTMkTbg4ocbdTKKxvikoQcDs0cp96ouMNbqoP6Kadm136w/sW0MWpfn/37DevMceqyK9kbXJyaXZuUIcorqtdQZc+boqmjwZaz3fbqQ7OcxqoPNeE+SRON+zBrlp16tVAsrBX/tNOQiFR5LA9nrf1nSq7cqeQqs9ymacqoB6BbkeRxlyR9/+mnn5iEBUQkFAqlE5M9fPhw3rx50rnnmQsCCASC/Px8PT29xMRE5kwm3X6jJuivZPMNDUj1DO3g7t27I0aMGDhwoI2NzcKFC52dnbdt2xYaGiq5OANzTUlJvHn//v0layvx8fFExFwOgog0NDS64dqK5JqYZmZmjx8/bnS0peEnJyfHxsYmJCRYWVm5u7szq5zXrl1jttJ0f4cPH5bsMEhLS2M2Espp37597u7uTk5O2dnZoaGhWFt5E2Bt5c2luHtPIiIiOqJaxZ0QAHkwmU2qasRX7768XrKg5GUuFbdpmv31lXafKv79Uc3thzV8HfZ8R01fT91W95IYGygZN9nmI9GnF+ddO/XzURVME4226jD0eyvxdTghv5QXlzVoqrGmjFbb4cVvdE52wcudTVfvVjGdHzaIu35Jr5vfGm07VnTl98rAoCJNNbblQK7HtGZiNLjKrEv7+674Ij82saawtN7Hrddn+16mkuVxWbIr2bqi96XblY+fi0aa8776vI++LoeI5Gy3vfrQLKbU9qCi8NuVwZHl2upsJsEN4+O5Ol+GlIjF9O54dVzEGhQRk/Q9KSkpMDBw6tSpZWVlcXFxHA5n5syZknP27t2bmpqalJQkyT2vra1tamqakpJib28/fvz4ixcvEpGLi0vfvn2bbYX5mHT69OmKigpbW9umqfEB5Ofh4XHu3Lnnz58/f/78zp07zs7ORKSrq5uZmUlEfD5fT08vPz9/2LBhzPlZWVmSskzIVTffI/P777/n5uYaGBgEBATExsYyC0Pa2tpDhgyJjY1tafhM2aCgoL179y5ZskRLS4sUPIut/OLj469duzZ58uSDBw/iwmRvCKytAAD0BI/ONP4u5fnFgdJ/1v4ua/+23UhVu5FyXfm7aUOMK1/1a3rjuR2tbE42NVL+9XAzBYmoIHKQ5HcvF52mJ+j14hxY04eoT9NDjQwZwI3+5u8vrj+Z/3dtsisx668sXbDVItJ9NjVSboj5e87b1gcZFfbW5uz9rM/ez5opZWqkbD2YF58ixIYgUFBM0veAgIArV65s27ZNU1PT0tJSOvE8EQUEBISHhzO55w8cOMDE7d++fXvTpk2RkZHBwcHGxsbe3t7r1q1rqZXVq1c/ePAgLi7u4MGDZWVlWFuB1xEUFPT1118nJSXl5+ePGDGCiF68ePHo0aOHDx+mpKSYmprevHnz9u3b7733HhGdPXtWkrJEUQiFwqVLl4aFhVlbWycnJ8fFxSkrKw8bNuz48eOxsbEtDZ8pe+rUqR07djCrnCUlJWFhYV05kk60dOlSc3NzGRcVlc7n8uTJk0avcqBwsLYCAADQc1y4UfHbw5pHz4QWJlynsbI2FgF0H01DR1vN6N9s7nl9ff2jR4/K2YSpqamMzzwArWK2lTEhCX5+flOmTBkyZMjQoUOLioouXry4cePGqqoqIho/fvzWrVunTp3q5uaWkZFx8ODBXbt2dXHX2+TKlSujRo1av379xIkTR48eXVRUdPPmzXPnzpHM4RNRQUHBzz///P777xNRcHAwc93iN0FGRgaTPaol0vlcpPc8goJidWb4GXMRkG4e8AYAnaNtVwVqqstfUpiBSAcUQA/QZ2pqYWn9+V2Gsyc0kyamO5vxafavsVXWprzjm/WHm/JaL9Dx2DbJ1A2eqm2A9y3dE5/PLywsDAsLmz17dlf3BV6BAj2h5HmLYmxsnJSUxOPxvL29Dx061Am9aqQdZ7K93pJ1hHZ/wHTnwcpDIZ5BbyzkAQWA7qKiokIs5ddff2Wz2UePHi0pKamoqHjw4IGmJrJyQicpiBzUEGOmcAsrRHTlq361v5s9OGXcTRZWAAB6ngMHDiQkJPB4vJKSkp9++qmruwMA3QJCjwCgu3B2dlZSUgoNDX3y5MnmzZsLCwsdHR2XL1++bNmytLQ0e3t7Rf+qAQAAXh9yz0OXMzMzq6qqunXrlp+fn+RyPwDwhsPaCgB0F9HR0UQkEokKCgp+/fVXIvrggw+I6K233kpJSQkICGAS6QMAAAB0oRkzZnR1FwCg28GeIADovi5duhQREbFq1aobN24kJycbGTW+YgsAAAAAAECXw9oKAHRfNTU1M2bMMDY23rx588CBA5kwFgAAAAAAgG4Fe4IAoPtycnKyt7ePiooqLy8n7LEHAAAAAIBuCWsrANB95eXlOTg4eHl51dXVnT59OjQ0tKt7BAAAAAAA0BirMy+RjWt8AEC768wXsWYxr2wNMWZd2w2Abottk0zd4KnaBsyzWxF7DtANKdATSiE+s7TjTHbn8bb7A6Y7D1YeCvEMemMh3woAAAAAAAAAQNt16p4gLLMBgISif28AAAAAAADAQNwKAAAAAAAAAEDbdWq+FQAAifaKW+nyFzEE4ADIo8ufqm2gQOkhALo/BXpCKcR/duRbaZvuPFh5KMQz6I2F6wQBAAAAtEjR34gDAABAJ8DaCgDAa8EXCAAAAAAAbzisrQAAAAA0AyunAAAAICfkWwGArtFj8q0AAAAAAMAbDmsrAAAAAAAAAABth2swAwAAAAAAAAC0HdZWAAAAAAAAAADaDmsrAAAAAAAAAABth7UVAAAAAAAAAIC2w9oKAAAAAAAAAEDbKXV1BwDgDdVe12AmXIYZAAAAAAC61GutrbTjRyM54RMUAAAAAAAAAHQr2BMEAAAAAAAAANB27bAnqHNiSTo/RgYAAAAAAAAAoFUKlm8FKywAAAAAAAAA0K1gTxAAAAAAAAAAQNspWNwKctkC9BgIQwMAAADoDhT9XRk+JEJ30CFxKxERESwWKy4uTv4iAoFgy5Ytjx8/7oj+AIBCsLW1FYvFHh4ejX5paGgoKSmJioqytbXt6j4CAABAD8fqxrp6bgCgRd0lbkUgEPj7+1tZWVlaWnZ1XwCge9mzZ09CQsKmTZsiIiLMzMwKCgq6ukd/w7scAHngG0UAAADo2Tow38rhw4d79+5tb29fU1MjEok+/PBDXV1dAwODwMBAIvrzzz+HDh3K4/GMjY2jo6OnTJlCRC4uLqamph3XJQBQRI8fPz558uTSpUu1tbXd3Ny6ujsA8Ebo6i+nAXqgrn5aAwB0oA6MW2GxWP7+/t7e3pGRkWlpad9///3Fixf/97//rV271t7ePiIiorS0NDY2tq6uTldXNyQkxM7O7ujRo7Nmzeq4LgGA4nry5AkRDRgwoKs70oyGGLOu7gJAN8W2Se7qLgAAAAB0uA5cW/noo4/4fL63t3dxcXFCQoKOjs677747fPjwtWvXJiQkzJ0799KlS2+99ZaRkVFQUFDfvn2JiAls6bguAYDisrCwIKKMjIyu7ggAvEGwmwmgXSBoBQB6vA7cE6SsrMy8jIrFYisrq5KSkitXrpw/f56Ihg0bxuFwQkNDnz9/LhKJzp49q6KiQkTPnj0TCAQd1yUAUESWlpaLFy8+duxYaWnpmTNnuro7AAAAAAAA/9BJuWxXrFgRFxfn5ubG5XL9/f3t7e1PnTq1bt26wsLCQYMGubu7GxsbT5kyxdfXNyoqKiIionN6BQAKYc2aNRUVFX/++eeSJUvy8vK6ujsAAAAAAAD/wHqdYFdJWEr79adbtAUAnaAdw4O78JWBGQXyrQC0hMm3ooj/vvHGA6AdKdYTqjvvYOqgOezOQ5aHojy0oGfrwD1BAAAAAAAAAAA9HtZWAAAAAAAAAADaDmsrAAAAAAAAAABth7UVAAAAAAAAAIC2w9oKAAAAAAAAAEDbYW2lddnZ2e7u7v369eNyuXw+38bG5siRI13dqRZNnz6dxWL5+fk1e9TKyor1F3V1dSsrq/3797ctsXZiYuKYMWPU1dVZLNbs2bNfr9c9Fp/PZ7FYFy5caMczAQAAAAAAoFtph7UVVqd4/X622fz580+ePCkSiRwcHAYPHvz06dNr1651YX9en4mJycyZM83MzBITE1evXr1r165XKl5bW0tEBw8evHv3rqmp6YEDB5YvX/6qxaGRUaNG2djY6OrqtleFmGcAAAAAaEcJCQlisfjQoUOSW9LS0sRi8Y4dO4iopKRE3JyQkBCmoFgsXrNmjaTs1atXmRv37NnTUv0ACgRxK62oqam5c+cOEV2/fj0yMjImJqagoGD79u3MUYFA4O3tbWpqqq6uPmLEiOPHj0sKJicnT5o0ibk9KCiIWSESCAREpKOjw2Kxdu7caWtrq6qqOm3atOzsbG9vby0tLWNj45MnT7ZaORPjsH379ilTpmhpaVlZWTGdtLKyunr1KhEFBgayWCwrK6tmBzVz5szw8PC4uLhJkyYR0fnz52W3KGlu3LhxXC43OjrawcHh6NGjRBQfH//JJ5/ExsYWFBSsXLly8ODB6urqlpaWgYGBQqGwpeLyT8KyZcsGDBigpqbG4/GGDh16+PDhVieBiIqLi9esWTNkyBBVVVUDA4MNGzbInk9pLZ0WFhbGYrF69+6dk5NTX18/btw4Fov12WefSXoSGBg4ceJELS2tt99+W9KTRmQM5969ezExMUVFRbKHJs+jQnqeAQAAAKAjDBs27NSpU5mZmUKh8MWLF5cvX54wYQJzqF+/fidOnMjOzhaJRAKBICYm5uOPPyaiqKgosVh89uxZ6Xpu374tFouDg4O7YAzt6o8//oiJiYmJiampqSGi3Nxc5s/k5GTJOd7e3hwOh4iGDRs2derULusrQAd4rbWVZhcmZWhbqaY1dCYVFZV+/foR0cKFC//zn/9cvHixqqrK3NyciEQikZ2d3aFDh/T19detW1dcXOzp6fndd98xh2bMmBEVFWVoaDhy5MjVq1c3rXnjxo06Ojo8Hi8yMtLKyurSpUuWlpaZmZkfffTRixcvZFQusXXrVktLS0tLy8TERHd3dyJavHjxgAEDiGj06NGffvrp4sWLZQwtKysrKyuLiLS0tGQPh+Hn56eiojJ37lwNDY25c+daWFgQkbW19aeffvqvf/1r/PjxR44c4XA4rq6u+fn5fn5+Hh4e0s1JF5dzEogoPj7ewcFhw4YNHh4eycnJXl5ejYKGmk6CSCSyt7ffu3evQCBYsGDB+PHjHz9+LM98yp6EOXPmeHl5FRUVLV26dMeOHXfu3Bk1apR0yM/u3bsdHBymTZv24MGD6dOn5+fnN53zVofT6tBaHUWz89w9teOLQ9cOpL1Yu6azbZK9dxdIbhn43nO2TfKGwwLmz8/2FfxrUYbulGcaE1KGu6UfOitqyVcAACAASURBVFsiXfy3h9Wz1+ToTUvljUvpN/P5wo25j5+LZDTEtknec7pYcuP0T7KZG9ceEBDRpI+z2DbJ8za8kC5otzyLbZO8cGNuew258834NJttk7zxm8KObigxVTR2WabGhBS2TfKctTkd3RwAAHSy6dOn379/f9GiRb169frjjz8yMzPt7e1dXFyYoz/++OOSJUuY77qePXtmYWExefJkIvr++++J6N133+3Vqxdz5uDBg8eNGyc5pNAcHR1tbW1tbW3T09OJ6Mcff2T+3LhxI3NCUVGRsbHxBx98QETM56PCwg7/jwzQaZS6ugMK4OzZsytWrIiPj3/06BERcbncTZs2+fr6Xrp06enTp6qqqsHBwerq6v3791+2bNkXX3zx0UcfMS+jbDb7xo0b/fr1++GHH5ouc/j4+OzcudPX13f79u01NTWxsbFaWlpaWlpCoTA+Pr6qqqqlyiU1rF+/3t/f/969e++8805KSkpJScn69eujoqLS09MdHR23bdvW0ogOHz4siZhgs9mff/45EckYDnOmp6fnsWPHmN/HjBlz9+7dp0+fjh079ssvvwwLC0tKSlJSUrp165a+vv6cOXOcnZ1DQkL27dtnaGjYtLick2BoaHjr1i0VFZWGhobq6uqcnJzw8PBffvmF+efU0iRER0cnJCSwWKzffvuNWQCqra0NDw9vdT5bnYQ9e/bcuXPn6tWrkZGRWlpaISEhXC5XUnbDhg3r168Xi8WDBg1KS0sLDQ318vJqNN5WhyOt6dCioqJaHUWz8ww9w9Gw0gEGyk5j1VOyamMTaz7ZU6CsxFoxR5uIfvylfNGm3PoGMujNGTtc5VGK6Exk+YUbFb8cMho7XKWlCg+Flvi49uKwKTFVFBlTJX3Ic5Z29IPqn29XFpc39NJkE9GzrNrfHlYTkee7Wh08UMVWWydWVmIdCi25m1Az3JS37D2twUbKr1q847oHAIzp06dfvXrV19dXxvslgJbweLygoCAej/fo0aPp06fn5OQQkba29pAhQ4hIRUVl7NixRDRp0iTJxwcTExMiOnfu3KFDh7S0tBYsWMAkcGS+P0tPT1egnANz584dNWoU87vkrb48goKCvL29fXx8oqOjFy5ceP/+fYFAMG3atI7pJkBnw56g1tna2j58+DA1NfXMmTPOzs4ikcjPzy85OTkjI4OIqqurTUxM+vTps2zZMiJKS0urqalh4kF0dXWZmJdm9+YwL0l8Pp+IBg0axOfzuVwuE2tQXl4uo3JJDcOHDycibW1t5s+Kigo5R8TkW3FxcVm3bl18fPz06dOJqNUWp0yZ0lKFTFk+n6+vry893szMTMk5zRaXPQkNDQ1btmwxMjJSVlbW0NAIDw8nory8POkamk4Cs1LO5/OZhRUiUlZWlmc+W50EHo/H7BEVi8Wurq6DBw+WLjtw4EAiYrFYzP9Opipp8gxH9tDkGYWMuwkU3eUv+z09O+CHrQa/H+s/3JRHRJF3q4ioWij22l1Q30CTRqk9CxsYdcQo6acBlgO51ULx/+1o8QGmq8XJyK376XoFEe0PLiai3tocydG5kzS01NlCkTgkspy55eTlMiIaYKA0+R21RlXtPV1s/kGaql2K9sRnI9wymDOJSFBS/8meArMP0jQmpIxcmHE8vExSJDmzdvLKLOb2oPAyJmRGUFJPRL0mP2PbJO88WTxmaaaaXcr0T7KzC+o+2VOgPfHZgFnP5am8z9RUtk3y9uNFjl7Z2hOfWbum34mvISJr1/Srd6uIKDCoiG2TbO2a3mggTMHAoKJJH2dpT3w2akkGU1B2i5Lmxn+UyRuXEv2geuLHWUfDSokoPkX46d6CPxJrCorrV+3KN30/TWNCyrD56YFBRUKRuKXics4AEX24Lc/kvefq9ikq41Ms56cfPlciewYYxeUNaw8ILFzS1exSDGek/udrgez5BOhknZAvDCnJ4DXZ2toaGBgQ0caNG5mFFSIqLS2NjY0lopqamuzsbCI6ffr09u3b33vvPTU1taSkJCKqrq4OCQmhv5ZUWCwW8/3r8ePHFSgOV19f3+Yv0l80tio3N/fMmTO2trbMytT+/fs7rpMAnQ9rK607fvx4RUXFwIEDFyxYIAn3KCoq6t+/PxGpq6tHRET88peff/5ZSUmJOVRUVJSbm0tEzIp1I8rKykTEpOllfpcmo3LJOcxmxUaJfpkbGxoaZIyIybcSGhq6c+fOYcOGydmiqqpqSxUyZQUCAbMRJjExkbndyMhIck6zxWVPwvnz53fu3FlWVvbDDz/cunXLwcGBiBr942k6CcbGxkxnmP9hRFRbWyvPfLY6CYWFhevWrSMiFRWVY8eONUqq8vDhQyISCoVPnjyRVPWqw5E9NHlGIeNuAoVw7nr5mKWZzM+LwnrpQxPffnnnNoipWthAREMGKBPR3YSawtJ6Ilrv3kuVxyIiXS3OJwt0iCjhmSjtRfMfITzf1eIqs748U5xfXH/6avnbFrxRQ3mSo6o81gJHTSI6eamMiMRiOnW5nIg8nLUa5RaPSaxZe0BQUd2wwUPXy0VngKHSs6xaIhLViu1XZB06W6Kvy1m7WLe4vH5pQN53F0qZQ06fZUfdrzbkK400533+ZQE1selooY4mm8dlRcZUWbumX/qt0nIgNzOvbvn2/BeCOhmVSwQcK7IcyLUcyE1MFXn45xLRohlaAwyUiGj0MJVP5+ssmtF8AM6eH4onvK02zVbtwf+EMz7Nzi+ulz0cxsZvClW47LmTNDRU2R9M0rAw4RKR9WDep/N1Rg5RsVuedeSnUg6bXKdq5hfXb/ym0HPrP5a9pIvLMwPMOfEpIoe31P7trusxUys5Q+S9u+DaH3/HHzWdAWYgE1Zk7j1dLCipXzBVc/wI1cepInnmE6DjNM0X1mpysWZznMmf+o3L5cqTnw6gJczCChFJZxKR5uLiEh8fb21tvWHDhgsXLuTl5fn6+jKHmPhiGxsbCwuLCRMmmJiYiMXioKCgzul5uzh8+LDkeiPMl5ry27dvHxE5OTllZ2eHhoZ2TAcBugb2BLXO09Nz5cqV5ubmenp6zOdnQ0NDa2vrkSNHmpubJyUlBQYGTp06taysLC4ujsPhzJw508HBwdTUNCUlxd7efsyYMRcvXnzVRmfOnNlS5bILMvlWTp8+XVFRYWtr6+bm1tEtEpGTk5NkvOPHj2fG6+Li0rdvXzlbb1Zd3cvPD0Kh8P79+3fv3pWnlJOT09ChQ588eTJ27NhZs2ZVV1cLhcIzZ87IMzrZk+Dh4ZGVleXp6TlhwgQPDw9XV9e4uDjJdtm9e/empqYmJSXl5eVpaGjMmzevXYYjf/egZ8grqs8rqpdxQm2dePn2/OTM2hFmvA0eukSUX/TyodVf7++XdGP9l4uV+UX1JobN7EnR781xnap54lLZ0q15QpH4M9deP1z5R6jC0lla314ojUmseZomyi2qT3tRy2KRx7vajeopr2wgIp4yy8RQydyY67tUV02FRUSXfqt8miZS5bFOBxiqq7D66yt9uC1vx4nij2ZrRz+ofpZVy2ZT9DdG/foo/XBFbcmWxjlcfFx1dnjx/Y4Ubj9eVCMUx5zrr6XO1p70TCgSx6eIqmoaWqpcUsP6xb22LO9970nNaI/MlKzakvKG9Ut6Rd+vSs+tcxytFvB/vVua4X+7665f0ksspsFz0tJe1Ib+Wu7loiNjOEwpT2et//rpM7+PsVaJSah5miYaO1xl/+o+YdEVSRkiJQ7r5rf99XU5sx0q312dE/JL+d7P+IZ8pabF5ZkBpuDNb41UuKyGBqoWNuQI6sJvV/4aWyUJLGo6Azqa7Mt3KhOeiVgsuv2dEbMAVFsnDr/dyugAOoGfn9/EiRPnzp3L5XLt7OyYfceLFy8+duyYp6dnbW2t9AbY3bt3+/j48Pn8c+fOTZ8+PSUlRVtbe/z48UlJSWZmZq6urhcuXPDz80tISDhz5kzTJqqqquLi4tLT00ePHj1mzJhX2tQAQFJxx2ZmZo8fP256wt27d0eMGDFw4EAbG5uFCxc6Oztv27YtNDQ0OTk5NjY2ISHBysrK3d2deexdu3btVVcoFFd8fPy1a9cmT5588OBBRJBBD4O4ldZt27bNxsYmPz//xo0bbDb7vffeu3r1KnOpl1u3bnl5eeXk5Gzbti0oKEgoFM6fP5+IlJWVL1++7ODgkJ2dHRcX98UXXzBVNYqSkEFG5bKtXr3axsYmLy/v4MGDkZGR8g+zzS0SkYqKyu3bt5cvXy4UCoODg3v37u3v7y+51k+bubi4rFy5UklJad26dTExMczepVYxA/Hx8dHV1Q0ODr5+/fqQIUPkHJ2M0/bv3x8eHj5o0KADBw64u7u7uLhkZGR4enpKygYEBOTk5Dx+/HjkyJGXL19mtke9/nDk7B70GKvm6jTEmDE/TJCFtMLS+mne2Sculdn/SzXqiJGmGpuI+vR6uZcnM79OcmZG3sv3K3q6HGqBj5sOEV2+U9mvj9K8KY2TH48epmI1mEtEJy+XnQgvI6LJ76g17dLEUWruM7WyC+o9/PPGLsvkOz7bF1xMRBl5dURULRQPfO+53rTUD7flEVHai9oakTgrv46IdLU4/fooERHTSiNvD1Uhot46HCIa1E+Zr8PhKrOYmI7yqgYZlUtqsDbjEZG2xsvhV1TLiuaTNrAvE09HJoZKRJSZWyd7OEypyaMbb5WSYMryddj6uhwishr8Mj4oM+/v+6tpcdkzQEQNDeT/XWF/5+fcccmaDs/Cb1cSkfTCXLMzkJ7LdIbDLKwQkbISS575BOhonp6e165dCw4OLioqkiQXW7Vq1ZYtW4hI8laKsWHDhs2bN4eGhpqYmJSXl4eGhl6+fFmS+u2///3viRMniCgkJIRJjd+oiQsXLjAbhx0dHb/88sv169d36lBB8f3+++9McHpAQIBkbU5bW3v06NHM7x4eHhoaGs+fPw8JCVm1ahVzo66uLvMLE6WyZMkSJqtrD8hi+0qWLl3q6OgofbnMRubOnXv3L4oV0QNvOMSttM7X11cSxdeInp7ewYMHmz3Up0+fqKgo5nfmSi66urpM7oySkr+v7vHZZ58x1/FlMBdpbrVy6dNMTU2l95WYmprKCIhISEho6ZCMFqWbk/jhhx9++OEHyZ/6+vrMVZnlLC7nJEin3ZVRbaNJ6N279759+5iYQwkZ8ynPaT4+Pj4+PpI/mwYxmpmZ3bhxQ0YnORyOnMORMTQ5HxUKodF2tmZvkYcC7U9+TYmpolmf5zzPqf3wPe3D6/pIMp6OsVbV1eIUldXvPFE8foSqKo9VVFZ/8McSIrIcyG02aIUx3JQ3+R21a39Uec3TaTZ/qoez1pqvBCcvlZdVNlALWWzZLArapP/1er3Hz0WXfqvc8m2h/3dFq916MUE06qrsczsMOX8t4wtrxUoc6q+vRERFZfW5hfUGvTmPnjVzPSNlJSIi5hHRtG8yKpdgbm/0mOKwWUQkc9MkxScL503REIrET9JERGSkryRPi8xurGYxZQUlDfnF9Xq9OImpLzcpMDW3VFz2DBDR+eiKnSeLNdXYp/wN+usrbfqmMPpBtfSzodkZMNZnOlOflCEyN34ZtyLPfAJ0NEm+MOnkYpKjTHIxFZWXybmlc5ylpaVlZGTU19dTC6nfJB99kZIM2otQKFy6dGlYWJi1tXVycnJcXJyysvKwYcOOHz/OpFwJCgr6+uuvk5KS8vPzR4wYQUQvXryQZAk4derUjh07mPjukpKSsLCwLhxL58vIyGial1Cavr6+5EtK+b+ZBuhyeLB2FH9//8TExLFjxxYUFDD7hFevXt22j44AAERktzyzpLxBTYWVmVc76/McIjI35n71eR9VHuvAmj5LtuRev1c1eM5ziwHc+BRRUVm9Cpd15N96sus85qeflCGytW7+WkKLZ2htOFyYI6gjIh1N9hyHZi7sfSe+etWugglvqRrwOQkpQvorUmbmOHVzY25Shmh7UJGjjVp5ZUNckpDDoZnj1B3eUjU1Uk7Jqp2wInOMterFm/Lm4ZaQUbnsggMMlYjo9NWyiuoGGysVt2maTc/ZG1ycml2blCHKK6rXUGXPm6L5Oi0SkdM49b/GmzV+hMrFm5VE5DJZoy//tf4F19W/XEcRisT3nwrvJtTIPv9lZ8aqDzXhPkkTjfswa5aderVQLKwVBwcYtHl0AO1Fki9Mklzsp59+YlKPEZFQKJT+iPXw4cN58+ZJ5zhjrh7ApH7T09NrNfWbPPnpAGS4cuXKqFGj1q9fP3HixNGjRxcVFd28efPcuXPMUT8/vylTpgwZMmTo0KFFRUUXL17cuHFjVdXLlFgFBQU///zz+++/T0TBwcGNLq3QnTVNTiS9BiohuaCEjIKMRnHcSH4ECg1rKx1l5MiR4eHht27dUlNTs7a29vb2lj/1CQBAU0xmk6oaMXOxGyJiLqxDRG7TNPvrK+0+Vfz7o5rbD2v4Ouz5jpq+nrrNbreRZmygZNxkm49En16cd+3Uz0dVME2ocJtZHdbvrcTX4YT8Ul5c1qCpxpoyWm2HF5+IeFzWzW+Nth0ruvJ7ZWBQkaYa23Ig12OaFhEpK7Eufdlvxfa82Mc1cUnC7Sv5q3blE5ESR97VZxmVy+bj1uvBU2FcsvBgaElZpVazaytbV/S+dLvy8XPRSHPeV5/3YTbytLlFIlLhsm5913/z0cLImKrgq+XGBspe83TWLe4l52Bb4jJZ83Zc9Zmr5esPCaa8ozZ9jPqFG62vUjED2R5UFH67MjiyXFudvXSW1uuMDqDdyZNcrGmOM21t7VdK/dbm/HQAEgkJCcxVfpoKDAwMDAyUUZbZDQQAPQmrM8PpmaiNNyeAHwBkaK8wrq59SWFG0RBj1oV9UERMXlXm9+8ulK74Il9Xi1MQOahrY/v6TE0tLK0/v8tw9oRmInSgbdg2ydTVz9O2wZuWTsPn8wsLC8PCwmbPns3ckp+fHxAQcOXKlaysLE1NTUtLS09PTw8PD8nJu3btCg8Pj42NtbCwOHDggJ2dHRHl5eVt2rQpMjIyLy/P2NjYzc1t3bp1zDaipk2kpKQsWrQoLi5OKBS6u7tLX4oIOoJiPaG6c6R5B81hdx6yPBTloQU9G+JWAACgs239b2FiqmjscJWCkvrj4WVE5OOmo+Dv6wCgjZrmC2s1RVqzOc5eKfWb7Px0AAAArwprKwAA0NlGmPPCb1feiqtWU2FbD+Z5zdNpdnsOAAAAAIBCwNoKAHQXFRUV6up/5868du3a1KlTjxw5Mn/+fCUlpaSkpAkTJpSXl3dhD6G9uM/Ucp/Z7dJ5FEQO6uouAAAAAIBCwtoKAHQXzs7OSkpKoaGhT5482bx5c2FhoaOj4/Lly5ctW5aWlmZvb6/om4EBAOA1Nd3dAwAA0B1gbQUAuovo6GgiEolEBQUFv/76K/2VRf+tt95KSUkJCAior6/v2h4CAAAAAAA0xe7qDgAAtOjSpUsRERGrVq26ceNGcnKykZFRV/cIAAAAAACgMaytAED3VVNTM2PGDGNj482bNw8cOJAJYwEAAAAAAOhWsCcIALovJycne3v7qKgoJoUtttkDAAAAAEA3hLUVAOi+8vLyHBwcvLy86urqTp8+HRoa2tU9AgAAAAAAaIwlFos7rzEWi4g6s0UA6Lba66I/XfuSwoyiIcasC/sA0J2xbZKpq5+nbYM3LQDtSLGeUN35uoQdNIfdecjyUJSHFvRsyLcCAAAAAAAAANB2XbAnSNGXRQEAAAAAAAAAJBC3AgAAAAAAAADQdp2abwUAQKIn5VsBANkU8c2GYqWHAOjmFOsJ1Z3/uSvKHAK8gXCdIAAAAIDmdeePWADQQbB+AQBtgLUVAIC2w9svAAAAAADA2goAAABAY1g5BQAAAPkhly0AAAAAAAAAQNshly0AAAAAAAAAQNshbgUAAAAAAAAAoO2wtgIAAAAAAAAA0HZYWwEAAAAAAAAAaDusrQAAAAAAAAAAtB3WVgAAAAAAAAAA2g5rKwAAAAAAAAAAbYe1FQAAAAAAAACAtlN6ncIsFqu9+iEnsVjcyS0CQAdpxxcQvDIAAAAAAEAXQtwKAAAAAAAAAEDbvVbcCqNzvjHu/BgZAAAAAAAAAIBWIW4FAAAAAAAAAKDt2iFupTMhegUAAAAAAAAAuhXErQAAAAAAAAAAtJ2Cxa3gaiAAPQbC0AAAAAAAoGfokLiViIgIFosVFxcnfxGBQLBly5bHjx93RH8AQCHY2tqKxWIPD49GvzQ0NJSUlERFRdna2nZ1H/+BBQBy6OpnKgDAq+nqV81uCrPaLo8u6MG6S9yKQCDw9/e3srKytLTs6r4AQPeyZ8+ehISETZs2RUREmJmZFRQUdHWPAKDnw9togPaF8HMA6Nk6cG3l8OHD58+fHzZsWGRkJJvNXrly5fnz57lcrre3t6+v759//unm5paamqqvr3/y5MlFixYRkYuLy+DBg1NSUjquVwCgcB4/fnzy5Mm0tLQbN264ubl99dVXXd2jf2iIMevqLgB0U2yb5K7uAgAAAEBn6MC1FRaL5e/v7+3tHRkZmZaW9v3331+8ePF///vf2rVr7e3tIyIiSktLY2Nj6+rqdHV1Q0JC7Ozsjh49OmvWrI7rEgAoridPnhDRgAEDurojAPAGwTftAK8PUWAA8CbowOsEffTRRzNnziSi4uLihIQEHR2dd99918XFhYgSEhLmzp2rp6f31ltvvf/++8+fP+fz+USkq6trYGDQcV0CAMVlYWFBRBkZGV3dEQAAAAAAgH/owLUVZWVlZpVaLBZbWVmVlJRcuXLl/PnzRDRs2DAOhxMaGvr8+XORSHT27FkVFRUievbsmUAg6LguAYAisrS0XLx48bFjx0pLS8+cOdPV3QEAAAAAAPiHTsplu2LFiri4ODc3Ny6X6+/vb29vf+rUqXXr1hUWFg4aNMjd3d3Y2HjKlCm+vr5RUVERERGd0ysAUAhr1qypqKj4888/lyxZkpeX19XdAQAAAAAA+AfW62wkloSltF9/ukVbANAJ2nH3dVe9MjBDQC5bgJYwuWwV9H833ngAtBeFezYhQUyzXvMe7AGzqkCPYegSHbgnCAAAAAAAAACgx8PaCgAAAAAAAABA22FtBQAAAAAAAACg7bC2AgAAAAAAAADQdlhb6UCJiYmLFy/u378/j8czNDR0cnK6ceOG7CJ8Pp/FYl24cKHpn40Otdn06dNZLJafn19Ljb6+7Oxsd3f3fv36cblcPp9vY2Nz5MiR9qq8q8g5S+0+mQAAAAAAANDNtcM1mHtAzueOEBERMXv2bKFQqK6u/s4774hEops3bw4aNGjChAnyVzJq1KiSkhJdXd1Xarq2tlZZWbmjW5Fh/vz5v/32G5/Pd3BwKC0tffr06bVr1z7++ON2qfxVR9fJ2n0yqdsPGQAAAAAA4A2HuJUOIRQKPT09hUKhtbV1UlLS7du3Y2Njs7OzlyxZQkTLli0bMGCAmpoaj8cbOnTo4cOHW6rn3r17MTExRUVFkluePXs2adIkNTW1t99++86dO8yNTKzE9u3bx40bx+Vyo6OjW2rFysrq6tWrRBQYGMhisaysrJq2UlBQsHLlysGDB6urq1taWgYGBgqFwkYNTZkyRUtLy8rKStIHiZqaGubG69evR0ZGxsTEFBQUbN++nTlaXFy8Zs2aIUOGqKqqGhgYbNiwQf5GJaMTCATe3t6mpqbq6uojRow4fvx4s7PX0mlhYWEsFqt37945OTn19fXjxo1jsVifffaZpK3AwMCJEydqaWlJT3IjLd2JjSZTxozJGEWzdygAAAAAAAB0T6+1tiJ+RW0r1bSG7u/u3bu5ublEFBAQ0LdvX+ZGbW3t0aNHE1F8fLyDg8OGDRs8PDySk5O9vLyuXbsmZ82+vr58Pn/IkCEPHjyYPn16fn6+5JCfn5+KisrcuXM1NDRaamXx4sUDBgwgotGjR3/66aeLFy9uVL9QKBw/fvyRI0c4HI6rq2t+fr6fn5+Hh4f0OVu3brW0tLS0tExMTHR3d29Ug4qKSr9+/Yho4cKF//nPfy5evFhVVWVubk5EIpHI3t5+7969AoFgwYIF48ePf/z4sZyNSkbH5XLt7OwOHTqkr6+/bt264uJiT0/P7777rlE3RCJRS6fNmTPHy8urqKho6dKlO3bsuHPnzqhRo3bt2iUpu3v3bgcHh2nTpjWdZIlXuhObzpiM7rV0hwK0xNo1nW2T7L27QHLLwPees22SNxwWMH9+tq/gX4sydKc805iQMtwt/dDZEsmZvz2snr0mR29aKm9cSr+ZzxduzH38XCS7IbZN8p7TxZIbp3+Szdy49oBg0sdZbJvkeRteSJeyW57FtkleuDH39UfaZ2oq2yb5wo2KRr83/bPNZnyazbZJ3vhNYbONAgAANMvCwoL5tMLn82WcFhERIRaLt23bJk+dr3SyQCAQi8WzZ89+pUPdUEJCglgsPnTokOSWtLQ0sVi8Y8cOIiopKWn2Q2JISIikrFgsXrNmjaT41atXmRv37NnTbP0A7aUd9gRBU8zCChGZmZk1PXrr1i0VFZWGhobq6uqcnJzw8PBffvll8uTJ8tS8du3agIAAkUhkYmLy4sWL0NBQLy8v5pCnp+exY8dkt7Jjx46oqKj09HRHR8dmX6kvX76clJSkpKR069YtfX39OXPmODs7h4SE7Nu3z9DQkDln/fr1/v7+9+7de+edd1JSUkpKSnR0dKQrOXv27IoVK+Lj4x89ekREXC5306ZNvr6+ly9fTkhIYLFYv/32m4WFBRHV1tbK2ahkdGFhYU+fPlVVVQ0ODlZXV+/fv/+yZcu++OKLjz76SLoPly5dknHanj177ty5c/Xq1cjISC0trZCQEC6XKym7YcOG9evXi8XiQYMGpaWlSU9y2+7EpjMWFRXV6iga3aE9j/RqKbO1KcRbbwAAIABJREFUUIHWTxXL0bDSAQbKTmPVU7JqYxNrPtlToKzEWjFH+8dfyhdtyq1vIIPenLHDVR6liM5Ell+4UfHLIaOxw1VkVHgotMTHtReHTYmposiYKsntnrO0ox9U/3y7sri8oZcmm4ieZdX+9rCaiDzf1WrfQY0ayispb9DV4rxqwdo6sbLSK2xlbXNDnexVxwXwJps+ffrVq1d9fX3l/NQKAN3EH3/8oampSUQjRoxQUVHJzc1NT08nouTkZOnTvL299+/fX19fP2zYsKlTp3ZNX+HNgz1BHUJfX5/5pdHznIgaGhq2bNliZGSkrKysoaERHh5ORHl5eXLWzOzi4XK5TCRIRkaG5NCUKVNevxWmQj6fzwyBaY6IMjMzJecMHz6ciLS1tZk/Kyoaf51ra2v78OHD1NTUM2fOODs7i0QiPz+/5ORk5rWPz+czCytExKQRkadRyeiYk6urq01MTPr06bNs2TIiSktLq6mpaTqQlk7j8XjMerZYLHZ1dR08eLB02YEDBxIRi8UyMTGhf05y26a36YzJMwrpOxTgdVz+st/TswN+2Grw+7H+w015RBR5t6paKPbaXVDfQJNGqT0LGxh1xCjppwGWA7nVQvH/7ZD1WqGrxcnIrfvpegUR7Q8uJqLe2i/XHeZO0tBSZwtF4pDIcuaWk5fLiGiAgdLkd9SaVlVc3rD2gMDCJV3NLsVwRup/vhYQ0Yfb8kzee65un6IyPsVyfvrhcyVNCxLRvSfCmMSaorJ66RufZdVOXpmlbp8yaknGnfiXzyYm8GT78aLxH2XyxqVEP6huqQlr1/Srd6uIKDCoiG2TbO2a3rShguL6VbvyTd9P05iQMmx+emBQkVAkbtSQo1e29sRn1q7pkj5I23u62PyDNFW7FO2Jz0a4ZTBT1NJsyNOcZFyCkvpP9hSYfZCmMSFl5MKM4+FlMu5HgO6J+dJF0ZuAN4SZmVlUVFRlZeWDBw9sbGykDx07diw9Pb2qqkooFD558mTVqlVElJCQMG3aNCLy9fUVi8UJCQmvenKzZ0obPHjw9evXq6qq7t+/P3bs2KZ95vP5Bw8eTElJqaysfPjwYaM48e7P0dHR1tbW1taW+Vjx448/Mn9u3LhRck5RUZGxsfEHH3xARKtXryaiwsLClioEaEdYW+kQY8aMMTAwIKKNGze+ePEyPL60tDQ2Nvb8+fM7d+4sKyv74Ycfbt265eDgQK/ydT2ziaa2tjYpKYmI+vfvLzmkqqoq+V1GKxwOh4gaGhqarZ+pUCAQMBthEhMTmduNjIwk5zA1yMhhfPz48YqKioEDBy5YsECSiIR5mWMqZzpPf72/kadRyeiYk9XV1SMiIn75y88//6yk9I8gLNmnFRYWrlu3johUVFSOHTvWKKnKw4cPiYj5p9Voklud3mY1nTF5RiF9hwK06tz18jFLM5mfF4X/WHGY+PbLx1KDmKqFDUQ0ZIDy3YSawtJ6Ilrv3kuVxyIiXS3OJwt0iCjhmSjtRYufPTzf1eIqs748U5xfXH/6avnbFrxRQ3nMIVUea4GjJhGdvFRGRGIxnbpcTkQezlpNXzBEteIJKzL3ni4WlNQvmKo5foTq41QREcWniBzeUvu3u67HTK3kDJH37oJrf1Q1LtwCv28K+TqcIQO4D/4nnPFpdn7x3/Ow8ZtCFS577iQNDVV2S00smqE1wECJiEYPU/l0vs6iGY1jbYQisd3yrCM/lXLY5DpVM7+4fuM3hZ5b/7EUFXCsyHIg13IgNzFV5OHfeCdUTGLN2gOCiuqGDR66Xi46AwyVnmXVtjQb8jQnGRdXiWW/IuvQ2RJ9Xc7axbrF5fVLA/K+u1Aq59QBdJVXyqomIy2a/InbuFxu09xzAK9KWVn58uXLDg4O+fn5T58+PXDggPTR4cOHR0dHf/HFF8ePHzczMzt06NDkyZNPnTrFrAjExsZ+9dVXp06detWTmz1Tut3AwECBQPC///3vrbfeioiI0NPTkz7K5XJv3brl5eWVl5e3a9euXr16BQUFNQr97ibmzp179y+SMHY5BQUFiUQiHx8fPT29hQsX3r9//969ex3UTwBp2BPUIXg83vfffz9nzpxHjx6ZmZmNHDmytrY2MTHRw8Nj/PjxzDlCofD+/ft37959pZp37dqVnJycnJz84sULDQ2NefPmNXtaXV1dS60w+VZOnz5dUVFha2vr5uYmXdDJycnU1DQlJcXe3n78+PEXL14kIhcXF0nWGHl4enquXLnS3NxcT0+PWacwNDS0trYeOXLk0KFDnzx5Mnbs2FmzZlVXVwuFwvPnz79SozNnzjQ3N09KSgoMDJw6dWpZWVlcXByHw5k5c6b8p3l4eGRlZXl6ek6YMMHDw8PV1TUuLq5Xr15M2b1796ampiYlJeXl5TU7yTKmV05yjgJAfnlF9XlF9TJOqK0TL9+en5xZO8KMt8FD9/Jvlczt/fX+/kdgrP/yilT5RfUmhs1fnUq/N8d1quaJS2VLt+YJReLPXHv9cOXv+Iils7S+vVAak1jzNE2UW1Sf9qKWxSKPd7Wb1nP5TmXCMxGLRbe/M7Iw4TI9JKKb3xqpcFkNDVQtbMgR1IXfrvw1tqrZsJem1i7qtXVFb1GteODstBeCutBfy71cXu5Y9HTW+q/fy4jClppYv6RX9P2q9Nw6x9FqAf/Xu9k+J2WIlDism9/219flzHaofHd1Tsgv5Xs/4xvyX07j+sW9tizvfe9JzWiPzJSs2pLyBh3Nv7/GKK9sICKeMsvEUMncmOu7VFdNhdXSbITfbr05ybjCoiuepolUeazTAYbqKqz++kofbsvbcaL4o9nNTD5Ad+Pn5zdx4kRJVrWnT5+OHTt28eLFx44d8/T0rK2tlf74t3v3bh8fHz6ff+7cuenTp6ekpGhra48fPz4pKcnMzMzV1fXChQt+fn4JCQlnzpxp2kRVVVVcXFx6evro0aPHjBnzqh/bABgODg6mpqYNDQ329vaZmZmenp7ff/+95KidnV1NTQ2bzVZVVe3bt6+zs7Ojo+O///3viRMnDhgw4JdffvHz82vDyc2eKZ3yb/fu3Rs3buRyuWlpaYaGhvPmzZNOLDJz5kwLC4vq6mo3N7fKysrMzMxjx45t2LChaeLCLqevry/ZB/CqcnNzz5w54+7uHhQUxOPx9u/f3zTFJEBHQNxKR5kxY8a9e/cWLVqko6MTGxubnp5ub28/d+5cFxeXlStXKikprVu3LiYmZvr06a9U7datW7OzsxMSEkaOHHn58uWWXnRktLJ69WobG5u8vLyDBw9GRkY2KqiionL79u3ly5cLhcLg4ODevXv7+/ufPHnylTq5bds2Gxub/Pz8GzdusNns99577+rVq8wldW7duuXj46OrqxscHHz9+vUhQ4a8aqNMJV5eXjk5Odu2bQsKChIKhfPnz5f/tP3794eHhw8aNOjAgQPu7u4uLi4ZGRmenp6SsgEBATk5OY8fP25pkl/zTpR/FADyWzVXpyHGjPlhgi+kFZbWT/POPnGpzP5fqlFHjDTV2H16vdzIk5lfJzktI+9luIqerqz0Ij5uOkR0+U5lvz5K86b8I9fy6GEqVoO5RHTyctmJ8DIimvyOWtP+EFF6bh0R8XU4zFICESkrsRoayP+7wv7Oz7njkjUdnoXfriQi2WtG0oYN4hIRV5llbqxMRJm5fw9t8uiXqzOv00RGHtNntr4uh4isBr8M2MnM+7shazMeEWlrvJzAiup/BAlOHKXmPlMru6Dewz9v7LJMvuOzfcHFLc2GPM1JxsWcXC0UD3zvud601A+35RFR2ovaGhHSGIEC8PT0vHbtWnBwcFFRkSQf2apVq7Zs2UJEX3zxhfTJGzZs2Lx5c2hoqImJSXl5eWhoqHTitv/+978nTpwgopCQEEnssHQTFy5cYPYmOzo6fvnll+vXr+/UoUJPwcQgFxYWMnvYHzx4IDnEZrO3bNmSlZVVW1tbUVHh7OxMUhkDGpH/ZHnOZLYOiUQiJkiciRmXYP5UVVVNS0srKChg8vqZmJioqMhKstYlDh8+zPoLE7/zSvbt20dETk5O2dnZoaGhHdBBgGYgbqUDWVlZSYL9pB0+fLil6y4LBIKW/pT8vnbtWtmliIjD4bTUiqmpaaM4i0bF9fX1jx492mr3TE1NW9oF4+vr6+vr2+yh3r1779u3j3m9a0OjDD09vYMHDzZ7sjyn+fj4+Pj4SP5s+oJrZmZ248YNGd2QMb0y7sFGMyZjFE2HDPA6ElNFsz7PeZ5T++F72ofX9WEyno6xVtXV4hSV1e88UTx+hKoqj1VUVn/wxxIishzIbSlohTHclDf5HbVrf1R5zdNpmj/Vw1lrzVeCk5fKyyobqOUstsb6SkQkKKlPyhCZG7+M1Lh4s3LnyWJNNfYpf4P++kqbvimMflAtf47jJ89FTD1JGbVEZKQvtcmO97Kf56MrZDTBYbOIqIVNky9jfAQlDfnF9Xq9OImpL3ccSDfEYRMRtbRpks2ioE36X6/Xe/xcdOm3yi3fFvp/V7TarVezsyFPc5JxMSerq7LP7TDk/PW9ibBWrNTdk/ACELWQVU1ylMlHJvn4J50WLS0tLSMjo76+nlpI3CYJS0EWM2hfWVlZRNSrVy9tbe3S0lLmYcl4//33169fX15evmjRoszMzICAAAcHB2ZvOPNYZbPZbThZxpkSlpaWRKSsrMxkZpTOXSj5s7Ky8oMPPmAqJyIejyeJyO4x4uPjr127Nnny5IMHDyLFEnQarK0AAPRwdsszS8ob1FRYmXm1sz7PISJzY+5Xn/c5sKbPki251+9VDZ7z3GIANz5FVFRWr8JlHfm3Xqt1HvPTT8oQ2Vo3803X4hlaGw4X5gjqiEhHkz3HofmLiDuNVR9qwn2SJhr3YdYsO/VqoVhYK3b5KwpGKBLffyq8m9BMLlgZdp0qTs6sTc6sfSGo01Blz5ui2fScunqxjCYGGCoR0emrZRXVDTZWKm7T/lGD0zh1UyPllKzaCSuyxo9QuXizkohcJmv05cv7z/ROfPWqXQUT3lI14HMSUoT0V4hQs7NxOsBA/uZmjlM3N+YmZYi2BxU52qiVVzbEJQk5HJo5Tl3OvgF0oaZZ1X766ScmWxkRCYVC6XxkDx8+nDdvnnRatH79+tFfidv09PRkJ26j1nLPAcgjOjo6NTV10KBB58+fj4iIkE4rK3m48ni8t99+29bWVnKICcFYuHChhobG3bt3g4OD5T9Z8ohteqbEunXrzMzMzMzMDA0NKyoqGn2DeOnSpaSkJHNzc19fX+ZamSNHjqyvr7906VK7zEm3snTpUnNz85Z27s+dO3fUqFHM70+ePJEOYAdoM+wJAgDo4ZgcH1U14qt3q5ifO/HVROQ2TTPqiJHzePXaOrr9sIarTPMdNWOPG9uNbD2PsrGB0pTRahqqzfwT6dOL867dy8/zbtM0VbjNh3DwuKyb3xr5uOroarGDI8uv36saMkDZZbLmyrnaShxaf0gQm1AzfcyrrQv4r+idXVCX8Ew40px36cu++s3tbJLdhI9bL5thKnlF9QdDS36JaZxDV4XLuvVd/+WztYUicfDV8t7anC3Le5/YYiB/D/V7K/F1OCG/lG8+WnTtj6opo9V+3G7Y0my8UnNMDV4uOjmCusCgoqDwMmGtuNnVJYDujMlHVllZGRgYePfu3cjIyF27dh06dEh6bWXv3r2urq5jx46VpEVjErfV1dXZ29t/+OGH7u7uJDNbnCT33CeffBIcHNw5Q4MeRiQSOTk53bx5c/To0QsXLpQOyj579uzXX39dV1e3a9cuGxubiIgIyaF9+/bFxMTo6+t7e3sz1waW/2QZZ0ps2rSpX79+VlZWcXFxTk5Oja5iKRQK7ezsDh061LdvXz8/P09PTx6P9+OPP7b/7HQDGRkZv/76a9PrmTL09fVt/mJtbd3JfYOeiiX/FWraoTEWi17lmjgA8IZQ0BcHptsNMWZd3RGAboptk0wK+NRmKOjrksLh8/mFhYVhYWGzZ89mbsnPzw8ICLhy5UpWVpampqalpaWnpydzpVjm5F27doWHh8fGxlpYWBw4cMDOzo6I8vLyNm3aFBkZmZeXZ2xs7Obmtm7dOmYbUdMmUlJSFi1aFBcXJxQK3d3dpS9FBB1B4Z5NMq6G+SZ7zXuwB8yqAj2GoUtgbQUAup6CvjhgbQVANqytQPtqukoCCkHhnk09YBWgI2BtRYEew9AlsCcIAAAAAAAAAKDtsLYCAAAAAAAAANB2uE4QAAAAACgAgUDQ1V0AAABoHuJWAAAAAAAAAADaDmsrAAAAAAAAAABth7UVAAAAAAAAAIC2w9oKAAAAAAAAAEDbYW0FAAAAAAAAAKDtsLYCAAAAAAAAANB2XXANZhaL1fmNAgAAAAAAAAB0BMStAAAAAAAAAAC0XafGrYjF4s5sDgAAAAAAAACgo7Gw3gEA0DbY4QggDwV9p8E8wRW08wDdisI9m/D/vVmveQ/2gFlVoMcwdIkuyLcCAAAAoBB6wIcBAAAA6ASIWwEAAABoDKsqAO0LHzoAoGfD2goAAAAAAAAAQNvhOkEAAAAAAAAAAG2HtRUAAAAAAAAAgLbD2goAAAAAAAAAQNthbQUAAAAAAAAAoO2wtgIAAAAAAAAA0HZYWwEAAAAAAAAAaDusrQAAAAAAAAAAtB3WVgAAAAAAAAAA2k7pdQqzWKz26oecxGJxJ7cIAAAAAAAAACDDa62tAAC0WTsuzmLVFQAAAAAAulA7rK10zqeazo+RAQAAAAAAAABoFfKtAAAAAAAAAAC0HdZWAAAAAAAAAADaTsHyrWBnEAAAAAAAAAB0K4hbAQAAAAAAAABouw6JW4mIiJgxY8aff/45cuRIOYsIBIJDhw7NmzfP0tJSxmm4GghAj9EDwtB6wBAAOgH+dwMAAECP113iVgQCgb+//+PHj7u6IwDQZWxtbcVisYeHR6NfGhoaSkpKoqKibG1tu7qPAPCmYAFAu+rq5zQAQMfqwHwrhw8fPn/+/LBhwyIjI9ls9sr/Z+/O42pK3wCAP7f9tmunnVKShBRSaoSyzaKMkMq+lG0sg2wjyyDGNjMmVHaJGMqu7DKWEKVNaV9UKtW93bq/P96Z87tzl9MtpZt5vh9/XGd5t9M995z3PO975s49e/asnJxcYGDgqlWrnj17NnHixMzMTF1d3cOHD0+ePBkAvLy8unXrlp6e3nalQgh1ONu3b09KSlqzZs3ly5fNzc1LSkrau0T/0phg3t5FQEhCSTmktXcREEKo2TpET1CrhARKeE1bN+xRwisrDowDlXBt2LfCYDDWr18fGBh49erVrKysQ4cOnT9//s2bN0uXLnV2dr58+fKHDx8ePXrE4XA0NDROnjzp5OS0f//+sWPHtl2REEId0evXrw8fPpyVlXXr1q2JEyfu2rWrvUuEEPqvwAtZhD7dF3BPixBCTWrDMUEzZswYNWoUAJSXlyclJamrq48ZM8bLywsAkpKSPD09dXR0+vbt+9133719+1ZLSwsANDQ09PT02q5ICKGOKzk5GQCMjY3buyAIIYQQQggh9C9t2LciKytLeqm5XK61tXVFRcWlS5fOnj0LAD179pSWlo6MjHz79i2bzT59+rSCggIAZGRklJaWtl2REEIdl6WlJQC8e/euvQuCEEIIIYQQQv/ymeaynTVrlp+f38SJE7ds2bJ+/XpnZ+fnz58PGTLEzMxMTU3N19fXyMjIzc1t1apVZOIVhBCiWFlZ+fj4HDx48MOHDydOnGjv4iCEEEIIIYTQv7TJfCvu7u7U+GTqw6FDhw4dOkRt4+Pj4+Pjw7vXtWvX2qIwCKGObsmSJdXV1c+ePZsyZUpRUVF7FwchhBBCCCGE/qUN57JFCKFmefjwITXdneAHhBBCCCGEEJJMn2lMEEIIIYQQQgghhNAXCftWEEIIIYQQQgghhFoO+1YQQgghhBBCCCGEWg77VhBCCCGEEEIIIYRaDvtWJJG1tTXjH4qKir169fr1119blpSWlhaDwTh37pzgKnd3dwaDERQU1OJyCqZAk52EELOEkl8RhBBCCCGEEEISohXeE4Rv8WgjVlZWAwcOTEtLu3379rx583r06OHq6trehWqCnZ1dRUWFhoZGexfkU7V6Rerr62VlZVsrNYQQQgghhBBCkgPjViSXq6vrgQMHbt68Se7Jk5KSAGDatGnGxsaKiory8vI9evTYt28ftX15efmSJUssLCyYTKaent6KFSt4U+NyuTNmzGAwGD169MjNzbW2tr5y5QoAbNy4kcFgWFtbA0BpaWlgYKCZmZmSklLv3r3Dw8Op3UNCQszNzRUUFFRVVW1sbA4fPiw0hcePHyckJJSVlZG9SPTHpk2b3NzcVFVVra2t79+/T1alpaV99dVXJKOwsDASpFNaWirYDkJLFR0dzWAwNDU18/PzGxoaHB0dGQzGwoULefPduHGjq6urqqpqv379qHx50TQmb0VoakHTYtRejo6OcnJy8fHxTR1whD6rXt7ZUg5pgdtKqCWmX7+Vckhbse/vr+HCHSV9Jr/TcMtQHpJuMzF77+kKast7z2u/WZKvMyJT3jFdf9TbSasLX79l02ck5ZC2/Vg5tdB9fh5ZuHR3KQB8NSdXyiFt/IoC3h2dZuZKOaRNWl3Il6D28Ewph7Rzt6o/ofZNJ+WxIE/KIW317+9bnLhgCq1YcoQQQgghJFE+qW+F20wt20swhf+I+Pj42bNnjxgxor6+Xk1Nzd3dHQBevHjh4uKyYsUKPz+/tLS0gICAGzduAACbzXZ2dg4JCSktLZ0wYcLgwYNfv35NJUU6Vg4cONCnT5/bt28bGBj4+PgYGxsDgL29/YIFC3x8fNhstpOT0969e3V1dZctW1ZeXu7v7x8aGgoACQkJS5Ysqa6uXrlyZUBAgLGxcXp6umAKoiry008/WVlZWVlZvXr1ytfXl5TWw8MjLi6uc+fOtra2ixcvFrWvqFJ9++23AQEBZWVlU6dO3bJly/379+3s7LZu3cq777Zt21xcXEaMGPH06VN3d/fi4mK+xEU1pvi1ENVilKCgIAUFBU9PT2VlZVEp/2e14smhfSvypdof/aGWxR05SMm6m3xSBnv+9pL90R8A4NS1KpfZuX/e+SgjDYNsFFhs7omrVf19391/UUef4N7IioZGAIBXmeyrCTW8q/zHqgHAhbsfy6sayZKM3Pp7z2sBwH+MahtUrn3Y9ZB36KmgoSrd3gVB6L/u04dFI4QQQnwwbkVyvXr1av/+/eRu39bWVk1NDQDu3LkTERGxatWqHTt2eHh4AMC1a9cAIDY2NikpicFg3Lt3LywsLCoqKioqikpq5cqVBw8edHR0jIuL09bWBoDly5dbWloCwLBhw3755Zfly5fHxMSkpKQwmczjx4/Pmzdv3bp1ALB582YAqKqqAgB5eXkTE5OxY8eeOnXqp59+EkxBVEWWL1++e/fuvXv3AkB6enpFRUV8fHxGRoaUlNStW7ciIiL27Nkjal+aUm3fvr1v375XrlxZvXq1qqrqyZMn5eTkePddsWLF2rVrIyMjTUxMqqqqIiMj+RIX1Zhi1oKmbBR/f/8bN24cP3584MCBolJGSDLF/qKfctr46E96Dw4a2pjJA8DVhzW1LG7AtpKGRvjKTjEj2jTuN4PUM8ZWpnK1LO7sLUU0qWmoSr8r5Jy5WQ0AO4+XA4Cm2v+7GDy/UlZVkmKxuSevVpElh2MrAcBYT2Zof0WhCWbk1g+dm6vknG435f/dOtODi0y+fqvknK4wON3q++x9Uf+PtSmvaly6u9TSK1vRKb2zR+bKX/8VJcflwsxNxVIOaVbfZ+cWc3p5Z195WAMAG8PKpBzSenlnA0BpRcP87SXm47KUh6TbTnoXfrGS2j3kWHn3cVlMp3Q114zeE98djq0UmsLjZFbCq7qyygayFwlj2RReNiwgT801o5d3NlWRtJz6oXNzSUZhFytJjE9pRUMTxwyhjq++vv4LyAL9RyQlJXG5XHJxqK+vHxERkZeXx2azS0tLExIS5syZ094FbE09e/Y8cuRITk4Oi8UqKCiIjY0dMmQIWSWq7nFxcVwu9/Tp07zp3L17l8vlHj9+vB3qIB7ew0pkZWVxudwtW7YAQEVFhdDnfCdPnqT25XK5S5YsoXa/cuUKWbh9+3ZRWaAvA/atSK558+Zxudzi4mJ7e/tbt24tXLiwsbFx3bp1BgYGsrKyysrKFy9eBICioiIAyM7OBgAtLS3S3wEAvLN7pKSkAMDUqVNJB41Q7969A4Da2loTExNtbe1p06YBQFZWVl1dnaurq6+vb15enq+v78CBAzU1NXfs2CF+RWxsbACAyrq6ujo3NxcANDQ09PX1AYCMJ2puqeTl5clpi8vlent7d+vWjW9fU1NTAGAwGCYmJlRSFJrGFLMWNGWj9nJzcxOjhRBqN1E3qwZOzSH/Ct7/69bdtR+TfGjkQi2rEQAsjGUfJtW9/9AAAMt9OzHlGQCgoSo9f4I6ACRlsLMKRN6x+I9RlZNl/HKivLi84diVqn6W8nY95Km1THnGhGEqAHA4phIAuFw4ElsFAH6jVUXN6BX0+3stdWkLY7mnb1geC/KKyxsA4EU626Wv4o++Gn6jVNPesQO3ldz4qwYA2PXcIbNyQo6Vl1Y0TBiuMrg383Xm/wcxcbkwc1PRgfMf+ljI3/rdwEBHZrKHqrGeDADY91RY8L36ZA9Vdj3XeVbu3tMVuhrSS300yqsapm4oCj33AQASXtUt3V1aXdu4wk8jwEvduLNMRm69YAqiWmbDwTIrUzkrU7lXmWy/9YWktCMX5sU9qe2sJWPbXf6HX0pE7YvQF0BwCG2T422FjvktKSmZO3dut27dlJSUrKysNm7cyGKxhGYhJycnOKgZoU906tSpKVOmkL/hjIwMS0vLoUOHtnehWo30h9RMAAAgAElEQVS7u/uTJ08mT57cqVOnv/76Kycnx9nZ2cvLi6wVVfdDhw4BwJgxYzp16kS27Natm6OjI7WqI/rrr78SEhISEhLIBX9hYSH5b1paGu9mgYGB0tLSANCzZ8/hw4e3T1nRZ9cKc9miNqWtrW1pafno0aPk5OSzZ8/+/PPPKioqR48eNTQ0XL16dXx8PBkQYWRkBAClpaWpqandu3eHf0+e6u/vf+TIkVmzZqmpqY0bN44sJF/4xsa/I/ANDQ0BQElJ6cyZM2QVALBYLBkZGSkpqfDw8N9+++3169cxMTFr165dt27d4sWL+VIQhWzGO+cxyausrKywsFBPT+/ly5ei9qUp1fv375ctWwYACgoKBw8enDJlyqBBg3j3ff78+fjx41ksVnJyMpUUhaYxm1ULoWWjtmEymfSNg1D7KiprKCqji4ao53BnbipOy6nvbS6/wk8j9t5HstxQ5/9/50a6f59qissaTDoLn7NZV1Pae7hKREzl1J+KWGzuQu9ORy9V8m4wdazqH+c+JLyqS8liF5Y1ZBXUMxjgN0Zkd/DSyZ1+mqXJrueafpNVUMqJvF4V4KV++w8DBTlGYyPUshrzSzkX7368/qhmaH/F2PsfkzLYDAbcDTWwNJEj9aKSWvXb+5QstmNv5sUdXdSUpQBg+ZRO8U9qsgs5w+wVN8zWBIDo+OqULDZTnnFsQ2clBYahrsz04KItEeUzvlGr+tgIAPKyDJPOMt2N5FZN1VBUYAAAXwqiLPfptG6m5uPkOnu/nPTc+oqqxkev6zJy66WkIP53A31tmaOXFKes4590BqEvTFBQkKurq6enp5ycnJOTU0pKyqBBg3x8fA4ePOjv719fXz9jxgxq423bti1atEhLSysqKsrd3T09PV1NTW3w4MGpqanm5ube3t7nzp0LCgpKSko6ceKEYBY1NTWJiYnZ2dn29vYDBw7s3Llze9QYfVEUFBTIVehXX31FLmvl5OTIsz0A0NbWXr9+/YgRI/T09LKzs48dO7Z9+3bS91daWqqpqblq1aqvvvrK3t7+3bt3M2fOFDpLYDuSl5cPCwuTl5d/+fKlu7t7fn4+AKipqVlYWABt3aOiovbu3auqqjphwoTffvsNAMiw+uzsbJph+BJu2LBh5ENKSoqFhcWpU6eoCR8pZWVlRkZG48aNi4yMJFMfvH//XlOT7mIAfRmwb0VykflWSkpKzp8/DwBOTk4cDoesYrFYT548efjwIbXxyJEje/TokZycPGjQoLFjx9bW1rJYrLNnz5K1Y8eOJbEn3t7eUVFRY8eOBQAyW8qxY8eqq6sHDBgwbty47t27p6ambty4cfjw4ZWVlYmJidLS0qNGjbp3797cuXOHDBnSuXNnctLU0dERTGHixIliVs3FxcXMzCw9Pd3Z2XngwIGkgkKNGjVKVKn8/Pxyc3P9/f2HDBni5+fn7e2dmJhI9YsDQEhISGZmZmpqalFRkbKy8vjx43lTpmlMMdGUrblJIdRe5nmq71mqTT6bfv02u5DDu/b9hwavHwvin9Y692Ge29ZFRVFKu9Pf3Yg5xRzSSQEA74r+DlfR0aCbSWTRRPWImMrY+x/1tWXGuynz9a3Y91Sw7iaXlME+HFtZUNoAAEP7K5LQD6F6dpUDADlZRncj2YJSTk4hp7ER1oe+P3qpquA9h+ryJT1HpF5a6tJUmWVl/t9PmpLFBgD/MaqkY0Wod0UcAKhlcU2/fkstzCqor2NzXe0UfUepHrtc5be+CAAU5BjBczQXT+wkKik+vczlAUBN+e+mq65tzC3mAICGqrS+tgwAWHeTo9kdoS+Dv7//wYMHASA6Opoab6ukpGRoaDht2rTNmzfz9q2sWLFi+fLlXC63a9euWVlZkZGR+vr6qampMjIyd+7c0dXV/fbbb0ePHn3y5MkdO3ZQXSdUFgDg7u6enZ09bNiw4ODgz19Z9OWpq6vLy8szMDA4duzYxYsXExISbt26lZqaCgDy8vJ3797t3r17WlraiRMnvvnmm+DgYGtra29vb2r3NWvW/PHHH8rKyg4ODhEREebm5u1XFSEGDBigp6cHAKtXryYdKwDw4cOHR48eAW3da2trT548OXPmTF9f399++43BYJD5GcPDwyV/sjxPT087Ozvyubk9sGFhYYGBgYsWLYqPj580adKTJ09KS0tHjBjRBsVEkgXHBEkuMt9KbGyssbHx0qVLf/75Zy8vr7lz58rIyCxbtiwhIYHMbkvIy8vfuXNn0aJFGhoax48fv3nzJulLpvj4+Ozevbu+vt7Ly+vSpUsAsHjxYgcHh6Kioj179ly9epWkEBAQkJ+fHxwcHBYWxmKxvv/+ewDQ1dXV0tI6efLkmjVrbty44ebmRuYu4UtB/KrJysrGxsa6uLjk5eUlJiZSc5TwRnzw1kuwVDt37rx48WLXrl13797t6+vr5eX17t07f39/3n03bNiQn5//+vVrW1vb2NhYXV1d3rU0jSkmmhZD6AvwKpNt75cT/7R2+tdq1/bqq6tIAcDAXkwyFevPEeW1LC4AlFU27DlVAQBWpnKiglYIGzN5MnlKwHh13q4Nit9oVQA4HFNFpmWhn8U2+S0bAOo53NR39QBgoCtzNr7658PllR8bj6zXu/2HgUtfJgCQizcjXRkAKK1oSH3391Ag3rgV/9GqMtKM2ZuLz8T9/w0+0lIMAKD6aEicjhJT6tIu/at7/v73Z0gXGWmQYkDYGt0Pcd3+ijBaN1Ozjs1dH1ommIIo0lIAALxDnwx1ZQCgrLKh8H0DALzMEPkOJoS+GNQQWnHG2wqO+SV7aWlpkd96aphPTk6OYBYItQUvL68XL1706tVrxYoV586dKyoqWrVqFQCMHDmye/fuHA7Hyclp+vTpJHBjwoQJvLfrP//88/z58wMCAgDAzMxMXV29vWohFOlYAQC+YS8UUXUHANKh6eDgYGlpOWTIEBMTEy6XGxYW9nlK/il0dXUd/sE3pWOTCgsLT5w4MWDAABLvs3PnzjYqJJI0GLciicjrloXat28f76uCeZFpUPhmQuF9q3FAQAA5axNmZmZ8wRo6OjpCp5U1MzMTGrknmALfS5R5/2tmZsbbRa2trR0XF0c+k3fraGhoCJ0ORlSpFi1aRH0WnKcWAMzNzW/dusW3kLdINI3JuxlNLUSVDQSaAgliCEykIbhEHJL/6KODcpqZU1HVqKjAyCmqH/tDPgB0N5Lb9YP27iXaU9YV3nxc0+3bt5bGci/S2WWVDQpyjN9+1GkyzYNBuqnv2AN6KQhd6+OhumLf+/xSDgCoq0h960L3dq2tR8rTcurTcuoLSjnKTKnxbipxT/5+9xCLzX2SwnqY9P87sZGDlHqYyCVnsR2n5451UqplcVn13DM//31dO8ZZycVO0W994cSgwtNbOo91UgIA484yAHDsSmV1baODtcI4V+XuRnKp79ibwsqGOShWfWxMTGVJS8MoR6V7z2vnbS0Z0peppyWdlM6Cf+J3+FKYOEKlyfYhXPoyzQxk03Prh8zKGdiLef42vrMZffmoIbTijLcVHPNLpm8rLS0tLi7W0dF59eoV2dLAwEAwCxAYFo3Qp3v48GHv3r1NTU0dHBwmTZo0evTo4ODgyMhIatg+mdSPusg3NDQsKCggn1+8eAEAHz58IP9VVlauqKgQkkc7oaYjNDc3530PKUVU3dPS0h49epSUlGRtbe3r60u6k27cuEGmiZRw+/bto+6bsrKySLS++Hbs2OHr6zty5Mi8vLzIyEiaF6qiLwnGraD2sX79+uHDh69bt27evHlkmOLixYtbdmuNEGoLZBqRmjrulYc15N/9F7UAMHGEStxvBqMHK9Vz4O7zOjlZ+H6YyqNwIyfbpmcXMtKTcbNXVGYK/+nR7iQ9xkmJfJ44QkVBju6EsH6WZl4JJymDZdtdPuaXLroa0l5DVeZ6qslIw/K9pY+S6twHKlEby8sxbv9hsMhbXUNV6vjVqpuPayyM/xVi4+OhsusH7XoOd/yKgkv3PwLAoomdHHoqFJU17ImsuJZQQ1II8FLPL+VsDCsLu1jJqueOd1MBAF1NGS116ZPXqtbuL7vxV42bveKpTZ0FU2iycSiyMoyYX/Rd+jLzSjiJqaxNc7XIchlpPEOiLx8Zb/vx48eNGzc+fPjw6tWrW7du3bt3L2/fSkhIiLe396BBg6gxvyNHjjQzM+NwOM7OzlRogJeXV5cuXYTmQg1qnj9/viS/rwR1IH5+fsrKym/fvj158uS8efPIQg0NDRI8paWlRQbU9+zZk6wiL3YgGhoaQIKfFT148KCwsBAANmzYQIXbqKmp2dvbk8+i6k4+kCiVKVOmkDkfO+4sts3y4sUL8mR6z549+G6y/w6MW0Htw9bW9uLFi3fu3FFUVOzVq1dgYKD407UghD7dyxP8T2Denjfl/W/9A5HjvZ1smeL0pIjKiLi0S19wYdSWpoc0l1ztSj4snfyvOU2kpWDvUp29S4WHz2iqSYcs1A5ZqC00KQAI8FIP8Pp/GLaZgeyDQ/+aAFunk/TuJdoA/0qBbHl9n5C6CKbAmx3ff80MZBsT/t/g2urSN3/7+3k7eRuRhqo0zXQwCH0xyHjbDRs2XLp0KTg4WEVFxcrKSnDM78WLF8mY3927d5NxQHfv3l2zZs3Vq1ePHz9uZGQUGBhIJrwXavHixU+fPk1MTNyzZ09lZSVegaBPFxYW9uuvv6amphYXF/fu3RsACgoKXr58+fz58/T0dDMzs9u3b9+9e/frr78GgNOnT1MTl0g+Fos1derU6OjoXr16paWlJSYmysrK9uzZMzw8nEy5IqruZPcjR45s2bKFdHRWVFRER0e3Y10+p6lTp3bv3p1mSkfeKV2Sk5P5TnSoI8K+FdQ+fH19yWOltoDjcRBCHdpPB96/ymQPslEoqWgIv1gJAIsmqmNgH/oiCf5k04y3JYSO+dXV1d2/f7+YWQgOakaoBcjgMhKVEBQU5ObmZmFh0aNHj7KysvPnz69evbqmpgYABg8e/NNPPw0fPnzixInv3r3bs2fP1q1b27nozXTp0iU7O7vly5e7urra29uXlZXdvn07KiqKrKWpOwCUlJRcuHDhu+++A4Djx4/zzp30ZaOmghJFV1eXmg5ScNJJ1BExPmf4GRnxIbEBbwihz6m1hoC14ymFVIE33AChVhERU7nxUFluMUdRQcrcUDZgvPrEESodsW9FyiENOuzvPl60SCAtLa33799HR0d/88037V0W1Awd7tskziWKkZFRamqqvLx8YGDg3r17P0Op+LRKe0r4ePzW/ZuR8MqKowN9if6bsIcMIYQQkiy+o1R9R9G9JgkhhFA72r17t5+fn7y8fEVFxZkzZ9q7OAghiYB9KwghhBBCqAPAMb9IQpibm9fU1Ny5cycoKIh63Q9C6D8OJ8ZDCEmK6upqLo/r169LSUnt37+/oqKiurr66dOnKirivsUWIYQQQqiNeHh46OnpjRo16tmzZ+1dFoSQpMC4FYSQpBg9erSMjExkZGRycvLatWvfv38/bNiwmTNnTps2LSsry9nZ+QsYKIsQQgghhBD68mDfCkJIUsTHxwMAm80uKSm5fv06AIwbNw4A+vbtm56evmHDhoaGhvYtIUIIIYQQQggJwjFBCCHJFRMTc/ny5Xnz5t26dSstLc3AwKC9S4QQQgghhBBC/LBvBSEkuerq6jw8PIyMjNauXWtqakrCWBBCCCGEEEJIouCYIISQ5Bo5cqSzs3NcXFxVVRXgGyIQQgghhBBCEgn7VhBCkquoqMjFxSUgIIDD4Rw7diwyMrK9S4QQQgghhBBC/BhcLvfzZYbv+EAItbbPeRLjQ85pjQnm7VUAhCSclEMatOuX9FOQL3gHLTxCEqXDfZs6xD1Lq7SnhNe0df9mJLyy4uhAX6L/JpxvBSGEEEIIIYQQQqjlPmvcCkIIUVrr6UG7x60ghOh10CuNDvekHSGJ1eG+TR3i9x3jVppLwisrjg70JfpvwvlWEEIIIYSE+wKuxRFCCCH0GWDfCkIItRA+PUAIIYQQQggB9q0ghBBCCAnCzlOEEEIIiQ/nW0EItY8vYL4VhBBCCCGEEAJ8TxBCCCGEEEIIIYTQp8C+FYQQQgghhBBCCKGWwzFBCCGEEEIIIYQQQi2HcSsIIYQQQgghhBBCLYd9KwghhBBCCCGEEEIth30rCCGEEEIIIYQQQi2HfSsIIYQQQgghhBBCLYd9KwghhBBCCCGEEEIth30rCCGEEEIIIYQQQi2HfSsIIYQQQgghhBBCLYd9KwghhBBCCCGEEEIth30rCCGEEEIIIYQQQi2HfSsIIYQQQgghhBBCLYd9KwghhBBCCCGEEEIth30rCCGEEEIIIYQQQi2HfSsIIYQQQgghhBBCLYd9KwghhBBCCCGEEEIth30rCCGEEEIIIYQQQi2HfSsIIYQQQgghhBBCLYd9KwghhBBCCCGEEEIth30rCCGEEEIIIYQQQi2HfSsIIYQQQgghhBBCLYd9KwghhBBCCCGEEEIth30rCCGEEEIIIYQQQi0n094FQAghhBBCCKFmYzAYnzlHLpf7mXNECHUUGLeCEEIIIYQQQggh1HIYt4IQQgghcbXiU2J8/IsQahWf52Ty+WNkEEIdC8atIIQQQgghhBBCCLUcxq2gDq/FjxHwkSlCCCGEEEIIoU+HcSsIIYQQQqijsra2ZjAYDAZj+/bt1MIRI0aQhUuWLGlWaikpKWTH0tJS+i21tLQYDMa5c+daUuj24O7uzmAwgoKChK6VhOqI3/gSy8XFhVRBUVGxe/fuM2bMSE1NFbWxnZ3d0aNHPz3T1kpHqKVLlzIYjEuXLvEudHFxEfWHxOVyu3Tp8ubNG/rtDQwMqDIvWrTIyspKUVHRyMho2bJlbDab2uzmzZuurq4qKiqKioqOjo4XL14UVc6goCBTU1N5eXkdHZ3JkycXFxcL3YwcoNOnT1NLYmNjGQyGnZ2dOJnyVUT84jVLmx5Q1KawbwV9IfKao70LixBCCKFWtmfPnoaGBgB49erV1atX27s4EqS+vr69i/AfsnLlyvr6+pKSklOnTtXX1/fp0+evv/4S3IzD4Sxbtsze3v4Ts+NwOJ+YAo36+vqIiAhNTc3Q0FAxd3n06JGqqqqFhYX4uVRXV+/fvz8nJ+fkyZMnT55cu3YtWX7mzBkPD4/hw4enpKRkZWV5e3uPHz/+0KFDQhMZM2bMjRs3SktLExISqqurZ82aJSo7S0vLgwcPUv89cOBAjx49qP+Kn2mziof+I7BvBSGEOhh8SEtP8NksX8lpKsK7imYv+se/NCShDb+AJ8MICdLQ0Hj37t2ZM2cAYMeOHQCgqanJu0FJScncuXO7deumpKRkZWW1ceNGFotFVqWlpbm6uiopKfXt2zchIYF3r9LS0sDAQDMzMyUlpd69e4eHh4tTmJCQEHNzcwUFBVVVVRsbm8OHDwMA+d7dunULAB4/fkz+W1dXB/+cGTZu3Ojq6qqqqtqvX7/79++TpGhW0VSK7LVp0yZHR0c5Obn4+Hhra+srV64AwMaNGxkMhrW1NX0VRFU8LS3tq6++IgvDwsL4Tiai9qLK4+bmpqqqam1tTdWCpvEBYOHChQwG45tvvhGn2YmHDx8yGAw1NbXa2loAWLBgAYPBmDx5MgDcvn3b3t5eWVm5R48e5KCQjUk5Bde2DIPBkJGRUVJS6tOnT3h4+MCBA5cuXUpWWVparl692snJycLC4u7du1u3bn306NHOnTsHDhzIW34VFZWPHz8CQGVl5Zw5cwwNDbW1tSdMmEC1M186vLkL3WXnzp0mJibq6updunQJDg4mWwpdyCc6Ovrjx4+hoaEXLlwoLCwUp/rnz5//+uuvm9VioaGhTk5OmpqagwYNmjJlyr179wCgoaFhwYIFgYGBK1as0NfX19HRCQgIWLVq1Q8//FBTUyOYiIODQ9euXVVUVAwMDLp06UITLjRu3LjHjx+/e/cOAIqLi+Pi4iZMmEBWiZ9ps4pXUFDg5eWlo6NjYGAQEBBAbWBpablu3TpXV9eePXva2dm9fPmSd6/du3cPGDCA+u/9+/dVVVXJHwaSWK3Qt0Jd5VM8PT1bnJqY152ScHkqySShffDyHaG2hg9pxWRnZ+fg4KChofEpW4qfCGotvNHXfJ/Fj7pvFYmJicrKyvTbYBR3+/L395eTk9u5c2dxcfGxY8f69evHG+TPYrEGDx7822+/SUtLe3t7FxcXBwUF+fn5AUB9ff3IkSPj4+N1dHQsLS3nz59P7cVms52cnPbu3aurq7ts2bLy8nJ/f/8mH+AnJCQsWbKkurp65cqVAQEBxsbG6enp1FqaSeK2bdvm4uIyYsSIp0+furu78w5qELqKplJEUFCQgoKCp6ensrKyj4+PsbExANjb2y9YsMDHx4emCqIqzmazPTw84uLiOnfubGtru3jxYnH2ojb46aefrKysrKysXr165evrS9/4n6iysvLs2bNsNvv48eNkSV5enoeHh4KCwrlz5+zt7X19fePj46nt6dd+ikmTJt27d48KHTp37ty5c+fevHnj4uJClkyePPnJkyfUSSwiIsLT01NJSYnsW1RU9OzZs3fv3qmpqfEeXMF0qOz4dklNTV2zZs3169crKiqSk5NHjhwJAEIXCgoNDfX09Pzmm2/09fXF7FhsQd8Krzt37vTr1w8AkpKS8vLy+P5QJ0+eXFFRIdgHRxw+fFhPT09FReXQoUMrV64UlYWCgsKECRPCwsIAIDw8/JtvvlFVVSWrxM+0WcUbN26crKxsRkbG48ePnz59yvvFiYmJiY6OfvXqlZeX17x583j38vHxefHixatXr8h/Dx069P3335M/DCS5uJ+sZ8+eAKCrq+vwjxUrVrQ4NfKQITo6mn6zESNGODg43Lp1q8UZ8WGz2a2V1GdItkliNmObSk5OJn9jJSUlbZoRyaUFY4LatFStpV+/fkeOHPn0dHr37n3ixIlPT6cFnj17pqSkRL+NqGpaWFg068+4uds3qzCSg5x1yU3+qVOnuFzu1KlT4Z+HtD/88APZrLi4eM6cOV27dlVUVOzRo0dwcHBdXR1ZlZqa6uLioqioSJ6q8X5VS0pKAgICunXrpqioaGNjExYWRuVLc2IpKyv74YcfunfvrqCgoKur++OPP9IXQE1NDQC2bNni4OCgoKAwfPjw3NzcgIAAFRUVQ0PDiIgI3hyDg4NdXFxUVFT69u177949smrq1KlGRkZMJlNOTs7S0nLv3r28jUPp2bOnYMnJf7dv3+7q6spkMnmT5d1S6F7R0dGCWZAnovPmzSNbxsTEAICenh6Hw+FrKN40aZo6NTXV1dWVLKcCjMU8QBs3bhw6dKiKikrPnj2petEc8QULFgDA119/LfIP7t8ELyQGDBjA5XL9/Pz4PjQ2NlZUVMTFxfE+eWvWFciQIUNWrVol9DOJuq+urn769Kmvr6+iouKjR4/ErEILfMp5DLU18pXctm0buV0nd4lHjx4dMWIEdUo8e/YsAMjIyBQWFnK5XGpahPz8fNIxLSUl9e7dOy6Xy/uNI3sxmcysrKySkhIyjsDU1JTkK+qUeO3aNQAwNjaOiIh48ODBx48fyXKS7J07d7hcLjVIpLa2lkpqy5YtXC63sbHRxMQEAPbs2UO/iqZSZK+pU6fyFow0CPU94sNbHVEVJ5EvUlJSubm5XC73yJEjvCcTmuYiia9Zs4a37uXl5TSNT0r1+PHjI0eOxMfHizr6gieTBw8eAICjo6Obm1tUVJS+vr6pqemkSZN27twJAOREUVVVJSsr6+fnRzYOCwsTurbJvATxnqkIUsfS0lIul2thYUH9WnF5Thpjx45duXIll8utq6vr1KkTqW9mZiaDwSguLiYbf/jwgcFgVFRU0KQjdJenT58ymczIyMjKykpql4yMDMGFfEhqcXFxXC53zZo13bp1a2xsFFVNIi0tTVdXt6GhQbBZmEymJg8pKSnBE+aaNWtMTU1JW12/fh0EbiLIVCwnT54UWuCampr8/Pzbt2/PmjXr6dOnQrcZMmTIhg0bnj17Zmxs3NDQQAJ/du7c2a9fP3EypSoufvFSUlIAgDooN27cUFBQIE1kYWERGhpKlr9+/ZrJZJLP1AH18fFZvHgxl8utrq5WUVF58OCB0EohydFqY4I8PT0f/mPTpk3R0dEMBkNTUzM/P7+hocHR0ZHBYCxcuJBsTB/fSJk2bZqxsbGioqK8vHyPHj327dtHrXr8+HFCQkJZWRnQxhmCGNGJVLQkX+40hWwyqpMvWZqkaCJUhQaUtiw2teMGdn55mjsstlXG4qJmaVabt+k4Z3qS85CWzWY7OzuHhISUlpZOmDBh8ODBr1+/bvKBKgCsXr1aXV1dXl7+6tWr1tbWMTExVlZWOTk5M2bMKCgooDYT+sD2xYsXLi4uK1as8PPzS0tLCwgIuHHjBgCI/2x21apVWlpaFhYWgo+I6QlmQZ41HTt2jPwQkFEJkyZNkpaWpmk0UU1N82RYnAP0OZ8M09u+ffv8+fMNDQ0vX76sra3duonTRN3zsrS0DAkJcXJyYjKZjo6OeXl5W7Zs0dXV1dbWJsNGQHS0dk5OzogRI9TV1a2trUmMOiEqSp8iTqQ9agvkyxIbG6uvrz9+/HjeVST+X0tLS1dXFwCoETE5OTk5OTkAoKmpaWhoCAB9+/bl26u2ttbExERbW3vatGkAkJWVRb7pori6uvr6+ubl5fn6+g4cOFBTU5P6Y6MI/e0wNTUFAAaDQTpQSO40q2gqRT64ubnRlJOGqIqTABwNDQ19fX3eHOn3oprLxsYGAEjfOgBUV1fTND7Rr1+/yZMnDxkypLlVIIEnmzdv9vHxkZKSAoD8/HwAIGdvZWVlTU1N3gn46Nd+itzcXBkZGSoyokuXLoLb+Pn5HTlypLGx8fz581Tuu9AAACAASURBVOrq6s7OzgCQlZXFYDD69+9vYmJiYmJiY2OjpqZGlUpoOkJ3kZeXP3z48K+//qqnp+fs7EzuX7p27Sq4kE9oaKiJiQlpfD8/v8zMzLi4OPrKnjt3bsyYMaTB+cyYMSORh46ODt8GGzZsOHr0aFxcHOmJ09LSAgC+o0AOE/lBCQ8Pp8ZMkLVMJrNz585OTk6TJk0aNWpUY2OjqHLa2tpqamquXbuWwWA4OjpSy5vMtAVb5ubmKisrUwu7du1aV1dH/XCQdEjha2tr+c4MM2bMOHr0aH19/enTpw0NDUU9qECSo9X6VqKiogb8Iyoq6ttvvw0ICCgrK5s6deqWLVvu379vZ2e3detW3l1oQh8JUZfOQgleTYIYl6G80ZJCk6UpJE1Up9BkBZOiufcQGlDa4tuejh7Y2V527NhhYWGhoqJiYmKyadMmarmlpeWKFStcXV379+/v6OiYlJTU5HLeYbFCr+Nzc3O1tbXJLRkAjB07dtKkSQBAxuJS6TR5kyC0zIGBga9fvw4MDDQxMSEHq8l7AzGza/V7EkHJyckODg5qamqDBg2impQmnbdv3w4ZMkRRUbFfv37UyFXBZqEZ3szb5kIrSDPO+XPS09Pz9vZ++PChv78/i8VatGgR79rY2NjU1FQZGZk7d+4cOHAgIiICAE6ePFlQUBAfH5+eni4lJXX79u3jx4//8ssv1F4xMTEpKSlMJvP48ePz5s1bt24dAGzevJm+JLGxsUlJSQwG4969e2FhYVFRUVFRUTQFoHZctGjR5cuXScdEXV3do0ePbt++LS8vz2azX7x4QW22YsWKtWvXRkZGmpiYVFVVRUZGAsCdO3ciIiJWrVq1Y8cODw8PACDPipcvX25paQkAw4YN++WXX5YvXy6q2EuXLo2MjExISOjcuTOVrDgEsxgwYEDfvn0rKirOnj3b0NBw/vx5AODrSOJD09Tx8fEZGRlSUlK3bt2KiIjYs2ePOHvxFm/37t179+4FgPT09IqKCpojDgA+Pj5Hjhzh+/tpFa9fvz58+PDUqVPV1NQmTpzY6unz4ou653Xw4MHff/+9pKRETU3NxcWlqqoqMzPzzJkzP/74Y0ZGBoiO1vby8jIwMCgsLIyNjf3999958xIVpQ9iR9qjtmBjYzN06FAACAwMlJWV5V1Fbt1LS0vJhRwVZm9gYGBgYAAA5eXlHz58AIC3b9/y7aWkpHT58uVr/7hw4YKMjAxNMaSkpMLDwysrKx8/frx+/fq6ujryVSXB/OXl5QDAe4qjPH/+HABYLBaJ/CW506yiqRT5wGQyedMnvb00N5xNVpx08ZSVlZGpN/imh2iyuUgBeK+faRqfePLkydGjR4Xe+TdZhSFDhjx58oT6hpK8srKyAKC6uvr9+/e8LUy/9lOcOHFi8ODB1B+k0NuH0aNH19bW3rx5MyIiwtfXl2xjZGTEYDCeP3+e9Y/y8nIrKyuadETt4unpGRcXV1pa6uHh8e2333K5XAAQupDC4XDCw8Pz8vK0tbW1tLT69+8PAE3edNAMCCIzoVD4nj38+OOPR44cuXXrFuneAgBra2t9fX0qNoo4evSouro66WLgDSwSzK6goIB+apLp06dv3LiR9ABSmsy0BVsaGBhUV1dTV6pv376Vl5enulToOTk5aWhoXLx48dChQ3xFRRLq00Nf+KKj4Z9Ixbq6OtL9zGAwVFVV09PTqV2aDH0kEYkkTrKhoaG6unr06NEAsHz5ct4UeMO2BeMMuWJEJ/JFS/KiKSTJgiaqky9ZUUnRBHMKDSilj00VzLHDBXa2DMmlLcYEnTp1KjMzs7Gx8enTp7q6ulFRUWS5hYVFjx49yN9YRESEkZERGelAs9za2pqEOHK53IEDB3p7e1dWVhYUFAwcOHDWrFlkeUxMTKdOnTIzM0NCQszMzEiUJm+QOUk/KSmpqqrKw8PDzMxs5cqV1dXVt27dkpWVJV8xUWXmGxM0evTob7/9tqSkpKamZubMmaNGjRKsvjjZiaqLg4PD1KlTa2trs7Ozra2tqVh6UfnSjAkyNjZ+9eoVm81et24d1aSi0rGwsLCwsHj27FlNTc3cuXOdnJxEHcri4mJZWdk3b96QDWbPnk39SPMWRmgF+Q7o50cFwJOrbQDQ19dns9m8AfDk/llPT4/sQq4XASAhIYF8nbW1tcmqxMRE6qvKd9dNMBgM3vObYAA82YtKkHeh0AJw/xkTFBkZyeVySVedlZUV2ZLkcvr0aeozGffE5XLJqPKlS5c2NDQsX75cX1+f9+EYdQQF496Fju6honbJQ7mlS5dyxRsTJDSLAwcOAMDQoUNJnDCJLhZEJULT1OQAaWlpkV2ePXvWrANEvvXU0P2cnByaIy60kPQECyBqTBC5pSH5Cj63h08bE0QTdc/LwsJi37595PPRo0fV1NSoYHVra+uoqChR0dpkyrCysjKy/PDhw+Q8RhOlT04d4kTao9ZFnRK5XG52dva1a9eqqqq4/3xPySmxtrbWzMwMACwsLKZNm0ZubLy8vLhcLovF6tq1KwB89dVXW7dupW7tSkpK6urqunfvDgBOTk4bNmxYunTpsGHD3N3dSb6iTol37961sbEJDAzctGkTmYKwW7duXC6XBCPY2NgsX76cit3g/fLKy8tPmDCBXDwrKyuTi0OaVTSVElq2OXPmAICRkVFgYOCxY8f4is27i6iKs9lskqO5ufmUKVOoWpCTCU1z8SaelpZGnZ1oGp+UqslBi4InEzLM59KlS1euXCHjKbp16zZp0qT8/HwlJaXBgwdfvnyZDOS8desWNSZI6Nom8xJEjVj8+PHjs2fPpk6dymQyyQ8fV2DkMu/1xvz584cNGyYrK5uZmUlt4OHh4enpmZ+fz+Vyi4uLyS8jfTqCuyQnJ9+8eZNcPoWGhuro6DQ2NgpdyFuR6OhoaWnpJ0+e5PwjNDRUXl6enGaFjgkqLi5WUVEhf9KCzcK3PemYIJ9JmHxmZmZtbW1tbS2LxSLLo6Ki5OTkNm3alJeXV1xcvHfvXiaTefDgQcH02Wz25s2bU1JSKioqHj16ZGdnN3jwYMHNuP+MCeJyudXV1deuXfvw4QOXy6XGBDWZKW9FxC/egAEDJk+eXFVVVVhY6OjoOHPmTLKc9ziSjsX6+nruvw9oSEiIra2tnJwc9buDJFmrxa1Q48y5XG5AQAD5GSCvq+Byud7e3t26dePbhSb0EQAaGxvXrVtnYGAgKyurrKxM+h2KiopEFUAwzhDEiE5sMlqSvpAgIqpTaLKCSdEEcwoNKG1ZbOoXENjZXsaPH29qaspgMPr06ePr68sbNjVr1ix1dXUAmDJlSkNDAzWmTNTy2bNnkwuLN2/ePHjwYNeuXSoqKnp6esHBwREREeQh0siRI6dPnz5q1Ki1a9eeOnVKRUVFsEgBAQE9e/ZUVlaeNGlSSUnJhg0blJSUnJ2dLSwsyP0STZkpb9++jYmJ2b9/v5aWFpPJ3LZtW2xsLHlk1KzsRNXlzZs3jx492r59u4KCgpGR0bJly5qbL685c+ZYWVnJysoGBQXV1dXFx8fTpxMYGGhra8tkMufMmfP48WNRh1JbW9vDw4MEU7BYrFOnTglGGdAcLOqAti8JeUhrZGRE8qLu5+vr65t8oAoApMzk4Rtf+XkJPrA9e/bszz//XFlZefTo0Tt37pA+F+4/F75iPpt9/fo1KSopdrMeUQpmMXHixE6dOt28eTMkJASaCloB2qYmq4Q+GRbnADX3yXCLHwuLicT4CP56ti6+qHteVOS8kpKSnp4e1R+npKRUVVUlKlo7NzdXU1OzU6dOZLm5uTn5QB+lD+JF2qO2Y2Rk5ObmJhiMrKCgcPfu3ZkzZ7JYrOPHj2tqaq5fv56MtpaTk4uNjXV2dn706NGxY8d4w3jl5eXv3LkTEBCQn58fHBwcFhbGYrG+//57+jLo6upqaWmdPHlyzZo1N27ccHNzI2Fx+/bt69+/f0ZGRlxc3Pr16wV33LBhQ35+/uvXr21tbWNjY8nFIc0qmkoJtXjxYgcHh6Kioj179tBPfy6q4rKysrGxsS4uLnl5eYmJiVTQHDkFtaC5aBr/0w0fPpyckInOnTtfvnyZzWaPGzfuyZMnR44ccXZ2Jpfx8vLyQte2LN9NmzbJyspqamp6enoyGIxnz56JM9DY39//2rVrjo6O5H6BOHHiRJcuXQYNGqSiojJgwABxQmUFd6mtrf3xxx91dHQ0NDQOHDgQFRVFeuQFF/KmExoaOm7cuL59+1KRJn5+frq6ujR/YBcuXHBzc1NQUGiykLwqKir27NmTnp7etWtXJpPJZDKp2Jxx48bFxsZeuXLF3Nzc2Nj42LFjp06dIhPM8WEwGA8ePBgyZIiOjg4pNhUJLoqSkhKZ3IBveZOZUg0lfvGioqJqampMTU379OljY2Mj9GGDKL6+vsnJyWPGjGn1obWoLdBdLn+i9+/fk3sqBQWFgwcPTpkyZdCgQbwbPH/+fPz48UJDHwGAXDqrqKgcPXrU0NBw9erVJN5BVHaCV5PAcxl65swZKvaMxWJRl6F80ZKChBZSSUnp48ePNFGdQpMVTIp0apB7Dx0dHd57DxJQ+ttvv71+/TomJmbt2rXr1q0jM57QVEcoUY1Ari/J5buenp6owE5RedFfvqupqQkN7ExOTiZRmjQFlhznzp3bsWNHdnY2AFRWVvJ2menp6fF+pi6sRS2nLu5FXceTcafz5s3bsWPH6NGjBXum+NIRepNAX2YKdW9ALSH3BlRPmZjZtfiepMl8eZG+SACQlpY2NDTMy8uTk5OjSYd8rQBAWVmZjFyVkZER2ix+fn4LFizYsGED7/BmXjRDZIWOc24Xhw4dSk1NFRyCO3LkSDMzs/T0dGdn58GDB5NRKl5eXl26dNHS0uratWtmZuZ3333n7u7OO5XVqFGjunfvnpqaunHjxuHDh1dWViYmJkpLS48aNYqmDCNHjuzRo0dycvKgQYPGjh1LHjodP35cVAGaW8eQkJDMzMzU1NSioiJlZeXx48dTQ75ZLNaTJ08ePnzIuz158nns2LHq6uoBAwaIGo2ydevWtLS0tLS0goICkqz4RRLMgslk+vn57dy589KlS3Jyck0OgaFpahcXF6rpBg4cSJquyb1o8nJxcRF1xAHgyJEju3bt+vrrr1v9zGxlZeXj47N69eoPHz6cOHGidRPnwxd1Lz4qWps886eitfX19SsqKlgslry8PABQLx+lQu5pTlyenp6enp61tbW//PLLt99++/79e5pBxOjTUcNF+Vy+fJn3v7q6uvv37xe6pYWFBW8vGO+gZh0dHd5xebxEDWs1MzMT+mDD2tqaGnAKACQig5e5ubmozjhRq0RVSmjZzMzM+M6WNLuIqri2tjZ1BiYjRDQ0NKivg6i9eBM3MzPjvZ6naXwA+OWXX4TG69EgcXO8S6j3NA0ePJh3NsC3b9/++uuvANCrVy/BtS1D/3YhEihHoZ4AAYCtra3gbY6amtquXbt27dolfjpCdxGsV58+fegrSyZl5yUjI0Ouo0BENc+fPz9u3DihqQlun5ubSz6oq6vT3N8NHTqUPEOiJyMjw/tbSUNoyRcuXEjNCkqfaUVFBe9wHjGLp6+vL7Svh/c4mpiYUO3Ad0DV1dVxQFBH0Sbzrfj7+wOAn59fbm6uv7//77//zuFwvL29SWcEJSQkxNvbe9CgQdTlMu9aKh5E6KWzmMhl6MePHzdu3Pjw4cOrV69u3bp179699J0RTRaSvBssKCjoxx9/pJ7JtyApcvPD4XCcnZ2nT59OJsIg9x7379/v3bv38uXLr169Sno9dHR0WlYdUXu5ubmZmZk1NjY6Ozv7+voGBgZ+YtORy3cOh/Pdd99t27aN9yRFHDlyxMfHh8zELvkKCgrGjx+/du3arKys7Ozs2bNn8576ySh9AOBwONnZ2dT9vKjl1IU1zajLhoYGX1/f8ePH37lzR8xfCPHLzDt0gn4Er/hE1YW6JyHLBe9JmpUv9UPe0NCQk5Ojr6/f3HRENYvQ4c3iVBBop1v6zCThIS3Za9GiRRoaGsePH79586aFhUVzH6jSEHxg6+XlNXfuXBkZmWXLliUkJLi7u/NuL+az2Z9++ikvLy8pKUnwEXGThGYxd+5c8ocxZsyYJl/VTNPUNE+GW3aA2vTJMI0lS5bs27evoKDA3d2dJuy0ZbhcLofDqampSUxMnDZt2t27d3/++ecWpGNhYTFgwIBFixZVV1cXFRWtXr3a19dXSkrK0tKyT58+5AeLxWJRNyrdunUbPnz49OnTycxBJSUlUVFRvAmmpKTExcWxWCwmk6mtrd2C7h6EJNn69euHDx++bt26efPmkSu9xYsXS85vYrNERkbGxcVt27aNL3YbtYyjo+OYMWPauxRthc1m37x5Mykp6TNPKHvgwAE1NTUywhF1AJ8+rEhwvpV+/fqRYKeuXbuSIa9eXl7AM1qSxNJv3brV2dlZQUHB1tb29u3bvKuio6M5HM7cuXM7deqkra3t7e1NXi7j6+vLtxlX9BhOsmVRURF5XSW5KXJ2diZvrGzyLcU0hXz58mX//v2VlJTs7e2p3nSa+QhokiosLJw5c6aJiQmTybSwsFi/fj1Jh7zBR1tbW0ZGplOnTm5ubk+ePKGpjtDCU8UQtRf1Vk4bGxvScw//TFUjZtPxNXhKSoqzs7OysnLv3r2pK1HxB822DMml1edbefPmjaysLBncmJeXZ2RkNG7cOLLKwsLCyMgoKSmpvr5+zZo1hoaG1Lwqopbz/kmIGnUZFBTUp0+furq66OhoDQ0NMm0N33wrVDrR0dEWFhZUmg4ODmFhYTRl9vDwWLduHbW9qBG8vJrMjqYu/fv337x5M5fLraurc3FxoeZbEZUvzXwrpqamycnJ9fX1GzZsoJpUnBHI1MhVmmYROryZtzBCK9gqb3pG4pCE18k3CxkseeHChU9MhzoPc7ncP/74AwA0NDT4BsO3l1a6+mj6JMylnW+FpKCgoNCtW7dp06alpKQITUGc81hubu53332npaXVuXPnOXPmVFdXkw2ysrLc3Nzs7Ozc3Nx27NhBnccqKirmz59vYmKirKzctWvXBQsWkOXk1PH06VN7e3tVVdVOnTo5ODhQP/cI0aA510naaTA8PNzMzExBQUFDQ8PBweHo0aPte2oS82TS4fJCkmbChAnk1W+fM1N1dXV9fX3BqX+QxGqfE4Sk/U4I1YqFlNj6SvLlu/jaqG+Fy+UGBwebm5sPGjTou+++CwgI4O1b2bx5s729vYqKioODQ2JiYpPLeY++0Ov469evq6mppaWlkW0CAgIcHR3r6+ub1bdCU+b4+Hhzc3N1dXWyRNS9Aa/Pf08itAybNm0SbFJR6YiaFUxUs5BZQl1cXHgz5S2M0Api38pnI7EnT0HR0dE//PADg8GwtLSkZkttsYULFw4bNmzt2rVz585VVFQEgM98PUfjc/atIIQQPexbQQhJDga3Va+TxKSlpfX+/fvo6GgSjSKZWrGQElvfRYsWvXr1atCgQSUlJeHh4TU1NcHBwatWrWrvcjUPiUTN+/cb5umRoTot/uO3tLTcsmWL4NEUtRwh1DISe/IU5O7ufv369V69ekVERJDolU8RERERHBycm5urqKhobm4eGBg4ceJECYm6b8VitMsVCELoS0LOSJ/nZPI580IIdURtOJctkny2trYXL168c+eOoqJir169yOV7exeqhaiZTRBCXwxRU0VKIL5ZMz+Rr68vmYELIYQQQgh1CO3Tt9IhLpdbsZASW1+8fEcIIYQQQgghhD5R+4wJQgghhFBHhGOCEEKS4/MPlsQTF0JIlFZ7BzNCCCGEEEIIIYTQfxDOt4IQQgghhBDqeDCKBCEkOXBMEEIIIYQQQgghhFDL4ZgghBBCCCGEEEIIoZbDvhWEEEIIIYQQQgihlsO+FYQQQgghhBBCCKGWw74VhBBCCCGEEEIIoZbD9wShDo/BYLRsR5zIGSGEEEKo42rxRWCL4dUjQkgU7FtBCCGEkLha604G708QQggh9CXBvhX0hYiLixN/Y1dX17YrCUIIIYQQ+mw+T1/t54+RQQh1LDjfCkIIIYQQQgghhFDLYd8KQiI9f/7cxsZGW1v766+/bu6+tra2J0+ebItS2dnZHT16tC1SFqrtKoIQQgghhBBCXwYcE4SQSAcPHuzdu/eLFy8+T3YcDkdGpumv5LJly2xtbT9DeQCAw+F8nowQQgghhBBCqOPCuBWEhJs2bVpYWNiff/5pYmJy6NAhAKisrJwzZ46hoaG2tvaECRNKS0sBYMeOHRYWFioqKiYmJps2bSL7BgYGvn79OjAw0MTExNfXFwD09PTu3r1L1j5+/FhZWZl8trS0XL16tZOTk4WFxd27d4VmwWfr1q2PHj2idg8JCXFycmIymY6Ojnl5eVu2bNHV1dXW1t6xYwe1zYoVK1xdXfv37+/o6JiUlESWFxQUeHl56ejoGBgYBAQE1NTUCBapT58+fBURWl+y17p161xdXXv27GlnZ/fy5UuyvKKiYu7cucbGxqqqqv369Xvz5o2olkQIIYQQQnxaPWDZxcWFwWAwGAxFRcXu3bvPmDEjNTW1rXNv07DrpUuXMhiMS5cu8S50cXEJCgpqoxwBwNLS8ty5c22XPuqIsG8FIeEOHjz4/fffz5gxIysra+rUqQAwadKkoqKiZ8+evXv3Tk1Nzc/PDwAMDAwuX75cWVkZHR29e/fuM2fOAMCePXusrKz27NmTlZUVERFBn9G5c+fOnTv35s0bFxcXoVk0Wc7ff/+9pKRETU3NxcWlqqoqMzPzzJkzP/74Y0ZGBtnm/Pnz0dHRf/3116xZs0aNGsVisQBg3LhxsrKyGRkZjx8/fvr06eLFiwWL9PLlS76KCK0vERMTEx0d/erVKy8vr3nz5pGF3t7e2dnZ9+/fr6ioCA8PV1FREdWSCCHEh29M4mceEdkkySlPYmIi1WUviuSUFqHP5uHDhwwGQ0pKSl1d3dXV9eHDh9QqKysrBoMRFhbGu2V4eHhrZd2Kkb/Lli2zt7dv3fRXrlxZX19fUlJy6tSp+vr6Pn36/PXXX4KbcTgcobk3V5vGQdfX10dERGhqaoaGhrZdLrw+pTqt1RSfmA5GprcR7FtBSCxv376NiYnZv3+/lpYWk8nctm1bbGzshw8fxo8fb2pqymAw+vTp4+vre+PGjeamPHv2bE1NTZos6HcPCAjo2bOnsrLypEmTSkpKNmzYoKSk5OzsbGFhkZiYSLaZNWuWuro6AEyZMqWhoSE+Pv7NmzcPHjzYtWuXioqKnp5ecHBwREREY2MjX5EE0dSXymXs2LGPHz8GgMzMzMuXL4eGhurr60tJSfXq1atLly4tqyZCSGJVV1dzeVy/fl1KSmr//v0VFRXV1dVPnz4lnaqfrlUu8VsFuSqVnPIghGgsWbJk9+7dOTk57u7uJSUlAJCcnJycnMxgMKKioloxIzGDkXNyckaMGKGurm5tbb13716qY1RUjDMVsCxm+jt37jQxMVFXV+/SpUtwcLDQojIYDBkZGSUlpT59+oSHhw8cOHDp0qVCa0HlvnPnzoEDB1IpPHz4UEVF5ePHj6KKwZcOb+7il1ycukRHR3/8+DE0NPTChQuFhYX0x+jUqVO9e/cmn+fPny8rK1tVVQUAx48fp87nQiO7aapTUVHh4uLi6+tbX18vtGo0+4qqozjptOIRQa0F+1YQEktWVhaDwejfv7+JiYmJiYmNjY2amlpeXt65c+ecnZ2NjY2NjY3/+OMP8pvdLF26dKHPQszdlZSU9PT0pKSkqP+SXwsA0NPTo7bX09PLy8vLzc1VVlbW1tYmC7t27VpXV0edfKk0BdHUV0tLi3xgMpm1tbUcDicrK0tRUZEvtZZVEyEksUaPHj1s2LDy8vL79+8PGzZs6dKlw4YNmzlz5uLFi8eOHfvnn3+27MWlgoMrmzsiUpzhhy2+qOUtT4vvcMSphajxm7y3Z/fu3aPSbHGtEfpSWVlZTZky5dChQx8+fDh+/DgAkC6V8ePHX79+vXWf7ogTjOzl5WViYlJYWHj58uUDBw60bvqpqalr1qy5fv16RUVFcnLyyJEjxUl20qRJ9+7dq6+vF8yF2mby5MlPnjyhRg9FRER4enoqKSnRBCMLTQdExC8LllzMuoSGhnp6en7zzTf6+vpNRh4NHTo0KSmJXL7euHHDxMTk1q1b5PPQoUPJNqIiu4VWJzs729HR0dHRMSIiQlZWVlRriGoKUXUUJ51WPCKotWDfCkJiMTIyYjAYz58/z/pHeXl5p06dxo8fv3bt2qysrOzs7NmzZ3O5XLI91cdBKCsr19XVkc98V7rUXYfQLKysrD698NTgIA6Hk52dra+vb2BgUF1dTZXk7du38vLyVOcI740Qb0UKCgpE1VcoY2PjmpqagoIC3oVtV02EULuIj4+/fv06m80uKSm5fv36s2fPyOPWvn37cjicDRs2VFZWtiDZJgdXNjkissnhh59yUcubzqfc4TRZC1FX+V5eXgYGBoWFhbGxsb///jt9YcSpNUJfth49egBAdnY2AJw5c6ZHjx7+/v5sNvvPP/9sxVyaDEZ+8+bNX3/9tW3bNgUFBQMDg+XLl7du+jIyMg0NDc+ePauqqlJTU+vbt684yRoYGHA4HOpcLTR+WVtb28PDg5yNWSzWqVOn/Pz86IORhaYjfsnFqcvbt29v3Ljh7+/PYDB8fX0PHDhAf2mqpaXVq1evmzdvFhYWlpeXz5kzh4RgU30rNJHdgtV5+vTp4MGDFyxYsHHjRpqqiWoKABBaRzHTaa0jgloR9q0gJJZu3boNHz58+vTppKegpKQkKiqKBIbY2NgwGIz8/HzyMITQ09Mj87YSdnZ2Fy5cAIC6urpdu3aJn0WrFH7//v2vXr0iNzlMv5IruwAAIABJREFUJtPFxcXCwmLAgAGLFi2qrq4uKipavXq1r68vX3+QYEVo6ktTo5kzZxYUFHC53KSkpIKCgrarJkJIQsTExFy+fHnevHm3bt1KS0szMDBoi1zoR0SKM/zwUy5qKZ94h0NfC1FX+W/evHn06NH27dsVFBSMjIyWLVtGX5gma43QFy8lJQUAjIyMMjIynj9/7ubm5uzsLC8v37oXIU0GI+fm5mpoaKiqqpLNTE1NWzf9rl27Hj58+Ndff9XT03N2diZBGU3Kzc2VkZGhSiUqftnPz+/IkSONjY3nz59XV1d3dnamD0YWmo74JRenLqGhoSYmJkOGDCHFy8zMjIuLo6+sm5vb9evXr1+//tVXX5HP6enphYWFjo6OpClERXYLVuePP/7Q19cnkZU0VaNpUqF1FD+dVjkiqBVh3wpC4jpx4kSXLl0GDRqkoqIyYMCAu3fvdu/efe3atSQUMDAwcOzYsdTGy5cvP3bsWKdOnTw9PQFgy5Ytz54969279//au/OwJq69D+BnVMRIMGwBZA2oBJULoiAIFXF5sCraWgJ1aQ3FpVKhvfq+F7XVotUqRQWtvdalanHDBate3FpL0Wpr6wbuoC+LCiKCioBANDjvH+fpPLlJZgiLC/T7efgjOTnL75xMkpnDmZnQ0FD6A2BgEy0SeXR0dFRUlIWFxY8//pienm5sbEwISUtLq6mpcXFx8fb29vT05Nafa9HsiEB/+ezcubNr166+vr4SiUSpVNLZmRfUTQB4TdTV1Y0YMcLJySk+Pt7FxSUsLOxFtCJ8RqQhpx82c6eWauYRjnAv+Pbyi4qKLC0tzc3NaXqPHj2Eg2mw1wBt2LVr17Zu3Tp58mSJRDJ+/Hg6mbJ69erOnTurVKoff/yRO4e6+RpcjGxvb//w4UOuxcLCQq6swBpnw+snhCgUiszMzPLy8hEjRowdO1Z4HQeVmpr6xhtvGBkZabWiJTQ0tLa29pdffklJSVEqlQzDCC9G1ltPoyIX7otarf7++++Li4ulUqmVlZWvry8hpMEr2g4dOjQjIyMjI2PYsGH/+Mc/7t+/v2PHjoCAAJFIRAgRWNmt251Vq1ZJpdLRo0fTszUFuiZwbqxuHw2vp0XeEWhBHV51AACvL62TYCUSyapVq7RWnXz22WefffaZbtlBgwZp3tBOJpP9+uuv3NM5c+bQB/S/KMJNaKGXidUt/vbbb7/99tvcU82L4bu7u//5559a9djb22ve5Udvnbod4euvZimZTMb9+Jmbm69fv14rsyHdBIDWa+TIkUFBQZmZmfT4ocn3Wde7mM5A3P6lRCIRyKZQKBQKRW1t7cqVK8eOHfvgwQOBgsLHCbr5dStv7H4tt5dP9+y5vXx7e/uKigqVSkXnyrnLNza519jhhjZs+fLlYrHY29t7y5YtNjY2e/fudXV1pbtAv//++4wZMw4ePNjY9SMN4lbpfv311127di0rKztx4oRCoXB3d+/Xr9/s2bOTk5PLy8sTExO5InSN87BhwwTWODdYf05OTklJCZ0pkEql3HSJFpZl1Wr106dPb9y4sXr16lOnTh0/frzBThkZGU2YMCExMfH48ePffPONQBgtEnlOTs69e/cE+nLw4MH79++fOXPG2tqaphw9ejQmJubBgwcCZ74EBQUVFRX98MMPX3zxBcMwgwcPXrFiBbf6j1vZ/e233z558kRgZTchxNjY+Icffpg4ceKIESMOHjzYhNHQ+34ZXk+LvCPQgrBuBQAAAFpSaWlpcHDw3r174+Pjt2/fvnv37qbVo3VyZaMYcvphTk5OZmamSqXSu1Nr4HmLfPn1Vt5YfOdvuru7e3t7JycnE0JUKhV3GNbkXgO0Sf7+/izLPn/+vLKy8sSJE/7+/oSQM2fO5OXl9enTp0+fPh999BHLsuPHj6c5da9P1Bx8q3T37Nlz8+ZNa2vr4cOHR0VFcfkNXOMsXH9tbe2cOXOsra0tLCy+++67tLQ0vTOnS5YsMTIysrS0VCgUDMNkZWUZeNezDz744NixY4GBgdxsVBMWIxsYeV1dnXBfNmzYEBYW1rdvX4e/REZG2tjYbNmyRaB1ExMTf39/W1tbR0dHQsjQoUMrKyuHDBnCZTBwZTdlZGSUmpoqk8mGDRv26NGjxo4G3/tleD0t8o5Ai2EBWjls/MLkcvm+fftedRQA0Ea8zN2P48eP9+jRw8zMLCwsjGXZfv36bd26lb6k+c22b98+uVzOlfLz89u8eTPLshUVFR9//LFMJhOLxa6urp988olW/RcuXOjfv3+XLl3Mzc39/Px+/fVXmq63oO53KReP3vx8lWsypBdFRUXvvPOOlZVV165do6OjuTteFxYWDhs2zMfHZ9iwYUlJSSYmJgLBa0ZrSGAArcXL3KN7EW1lZWVxH14AaO0YtuX2kwBeiSavZMbGDwDQWC118gi+gQGg+eg30sv5PnkRbWVnZ7/xxhvV1dUtWCcAvCq43gq0ethBBwAAAAAAgFcI61YAAADAUFi3AgCvj9a+bgUA2hJcyxYAAAAAAAAAoOkwtwIAAAAAAAAA0HS43goAAAAAALRWLXWuIgBAc2DdCgAAAAAAAABA0+FatgAAAAAAAAAATYdzgqDVa/JCUEwsAgAAAAAAQPPhnCAAAAAAAAAAgKbDuhVoI+Lj4w3PvHDhwhcXCQAAAAAAAPytYN0KAAAAAAAAAEDTYW4FgNfFixc9PT2lUulbb73V2LJ9+vTZuXPni4jK3d19//79Ly6/Xj4+Ptu2bWtmJQAAAAAAAG0SzgkC4LVx40YvL69Lly69nObUanWHDq/pRzIuLq5Pnz6G53+d+wIAzdLUy4drw9XEAQAAoA3BuhUA/SZPnrx58+b//Oc/Mpls06ZNhJDKysro6GhHR0epVDpu3Ljy8nJCSFJSklwuNzU1lclkS5YsoWVjY2OvXbsWGxsrk8mUSiUhxNbW9tSpU/TVc+fOicVi+tjd3X3+/PkDBw6Uy+WnTp3S24Su69ev+/n5SSSSgICAK1eu0ESBsgUFBYMGDercuXO/fv0uX75ME/VGnpycPGDAAK7gH3/8YWpq+uTJk8TExDNnztDEkpKS8PBwa2trBweHmJiYmpoavX1p3vADAAAANIB56V51jwHg9YW5FQD9Nm7c+O67706dOrWwsDAqKooQMnHixNLS0qysrNu3b0skksjISEKIg4PD0aNHKysr9+3b9/XXX+/du5cQsnr16l69eq1evbqwsDAlJUW4of379+/fvz83Nzc4OFhvE7rWrVu3efPm8vLy4cOHjxo1SqVS8YXH5V+1atWDBw/8/f1nzJhBE/VG/t57750/f/7GjRs0T0pKikKhMDEx0Ww9LCzMyMgoLy/v3LlzFy5cmDVrlt6+GD7UAAAAAAAArRoW7QMYpKCg4NChQ6WlpVZWVoSQZcuWmZmZPX78OCIigmbw9vZWKpUZGRlhYWGNqnn69OmWlpYCTUgkEq0i0dHRvXr1IoTMmzdvzZo1x48fd3NzEygbGxtLz+iJjo7u378/rURv5FKpdMSIESkpKV9++aVKpdq1a9e+ffs0m87NzT19+vSBAwdMTU1NTU0XL148atSoNWvWtGvXTrMvAAAAAC8H+1LOMcSiFQAQhrkVAIMUFhYyDOPr68ulSCSS4uLizMzMpKSkW7duEUIqKyuHDRvW2Jrt7OyEm9CdW5HJZPRB+/btHR0di4uLO3bsKFDW3t6eJorF4traWnoxlP379+uNPDIy8pNPPlm0aNGBAwfMzMyCgoI0my4qKhKLxVKplD51dXWtq6srLy+3trbW7AsAAAAAAMDfB+ZWAAzi5OTEMMzFixc1ZzpKSkoiIiKOHDkyZMgQhmHmzp178+ZN+hJdx8ERi8V1dXX0sdZVVLh/g+htQi86IUIIqa+vv3Pnjr29veFlG4w8NDR02rRpv/zyS0pKilKp1PovjYODQ3V1dXl5OV0gU1BQYGxsTB8T/EsHAAAAAAD+lnC9FQCDdOvWLSQkZMqUKSUlJYSQsrKytLS0qqoqQoinpyfDMHfv3t2xYweX39bWNjc3l3vq4+OTnp5OCKmrq1u1apXhTejNuXbt2pycHLVavXTpUmNj4+DgYMPLUgKRGxkZTZgwITEx8dixY5MmTdIqKJfL/f39Z86cWV1dXVpaOn/+fKVSqTWRBC+Hh4cHd2k9ExMTDw+P5OTkpq2LtrKyYhhG7726BV56beXk5NBh4bsadKO8+eabDMPMmzePS2mNYwIAAH8Hffr02blzp+H5g4OD6S9m586d3dzcpk6dyl10T5ePj8+2bduaH2RL1aPXv/71L4Zhjhw5opkYHBys+TvOcXd3b+aveWMHnBCSnZ3N3dTiRTCk/hf6FjQ2GEPyNGGcX3KFFI6IAAyVmppqZ2cXEBBgamrq7+9/6tQpNze3+Pj4wMDAwMDA2NjYMWPGcJlnz569fft2c3NzhUJBCElISMjKyvLy8goNDR00aJDhTejNNnXqVKVSaWFhcfDgwfT0dGNjY8PLUgKRE0I++OCDY8eOBQYGuri46JZNS0urqalxcXHx9vb29PRMSkoSHDZ4sWQy2ahRo3r06HH16tVZs2YlJibqzfbs2bOXHFgb5uPj4+fnZ2Fh8aoDef34+xOWJSxL6urIL78QB4dXHRAhBuxBvrRdzBfhBe0d6sV3ENKqBxCgLVGr1U0o9emnnz579qysrGzXrl3Pnj3z9vY+e/as3prj4uK4y/a95CAN9OzZs5SUFEtLyw0bNry4VgB4sQCtHN2S4xsDGz+0Ab179yaEzJgxgz4dMmQIIaR///70Kb2o8JdffhkQEGBkZPTTTz/dv38/Ojra1dW1c+fOPXv2XLx4cV1dnWbm5cuXDx48WCQS9e3b97ffftN8ad++ffRpWVlZTExMt27dOnfu7OnpuXnzZppOT0ZLSEjw8/Pr1KlTSEhIUVFRTEyMqampo6NjSkoKzbZ8+fLu3bsbGxubmpr+4x//4NI1RUVFOTk5iUSijh07uru7f/PNN9xLXKeGDh1qamrau3dvLs4bN24EBwd37tzZ29v7+++/p5/xsrIyw+t/+PDh//zP/7i5uXXq1MnGxmbOnDncIHN69+6tNSYNjqreaPmG4pNPPiGEvPXWWwZuA68GIbx//v4sy7KLFrFjx7IqFbttm1Dml6Vfv35bt24VyLBr167c3NxmtvLs2bMmv9qcFr28vFJTU19Ezbrkcjn3VaCpwREGeHH49ujosoWsrCzDqyorK4uPj7969Wpj2+IkJSU5OztLJJKuXbsuWrSIJtrY2Jw8eZI+Pnv2rImJCX0sl8vnzJkTHBzs4+MTEBBw+fJl4fS7d+8qFAqpVGpvbz9jxownT55w+efNm/fGG2+4ubl5eHgYGRlZWVk5OztPmjSJLyRNgwYN+uyzzzRThg4dOmjQIK2aMzMz2b8+7ElJSf70255lWZY9ffq0WCyurq5mWfbx48fTp093cHCwsrJ69913uV9hraq4Lw2+/HrDbrAvLMvu2rWrc+fOP/zwQ4cOHUpKSgS6yQWWlJQUFBREd34uXbokMNq6AcTExGgNOF+Pbt++HRISIpFIevfu/c0333CbgVYwy5cvf+ONNzp16hQQEFBUVLR06VJra2srK6sVK1Zw2fTGxlc/Xzya39sNDuyKFSvc3NzEYrGzs/OXX37JRRsfHx8cHNyrV69+/fpxQ2dgZ/ny6G1Ld5z1ZtPbEb0joFuhIVuXIXB4Ca0e5lbg70lzbuXOnTtubm6EkGHDhtFX6YE9wzBDhgwZP358ZmYmzdCjR4/JkyfTV8eNG6eZ2djYODw8nN5SytTUtLS0lP3veQSVSuXu7k4ICQgIWLBggaOjIyFk/fr17F9zK0ZGRsOHD6ePzczMXFxc/Pz8CCEdO3a8e/fuH3/8QQixtbVduHDh3LlzQ0ND58+fr9svHx+fSZMmffHFF9OmTWvfvj0h5Oeff9aKMzY2ltbcvXt3lmWfPn3avXt3QohMJhs/fnyXLl0E5lb01q9SqTw8PAghFhYWkZGRYWFhY8aMYVk2ISHB2dmZTlp98sknCQkJmmNSV1dnyKhqRcuyLN9QtJG5lchIlhA2K4u9dKk5cyt694fu3LljZWWVlpZG84wePXrChAks/9EIq7EHqXdXjP3vXUy+nUW9wegeeGjSelXgkMPw4yitOpuwdyhwOKdZM1/rS5Ys6d+/f5cuXQYMGMCV1RxAgW42eMzQUru28LfSgnMr169fJ4Ts2bOnsW1Rubm5YrH45s2bLMtWVFScP3+epgvMrfTs2fPRo0csy6akpDg5OdGpeb70AQMGjB8/vrKysqSkZMCAAR9++CFXj4eHR3l5OX2qOd/KF5Im3UmHTZs2dejQ4enTp1o1s3992O/fv29kZMRNSU+fPj0yMpI+Dg0NHTt2bFlZWU1NzbRp00aNGqU3SO5LQ29+vWEb0heWZYcNGzZp0qTnz587OzsvXbpUoJtcYHK5PCsrq6am5qOPPho4cCBN1x1tvgC0Jrj5RsDPzy8qKqq2tvbWrVseHh58cys9e/a8cuVKVVXViBEjunfv/umnn1ZXV584ccLIyOj//u//+GITqJ8vHu4tMGRgd+3alZ+f//z58wsXLtjY2NCfYLlc7uPjQzfUhIQEbugM7CxfHr1t6Y6zbja+jvCNQGM/KQbC4SW0eoat0NLjVQcO0CxaSyoIIe3atTty5Ah9lR7YR0VF0ac//PADIaRDhw737t1jWfbgwYO0yN27d7nM8+bNY1lWpVJ17dqVELJ69Wr2v+dWaCUikaiwsLCsrGzjxo2EEBcXF/avuZW4uDiWZT/99FNCSKdOncrKylQqFT1n7ejRo8eOHSOEODs7p6SknD59mvufm5ba2lqWZevr66urq0NDQwkhs2fP1uzU559/zrIst2L50aNHP/30E+3+7du3WZbdtGmTwNyK3vrpvcYZhrl+/TrN9vTpU/pg+PDhhBDNfTJuTAwZVd1oWZblG4pz585t3br1+PHjjd4aXqYG51aiolhHR7aigj10qDlzK3z7Q4cOHTI3N8/Pz1+xYkX37t0rKytZ/qMRVmMPkm+PTWtuRe/Oot5gdA88NGm9KnDIYfhxlG6Ljd07FDic06yZr3VnZ+erV68+ffp0wYIFXFnNARTupsAxQwvu2sLfivDcypQpUywsLAYOHFhbW6tSqSZPnmxubm5jY7N48WKWZS9cuODu7t6xY0dHR8fMzEzupordunVrVFtUXl6eSCTavXs3/VLiCMytrFy5kstmb29/9OhRvvScnBxCyP3792liRkZGp06d6uvraX7NBZ6a3wl8IWnSnXSgP6nl5eVaNbMaH/YxY8Z8+umnLMvW1dWZm5vTn638/HyGYbggHz9+zDBMRUWFbpC0Hr78esM2pC+0QjrT/fnnn3fr1u358+d83aQ0A7t8+bJIJGJZVu9o37x5U28AmgPO1yN6GbiHDx/S9C1btvDNrfz73/+mj7dt2yaRSOhbzLKsh4cH/c3SG9u1a9f01i/wjnBvpSEDqykuLi46OppGu2HDBpp47do1OnQ0wgY7a+CAcG2xgos0aTa9HREYgcZ+UgyE660AALRi9Hor4eHhcXFxly5devPNNzVf5W6tffv2bUKIlZWVjY0NIYSu0SCE3Llzh8tMEzt27EjXYtAimmhKbW2tTCaTSqWTJ08mhBQWFnL3wPLx8aGtEEJcXV2trKw6duxIr09WVVU1ePBgpVJZXFysVCoHDBhgaWmpe7Ge58+fL1iwwMHBwcjISCwW09mK0tJSzTyenp6EEO6WWNXV1bQXlpaWdClN3759+YaLr3567y0rKyu6MIcQYmRkxFeJ1oAIj6putIQQvqHo16/fe++9J3BJptZh40Zy+zZ59owsXNjkOgoKCg4dOrRu3TorKyuRSLRs2bLDhw8/fvyYEDJy5MgpU6aMGjUqPj5+165dpqamtMiHH35oZmZGCJk0aVJ9ff3x48e16oyIiHBxcWEYxtvbW6lUZmRk6G2aq2fMmDHnzp0TDmb69Ol0Ek0v7lWBGvRGnpube/r06VWrVpmamtra2i5evDglJeX58+fCLXbo0KG+vj4rK6uqqkoikfB9EPgGiqtZoPXo6OhevXoZGRnNmzevrq5Oa5CFuxkTE9O7d2+xWDxx4sSysrJFixaZmJgEBQXJ5fLs7GwDgwdoFIZhFi5cePLkyZ9++mnt2rWbNm1KSUn53//933nz5p08eTItLe3x48dnzpzZt2+fs7MzvXTRunXrhK9Yx8fV1XXLli1r1qyxtbUNCgo6ceJEg0VsbW01HxcXF/OlFxUVicViqVTKtVVXV8ddr93Ozq6lQiKEFBUVdejQga4A5as5MjJy69atz58/P3DggJmZWVBQECGksLCQYRhfX1+ZTCaTyTw9PSUSCdcp3ar48usN25C+bNiwQSaT0d/QyMjI/Pz8zMzMBvvLzamJxeLa2lq1Wq13tLt06dJgAHw9KioqsrS0NDc3p9l69OjBFww3SiYmJra2ttzNIkxMTOhtKPTGlp2drbd+4XeEq6HBfu3fvz8oKMjZ2dnZ2Xn9+vVlZWU0nbtPqEgkokNHI2ywswJ5+NpqMCS9HTFkBAwcBANhbgVavSbPLL7qwAFawKhRow4ePLh79+6vvvpKdyWLSCSiD+ikQ3l5+f379wkhV69epekOGlcbvXbtGiHk2bNn9B4BtIgmmmJiYkIXoVDp6ekdOnSgGeh8BL0Vt+7cRLt27b7//vvKyspz584tXLiwrq5uwYIFWnl++OGHr776qrKyctu2bSdPngwODiaEaH1a6Yk8mjf8pr149OgRPYorKCjgGy6++p2cnOj4cPdH4K7+S5ujB5Z6B0R4VHWjFRiK8+fPb9u2rTk/6q+FhQuJpydxdCRnzjS5DuH9oRkzZty4cWPo0KGaR+B8RykcA/fYdHcWBYLhO/CguFeFu9Oo4yiBFg3cO+QbKK5mgdZlMhlNbN++vaOjo9YgG3hkpfeYoQV3bQE4U6dOHTVqFCHk0aNHV65cMTMzGz16dHh4OCHkypUrCoXC2tq6b9++77zzTkFBAf3sW1hYaH5GGkWhUGRmZpaXl48YMWLs2LH0x0ssFnP/gdC6e11eXh59oFarb926xR3k66Y7ODhUV1dzxQsKCoyNjbkvK82fGK1bN+oNSVhqauobb7yh+YOuKzQ0tLa29pdffklJSVEqlTSbk5MTwzAXL14s/MujR4969eqlGyQlkF9v2MJ9UavV33//fXFxsVQqtbKy8vX1JYQ07Yq2fKOtNwDNAefrkb29fUVFhUqlotnu3bvXhKgEYvPy8tJbv/A7whEe2JKSkoiIiPj4+MLCwlu3bk2fPl14KzKks3x5BNrSHGe+bLodERiB5n9S9MLcCgBA2zdy5Mju3bur1eqgoKApU6YolUpCSHh4uOZxWmJi4oQJEwICAkpKSsRicUREhFYlo0aNcnNze/LkyZdffvnHH3/89NNPiYmJ33zzDTe3Iuz333/38vKaPXv2Tz/9dPnyZUKItbW1Vh7u9gEqler8+fP0uiQNCg4OdnV1VavV77zzzrJly/75z3/y5eSrf+TIkT179mRZNiAgICoqavz48e+++y59iV5vZfv27R9//LHm3cqJYaPaqKHYunXr+++/n5ycbEivX1+FheTyZfLXsUTTCOwP1dfXK5XKiIiIkydPHjhwgCvCd5RCNXbv0MBg+A48KO5V4R3cRh1H6c7TaT41ZO+Qb6C4mgVapyu8CCH19fV37tzRGmQD9+P5tNSuLQDHyMiIbtgsy3p4eFRUVBw5coSeztm7d+/27dvv3r27oKDg6dOne/bs6dSpEyEkLy9PawbEQDk5OZmZmSqVSiQSSaVS7h8MPj4+6enphJC6urpVq1ZpFlm3bt3Vq1fVavWiRYtEIhGd7tebLpfL/f39Z86cWV1dXVpaOn/+fKVSqfXxp2xtbXNzc4VD0sKyrFqtrqmpyc7Onjx58qlTp7766ivhzhoZGU2YMCExMfHYsWOTJk2iid26dQsJCZkyZUpJSQkhpKysLC0tTaASvvx6w26wLwcPHrx///7p06ez/7J+/fp9+/Y9ePBAuC+69I72jRs39AagOeB8PXJ3d/f29qY/7iqVSmszaH5svXr10lu/Ie9IgwNL18t4enoyDHP37l2tHSFdhnSWL49AW5rjrDeb3o4IjEATPimGwNwKAEDb16lTp1OnTk2bNk2lUu3YscPS0nLhwoVbtmzRzPPFF18UFxdfuXKlT58+hw8fpue5aDI2Nj558mRMTMzdu3cXL168efNmlUrFzUE0yMbGxsrKaufOnZ9//nlGRsawYcN2796tlSc8PPyjjz7q0KFDXFzcn3/+qXWKE5+OHTsePnw4KCjozJkz27dvnzVrFl9Ovvpp12bOnGlhYbFjx45ffvlFLpfTl2bNmuXn51daWrp69Wp6FjrHkFFt8lC0Sn/8QRiG/HWfpuYQ2B9asGBBZWXl5s2bN27cGBUVxZ2BxXeUQjV279DAYFqkhpd8HCU8UIRn3522vnbt2pycHLVavXTpUmNjY62yzRmoFty1BdDrww8/jIyMnDBhQkJCwsKFC4OCgi5evDho0KDu3btLJBKlUunk5DRs2LDPPvvsvffea0L9tbW1c+bMsba2trCw+O6779LS0ui0TkJCQlZWlpeXV2hoqNYpn9HR0VFRURYWFj/++GN6ejq9PBlfelpaWk1NjYuLi7e3t6enp+5JtdTs2bO3b99ubm6uUCj4QtKyZMkSIyMjS0tLhULBMExWVpYhN1r+4IMPjh07FhgY6OLiwiWmpqba2dkFBASYmpr6+/s3eHaV3vx6w26wLxs2bAgLC+vbt6/DXyIjI21sbAz5UdalO9p8AWgOuMAI7NmzJyMjw9fXNzQ0dMyYMU0ISSA2gfobfEcaHFjBqg+PAAAZ5klEQVQ3N7f4+PjAwMDAwMDY2FhDgjeks3rzCLSlOc56s/F1hG8EmvBJMUiTz6cAAACAvx2By9M26q8hFRUVH3/8sUwmE4vFrq6un3zyCcuyP//8s0QioVc8ZVk2JiYmMDDw2bNncrl86dKl/fv3NzU19fPzy87O5urhrti3ePHiHj16BAQEvPPOOzExMWFhYVoZ2P++zTA9uYzekFhvMHz3JNatiq8Gmk1v5EVFRe+8846VlVXXrl2jo6Pp/U11Wzx+/HiPHj3MzMzCwsIuXLhAb+Jjbm7u5+f366+/6o1Kb3NaNfO1Tu8TpFVWcwAFusnVv2/fPrlczrXl5+e3efNmQ4IH0PUyD2dati2+LxDhLxYAeJ0xLJZcAgAAgIGa/M8cLS26++Hu7p6QkPD222+3YJ0vx0uOvPUOFIBe3Ck/ra4tvg8jPqQArRfOCQIAAAAAAAAAaDqDLkAIAAAAAAAALSInJ6dR6QDw+sM5QQAAAGCw1/KcIAD4e2q95wQBQNuDdSvQ6jX5Ss74dQQAAAAAAIDmw9wKAAAAAAC0Vk2/YSoAQMvB3Aq0EaNHjzY8c3p6+ouLBACgLcOKPwAAAAAduE8QAK+LFy96enpKpdK33nqLL4+Pj8+2bdteZlQAAAAAQAhhX7pX3WMAeH1h3QoAr40bN3p5eV26dEkgT1xcXJ8+fXTT1Wp1hw74fAEAAAAAALR9WLcCoN/kyZM3b978n//8RyaTbdq0iRCSlJQkl8tNTU1lMtmSJUtotsTExDNnztDH7u7u8+fPHzhwoFwuP3XqlLu7+4IFCwYPHty7d28fH5/Lly/TbJWVldHR0Y6OjlKpdNy4ceXl5TQ9OTlZJpOZmZnZ2dktXrxYIBEAAAAAAABeH/i/OoB+GzduZBjGzMxs+fLlNMXBweHo0aMymSw7O3vEiBFyuTwsLEyr1P79+48fP25paUmfHjp06NixY2ZmZl999dWMGTN+/fVXQsjEiRONjIyysrJMTEz++c9/RkZGHjx48MaNG59//nlWVlb37t0fP36cl5dHCNGbCAAAAAAAAK8VzK0AGCoiIoI+8Pb2ViqVGRkZunMr06dP5yZWCCEffvihmZkZIWTMmDELFy4khBQUFBw6dKi0tNTKyooQsmzZMjMzs8ePH3fo0KG+vj4rK8vGxkYikfTt25cQojcRAAAAAAAAXis4JwjAUPv37w8KCnJ2dnZ2dl6/fn1ZWZluHjs7O82ndAKFECISiWpra9VqdWFhIcMwvr6+MplMJpN5enpKJJLi4mJXV9ctW7asWbPG1tY2KCjoxIkThBC9iQAArxLDtMwfAAAAQBuCdSsABikpKYmIiDhy5MiQIUMYhpk7d+7Nmzd1szENHTA4OTkxDHPx4kWJRKL1kkKhUCgUtbW1K1euHDt27IMHDxiG0ZvYYr0CAAAAAACAZsO6FQCDVFVVEUI8PT0Zhrl79+6OHTuaVk+3bt1CQkKmTJlSUlJCCCkrK0tLSyOE5OTkZGZmqlQqkUgklUqNjIz4EgEAAAAAAOC1gnUrAAZxc3OLj48PDAyUSqW2trZjxoyhkyNNkJqa+vnnnwcEBJSXl1tbW48ePZquTJkzZ05OTk779u3d3NzS0tIYhtGb2LL9AgAAAAAAgGZiWJZ91TEANAudbhg9erThRdLT0wkh2PgBABqtpWZ48Q0MAAAAbQjOCQIAAAAAAAAAaLr2CxYseNUxADQLvbfxjcagBbHxAwA02sKFLVPPy/oG9vHx6dy5s6enZ5Mz/M316dPHzMzMw8ODPm3Vw5Wdnd29e/dPP/1UII+7u7uDg4O7u7tWetM6/gqHS+uNM5AhAbfqbQAA4MXBuhUAAABoOdeuEZYlH3zwquMwVFxcXP/+/ZtZiVqtbpFgXs/mNLXIcLVGTev4SxuuRm0SApkNCfhvuw0AAAjD3Aq0emxTverAAQDanJ49Sc+ehGWJQtHMmiorK6Ojox0dHaVS6bhx48rLywkhRUVFUql07969NM+YMWMmTpxIH7u7u8+dO3fw4MG+vr6BgYFXrlzRqjApKUkul5uamspksiVLlnDpiYmJZ86c4SpZsGDB4MGDe/fu7ePjc/nyZYFg3N3d58+fP3DgQLlcfurUKa3mKioqPvroI2dn5y5duvTr1y83N1cgBr52NWk1Z2tryzV67tw5sVhseFV8YejGHBsbe+3atdjYWJlMplQqtYarpKQkPDzc2trawcEhJiampqZGOIbk5GSZTGZmZmZnZ7d48WKtqPQOspZbt26FhISYmZl5enp+++23XK8F3LlzZ/jw4XQFx2+//WZIc9evX/fz85NIJAEBAdyGxHWcb/T09o4rxTcmBvZI71ALbIG6b5xW5gY/DgIbUkt1CgCgrWnycSkAAAD87RAi9Dd/Psuy7M6drErFSiRCORsSGho6duzYsrKympqaadOmjRo1iqYfOnTI3Nw8Pz9/xYoV3bt3r6yspOlyubxnz56PHj1iWTYlJcXJyamuro5l2X79+m3dupVl2V27duXn5z9//vzChQs2NjZpaWm0IJeBVuLj40MrSUhIGDhwoEAwcrncw8OjvLxcb/xvvvnmyJEji4qK6uvrL126VFxcLBADX7uatJqzsbE5efIkfXz27FkTExPDq+ILQ2/MXl5eqampXFnN4RowYMD48eMrKytLSkoGDBjw4YcfCsSQm5srFotv3rzJsmxFRcX58+e1ouJ7xzX5+PhER0fX1dUVFxf7+PhwvRbg5+cXFRVVW1t769YtDw8Prghfc3K53NnZ+erVq0+fPl2wYIGBGxJf77hSfO+LgT3SO9TCW6DWG6eVucGPg8CG1FKdAgBoYzC3AgAAAAYTnlvJzmavXWOHD2dZln3//SbPreTn5zMMc//+ffr08ePHDMNUVFTQp//617969uwpFos1j8/lcvnKlSu5p/b29kePHmX/ey6AExcXFx0dTR9rza1s2LCBPr527ZpIJBIIRi6Xf/PNN3rjz8vLI4TQuQk+mjHobVeLVnMCcysNVqU3DL6Y+eZWcnJyCCHcsGRkZHTq1Km+vp4vhry8PJFItHv3bm46TJPwO05dv369Xbt2XPFdu3Y1eNCek5PDMMzDhw/p0y1bttAiAs3J5fKEhASarlarra2tDdmQ+HqnOQ2hOyYG9ohvqAW2QFbf3ApfZr0fB4ENqUU6BQDQ9uCcIAAAAGgJ3boRLy/y88/k11+JStWc04IKCwsZhvH19ZXJZDKZzNPTUyKRFBcX01dnzJhx48aNoUOH9u3bV7OUra2t5mMuP7V///6goCBnZ2dnZ+f169eXlZXpbdrKyoo+EIlEtbW1arVaIBg7Ozu++Dt37qz7qkAMuu3qVsvXXINdMCQMvpj5FBUVicViqVRKn7q6utbV1XFn1ujG4OrqumXLljVr1tja2gYFBZ04cUKzNuF3nCouLrawsDA1NaVPnZycDAnS0tLS3NycPu3Ro4chzclkMvqgffv2jo6OhmxIwr3jGxMDeyQw1Ia/X1qZDfk4GLJNNrlTAABtT4dXHQAAAAC0CXQyJTaWxMYSQsjw4cTUlFRVNaEmJycnhmEuXrwokUi0Xqqvr1cqlRERET/++OOBAwfeeust7iW68oIQolarb926ZW9vz71UUlISERFx5MiRIUOGMAwzd+7cmzdvNj8YhmH0FnF2dq6pqSkpKenatWuLxKDbnFgsrquro4/1XpqED18YemMmhLRrp///cA4ODtXV1eXl5fTouqCgwNjYmDvS1kuhUCgUitra2pUrV44dO/bBgwdcjwQGmWNvb//w4cOqqip63H779u0GO2tvb19RUaFSqYyNjQkh9+7dM6S5W7du0Qf19fV37twxcEMS6F0zeyQw1AJN6L5xXObmb4oCmvA2AQC0DVi3AgAAAC0hLIzk5xNvb+LtTWbMIMbGJDS0aTV169YtJCRkypQpJSUlhJCysrK0tDT60oIFCyorKzdv3rxx48aoqKg7d+5wpdatW3f16lW1Wr1o0SKRSBQcHMy9VFVVRQjx9PRkGObu3bs7duxokWCEi0ybNq2kpIRl2StXrpSUlDQnBl0+Pj7p6emEkLq6ulWrVhlekC8MvTETQmxtbemFeLXI5XJ/f/+ZM2dWV1eXlpbOnz9fqVTyTcQQQnJycjIzM1UqlUgkkkqlRkZGmq8KDPKuXbu+//572qK3t/fs2bNVKtXdu3eXLVvGFefyaHF3d/f29k5OTiaEqFQqbqCE39O1a9fm5OSo1eqlS5caGxsbsiEJ946PQI+0sjVqqCm+N06gFy3CwE4BALQ9mFsBAACAltC/P+nWjWRnk+xssmYNYRiSmtrkylJTU+3s7AICAkxNTf39/eltUDIyMlavXr17925jY+O33357woQJ48eP505ViI6OjoqKsrCw+PHHH9PT0+lSBcrNzS0+Pj4wMDAwMDA2NnbMmDHND0bYzp07u3bt6uvrK5FIlEplVVVVM2PQkpCQkJWV5eXlFRoaOmjQIMMLCoShGzMhZPbs2du3bzc3N1fonOSVlpZWU1Pj4uLi7e3t6emZlJQk0G5tbe2cOXOsra0tLCy+++67tLQ0rTUXfIN85MiR/fv3E0IYhklLS7tx44a1tXVISMjEiRO5t5jLo2vPnj0ZGRm+vr6hoaGanRV4T6dOnapUKi0sLA4ePGjghtRg7/QS6JGWRg01JfDGteymqMXwTgEAtDEMizvRAgAAgIEMOGI0SEvvfri7uyckJLz99tstWy28tlJTUxMTE7Oysl51IC2m7fWItNFOAQDohXUr0OoxTfWqAwcAAIBGOHv2bH5+PiEkPz9/yZIl48aNe9URNVfb6xFpo50CAGgQrmULAAAAAK1AUVGRQqF4/PixRCIZN27czJkzX3VEzdX2ekTaaKcAABqEc4Kg1aMrUDRvvdkgeqcAbPwAAI32up4TBAAAAPAK4ZwggBcuOztbLBYL5/Hx8dm2bVvz63F3d+e7nl+LMCROAAAAAACAvxWcEwQAjRAXF9enT59XHQUAAAAAAMBrBOtWAHi5u7uvWLFi4MCBIpEoMDCwuLg4ISHBxsZGKpVq3v6wpKQkPDzc2trawcEhJiampqaGEHLnzp3hw4ebmZl5eHj89ttvmtVWVlZGR0c7OjpKpdJx48aVl5cLxMBXj0Al169f9/Pzk0gkAQEBV65coYm2trbcDSbPnTvHrX+5detWSEiImZmZp6fnt99+22B6YmLimTNnuPFZsGDB4MGDe/fu7ePjc/nyZeGyAAAAAAAAbRLmVgCEbNy4ce3atWVlZRKJJDg4uKqqKj8/f+/evXPmzMnLy6N5wsLCjIyM8vLyzp07d+HChVmzZhFCwsPDHRwc7t27d/jw4bVr12rWOXHixNLS0qysrNu3b0skksjISIEA+OoRqGTdunWbN28uLy8fPnz4qFGjVCqVQP0KhaJ79+6lpaVHjx7dtGlTg+laDh06tG/fvqtXr4aHh8+YMaNRZQGgVWLZlvkDAAAAaEtYgFaObsm2jWHgxi+Xy//973/Tx9u2bZNIJPX19fSph4dHWloay7I5OTmEkPv379P0jIyMTp06Xbt2jWGYhw8f0sQtW7aYmJjQx/n5+QzDcPkfP37MMExFRUW/fv22bt2qFUBOTo7eevgqoTEnJCTQdLVabW1tffToUZZlbWxsTp48SdPPnj1L67l+/Xq7du0qKytp+q5du4TTWZbVjFMul2/YsIE+vnbtmkgkEi4LAAAAAADQJuF6KwBC7Ozs6AMTExNbW9t27dpxT6uqqgghRUVFYrFYKpXSdFdX17q6uuzsbEtLS3Nzc5rYo0cPrsLCwkKGYXx9fbkUiURSXFyst/WioiK99fBVIpFICCEymYwmtm/f3tHRka9yQkhxcbGFhYWpqSl96uTkJJyuy8rKij4QiUS1tbVqtdrwsgAAAAAAAG0D5lYAmsXBwaG6urq8vJzOMhQUFBgbG3t5eVVUVKhUKmNjY/LXLZ8pJycnhmEuXrxI50GE2dvb661HuJJbt27RB/X19Xfu3LG3tyeEiMXiuro6ms5dnMXe3v7hw4dVVVV0KuT27dvC6YZoTlkAAAAAAIDWCNdbAWgWuVzu7+8/c+bM6urq0tLS+fPnK5XKXr16eXt7JycnE0JUKtWqVau4/N26dQsJCZkyZUpJSQkhpKysLC0tja9yd3d3vfUIV7J27dqcnBy1Wr106VJjY+Pg4GBCiI+PT3p6OiGkrq6Oq0cul3t7e8+ePVulUt29e3fZsmXC6QYOSJPLAgAAAAAAtEaYWwForrS0tJqaGhcXF29vb09PT3oLoT179mRkZPj6+oaGho4ZM0Yzf2pqqp2dXUBAgKmpqb+/P3f7Hr346hGoZOrUqUql0sLC4uDBg+np6XTNS0JCQlZWlpeXV2ho6KBBg2hOhmHS0tJu3LhhbW0dEhIyceJEmpkv3RDNKQsAAAAAANAaMSyu1Q+tHMMwhBDuCrWGoCfXYOPXkpqampiYmJWVZWB6c+oEAAAAAABoM7BuBeBv7ezZs/n5+YSQ/Pz8JUuWjBs3Tji9OXUCAAAAAAC0SbiWLcDfWlFRkUKhePz4sUQiGTdu3MyZM4XTm1MnAAAAAABAm4RzgqDVo+cENQE2fgAAAAAAAGg+nBMEAAAAAAAAANB0WLcCAAAAAAAAANB0WLcCAAAAAAAAANB0mFsBAAAAAAAAAGg6zK0AAAAAAAAAADQd5lYAAAAAAAAAAJoOcysAAAAAAAAAAE2HuRUAAAAAAAAAgKbr8KoDAGguhmGaVhA3IAcAAAAAAIDmw7oVAAAAAAAAAICmw7oVaCMatQilyUtdAAAAAAAAALRg3QrAC5ednS0Wi4Xz+Pj4bNu2rZkNaVbCV6FmMFyePn367Ny5s1FtNaFIMwsCAAAAAAC8nrBuBaBtiouL69OnT/PzAAAAAAAAgDCsWwHg5e7uvmLFioEDB4pEosDAwOLi4oSEBBsbG6lUmpSUxGUrKSkJDw+3trZ2cHCIiYmpqakhhNy5c2f48OFmZmYeHh6//fabZrWVlZXR0dGOjo5SqXTcuHHl5eV8ASQnJw8YMIB7+scff5iamj558oQQkpSUJJfLTU1NZTLZkiVLdMsmJiaeOXOGPuYLhuaJjY29du1abGysTCZTKpVff/21v78/l+f333/v0qULbZSjVUSgU8nJyTKZzMzMzM7ObvHixboFAQAAAAAAWjvMrQAI2bhx49q1a8vKyiQSSXBwcFVVVX5+/t69e+fMmZOXl0fzhIWFGRkZ5eXlnTt37sKFC7NmzSKEhIeHOzg43Lt37/Dhw2vXrtWsc+LEiaWlpVlZWbdv35ZIJJGRkXytv/fee+fPn79x4wZ9mpKSolAoTExMCCEODg5Hjx6trKzct2/f119/vXfvXoFeCARDCFm9enWvXr1Wr15dWFiYkpLy/vvvX7p06erVq/TVTZs2vfvuu7RRviJ8nbpx48bnn3/+888/V1RUXL9+feTIkboFAQAAAAAAWjvMrQAIiYmJ6d27t1gsnjhxYllZ2aJFi0xMTIKCguRyeXZ2NiEkNzf39OnTq1atMjU1tbW1Xbx4cUpKyvXr18+cObN8+fJOnTo5OTnFxcVxFRYUFBw6dGjdunVWVlYikWjZsmWHDx9+/Pix3talUumIESPoHIRKpdq1axc3ERMREeHi4sIwjLe3t1KpzMjI4OtCbm4uXzB6mZubKxSKTZs2EUKePHmye/fuyZMnCxfh61SHDh3q6+uzsrKqqqokEknfvn2F6wEAAAAAAGiNMLcCIMTOzo4+MDExsbW1bdeuHfe0qqqKEFJUVCQWi6VSKU13dXWtq6vLzs62tLQ0NzeniT169OAqLCwsZBjG19dXJpPJZDJPT0+JRFJcXMwXQGRk5NatW58/f37gwAEzM7OgoCCavn///qCgIGdnZ2dn5/Xr15eVlfHVUFRUxBcMn6lTp27btu3Zs2d79uxxdHTUPEVIL75Oubq6btmyZc2aNba2tkFBQSdOnGiwaQAAAAAAgFYH17IFaBYHB4fq6ury8nIrKytCSEFBgbGxsZeXV0VFhUqlMjY2JoTcu3ePy+/k5MQwzMWLFyUSiSH1h4aGTps27ZdffklJSVEqlfTu0SUlJREREUeOHBkyZAjDMHPnzr158yZfDfb29nzBcLg5I2rgwIEWFhYHDx7ctGkT36IVzSICnVIoFAqFora2duXKlWPHjn3w4IFWWwAAAAAAAK0dDnIAmkUul/v7+8+cObO6urq0tHT+/PlKpbJXr17e3t7JycmEEJVKtWrVKi5/t27dQkJCpkyZUlJSQggpKytLS0sTqN/IyGjChAmJiYnHjh2bNGkSTaRLZjw9PRmGuXv37o4dOwRqcHd35wuGY2trm5ubq5kyderUL7744s8//3z//ff1VqtZhK9TOTk5mZmZKpVKJBJJpVIjIyO9bQEAAAAAALRqmFsBaK60tLSamhoXFxdvb29PT096C6E9e/ZkZGT4+vqGhoaOGTNGM39qaqqdnV1AQICpqam/v/+pU6eE6//ggw+OHTsWGBjo4uJCU9zc3OLj4wMDAwMDA2NjY7Xq1yUQDDV79uzt27fTK63QFKVSef369dGjR3OnOwkX0dup2traOXPmWFtbW1hYfPfdd2lpaQzD6LYFAAAAAADQqjEsy77qGACahZ4m06gtuQlF/m7UarWDg8PmzZtHjBjxqmMBAAAAAAB4rWHdCgDo8d1330kkkuHDh7/qQAAAAAAAAF53uJYtAGgzNzc3MTHZsWMHrjsLAAAAAADQIJwTBK0ePcGnCbDxAwAAAAAAQPPhn9IAAAAAAAAAAE2HdSsAAAAAAAAAAE33/8LbYKby0pmwAAAAAElFTkSuQmCC",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import Image\n",
    "Image(filename='../../docs/source/images/explainers-overview.png')"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "bb54d522-bd46-499e-a27c-15332bfbeace",
   "metadata": {},
   "source": [
    "**Table of Contents**\n",
    "\n",
    "* Pre-requisites \n",
    "  * Driverless AI Connections Configuration\n",
    "  * Driverless AI Experiments Preparation\n",
    "* Interpretation: Python API\n",
    "   * Drift Detection explainer\n",
    "   * H2O Model Validation explainers that use Driverless AI (regression)\n",
    "   * Calibration Score explainer (classification)\n",
    "* Interpretation: CLI\n",
    "   * Drift Detection explainer\n",
    "   * H2O Model Validation explainers that use Driverless AI (regression\n"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "3340a5b3-c4df-4791-a90e-4557b18088f3",
   "metadata": {},
   "source": [
    "# Pre-requisites"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "69f414e3-bc88-478b-bed5-890352b1041a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<style type='text/css'>\n",
       ".datatable table.frame { margin-bottom: 0; }\n",
       ".datatable table.frame thead { border-bottom: none; }\n",
       ".datatable table.frame tr.coltypes td {  color: #FFFFFF;  line-height: 6px;  padding: 0 0.5em;}\n",
       ".datatable .bool    { background: #DDDD99; }\n",
       ".datatable .object  { background: #565656; }\n",
       ".datatable .int     { background: #5D9E5D; }\n",
       ".datatable .float   { background: #4040CC; }\n",
       ".datatable .str     { background: #CC4040; }\n",
       ".datatable .time    { background: #40CC40; }\n",
       ".datatable .row_index {  background: var(--jp-border-color3);  border-right: 1px solid var(--jp-border-color0);  color: var(--jp-ui-font-color3);  font-size: 9px;}\n",
       ".datatable .frame tbody td { text-align: left; }\n",
       ".datatable .frame tr.coltypes .row_index {  background: var(--jp-border-color0);}\n",
       ".datatable th:nth-child(2) { padding-left: 12px; }\n",
       ".datatable .hellipsis {  color: var(--jp-cell-editor-border-color);}\n",
       ".datatable .vellipsis {  background: var(--jp-layout-color0);  color: var(--jp-cell-editor-border-color);}\n",
       ".datatable .na {  color: var(--jp-cell-editor-border-color);  font-size: 80%;}\n",
       ".datatable .sp {  opacity: 0.25;}\n",
       ".datatable .footer { font-size: 9px; }\n",
       ".datatable .frame_dimensions {  background: var(--jp-border-color3);  border-top: 1px solid var(--jp-border-color0);  color: var(--jp-ui-font-color3);  display: inline-block;  opacity: 0.6;  padding: 1px 10px 1px 5px;}\n",
       "</style>\n"
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import json\n",
    "import os\n",
    "\n",
    "import datatable\n",
    "import webbrowser\n",
    "\n",
    "from h2o_sonar import config as h2o_sonar_config\n",
    "from h2o_sonar import interpret\n",
    "from h2o_sonar.lib.api import commons\n",
    "from h2o_sonar.lib.api import explainers\n",
    "\n",
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "7cb20c8f-05de-47cb-94b6-cef937baee50",
   "metadata": {},
   "outputs": [],
   "source": [
    "from h2o_sonar.explainers.drift_explainer import DriftDetectionExplainer\n",
    "from h2o_sonar.explainers.adversarial_similarity_explainer import AdversarialSimilarityExplainer\n",
    "from h2o_sonar.explainers.size_dependency_explainer import SizeDependencyExplainer\n",
    "from h2o_sonar.explainers.segment_performance_explainer import SegmentPerformanceExplainer\n",
    "from h2o_sonar.explainers.calibration_score_explainer import CalibrationScoreExplainer\n",
    "from h2o_sonar.explainers.backtesting_explainer import BacktestingExplainer"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f756429b-a82d-4ecc-a590-8488e9e84015",
   "metadata": {},
   "source": [
    "## Configuration\n",
    "\n",
    "Certain MV-based explainers require Driverless AI connection and/or Driverless autoML connection (see the documentation). Driverless AI connection must be configured prior running the interpretation:"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "edca54f8-9b8d-4b94-9c20-9202a028420b",
   "metadata": {},
   "outputs": [],
   "source": [
    "# CHOSE and UPDATE one of the Driverless AI server options:\n",
    "\n",
    "# OPTION: H2O AIEM hosted Driverless AI\n",
    "AIEM_DAI_WORKER_CONNECTION = h2o_sonar_config.ConnectionConfig(\n",
    "    connection_type=h2o_sonar_config.ConnectionConfigType.DRIVERLESS_AI_AIEM.name,\n",
    "    name=\"My H2O AIEM hosted Driverless AI \",\n",
    "    description=\"Driverless AI server hosted by H2O Enterprise AIEM.\",\n",
    "    # H2O AIEM server URL\n",
    "    server_url=\"https://enginemanager.cloud.h2o.ai/\",\n",
    "    # name of the Driverless AI server in H2O AIEM\n",
    "    server_id=\"new-dai-engine-42\",\n",
    "    # H2O.ai environment URL\n",
    "    environment_url=\"https://cloud.h2o.ai\",\n",
    "    # valid H2O.ai Cloud user name\n",
    "    username=\"firstname.lastname@h2o.ai\",\n",
    "    # H2O.ai refresh token for given environment which might be obtained from:\n",
    "    # H2O.ai Cloud > User > CLI & API Access > API Token (generation)\n",
    "    token=os.getenv(\"H2O_SONAR_AIEM_REFRESH_TOKEN_CLOUD_QA\"),\n",
    "    # REFRESH_TOKEN as token use type\n",
    "    token_use_type=h2o_sonar_config.TokenUseType.REFRESH_TOKEN.name,\n",
    ")\n",
    "\n",
    "# OPTION: H2O Enterprise Steam hosted Driverless AI\n",
    "STEAM_DAI_WORKER_CONNECTION = h2o_sonar_config.ConnectionConfig(\n",
    "    connection_type=h2o_sonar_config.ConnectionConfigType.DRIVERLESS_AI_STEAM.name,\n",
    "    name=\"My H2O Steam hosted Driverless AI \",\n",
    "    description=\"Driverless AI server hosted by H2O Enterprise Steam.\",\n",
    "    # name of the Driverless AI server in H2O Enterprise Steam\n",
    "    server_id=\"my-demo-driverless-ai\",\n",
    "    # H2O Enterprise Steam server URL\n",
    "    server_url=\"https://steam.cloud.h2o.ai/\",\n",
    "    # username (needed by MV, but actually overwritten after contacting Steam)\n",
    "    username=\"firstname.lastname@h2o.ai\",\n",
    "    # H2O Enterprise Steam client access token which might be obtained from:\n",
    "    # Enterprise Steam > Configurations > Personal Access Token > Get token\n",
    "    token=os.getenv(\"H2O_SONAR_STEAM_TOKEN_CLOUD\"),\n",
    "    # ACCESS_TOKEN as token use type\n",
    "    token_use_type=h2o_sonar_config.TokenUseType.ACCESS_TOKEN.name,\n",
    ")\n",
    "\n",
    "# OPTION: local/private Driverless AI installation\n",
    "LOCAL_DAI_WORKER_CONNECTION = h2o_sonar_config.ConnectionConfig(\n",
    "    connection_type=h2o_sonar_config.ConnectionConfigType.DRIVERLESS_AI.name,\n",
    "    name=\"Local Driverless AI server\",\n",
    "    description=\"Driverless AI server running on the localhost.\",\n",
    "    server_url=\"http://localhost:12345\",\n",
    "    username=\"h2oai\",\n",
    "    password=\"h2oai\",\n",
    ")\n",
    "\n",
    "# MAKE YOUR CHOICE fro one of the options above:\n",
    "DAI_WORKER_CONNECTION=LOCAL_DAI_WORKER_CONNECTION"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "6544f951-c69e-4089-9698-b7370c542ae8",
   "metadata": {},
   "outputs": [],
   "source": [
    "# add Driverless AI connection to the H2O Sonar configuration\n",
    "h2o_sonar_config.config.add_connection(DAI_WORKER_CONNECTION)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "82d80586-9e50-484b-b738-4423502bdcbf",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'h2o_host': 'localhost',\n",
       " 'h2o_port': 12349,\n",
       " 'h2o_auto_start': True,\n",
       " 'h2o_min_mem_size': '2G',\n",
       " 'h2o_max_mem_size': '4G',\n",
       " 'custom_explainers': [],\n",
       " 'look_and_feel': 'h2o_sonar',\n",
       " 'per_explainer_logger': True,\n",
       " 'create_html_representations': True,\n",
       " 'connections': [{'key': '0d860575-9d4b-46e5-9870-97ebd73780a1',\n",
       "   'connection_type': 'DRIVERLESS_AI',\n",
       "   'name': 'Local Driverless AI server',\n",
       "   'description': 'Driverless AI server running on the localhost.',\n",
       "   'auth_server_url': '',\n",
       "   'environment_url': '',\n",
       "   'server_url': 'http://localhost:12345',\n",
       "   'server_id': '',\n",
       "   'realm_name': '',\n",
       "   'client_id': '',\n",
       "   'token': '',\n",
       "   'token_use_type': '',\n",
       "   'username': 'h2oai',\n",
       "   'password': 'h2oai'}],\n",
       " 'licenses': []}"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "h2o_sonar_config.config.to_dict(encrypt=False)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "42b1ba13-bff4-4875-b0f9-5164ed7558b0",
   "metadata": {},
   "source": [
    "## Driverless AI Experiments\n",
    "\n",
    "Prepare time series and multinomial Driverless AI experiments from below to be \n",
    "EXPLAINED by H2O Model Validation based explainers provided by H2O Sonar."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "642a8ab0-eaf1-4c0d-a593-bcac532e099e",
   "metadata": {},
   "outputs": [],
   "source": [
    "# TIME SERIES Driverless AI experiment\n",
    "#\n",
    "# 1) create a time series experiment in Driverless AI to build the model to be EXPLAINED, for example:\n",
    "#    train dataset: walmart_tts_small_train.csv\n",
    "#    test dataset : walmart_tts_small_test.csv\n",
    "#    target column: Weekly_Sales\n",
    "#    time_col     : Date\n",
    "# 2) set UUIDs of the Driverless AI experiment and dataset(s):\n",
    "ts_model_uuid = \"b78cb888-f658-11ed-9ecf-0242709d15f7\"\n",
    "ts_dataset_uuid = \"a407dd4c-f658-11ed-9ecf-0242709d15f7\"\n",
    "ts_testset_uuid = \"a4077500-f658-11ed-9ecf-0242709d15f7\"\n",
    "ts_target_col = \"Weekly_Sales\"\n",
    "ts_time_col = \"Date\"\n",
    "\n",
    "# 2) use HANDLERs to reference ^ datasets and model to be explained\n",
    "ts_model_handle = commons.ResourceHandle(\n",
    "    connection_key=DAI_WORKER_CONNECTION.key,\n",
    "    resource_key=ts_model_uuid,\n",
    ")\n",
    "ts_dataset_handle = commons.ResourceHandle(\n",
    "    connection_key=DAI_WORKER_CONNECTION.key,\n",
    "    resource_key=ts_dataset_uuid,\n",
    ")\n",
    "ts_testset_handle = commons.ResourceHandle(\n",
    "    connection_key=DAI_WORKER_CONNECTION.key,\n",
    "    resource_key=ts_testset_uuid,\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "effcb2ae-be4b-466d-be9f-37f87b84c9d3",
   "metadata": {},
   "outputs": [],
   "source": [
    "# MULTINOMIAL Driverless AI experiment\n",
    "\n",
    "# 1) create a multinomial experiment in Driverless AI to build the model to be EXPLAINED, for example:\n",
    "#    train dataset: creditcard.csv\n",
    "#    target column: EDUCATION\n",
    "# 2) set UUIDs of the Driverless AI experiment and dataset(s):\n",
    "m_model_uuid = \"4b000748-fba7-11ed-8ea4-8c1d96f410ff\"\n",
    "m_dataset_uuid = \"f330f964-ab83-11ed-aa8e-a0cec818dc16\"\n",
    "m_target_col = \"EDUCATION\"\n",
    "\n",
    "# 2) use HANDLERs to reference ^ datasets and model to be explained\n",
    "m_model_handle = commons.ResourceHandle(\n",
    "    connection_key=DAI_WORKER_CONNECTION.key,\n",
    "    resource_key=m_model_uuid,\n",
    ")\n",
    "m_dataset_handle = commons.ResourceHandle(\n",
    "    connection_key=DAI_WORKER_CONNECTION.key,\n",
    "    resource_key=m_dataset_uuid,\n",
    ")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "90d401d2-14cd-4686-982f-3cac9e9f5eb7",
   "metadata": {
    "tags": []
   },
   "source": [
    "## Interpretation: Python API\n",
    "\n",
    "Lets run MV test explainers using the H2O Sonar's Python API."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "de825ceb-4e0f-4cd2-a97e-a6ddd7d4ed13",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['h2o_sonar.explainers.drift_explainer.DriftDetectionExplainer',\n",
       " 'h2o_sonar.explainers.adversarial_similarity_explainer.AdversarialSimilarityExplainer',\n",
       " 'h2o_sonar.explainers.size_dependency_explainer.SizeDependencyExplainer',\n",
       " 'h2o_sonar.explainers.segment_performance_explainer.SegmentPerformanceExplainer',\n",
       " 'h2o_sonar.explainers.calibration_score_explainer.CalibrationScoreExplainer',\n",
       " 'h2o_sonar.explainers.backtesting_explainer.BacktestingExplainer']"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# list H2O Model Validation explainers\n",
    "mv_explainers = [e.id for e in interpret.list_explainers() if \"h2o-model-validation\" in e.keywords]\n",
    "mv_explainers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "8d8fd978-4459-4434-bb34-f958c548126e",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'id': 'h2o_sonar.explainers.drift_explainer.DriftDetectionExplainer',\n",
       " 'name': 'DriftDetectionExplainer',\n",
       " 'display_name': 'Drift Detection',\n",
       " 'description': \"Drift detection refers to a validation test that enables you to identifychanges in the distribution of variables in your model's input data, preventing model performance degradation. The explainer performs drift detection using the train and another dataset captured at different times to assess how data has changed over time. The Population Stability Index (PSI) formula is applied to each variable to measure how much the variable has shifted in distribution over time. PSI is applied to numerical and categorical columns and not date columns.\",\n",
       " 'model_types': ['iid', 'time_series'],\n",
       " 'can_explain': ['regression', 'binomial', 'multinomial'],\n",
       " 'explanation_scopes': ['global_scope'],\n",
       " 'explanations': [{'explanation_type': 'global-work-dir-archive',\n",
       "   'name': 'WorkDirArchiveExplanation',\n",
       "   'category': None,\n",
       "   'scope': 'global',\n",
       "   'has_local': None,\n",
       "   'formats': []},\n",
       "  {'explanation_type': 'global-feature-importance',\n",
       "   'name': 'GlobalFeatImpExplanation',\n",
       "   'category': None,\n",
       "   'scope': 'global',\n",
       "   'has_local': None,\n",
       "   'formats': []}],\n",
       " 'parameters': [{'name': 'worker_connection_key',\n",
       "   'description': 'Optional connection ID of the Driverless AI configured in the H2O Sonar configuration. Only Driverless AI servers with username and password authentication are supported.',\n",
       "   'comment': '',\n",
       "   'type': 'str',\n",
       "   'val': None,\n",
       "   'predefined': [],\n",
       "   'tags': [],\n",
       "   'min_': 0.0,\n",
       "   'max_': 0.0,\n",
       "   'category': ''},\n",
       "  {'name': 'drop_cols',\n",
       "   'description': 'Defines the columns to drop during the validation test. Typically drop columns refer to columns that can indicate a drift without an impact on the model, like columns not used by the model, record IDs, time columns, etc.',\n",
       "   'comment': '',\n",
       "   'type': 'list',\n",
       "   'val': [],\n",
       "   'predefined': [],\n",
       "   'tags': ['SOURCE_DATASET_COLUMN_NAMES'],\n",
       "   'min_': 0.0,\n",
       "   'max_': 0.0,\n",
       "   'category': ''},\n",
       "  {'name': 'drift_threshold',\n",
       "   'description': 'Drift threshold.',\n",
       "   'comment': '',\n",
       "   'type': 'float',\n",
       "   'val': 0.1,\n",
       "   'predefined': [],\n",
       "   'tags': [],\n",
       "   'min_': 0.0,\n",
       "   'max_': 0.0,\n",
       "   'category': ''}],\n",
       " 'keywords': ['compliance-test',\n",
       "  'explains-feature-behavior',\n",
       "  'h2o-model-validation']}"
      ]
     },
     "execution_count": 12,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# get explainer name, description, parameters, keywords, ...\n",
    "interpret.describe_explainer(mv_explainers[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "f66a6e6a-c801-4fa6-9bd5-a8d261269bfc",
   "metadata": {},
   "source": [
    "## Run Drift Detection explainer\n",
    "\n",
    "Lets run Drift explainer - which requires two (local or remote) datasets, but does **not** need model."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "bee44e67-307e-4de3-970e-81fdf1c691cd",
   "metadata": {},
   "outputs": [],
   "source": [
    "# prepare PRIMARY DATASET\n",
    "primary_dataset_path = \"../../data/creditcard.csv\"\n",
    "\n",
    "# prepare SECONDARY DATASET (just randomize/sub-set primary dataset for the DEMO purpose)\n",
    "secondary_dataset = datatable.fread(primary_dataset_path)[:500, :]\n",
    "secondary_dataset_path = str(\"/tmp/secondary_dataset.csv\")\n",
    "secondary_dataset.to_csv(secondary_dataset_path)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "70188a3b-5deb-496b-828f-c224efc4d7a5",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:33:16 # \u001b[94mDEBUG\u001b[39m H2O Model Validation LogLevel: DEBUG - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/logging.py:97\u001b[39m\n",
      "2023/09/26 17:33:17 # \u001b[33mWARNING\u001b[39m IoC/DI: No implementations found for >>IMVDatabase<< - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/ioc.py:35\u001b[39m\n",
      "2023/09/26 17:33:17 # \u001b[91mERROR\u001b[39m Can't save test Drift Detection: no database connection - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_test.py:529\u001b[39m\n",
      "2023/09/26 17:33:17 # \u001b[94mDEBUG\u001b[39m Selected database: test-db - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:255\u001b[39m\n",
      "2023/09/26 17:33:17 # \u001b[92mINFO\u001b[39m Initialize MVDatabase: test-db\n",
      "2023/09/26 17:33:17 # \u001b[94mDEBUG\u001b[39m SQLDatabase: results-drift/h2o-sonar/mli_experiment_ea02c00c-8cb5-4596-bde0-1d602a862c1e/tmp/test.sql_db.sqlite - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:187\u001b[39m\n",
      "2023/09/26 17:33:17 # \u001b[94mDEBUG\u001b[39m ObjectStorage: results-drift/h2o-sonar/mli_experiment_ea02c00c-8cb5-4596-bde0-1d602a862c1e/tmp/test.obj_storage.sqlite - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:188\u001b[39m\n",
      "2023/09/26 17:33:17 # \u001b[92mINFO\u001b[39m LocalPlatform local-platform created\n",
      "2023/09/26 17:33:17 # \u001b[94mDEBUG\u001b[39m Database cache is enabled - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:67\u001b[39m\n",
      "2023/09/26 17:33:17 # \u001b[94mDEBUG\u001b[39m Deleting cache entries that are older than 24 hours - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:68\u001b[39m\n",
      "2023/09/26 17:33:17 # \u001b[94mDEBUG\u001b[39m Import dataset obj-3a12f131-56e4-4ecc-b26f-a4bd5aae8ddc from LocalPlatform - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/local/platform.py:64\u001b[39m\n",
      "2023/09/26 17:33:17 # \u001b[94mDEBUG\u001b[39m Import dataset obj-1f7e53f1-2c5b-4477-a027-0acd1092f43a from LocalPlatform - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/local/platform.py:64\u001b[39m\n",
      "2023/09/26 17:33:17 # \u001b[94mDEBUG\u001b[39m Save Drift Detection: Drift Detection - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_test.py:531\u001b[39m\n",
      "2023/09/26 17:33:17 # \u001b[94mDEBUG\u001b[39m Folder data/temp/mvt-701ad883-23dd-4e4f-ae35-e5e0e66cb86a/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:69\u001b[39m\n",
      "2023/09/26 17:33:17 # \u001b[94mDEBUG\u001b[39m Folder data/artifacts/mvt-701ad883-23dd-4e4f-ae35-e5e0e66cb86a/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:82\u001b[39m\n",
      "2023/09/26 17:33:17 # \u001b[92mINFO\u001b[39m Drift Detection 'Drift Detection': Running\n",
      "ic| 'Creating feature distribution'\n",
      "100%|████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 25/25 [00:00<00:00, 188.17it/s]\n",
      "100%|█████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████████| 25/25 [00:00<00:00, 47.32it/s]\n",
      "2023/09/26 17:33:19 # \u001b[92mINFO\u001b[39m Drift Detection 'Drift Detection': Completed\n",
      "2023/09/26 17:33:19 # \u001b[94mDEBUG\u001b[39m Save Drift Detection: Drift Detection - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_test.py:531\u001b[39m\n",
      "2023/09/26 17:33:19 # \u001b[94mDEBUG\u001b[39m Folder data/temp/mvt-701ad883-23dd-4e4f-ae35-e5e0e66cb86a/ deleted - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:227\u001b[39m\n",
      "2023/09/26 17:33:19 # \u001b[94mDEBUG\u001b[39m Folder data/artifacts/mvt-701ad883-23dd-4e4f-ae35-e5e0e66cb86a/ deleted - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:227\u001b[39m\n"
     ]
    }
   ],
   "source": [
    "# run the interpretation\n",
    "interpretation = interpret.run_interpretation(\n",
    "    dataset=primary_dataset_path,\n",
    "    testset=secondary_dataset_path,\n",
    "    model=None,\n",
    "    target_col=\"default payment next month\",\n",
    "    explainers=[DriftDetectionExplainer.explainer_id()],\n",
    "    results_location=\"results-drift\",\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "63742017-22c2-4d1b-a3c0-2b9d63454315",
   "metadata": {},
   "outputs": [],
   "source": [
    "# get Drift Detection explainer result\n",
    "drift_result = interpretation.get_explainer_result(\n",
    "    DriftDetectionExplainer.explainer_id()\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "a592d714-9aa3-47ed-8816-54bea5a4b345",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "drift_result.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "e105dc93-b667-4046-aa33-1b642ababf73",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div class='datatable'>\n",
       "  <table class='frame'>\n",
       "  <thead>\n",
       "    <tr class='colnames'><td class='row_index'></td><th>Feature</th><th>PSI scores</th></tr>\n",
       "    <tr class='coltypes'><td class='row_index'></td><td class='str' title='str32'>&#x25AA;&#x25AA;&#x25AA;&#x25AA;</td><td class='float' title='float64'>&#x25AA;&#x25AA;&#x25AA;&#x25AA;&#x25AA;&#x25AA;&#x25AA;&#x25AA;</td></tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr><td class='row_index'>0</td><td>PAY_2</td><td>0.0915187</td></tr>\n",
       "    <tr><td class='row_index'>1</td><td>PAY_5</td><td>0.0789593</td></tr>\n",
       "    <tr><td class='row_index'>2</td><td>PAY_3</td><td>0.0786285</td></tr>\n",
       "    <tr><td class='row_index'>3</td><td>PAY_4</td><td>0.0645408</td></tr>\n",
       "    <tr><td class='row_index'>4</td><td>BILL_AMT5</td><td>0.041096</td></tr>\n",
       "    <tr><td class='row_index'>5</td><td>PAY_AMT2</td><td>0.0371336</td></tr>\n",
       "    <tr><td class='row_index'>6</td><td>BILL_AMT1</td><td>0.0291849</td></tr>\n",
       "    <tr><td class='row_index'>7</td><td>EDUCATION</td><td>0.0275294</td></tr>\n",
       "    <tr><td class='row_index'>8</td><td>BILL_AMT6</td><td>0.0252464</td></tr>\n",
       "    <tr><td class='row_index'>9</td><td>PAY_6</td><td>0.0249008</td></tr>\n",
       "    <tr><td class='row_index'>10</td><td>BILL_AMT3</td><td>0.0224872</td></tr>\n",
       "    <tr><td class='row_index'>11</td><td>PAY_AMT1</td><td>0.0223227</td></tr>\n",
       "    <tr><td class='row_index'>12</td><td>LIMIT_BAL</td><td>0.0182056</td></tr>\n",
       "    <tr><td class='row_index'>13</td><td>BILL_AMT2</td><td>0.0177724</td></tr>\n",
       "    <tr><td class='row_index'>14</td><td>PAY_AMT5</td><td>0.0167203</td></tr>\n",
       "    <tr><td class='row_index'>15</td><td>BILL_AMT4</td><td>0.0162574</td></tr>\n",
       "    <tr><td class='row_index'>16</td><td>PAY_0</td><td>0.0159517</td></tr>\n",
       "    <tr><td class='row_index'>17</td><td>PAY_AMT6</td><td>0.0150096</td></tr>\n",
       "    <tr><td class='row_index'>18</td><td>PAY_AMT3</td><td>0.014</td></tr>\n",
       "    <tr><td class='row_index'>19</td><td>PAY_AMT4</td><td>0.013425</td></tr>\n",
       "    <tr><td class='row_index'>20</td><td>AGE</td><td>0.0133567</td></tr>\n",
       "    <tr><td class='row_index'>21</td><td>MARRIAGE</td><td>0.00629553</td></tr>\n",
       "    <tr><td class='row_index'>22</td><td>SEX</td><td>0.000531411</td></tr>\n",
       "    <tr><td class='row_index'>23</td><td>default payment next month</td><td>9.55195e-05</td></tr>\n",
       "    <tr><td class='row_index'>24</td><td>ID</td><td>0</td></tr>\n",
       "  </tbody>\n",
       "  </table>\n",
       "  <div class='footer'>\n",
       "    <div class='frame_dimensions'>25 rows &times; 2 columns</div>\n",
       "  </div>\n",
       "</div>\n"
      ],
      "text/plain": [
       "<Frame#7f736f43b5a0 25x2>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "drift_result.data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "bd3efb21",
   "metadata": {
    "tags": []
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'id': 'h2o_sonar.explainers.drift_explainer.DriftDetectionExplainer',\n",
       " 'name': 'DriftDetectionExplainer',\n",
       " 'display_name': 'Drift Detection',\n",
       " 'description': \"Drift detection refers to a validation test that enables you to identifychanges in the distribution of variables in your model's input data, preventing model performance degradation. The explainer performs drift detection using the train and another dataset captured at different times to assess how data has changed over time. The Population Stability Index (PSI) formula is applied to each variable to measure how much the variable has shifted in distribution over time. PSI is applied to numerical and categorical columns and not date columns.\",\n",
       " 'model_types': ['iid', 'time_series'],\n",
       " 'can_explain': ['regression', 'binomial', 'multinomial'],\n",
       " 'explanation_scopes': ['global_scope'],\n",
       " 'explanations': [{'explanation_type': 'global-feature-importance',\n",
       "   'name': 'DriftDetectionExplainer',\n",
       "   'category': 'COMPLIANCE TESTS',\n",
       "   'scope': 'global',\n",
       "   'has_local': None,\n",
       "   'formats': ['application/vnd.h2oai.json+datatable.jay',\n",
       "    'application/json']},\n",
       "  {'explanation_type': 'global-model-validation-result',\n",
       "   'name': 'DriftDetectionExplainer',\n",
       "   'category': 'COMPLIANCE TESTS',\n",
       "   'scope': 'global',\n",
       "   'has_local': None,\n",
       "   'formats': ['application/zip']},\n",
       "  {'explanation_type': 'global-work-dir-archive',\n",
       "   'name': 'DriftDetectionExplainer',\n",
       "   'category': 'COMPLIANCE TESTS',\n",
       "   'scope': 'global',\n",
       "   'has_local': None,\n",
       "   'formats': ['application/zip']},\n",
       "  {'explanation_type': 'global-html-fragment',\n",
       "   'name': 'Drift Detection',\n",
       "   'category': 'COMPLIANCE TESTS',\n",
       "   'scope': 'global',\n",
       "   'has_local': None,\n",
       "   'formats': ['text/html']}],\n",
       " 'parameters': [{'name': 'worker_connection_key',\n",
       "   'description': 'Optional connection ID of the Driverless AI configured in the H2O Sonar configuration. Only Driverless AI servers with username and password authentication are supported.',\n",
       "   'comment': '',\n",
       "   'type': 'str',\n",
       "   'val': None,\n",
       "   'predefined': [],\n",
       "   'tags': [],\n",
       "   'min_': 0.0,\n",
       "   'max_': 0.0,\n",
       "   'category': ''},\n",
       "  {'name': 'drop_cols',\n",
       "   'description': 'Defines the columns to drop during the validation test. Typically drop columns refer to columns that can indicate a drift without an impact on the model, like columns not used by the model, record IDs, time columns, etc.',\n",
       "   'comment': '',\n",
       "   'type': 'list',\n",
       "   'val': [],\n",
       "   'predefined': [],\n",
       "   'tags': ['SOURCE_DATASET_COLUMN_NAMES'],\n",
       "   'min_': 0.0,\n",
       "   'max_': 0.0,\n",
       "   'category': ''},\n",
       "  {'name': 'drift_threshold',\n",
       "   'description': 'Drift threshold.',\n",
       "   'comment': '',\n",
       "   'type': 'float',\n",
       "   'val': 0.1,\n",
       "   'predefined': [],\n",
       "   'tags': [],\n",
       "   'min_': 0.0,\n",
       "   'max_': 0.0,\n",
       "   'category': ''}],\n",
       " 'keywords': ['compliance-test',\n",
       "  'explains-feature-behavior',\n",
       "  'h2o-model-validation']}"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# summary\n",
    "drift_result.summary()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "43e0f191-2040-44ee-b564-9734723cf104",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# open interpretation HTML report in web browser\n",
    "webbrowser.open(interpretation.result.get_html_report_location())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "da4e2b28-96d7-440e-bfea-41cb694a52d4",
   "metadata": {},
   "outputs": [],
   "source": [
    "# save the explainer result as zip archive\n",
    "drift_result.zip(file_path=\"./drift-detection-demo-archive.zip\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "c0540819-f896-481a-b470-b9d53a243b0a",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Archive:  drift-detection-demo-archive.zip\n",
      "  Length      Date    Time    Name\n",
      "---------  ---------- -----   ----\n",
      "     3757  2023-09-26 17:33   explainer_h2o_sonar_explainers_drift_explainer_DriftDetectionExplainer_dfff02c1-2007-4b47-a6fa-70d4013b42f4/result_descriptor.json\n",
      "      110  2023-09-26 17:33   explainer_h2o_sonar_explainers_drift_explainer_DriftDetectionExplainer_dfff02c1-2007-4b47-a6fa-70d4013b42f4/global_html_fragment/text_html.meta\n",
      "      358  2023-09-26 17:33   explainer_h2o_sonar_explainers_drift_explainer_DriftDetectionExplainer_dfff02c1-2007-4b47-a6fa-70d4013b42f4/global_html_fragment/text_html/explanation.html\n",
      "    30716  2023-09-26 17:33   explainer_h2o_sonar_explainers_drift_explainer_DriftDetectionExplainer_dfff02c1-2007-4b47-a6fa-70d4013b42f4/global_html_fragment/text_html/fi-class-0.png\n",
      "      164  2023-09-26 17:33   explainer_h2o_sonar_explainers_drift_explainer_DriftDetectionExplainer_dfff02c1-2007-4b47-a6fa-70d4013b42f4/global_model_validation_result/application_zip.meta\n",
      "   347752  2023-09-26 17:33   explainer_h2o_sonar_explainers_drift_explainer_DriftDetectionExplainer_dfff02c1-2007-4b47-a6fa-70d4013b42f4/global_model_validation_result/application_zip/explanation.zip\n",
      "      142  2023-09-26 17:33   explainer_h2o_sonar_explainers_drift_explainer_DriftDetectionExplainer_dfff02c1-2007-4b47-a6fa-70d4013b42f4/global_work_dir_archive/application_zip.meta\n",
      "       22  2023-09-26 17:33   explainer_h2o_sonar_explainers_drift_explainer_DriftDetectionExplainer_dfff02c1-2007-4b47-a6fa-70d4013b42f4/global_work_dir_archive/application_zip/explanation.zip\n",
      "        0  2023-09-26 17:33   explainer_h2o_sonar_explainers_drift_explainer_DriftDetectionExplainer_dfff02c1-2007-4b47-a6fa-70d4013b42f4/log/explainer_run_dfff02c1-2007-4b47-a6fa-70d4013b42f4.log\n",
      "      185  2023-09-26 17:33   explainer_h2o_sonar_explainers_drift_explainer_DriftDetectionExplainer_dfff02c1-2007-4b47-a6fa-70d4013b42f4/global_feature_importance/application_vnd_h2oai_json_datatable_jay.meta\n",
      "      143  2023-09-26 17:33   explainer_h2o_sonar_explainers_drift_explainer_DriftDetectionExplainer_dfff02c1-2007-4b47-a6fa-70d4013b42f4/global_feature_importance/application_json.meta\n",
      "      738  2023-09-26 17:33   explainer_h2o_sonar_explainers_drift_explainer_DriftDetectionExplainer_dfff02c1-2007-4b47-a6fa-70d4013b42f4/global_feature_importance/application_vnd_h2oai_json_datatable_jay/explanation.json\n",
      "      944  2023-09-26 17:33   explainer_h2o_sonar_explainers_drift_explainer_DriftDetectionExplainer_dfff02c1-2007-4b47-a6fa-70d4013b42f4/global_feature_importance/application_vnd_h2oai_json_datatable_jay/feature_importance_class_0.jay\n",
      "      688  2023-09-26 17:33   explainer_h2o_sonar_explainers_drift_explainer_DriftDetectionExplainer_dfff02c1-2007-4b47-a6fa-70d4013b42f4/global_feature_importance/application_json/explanation.json\n",
      "     1802  2023-09-26 17:33   explainer_h2o_sonar_explainers_drift_explainer_DriftDetectionExplainer_dfff02c1-2007-4b47-a6fa-70d4013b42f4/global_feature_importance/application_json/feature_importance_class_0.json\n",
      "        2  2023-09-26 17:33   explainer_h2o_sonar_explainers_drift_explainer_DriftDetectionExplainer_dfff02c1-2007-4b47-a6fa-70d4013b42f4/model_problems/problems_and_actions.json\n",
      "---------                     -------\n",
      "   387523                     16 files\n"
     ]
    }
   ],
   "source": [
    "!unzip -l drift-detection-demo-archive.zip"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "ab51051c-4896-463a-88b3-ed19b071fec7",
   "metadata": {},
   "source": [
    "## Run Adversarial Similarity explainer\n",
    "\n",
    "Adversarial Similarity explainer can use remote Driverless AI worker to explain local datasets."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "8da024d0-7c89-42be-8398-ad74d9ad0c7e",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:33:44 # \u001b[94mDEBUG\u001b[39m Save Adversarial Similarity: Adversarial Similarity - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_test.py:531\u001b[39m\n",
      "2023/09/26 17:33:44 # \u001b[33mWARNING\u001b[39m Singleton MVClient already initialized, ignoring: args=(), kwargs={'data_folder': 'results-adversarial/h2o-sonar/mli_experiment_c9e66eb3-3deb-4a2c-9785-2d95ff2f38fb/tmp'} - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:296\u001b[39m\n",
      "2023/09/26 17:33:44 # \u001b[94mDEBUG\u001b[39m Selected database: test-db - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:255\u001b[39m\n",
      "2023/09/26 17:33:44 # \u001b[92mINFO\u001b[39m Initialize MVDatabase: test-db\n",
      "2023/09/26 17:33:44 # \u001b[94mDEBUG\u001b[39m SQLDatabase: results-drift/h2o-sonar/mli_experiment_ea02c00c-8cb5-4596-bde0-1d602a862c1e/tmp/test.sql_db.sqlite - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:187\u001b[39m\n",
      "2023/09/26 17:33:44 # \u001b[94mDEBUG\u001b[39m ObjectStorage: results-drift/h2o-sonar/mli_experiment_ea02c00c-8cb5-4596-bde0-1d602a862c1e/tmp/test.obj_storage.sqlite - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:188\u001b[39m\n",
      "2023/09/26 17:33:44 # \u001b[92mINFO\u001b[39m Local Platform already exists\n",
      "2023/09/26 17:33:44 # \u001b[94mDEBUG\u001b[39m Database cache is enabled - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:67\u001b[39m\n",
      "2023/09/26 17:33:44 # \u001b[94mDEBUG\u001b[39m Deleting cache entries that are older than 24 hours - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:68\u001b[39m\n",
      "2023/09/26 17:33:44 # \u001b[94mDEBUG\u001b[39m Save credentials: DriverlessCredentials(address='http://localhost:12345', username='h2oai') - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:33\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "A service ('Driverless AI') is running on localhost:12345 and it is accessible\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:33:44 # \u001b[92mINFO\u001b[39m Adding connection to Driverless AI server 'http://localhost:12345' for user 'h2oai'\n",
      "2023/09/26 17:33:44 # \u001b[94mDEBUG\u001b[39m Save credentials: DriverlessCredentials(address='http://localhost:12345', username='h2oai') - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:33\u001b[39m\n",
      "2023/09/26 17:33:46 # \u001b[94mDEBUG\u001b[39m Worker set: <h2o_mv.platforms.driverless.platform.DriverlessPlatform object at 0x7f73b36a0250> - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:331\u001b[39m\n",
      "2023/09/26 17:33:46 # \u001b[94mDEBUG\u001b[39m Dataset with hash 3c4c14a0ee50160d already stored in LocalPlatform - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/local/platform.py:113\u001b[39m\n",
      "2023/09/26 17:33:46 # \u001b[94mDEBUG\u001b[39m Import dataset obj-3a12f131-56e4-4ecc-b26f-a4bd5aae8ddc from LocalPlatform - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/local/platform.py:64\u001b[39m\n",
      "2023/09/26 17:33:46 # \u001b[94mDEBUG\u001b[39m Dataset with platform_obj_key 'obj-3a12f131-56e4-4ecc-b26f-a4bd5aae8ddc' already in DB - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/databases/sql_db.py:314\u001b[39m\n",
      "2023/09/26 17:33:46 # \u001b[94mDEBUG\u001b[39m Dataset with hash ad452137d213fa4c already stored in LocalPlatform - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/local/platform.py:113\u001b[39m\n",
      "2023/09/26 17:33:46 # \u001b[94mDEBUG\u001b[39m Import dataset obj-1f7e53f1-2c5b-4477-a027-0acd1092f43a from LocalPlatform - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/local/platform.py:64\u001b[39m\n",
      "2023/09/26 17:33:46 # \u001b[94mDEBUG\u001b[39m Dataset with platform_obj_key 'obj-1f7e53f1-2c5b-4477-a027-0acd1092f43a' already in DB - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/databases/sql_db.py:314\u001b[39m\n",
      "2023/09/26 17:33:46 # \u001b[94mDEBUG\u001b[39m Save Adversarial Similarity: Adversarial Similarity - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_test.py:531\u001b[39m\n",
      "2023/09/26 17:33:46 # \u001b[94mDEBUG\u001b[39m Folder data/temp/mvt-4c3821f7-7468-454b-bf32-0ede503ca1ca/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:69\u001b[39m\n",
      "2023/09/26 17:33:46 # \u001b[94mDEBUG\u001b[39m Folder data/artifacts/mvt-4c3821f7-7468-454b-bf32-0ede503ca1ca/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:82\u001b[39m\n",
      "2023/09/26 17:33:46 # \u001b[92mINFO\u001b[39m Adversarial Similarity 'Adversarial Similarity': Running\n",
      "2023/09/26 17:33:47 # \u001b[94mDEBUG\u001b[39m Uploading adv-mvt-4c3821f7-7468-454b-bf32-0ede503ca1ca to worker instance - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:664\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Complete 100.00% - [4/4] Computed stats for column BILL_AMT6\n",
      "INFO - Experiment launched at: http://localhost:12345/#/experiment?key=171de292-5c82-11ee-9192-00e04c68003f\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:36:41 # \u001b[92mINFO\u001b[39m Adversarial Similarity 'Adversarial Similarity': Processing results\n",
      "2023/09/26 17:36:41 # \u001b[94mDEBUG\u001b[39m Folder data/temp/7a52a2bd-13d0-4a55-8237/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:69\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Downloaded 'data/temp/7a52a2bd-13d0-4a55-8237/train_preds_train.csv'\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:36:42 # \u001b[94mDEBUG\u001b[39m Folder data/temp/7a52a2bd-13d0-4a55-8237/ deleted - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:227\u001b[39m\n",
      "2023/09/26 17:36:42 # \u001b[94mDEBUG\u001b[39m Folder data/temp/2ab76e11-e81b-4736-a2e7/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:69\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Downloaded 'data/temp/2ab76e11-e81b-4736-a2e7/h2oai_experiment_summary_171de292-5c82-11ee-9192-00e04c68003f.zip'\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:36:42 # \u001b[94mDEBUG\u001b[39m Folder data/temp/2ab76e11-e81b-4736-a2e7/ deleted - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:227\u001b[39m\n",
      "2023/09/26 17:36:42 # \u001b[92mINFO\u001b[39m Adversarial Similarity 'Adversarial Similarity': Completed\n",
      "2023/09/26 17:36:42 # \u001b[94mDEBUG\u001b[39m Save Adversarial Similarity: Adversarial Similarity - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_test.py:531\u001b[39m\n",
      "2023/09/26 17:36:43 # \u001b[94mDEBUG\u001b[39m Worker cleanup - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:801\u001b[39m\n",
      "2023/09/26 17:36:43 # \u001b[94mDEBUG\u001b[39m Deleting temporary experiment 171de292-5c82-11ee-9192-00e04c68003f from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:803\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Driverless AI Server reported experiment 171de292-5c82-11ee-9192-00e04c68003f deleted.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:36:44 # \u001b[94mDEBUG\u001b[39m Deleting temporary dataset 162c074c-5c82-11ee-9192-00e04c68003f from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:808\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Driverless AI Server reported dataset 162c074c-5c82-11ee-9192-00e04c68003f deleted.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:36:44 # \u001b[94mDEBUG\u001b[39m Folder data/temp/mvt-4c3821f7-7468-454b-bf32-0ede503ca1ca/ deleted - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:227\u001b[39m\n",
      "2023/09/26 17:36:44 # \u001b[94mDEBUG\u001b[39m Folder data/artifacts/mvt-4c3821f7-7468-454b-bf32-0ede503ca1ca/ deleted - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:227\u001b[39m\n"
     ]
    }
   ],
   "source": [
    "# run the interpretation\n",
    "interpretation = interpret.run_interpretation(\n",
    "    dataset=primary_dataset_path,\n",
    "    testset=secondary_dataset_path,\n",
    "    model=None,\n",
    "    target_col=\"default payment next month\",\n",
    "    explainers = [\n",
    "        commons.ExplainerToRun(\n",
    "            explainer_id=AdversarialSimilarityExplainer.explainer_id(),\n",
    "            params={\n",
    "                AdversarialSimilarityExplainer.PARAM_WORKER: DAI_WORKER_CONNECTION.key,\n",
    "            },\n",
    "        ),\n",
    "    ],\n",
    "    results_location=\"results-adversarial\",\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "236e1778-eae5-4444-8168-f32b91ddf30d",
   "metadata": {},
   "outputs": [],
   "source": [
    "result = interpretation.get_explainer_result(\n",
    "    AdversarialSimilarityExplainer.explainer_id()\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "97bca1f4-efde-4aee-b7f3-b1f230e329c8",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "result.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "49e9a8b1-60ac-484a-a810-f83346a5b44c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div class='datatable'>\n",
       "  <table class='frame'>\n",
       "  <thead>\n",
       "    <tr class='colnames'><td class='row_index'></td><th>x</th><th>y_group_1</th><th>y_group_2</th></tr>\n",
       "    <tr class='coltypes'><td class='row_index'></td><td class='float' title='float64'>&#x25AA;&#x25AA;&#x25AA;&#x25AA;&#x25AA;&#x25AA;&#x25AA;&#x25AA;</td><td class='float' title='float64'>&#x25AA;&#x25AA;&#x25AA;&#x25AA;&#x25AA;&#x25AA;&#x25AA;&#x25AA;</td><td class='float' title='float64'>&#x25AA;&#x25AA;&#x25AA;&#x25AA;&#x25AA;&#x25AA;&#x25AA;&#x25AA;</td></tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr><td class='row_index'>0</td><td>0.0259993</td><td>0.5837</td><td>0.2</td></tr>\n",
       "    <tr><td class='row_index'>1</td><td>0.0380496</td><td>0.1214</td><td>0.068</td></tr>\n",
       "    <tr><td class='row_index'>2</td><td>0.0500998</td><td>0.0346</td><td>0.04</td></tr>\n",
       "    <tr><td class='row_index'>3</td><td>0.06215</td><td>0.1397</td><td>0.17</td></tr>\n",
       "    <tr><td class='row_index'>4</td><td>0.0742002</td><td>0.021</td><td>0.04</td></tr>\n",
       "    <tr><td class='row_index'>5</td><td>0.0862504</td><td>0.0171</td><td>0.03</td></tr>\n",
       "    <tr><td class='row_index'>6</td><td>0.0983006</td><td>0.0163</td><td>0.108</td></tr>\n",
       "    <tr><td class='row_index'>7</td><td>0.110351</td><td>0.0267</td><td>0.09</td></tr>\n",
       "    <tr><td class='row_index'>8</td><td>0.122401</td><td>0.0105</td><td>0.046</td></tr>\n",
       "    <tr><td class='row_index'>9</td><td>0.134451</td><td>0.008</td><td>0.052</td></tr>\n",
       "    <tr><td class='row_index'>10</td><td>0.146501</td><td>0.0034</td><td>0.022</td></tr>\n",
       "    <tr><td class='row_index'>11</td><td>0.158552</td><td>0.003</td><td>0.034</td></tr>\n",
       "    <tr><td class='row_index'>12</td><td>0.170602</td><td>0.002</td><td>0.008</td></tr>\n",
       "    <tr><td class='row_index'>13</td><td>0.182652</td><td>0.0019</td><td>0.006</td></tr>\n",
       "    <tr><td class='row_index'>14</td><td>0.194702</td><td>0.0033</td><td>0.022</td></tr>\n",
       "    <tr><td class='row_index'>15</td><td>0.206752</td><td>0.0012</td><td>0.012</td></tr>\n",
       "    <tr><td class='row_index'>16</td><td>0.218803</td><td>0.0025</td><td>0.012</td></tr>\n",
       "    <tr><td class='row_index'>17</td><td>0.230853</td><td>0.0013</td><td>0.012</td></tr>\n",
       "    <tr><td class='row_index'>18</td><td>0.242903</td><td>0</td><td>0</td></tr>\n",
       "    <tr><td class='row_index'>19</td><td>0.254953</td><td>0.001</td><td>0.018</td></tr>\n",
       "    <tr><td class='row_index'>20</td><td>0.267003</td><td>0</td><td>0</td></tr>\n",
       "    <tr><td class='row_index'>21</td><td>0.279054</td><td>0.0003</td><td>0</td></tr>\n",
       "    <tr><td class='row_index'>22</td><td>0.291104</td><td>0</td><td>0</td></tr>\n",
       "    <tr><td class='row_index'>23</td><td>0.303154</td><td>0.0008</td><td>0.008</td></tr>\n",
       "    <tr><td class='row_index'>24</td><td>0.315204</td><td>0.0003</td><td>0.002</td></tr>\n",
       "  </tbody>\n",
       "  </table>\n",
       "  <div class='footer'>\n",
       "    <div class='frame_dimensions'>25 rows &times; 3 columns</div>\n",
       "  </div>\n",
       "</div>\n"
      ],
      "text/plain": [
       "<Frame#7f7320a0e4e0 25x3>"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result.data()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "66563f96-e643-4eda-9cb7-e4c085c133fe",
   "metadata": {},
   "source": [
    "## Run H2O Model Validation explainers which use Driverless AI server\n",
    "\n",
    "Lets run a few H2O Model Validation based explainers which explain **remote** models and datasets hosted by a Driverless AI server."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "3f029ff4-be42-4a41-9842-5836e06e8830",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['h2o_sonar.explainers.drift_explainer.DriftDetectionExplainer',\n",
       " 'h2o_sonar.explainers.adversarial_similarity_explainer.AdversarialSimilarityExplainer',\n",
       " 'h2o_sonar.explainers.size_dependency_explainer.SizeDependencyExplainer',\n",
       " 'h2o_sonar.explainers.segment_performance_explainer.SegmentPerformanceExplainer',\n",
       " 'h2o_sonar.explainers.calibration_score_explainer.CalibrationScoreExplainer',\n",
       " 'h2o_sonar.explainers.backtesting_explainer.BacktestingExplainer']"
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# list H2O Model Validation explainers\n",
    "all_mv_explainers = [e.id for e in interpret.list_explainers() if \"h2o-model-validation\" in e.keywords]\n",
    "all_mv_explainers"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "10efde0a-d2dd-4f68-881c-99e9e3f3adf3",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'resource:connection:0d860575-9d4b-46e5-9870-97ebd73780a1:key:b78cb888-f658-11ed-9ecf-0242709d15f7'"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# (model) handle is a string that identifies REMOTE model in the H2O Sonar runtime\n",
    "str(ts_model_handle)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "id": "f21230cc-438a-4a83-97c1-d940a7cc694e",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'resource:connection:0d860575-9d4b-46e5-9870-97ebd73780a1:key:a407dd4c-f658-11ed-9ecf-0242709d15f7'"
      ]
     },
     "execution_count": 28,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "str(ts_dataset_handle)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "id": "01d9ba42-d95c-4f1c-815e-b479f5d640b0",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'resource:connection:0d860575-9d4b-46e5-9870-97ebd73780a1:key:a4077500-f658-11ed-9ecf-0242709d15f7'"
      ]
     },
     "execution_count": 29,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "str(ts_testset_handle)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "id": "a9f0ae70-7692-4771-bf25-2a220b042f5f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'0d860575-9d4b-46e5-9870-97ebd73780a1'"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "DAI_WORKER_CONNECTION.key"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "id": "06c879ed-6bb7-410f-a242-29a3a1e02953",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'h2o_host': 'localhost',\n",
       " 'h2o_port': 12349,\n",
       " 'h2o_auto_start': True,\n",
       " 'h2o_min_mem_size': '2G',\n",
       " 'h2o_max_mem_size': '4G',\n",
       " 'custom_explainers': [],\n",
       " 'look_and_feel': 'h2o_sonar',\n",
       " 'per_explainer_logger': True,\n",
       " 'create_html_representations': True,\n",
       " 'connections': [{'key': '0d860575-9d4b-46e5-9870-97ebd73780a1',\n",
       "   'connection_type': 'DRIVERLESS_AI',\n",
       "   'name': 'Local Driverless AI server',\n",
       "   'description': 'Driverless AI server running on the localhost.',\n",
       "   'auth_server_url': '',\n",
       "   'environment_url': '',\n",
       "   'server_url': 'http://localhost:12345',\n",
       "   'server_id': '',\n",
       "   'realm_name': '',\n",
       "   'client_id': '',\n",
       "   'token': '',\n",
       "   'token_use_type': '',\n",
       "   'username': 'h2oai',\n",
       "   'password': 'h2oai'}],\n",
       " 'licenses': []}"
      ]
     },
     "execution_count": 32,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "h2o_sonar_config.config.to_dict(encrypt=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "id": "ad225728-14c0-4da9-b10e-b0fd82154608",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:36:45 # \u001b[94mDEBUG\u001b[39m Save Backtesting: Backtesting - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_test.py:531\u001b[39m\n",
      "2023/09/26 17:36:45 # \u001b[94mDEBUG\u001b[39m Save Size Dependency: Size Dependency - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_test.py:531\u001b[39m\n",
      "2023/09/26 17:36:45 # \u001b[94mDEBUG\u001b[39m Save Segment Performance: Segment Performance - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_test.py:531\u001b[39m\n",
      "2023/09/26 17:36:45 # \u001b[33mWARNING\u001b[39m Singleton MVClient already initialized, ignoring: args=(), kwargs={'data_folder': 'results-all-mv-explainers/h2o-sonar/mli_experiment_c292bbf2-8aef-4ee7-86c4-c8c6381f70e9/tmp'} - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:296\u001b[39m\n",
      "2023/09/26 17:36:45 # \u001b[94mDEBUG\u001b[39m Selected database: test-db - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:255\u001b[39m\n",
      "2023/09/26 17:36:45 # \u001b[92mINFO\u001b[39m Initialize MVDatabase: test-db\n",
      "2023/09/26 17:36:45 # \u001b[94mDEBUG\u001b[39m SQLDatabase: results-drift/h2o-sonar/mli_experiment_ea02c00c-8cb5-4596-bde0-1d602a862c1e/tmp/test.sql_db.sqlite - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:187\u001b[39m\n",
      "2023/09/26 17:36:45 # \u001b[94mDEBUG\u001b[39m ObjectStorage: results-drift/h2o-sonar/mli_experiment_ea02c00c-8cb5-4596-bde0-1d602a862c1e/tmp/test.obj_storage.sqlite - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:188\u001b[39m\n",
      "2023/09/26 17:36:45 # \u001b[92mINFO\u001b[39m Local Platform already exists\n",
      "2023/09/26 17:36:45 # \u001b[94mDEBUG\u001b[39m Database cache is enabled - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:67\u001b[39m\n",
      "2023/09/26 17:36:45 # \u001b[94mDEBUG\u001b[39m Deleting cache entries that are older than 24 hours - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:68\u001b[39m\n",
      "2023/09/26 17:36:45 # \u001b[94mDEBUG\u001b[39m Worker set: <h2o_mv.platforms.driverless.platform.DriverlessPlatform object at 0x7f73b36a0250> - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:331\u001b[39m\n",
      "2023/09/26 17:36:45 # \u001b[94mDEBUG\u001b[39m Save credentials: DriverlessCredentials(address='http://localhost:12345', username='h2oai') - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:33\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "A service ('Driverless AI') is running on localhost:12345 and it is accessible\n",
      "A service ('Driverless AI') is running on localhost:12345 and it is accessible\n",
      "A service ('Driverless AI') is running on localhost:12345 and it is accessible\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:36:46 # \u001b[92mINFO\u001b[39m Adding connection to Driverless AI server 'http://localhost:12345' for user 'h2oai'\n",
      "2023/09/26 17:36:46 # \u001b[94mDEBUG\u001b[39m Save credentials: DriverlessCredentials(address='http://localhost:12345', username='h2oai') - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:33\u001b[39m\n",
      "2023/09/26 17:36:46 # \u001b[94mDEBUG\u001b[39m Platform with address 'http://localhost:12345' for user 'h2oai' already in DB - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/databases/sql_db.py:234\u001b[39m\n",
      "2023/09/26 17:36:47 # \u001b[94mDEBUG\u001b[39m Worker set: <h2o_mv.platforms.driverless.platform.DriverlessPlatform object at 0x7f73cde1d700> - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:331\u001b[39m\n",
      "2023/09/26 17:36:47 # \u001b[92mINFO\u001b[39m Connection for <class 'h2o_mv.platforms.driverless.platform.DriverlessPlatform'> not supported\n",
      "2023/09/26 17:36:47 # \u001b[94mDEBUG\u001b[39m Import dataset a407dd4c-f658-11ed-9ecf-0242709d15f7 from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:454\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Downloaded 'data/temp/000-walmart_tts_small_train.csv.1684509609.0011456.csv'\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:36:49 # \u001b[94mDEBUG\u001b[39m Import experiment b78cb888-f658-11ed-9ecf-0242709d15f7 from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:628\u001b[39m\n",
      "2023/09/26 17:36:49 # \u001b[94mDEBUG\u001b[39m Import dataset a407dd4c-f658-11ed-9ecf-0242709d15f7 from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:454\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Downloaded 'data/temp/000-walmart_tts_small_train.csv.1684509609.0011456.csv'\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:36:51 # \u001b[94mDEBUG\u001b[39m Dataset with platform_obj_key 'a407dd4c-f658-11ed-9ecf-0242709d15f7' already in DB - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/databases/sql_db.py:314\u001b[39m\n",
      "2023/09/26 17:36:51 # \u001b[94mDEBUG\u001b[39m Import dataset a4077500-f658-11ed-9ecf-0242709d15f7 from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:454\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Downloaded 'data/temp/000-walmart_tts_small_test.csv.1684509608.881174.csv'\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:36:52 # \u001b[94mDEBUG\u001b[39m Folder data/temp/61793d0c-b9ee-4884-94aa/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:69\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Downloaded 'data/temp/61793d0c-b9ee-4884-94aa/h2oai_experiment_summary_b78cb888-f658-11ed-9ecf-0242709d15f7.zip'\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:36:52 # \u001b[94mDEBUG\u001b[39m Folder data/temp/61793d0c-b9ee-4884-94aa/ deleted - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:227\u001b[39m\n",
      "2023/09/26 17:36:52 # \u001b[94mDEBUG\u001b[39m Save Backtesting: Backtesting - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_test.py:531\u001b[39m\n",
      "2023/09/26 17:36:52 # \u001b[94mDEBUG\u001b[39m Folder data/temp/mvt-611575f1-cf5b-4fc9-a0ff-5685538ed42f/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:69\u001b[39m\n",
      "2023/09/26 17:36:52 # \u001b[94mDEBUG\u001b[39m Folder data/artifacts/mvt-611575f1-cf5b-4fc9-a0ff-5685538ed42f/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:82\u001b[39m\n",
      "2023/09/26 17:36:53 # \u001b[92mINFO\u001b[39m Backtesting 'Backtesting': Running\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Downloaded 'data/temp/e1b7dfaa-3020-458e-a406.tmp'\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:36:56 # \u001b[94mDEBUG\u001b[39m Uploading bt-train:0-mvt-611575f1-cf5b-4fc9-a0ff-5685538ed42f to worker instance - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:664\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Complete 100.00% - [4/4] Computed stats for column sample_weight\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:36:58 # \u001b[94mDEBUG\u001b[39m Uploading bt-test:0-mvt-611575f1-cf5b-4fc9-a0ff-5685538ed42f to worker instance - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:664\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Complete 100.00% - [4/4] Computed stats for column sample_weight\n",
      "INFO - Experiment launched at: http://localhost:12345/#/experiment?key=88d181f0-5c82-11ee-9192-00e04c68003f\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:37:01 # \u001b[94mDEBUG\u001b[39m Uploading bt-train:1-mvt-611575f1-cf5b-4fc9-a0ff-5685538ed42f to worker instance - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:664\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Complete 100.00% - [4/4] Computed stats for column sample_weight\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:37:03 # \u001b[94mDEBUG\u001b[39m Uploading bt-test:1-mvt-611575f1-cf5b-4fc9-a0ff-5685538ed42f to worker instance - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:664\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Complete 100.00% - [4/4] Computed stats for column sample_weight\n",
      "INFO - Experiment launched at: http://localhost:12345/#/experiment?key=8c0f9028-5c82-11ee-9192-00e04c68003f\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:37:41 # \u001b[92mINFO\u001b[39m Backtesting 'Backtesting': Processing results\n",
      "2023/09/26 17:37:41 # \u001b[94mDEBUG\u001b[39m Folder data/temp/8be2e5a2-2323-4cbf-9dea/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:69\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Downloaded 'data/temp/8be2e5a2-2323-4cbf-9dea/h2oai_experiment_summary_8c0f9028-5c82-11ee-9192-00e04c68003f.zip'\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:37:41 # \u001b[94mDEBUG\u001b[39m Folder data/temp/8be2e5a2-2323-4cbf-9dea/ deleted - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:227\u001b[39m\n",
      "2023/09/26 17:37:41 # \u001b[92mINFO\u001b[39m Backtesting 'Backtesting': Completed\n",
      "2023/09/26 17:37:41 # \u001b[94mDEBUG\u001b[39m Save Backtesting: Backtesting - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_test.py:531\u001b[39m\n",
      "2023/09/26 17:37:41 # \u001b[94mDEBUG\u001b[39m Worker cleanup - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:801\u001b[39m\n",
      "2023/09/26 17:37:41 # \u001b[94mDEBUG\u001b[39m Deleting temporary experiment 8c0f9028-5c82-11ee-9192-00e04c68003f from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:803\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Driverless AI Server reported experiment 8c0f9028-5c82-11ee-9192-00e04c68003f deleted.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:37:43 # \u001b[94mDEBUG\u001b[39m Deleting temporary experiment 88d181f0-5c82-11ee-9192-00e04c68003f from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:803\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Driverless AI Server reported experiment 88d181f0-5c82-11ee-9192-00e04c68003f deleted.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:37:44 # \u001b[94mDEBUG\u001b[39m Deleting temporary dataset 89a7f96a-5c82-11ee-9192-00e04c68003f from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:808\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Driverless AI Server reported dataset 89a7f96a-5c82-11ee-9192-00e04c68003f deleted.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:37:45 # \u001b[94mDEBUG\u001b[39m Deleting temporary dataset 8b135330-5c82-11ee-9192-00e04c68003f from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:808\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Driverless AI Server reported dataset 8b135330-5c82-11ee-9192-00e04c68003f deleted.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:37:45 # \u001b[94mDEBUG\u001b[39m Deleting temporary dataset 866ecc1a-5c82-11ee-9192-00e04c68003f from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:808\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Driverless AI Server reported dataset 866ecc1a-5c82-11ee-9192-00e04c68003f deleted.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:37:45 # \u001b[94mDEBUG\u001b[39m Deleting temporary dataset 87db7e86-5c82-11ee-9192-00e04c68003f from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:808\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Driverless AI Server reported dataset 87db7e86-5c82-11ee-9192-00e04c68003f deleted.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:37:46 # \u001b[94mDEBUG\u001b[39m Folder data/temp/mvt-611575f1-cf5b-4fc9-a0ff-5685538ed42f/ deleted - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:227\u001b[39m\n",
      "2023/09/26 17:37:46 # \u001b[94mDEBUG\u001b[39m Folder data/artifacts/mvt-611575f1-cf5b-4fc9-a0ff-5685538ed42f/ deleted - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:227\u001b[39m\n",
      "2023/09/26 17:37:47 # \u001b[33mWARNING\u001b[39m Singleton MVClient already initialized, ignoring: args=(), kwargs={'data_folder': 'results-all-mv-explainers/h2o-sonar/mli_experiment_c292bbf2-8aef-4ee7-86c4-c8c6381f70e9/tmp'} - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:296\u001b[39m\n",
      "2023/09/26 17:37:47 # \u001b[94mDEBUG\u001b[39m Selected database: test-db - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:255\u001b[39m\n",
      "2023/09/26 17:37:47 # \u001b[92mINFO\u001b[39m Initialize MVDatabase: test-db\n",
      "2023/09/26 17:37:47 # \u001b[94mDEBUG\u001b[39m SQLDatabase: results-drift/h2o-sonar/mli_experiment_ea02c00c-8cb5-4596-bde0-1d602a862c1e/tmp/test.sql_db.sqlite - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:187\u001b[39m\n",
      "2023/09/26 17:37:47 # \u001b[94mDEBUG\u001b[39m ObjectStorage: results-drift/h2o-sonar/mli_experiment_ea02c00c-8cb5-4596-bde0-1d602a862c1e/tmp/test.obj_storage.sqlite - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:188\u001b[39m\n",
      "2023/09/26 17:37:47 # \u001b[92mINFO\u001b[39m Local Platform already exists\n",
      "2023/09/26 17:37:47 # \u001b[94mDEBUG\u001b[39m Database cache is enabled - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:67\u001b[39m\n",
      "2023/09/26 17:37:47 # \u001b[94mDEBUG\u001b[39m Deleting cache entries that are older than 24 hours - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:68\u001b[39m\n",
      "2023/09/26 17:37:47 # \u001b[94mDEBUG\u001b[39m Worker set: <h2o_mv.platforms.driverless.platform.DriverlessPlatform object at 0x7f73cde1d700> - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:331\u001b[39m\n",
      "2023/09/26 17:37:47 # \u001b[94mDEBUG\u001b[39m Save credentials: DriverlessCredentials(address='http://localhost:12345', username='h2oai') - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:33\u001b[39m\n",
      "2023/09/26 17:37:47 # \u001b[92mINFO\u001b[39m Adding connection to Driverless AI server 'http://localhost:12345' for user 'h2oai'\n",
      "2023/09/26 17:37:47 # \u001b[94mDEBUG\u001b[39m Save credentials: DriverlessCredentials(address='http://localhost:12345', username='h2oai') - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:33\u001b[39m\n",
      "2023/09/26 17:37:47 # \u001b[94mDEBUG\u001b[39m Platform with address 'http://localhost:12345' for user 'h2oai' already in DB - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/databases/sql_db.py:234\u001b[39m\n",
      "2023/09/26 17:37:49 # \u001b[94mDEBUG\u001b[39m Worker set: <h2o_mv.platforms.driverless.platform.DriverlessPlatform object at 0x7f73b37bd6a0> - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:331\u001b[39m\n",
      "2023/09/26 17:37:49 # \u001b[92mINFO\u001b[39m Connection for <class 'h2o_mv.platforms.driverless.platform.DriverlessPlatform'> not supported\n",
      "2023/09/26 17:37:49 # \u001b[94mDEBUG\u001b[39m Import dataset a407dd4c-f658-11ed-9ecf-0242709d15f7 from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:454\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Downloaded 'data/temp/000-walmart_tts_small_train.csv.1684509609.0011456.csv'\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:37:50 # \u001b[94mDEBUG\u001b[39m Dataset with platform_obj_key 'a407dd4c-f658-11ed-9ecf-0242709d15f7' already in DB - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/databases/sql_db.py:314\u001b[39m\n",
      "2023/09/26 17:37:50 # \u001b[94mDEBUG\u001b[39m Import dataset a4077500-f658-11ed-9ecf-0242709d15f7 from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:454\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Downloaded 'data/temp/000-walmart_tts_small_test.csv.1684509608.881174.csv'\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:37:52 # \u001b[94mDEBUG\u001b[39m Dataset with platform_obj_key 'a4077500-f658-11ed-9ecf-0242709d15f7' already in DB - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/databases/sql_db.py:314\u001b[39m\n",
      "2023/09/26 17:37:52 # \u001b[94mDEBUG\u001b[39m Import experiment b78cb888-f658-11ed-9ecf-0242709d15f7 from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:628\u001b[39m\n",
      "2023/09/26 17:37:52 # \u001b[94mDEBUG\u001b[39m Import dataset a407dd4c-f658-11ed-9ecf-0242709d15f7 from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:454\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Downloaded 'data/temp/000-walmart_tts_small_train.csv.1684509609.0011456.csv'\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:37:54 # \u001b[94mDEBUG\u001b[39m Dataset with platform_obj_key 'a407dd4c-f658-11ed-9ecf-0242709d15f7' already in DB - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/databases/sql_db.py:314\u001b[39m\n",
      "2023/09/26 17:37:54 # \u001b[94mDEBUG\u001b[39m Import dataset a4077500-f658-11ed-9ecf-0242709d15f7 from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:454\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Downloaded 'data/temp/000-walmart_tts_small_test.csv.1684509608.881174.csv'\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:37:55 # \u001b[94mDEBUG\u001b[39m Dataset with platform_obj_key 'a4077500-f658-11ed-9ecf-0242709d15f7' already in DB - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/databases/sql_db.py:314\u001b[39m\n",
      "2023/09/26 17:37:55 # \u001b[94mDEBUG\u001b[39m Folder data/temp/eea1a0db-99b5-42e0-8c8f/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:69\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Downloaded 'data/temp/eea1a0db-99b5-42e0-8c8f/h2oai_experiment_summary_b78cb888-f658-11ed-9ecf-0242709d15f7.zip'\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:37:56 # \u001b[94mDEBUG\u001b[39m Folder data/temp/eea1a0db-99b5-42e0-8c8f/ deleted - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:227\u001b[39m\n",
      "2023/09/26 17:37:56 # \u001b[94mDEBUG\u001b[39m Model with platform_obj_key 'b78cb888-f658-11ed-9ecf-0242709d15f7' already in DB - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/databases/sql_db.py:424\u001b[39m\n",
      "2023/09/26 17:37:56 # \u001b[94mDEBUG\u001b[39m Save Size Dependency: Size Dependency - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_test.py:531\u001b[39m\n",
      "2023/09/26 17:37:56 # \u001b[94mDEBUG\u001b[39m Folder data/temp/mvt-920752b2-07cb-4c8d-88b6-f7378ed23238/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:69\u001b[39m\n",
      "2023/09/26 17:37:56 # \u001b[94mDEBUG\u001b[39m Folder data/artifacts/mvt-920752b2-07cb-4c8d-88b6-f7378ed23238/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:82\u001b[39m\n",
      "2023/09/26 17:37:56 # \u001b[92mINFO\u001b[39m Size Dependency 'Size Dependency': Running\n",
      "2023/09/26 17:37:56 # \u001b[94mDEBUG\u001b[39m Get cached dataset: 000-walmart_tts_small_train.csv (mvid='ds-2752daf9-4593-4229-9614-3bb26d1cccdb', dataset_key='a407dd4c-f658-11ed-9ecf-0242709d15f7') - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:495\u001b[39m\n",
      "2023/09/26 17:37:57 # \u001b[94mDEBUG\u001b[39m Uploading sd-train:0-mvt-920752b2-07cb-4c8d-88b6-f7378ed23238 to worker instance - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:664\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Complete 100.00% - [4/4] Computed stats for column sample_weight\n",
      "INFO - Experiment launched at: http://localhost:12345/#/experiment?key=ac723f46-5c82-11ee-9192-00e04c68003f\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:38:01 # \u001b[94mDEBUG\u001b[39m Uploading sd-train:1-mvt-920752b2-07cb-4c8d-88b6-f7378ed23238 to worker instance - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:664\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Complete 100.00% - [4/4] Computed stats for column sample_weight\n",
      "INFO - Experiment launched at: http://localhost:12345/#/experiment?key=ae891ef8-5c82-11ee-9192-00e04c68003f\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:38:49 # \u001b[92mINFO\u001b[39m Size Dependency 'Size Dependency': Processing results\n",
      "2023/09/26 17:38:49 # \u001b[94mDEBUG\u001b[39m Folder data/temp/3aea884f-1cb3-48ad-a009/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:69\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Downloaded 'data/temp/3aea884f-1cb3-48ad-a009/h2oai_experiment_summary_ac723f46-5c82-11ee-9192-00e04c68003f.zip'\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:38:49 # \u001b[94mDEBUG\u001b[39m Folder data/temp/3aea884f-1cb3-48ad-a009/ deleted - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:227\u001b[39m\n",
      "2023/09/26 17:38:49 # \u001b[92mINFO\u001b[39m Size Dependency 'Size Dependency': Completed\n",
      "2023/09/26 17:38:49 # \u001b[94mDEBUG\u001b[39m Save Size Dependency: Size Dependency - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_test.py:531\u001b[39m\n",
      "2023/09/26 17:38:49 # \u001b[94mDEBUG\u001b[39m Worker cleanup - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:801\u001b[39m\n",
      "2023/09/26 17:38:49 # \u001b[94mDEBUG\u001b[39m Deleting temporary experiment ac723f46-5c82-11ee-9192-00e04c68003f from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:803\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Driverless AI Server reported experiment ac723f46-5c82-11ee-9192-00e04c68003f deleted.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:38:51 # \u001b[94mDEBUG\u001b[39m Deleting temporary experiment ae891ef8-5c82-11ee-9192-00e04c68003f from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:803\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Driverless AI Server reported experiment ae891ef8-5c82-11ee-9192-00e04c68003f deleted.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:38:52 # \u001b[94mDEBUG\u001b[39m Deleting temporary dataset ab34fc90-5c82-11ee-9192-00e04c68003f from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:808\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Driverless AI Server reported dataset ab34fc90-5c82-11ee-9192-00e04c68003f deleted.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:38:53 # \u001b[94mDEBUG\u001b[39m Deleting temporary dataset ad4f88d8-5c82-11ee-9192-00e04c68003f from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:808\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Driverless AI Server reported dataset ad4f88d8-5c82-11ee-9192-00e04c68003f deleted.\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:38:54 # \u001b[94mDEBUG\u001b[39m Folder data/temp/mvt-920752b2-07cb-4c8d-88b6-f7378ed23238/ deleted - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:227\u001b[39m\n",
      "2023/09/26 17:38:54 # \u001b[94mDEBUG\u001b[39m Folder data/artifacts/mvt-920752b2-07cb-4c8d-88b6-f7378ed23238/ deleted - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:227\u001b[39m\n",
      "2023/09/26 17:38:55 # \u001b[33mWARNING\u001b[39m Singleton MVClient already initialized, ignoring: args=(), kwargs={'data_folder': 'results-all-mv-explainers/h2o-sonar/mli_experiment_c292bbf2-8aef-4ee7-86c4-c8c6381f70e9/tmp'} - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:296\u001b[39m\n",
      "2023/09/26 17:38:55 # \u001b[94mDEBUG\u001b[39m Selected database: test-db - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:255\u001b[39m\n",
      "2023/09/26 17:38:55 # \u001b[92mINFO\u001b[39m Initialize MVDatabase: test-db\n",
      "2023/09/26 17:38:55 # \u001b[94mDEBUG\u001b[39m SQLDatabase: results-drift/h2o-sonar/mli_experiment_ea02c00c-8cb5-4596-bde0-1d602a862c1e/tmp/test.sql_db.sqlite - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:187\u001b[39m\n",
      "2023/09/26 17:38:55 # \u001b[94mDEBUG\u001b[39m ObjectStorage: results-drift/h2o-sonar/mli_experiment_ea02c00c-8cb5-4596-bde0-1d602a862c1e/tmp/test.obj_storage.sqlite - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:188\u001b[39m\n",
      "2023/09/26 17:38:55 # \u001b[92mINFO\u001b[39m Local Platform already exists\n",
      "2023/09/26 17:38:55 # \u001b[94mDEBUG\u001b[39m Database cache is enabled - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:67\u001b[39m\n",
      "2023/09/26 17:38:55 # \u001b[94mDEBUG\u001b[39m Deleting cache entries that are older than 24 hours - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:68\u001b[39m\n",
      "2023/09/26 17:38:55 # \u001b[94mDEBUG\u001b[39m Worker set: <h2o_mv.platforms.driverless.platform.DriverlessPlatform object at 0x7f73b37bd6a0> - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:331\u001b[39m\n",
      "2023/09/26 17:38:55 # \u001b[94mDEBUG\u001b[39m Save credentials: DriverlessCredentials(address='http://localhost:12345', username='h2oai') - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:33\u001b[39m\n",
      "2023/09/26 17:38:55 # \u001b[92mINFO\u001b[39m Adding connection to Driverless AI server 'http://localhost:12345' for user 'h2oai'\n",
      "2023/09/26 17:38:55 # \u001b[94mDEBUG\u001b[39m Save credentials: DriverlessCredentials(address='http://localhost:12345', username='h2oai') - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:33\u001b[39m\n",
      "2023/09/26 17:38:55 # \u001b[94mDEBUG\u001b[39m Platform with address 'http://localhost:12345' for user 'h2oai' already in DB - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/databases/sql_db.py:234\u001b[39m\n",
      "2023/09/26 17:38:57 # \u001b[94mDEBUG\u001b[39m Worker set: <h2o_mv.platforms.driverless.platform.DriverlessPlatform object at 0x7f73cdc9a730> - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:331\u001b[39m\n",
      "2023/09/26 17:38:57 # \u001b[92mINFO\u001b[39m Connection for <class 'h2o_mv.platforms.driverless.platform.DriverlessPlatform'> not supported\n",
      "2023/09/26 17:38:57 # \u001b[94mDEBUG\u001b[39m Import dataset a407dd4c-f658-11ed-9ecf-0242709d15f7 from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:454\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Downloaded 'data/temp/000-walmart_tts_small_train.csv.1684509609.0011456.csv'\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:38:59 # \u001b[94mDEBUG\u001b[39m Dataset with platform_obj_key 'a407dd4c-f658-11ed-9ecf-0242709d15f7' already in DB - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/databases/sql_db.py:314\u001b[39m\n",
      "2023/09/26 17:38:59 # \u001b[94mDEBUG\u001b[39m Import dataset a4077500-f658-11ed-9ecf-0242709d15f7 from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:454\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Downloaded 'data/temp/000-walmart_tts_small_test.csv.1684509608.881174.csv'\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:39:00 # \u001b[94mDEBUG\u001b[39m Dataset with platform_obj_key 'a4077500-f658-11ed-9ecf-0242709d15f7' already in DB - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/databases/sql_db.py:314\u001b[39m\n",
      "2023/09/26 17:39:00 # \u001b[94mDEBUG\u001b[39m Import experiment b78cb888-f658-11ed-9ecf-0242709d15f7 from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:628\u001b[39m\n",
      "2023/09/26 17:39:01 # \u001b[94mDEBUG\u001b[39m Import dataset a407dd4c-f658-11ed-9ecf-0242709d15f7 from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:454\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Downloaded 'data/temp/000-walmart_tts_small_train.csv.1684509609.0011456.csv'\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:39:03 # \u001b[94mDEBUG\u001b[39m Dataset with platform_obj_key 'a407dd4c-f658-11ed-9ecf-0242709d15f7' already in DB - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/databases/sql_db.py:314\u001b[39m\n",
      "2023/09/26 17:39:03 # \u001b[94mDEBUG\u001b[39m Import dataset a4077500-f658-11ed-9ecf-0242709d15f7 from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:454\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Downloaded 'data/temp/000-walmart_tts_small_test.csv.1684509608.881174.csv'\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:39:04 # \u001b[94mDEBUG\u001b[39m Dataset with platform_obj_key 'a4077500-f658-11ed-9ecf-0242709d15f7' already in DB - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/databases/sql_db.py:314\u001b[39m\n",
      "2023/09/26 17:39:04 # \u001b[94mDEBUG\u001b[39m Folder data/temp/6f1bb3c9-be39-4536-a430/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:69\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Downloaded 'data/temp/6f1bb3c9-be39-4536-a430/h2oai_experiment_summary_b78cb888-f658-11ed-9ecf-0242709d15f7.zip'\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:39:04 # \u001b[94mDEBUG\u001b[39m Folder data/temp/6f1bb3c9-be39-4536-a430/ deleted - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:227\u001b[39m\n",
      "2023/09/26 17:39:04 # \u001b[94mDEBUG\u001b[39m Model with platform_obj_key 'b78cb888-f658-11ed-9ecf-0242709d15f7' already in DB - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/databases/sql_db.py:424\u001b[39m\n",
      "2023/09/26 17:39:04 # \u001b[94mDEBUG\u001b[39m Save Segment Performance: Segment Performance - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_test.py:531\u001b[39m\n",
      "2023/09/26 17:39:04 # \u001b[94mDEBUG\u001b[39m Folder data/temp/mvt-41d6e90c-b8bd-4eb1-ba3c-1f11faa22670/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:69\u001b[39m\n",
      "2023/09/26 17:39:04 # \u001b[94mDEBUG\u001b[39m Folder data/artifacts/mvt-41d6e90c-b8bd-4eb1-ba3c-1f11faa22670/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:82\u001b[39m\n",
      "2023/09/26 17:39:05 # \u001b[92mINFO\u001b[39m Segment Performance 'Segment Performance': Running\n",
      "2023/09/26 17:39:05 # \u001b[94mDEBUG\u001b[39m Get cached dataset: 000-walmart_tts_small_train.csv (mvid='ds-2752daf9-4593-4229-9614-3bb26d1cccdb', dataset_key='a407dd4c-f658-11ed-9ecf-0242709d15f7') - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:495\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Complete\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:39:10 # \u001b[92mINFO\u001b[39m Segment Performance 'Segment Performance': Completed\n",
      "2023/09/26 17:39:10 # \u001b[94mDEBUG\u001b[39m Save Segment Performance: Segment Performance - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_test.py:531\u001b[39m\n",
      "2023/09/26 17:39:10 # \u001b[94mDEBUG\u001b[39m Folder data/temp/mvt-41d6e90c-b8bd-4eb1-ba3c-1f11faa22670/ deleted - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:227\u001b[39m\n",
      "2023/09/26 17:39:10 # \u001b[94mDEBUG\u001b[39m Folder data/artifacts/mvt-41d6e90c-b8bd-4eb1-ba3c-1f11faa22670/ deleted - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:227\u001b[39m\n"
     ]
    }
   ],
   "source": [
    "# run explainers that support REMOTE datasets / models in order to explain a regression experiment\n",
    "interpretation = interpret.run_interpretation(\n",
    "    dataset=ts_dataset_handle,\n",
    "    testset=ts_testset_handle,\n",
    "    model=ts_model_handle,\n",
    "    target_col=ts_target_col,\n",
    "    # schedule for run all MV explainers + specify parameters they need\n",
    "    explainers=[\n",
    "        commons.ExplainerToRun(\n",
    "            explainer_id=BacktestingExplainer.explainer_id(),\n",
    "            params={\n",
    "                BacktestingExplainer.PARAM_WORKER: DAI_WORKER_CONNECTION.key,\n",
    "                BacktestingExplainer.PARAM_TIME_COLUMN: ts_time_col,\n",
    "            },\n",
    "        ),\n",
    "        commons.ExplainerToRun(\n",
    "            explainer_id=SizeDependencyExplainer.explainer_id(),\n",
    "            params={\n",
    "                SizeDependencyExplainer.PARAM_WORKER: DAI_WORKER_CONNECTION.key,\n",
    "                SizeDependencyExplainer.PARAM_TIME_COLUMN: ts_time_col,\n",
    "            },\n",
    "        ),\n",
    "        commons.ExplainerToRun(\n",
    "            explainer_id=SegmentPerformanceExplainer.explainer_id(),\n",
    "            params={\n",
    "                SegmentPerformanceExplainer.PARAM_WORKER: DAI_WORKER_CONNECTION.key,\n",
    "            },\n",
    "        ),        \n",
    "    ],\n",
    "    results_location=\"results-all-mv-explainers\",\n",
    ")\n",
    "\n",
    "# HINT: Calibration Score explainer will be run later as it requires a CLASSIFICATION experiment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "id": "09760ca2-412f-46f8-bbec-e0e681ce8d23",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# open interpretation HTML report in web browser\n",
    "webbrowser.open(interpretation.result.get_html_report_location())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "id": "3a62ea80-00ec-4d54-9ea0-8a1d8d9dc46d",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['h2o_sonar.explainers.backtesting_explainer.BacktestingExplainer',\n",
       " 'h2o_sonar.explainers.size_dependency_explainer.SizeDependencyExplainer',\n",
       " 'h2o_sonar.explainers.segment_performance_explainer.SegmentPerformanceExplainer']"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "interpretation.get_scheduled_explainer_ids()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "id": "87963b3f-bd29-4852-a281-d119ffdd0604",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['h2o_sonar.explainers.backtesting_explainer.BacktestingExplainer',\n",
       " 'h2o_sonar.explainers.size_dependency_explainer.SizeDependencyExplainer',\n",
       " 'h2o_sonar.explainers.segment_performance_explainer.SegmentPerformanceExplainer']"
      ]
     },
     "execution_count": 36,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "interpretation.get_finished_explainer_ids()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "id": "593f3f80-3a3b-4eca-90b9-b863e269bfe4",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['h2o_sonar.explainers.backtesting_explainer.BacktestingExplainer',\n",
       " 'h2o_sonar.explainers.size_dependency_explainer.SizeDependencyExplainer',\n",
       " 'h2o_sonar.explainers.segment_performance_explainer.SegmentPerformanceExplainer']"
      ]
     },
     "execution_count": 37,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "interpretation.get_successful_explainer_ids()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "id": "c31a852d-304b-4035-a203-3cb00a8914c8",
   "metadata": {},
   "outputs": [],
   "source": [
    "result = interpretation.get_explainer_result(\n",
    "    SizeDependencyExplainer.explainer_id(),\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "id": "74de867d-39e9-4561-8895-c8e7b65bd498",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "result.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "id": "1f15182b-5950-44b3-bd2d-bdcbd0313ae6",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'Dept': {'73165': 0.0, '48805': 0.0},\n",
       " 'MarkDown1': {'73165': 0.0, '48805': 0.0},\n",
       " 'MarkDown2': {'73165': 0.0, '48805': 0.0},\n",
       " 'MarkDown3': {'73165': 0.0, '48805': 0.0},\n",
       " 'MarkDown4': {'73165': 0.0, '48805': 0.0},\n",
       " 'MarkDown5': {'73165': 0.0, '48805': 0.0024520653},\n",
       " 'Store': {'73165': 0.0, '48805': 0.0052142237},\n",
       " 'sample_weight': {'73165': 0.0242297749, '48805': 1.0},\n",
       " 'IsHoliday': {'73165': 0.0389024862, '48805': 1.0},\n",
       " 'Date': {'73165': 1.0, '48805': 1.0}}"
      ]
     },
     "execution_count": 40,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result.data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "id": "a9be6935-ee96-428d-a302-53bb167b9f82",
   "metadata": {},
   "outputs": [],
   "source": [
    "result = interpretation.get_explainer_result(\n",
    "    SegmentPerformanceExplainer.explainer_id()\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "id": "be350709-aceb-4085-944c-85d4855235df",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1440x1200 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "result.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "id": "083c9e37-cfcc-494f-83c5-21f9adb4ae5c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div class='datatable'>\n",
       "  <table class='frame'>\n",
       "  <thead>\n",
       "    <tr class='colnames'><td class='row_index'></td><th>value_1</th><th>Freq</th><th>Metric</th><th>value_2</th><th>var_1</th><th>var_2</th></tr>\n",
       "    <tr class='coltypes'><td class='row_index'></td><td class='str' title='str32'>&#x25AA;&#x25AA;&#x25AA;&#x25AA;</td><td class='float' title='float64'>&#x25AA;&#x25AA;&#x25AA;&#x25AA;&#x25AA;&#x25AA;&#x25AA;&#x25AA;</td><td class='float' title='float64'>&#x25AA;&#x25AA;&#x25AA;&#x25AA;&#x25AA;&#x25AA;&#x25AA;&#x25AA;</td><td class='str' title='str32'>&#x25AA;&#x25AA;&#x25AA;&#x25AA;</td><td class='str' title='str32'>&#x25AA;&#x25AA;&#x25AA;&#x25AA;</td><td class='str' title='str32'>&#x25AA;&#x25AA;&#x25AA;&#x25AA;</td></tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr><td class='row_index'>0</td><td>2010-09-17</td><td>0.00858334</td><td>16080</td><td>2010-09-17</td><td>Date</td><td>Date</td></tr>\n",
       "    <tr><td class='row_index'>1</td><td>2011-09-02</td><td>0.00856967</td><td>18075.6</td><td>2011-09-02</td><td>Date</td><td>Date</td></tr>\n",
       "    <tr><td class='row_index'>2</td><td>2011-09-09</td><td>0.00856967</td><td>18049.8</td><td>2011-09-09</td><td>Date</td><td>Date</td></tr>\n",
       "    <tr><td class='row_index'>3</td><td>2011-09-16</td><td>0.00858334</td><td>16347.4</td><td>2011-09-16</td><td>Date</td><td>Date</td></tr>\n",
       "    <tr><td class='row_index'>4</td><td>2011-10-14</td><td>0.00856967</td><td>16272.6</td><td>2011-10-14</td><td>Date</td><td>Date</td></tr>\n",
       "    <tr><td class='row_index'>5</td><td>Other</td><td>0.957124</td><td>19211.3</td><td>Other</td><td>Date</td><td>Date</td></tr>\n",
       "    <tr><td class='row_index'>6</td><td>2010-09-17</td><td>0.00184514</td><td>20805.9</td><td>1) 0.999 - 3.0</td><td>Date</td><td>Dept</td></tr>\n",
       "    <tr><td class='row_index'>7</td><td>2010-09-17</td><td>0.00181781</td><td>13502.1</td><td>2) 3.0 - 6.0</td><td>Date</td><td>Dept</td></tr>\n",
       "    <tr><td class='row_index'>8</td><td>2010-09-17</td><td>0.00184514</td><td>16127</td><td>3) 6.0 - 9.0</td><td>Date</td><td>Dept</td></tr>\n",
       "    <tr><td class='row_index'>9</td><td>2010-09-17</td><td>0.00184514</td><td>12324.2</td><td>4) 9.0 - 12.0</td><td>Date</td><td>Dept</td></tr>\n",
       "    <tr><td class='row_index'>10</td><td>2010-09-17</td><td>0.0012301</td><td>16356.6</td><td>5) 12.0 - 14.0</td><td>Date</td><td>Dept</td></tr>\n",
       "    <tr><td class='row_index'>11</td><td>2011-09-02</td><td>0.00184514</td><td>24187.6</td><td>1) 0.999 - 3.0</td><td>Date</td><td>Dept</td></tr>\n",
       "    <tr><td class='row_index'>12</td><td>2011-09-02</td><td>0.00180414</td><td>14575</td><td>2) 3.0 - 6.0</td><td>Date</td><td>Dept</td></tr>\n",
       "    <tr><td class='row_index'>13</td><td>2011-09-02</td><td>0.00184514</td><td>18506.3</td><td>3) 6.0 - 9.0</td><td>Date</td><td>Dept</td></tr>\n",
       "    <tr><td class='row_index'>14</td><td>2011-09-02</td><td>0.00184514</td><td>13829.7</td><td>4) 9.0 - 12.0</td><td>Date</td><td>Dept</td></tr>\n",
       "    <tr><td class='row_index'>&#x22EE;</td><td class='hellipsis'>&#x22EE;</td><td class='hellipsis'>&#x22EE;</td><td class='hellipsis'>&#x22EE;</td><td class='hellipsis'>&#x22EE;</td><td class='hellipsis'>&#x22EE;</td><td class='hellipsis'>&#x22EE;</td></tr>\n",
       "    <tr><td class='row_index'>334</td><td>4) 27.0 - 36.0</td><td>0.0151028</td><td>16763.2</td><td>5</td><td>Store</td><td>sample_weight</td></tr>\n",
       "    <tr><td class='row_index'>335</td><td>5) 36.0 - 45.0</td><td>0.183407</td><td>16224.6</td><td>1</td><td>Store</td><td>sample_weight</td></tr>\n",
       "    <tr><td class='row_index'>336</td><td>5) 36.0 - 45.0</td><td>0.0152942</td><td>17821.4</td><td>5</td><td>Store</td><td>sample_weight</td></tr>\n",
       "    <tr><td class='row_index'>337</td><td>1</td><td>0.923105</td><td>18956.6</td><td>1</td><td>sample_weight</td><td>sample_weight</td></tr>\n",
       "    <tr><td class='row_index'>338</td><td>5</td><td>0.0768947</td><td>20998.1</td><td>5</td><td>sample_weight</td><td>sample_weight</td></tr>\n",
       "  </tbody>\n",
       "  </table>\n",
       "  <div class='footer'>\n",
       "    <div class='frame_dimensions'>339 rows &times; 6 columns</div>\n",
       "  </div>\n",
       "</div>\n"
      ],
      "text/plain": [
       "<Frame#7f736308d600 339x6>"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result.data()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "id": "1969268a-a63d-4d5b-b7d5-75340341d60b",
   "metadata": {},
   "outputs": [],
   "source": [
    "result = interpretation.get_explainer_result(\n",
    "    BacktestingExplainer.explainer_id()\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "id": "5df0d4eb-b17f-4730-a48e-a236f076405a",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "result.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "id": "1bbaa74f-d1c3-408c-8ab5-61c682b65937",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'48755': {'Dept': 0.0,\n",
       "  'IsHoliday': 0.0,\n",
       "  'Store': 0.0,\n",
       "  'sample_weight': 0.3120111525,\n",
       "  'Date': 1.0}}"
      ]
     },
     "execution_count": 46,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result.data()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "a99fcd50-d374-4df0-84d7-1a6731f89c5c",
   "metadata": {},
   "source": [
    "## Calibration Score explainer\n",
    "\n",
    "Calibration Score explainer requires **classification** (in contrast to the explainers which were run in the previous section to explain a **regression** problem)."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "id": "1523b50a-29b4-42a7-959e-4a13ba952f65",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:39:30 # \u001b[94mDEBUG\u001b[39m Save Calibration Score: Calibration Score - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_test.py:531\u001b[39m\n",
      "2023/09/26 17:39:30 # \u001b[33mWARNING\u001b[39m Singleton MVClient already initialized, ignoring: args=(), kwargs={'data_folder': 'results-calibration-score/h2o-sonar/mli_experiment_ff0c6f6e-cdca-4098-bf85-aa8584955177/tmp'} - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:296\u001b[39m\n",
      "2023/09/26 17:39:30 # \u001b[94mDEBUG\u001b[39m Selected database: test-db - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:255\u001b[39m\n",
      "2023/09/26 17:39:30 # \u001b[92mINFO\u001b[39m Initialize MVDatabase: test-db\n",
      "2023/09/26 17:39:30 # \u001b[94mDEBUG\u001b[39m SQLDatabase: results-drift/h2o-sonar/mli_experiment_ea02c00c-8cb5-4596-bde0-1d602a862c1e/tmp/test.sql_db.sqlite - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:187\u001b[39m\n",
      "2023/09/26 17:39:30 # \u001b[94mDEBUG\u001b[39m ObjectStorage: results-drift/h2o-sonar/mli_experiment_ea02c00c-8cb5-4596-bde0-1d602a862c1e/tmp/test.obj_storage.sqlite - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:188\u001b[39m\n",
      "2023/09/26 17:39:30 # \u001b[92mINFO\u001b[39m Local Platform already exists\n",
      "2023/09/26 17:39:30 # \u001b[94mDEBUG\u001b[39m Database cache is enabled - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:67\u001b[39m\n",
      "2023/09/26 17:39:30 # \u001b[94mDEBUG\u001b[39m Deleting cache entries that are older than 24 hours - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:68\u001b[39m\n",
      "2023/09/26 17:39:30 # \u001b[94mDEBUG\u001b[39m Worker set: <h2o_mv.platforms.driverless.platform.DriverlessPlatform object at 0x7f73cdc9a730> - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:331\u001b[39m\n",
      "2023/09/26 17:39:30 # \u001b[94mDEBUG\u001b[39m Save credentials: DriverlessCredentials(address='http://localhost:12345', username='h2oai') - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:33\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "A service ('Driverless AI') is running on localhost:12345 and it is accessible\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:39:30 # \u001b[92mINFO\u001b[39m Adding connection to Driverless AI server 'http://localhost:12345' for user 'h2oai'\n",
      "2023/09/26 17:39:30 # \u001b[94mDEBUG\u001b[39m Save credentials: DriverlessCredentials(address='http://localhost:12345', username='h2oai') - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:33\u001b[39m\n",
      "2023/09/26 17:39:30 # \u001b[94mDEBUG\u001b[39m Platform with address 'http://localhost:12345' for user 'h2oai' already in DB - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/databases/sql_db.py:234\u001b[39m\n",
      "2023/09/26 17:39:32 # \u001b[94mDEBUG\u001b[39m Worker set: <h2o_mv.platforms.driverless.platform.DriverlessPlatform object at 0x7f7363f8a430> - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:331\u001b[39m\n",
      "2023/09/26 17:39:32 # \u001b[92mINFO\u001b[39m Connection for <class 'h2o_mv.platforms.driverless.platform.DriverlessPlatform'> not supported\n",
      "2023/09/26 17:39:32 # \u001b[94mDEBUG\u001b[39m Import dataset f330f964-ab83-11ed-aa8e-a0cec818dc16 from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:454\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Downloaded 'data/temp/creditcard.csv.1676281872.9771898.csv'\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:39:33 # \u001b[94mDEBUG\u001b[39m Import experiment 4b000748-fba7-11ed-8ea4-8c1d96f410ff from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:628\u001b[39m\n",
      "2023/09/26 17:39:34 # \u001b[94mDEBUG\u001b[39m Import dataset f330f964-ab83-11ed-aa8e-a0cec818dc16 from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:454\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Downloaded 'data/temp/creditcard.csv.1676281872.9771898.csv'\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:39:35 # \u001b[94mDEBUG\u001b[39m Dataset with platform_obj_key 'f330f964-ab83-11ed-aa8e-a0cec818dc16' already in DB - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/databases/sql_db.py:314\u001b[39m\n",
      "2023/09/26 17:39:35 # \u001b[94mDEBUG\u001b[39m Folder data/temp/8db7fd2a-ceb2-411b-ab34/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:69\u001b[39m\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "INFO - Downloaded 'data/temp/8db7fd2a-ceb2-411b-ab34/h2oai_experiment_summary_4b000748-fba7-11ed-8ea4-8c1d96f410ff.zip'\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:39:35 # \u001b[94mDEBUG\u001b[39m Folder data/temp/8db7fd2a-ceb2-411b-ab34/ deleted - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:227\u001b[39m\n",
      "2023/09/26 17:39:35 # \u001b[94mDEBUG\u001b[39m Save Calibration Score: Calibration Score - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_test.py:531\u001b[39m\n",
      "2023/09/26 17:39:35 # \u001b[94mDEBUG\u001b[39m Folder data/temp/mvt-f7cb0574-2afb-4899-9ff5-3ce2e2cb6087/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:69\u001b[39m\n",
      "2023/09/26 17:39:35 # \u001b[94mDEBUG\u001b[39m Folder data/artifacts/mvt-f7cb0574-2afb-4899-9ff5-3ce2e2cb6087/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:82\u001b[39m\n",
      "2023/09/26 17:39:36 # \u001b[92mINFO\u001b[39m Calibration Score 'Calibration Score': Running\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Complete\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "2023/09/26 17:39:40 # \u001b[92mINFO\u001b[39m Calibration Score 'Calibration Score': Completed\n",
      "2023/09/26 17:39:40 # \u001b[94mDEBUG\u001b[39m Save Calibration Score: Calibration Score - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_test.py:531\u001b[39m\n",
      "2023/09/26 17:39:40 # \u001b[94mDEBUG\u001b[39m Folder data/temp/mvt-f7cb0574-2afb-4899-9ff5-3ce2e2cb6087/ deleted - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:227\u001b[39m\n",
      "2023/09/26 17:39:40 # \u001b[94mDEBUG\u001b[39m Folder data/artifacts/mvt-f7cb0574-2afb-4899-9ff5-3ce2e2cb6087/ deleted - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:227\u001b[39m\n"
     ]
    }
   ],
   "source": [
    "# run the interpretation\n",
    "interpretation = interpret.run_interpretation(\n",
    "    dataset=m_dataset_handle,\n",
    "    model=m_model_handle,\n",
    "    target_col=m_target_col,\n",
    "    # schedule for run all MV explainers + specify parameters they need\n",
    "    explainers=[\n",
    "        commons.ExplainerToRun(\n",
    "            explainer_id=CalibrationScoreExplainer.explainer_id(),\n",
    "            params={\n",
    "                CalibrationScoreExplainer.PARAM_WORKER: DAI_WORKER_CONNECTION.key,\n",
    "            },\n",
    "        ),\n",
    "    ],\n",
    "    results_location=\"results-calibration-score\",\n",
    ")\n",
    "\n",
    "# HINT: Calibration Score explainer will be run later as it requires a CLASSIFICATION experiment"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "id": "30e4d25d-0e2f-4b1f-b67d-1cbbb39d3552",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# open interpretation HTML report in web browser\n",
    "webbrowser.open(interpretation.result.get_html_report_location())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "id": "739e1b6e-448f-4bfa-b06d-49f4f020ff13",
   "metadata": {},
   "outputs": [],
   "source": [
    "result = interpretation.get_explainer_result(\n",
    "    CalibrationScoreExplainer.explainer_id()\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "id": "6773b932-3b46-4057-83c9-5bfb14ff754c",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 640x480 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "result.plot()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "id": "fdc823a4-51f2-415c-80fd-ad0c82b43195",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "{'classes_labels': ['0', '1', '2', '3', '4', '5', '6'],\n",
       " 'classes_legends': {'0': 'EDUCATION.0',\n",
       "  '1': 'EDUCATION.1',\n",
       "  '2': 'EDUCATION.2',\n",
       "  '3': 'EDUCATION.3',\n",
       "  '4': 'EDUCATION.4',\n",
       "  '5': 'EDUCATION.5',\n",
       "  '6': 'EDUCATION.6'},\n",
       " 'plots_paths': ['results-calibration-score/h2o-sonar/mli_experiment_ff0c6f6e-cdca-4098-bf85-aa8584955177/explainer_h2o_sonar_explainers_calibration_score_explainer_CalibrationScoreExplainer_74573bc7-16c5-487e-a565-fa001c02eac1/work/calibration_curve_0.png',\n",
       "  'results-calibration-score/h2o-sonar/mli_experiment_ff0c6f6e-cdca-4098-bf85-aa8584955177/explainer_h2o_sonar_explainers_calibration_score_explainer_CalibrationScoreExplainer_74573bc7-16c5-487e-a565-fa001c02eac1/work/calibration_curve_1.png',\n",
       "  'results-calibration-score/h2o-sonar/mli_experiment_ff0c6f6e-cdca-4098-bf85-aa8584955177/explainer_h2o_sonar_explainers_calibration_score_explainer_CalibrationScoreExplainer_74573bc7-16c5-487e-a565-fa001c02eac1/work/calibration_curve_2.png',\n",
       "  'results-calibration-score/h2o-sonar/mli_experiment_ff0c6f6e-cdca-4098-bf85-aa8584955177/explainer_h2o_sonar_explainers_calibration_score_explainer_CalibrationScoreExplainer_74573bc7-16c5-487e-a565-fa001c02eac1/work/calibration_curve_3.png',\n",
       "  'results-calibration-score/h2o-sonar/mli_experiment_ff0c6f6e-cdca-4098-bf85-aa8584955177/explainer_h2o_sonar_explainers_calibration_score_explainer_CalibrationScoreExplainer_74573bc7-16c5-487e-a565-fa001c02eac1/work/calibration_curve_4.png',\n",
       "  'results-calibration-score/h2o-sonar/mli_experiment_ff0c6f6e-cdca-4098-bf85-aa8584955177/explainer_h2o_sonar_explainers_calibration_score_explainer_CalibrationScoreExplainer_74573bc7-16c5-487e-a565-fa001c02eac1/work/calibration_curve_5.png',\n",
       "  'results-calibration-score/h2o-sonar/mli_experiment_ff0c6f6e-cdca-4098-bf85-aa8584955177/explainer_h2o_sonar_explainers_calibration_score_explainer_CalibrationScoreExplainer_74573bc7-16c5-487e-a565-fa001c02eac1/work/calibration_curve_6.png'],\n",
       " 'data': {'0': {'brier_score': 0.00023394372737651364,\n",
       "   'calibration_curve': {'Target': ['0', '0', '0', '0', '0', '0'],\n",
       "    'prob_true': [4.168577264579599e-05, 1.0, 1.0, 1.0, 1.0, 1.0],\n",
       "    'prob_pred': [0.0018975235771166031,\n",
       "     0.17078932666666669,\n",
       "     0.274813025,\n",
       "     0.33970129000000004,\n",
       "     0.5294499500000001,\n",
       "     0.60520023]}},\n",
       "  '1': {'brier_score': 0.16110014195450903,\n",
       "   'calibration_curve': {'Target': ['1',\n",
       "     '1',\n",
       "     '1',\n",
       "     '1',\n",
       "     '1',\n",
       "     '1',\n",
       "     '1',\n",
       "     '1',\n",
       "     '1',\n",
       "     '1'],\n",
       "    'prob_true': [0.02574389836175192,\n",
       "     0.09498680738786279,\n",
       "     0.1849246231155779,\n",
       "     0.32599795291709316,\n",
       "     0.4733868569729408,\n",
       "     0.6404358353510896,\n",
       "     0.7366273798730735,\n",
       "     0.9158279963403476,\n",
       "     1.0,\n",
       "     1.0],\n",
       "    'prob_pred': [0.07172593633467063,\n",
       "     0.14804822925435335,\n",
       "     0.2511202867386934,\n",
       "     0.34945573730040963,\n",
       "     0.44801741561700625,\n",
       "     0.5466558755447947,\n",
       "     0.6487950818087036,\n",
       "     0.7382659015828005,\n",
       "     0.8268039368253969,\n",
       "     0.90072805]}},\n",
       "  '2': {'brier_score': 0.19774319659796083,\n",
       "   'calibration_curve': {'Target': ['2',\n",
       "     '2',\n",
       "     '2',\n",
       "     '2',\n",
       "     '2',\n",
       "     '2',\n",
       "     '2',\n",
       "     '2',\n",
       "     '2'],\n",
       "    'prob_true': [0.0,\n",
       "     0.016511867905056758,\n",
       "     0.1401360544217687,\n",
       "     0.25562483057739227,\n",
       "     0.4377665761923226,\n",
       "     0.5962504184800803,\n",
       "     0.7625843279709393,\n",
       "     0.9038461538461539,\n",
       "     1.0],\n",
       "    'prob_pred': [0.08570079768333334,\n",
       "     0.16363757540247661,\n",
       "     0.25424372374829934,\n",
       "     0.3513734156058554,\n",
       "     0.4523625681930974,\n",
       "     0.54923692184466,\n",
       "     0.6438439597457196,\n",
       "     0.734262823176923,\n",
       "     0.8187404943636363]}},\n",
       "  '3': {'brier_score': 0.10323703991438297,\n",
       "   'calibration_curve': {'Target': ['3',\n",
       "     '3',\n",
       "     '3',\n",
       "     '3',\n",
       "     '3',\n",
       "     '3',\n",
       "     '3',\n",
       "     '3',\n",
       "     '3'],\n",
       "    'prob_true': [0.028092608737770027,\n",
       "     0.1252999873689529,\n",
       "     0.3228376327769347,\n",
       "     0.41272965879265094,\n",
       "     0.5710955710955711,\n",
       "     0.7165178571428571,\n",
       "     0.9215686274509803,\n",
       "     1.0,\n",
       "     1.0],\n",
       "    'prob_pred': [0.06785811246391579,\n",
       "     0.14155374299734794,\n",
       "     0.24225255552731428,\n",
       "     0.34774164095800536,\n",
       "     0.4448828712354311,\n",
       "     0.5434934320758932,\n",
       "     0.6411001715196082,\n",
       "     0.7410142480327868,\n",
       "     0.8383238139285715]}},\n",
       "  '4': {'brier_score': 0.0027960851690444323,\n",
       "   'calibration_curve': {'Target': ['4', '4', '4', '4', '4', '4'],\n",
       "    'prob_true': [0.001837160751565762,\n",
       "     0.9444444444444444,\n",
       "     1.0,\n",
       "     1.0,\n",
       "     1.0,\n",
       "     1.0],\n",
       "    'prob_pred': [0.004683773584626303,\n",
       "     0.1515408047777778,\n",
       "     0.2387325681818182,\n",
       "     0.34578425909090915,\n",
       "     0.4457971857142858,\n",
       "     0.5425023]}},\n",
       "  '5': {'brier_score': 0.006712204660318849,\n",
       "   'calibration_curve': {'Target': ['5', '5', '5', '5', '5', '5', '5'],\n",
       "    'prob_true': [0.0052275008364001336,\n",
       "     0.8717948717948718,\n",
       "     1.0,\n",
       "     1.0,\n",
       "     1.0,\n",
       "     1.0,\n",
       "     1.0],\n",
       "    'prob_pred': [0.008901456397448996,\n",
       "     0.14605884664102564,\n",
       "     0.250982425,\n",
       "     0.3392365107692307,\n",
       "     0.4529326545454545,\n",
       "     0.5299739685714286,\n",
       "     0.67699438]}},\n",
       "  '6': {'brier_score': 0.0008094460582697336,\n",
       "   'calibration_curve': {'Target': ['6', '6', '6', '6', '6', '6', '6', '6'],\n",
       "    'prob_true': [0.000208620186089206, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0],\n",
       "    'prob_pred': [0.002580000844333881,\n",
       "     0.145560698,\n",
       "     0.23762778750000002,\n",
       "     0.3246545966666667,\n",
       "     0.43545038,\n",
       "     0.5143077,\n",
       "     0.6134635366666666,\n",
       "     0.722229515]}}}}"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "result.data()"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1d9d8616-11c4-4f67-8644-1421f224e9d5",
   "metadata": {},
   "source": [
    "# Interpretation: CLI"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "id": "0ee4d302-7132-4655-8b57-ca23c443259e",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "usage: h2o-sonar [-h] [--dataset DATASET] [--target-col TARGET_COL]\n",
      "                 [--results-location RESULTS_LOCATION] [--model MODEL]\n",
      "                 [--validset VALIDSET] [--testset TESTSET]\n",
      "                 [--use_raw-features USE_RAW_FEATURES]\n",
      "                 [--weight-col WEIGHT_COL] [--drop-cols DROP_COLS]\n",
      "                 [--sample-num-rows SAMPLE_NUM_ROWS]\n",
      "                 [--used-features USED_FEATURES] [--model-type {pickle,mojo}]\n",
      "                 [--explainer EXPLAINER] [--explainers EXPLAINERS]\n",
      "                 [--all-explainers] [--explainers-pars EXPLAINERS_PARS]\n",
      "                 [--config-path CONFIG_PATH] [--config-type CONFIG_TYPE]\n",
      "                 [--config-value CONFIG_VALUE]\n",
      "                 [--encryption-key ENCRYPTION_KEY] [-d]\n",
      "                 [--args-as-json-location ARGS_AS_JSON_LOCATION]\n",
      "                 [--log-level {error,warning,info,debug}]\n",
      "                 action entity\n",
      "\n",
      "H2O Sonar Python library for Responsible AI.\n",
      "\n",
      "H2O Sonar is Python package that enables a holistic, low-risk, human-interpretable,\n",
      "fair, and trustable approach to machine learning by implementing various facets of\n",
      "Responsible AI.\n",
      "\n",
      "model, dataset, validset or testset handle schema:\n",
      "\n",
      "  resource:connection:\"<connection_configuration_name>\":key:<resource_key>\n",
      "  [:version:<resource_version>]\n",
      "\n",
      "optional arguments per action and entity:\n",
      "\n",
      "  show version:\n",
      "                      show H2O Sonar version\n",
      "\n",
      "  add config:\n",
      "    --config-path     path to JSon or TOML file with H2O Sonar config to be changed\n",
      "    --config-type     config item type: 'CONNECTION' or 'LICENSE'\n",
      "    --config-value    config item value (serialized as JSon) to add to the config file\n",
      "    --encryption-key  secret key to encrypt config fields with sensitive data\n",
      "                      (alternatively set H2O_SONAR_ENCRYPTION_KEY environment variable)\n",
      "\n",
      "  show config:\n",
      "    --config-path     path to JSon or TOML file with H2O Sonar config\n",
      "    --encryption-key  optional secret key to decrypt config fields with sensitive data\n",
      "                      (alternatively set H2O_SONAR_ENCRYPTION_KEY environment variable)\n",
      "\n",
      "  list explainers:\n",
      "    --detailed        show detailed descriptors (only IDs are shown by default)\n",
      "    --args-as-json-location\n",
      "                      optional JSon file which overrides filtering CLI arguments\n",
      "\n",
      "  describe explainer:\n",
      "    --explainer       explainer ID\n",
      "\n",
      "  run interpretation:\n",
      "    --dataset         path to dataset\n",
      "    --target-col      target column\n",
      "    --model           path to the serialized model, URL or locator\n",
      "    --results-location\n",
      "                      optional path to the interpretation results location (directory)\n",
      "    --validset        optional path to validation dataset\n",
      "    --testset         optional path to test dataset\n",
      "    --use_raw_features\n",
      "                      force the use of transformed features in surrogate models\n",
      "                      with 'false', by default the original (raw) features are used\n",
      "    --weight-col      optional dataset column name with examples weights\n",
      "    --drop-cols       optional list of dataset columns to drop\n",
      "    --sample-num-rows\n",
      "                      optional number of rows to sample from dataset (default: sample\n",
      "                      based on the RAM size, 0 do not sample, >0 sample to the specified\n",
      "                      number of rows)\n",
      "    --all-explainers  run all explainers (only the most important are run by default)\n",
      "    --used-features   optional comma separated list of features used by the model\n",
      "    --model-type      optional model type: 'pickle' or 'mojo'\n",
      "    --explainers      optional comma separated list of explainer IDs to be run\n",
      "    --explainers-pars optional dictionary with explainer parameters\n",
      "    --config-path     path to JSon or TOML file with H2O Sonar config to be changed\n",
      "    --args-as-json-location\n",
      "                      optional JSon file which overrides CLI arguments\n",
      "    --log-level       optional log level: 'error', 'warning', 'info', 'debug'\n",
      "\n",
      "  list interpretations:\n",
      "    --results-location\n",
      "                      path to directory, URL, location of interpretation results\n",
      "    --log-level       optional log level: 'error', 'warning', 'info', 'debug'\n",
      "\n",
      "positional arguments:\n",
      "  action                action to take: 'list', 'run' or 'describe'\n",
      "  entity                entity on which to perform the action:\n",
      "                        'interpretation'(s) or 'explainer'(s)\n",
      "\n",
      "optional arguments:\n",
      "  -h, --help            show this help message and exit\n",
      "  --dataset DATASET     location of the dataset\n",
      "  --target-col TARGET_COL\n",
      "                        target column\n",
      "  --results-location RESULTS_LOCATION\n",
      "                        location where to store the interpretation results\n",
      "  --model MODEL         location of the model\n",
      "  --validset VALIDSET   location of the validation dataset\n",
      "  --testset TESTSET     location of the test dataset\n",
      "  --use_raw-features USE_RAW_FEATURES\n",
      "                        force the use of transformed features in surrogate\n",
      "                        models with `false`\n",
      "  --weight-col WEIGHT_COL\n",
      "                        optional dataset column name with examples weights\n",
      "  --drop-cols DROP_COLS\n",
      "                        optional list of dataset columns to drop\n",
      "  --sample-num-rows SAMPLE_NUM_ROWS\n",
      "                        optional number of rows to sample from the dataset\n",
      "  --used-features USED_FEATURES\n",
      "                        optional comma separated list of features used by the\n",
      "                        model\n",
      "  --model-type {pickle,mojo}\n",
      "                        model type: 'pickle' (.pkl) or 'mojo' (.mojo)\n",
      "  --explainer EXPLAINER\n",
      "                        ID of the explainer to describe\n",
      "  --explainers EXPLAINERS\n",
      "                        comma separated list of explainer IDs to be run (only\n",
      "                        the most important explainers are run by default)\n",
      "  --all-explainers      run all explainers (only the most important explainers\n",
      "                        are run by default)\n",
      "  --explainers-pars EXPLAINERS_PARS\n",
      "                        optional dictionary with explainer parameters - the\n",
      "                        dictionary key is explainer ID and value is dictionary\n",
      "                        with parameters; parameter dictionary has parameter\n",
      "                        name as the key and parameter value as the value\n",
      "  --config-path CONFIG_PATH\n",
      "                        path to JSon or TOML file with H2O Sonar configuration\n",
      "                        to be used to override defaults - specify only items\n",
      "                        you want to change (please refer to\n",
      "                        h2o_sonar.config.H2oSonarConfig for more details)\n",
      "  --config-type CONFIG_TYPE\n",
      "                        configuration item type - 'CONNECTION' or 'LICENSE'\n",
      "  --config-value CONFIG_VALUE\n",
      "                        configuration item value represented either as\n",
      "                        dictionary or as string with JSon serialization of the\n",
      "                        configuration item - it is expected that the config\n",
      "                        item is NOT encrypted\n",
      "  --encryption-key ENCRYPTION_KEY\n",
      "                        encryption key to be used for encrypting/decrypting\n",
      "                        sensitive data in the configuration. If not specified,\n",
      "                        shell environment variable H2O_SONAR_ENCRYPTION_KEY\n",
      "                        with the encryption key is used.\n",
      "  -d, --detailed        show detailed descriptors (only IDs are shown by\n",
      "                        default)\n",
      "  --args-as-json-location ARGS_AS_JSON_LOCATION\n",
      "                        location of the JSon file with all command arguments\n",
      "                        (replacing command line arguments) allowing to load\n",
      "                        them from the filesystem\n",
      "  --log-level {error,warning,info,debug}\n",
      "                        log level\n",
      "\n",
      "examples:\n",
      "\n",
      "  h2o-sonar --help\n",
      "  h2o-sonar show version\n",
      "  h2o-sonar list explainers\n",
      "  h2o-sonar list explainers --detailed\n",
      "  h2o-sonar describe explainer\n",
      "    --explainer=h2o_sonar.explainers.dia_explainer.DiaExplainer\n",
      "  h2o-sonar run interpretation\n",
      "    --dataset=dataset.csv\n",
      "    --target-col=PROFIT\n",
      "    --results-location=/home/user/results\n",
      "    --model=model.pickle\n",
      "    --all-explainers\n",
      "  h2o-sonar run interpretation\n",
      "    --dataset=dataset.csv\n",
      "    --target-col=PROFIT\n",
      "    --results-location=/home/user/results\n",
      "    --model=model.pickle\n",
      "    --used-features=FEATURE_1,FEATURE_2,FEATURE_3\n",
      "    --explainers=h2o_sonar.explainers.dia_explainer.DiaExplainer\n",
      "    --explainers-pars=\n",
      "      \"{'h2o_sonar.explainers.dia_explainer.DiaExplainer':{'cut_off': 0.5}}\"\n",
      "    --drop_cols=COLUMN_1,COLUMN_2,COLUMN_3\n",
      "  h2o-sonar run interpretation\n",
      "    --args-as-json-location=h2o-sonar-args.json\n",
      "  h2o-sonar list interpretations --results-location=/home/user/results\n",
      "\n",
      "H2O Sonar JSon configuration example:\n",
      "  {\n",
      "    \"h2o_host\": \"192.168.1.210\",\n",
      "    \"h2o_port\": 57561,\n",
      "    \"h2o_auto_start\": true\n",
      "  }\n",
      "\n",
      "Interpretation arguments JSon file example - see interpret.py::run_interpretation():\n",
      "  {\n",
      "    \"dataset\": \"dataset.csv\",\n",
      "    \"model\": \"model.pickle\",\n",
      "    \"target_col\": \"PROFIT\",\n",
      "    \"results_location\": \"./results\"\n",
      "  }\n",
      "\n",
      "Explainer listing arguments JSon file example - see interpret.py::list_explainers():\n",
      "  {\n",
      "    \"experiment_types\": [\"regression\"],\n",
      "    \"explanation_scopes\": [\"local_scope\"],\n",
      "    \"keywords\": [\"explains-fairness\"],\n",
      "    \"explainer_filter\": [{\"filter_by\": \"filter-name\", \"value\": \"v\"}]\n",
      "  }\n"
     ]
    }
   ],
   "source": [
    "# check H2O Sonar command line interface (CLI) help\n",
    "!h2o-sonar --help"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "id": "b919d633-d64e-4f18-ac6c-e98037f6bb8c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{\n",
      "  \"explainers\": [\n",
      "    \"h2o_sonar.explainers.fi_naive_shapley_explainer.NaiveShapleyMojoFeatureImportanceExplainer\",\n",
      "    \"h2o_sonar.explainers.summary_shap_explainer.SummaryShapleyExplainer\",\n",
      "    \"h2o_sonar.explainers.dt_surrogate_explainer.DecisionTreeSurrogateExplainer\",\n",
      "    \"h2o_sonar.explainers.pd_ice_explainer.PdIceExplainer\",\n",
      "    \"h2o_sonar.explainers.dia_explainer.DiaExplainer\",\n",
      "    \"h2o_sonar.explainers.transformed_fi_shapley_explainer.ShapleyMojoTransformedFeatureImportanceExplainer\",\n",
      "    \"h2o_sonar.explainers.residual_dt_surrogate_explainer.ResidualDecisionTreeSurrogateExplainer\",\n",
      "    \"h2o_sonar.explainers.residual_pd_ice_explainer.ResidualPdIceExplainer\",\n",
      "    \"h2o_sonar.explainers.fi_kernel_shap_explainer.KernelShapFeatureImportanceExplainer\",\n",
      "    \"h2o_sonar.explainers.pd_2_features_explainer.PdFor2FeaturesExplainer\",\n",
      "    \"h2o_sonar.explainers.friedman_h_statistic_explainer.FriedmanHStatisticExplainer\",\n",
      "    \"h2o_sonar.explainers.morris_sa_explainer.MorrisSensitivityAnalysisExplainer\",\n",
      "    \"h2o_sonar.explainers.dataset_and_model_insights_explainer.DatasetAndModelInsightsExplainer\",\n",
      "    \"h2o_sonar.explainers.drift_explainer.DriftDetectionExplainer\",\n",
      "    \"h2o_sonar.explainers.adversarial_similarity_explainer.AdversarialSimilarityExplainer\",\n",
      "    \"h2o_sonar.explainers.size_dependency_explainer.SizeDependencyExplainer\",\n",
      "    \"h2o_sonar.explainers.segment_performance_explainer.SegmentPerformanceExplainer\",\n",
      "    \"h2o_sonar.explainers.calibration_score_explainer.CalibrationScoreExplainer\",\n",
      "    \"h2o_sonar.explainers.backtesting_explainer.BacktestingExplainer\"\n",
      "  ]\n",
      "}\n"
     ]
    }
   ],
   "source": [
    "# list explainers\n",
    "!h2o-sonar list explainers"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "c6d5f82e-fc08-490d-ac59-4d1fb9fe37e1",
   "metadata": {},
   "source": [
    "## Run Drift Detection explainer using CLI"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "id": "ba320bf0-6e80-4c24-9254-71c23e81ab13",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'../../data/creditcard.csv'"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "primary_dataset_path"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 55,
   "id": "da4ddf1e-1384-49ee-9b23-c8e73dbdabb9",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'/tmp/secondary_dataset.csv'"
      ]
     },
     "execution_count": 55,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "secondary_dataset_path"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "id": "391989aa-9a1e-4bbd-b713-7af20194eea7",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Stratified/random sampler: loading the original dataset '../../data/creditcard.csv' for sampling...\n",
      "Stratified/random sampler:   -> did NO sampling as the sampling limit is smaller than the number of rows in the dataset: 10000 <= 25000\n",
      "Stratified/random sampler: loading the original dataset '/tmp/secondary_dataset.csv' for sampling...\n",
      "Stratified/random sampler:   -> did NO sampling as the sampling limit is smaller than the number of rows in the dataset: 500 <= 25000\n",
      "2023/09/26 17:39:47 # \u001b[94mDEBUG\u001b[39m H2O Model Validation LogLevel: DEBUG - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/logging.py:97\u001b[39m\n",
      "2023/09/26 17:39:48 # \u001b[33mWARNING\u001b[39m IoC/DI: No implementations found for >>IMVDatabase<< - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/ioc.py:35\u001b[39m\n",
      "2023/09/26 17:39:48 # \u001b[91mERROR\u001b[39m Can't save test Drift Detection: no database connection - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_test.py:529\u001b[39m\n",
      "2023/09/26 17:39:48 # \u001b[94mDEBUG\u001b[39m Selected database: test-db - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:255\u001b[39m\n",
      "2023/09/26 17:39:48 # \u001b[92mINFO\u001b[39m Initialize MVDatabase: test-db\n",
      "2023/09/26 17:39:48 # \u001b[94mDEBUG\u001b[39m SQLDatabase: results-drift-cli/h2o-sonar/mli_experiment_532e030e-57fa-43a2-82ab-20f641f071ad/tmp/test.sql_db.sqlite - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:187\u001b[39m\n",
      "2023/09/26 17:39:48 # \u001b[94mDEBUG\u001b[39m ObjectStorage: results-drift-cli/h2o-sonar/mli_experiment_532e030e-57fa-43a2-82ab-20f641f071ad/tmp/test.obj_storage.sqlite - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:188\u001b[39m\n",
      "2023/09/26 17:39:48 # \u001b[92mINFO\u001b[39m LocalPlatform local-platform created\n",
      "2023/09/26 17:39:48 # \u001b[94mDEBUG\u001b[39m Database cache is enabled - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:67\u001b[39m\n",
      "2023/09/26 17:39:48 # \u001b[94mDEBUG\u001b[39m Deleting cache entries that are older than 24 hours - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:68\u001b[39m\n",
      "2023/09/26 17:39:48 # \u001b[94mDEBUG\u001b[39m Import dataset obj-99ae974d-bd4b-4580-8d0f-7c17396905fb from LocalPlatform - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/local/platform.py:64\u001b[39m\n",
      "2023/09/26 17:39:48 # \u001b[94mDEBUG\u001b[39m Import dataset obj-0f44a792-d6e8-4f84-a495-be592eebb96c from LocalPlatform - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/local/platform.py:64\u001b[39m\n",
      "2023/09/26 17:39:48 # \u001b[94mDEBUG\u001b[39m Save Drift Detection: Drift Detection - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_test.py:531\u001b[39m\n",
      "2023/09/26 17:39:48 # \u001b[94mDEBUG\u001b[39m Folder data/temp/mvt-d505b3c7-f16a-4d0d-9667-2f4c0d2ad957/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:69\u001b[39m\n",
      "2023/09/26 17:39:48 # \u001b[94mDEBUG\u001b[39m Folder data/artifacts/mvt-d505b3c7-f16a-4d0d-9667-2f4c0d2ad957/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:82\u001b[39m\n",
      "2023/09/26 17:39:48 # \u001b[92mINFO\u001b[39m Drift Detection 'Drift Detection': Running\n",
      "\u001b[38;5;247mic\u001b[39m\u001b[38;5;245m|\u001b[39m\u001b[38;5;245m \u001b[39m\u001b[38;5;36m'\u001b[39m\u001b[38;5;36mCreating feature distribution\u001b[39m\u001b[38;5;36m'\u001b[39m\n",
      "100%|██████████████████████████████████████████| 25/25 [00:00<00:00, 449.32it/s]\n",
      "100%|███████████████████████████████████████████| 25/25 [00:00<00:00, 56.45it/s]\n",
      "2023/09/26 17:39:49 # \u001b[92mINFO\u001b[39m Drift Detection 'Drift Detection': Completed\n",
      "2023/09/26 17:39:49 # \u001b[94mDEBUG\u001b[39m Save Drift Detection: Drift Detection - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_test.py:531\u001b[39m\n",
      "2023/09/26 17:39:49 # \u001b[94mDEBUG\u001b[39m Folder data/temp/mvt-d505b3c7-f16a-4d0d-9667-2f4c0d2ad957/ deleted - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:227\u001b[39m\n",
      "2023/09/26 17:39:49 # \u001b[94mDEBUG\u001b[39m Folder data/artifacts/mvt-d505b3c7-f16a-4d0d-9667-2f4c0d2ad957/ deleted - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:227\u001b[39m\n",
      "pandas.Int64Index is deprecated and will be removed from pandas in a future version. Use pandas.Index with the appropriate dtype instead.\n",
      "\n",
      "Interpretation FINISHED.\n",
      "  Results directory:\n",
      "    file:///home/dvorka/h/mli/git/h2o-sonar/examples/explainers/results-drift-cli\n",
      "  Interpretations index:\n",
      "    file:///home/dvorka/h/mli/git/h2o-sonar/examples/explainers/results-drift-cli/h2o-sonar.html\n",
      "  HTML report:\n",
      "    file:///home/dvorka/h/mli/git/h2o-sonar/examples/explainers/results-drift-cli/h2o-sonar.html\n",
      "\u001b[0m"
     ]
    }
   ],
   "source": [
    "# run Drift Detection explainer\n",
    "!h2o-sonar run interpretation \\\n",
    "  --dataset {primary_dataset_path} \\\n",
    "  --testset {secondary_dataset_path} \\\n",
    "  --target-col \"default payment next month\" \\\n",
    "  --explainers h2o_sonar.explainers.drift_explainer.DriftDetectionExplainer \\\n",
    "  --results-location results-drift-cli "
   ]
  },
  {
   "cell_type": "markdown",
   "id": "412142aa-1379-44c9-9996-04dfcddce21f",
   "metadata": {},
   "source": [
    "# Use CLI to run H2O MV based explainers\n",
    "\n",
    "Use command line interface to run H2O MV based explainers which depend on Driverless AI."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "id": "70c0cc3b-cf5d-4fa8-9713-65493f28651f",
   "metadata": {},
   "outputs": [],
   "source": [
    "# configure H2O Sonar by adding Driverless AI connection to the configuration\n",
    "h2o_sonar_config_path = \"/tmp/h2o-sonar-config.json\"\n",
    "\n",
    "# secret key to encrypt sensitive fields in the configuration\n",
    "encryption_key = \"my-demo-3ncr1pt10n-key\"\n",
    "\n",
    "# custom Driverless AI server connection ID\n",
    "dai_connection_key = \"demo-dai-connection\"\n",
    "\n",
    "# initialize configuration file with the DEFAULT H2O Sonar configuration\n",
    "!h2o-sonar show config > {h2o_sonar_config_path}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "id": "1cd4de63-1a62-4a99-9308-cd92d02607a4",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{\n",
      "  \"h2o_host\": \"localhost\",\n",
      "  \"h2o_port\": 12349,\n",
      "  \"h2o_auto_start\": true,\n",
      "  \"h2o_min_mem_size\": \"2G\",\n",
      "  \"h2o_max_mem_size\": \"4G\",\n",
      "  \"custom_explainers\": [],\n",
      "  \"look_and_feel\": \"h2o_sonar\",\n",
      "  \"per_explainer_logger\": true,\n",
      "  \"create_html_representations\": true,\n",
      "  \"connections\": [],\n",
      "  \"licenses\": []\n",
      "}\n"
     ]
    }
   ],
   "source": [
    "# verify the configuration content\n",
    "!h2o-sonar show config \\\n",
    "  --config-path {h2o_sonar_config_path} \\\n",
    "  --encryption-key {encryption_key}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "id": "43c8873c-8ac4-4ff5-98a3-4e52096c24bf",
   "metadata": {},
   "outputs": [],
   "source": [
    "!h2o-sonar add config \\\n",
    "  --config-type CONNECTION \\\n",
    "  --config-path \"/tmp/h2o-sonar-config.json\" \\\n",
    "  --encryption-key \"my-demo-3ncr1pt10n-key\" \\\n",
    "  --config-value \"{\\\"key\\\": \\\"demo-dai-connection\\\", \\\"connection_type\\\": \\\"DRIVERLESS_AI\\\", \\\"name\\\": \\\"Local Driverless AI server\\\", \\\"description\\\": \\\"Driverless AI server running on the localhost.\\\", \\\"auth_server_url\\\": \\\"\\\", \\\"server_url\\\": \\\"http://localhost:12345\\\", \\\"realm_name\\\": \\\"\\\", \\\"client_id\\\": \\\"\\\", \\\"token\\\": \\\"\\\", \\\"username\\\": \\\"h2oai\\\", \\\"password\\\": \\\"h2oai\\\"}\"\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "id": "915d49ab-fac7-4abe-9053-1554b1f1a64c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{\n",
      "  \"h2o_host\": \"localhost\",\n",
      "  \"h2o_port\": 12349,\n",
      "  \"h2o_auto_start\": true,\n",
      "  \"h2o_min_mem_size\": \"2G\",\n",
      "  \"h2o_max_mem_size\": \"4G\",\n",
      "  \"custom_explainers\": [],\n",
      "  \"look_and_feel\": \"h2o_sonar\",\n",
      "  \"per_explainer_logger\": true,\n",
      "  \"create_html_representations\": true,\n",
      "  \"connections\": [\n",
      "    {\n",
      "      \"key\": \"demo-dai-connection\",\n",
      "      \"connection_type\": \"DRIVERLESS_AI\",\n",
      "      \"name\": \"Local Driverless AI server\",\n",
      "      \"description\": \"Driverless AI server running on the localhost.\",\n",
      "      \"auth_server_url\": \"\",\n",
      "      \"environment_url\": \"\",\n",
      "      \"server_url\": \"http://localhost:12345\",\n",
      "      \"server_id\": \"\",\n",
      "      \"realm_name\": \"\",\n",
      "      \"client_id\": \"\",\n",
      "      \"token\": \"\",\n",
      "      \"token_use_type\": \"\",\n",
      "      \"username\": \"h2oai\",\n",
      "      \"password\": \"h2oai\"\n",
      "    }\n",
      "  ],\n",
      "  \"licenses\": []\n",
      "}\n"
     ]
    }
   ],
   "source": [
    "# verify the configuration content\n",
    "!h2o-sonar show config \\\n",
    "  --config-path {h2o_sonar_config_path} \\\n",
    "  --encryption-key {encryption_key}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 61,
   "id": "ff2a4f43-1a68-425c-9273-d6193072057f",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{\n",
      "    \"h2o_host\": \"localhost\",\n",
      "    \"h2o_port\": 12349,\n",
      "    \"h2o_auto_start\": true,\n",
      "    \"h2o_min_mem_size\": \"2G\",\n",
      "    \"h2o_max_mem_size\": \"4G\",\n",
      "    \"custom_explainers\": [],\n",
      "    \"look_and_feel\": \"h2o_sonar\",\n",
      "    \"per_explainer_logger\": true,\n",
      "    \"create_html_representations\": true,\n",
      "    \"connections\": [\n",
      "        {\n",
      "            \"key\": \"demo-dai-connection\",\n",
      "            \"connection_type\": \"DRIVERLESS_AI\",\n",
      "            \"name\": \"Local Driverless AI server\",\n",
      "            \"description\": \"Driverless AI server running on the localhost.\",\n",
      "            \"auth_server_url\": \"\",\n",
      "            \"environment_url\": \"\",\n",
      "            \"server_url\": \"http://localhost:12345\",\n",
      "            \"server_id\": \"\",\n",
      "            \"realm_name\": \"\",\n",
      "            \"client_id\": \"\",\n",
      "            \"token\": {\n",
      "                \"encrypted\": \"\"\n",
      "            },\n",
      "            \"token_use_type\": \"\",\n",
      "            \"username\": \"h2oai\",\n",
      "            \"password\": {\n",
      "                \"encrypted\": \"gAAAAABlEvtKI_IBN3Jppr5b4bJxBxJwFOzLWlW2vVF3z_SGQL9f9fFX9PIAuAibmZGnCheM6uo-sM8hqu2OyyqzFVB0ynNJng==\"\n",
      "            }\n",
      "        }\n",
      "    ],\n",
      "    \"licenses\": []\n",
      "}"
     ]
    }
   ],
   "source": [
    "# note that the configuration is ENCRYPTED on the filesystem and DECRYPTED in memory\n",
    "!cat {h2o_sonar_config_path}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "id": "d8f34f19-8c07-4d5c-9b96-60c994e6319d",
   "metadata": {},
   "outputs": [],
   "source": [
    "# if there are many CLI arguments and explainer parameters,\n",
    "# then it is convenient to prepare H2O Sonar arguments as JSon\n",
    "run_interpretation_args_path = \"/tmp/run_interpretation_args.json\"\n",
    "\n",
    "run_interpretation_args = {\n",
    "    \"model\": str(ts_model_handle),\n",
    "    \"dataset\": str(ts_dataset_handle),\n",
    "    \"testset\": str(ts_testset_handle),\n",
    "    \"target_col\": ts_target_col,\n",
    "    \"explainers\": [\n",
    "        {\n",
    "            \"id\": BacktestingExplainer.explainer_id(),\n",
    "            \"params\": {\n",
    "                \"worker_connection_key\": dai_connection_key,\n",
    "                \"time_col\": \"Date\",\n",
    "            },\n",
    "            \n",
    "        },\n",
    "        {\n",
    "            \"id\": SizeDependencyExplainer.explainer_id(),\n",
    "            \"params\": {\n",
    "                \"worker_connection_key\": dai_connection_key,\n",
    "                \"time_col\": \"Date\"\n",
    "            }\n",
    "        },\n",
    "        {\n",
    "            \"id\": SegmentPerformanceExplainer.explainer_id(),\n",
    "            \"params\": {\n",
    "                \"worker_connection_key\": dai_connection_key,\n",
    "            }\n",
    "        },\n",
    "    ]\n",
    "}\n",
    "\n",
    "# save interpretation arguments to a JSon file\n",
    "with open(run_interpretation_args_path, \"w\") as f:\n",
    "    json.dump(run_interpretation_args, f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "id": "0c909138-7ca7-46cd-a714-c7ad94737607",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "{\"model\": \"resource:connection:0d860575-9d4b-46e5-9870-97ebd73780a1:key:b78cb888-f658-11ed-9ecf-0242709d15f7\", \"dataset\": \"resource:connection:0d860575-9d4b-46e5-9870-97ebd73780a1:key:a407dd4c-f658-11ed-9ecf-0242709d15f7\", \"testset\": \"resource:connection:0d860575-9d4b-46e5-9870-97ebd73780a1:key:a4077500-f658-11ed-9ecf-0242709d15f7\", \"target_col\": \"Weekly_Sales\", \"explainers\": [{\"id\": \"h2o_sonar.explainers.backtesting_explainer.BacktestingExplainer\", \"params\": {\"worker_connection_key\": \"demo-dai-connection\", \"time_col\": \"Date\"}}, {\"id\": \"h2o_sonar.explainers.size_dependency_explainer.SizeDependencyExplainer\", \"params\": {\"worker_connection_key\": \"demo-dai-connection\", \"time_col\": \"Date\"}}, {\"id\": \"h2o_sonar.explainers.segment_performance_explainer.SegmentPerformanceExplainer\", \"params\": {\"worker_connection_key\": \"demo-dai-connection\"}}]}"
     ]
    }
   ],
   "source": [
    "!cat {run_interpretation_args_path}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "e9bfab76-0926-4a17-b1fb-4a7a36d14a1c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "A service ('Driverless AI') is running on localhost:12345 and it is accessible\n",
      "2023/09/26 17:39:58 # \u001b[94mDEBUG\u001b[39m H2O Model Validation LogLevel: DEBUG - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/logging.py:97\u001b[39m\n",
      "2023/09/26 17:39:58 # \u001b[33mWARNING\u001b[39m IoC/DI: No implementations found for >>IMVDatabase<< - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/ioc.py:35\u001b[39m\n",
      "2023/09/26 17:39:58 # \u001b[91mERROR\u001b[39m Can't save test Backtesting: no database connection - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_test.py:529\u001b[39m\n",
      "A service ('Driverless AI') is running on localhost:12345 and it is accessible\n",
      "2023/09/26 17:39:58 # \u001b[33mWARNING\u001b[39m IoC/DI: No implementations found for >>IMVDatabase<< - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/ioc.py:35\u001b[39m\n",
      "2023/09/26 17:39:58 # \u001b[91mERROR\u001b[39m Can't save test Size Dependency: no database connection - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_test.py:529\u001b[39m\n",
      "A service ('Driverless AI') is running on localhost:12345 and it is accessible\n",
      "2023/09/26 17:39:58 # \u001b[33mWARNING\u001b[39m IoC/DI: No implementations found for >>IMVDatabase<< - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/ioc.py:35\u001b[39m\n",
      "2023/09/26 17:39:58 # \u001b[91mERROR\u001b[39m Can't save test Segment Performance: no database connection - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_test.py:529\u001b[39m\n",
      "2023/09/26 17:39:58 # \u001b[94mDEBUG\u001b[39m Selected database: test-db - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:255\u001b[39m\n",
      "2023/09/26 17:39:58 # \u001b[92mINFO\u001b[39m Initialize MVDatabase: test-db\n",
      "2023/09/26 17:39:58 # \u001b[94mDEBUG\u001b[39m SQLDatabase: results-mv-dai-cli/h2o-sonar/mli_experiment_5389c272-1756-4a68-a6fc-463fdd95e673/tmp/test.sql_db.sqlite - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:187\u001b[39m\n",
      "2023/09/26 17:39:58 # \u001b[94mDEBUG\u001b[39m ObjectStorage: results-mv-dai-cli/h2o-sonar/mli_experiment_5389c272-1756-4a68-a6fc-463fdd95e673/tmp/test.obj_storage.sqlite - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:188\u001b[39m\n",
      "2023/09/26 17:39:58 # \u001b[92mINFO\u001b[39m LocalPlatform local-platform created\n",
      "2023/09/26 17:39:58 # \u001b[94mDEBUG\u001b[39m Database cache is enabled - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:67\u001b[39m\n",
      "2023/09/26 17:39:58 # \u001b[94mDEBUG\u001b[39m Deleting cache entries that are older than 24 hours - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:68\u001b[39m\n",
      "2023/09/26 17:39:58 # \u001b[94mDEBUG\u001b[39m Save credentials: DriverlessCredentials(address='http://localhost:12345', username='h2oai') - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:33\u001b[39m\n",
      "2023/09/26 17:39:59 # \u001b[92mINFO\u001b[39m Adding connection to Driverless AI server 'http://localhost:12345' for user 'h2oai'\n",
      "2023/09/26 17:39:59 # \u001b[94mDEBUG\u001b[39m Save credentials: DriverlessCredentials(address='http://localhost:12345', username='h2oai') - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:33\u001b[39m\n",
      "2023/09/26 17:40:00 # \u001b[94mDEBUG\u001b[39m Worker set: <h2o_mv.platforms.driverless.platform.DriverlessPlatform object at 0x7f0ac32cc8e0> - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:331\u001b[39m\n",
      "2023/09/26 17:40:00 # \u001b[92mINFO\u001b[39m Connection for <class 'h2o_mv.platforms.driverless.platform.DriverlessPlatform'> not supported\n",
      "2023/09/26 17:40:00 # \u001b[94mDEBUG\u001b[39m Import dataset a407dd4c-f658-11ed-9ecf-0242709d15f7 from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:454\u001b[39m\n",
      "INFO - Downloaded 'data/temp/000-walmart_tts_small_train.csv.1684509609.0011456.csv'\n",
      "2023/09/26 17:40:02 # \u001b[94mDEBUG\u001b[39m Import experiment b78cb888-f658-11ed-9ecf-0242709d15f7 from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:628\u001b[39m\n",
      "2023/09/26 17:40:02 # \u001b[94mDEBUG\u001b[39m Import dataset a407dd4c-f658-11ed-9ecf-0242709d15f7 from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:454\u001b[39m\n",
      "INFO - Downloaded 'data/temp/000-walmart_tts_small_train.csv.1684509609.0011456.csv'\n",
      "2023/09/26 17:40:04 # \u001b[94mDEBUG\u001b[39m Dataset with platform_obj_key 'a407dd4c-f658-11ed-9ecf-0242709d15f7' already in DB - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/databases/sql_db.py:314\u001b[39m\n",
      "2023/09/26 17:40:04 # \u001b[94mDEBUG\u001b[39m Import dataset a4077500-f658-11ed-9ecf-0242709d15f7 from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:454\u001b[39m\n",
      "INFO - Downloaded 'data/temp/000-walmart_tts_small_test.csv.1684509608.881174.csv'\n",
      "2023/09/26 17:40:05 # \u001b[94mDEBUG\u001b[39m Folder data/temp/0fee15f1-af18-491b-acd9/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:69\u001b[39m\n",
      "INFO - Downloaded 'data/temp/0fee15f1-af18-491b-acd9/h2oai_experiment_summary_b78cb888-f658-11ed-9ecf-0242709d15f7.zip'\n",
      "2023/09/26 17:40:05 # \u001b[94mDEBUG\u001b[39m Folder data/temp/0fee15f1-af18-491b-acd9/ deleted - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:227\u001b[39m\n",
      "2023/09/26 17:40:05 # \u001b[94mDEBUG\u001b[39m Save Backtesting: Backtesting - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_test.py:531\u001b[39m\n",
      "2023/09/26 17:40:05 # \u001b[94mDEBUG\u001b[39m Folder data/temp/mvt-c8bd7613-4246-4d68-8f18-2f340f0549d2/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:69\u001b[39m\n",
      "2023/09/26 17:40:05 # \u001b[94mDEBUG\u001b[39m Folder data/artifacts/mvt-c8bd7613-4246-4d68-8f18-2f340f0549d2/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:82\u001b[39m\n",
      "2023/09/26 17:40:06 # \u001b[92mINFO\u001b[39m Backtesting 'Backtesting': Running\n",
      "INFO - Downloaded 'data/temp/a1d8ce02-22da-42a5-a83a.tmp'\n",
      "2023/09/26 17:40:08 # \u001b[94mDEBUG\u001b[39m Uploading bt-train:0-mvt-c8bd7613-4246-4d68-8f18-2f340f0549d2 to worker instance - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:664\u001b[39m\n",
      "Complete 100.00% - [4/4] Computed stats for column sample_weight\n",
      "2023/09/26 17:40:11 # \u001b[94mDEBUG\u001b[39m Uploading bt-test:0-mvt-c8bd7613-4246-4d68-8f18-2f340f0549d2 to worker instance - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:664\u001b[39m\n",
      "Complete 100.00% - [4/4] Computed stats for column sample_weight\n",
      "INFO - Experiment launched at: http://localhost:12345/#/experiment?key=fb9f2854-5c82-11ee-9192-00e04c68003f\n",
      "2023/09/26 17:40:14 # \u001b[94mDEBUG\u001b[39m Uploading bt-train:1-mvt-c8bd7613-4246-4d68-8f18-2f340f0549d2 to worker instance - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:664\u001b[39m\n",
      "Complete 100.00% - [4/4] Computed stats for column sample_weight\n",
      "2023/09/26 17:40:16 # \u001b[94mDEBUG\u001b[39m Uploading bt-test:1-mvt-c8bd7613-4246-4d68-8f18-2f340f0549d2 to worker instance - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:664\u001b[39m\n",
      "Complete 100.00% - [4/4] Computed stats for column sample_weight\n",
      "INFO - Experiment launched at: http://localhost:12345/#/experiment?key=fed1a4c0-5c82-11ee-9192-00e04c68003f\n",
      "2023/09/26 17:40:59 # \u001b[92mINFO\u001b[39m Backtesting 'Backtesting': Processing results\n",
      "2023/09/26 17:40:59 # \u001b[94mDEBUG\u001b[39m Folder data/temp/7d547b95-c3d3-4013-bd2a/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:69\u001b[39m\n",
      "INFO - Downloaded 'data/temp/7d547b95-c3d3-4013-bd2a/h2oai_experiment_summary_fed1a4c0-5c82-11ee-9192-00e04c68003f.zip'\n",
      "2023/09/26 17:40:59 # \u001b[94mDEBUG\u001b[39m Folder data/temp/7d547b95-c3d3-4013-bd2a/ deleted - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:227\u001b[39m\n",
      "2023/09/26 17:40:59 # \u001b[92mINFO\u001b[39m Backtesting 'Backtesting': Completed\n",
      "2023/09/26 17:40:59 # \u001b[94mDEBUG\u001b[39m Save Backtesting: Backtesting - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_test.py:531\u001b[39m\n",
      "2023/09/26 17:40:59 # \u001b[94mDEBUG\u001b[39m Worker cleanup - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:801\u001b[39m\n",
      "2023/09/26 17:40:59 # \u001b[94mDEBUG\u001b[39m Deleting temporary experiment fed1a4c0-5c82-11ee-9192-00e04c68003f from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:803\u001b[39m\n",
      "INFO - Driverless AI Server reported experiment fed1a4c0-5c82-11ee-9192-00e04c68003f deleted.\n",
      "2023/09/26 17:41:00 # \u001b[94mDEBUG\u001b[39m Deleting temporary experiment fb9f2854-5c82-11ee-9192-00e04c68003f from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:803\u001b[39m\n",
      "INFO - Driverless AI Server reported experiment fb9f2854-5c82-11ee-9192-00e04c68003f deleted.\n",
      "2023/09/26 17:41:02 # \u001b[94mDEBUG\u001b[39m Deleting temporary dataset fc73d37e-5c82-11ee-9192-00e04c68003f from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:808\u001b[39m\n",
      "INFO - Driverless AI Server reported dataset fc73d37e-5c82-11ee-9192-00e04c68003f deleted.\n",
      "2023/09/26 17:41:02 # \u001b[94mDEBUG\u001b[39m Deleting temporary dataset f953f8b8-5c82-11ee-9192-00e04c68003f from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:808\u001b[39m\n",
      "INFO - Driverless AI Server reported dataset f953f8b8-5c82-11ee-9192-00e04c68003f deleted.\n",
      "2023/09/26 17:41:03 # \u001b[94mDEBUG\u001b[39m Deleting temporary dataset faaf950a-5c82-11ee-9192-00e04c68003f from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:808\u001b[39m\n",
      "INFO - Driverless AI Server reported dataset faaf950a-5c82-11ee-9192-00e04c68003f deleted.\n",
      "2023/09/26 17:41:03 # \u001b[94mDEBUG\u001b[39m Deleting temporary dataset fddeb026-5c82-11ee-9192-00e04c68003f from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:808\u001b[39m\n",
      "INFO - Driverless AI Server reported dataset fddeb026-5c82-11ee-9192-00e04c68003f deleted.\n",
      "2023/09/26 17:41:03 # \u001b[94mDEBUG\u001b[39m Folder data/temp/mvt-c8bd7613-4246-4d68-8f18-2f340f0549d2/ deleted - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:227\u001b[39m\n",
      "2023/09/26 17:41:03 # \u001b[94mDEBUG\u001b[39m Folder data/artifacts/mvt-c8bd7613-4246-4d68-8f18-2f340f0549d2/ deleted - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:227\u001b[39m\n",
      "pandas.Int64Index is deprecated and will be removed from pandas in a future version. Use pandas.Index with the appropriate dtype instead.\n",
      "2023/09/26 17:41:04 # \u001b[33mWARNING\u001b[39m Singleton MVClient already initialized, ignoring: args=(), kwargs={'data_folder': 'results-mv-dai-cli/h2o-sonar/mli_experiment_5389c272-1756-4a68-a6fc-463fdd95e673/tmp'} - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:296\u001b[39m\n",
      "2023/09/26 17:41:04 # \u001b[94mDEBUG\u001b[39m Selected database: test-db - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:255\u001b[39m\n",
      "2023/09/26 17:41:04 # \u001b[92mINFO\u001b[39m Initialize MVDatabase: test-db\n",
      "2023/09/26 17:41:04 # \u001b[94mDEBUG\u001b[39m SQLDatabase: results-mv-dai-cli/h2o-sonar/mli_experiment_5389c272-1756-4a68-a6fc-463fdd95e673/tmp/test.sql_db.sqlite - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:187\u001b[39m\n",
      "2023/09/26 17:41:04 # \u001b[94mDEBUG\u001b[39m ObjectStorage: results-mv-dai-cli/h2o-sonar/mli_experiment_5389c272-1756-4a68-a6fc-463fdd95e673/tmp/test.obj_storage.sqlite - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:188\u001b[39m\n",
      "2023/09/26 17:41:04 # \u001b[92mINFO\u001b[39m Local Platform already exists\n",
      "2023/09/26 17:41:04 # \u001b[94mDEBUG\u001b[39m Database cache is enabled - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:67\u001b[39m\n",
      "2023/09/26 17:41:04 # \u001b[94mDEBUG\u001b[39m Deleting cache entries that are older than 24 hours - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:68\u001b[39m\n",
      "2023/09/26 17:41:04 # \u001b[94mDEBUG\u001b[39m Worker set: <h2o_mv.platforms.driverless.platform.DriverlessPlatform object at 0x7f0ac32cc8e0> - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:331\u001b[39m\n",
      "2023/09/26 17:41:04 # \u001b[94mDEBUG\u001b[39m Save credentials: DriverlessCredentials(address='http://localhost:12345', username='h2oai') - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:33\u001b[39m\n",
      "2023/09/26 17:41:05 # \u001b[92mINFO\u001b[39m Adding connection to Driverless AI server 'http://localhost:12345' for user 'h2oai'\n",
      "2023/09/26 17:41:05 # \u001b[94mDEBUG\u001b[39m Save credentials: DriverlessCredentials(address='http://localhost:12345', username='h2oai') - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_cache.py:33\u001b[39m\n",
      "2023/09/26 17:41:05 # \u001b[94mDEBUG\u001b[39m Platform with address 'http://localhost:12345' for user 'h2oai' already in DB - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/databases/sql_db.py:234\u001b[39m\n",
      "2023/09/26 17:41:06 # \u001b[94mDEBUG\u001b[39m Worker set: <h2o_mv.platforms.driverless.platform.DriverlessPlatform object at 0x7f0ac267e100> - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_client.py:331\u001b[39m\n",
      "2023/09/26 17:41:06 # \u001b[92mINFO\u001b[39m Connection for <class 'h2o_mv.platforms.driverless.platform.DriverlessPlatform'> not supported\n",
      "2023/09/26 17:41:06 # \u001b[94mDEBUG\u001b[39m Import dataset a407dd4c-f658-11ed-9ecf-0242709d15f7 from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:454\u001b[39m\n",
      "INFO - Downloaded 'data/temp/000-walmart_tts_small_train.csv.1684509609.0011456.csv'\n",
      "2023/09/26 17:41:08 # \u001b[94mDEBUG\u001b[39m Dataset with platform_obj_key 'a407dd4c-f658-11ed-9ecf-0242709d15f7' already in DB - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/databases/sql_db.py:314\u001b[39m\n",
      "2023/09/26 17:41:08 # \u001b[94mDEBUG\u001b[39m Import dataset a4077500-f658-11ed-9ecf-0242709d15f7 from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:454\u001b[39m\n",
      "INFO - Downloaded 'data/temp/000-walmart_tts_small_test.csv.1684509608.881174.csv'\n",
      "2023/09/26 17:41:09 # \u001b[94mDEBUG\u001b[39m Dataset with platform_obj_key 'a4077500-f658-11ed-9ecf-0242709d15f7' already in DB - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/databases/sql_db.py:314\u001b[39m\n",
      "2023/09/26 17:41:09 # \u001b[94mDEBUG\u001b[39m Import experiment b78cb888-f658-11ed-9ecf-0242709d15f7 from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:628\u001b[39m\n",
      "2023/09/26 17:41:10 # \u001b[94mDEBUG\u001b[39m Import dataset a407dd4c-f658-11ed-9ecf-0242709d15f7 from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:454\u001b[39m\n",
      "INFO - Downloaded 'data/temp/000-walmart_tts_small_train.csv.1684509609.0011456.csv'\n",
      "2023/09/26 17:41:12 # \u001b[94mDEBUG\u001b[39m Dataset with platform_obj_key 'a407dd4c-f658-11ed-9ecf-0242709d15f7' already in DB - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/databases/sql_db.py:314\u001b[39m\n",
      "2023/09/26 17:41:12 # \u001b[94mDEBUG\u001b[39m Import dataset a4077500-f658-11ed-9ecf-0242709d15f7 from Driverless Server http://localhost:12345 - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:454\u001b[39m\n",
      "INFO - Downloaded 'data/temp/000-walmart_tts_small_test.csv.1684509608.881174.csv'\n",
      "2023/09/26 17:41:13 # \u001b[94mDEBUG\u001b[39m Dataset with platform_obj_key 'a4077500-f658-11ed-9ecf-0242709d15f7' already in DB - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/databases/sql_db.py:314\u001b[39m\n",
      "2023/09/26 17:41:13 # \u001b[94mDEBUG\u001b[39m Folder data/temp/4bb9bf64-6fc8-4f7b-b369/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:69\u001b[39m\n",
      "INFO - Downloaded 'data/temp/4bb9bf64-6fc8-4f7b-b369/h2oai_experiment_summary_b78cb888-f658-11ed-9ecf-0242709d15f7.zip'\n",
      "2023/09/26 17:41:13 # \u001b[94mDEBUG\u001b[39m Folder data/temp/4bb9bf64-6fc8-4f7b-b369/ deleted - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:227\u001b[39m\n",
      "2023/09/26 17:41:13 # \u001b[94mDEBUG\u001b[39m Model with platform_obj_key 'b78cb888-f658-11ed-9ecf-0242709d15f7' already in DB - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/databases/sql_db.py:424\u001b[39m\n",
      "2023/09/26 17:41:13 # \u001b[94mDEBUG\u001b[39m Save Size Dependency: Size Dependency - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/mv_test.py:531\u001b[39m\n",
      "2023/09/26 17:41:13 # \u001b[94mDEBUG\u001b[39m Folder data/temp/mvt-26fb87ed-bb10-4582-a0ec-2639e345d9aa/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:69\u001b[39m\n",
      "2023/09/26 17:41:13 # \u001b[94mDEBUG\u001b[39m Folder data/artifacts/mvt-26fb87ed-bb10-4582-a0ec-2639e345d9aa/ created - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/core/utils.py:82\u001b[39m\n",
      "2023/09/26 17:41:14 # \u001b[92mINFO\u001b[39m Size Dependency 'Size Dependency': Running\n",
      "2023/09/26 17:41:14 # \u001b[94mDEBUG\u001b[39m Get cached dataset: 000-walmart_tts_small_train.csv (mvid='ds-9640ed1e-24a2-4b64-a461-90a7565907af', dataset_key='a407dd4c-f658-11ed-9ecf-0242709d15f7') - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:495\u001b[39m\n",
      "2023/09/26 17:41:15 # \u001b[94mDEBUG\u001b[39m Uploading sd-train:0-mvt-26fb87ed-bb10-4582-a0ec-2639e345d9aa to worker instance - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:664\u001b[39m\n",
      "Complete 100.00% - [4/4] Computed stats for column sample_weight\n",
      "INFO - Experiment launched at: http://localhost:12345/#/experiment?key=223bdc3c-5c83-11ee-9192-00e04c68003f\n",
      "2023/09/26 17:41:18 # \u001b[94mDEBUG\u001b[39m Uploading sd-train:1-mvt-26fb87ed-bb10-4582-a0ec-2639e345d9aa to worker instance - \u001b[36m/home/dvorka/h/mli/git/h2o-sonar/.venv/lib/python3.8/site-packages/h2o_mv/platforms/driverless/platform.py:664\u001b[39m\n",
      "Complete 100.00% - [4/4] Computed stats for column sample_weight\n",
      "INFO - Experiment launched at: http://localhost:12345/#/experiment?key=2446124a-5c83-11ee-9192-00e04c68003f\n"
     ]
    }
   ],
   "source": [
    "# run the interpretations\n",
    "!h2o-sonar run interpretation \\\n",
    "  --args-as-json-location {run_interpretation_args_path} \\\n",
    "  --config-path {h2o_sonar_config_path} \\\n",
    "  --encryption-key {encryption_key} \\\n",
    "  --results-location \"results-mv-dai-cli\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "345129a9-d874-4701-8c4d-ec70f6d31343",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "h2o-sonar",
   "language": "python",
   "name": "h2o-sonar"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}