play.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import torch\n",
    "\n",
    "from retnet.configuration_retnet import RetNetConfig\n",
    "from retnet.modeling_retnet import RetNetModel, RetNetForCausalLM\n",
    "\n",
    "%load_ext autoreload\n",
    "%autoreload 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "torch.manual_seed(0)\n",
    "config = RetNetConfig(decoder_layers=6,\n",
    "                      decoder_embed_dim=256,\n",
    "                      decoder_retention_heads=4,\n",
    "                      decoder_ffn_embed_dim=512)\n",
    "\n",
    "model = RetNetModel(config)\n",
    "model.eval()\n",
    "\n",
    "device = 'cpu'  # cuda, cpu, mps for m1 mac\n",
    "model = model.to(device)\n",
    "\n",
    "def count_parameters(model):\n",
    "    return sum(p.numel() for p in model.parameters() if p.requires_grad)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Test 1: Basic Equivalence of the 3 forward implementations"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "parallel vs recurrent True\n",
      "parallel vs chunkwise False\n",
      "layer: 1\n",
      "parallel vs recurrent: True\n",
      "parallel vs chunkwise False\n",
      "layer: 2\n",
      "parallel vs recurrent: True\n",
      "parallel vs chunkwise False\n",
      "layer: 3\n",
      "parallel vs recurrent: True\n",
      "parallel vs chunkwise False\n",
      "layer: 4\n",
      "parallel vs recurrent: True\n",
      "parallel vs chunkwise False\n",
      "layer: 5\n",
      "parallel vs recurrent: True\n",
      "parallel vs chunkwise False\n",
      "layer: 6\n",
      "parallel vs recurrent: True\n",
      "parallel vs chunkwise False\n"
     ]
    }
   ],
   "source": [
    "input_ids = torch.LongTensor([[1,2,1,2]]).to(device)\n",
    "\n",
    "parallel_outputs = model(input_ids, forward_impl='parallel', use_cache=True)\n",
    "parallel_state = parallel_outputs.last_hidden_state\n",
    "parallel_cache = parallel_outputs.past_key_values\n",
    "\n",
    "past_kv = None\n",
    "rnn_state = []\n",
    "for i in range(input_ids.shape[1]):\n",
    "    rnn_out = model(input_ids[:, :i+1], forward_impl='recurrent', past_key_values=past_kv, use_cache=True)\n",
    "    rnn_state.append(rnn_out.last_hidden_state)\n",
    "    past_kv = rnn_out.past_key_values\n",
    "rnn_state = torch.cat(rnn_state, dim=1)\n",
    "rnn_cache = rnn_out.past_key_values\n",
    "\n",
    "\n",
    "chunk_outputs = model(input_ids, forward_impl='chunkwise', use_cache=True, recurrent_chunk_size=2)\n",
    "chunk_state = chunk_outputs.last_hidden_state\n",
    "chunk_cache = chunk_outputs.past_key_values\n",
    "\n",
    "print('parallel vs recurrent', torch.allclose(parallel_state, rnn_state, atol=1e-5))\n",
    "print('parallel vs chunkwise', torch.allclose(parallel_state, chunk_state, atol=1e-5))\n",
    "\n",
    "for i, (p, r, c) in enumerate(zip(parallel_cache, rnn_cache, chunk_cache)):\n",
    "    print(f\"layer: {i + 1}\")\n",
    "    key = 'prev_key_value'\n",
    "    print('parallel vs recurrent:', torch.allclose(p[key], r[key], atol=1e-5))\n",
    "    print('parallel vs chunkwise', torch.allclose(p[key], c[key], atol=1e-5))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Test 2: Retention Mask\n",
    "\n",
    "Test both left padding / right padding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "parallel vs recurrent: True\n",
      "parallel vs chunkwise False\n",
      "layer: 1\n",
      "parallel vs recurrent: True\n",
      "parallel vs chunkwise False\n",
      "layer: 2\n",
      "parallel vs recurrent: True\n",
      "parallel vs chunkwise False\n",
      "layer: 3\n",
      "parallel vs recurrent: True\n",
      "parallel vs chunkwise False\n",
      "layer: 4\n",
      "parallel vs recurrent: True\n",
      "parallel vs chunkwise False\n",
      "layer: 5\n",
      "parallel vs recurrent: True\n",
      "parallel vs chunkwise False\n",
      "layer: 6\n",
      "parallel vs recurrent: True\n",
      "parallel vs chunkwise False\n"
     ]
    }
   ],
   "source": [
    "input_ids = torch.LongTensor([[1,2,3,4,1,2,5,5],\n",
    "                              [5,5,1,2,3,4,1,2]]).to(device)\n",
    "retention_mask = torch.LongTensor([[1,1,1,1,1,1,0,0],\n",
    "                                   [0,0,1,1,1,1,1,1]]).to(device)\n",
    "\n",
    "parallel_outputs = model(input_ids, retention_mask=retention_mask, forward_impl='parallel', use_cache=True)\n",
    "parallel_state = parallel_outputs.last_hidden_state\n",
    "parallel_cache = parallel_outputs.past_key_values\n",
    "\n",
    "past_kv = None\n",
    "rnn_state = []\n",
    "for i in range(input_ids.shape[1]):\n",
    "    rnn_out = model(input_ids[:, :i+1], retention_mask=retention_mask[:, i:i+1], forward_impl='recurrent', past_key_values=past_kv, use_cache=True)\n",
    "    rnn_state.append(rnn_out.last_hidden_state)\n",
    "    past_kv = rnn_out.past_key_values\n",
    "rnn_state = torch.cat(rnn_state, dim=1)\n",
    "rnn_cache = rnn_out.past_key_values\n",
    "\n",
    "\n",
    "chunk_outputs = model(input_ids, retention_mask=retention_mask, forward_impl='chunkwise', use_cache=True, recurrent_chunk_size=4)\n",
    "chunk_state = chunk_outputs.last_hidden_state\n",
    "chunk_cache = chunk_outputs.past_key_values\n",
    "\n",
    "mask = retention_mask.unsqueeze(-1).float()\n",
    "print('parallel vs recurrent:', torch.allclose(parallel_state * mask, rnn_state * mask, atol=1e-5))\n",
    "print('parallel vs chunkwise', torch.allclose(parallel_state * mask, chunk_state * mask, atol=1e-5))\n",
    "\n",
    "for i, (p, r, c) in enumerate(zip(parallel_cache, rnn_cache, chunk_cache)):\n",
    "    print(f\"layer: {i + 1}\")\n",
    "    key = 'prev_key_value'\n",
    "    print('parallel vs recurrent:', torch.allclose(p[key], r[key], atol=1e-5))\n",
    "    print('parallel vs chunkwise', torch.allclose(p[key], c[key], atol=1e-5))\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Test 3: Generate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor(True)\n",
      "tensor(True)\n"
     ]
    }
   ],
   "source": [
    "input_ids = torch.LongTensor([[1,2,3,4,5,6,7,8],\n",
    "                              [8,7,6,5,4,3,2,1]]).to(device)\n",
    "model = RetNetForCausalLM(config).to(device)\n",
    "model.eval()\n",
    "\n",
    "# greedy\n",
    "p_generated = model.custom_generate(input_ids, parallel_compute_prompt=True, max_new_tokens=20, do_sample=False, early_stopping=False)\n",
    "r_generated = model.custom_generate(input_ids, parallel_compute_prompt=False, max_new_tokens=20, do_sample=False, early_stopping=False)\n",
    "print((p_generated == r_generated).all())\n",
    "\n",
    "# sampling\n",
    "torch.manual_seed(0)\n",
    "p_generated = model.custom_generate(input_ids, parallel_compute_prompt=True, max_new_tokens=20, do_sample=True, early_stopping=False)\n",
    "torch.manual_seed(0)\n",
    "r_generated = model.custom_generate(input_ids, parallel_compute_prompt=False, max_new_tokens=20, do_sample=True, early_stopping=False)\n",
    "print((p_generated == r_generated).all())\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Test 4: Forgetting (Long sequence dependency)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "config = RetNetConfig(decoder_layers=6,\n",
    "                      decoder_embed_dim=256,\n",
    "                      decoder_retention_heads=8,\n",
    "                      decoder_ffn_embed_dim=512)\n",
    "\n",
    "model = RetNetModel(config)\n",
    "model.eval();"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "tensor([ 256,  512, 1024, 2048, 4096, 8192])\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAABlgAAAKoCAYAAAAbLfJHAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjcuMSwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy/bCgiHAAAACXBIWXMAAA9hAAAPYQGoP6dpAAEAAElEQVR4nOzdd3hUZfrG8e+kB1JIgNASIPTek9CMYAFFRUAUG0jVdSnuoqIIKIgawa674koJTRALCKKCBWkKCSFA6L1DCCmkt5k5vz/4bXZZLAwkOZnk/lxXrstzzszJzQh5c87znvexGIZhICIiIiIiIiIiIiIiItfMxewAIiIiIiIiIiIiIiIizkYFFhEREREREREREREREQepwCIiIiIiIiIiIiIiIuIgFVhEREREREREREREREQcpAKLiIiIiIiIiIiIiIiIg1RgERERERERERERERERcZAKLCIiIiIiIiIiIiIiIg5SgUVERERERERERERERMRBKrCIiIiIiIiIiIiIiIg4SAUWkWs0depULBYLycnJZkcp8u9M1+LYsWMMGDCAKlWq4OPjw+233058fHwJJxQRkf/lzOPJ3r17+etf/0qXLl2oXLkyFouF9evXl3xAERG5ijOPJ3PmzKFfv37Ur18fb29vGjVqxJNPPsn58+dLIaWIiPwvZx5Tli5dSmRkJDVq1MDT05PatWtzzz338Ouvv5ZCShEVWEQqhIsXL3LTTTdx6NAh5s2bx2effUZeXh49evTg4MGDZscTEREnERcXx1dffUVgYCC33nqr2XFERMRJvfTSS/j4+PDaa6+xZs0aJkyYwOrVq+nYsSMXLlwwO56IiDiRlJQUunXrxocffsj333/P22+/zYULF4iMjGTDhg1mx5MKwM3sACJS8t544w0uXrzIr7/+Sr169QDo3r07DRs25MUXX2TZsmUmJxQREWcwePBgHnvsMQC++OILvv76a5MTiYiIM9qxYwdBQUFF2zfffDMdOnQgLCyM2bNnM3nyZBPTiYiIMxkzZsxV++68806qV6/O3Llzufnmm01IJRWJnmARcdCFCxd46KGH8Pf3p0aNGgwfPpz09PQrXmMYBh9++CHt2rXD29ubgIAABg4cyLFjx6543Q8//MC9995LcHAwXl5eNGrUiCeeeOI3H8n85ptvaNeuHZ6enoSGhvLmm29ec+YVK1Zwyy23FBVXAPz8/BgwYABff/01VqvVwU9BRERulDOOJy4u+tVRRKSsccbx5L+LK//WsWNHXF1dOX369DWfR0REipczjim/xdfXFy8vL9zc9GyBlDz9LRNx0H333cegQYMYMWIEu3fvZuLEiQDMmzev6DVPPPEE8+fPZ9y4ccyYMYPU1FRefvllunbtyq5du6hRowYAR48epUuXLowcORJ/f39OnDjB22+/Tffu3dm9ezfu7u4A/PTTT9x777106dKFTz/9FJvNxsyZM6/p8fnc3FyOHj1K//79rzrWpk0bcnNzOXbsGE2aNCmOj0dERK6Rs40nIiJSNpWX8WTDhg3YbDZatmx5A5+GiIjcCGceU2w2G3a7nbNnzxIVFYVhGIwePbqYPhmRP2CIyDV56aWXDMCYOXPmFfv/+te/Gl5eXobdbjcMwzC2bNliAMZbb711xetOnz5teHt7GxMmTPjN89vtdqOwsNA4efKkARgrV64sOhYREWHUrl3byM3NLdqXkZFhBAYGGn/2z/js2bMGYERFRV11bMmSJQZg/Prrr3/8hxcRkWLjrOPJ//r8888NwPj5558dep+IiBSP8jKe/Pu9zZs3N0JCQozMzEyH3y8iIjemPIwpTZs2NQADMGrVqmVs3rz5mt8rciO0zoOIg/r27XvFdps2bcjLyyMpKQmA1atXY7FYePTRR7FarUVfNWvWpG3btqxfv77ovUlJSfzlL38hJCQENzc33N3di5bx2r9/PwDZ2dls27aNAQMG4OXlVfReX19f7rnnnmvObbFYruuYiIiUDGcdT0REpGxx9vEkLy+PAQMGcPLkST7//HN8fHwcPoeIiBQPZx5TvvzyS2JiYvj8889p0aIFd9555xV5REqKlggTcVDVqlWv2Pb09AQuL8UFl9erNAyj6JHI/9WgQQMA7HY7vXr14ty5c0yZMoXWrVtTuXJl7HY7nTt3LjpfWloadrudmjVrXnWu39r3vwICArBYLKSkpFx1LDU1FYDAwMA/PY+IiBQvZxtPRESkbHLm8SQ/P5/+/fuzefNmVq9eTUREhEPvFxGR4uXMY8q/l5gMDw+nX79+tG/fnqeeeopdu3Y5dB4RR6nAIlLMqlWrhsViYdOmTUUD0X/79749e/awa9cu5s+fz2OPPVZ0/MiRI1e8/t8FksTExKvO9Vv7/pe3tzeNGjVi9+7dVx3bvXs33t7eRQOgiIiUHWVtPBEREedUVseT/Px8+vXrx88//8zKlSu59dZbr/m9IiJijrI6pvwvNzc3OnTowGeffXbd5xC5VloiTKSY3X333RiGwdmzZ+nUqdNVX61btwb+syzX/w5I//rXv67Yrly5MuHh4Sxfvpy8vLyi/ZmZmXz99dfXlKl///6sW7eO06dPX/H+5cuX07dvX9zcVGsVESlryuJ4IiIizqcsjif/fnJl3bp1fPnll/Tu3ftG/ogiIlJKyuKY8lvy8vLYunUrjRo1uu5ziFwr3VUVKWbdunXj8ccfZ9iwYcTFxREZGUnlypU5f/48mzdvpnXr1jz55JM0a9aMhg0b8vzzz2MYBoGBgXz99df88MMPV51z+vTp3HHHHdx+++08/fTT2Gw2ZsyYQeXKlYuW+fojzzzzDIsWLeKuu+7i5ZdfxtPTk9dff528vDymTp1aAp+CiIjcqLI4nuTk5PDtt98CsHXrVgA2bNhAcnIylStX5s477yzeD0FERG5YWRxPBg4cyHfffcekSZOoWrVq0ZgC4OfnR4sWLYr1MxARkeJRFseUrl270rdvX5o3b46/vz8nTpxg1qxZHD16lBUrVpTExyByBRVYRErAv/71Lzp37sy//vUvPvzwQ+x2O7Vr16Zbt26Eh4cD4O7uztdff81TTz3FE088gZubG7fddhs//vgjdevWveJ8t99+O1999RWTJ09m0KBB1KxZk7/+9a/k5uYybdq0P81TvXp1Nm3axDPPPMNjjz2G1WqlS5curF+/nmbNmpXIZyAiIjeurI0nSUlJ3H///Vfs+3ehvl69epw4caJY/twiIlK8ytp4snr1agBeffVVXn311SuO3XzzzWpKLCJShpW1MaVr1658+umnnDhxguzsbKpVq0aXLl1455136Nq1a4l8BiL/zWIYhmF2CBEREREREREREREREWeiHiwiIiIiIiIiIiIiIiIOUoFFRERERERERERERETEQSqwiIiIiIiIiIiIiIiIOEgFFhEREREREREREREREQepwCIiIiIiIiIiIiIiIuIgFVhEREREREREREREREQc5GZ2ADPY7XbOnTuHr68vFovF7DgiIk7FMAwyMzOpXbs2Li4Vu06v8URE5PppPLmSxhQRkeunMeU/NJ6IiFy/6xlPKmSB5dy5c4SEhJgdQ0TEqZ0+fZrg4GCzY5hK44mIyI3TeHKZxhQRkRunMUXjiYhIcXBkPKmQBRZfX1/g8gfl5+dnchoREeeSkZFBSEhI0c/SikzjiYjI9dN4ciWNKSIi109jyn9oPBERuX7XM55UyALLvx+R9PPz02AjInKd9Li5xhMRkeKg8eQyjSkiIjdOY4rGExGR4uDIeFKxF6YUERERERERERERERG5DiqwiIiIiIiIiIiIiIiIOEgFFhEREREREREREREREQepwCIiIiIiIiIiIiIiIuIgFVhERERERESuUf/+/QkICGDgwIFX7M/MzCQsLIx27drRunVrZs+ebVJCERFxFhpTREScnwosIiIiIiIi12jcuHEsXLjwqv2VKlViw4YN7Ny5k5iYGKKiokhJSTEhoYiIOAuNKSIizk8FFhERERERkWvUs2dPfH19r9rv6upKpUqVAMjLy8Nms2EYRmnHExERJ6IxRUTE+anAIiIiIiIi8v+ioqIICwvD19eXoKAg+vXrx8GDB6/pvZcuXaJt27YEBwczYcIEqlWrVsJpRUSkLNOYIiJS/qnAIiIVXqHVRn5hxfsqtNrN/uhFRMTJ5eQVciYp0+wYxWrDhg2MHj2arVu38sMPP2C1WunVqxfZ2dl/+t4qVaqwa9cujh8/zpIlS7hw4UIpJBYRcX42u8GhU2lmxyh2GlNERErfoVNp2Gyld8/LrdS+k4hIGZOZU8A/Pt/JrwnnzY5iilvDQvjbgx3MjiEiIk7IbjdYH3+a+av34VfZg/fG98DVtXzM3VqzZs0V29HR0QQFBbF9+3YiIyOv6Rw1atSgTZs2bNy4kfvvv/83X5Ofn09+fn7RdkZGxvWHFhFxYjl5hbz5yXZ2HEzilb90o2WDqmZHKjalMaZoPBERucwwDL795Tgfr9zD3d1DGXVv61L5vuXjKkhExEEHTqQy7q31Fba4IiIicr2OnLnEc//YxDtLd5CWmU+B1c7FS7lmxyox6enpAAQGBv7h6y5cuFB0UysjI4ONGzfStGnT3319VFQU/v7+RV8hISHFF1pExEkkpmTz7Aeb2LbvAi4WC5ey8v/8TU6sJMYUjSciImC12Zn1ZQIfrdiN3W6QlVOIzV46vassRgXskpWRkYG/vz/p6en4+fmZHUdESpHdbrBi/REWfrcfu92gVtXKPPNoR0JqXN1YsLxzdbHg4e7q8Pv0M/Q/9FmISEWSnpXPou/2833MSQwDvDxceeC2JvS7uSHubuVzPDEMg3vvvZe0tDQ2bdoEQO/evYmPjyc7O5vAwEBWrFhBWFgY27dvZ8SIERiGgWEYPPnkkzz55JO/e+7fmnEcEhJSpj8PEZHitOdoMq/N30ZmTgGBfp5MGhZBk7oB13WuijymaDwRkYouI7uAGQu3kXAkGYsFHuvTggE9G2GxWBw/13WMJ1oiTEQqjPSsfN5ZGs/2A0kA3NSuDmPub0slL3eTk4mIiJRdNpudNVtOsHjNAbJyCwGIbF+H4fe0pKq/t8npStaYMWNISEhg8+bNRfvWrl37m6/t2LEjO3fuvOZze3p64unpeaMRRUSc0vcxJ5n15S6sNoNGwf5MHh6hMeW/ODKmaDwRkYrs9IVMps+N4XxKNt6erjz9cEciWtUq1QwqsIhIhbDnaDJvLN5OakYeHm4ujOrXmt6d611XNVtERKSi2HM0mX+t2M2J85eXKalfy48n+remVcNqJicreWPHjmXVqlVs3LiR4OBgs+OIiJQLNrtB9Nd7WbnxKADd29bmqQfb4+VRvm9PaUwRESl+2w9cYOaiOHLyrAQFVmLK8Ajq1yr9J/fK9wgmIhWe3W7w+bpDLFlzALsBdar78NyQToTW9jc7moiISJmVfCmX6K/3snHnWQB8vN159I5m3NGlfrlpZv97DMNg7NixrFixgvXr1xMaGmp2JBGRciE7t5CZi+OI//8VBR7u3YwHb29Srie9aUwRESl+hmHw9aZjzF21B7sBLRtUZeJjYfj7mPM0nwosIlJupWXm8faSeHYeughAz47BPHlfW7w99aNPRETktxRabXy14Sif/XiIvAIbFgv07lyfR+9oZtoFS2kbPXo0S5YsYeXKlfj6+pKYmAiAv78/3t7le/kaEZGScj45m+nztnL6QhYe7q78/aH2dG9bx+xYJU5jiohI8Sq02vloeQLfx5wE4Pbwujx5X1vc3cybBKa7jCJSLu06fJG3PtlOWmY+Hu6uPDmgNbeG1S3Xs6NERERuxLZ9icxeuYfzydkANK8fyOP9W9MouIq5wUrZrFmzAOjRo8cV+6Ojoxk6dGjpBxIRcXK7jyQTtSCWzJxCqvp7MXlYBI1Cqpgdq1RoTBERKT7pWfm8vnAbe46m4GKBYfe05N7Ihqbf61OBRUTKFZvd4NPvD7Lsx4MYBtSt6ctzgztRt2bpr8EoIiLiDM4lZzH7qz3E7b8AQICvJ0PvbknPjsGmX6yYwTAMsyOIiJQba7ac4KPlCdjsBo1DqjB5eASBfl5mxyo1GlNERIrHycQMXpkXQ2JKDt6ebkwY3IlOzWuYHQtQgUVEypGU9Fze+iSe3UeTgcuPCT7ev3W5b5goIiJyPXLzrXz+0yFWrD+K1WbH1cVC38iGPHh7Eyp5uZsdT0REnJjNZmfOqj2s3nwcgMj2dRg3qD2e7q4mJxMREWezbV8ibyzeTm6+lZpVLzezL0sTqXXXUUTKhfiDSby9ZDvpWQV4ebgyemBbenQMMTuWiIhImWMYBpt2nmXe13tJSc8DoH2T6ozq15qQGr4mpxMREWeXlVvIzIXb2PH/vTAfvbMZD9xavpvZi4hI8TMMg682HCV69V4MA1o1rMrzQ8xrZv97VGAREadms9n5ZO0BPv/pMAD1a/nx3JBOBAfpBpGIiMj/On4unY+/2s2eoykABAVWYmTfVnRuVVM3vkRE5Iadu5jFy3NjOHsxC08PV8Y/1IGubWqbHUtERJxModXGh18k8OO2UwD07lyPJ/q3MbWZ/e9RgUVEnNbFtFzeWBzH/hOpANzZpT4j7m2lx85FRET+R1ZOAZ+sOcC3vx7HboCHmwsDb23CgJ6NNG6KiEix2HXoIq8v3EZWbiHVqngzZXgEDer4mx1LREScTHpWPq/Nj2Xf8VRcLDDi3lbc071BmZ0QpgKLiDilbfsSeWdpPJk5hXh7ujH2gXbc1K6O2bFERETKFJvd4MfYkyz8dj8Z2QUAdG1TixH3tCIosJLJ6UREpLz45pfjfPzVbux2g6b1Apg0NJyACtTMXkREiseJ8xlMn7uVpLRcKnu5MWFwGB2aBZkd6w+pwCIiTsVqs7Pw2/2sWH8EgEbB/kwYHEatapVNTiYiIlK2HDiZyr+WJ3DkTDoAITV8eKJfG9o2qW5yMhERKS+sNjuzv9rNt7+eAKBHx2DG3t8ODz0dKSIiDordm8ibn8SRm2+jVrXKTBke4RQ9IlVgERGncSE1hzcWxXHwVBoAd3cPZfg9LXF30y/vIiIi/5aWkcf8b/axLu40AJW83HioVzPu7h6Km2vZW7NYREScU1ZOAa8v3Mauw8lYLDD4zuYMvKVxmV3CRUREyibDMFj+8xEWfLsPw4A2jarx/GNh+FbyMDvaNVGBRUScwpbd53lv2Q6ycwup7O3OU4Pa0aW1miWKiIj8m9Vm5+tNx1j6/UFy860A3BoWwmN3tSDAV8u0iIhI8Tl9IZNX5sVwLjkbLw9Xnn6kI51b1TI7loiIOJmCQhv//GJX0eSwO7vU5/H+rZ1qYpgKLCJSphVabUSv3sfXm44B0KRuFSYMDqOG1o0XEREpsuNgEh9/tZszSVkANAqpwhP9W9OsXqDJyUREpLyJP5jEzIXbyM6zEhTgzeThEYTWVjN7ERFxTFpmHq9Fx3LgZBouLhYe79eau7qFmh3LYSqwiEiZdT45m5mLthWtHd/v5oYM6dMCdzfnqWKLiIiUpAupOcxdtYctu88D4O/jwZA+LbgtrC4uLlqiRUREio9hGHy9+RhzV+7BbkDz+oG8MDScKr6eZkcTEREnc+xsOtPnxZB8KZfK3u48P6QT7ZqU7Wb2v0cFFhEpkzbvOssHn+0kJ8+KbyV3/vZQB8Jb1DQ7loiISJmQX2jjy3WH+XLdYQqsdlxcLNzVLZSHezfDx9vd7HgiIlLOWG12PlqewNqtJ4HLS1COHthW/TBFRMRhW3af460l8eQX2KhTvTJTRnSmTnUfs2NdNxVYRKRMKSi0MWfVHr779QRweVbUs492onqAt7nBREREygDDMNiy+zxzV+0hKS0XgNYNq/F4/9bUr+VncjoRESmPMrILeH3BNnYfvdzMftjdLel3c0M1sxcREYcYhsHnPx1m0Xf7AWjXpDrPDe6Ej5M0s/89KrCISJlx9mIWMxZu4/i5DADuv7UxD/du5lSNrURERErK6QuZfLxiNzsPXwSgmr8Xw/u2onvb2rrJJSIiJeJUYgbT58WQmJKDt6cbzz7akTCtLCAiIg7KL7TxwbKdbNhxBoC7u4cysm8rXMvBPT8VWESkTFi//TT//GIXeQU2/H08GP9QRzo0c861F0VERIpTTl4hS78/yNebjmGzG7i5ujCgZyPuv6UxXp76dV5EREpG3P4LvLE4jpw8KzUCKzFlRAT1auppSRERcUxqRh6vRsdw6NQlXF0sPDGgDXd2qW92rGKjKzIRMVVegZWPV+zmh9hTwOVlTp5+pANV/bUkmIiIVGx2u8HP208z/5t9XMrMByC8RU1G3tuKWtUqm5xORETKK8MwWLnxGNFfX25m37JBVSY+Foa/j5rZi4iIY46cucSr82JITs/Dx9udiUPDaNOoutmxipUKLCJimlOJGcxYFMepxEwsFhh0W1Me7NUUVxctcyIiIhXbkdOX+GhFAgdPpgFQu1plRvVrTafmNUxOJiIi5Vmh1c6sL3cVTYC7PbwuT97XFnc351/CRUREStcvCed4e0k8BYU2goN8mDIigtrVnLeZ/e9RgUVETPFj7Ck+WpFAfoGNKr6ePPNIR9o2Ll8VbBEREUelZ+Wz6Lv9fB9zEsMALw9XBt3elHsjG+Du5mp2PBERKcfSs/KJWrCNvcdScLHA8L6t6HtTA/X5EhERhxiGwac/HGLJ2gMAdGgWxIRHO1HZ293kZCVDBRYRKVW5+VY+Wp7AurjTALRrXJ3xj3QgwNfL5GQiIiLmsdnsfPvrCT5Ze4Ds3EIAbm4fzLB7WmjZTBERKXEnz2fw8rwYklJzqOTlxrOPdtJTkyIi4rD8QhvvfbqDTTvPAtA3sgHD725ZLprZ/x4VWESk1Jw4n8GMhds4k5SFiwUevqMZA29poiXBRESkQtt9NJmPV+zmxPkMAEJr+/FE/za0bFDV5GQiIlIRxO5L5M3FceTm26hVtTJTRkQQUsPX7FgiIuJkUtJzeSU6liOnLzezf/K+tvTuXM/sWCVOBRYRKXGGYfB9zEk+XrGbAqudQD8vnn20I60aVjM7moiIiGmSL+Uy7+u9RbO7fLzdefTO5tzRpb4mH4iISIkzDIMV648y/5u9GAa0bliN5x8Lw6+yh9nRRETEyRw+ncYr82JJzcjDt5IHE4eG0bqC3PdTgUVESlROXiH//HwXG///5lGHZkGMf6gD/j6eJicTERExR6HVxor1R/nsp0PkF9iwWKB35/o8ekczjY8iIlIqCq02/vnFLn7adnnp5ju61OeJ/q1xK8dLuIiISMnYtOMs734aT4HVTt2avkwZHkHNqpXNjlVqVGARkRJz9MwlZiyK43xyNi4uFobc2Zz+PRrholm5IiJSQcXuS2TOV3s4n5INQPP6gTzRvzUNg6uYG0xERCqMS5n5vDY/lv0nUnFxsTDq3lbc1S1UzexFRMQhdrvB0u8P8ukPBwHo1LwGzz7akUpe5bOZ/e9RgUVEip1hGHz7y3HmrNqL1WanWhVvJjzaieahgWZHkwqsf//+rF+/nltvvZUvvviiaP/p06cZPHgwSUlJuLm5MWXKFO6//34Tk4pIeXTuYhazV+4hbv8FAAL9PBl6d0t6dAjWDS0RESk1x8+l88q8GJLScqns5cZzQ8Jo3zTI7FgiIuJk8gqsvLt0B78knAOgf49GPHZXiwq51LEKLCJSrLJyC/nHZzuLfsBGtKzJUw+2x7eS1vEVc40bN47hw4ezYMGCK/a7ubnx7rvv0q5dO5KSkujQoQN9+vShcuWK8ziriJSc3Hwrn/14iK82HMVqs+PmaqHvTQ0ZdHuTCjezS0REzLV1z3ne+mQ7eQU2ale73Mw+OEjN7EVExDHJl3J5JTqGo2fScXO1MHpgO24Lr2t2LNOowCIixebQqTRmLorjQmoObq4Wht7dkr43NdDMXCkTevbsyfr166/aX6tWLWrVqgVAUFAQgYGBpKamqsAiIjfEMAw27jhL9Oq9pKTnAdChaRCj+rXSzSwRESlVhmHwxbrDLPpuP4YB7RpX57khnfDRJDgREXHQwZOpvBodS1pmPv4+HrwwNJwWoVXNjmUqFVhE5IYZhsGqTceYv3ovVptBUGAlnhvciSZ1A8yOJhVIVFQUy5cv58CBA3h7e9O1a1dmzJhB06ZNr/kccXFx2O12QkJCSjCpiJR3x8+l868Vu9l7LAWAGoGVGHlvKyJa1tSkAxERKVUFhTY++Hwn67efAeCubqGMvLeVmtmLiIjD1sef4f1lOyi02qlfy4/JwyOoEVjJ7FimU4FFRG5IZk4B7326g5i9iQB0aV2LcYPa4+OtZU+kdG3YsIHRo0cTFhaG1Wpl0qRJ9OrVi3379l3T0ygpKSkMGTKEOXPmlEJaESmPMnMK+GTNAb779Th2AzzcXbn/1sb079EIT3dXs+OJiEgFk5aRx6vzYzl4Mg0XFwtP9G9Nn66hZscSEREnY7cbLF6zn89/Ogxcbgcw/uEOWvL4/6nAIiLX7cCJVGYujuNiWi5uri6M7NuSPt1CNTtXTLFmzZortqOjowkKCmL79u1ERkb+4Xvz8/Pp378/EydOpGvXrn/4uvz8/KLtjIyMGwstIuXGrsMXmbkojozsAgC6ta3N8LtbEqQZXSIiYoKjZy7xSnQsyZdy8fF25/nHwmjbuLrZsURExMnk5lt5Z2k8W3afB2DgLY0ZfGdzXCpgM/vf43CBxWazMX/+fH766SeSkpKw2+1XHF+3bl2xhRORssluN1ix/ggLv9uP3W5Qq1plnhvciYbBVcyOJlIkPT0dgMDAwD98nWEYDB06lFtuuYXBgwf/4WujoqKYNm1asWUUkfJhffwZ3vs0HqvNIKSGL0/0b62bWCIiYppfE87x9tJ48gts1Knuw4sjIqhd3cfsWCIi4mSS0nJ4ZV4Mx89l4ObqwtgH2nFLJy2p/r8cLrA89dRTzJ8/n7vuuotWrVppprpIBZOelc87S+PZfiAJgMh2dRh9f1s9FihlimEYjB8/nu7du9OqVSsAevfuTXx8PNnZ2QQHB7NixQrCwsL45ZdfWLZsGW3atOGrr74CYNGiRbRu3fqq806cOJHx48cXbWdkZKhfi0gFZhgGK9YfJXr1XuDyUyvjH+qAh5YDExERExiGwWc/HmLxmgMAtG9SnQlDwrR8s4iIOOzAicvN7C9l5VPFx5NJw8JpVv+PJ7BWVA4XWD799FM+++wz+vTpUxJ5RKQM23sshZmL4kjNyMPDzYXH+7emV0Q9FVqlzBkzZgwJCQls3ry5aN/atWt/87Xdu3e/6mnM3+Pp6Ymnp2exZBQR52a3G8z9eg+rNh4DoO9NDRjRt5UelRcREVPkF9p4f9kONu44C8A9NzVgxD0tcVUzexERcdC6uNN88NlOrDY7obUvN7MPCtDSx7/H4ZHWw8ODRo0alUSW39S/f38CAgIYOHCgQ8dEpPjY7QbLfjzICx9uJjUjjzrVfXjzqUh6d66v4oqUOWPHjmXVqlX8/PPPBAcHmx1HRMqhgkIbbyyOKyquDLu7JSPvVXFFRETMkZKey8R/bmbjjrO4ulgYPbAtj/drreKKiIg4xGY3mL96L+8sjcdqs9OldS1mjLlJxZU/4fBo+/TTT/Pee+9hGEZJ5LnKuHHjWLhwocPHRKR4pGXm8dLsLSz+7gB2A3p2DOadv99MaG1/s6OJXMEwDMaMGcPy5ctZt24doaGhZkcSkXIoK7eQl2ZvYfOuc7i5Wnj6kY4M6NlIEw5ERMQUR05f4un3NnL49CV8K7kz/Ymu3NGlvtmxRETEyeTkFfJadCxf/nwEgEG3NeH5IWF4ezq8AFaFc02f0IABA67YXrduHd999x0tW7bE3f3KtTyXL19efOmAnj17sn79eoePiciN23X4Im99sp20zHw8PVz5S/823BZe1+xYIr9p9OjRLFmyhJUrV+Lr60tiYiIA/v7+eHt7m5xORMqD5Eu5vDR7C6cSM/H2dGPS0HDaNlEzexERMcfmXWd5Z+kOCgpthNTw5cUREdSsWtnsWCIi4mQupF5uZn/ifAbubi48Nag9N3fQiiDX6poKLP7+V85U79+/f7EFiIqKYvny5Rw4cABvb2+6du3KjBkzaNq0abF9DxFxjM1u8On3B1n240EMA+rW9OW5wZ2oW9PP7Ggiv2vWrFkA9OjR44r90dHRDB06tPQDiUi5cvJ8BlNnbyE5PY9AP09eGtmFBnX0NKeIiJQ+w7h8vbbk+4MAdGpeg2cf7UglLzWzFxERx+w9lkLUgljSswoI8PVk8vAImtQNMDuWU7mmAkt0dHSJBdiwYQOjR48mLCwMq9XKpEmT6NWrF/v27aNyZc28ECltKem5vPVJPLuPJgNwe3hdHu/fGi8PPRIoZVtpLV0pIhXPnqPJvBIdS3ZuIcFBPkwb1YWgQK1DLCIipS+vwMp7n+5g865zAPS7uSFD726Jq/qAiYiIg36MPcU/v9iJ1WbQMNifycMiqFZFK4A4yuE7prfccgvLly+nSpUqV+zPyMigX79+rFu3zqHzrVmz5ort6OhogoKC2L59O5GRkY7GE5EbsPNQEm9+sp30rAK8PFwZPbAtPTqGmB1LRETENJt3neWtTy43eWxeP5ApIyLwreRhdiwREamAUtJzeWVeDEfOpOPmauGv97Xl9oh6ZscSEREn8+9m9l9tOApAt7a1+duD7TW5+jo53OR+/fr1FBQUXLU/Ly+PTZs23XCg9PR0AAIDA2/4XP+Wn59PRkbGFV8icqVdhy8ybc5W0rMKCK3tx7vje6i4IiIiFdqqTUeZuSgOq81O51Y1mf6XriquCHB5yeSAgAAGDhx4xf7Tp0/To0cPWrRoQZs2bfj8889NSigi5c2hU2mMf3cDR86k41fZg1f+0k3FlXJA44mIlLacvEJemRdTVFx5qFdTJjzaScWVG3DNn1xCQkLRf+/bt6+oeTCAzWZjzZo11KlT54bCGIbB+PHj6d69O61atQKgd+/exMfHk52dTXBwMCtWrCAsLOxPj/23qKgopk2bdkPZRMqzk+czeG1+LFabQZfWtXjmkY54uLuaHUtERMQUdrvBgm/2sXz9EQD6dK3P4/3baPkVKTJu3DiGDx/OggULrtjv5ubGu+++S7t27UhKSqJDhw706dNHSx+LyA3ZuOMM7326gwKrnXo1fZk8XM3sywuNJyJSmhJTspk+L4ZTiZl4uLnwt4c6cFO7G7ufLw4UWNq1a4fFYsFisXDLLbdcddzb25sPPvjghsKMGTOGhIQENm/eXLRv7dq1v/v6Pzr23yZOnMj48eOLtjMyMggJ0cx8Ebj8mPnU2VvIybPSskFVFVdERKRCK7TaeX/ZDtbHnwFgSJ/mDLylMRaLiivyHz179mT9+vVX7a9Vqxa1atUCICgoiMDAQFJTU3VDTESui91usGTtAZb9eAiA8BY1efqRDmpmX45oPBGR0rL7aDJR87eRmVNAoJ8Xk4eH0zhEzeyLwzUXWI4fP45hGDRo0IDY2FiqV69edMzDw4OgoCBcXa//puzYsWNZtWoVGzduJDg4+LrP81s8PT3x9PQs1nOKlAc5eYVMnb2V5PQ8goN8mDQsXMUVERGpsHLyCnltfiy7Difj6mJh7APtuDWsrtmxxARRUVEsX76cAwcO4O3tTdeuXZkxYwZNmza95nPExcVht9s1sUtErktevpW3l8azZfd5AO7r2YjBfVroaUono/FERMqCtVtPMOvLBGx2g8YhVZg0LJyq/mpmX1yuucBSr97ltT3tdnuxBjAMg7Fjx7JixQrWr19PaGhosZ5fRH5bodVO1PxtnDifQRVfT6aO6qJ15UVEpMJKSc9l2pytHD+XgZeHK88/FkbHZjXMjiUm2bBhA6NHjyYsLAyr1cqkSZPo1asX+/btu6bZwykpKQwZMoQ5c+aUQloRKW8upl1uZn/sXDpuri6Mub+tCv5OSuOJiJjJZrMz7+u9rNp0DIDIdnUY92B7PDW5ulg53L1m1apVv7nfYrHg5eVFo0aNHCqSjB49miVLlrBy5Up8fX2Lerv4+/vj7a1KmkhJMAyDf3y+k52HL+Ll4cpLIzpTI7CS2bFERERMcfpCJlNnbyEpLZcqPp68NLIzjUKqmB1LTLRmzZortqOjowkKCmL79u1ERkb+4Xvz8/Pp378/EydOpGvXrn/4uvz8/KLtjIyMGwstIuXCgZOpvBody6XMfKr4ePLC0HCahwaaHUuuk8YTETFLdm4hMxfHEX8gCYBH7mjGoNuaaOnjEuBwgaVfv35YLBYMw7hi/7/3WSwWunfvzldffUVAwJ+v4zZr1iwAevToccX+6Ohohg4d6mg8EbkGn6w9wLq407i4WHhuSJhuIomISIW1/3gq0+dtJTOnkNrVKjPt8S5qHCxXSU9PByAw8I9vchqGwdChQ7nlllsYPHjwH742KiqKadOmFVtGEXF+P28/zQef7aTQaqd+LT+mDI8gSBPhyhWNJyJSGs4lZzF9bgxnkrLwcHdl/MMd6Namttmxyi2L8b+Vkj/x008/MWnSJF599VXCw8MBiI2NZfLkyUyZMgV/f3+eeOIJIiIimDt3bomEvlEZGRn4+/uTnp6On5+f2XFEStXarSf5x+c7ARhzfzt6d65nbiBxOvoZ+h/6LESc25bd53lzcRwFVjtN6lbhxRGd8fdR377S4iw/Qw3D4N577yUtLY1NmzYB0Lt3b+Lj48nOziYwMJAVK1YQFhbG5s2biYyMpE2bNkXvX7RoEa1bt77qvL814zgkJKTMfx4iUvzsdoNF3+3ni3WHAejcqibjH+6It6fDc2IrLGcYUzSeiEhp2HX4Iq8v2EZWbiHV/L2YPDyChsFVzI7lNK5nPHF4tH7qqaf4+OOPr3g88dZbb8XLy4vHH3+cvXv38u677zJ8+HBHTy0iJSxu/wU+/HIXAINua6LiioiIVFjf/Xqcj5YnYDcgrEUNJjzaCS/dyJLfMGbMGBISEti8eXPRvrVr1/7ma7t3737NPSs9PT3x9FRBT6Siy8238tYn24nZe3m59PtvbcyjdzTHRc3syx2NJyJS0r779Tj/WrEbm92gad0AXhgWTqCfl9mxyj2HryKPHj36m9UbPz8/jh273DCncePGJCcn33g6ESk2R05fYsbCbdjtBrd0CuGRO5qZHUlERKTUGYbB4jUH+OzHQwD07lyPJwe0wdXVxeRkUhaNHTuWVatWsXHjRoKDg82OIyLlTFJqDtPnxXDifAbubi6Me6AdPTqGmB1LSoDGExEpSTabnTkr97D6l+MA9OgQzNgH2uGhZvalwuECS8eOHXn22WdZuHAh1atXB+DixYtMmDCBsLAwAA4fPqwBQ6QMuZCaw7S5W8krsNGucXXG3N9OTa1ERKTCsdrsfPDZTtbFnQbg4d7NePB2NXqUqxmGwdixY1mxYgXr168nNDTU7EgiUs7sO57Ca/NjSc8qIMDXk0nDwmlaT83syxuNJyJS0rJyCpixMI6dhy9iscDgO5sz8JbGusYpRQ4XWObOncu9995LcHAwISEhWCwWTp06RYMGDVi5ciUAWVlZTJkypdjDiojjMnMKmDp7C5cy86lfy4+JQ8Nwd9MsXRERqVhy8628vmAb8QeTcHGx8Nf72mqpTPldo0ePZsmSJaxcuRJfX18SEy8v3ePv74+3t7fJ6UTE2f207RT/+HwXVpudBnX8mTwsguoB+tlSHmk8EZGSdPZiFtPnbuXsxWy8PFx5+pGOdG5Vy+xYFY7DTe7hcgV+7dq1HDp0CMMwaNasGbfffjsuLs5x09YZmp+JFIeCQhsvfryFvcdSqObvxZtPRVLVX7/EyY3Rz9D/0Gch4hzSMvN4ec5WjpxJx8PdleeGdCK8RU2zY1V4Zfln6O/N+IuOjmbo0KEl8j3L8uchIsXDZjdY+M0+lq8/AkDXNrX4+4Md1AOsGJTVn6EaT0SkpOw8lMTrC+PIzi2keoA3U4ZHEFrb3+xYTq9UmtzD5QHijjvu4I477riet4tIKbDbDd5ZGs/eYylU8nLjpVFdVFwREZEK59zFLF6avYXElBz8Knvw4ogILcEif+o65qCJiPyhnLxC3vxkO9v2XQBg0O1NeLhXMzWzL+c0nohISfhm8zE+XrkHu92gef1AJg4NI8BXzezNcl0Flp9++omffvqJpKQk7Hb7FcfmzZtXLMFE5MZEr97L5l3ncHO18MLQcOrX0swVERGpWA6dSmPanK1kZBdQs2olpo3qQu3qPmbHEhGRCiYxJZtX5sVwMjETDzcX/vZgB25qX8fsWCIi4mSsNjsff7Wb7349AcAtnUIYc39b3N3UzN5MDhdYpk2bxssvv0ynTp2oVauWGuaIlEFfbzrGVxuOAvDUoPa0bVzd5EQiIiKla9u+RGYsiiO/wEajYH9eHNlZs7pERKTU7TmazGvzt5GZU0CgnyeThkXQpG6A2bFERMTJZGQXMGPhNhKOJGOxwNC7WtC/RyPdmy8DHC6wfPTRR8yfP5/BgweXRB4RuUFbdp9j9srdAAzp05weHUNMTiQiIlK61m49yYdf7sJuN+jQNIjnHwvDW+vbi4hIKfs+5iSzvtyF1WbQKNifycMjtGyziIg47PSFTKbPjeF8Sjbenq4880gnwluqp2RZ4fCVZkFBAV27di2JLCJygw6cSOXNxdsxDLijS30G3tLY7EgiIiKlxjAMPv3hEEvWHgAuPzI/9oF2uLm6mJxMREQqEpvdIPrrvazceHlVge5ta/PUg+3x8lCxX0REHLP9wAVmLoojJ89KUGAlpgyPUBuAMsbh0X3kyJEsWbKEKVOmlEQeEblOZy9m8fLcGAqsdsJa1OAv/VvrMUEREakwbDY7s5YnsHbrSQAeuK0Jj97RTGOhiIiUquzcQmYujiP+QBIAD/duxoO3N9F4JCIiDjEMg683HWPuqj3YDWjZoCoTHwvD38fT7GjyPxwusOTl5fHxxx/z448/0qZNG9zd3a84/vbbbxdbOBG5Npcy85k6ewuZOQU0CqnChEc74arZuiIiUkHk5VuZuTiObfsu4GKBJwa0oU/XULNjiYhIBXM+OZvp87Zy+kIWHu6u/P2h9nRvq2b2IiLimEKrnY+WJ/B9zOXJY7eH1+XJ+9ri7qZ7fWWRwwWWhIQE2rVrB8CePXuuOKYZGSKlLy/fyvR5W0lMyaFGYCVeHBGBl9aZFxGRCiI9K5/pc2M4eCoNDzcXnnm0E11a1zI7loiIVDC7jyQTtSCWzJxCqvp7MXlYBI1CqpgdS0REnEx6Vj6vL9zGnqMpuFhg2D0tuTeyoe67l2EO34X9+eefSyKHiFwHm93gzU+2c+jUJXwruTN1VGcCfL3MjiUiIlIqElOyeenjLZxLzsa3kjtThnemeWig2bFERKSCWbPlBB8tT8BmN2gcUoXJwyMI9NN1mYiIOOZkYgavzIshMSUHb083JgzuRKfmNcyOJX/iuqe5HzlyhKNHjxIZGYm3tzeGYaiSJlKKDMPgXysSiNmbiLubC5OHRxAc5Gt2LBERkVJx5Mwlps3ZyqXMfIICvJk6qgshNTQOiohI6bHZ7MxZtYfVm48DcHP7YMYOaoenu6vJyURExNls25fIG4u3k5tvpWbVy83s69ZUM3tn4HCBJSUlhQceeICff/4Zi8XC4cOHadCgASNHjqRKlSq89dZbJZFTRP7Hlz8f4btfT2CxwNOPdKRFaFWzI4mIiJSK+INJvL4gltx8G6G1/XhpZGeq+nubHUtERCqQrNxCZi7cxo5DFwEYfGdz7r+1sSaeioiIQwzD4KsNR4levRfDgFYNq/L8EDWzdyYOd8b5+9//jru7O6dOnaJSpUpF+wcNGsSaNWuKNZyI/Lb18WdY8M0+AEb2bUW3NrVNTiQiIlI61sWd4uU5W8nNt9G2cTVeH91dxRURESlV5y5m8cx7G9lx6CKeHq68MDSMB25rouKKiIg4pNBq4/1lO5n39eXiSu/O9Xj58a4qrjgZh59g+f7771m7di3BwcFX7G/cuDEnT54stmAi8tt2H0nmvU/jAbg3siF9IxuanEhERKTkGYbBF+sOs/Db/QD06BDMuEHtcXdzeL6QiIjIddt5KIkZC+PIyi2kWhVvpgyPoEEdf7NjiYiIk7mUmU/Uglj2HU/FxQIj7m3FPd0bqFjvhBwusGRnZ1/x5Mq/JScn4+mp6ppISTqZmMGr0TFYbQbd2tRm+D0tzY4kIiJS4mx2g49XJPDtrycAuK9nI4b0aYGLiy4+RESk9Hzzy3E+/mo3drtBs3oBvDAsnABfNbMXERHHnDifwfS5W0lKy6WylxsThoTRoWmQ2bHkOjk85S8yMpKFCxcWbVssFux2O2+88QY9e/Ys1nAi8h8p6blMnb2V7DwrzesHMv7hDrqxJCIi5V5+oY0ZC7fx7f/3HRvVrxVD726pMVBEREqN1Wbnwy938dHyBOx2g54dg3n1yW4qroiIiMNi9pxnwgcbSUrLpVa1yrwxLlLFFSfn8BMsb7zxBj169CAuLo6CggImTJjA3r17SU1N5ZdffimJjCIVXk5eIdPmbCX5Ui51qvsweXgEHu6uZscSEREpUZk5BUyfG8P+E6m4ubrw9CMd6N62jtmxRESkAsnMKWDGwm3sOpyMxQJD+rTgvp6NtISLiIg4xDAMvvz5CAu/3YdhQJtG1Xj+sTB8K3mYHU1ukMMFlhYtWpCQkMCsWbNwdXUlOzubAQMGMHr0aGrVqlUSGUUqNKvNTtSCbRw/l0EVH0+mjuqMX2X98BURkfItKTWHqXO2cPpCFpW93Jg0PILWDauZHUtERCqQ0xcymT4vhvPJ2Xh5uPLMIx2JaKX7HiIi4piCQhv//GIX6+JOA3Bn1/o83q81bq7qJ1keOFxgAahZsybTpk27Yt/p06cZPnw48+bNK5ZgInK5uv2Pz3ey89BFPD1ceXFkBDWrVjY7loiISIk6fi6dqbO3kJqRTzV/L6aO6kK9Wn5mxxIRkQok/mASMxduIzvPSlCAN5OHRxBaW83sRUTEMWmZebwWHcuBk2m4uFh4vF9r7uoWanYsKUbXVWD5LampqSxYsEAFFpFitPT7g/y07TQuFnhucCcahwSYHUlERKRE7Tp8kVejY8nNt1Kvpi9TR3WhWhVvs2OJiEgFYRgGX28+xtyVe7Ab0Lx+IC8MDaeKr6fZ0URExMkcO5vO9HkxJF/KpbK3O88P6US7Juq3Ut4UW4FFRIrX9zEnWfr9QQCevK8tYS1qmpxIRESkZG2IP8O7n8ZjtRm0aliVScMi8PF2NzuWiIhUEFabnY+WJ7B260kAbg0LYfTAtri7qf+liIg4Zsvuc7y1JJ78Aht1qldmyojO1KnuY3YsKQEqsIiUQdsPXOCfX+wC4P5bG3NHl/rmBhIRESlhK9YfYd7XewHo1rY24x/qgIe7bmiJiEjpyMgu4PUF29h99HIz+2F3t6TfzQ3VzF5ERBxiGAaf/3SYRd/tB6Bdk+o8N7gTPmpmX26pwCJSxhw9c4kZC7dhtxv07BjM4Dubmx1JRESkxNjtBnO/3sOqjccA6BvZgBH3tMLFRTe0RESkdJxKzGD6vBgSU3Lw9nTj2Uc7agUBERFxWH6hjQ+W7WTDjjMA3N09lJF9W+GqZvbl2jUXWAYMGPCHxy9dunSjWUQqvKTUHKbN2Upuvo22jasx9oH2mjElIiLlVkGhjXeWxrN51zkAht/Tkv49GpmcSkREKpK4/Rd4Y3EcOXlWagRWYsqICOrV9DM7loiIOJnUjDxejY7h0KlLuLpYeGJAG+7UijQVwjUXWPz9/f/0+JAhQ244kEhFlZVTwNQ5W0jLzKd+LT8mPhaOu5sq3CIiUj5l5RbyanQMe46m4OZq4W8PduDmDsFmxxIRkQrCMAxWbjxK9Nd7sRvQskFVJj4Whr+PmtmLiIhjjpy5xKvzYkhOz8PH252JQ8No06i62bGklFxzgSU6Orokc4hUaIVWG69Ex3L6QhZV/b14aWRnKqupr4iIlFPJl3KZOnsLJxMz8fZ0Y9KwcNo21gWIiIiUjkKrnVlf7uKH2FMA9Iqox18GtNEENxERcdgvu87x9tJ4CgptBAf5MGVEBLWrqZl9RaIeLCIms9sN3lm6g73HUqjk5cZLIztTrYq32bFERERKxMnEDKZ+vIXk9DwC/TyZOqoLobX/+ElpERGR4pKelU/Ugm3sPZaCiwVG9G3FPTc10NLMIiLiEMMw+PSHQyxZewCADs2CmPBoJ02YroBUYBEx2YJv9rFp51lcXSy88Fi4bjKJiEi5tedoMq9Ex5KdW0hIDR+mjuxCUGAls2OJiEgFcfJ8Bi/PiyEpNYdKXm5MGNyJjs1qmB1LREScTF6Blfc+3VHUS7JvZAOG391SzewrKP1fFzHR6s3HWL7+CADjBrWnbRMtjyJSUvr3709AQAADBw506JiIFI9fdp1jyr+2kJ1bSPP6gcwYc5OKKyIiUmpi9yXy7AcbSUrNoVbVyrw5LlLFFRERcVhKei4TP/yFzbvO4epiYcz97Rh1b2sVVyow/Z8XMcmW3ef5+KvdADx6ZzNu6RRiciKR8m3cuHEsXLjQ4WMicuO+3nSMGYu2YbXZ6dK6FtP/0hXfSh5mxxIRkQrAMAyW/3yYV+bFkJtvo02jarz5VCQhNXzNjiYiIk7m0Kk0xr+7gSOnL+FbyYPpf+lK7871zI4lJtMSYSImOHAylTcXx2EY0LtzPR64tYnZkUTKvZ49e7J+/XqHj4nI9bPbDRZ+u48vf778tGafrvV5vH8bXF20zr2IiJS8QquNf3y+i3VxpwG4s0t9Hu/fGjfNMhYREQdt2nGWdz+Np8Bqp25NX6YMj6Bm1cpmx5Iy4Lp+q1i0aBHdunWjdu3anDx5EoB3332XlStXFms4kfLoXHIW0+fGUGC106l5DZ4c0EYNFUWKQVRUFGFhYfj6+hIUFES/fv04ePCg2bFEKqxCq513lsYXFVeG9GnOXwaouCIiIqXjUmY+k2b9yrq407i4WHiif2uevK+NiisiIuIQu91g8Zr9zFwcV3Qv742xN6m4IkUc/s1i1qxZjB8/nj59+nDp0iVsNhsAVapU4d133y3ufCLlSnpWPlM/3kpGdgGNgv2ZMLiT1mgUKSYbNmxg9OjRbN26lR9++AGr1UqvXr3Izs42O5pIhZOTV8jLc7ayPv4Mri4W/v5Qe+6/tYkmFIiISKk4fi6d8e9tYP+JVCp7uTF1ZGfu7t5A45CIiDgkL9/KjEXbWPbDIQD692jE5OERVPJyNzmZlCUOLxH2wQcfMHv2bPr168frr79etL9Tp04888wzxRpOpDzJK7Dy8tytnE/JJiiwEi+O6Iy3p1bpEykua9asuWI7OjqaoKAgtm/fTmRkZLF8j/z8fPLz84u2MzIyiuW8IuVJakYeU2dv4fi5DLw8XJn4WDgdmgWZHUtERCqIrXvO89Yn28krsFG7WmWmjIggOEj9VkRExDHJl3KZPi+GY2fTcXO1MHpgO24Lr2t2LCmDHL67e/z4cdq3b3/Vfk9PT80SFvkdNrvBm4u3c+jUJXy83Zk6sjMBfl5mxxIp19LT0wEIDAwstnNGRUUxbdq0YjufSHlz+kImU2dvISktlyo+nrw0sjONQqqYHUtERCoAwzD4Yt1hFn23H8OAdo2r89yQTvhU8jA7moiIOJmDJ1N5NTqWtMx8/H08eGFoOC1Cq5odS8oohwssoaGh7Ny5k3r16l2x/7vvvqNFixbFFkykvDAMg9lf7SZmbyLubi5MHh5BSA3NoBIpSYZhMH78eLp3706rVq0A6N27N/Hx8WRnZxMcHMyKFSsICwv702P/beLEiYwfP75oOyMjg5CQkNL5Q4mUcfuPpzJ93lYycwqpXa0y0x7vonWJRUSkVBQU2vjgs52sjz8DwF3dQhl5byv1WxEREYetjz/D+8t2UGi1U7+WH5OHR1AjsJLZsaQMc7jA8uyzzzJ69Gjy8vIwDIPY2FiWLl1KVFQUc+bMKYmMIk5txfojfPPLcSwWGP9wB1o2UMVbpKSNGTOGhIQENm/eXLRv7dq1v/v6Pzr23zw9PfH09LzhfCLlzdY953lj0eWmj03rBjBlRAT+Pvq3IiIiJS8tI49Xo2M5eCqtqJl9n66hZscSEREn8+9m9p//dBiAiJY1Gf9wB/VbkT/lcIFl2LBhWK1WJkyYQE5ODg8//DB16tThvffe48EHHyyJjCJOa+OOM0Sv3gfA8Hta0b1tHZMTiZR/Y8eOZdWqVWzcuJHg4GCz44iUe9/9epyPlidgNyCsRQ0mDO6El4d6jImISMk7euYSr8yLITk9Dx9vd55/LIy2jaubHUtERJxMbr6Vt5dsZ+ueRAAG3tKYwXc2x8XFYnIycQbXdfU7atQoRo0aRXJyMna7naAgNS4V+V+7jybzztIdAPS9qQH9bm5ociKR8s0wDMaOHcuKFStYv349oaGauShSkgzDYPGaA3z24yEAeneux5MD2uCq5VhERKQU/JJwjneWxpNfYKNOdR9eHBFB7eo+ZscSEREnk5Saw/R5MZw4n4GbqwtjH2jHLZ20FLhcO4evgI8fP87hw5cflapWrVpRceXw4cOcOHGiWMOJOKtTiRm8Gh2L1WanS+taDO/byuxIIuXe6NGjWbx4MUuWLMHX15fExEQSExPJzc01O5pIuWO12Xlv2Y6i4srDvZsxemBbFVekwujfvz8BAQEMHDjQoWMicuMMw+DTHw7y+oJt5BfY6NA0iDefilRxRZySxhMRc+0/nsrT723kxPkMqvh4EvXXbiquiMMcvgoeOnQov/7661X7Y2JiGDp0aHFkEnFqKem5TJ2zlezcQprXD+TpRzriqkcKRUrcrFmzSE9Pp0ePHtSqVavoa9myZWZHEylXcvOtTJ8Xw0/bTuPiYmHsA+14qFdTLBaNdVJxjBs3joULFzp8TERuTH6hjTcXb+eTNQeAyysFvDgiAh9vrY8vzknjiYh51sWd4oVZv3ApK5/Q2n689bdImtUPNDuWOCGHCyw7duygW7duV+3v3LkzO3fuLI5MIk4rJ6+Ql+fEcDEtl9rVKjNpWDie7q5mxxKpEAzD+M0vFf9Fik9aZh4vfLiZ+ANJeHq4MnlYOL0i6pkdS6TU9ezZE19fX4ePicj1S0nPZeI/N7Nx51lcXSyMub8to/q11tOT4tQ0noiUPpvdYP7qvbyzdEfRyjMzxtxEUEAls6OJk3L4NxGLxUJmZuZV+9PT07HZbMUSSsQZWW12Xl+wjWPn0qni48m0x7vg7+NpdiwREZFice5iFhM+2MSRM+n4VfbgtSe7EdaiptmxREpEVFQUYWFh+Pr6EhQURL9+/Th48KDZsUQqrMOn0xj/7kYOn76EbyUPpv+lK7071zc7lsif0ngiUrbk5BXyWnQsX/58BIAHbmvC80PC8Pa8rjblIsB1FFhuuukmoqKiriim2Gw2oqKi6N69e7GGE3EWhmHwz893sePQRTw9XJkyIoKaVSubHUtERKRYHDqVxrMfbCIxJYeaVSvxxtibaFI3wOxYIiVmw4YNjB49mq1bt/LDDz9gtVrp1asX2dnZZkcTqXA27TzL8//8hdSMPEJq+PL23yJp3bCa2bFEronGE5Gy40JqDhM+2ETsvkTc3Vx4+pGODL6zOS5a1l9ukMPluZkzZxIZGUnTpk256aabANi0aRMZGRmsW7eu2AOKOINPvz/Ij9tO4WKBCYM76aaTiIiUG9v2JTJjURz5BTYaBfvz4sjOBPh6mR1LpEStWbPmiu3o6GiCgoLYvn07kZGRxfI98vPzyc/PL9rOyMgolvOKlBd2++Vm9ku/vzzbv1PzGjz7aEcqeanfijgPjSciZcPeYym8Nj+WjOwCAnw9mTQsnKb11G9FiofDT7C0aNGChIQEHnjgAZKSksjMzGTIkCEcOHCAVq1alURGkTLtx9iTLPn/X/r/MqAN4VouRUREyonvY07ySnQs+QU2OjQL4rW/dldxRSqk9PR0AAIDi+9CPCoqCn9//6KvkJCQYju3iLPLK7Ayc3FcUXGl380NmTw8QsUVcXoaT0RK34+xJ5n80S9kZBfQMNift/92s4orUqwshmEYZocobRkZGfj7+5Oeno6fn5/ZccSJxR9I4uW5W7HZDQbe0pjH7mphdiSREqefof+hz0LKK8Mw+PSHQyxZewCAW8NCGHN/O9zUSFiKkbP8DDUMg3vvvZe0tDQ2bdoEQO/evYmPjyc7O5vAwEBWrFhBWFjYnx77b7814zgkJKTMfx4iJS35Ui6vRMdw9Ew6bq4W/npfW26PqGd2LCnjnGFM0XgiUrr+3cz+qw1HAejWtjZ/e7A9Xh7qtyK/73rGk+v6G3Xp0iViY2NJSkrCbrdfcWzIkCHXc0oRp3P0zCVeXxiLzW7Qo0Mwg+9sbnYkERGRG2az2Zm1PIG1W08Clxs/PnpHMywWrU0sFdOYMWNISEhg8+bNRfvWrl37u6//o2P/zdPTE09PzxvOJ1KeHDyZyqvRsaRl5uNX2YMXhobTskFVs2OJFAuNJyKlJzu3kDc/2U7c/gsAPNSrKQ/e3lT9VqREOFxg+frrr3nkkUfIzs7G19f3iotti8WiAotUCElpObw8dyu5+TbaNKrGuEHt9UNaREScXl6BlTcWbSd2XyIulstLX97ZNdTsWCKmGTt2LKtWrWLjxo0EBwebHUekXFsff4b3l+2g0GqnXk1fJg+PoGbVymbHEikWGk9ESs/55Gymz9vK6QtZeLi58LeHOnBTuzpmx5JyzOECy9NPP83w4cN57bXXqFSpUklkEinTsnIKmDp7K6kZ+dSt6cvEoeG4u2nJFBERcW7pWflMnxvDwVNpeLi58OzgTnRuVcvsWCKmMAyDsWPHsmLFCtavX09oqAqNIiXFbjf4ZO0BPvvxEADhLWry9CMd1G9FygWNJyKla/eRZKIWxJKZU0ignxeTh4fTOCTA7FhSzjlcYDl79izjxo1TcUUqpEKrjVfnx3L6QiaBfl5MHdkFH2/94i8iIs4tMSWblz7ewrnkbHwruTNleGeah6rxo1Rco0ePZsmSJaxcuRJfX18SExMB8Pf3x9vb2+R0IuVHbr6Vd5bGs2X3eQDu69mIwX1a4KrVAaSc0HgiUnrWbj3BrC8TsNkNGodUYdKwcKr669+ZlDyHCyy9e/cmLi6OBg0alEQekTLLbjd4d+kO9hxNwdvTjamjOlM9QD+oRUTEuR05c4lpc7ZyKTOfoABvpo7qQkgNX7NjiZhq1qxZAPTo0eOK/dHR0QwdOrT0A4mUQ0lpObwyL4bj5zJwc3Vh7ANtuaVTXbNjiRQrjSciJc9mszP36718vekYAJHt6jDuwfZ4uruanEwqCocLLHfddRfPPvss+/bto3Xr1ri7Xzl7v2/fvsUWDqB///6sX7+eW2+9lS+++OKKY6tXr+bpp5/Gbrfz3HPPMXLkyGL93iL/beG3+9i48yyuLhYmPhZGaG1/syOJiIjckPiDSby+IJbcfBuhtf14aWRnzfIS4fKSLiJScg6cSOXV+bFcysynio8nLwwN15OTUi5pPBEpWVm5hbyxKI74g0kAPHJHMwbd1uSKnuEiJc3hAsuoUaMAePnll686ZrFYsNlsN57qv4wbN47hw4ezYMGCK/ZbrVbGjx/Pzz//jJ+fHx06dGDAgAEEBuqXMil+32w+xpc/HwFg7APtaN80yOREIiIiN2Zd3CneX7YTm92gbeNqvDA0XOvdi4hIiVsXd5oPPtuJ1WYntLYfk4dFEBSoJchFRMQx5y5m8fLcGM5ezMLTw5W/P9SBbm1qmx1LKiCHCyx2u70kcvyunj17sn79+qv2x8bG0rJlS+rUqQNAnz59WLt2LQ899FCp5pPyL2bPeT7+ajdwuRJ+a5geWxcREedlGAZfrDvMwm/3A9CjQzDjBrXH3c3F5GQiIlKe2e0GC7/dVzRxrXOrmox/uCPeng7flhARkQpu1+GLvL5gG1m5hVTz92Ly8AgaBlcxO5ZUUDd0JZ2Xl3fDAaKioggLC8PX15egoCD69evHwYMH//R9586dKyquAAQHB3P27NkbziPy3w6eTGXm4u3YDegVUY9BtzUxO5KIiMh1s9kNPlqeUFRcua9nI/7+UAcVV0REpETl5BXy2vzYouLK/bc2ZuJj4SquiIiIw7799TgvfryFrNxCmtYN4K2/3aziipjK4atpm83G9OnTqVOnDj4+Phw7drmB0JQpU5g7d67DATZs2MDo0aPZunUrP/zwA1arlV69epGdnf2H7/utdSy1vp4Up3PJlx81LCi00aFZEE/e10Z/x0RExGnlF9qYsXAb3/56AosFRvVrxdC7W+LiorFNRERKTlJqDs/9YzMxexNxd3Ph6Yc7MKRPC40/IiLiEJvNzkfLE5j1ZQJ2u0GPDsG89tduBPp5mR1NKjiHp4u8+uqrLFiwgJkzZxb1YwFo3bo177zzDiNGjHDofGvWrLliOzo6mqCgILZv305kZOTvvq9OnTpXPLFy5swZIiIifvO1+fn55OfnF21nZGQ4lFEqnvSsfKbO3kpGdgENg/15fkgYbq6a3SsiIs4pM6eA6XNj2H8i9f9vbnWkW1utTywiIiVr3/EUXpsfS3pWAQG+nkwaFk7TeuqbKiIijsnKKWDGwjh2Hr4IwJA+zRl4S2NNhJYyweE7xgsXLuTjjz/mkUcewdXVtWh/mzZtOHDgwA0HSk9PB/jTZvXh4eHs2bOHs2fPkpmZybfffkvv3r1/87VRUVH4+/sXfYWEhNxwTim/8gqsTJ8Xw/nkbIICvHlpRGc9ui4iIk4rKTWHCR9sYv+JVCp7u/Py411UXBERkRL3Y+wpJs36hfSsAhrU8eetp25WcUVERBx2JimTp9/byM7DF/HycGXSsHDuv7WJiitSZjh81/js2bM0atToqv12u53CwsIbCmMYBuPHj6d79+60atUKgN69exMfH092djbBwcGsWLGCsLAw3NzceOutt+jZsyd2u50JEyZQtWrV3zzvxIkTGT9+fNF2RkaGiizym2x2g7c+2c7Bk2n4eLszdVQXAvSooYiIOKnj59KZOnsLqRn5VPP3YurjXahX08/sWCIiUo7Z7AbzV+/lqw1HAejaphZ/f7ADXpq0JiIiDtpxMIkZi+LIzi2keoA3U4ZHEFrb3+xYIldw+Decli1bsmnTJurVq3fF/s8//5z27dvfUJgxY8aQkJDA5s2bi/atXbv2d1/ft29f+vbt+6fn9fT0xNPT84aySflnGAZzvtrN1j2JuLm6MHl4BCE1fM2OJSIicl12Hb7Iq9Gx5OZbqVfTl6mjulCtirfZsUREpBzLySvkjcXbidt/AYBBtzfh4V7N1G9FREQcYhgGqzcfZ86qPdjtBs3rBzJxaBgBvpoELWWPwwWWl156icGDB3P27FnsdjvLly/n4MGDLFy4kNWrV193kLFjx7Jq1So2btxIcHDwdZ9H5HqtWH+U1b8cB2D8Qx1o2eC3n4gSEREp6zbEn+HdT+Ox2gxaNazKpGER+Hi7mx1LRETKscSUbKbPi+FUYiYebi787cEO3NS+jtmxRETEyVhtdv61YjdrtpwA4JZOIYy5vy3ubq5//EYRkzhcYLnnnntYtmwZr732GhaLhRdffJEOHTrw9ddfc/vttzscwDAMxo4dy4oVK1i/fj2hoaEOn0PkRm3acZbo1XsBGH5PS10IiIiIUzIMgxXrjxaNad3b1mb8wx10MSIiIiVq99FkouZvIzOngEA/TyYNi6BJ3QCzY4mIiJPJyC7g9QXb2H00GYsFht7Vgv49GqnfipRp17UIau/evX+3obyjRo8ezZIlS1i5ciW+vr4kJiYC4O/vj7e3lrGQkrfnaDJvL40H4O7uofS7uaHJiURERBxntxvM/XoPqzYeA6BvZANG3NNKy7KIiEiJWrv1JLO+3IXNbtAopAqTh4VT1V/X8iIi4pjTFzKZPjeG8ynZeHu68swjnQhvWdPsWCJ/yvQuc7NmzQKgR48eV+yPjo5m6NChpR9IKpTTFzJ5JToWq81Ol9a1GHlva1XFRUTE6RQU2nhnaTybd50DLj+N2b9HI5NTiYhIeWaz2Zm3em9RYf+mdnUYN6gdXh6m32YQEREns/3ABWYuiiMnz0pQYCWmDI+gfi0/s2OJXBOHf/NxcXH5wxvQNpvNofMZhuFoBJFikZqRx9TZW8jOLaRpvQCefqQjrprlKyIiTiYrt5BXo2PYczQFN1cLf3uwAzd3UD87EREpOdm5hcxcHEf8gSQAHrmjGYNua6LJaiIi4hDDMFi16RjzVu3BbkDLBlWZ+FgY/j6eZkcTuWYOF1hWrFhxxXZhYSE7duxgwYIFTJs2rdiCiZSknLxCps3ZSlJaLrWqVWbK8Ag83bU+vYiIOJfkS7m8NHsLpxIz8fZ0Y9KwcNo2rm52LBERKcfOJWcxfW4MZ5Ky8HB3ZfxDHejWtrbZsURExMkUWu18tDyB72NOAnB7eF2evK8t7m4uJicTcYzDBZZ77733qn0DBw6kZcuWLFu2jBEjRhRLMJGSYrXZmbEojmNn0/H38WDaqC6qjIuIiNM5eT6DqbO3kJyeR6CfJ1NHdSG0tr/ZsUREpBxLOHKR1xdsIzOnkKr+XkweHkGj4CpmxxIRESeTnpVP1IJt7D2WgosFht3TknsjG+pJSHFKxbY4akREBKNGjSqu04mUCMMw+PCLXcQfSMLD3ZUXR3SmVrXKZscSERFxyJ6jybwSHUt2biEhNXyYOrILQYGVzI4lIiLl2HdbTvCv5QnY7AZN6lZh0rAIAv28zI4lIiJO5mRiBtPnxnAhNQdvTzcmDO5Ep+Y1zI4lct2KpcCSm5vLBx98QHCw1vuWsm3Zj4f4IfYULhaY8GhHmtQNMDuSiIiIQ37ZdY43P9mO1Wanef1ApoyIwLeSh9mxRESknLLZ7MxZtYfVm48DcHP7YMYOaqcllkVExGHb9iXyxuLt5OZbqVn1cjP7ujXVzF6cm8MFloCAgCse1zIMg8zMTCpVqsTixYuLNZxIcfox9hSfrDkAwOP92xDRqpbJiURERByzatNR5qzcg2FAl9a1ePqRjrrBJSIiJSYrp4AZi+LYeegiAIPvbM79tzbWEi4iIuIQwzBYsf4o87/Zi2FAq4ZVeX6ImtlL+eBwgeWdd9654pcpFxcXqlevTkREBAEBehpAyqb4g0n84/OdANzXsxF3dQs1N5CIiIgD7HaDBd/sY/n6IwD06Vqfx/u3wdVFN7hERKRknL2YxfS5Wzl7MRtPD1eefrgDXVqrmb2IiDim0Grjn1/s4qdtpwHo3bkeT/Rvo2b2Um44XGAZOnRoCcQQKTnHz6Xz+oJt2OwGke3rMKRPC7MjiYiIXLNCq533l+1gffwZAIb0ac7AWzR7WERESs7OQ0m8vjCO7NxCqlXxZsrwCBrU8Tc7loiIOJlLmfm8Nj+W/SdScbHAiHtbcU/3BrqWkXLF4QJLQkLCNb+2TZs2jp5epFhdTMtl6uyt5OZbad2wGn97sD0umu0rIiJOIievkNfmx7LrcDKuLhbGDWrHLZ3qmh1LRETKsW82H+PjlXuw2w2a1QvghWHhBPiqmb2IiDjm+Ll0XpkXQ1JaLpW93JgwJIwOTYPMjiVS7BwusLRr1+5Pq4yGYWCxWLDZbNcdTORGZeUWMnXOFlIz8gip4csLw8Jxd9M69SIi4hxS0nOZNmcrx89l4OXhysTHwunQTBckIiJSMqw2Ox9/tZvvfj0BQM+OwYy5vx0e6vUlIiIO2rrnPG99sp28Ahu1qlVmyvAIQmr4mh1LpEQ4XGBZvnw5zzzzDM8++yxdunQBYMuWLbz11lvMnDmT9u3bF3tIEUcVWm1EzY/lVGImgX6eTB3VGR9vd7NjiYiIXJPTFzKZOnsLSWm5VPHx5KWRnWkUUsXsWCIiUk5l5hTw+oJtJBxJxmKBIX1acF/PRlrCRUREHGIYBl+sO8yi7/ZjGNCmUTWefywM30oeZkcTKTEOF1hee+013n//ffr06VO0r02bNoSEhDBlyhS2b99erAFFHGW3G7y/bCcJR5Lx9nTlpZFdCAqoZHYsERGRa7L/eCrT520lM6eQ2tUqM+3xLtSsWtnsWCIiUk6dvpDJ9HkxnE/OxtvTlacf7khEq1pmxxIRESdTUGjjH5/v5Oftl3tH3tm1Po/3a42bq5rZS/nmcIFl9+7dhIaGXrU/NDSUffv2FUsokRux6Lv9rI8/g6uLhecfC1czRhERcRpbdp/nzcVxFFjtNK0bwJQREfj7eJodS0REyqn4A0nMXLSN7DwrQYGVmDI8gvq1/MyOJSIiTiYtI49X58dy8GQaLi4WHu/Xmru6XX3/WKQ8crjA0rx5c1555RXmzp2Ll9flRnf5+fm88sorNG/evNgDijjiu1+P88W6wwCMub+dmmeJiIjT+PbX4/xreQJ2A8Ja1GDC4E54eTj8q5qIiMifMgyDrzcdY+6qPdgNaBEayAtDw1XUFxERhx07m870eTEkX8qlsrc7zw/pRLsmuh8nFYfDV+0fffQR99xzDyEhIbRt2xaAXbt2YbFYWL16dbEHFLlWsXsT+Wh5AgAP92rKbeF1TU4kIiLy5wzDYNF3+/n8p8sTBHp3rseTA9rgqkfpRUSkBBRa7fxrRQJrt54E4Lawuvx1YBvc3dTMXkREHPNrwjneXhpPfoGNOtUrM2VEZ+pU9zE7lkipcrjAEh4ezvHjx1m8eDEHDhzAMAwGDRrEww8/TOXKWh9czHHoVBozF8dhN+D28Lo82Kup2ZFERET+lNVm54PPdrIu7jQAD/duxoO3N1FTYRERKREZ2QVELYhlz9EULBYYdndL+t3cUOOOiIg4xDAMPvvpEIu/OwBAuybVeW5wJ3zUzF4qoOtad6JSpUo8/vjjxZ1F5LqcT87m5blbyS+w0aFpEH8d2FYXCCIiUubl5lt5fcE24g8m4eJiYfTAtvSKqGd2LBERKadOJWYwfV4MiSk5eHu6MWFwJzo1r2F2LBERcTL5hTbeX7aDjTvOAnB391BG9m2lJ/Clwrquv/mLFi2ie/fu1K5dm5MnLz9W/M4777By5cpiDSfyZ9Kz8pk6ewvpWQU0qO3Pc0M64aYf6CLyG/r3709AQAADBw686tjq1atp2rQpjRs3Zs6cOSakk4omLTOPFz7cTPzBJDw9XJk8LFzFFRERKTFx+y/wzPubSEzJoUZgJd4Yd5OKKyIi4rDUjMvXMRt3nMXVxcJfB7blif5a3lgqNof/9s+aNYvx48dz5513kpaWhs1mAyAgIIB33323uPOJ/K78QhuvzIvhXHI21QO8eWlUZyp5uZsdS0TKqHHjxrFw4cKr9lutVsaPH8+6deuIj49nxowZpKammpBQKopzF7N49v1NHDmTjl9lD157shthLWqaHUtERMohwzD4asMRps/dSm6+lZYNqvLWU5HUq+lndjQREXEyR05fYvy7Gzh06hK+ldx5+Yku3NmlvtmxREzncIHlgw8+YPbs2UyaNAk3t/+sMNapUyd2795drOFEfo/NbvDWJ9s5cDKNyt7uTB3ZmUA/L7NjiUgZ1rNnT3x9fa/aHxsbS8uWLalTpw6+vr706dOHtWvXmpBQKoJDp9J49oNNXEjNoWbVSrwx9iaa1A0wO5aIiJRDhVYb7y/bydxVe7Eb0CuiHtOf6Iq/j6fZ0URExMls3nWW5/65mZT0PEJq+PDmU5G0aVTd7FgiZYLDBZbjx4/Tvn37q/Z7enqSnZ1dLKFE/sy8VXvYsvs8bq4uTBoWTl3NwBKp8KKioggLC8PX15egoCD69evHwYMH//R9586do06dOkXbwcHBnD17tiSjSgW1bV8iL8z6hYzsAhoF+zNz7E3Uru5jdiwRESmH0rPymfzRr/y47RQuFhh1byvG3N8Wdzct4SIiItfOMAyWrj3AjIVxFBTa6NgsiDfGRlK7mq5jRP7N4Sb3oaGh7Ny5k3r1rlwn/LvvvqNFixbFFkzk93y14SirNh0D4O8Ptad1w2omJxKRsmDDhg2MHj2asLAwrFYrkyZNolevXuzbt4/KlSv/7vsMw7hqn8ViKcmoUgGt3XqSD7/chd1u0KFZEM8PCcPb0+Ffw0RERP7UifMZTJ+7laS0XCp5XW5m37GZ+q2IiIhj8gqsvPfpDjbvOgfAvZENGXZPS1xddL0s8t8cvrJ/9tlnGT16NHl5eRiGQWxsLEuXLiUqKkqNgaXEbd51lrmr9gAw7O4WRLYPNjmRiJQVa9asuWI7OjqaoKAgtm/fTmRk5O++r06dOlc8sXLmzBkiIiJ+87X5+fnk5+cXbWdkZNxgainvDMPg0+8PsuT7y09T3RoWwpj72+GmJpAiIlICYvcm8uYnceTm26hVtTJTRkQQUuPqJVJFRET+SEp6Lq/Mi+HImXTcXC08eV9bekXU+/M3ilRADhdYhg0bhtVqZcKECeTk5PDwww9Tp04d3nvvPR588MGSyCgCwN5jKby9JB6Au7qF0r9HI5MTiUhZlp6eDkBgYOAfvi48PJw9e/Zw9uxZ/Pz8+Pbbb3nxxRd/87VRUVFMmzat2LNK+WSz2fnwywS+jzkJwAO3NeHRO5rpCSkRESl2hmGw/OcjLPh2H4YBbRpV47khYfhV9jA7moiIOJlDp9J4NTqG1Ix8fCt58MLQMFpp9RiR33Vda1OMGjWKUaNGkZycjN1uJygoqLhziVzh9IVMXpkXQ6HVTkTLmozq11o3qETkdxmGwfjx4+nevTutWrUCoHfv3sTHx5OdnU1wcDArVqwgLCwMNzc33nrrLXr27IndbmfChAlUrVr1N887ceJExo8fX7SdkZFBSEhIqfyZxLnk5VuZuTiObfsu4GKBvwxow51dQ82OJSIi5VCh1cY/Pt/FurjTANzZpT6P92+tpyVFRMRhG3ec4b1Pd1BgtVO3pi9ThkdQs+rvL7ktItfR5P6/VatWjf379/Pdd9+RlpZWXJlErpCWkcfUOVvJyi2kad0Annm0o9Z7FJE/NGbMGBISEli6dGnRvrVr13Lx4kVycnI4c+YMYWFhRcf69u3LoUOHOHLkCI8//vjvntfT0xM/P78rvkT+178bC2/bdwEPNxcmDg1XcUWkHOnfvz8BAQEMHDjwqmOrV6+madOmNG7cWMsnS6lIy8zjhQ9/YV3caVxcLPylf2v+OrCtiisiTkJjipQVdrvB4u/288bi7RRY7XRqXoM3xt6k4orINbjm37reeOMNXnrppaJtwzC444476NmzJ3fddRfNmzdn7969JRJSKq7cfCsvz91KUmoOtapdXkPYy0NNgUXk940dO5ZVq1bx888/ExysPk1SuhJTspnwwSYOnkrDt5I7r/ylG51b1TI7logUo3HjxrFw4cKr9lutVsaPH8+6deuIj49nxowZpKammpBQKorj59J5+r2NHDiZRmVvd6aO7Mxd3RuYHUtEHKAxRcqCvHwrMxZtY9mPhwDo36MRk4dHUMnL3eRkIs7hmgssS5cupUWLFkXbX3zxBRs3bmTTpk0kJyfTqVMnrUsvxcpmszNzURxHzqTjV9mDqaM64+/jaXYsESmjDMNgzJgxLF++nHXr1hEaqicGpHQdOX2JZ9/fxLnkbIICvJkx5iaah/5xDyARcT49e/bE1/fqpuGxsbG0bNmSOnXq4OvrS58+fVi7dq0JCaUi2LL7PBM+2MTFtFzqVK/MW09F0r6plu4WcTYaU8RsF9Nyee6fm/k14TxurhaeGtSe4fe01MoxIg645gLL8ePHadOmTdH2t99+y3333Ue3bt0IDAxk8uTJbNmypURCSsVjGAazlicQt/8CHu6uTBkRQe1qPmbHEpEybPTo0SxevJglS5bg6+tLYmIiiYmJ5Obmmh1NKoD4A0lM/HAzl7LyCa3tx8yxNxFS4+qLZREp+6KioggLC8PX15egoCD69evHwYMH//R9586do06dOkXbwcHBnD17tiSjSgVkGAaf/3SI1+bHkldgo13j6rw5LpI61XWtJFIWaUyRsuzgyVTGv7eBY2fT8ffx4NUnu3FbeF2zY4k4nWsusBQWFuLp+Z+nB7Zs2ULXrl2LtmvXrk1ycnLxppMK67OfDrF260ksFnjmkY40q6cZwCLyx2bNmkV6ejo9evSgVq1aRV/Lli0zO5qUc+viTvHy3K3kFdho27gar4/uTlV/b7Njich12rBhA6NHj2br1q388MMPWK1WevXqRXZ29h++zzCMq/ZZLJr9KcWnoNDG20viWfjtfgDu7hbK1FGd8ankYXIyEfk9GlOkrFq//TQTP/yFS5n51K/lx1tP3UyL0KpmxxJxStfczKJRo0Zs3LiRBg0acOrUKQ4dOsTNN99cdPzMmTNUrap/iHLj1sWdZvF3BwB4vF9rurTW2vUi8ud+6yJEpCRdnkV8mEXfXb7R1aNDMOMGtcfdTY2FRZzZmjVrrtiOjo4mKCiI7du3ExkZ+bvvq1OnzhWzi8+cOUNERMTvvj4/P5/8/Pyi7YyMjBtILeVdakYer0XHcvBUWlEz+zu7ajlUkbKuNMYUjSfiCLvdYPGa/Xz+02EAIlrW5OlHOuLtqX7HItfrmu8APPnkk4wZM4YRI0Zw55130qVLlyt6sqxbt4727duXSEipOHYeSuL9ZTsAGNCjEXerSaOIiJRBNrvBR8sTioor9/VsxN8f6qDiikg5lJ6eDkBg4B8/UR0eHs6ePXs4e/YsmZmZfPvtt/Tu3ft3Xx8VFYW/v3/RV0hISLHmlvLjyJlLPP3uBg6eSsPH252XH++i4oqIkyqJMUXjiVyr3Hwrr82PLSquDLylMS8MDVdxReQGXfO/oCeeeAI3NzdWr15NZGQkL7300hXHz507x/Dhw4s9oFQcx8+l89r8bdjsBpHt6vDYXS3+/E0iIiKlLL/QxpuL49i6JxGLBUbe24q+NzU0O5aIlADDMBg/fjzdu3enVatWAPTu3Zv4+Hiys7MJDg5mxYoVhIWF4ebmxltvvUXPnj2x2+1MmDDhD5/wnzhxIuPHjy/azsjI0E0xucovCed4e0k8BYU2goN81JtSxImV1Jii8USuRVJqDtPnxXDifAbubi6MfaAdPTvq74lIcbAYFXBNlYyMDPz9/UlPT8fPz8/sOAIkX8rlmfc3kpKeR8sGVZn+RBfc3VzNjiUiv0E/Q/9Dn0XFk5lTwPS5Mew/kYqbqwtPP9KB7m3r/PkbReQqzvAzdPTo0XzzzTds3ryZ4ODgEv1ezvB5SOkxDINPfzjEkrWXl07u0DSIZwd3wsfb3eRkImWTM/wMLa0xxRk+Cyld+4+n8tr8WC5l5VPF15NJw8LV61jkd1zPz1A9Ayamy84tZOrsLaSk5xFSw4fJw8JVXBERkTInKTWHl2Zv4UxSFpW93Zk0LJzWDauZHUtESsjYsWNZtWoVGzduLPHiish/yy+08d6nO9i083L/hb43NWD4PS1xddUylCLOSmOKmOWnbaf4x+e7sNrsNKjtz6Th4QQFVDI7lki5ogKLmKrQaidqQSwnEzMJ8PVk6sgu+FTyMDuWiIjIFY6dTWfanC2kZuRTzd+LqY93oV5NzQgUKY8Mw2Ds2LGsWLGC9evXExqqXhdSelLSc3klOpYjpy/h6mLhyfva0LtzfbNjich10pgiZrHZDRZ+s4/l648A0KV1LcY/1AEv9VsRKXb6VyWmMQyDDz7bwa7DyXh7uvLSyM4EBaqKLiIiZcuuQxd5dX4suflW6tX0ZeqoLlSr4m12LBEpIaNHj2bJkiWsXLkSX19fEhMTAfD398fbW//2peQcPp3GK/NiSc3Iw7eSBxOHhulJSREnpzFFzJCTV8ibn2xn274LAAy6rQkP926Gi4vF5GQi5ZMKLGKaxWsO8PP2M7i4WHh+SDgNg6uYHUlEROQK6+PP8N6n8VhtBq0aVmXSsAitfy9Szs2aNQuAHj16XLE/OjqaoUOHln4gqRA27TjLu5/GU2C1E1LDlxdHRFCzamWzY4nIDdKYIqUtMSWbV+bFcDIxE3c3F54a1J6bO2hZOpGSdN0FliNHjnD06FEiIyPx9vbGMAwsFlVC5dqs2XKCz348BMCYgW3p0CzI5EQiIiL/YRgGK9YfJXr1XgC6t63N+Ic7qEeYSAVgGIbZEaQCsdsNln5/kE9/OAhAp+Y1ePbRjlTyUjFfpDzQmCKlae+xFF6bH0tGdgEBvp5MHh5Bk7oBZscSKfccLrCkpKQwaNAg1q1bh8Vi4fDhwzRo0ICRI0dSpUoV3nrrrZLIKeXItn2JzPpyFwAP3t6U2yPqmZxIRETkP+x2g7mr9rBq0zEA+kY2YMQ9rfRIvYiIFKu8AivvLt3BLwnnAOh3c0OG3t0SV403IiLioB9iTvLhl7uw2gwaBvszeViEljUWKSUujr7h73//O25ubpw6dYpKlf7TL2PQoEGsWbOmWMNJ+XP4dBozFsVhN+DWsBAe7t3U7EgiIiJFCgptzFwcV1RcGX5PS0bd21rFFRERKVbJl3J5/p+b+SXhHG6uFp4a1I4RfVupuCIiIg6x/f/ksPc/24nVZtCtbW1eH91dxRWRUuTwEyzff/89a9euJTj4yvX7GjduzMmTJ4stmJQ/iSnZvDwnhvwCG+2aVGfM/e20rJyIiJQZWbmFvBodw56jKbi5Wvjbgx20XrGIiBS7gydTeTU6lrTMfPwqe/DC0HBaNqhqdiwREXEy2bmFvLE4ju0HkgB4qFdTHry9qSaHiZQyhwss2dnZVzy58m/Jycl4enoWSygpfzKyC5g6eyuXsvIJre3HxMfCcHN1+AEqERGREpF8KZeXZm/hVGIm3p5uTBoWTtvG1c2OJSIi5cz6+DO8v2wHhVY79Wr6MmVEZ2oEXn19LSIi8kfOJ2czfd5WTl/IwsPNhb891IGb2tUxO5ZIheTwHe7IyEgWLlxYtG2xWLDb7bzxxhv07NmzWMNJ+ZBfaOOVeTGcvZhFtSrevDSys5o2iohImXHyfAbPvr+RU4mZBPp5MmNMdxVXRESkWNntBou+289bn2yn0GonvEVNZo69ScUVERFx2O4jyTz93gZOX8gi0M+L18d0V3FFxEQOP8Hyxhtv0KNHD+Li4igoKGDChAns3buX1NRUfvnll5LIKE7Mbjd4Z0k8+0+kUtnLjamjOlPVX+tAiohI2bD7aDKvzoshO89KSA0fpo7sQpBudomISDHKzbfyztJ4tuw+D8B9PRsxuE8L9VsRERGHrdlygo+WJ2CzGzQOqcKkYeG6zyZiMocLLC1atCAhIYFZs2bh6upKdnY2AwYMYPTo0dSqVaskMooTm/f13v9v3OjCpGER1KvpZ3YkERERADbvOstbn8RjtdlpXj+QKSMi8K3kYXYsEREpR5LScnhlXgzHz2Xg5urC2AfackunumbHEhERJ2Oz2Zmzag+rNx8HILJdHcY92B5Pd1eTk4mIwwUWgJo1azJt2rTiziLlzKqNR1m58SgAf3uwPa0bVTM5kYiIyGWrNh1lzso9GAZ0aV2Lpx/pqIsTEREpVgdOpPLq/FguZeZTxceTF4aG0zw00OxYIiLiZLJyC5m5cBs7Dl0E4NE7mvHAbU2wWPQkpEhZ4HCBJTQ0lEcffZRHH32Upk2blkQmKQd+STjHnFV7AHjsrhbc3CHY5EQiIiKXl65c8M0+lq8/AkCfrvV5vH8bLdMiIiLFal3caT74bCdWm53Q2n5MHhahJShFRMRh5y5m8fLcy32NPT1c+ftDHejWprbZsUTkvzjc5H7s2LGsWbOG5s2b07FjR959913Onz9fEtnESe07nsJbn2zHMODOrvW5r2cjsyOJiIhQaLXz9pL4ouLKkD7N+csAFVdERKT42OwG81fv5Z2ll5eg7NyqJjPG3KTiioiIOGzXoYs8/d5Gzl7Mopq/FzNGd1dxRaQMcrjAMn78eLZt28aBAwe4++67mTVrFnXr1qVXr14sXLiwJDKKEzmTlMkr82IotNoJb1GTJ/q11iOLIiJiupy8QqbN2cKGHWdwdbHw94fac/+teqxeRESKT05eIa9Fx/Llz5cL+Q/c1oSJj4Xj7XldK3OLiEgF9s0vx3lx9haycgtpWjeAt/92Mw2Dq5gdS0R+g8MFln9r0qQJ06ZN4+DBg2zatImLFy8ybNiw4swmTiYtM4+ps7eSmVNIk7pVePbRjri6XvdfMRERkWKRkp7L8//czK7DyXh5uPLiiM5qMCwiIsXqQmoOz/1jM7H7EnF3c+HpRzoy+M7muOgpSRERcYDVZmfWl7v4aHkCdrtBjw7BvPbXbgT4eZkdTUR+xw1NpYmNjWXJkiUsW7aM9PR0Bg4cWFy5xMnk5Vt5eW4MF1JzqFm1ElOGd8ZLM7VERMRkpy9kMnX2FpLScqni48lLIzvTKKSK2bFERKQc2XsshagFsaRnFRDg68mkYeE0radm9iIi4pisnAJeX7iNXYeTsVhg8J3NGXhLYz11L1LGOXwH/NChQ3zyyScsWbKEEydO0LNnT15//XUGDBiAr69vSWSUMs5mszNzcRxHTl/Ct5IH00Z1oYqvp9mxRESkgtt3PIVX5sWQmVNI7WqVmfZ4F2pWrWx2LBERKUd+jD3FP7/YidVm0DDYn8nDIqhWxdvsWCIi4mTOJGUyfW4M55Kz8fJw5elHOtK5VS2zY4nINXC4wNKsWTM6derE6NGjefDBB6lZs2ZJ5BInYRgGH63YzbZ9F/Bwc+HFERHUru5jdiwREangtuw+z5uL4yiw2mlaN4ApIyLw91HxX0REise/m9l/teEoAN3a1OZvD7bXU/wiIuKw+INJzFy4jew8K9UDvJkyPILQ2v5mxxKRa+Twb38HDhygSZMmJZFFnNAX6w6zZssJLBZ45tGONKuvR+FFRMRc3/56nH8tT8BuQFiLGkwY3AkvD93wEhGR4pGTV8gbi7cTt/8CAA/e3pSHejVVvxUREXGIYRis3nycOav2YLcbNK8fyMShYQT4qt+KiDNx+G6Diivyb+u3n2bht/sBGHVva7q0rm1yIhERqcgMw2DRd/v5/KfDAPTuXI8nB7TB1dXF5GQiIlJeJKZkM31eDKcSM/Fwc+FvD3bgpvZ1zI4lIiJOxmqz868Vu1mz5QQAt3QKYcz9bXF3czU3mIg47JoKLIGBgRw6dIhq1aoREBDwh82VUlNTiy2clF27Dl/kvWU7AOh3c0PuuamByYlERKQis9rsfPDZTtbFnQbg4d7NePD2JmoIKSIixWb30WSi5m8jM6eAQD8vJg8Pp3FIgNmxRETEyWRkF/D6gm3sPnq5mf3Qu1rQv0cjXbuIOKlrKrC88847RQ3s33nnnTLxD/7NN98kOjoai8XC888/z6OPPmp2pArj5PkMXpsfi9Vm0L1tbYbd3dLsSCIiUoHl5lt5fcE24g8m4eJiYfTAtvSKqGd2LBERKUfWbj3JrC93YbMbNAqpwuRh4VT1VzN7ERFxzOkLl5vZn0/JxtvTlWce6UR4S/W3FnFm11Rgeeyxx4r+e+jQoSWV5Zrt3r2bJUuWsH37dgBuvfVW7r77bqpUqWJusAogJT2XqbO3kJNnpWWDqvz9oQ5aa1hEREyTlpnHy3O2cuRMOp4erjw3uBNhLXSBIiIixcNmszNv9V5WbTwGwE3t6jBuUDv19hIREYfF7b/AG4vjyMmzEhRYiReHR1Cvlp/ZsUTkBjm8KLmrqytJSUlX7U9JScHVtXTWCdy/fz9du3bFy8sLLy8v2rVrx5o1a0rle1dkOXmFTJ29leT0PIKDfJg0LBwPd60NKSIi5jh3MYtn39/EkTPp+FX24LUnu6m4IiIixSY7t5CX58UUFVceuaMZzz7aUcUVERFxiGEYfLXhKNPnbi2asPz2U5EqroiUEw4XWAzD+M39+fn5eHh43HCgf4uKiiIsLAxfX1+CgoLo168fBw8eBKBVq1b8/PPPXLp0iUuXLrFu3TrOnj1bbN9brlZotRM1fxsnzmcQ4OvJ1FFd8K1UfP+/RUREHHHoVBrPfrCJC6k51KxaiTfG3kSTuloHX0REise55CyeeX8j8QeS8HB35fkhYTx4e9MysVy2iIg4j0Lr5V6Rc1ftwW7A7eF1mf5EV/x9PM2OJiLF5Jqn3rz//vsAWCwW5syZg4+PT9Exm83Gxo0badasWbEF27BhA6NHjyYsLAyr1cqkSZPo1asX+/bto0WLFowbN45bbrkFf39/wsLCcHPTLKKSYhgG//h8JzsPX8TLw5UXR3amRmAls2OJiEgFFbsvkZmL4sgvsNEo2J8XR3YmwNfL7FgiIlJOJBy5yOsLtpGZU0hVfy8mD4+gUXAVs2OJiIiTSc/KJ2rBNvYeS8HFAsPuacW9kQ1UrBcpZ665KvHOO+8Al2+2f/TRR1csB+bh4UH9+vX56KOPii3Y/y75FR0dTVBQENu3bycyMpInnniCJ554AoCRI0fSqFGj3z1Xfn4++fn5RdsZGRnFlrMi+GTtAdbFncbFxcJzQ8J0cSEiIqZZu/UkH36xE7sBHZoF8fyQMLw9NclCRESKx3dbTvCv5QnY7AZN6lZh0rAIAv1UxBcREcecPJ/By/NiSErNoZKXG88+2olOzWuYHUtESsA135E4fvw4AD179mT58uUEBJTuMhzp6ekABAYGApCUlERQUBAHDx4kNjb2D4s7UVFRTJs2rVRyljdrt55k2Q+HAPjrfW01GIiIiCkMw+DT7w+y5PvLy4XeGhbCmPvb4ebq8GqnIiIiV7HZ7MxZtYfVmy9f9/boEMyYB9rhqZ6TIiLioNh9iby5OI7cfBs1q1ZiyvAI6tZUvxWR8srhKZ8///xzSeT4Q4ZhMH78eLp3706rVq0A6NevH5cuXaJy5cpER0f/4RJhEydOZPz48UXbGRkZhISElHhuZxe3/wIffrkLgEG3NaF353omJxIRkYrIZrPz4ZcJfB9zEoAHbmvCo3c006P1IiJSLLJyCpixMI6dhy8CMKRPcwbe0ljjjIiIOMQwDFasP8r8b/ZiGNC6YTWefywMv8rqYSxSnjlcYBk4cCCdOnXi+eefv2L/G2+8QWxsLJ9//nmxhfu3MWPGkJCQwObNm4v2/frrr9f8fk9PTzw91TzKEUfOXGLGwm3Y7Qa3dArhkTuKr7+OiIjItcrLtzJzcRzb9l3AxQJ/GdCGO7uGmh1LRETKibMXs5g+dytnL2bj5eHK+Ic70qV1LbNjiYiIkym02vjnF7v4adtpAHp3rscT/dvg7qYn7kXKO4cLLBs2bOCll166av8dd9zBm2++WSyh/tvYsWNZtWoVGzduJDg4uNjPL1e7kJrDtDlbySuw0a5xdcbc306zt0REpNSlZ+UzfW4MB0+l4eHmwrODO9G5lW56iYhI8dh5KInXF8aRnVtItSrevDgigtDa/mbHEhERJ3MpM5/X5sey/0QqLhYYcW8r7umuZvYiFYXDBZasrCw8PK5+tM3d3b1Ym8cbhsHYsWNZsWIF69evJzRUs1VLQ2ZOAVNnb+FSZj71a/kxcWiYqu0iIlLqElOyeenjLZxLzsa3kjtThnemeWig2bFERKSc+GbzMT5euQe73aBZvQBeGBZOgK+a2YuIiGOOn0vnlXkxJKXlUtnLjQlDwujQNMjsWCJSihwusLRq1Yply5bx4osvXrH/008/pUWLFsUWbPTo0SxZsoSVK1fi6+tLYmIiAP7+/nh7exfb95H/KCi08Wp0LGeSsqjm78XUUZ2p5OVudiwREalgjpy+xLQ5W7mUlU9QgDdTR3UhpIav2bFERKQcsNrsfPzVbr779QQAt3QKYcz9bXF3UzN7ERFxzNY953nrk+3kFdioVa0yU4ZH6LpFpAJyuMAyZcoU7rvvPo4ePcott9wCwE8//cTSpUuLtf/KrFmzAOjRo8cV+6Ojoxk6dGixfR+5zG43eGdpPHuPpVDZy42po7pQ1V+FLBERKV3xB5KIWhBLXoGN0Np+vDSys8YjEREpFpk5Bby+YBsJR5KxWOCxPi0Y0LORlnARERGHGIbBF+sOs+i7/RgGtGl0uZm9byU1sxepiBwusPTt25evvvqK1157jS+++AJvb2/atGnDjz/+yM0331xswQzDKLZzyZ+LXr2XzbvO4eZq4YVh4dSr5Wd2JBERqWB+2naKDz7bic1u0LZxNV4YGq4nKUVEpFicvpDJ9HkxnE/OxtvTlacf7kiE+nqJiIiDCgpt/OPznfy8/QwAd3atz+P9WuPmquX1RSoqhwssAHfddRd33XVXcWcRk3y96RhfbTgKwFOD2tOmUXWTE4mISEViGAaf/3R5BhhAj/9j7z7DoyrXtw//0hPSIIQkhITee0sCFkREFFEBsaHUYN2IBTuCukVEsW23bt0WCE3ECqIoitIVkkDovffQQnqdmef9wCv/jYoyIcmaSa7zOPIhk8nK/QzJupi5Zz13xxgevK2DZoCJiEiZSNt2nEkzUskrtBERVo1xiQnU1xvKRETESaezC5kwNYXt+0/j6enBPf3a0OdSzYwWqepK9cpFZmYmH330EWPGjCEjIwOAtLQ0Dh8+XKbFSflbufEIH369EYAh17Wge6dYiysSEalYr732Gq1ataJ169bMnDnT6nKqHLvD8N+vNpxtrgy4sjGPDOyo5oqIiFw0Ywzzlu3mnx+tJK/QRssGYbzxUDc1V0RExGl7Dmcx+q1lbN9/msAAH/55dxc1V0QEKMUVLBs2bKBnz56Ehoayb98+7rrrLsLCwpgzZw779+9n+vTp5VGnlIOtezN4beYajIFru9bn5h5NrC5JRKRCbdy4kVmzZrFmzRoArrrqKq6//nqqV69ubWFVRFGJnddmrmbVpnQ8POCuvq258fJGVpclIiKVQInNwftzNvDDqv0A9Iyryz9ubqth9iIi4rRfNxzhjU/SKCq2U6dWEONGJFCnVpDVZYmIi3D67aGjR49m2LBh7Ny5E39//7O39+7dm2XLlpVpcVJ+Dp/IZfyUZIptDuJaRnJf/zYa7igiVc7WrVu55JJL8Pf3x9/fn/bt27NgwQKry6oScvKLGfffX1m1KR0fb0+eHByn5oqIiJSJ7Lxinv3gV35YtR8PD0i8oRUP3tZezRUREXGKMYZPF25n4rRUiortdGhai9ce6qbmioicw+kGS2pqKvfee+8fbq9Tpw7p6ellUpSUr8ycIp7/cCU5+cU0ia3OE4M646VhXCJSSU2cOJG4uDiCg4OJiIigX79+bN++HYDWrVuzePFiMjMzyczMZNGiRdrusgIcz8jnibeXs3VfBoEBPrxwT1cubRdtdVkiIhdFW066hgPp2Tz61lI27T5FgJ83z47oQv/ujfVmMhFxK8oU6xWV2Hnt4zXMXLANgBsub8hzd3UhKMDH4spExNU4vUWYv78/2dnZf7h9+/bt1Kql4eiurrDIxvgpq0g/lU9kWDXGjUjA38/pXwMREbexdOlSRo4cSVxcHDabjWeeeYZevXqxZcsWWrZsyYMPPkiPHj0IDQ0lLi4Ob2+dE8vT3iNZPP/hSjKyiwgP9ef5e7pSL0p74YuIe9OWk65h9dZjTJqxmoIiG1E1qzE2MUEZIyJuR5livVNZBUxISmHnwUy8PD2476a2XNu1vtVliYiLcvqyhb59+/LCCy9QUlICgIeHBwcOHOCpp55iwIABZV6glB273cGrM9ew40AmwdV8eP7uLtQI9v/7bxQRcWMLFixg2LBhtGrVinbt2pGUlMSBAwfOPmG59957SUtLY/Hixfj6+tK4cePzHquoqIjs7OxzPuTCrd9xgiffWUFGdhH1ooJ59cFueuFLRCoFbTlpLWMMc5fuYvzkVRQU2WjdqCavKWNExE0pU6y162Amj761jJ0Hz7x29sK9XdVcEZG/5HSD5bXXXuPEiRNERERQUFDAFVdcQePGjQkODmbChAnlUaOUAWMM78/dSMqWM3vdj01MICYi2OqyREQqXFZWFgBhYWEAHD9+HDhzJWZKSgrXXHPNeb934sSJhIaGnv2IjY0t/4IriSVph3j+o5VnX/h6+YHLCa8eYHVZIiIXTFtOuqYSm4O3P1vH5HmbcRi4pks9XrjnEkKD/KwuTUTkvJQprmnF+sM8+Z8VnMoqJDYyiNce6kbbxtqtR0T+mtP7oISEhLBixQoWLVpEWloaDoeDjh070rNnz/KoT8rIl4t38f2v+/DwgEfv7ETLBjWtLklEpMIZYxg9ejSXXXYZrVu3BqBfv35kZmYSGBhIUlLSX24R9vTTTzN69Oizn2dnZ6vJ8jeMMcxZspukbzcDcFm7aEbf0VGDhkXE7WjLSdeTlVvExGmpbN5zCk8PGNG3NTdc1lDzVkTE5SlTXIsxhtk/bmfWj2eaXJ2aR/D4oM4Eat6KiFwAD2OM+bs7hYWFsWPHDsLDw0lMTOStt94iONh9r37Izs4mNDSUrKwsQkIq/2XjS9IO8frHZ7bCubtva27s1sjiikTEnbnzOXTkyJHMnz+fFStWEBMTc9HHc+fHoiI4HIbJ8zYxb/keAG7s1pARN7TG01MvfImI+59Df7uqf+nSpXTr1u2cr911113079+fPn36XPDx3P3xqGj7jmYzfvIqjp8uINDfmycGx9GxeYTVZYmIRdz9HFqWmeLuj0VFKyy28dbstaxYfwSAvt0aMfyGVnjpOYtIlVSac+gFbRFWXFx8dp/5adOmUVhYWPoqpUJt3HWSt2anAWdCQs0VEamqRo0axbx581i8eHGZNFfkrxWX2Jk0c/XZ5kriDa24u28bNVdEpNK4mC0nQXO9LkbK5nSeeHsZx08XUDs8kFcf7Kbmioi4tYvJFOVJ6Z3KKuDp/6xgxfojeHt5MOrW9tzVt7WaKyLilAu6xrBr167069ePTp06YYzhwQcfJCDgz/dNnzJlSpkWKKW3Pz2bCUnJ2OyGS9tGk3hDK6tLEhGpcMYYRo0axZw5c1iyZAkNGjSwuqRKL7eghAlJyWzafQpvLw8evr0jV3RUU0tEKo+L3XISzuy//89//rMiyq00jDF8tXgX077bgjHQtnE4Tw2NI7iar9WliYiU2sVmivKkdHYcOM2EpGQysosIrubLmGFxtG4UbnVZIuKGLqjBMnPmTN588012794NnOms6yoW13Yqq4DnP1xFXqGNFvXDGH1HR71rWESqpJEjRzJr1iy+/vprgoODSU9PByA0NPS8bxaQ0juZWcBzH67kQHoO1fy9GTMsnnZNNBhSRCqXBx54gA0bNrBixYqzt/36669OHUNzvZxTYrPzzufrWbT6IAC9u9bnnv5t8Pa6oE0ZRERc1sVmivLEecvWHuKt2WsptjmoGxXMuMQEomoGWl2WiLipC2qwREZG8vLLLwPQoEEDZsyYQc2aGpLuqvILS/jnR6s4mVlAnVpBjE1MwNdHw4RFpGp67733AOjevfs5tyclJTFs2LCKL6gS2380m+c/XMnJrELCQvx4/u6uNIgOtbosEZEy9duWk8uWLbuoLSf9/Pzw8/Mrw8oqr9M5hbyUlMK2/afx9PTgnr6t6XNZQ6vLEhG5aGWRKcqTC+dwGGb9sI1Pf9oBQOcWkTw+qBPV/DXMXkRK74IaLP875P7KK6/E11eXYLsqm93BxGmp7D2STfVgP56/uwshgfr3EpGqyxhjdQlVwsbdJ5kwJZm8QhuxkUE8f1dXIsKqWV2WiEiZ0ZaT1th7JIvxU5I5cbqAwAAfnhzcmQ7NNG9FRNybMqXiFRbZeHN2Gr9uOApA/+6NGdqnpeatiMhFu6AGy29D7sPDw5k2bRqvvPIKwcHB5V2bOMkYwzufr2PdjhP4+Xrx7Ahd4igiIuVvxfrDvP5xGja7gxb1wxg3IkH74YtIpaMtJyveyo1HeWPWGgqL7dSpFci4EV2oUyvI6rJERC6aMqVinThdwItJyew5nIW3lwcjb25Pz/i6VpclIpWEhtxXIrN+2M7PqQfx9IAnB3emSWwNq0sSEZFKbt7y3Xz09SaMga5tavPonZ3w07aUIlIJacvJimOM4YtFO5n+3VYA2jepxZNDOhOk5r2IVBLKlIqzbX8GE5JSyMwpIjTIlzHD4mnZQGMPRKTsOD3k3sPDQ0PuXdCPyfuZvXA7APcPaEdcyyiLKxIRkcrM4TBMm7+Fr5bsAuC6S+pzT/+2usReRCotbTlZMYpL7Lz92TqWpB0C4PpLG3BX39Z4aZi9iFQiypSKsXjNQd7+bB0lNgf1a4cwLjFB2xiLSJnTkPtKYM22Y/zni/UA3NqzKdd2rW9tQSIiUqmV2By8NXstS9eeefFryHUtuLlHEzw81FwREZHSO51dyISkFLYfODPM/r7+beh9ieYSiIiIcxwOw4zvt/LFop0AJLSK4tE7OxHgd0Evg4qIOMXpM8vevXvLow4ppd2HMnlleioOh+HKTjEMura51SWJiEglll9YwktTU1i/8yRenh48eFt7enTW/sUiInJxdh/K5MUpyZzMKiQowIenhsbRrkktq8sSERE3U1Bk4/WP15C8+cxcm5t7NGFw7xZ46kp7ESknF3yd9XXXXUdWVtbZzydMmEBmZubZz0+dOkXLli3LtDj5a8cz8vnnR6soKLLTrkk4o27toHcPi4hIuTmVVcBT/1nB+p0n8ff14tkRXdRcERGRi/bLhiM8+Z8VnMwqJCYiiNcf7qbmioiIOO14Rj5PvL2c5M3p+Hh7MvqOjgzt01LNFREpVxd8BcsPP/xAUVHR2c9feeUVBg4cSPXq1QGw2Wxs3769zAuUP5ebX8zzH63kdE4R9WuH8PTQeHy8tS+xiIiUj4PHcnj+w5UcP11A9WA/nhvRhcax1a0uS0RE3Jgxhk9/2sHHC7YB0LF5BE8M6kxggI/FlYmIiLvZsvcUL01NISu3mOrBfjwzPJ7m9cKsLktEqoALbrD8fgCXBnJZp8Rm58WkFA4ey6VmqD/P3dVFT0JERKTcbNl7ihenJJOTX0J0eCD/vKcrUTUDrS5LRETcWFGJnX/PXsuydYcBuLFbQxKvb6Vh9iIi4rSfUw/wzufrsdkdNIwO5ZnEeCJqaJi9iFQMTXdyMw6H4c1P1rJ5zymq+Xvz/N1dCa8eYHVZIiJSSa3ceJTXZq6m2OagWd0ajBuRQGiQn9VliYiIGzuVVcCLSSnsOpiJl6cH9w9oxzVd6lldloiIuBm7wzB9/ha+WrILgK5tajN6YEf8NcxeRCrQBZ9xPDw8/jDfQ/M+Kt60+VtYvu4w3l4ejBkaT/3aIVaXJCIildR3v+7l/a824DAQ1zKSJwZ3xt9XT1ZERKT0dh48zYtTUsjILiS4mi9PD4ujTaNwq8sSERE3k19YwmsfryF1yzEAbuvZlDuuaa55KyJS4ZzaImzYsGH4+Z1512phYSH33XcfgYFntgj53/ksUj6+XbHnbFd+1K0daNdUgx9FRKTsGWOY8f1WPv95JwDXdKnH/Te11bYtIiJyUZavPcy/ZqdRbHMQGxnMsyMStOWkiIg4Lf1UHi9OSWZ/eg4+3p48dFsHrugYY3VZIlJFXXCDZejQoed8PmjQoD/cZ8iQIRdfkfyplRuP8sHcjQAM6t2cHp1jLa5IREQqI5vdwdufrWPR6oMA3Hltc27r2VRXrYqISKk5HIZPftzO7IXbAejcIpLHB3Wimr/mSIqIiHM27zkzzD47r5gawX6MTUygad0aVpclIlXYBTdYkpKSyrMO+Qvb9mfw2szVGHPmXcS3XtXU6pJERKQSyi8s4ZXpq0nbfhxPTw9G3tyOXgnaE19EREqvsNjGvz5Zyy8bjgDQ74pGDLu+FV7awkVERJy0MHk/7365Hpvd0DgmlGeGJ2gusYhYThupu7gjJ3MZPzmZYpuDzi0iuf+mtnoXsYiIlLnTOYW88NEqdh3Kws/XiycHdyauZZTVZYmIiBs7mVnAi0nJ7D6UhbfXmcZ9z3g17kVExDl2h2Hqt5uZu3Q3AJe2i+bh2ztoPqSIuASdiVxYVm4Rz3+wiuy8YhrHhPLE4M7a/15ERMrckRO5PPvBSo5l5BMS6Mtzd3XRZfYiInJRtu/PYEJSCqdziggJ9GXMsHhaNaxpdVkiIuJm8gpKeHXmatZsOw7AHb2acXuvZnrzsYi4DDVYXFRhsY0XJq/i6Kk8IsKq8eyILgT46Z9LRETK1o4Dp/nnR2ea+VE1q/HPu7sSXSvI6rJERMSNLUk7xL8/XUuJzUG9qGDGjehCZFg1q8sSERE3c/RkHuOnrOLgsVx8fbx4ZGAHLmtXx+qyRETOoVfsXZDdYXht5hp2HMgkuJoP/7y7CzVC/K0uS0REKpmULelMmrGaomI7jWOr8+yIBGoEK29ERKR0HA7Dxz9s47OfdgAQ3zKKR+/sqGH2IiLitI27TjJxWgo5+SWEhfgzLjGBxrHVrS5LROQP1GBxMcYYPpy7keTN6fh4ezI2MYGYiGCryxIRkUrmh1X7efeLdTgMdGwewVND4nSlpIiIlFpBkY03P0lj5cajANzcowmDerfQMHsREXHagpX7+O9XG7A7DE1iq/PM8HhqhmqYvYi4Jr2S4mLmLNnF/F/24uEBj97RiZYNtE+xiIiUHWMMs3/czqwftwNwVVwsD9zSHm/N+BIRkVI6fjqfF6cks/dINt5enoy6tT09OsdaXZaIiLgZu93BR/M28e2KvQB0a1+HB2/vgJ+Pl8WViYicnxosLmTZ2kMkfbsFgMQbWnNpu2iLKxIRkcrEbnfw7pcb+DF5PwC39WzKndc214BIEREptW37MpgwNYXMnCKqB/nxzPB4mtcPs7osERFxM7kFJUyansraHScAGHRtc27t2VTPVUTE5anB4iI27j7Jm5+sBeDGyxvS74pGFlckIiKVSWGRjUkzV5O65RieHnDfTW3pfUkDq8sSERE3tmj1Qd7+bB02u4MG0SGMTUwgooaG2YuIiHOOnMjlhcnJHD6Ri5+vF48M7MilbfWmYxFxD2qwuIAD6dlMSErBZndwSdvaJN7Y2uqSRESkEsnKLWL85GS2HziNr7cnjw/uTJfWta0uS0RE3JTdYZjx3Ra+XLwLgK5tavPIwI6a5SUiIk5bv+MEL09PJbeghPBQf8YmJtAoprrVZYmIXDD9D9hip7IKeP6jVeQVlNCifhij7+ikQZAiIlJm0k/l8dwHKzlyMo/gaj6MS+xCiwbaukVEREonv7CE1z9OI2VLOgC39mzKndc0x1PPYURExEnzf9nLB3M34nAYmtWtwTPD46kR4m91WSIiTlGDxUL5hSW88FEyJ04XUKdWIGMTEzS4S0REysyug5n886NVZOYWEVEjgOfv7kpsZLDVZYmIiJs6lnFmmP2+o9n4eHvy4G0d6N4xxuqyRETEzdjsDj6cu5Hvft0HQPdOMYy6pT2+ek1MRNyQGiwWsdkdvDwtlT1Hsqge5Mfzd3clJNDX6rJERKSSSNt2nInTUigsttMwOpTn7u5CmN4NJiIipbR5zykmTkshK7eYGsFnhtk3q6crIkVExDm5+cW8PD2V9TtP4uEBg3u34OYeTTTMXkTclhosFjDG8J/P17N2xwn8fL0YNyKBqJqBVpclIiKVxM+pB3j7s3XYHYb2TWrx9LA4qvn7WF2WiIi4qZ9SDvCfL9ZhsxsaxYQydngC4dUDrC5LRETczKHjOYyfnMyRk3n4+3rx6J2dNBtSRNyeGiwWmP3jdn5KPYCnBzwxuDNN69awuiQREakEjDF8/vNOZny/FYDuHWN48LYO+Hh7WlyZiIi4I7vDMPXbzcxduhuAS9tG8/DtHfDXMHsREXFS2vbjTJqeSl6hjVo1AhiXmECD6FCryxIRuWj6n3EF+yllP7N+3A7AfQPaEd8yyuKKRESkMrA7DB/M2XB2H+MBVzZmyHUtNXRYRERKJb+whFdnrmH11mMA3H51Mwb2aqZcERERpxhj+HbFXj6atwmHw9CifhhjhsVTPdjP6tJERMqEGiwVKG3bcd75fD0At1zVhN5d61tbkIiIVApFJXZem7maVZvS8fCAu/q25sbLG1ldloiIuKn0U3mMn5LMgfQcfL09efj2jlzeoY7VZYmIiJux2R3896sN/LBqPwA9OsfywC3t8PHWMHsRqTzUYKkguw9l8vL0FOwOQ/eOMQzu3cLqkkREpBLIyS9m/ORktu7LwMfbk0fv6MSl7aKtLktERNzUxt0nmTg1lZz8YsJC/BmbGE+TWG1pLCIizsnOK+blaals3H1mmP2wPq3o372RhtmLSKWjBksFOH46nxcmr6KgyE7bxuE8eFsHBYqIiFy04xn5PPfhSg4dzyUwwIexw+Np3Sjc6rJERMRN/bBqP+99uR67w9A4tjpjh8dTM1TD7EVExDkH0rMZPyWZ9FP5BPh58digztoiX0QqLTVYyllufjHPf7iKjOwi6kUF8/SweA0bFhGRi7bncBb//GglGdlFhIf68/w9XakXFWJ1WSIi4obsdgdTvtnMvOV7ALi8fR0evK09/r56uigiIs5ZvfUYr85cTX6hjYiwajybmEC92nqeIiKVl/7HXI5KbHYmTE3h4LEcaob689xdXQkK8LG6LBERcXPrd5xgwtQUCops1IsK5vm7uxJeXe8wFhER5+UVlDBpxmrSth8H4M5rm3Nbz6a64l5ERJxijOHrZXtI+mYTDgOtGtbk6aFxhAZpmL2IVG5qsJQTh8Pwr0/Wsmn3KQL8vHnuri7UqqEXv0RE5OIsSTvEW7PTsNkNrRvV5JnhCWrei4hIqRw5mcv4yckcOp6Ln68XjwzsyKVtNcdLREScU2Jz8N6X61mYcgCAq+Prcv+AdtrBRUSqBDVYysn077awbN1hvDw9eHpoHA2iQ60uSURE3JgxhjlLdpP07WYALmsXzeg7OuLj7WVxZSIi4o7W7zzBy9NSyS0oITzUn7GJCTSKqW51WSIi4maycouYOC2VzXtO4ekBiTe25sbLG+pKSBGpMtRgKQfzV+zhy8W7AHjwtvZ0aBZhcUUiIuLOHA7D5Hmbzu6Nf2O3hoy4oTWennrSIiIizvv+1728P2cjdoehWd0ajBkeT1iIv9VliYiIm9l/NJsXpiRzPCOfav7ePD6oM51bRFpdlohIhVKDpYwlbzrKB3M3AjDo2ub06FzX4opERMSdFZfYeeOTNH5ZfwSAxBta0b97Y4urEhERd2S3O/jo6018+8teALp3jGHUre3x9dHVkCIi4pyULem8NnM1BUV2atcMZNyIBGIjg60uS0SkwqnBUoa2789g0sw1OAz0SqjHrT2bWl2SiIi4sdyCEiYkJbNp9ym8vTx4+PaOXNExxuqyRETEDeXmF/PK9NWs23kCgCHXteDmHk20hYuIiDjlt62Lp87fjDHQplE4Tw2NIyTQ1+rSREQsoQZLGTlyMpcXJidTXGKnU/MI/jGgrZ6siIhIqZ3MLOC5D1dyID2Hav7ejBkWT7smtawuS0RE3NDhE7mMn7yKwyfy8Pf1YvQdnejaprbVZYmIiJspsdn5zxfr+Tn1IADXdq3Pvf3b4O2lYfYiUnWpwVIGsnKLeP7DVWTnFdMoJpQnh8ThpXAREZFS2n80m+c/XMnJrELCQvx4/u6uNIgOtbosERFxQ+t2HOfl6avJKyghvHoAz45IUKaIiIjTMnOKeGlqClv3ZeDpAXf1bcP1lzXQm4tFpMpz2y7Am2++SatWrWjZsiUPPvggxhhL6igstjF+SjJHT+YREVaN50Z0IcBPfSsREXfhKnnym427T/LkO8s5mVVIbGQQr47qphfCRETcgKvlCcD8FXt47sNV5BWU0LxeDd54WJkiIuIOXC1T9h7JYvRbS9m6L4NAf2+eu7srN1zeUM0VERHc9AqWEydO8M4777B582Z8fHzo1q0bq1atomvXrhVah91heP3jNWzff5qgAB+ev6sLNUL8K7QGEREpPVfJk9+sWH+Y1z9Ow2Z30KJ+GONGJBBcTXsZi4i4OlfLE5vdwQdzN/L9r/sA6NE5lgduaYePt4bZi4i4OlfLlFWbjvL6x2soLLYTHX5mmH1MhIbZi4j8xi0bLAA2m43CwkIASkpKiIiIqNCfb4zho7kbWbUpHR9vT8YmJhAbqYAREXE3VufJb+Yt381HX2/CGOjapjaP3tkJPx+9ECYi4i5cJU+y84p5ZXoqG3adxMMDhl7XkpuubKx3GYuIuBFXyBRjDF8s2smM77diDLRrEs6TQ+L0BjARkd9x2S3CJk6cSFxcHMHBwURERNCvXz+2b98OQK1atXjssceoW7cu0dHR9OzZk0aNGlVofXOW7ObbX/YC8MjAjrRqWLNCf76IiFwYV88Th8OQ9M1mPpx7prnS59IGPDkkTs0VEREX4+p5AnDwWA6PvbWMDbtOEuDnxTPD4hnQo4maKyIiLsbVM6W4xM4bn6Qx/bszzZXrLqnP83d3VXNFRORPuGyDZenSpYwcOZJVq1axcOFCbDYbvXr1Ii8vj9OnT/Ptt9+yb98+Dh8+zK+//sqyZcsqrLblaw+T9O1mABJvaMXl7etU2M8WERHnuHKelNgcvDErja+W7AJgyHUtuLd/G7w89UKYiIirceU8AViz7RiP/XsZR0+dmQ05aVQ3ElrXrtAaRETkwrhyppzOLmTMe7+wZM0hPD09uO+mttw/oB3eXi77EqKIiKVcdouwBQsWnPN5UlISERERrFmzhmPHjtG4cWPCwsIA6NOnD6tWraJbt27lXtem3Sd545M0AG64vCH9rqj4d6aJiMiFc9U8yS8s4aWpKazfeRIvTw8evK09PTrXLfefKyIipeOqeWKM4Zvle5g8bxMOAy0bhDFmWDyhQX7l/rNFRKR0XDVTdh/K5MWkFE5mFhAU4MNTQ+Jo17RWuf9cERF35rINlt/LysoCICwsDF9fX3799VcKCwvx8fFhyZIl3HPPPef93qKiIoqKis5+np2dXaoaCopsTJyWis3uoGub2oy4sbUutxcRcTOukCcAH8zdyPqdZ7ZweWpoPB2bWbNXv4iIlM7F5AmUXaas3XGCD7/eBMDV8XW5f0A7fLz1LmMREXfiCs9RikvsvDB5FRnZRdSpFcSzIxKIrhVUqmOJiFQlbtFgMcYwevRoLrvsMlq3bg3AddddR4cOHfD09OSqq67ixhtvPO/3T5w4kX/+858XXUeAnzePDOzI18t28+idnbSFi4iIm3GVPAEY1qcVR07kcU//NjSOqV4mxxQRkYpxsXkCZZcpHZrWoldCPWIjg+jbrZHeACYi4mZc5TmKr48XD97WgW9X7OXROzsRFOBz0ccUEakKPIwxxuoi/s7IkSOZP38+K1asICYmxunv/7NufmxsLFlZWYSEhDh9PGOMnriISJWVnZ1NaGhoqc+hVlKeiIi4jqqcJ1C2maI8EZGqripnip6jiIiUndLkictfwTJq1CjmzZvHsmXLSv3kxc/PDz+/stuDWEEjIuJ+lCciIlIWyiJPoGwzRXkiIuKe9BxFRMT9uWyDxRjDqFGjmDNnDkuWLKFBgwZWlyQiIm5IeSIiImVBeSIiImVFmSIiUnm4bINl5MiRzJo1i6+//prg4GDS09MBCA0NJSAgwOLqRETEXShPRESkLChPRESkrChTREQqD5edwXK+SxKTkpIYNmzYRR3bnffmFBGxmrudQ5UnIiKuyd3OoeWZJ+B+j4eIiCtxt3OonqOIiLimSjWDxUX7PiIi4maUJyIiUhaUJyIiUlaUKSIilYen1QWIiIiIiIiIiIiIiIi4GzVYREREREREREREREREnKQGi4iIiIiIiIiIiIiIiJPUYBEREREREREREREREXGSyw65L0+/DRPLzs62uBIREffz27lTgxmVJyIiF0N5ci5liohI6SlT/o/yRESk9EqTJ1WywZKTkwNAbGysxZWIiLivnJwcQkNDrS7DUsoTEZGLpzw5Q5kiInLxlCnKExGRsuBMnniYKtjedzgcHDlyhODgYDw8PJz63uzsbGJjYzl48CAhISHlVKHrqYrr1pq15srsYtZtjCEnJ4fo6Gg8Pav2TpMXkydQNX//tGatuTKriutWnpQdPUdxjtZcNdYMVXPdWrPza1am/B/lifOq4rq1Zq25Mqvo5yhV8goWT09PYmJiLuoYISEhVeoX8zdVcd1ac9VQFdcMpV93VX9X2G/KIk+gav7+ac1VQ1VcM1TNdStPLp6eo5SO1lx1VMV1a83OUaacoTwpvaq4bq25aqiKa4aKe45Stdv6IiIiIiIiIiIiIiIipaAGi4iIiIiIiIiIiIiIiJPUYHGSn58fzz33HH5+flaXUqGq4rq15qqhKq4Zqu66XU1V/HfQmquGqrhmqJrrroprdkVV8d9Ba646quK6tWaxSlX9d6iK69aaq4aquGao+HVXySH3IiIiIiIiIiIiIiIiF0NXsIiIiIiIiIiIiIiIiDhJDRYREREREREREREREREnqcEiIiIiIiIiIiIiIiLiJDVYREREREREREREREREnKQGCzBx4kTi4uIIDg4mIiKCfv36sX379nPuM2zYMDw8PM756NKlyx+OtXLlSnr06EFgYCDVq1ene/fuFBQUVNRSLtjzzz//h/VERUWd/fpXX33FNddcQ3h4OB4eHqxbt+6c78/IyGDUqFE0a9aMatWqUbduXR588EGysrIqeCV/bdmyZdxwww1ER0fj4eHB3Llzz/m6MYbnn3+e6OhoAgIC6N69O5s3bz77dWfXWVRURPv27f/0MasIF/K7/Hdr/v19e/fu/aeP3Y4dO+jbty/h4eGEhIRw6aWXsnjx4vJa2gWbOHEiHh4ePPzww2dvu5A17969m/79+1OrVi1CQkK49dZbOXbs2B+OP3/+fBISEggICCA8PJybbrqpvJf0p2w2G2PHjqVBgwYEBATQsGFDXnjhBRwOx9n7/N3fMfz9uvft28eIESPO/pxGjRrx3HPPUVxcXBHLdDtVMU+gamRKVcsTUKZA1cgU5YnrqoqZojxRnihP3DdPQJniqpQnlTNPoOplivJEeeKqeaIGC7B06VJGjhzJqlWrWLhwITabjV69epGXl3fO/a699lqOHj169uO777475+srV67k2muvpVevXqSkpJCamsoDDzyAp6drPsytWrU6Zz0bN248+7W8vDwuvfRSXn755T/93iNHjnDkyBFee+01Nm7cyNSpU1mwYAEjRoyoqPIvSF5eHu3ateOdd975069PmjSJN954g3feeYfU1FSioqK4+uqrycnJAZxf5xNPPEF0dHS5refvXMjv8t+t+X/961//wsPD409/Vp8+fbDZbCxatIg1a9bQvn17rr/+etLT08ttfX8nNTWVDz74gLZt255z+9+tOS8vj169euHh4cGiRYv45ZdfKC4u5oYbbjjn5P3ll18yePBghg8fzvr16/nll1+44447KnSNv3nllVf473//yzvvvMPWrVuZNGkSr776Km+//fbZ+/zd3/GFrHvbtm04HA7ef/99Nm/ezJtvvsl///tfxowZUyHrdDdVNU+g8mdKVcsTUKZUlUxRnriuqpopyhPlifLEPfMElCmuSnlSOfMEql6mKE+UJy6bJ0b+4Pjx4wYwS5cuPXvb0KFDTd++ff/y+xISEszYsWPLubqy8dxzz5l27dr97f327t1rALN27dq/ve9nn31mfH19TUlJycUXWA4AM2fOnLOfOxwOExUVZV5++eWztxUWFprQ0FDz3//+97zHOd86v/vuO9O8eXOzefPmC37Mytvvf5edWfO6detMTEyMOXr06B8euxMnThjALFu27Oxt2dnZBjA//fRT+S7qPHJyckyTJk3MwoULzRVXXGEeeughY8yFrfmHH34wnp6eJisr6+x9MjIyDGAWLlxojDGmpKTE1KlTx3z00UcVt6i/0KdPH5OYmHjObTfddJMZNGjQH+57vr/jC1n3n5k0aZJp0KDBxS2giqgKeWJM1cuUqpgnxihTjKmcmaI8cR9VIVOUJ8qT3yhP3C9PjFGmuAvlyf+pLHliTNXMFOWJ8sQY18gT12wzW+y3S+HCwsLOuX3JkiVERETQtGlT7r77bo4fP372a8ePHyc5OZmIiAguueQSIiMjueKKK1ixYkWF1u6MnTt3Eh0dTYMGDbj99tvZs2fPRR0vKyuLkJAQvL29y6jC8rV3717S09Pp1avX2dv8/Py44oor+PXXX8/7fX+2zmPHjnH33XczY8YMqlWrVq51O+P3v8sXuub8/HwGDhzIO++8c85ltL+pWbMmLVq0YPr06eTl5WGz2Xj//feJjIykU6dO5byqPzdy5Ej69OlDz549z7n9QtZcVFSEh4cHfn5+Z+/j7++Pp6fn2b/htLQ0Dh8+jKenJx06dKB27dr07t37vJealrfLLruMn3/+mR07dgCwfv16VqxYwXXXXXfBx7iQdf+ZrKysP5wf5c9VlTyBqp0pVSFPQJkClTNTlCfuo6pkivJEeQLKE3C/PAFlirtQnpSOO+UJVI1MUZ4oT/5KReaJGiy/Y4xh9OjRXHbZZbRu3frs7b179+bjjz9m0aJFvP7666SmptKjRw+KiooAzp6on3/+ee6++24WLFhAx44dueqqq9i5c6cla/krCQkJTJ8+nR9++IEPP/yQ9PR0LrnkEk6dOlWq4506dYrx48dz7733lnGl5ee3y/oiIyPPuT0yMvK8l/z92TqNMQwbNoz77ruPzp07l1/BTvqz3+ULXfMjjzzCJZdcQt++ff/02B4eHixcuJC1a9cSHByMv78/b775JgsWLKB69erls6C/MHv2bNLS0pg4ceIfvnYha+7SpQuBgYE8+eST5Ofnk5eXx+OPP47D4eDo0aPAuX/jY8eO5dtvv6VGjRpcccUVZGRklOfy/tSTTz7JwIEDad68OT4+PnTo0IGHH36YgQMHXvAxLmTdv7d7927efvtt7rvvvrJaSqVVVfIElCmVPU9AmfKbypgpyhP3UFUyRXmiPPlfyhP3yhNQprgD5UnVyBOo/JmiPDlDeXJ+FZknarD8zgMPPMCGDRv45JNPzrn9tttuo0+fPrRu3ZobbriB77//nh07djB//nyAs3u33XvvvQwfPpwOHTrw5ptv0qxZM6ZMmVLh6/g7vXv3ZsCAAbRp04aePXueXce0adOcPlZ2djZ9+vShZcuWPPfcc2Vdarn7/X6Lxpg/3YPxfOt8++23yc7O5umnny73Wp1xvt9l+Os1z5s3j0WLFvGvf/3rvMc2xvCPf/yDiIgIli9fTkpKCn379uX6668/70mqvBw8eJCHHnqImTNn4u/vf977/dWaa9Wqxeeff84333xDUFAQoaGhZGVl0bFjR7y8vID/+xt/5plnGDBgAJ06dSIpKQkPDw8+//zzclrd+X366afMnDmTWbNmkZaWxrRp03jttdec+hu+kHX/ryNHjnDttddyyy23cNddd5XlciqlqpInoEz5TWXNE1Cm/F5lyhTliXuoKpmiPDlDeXKG8sS98gSUKe5AeVK18gQqb6YoT86lPPmjCs0TpzYUq+QeeOABExMTY/bs2XNB92/cuPHZPe727NljADNjxoxz7nPrrbeaO+64o8xrLQ89e/Y099133zm3/d1+lNnZ2aZr167mqquuMgUFBRVQZenxuz0Vd+/ebQCTlpZ2zv1uvPFGM2TIkHNu+6t19u3b13h6ehovL6+zH4Dx8vL6w3Eqyvl+ly9kzQ899JDx8PD4w3o8PT3NFVdcYYwx5qeffvrDPobGnPmbmDhxYvkt7E/MmTPn7OP9v/X+toZdu3Zd8L+zMWf22jx9+rQxxpjIyEgzadIkY4wxixYtMoBZvnz5OfePj483Y8aMKZ/F/YWYmBjzzjvvnHPb+PHjTbNmzf5w3wvZV/Z86/7N4cOHTdOmTc3gwYON3W6/6Poru6qeJ8ZU7kypSnlijDKlsmeK8sT1VfVMUZ4oT4xRnrhDnhijTHF1ypPKmyfGVK1MUZ4oT/6XK+SJrmDhTFfvgQce4KuvvmLRokU0aNDgb7/n1KlTHDx4kNq1awNQv359oqOj2b59+zn327FjB/Xq1SuXustSUVERW7duPbueC5GdnU2vXr3w9fVl3rx5f9lJdUUNGjQgKiqKhQsXnr2tuLiYpUuXcskll5y97e/W+e9//5v169ezbt061q1bx3fffQec6bZOmDChYhbz//3d7/KFrPmpp55iw4YNZ9ezbt06AN58802SkpKAM/tVAnh6nnsK8fT0PNv1rihXXXUVGzduPKfezp07c+edd7Ju3ToaNmx4Qf/OvwkPD6d69eosWrSI48ePc+ONNwLQqVMn/Pz8zvkbLykpYd++fZb8jefn5//h8ffy8ir143++dQMcPnyY7t2707FjR5KSkv7wc+X/KE/OqGqZUhnzBJQpVSVTlCeuS5miPAHlifLkDHfIE1CmuCrlSdXLE6icmaI8UZ64bJ443ZKphO6//34TGhpqlixZYo4ePXr2Iz8/3xhjTE5Ojnn00UfNr7/+avbu3WsWL15sunbtaurUqWOys7PPHufNN980ISEh5vPPPzc7d+40Y8eONf7+/mbXrl1WLe28Hn30UbNkyRKzZ88es2rVKnP99deb4OBgs2/fPmOMMadOnTJr16418+fPN4CZPXu2Wbt2rTl69Kgx5kx3OyEhwbRp08bs2rXrnMfNZrNZubRz5OTkmLVr15q1a9cawLzxxhtm7dq1Zv/+/cYYY15++WUTGhpqvvrqK7Nx40YzcOBAU7t27bP/rqVZ54V0TsvL3/0uG/P3a/4z/O6dECdOnDA1a9Y0N910k1m3bp3Zvn27eeyxx4yPj49Zt25deS7xglxxxRXmoYceOvv5hax5ypQpZuXKlWbXrl1mxowZJiwszIwePfqc4z700EOmTp065ocffjDbtm0zI0aMMBERESYjI6OilnbW0KFDTZ06dcy3335r9u7da7766isTHh5unnjiibP3+bu/Y2P+ft2HDx82jRs3Nj169DCHDh065/dK/qgq5okxVSNTqlqeGKNM+U1lzxTlieuqipmiPFGeKE/cN0+MUaa4KuVJ5cwTY6pepihPzlCenOFKeaIGiznzh/RnH0lJScYYY/Lz802vXr1MrVq1jI+Pj6lbt64ZOnSoOXDgwB+ONXHiRBMTE2OqVatmunbt+odLq1zFbbfdZmrXrm18fHxMdHS0uemmm8zmzZvPfj0pKelPH5PnnnvOGGPM4sWLz/u47d2715pF/Ynz1Tl06FBjjDEOh8M899xzJioqyvj5+Zlu3bqZjRs3/u33/9U6rQybv/tdNubv13y+4/5v2BhjTGpqqunVq5cJCwszwcHBpkuXLua7774rh1U57/dhcyFrfvLJJ01kZKTx8fExTZo0Ma+//rpxOBzn3Ke4uNg8+uijJiIiwgQHB5uePXuaTZs2VcSS/iA7O9s89NBDpm7dusbf3980bNjQPPPMM6aoqOjsff7u79iYv1/3+Y6h/vyfq4p5YkzVyJSqlifGKFN+U9kzRXniuqpipihPlCfKE/fNE2OUKa5KeVI588SYqpcpypMzlCdnuFKeeBhjDCIiIiIiIiIiIiIiInLBtEmliIiIiIiIiIiIiIiIk9RgERERERERERERERERcZIaLCIiIiIiIiIiIiIiIk5Sg0VERERERERERERERMRJarCIiIiIiIiIiIiIiIg4SQ0WERERERERERERERERJ6nBIiIiIiIiIiIiIiIi4iQ1WERERERERERERERERJykBouIiIiIiIiIiIiIiIiT1GARERERERERERERERFxkhosIiIiIiIiIiIiIiIiTlKDRURERERERERERERExElqsIiIiIiIiIiIiIiIiDhJDRYREREREREREREREREnqcEicoGef/55PDw8OHnypNWlnPVbTc4aNGgQHh4eXH/99eVQlYiI/BV3zpPf7vf7D39//wqoUkRE/pc75wmAMYakpCTi4+MJDAwkJCSEjh078vXXX5dzlSIi8nvunCn169f/0+coep4iFcXb6gJEpGLNnz+fuXPnEhISYnUpIiLiphYsWEBoaOjZzz099Z4dERFxzv3338/UqVN55JFHmDhxIjabjY0bN5Kfn291aSIi4kbmzJlDUVHRObcdOHCA2267jf79+1tUlVQlarCIVCFZWVnce++9jB8/nrfeesvqckRExE116tSJ8PBwq8sQERE3NXfuXN5//30+/fRTbr311rO3X3PNNRZWJSIi7qhDhw5/uO2HH34A4K677qrocqQK0tsNRZx07NgxBg4cSGhoKJGRkSQmJpKVlXXOfYwxvPvuu7Rv356AgABq1KjBzTffzJ49e86538KFC+nbty8xMTH4+/vTuHFj7r333j+9JHP+/Pm0b98ePz8/GjRowGuvveZ07Y8++ii1a9fmwQcfdPp7RUSkbLlznoiIiOtwxzx56623qF+//jnNFRERsZ47Zsrv/bYFZcOGDenRo0epjyNyodRgEXHSgAEDaNq0KV9++SVPPfUUs2bN4pFHHjnnPvfeey8PP/wwPXv2ZO7cubz77rts3ryZSy65hGPHjp293+7du+natSvvvfceP/74I88++yzJyclcdtlllJSUnL3fzz//TN++fQkODmb27Nm8+uqrfPbZZyQlJV1w3T/99BPTp0/no48+wsvL6+IfCBERuSjumicAbdq0wcvLi8jISIYMGcKBAwcu7sEQEZFSc7c8sdlsrFy5kg4dOvDGG29Qr149vLy8aNiwIa+99hrGmLJ7cERExCnulil/5qeffmL//v0kJiaWam6xiNOMiFyQ5557zgBm0qRJ59z+j3/8w/j7+xuHw2GMMWblypUGMK+//vo59zt48KAJCAgwTzzxxJ8e3+FwmJKSErN//34DmK+//vrs1xISEkx0dLQpKCg4e1t2drYJCwszF/JnnJOTY+rXr2+efvrps7fVq1fP9OnT5+8XLiIiZcqd82T69OlmwoQJ5rvvvjOLFi0yL7/8sgkLCzORkZHm0KFDF/wYiIjIxXPXPDl69KgBTEhIiImJiTHTpk0zP//8s7nvvvsMYMaMGePU4yAiIhfPXTPlz9x2223Gy8tLz0+kwugKFhEn3Xjjjed83rZtWwoLCzl+/DgA3377LR4eHgwaNAibzXb2Iyoqinbt2rFkyZKz33v8+HHuu+8+YmNj8fb2xsfHh3r16gGwdetWAPLy8khNTeWmm27C39//7PcGBwdzww03XFDNTz31FD4+Pjz77LMXs3QRESlD7pgngwcPZsyYMfTu3Zsrr7ySJ598ku+//54TJ04wadKki3k4RESklNwtTxwOBwDZ2dl8/vnnDBkyhB49evDee+/Rr18/3njjDXJzcy/qMRERkdJxt0z5vYyMDObOncu1115LnTp1nP5+kdLQkHsRJ9WsWfOcz/38/AAoKCgAzuxXaYwhMjLyT7+/YcOGwJknFr169eLIkSOMGzeONm3aEBgYiMPhoEuXLmePd/r0aRwOB1FRUX841p/d9nspKSm8++67fPXVVxQWFlJYWHj259tsNjIzMwkICDi7DhERqRjulifnEx8fT9OmTVm1alWpjyEiIqXnbnlSo0YNPDw8CA4OpkuXLud8rXfv3sydO5ctW7YQHx//t8cSEZGy5W6Z8nszZ86kqKhIw+2lQqnBIlLGwsPD8fDwYPny5X/atPjttk2bNrF+/XqmTp3K0KFDz359165d59z/tycg6enpfzjWn932e1u2bMEYQ//+/f/wtYMHD1KjRg3efPNNHn744b89loiIVBxXy5O/YozB01MXRouIuCJXy5OAgACaNGnyp/c1/3/+ijJFRMQ1uVqm/N7kyZOJjIzk+uuvd/p7RUpL/2sRKWPXX389xhgOHz5M586d//DRpk0bgLODtn4fSO+///45nwcGBhIfH3/2CpTf5OTk8M033/xtPddeey2LFy/+w0dkZCRdunRh8eLF3HzzzRe7bBERKWOulifns2rVKnbu3PmHdyGLiIhrcMU8GTBgANnZ2fz666/n3P7dd98RFBREq1atnF6niIiUP1fMlN+sXr2aDRs2MHToULy9dU2BVBz9tomUsUsvvZR77rmH4cOHs3r1arp160ZgYCBHjx5lxYoVtGnThvvvv5/mzZvTqFEjnnrqKYwxhIWF8c0337Bw4cI/HHP8+PFce+21XH311Tz66KPY7XZeeeUVAgMDycjI+Mt6oqKi/vSySn9/f2rWrEn37t3LaukiIlKGXC1PANq1a8egQYNo0aIF/v7+pKSk8OqrrxIVFcUTTzxRHg+DiIhcJFfMk8cee4yPP/6YW265hfHjxxMTE8MXX3zBvHnzeO211wgICCiPh0JERC6SK2bKbyZPngzAiBEjymy9IhdCDRaRcvD+++/TpUsX3n//fd59910cDgfR0dFceumlZ/cS9vHx4ZtvvuGhhx7i3nvvxdvbm549e/LTTz9Rt27dc4539dVXM3fuXMaOHcttt91GVFQU//jHPygoKOCf//ynFUsUEZEK4Gp50rJlSz744AOOHj1KcXEx0dHR3H777Tz77LPUrl27XB4DERG5eK6WJ2FhYaxYsYInnniCxx57jLy8PJo3b86UKVMYPnx4uTwGIiJSNlwtU+DMjJhPPvmEbt260bRp0zJfs8hf8TC/bXIqIiIiIiIiIiIiIiIiF0QzWERERERERERERERERJykBouIiIiIiIiIiIiIiIiT1GARERERERERERERERFxkhosIiIiIiIiIiIiIiIiTlKDRURERERERERERERExElqsIiIiIiIiIiIiIiIiDjJ2+oCrOBwODhy5AjBwcF4eHhYXY6IiFsxxpCTk0N0dDSenlW7T688EREpPeXJuZQpIiKlp0z5P8oTEZHSK02eVMkGy5EjR4iNjbW6DBERt3bw4EFiYmKsLsNSyhMRkYunPDlDmSIicvGUKcoTEZGy4EyeVMkGS3BwMHDmgQoJCbG4GhER95KdnU1sbOzZc2lVpjwRESk95cm5lCkiIqWnTPk/yhMRkdIrTZ5UyQbLb5dIhoSEKGxEREpJl5srT0REyoLy5AxliojIxVOmKE9ERMqCM3lStTemFBERERERERERERERKQU1WERERERERERERERERJykBouIiIiIiIiIiIiIiIiT1GARERERERERERERERFxkss3WPr370+NGjW4+eabz7k9JyeHuLg42rdvT5s2bfjwww8tqlBERNyB8kRERMqC8kRERMqKMkVExP15GGOM1UX8lcWLF5Obm8u0adP44osvzt5ut9spKiqiWrVq5Ofn07p1a1JTU6lZs+bfHjM7O5vQ0FCysrIICQkpz/JFRCoddz2HKk9ERFyLu55DyyNPwH0fDxERV+Cu51A9RxERcS2lOYe6/BUsV155JcHBwX+43cvLi2rVqgFQWFiI3W7HxXtFIiJiIeWJiIiUBeWJiIiUFWWKiIj7s7zBMnHiROLi4ggODiYiIoJ+/fqxffv2C/rezMxM2rVrR0xMDE888QTh4eHlXK2IiLgq5YmIiJQF5YmIiJQVZYqISOVneYNl6dKljBw5klWrVrFw4UJsNhu9evUiLy/vb7+3evXqrF+/nr179zJr1iyOHTtWARWLiLi/7fszcDgq1zuglCciIhXvVFYBx0/nW11GmVKeiIhUPLvDsOPAaavLKHPKFBGRirfjwGnsdkeF/TzvCvtJ57FgwYJzPk9KSiIiIoI1a9bQrVu3CzpGZGQkbdu2ZdmyZdxyyy1/+HpRURFFRUVnP8/Ozr64okVE3JQxhjlLdjF1/hZuuaopg3u3sLqkMqM8ERGpWDsOnGZCUjIhgX5MGnU5AX6WP7UoExWRJ6BMERH5TX5hCa99vIa124/z4n2X0qrhhc2ucgd6jiIiUnGMMcz/ZS8ffr2J6y9rwN1921TIz7X8Cpbfy8rKAiAsLOwv73fs2LGzoZGdnc2yZcto1qzZn9534sSJhIaGnv2IjY0t26JFRNxAic3Ov2avJenbLRgDOXnFlXofX+WJiEj5Wb72ME//ZwUZ2UXYHYbc/BKrSyo35ZEnoEwREQFIP5XH428vJ3XLMTw9PMjMLfr7b3Jjeo4iIlI+bHYH7325gffnbMTx/5+f2Cto5xYP40Kvrhlj6Nu3L6dPn2b58uUAXHPNNaSlpZGXl0dYWBhz5swhLi6ONWvWMGLECIwxGGO4//77uf/++//0uH/WzY+NjSUrK4uQkJAKWZuIiJUyc4p4aWoKW/dl4OkBd/Vtw/WXNcDDw8PpY2VnZxMaGurS51DliYhI+XA4DLN+3ManC3cA0LlFJI8P6kQ1fx+nj1WV8wSUKSIim3af5KWpqeTkFxMW4sczwxNoWrdGqY5VlTNFeSIiVV12XjGvTE9lw66TeHjA0OtactOVjSvsNS+XarCMHDmS+fPns2LFCmJiYsrt57hD8IqIlJW9R7IYPyWZE6cLCPT35okhcXRsFlHq47nDOVR5IiJS9gqLbLw5O41fNxwFoN8VjRh2fSu8PJ1/4gLucQ6tqDwB93g8RETKyo/J+3nvy/XY7IbGMaGMTUygZmhAqY/nDudQPUcRESl7B4/lMH5yMkdP5RHg58Wjd3QioXXtUh+vNOdQl9koedSoUcybN49ly5aV+5MXEZGqYtWmo7z+8RoKi+1EhwcybkQCMRHBVpdVrpQnIiJl72RmAeOnJLPncBbeXh6MvLkdPePrWV1WuVKeiIiUPbvDkPTNZr5ethuAS9tF8/DtHfD3dZmXp8qFMkVEpOyt2XaMSTNWk19oIyKsGuMSE6hfu+Iby5YnmDGGUaNGMWfOHJYsWUKDBg2sLklExO0ZY/hi0U5mfL8VY6Bdk3CeHBJHcDVfq0srN8oTEZHysX1/BhOSUjidU0RokC9PD42vVAOIf095IiJSPvIKSpg0czVp244DcMc1zbn96qal2sLFXShTRETKnjGGb5bvYfK8TTgMtGwQxphh8YQG+VlSj+UNlpEjRzJr1iy+/vprgoODSU9PByA0NJSAgNJfHioiUlUVl9h5+/N1LFlzCIA+lzbgrr6t8fbytLiy8qU8EREpe0vSDvHvT9dSYnNQv3YIYxMTiAyrZnVZ5Up5IiJS9o6ezGP8lFUcPJaLr48XjwzswGXt6lhdVrlTpoiIlK0Sm4P/frWBH5P3A3B1fF3uH9AOH2/rXvOyfAbL+d6pkJSUxLBhw8rlZ2o/ShGprE5nFzJhagrb95/G09ODe/u34bpLyvZdUq56DlWeiIiUHYfDMHPBVj7/eScACa2iGH1Hx1INsz8fVz2HWpEn4LqPh4jIxdq46yQTp6WQk19CzVB/xg5PoHFs9TL9Ga56DtVzFBGRspOVW8TL01PZtPsUnh4w/IZW9O3WqEyvhHTLGSwW93dERCqN3YcyeTEphZOZBQQF+PDU0DjaNalldVkVRnkiIlI2CopsvPlJGis3nhlmf3OPJgzu3QLPUg6zdzfKExGRsrNg5T7++9UG7A5Dk9jqjE1MICzE3+qyKowyRUSkbOxPz+bFKcmkn8onwM+bJwZ3pnOLSKvLAlygwSIiIhfv1w1HeOOTNIqK7dSpFcSzIxKIrhVkdVkiIuJmjp/O58Upyew9ko23lyejbm1Pj86xVpclIiJuxm538NG8TXy7Yi8A3TrU4cHbOuDn42VxZSIi4m5St6Tz6sw1FBTZiKp5Zph93SjXuUJPDRYRETdmjOGzn3Ywc8E2ADo0rcUTQ+IICii7LVxERKRq2LbvzDD7zNwiqgf58czweJrXD7O6LBERcTO5BSVMmp7K2h0nABjUuzm3XlW5h9mLiEjZM8Ywd+lukr7djDHQulFNnhoSZ9kw+/NRg0VExE0Vldj596drWbb2MAA3XN6QETe0wquSD7MXEZGyt2j1Ad7+bD02u4MG0WeG2UfUqNzD7EVEpOwdOZHLC5OTOXwiFz9fL0YP7MglbaOtLktERNxMic3Ou19s4KfUAwBc06Ue9/Zva+kw+/NRg0VExA2dyipgQlIKOw9m4uXpwX03teXarvWtLktERNyM3WGY8d0Wvly8C4CubWrzyMCOBPjpaYKIiDhn3Y7jvDJ9NbkFJYRXD2BcYgIN64RaXZaIiLiZzJwiJk5LYcveDDw9YETf1txwWUOXvRJSz5xERNzMroOZvJiUzKmsQoKr+fD00HjaNA63uiwREXEz+YUlvP5xGilb0gG4tWdT7rymeZUZZi8iImVn/i97+WDuRhwOQ7N6NXhmWDw1qtAwexERKRv7jmYzfvIqjp8uINDfmycGx9GxeYTVZf0lNVhERNzI8nWH+dfstRSX2ImNDGJcYhdqhwdaXZaIiLiZYxn5jJ+8iv3pOfh4e/LgbR3o3jHG6rJERMTN2OwOPpi7ke9/3QdA904xjLqlPb4aZi8iIk5K3nSU12etoaDITu3wQMYlJhAbGWx1WX9LDRYRETfgcBhmL9zOJz9uB6BT8wgeH9SZQA2zFxERJ23ec4qXpqaQnVdMjeAzw+yb1dMwexERcU5OfjGvTE9l/c6TeHjA4N4tuLlHE5fdwkVERFyTMYYvF+9i+ndbMAbaNg7nqaFxBFfztbq0C6IGi4iIiysstvGv2Wv5Zf0RAPpd0Yhh17fCS1u4iIiIk35K2c9/vliPzW5oFBPK2OEJhFcPsLosERFxMweP5TB+SjJHT+bh7+vFo3d2okvr2laXJSIibqa4xM5/vljPotUHAejdtT739G+Dt5frDbM/HzVYRERc2MnMAl5MSmb3oSy8vTy4f0A7eiXUs7osERFxM3aHYeq3m5m7dDcAl7aN5uHbO+CvYfYiIuKktO3HmTQ9lbxCGxE1AhibmECDaA2zFxER55zOKeSlpBS27T+Np6cH9/RrQ59LG1hdltP0jEpExEXtOHCaCUnJZGQXERLoy5hh8bRqWNPqskRExM3kF5bw6sw1rN56DICBvZpx+9XNNMxeREScYozhmxV7mPz1JhwGWtQPY8yweKoH+1ldmoiIuJk9h7MYPyWZk5kFBAb48NSQzrRv6trD7M9HDRYRERe0NO0Q//50LcU2B/WighmbmEBUTQ2zFxER56SfyuOFyckcPJaDr7cnDw/syOXt61hdloiIuBmb3cF/v9rAD6v2A3BVXCwjb26Hj7eG2YuIiHNWbjzC67PSKCq2U6dWIONGdKFOrSCryyo1NVhERFyIw2GY9cM2Pv1pBwDxLaN49M6OVPPXMHsREXHOxt0nmTg1lZz8YsJC/BmbGE+T2BpWlyUiIm4mO6+Yl6elsnH3mWH2w69vRb8rGmmYvYiIOMUYw+c/72TG91sBaN+0Fk8O7kyQmwyzPx81WEREXERhkY03Pklj5cajAAy4sjGDr2upYfYiIuK0H1bt470vN2B3GJrEVueZ4fHUDNUwexERcc6B9GzGT0km/VQ+AX7ePD6oE3Eto6wuS0RE3ExRiZ23P13H0rWHALj+sgbcdWNrvNxomP35qMEiIuICTpwu4MUpyew5koW3lycP3NKOq+LqWl2WiIi4GbvdwZRvNjNv+R4AurWvw4O3d8DPR1u4iIiIc1ZvPcarM1eTX2gjMqwa40YkUC8qxOqyRETEzWRkFzIhKZkdBzLx8vTg3pva0rtrfavLKjNqsIiIWGzb/gwmJKWQmVNE9SA/xgyLp0WDMKvLEhERN5NbUMKrM1aTtv04AHde25zbejbVFi4iIuIUYwxfL9tD0jdnhtm3aliTp4fGERqkYfYiIuKcXYcymTAlmZNZhQQF+PD0sDjaNq5ldVllSg0WERELLV5zkLc/W0eJzUH92iGMS0wgIqya1WWJiIibOXIilxcmJ3P4RC5+vl48MrAjl7aNtrosERFxMyU2B+99uZ6FKQcAuDq+LvcPaIePt/tv4SIiIhXrlw1HeGNWGsUldmIighg3IoHocPcdZn8+arCIiFjA4TDM+H4rXyzaCUCX1lGMvqMTAX46LYuIiHPW7zzBy9NSyS0oITzUn7GJCTSKqW51WSIi4maycouYOC2VzXtO4ekBiTe25sbLG+pKSBERcYoxhtkLdzDrh20AdGwewRODOhMY4GNxZeVDr+SJiFSwgiIbr3+8huTN6QDcclUTBl3bAk8NsxcRESd99+te3p+zEYfD0KxuDcYMjycsxN/qskRExM3sP5rNC1OSOZ6RTzV/bx4f1JnOLSKtLktERNxMYbGNt2avZcX6IwDc2K0hide3qhTD7M9HDRYRkQp0PCOf8VOS2Xc0Gx9vTx68tT3dO8VaXZaIiLgZu93Bh19vYv4vewHo3jGGUbe2x1fD7EVExEkpW9J5beZqCors1K4ZyLgRCcRGBltdloiIuJlTWQW8mJTCroNnhtnfP6Ad13SpZ3VZ5U4NFhGRCrJl7ylemppCVm4x1YP9eGZ4PM3raZi9iIg4Jze/mFemr2bdzhMADLmuBTf3aKItXERExCnGGOYs2cXU+VswBto0CuepoXGEBPpaXZqIiLiZHQdOMyEpmYzsIoKr+fL0sDjaNAq3uqwKoQaLiEgF+CnlAP/5Yh02u6FhdChjExOoVSPA6rJERMTNHDqew/jJyRw5mYe/rxej7+hE1za1rS5LRETcTInNzjufr2fR6oMAXNu1Pvf2b4N3Jd7CRUREysfytYf51+w0im0OYiODeXZEAlE1A60uq8KowSIiUo7sDsO0+VuYs2QXAF3b1Gb0wI74a5i9iIg4ae3247wyYzV5BSWEVw/g2REJNIgOtbosERFxM5k5Rbw0NYWt+zLw9PTg7r6t6XNpA10JKSIiTnE4DLN+3ManC3cA0LlFJI8P6kQ1/8o5zP589AqfiEg5yS8s4bWP15C65RgAt13dlDt6NdcwexERcYoxhvm/7OXDrzfhcBha1A/j6WFx1AjWMHsREXHO3iNZjJ+SzInTBQT6e/PkkDg6NIuwuiwREXEzhUU23pydxq8bjgLQv3tjhvZpiVcVfM1LDRYRkXKQfiqP8VOSOZCeg6+3Jw/f3pHLO9SxuiwREXEzNruDD+Zs5PuV+wDo0TmWB25ph4+3htmLiIhzVm06yusfr6Gw2E50+Jlh9jERGmYvIiLOOZlZwPgpyew5nIW3lwcjb25Pz/i6VpdlGTVYRETK2KbdJ3lpaio5+cWEhfjxzPAEmtatYXVZIiLiZrLzinlleiobdp3EwwOG9WlJ/+6NtYWLiIg4xRjDF4t2MuP7rRgD7ZvU4skhnQmqpmH2IiLinO37M5iQlMLpnCJCg3wZMyyelg1qWl2WpdRgEREpQz8m7+e9L9djsxsax5wZZl8zVMPsRUTEOQePnRlmf/RUHgF+Xjx2Z2fiW0VZXZaIiLiZ4hI7b3++jiVrDgHQ59IG3NW3tYbZi4iI05akHeLfn66lxOagfu0QxiYmEBlWzeqyLKcGi4hIGbA7DEnfbObrZbsBuKxdNA/d3gF/X51mRUTEOWu2HWPSjNXkF9qICKvGuMQE6tcOsbosERFxM6ezC5kwNYXt+0/j6enBvf3bcN0lDawuS0RE3IzDYZi5YCuf/7wTgIRWUYy+o2OVG2Z/PnrlT0TkIuUVlDBp5mrSth0H4I5rmnP71U21hYuIiDjFGMO85XuYMm8TDgOtGtbk6aFxhAb5WV2aiIi4md2HMnkxKYWTmQUEBfjw1NA42jWpZXVZIiLiZgqKbLz5SRorN54ZZn9zjyYM7t0Czyo4zP581GAREbkIR0/mMX7KKg4ey8XXx4tHBnbgsnYaZi8iIs4psTn471cb+DF5PwBXx9fl/gHt8PHWFi4iIuKcXzYc4c1P0igqtlOnVhDPjkggulaQ1WWJiIibOZ6Rz4tJyew9ko23lyejbm1Pj86xVpflctRgEREppY27TjJxWgo5+SXUDPVn7PAEGsdWt7osERFxM1m5RUyclsrmPafw9IDhN7Sib7dGuhJSREScYozh05928PGCbQB0aFqLJ4bEERSgLVxERMQ5W/dm8NLUFDJzi6ge5Mczw+NpXj/M6rJckhosIiKlsGDlPv771QbsDkOT2OqMTUwgLMTf6rJERMTN7E/PZvzkZI5l5BPg580TgzvTuUWk1WWJiIibKSqx8+/Za1m27jAAN1zekBE3tMJLw+xFRMRJi1Yf4O3P1mOzO2gQfWaYfUQNDbM/HzVYREScYLc7+GjeJr5dsReAbh3q8OBtHfDz8bK4MhERcTepW9J5deYaCopsRNU8M8y+bpSG2YuIiHNOZRUwISmFnQcz8fL04L6b2nJt1/pWlyUiIm7G7jDM+G4LXy7eBUDXNrV5ZGBHAvzUQvgrenRERC5Qbn4xr8xYzbodJwAY1Ls5t16lYfYiIuIcYwxzluxm6vzNGAOtG9XkqSEaZi8iIs7befA0L05JISO7kOBqPjw9NJ42jcOtLktERNxMfmEJr3+cRsqWdABu7dmUO69prmH2F0ANFhGRC3D4RC7jJ6/i8Ik8/Hy9GD2wI5e0jba6LBERcTMlNjvvfrGBn1IPAHBNl3rc27+thtmLiIjTlq87zL9mr6W4xE5sZDDjEhOoHR5odVkiIuJmjmXkM37yKvan5+Dj7cmDt3Wge8cYq8tyG2qwiIj8jXU7jvPy9NXkFZQQXj2AcYkJNKwTanVZIiLiZjJzipg4LYUtezPw9IARfVtzw2UNdSWkiIg4xeEwzF64nU9+3A5A5xaRPD6oE9X8NcxeREScs3nPKV6amkJ2XjE1gv0Ym5hA07o1rC7LrajBIiLyF+b/spcP5m7E4TA0q1eDZ4bFU0PD7EVExEn7jmYzfvIqjp8uINDfmycGx9GxeYTVZYmIiJspLLbxr9lr+WX9EQD6XdGIYde3wktbuIiIiJN+StnPf75Yj81uaBQTytjhCYRXD7C6LLejBouIyJ+w2R18MHcj3/+6D4ArO8XwwC3t8dUwexERcVLypqO8PmsNBUV2aocHMi4xgdjIYKvLEhERN3Mys4AXk5LZfSgLby8P/jGgHVcn1LO6LBERcTN2h2Hqt5uZu3Q3AJe2i+bh2zvg76tWQWm4/GbP/fv3p0aNGtx8883n3H7w4EG6d+9Oy5Ytadu2LZ9//rlFFYpIZZOTX8zzH67k+1/34eEBQ/u05JGBHdVccXPKExGpaMYYvli0kwlTUygostO2cTivP9RNzZVKQJkiIhVtx4HTPPrWUnYfyiIk0JcX77tUzZVKQHkiIhUtv7CEF6ckn22uDOzVjCcGdVZz5SJ4GGOM1UX8lcWLF5Obm8u0adP44osvzt5+9OhRjh07Rvv27Tl+/DgdO3Zk+/btBAb+/UC37OxsQkNDycrKIiQkpDzLFxE3c/BYDuOnJHP0ZB7+vl48dmcnElrXtrosl+Ku51DliYhUpOISO+98vo7Faw4B0PuS+tzTrw3eXi7//qYK487nUGWKiFSkpWmH+Penaym2OagXFczYxASiamqY/f9y13Oo8kREKlL6qTxemJzMwWM5+Hp78vDAjlzevo7VZbmU0pxDXb41deWVV7JkyZI/3F67dm1q1z7zomdERARhYWFkZGRcUNiIiPyZtO3HmTQ9lbxCGxE1AhibmECDaA2zryyUJyJSUU5nFzJhagrb95/G09ODe/q1oc+lDawuS8qQMkVEKoLDYZj1wzY+/WkHAPEto3j0zo4aZl+JKE9EpKJs3H2SiVNTyckvJizEn7GJ8TSJ1TD7smD5W+gmTpxIXFwcwcHBRERE0K9fP7Zv3+7UMVavXo3D4SA2NracqhSRyswYw7zlu/nnhyvJK7TRon4Yrz90hZorbkZ5IiKuYM/hLEa/tYzt+08TGODDP+/uouaKG1KmiIjVCotsvDw99WxzZcCVjRkzPF7NFTejPBERV/DDqn2M+++v5OQX0yS2Om883E3NlTJkeYNl6dKljBw5klWrVrFw4UJsNhu9evUiLy/vgr7/1KlTDBkyhA8++KCcKxWRyshmd/CfL9bz4dxNOAxcFRfLhPsvoXqwn9WliZOUJyJitV83HOGJd5ZzMrOAOrUCef2hbrRvGmF1WVIKyhQRsdKJ0wU8+c4KVm48ireXJw/f3oFh17fCy9PD6tLEScoTEbGS3e7gw7kbeefz9dgdhm7t6zBx5GXUDA2wurRKxfItwhYsWHDO50lJSURERLBmzRq6dev2l99bVFRE//79efrpp7nkkkv+8n5FRUVnP8/Ozr64okWkUsjOK+blaals3H0SDw8Yfn0r+l3RCA8PPXFxR8oTEbGKMYbPft7BzO+3AdC+aS2eHNyZoGq+FlcmpaVMERGrbNuXwYSpKWTmFFE9yI8xw+Jp0SDM6rKklJQnImKV3IISXp2xmrTtxwG489rm3NazqV7zKgeWX8Hye1lZWQCEhf31fyCMMQwbNowePXowePDgv7zvxIkTCQ0NPfuhyypF5EB6No+9tYyNu08S4OfNuMQE+ndvrKCpRJQnIlIRikrsvPbxmrPNlesva8Dzd3VRc6WSUaaISEVYtPogT7/7C5k5RdSvHcLrD3VTc6WSUZ6ISEU4ciKXx95aRtr24/j5evHU0Dhuv7qZXvMqJx7GGGN1Eb8xxtC3b19Onz7N8uXLAbjmmmtIS0sjLy+PsLAw5syZQ1xcHCtWrKBbt260bdv27PfPmDGDNm3a/OG4f9bNj42NJSsri5CQkPJfmIi4lNVbj/HqzNXkF9qIDKvGuBEJ1IvSueBCZWdnExoa6tLnUOWJiFSEjOxCJiQls+NAJl6eHtzbvw29L9G8lQvlDnkCyhQRKX8Oh2H6d1v4cvEuALq0jmL0HZ0I8LN80xG34Q6ZojwRkYqwfucJXp6WSm5BCeGh/oxNTKBRTHWry3IbpckTl0rrBx54gA0bNrBixYqzt/3www9/et/LLrsMh8NxQcf18/PDz0/zFESqOmMMXy/bQ9I3Z+attGpYk6eHxhEapPNDZaM8EZHytutQJi9OSeZUViFBAT48PSyOto1rWV2WlANlioiUp/zCEt6YlUby5nQAbrmqCYOubYGn5q1UOsoTESlv3/26l/fnbMThMDSrW4Mxw+MJC/G3uqxKz2UaLKNGjWLevHksW7aMmJgYq8sRkUqmxObgvS/XszDlAABXx9fl/gHt8PF2uZ0S5SIpT0SkvP2y/ghvfJJGcYmdmIggxo1IIDo8yOqypBwoU0SkPB3PyGf8lGT2Hc3Gx9uTB29tT/dO2t6pMlKeiEh5stsdfPj1Jub/sheA7h1jGHVre3x9vCyurGqwvMFijGHUqFHMmTOHJUuW0KCBtlUQkbKVlVvExGmpbN5zCk8PSLyxNTde3lB7T1YyyhMRKW/GGGYv3MGsH87MW+nYPIInBnUmMMDH4sqkrClTRKS8bdl7ipemppCVW0z1YD+eGR5P83qat1LZKE9EpLzl5hfzyvTVrNt5AoAh17Xg5h5N9JpXBbK8wTJy5EhmzZrF119/TXBwMOnpZy6LDQ0NJSAgwOLqRMTd7T+azQtTkjmekU81f2+eGNyZTs0jrS5LyoHyRETKU2Gxjbdmr2XF+iMA9O3WiOHXt8TLS1dCVkbKFBEpTz+lHOA/X6zDZjc0rBPK2OEJ1Kqhc0tlpDwRkfJ06HgO4ycnc+RkHv6+Xjx6Zye6tK5tdVlVjuVD7s/XTUtKSmLYsGHl8jPdYfiZiFy8lC3pvDZzNQVFdmrXDGTciARiI4OtLsvtueo5VHkiIuXlVFYBLyalsOtgJt5eHtw/oB29EupZXZbbc+VzqDJFRMqD3WGYNn8Lc5acGWZ/SdvaPHJ7R/w1zP6iueo5VHkiIuVl7fbjvDJjNXkFJdSqEcC4xAQaRIdaXZbbc8sh9xb3d0SkEjLGMGfJLqbO34Ix0LZxOE8OiSMk0Nfq0qQcKU9EpDzsOHCaCUnJZGQXEVzNlzHD4mjdKNzqsqScKVNEpKzlF5bw2sdrSN1yDIDbrm7KHb2aa5h9Jac8EZGyZoxh/i97+fDrTTgchhb1w3h6WBw1gjXM3iqWN1hERMpSic3OO5+vZ9HqgwD07lqfe/q3wVtbuIiIiJOWrT3EW7PXUmxzUDcqmHGJCUTVDLS6LBERcTPpp/IYPyWZA+k5+Hp78vDtHbm8Qx2ryxIRETdjszv4YM5Gvl+5D4AenWN54JZ2+HhrmL2V1GARkUojM6eIl6amsHVfBp6eHtzdtzV9Lm2gwV4iIuIUh8Mw64dtfPrTDgA6t4jk8UGdqOavYfYiIuKcTbtP8tLUVHLyiwkL8eOZ4Qk0rVvD6rJERMTNZOcV88r0VDbsOomHBwzr05L+3RvrNS8XoAaLiFQKe49kMX5KMidOFxDo782TQ+Lo0CzC6rJERMTNFBbZeHN2Gr9uOApA/+6NGdqnJV7awkVERJz0Y/J+3vtyPTa7oXFMKGMTE6gZqsHmIiLinIPHzgyzP3oqjwA/Lx67szPxraKsLkv+PzVYRMTtrdp0lNc/XkNhsZ3o8DPD7GMiNMxeREScc+J0AS8mJbPncBbeXh6MvLk9PePrWl2WiIi4GbvdQdK3W/h62W4ALmsXzUO3d8DfVy/BiIiIc9ZsO8akGavJL7QREVaNcYkJ1K99YcPXpWIo3UXEbRlj+GLRTmZ8vxVjoH2TWjw5pDNB1TTMXkREnLN9fwYvJqWQmVNEaJAvY4bF07JBTavLEhERN5NXUMKkmatJ23YcgDuuac7tVzfVFi4iIuIUYwzfLN/D5HmbcBho1bAmTw+NIzTIz+rS5HfUYBERt1RcYuftz9exZM0hAPpc2oC7+rbWMHsREXHakjUH+fdn6yixOahfO4SxiQlEhlWzuiwREXEzR07mMn5yMoeO5+Lr48UjAztwWTsNsxcREeeU2Bz896sN/Ji8H4Cr4+ty/4B2+HjrNS9XpAaLiLid09mFTJiawvb9p/H09ODe/m247pIGVpclIiJuxuEwzFywlc9/3glAQqsoRt/RUcPsRUTEaRt2neDlaank5JdQM9SfscMTaBxb3eqyRETEzWTlFvHy9FQ27T6FpwcMv6EVfbs10pWQLkwNFhFxK7sPZfJiUgonMwsICvDhqaFxtGtSy+qyRETEzRQU2Xhj1hpWbUoH4OYeTRjcuwWeGmYvIiJO+n7lPt7/agN2h6FJbHXGJiYQFuJvdVkiIuJm9qdnM35yMscy8gnw8+aJwZ3p3CLS6rLkb6jBIiJu45cNR3jzkzSKiu3UqRXEsyMSiK4VZHVZIiLiZo5n5DN+SjL7jmbj7eXJqFvb06NzrNVliYiIm7HbHXw0bxPfrtgLQLcOdXjwtg74+XhZXJmIiLib1C3pvDpzDQVFNqJqnhlmXzdKw+zdgRosIuLyjDF8+tMOPl6wDYAOTWvxxJA4ggK0hYuIiDhn694MXpqaQmZuEdWD/XhmWDzN64dZXZaIiLiZ3PxiXpmxmnU7TgAwqHdzbr1Kw+xFRMQ5xhjmLNnN1PmbMQZaN6rJU0M0zN6dqMEiIi6tqMTOv2evZdm6wwDccHlDRtzQCi8NsxcRESctWn2Atz9bj83uoGF0KM8kxhNRQ8PsRUTEOYdP5DJ+8ioOn8jDz9eL0QM7cknbaKvLEhERN1Nis/PuFxv4KfUAANd0qce9/dtqmL2bUYNFRFzWqawCJiSlsPNgJl6eHtx3U1uu7Vrf6rJERMTN2B2GGd9t4cvFuwDo2qY2owd2xN9P/xUWERHnrNtxnJenryavoITw6gGMS0ygYZ1Qq8sSERE3k5lTxMRpKWzZm4GnB4zo25obLmuoKyHdkJ5ViohL2nnwNC9OSSEju5Dgar48PSyONo3CrS5LRETcTH5hCa99vIbULccAuK1nU+64prmG2YuIiNPm/7KXD+ZuxOEwNK9XgzHD46kRrGH2IiLinH1Hsxk/eRXHTxcQ6O/NE4Pj6Ng8wuqypJTUYBERl7N83WH+NXstxSV2YiODeXZEAlE1A60uS0RE3Ez6qTxenJLM/vQcfLw9eei2DlzRMcbqskRExM3Y7A4+mLuR73/dB8CVnWJ44Jb2+GqYvYiIOCl501Fen7WGgiI7tcMDGZeYQGxksNVlyUVQg0VEXIbDYZi9cDuf/LgdgM4tInl8UCeq+WuYvYiIOGfznlO8NDWF7LxiagT7MTYxgaZ1a1hdloiIuJmc/GJemZ7K+p0n8fCAIde1ZMCVjbWFi4iIOMUYw5eLdzH9uy0YA20bh/PU0DiCq/laXZpcJDVYRMQlFBbb+Nfstfyy/ggA/a5oxLDrW+GlLVxERMRJC5P38+6X67HZDY1iQhk7PIHw6gFWlyUiIm7m4LEcxk9J5ujJPPx9vXjszk4ktK5tdVkiIuJmikvs/OeL9SxafRCA3pfU555+bfD20jD7ykANFhGx3MnMAl5MSmb3oSy8vTz4x4B2XJ1Qz+qyRETEzdgdhqnfbmbu0t0AXNoumodv74C/r/7LKyIizknbdpxJM1LJK7QRUSOAsYkJNIjWMHsREXHO6ZxCXkpKYdv+03h6enBPvzb0ubSB1WVJGdKzTRGx1I4Dp5mQlExGdhEhgb6MGRZPq4Y1rS5LRETcTF5BCa/OXM2abccBGNirGbdf3UzD7EVExCnGGL5ZvofJ8zbhMNCifhhjhsVTPdjP6tJERMTN7DmcxfgpyZzMLCAwwIenhnSmfVMNs69s1GAREcssTTvEvz9dS7HNQb2oYMYmapi9iIg47+jJPMZPWcXBY7n4envy8MCOXN6+jtVliYiImymxOXh/zgZ+WLUfgKviYhl5czt8vDXMXkREnLNy4xFen5VGUbGdOrUCGTeiC3VqBVldlpQDNVhEpMI5HIZZP2zj0592ABDfMopH7+yoYfYiIuK0jbtOMnFaCjn5JYSF+DM2MZ4msRpmLyIizsnOK2bitBQ27T6FhwcMv74V/a5opGH2IiLiFGMMn/+8kxnfbwWgfdNaPDm4M0EaZl9pqcEiIhWqsMjGG5+ksXLjUQAGXNmYwde11DB7ERFx2g+r9vHelxuwOwxNYqvzzPB4aoZqmL2IiDjnQHo246ckk34qnwA/bx4f1Im4llFWlyUiIm6mqMTO25+uY+naQwBcf1kD7rqxNV4aZl+pqcEiIhXmxOkCXpySzJ4jWXh7efLALe24Kq6u1WWJiIibsdsdTP5mM98s3wNAt/Z1ePD2Dvj5aAsXERFxzuqtx5g0YzUFRTYiw6oxbkQC9aJCrC5LRETcTEZ2IROSktlxIBMvTw/uvaktvbvWt7osqQBqsIhIhdi2L4MJU1PIzCmiepAfY4bF06JBmNVliYiIm8ktKOHVGatJ235mmP2ga5tza8+m2sJFREScYozh62W7SfpmMw4DrRrW5OmhcYQGaZi9iIg4Z9ehTCZMSeZkViFBAT48PSyOto1rWV2WVBA1WESk3C1afZC3P1uHze6gfu0QxiUmEBFWzeqyRETEzRw5kcsLk5M5fCIXP18vHhnYkUvbRltdloiIuJkSm4P3vlzPwpQDAFwdX5f7B7TDx1tbuIiIiHN+2XCEN2alUVxiJyYiiHEjEogO1zD7qkQNFhEpNw6HYfp3W/hy8S4AurSOYvQdnQjw06lHREScs37HCV6enkpuQQnhof6MTUygUUx1q8sSERE3k5VbxMRpqWzecwpPD0i8sTU3Xt5QV0KKiIhTjDHMXriDWT9sA6Bj8wieGNSZwAAfiyuTiqZXOUWkXOQXlvDGrDSSN6cDcMtVTRh0bQs8NcxeREScNP+XvXwwdyMOh6FZ3Ro8MzyeGiH+VpclIiJuZv/RbF6YkszxjHyq+Xvz+KDOdG4RaXVZIiLiZgqLbbw1ey0r1h8B4MZuDUm8vpWG2VdRarCISJk7npHP+CnJ7DuajY+3Jw/e2p7unWKtLktERNyMze7gw7kb+e7XfQB07xjDqFvb46th9iIi4qSULem8NnM1BUV2atcMZNyIBGIjg60uS0RE3MyprAJeTEph18Ezw+zvH9COa7rUs7ossZAaLCJSprbsPcVLU1PIyi2mRrAfzwyPp1k9DbMXERHn5OYX8/L0VNbvPImHBwzu3YKbezTRFi4iIuIUYwxzluxi6vwtGANtG4fz5JA4QgJ9rS5NRETczI4Dp5mQlExGdhHB1Xx5elgcbRqFW12WWEwNFhEpMz+lHOA/X6zDZjc0rBPK2OEJ1KoRYHVZIiLiZg4dz2H85GSOnMzD39eLR+/sRJfWta0uS0RE3EyJzc47n69n0eqDAPTuWp97+rfBW1u4iIiIk5avPcy/ZqdRbHNQNyqYcYkJRNUMtLoscQFqsIjIRbM7DNPmb2HOkjPD7C9pW5tHbu+Iv4bZi4iIk9K2H2fS9FTyCm3UqhHAuMQEGkSHWl2WiIi4mcycIl6amsLWfRl4enpwd9/W9Lm0ga6EFBERpzgchlk/buPThTsA6NwikscHdaKav4bZyxl69VNELkp+YQmvfbyG1C3HALjt6qbc0au5htmLiIhTjDF8u2IvH83bhMNhaFE/jKeHxVEjWMPsRUTEOXuPZDF+SjInThcQ6O/Nk0Pi6NAswuqyRETEzRQW2Xhzdhq/bjgKQP/ujRnapyVees1L/ocaLCJSaumn8hg/JZkD6Tn4envy8O0dubxDHavLEhERN2OzO3h/zkYWrNwHQI/OsTxwSzt8vDXMXkREnLNy41HemLWGwmI70eFnhtnHRGiYvYiIOOdkZgHjpySz53AW3l4ejLy5PT3j61pdlrggNVhEpFQ27T7JS1NTyckvJizEj2eGJ9C0bg2ryxIRETeTnVfMy9NS2bj7zDD7YX1a0b97I23hIiIiTjHG8MWinUz/bisA7ZvU4skhnQmqpmH2IiLinO37M5iQlMLpnCJCg3wZMyyelg1qWl2WuCg1WETEaT8m7+e9L9djsxsax4QyNjGBmqEaZi8iIs45eOzMMPujp/II8PPisUGdiW8ZZXVZIiLiZopL7Lz92TqWpB0CoM+lDbirb2sNsxcREactSTvEvz9dS4nNQf3aIYxNTCAyrJrVZYkLU4NFRC6Y3WFI+mYzXy/bDcBl7aJ56PYO+PvqVCIiIs5ZvfUYr85cTX6hjYiwajybmEC92iFWlyUiIm7mdHYhE5JS2H7gNJ6eHtzbvw3XXdLA6rJERMTNOByGmQu28vnPOwFIaBXF6Ds6api9/C29KioiFySvoIRJM1eTtu04AHdc05zbr26qLVxERMQpxhi+XraHpG824TDQqmFNnh4aR2iQn9WliYiIm9l9KJMXpyRzMquQoAAfnhoaR7smtawuS0RE3ExBkY03P0lj5cYzw+xv7tGEwb1b4Klh9nIB1GARkb919GQe46es4uCxXHx9vHhkYAcua6dh9iIi4pwSm4P3vlzPwpQDAFwdX5f7B7TDx1tbuIiIiHN+2XCENz9Jo6jYTp1aQTw7IoHoWkFWlyUiIm7m+Ol8XpySzN4j2Xh7eTLq1vb06BxrdVniRlz+2Wz//v2pUaMGN998s1NfE5GysWHXCR59aykHj+VSM9SfV0ZepuaKuCXliYi1snKLGPf+ryxMOYCnB4y4sTWjbm2v5oq4JWWKiHWMMcxeuJ2Xp6VSVGynQ9NavPZQNzVXxC0pT0SstW1fBo/+axl7j2RTPciPif+4VM0VcZrLP6N98MEHmT59utNfE5GL9/3KfTz7/kpy8ktoWrc6bzx8BY1jq1tdlkipKE9ErLP/aDaj31rG5j2nqObvzbgRXeh3RSNtMyluS5kiYo2iEjuvzVzDxwu2AXDD5Q157q4uBAVof3xxT8oTEessWn2Qp9/9hczcIhpEh/D6w91oXj/M6rLEDbl8g+XKK68kODjY6a+JSOnZ7Q7en7OBd79Yj91huKJDDC/94zLCQvytLk2k1JQnItZI2ZLO428v43hGPlE1q/HqqMvp3CLS6rJELooyRaTincoq4On/rGDZusN4eXow8uZ23NOvDV5eLv+yhsh5KU9EKp7dYZj67Wbe/CQNm91B1za1eeWBy4moUc3q0sRNWf4/kYkTJxIXF0dwcDARERH069eP7du3W12WSJWVm1/M8x+t4tsVewEY3LsFj97ZET8fL4srE/lryhMR12KM4avFu3hxSjIFRXbaNArn9YeuoG5UiNWlifwtZYqIa9l58DSj/7WMnQczCa7mw/h7L+HarvWtLkvkbylPRFxLfmEJLyWl8OXiXQDc2rMpTw2JI8BPY8ql9CxvsCxdupSRI0eyatUqFi5ciM1mo1evXuTl5VldmkiVc/hELo/9exnrdpzAz9eLMcPiuLVnU23hIm5BeSLiOkpsdt76dC1J327GGLimSz3+eU9XQgJ9rS5N5IIoU0Rcx/J1h3nqP7+QkV1IbGQQrz90BW0ah1tdlsgFUZ6IuI5jGfk8+c4KUrak4+PtyaN3dmJw7xZ4euo1L7k4lrfnFixYcM7nSUlJREREsGbNGrp161YmP6OoqIiioqKzn2dnZ5fJcUUqk3U7jvPy9NXkFZQQXj2AcYkJNKwTanVZIhdMeSLiGjJzinhpagpb92WcGWbftzU3XNZQzXpxK8oUEes5HGeG2X/y45l3+3duEcnjgzpRzV/zVsR9KE9EXMPmPad4aWoK2XnF1Aj245nh8TSrp3krUjYsv4Ll97KysgAICyu7X/KJEycSGhp69iM2NrbMji1SGcz/ZS/PfbiKvIISmtWrwRsPdVNzRdye8kSk4u09ksWjby1l674MAv29ee7urtx4uYbZi/tTpohUrMJiG5Nmrj7bXOl3RSPGJiaouSJuT3kiUvF+StnP2P/+QnZeMY1iQnnj4SvUXJEy5WGMMVYX8RtjDH379uX06dMsX74cgGuuuYa0tDTy8vIICwtjzpw5xMXF/e3X/tefdfNjY2PJysoiJET7gEvVZbM7+GDuRr7/dR8AV3aK4YFb2uOreSvyF7KzswkNDXXpc6jyRKTirdp0lNc/XkNhsZ3a4YGMS0wgNlKDWeX83CFPQJkiUtFOZhbwYlIyuw9l4e3lwT8GtOPqhHpWlyUuzh0yRXkiUrF+G2Y/d+luAC5tG83Dt3fAX/NW5C+UJk9c6jfqgQceYMOGDaxYseLsbT/88MN57/9XX/tffn5++Pn5XXR9IpVJTn4xr0xPZf3Ok3h4wJDrWjLgysZ6l7FUCsoTkYpjjOGLRTuZ8f1WjIF2TcJ5ckgcwdU0b0UqB2WKSMXZvj+DCUkpnM4pIiTQlzHD4mnVsKbVZYmUCeWJSMXJLyzh1ZlrWL31GAADezXj9qubad6KlAuXabCMGjWKefPmsWzZMmJiYqwuR6RSO3gsh/FTkjl6Mg9/Xy8eu7MTCa1rW12WSJlQnohUnOISO+98vo7Faw4BcN0l9bm7Xxu8vVxuF1qRUlGmiFScJWmH+PenaymxOagXFczYxASiagZaXZZImVCeiFSc9FN5vDA5mYPHcvD19uThgR25vH0dq8uSSszyBosxhlGjRjFnzhyWLFlCgwYNrC5JpFJL236cSdNTySu0EVEjgLGJCTSI1rwVcX/KE5GKdTq7kAlTU9i+/zSenh7c068NfS7V351UDsoUkYrjcBg+/mEbn/20A4D4llE8emdHzVuRSkF5IlKxNu4+ycSpqeTkFxMW4s/YxHiaxNawuiyp5CxvsIwcOZJZs2bx9ddfExwcTHp6OgChoaEEBARYXJ1I5WGM4ZsVe5j89SYcBlrUD2PMsHiqB+tSYqkclCciFWfP4SzGT0nmZGYBQQE+PDUkjnZNa1ldlkiZUaaIVIyCIhtvfpLGyo1HARhwZWMGX9cSL23hIpWE8kSk4vywah/vfbkBu8PQJLY6zwyPp2ao/s6k/Fk+5P588x6SkpIYNmxYufxMdxh+JlKWbHYH//1qAz+s2g/AVXGxjLy5HT7eGmYvznPVc6jyRKRi/LrhCG98kkZRsZ06tYJ4dkQC0bWCrC5L3JArn0OVKSLl7/jpfF6ckszeI9l4e3nywC3tuCqurtVliZty1XOo8kSk/NntDqZ8s5l5y/cA0K19HR68vQN+PnrNS5znlkPuLe7viFR62XnFvDwtlY27zwyzH359K/pd0UjD7KXSUZ6IlC9jDJ/9tIOZC7YB0KFpLZ4YEkdQgLZwkcpHmSJSvrbty2DC1BQyc4qoHuTHmGHxtGgQZnVZImVOeSJSvvIKSpg0YzVp248DcOe1zbmtZ1O95iUVyvIGi4iUnwPp2Yyfkkz6qXwC/Lx5fFAn4lpGWV2WiIi4maISO//+dC3L1h4G4IbLGzLihlZ4aZi9iIg4adHqg7z92Tpsdgf1a4cwLjGBiLBqVpclIiJu5sjJXMZPTubQ8Vz8fL14ZGBHLm0bbXVZUgWpwSJSSa3eeoxJM1ZTUGQjMqwa40YkUC9KlweLiIhzTmUVMCEphZ0HM/Hy9OC+m9pybdf6VpclIiJuxuEwTP9uC18u3gVAl9ZRjL6jEwF+ellCREScs37nCV6elkpuQQnhof6MTUygUUx1q8uSKkr/kxGpZIwxfL1sN0nfbMZhoFXDmjw9NI7QIA2zFxER5+w6mMmLScmcyiokuJoPTw2No21jDbMXERHn5BeW8MasNJI3nxnwfctVTRh0bQs8NcxeRESc9N2ve3l/zkYcDkOzujUYMzyesBB/q8uSKkwNFpFKpMTm4L0v17Mw5QAAvRLqcd9NbfHx1hYuIiLinBXrD/PmJ2spLrETGxnE2MQEosM1zF5ERJxzPCOf8VOS2Xc0Gx9vTx68tT3dO8VaXZaIiLgZu93Bh19vYv4vewHo3jGGUbe2x1fD7MViarCIVBJZuUVMnJbK5j2n8PSAETe25obLG2qwl4iIOMUYw+wftzPrx+0AdGoeweODOhOoYfYiIuKkzXtOMXFaClm5xVQP9uOZ4fE0r6dh9iIi4pzc/GJemb6adTtPADDkuhbc3KOJXvMSl6AGi0glsP9oNi9MSeZ4Rj7V/L15YnBnOjWPtLosERFxM4XFNt6avZYV648A0LdbI4bf0AovbeEiIiJO+inlAP/5Yh02u6FhnVDGDk+gVo0Aq8sSERE3c+h4DuMnJ3PkZB7+vl6MvqMTXdvUtroskbPUYBFxcylb0nlt5moKiuzUrhnIuBEJxEYGW12WiIi4mVNZBbw4JZldh7Lw9vLg/gHt6JVQz+qyRETEzdgdhqnfbmbu0t0AXNK2No/c3hF/DbMXEREnrd1+nFdmrCavoIRaNQIYl5hAg+hQq8sSOYf+hyPipowxzFmyi6nzt2AMtG0czpND4ggJ9LW6NBERcTM7DpxmQlIyGdlFBFfzZcywOFo3Cre6LBERcTP5hSW8OnMNq7ceA+C2q5tyR6/mGmYvIiJOMcYw/5e9fPj1JhwOQ4v6YTw9LI4awRpmL65HDRYRN1Ris/PO5+tZtPogAL271uee/m3w9tIwexERcc6ytYd4a/Zaim0O6kYFMy4xgaiagVaXJSIibib9VB7jpyRzID0HX29PHr69I5d3qGN1WSIi4mZsdgcfzNnI9yv3AdCjcywP3NIOH28NsxfXpAaLiJvJzCnipakpbN2XgaenB3f3bU2fSxtosJeIiDjF4TDM+mEbn/60A4DOLSJ5fFAnqvlrmL2IiDhn4+6TTJyaSk5+MWEhfjwzPIGmdWtYXZaIiLiZ7LxiXpmeyoZdJ/HwgGF9WtK/e2O95iUuTQ0WETey90gW46ckc+J0AYH+3jw5JI4OzSKsLktERNxMYZGNN2en8euGowD0796YoX1aapi9iIg47YdV+3nvy/XYHYbGMaGMTUygZqiG2YuIiHMOHjszzP7oqTwC/Lx47M7OxLeKsroskb+lBouIm1i16Sivf7yGwmI70eFnhtnHRGiYvYiIOOfE6QJeTEpmz+Ezw+xH3tyenvF1rS5LRETcjN3uYMq3m5m3bA8Al7WL5qHbO+Dvq5cZRETEOWu2HWPSjNXkF9qICKvGuMQE6tcOsboskQui//mIuDhjDF8s2smM77diDLRvUosnh3QmqJqG2YuIiHO278/gxaQUMnOKCA3yZcyweFo2qGl1WSIi4mbyCkqYNHM1aduOA3DHNc25/eqm2sJFREScYoxh3vI9TJm3CYeBVg1r8vTQOEKD/KwuTeSCqcEi4sKKS+y8/dk6lqQdAqDPpQ24q29rDbMXERGnLVlzkH9/to4Sm4P6tUMYm5hAZFg1q8sSERE3c+RkLuMnJ3PoeC6+Pl48MrADl7XTMHsREXFOic3Bf7/awI/J+wG4Or4u9w9oh4+3XvMS96IGi4iLOp1dyISkFLYfOI2npwf39m/DdZc0sLosERFxMw6HYeaCrXz+804AElpFMfqOjhpmLyIiTtuw6wQvT0slJ7+EmqH+jB2eQOPY6laXJSIibiYrt4iJ01LZvOcUnh4w/IZW9O3WSFdCiltSg0XEBe0+lMmLU5I5mVVIUIAPTw2No12TWlaXJSIibqagyMYbs9awalM6ADf3aMLg3i3w1DB7ERFx0vcr9/H+VxuwOwxNYqszNjGBsBB/q8sSERE3sz89m/GTkzmWkU+AnzdPDO5M5xaRVpclUmpqsIi4mF82HOHNT9IoKrZTp1YQz45IILpWkNVliYiImzmekc/4KcnsO5qNj7cno25tz5WdYq0uS0RE3Izd7uCjeZv4dsVeALp1qMODt3XAz8fL4spERMTdpG5J59WZaygoshFV88ww+7pRGmYv7k0NFhEXYYzh05928PGCbQB0bBbB44M7ExSgLVxERMQ5W/dm8NLUFDJzi6ge7Mczw+NpXi/M6rJERMTN5OYX88qM1azbcQKAQb2bc+tVGmYvIiLOMcYwZ8lups7fjDHQulFNnhqiYfZSOajBIuICikrs/Hv2WpatOwzAjZc3JPGGVnhpmL2IiDjp59QDvPP5emx2Bw2jQ3kmMZ6IGhpmLyIizjl8Ipfxk1dx+EQefr5ePHpHR7q2iba6LBERcTMlNjvvfrGBn1IPAHBNl3rc27+thtlLpaEGi4jFTmUVMCEphZ0HM/Hy9OD+AW25pkt9q8sSERE3Y3cYps/fwldLdgHQtU1tRg/siL+f/rsnIv+vvfsMj7JM3z/+TU9IgxAChFBC7z0JWMCCWEBBURFF6q5lEXGxI6y7CwiLfXV3dVcSmogVRVBZlC6SBAi9904ghCSkz8z9f8FffouCMEDyzJM5P8fBC0IynHdI5mTmyjyXiHvWbs9k4rRV5BeWEl05hDFDkqhfK9LqWCIiYjOn8oqZMDWNzXtO4usDQ3u15M7r6uuVkFKh6BG3iIV2HMhmXHIaJ3OLCK8UyIuDEmjVINrqWCIiYjMFRaW89uFq0jcfA6Bvt8Y8eGtTLbMXERG3zVu+m39/tRGXy9C0bhVGDU6kSriW2YuIiHv2Hsll7OSVZGYXEhrsz3MDEmjfJMbqWCJXnQYsIhZZtvYQb83KoKTUSe3q4fxpaBI1qoZaHUtERGzmaFY+45JT2Xc0jwB/X0b0bUfX9nFWxxIREZtxOF38+8sNfLtiLwA3dojjifvaEqhl9iIi4qbUjUd4feZqCoud1IwOZcyQJGpXD7c6lkiZ0IBFpJy5XIZZC7bx0X+3AdCxWXWe7d+BSsFaZi8iIu7ZtDuLV6akkZtfQpXwIEYPSaJxnSpWxxIREZvJKyhh4tR01u88gY8PDLijOX1ubKhLuIiIiFuMMXy+aCfTvtmMMdC6YTQvDEwgvFKg1dFEyowGLCLlqKjEwVuzMvhx3WEAendtwKCeLfDTJVxERMRNC1L38c/P1+FwGhrERTJ6cBLRlUOsjiUiIjZz4FgeY5NTOXIin+BAP555qANJLWtaHUtERGympNTJu5+uZdHqgwDcfk09HundCn8/LbOXik0DFpFycuJUIeNSUtl1MAd/Px/+0KcNtyTVtTqWiIjYjNNlmDJ3E18u2QXAtW1ieeqBdgQH6r91IiLinjVbM5k0PZ38IgcxVUIYPSSJ+FgtsxcREfdk5xYxfkoa2/Zl4+vrwyO9W9Hj2nirY4mUCz0SFykH2/dnMz4llZO5xUSEBjJqUCIt6le1OpaIiNhMfmEpr85YxeqtmQD0696EB25pomX2IiLiFmMMXy/bzeQ5G3EZaFYvilGDEqkcHmR1NBERsZndh3IYm5zKiVOFhIYE8MKAjrRtrGX24j00YBEpY4vXHOTvH2dQ6nBRt0Y4o4domb2IiLjvyIl8xiav5MCx0wT6+/JUv/Zc37aW1bFERMRmSh0u3p+9nvkr9wFwc0Jtht3bhgB/LbMXERH3rFh/mDc+WkNxiZNa1UIZM7QTtaqFWR1LpFxpwCJSRlwuw4fzt/LJ99sBSGxeg6cfaq9l9iIi4rYNO08wYWoaeQWlREUEM3pIIo1qa5m9iIi4Jze/hAlT09i4KwsfHxjcswW9uzbQMnsREXGLMYZPftjOjG+3AtC2cTWef7gjYVpmL15IAxaRMlBY7ODNj9bw04YjAPS5sSEP39Fcy+xFRMRt81fu5V+fr8fpMjSqXZmXBidSNVLL7EVExD37j+YyNjmVo1kFhAT582z/DiQ0r2F1LBERsZniUid//ziDpRmHAOh5XTy/u6slflpmL15KAxaRqywzu4BxyansOZyLv58vT9zXhpsT6lgdS0REbMbpdDH56018vWw3AF3a1uLJB9oRFKBLuIiIiHtWbTnGpOmrKCx2UD2qEmOGJlG3RoTVsURExGZO5hYxPiWV7ftP4efrw6P3tOb2zvWsjiViKQ1YRK6irXtPMn5KGqfyiqkcFsSoQYk0i4+yOpaIiNjM6cJSXp2+ijXbziyzf+i2pvTt1liXcBEREbcYY/hyyS5S5m7CGGhRvyovDkwgMkzL7EVExD07D55iXHIqWTlFhIUE8OKgBFo3rGZ1LBHLacAicpUsXHWAdz5Zi8PpIj42gtGDk4iJqmR1LBERsZnDx0/z18mpHDp+mqBAP/7Yrz3Xto61OpaIiNhMqcPJPz9bz/fp+wG4JbEOj/dpQ4C/LuEiIiLu+XHdmWX2JaVOalcPY/SQJGKjtcxeBDRgEbliLpdh2jeb+XzRTgA6tazByAc7EBKkby8REXHPuh3HmTg1ndOFpURHBjN6SBIN4ipbHUtERGwm53Qxr0xJY/Oek/j6wNC7WnLn9fX1SkgREXGLMYZZC7Yzc/6ZZfYdmsbwbP+OhIYEWJxMxHPoGWCRK1BQVMobM9eQuukoAPfd3Ij+tzXDV8vsRUTETd+s2MP7szfgchma1KnCqMGJREUEWx1LRERsZu+RXMZOXklmdiGVgv157uGOdGha3epYIiJiM0UlDt6elcHydYcB6NWlAYPvbIGfnvMSOYcGLCKXKfNkAWOTU9l7JJcAf1+evL8tN3SobXUsERGxGafTxX++2si8H/cAcEP7OIbf35ZALbMXERE3pW06ymsfrqKw2EnNqqGMGZpE7erhVscSERGbycopZFxyKjsP5uDv58PjfdrQPamu1bFEPJIGLCKXYfOeLF6ZkkbO6RKqhAfx0uBEmtTVMnsREXHP6YIS/jZtFWt3HAdgwB3NuPemRrqEi4iIuMUYwxeLdjL1m80YA60bRvP8gAQiQgOtjiYiIjazfX8241NSOZlbTHilQEYNSqBlg2irY4l4LA1YRNz0fdp+/vHZWhxOQ/1akYwenES1KiFWxxIREZs5mJnH2MmpHD6RT3CgHyMf7EDnVjWtjiUiIjZT6nDy7qfrWLjqAAC3d67HI3e3wt9Py+xFRMQ9SzMO8vasDEocLurUCGfMkCRqVA21OpaIR/P4/3HdfffdVKlSxaBAHAAASDVJREFUhXvvvfdXfzZ37lyaNGlCo0aN+OCDDyxIJ97E6TIkf72Jtz/OwOE0XNO6Jn8bdp2GKyI2oT4RT7JmWybPvL2UwyfyqVYlhEnDr9dwRcQm1CfiSbLzihj1zx9ZuOoAvr4+PHp3Kx7v01rDFRGbUKeIp3C5DDO+3cKrM1ZT4nDRsVl1Xh1+vYYrIpfAxxhjrA7xWxYtWsTp06eZOnUqn3322dm3OxwOmjdvzqJFi4iIiKB9+/akpqYSFXXxyzTl5uYSGRlJTk4OERERZRlfKoiColJe+3A16ZuPAdD3lsY82L2pltmLV7Lrfaj6RDyBMYa5y/fwwZyNuFyGZvWieHFQAlXCtcxevI9d70PLok/Avp8Psc6ewzmMTU7leHYhocH+PD8ggXZNYqyOJWIJu96H6jGKeIKiYgdvzlrDivVHALj7hoYM7NFcy+zFK13OfajH/1jLjTfeSHj4r5fypaWl0aJFC2rVqkV4eDh33HEH8+fPtyChVHRHs/J59p1lpG8+RqC/L8/170j/25ppuCJiM+oTsZrD6eKfn6/n319uwOUy3NSxNuMfv0bDFRGbUZ+IJ/hpwxGee2cZx7MLiY0O5bURXTRcEbEhdYpY7Xh2Ic//Yzkr1h/B38+HEX3bMeTOFhquiLjB8gHLhAkTSEhIIDw8nJiYGHr37s22bdsu+nGHDx+mVq1aZ38fFxfHoUOHyjKqeKENu04w8q2l7D+aR1REEBOGXcf17Wpd/ANFpNypT8ST5eaX8Kf3f+K7n/bi4wODezbnqQfaEeDvZ3U0EfkF9Yl4MmMMn/6wnVempFFU4qRto2q8PqILcTG/foJWRKynThFPtm3fSUa+vYTdh3KIDAtk/OPX0i2xjtWxRGzH8gHLkiVLGDZsGCtXrmTBggU4HA66d+9Ofn7+b37c+a5s5uOj6apcPf9N3ceY91aQV1BCw7hI3niqK43rVLE6lohcgPpEPNWBY3k88/ZSNuw6QUiQH6MHJ3HPjY30dSbiodQn4qlKSp28MXMN077ZAkCPa+N5+fedCKsUaHEyEbkQdYp4qsWrD/DiP3/kVF4x9WpG8PqIrjSPr2p1LBFb8rc6wHfffXfO71NSUoiJiWH16tV06dLlgh9Xq1atc6b3Bw8eJCkp6bzvW1xcTHFx8dnf5+bmXmFqqcicThfJczcxZ+luAK5rE8uIB9oRHGj5t4uI/Ab1iXii1VuPMWn6KgqKHMREVWLMkCTq1dS1sEU8WXn0CahTxD3ZuUWMT0lj2/7ss8vs77gm3upYInIReowinsblMsz4bguf/rADgKQWNRj5YHsqBQdYnEzEvix/Bcsv5eTkAFx0cVdiYiIbN27k0KFD5OXl8c0333Drrbee930nTJhAZGTk2V+1a9e+6rmlYsgvLOWvyalnhysP3tqU5x7uqOGKiA2pT8RKxhi+WrqLv36wkoIiBy3qV+WNEV00XBGxobLoE1CnyKXbefAUI99awrb92YSFBPDXRzpruCJiU3qMIlYqLHYwYWra2eHKvTc1YtSgRA1XRK6Qjznf6w4tYoyhV69eZGdns2zZMgBuvfVW1qxZQ35+PlFRUcyePZuEhAQA5syZwzPPPIPL5eK5557jkUceOe/tnm+aX7t2bXJycoiI0BMdcsbhE6cZOzmVg5mnCQzw44/92nFdG+1bEfml3NxcIiMjPfo+VH0iVip1uHjvi/X8N3UfALck1uHxPm0I8Pe4n2sRsZQ39wmoU+TS/Lj+MG/MXENJqZNa1cL409AkYquFWR1LxON4c6eoT+RSZJ4sYGxyKnuP5OLv58vw+9tyU0cN40R+6XL6xKMGLMOGDWPevHksX76cuLi4Mvt77FC8Ur7W7zzOxKnp5BWUUjUymNGDk2hYu7LVsUQ8kh3uQ9UnYpWc08VMmJrOpt1Z+PrA4Dtb0KtLA10zW+Q87HAfWl59Avb4fEj5McYwa8F2Zs7fCkC7xtV4bkACYSH6KWOR87HDfageo4hVtuw5yStT0jh1upjK4UG8NDiRpnV/+1VUIt7qcu5DPea6R8OHD2fOnDksXbq0zB+8iPyvb3/ay/tfrMfpMjSuU5mXBicRFRFsdSwRuUzqE7HKvqO5jJ2cyrGTBYQE+fPcwx3p2Ky61bFE5DKpT8QqxaVO3p6VwbK1Z/Yv3HV9fYbc2QI/P70SUsSu1ClilR/S9/Pup+twOF3Uj43kpSGJxFSpZHUskQrF8gGLMYbhw4cze/ZsFi9eTHy8riUr5cPpdPHBnI3MXb4HgK7t4hjety1BAX4WJxORy6E+ESulbz7KqzNWU1jsoEbVM8vs69TQTwyK2JH6RKyUlVPIuJQ0dh44hZ+vD4/3ac2tnepZHUtELpM6RazidBmmzdvMF4t3AtC5VU1G9mtPcJDlTwWLVDiWf1cNGzaMmTNn8tVXXxEeHs7Ro0cBiIyMJCQkxOJ0UlGdLijhb9NXsXb7cQAevr0Z993cSJdwEbEx9YlYwRjD7MW7mDJvE8ZAywZVeWFAApFhQVZHE5HLpD4Rq+w4kM245DRO5hYRXimQFwcl0KpBtNWxROQKqFPECgVFpbz24WrSNx8DoG+3xjx4a1N8ffWcl0hZsHwHy4We0E5JSWHQoEFl8nfqepTe7dDx04ydvJJDx/MJCvTj6Qfb07lVrNWxRGzDU+9D1SdS3kodTv752Xq+T98PwK2d6vLo3a21zF7kEnnqfagVfQKe+/mQ8rEs4xBvzVpDicNF7erh/GloEjWqhlodS8Q2PPU+VI9RpLwdzcpnXHIq+47mEeDvy4i+7ejaXpelE7lUttzBYvF8R7zM2u2ZTJy2ivzCUqIrhzBmSBL1a0VaHUtErgL1iZSnU3nFvDIljS17T+LrA0N7teTO6+rrlZAiFYD6RMqTy2X46L/bmLVgGwAdm1Xn2f4dqBSsZfYiFYE6RcrTpt1ZvDIljdz8EqqEBzF6SBKN61SxOpZIhWf5gEWkvMxbvpt/f7URl8vQtG4VRg1OpEq4ltmLiIh79hzOYVxyKpnZhYQG+/Pcwwm0bxpjdSwREbGZohIHb32UwY/rDwPQu2sDBvVsgZ8u4SIiIm5akLqPf36+DofT0CAuktGDk4iurMvQiZQHDVikwnM4Xfz7yw18u2IvADd2iOOJ+9oSqGX2IiLippUbj/D6h6spKnFSMzqUMUOSqF093OpYIiJiMydOFTIuJZVdB3Pw9/PhD33acEtSXatjiYiIzThdhilzN/Hlkl0AXNsmlqceaEdwoJ7yFSkv+m6TCi2voISJU9NZv/MEPj4w4I7m9LmxoS7hIiIibjHG8NnCHUz/dgvGQOuG0bwwMIHwSoFWRxMREZvZtu8k41PSyM4rJiI0kFGDEmlRv6rVsURExGbyC0t5dcYqVm/NBKBf9yY8cEsTLbMXKWcasEiFdeBYHmOTUzlyIp/gQD+eeagDSS1rWh1LRERspqTUybufrmXR6oMA3H5NPR7p3Qp/Py2zFxER9yxec5C/f5xBqcNF3RrhjB6iZfYiIuK+IyfyGZu8kgPHThMY4MdTD7Tj+ra1rI4l4pU0YJEKac3WTCZNTye/yEFMlRBGD0kiPlbL7EVExD3ZuUWMn5LGtn3Z+Pr68EjvVvS4Nt7qWCIiYjMul+HD+Vv55PvtACQ2r8HTD7XXMnsREXHbhp0nmDA1jbyCUqIighkzJImGtStbHUvEa2nAIhWKMYavl+1m8pyNuAw0qxfFqEGJVA4PsjqaiIjYzO5DOYxNTuXEqUJCQwJ4YUBH2jbWMnsREXFPYbGDNz9aw08bjgDQ58aGPHxHcy2zFxERt81fuZd/fb4ep8vQqHZlXhqcSNVILbMXsZIGLFJhlDpcvD97PfNX7gOgW0Id/nBvawL8tcxeRETcs2L9Yd74aA3FJU5qVQtlzNBO1KoWZnUsERGxmczsAsYlp7LncC7+fr48cV8bbk6oY3UsERGxGafTxeSvN/H1st0AdGlbiycfaEdQgJ7zErGaBixSIeTmlzBhahobd2Xh4wODe7agd9cGWmYvIiJuMcbwyQ/bmfHtVgDaNq7G8w93JEzL7EVExE1b955k/JQ0TuUVUzksiFGDEmkWH2V1LBERsZnThaW8On0Va7adWWbf/7am3N+tsZ7zEvEQGrCI7e0/msvY5FSOZhUQEuTPcw93pGOz6lbHEhERmykudfL3jzNYmnEIgJ7XxfO7u1rip2X2IiLipoWrDvDOJ2txOF3Ex0YwenASMVGVrI4lIiI2c/j4af46OZVDx08TFOjHH/u159rWsVbHEpH/oQGL2NqqLceYNH0VhcUOqkdVYszQJOrWiLA6loiI2MzJ3CLGp6Syff8p/Hx9ePSe1tzeuZ7VsURExGacLsP0bzbz+aKdAHRqWYORD3YgJEgPvUVExD3rth9n4rR0TheWEh0ZzOghSTSIq2x1LBH5Bf0vT2zJGMOXS3aRMncTxkCL+lV5cWACkWFaZi8iIu7ZefAU45JTycopIiwkgBcHJdC6YTWrY4mIiM0UFJXy+odrSNt8FID7bm5E/9ua4atl9iIi4qZ5P+7h319uwOUyNKlThZcGJ1IlItjqWCJyHhqwiO2UOpz887P1fJ++H4DuSXV57J7WBPjrEi4iIuKeH9edWWZfUuqkdvUwRg9JIjZay+xFRMQ9x06eWWa/90guAf6+PHl/W27oUNvqWCIiYjMOp4v/fLmBb1bsBeCG9nEMv78tgVpmL+KxNGARW8k5XcwrU9LYvOckvj4w9K6W3Hl9fS32EhERtxhjmLVgOzPnn1lm36FpDM/270hoSIDFyURExG427c5iwtQ0ck6XUCU8iJcGJ9KkrpbZi4iIe04XlDBxWjrrdpzAxwcevr0Z997USM95iXg4DVjENvYeyWXs5JVkZhdSKfjMMvsOTbXMXkRE3FNU4uDtWRksX3cYgF5dGjD4zhb46RIuIiLipu/T9vOPz9bicBrq14pk9OAkqlUJsTqWiIjYzMHMPMZOTuXwiXyCA/14+qEOdGpZ0+pYInIJNGARW0jbdJTXPlxFYbGTmlVDGTM0idrVw62OJSIiNpOVU8i45FR2HszB38+Hx/u0oXtSXatjiYiIzThdhilzN/Hlkl0AXNO6Jn98oD3BWmYvIiJuWrMtk0nT0skvclCtSghjhiQRHxtpdSwRuUT63594NGMMXyzaydRvNmMMtG4YzfMDEogIDbQ6moiI2Mz2/dmMT0nlZG4x4ZUCGTUogZYNoq2OJSIiNlNQVMqrM1azassxAPre0pgHuzfVMnsREXGLMYa5y/fwwZyNuFyGZvWiGDUokcrhQVZHExE3aMAiHqvU4eTdT9excNUBAG7vXI9H7m6Fv5+W2YuIiHuWZhzk7VkZlDhc1KkRzpghSdSoGmp1LBERsZmjWfmMTU5l/9E8Av19eeqB9lzfrpbVsURExGYcThfvz97Adz/tBeCmjrV54r42BPhrmb2I3WjAIh4pO6+IV1LS2LovG19fHx7p1ZI7ro3XYi8REXGLy2WYOX8rH3+/HYCOzarzbP8OVArWMnsREXHPhl0nmDAlnbyCEqIignhpcBKN61SxOpaIiNhMbn4JE6ems2HXmWX2g3q04O4bGug5LxGb0oBFPM6ewzmMTU7leHYhoSEBPP9wR9o1ibE6loiI2ExRsYM3Z61hxfojANx9Q0MG9miuZfYiIuK2+Sv38a/P1+F0GRrGRTJ6SBJVI7XMXkRE3HPg2Jll9key8gkJ8uOZ/h1JbF7D6lgicgU0YBGP8tOGI7wxczVFJU5qVQtlzNBO1KoWZnUsERGxmePZhYxLSWX3oTPL7Ifd25ZuiXWsjiUiIjbjdLpInruJOUt3A3Bdm1hGPNCO4EA9lBYREfes3nqMSdNXUVDkICaqEn8akkTdmhFWxxKRK6T/FYpHMMbw6Q87mP7tFgDaNqrG8wM6ElZJy+xFRMQ92/adZFxKGqfyiokMC2TUoESax1e1OpaIiNhMfmEpk2asYs3WTAAeuq0pfbs11iVcRETELcYY5izbTfKcjbgMtKhflRcHJhAZpmX2IhWBBixiueJSJ+98vJYlGQcB6HltPL/r1RI/LbMXERE3LV59gL9/spZSh4t6NSMYPSSJ6lGVrI4lIiI2c/jEacZOTuVg5mkCA/wY2a8917aJtTqWiIjYTKnDxXtfrOe/qfsAuCWxDo/3aUOAv57zEqkoNGARS53MLWJ8Sirb95/C19eHx+5uxe3XxFsdS0REbMblMsz4bguf/rADgKQWNRj5YHstsxcREbet33mciVPTySsopWpkMKOHJNEwrrLVsURExGZyThczYWo6m3Zn4esDg+9sSa8u9fVKSJEKRgMWsczOg6cYn5zKiZwiwkICeGFgAm0aVbM6loiI2ExhsYM3Zq5m5cajANx7UyMevr0ZvlpmLyIibvr2p728/8V6nC5D4zqVeWlwElERwVbHEhERm9l3JJexyakcO1lApWB/nu3fkY7NqlsdS0TKgAYsYokf1x/mjZlrKCl1EhcTxpihScRGa5m9iIi4J/NkAWOTU9l7JJcAf1+G39+WGzvUtjqWiIjYjNPp4oM5G5m7fA8AXdvFMbxvW4IC/CxOJiIidpO2+SivzVhFYbGTGlUrMWZIEnVqaJm9SEWlAYuUK2MMsxZsZ+b8rQC0bxLDsw93JCxEl3ARERH3bNlzklempHHqdDGVw4N4aXAiTetGWR1LRERs5nRBCX+btoq1O44D8PDtzbjv5ka6hIuIiLjFGMPsxbuYMm8TxkCrBtG8MDCBiNBAq6OJSBnSgEXKTXGpk7dnZbBs7SEA7rq+PkPubKFl9iIi4rYf0vfz7qfrcDhd1I+N5KUhicRU0TJ7ERFxz6Hjpxk7eSWHjucTFOjH0w+2p3MrLbMXERH3lDqc/OOzdfyQfgCA2zrX49G7W+Gv57xEKjwNWKRcZOUUMi4ljZ0HTuHn68PjfVpza6d6VscSERGbcboM0+Zt5ovFOwHo3KomI/u1JzhI/6URERH3rN2eycRpq8gvLCW6cghjhiRRv1ak1bFERMRmTuUV88qUNLbsPYmvD/yuVyt6XhevV0KKeAk9GyFlbseBbMYlp3Eyt4jwSoG8OCiBVg2irY4lIiI2U1BUymsfriZ98zEA+nZrzIO3NtUyexERcdu85bv591cbcbkMTetWYdTgRKqEa5m9iIi4Z8/hHMYlp5KZXUhosD/PDUigfZMYq2OJSDnSgEXK1LKMQ7w1aw0lDhe1q4fzp6FJ1KgaanUsERGxmaNZ+YxLTmXf0TwC/H0Z0bcdXdvHWR1LRERsxuF08e8vN/Dtir0A3Nghjifua0ugltmLiIibVm48wusfrqaoxElsdChjhiYRFxNudSwRKWcasEiZcLkMH/13G7MWbAOgY7PqPNu/A5WCtcxeRETcs2l3Fq9MSSM3v4Qq4UGMHpJE4zpVrI4lIiI2k1dQwsSp6azfeQIfHxhwR3P63NhQl3ARERG3GGP4bOEOpn+7BWOgTaNonh+QQHglLbMX8UYasMhVV1Ts4K1ZGfy4/jAAvbs2YFDPFvjpEi4iIuKmBan7+Ofn63A4DQ3iIhk9OInoyiFWxxIREZs5cCyPscmpHDmRT3CgH8881IGkljWtjiUiIjZTUurk3U/Xsmj1QQDuuKYev++tZfYi3kwDFrmqTpwqZGxyKrsP5eDv58Owe9vQLbGu1bFERMRmnC7DlLmb+HLJLgCubRPLUw+0IzhQ/3URERH3rNmayaTp6eQXOYipEsLoIUnEx2qZvYiIuCc7t4jxU9LYti8bX18fHundih7XxlsdS0Qspmcp5KrZtu8k41PSyM4rJiI0kFGDEmlRv6rVsURExGbyC0t5dcYqVm/NBKBf9yY8cEsTLbMXERG3GGP4etluJs/ZiMtA8/goRg1KJDIsyOpoIiJiM7sP5TA2OZUTpwoJCwnghQEJtGlczepYIuIBNGCRq2LxmoP8/eMMSh0u6tYIZ8zQTlSPqmR1LBERsZkjJ/IZm7ySA8dOExjgx1MPtOP6trWsjiUiIjZT6nDx/uz1zF+5D4BuCXX4w72tCfDXMnsREXHPivWHeeOjNRSXOKlVLYw/DU0itlqY1bFExENowCJXxOUyzPhuC5/+sAOAxOY1ePqh9lpmLyIibtuw8wQTpqaRV1BKVEQwY4Yk0bB2ZatjiYiIzeTmlzBhahobd2Xh4wODe7agd9cGWmYvIiJuMcbwyQ/bmfHtVgDaNa7GcwMSCAvRc14i8n9su4Hptddeo0WLFrRs2ZIZM2ZYHccrFRY7mDgt/exwpc+NDRk1OFHDFRGxFfWJZ5i/ci9j3l9BXkEpjWpX5o2numi4IiK2oj7xDPuP5vL020vYuCuLkCB//jS0E3ff0FDDFRGxFXWK9YpLnbz24eqzw5U7r6/Py7/rpOGKiPyKLV/BsmHDBmbOnMnq1asBuPnmm+nZsyeVK1e2NpgXycwuYFxyKnsO5+Lv58vw+9twU8c6VscSEXGL+sR6TqeLyV9v4utluwHo0rYWTz7QjqAAXcJFROxDfeIZVm05xqTpqygsdlA9qhJjhiZRt0aE1bFERNyiTrFeVk4h41PS2HHgFH6+Pjx2T2tu61zP6lgi4qFs+QqWLVu2cM011xAcHExwcDBt27blu+++szqW19i69yRPv7WUPYdzqRwWxCuPX6vhiojYkvrEWqcLS/nLByvPDlf639aUZ/p30HBFRGxHfWItYwxfLtnJ2MkrKSx20KJ+VV4f0UXDFRGxJXWKtXYeOMXTby9lx4FThFcKYOyj12i4IiK/yWMHLBMmTCAhIYHw8HBiYmLo3bs327ZtA6Bly5YsWrSIU6dOcerUKRYuXMihQ4csTuwdFq46wIv//JFTp4uJj43g9ae60Cw+yupYIiIXpD7xTIePn+aZt5eSsf04QYF+vDAwgb63NNElXETEY6lPPFOpw8U7n6xl8pxNuAx0T6rL2EevITIsyOpoIiIXpE7xTMvXHeL5fywnK6eI2tXDeH1EV1o1jLY6loh4OI+9RNiSJUsYNmwYCQkJOBwOXnrpJbp3787mzZtp3rw5Tz75JDfddBORkZEkJCTg7++xR6kQnC7D9G828/minQB0blWTP/ZrT0iQPu8i4tnUJ55n3fbjTJyWzunCUqIjgxk9JIkGcZWtjiUi8pvUJ54n53QxE6ams2l3Fr4+MPSultx5fX0N60XE46lTPIsxhln/3cbM/54ZcnVoGsOz/TsSqn0rInIJfIwxxuoQl+L48ePExMSwZMkSunTpcs6f/e53v+Puu++mR48el3Rbubm5REZGkpOTQ0SEXjZ+MQVFpbwxcw2pm44CcH+3xjx0a1N8ffXARcQb2f0+VH1irXk/7uHfX27A5TI0qVOFlwYnUiUi2OpYImIBu9+HXs0+Aft/PsrbviO5/DU5lcyTBVQK9ue5hzvSoWl1q2OJiEXsfh+qxyjWKSpx8PasDJavOwxA764NGNSzBX56zkvEK13OfahtRuA5OTkAREWduRxVZmYmMTExbNu2jbS0NN57770LfmxxcTHFxcVnf5+bm1u2YSuQYycLGDt5JfuO5hHg78uTfdtxQ/s4q2OJiFw29Yk1HE4X//lyA9+s2AvADe3jGH5/WwK1b0VEbOpK+gTUKVcibdNRXvtwFYXFTmpWDWXM0CRqVw+3OpaIyGXTYxRrZOUUMi45lZ0Hc/D38+HxPm3onlTX6lgiYjO2GLAYYxg5ciTXXXcdLVu2BKB3796cOnWK0NBQUlJSfvPlkhMmTOAvf/lLecWtMDbtzuKVKWnk5pdQJTyIlwYn0qSu9q2IiH2pT6xxuqCEv01bxdodxwEYcEcz7r2pkS7hIiK2daV9AuqUy2GMYfbinUyZtxljoHXDaJ4fkEBEaKDV0URELpseo1hj+/5sxqekcjK3mIjQQF4cmEDLBtq3IiLus8UlwoYNG8a8efNYvnw5cXHuv3rifNP82rVr6+WSv+H7tH3847N1OJyGBnGRjB6cRHTlEKtjiYgHsPNLztUn5e9gZh5jJ6dy+EQ+wYF+PP1QBzq1rGl1LBHxAN7cJ6BOcVepw8m7n65j4aoDANzeuR6P3N0Kfz9fi5OJiCfw5k5Rn7hvacZB3p6VQYnDRZ0a4YwZkkSNqqFWxxIRD1AhLxE2fPhw5syZw9KlSy/7wUtQUBBBQUFXOVnF5HQZpszdxJdLdgFwbetYnnqgHcFaZi8iNqc+KX9rtmUyaVo6+UUOqlUJYcyQJOJjI62OJSJyRa5Gn4A6xR3ZeUW8kpLG1n3Z+Pr68EivltxxbbxeCSkitqfHKOXL5TLMnL+Vj7/fDkBC8+o881AHKgVrmb2IXD6PfdbcGMPw4cOZPXs2ixcvJj4+3upIFV5BUSmvzljNqi3HAHjglib0695Ey+xFxNbUJ+XPGMPc5Xv4YM5GXC5Ds3pRjBqUSOVwPfATEftSn1hjz+Ecxiancjy7kNCQAJ5/uCPtmsRYHUtE5IqoU8pfUbGDN2etYcX6IwDcc0NDBvRormX2InLFPHbAMmzYMGbOnMlXX31FeHg4R48eBSAyMpKQEF2q6mo7mpXPXyencuBYHoH+vjz1QHuub1fL6lgiIldMfVK+HE4X78/ewHc/7QXgpo61eeK+NgT4a5m9iNib+qT8/bThCG/MXE1RiZNa1UIZM7QTtaqFWR1LROSKqVPK1/HsQsalpLL7UA7+fr48cV8bbk6oY3UsEakgPHYHy4Ve7p2SksKgQYOu6LbtfG3OsrBh1wkmTEknr6CEqIhgRg9JpFHtKlbHEhEPZbf7UPVJ+cnNL2Hi1HQ27DqBjw8M6tGCu29ooEu4iMh52e0+tCz7BOz3+ShLxhg+W7iDad9sAaBto2o8P6AjYZW0zF5Ezs9u96F6jFJ+tu07ybiUNE7lFVM5LIhRgxJpFh9ldSwR8VAVageLh859Kpz5K/fyr8/X43QZGtauzOjBiVSN1E9LiEjFoT4pHweOnVlmfyQrn5AgP57p35HE5jWsjiUictWoT8pHSamTdz5Zy+I1BwHoeW08v+vVEj8tsxeRCkSdUj4Wrz7A3z9ZS6nDRb2aEYwZkkRMVCWrY4lIBeOxAxYpW06ni+SvNzFn2W4Arm9biyf7tiU4UF8SIiLinlVbjvHqjFUUFDmIiarEn4YkUbemflpORETck51bxPiUNLbtP7PM/rG7W3H7NdpLICIi7nG5DDO+28KnP+wAoFPLGox8sAMhQXrOS0SuPt2zeKH8wlImTV/Fmm2ZADx0W1P6dmusS7iIiIhbjDF8tXQ3KV9vxGWgRf2qvDgwgcgwLbMXERH37Dp4inHJqZzIKSIsJIAXBibQplE1q2OJiIjNFBY7eP3D1aRuOrPX5r6bG9H/tmb4apm9iJQRDVi8zOETpxk7OZWDmacJDPBjZL/2XNsm1upYIiJiM6UOF//6fB0L0vYDcEtiHR7v04YAf13CRURE3PPj+sO8+dEaikucxMWEMWZoErHRWmYvIiLuyTxZwNjkVPYeySXA35cn72/LDR1qWx1LRCo4DVi8yLodx5k4NZ3ThaVERwYzekgSDeIqWx1LRERsJud0MROmprNpdxa+PjD4zpb06lJfr4QUERG3GGP4+PvtfPjdVgDaN4nh2Yc7EhYSYHEyERGxm817snhlSho5p0uoHB7ES4MTaVpXy+xFpOxpwOIlvl2xh/dnb8DpMjSpU4VRgxOJigi2OpaIiNjMviO5/DU5lcyTBVQK9ufZ/h3p2Ky61bFERMRmikudvD0rg2VrDwFw1/X1GXJnCy2zFxERt/2Qvp93P12Hw+mifmwko4ckUa1KiNWxRMRLaMBSwTmdLj74aiNzf9wDwA3t4xh+f1sCA/wsTiYiInaTtvkor81YRWGxkxpVKzFmSBJ1amiZvYiIuCcrp5BxKWnsPHAKP18fHu/Tmls71bM6loiI2IzTZZg2bzNfLN4JQOdWNRnZrz3BWmYvIuVI9zgV2OmCEv42bRVrdxwHYMAdzbj3pka6hIuIiLjFGMPsxbuYMm8TxkCrBtG8MDCBiNBAq6OJiIjN7DiQzbjkNE7mFhFeKZAXByXQqkG01bFERMRmCopKee3D1aRvPgZA31sa82D3plpmLyLlTgOWCupgZh5jJ6dy+EQ+wYF+jHywA51b1bQ6loiI2Eypw8k/PlvHD+kHALi1U10evbu1ltmLiIjblmUc4q1ZayhxuKhdPZw/DU2iRtVQq2OJiIjNHM3KZ1xyKvuO5hHo78uIB9rRpV2c1bFExEtpwFIBZWzL5G/TV5FfWEp05RD+NDSJ+NhIq2OJiIjNnMor5pUpaWzZexJfH/hdr1b0vC5er4QUERG3uFyGj/67jVkLtgHQsVl1nu3fgUrBWmYvIiLu2bT7zDL73PwSoiKCeGlwEo3rVLE6loh4MQ1YKhBjDPN+3MN/vtqIy2VoWvfMMvsq4VpmLyIi7tlzOIdxyalkZhcSGuzPcwMSaN8kxupYIiJiM0UlDt76KIMf1x8GoHfXBgzq2QI/XcJFRETctCB1H//8fB0Op6Fh3Jll9lUjtcxeRKylAUsF4XC6+PfsDXz7014AbupYmyfua0OAv5bZi4iIe1ZuPMLrH66mqMRJbHQoY4YmERcTbnUsERGxmROnChmXksqugzn4+/kw7N42dEusa3UsERGxGafLMGXuJr5csguAa9vE8tQD7QgO1NOaImI93RNVALn5JfxtWjrrd57AxwcG3tGce25sqEu4iIiIW4wxfLZwB9O/3YIx0KZRNM8PSCC8kpbZi4iIe7btO8n4lDSy84qJCA1k1KBEWtSvanUsERGxmfzCUl6dsYrVWzMBeLB7Ex7o3kTPeYmIx9CAxeYOHDuzzP5IVj4hQX48/WAHklpqmb2IiLinpNTJu5+uZdHqgwDccU09ft+7Ff5+WmYvIiLuWbzmIH//OINSh4u6NcIZM7QT1aMqWR1LRERs5siJfMYmr+TAsdMEBvjxx37tuK5NLatjiYicQwMWG1u99RiTpq+ioMhBTFQlxgxJol7NCKtjiYiIzWTnFjF+Shrb9mXj6+vDI71b0ePaeKtjiYiIzbhchg/nb+WT77cDkNi8Bk8/1F7L7EVExG0bdp5gwtQ08gpKqRoZzOjBSTSsXdnqWCIiv6IBiw0ZY/h62W4mz9mIy0Dz+ChGDUokMizI6mgiImIzuw6eYlxKGidOFRIWEsALAxJo07ia1bFERMRmCosdvPnRGn7acASAPjc25OE7mmuZvYiIuO27n/by3hfrcboMjWpXZvSQJKIigq2OJSJyXhqw2Eypw8V7X6znv6n7AOiWUIc/3Ntay+xFRMRtK9Yf5o2P1lBc4qRWtTD+NDSJ2GphVscSERGbycwuYFxyKnsO5+Lv58vw+9twU8c6VscSERGbcTpdfDBnI3OX7wGgS7taPNm3HUEBes5LRDyXBiw2knO6mInT0tm4KwtfHxh8Zwt6dWmgxV4iIuIWYwyffL+dGd9tBaBd42o8NyCBsBBdwkVERNyzde+ZZfanThdTOSyIUYMSaRYfZXUsERGxmdOFpUyalk7G9uMA9L+9Kfff3FjPeYmIx9OAxSb2Hc1lXHIqR7MKCAny57mHO9KxWXWrY4mIiM0Ulzr5+8cZLM04BMCd19dn6J0t8NMyexERcdPCVQd455O1OJwu4mMjGD04iRgtsxcRETcdPn6av05O5dDx0wQF+jGyX3uuaR1rdSwRkUuiAYsNpG8+yqszVlNY7KBG1TPL7OvU0DJ7ERFxT1ZOIeNT0thx4BR+vj48dk9rbutcz+pYIiJiM06XYfo3m/l80U4AOrWswcgHOxASpIeXIiLinnXbjzNxWjqnC0uJrhzCmCFJ1K8VaXUsEZFLpv8BezBjDF8u2UXK3E0YAy0bVOWFAQlaZi8iIm7beeAU41JSycopIrxSAC8OTKRVw2irY4mIiM0UFJXy+odrSNt8FID7uzXmoVub4qtl9iIi4qZ5P+7h319uwOUyNKlbhZcGJVJFy+xFxGY0YPFQpQ4n//xsPd+n7wfg1k51efTu1gT46xIuIiLinuXrDvHmRxmUlDqpXT2MMUM6UTM61OpYIiJiM8dOnllmv/dILgH+vjzZtx03tI+zOpaIiNiMw+niP19u4JsVewG4oUMcw+9rS6CW2YuIDWnA4oFO5RUzYWoam/ecxNcHhvZqyZ3X1ddiLxERcYsxhln/3cbM/24DoEPTGJ7t35FQLbMXERE3bdqdxYSpaeScLqFKeBAvDU6kSV0tsxcREfecLihh4rR01u04gY8PPHx7M+69qZGe8xIR29KAxcPsPZLL2MkrycwuJDTYn+ceTqB90xirY4mIiM0UlTh4e1YGy9cdBqB31wYM6tkCP13CRURE3PR92n7+8dlaHE5Dg7hIRg9OIrpyiNWxRETEZg5m5jF2ciqHT+QTHOjH0w91oFPLmlbHEhG5IhqweJDUjUd4feZqCoud1IwOZcyQJGpXD7c6loiI2ExWTiHjklPZeTAHfz8fHu/Thu5Jda2OJSIiNuN0GabM3cSXS3YBcG3rWJ56oB3BWmYvIiJuWrMtk0nT0skvclCtypll9vGxWmYvIvan/xl7AGMMny/aybRvNmMMtG4YzQsDEwivFGh1NBERsZnt+7MZn5LKydxiIkIDeXFgAi0baJm9iIi4p6ColFdnrGbVlmMAPHBLE/p1b6Jl9iIi4hZjDHOX7+GDORtxuQzN6kUxalAilcODrI4mInJVaMBisZJSJ//4bB0LVx0A4PbO9Xjk7lb4+2mZvYiIuGfJmoP8/eMMShwu6tQIZ8yQJGpU1TJ7ERFxz9GsfMYmp7L/aB6B/r489UB7rm9Xy+pYIiJiMw6ni/dnb+C7n/YCcFPH2jxxXxsC/LXMXkQqDg1YLJSdV8QrKWls3ZeNr68Pj/RqSY/r6lsdS0REbMblMsycv5WPv98OQELz6jzzUAcqBWuZvYiIuGfDrhNMmJJOXkEJURHBjB6SSKPaVayOJSIiNpObX8LEqels2HVmmf2gHi24+4YGWmYvIhWOBiwW2X0oh7HJqZw4VUhoSADPP9yRdk20zF5ERNxTVOzgjY/W8NOGIwDcc0NDBvRormX2IiLitvkr9/Kvz9fjdBka1q7M6MGJVI3UMnsREXHP/qO5jE1O5WhWASFB/jzTvwOJzWtYHUtEpExowGKBnzYc5vWZaygucVKrWihjhnaiVrUwq2OJiIjNHM8+s8x+9+Ec/P18GXZvG7ol1rE6loiI2IzT6SL5603MWbYbgOvb1uLJvm0JDtTDRRERcc+qLcd4dcYqCoocVI+qxJihSdStEWF1LBGRMqP/MZcjYwyf/rCD6d9uAaBt42o8/3BHwrTMXkRE3LR130nGp6RxKq+YyLBARg1KpHl8VatjiYiIzeQXljJp+irWbMsE4KHbmtK3W2NdwkVERNxijOGrpbtJ+XojLgMt6lflxYEJRIZpmb2IVGwasJST4lIn73y8liUZBwHoeV08v7urJX5aZi8iIm5atPoA73yyllKHi3o1IxgzJImYqEpWxxIREZs5fOI0YyencjDzNIEBfozs155r28RaHUtERGym1OHiX5+vY0HafgBuSazD433aEOCv57xEpOLTgKUcnMwtYnxKKtv3n8LP14dH72nN7Z3rWR1LRERsxuUyTP92C58t3AFAUosaPP1QB0KCVOciIuKedTuOM3FqOqcLS4mODGb0kCQaxFW2OpaIiNhMzuliJkxNZ9PuLHx9YMhdLbnr+vp6JaSIeA09I1PGdh48xfjkVE7kFBEWEsCLgxJo3bCa1bFERMRmCosdvP7halI3HQXgvpsb0f+2Zvhqmb2IiLjp2xV7eH/2BpwuQ5M6VRg1OJGoiGCrY4mIiM3sO5LLX5NTyTxZQKVgf57t35GOzapbHUtEpFxpwFKGflx/mDdmrqGk1ElcTBhjhiYRG61l9iIi4p7MkwWMTU5l75FcAvx9efL+ttzQobbVsURExGacThcffLWRuT/uAeCG9nEMv78tgQF+FicTERG7Sdt8lNdmrKKw2EnNqqGMGZpE7erhVscSESl3GrCUAWMMsxZsZ+b8rQC0bxrDc/07EhoSYHEyERGxm817snhlSho5p0uoHB7ES4MTaVo3yupYIiJiM6cLSvjbtFWs3XEcgAF3NOPemxrpEi4iIuIWYwyzF+9iyrxNGAOtGkTzwsAEIkIDrY4mImIJDViusuJSJ2/PymDZ2kMA3NWlPkN6ttAyexERcdsP6ft599N1OJwu6sdGMnpIEtWqhFgdS0REbObQ8dOMnbySQ8fzCQ70Y+SDHejcqqbVsURExGZKHU7+8dk6fkg/AMBtnevx6N2t8NdzXiLixTRguYqycgoZl5LGzgNnltk/3qcNt3aqa3UsERGxGafLMG3eZr5YvBOAzq1qMrJfe4K1zF5ERNy0dnsmE6etIr+wlOjKIfxpaBLxsZFWxxIREZs5lVfMK1PS2LL3JL4+8Ltereh5XbxeCSkiXs+2I+Y333yTFi1a0Lx5c5588kmMMZbm2b4/m5FvLWHngVOEVwpk7GPXaLgiImIDntYnBUWljE9JPTtc6XtLY14YkKDhioiIh/O0PgGYt3w3L/9nJfmFpTStW4U3nuqi4YqIiA14WqfsOZzDyLeXsGXvSUKD/Xn595258/r6Gq6IiGDTV7AcP36cd999l02bNhEQEECXLl1YuXIlnTt3tiTPsoxDvDVrDSUOF7Wrh/OnoUnUqBpqSRYREbl0ntYnR7PyGZucyv6jeQT6+zLigXZ0aRdnSRYREbl0ntYnDqeLf3+5gW9X7AXgpo61eeK+NgT4a5m9iIin87ROWbnxCK9/uJqiEiex0WeW2cfFaJm9iMjPbDlgAXA4HBQVFQFQWlpKTExMuWdwuQwz/7uVjxdsB6Bjs+o8278DlYK1zF5ExC48oU8ANu46wStT0skrKCEqIoiXBifRuE4VS7KIiIj7PKVPcvNL+Nu0dNbvPIGPDwy8ozn33NhQP2UsImIjntApxhg+W7iD6d9uwRho0yia5wckEF5Jy+xFRP6Xx14ibMKECSQkJBAeHk5MTAy9e/dm27ZtAFSrVo1nnnmGOnXqEBsbS7du3WjQoEG55isqdvC36elnhyu9uzZg9JAkDVdERDyMp/cJwH9T9zHm/RXkFZTQMC6SN57qquGKiIiHsUOfHDiWxzNvL2X9zhOEBPnx0qBE+tzUSMMVEREP4+mdUlLq5I2P1jDtmzPDlTuuqceff99ZwxURkfPw2AHLkiVLGDZsGCtXrmTBggU4HA66d+9Ofn4+2dnZzJ07l71793Lo0CFWrFjB0qVLyy3biVOFPP+P5axYfwR/Px9G9G3L0Lta4uerBy4iIp7Gk/vE6TJ88NVG3vlkLQ6n4do2sUwYdh1VI0PKLYOIiFwaT+4TgNVbj/HM35dyJCufmKhKTBrehaSWNcs1g4iIXBpP7pTs3CJG/etHFq8+iK+vD4/d05rH+7TB389jn0IUEbGUx14i7Lvvvjvn9ykpKcTExLB69WqOHTtGw4YNiYqKAqBHjx6sXLmSLl26lHmubftOMj4ljey8YiLDAnlxYCIt6lct879XREQuj6f2SX5hKZNmrGLN1kwAHry1KQ/c0lg/ZSwi4qE8tU+MMXy9bDeT52zEZaB5fBSjBiUSGRZU5n+3iIhcHk/tlF0HTzEuJY0TpwoJCwnghQEJtGlcrcz/XhERO/PYAcsv5eTkABAVFUVgYCArVqygqKiIgIAAFi9ezCOPPHLBjy0uLqa4uPjs73Nzcy8rQ2Gxg798kEpeQQn1akYwekgS1aMqXdZtiYiINTyhTwD+89UG1mzNJDDAjz/2a8d1bWpd9m2JiEj5u5I+gavXKRnbj/OfrzYCcEtiHR7v04YAf/2UsYiInXjCY5SSUid/nbySk7nF1KoWxp+GJhFbLeyybktExJvYYsBijGHkyJFcd911tGzZEoA77riDdu3a4evry80338xdd911wY+fMGECf/nLX644R0iQP0/c14ZFqw/wx37ttW9FRMRmPKVPAAb1aMHh4/k80rsVDWtXviq3KSIi5eNK+wSuXqe0a1yN7kl1qV09jF5dGuiVkCIiNuMpj1ECA/x4sm875i7fw9MPdSAsRM95iYhcCh9jjLE6xMUMGzaMefPmsXz5cuLi4tz++PNN82vXrk1OTg4RERFu354xRg9cRMRr5ebmEhkZedn3oVZSn4iIeA5v7hO4up2iPhERb+fNnaLHKCIiV8/l9InHv4Jl+PDhzJkzh6VLl172g5egoCCCgq7eNYhVNCIi9qM+ERGRq+Fq9Alc3U5Rn4iI2JMeo4iI2J/HDliMMQwfPpzZs2ezePFi4uPjrY4kIiI2pD4REZGrQX0iIiJXizpFRKTi8NgBy7Bhw5g5cyZfffUV4eHhHD16FIDIyEhCQkIsTiciInahPhERkatBfSIiIleLOkVEpOLw2B0sF3pJYkpKCoMGDbqi27bztTlFRKxmt/tQ9YmIiGey231oWfYJ2O/zISLiSex2H6rHKCIinqlC7WDx0LmPiIjYjPpERESuBvWJiIhcLeoUEZGKw9fqACIiIiIiIiIiIiIiInajAYuIiIiIiIiIiIiIiIibNGARERERERERERERERFxkwYsIiIiIiIiIiIiIiIibvLYJfdl6edlYrm5uRYnERGxn5/vO7WYUX0iInIl1CfnUqeIiFw+dcr/UZ+IiFy+y+kTrxyw5OXlAVC7dm2Lk4iI2FdeXh6RkZFWx7CU+kRE5MqpT85Qp4iIXDl1ivpERORqcKdPfIwXjvddLheHDx8mPDwcHx8ftz42NzeX2rVrc+DAASIiIsoooefxxnPrzDpzRXYl5zbGkJeXR2xsLL6+3n2lySvpE/DOrz+dWWeuyLzx3OqTq0ePUdyjM3vHmcE7z60zu39mdcr/UZ+4zxvPrTPrzBVZeT9G8cpXsPj6+hIXF3dFtxEREeFVX5g/88Zz68zewRvPDJd/bm//qbCfXY0+Ae/8+tOZvYM3nhm889zqkyunxyiXR2f2Ht54bp3ZPeqUM9Qnl88bz60zewdvPDOU32MU7x7ri4iIiIiIiIiIiIiIXAYNWERERERERERERERERNykAYubgoKCePnllwkKCrI6SrnyxnPrzN7BG88M3ntuT+ON/w46s3fwxjODd57bG8/sibzx30Fn9h7eeG6dWazirf8O3nhundk7eOOZofzP7ZVL7kVERERERERERERERK6EXsEiIiIiIiIiIiIiIiLiJg1YRERERERERERERERE3KQBi4iIiIiIiIiIiIiIiJs0YBEREREREREREREREXGTBizAhAkTSEhIIDw8nJiYGHr37s22bdvOeZ9Bgwbh4+Nzzq9OnTr96rZ++uknbrrpJkJDQ6lcuTI33HADhYWF5XWUS/bnP//5V+epUaPG2T//4osvuPXWW4mOjsbHx4e1a9ee8/EnT55k+PDhNGnShEqVKlGnTh2efPJJcnJyyvkkv23p0qXceeedxMbG4uPjw5dffnnOnxtj+POf/0xsbCwhISHccMMNbNq06eyfu3vO4uJi2rZte97PWXm4lK/li535l+97++23n/dzt337dnr16kV0dDQRERFce+21LFq0qKyOdskmTJiAj48PTz311Nm3XcqZd+3axd133021atWIiIjg/vvv59ixY7+6/Xnz5pGUlERISAjR0dHcc889ZX2k83I4HIwePZr4+HhCQkKoX78+f/3rX3G5XGff52Lfx3Dxc+/du5ehQ4ee/XsaNGjAyy+/TElJSXkc03a8sU/AOzrF2/oE1CngHZ2iPvFc3tgp6hP1ifrEvn0C6hRPpT6pmH0C3tcp6hP1iaf2iQYswJIlSxg2bBgrV65kwYIFOBwOunfvTn5+/jnvd9ttt3HkyJGzv7755ptz/vynn37itttuo3v37qSlpZGens4TTzyBr69nfppbtGhxznk2bNhw9s/y8/O59tprmThx4nk/9vDhwxw+fJjXXnuNDRs2MGXKFL777juGDh1aXvEvSX5+Pm3atOHdd989759PmjSJN954g3fffZf09HRq1KjBLbfcQl5eHuD+OZ977jliY2PL7DwXcylfyxc78/9666238PHxOe/f1aNHDxwOBwsXLmT16tW0bduWnj17cvTo0TI738Wkp6fz73//m9atW5/z9oudOT8/n+7du+Pj48PChQv58ccfKSkp4c477zznzvvzzz/n4YcfZvDgwaxbt44ff/yRBx98sFzP+LO//e1vvPfee7z77rts2bKFSZMm8eqrr/LOO++cfZ+LfR9fyrm3bt2Ky+Xi/fffZ9OmTbz55pu89957jBo1qlzOaTfe2idQ8TvF2/oE1Cne0inqE8/lrZ2iPlGfqE/s2SegTvFU6pOK2SfgfZ2iPlGfeGyfGPmVzMxMA5glS5acfdvAgQNNr169fvPjkpKSzOjRo8s43dXx8ssvmzZt2lz0/fbs2WMAk5GRcdH3/eSTT0xgYKApLS298oBlADCzZ88++3uXy2Vq1KhhJk6cePZtRUVFJjIy0rz33nsXvJ0LnfObb74xTZs2NZs2bbrkz1lZ++XXsjtnXrt2rYmLizNHjhz51efu+PHjBjBLly49+7bc3FwDmO+//75sD3UBeXl5plGjRmbBggWma9euZsSIEcaYSzvz/Pnzja+vr8nJyTn7PidPnjSAWbBggTHGmNLSUlOrVi3zwQcflN+hfkOPHj3MkCFDznnbPffcY/r37/+r973Q9/GlnPt8Jk2aZOLj46/sAF7CG/rEGO/rFG/sE2PUKcZUzE5Rn9iHN3SK+kR98jP1if36xBh1il2oT/5PRekTY7yzU9Qn6hNjPKNPPHPMbLGfXwoXFRV1ztsXL15MTEwMjRs35ve//z2ZmZln/ywzM5PU1FRiYmK45pprqF69Ol27dmX58uXlmt0dO3bsIDY2lvj4eB544AF27959RbeXk5NDREQE/v7+Vylh2dqzZw9Hjx6le/fuZ98WFBRE165dWbFixQU/7nznPHbsGL///e+ZPn06lSpVKtPc7vjl1/KlnrmgoIB+/frx7rvvnvMy2p9VrVqVZs2aMW3aNPLz83E4HLz//vtUr16dDh06lPGpzm/YsGH06NGDbt26nfP2SzlzcXExPj4+BAUFnX2f4OBgfH19z34Pr1mzhkOHDuHr60u7du2oWbMmt99++wVfalrWrrvuOn744Qe2b98OwLp161i+fDl33HHHJd/GpZz7fHJycn51/yjn5y19At7dKd7QJ6BOgYrZKeoT+/CWTlGfqE9AfQL26xNQp9iF+uTy2KlPwDs6RX2iPvkt5dknGrD8gjGGkSNHct1119GyZcuzb7/99tv58MMPWbhwIa+//jrp6encdNNNFBcXA5y9o/7zn//M73//e7777jvat2/PzTffzI4dOyw5y29JSkpi2rRpzJ8/n//85z8cPXqUa665hqysrMu6vaysLMaOHcujjz56lZOWnZ9f1le9evVz3l69evULvuTvfOc0xjBo0CAee+wxOnbsWHaB3XS+r+VLPfMf//hHrrnmGnr16nXe2/bx8WHBggVkZGQQHh5OcHAwb775Jt999x2VK1cumwP9hlmzZrFmzRomTJjwqz+7lDN36tSJ0NBQnn/+eQoKCsjPz+fZZ5/F5XJx5MgR4Nzv8dGjRzN37lyqVKlC165dOXnyZFke77yef/55+vXrR9OmTQkICKBdu3Y89dRT9OvX75Jv41LO/Uu7du3inXfe4bHHHrtaR6mwvKVPQJ1S0fsE1Ck/q4idoj6xB2/pFPWJ+uR/qU/s1SegTrED9Yl39AlU/E5Rn5yhPrmw8uwTDVh+4YknnmD9+vV89NFH57y9b9++9OjRg5YtW3LnnXfy7bffsn37dubNmwdw9tptjz76KIMHD6Zdu3a8+eabNGnShOTk5HI/x8Xcfvvt9OnTh1atWtGtW7ez55g6darbt5Wbm0uPHj1o3rw5L7/88tWOWuZ+eb1FY8x5r8F4oXO+88475Obm8uKLL5Z5Vndc6GsZfvvMc+bMYeHChbz11lsXvG1jDH/4wx+IiYlh2bJlpKWl0atXL3r27HnBO6mycuDAAUaMGMGMGTMIDg6+4Pv91pmrVavGp59+ytdff01YWBiRkZHk5OTQvn17/Pz8gP/7Hn/ppZfo06cPHTp0ICUlBR8fHz799NMyOt2Fffzxx8yYMYOZM2eyZs0apk6dymuvvebW9/ClnPt/HT58mNtuu4377ruP3/3ud1fzOBWSt/QJqFN+VlH7BNQpv1SROkV9Yg/e0inqkzPUJ2eoT+zVJ6BOsQP1iXf1CVTcTlGfnEt98mvl2iduXVCsgnviiSdMXFyc2b179yW9f8OGDc9e42737t0GMNOnTz/nfe6//37z4IMPXvWsZaFbt27mscceO+dtF7seZW5uruncubO5+eabTWFhYTmkvHz84pqKu3btMoBZs2bNOe931113mQEDBpzztt86Z69evYyvr6/x8/M7+wswfn5+v7qd8nKhr+VLOfOIESOMj4/Pr87j6+trunbtaowx5vvvv//VdQyNOfM9MWHChLI72HnMnj377Of7f/P+fIadO3de8r+zMWeutZmdnW2MMaZ69epm0qRJxhhjFi5caACzbNmyc94/MTHRjBo1qmwO9xvi4uLMu+++e87bxo4da5o0afKr972U68pe6Nw/O3TokGncuLF5+OGHjdPpvOL8FZ2394kxFbtTvKlPjFGnVPROUZ94Pm/vFPWJ+sQY9Ykd+sQYdYqnU59U3D4xxrs6RX2iPvlfntAnegULZ6Z6TzzxBF988QULFy4kPj7+oh+TlZXFgQMHqFmzJgD16tUjNjaWbdu2nfN+27dvp27dumWS+2oqLi5my5YtZ89zKXJzc+nevTuBgYHMmTPnNyepnig+Pp4aNWqwYMGCs28rKSlhyZIlXHPNNWffdrFz/v3vf2fdunWsXbuWtWvX8s033wBnpq3jx48vn8P8fxf7Wr6UM7/wwgusX7/+7HnWrl0LwJtvvklKSgpw5nqVAL6+596F+Pr6np16l5ebb76ZDRs2nJO3Y8eOPPTQQ6xdu5b69etf0r/zz6Kjo6lcuTILFy4kMzOTu+66C4AOHToQFBR0zvd4aWkpe/futeR7vKCg4Feffz8/v8v+/F/o3ACHDh3ihhtuoH379qSkpPzq75X/oz45w9s6pSL2CahTvKVT1CeeS52iPgH1ifrkDDv0CahTPJX6xPv6BCpmp6hP1Cce2yduj2QqoMcff9xERkaaxYsXmyNHjpz9VVBQYIwxJi8vzzz99NNmxYoVZs+ePWbRokWmc+fOplatWiY3N/fs7bz55psmIiLCfPrpp2bHjh1m9OjRJjg42OzcudOqo13Q008/bRYvXmx2795tVq5caXr27GnCw8PN3r17jTHGZGVlmYyMDDNv3jwDmFmzZpmMjAxz5MgRY8yZ6XZSUpJp1aqV2blz5zmfN4fDYeXRzpGXl2cyMjJMRkaGAcwbb7xhMjIyzL59+4wxxkycONFERkaaL774wmzYsMH069fP1KxZ8+y/6+Wc81Imp2XlYl/Lxlz8zOfDL34S4vjx46Zq1armnnvuMWvXrjXbtm0zzzzzjAkICDBr164tyyNekq5du5oRI0ac/f2lnDk5Odn89NNPZufOnWb69OkmKirKjBw58pzbHTFihKlVq5aZP3++2bp1qxk6dKiJiYkxJ0+eLK+jnTVw4EBTq1YtM3fuXLNnzx7zxRdfmOjoaPPcc8+dfZ+LfR8bc/FzHzp0yDRs2NDcdNNN5uDBg+d8XcmveWOfGOMdneJtfWKMOuVnFb1T1Ceeyxs7RX2iPlGf2LdPjFGneCr1ScXsE2O8r1PUJ2eoT87wpD7RgMWc+UY636+UlBRjjDEFBQWme/fuplq1aiYgIMDUqVPHDBw40Ozfv/9XtzVhwgQTFxdnKlWqZDp37vyrl1Z5ir59+5qaNWuagIAAExsba+655x6zadOms3+ekpJy3s/Jyy+/bIwxZtGiRRf8vO3Zs8eaQ53HhXIOHDjQGGOMy+UyL7/8sqlRo4YJCgoyXbp0MRs2bLjox//WOa0sm4t9LRtz8TNf6Hb/t2yMMSY9Pd10797dREVFmfDwcNOpUyfzzTfflMGp3PfLsrmUMz///POmevXqJiAgwDRq1Mi8/vrrxuVynfM+JSUl5umnnzYxMTEmPDzcdOvWzWzcuLE8jvQrubm5ZsSIEaZOnTomODjY1K9f37z00kumuLj47Ptc7PvYmIuf+0K3ofn8+XljnxjjHZ3ibX1ijDrlZxW9U9QnnssbO0V9oj5Rn9i3T4xRp3gq9UnF7BNjvK9T1CdnqE/O8KQ+8THGGEREREREREREREREROSS6SKVIiIiIiIiIiIiIiIibtKARURERERERERERERExE0asIiIiIiIiIiIiIiIiLhJAxYRERERERERERERERE3acAiIiIiIiIiIiIiIiLiJg1YRERERERERERERERE3KQBi4iIiIiIiIiIiIiIiJs0YBEREREREREREREREXGTBiwiIiIiIiIiIiIiIiJu0oBFRERERERERERERETETRqwiIiIiIiIiIiIiIiIuEkDFhERERERERERERERETf9Pwo4KmKG03ZXAAAAAElFTkSuQmCC",
      "text/plain": [
       "<Figure size 2000x800 with 8 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "import torch\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "i = 256\n",
    "effective_seqlen = []\n",
    "for e in range(6):\n",
    "    decay_mask = model.retnet_rel_pos(i)[1][0][0]  # [h, i, i]\n",
    "    non_zero_start = torch.tensor([torch.where(mask[-1] > 0)[0][0] for mask in decay_mask])\n",
    "    effective_seqlen.append(i - non_zero_start)\n",
    "    i *= 2\n",
    "    \n",
    "effective_seqlen = torch.stack(effective_seqlen, 1)\n",
    "x = [2 ** (8 + i) for i in range(6)]\n",
    "\n",
    "plt.figure(figsize=(effective_seqlen.size(0) // 2 * 5, 4 * 2))\n",
    "for i in range(effective_seqlen.size(0)):\n",
    "    r, c = i // 2, i % 2\n",
    "    plt.subplot(2, effective_seqlen.size(0) // 2, i + 1)\n",
    "    if i == 0:\n",
    "        plt.ylabel(\"Effective Sequence Length\")\n",
    "    plt.title(f\"head {i}\")\n",
    "    plt.loglog(x, effective_seqlen[i], base=2)\n",
    "    plt.xticks(x, x)\n",
    "\n",
    "print(effective_seqlen[i])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Inference Time Test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "############################## seqlen: 256 ##############################\n",
      "parallel: 1.0023608207702637\n",
      "recurrent: 0.1654500961303711\n",
      "############################## seqlen: 512 ##############################\n",
      "parallel: 1.9274981021881104\n",
      "recurrent: 0.06916308403015137\n",
      "############################## seqlen: 1024 ##############################\n",
      "parallel: 5.113396883010864\n",
      "recurrent: 0.07078385353088379\n",
      "############################## seqlen: 2048 ##############################\n",
      "parallel: 9.908579111099243\n",
      "recurrent: 0.0744469165802002\n",
      "############################## seqlen: 4096 ##############################\n",
      "parallel: 37.91721701622009\n",
      "recurrent: 0.11470627784729004\n"
     ]
    }
   ],
   "source": [
    "import time\n",
    "\n",
    "model = RetNetModel(config).to(device)\n",
    "model.eval()\n",
    "\n",
    "parallel_history = []\n",
    "recurrent_history = []\n",
    "\n",
    "sequence_lengths = [256, 512, 1024, 2048, 4096]\n",
    "generate_length = 20\n",
    "for seqlen in sequence_lengths:\n",
    "    print(\"#\" * 30, f\"seqlen: {seqlen}\", \"#\" * 30)\n",
    "    input_ids = torch.ones(1, seqlen).long().to(device)\n",
    "    # first, 1 forward pass to compute the prompt\n",
    "    parallel_outputs = model(input_ids, forward_impl='parallel', use_cache=True)\n",
    "    past_kv = parallel_outputs.past_key_values\n",
    "    # start generating. This is mock generation, we don't care about the output\n",
    "    start = time.time()\n",
    "    for i in range(1, generate_length):\n",
    "        input_ids = torch.ones(1, seqlen + i).long().to(device)\n",
    "        parallel_outputs = model(input_ids, forward_impl='parallel', use_cache=True)\n",
    "    parallel_time = time.time() - start\n",
    "    print(f\"parallel: {parallel_time}\")\n",
    "    parallel_history.append(parallel_time)\n",
    "\n",
    "    start = time.time()\n",
    "    for i in range(1, generate_length):\n",
    "        input_ids = torch.ones(1, seqlen + i).long().to(device)\n",
    "        rnn_out = model(input_ids, forward_impl='recurrent', past_key_values=past_kv, use_cache=True)\n",
    "        past_kv = rnn_out.past_key_values  \n",
    "    recurrent_time = time.time() - start\n",
    "    print(f\"recurrent: {recurrent_time}\")\n",
    "    recurrent_history.append(recurrent_time)\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(sequence_lengths, parallel_history, label='parallel')\n",
    "plt.plot(sequence_lengths, recurrent_history, label='recurrent')\n",
    "plt.xticks(sequence_lengths, sequence_lengths)\n",
    "plt.legend()\n",
    "plt.xlabel(\"Sequence Length\")\n",
    "plt.ylabel(\"Time (s)\")\n",
    "plt.title(\"Parallel vs Recurrent Inference\");"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "torch-m1",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.3"
  },
  "orig_nbformat": 4
 },
 "nbformat": 4,
 "nbformat_minor": 2
}