{ "cells": [ { "cell_type": "markdown", "id": "a8650ad0", "metadata": {}, "source": [ "# Time Series Super-Resolution\n", "\n", "In this tutorial, we will go through how to perform time series super-resolution with `GenTS` models and datasets.\n", "\n", "## Problem setting\n", "Given a low-resolution time series $\\mathbf{x}_{\\text{lr}} \\in \\mathbb{R}^{(T / r) \\times D}$, where $r$ is the super-resolution factor, we are interested in learning the conditional distribution $p(\\mathbf{x}_{\\text{hr}} \\mid \\mathbf{x}_{\\text{lr}})$. From this distribution, we can sample possible high-resolution reconstructions $\\hat{\\mathbf{x}}_{\\text{hr}} \\in \\mathbb{R}^{T \\times D}$.\n", "\n", "Two downsampling strategies are supported:\n", "- **subsample**: take every $r$-th point from the original series\n", "- **average**: average every $r$ consecutive points into one" ] }, { "cell_type": "markdown", "id": "e899ca45", "metadata": {}, "source": [ "## Implementation\n", "### 1. Import modules" ] }, { "cell_type": "code", "execution_count": 1, "id": "d41009be", "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "/home/wcx/anaconda3/envs/gents/lib/python3.10/site-packages/tqdm/auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n", " from .autonotebook import tqdm as notebook_tqdm\n", "CUDA extension for cauchy multiplication not found. Install by going to extensions/cauchy/ and running `python setup.py install`. This should speed up end-to-end training by 10-50%\n", "Falling back on slow Cauchy kernel. Install at least one of pykeops or the CUDA extension for efficiency.\n", "Falling back on slow Vandermonde kernel. Install pykeops for improved memory efficiency.\n" ] } ], "source": [ "import torch\n", "import matplotlib.pyplot as plt\n", "from gents.dataset import Spiral2D\n", "from gents.model import VanillaDDPM\n", "from lightning import Trainer" ] }, { "cell_type": "markdown", "id": "fdf45e87", "metadata": {}, "source": [ "### 2. Setup datamodule and model\n", "Here, we set $T=64$, and the super-resolution factor $r=4$, so the low-resolution condition has length $64/4=16$.\n", "\n", "Note that `condition='super_resolution'`, `sr_factor` and `sr_type` are required for both datamodule and model." ] }, { "cell_type": "code", "execution_count": 2, "id": "a2adddb2", "metadata": {}, "outputs": [], "source": [ "sr_factor = 4\n", "sr_type = \"subsample\" # or \"average\"\n", "\n", "dm = Spiral2D(\n", " seq_len=64,\n", " batch_size=64,\n", " data_dir=\"../data\",\n", " condition=\"super_resolution\",\n", " sr_factor=sr_factor,\n", " sr_type=sr_type,\n", ")\n", "model = VanillaDDPM(\n", " seq_len=dm.seq_len,\n", " seq_dim=dm.seq_dim,\n", " condition=\"super_resolution\",\n", " sr_factor=sr_factor,\n", " sr_type=sr_type,\n", ")" ] }, { "cell_type": "markdown", "id": "d5a43006", "metadata": {}, "source": [ "### 3. Training\n", "Utilizing `lightning`/`pytorch-lightning`, one can easily set GPU devices, training epochs, callbacks, etc." ] }, { "cell_type": "code", "execution_count": 3, "id": "d5540f65", "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "GPU available: True (cuda), used: True\n", "TPU available: False, using: 0 TPU cores\n", "HPU available: False, using: 0 HPUs\n", "You are using a CUDA device ('NVIDIA GeForce RTX 3080 Ti') that has Tensor Cores. To properly utilize them, you should set `torch.set_float32_matmul_precision('medium' | 'high')` which will trade-off precision for performance. For more details, read https://pytorch.org/docs/stable/generated/torch.set_float32_matmul_precision.html#torch.set_float32_matmul_precision\n", "LOCAL_RANK: 0 - CUDA_VISIBLE_DEVICES: [0,1,2,3]\n", "\n", " | Name | Type | Params | Mode \n", "------------------------------------------\n", "0 | backbone | DiT | 1.2 M | train\n", "------------------------------------------\n", "1.2 M Trainable params\n", "1.0 K Non-trainable params\n", "1.2 M Total params\n", "4.703 Total estimated model params size (MB)\n", "85 Modules in train mode\n", "0 Modules in eval mode\n", "`Trainer.fit` stopped: `max_epochs=10` reached.\n" ] } ], "source": [ "trainer = Trainer(max_epochs=10, devices=[0], enable_progress_bar=False)\n", "trainer.fit(model, dm)" ] }, { "cell_type": "markdown", "id": "513a84a7", "metadata": {}, "source": [ "### 4. Super-resolution on the test set\n", "Here `n_sample=10` means we sample $\\hat{\\mathbf{x}}_{\\text{hr}} \\in \\mathbb{R}^{T \\times D}$ 10 times for each low-resolution input. The tensor shape of `gen_data` is `[batch_size, T, D, n_sample]`." ] }, { "cell_type": "code", "execution_count": 4, "id": "dbfe94f6", "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Low-res condition shape: torch.Size([100, 16, 2])\n", "Generated high-res shape: torch.Size([100, 64, 2, 10])\n" ] } ], "source": [ "dm.setup(\"test\")\n", "real_data = torch.cat([batch[\"seq\"] for batch in dm.test_dataloader()])\n", "cond_data = torch.cat([batch[\"c\"] for batch in dm.test_dataloader()])\n", "\n", "gen_data = model.sample(\n", " n_sample=10,\n", " condition=cond_data,\n", ")\n", "print(f\"Low-res condition shape: {cond_data.shape}\")\n", "print(f\"Generated high-res shape: {gen_data.shape}\")" ] }, { "cell_type": "markdown", "id": "f2e01578", "metadata": {}, "source": [ "### 5. Visualization\n", "We plot the ground truth high-resolution series and the super-resolution results (mean and 95% prediction interval)." ] }, { "cell_type": "code", "execution_count": 5, "id": "445c33ad", "metadata": {}, "outputs": [ { "data": { "image/png": "iVBORw0KGgoAAAANSUhEUgAABLsAAAHICAYAAACmiOgzAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjMsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvZiW1igAAAAlwSFlzAAAPYQAAD2EBqD+naQAA4n1JREFUeJzs3Xd4VGX6xvHvmcmkkkYqgUCAkNC7ID00QUWxoFh2ETusrl13XXft+3Ntu7rqYltFFCsqdlBAVKq00AKhJwFCeu+ZOb8/RrJGOiQ5Kffnuua6mDNnZu4Jgbx5zvs+r2GapomIiIiIiIiIiEgzYLM6gIiIiIiIiIiISF1RsUtERERERERERJoNFbtERERERERERKTZULFLRERERERERESaDRW7RERERERERESk2VCxS0REREREREREmg0Vu0REREREREREpNnwsDqAiEhz53Q6qaqqsjqGiIiINCIOhwO73W51DBGRZknFLhGRemKaJocOHSI/P9/qKCIiItIIBQUFERkZiWEYVkcREWlWVOwSEaknhwtd4eHh+Pr6aiArIiIigPuCWGlpKZmZmQC0adPG4kQiIs2Lil0iIvXA6XTWFLpCQkKsjiMiIiKNjI+PDwCZmZmEh4drSaOISB1Sg3oRkXpwuEeXr6+vxUlERESksTo8TlBvTxGRuqVil4hIPdLSRRERETkWjRNEROqHil0iIiIiIiIiItJsqGeXiIiIiJy5qnJwNdBSLJsDHN4N814iIiLS5KjYJSIiTdrs2bO54447yM/PtzTH9OnTyc/PZ/78+ZbmaCoSEhLo27cvzz33nNVR6lSL/T6oKoe9P0JFYcO8n1cAdBypgpeIiIgclZYxiohIs7Zv3z4MwyAxMbFRvp40L88//zyzZ89u8PedPXs2QUFBDf6+NVxV7kKX3RO8/Ov3Zvd0v9cpzCIrKirijjvuoEOHDvj4+DB06FDWrFlT65zp06djGEat28SJE2ser6io4Pe//z0BAQHExcWxaNGiWs9/+umn+eMf/3jCLA8//HDN63t4eBATE8Odd95JcXExoP9jGqvS0lLuv/9+OnfujLe3N2FhYYwaNYrPPvus5pyEhISav1tvb2/i4uJ44oknME3TwuQiIi2TZnaJiMgZqaysxNPT0+oYZ6y5fI7TUVpVypykOczbMY+ssizCfMKYEjeFad2n4etovjuKVlVV4XA46vQ1AwMD6/T1mhwPL2iI7xln5SmdfsMNN7BlyxbefvttoqKieOeddxg3bhxJSUm0bdu25ryJEyfy5ptv1tz38vKq+fOrr77KunXrWLlyJd988w1XXXUVGRkZGIbB3r17ee2111i7du1J5enRoweLFi2iurqa5cuXc91111FaWsorr7xySp9LGs6MGTNYvXo1L7zwAt27dycnJ4cVK1aQk5NT67wbb7yRRx99lIqKCpYsWcJNN91EUFAQM2fOtCi5iEjLpJldIiINxDRNKqqdDX47lSvKRUVFXH311fj5+dGmTRv+9a9/kZCQwB133FFzTkxMDI899hjTpk0jICCAm266CYCPP/6YHj164OXlRUxMDM8++2yt1zYM44ilXUFBQTWzYA7PZvjkk08YPXo0vr6+9OnTh5UrV9Z6zuzZs2nfvj2+vr5cfPHFR/yi8VsdO3YEoF+/fhiGQUJCAuCexXHRRRfx97//naioKOLj408q57Fe77BnnnmGNm3aEBISwi233NLot5MvrSpl+oLpzEqcRUZpBi7TRUZpBrMSZzF9wXRKq0obJEdeXh7Tpk0jODgYX19fzj33XHbu3Am4/+2EhYUxb968mvP79u1LmzZtau4vW7YMLy8vSkuPnvfw99cHH3zAqFGj8Pb2Zu7cuQC8/vrrdOvWDW9vb7p27cp//vOfmudVVlZy66230qZNG7y9venQoQNPPPHEMT/H4e+rwxISErjtttu47777aN26NZGRkTz88MO1nmMYBrNmzeLcc8/Fx8eHTp061fqsS5cuxTCMWkt1ExMTMQyDffv2sXTpUq699loKCgpqZpX89j1asrKyMj7++GOeeuopRo4cSWxsLA8//DCxsbHMmjWr1rleXl5ERkbW3IKDg2se27ZtGxdeeCE9evTglltuISsri+zsbABmzpzJk08+SUBAwEll8vDwIDIyknbt2jF16lSuvvpqPv/887r70KfDNKGyxJrbKfycWrBgAcOHDycoKIiQkBAmTZrE7t27ax4fOnQof/rTn2o9JysrC4fDwY8//ghAeno6559/Pj4+PnTs2JF3332XmJiY4y6r/vzzz/nLX/7CeeedR0xMDAMGDOCPf/wj1113Xa3zfH19iYyMpEOHDlx77bX07t2b77777qQ/n4iI1A3N7BIRaSCVThcPf57U4O/78IXd8fKwn9S5d911F8uXL+fzzz8nIiKCBx98kPXr19O3b99a5z3zzDM8+OCDPPTQQwCsW7eOyy+/nIcffpipU6eyYsUK/vCHPxASEsL06dNPKe8DDzzAM888Q5cuXXjggQe48sor2bVrFx4eHqxevZrrr7+eJ554gosuuogFCxbUZDiWn3/+mUGDBrFo0SJ69OhRa/bW4sWLCQgIOKVfRI73et9//z1t2rTh+++/Z9euXUydOpW+ffty4403ntLXoCHNSZpDcm4yLly1jrtwkZybzJykOczoM6Pec0yfPp2dO3fy+eefExAQwJ/+9CfOO+88kpKScDgcjBw5kqVLlzJlyhTy8vLYtm0bPj4+bN++na5du/LDDz9w1lln4et7/FlFf/7zn3n22Wfp169fTcHrwQcf5MUXX6Rfv35s2LCBG2+8ET8/P6655hr+/e9/8/nnn/Phhx/Svn170tLSSEtLO6XP9tZbb3HXXXexevVqVq5cyfTp0xk2bBjjx4+vOedvf/sb//jHP3j++ed5++23ueKKK9i8eTPdunU74esPHTqU5557jgcffJDk5GQAWrVqdUoZm7Pq6mqcTife3rX7e/n4+LBs2bJax5YuXUp4eDjBwcGMGTOGxx9/nJCQEAD69OnD22+/TVlZGQsXLqRNmzaEhoYyd+5cvL29ufjii087o4+PD5WVpzZbrc5VlcL/RVnz3n85CJ5+J3VqSUkJd911F71796a4uJgHH3yQiy++mMTERGw2G1dffTVPPfUU//jHPzAMA4APPviAqKgoRowYAcC0adPIzs5m6dKlOBwO7rrrLjIzM4/7vpGRkXz99ddccskl+Pv7nzCnaZosW7aM7du306VLl5P6bCIiUndU7BIREcA9q+utt97i3XffZezYsQC8+eabREUd+cvPmDFjuPvuu2vuX3311YwdO5a//e1vAMTFxZGUlMTTTz99ysWue+65h/PPPx+ARx55hB49erBr1y66du3K888/z8SJE7nvvvtq3mfFihUsWLDgmK8XFhYGQEhICJGRkbUe8/Pz4/XXXz+l5YvHe73g4GBefPFF7HY7Xbt25fzzz2fx4sWNutg1b8e8Iwpdh7lwMW/HvHovdh0uci1fvpyhQ4cCMHfuXKKjo5k/fz6XXXYZCQkJNUu8fvzxR/r160dkZCRLly6la9euLF26lFGjRp3wve644w4uueSSmvsPPfQQzz77bM2xjh07kpSUxCuvvMI111xDamoqXbp0Yfjw4RiGQYcOHU758/Xu3bumKNulSxdefPFFFi9eXKvYddlll3HDDTcA8Nhjj/Hdd9/xwgsv1Jpldiyenp4EBgZiGMYR35MC/v7+DBkyhMcee4xu3boRERHBe++9x8qVK4mNja05b+LEiVxyySV07NiR3bt385e//IVzzz2XlStXYrfbue6669i0aRPdu3cnNDSUDz/8kLy8PB588EGWLl3KX//6V95//306d+7MG2+8UWt55PGsW7eOd999lzFjxtTXl6BZufTSS2vdf+ONNwgLCyMpKYmePXty+eWXc8cdd7Bs2bKa4ta7777LlVdeiWEYbN++nUWLFrFmzRoGDhwIuGd3nqgg9eqrr3L11VcTEhJCnz59GD58OFOmTGHYsGG1zvvPf/7D66+/TmVlJVVVVXh7e3PbbbfV4VdAREROhopdIiINxNNu4+ELu1vyvidjz549VFVVMWjQoJpjgYGBNcv7fu3wLwiHbdu2jcmTJ9c6NmzYMJ577jmcTid2+8nNLAN3YeCww8vUMjMz6dq1K9u2bTti9sSQIUOOW+w6nl69etVpn64ePXrU+qxt2rRh8+bNdfb69SGrLOuMHq8L27Ztw8PDg8GDB9ccCwkJIT4+nm3btgEwatQobr/9drKysvjhhx9ISEioKXZdf/31rFixoqYIOmPGDN55552a1zrc+Btqf++WlJSwe/durr/++loFyerq6preW9OnT2f8+PHEx8czceJEJk2axDnnnHNKn+/X39Pg/r747SySIUOGHHFfDcrrzttvv811111H27Ztsdvt9O/fnyuvvJJ169bVnHPFFVfU/LlXr1707t2bzp07s3TpUsaOHYvD4eCll16q9brXXnstt912Gxs2bGD+/Pls3LiRp556ittuu42PP/74mHk2b95Mq1atcDqdVFZWcv755/Piiy/W/Qc/FQ5f9wwrq977JO3cuZMHH3yQ1atXk52djcvlLtanpqbSs2dPwsLCOOecc5g7dy4jRoxg7969rFy5sqZYnpycjIeHB/379695zdjY2FpLVo9m5MiR7Nmzh1WrVrFixQoWL17M888/zyOPPFJzoQfcF38eeOAB8vLyeOihhxg6dGhNEV9ERBqOil0iIg3EMIyTXk7Y2Pn5ndxyk18zDOOI/mFH62f164bhh5egHP5lpq4d7XOcbM6j+W2zc8Mw6i17XQnzCSOjNOO4jzcGvXr1onXr1vzwww/88MMP/P3vfycyMpInn3ySNWvWUFVVVfML5aOPPso999xz1Nf59d/54SLYa6+9VqvQBtQULfv378/evXv55ptvWLRoEZdffjnjxo2r1VPrRM70+8Jmcxesf/192dh7wTU2nTt35ocffqCkpITCwkLatGnD1KlT6dSp0zGf06lTJ0JDQ9m1a1fNbNdf+/7779m6dSuvv/469957L+eddx5+fn5cfvnlJyxcxcfH8/nnn+Ph4UFUVFTj2BzDME56KaGVLrjgAjp06MBrr71GVFQULpeLnj171loGevXVV3Pbbbfxwgsv8O6779KrVy969ep1xu/tcDgYMWIEI0aM4E9/+hOPP/44jz76KH/6059q/g4DAwNrZgx++OGHxMbGcvbZZzNu3Lgzfn8RETl5alAvIiKA+xc7h8PBmjVrao4VFBSwY8eOEz63W7duLF++vNax5cuXExcXV1M0CAsLIz09vebxnTt3HrOZ+PHeZ/Xq1bWOrVq16rjPOfwLiNPpPKn3OFHOU329xm5K3BRsxxgO2LAxJW5KvWfo1q0b1dXVtf5uc3JySE5Opnt392xIwzAYMWIEn332GVu3bmX48OH07t2biooKXnnlFQYOHFhTyAoPDyc2NrbmdiwRERFERUWxZ8+eWufHxsbWbEQAEBAQwNSpU3nttdf44IMP+Pjjj8nNza3Tr8Fvv49XrVpV06/r8NLZX39f/nbWl6enZ7P5nqxPhzffyMvLY+HChUfMSP21/fv3k5OTU2sjhMPKy8u55ZZbeOWVV7Db7TidzpoCZFVV1Qn/Ljw9PYmNjSUmJqZxFLqaiMP/L/z1r39l7NixdOvWjby8vCPOmzx5MuXl5SxYsIB3332Xq6++uuax+Ph4qqur2bBhQ82xXbt2HfV1TqR79+5UV1dTXl5+1MdbtWrF7bffzj333HNKm8WIiMiZ08wuEREB3H1trrnmGu69915at25NeHg4Dz30EDabrWaG1bHcfffdnHXWWTz22GNMnTqVlStX8uKLL9bqNzRmzBhefPFFhgwZgtPp5E9/+tMRM15O5LbbbmPYsGE888wzTJ48mYULF55wCWN4eDg+Pj4sWLCAdu3a4e3tXbNE7WhOlPNUX6+xm9Z9GktSlxzRpN6GjfjW8UzrPq3eM3Tp0oXJkydz44038sorr+Dv78+f//xn2rZtW6sYkZCQwN13383AgQNrGrCPHDmSuXPncu+9957Wez/yyCPcdtttBAYGMnHiRCoqKli7di15eXncdddd/POf/6RNmzb069cPm83GRx99RGRkJEFBQXXx0Wt89NFHDBw4kOHDhzN37lx+/vln/vvf/wLuJVbR0dE8/PDD/P3vf2fHjh1H7HYaExNDcXExixcvpk+fPvj6+p6wWX+9qK5olO+xcOFCTNMkPj6eXbt2ce+999K1a1euvfZawD3L75FHHuHSSy8lMjKS3bt3c9999xEbG8uECROOeL3HHnuM8847j379+gHuZdv33nsv1157LS+++OIRfZykbgQHBxMSEsKrr75KmzZtSE1N5c9//vMR5/n5+XHRRRfxt7/9jW3btnHllVfWPNa1a1fGjRvHTTfdxKxZs3A4HNx99934+Pgc92ddQkICV155JQMHDiQkJISkpCT+8pe/MHr06OPuwnnzzTfz2GOP8fHHHzNlSv1fPBARETfN7BIRkRr//Oc/GTJkCJMmTWLcuHEMGzaMbt26HbGL2W/179+fDz/8kPfff5+ePXvy4IMP8uijj9ZqTv/ss88SHR3NiBEjuOqqq7jnnntO+Zfxs88+m9dee43nn3+ePn368O233/LXv/71uM/x8PDg3//+N6+88gpRUVHHnclxMjlP9fUaO1+HL7MnzmZm35lE+EZgM2xE+EYws+9MZk+cje8p9NI5E2+++SYDBgxg0qRJDBkyBNM0+frrr2sVGkeNGoXT6SQhIaHmWEJCwhHHTsUNN9zA66+/zptvvkmvXr0YNWoUs2fPrpnZ5e/vz1NPPcXAgQM566yz2LdvH19//XXN0sK68sgjj/D+++/Tu3dv5syZw3vvvVczq83hcPDee++xfft2evfuzZNPPsnjjz9e6/lDhw5lxowZTJ06lbCwMJ566qk6zXdCNgd4BYCzEiqK6vfmrHS/l+3ki+UFBQXccsstdO3alWnTpjF8+HAWLlxY8/1lt9vZtGkTF154IXFxcVx//fUMGDCAn376CS8vr1qvtWXLFj788EMeeeSRmmNTpkzh/PPPZ8SIEWzatInnn3++br6uUovNZuP9999n3bp19OzZkzvvvJOnn376qOdeffXVbNy4kREjRtC+fftaj82ZM4eIiAhGjhzJxRdfzI033oi/v/9xf9ZNmDCBt956i3POOYdu3brxxz/+kQkTJvDhhx8eN3Pr1q2ZNm0aDz/8cKNf1i4i0pwYpubUiojUufLycvbu3UvHjh1PWChqzEpKSmjbti3PPvss119/vdVxRJolwzD49NNPueiii6yOcmaqysHVQL3EbA5wNN3/W6Vx2b9/P9HR0SxatOio/dnqU3MZL4iINDZaxigiIjU2bNjA9u3bGTRoEAUFBTz66KMATX72kog0AIc3oF/WpfFbsmQJxcXF9OrVi/T0dO677z5iYmIYOXKk1dFERKSOqNglIiK1PPPMMyQnJ+Pp6VmzjCc0NNTqWCIiInWiqqqKv/zlL+zZswd/f3+GDh3K3LlzT7mPpIiINF5axigiUg+0LEFEREROROMFEZH6oQb1IiL1SNcTRERE5Fg0ThARqR8qdomI1IPDSyFKS0stTiIiIiKN1eFxgpZQiojULfXsEhGpB3a7naCgIDIzMwHw9fXFMAyLU4mIiEhjYJompaWlZGZmEhQUhN1utzqSiEizop5dIiL1xDRNDh06RH5+vtVRREREpBEKCgoiMjJSF8REROqYil0iIvXM6XRSVVVldQwRERFpRBwOh2Z0iYjUExW7RERERERERESk2VCDehERERERERERaTZU7BIRERERERERkWZDxS4REREREREREWk2VOwSEREREREREZFmQ8UuERERERERERFpNlTsEhERERERERGRZkPFLhERERERERERaTZU7BIRERERERERkWZDxS4REREREREREWk2VOwSEREREREREZFmQ8UuERERERERERFpNlTsEhERERERERGRZkPFLhERERERERERaTZU7BIRERERERERkWZDxS4REREREREREWk2VOwSEREREREREZFmQ8UuERERERERERFpNlTsEhERERERERGRZkPFLhFp9AzD4NZbb7U6Rp0xDIOHH37Y6hgiIiIix6Txl4g0ZSp2iYhldu/ezc0330ynTp3w9vYmICCAYcOG8fzzz1NWVmZ1PMvNmjWLyy67jPbt22MYBtOnT7c6koiIiDRxGn8dW1paGo888giDBg0iODiY0NBQEhISWLRokdXRROQUeVgdQERapq+++orLLrsMLy8vpk2bRs+ePamsrGTZsmXce++9bN26lVdffdXqmJZ68sknKSoqYtCgQaSnp1sdR0RERJo4jb+O77PPPuPJJ5/koosu4pprrqG6upo5c+Ywfvx43njjDa699lqrI4rISVKxS0Qa3N69e7niiivo0KEDS5YsoU2bNjWP3XLLLezatYuvvvrKwoSNww8//FAzq6tVq1ZWxxEREZEmTOOvExs9ejSpqamEhobWHJsxYwZ9+/blwQcfVLFLpAnRMkYRaXBPPfUUxcXF/Pe//6010DosNjaW22+//Yjj8+fPp2fPnnh5edGjRw8WLFhQ6/GUlBT+8Ic/EB8fj4+PDyEhIVx22WXs27ev1nmzZ8/GMAyWL1/OXXfdRVhYGH5+flx88cVkZWXVOjcmJoZJkyaxbNkyBg0ahLe3N506dWLOnDlH5MvPz+eOO+4gOjoaLy8vYmNjefLJJ3G5XKfxVYIOHTpgGMZpPVdERETk1zT+OrEePXrUKnQBeHl5cd5557F//36KiopO+TVFxBqa2SUiDe6LL76gU6dODB069KSfs2zZMj755BP+8Ic/4O/vz7///W8uvfRSUlNTCQkJAWDNmjWsWLGCK664gnbt2rFv3z5mzZpFQkICSUlJ+Pr61nrNP/7xjwQHB/PQQw+xb98+nnvuOW699VY++OCDWuft2rWLKVOmcP3113PNNdfwxhtvMH36dAYMGECPHj0AKC0tZdSoURw4cICbb76Z9u3bs2LFCu6//37S09N57rnnzuyLJiIiInIGNP46fYcOHcLX1/eIzyIijZgpItKACgoKTMCcPHnyST8HMD09Pc1du3bVHNu4caMJmC+88ELNsdLS0iOeu3LlShMw58yZU3PszTffNAFz3Lhxpsvlqjl+5513mna73czPz6851qFDBxMwf/zxx5pjmZmZppeXl3n33XfXHHvsscdMPz8/c8eOHbXe/89//rNpt9vN1NTUWp/noYceOunPb5qm6efnZ15zzTWn9BwRERER09T46/DnOdXxl2ma5s6dO01vb2/z97///Sk/V0Sso2WMItKgCgsLAfD39z+l540bN47OnTvX3O/duzcBAQHs2bOn5piPj0/Nn6uqqsjJySE2NpagoCDWr19/xGvedNNNtZYJjhgxAqfTSUpKSq3zunfvzogRI2ruh4WFER8fX+u9P/roI0aMGEFwcDDZ2dk1t3HjxuF0Ovnxxx9P6fOKiIiI1BWNv05PaWkpl112GT4+PvzjH/84o9cSkYalZYwi0qACAgIATrnnQfv27Y84FhwcTF5eXs39srIynnjiCd58800OHDiAaZo1jxUUFJzwNYODgwFqvebJvvfOnTvZtGkTYWFhR82fmZl51OMiIiIi9U3jr1PndDq54oorSEpK4ptvviEqKuq0X0tEGp6KXSLSoAICAoiKimLLli2n9Dy73X7U478eUP3xj3/kzTff5I477mDIkCEEBgZiGAZXXHHFUZuUnsxrnux5LpeL8ePHc9999x313Li4uKMeFxEREalvGn+duhtvvJEvv/ySuXPnMmbMmNN+HRGxhopdItLgJk2axKuvvsrKlSsZMmRInb3uvHnzuOaaa3j22WdrjpWXl5Ofn19n73EsnTt3pri4mHHjxtX7e4mIiIicKo2/Tt69997Lm2++yXPPPceVV15Zp68tIg1DPbtEpMHdd999+Pn5ccMNN5CRkXHE47t37+b5558/5de12+1HXBV84YUXcDqdp531ZF1++eWsXLmShQsXHvFYfn4+1dXV9Z5BRERE5Fg0/jo5Tz/9NM888wx/+ctfuP322+sipohYQDO7RKTBde7cmXfffZepU6fSrVs3pk2bRs+ePamsrGTFihV89NFHTJ8+/ZRfd9KkSbz99tsEBgbSvXt3Vq5cyaJFi2q2xq5P9957L59//jmTJk2q2Ra7pKSEzZs3M2/ePPbt20doaOgpveYXX3zBxo0bAXfD102bNvH4448DcOGFF9K7d+86/xwiIiLSPGn8dWKffvop9913H126dKFbt2688847tR4fP348ERERdf0xRKQeqNglIpa48MIL2bRpE08//TSfffYZs2bNwsvLi969e/Pss89y4403nvJrPv/889jtdubOnUt5eTnDhg1j0aJFTJgwoR4+QW2+vr788MMP/N///R8fffQRc+bMISAggLi4OB555BECAwNP+TU//vhj3nrrrZr7GzZsYMOGDQC0a9dOxS4RERE5JRp/Hd/hi4w7d+7k97///RGPf//99yp2iTQRhvnbOaciIiIiIiIiIiJNlHp2iYiIiIiIiIhIs6Fil4iIiIiIiIiINBsqdomIiIiIiIiISLOhYpeIiIiIiIiIiDQbKnaJiIiIiIiIiEizoWKXiIiIiIiIiIg0Gx5WB2hILpeLgwcP4u/vj2EYVscRERGRJso0TYqKioiKisJm07XD49H4S0REROrKyY7BWlSx6+DBg0RHR1sdQ0RERJqJtLQ02rVrZ3WMRk3jLxEREalrJxqDtahil7+/P+D+ogQEBFicRkRERJqqwsJCoqOja8YWcmwaf4mIiEhdOdkxWIsqdh2eOh8QEKDBloiIiJwxLcs7MY2/REREpK6daAymJhMiIiIiIiIiItJsqNglIiIiIiIiIiLNhopdIiIiIiIiIiLSbDSZYtcTTzzBWWedhb+/P+Hh4Vx00UUkJydbHUtERERERERERBqRJlPs+uGHH7jllltYtWoV3333HVVVVZxzzjmUlJRYHU1ERERERERERBqJJrMb44IFC2rdnz17NuHh4axbt46RI0dalEpERERERERERBqTJlPs+q2CggIAWrdufcxzKioqqKioqLlfWFhY77lERERERERERMQ6TbLY5XK5uOOOOxg2bBg9e/Y85nlPPPEEjzzySAMmExERERERETl5pVWlzEmaw7wd88gqyyLMJ4wpcVOY1n0avg5fq+OJNEmGaZqm1SFO1cyZM/nmm29YtmwZ7dq1O+Z5R5vZFR0dTUFBAQEBAQ0RVUREmgANMuVUFRYWEhgYqDHFSdDXSkTk2EqrSpm+YDrJucm4cNUct2EjvnU8syfO1lhE5FdOdlzRZBrUH3brrbfy5Zdf8v333x+30AXg5eVFQEBArZuIiMivHR5kzkqcRUZpBi7TRUZpBrMSZzF9wXRKq0qtjigiIiLN1JykOUcUugBcuEjOTWZO0hyLkok0bU2m2GWaJrfeeiuffvopS5YsoWPHjlZHEhGRZsA9yNyuQaaIiIg0CNM0Ka9ykl1cwfvbPzxiDHKYCxfzdsxr4HQizUOT6dl1yy238O677/LZZ5/h7+/PoUOHAAgMDMTHx8fidCIi0mQUplN9YD0le9fhPLCBj+w7cNmPfu3n8CBzRp8ZDRxSREREmpu92SXM33CA3JJKql3ubkI55dnHfU5WaRamaWIYRkNEFGk2mkyxa9asWQAkJCTUOv7mm28yffr0hg8kIiJNSknKBlzzrse/aDceQOAvx7Njoo/7vKzSrHrPJiIiIs1b8qEiPlqxjQGZnxBZvofgyoOEVB5kTRsvMjyO8Wu5CZ62QJ5emEzvdoH0ahdEVKC3Cl8iJ6HJFLuaYB99ERFpBKqdLrb/8CFxy+7A01WGC4NMR1syvTtRFhhLoG0leWbJ0Z9smoQ6q3Eu+zf2wTeBw7thw4uIiEiTt2l/Pl+s3Mw1e+4lumxbrcemFAUwKygQ11EKWDZMpuft56y8x1mbPYkfkrvTrrUv1w3riI+nvaHiizRJTabYJSIicipM02TbwUIyv3uWkfv+jQ2T/X49KRjxEG27DqZXYACGYXDVxpeZlTjrqP0yDOCyokLsi/5G1YoX8Rj/MEbfK0FXVEVEROQkrN2Xy+LV67hxz12EV6RievpjdJ8MgW0hOIZpQR1YsvGfJOfv+s1ujBBfbXJtfi6+5hcMyvuCDK8YloVdwVz75Vw7vBN2m8YjIseiYpeIiDQ7GYXlfJ2YSo/ER0nI/QKArHbjaXvRP2gXGlvr3Gndp7EkdclRt/xu7x9LuLM7BUXvEFiaAZ/NpCo/DcfoPzXo5xEREZGmZ9nObNasWcHNe+8mqCoT0zcEY/xj0OdKsLn7hfoCs9v2Y07SHObtmEdWWRZhPmFMiZvCtG6/x3f3EljzX0hZTkTFPi7d/w++rcjgy4B7mdy3rbUfUKQRM8wWtD6wsLCQwMBACgoKCAgIsDqOiIjUsdLKahZty2TTjr1cue+vdC5Zj4lBdf/rcIx9APxCjv68qtKjDzK7T8NuePHDllS8fvo/RuV8iAsbxlUfYsSNb+BPJ42JxhQnT18rEWlpTNNk0bZMdq5bwvR99+HrLMQMaIsx4R/Q/YLTmyFekgNLHoN1bwLwZeQfCBl/D0M6H31sI9Jcney4QsUuERFpFiqqnbz0/W7M7J1cs/c+Qiv3Y3p4Y4y8F4bcesb9ttJySsh++xr65X9HpYc/nrcsg+CYugkvTY7GFCdPXysRaUlM0+TLTelkJ37F1fv+iqdZjhkSi3H+P6HTqDN/g8WPwk/PAvBp23voNfkOYsP9z/x1RZqIkx1XHH2vdRERkSZmwZZDtEpfxR92zSC0cj/4hWGc+xQMv7NOGstHh/hhXvBv9nvF4lldRPmbF0Fl6ZkHFxERkWbBNE0+Xn+AkrXvMW3vn/A0y6FNH4xL/1s3hS6AMX/DPOsmACYfeJZNX71KVlFF3by2SDOiYpeIiDR5yYeK2L59K9P2/RkfZyGExsHk/8CAa8BWd7sV9e8cRdKolym2BeBduJfy96dDy5kgLSIiIsexdEcWnute44q0R7HjhA7D4dI3IKpv3b2JYWCc9xSuPldhw2RyyuOs+OINSiur6+49RJoBFbtERKRJK62s5pN1KVye+hjerhII6wpT3oC4c+rl/cYMHsCSXk/jxIb3noVUL32qXt5HREREmo4dGUXs+vkbLjz4nPtA/Plw6Wvwm41x6oRhYJv8ElXxk7Hj4vwdf+WHr97D6dIFOJHDVOwSEZEmyzRNPt1wgP6ps+lYugnTwwdG3AORvertPT3sNhLOvZQl0bcBYPvhH5g7v6u39xOpSz/++CMXXHABUVFRGIbB/Pnzj3v+0qVLMQzjiNuhQ4caJrCISBOQW1LJvFU7uSTtSfeB2HFw4QsQEFV/b2qz4bj8DcpixuJBNWM33cWKxfPr7/1EmhgVu0REpMnakJZP/s5VjMt4AwBj8AzoeWm9v2+At4O4yfeyIWg8NlxUfXgd5KXU+/uKnKmSkhL69OnDSy+9dErPS05OJj09veYWHh5eTwlFRJqWymoX76xKYXjaq4RUHsD0CYFRfz7mDtB1yu6Bz+/ep7jNEDzNSs5aMZPNqxbX//uKNAEqdomISJOUX1rJgnW7mJr6S1+MmBEw9I9ga5gfbR1CW1F13vPuhvVVhZTPvkgN66XRO/fcc3n88ce5+OKLT+l54eHhREZG1txsDfTvTESkMTNNk/kbDmBP38Dw7A8AMIb8AdoNbLgQHp60uu5T8kP64W2WEb3oZgoKCxvu/UUaKY1URESkyTFNk4/W7md86nOEVu7H9A2FhAa6ivorZ3WJYtPwWe6G9QV7qPjw+gZ9f5GG0rdvX9q0acP48eNZvnz5cc+tqKigsLCw1k1EpDlauTuHTSlZXLr/CWy4IGYkDLoJDKNhgzh8CLzxc4odoQRVZ7H/k7827PuLNEIqdomISJOzfFcOPru/5qy8LzExMEbcBTHDGzyHYRiMHzqQ73o9hRMbXru+xrVjUYPnEKkvbdq04eWXX+bjjz/m448/Jjo6moSEBNavX3/M5zzxxBMEBgbW3KKjoxswsYhIw9ibXcJXm9MZlfUOkeV7wMsfRt4N3gGW5DG8Aygf+zgAcSnvkr4nyZIcIo2Fil0iItKkZBSWs2LDZi7Z724Ca/S8BAZeZ1keh93GmAlT2BhyHgAl3zwIpnZDkuYhPj6em2++mQEDBjB06FDeeOMNhg4dyr/+9a9jPuf++++noKCg5paWltaAiUVE6l9BWRXvrk4hrGwPYzLfch886yboOMrSXKGDryAzuD8Os4qyz+7E1HhEWjAVu0REpMmodrr44OcULk59HF9nIWbrTu4msA4fS3MF+jooH34fVYYD/7ytOLd8amkekfo0aNAgdu3adczHvby8CAgIqHUTEWkuqp0u3l2dSkl5JVekP4ndrHb36BpyS8MvX/wtw8Dn4udxYqNTwSr2Lp9nbR4RC6nYJSIiTcYPO7LovHsOXYrXYto9MUbeB2FxVscCYECvnqwLd+8EWf7tY+ByWZxIpH4kJibSpk0bq2OIiFhi4dYMUnNLGZX3MZFFW90X3Ibf0+B9Q4/Fv31vDsZeBUDAjw9TVVlhbSARi6jYJSIiTUJheRXJG1cy4dArABhn3QC9LrM41f94edixjbibcsMHv6I9VK+dbXUkkSMUFxeTmJhIYmIiAHv37iUxMZHU1FTAvQRx2rRpNec/99xzfPbZZ+zatYstW7Zwxx13sGTJEm655RYr4ouIWKqovIpVe3JoXXGA8emvug8OuBbiJlgb7DciLnqcEnsgoZX7SZn/mNVxRCyhYpeIiDQJi5MOMTHtX3iYVZjtBsGwO8HuYXWsWvp3i2VN1NUAVC59GpzVFicSqW3t2rX069ePfv36AXDXXXfRr18/HnzwQQDS09NrCl8AlZWV3H333fTq1YtRo0axceNGFi1axNixYy3JLyJipRW7c6h2urgq42lsznKI6Okej9ga16/Vnq2CyR3yFwDabXudwozUEzxDpPkxzBbUta6wsJDAwEAKCgrUP0JEpAnJLCzny/lzuW7PXZg2B8aUN6D7hVbHOqr1O1OJe38YrZyFVI1+CMeou6yOJPVAY4qTp6+ViDQHFdVOnvwmmR6H5nPpgSfB7gmTZ0HvKVZHOyrT5STz2aFElGxnb9hYOt7yidWRROrEyY4rGlcJWkRE5CgWbknnnPRfli92PQ/iz7U40bH17RzNmnbu3SGdK16EqnKLE4mIiMiZWpeSR3VFMRMzf1m+2Ocq6HGRpZmOx7DZ4fxnAeiYtZiMjd9anEikYanYJSIijdre7BKM5K9oV5aMy8Mb+l8DdofVsY7JZjMISZhJvkcY3hU5VC55wupIIiIicgZcLpPlu7IZnDMfv6o8aBUOw25rdO0Ufiui+3D2tnXPhDcW/AlT7RWkBVGxS0REGi3TNFm4+QDjD70GgK37RdApwdJMJ6NXTARrY25231nzXygvsDaQiIiInLatBwspKipkVNa77gO9pkJIZ2tDnaSQi/9Buc2X8LI97F/wT6vjiDQYFbtERKTR2nqwkJDd84mo2IfL0x8GTAeb3epYJ2QYBlEJ15Pp2Q7P6iIqFjxsdSQRERE5DaZp8uPOLAbnzKdV9S+zugZeZ3WskxYQ2pbU3rcDELL+BaqKcixOJNIwVOwSEZFGyekyWbQ5lXEZ/wXA1nMKRA+2ONXJ69Y2mA2xtwJg3/QuFGVYnEhERERO1b6cUjJy8v43q6vn5RDSydpQpyjm3DvJ8u6Ar7OQ9E8fsDqOSINQsUtERBqln/fm0jH1Y4KrDmH6BMPAaxvd1t7HYxgGnUddzQHvLni4yin/+i9WRxIREZFT9NOvZ3X5hcNZ11sd6ZR5enlRMOTPAISmfIVZWWpxIpH613R+axARkRajvMrJT1v3MSbzLQCM3ldAm94Wpzp1sREBbOp6BwCO7Z9B7j5L84iIiMjJyywqZ9eBLEZlzXUf6NX0ZnUd1n7IFAocYfg6Czm0eJbVcUTqnYpdIiLSKJRWlfLyxpcZ99E4Br3bn89zZjC3VTUlrSLcs7oMw+qIp6XH8IvZ49sbu1lF2Vf3Wx1HRERETtKyndkMyvnsV7O6mk6vrt/y9PTkYNzvAPDY/J7FaUTqn4pdIiJiudKqUqYvmM6sxFlklGZg4qLAVs6soECujYqgNCja6oinrX2oHzt73AaAx94lUJZncSIRERE5kaLyKrbsO/SrWV2XNZkdGI8lbOSNVONBWOlOCrd+Z3UckXqlYpeIiFhuTtIcknOTceGqddxlGCRXFTAnaY5FyepG97PPI9OzPQ5XOZXLXrI6joiIiJzAyt059M+aj391LmYT7dX1W6ERbdkTOQGAoh9etDiNSP1SsUtERCw3b8e8Iwpdh7kwmbdjXgMnqlvtWvuyve2lAFRtmgemaXEiERER+a3DLRXGfjSWu9eewz89PuXloADKelzc5Gd1HWY/ewYAEVnLqM7bb3EakfqjYpeIiFguqyzrjB5v7AzDwGvg76gyHPgV7YU931sdSURERH7l1y0VMkszAZMsDxuzggKZXpFMaVXz2MGwY+8RHPTthodZTda3z1gdR6TeqNglIiKWC/MJO6PHm4JeXWLYGjQagJIftZRRRESkMTluS4XCfU2+pcJhNptBXq9rAfDf9Tk4qy1OJFI/VOwSERHLTYmbgsHRd1u0YWNK3JQGTlT3fD09yI6/CgCvtJ+gVI3qRUREGovjt1RwNfmWCr/WbvjVFNsDaVWVQ96KN62OI1IvVOwSERHL/b7b7wnxiMFmmrX6WdmwEd86nmndp1mYru7E9BtLhmd7PFwVOJe/YHUcERER+UVzb6nwa4H+rdjX4XIAqtfMtjaMSD1RsUtERCyXWQjXlw9mZn4B4S4TGwYRvhHM7DuT2RNn4+vwtTpinegc7s+WNu5G9ZWbPlajehERkUaiJbRU+LVWw27EiY2wwi1U7P3Z6jgidU7FLhERsdyqPTmMyv6CGfmFLA4ZzcZpG1l02SJm9JnRbApd4G5U7xhwFVWGJz5F+2D3EqsjiYiICC2jpcKvdegUx+7WowDIX/ysxWlE6p6KXSIiYqnC8iryd64kumwbps0Del4CxtEHm81Bn9iObA50N6ovX6ZG9SIiIo3B5I5XEkZYs2+pcJhhGFT2vwGA1ge+xyzJsTiRSN1SsUtERCy1dl8ug7M/AcCIGQ4dhlucqH4F+jo41OUKABypP4EGlyIiIpbbnFbGkznezMwvIMJwYDNszbKlwq91PmsiGV4dcZgV5C76l9VxROqUh9UBRESk5XK5TDbt2M0t+b8s5+t6AXh4WhuqAbTvM4ZDm2KIrNiHa+UL2MY9bHUkERGRFqvK6WL79iRuLVzBQFzMSHgK+l5pdax65+PlQWLclXya9iIf5swnZ84XhPmEMSVuCtO6T2uWBT5pOTSzS0RELJOUXkjXg/NxmJWYrWPdSxhbgK5tAtgY4f6slRvVqF5ERMRKG9Py6ZXxCTZcmBE9odsFVkdqEKVVpTxl+5lZQYFk2Q1cpouM0gxmJc5i+oLplFaVWh1R5LSp2CUiIpZZvTuTwTnzATC6TQLf1tYGaiAedhu2PldQZXjiXZQKOxdZHUlERKRFMk2TVTvSOSvnCwCMrpPAq5XFqRrGnKQ57Cnches3vVJduEjOTWZO0hyLkomcORW7RETEEpmF5Th2f0dwVQYuT3/odbnVkRpUn7gObAocA0Dliv9YnEZERKRl2p1VQljaAlo58zF9WkOfK6yO1GDm7ZiHC9dRH3PhYt6OeQ2cSKTuqNglIiKWWLU3lyE5HwNg63IOhHezOFHDCvf3JrXTVADsqcugONviRCIiIi3Pit3ZnJ3zy0Y5cRMgOMbaQA0oqyzrjB4XacxU7BIRkQZXXuUkNXkDXYrXYmKD7heCreX9SGrfexTpXh2xuyoxV/zb6jgiIiItSnZxBQV71tGhdAum4QG9psBvlvQ1Z2E+YWf0uEhj1vJ+sxAREcttTMunf4b7KirtBkKXc6wNZJFOEZ48HdGVcdFR9D34KeM+GsfLG19WQ1gREZEGsHJ3Dmdn/zKrq/3ZEDPS4kQNa0rcFGzHKAnYsDElbkoDJxKpOyp2iYhIgzJNk7U7Uumf9zXwS2N6z5a3tXVpVSk3L7qeRfYkMjw8cBloByQREZEGUl7lZOvuFPrkf+c+0O0C8PC0NlQDm9Z9GvGt42sXvEwTA4hvHc+07tMsyyZyplTsEhGRBrUvp5R2qZ/j7SrF5d8Wel1mdSRLzEmaQ3JuMiZmrePaAUnq048//sgFF1xAVFQUhmEwf/78Ez5n6dKl9O/fHy8vL2JjY5k9e3a95xQRqW9r9+XRK+tLPM0KzKAOLXI84uvwZfbE2czsO5MI3wgMDCKcTm4ocjJ7wpv4OlrexUhpPlTsEhGRBrXqV41gbV3Ph4AoixNZQzsgiRVKSkro06cPL7300kmdv3fvXs4//3xGjx5NYmIid9xxBzfccAMLFy6s56QiIvXH5TJZsSuTs3M+BcCIPx/8QixOZQ1fhy8z+sxg0WWL+GTst3y9P4fbsg/gnbLK6mgiZ8TD6gAiItJylFRUU7JjKREV+3DZvbH1arm9ILQDkljh3HPP5dxzzz3p819++WU6duzIs88+C0C3bt1YtmwZ//rXv5gwYUJ9xRQRqVdJ6YWEZS4npPIApsMXo++VVkdqFDpGRbA9aDg98hZTsPItgruMtTqSyGnTzC4REWkwmw8UMPiXRrC2zqOh7QCLE1lHOyBJU7By5UrGjRtX69iECRNYuXLlMZ9TUVFBYWFhrZuISGOycncOQw43pu88FiJ6WpyocbDbDAo7TwbAe/9P4HJanEjk9KnYJSIiDSZ59266F/zkvtP1fLC33AnG2gFJmoJDhw4RERFR61hERASFhYWUlZUd9TlPPPEEgYGBNbfo6OiGiCoiclIOFZSTd3AncUW/FO17XAw2/Vp8WGj/8ymz+eFTmYtr25dWxxE5bfpXLSIiDSKvpJKQPV9gx0l16zjofqHVkSx1rB2QbKZJvG+kdkCSJuv++++noKCg5paWlmZ1JBGRGutS8hicMx8bJrTpA/ETrY7UqHSKDCE5OAGA4jXvWppF5Eyo2CUiIg0icX8+/fIXAOAROwa8Ay1OZK3f7oBkw0aw6WBmfgH/rQrWDkjSKERGRpKRkVHrWEZGBgEBAfj4+Bz1OV5eXgQEBNS6iYg0Bk6XyZaUdM7K/WXGUtcLwNPP2lCNjN1mUBTrXsrotX85VFdanEjk9KjYJSIi9c40TdK2r6Nt2Q5chgd0m2R1pEbh1zsgJU5L5HafW5iRX+jeAUl9MqQRGDJkCIsXL6517LvvvmPIkCEWJRIROX07MorokPk9vs5CTN9Q6HOF1ZEapYi+51BkD8KrugjX5o+sjiNyWlTsEhGRepdeUE77/V8AYLYdANGDLE7U+BiGgW+38ZTZ/HBU5MH2r62OJM1QcXExiYmJJCYmArB3714SExNJTU0F3EsQp0373xLaGTNmsGfPHu677z62b9/Of/7zHz788EPuvPNOK+KLiJyR9al5DMz7CgAjdhwEtrM4UePUOSKIba3dOzGWrFexS5omFbtERKTeJabm0C//WwDssWPAw8viRI1T93ahbAsYAYBr4/sWp5HmaO3atfTr149+/foBcNddd9GvXz8efPBBANLT02sKXwAdO3bkq6++4rvvvqNPnz48++yzvP7660yYMMGS/CIip6u0sppD+7YTW7wOEwO6XQiGYXWsRsluMyjp8suujAdXQUWxxYlETl3L3QZLREQahMtlUpC0mMCqLKodrfDoPtnqSI1W+9a+rA4fT//8Bbj2LcfmcoLNbnUsaUYSEhIwTfOYj8+ePfuoz9mwYUM9phIRqX+Jafn0zXHPmjba9IbOoy1O1LhF9Uog/+cwgqqzcCW+i23wTVZHEjklmtklIiL1ak92CfEZ7sGlLWY4hMZbnKjxMgwD765jKbP54aGljCIiInUmMSWHAXm//FyNPQc8tRHM8XQO9ycp5BwAyjZ+anEakVPXpIpdP/74IxdccAFRUVEYhsH8+fOtjiQiIiewZe9BehT8AIAtdizYmtSPngbXvV1YzVJGc5OWMoqIiJypjMJyvNN+IqgqE9OzFfS61OpIjZ6H3UZJ3C9LGQ+thdJcixOJnJom9RtHSUkJffr04aWXXrI6ioiInIQqpwtX0hd4ucqo8mvj7o8hx9Ux1I8doe6msM59K7Qro4iIyBlan5LHwNxfGtN3HKlZ5icpuvvZZHm2w+6qxFz3ltVxRE5Jk+rZde6553Luueee9PkVFRVUVFTU3C8sLKyPWNKClFaVMidpDvN2zCOrLIswnzCmxE1hWvdp+Do0FVrkt5IPFdEr5xsAPGITICDS2kBNgN1m4Bk3jvI9vniX50LyN9BtktWxREREmiSXy2T7nhTGF/7kPhA3UbPMT1JsuD/LQs4hIf0NyrZ8ge8I7cQrTUez/lf+xBNPEBgYWHOLjo62OpI0YaVVpUxfMJ1ZibPIKM3AZbrIKM1gVuIspi+YTmlVqdURRRqd7TuT6Vy8DgCjy0SL0zQd3aJDSTq8lFG7MoqIiJy2HZlFxGZ8jYdZhRncEbprlvnJ8rDbKIu/CADvzEQoTLc0j8ipaNbFrvvvv5+CgoKaW1pamtWRpAmbkzSH5NxkXLhqHXfhYntuMs+veZ2dGUVsSy9ky4ECsosrjvFKIi1DWaWTVjs+xYaLypBu0GWs1ZGajNjwVmxvfXgp43ItZRQRETlN6/blMjD3SwCMLuPBJ9jiRE1LTHw/DnrHYjOdmGtetzqOyElrUssYT5WXlxdeXl5Wx5BmYt6OeUcUug4zcfHJro8pPPS/LYwNA/pGBzGuWwSt/TwbKqZIo7F5fz59cxcA4NllLHj5W5yo6XDYbdi7jKV83y9LGXcsgK7nWx1LRESkSSmtrKZwz1ralO/GtDkwel5idaQmp0t4K75vfQ5RB3dRkbQA77F/szqSyElp1jO7ROpKQWkVmaVZxz2nwsyjTaA37YJ9aBfsg2nChtR8nv02mc8SD1BQVtVAaUUah7RtPxNZvgeX4VDPqdOgpYwiIiJnZmNaAf1yvgDAaH82tB1ocaKmx8NuoyLevSujV04SFBywOJHIyWnWM7tEzlR6QRk/7chm4/58vIwgys1jb7kb7hvObWO71NxPyy3lu6QMdmYWs2pPLutS8hjSKYSRcWH4eemfnjRvBaVVROz9FIDqtmfh2XaAxYmanrgIf+YFj6F//kJce5dhd7nUUFdEROQUbNybzjV5i9x3upwDdoe1gZqoTl26k76iE20q9mBufA9j5D1WRxI5oSY1ai4uLiYxMZHExEQA9u7dS2JiIqmpqdYGk2aloLSKdSm5vLFsL/9evIsNafm4TOgbNBHjGP9kbNiYEjel1rHo1r5cN7wjN47oSIcQX6qcJj/uzObphcksSsqgrFI9eKT5SkzNpk/+dwB4xo0FDy3lPVXeDjvEjqHc5ou9PBd2fGN1JBERkSYjo7Cc4JSF+LiKcfmGQa8pJ36SHFWXiFbsCHLPNi9LXmJxGpGT06Sml6xdu5bRo//XE+muu+4C4JprrmH27NkWpZKmrrLaxd7sEnZmFrEzo5jMov81ljcM6NU2kBFdQmnd6namL9hwRJN6GzbiW8czrfu0o75+p7BW3DzSjx0ZxXy79RAHC8pZvD2TZbuyGdyxNUNjQwn00VUmaV7yNy/EvzqXKkcAju6TrY7TZHVrF8q2gOH0y/8WNr6vvl0iQmlVKXOS5jBvxzyyyrII8wljStwUpnWfhq/D1+p4Io3G+pQ8BuR9BYAtdiz4t7E4UdPlsNso7TgeMt7C89B6qCoDh4/VsUSOq0kVuxISEjBN0+oY0gyYpsmqPblsPVhASk4p1a7/fV8ZBkQH+9IlvBX9OwTXai4/e+Js9wAz+SOyyrIJ827NlM6T3QNMj2P/h28YBvGR/sRFtGLLgUIWb88go7CCH3dms3x3Nn2jgxnRJZSIAO96/dwiDeFQQTkdDrj7Y9BxJLTubG2gJqxbmwA+DhpDv/xvce5brqWMIi1caVUp0xdMr3XhLaM0g1mJs1iSuoTZE2er4CUCuFwm+3Zu5dzidZgYGN0udA/y5bSFdx1C0Zpg/J15kPQZ9LnC6kgix9Wkil0idcE0TT5LPMjqvf/rvxXo4yAuohVxEf50CvPD1/M3/zQKDkDaanz3r2FG2mpmHNoMzkpgH2xbD18+5D7P7gC7F3h4QWgc9JwCPS4Cv1DAXfTq1S6Qnm0DSM4o4qcd2ezJLmFdSh7rUvLo1safEV3CiAnxxdAPZGmiNu9JI6HgRwAcXcaqOHMGfD09qI5JoCLFB6+yHNjzPcSOtTqWiFhkTtKcI2aYA7hwsT03mX+seIUhra/gUGE52cUVtPJyEBnoRWSAD5GB3kQGeOPjabcovUjD2ZlZTNwh94U3M7I3RufRJ3iGnEh8myC2BwxlYN5XVGz5Ai8Vu6SRU7FLWhTTNPlmyyFW783FMGB8twh6tg0ktJVn7eJSWZ57yVDqKkj7GYoOHvliNg9wOYFfzTZ0VrlvlcWQutJ9++Y+6DAUek9170jnE4xhGHSNDKBrZABpuaX8uDOLrQcL2ZZexLb0IjqE+HLloPZa3ihNjmmaVG2ej8OspNyvHd7dLrA6UpPXNTqcnf4D6Vn4E2z9RMUukRZs3o55RxS6DjNx8U3KfKpyxtUcyy2pIjW3FMirORbk6yAywJu2QT4M6RyiTXOkWdqQks3EvK8BsHU5Bzw14/FMtfLyILvtWMj7CjPtZzBNzZaTRk0/3aRFWbI9k592ZgNwSb+2DIxpXfsE04Stn8I3f4KSzP8dN2wQGO2erRXeHdqfDW36gs0OzgqoqoDqcqgug6oSqCiGfT/Bzu8gd4/7z/t+gi/vgE4J0Oty6D4ZHN5Et/bl6sEdyC6uYNnObNal5JGSU8o7q1K4aWQnHHbNipGmY19OKfEZ7sGlo8toaBVmcaKmr3tUAIv8h9Kz8CeqU9foB7dIC5ZVmnXcxyvMPIZ0DiHC34swfy+KK6pJLygno7Cc9IJy8kuram7bDxWxJiWXywZEExveqoE+gUj9q6h2UrljCUFVmbgcrbD1utTqSM1Gq+7jqNrqwLs8E1JWQMwwqyOJHJPGzNJiLNuZzaJt7gLWpN5tjix05afB1/fAjgXu+wFR0HE0RPaETqMgsB14tnIXuE5GtwtgogkHN8D6t2HXd1CQBrsWuW/f/x0mvwQd3TubhLby4qJ+bRneJZRZS3ezP6+Mj9ftZ+pZ0VrSKE3G9p07mVCyAQB73ASL0zQPgT4OitsnwIEnsefscP9fFRRtdSwRaSBOl8nWgwWs2J2DpxFEuZl7zHPDfcO5sE9UrWO92/3vz2WVTg4VlpNeUMaqPblkFVXwxvK9jOwSyvjukdhtGm9I07fjUDF9s78EwOg4AkLjLU7UfMRHR7K71QC6Fq2iauOHOFTskkZMxS5pEdbsy+WrzekAjO8ezrDY0P896HLCz6/BkkehssRdzOpxKYy4G8Liz2x6rmFA2/7um2lC2mpYPwe2fwX5KfDWJOh7NUz4P/AJAtxFr6sHt+e/y/aycX8BEYHejI4PP4NPL9Iwqp0ubEmfYsOkLDgen84JVkdqNtp3iOXghs5Ele+GzR/BiLusjiQi9ay0sppVe3JYvTeXwrJqADp4jmFHxceYHLlhkw0bU+KmHPc1fTztdAz1o2OoHwM6BPP15nR+3pvHDzuy2Z1VwtSzoglt5VUvn0ekoexISWNy4U8AGHET1Tu0DoX5e7ExbBRdi1ZRuXclargijZn+5UuztzEtn083HABgZJfQ2oWjQ1vgv+NhwZ/cha7QeLjgBZj8IoR3rdt16IbhXv540X/gtkT3MkaAxLnwQn9I+rzm1E5hrWquzH6XlEHSwcK6yyFST3ZmFtMt51sAvGJHgZe/xYmaj25tAkj2HwqAc/dSa8OISL3bcqCAf323g++SMiksq8bf24OxXcN58fw76Nq6K7bfDOFt2IhvHc+07tNO+j28POxc3K8dVw9uj4/Dzv68Ml5csov1qXna/VyarCqnC0fyF3iYVVT6t3evtJA6ZY8/FwDf/B2Ql2pxGpFjU7FLmrWkg4V8uDYN04TBHVszsWeke0mgswoWPwqvjoID68DhA4Nuht9/Cv2udu+mWJ98g+HyOfC7TyCgLZTmwIe/h3enQqF7BtrgTiGc3ak1pgkfrk0jo7C8fjOJnKFdyZtpX5qEiQ1bvJYw1qWIAC8Ohg0HwDywDpzVFicSkfpQXFHNu6tTmbs6leIKJ+H+Xkw9K5r7JsQzrnsEkf6BzJ44m5l9ZxLhG4HNsBHhG8HMvjOZPXE2vo5Tb8Lds20gt4/tQsdQXyqqXXy0dj8frk2jvMpZD59QpH7tyiymZ467JYkjNgH8QqwN1AzFdI7jgHcXDExciXOtjiNyTFrGKM3Wrsxi3vs5FZcJ/aKDmNw3yl3oKi+Aj6bD7iXuE9sNguF3QZfxYG/gfxKxY+GP62DhA7Butrtf2IsD4JzHYcC1TOodRWZhBXuyS5izch+3jI7F11P/bKXxKa9y4p38GQAVYT3xbj/U4kTNi2EYBHYZSun2VvhWFbv/r+g2yepYIlJHTNNk84ECPk88SEmlE5sBo+LCGNM1HI/fbFTj6/BlRp8ZzOgz4/CT3WObvBQoPAhF6e5bcSb4hbtbMoR3g+COxxznBPo6uGF4J37YkcWibRkkphWQmlvKFWe1J7q1drGTpmP3ru1MKkkEwOh6nrVhmqkOrX35KXgkbdN3UrZjKX6j77c6kshR6bdmaZayiip4Z1UK1S6THlEBTBnQzl3oyk+FuZdD1jawe8HQ22DIH8C39YlftL44fGDSP6Hf7+CTmyBnJ3x5J6StwT75Ja4a3J7/LN1FZnERt3/zNHsqF5NdlkWYTxhT4qYwrfu007qSK1KXtqUX0jNvEQBenUdqi+960LVta3b6D6ZPwWLMpM8wVOwSaRaKyqv4LPEgW39pWdAm0JspA9oRFeRz7Ce5nPDdg5D8tbuwVVV24jeyOyAkFsK6uXeWjuwJncfUzGa32QxGdw2nU5gfH6xJI7ekipd/2M2EHpGM6BKqzXKk0XO6THy2fwJAWUgPfDqOsjhR82SzGVR2PgfS/4tXRqK7FYynn9WxRI6gYpc0OxXVTuauTqGi2kVMiC9XnBWNzWa4lyu+ewWUZIJPMIx5EPpPa/jZXMfStj/cshq+/z/46VnY+C5UFOJ32WwuOyuc3399H/nOffBLU9qM0gxmJc5iSeqS0166IFJX9m1bS7/y3bgMD2zdzrc6TrMUE+LL50FD6VOwmKrUtXhaHUhEzohpmiSm5fPlpnRKf5nNNaZrOKPiwo6YzXWEpf+AlS/WPubpBz6t3WMc39bgHQzlef+b8eWsgMxt7ttWd0EAv1AY/Ac46zr384AOIX78cUwXPt1wgM0HCvhmyyF2ZxVz2cBoWnk1kjGTyFHszSqiZ/Y3AHh1Ga0Lb/WoTdezKVgVQmB1DuaWjzH6n3y/QJGGop9Y0qyYpslnGw6SUViBv7cHVw5u7x4wbvsCPr4RqssgqD2Mfxy6X1i3Dejrgs0OY/8GIZ3hs1th+5cwdwoLeo6nwJkCv9l9yYWL5Nxk5iTN+d9yBpEGVlxRTdDuLwCoiuyHV9sBFidqnjzsNogdiyvlcTwL9kD2LgiNtTqWiJyGaqeLzxIPsjYlD4CoQG+mDGxHm8DjzOY6LHkB/PiU+88Db4C4cyA4BvzCwO7pnqll/9UeaS4XVJVA1nY4sB4ytrp3hE7fCCXZ7t2of3zaPcN86K0QHIOPp50rB0UTu68VX246yI6MYv69eCcX9G3N8qxPmbdjHlmaZS6NTNq2nxldsQ+n4cDeY7LVcZq1LpH+bAwczqCczyjb8jW+KnZJI6RilzQrq/fmsiEtH5sBVw5qT4CXByz/t3uqPya06QfnPQXRg6yOenx9rwKvAJh3LexZyjzXPkzDddRTXbiYt2Oeil1imU1pefTO/2UJY+yo+t/goQXrFNORAz7xRJdth80fgfpkiDQ57ib0KezNLsUwYHy3CEbGhWG3ncQFuNy98OlN7j/HTXRfIPMJOv5zbDb37rjtznLfDqsqgzWvw8+vQ/4+WPMarP0vdD0fht+J0XYAgzq2pkOIL+/9nMqBggJmLr6PQlcKpmaZSyPjcpn4bJsHQGnEQPyj+lmcqHnz8rBTGD0Wcj7DdmCNu6hu09530rjoO1KajbTcUr7cdBCACT0i6Rjs5e599d3fABO6TITL3mz8ha7Duk2Cqz4CDx+yOP6OSFllWQ0USuRIB5NWElJ5AKfNS1t817P4CH92BLib/1fu/sHiNCJyqjIKy/nP97vYm12Kl4eN6UNjGN01/OQKXVVl7p2bywsgpAuMf/TEha7jcfjA0D/CbRtgypvuC4Kmyz0b/rUx8OZ5kJ9GRIA3f0iIpdLvewp+Veg67NezzEWskpJdRPecbwHw7Tq69uxGqRcB3cdRaXjhXZEN+360Oo7IEVTskmahpKKauatTcbqgR1QAI2L84P0rYd2bgAEDroOL/gOtO1od9dR0ToBrviDMaR73tDCfsIbJI/IbuSWVRKZ+CYCr7VkQ0dPiRM2bj6edwugxANjSE6GqwtpAInLSth8qZNbS3eSVVhHi58kfEjoTF+F/ck82Tfjqbji02T3ze+yD7l0W64LNBj0vgZuXwnUL3U3rDTukLIf/nA3bvsLTw8bWom/5bTuFww7PMhexSsambwmozqHC3gp7z0usjtMixEeHs6vVQAAqEj+yOI3IkVTskibP5TJ5f00aBWVVhLbyZErvEIz3r4Kd37p3XEz4M0z4O/iFWB319ESfxZSuU7GZRx9gGhhMiZvSwKFE3Dam5tIrfzEAjthRjWfDh2YsLG4wxfZAPJxl7r5+ItKomabJjzuymLPSvXlOp1A//jC6M+EB3if/IutmQ+JcMGww4l7oWk+7sbY/G37/Kcxc4Z49VlkMH1wFX91DVunxZ5FrlrlYxTRN/H7ZhbEkaii07mxxopYh0MfBoTajAajet8riNCJHUrFLmpzSqlJe3vgy4z4aR585fRj5/hi+TpuDzVbB7waE4T3vatjzPXh4w7iHYMQ9TX43lmmD7iE+MNZd8KpV9DLwt3Wgu189DXpFjsM0TbK2fk9gdTbVHr5awthAukUFsiNgCABVSSp2iTRm1U4XH68/wDdbDmGaMKhjMNcOi8HX8xQuDBxYB9/c5/5z36tg8I313xsnvKu74NX3avf9Na8R5tIsc2mcDmTmEJe7BICAbmPVO6oBeXY7FwC/wl3ujXNEGhH9TyBNSmlVKdMXTGdW4iwySjNwmS4KqrLZUfkxm6r/jv+X02DPUneha+xDMOjmZjHTxNfhy+xJ7zKz+zVEuMBmmkS4DM5v+zuG+j7It1vySM0ptTqmtDDpBeV0OOje4pvosyE0ztpALURIKy/Sw0YAUJ221uI00hS99NJLxMTE4O3tzeDBg/n555+Pee7s2bMxDKPWzdv7FGYktWDlVU7eWL6XdSl5GAZc0LsNF/Vt695Z9WSV5MAH08BZ6W4uP/pv7l5bDcHD090C4tL/gqcfUwryNMtcGqXs9fPxcpVR7BmGR6+LrY7TonTu1Jk0n64AVCe+Z3EakdpU7JImZU7SHJJzk3Hx250JTVJLdjEnb+Mvha6HYXDzKHQd5uvwZcage1l0wSdsPJDNopQUnijYRd924ThdMHd1CoXlVVbHlBZkU0oWPQuWAuAROxpsdmsDtSBeXcfjwoZPcSpkbLU6jjQhH3zwAXfddRcPPfQQ69evp0+fPkyYMIHMzMxjPicgIID09PSaW0pKSgMmbpqqnC7eXplSqxH90NhQDOMkGtEf5nLCx9dD4X7wbwPjH4eAyPoLfSy9psCMFUzziCS+svIoBS+DAFsHEiIvbfhs0uKZpon/jl+WMEaPglbhFidqWSIDvNkbMgqAsp3aOEcaFxW7pEmZt2PeUQpdbi7TZJ6/P4x7BAbf1Hx/8Q6Lg0v+C4Cx+QMud35FuL8XheXVvLs6lWrn0b8+InXJ5TIpSPoOP2cB1Z6B7t1DpcF06dCOFN8eALg2qSmsnLx//vOf3HjjjVx77bV0796dl19+GV9fX954441jPscwDCIjI2tuERERx32PiooKCgsLa91aEqfL5N3VqezJLsHLw8b1wzuefCP6w/JS4OMb3G0Z7F4w+gHocHb9BD4ZrWPwnbGc2RFjmZlfQER1NTYTInzCGBF6FUN8H+STddkUlOqimzSszEP7iclbCUBQzwlwKgVlOWOGYeCKmwiAT9ZGKG9Z/99L46ZilzQpx21+ahhkeXjAoBubb6HrsG7nw0h3/w7Hor9xbbt0vB02UnJK+WpzusXhpCXYl1NCl4yFANhihkJwE9vptImLDvZlX/AwAMp2LbM4jTQVlZWVrFu3jnHjxtUcs9lsjBs3jpUrVx7zecXFxXTo0IHo6GgmT57M1q3Hn034xBNPEBgYWHOLjo6us8/Q2LlcJh+uTWP7oSIcdoNrhsYQ3foU+oYWprt3XXxhAGx1z1ZhyC3Q58r6CXwqPDzxnTyLGeP/zaJD+Wzcl8qinAr+lXAz7QKDKCqv5p3VKVTpops0oLyfP8COi2zfznh1m2B1nBapXfxA8j3C8HBVYm7WrqzSeKjYJU3KiZqfhvmGNf9C12EJ90OX8eCqJujz6VzVzYFhwKo9uaxLybM6nTRzm/YdokfhjwDYOo/WldQGZrMZVHceD4BX1iaoKLY4kTQF2dnZOJ3OI2ZmRUREcOjQoaM+Jz4+njfeeIPPPvuMd955B5fLxdChQ9m/f/8x3+f++++noKCg5paWllann6OxMk2T+YkH2LS/ALsNrh7cgY6hfif35JJsWPgA/LsvrHkdXFUQ0RPOfRpG3tO42jL0uBiu+RI8/SBjC15vnc+0/sH4etrZn1fGpxsOYB6jt5dIXQvc9SkAZdGjwDvQ4jQtU0xoK3YEjwSgeOtCi9OI/I+KXdKkJERdCBz9l2obNqbEXdawgaxks8GUN90zaspy6fLtNMZ1CQLg683plFU6rc0nzVa100XltoV4ucqo8gmFrlrCaIWo+IEUeITg4arATJpvdRxppoYMGcK0adPo27cvo0aN4pNPPiEsLIxXXnnlmM/x8vIiICCg1q25+e3O0OM+Gsfd3z3Dyr3pGAZMHdie+MiTWLpYlgeLH4PnesHKF6G6HELj4Zy/w/Qv3W0ZPE+yYNaQ2g2Aa74Cz1aQuZXg9yZxdZ9AbAZsSM1n+a4cqxNKC5Cbuo02RVtwYiO0v8YiVvGw2yjrMAYAe/r63+wcL2IdFbukycgqqqAiZxgBtg4YJrX+I7VhI751PNO6T7MuoBW8/OF3H7sHwpnbGJ30V8JbeVJa6eSHHcduNixyJpIziuiW8y0AHh1HQECUxYlaptgIf3YGDAWgdOsCi9NIUxAaGordbicjI6PW8YyMDCIjT67xucPhoF+/fuza1XK3mD/aztAZpRl8l/42K0of5fzewfRqd4IZJqYJG96B5/vAT89AVan74tWYh2D61zD0VvAJbpgPdLra9nNn9fKHrG10+vxiLohzL9n8eks6uzI141TqV8HPcwFI9++NT6dhFqdp2QK7JVCNB74VmXBwg9VxRAAVu6SJKK9y8s6qFKqrPXnM7MMf8vOJcDqxYRDhG87MvjOZPXE2vo5T6IvRXIR0himzAQMjaT5XuL4AYPmuHPJKKi2NJs3Ttr0H6Fq4AgCjU4KWMFrEy8NOYfRoAIwDay1OI02Bp6cnAwYMYPHixTXHXC4XixcvZsiQISf1Gk6nk82bN9OmTZv6itnoHW9n6CJXCpuLvzz+CxRnwvtXwWe3QHkBBLSDkX+GaxfAyLugVWi9Za9zUX1+KXgFQFYyg5dMZXCkgWnCvHX7Ka/SLHOpJ6ZJ61+WMFZ0SACHt7V5WrhObSNI8esJQPnGTyxOI+KmYpc0eqbpbvaaWVTBwLJljN7zIjPyC1kUMZGNV61m0WWLmdFnRsssdB0Wd457pyYgcvX/McS2jWqXyXfbMk7wRJFTU+V0YdvxNQ6zkqpWbaHr+VZHatECuo11X0ktS4f966yOI03AXXfdxWuvvcZbb73Ftm3bmDlzJiUlJVx77bUATJs2jfvvv7/m/EcffZRvv/2WPXv2sH79en73u9+RkpLCDTfcYNVHsNzxdoY2MZm34zgNmrd9Af85G5K/BpsH9P0dTP8KRv8ZAk5udl2j06Y3XPsNeAdi5OzkwrXX0NazlIKyKr7cpE1zpH4U7lpJYPl+KgwvIs662Oo4LV6At4OMMPfsuop9qy1OI+KmYpc0eou3ZbItvYgO5du5eO/DGJgQdx6M/is4fKyO13iMvAfiz8MwnZyXfD9ezhIS0/I5mF9mdTJpRnZkFNE91z0rxKPjcGh1/E0jpH516dCWfX69ASjfqB2Q5MSmTp3KM888w4MPPkjfvn1JTExkwYIFNU3rU1NTSU//X4EiLy+PG2+8kW7dunHeeedRWFjIihUr6N69u1UfwXLH3Rn6WI+XF8CnM+CD30FpDgS2h/P/CZP+Ca1jmv4M2cie7plp3kHYcndx084Z+DnzWZeSx7b0QqvTSTNUvOZdAFIDB9EqupfFaQTA1ck929w3ezNUlVucRkTFLmnkthwoYPH2TIIqD3F92v3Yqsshqj9M/D/wC7E6XuNiGHDJaxDQFo+ybH6f/Rym6W5Wr12RpK4k70sltvhnAIzY0RankUAfBxkR7h2QKvassDiNNBW33norKSkpVFRUsHr1agYPHlzz2NKlS5k9e3bN/X/961815x46dIivvvqKfv36WZC68TjhztC/fXzPD/CfIbDxPffP6u4XwdUfwYBrwMOr/oI2tIjucN1C8AnGs2APf0i7D7urkk83HKC0strqdNKcOKsI3vM5AFUxCS1nJ/ZGLiJuECX2AByuMszkb6yOI6JilzReGYXlzFu3Hy9nMTMP3o+jLAuCOsC5T0LrjlbHa5y8WsElrwLQOeMbuhYuZ3dWCTvVJFbqQJXThS35GzzMair920OXCVZHEsAeNx4Av9wkqNC/dZH6dmGnSzj+ztBT3HeqK2HB/TDnQig8AK0i4Jwn4KJZEN614QI3pPCu7hleDj9aFyZxxcEnKCqr4ouNB61OJs1I3uZv8KkuoMgeSIchWsLYWMSE+bPb/ywASreq2CXWU7FLGqXDDemrqiq5Pv1RAgp3unclGv8oRA+yOl7jFjMcBl4PwOUHn8LLWcTXm9NxuTS7S87MzoxiuuUuAsDRcRj4trY4kQC0i+tHob01HmYlru1fWR1HpFmrcrqgcCQBtg78tuBVa2fowoMw+zxY9R/3g7Hj4Mr34ewZ4NnMe4yGd4XL54Bho2fud4zMmENiWgFbDhRYnUyaibK17wGwP2Q4fuGdLE4jhznsNvKjRgDgOrDe4jQiKnZJI2SaJh+tTSO7qIIpGc8TnbsC7F4w9kHoPtnqeE3DOY9BQDt8KnO4cP8/ySisYH1qntWppIlL3ruP2CL3rn9Gl7EWp5HDooJ82RfovpJalLTI4jQizZdpmny8bj/peSZjAx9herebiPCNwGbYiPCN+N/O0PvXwSsjYf8acPhCwl9gypvQtn/T7811srqMg3MeB2Bi5ut0zf+RzxIPUFyh5YxyZlxlBYQdcP+s84wb03L+TTURnnHu8aFf4W530V/EQip2SaOzNDmLpPQihud+TL/MTwHD3Xy93zT9QDtZnn5w6WuAQf+C74jL/4nvtmVQUa0twOX0VDld2JO/wo6TyoAY9ywFaRRsNoOStsPddw4mWppFpDlbsj2TjfsLsBkw7ew47h50K4suW8TGaRtZdNkiZvS+Gd+fX4c5k6EkC4Law6TnYMTd4B1gdfyGd/YfoO/VGJhcuf9R/PK381niAfURlTOS+fM8HGYl2Y4oOgy+yOo48hsdOnYhw6sDNlw4N39idRxp4VTskkYl+VAR323LoHPRWs47+IL74IBrYegfwe5hbbimpsNQGHQjAFMOPk1lcS4rduVYHEqaKvcSRvcujI5Ow93LiqXR8I5zbxbgX7QLig5ZnEak+UlMy2fRtkwALurXltjwVrVPqCiCj66B7/4GphNiRsDlc6HP1JY7fjEMuOB5iB6Mp6uca/bex96UfWzar+WMcgY2fQBAVptRePiHWhxGfisywJuUoCEAlOz4weI00tKp2CWNRk5xBR+sSSOo/ADT9j+EYTqh02gY/QA4fKyO1zSNewQC2+Nfncv5qc/yw44sLSGQ07Jzz146Fbv7LxhdNKursYnuGEemZzQ2XFRt/szqOCLNSkpOCR+v2w/AyC6hnBXzm36FWTvgtTGQ9Jl7V7izboQpsyGqd8OHbWzsDnevsoC2BFdn8bs99/Hl+n0UlldZnUyaoPKcNMJz3DtCB/WcaHEaORrDMChvPwoAx6EN4HJZnEhaMhW7pFGorHYxd3UqzvIirt//AJ5VBRASC+f8HVrpqs1p8/SFS1/DxGBg0WJispeyeFuG1amkiXHvwvgldpxUBHaCzmOsjiS/EeLnyf7gwQCU7FhqbRiRZiS3pJK3V6ZQ7TLpHhXAxJ6RtU/Y9gW8lgDZO9wzXs/5P3evKo1d/se3Nfx+PqbDj5jy7UzY/Rjz1+/XckY5ZZkr5mLD5IBPVyJ7q3doYxXQdRTVhgc+FVlwYJ3VcaQFU7FLLGeaJp9u2M+h/FKuPPB3Qkp2gXcQjH0YIntYHa/pa382xuCbALj4wNNs2rmPrKIKi0NJU7Izo5huee4ljJ6dR4B3oMWJ5LcMw6AieiQAHhkbLU4j0jyUVFTz5vK9lFQ6aRvkzeUD22Ec7h1qmvDTs/DB76CyBMK6wcWvwaCbwOFtbfDGKCwO4/I5mIaNgYWLCE18ia0HC61OJU2Mb/LHAJREj8Lw9rc4jRxLp7bh7PN1z2yt2PypxWmkJVOxSyy3cncOiWkFjMl8i675P4DNw710sdskq6M1H+MegaD2BDrzOC/1Gb5NUk8fOXk79+ymU/EGAAw1pm+0WnVNwIWNVmUHIHO71XFEmrTKahdzVqaQXVxJkK+D3w+JwcvD7n6wugLmz4TFj7rvx02EK9+FuPFg09D6mLqMwxj/GODeoXHLT/OpdmqJk5ycvL0bCC3eQTV2IgdeaHUcOY4AbwcZYUMBKN/7s8VppCXTT2Sx1N7sEr7anE73gh8Zl/Ff98Gz/wADr9XOi3XJ4QOXvI6JwYCiJVRu/YrUnFKrU0kTUOV04ZH8BTZcVATFQqcEqyPJMXSMbssBn3gAyjfpSqrI6XK5TD5Ym0Zqbik+DjvXDo0h0MfhfrAkx73b4sb3wLC5+3Nd/DK07mRt6KZiyC04e16GDZOJOx9hddJeqxNJE1Gwei4ABwL7E9BpkMVp5ETMX1pe+OZshqoyi9NIS6Vil1imqLyK935OJaxsD1fsf9x9sOskGHmPu6Gp1K32gzEGuZcznn/wBRZuTlW/DDmhX+/C6F7CGGBxIjmWVl4eHAo9G4CyPSstTiPSNJmmyRebDpJ0sBAPm8Hvh3QgPOCXZYlZyfD6GEhd6b6INPpvcM5j2p32VBgG9gueo8I3iiBnDn7f3aONc+SETJeT1nvcm6+4Oo0GD0+LE8mJRHQZSLE9EIerHHP711bHkRZKxS6xhMtl8sGaNKqLc5meej8OZylE9oLxj6ofUH0a81dc3kGEVx8kMulNkjOKrE4kjdyu3TuJKXH3gDK6nGNxGjkRZ4x7BySfrI3aAUnkNPy4M5tVe3IxDJh6VjQdQ/3cD+xaDK+Pg7x94BcO5z4Dw27TbtGnw6sVnle8iQsbfQuXkLTwv1YnkkYufdNiAiozKbf50nbQJVbHkZMQE+bP7gD3DLzSrQssTiMtlYpdYokl2zPZm1nA1WkPEVR+AFpFwPjHIKSz1dGaN+8AbGMfAmBs9tt8v347Lpdmd8nRVTtdOJI/x4ZJeXA8dBxhdSQ5gdbxw6k0vPCuysdMWWF1HJEmZUNqHgu2uHtant+rDT3b/nLxbc3rMPcyqCiE0Hi45BXodzXYPSxM27QZ7c+moO/NAPRIfIysgykWJ5LGrGL9+wCkhwzBMyLe4jRyMhx2G/lt3ONG18H1FqeRlkrFLmlwuzKLWZKcyfhDr9O5aA3YvdwN6TuPtjpay9B/Gq6QLvi6Suid/G82pOVbnUgaqZ2Z/1vC6NV5OHhp56PGrkNka1L83DsglW7+wuI0Ik3HrsxiPl6/H4DhsaEMiw11z45c8Bf46m4wnRAzAi5/CzqPUV/ROhA86VFy/WLxcxVR/sF17h0uRX6joryENvvdM4O848doE4gmxCtuLACtCndD/n6L00hLpP8tpEEVlVfx4do0uuX/SELWO+6DQ2+Dfr+zNlhLYvfAdt7TAJxd8DXr1qygSrshyVHs2pVMTOlmAIy4CRankZPh5WEnJ8K9A1Jl2jqL04g0DekFZbyzKgWnC3q3C+S8XpFQVQ7zroVVL7lP6nMlXPoGhHezNmxz4uGJMeW/VONBdMFaMr79l9WJpBE6uHo+3q4SCj1CiDzrYqvjyCno0LEzh7w6YmDi3PKJ1XGkBVKxSxrM4T5dXgV7uHz/390Hu14Aw28Hm93acC1N59G4Oo/FjpMRu59l5e4cqxNJI1FaVcrLG19m7EdjeWD3dYyLjuLFNp0pbdvf6mhykmy/7JjZKmeL+xd2ETmmg/llzF6+j4pqF51C/bhsQDuMsjx4+yJImu8enwy7A857GvzDLU7b/AR37MuOHne4/7zqH7gyd1obSBod25YPAchtMxIjIMriNHIqIgO8SQl2b5xTsuMHi9NIS6RilzSYJdszSc3I5vcpD+DlLIGwbjDuYS2Nsojt3CcxDTvdytaRuupTyiqdVkcSi5VWlTJ9wXRmJc4iszQT04AMDw9e865i+pJbKK0qtTqinISI+LMotfnjcJVh7vjW6jgijda6lDxe/mE3heXVRAR48buzO+BRmAZvTPhlx0VfGPMwjP6Lxir1qOOFf2afXx88zQpK5v4OXBqPtHSHL7yN+WA0F/rtZlx0FPMj/SitLrM6mpwCwzAob58AgCMjURvnSINTsUsaxK7MYpZsz+CS/U8SXr4XvINh3EMQGmt1tJYrtAsMvB6AcWn/5sftBy0OJFabkzSH5NxkXNQejLiA5Nxk5iTNsSaYnJJ2rVuxN+AsAIq2fGNxGpHGp9rpYv6GA8xbt58qp0l8RCtuHtkZn+zN8N/xkL0DfFrDxCdhyB/Aw8vqyM2aj5eDnHNeoNzwxb9gO1XfPmJ1JLHQry+8ZZVn4zIMMjw8eDP7J6YvmK4Lb01MQNeRVBkOfCqyYf8aq+NIC6Nil9S7wvIqPliTypDsefTNXwSGHRL+BHETrY7W4hmj78fp8CeyKo2Kla9SUFpldSSx0Lwd844odB3mwsW8HfMaOJGcDrvNoDBqGACug4nWhhFpZApKq3j1pz2s3puLYcC4buFcMzQGn5Tv4c3zoDgDAtvDBc+7+4lqx8UG0bdXb5Z2ugcA++qXYL96DrZUx77wZurCWxPUOSqMfX59AKjY9KnFaaSlUbFL6pXLZfLBz2mE5G7gvPQX3QcHTIeB12kno8bAtzW2MQ8AMCZjNks3qVdGS5ZVlnVGj0vj4egyBoCAgmQoVU8+EXDPMn/x+52k5Zbh47BzzZAYxnaLwNjwDrx7OVSVQERPuORV6HaBdn1rQHabQYexN7DFfzg2s5rqj67TcsYWShfempcAbwcZYe6NcypSNLNLGpZ+iku9+j45k8z0VK5O+Rt20wkdhkPCX7QkoBExBt1AVWAMrVyFBK9+isxCNbRuqcJ8ws7ocWk8ojt3J9cRiQ0n1Vu+sDqOiKVM0+THHVm8sXwvxRVOogK9uXVMLPHhvvDtX+HzW8F0QswImPIGdBiiC3IW6NomgI39HqXM5odHwT5YN9vqSGIBXXhrfsxO7gtwvjmboaLE4jTSkqjYJfXmYH4ZS7cd5MrUh/CvzoHAaDjnMWgVanU0+TW7A8f5TwMwNO9zlq9eZXEgscqUuCnYjvFjwYaNKXFTGjiRnK5wfy9SggYBUJy8xOI0ItZxuUw+XJvGN1sOYZrQv30QMxI609pWAnMvgxUvuE/scQlc8hqExVsbuAUzDIMxA7qzLOwKACp++Jdmd7VAuvDW/ER26U+RPRgPVwVm8ldWx5EWRMUuqRdOl8nH6/ZzzsFZdCpJBIcPjH4A2va3OpocTZfxVEYPxwMnXTb8H/vz1PyzJZrWfRrtW8ViM00wzZrjNmzEt45nWvdpFqaTU2EYBuXtRgBgP5RobRgRi5imyWcbD5CYVoDdBhf1jWLKgHY4cnfCa2Nh92Kwe8Lwu+CC5yCgjdWRW7w2gT5Un3UzZTY/vIrTqF7zptWRpIFNiZuCwdFnVurCW9MUE+bPXn/374Al27+3OI20JCp2Sb34cWcWoSlfMiL7A/eBoXdA76mWZpLjMAw8Jz2NCxs9S1ezZdViqxOJBXwdvtzouIiZ+QWEO13YMIjwjWBm35nMnjgbX4ev1RHlFPh2HQ2Af0kK5O61OI1Iw/suKYOf9+ZhGHDFWe0Z3CkEY8dCeG0M5O4G3xA49ylIuB+8A62OK78Y3SeWNW2uAqD8h+c0u6uFmdbt93SpMn+58Pa/47rw1nQ57DaKwt27RFcf2mpxGmlJVOySOpdZWM7W9Su5NO0f7gM9LoWht6rRa2MX0Z3yLpMACN/6GvmllRYHkoZWWllN9N4vmJFfyJcePdj4+0QWXbaIGX1mqNDVBMVEx3DQOxaASu2AJC3Mil3ZfJ/s7u1zUd+29IwKgJ+ehfeugMpiCOsKF70M/a8BD0+L08qveTvstB57G2U2P1qVplG4/HWrI0kD8ti3ircPpHFTfgmhnkHYDJsuvDUDjk7uXaL98pOhSv2BpWGo+iB1yuUy+Xz1Nq7cez+eZjlmZB8Y9zB4tbI6mpwE31G3A9C7eDlrNunKS0uTtGc/PfKXAuDT7RwVqJu4QF8HB1ufDUDJruUWpxFpOIlp+XyxKR2Ac7pHMKitF8y7DhY/CpgQOx4unwNx+n+userRMZqkGPcMHufyf2M6qy1OJA2l9KeX8DVNhhqDWDL1ezZO26gLb81ARGw/ymx+OFzlmHu0lFEahn7CS51asSuL4Zv+QkjlAVy+YRjjH4Hg9lbHkpPVbiBlYb3xoBqPNa9QVqmlAy1JSeIneJoVlPpEQfeLrI4jdaCqg7tvl1dmYq0+bCLN1Y6MIj5amwbAkM4hJARnw+vjYesnYNhh0M1wyatqRN/IGYZBzPl3U2prRXD5flIXv2J1JGkIuXsITHNvqlIZfyGG3cPiQFJX2rVuRVqr3gAUJaldijQMFbukzmQXV1C55B90LVqBy+aJbcxfoPNoq2PJKfIecRsAA/K/Ye2udIvTSEMpKK2ifdp8AIzYsdBKux01B4FdR1GNB76V2XBwg9VxpJF46aWXiImJwdvbm8GDB/Pzzz8f9/yPPvqIrl274u3tTa9evfj6668bKOnxlVaV8vLGlxn30Tj6zOnD6A/G8tel/6LSVU6ftgFcUP4FxmujIXMreAXA+EfdN9/WVkeXkxASEkZ6t+sB8Fv7EuUVaq/Q3JUvfwUDk2SfvnQdcr7VcaQO2W0Ghb/07ao6uNniNNJSqNgldcI0TdZ8+x4Jh94AwDh7JvRTA8mmyOhxEZXeYQQ488lb+RbVTpfVkaQB7Ni+kY4lGzEx8Ok5yeo4Ukc6tgkn1a8nAGUb1bdL4IMPPuCuu+7ioYceYv369fTp04cJEyaQmZl51PNXrFjBlVdeyfXXX8+GDRu46KKLuOiii9iyZUsDJ6+ttKqU6QumMytxFhmlGbhMF9nlmWwrn8eGikeYtONOjG/ug+pyaNMXLnkNzp4JDm9Lc8upaX/enZTZWxFaeYCtX/3H6jhSnyqK8dj0DgB72kwiMDDI2jxS52wxQwHwy0sCl36/kPrX5Ipdp3o1UurPb6+oPpP3FK8G+ZPb6RyMkXeDph43TXYH9rNvAqB/xsds3p9vbR5pGBvfB6CgdW/oNMriMFJXfDztZIa5B5fl+/TzUuCf//wnN954I9deey3du3fn5ZdfxtfXlzfeeOOo5z///PNMnDiRe++9l27duvHYY4/Rv39/XnzxxWO+R0VFBYWFhbVudW1O0hySc5Nx8dtfmEyyKvfxTs4asDlg4PVwxXsQPxFs9jrPIfXL4RdMSf8ZAERve5UDucUWJ5L6Yia+i0dVMVkebQgbMNnqOFIPQuPOpspw4F1diJm22uo40gI0qWLXqV6NlPrz2yuqJiaZHnZmBQUxw7+SUrvD6ohyBuxnXYfT5qBd5W52/vwNpnr9NGvZRWXEHvwCAO/4ceDwsTiR1KlflpP75WyCai0DaskqKytZt24d48aNqzlms9kYN24cK1euPOpzVq5cWet8gAkTJhzzfIAnnniCwMDAmlt0dHTdfIBfmbdj3lEKXW4uYF5AIJz/L5j4BARG1fn7S8MJHXs7FR6taFV9kH99OZOxv1xkHffROF7e+DKlVaVWR5Qz5XJRveplAFYHnkvPuM4WB5L60DY0iIO+3QEo2LLQ4jRSX9buy+X9n1PZfqjuL3SdqiZV7DrVq5FSf451RdVlQHLhPuYkzbEomdQJv1DMHlMA6LrvbXZn6Upqc5ay/juCqw5RafPBu++lVseROhYeN5hSWys8naWYOzS4bMmys7NxOp1ERETUOh4REcGhQ4eO+pxDhw6d0vkA999/PwUFBTW3tLS0Mw//G1llWcd+0DDIstug/+/Aw6vO31samHcAJYNmML1NON+6NpD5y7LVjNIMZiXOYvqC6Sp4NXV7luDI20254UNZl8l4OzQLszmy2wzywwcCULE/0dowUm9sK55n6NKpmL+sGrE0i9UBTtbpXI1siGn0LdXxr6i6mLdjXgMnkrrmMexWAHqUrGZ9YqK1YaTemKaJ9xb3D6PCNsMgrJvFiaSutQ/1Z2/AIACKNn9lcRppCby8vAgICKh1q2thPsffRCPMNxwMo87fV6zxYUgAyZ6emL/5O3XhIjk3WRdZmzhzlXu3zbX+o+nVu6+1YaReGR3crRV8c5IsTiL1wTRN/DPX0r40ifDKA1bHaTrFrtO5GtkQ0+hbpPICskozjnvKca+4StMQ2ZOqdkOw4yJ863/JKCy3OpHUg0NZOcTmuLf59u9xDtiazI8FOUkedhsFbUcAYB7QjowtWWhoKHa7nYyM2j/DMzIyiIyMPOpzIiMjT+n8hjIlbgq2YwxjbdiYEjelgRNJfZq35wtcxyhe6iJrE5e9C2PXt7gw2BRyPl0iA61OJPWodfwIXNjwr8zAlbHd6jhSxwpKK4kodhcyA8LaWZymCRW7TkdDTKNvUVwuWP82vDCAsOrq4556oiuu0jQ4hv8RgEFF37JiW4rFaaQ+ZK3+AC9XGYVebfDqdbHVcaSeeMWNBcC/cAcUHf9ihTRfnp6eDBgwgMWLF9ccc7lcLF68mCFDhhz1OUOGDKl1PsB33313zPMbyrTu04hvHX9EwcuGjfjW8Uzrrh2hm5MTXUTVRdYm7OdXAUj26Ud0z2HYbZqR2ZxFRYST4ePuyZa/6WuL00hdO7h/LwHVubiw4eg41Oo4p1/s2rVrFwsXLqSsrAyg3htYn87VyIaYRt9ipP0Mr4+Bz2+FkizOLzUwjvFXriuqzUjcRKpbtcXXVYKxfg6F5VVWJ5I6ZJomgb9cDS9rPxr8wy1OJPWlfaeuZHpGY8NF1eZPrY4jv9LQ46m77rqL1157jbfeeott27Yxc+ZMSkpKuPbaawGYNm0a999/f835t99+OwsWLODZZ59l+/btPPzww6xdu5Zbb721XnOeiK/Dl9kTZzOz70wifCOwGTYifCOY2XcmsyfOxtfha2k+qVsnXLaqi6xNU3khZuI7AKwIPI9+nSJO8ARp6mw2g/wwd9+u8pQ1FqeRulay173zd5FvNLTuZHGa0yh25eTkMG7cOOLi4jjvvPNIT08H4Prrr+fuu++u84CHnc7VSKkDlaUw/w/w3/FwcAOmhw+ro35PRut/EeTopCuqzZ3NjsewWyg1DHZVfcZ5n5yjHZCakf17t9OhaD0mBq37TbI6jtSj0FaepLV2/6wsSf7e4jQC1o2npk6dyjPPPMODDz5I3759SUxMZMGCBTVtIlJTU2uyAAwdOpR3332XV199lT59+jBv3jzmz59Pz5496y3jyfJ1+DKjzwwWXbaIjdM2suiyRczoM0OFrmZIy1abqcR3MSpLyHC0paDtGNoGaTfoFqG9ezzirb5dzY5xMBGAqtZdGsXu7qdc7Lrzzjvx8PAgNTUVX9//DSamTp3KggUL6jTcb53oaqTUsaJDMPs8SJzrvt9pNN8OfI35ITfi1aotb054Q1dUW4DSnpcwPSqStwOgoCpbOyA1IyU/vw1AZkBPHLGjLU4j9ckwDCo7jALAkaG+XY2BleOpW2+9lZSUFCoqKli9ejWDBw+ueWzp0qXMnj271vmXXXYZycnJVFRUsGXLFs4777x6zSfyW0ddtmqaGBi6yNpUuVzws7sx/Qr/c+kXF42hTSVahMCuIwEILkvBlX/Q4jRSV6qcLgLzNgPgEx5rcRo3j1N9wrfffsvChQtp1652w7EuXbqQklK/PX2mTp1KVlYWDz74IIcOHaJv3761rkZKHTq0Gd69Agr3g6c/JNzP6pDJLN2aBwZcNjCazqGBdA6dwYw+M6xOK/Vozp7PSPZ0HLH35q93QNL3QNPjdDqJ2ONezmZ2HgueKlA3d4FdR+NcbcevPAMObIC2/ayO1KJZOZ4SaWoOL1udkzSHeTvmkVWSQZjTyfCKYG6++DVdZG2KUpZB7h7KDR8SgydwR7sgqxNJA4mMiiHXM4rWlQfJ3vg1YaNusDqS1IH0vDLalro3HfBta/3sbziNmV0lJSW1rkAelpubi5eXV52EOp7jXY2UOrLjW3hjgrvQ5R8F5z/LnthpfJ6UB8A53SPo2VY7pbQU83bMO6LQdZh2QGq6DmxcQnDlQSpsPoQP0vKPliCmbQSpfu7BR9km9e2ymtXjKZGmptay1fFvsSjtIA9nbmLn6qVWR5PTsd49u3yj3zDate9EsJ+nxYGkodhsBrmh7r5dZXtWWJxG6kpGajK+zkKchgdGh8bRZuqUi10jRoxgzpw5NfcNw8DlcvHUU08xerSWwTR5q1+B96ZCZQmE94BLXyOn02Tm/pyGy4Q+7QJJiFcT0JZEOyA1T84N7oawh0KHYIvobnEaaQi+nh5khg8DoHzvaovTiMZTImegbX8Kf+nT1XHtYxSUVlgcSE5JWT7mts8BWBswhn4dWlscSBqa2d69U59Xtvp2NRcVqWsBKGkVA0HR1ob5xSkvY3zqqacYO3Ysa9eupbKykvvuu4+tW7eSm5vL8uXL6yOjNARnNSy8v2b7XzqPgYlPUtCqI28t20tppZN2wT5cOqCd1tO3MGE+YWSUZhz3cWlaKstLaXPgOwC84saBzW5xImkwncfA3v/gl7MJqivBQ1fSraLxlMiZ8b/g71T+6yuiK3ay4ZsX6Xdp/W3sIHVsyzyM6nIOOaLJCB5EjyitGGlpAuJHwCoIKdmJq7QAm6++B5o6R0YiAGZILHg0jhnqpzyzq2fPnuzYsYPhw4czefJkSkpKuOSSS9iwYQOdO3euj4xS3yqK4P0r/1fo6vc7uPhV0j2j+c/SXWQVVRDg48Hvzu6Aw37K3zLSxGkHpObn4JrP8HaVUOgRQsRZF1sdRxpQRNxgSm3+eDpLcSXXbxN0OT6Np0TOjOEfSeFZtwPQOekFigpyLU4kJ+2XJYxrW42he4cIfDx10a2lCe/QnWKPIDyoJmvTQqvjyBkqKK0itNA9S8+3TbzFaf7nlGd2AQQGBvLAAw/UdRaxwv618OnNkLML7A4YdicMv4PkXBfv/byHimoX4f5eXDM0hkAfh9VpxQLTuk9jSeoSknOTcR3u3mWaGIZNOyA1VVs+BiA3chgBAW0sDiMNKTrUn20BZ9EzfwnFm78ioMeFVkdq0TSeEjkzIePupCBxDoEVB9n16QP4T59ldSQ5kUObIT0RJ3Y2BI3j8vbBVicSCxg2GzkhA2mVsYiSHT/A2ZdbHUnOQFpuMbFlyQA42jSO5vRwGsWuH3/88biPjxw58rTDSAOqroQfnoRl/wTTBT7BMPoBGDCdVSmFfL7xIKYJncP8uHpwB11xacGOtQPSkIpQLh79b+2A1MQUF+bSNvMHAAK7jQEtS25R7DaDorYjIH8J5sENVsdp0TSeEjlzhsOb4oRHCVx4Ax1SPqLkwO34te1qdSw5nl9mdW31PQt7SCe6hLeyOJBYxdnubMhYhGf2VqujyBnKTkmip6uUapsnHtGNZwPBUy52JSQkHHHs1z2cnE7nGQWSBpCx1T2b69Bm9/0OwyDhflzth/HN1gyW7coGYECHYC7qG4WHli62eId3QJrRZwakroI3JuDkEO9t3Ey/6Aj1cWtC0lfNo4tZSZ5nFMF9NaunJfKKHwdbH8G/cCcUHQL/SKsjtUgaT4nUjajBl5K2ahbRBevI+fQO/G7VEu1Gq6ocNn0AwDr/MfTrEILNpjFkS+UfP+L/27vz8KjKu//j7zMz2Sb7vpCEJEAIyCq7KKJEwX0DtdVS1GqhWlur7aP9Pa3dtZu22hZqHwWxrrhvRdndkE32JeyEJfueTJbJzPn9MRiLAhJIciaTz+u6ziU5s33mGMid79z394Z1kFC/A09LM/Zg/+jzJO3nOehrTt8UlUNEVJrFab7Q7ipGVVXVMUdpaSkLFy5k1KhRvP/++52RUTqK1wMfPQpPTPQVuoIjYcKP4cZnackYz3NrDrYVui4emMx1Z/dSoUu+KnMsrVkTsePlrB2Psau03upE0g4hO14D8M3uCY+3OI1YoXdOf0qDM7DhpWXTq1bH6bE0nhLpGIbNhvvih/FikFK+ksYt71odSU5kx9vQVE21PZ5dMedydm8tYezJEvqMoNkWRqi3kdKty6yOI6ep1ePFWe6bRGNP7OtrjeQn2l3JiI6OPuZISEjgoosu4ve//z0/+clPOiOjdISKPTD3Ulj8C/C0QNoIuO5JmPhT6owI/vXhXrYeqcVhM7hxVAYX5CVpto6ckOPiXwAwtOFD1q/+ENM0rQ0kp6Si9DC9Kj8FIG7IxRanEavER4RwMG4cgK9PhlhC4ymRjpM9cBRbUq8FoPU/9/s+4BX/s963hHFdxETSkpNIjgq1OJBYybAHURE7DIC6HUutDSOnraimiTTXdgBCk/2nOT2cRrHrRJKTkykoKOiop5OO0lABS38Dc86Fg5+CIwzG3gXffBH6T6aysZU5K/ZwqKoRZ7Cd287NZmhGjNWpxd+lDcfddzI2TAZu/yv7yhusTiSnoGzVS9jxUh6WTcQAFbt6spbeEwEILlbfLn+j8ZRI+xmGQchFD+KyRRDZcIDmFX+xOpJ8WdUB2LscgHVR+ZytxvQCtKb7+js5SjZbnEROV2F5LWmNOwEwUodYnOZY7e7ZtWnTpmO+Nk2ToqIiHn74YYYNG9ZRueRM1RbByr/B2rngPlqISDoLJtwHA68Cm52S2iae+mgftU2txIUHccv4bBIitFZaTk3QRb/A3P0+g12f8vqnS8i54iqrI8lJmKZJxK7XAWjOnAChUdYGEktFD7iA1lUOwptL4PB66DXc6kg9jsZTIh0rNzuTD7O+y4S9f4ZP/grjvgOh0VbHks9teBaA3aGDqYkZwBB9uC5ARO4E2PgXEuq20+puxRHU7vKEWKy2cDNBZgutdieOjNFWxzlGu7+bhg0bhmEYX1m2NHbsWJ566qkOCyanqeoAfPwXWP9v33JFgNhsGHw9jLwFolIBOFjpYt4n+3G1eEiOCuHWc7OJCvWf9bXSDSQPpCXvKkJ2vE7etr9QeM5FZMZrZ0Z/VVS4m/TaDXgxSBh2mdVxxGLZackUhg8ip2EDDRtfJVzFri6n8ZRIxzIMg/gLv0fZoRdIbDmM+z8/Jeiav1sdS8C3rHS9r9i1NuJC8tJiiQhRUUMgJGsos2NieCUyjLLnzibRmcTU3KlMHzhdO753F0d3926J7YMjItniMMdq978y+/btO+Zrm81GYmIioaFac22psgL46C++HU7Mo30KEvNgyA0w7GaITGq7656yep5ZeYDmVi8ZcWHMOCcLZ7B+4Ej7heT/DO+ON8lr/Ix3P3mHzCumWR1JTqBqzYukAeWRA0jqc57VccRiYcF2ypLGk7NvAy37VxFudaAeSOMpkY43sFc87/T9EZdvuxfbphdhwr0Qn2N1LNm7HGoP0WgLZ2vMBdygJYwCuNwubln6PXbERGEaACYlrhJmb5jN0sKlzJsyTwUvP1fT6Ca2eisAwYn9wGa3ONGx2l3h6N27d2fkkPYyTSjaCDve8R2lW7+4LWUIDL0Rhtz4ld3Wth2p5fnVhbR6TfokhvOtcb0JcfjXN6V0Iwl9aT5rGmFbXyR36185ct7lpMWEWZ1KvsTjNYnb+yYA3qzzIFgDBwGjzwWw7++EV2z2bQcfpCJLV9J4SqTjGYZBzvhp7Nn/DH1cG2h984c4bnnT6lhytDH9+vDzCI5OIi8l0uJA4g/mb5tPQWXB0ULXF7x4KagsYP62+cwcOtOacHJKDla6SG/0Nad3JOVanOarTqnY9dhjj53yE959992nHabbW/YQRCT5ZlQl5n2l0HTGPG448PHRAte7UHvoi9sMG/Qa4StwDZl23B4F6wureHndIbwmDEyL4sZRGQTZO2yPAumhwvL/H55tr9C3aTOLV7xC2lU3Wx1JvuTAzo3kuArwYCdp+BVWxxE/kdJ/LK6lkTg9dXh3voftLPXd62waT4l0vgFpUbyY92OyP7sZx4EVvllFOROtjtVzuSp9v7sAayMvZGhmLA79/iHAyztfxov3uLd58fLyzpdV7PJzh8uqyWva4/ui19nWhjmOUyp2Pfroo6f0ZIZh9NzBmbsRPvgDmP/1F9YZD4n9IXHA0QJYfwhPgKAwCAo/+l8n2B3HPk/tEag55DtqD3/x58Nroanmi/vaQyBtOGSOg9zJvj+f4JP5lXsqeHPjEQCGZ8Yw9ex0bDbjuPcVaZfY3jQPvgnnpqfpu+0xSideR1K0Znf5E9e6FwAojxlCcqZ/NY4U68RF2fhTUh8+CCqlbO3/krjtcfXJ6GQaT4l0PsMwGDpqAuv2XsKo6nfxvPMT7HetAkPjXktsehE8LRwJzqYo5myuzYyxOpH4ibLGsjO6XazXcHAjDrOV1qBIHKn+1//1lIpdX+4rIcfR2gxj74TClVBdCA2l4KqAA5/4jpOxBfkKXzY7NFad/L4hUZA+CnqfA3mXQnzuscWyL8fyeFm4tZiPd1cAMK5PPFcMScXQD3zpQM78B2jd/DxZzQV8tOIFkq68xepIclRTSysphW8D4OgzERzacVV8fTJuff8WdoRVYR4dCqhPRufTeEqka+SlRDIv726GrFpKSEWBb3fyUbdaHavnMU34zLeEcU3EhSTFRtNL7S7kqMSwREpcJSe9XfxXq8dLUMlGALxxfX2TevyMupJ3lLAYmPwb35+9XmgogyOfQdEGKN8N1Qeg5jC4G3yFsdZm4OgOTF43NLu/eC57iO+bJTwBnAm+pZHhiRCbBf0uhuj0U/p0qrS2iRfWHKSopgmASXlJTBqQpEKXdLyoVFxDvk3Uxn+RvfVxXJNvxhmi3T39wb4tKxnQfBC3EUTc8CutjiN+oq1PBsfuBKg+GSISCAzDYNyws/hg701cVPok3mW/wzbsG74Pl6XrHPkMSrfSagSxIXoSE3vH6vcQaTM1dyqzN8w+7lJGGzam5k61IJWcquLaJlIbtgEQlNQPbP63PPm0il2HDh3izTffpLCwkJaWlmNue+SRRzokWLdms0FkMvS/xHd8zjTB0+LbftfjhpYGaK6F5jpoqfM1CI5MgehMCI0Ee/BpTbk2TZM1+6t4e9MR3B6T8GA7141IZ0BqVAe+SZFjRV70E1o2P0Ov5j1sWfY0g6Z8x+pIArRufAmAyviRJKcOsTiN+Av1yfAPGk+JdJ7+yZGsyL2FmsrXiXaVwdLfwOTfWh2rZ/lsPgBbnGNojkhnWEaMtXnEr0wfOJ2lhUspqCz4YkximtgMO/3j+jN94HRrA8pJFVa6yG4sAMBI8L/m9HAaxa4lS5Zw5ZVXkpOTw44dOxg0aBD79+/HNE3OPtv/mpL5FcM4dglRWDSQdkZP6XK7mL9tPi/vfJmyxjISQhPIdV5ERPMkHEYofZMimDYynahQzbKRzmVEJFExcAapW+aQuP5xvBfdgs2unT6tVNfYTMaRhQCE5Z5/0iXP0rOoT4b1NJ4S6VyGYTBxUBbvHZjJ9Qd/i7nmSYxxd0FUqtXReobGKtjk+8BtdWQ+/VKiiA7T7yPyBWeQk3lT5vl+ly14iTJXKYkeD/kZV3L3Bf+rdgp+7nBJOWObjrZnyBhlbZgTaPdcswceeID77ruPzZs3ExoayiuvvMLBgwc5//zzmTZtWmdklBNwuV3MWDiD2RtmU+IqwWt6KW0s5aOK51jp+hUXDIjm1vFZKnRJl0mY/GOabWEkN+/n0PK5Vsfp8favX0qMu5RmWxhRw66xOo74ka/rg6E+GZ1P4ymRzpebHEF5ztUcCs3FaG2Ed39sdaSeY/2/we2iNKQ3+2LO4ezMWKsTiR9yBjmZOXQmi69fyuLaZBYfPMIV5ajQ1Q24D2/EhpfWkFhI8c/VI+0udm3fvp3p031TCh0OB42NjURERPCrX/2K3//+9x0eUE7s854rX12KYlLrLWRvy7taFy9dKigygUO5MwCIWPNXX/86sYyxeQEANUljwU+nF4s1puZOxXaCIYD6ZHQNjadEOp9hGOQPTOWdXj8AwCx4Bw5/ZnGqHsDrgdVPAPBhxCWEOCMYmKZ2KnJy7qwJAISUrLM4iXyduiY3UZVbADAS+kGYfxaz213sCg8Pb+srkZqayp49e9puKy8v77hk8rVeKlhwwp4r5tGeKyJdLeHie2m0hRPXVEjVh/+yOk6PVVpdR3bpYgAi8y7wy6aRYp3pA6fTP67/sQUv08QG6pPRRTSeEuka/ZIi8GaMY3PUBAzTC+/8yNdHVzrPzoVQXUizPYINsRczJD2aILvGIXJyUQPyAUir20JTU7PFaeRkCitd9GrcAYA9sd9p9RnvCu3+V2fs2LF89NFHAFx66aXce++9/Pa3v+XWW29l7NixHR6wO3C5XczZOIf8BfkMnT+U/AX5zNk4B5fb1SmvV1bXzCvrDqnnivil6LhEdmYfna3w6V99n+5Jlzu09l3CPTU0OqIJG3qV1XHEz3zeJ2PWsFkkO5MxTEj2eLjFE8+8KfO0fKALaDwl0jUMw+CigUksTL2TVhxwZD1s1gfCnWrVHN9/IvNpDUtkZO84iwNJdxCRM5oWWxhObz1FGxdZHUdO4mCli3TXdt8Xfrx6pN3dih955BHq6+sB+OUvf0l9fT0vvvgi/fr165E7B33eN+u/lxOWuEqYvWE2SwuXdugvDYUVLlbsKmN7US2mCaFGLE1m5Qnvr54rYpXoC3+Aa98zRDYepvmTfxJy7vesjtSjmKZJ2HZfU9j6tPGExfS2OJH4o8/7ZMwcOpOV7z3PuJUzcYW0qtDVRTSeEuk6fRIjiEzrx8cV0zi//HlY9DMYeBU4gq2OFnhKtsG+DzANG59EXUJaTBgZcWFWp5LuwB5EZcIoUko/oGnHIhhzudWJ5ASKSktIbDno+yLDfz+ga/fMrt/97ndUVvoKLOHh4cyZM4dNmzbxyiuv0Lt3z/uF6kR9s7x4KagsYP62+Wf0/GV1zSwvKOXvy3Yze8Ueth3xFboGpkZyXT/1XBH/1DsthQ0Zvtldno8fA0+rxYl6lkOHD9KvcgUAMYMm++3UYvEfUbnn48GOs7kUjmywOk6PoPGUSNcxDIP8AUksTb6FOnss1BXBst9YHSswrf4nADsjRlMT1Z8xOfHqISynzJt9PgDOsg3WBpET8npNOLIRAI8zCZLyLE50Yu0udpWVlTFlyhQyMjL48Y9/zMaNGzsjV7fx8s6XT9g3y4uXJzc+zy/e3MpD/9nOI+8XMPfjffxncxGfFVZxpLqRVs+xjzVNkyPVjby/tZhHF+3kkUU7eW9rCYeqGrHbYETvWO7J78e3xmVx98jbvtpzBV+hSz1XxEqGYRB67izq7VE4G4vwfvoPqyP1KNWfPovDdFPpzCFo8NVWx5FuoHdaEgedAwFo2vyGxWl6BivGU5WVldx0001ERUURExPDbbfd1ja77EQmTpyIYRjHHDNnzuz0rCIdrU9iBJmpSSxMPfr9u/oJqD1ibahA46qEjS8CsDziUkKCgxmaEW1xKOlOos+6CIDUhu001NVanEaOp6i2iZR63xJGW0JfCPXfv+PtXsb4xhtvUFVVxYIFC3juued45JFHyMvL46abbuKb3/wmWVlZnRDTf31dX6wms4rmVi/Nrb6iVll9CztLvhhY2gxIiAghJToUZ7CdguI6qlzuY27vmxTBWWnRDEiNJDI0qO22z3uuzN82n5d3vkxZYxmJYYlMzZ3K9IHTtRRFLDU4J50VKd8i//Dfaf34bwSPnQX2oK9/oJyR1lYPyXt8Sxhb+1wEYTHWBpJuISLEwZb4MWS5NtO0fw2hVgfqAawYT910000UFRWxaNEi3G43t9xyC3fccQfPPffcSR93++2386tf/arta6dT4wvpfgzD4IohqTxWegljy18lo6kA3voh3PSS1dECx/pnoLWRyrAs9seMY1zvWEIcdqtTSTcSnjGEBkcs4a1VHPrsHXLO/4bVkeRLCitcpDf6il1Ggv82p4fTKHYBxMbGcscdd3DHHXdw6NAhnn/+eZ566il+/vOf09ras5YrJYYlUuIqOent916cS0urlya3h7K6ZoprmyipbaK4pplGt4fSumZK677YcSLIbpCbHMlZaVHkpUQRFnziHxL/3XNFxJ8E2W14R32HuuJniXSVwCePw3k/sjpWwDu45SOyG/fiNoJJGHWd1XGkG/FknQcH/4/Q8k3g9WoHzy7QleOp7du3s3DhQtasWcPIkSMBePzxx7n00kv505/+RFpa2gkf63Q6SUlJOeXXam5uprn5i3FNba0+nRf/kBQVyri+SbzVcA/f2zMTdr0H+z+GrPFWR+v+PK2w2rcL99LwSyEojLHZakwv7WQYVCWPI/zwu7TsXgoqdvmdwsoGLnIV+L7w4+b0cBrLGP+b2+1m7dq1rFq1iv3795OcnNxRubqNqbkn75t1ff9pJESEkBYTRk5iBGNy4rlqWC/umNCHn10+gPsvyeOW8VlMGZTC+L7x3Dw2k/+9bCA3j+3N8MzYkxa6RPzdqH7prEj6FgCelf+A1haLEwU+z9qnAShJGIet1wiL00h3EtPvHFqMEELdNXDgY6vj9ChdMZ5auXIlMTExbYUugPz8fGw2G6tWrTrpY5999lkSEhIYNGgQDzzwAC7XyXebfuihh4iOjm47MjIyOuQ9iHSESQOSqI4fxmcxvuVSvHMPmKa1oQJBwbtQc5AWRxQbYy4iJyGcpCjNE5b2M/r4+nZFlvfsdkn+qrT4MHHuIt8Xvc+xNszXOK1i17Jly7j99ttJTk5mxowZREVF8fbbb3Po0KGOzuf3pg+cftp9swzDIDosiNzkSM7PTeTyIWmclRZNsEOfpktgiA0Ppm7Qt6ixx2N3lcFH2mGsMzU11JBx5F0AQgdMBvtpTd6VHio7JY4D4YMBaNr6jsVpeoauHE8VFxeTlJR0zDmHw0FcXBzFxcUnfNw3v/lN/v3vf7Ns2TIeeOABnnnmGW6++eaTvtYDDzxATU1N23Hw4MEOeQ8iHSE0yM6UQSksTL2TFiMUygp8/bvkzKzyNaZfEzWJ1rAExubEWxxIuqv4QRcDkNK4m9rKE6+gkq5X0+gmomIzAN7INEjoZ3Gik2t3VaVXr15ceumllJeX88QTT1BSUsJTTz3FpEmTeuROG5/3zZo1bBbJzmRsho1kZzKzhs1i3pR56pslPd6Y3F4sT/YVfc1VT4C7yeJEgavok+cJ8TZSFZxK/Ejtxirt4wx2UJrg2z66uXCdxWkCX0eNp+6///6vNJD/8rFjx47TznnHHXcwefJkBg8ezE033cT8+fN57bXX2LNnzwkfExISQlRU1DGHiD8ZnhFDbHImS5OOfii9/GForrM2VHdWvBkOfIRp2Pgg4hIiw4IYmKa/93J6QpNyqAlJxY6H8rWvWx1H/svBShcZrm0A2BJzISTS4kQn1+6P/X/xi18wbdo0YmJiOiFO96S+WSInlp0QzrtZ06gu/TcxjWWw4veQ/6DVsQKSc8uzAFSm5xMb1fOWlcuZ82ZNgP1/w1mxBVqbwRFidaSA1VHjqXvvvZcZM2ac9D45OTmkpKRQWlp6zPnW1lYqKyvb1Y9rzJgxAOzevZs+ffq0O6+IPzAMgyuHpfHPihsYVfkm8Y3F8N7/gysfszpa93R0VtfeqDHURuVyYVYcdlvPmwQhHac29Vyi9y/As+9D4LtWx5GjDlS46Ofa6vsiMc/aMKeg3TO7br/9dhW6ROSUGYbB6H6pLE65DQBz9b+gsdraUAGornALyTWb8GAj/uyrrI4j3VR839E02sIJ8rhgzxKr4wS0jhpPJSYmkpeXd9IjODiYcePGUV1dzbp1X8zaW7p0KV6vt62AdSo2bNgAQGpq6hlnF7FSr5gwzs5J5d207wNgbngOynZanKobaqiAzQsAWOS8FJvdwegsNaaXM2PvOxGA6MpN1gaRYxRW1LfN7CJ5sLVhToGaQ4lIpxuWEcP2pMsoDc7AaKmDRT+3OlLAqf7kSQAKo0YQkzfB4jTSXWUnRbEvYjgATdsWWpxGOtKAAQOYMmUKt99+O6tXr+bjjz/mrrvu4sYbb2zbifHw4cPk5eWxevVqAPbs2cOvf/1r1q1bx/79+3nzzTeZPn06EyZMYMiQIVa+HZEOcfFZyeyLv4Bd4WdjeN3w1t1WR+p+PnsaWpuoDs/hQMxYBqRGEe0MsjqVdHNxg/IBSGo+QPXhXRanEQC3x4u7ZAdh3npMezBkn2t1pK+lYpeIdLpgh41xfZN5L9W31Nfc+ALUHLY4VQBpbSZ+96sAtPS5WEvP5LSFBdspTxoHQMuhDdaGkQ737LPPkpeXx6RJk7j00ks599xzeeKJLxpzu91uCgoK2nZbDA4OZvHixVx88cXk5eVx7733ct111/HWW29Z9RZEOpQz2MFFZ6Xwdq8f4sUGhSth+9tWx+o+PK2w5v8AWBZ+CQSFMTZHs7rkzIXGpFDu7AtA5fo3LE4jAEeqG+lVv8X3RXw/iEq3NtAp0FZdItIlzukbzx92TaSwdACZjdth4f1wwzNWxwoIVetfJ7a1mlp7LJnj1JhezoyRPQF2/5Hwyu3QXA8hEVZHkg4SFxfHc889d8Lbs7KyME2z7euMjAxWrFjRFdFELDMmO441+wfyafnVnFP5Kiz8H+h3kT44OhXbXofaw7iDovgs6iISIoLpk6ifGdIx6nudS8Ku3ZgHPrU6iuDr15VxtF+XkZjXLXZ918wuEekSzmAH4/omsDD1ewCYO96B0u0WpwoMrWvnA3Ag8ULCErMtTiPdXVLOMGrtcdjNFih41+o4IiKdymYzuHJoGotSvkOdPQZqDsH7/2t1LP/n9fg2HQLWxUyh1RnPmOz4du0mK3Iywf0uACCueuMxH8SINQorXWQ2HJ3ZlTTA2jCnSMUuEeky4/smcDj6bAoiRmOYHnj3J1ZH6vbMqgMklHwMQPBZl4IGmXKGeieEszdyBACNOxZbnEZEpPNlJYQzIDuTt3rdA4C59iko3mJxKj+3eQGU78QTFMHCyGsJstsY0TvW6lQSQBIHXYAHO7HuUir3rLU6To9mmiYlpaUkNe/3ncgcZ2meU6Vil4h0mYgQB2Nz4lmYOgsTA/Z/AAdWWh2rW6v6ZB4GJvvCBtFn5BSr40gACA2yU5U8FgDPkY0WpxER6RqTz0phe+wktkeMxfC2wqvfAa/X6lj+yeOG5Q8DsDX5KprDezEkPYawYLvFwSSQBDmjKYs6C4CajeqlZ6Uql5uYqs3YMDEjkiFlkNWRTomKXSLSpc7tl0BFRC4boif5Tix8wNpA3ZnXQ8iW5wEoT78IR5j6ZEjHsOVMBMBZswvqy60NIyLSBaKdQYzOieeN9PtosYX6Wi18/FerY/mnjc9D1T68IVG8EXY1GIYa00unaEw/DwD74TUWJ+nZDlQ0kOnyzXY1EvpDWIy1gU6Ril0i0qUiQ4MYnR3HotQ78BgOKFqvnY9OU+vupYQ3FuGyhZM48hqr40gAScvqT2VQCjbTg7ldO++JSM9wfv9EXGGpvJd8h+/EB3/w9fCSL7S2wIo/AlCQdg2usBTSY8NIj3VaHEwCUWh/34fj8TVbMD0ei9P0XIWVLjKPNqcnKc/aMO2gYpeIdLnz+iVSF9qLVXFX+U4sfhDUeLLdGlY+BcC2qAn07ts9phNL95AZ72Rv5CgAmnYtsziNiEjXiAoNYkx2PCsTrqMkPA/cLnjtu1bH8i/r50NNIZ7QWF50XAUY5A9ItjqVBKjEAeNpMUKI8NRQsW251XF6rMLy+radGEkZbG2YdlCxS0S6XHRYECOzYlmaPIMWWxhU7IbP5lsdq3upLyNi//sANPeZjGFXnwzpOCEOO9Upvuaj3qJNFqcREek6E3ITCHI4eCHtfkzDDvs/go0vWB3LP7ib4IM/A7Au6VqawxLJTY4gN1ltFKRzOIJDKY09G4C6Lf+xOE3P1OT20Fq2C6enDtMeDL3HWx3plKnYJSKWOD83kabgWD5IuMF3YsXvwdNqbahupGXlHOxmK4Uh/eg39nKr40gAcvQ5H4Dwun1QXWhxGhGRrhEZGsS4PvEUh/VlXa+bfScXPgBNNdYG8wfr5kLdEdyhCbwZfCU2w+CywakY2glaOlFz5gQAgorWWZykZzpU1Uh6g29WlxHfF2IyLU506lTsEhFLxDiDGdE7lg8Tv0mjIxpqD8MnagR7SprrYfW/AN8uSEmJiRYHkkCUnpFFcUgWAObWN6wNIyLShc7rl0iIw8YbMdNpjkiHxkp464dWx7JWiws+fASAD+Kuw+OMY0xOPElRoRYHk0AXNehiAJLqtuJurLc4Tc9TWPlFc3oS+oM9yNpA7aBil4hY5vzcJFodThYnzfCd+ORxaKq1NFN30LTmaYLdNZQ7Usg59war40iA6h3vZN/nfbv2fGRxGhGRrhMe4mBcn3habSG8k/VT38mtr8He5ZbmstSaf0FDKU1hSSyNuIywIAf5A5KsTiU9QELOcOodsQSbzRxZ86bVcXqcAxX/3Zx+gLVh2knFLhGxTFx4MMMzY1kVdxW1ob2gsQoW3m91LP/mceP9+HEAtiZfSW6/fhYHkkAVZLdRl+rr20XxRmvDiIh0sfP6JRDisLHGNoSqvtcCJrxxl283wp6muQ4++gsAi6Km4g2NI39AEs5gh7W5pEcwbHaqUs4FoKlgkcVpehbTNCkuKyO5aa/vROYYawO1k4pdImKpif0T8dqDeTXlh74TG1+A4q2WZvJnrvUv42wsos4WTfr4b6hPhnSq4L7n48FGmKsIirdYHUdEpMs4gx2c2zcBgBfiZmGGxkDNQfjPj60NZoVVc6CxkvrQND6NuYzEiGDG5MRbnUp6EEfuJACiyjdgagf3LlNa10xizVZsmJjhiZA61OpI7aJil4hYKiEihGHpMRREjeNQ/HgwPfDmnaAfZF9lmrg/eBSAHUmXkpM3zNo8EvB690rlUJhvyrq59TWL04iIdK3xfRMIC7JT2BTGgXN+5zu5bh5sf9vSXF2qsdrXZgJ4N2oq3pBoLhuSht2mD9uk6yQMmQxAcvM+igp3WZym5zhQ4SLj6BJGIzEPQmOsDdROKnaJiOUm5iViGPBCwvcxbQ44sh42vWh1LL9Tu/V9omsLaDZCSBp7I4ZN/4RL58qMc7Iv2jdlvXHvSovTiIh0rbBgO6NzwtnZ/Co3HHiCodmZ5GekMWfxPbgq9lgdr2t8+g9oqqE6NJ0NsVPITY4gNznC6lTSwwTFpFEV0RcbJhVr9eFbVymsdNH7835diXnQzVaU6DclEbFcUmQoY7LjqAhJZ23a0W2+F/0M3I3WBvMzjct9s7p2xefTe8h5FqeRnsBuM2jp7dvy21GyCTytFicSEek6LreLJ/f8mJ0tr+DyVuIFShwOZkeGMOPN63A111kdsdO43C7mrP4j+XufZWhWBpcmh7PTWMyFA6LVQkEs0Zx1IQCOQ/rwrasUlte3zewi+Sxrw5wGFbtExC/kD0gmLMjOWzE30RyaCPWlsOhBq2P5jao9a0guX4kHG7Gjrgeb3epI0kPE551Dk81JcGsd7F1mdRwRkS4zf9t8dlYVAMe2VvAaBgVGK/PfvsWaYJ3M5XYxY+EMZm+fT4nDhtcwqLM1srP5FX788UxcbpfVEaUHih7kW8qYVruBmoZmi9MEvobmVryVewn31GDagiCr+33QrmKXiPiF8BDfFtZuWxhvp93tO7luLlTsszaYn6hb8mcA9seOp9eISyxOIz1Jbmoce8OHA9C8tQf1qRGRHu/lnS/jxXvc27zAy9VbYdfirg3VBeZvm09B5Y6vvHMTk4LKAuZvm29JLunZwvqcS6sRTLSnigMbV1gdJ+AVVrrIdPk2JzLi+0BMpsWJ2k/FLhHxG2Ny4kmKDGFt+ETK40aAp8XXrL6HKz9YQPqR9wCIGD4VHCEWJ5KeJDI0iPLk8QA0F66zOI2ISNcpayw78Y2GQZndDq9+B+pKuy5UF3i54CW8HH+jIC9eXt75chcnEgGCQqlJHg1A8/aFFocJfIWVLjIbPu/X1R8cwdYGOg0qdomI37DbDK4YmgqGwXMJd2MaNjjwMex41+polqpY/Cg2vByOGkrymOusjiM9kL2vr09GeNV2385cIiI9QGJY4slvNw1orIIXbwLv8WeAdUcnLfKdwu0inSUo9yIAYsrX0dIaOH/n/FFhhYvMtub0A6wNc5pU7BIRv9I3KZKBqZEUhfZhe69pvpP/+Ql43NYG60Iut4s5G+eQvyCfoU8P5XveD5gTE4V38JUQEml1POmBMvsNodqRiN1sxbPtTavjiIh0iam5U7Gd4NclA4Op/aaCPRgOrYZlv+vidJ1k73ISW0++GcnXFQFFOkvkWb5iV++mbew5VGJxmsDl8ZqUlJeT0nR019mM0dYGOk0qdomI37lkcCp2GyyInkFrSAzUHITlD1sdq0u0NYXdMJsSVwlevJQ67PwjJoZ7XSvVFFYs0SvWyf6YMQDUbQu8/jQiIsczfeB0+sf1P07ByyDS1pvcpFth8tHxyUePwL4Puzxjh2pxwVs/YGpdPcbxVzFiw8bU3Kldm0vkKCNpII0hiQSbLZRvVB/RzlJU00hS/XZseDGd8ZA23OpIp6XbFLt++9vfcs455+B0OomJibE6joh0ooSIEM7tm0CTPZIlvY727Pr071BbZG2wLuBrClvwlYa4pgEF1bvVFFYsYbMZNGZMAMBetMHaMCIiXcQZ5GTelHnMGjaLZGcyNsNGsjOZKzJncI7z5yzbVsOhvjdC3uVgeuDlW6DmiNWxT9/y30HVfr7lDiLW1gswjrnZho3+cf2ZPnC6NflEDAN31kQAHPs/xDRPUJWVM/LfSxiNxP4QFmtxotPTbYpdLS0tTJs2jVmzZlkdRUS6wMT+SUSGOlgRPpm62IHgboTXA//v/8l3flJTWLFO5IB8339dB6Bij8VpRES6hjPIycyhM1k8bTEbp29k8bTF/HbiPQzplUSr1+T5NQdpvOxvEJ0ODWUwdwo0VFgdu/2OrIeVfwegsN9MRof/iiHh15PkTGor8s0aNot5U+bhDHJaHFZ6svCBFwPQu34Dh6oaLU4TmHzN6X07MZI4AAzj5A/wU92m2PXLX/6Se+65h8GDB1sdRUS6QGiQnclnJeOmhXviR5KfkcZQ727ynzuHORvnBOxyPjWFFX+Vk9Wbw6H9AGjY8IrFaURErGMYBlPPTifWGURlg5tXt9ViTn8LwuKg+gDMvQSaaq2Oeeo8bnjj+2B68fY+jzfsF+Mwwvj+2bNYMm1JW5Fv5tCZKnSJ5T7fNCfNfYC9u7ZZnCYwHaho+KI5ffJZ1oY5A92m2HU6mpubqa2tPeYQke4jLzWEtc2/YZVnBSUOB17DoMRdx+wN/2DGwhkBWfD62p2f1BRWLOIMdlCaNB4A1+6PLU4jImKtsGA73xidid0GWw7XsrI6Cma849tIprwAnr7C1wOrO/jkMSjZDMERrO/zPWrcdqLCHIzs3T2XLkmAC0+gIW4gAM3bevaO7Z2hqqEFW/UBIjzVmDYHZJ1rdaTTFtDFroceeojo6Oi2IyMjw+pIItIOz2x/hgr3fuDY9fheTAoqCwKyf9XU3KkYHH+qsJrCitWMPhcAEF6+Ebza8ltEeraMOCeXDEoF4D+bizkUnAXfeh2CnFC0Af59LbS2WBnx65XvhuW/B8A78jssacgG4Px+iTjsAf2ronRjQf19uzImlq+mxtVzdmzvCrvL6r/o1xWXA7FZ1gY6A5b+C3b//fdjGMZJjx07dpz28z/wwAPU1NS0HQcPHuzA9CLS2V7e+TJmD+tfdVPet8j2BGMzzWNqfGoKK/4gedD5uI1gnO4q3Ps0u0tE5Jw+8QxMi/L171pdSFPycPjmS2APhsKV8MI3wNNqdczja66DV24FTzOkDuezjG9T1dhKZKiDUdlxVqcTOaHgXF+xq2/jJnYUVVsbJsDsKa0n0/V5v67+4AixNtAZsLTYde+997J9+/aTHjk5Oaf9/CEhIURFRR1ziEj38bX9q1yB17/q4LplPH9wNzOra0kKilBTWPErKXExHIwcBkDV+jetDSPtdjo7W5umyc9//nNSU1MJCwsjPz+fXbt2dW5QkW7ky/273t5UBNnnwfX/BpsDdi/2FZT8bTZsazO8eDMUbYSQKLzn3M2yQt8stPP6JRCkWV3izzJG47GHEemtoXjrh1anCRimabKn7L+LXQOsDXSGHFa+eGJiIomJ6j8jIseXGJZIiavk+DeaJolH/9tddwj5suaWZhI+/DlO0+Sa8HHMuv5pCAq1OpZIG8MwaMiYAFtXYx5ea3UcaafPd7YeN24cTz755Ck95g9/+AOPPfYYTz/9NNnZ2fzsZz9j8uTJbNu2jdBQ/fskAr7+XdNGZvCvD/ey7kAVeSmRDOo/Ga55Al79Dmx7A966G6583D/GLF4vvDYT9i4HRyhM+gUboiZSuesw4cF2RmtWl/g7RwitGedg37+EkANLaW69ihCH3epU3V5xbRMtrjpSGo/uup0+ytpAZ6jblOwLCwvZsGEDhYWFeDweNmzYwIYNG6ivr7c6moh0kqm5U7Gd4J8pGzC1uhLW/F/XhupE+xc+TlLjHhptESRc+D0VusQvheXlAxBXvRWzST+Du5P27mxtmiZ/+ctf+N///V+uuuoqhgwZwvz58zly5Aivv/76CR+nDYKkJ8pOCOf8XN+H+K+tP0xNoxsGXweXP+q7w/pn4M27fTOqrGSasPB/YOurvplnE3+K9+xvs7zAN1v+vNxEFQ2kWwjO8y1l7NOwgT2lDRanCQx7Shvo7dqCHQ+EJ0KvEVZHOiPdptj185//nOHDh/Pggw9SX1/P8OHDGT58OGvX6pNlkUA1feB0+sf1P07By6AX0UyvqYNFD0J19+/HV19VQubGvwBQkXcTjpwJ1gYSOYGMvJHU22MIMpup2fS21XGkE+3bt4/i4mLy8/PbzkVHRzNmzBhWrlx5wsdpgyDpqSblJdErJhRXi4dX1h3CNE0YMQMu/o3vDuvnwz/GwaF11oX84E+w+gnAgPE/hLGz2FxUT1l9C85gO2M0q0u6CaPPJACymnaw8+ARi9MEht2ldWTXr/d9kTwInN3734NuU+yaN28epml+5Zg4caLV0USkkziDnMybMo9Zw2aR7EzGZthICktiSPj1DI74E66IPHA3wIJv+z6p7MbK3/gZYZ46ykN70yv/Tv9Y5iByHCFBQRQnjAOgbtsii9NIZyouLgYgOTn5mPPJyclttx2PNgiSnspht3H9yAyC7Aa7SutZubfCd8M53/ctaQyNhso98GQ+vP/zrt+pce1cWHa08DbqO3DejzDtQSzdUQrAuX0TCA3SrC7pJhL64Q5PIQg3TdsX+YrLcto8XpP9FS5yGo4Wu1KGdPvfR7pNsUtEeiZnkJOZQ2eyeNpiNk7fyJLrl/DwhT8iyBHB/NT/h9cWBIfXwcq/Wx31tFXvWUvm/pcAaBpxB0Zcb4sTiZycN+cCAEJLN1gbRDp9Z+vToQ2CpCdLigrlkkGpACzcUkxpbZPvhqE3wJ2rIWcimF745K8w+xxfg/iusO1NeOdHvj8PmgYX/i8Eh7P1SC2ldc2EBtkY1ye+a7KIdATDwJ57MQDplSs5VNVocaDu7VCVC7O5nnTXdt+JzLHWBuoAKnaJSLeTEefk0sGplIVmsTD5Dt/Jpb+Bsm64Q5hp4n77PmyY7Ik9l/QJM6xOJPK14gdP9v3XtYfmCs3asVJn7mydkpICQEnJsRuFlJSUtN0mIl81NieO3OQI3B6TF9ccpNVzdCfGyBSY/gZc8RiERELFLnjiAljya/C4Oy/Q/o/gle/4imx9L4bJv4OwGKoaWnhrk2/51/g+mtUl3Y+t74UA9G3cxI7iOovTdG+7S+uP7deVMcbqSGdMxS4R6ZbG5cQzuFc0HyVcz8HwQdDaCPOvgPoyq6O1S8Wnz5JYtZ4WI4SI8XdASITVkUS+Vlxqb8pDs7BhUrr2Favj9GiJiYnk5eWd9AgODj6t587OziYlJYUlS5a0nautrWXVqlWMGzeuo96CSMAxDIPrRqTjDLZzpKaJxdu/tLP0iG/DrE+h93gwPfDhn2DOeVCyrePDFPwHnr8RPM2+X14v/SNEJlHf3Mrcj/dR29hKUmQI5/ZL6PjXFuls2edjYpDiPsihPZ3w96cH2VNWT3ZD4PTrAhW7RKSbMgyDa8/uRUJkGPMzf0ddcDLUFcHTl0FLN9mRpbme0OW/AKCg13Ukn325tXlETpFhGNT0Og+A1j0fWZxGTtWp7Gydl5fHa6+9Bvj+P//whz/kN7/5DW+++SabN29m+vTppKWlcfXVV1v0LkS6h6jQIK4Z3guAD3aVs6/8S2OTmHSY8Q5c+ggEh0PZdvjnBPjwEfB6zzxAxR54dpqv0NVcB0lnwWWPQFwWTW4PT3+yn7L6FmKcQdw6PluzuqR7csbhTRkKQOTBJb5dUKXdmls9HKhwkV2/wXciAPp1gYpdItKNhQbZ+eaY3jSHxvNE9qO0BEVCWQE8dz14Wq2O97Wq3vwp4c1lVDqSSc//Htg00JTuI7S/b4e+2KoNmB3xi5l0ulPZ2bqgoICampq2r3/yk5/w/e9/nzvuuINRo0ZRX1/PwoULCQ0NteItiHQrg3pFM7J3LKYJL609SJPbc+wdDANG3+ab5ZU+GrxuWPJLeOpiqDpwei/aXA+Lfwn/GAu73gebA866Gq6eDSmDaPV4+fenBzhU1Uh4sJ1bx2cT7Qw64/cqYhV7/ykA5NWvZmeJljKejv3lLuyeRjIaj86OC4B+XaBil4h0cynRoVw1LI3ykEyezPwDXluwrzfF6zP9eofG2k+eJHbr0wDs7n87sVmDLU4k0j5Jgy+kFQcx7jLKdq22Oo6cglPZ2do0TWbMmNH2tWEY/OpXv6K4uJimpiYWL15Mbm5u14cX6aYuG5JKXHgQ1S43C9b+V/+u/xabCbe9D/m/BEcIHFoD/xgD654+9bGMacLml+Fvo+CjR8DTAqnD4Mq/+XaCTBuK12uyYN0h9pQ1EOKw8e1zskiMDOnQ9yvS5XJ9fUT7NW5k56Fyi8N0T3vK6undsAW7GTj9ukDFLhEJACN6xzGidyyF4YN5NfuXmBiweQEs/bXV0Y6rYffHhC/6CQCrkm9k2GWzLE4k0n5BYVGUxwwBoGr9mxanERHxT6FBdm4YmYndBtuK6njq4324Wo4z+9ww4Nwfwh0fQOIAcDfCW3f7liKerB9pUw0UroJ5l8Mrt0HdEQhPgok/hZtegWHfAEcIpmny1qYjbDpUg90GN4/NJCPO2WnvW6TLpA7DExZPiNmEZ9ci3McrKMtJ7S6tJ6fhM98XAdKvC1TsEpEAceXQNJKjQlgXfh6f9LvPd/LDP8Oap6wN9iVNFYXw4rewm60URI7lrGkPEuxUU3rpntxZFwAQfGSVxUlERPxXZryTGedkE+Kwsa/cxZzle6hsaDn+nZPyYNbHMP6HvvYGuxfB30bCytm+fl5v/wievR7+MQ4eSoeHM33LHg98BPZgGPoNuPlVOP8nEPFF0/mlO0r5dG8lhgHXj8ygb1Jk17x5kc5ms2HrfwkAfao++mp/PDmp+uZWimqaAq5fF6jYJSIBIthh45ujMwlx2Hg77Cr295vhu+Hd+6DgPUuzfc7T7KLu6RsId1dQEpxJwpW/ISIhzepYIqctZtgVAPSq20R9TZXFaURE/FffpAhmTexDdFgQZfUtzF6+m4OVruPf2WaHi34Jty6CmCxoqob37vf181r7JOx6D0q3+RrPA4REQu9zfcsVr3wcUgcf88vqp3srWLy9FIArhqQxJD2mU9+rSFczPu/b5VrHjqJai9N0L3vL6gnyBl6/LlCxS0QCSFKUr38XwL/CbqUu51Lflt4Lvg2H11uazfR6OTj/dhJrt+GyReC54H+J7zfK0kwiZyqy9zDqgpMINlsoWrXA6jgiIn4tOSqUWRP7kBYdSn2zh399uJdtR07yi3n6CLhrNYy6A1IGQ/YEGDQVxt7l6+91/TMwayX8YBN86zUYdA3Yv2g2X1rbxGvrD/HmxiMATMpLYlyf+M5+myJdL2cipuEgvrWE0t3rMP24b6+/CdR+XaBil4gEmOGZsYzOjsWLjb/F3k9r2ihobYT5V8D2tyzLtfON35N1+G082KgY/RPSxl5vWRaRDmMY1GT4dmX07lpscRgREf8XHRbEHefn0D85ArfH5N+rDvDJ7uM31Xa5XczZOpd813qGOuvIdzYyp98oXJP+n6+/18ArIXmgr7+OIxjwbTKxs6SOpz7ax6OLd7F6XxWmCWNz4pg0IKkL36lIFwqJxOw9HoDUoiWU1TVbHKj7CNR+XaBil4gEoMuHpJEaHUqt28Yzff+EmTjAN9X/xZvhtVnQ0rVr+bd/9Dr9Nv4BgMN5t5KRPzNg1sKLhA26FICUyrW4Wz0WpxER8X8hDjvTx2UxJjsO04S3NhXx9qYjVLta8Hp9M1JcbhczFs5g9obZlLhK8JpeSlwlzN4wmxkLZ+ByH7sE0u3xsnpfJX9ZvIu5H+9nV2k9hgFnpUXx3Qk5XDk0DUNjDwlgtjzfeCSvYQ07iussTtM9VDa0UNngJicA+3UBOKwOICLS0YLsNr45JpO/Ld3NzmpYNOElLt7/Z98W3hufg/0fwvVPQ68RnZ5l1/aN9F52Fza8HE6+kMwrfurbVlwkQMSdNQn3myFEeyrYt2Ex2SMnWx1JRMTv2WwGVw1LIzY8mIVbivl4dwUf767AZvhmf+1qfo0d1TswOXY5lhcvOyp38OsP55CfehNuj5cmt5cth2toaPF94BDisDEyK5Zz+iQQFx5sxdsT6Xq5F8PC/yGraQcfFh5kQm6i1Yn83p6j/brSA7BfF2hml4gEqISIEK49uxcAy3bVUDDqN/DNBeBMgJqD8H8XwfKHwdt5M1H27N1F9BvTcXrqqIjIJe2630G4emVIYDGCnZQnnwNA46bXrQ0jItKNGIbB+bmJfGN0BokRwdgM8JpQ5XKzoWbhVwpdnzMxWXzoDRZvL2XFznJW7aukocVDrDOIywancv8leVw+JE2FLulZ4nLwxOZgx0PI3vdobNFs86+zuzRw+3WBZnaJSAAbkh7DvvIGPt1byfOrC7lx9Bjy7loLL98Ke5fC8odg53swbR7E9u7Q197z0UukLLuPcE8NrqBYYi77JUbSgA59DRF/Yes/BYqXEV26BtM0tVRGRKQdhqTHMCQ9Bq/XpK6plerGFt55u/qkj2k2qxmTHUeQ3YbDbtArJoyBqVHYbPr3V3oue96lsPJv9K9dya7SOu08ehKmabK3rJ5zArRfF2hml4gEuMsGp5KTEE5zq5f5Kw+w7KAb8+ZX4LI/Q1AYHPkM/jEGPvs3dMDOLWZLA0eenUWfxbcT7qmh0plN8KUPYc+7pAPejYh/ih9+BQCpTbspOlBgcRoRke7JZjOIdgbROz6cJOfJl2AlORO5engvLhuSyuSzUhjUK1qFLpHcKb7/uD6j4EiVxWH8W3FtE/XNHvo0bPCdCLB+XaBil4gEOIfdxi3jv2gC+/7WEp5bc5Dm4bfAzE8gaSC4G+HNO2H+VVBz+LRfyyzeTP3jE0jb9RwAe9OuIHbGCziGfyPgfniI/DdHTC8qIgdgw6Rq9UtWxxER6fam5k7FdoJf1WzYmJo7tYsTiXQDmWPxBEUQ4a2jruCDtg0f5Kv2lDb4+nW5ArNfF6jYJSI9gMNu4+rhvbhmeC/sNthyuJY5y/dSGZoO3/0Qxv8QbA7YtwL+NtLXyL49s7xME88n/8D7zwuIrNtNrT2GTcMeJGfGExhJeZ32vkT8SUufiwEIOfSRxUlERLq/6QOn0z+u/1cKXjZs9I/rz/SB0y1KJuLH7EEY/fIByCpfzqGqRosD+ReX28WcjXPIX5DP9z65kEUNP+CJ6HBcEYHXrwtU7BKRHmR0dhy3n5dDZKiD4tom/r5sN7srmuCiX8IdKyChP7hd8NbduJ6+jDmrfk/+gnyGzh9K/oJ85myc85Wtvqk9guffU7G//wB2082OsLM5MOmfDLnqHgh2WvNGRSwQe3QpY0bdRqqqq60NIyLSzTmDnMybMo9Zw2aR7EzGZthIdiYza9gs5k2ZhzNIYwyR47H197UOyWv8jO3FtRan8R8ut4sZC2cwe8NsSlwlmJjUU8/smGhmJEbjCgq1OmKHM0yzA5rUdBO1tbVER0dTU1NDVFSU1XFExCI1Ljf/XnWAQ1WN2Ay4dHAq5/SJxzC9sOy3uFb+jRlJMRQEB+P9r+WHNgz6h/diXuJEnMVboGgDVBcC4CaI9xO+Rb9Lf0BuTpY1b0zESl4vDQ/1JdxdwdZxj3DW5NusTtSpNKY4dbpWIiLSZRrKMf/YFwOTp85+jVuvvNDqRH5hzsY5zN4wGy/er9xmA2YNu5OZQ2d2fbDTcKrjCs3sEpEeJ9oZxB0TchieGYPXhLc3FfGH9wqYt7KQhcm386dR36UgOOSYQheAF5OC+oPMX/832P5mW6GrMLgvc3s/zJAbf6VCl/RcNhu1mZN8f9692NosIiIi0jOFJ+BNPRuA2IPvU+NyWxzIP7y88+XjFroAvEdvDzQOqwOIiFghyG5j2oh0esWE8Z8tRVS73FS73BSU1LO4/gO8J+gn7wWei4oj07iSwyF9OBI+kOjEXtx8bi4JESFd+h5E/I1z0GWw5yVSq9bS1NJKaLCGGSIiItK17AMuhaJ19G9YzY7iWsbkxFsdyXJljWVndHt3pFGoiPRYhmEwvm8CZ2fGUlzbRMnR452NJ9mq2DCotps4pzzIhOgwUqJDiQp1YGi3RRGiz7qI1jeDiGstZeemFeSOnGR1JBEREelpcifD0l/Tt3EzCw6VqNgFJIYlUuIqOentgUbLGEWkxwsLtpOdEM7YnHiuGtaLJGfSSe+f5ExiYv8k+qdEEh0WpEKXyOeCw6lM8m1d3bTpDYvDiIiISI+UPAhPeArBZgveXe/j9hx/+V5PMjV36ld2d/2cDRtTc6d2caLOp2KXiMiX9MQfBiIdxZbn2wUppvRTvN4esweOiIiI+AvDwJZ3KQB9qz9hb1mDxYGsN33gdPrH9cfgvz6kN03fBlxx/Zk+cLp14TqJil0iIl/y+Q+DLxe8bNgC9oeBSEeJG3YFAOlNOyncv9viNCIiItITGbmTAchzrWNHUY3FaaznDHLy5MVzGRA6lQgisJkmyV6TWQNnMG/KPJxBTqsjdjgVu0REvsQZ5GTelHnMGjaLZGcyNsNGsjOZWcNmBewPA5GOYovNpDqiHzZMqtcusDqOiIiI9ETZE/Dag4n1lFO+cxWmqdnmpbUmOUHX8Mem0Wzcf5DFjn7MHHlPwP5uowb1IiLH4QxyMnPoTGYOnWl1FJFup6XvZNiwi9CDH2KapvraiYiISNcKdkLWBNizmF4lSymquZS0mDCrU1lqR1EdAAMa1/lOpAyBAB6jaWaXiIiIdKjoo0sZs+rXU1ZZa3EaERER6Yk+7yOa17CGzYe1lLGguJYIdwXx1Zt9J3LOtzZQJ1OxS0RERDpUSOYoGoNiCDUbKVqnXRlFRETEAv18fbsym3eye8/uHr2UsayumbL6FgbUrfSdiO8LGWOsDdXJVOwSERGRjmWzU5dxoe/PO9+zNouIiIj0TDEZeJMHYcMk9eA7FNU0WZ3IMgXFviWMZzd96juRPhpCIixM1PlU7BIREZEOFz7Yt5SxV/U66ptbLU4jIiIiPZFt8FQAhtR9yKZDPXcp447iWhzeZjKqVvlO9B5nbaAuoGKXiIiIdLjwAfl4DAcJrUXsXv+h1XFERESkJxp4NQA5TVvYu3dXj1zK2OT2sK+8gT71n2H3NIIzvm2JZyBTsUtEREQ6XmgUtcljAWjZ+JLFYQTgt7/9Leeccw5Op5OYmJhTesyMGTMwDOOYY8qUKZ0bVEREpKPEZeNNHtK2lPFID1zKuKukHq8JwxuP9utKHw0RSdaG6gIqdomIiEinCB1xIwBZ5Ssoq+15g0t/09LSwrRp05g1a1a7HjdlyhSKiorajueff76TEoqIiHQ825DPlzJ+wOYeuJRxe3EtmCa5tZ/4TmSMAcOwNlQXULFLREREOkXY4KvwGEEkuQ+z+7MlVsfp8X75y19yzz33MHjw4HY9LiQkhJSUlLYjNja2kxKKiIh0gqNLGbObtrF3984etZTR6zXZWVxHatMuwhqLwREK/XvGDG0Vu0RERKRzhEZRlzkJANvml3rU4DKQLF++nKSkJPr378+sWbOoqKg46f2bm5upra095hAREbFMbG+8qcOwYdLr8Ns9ainjoapGGlo8DKk/OqsrdSjE97M2VBdRsUtEREQ6TcSobwDQv/pD9pXVWZxG2mvKlCnMnz+fJUuW8Pvf/54VK1ZwySWX4PF4TviYhx56iOjo6LYjIyOjCxOLiIh81ee7Mg6u+5DNh6qtDdOFthf7PnAa7DrarytjDNgdFibqOip2iYiISKdx9J+M2x5OjKeCwtVvWx0n4Nx///1faSD/5WPHjh2n/fw33ngjV155JYMHD+bqq6/m7bffZs2aNSxfvvyEj3nggQeoqalpOw4ePHjary8iItIhBl4FQFbTdvbuLugxs80LiuuIdJcTX7MVMKDPJKsjdZmeUdITERERawSF0dT3UoIKFhC+83VaWm8k2KHP2jrKvffey4wZM056n5ycnA57vZycHBISEti9ezeTJh1/wBwSEkJISEiHvaaIiMgZi8nEm3o2tqLPSD/0DoerR5Me67Q6VaeqdrVQVNPEqLqjSxgT+kHGKGtDdSEVu0RERKRTRYz6BhQsYED9J2w/VM7QrMDf7rqrJCYmkpiY2GWvd+jQISoqKkhNTe2y1xQREekItiFToegzBtd9yJbDNQFf7NpR7GsfcXbjp74T6aMhONzCRF1LH62KiIhIpzKyz6c5OJYIbx0lq1+xOk6PVVhYyIYNGygsLMTj8bBhwwY2bNhAfX19233y8vJ47bXXAKivr+fHP/4xn376Kfv372fJkiVcddVV9O3bl8mTJ1v1NkRERE7P0aWMvZt3sG/3joBfyrijqBaHt5mM6tW+E73PsTZQF1OxS0RERDqX3YF34DUAJBS+S02j2+JAPdPPf/5zhg8fzoMPPkh9fT3Dhw9n+PDhrF27tu0+BQUF1NTUAGC329m0aRNXXnklubm53HbbbYwYMYIPP/xQyxRFRKT7iU7HmzYSGybph97mcHWj1Yk6TXOrhz1lDfStX4vd0wThidD3IqtjdSktYxQREZFOF3b2jbDhKc5qWM2avYc596wsqyP1OPPmzWPevHknvc9/f8odFhbGe++918mpREREuo5tyFQ4sjbglzLuLWug1Wsy9PNdGNNHQUTXtT3wB5rZJSIiIp0vYzRNzjRCzCZq1i4I+KUDIiIi4oc+35WxuYB9u7cH7HhkR3EtmCb9a482p88YA4ZhbagupmKXiIiIdD7DwD70egCySt6jqKbJ4kAiIiLS40Sl4U0fDUDGobcCcimjaZrsKK4jrbGAsKZScIRC7iVWx+pyKnaJiIhIlwga5it29XetZ/POvRanERERkZ7INngqLsNgi7mcG/9zGUPnDyV/QT5zNs7B5XZZHe+MFdU0UdvYyuD6o7O60oZDQl9rQ1lAxS4RERHpGsln0RTTFwetuDe8iNcbmEsHRERExH+5+l3EjNRkno90U+Mux2t6KXGVMHvDbGYsnNHtC16bDlUDMPjzfl0Zo8Fmty6QRVTsEhERkS4TPPwbAPSvXMqu0nqL04iIiEhPM//g+xQEB+P9Ug8rL14KKguYv22+RcnOXKvHy7oDVUS5y4iv3Q4YPW4Xxs+p2CUiIiJdxjb4OlyGweLQQm5dHHhLB0RERMS/vbzzZbwn6NXuxcvLO1/u2kAdaEdxHfXNHoY1fuo7kZgLvUZYG8oiKnaJiIhIl3FFJvPtjCzmxERR66kKuKUDIiIi4t/KGsvO6HZ/tmZ/JQAjmlf5TqSPhmCnhYmso2KXiIiIdJn52+az0+4NyKUDIiIi4v8SwxLP6HZ/VdXQwq7SeoK8jSSWHu3XlXWutaEspGKXiIiIdJmXd76Ml+M3pu/uSwdERETE/03NnYrtBKUQGzam5k7t4kQdY83+SkwTJtg2Y3iaISIZ+uRbHcsyKnaJiIhIlwnkpQMiIiLi/6YPnE7/uP7YMMA8+gGcCQY2+sf1Z/rA6dYGPA1er8m6wioARtUs9J3MHAcRCRamspaKXSIiItJlAnXpgIiIiHQPziAn86bMY9aw75FsC8FmmsR5bYyNvZF5U+bhDOp+Pa52FNdR29hKqlFBVOES38ncS6wNZTEVu0RERKTLBOrSAREREek+nEFOZg6dyeKLn2bj/oMsLSwku2EkNa4TbNPo5z5vTD/FvQTD9EDSWTDgMotTWUvFLhEREekyXywd+K8hiAkGRrddOiAiIiLdVMogyBiDHS9jSl/kw13lVidqtxqXm4KSOgzTQ5+Dr/pO5k6BkEhrg1msWxS79u/fz2233UZ2djZhYWH06dOHBx98kJaWFqujiYiISDt8sXRgFsmh8dhMk2RPK+OCJzF38txuuXRAREREurGx3wNgVM37bDlQSk2j2+JA7bP2wNHG9MZG7HWHIDgSBmumfLcodu3YsQOv18s///lPtm7dyqOPPsqcOXP46U9/anU0ERERaae2pQM3LOezoEEsPniE7xzeT1G11+poIiIi0tPkXQ4RyUR6axhU9g4r93Sf2V1er8ma/b7G9OOq3/ad7HMBJOZZmMo/dIti15QpU5g7dy4XX3wxOTk5XHnlldx33328+uqrJ31cc3MztbW1xxwiIiLiP+zn/RCA4fUfsH7LDmvDiIiISM9jd8DoOwAYV/k6q/ZV0uT2WBzq1Owqraem0U0yFUQdPNqYvv9lYOsWpZ5O1W2vQE1NDXFxcSe9z0MPPUR0dHTbkZGR0UXpRERE5JRkjMGdPBQHrcRvfoKqBrUoEBERkS42YgamPYiMlj0kVnzGugNVVic6JauPNqa/xL1Yjem/pFsWu3bv3s3jjz/Od7/73ZPe74EHHqCmpqbtOHjwYBclFBERkVNiGARNuAeAMTX/Yc3OQxYHEhERkR4nPAHjrOsAGFf2Ih/vLsfrNS0OdXK1TW52FNVimB76HlJj+i+ztNh1//33YxjGSY8dO45d0nD48GGmTJnCtGnTuP3220/6/CEhIURFRR1ziIiIiJ/Ju4KW8DTCvXW4Vz+J26PeXSIiItLFxs4EYEjDJ7irj7DlSI3FgU5u3YEqvCaca2zEXndYjem/xGHli997773MmDHjpPfJyclp+/ORI0e44IILOOecc3jiiSc6OZ2IiIh0CbsDx7nfh/ceYHT5K2w6eA8jshKsTiUiIiI9SdpwSDsb+5HPGF36Eh/uymRwr2gMw7A62VeYpsnao0sYz6l+y3dSjemPYWmxKzExkcTExFO67+HDh7ngggsYMWIEc+fOxaaGayIiIgHDdva3aF3yGxLdRaz65AXM3nf65eBSREREAti4O+GV2xhTs5AVFbdzoMJFVkK41am+Yk9ZPZUNbhLNCqIPLvWdVGP6Y3SLK3H48GEmTpxIZmYmf/rTnygrK6O4uJji4mKro4mIiEhHCInEO+JWAAYWPsvBykaLA4mIiEiPM+BKcCYQ5anirIr3+HBXmdWJjmv1Pl8DfTWmP7FuUexatGgRu3fvZsmSJaSnp5Oamtp2iIiISGAIPmcWXsNOTtM2CtYssjqOiIiI9DSOYBj1HQDGVb7O9uI6yuqaLQ51rKqGFrYV1WCYHnIPqzH9iXSLYteMGTMwTfO4h4iIiASI6F409rsSgKTN/6SuyW1xIBEREelxRt4KNgdZzQWk1m1h4ZYiv6o9vLO5CI8XJtjUmP5kukWxS0RERHqG8At+BMCghpVs3rzB2jAiIiLS80Qmw4CrABhf+gLbiurYfNg/dmbcVVLH1iO12AyYWP+O72SfCyBpgLXB/JCKXSIiIuI/UodQlzIGO17sKx/H6/WfT1JFRESkhxg7C4ChDR8T7q7krY1HcLW0Whqp1ePlrY1HALggrZXQvUdbPvS/DLSpz1eo2CUiIiJ+xTnRN7traNX77Nh30OI0IiIi0uOkj4SUIdhNN5PLn6a+2cPbm4osjfTJngrK6luICLFzfsNCML1qTH8SKnaJiIiIX7HnXkx9RBahZiNli/9CS6vX6kjd3v79+7ntttvIzs4mLCyMPn368OCDD9LS0nLSxzU1NXHnnXcSHx9PREQE1113HSUlJV2UWkRExCKGARMfAGBk+eskN+5mfWE1O0vqLIlT0+hm6Y5SAKYMTCRow799N6gx/Qmp2CUiIiL+xWYj6Hzf7K7RJc+zbM1GiwN1fzt27MDr9fLPf/6TrVu38uijjzJnzhx++tOfnvRx99xzD2+99RYLFixgxYoVHDlyhGuvvbaLUouIiFgo71LodxGG6WF6ycMYppfX1x+mudXT5VEWbimiudVLZpyTs4sXQO0hCIlSY/qTULFLRERE/E7IiJtpij+LMK+L1I9+atknqYFiypQpzJ07l4svvpicnByuvPJK7rvvPl599dUTPqampoYnn3ySRx55hAsvvJARI0Ywd+5cPvnkEz799NMuTC8iImKRy/8CjlDiancwoeYNqlxu3t/atTOc95U3sOFgDYYBV/ezYyz7ne+G4TerMf1JqNglIiIi/sdmJ3TqHLzYGNLwCRve/7fljWEDTU1NDXFxcSe8fd26dbjdbvLz89vO5eXlkZmZycqVK0/4uObmZmpra485REREuqXodLjg/wFwUdE/CW+tZOXeCgorXF3y8l6vyZsbfE3pR2fFkfrpb6GlDuL7wtg71Zj+JFTsEhEREf+UOgRz9HcBuGj/n3hn7W5MU7szdoTdu3fz+OOP893vfveE9ykuLiY4OJiYmJhjzicnJ1NcXHzCxz300ENER0e3HRkZGR0VW0REpOuN/R4k5mF313Nz2aOYJrzy2SFaPZ3fU/TTfRUU1zbhDLYzJWIXbH4JMHyZYtI7/fW7MxW7RERExG/Z839Ga3gKsZ5yUlY/xMZDNVZH8iv3338/hmGc9NixY8cxjzl8+DBTpkxh2rRp3H777R2e6YEHHqCmpqbtOHhQO2qKiEg3ZnfAVf8ADLLKlnFW4xpK65pZXlDWqS9b1+Rm0TbfksnJeXGEvv8/vhtyJ8Owb3bqawcCFbtERETEfwWH47j6bwCMr36DVR8tosbltjiU/7j33nvZvn37SY+cnJy2+x85coQLLriAc845hyeeeOKkz52SkkJLSwvV1dXHnC8pKSElJeWEjwsJCSEqKuqYQ0REpFtLHwEjZgAw7fAfcXibWb6zlOKapk57yfe2ltDk9tIrJpSRxS9C2Q5fU/pzvg9BYZ32uoFCxS4RERHxb/0uwsy7Ehsml+37HS+v2afljEclJiaSl5d30iM4OBjwzeiaOHFiW6N5m+3kw8ARI0YQFBTEkiVL2s4VFBRQWFjIuHHjOvV9iYiI+J2LfgnOeEJcR5ha8zQer285Y0trxy9nLKxwse5AFQDX9DGwrfi974YRM6D3+A5/vUCkYpeIiIj4PePyP+MNiiC9ZS/JW/+PT/ZUWB2pW/m80JWZmcmf/vQnysrKKC4uPqb31uHDh8nLy2P16tUAREdHc9ttt/GjH/2IZcuWsW7dOm655RbGjRvH2LFjrXorIiIi1giNhkv/BMCQw8/Rq7WQQ1WNPPHBng6ddX6gooHnVhcCMKJ3LL1W/RrcDZCY5+vVpab0p0TFLhEREfF/EUnYJv8GgIvK5rFy3XpKaztv6UCgWbRoEbt372bJkiWkp6eTmpradnzO7XZTUFCAy/XFDlOPPvool19+Oddddx0TJkwgJSWFV1991Yq3ICIiYr2zroHs8zG8rdxW+jDhQTYOVzfxt2W7OFDRcEZPbZomywtKeeKDvdQ0ukmICOYy53bY9joYNl+hKyr1a59HfAyzB60DqK2tJTo6mpqaGvWPEBER6W68Xswn8zEOr2ObcyRLzp7NzAv6EmTv+s/uNKY4dbpWIiISUKr2w99Gg6eZhkkP82RLPkU1TdhtcPWwXozMimv3U9Y1uVmw9hC7SusBGJYRzVWDEgj917lQuQfyLofrnoSg0A5+M93PqY4rNLNLREREugebDePq2Zg2BwNda4nd/xZ/X7abI9WNVicTERGRniI2C87/CQDhH/2WmX2rGdQr6mgPr8O8ufEIXu+J5xS53C7mbJxD/oJ8hs4fysQXLuS7bz3M9pJyguwG153di+tHZhC65u++QldoLIz7vgpd7aRil4iIiHQfif0xzvkBAFcVPU5dZTH/WL6bZQWlJx1YioiIiHSY8T/w9dBqriP46Sl8k/e4aEAiACv3VPDUx/twtbR+5WEut4sZC2cwe8NsSlwleE0vFc1lbHYtYE3zb7j1vDRGZsVhVBfCB3/2PWjkLZA5pivfXUBQsUtERES6l/N/AjG9ifRU8YM9dxDnOsD7W0v45wd7Ka9vtjqdiIiIBDp7ENz6HmSdB95WjP/8mAs3/4RvnR1PiMPGnrIG/r5sN2v2V7J6XyWf7Cnno13l/OrDOeyo3IGXL+/gaFLl3s/Cgy9BXQm88h1obYTkQTBmlprSnwb17BIREZHup3QHPH0FNJTS6nDyQu9fszV8DMF2g0sHpzI6Ow6jEweGGlOcOl0rEREJWKYJyx+CD/4Iphdisyi74hnm7QqmsuGrOzQurr+LJrPyhE+XHBzD4kPF0FAKjlC47BEYflNnvoNuRz27REREJHAl5cH3VkLyYBytLm7a+2OudL1KS6uX1zccYd4n+6lp7LhtwEVERES+wjDggp/Ct16HsDio2k/icxfx/YTPGJMdR//kCAamRjK4VzTDMqJpNqtP+nRlzVW+Qld0Jlz+CAy9sUveRiDSzC4RERHpvlqb4dXbYdsbABRnX8OcyB/QjIOwIDu3nptFeqyzw19WY4pTp2slIiI9Ql0JPH8DHFnv+3r4t+CyP4MjpO0u+QvyKXGVHP/xpkmyx8Pi4DyY9EtIG9IFobsfzewSERGRwOcIgWlPw8SfAgYp+17jp6U/om9EExGhDpKjtHORiIiIdIHIZPjOEhgz0/f1+mfgj33huRvg48fg0Fqm9r0W2wnKMDZgasxgmPqUCl0dQDO7REREJDBsf8s3y8vdiBmZSt20l4jK7JzBosYUp07XSkREepztb8Ebd0FT9TGnXUFhzOiVRoHh/qJFvWliA/qHpTDvqldxhupn5cloZpeIiIj0LAOugO8shYgUjLoiop65GA6tszqViIiI9DQDroD7dsENz8HI2yB9NARH4HQ3Mu/AXmZVVZPc2orNNEk2HMzqcy3zrn1Tha4O5LA6gIiIiEiHSR4Isz6Bf18LLfUQl211IhEREemJHMEw4DLfAeBxQ+FKnLsWM/PIembWl0DffDjvXgiPtzZrAFKxS0RERAJLeDzctggaq8EZZ3UaEREREbAHQfYE3wHg9fp2czQMa3MFKBW7REREJPA4giEyyeoUIiIiIsdnU1epzqSrKyIiIiIiIiIiAUPFLhERERERERERCRgqdomIiIiIiIiISMBQsUtERERERERERAKGil0iIiIiIiIiIhIwVOwSEREREREREZGAoWKXiIiIiIiIiIgEDBW7REREREREREQkYKjYJSIiIiIiIiIiAUPFLhERERERERERCRgqdomIiIiIiIiISMBwWB2gK5mmCUBtba3FSURERKQ7+3ws8fnYQk5M4y8RERHpKKc6ButRxa66ujoAMjIyLE4iIiIigaCuro7o6GirY/g1jb9ERESko33dGMwwe9BHkl6vlyNHjhAZGYlhGB3+/LW1tWRkZHDw4EGioqI6/PkDka5Z++matZ+uWfvpmrWfrln7dedrZpomdXV1pKWlYbOpK8TJaPzlf3TN2k/XrP10zdpP16z9dM3ar7tfs1Mdg/WomV02m4309PROf52oqKhu+U1jJV2z9tM1az9ds/bTNWs/XbP2667XTDO6To3GX/5L16z9dM3aT9es/XTN2k/XrP268zU7lTGYPooUEREREREREZGAoWKXiIiIiIiIiIgEDBW7OlBISAgPPvggISEhVkfpNnTN2k/XrP10zdpP16z9dM3aT9dMOoK+j9pP16z9dM3aT9es/XTN2k/XrP16yjXrUQ3qRUREREREREQksGlml4iIiIiIiIiIBAwVu0REREREREREJGCo2CUiIiIiIiIiIgFDxS4REREREREREQkYKnZ1oL///e9kZWURGhrKmDFjWL16tdWR/MYHH3zAFVdcQVpaGoZh8Prrrx9zu2ma/PznPyc1NZWwsDDy8/PZtWuXNWH9wEMPPcSoUaOIjIwkKSmJq6++moKCgmPu09TUxJ133kl8fDwRERFcd911lJSUWJTYerNnz2bIkCFERUURFRXFuHHj+M9//tN2u67X13v44YcxDIMf/vCHbed03Y71i1/8AsMwjjny8vLabtf1Or7Dhw9z8803Ex8fT1hYGIMHD2bt2rVtt+tngJwJjb9OTOOv9tMYrP00BjszGn+dGo3BTk9PHoOp2NVBXnzxRX70ox/x4IMP8tlnnzF06FAmT55MaWmp1dH8QkNDA0OHDuXvf//7cW//wx/+wGOPPcacOXNYtWoV4eHhTJ48maampi5O6h9WrFjBnXfeyaeffsqiRYtwu91cfPHFNDQ0tN3nnnvu4a233mLBggWsWLGCI0eOcO2111qY2lrp6ek8/PDDrFu3jrVr13LhhRdy1VVXsXXrVkDX6+usWbOGf/7znwwZMuSY87puX3XWWWdRVFTUdnz00Udtt+l6fVVVVRXjx48nKCiI//znP2zbto0///nPxMbGtt1HPwPkdGn8dXIaf7WfxmDtpzHY6dP4q300BmufHj8GM6VDjB492rzzzjvbvvZ4PGZaWpr50EMPWZjKPwHma6+91va11+s1U1JSzD/+8Y9t56qrq82QkBDz+eeftyCh/yktLTUBc8WKFaZp+q5PUFCQuWDBgrb7bN++3QTMlStXWhXT78TGxpr/93//p+v1Nerq6sx+/fqZixYtMs8//3zzBz/4gWma+j47ngcffNAcOnTocW/T9Tq+//mf/zHPPffcE96unwFyJjT+OnUaf50ejcFOj8ZgX0/jr/bRGKz9evoYTDO7OkBLSwvr1q0jPz+/7ZzNZiM/P5+VK1damKx72LdvH8XFxcdcv+joaMaMGaPrd1RNTQ0AcXFxAKxbtw63233MNcvLyyMzM1PXDPB4PLzwwgs0NDQwbtw4Xa+vceedd3LZZZcdc31A32cnsmvXLtLS0sjJyeGmm26isLAQ0PU6kTfffJORI0cybdo0kpKSGD58OP/617/abtfPADldGn+dGf3dOzUag7WPxmCnTuOv9tMYrH16+hhMxa4OUF5ejsfjITk5+ZjzycnJFBcXW5Sq+/j8Gun6HZ/X6+WHP/wh48ePZ9CgQYDvmgUHBxMTE3PMfXv6Ndu8eTMRERGEhIQwc+ZMXnvtNQYOHKjrdRIvvPACn332GQ899NBXbtN1+6oxY8Ywb948Fi5cyOzZs9m3bx/nnXcedXV1ul4nsHfvXmbPnk2/fv147733mDVrFnfffTdPP/00oJ8Bcvo0/joz+rv39TQGO3Uag7WPxl/tpzFY+/X0MZjD6gAicnJ33nknW7ZsOWZNuhxf//792bBhAzU1Nbz88st8+9vfZsWKFVbH8lsHDx7kBz/4AYsWLSI0NNTqON3CJZdc0vbnIUOGMGbMGHr37s1LL71EWFiYhcn8l9frZeTIkfzud78DYPjw4WzZsoU5c+bw7W9/2+J0IiInpjHYqdMY7NRp/HV6NAZrv54+BtPMrg6QkJCA3W7/ym4PJSUlpKSkWJSq+/j8Gun6fdVdd93F22+/zbJly0hPT287n5KSQktLC9XV1cfcv6dfs+DgYPr27cuIESN46KGHGDp0KH/96191vU5g3bp1lJaWcvbZZ+NwOHA4HKxYsYLHHnsMh8NBcnKyrtvXiImJITc3l927d+v77ARSU1MZOHDgMecGDBjQtvRAPwPkdGn8dWb0d+/kNAZrH43BTp3GXx1DY7Cv19PHYCp2dYDg4GBGjBjBkiVL2s55vV6WLFnCuHHjLEzWPWRnZ5OSknLM9autrWXVqlU99vqZpsldd93Fa6+9xtKlS8nOzj7m9hEjRhAUFHTMNSsoKKCwsLDHXrPj8Xq9NDc363qdwKRJk9i8eTMbNmxoO0aOHMlNN93U9mddt5Orr69nz549pKam6vvsBMaPH09BQcEx53bu3Env3r0B/QyQ06fx15nR373j0xisY2gMdmIaf3UMjcG+Xo8fg1ndIT9QvPDCC2ZISIg5b948c9u2beYdd9xhxsTEmMXFxVZH8wt1dXXm+vXrzfXr15uA+cgjj5jr1683Dxw4YJqmaT788MNmTEyM+cYbb5ibNm0yr7rqKjM7O9tsbGy0OLk1Zs2aZUZHR5vLly83i4qK2g6Xy9V2n5kzZ5qZmZnm0qVLzbVr15rjxo0zx40bZ2Fqa91///3mihUrzH379pmbNm0y77//ftMwDPP99983TVPX61T9925Apqnr9mX33nuvuXz5cnPfvn3mxx9/bObn55sJCQlmaWmpaZq6XsezevVq0+FwmL/97W/NXbt2mc8++6zpdDrNf//732330c8AOV0af52cxl/tpzFY+2kMduY0/vp6GoO1X08fg6nY1YEef/xxMzMz0wwODjZHjx5tfvrpp1ZH8hvLli0zga8c3/72t03T9G17+rOf/cxMTk42Q0JCzEmTJpkFBQXWhrbQ8a4VYM6dO7ftPo2Njeb3vvc9MzY21nQ6neY111xjFhUVWRfaYrfeeqvZu3dvMzg42ExMTDQnTZrUNsgyTV2vU/XlwZau27FuuOEGMzU11QwODjZ79epl3nDDDebu3bvbbtf1Or633nrLHDRokBkSEmLm5eWZTzzxxDG362eAnAmNv05M46/20xis/TQGO3Maf309jcFOT08egxmmaZpdN49MRERERERERESk86hnl4iIiIiIiIiIBAwVu0REREREREREJGCo2CUiIiIiIiIiIgFDxS4REREREREREQkYKnaJiIiIiIiIiEjAULFLREREREREREQChopdIiIiIiIiIiISMFTsEhERERERERGRgKFil4j0GMuXL8cwDKqrq62OIiIiItJjaAwmIl3NME3TtDqEiEhnmDhxIsOGDeMvf/kLAC0tLVRWVpKcnIxhGNaGExEREQlQGoOJiNUcVgcQEekqwcHBpKSkWB1DREREpEfRGExEupqWMYpIQJoxYwYrVqzgr3/9K4ZhYBgG8+bNO2YK/bx584iJieHtt9+mf//+OJ1Opk6disvl4umnnyYrK4vY2FjuvvtuPB5P23M3Nzdz33330atXL8LDwxkzZgzLly+35o2KiIiI+BGNwUTEH2hml4gEpL/+9a/s3LmTQYMG8atf/QqArVu3fuV+LpeLxx57jBdeeIG6ujquvfZarrnmGmJiYnj33XfZu3cv1113HePHj+eGG24A4K677mLbtm288MILpKWl8dprrzFlyhQ2b95Mv379uvR9ioiIiPgTjcFExB+o2CUiASk6Oprg4GCcTmfbtPkdO3Z85X5ut5vZs2fTp08fAKZOncozzzxDSUkJERERDBw4kAsuuIBly5Zxww03UFhYyNy5cyksLCQtLQ2A++67j4ULFzJ37lx+97vfdd2bFBEREfEzGoOJiD9QsUtEejSn09k2yAJITk4mKyuLiIiIY86VlpYCsHnzZjweD7m5ucc8T3NzM/Hx8V0TWkRERKSb0xhMRDqTil0i0qMFBQUd87VhGMc95/V6Aaivr8dut7Nu3Trsdvsx9/vvwZmIiIiInJjGYCLSmVTsEpGAFRwcfExT044wfPhwPB4PpaWlnHfeeR363CIiIiKBQGMwEbGadmMUkYCVlZXFqlWr2L9/P+Xl5W2fDJ6J3NxcbrrpJqZPn86rr77Kvn37WL16NQ899BDvvPNOB6QWERER6d40BhMRq6nYJSIB67777sNutzNw4EASExMpLCzskOedO3cu06dP595776V///5cffXVrFmzhszMzA55fhEREZHuTGMwEbGaYZqmaXUIERERERERERGRjqCZXSIiIiIiIiIiEjBU7BIRERERERERkYChYpeIiIiIiIiIiAQMFbtERERERERERCRgqNglIiIiIiIiIiIBQ8UuEREREREREREJGCp2iYiIiIiIiIhIwFCxS0REREREREREAoaKXSIiIiIiIiIiEjBU7BIRERERERERkYChYpeIiIiIiIiIiASM/w9aLw9lubcrSgAAAABJRU5ErkJggg==", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "q = torch.tensor([0.05, 0.95])\n", "gen_quantiles = torch.quantile(gen_data.cpu(), q, dim=-1)\n", "gen_mean = gen_data.cpu().mean(dim=-1)\n", "\n", "sample_id = torch.randint(0, len(real_data), ())\n", "n_channel = min(3, real_data.shape[-1])\n", "fig, axs = plt.subplots(1, n_channel, figsize=[12, 4], layout=\"constrained\")\n", "if n_channel == 1:\n", " axs = [axs]\n", "\n", "t_hr = range(real_data.shape[1])\n", "t_lr = range(0, real_data.shape[1], sr_factor) # for subsample\n", "\n", "for i in range(n_channel):\n", " ax = axs[i]\n", " # ground truth\n", " ax.plot(t_hr, real_data[sample_id, :, i].cpu(), c=\"C0\", label=\"ground truth\", alpha=0.6)\n", " # low-res condition\n", " ax.scatter(t_lr[:cond_data.shape[1]], cond_data[sample_id, :, i].cpu(),\n", " c=\"C2\", zorder=5, s=30, label=\"low-res input\")\n", " # super-resolution output\n", " ax.fill_between(t_hr, gen_quantiles[0, sample_id, :, i],\n", " gen_quantiles[-1, sample_id, :, i],\n", " color=\"C1\", alpha=0.25, label=\"95% PI\")\n", " ax.plot(t_hr, gen_mean[sample_id, :, i], c=\"C1\", label=\"avg SR\")\n", " ax.set_xlabel(\"time\")\n", " ax.set_ylabel(\"value\")\n", " ax.set_title(f\"Channel {i + 1}\")\n", "\n", "fig.legend(*axs[0].get_legend_handles_labels(),\n", " loc=\"upper center\", ncol=4, bbox_to_anchor=(0.5, 1.12))\n", "plt.show()" ] } ], "metadata": { "kernelspec": { "display_name": "gents", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.10.19" } }, "nbformat": 4, "nbformat_minor": 5 }