semseg.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "\n",
    "Multi-GPU: Semantic Segmentation\n",
    "=====================\n",
    "\n",
    "Overview\n",
    "----------------\n",
    "1. Memahami struktur dataset cityscapes (https://www.cityscapes-dataset.com/)\n",
    "2. Mendesain arsitektur sebuah model CNN encoder-decoder sederhana\n",
    "3. Membuat beberapa fungsi pendukung (dataloader, train-val-test, dll)\n",
    "4. Train-val-test model\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Defaulting to user installation because normal site-packages is not writeable\n",
      "Requirement already satisfied: pip in /home/aisl/.local/lib/python3.8/site-packages (22.2.2)\n",
      "Defaulting to user installation because normal site-packages is not writeable\n",
      "Requirement already satisfied: opencv-python in /home/aisl/.local/lib/python3.8/site-packages (4.6.0.66)\n",
      "Requirement already satisfied: matplotlib in /home/aisl/.local/lib/python3.8/site-packages (3.5.2)\n",
      "Requirement already satisfied: tqdm in /home/aisl/.local/lib/python3.8/site-packages (4.64.0)\n",
      "Requirement already satisfied: numpy in /home/aisl/.local/lib/python3.8/site-packages (1.22.4)\n",
      "Requirement already satisfied: pandas in /home/aisl/.local/lib/python3.8/site-packages (1.4.3)\n",
      "Requirement already satisfied: ipywidgets in /home/aisl/.local/lib/python3.8/site-packages (7.7.1)\n",
      "Requirement already satisfied: python-dateutil>=2.7 in /home/aisl/.local/lib/python3.8/site-packages (from matplotlib) (2.8.2)\n",
      "Requirement already satisfied: pyparsing>=2.2.1 in /home/aisl/.local/lib/python3.8/site-packages (from matplotlib) (3.0.9)\n",
      "Requirement already satisfied: pillow>=6.2.0 in /home/aisl/.local/lib/python3.8/site-packages (from matplotlib) (9.2.0)\n",
      "Requirement already satisfied: packaging>=20.0 in /home/aisl/.local/lib/python3.8/site-packages (from matplotlib) (21.3)\n",
      "Requirement already satisfied: fonttools>=4.22.0 in /home/aisl/.local/lib/python3.8/site-packages (from matplotlib) (4.34.4)\n",
      "Requirement already satisfied: kiwisolver>=1.0.1 in /home/aisl/.local/lib/python3.8/site-packages (from matplotlib) (1.4.4)\n",
      "Requirement already satisfied: cycler>=0.10 in /home/aisl/.local/lib/python3.8/site-packages (from matplotlib) (0.11.0)\n",
      "Requirement already satisfied: pytz>=2020.1 in /home/aisl/.local/lib/python3.8/site-packages (from pandas) (2022.1)\n",
      "Requirement already satisfied: traitlets>=4.3.1 in /home/aisl/.local/lib/python3.8/site-packages (from ipywidgets) (5.3.0)\n",
      "Requirement already satisfied: jupyterlab-widgets>=1.0.0 in /home/aisl/.local/lib/python3.8/site-packages (from ipywidgets) (1.1.1)\n",
      "Requirement already satisfied: ipython-genutils~=0.2.0 in /usr/lib/python3/dist-packages (from ipywidgets) (0.2.0)\n",
      "Requirement already satisfied: widgetsnbextension~=3.6.0 in /home/aisl/.local/lib/python3.8/site-packages (from ipywidgets) (3.6.1)\n",
      "Requirement already satisfied: ipykernel>=4.5.1 in /home/aisl/.local/lib/python3.8/site-packages (from ipywidgets) (6.15.1)\n",
      "Requirement already satisfied: ipython>=4.0.0 in /home/aisl/.local/lib/python3.8/site-packages (from ipywidgets) (8.4.0)\n",
      "Requirement already satisfied: pyzmq>=17 in /home/aisl/.local/lib/python3.8/site-packages (from ipykernel>=4.5.1->ipywidgets) (23.2.0)\n",
      "Requirement already satisfied: jupyter-client>=6.1.12 in /home/aisl/.local/lib/python3.8/site-packages (from ipykernel>=4.5.1->ipywidgets) (7.3.4)\n",
      "Requirement already satisfied: tornado>=6.1 in /home/aisl/.local/lib/python3.8/site-packages (from ipykernel>=4.5.1->ipywidgets) (6.2)\n",
      "Requirement already satisfied: debugpy>=1.0 in /home/aisl/.local/lib/python3.8/site-packages (from ipykernel>=4.5.1->ipywidgets) (1.6.2)\n",
      "Requirement already satisfied: matplotlib-inline>=0.1 in /home/aisl/.local/lib/python3.8/site-packages (from ipykernel>=4.5.1->ipywidgets) (0.1.3)\n",
      "Requirement already satisfied: psutil in /home/aisl/.local/lib/python3.8/site-packages (from ipykernel>=4.5.1->ipywidgets) (5.9.1)\n",
      "Requirement already satisfied: nest-asyncio in /home/aisl/.local/lib/python3.8/site-packages (from ipykernel>=4.5.1->ipywidgets) (1.5.5)\n",
      "Requirement already satisfied: pickleshare in /home/aisl/.local/lib/python3.8/site-packages (from ipython>=4.0.0->ipywidgets) (0.7.5)\n",
      "Requirement already satisfied: decorator in /usr/lib/python3/dist-packages (from ipython>=4.0.0->ipywidgets) (4.1.2)\n",
      "Requirement already satisfied: stack-data in /home/aisl/.local/lib/python3.8/site-packages (from ipython>=4.0.0->ipywidgets) (0.3.0)\n",
      "Requirement already satisfied: backcall in /home/aisl/.local/lib/python3.8/site-packages (from ipython>=4.0.0->ipywidgets) (0.2.0)\n",
      "Requirement already satisfied: pygments>=2.4.0 in /home/aisl/.local/lib/python3.8/site-packages (from ipython>=4.0.0->ipywidgets) (2.12.0)\n",
      "Requirement already satisfied: pexpect>4.3 in /home/aisl/.local/lib/python3.8/site-packages (from ipython>=4.0.0->ipywidgets) (4.8.0)\n",
      "Requirement already satisfied: prompt-toolkit!=3.0.0,!=3.0.1,<3.1.0,>=2.0.0 in /home/aisl/.local/lib/python3.8/site-packages (from ipython>=4.0.0->ipywidgets) (3.0.30)\n",
      "Requirement already satisfied: setuptools>=18.5 in /home/aisl/.local/lib/python3.8/site-packages (from ipython>=4.0.0->ipywidgets) (63.4.2)\n",
      "Requirement already satisfied: jedi>=0.16 in /home/aisl/.local/lib/python3.8/site-packages (from ipython>=4.0.0->ipywidgets) (0.18.1)\n",
      "Requirement already satisfied: six>=1.5 in /usr/lib/python3/dist-packages (from python-dateutil>=2.7->matplotlib) (1.11.0)\n",
      "Requirement already satisfied: notebook>=4.4.1 in /home/aisl/.local/lib/python3.8/site-packages (from widgetsnbextension~=3.6.0->ipywidgets) (6.4.12)\n",
      "Requirement already satisfied: parso<0.9.0,>=0.8.0 in /home/aisl/.local/lib/python3.8/site-packages (from jedi>=0.16->ipython>=4.0.0->ipywidgets) (0.8.3)\n",
      "Requirement already satisfied: jupyter-core>=4.9.2 in /home/aisl/.local/lib/python3.8/site-packages (from jupyter-client>=6.1.12->ipykernel>=4.5.1->ipywidgets) (4.11.1)\n",
      "Requirement already satisfied: entrypoints in /home/aisl/.local/lib/python3.8/site-packages (from jupyter-client>=6.1.12->ipykernel>=4.5.1->ipywidgets) (0.4)\n",
      "Requirement already satisfied: Send2Trash>=1.8.0 in /home/aisl/.local/lib/python3.8/site-packages (from notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (1.8.0)\n",
      "Requirement already satisfied: terminado>=0.8.3 in /home/aisl/.local/lib/python3.8/site-packages (from notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (0.15.0)\n",
      "Requirement already satisfied: argon2-cffi in /home/aisl/.local/lib/python3.8/site-packages (from notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (21.3.0)\n",
      "Requirement already satisfied: prometheus-client in /home/aisl/.local/lib/python3.8/site-packages (from notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (0.14.1)\n",
      "Requirement already satisfied: nbformat in /home/aisl/.local/lib/python3.8/site-packages (from notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (5.4.0)\n",
      "Requirement already satisfied: nbconvert>=5 in /home/aisl/.local/lib/python3.8/site-packages (from notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (6.5.0)\n",
      "Requirement already satisfied: jinja2 in /home/aisl/.local/lib/python3.8/site-packages (from notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (3.1.2)\n",
      "Requirement already satisfied: ptyprocess>=0.5 in /home/aisl/.local/lib/python3.8/site-packages (from pexpect>4.3->ipython>=4.0.0->ipywidgets) (0.7.0)\n",
      "Requirement already satisfied: wcwidth in /home/aisl/.local/lib/python3.8/site-packages (from prompt-toolkit!=3.0.0,!=3.0.1,<3.1.0,>=2.0.0->ipython>=4.0.0->ipywidgets) (0.2.5)\n",
      "Requirement already satisfied: pure-eval in /home/aisl/.local/lib/python3.8/site-packages (from stack-data->ipython>=4.0.0->ipywidgets) (0.2.2)\n",
      "Requirement already satisfied: executing in /home/aisl/.local/lib/python3.8/site-packages (from stack-data->ipython>=4.0.0->ipywidgets) (0.9.1)\n",
      "Requirement already satisfied: asttokens in /home/aisl/.local/lib/python3.8/site-packages (from stack-data->ipython>=4.0.0->ipywidgets) (2.0.5)\n",
      "Requirement already satisfied: MarkupSafe>=2.0 in /home/aisl/.local/lib/python3.8/site-packages (from nbconvert>=5->notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (2.1.1)\n",
      "Requirement already satisfied: tinycss2 in /home/aisl/.local/lib/python3.8/site-packages (from nbconvert>=5->notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (1.1.1)\n",
      "Requirement already satisfied: defusedxml in /home/aisl/.local/lib/python3.8/site-packages (from nbconvert>=5->notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (0.7.1)\n",
      "Requirement already satisfied: nbclient>=0.5.0 in /home/aisl/.local/lib/python3.8/site-packages (from nbconvert>=5->notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (0.6.6)\n",
      "Requirement already satisfied: jupyterlab-pygments in /home/aisl/.local/lib/python3.8/site-packages (from nbconvert>=5->notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (0.2.2)\n",
      "Requirement already satisfied: bleach in /home/aisl/.local/lib/python3.8/site-packages (from nbconvert>=5->notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (5.0.1)\n",
      "Requirement already satisfied: beautifulsoup4 in /home/aisl/.local/lib/python3.8/site-packages (from nbconvert>=5->notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (4.11.1)\n",
      "Requirement already satisfied: pandocfilters>=1.4.1 in /home/aisl/.local/lib/python3.8/site-packages (from nbconvert>=5->notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (1.5.0)\n",
      "Requirement already satisfied: mistune<2,>=0.8.1 in /home/aisl/.local/lib/python3.8/site-packages (from nbconvert>=5->notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (0.8.4)\n",
      "Requirement already satisfied: fastjsonschema in /home/aisl/.local/lib/python3.8/site-packages (from nbformat->notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (2.16.1)\n",
      "Requirement already satisfied: jsonschema>=2.6 in /home/aisl/.local/lib/python3.8/site-packages (from nbformat->notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (4.9.1)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Requirement already satisfied: argon2-cffi-bindings in /home/aisl/.local/lib/python3.8/site-packages (from argon2-cffi->notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (21.2.0)\n",
      "Requirement already satisfied: attrs>=17.4.0 in /home/aisl/.local/lib/python3.8/site-packages (from jsonschema>=2.6->nbformat->notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (22.1.0)\n",
      "Requirement already satisfied: pkgutil-resolve-name>=1.3.10 in /home/aisl/.local/lib/python3.8/site-packages (from jsonschema>=2.6->nbformat->notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (1.3.10)\n",
      "Requirement already satisfied: importlib-resources>=1.4.0 in /home/aisl/.local/lib/python3.8/site-packages (from jsonschema>=2.6->nbformat->notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (5.9.0)\n",
      "Requirement already satisfied: pyrsistent!=0.17.0,!=0.17.1,!=0.17.2,>=0.14.0 in /home/aisl/.local/lib/python3.8/site-packages (from jsonschema>=2.6->nbformat->notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (0.18.1)\n",
      "Requirement already satisfied: cffi>=1.0.1 in /home/aisl/.local/lib/python3.8/site-packages (from argon2-cffi-bindings->argon2-cffi->notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (1.15.1)\n",
      "Requirement already satisfied: soupsieve>1.2 in /home/aisl/.local/lib/python3.8/site-packages (from beautifulsoup4->nbconvert>=5->notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (2.3.2.post1)\n",
      "Requirement already satisfied: webencodings in /home/aisl/.local/lib/python3.8/site-packages (from bleach->nbconvert>=5->notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (0.5.1)\n",
      "Requirement already satisfied: pycparser in /home/aisl/.local/lib/python3.8/site-packages (from cffi>=1.0.1->argon2-cffi-bindings->argon2-cffi->notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (2.21)\n",
      "Requirement already satisfied: zipp>=3.1.0 in /home/aisl/.local/lib/python3.8/site-packages (from importlib-resources>=1.4.0->jsonschema>=2.6->nbformat->notebook>=4.4.1->widgetsnbextension~=3.6.0->ipywidgets) (3.8.1)\n",
      "Defaulting to user installation because normal site-packages is not writeable\n",
      "Looking in indexes: https://pypi.org/simple, https://download.pytorch.org/whl/cu116\n",
      "Requirement already satisfied: torch in /home/aisl/.local/lib/python3.8/site-packages (1.12.1+cu116)\n",
      "Requirement already satisfied: torchvision in /home/aisl/.local/lib/python3.8/site-packages (0.13.1+cu116)\n",
      "Requirement already satisfied: torchaudio in /home/aisl/.local/lib/python3.8/site-packages (0.12.1+cu116)\n",
      "Requirement already satisfied: typing-extensions in /home/aisl/.local/lib/python3.8/site-packages (from torch) (4.3.0)\n",
      "Requirement already satisfied: numpy in /home/aisl/.local/lib/python3.8/site-packages (from torchvision) (1.22.4)\n",
      "Requirement already satisfied: requests in /home/aisl/.local/lib/python3.8/site-packages (from torchvision) (2.28.1)\n",
      "Requirement already satisfied: pillow!=8.3.*,>=5.3.0 in /home/aisl/.local/lib/python3.8/site-packages (from torchvision) (9.2.0)\n",
      "Requirement already satisfied: urllib3<1.27,>=1.21.1 in /usr/lib/python3/dist-packages (from requests->torchvision) (1.22)\n",
      "Requirement already satisfied: idna<4,>=2.5 in /home/aisl/.local/lib/python3.8/site-packages (from requests->torchvision) (3.3)\n",
      "Requirement already satisfied: charset-normalizer<3,>=2 in /home/aisl/.local/lib/python3.8/site-packages (from requests->torchvision) (2.1.0)\n",
      "Requirement already satisfied: certifi>=2017.4.17 in /usr/lib/python3/dist-packages (from requests->torchvision) (2018.1.18)\n",
      "Defaulting to user installation because normal site-packages is not writeable\n",
      "Requirement already satisfied: torchsummary in /home/aisl/.local/lib/python3.8/site-packages (1.5.1)\n",
      "Requirement already satisfied: torchviz in /home/aisl/.local/lib/python3.8/site-packages (0.0.2)\n",
      "Requirement already satisfied: tensorboard in /home/aisl/.local/lib/python3.8/site-packages (2.9.1)\n",
      "Requirement already satisfied: graphviz in /home/aisl/.local/lib/python3.8/site-packages (from torchviz) (0.20.1)\n",
      "Requirement already satisfied: torch in /home/aisl/.local/lib/python3.8/site-packages (from torchviz) (1.12.1+cu116)\n",
      "Requirement already satisfied: setuptools>=41.0.0 in /home/aisl/.local/lib/python3.8/site-packages (from tensorboard) (63.4.2)\n",
      "Requirement already satisfied: grpcio>=1.24.3 in /home/aisl/.local/lib/python3.8/site-packages (from tensorboard) (1.47.0)\n",
      "Requirement already satisfied: absl-py>=0.4 in /home/aisl/.local/lib/python3.8/site-packages (from tensorboard) (1.2.0)\n",
      "Requirement already satisfied: protobuf<3.20,>=3.9.2 in /home/aisl/.local/lib/python3.8/site-packages (from tensorboard) (3.19.4)\n",
      "Requirement already satisfied: requests<3,>=2.21.0 in /home/aisl/.local/lib/python3.8/site-packages (from tensorboard) (2.28.1)\n",
      "Requirement already satisfied: google-auth<3,>=1.6.3 in /home/aisl/.local/lib/python3.8/site-packages (from tensorboard) (2.10.0)\n",
      "Requirement already satisfied: google-auth-oauthlib<0.5,>=0.4.1 in /home/aisl/.local/lib/python3.8/site-packages (from tensorboard) (0.4.6)\n",
      "Requirement already satisfied: tensorboard-plugin-wit>=1.6.0 in /home/aisl/.local/lib/python3.8/site-packages (from tensorboard) (1.8.1)\n",
      "Requirement already satisfied: werkzeug>=1.0.1 in /home/aisl/.local/lib/python3.8/site-packages (from tensorboard) (2.2.1)\n",
      "Requirement already satisfied: markdown>=2.6.8 in /home/aisl/.local/lib/python3.8/site-packages (from tensorboard) (3.4.1)\n",
      "Requirement already satisfied: wheel>=0.26 in /home/aisl/.local/lib/python3.8/site-packages (from tensorboard) (0.37.1)\n",
      "Requirement already satisfied: tensorboard-data-server<0.7.0,>=0.6.0 in /home/aisl/.local/lib/python3.8/site-packages (from tensorboard) (0.6.1)\n",
      "Requirement already satisfied: numpy>=1.12.0 in /home/aisl/.local/lib/python3.8/site-packages (from tensorboard) (1.22.4)\n",
      "Requirement already satisfied: rsa<5,>=3.1.4 in /home/aisl/.local/lib/python3.8/site-packages (from google-auth<3,>=1.6.3->tensorboard) (4.9)\n",
      "Requirement already satisfied: six>=1.9.0 in /usr/lib/python3/dist-packages (from google-auth<3,>=1.6.3->tensorboard) (1.11.0)\n",
      "Requirement already satisfied: pyasn1-modules>=0.2.1 in /home/aisl/.local/lib/python3.8/site-packages (from google-auth<3,>=1.6.3->tensorboard) (0.2.8)\n",
      "Requirement already satisfied: cachetools<6.0,>=2.0.0 in /home/aisl/.local/lib/python3.8/site-packages (from google-auth<3,>=1.6.3->tensorboard) (5.2.0)\n",
      "Requirement already satisfied: requests-oauthlib>=0.7.0 in /home/aisl/.local/lib/python3.8/site-packages (from google-auth-oauthlib<0.5,>=0.4.1->tensorboard) (1.3.1)\n",
      "Requirement already satisfied: importlib-metadata>=4.4 in /home/aisl/.local/lib/python3.8/site-packages (from markdown>=2.6.8->tensorboard) (4.12.0)\n",
      "Requirement already satisfied: certifi>=2017.4.17 in /usr/lib/python3/dist-packages (from requests<3,>=2.21.0->tensorboard) (2018.1.18)\n",
      "Requirement already satisfied: idna<4,>=2.5 in /home/aisl/.local/lib/python3.8/site-packages (from requests<3,>=2.21.0->tensorboard) (3.3)\n",
      "Requirement already satisfied: urllib3<1.27,>=1.21.1 in /usr/lib/python3/dist-packages (from requests<3,>=2.21.0->tensorboard) (1.22)\n",
      "Requirement already satisfied: charset-normalizer<3,>=2 in /home/aisl/.local/lib/python3.8/site-packages (from requests<3,>=2.21.0->tensorboard) (2.1.0)\n",
      "Requirement already satisfied: MarkupSafe>=2.1.1 in /home/aisl/.local/lib/python3.8/site-packages (from werkzeug>=1.0.1->tensorboard) (2.1.1)\n",
      "Requirement already satisfied: typing-extensions in /home/aisl/.local/lib/python3.8/site-packages (from torch->torchviz) (4.3.0)\n",
      "Requirement already satisfied: zipp>=0.5 in /home/aisl/.local/lib/python3.8/site-packages (from importlib-metadata>=4.4->markdown>=2.6.8->tensorboard) (3.8.1)\n",
      "Requirement already satisfied: pyasn1<0.5.0,>=0.4.6 in /home/aisl/.local/lib/python3.8/site-packages (from pyasn1-modules>=0.2.1->google-auth<3,>=1.6.3->tensorboard) (0.4.8)\n",
      "Requirement already satisfied: oauthlib>=3.0.0 in /home/aisl/.local/lib/python3.8/site-packages (from requests-oauthlib>=0.7.0->google-auth-oauthlib<0.5,>=0.4.1->tensorboard) (3.2.0)\n"
     ]
    }
   ],
   "source": [
    "!pip3 install --upgrade pip\n",
    "!pip3 install opencv-python matplotlib tqdm numpy pandas ipywidgets\n",
    "!pip3 install torch torchvision torchaudio --extra-index-url https://download.pytorch.org/whl/cu116\n",
    "!pip3 install torchsummary torchviz tensorboard"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "INSPEKSI DATA\n",
    "=====================\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/home/aisl/BISA_AI/ai-hw-acc/dataset/cityscapes/leftImg8bit_trainvaltest/leftImg8bit/\n",
      "/home/aisl/BISA_AI/ai-hw-acc/dataset/cityscapes/gtFine_trainvaltest/gtFine/\n"
     ]
    }
   ],
   "source": [
    "#Download data dari: https://www.cityscapes-dataset.com/downloads/\n",
    "\n",
    "rootdir = \"/home/aisl/BISA_AI/ai-hw-acc/dataset/cityscapes/\"\n",
    "dir_rgb = rootdir+\"leftImg8bit_trainvaltest/leftImg8bit/\"\n",
    "dir_seg = rootdir+\"gtFine_trainvaltest/gtFine/\"\n",
    "part = 'train/' #train val test\n",
    "town = 'aachen/' #pilih salah satu kota\n",
    "\n",
    "print(dir_rgb)\n",
    "print(dir_seg)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Part 1: visualisasi gambar dan inspeksi dimensi\n",
    "----------------"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1024, 2048, 3)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#import library yang diperlukan\n",
    "%matplotlib inline\n",
    "from matplotlib import pyplot as plt\n",
    "import cv2\n",
    "import os\n",
    "\n",
    "#list semua file yang ada pada salah satu town di direktori RGB\n",
    "list_file = os.listdir(dir_rgb+part+town)\n",
    "list_file.sort() #urutkan\n",
    "idx = 10\n",
    "filename = list_file[idx] #ambil salah satu file\n",
    "\n",
    "#tampilkan\n",
    "image = cv2.imread(dir_rgb+part+town+filename)\n",
    "# image = image[:, :, [2, 1, 0]]\n",
    "print(image.shape)\n",
    "plt.imshow(image)\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Part 2: load GT segmentation\n",
    "----------------"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1024, 2048, 3)\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#list semua file yang ada pada salah satu town di direktori SEG\n",
    "list_file = os.listdir(dir_seg+part+town)\n",
    "list_file.sort() #urutkan\n",
    "idx = 10\n",
    "idx = 5*idx + 2\n",
    "filename = list_file[idx] #ambil salah satu file\n",
    "\n",
    "#tampilkan\n",
    "seg_gt = cv2.imread(dir_seg+part+town+filename)\n",
    "print(seg_gt.shape)\n",
    "plt.imshow(seg_gt)\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Part 3: visualisasi GT segmentation\n",
    "----------------"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "#Define colormap dan konversi label dari 35 class ke 20 class\n",
    "#BACA https://github.com/mcordts/cityscapesScripts/blob/master/cityscapesscripts/helpers/labels.py\n",
    "SEG_MAP = {\n",
    "    'colors'        :[[0, 0, 0], [128, 64, 128], [244, 35, 232], [70, 70, 70], [102, 102, 156],  \n",
    "                    [190, 153, 153], [153, 153, 153], [250, 170, 30], [220, 220, 0],\n",
    "                    [107, 142, 35], [152, 251, 152], [70, 130, 180], [220, 20, 60],\n",
    "                    [255, 0, 0], [0, 0, 142], [0, 0, 70], [0, 60, 100], \n",
    "                    [0, 80, 100], [0, 0, 230], [119, 11, 32]],  \n",
    "    'classes'       : ['None', 'road', 'sidewalk', 'building', 'wall',\n",
    "                        'fence', 'pole', 'traffic light', 'traffic sign', \n",
    "                        'vegetation', 'terrain', 'sky', 'person', \n",
    "                        'rider', 'car', 'truck', 'bus',\n",
    "                        'train', 'motorcycle', 'bicycle'],\n",
    "}\n",
    "\n",
    "mapping_20 = { \n",
    "        0: 0,\n",
    "        1: 0,\n",
    "        2: 0,\n",
    "        3: 0,\n",
    "        4: 0,\n",
    "        5: 0,\n",
    "        6: 0,\n",
    "        7: 1,\n",
    "        8: 2,\n",
    "        9: 0,\n",
    "        10: 0,\n",
    "        11: 3,\n",
    "        12: 4,\n",
    "        13: 5,\n",
    "        14: 0,\n",
    "        15: 0,\n",
    "        16: 0,\n",
    "        17: 6,\n",
    "        18: 0,\n",
    "        19: 7,\n",
    "        20: 8,\n",
    "        21: 9,\n",
    "        22: 10,\n",
    "        23: 11,\n",
    "        24: 12,\n",
    "        25: 13,\n",
    "        26: 14,\n",
    "        27: 15,\n",
    "        28: 16,\n",
    "        29: 0,\n",
    "        30: 0,\n",
    "        31: 17,\n",
    "        32: 18,\n",
    "        33: 19,\n",
    "        -1: 0\n",
    "    }\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import numpy as np\n",
    "\n",
    "#visualisasi segmentation\n",
    "def vis_seg(seg, SEG_CMAP):\n",
    "    #buat array kosong untuk menyimpan output gambar\n",
    "    imgx = np.zeros((seg.shape[0], seg.shape[1], 3))\n",
    "    #warnai setiap label\n",
    "    for cmap in SEG_CMAP['colors']:\n",
    "        cmap_id = SEG_CMAP['colors'].index(cmap)\n",
    "        imgx[np.where(seg == cmap_id)] = cmap\n",
    "#         imgx = imgx[:, :, [2, 1, 0]]\n",
    "    return imgx\n",
    "\n",
    "#konversi dari 35 class ke 20 class\n",
    "def encode_labels(mask, mapping_label):\n",
    "    #inputnya adalah HWC hasil baca opencv\n",
    "    mask = mask[:,:,0:1].reshape(mask.shape[0], mask.shape[1])  #ambil 1 channel dan reshape dari 3D ke 2D\n",
    "    label_mask = np.zeros_like(mask)\n",
    "    for k in mapping_label:\n",
    "        label_mask[mask == k] = mapping_label[k]\n",
    "    return label_mask\n",
    "\n",
    "\n",
    "#visualisasi GT segmentasi\n",
    "seg_gtx = encode_labels(seg_gt, mapping_20)\n",
    "seg_color = vis_seg(seg_gtx, SEG_MAP)\n",
    "plt.imshow(seg_color.astype(np.uint8))\n",
    "plt.show()\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Part 4: Inspeksi setiap class channel pada GT segmentation\n",
    "----------------"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "shape GT segmentation yang telah dikonversi dari 35c ke 20c:  (1024, 2048)\n",
      "shape GT segmentation one hot encoded:  (1024, 2048, 20)\n",
      "idx class:  0\n",
      "nama class:  None\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "idx class:  1\n",
      "nama class:  road\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "idx class:  2\n",
      "nama class:  sidewalk\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "idx class:  3\n",
      "nama class:  building\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "idx class:  4\n",
      "nama class:  wall\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "idx class:  5\n",
      "nama class:  fence\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "idx class:  6\n",
      "nama class:  pole\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "idx class:  7\n",
      "nama class:  traffic light\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "idx class:  8\n",
      "nama class:  traffic sign\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "idx class:  9\n",
      "nama class:  vegetation\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "idx class:  10\n",
      "nama class:  terrain\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "idx class:  11\n",
      "nama class:  sky\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "idx class:  12\n",
      "nama class:  person\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "idx class:  13\n",
      "nama class:  rider\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "idx class:  14\n",
      "nama class:  car\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "idx class:  15\n",
      "nama class:  truck\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "idx class:  16\n",
      "nama class:  bus\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "idx class:  17\n",
      "nama class:  train\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "idx class:  18\n",
      "nama class:  motorcycle\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "idx class:  19\n",
      "nama class:  bicycle\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#konversi dari class id ke representasi one-hot encoding\n",
    "def cls2one_hot(ss_gt20, n_class):\n",
    "    one_hot = (np.arange(n_class) == ss_gt20[...,None]).astype(int) \n",
    "    return one_hot\n",
    "\n",
    "print(\"shape GT segmentation yang telah dikonversi dari 35c ke 20c: \",seg_gtx.shape)\n",
    "seg_one_hot = cls2one_hot(seg_gtx, n_class=len(SEG_MAP['classes']))\n",
    "print(\"shape GT segmentation one hot encoded: \",seg_one_hot.shape)\n",
    "\n",
    "for i in range(seg_one_hot.shape[2]):\n",
    "    print('idx class: ',i)\n",
    "    print('nama class: ',SEG_MAP['classes'][i])\n",
    "    img_class_r = seg_one_hot[:, :, i:i+1] * SEG_MAP['colors'][i][0]\n",
    "    img_class_g = seg_one_hot[:, :, i:i+1] * SEG_MAP['colors'][i][1]\n",
    "    img_class_b = seg_one_hot[:, :, i:i+1] * SEG_MAP['colors'][i][2]\n",
    "    img_class_rgb = np.concatenate([img_class_r, img_class_g, img_class_b], axis=2)\n",
    "    plt.imshow(img_class_rgb.astype(np.uint8))\n",
    "    plt.show()\n",
    "\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "MENDESAIN ARSITEKTUR UNET DAN DEEPLABV3\n",
    "=====================\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "scratch_unet(\n",
      "  (conv1_rgb_f): ConvBlock(\n",
      "    (conv_block0): ConvBNRelu(\n",
      "      (conv): Conv2d(3, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "      (bn): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU()\n",
      "    )\n",
      "    (conv_block1): ConvBNRelu(\n",
      "      (conv): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "      (bn): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU()\n",
      "    )\n",
      "  )\n",
      "  (conv0_rgb_f): ConvBlock(\n",
      "    (conv_block0): ConvBNRelu(\n",
      "      (conv): Conv2d(32, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "      (bn): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU()\n",
      "    )\n",
      "    (conv_block1): ConvBNRelu(\n",
      "      (conv): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "      (bn): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU()\n",
      "    )\n",
      "  )\n",
      "  (conv1_ss_f): ConvBlock(\n",
      "    (conv_block0): ConvBNRelu(\n",
      "      (conv): Conv2d(96, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "      (bn): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU()\n",
      "    )\n",
      "    (conv_block1): ConvBNRelu(\n",
      "      (conv): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "      (bn): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU()\n",
      "    )\n",
      "  )\n",
      "  (final_ss_f): ConvBlock(\n",
      "    (conv_block0): ConvBNRelu(\n",
      "      (conv): Conv2d(32, 32, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "      (bn): BatchNorm2d(32, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU()\n",
      "    )\n",
      "    (conv_block1): Sequential(\n",
      "      (0): Conv2d(32, 23, kernel_size=(1, 1), stride=(1, 1))\n",
      "    )\n",
      "  )\n",
      "  (sigmoid): Sigmoid()\n",
      ")\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/aisl/.local/lib/python3.8/site-packages/torchvision/models/_utils.py:208: UserWarning: The parameter 'pretrained' is deprecated since 0.13 and will be removed in 0.15, please use 'weights' instead.\n",
      "  warnings.warn(\n",
      "/home/aisl/.local/lib/python3.8/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and will be removed in 0.15. The current behavior is equivalent to passing `weights=ResNet34_Weights.IMAGENET1K_V1`. You can also use `weights=ResNet34_Weights.DEFAULT` to get the most up-to-date weights.\n",
      "  warnings.warn(msg)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "resunet_seg(\n",
      "  (rgb_normalizer): Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])\n",
      "  (RGB_encoder): ResNet(\n",
      "    (conv1): Conv2d(3, 64, kernel_size=(7, 7), stride=(2, 2), padding=(3, 3), bias=False)\n",
      "    (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "    (relu): ReLU(inplace=True)\n",
      "    (maxpool): MaxPool2d(kernel_size=3, stride=2, padding=1, dilation=1, ceil_mode=False)\n",
      "    (layer1): Sequential(\n",
      "      (0): BasicBlock(\n",
      "        (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "        (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      )\n",
      "      (1): BasicBlock(\n",
      "        (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "        (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      )\n",
      "      (2): BasicBlock(\n",
      "        (conv1): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "        (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      )\n",
      "    )\n",
      "    (layer2): Sequential(\n",
      "      (0): BasicBlock(\n",
      "        (conv1): Conv2d(64, 128, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "        (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (downsample): Sequential(\n",
      "          (0): Conv2d(64, 128, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
      "          (1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        )\n",
      "      )\n",
      "      (1): BasicBlock(\n",
      "        (conv1): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "        (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      )\n",
      "      (2): BasicBlock(\n",
      "        (conv1): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "        (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      )\n",
      "      (3): BasicBlock(\n",
      "        (conv1): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "        (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      )\n",
      "    )\n",
      "    (layer3): Sequential(\n",
      "      (0): BasicBlock(\n",
      "        (conv1): Conv2d(128, 256, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (downsample): Sequential(\n",
      "          (0): Conv2d(128, 256, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
      "          (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        )\n",
      "      )\n",
      "      (1): BasicBlock(\n",
      "        (conv1): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      )\n",
      "      (2): BasicBlock(\n",
      "        (conv1): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      )\n",
      "      (3): BasicBlock(\n",
      "        (conv1): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      )\n",
      "      (4): BasicBlock(\n",
      "        (conv1): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      )\n",
      "      (5): BasicBlock(\n",
      "        (conv1): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "        (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      )\n",
      "    )\n",
      "    (layer4): Sequential(\n",
      "      (0): BasicBlock(\n",
      "        (conv1): Conv2d(256, 512, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "        (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (downsample): Sequential(\n",
      "          (0): Conv2d(256, 512, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
      "          (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        )\n",
      "      )\n",
      "      (1): BasicBlock(\n",
      "        (conv1): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "        (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      )\n",
      "      (2): BasicBlock(\n",
      "        (conv1): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "        (relu): ReLU(inplace=True)\n",
      "        (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "        (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      )\n",
      "    )\n",
      "    (avgpool): Sequential()\n",
      "    (fc): Sequential()\n",
      "  )\n",
      "  (conv3_ss_f): ConvBlock(\n",
      "    (conv_block0): ConvBNRelu(\n",
      "      (conv): Conv2d(768, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "      (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU()\n",
      "    )\n",
      "    (conv_block1): ConvBNRelu(\n",
      "      (conv): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "      (bn): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU()\n",
      "    )\n",
      "  )\n",
      "  (conv2_ss_f): ConvBlock(\n",
      "    (conv_block0): ConvBNRelu(\n",
      "      (conv): Conv2d(384, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "      (bn): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU()\n",
      "    )\n",
      "    (conv_block1): ConvBNRelu(\n",
      "      (conv): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "      (bn): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU()\n",
      "    )\n",
      "  )\n",
      "  (conv1_ss_f): ConvBlock(\n",
      "    (conv_block0): ConvBNRelu(\n",
      "      (conv): Conv2d(192, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "      (bn): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU()\n",
      "    )\n",
      "    (conv_block1): ConvBNRelu(\n",
      "      (conv): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "      (bn): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU()\n",
      "    )\n",
      "  )\n",
      "  (conv0_ss_f): ConvBlock(\n",
      "    (conv_block0): ConvBNRelu(\n",
      "      (conv): Conv2d(128, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "      (bn): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU()\n",
      "    )\n",
      "    (conv_block1): ConvBNRelu(\n",
      "      (conv): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "      (bn): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU()\n",
      "    )\n",
      "  )\n",
      "  (final_ss_f): ConvBlock(\n",
      "    (conv_block0): ConvBNRelu(\n",
      "      (conv): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1))\n",
      "      (bn): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU()\n",
      "    )\n",
      "    (conv_block1): Sequential(\n",
      "      (0): Conv2d(64, 20, kernel_size=(1, 1), stride=(1, 1))\n",
      "    )\n",
      "  )\n",
      "  (sigmoid): Sigmoid()\n",
      ")\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/home/aisl/.local/lib/python3.8/site-packages/torchvision/models/_utils.py:223: UserWarning: Arguments other than a weight enum or `None` for 'weights' are deprecated since 0.13 and will be removed in 0.15. The current behavior is equivalent to passing `weights=DeepLabV3_ResNet101_Weights.COCO_WITH_VOC_LABELS_V1`. You can also use `weights=DeepLabV3_ResNet101_Weights.DEFAULT` to get the most up-to-date weights.\n",
      "  warnings.warn(msg)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "dlv3_seg(\n",
      "  (rgb_normalizer): Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])\n",
      "  (net): DeepLabV3(\n",
      "    (backbone): IntermediateLayerGetter(\n",
      "      (conv1): Conv2d(3, 64, kernel_size=(7, 7), stride=(2, 2), padding=(3, 3), bias=False)\n",
      "      (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (relu): ReLU(inplace=True)\n",
      "      (maxpool): MaxPool2d(kernel_size=3, stride=2, padding=1, dilation=1, ceil_mode=False)\n",
      "      (layer1): Sequential(\n",
      "        (0): Bottleneck(\n",
      "          (conv1): Conv2d(64, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "          (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "          (downsample): Sequential(\n",
      "            (0): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "            (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          )\n",
      "        )\n",
      "        (1): Bottleneck(\n",
      "          (conv1): Conv2d(256, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "          (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "        (2): Bottleneck(\n",
      "          (conv1): Conv2d(256, 64, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(64, 64, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "          (bn2): BatchNorm2d(64, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(64, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "      )\n",
      "      (layer2): Sequential(\n",
      "        (0): Bottleneck(\n",
      "          (conv1): Conv2d(256, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(2, 2), padding=(1, 1), bias=False)\n",
      "          (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "          (downsample): Sequential(\n",
      "            (0): Conv2d(256, 512, kernel_size=(1, 1), stride=(2, 2), bias=False)\n",
      "            (1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          )\n",
      "        )\n",
      "        (1): Bottleneck(\n",
      "          (conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "          (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "        (2): Bottleneck(\n",
      "          (conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "          (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "        (3): Bottleneck(\n",
      "          (conv1): Conv2d(512, 128, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(128, 128, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "          (bn2): BatchNorm2d(128, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(128, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "      )\n",
      "      (layer3): Sequential(\n",
      "        (0): Bottleneck(\n",
      "          (conv1): Conv2d(512, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "          (downsample): Sequential(\n",
      "            (0): Conv2d(512, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "            (1): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          )\n",
      "        )\n",
      "        (1): Bottleneck(\n",
      "          (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
      "          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "        (2): Bottleneck(\n",
      "          (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
      "          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "        (3): Bottleneck(\n",
      "          (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
      "          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "        (4): Bottleneck(\n",
      "          (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
      "          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "        (5): Bottleneck(\n",
      "          (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
      "          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "        (6): Bottleneck(\n",
      "          (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
      "          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "        (7): Bottleneck(\n",
      "          (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
      "          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "        (8): Bottleneck(\n",
      "          (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
      "          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "        (9): Bottleneck(\n",
      "          (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
      "          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "        (10): Bottleneck(\n",
      "          (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
      "          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "        (11): Bottleneck(\n",
      "          (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
      "          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "        (12): Bottleneck(\n",
      "          (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
      "          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "        (13): Bottleneck(\n",
      "          (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
      "          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "        (14): Bottleneck(\n",
      "          (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
      "          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "        (15): Bottleneck(\n",
      "          (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
      "          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "        (16): Bottleneck(\n",
      "          (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
      "          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "        (17): Bottleneck(\n",
      "          (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
      "          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "        (18): Bottleneck(\n",
      "          (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
      "          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "        (19): Bottleneck(\n",
      "          (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
      "          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "        (20): Bottleneck(\n",
      "          (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
      "          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "        (21): Bottleneck(\n",
      "          (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
      "          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "        (22): Bottleneck(\n",
      "          (conv1): Conv2d(1024, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
      "          (bn2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(256, 1024, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(1024, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "      )\n",
      "      (layer4): Sequential(\n",
      "        (0): Bottleneck(\n",
      "          (conv1): Conv2d(1024, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(2, 2), dilation=(2, 2), bias=False)\n",
      "          (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "          (downsample): Sequential(\n",
      "            (0): Conv2d(1024, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "            (1): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          )\n",
      "        )\n",
      "        (1): Bottleneck(\n",
      "          (conv1): Conv2d(2048, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(4, 4), dilation=(4, 4), bias=False)\n",
      "          (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "        (2): Bottleneck(\n",
      "          (conv1): Conv2d(2048, 512, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn1): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv2): Conv2d(512, 512, kernel_size=(3, 3), stride=(1, 1), padding=(4, 4), dilation=(4, 4), bias=False)\n",
      "          (bn2): BatchNorm2d(512, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (conv3): Conv2d(512, 2048, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (bn3): BatchNorm2d(2048, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (relu): ReLU(inplace=True)\n",
      "        )\n",
      "      )\n",
      "    )\n",
      "    (classifier): DeepLabHead(\n",
      "      (0): ASPP(\n",
      "        (convs): ModuleList(\n",
      "          (0): Sequential(\n",
      "            (0): Conv2d(2048, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "            (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "            (2): ReLU()\n",
      "          )\n",
      "          (1): ASPPConv(\n",
      "            (0): Conv2d(2048, 256, kernel_size=(3, 3), stride=(1, 1), padding=(12, 12), dilation=(12, 12), bias=False)\n",
      "            (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "            (2): ReLU()\n",
      "          )\n",
      "          (2): ASPPConv(\n",
      "            (0): Conv2d(2048, 256, kernel_size=(3, 3), stride=(1, 1), padding=(24, 24), dilation=(24, 24), bias=False)\n",
      "            (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "            (2): ReLU()\n",
      "          )\n",
      "          (3): ASPPConv(\n",
      "            (0): Conv2d(2048, 256, kernel_size=(3, 3), stride=(1, 1), padding=(36, 36), dilation=(36, 36), bias=False)\n",
      "            (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "            (2): ReLU()\n",
      "          )\n",
      "          (4): ASPPPooling(\n",
      "            (0): AdaptiveAvgPool2d(output_size=1)\n",
      "            (1): Conv2d(2048, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "            (2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "            (3): ReLU()\n",
      "          )\n",
      "        )\n",
      "        (project): Sequential(\n",
      "          (0): Conv2d(1280, 256, kernel_size=(1, 1), stride=(1, 1), bias=False)\n",
      "          (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "          (2): ReLU()\n",
      "          (3): Dropout(p=0.5, inplace=False)\n",
      "        )\n",
      "      )\n",
      "      (1): Conv2d(256, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (2): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (3): ReLU()\n",
      "      (4): Conv2d(256, 20, kernel_size=(1, 1), stride=(1, 1))\n",
      "    )\n",
      "    (aux_classifier): FCNHead(\n",
      "      (0): Conv2d(1024, 256, kernel_size=(3, 3), stride=(1, 1), padding=(1, 1), bias=False)\n",
      "      (1): BatchNorm2d(256, eps=1e-05, momentum=0.1, affine=True, track_running_stats=True)\n",
      "      (2): ReLU()\n",
      "      (3): Dropout(p=0.1, inplace=False)\n",
      "      (4): Conv2d(256, 21, kernel_size=(1, 1), stride=(1, 1))\n",
      "    )\n",
      "  )\n",
      "  (sigmoid): Sigmoid()\n",
      ")\n"
     ]
    }
   ],
   "source": [
    "import torch.nn as nn\n",
    "import torch.nn.functional as F\n",
    "from torch import cat\n",
    "from torchvision import models\n",
    "import torchvision.transforms as transforms\n",
    "from torchvision.models import segmentation #deeplabv3_resnet101\n",
    "\n",
    "class ConvBNRelu(nn.Module):\n",
    "    def __init__(self, channelx, stridex=1, kernelx=3, paddingx=1):\n",
    "        super(ConvBNRelu, self).__init__()\n",
    "        self.conv = nn.Conv2d(channelx[0], channelx[1], kernel_size=kernelx, stride=stridex, padding=paddingx, padding_mode='zeros')\n",
    "        self.bn = nn.BatchNorm2d(channelx[1])\n",
    "        self.relu = nn.ReLU()\n",
    "    \n",
    "    def forward(self, x):\n",
    "        x = self.conv(x) \n",
    "        x = self.bn(x) \n",
    "        y = self.relu(x)\n",
    "        return y\n",
    "\n",
    "class ConvBlock(nn.Module):\n",
    "    def __init__(self, channel, final=False): \n",
    "        super(ConvBlock, self).__init__()\n",
    "        #conv block\n",
    "        if final:\n",
    "            self.conv_block0 = ConvBNRelu(channelx=[channel[0], channel[0]], stridex=1)\n",
    "            self.conv_block1 = nn.Sequential(\n",
    "            nn.Conv2d(channel[0], channel[1], kernel_size=1)\n",
    "            )\n",
    "        else:\n",
    "            self.conv_block0 = ConvBNRelu(channelx=[channel[0], channel[1]], stridex=1)\n",
    "            self.conv_block1 = ConvBNRelu(channelx=[channel[1], channel[1]], stridex=1)\n",
    "\n",
    " \n",
    "    def forward(self, x):\n",
    "        #convolutional block\n",
    "        y = self.conv_block0(x)\n",
    "        y = self.conv_block1(y)\n",
    "        return y\n",
    "\n",
    "\n",
    "class scratch_unet(nn.Module): #\n",
    "    def __init__(self):\n",
    "        super().__init__()\n",
    "        #RGB\n",
    "        self.conv1_rgb_f = ConvBlock(channel=[3, 32])\n",
    "        self.conv0_rgb_f = ConvBlock(channel=[32, 64])\n",
    "\n",
    "        #SEG\n",
    "        self.conv1_ss_f = ConvBlock(channel=[64+32, 32])\n",
    "        self.final_ss_f = ConvBlock(channel=[32, 23], final=True)\n",
    "               \n",
    "        #aktivasi final\n",
    "        self.sigmoid = nn.Sigmoid()\n",
    "\n",
    "    def forward(self, rgb):\n",
    "        #bagian downsampling\n",
    "        rgb_f_1 = self.conv1_rgb_f(rgb)\n",
    "        down_rgb_f_1 = F.max_pool2d(rgb_f_1, kernel_size=(2,2))\n",
    "        rgb_f_0 = self.conv0_rgb_f(down_rgb_f_1)\n",
    "                \n",
    "        #bagian upsampling\n",
    "        up_rgb_f_0 = F.interpolate(rgb_f_0, size=rgb.size()[-2:], mode='bilinear', align_corners=True)\n",
    "        ss_f_1 = self.conv1_ss_f(cat([up_rgb_f_0, rgb_f_1], dim=1))\n",
    "        ss_f = self.sigmoid(self.final_ss_f(ss_f_1))\n",
    "        \n",
    "        return ss_f\n",
    "\n",
    "print(scratch_unet())\n",
    "    \n",
    "    \n",
    "class resunet_seg(nn.Module): #\n",
    "    def __init__(self):\n",
    "        super().__init__()\n",
    "        n_fmap = [64, 64, 128, 256, 512]\n",
    "        \n",
    "        #RGB\n",
    "        self.rgb_normalizer = transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])\n",
    "        self.RGB_encoder = models.resnet34(pretrained=True) #resnet18\n",
    "        self.RGB_encoder.fc = nn.Sequential() #cara paling gampang untuk menghilangkan fc layer yang tidak diperlukan\n",
    "        self.RGB_encoder.avgpool = nn.Sequential() \n",
    "\n",
    "        #SEG\n",
    "        self.conv3_ss_f = ConvBlock(channel=[n_fmap[4]+n_fmap[3], n_fmap[3]])\n",
    "        self.conv2_ss_f = ConvBlock(channel=[n_fmap[3]+n_fmap[2], n_fmap[2]])\n",
    "        self.conv1_ss_f = ConvBlock(channel=[n_fmap[2]+n_fmap[1], n_fmap[1]])\n",
    "        self.conv0_ss_f = ConvBlock(channel=[n_fmap[1]+n_fmap[0], n_fmap[0]])\n",
    "        self.final_ss_f = ConvBlock(channel=[n_fmap[0], 20], final=True)\n",
    "                \n",
    "        #aktivasi final\n",
    "        self.sigmoid = nn.Sigmoid()\n",
    "\n",
    "    def forward(self, rgb):\n",
    "        #bagian downsampling\n",
    "        in_rgb = self.rgb_normalizer(rgb)\n",
    "        x_RGB = self.RGB_encoder.conv1(in_rgb)\n",
    "        x_RGB = self.RGB_encoder.bn1(x_RGB)\n",
    "        RGB_features0 = self.RGB_encoder.relu(x_RGB)\n",
    "        RGB_features1 = self.RGB_encoder.layer1(self.RGB_encoder.maxpool(RGB_features0))\n",
    "        RGB_features2 = self.RGB_encoder.layer2(RGB_features1)\n",
    "        RGB_features3 = self.RGB_encoder.layer3(RGB_features2)\n",
    "        RGB_features4 = self.RGB_encoder.layer4(RGB_features3)\n",
    "        up_RGB_features4 = F.interpolate(RGB_features4, size=RGB_features3.size()[-2:], mode='bilinear', align_corners=True)\n",
    "        \n",
    "        #bagian segmentation\n",
    "        ss_f_3 = self.conv3_ss_f(cat([up_RGB_features4, RGB_features3], dim=1))\n",
    "        up_ss_f_3 = F.interpolate(ss_f_3, size=RGB_features2.size()[-2:], mode='bilinear', align_corners=True)\n",
    "        ss_f_2 = self.conv2_ss_f(cat([up_ss_f_3, RGB_features2], dim=1))\n",
    "        up_ss_f_2 = F.interpolate(ss_f_2, size=RGB_features1.size()[-2:], mode='bilinear', align_corners=True)\n",
    "        ss_f_1 = self.conv1_ss_f(cat([up_ss_f_2, RGB_features1], dim=1))\n",
    "        up_ss_f_1 = F.interpolate(ss_f_1, size=RGB_features0.size()[-2:], mode='bilinear', align_corners=True)\n",
    "        ss_f_0 = self.conv0_ss_f(cat([up_ss_f_1, RGB_features0], dim=1))\n",
    "        up_ss_f_0 = F.interpolate(ss_f_0, size=in_rgb.size()[-2:], mode='bilinear', align_corners=True)\n",
    "        ss_f = self.sigmoid(self.final_ss_f(up_ss_f_0))\n",
    "        \n",
    "        return ss_f\n",
    "\n",
    "\n",
    "print(resunet_seg())\n",
    "\n",
    "\n",
    "\n",
    "class dlv3_seg(nn.Module):\n",
    "    def __init__(self):\n",
    "        super().__init__()\n",
    "        self.rgb_normalizer = transforms.Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])\n",
    "        self.net = segmentation.deeplabv3_resnet101(pretrained=True) #pretrained di MS COCO 2017\n",
    "        #ganti layer terakhir, point wise conv nya memiliki n_class=20\n",
    "        self.net.classifier[4] = nn.Conv2d(256, 20, kernel_size=(1,1), stride=(1,1))\n",
    "        self.sigmoid = nn.Sigmoid()\n",
    "        \n",
    "    def forward(self, rgb_f):\n",
    "        in_rgb = self.rgb_normalizer(rgb_f)\n",
    "        ss_f = self.sigmoid(self.net(in_rgb)['out'])\n",
    "\n",
    "        return ss_f\n",
    "\n",
    "\n",
    "print(dlv3_seg())"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "MEMBUAT FUNGSI-FUNGSI PENDUKUNG\n",
    "=====================\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A. DATALOADER DAN FUNGSI PENDUKUNGNYA\n",
    "----------------\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "from torch.utils.data import Dataset\n",
    "\n",
    "#Define colormap dan konversi label dari 35 class ke 20 class\n",
    "#BACA https://github.com/mcordts/cityscapesScripts/blob/master/cityscapesscripts/helpers/labels.py\n",
    "SEG_MAP = {\n",
    "    'colors'        :[[0, 0, 0], [128, 64, 128], [244, 35, 232], [70, 70, 70], [102, 102, 156],  \n",
    "                    [190, 153, 153], [153, 153, 153], [250, 170, 30], [220, 220, 0],\n",
    "                    [107, 142, 35], [152, 251, 152], [70, 130, 180], [220, 20, 60],\n",
    "                    [255, 0, 0], [0, 0, 142], [0, 0, 70], [0, 60, 100], \n",
    "                    [0, 80, 100], [0, 0, 230], [119, 11, 32]],  \n",
    "    'classes'       : ['None', 'road', 'sidewalk', 'building', 'wall',\n",
    "                        'fence', 'pole', 'traffic light', 'traffic sign', \n",
    "                        'vegetation', 'terrain', 'sky', 'person', \n",
    "                        'rider', 'car', 'truck', 'bus',\n",
    "                        'train', 'motorcycle', 'bicycle'],\n",
    "}\n",
    "\n",
    "mapping_20 = { \n",
    "        0: 0,\n",
    "        1: 0,\n",
    "        2: 0,\n",
    "        3: 0,\n",
    "        4: 0,\n",
    "        5: 0,\n",
    "        6: 0,\n",
    "        7: 1,\n",
    "        8: 2,\n",
    "        9: 0,\n",
    "        10: 0,\n",
    "        11: 3,\n",
    "        12: 4,\n",
    "        13: 5,\n",
    "        14: 0,\n",
    "        15: 0,\n",
    "        16: 0,\n",
    "        17: 6,\n",
    "        18: 0,\n",
    "        19: 7,\n",
    "        20: 8,\n",
    "        21: 9,\n",
    "        22: 10,\n",
    "        23: 11,\n",
    "        24: 12,\n",
    "        25: 13,\n",
    "        26: 14,\n",
    "        27: 15,\n",
    "        28: 16,\n",
    "        29: 0,\n",
    "        30: 0,\n",
    "        31: 17,\n",
    "        32: 18,\n",
    "        33: 19,\n",
    "        -1: 0\n",
    "    }\n",
    "\n",
    "\n",
    "#konversi dari 35 class ke 20 class\n",
    "def encode_labels(mask, mapping_label):\n",
    "    #inputnya adalah HWC hasil baca opencv\n",
    "    mask = mask[:,:,0:1].reshape(mask.shape[0], mask.shape[1])  #ambil 1 channel dan reshape dari 3D ke 2D\n",
    "    label_mask = np.zeros_like(mask)\n",
    "    for k in mapping_label:\n",
    "        label_mask[mask == k] = mapping_label[k]\n",
    "    return label_mask\n",
    "\n",
    "#konversi dari class id ke representasi one-hot encoding\n",
    "def cls2one_hot(ss_gt20, n_class):\n",
    "    one_hot = (np.arange(n_class) == ss_gt20[...,None]).astype(int) \n",
    "    return one_hot\n",
    "\n",
    "\n",
    "#untuk mereduksi dimensi spasial supaya komputasi tidak terlalu berat\n",
    "def resize_matrix(image, resize):\n",
    "    #inputnya HWC\n",
    "    WH_resized = (int(image.shape[1]/resize), int(image.shape[0]/resize))\n",
    "    resized_image = cv2.resize(image, WH_resized, interpolation=cv2.INTER_NEAREST)\n",
    "\n",
    "    return resized_image\n",
    "\n",
    "\n",
    "#buat class dataset\n",
    "class myData(Dataset):\n",
    "    def __init__(self, root_dir, part):\n",
    "        #buat list kosong dulu\n",
    "        self.rgb = []\n",
    "        self.seg = []\n",
    "        \n",
    "        #define folder masing2 untuk rgb seg dep\n",
    "        rgb_folder = os.path.join(os.path.join(os.path.join(root_dir, 'leftImg8bit_trainvaltest'), 'leftImg8bit'), part)\n",
    "        seg_folder = os.path.join(os.path.join(os.path.join(root_dir, 'gtFine_trainvaltest'), 'gtFine'), part)\n",
    " \n",
    "        # list semua kota yang ada di RGB leftimg8bit\n",
    "        town_list = os.listdir(rgb_folder)\n",
    "        town_list.sort()\n",
    "        for town in town_list:\n",
    "            rgb_dir = os.path.join(rgb_folder, town)\n",
    "            seg_dir = os.path.join(seg_folder, town)\n",
    "            \n",
    "            #list semua file\n",
    "            file_list = os.listdir(rgb_dir)\n",
    "            file_list.sort()\n",
    "            for filename in file_list:\n",
    "                #append ke list kosong tadi\n",
    "                self.rgb.append(os.path.join(rgb_dir, filename))\n",
    "                self.seg.append(os.path.join(seg_dir, filename[:-15]+'gtFine_labelIds.png'))\n",
    "\n",
    "    def __len__(self):\n",
    "        return len(self.rgb)\n",
    "\n",
    "    def __getitem__(self, index):\n",
    "        data = dict()\n",
    "        data['rgb'] = np.transpose(np.array(resize_matrix(\n",
    "            cv2.imread(self.rgb[index]), \n",
    "            resize=4)), (2, 0, 1)) # (H, W, C) --> (C, H, W)\n",
    "        data['seg'] = np.transpose(np.array(resize_matrix(cls2one_hot(encode_labels(\n",
    "            cv2.imread(self.seg[index]), \n",
    "            mapping_label=mapping_20), n_class=len(SEG_MAP['classes'])), resize=4)), (2, 0, 1)) # (H, W, C) --> (C, H, W)\n",
    "\n",
    "        return data\n",
    "    "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "B. TRAINING-VALIDATION\n",
    "----------------\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "from tqdm.notebook import tqdm\n",
    "\n",
    "#Class untuk penyimpanan dan perhitungan update loss\n",
    "class AverageMeter(object):\n",
    "    def __init__(self):\n",
    "        self.val = 0\n",
    "        self.avg = 0\n",
    "        self.sum = 0\n",
    "        self.count = 0\n",
    "    #update kalkulasi\n",
    "    def update(self, val, n=1):\n",
    "        self.val = val\n",
    "        self.sum += val * n\n",
    "        self.count += n\n",
    "        self.avg = self.sum / self.count\n",
    "        \n",
    "\n",
    "#FUNGSI TRAINING\n",
    "def train(data_loader, model, writer, cur_epoch, device, optimizer):\n",
    "    #buat variabel untuk menyimpan kalkulasi loss\n",
    "    loss_total = AverageMeter()\n",
    "    loss_seg = AverageMeter()\n",
    "\n",
    "    #masuk ke mode training, pytorch\n",
    "    model.train()\n",
    "\n",
    "    #visualisasi progress training dengan tqdm\n",
    "    prog_bar = tqdm()\n",
    "    prog_bar.reset(total=len(data_loader))\n",
    "\n",
    "    #training....\n",
    "    total_batch = len(data_loader)\n",
    "    batch_ke = 0\n",
    "    for data in data_loader:\n",
    "        cur_step = cur_epoch*total_batch + batch_ke\n",
    "\n",
    "        seg_gt = data['seg'].to(device, dtype=torch.float)\n",
    "        image = data['rgb'].to(device, dtype=torch.float)\n",
    "\n",
    "        #forward pass\n",
    "        pred_seg = model(image)\n",
    "\n",
    "        #kalkulasi loss\n",
    "        bce_loss = F.binary_cross_entropy(pred_seg, seg_gt) \n",
    "        total_loss = bce_loss\n",
    "\n",
    "        #backpro, kalkulasi gradient, dan update bobot2 pada network model\n",
    "        optimizer.zero_grad()\n",
    "        total_loss.backward() \n",
    "        optimizer.step()\n",
    "\n",
    "        #hitung rata-rata (avg) loss, dan metric untuk batch-batch yang telah diproses\n",
    "        loss_total.update(total_loss.item())\n",
    "        loss_seg.update(bce_loss.item())\n",
    "\n",
    "        #update visualisasi progress bar\n",
    "        postfix = OrderedDict([('train_loss_total', loss_total.avg),\n",
    "                               ('train_loss_seg', loss_seg.avg)])\n",
    "\n",
    "        #tambahkan ke summary writer\n",
    "        writer.add_scalar('train_loss_total', total_loss.item(), cur_step)\n",
    "        writer.add_scalar('train_loss_seg', bce_loss.item(), cur_step)\n",
    "\n",
    "        prog_bar.set_postfix(postfix)\n",
    "        prog_bar.update()\n",
    "        batch_ke += 1\n",
    "    prog_bar.refresh()\n",
    "\n",
    "    return postfix\n",
    "\n",
    "\n",
    "#FUNGSI VALIDATION\n",
    "def validate(data_loader, model, writer, cur_epoch, device):\n",
    "    #buat variabel untuk menyimpan kalkulasi loss\n",
    "    loss_total = AverageMeter()\n",
    "    loss_seg = AverageMeter()\n",
    "\n",
    "    #masuk ke mode eval, pytorch\n",
    "    model.eval()\n",
    "    \n",
    "    with torch.no_grad():\n",
    "        #visualisasi progress training dengan tqdm\n",
    "        prog_bar = tqdm()\n",
    "        prog_bar.reset(total=len(data_loader))\n",
    "\n",
    "        #training....\n",
    "        total_batch = len(data_loader)\n",
    "        batch_ke = 0\n",
    "        for data in data_loader:\n",
    "            cur_step = cur_epoch*total_batch + batch_ke\n",
    "\n",
    "            seg_gt = data['seg'].to(device, dtype=torch.float)\n",
    "            image = data['rgb'].to(device, dtype=torch.float)\n",
    "\n",
    "            #forward pass\n",
    "            pred_seg = model(image)\n",
    "\n",
    "            #kalkulasi loss\n",
    "            bce_loss = F.binary_cross_entropy(pred_seg, seg_gt) \n",
    "            total_loss = bce_loss\n",
    "\n",
    "            #backpro, kalkulasi gradient, dan update bobot2 pada network model\n",
    "    #         optimizer.zero_grad()\n",
    "    #         total_loss.backward() \n",
    "    #         optimizer.step()\n",
    "\n",
    "            #hitung rata-rata (avg) loss, dan metric untuk batch-batch yang telah diproses\n",
    "            loss_total.update(total_loss.item())\n",
    "            loss_seg.update(bce_loss.item())\n",
    "\n",
    "            #update visualisasi progress bar\n",
    "            postfix = OrderedDict([('val_loss_total', loss_total.avg),\n",
    "                                   ('val_loss_seg', loss_seg.avg)])\n",
    "\n",
    "            #tambahkan ke summary writer\n",
    "            writer.add_scalar('val_loss_total', total_loss.item(), cur_step)\n",
    "            writer.add_scalar('val_loss_seg', bce_loss.item(), cur_step)\n",
    "\n",
    "            prog_bar.set_postfix(postfix)\n",
    "            prog_bar.update()\n",
    "            batch_ke += 1\n",
    "        prog_bar.refresh()\n",
    "\n",
    "    return postfix\n",
    "\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "TRAIN-VAL-TEST MODEL\n",
    "=====================\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "A. KONFIGURASI TRAINING-VALIDATION\n",
    "----------------"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Jumlah sample training: 2975\n",
      "Jumlah sample validation: 500\n"
     ]
    }
   ],
   "source": [
    "from torch import torch\n",
    "from collections import OrderedDict\n",
    "import torch.optim as optim\n",
    "from torch.utils.tensorboard import SummaryWriter\n",
    "from torch.utils.data import DataLoader\n",
    "import pandas as pd\n",
    "import os\n",
    "\n",
    "#IMPORT MODEL UNTUK DITRAIN\n",
    "model = resunet_seg() #scratch_unet resunet_seg dlv3_seg\n",
    "\n",
    "#KONFIGURASI OPTIMIZER\n",
    "optima = optim.Adam(model.parameters(), lr=0.0025)\n",
    "n_epoch = 1\n",
    "lr_sch = optim.lr_scheduler.CosineAnnealingLR(optima, T_max=n_epoch, eta_min=0.0001)\n",
    "\n",
    "\n",
    "#BUAT DATA BATCH: NxCxHxW\n",
    "rootdir = \"/home/aisl/BISA_AI/ai-hw-acc/dataset/cityscapes/\"\n",
    "train_set = myData(root_dir=rootdir, part='train')\n",
    "val_set = myData(root_dir=rootdir, part='val')\n",
    "print(\"Jumlah sample training: \"+str(len(train_set)))\n",
    "print(\"Jumlah sample validation: \"+str(len(val_set)))\n",
    "dataloader_train = DataLoader(train_set, batch_size=4, shuffle=True, num_workers=4) \n",
    "dataloader_val = DataLoader(val_set, batch_size=4, shuffle=False, num_workers=4)\n",
    "\n",
    "#buat dictionary log untuk menyimpan training log di CSV dan tensorboard\n",
    "save_dir = os.getcwd() +\"/resunet_seg/\"\n",
    "os.makedirs(save_dir, exist_ok=True)\n",
    "log = OrderedDict([\n",
    "    ('epoch', []),\n",
    "    ('val_loss_total', []),\n",
    "    ('val_loss_seg', []),\n",
    "    ('train_loss_total', []),\n",
    "    ('train_loss_seg', []),\n",
    "    ('elapsed_time', [])])\n",
    "writer = SummaryWriter(save_dir)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "B. SET DEVICE DAN PLOT MODEL\n",
    "----------------"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "cuda\n",
      "Available GPU = 1\n",
      "NVIDIA GeForce GTX 1080 Ti\n",
      "----------------------------------------------------------------\n",
      "        Layer (type)               Output Shape         Param #\n",
      "================================================================\n",
      "         Normalize-1          [-1, 3, 300, 400]               0\n",
      "            Conv2d-2         [-1, 64, 150, 200]           9,408\n",
      "       BatchNorm2d-3         [-1, 64, 150, 200]             128\n",
      "              ReLU-4         [-1, 64, 150, 200]               0\n",
      "         MaxPool2d-5          [-1, 64, 75, 100]               0\n",
      "            Conv2d-6          [-1, 64, 75, 100]          36,864\n",
      "       BatchNorm2d-7          [-1, 64, 75, 100]             128\n",
      "              ReLU-8          [-1, 64, 75, 100]               0\n",
      "            Conv2d-9          [-1, 64, 75, 100]          36,864\n",
      "      BatchNorm2d-10          [-1, 64, 75, 100]             128\n",
      "             ReLU-11          [-1, 64, 75, 100]               0\n",
      "       BasicBlock-12          [-1, 64, 75, 100]               0\n",
      "           Conv2d-13          [-1, 64, 75, 100]          36,864\n",
      "      BatchNorm2d-14          [-1, 64, 75, 100]             128\n",
      "             ReLU-15          [-1, 64, 75, 100]               0\n",
      "           Conv2d-16          [-1, 64, 75, 100]          36,864\n",
      "      BatchNorm2d-17          [-1, 64, 75, 100]             128\n",
      "             ReLU-18          [-1, 64, 75, 100]               0\n",
      "       BasicBlock-19          [-1, 64, 75, 100]               0\n",
      "           Conv2d-20          [-1, 64, 75, 100]          36,864\n",
      "      BatchNorm2d-21          [-1, 64, 75, 100]             128\n",
      "             ReLU-22          [-1, 64, 75, 100]               0\n",
      "           Conv2d-23          [-1, 64, 75, 100]          36,864\n",
      "      BatchNorm2d-24          [-1, 64, 75, 100]             128\n",
      "             ReLU-25          [-1, 64, 75, 100]               0\n",
      "       BasicBlock-26          [-1, 64, 75, 100]               0\n",
      "           Conv2d-27          [-1, 128, 38, 50]          73,728\n",
      "      BatchNorm2d-28          [-1, 128, 38, 50]             256\n",
      "             ReLU-29          [-1, 128, 38, 50]               0\n",
      "           Conv2d-30          [-1, 128, 38, 50]         147,456\n",
      "      BatchNorm2d-31          [-1, 128, 38, 50]             256\n",
      "           Conv2d-32          [-1, 128, 38, 50]           8,192\n",
      "      BatchNorm2d-33          [-1, 128, 38, 50]             256\n",
      "             ReLU-34          [-1, 128, 38, 50]               0\n",
      "       BasicBlock-35          [-1, 128, 38, 50]               0\n",
      "           Conv2d-36          [-1, 128, 38, 50]         147,456\n",
      "      BatchNorm2d-37          [-1, 128, 38, 50]             256\n",
      "             ReLU-38          [-1, 128, 38, 50]               0\n",
      "           Conv2d-39          [-1, 128, 38, 50]         147,456\n",
      "      BatchNorm2d-40          [-1, 128, 38, 50]             256\n",
      "             ReLU-41          [-1, 128, 38, 50]               0\n",
      "       BasicBlock-42          [-1, 128, 38, 50]               0\n",
      "           Conv2d-43          [-1, 128, 38, 50]         147,456\n",
      "      BatchNorm2d-44          [-1, 128, 38, 50]             256\n",
      "             ReLU-45          [-1, 128, 38, 50]               0\n",
      "           Conv2d-46          [-1, 128, 38, 50]         147,456\n",
      "      BatchNorm2d-47          [-1, 128, 38, 50]             256\n",
      "             ReLU-48          [-1, 128, 38, 50]               0\n",
      "       BasicBlock-49          [-1, 128, 38, 50]               0\n",
      "           Conv2d-50          [-1, 128, 38, 50]         147,456\n",
      "      BatchNorm2d-51          [-1, 128, 38, 50]             256\n",
      "             ReLU-52          [-1, 128, 38, 50]               0\n",
      "           Conv2d-53          [-1, 128, 38, 50]         147,456\n",
      "      BatchNorm2d-54          [-1, 128, 38, 50]             256\n",
      "             ReLU-55          [-1, 128, 38, 50]               0\n",
      "       BasicBlock-56          [-1, 128, 38, 50]               0\n",
      "           Conv2d-57          [-1, 256, 19, 25]         294,912\n",
      "      BatchNorm2d-58          [-1, 256, 19, 25]             512\n",
      "             ReLU-59          [-1, 256, 19, 25]               0\n",
      "           Conv2d-60          [-1, 256, 19, 25]         589,824\n",
      "      BatchNorm2d-61          [-1, 256, 19, 25]             512\n",
      "           Conv2d-62          [-1, 256, 19, 25]          32,768\n",
      "      BatchNorm2d-63          [-1, 256, 19, 25]             512\n",
      "             ReLU-64          [-1, 256, 19, 25]               0\n",
      "       BasicBlock-65          [-1, 256, 19, 25]               0\n",
      "           Conv2d-66          [-1, 256, 19, 25]         589,824\n",
      "      BatchNorm2d-67          [-1, 256, 19, 25]             512\n",
      "             ReLU-68          [-1, 256, 19, 25]               0\n",
      "           Conv2d-69          [-1, 256, 19, 25]         589,824\n",
      "      BatchNorm2d-70          [-1, 256, 19, 25]             512\n",
      "             ReLU-71          [-1, 256, 19, 25]               0\n",
      "       BasicBlock-72          [-1, 256, 19, 25]               0\n",
      "           Conv2d-73          [-1, 256, 19, 25]         589,824\n",
      "      BatchNorm2d-74          [-1, 256, 19, 25]             512\n",
      "             ReLU-75          [-1, 256, 19, 25]               0\n",
      "           Conv2d-76          [-1, 256, 19, 25]         589,824\n",
      "      BatchNorm2d-77          [-1, 256, 19, 25]             512\n",
      "             ReLU-78          [-1, 256, 19, 25]               0\n",
      "       BasicBlock-79          [-1, 256, 19, 25]               0\n",
      "           Conv2d-80          [-1, 256, 19, 25]         589,824\n",
      "      BatchNorm2d-81          [-1, 256, 19, 25]             512\n",
      "             ReLU-82          [-1, 256, 19, 25]               0\n",
      "           Conv2d-83          [-1, 256, 19, 25]         589,824\n",
      "      BatchNorm2d-84          [-1, 256, 19, 25]             512\n",
      "             ReLU-85          [-1, 256, 19, 25]               0\n",
      "       BasicBlock-86          [-1, 256, 19, 25]               0\n",
      "           Conv2d-87          [-1, 256, 19, 25]         589,824\n",
      "      BatchNorm2d-88          [-1, 256, 19, 25]             512\n",
      "             ReLU-89          [-1, 256, 19, 25]               0\n",
      "           Conv2d-90          [-1, 256, 19, 25]         589,824\n",
      "      BatchNorm2d-91          [-1, 256, 19, 25]             512\n",
      "             ReLU-92          [-1, 256, 19, 25]               0\n",
      "       BasicBlock-93          [-1, 256, 19, 25]               0\n",
      "           Conv2d-94          [-1, 256, 19, 25]         589,824\n",
      "      BatchNorm2d-95          [-1, 256, 19, 25]             512\n",
      "             ReLU-96          [-1, 256, 19, 25]               0\n",
      "           Conv2d-97          [-1, 256, 19, 25]         589,824\n",
      "      BatchNorm2d-98          [-1, 256, 19, 25]             512\n",
      "             ReLU-99          [-1, 256, 19, 25]               0\n",
      "      BasicBlock-100          [-1, 256, 19, 25]               0\n",
      "          Conv2d-101          [-1, 512, 10, 13]       1,179,648\n",
      "     BatchNorm2d-102          [-1, 512, 10, 13]           1,024\n",
      "            ReLU-103          [-1, 512, 10, 13]               0\n",
      "          Conv2d-104          [-1, 512, 10, 13]       2,359,296\n",
      "     BatchNorm2d-105          [-1, 512, 10, 13]           1,024\n",
      "          Conv2d-106          [-1, 512, 10, 13]         131,072\n",
      "     BatchNorm2d-107          [-1, 512, 10, 13]           1,024\n",
      "            ReLU-108          [-1, 512, 10, 13]               0\n",
      "      BasicBlock-109          [-1, 512, 10, 13]               0\n",
      "          Conv2d-110          [-1, 512, 10, 13]       2,359,296\n",
      "     BatchNorm2d-111          [-1, 512, 10, 13]           1,024\n",
      "            ReLU-112          [-1, 512, 10, 13]               0\n",
      "          Conv2d-113          [-1, 512, 10, 13]       2,359,296\n",
      "     BatchNorm2d-114          [-1, 512, 10, 13]           1,024\n",
      "            ReLU-115          [-1, 512, 10, 13]               0\n",
      "      BasicBlock-116          [-1, 512, 10, 13]               0\n",
      "          Conv2d-117          [-1, 512, 10, 13]       2,359,296\n",
      "     BatchNorm2d-118          [-1, 512, 10, 13]           1,024\n",
      "            ReLU-119          [-1, 512, 10, 13]               0\n",
      "          Conv2d-120          [-1, 512, 10, 13]       2,359,296\n",
      "     BatchNorm2d-121          [-1, 512, 10, 13]           1,024\n",
      "            ReLU-122          [-1, 512, 10, 13]               0\n",
      "      BasicBlock-123          [-1, 512, 10, 13]               0\n",
      "          Conv2d-124          [-1, 256, 19, 25]       1,769,728\n",
      "     BatchNorm2d-125          [-1, 256, 19, 25]             512\n",
      "            ReLU-126          [-1, 256, 19, 25]               0\n",
      "      ConvBNRelu-127          [-1, 256, 19, 25]               0\n",
      "          Conv2d-128          [-1, 256, 19, 25]         590,080\n",
      "     BatchNorm2d-129          [-1, 256, 19, 25]             512\n",
      "            ReLU-130          [-1, 256, 19, 25]               0\n",
      "      ConvBNRelu-131          [-1, 256, 19, 25]               0\n",
      "       ConvBlock-132          [-1, 256, 19, 25]               0\n",
      "          Conv2d-133          [-1, 128, 38, 50]         442,496\n",
      "     BatchNorm2d-134          [-1, 128, 38, 50]             256\n",
      "            ReLU-135          [-1, 128, 38, 50]               0\n",
      "      ConvBNRelu-136          [-1, 128, 38, 50]               0\n",
      "          Conv2d-137          [-1, 128, 38, 50]         147,584\n",
      "     BatchNorm2d-138          [-1, 128, 38, 50]             256\n",
      "            ReLU-139          [-1, 128, 38, 50]               0\n",
      "      ConvBNRelu-140          [-1, 128, 38, 50]               0\n",
      "       ConvBlock-141          [-1, 128, 38, 50]               0\n",
      "          Conv2d-142          [-1, 64, 75, 100]         110,656\n",
      "     BatchNorm2d-143          [-1, 64, 75, 100]             128\n",
      "            ReLU-144          [-1, 64, 75, 100]               0\n",
      "      ConvBNRelu-145          [-1, 64, 75, 100]               0\n",
      "          Conv2d-146          [-1, 64, 75, 100]          36,928\n",
      "     BatchNorm2d-147          [-1, 64, 75, 100]             128\n",
      "            ReLU-148          [-1, 64, 75, 100]               0\n",
      "      ConvBNRelu-149          [-1, 64, 75, 100]               0\n",
      "       ConvBlock-150          [-1, 64, 75, 100]               0\n",
      "          Conv2d-151         [-1, 64, 150, 200]          73,792\n",
      "     BatchNorm2d-152         [-1, 64, 150, 200]             128\n",
      "            ReLU-153         [-1, 64, 150, 200]               0\n",
      "      ConvBNRelu-154         [-1, 64, 150, 200]               0\n",
      "          Conv2d-155         [-1, 64, 150, 200]          36,928\n",
      "     BatchNorm2d-156         [-1, 64, 150, 200]             128\n",
      "            ReLU-157         [-1, 64, 150, 200]               0\n",
      "      ConvBNRelu-158         [-1, 64, 150, 200]               0\n",
      "       ConvBlock-159         [-1, 64, 150, 200]               0\n",
      "          Conv2d-160         [-1, 64, 300, 400]          36,928\n",
      "     BatchNorm2d-161         [-1, 64, 300, 400]             128\n",
      "            ReLU-162         [-1, 64, 300, 400]               0\n",
      "      ConvBNRelu-163         [-1, 64, 300, 400]               0\n",
      "          Conv2d-164         [-1, 20, 300, 400]           1,300\n",
      "       ConvBlock-165         [-1, 20, 300, 400]               0\n",
      "         Sigmoid-166         [-1, 20, 300, 400]               0\n",
      "================================================================\n",
      "Total params: 24,533,268\n",
      "Trainable params: 24,533,268\n",
      "Non-trainable params: 0\n",
      "----------------------------------------------------------------\n",
      "Input size (MB): 1.37\n",
      "Forward/backward pass size (MB): 714.57\n",
      "Params size (MB): 93.59\n",
      "Estimated Total Size (MB): 809.53\n",
      "----------------------------------------------------------------\n"
     ]
    }
   ],
   "source": [
    "#setting device dan inspect arsitektur\n",
    "import os\n",
    "import time\n",
    "from torch import torch, nn\n",
    "from torchviz import make_dot\n",
    "from torchsummary import summary\n",
    "\n",
    "\n",
    "#SET Device yang akan dipakai\n",
    "device = torch.device(\"cuda\") #cuda cpu\n",
    "print(device)\n",
    "\n",
    "os.environ[\"CUDA_DEVICE_ORDER\"]=\"PCI_BUS_ID\" \n",
    "os.environ[\"CUDA_VISIBLE_DEVICES\"]=\"1\"\n",
    "\n",
    "if torch.cuda.device_count() > 1:\n",
    "    print(\"Available GPU =\", torch.cuda.device_count())\n",
    "    for i in range(torch.cuda.device_count()):\n",
    "        print(torch.cuda.get_device_name(i))\n",
    "    model = nn.DataParallel(model) \n",
    "else:\n",
    "    print(\"Available GPU =\", torch.cuda.device_count())\n",
    "    print(torch.cuda.get_device_name())\n",
    "model = model.to(device, dtype=torch.float)\n",
    "\n",
    "\n",
    "#visualisasi model\n",
    "#buat dummy tensor dulu: batch x channel x H x W\n",
    "X_in = torch.zeros(1, 3, 300, 400, dtype=torch.float, requires_grad=False, device=device)\n",
    "#inference 1x dan ambil graph arsitekturnya\n",
    "Y_out = model(X_in)\n",
    "model_graph = make_dot(Y_out[0])\n",
    "model_graph.format = 'png'\n",
    "model_graph.render(save_dir+\"model_graph\")\n",
    "\n",
    "#visualize arsitektur di tensorboard \n",
    "summary(model,input_size=X_in[0].size())\n",
    "if torch.cuda.device_count() == 1:\n",
    "    writer.add_graph(model, X_in) #visualisasi model harus dengan 1 GPU saja.\n",
    "#     writer.close()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "C. MULAI TRAINING DAN VALIDATION\n",
    "----------------"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch: 00000------------------------------------------------\n",
      "current lr untuk training:  0.0025\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "188f9e083577467495cc5268e8084899",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "a61304666d0142fca9d630583440da99",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "best model disave!\n",
      "OrderedDict([('train_loss_total', 0.07655044099796683), ('train_loss_seg', 0.07655044099796683)])\n",
      "OrderedDict([('val_loss_total', 0.05859080305695534), ('val_loss_seg', 0.05859080305695534)])\n",
      "305.7993187904358\n"
     ]
    }
   ],
   "source": [
    "#define loss terkecil\n",
    "lowest_loss = float('inf')\n",
    "for epoch in range(n_epoch): #ulang sebanyak n epoch\n",
    "    print(\"Epoch: {:05d}------------------------------------------------\".format(epoch))\n",
    "    print(\"current lr untuk training: \", optima.param_groups[0]['lr'])\n",
    "    \n",
    "    start_time = time.time() #waktu mulai\n",
    "    train_log = train(dataloader_train, model, writer, epoch, device, optima)\n",
    "    val_log = validate(dataloader_val, model, writer, epoch, device)\n",
    "    lr_sch.step() #update learning rate\n",
    "    \n",
    "    #save best model saja\n",
    "    if val_log['val_loss_total'] < lowest_loss:\n",
    "        print(\"best model disave!\")\n",
    "        torch.save(model.state_dict(), save_dir+'model.pth')\n",
    "        lowest_loss = val_log['val_loss_total']\n",
    "    elapsed_time = time.time() - start_time #hitung elapsedtime\n",
    "    \n",
    "    print(train_log)\n",
    "    print(val_log)\n",
    "    print(elapsed_time)\n",
    "    \n",
    "    log['epoch'].append(epoch)\n",
    "    log['val_loss_total'].append(val_log['val_loss_total'])\n",
    "    log['val_loss_seg'].append(val_log['val_loss_seg'])\n",
    "    log['train_loss_total'].append(train_log['train_loss_total'])\n",
    "    log['train_loss_seg'].append(train_log['train_loss_seg'])\n",
    "    log['elapsed_time'].append(elapsed_time)\n",
    "    pd.DataFrame(log).to_csv(save_dir+'trainval_log.csv', index=False)\n",
    "    \n",
    "    epoch += 1\n",
    "    \n",
    "torch.cuda.empty_cache()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "D. TEST MODEL\n",
    "----------------"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/home/aisl/BISA_AI/ai-hw-acc/dataset/cityscapes/leftImg8bit_trainvaltest/leftImg8bit/\n",
      "/home/aisl/BISA_AI/ai-hw-acc/dataset/cityscapes/gtFine_trainvaltest/gtFine/\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAADKCAYAAACohkc8AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8qNh9FAAAACXBIWXMAAAsTAAALEwEAmpwYAAEAAElEQVR4nOz9Z7BlWZbfh/22Oea6502aysyqyjJdpquru9r3OExPN2YGbgaQKA0VIAlKhECCEaJCH0Tpm74xpAgqQAqBIBSUQmQgCA4wA8wAmOkxPT3tTbWv6qouk2Wz0rx89vpzzjb6sPc+976sbAeIjdJM7oqs9969x5+9l/mv/1pLeO+5O+6Ou+PuuDv+/A75b/sC7o674+64O+6Of7vjriK4O+6Ou+Pu+HM+7iqCu+PuuDvujj/n464iuDvujrvj7vhzPu4qgrvj7rg77o4/5+OuIrg77o674+74cz5+6opACPHLQogXhBAvCyH+85/2+e+Ou+PuuDvujtND/DTzCIQQCngR+ARwFXga+A3v/XM/tYu4O+6Ou+PuuDtOjZ+2R/BB4GXv/Sve+xr4x8Bf+ylfw91xd9wdd8fdsTT0T/l854E3l/6+CnxoeQMhxN8G/jZAlumnNrdWCT6LAPxtPxfDOUdtDN6BlAohRNxEwNu2/p9qhCvVUiDk4kp/6ObxwsTbPl7ee/GtEIJO2WFtbQWEiN9IZlXFjRu3qKsZ1loyrSjKko31DbrdAilVeyzvHY0xeBz7x8eYeYO1Bucc3jqklOR5Rm0M09mcTCocgjzPsNaihEBphZQSKQRCZuxsbaMzjZACj0DGK7Pe8fpbb3Fhd5csy5du2eO9BwHz+Yz5rGZ1dcDxdMz+zVtIKairim7ZYWtji3JlBes8QghkfDRCgItPSZ5+nO3P5ee//JkA/NLjTU9bASZ+L+JxzdI+Cqg95EsvzMbPl99hOp7j7de1/JmMP226B+955corNHXN9kMPsaYkh7f2Odi/iceDdyghQQiMswjvQQqyLCcvuzR1Q6E1Sgi8VsyMRdQ1SgpUliGFZGItCmhMw1rZYWYMtmmwxiCyDOcczjk6RYkU4Fw4h/ceKQTOWLwIz6WT52AdHqhNgxICKSWZzmisxZoGpTOsc0jvQYWZIYSgqmvyLAPvEVJSOYd3jlxKrPN4ATiPcw4JWGsRWpNrjcPjncc2Dd5aZrMZQinwnjzPcc5grUNJCUJijQnXAJApMBYXX0I56NERnsYLGmPBe1ZWBkilcdUcIwWZEFgEsqrJe12mdYUsSjKlEB4Qgo1Ol1snR0ynk7CWnCXPcjKlEUKQ6YyD4YimqsKzVIqN9VWGh8dUztFd1RRSQdmn1AUMTzCTGY339He38cIya2Z4ocilZjIaU1SO/uYKRmc4INMZpmm4dnVv33u/zY85ftqK4EcO7/0/BP4hwMZ2z//V3/gg/c4OUsSltuTDCBEmp/CeWdPw1vUDZnNPp9cny3KkDAtGiIXkWAjY8HOxQEVYaIj2WwF4IRDe4+MWS2dPV7z4vwtLfKOv6JRqSZQ78KKVPkH+LTSAaI+wuBoftxVRygkRblwJxbsuP8av//VPUJYdpNAgMl65scd/9ff/e1556TtMhidMxmOEctx73z384q/8Ir/2iV9he/UMQkicd0ymR2S6ZFo1/N5nPs3TX/k6R/s3mY3HzKZTiiJjc2uDk9GU19+4jnMSnWWsbawghSQTim63Q9kpKfMOj737vXzilz/J2d2zFLqDFnkQet7zje9/h3pe8+H3vB8ZpXjjGwQGhedb3/sev/Obf8h//J/8r9C9Pv/vf/Q/8tx3vs6Na2+Sy5y/+Bf/Br/+t/59sjxHAmrp3bn4VtTyW12aJh6CsAwTplUAEo9bfpsiPPvMQ4VvBbtE0AA+ni8HpkBn6VgWyMLh23mTzp2USJox0ntsvG6Bj/uFc2hgNpvxd/7Of8bx4TX+i3/yW1zUiv/y//Z/51/99n+L8wZpa7pZjheSaTNH4hFasb6xw6UHn+DmrX3OrG2yPehS54qb0xp5eEhZlqxsb4CFl0+GZNZy6+iQj15+kOvTEYd7t6jGU2S/T2MMddNwbnuHXEjmTQ1ZhrOWTCrq2RQHjIXg3MYGfjbH4TkYDekqTV4UrK+scDydMRudUA5WqeqazBooSqQQSAmHJyes9QfgHHmRs1fX2HnNQEnGjQEhEU3DrKrIgdlsTjYYsL7Sx1hHUzfMhyeY4ZCnv/4N1GCAMA1nz+ximglVU1GInMpLxoeHMJ+xIaF56B7K4SE3ZjV0Cx5/z/u53M+5MZpz7eY+wtT8wi98jP76FtXeDUZas5pppgg6N66z9eDDvHa0T757no2yi/AOKTUfv/d+/tEf/gtefOk5qnpOU885s7HDWjlAoNla2+GffOnLDK9fYzat6K2s8Mu/8rN88V/+IW9UU37mL9/DPUUPLj7J2c5Z3NNf4uSZZznxhif/g79Bpae8efA6Luuwka/w7Je/xvnXRrz7f/ELnKxuYbxkdXWb45t7/F/+T//gdX6C8dNWBG8BF5b+vid+dsehpKbMB2GRtnIy+gdiYX/5KOyFlIAFH7byrRJIgndZ7IefLB0niXp56pO0wN8urpcvyRM39Av/JVk+3stozQE+nkUsjp/iNKedGN9enkiWvxA4oG6C54NIdrBja22Nj/7Ch2hkxfXXXkNlOdPRkNdffJ3fO/ldxKjL3/qbv47OOggEnbKLFIoiX+HXPvlJ+r0Of/onn+E4PrPxaMhb126wu7vJQw9c5MUrbzGfzTjYt2ye3aapLJPJNFiP1vHMM9/CqoK/8pc/ye7GDlrmUckJ3vfwuzk8PmifmsPhvEPGe9Q65/B4yOHBMY/uXuJX/8LPc+vaNZp6xt71G3zl61/g/qee4v3vfx9KLL07Fhb18jtZxjtlmgcsBLKMD3nhIy2GF5Ave2AE4R+8nDB6y/sJ2uO0x1j6PrvtO8TyeRfnKQjz4PUb17n00GXOVufYdAaFpPECJRXae5Qu8UIGz1B2EBKUCqoR59BSYaSkWFlhNjphkOWorS2yLKPT7TObTFlRCu+hqxQTaxmUHebdHpnO8UrRKUucdcFoEAKpFN57lAzrQ+cFDs8AgambYP3iKfMCGb2VaR2sXp0XgEcrhZAizmkRPI6yA0LgpcR6T0dKXJYFL4K4xpWmLEDhQSpkluGcb72TvNNBCdC9HlopHI6qqVE6o1vk2Mpi5w1OCS7eu8PWoIM9c5as2aRb1witGRQl+WCFrhlSaB3nZ1D8Wa9PT0qUkpRAtr2NFYJepxeEZ1Pj8AjX8Mb1q0xGJwg8WkpkVmCtZVLNcVbg/SG2adBZhtANQniGoxMaLejkms3uCmXWBdtgRyc4pVEbA/r9htpUWGHD9ypDAuu9knJ1xhyBMw3OC6rpGGOa22fdjxw/bUXwNPCgEOI+ggL4XwL/7g/aWAiPFPa0cHTJ7IriIMpegUBKQdIZ7TFYWNOn9Mnt/xdhryBak3+QVENSAvKUVxB2Cfu2VifgXVRAPoED4W+Hj9veBlyIhcV6SmIt7gAEcXJ6KuOwLhzbIZAe8iznPU88yvb2Ol/76tM88+3vcHTzJtXoBC0Eo1tHWGPRmYunkgihQHh65YBP/uzPk2nBpz/9+Vapjk+GXL16kzO7Z3nw/gu8+to18BmqkQgJxljmszlaaXxR8+oL3+Yz/ZJf/eQvUa502/uWKLbWd+IdWzwuwDre4fB0egW9QQ/jLAAPXb6Pn/vZj/Kvfu+Yk+Mhr752hc//wR/x6GOPsNLpLBTj0hO67WktP907fre8zbIHKLjzuH2/27+7HawUd/juTtvdfpxev89/8nf+Q77zzZd44fmX+MB7HqerNUpKBBIZBamUAhVcRZQKc815j1IKLzzHszGNadA6Q+YFOs9QOsMLSaEUxjsUgsYZcp2H7wgGi9KagJt4jHPRUPEgJM57pJLI6Ik3TY3wHuch0zrMU2Be14BEaIVzLioRFZatCHBTnmU4H/xw4xy5EDilmNd19MZAaIWMsFSuc4QMSoR4HVlZovOMey/fx971mxjhqZuGbtmh0y0ZmQnOVayf2eLi4w/RyQtEnoN3DKzDe4dAkfX6dCYVudI03gU4zDlU2aGMcKQUAqk0jYBu2UE4B77BeYepap49uMHJyRF4HxS3VDTW0DRT6sZxeDKBpibLMvqrXXorfQ6PD8k2Si7v9FntrFCoEtHU2MkBTirUuV06g4z5fAJKUmYlIhoTW7sbZEXGTIA1Fc4J5m6Iad7hisB7b4QQ/ynwBwRD6v/lvf/eD9reeYNzFagB7fKJP3x0rluIR4iAV0drKf1LwvW0LyCWHQFOHfhtnyyW79sWd/wlfesTEOQ9Pln+AePBCxG8geQMnFIwSRdEcMq3N5mA4yUQy9OYBtMYIjxMUlGZ1pzb2eUv/tLHuf/++/jyl77Mq9/7Hp0inKpd0CikyOJJLQjo5F1+6WM/Q391wO//wWfYe/MaUkqmozFrnRX8YJWVwzFeaNY2tpjPJjTVHLzn0vl7eezxJxj0BwwGXcbDKdsrS888PmzvfXin2GBhR48mU5osU9R1BRgm8zEf/sDjvPHmW4yGJ/R7Ja+/9SpfefZbfOzJp+hlBeq2F3gnAXu78L6ToL/97x8k0H/Y73AbFHXbNslvu/33tyksIbh/K8C6gw93+Z1PfZ7LD9yH1joYOUiUDNtJKdp5keaGS5Cj93jrUFLhvA94erInBEgp0EoHpeItaf2IJSsqebDee5AyvDtrgzElZJz/wWAQgLc2eL8ED8x7j8CGk3qHJcSTiN5A2DeezznS1HTOhvsjCN4QmxCtV+69DyaaENFQAy8l99x3L0WZcf3qNeazOYUv43MRZIVm0O9ya+8gnDM+MxlfWtHpcvbcGTKtowcvcN7H2IVCi2goChGUISCVBGex3jCajLh17SrDwwOaum6NRe89s6phOjeMRjNm0waJp7vapdvLyAuBLmo+9IH72NgYUOY5UmjccIade2QvR/e3Uf2Mpp6hOiWZyloDtL91Fru6zRyBtAbTeBpn8f4nF+s/9RiB9/73gN/7cbY1RlDVkjwHWqGaDhQn49IkFvK0gD+18JfMMXFqKfpTRvipBe6T1E64/iIeEMV7CwV5wHsH3uG8xMsQzAoLM55naZGlvcLh40QXxIV9WsmcslgFeFvhrFu6Wo8QNlosirLo8vD9D7C7scE/HlXUx2/gRQgAQgyko3DeYN0ULUsgI9d9PvTYk2ytrPOpz3yGl773AlrnrG9vMfYFw5MJg0GfbqkRso9Sikzk/Mon/wof+ujPhJgMwZM7bQMvnu+b+zcY9Eq6ZQ8lQCKRSiKUwBgLGJ75/gusrvT5lU/+PFeuXGE+OcK4Kf/0v/m/8r33f5Rf+7W/ycUzZ8iix+Hbp3D63XPb3z/Mcr/T33ey6tt3fYd9bx+3X9fyPEufv83TiBN4td/hiUce5Ovf+C7GNiAIgVgVDJ4QaBftLsEBdUgRYJMsK2h8DT7EOaT3OJ+EqiTFrawPcQoZJ1wQ+D5ayoBUizWmFD5a4wIZBXmwRqSSYb46t7CSogHm4+deSYSPSkIIvLNh7gPWphB9OLfzHuFlfNhusSDaZ+ej8vCcHB9xdLiPkpKVzXXkcBiVZ/BINjd3yCTgfDxUkCVBYTkaFRSWlAEyWxhM8dw+PGAhg0GXoLD5bMStm9c4urWHqYIV7j1Y46mNZTpvGA5nVLMG5z06U6yslHS6kvWNjLNnBqz2OygjkDKHOs6CyqMKje4qzNTga4/sJwJMYKF4AK3xQmGtwjfgrUDpDK1vByt/9HjHBYuXhxQeIaIQFkvLUCyWZrCgg8ANgsiDt+0LTHCO98m+99FyIrzcqF8WcNBtkIJP9sOdF3SySpa/9QEBai01v7SDT0dIEFFaLA6c8EgZXHCfri/u750PjBzvccbhkyWHTodrPRgpQEpJWZb0en3sWEMUEMtqZVrN+PK3v8p7HnkX2yu7gCLPCt513/1sb6zx6bNP85Uvfp3+yiovf//7lIVme2tAt1BoLZkIQSb7FJ0uKjIjTo/bwRnPlStXGR7v8bGPfZi13gAtApNFeI8xBo/n+HDMM1/9Pv/e3/p13v3Y4zzzzLfJc0mZ57z4lS/yX73yCp/8tV/n5z/wC/SyYklJpifww72DOwngZaF9uzL4QcrjTspg2TPgtu+XP/uBxkf6TAgeefgi333+eb7/0reRIkAhKqBBOJ/iHkEIi+hdJfw9CW5haBlcRGjRRwMDAca66MEuBGB4jItn6ZcMFRHXmPMeKWQbIxJCRgtaBFjF+1PHcM4FkDQpsvRNPJeUCoHAWNs+GOc9wtEK/LRtwni9C2yl0a1j9q5fD0aYCy9BSoktgzF268YBUgm2zuwy6PbaZ4UPgl1IGTwRAnspmS5eLMBh76PycA3T0QlHRzc52t+nmdXhWl2Aa+e1ZTKtGY1n1PMGPCgt6JSK1bWMM2e6bG/16HUztNIooTD7NaIjUHmGlA4xKHA2PHNZKFxlkGXWPi9rHY0FU+fUNqdbZmSywhMgKaX+/8Aj+ElGWXRZX9nEGNFOXJLA9oGimSa1ENHK8GHier8Q0gthneaTWKy+ZSfj1NnF7V+/7S+fJHA76SUeGyyGNJlvO7rwIrJawjUkCMnLU1fZTv4gWCReEhlMYJ3HuoC1hwktEcTFTwgoIwRSS3plENjW+2CttVcksA6+8CfP8tWvPs1f+yu/zCOXHkSIHCkKttfO8uuf+EV2N9f48pe+x+GtITorGKysYL1FSegWAqEKVJ5Ilm8LjZ56ch4oioIvfOHLeOX4iz/3i+R5h7LoorIMb8NiLIuCeTVFILnn4llefOH7yMzRH6yRFx2ODvf4H/7+f80Lf+F5/tJf+Rtsr+6wXhZt4Pjt7/JOb+9HC+M7QTh3UiB3Eva3H/NOCudOo/3ee6RWPP6+R/gf/tEh1nmKXEbWFSgPSmtW1lY4c/4e9q7fbJVhUA4K6TxGmLgmwjGjcRuOIwTGO6wzCDxSSowNc0pImSRwC7F6FyznMM+CYeFjfEDKaGQ4SPEK2SqOaMnCEiwUju/j2kmQT7S82hiFkDLgn37hRydPecH/Jcl1vHfY2lHXNaYp8d5j65p5Y2k2HJ3t/sI4dMT1Giz4ZHiFYDQI5/ESEB5nG+azKYcHNzjcCx6Ad+E+G+OYVw3jac14NMfUBiEgywVlR7G2nnFmp8vmZpeyKJAyvB8hJcJLhArvQymBa8BVBvoFPleorkRrCVJSG8fB0TE39qasrJ1hc6WPkppOrlA+KE4JyCz/AbPrB493tCKAIChPmXhRji0sNYGP+J2UC475Yiwv3eh6CnHbQl7mCy15Gq1tECZOGwBewlJb4yl+LlrPIx2dOEkF3kVXod18saBa6z/M8FYN+fYAorVgGutxxmKtQygTeeUqmmzBbdciuPt5ZGZY12CMWX6yKKmYzyo++9lPc+3Na/zaX/9lPvTEB1jprQM5RbbCu9/1CH/62W9T5BIjcqazOUWRheClEGTqtLv+9jDuadG4vr6CFgXf/er3+bmnPkIvH6BVjtAFlQkud6fbCXkP3rO+voaUGZIKJ0NQfGPzDAdHV/jNf/D/5Omvfp2/9Jf/Z/zKJz/JZqf7Nq/kdsF9u7dwJ6v+hymTtP0PteZv2//2J3Lao1w6dqSWnjRzXrt5jW9+9zt89Utf5Pj4ANs4VC8Lz7zI2NzaZnN3F6E00GE4fg3dqQlGRFgkIa4QhCOOiNtH6zxqB2ssNrNtwoQgWb9JHgevXMT/WCJeEOfughhB+13MFog2kidBUc4HhSN8sNZbkNT5eOjFecJh4/qI8ZBk7AlOe8qnnqX3eONQKgp2G5WKVpTdXjh2Uj7OYUyNjQownM+1sRVszWw25saN6wxPDqlnc2wka1jjmFWG0XjObFpjjUUpQVkIen3NxlbBmZ0eK/0c6QVUDlkmeE8ihUI4EFKBVjgkSI/uavRqDkLQGMdoWlNkA6YnDS88e5PaSAarCqEkeZZTlhpbNwjvUV4g1Z8xRTCd1RwcD1kZrERrh2AZt+ImTiMXQ2ZCBgsiBb6iGxmE9ek8gDSRFgszCXnSX0vCI04kUtDWtccIW7P0W4ShnG8Fc3uGJdzAJ4Vwm+cRzhsVVjpfiwd7cB7jLE0TEouk0HGPiL8i8RggwAUy78VjOGxk5aShJfRXVnBG842vfYebN25w5ROv81d/9ROc3bqMQFHkXS5fup/XX32V/b1bIcHHhkC9kAIpPHVV41tG/51E42KUWUZvZYCSOr49S6Ygzwsa5zDeUuQah8Z5T57leAFNTDbCB9pgp9NhPq74/he/wq3Xr7B3+Aa/8df/Pc6urrWB5NuF/p3e2fJz97f9/oPu5E6ewe3f/6DvftC21nte2nuLL33zab719Nd55fvf53jvFvV0incOLQVKKXResLqzxdbORa5ev4pzgnvOXgAPJ8cHdHvruMaEhKxoJTsR8xaEjzh5SHYSgHcG42y0yAOtV0WCa1AChHUFMYnLtlfeCmTvAx04BpWT1wARppJBz7j47mj3SccP68EHyR7Wh3M4GTzdtD1L+6YRvJtIEPG0UE+IF55+A9679rN0rV4IpJXYpkHakFBpjaVpak6OD9i7dZOjWwfMpnNms5rRcEpdG5z1WOdomqBAskzR6UrW1jK2dkp2t7v0OhnSCpq9OWZmkUqR9zqITLXeu60iuqE1aqVEaIkXgnE158bekDde2eNoIvjEL34Mbz3jiafT0yE+l2V0y5JcS6wvsKbCe4ESnp90vKMVQZ4Jul3dBoBbpH5ppYmIbyIIjArChE7BXeFdK/V9KyCiJeJPqZPwe/x1we5YiIjWS0gbLXumaQITdICPUBVIhPTxw8V9tKDVbQFp4oINsBOt0kj8EInA25qqsUs0So/AoIRHuBhXkSEQW+T5Qoe403aoUprzF86xdeYcB3vXefPVm/zeb3+KN64e8r//j/5Dtrd3KbKS9Z1thMpxpkEpjdYaoRRKZoGZYsOTaVyFFhrBneIFYWit6PYKbB1YKCBCkG+gyaRAodCZjtaaJcsUOIOzBu8szlk8Hq0VZZkxGk259cYtPv9P/jn39M/xa3/91ymVCgyy2879wwTzD/MMftz9l7f5YcridoUDcDCa8P/4e/+Ab3/jC7jaRIFkyDOJQONqQ55pZFYynYBA0cw9UkeFLxVFhB2styELe2FKLF1UoIgmg8k7x40bN+gWPaTS4B3ehXibUMFydy4I+sUNyTh/HVIqnLMxfyAEoVsvAd9SP72PsJBSIfFStj5GxOfFYv5LwN7mbcTjOX+71xmEOmlfliHi5LaH+8mVbNek8yG2EWxFQUyeDsl0s4orL12hns+pZjV1bajmDd4J5tOGxiwCwx7o9hU7uwVnz3TYXCnIo6B3Rw31yKK0RGUavVrGWN7CSJUdheoWqE5O4zxHB8dcfWOf/ZtjxhNLPxPgSoTU8YQOoSRKKZTOKTMdkm1Vp4XcvH2H00d/0hGyBy1aEaheyT2MwrPl8af/J4shcuy9T3UeIpsoWiBhQSzcSXFKCiznECwrgzuMU6s+MTGiX5DOcZvJGa7Tc0dREa/Rtcojnt9B8CFFVDQ2WDAy5b8qkgcQmBzBzxfC0yk1EoHxEmvTeaMikAU/97MforE1n/3sn3LjzTcxVc3+K1eZDCdsbwfBr0TG+mCDPd6IpSU0UmmU0IEtIsIlvvDGKzg/59FLj6DpLDG4Fg9KSUWWd7HNNLKYHNZ7zp7bYNAvkHFKWhugr1wFjwcXgojOOhwOpST9fsl8VlFXDVffvMbnPv0pHnviCR59+KE7CuHTINXiDaTf7wTXLAd/f5AXcSeFcSdv5PansXyM4cmIwzevU49maK0oMoHulvT7PeZzw/hoFvj7UqCyAo9sBUPKLSjKDhKBc9EIas8U2GLpWmUKrgF7N24wGo649/7LSCEwMTcBBN7aYDxFKDQdb9mSD+9QBJrncs5BMl2ciTkvcZ2KBMkkSCcxeQhzP5a3aIwJQeVI6UwxgiDsYh55DEoLucznYwlSojWWfPwirHePtQbvXci7kJK6rmiqOfPZHOcd8/GU2bRiOJziLOS5DnEXuQh2CwFCCS4/vMa95zrkmcAeGexwjigypJfoQiE7Gj832FGF2OxEHRCVUqaoasfh9WPeuj5kejSiJzRdmWM6nu3tPnv7M1648jLrRYembpjNaprG4K2lchXaEmE/jbf2Ngj4xxvvaEWgtaIs8iQfW6mdSkEsVmaaAJEf4ZNFvdC8cYOlVb60jAW83c5I+kK0i2YBDS0E2x3Ji37ps2iZJLpborqm0hiBXupCDCFqJBFpbsviRxDiDxKBs2Ca0y87JdSlBeiEQkgoCx3v22FNQxIK6Y5X11b4C7/wc9x37z18+jOf4coz3yLPFaY28ZF5zpzd5tzuFs9/z4GQKKWRQiOEjDVngnussw6/+4ef5eSpIz7y2IfRsscCkgvPWWtNnpfMxtNIF4VCaTo6pzIN3gvyvAAF1jVorQEdrEkIwsNBnuf0+32aquHgcMR0Mue5736HT//+v+KeC/ew2u0uso5v8w5uV8O3O9LL2y6TdJcVxg9SAj9oLFv/d/p8sNpnZW2VTkejMwXC0OlmlKXGO8dEJms2MmJ8Cq3G3JlYZ0fJaHkniAaCgPVugbcTIRQEo5MxG5vrSKkXWLz3gXkTsUzrXPucQoJg8C5boQ7BE116esladh5kvDZaRlPYzqdYRfK147m9i5avjms/egzJa3YR4knPM0FDp2LH3pOYdQjw1jE8GWONbRVhqEmmcK7G1A3NdIZ3nvms4dB5bG1whgA1CRlqayXYyQe4TSrJYKWkKDRSeExtcJUlyzOwHpREKoFYzRFKInOFRzA3htG0pp57bAOTmcNYybnzu/Rkzry2FMazvtZlONrj4OoefnOd+aymXOkhJQgc9WyImTfkeUmZ5QGZaDHoH3+8oxWB92Ctb70BINIyWXIXWZLnCybR7VZXu5hFO+3irv7UAl1UHCLKcd+uYH/qApLIXVxrslpSCjwO3jgYMzmpwFuEFGRKcv+FjehKc2ry3zjYZ3t9g7yILxTRni5do4vWd2MjzzsqP+8h1zJwiL1BWodDkhUdPBqPjYrgtEiSQpPnHe679z4++Us5v3ntLYoij4tPIxBsb22yc3aLoIYEQupA+RQKiQyLC8ED587zlz7+i/zRZ/6Yad3ws09+jI7qAFnApwm1kjqdghMRcH/waJ3zkQ9+mKe/8U2mswlFXtApJcYaUIvkIyEFygZBl2U5K2urmKZmXjUMR3MO9g/58uf+iI0zG2yfv4D1gqeeeA/n1jdPz4HbfvoYGAyRlvB8HQGGS5h2+/6XnuAph/C2Y97++bLy4LZt5tbwnZeeYXiyj5OWbqm4cOYMvf6Asuxya3/E4f6tJbgtCFSLoLGWxvnI9gnHS5RQKVOeRYQgIy5PnKMAvZVVBisbaK1pGhM9hiik4++tAlyCfBJLz1kXIdcYoxNEgoeIHrholUtrpQNKSZwLsFAwYELcInkVSmeopEDS8dq5vjhegsVuf+4qJo4ppdvjpEKLCBW9isUO1obrkFKGYnbGRQWQ7jn8S4UWPSB9TOgUURkBMpOgQyE6bw16pUu+WiIyhUVwNK65dmPKwUHD6tqAh+/bxlvHtBmztbnDgw9cZpAVXHnhBbKioPQWKRSz6ZxbfohUnjyXZFko/KhR2GbOdD6jsQKJDRDfTzje0Yogz3K63R5NY1sJnsS2EKLNLxMuMQoinc25MMH8ovDb7dZbsrOXPzm9SJOnQCvETknkJSsm/RlwULBycb7JcM7etTEy12R5hpuMuf/cBiKLdnK0Loy3HOyN2F3diotnOcch/hJP5KylaWw8h6MNYDuPlgqhJY0DKQ067wTL3RmstXFRLQAPLQLNT0lFp9ujLPsoZdvAskDQLXP6K2tkRTdclxAh0UgESMBG2qcSinddeIjer3b43T/+Q8z0C3z8Yz9LIXXrb+VZzmC1y9FhFpPigjW3MljjZz/2UTKd0Z/POLO7TScb4N0cqcLiTBak82GRFp0u3V6XtaZhPg/u8quvvs6n/8U/Q6qMbjHgwv9ul7NrG63RcPs8MN5zUk949eobZJnh/PYFhM+4erzH9uo6Z7prIbP0tv3upFiWj78s9O+kKJY/G9cN45MpF+69zNXrb/LuR97FSr+LLgpU3mFW3STlyVgCtBFw6wqEZTIdRcpyoHL6ZIiI02cN34f3mkomZFlOXhSxQGOEfayNpSQSEy14lME6d0tKIRkxFi8kUi2s9cC5d62l7mJAOmX2JrRoOagbtot+TvxciEB9Xr4VIWRYaG2trduGj8woFxSeRCAyjTeGuqrIs2XYSKGkxJoAN6o2jhCvS8T7jc9BtArJLWRIetbOY0cGtZKTbRboIkdkOYicw2HNjZtjJiOPtRnra5vcc37A9plzFOUa2+ct3kv6K6tktkIpzWQyQRRBRCulUNYghEPIALEKrfA2UIK9rWkQKF2gnHv7M/kR4x2tCIypmc+nKFnE1bVYPj5CDt4lZk3rCLfMhWBMpE9DVuWyVddKWpGAksV3C09gkcLPknfA0jY+agHvgjcoUExrD76h11OcPZMx6Pcw3jOdOIzOMEIhfRDkxlQoBEprjA+c4KSAfDxtuHUXXe6Gej6PixGILCHrHFUsOSwzhXYZnW6GVhK8pTENKejc3qWgrW6aa4kuOgg7wZpUGFlR6MDq0boACLS3hNVKFQW6j9/Bxa3z/I1f/lW++43nMI0hKwzCZwgBWkt6nZyiLNpgcYDxHHneBcIkX19bQWclsrEoHYqeWRdwXe89Sml6/T7VfIa1hrXpnP3DCePhhBvXb3Luwlk8jtl4cse55QlK4NW9t/jKt7/C6PAIY8Z89CMfR6suxjd8/plP84H3fYT7ts6d8gyWlcAP8waWvxe3fb7sjawWBT/7oY9wtLfHt77xZQarWwwGnVBnR0vyTkkSTFKC1oLa1PT7HfAWW1coIQOrSoV5ENg/yWgKAt4ag/cGKUVISJMCnWmyLAvP1flW0BGfs40lplOhvrQiWg+cUG5BRMNAehX1hmtvWIhF4bqwVDzemnZdJYw/sHrCBVtjUFK1s9X7SAYRslVyUgq8g06vQ7ffp6krrA2EApUptFbY2uIgUDvTtUiFj/WEjDVIoTCNx7iGrNBkmcZYG+C2dA9Lbz95Pe0jicaR0ILifA/VzTBI9o9rDk6mZMUmk5MpILh0+T7Ob++yurJJZ1BSlFkIgA4rTBPKZms0zsDwYMiRt1SNxRFIAC1YKWQswJgSSz3YBq/yJVLIjz/e0YrA+RAjVIoW70kPP2GTi7GM3vtATGgx0dvts/hbesFRmN/ZfV+W/MsU0sXXCa31wGxWMR4N2VeCuo413oHjcRUsD+d45rnX0VKhtMTYhm6h6Pe7rO1ug87xmMCagNNSxge6m/MhgaWV4nFyxLgtdWPBG4QTrPb7ZHkX7z1NY5YOmCiw0dJREpVl6CzDmZjh2RY1U2il0CpaSlK0rBMpRCSOqphDESbp5mCDtc0NrHfcGt4iFxkbgy0m8ylHoxH9bqel0C7enwcknbxkXs85ON6nM+jGgmvtmgMP2xs7PHj/Kl/88tfAWUwTUvqn04q9G/usb6zSK/vsH+y97c17Aix1Zf8az3/3Wa488yKdbsPetdcx1ZzzF+6lMYLXXnmZ6XSf+iOfYLW/wUY5oNT6bZ7BnUaaR+K2bW5XCKOm4Z//8R/y+ovP8NJLr2Kcpez36PY7EMs4FGU3CloZhSM4U4E3WGsw1qCyQMcNKE2stb8kEKz3GGs5GY1AqFDAjfBuhRQ4s2DpCLnoJ9F+Fr3IRMFMxlIiC4Tv4hxN9ycWnH9iDEEIifWmLdMQWEeBu++J1M8IewYFtJQLgY7QUYguK6GwODr9Hg8//mjoOWAN1XzOdDamqSuEEAzWBnQ7XYpel6IoQ0a0D/IhTn+kzJjNRjg8ZbdgMp5RNyYSU0RaKQvHPLFORKqvFBI9h9Zz/eVjbt6YMxtbcIKmvoHScOHyDgzW2Tp3D6qxlGVBmetwHhEqC1hvAz3cNDhrmc+qsBYFASolOWkJhnNR0eYoXaKzDimP5CcZ72hF0JiKyszI8mJh7bf/gzY6FL0zR8BHjRUoa8MkhaWJnCZj2EkmIUi0vqPFsPAtwsnC9kvBh7ZmRGIJRY8AT6YkuRQM+mvcunWIVhl5kWOtRasM1xiccTTOUW70OLg2Zm2zxDeW2XjIalbSKUqsMEGwxnN6u9BaHk9dWVLNoJbsqjRSeTQG0xisjVUcdYb3ouU8p/uCWMVSgBPJFQePpDHx/giTTipJnoX7bIOvCYOOAmfYjPDGMOisB/dZgrcGYwzfevbb/OKH/wJYuPLyq+zsrMcJm3yxhcbr91Z46j1PcnRyzMpqL1BUlyxMAK1Knnrvu3nuxddAWPCBLTFrHKOTMXs391lf3+Dw1iGNtQEnXnr3VVPz5S9+nldefpZrb77O2kqXra2LHB0csXvPvTz6yJPMRwd87St/wnPf+zYf+dgv8ys/86sY7xk3M3Ih6OsyCgLPYnYtmEZ38hq47bv9W4d85VOf4q1rrzAZT/GEgGRWlgiC9ZvpLEA3UiFUgRdwcHCL+XSEzDJW1Rrb2xvMZvOAaftTPmxr3Y/Gh1x/8012zt0DOtCKbYSUbIJxnG3po2l/H4V261unQG3Ed1xkF6XgsjcuKvmwTyD3JJgoCDNvI6QZPX1BgokkSuvANFpiCqVCcHKZALIE1+g8Q2qNryreeusKdT2n2+3gvKfbH3DuwsUAdQGh+q1EqKQIJHmny/HBfjCWRPRSjI/B7eXC9MvrX4AXTGeW16+OuXVjwtFhzXwWkAqpJVtbfU72x1jhaOYNz37laYbjKT/3vscpuoFK67yLitvS1DCbjTkazyJtN0ijppqjlYu3HJ6b8zakXciSbneDPCsC7ZjT+UI/znhHKwLvFNj08gSLpeaXcgKC1VI7y2ha4RoXam04S6bKWB0xLVa/sJR8yjfw4DyNM5wMT1A+1l13ITPROxcUkZBhMkeIJh6IBDbF5EYcBcbMODrxqFKFJioqlIMwJnRRqhrD+koPlKe/2sO4UC449x2a+RhWstNlrRPTA0hUWONNqxUFAodjODzAuAydFSgycLQWngeqepEMlPzdUIpaIrEI4ekUGeOpx0WGkUChhQ711YXA2hpvq6B0vGE2q3HMcK6mqiu++M2neejyJS7u3hcgBgQ765vce+kyzlm21tb55V/8KH/4uc/x1o09HrzvfrQoIbq4xEV69sxFzuzuMJ5N0TpkMjvnYwAzZHoPVgY8+eTjfOubVRDE1iHzgvlmRWNrrDMcHt1gVM25WU3JpGV3dRcJlHnBB5/8IK+++jz9TkaZDzgZHdMbbKKV4tVrVyDr0y1WEN5xfLLHjfENJvMpk9GY85u7lBvnGVVjjPWslT1yKVk2IdL4QV4DPpR5rud18BxlCL4KKcmyAu8d87qhbqYsWGcS09QYM2RzO0dJhamOGR4P6fQ2cDZ04iImlIVEwsAYe/P114PQa0sQLGJci6Cui0zdIJRlVHTeOZAK70XIaI/rz7lUylniRcDjUzez5f5rIpX69EGxCCUhFaAjxhAQoZyCj7RUISJmH7xeZy1Sq7YUdYg5xCzleJzjvQNODo5ROuxb5IF4sShIufB0FvkH4f5NFd6F854iD57ffFrjRWJcsYhNRmq393D9jTn1ZB77LWT0N/pUozFSS4qexu8Hz6FpNF1mfPfTX2S1LHjPIw+GjmIm5IwgwBjDdFZjE5tKRG/JOVC0FVy1lCihcbqg09+gyIvw3BRx7f5k4x2tCIoio+wU8a/Y0ynRfiJuEhaIZz6vqcYTSu3xVPhmzux42LqwwY5IcEhyoYNWFwKUgN1eRpbpOKkDHut9yDR0to7WsIwFACP9TMYeBZEXPLcCfIY3DVJIjPPMZzV5rLlijGUgBX5uMI0mcxZTFLhmTt7JQiVBL5djw60b2goUT8jm9Qu833vP17/+DNVswkPvepDB6iaZ0hSFjpCRoWnm8QCKZO2nAnUuxQm6q/ijWzRtddPwrLK8QOcdFJZOt0ev26fs9VlbWeH7V1+leOYrvPuhR9jZvIfPf/HrXLr8JmLexdlzrcUVknY073nkSTbWNnjz9bfCgo+lBRaCIymoEiFnoAJW7IiBYgTWWTKtefD+SwwPD3npisWZJgS584JcZhhjuP7Wy/zuH/w2r7/+MvedP8/j7/0Aw9ER6+ubqLzHr//13+AbX/wCN66/ytjUPP74o5w5e4HP/9Hvk3VKLj/0Pvb332Q8POIf/+Z/w7yu6RRdnnjig6h3d3n2+W8yGc74wBNPcXHzXGsQ/LjjYP+Y+WyGcQ5dZKz0+oBA6wznLLduvslbb74eva4QJ9neXKHX3eTk8IjhZALWkumA8xvrsbEGjncBInLOsnfzGk1t2NjeiQKemOxFawylwoTGNmBigFLJgIerEDdoS023XnDwIFPWNzIEctv+GwSYVkXWV5rSkBIog3BVIrSnDBi/i4X0CNfkfMwEXvT2WBQYTEXzHM4Yjg4OcTZ0N3PGIkoRvJ5oMAUyiW+Ly/lIshCEOIaz0Bt0UEowHs4wtQmNoNIUZUHiSHLIOkEo8QJOaAY7Z1nbmHOyf8CNN0+oa0tWSKRr0BmY6Yxnv/ks9547R7/TDxnbSdgn405EYzRWUmgR7hbKUnR7fYp8gHOx8VX7ZP+MeQSpYEILj/j2i+iiJasHOplmc63AVBVh8i1gFaIlJeUi4zVxjyE1nJCha5dMHkhcINFzSHhrqr++6HUgFpztiB5trrsA5XjBxBiuvLrHoBu8k5NpxdmtFXrdDB/bb0oJRZ5x/eYBGIc4s5kM9nbCtY65DwGiajbFWoNUBakuS6ZzPvXZ3+WZZ7/OI488zr33PUBWdijKLlTHUXkkeCQsItUqRUkhFUXRQWBpmoYQhA6wmVAKoUqcm6F0icpKut0ujz/6BNvnt/jSN77Bp27e5PL97+by2Qv8/j/7A86fu8h7H34AAC1kENJCkMmS+87fz5mtHbQqWLD1l0GUZKmGTOPGNkCkFwqBsRUqy1gZrPCBp97HtLJ4ZzgeH1NFC9uMaopc8flP/RZeGw6uf5fPfv53aCY1WafP7u69XLr/MgfXb3DljStkuuHpr1pWN7dY29nh8cef5PWrN9nbv8YLzz+DcBX97iZNNufl559h7+Yt5pMhqhjw8rWX2VjfZiCzNrD8o3SCB87sbIZ6QbbhzM423W6XIs8jjREy2QUf8H9nDU19zBuv7NNUdehBkakQ28lD79+6NjS1Cc8q4Dc0VRX6SmxskBclEILHiGCdW2tJ1TWdjfG1CJWmqr1EOqlSoq1b1LKAfCoXEfsWuGCBp7W0MKgCbVTKWLU0ZuMrKWJFlgjhCvBzE5QPyZJPRpuIDDIbKL5x0XnrcNZTVxXeQ56HrHdTG0RsxhS8jDTVfEtVlTERM8+zsG6V5uK9Zxgdjnlx8jpVFZqVChJriEXNM0Ebr/QeVJHRLQu8LbH2EJk2d4B3jCcNxjqObx0ynIzpFZ2WpmyXrimV/Ah27yL/IwXes7ykt9pHeUNdNYxOxsznI7wX6Dz10PvxxztaEQTNuKwKF5/TgjLh/1mmWBms4Hsu1RaMCyE+QBbHWdQh8i0RKSxeGagJkd3QjthAY+FZpmPJdlItX1eeap97z3QCN27OGQ7DYWrvubCh6A36SzLPI7xgbW2F1CVKELqHiRbWT/GNcF+1sVhryGKTF4Hilz72PjYHmk9/7nP8yR/9AVs73+DhRx5nfW3AbGyCu+kXyTntU3EuWJVFRtHJoovaEFI+PQIZ67urkHhjHcba1gq5Z/sefu3jO3zruW/xpS99nnvv2WF9YxOBwJiGmZ2zf3JAPyviO5VARlH0l6CUlOh2WnxKAbotnubbPtTWOZRQdMsOUkqeePJRprMh48mI9d46H/iZj/DHn/o9msZQ9kOC2mQ04vq1I2zjKTtDZuMj9vdexnvL3vUDTG0osy0mkwn72XXysmRr/Twr5QqvjTxSZ7z/fe+louHg+nXOnX8XW2vbTKpDrr7yPJ3OKu+7/9101ILXfqf4QBoCyIscnRVsrK5w8ewug9VVumXeVqfVmYrJXnEOxnfVLfKQ5a0CfJOCyFmeo/Qc5xxGeJyFLC/ZPnseISRaaYKhFC0Cv8g/iLhNsNBjXEIIgTEGGckNAfP3LXSTKKJt+Ngl+yusMe9te+3JGldREeBDgqGNTW2UUi3dU8kQK1BaRbZTfGYRI1dSBRaJi55NfA4iKo4sD3kIIpY4D/NILd1zuH8RW2viIcvy6PErts/sIp1alLZPLyxu65feonMLJaXznLLIqGdNKHqX59SVjY2BHONZKLVuqprhZMbZrbC/s6F0fvJTrIklZFpCRYTbBEilKbt9iqLAVhaBo5qdUM+ngMTUf8Y8AkRyJyPemObuInQPRLeR0JFJKN0Kjbjx0luLLx/RNocPeEU8VsuGiC8gWvwBnYmRhgRJpSBz/C7gqCJ2fUo5DoJeoXnPozvgZbRmJb1+keZTcGuNp5spHJ75ZMjq6qC1MlLnskRSSESQxtjICQ/pT0IIumWPj37ggzz60GWef+EFvvjlr/GVL36B/kqPdz30KFlH05iaXKYEr7BfYxxSCTIhKIsOAkVTBwWDtwg8eZ5RForjsQ0BeWNpqoamtgg0ZVbwvsfeR170+ZMv/Qmq49E64Mnzec3zz77I2Y9sLb/c9v3R/j9lIS8Uq1aKXhljNO024EyoPaRzSak73HfvvRwcHDEbDxn01/nAh97H01/5BsPRTWzmmdVHeOGp6wAtSQmHB0OOj8YUpcbUDZnKOD68QVNPGaxt8K1vfJF7H3ica4evcf78eab1lNeuvcDx8RG9osOrb7zIe5/8KC+/9Rq76wN6vQFXD/e4f+tMKNuWXK8fMAzwze8/z7zaZ3OtixCSweoaWR6otsReuVKqQOP04T1kWqFUEJahnHQM0koRbZNoLDQNSgbFHoScaPM/Um0s513b+jEINNkKOk/q+heUUkgeC+tGioVCSBnG0i88hTa4Hw0Y22L6gaCQaY1WAhWD1plW5LmO8SCF9Q6dZS3Dx+PR0XNPcJZP3oIPplA9m/LSd77HMOEnUQ4UZc7ly5fIdYiNqLh+nVsEo6V33Mw01958Ixbl8+RKIUU6Xys+FnIlzdRWVotQTt0HT0tKQoE5BE1jqY0jJNPHa7Ph2XsXevYFTywE7G0sZR+JQoHsUTus9VgLQmqUCJWIvXWx9IVC6IxO2f+h8+5O4x2tCJwTGCPQmkVxOdUS2+JWkRmUPIbEaEjueYJWhFjaJ02gsE+rJ9qJsXScNHzk3CdxFL/0IuUik5oHpc3BQ54pLlzYjEwO2XoXxMntnMOMZwybGS7rs1l2mB4fU6pwbVaGYmxGCDQJj4SmaWgag8MgPSBUtLQ162tbfOj9K7znscd46dXX+PyXvsYzz3yXk+MZ9126xMULJcH6tqF0bQzI+dqF8g6I0D/Ye1I7S6Uy8kwGhoizOOuoGkddR89BSnLd4T0PPcq5rQ0+97WvcbA/orGOLpAh8bY+ZSILJJ4GvGRcjyh0QSY7LBheDglkRRGqrPpAf/Xp3XlPYyxlr0/p4KHL93Pj+g1y5cjxrK2t8vLLz3BPeY7ZrCIrBEWeMTyeIQhB1ExLxmOLt55OmaOVohaO69feItPXObj5OvOZ5dylXcys5ta1io3NTVa3zvDaC88hMTz00OOUecnLLz3DcHhC9d6nOLN1kW7WpRet0CVUsx3TecVX/vizDEqBViEKVuQ5Sko8MnhjUiK1bOmWmdaACqWMEa1Mct4jZMTivaepa27t77F75jxaSXSWkRLOXMoZiNM55Q84H+iLCBlrAQW4SCkfA64hQ9e3BhUtzBNyOzLwDuNDjao8z1FCoHVoZJRlOUWRk+lQqyoU0dMt3BqWcMgrcc4jdfQ0kjvg45JNcQpHSO+NX9d5RqazKDs8QoftrfUUnS69otMGfBfrPj0Hz8r6BKV0zE2CvMyjVyHeBhKk0ULEcSgVEte8Na2CEsGeYjZPLDzQuaZX5HgX8mZMLKrY+ECFDglxAVJKeVKmCV0bnRXMxkMm0zVoDEjFymCTqRqRFT2k/DOWWYy3ICyenKSOUxJXssRP1R6CJUEjSLiPj6sl/BnLcfkFOBKdWlJQuQ3MeJZ4xGmHKKTSsaK7tgwFLOzZCOd4gssZX6gXS9a+E4im4Y2jGpPNGRY5w8khdj6howuKbknpDbLbY10aTNFnkAlm1ZSXX3yRC5cvsb2xG+3kUPbB+wylBL1+n/c8vsoD91/m+o3r3Lx+yJmd7fY68aEWe5kXsSidpdPNA14ZewNINEH/KrI8BGC9DfiuNYZ5NQ8JPDJW/JQlu1vneeC+h3H1yyHwJyRZkcUEsgABLZZOUDTPv3yFK9fe5K/+3C/SzQchWIwEoVFa0+Kz6cnGBX3p4lnevHaLfrfLoN8lz3I6mcWYMffdu8v3XlSMqqN4PsnqhkJrx3TcUFUW2xBT9hWb21tYW/Dma/tkWrO62WXWeEBztHfMyvou9z/0MPvXX+PG4VUs8Ny3vsrsZM7j736UyjgO37zJ97Pv8cb6Tba2dnnowoNkQoYy0kK08AXApKo5PjoM5Tx8EI5SRy+AWPZZiFjOQ7b3nEpHtHTfaMHiIYtB1qPjQ2zTtEJMChmCt7GjWDpGO9F9jJXFJ5wK2SEI/a1bK9tHKEqClIFNJkErRVmWZHlGlgXShYoF42Rac1KEuvxCtAJcICIz2y+8/WVn/k5WuCfUVfQLWRBsIUVeFHGHYEQ0xtLUNdaaGJNYzkYWSz6oCLBaTHwL8ZBUnoLFPste3u2fC4lEUMrQOzjBWFIJML6lWQshWFlfpd/vYJ3BWh+qIyRPygcGXyiHsUCkhCA2tRHYpqFqLJkP8J8TGVkxCO/sThrrR4x3tCIIjAfQGhJtMlnabRu5ZUMfWqsiVB9dlKEI30UvIe0vWFQ+hFbQBNe6nYEAeC9P9xOg1TOnRjs30u5LCip5FH7ZvBCCeT5gWE25tLtNp9ejORTURckjjzxBp9Pj5OiAGzdv8P39W9T1dVRTsXY0540XX+E3/ud/le2P78TTJFZFyAgOvxf0e5oH71/h/ks1KnLfIbBLXn3jdXQ5YGtrlzzT9LpdpMqxtmkpa3iPlmGhJcs03IOjaRpsi23KuMYVvV6HshNqFgkBWZFTmwg9LD2tpCzNdM4XfucLvP/yIzxw3yDeiwXhQ4BTeCSp3n2YG8YY8izj/O4ux8fHFJlCqQyJ5fhwytr2Tuh7PZxQFJosV9jGoDIBWrHR7cZ+wIJ63vD6G68iEfT6kt3NHVw+x/qMzTOXWM1zXn79Gc6fP8f5Cw/Q6a7x/Le/zqCzycnhiH/5W7/N6soaxjRcf+sG69u7fOBDH6PZnlN7Ac5hU3MW52mAg+EBQhqquaXbCUJJKR3ZaCBtsPDruoLYFnLBmAnzSiuNxaQy/ojoGfb6q2QRdkkMnrY8ByljOEhRH8s1CJmw9WitS0me52RaIfOcXCtyrZERf0+sIklM/IqZvin9ZdlAAtEK+7QUwxKLRtGpdRznl0uNpuIabDeKi0m2BwnCUin6vU5QDM6DkCgtWuHoAS+CggjOrjjtreUFSmka12CaGlzTQrhJ7qR7WRInp8bcOCZVhZKGsuMZ9DWX7lnnlZdOqAn5GTrLue/yJUpdoIBep0R7x82jOfW8ompCHpAxFmsXCgQRMvMDHOgxpkF5sMbhjMV7g7c+Niv6ycY7WhEURUGn08GaOPFFFNKw4Cq3L3Rp5rXPbeH8haqV0XxI7mAkq6QGNa2AbqGLOCSLoG37oW8rOIYJRqChiaTC41ZxFvn0R1rE6XKFZ6AF89mc2WxGb9BDCsX6YJMPfuDDXLh0nvF4yne+9wzf/ca3OTk5YTaacHJSYd0UaT1CpNdo4kUmemhKohIgCpTICSnpwXuQwlLVlt/+vd/nIx96hMv3PxStogyXciwID0priRBFDGxFFpUNZQucM+CzpUXlmc7nvHHzJg/cfwHhHXmexZjGQmsLoHE1uVT0uh3yQmKaBmgw1qKkRuAoiqAAlj036yzTqkH4ULKims2oTIVE45xlPDykLDxnd7Y5Gp3QuFkLDyml8LbGeMPKesl0MkcKTW9VU80q+itdvBgxmdSsrCi8P2ZcryHIOBlOuXjvuxDA6u4u919+nM7aGsO9q9x3/jxrG1ts75zjwsYOq90B+gdUgvRASc17P/wBnv7ChLkZ4vCh2xwEaMg7iC1JiyJvMW0XJ6yWkrzMsUZj/bztPOaBLAu5KAvjJs77BO0swS3Ohffb6ZTsbm+idUFexGBrpD0n+EiolHSYeGynraHWx5BL9NClb32yoDyn1jDEOkNLy0cIaDsI+6QIaAW9JzgWybBTSrK1vYl8UUVFkB62C31JWoEa5IWLxmJEwBCEBkC1i+U4okeVArZ+ae0vOShAYEYpZempId2uZntNUubrdMsMieTWW1MOpx6hBDvnt3jX5XvZ7nQ409NUCJrKcK2ZM51NGI1OmFVz6rpuqwEYY2LcSGAiIxHjaZwLvTq8oWlmGGPRf9ZaVToXgpKihRJSU/e4gUiTYFk+J6AocmuTsI8ThuVyz0H6B1G+ZJa0VRZjINm7JO7TSdMJF8Hj2Ci4FVjhW1jwzIimUrTaCefzDjJXk2eCg70bZNqFYnu9AcZ5VFEwUIrt7W3uuecetjfWmc5mHJ8c0xzPydfWF9fVsm6S5Z3MmGTBCJbZOUJoLt1znjNbXf7lv/gXrG9u8ODlB7jn0k4oX72kSqXUZFkeg1uheJnzNkxWa8mJ6fhR+Tx++SG08NimxgOdMksvIF5PsPhTxKfT7VKUnZCViuVz3/k6589s88CZS3TynGWsOCQ9zZnNJhjbkEnFxsYGJ9MJUnpuHh7z7Re+R391QNHp060ttw4qrK3pbxSEWEOAxcZDg3HgakenUjRzy7XhiNWNHmWueePVW5T5kEsX76PX6fHou97DxpltPvu5L7K60uPg4BY7omI6HtFgePihhxj0NunJ7AcqgfS2zm3u8hu/9jcY3jzktStPczweMm6mZK5Eeo1FMK+mrPSCVQ6p/3AQRDJTKJ2RF5reyhr7+0ehkU+cyllexKftWxgFiI2EggshY7mQTjdnd/cM3W4vQklhHqW4l5Oy7UMg4zROhIh0bC9pHb6W0OEXs3C5s1iCdGDJPosKZwFbhWq7EM8XYViWzpsqYqXz7JzbpdA5xplYcyg0dBlNJ2idtQZZKMvtYqOboNiaahab4oSk1LKXhzpd8VzLzMPlG9GZ555zOVvrOaurBWXmEJXDDR2uBr3WCcwmAStbK3z4w09xqV/QlZ5VKbg6PObNGzd49qXXuXbzgNl4QjUL7TOtdYGxl+Wh8peUkVLvmE8nCFeH8tdmxnw6xhnzZ7FnsUEIC05FKybiOMuYvQd8qsC5cD1biAbCS2zd1YTbLA6XfL8UhIr6od23tf4X/uxtSSVRy0TlkWDQFHxO9FBSl6fYDslbEE7gVzdZ25gjvUFLzfbuOsOJ59aNG1y8fAmtRHDZpabo9sk7XfJuh73x60ibegwEK38hZCExipZsLFIvgsQY2lhd49/99b/Eq0+9m8995Wv86ec+w/bGGrvnzjKcTdjIQvBNa8Wgu6j1kgRSYyqctdh4TBnrE+W64IkHn8D6ipPZiBu3brDSL/AxMRCCOpRtuWCBlDpkWCJ44+U9vvOV7/Mf/2/OhDLk0VSUPpbZ8o7peMrh8ISdjcAuuXjuLN1+l29/+hmO9/cpuh2cdRzcOmR1pc/qxg7Hk2PmoqGah7LcW9sdmlmDaRyHeIR1WOMZHc6R6wVKevJeh6PxOLBSXvk+w6//EUfHNSedLpcuXeTW/gxpPQ/e/zCH42Nevfoaj9z7Hrbv0EP51IhzWUhFrguKMuPWrX0yWbK61kFK2NraoFPkXHnlOqYKTeytswxW+mzv7HL23C7d7iq39g45PDyJQtXH4K0K70tGl7btSRBYYkIE6Gllrcf53TNoXb6toKdPBlIy0VvMZ7EGffKL05pIQhvi8RZuesuGlDGXIa3TdMi4Bl082FIF+gVM48KBWqFOmtaC9TNb7G5v8ebeNaqqoVQShOUbX/gi48mMRW5QuMbkWbTtNi2U3ZLaGHKdtYXnEiFkATpET8FDry+4fG8ff9Sg5g4zrBFkaK1QsWR7UWp2tlZ48KEHUALeevNN+rlmz3veGI45ORnzxotXOYxFEhePPNBoUzxSSIlWoRqAayqq6X6ILUUSjZDqlKfy4453tCJomtA7NDABFinq7aRqqTuLlxQM9CSwaSdZarbRCu7ga7YTPbmbSUEs2AgLDjsxy08kAS8Wx2HZFSVOsLY9pccLueSaerALjVU0Fev9FR5976McTy3ntvp885nXONzfZz4e0+13MS6UI3Cp12xkIaVmNEsPZ3HT7e+pMF3yGBY1h4RQdLsbPPJwj0sXL/HeJ57gTz/7Gb759Hd58tFH2FhZAyGQSseEuZCA4xEgJY0xIZ7gPU5YAlgWTUMhEeRopdne2OT69Vs89mBFpnuk4n0iKiapVSi/UYXs50E352hW4VyD1IoUYE53I4Wg3+milGI0HbPWW0FJzaA/oNdd5cGPPIhWjpdeukKmNWe3t+j2C9w1uL5/I/Z8DqrLOE9RqIAnZwVuXoPwHOyNAbh5dZ9O94RLF+9ldHzMbOzY3NpmNNzntSvf5dy5B5nXEw5mc/porrz4Mg/d8y6ge+d57Rxza7DOsn98wtH+zQj1ZJzZ2GF/7zDwxLWmLHt4JJmW1FMXqrl2V1g/ew9r62dY315DCUl9jZYJpCUsggYsZawue4aByqkyzc7OLllRBk/BLazxJPOd9bFEQzB2nF+aX9Fq8kIs6mG1HnvwvlMTG5HiB3hEorq3y3qhTEKmcGQyke7LR+UhUN5DLE3eBrBjkLkoOrz7vY+z/+kjHDZWVoXZcMbxwcGyjlvIkhRD8J6y28EDdVUxrqqFp7KwH1shnSqQyri/qxu8Ukg0WSe0khQKhNKcf+A8Tq0wunXE1cNXOOOBTCN6XTCOI2tpXDSSEuggFuBb6hERTuSCR+4szoZ/QqX3kQTYTzb+tRWBEOIC8N8Bu/Fa/6H3/u8JITaA/xG4F3gN+He890cizK6/B/wqMAX+A+/9N3/YOZTSKJlHmZ4E14Kume637U7UmvO0uGHYME78JPhhYVpEpeEjbhjmYkzkiidZOAdLCyT+kgymMEkWJ1/uU7DkupC0VWtheIkUkgcfup9Ob41bh69ixAp7+3vsrvaoqzlFpxPhmGXuN+hOTrm6srjpVuDL2z5btkrT9+n3sOCUyOl3M5564j3cf8897N/a59zZc+0xtQxdwVpFK4hVUGsak5QTeBEzimKfAYEnVxkffPIJvHVkKoPoP4Qr8ChcgMMby6yucd5RlllgV0XvyRHKgAQ5E4qQGevYWt3meHzMpJrRLbqsra3w3vc8xWNP3M+NWzd57rnn2dgcMFjvIzxsbKwznI6w0xG9bgZOcPbsGsdHY0ylAnuqtkjpULlGyg5NPQEhOZ7O2Xv5Oba21rFDyWR8gmumlN09yjLnxZe/w6OPvoePvO8jbHYHd/YGvOelq1f5zX/6Tzg+uIE1NWZ4TL+X47wjz3NWV/vcunXI2d1tovwL9E8inVSV3HzriNX+Nk3dYKQGmbX0TiFlbLSS+g8nSjVRyIQ4SaYFw+EwdKMrCLV8vA4Qz20WfLsmIsnCuWXm22JKtbZXihel3K3oNbc076VtXTTMWgfbEVgiinDg1qtINb3SyYKVnCDgNAfPPXAv75uPuXn9GvN5xXgyiz2XF7k/LCk7QfTil+xKZywqyyK1eskzWtIkyTPw3mMbH7qZ5WHe56s5Ms9wQuJUl3sfuEyns4UzNc9/51mGN29yUs2oxjNmkzl13YQqpPL0OYhec8g8pu3L4SPdNBBm/EK8ELKuf9Lxg0HMHz0M8H/w3j8KfBj4u0KIR4H/HPi09/5B4NPxb4BfAR6M//428A9+1AlCrGxhZbSWeZpcEXZZCN2gLXEW7x3YCGM4AzYGVawN/9zpf3gbXNpI4wpuo1uwlMKMJXb+aBVEG1+IFkVwOJKfG7MWRQy6SR2iSsFMAKlAajo753ngXQ9z+d6LPPbY43zxq9+jqR1z67l1eMR4MqKqKqLP3d6zFpCdepUJHoLTs8ktfbYcO1geAiEUmS7Y2d7m4Xc9SJbnrfWodY4qBxjr21aIzlrqytBUdWhAHpt42FOQVBhl3qPfXQMUztcYX2OJlD5CbR0RmTHgUFK3jUUypRNuEONBofn4aDrheHTASq9P1dRMmjn9XofNrQ12NjcZDofc3DvitVff5PsvvsqFc5fZ2D5LVkqwjtF4Djqyk6RmPKsZn5xQVZZqZrGNRynP5tY29axi7+qbdDOLqaYMBl0eec8HmFcG5SWbG7scHZzw7ae/g+h0Ywbvncd8MufKM9/luW99hVe+9yyz+RRkVHrCsTLoYauaqq5jMlWwLpOTK4UnU4pM5wxWtlhbWSPLMpTKSV27rFtY1WFZxI52HvJccObMNusbG9F4saGkROAWtUK5RUO8AJvqEoWeyKeWXLquJZujFdA+Cu9UM4dAjk95QSkTedlWEvjWyk4NdIgZwD7Q0uKUTtn2cW7EctZKKx5/8gkef+xxMp0HOeKXGFeyxaHC80lKIb2gsIjxMha9a1UUpzyeaI7iLKAg3ymRgwzT14yt4sU3T3jh1QPmRpPlZegolpesrKxwbVSxfzzm+HCIaUwMeidqbkwWTH2NEW0WuVCyVWLpobeU0Tah8A4GyI8Y/9oegff+OnA9/j4SQjwPnAf+GvALcbP/D/CnwP8xfv7f+SC1vyKEWBNCnI3HufMQ0bR3oXFgyN5lyfrwi7cXsX8fg7anFIdPVkq69tZ3jZ5hnKlSxlOGBZneuI9p6mGP6JEIYtndhQZPsUEfzYugreO1xOYVbcwgPUfjELMhddNQzSccj8dMJ2NyHfrQHhzs0zQTDg5GARaKlpb3Fp0psk4PKFkI3gUrKAz7Az5PykEsfR+T0kSGlpE7nnj/wscyE6HSqfMOaw3zasJwdMxgY4UcTaZU9AxCQhiETlciKUXvee7VK1hfcf+5i2ilybIAfchMxUCmpVNkkaFkKDLd0gWES4G+mvlsxo1r+4h7+6x01ziphpSDjO2dLR5++EGOhlP65e9TmRnHRyN0kbO5uUrWEazv9jk6nHJyVGOMp5k1oX5PnuOsQYuQFf3AfQ8w2NrhG1/7PLn2lEUXJwteev47lG+th4QoNePw4BZieoRcv4dmOn7bUlwSJWgpEc6hlUY6g3QGnA/Z0tYjCkGn32U0GrGxthqUvtYRWRQo4eO0C4bH1Zs3ODraZzg6YrCyjUC07VLbyrVxenS6OefObTHo9dm7dYzHM56Nmc/GrK9uBmOFGMlJNkPsYrds/C/fDyJ4dq1V6Vt7JW6XvPFglSsBpxq7hEUDUraBaBfXi0sKIx2YNP+X/oqBCY8/ZR3rLI/F5XykMCwFQJL1fNt7SrToujGhwWusT+SXIIdWocRnW9We6/sV1bzC+zGD1R4Xz/S4uLtCp1uS5QZvashD68veoEeuFBMvKDoFeaap54aqju060/3FeE6n0+NdD9/LwdUDDmfTkNUdr0u0IICIHtJyvs2PP/5/EiMQQtwLvBf4KrC7JNxvEKAjCErizaXdrsbPTikCIcTfJngM9AcdfIJzfDQPlGxhIuJkTZl3EOhZC0WR2D7JAqC1MtrgcnqawbRoawQt5wwsJ2kI/NtXQfx7MVmXvlqCZ1PWcqpvLvBYKfHGU03HjGdTPvfZL/Pys8+w3h9w/6XzOAzHJ0dMTiYoL0IZAR/SygVEep9iQRm9k5Pnl75P1yfxsW3hop7J4rulFRqtUBVKShjD8fEtnJlhOjlNNea5F58j6+Zsrm1ii1BLRmoBZHFfGRRBfFD7B0P+2T/7LZ76yJP85Z/7RTrZClmm6XQKvDfhvWmFwVM1ljzLI/QZqaPegVA0TSiEdu3ade47f461ssd1oTGmQkpFf3Wdja1tVtf6zKoZ3V6X8fGM+Rxkpun0MiZHFd3dAU4H5T2bWqraUAkoJIyqOeJoH+/CeQ/HY9bXc2bzBq1mbO5sc/PWHqbZI89L7v/5y/SK7G2KYFngCCnxUsWSN7Eej7U46bAuJCv2ugW3bp3gV4PC1jJrO08lq907SzWZ8t2nnyXPCkxdt1M9NSdJ5SY8nl6vw/lz23Q64Xmq6C28ceVVNtbXWV3ZjM1aZJRxbanDZIAvS/c0xYmI0bJN1FpeKTkrzSPS2ovC1TuxgHGT0WaDd2JJpaJFy8VI1UiTtvHEfRO1NF6HjTE9Ecuny1gdtL3qJQWUbkcg2nIWSivW1lbJs1uxhM1CBC1eZPjDGI8SjnvO9FhZ6VKWBVqB1EXICRGhqUxtarSUlGXO6kqf4Xweu6UFgW99YEkG4ykkOWZ5wdaZLe67716aYc1xNUdlGVI4jAv9jL2Osb/YwtO3Rt6PP/6NFYEQog/8FvCfee+Hy9rIe+/FQsr8WMN7/w+BfwiwtbPuhSyQOqedRXLJKk/Yv1iI9bhVO1+X5Fn4kRT5UjmIGMFKs4GUASziAdqOP3HGiFhHPdx/0gAiZgyL2Cg8uWsLpSXT/JXEbl7x3IMV5tbw3JW3eOOVt+h2+6ysreFwjI5HmNoGPrEzcdIInGugdlCnhvQQhP2dcgqWWty93Qa67e8lb6n1IgJOubbSQeG4efUah5mm2y0oy4L9/T2+9/3n+OhHP8y7Lr+LwWANJTVKumhF+hg+Du9ta3MV7wXf+eJz/Nx7PsjGygpKKYqyRMg4jYUIwtE4EjLk4+NyEQJ0JtS1l87z1vVrbO3uUBQFN27t8exLz1LmAzZ3dtDC0ut1cN6hvWB6Mmc8nZLngvWdDs28RmsR4xKWem5pKosQhisvPI9SGZNhaPpSdir8rKEsMt716KM88eQH+cKXv8De9VfpFj3mU8vU1neY2emxxqqcQmEc4EIDeu8FKxsbES6xaJ1hbIXF4YVDZ3LRgtCnZEtHbWybgKbzLEJtsb59nF9CwOp6jwtnt8nzjNBZNGTBCjxZXoS+xUSjy4V5HSzy1sYiZTi4tiSuaJuqCCUX25Kmf+w61grxWJ1UJIMqBKC9Sw1nQr2dkH+35NGIlP/Ast+BT3kIyw6+h5TYmCqdLsclln4Ja3lpTYREyVjm3Hqqqm6prMv31h4j/shy2Nnq0Ck1WawthKeVEd57nKmxWYPUOnjIEqyziMaDDu+o083Jsx6dfsna2oBOtxtEXh72KXsFvaqDkAqddSjLFWylENLHHuPBq3L/GjGCfyNFIITICErgH3nvfzt+fDNBPkKIs8Be/Pwt4MLS7vfEz37ICCpYCFqIJX4aJjiLSeqhpaoFy3tRWE4sTZRFaQlO/ZK8DqFgcaKF1+Bbvmj8fOEIxGcRbYpk/ieXLf29bEk4R+o54G2DGB8yHfR45aXX8M7QKbv0+z1wlmo2w1kZ2tVFzBfvsY1l0OuQdUuSsCaxdVrPIP0MHPTE8YdlysaSBlxcIKfvziJlRr/XCVi1tTTGcnw8QusZncmM2WTK4c09vvvgc3zsox/h0QceQ3cyHE1Y1LH/LDh6ecnG2hpmlq4jPLfuRhFZJpZep0DKSOmTi9Ii4V2Ff8aYkD2LwDrDm29dJZeas9u7NA0oNeajH3mKr3z5S7gG3rr2GnU1x9QV80mFrxWzXLCy3qXMAzY7rwyDVY03GcZ7ZqMaqRuscQxWSs7fc56i20dkilF1jNeeSxfvwVf7fOhn/gK4gpPhMc1ZSyZOt7Vs/a0QbUVag4mWXF6WaJXhncG5DKEkwit849i7dY3p9DjGT1L8yjOZzbixd5NpXZHZGlMbnHWoCB1450BL1tdXOXtmM1T3NLGGlAwYeKBBBkgwwHkeJ1xrtCwnjSVDPOD9C169xC8SzuJU9Mufpfv3QTXIJSHZGlvOR0tWttPZO9fSSBME5AmlzNvqARAsK2xUTLRGh0TG+EKUH0sR4dPmTywa6WljhNZadJ6hY5kMlozBhdxZxGBaVeJ9NFQ8ZtZQmYrBSo+W1BM947qxnL1nm26nQ6/fIVOK2jqmsznz2ZRypcdqd8BrL79BvtoP2ei1pZ43GGOwHnRWksnQ2937Ch/LaHj7E9newL8Za0gA/y3wvPf+v1z66neBfx/4L+LP31n6/D8VQvxj4EPAyQ+NDxAglIB3RiGfLH6xMLiTCwoLL6DNpkxC+FTc0i+C/8l1SL/fhq0tEETRKp7k0iamEBHDDHJ42SSK+yfOc+yutXCtA2ZkPTRk6Nrw4IUd1roZeVHQ6ZRondFYMGYei1OJtrJpbS06up0h0SgVnUs3k5RCgoTSZ0lpOIRImb53CjAv3Xt6bioDmSNcHbq3xYJj02nNxfMP8Muf+CW6/QF91aOZz6GzAn4JyvNB6GVZRq8z4GR+QtMYPJZOkbO51uPK62/y3iefoNst0ZnGOYPSGYis1eKJaT6vZqjYRct4jXSeTIaaN48//AjPv/gc3c2SvOgwn4w4Ptjnrbeu4y0UmULniqbyzKaGupF0exn9lU4Mgjv6haCaN1SzhizT5LmmtlOYG7pKMh95nv/uF5hNJshM8/ILz7CyfhGUXliDS0+wFakClJAhi9jVCARlpxNbEwarOMPjvaSxDXtX9zg5mQSFKiSVtWCmnBzeYjpSzGdTRNmhMSGJSshgnQsp2Nne4OzZ7YB6WkdtKvb2Dzl/5hzLQcVUjVOqQI8mdRRrZaCL3kWShHIpZrl0l36xpGSa7xGm8cRaQ+ElklI/g6fnIvPHRyUUKddpJkYl4mMxREv0iqIAN84FWAvodbooGdtXpg6F3p62dU69mUReXQjyEJdRKL0oOrdQjEl1xGuLpIbh2FDVBkROnitmozm1MTzwUEmRu9CLOYslvnWGKgVSCw5PxkgBXS1585XrqFxz4b6HuLCzy623DlBlGRS0saGcRIxjWNO0MsA7G3oa+9jn4Scc/yYewceAvwk8I4T4dvzs/0xQAL8phPhfA68D/0787vcI1NGXCfTRv/WjTuB8qHkvAnWA9OIS5Stp+kD/ui1WHqNjPk4uL0SQuuHLAD24JYEuiAKbRRwVf2r1LqyAwGbyzscqgRaJOkWha60M50/FJFrbIR7aOUEjBKaq2Fhf5cH7LyKzjOm84treEc6FDmneGFRRREM/NKJvHEzHI6wbo2SB9yVCLGcXw2nhn86fsfAWFlb5bT7Oqc+8dxgXBHmoFhm8Ca9AS82jDz3Ox3/pE0iVE+qkRmtRKDxNDBQGi1ZrTbcsGAraBLI8K/jVX/gFbu7tMeitcTQcsb5WIIWnk+fkWcas8S3LSAhPXc+DBa01WmmE80gZrKqj4YjHH3mU8TMV/X4PU8+o50dcuHie8bTirVu3aGrT1tyRGowp6BQZs2mDqRuczUPBvFjrXmrJfDbF2TllZw0lHTf2XqWaWTY3zjCdH9OrexhTMXeGXCa20+lR1w1OQNHvYW0RFrpShIxpF0p24AM05Hrk5SB2CgPvBbaJrR6FxDY1nU5Jr9ulmQ9wLmTFCq24cPEcuzubKAGNMRycHHB464D5tObM9tl27nu/EIrB2BEL3D8ZOtBa+2EKhs8DeUFEeFYs2VUx4NzefrC+UnZyUgHJtqcV1sG4sM6CUiFojOfam9e4fnCTyXiGtS4ylyROC/r9DvPxhEG/R9V4PviB96BkjpCq7VUtl4R9ONttCAPxXlpPJUJa6ZlEiKq1+pYzpz3Uc8fLL40oypzVtZzBuQFTXCh610CuLdbWGKdDLw0pqSZzGuYU3S5lp8+FM1uMJhalk+fuQ2mRaIHKxGISEik8dT3ERW+4aipcE0gH5qcJDXnvv0D7KN82Pn6H7T3wd3+Sc+S6JFN562otEl1EO5GCaIsBXljK7PXxJcWJZqOQj5ZB0uQL3m60OJZYRvG64y+LvxMvu+0mBAhcyBLGLxKgWyvDtdmQy0MAqjFM9mbMygFl2WV1dZXh4TFCCuaTKWUnC8LSOpxViFgNtGlqxKBkXNXMZjO0bsiyBiVzIFtiDqRkvOUFnRreSxbwUNreAQ2nFYhEYlBSorWk8VnsSGiR3ofMxqxYlOlebibkPVqkiR3+llKgMhmsF5fiF5p+b43yQij8lWUlU6HZP5mwvbERql0Snmmi1BmTYCJBJhRSGIQXWGu4dfMALQQPnX+Ar/WfYzqeMB0dMD855J57dhhNpoyaCc45ZnNHnguEbCh0Fpqne4d1sLpR0jQ5ZaHp9EuaukEqwcHBMYPVnCwDZxUqL7l28y3ms4rhwQHH544ZrJ5Z2BFL71162NxcZ+dMn6PjE5pqGize2PvBC49xhvl0jreGLNdoHRLvvLMI5ej3u5RFQZF16PYd83qO8zXK52it2LlwlnO724SlYzG15fWXX2dlsIbOYu2mpe5iYTm4hVJIhr8PdNVTFPoEt7hQ4EF4F+IGculuW3c9rbWgwAONU7YxsnbORGPLJeNLLgLW4KlmNfVkhM5yGgSi20FOphzuHdKT62R4tDBUVoS2n84hRCpXsST4FzextKgXXkz629nQAU3I4Fmlltp+oRkWm0ePwBlB3UjGc8jyPvNqzGwuqayg6wNclU7UX+2z2VtFOIlFUNU+ZAU7RzOvmc7m+CbEruYxgBwUW1iDwlZUowpbhIb13tjb7u0nG+/ozGLrLfO6aq3+1AkozdBWkwfTJAgWINnfrZCLTIiWQ0fEQlvBuEg0SQyLMDEjFNHWxw8fGxuuY5HkliowRoqrDF2WQpDJ4YQkRcxu5/g6K+hkisl4yHgyZM9MGE4bMgWjyZx6GLJfi6KDdyb8lJLpaIh1huloEuqPe0HTNBjZhOAhIhRtE46gGDxhNls82dJ13E4GFCwayS+eoRCCXEty5ZlNZ+isCNnGEBRB27A8EsOXuM7CLywxEAjp0Z0uSmtMig7KYGVpXQKes9ubfOJnnqJb9skyicwS3LLwqprG4kVw4bXwCC1wAqxpcKZhb38fU9fkWrOxuY1UjutvvM7VgzdpnKEaG4SKHbnyHNeEPgxKEOijeGpj0FmOl9DUFbO5waKwtcUfOta3SlZWNJPjfapjgy9hWlVs9jfCTFxipURzm3vuOcPjT72fz37uD7l+/U0ubmwgRWCOWTMAF3pCCGExNgJ8sXuX85ZMWPJMcHR8k/loGDtTOYq8oNPNcVZS5ip6zT56sYH6qvOcZl5hY65NkYcmN01TMxwN6XQsZdlFedUKfKJH6+OasD7x92VreAUY1EcWW1o/yUhbyE6JWBAlYDEvogJaLpTnFyuMzZ11jJ2H48YS2GJ1wMULZ8l06LwmtELrjFyrxXqTqTWmb4/ZVtBaUlbpN0m47elkxtW3bnE0HLfmpvOeha2+UHpJ2hgrMFaRGcHByZi80MzrMG9D/yBByB8KXdCsKBC2wglB40AoibeO0cExg615MAyURhPIBFJJym6XMvZ0UG3uSXgXxtQ0xjCZzvhJxztaEThnqJsqCuIlDnAcKdmpZeqQ0JlUwCAFjUW0JFNySDJAll3TsF8qi7vMMJZL26RG02l7YFFjPppNoj1fUBeyvZ5lVzkeO1fMA+JCNa+4eWPC4bjiTDejKHus4hg1jp5oOJkdk3cKRvOKdT9DipLJZIbDoVW+gBdid6jGNEhZoVUWPYUUQE5Mo/Rc05NL95wtXWHyaixagVaCsugyN3NMNQudpjohuNnYGoenUGpxrIgJOxwqQlFSSjrdLqooScl36U2mxaVkxgP33g8ChtMJKsuXStYE4dIYE2vf5whvsE7gvWQ2m2Ntg7aSujZkSnLmnvN88AOP8ak/+gxX3rjOaDQJvXqtJ8sl41FFp5dRVbI1AKx11JVlMp+Ql5r+isJZwf71CZ2OQpYwH2qEySiKjJX+Kq4WvPC97/O+xz7IzvpW+3w9oTpmoTKckrxx/Sb7N6/zrgceoNAZ83mFmc+ojaXrF02BnDMxxUVjbUjiOzk6Yv/mTZQQFIVGxy5mWZaxs7vLwd4JzhiGkxHz2ZzNrfXgGUvFaDhmNq944fkrjIYz5hOPcDWToyPGqzNWtzY4d6aDTzEB307rBWyiZHTuFgFiF6nXeEd646kQI8k7jkrJEzNik4225BUEr54wjwkLQwrByvo6g7W11sBIfngwNIIFnxwREa81KIHQJ8O0q9W386f1WCIKIDyBnCIlk9GU0fGI4cmEqq7ba/WnjtP+ET6XMhS5i+cw1i95WWmNhYoJAk9TzekLx2RegeggpSIvM0QeSn0LlTGfN+iuRorQjKqqKqwHqQRVXTEcTZnNa4bHJ9y8cYAxhnpu+EnHO1oRyFhwSfiloBHJGl+M0/bm0qS9g4eU9j5Vlbr9zr/tL3lK8y+dpLUsxPJKWXxNgCzSSRfTd2k/ABxzDxMn+dozBzSzGdZUvLpScPmhB7n4xCMM8JzcuoqeDpk1oU9A3ss4oea5V69w9vIO585sU8puuF6l2pR0ay3OG1ImnhIqrrnkA9m4uBLunwLMy89CImWDVBJdlHzsfe9hY3eDV159g1euvMroZMRkPI5Z4A6nAoMj+V4u9UkWIQCTZyWbqwNudvP4LNI5E0wUrkPrAh+tIRESRhbvy0NVT/DW0O90qZpZ8C5csFitF3gb5lCehcX9wSceJyu7fP97VziejCh6BU1Vh/Z/DprawckcA2SZprcSymEoobAWjJF0+xLb5DhnODqp0dmExs4oXcZkanEWLj70HoyCQdE9FYY33nJzesyXv/FNvvH5z3DpwgV2d+5hdHzAbHIckvWMDfi383iZYa2J1qRsXdZMCoSWbU8AqWTo4YukqmuyPAQ5RydzGlOxbiN0KiTH+8esrnRQ0iF6io3VAUoLpPTMZob5fBYho/CkLYDzISG+lboh5hYEY8gSTms0bJ485RgT8AKXEj7TyvE+rgbi+w24S/Lo0xpu03/EQhHhEvbqW1hmgdsk5yJMltDWU4bEzXj5LarAAqrxy4JACOaTivm8pprXOJOSyhxtYmQ7D5NSic/LOKz2IFRQ0DHuYGOuiAd0psmUxuCwSuJE8EqlgF6vx/BkGLcr6Ax6FIMOeR4gvzLTzKZzmrph7809XrxyjXkd1vd8NkNn/3oi/R2tCDjlRp623JOgEum7OFES+COWbP07HSGMdiqShL1rJ+hp3Z/IbeDaiZdK2DqXJvRpSyMEkn1sGp8qP/pAi4vbOhvYQ0oVrG1tcnJrSDWfMRxZ3rhyzJMfPcfXv32DyytrdIqauqrodnOcs+ispCwEn/rMF6GuuXDhLPffdy9nzuyw2u8HL8GHEtIJlgp3Z7DWIFAhK10qAk9FEHobJLgoJSUJpCxDKWJV8sbVt5CdjPsfupf3v/9JDo9GNKMgqA5HI15+/VkevHie9ZU1Mp2HwJ8XaB0WVJFpzp/Z4mB/IwgUJ2IhIQl+4Xwnr0UKQZllSypUtDTT+XzOyfiQlZU1vHOYOhSTC3VYAiQisUwmYw4Pj3nyoQd4+OF38dJrL2NtFbJYnUcpEZJ5IhwiSkVZ5EityITDGB/q/6ucEQ2Z1tR1zdHhlM3tDnlZkpea+URx7a1n+fyf/hGXfv1vIoVgUs04Hg+5uX+TL3z18/zub/1zCq3orzyIkllIVjMGqVRsBRpKo2RSxcRBiUydsnxo+C5VeJ9ay1iiOszjN157nW5eMFjdROsQIg3NhAAh6K12WN8e0FQVsgo9dKs6tEdsGoMsVWQNhfyAlLHaemwpYiwiXcAFby9BOgIZckFcFORRAbUQRvKQfFxRwZRujbBgmvg2VuGdxUW4SYrEMBJxPcX1Ha+pVQDxh4guZChL45ZWeZpiC2NHpLUZb9ku0dVThnMyoJaA5LCGY6Jjyjew1jEcTRGNCQmCgGyrvwbWX1GUDMcTho2jMaBVRC+cp1NIShq8s+yePxvnRqCxqphQWzcVtjGh1lRjyQpFURatd/OTjne4IliAPK04FmFyLNcySS/ER2ErSDSwhQ0OPnr86eXFxtGtJR8/9y4EX5bYEh4WFDLnYsmhgI1KQqMWSXC9fdTsUqaysIFlExqtiED3kgEkUToDHwJu22ubPPmzj/P0V59mNhphrWVQ9vj5d1/knt1dXn32G5i6F2r+KEllGgbddT72wQ9RrnT5r//+f89n/uiP6Q967Jw5z/33X+SBy5c4c/Yc589u0+8OyLPylAvkfMCJnTcI0QQ+OTr+W7KaUAuvQWh2ts6Qqy6vvPQGz8xeQHjLpd2HwDvqquJf/fM/pDbHvPepJ3j/e97LxXP3kOlOgAS8pKpqjkYHnD23Hr0FA2hEex4P2DbmoISPiTrtpGg9sO2NNXSRc3xywEpvFS1B5XkEwYLl6AVU0zHXr90kywu6ZR5ItR6cFORFoODZxpFlirKjqJ3D1BYlQlBVaUVTzxG6Q6Y11nmkUHR6GXVjMY1HSShKz8nxPnNT8erBmzz95a9y/doNZtMxzz/zLK+8coXpeMJqvx9af/pQvTWxaZz1GAs6Cr7GOqqmoihC4x9H1JkxauldiJVIJcF76umUXCpqY+gIIkwYTSMhGA9rpvNbKCTDo0kr3KSCbldHKoInZD2Ghx0UsMAL28Y8hE84vA9EBoKQl3iEE62lHAyklP8g2vfqEmMvvcxTjEffKoZTHnRYiAgcPvVVxreMQhHXaejB5pDCIZcnvGdRYXhJWLZrPd6OUrEbXLuvb6+KJQVx6gp9CF00xuGVwzZNiMRFtqJAgAqwo4fwvmxNmTkQJhhBUuFtg/YGTBMri6b8hXD/QmnKIo9sNxGS06wPtabiHHm7wfujxztaEdRNzfW9WwghqZrwcLxQsR5HePmyxfk9jpgU40OSC97jY2AmCGaPEgFmsHGWaq1DfRc8QimsDU24QxE0AnWTEKzTWoX6QiJYWaF/qMQYG5t0axprYxp6cAnzyEKp6jpUyEiN4r1HK433IcNQlz12ts9y/tL9vPTc91HS4qRgPJrw+OOXuPbmC4yuXcU3NZjAjJmLKfPplJWtFTq9VWa14/itPV5//Rrf+NrXKfOMtbVVts+e5cMffD/vff8T5EVBt1Oy2huQ6RypdbBU4rS2tkJQh+flZYzPaLQQFErRKxRaSMpOxtmdM4zHFdfefI15VdHUNf2yw7sef4iv/Omn+dTv/B5f+fy3+Lv/2/+Ii5fOM54PWe+tIYTn9TeucuXKizz6yJM88sAlSpUhSElnHtHCUx4hVAiMeRa1ZAjUW+cFqyur9Ho9Dk+OY7UQSd3UyJgM5FSGtXOcd+zt3QoZ2iJAQaEVoCaTkBWKpjZUs9BHtpASoaGaGhCOrFCcDI9RMmOlU1KNDZNRRVZIup2GotRI1WC84E+/+Pt8/8VvU9o+81nIUn79tTc4ORjS7RVsbq5ST8d0Ox10VtAkIyd6MTiLkgEXNnZOLm2YKzb0gA7YenhrLpZj9kBeFDgEjakRRM9TOJDBMl9d79LUFWWpWRtowGK8Yt6EQnY6z1LcNrBmXNivpVYnmmg8NrhTjWVaWCcmZiEWwG36GXoj+xhcju8SwLtoB6R9Y/wtWtE+rekkIEQU4JH9lzwG4tmEUHG9JYrrMnREa0Sy/CcenWU0JmWHJ+vfLaBJsXy3iz0z4ZHCUWKR3gUjUPg2rzRdn4DYm7rFDhACpFLMphWNykHnKB36PjvvUBJwjrpu6EbISyvRNhYqe13KTMdig/8WSkz8TzmM8QyHFUWWoYqCTCpOjEPZwLXNy4IsizipDIkfwWs97TIGDnZMElEqussWKULVy6YJJV2V1hhrUCLw2oWAqgkceC3VqSbqNhaAU1JRqwYJZDpDGIOSC/w0UyGYmSZopnRUBC5es6Spa7x3fOYLr/LmNc3RfomZ7FNkNd9+/hWaV4fs35iQxaYkkpCYIxDggztZlAWdbg8pJU1VMZvOGc1nZMJTz2eMLz3GmY1tDofH/KN//Ptgxjz0yIO866EHObO5RlZ0KItuYP+4ICCdCwpBSoV1weLNtUM0HickGZpeD7plSUj886hMc/7iRT70vo/w/EvPUVcGM6+Zz+f809/9Y37xY0/x8AMP8qsf/ziH732C+cwgVU6b9ObBB5SeYOUFRZqVnZZqKCIs4Wwo0iWEJNcddjY0k8kJ0npef+11zpzfod9bQ8oCxxiEROY5RaHw1uNiQ/Fm3mCloJfB6kbBeGgx45rhsKYsMoYnc/I8wCxSBPbQPIeyp6lnNUVeUs8bbAVO1fR6Bc1syK3rjjOr9+LIqCcjpNB4oZG6YHtzE4TANM1S4hUYawNU5Fwbx7908RJlMUB86/lQe16HHIJEkRZCRoVAS212sVd0KhUqhUcJT6/fY14pVnuKupriZUbuFLN6htIq1u+PhlGCL3Gx6XyEQQQt9TR5DIjYG6PFXxbUaSlkm+XvPaFellz0nxYuClUhWscg+QHLMFD7bYowBzw4oEE2egltQDp4KVJplFZh/S0byn5xqKU/AUGWKeazBIHRaonghCbGFCyXpffek+HIpSOTPjSTyhTGqyUmYyqQIWLhuMAWms4Nq+sKnWUMVgY41WX7/C7rO6tsldtI4zg6PEHnmszm5HlOUXYoO+XCk4nPVOApy85PIGXDeEcrAqUU3W43JPWUHQrvqRqDshqdZxRFHkq7qgC3AAvGDgtIRwmJJWDKmdR4ITCmQUoR6uMLibOWTGdIpcmkRKngZhkP3rngESjdloYV1iBFoK0liyjTeWicIZNA82idIbMMbS14F5uTB9dT6pja7yzOe+47U/HA2RVU9mG+8+2vc/36Ps+8cJWVdclKr0TNc/Aho9gaixLBYu+UJYOVbmjyHukNxgYXVec5RVkgpEArxc7mFt7An376c/zhH/4J/f4K5y9c4MEHLvLoYw9z6cI5djc3KTvdoGRS1Uhv6ZYanWXMq4bKOIra4nLZNglxXqERrHU7HK702Tx3juHRkMbY0FvCSeaTBolgbbDK6mCVummQUiOWktsCq2TBYtI6oyi7MagIIf/A09hQUycuR6RQFEWHlZUO3c46h+MDTqZzhDd45xBehPLNeRkarUgB8XMENMYzm4TmL1JJlHaUA4HIOswmNaaxrG52mR5VDI9nKCnIMonKBN1Bh+HxjF5RUOQZXjiE9kgFpTSMxiPOXzrP0ck4JOYpHYSBtVhrAInDM5tNMWYD7UL4vnbBsCjzHJVKQdDKwGBpRqHqvEN4ibMGZ5rWW/BOxOzgxb7OBwvUWI+3gWUSKhwHem9r80bvNsS03EKWRqszxMnEwjpPDlsMLjufMmGjNwCLC4/H96l/x21QLK3yCB4I3seS0XGNe48TiYm3gINT+8BkZSfDUEmJTfTWCFWl55hwLEHo7CZFYv1F0CluEy57WaMkZRC8FotAZ7FjWCwn7RPtPRS9QApBrhV5kXPmzBqZ1vTW11ntrfKBn32Ksswoii5Frsk6JUdHwwhZ5XRKT5HlCBc9da3wlQnrzxocgrJ/54ZIP2y8oxWBlJDloflzrnOUtZSZxAmPUkEoKaXRKos8dqK1GF9U1AZKKbAOKWXg1ivVBqGD6xUmlpQ6INVSxbIGEmUsCItWOtQdQcZqhOE8UmeoyFaRmUY5WrjKOhcnhEZKHRJwQuAgYPMuufdhoZ3sv8VgMOXjP/dLPPbEeX7vX/0RZ7dWyfUBk/EoFOtSAeoSNsBjTWMCvW5lhbX1DYYnR7hYxdD4JNAU1plAJ9WS3qDP+sYGB0cnXLt6k5vXbvKdb3yLLNOsbW7wwIOXePzdj/LYw5c5f/Ycq4NBiE1kCln2mc8riqKDOalwnRwj8xC8xSGEpptprBTkvVX0uKY2DQBnz2+jdAhEKxG48k4RLb6A6vu2j3GCHUJ/YS0DBp7S/4PEsTQ2EgOTdSgC1/rCxfM8UN7PlSuv0ZwcMSQkYykPiCLmN4l2X289pnbMZEN/kDEbW5TUnBxUbO526ZSK4bAmE5o891TzGiFCa85OniF8zuZKgXMCW0uccgzEenzvhqqeceNgP1TTROKcwnlL1VTUxnIymVMaQAX3PsxbDSZUFFWxQ1go/x0hk2hBpzwua0NfBecczlg8EhuTHpMZa6PAlUojtGY+OkS40Nd3GfZIBeNIjB5YImOIGISPiywlby4DM/E9pXwaT8rb8QECSl5DkpFECKlFYRy4EHhOwV4nPIhY8oFgBWu/KA3RLvk2iOxjcBW8ta1gTyqpTTpl2cIXyGw5ITNcc1QdC8Ww/G1kO4k8w6sSiiIYqFJi0rsLyz6cXUqKbpcN4XjosUfoFgWzec3hwT6z2ZTZdM5wNEZ4T9nrgsxZ7w9a0klS0lKFiqrWGBDQGfQoux36g8GPkKxvH+9oRSBiXRqBpOiU2OksMGBsoDJKFTBA1Wr98MBlpBom11opDdJhTbAQpFI4pxHeo6TGKxlKsmmFb0RQGLHkpdIab4PCCFCOCniki3hm3NZ5Q6CMxYSqRMETCqGCsqlNg7QCqSyNaRAItJJ4L1BFB98EWEApxZndbZ566r2c7O9zdHSVajIK12c91jfUtgExYzSZ4pxhMOgwWFlFKUFZZJwoyehkTJbl5Hmo05NlGVJLur0OOi/Isow8z1lZ6VE1NaPhFGsPmE8nfOebz1KUOVubG1y8/zIfeO8j7Jy9F2RBZcZ08y579ogi7yBlHjJCXbJQw2oulCJXMnhfQtDp9hZwgsjwSJS04ENew7WDW3z1W9/kY+97LzsbuyQXWhBqA7nYuMXF1ovW1RiTwITILIkWmHOezc1NVlZXGJ5MuHZjCMIjpQpxIhIE5lA+eAfOOJqZZy4FUiuGhzOsceSlwBuwxmNFGYX1HJVB3pHM5zXGnuCkZzKsWNvoUM8U1cSy0h3gGsnJaE7WlQwazWw4B2+YzSaMDvd56+Y+uZ2yvroCqqCqZhRFHmBD40IyvAi4u7UhY9Z74tyXIW4gFcZFT9i5lsDgoleXIDUR2VTVfIbznqOjMc55dN4PmeMpSUwmyumibMRCmkW4hhQTWIjX9m+f4B0fjyeizxM2TXTQFGRNxIDUdyTBO8mZCHWGPNLb2IM5FrtrCbqLs7cluL1b0Hdvi1fAwksJLDRoIwwxjriw/Jd+Jqgo3VvrXdDGKaQUFJlGC4VRTaT2huxfpyyubphOR/9f7v4s1pYsze/DfmuIae99xjvkHXKoeWJX9UCqWy3aNCmbD6YoEgYoS7ZgSLYhvQqwBUvykx9swH6yCcuQQIiSKMMAZUm05QfLlkyySTbZxW52d3VNmZWZVTnd+dwz7iki1uSHb63Y+1Z3s6tI2Uh0FLLOuefsEzt2xFrf8P/+3/9jfXPF3//Vv8foIipGVus1i6MFB+2MWdvQti2Lg0N0VRO93DvnfGYpimx1lddA3XWc3rlD17WZBfjTHZ96R2CqGhK5tXxAVTXKbWVDa4tREm3ve3Cdh1EI9zxrmqNIJklEbQw6QzVKSSFZx5C/TxPTRyIaTcRMhVOlclejDnlfWLTOBTkk+i8UCG1KwUgxBIdzjkQS6qHPka+qsSpnGdYQIwwucnwy4xe+8VWenp3xd//mIzabdcZ8JVoKIUGt2a63jIPj4GDOvYcPWa+O8dsVi4Nzfth/KPLOVcVU7FKGu6/d5sHrr1O1NU1VcXp8zNoNpPSco8UMVVmG7Zbtpuc7j37A048+Ri23/Mk/K3r1bhiZ1ZYxek6bjkYnyQiCGJlohB01t3ClEm4QmKKujeD6+7RUpSktSGEI/H/+87/DnfqQu3/iNk+vLqgIHM4P0EZN93VKx6PAOJOEBmLYh9Hx6OlT7j04QesKXdeEZGgWB1RNTdPImopemCtaS/HfWEvT1fTbkRic1BBIrK4GtFYcntR4d8O2F5XH0YEePHXXZQ0kTdsZGlux8ZGbZc8wjMxnCw4Xc95+50d0h3MWbQ0qUtUdH3z/RygVWK826GioQ+LDH7zN577yFYxt8c4JB13pXPwOBJ8wdc3xySFdK5OvtDZ4F7i8WUlGmCPtlEKeKJeZP1nioZ3PsKbl8aNzrpfX3Ls/l/GVVVHbzIauJF+aTByQJrOInFuieikki66PKojOjxldpudcMnbRnyI7jZh7DrLKLnE6r0q5aG2qV/oJdgXmolq0Z9yVoPLConmV6y9OZmcv0vSz7Ay0RPF7S236u3Lu3V/me5XEUZHJVt5HTKVzFiIfMsZIUhJ8dt2MpBLz2YyqEvlw7wJ1O6eqq2zHDG1t8MHTp4g1Gj84XAxkagymtihlcvNdhvz2KVE/4fGpdgQxhYwVKmkhz9jaoNQkfCYRelbhTBKwqCw0xSRJoVFWaG3lGUbStPA8EecdgYTzHhMUVRLcXsUclWp5L7LYl86LT06vc2u9NPoUoUOtJDNIWdtHGwvZ+Gmt8mi6ktFolFEkFYnOsWg67GJO07T8zuKE4eZ8LyVNeB9pahHWatqO0+NjvvCZt3h5ecGLp89Z92uqqgYStjKMThgHrbWcHi14/fU3uHV6h0/aD5g3Hb/45bd494cf8/LJUwYvBfJIwlRFild0fFobWPlAXQsGOq8VlYpUucistIGqpa1qYlODVjgfIEaaLCkRU8QQsrpk6VmAxbzhzp1jvJeF/g+++Tu8+857/Cv/43+OlKUNytzihAwfGQZPTHoSFUNJheHZs6c8ejTn6OCQUnOoZzPq+YLZwSxnfQqdRH5AaU0MEdc73Bgyriv32vWKqjEsrzxRBcLoMcbith6/Bbe95vXPHZFiYlwnrq8c2600l7lxRHeBxeERlTa46yXtwQGffPIIbEO/2fDanUNWw8CzZ+fcPlpweO8hqrbZ0Y2oDAcUUTSjNYuDQw6PjifDLQmxkwJvJM+RFgOaMjtHEjWh1Jp6TltbjNE7yZSSLcvdFbv3yvxJMjwXKVP9UvqxjvkM3xXcniRdx4qIji4b65A7zX2etJWo9ch2uaQ2hkoHcBt0HAhbxxrN8bzh1uED1vY2z5zOjmjndCbS5J6jUEoyu/2A/tVv9jKZDC1KMi8Neq/AQGl33n3WUEw7GGyC0GJk3PbY1ORxq9JQaaoKo0WO5eD4hGN7CxUFsUgKtv0WU7XURmjURkufiAmOYfQyya+CutbouqJuDVWeqSw+T2Wz94fMEezyMIXNcwKsMQwlbVNyE3RhEABTU0uJR/TuXH3wuHWPtoZtP5JCoK4tERj7UYaBeI/VMkIONI2tJRpV7KjVRTMkT5MSJ1HeRvB/H0QWdmZlWpXRBm1sjsrS9N8U3eYIyDvHdr0mJEc/QiJQV5U4vIw26ZxOK2C1XvGjJ89om477D17HNg2r1Rr1UotyYRgwuSh4c7OWaUpG2CG+jdhmgdKWk9O73Osjbut48OAOy6sbLpc3DMO72AqUsdSVRCgpKrCWxspUMWUstjJSE1GKxgrVVjWNRHGZglhVFjc6XHAENVKZJmP+GVazlnZ2kB0BdG3DzcWSEGA+74CKlNxUJ/A+sN6s+fDD97n12mscdIdoFWibisuLxMvrJavtiHdblIaHt29zcHzE8dEJ1lYkNqDA2J3cgXeloSvz3lPC1mBrzep6hCS4s2ckhShMjkrz/OmKg6OGN996yLpPqMtLrs63bPueoTbEzYZZUzH0gX69xdYGFQYMnh+8/UNA8da9E5r5HG0bfDDoGNhuR5wPrDYrUeUMMc+1zZlskVlmF3xUdQXBTVF5ATC0zpRQNDoljLKi5Jl17AvNIubnNekxFqNOml4j3xUKaS6sKqksSHS9+zkkkr+mWr6LCYHgPGFWs+kdl6uB+UHHoYWrZzfMO0vXVig/omxkc36DszVz09JrqI6OsKYTZ52Qpor84XbaREzAfwm+SGnfFOwF+vvf7dratNlBTtNnyL+byCiTIyqOIL/1VH9iWqflKcQYpBZZ1cxm7STGl0jTfVcJdKs5PLT0W0hDRjViZBhHvPMkN+D7HjcOTLAfhX78ag3jJzk+1Y5AhpTLwpIRbogKY9oVAApeKAtWZVErJn0hCWLUlOK6vJGClw3vXMxRw85Ah6SkIKd3JaLJcKeE0nFiaUjHokgUEBwpKNEDiYLl1daStJmmBqlp3eoc1UZMMgQ3SpOIGvneO9/h7OYFddfS1DXbfj25ROFDp1w4lJ6Jt167Sx803YGmnrVEIuv1NWdnz3FrN8kT1LWl6VqOT05p5gckDfP5AW5d5CFk5vDxrWO6+QHH7jWePHtJ2lwTosdoco1hwf2HD3Cbke7okGaxoG7aHAlqZlZgu7prqGupz2ij0daQ+pHBjTx7/oSqkmeZYmA+n1HZigRsNz0pJeaLDmMqcYbNXNZDLsEITtrjx55t3/P40WOa2QW3jo5pa8vDew/4+pe/yuPnT9GIqmNI0DUtJ8dHNHXNRkkPSFJQVYYQxNCSOfAq94lELy1KxopsRfCZQaOyfk5Vcft2jdYJn5ZoW9EtNFcXiRAcL5+9YLPc0DQ1wzBweDjni198E5egbgy37p4Q+wGfAuvNQG0tjTV4JwHF1dWad773HYIfCTpRIXTIGIUdpvcK6SEERu9pu+4VyCKWBrTMkApJoNGmzvTmMuGKPEGNCYfM0Wbp/s14emYDSTYgBqqyTZ5TwV5GxWSk/DAy9Bs2Y2IMnnGUNT9s4FJbotU4FRm2G/w4Sg1AG1SCKzfn6PZnOHl4j8/du8+3v/UeZ9c904zyYkTLKFdAJREPLA1fv6d9nIoWe1G0EsP7468vJAVVUo+9+kCRuy/PRBVoKcbMykpCwc6wVFVVNJWWEbBjwEVAmUmCetYYooN+K4a/0IrXy5HtdpSOZS2NCD4ExsExOIexFd2s/ckM7N7xqXYEIMMYYlKZBROplZbpPHtRdcyTa1RKBC066EVatMwcyPwFGXaSo/gd00DRNC0+BCkmq9JSroTBQSSGhPF5gTimUXo686pDjMQx4JPwr622DMnjQwTnskNK2KJQmqOUImmR8qSNBLz74cf8yt/+Nd76zH3qyhAHEUiLMcv9pkiMDqMs11dXjMOWqj2hXixYHB5wcnrC0ektPvn4Jet+izGGlBR1ZWlay8x33L51zNWFZ9Fp1kPN7du3cClw8YnlYL4gxDUhQVO3DNsV0Yu6qTOW2cEhX3rjIV/9zFsMPnDnaMG773+C1pIJNNai6prW1sxbIxTCPDCFlPBu4Ld/57scdVIQu7q+5oNPnqK14mZ9RuANSImu62RjJJjVlYiHhV3zW0oylyFEsFoxrjY8XW0gjgxbx6xr+eybbzFuA++8/YiYgXOZBSuQQekN8M5DAlMLXDVs/KQTkyIsL0aBV6wm+UAIor8TfWS42eAOLKubLcEberdldTkSgxiGtqpx/TkJmU+LMZiqYrv1nJycyMCY1ZrWQlvDD3/wHT77hS9ydHic+wEMxrYobYhRMqKIrDkVi4EXqfPRRfrBsQgShWbCTY7YTZbTkLXm3MDFxZnw71Nis91SNZ2szrK/NCI5UYItUo6WNTUq9w44NtuBedsxn9U7OrDVVDbR1Ym5ncNbB2zXF6y3ju0YGb0XCEuBioEQNvRuxTgMxFpUZInyvkt3yTvvfRP70dvcffgzfPmLb3LzA4fL+mqKnaU3xuaCecjIVslYSpS+sy9lHaUMe6GkL8ea4gzLf3uRf04pymhaIDMPofRgxOyMUvCUCWclkzJWmJD9IDWWGLLj0na6rKbV+CEyDmPeO4p2NtuhH7qmqluMNkQX8F7ex9qGw6PTn9rSfqodgTWGk6MTKYyYiphk4bnRYWrL6B0BkYVWWkljitJSJ0g53VfTo5LJT6bKmKhDKZ2V8BV1VYnRL+JYe05GVEvzYkulpJSmyCgl4UuHGKisRUUlaow5IpPiVylu7TbUfjGtQEMljQ+RzBSIWC3U1RgTKoKPIz4oqkyh7dqOaA1BKZQSKd7Z/IB2NmedMxAfBsahJ8YF867lZ37257m4uODiesPm+iknhwsODg54+eglp7fu04cnOHczpcghSgNWZTp+4+3f4N/9D/49vvCZL/Dg9be4e+uQ+awhJqktdLMZn/3sQ5z3hM0KYxpICqttNkCRF88vuVQ93axjsZjxy//kL1BVNS9fntGvRmKUJr6UEqMPNLUmsV9clEcxuCjZnyCFE6z07Oya99//gNNbRwRiptQKjdVWDZXJ+HjWF0pRSTSbJRuMVlRGkfKglJQUVWvo2hoiLFdb6trifGDYRp5+dMPssJWJc17hXSmEQ9QVz642LA4WgEzcu1oPxCR1nqPFAS+GkdZCpS1Hx7fp5rMM0wj/3Vgj9ab0Y0FsynpV5CFJSeZDjD5OHa3SDFjWWpCAKUSUhehGkq7RWmOt4eziOacnd7l3+xZKgbXCwFvfvODZ05c8+OxX+eLn71JbQ9tUnBzPaSz0w4bFbMbBvEFbm6FQJRCmjrmjPwuwOU+IgwRCUZ5XCJ71asv7H37A3/hbf5fLi5Ghl6ZCU9VYozFWcdLdoTFQdUf8t/+bDxmHxHY7st46+lGhTMPRwQEpBD56/Ak35y9Fvz87st3e+7HSQQ4KSv2tMnusob36h/x7VyOY7EQi96WU+iUoPBpHTCY75UCMIinSNhVW4iKMkfVrjDChohY2YSgwT5J1aetaVHijSFiEHGCqfG1KaazWxJ9efPTT7Qh88NzcnNM0LW0nUqxlwHZKWT0052U67ppMpCgsYZA0e5UlkDOA7OQlfcwGWYlERMj8/BIlpL1Fk1LKEhZ5QaTS4CKj8iolqV7SRYRrz9QnERKLKuzm06T91HKHM5YhLKKpJEYglZqdjhifi886F/pSxmdLe3mMIglhMwylFDGJbAIpolVi1s0ZFp5mfsLN5h3+L//Zf8K86bh3/w3uPXid2/fu8eL5M7739gf0l2eAGIXGWA7qiq9/4UskDD94+/v8yrNH9KPnYDbj5OSE8+tLusZy//YpSmueffSYlJAuXZUIKXJzfcn1y48xtuYX/9g/xWcfvA5G8fb332Vce5wbScic1uDGHNLubz6AJINi8i8iZEmRBu829EPg2bNLzs+vGf2G5eqaEEYqK12dJDg+PsA5x/qmx1SG+6/fZ73ZcP7ygnpm6GY1243QW41VHB53WGW5c+eI2eGCs/MXDL1n6AM3VyPbrWhHua0MPt+utqgmUtciF5BywNDWNdoY3OgY+h7CiKpkyEg7m5MCJKMyy0qxXm3o+4Gukgyl1Ee00bhRsgScQ+QKcqNYzoZjFENsVCIFTzNecxN6Fvff4MGDW5zcvs0XP/869147RqUt885w7/4bzNpWhuLoxIfvw//7r/2AhXb8t37xds4yhRYZQiT4RoIN3xMGYXP1zjNuRkYniqqjDzLE3Xu8jwxO6mgxBYgOQsK5njsnc44XHWCx1Yy2PcFUMI4buu4uRwctfnuFUyvaruX0wQHd4gHHixNse495dxuN4u0PP2TcDHz/u29TGtSmGcbTpi4GQYKsRKLve1CWvdIjkxVJir3qAcUlxKJ7lGTPpjBgdIM4aBC6regsaQwxyLhNCSJF68kHWdd1p6isZiBO2kikRPCOcRgYR4+PUbrPVZGrUNS1oWosIfwhm0eQUiL4yO37pyTdcHN9g9EZ9dmH9Pb/JodK0/PNhvaVAor63a8nkaOW/XPvG/y0i/D36gkScQXhvmS6KnuwVSk8lulnO7G8NEV2opkl0dM02jLPJA1Vlw2gKkAkUJp8Em4U+MmkSMgFQmkilRRUhplEfHDCmdeWpEUPXSsRWbN1w8P7dxn6nicfP8b+CcPh6TFN23J85z4Xn7wnvPIIJM9idspXv/Z1Dk8OiSGyWa9YLVecnNyiqiuePX/J7/z2tyFa6jrQHdQM45cIKnC9XtJvNjjnCDl1Pzw85LXX7hJTwI3geo9GYazGRem8tbUFVU01I3EEEecGwaC1QmFRKWC1IgaXKe8GowRmePzkE+7fv83Qb9AKTk6OuXv3NoPzfPj+B7z5udf54le/wsXFBYcnB9ysLyQ1Q9HNRXuqri2/9PP/JGcvXjCmwIvzC8mWMr2v0Q1GKYZWoawipYBzjtNbx/RDkNGFo8f1wyS7UFnL4eEBdfQM6yWPH31M4D737z4gRi/BjJIGpRBEpXS93bDdJg4OFqw3veDaIco9tRXD0OO86FupJPILWmuapuLWwTHrccXV2bvMZw0np4am8ayW11iTGEdYjx/R1BXzzuCj429+85ucXXyCfn6PX/tVC0hgMYYkndFJcOxxHBnHgXEcWW+XbAYZ1O7HUaJ353LBc8D7iA8JH3XuEhfFzZgb6ibJZwXHp7ckD/fvMQw988MD2tkB3XzO8WzO4cGM27fv8vrnHvKVr/9xPnP6Wb7ymXv4P/EL/P1f/VUePX5G3VnqapcHaJPZe5l0UuiyJo5s3TDZigkcKkGl2ufvZeg5B3XRDSgTAENKlTT25TqiCgnTiMSMqTRNbaY+D60UWnmGMWCNRsWIVhLaxxRRObtK3lMkb2hrZrOOdjbw+psPeOPB67K/f2wS4k9yfKodQfCB9aYX7Z7GEtWOIVGyginNewUGUjndj3vRvJyztLJrpUhaZY+bO/8mB7EfLuSHPb1JxiKzUfdRCkPS+bq7drUDk/K/9S7STwJBqQkKKq3u8p6lFy2GQEwwei8RvxOxtOjjNGHq7PKci5trFroi2iazDkRHSdczyBRbbYwMvg7iiMYoOHdtE42t+OIXvoJLkXe/9c7kOLWxdG2d/y2FuOvlErSm7WbM54eQErPZgtNTj7VCNf2Fr3+NP/Klz0uRa3CMwbOYLXh+fsnV1Q2DixhT5/4O6YyU56Np247hpgetcue4FnVNrVAYUixwh1Aqh77HmERjhQFzuVzy6PHHDL2cUysEYgqRx49f8vDeCz766BHWJqKqBP9frjk+OeLr3/hZtK7oHszoujnf+u0zgV1ioraa23dmnByfcvvWKRcvLwmbETfKego+4Z3HjYGDA+GG++hxPtJVBtvNGYalyA8gGPTx8Qm+X6OrmvVmSaw7tEu8+OgjZpXm3u0HJBlOIIPUjSX6keUwsL64xtQ151drNusVFpmkdny0oDUN3jsuL86wtsk06IS2lmbW8WLc8PLFFlLg/GLDDz5+Cnw/Q5gRoxN37xiOFzV1pdHe8eEnz7l955Tjww2PPnmb0Q9shp7z6y3X11tubnquV471qmez8figQBvqtqNEb8urDXVrUToXSxvpLg8hSH1OKUJQuUFOT41mQ+94cf6Ctq5lz6VIs1yhWKGNobKabt5x996Kry5b7MF73D98wDs/fI9f/81fZ7VdobTma9/4HEdd4vmLPotLWozNzCulJhYWwMWLGz785Pm0j5mM/c6+7GeoMQdw3gdCNJLh531cqslSCwyk6HGDwJ4xQNVU0j8BNJWmrkXyBrQMAYoJozTW1pnUoVBGUzUtXdtCyuSXMKKS2pt5/JMfn2pHoIwSjW1tIWsExZQy/p5vbirNRMX47qL4cmTezys/VlpNTTbFCahsANMrf7t/QYoy1lKckTzcStlJWCqbdSa0v9QI9lQzd06MXD7aaW7KtcnCDEE0fhRSsBMWS44MUyB6R20Tra2wSuaOgcAjWmeGlZJCoveO9XbLMA6oSk1aM3XVUGnLfLEgKAiBPBxcLrCtiuMVSea7tw55/uELjDJUSqIPqy2pqqfPbUzFrLPomUEj8w16v+Lzb77B5x4+QNuKz33xC3zn6jmkxDh4UvSiEmsUqEhKQUYQVjUJRWNt3qSFvSXPdfReVGALg4eAj4nRFQ2iLPORIimItK8KjrqusKpmdCPX51d0XcVv/fbv8MaDB8zmc24uLhm3EuVaa1C2oh8jB4sZq+UFJJ+fkcoKphIszA9m1E3DervBO8/QD4Rmhq0qcWxZblOZirDe8vLDpywOO1Rj0W2FrSJHs27SPyqccGMNSmtuNgPjepuL2xsUCauzrDqeuq6o6rn0IPgRY2vJEoxlsA0vzlf84Nvvcn72nBQDIUDTtaAt2ta4fkQlz+27nyPqBR5FVW14+MDRtA03q5fcrCIfPr7i17/9gu1GoW1NM+to2pb19cDV2Uoynabl6FbN7Ycn1LZmdfMxIUBlNMd373B8dIRSMA4j235L1zSMo2N1s8JWlhC8DIaJsLze4EZ5potbR9x5/TVUgnF0XLx4ydn5OatxZLsZePHsKb/za+/y5MUjHj9+zHbYolLizp3b/Jk//XXcOPCjD654erbBO6ZMXzIQgXtVfM7Hj87ytt9BCEk2LeVLKrXIlDCVwlYRWyW0digcpgZVyXuoIk1jLNoauk5ELKNqUcmDEYZRk1sflEqk4Bj7LREYxh7vA86PeDx1hotVSvTrDW7osW0tkhM/5fHpdgSIXk/wAdMI/r5rG98v+OyM/CvjK/N50t7vdzzb8t/OQE9HeuXLzkjHRDK7LEFavHNDT8rFY5UFsvayCcrwb6X2z7aDiUp0oTKrIA/cjl6YLGXkHj6Ktw8BsKgUCVHmE1ut8Rp5ryifvm2qnHKqbKgyc6d8KqXRVYNL8OTZE3StGWLCez85r6au8gIW/aKgNT4pdErS2zCNApKnEaKI8enSLLaXPdW2wrYtWlt+7htf5b3vfot+2MpIvxilQGs1qkoMw0DTtFQGghsxpsuOf3orMQRZi9/WljgEtDbcOj7k+fNrMqsdHyLbQbq60Yqmq9kMju12YLvZMI4jn/38m1B33Lv/Oient6jbGe++9x4qSXH7+mpN9A2r5cjDezPqpma5GXCjZB5GSzF3eb1idb2cdHXcMKLy3AvvBJ4Dge98gKANLiSsHzGjGHVrBG5E7iDBy+S2+azh6QvB1bVRufiYqIw0w1krX2PKUg654SukyOgDFzdXbJbnzDrD7PV73Kz7qS/BJ1mVQwjcPu04PDgAtlTGoBipW1Hj7YeRs7M1f+dXH3O1Sdx6/S7Ht0+xdSWjCQL4pxdSa1OOftMzbgf6uBZl32SoG5E+H90oMjG2wlQBKsu42ZBUYnF6SNu0jP3I8uqGfi0D3ZXSjM/PMbXm4GCBAmazhhQcw7bn8YuPefzsY8x3fxvve6IP6Myg+s733mc7DHzp8/f5yhce8KUva374wRVPn67JPAJRWlWKum0nhs6uVix3tdQOdgiyQJX3HnY80NIvY60R2NUc5xniwnuOMTCOG7bDhpsb6WpXtiKMjvV6TQw9x3NhC203K957/znL5SCy8+ue9WZN9dJw8fKI+7dOMZWhndcYPGHscaYqeoA/1fGpdgQg6VpwHpPyAOgYJwmJqVqfi7bS4JJ1QqaYfN/K70E7e0dMO8NdnnpGb6YzqOJcsjOR5im9C/rzK1Lc9R1M7wl7boksy5uEFZAdzMREUHlgZGG0IMZAZ0cikwLFmQmd0rPZDMyOxckI3ztIWm53jSohehmEruVadUpSVK5rktZYazF1SzAt3u8Mia5neRZwytTDCm2MGNmpw1QBJtcwSsf1DsYDJv0TmZ2s6JoWY2v02DN6ofwapXn99buMXsYEVlXLa/cXrLY9h11Hmjaf3PSYwI2Buqppm46kDFXTM18sWC6XrNdLutmCImUQc+TeVA2Vbdmknn6zRStNU9eYWTNBMZW11E1H187ZbLeY3Kh22B0wmy1kLrIPMn1TkQuziqZtJcQwRe/KMoaACQ5lLfOTYw4O52xDop5ZmBuO791m3a+xrcWRuH5+SRNCphMrVtsV3vWZwQaKSO7BQ2lF8IlKG7ra0NSWbt7SdQuqyrLtt6jNEmUqjmrgqGVz07LdDjKYJyaSKbpLCTcM3Lt3D5U8WteSyYYNZe7A0Ae+/4Nr9GzByVyKnuvlRoQNfcpsIVHHjUG62Yd+ZBy2krWERBNrhn4QPN4KI+7yxTmbzUomsinF9cUVWMtnv/xF7r3xOtvVwOqjR6QUUA6WL2+wKWGbSnD/rqY5PCDpOkuoR1Q1JwWPz6Mfb5Y9P/jhE370oyf8+m9+n2987U1+7mc+x+ffbHn7hytenI3oYAR2qes9aXv1ezRpFWRAisEhJKzV1HUWp1T7NQjw45a3v/eb9EMUJx4hJE1Td1hTAVpUjpWnmUsvz/pizdnTS0IWzBt7kam5urjhxdklc6N5+vgF5+dXjFqzOF0y74OID/6Ux6faEaSUcD4wjgNNCjLJMBZVxN1rROMj70j1KjK///hUwX9IO0ZA3gyZC4oCoipQkpr+N5kgtWNhQGntz6eNkHR65X0lgpWzhSQUMq2kV0GuMH+GqUEJUApTNaQ4TIa81DwC0lyndCAGzzisccMWHxwxwwSaiFZQ1/W0iI3RkxZQIqC0FD+rqgYjWG3TNFRGxmCqJJGRrpqpe9Uaw6xuM/1wrx6z91/hvecbPt03Y6Spa4KPrMEjVN6QgX+tDD/3R77KV77wGRbzQ0Lw/PFf/qO0sxnKWEyVGwzL0BNkQI9zIy6OVLXl4PAAN/ZU2hKcw/U9KY50swbnxYGbynJ0sCDgmc8PqaqWo5Njgq547+NnLM6vuXd6wPHJLWpraa/OaTvLYtExb+Z0tiYGpAkrJRlspBUGTVXVHB0eUbctTdPSdJUwfxLUc4NLhpvVyPnVxwz9hnu3Z3yuusuzpzfcrDfE4Oi04vzykvc++oCYIrOu46tf/ioxvMfjJ09pbY3rPQqFrhTeR2ZdTdcIJfa11+5xcnKXYdiyPHuBtQ137h/i3JbLywtulmuur9b5kWkpXJPYLge6znDnzi1QicomSFuKNlQKiY8+WnE9GO68fpe6afEh4nKNZxw9XVszP5zRr3Itx3mG9QZloG4q3BDwznNzfpX7MixKa4bNQHCOcSsjGIMXiuvx6S1mb805unXC00+eMowjoFjfrGm7ijoEfG78M6aCLG6HkaAoeljl8N0Yw+HhIZvVho8+uOa9t/8ef+/Xv8sv/9LP8I2vfpaH9yI//OCa1TpNYpZiyPehoV3QVmK9mFKmpMtnFskhec+YREYvxER/s6R3iXrWcefuaxhT44eI916ciU/4KJlid6AJQQtrLvppTojrZTTl9eUNm0XD8mrFsBqoFwPD6KgqL3WEn/L4VDuCYlBdRPjcGGko2isF5JfJK0saN2FGu3ygSBsXAYpSuCVJEVcicU1lK0mHlcBS+UXTeUIQsThlNPvFI0UerZLfO+0vnkIdjYL3J5UF4PIUo5CEUmq1mfSRrDXEQd7fe48OERdihowS2qhMw3NCz/MZY0SuXSmFqltS0oQYcCGw3o6MmTcfjShxtpWmsxWH8xMWR4fMGhnVKRi1qChqBT6CC4m2tqRYII7iAPafVy7Am+ImBFYSobJyv2KGqIxotUQpjEUSddNINKYUSld84XOfJyY4u7qQUZv5/DEltE5EP9DUBjducC5IL0WKpOh590cfigxxFLnolDxWQVNb2sail3B0esj88BBTG66vNrw8W3Fzteb+6QF37r7GsNmwtSJsWM9uo9sjtmPP6DwpCRarFMy7GZ/7/Od4+MZbtE2HsTXWWkY3ErzDJ5g1FdttnxlTns1mjfEb3LCF4PCD1DD0QcvZ2TXPz3+bNz7zBil46rajqmuSj1StoVlIJ7SxlmdX16ikOJg1ECPXyyWHhyfYqubuvYecnN4hqZHnZ2c8e3LGzfWGBMxyX0O/hn7rcKPjwf0DrOohaqFqhyVlTsRqOfLhJyuqg2OquhZee4pUQSbtVWakaRsWixkET/DSue9d5GA+E2mE4HAuN1mFhI5CFohRApxhK9IdKKl1uRBxzlF1DU3X0vciqTD2I6vrFcYqbF1j25Y+XktHddhgqxZtm1wjyzZAJZpOpvIN28DY91y+HPi//bW/za+/+T1+6Rd/hq9+4QHXy8Db22XuXTCZkVZ6ghDnz54ziInNJnJ+6TEm0nQaHzzdzGIZqaxBo3n45iEApmqpukPGTSKOI1UjirljiuDE1vhBuq5lpoEEjl4FmW+gpNE1kaVmdLF7+8HZT3f8YzsCJeHfPwAep5T+rFLqs8BfBW4Bvwn8j1JKo1KqAf4j4I8C58A/n1L68A84NzI0Jk6FXPJEqsIl31UEXi24kvY0NzJ+DzlbV5quqggJYlXt6d8jkhAZF4pOMPzSLe9SEPw7z5grUWl5/X7PgUpZjSVJj0EIIeOxuRciS+zqlHYLDeQ8Ueb0piRReGMr+WSjsGmiEtbJgMhk26waGZI4nSLkprUBpYVxoGUAjlFaxN7yDbG2wmgppErTlxRq0QJDdXWWwvABPw5Yq1AqZMmC0tyQ0yFAKdHZ3x0iKhZjLn5rUMj1HixqhrURDnkKKORz6iLSpwK6akgkdDKiNIsMFk85kxvHgXnXsZjPAdHbf4ni+PiIr3zx8yhj2SxveP78HIXl4OiI7eipq5rKKtara2bzCqMW3D454Ga5ZNGZvHklq9K6xlYtSle88dZbzOdWOkejxyiJat/8wlvcffNNTLfAdDNmswVt1+K95+bmmnHdM791ShMi65trri/OMSqSVMVVTHgl8xhCVvQ8mM3Y9FseP37Gw4cPaduWyhpmtXS8liYppRSN1dMM25AiVy8vuf/aQ6zVtLNDVKN5+ugFP/jeByyX4gTaRSeDaUaPtpax36CI3Lm9wNYH1DqhGCjzhmNIfPDRDZerQKO3tIcD9WyOURavfG7IiyJSWDVSP1JSxxidF5qk0pjK4jaeWAmcoqJAYCEE+sHjRMVNuqhJuHEgeI81mnbWcnVxJasqs2cODmacn1/DaisT2oxQM23Tcnz7mNrWKCVYvMsRN9pgm4rXXn+Dr/7Rr/OdX/lbPP7wgr/6/n/Fl7/2gF/+J36WL33xkN/8puHyJhelVMkAUg44dw2hKSYuX3quzm9Q2vLa64f02y0PP3NAGy5pm5Gua5nNm0ks0xGzhL1GVQJ128ZycbZku9miKuiXa1bbkVu3FxAi4+Co2iqTSLLce0ZEgg95TKWgJj/t8V9HRvCvAW8Dh/nf/zvgf59S+qtKqX8X+J8C/07+eplS+oJS6l/Ir/vn/6CTF25+KaKmELGqFH1zE1aeDgSv+sIdvTTXE/LA+hi9YKtBpvqUotDUIxDTbvBHChPur9Gij5PPSq4XlMEfJTksLIISN2hkEpqhKI/qUkPFap2bc0QiQGbcCMujOzrhzr37JAL9zRVuHDEoZJSjRNsxR1SaNClPlppGXaJ573F+ZBwGaVLTWgaRI5mFzw7LGIXXWpq5ksT3bVPJHANTsHPp7g7BM+Fi5M2C1FvEQe7rxBeudaHzKowxWW8rMPoR50dMELqE1TvHXNymygVoF8pAH0ApBi+9JnlYMRaYHcwI1Dx4+JDT0yO224H3f/SUpmpZHM6I0fGNr/8sXwmf5e/8+neoaxF6W/dLDg8cb73xgHbewNiTVmsO5zOSGxi2Gx598hGnR4fMupYYHE1dc3pyzKLV/N2/8beIXpxUiCFngAofRH2yqqwoUOa51UYr3LBBaccbd29RLw4IyxUXlzesNiO9d4wjXH3xklsnxySfsNUBwzhibWS2WOBTxZ3XOkYf6IPB3WwxbSDESGMquuMD3v7u7/Cj9z9kHDMbq2uo25roA24c8D4xbgcW84rjwwY/nNE0lhTWgAwGur4e+ODjFT5qws2WfvOY7d0NR6eSeYQQpgw1xID3IXdkiwx53w+QZcPHYaTvxcBP9bwc6JV6G1GaqZ5/9ITV5RUHR4eE5Cm9O6TEzeUKrOEzn/8c737vPbarLabWdJ2IFVaNpjo6nNZS8NLToY2hbg0fv/tDzj55xD/75/5ZfuHnv8b/8d/+S/zWr7/Lu+8+5htffR039rnRc2dQ0uQXdphwCUpSBG1F9gRtCFHW8WY14gYJQOvWMusqxn4gxBplNWM/4oInes96tWYMiVkIDMNIvx0ZnacqbDOy/lXcZ0oWeLkExP9/hoaUUq8D/wzwvwH+Z0pC8H8a+B/ml/wV4H+FOII/n78H+E+Bf1sppdI+V/N3HRmzD57SuBGi6AHhA3uZ2pQWFVG28vfl9yEGxmEkIaiFUjpHdSFLQYu8tEFJNJx5xfK6SHCeqtbTjZ+U/ig4/84JFS2SKV1TUBmDTyIhXFg4qmQtSiQOyiLTSjPrOlbnF5w9f4aJDmvz3FcluKNBDKsPjn4QVkFUAWUsKI3G0GSnRQKrEMzcBZTReB3EgGpDtK0MJzGWVHcMg0hQoxTa1ERVEUJgGD3RWEISBykXLFmH7I5cQFcRprtDrpGUgTKSPdVWM5u1XCqD91JUrJsao2VYt86d0mUMo9GRqtYMTnjYhTK67Uecc6g80YqcXcWoIEDXdqSkiMpgTM3x8S2ads6Xv/J53nzjFpvQ0dWG1157jY+efsQ/+AdXvDh/RlO1XK5u0DERk0fVNZ994yGtrQljL7gwmqPDBXdv3+Lvf/M7XLy4lFWRACV9EEeLCj961psRbWUSWAw+S0PLMBxtFTbA/ddu01YVLy+v+eDJJYdHcxKa9370IQeLhpdPn3B2tiLEwOmJZXY4Z7XqgYqAYrPc4MeB+eGI945k4fLlUz7+0Uesl2tx0FpRYxk3S4bB4X1i2ErT3p37b3Bw8nlUciTtiH4DSTR7PvzwhpuVQ1dVljhInD96wcWzl9iqop61aMDWFVcXF2zX6wxdGPw40q9uqKrEdh2EoaNKViN7MYQgNOK8qeXnin69pV9vOX96LgYwFFgx99P4wKOPn7BZbohRhiGREuNm4PF7j1jeOeL24WEelpS4feuE2cEhRMfVs0tefPyEH7zzHdp5oJtZDk9PuDm/5m/9yvscHBzsaoklvCobvQRBqQRACWMrbFXjR0fVNkCeS2AUs06mKGojn3VYb0gmYSuZZqYMbAcHSvoKTm4fYYgsVjeMm4FgxG6YrEthc82gIBMxF8it0eif3g/8Y2cE/wfgfwGU2Wi3gKtUwFN4BDzM3z8EPgFIKXml1HV+/cv9Eyql/lXgXwXoZi1JgYs7GChm1b2YQ99XMgDYM8+7n4A0kNV5BJ3ORl5YPlKkNXo3VCNfx2Sk+6EX5c0CNe0sfv6aptGX5R1LVjBdWywLiUlcDhDu+0QvLawj0V4R2qA0KkkXZ8SqLGubAuMYUHrL8mbFph+gmaGz9jxaumCV1vgQcN6z6R3OS5FRKYlSrTFUpiIEUFrTVC0uKx1qBbquJVKLie1mw9PLax49e8K3vvNbeeZBg7Wib6S1QWXF1lnTiiPVUgR2UYr+BplZEIKnspraiDT1ZrOlqmKmnioSAmFVVTVdf1Lw8sVLrq5vOD465OBgxmpzxXJ1zfHJHK1rdJI5xmfPn/D+++8xW+RIcBjoMZBnAK+XA5WC1hoe3L3Hn/pT/xRj+AVCH/kHf/9Xqbjk4uUlldH4ELl355DbJ7c4OpxxfXXOph948eKCtq5Yblf4IdC0XW4AiqAVXV2jFcyaGlXPpmfuRyepvOuFdgmsllvOzDm3F3OZWAVsNkKX/MH3f8Tq5UtM1Dg/EGJAJcPq4oxkO1yI+HHEjx43OsZh5OLiBe+++w7XV0turtcolZjNpPaixy3OB7Zrhw8K7yJWJxbtwOryB8wbj26yXo42kDRJV5ye7AyjzGdSaF1hlDRmWQNKRw7vH6HuH2GsFmaVhqatadsaqEVgUYsekUTTwjCSdZ1yDSjPpXZOJBWc1LlCCDmAE10l4xPO98znHUorTu8eMG5H+q2T/g5XGHaafrXl7W+/Iz1JMXB5fk4/DnzrW7/Bo0++y3qsKPFNcFli+8ep5q8cu38rpWhmHQbDOIxYYxmHHu0iIx6rR5kxnmdl91sRqKw7kccetgM+BCKJpqsgqsyGq/HjMCEMtrJURtO0NXVTY62mshqjEyo4dBLE46c9/pEdgVLqzwIvUkq/qZT6k/+o5/nxI6X0l4C/BHB06ziR0y4VJY0MAEZLkZFcNExpV6nVit3gACjcc60UytodzECemKT3Ow5gn+lCErleY/SOm1uK0HE/JVO7NZGSdDjnDZOmKDjPZVLspHL3oTzJMvNrxOPHJNQ0rQ1NpTDaY7Rm1JHaGhQCTRhjaaqasQw2zxeja6FoGuOxlXR0NnWFMjXKyHjJxho6m+ElpTisIY7jLnq0FZWR/MfHwIsnH/HiySf8P/+zv8bJ7dvS5JYzp9LvYLTJDI4M9RiRpU542qbD2gplIpc315hac371kn//P/wrzNtGPmsrIzSbtmE+O0ABW7fh/e9/j5cvLlnerLl4uaSpKz766Al1Y3nrwWs0TUfbdvjoeOedb/Pd7/8mX/zcfT7z+hd4+egR2/kRV1cXVFVNV9fM53OOT4/49vfewTuHsqIOa42mtrnzU8GsNrw2A+duuB4Mj69uOF9uSEZxeLLg5dkVSUmjkNJK5jNoxRfnFTOd6O7c4pG3fPzoBZDQrSWGwBB8zgwV/ejxjcUfdNgLgzVamuIyJLBe9RzMukwXjWz7EWU8zeGCcbthGPrcfaz4wmfu8ebrJ8zbwFWXWB1UaB2xlcy0aCpFCD0pNdR1zeam5+zDNfHjCy6facz9luaNI2ZHBxwf3eP0+ICvf72GqNCVxkp/J5WWLNKYGm0qJp02LemwJrDZDkQCXdvR1K2sBYXIS6NBZdoxmeXDLujbReEmZ8wxOwNPiImhoJMxMQyOfhgZh57NMLDdbFmtRzbbgR/+6BNWVxtMUtw+aPCj4+rqmtAvMTrivOL0tc/yCw/fwAV48fKCfrXh5vqCs7NryuCfuBfcFZnpsu2Vknpda2u2Q08MkWE9cHhombWKRdfsh5ncXIwEsry50thK+lCqumJ+OKedteD9lBGXoDdmUcXGaNrKcu/eMW1jaI4Pub1o0c6RhtJa+pMf/zgZwR8H/pxS6s8ALVIj+IvAsVLK5qzgdeBxfv1j4A3gkVLKAkdI0fj3PUql3ocg0XFWgqy0xmUMcYeNZSApkaEhldOmgtdnS0tZZPvgTn4/tY9454lYiKJnmGoNFBnRXcQ/4VO7wvUONS8F6umNdxll2ssg2MsSMiHHeZ+phwVzV0zopEKYSQlGH9GJ7OTUdF6Zu2zwbktwHj+OAqvYSsbcGYOxFmMTKTpi8qyWNzR5xmpMEm0orXHe450D0+JGz3az4XCU1n0f93TrUdNnUq9Q7xRllmxKYGzN/NZdTCXjGL/1G7/FwazF2gyVkbO2VOQ3Ep989JgU5LmPzuEGx9gP/OjtH3Dx9HGGpUSy4MXzJzLUZbMmjYmoLOsXH/P86QvunJ5y77Vj6sqKwXcbHv3wHd79+ENG59FoNssrZouO2cGCwyqxXl3z9jvv4KjYjA43Rm4fHVBVNVeXl9IZCpA0zWyOjZ5GweL2LW5/5WusHj3n7PySrFghkgrWEoq4nvMsXy6JQ8JtBspKDSFMc4ldSFKfiY6LpWc2n1EnqZH0257DowMevHGHz3/mLl0bae42vH7ntsBmmQCQ4ih6+XSiQ+UjH55tUcZwu6240zVcPV4z3IyEz49UOtFvX6JV7s7ee8bFOqlpXSZSKjOH5bm5MOKCBBTG6CzjIM+47NkQYJo5bhRNJd3z2igiFue0TPayFlPVtHXNrG2wNuFTYtHNOD2xaNOi1DGrjRchPESiJfSOD370guPT2/zr//N/kYPG8Ld//df5K3/5v2C19HzjG3+Mf+lf+Qv80s99lUXXsrx5TnI9//f/9Ff5P/0Hf21v08YdA2ma3JbvhNDfpNZmLbODOQeHM7S5QRvQlcEoA8rmfooNQXliCKz6Hu8dwzCwXG/wOFLcUpX5CtkRlfcZtiNus2F59pKZClSLCm08Vb9E6S2qH/5hZvX3PP6RHUFK6d8C/i2AnBH86ymlf1Ep9Z8AfwFhDv1LwH+e/+T/kf/9a/n3f+MfXh/IEX+MwhMfNhAdKQ/qLkqdau/Vk/FXZIVQNUXhlH8ppnpDqTGU99qPQ0KMUtw1ZurynICn6e/E7MfJEb36cYqxL0WmKVrPGYwMwIqkqKc6R8p/qJQSnDeWZrdS78jfl68kghun9FrbHYOqri1VZYRxYw0hJXwAFZNIMyuRHk7GMI4Dm82WDz56wuLLC8oNqoyI16W0FYE30wHk+QgJzd71RZWb88TxiUa7YifbEXeCgcqgMMJiyl255AGGkgVmAbWp/lBYVbuoLKWUO89l40cVkWFGcTd5K0onblVbtmNgu90yuBGtItc311xcXKFVYHZyG/2o4vLqJXVVsxk23L214PlLx93XW67XluQCmEToR2aN5Quff4urmyvu3Tlm0VW4oNDKSkHbK5qTBbe/+jVM22KMghQwgEoRoyL1YYdzUjg0WqMirC+XGLLExxQt5HqN1TSdZbkWGZDNpmd+6MTgOscwjIIPq8jl1UAMA9GPYMRIVQa8H8RxjIH5TKFcwo+wOJrRtZZrF1jbBn+xRo0vCWrAZqG2UqDfbTJh74xXHpzOMzU8qWhCxQSzAAspjMtw9/I8ZT++fNnz228vcU4+a9Vo/ug3jjk5kLm9l5cjv/GtKwaX93Me3zibW+4/aHl4p2XeVVPtbb0O/INvXbDaCORqasPd20cAXJ2f8V/9F/8l43bkb/7697m8kLrNs7Mn/If//v+Vv/651/hv/MJnOJ1VBD/y6ON3M7W5rHAgs/2KvPxurwuLLEZH1dRUdUU360irG0Y3EKoK3RicF2ZcjJGt20qdKIB3I9vNkuAcN5cjVRU5PJhz+84hzx8N+DANXhAYbQi4fouJUeatjwm/3mDbRkbk/pTH/y/6CP4N4K8qpf7XwG8Dfzn//C8D/2el1PvABfAv/EEn8qPnox9+iNaKfrPh6PYtyuSoaZwgTJHFLhYlr9W9B7UH96DUJAHAK75IXhWTRJ3aGoqxLUY4qZ3TkE7j9CqEmHavJ+0YRdNSSiViz7+LEFXMfxpz7UM6CWPWGgpkyd7cAzC6QFvnzuMUWa+X9OOAajpSlnxQCMvKKBm/GLyj7zfSk6BHUiUyyCa3uI8+4aOiaWspShnJvrQ1KF2T4prgE7Na4AUfIterFcPgOFp0zGYziHFiMzw5O+fqesVn33jAvGuBhPORd374CSeLjtffeDjBZFpnHZasOmmsJgaRfo7E3FldCsdMoypBjE2h5craSGz7Aec8/eCks3ToieOAH2HoR9wwMoyO8WJkdXNB0844PFlws+n5xs//DI+ffYgOgVunh9y+pehmc+7cfsDTlxe8+4MPSDHQdS0medbrNVXTYHpH3/eoSrHdbNluHPOf+wbVbM6w3ZA2K+7dPcJYy9nZFZvVkhgTw1acwOmtQ9qu5eZqyWo14mPIvRqJuqmYHx/ihnFSbFVK453n4uKlwG4x4saBqhLJkRAixjSMDsIgv9ssl2zHQZZqjNT3O5SD5vZtYjIMTc2q9xw8eAAvnnD1/g/ZLD3zo9xhWwKfKA1SWol0s7tW2Nktkc32vcxhCA7Q+NUlqokEBdshMjpIynD33pucHmo++uhdbq5GfOZom43m+YuBgy6PoSXhXczy3mWfJa6vE8+fLzl70PK1Lx5QGY0LCRcMd+8dcvXBGcvzNShFXRmssgzDwMcfvMOPfnDJk6cDWitunWqWZx/z8Xsf8Fvf1Pzwd0758lsyDe/ps53dUJA1J/eygOITsx1KITH4LW0tEtYhiF6WRmfHiAg/RsmAtLUsjhZ45/B9Yhgq3nh4j6q28ndKM6ZdVi2om8q6UkkCyAi2MxAV4+BIlfSv/LTHfy2OIKX0K8Cv5O9/BPzi7/GaHvjnfrrzwtXFNUZrGtvQzeYYo7FIQ0WMCZ0Hw+xulRjbYuenE+VD/a5vcrQ+LXTJBipjpj9N7B7+XjWBVwGdfW7xzj2Vcyi9+4vd4pHxidMQndx4logT3TJmTfnBeVKmSaYYqasWhZuGt2hjpjnjZWGaycgqrLFAEhxXi7HQaLS2VLrCu0RtDa/de43gxFCI0J1GGUNwkeAjbQOFavv2+4948uglP/fVz/DFL72RszTR0D87X/Po4+c8vH2bWVvnz6548fKaNEbeeD2hkqRFCikKh+hJPvDy/JLjowNk5oLo2F9eyjooTnfnxxPeuynzizHhs9PU2blopWhqw3K9JuZJTsENvHh2zmZ5Tu8DFy80904bDCNt3TG3FQfzU2xd40YZuPPg7h0qYzk6nFG1Ddc3Sz780RMuzm8IIVDXFZ2S+w4wxsAwbDl7+13Wz59z3bQ0bcMwjGw3IsdcmB9Pn7ykqmtOTg9pupb1apiCltpaUu9wPmArmWtAzmrdtqcPUlxuO7nW5dUzTG2Zdwc0NaQUICSGWU0/qEmCoGk065UjeM/sYEbbNOjKCxQ4m6Grms3FSHfQoXSagpOQh+WEEDJbJUwBlKpq4cfXNXHoiX1CB3Ah8KMPN4xOoJObzRmnJy1XN+6VYesxJM7OB9560EotQoPRKa9xxRsPG6yKPHk+su3hyeOeppLNtelhfnLMF7/0Nfr0I969fg/n8x6qRPJjuwmcX46kmDg4tHztS0eoEPnWt6+5WHk+/viae6eaxiiGPk1w5s4m7SikrwAASXoY2q6l7lpShGGQddk0GmXFHmhcNkfyedwYWBzWBBO4upRObKsVIThcFOptoeGWw4+eq+WaWwtxMjjJoK9XI1yssVXFT3t8qjuLjTEsDucoEtpErAGfEpXajYoUudc0wQYAU/GJYjBy5J/nG+8bc3EiaXo4Icap07hkwtMgmiKwVuYO8MqJJgBn/zeld2Ff1TSqREyZZ54NPpCnaPmM18s8Y+89RiesUUQjhVeU3XN84IaYZR+8QANaImijNcrUDM7jw8jQZ2437KA1pWUakx+preb45Bh/M2SWU6IyRqirKTK6LSkNxCATvG7fOebp0wvGsYjU7aKX01uHPH70PLOy5F5Yo5jPO5nfLGFpvscqSyzLYJVt7zk63CVSQhtOKFOei5oMUUrSKESGoJSCpqlkspqTgME7R9NUbLYbbm5u8P4Wy9WS9z/8gK6rmStL9J6jgwUnx7e4dXyP9XbNzXrD448/YuxHQlKijukcL6/WWAPDZs3Zswupd4bIeuXot32OmsH1G569+5LVx5+w7R2Ptxfs91KkLAVeGq9QTp63kSL7RFILEd8PRK2oZy1VVWWERgKH9XqAPLQ+hIhKjtpo0raHEFExoHA0VgtFOJmpB8c5CSKM99hM9Ux+IKpEdXrM+uoF/kGDzSq0PhQ9rwyvArSOfvWYFJKwjYzslaRBHwEY6trwlc/PQctAmxjWqLQkZnmKttEQoXeRmxvHpg8czEymXu9W+8lxw+t3aw5nG77zgzXeJ549H/iZryzoWks/Dnz7m3+DyyvH4Vxxfk3ukZE9u147+lH2YdMaKiNd/LdPLUMeLjMMAdOYSdvpx/d5CUDKofJ+GjLWP1Oaps7OwEfmncBWSkNdZ0hVRYb1hrqqODFzKbYrNUlSxCBQ76QvNZXaJADuh8CwGbI9SqSQeHG25mbM9Pqf8vhUOwKt4XNfeAutFYeLGZgKv3UCAxSXvF8nyIG4oDFT318e8r77mdrriJ18RxK9kMoIninxem5Wy5ieynE0SqQiikZOiUZTlCEUSksjUQxeOibHXnT1M3XTWIN3npiHWpu9B+e9NJaU6NY5T7LSPaqMxposCx1DptiJxovIDgdsjJQZSsYYorJohDLrvMO7iLYer0cxWCHmhbclhYQ1wtDxzqN1BCLWKHrv2W6XVIsFylqcC9x9cIeq/kTmMmfstzjPtm0ylBMnp5VSorKG6OMOQhPlX4zWeRC7yVPnYJ8QbduK5PMzUFJSKOPWpasy5PtY2u0z0SBGQhxJqSES+PjJI5JxnF1eYirD4eExIST6fkDblqvrDavtludnF5JqR7mPTVWhtGHeVSRgdXPDzVU2ZEPAxcTodniyVpqX775LMzg00BzPCKut1Jv2Ic2UB8sn6a5dLdd0sw5jjIiN5fRn7AfirKUzmm7WEmNk6IVeqpU4vOAznKQbSBDUCHhiGKUB0JJXdS7aRxi3jnXfY6uGJkWSVti6xY0j1WJBfP6csd8zLkmyRbTKDYGgDis2ZqS2Cp81wUraZm2hoWqqvK9IHqwieMW2l5McLCyWxLPzyDBErpeOeaum+K1k7s7LtZ+e1MyaDTebRN9HRg+njeLq5Zanz3raznD7pOJqKdmK1tJ8ud3m7uKCIGd23FtvzXnjDXEYVaX2SBr5UKW+CKi0G05fzE7+pzaWcXBs1huUApMcPuZMNcokMgBTGbQOtK0l9gNxlL4oMnUabVE4dBCKt8zfyFFRKaxbQ1NJUBhT4uAwEQfpFP9pj0+1I0AxDUSvqgZfHkzZTNKGKxlBrh2UYd5Qnl2iFGtjzOqcKk5Yf5n3GrLqYTJW1Df3FoELjrEXHrA2euonCDklNlN3ZCKFkFUHTcZSIw5yFLQrrBaPL+5INEXkGnNvQ15o3jmUqaQYjMIhdYGoFM4FooLV+ophGBmSZt4G0GFKt7XWbMaRoe/pt0uGvkdXmqCN6Ae5EZek2c4HD0qzHQY26xtUigzrG5T29MPIsPHMTufo3ACkEBZLaRTbTXcQzXnBSROJzJ+OQcY1+pEC4WmliVqhrcmqqnB6+0DOmwSaspXlzvEBL86v84ZTe5Ga3GPJFBK2MmjTcHQyg+tE0pp1SPSrntnRCb/9/ff5rbd/SELktqNQVnA+UNU13nnWaxkLKRkHOfo1mSIpRnu7XTOMLl9Bjj72w48U2V4sOWgrZg8fkO7eQT/95kQ8yFA3r0CKSTGOHmNcDmay0zcWTyT6KHINIeCd+119LTEEojZc9zNGKk6ODjg8OkQxcH1zw2q1IvgR70aGcST6wEhHozVVijKqNXk2n3yIX69Iw0DbtQwDVN0uaCpQpVKKyhqqSnO7NlNW/QqOHmEY4959yk8tJZxPbHuJeLvOMK8Uzy8cIcDlteO105pEyrLssq9DkH1mraapNWxENXW18cTjim0vGUtVaQ4OKmo7yD7Lfz+6oni1229RgTVaui5zNrfrKN5LCUqZb4KL96DoiAjCBYWZ1YTBY2aKOkLfB7pGYVKFDlKvM8ailGezjQRXyZhMbWhnDYuF9GuM2w2jHqisxQUP5GJ7rpElranqaurzmc0Sqakxdf0PMaq/9/HpdgR7hRhrKpQxrDdDZpRIV6E2Ydp8Ou685Q6fz0a3yEbnkLV40ZBHyQk8YfBOsMR9qecQA24cSdpgkp2GpRetIWGrpAmBElw/7kVPenc9MRFU3IOMJPrVccpNJsirOCgbBZtNSolglxbVUJXNrs+dqlfLa0zXobWS2QHOoVLIao5eRL+ydkxwW9Z9ZNtviLFnuV6LGJpSbDYrzh5/Ipsnet68f5fNs0fCTnJx2gyl3V0yIWlkK+5Ta+mQlHrHDrq5/+A2Ril0JUuvjAc1laWkdG1VU2jBATBTQVhl5wNJa0zuPNZGsTicc+vWCbdPTjBVxde/ARfXN7jRZzhEGoS8Dxir6eZzzi8u+OjR05yqK9wwyrUrmHW1RGb5qmIUEcIYZDKc956bqyUqRE6NAmu4qi0+7TlGY+hee43bf+RrXK83ogabn6/aPW66FNAJ+iRqu4myFr1AI10jReAcHBij8V5R13LPnKsYhlHqZyaShhds+gjjAcP6gBQGzq82XC97kTApGh1Kga5pTxe8dI5nV1dUbc3R7ROGRuZ3oyCmy1z83WXgNkssl0xaQd5femo0TymPaIVJFqQ4CwA3RvoxO4LWcNAqrFb4CKtVENl09nWV5Dxp0skqMs8K7xIhIjLPSlE3hm5m6Bo9rXlQhFx8VUjBO0SyZA0TFjk56amxZ3pU2WGnV6BhVV5XmG3GkHTF6CNtrehaw3xeURsFqiJFw+AidjUKJGhmRF1jmw3KCM06+AAhUjWWbjHDq5AlZCJ1IzMdBERSFAbX0WHDAfoPnyMoDWOJbAiNCKu5EOj7rczcdXUWVyuNTSV62DF3ypxbU2oE+XUpRuFUp8xIIRbpNFk8SpF0rgfEJJlEnv8qq6IshkybLAY8qTxNLTuyFCHpqRisMmyRMi9ZqbQTxStGLzflxBhJeQaD0iKqVtQGhWKcCD4x+pHHnzwihFGMSPRcXpyhlGO53FCZyKLeTEZqs1qxWQ4o5fns3QNWz58zrG+I3rNaX/Od3/omb9w+ojOKe3Zk8cWHhLDmxQe/IyJxSLZRt3VmVzHlyFppbt26zT/9px9g0wjjRhyignu3j6S/IUYUco8lsrQZUimGqhgUKarVTYNPUNmar331LZq2IYYsrFdr6tmcZR+4enTGMI4Cl2XGls60RWstKDg8PmDVb7i8uiFG2GwHxkHkN2KMFOZmaZATzSmhs6psxJdXN6xWW0bnWCZknoM1FGqyNobq1imnX/8j2LZDr7dUVhOjmZ59mQK3iZBCkGbJnKnOZg0+SNTddLWMQkXujTIiz6GnrFFgQmMtb71xxLz1aJXwqeflhRAEABbzhqbVNGYghR5tVFa/vZqYLN4tieMZ8wqSlYcqoyTFAUuZTMDHGEXuweRAp8zHnox9ZkjobCAn9c4cjfdbgXSUUrStoZtp6krjx8RmGxlcoqsVxuxqT1lvTZ7rXlZYuo37QSL+qhIHMu8M6yzjLmy4HCsm6V+IMRGNMKGGUbrtm1pEGhWptB7tQcjy1ZSbys4UFBl5a430qFQm7+PEer1mk71oinB+dsOTpxtmx5GjQ8sQKlYXG7yGjVlTW43b9Gyjxw2Bm5WoG4QQmS9arLVcbQOeQG0NWgmEaY3eq2n+5Men2hGoEvMm0WdR0Wb4RTZRCAFTom90Lr7qV/C7DKhn3n6aokqJ+nkF3iiFGp27gEOMWF4t+JpUIoZdZFeGtkhPgODYEh3I4oxBTc0jJNlL5ZzSqEMJsVEqN3ppUTlNeZZrjJIiS/PcvgQ2udfC8ezZM1LoqWtDW89oQuJnv/pZ1i8eMW8qjHWsV5fENLC6vmH58gI/bBhXN5zGGz765n/JePRZ1qs17/z93+SPvnXE3ZM5AZ17AiJsR2ksInHn9IR/5s/8Mm7jIMsm6FJEHHtqK30fITOfNKJoWnojvBtwrgcSda4VaCOFTGuttNNXhqapabuOe6pBG83BfIGPJUjYwXQ6r5l1P7DerBnGnqA0T548Z70ec7aWeXw56xOoYdc2rpWaROF0lg8OMYj0d4pTsBFC0ZHPGR0pFz5zzSdBfXpE1bX4oefy/R9K81hJ66PUNmR5SvMisooxWjGbd6DVlPlGlUMUpVBaZJ9D/vTdYkHVNMxnFUcH0kBYVRbvEjerxGJWk7AyX9slukoEG21taNpWolLyEKgBtts+yz1IDSqoNNE7y30ir0cFIjXtI9vesd4E+jFMxn0xq2SCGkJ4SCXITrDtRTFTG6lxuBCloXCI9H3gZjkyu9VkxVsx4P0Q2Y6R2qo8LyBgK/LeEFIBwM1N4Icf9Sz7RKolOzNa07Sa2mp6F3EusRkSioAbE++8v2bwic9+Zs69k2qqd+0KCrJPp8zgxw5bWzA2R+5yz7xSjF7JTOKi5IrCeQlCU5AaTlQNIRn84NAzgZq0UcQ+MvaOzXpE55GkAKOL9ENguXYCaynJEIWh94fMEUgyJzbSx0glJjSP48vevFTVjUKki9lH7ybNEpULcuJI0oTrviLQlIvKWk1d8sJh1xpdVZPcc3E0oiBpMJSag+SY2qhppoEEySWClysrEY3Jq7sYsKikOYoMiZxfXRNDwNaGGF1WHk1QlEpTJIXEOCy5WV5xeXXNm6czDrD4zRWEntO45U//0a8Q/JZh0/Pu3/3rRGVE72Z7QwyjvLcWxomvTkTXJxefE4qUhGaqVaKpMwNCadY311SNzF11vojQqVxkz/WXlKZOUrIDJhfoVv2GIQq9tZ21zA9m3Ll7QtU0HCwOWcwXWKUYQ2S9WXOz2rBZrfno8XOW6y3Rh+lJF7ijqiwYw3Yc8OPI4cGc6OPkbHWGB4qmkzYyF6E0G6Lks6oyHznKHInikAsdd3+wUEgRN/pXCpDJGHEUo+f8u2/TP3kmUt45IytquCntIuWSzY5jwFZBZhpUu59LkXwHf0pUut88qBiHEYOwqbQOLOaGqoKY8iDcMBKDy9F0ZivlNeWHUcaU5g8WAlNWpch2TO2aNH2I+DFydTPy0eM1Zy9HhiEPUkKhNczmhvv3Wx7c6egakSsp2cF6m+tyEd774QalEt4XFhVcXHtOjyrZf3kvBS8Z8OATo5css9Ck19vI6EQja7v2bNeehGJe5UejFEZDW4tx9i6yXDms0qxXgc02iDS9L2u/yMKUIK3AwflevGJwE9qKQXajZww9ymjaNmRDJpz/ECKD84QkBIdx9GyHADONMjID3DuHi5HgAqt1L9mt5Jq5PiFRbAxRoGKYpOyvfhfN6Sc7PtWOAHLzUBBeuMkRd/ABYw1aa6zeRc+icyPUQlk8k81BKxn+LTN8xTgp5Gem5H9a/k/s1pQHExE9I5PlX30Rv4oBkElhwmjwwubJxewSOQpzRmAdP0pTkIw2FA62ijK1LIUyZk6yk5tlT1jM6D9eMa+0MG5Ik8xCTIHkA/H6hu/9Grz3m9/l6OKU41lDRE/psooejCUFz2gblKkK4InEskC+P8E5XBiJSeFVTWqOeXpxhU2eB/fvUauBWDZliGgvc4J11tAQHaXsdDN0oJSZPKtCqKJaaUJlOTpu8G7k9PQer99/wGzeSd0hSdEu+kRtE6aaMT8Qhzv2G569fMmz5y9YLze4LHmMNgSjuXh5CT5gbMXQL8XRJYHcxAFkqICdY055jWgjaqop5YAjeJHl2DM4qjBIUglWFCaLf4USpITAerni2fffYf34CS6GPUcAUIqRO+MyMU+0UG13Y0D3fpkx4XL/94vnIZLpoPJZ6yYyX2xz+GBlzWS13eCl3uF6P6l6+swqk2w7ZbhIjKExaTJCsidhGAIfPV3xww/XDP3eJeaAKARYXntWyxUvL0a+8rlDDmcWazXeiSMoPTrB7/H1c/S33gSGIeBc2IP5IKXI+VXgeuWnwOv4wLJZe/qsbFpVZcllBKAwbQwsZoabrRSZL64cKlnOzgacF8e1mJk9J/jj0XUqXqBcKIXZFYPAx3WtGBKAqOhaLVYELcFfZys2N9KDMbqEHQaaToKCiHT6K0Zu1mu2g8N76TcCSEpNxfiUkvQWabB7M1peFTH7yY5PvSNwIaCiTPypUmIYR4bthgd3junmM8EnS1OZyjhdDHm0W8rsFojR410UnfMkPw95YxC8pGlRII8UXO4iFEcUQ8jnRLD+Ykz8KDS6DFyakvplDFhpg8qYLhn3ZpKrQHByo0le2DREJ0JtMdchHHjfYoNjs9wK66aqSSGgtKEkqdvgWT6dkfoeHT0pWGKe36xTlEEvOrf9qwqSRGteV4wJ3CgDSOaLE5IRlhZakZQlpMi333vMsVbcOTnIM3o12gh0U6AObc10/8mU2xQypVMLlm+rhsNFRzvrsLpi8AGjDYtb3W5Gc5RpaMU4m1qjY0RHKaiHBF074437D7h3+xamtrgQePToKWcvLxn6ntPDAyEU5HqLc26C5QoNT6TAM86rRE+qyBsXZ1709cn1jRAVygs8N465MSjJCMbDo7k0W4WdPMTZi3NuXl5Stw2rzRZdWzqYMsoS5a5XW/Ybl5Q2uLBXG5oyAp0bAmXKnrCZVI4GpXA6+kRtVB4G0+PdKPBjeU8t95BSI4sJFXbG3+e5zmIE9X4ZbIL3Ukz0o+eDD1f88OMNITI5SJ0j0zhlPHK/Ls8c76UVX/nCIfMWxjGy6SUbmc81xwuD1olhSJxdCFtrGCM+7DKTlGC19vzoo8jltWf04hhun9a0leZiPRBComk1n/9MR1dLEfnRGRQ9spgUJ8c1L688g4u8PBs4fzmQksJaxe3TCqME7prWx96xV+Z49eeQWWaRUMt+ruoS7EHKDkzOkZ1DZgganWj1yFYhA6ZiIrk4weIpCOEjealVTQFmPpexWfRxby38tMen2hGkHFWiSrrm2F5dMFxdsbm5zBGI3ChgagRLyaNIwhDI+DQhSPdwFCNKDHmRBnlNNtZaG1KUwRIphFwYzl2wuqTXEYxBuQFsBUEw0VjYMVoMftLSJKWMlfc1dtIuQWlp9ElGpjQlI1kCshEj0g+QYsTWhjClhTu4iYyR+xAZciOStjVBGYlTZOWTlGRLqq7Q1RxlDWhLHQNPz8755rfe4/5ixi/+EyccdC0HhwckFHEcUUkatEwEWy+wOs9PJRfMsvGXucZk3FZSXGUMs65j3s1o6hqMpdGGlCQ7kr4pJZK/rpC7E7oIlvmYczLJxhYHHU3X4n3AuUGKezFxdX3DQddR37eMo6MfpEdDW3EuL8+v6Dd9Nt5i2WSuq6wFnYua+zpOKYrU945sQA44SvE/1z0yTLRd97nuU6ARMVwbozlpW5bbMeO7O8uqMggtHdBF3RL8OHJ0eiQZbe6E3gm1qT14SNb/JEGgFMMIRwf5HgVysOMzREeufySMkmKqDGDKDgUx5qIAqqbmQyiOSAKg9cbz3XdvOHs5SLBhFLNFy6xr2W56cY5tQwyBm5t1boxKXF6NPHvR8/q9ln4bGUYxaq/dbnhwRyaJ3Sw9l9drRi+som0fMp1b7txq6VktZaUYC7duVdw+lr6W5UoglLbVtLXKMtkJW6I6JfMBFq3mtTsVT18OeBkBQN3CnVsVpwcmy0AkivLrj1klSoKxY0HJOWYHc9qmYfAKZwxdW6P9yGY90M1qsUcp4MdQOjmIIRESDOs1MXi8N5ycHKKY431iuRmn+SjlnWQ+s8b7POMjSkdymW3+jwIOfaodASnhRofSmtEYrK/w1xeooSeGfmps0Umi7UIk1YoJAyYCxkDwKGNQmWstJRsxaCJTYdAxEo04FmUySKpLNCIbaRKcK2HB9A81bfF86Tl8yNFVXsnFmKts2HWBJ9ReTSSfQKtssLSZHIBOEqBiJNIAGENkO4xYU2GbBaqpMKomgQzKcAMoRQxiCG2IeDcQ04AbejbbAF2g1hVVI2Mfj+7eozmdY9oFddNiY6LqjoUDrUTHqGkacbxa09YNtbXUtSVpgxvzbGQUy/XA9XILiBFKSWc11jwTgiSwWvAye8BFRh/wwZGQjeJj5MtfeIvjwwXL5ZrtesPNekvfDwzDyHrbsx1Ei3+96fFOsrxXG4PUK5tbiAUGcgARMxEhFWGYJNG3UbuGP2s0r712i6urKy5etjuSQCr04TRh6GWehTGGu/duCXwQQ46+4xQ1lw5Suabc/OQch7dOJaAZhtxBn9fWZKCZ1p1SknAODsZB+lC2LnK8qChdzGXNg1CRYyZQqFx4FwG/fG2xwEDqFacXU+LJs4Gzl+OUCRwdH3D/4R0IiU9Wz1lvthyenHBycsD88pLnT85xSbqoz857To4s241k5NYq2kbLfPDc5W9NLvqmyJALynU1oTt50Ivm6NAyb4UavN1GokocLhRHC4G+AlJH7LodhBN8xNvI8aGhaxuGMWV5dkOdG8mCVxirMnxbgn817esCB+7SgjQ9g6apCDEw+CizhvO8YaG8y+eLKYLK5I+sDBuSjIsd+khUB1htaedzZqPn+uI6Z2gigmlK9m3EPhZbp14JEH6641PtCEIIrFdrtNHYlOgOu/xgpFgqFMQiDCfHNEwiR8vS5JQhn4KzkQ36VDADktopYyY5d3md/NEuEihmf5K9Ru2G1aedE4jl9yUgKfrS8rbyeylfUqKuqHX+uwSpPHBDNEI7DFqTtM5SxtKjoKPmJtQo6wnGonzE6AGix/lN1oKRqV1aR3TTMQaBDIytaBuZ3+yc487JLf4Hf+pPkP57f471asl3f+3X+PmvV1RJmvuqCj73uQecHh5z57XXcjNPz+VyyeXVFZv1wLbv8U6KbfsNVlMUXCLM/OSEdrh38xRUbUNQiuXNivVqQ3CeTz58xOnJIcPoJfIffR7aEzMcVx5TeQg/tolh2igFalEZjw7ZQAuGnvtEUAKrZe9fnNJ6u6Wbz7hbN7hRhrHHGPJz3jHJylpJMWLMDGNs1uzXe8tMioOb7QApYow88xQjYdjSHRzStB3bzWrSSdK1zew3cQJF8VUBg0v0Y5L79WzJFz9ziFIgbRrl/Nn9KiQbQBySz8PmQ5T9VfpkUIXoIF28Z+ejMMFQVFnXxo2Brq2589oJzabl6OQQYzSzWcdi0XF9vSYmYQqtNx4N3L9bUVeKpgLvC5wKb75ek9nbVFXi7t2KO6nKWP8OGgOpLRRY+P5rzeS0ShYBibaxuLx5Y5L3AmlKE0axmorWLsWcNYm4Y0qZcq5266jU//a2PColVlcrVlcrqcG1Hav1mmrc0hwGtMtBocoMp0rTNIZYKU4XiqbxvFg7kaEwI1bDaOH45JB+uWYYxymDtHVF3VQQI/02O73M8tK2oukaXj75yWxsOT7VjiAl6LdbjLU4IwsuZdyyRM0kYf4oveMwx7wR0TpP7Cl7ORuHUvnPD3QyGTlan0pMSospy+FWUirLGmSej9EUpbcSacbcQJZiAmVkAIfWRCVF0pQQSMjKzF6tNSFqFIaUgigVskuFSx/FduvQKWBTysMpcm1CabRN3Cy3tDGSho1cb5WLg8aiVC2SzBgSGpe7lGNUtF3Ha/fvcDpr0e2Ms2fPeGvzedkIfkSbjjtHpyRlCMAwDnz1i19kuVrzvXfeZ7PeMoxeGqFyVCn4vOD/UwQ+dRmVWx0no907x6bvqWzNYtFRNS3rzYaXZ+eMW0dVW2Zdw2YjDArp+RCmkmxis+s/KCGBUrt1QvlxCQZ2DYU6946UIpsudaa0H1ztxElSilxdrsnpwgQRKHSGu4RZZnL6WGQflNITb19NhlWMrLFSpB4HmSBnlKLpOqxWjJsVB8cndKeneC9T3koWEUpNC+mTkMEtiqfnAT841uvAk+c91ipmrQyXb2pLbUXMT+xKlpwoBeKSDUxZ1A77SEmxXHvWWynSosFWlpvLFTEm7t2/jU+BO3eOSTEwuMCzJ+c0raWqLMPo8S6xHSLHC0PTCLkjxYSbMuZEU2UCQ26UMibtGeICYZXnl1lpSaNVeUrsArts5L0sCclwihPN7+eCOIYU5XuZ3wCjV4RYTWtgMkrTetn9JqUoM8W1QUWhGi+fDagUid4wa23OdgqOr1gsapZDhr6DOOukIjpsCMOABhazxMlJy2YcBH41mhgCwSuatuGoa0gp4UfPOI5U3Yy3PvcZPnrn/T/Aur56fKodQYkQFcKjpxhXQJStIkXYQFNSo/KwBf+MSoyGFFfzxtV6ZyCyUREtlZiLzPJ6MRxiERIhaxblhqMgTkYlieo0IT+kiExEFKNOiAS00L2in+oUSWtS8CStiD4QdS5K+lLElvQ9aakWuNFhlGCKiSTizCmBlsLS9fWa+fFMzqsNAdGq8Rhsrh0YW2FMRd3U4AZSMtw/nfGZz36JfpCmqtrCJx9/xNAvWS63bIYtfd9zvVpycbNktdowjG6CXFKCpGQ6EwpMyowV2IvS02RY9lnYUSVsVdFYS8xI3NXNBueu6TcbFHBwMKNrG+kvyAwnH4sT3xUzdzTHssnL++xfQzHuu+S+YKpSSy0aVK+swCkTFFbkXpd42jeY8r0PMkIUtTuTrMuE8rlHJUNRhdkmvQwGbUWuJPjA6AJeaRSR5fWVTHXTZpphEVJEm+JUdtAWSaCh7ZBYbwPPXo5ThhNytlNZsFbE3JoW7p4YmkoYcaWfQWb87OPNIglxde0YndxrKRDLLUu52/3y/IrNasPxyRHDZmS93jJbnNA0NaPzJKQQC9IpPY2iQGdnk6tCe9l7KTiXjK+MqS39OzFJcVlqNOXnu69oS3ssn6fvA/11j49CTIj53DFJ6S/GndZwjFDP7G5NlLWToYIpsFRF/6tYpIiOHt+LLMTlueM6N4TajO/ryuY5IZblNtJUmnreMbjAuodKSy2gspGT0zmbURxGytmzzC4WIkjV1OhZh60sTdvSNX/IOottVXHrtbtUtqKupSUwaSODyDNEorSW6js5NCt9BRmKKdRNlaLg/mLFSXnikuwQD0ZLbQCVJSck9k8hSct4npAm+Kya5gszGZxdlFByDK0gs3wnWAGQBRjloaYgzUWElGGkTD/0kaACgcDgEyFJJEDZoJndmpDFd711vHm7IlKRqhrqGWqm6WxLXSmUkm5eySLA2Batayo7YxgCQ6zoY+Dl2Tk/+PAT1pstqxyBxyjOqXTDCl4t6bnICEBRfN3dhzQVVkvkH1PC1pZm1hEiLJcbrq9WjP0gIndKTQZyMZ/nwrOR7CUAE6Aixw4GyiBTifqnrSxGRCHOvxh92bgFWCn/X85ZDKHae7by/Y5JmJjQu1fSjnKNau97XrmukCFNOcL0uqLSpJVB5ZGS8huDi4pxEKpySmU17TKfksEIHCrXo+sZ3YHi6vKS6PI+yEXncp/L3Xx5rHjzXoPRMAzCqtKFvjplUpIxXF67qZ6BUhwezXnwxl3q2nJzc0NV17RdDSqxOJlzzyQqW01GUWXHt9lK9sXOr1J61lKO+XJNXmopKUm0rXPfR6aC+5AY+pFxlM7wEHc015REakWbwOszDynixsTVuUPVijt3TujaGbYy1FWV2VLiYOumASJPn1xwc7PZex6llycjAVrUS5XSkzpuvjXymDOcExPgMwORgNr6nE0qXD+yzoZfGYNzUYI3rTFGnpmpa7E90hCDtZaijpCS2MrZfE7T/vROAD7ljqCqKu7eu5ejKEhRcLIUfB5OE6a0TOWUL8U4sXRi7vBURsmkvozRJcg4bI4qywZNKhf/1CT8VtDlACINUahyAKiM8xejoSEFVJKfh5T/LlNWpaYt1C+lxQnI50r4rMgp4wDlNQJ0iOCZk0RDso9s51JUkxEkJHRVkZoZdn6EVhEbApVS1PWMmCpRS63E/ITg2Y6O9dVLrpZLzi9v2G56nPcCERSDk9PznWnN0Vm+J0VIT6LtzHSaojYxV8Yo2oM5ylgub5ZcPj/Dj1Ic9M5J70ddM9EkdaZz7uwQJTv88VqPOJzsCMpPsy0uTJgCBVCuP39XPiOoqRCbiwb55LsPvnP3qgSE0/l2YXO5S3sZxyu/lyVSJqmWesi0fpSafjZde05HdpkPqF3havoZ+Z6rfHFKGRYHFe2sY9huGfqetrKimxU829XIdtsLFXSj2G5E8kMoqyo3zzFBX0Xh9GZdGC/ynsZaaulYo64rTFfnDDwSfeBg0aGV5ioEqsZy++4RJwcLGXZvDd2sYTabY7TBh0DbFdVaea5GaVx0hJQ4PjymqqrMBhI4NkWZOrdabVguV2y3IhY4jjKc6NEnz9msh12jHJJFzGcd88UClYRZ5l0/SX4AqPUgHc9h72cU5y+9C00rmlj7I1kLo0fMgWI2awijFzpvzlJKYqkSE3vI+0hy0gDXl/Wzp9GldG4gVHqixielBS1RMsM5oIgpUHd/yPoIppSezOaImQI6RZ4qbwCyXpB8r6efFTOeN1ASFoQqDiGBzkaamCZjnXI0AlqasbTAETpz0UujT3ECZf+Wzr8spkCM4EJugkEYC9qkLJ7qiT4XssPedezVL0JKNFaKWoOPMvSb3AKmShorGGmrDKZp0aZC61boopWGyuKsZrsdWPU9VzfXXC+XrNcbhsHlxiHB9iFHy4XNkxCoYIo2yfRXtfd84t6/E1plmQ8l90dVitnBgsEFzl++ZLva5EKn1A5E/2cPplE7w6211GFSFC0oyI5QS0Q/9EM20jvoqRTUSrF3F8lJd3nKuy/ldKoMJNrP+l+x9HHPxSR56lPBmbIWdqwRseKFC7a3ltlBLUWTZpc5qt05lWx4Y21W3US61/PUOGlq0zlbYlp/WmupNajyNcOleqfBZZVkj2PwfPzDR/zGb3yb7VZqDjfL3LjkM6xVPguinfPg9bvcnN8IQSHf65gS69WWqtK8PLsG4I03H6IUrNdbVpsNx8cHBO8Z+pG792/xp/7UL9PWDTmimgwoKRGSKKwWyqtz0mCpnPRFnJ2dMw6jKO0OA9vtyDiOuNExDI5xGHB5KFHIQpMppjxulhJLkBJcPL/m8sXN72F1dhG90prD44NXn2T2ugrh/BdHvr8OYsxP1ifW6yHDx3v3NOU1mLKh1xptd2tgyiBTaRxNJB+nYHG6vuk9M6X06oZLrf/wyVB7N/LiyRPZ6DHSGngwq5gfvc5aaZ4+e8GQB32nKIPmUxTsOYXM2ohAkH0dIpQHkZLKEI3giFoJ/ILSU9ORVbnekKNcH0uvguDzkexUEN0YhbAndEokF1BWjL9S+bUx4EOGGShMhZRTx4Q0Yqld80/ekcZamvmMSitsZbFGiRaPNVhb0XYtXdtQHd+mVzXbfs0YItc3a9brPlMsB2ltDyVa2EW3O+cTKYwWOYoEgXggrRVonWU9clE7X2OJUsRyiQBbjI7VZsXlzVJuHoba1llyW1FVgm+SFLYyuV1esiGUoqorbGXZZDmJ0QlVDkQ5kkIRJsff0+bIP9sLtHfHLlpPahfJKbnhuzh+/4/U7j32HVbZjOK0yCJ1gvvrLMamsyG2RprBtDHCHjKWygjWq7Wd2Dwy3N3uYDJUxuvTKx9GoV4VeGPHgJIsUu6xwC+7NYeCTsGXv/oFNmHDd7/1Lv3G0ffu1XuTz6trw2e//CZ/7Ge/zm/++rf55MmzTIGVFy2vV4TgMzlAhNqapiaGhsePnjMODhMVo/coo/n44ydE5xhdYBhHnHNizAfP4Bxu9HjnRGMpiLxH6UOYRpTuBUs7eHD3fPfKNpTlLWtb7oXad0B7z/jHfrL7xe91PsQR7vK2vV2jskhBKaCovHfKc2H3nCB3iAfEGeSahcpS9kkhtHcF03jeDAdNY3LzPQlBmiAZx9/zk/zDjk+5I/CcPXvGODgOasvD2wuOFzUHh3Nuv3aPv+E1Tz78SCLxmBiTcKKJUaLtGIkhYauY6WgJmeYT0ZUhuig9BD5hskyxyTMGRC89Zp0aeY1XkRQixubBNKWwpLUMTcn1XmsSSVmRLKjEuexLXxhrckQnkV9lRRmzbRuqylLVFU1VYyz84L2PuXVwyM/8d74kkaE1GR+U9FkbM80HGLzj+fWS86slF5dL+u0gEd4ezDEVAPOqLwJ7O2wzTAZOohvZ9DIRjDyzQLowjbU0bY2pLD7Acrnm6vKcMh7SakuBzFISMLiqTI7QZECI9lHkxdc93rkpfdZas930FPnoCXHZfRRhbeUfvGqg1W5jq5K1FWteAoEyd1c+ayniiYKjwdisI2WyETdK5hFkhccqM32sNaJUmRu+DK+KHiqtpF5VMg+lczZQrknS+V3mIdSHMrcg41ckZRDR+skroaLJFidNmWpSCZWnngnfYWegYo5mC8b/M1/5MocHC77z7Xd5+exylw1n+2Jqw+e+/JAvf/5Ntpst99+4i240z5++5PpyJXvUB/p+5Oh4ga1Eurrf9jx79pLF8QKD4vLlNSEEHn/4jCcfPZvqKpPxTHtw349bcfnp7zLuv/+hpkypPG41iTVKJlrVlqR24pAF2tHl+5J9GdmX41D0l34sxyvXlOt20zWmvTrV/ucq63byDyo/Wz3d9/IZ4iSGmNAhYIwES1XV0G9HqbdZnftBdqnOq07xJz8+1Y6g3KyUBA5ZXi159GhLFeGPf+ma1o3ToIqEyAZIR6CwhZRSYAwhaant5JMqo4XJY2p8koYolSR6CzlNn+5lJa3myWb5YC0YuxgIkxUyLbUVI11VFVVVUVeWpjHUdUNd1djK0NUVtqrQVqJGWxpO9tN9JTSzCPgY+ORsCSi6WUNKgaACLjl659lsVoSgGXvParVludrSD57gyvCTnGqiJkb/VGMoGYmVr0UWI+VicGlbl85RnZlBmoqq5KX4EFldLiUlH4X6qBUydANxHPvxkgKG0UMqAm2KoR/3DPbuKGqvSglOPL1G7TYtpa1e7eAVrbV0zBox0KIBVE3O1hgrg4OA0TsM0rXcNg1VVeWCpNlzJDunqORmTpitBHn5hVMtJe2gxbzhsy2fPqLYqZ0RNPk5hCjBS3FUEsEiGWimSKbSQ6MyvBV3wooJmd88jiPjIL0W4zjmyNszDI5hGBkHIQEMw8A4Orbbfuc8SwCcJKv++L2nfPzeswyrFHhwZ2lSSgy94+L8hm7eEnzMjtQyDCPXN2uRb0mJ5P0Evb2yKn7faLwYyx3E90owk9fEfjaoSxZmVN5bGlNp6soSXKJqKt76wgOM1tR1Q9NUVHVFXVfUVZWz0IqmqtAa3vveB7z7w0cUymp5ZpKE7Zz5brcxfZhUNLemm8UEuU7n2buP8hn0dIoieQISdIYx4kYJjlICFwOmsigFXddIs1yWJvlpj0+1I0gp0Q8jMSYGF6TZyidcxnabpmNQ1asYrjb5IVmmecSRaTFN2VSAMkllSu8jE0ZdWBPCWJJofdY1zGcdbVszm7U0TUtjTcYKcyciYGyFtZbK2ldoePlqCCmrmEwFwjw0JQQqo/AJRieNYJ9/eJu21iz75QRLXdxc8fL8is3Gs115kSkm8xlSMZ5Cs1UZ7hCJbpUhLjE0xmpiSChjidEToxeHaCu0FcnG4IMoi6LABfqhp+/XE8PKmhprLVpZmroMl9mt8f1ILhWDyf6m1XtGSO8Z2V2dQCa+5azKGKopixLnWxubI3hxpvJfhvAoEV+OwovVKZH6lE2kyZgnCuSnpnVY9nupWxETQXhT0/rbNxRT4VGTZT7kj6V9ME0QoFyKfLUKUp6tW2kZT+i9Z/TDNFxpGB3bfsQNA/0gbJlhGOlHh88Gfxw9zuV1EUP+W4EUYhZujBPMkHJtR5H0ZMGm2okbPRMhofy6RNuUBEt47Cu3Zn2zmW5UwbfLa8l1i2K02ctOd5ncDnLbZ5Fpo6YMrKoEFq1qS4oyszkBdVfz4MEdTg4WVLWltvUkZf7Jxy/45NEz6sbyS7/081KnyM9PWbsXaOQ1IbAA8+7FtD7LDUjlLqhcTd8z7iov9LT70LmHafLe0xnK2io/kaA37jJ4dh3kJeBJlH1C3p/yNFaun85p7a4T/ic9PuWOgCmqDEkRKJV0jWlaaq1RpmLSftn32pRnJOa3/CCryky687YytE2djX1FXdV0XU3T1DSVpWlq5l1H03YTW0EYPSkPO1FU2sq8BF1NjIvS7FM2nMpqoHiPJ1BZLdBVFgszeQiKNqI5VFsLyjK7O6M2lq1LVFbhIpAG2iqSGk+KG2nU8YFx7LHGkLKwmMmFI4n4xUiL9pL0SJDIUtaR4GUGbkpRBscoNV17DEHOq8ReqqjQqpYFp/Q0gKek+HsZ825jZZZHSb+NVlR1I92dpmRTxchbKqP3oJesx6NkfqvRJbfOhmJv402GHqZaRHEC+dGzo0Wq3b7MQnmiASUbMBY9kQgQMsGgGEEJ+3dFajVF+UI3FQaPBBxiLESiQeofZfyimkCExDD2XF+v2GwH+n5kvR1yYXTEh32ufMqSBX7vZ6WwWGTay2yw/P6CT6HQoLPgXYzZLxfsnVeghSkLYu+bHLyUZszdZi33P0e1xZib8szzOjd6D3oRCK7ae/51U1NXJZASY9+0LU1Ti4SJrWTer5I941PKGQ3UdU3bNFNwkfLnjTEyb5fT/jW2wlQ1xCKdIfBtYebILRF1Uumc3j2rKVvag9B2pnz3rPMpxHhPAdB+JeHVWyf3Lz+DEh/uflE8zC4QUWQyTCEqlEUP9g/b8HpgEjND5U5fpQkx0W8Gbt1/jap+IYO5dysVKGyLArkIzlvXlqapmM1aFrOWtmt3hijjwVpLtNm2LUZnITidm6NiYrPd5oWnWSxaKlszuEGYAqnPkYyoOuokUas2NrfFg1eakAL9uGbezql0jUczlJU1DIw+EKMT0Trn0ERcjPg8IWu1WYuV1YLZ99s+p/gbbGMLJw1r62nOsi9wUQ7jgg8ZZ5RFo5HRf+UwpiZ4JzxsVbKrnCVVFfBq+lk2u0iCy3/WGCqrqSpL19bUdU2Vi9xaSeRSJKlLt6VSCpWKQ88boBhboV/JWqAYor3NpYSNIwwjTQheZBkyRCDFvQSx6MDoXdRfTpFifs9SaGWH+U/4vJo6mXeF0+IUJJIr2UzSatq0CohKYDqV4R9hLUvE8/j5GZ88OpPaTi6QToNscoQ6Gerdvs9r10zRZIwB7zXR7IqLpDTVCMSQleizMK4y+6REsSUr06VhLe/FHLRMhryytLXNRl2eZ11ZrBViQ2WrbPjNxHqy1mKNnQy21LnKXt/d47KPi9Of3PkuFcGimLXzLJeRWYM5AtHkzB5Dk2U5QHFweMDhfI7VEWMaqf/YPN2LwsyRORjPP3w+zZ6YUsLp+N2GffdrOce+s5hcfipOOd9r0iR3I8dOMmT3ZPacbj79K05b7RbFjhDykx//WI5AKXUM/HvAz8il8T8BfgD8x8BngA+B/35K6VLJrv2LwJ8BNsC/nFL6rT/g/DsPrRRRwGB0Al3XdF1L3VgUYlya2tC2DU1d0dSWrquoqiYXfYzMLsjTfCptwWiC95ThNil7fR8CwzBgrSGFiHMjo/cyD0Ab0aMxIvswjJH5vEZVBktE61qi5AhhdGzHkTEMqDTiwshyu2U7bFgvV8So0UmGqbgI6+VIHD1VDVWtIUkjUd3YHGGKRMXWjZA8ShnGMUizVYqyoMkFZK3punradIp877LBLmuoTHrzWQpBmsYipCh1jrraFbaNFLrtVFTVE7xjcvQum91kCuOOZSW9F/rHnm/Z2tkYKzJ8IcZ6R+0UCQWVDZdAoEXvJavDKunm3o5OlE6TYeg3QgCwZso+pDpgpshKpSQwUbmKjBClbKyliLsfZMi6FNJMmXGQWRxZPqJ8liJdoUrRIEHSWthrqvRZpClDeXjnLneOj3M20LPJ/w39KJ3GXoaZTJu/vD9SOKyrljLERYLynZlSWuYlGG2yzIEIAIoFyk1UlWDl1lbZUapdFK8y6cGoKepXxfkXSKfYJaWmIrb0g2QjlRB8e8rO5P1VqYNk1lnKMJSiPBuT+ZflPZnwc2u01N2MRRs1ORtZjxqrFMEnrp9fyHOKkaapOVp0OduNGf70YKWRrKpz0KAUTVO9mvnAHhi4ywpeyZj23YMqT+GV/GoK+3fiJeX1eU8UQ0+JhdLeXtnRpac/S1O+kgPjn+74x80I/iLw/0op/QWlVA3MgP8l8NdTSv9bpdS/CfybwL8B/HeBL+b/fgn4d/LX3/dQ5KhRI7BBrSVijo7ri2vuvnbEH/nqm7RVTV1XVFnmoDy4mGQQRIjSqFGKXVZJkRclTiVZ2VAhBtZDL8wgI6ycWBZLjs68D4xOJgPpyuYpQdKkE5yDJHzfiMK7EWuBSgwiOrDZDKSo6DrR83djRPcwn3c0s8TWB7wDo8HWNfNZQ91YFgdziGrqrkxJ7aJrrSeZAcrmtHoaxlMiZ9jbwHkJ+uAZ+p71esO274Wiagx1Xe/msqb9CHxXnFV6NzKzRPOvRE/8HvDClEG/+jpjLdoqxsERvcNocDFmDRadjba8Vud8e4qoCmMnQmsTVc7ArJ3tYCCVZ7mqLFIGRJ93cSzUyj1oZ//ic4SoMllA5VqSIkMQ6Cw5JedISYmd11CYPiIRXK5BsgSpXWUJFJUw1tIaw6ybcXpSYCtxzDIMycvayt3yg3dst0MebSky0DKXIDtfRe49yQ2CSRx/UjtiQIgCL1krDr/c3wI3lOBbbPSeUVNa2Hlp3+HI5yrwH6nEsio3gEZi0vn57e55QvZpyhTLiDB/ZDyrxtYVTWUyw072jYyXFf/QdBZrTC6MOxSKcRhYr5wEfBoKNJYylNTPZpDKfF+Bq3zS+CFgsoYVaedw9hGHaab95AVkj+zZ6d3yKcXiEmW8cuzqUK8ee5nwnpdJBT76vV6fnxNJ6no/7fGP7AiUUkfAnwD+ZbnINAKjUurPA38yv+yvAL+COII/D/xHSSzHN5VSx0qp+ymlp7/fe7RtzR/56pt0bUvTVAw3N7xYrRl72PYjc1tzdzbPm4kpMiyGbiJuTCmosC+C89xsNris6LfcbDm/uGGz7nHjOEXHRU9HZY68NZrRSedtZS3NvCElRcij/4QLvjOOpojSKXCjx1jDvFugsHRtS9dWVHWDVobFrMNWlTSp5e7l3caUbCUVqd6y2PYMvM6wmUKhrM74OFKoDCIsprMQmguimOn8KDMXErSVZd4cUDoXp8VFhhEomjCvLIIp3Z7E5VIiTf2S8v+pRDelGA8SmedFrkBa75MwaLQVXLbWCWXZQSP5iva/7BxUJCaH0TZH6PlelL8rmncFxilOM4exESZHowocWeQPdN7nIjY0GbqY5c91pTLenPleOlvOBCrT1VTBoYu/URpdidMp98n8GOYuHAeNTqLJVH4+GeOUiMVgTE4rG2KbacE58yEJ804Zcu3HEKKTnoVMQdZG5ciS7NT37VrcvQHy+dIeHaoYe3EEu+FNUx9DeWb5Pqm8sIuOUFQivle1FV0jWb22GqOEhWbzWE7nAzGJrIy18myMVhB9Zn+1oGQ9uyCd7jEllKnleaDwySDKvrsgYi93Ivj86FXRU9pBUuVWp33A5tXHxgT77EX1JSPaP353DeCV31L2xuRq916zV7F4ZR+l39NR/MHHP05G8FngDPgPlFI/C/wm8K8Br+0Z92fAa/n7h8Ane3//KP/s93UEVWW5d/e2rLuYSE1NGTUZYsL1A9E2oj+UJHoA2QjGyoCZYfT0/cj1eoMfRjb9wHq9Zb3aMOZhIYUOJxgyssmz4e9mDaYxDL0jJcVs0QlFsappuirTFNX/t70/D5osy+7DsN+5972Xmd9ae3f1Mj2DnaAsggMQgEUStEUGt5ABSbQp2g4LouhAOEKSLTscNm06TP4hR4he5JCDDjLgICVSogVRNCUiaDpEioYokSZAzIAzmMFgpve9tq/q2zPzLfce/3GWe/Or6u6pmenuCqBudNeX+fIt993lnN/ZcWlvG6HtgCTpmr34uqJ2QT+Wq6jypiA4kTN/eYYlz1JdX6UKoGrjmU7VGB/Z6rNNyhLyNSVxFRQkXIyITYwIrYm+ssk3jK36rxHvWiNaJGBSxJi9j8IIN/WUct9iuDX9JkPEWGIAuUhtxmSKbrT0gardJwVbGGkSjyfP31DpvSlADcDCXJmAcZSyouImCyUYYjsgYzymsoiEWKFX0g0s7rWTMOBYAupsDbkwr4w8WtCX/i88yMZd9eMe/AV4DTou7r3UiHILWRmoZnE1hgUKiNkS5MGZMLPUVs6qYgEzZm2L2ElumjQJALLFyCAkRfchNIUZyEMrbyCL+DbX4+zrxeZJkuWxRoiL6rJrZyBV68zUXz+ozaFrSPloBjQj6pCKQZuC6NQzA0iS4VV5q4+FOXIAAZEZs/lcDcURbQxoW81fFazGNoqUJiKKSL6NvgfxBcBfwBIgtoi2ieiHVKlteGOvVKTbhEy92hg5VT9U621Tri7bwfaf/2wZtB6fGXw7jKAB8HkA/wYz/xIR/bsQNZA3Zmaix2NRRPQzAH4GAHZ2tzGypsedRtmoqpfOYHRggDK4ESMPU8DZusf9gxOsztZYDwPSOIgYTBGRxDC5NZ9hd2dLRExaACFg1mb0wxqcRU2TofrUJmimvxIhWtzGhKiJoatRtQ00wENTBDtBZ1hREq42G8GIhy4CEn24LzdzFdOdbm52gsSL+ys0UCmA3Ig5TpP6kIuvftM0mqtF3SqJ1etECPdFUFKbqDySVntia12IDBtA9SvL2lVCp8iVyXTTQMmlApWkCES2E7VvSiqDZgYNHpClzEV96ANE+jJXUZhKCFLXYj0MarOIiA2h7wckBpopYtaKfSOgRrLCuJoYPM1FduaVq+dDJLUgSRFDNEJu6Lio4VL2cuheo8I5ha6mAALUOcGdljj5cOachVFlEw2kX+LUIIwwk6RkRiaXwiIBHKJ7nIBISnzpmmuaDrEVojOmhGmQ8p5izJ58rTsAaaPMrBLorLYdK4ADZlBsMe9azOczhKZF03ZomoDZrMFi0ZVrk4xnZiCPGevR1pitG4UVyoREurXEjVnWhHt9iavOYtHh0uXLaJuA9XqJnNfY2l1gZ38LMSSkPGIYM5ikSFNsAgIaDOsBy36FbraFYb3C/QeHG/p4+6jLo0gVkDxDQNrYM+wXVZYFc8f1PURK922WC6txvKX7wXBa4TVmaK723rfQvh1G8C6Ad5n5l/T7X4cwgjum8iGimwDu6u/vAXixuv4FPbbRmPlnAfwsANy4eZ3DbBuRAro2Ylz3OF68h7gekNqI+aV9fNd3fy9yDLIPmTBOEdOo9UPTCIa42S2HHnmafCOOaQJzAucgyEALzrMh7KoUmXtvQFMH1Lpg77lGBybzAw4G0AAERWgMRlIOXlB8IItKVGIfimjpagq5xNNOALKeLAIxcwImqc87jANSEt//JkZszRdOyDMXrwJD7y5VqLbGMkXovlNmFVAElEpV44uzoEAh6L58ZUM4cRS1DRSFMcnxZFW3VN0gXlZK4FFoJTOXMQgZnBICdar2Qwm7V+YIFjffGBpJQqXvtrPTSNBhhWA95w+zj5dco9eZ/ywbcbclwpr/iYEsDCubAZts/ITxidTHGwnHZK4DQJPrxn2sDU6ySRnw9CnCB8vvjsYbmTgKcAcBzkBiS9Yor5SDuhMDwsi0N13bog0R0zgImJhGBAqqsxeYTKOmTskMUFQX6AigQdsAbTNDt5hjez6X3DcaTxMCSSQ2JDNo1j3n3lFZE6rZGMI2kWTeDCpVmc2cQepaK/tzHNeITYPl+SnefOtdnJ6eYLmU/FY//s/8MJ65dhUNCMN6wun5EQ6PD9APCbOmw/K8x9HhOe4/OMaqn7BarbBerjHlydeDBKqhAjky/1k9GQ0k2rogFLrv4OSCm6jPn0oBfnq9wRxA1EzJ1o6LFnr2RUj30e1bZgTMfJuI3iGi72fmbwD4vQC+pv//NIB/W//+Tb3k5wH860T0cxAj8fGH2QcAYNZ1+J7vfgEUApoQ0PcDMPZYnZ7h2Weu43f80X8eX377Pt78+leRhh5bswhqWnBi5DRgmiQH+zRKGURJoKZudloTGGBXBRPpUTJ9P4DMyMHQOnTRisugqDeUiDjHJzeGackPF6MBEbPlk11HkKI6NrBFQnR0C7hXwAQGcZYMi9OA89MzMIvKQ3S8wqxmXSdP8QRy6kZYo3wA1QNgPnhq1qjETNm4bMzRAvU8xFUXsBN8+FiYOcwQHpJtbtYNLgF2dTpnI1wclGFA1BrQd1M/IuSkjEKNn5RQsssyQJT8WWbzMz4htSnUbsKSa4pZVFxBz5NSkgB41E0rA2O/s9kmsjFUMUQbA/INbQuMdVyUIDgtIQJYalW4cdHmhss8UGLRqxco6fEZ2YCEqaYAiWEIgJdtVTjJOSE0onePFfMTW1tSJhLE06qbYQ4W9O7zyi4FNyphiroqI7YSld3NFmg7CdSaEiTanQizNmBgLUE6GSVU9VKSdSSxEBLINqUJ1ET0/RrDcAZCh52tLUmNPq5wtlqBELDqR+QkubSaplGPtwbPXNlDc+0KiAh9v8bt925jtZJ4jePTU5yeLXF2vpKo3HFyN3FR30qKGVEpBYQopVmjpRThEv0u6xLOwMUltIAKm7CC9YGLyhJ2cEQKSut71bJC+VCC+wyckfOnx2nfrtfQvwHgr6rH0OsA/jhkVv8aEf0JAG8B+KN67t+GuI6+CnEf/eMfdfMQAxZb2+L9Ain+fuX5l7A6O8HO3hV0i32M61tYn5wjUMb5SGCS4tlQpJSZ1WfeEjVl5+6ueGDA0i03QfS1QZE8IoNMd80liEeainRkoFEIcTKmQABBNhvbMUZZGBcmM7N4hIRkaYADYmywmLcgajBm8fBZrXus+x79eo0ISd4WY+M6WcvUCdKIYhTiQSrJbCwW248EmPyZDbJ4RxXhsyKRanH7PR3ZlJezs9zQTPVyNkaijIohro1EhbgqI7M+ilFSajO0bSuqEuiYEns6ciN4qqUrNgxYVk7Z8FntAJJ0U9Us2kdigFXXD7bst3APsqILF0pvhl8yiUINjabr5mp8y66WdWVeUaaJpoYQ2zkiEcaxVz2IjqdGRbMxEUvBgYw2RoBFFy5LOzv6BEsVM47wvFdse0UiBcEUxFmBRDIgRIQm+ruapCOealFiVIYVxql3tZhUB5P/AzPW615iVJiw7id0baNqrCQ69TxqX7KmZwmYouaq0rghTAEn6wGHhw/QkERLy1QHdE2LWddie2uODGBcDzg+PMHp2TmWyx6nZ0sMU5LMpcPoVfRMveZg5aIKSFW65j6da+aua8SuCTGIlEfsHlYGaMrekTVa9gWVZxYRwB7uBwzwb7AN0u0DrraU7oGHvDo+un1bjICZvwTgRx7x0+99xLkM4F97nPuP44Rb774PsPjGciA07ba4eo0TUhrFZkAlOVpO2cXvXCEhE8Fk35hRUeOMNWiMGVoyj4GBHR1xRUgd8dqmZlQuk4aktToaG/k30qgBP6pTdWJauSMGihIrpmqAKU04OV1LCoE0ahoLSSlMs9kF5l88HITGGOUwSkBl4XO98C3vuoq/oSJw+p5s+c6opEQoC764HHp6ZfNgqogU6/k+nq6aEsO1JLeTcaBQ/M2D5tqx3D6BCO1sJuqgrMXmLUeLMaKglcxI5prMS4l1XEJEVMbJiropEqLpc7m6mdu9SQzxuulM0mOeYIVfTGVEKpoEssRwav/QcTHHARlnG2vyOSAKwDQik9iqREwrXiRBQY6sZ3GWCGgQo94jBGUWwqSyLCYpnUqErrEyosb+TOURHNyIzSMgc5Lkixp7ECAJFYkY/eoUX//G14Em4oXnnsfW1jaIM07P7gmjThntfAfzeYdIYr+KjQR3TaQMK08YRtm/w7DGaj1g3i2wRsI4DOjHAeMEzGZzPHtjjvPlGQ4PD4CQgByxWq6wWg44Plni5GyFYRg0vboWRTK1o6Nm/ct+FBVprvZ5tVZRiD9V1wJC7MexuGy6qlFzdFmgn0iZ2CT6+qzCjDZ/s+3rX+oTbG96AKa07a05Hrc90ZHFKSWcnJwK0iZR14R2G0wB0ziAMjDb3pJi5tMAUewwmKtUyqTTaWoPEPKkG5CT+tBoFTPjwJhg5DRUi6H8XlvqASUjvtgKapBF5otPV4Ax78ys+VQ04IrFGJzSiFU/YLVaeaGQqJ5IgUp5xui0zRYxlNDnql/VMrex0F/NTbRG3vbe4n9f0DZpBDJBXf0oqm3C0L6o74TAZ1VXyP3MNTOQgRg1xoYiaU0pQ9LsBMkMqcxlnCS9cQwERizxElD7CISRB05yP1J7w8hw5ZwZPTQJGBGBteqae0xRAE+SjwkhiO1C1QQhBPFoItIiRwzm5KNuKN6RojJHC9yjEMSekUvfmVk8gNhWhiXSgyN+i+iFejQF1bFnFsNwMEZKcL27rQP3gGNZxZwmFUSE8aUExNYcDwShCx8IxWtKGaVIY0VNR+pRw8x44/W38U++/AqaNuDe0TH+qe/9Hmxv7YKZcHZyhCGdoz8k3Lj8DK5duYbZrAWyZFwlApAyTk7O8c57t3B2doYQMeodDgAAfkVJREFUW8wXM+zubmMx38b2Voe9nV1Mw4iz8zMcnZzi3v0j3L17iKOjY6xWa/T9BIoRbdNuwOac0saONNRfS6z2sQZ6Pj96tLbT2F4urQAvkVZ1bWYGpuRJ8IRmsADLXMbXhtgLpJSd7BI+KkbhHYZJmw+3b6VK2RPNCHLOOK3ypLcxYntbfPannDEOE65e2gPFGdrI2JovJI1B26FVcTbEBvPFAotZC4qMcRxwenaOk+MznJ2cYT0M7j9uhLOe3Jq4lyVVEVe289h/sTwvlnjM9rcRCisoAzBSBlIaAZYI38wZ0yA6aStQEnTFpJzFzz4GRyvkRFs6I27Rhhi017nyh64CygKJGitrygVD50WSsvKGQVVr5GKoF5/nhCGJdBXdjxyglGHbMMOMskKQjZFGVbaLl5W+Q0DRtzI0elnmo2kkVkK5kwMkSVcgCDdPo8CBADHces7+LG6MNq+5eOVAGQ9DBQvW1OMMVQdWdg3LMqoqFEH/7ETWxHOTyDgzQmaRXkxdpC9qsS0uZTKBLdU0kabZZkQCCNGZaQmOyhbA6+kjgsXU6DlWg9ckDRU3XNed1bDv7pNTkvz3MfgarcEBaVBF0zSIAdjav4R2e4FxucLbr76Pg9sH+K7vegHPXLuq89YBeYk3330NL7/xMm5cvYnFbAecB6Qs4xUD4erVS7h25RJSGnByusLtd2/h6PgU6/WEdd9jterRj6Mzf+jcMCQJHYPdZbpG+DYP9Q41lQqMkCrjq1F1kfiK+qfcDxugxja31RRxSSMDCVo6V+czBI04zyoxVLFBzKxld+vOyGdiW6wV4XEAWiQb2VcBj9ueaEYAhnvYyESLlwhIykCOw4Bnrl/BD33+BzGfRexsb2Mxn6OJEYlFzBzGATlNWK0HvH/3Hu7euYfV2QrjNIqHjw6keO4kR+vqlF11Rme7UEKY8q6YdXRBVQgvUI2z5RwiqbWQswRzjeOEpInduq7BfHuhekzZxDFEdU3TGIloxXNkm0YLbtLOE0RtYAnyjNhMU/b02QzGpEg5hqjaC8KUE4gksRtzBCAR146EWNQOowaiEWdNMwAEMmN0UKlB0GcbgtR9lgpAlYFXNpGpoqxEIWn0KSNj4gxkSP7/JiJqCcus6QKiq7EARsIwjW6vYaiRGYyGGiUejIk15iRSmU8DAe6BIWg8qhtADA2oypcDRXiiZmwQiLW+bIfMCednp1j3g6gcs+nWRQLau7yPG1evoGlaIEo+nsyWaBCAxTNozWbKklI6tC0afaej0yO88+5tpGGAp0cCK9ImNNTiyo3L2NnekzTeOYucGwLGccQ49lh0LSxBuVWoy1rLIAHo10usz5eq4rT1JYBlGAnzrsV3vfQCulnAL/3jL+HenQOcHi3x1a+8glvX7uK5555BE1qAJfdWzyPuH76Pppljq+1ALCrJvh9xvlzj5GSF07MlVusBfT9JQKfuPVevoEbtcDJoBl7bikZYa6YAR/QVWPID8DNtrROZE0ch7rLOKunfeL/BewKo8vzL6gVXwBuDIqFtxXEka8JHc6FlMs2E0hILDLUumBRRURRRdZIDm3745FNMfKzN5sB85mMISFndx3jEarVCFxucrVc4vH8MBmGaMsZh8JJ2Q0rIaapy6VSKFNPlwaKByQ2SckatEpLf67S6bgFQX2bT7Tr7UNTmfjqcMWo8xNCLn7aphRbbW5i10cU9dzNUERBMGjykWN2ymoYATnA3QtEVQxFz8IVoQXMpJVhiMUtVMJmqBpbHJxc7QfUekq5Y/MebaHl7HGSCmVzFNXHWilUEClCGFEWtoJ41jtRZJBMBrMqASbO76mK3bJTMohKJTePzIZWdhKE2bafIW3IhpSypmDstj3h2ugRBvMkajhq1HUEMzBdb+N7v/x7s7W6jbWcIMaKNDdpIGn8gBYEsQZm4pUrsACvLIJWeXnvjHfz//uEvIqUeTBmcFYk3Abu7V/BjP/qjWMxnPg7DMHoWWtaN3bYdWpUkckoyturEME4Dfu3VV/ClX/4yVssldCRhOY3yJKUff+C3fC9uXrvq+aaYZI5Pz8/BmTGfz1VKlDWQs0mqCSenx3jzzTfx1tu3sF6tURcxIhDGkdF2AS89/wLC5xm/+OVfxnq9xvHBCnfeO8S928fYvrSF73rpJrpmBh4JZ/cnLNdHWK16TINVJUtes6Daau7lZuvaiKOtOeg+eVQAl6tzVNSzHEc1jrO/jqvtoO9gcolSGEQlGVgf7fpIqhpVO5umE58my9/F3j+JZo4ae5FBUWxjllLan6PIvzglkO/juovGvJQPYb0qKam/2fZEM4IQAp69cQ2Xr1zB/t4eFltbuPvgGHdv3UJmxunZElvLc1y+tI/7aY3l6SmOjs8kVcQ4iV91xT8BYyq2Guyv6cKVsIE3O1LFFMi1Wf2Z1S+wiR7VusGticTjCIxhHDCMA9a9lJGbtS1C13hSPQJhGCZHyEKszIYgeu+gov2UbMFnqYxGQK713+rG6GoBdfOMJKqocUyeNqNrG5gXFeeANgJMrFKCBm8xI2tmW6IgGJKBxJIjP8YI8bSKghyJEKkBBVbvlAYUhTF0s07GLhCIolQ8y8Byfe7pnzmrgZhEOuhig89+9iVcurQPaucAJ+wuttC1EU03Q+aEk7MlFosFuqYFUUY/Dehii6bpkNKIxdYCb717C1/4xS+C8wTOpB5HMm5MAZwY3/u5z+Ly5csgleSMWZIRE5U2AmkOoJwwDQwgASxMJ+WExVyKoSyHJYpHFSMw4f133sYXf3WO565fEcmHAsYhoR8ztroGmSckJjQhehZQtpzGEEIhdaoJV69cwXvLlUi1MkESR5YSDt69i3+w/kf4ge/7LszaVvK2UUCGBIyt1j12d7c8NUpmYN61fv8mRjz7zLMgIrzyyttI4yArNQgjFY+eiNi2uPbMM/gdn/9teO/W+/jS8SvgQdDwycEpvnx4jhCiEz9WycT2GcGqbGVDHhV6NzWLApqK8NeSgb66tkqjbufYcdZjlrrcpB2/gXp5Fef/Cyoj4y01UxBX2KSR1yEQIgeVFuER18xAo1lQLSOsMS1hHjIueZLynCoS+GtIsj3rTDESm1ebHCUsfqMZi7uuxQsv3EQ320ITGwmUGnvxaifCsBYPgVdffgOHD+5KGoXKw8OlflRIg4NzBTd/eYKsDbNraRVfaIhApCltTeFX63UJMAtqPwxYrc4xTUmILzR5XiO5/FNKovcPBIry2Fy5tAEsXh6mTkpJ9X8Bk0bUGCOxnP3mFy5usxPMhpCUkMy6FgBL7QO1F5gHUw4mGSjKMWMnmZqHXSpJnLF3+TKuX7qC2aJD08zQNg22FnO0XSf5YdqI2WyBrm0QW3GFbRv1OycGiHF2doo333ofX/qVrwhyRkn7IP1IePGlz+BHf+x3YHu+QKCg6qtiZD85O8Hu7h4WsxmGYYmz8xME6vDg+D62uw7LoUc4bvHm27fQDytEUPFyMirCjL5f45e/8GV0bQeLll2NpwBnBI7IiZEzYdLo3pQHpDQiZRL9vkpTggAHDEMqQV5E4nqaJoATfv0rX8M3ggRKmTohqy1Hpx6uKzY9vhMqdr1zQcoEqGrTopkzGIf3D/CLv3hY+qHrPMTgUmXTRE9vYjmHAklxI3FDJYzTJG6cLO8kSRoh6dKZ0TSEWZjjM89+FvinO7z8yjs4Oz1TVJzAeRKDdJRCUsTsbs6M5PYYeHxO2cMiOFakV9/d9PoF5T8sHTh0R0HO5RBvMB7BhpUKyJlBTQuUSpg0oQ93jQKbV2ASVSlpNHyU55pXY0bxNtRpcVUr1DnCsxBosKsFK5ImFeSqT5J0Sl4nxAv065toTzQjkEItMwDwtAmE5H7ewzjhypXLmMagdTyzbwITXxOSqDAYbki1HD4m5gLi5ucuXACgKqE6srcwEyGKEQAH9Yxh9jqz0zSh76VoPCBGzr3dhZaJDJqeOCvj0ICqpAY8NY4GCur7LjaACBEfpVqYVLASHqb9t8AiiYDyvDcMQZZN08J8302lUwEucSAk8cxy45WvUOmvjVqIHV586Tn82I/9CC7v7ilKEUmpaaR2gQSIyTvmnIWBaKWsaRowpoxh6vHK2+/hG199GSkNWkUtIBOrIU3mrmkCfv3XX5Z5yIDUtmdwltoN6/UaIXaYUsL52QOklDEMwHpYoSEhLmOSCFQwS2lSY9rCfUU5wBlvvfWGIs+SO0c2rYxTLbabcdj00WaoNY8TszOYqyxYPY8ApGlCQkG9QdGjpPuRYkDTlNAPI4ZhlGRrqkpo24jtrQW2t7ZdaoDZLWxt6b0lHkSrlPGEfj1i3Y9Yr3vNw9VLFbRxwjRNCJEwn88x7zo0AeohpLabKHmYpLiMVQtr0YQGbaeyNzOuX72My5f28PLrb+Pdt+8i94POvzBIMZJLBTZD4aauFBUObLAdLbvuvCL8QcfWzGPGHD0Znh1E2bO1eqeofuE3NVVPdVoFzMr9BCTVlMIWhd6D4Zles17nWVlDUDfl6gFBPMgkFsMepW4F1LhNDFB7g8WuWBQ2jK4xpPDU47UnmhHMZx2uP3MF9+4/wN3bt3F0/xAERtd2yNOAowf3MZ2deRrmAKBkCpSpaanRzSGTRIrobXO6n5AaL80GENQm4eHtDFgpSlLxzwxUiSW1w3K91MRdokPf2Vp44ZcYhNEEsgUo/XOVDwQVmj4+ECFCpAFTw7RRNp357xvar4uSm+eAPDO4btiMlabnLa6JgJT15I10C4DUxk0msgMwj6PFYobr1/Zx6913cYs6zdi6xnpYYxrFt3+ceoxDj2FijJPeaxpE7ZGSIGAG+nHtaQ5YVQMUUIxngfHaa6+JmkvcZ9AEqT3MOaMfegCazM/kZ7bUBVntaMaoVPBnKbZj0psAv6wRtSzI37KPkaWDk7mOGoRlDga+6cS5HETRiymJey072jRIyjq+kl6atfyn1BwYp8nrDFvhoFnXYbGYYWsuFbhms5kyXKlRnFRSy5MY9sdREi2u1yP6UYj+erXGuh+RkpQeTSk7oXFCx6yqqFOXQJz46p4isrVDuvbgDN/OMyDGANq2EyI89KrukJgfzmIzQohFvWEp001tYjTcELgydWOmIrkErdPAVQoUnTrYMfMK0n6Z/eBRTZ9rAY5FeFBgZULaBZuF6xK4ZijmBSS0J4GApEMVVMoLqgpQicKGUgQBDYjVrpHlOoOOIVCM5DonDwWLfpPtiWYEZ+fn+Pt//x9hvVrDKjpFTrj77js4u/8AByEgfv2rWC52BOOqv27ijMBGcHXTQvdjhUJkwcrAJtUFN4GcuMrmyH6O8QVmILMmdRsHrNYrIEtSsVknxVza2GjVqIImY2hKFkmoNKJpFFKSALk0JX9m0zYqJgbVsYoxseHaC8X2IEleHBSjqy8oZC8QYwhW3l3G1PWVpHYFVg/8lJA4KT0M7p10dHSEL3zhS3pPkSDqjWXBVWLXyh4vYQRUWd4mimNWfTwAlvKTFKRwvXn+ZI3+RStJCIdJUoITJZR0F6YrrRm96QOEIXsiv+SDBzO8A9hglEVqEmWLLDNy6cgS0BX7QQJYUkaL1JJ1rjSNQk4YxglnyxUOD09xdHKOXtOhz2Yddne3sLMzx7X5HrpWKsylzFqBrsfxydqLzg+j1CaeTFrIrDYK1nWmI2FqFUXXHkhpTIlRkDfDJVsjnD6vRR8CCyQ0HTc2lkBB1hQIoWnQEjAOKvVppP40TQCSFICKsSTkUzUatOgOs82lLhilztNkNjD25278Mf6vEq5JEsakiOzainJS6QMMVCmjkp9qBuL+PX4H9hWkFxrwM8Zhv0wK2oJKoyK7yR2rLHK2vwEACUhjeZBJtMIIWPfjb8CaxdM4YViuLFcYiESkPr13H+cPjoGuRT8sEbd2dehlSKNGaQrZs9S/XOaQTbdOsCxBkakgBpjzqE4bRSDIZE/TJAhu6LFeDRpwRJh1DbquVR2+GNpEDaTEFlyModoJUrR6tlohTaNWhdLyjUq4zT1S9PKMQd1MpS4yO5HNnBWJBlWBBZU2bGErB9MU3hYvIU3+5jy6isxKUwaIOqNVL52UBo3wFeBGCnEIpAZEuR/ps3JOUiNA4xCielMwAUHtNZvqESH+go4E6eoA6LhIX8dU0od75Kg9z4gKBc/RBBKX42C5lxQpbqSoUFfZ4uEhunEdwCKySw4GeFUtPds8xFJQtR9k6w/DiH6YcHa2xOn5GudnSyzXkvZZMnHO0LVSLjX1CQdnx+jHSdRBKSuCT0Vi3aR3TuD8OyomoGvH1FgSaJY37qErTK5RCckJYKhImzEBXdPi1KNv6RHXGz3TwmISz9PNgkgw4yg2CwVF08iglIFG0TGJowg1we0RMlfsKqL6nUtshvaPamOxvaCpmWrGJvuiBgr+q9+f4Rij4gGs1MYYDfuDlJyb5OHiCT10fyP6clvdF85cAMeJMIZbnu/nEvydsoLcTzzFxCfVHPESgUODOOtAJHtwGAZIcrFW1S5G+EXElJqllhKOdbBE5RGJtVoTgMy6gACLNHbRi0fJ6T8O6NdrpEnQX9s2mM1bxKjIHHBUO6aMQFGDs6rC5RBXQFKDUBdb7G/vajbUjOhGIGFHBBNJzXVUApwylbiCJgRwEGSecxKgq9czBN2mPGJKWQtbC5FkTmpXiUhqgDVdraMvqL0iSzZTWXhR+qeqHKu+xcga9CSeKZzYGRqz2jFYvCvEMwhgjVtgSKrkGKITUBBpYBPQbXWYdQtQJCzPV7i0P5N60b0EHI2abpuhRnLbiFSqpolEYxIOObEunkGEgIjQkKqashZEMfWbeSJEiFthghkULVAw54xhWGG97nFytsLp+RqrVY9hEB18oeNy3TgmnJ31cEKs426SK9V72tB3Db8JwrR9jei7Vfrwi0ZU/1RuUX6TBSPnZ4gfRE1wzA3TekCV54rusU10LPe0crBN2yI2EXlKmMbRiz/lJFJNCAEhRpkTEqAmez1r0kNWgUa9ylztR+ZgU96VH37Jii/4WBT5xaSeCiQpY7HxKpJC8S0UXmA2ElUibjANlKyhPqH6NC6Mgyr1mtsfMnwdFwZRMxqjVuYMQBKf8pjtiWYERFLsQcm7LNAAxK5T8ZFAsy20cQ5KolJx/TmRBl0xAhXC7rnrmTER4DlwVL9M0YQHmYj1sMJyucI0TYI0KWK21UrBeyWqsj+yuE8aUVGubOUes5b1iwEalRm1qErwUHjT38s/VioT7kIq679xxGdoKGUpYRg02MoWF7OkDZ5U7cKsAWMgQDWP2Qy6LPFVFpAEFIIlRlzZ7I2lK1aPnabpxBskAPM2IiVG4oDVeo1x6BVRZ5fIoqahcD1zaADbfLpxqY3YWrQANdjenmuxe0GX40SY0hkILZpmhsVsga3dq7h8+Sreu30Xr379FfTjoGKyQIDEopYjmBAuqrWISuph8YKhQCAt+hJIgs6yGDgQQ9RcOcA0Zg18krrUp6dLLM97DKMZXqvaE2S69mCU1VVZQjhrwqO/GeEyQuGbotogDF+7GaxjzMqc7NoiHRSpoSJssKhkuyVXuZXqxxn6Leeak0BRIZW7uxNGIHTzGaTYk1j6p1HjQEhrhk+Tq9DSJIFkpAXt67cnBQkGUjhDGSaKxOYE2t63JvLQdCPVGBrj0sBIuUYZqY0Im9Qh77+Z/1/EJ/8NF3+78J3r519g9BvP1R5fUM/Zp1yPddVXIGNvbw+P255oRmDRsATCpHsInHxwOGek1TlCHiGorQyMEdRqC0kAUKsSgiIm5izpekkmlLO4aU6TJLtarQbJ6b+YowkBU5owDiNG7tE2LWazDrERX+qpigY2ImBF20NA0XtrXpcpJYgpiDVFgixsCsJwvH+cxSCo7qPm8ZJyxrofvGCOrlhHlCYZcU5oSKJ9s6pRTE8uhCBp7iAlS6puIlhuGyA0EdvbW9jd2cFs0SmCltrNOXfImBB4QtfMMN/Zwv6lPUwJePWVl3H/3n1ljiIZRCJ/X0A2OKcJs8UCV65fwnzeYb3qcXx4gru3z8E5o5nPMI0TxqFHShnnZ2fuXdXOOvz477iClz7zAt587R2N2Bbph0CSRoM3DcfCTKXwStags3GcMI4S3JRSQt8P6IeEYRgwjkmJvBhc05SRstWyhquTLqoa/FkVQaqVBFm9ZEJNI2weDdU/ApEW7yXTL5MSBEWsvurtfEAISygM13++IGXYlZU6bPM3KjTNVK5ka0pBitq5mIGhHxxtE8OjmAlSq1p850dwSup+KYZvNNFTivg+1ueyqsqsj6acqUm/RXXrk4vMwg+/re3ZemyJzCUYDlR8FgM8iNPu8BAbIJv9wpSLsbK60hheZUcozew6Zuva+LFaL+TzcX5+hsdtTzQjaNoOe1euIA0jUp50EBK67V2AxJ97dXyC+c0RgKB0KxFoxTuMmArl1FHUzUdBrfYkSc+mPKJfixgPSJj/7vYCFAOW5+dYqdTRNGIPkILfEpDlZQABjNOIaZwAZvcKIgK6tkE3mwOxUTEYEmxFpr4CEKWPdi+QEKoALbMI0ZVOOYEzawEaYFRJpHURkyR3UZalO8EYaxJVEolILtlbyTd0ZrGxmHtau+jwzM3ruHz5CsZhjcODB3jv3QcYhwE7+7u4emkfRBMCB2R0OO2XYBrRtMCLz30G7Q98H06eP8N8PsPp6TmOT05BecLp2Qrr85VklZ0GbC228EM/9E+DEfCrX/4yVufnSDlhNp/juc+8gP29BV5++XWsUxJVlu7QZtbgs597HodnR3j7a+9iGlcVcA2uhpqmQSLLc0Y/JtXbjxj6AetePo/jhGGYNAJdVXC1x5QSUtPVE4CHVC5sTgZc1ABcCLFqbVA+yOfseXKgVKpEkcvtVD9eEeYi/enzKwq3QeTICDcJ8gWVXPmAvJOea553yitQcuYXQlzDWPMsqg3Lfs9stjF28kaKoNleNJCneQ4pYZpGhCSJ3HOSamchanAWjGKqytBfvUb/ZTYE1NgAyuWkYsbmcWWgmksqmGpMGZ45qmwyVfIn1WylTiqYs8V7lFP8zErAsr7YskB9f5PyKuGCKwZjUoN9zAwc3nuAx21PNCPY3dnC7/6JH0M/9OAkBHbol/hKaPHgtTfAw4ABhN3U49qzN9FPjDQOiiakGAlnIcai+2TXOeeckNKEceyx6nvJvzIkzOYzKWqhEbdQXdylvf0q2MqCRjQ/iyGIENR7KEGKoURhGK0EkDVNgykZEZEFmXICckZSpNRoSuGURVxumgaRBPFPOWO17gWpa+EcTzCVGRxJSzIS2raR1AdKRCIABCBrQreN9eMqBEEdWY/t7OzjR3/881idL/Hlr/walienyk8JbWzwg9/3W/CDP/j9eP3N1/CVL38Fy/WIzAmcF3jxpefx3sEdnJ2c4eqVSzh8cB/vvHsbSBlXL+/g2Wcu4eh0Bk6McZowrs5w+OAubj77Ap5/8Vncv/sAoSE898KLWA8Jr73+KvLE2N7ZwWIxw2wRgRxxeHiCN958D+fnS8n5r1LK+fkKU8pY9SP69YB1v8Y4SoqPNImazgKqZE51A+u7G0qzPPTQcYHqrgPgTNau86ZjbozCiKq5notu13Y2AdlhJww4OpUwhuE3gTOchwHihXep1RA23+ZPXz2i3KOgzkLeas4iNyn6brtMfdgvuC5SFahZn24xLpnFSMym+owRDUmxx5Qm9+9IU0IIMoAlrsfUasbkSoCau7RWUlk9DkZwcy7zBL2X82dnLFzmyRiRjb+/O/ym7uFkqe4NE1SzI8GMNgdlDdk5gcwHweZCYyOMMdlb2bTovJICiN9wxuLVeo2vfvnrIBrRdRGzbo4QCWkaYWX7xiHhczc/g5d+/+/FOgHLszOMo+jzm6bB+eoU7757C/fu3EfbBoSQsFqvcHD3GOfnK+Qpi692G7G3LRHM5AFVEAOiPguaUiCxqARySpLELZibKAGRsbO17WIxIKh93Q8I0whxgMjo2gY5swYKiaGsaVrkMIkLKYC263QpMqYsqoi2bTyAxvzMQyBPYGeRqa4K1Q2R1GPHCt5UK9ohSwjmcir+6Zf29/DstWt4N93HrJsjz8XgTCzpdb/x8ut48YWbeOkzn8VyOWLZn2PRBhyenuDeg0PMZlsIIeDLv/prODs6ASB9WC7XIGLMmoirV6/g8t4uOOzh3tEpzpZvoIkBz730AvYWc5yeHGO9PMd8a4b9nRlW6zWOHhzh+HyJ8zPJQ79a9+JKOmVMk6hyPHeNosKs4yBeVsGZuREAN05S8a2vN6gMlWIwzoXGGBIv3wqCr44D5N4nzjyyqE9qYin9ZXjsBgcnGQJgirO81eB5VJBU6bJBTvLfqpkvN0KRckqqt3JvU+t4Tj6HuJJug8ginssa8vWIQgwLQZZxKgWQZM8FCqCuA00k+yBbUCIDGp1MUW0FlV3C6zWnXPUZhY4DxYzgNLwYXUm5IvNF11LttFP7ulWUmOoRJQcMPtZkkghXEpFeQsYSqn5vSBpmSzCWpcxG6VNtCxKL028wRjBNEw6PDsSAlCVLX5omHN29AwnGAKZxxHoc8e67b2Ox2ML+/i7295+RYtwxYJhGPPfMNRw8OEUMwOlyhYM79zGbbSGS1CuehkF0z9MkOXtUv8/Z/LOz+9WLNEtgyoia6loy/gZV68o0ppwxZVEzTOMk6qFEyFDFpek2CeIWGUkL0gChJU1VK15NxUgmFD5GqR8cmwjzx5ZqZqGoGHLClMVYHKwalbq6ApWuXL19JOpZ9d26md5771383b/3C3j22hV892eflzQDSTZbIEaIEW+99Rpe+sz34IXnn8c//OIvIyLjhedu4uz8BG+++iru3T+W1LixgdWbJZJoa6IG9w/PcffeMWJkLLYW2F40mHLAO++8gzfffB9HDw6x7kesVqN43iSLkpVEduaVBdMX68a2lApsyM700vWxCpOZZw1nUqJiSNg8giqiyAWJWpNrWdxKqzVc8F65vhAw8mNgczCwK/Q9FO359U77aumhuk6ZygYPgp3P/rurFlhlQJU+LHW6kv7ynIK/L4wfPBCMGaCQNWZE7mfz4KPCGe7es4GYyx1DCKC2Q4gZSetvG50VL75J95syas1LBZDUi8jQzB1c5tY0rQRfJ06YUdF8rZpnsRe1obaea5vHsoLgPFW6FWo67s8uqiBj0uU3lBGtjhV5ZaPRw8fs2rb7DeY1lFPC+nwlCDZIrpoYA7Z393CkiLufMs6O7uG9V95EaBd4/voenn1uwGw2B4eId27dxvHRCXa6jOdu3MSV3Tk+88z3Y3vrt6Fr58iI6IcB677HcrXC+dkaq9U5zk9PcXJ6jpOTYyzPznC+Wrp3AyD6P0nrIPkxpzRiXIt7YAyyqQ2dS+QzJL+8JhOLIbhfO1CQghBZuHuaPY+zlCGUiFIpRB4paNyBBpAwo1F1UM4sbtlgTckhefsntVkYYjPIZAFJIQRJY0GMKU24ffsObt++gyY02l8VkUltL5lxetzjh3/H5/FjP/x5DMtTnK/XODwKiO0Wvue7L2HRRQRqMHHA2+++h3krCdnOzlc4Ol/j7r1TLE9PkJGxs72Lk7MlHtw/wjiMmFKqnkkelwBo5slc8uvY6zgRMmKtSNxLMwJ+P1Sb2qQHXX0ypOpXXzMBi4kwqdRQJ0unyq6upANL0yDEwjy3auxOVahLQYEbUkXNeKpPwpg0JsMoj6oJzJ2xECH1WlIUQrpuYKcrmn3oKfpaTiD92RKwKG+jIEoJbR24RWAn3BZwaHJF4S7iJSe/K0Po5ghpQpommWtXiWVJy7IRQyD/BCJwMEZp9sACvByEo5oqeyN75Zpq1/MJkySMgZl0d4HB1tKCMdJHAXUqf8q4Ojcp67O6hDeYizFs46tZiv88ZnuiGQGgufLbgL3Ll3Dl0hXs7G7h7p0j3PnK15DHEakL2Nm/jt/+E/8MtnZ2EAhoO/Gpv394hOt7CwzTGm+8/hZefvkN9KsVZt02dnc67O9dwdUre7h8ZR97+5dxfX8HN29c16R2slmGocdqtcbB0QnWp0ucLtc4O32A09MVzs/PkNKE8/NzYCLNJx+BnDAlRemtJEkLJLnXA2RzJ1UL5JxEdwwCkubXJwJPmvY5SO3WkpRO0FvbNAWFVWgiaCRuqU1BKixqZEHQvEia8A5kKRQA0/VGzS9j2TmZCFOexN6i6pVZN8OVy3toZwvERYvz5Qlu372NN155D8vlCQ4fHOHo+EzC/6ekifdGnK/WwizbVhhaykhqgA8x4sGDW1ivB5gaRwhnQfmi5poAd7UtyM6+W5nGbNeQBNoVdY0Qx4BiL7GtaCi5eKlUaNwHWwitMYlSPxhl//tfE9kVaToKzNWYq154A8nLeaRrwwi20wAj3ioNeCwElWtFFVUYWEGfF5iQqqdE1cCApsLm6jmOjO19vSP0MKhwSksFgbOAG6/jnAX4WEoYI3c1QSWNK2haidKfphE0TZ6inZnBCUgkzw1cKsB5nzXLaUHh9qzKR9bfb4OMq2BmDKSsA/Z1AEgsSUXAH2pmbyqM6FGtVg36aFSxCptLz6oJAoUVQU00klbmcdsTzQhi0+D7f+v346XPvIDnn7uJxWIbjIyvv/EuXv3qDyKtTvDSZz+DH/4jPwnM53j3/fdxePcAp2dLPDh8gJOTE4TMOFut0a97z/9/fHqGg/vAlN5G1wRNbhcx6xrs7V/FYh6wu7OH/b3L2Nvfw2K7w9XdHfDutqKBF5EzgShizBOGfsQ4DuiHHifHZ7hz/xAnD+7j5PgY/XoFZkkgF2OQxGdZN0BDkpVUiUqIhGnIIv1Q8CLfFswFloCr7LtQvH0sstgWawyWNkPQqySNTAixLGLSJFcAa0AcAxpjMI6TG6NJg8CmacS615QGKSOGFcZhwpQO8NWvvor/4u/811guJZ9N3/dYrtYYp1QRoJJmIsSAtsu+mRMD5+MShSRIf7LWedgIEAJ8R6ScytYjibQOoUa/vo1QJ2Kz7OGClOsNXj/mAkU3PgCUkxTJFoQpVJLAbl8ytQ0TfN6tdTOpnTCsJRDOqQ8ZNa0YA9V/yLiBg2kn7maHcMJnJNb6Zm7G6u2D8g414QOpUdh06Havij9aoJlbFrhG+BsjKX8thocK4BYGVBFvz74qfcqpuEGLDU0YAmu0sQMECFO0WuFg9iyd0kONUmeU9zOVGhFKCpQy5+YcUvNm88CyPtq1ZW4YFgjq18GgR8UFyCTAwqC1W/6lTj2+KZRUa/kR63exu4vHbU80I7h8aR+/75/9PR6odXh6jAcHBzg9OsClG9eQhm10e1dwsu7xxV/+Kt5++VfFKyQlNE0DICA0HXgc0UCLZYeyMFpFxzknEIsxNvMD9DvboNkC85ywniYsj3vcunuM5dkaq+Ecq+UJAoDnn3kGV65cxtVr17C7exVduwBiwDgmnC6XOD48wr3DIxzcPcC9g3t4cO82uO8xDpMEKk0TzO/asjy2XUTKkCLkTtij052gvtpZA3CiIncrZG66bkCK0EjQzqgbWRK9iQo8gDUtRc6CKKVqmqSgGM9X6k45Yr0esFoPWK1HjBrw46kOfIHKBnCtxMZvIsIm3dxplGe5C6NuSMuJL8wrAUwV05OWWUo/EllOIaj/vG7hnCBFYGzTmjEWMOOgqCyy999og/TdUKEQTs5WpcrGNvuONQ1KjEFdHRWpOqGGM5D6PVkJwNiPTtAt1bSyQe2PU0s9Tt4PafacShypm76P3cPny45XE2j03yQY/144nlbLYleJWSCeSERhQ+VIVZdqhpx0HCzbqjEa64MZrsu8GgMq0kXbdQBnDP0gYUVq3s5TAofs3kPYGCdT1Yr6FQ4CFHTYOGd7lql/qnGt+goUIlyDF2lSZsniHModMjIX5gtg436ARenbBVWsR7UHaitUGVmTBoEf+m98H/7O3/jbeJz2RDOCzIzX3nkLh/cOcHayxO0HB7h/+z1wGoEsBtPl8gTHJ8c4Xa3AnLDY6tAPCW1DCGgwDCJmhgA0JAQ1WnWpbo7FYo7LVy7jyuUr2N3ZwZXLe9jf30PbaIK4acLJ2Rnu3DvEenWEuwf3cHJ0ivVqhddeeQMgxvZiB1ev7uGZG9fw/PMv4Nlnn8Hu9g72Fjfx0gvPg7Wi1O379/Daa6/jjddex/nJCaZhFH94Rcbi0RPRBHhyOqlMpp4tABjZVSXEGYmEgUkgmqjSUpa4iDQKQySyYC4GkLHuxYjd9xooNYxqkO1VLy8unWxGcyN8igzN/dWK2wDQhV2C/VRe8U0WUbKkAkqUbaObO6FleVXib5sxc3aia6l8LROsUgchtNWGBXQbWV/dQKmMxqQy6YD8pASaHOkbI8ne70LT2FUcaZo2hBXrl7uh1pveEXWxGdj9jPDWZS2LkZU8VbgT5lpXbUTLDLAo97V7ybAWXFqrc4xZwQiXqV8gjMljEew9rWdUXE4NjDgzsTkwpqzvAYjbdKAABLhx2e3U1XiUOgylv8bgm66T6oMjAHUFtiAvr3RnqhztawB5CmlPVw1TufrDUSaKUQ1QxUR9OW1In+VZXAi+M4zqygqcPNxy6YHvhXpM7cllrMshwvHZb7CAsmEY8KUvfxV33rqFMa1wdHaK1dk5Zm2L3a0FQISTkyO8/PLXsF6eY8IS3LfgDCRuMDCBMiF2Lbr5HJcvX8L+pcvY39/D/u4+9vb2MJ+Lj/961ePo+BDHx8d47bXXcHh4iOOTJVbnpxiGhCmNnqCsdjtMOWO1vo879w/wta+/gq2tLdy4fh2f+9yL+C2/5bfg2es3EGPAvNvC3vZn8T0vvIj7P/Tb8Ouvv46vfekrOLx3z4l70d0arC4iJ4WIwKY/VcMbi467iQ1ykLw3zIIq0iRRx8PJmRD6YcSggVPDKLUdxikVDyknHrbwoO9aY5HKU8VSBmjz3Cc5W+EHmBukRduKfzShRlJElghNkbkT6SB5oLKMAEORHhuhY/WCUo8VGynbmS6W2IhSdV7xnnKCYJcrZbW+iDRDJiC4667QBCWOXIgrXYjcZe9WURb4WNUIvlaPGBut10StnlEmFaiodJyoGzKsiJ/9rXoA03XXTMKIu4NSF2yKEb3Gt3BxpwKt1SuZMEf1qaRzYQhaww+8f8ZMqNyM/Wb2snCpgqgFhYhU5S0SO0MC5YzYSrrxaXIkUYDRhUjiMo4EM7Tb+tlE7uX86iKfW2fW+uLl53qANr3LNuaIRerfHE+qT9Px5Yt3AAhYrceHj39Ee6IZwTQNOD66jcPTe1ivl1r/M+F81WMc1ui6FqsV4e13CA9OVxjXJ7i0fxld2+HK1Wu4dPkqnr1xE9evP4OdnR3MNOHVMPbolyucnC7xzrvv4P33b+Hg/iHOTk+QNFXulDRy1d0tnf7BqwP55hTzTWhbTOOI23fv4ODgHr7+jVfxAz/w3fjhH/rt2N+7IsFhscGzV6/i6v5lfN/nPot/9IUv4PWvvYzl8rxkWQX7RpBKYgRJICd9ssjXaZIcQ8M4Ynm+Qj9MWK56LWYyYRrVfTXlCrrAF7Bsft5Ar2XrF6IqhsiSM8kSnJn+NoSwsTfI7k1AMMOVficn5EoQjXDpoueaYOdCOD0lhap6AhFy2CTyGwTXNpES6vp/u6dcw3CjgSPeeotyYT6qsyU2/3d9vkWrguHuh+UlXaUBQ47OB8l/lx+LcdCa/CY651r9F8qPUNzt//qAKuHlet69nzZZXC5xys1Vf+HnuqYKRiD1/aD39NKJbCuoDGFN2ApvkrnWe9drzk/U76X7DGQqBJ2kohpRQJ40sZ9JB8zgcULTRMRGpVdWySVoTRFQCeirJRp/J32HskQqJmXvi+qfypNM95MsL67G15eGP9efVTMVfzyhDHuZY++FM2xpv/Rf/2M8bqNHcpVv9mKi/zmA/7H25ysA/jiAmwB+DsBVAF8E8D9i5oGIZgD+CoAfBnAfwL/EzG9+2P2vP3uV/8Af+T0Y+jWGSVwVh2FEICmOsrVYYGt7B5cvXcaNm8/iheefx7Xr15HThP2dHSwWUrtztR5weHSCW++/j3sHD3D44AHOT44xTkDKkte9URQSQ3S0FqgEawE2T4yUykTEtsF8tsBivkBmQr8+F0YyjpqsjfDss9fxE7/zd+IzL30WFFs0RGhiCxDjZHWKX/7Kr+JX/uEXsTo7liRymjbCVCTTlDCNE1b9gNWq16IjPfpe/OqnlCXGgdWMpHOas+V7qRAcyseH0Ytu8Fx0qLZ268v9LrVztq7IoCooq5hWctbX59a7C4DaPOBZTOGbw8iJPII2iAJIqtgB4s4rWV3JE9o5kvQ+qCGXJXeSI2zHwsXWYKi5Vi0wc8kV40iwuGf66HjUlTHbzTF04uFDXwL8Pnw7cunXxjHrT/m+aYQ0Yl/IRY2C6/M2DJgXfqu64P2gjd8uvkP5flGlJcNDvo8MDYMITRMkZkiBRvWGfrPimlpJTqx1PSaJNfH05IAmRrR6HAoaXIWIMs+qhqzVSeXhRVIu1+HhPlajYpJPDbQ2Gcdme4i1U+HAD19X3HONwRMIe5d28Gtf/MoXmflHHnrAB7RvWSIgoucB/E8B/CAzr4jorwH4YwD+MID/KzP/HBH9BQB/AsCf17+HzPw9RPTHAPxZAP/Shz2DOaOLEe1igSvdDE03w9bWLnZ293D12jXcfPYGrl65JoXtFzPEAJyvVjg4fIBf+/WXcXh0gvv37+H4/jH6Yal1gwlN2wixZc2OGKX4RwA0RQI0q2fClMTVMsYGsY3YWmxhb2cXV69fxzM3buD6jWvY39nGfGuBIWWslqc4OjrGO++9h9deexOH9w7w7nu38bf+1n+On/zn/iBe+OxngbYBB8k138aA3/o934333z3AV77wy+KuuhRiL/ENkvoiKcFnTXXsbpHVptyAW9os/W8hSLb4FfPowjIjKUPQrqGRi26HtCHClw3JgBsLpXJiFR+BGhHb5AJW1Wtj48lFcs/MHnsknqMmjehxLh5VVqaPU0YOtneqzaNSjEVWx9ioFxlVw0bluy90AFypMsBqExfUzaiYnDGGmmgQzJ8SJqmUPrFGhpYkbABgBeszG1Oub3+BqXOZTR9TM7jaHNU3L2fDmZaBzOp9q6RG1UAAJSVzMWQS6jGE/64X1ILLxknSXeu7MWHJiyWXmmTF9iAfc39v5o0aBSE2oiqaRiTN9yVRyZL6hQJpIGjwXEBQECivHgrD8Zcq47FpZK8Yn7+Wz2L1qR6HyvaVlPHo2axcpVYbbtiRLs7dxkCSeqkBi0WHx23frmqoAbAgohHAFoBbAP5ZAP8D/f0vA/gzEEbwU/oZAP46gD9HRMQfIpK0TYvv/97vw/6lq7h+/Sr2r1zF/u4ltK0UbwGkzN+dg1s4uHOAO7fv4s69Bzg7PUYaRwzj6ETTvBSIIsZhlGRXShQ86yAInKTCT4gN2lmH3b19XNrfxTPPPINnb1zHlStXsLu9ha7rQAxMOWHd9zi4f4AHBwe4c/c+7ty9h4N7d3B+vsLQ92hixL31A/zCL/wD/PjvWmHVZxweHuHB0RInx/dweNrj8Og+3njvAGdHJ0jqdpk1R1LZOlXcQCgpLMr+IkcGhlTsOiqgsyAy+aK6Vlb1F5VKXSwGPIkCNjVOKQuZwb7Z3FdcN1UyQlGhIfPHUSju71JUF3KG5Tq66BpXo1wJgbDiNFS2bKhKcJIYpC1dshAMUWPlNJXxYwansqk39Pfmj6mEohCnkg+mSA62imyMK4Ks99hc7kZMak8b6bykKI/iWRYCYtuinW9JDqkQkPpzTOOIcTDduPTFpJBaFNmgw9W/VH/Y6JZJO2V9OSNHPW/2HHuGYtJq3T0keFAl55m0VklXZFk9jYl60RdU82Wd1zvz5jsFIlDTaBlLSSDJlfoljQkgLeQUKqlX1X7u9etcG9UD2fmnDHctVRReV7M7L7npJ7FWU+T6rA1m4G9Tr0XvBfmnDdxHoqb9rb/1e/GFf/BFPE77lhkBM79HRP9nAG8DWAH4OxBV0BEzG1N/F8Dz+vl5AO/otRMRHUPURwf1fYnoZwD8DABcv3ENP/F7/tvY3tlFbAJSyliv1rhz7wAP7t/HrTv38PY772F9duZl8Bhwz4ANUYzIjUghBCADI0+SyjhEtLMZ9vb3cP3aNVy7dhXXrz+DK5f3sbu3I2qcnAEkrNY97h8+wOH9I9w9uIs7tw5weHSIB4cPkKYRgcRtc7UegZwxTBOGfkI/jLh9+z5efvNd7Oztu+dPhkgiXbvAfLGF08NTcNXnGh0XlEZOQImiByFaec2gBkvbrPWC0/GHIVclVbJBWEbLGJBuUXAIYNWpZs94atJEhaJD6bNcXT0TZX3nZHl9CpE3TmVEk51I+MJwwkUaJGTH7aPZKhzkMzzgSk610Dl2icr2mTMk3ahWaMS9jcCSLps3Cb8O6Aaxf1i/UxhVRYJhTMDm05m1rlWKGbuXtzHb2gWjwzRMmIYemRPaQNjd3wJF4OT4HGfng0iyGxSJq6dUajZ/Ohc07z3hDW8aP79Wj+ERr7jBEAqBvDgKJnvKf/ruNo8Xh4fkPsTi6qws1gTJ8mZej1wOZX1+bKKqWrNGJueyH1jyd0VEcf/1bK9mOEfpW9WtTQGyAiDaXwM1vlOpZhblh02vNz23ijvZVC8qY9S95pKuPteTWOqzXnjx5kNj/1Ht21ENXYag/M8BOALwnwD4g9/q/awx888C+FkAuPHsNf6b/9nfxqybY2d7hrO+x9GDBzhfrTAOI2KAGEJZ8u9YtGk/TogSpylcUqcqNgExRnTdDIvtHVy+tIcrV6/i6pVruHbtCi7t7WJne0sqJKUJy9USB/fu4969u7h39x7uH9zHvYP7OF/1yNPg0saYEtarNXrNdHm+6tH3A6ZR6gjIusto2w472/vY3mMn9syMUX3QJzXE2n1lretSM8Sih5ikeLctxoddEStkCF3AG9BQI1FB7lftCMR1uIJ6cz/oIqtrGNSTprn+PZkdeQS13Ei3rG7CenMk65/2y8gVuXFBmZ67BTrrcqIuc1yYWkH19pxCoDeYIpuXiRW+r6CxF47RcVfVkgG6asH6RkX5o8PMPqzWN7e7GJHegMyWmRKIbcRnvus6ru3t4o3X7uDe/TsAMwKkdGXQtCJtG7C93WFr3mC5HFVtRkBQ2cSQtuLNXLqEi7TaVRfe9/I2jPI+NdEr1pVqUVxIXc26/otEaG7IpQ+OwKsruV4bFDYKRoFQMmFYZUGNv7Lo6KCZhpsQBEiOkpgwmw0MktjOgjlrT6qyZgqtrr2ruOpfZbXSd6FqzsuCCPVevfC7j3LNROzehVe4h501tudxWffj+Ml6Df0+AG8w8z1IZ/4GgN8J4BIRNSoVvADgPT3/PQAvAniXiBoA+xCj8Qe2nDPev3MLyJLgLGkSOEdjTfRBGyebXN2+IWDezdDNOnTdHHuX9nHjxnVcvXYNz1y9gr39fcw6CTqbhhGcR5ydH+PW++/hzt3buHPnAPcfHGF1vsQ4jkhpxLof0A8j+n7E2A8Yp4RhSKLHH0dMYyqLhUpec1sm4yilIYU5SLBXDJKFdBoHKXij6qx6BZES4drzpU5B6xui3rSG0AQsbahfTC9ZJ2YzpCP3VrdMkG9a6U5GbBo0bcTQj+AsUcLMpDV1SzbNXBFH8zTyKk+6cN0Nl6wylLybpfW2hMMgPCJmoRatNYFDxemEFhhWKwSx5LmBj6WhLVMfOCl0osnVGNtY1EQDZQPbcyxVAtHGvYGKEdjYo0h9GaKK+4Hv/wy2duf41V95GcvlaMMAikJUgQgQYZoYR8eSzbWNUYGRnC28xchW+dfJq6NPGKW7gPS1jyjzsUHiq8t11GFMB7DcTjZwwRmvj4EBg7A5DjYfIHGRhY2hr292Au2aE5sG5+MlkZ6tsdi2CCTpX0ydRnpuAiE2jSdurF2KC/MxoGTvTwU31HPPXI1VpcryPUsetV0PuD9HwtDBPmdljFEzmXqyqERChybgcdu3wwjeBvDjRLQFUQ39XgBfAPALAP67EM+hnwbwN/X8n9fv/0h///9+mH0AkJc3osiTIPygVcI4M8apbLQQA2azOeZb27h89SquX7uCm88+i6tXLmN3R+IFZm2DzMDQr3F+forbt+/j1p0HeHD3APcf3Mf5+TnyJDVwV+sBaZqwXg8YhgmrdY/VapCKXFpTgLmsQEO5oiPlirDKu8j5Se6bMpq28beMISDHgK3tBU7MVx4MIHhyu5I7n3Sx6nh4UrGCYBh1zipd7RmbC4+NVJAWyCFYPQLX6duGrCYkaUH1rMxOEtmJzzaY3bffWubsHkTyRtIFq+ewkQiOzH9G02oTbYwjw8p1hjL2NiZmBzAEXOdk5+yEppYKNiQHfUF/X3PtdNSHgg5FZKkKlpTNLhvUUgCU9fFI9FgTAZ2mEAjP37yOeezwT37pZQzTWK4whmXSUTUvzMCYJ1VzlFvX6tHCGCq0Wai4j8MmbTWia0yZURuY5SQdQ5ckSQEA+6vKIU3HYLeo1Xb+cNpgTOpIVvrr92e/rw8l6ZtWenuJB6vWRYxoY9Aqc6mkp2BGniYEarwGxeb8VdNmtrvCB+COE9VxG18DHMHVjFXg5AVmbCoepyvV/nNgQVBX2LI2/RhlvPqNV/C47duxEfwSEf11AL8CYALwTyAqnf83gJ8jon9Lj/1FveQvAvgPiOhVAA8gHkYf8RD9n4K4jUYCkqqBuoit+cwJ/2deeA7PP3cTV69fx872NrootX1TTliv1zg5OcbdO3fw/vu3cXDvPs7OTnC2XGIaRgzqd79er3C+HHC+WmO9HkSMVE8dQ9dFp11RKF/PlerGiClYglcgbo3zWSdifVVblSGeDUPfb6gZCmhQpAFB66V4WZD8KTWxJkFutQgpRJQlNa8uMtd1kvwj99QbazI18+60jSRIJ/hNmRnDMGxsho2+QFBKBoM1ktSSs9UieBFtDZFWKiDW/kCkDyf2ej3rc8UTTMqOCm12vYG+82bEbq3+Uswo5UFRfPIFxZuqCr5fre8u8enc+32rlNf+nAqhg01tU55vGai7rsW1K7v49a+9gWGSEqxhY67stQJiG0WtxxqAl3W9WlQtSgS1qRB9axU8UGaMqj8XGARQEUbX12+iUmc6F2/qsQp6DWFTdWhrW0HURkfYCFx1z6rvxXvN1nn1VlwMtVRdKLEH8jfnhDwl27SYRklxDYIXwSnLRSfLOQAXaRuFY5XvMgcyx8Wlmfx964E3CbQCBzAWboNrLruV04EOr59DhNNlj8dt35bXEDP/aQB/+sLh1wH86CPOXQP47z3uM1KWZG2xbbC/u4OdS/u4cf0yrt54Ds8/8ywu7Ytev5t3EkA19Tg/O8Z7x8c4PT7Be7fv4db772N5vsT52RKr5RnOziWj6HLVY1gPGMbkG0qIFvtA2zYuhkRbFJpaOsSNqZP4pAIVuCpCEYjQzlpPCpc1v1FKUkxleb6WjYyLqqHis03VBiu56v1hTvRKQeuKeRkytvsZsKoojD2HwciJPRzfCKsVty9BdYDlyHF0U3+sUH7JgGAeEhIYVwBRwX2uYoLWTybSYM/SF+FsErUsqiolwJU0UAi1jJm7DBpCrFQ70A1rXLiO7DWkVpBitZU9E6r+Eyquoch3Y025aqokFYO6kcZAeOvNWzhdr30u6/iJ8k4ZRE1RsVTvkzXdiKlUilRQswJ732rOCtnxZ9dE2eenUndtCJkMkKGNjdqKNjiF0ZvrszDPgpw35qPqJ8NPK1jZAVP1Zlw9Vuhw1R3yuScKiJHBgTCBtMYFKyDJzqyC1r3O5p6s+6eMqdrXqAAal0hqbguAUdSS9vumdmjT48pVhjpE9mJ1RL/1xxg1gbCz+GRVQx97CyHghc+8iGvXn8Fzzz+LF24+i/1Ll9DNZ8ij1AU+XZ3i3Vfexa07d3Bw6y6WyxVWqzNJD7FcqUqnR78eRLc/SkUwSSZm+fV1s3kKXKAQCXYgAAQXU2URGKo0l0LyjS+J4SokDyDGuFF1zBc+ichadrXc3z4b8TFjn9eRdSKQAQ4AskvWtiG4sgcwmwoIZXMZ1KoIHlmtPDss8MhXLOszSdG+oSJSxmSBxj5+3kd5QkHmcJVXDKRI1ja2EnyioudydGT91tE1JFWrBGDvY1eyIz0fICcgtnFrcqkvnw1jlqpkdk/zniKEovNCvT703T2wiVy6gUsUtobk92ma8OB87US8SCil2pdRh2kw5zwlUFqEKKXJiUIZtWrNEDYIElB+srVev4kQR/L5rWhVGelq/lmp9oZHkjNBPcdSkVRdq88rd74wL9W6JcDtC27vQWULqeZLlh87oSUKyEGkrdg0UsJ2KoWowAROGROLqlNsYSxxKhWzETohvyWe9GnmBWadveDtY2CLNiV3M3r7MjdQd2GS3P6nz5FkeravgLfefPDQ3H5Ue6IZwaUrl/Av/gs/idlsBnDG+fk53n37Hdw7OMDr77yD47uHWK7PcX6+RL/usVr3WK+M4ItBaEoJhW4WEgGSoCfOovoxUBwoyn7NJYth4fS22SKaNoohetZhmBKGdS/5TrJkN6wrPdn124sFZou5IzpD300I6EEY1r3R9rKcmdXYWtAB15tL+9W2AV0XMU0T+lH19bYZ2JaTEMOi7y0GQIbdyxij9VsYT5EUsHF9WbTK/Kpx9pKOBFj1rjpAikjsHULsCpOR3wBQrJgsu8un2RxcVWcjzVqnWmsP1KqUggYLkSM1RBZlUDmnZhRuUNZ9bd4fdVRrhd/8RFs2xvgKbzeQYeyqqPdYpaCi9hLpomkjthYd0jgiLRrM5g3GCcimCtNFHpsoTgsoai1nxGUGfcw2WUUBPoUJ2OBmv58RXgfvujBNNaMaEWQ7uXoWb3ypemHM39aXq9Uednu1/jFYeYkClWptlc6zM1vnSwocQgaSMsVAUvdEKiFKAKeoU60uRkaIUjODc9nfxoycabvaCAVEVBKPTYjtS18oRB6Fvzn0RY3rr19J3mANTEOZg+X6E1YNfdxtvVrh53/+b2F1PqLvz7DuJ5wvlxj6AX0/oF/3kj1zHDGNGZmzlxQsLoHVeuOyPWxjCOdWB1P1gzc0FQJp7vwObddisTXD9s42FosFQiDE0LhxbppGDMMgnkXrNc7OzrE8WQpzUMlja3uhGyD5pjAkc352jrGXgiyspRgBFMJs6gjfaUb45HNmRj9MkorXUAaJEVXGQwiYlNlkZX5WmN0eVVZgraO3CF9mU0fpgmZZ/k4clLiKdKUpAqwQCQGcE0RXulkK0ryNihdPheN108r8VBKDel2Ve2jH1dfImGbtblrsNv7GvgZqCuWuhO5mqD870Qc4Z0lprkyO6yyZUKSaawJsP5Cmb9a+BUCK5pTXZpaaEoby57MWi5nYA9ZTRmxnaNsGbZMxThnDkJDV9XTWtejXoyL7QlWKBHBhbHXy5L0KIq2RuTFiI8u2LKtTnGHXxNqM6Q8TsWJvK5x6U9LYRNzWb974d4OpK0E19U71wDK5XN+1Qt9c9n6IUejBRL6XwKImzXkEKKBpI0IQxwmnL0qty0cFT6xPU2K+EbBN9bwow9LuemyAwySGM0VHOFzGhyEqO81A/LjtiWYE52dLfP1r38AwMcZhxEp99a1WrXsUOWGylMVKUHSm3ViWK8QGJwO6wSGEs2nQdRHzxQLbu9vY3d3BfD5DjJLFUFSF5lUjhCBkMdBFSLGX1I8YVj2mcdRiJBIlfenqJZh+nUiIZYaokI5PTouO24xV0EynLgbyxhqHxhEIYVIJAKr+Ma8CfVdBsRlJ8yZJ4RnV0yfTy6tXhwFDZRpyKKjmS1EO62ZiLk4k2XzvRSIR4642BkpdXgBuByn65kIw2evgWoyDWRX8VioluX5ed6HUXi7bxIuU6IQXUIBCJNyhXYurcIV8cznd+uYExGmrbdiaOMmaDO7hZBfoeSbpmPYLlRu+ntO0EVvzFpEYy2WPccxIGQBlnC9HxBgwmzWIkdAPVmrU1rbaC1xdVYigk8+6/+zTUZB9TcIr4rOxBjfveGFgbO1iY+DZOJAxg1oMZBub2n5SmOwFnq1VvBgXe+A2E9SMqNzT1q9fkeBeaoDQAQqSyE7sdrZOJkxjURcBjJSyjbiOYzWYJfqtjF3FHAvzo42Xq+s+X6BWVQvg6p42R/PfaKUq+37EW2/dRaqilxjsPr51lkEATvxtek1k30QdgvJDFEV200S0sxbzrQV293awWGxhoYSfNe+QEVapuJXEKydnqXN8do6z0zOsNetnTsnFRrMBzGcdbr7wLBY7W4IumAUxa+BYP4xYnp4XG6OGn7OqUS2xGoPVnbG4qclzALNJlAUux5Pm0ymElj2dc86bHkdiT7DEXIymayW4LjMmTWWgw4jaCMtW8pIKY0IgX7YpMYCs0d6GttmfUxpvbIacM1KSFAtWFAXMJWqcAPO3pspjw/Pi6y3rrKFQlYvlRbJrqNphmyj2wvoCSu1dmGRFhRFfNE6GQiBM2kS1ZoUoKCdgwpg1HxIxZm3ENGWcrsciKXmtaEnjsZwSACkvurXoMA1SFMWlDjjoLcjRdgWZ2sROLATMkHe1Ojb7TIU0WfU157Z2L58DrpeZSwXQPtjgezryKh6lnhOUpaHv8rAB3ACNrzObW0spXUs2ZLNo9yv2IbO3UNOAQgbn5ASfU/bgzxCip6nZVG2V+7nTgX43F16TCKh6v/Jym3O3wYBp40+RFxS5mOT8OO2JZgSZGcMwFb7NNtgFmZBSS0f21fUuWrH48IYmoO1azLfn2Nndwfb2NubzOZqmQRMbSFIyGcR+kNKWIUSkUWruni2XWJ8tsVqvsV6tMQ4jkoWKc0EiJqYaI2pjg/29PfSaG6aNEaGRIsTTNOHo+BhTv5aSkRmeV52IQSHCtPgwIOWEqbxr4lzE4irRp3bNx8xQYr34yIiWEyqgbRtcv3EVs9kMp8sVDu7cB7jK4x8C0pg8BbAQYCAQb2xCSaVh1FG8wKRPStSNoBDBVEtE4qmROCHGoGMghWti04Bzcq16KYhug5HF24nIGZ7ZekwKsvG1MSWPiPaF5ve0iGIZ2hJoVxwo4fcqnh3s3lGcLXUGFX12hRCdOWT4WmUAlDLOz9davwEFLSvB9alXBr/YmuO5567htZffl3cENjzHakZnShwHDUZ4TDJQwu32byOmBthNJ84GPKr+2L6sjzqxQ9UMCJTjtTTq96EyD+W3ij1dSIfBzBsZaC8+0ZhYmTfpW9b3AhuDLGMWQtC017nYDvTdxYvOBriSCKyXepy5eOCZcwX7XNaopRoo49wOaHXNqyRiabUvIMJPJencx9uY1S+6LGoiW4gBsvNKaolaNCdIhF3XtVhszbGzt4PF1jYW8zlmXech+jbwKUuqZwBicF2vMfQjVqsl+tVao4cncJpQgjoyorlR6uIQmkO+cGOI2N7eQmgaUVsRYzX0yOuMJorr6enpubuvsr4kEam6QBlZFHdVUNB6BAXF5lwWUCACRdvYUA2MjZ3q2M290M7xRWZnMqYp4fZ79xCIMKlNw1E5mcdGIeJ1cHNtXAW4SkXNAAqzLYXVRXIJgRBjBCOpFAEngARF8Tl5rWJPK2wVywiYkvTHiH4h6ja+7AWGHFkZjdW03Rt4mCBFgRyC6DHzfFFUZwKXEzyisiZzdqnOV+cGPzDgYLMgmW9DFDtVTnlTVWCMQG+12Gnxfd/7It5/6wDDqNXS9DwCVFopRFMklLxBhDZ05tq3jeRrKPPuQ0e2YgioKmj5QiBVzShxJFRSSnWyvFYhhvVnANU8XiDsRnRrugmIZ1YozNbpxsZd2QeRL6hlfAjYAIocDE1EiCSpKrKtaVUnEhTEBB8XJQkbNxY9vkr2GlQjr1ev1dJfduZS+p+nXO6JYhcxJ4TYPj5Zf7IZAWRALFWAMz9lDMLdNf8NiRtY27aYzTvs7u1ga3vH1TyGEG340jRhnCYM6zXSmND3PYa+R681hft+cIJRjM8o6Z+NltlWihKUZYTCCofMZzPceO4GQoyI1KJru4KKmXF4eIzzkxOAkxu6Y6zeN2sJP0XXtkBzmtA2DRhKIE01o2TMUHAmIbyetiEQiKJIO1HdYbnokq24OJglJbcRYqXmxuAsX4sQ4gDKorKoyGXZqHoPc3NzpA3SqmUyzmbkarsO0zhWUcpGmeSe7unEhckYgmQu716rdIzZAOoR5gQf6iOuTcRMOHO0ewMu4RmRgxnllCnLqdkLkagfkmaahDPJQIXV6KToWMkD5beAEIDZvEVKGeNgxvFCyEKM2Nmb4caNXbz+5nu4f7AEQqlu4ATFGDiAi+qUTeKoBNsIWKE1LimiOr8ESFU/2GwZ4a24ywZppnK89oOX+RGpupRXLffZ8JipUPjGs8GePgVAKRxkQOtCXwsXZ78f6zooAEruFWMAoQE0Z5E7KrFFQScQoo+BAbGqcy6JVCMp+8C6Avao7OKVVkut1cTYvwqGY4x47sUX8LjtyWYERKWgBBmikQUSm4C2bTFfzLHYXmBrewttK2UnrSZxrX9L04R+mrBarrBeLjVJnOT3MQJvBCEnJVpkPspCqAgAxVC5DxoRIWgtdSTV8SaIeuXGszfQzedw4515MzGwWq9x9/Zd8JT0dQ3tkaRPJslJH4L8jR4DwLpezdsGko8pqDcOVRufKjRB5OgQdsz8qyEb0HKtAIokKTuBJnX/JCfgVDEIAAjgLHWLLXCLAKn1oAR7E9PZHiyeOiBCmtTgr944m3pgQ13VMXUCMF8AMXMwSp0DdZuFviuFanOSRlJTYSB6oxCCuv6poVr7Z3NQbU1fD3bUXYR9Ldt42NiQM0i5ezGeG2FKCejXI7a2Zpgv5qKPbufIbOqkhKlf4Y1X72JSl2FnUjXR3qBDF9QsG1JTQeZyi8JIC92qV4+xMtXF1yyBZX3Cdo7fo6ikikuz0fky08wAJw+h9375WmZB4Pp0IcSaj8JjIHXvkBP58hx386RqZCp7CZmHnCwIGMPKTKAYESkgqKtpSf8CXXOqzlba5bjCeVi1C5SxWSp8sbkEfzd7V3Pa4IvXl7PADDRtg9/13/zt+A///F/F47QnmhEQgKZr0HYtunmHrpthvujQdZ1Y9VmMvZ2penSj5pQwjD1W6zVWZ6LaWa1WHltgYluJD7DNC1UxKDrhYuxLKcvizQzPo3gBjZYJEoP01WtXsL23I6J9hPgkW8bSMeHenXuY+l4vMwMuQKHoD43QilqkeIZQ08AKdrsnjhEpFPUHkehepT6ALS4CxaA5/XXMNAWER94CKuVIqcqAUNJHw9xv4YvUfvO+aD+SbsZAkoAshKhISYN2SDXuOYOiIuOcnN4YoDTCYQF1QefO1DmZzTCuolrl3VVIIHzjuQFXjwVYdTUAoTjsoRIW5L0LUXTpvu5sRcweCtqqkGAdA2JzYmo5UEkTkibG6ekaIQS0TYOdvRb9qscwDEjTWDHJEntQ3tE6Up5rnl4GZuz9a0nI1S1Ovw35E1w6oOp1zMuqHikC3H9fp8AMyzJkRc3EXF8EVTkqMlfG9hCA0H9qlO+qT7u39tVYjgAYVWdVc0Y+n2V/iEpPR8N1sHDPInEEiEAbQClpoSsuRFvXAtTwTaFIsWVA6s+wwSxqcH9PAzOVraVWFULWWiBgGkf8/V/4B3jc9kQzgm7W4Xu+/7vRdq0YWsz7q9p8mTPWq5WodtZrrFc9lss1pnGCeP2UXEEukpEZiCwghasFxY4kMjPSKGH/G0Y1RTX1xi4AQ9Dt7u4utnZ3sF6v0LYdWuqk7i6AMSU8eHCA08NjMCcnvpaNM6eCtMCkUY0ZeVLdNgvhSqqqcL01kRPZYChK11pQ8VgkDHkLjgQzrkpMvaqVbJBVS5SZkTb8KO1NjWmiEKGsjEmJSiQoAc+O2mRTUOXqGFQFBY/4JnsfC9wxe0RFbM1biZmRskR8NrHMjc+pvrMzFdt0hrIyRH2jJ9jvhZw6xUftneQ6cL2duy07gr2oijBJsjADMXPJHJYUB8oMlOiYC6rYsQZM0wCijKaN+rvcXWpF6Htn0hgrY5raRx3DjYAoZeCWdHCzOFlZ90UfX3GBioiTvdPmZYVoGxOA2kNM3USF8Pm82zj6/ipSrq0zsxGWBItwla1dUxgzedepWmtVD2uOYNMFrgaNWbKK1g8Rm5zEHuQkqWqq0dK9nRBIaiOQOoRUAhLKatPHa+U6Z7/OvIoajeqLURjMOCa88o138LjtiWYEsYmYzxdiZFUD6TAMWK6WODs+QWBgnESfz8xSeQhAMUyxowtRy8C5qqOuLDo96JUuNesiDebKyTrpmgkUgOvvXFcNgELAzv4Obj73LGInVZLMOJqzBIrdvnUXRwcHSFNWxkbwEH57A3UdZWgtXsArbdlikMjosmjtjzAyS5xVjMmyEWx0ynPqHctgTLbhkndHn13UNLYoc8bGBjUkZ8vYmJxsWYIZFcFsjjIlqE2Jixt/63QehoqCEm5QCTCzdzYGphMcgkhyQQMLLhqB3VAabOxrMKAqJHZWoOsIhZgrMgPBx4UIxX2PqbxDJUltGEYVmIAlLMTGwRgAKQjJaitardbOhGIT0ERCtNz7FDCMI4Z+9HGJXO7J1g9DrhsY3k4zWwvboG6sKwMsvoJMzVLPv60RAwtOsCqpp0bGFbdkHSNUP8t8aMYqU58SKgmgut6BCsp8oOqfY0HSIMn6QebtVRh5ubGtXfaB8DOIPB8WUZbqdyixIcxQg3/0SPpsKlpjyEZ4QCU2omLg5BkgZfyyOeL5/i8u5XfvHT00rx/VnmhGkFLC/bt3se57pHHCMAxSG8Aj/ozQQRPAGeGRzUVKHDKzRoLaYMORaTCfdNZFmIUw8AZnNxfCrH7ipOnVRYVgOskQCJeuXsLN526ibTugWkySr59xeHSM4/sP3BOEVPLwOsT6zGCupCrGZ924LhEBMLjr6C6UhSPIrhiBjVgCMlZOwF20tDGAP88WnRjaAVCVnpdYo1kBI9QFWSoKzUCtSE9ZVEymlMhsRe6NqMDMH2B3kQOIisG8iRFJk9W5nt3mIcYNtA9ICgv36qlgsZMmIlgaB7MRyfekXlpUzWNRC8hA6f3cQwXVecaZ6cKztHtmw6qAihGDYdlj7/I2urYVm5eCkZwzDg/PcHK8xP7+Fp57/jrms1ZSqYxij1ivexzcO8TW7sJJcMrsaRNykohviWEBLNWJ0Td30ebCsLl+NUgYk0N4trVVVI8bk84FzQadh9rLyJrHYtRMoKAXUZdyea7HCChzIV/EVVeZnYCXBwElCIKq51aAqtD6jXkt264QXx8H/RpiAKgFW30OWF8z8sQAx5K7KOeyPz0C3QAXXGqQZWlu17Z+lHFV+YxAmj0gbnT8m2pPNCMY+gG33nnfyzoKKtYUBUoMsqJD8zwpRknWpE4opehyUReY2A0IIalRhbtyGvRAUWd4STzAy2cii2vZtevXcPX6NQkwUbdXQ7I5Zxwen+Le+7eQU9KFxs7RA5GjL2bj9LLQMxcDrfye1UMJgkYsjUSlCtlMZqWoW5mkRGWjICI2PWq1Qcy9MBfEKv2C7xACaxqJUvuV3TDMDs6DFfuIlqVTkbaqimJdX8AIJat3lk5BsPnMIkW5y6w+14LVfGPob64228CFeAShJkSE6t0aZQSFgJOOdQgijUR1HCjH9XsIaJqgeWkK+qUqR1KggNm8xXzWoQmG+Bj9mPD2O7fwwvPXsJgtAAQtYiIEnXLA2ckK89kMi3YGnooqMaeM9bLHrOtw9dI+FosZ5l3rtg/TjU/ThCknCPpkTBoslRJrNlzxl5+mCcM4YRwniebX37PtI/Wky/UxlzY29+EG1Wcjdf5FPkUI8jfm44xI7R9cpHFjyuafX1yU5cKiFmJNUlv21qbEUKT62ptnM9ARFwUXpQ8lotj2ImBqyAjOonGoXrGKEaIN4KTbDa5qrSQrZ14+UN5N3zPmTs4gT974OO2JZgSymb1OFYpPL4ubXGjKwLo4rhsOJpopp62QiPlkmz7PGCoAjxQtop+JHUIFTVUkWSQlKCzEgBvPXsflq1cAChhTBufRJ4/BOD45xr3375YcSGBMZnxlkoXHJSOq0m1BWUqRgnI6qx1QQuILfDHhm0DuQrdhGIUSZ82tYpuukENDiTbmRvDLxjW1UI1ia0DGOld2k+zJ50xyYI+idjUtqbPPRobQ4GmoQ0PidqsMLoSm5DEyHkKlTrXxeIt2rvXtFOBE29QqIMnvI4gquPdZE6NLLTEqmgsBMRIiNf7uHlIMlUR1yZwu1xhVZemMyDZyIvSrEb0xcYjkc+3aFRwen+EES3TtzNd9tgjuQEhpwuHRidhU1KB/vlxhueyxsz3HvXvHaCJhNm8x6zrPp2WDHWLQDCUBoIguRlATXLVltIeVMRgTMaJMEAlvve6x7kes170yj4Q0CcOYJkvtXBB8AStmlFaJQYG7AYV6wTrTYJHw6lbR7s16zSQePgA7qHJAY1vGrzdkXdas3/fCvrEfC11mB41lB+me0HUCTX7pl+dqHA3IeanM8ta1RsGIPJwJc7W+N/v8kJPCN9GebEYAG2v2z+b+aIXKnVObzMTmCimpFUQEq9Iv+yLJ1USYrhGIKubnSfzBuYICshl1glRNEhrCjZs3cO36dUe+OSVQjEhJJIGD+/dx/+59uSf5WxWCEMp7SXCWTbKIj84YACUyusid2JJn9iSUhUBRU1OwIg3IhsxGvcxNUuMWshpcjUcIY4LtTkfWtrE9q6b/bCoY8Qgym4aMWXEFDKGgsAb2MPHVFg9YJQSGwhVlmyolBkmmF2Kjft2aKqSNaNsGwzjBjIgWpOaondQ91XlnsbnYVu+6BluLub6/1ZBOqu9l5EnsUTmP7rKaUtL5KIjVJtnjP4yJ6vxPqi68uN4JwLxdYJwGLNdrDP2EdT/q/Eheq9OzHmfnkqQwREI3awAK2NpZgAmY8oQxMZZ975JwoIAQRNUU1PZAyghjiJCy38HVpTWjL8xU/ppbZhMbtDEhbs1dKjKNCavECBYJ1KTyKQmTSPp3miZhGkl15znLGmWUdDIXkLx/o41vOvTkdJRsb9jvGjcj42wMArBaHhsEXoGj0Q1LG2JGc/K1q79vSBpKOwAgSDknK7NrtzZaQkyI0Wxjch9fJVRJHGDHpGXf1zYYqHrqNyIjqHyJwSXIpE7qKLTV0Ibp9mVG05QVBUqqBiJR4yyaBbrZDPPFDLP5HG3bgcFYrlY4OT7B+myNPI0gd89OyClB8u2Loa6ddXj2uWewt79fomUFIGCaJqSccP/BIe7fOQBSybBJACgSghqpiVR8RtF76q3cdz+G6LaE4oMuRJMhOmCzOfg4uIioenCVP028NmJoaLn2ZoEv5OKWaB4udQptM2iCLfAPACwXkDCCCLgXE4VC2IWYRyHYUQhQ20VszefeL3NzbLSebIhRNg3D87zYnItHlHyfkqieLE/VNE3KbO0dNW8MF/2y9X7dT1j3Z77pjQ86soXhPiPpauPxVesn61f7/sgVDvdMMVKia9mBbBA32yQrGN18hqaJIM5iLNa4GZOuzM1Y9oWsFQoy94kTpqkkbUzThGFIPj6EoJl1VWoIKh01QeZAGUdQqcHydklciaL6ENDoOaEVv36zx4k0JZXghDBmTCljGCekNCElyemVMmNSe+AwJUyj9FFUVCql5AJujFD7sJORDEeJ1f4swM43QjV1hYXDpZ9aM1PXI3ZpwEt4Vl5nBLCOvwfFii4NbteQCcM0mT2rgFIYLdioU1CAi7m41/EYgQLmi9mjFtqHtieaEfgkG1c0WquEuXBMOVuFKM0LokUn2hazuRD8ra0FutkcXdtItHEwjx4JOMspY74/w6W9fUzThHEccHZ2jqP7hxiW6gOvUKOdzXDzuWcw395SNU5GnoCcEsYpYZoGHD44wtHBA2jKSBhpdY+TStTOXAxlpFlFOZV6vyBT9cAXa8qMGDfzfU0pSXBZtCpoKHp1peZmoDSvHWMIrisFaT4jcwUt9glTi8DQoapPYgxo28bzvDSt1IaNMbh6r/F3YcQQndAY8UtJOkMbTFWO5UlURDEGiepkQb1CMJWwaZ8NVRmKNN2xITlbJrX7oEtRtr78i91LN/3Gd1xQeVgZ0eLlAzfGckmRXl5O5x3VOfY3e782jKS6J6ah9q55mLD5KxAQA1RXTcJAKMLqaqcxYRwGYQTVeNUtRELXSixPDAH9MCJNWQrjsBF5IfRN0yA20RlB1DkOqmKTdSLrJeq6cRAQO8w7kyqKBGruwVOaXJIYxgnTNAqDqCSMZH81Q7Exxay2GpciNhhvhaC5gAWXiJ3olPPqISJTFSk32sjSCziwo6AeQUSFGaCALKksB1VxlvXlz3GOBN1/UOZdqF/TNnj+My/g1/7xl/E47YlmBAQxyHrEKKtuO5BOsuqaAxBjg67p0HQRi8UcW9tbmM1maNoWkWIRWdl0cxpdSpAsossVKLPkIdL0s1M/ISTxVFkrIaBA2NnbxrPP38R8tvCAM/HhFgN2ShPu3r6Hs+NTV9eIQiao+2PxdgCjqJd1EYm/stuOYYY52aQFtQuqlTw4Zvh1mwI71oQRRxPrbQGqR+iGeixEbKgBoiLOpmkU+cm5sQlom4Cu67C1NcdCicQ4JnX9VCbAwJQmeV4dbMaM0YLrpoxpAqYpKcqbyvj4BQT0YyUF1lsN/p6mvnFRH/DYhNJkPJLaa7IbPWWRpVz0ysLk2XXxJg2ZV5UR6WzocOM5FYzkgl6FMGTf6NLnrM/IzmxiCJLfxkUScvCw2SouYd/1mq4jLGZyKOUMSZw2SJ8zo4nA1gIAGkhuNfGiSUY8tWXOwNQDMYByEkZg9rlMiG0EpoQxZ8zQYEwKVlIxJMv+ABZbM7RNdGnMssIGUtuLgQQFF23bqFQiaq1Z02LedQ5eSrH5vDGehWkknJ6ciVsticQiKqmkkkWpY+LM3jCDSXrMvvYeNfobY+/zbmvw4hUoruD+b7U21TmEhIOXFe44xtgLeaCa2UdCCJjLhD5We6IZAWDojMAs6XUN7camQdM12N7ZRjebYz6fo2slbXKjbqLZN3qqCKcY2gIFDMOIQKLfTf2IYd3jeDjCarXCNArayElcvIKi4J29Xbz40gvoug7MVblCXdE5Mw6PjnCqhjxbCM4MOPh382SQJFLws9KUgWjEWd6FTUKIwaOGY4xqwy73CxSd4TELgm+K+4vr50V/LmoAGU8Zt6AoLYag6puA+bxD1zaix1XiYIFKTRPRxQZpzBizBPBNqs4z5mjR3OYOyzUxq/445uIK3TuBhhNB27SW+rd4gxUi7sjbUXxB3bDPAKTYDmsRlYdWn//ZUCfoZwF2Se0FRbpzX3M9P8aIthHlbhoncSYguPhv3ipFjaFsrSFEJEzmQRbjJsnwDpfrs8VNACqtNWiIwcRom4DMEbPWnmPJ9zR5n463M4IakbpuhNE1AU0ETrMwcKCK4dB+5WQSXAQFKQrftBHDesTQD2jiDF1DTvBBpuIAUhJ1VcpwIl2oYJBkfGariCJ9kBr6m0aYR9e18jkGdIsG69M1Tpbn6LZafPbFG+iaBpNK71kN4tMoXlL9MGKaMqZhxPJ8jWHQ+CRLMmhM2aSAjVmp1o2ueVks9W+FafPFS/XEDAalekHyhXMryQVwrUlmxvJsffGmH9meaEYg8y7Epmk7ySc067C9tYUmRjRtCwrRRXypGMXqmROUEEqGz5wS8pQwjANODk9AYCxXa/TLNQB2NOZeDjmr6kR93RGwd3UXzz13E03TuA+wu7OR6DQP7h/i6OC+Sg/QAJXi/mq65KyeK1Z4RYxG6uYp+iNXLVj+/aYNzlxMfDa3Oj9mxF3VXkEDjqLq1s3zhaDnK2oxhWVBJ0XVNvRJisMbUnIIJGecnKIsbJNyUAivZWhkRc0mqtfqFRtrj6ngwsxs89SJ7mr2WhZLee7GOuJyRUFVmwvNUZf/IVdLFLG/epCeOo4JYz84iveART2HAGxtzbG3t42cEtbnaxyfrUofDe0rOnZvGlWPpSkJkmdCZJQ8V5E2xokIjm6bNpbuUhR1oAZ1JBZPN0m5nkCk6FONjipMSd89w2bdT+na1iIghAbLVfLCUYKAZa2FSBiHCbnPIk0SSYoXZvS91BtfzMVOBJgKqzgYiFAjvzWK/BlAiIxWAZ+tQeYJ45SRpqzeSrreVEJmlv0wThPG04T33r6LrmnQtC26Lgp9iRHzrYWozjQYM6eEO+/fxzvvHbhNAmBVc8lcuZ5fJ1NWJUNjkFFHTT+8xjZ+QFFL1aDDzrWJebiZCSoTMKx7vP3a648878PaE80ImrbB8y89j62tLcQQVc9duX9mBtLkgUSillGvHcoYxgF5HKWe8flKM4uOmKYRVhjF0FhW7xwigC0wTUWv0ATsXtrBc8/dRIxN8RBRlJ2TZC89ODjA+dGpb06hQBoYBY0ErpBgTurLHUrBlojotxZPD6mhHKN4fBjBb6KoDRrV2Rt6N4NRcSFzLaSPq4u9jopFWhHffu2tvYP3txw3gl2L1MzGGNmlKLu/nV8Hjm3EOQC+mjNraRw1KFMo2ydq7YP5vMWsEy8ZYfLZUeg4TUoIKt0wBB2LmgI+HmbElD5zRfCcL5a4CYobe9YGpc2N2JdylnxFEJxdct4Ded2jZ0a7mKOdz8FnKx9Pv101DATSFNTsyeQYlhRQR4NR8l/BvLXKzdwFOnZIyIghS0WzRirtjcOIcSKIZysLETYwYAnyAtRZQwmu3tzGadZFzLqAKWX0I2MYhAgPQ+9MVKTCQUAQi+twjITZjDDrZCbOz2W9zLuAriVn5n0vjC5EQqOpxHIG1rSunBzI7V9mnA26vi0L7GpMCLEVSYczbt++72lnRHpoHFi5+3CMiCFgvRoqUKIjwKqiNQOxTknBC/JbPScXsL2uJZMMyq+1QnTzN6oPwqGN7VP9KRNhtfwNJhG0bYudnR11KxMVjCQt0xiAGLHuB8xaqdA0DJZddIXVao2hHxRRJRWZBS2bfcE5b4XOiipCBj7GBpevXsKly5cwDSNGmpzo2mXD0OPW+3fQny9FLeFJQINvSM7QIi7qKaN69thExCioBNBCPCnJ7wGg2IiR1Qm9+iw48qRCcGEIzpAiIIulXpGbGstyjiLznL3KmunGhXCXCFQbMwt0o7IXRIWVBRUZOqYAT+pm0ghIiHDUOYmqIjApp1GXRpUV5DibagmVG+IkdSSUsDYENDPRKYOLqyIzI3RRE+tlzGYt5vMOIGC9HtCPpcaBrJUoxskkAUFJuae8L7lk10QCUcZ6PZagCEXottmbxQwUCP1yiSkJQ0NDG6oyq2hIsQpMY0ZS4mbAgEnyN7kKjBRNN+KBNo0TGMXNMwZx8WzCBBAjp9FVUE0AJhBSZvQDI41FnSSSaBRDfJpEV00ZsYF7rRAJWOpaJeALUSlNyYLMIG7OxjnMgSMK+peUV4Q0B87OE5brhHESL6OcgWGUuW4i0G0HjBNjuZb1EAMQG137LB5rTUOIARgncTyIemzWijw0JvHumc80kjwziJLueWWuAHgsThPTZCplZYa1ikfVioBJ/v6LrpWLzejNJgiofNNgIoBJr4Y+aofXstdQfbNPVDwYH6M90YwALAstNqHyLwf6XlQ6zBLQcrBcYRpGDOOINE4uRkqlKrlVUMtoznDjlKMbhid2s1kMJPUN9q7sY3dvF1POrn+lELAaRzQhou/XuH9wiLEfQBTQdELga5//2DSaJbXBYtFhZ2cb3WyGhhgpAeM0YJoShmHE1E3aB6pWl2FRcjWKewtUXinSjCo7jHeE5Uyi8unO2XT4xaZiIjWFqK6eEtlrxjlBj8A4iU6ya2JRLSksIpAbiy0aN6pHkQ2/MYXEQGxaMEvU6/JsJUV7VNKwd4D2bUoZq/XgEsg0ie2nbQK2t+eIIWFKYgg8X/ZY9xOahrC3s4WmJZyserxw8ypuXNvH2fka/TTh4PAM/XrEpf0d3Ds4UmAxOrHjJF5PyBMyBTc4hhBx+fIuzpdSyMgicG08UspoohjwMkvt4bgWQpCJEVSyYmUsspEJTdsgDZMw/lAkmHbWgDIw9knVMJpSwCSWxlQ8VkpR8xXFFuM4gMhUdTZP8n/bEnhijT5nBASkTKCmBWXG1I9Ay2i7ErSX3YW7qHIihEBboN8m6JA16OYE/XExCxjHjPVa7BKLGWG1slgPQtsFtK3IWroMwCDMZxEhMJbLjHXPCCNjMSe0DWEYMoYRmLEy3kAwD66RGbOO0ESSaOqxVEH0tB9KacmpbQnVlE5gk1jbfq33YHlz2GiXO130dnz43/pjOa7SBtiGQ38zQge3PT5O+0hGQER/CcA/B+AuM/9TeuwKgP8YwGcBvAngjzLzIQlU+HcB/GEASwD/CjP/il7z0wD+d3rbf4uZ//JHd09WTIBM4mp5hoODQyzPzpCG5F4fdVCB5IAvk2VohFStkJKIyNnzFcnKNCJoczrbWuDaM1exu7uHaRK0OEJUSuM4ggAsl0ucn56haxss5q26zwXffIGkbNxs1qGNopslZkwT4/z4FGmq7Qzs6NAopalhfNnISlUGIAYl8ygyAYe5ypGOojaR1yrMLwQtwE1AbKAqJiHojaL0rZ1txLYRYr3usV6uMY4jdhYzhEB4594phj5h9/LMPXSsjKaLrkCJfgUDaiw3e0GMEbHtEDuJ81gve5yerdWwWdIli3TB7gO/XI8YJ0n9a+6zRJ3rzaFMbtK8Og01rv8NIWBnayFRu+hx/coe+kRYr47QBHE3zsxYr0dsb81wdt67cXzWAN1Wh739fUzTiJPTJYZ+wqxtMQ5J1URQpC4oeH3eI69HUCfeNOOk+QSM8XraBCV8bSNMswlIWavg6Vq6ceMSluc9xoMTZaaaQkXXhKkHKYhqAyAgNKDAiDFL+urEGlTHSsMJbWTEFhhWCZQyWiJkBCBH8CjSRDcPVaxIyXpazzOPwpQyTOI2KTJLBjwqKRVq8LCYRwyD7I1xlFoM9oz5TNV4sDUuTCUGkci6ltD3IgWse2BvJ2Axjzg9T+gHRjcTSUGXhTAtErS/VvVTbET9lLOk/o4xoOukHzV2NxGsACf4O0DVVKTSD6DqqYoekYkKNYe8QPFMgOILxzf4Cz/idzDEmP7xSAT/PoA/B+CvVMf+JIC/x8z/NhH9Sf3+vwbwhwB8r/7/YwD+PIAfU8bxpwH8iPb2i0T088x8+KFPJiMAwDQNWC1PcXx8DJ6SpmkVZGvB88YTGLYTNUdPCKiVts7jiYAQwSnBU1JrmoJrV67h8qVLSFndIYkRiCXl9WjFtTP29y9hZ38bi46BKYOaViJCzwdBW4mwOh+w5MqXmYtKwPoNOP2vhAHdSJkBFILui4BE15stlXT2XSJRtk1AO28QidBoRC4RIareIpr9Qr2QHpz3GNYj9vcWiF2Lppvh6OgUed3j0iIg9T1Wqx47DYHVn3ycGIMy1TQlzNqozEolMAaW60G8j2YNxG2SsexHrFcDLu9vI7Tk4zFNycciJalRUOcFsnxmxtrcQEtA24m+l1TVBCJMiRHbiP39LVza28LYJwyZcenSPrqtBU7fu4fFVofVckBsG2ztbgMNcP/+KZouoFvM0CXG2I8gEKacwf2EcRixszPH6ck51v2IaUpoZy2GYVLbj+qRAfQBGDMDq0FSZjRWRlO8Cay2AkEykJLOeYyExbwBM6GZtdjZWYiThOr5jRBPk0TRSyI5RqcqsLaLmpKBVM3SSL6nyOhHiVg2j64YCG0DIDPGMSO0EQ0lpDygaYDZdotuYdJALWFy+Z4YOSjhN7UQxMHBEtxZJb9J12oGoesadF1G1yWs+4zVWqWBQJjNAro2wGphGxE1oi5xMpXgrGun6wK6MWMYS0U4QPZx04jU0K8lcI2I0HUBi5lQ32HMGCZgSiajVZp7Q/1ugytSt2kLwJAgM03z7n33MdMrVQVt6bEctBYFAJyR1iqpQh5RjhawuGnA/ubaRzICZv6viOizFw7/FID/ln7+ywD+Swgj+CkAf4VldfwiEV0iopt67t9l5gcAQER/F8AfBPAfPU5nLzJRrv7SxR/o4fM3TnCqK00Zut8rxAjzorATJO+KRD9GdWFrmgbz+RyzdsKUBlCMGCfbKJJDPjM8TYQtyKyr14siVWH9qM6hzJ65dJoSiFliJmYzUEOqv5VcPA0BqR+QUsLubIbteYM2kmZCNJFBIE7O8o5m1MosEZ2kkIQz4+DwHCfHZ5jlhP1uge1ZRBc6Te0ckdgkEsbxqbjc3ri66xuHIaq407M1tuczzNrGx3jUPDTK6oqveU5oGjGGN0owLQo5alDS6fkaoR914zWIUTzLXnzxOq7fuISmaeRZIeLXv/4W9na38fzz10AgvPXOXYyKRsfViJOzNVbrHvfun4Ao4PU33kVKGWfLtUTAZjEBhxhF/54zEBOG9RpjG1ztNaaEaZWxtT1DgqjAQtO6O67VFMhZvV5CwDQMGIcBOU2gzJi3hDFlrckstTAIwHx7jkARIUScniyx7gch+qgQqY4liDQvlBCGaRyRs9h8mlZibSgAs9kcq3WP8/MVhvUo7qKjjPUEBpqI+TyKpBCg2T9FyrB0EBTgRLWoqMrGc4mW4bmHKiEdQEkeyBwQGsJyNWnyO/l9ZysKg0JAnGRtmJ1otVZJfRKpNsaA+UyNxwC2duaYZ2C1mtAqgwqRRH2kfe/Hsi+jUuRZF9F1ADNhuTLJxshFxQDsc8UcBFwKAwxNQBOANiZxcaXKdpIZQ88CVJrgjCIlVtuHMTj2dVMPnFfOc42QOCiI7fGTq1n8DDPf0s+3ATyjn58HUFdFeFePfdDxhxoR/QyAnwGArZ3tiqhgEzrX13yLL0Fkzl4PM42NxE0f8QATWTcgSb0Zvslm+KOW+/ihMzY7ZOajovbZNA6L1FQ6wzD1zcMvFRRmUrVJSY+LA60tWIAoovJ9cLfbTdbKrifeEJH1dzcOQ139tJ8xRqlAp8bkppWqdPPZTHLqHB5htRpAIaProhuab16/givXr4JYnPfSmNG0kjeIk7g5GlM9Pj4G0OD0fInj0yUIQNu0KLViJUIWLKmdOYthWsYwoW+AranTvD3q6pkyzs/XCCHKvaaC+My1lwHktZRIndIkTC8EDP0ESgExAo3GC1CagBBwaX8PZ2crxBAwAGITqYiS59iHrOkMMT5PU8LZ2QqzxQzzrpOI76bRAvdZ4m52As5piUFzGQFS/HzdJ8zmLZjU3pVHpJQwEGO5GnF6Ngjxjuq6TCX4UFSPhDaKbSlrgCGB0TQBsy6iaTUXVKu2jzbisurms8YxqKzheY1Szrg6JEwZmmZC7EVKHERNFpSgMpAmYL1O6C0/D4D5PGK+EBXtNhHGNGIcGcOY1YhMKpFoevmH9onG7GzsR/8Jlt7eHERMRRgbUbkC0b3Z8jSCYYkPowY+TuhmLWKQnFg5ZeQsktJG8jySuQYA8zg0L0cKH49q6EMbMzM9lLP127rfzwL4WQC4cuMaX7Sn12NfJuRCewTT3viJL5748Mm0+dSHPpYTH9WDR5xIFw7zhd8+ovHFe9Kjjn70NHzQo4x3hWpQyzA9fJUxZh8lF3dquba8ND3qyXaIql+ZN2dBQabovU1NRDCR3T2TqBBkk7QMfRIEiblNSfub8+S2IWiRH/HJJ8QmOnLPVumONX04ByRN3ex9sDHLjClJjhyRuORYEyKaLkrW3JzFOyxnzGYLLLoOnEzFpRKajoUhe3EllrgPq7kRY9Q4FLjEQZB3jyRSSaoYrK2ZJgYs1yOYAjiLVCsuvwHTmFzlc3a2xt7eAggBMXagPABI2JoHAA3OztU2J4YKTCy9Nc/KYcgYh+SpKFiRQqPZXqGux5YQUKRsuIoDJLYrYR5ShGdMotrpmoCGGLNWkg1SCKAo5zSaRZWzqC3Pzie8f+scALB3aQvPXttyZ4nLyxG3bp/g7GTE8WnCfGuGnctbaBoCccLR/ZI9llEAYtlpNd1gzxIgBudq0+uYli8icbgfEAUgEGKQeYZqERKLR1MgAoeiaTBJ1JUbet/ZrMPV65cf3msf0b5VRnCHiG4y8y1V/dzV4+8BeLE67wU99h6KKsmO/5ffzIOcLrlMWdqjkLwRR6cxj7rhhZtcpNFGeDYeRI96Vg3MqqXBm8vjQ1uRKj/0iod/udjrclYhk49+3Ac+paYXKqo+ChNt3ItoM87loZMfPRiGYC5eUqSii9IFbVz7qFZnQq3vWM+lCG4udz3MYB/VLhjAbXytLw8vzc2xv/j5kb9RRTjo4jNRTXU5bkwIULdXy0WlDBSEykOuZqAMcAbFVh+jkhmXyGgi8pw9gRqd6+B1KhZdRN8zRs3sG8w4CpJcXxBQMVHZI8KUyRNHiuFY3TcTIxOJo0OydCrVSJK4yMZWgjnTJAGiNpdEkuaiaaOkhiGAAhWVjBLik+MBPKmLbpT8NIvZAutGbFwM4NLlHVy7tIOGGF87fQ+r9arMBG+u0c1m2XPNfZiVcWMj4eHGdNocqjRhqtY6R5WDI4KkRtHcZY26GRuDTQno2gY3bj77iL59eHt8GULazwP4af380wD+ZnX8XyZpPw7gWFVI/zmA309El4noMoDfr8e+qcYfsmXLxvrmhRLWCy9e8dDk4OIG50fQ3m+C3Nc0+9Hc64MOf1AHP4A4V2N0kX48VttUlxW880EH5Msm4a5P+WBm8hCR/qgOU1kPFw5vqvMAmH66foihZuFzfBFbbN6veoNvcoYe2Yom7nFXanUPKNN9iCFstkeP88WQwguI9lvp0CPp4CMObgoj1akFqlnsyAYYKkAabEyBAtqmQdO0qNPKmwGWVZKYEmM9JPRDwnI94mw1umdWAmGdA07XCcfLCadrYOQGsRFmFwPh3u1j/Oqvvosv/pP3cbqcPuS9yyoxyTSoa3UThQk1mpm1nCoo06U0TcQYkIE8adZiBmJE7Fq0bXQvIGYbC01uyMZsDDswTk/OcHT/6FEz9qGNPghd+QlE/xEEzV8DcAfi/fOfAfhrAD4D4C2I++gDdR/9cxBD8BLAH2fmL+h9/lUA/1u97f+Bmf+9j+wc0SmAbzz2W32y7RqAg0+7Ex/RnvbxO9Oe9vE70570Pj7p/QM+uo8vMfP1b/ZmH8kIPs1GRF9g5h/5tPvxYe1pH78z7WkfvzPtaR+//fak9w/4zvfxW1UNPW1P29P2tD1tv0HaU0bwtD1tT9vT9pu8PemM4Gc/7Q58E+1pH78z7WkfvzPtaR+//fak9w/4DvfxibYRPG1P29P2tD1tH3970iWCp+1pe9qetqftY25PGcHT9rQ9bU/bb/L2xDICIvqDRPQNInpVM5x+Wv14kYh+gYi+RkS/RkT/Mz3+Z4joPSL6kv7/h6tr/jfa728Q0R/4hPr5JhF9RftisRtXiOjvEtEr+veyHici+r9pH3+ViD7/Mfft+6tx+hIRnRDRv/lpjyER/SUiuktEX62OPfaYEdFP6/mvkKRb/7j7+H8ioq9rP/5TIrqkxz9LRKtqPP9Cdc0P6/p4Vd/jmwp8/zb6+Nhz+3Hu+Q/o439c9e9NIvqSHv+0xvGDaM3HvyYtjPlJ+h9ABPAagO8C0AH4MoAf/JT6chPA5/XzLoCXAfwggD8D4H/5iPN/UPs7A/A5fY/4CfTzTQDXLhz7PwL4k/r5TwL4s/r5DwP4/0ACEn8cwC99wnN7G8BLn/YYAvgJAJ8H8NVvdcwAXAHwuv69rJ8vf8x9/P0AGv38Z6s+frY+78J9/rH2m/Q9/tDH3MfHmtuPe88/qo8Xfv+/APjff8rj+EG05mNfk0+qRPCjAF5l5teZeQDwc5AU1594Y+ZbrMV1mPkUwK/jAzKnavspAD/HzD0zvwHgVcj7fBrtpyBpwqF///nq+F9hab8I4BJJzqhPov1eAK8x81sfcs4nMobM/F8BePCIZz/OmP0BaIp1lvoalmL9Y+sjM/8dZrbcB78Iyd31gU37ucfMv8hCKf5K9V4fSx8/pH3Q3H6se/7D+qio/o/iI9LifwLj+EG05mNfk08qI/im01Z/ko2kLsNvB/BLeuhfV5HsL5m4hk+v7wzg7xDRF0lSeQOPny78k2h/DJsb7kkaQ+BjTLH+MbV/FYIKrX2OiP4JEf19Ivrdeux57Ze1T6qPjzO3n+Y4/m4Ad5j5lerYpzqOF2jNx74mn1RG8MQ1ItoB8P8C8G8y8wmk+tp3A/ghALcgouWn2X4XM38eUiXuXyOin6h/VATzqfoKE1EH4CcB/Cd66Ekbw432JIzZhzUi+lMAJgB/VQ/dAvAZZv7tAP4XAP6fRLT3KXXviZ7bC+2/j01w8qmO4yNojbePa00+qYzgg9JZfyqNiFrIxPxVZv4bAMDMd5g5MXMG8P9AUV18Kn1n5vf0710A/6n2546pfOibSxf+cbc/BOBXmPmO9vWJGkNtjztmn0pfiehfgdQS/x8qcYCqW+7r5y9CdO7fp/2p1Ucfex+/hbn9tMaxAfAvQmqwA/h0x/FRtAafwJp8UhnBLwP4XiL6nKLIPwZJcf2JN9Uf/kUAv87M/051vNap/wsAzBvh5wH8MSKaEdHnIPWb//HH3MdtItq1zxBj4lfx+OnCP+62gbyepDGs2ieaYv1baUT0BwH8rwD8JDMvq+PXiSjq5++CjNvr2s8TIvpxXc//cvVeH1cfH3duP609//sAfJ2ZXeXzaY3jB9EafBJr8jtl8f5O/w+xiL8M4cZ/6lPsx++CiGK/CuBL+v8fBvAfAPiKHv95ADera/6U9vsb+A56FXxIH78L4mXxZQC/ZuMF4CqAvwfgFQD/BYArepwA/N+1j18B8COfQB+3AdwHsF8d+1THEMKUbgEYIXrUP/GtjBlET/+q/v/HP4E+vgrRAdt6/At67h/R+f8SgF8B8N+p7vMjEGL8GiRVPH3MfXzsuf049/yj+qjH/30A/5ML535a4/hBtOZjX5NPU0w8bU/b0/a0/SZvT6pq6Gl72p62p+1p+4TaU0bwtD1tT9vT9pu8PWUET9vT9rQ9bb/J21NG8LQ9bU/b0/abvD1lBE/b0/a0PW2/ydtTRvC0PW1P29P2m7w9ZQRP29P2tD1tv8nb/x/TXYX0WbwuhQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "rootdir = \"/home/aisl/BISA_AI/ai-hw-acc/dataset/cityscapes/\"\n",
    "dir_rgb = rootdir+\"leftImg8bit_trainvaltest/leftImg8bit/\"\n",
    "dir_seg = rootdir+\"gtFine_trainvaltest/gtFine/\"\n",
    "part = 'val/' #train val test\n",
    "town = 'frankfurt/' #pilih salah satu kota\n",
    "\n",
    "print(dir_rgb)\n",
    "print(dir_seg)\n",
    "\n",
    "#------------------------------------------------\n",
    "#load input dan outputnya\n",
    "\n",
    "#RGB\n",
    "#list semua file yang ada pada salah satu town di direktori RGB\n",
    "list_file = os.listdir(dir_rgb+part+town)\n",
    "list_file.sort() #urutkan\n",
    "idx = 0\n",
    "filename = list_file[idx] #ambil salah satu file\n",
    "image = cv2.imread(dir_rgb+part+town+filename)\n",
    "imagex = image[:, :, [2, 1, 0]]\n",
    "\n",
    "#SEG\n",
    "#list semua file yang ada pada salah satu town di direktori SEG\n",
    "list_file = os.listdir(dir_seg+part+town)\n",
    "list_file.sort() #urutkan\n",
    "idx = 0\n",
    "idx = 5*idx + 2\n",
    "filename = list_file[idx] #ambil salah satu file\n",
    "seg_gt = cv2.imread(dir_seg+part+town+filename)\n",
    "\n",
    "\n",
    "#visualisasi image\n",
    "plt.imshow(imagex)\n",
    "plt.show()\n",
    "\n",
    "#visualisasi GT segmentasi\n",
    "seg_gtx = encode_labels(seg_gt, mapping_20)\n",
    "seg_color = vis_seg(seg_gtx, SEG_MAP)\n",
    "plt.imshow(seg_color.astype(np.uint8))\n",
    "plt.show()\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input image shape torch.Size([1, 3, 256, 512])\n",
      "output prediction shape (256, 512)\n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXkAAADKCAYAAABAKjBSAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjUuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8qNh9FAAAACXBIWXMAAAsTAAALEwEAmpwYAABDb0lEQVR4nO29eXgc13mn+56q6n3Bvm/cd4oUJZGyrIWSJVmSHcuOYsWyEyuJr+WZ69wk93EyYyeZJDOTZDL3Zr2ZXOfKsR07dmI53iRLtiRakilZsiSK4g5uAAkQALGvjd6r6tw/ugECBEA0wG50Azgvn37YdepU1VfVqF+d+s53viOklCgUCoViZaLl2wCFQqFQ5A4l8gqFQrGCUSKvUCgUKxgl8gqFQrGCUSKvUCgUKxgl8gqFQrGCyZnICyEeEEKcFUK0CCE+n6vjKBQKhWJuRC7i5IUQOnAOuA/oBA4Bj0kpm7N+MIVCoVDMSa5a8nuBFinlBSllAvgW8HCOjqVQKBSKOTBytN86oGPKciewb67KQgg17FahUCgWzoCUsuJaFXIl8vMihHgCeOLq8vXr1xMIBPJgkSKrVO+EnhML28bphbV3gr50f5ajHWfoOPR81vZnJaJI28ra/hSKeWifr0Ku7qYuoGHKcn26bBIp5ZPAkzC9JV9cXExZWVmOzFIsGbIfKisXvl3AAY1zvvRlnYpNe1l398eztr9Lb/6QcH/H/BUBpIRQNwRrU8ujnRAbm3ezZDJJf3//dVipWE3kSuQPARuFEGtJifvHgOzdSYrCR9qL267vNFRuBXcwu/bMgdA0dM2Ztf2tvf2RzCtLCZFB8JWnlk8/B4Ot825mmiYjIyOLM7CA6enpYWBgIIOaEt1pk4jZCA2chgdd1wmHw9j2tf/uPB4PhnFF9hKJBAsNPkkmkwveJp/kROSllKYQ4jeBFwAd+IqU8lQujqVYYcTGYPAC1O3OtyW5R4grAr8ADMOgvHzh2xU6ZWVlGYmnZtjU7umhvaWf8pJabmz6ZYqLi3nmue/QOfwOZlxnvNc3Yzu/388jjzxCXV3dZNnBgwcZHBxckJ1dXV2EQqEFbbNQent78Xg8BIOpxk53dzeWZVFbW4umpeJlBgcHiUaj8+4rZ85PKeWPgB8tZJvy8vLJk1KsYtrfgMEWWL9/USKoWJ4IIRBCzFtPk4LdCTc77/QTktspc6bcu7/64bvY427FNjXiodTbmS3h2yHBQGsJQc8mGhsbp+9L0yZFM1MaGhrmr3Sd1NfX43A48Hg8SCmpq6vDtm2Kioom7T1+/Hh+RX4xOJ1OHA5Hvs1Q5BvbhLHL0PxDqN0N1dtBz55LRbG8cRs2t9VF2FBm8eyYZELmnALWuyS4LPBFMR0mkZIIny2J4B300X+5iNfCeTU9Y6Y2doUQFBUVLXpfKq2BonCJh+Dia9DyCpjxfFujKBAiSZ0vvlHF2LiDbZ4JH7zEI6YreNKdZKR+hKQvwWjDMGO1o0tvbAFQUC15hWJW+s+ClYSKzVCxMd/WZJ/BC6kO2Ohwvi1ZNtxUH6YkkOTN0JV26j7vi3NvIMB0mktgWeGhRF6xPBi6AKMd0HUYNj8AnuJ8W7R4EuGUO6rjUGo5FgJLvalkStBt8pGdQ0RssO3jBLWdjNmlM+q5wi48Ix6ixfP7rVcyyl2jWD5YSRjvg7Y3UuGHy5XOd+DMjyE8kPoogV8QhiYp85p4NLjBM44uZrbQ3+nwEY/r+AZ9CFuADZ5RTx6szT9K5BXLj5F2GLmUbysUecYpoMpx5WH/dvReBsMGSQsOd/o42+ch2e9H2AKBwBldnZ33SuQVyw8rCb2nYTmmD1jObyAFQiSh83aHnxETvjpgMGYBCCJ2gGdOlXCmz0Nzj5c//HEjf3uwhpgpsIVktFp1vCquG0lj4xhCwKVLQaScP+ZXsUgGzoHQoGZnatlbCoYrvzZlQngg9YBSLJqYqfHXP63B67IYsQT7b3dSFIRLw05eO13MT84VMRpLSduhdj+tf3ob99zTztateTY8TyiRzxqSvXu7+cQnTiEEvP56Pa+9Vs/ly35AiX1O6D+T+gBUbYMN70uNIp0TizLn8wiSDCYeROIEJlrWgiX5naQFViL3x1nhRJI6kaQOSKSd+t2SliCS1AhUjMPkfScYGvLw/e9vYmTEz3WEmy9blMhniQ0bhvn4x5txpH2Ed93Vwc0393DqVBmvv14PQEdHgFjMQXX1OD09PpT4Z5Hhayfj04hS4fouAeMdQOLQ+hhN3k6l6zsITCLWBvrjv4SNN7d2uouguFH1KWQJX0UEhyfV8WoYBpXrbLz1Q1hxncjgld/SsjSOHfOxb18cl2sZvPFlESXyWUAIyR13dOJ0Tk+O5PMl2bu3h717ewA4dqyCsTEXmzcP8r3vbeb48QqU0C8FFhWu7xF0HJos8Run8BtX0ikFtcMA9MU/jkQn67+LlKnkY/1nlMBnDUls0IsZM8AFgUCAYmMrHSfjRIdmPqwHBvoYHu6nurqO1XTfqY7XLCCEZPPmoXnr7drVzx13dFJZGeXXf/04FRWRJbBOUep8gYBxaN56AeNd1vj+CJeWYarghRAZgnMvpgY+KbLC3sZxvvZYC43F091fUwU+dW8Osm3bAA891MqHP/z3+P1Xft/rSRewXFAt+TzhdNpomoq0yCrSTnXGYpNKfpoibG6nyHgDQ1w7c6AQEoMwxY6D9MV/mVLnAeJWHePWLsBGYCPRpu17QbbZq3PEZa54dNcgxR4LLSnhqkArw7D5yEfO4vWa3HRTD4Zx5V5rbHyJ5ubHAY3Nmzdz4cLKfvAqkc8CUgouXQqyY0cmubAVOSERxtH5HJ515Xj0VsbNXVdW2TVcjj1BjfurOLT537gCxmF8xkk04kiHjh4fx6O34jVOE0rezEDiYSQqkV6h4vUm+MQnmtm9u2/WfviqqkNcvnwbIyObl964PKBEPgtIKXj55SYl8nnGJ45S4U7lBg863pksj1t1JOzKdCt8foSQ6MRS3zGpdP/75Loix88YTt6NKdXsZflE11MThgDopN6QIpFxhoY6eOyx09x4Y9+c22qahRCLnNRmGaJ88llibMzJwMDMYdP9/R4OHarm0KFqxsdV6y8fuPQuAo4jOLVsPIQlla7vZGE/iuvhve/tpLExNVXibd4fA5KKiiH+4A+eZc+e3vwaV2ColnyWuHw5wJe/fANPPHEUpzPlIHzllSaOH6+gszOVG3r37l4+/eljk6+QLtcyHLFZoBjuJMH63M7WA6kwfMEifOt2MvvGrGJ0XU72aRkiiY7JmjUH1D01C0rks0h7e5A//uM7mBhgY1natFGvly5NnQgAPv1rJ/mnv7+J9mH3Upu64qja0Y/Tn/tBRlJC1FqPW7tIzG4go1tISmj5aa5NW1V0d/s5WhtlrxuC2hB3+p7BCp7Lt1kFiXLXZBWBaWqYpo5p6tdOayBhU8TNb93eTXlRnDgaV8faaJqGbdvLatLgvKHJaw92zRInXx/l5X9+Fbvri/iSTxMfHyEZy2C6IalamNnkzJkynm73IXUbIaDJeRqPNp5vswoSJfJLSCym03WqAu+QF9cJcPc6+cbLAeLuEV5NVtBmBOkaC7FvXwd33NHBzp2jtLWdZ/36hU00rMgdbp9ObU07Xs8Yz//N13n1Lx/n9A//33ybtSqpqw0xsHYAS1/8A9QwDEpLZ+aiX0kod80SEok4OfuzRrZu7OPsn+sEqkMEisr5wferkN5BXLeE+cPPH2X3liiGAV1dcOAAHDuWJBBQE1pfCzuhI+U8qWuuAymhP/4ovi1BqpucDCVg52Mm0pI4fBkMqHF41MxPWWZ7Y4hYUYy+jX1otobltJAyFQThclm43dPFP5HQJoMfTDPVvnU6ndTW1jI0NH9o7XJFiXweMGOC5wdK+Xm/xa/GzmETZjBicLY9gv12lK+/nap3+jQcPQrB4Di7dkXxeFbnpAeZ0H+mHF9VbmdpLnG+RFKWoItbGHPeSrE/ww2FSM1mdfo5GFeRH9km6U0yMrKOsc619PcP8q1vFbNx4zDr149Mq9fREeDQoRoAbr+9mMrKPBibB5TI55FRYfBNVwWfjV1CAkNtJvafw38AxqbUGxsbIxwOK5G/BmZcZ6i1hEBtCN2w0Z3ZjYMWAhxiEAeDIDXGzH3oYhxLBjLbgcsPxQ1K5LNIKORESpBSp7f3Fjo77+HEiROEw+c4erSKo0er5tw2FosjpUQsRUdOnlEin2dGhc7zzhJO6l6+EO3EKS31oywGKRhqKWGopQRPaYzqXT0YrlwOeLGocX+ZvtjHSMjq+asnIjDWnUN7Vh//8i87uPfeNoQowzDuXtC2x48fp76+Xon8fAgh2oAQqcwRppTyZiFEKfAUsAZoAx6VUmbkjFzpKUB1ISlyW4xNvRpCcNzwIaREAj3MSMOhyJjUDRsd8mAlDAxXbkIq3Xo7jZ6/xJJ+TBm8duWxy9D6U7BMiI3kxJ7VSiTi4JlnNqJpGrt2XWTdunX5NqkgyUZ0zd1Syt1SypvTy58HXpJSbgReSi/Pi2EYrF+/PgvmFC5lvtQs813fjDIjXjLNnwJXT1IWCATwenOc51yRMZqI49K7sXHNn3++9WBqNigl8DnDtm3i8bgKNZ6DXIRQPgx8Lf39a8CHc3CMZUnSEgxGDHab4wSuipveY4c5/mA5b2szB0YpkVco5icUCnHx4sV8m1FwXK/IS+BFIcRhIcQT6bIqKeWE87EHmLX3QwjxhBDiHSHEO7OtX4kMRw1+eKqEUmnyiXg/DySGWGdFQUoOaT4++2KY43Ys32YqFMsSy7JIJlX6iKu5XpG/XUq5B3gQ+KwQ4s6pK2Xq/WnWdygp5ZNSypunuHlWFWvsOPuTY/xKrJ91dow3kqOcMNWIvWzRf7oc9fa+OvD5fCve1Xs9XJfISym70v/3Ad8H9gK9QogagPT/c+f8XKVYHwpiPV6M9QsBPNjsjQ9w3grP5aZXLAIrsYiJPRTLkkQisaIHM10vixZ5IYRPCBGY+A7cD5wEngEeT1d7HHj6eo1cSZhOE+tOD/ZDAeRWF7JM5zekRZ/KUqhQLIpkMsnw8OJGE1uWRTic20F0+eZ6WvJVwM+EEMeAt4HnpJTPA38B3CeEOA/cm15WpAkXR0k6TaQA7fUI1v9ZzpBfpRBSKDJl//5LPProacrKopNl3d3dxOPxBe8rHo+v+M7aRcfJSykvALtmKR8E3rfQ/QUCGY4cXOYcPNjAw0XtlG0ewLdJIDY6wSNAueOzim0KklEDpzd386rqIoQuRrHkyp8MupB4660adF0SiVyRr+HhYfr7+/NoVeFSME3IpqYmNK1gzMkZyaSObQncITfiAyUZbWOoIbALxowZjPdkmlxmcXj0djx6y7UraYXRN6DrNpq2Mqa8i0YdjI87se3penHunMonPxtKPvJALBDDdJoYcQMB3LfLQ1cihukysW04eBDMdAPU4xF8/OMB3n03ryYvO3SXha88km8zoGobhHqW/LC6brN9+wBCpLrzH3zwAr29Pt59t4qeHj+9vb4lt6kQ0XWdYDDI2NjY/JWXKUrklxhdtxltGGagbpzqM9UYcYO//W0/A2tjmG5IJGD/fujrS/0B1tQ0cfhw45JMiLGSsOIG4X4vrmBuZ4tKTQUomUipMImUqSn/uo/n9Phzcf/9F/nAB1qZ+nLc2Bjillt6uHzZz9///R5GR1fOjGR1dXUMDg4Siy1snInL5aKxsZGTJ0/myLL8o0R+idm8eYji4lQHUe/m3kl9kOn5Kh0O+C//Bf7oj0rYtWsXDodjVSRRWq5UuH7AuLkbyZS8S9FhGL4EbW/kZW7X0tIou3f3MZf3s6ZmfEXNhWoYBg0NDYyOXp0QRAFK5Jec5uZyhofd+P1JpJgZGS8EVFQE2blzJ06nMw8WKhaCwEQTCay4Ba2vgLQhNgaR/M3m1dQ0Sm3t6unJ93g81NbWLrg1blnWqugHXPlnuAxxuZy43SvnVXolI0hQ4/4yxsXvw2ArDF3Mq8ADHD1axZkzZXOu7+wMEIut7vZdPB6nubk532YsCUrkFYrrQAjw6BdxiUv5NmUSKQUvvLCWkREX4+OOaekdhobcfOlLuxgbW9lpvTNBSomUEtPMXZhtIbC6H+d54vTpMurrQ6ozVZEzWlqK+cM/vAOPx+SDH2zlxht7CQYTHDtWycCAmmFsgmg0ytmzZ/NtRk5RIp8HnntuPYZhc9ddHej6dL98KOTgxz9Wkx9kg9GOIIG6EA53bjsZI4MeosOF5l4T2LYgHHby1FNbef31Olwui46OADMigVY5Kz0PvRL5PJBM6nznO5uxbcG997YDqYi7wcHtvPCCpK1NjaDMBmbMQFoauZ5ry0po2MnCGPQ0F52d88xgpVixKJ983hC8/XbNtOUzZx7h8uXtqJbW8sJbFsVTEp2/oiKnbN++Pd8mFCRK5AuI3t4IFy+qWN/lhu600RwrI2XAcsbnW/goXsMwqKioyIE1hYMS+TySTOqMjzsASCSCSKl+DsXKQzMsqnb20nhbB76KMHNOcJwHnE4nVVUzJ6/TNA2Hw5EHi7KP8snnkd5eL1/+8g34/WcxjP1omop6UKw8DI+JuziOw5vEUxYl3D8xX7GkqGEM3WVhxgzGOq90Ck/0heYrAq24uJiSkhJaW1vzY0AWKViR37VrF2fPnl1wLorlheDMmVLi8T243VHc7iP5NkihWBSxmItEYo4R2mMBRgaL0XSJGTOwLYFuWDTc0EnpumGEBomog0jIgzmaius/fXo7JSVDVFd3zyv04XCY5uZmNm3axH333cdrr722IN1wu93ouo5lTe+g93g8CCGWffRNwYp8SUkJGzZsWNGJgwCEEJOjW1f2A01RaFytXX19VYyOXonsqq29jGVp9PZWz7uvoaEyxsYyjwpzeuJEnH68XRFKq4Y49+4m+i9UEAyOUlnZS1NTG8eP78KydOrqOufcz/BwKWNjQSzrDI2NjQQCARobGzNOOyylpKmpiZaWFkZGRqat27RpEy0tLcv+vixYkVcoFNknFPJjWalwz56eWvr7r3Q6JhJOLOuKJFy+XIeUgmQy+zmUElEXR17eg6ZbGA6TRCw1AndoqIyurnqEkCQSLs6c2crFi1fGjZSWDlFZ2cOZM9sASCYdWJZOIuHivvs0hBDU1NRkJPLt7e00NTVRUpJKBnjw4MGsn2choEResSqRkslc63OFrN60fhif2+Llgw0Y5dNDJKUNgy2l+KvCuIILn3YuQyunLekui5K1IwCE+7xEhzzMZftEK900DVpbN0yWd3fXZizaiUTuUx/Ylk7Cmj7GYKp9tq0TjXonl7u6vHR11c/YT1vbWqJRD0ULGGJimuaki8blmn6ujY2Nme+owFEiXwAYhkFxcTFDQ0PY9soMxQsEAoRCoXybAaQEcKS9iHtu6ua1czr+stnFzGtptLxTSU9LKfXlXdPWDZwrY6StiNFLQYQusZPZjYzSXSbu4hgVWweuyLgAI50iOFgXov1nDVjx2W/h0dFijh/fhW1raV/56hl74XA4cLvdC3Kz6LqO1+slEklNNFNdPb+LarmgYvYKgGAwyJ133snGjRvzbUpO0DSNzZs359uMSaQtGGop4dAb1Rx+ae4Zgb7xze0cPNhIPOwgfFW+l5SoC2xTx4obSDt7t5LQbapv6KX2xl4cbgtj4jMlB7xm2Aht9g7B0dEijh/fRSzmSbfGV4/AAxQVFVFbW5tR3ba2NqSUeL1ebr311snO1pVEQYu8z+fDWCUTnAohqKyszLcZq4LEuAMpBRculNNUvx0zfu3bwE7qxJcwa6OmSTwl87dCK7cOzCizbUFfXyWxmArHzYS+vr7J78XFxdx2223s2bMHr9fL2NjYinizLmiRr6+vX9QotuXGunUrPyFZobSOpC0ZbitCWqlOuuRYgNhV0+BJCeO9XswCz7ku9JktedM0aG9fs/TGFBh9fX2cPDnAqVPbSSQcMyKJpjIRIimEoLi4mDVr1iCE4OLFiyQSuZ0+ciko7L/iVUJ5eXm+Tcg5lZWV1NXV0dXVNX/lHCGl5MhPR+g5ZbF+vX9yVqDYiBvklYeQldDoP11+TReMRNDfX4Ftz//w8vnC+P3h6z+BKSTGnfSemDkc/+zZLdhZdB0tV+666y4OHLAIherp7a2moeESGza0TOlsTxGNRjl9+vSKzntTsCLf0tJCaWlpvs1QZAlN0/Iy1dpIexGV21JujbHOILHLAQKB6Tf68IWSjPdnJnRe+8EdjPUE6WqvyygVhc8Xorx8gI0bz6FN9aNLidNOeczjGpPDO50Jwc6jXs7siGI6JBLQTIHUJWiQiBh0H6nCjDmm7oqRkRKGhkpZbT742RDCBdwFgGk6aGtbSzjso6mpjeLikWkDrFaCS+ZazCvyQoivAB8E+qSUO9JlpcBTwBqgDXhUSjksUu/kfwc8BESAX5NSvrsYw4aGhhazmUIxBUFs5IovPT7moqJ88Wmco+NuXv7WPbQ1r2EhQhoOBwiH/SQSTrZubcYwUh2oa8M2n7kQR5fwT2tdnA2mQgl1U7D/QJBbX/fz2t1jWEmd6neKiO4cpW19nJawj2RkehjkyEgJR47swTRXRr6VxdDaalFdnXroPvNMgqnBXFJq9PVVMz5czMdufoXOwOp5EGbStPpn4IGryj4PvCSl3Ai8lF4GeBDYmP48AXwxO2auDoQQK3pi4fXr1y/5+SUjDi6/W83ld6sJD1zJmaIZFswykfpcmEmDV759N23Na1lcS1nQ3V3L+fObJktuHbIImuCz4PH2BFvGLLym5L7eJIYlCIR0HnqmhF/4cZBb+iV3vhzkA98rQTt55c1DShgeLubEiRtWtcADvPLKlWn85prRTzc1HrkYoSpyJVLJsqwV3Zqf946TUr4KXN2sfhj4Wvr714APTyn/ukzxJlAshKhBkRF+v5+SksxdB8uNoqKiJRd529QJ9/kI9/lo6Nd4z4DJe4ZM7mjqwV+ZuZ/88IGbuHhy7XVaIxgYqGB8fGYwQXFS8r+3xvkfJ6K8r8+c8zFSOmjwydbk5PLISAlHj+5R0TQZEpLwibFxLtpXBrdduHCBsbG5Q2mXO4v1yVdJKbvT33uAiVyddUDHlHqd6bJurkII8QSp1r4iTTgcZnh4ON9mrFj2DVncOZBq4kX6gvSsc3J2nsGfUkIoFOTy5Vqy4euORr309VUR9F7AaU9/kzAyeLEQQE3UZvewyWtGMceP35CTtAMrFwdR6wNE48fRHSEcDgeapi37JGTX4ro7XqWUUlzdZZ3Zdk8CTwIsZvuVSFlZGQ0NDbS3t+fblBWPN6LjC2uQgT6eOLGTWMw7f8UMudC6gf8g2rhleHHTEnpsuHnY4odmNfG4asEvjBPE49/g8GELIVLpC2bLJ7+SWOy7c++EGyb9/8SIgi6gYUq9+nSZIgOEEAUTT54LdF1n06ZN81fMAXURm10j0x219/ckMa5qTbssya+0x3lPjyQWc3PpUtO03CnZwJaCdzs2XNcgld3DFsOXMhvVuRrZVxfDY3wZOAyalfpwFPg2YAISKSXt7e0cO3ZsMp3BSmSxLflngMeBv0j///SU8t8UQnwL2AeMTnHrKFY5Qgj8fn9eju22JUVXdcZVxiWaTHW8dXZ24hVe/lR7jK1Dr3EiXM6rkR3pmtl+8AouWIEs7/P6EcJm+3tOoTssRvqK6Txfz/b3nEJokoGucrpaZiYGu17c3iibbz7L6be3TmaiXCw7d15JdFbisdFEG3AW3EUgzkL4ICmBn048Hp+RS34lkUkI5b8B+4FyIUQn8MekxP3bQohPAe3Ao+nqPyIVPtlCKoTy13Ngs2KZIaUkFovhdObJdywlv3A5OaO45UYff9L1eYTlIDQ+hkM4uMG5g173bbw+dpg63UWXlZsWXkxqjNsGfm2OMJA8cNO9h7nl/kNouuT8kQ1cvlDL9ttOYSV1Lp643k7n2fEGI7z34dcJlIT42dN3XNe+9u27ImeaS0NoAjDBOAiuGIRnv9aGYcwaELBx40Y6OjqW/ajXeUVeSvnYHKveN0tdCXz2eo1SrCyklJw8eZLbbrstL8ffHLJpiEwNkdPwOh/irrN3Y+iNCARMyWxQ5VjPC3Xr6WCUBwaf4dJo9scMHjVLeTZRz8fcbYva/ufJCsIyuyGTw30l2HaqE7L9dBPJuIN/+59z3f7ZITruoaetmhM/25nV/fq3+9FcGsSBZBcUrQWhgbRZV1fHrTtTx3vxzTfxFxfPmj7F6/WuiJDmgh3xqlhZJJPJvKUa3hnyU6ztwu+deOEUGFoDQly5gWMiTlhPtdpfLzrEsDGKKSwerY7yT+/6GIll+2YX18inoqFrNVh2PzB7K/JgoppQlkX+4sm1jA4WUVQ6Sk9bNUsxcjY67uX5f36AWMQ9f+UFIISgvrKSkNdLV28vdkdqak2v281/+4//kdqKVEqIvTt28Kdf+QrJZd5avxYF+5hKJpPTMsQpljcDAwMcOHCAjo6O+StnA5n6PDRyH3vk71Ls/T0c+pr0p2mawCeFyfOlr/DV6qf4avVTnPNcoN85yLBjhP6wRjxHHpWfJquJSn1GuRB+yv1/hUPfMMtWucO2dN587lZ0h8W+B99asuNGQj5sa+Z1uB40TeNvf/d3+dIf/iFez5UIJNOyaOnomAyZrCot5d5bbqFhBWeALdiWvBL5lYeUku7u7PfDByw/5eb0PEdlZimfHPwY1clKDIwZjdK4iDNipN4s3g4eocXTNmvDdSCikbBy06I9mKgmJnU8YqLTT2BojfjdvwToBN2fZDD8Ba6eISoqdQZlblIfJ2JOwmM+3vrxvpzsP5cMDtqTaQ0ATCEYAaY66hLJJP/3179OLJFgbW0tQ6OjE9OELbW5S0bBirxCMckUjXtveB9r403TVq+NN3Jj9IY5Nk39G9VDHPc3T5YPG6Oc916c99ClHhunLoma2RcBG8GP4vV8wnMBAI/jboKezyBEOoeNXofLuIm4+c607c6bAZ6O52Z6usutdbzy1N2Ehgsv+mc+vv/9JNu3X5G0ASk5etQkkZj+kAxHo/zFV786ubxlzRpuHhkhUFy8VKYuKUrkC4Dx8fEVnzff5/MRDs+fRqDULJnRKq9P1PHR4YcBCFoBPHJ+/21SmPQ7Bhk1xnit6G0sYU363BfCpVGDqJkbr6aGxgdKfg+v/WPAwu9+bFLgATThocj7W9gy9cZhWd0MR/40J7ZM5dKZpvkrLQNOtrbyxX85gWleOy/NmbY2hsfGlMgvNQ6Hg5qaGnp7e/NtSs5pbm5e0aPuNE1j3759HDly5EraBkkqqgXw2z5+afhDCAQbY+vYHL/2NIgTrXOAE77TDDpGZtSJaTGavecLOuuuACqMIEH9k3PXER40kX6oCQ8O4wZIXloaA5c5iWSSaPz5fJsxjakTlCwVBS3yHo8Hc650corrZvfu3Rw9enRaWSQSwTRNgsEgwWAQp9PJwMDMaeYyQsLm+Aaq9tRQXFzMA5vu59SrJ7Fsi23RLdw/djeQatEWW8FJ0b8WNpLDgeMc9h8DIKrHscTyHMgigS5XN2F97iiZi+4OTnvPTS7bVOFpi8DoEhioyDrnzp2jurqaoqLFp7xeKAUr8mvWrKGzszOjV3zFwqmoqMDn89Hf308gEODChQusW7eOjo4ORkdH2bt376Iy8+lSY2/4Jm6IbkdIwV3jtzGihRi7NE51ezkfHfnANbeXSH4ePMz4Va6VLZH11MWrOek7y2tFb2GL5Z8a1hY2Pyh/Aa9jYambzleuS6X+UywrxsfH6e/vJxqNsmvXriVrzResyFdWVi6+BamYl6KiInw+H6WlpSQSCQYHB+np6cG2bYLB4IL2VZWsoDZZDcCvDv4yNckqXFOiP6o6XFR1ZD7F4ZAxwhlfy7Sy097zGFInoSXzLPCSNcUWuoCLIzq2nP1GNbRUvQniJnSMFeztturp6OigoaEhp8Ibi8WIRCJEo1H6+/upXKKwzYL8q6usrMTn8ymRzzFCCHRdx+PxsOeGXdxxfogWv86F8mtkNpTgkA68todfGXoUDUFjvJ71iewMexcI7h2+A1OY08Iak1qSJDNTE+QaTaQC7jUBZR6b2xsTbKtIomtwqs/g0GUnl0MzY7zvaopze2NiMjIvmoSnz3o4M2BQ0B0Fq5QXXngh5yOyJ+aLGB0dXdLUxgUp8oFAAMMwVnQi/0JiU8ji1mEXlxvq2OgQRIv1aXJaYhazMbYeALd08RsDn8CQOh7pyciPvlDc0sUHhu7lq9XfYswYz/r+F8JtDQlCCY2mIpNNZSauKXfMziqTzeUm9iz3q1OfHnrtccAjW6P8uMVNJJlaoQnQVZbtVYPb7WbPnj3Yto2uZ3fw17UoSJEHSCQSdHauPsfjUubKMBMmGy6s4T+ZD+CxJEQBASdiOokpZpRZpWyNLV2KYBuJxKYQWrwOHT64KTbneucC7lWHDh/aPPe+MqXUY7O1PMlp9VaQc7ItxrquL6nAQwGLvMfjYcuWLTQ3N89feQWxY8cOOjo6SCZz75qoFTV82n4Mp+6EKX93e6Nzb5NrBowhLngu8fPgOySFiqyaDY8DHtkW5XunPTT3r+55XXPN7t27ef311/NtxnVRsCK/0ie1notcnLPLdrEjuoX70iGLAL4xH6VnSnCI/ImERBIXqcRQFz2XOONtYdgYZWiWuHfFdAwt9YYhJapFnyOuV4OklCSTSc6dOzej0dbQ0LCoMErDMBBCYNt2xpOPF6zIKxbP5tgG6hJX5k+/b+xu1sfXpHK4TJCHiXBsbM54W7DS0TEJkeS1oreQQmJjI5V/ekF4HZJHtkU50OoibglGYxoXRwxK3Db1QYuTfQZyFYl/OCzp6rKpq8tSQ0nC5rb1OObpGjzsPcaocaXSRN6taDRKW1vbrGI8MDCwqEiejRs3YhgGoVAo49xeBS3yE0+tlTzJ7kJxOp2srVlDT2vPZNlt4b3cEr5xcrk+UUu5VZYP86YhkUS1GMd9p+lwX0Yi6XBdVmKeRQwNHtwYB2A8IegZ1wg4JeVem2K3i8PdjnRH73ISe4nXMffjKWqKWUNXh4YkJ0+a1NVlZ3IaIWH/offwUPiea9Y76zzPP0W+zrfHvweAaZqMjl57tJqUclG6dvbs2QVvU9Aiv27dOi5evDjvBVtN6LrOI/aH2N2+fbJMQ6AVUNboy84e+hyDSCF5PXiImBZfXhqzTPE7JRtKr8Tm37M2zk21Cf7uTT9L/VjVhGR3dRLtqt89YcHxXgcuHXZWzd7vpAm4e01s1k7t/ojGt095GIoubefltfBFfHyp/auYs0wtWAgUtMjLeAI9nlxxqUB1qVFsFU8uB60rg49ccSflZikJM3UD/MLIA2yIX4lB1zSNuo5aDArnjxzAwiasRxjTQzxb9hPGDTVSOd8IkRL+2xsTvNHhxJpj4FYuuKMxwV1r4jNE3rThppokhiapDdgLvq37whdIWueAX2S+6TCamprYsWMHLadaKNNKicooQ/bwwg6YAUPWEBaFm1qj4ERes2yczRcx165n/N1TVL3RzNCN+c+K57U83BO6MysN0iIryIenDO8fSA5zgJ8BsOfVHXy05aHJdSL9r1CJiwQnfGeI6BEOBY6mWoyFa+6qw9BSLXqfUzISExzvdRBJpsSxscikNjC7OA1GNM4PLb5TvtRjzxD4CXuaihcviDsrm9hY2sSr7clpbycdozpdoely5vV6+Uj1h6juLedXAr/M8cRJngp9d9b9/lRffATN7w38l8mEeYVIwYm8M5bEdbQF8z234ETDJ5amxRq0AnjsKylsPbaHJ/ofn3SDOKSDNYmGnAhuSX8RjWdqGS+K0HC+pqBcLwBJkSSsRehy9XLEf3LaOktY9DsGlbAXMELArfWpKKYbq5Mk0/2ARS5JwDW7OEWSgt7xBM+eczMa1zDtzH5gv9PmF7dGqfbnJvWEEAKPA96/IT6tfCwuuDQ6itOIAKn+qWCPn8cHHsNd4iTEOGuNNXze+zk63Jc55js9bXuvt3jBtljS4muhf+VY/PhiT2dJKDiRr7vYP/l9i+7jFmc571yj/mLZGdnGxvi6yeX3jO9lXWLmG8NStKIdSQPPuJuKrlLc0dzM+LNQJJJ3/SdICpNBxzCnvedTK5SYL2uqMhRfryOVe+c394Y50uNgKHql4XG0x8F44uqGiGRvXZJdVYlFuWGul6BLsr0iQLC2brKsn0G+UfRdmC1SMW2f03ZQlagg4F34JCnfCD3F7/T/Z2wKO1leQYm8blrotj1NR4SUGKaF6TBAQrVZiSFTZmtS4zMDvzatBZ4pJWYxRfbCEnHlkm1vb8AdyY/Aj+hjk+l6JZKXSn5GVIsx6BhWkTCrmAmh3lMzvYP0xuokx3sdnOq/Ih+7q5O8pz6BnseXUCEEdsxG2hIx4Su66mEjpKDETKn+2lgDN4xvw2O76RMDvM67GR9r1BrlG6FvFbzAQ4GJfNHgON7Q9GHfrliS/ReaKKu4HYAPjt6P1/ZOq1PIPutM8Y5fIynYIpnqJ+x3DHLeM/t0d8f8zUS0q4a5Lv9LqsgRZV6b/Wvi7F8z3WVSCLER4TNhfJt86L7Z3bzbw5u4f/iuSc2Y+L80WZzxMWJ2nM/2f46fx96+bnuXgnlFXgjxFeCDQJ+Ucke67E+ATwMTvpXfl1L+KL3uC8CnAAv4LSnlC5kYoqPRYK+l0vDRb6bmvDxrhXFIk+3RLdw1/JGFndkKZ8gYmXfY/1H/SbpcqXj6uJZY1PR3CsVsFIKgL4bKZPl1BzO0JFv5UfjFLFqVWzJpyf8z8L+Ar19V/jdSyr+cWiCE2AZ8DNgO1AI/EUJsklLO251+79h+fl98Dqs0wQ9G/5iRSIzR0+dxiSJqXbdkdDLLgfl64c97LtLjnH8k20nfWSJ6HpPMKBTZYAHeQIc02De2h5O+M4xMGWG6EL0+WPwmXsvDluiGzDeawsl4M7/S+7+RILGo7fPBvCIvpXxVCLEmw/09DHxLShkHLgohWoC9wM/n23BPZBe6T0fXPKxx3sKBb30ROxzB6aqlTHYCuzM0Ib/ERIwhx9yDt1o9bZz2tsy5PqrFSGpLnzddocg1XstDkTm9gzNg+blr5NaMthcIApafHeEtWOm32BZPGyd9Z+eN8CpOFuGxXTikg/p4zYz1uqnjDbmJBK+4i31jXjT7SieDJS1eiLxES/JCRvYWCtfjk/9NIcQngXeAz0kph4E64M0pdTrTZQvGDqdcCxITWy7NwJqpE0RnQqunnQ7X5WllI8YYFzzt2TZNoShs5PS+sa2RDVQlKqZVqUiW0RhflBxMwz+lT+6m8RvYFtlEc3oe3FZPO7H0bAglySJ2DW8BYG2skVKzeM59+kJe1h9v4sTtZyffLra8s35atJuNTVALsNO5nZOJ5oKOjZ/KYkX+i8B/J3U5/jvwV8BvLGQHQogngCcWefxr0ucYJKYtPG93j7Ofd/2Zx7zGtQRJrTCHMisUS0GRGSBoBtgW2ciaWONkudt24ZBLE9fhsd3cNH4DADsim3nO+ilRGaO+r2ayPFOMhM6NP91OJBBl3YnGaescwsGni36NR/wP81LkIF8Z+zrdVk/Bt+wX9StIKXsnvgshvgQ8m17sAhqmVK1Pl822jyeBJ9P7mPWR6Nm0Fm1UQ07vxMfGZsgY4d3Aydk246K7nZAaVq9Q5AQhBetijayLNlKdqKIqmfn8vbnGaTu58dVtmA6Ltaca5t9gCnWt1RQNBGg6d+23jVK9hI8GPsxHAx/mdOIsn+j5FOeSc7tg882iRF4IUSOl7E4vfgSYUNtngH8VQvw1qY7XjcCC44yK9Grcwo/mdvL8UD+b6GS7+9Lk+mbveS542olry6fzQ6FY7uhSpyFWy3vHbqE8WYJDFt6EJQLB2uaFifsEZb3FlPUWL2ibrc7NfLP6y+zt2F+w7ptMQij/DdgPlAshOoE/BvYLIXaTcte0AZ8BkFKeEkJ8G2gGTOCzmUTWXM1m9x28X36Op4//VwBO1JzldPlzOZ1JXaFQXJvbRm9ib+jGFTEuJZs0Gg38sv8Rnhr/bkEKfSbRNY/NUvzla9T/M+DPrscogEbnbkDgaqgi3nkS4SnCKFt/vbtVKBSLoMgMsCG6Vgn8LPg0L/9Q+Vd4hJuvhr6Rb3NmUFiZsKYgEOg4KNp/K2gCs3t2/7tCocg9a6ONlJkl+TajYHEJF/+j/E94zP9LeET2R69fDwUr8kGtkjv8v46jvITKxz6Id6tqxSsU+WLYMToz9YViGn7Nzz9U/g2v1x9gi2NTvs2ZpGBFfswY54zvAgjw79lG6Qf259skhWLV0u7u5OnyFzjqO7UsknLlC6dwsMm5gW9U/xMbHetxiexMRXg9FFSCsmFjhM50rpVDgaMMDAxRRT1SShJ9KpJGocgnXa4eLjt7iehR1kebqEpWzL/RKmWzYyNvNrzC98d/yFPj3+VA5OW82VIwIv9G0Tv0lY8w6Jh9eq5wi4p7VyjyjRSSN4re4ZTvHI2xWu4ZeW9BhlLmGyEELpx8LPAI7/e+j0/3/SbPR36SF1sKRuT7HAOUOMrybYZCociAUWOME74xYlqcrZGNbIqum3+jVUqJXswv+x8ByIvQF6xPfioylsDsbs63GQqFYioCznsv8nzpKxz3zTIngWKSjwY+wpcq/xcf9X+EUm1po5SWhciPvnYIM6QyMyoUhUhCS/Jiyav8W+UPUtkgFbNSohfz5cp/4Nnaf2e9Y+nefApW5KW0wYwjLRsrHMXsOoI5ULj5IRSKVY1IhVl+v/x5Xiw5yIg+VpCjP/ONJjRucO3gezXf4M/L/oS1RhNlWmlOjymkzP8PIYSQj+14jM3lmyfL7Pg48ePfwbd7K+GjqZnVjfo9OGoWllVOoVAsMRI0NB4avGfRk3OsBia093D8CD8MPz93PSRfGv1nQjI02+rDUsqbr3Wcgul4vRrh8KBXbSN8VPniFYplhUhlij1Q+iqHkye4b/gOis0inCoKZxoTubhudu/hZveeOetJKfl44KN8ZezrfCP0FC7h5DPBT/Gnw/9XRscpXJHXdITDg+bz4KyuINZ6af6NFApFwRDXEnQ7e/l61XfYGtnAlsgG1sWaVP6bBSKEYItzE39R9t+43X0bTuFkXI5nvH3BivwElY99kHhHtxJ5hWI5ktbz074WLnguce/wHdTHawhY/vzatQzRhMaH/A8BMG6Pc7/3fbwYeWn+7XJt2PWiB9Ufg0KxEohrCZ4re4mny15gTM+8JaqYiV/zE9Ay08aCFnnh9IJe8C8bCoViAfS4+vlB+Y+57OwhKdT0mbmmoEXeKFuP5i4CYIPunae2QqFYLvQ5B/nXqh/wevAQNrYKt8why6aZfMlSo+kUipXGu4HjnPGeZ//IbdQmqgkqX33WKeiW/FQSSKzhS0gzlm9TFApFlrCFZNyI8Gz5T3i27AARLapa9Vlm2bTkAbASkOHgrdHYKH3hvsnldY41Vw0llrwWfYOgt4iitEtIoVDkj8vOXr5S/S12hrdy++gt6Oj5NmlFsGxEXjh9ONffnYqdlwJxjZeQPaEd/Kz3dV7q+Ql/UvoFBIIdrm3scu2crCOlRAvbvMRBXtNeQ9M07lt3Hw5dDdhQKPKCgJge51DgKGNGiF3j26iLVyuxv04KRuRrElVsiKyZXO50deO23bhsJ/HKMjRfAM1bgttycd/wnTTF6ufcl0M6uNV1E/9H42cICP/kyLKpCCH4Rf+HuF++j3+s+hdiegxDK5jLoVCsXgSc9bbS6m6nNlHFg0P3ELB8+bZq2VIwqrZv7EY+YL1/crnd1YnHdjP+ZoRD92nsfes9eAbq8dhumuJzC/wEunBmNPWWT3jZONDEa743KQ9UYKiQTYWiIDA1k0vuLn5Y9iKbIuu5IbwVhzTUiNkFUrCKNiHkFackDedqMZJ6Rj/uiD5Gt6t3WtmaWAMe2z1rfYHgA9b9bI1t4s8if83O2p2z1lMoFPnhsquXbmcfbweP8PDA+6lLVOfbpGVFwYr8BAKBI5m5md2uXn5S8hpxkZgcUr0pso77h+6EKQ8JKSQ/LX6DmBYHUjNT7TSUwCsUhYgUkoge5bmyn9AQr+V9w3eohGcZMq96CiEagK8DVYAEnpRS/p0QohR4ClgDtAGPSimHRcoB/nfAQ0AE+DUp5bvZNFoiOetpxRQWWyIbMKZ0zGyOrKcpVs/T5S/QlZ4U/JznAq117TP2Y2Gh3vwUiuXDmDHOKf0cES3KjvAWNkXXKffNPGTSRDaBz0kp3xVCBIDDQogDwK8BL0kp/0II8Xng88B/Bh4ENqY/+4Avpv/PCIkkosUAyQXPJTpdl7lz5FY8thttSkSNz/ZiYRHTYpM/8ruBE/Q4+wGIalPi6UVa0BUKxfJHwEVPBx2uy4RHI2yNbJzTHavIQOSllN1Ad/p7SAhxGqgDHgb2p6t9DfgpKZF/GPi6TGXEf1MIUSyEqEnvZ04uui9x1HeKuJbgjaJDyPSQCKd04LSd3Dq2B5/tJaxFOO+5CIAtbJ4pP4ApUlMD2sictcyb+5sJJ8IA1AXrqA3U5uZACsUUesZ78Dv9+J1qJOjVmJrFy8Wv83bgKB8eeD/Vycp8m1SQLMgnL4RYA9wIvAVUTRHuHlLuHEg9ADqmbNaZLrumyB/zNxMrndnavm3kZjZG12EJm0OBozR7z9PvXPp5JI/1HKN7PHUKtzXcpkRekVWSVpLR+CjPnXsOS165D0LxEC7DhVN3EnQFuW/dffidfnRNxY4DIGDcCPNM+YvUx2u5fXQvfss77a1/tZOxyAsh/MB3gd+RUo5NjT2XUkohxILGIgshngCemK/eK8Vv8NPiN1LHgbz50D+242OTw62VD1CRTc4Pnudk30mO9x6fd0j/qb5TvH/D+7m1/tYlsm55MGaM06yf47T3PO8bvp3d4e35NqlgyEjkhRAOUgL/TSnl99LFvRNuGCFEDTCRQ6ALaJiyeX26bBpSyieBJ9P7n/svW5DXTBaheIiknSToCqrBUoqsIqXkzMAZnjn7DFEzswR8EslLF15CILil7hY0oVqsk4jU9TlY/CanfOd4YGg/ZWZJvq3KO/P+haSjZb4MnJZS/vWUVc8Aj6e/Pw48PaX8kyLFrcDofP74Qubc4DmOdB+Z9McrFNni/NB5vnf6exkL/ARJO8nzLc8Tis86sfOqJ6kl6Xb28v3yH/N24CgmqztnfSZN0/cCvwqcEEIcTZf9PvAXwLeFEJ8C2oFH0+t+RCp8soVUCOWvZ9Pgpeam2pvybYJiBZKwEhztOUrSTubblJWJgBHHGK8Wvckp31luDu1iR3jzqnS1ZhJd8zPm9oS/b5b6EvjsddqlUKxousa6aO5vzrcZKx8Bg45hDpS8yogxyrbwJkrN4lUl9sqhp1AsQySSF1tfzLcZywZb2LwVPMIPyp+nx9lPQiTybdKSoUReoVhipJRZmRgjlFA++YUy7Bjlm1Xf44XSg7S4L66KCUqUyCsUS4xE8kLLC/k2Y1Vz1tvKj8pe5pT3HGP6yn5YqphAhSIPRJKRfJuw6kloSZ4ve4WqRDkbI+vYF7pxRfrqlcgrFMuUCm9Fvk1YEfQ6B+hzDBLWI2wPb1oW6REGrSH6rYGM6ip3jUKxxAgEN9ZcX6tRExp3Nt2ZRatWN1JIjgRO8p2K5zhQ8iphrbDftH4ee5tXo69nVFeJvEKxxAghuKvpLm6oumFR27t0Fw9vfpiAK5BlyxQxPc4xXzPfL3+e0RXiq1cir1DkAV3T2VOzB7exsBS5mtB4YMMD7KrepVIa5AoBPa4+flj2Im8G3sXEXNZROOqvRKHIE41FjTy24zGCrmBG9dcWr+XhzQ+zq3pXji1TAPS4+nm96BD/WPsvnPaeZ0wfz7dJi0J1vCoUeUIIQVNxE49uf5Tm/mbeufwOlm1NSzUM4HP4eHDjg2wo3bDglr/i+pBCEtPj/KjsZepjNXxo8H7ctmtZpTJWIq9Q5Jn6YD21gVr2r9nPid4T/PDcD2fUkVJyfvD8tDKf08e6knVLZeaqp9PVzZM13+TG8e1sjK6jNlE1/0YFgBJ5hSLLmLZJ0sos8VhvuJefd/x8cnk0PjqjTjgZ5runvzuj3GN4aCxqvOb+39v43usKtXQbbqbOHbGqEWAKk0PBY5zxtvDQ0PuoTlTgKPAJxZXIKxQZIKWkub+ZhDV/zpO2kTZO9p3MbL9IbGkvyqaoGeXs4Nlr1jk/dH7RoZoO3cF96+6bt4PX6/CyuXzzoo6xXAkZYf694lka4rXsHN/C5uj6gh1IpUResWJIWkniVnzR2//s0s/oD/fPub59tB3TXl65yRf7AAGwTGtW19HVuHQX9cH6eeutLVnL7urd16yjCx2Pw5OpiXnFFjbt7k46Xd2Mj0TYHtmE23YVnNgrkVcsK2xpc7TnKJY9cz7giyMXOTNw5rr2rVg4cStO63DrvPVah1t5+eLL16xT6illX92+OddX+CpYU7xmoSbmFEtY/LT4DQ4FUxOK1xSYr16JvCJvhBPhGS3jMwNnONV/as5tpJR0jnUu67jl1cx8D9KByADPnX9uzvV+p59ST+m0snvX3UuRq2hyWdd0/E7/9Rm6UASE9QjPlB2gPl7NHaO3FsyE4krkFTmndaiVvnDfjPLD3YcZjAxOK1PirbgW44lxxhPT49W/cuQr01wkxe5i9tbtnVwu8ZSwpXzLktgXMsY5rbdwxtvK3tBuPJabgOVnc3T9khx/NpTIKxbFxATnV/Ni64sz5h4djg2rrIuKnDK1cTAcG+aF1iupnN2GmzJP2ZzbbirbxM6qnfMew+fw4TJc8xuTnlD8reARAJy2g0PJY5OraxKV3BS6cjyndOK1c9cPoURecU1safNW51sz3CpHeo4wFB3Kk1UKRebEzBhdoa4513eFunil7ZV597O9YjvV/upF2bChdAM1gRoAepx9HPFfib6qTlSyIbpm1u18tped4et7C1kWIi+lZDg2PGtn2wSdY5282fnmktizvXI7W8u3Ti77nf5lExEwQSQZIZwITysbjg3z0oWXppVJJP3hfuVGUax6TvWfumZ/0bV4u+ttvA7vnOvvbLqTSt/MFMeG1HlBKwZASMFtkZsZKh6lY6wz42OL1Lzb+UUIIR/b8RibyjYBqSfr1aP7Dl0+VLCv/BtLN1IbqJ1RvrNq5zVfE7NNy1ALnRn++J1jnRlFRCgUisJhjdHER/0f5u9H/5GYjAMcllLefK1tCkbkA84ATt0JpF6vwsnwPFsVPlPPaSkIJ8PEzNiSHU+hUOSdeUW+YNw1K3FS4pV4TgqFYnmR/yBOhUKhUOSMeUVeCNEghHhFCNEshDglhPjtdPmfCCG6hBBH05+HpmzzBSFEixDirBDi/bk8AYVCoVDMTSbuGhP4nJTyXSFEADgshDiQXvc3Usq/nFpZCLEN+BiwHagFfiKE2CSlnDs0RqFQKBQ5Yd6WvJSyW0r5bvp7CDgN1F1jk4eBb0kp41LKi0ALsPca9RUKhUKRIxbkkxdCrAFuBN5KF/2mEOK4EOIrQoiSdFkd0DFls05meSgIIZ4QQrwjhHhn4WYrFAqFIhMyFnkhhB/4LvA7Usox4IvAemA30A381UIOLKV8Ukp583zhPwqFQqFYPBmJvBDCQUrgvyml/B6AlLJXSmlJKW3gS1xxyXQBDVM2r0+XKRQKhWKJySS6RgBfBk5LKf96SnnNlGofASaSMTwDfEwI4RJCrAU2Am9nz2SFQqFQZEom0TXvBX4VOCGEOJou+33gMSHEbkACbcBnAKSUp4QQ3waaSUXmfDaDyJoBIJz+fzVTjroG6hqkUNdBXQOY/xo0zbeDgkhrACCEeGe1++fVNVDXYAJ1HdQ1gOxcAzXiVaFQKFYwSuQVCoViBVNIIv9kvg0oANQ1UNdgAnUd1DWALFyDgvHJKxQKhSL7FFJLXqFQKBRZJu8iL4R4IJ2tskUI8fl825NL0ukf+oQQJ6eUlQohDgghzqf/L0mXCyHE/5O+LseFEHvyZ3n2uEZW01VzHYQQbiHE20KIY+lr8F/T5WuFEG+lz/UpIYQzXe5KL7ek16/J6wlkESGELoQ4IoR4Nr28Gq9BmxDiRDqb7zvpsqzdD3kVeSGEDvwD8CCwjVTs/bZ82pRj/hl44KqyzwMvSSk3Ai+llyF1TTamP0+QSiOxEpjIaroNuBX4bPo3X03XIQ7cI6XcRSotyANCiFuB/0kqs+sGYBj4VLr+p4DhdPnfpOutFH6bVNLDCVbjNQC4W0q5e0q4ZPbuByll3j7Ae4AXpix/AfhCPm1agnNeA5ycsnwWqEl/rwHOpr//f8Bjs9VbSR/gaeC+1XodAC/wLrCP1KAXI10+eW8ALwDvSX830vVEvm3PwrnXpwXsHuBZQKy2a5A+nzag/KqyrN0P+XbXZJSxcoVTJaXsTn/vAarS31f8tbkqq+mqug5pN8VRoA84ALQCI1JKM11l6nlOXoP0+lFg6WaIzx1/C/wnwE4vl7H6rgGksga8KIQ4LIR4Il2WtfuhYOZ4VYCUUgohVkW409VZTVMpklKshusgU6k+dgshioHvA1vya9HSIoT4INAnpTwshNifZ3Pyze1Syi4hRCVwQAhxZurK670f8t2SVxkroXci2Vv6/750+Yq9NrNlNWUVXgcAKeUI8Aop10SxEGKi4TX1PCevQXp9ETC4tJZmnfcCHxJCtAHfIuWy+TtW1zUAQErZlf6/j9QDfy9ZvB/yLfKHgI3pHnUnqWkDn8mzTUvNM8Dj6e+Pk/JRT5R/Mt2bfiswOuX1bdkixOxZTVlF10EIUZFuwSOE8JDqkzhNSux/KV3t6mswcW1+CXhZph2yyxUp5ReklPVSyjWk7vuXpZSfYBVdAwAhhE+kplVFCOED7ieV0Td790MBdDo8BJwj5ZP8g3zbk+Nz/TdSE6wkSfnSPkXKr/gScB74CVCaritIRR61AieAm/Ntf5auwe2kfJDHgaPpz0Or6ToANwBH0tfgJPBH6fJ1pNJytwD/DrjS5e70ckt6/bp8n0OWr8d+4NnVeA3S53ss/Tk1oYHZvB/UiFeFQqFYweTbXaNQKBSKHKJEXqFQKFYwSuQVCoViBaNEXqFQKFYwSuQVCoViBaNEXqFQKFYwSuQVCoViBaNEXqFQKFYw/z+0TS8j04DHuwAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "#PREDIKSI\n",
    "model = resunet_seg() #resunet_seg dlv3_seg\n",
    "model.load_state_dict(torch.load(save_dir+'best_model.pth'))\n",
    "model.to(device, dtype=torch.float)\n",
    "input_img = np.expand_dims(resize_matrix(image, resize=4).transpose(2,0,1), axis=0) #buat jadi NxCxHxW\n",
    "input_img = torch.from_numpy(input_img).to(device, dtype=torch.float)\n",
    "print(\"input image shape\", input_img.size())\n",
    "pred_seg = model(input_img)\n",
    "pred_seg = np.argmax(pred_seg.cpu().detach().numpy()[0], axis=0) #ambil batch 0 dan lakukan argmax\n",
    "print(\"output prediction shape\", pred_seg.shape)\n",
    "\n",
    "\n",
    "#visualisasi prediksi segmentasi\n",
    "pseg_color = vis_seg(pred_seg, SEG_MAP)\n",
    "plt.imshow(pseg_color.astype(np.uint8))\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "E. TEST KE SELURUH SAMPLE\n",
    "----------------"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Jumlah sample test: 500\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "570accf5c1fa4a4db49708837b48f070",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "0it [00:00, ?it/s]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "OrderedDict([('tm', 0.1751321575641632), ('iou', 0.8248678424954414)])\n",
      "41.188594579696655\n"
     ]
    }
   ],
   "source": [
    "#FUNGSI IOU\n",
    "def IOU(Yp, Yt):\n",
    "    #.view(-1) artinya matrix tensornya di flatten kan dulu\n",
    "    output = Yp.view(-1) > 0.5 #maksudnya yang lebih dari 0.5 adalah true\n",
    "    target = Yt.view(-1) > 0.5 #dan yang kurang dari 0.5 adalah false\n",
    "    intersection = (output & target).sum() #irisan\n",
    "    union = (output | target).sum() #union\n",
    "    #rumus IoU\n",
    "    iou = intersection / union\n",
    "    return iou\n",
    "\n",
    "#FUNGSI TEST\n",
    "def test(data_loader, model, device):\n",
    "    #buat variabel untuk menyimpan kalkulasi loss\n",
    "    metric_tm = AverageMeter()\n",
    "    metric_seg = AverageMeter()\n",
    "    \n",
    "    #buat dictionary log untuk menyimpan training log di CSV\n",
    "    logx = OrderedDict([\n",
    "        ('batch', []),\n",
    "        ('test_tm', []),\n",
    "        ('test_seg', []),\n",
    "        ('elapsed_time', [])])\n",
    "\n",
    "    #masuk ke mode eval torch\n",
    "    model.eval()\n",
    "    \n",
    "    with torch.no_grad():\n",
    "        #visualisasi progress training dengan tqdm\n",
    "        prog_bar = tqdm()\n",
    "        prog_bar.reset(total=len(data_loader))\n",
    "\n",
    "        #training....\n",
    "        total_batch = len(data_loader)\n",
    "        batch_ke = 0\n",
    "        for data in data_loader:\n",
    "            seg_gt = data['seg'].to(device, dtype=torch.float)\n",
    "            image = data['rgb'].to(device, dtype=torch.float)\n",
    "\n",
    "            #forward pass\n",
    "            start_time = time.time() #waktu mulai\n",
    "            pred_seg = model(image)\n",
    "            elapsed_time = time.time() - start_time #hitung elapsedtime\n",
    "\n",
    "            #kalkulasi metric\n",
    "            iou = IOU(pred_seg, seg_gt)\n",
    "            tm = (1-iou)\n",
    "            \n",
    "            #hitung rata-rata (avg) loss, dan metric untuk batch-batch yang telah diproses\n",
    "            metric_tm.update(tm.item())\n",
    "            metric_seg.update(iou.item())\n",
    "\n",
    "            #update visualisasi progress bar\n",
    "            postfix = OrderedDict([('tm', metric_tm.avg),\n",
    "                                  ('iou', metric_seg.avg)])\n",
    "            \n",
    "            #simpan ke log\n",
    "            logx['batch'].append(batch_ke)\n",
    "            logx['test_tm'].append(tm.item())\n",
    "            logx['test_seg'].append(iou.item())\n",
    "            logx['elapsed_time'].append(elapsed_time)\n",
    "            pd.DataFrame(logx).to_csv(save_dir+'test_log.csv', index=False)\n",
    "            \n",
    "            batch_ke += 1  \n",
    "            prog_bar.set_postfix(postfix)\n",
    "            prog_bar.update()\n",
    "        prog_bar.refresh()\n",
    "        \n",
    "        #ketika semua sudah selesai, hitung rata2 performa pada log\n",
    "        logx['batch'].append(\"avg\")\n",
    "        logx['test_tm'].append(np.mean(logx['test_tm']))\n",
    "        logx['test_seg'].append(np.mean(logx['test_seg']))\n",
    "        logx['elapsed_time'].append(np.mean(logx['elapsed_time']))\n",
    "        pd.DataFrame(logx).to_csv(save_dir+'test_log.csv', index=False)\n",
    "\n",
    "    return postfix\n",
    "\n",
    "\n",
    "#BUAT DATA BATCH: NxCxHxW\n",
    "rootdir = \"/home/aisl/BISA_AI/ai-hw-acc/dataset/cityscapes/\"\n",
    "test_set = myData(root_dir=rootdir, part='val')\n",
    "print(\"Jumlah sample test: \"+str(len(test_set)))\n",
    "dataloader_test = DataLoader(test_set, batch_size=1, shuffle=False, num_workers=4)\n",
    "\n",
    "start_time = time.time() #waktu mulai\n",
    "test_log = test(dataloader_test, model, device)\n",
    "elapsed_time = time.time() - start_time #hitung elapsedtime\n",
    "print(test_log)\n",
    "print(elapsed_time)\n",
    "torch.cuda.empty_cache()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "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": "3.8.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}