yolov3-rsna-starting-notebook.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "cc075cd8b1f3912c3e65651871a4ce263f772750"
   },
   "source": [
    "## Contents\n",
    "### YOLO v3 for image detection\n",
    "0. <a href='#0.-Introduction'>Introduction</a>\n",
    "1. <a href='#1.-Clone-and-Build-YOLOv3'>Clone and Build YOLOv3</a>\n",
    "2. <a href='#2.-Data-Migration-for-YOLOv3'>Data Migration</a>\n",
    "3. <a href='#3.-Prepare-Configuration-Files-for-Using-YOLOv3'>Prepare Configuration Files for training</a>\n",
    "4. <a href='#4.-Training-YOLOv3'>Training model</a>\n",
    "5. <a href='#5.-How-to-use-trainined-YOLOv3-for-test-images-(command-line)'>How to use trained model for test images (command line)</a>\n",
    "6. <a href='#6.-Generate-Submission-Files-with-YOLOv3-Python-Wrapper'>Generate Submission Files Using YOLOv3 Python Wrapper</a>\n",
    "7. <a href='#7.-Future-works-&-Etc'>Future works & Etc</a>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "c828e6dd0b8dc2e80806ef2d3f4137ae4a3ea670"
   },
   "source": [
    "## 0. Introduction\n",
    "* I'll introduce super easy and quick way to train [YOLOv3](https://pjreddie.com/darknet/yolo/) on RSNA and to generate submission file (to be honest, not super easy ...!).\n",
    "\n",
    "\n",
    "* The purpose of this competition is 'object detection'. Generally, object detection algorithms with deep learning take a long time to train model and require a lot of gpu resources. Most individual participants use one or two gpu (... or zero). Therefore, there is a need for **algorithms that works quickly with less gpu resources.**\n",
    "\n",
    "\n",
    "* I tried to use Mask R-CNN, UNet, Fast R-CNN and FCN algorithms, But eventually switched to YOLOv3.\n",
    "\n",
    "\n",
    "* In comparison to YOLOv3, Other algorithms(Mask R-CNN, UNet, FCN, ...) which contain instance/sementic segmentation tasks are very slow, and require more gpu resources, redundant parameter tunning and post-processes. Therefore, if you try to use these algorithms, you may experience difficulties in terms of training time and gpu resources. (Please see [YOLOv3 paper](https://pjreddie.com/media/files/papers/YOLOv3.pdf) for details)\n",
    "\n",
    "\n",
    "* In addition, **YOLOv3 was able to obtain high score (LB: 0.141) without additional processes(data augmentation, parameter tunning, etc...)** compared to other algorithms. So I think YOLOv3 has sufficient potential for this competition.\n",
    "\n",
    "\n",
    "* In this notebook, I'll introduce how to simply apply YOLOv3 on RSNA data. I hope this notebook would be helpful for everyone."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "_uuid": "ce28f29df7b9f64725998f8133be50068c53cb69"
   },
   "outputs": [],
   "source": [
    "import math\n",
    "import os\n",
    "import shutil\n",
    "import sys\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "import glob\n",
    "import pydicom\n",
    "import cv2\n",
    "from sklearn.model_selection import train_test_split\n",
    "from tqdm import tqdm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "# !pip install pydicom\n",
    "# !pip install tqdm"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "_uuid": "c4b33affdc049734daa58852c4ed06e37c344063"
   },
   "outputs": [],
   "source": [
    "random_stat = 123\n",
    "np.random.seed(random_stat)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "97d22b2e1c828e655e7df04474169cd0d7a5a031"
   },
   "source": [
    "## 1. Clone and Build YOLOv3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "# !git clone https://github.com/pjreddie/darknet.git\n",
    "# I built it manually from the terminal"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "_uuid": "558964006de8b4e454d76c91a5d0a605e397a93d"
   },
   "outputs": [],
   "source": [
    "#!git clone https://github.com/pjreddie/darknet.git\n",
    "\n",
    "# Build gpu version darknet\n",
    "#!cd darknet && sed '1 s/^.*$/GPU=1/; 2 s/^.*$/CUDNN=1/' -i Makefile\n",
    "\n",
    "# -j <The # of cpu cores to use>. Chang 999 to fit your environment. Actually i used '-j 50'.\n",
    "#!cd darknet && make -j 50 -s\n",
    "#!cp darknet/darknet darknet_gpu"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "9f77f087a15c73fae7a710fd560244b8f7884197"
   },
   "source": [
    "## 2. Data Migration for YOLOv3\n",
    "It might take a while."
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "5702330ac23995099c3f184ef0424731d91a303c"
   },
   "source": [
    "### 2.0. Make subdirectories"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "_uuid": "55f5d67ac085afcc445db23227ba60748eb269fa"
   },
   "outputs": [],
   "source": [
    "# DATA_DIR = \"../input\"\n",
    "DATA_DIR = \"./downloaded_data\"\n",
    "train_dcm_dir = os.path.join(DATA_DIR, \"stage_2_train_images\")\n",
    "test_dcm_dir = os.path.join(DATA_DIR, \"stage_2_test_images\")\n",
    "\n",
    "img_dir = os.path.join(os.getcwd(), \"images\")  # .jpg\n",
    "label_dir = os.path.join(os.getcwd(), \"labels\")  # .txt\n",
    "metadata_dir = os.path.join(os.getcwd(), \"metadata\") # .txt\n",
    "\n",
    "# YOLOv3 config file directory\n",
    "cfg_dir = os.path.join(os.getcwd(), \"cfg\")\n",
    "# YOLOv3 training checkpoints will be saved here\n",
    "backup_dir = os.path.join(os.getcwd(), \"backup\")\n",
    "\n",
    "for directory in [img_dir, label_dir, metadata_dir, cfg_dir, backup_dir]:\n",
    "    if os.path.isdir(directory):\n",
    "        continue\n",
    "    os.mkdir(directory)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "_uuid": "f0b2173fa0ff6a1bfdf304310466d01052920fe6"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "total 166M\r\n",
      "304K -rwxrwxrwx  1 qinghe qinghe 301K janv. 22 12:01 yolov3-rsna-starting-notebook.ipynb\r\n",
      " 40K -rw-r--r--  1 qinghe qinghe  38K janv. 21 19:08 loss.png\r\n",
      "8,0K -rw-r--r--  1 qinghe qinghe 6,4K janv. 21 18:41 submission.csv\r\n",
      "4,0K drwxr-xr-x  2 qinghe qinghe 4,0K janv. 21 17:23 __pycache__\r\n",
      "8,0K -rw-r--r--  1 qinghe qinghe 5,1K janv. 21 17:23 darknet.py\r\n",
      "4,0K drwxrwxrwx  2 qinghe qinghe 4,0K janv. 21 16:56 cfg\r\n",
      "4,0K drwxrwxrwx  2 qinghe qinghe 4,0K janv. 21 16:54 backup\r\n",
      "224K -rwxrwxrwx  1 qinghe qinghe 222K janv. 21 16:54 test.jpg\r\n",
      "8,9M -rwxrwxrwx  1 qinghe qinghe 8,9M janv. 21 16:54 train_log.txt\r\n",
      "4,0K drwxr-xr-x 13 qinghe qinghe 4,0K janv. 21 11:47 darknet\r\n",
      "4,0K drwxrwxrwx  2 qinghe qinghe 4,0K janv. 21 11:18 metadata\r\n",
      "424K drwxrwxrwx  2 qinghe qinghe 424K janv. 21 11:18 labels\r\n",
      "616K drwxrwxrwx  2 qinghe qinghe 616K janv. 21 11:18 images\r\n",
      "4,0K drwxrwxrwx  4 qinghe qinghe 4,0K janv. 21 11:17 downloaded_data\r\n",
      "   0 -rwxrwxrwx  1 qinghe qinghe    0 janv. 21 10:28 0\r\n",
      "   0 -rwxrwxrwx  1 qinghe qinghe    0 janv. 21 10:28 darknet53.conv.74\r\n",
      "   0 -rwxrwxrwx  1 qinghe qinghe    0 janv. 21 10:28 detector\r\n",
      "   0 -rwxrwxrwx  1 qinghe qinghe    0 janv. 21 10:28 train\r\n",
      "272K -rwxrwxrwx  1 qinghe qinghe 271K janv. 17 15:44 pneumonia-V0.ipynb\r\n",
      "155M -rw-r--r--  1 qinghe qinghe 155M mars  25  2018 darknet53.conv.74.1\r\n"
     ]
    }
   ],
   "source": [
    "!ls -shtl"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "ba6a35b97d0a42060e6fdd68a6da23454aa40568"
   },
   "source": [
    "### 2.1. Load stage_2_train_labels.csv"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'./downloaded_data/stage_2_train_labels.csv'"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "os.path.join(DATA_DIR, \"stage_2_train_labels.csv\")\n",
    "# changes stage 1 to stage 2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "_uuid": "e242b29961c22433703e442a84572783b9459de9"
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>patientId</th>\n",
       "      <th>x</th>\n",
       "      <th>y</th>\n",
       "      <th>width</th>\n",
       "      <th>height</th>\n",
       "      <th>Target</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0004cfab-14fd-4e49-80ba-63a80b6bddd6</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>00313ee0-9eaa-42f4-b0ab-c148ed3241cd</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>00322d4d-1c29-4943-afc9-b6754be640eb</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>003d8fa0-6bf1-40ed-b54c-ac657f8495c5</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>NaN</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>00436515-870c-4b36-a041-de91049b9ab4</td>\n",
       "      <td>264.0</td>\n",
       "      <td>152.0</td>\n",
       "      <td>213.0</td>\n",
       "      <td>379.0</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                              patientId      x      y  width  height  Target\n",
       "0  0004cfab-14fd-4e49-80ba-63a80b6bddd6    NaN    NaN    NaN     NaN       0\n",
       "1  00313ee0-9eaa-42f4-b0ab-c148ed3241cd    NaN    NaN    NaN     NaN       0\n",
       "2  00322d4d-1c29-4943-afc9-b6754be640eb    NaN    NaN    NaN     NaN       0\n",
       "3  003d8fa0-6bf1-40ed-b54c-ac657f8495c5    NaN    NaN    NaN     NaN       0\n",
       "4  00436515-870c-4b36-a041-de91049b9ab4  264.0  152.0  213.0   379.0       1"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "annots = pd.read_csv(os.path.join(DATA_DIR, \"stage_2_train_labels.csv\"))\n",
    "annots.head()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "9b26d929bb119deca558ac983a5ce4e95fe9b7cc"
   },
   "source": [
    "### 2.2. Generate images and labels for training YOLOv3\n",
    "* YOLOv3 needs .txt file for each image, which contains ground truth object in the image that looks like:\n",
    "```\n",
    "<object-class_1> <x_1> <y_1> <width_1> <height_1>\n",
    "<object-class_2> <x_2> <y_2> <width_2> <height_2>\n",
    "```\n",
    "* <object-class\\>: Since RSNA task is binary classification basically, <object-class\\> is 0.\n",
    "* <x\\>, <y\\>: Those are float values of bbox center coordinate, divided by image width and height respectively.\n",
    "* <w\\>, <h\\>: Those are width and height of bbox, divided by image width and height respectively.\n",
    "\n",
    "* So it is different from the format of label data provided by kaggle. We should change it."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "_uuid": "5204d3b142e17927c06e9e992dd4e4f4a3d0c8ab"
   },
   "outputs": [],
   "source": [
    "def save_img_from_dcm(dcm_dir, img_dir, patient_id):\n",
    "    img_fp = os.path.join(img_dir, \"{}.jpg\".format(patient_id))\n",
    "    if os.path.exists(img_fp):\n",
    "        return\n",
    "    dcm_fp = os.path.join(dcm_dir, \"{}.dcm\".format(patient_id))\n",
    "    img_1ch = pydicom.read_file(dcm_fp).pixel_array\n",
    "    img_3ch = np.stack([img_1ch]*3, -1)\n",
    "\n",
    "    img_fp = os.path.join(img_dir, \"{}.jpg\".format(patient_id))\n",
    "    cv2.imwrite(img_fp, img_3ch)\n",
    "    \n",
    "def save_label_from_dcm(label_dir, patient_id, row=None):\n",
    "    # rsna defualt image size\n",
    "    img_size = 1024\n",
    "    label_fp = os.path.join(label_dir, \"{}.txt\".format(patient_id))\n",
    "    \n",
    "    f = open(label_fp, \"a\")\n",
    "    if row is None:\n",
    "        f.close()\n",
    "        return\n",
    "\n",
    "    top_left_x = row[1]\n",
    "    top_left_y = row[2]\n",
    "    w = row[3]\n",
    "    h = row[4]\n",
    "    \n",
    "    # 'r' means relative. 'c' means center.\n",
    "    rx = top_left_x/img_size\n",
    "    ry = top_left_y/img_size\n",
    "    rw = w/img_size\n",
    "    rh = h/img_size\n",
    "    rcx = rx+rw/2\n",
    "    rcy = ry+rh/2\n",
    "    \n",
    "    line = \"{} {} {} {} {}\\n\".format(0, rcx, rcy, rw, rh)\n",
    "    \n",
    "    f.write(line)\n",
    "    f.close()\n",
    "        \n",
    "def save_yolov3_data_from_rsna(dcm_dir, img_dir, label_dir, annots):\n",
    "    for row in tqdm(annots.values):\n",
    "        patient_id = row[0]\n",
    "\n",
    "        img_fp = os.path.join(img_dir, \"{}.jpg\".format(patient_id))\n",
    "        if os.path.exists(img_fp):\n",
    "            save_label_from_dcm(label_dir, patient_id, row)\n",
    "            continue\n",
    "\n",
    "        target = row[5]\n",
    "        # Since kaggle kernel have samll volume (5GB ?), I didn't contain files with no bbox here.\n",
    "        if target == 0:\n",
    "            continue\n",
    "        save_label_from_dcm(label_dir, patient_id, row)\n",
    "        save_img_from_dcm(dcm_dir, img_dir, patient_id)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "_uuid": "1e907f30b11bf6956884061138d615bee60c2a15"
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "100%|██████████| 30227/30227 [00:01<00:00, 17646.79it/s]\n"
     ]
    }
   ],
   "source": [
    "save_yolov3_data_from_rsna(train_dcm_dir, img_dir, label_dir, annots)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "_uuid": "19d2c4ab8448c0572ac57b49eff52c2230c9f079"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "1,5G\timages\r\n",
      "24M\tlabels\r\n"
     ]
    }
   ],
   "source": [
    "!du -sh images labels"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "17868a4d327c2e22adae76cc7ce163c5546e1743"
   },
   "source": [
    "### 2.3. Plot a sample train image and label"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'c040e4f2-467a-4078-beef-0ff693cfa952'"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_dcm_fps[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'00436515-870c-4b36-a041-de91049b9ab4'"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "annots[annots.Target == 1].patientId.values[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "_uuid": "8ebf845a4f3b51555991292d10ddc272b226bf7c"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0 0.36181640625 0.33349609375 0.2080078125 0.3701171875\n",
      "\n",
      "0 0.36181640625 0.33349609375 0.2080078125 0.3701171875\n",
      "\n",
      "0 0.36181640625 0.33349609375 0.2080078125 0.3701171875\n",
      "\n",
      "0 0.673828125 0.36962890625 0.25 0.4423828125\n",
      "\n",
      "0 0.36181640625 0.33349609375 0.2080078125 0.3701171875\n",
      "\n",
      "0 0.673828125 0.36962890625 0.25 0.4423828125\n",
      "\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": [
    "ex_patient_id = annots[annots.Target == 1].patientId.values[0]\n",
    "ex_img_path = os.path.join(img_dir, \"{}.jpg\".format(ex_patient_id))\n",
    "ex_label_path = os.path.join(label_dir, \"{}.txt\".format(ex_patient_id))\n",
    "\n",
    "plt.imshow(cv2.imread(ex_img_path))\n",
    "\n",
    "img_size = 1014\n",
    "with open(ex_label_path, \"r\") as f:\n",
    "    for line in f:\n",
    "        print(line)\n",
    "        class_id, rcx, rcy, rw, rh = list(map(float, line.strip().split()))\n",
    "        x = (rcx-rw/2)*img_size\n",
    "        y = (rcy-rh/2)*img_size\n",
    "        w = rw*img_size\n",
    "        h = rh*img_size\n",
    "        plt.plot([x, x, x+w, x+w, x], [y, y+h, y+h, y, y])\n",
    "        plt.annotate(\"pneunomia\", (x-20, y-20), color=\"magenta\")\n",
    "        plt.savefig(\"test1.png\", dpi=300)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "42551c1e11208ee88f9b93bea40c23dcad295b8a"
   },
   "source": [
    "### 2.4. Generate train/val file path list (.txt)\n",
    "* We should give the list of image paths to YOLO. two seperate list textfiles for training images and validation images."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "_uuid": "d62f98b287e73a66c234c4c9f9e41503fc0d4bca"
   },
   "outputs": [],
   "source": [
    "def write_train_list(metadata_dir, img_dir, name, series):\n",
    "    list_fp = os.path.join(metadata_dir, name)\n",
    "    with open(list_fp, \"w\") as f:\n",
    "        for patient_id in series:\n",
    "            line = \"{}\\n\".format(os.path.join(img_dir, \"{}.jpg\".format(patient_id)))\n",
    "            f.write(line)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "_uuid": "4c3982b42711fe4a20600b90ed3d123384da7096"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "The # of train set: 5410, The # of validation set: 602\n"
     ]
    }
   ],
   "source": [
    "# Following lines do not contain data with no bbox\n",
    "patient_id_series = annots[annots.Target == 1].patientId.drop_duplicates()\n",
    "\n",
    "tr_series, val_series = train_test_split(patient_id_series, test_size=0.1, random_state=random_stat)\n",
    "print(\"The # of train set: {}, The # of validation set: {}\".format(tr_series.shape[0], val_series.shape[0]))\n",
    "\n",
    "# train image path list\n",
    "write_train_list(metadata_dir, img_dir, \"tr_list.txt\", tr_series)\n",
    "# validation image path list\n",
    "write_train_list(metadata_dir, img_dir, \"val_list.txt\", val_series)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "482fc4b015279312504c910b6c910a5d58960159"
   },
   "source": [
    "### 2.5. Create test image and labels for YOLOv3"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "_uuid": "7394207f5f23778aee12d5f4dadae632b340ce62"
   },
   "outputs": [],
   "source": [
    "def save_yolov3_test_data(test_dcm_dir, img_dir, metadata_dir, name, series):\n",
    "    list_fp = os.path.join(metadata_dir, name)\n",
    "    with open(list_fp, \"w\") as f:\n",
    "        for patient_id in series:\n",
    "            save_img_from_dcm(test_dcm_dir, img_dir, patient_id)\n",
    "            line = \"{}\\n\".format(os.path.join(img_dir, \"{}.jpg\".format(patient_id)))\n",
    "            f.write(line)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "_uuid": "e99463093b202e2aabfcd3979701b7e47ba91c3e"
   },
   "outputs": [],
   "source": [
    "test_dcm_fps = list(set(glob.glob(os.path.join(test_dcm_dir, '*.dcm'))))\n",
    "test_dcm_fps = pd.Series(test_dcm_fps).apply(lambda dcm_fp: dcm_fp.strip().split(\"/\")[-1].replace(\".dcm\",\"\"))\n",
    "\n",
    "save_yolov3_test_data(test_dcm_dir, img_dir, metadata_dir, \"te_list.txt\", test_dcm_fps)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "42c007da9fd2009310dcf697bca943c4dd7a33f7"
   },
   "source": [
    "### 2.6. Plot a sample test Image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# # ex_patient_id = annots[annots.Target == 1].patientId.values[0]\n",
    "# ex_patient_id = test_dcm_fps[0]\n",
    "# ex_img_path = os.path.join(img_dir, \"{}.jpg\".format(ex_patient_id))\n",
    "# ex_label_path = os.path.join(label_dir, \"{}.txt\".format(ex_patient_id))\n",
    "\n",
    "# plt.imshow(cv2.imread(ex_img_path))\n",
    "\n",
    "# img_size = 1014\n",
    "# with open(ex_label_path, \"r\") as f:\n",
    "#     for line in f:\n",
    "#         print(line)\n",
    "#         class_id, rcx, rcy, rw, rh = list(map(float, line.strip().split()))\n",
    "#         x = (rcx-rw/2)*img_size\n",
    "#         y = (rcy-rh/2)*img_size\n",
    "#         w = rw*img_size\n",
    "#         h = rh*img_size\n",
    "#         plt.plot([x, x, x+w, x+w, x], [y, y+h, y+h, y, y])\n",
    "#         plt.annotate(\"pneunomia\", (x-20, y-20), color=\"magenta\")\n",
    "#         plt.savefig(\"test1.png\", dpi=300)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "_uuid": "f057a0fc9938308f69a5d402cf51ba340af592c0"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7fe3a94c3780>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAQ8AAAD8CAYAAABpXiE9AAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nOy9e4xk6Xne95y6V3VXV1f39HTPzs5wd0lqqctSIGSYEiTAcmQZURJEERAFiqLIDkQIAmQ4hgXQjuA/Q8j6R5RlQTHkMLAsCJCMyJBNRbAQSFEMCg4lkwwsMgJ3h8vh3Ke7q7q6u25dt5M/an5vPeds7613PRyu5gCN7q7LOd/5vu993ud93vf7TpKmqZ4eT4+nx9Pj7R6Fr3cDnh5Pj6fHN+bxFDyeHk+Pp8eFjqfg8fR4ejw9LnQ8BY+nx9Pj6XGh4yl4PD2eHk+PCx1PwePp8fR4elzoeOzgkSTJf5okyZeTJLmRJMnff9zXf3o8PZ4e786RPM46jyRJipJelvT9ku5I+lNJ/22apv/fY2vE0+Pp8fR4V47HzTz+sqQbaZq+mqbpRNJvSvrBx9yGp8fT4+nxLhylx3y9q5Ju2/93JH3UP5AkyU9K+slH/35HkiSPqWlv7UjTVBdp00UYXpIk5/4UCoU4H//7e/yUSqV4z9ueb0v+fjiXJC0Wi9d8x8/3et/Pv79YLOLvNE3P7Q9eT9NUi8VC8/k8vufvcz7Oxf98drFYxDn8+2/3+AtUfX2YpunO2/3S4waP86wuM0Jpmv6qpF+VpEKhkJZKj7uJrz3cGN4IPM4zptcz2PO+lzfQUqmkYrGocrmsWq2mer2uUqmkcrmsYrEoSQES5XJZ1WpVklSpVLS1taVCoaD5fK5isai9vT01m03N53MlSaL5fK5KpaJisajJZKJarabZbKb5fK5yuay1tTWlaarpdKrZbBbGXK1WtVgsdHZ2pmKxqNlspmq1GufkPJJUKpW0WCzi+4vFQpVKRfP5XNPpVIvFQoVCQYvFQkdHRzo7O9NsNtNsNstcdzweRx/O53MVCgWNRiOtra1pPp9rNBppOp1qMpkoTVOdnZ3p7OxMg8Eg8908mPhYet+fBzgOeA5yfJ7zvJljeb33v55Ocjabfe0i33vclnlH0jX7/1lJ9x5zG9724RMlP8ivByavNzHP+9x5rxUKy4iyUCioVqsFYMA6SqWS0jRVuVwOIwUQFouFJpNJfCZNUz18+FDdblfPPvts3AfXKRQKAQbr6+txLmcJtVpNp6enms/nmkwm0dbpdBpMpVwuazweq1QqaTweq1wuazKZBEgVCgXNZrNgSIvFQqenpxqNRvE/LKJYLMZ9co35fK5SqaTJZKJGoxH3P5lMNJ/PJUnFYjEDvICsMykfU1gK/ZqmaXzHx71YLMY1fNzy3329efJW5svrfe/NmOLX63jc4PGnkj6YJMnzku5K+hFJP/qY2/CWDx/oNwOI1/v+m53/jd6rVquqVquqVCpxLTz/bDYL4JjNZmEsGPF0Oo2whUk9nU518+ZNNZtNbW9vq1KpSFJ48GKxqFqtFqzA7/vs7CzYAIaEoblRlctllcvlAB6ADwZQq9WCYXQ6HSVJEmAE66AtsKJKpaLZbKZKpaLpdKpqtap6va7ZbBZso1wuBwClaarxeBzn8DAPQAWM6GtnZJKCFfE+QMb5zwsFASNez/fh6417nqm+2fFmn3lc4PJYwSNN01mSJH9L0u9LKkr639I0/dLjbMPbOd4KYJwHIG8W3vCZN/q/Wq1qY2MjaHij0Yj3ABP0DACFNkPtp9NpGBWevFAoqNfrabFY6MqVK1pbW1O/3482nJ6eam1tTUmSaDqdSlqyi/l8HoAFS+E6kiIkIIwol8vBIGAOaZqq3+9rPB7r+PhYi8VC0+lUlUpFw+Ew0w+wEJgURg97kRSspFqtZtgJ7/Hj2ogDDPfg56effBz5m/vh+/nPSMqwlvMYyevNhXfT4N/t873e8dgFhTRNf0/S7z3u676d460Y/uuFL29GW/ns6/3PBG40GkHJ854zSZJgGfV6PQwNo+J9zofB4Mk578HBQXhUDOn09DTuBWZTLpclSc1mU2dnZ6rVahEacG6uy/+LxUL9fj/CoHK5rDRNdXJyon6/H58DjFyk9XPSPliGJNVqNaVpGn3D/XgYx307++L8DhCAn7QCLFiV6yx+eHjlPw4UHhb6668HIu+EjXy9wpivvxr5hB1vBhxvJH6+G5STyQ9FlhQGQjjAawiH6+vrEaZIK+/H4ZO/VCoFU1ksFup2u6pUKqpWqyoUCvF3kiRB/QEPjJHzVSqVYCgAEyBQLBZVr9fjuqPRKEPt8fh+P4Qs3AfnAhToH0IZ+sf/L5fLGg6HGRDx0M4zNXkGMZvNgqXxmgMfPwjReUDgvjjXec7krQDJW5kn+c++VS3u3TyegkfueCsdngeQtzrQ533OPWClUolQxBV/jJFJTUjAe7CBQqEQ4mGpVAqDQkjFy8I+YA+TyUTD4TADGpVKJYRPPof3d8+9sbGh8Xis8XgcYQzMAdGWkAL2U61WdXJyEmwH1uQpZ9oPmyIjNJ1OVa/XI9yZTqcBaoRp5XI5Qiz6De3DmRGgAJgAWnlQkZTRRyRFf7yeEMt56X/A0tPIfD6vo7yV4zOf+Yy+8zu/M9r89QCQp+Bhx+t1tnuPt+MRzjvHef8Xi8XISGBI7pnwmIQF0op5QIv5e21tLQxxNBqF4WCkeVqOMFgsFnV2dhb3KSm0DSmbjsTTp2mq4+NjlUqlSI3CYNI01XA41Pr6usrlcqSAZ7OZGo2Gzs7OgqnwHgyGe+f63B/9M5lMQrgFSPk830HHAFwAYMJCB1b6xxlbXuz0dC/9z/nz43oeoOSBg3AILclff7PQN01TffGLX9S3fuu3xri81ZD53TyegocdbyVN9naPNwKNQqGgarUaRuEshElYqVTCg3q60o2hUChoPB5rY2Mjai3w9q6X5O8TrSFN02AblUolwpXFYqHBYBBeu1wuazQaabFYqNlsam1tTUdHR5GlKZVK6vV6oTO0Wi31+301Gg0VCoU4F8brdSXcN212wZfQI58JIeskKcI2zill06swM4CVsaAPATgOB00HEgRgvsu58myRceEcjJ/rOZ7RkRT1LwjU+WI3n5+/8zu/ox/90R/NjKdfO58N+o9xPAWPcw7v8LebRnujzzo4eDbAP+uVk/55N2Cn4Ey+YrGo8XgcmRLec8biVNwzE54adeBxRsP5EVIx6u3tbaVpqkqlosPDQ9VqtQCiJEnUaDRULBaDlVD/QR+4Nwew+M01GAPYCcADqHBPo9Eok0r296RVBiYfItHHMAGunw/DMHbaRf/lBVlndcwhF65duHXAcXE6TVcFejAtB5PPfOYzwSY5J0V757WH490Ekqfgcc7xZnn51zveTEjFYDDUPCV2L+ReOF/x6MKdhx4YAEbioQrfgebnS9wJmxaLhVqtlhqNhkajkaRs2pTzdzodTSYTjcdjbW9vq1wu6+TkJDy8pPDQsAu8NeEJBkHYAzC5AHlepsQNxLMdsKbFYhFFbYAQ33Xm5u10IMjPg3w2iTGgvS4Ue5vzbef8ecB0EOfzzjwZS36m06lOTk40mUy0tbWlbrcbAJ4vZHujufxOgeQpeJxzvN1Q5a2o5sViMYDD15v4wQR0UPEfF+LOq4j03xjTaDRStVrNiKwYQLVa1Ww2U71ej4kK8wDoXBSkTYPBQNPpNArAjo+Pw0uTdYEBUdw1Ho8lKXQPWIyHEl6FSntpQ7FYDA3HjdoZ1nA4zGRC6vW6xuNxxuB53+s1nKF4Za2kDFDlw0p+zmOPecPNsxJnfu4M/G/XQBB0XcNaW1vTK6+8kmFKly9fflthyzsJy5+Cxxscb0UgfTO2gVG64OeAkhctnUbzfSaOi2OEBU6rPZXK51nPAusgZYnAWKvVMkVljUYjPJh71LOzsyg9l5Zp2nq9Hp8tFJYl7rCVfr8f9RiFQiFqRKDjtIksEzUb3AdGS3sJd7gPjI/fhFreT5zfq0q9QMyB5rzwzkHKmWE+5KPPHczz8wJtJi+IEm4AlrSRc0wmE52dnZ1bSwK4Ekr923/7bzPCbp4l5Y+8jvJ2j6fg8eh4I43iIp/3EMWzIrzn1BRWcp63cH1EWk04PDoT3jMIGAUT0jMLfB4xUloVTRFSYLhuEGgNLiw6mzo7O4sU6mKxXMx2dnYWC/bIynB/hBOEPkmSBAB4fwBirofQf1634ZkVBxDa6xknDyc8WyMtQfHs7CxTI0JbPd3KOQB8xtNL2H1OONNB8wEMuZ+zs7MMkyoUClpfX9fx8bH6/X4GlOr1ura3tzMsBkBn7BhLZy8cznAverxnwSPfWefFehehbG/2nXxNQV6k9DCDz+dZyXkgwgRx0S0vlnG+fLx9Hj3287FalrUlZFXW1tZiTUupVNLa2lpGWOT6W1tbOjk5iayPJK2vr+vw8DCAysVR1zbc2BBbuV9CFimrFfjEB+ioXYHdeLiH4XuY5Ock7QwweB/yeT8n7wN49L2zBgCe+ydF7U7CmQTnYS7QlrW1tQjH3Pn4gkhJAczMD0Cd7JMvZvR5/JR5nHO823HeW2EaeJR8yOHeic86yEgro8/Ta7yehy8eE/v1MCAqSHk9nzGQlDFoqkEJDVhWPxwOwwharVa0m3bNZrPQNtbX18NjS8uwhQV4k8kkMgZc27WUarUan3GxMV+lSZ9xba/8rFarwXQAEx8bskUAKZ7eQz8vNIN1ATrej3mg8dQxfe4spVKp6Pr161FHA2tcLBYaDocBHEmSRF9w/9PpNADR60a4BxyBC68czBeAZDKZnCu+X/R4z4KH9M5rNN7oHBi5C32u1EtZAPPPes7fxVO+j+dgUrgWwuTOx9W+qhatgslMatMNsVqtRoGSl3q7qAqN9k2FCAE439raWiyEo1+4r8FgEN/NV3oSRjmb4D0AwTMjGGK+rJ224uXztRauHXghmms/AI0vAfBUsK8y5r4ZP2eDXJN+3dnZ0d7eXlyzVqtpfX1do9FI4/FYaZqGbsTY+pju7e1pc3NTX/jCF/Tw4cNYPOhVt7APB8J8ihr2BVMiPHynx3saPN7p8Ubg44aQjx8ZREmvMRynvQ4KeDCPybk+egOTggnj5dR+ft7nPGQ/YAY+6WgvYigU18U2wArhDhBx8bNQWFa3wgpYMg940JbxeJwxQBcu3dtzXQdpF0AXi0WGXbhYyf+1Wi0Ax1PFksKYXJ9wVuJjCRPxUM3Hx8PB+XyuVqul559/XvV6PfqIkvzBYKDhcKh6vR5AnCRJ9BfMqVKpaG1tTTs7O/q2b/s2fehDH1Kj0dDu7q663a6Oj4+VJElseDQcDtXtdnVwcKDBYPCa0I22M2feDQB5z4HHO2UbbxSe0PnnAQbvu9iVZxrnhTDSqkCJgZWySrjH1LxXrVYzmgHe0Zfrc+5SqaThcJgRVDFS1od4jD6ZTKLyNU1TtdttjUaj0CD4Lp4azz8YDIIZUKbebDaDLqOtuOibpmn8RrDFeNwg6Qe8Lv+js7hxe6rXPyutwjX60RmLaxjOJPKZC/rdy8ol6X3ve58uX76cCdUAjLt370Z41263M+clde0MkLT04eGhBoOB9vf3VSwWdfv27cwCwjRNtba2pvX1de3t7Wk4HGo2m+nWrVu6f/9+AHF+DjlIX/T4hgaPvOjj/+c75fXE07fymguaHprkAQBDdZBwduJaiH+H9rrQ6UDC5PbMgU9oKbv+xel2rVbLhDFcP0mSWFrvBVvNZjNSh41GIzyix/7z+TyW2o9Goyg7ByBIvbJXCIBRKBR0fHwcGQ1fiMd5vRTfmZizGA+tPIRgWT7fAxjoJwc+p/ic29PDPh685u3xEvK9vT29+OKLajQaGg6HevjwoQ4PDzPVtBsbG5G6Hg6HwQgrlYrW19clrdYMAcRc7/j4OPo4TZd7opycnASzKRaLASDT6VTtdltXr16VJHU6nUxq3EM5+uSixzc0eJwHEtIb743wRp/Lv5c3fn+d/30S5tNs/r3z2n5eqJPPlOAx3PNx317klAccF9+cyhMOwFYAJCpMeQ8Ww/0joiIMe1YDZuFMAVAgvECwY8JT+5KmqTY3NyMkQvzMLw7EYAEWdifz7AaxP/cEK/PPeFbL75sw0H/TjxSpQf8Bkul0qo9+9KO6dOmSFouFbt++rX6/H33TarWi7Z45ajab6nQ6KhQKqtfrsYUAAimv93o9SVKv1ws2hSNYX1/PVBCfnJzE+Vnn9Oyzz8Y5PBQmbHTAv8jxDQ0e0tsv5KKz86qzf+48odK/6+LneezCvZ2n9Ty7kq9KzAuggIGHKoiivO/XzVdkeoyLKOoG6RvooHMQhzsz8gVanmkpl8vq9XqxBQDXQv9gIhN+YIC0HaNeLBZR1o7B44EJSfIaEOEG5+IHkMivDvZwzVPafl8ODnlh1rMvfLder+u7vuu7tLW1pbt376rf7+v09DSqbkulUlTubm5uSlplwwaDgT7wgQ9EGx88eKCTkxNVq9UArvF4HKC6vb2tu3fvBpskpe796iGfF8y1Wi11u92onaH/vD/+woLHW43Z8qFBXiHnM27gvOfA4OGLgwff99SZf1fS64LReZkFp9oOInzP168wYZh4CIW870bvGySzEta/O5vNInTxvUHoK8TW4XAYXtB1F4ySSQkIjsfjyL54bYKnbL3+xXdFk1Zl7S5uulZBWwqFQuwi70zFWZXXvfjhwMnYcX6vU3nf+96nK1euqFwu68GDBzo7O9N8Ptfe3l5cf21tTdVqVbVaTcPhUMfHxzo7O1Ov19PZ2Zn29/c1HA61tbUV4SOgh3jd6/U0Ho+1t7ener2uvb29WHw4HA41n89Vr9cz/ebhKBkvD009vU0fXvR4rE+Me7tH4XUevXCRNr9ZeJMPT5xleCo2DxB8Nq+J5EXV/DmkFZvw63m5NK+z5oT/8VAAD6k+FqURWjBZ/HpOvZNkueqVbECaplGe7vSe+wRM8oVH7D7mm/NQ1+FVp55dIA7PFzNh+IBOsViMkncPlfC6MAR/lIOX1nPfeGfAj/uhTwhPnJWmaRrXlJaA9sILL2hra0tJkujg4CC2MWDFcLvdjpL94XCowWAQ4UxeQxuNRmo2m5GCbTabOjg4CMcAwHt9SKPRULvd1sbGhra3t9Vut1UqlXR6eqpPfvKT+sf/+B+rWq3q4cOH+vjHP64XX3wx7o8+p3/4fzwefy5N07/0tgxK7wHm8VaONwMbFxslZQCASSdlAcYr+jg8jsyDiMfleaDx6zL5nSI7kOUBz9vuQIRX4T20AAyV1C8eFmND5GOdiwuP+W36isXloxrSdLk0v1qtqtvtRp/X6/UwUE+xDgaDTIjEb4zbwwgXJhFv82laAE1aLZbz1KyDgQMjrwEaXh/h2RCM98qVKyFu9vt9ra2tZcav1WpFSDEajXT37l3t7++rVqtlmCh9yLUZl+FwmBlDNnbmfjz0XCyWz7qZTCbqdDqaz+f6p//0n+pXfuVX9Hf/7t/VP/tn/0y///u//5plD/RRvm8ucjzx4OGe5q18luO873h44izAQwIAIp9iZXL5Z51FEA44xc1/n/PnaboDEHSf89BO+gEmAaUej8eRVnUvyufx7giexN14bddAOA8FR274GJjrOV5AxU+j0YjMACEQ2QW2K/Tdzv27VLdKK13DF+N5WwuFgvr9fqb/CfFc8HU2wXnoY99g2mtPmCMupF69elWXL19WtVqNDJKfg76ZTqfq9/t69dVXMyEnzM7XGwGWzBXai0Dd7/djLrVarRBJO52OXnnllQC1RqOhP//zP9f3fM/36OHDh0qSRK+88op+8Rd/Ub/yK78S89QzdbCji+od0jcAeEhvvnL1jb7nqMuEcI8rZQXSvOaBh8jrGlIWPPgfo3YP7e9xDn7zwwTOawCcJ6+dOMOZz+eRBnRNwb+Hp4ZJuECZ12gIYwg9EAARMclqEIp4GIGBYPSz2SxKxwuF5UIvFsIBhqSUicNd/HSRmAPAoiSecIf2wLK8BJ92c58YTR5gHCzJely9elWtVkulUikTOpFqZgf7fr+v/f392KKAEIxx89oZti8krMO4a7Va3MvW1laAzGw2040bN+KRFfRdpVLRV77yldiD5Y/+6I/0J3/yJ3r++eeD7UnZvVHyDumixzcEeJx3vBlwOFV1L+5swI3ZwxA6NF8aTYf75857zUMJN0quxeHfh457qhbD9c+yWpLX8MK0z1kJ5+NeODBGJqan/Dgvk59MA31Xr9djERxGSPUqbME1Gda/wD44L/ciKWpRXNiVFBWxhFK+gxc6x8bGhra2tjQcDqN6FQDN97XrGaSDC4VVQdqj+D+0mlKppPX1dbVarWAcABB9C4gMh0M9ePAgI1BKivFDDyIr4qEsYIGoS3uphzk6OtKdO3deM7cA8Z/+6Z/Wpz/9aUnSf/gP/0Gf/exnMxoZjAO7YCw9vL7I8UQLpkmSvEYwfTMWkqedecBwYPDP5+Pg8wRDP2+ewvOex7T57zvQYChuDN52SQEMnKtUWj02wWP3JEkyYqm316+NYXOPnqXBu/G5VqsVgh0hGUY+nU5jyT33uLGxobW1NU2nU21ubqrb7QblB3DYQoDJzOMrYSmu93jqmv5gsvNcWgqsyP4kSaJ+vx/VsPQp4ON1Kr6CWFoVkPm1m82mNjc3w+A9G8MYkNF4+PBhRhimr2F3eT0LLYNxlpQxeDaU3t/fz5TFO2i7XnReOO2OCKAGdGnjZDJRt9t97wqmbwRweZ0jzzDyxi1lBVE6mYFxMeu8cCV/TkKFPBPhu6RL84uViHfzYp2Liw5GnN9DGyYUgODrVnydhJd3ezvy9SAe7nhdhusCKPt+7wiCFF3x7FnuhwImaL8DOR6ce3JPjTDqa2XoV9LMhEDS8rGZtJn74rm2tBdGxpj52DoDabfbUafh3/WxpjL29PT0Nc/sdc0H5+CbMfk85eDJeWmaxhoVKbsOxxmmz+nz2sfhqXPXcvLXf7vHEw8eb0XvYPIzaLzm74PCbvwuMrrC7kDCwXn5nJTdmNiv4zqJhy60CcT3cIa/MR6/FtdjYgMITkfxjvl7cyMmnGG/Dk+lEjbwGYCA3/SvtFqYx/+sm5GWBgCj4LtedFUqlaL02uk6hufgBoCyQ9l5wh99Xi4vH/bUaDS0sbERlJ6+IHwiZUrNCazL07j1el2NRiPqKTw7wThSgt/v93V4eJgBcJ8LznRhb9wfoYyn3SeTiU5PT+MREu5UHHAZJzQd+ot5mdfVPFXuGtB7WvN4o2wLk+08FHfA4H8mCcbARJSyez7k6wnyIOIA5SEGg+QDx8TOZzf8fO71aUueivr33HC4ju8M5svNARt+2KOU9jJJvXiIJ73BomgfmRiyIPQdhlgsFtVsNkMgpT8IG7hHvsdeFZLi+S5+/1yTTYKc/ZFuRmzlnGSTSqWSGo1GRqBtNpu6evWqqtWq7t69q3v37oXRO/PY3d0NgCGkQnBuNBohyj6i/K9heDyYivnltTaEXez65aHHeDzWwcFBhkEwzg6UiKGMn6dvDw4OXhPmuZDvji/vIN/u8cSDh/Ra9gGtx+ABAhR26bWZDNR8jIdz4jH5DE+MZ+EXsac/y+Q8PcXTtM5kGGBP6XE9zpOPW/MiLPfoYZZ7VdruMbFPNry/i5HuRX07Qt5nDQrtZj2L6xOSYiVtvgCMMUA3cebBvWJYgA0hGUBEtoMtDr3uwlkQ8btrKoRWFGBhwKVSSYeHh0rTVN/0Td8UomSaLhecYdDst8GcIvtEgdXZ2Vksi+dIkqX4yUOtisVi6B7e7mazGQI5oDoajXT//v3MvGJ8GKNv+qZvCoMfDoexuNA1k/e9731qNpv60z/901ia72G5ryLOhzdv9/iGAA8OBh+hym/eQwgvguJv/7yzB8/9NxqNTJYCtX1vb0/z+VyDwSBKhl2IcsbD4DhlxaCcXtJWDCgf5vimP84eACpfwCYp86wUzkP5sguPgCnGjPHg5SUFSLTb7TAiQhxJ4YWp2+CcUH/vewckzsPfLNenLiRJlpsLUT9SLBaj1gGRVlJsoIMRIJZyXq/mJG0sSScnJyqVStrZ2cnoD2m6XNZOdW23243+wfgJ6xBFe71ehqXxXUJH6joIhfgsr3Mu6kKOjo5eEwbTB/V6PdbCuMCOhsQ4w4ZOTk700ksvhROUlmzv9u3bevXVV2P8/sKAByierwb1icskgwKPRqNgGpwDYwc0MEJCFhdNEQKZDDs7OxqPx1HVd3h4KCm7SI3/aZszEIzKY3fWITCRYSG87pvkAhbOcLzWAhUd+p3fvwMPhxcm0+GZISblfD6PBWtkWQAVwhbfCMgpNtd3VsU44L3X1taiZgGjIrSTVovyeB0twlkVISt9Tkk4IQjGCVNFvDw9PVWv14sQljUpa2trGo/H2tnZ0Xw+18OHDzPp5DRdFrrRbgdJmBl964zw6Ogo+oS+cOfT6/UywOGCe6FQ0PPPPx9znv7nXIDTaDSKbRBoO/U8aEGS9O3f/u368pe/rF6v945TtU88eEAFiSsBCwzUKTSvsVgIY8NL+f6WMBQmOyjv6jwTGCW90+lEXcHOzo5ms5n6/b56vZ729/fjvB7juqCJkfmSce4JwwAwXAfxjXHOAyhnLJIClAABDj7H5PPwD2+WBx1CBo+jfdIRkkDxSSfDnDg/IMXneDQD45TfOdwzXXhwfw4Lq0thDNSYMIaMK20fDoc6OTmJfkefof3oH+gblUpF73vf+2JvDepICIOcAa+vr2cEU4CAcKLRaMQy+XJ5+XAsacng7ty5o+3tbW1sbIRe4UD+oQ99KM7jyxakFZPC+RH+ucAuKe4Bh9BqtdTpdF6j573d44kGD2isC4gunnpcCLV0L12pVGJDG6eYGIik2ETl7Ows1haQC+fAe1YqFR0fH+vo6Ej1el1bW1u6dOmSqtWqtre3g36yNJs2unCFB2USOhh62/Ho+a34JGVAAW/Jd7hPvBpgQXk65+U8HspRsQiQevWpr3LlntrtdmZxmae4aaOXY7v4ygO5ea1QWJabU79CX6CFEPaRCgakMAD0rA0AACAASURBVMpSqZTZeGh9fT321gBwuPfJZKJWq6VmsxmCLiGS19QcHx+r2WyqXq/H/hr9fj8YhYcxOAD+9vGWpEuXLklSrIYdj8fBOAjhACja+NJLLylJknh6H+B4enqqarWqVqsV5f6uxXBdD3GopKXN7XZbR0dH792wJa8y54VKCnE8xVgqlWK1IaIVXrharWo0GmljYyM2YFksFkEroclMBACEbfl8Y51isRhPTsMD1mq1WNvx4MGDDFOSVvUlxJwYmN9bmi5Xtrp24lQdSu7gcfnyZR0cHMR3XMfxyQG9ZhK5HsTnfZu+0WgU4MNDm7yegxoPdBG0C0CHe/UUoot2XssiKR6IzVjiLQGu09NTFYvFeKA3rNCZFmPkD4mq1WoaDAaxQnV3dzf6EPYJI3GvnyTLojPCgUqlomeeeSac1MnJiRaLRbTHNR9ptU8o9wLLoT0OvLVaTVtbW7p3754kZZ78Nh6PYyx8f1nfhImxzmdWPDUPiIzH43Bu74R5PNEVpsViMaXwx4U+1yxefPHFqG4kLm02m5KWFJu9HyktZq+FwWAQ9NFFVBZnMZkwDH7wgGz51mw24wHTXrQ0nU51cHAQrIAJlS9Owsioo3Cg5DMADADomQaPnzudjlqtVoCif1dSGBMZhDxoAAwACB6QtCeAMZ1ONRgMMmzE9YrNzU3t7Ozo4OAgrufek7GhTWQzyuWy7ty5Ex7UmRJ970VezWZTly9fjtDWN4pGSASs6F+Mx9sBaHu76BeEcjz87u5uhE2np6cB3l7DIinzvBs0HlK8tVpNr7zySgi57XZb0hLcz87O9Ou//usqFAr62Z/9WQ2HQ/3yL/+yJOljH/uYbt26pU9/+tMql8v6sR/7Mf3Gb/xGANRf+2t/TX/4h3+o7/3e79Xh4aG++MUvajwe6wd+4AeCLQ4GA/27f/fvJClSvjdv3rxQhemFwSNJkmuS/rmkPUkLSb+apuk/SpJkS9JvSXpO0k1J/02apkfJ0hr+kaT/TNJQ0t9M0/Tzb3QNwMNDlmKxqBdeeEEvvfSSpJV3RI2nIxGySGdRD0BtQaPRUK/XC/qJYTYajYwR+wQAwYmhvfAG8HAkH41GGg6HOjw8zLAmD8Ncy/ACI4CS+2FhFqDgQi/fwQi8ahSG4SIr4YCkCAvRDKSVR3PDc/GY1C0aghdY9ft9bW5u6vr16+r1evE+egVGCLDR5m63G+/jNR2QmKeAIYwS9rCxsRH7dvomQNPpNONYWHsCCPljFzwEhqFR9+MhSqVSUbvdDsFVWlXEUiHqGt1isdwbFgF3Nptpf39fo9FI8/lczWZTSbLcCf309FS/+7u/q5/+6Z/Wxz/+cf3Yj/2YvuM7vkPdblef+tSn9P3f//26fPmybt26pdPTU33kIx+JOfn5z39e0+lUly5d0tnZma5evaqTkxOtra1FyLe/vx99QpHbjRs3Hnt5+kzSz6Rp+vkkSZqSPpckyf8p6W9K+oM0Tf9hkiR/X9Lfl/T3JP2ApA8++vmopP/l0e83PcrlsjY2NrSzs6Nnn31WH/nIR3Tnzh2NRiONRiMdHR1ldILFYqGdnZ0wDGnpUY6Pj2OSQOfW19dDF/G1CNJrKZ3H8hQLpWmqXq8Xu10DIoQfCGsPHz4M9sK5vWrQ43yMG2+5vb0tSRkdhtifg+uwaI2UIOcmncdeE9wP3ppJjjEAHpyHceAeer2eNjc31ev14hqIqyj8eHmMGVZAGTa0HiPyxYEYqt+360K+JgdB9d69eyoWi7F7+WKxXJkK6HrFa5qmsUDPF+URdqAhAKiSIuyZTqfqdDqaTCba3NxUkiS6dOlSBiQRLhFSSYevra3p9u3bmkwmarfbEbIcHx8rTZdrhCTpH/yDf6C1tTX9+I//uD73uc8F+BcKBf3Wb/2WXn75ZX3iE58I9jIajfRP/sk/0c///M/rpZde0qc//Wn9m3/zb1QsFvVX/spfiXuTpM9+9rNqNBp66aWXMvuivN3jwuCRpul9Sfcf/X2aJMmfS7oq6Qclfe+jj/2apD/SEjx+UNI/T5cz8f9JkmQzSZIrj87zuke1WtX169f1gQ98QM8995xGo5E+//nPq9PpxGRDWOTver2uVqulXq8X1NJz5+VyOQTPSqWio6MjHR0dhY6BFrC1tRVg4JkYGAyGD3VmwgIkhEgYTKfTidgTBuAio4PH5cuXYyIx8M46jo+Po3oSr+rCpcfuxWJRrVYrBEFptcqTdrMbFn1KeMKk5X28J4KnpMw6GARbT3mm6WpHLsJARGV2X6eoy7NbMCheh0UB8DBI7vf09FSbm5u6c+eOyuWynnnmmWBUrm3QVsAScEXbms/noXXwWT8HYOR7iAyHw/D4Ls76s3KazWZoMJLiO+wUNxgMYp5+4hOfUJIk+p3f+R399b/+1yOlO5/P9eM//uP6pV/6pehX5v6HP/xhdTodffKTn9Tv/d7v6Yd+6If02c9+NoR9apm+53u+JzP/Lnq8K4JpkiTPSfqIpM9K2gUQ0jS9nyTJ5Ucfuyrptn3tzqPXMuCRJMlPSvpJaemdn3322eiUP/uzPwuhCZrMgENDmZD7+/ux3oFCpHq9HgY5m810586deM4FbAHdYjabqdvthnK/ubmpRqOhNE0jfkVE9by8pNgEl3Co0WjE5Dk5OdG9e/fC4DB2j42vXLkS9SR4G3QWL+cuFAqhucAy3FMS4xO3S4o6GErKKRLjPugL0p14T4wY9kDGAJaA0bP8HtoOgEkKHYM+ItT0MIG0K0aXJEl4ZSY8DAjdhnUmFFxtb28H++n1emq328GgEMI9pHCAHQ6H4Uhc4wFgPIwiK5emy8cpvPrqq3rmmWdi/1LWvJCyJqPH9ZkXgCTj+eUvfzn0kx/6oR/S9evX9bGPfUw/9VM/FTuYfexjH9Pf+Tt/R7/0S7+khw8fqtFo6K/+1b+qa9eu6ZlnnlG73dbP/dzP6Rd+4Rd069Yt/fIv/7KKxaJ++Id/WL/927+t8Xisv/23/3aGzb5tu3+ngmmSJOuS/m9Jn0jT9F8mSdJL03TT3j9K07SdJMn/Ienn0jT9zKPX/0DSx9M0/dzrnbtSqaTf/d3fHQujMESviERlB+UR5citQ4u9NgRPjdDmg+f9AZPBo89mM62vr2t3d1f1ej3iWCYAYJLfSwEvhbcfDAbqdrtRqITe8a3f+q3q9/txbX8+R7vdjifNE79CwZ290A882hBPiAEUi8sHCUH7P/zhD+uVV16JzBVAAQOBhQDQ6+vrunfvnpJkWXbPGhZpxTr29vYiQ8FuX97XCIeEAKPRKFK5hFWuBzGmvqgM8GIbRCpgW61W1PkQkg0GgwDzdrsdY+rrgfDijBel3Yyj94kzIs/+AWytVivmDM9yKRQKEeIxZmT1vPiu3W5nrsMiSowc0AN0nWHCjug32I9/DjtC+5nP5/r3//7fP/4l+UmSlCX9tqTfSNP0Xz56+SHhSJIkVyTtP3r9jqRr9vVnJd17w8Y9ivnJlDBh8Uj1ej30kK9+9auZLMHJyUlMMpAfAwckML58ERZCm5cV+85Yt27diqd+oWtAN4nzOT+MhIGrVqtaX1+PvSI6nY6eeeYZSYoJi5A1n89js128r7QSD/G4j8YiJkqpVAoGQJzPRCwWl3tlFotFXblyRV/96lcDIDy1C9DQ7xwernHfR0dHAT6j0SgKkZj4hHC0lfCDoisM3cMs2Fir1dLx8XGACvdYqVTCw5O+9IwQTMrXd1AZvLm5GeHKcDh8zW7rgCRgkgcK5oizkLOzs9CLMNDxeKx+vx/1GGhp0ioM4np5PcwBgfEFgEejUSY962yRz/KbsNFTtIRf6H4XPS4MHo+yJ5+S9Odpmv6CvfWvJf0NSf/w0e9/Za//rSRJflNLofT4zfSONE318OFD1Wo1tVotFYtFXbp0Sdvb2xGSHBwc6Pbt2wEmGIJvQwdo0HkMFuXN7jn5HgPCACDyMblPTk40mUy0tbUVgODL2jGIarWqer0eA857xWJRL774YhjY0dFRZt8Lah6IxSnj5p5c1PV2N5vNAD1Cgfv372t7e1vNZlO3b9+OdrGBzWAwiOfMAmxMMPdc0G2qPdGBMCKf6AAlu7NzwARgLLSVceMeGTPfOV1SLNSTFAVe169fz2g0ACW/PezCsSTJ8nkq29vb6na7sRKWbBDOAOd1nnF7Nor5M5/PNRqNtLa2FpXEVCbzvoM/oFYoFGJjIzQzPkv/samSMyGuzdomamj82cOSMpWxzFEPtS9yvBPm8d2S/ntJf5Ykyf/76LWf1RI0/kWSJD8h6ZakH3703u9pmaa9oWWq9n94swsgLj3//PPx8BooL5V4ly5dithRUtA9OpWFVRRyjUYjbW5uxkBiGE7xqPegg6GJvrKTuPzg4CDSdgADXk1SZh9QzwRhaKRBvfCp1+tFGhRNhMMFYiYKE217ezsMnBAHyj6dTvW5z30uJv0HPvABfeUrXwnBtFQqxSMRCSN8f1DaQhiEfoFRE+IQVvrCPU9DA0oAKNdYX18Pb+zl+YAKWQlPwzebzahroS30B+fwgjoP78rlsjqdjm7evKnd3V1tb2/r6OgotJE8C2Ee5YuvmI+EdQBiv9+PDYur1WoIwty3tFrJ7UV73ANOylfBUpAG4OMkSZ0zD2gbQMz9oNUxlhRYXvR4oovEyuVyevny5egYxDQGQlpOxlarpeeeey7iWa+SlFYeik7EE+HR+QwpSzwBMbGnswh1SJFBYzc3N9VutzNVqgwMr0nS7u5uGIBnELy4bDqdhrCGgfm5HIigneTymdzj8ViHh4daX1+PVDYh397eng4ODrS9va0HDx5EdqlWq+nKlSuSFO1AIAZEF4vVilJptQKXMIGxarVaqtVqsbaE+0eIHQwGmf0+OTdeH/YBdQcUYRStVisefUB5OdkKjJpz5bNalJ7DagqFgra3t7W3t6d+vx8gBvsgTEF4B4wxcPoJ8RYw9LoQ6i0APoToYrEYrNSXBwBAhLvMQ74P88BRck7msgu6OEFEcxzn2dmZms2m/uAP/uC9tw2h0+bFYhE1DJ7fJ+Z9+eWX4/N8l5JfqlR5MBEI7YONd/KYmwGV9Jqt81utVsT6ePokSSL7gYAHylPkRRERYiETDHqPFyV0kFardjmve1M8Hh67XF4+zhB94ObNm7p+/boGg0F8tl6va3d3V91uNzz3aDSK2hVnBUxMDBbDQbSWFJ6Z/qT2gTUnksKgyUy41sR4SKvQx+8H9lYqLZfTA7z80B/OTHAUGDXPoCHUwjBhE2hK6+vrEWYCFHyPUn7W4ACEtIFzAYbuwDY2NiJElVahj6/69XmPA+v3+5FdYi5Iq2I5/nf2AgOCuZMmd1bIOiB0moscTzx44HExYuI9BDV0BoADj0zHsnS+Uqno0qVLUSoNAHEOqg2dLnOdRqMRNQ7SauB2dnbiKeR4l6OjI7Xb7fhOtVrVlStXIkfP4ifCHGg9nsIXOpFhYFs9ryiVFLuWw5jITIxGIz148EDlclm7u7vhCefzuT70oQ/p5s2bKhaXa0SgzugoTr/JStH3o9EodAUvhqOYTFKENFtbW3GP9A9elnuRshs9edrXQWM0GsXiNNLKbGUIs8FYqTp1jywpmBcsj7HxtlDc1Ww21Ww2A8gZbxgAcwGAc3F5MBhoa2srxgWAYjyZw7AZNA80F0mh9VBoB0B52pjDmZWzWeaW1/N4xa4zmIseTzR4SNn1H0wQBwqMnfehrVL24dGIr+VyWZubm1ESjBLtaj7XGAwGsYya8zHxmDi7u7u6f/9+ZtCYgFTFnp6eajQaBXNwxZu2uVKeXzMCaPAMWWkFYEwszjkcDiPvj7fEQBBPYXOEPx6zA1ykewFhzkOK1FevkoqkP7knaVWly3fI0ACaGF2+CIwfWBmMT1L0CcwCQEJsZBGbtHouC84BAdUpPUZHytdXp3IPVI9i0BSo+SZB9Clhh6dRaQ9tgUlxf8wNNi+in/if7+AouQfqmGAuJAZcvJcU9+NVwj4+FzmeaPBIkiSzO5akzMQDLByJ+byr+JyLju/1eppOp7F6lXDB94WYTqfa29tTu92OykxCHVerPZ5mwyDWWMxmM33pS1/KCIR4Dw6+yz0ACK7MAxCwAUlRsEYYM51O1e1242nrTBLi3Fqtppdeekk3b97MpHIR3UgtejjhB3qCM0FoO+wCqo3n3NnZCYaAPuOl8R42QNUxLtLwGA+G7vG8h6+MB8VYjCmfJfvD/w5Eno3xtDoATGjKOTF4jBl2RkjAPPEwl7VVpI4R3+kX/1tSpv0eYgPkzAfYCkDr5yY8ZnwBNcLRSqXy9UnVPo6jXC7r6tWrgeh0MutZELqk7L6m7KokZZe98z5eH4ZAtgRQmM/n2tzcVL1eV7fb1cbGhk5OTuI3sTV0mqKxVqsVoufh4aF6vV4YGHqBp1klZcBBUkbP4HXaBkWm7YRUhF2tVitqCnwH83K5rCtXruju3buR1QF0XGhjXYpXfLp6j6H7Vnr9fl+XLl2KBzTjhVl4yLNcMCo8MP3kDEKStre3M8CJZ/UsCQCHsElqGyOn31iDwlwClHhIk7R6zq2vywHcARZnU/QNTMb1Gq7NOOW9u+80xpIATy3THgdD2uxODhCAkeZ1OkCNsWPbBGwBwKDdFz2eaPAoFJZbw4GwxM5ra2shhI5GI52enoYH9eKxJEl08+bNTKpLWopX/BQKBZ2enoaBFwoF7ezsSJL29/cj3QlVxavPZjM988wz2t7e1mKx0Msvv6zhcKh6vR77ROI5SqWSut1uZCE8fnflnokLiGHYTEYMxQvfeGaIeyaEMq9UJE3Ha1Syund2rcAzOpzLQxraRfjXbDZ1enoa7ZBWzJElBb53CNTbPS394GwRr0v76B+AwvsEIKYYjNcwKlgQ/cA9AGpoTh5+whalbB1QqVTSpz71qcdnDE/g8USDBwIaQFAul2NpOGW/GD0ITBYA7/Qt3/ItUX3K5CKFV6vVdHJyosFgEOfDo/E5aD8TmrJwPCQ0lCzHgwcPMnUGnIvJ6c8fwYPhNUhX4sHzBUjsa4pXYh9Knm/KOh3ai7GxiK3dbkeREUZFFopqS8Q1YuQkSWKtim8iwybEk8nyOSNeKQpY4S3r9XpUCVNDQiET9NrTjtRFwHLYBMhDHRgmOlSz2QwwARhpv4es3KdrHIVCIeonPAzj+oAyQItXd43nL+LxRIOHlF0fwoQB/SkUIx2FwUqrrfbTNNW1a9d0dHQU6yyg4ZPJRKPRKGLkZrOp7e3tOEc+/q9UKjFJKY6q1WpR4epPSgPMELYIqwAUAAvj9cIdD7NgOdJK7yGNSg0GYOmA4U+Qf//736/j4+MIo6rVamwiJK12NsPA0C98R3Df9wLR1AVGr5uBKSEsA9qSIn3tD4qSsqlGdhDjXlutlgaDQSwGdI3J73k+X25u7BkiBwK/V+7T6z7W19eD4jN+3CcMi3H6iw4c0hMOHgg+DBppMq/LwLvhlTw+hUJjpGQ88LjQ03xacHNzMyb95uZmXG82m6nX60Uc3G63dfPmTc1mM3U6ncx+ILAmvKzn/8mOIEJ6ChYKDvvgXrzNx8fHGg6HETKhwTChvR/a7XZszozBAWQ8+8NFSy9oA1w9u0VxXH5dBjusS4oyb2ofMECqSTc2NmJcASCAxscOcW9jYyP6HOMFpDyTUq1WYxc5Urqcj9AM1sb/ABGiLDvuI6IWCoVYxe1A/05Wo75XjovnaR7DgdF4npzFVC5Y+foUjJG/mbzFYlEf/OAH49xJkgQN91JdJli9Xtf29nZGhIKOo6vgyWez5WrbYrEYTICwBO/O3y78UlE4Ho9jvYdv7gvYoKh7yThraQAJDI/FWNLSiK9evRp9QJ8Ui0V1Op1MJoT2kGnwVDjMAe8OM+Lw9Ot0OtXJyUn0DfUKrIwlnQwLAyi9DTxcG7ADyBGPfaUwu6J7SINjYTUyWQXmiQM5TMrbw+bE3mcwVNrx9HjCwcNTlo1GI9KTPsER8PAo1FFAnZ2VSNK3fMu3hEfHk1y6dCkqC3u9XiydppBIUizywogIU6RVeph6DUQ+KbslHiELk9jTg9wH32dCM/F9YR5FXRReedaCMnMo/cOHDwOA1tfXNRqNdHx8HCDi4iP3QN8BgrAMAA3Pm69f6Pf7ATAI0Rg+Wg5gRFh06dKl14AtoQhpc4ydXeGphD08PNTJyUkwx/l8HruSecEcY8FeLlRY+rUcEAjD+v1+6GroQOg0njH7i3o80T3glYuukOPBvciK972qUFKo/Hi2UqkUK095zes7KpWK+v2+NjY24vGDTH4vvjk+Po5l854hgPbjNfMl5WQuADNX79kO8ezsLPbx2NzcDFGWdBuCrBcEUfXJps7ScnNo9sr01a8YP/3lRXBUTwI4ZKIAcE/bEkKgi9Tr9Qg1fOw4L5kXaQm4rJHhmgAZDKJYXO2NiqZVr9djP5R8jQasAmeRDy98pzAe4AXYSStRHFYKMMPkyMZR0/EzP/MzsVs+12U+ek0J9+s1KIA9Ar+L657hKhaLwWphnQAq6XlYuJfxe6hJKprx8sxXkiT6tV/7tQvZ5xMNHqjzLIDCA0FfMQhCC19FSKzqZdR4i93d3aChpVIpljr79v2UmZ+enkbsfHx8HGIjYQztY7Dm83kMDnH9eXoG90OKEsZAWTOfGQwG4QHJKmEwvjPXZDLR+vp6GHuaLh8NgAFIyjwwCYOh6hNRFAaAnkDBF5mNJEliK0ZCSBhfnllRjEbFptcvoAkRUrknR5AGrCRFtWyv19Ph4WEU9RF2YgywU+YBTgGj9Ape2BvhsL8OoGLwSZJE7QzGzHXW19d1fHwcfY02hKbm4bNnuvifueBpeB4dUa1WAygBFd8wGQDywkhptYDSM00uQqMHvpMQ7IkGj0KhoOvXr2dW0fpKVzoeY2LiO/Pw9BuUE11CUlBc1khgvPP5PPOYQFgDlBsPwuTjPOgZXvbM9fGmeFhApdVqxXmYOGzi2+12Izzi84QZvgMXHhXxdGdnR9PpNGovOIeHc5cvX46qUK/S5Rr0rz/ACk/Jgi2Mle8zQQFlrz+BPVSr1QgNJYXhz2azqFshtKD4CyOCPfgYcD+wN9LphHqMpy+zd32kUCjEg8199TVMiTHzR0fglAg3cSherAhoY9TMW+all6gDbpIy4ErtEKCELlapVHT58uVY8Mg8k1YMDAAmdHOnk3dsFzmeaPAAXWEIPDgHb+ArGev1emwB6EVNiH3EtR66+OZAnmKFGubjfmcwsBAPBwAJRDjADONikBE6CQ1gO9LS+Or1uo6OjmKCoXvgVcgK4M0xGkKmQqGgdrutTqcTExaBlftnn1RiftpOG5xJuTFAdVnM5cveSdvm16F41giQJJRknJwh4ZV9I6HRaKTbt29nQisvEnNqLq0WUvb7/VhnVCgU1O12M8IpzgDH0Gq1AnT9wUiAEwB3fHysF154IbQTfiTFfQHKXMv7ktoV5gvX4bMwNMabUBVGnaap7t+/r+eee06FQkGdTif61DdOgpl7+IbN8PdFjycaPBDY6HBJEae62AiFRNCUFEYPo8BYfaWjZ3FINTrQSAqRkoHDQBhYBx4GHAqJ6Jpf+YtGgLApKeO92RsVY0AA9mpMDJCJjTFzr1/+8pdDaEXUJNTDGzGhaQMejYpL7hMj88wX36Xvm81mgJWnVTc2NjIPiELD4Hu+ToMsDkDqe8/2er0AHs4Bu5FW7BMRdDabRSFbv98PrQVmUq1W43mtOClCtFKpFAsivXQeyo/esL+/H89GAZQxWIyV87sTkxRzDqbjGUPGCoB8//vfHyXq1WpVX/rSlwKU7969q52dndDGfPEc2gog5iEyx/r6+sXt88LffAwHMRsG4mXa7XY7M0CVyvI5JpcuXQrRzWNxL7qCsdDJvuuTPy3ON4thcrjhe3pTym6GK60QnoIsrkeY4cvYyZw0Go145OB8Pg+dww0OTQWtwouyut1uVOEeHR2FV6UkvtVqRb9ICmrsXtMzQQ64XF9a7SPi+grj5KtKAWLA11OkhUIhnqWTDynyNR0AMmES90hY4qGGp38ZL0TcTqcTDmd3dzfai1AtLUGLdUrobYCPZ2gmk4kODw+1sbERYAx4cNA2HAj6Cilor98BnADP2Wyma9euqVwu68aNG6rVanrhhRfivujfTqejy5cvR5W0h1+eUmeO0teUCVz0eKJTtdQEEJej8heLyz1EEfEwTkq0iUGhnS4kuTjope58Vso+Cd6LkWiTr/DEoBjIPFNykdA9OOs6oI2FwnLbfbb9Z3WstKKfhGZsbkQGBgODkVy5ciWzYAyQnUwmUTCGEbK3CW0BiEgbE1JBm2EH9Dv9CcPjfgBSQkoXh4m1T09PQxcpFJYbKUvL+oy1tbVgYGhBhEhoALAtwlp0LTcaf0Ic/YBOcffuXe3v72e+DzBTk+IZObQyNC6yG76xMJoatSqEvIQTnJ/Di9ckZYrw9vb21Ov1YnOnW7duhZPzbBILAD0T6Xbh89CzlDDMix5PNHjgYaRVPMjh3stjOp8IDDqeL686c25EQJAYTQHAcKNgIoPceGkfEPeYMBkHGz+XtCqyYqMgJpAzDl/05iliVol65gVgJcyCmjLh9vf3dXJyEruIu+ruBu4PrBoOhyEe0id4czY+Imz0yU3fc5+SIr2MB/dSf3QW+nJtbU2NRiO29ePZJ/Snr5NxduSZCzw/YSlAApifnJzo8PAwGAcPZ8Ip+CptGAOMi2wTYEk/e2kA77lAyZiiD3FO2nv9+vV42Bk79V+7dk0nJyeZFPPZ2ZmOj491cHCg4+PjAAOYF3Pea5HQmNDMLno88eABpXcm4Io2EwHvhXeCmmNo+TiUc9Oh+WwEOgPGwKBgIAyOi4UwChehnLH4PdF+n1Ss4mT7O1/nQm0BQhwARcahUChoMBhoA0HqtAAAIABJREFUe3s70oRpmmZqBNBJNjc3NZ1Odfv27YxHxqPyt2eWCJNgVwAubfNCOMCcfibL4iwNLwirQRhn3PgM5ydFjiHymND19fUAT0+Lutbg+hfzAHDjs/P58ilx7OnK7uf0mdeTEALSR7Abr51gHrJOB6bCPblO5hmX+Xz57FpqNzz8a7fbAS5UEnuG5sGDB5kCQvqL6zCusFU0l4seT7Tm4XGvC6T8ZnIwIBgkkwQtRFptnuPVnbAWhDcXTxkU/038jDECXpIyGgQg4uXNtBljAJC8ArNarca+GIQbzWYzluqzOS/hBnE4XrdQWG7EzNPpWSnqbYWJESLcv39fV65cyWwhADiRnnSdhfRlXmh0RietVt3S3yzUI+73xXaMB6yR+6Wv0FLa7XbmKfQslGP8WBUMU+La9JmzMwDO61zwyt1uV+VyOZ4lS3bJV0HPZrNYRyMpdjMHMBByGX/mHozEC9KYizAjWMLx8bHm87muXLkSjowUK1oX85IQ6uDgIJ4OwH3TFuZdPjt00eOJBg9p9egCioGcFuPhCT8wfgaB2BjazAQAkJzS8ZrHyxgptRcOYi4weqqNiQE4OHXEKL3ilEkEHcZrcE7ibEq9AU/qIVxkLJfL6na7YQTE7YQ2iLGtVkvtdjvaf3h4qK2traDPnAsGQz9jYP4ZHo7k2Sw2D6ZPvHrSaxt8L1AmNSzChT3GxWs1MGj6wxcgTiaTAETCk/l8HmIi9wOg4PV5bgpzCtbAxlBkv2Cano5FR0Islla71wH8m5ubGcBkruDguEf6HV1OWmbghsNhbEXptU44n1qtpps3b+qDH/xgZOf8/pm/6C++YvwixxMdtkgKysgEzFdGkgp18dTz6kxYBEdfe+HxqLRiNdJKh4C2Y8R4L6/U81QcPy5QObvA6Dz8oo3oDQDW5uamCoXVoiwACQPGe+fpt7R6MDTC7GKxXOcirXQhLxTivtbX1zPemnYT1gGC7i3pR0r5AXhoNe2mPbQPlsi5SI8Csi74MTaAIkB3enoa4OUhEe2lqpYHm7M0gVAiL2QDPmgr8/nyOb9UwcIAvIaIvkJjYa5Kq1CQHd8Yf+7JQxjOA0t79tln1Ww2gzmyQz9jT22LpGCUpKZhYrBJ5ivAC1t8z9Z5OKXDw0O9pVVBFxSMkl5P6bpQyYDyWAPYi5dU07kcqNsMBBWPsAbawDUlxeTHoD2+9lDFwQdq22q1MmGAVzBKiqrX/D4jpVJJ29vb6vf7AaTSakJ2Op0ohDo9Pc2stUAopr8cgDm/pKiFob+I4wkJCdO4N87B/fskpt/RM0gZ+9J3+tLFcPqRSlNAiXtiM2FAg/aSyqWPEYXx9NybVwEzdlz/5OQkCs4864azQJeCmXghnLNlzufz1IvQAPNer6fNzc24Xyqt2Rv2wx/+sO7fv6/Dw8O47ny+fGjYtWvXAmhhuQAVIEn7Lno88cyDwUfgAUkZeIyKzvBl615TgCrOYHsKjclcqVRim0AmNqlfqKK0ouLONvy8IDrGAqV2oMBAJYUm4Kk0QhcfdB5u7F7TMw3s2AVIEuuy8RD9B1A6ZW21Wtra2or78wyJU3GMnb8lvcaI+B/W5eyB+yuXy1pbWwtD8zVJ/ACQaACe/gUQaC+bNLHeRlo9N3g8Hkf2DPbItoyFwvKpgzAggAuWwPxik6HhcBgPKKcvCRMBNPfytM9rZGAwtCd/30mSqNPpqNPp6OjoSN1uN/SV/f19ffM3f7O+8zu/U7du3VKxWNS1a9fiub3S0mF0Oh01m80AWE+hewj9ng5b2u12DKwXRDEweFkAwDMheA6AgQHHqD1DAw2kBF5abRCLV/LY3nP9GEQ+ZSdlRVLaiIExoPn6D2pF3FuyDyg0GqPEs6RpquPj4xBIm82mjo+PM3tQoLe4vrO5uRl7ttIWvofXd4OlD73wzJ//y3NmHES4L/rcdSYpu5aDNvi1PLSgb+fzeWajIsKMo6OjqMmg/sSrLAGPk5OT0DtgM+12O7w/8wrnhbHBBClOwxj5rC9+Q2/gnmB+vOYg7p/3YrWbj56xs7Ozo0JhWQvzxS9+UV/4whe0WCwfvXB4eKi9vb1MAaRnURhv5hR1Qi7cXuR4osMWR2rXBNrtdqS9vJqPEILBhH14B/kERpDCy5B7J0Zlw2N/ulaSrJZreyrV9RIPZQAXD2EQCxG8EIRdZwEYGGB0Ds45GAwyz2lNkkRHR0dK0+XDqHq9XpyDiem7e5HSlRR7YNDnZIp42LZnqZxt0b+SooCNUMX7y2Ns739CH6oz2feUdngoRD+cnp4GY4EJ3Lt3L9ojZR/exGrkUqmkXq8XIRZtJSQAGBCWCVW9XoLQmWJEDHFjYyPGmbHMjx8AJ632f3EGCVj6PAR8b9y4ofe///369m//ds1mM331q1+VpLgeT/8jJOZer1y5EtkxbIVQBQ3nPQsepKZAeK9BYOARVFHM0T1QuQkNEAdhMLzG8zxKpeWjDFnu7VWGiI6uMQA+gA0eHSrvz1Uh/MCjulgFhYftzGaz+C4e0tcoIJJB6xeLRSjweFoWSR0cHAQzgMKzWtdFR/oWbQP2AjBsbGxESTX97MDEjwuB/PY6BK9lYGzw6oVCIVaQYqwIlrAvQo/j4+MQcamxgB0xRgAIug5FaWSVdnZ2YlMkGBbjwUPNCVvRQvgczAkQYluGyWQS33NNDF2D9nEt/uZ8zHmKFJkTSZLoxo0bMTYbGxtqt9vx/CH6Q1oV4uUzPF6U56Lue7bC1A0dpCSEkRT7TECzmfAYtgtEFBm5eOf7eSDGejo2rw94uhHj93oEQItJhtF7pSHMiMFmMDFE96jsho6xe0UqdSn0AUvwYUVUZ7ouAHDAXhxcXTTmO3g+Jp5XdkqKGBvDwpgAUt/tje+gR6BP5FO4PqHJctCfaZrGjnKTySTAhYNrYCznpbIfPnyobrerTqeTaTd9wDU9tKXiFI2EcxLycq+87vU9zBHYMK/DvmBRXJ/3fVOmxWIRfzMe6Djz+VwnJydR44LWVK/X9fDhw3BwZPAYP08KXPR44sEDj+kCFfSxXC6H13Ua6KGLx8qka5kYrmKXSiV1Oh0Nh8PIWHjKFQrP4PCbSeSelwFyEdHDFddofC8G7hHRj7QyXgtwKBQKunz5cmaDnYODg8zGRTAKqjDzK46ZoIvFIrYFIAyBrUhLYGDjaO4H0EAMBtig7B7a8Dl0BdLlgC9hAuBMZgjtCcOkLa411ev1MMi82Mn9UApPaFKtVkMPOTg40NHRkZrNZhS1IdAWCoUQsavVaoSIzjL5G0dBuAlzw/C9FgWAdtHXVwQ7cDuQO+CgiflqaddSmDfMBc/ieNl+vhr67R5PPHhgpF7d6ZkNvBGhBBkXxFXPgHhMOZksn5TO5sOdTkc7OzvBRjDwzc3NMIj19fXodAqveJ9JSVkzHscZDfeEcXnOHSaEcCutnmtKetj3AYGuS4pKUKgwXpdaEWm1ZoUJz/nYRY04G49HAZi0elYNYqELwnhifjxtjfiXJKt9S9BPJGUYH2FKv9+PfpBW6Ux2ZD88PNS9e/cinHHNCabqDMrXb3iKuFqtxv3dvXs3Njfa2tqKPqK9vm+KrxPybB7OibS6Z/xwWJyHe8P4cR5k2DifC91ci4WEzpqc1QCUfIYUNrvdYT/033ta83A67AIcHQ2iutrPs128PgTqDOPAYCQFJf/85z+v7e3tWFuCwaEteEUfngDazeACPNSi4EGoz8inmz3uJ9wi5eeZEuoXvNIRYZPvsmZlMpmESOj3DKvx+NcnKalgvKmHWIQ6HrP7HqUYlwvZgAyswkMQSXE+aDgCqWtbbO8HG2SyO9uDrtPfnrImzGV+jMfj2IvUd6qniK5QWK6boZbi9PQ07h2vzSI9xp/X6W9WRKNDOcvzMM5ZBCDBvEcn4Xu+kTNaoLMR5ihzCuAbDAaxoJH2IvhSB3LR44lmHo7u0moTHAzOxTbXEDxD4UU5DLCvqISikh2o1WqRafEUJUCWL77Cu85ms1gV2+v1Yr9LBnRjYyPidbwzbaednIMQY21tLUKIcrkcz6Hd3d2NPuj3+2q32+p2uxHqsCQcMQ+RdTAYqNlshmGT0gOAYRf+6AO+6+lu71PXFugfHxdniYwhr5HRcIovrR70BcDzGfqCBV9uNIQSsDhpVawH2JMuBaxgkegVhE/D4VCvvvpqiNGwNA+FGUvu17M9aEMennjRGawBIGWOeAo8r+/hlLgOwMh7nM9X6kqr+hQpy7zcHi56vGPwSJKkmCTJF5Ik+d1H/z+fJMlnkyR5JUmS30qSpPLo9eqj/288ev+5t3DuGHziX0mR1qQz6Hz2pkCjyK8tSdNUly5d0gsvvKDv+77v087OToQ9169f13d913ep0+lobW1NrVYrPLxnSjxDwqRnMrtuAJtJ01S3b98ODYXyc2m1ua60opuo/RglQi/MAO1lMBjE7los0R4Oh7Fz1507d9TtduNxmtKyQvLBgwcajUbqdDrhzZhQs9myPD0vDlLyjIiLtsCu4fQtnlVaxeiwE/qM1abSajJLy6pfCvwkRXEXIRMCq1N6xvXZZ5/NjEu/3482OsgwZtTsAJywMC85R0c6ODiIGgrGxIVML5enXdPpcn8NB1CyaJ6yp59gtfQh+gzj4MV1Doi+SA7wpl9gVMwxvu+OA6Z40ePdYB7/o6Q/t/9/XtIn0zT9oKQjST/x6PWfkHSUpukHJH3y0efe8ICW4gGpAAXt8Y7EiXiA7e3tTIiA8VcqFb388svq9Xr64z/+Y41Go9iL4vbt2/rjP/7jUPVZqchAuzEhcPkydc8IMIDFYlEHBwdhAHnlHFD0yYFXGY1GEYcDlEyGdrsd2aXZbKZut6tCYfmAbkqYqbsoFApqtVq6e/duPJMGKks4g5H5s2MxREIpPCJCqi8XJ8uFR3MwpRDLax08I4HICgjj+RG3GTcK0orF5YO1WPFK7L63t6fd3V21Wi3t7OxkNADmiNeQ8NvrMzzti+6SJIm+9rWvhfBMewnrpGW5OPPg+Pg45iHbOvI/jMCLCn1za/oewOZv3/4AMPRw0quqvVDSQZA5Sh9Si/J1Yx5Jkjwr6T+X9L8++j+R9J9I+t8ffeTXJP1Xj/7+wUf/69H735fAyd74GrFLE6IZSOtFShQZsTclRoIRLxaL8NTslD0cDqPI5+rVq7p27ZrW19djpai0Ei09DuVv93bSages/OsbGxvxuoc9UEtWc7oiTmzK35w7SZaPAMD42+12TCJADyPZ2NhQpVLR3bt3M4CGtzo+Po6NnGmzlF0UCBWGgQGgCNkYGCIrnpExwUtiKG5M3CPnRbD1gjVpJZoy2d0r85s1MS4Qkp50Y6NdhBWMKSlk2kX2hXl2dHSks7Mz7ezsxDmlpdH3er0QU6k8xUnQ3/l+oU8IkWHI3I+PA2wF7cPraAAhwm6Yi4eO3L+He2hMb8EEX/d4p4LpL0r6uCQ2NtiW1EvTFBXmjqSrj/6+Kum2JKVpOkuS5PjR5w/9hEmS/KSkn5QUFBrv48iap9WlUkn9fj/z0CMmF0bM9+bzuV588cX43snJSTz+YHNzU+PxWAcHB0FVuc5oNIpqQsDBJwQeFQM5OTkJD+q5fLy1pBCvisWims1mhAeIcp59YZJ85StfCY+5s7MTnoXFe7SHZftMPhhTkiSRNej3+zo+PlaappGyZNJzn/QdEzl5lCFy0EKMRUfBc6KV+FoO+g3aTehZq9W0u7sb18ArSwqvia5UKpViv5JWq5XJSoxGowj5oPKtVitCQ58Pro8xVjAwDJ7xpCz90qVLId5Op9NgTsPhMFgJ4839cr5araZ+vx8g6HoHbad/uB8XYznncDjU1taWDg4OwrkCNh7iO/Az3wFFhPOLHhdmHkmS/BeS9tM0/Zy/fM5H07fw3uqFNP3VNE3/Upqmf8mVcs9GMCAo4+VyWffu3Yvt7EmPOfV278Pg5fcuZTAPDg7iPJ7bd+rrBV8MCDE+mY3j4+NI5cKY8E6kKYmHEV5LpVJs5sJ3vGgKAER4PT09DWORljSd6kmMgbQyE8wNnp3UidWdKRDSoQWgNTkTAyAGg0EAhtd6oBXh2WFH7j2p4OX33t5e/I0RInJ66LRYLOIRmvlSf/QFwJo+39jYyKwxIfQDJNkvlDAAgGBcJ5OJ7t+/r2azGaBF/Q3sB+Ml8yWt9r717A7MAFEY8KHPYBEAAPoITFpapcrpD8/YeI0TYwxTkVY7zl/0eCdhy3dL+i+TJLkp6Te1DFd+UdJmkiQwmmcl3Xv09x1J1yTp0fstSd03u4jT1VKpFJkA6hBms5nu3r0raVUMg6elsz3MSNM0HnPIxOdvgMDXo+RjdU9veuoMoMGA8Brb29sBPOT//QFBtAsDYNJubm7q0qVLcT0m4dWrV2MCXLlyJQRAJm2pVNLLL78caULSo2xHSDsRXvH4zWYzVmFifFJ2OT5hF+FFuVyO3cVZbAWrog9c43Gx1DNciLWAHZklNA/PxpDJ8FQ3AIe2ROUllZakTz1c2NjYiKwG+hm6FQbl7carwwIODg4yGQ+qXvlxDckzIIAIrBbtiJADh4Tjc8YhrR5PikDuYQmfx25c/D05Oclk1WBSCNoXOS4MHmma/k9pmj6bpulzkn5E0h+mafrfSfq/JP3Xjz72NyT9q0d//+tH/+vR+3+YAoFvcEDv6SBQlsFE/AME3OtZW8PgoIs3b97UdDrVYDDQ1tZWnLfX62ViaQaW8ILv+3kxEqea4/E4WAeeAhDiftxopJVBIZiur6/HZjB8/86dO1HMxLoFDLzRaOj+/fuZ53egBxwfH8c+oTwTd3NzM2pDKMsmLcqkZYJ5JsQNl+sCLLBCaDigk/dwAA/ncQfB3zyPBKAiTcuq2clkEqtj0zSNjYH4PJqXgwP9iLhN9SjhhNeYYJgumAN+hMmELfQL14DZetGWr8JFg3IdI/94Tb82n/W5Qj8y72DlDqowW+Y07SSZ8E6O/xhFYn9P0m8mSfI/S/qCpE89ev1Tkn49SZIbWjKOH3mzE8EEpFXmRVJ40sFgoHv37gUNpHNR+b1wi8nCsVgs9ODBg9jYhVTdwcFB0Dm8lGd9qJXgHF64xqSBmpfL5ZgYzoZceecgrcvg4x2cLifJct+Pra0tjcfjWDW6WCxLzF9++WWVy+VgLDwVD4CTVru7k8pO0zSzrJ4fMj30NzoJtBkDp06BNLKfh3sEtKXsw7Q5h9fSuNBMuFCtVnX//v2g8rVaTZ1OJ8aIBx5hKP1+P6p+GR8pmyHyal3mE/fBeOcFS7JzOBB0mdPTU125ciVACQfQ7/fVarUy16aqlwWf9InXg3CfXNcZMP0Kg4FpwyjJTHo/OgN3LcbnxUWOdwU80jT9I0l/9OjvVyX95XM+M5b0w2/nvPm4kM6p1+sajUa6ceNGpv4inw2QViIprMWNl4dbQ3fZ5o2wBVSHJdAWHkXoxWR+bs+iQLEbjYZOT08zno/BdFbl6yNY9dloNNRsNrWzs6MHDx5EARqTdG1tTTdu3AhDwYjwPrAraTmJO52OptOpdnd3Mx4Voyfj4doNJfKSov99S0GvpmT/T0nBWOg/r2nAKABbru9MjQVwxWJRDx8+zIQZGBnME+9KNSoiMMVc3W43xGvXgAiHkiSJjB3Cu4uiOBLqTWAzs9ksnpYH8BCawChOTk6C6UiK6/n8pC8f2UuEP5IyqXAcIyIoWpCXpcNq/UBQ9yUB79kKUwzR4zI8+sHBgaTVU+W8AhDx0MVWBsHrMNJ0uQjrzp07unv3bpQiu6jlKTmPtdM0+0gB9wyEOkxMDIQB4x7y3ltabYTDhJpOp+p2u9ra2grh0sGOUIU2Eyahr6RpGt/lea3UNPjDpskYNJvNCCmoot3c3Aw2JCmW5ztAc0/0P0wD7cF1IfrLwcJZmfcxY8m9SYpQzTNxPrZ+DdbD3L59O2qEAC/axuc9lb61tRVjCJP1IjCMHnbY7/cjTcx8gwWgOxFaAOwehni4wf/uZAAL+ssrUz0r5CzFATt9VD8DwHux2UWPJxo8JGW2d2PyjUYj7e/vZ0IGaVXVKK3oeb4eJF+XwMGAz+fzYAxevstvPz8T/LzqPeJ5SSGQ0l4mlYdFFO1w7el0Go9wxMgQXl0xf/XVV0MM9KrUYrEYsXytVtPVq1fDg7L+h70qNjY2Ikb3VagAJuGf35+33Xc/c8rN53x1J2IfbIDfHsYBnmwEDKCz1aDrBDAWGGmxWIy6GdoP86FiFEbjoIFoDTAChoAoY4f+xFYI9BWaEVsUSgom4owAJkP/0rfcc77Gw0GKe8qzcOYjv3FeZBxxDFTUut5CWy9yPPHgQXUjRUnlcll37959TWzom/b6hGCC48FeLxOD94TCTiaTyCRIivUN0qrmwSsJvegIg3Iv4SIu12dCktZDFSd1R3ESomGn04mHHG1vb+vo6ChWbELbKbumPygUOzg4CGEUsKBdCKmEcJ5NIn3rXtmfa0Of41GdXUjKgA+AXavVoo9pK2BNHO87u9NnvmGSOwEHcFZJe+oaw8U4KfgifKHfGo1GrB/yh3Pz/GNPfUoKYZwwgve73W5kn2ALDvqUqnO4CMr8BKi4d/rOGZr3k6TMFoj0p7RiMYTh2AT3cdHjiQYPxE46H4GQoiZAwzuDDsoXVzn7QDRzcQkj8JJ3X8RFaph2AT5OH13448Cjcm4PYWiDF+0UCoWYBOVyWRsbG0qSRPv7+/GUsCtXruj27dvhvQAQaVlVSi1DnioXCoVIT8IyqB3BkPOCIoZFWIgRYix8HmAnHHMBm+/5pAdQeKSFjzdA7syH73JuVh0TosEiWVToj6rIgxd1E6RwMWYvyALcJYWm4Lqas17YFEY/mUxiyQBzk7HKrwbnmjAIDytgdN5+zu9aBe8TSlIwKCnDZDzV7c7goscTDR6SYk0FA3Tv3r3MoPiE84wHHcT7LjxBA92AMeg8IwCUHBRc3PJCL64DNXQBzetHPFvh7eB6vtDKV41iLK+88oqeeeaZEMVgODs7O+HxXEcAhCgWGw6HsQMb6c7T09P439PK9CuLtUgj014MVFJ81xkh+gx97xoHDMvHLJ/udHCiP9ybk06mfZIi8wIgAyb8zOfzKOTivjudTgYUMUgAkBW3a2trwe7YDwZAop/of/a+ZUx5jznpZenT6TQ0KtfJAAPu2fUw+gkWw//0db5kwWtEPENz0eOJBg8Xz7a2tjQcDtXtdl/jiWAhhCsYs7TaO5PDU4ZuZHzWxSwMx7MurlZTkISx0WbYA3QTr0+GxFfKck1KsZkgkqIU3sW7g4ODeLI7WxFWq9XYUY0+wGhcDyLEYid2ZyxMLO4B4GOCwWTwnvTlfL7awJgCL87rLITr0P/eV77TlxusC8zFYlEbGxtqtVrRL5TyU+tDqIE4CMB5uhgwI1TjemRpDg8P4z4pMPMtAefzuVqtVhg2fQ1jYDy9chVQAji8PN4dF/MaRgvbcLGePiSkI7Shz2AtCOd818Vo5h1AedHjiQYPKD1gwDNEGQRPczIB3Gv7ykfQXso+mtCNi0HEwxKnemrQz+lrP3w9g7QSyzxzQ3ztbIYB5PwYEAVGlUpFJycnQfFJS/I8VIzVJyTXxOsVCgU9ePBA+/v7ERpxz/n0MmCBl6fqlfFALOS3ZxboN1KCrnt425xpAaSwMz4DvSeEyYd/CIEACwBNeTp974xysVjt5+EZDNdMJpNJ7IHB+ABU0ip1i7DNuV3EZn7ArChoY276gjl+XID2uc08zrNo15pcHOb8MKz8NahslRTM7qLHEw0e3DSGtb+/fy5wMAGoS2CS8TmvwPP1Gy5OedpQWlFATxt6Ws8HC4N0QYv3ec0nlKcqXdDt9/uxRoSQhM1qGo2Gbt++rWKxqI9+9KMqFJa7jV+7di3zWEZEQ9o9n89169at2EzZ97Nwo2ELA9qBoOhaDkIwz2Xh0ZQOzHg997jSqhLSjZlxcVbngOf34FkJ+su1kvX19czS/Xa7HazPMxBsjuRMwPfkQPsYjUY6PDwMTYUfxg1nQa2PbzBFn9J2UuIOSMw9z9Q5E0Yn8QyUA7SDDGPlgqnbEL9hebAiQr+LHk88eLBBrm+l5loDIhidQceC1FBLF/GYvExSfs4T5jzFxuF0D08NmuezKv4MXT8XZdvQR86FYbpIu7W1pa997WsxsbrdbiYD4utW8LylUkndblcPHjwIjwpNZ2HgYrHI7DTFQi/K8SUFqNBv3Bf9x3tcm8kIkHN/vrgM5uAZM8bAU+CewaAtiIKMkact+btSqWhra0tXr17V9vZ2VNACni7ewtw85J3P57HlwcHBgQ4PD8PwATeA8Oxs+fQ5hHZCDRYrAl6875oG5ehcm/mL06GNgDF95EsaCJskRdjp4ZGHjvQhrIT2XPR4osGDyTGfz3Vw8P+T9y4xkq7pndf/i8jIqszIjIzIzLqf033czZFpGRlhtwbDAiGMBoaNvWCkWU0LWfIGgcRudt6wQUIaMRtLFgY8CHGRhWRLIJBlQCwsW2oYezyesdru7tNddaoqq7LyEpH3zIhgEfl7v9/3VZ3uc7I87ezmk1KZGZfv8r7P5f/8n+d93tcNxTcUIyRI8lZnJIcmJveIhZPaKyI8nJ88uZfl25NxGP1QY9Em+pKUGBVBZuEU93///v3CI6B86+vrZZ9ZskivXr0q/7uXBNc/OzvLq1evyljh7c7OzkohU5JGHA80pqCM1/HWSQovANw1KYsn5t6NKEweowScDyNiI8F5IE0ZTwzjYDAoLQt4nec8ODgoy+/pur+2tlaaU2P0TNpakSBDkQsWoB0cHBTZssNAUW2gkhRUBoKgb4oJaK858Rgxdoyt9wD2GIGAkroze5vrc9huApZn/2tZGPejOObzeSmLdtduBoqBQHiBYQ5FTKhCMMJ5mAx1mML/SRpI5vT0tJS1mF/lAAAgAElEQVQ3t8Mdw0+YdQwHisD1EFA8MEJ1cVFvIAQBiREwTN7c3GxkDoCrKNDu7m4pCMKYGXXhVfFwDx48KM/qeg0guTcWX1paKuk+r2bl+ZNmsx3H6g5XnA3DMHtuqUUxCelanbt375bGTYQxvPf973+/8FRcF4VZW1srGTyegVCO9UVkJFxGT40MxpFCNxsDDFW7xSFoBN6EsWYcHK6trKyU0MltG1wEhqy2U9AYDlffMteEwdQvEca+z3GrjQfK6IIhvGaSQuoxkI6BETg+Z49DuTCGg+84hYonuHPnTjE67RW7jvfxYFwPqA5/gcEipDD0NN/gzlrD4TBPnz4tSrSyspKvfvWrBZW4HJ2MACljk8jdbrdkIqbTaba3t/Po0aMMh8PSz3Q2m2V/f/+tKlbuB16AzIs9GcYTAtOVkCYC+ayL98h+mAtifPCyzDsx+2w2K1WkXkYAeci6I8Y1SaltMd+Fcbh3714xrOZDyACxxJ8ao7Ozs2LgUV74mKpaZAaR3aSuRgYJ8TzmRRza2DBg4NbW1t4q+oL4nc/npUaH8BTnxwpjc3U8i9HeTY5bbTx40P39/TKYTDwT7VAD78VnzFsg+EycJ8KC67UE7h3pvDifI763YpjEMiTGEzsd6vukyAhvs7m5WRaXgQaogCQLQ3gwn8+zs7OT1dXVtxZcsWQdyO99SRBUqiSJ1WnhiAFmfOEbMNIIaK/Xa+wpy3sYWZYY2LhCbhoVue0iRo/x4ZmNXDAGzDFd5UejUVZXV3N5eVnaLfA9DC7now3B+vp6qcZ1Nsj8AQ5hb28vu7u7xYC5EI1r0EjJmTmMlUMPj6e5pKTevMkhLOuTPM6z2ayEwxw4QRIFzujwN1m0mx632ngkCyWjK5Y5EAyCOQ0TbygjYQhC6tSchcQkrIWFFCJ8hxvNvguFoAgQia7aNBGYNEvl+Z9U5KefflqEBk4DKG1EwaKsnZ2doujmHwg7tra20u/3C5E3mUzKOhYz7hjX2WzW2AneKdIkBYUxni6s4z6SNBCLn6eq6vJrFC9JY3x4Bs7H6lobRVYeX15e5uDgoKyHIYygBsZ1EyixZQRn5PUsPo8RKud5+fJlkSWeja5xTqG6dwhpd66X1D1ajXww3IypDQz1J/zvwj5eQ0a80trc33Q6bYRRNzluvfFgwZEzE073OaOCoLb5Dl5rw+gkhX134U1S5+qTeuNnJgMYyu8kDYOF8cGDAkcNXzFkVVWvjcFD0FPk6dOnjbCkqqqCCAaDQYbDYWkOxCra7e3t4r3X1tbKjnFU6WIYCMfw0A7VHHrwm/tgvQ/KCFzHYKEI5nl43RkbvpvUK5b5m/E3B9XtdvPmzZscHh6W4jAjyqurq2IQ2aNmf38/h4eHjbqVNsJ0wxzX8IDSCJPx7p77s7Oz0saBZwAN4/FtQOAcMF5GB7zvokJX03r/XAwRsubqZQweaBWZ4jzMiefopsetNx5OKzLQkElOrSJkQFsXwJDiw3ua3HNpr2Gdi5mS2jsAd9vxp8lXFMppSzI2GBliUeJVeBGUh3U029vbjdiX1GCvt1gguLy83KgxODs7K9kIOoVR2wD/sr6+XrwrfT3hHmjy63iY3wj9yspKg08CaWFgeR2y28acwxkzPC7CDbltdImyuQ1hkmxtbZV1HLPZooO+0d7u7m5evHiRp0+fZmdnpxg5jDcG3/eGsTk/P8/W1lYxMA6bqqoqcrC7u1sMLnLC/YGWkAG8vsNY5BhHB2LkdZPLdpzOnCQpjsU6geEiPE/SCD+R75sct9547O7uNoq7XCxkXoOBdmo0qXfaSppwziw5k+XaCbgJOj7NZvXKUQyF04zuUmU4z6bKSb3dAhwLQs41B4NBPv300/R6vYIgyAR4P9mlpaW8evWqPBNNe+E1Hj9+nCSFuxgMBtna2ir3BdKgWzpVq6AgnscEJV4Ko4mCONPDfGAU2+Q2BhBjgRAvLS1aGDgcREFZq9PpdEqG6vz8vGwSBXog5UgGhDnGaBq50VCJFcooLmiMeWLe7969W7gmjCabUsETgXCciYFIhgex7MJPIIc8g8sBvEgO5NauCGU/GeSQNThGeMhbUvNTcDXwajc5brXxAF66+W7S7GeKQUEAiKOZRMNM4KC9I5DXGRqEG2/D+Uza4jFQJowJwm4vhMBhrPDIS0uLPph4/Z2dnVJwdHBw0CjF73a7JaX6/e9/v5GtgEyjmxTpOFeT7u3tvUWWjcfjktZ9/fp1KRizAfRGUEtL9faUwHKeizAAxAbURqGYT8bVGRJCAiuPxwsjxtiyBST3YOVnLieTSWnnsLW1lU6nk/X19VKsd3p6mu9///t58eJFCctMYLa7tJ+dnWVzc7PUj/A616X/CnLFPTEWyMnVVb2Oic/Cc4AcQAMuNjPyvrq6KtwOIbVDWxt6jIgdmNfqvA9pequNx3xebzNpi+vMCsKGwLVjZsM6DIpfd1rMUBbvSRyKIXKqz2laiCvCEpTDUBKUAaEGgllZWcne3qKRfFVVpcfl0tJSWTJP7cPOzk5R/qqqd1m7f/9+koVA4wXxSGRgvvzlL2c0GuXw8DBPnz4taAuPy4I2zsteuXAhJgeTulahnUHg+0kaBhCYj+G8vLws5zcC8Fg5nOx0OiXDAYezsrJSmgRxDQw3dTOsbqVADuOOjLx+/boYp16vl62trVRV1Vh0R8h07969Qli6mpdrvXnzpoQ3PL9lAiId5+P1URD0Rk4YTrJh8BigIuQSg+jNwWwMcbgm7zHONz3+eTRA/is78FJAyTbysEc0GQfEA8o6VnTGwMbDRBL/G5kgnG2vY16DTYhQaJO7KAChCKSmqwIJlzCYpIqrqsrGxkaePXtW7o804HQ6bTQeptjpgw8+yKefflqINvpWsEs84cn6+noZO/ZwYeMjdnjDQ5k8bPMXKKVRGYjC7zEH8CdGfQj8yspK8eIWeupLqPwkI0RqFLTEc3quMEgOiQhx4A52dnaK0UWxbPTw6mRwvHI3SRmvN2/eZHt7u8GPYPxoXXl0dFR6hIBsaJKEbHLvODSjYxAynfBBRDYU8GoYCEI45M9Zrpsctxp5MHBGHs6wOMVl1hgo3N5SzwiD8+NBUG68qTMj5lUQ0qRedISxYEEZcBkmnnvG28Kz4MVBHRiG4XBYDAvdul+/fl1KpG2MUEwUBQHe399Pr7fYUW53d7dkHijvppDJBUUO/WjJ5yZAXv6NgHr9BkIKLLaHR8n435kq5oS5sNfEmCTNDvqEiJSE9/v90tWeMJYwhmY+jx49ymQyKfPNPc7ndX/P4+PjvHnzpqFwyCGpengMrocy0y+FNK7v27JMqO30bDuVytiau4CrMEI7Pz/P/v5+QTXIDUgFo7O8vFwK2wjRMOA3PW418kiahTRwDFhgdtjic7yHICIYCLGhNZPn1C0K0el0SsWnU6rARLwFAghioTuXWWw+47QbgnLnzp3s7e1laWmpLGNfWloq6yCsuKCB6XRathWgMzqIBaGkgnA4HOaTTz7JgwcPMpvNyu5yQN6HDx9mMBjk8ePHpb7h9PQ0L168KOPqDukoAQYMFAA688pZxtuZKZPPvA7RzGuuD2G8CSMpl8cIm29AqRmnpLlrGmRnO2OCfOG58dQYd3tvozCeneX2hBY8Y6fTKe0IHVYwXshh2ykSrjHGjE2yyMC963W+f3R0VGSdFDzjwA9FiOfn52W7iRvr5o2/+SM6MBzt/81kgygQFntKYkg8SFKvV0mamwkndRYBIcBQ8T3OxXctDO6V4VQuJCgxMopBAZLXQdy9ezfj8bjA17W1tRweHubhw4dlEyiqDDkwCJPJpHzv1atXpe8EhOzW1lZZE0JMz8rbnZ2dgk6Imbk3DGKSRnex1dXVDAaDBmnN3xgQxt/z5ZoQjy9Kyxxi1B0+EJ4Nh8Pi+U10w5PR/Nlw//T0NKPRqFFkdXJyUrZFoLzea40uLxdbYHhxHZwOqGFpaansz8IY4HT29/fL+hvCblCx09NcC6SEcQMBYWhxiq4H4ZrwaWSmQGcYbMYArsuLAG+kmzf+5o/oaBNxCCGW34Qon8eTGHE4XHEpsYvBnH5N6hb/rgswoQcngWdESPGGNnxwNihjv9/P0dFRY6UmcNRw2svBKTPv9/vZ3NwsWQE8cre7WJhF5SMpQ7w3MJrsDPutPnjwoCzoYpxAZkZQ5iaSetMrr3dhvLxyFyHnnFTLorAoEOlDcygmGZMU5AWaGwwGBR2wnoUersgI6Ono6KixWRQKy/WNcpEzQjJCGYroGFOHuuvr640wDWfy/PnzEn6CXJAl5h0kgcwyHoy3C+kYZ1dJI0M4RkI2ZIzSf8aO+Winfr/I8WMRtjCxKK6hpRdGMRl83krowSdzkDSJKYQbRTfKcEiT1BWWbcTisMpVlRi/6XTR7JZtLVFWVqcOh8NSpMW9WEgGg0GSGgURzx8eHmZlZSUvXrwoLQrPz8/z5MmTjMfjjEajzGazjEajUvHa6XQyGAxyfn6ejY2NIlx7e3sFtZFadtqbeyGDQdjVJiARejfbSdI4j4lwFBGUxhgSKoJkOCgKdANkkCSZItax0FcFkhWPS5GglZ0MCXNGWDKdLnqe3r9//y2+gvkCuRCi8FyHh4fZ3Nwsm7G3DTD8HLLKGFiu2wVuRnmgjU6nU67hcSa167oiwsob6+aNv/kjOhwKmKRj8IllgckILjuGobhASbwFA8+EEGdyPay/i5+YRGcAECDO52xOkgKRgY2gIrIq9qzsKEcFKoaL9RJ4VZSUBV3dbrcYjvl83tgg6vLyMqPRqAgyRpS9aoH4a2trZfVwr7domNzv90t5O+PEvaLUCCfPCbHouaLy1ClbE58mJamidfYlScNgkRGhDH19fb3BW1EfYb6I/W2Zm42NjQyHw7eyRsx1UjsWZ2DOzs7y7NmzYsSTeq0IpOt8Pi/bmFK8R62NVwdjMPjNgSybVLUs81lkhM+AVJzmxqg5NMeYnJycZH9//8a6eeuNhwkuDyivG2q6qMn5bQYPYXK9PwLKgJtABaG02W2z4kyOswUuSDNZRXyMV0IgILbu3r2b/f39An1Rfta+rK2tlT4MGCvIMYSA9B+IjIVxCC+LDEejURkjr71B6Ph59OhR7t+/X5QdBMLYujkzrzFfScq9eEmAESGhBYYB0tJIhTCGtDbP5dAGTgAi/eTkJLu7u2UhWrfbLUaSuR0Oh9na2irbWzDeSd0EGoOHQWNOP/nkk8KR+LmRHa/HwgmAXjg313JbAiNivkuIlaQxLoTLLvTiGbz2hedhTiCp3/e49caDQhqXfkMqOaRgUkETy8vLxUPjNSGcqJHw+UAqXJNYHYPg9SNMgkk9CKokxTMntReBWLX3cAoPJNLmTyA4TRySXiW//+rVq7K1AorLs1C5eXFxka2trXz00UcZjUaNegi3IqRgrNPplP1137x5k48//jjb29tZXV0tsB4CDmX3c4IuXLTnDJeViHlzuEk60UaYeg5X75KhYuEexpbtJTCCpDQxuFdXi6ZJZKA4F94c0hxD7dWpcFeffPJJMd4uu6dGBkRoopTG1Q5XcCDIJ7KFLJJ2dmEk98I1k7ogjXPMZrOCfEzYMp7I102PW2885vN5MQLAYa+JQKERfhQHz0RaD1LP1tyoAlRCDGg4S0lvUhsWJhzhTGqj4Xty9qaqFv1DWQrNdT35fJeJXl1dbXAONjzz+TwvX77M7u5uqSo1z5CkNKBhrczh4WH29vayt7eXo6Oj4qn29/eLAPP8hvD0C7l//36ePHlSSNak5hlAMG0hZeyT2lsmNc9gb5qk0cAHLgj0kqSkeUEwNGV2WwAUjEbIZBb29/fLnsRJ3QEf40HKnfvDUTAecAkYmu9973sNWer3+2UMIO0Jq1BYSgyM4DAQ7cygeSLkz86ScIXvgL5OT08LuiCsMplKePUTW2Ga1IVcy8uLXpoQYI5xGUQYcKMPPD/IwAx2UvcvbS88MgrB4HAfhvUoF+QmhJ/XzCQpyMce1U16MJL7+/u5uLhorK0AeXCfnPfp06dJUjiUXq+X0WjUWESHco3H48JpILAYI4wGCpQ0t6pgzCg8g5jDaJgYdvYK4+psF701bHgdznA/PhewHEM0n89L/QTPcHJyUsYTL40yT6fTRjYJcpgS9ySlyIu5b29fSgocx8W1l5eXs7u7m6WlpXz00UclfERWQFPep4caC66F7LpsgPHE4PBZDAXPzbhcXFyUzBMl7B5b7snOiu/e9LjVxuNdKIHDg53UXh/lwKqiMIQ9nNfKAw9BhgJ4zXkREhS3XXKNMSDOTOq40tkEK4yVm8KndhtAPAn3gdD2er3s7OxkY2OjrA7d29vLxcVFWUvDmg885XA4LAVlKMP5+Xl2dnYyGo2SpFEUBXu/vLzoRI7Buri4yPr6egkb3ejXZekY9aT2migK1zJCgyz1Qjc+6+pWe1aIcRzL6elp9vf3C3RfX1/Pzs5OZrNZWVRHytqEKs96cnLyVo/atbW1HB8fFyLXPAUhMWP6+vXrbGxslGItUrs8p7kctw1waEaWxTpgDslo2YeRMA7MPBRzYZnEUN/0uNXGg8Nko4W01+sVeIYxMBlnopPXeQ1hRZmBqsSuGB9ialAJXoT7wnixtB2v5wpIw1/zKHApS0uLTmHssA58Bk3Y4HS73Tx79qwozPLyclkHwvJssh8Yuqqq8vDhw1xeLrptEcocHh6WkIjaA56BJfwcrsGoqrpIytkjQ3aeH6PO7vM8tzMAXvbPnKJIEJVLS0tlW0wUAl6BOgYaFOPxMb6kkal/IcMEJwG3w058yAIZFM6J0WRtjkMkjNfZ2VkeP36cyWSS9fX1EpZhxOl7YqfExlAm2tuouk26m0BlDty5H1kmbetsn3+c+v6ix602Ho7tkrq2gdfx2HgA10Y4beriGRTCRgSBx9PxN4rrtDBoCNTA/aAwGIl+v9/gZFxN6KyOJ9bkGMSu2fTl5eXS7wPBOzg4KA1yyCBgPHZ3d0tzn8FgkPX19RwdHRUe5ejoqKyZmc1mefz4cUkdHx8fp9fr5eTkpOznQv0JBgrjwHc8TwgnSJDCKu7TNQwYb56T0AT0xnoUPg8RmKTBbXEO5on3QEfD4bB0GyPsgAuqqqoYDc8Riouxxpk4JQ2yBJnt7+/n4cOHpSDNoSDIx2S/szKMm8NZyHzQi7kvUJbvFVnFOWF8kVeu5ZDoJsetJkzb1tdQ1p8hLElSBitJIT9dIGZvaIV2yjVJ43oYBISUBUhMMkgCZIJH8k7lGC9qBXiNa1NyTtYgSeF38Or7+/sN4hCP0+v18uTJk4xGo/IMb968aRjBo6OjXF1dZTAYZHd3N9/61rdKgRkIYjKZlLDNu755DngeiqzwzAg4HtaGmXHCKzIfNiK8T5YL5beDcDrefEuSomxra2sNAno6nZZ0MoaCXienp6d5+fJlQTpJbQCYF1CgFRAS1PwE43F1dZX9/f0SVjJPzixB+uNsnMkBLeE0vGQfg4GsOdvHd0GoyaKzGPPIZ5kT68ZNj/cyHlVVDauq+u2qqv68qqp/VlXVv1ZV1WZVVb9XVdVfXP8eXX+2qqrqH1RV9ZdVVf3jqqp+7nOcP0lzsxuHCiAKLG7bo0OgMQmu6wC6G50gIElKOa8LaziH741J49ykTzkXiuFMyGAwKIrDRDptRohFh69er5fnz5+XJeAIyOHhYZJFq8IkGY/HOTw8LF6Z/UlQmMlkkt3d3YJMkjqFirevrlPBbNVAyGHOCQEkxQuy4PkwsgimC/1QbMJPDAux+PLycuEdQHcQgKAYxp37PTo6aqBGnh3CFNmghJ0ULgboxYsXhW+isxbvEc657oZiQowIiMnPdnBwkPF4XErxLT/UeFDBC8pymhVDawPBuXEKODA7PcIZeK92WII8vs+aFo73RR7/RZL/bT6f/4tJ/uUk/yzJ30vy+/P5/OMkv3/9f5L8rSQfX//8apJf/7wXaROYzmKglHjypNksiAlBOci6AAmTGk6aoGsjAwwEAoWx8cIlKwRCgbDYQFEejVe4uLgoJBvl6+PxuAiP+2BipCaTSe7cuZMHDx5kaWnRkYysA96Me1taWizpR8FRLDdXJrWLwCGQl5eXjVWqyWIhnit7gfKMDQaBClQXiHk8nS1oh26kdzk3u711u93CDyQLhV1fXy8oA3lhiwX4rsvLy+zu7jbSstvb29nY2EhVVRmPx6XoCw4ERAtvZVRJNo99hN3dDLnZ29vLmzdvGujI4RjywfzaCSZN5G0jypzauDCOjKkzjJZd5MeFiDc9bmw8qqoaJPk3kvzm9YNezOfzgyS/lOS3rj/2W0l++frvX0ryD+eL4w+TDKuqevTDroNC4s0YUKxuUu+ihQczWQcEdWMd3sfLMphMYJugIlZsF+kYProOhJgZT2llweNx3xhCvD1/o8BUP0JYTqfT0hmLZ4WMTOoWBhgbiLzxeFzg9Ww2K1WYP//zP1/4kKTZvwLB8oK5JIW/oDKSeUL5nS3wmDNWjJcNLBwP58LousOZQ08MEmNJCMYm1xhJ3qMrmIlFk9WkwzudTg4PD4vsVNWiGxzhFEqLQcHpwIswZowh2S+QlT0/82SFNgd2rWdFVjGKDqfN4VFlyzwS8nlM0ROjpJse74M8vpLkdZL/uqqqf1RV1X9ZVVU/yYP5fP7i+sZeJLl//fknSZ7q+8+uX2scVVX9alVV36yq6psMMN7ZKS2HNF48hYIzKPzPgBn+4ln8HSaXCeW7Sb0s2oqUpIRBjmkxIF5Ah1C048+tra1G+TcCScjBOgi2DqQSlfUJCD2Gx2EIGQDO7XT06elp/viP/zj379/PV7/61RLmJGl05uI7VG1S4k7oZaKPZyNMsNDicT2ONijj8fgtBWBe2WqCe6RwDGU2ugPhmT8xj8FcIwc8A2Ga08XwKxj/+XUGDUfG9UB7GxsbxbBQbTsej9Pp1D1Uec+peMbDSJbMDGgHA814OaS2kUH+eA00Z1IfQ+a6nC96vI/xWEryc0l+fT6f/ytJjlOHKO863lWN8pbZm8/nvzGfz78+n8+/jvIkdQyKIKKIDjeITzE6kEsIszkM4DUNZjE2GA7DSL7PBDhMwpInNUHnbAy8i4kuCpSqqio9OZn4JKWoiXUsxNRsk4Cxwtt53QeCTN9NipEIf1ZXV7O9vZ3hcJj79+9na2sru7u7ef78eX7u536uNAgCIVHc5PUQd+/eLYv1EEKnD90Pg2ub4+HzkIYYTtAPfUE5nLHZ2toq33WvCqeyyZJhfBhDZz2YF8hJlLrbXTSa9mpTG2dkxiEGqDZZKDLZLBwUKXLO41W6NnTmYQgBHZZgPFySkNSFjNYJvoPxc0Ed48Oz3/R4H+PxLMmz+Xz+R9f//3YWxmSHcOT69yt9/kN9/4Mkz3/QBRgox8SGr8AyJg6lhrTDQzjetrcAwlq5IerwRAgJyoNQGv3Y0jvzw7lMuM1mswJlgemG+YQn1Cd873vfayAuFAlyFmG5vLwsqUYE9ejoqDRBRjCB1qurq6XBEJ776dOnWV1dLStpHcJhuDF+GFEXy0F0OpXqbAuZHdAYZDGFSu4zASJgXnhu+A8cCw2JGGeQAzxCO3VrXgRkxjWcmsdYtb21kSO/vRaJ+/Z2CFZ6IyNkys9bVXXrS3gpyz1ELedAfvgMP659smPifcLE9ylPv7HxmM/nL5M8rarqp69f+sUk/zTJ7yb5xvVr30jyO9d//26Sv1stjl9Ickh481mHeQcEGKV2iJE0i6/MTl/fa8M7kf5MaoOCQbq8vCx8AOdHiVBSFzlxXjwrSuaFdDYqTnd2u4vGOOPxuNGwBQHypOMV8XwIkNdDsH5jZWWlrJ9wRymMFcf5+XkhBpMUovDly5cFuqNUcCgmg11kR2oRRQQB+HWUAs7B8+LiKO6H8XJHN69FsefmmldXV9nY2MjGxkYjM4ITYtzY9OrycrH1prkLDDHZF+b/6mrRXIfzMjcOE5E3CHCK8JBLVrwynsiinRKvc7/8xlDZoDIGyLIRM++DgjCg6Ey7FcAXPd63SOw/SvLfVVW1nOQ7Sf6DLAzS/1RV1a8k+X6Sv3392f81yb+X5C+TnFx/9gceFi7Kg2Hfk5TYvU0ewbLb6iN8bc/hEnYrC4OPx+bzrMi1l7SFJ0Y3jMTDAnfxDEkaBgDEBHJIUuoW2kJjwRoMBmWhWzvEWFpaKq32hsNh1tbW8t3vfjfD4bDA99PT0xwfH+dnfuZnGkQj101SPDRVk4wDz2ZDCteDESCcc1tAfsOpYORMFIJsWC/iOSZNilLyneXl5bLL4OvXrwsSoUGPkRRhE7JGH1nGmjS1+4sw/qyzAu0lyWAwaKyMxdmwink8Hufy8rKQs4RsjBfFcowbDoexMD9hY+I1QaAnZMRdy6bTaZE1xgADeJPjvYzHfD7/4yRff8dbv/iOz86T/Idf5Pw8JINAnGfP5PQpHtITmDQrHm0UECRbbO8fQqxua+7MhnkQPAATh6dFwYHxKJGJVyaSXe4xOAgqYQXe1RsUnZ2dZXd3N0lKXQgFU3AsEJ4oJ+s8Tk5OGmiCe+C6EJPE9HTfhmewUWUZPSlCngX0wDOanAaW+1ygB+bp4uKiNGa2A8CwseSdzJLRKsgL1OCUK7KE0cIx0HMU4pj1Q84Wce88M/zVyclJwxm5YpXnwqjaGST15k/wOc56MabOZDFeGDZky6QwsndxcVFCcZ7DvN5Nj1tdnu7wA6/m1YL8dkdwJsakkAXX0JC/uQYCSDrMJB+CQwyM5UdxiK9NSHFuFBeBwyhNp9NSGk4jHNJ4hGcsSbc3Zz1Er9crtQt42e3t7eIRR6NRjo+Ps7W1laurq+zt7WUymeT4+Lgodre7aJJz7969Ur2K4CPQoIukFlAMKEYCL80+IjayxOkWVI8jBB8FdiiAa3NQQAy4F8+7B6kAACAASURBVEHCC2xubpYVq3SP//M///PCj2BsMeAuTmP+6Vfi0BPFa1fEep+V0WiUs7Ozcm2jEDItGHEcHaHyxcVFCS1MvIIKnO1zKG0ew2g6qeubGDPXejiU/old28JgmoDEs+NNkpootVKbeSdUMDFpwSD2o0iIa9k6IzQYIOC5vSaIAqEEgbD0G+/GufCE8Af05MDr9fv9Rgv9Xq9XyNaNjY3Se/Tk5CTb29vl8wj2dLpYTUtowv1YYRDa9fX1ci3uEQ/qmNw7+BGaeTtQUCHK6dCHZyZ7xBj7s8ydl+Vj7OEfeEbkw+EURPLR0VHpNE94ggeGG2IO4Z5OTk6yvLzYtAkHgtPhvlDeXq9XSNrDw8NMJpOMRqMyJyA7ZNB9UuBOXJtDSOECMkINIx7XgODsGC9CGIwGr7eNNOflujc9bvXaFjycPRbChDWl8AmFbBsbvBaD7oIeJp9ww8VlzhIY5aBIDp24L1t+DjPuGB04FfgMe4tkYTQePHiQzc3NwvIndU1Lp9MpRV9LS4sNq7e2trK2ttaIa3u9Xvb39/Py5cucnZ1la2srKysrZYUtxByelQV2SYpXNWqyAQUVmGRMUshAG1yMvr0f88scYLjxjFYYFJZsDuiTz1MIR1Fdr7foDj+dTvPBBx8kWazBWVtby+bmZkajUTG2DjUxmktLSyVUQfExYBgnL3knpKavx2w2KxXCzDvPwVij4IwDm2a3QzyjZL6X1A7Tsk7BoM/NvLXDcWd8bnrcauORpBGjMbFJnX1gYBFQewtXRppU5VxMDkqQNEuCk5rXwCNY0Q0hPVl4YASciXY2yMaHZ6AEG89ECIOiEg97IReef319vRgkcxooBUQiiooRcscpl1ETgsHYsziPGhY+j+duf9Ye2ylfp18hZkFmjBEH3t5FYEnd1xZlM6noEI8tHan9IcTg+UejUR4/fpyNjY2yRGBjY6NU9V5eXhauyul/Whi4UplnYgUw80oYhLwmi3oWQkfzehgiUAOH5RpZ8HWT2qkSKvF7Pp+/RdQjs4zvTY9bbTys1AgqVtnEGZ+jqMnhgBWcOgfOYWV2WtWCzvXx0jYyDl8wGLDoTKaNng2ZYa3ZbysPS8BRSNAByuS4eWlpqWzwxPWn02nZ/IhzkapDCFnXAlJxxoT3WKwG3+KQ0GQyYwv8tuIzZjgDPDCkqPd9YTxZZ0MoBfmLR/cevYzdbDZrfO/w8LAYO3vl+Xyxfy/74vCMdFy7f/9+2a4CI2VUxfld+QyhfHBwUEISQi0r9tXVVeFEMAqESXZGDi0sS0ZyLl8wz8F7SRqhOAey6lD+ix632ngw4Ay6CUwz2FQIYnEdMzpGBNYxuEyqiSnXabj82ecCjST1att2Jqe9FqeNkOAO2ilYKzJeg93dQQOMy/n5edbX17O2tla4BOAvTXiPj48L4oAMZun62tpaPvjgg6yvr2dzc7OMkWtUDKdtTHkfLseIjFDQBWVJM73KmDgcBCEg2MwlHBDX4XvT6bSgS7wrXAYhGnwGy/B9DbcUJCzq9Xp58eJFptNFh3nvk2NZdL0E52D9DDyIye53heDLy8s5Pj5uFNoxt5w3qRGYt9MgAwaKdT0I98b/vNZGxOaobnLcauPRRgQoT5td9sA7FdjmC5J6ZaghstOGGA0X/TBRbaMBGnLsa1KN0MhhEud1nwpzL4QF9ItAWdgkyrvM4ZF6vV4mk0mJzQ1nQSx4SJO1m5ub2dzcbNwju655payJVj6LMeEejUAwkIw98N6I0f9zDp6FcXFGhTG3cnlbxtms3syJOplHjx4VwnprayvD4bAQxzwb6ORdXAtzNBwOi5KBrrgm30vqXegx3izJNy/keUD2bJTMFb2L9zH/xg/zw3VYy4Ssmo/j/Mz3T2y2JUlJ+RnK8fD2hhiLdqjCYaVCaICLdBHnsFBYaJyyNHIBfQBFUT7uA9RhpaGbFPdsw3V0dFSWy2MwCHMggZOFJ4f/IKy4uroqCpKkxOxGUU+ePEm3283R0VEmk0lRKioiMTagCsaD8Mc7w5l7ARlChBpaOySiwMxZG8bYBW5eCe35QWGQASM/0ActFtkMC+POvTnjQPtE6kJcGAef8/jx48zn8zI3Dm2NZB2CUL3K/xhZDB7Gij10XZOC0abtoZ0eSITrwS8lTYeLQ7JxwXkxhj/RhCnt8MwdMGC2+hgKBJZQoJ39YFLxjCwOwkAZKtrSEyY47jTbvbq6Wkg1k6dtJeNvdle38hh5OKszmUzy4sWLHB8f5/j4uBCXFxcXZZk55zfRmtTdzxg3trokrKDIjD10WcFLBotSbcYDMjapVyrzDIwlhsKGHcWxxzRP4T4V7XS353k+nxcj6vVLzBVFU7PZYg3RxsZGQVcY5t3d3WJ87JAgijGss9ksh4eHOTo6ysHBQTEULGHAeMERsacMTgZDBt9kJJmkQbh6kyjG9uLiotHyMal7dpjvgz+ygWXunZ3xvsAOfW563GrjYeuIwLkeAG4CYcGjJXUdByENioCQu1rVRBXXTJqrKSFjzaW46hM4zcQ67myjGpZpm3zjungajByl1vfu3StGAThNfcRkMsnz58/z4sWL4r1QYBaO8dsw9/DwsCj1yspKNjY2Mh6PS2MhDBjEqMMRh2wQoGQ5kjQMPPwCY4eCYxggVZkv11XgOJJaoQjzbFBALQ4X4UI+/PDDspz//v37mUwmefbsWQlFMSyXl5cFMVE3wnNjtK6urvLkyZOsr6+XjmRJGtyVM4LMI2NNuMIzIodeeMk42GjyeY75vF7ngtwmNXqDL0NveBbQWZJGL9WbHLfeeDAh3u80ebuBcVVVpdLRsSOrKvHqCKIRig2Ef6MoNirAXngOBNcGhPu0t2iHQhgA7tHehQIkajmSep0JXoM6Bkg+Fn+hyL1er3TJ4n7H43GOj49zfn5ejBJFaVdXVzk8PGyMB0jK/wO1HSowxmRS8KAgh6WlpUb3Nd7391GWdjiIgQUdMNbmoNrL/eFLQIiHh4d5/Phxer3FHrzb29vpdrvZ2dnJZDIp42veKqk5Fu4TZT05OSld3OAySHmDQJEVh7W0RHDogtzYWM9ms4JWkDcOnhlZYXwZE8ubyWyHNxgaZOamx602HigmjVyTGuq5n4G9F5YYAopJt0dkcBE2Xjcn4c86ljeKwXPCMRgdJXU3KSYaxUKw+B6IgzL79fX1UqDlz7SzMevr66X5zNbWVjY3NzMYDEpnMK7d6y0aIB8fH5fmQMTmVNg6w8J+L0598x6vt8cQo2uEYq/pCkeMAvOJ4lipGD++byPhxXU2ylZA8wLJor/r48ePc3V1lfv37xfjTyjInJjPuHv3bml5CDokhO12u9nc3Mza2loJV8hU0aYR7094dH5+XhoewbEgVyBslgUk9VJ75MZjntQ8HuEmOkO9ENk33zv8iY30TY9bbTySer9NUIhjYKct+SxxHoOVNNlmBg6rb6LJsNzElZdH8x6GCm/Kfb3LOLnQjPtkIkECeNVOZ9GH9ODgoAh1WxkuLy/Le3T3gii9uFi0/j88PCwVmWxviaEARVE49vz589IUiQpNenkihIbUjJWbFZsg9opZXoeE9BYN83ndlYuxBR0ylgg/aIr5dDaMNC1e2DyEw9Dz8/OMRqNibJM0HANZj36/XxBZp9PJ/fv3S0iM8T47O8ve3l5BG6PRqMy3kWz7Hmg12e12SzIAp8XyAxylUbSLFDHAOCmPhTOPODYWW8KHcT1zYzc5brXxQOicXnIIQOrWm/X6M/wkdQzOeVF4PovQeSk5h9OzxLNGIa7MxLAktcd1wZS9qpeAr62tFcP18uXL4glRLCacUGI+n5eqSYzI0tJi+X17LQNdtwaDQVkxyrJw1tPwOaMDuA4rJO8lKSiG+TFpZ54C4hAOAvSFwYAHwSC5/oNnZf7wnIRNnt82X8UcswqZ5kgghgcPHhRUA0KAQwLi01VtNpuVDaKYA9AG8/Tw4cNGZzeQM4qMo9rf3y/ziuNATpwp5BmdOTNn4rDH3eTQG65Jf1zkx1wU8nyT41YbD7wJcTFCigKipCbakrqxMCXE5iIgTCHjQCGu3LSnJ0RCQFw7wkQkdUt7BJjCJmA0E9kOsVBQagAQKi+G86rUjY2Ncj6MplcOV1VVYHSSRihBzQcCS31Dm5iElGRsMST2qIbThEDuJMb1MTYOQTgf6Id7Zv7alavcE/dDpSfhlJXTLQy9jJ7zTCaTUkG6tbXV2Nfk6OioOBN6xpJuJdsGUnRmbDpdFODt7e01EA/Pa9ISx3VwcNDo4ma+w8Q7RtbIK2k2v3LaGtRB+hlZxcEYHSb5yUYeKCdNZH0gUFa2pJlC5HMmkbzKEYVnubQNFhPJhJjFT2qj5QMPjWVvZ1t8Lq5hY8RyexQUpe31emXTIjwM6T0gMwfZkqReN4JRI+1Kn83V1dUS1zsObhdnMY48m58TxESYR+0C5CYkrolIp0lRFHvedusFDH+b1PZiOiNBjAsKR6k9yGJ9fT2j0aiU7zMWZFzYzsHpUCMNrs1zOKz0GhgjK54RfsMrawm3HX5Y8a0TNoYeU+QLueWZGEfGiMZOzNtNj1tvPPjt3cxM8jBhIA8mEk9r69w+t1EMHg1htdcDHvI5hxBsEmThIXY2KWmi0GSi4/PxeFyQjslY17dcXi46UQ2HwxIvI5AIBbUBoAvIYxOgXmjG9VEQt+hj+bwhMrE/np/Qxdkwtn70fPEcGESuyzOYyyJU4ocxxEgZzTCGHCgEY4+CQ1CyKI2M1ObmZqOJD+X83W63VK0S/tD8h+fm+lyvqqoSItAVDONj3q3b7WZ/f78RXrkNgfkzowujbpPGjI35PcaWefFOhXTDhw+7yXGrjUfSZNop0baHM0OP5zDZ2RaspLkU3ClAF/04pesUoku2icGZZJOwKCzFVjyLiSo8KdWl8/k8+/v7JZwiXFlbWyvcCAgJQ+NnQpiSuqLUMTS7p0E4YnxAaCgbBCljwG8vm3cGC0/r7AYC6x4SCDw/jDdj5j1JUICVlZVGhSbKa0TBylU/N9dk3JeWlsq6HozinTt3Ss+PtbW1guROTk7y5s2bssUE92Y04esS9hkp21iDpJK6VSKfbW8Jad7NssazGXEgV6BRDLUzgoSwPr97leAcbnL8WBgPvPHu7m5hjl1/YcaerEbbM6EMjj2BdXg5rmfLz+uuKkVYnJJk0hBAFl2ZS2DiaDrDvRMDf+lLXypGDF7CUJasAuENxmIymZSMBEYVDsCEIX03qdNwxaxXdXKv8B5cHzKWzzmM8GIujJ89oD9ng8FYmwinAA1E4noFYL4RnN/jngkDCDWNqmigRAl7v9/PaDQqzZzMLWAADg4OymZQyAPL7zEcyCArf0G2GxsbBTE5jGT+OAfKzL1zLZf2E1q3F88xT+b+OL85GLfGvHPnTsbj8Y1189Ybj6TeSbzT6eSTTz5prGwEpjklhoIjkEBCiFcOvP9nxZftakmsNffknhdJnXrkHjgHjDeGwSQjHczdxwFDwPnc3hClXVtbK3E0ggfRifeCIAOSG+FQBzKdTkumx0bHRpYVwEm9r81sVtdEMD5OLbpTO3OAUeDZeA/lYR5RGowhnh+C2N3XMaJOndOsxwYVI8oKbDZUwqBubGyU7SfpPUr6m61B3ZMVDs6hgzNPGEVklfVA3lMIA0t2x+lwVkB7zB3mIYPIk7NgIFvml/dNqmN8CLtvctx64+FKPSYDD+A4j/UgRh0YjKQuuLFhAVajAPzGkNj48D0OpySdGXCWIUlDyPGkoBgOlmU/f/68rKU4Pj4uBoJwA/htg4VAzufzkkrEmPDcGC2HMC59hwMhg2P+A6Uwp+Rsho0q3pVx91J6DChIDM+L8jsjZJ7EBqldZMYiPhsYZ2ogaVlW4II9zr23t5eLi0V3dQyIERMciTNz83m9KblDX2QFIwa/Q/i5urpaCHgyTsmCiCYMxWFQletamXeFf4yVSVGQJYiLLJ6TAZ1OpxFS3+S49cbDdQ4c3/nOdwopxeKhpLn3q8lJQ7l31Vv4PYcl/G1YndRK4/4MTKLDG4cWwGfSuJwTHsdKeH5+XnqPYtAODw/LOciYVFVdNzEcDgu6QHk6ncWOcaPRqNyTMw6EbUlKjw+UDE6Ac/I6B4rKWDCubjbMXNhT8hu+CCfglCTj6WwLhhGZ4DNwQXwG5fJ1knrX+MFgUMYQw4JxgEAdDoeN5f7Ly8s5PDwsoS8ZC8hnHJzRGbIJiY0xYikAhoDx8WI/jDCIDBm9uroqPWdBW8ijkQdyDTo3IY6sIbv+zhc9br3x8JJhCKOqqvInf/In2d3dLZ4VBU3qUGE+n5eKPl6H/4CUgk/hx4gE5bendJoM5cDSG4o7BUubQKA+ApGkCPCTJ09KWm17e7vEqmypSMhhWGuOAzicpBQqQbTijYyCYNtBKtwTpKhDJ38fwcWYglAwnJ43FJbnBCVZYCGEIZkdFiZ1yEjX8aTZYY654rwYJAwn3htj1ust2g+ysz3L9hkXslntMnEXrrVDknZ2iPJwEA1yRgjB9fg+4027AxsDyxvzjbF2ONMOZTDKzKvLBjC23h/oJsetNx72HEYTnc6iCTALiBBgx4VAZxNnDj0wJHzWJF7ydoGX40aHHlh3rwFxkRjrVsyHYAQc3gDr8SpecAehSeEbY2OBMrJI6rALEpBCJgqUHj58WGA+iAhlI7PA8/LMzAOCzd8mh52GtRGGOHWFLPCZuUX4gf4gN4wYUBsCmLQjBoaxASXYWPseIEoJEUhZIgebm5slhGMMvfjPxYQ2ClW1KCuw7LkQjvluo1naK4D44OkwwvAYjAcGAqNmmeYwxweahMe6vLzM/v7+W3VKX0g3b/zNH8FhT45nQzHm83mePXuWs7OzsikQE4lAYtmT2iKjXBgThJ9rIXyci8+6GMilwCgfE4V1529DdeJ2DMPV1WJP09PT07x8+bLEu6ABsgcYDyoSO51FA5tksSlRUq/fIWwgnUk4gfFaXl4uu8WRZUHx8f4YEyMzGuM4lWh+AuTm1HfSXHCIh8Uw2NCbS2n38+SZWW+DTCQpnA1jjXEj1ncYxHs+HwhtZWUlx8fHWV1dLeX8y8vLjc20IYin02kZf9deMDbwRBSDWWYxGmRdHLItLS26treRIj1U2mQs88T5MSKMqUldp8Lhlg4PD99LP2+18UAYTBbZsnY6nbLpEcrsFK4NApOEoXBOnGuZYHXqzd4B78ck0JEblMJEuTAMD8b5EGx39Mbzg6KGw2HZcwSvj+cgvk9SUtcIPegABNPtdkt4x3aTNOXF8OAR8dTcD9kIQkIvuUdZMTJtFNX2jjbA3CNID0OKYlpBnG43r+BwEUVyivf4+LiML7IDwcp6IIwjxpitMOhb+ujRo2LgyXrx3CxCdI0GBpT7hY9yKMz9XV1dlWI85AQEw7hheJzRurq6aqz85TXO67DPITvz6+Iyh7o3OW618UiaO8gnaQgUAshgI+QMJoPkVKKzIY6xGXgm0kJgqM73OSdeyQU8bf4EIQVeQ2Rxnnv37hWPikfnflAML3irqqqUoDtTRJqS1Cr9PEEsEIPAdYhDPDX7n7QVFISAwXLakM8RakGkWhlYPcr4Et5xTV+H53DRFQYLeXBRGWPe3l+XzFGbuEV2IHup7UgWe81CKhJysBcOr9vBMC/MrdPYXIdMFSEVIRey0e/3yzkxvu5Lg6MyYU04yNiD4hwSM08+L0iIqlkM002PW288khqiGv7x9+rqaqN4B9ILwUXokrqGw6kyBpuBtjey8lJngPA6PYbwOR16586dYtlBLybugI4IiglFM/O9Xq+gE8IZE2BUSZJqpFR9Mpnk1atXqapFkRJd1lkw1ev1ClqZTCalWAiobC9GVsH/44V5btcfJCkrhclWWJCdYXEKES6BOWfcCJ2S5r44FOQ5G+eMGWEqc3VwcFBqZajdwDAQrq2srJSFc73eonkQ94DHp/5iNptlf3+/8BVGtDgu+AtnjlBy7o97NXoxH+d1W4yLn5kxJPRxmto6g0PlPpnnmx632ngQsmDJDVt5v9fr5c2bN5lOp2/xHhgZBp8FYG0C1N4ET57UKTeU2gQi5BkC4LAKEgoSD4Tjwh2U02EGiMnrTqbTaTY3N7O6upqNjY0Sg3uDZ5SFPiDPnj0rGzbDY9AyD3QA7/HixYsGQiKubmdVGBu4HWd/TILym/lg3jDUfm7zUYb1GM/2+64vQQHw9sxFUvcM4TooEv1TzOk4e8IqY14bjUaFBCUrNBgM8ubNmxwdHRVyk9W3XjRHpgz0QC0PcsFcXF5eFidh4t+1NSQFXP/CuJhzMRmLLL1586bMIY6On/c9fiyMhxXd8SGDxNoBD6RRhBdVOYXFYDNpDmEIa1yGbWuNYlnYXRUKgkDo4SnoEYqSfvzxx+V+QRv2KpSRo0gQqnjzyWSSs7OzHB0dlVWa9KtAQaghQPFBZ+6EjlJ79znuqb0cH+QBXJ7NFh257F1JQUMOotg8t5fOo0ied+/f4tQkhghFgnw2t+SQkXsi5GPcGF9IXKMJUGFVLZa2f/nLX06SMtbwH67ZMDdGkZ+R1Ww2y97eXpEd5oBQxtnCJMXIgVQYH+6d/xl/83yMGbwLYRv3Cpf1Pqgj+TEwHuYcGDj+R0BQcuK4NvTmHIZsXl/gayR1DE/BFV6N6xlZWCncJ4Rz8FmXZid1CEUcilInzcVda2trJTaHxKQJMp/rdDqlzNjl2s6OcE+np6c5PDxsrFPBONrrwcGgeCgon3GBkjkLt8Pj+y79N1fC+RyeOM3rWJ3xNyLic7zGnHievdMd77ty1gjGJeSMFwgNcpxx6nQ6xVgTzkAMw/NA2hK+Ud/hIjeMM0bs+Pi4QR4nKXyJwxLzLv6N8TECpnrV6MOG5qbHrTYebTjG320CKUkJD+AbvFQZr+9yahuUpAmFndNP0rDqkF+GvhgRQ+g2N0K2hNQiqVOvtUEJvX1AksL2Q9CCqhBUsidAbD5LSHN2dpanT5/m9evXefr0aeFJCIMolELwTCY6NEIRbYgJmVBo7ptQzQQ1BWjmN5x+x4iAEFwJjOGw4rXRSJLiUanlIF3r+0CRMaw4CvdugdMZj8fp9/u5d+9eBoNBQY1ehc35bfTIYCEzLIj0lhbIrccQ+ex2u2W/nocPH5b3vJzCsse4mfvBebH0n+9RV+MQ6CbHrTYe7VClDc0YdLqBE4o4/cRk2BCYOGLQjUCc+kOJeM+ew58HinIumH6MDAdKX1VV2QoB4cLbuZiH76CQXk5+9+7dImCj0ajULfjZQDyXl5cZj8cFwXQ6ndJiz6FZu1LWYZcJSs7rCsqkVmoU17UH/Hj5volp5pT5MtIjLEFBMNRGeA5RCX2oxyAEoaM8hhhFtrHl+ycnJwVtrK6uZnt7u4RihBztdLJDCojYvb29BmkOX+IMIDLhDBH9QnZ3d0udDQSwQ3kMCOOHvFj+2fyL2iejj5se72U8qqr6T6qq+rOqqv5JVVX/fVVVd6uq+qmqqv6oqqq/qKrqf6yqavn6s3eu///L6/c/+hznb4QFRh0OCZJkb2+vWFSjAAbVlrjdHcxW2GlKBtiGJanjzPZkAuc5J9dxv4vl5UUTYkgyJpeJZHk98LTT6RQk4XTnfD4vcfHKykrW19ezvr5eBL1tACAEu91unjx5kjt37pQsFedMUoSXlakIOGGQPRjjAOzGW6OEGHM3+kFZMfzOFLjcnnFl3EFphGBJsys7aM9pcUIQrkGNh1f82hH53jECyAvFd6Rz3aHLyg+ysIFCJthCFITcnmdkBVlcWlpqtG/A8dk4EM4wxzgs5BanRvNjnDAVtX8txqOqqidJ/uMkX5/P5/9Skm6Sv5PkP0vy9+fz+cdJ9pP8yvVXfiXJ/nw+/xeS/P3rz/3Qw4LQzqG7SQycwbvy3cBT8x98r81zOA0K+5/U/VKTNLyxm7ygHMBL7gUIOpvNivCCWBBSnsF9SVl/kqRUH47H4wbc3N7ezmAweKvpTL/fL2EM4c3W1lY++OCDkvnAk5sfYdzMLXBe7hmB595BLiAvuB/Kq60Y5jUYJ+boszJqGHKPg+s+eA5I2qTe0AhlAn0RJrnOATRFdghD79CMDukrKytlzcqTJ09KXxDkwYbJWzrwG0NYVVX29/fLOBBeul8tryUpa5nMG7lTmksAzP8wThQKskaIfq1/nZzHUpKVqqqWkqwmeZHk30ry29fv/1aSX77++5eu/8/1+79YfY6gy5Dd4QWcAmShY8B2iON43BkZBpf3DGWZPGJMvofBcJ6e+8HAcC4ThSgwaMmehO5YEKGG9N1uN+PxOCcnJxkMBiVV2+v1srm5WfZY4ZkJS8bjcaPUvNPpZDQaNe6Vfh8gMxBSOyXqFC/3Rb2JY3Q8Jc/Z7XZLDQrXNdk6m9XdzfghLCCMcgYHo++MFNwFY3p8fNxAD7zneab+gjBvNps1nI+LCeE0Dg4OMhwOy5qXJKXC1J/nnkGl7apTFz26UNC8R9JcVczhMBjEgQyBAN+FypFdCGIMiMP7mxw3/vZ8Pv80yX+e5PtZGI3DJP9PkoP5fM4IPUvy5PrvJ0meXn/36vrzW+3zVlX1q1VVfbOqqm86/sart8nO4+Pj7O3tlUwEnpLB9wpWDmcEHEOTPuM3k8XnzA9QWWhPyXlASXgVQ1fStZubm+X+qqoqvTXwkAg3r11dXTXSf1Z6MiPX41e8PGTh6elpaVvgmpP19fWCHrg/E51OiTqmxnsb+TFe/X6/kS517YMF2+exU/B3MQ48M3MAWqLWhB+exYsFvd7GYQLoi/J70A2hI/fT7/eLsl1cXJTqXHgnZGsymRSlxkCZ58ExIDNVVWVtbS0XFxfZ2Ngon+MeKJVP6obWvOfCRzvCtgw4jOl2u3nx4kWDL3rf433CllEWaOKnkjxO0k/yt97xUXDRu1DGBWTFuQAAIABJREFUW5hpPp//xnw+//p8Pv866TWHGYZ/ZBnYJhHBQ3gIExAMJg1GnEn2d+EhuDaZFwYbxdrc3CwCjUKg5OZaCEfwrAgr0DhJyb64UTH3wXmJt1Gg5eXlbGxsFEVi2T7KCFp7+fJlhsNhkrrqk3PcvXs3g8GgpIsxWBTEYSjb49ROB2LMeG6jO6e4UWTIbdd/ODzh2ZO8tSoX50BKs6qqQjZTW2MDCDdhfqnTqff8gcsgjW4eBoXlXtnx/smTJw20wthy/U6n3gcGY0M4g9Fifc3V1VX29/cbNSomYafTaUnDg2A4lz9n1MF4u0wgSZEx16a8T9hy8x1fkn87yXfn8/nrJKmq6n9O8q8nGVZVtXSNLj5I8vz688+SfJjk2XWYs5Fk7wddACvpeJgfjAcrAxFeeAOE0rExh4t3vIbCTLtTfYbyDDppV8enKDIe1XuRcA16aLBoiVQrcTBCjJcBngLP+/1+WWVpHgJFJq4FdXDvbNuAt+V81XUWAiHFsJL6RThdJ2ESFoPgloROqdLFykbCCI57tjdNmi0imXPXqzgkQWFQoLZ3xhAyT/AO5hE4L3JC5SdzjSOaz+e5d+9e7t69WxoCuXqWxXDIoJFNktJYyW0CIC/X19fT7/eL8ev3+8UgOaTieTiXuTvGljEBzTD2n376aZkPEPZNj/cJer6f5Beqqlq95i5+Mck/TfJ/Jvn3rz/zjSS/c/33717/n+v3/4/5DzF7FhD+Nqs+nU6zv7/fSNuhPE4/MkkIui12+2+nW1FcCDYbM3so0rluxc+KWDPrCNijR48KX3JyclLOR5gA8YUhYc9U7/QGpDdngNCReRmNRhkOh4XrYIxY34JHN9/gBjdu1e8YnPFC6TEsKAzzAF+AsUcBnGYETTA3zLOX7xv1mPCGewA5YGja6WNQQlJXlJ6cnBRCGRKU8zAWPLv5MOTvK1/5Srlvnt3E7YcfflgWGoIoOA9oh/uaTqclE4aRce3L/DpjRdjYJkUZD358OGzxnJp7uulxY+Qxn8//qKqq307y/ya5SvKPkvxGkv8lyf9QVdV/ev3ab15/5TeT/LdVVf1lFojj73yOaxT4bYNha4ryM6BMLgODYphIbYdAXl/CYaLOjLg33CbmBwW0JxTv4MKxJCVnf3p6WmJn6hE4t0umMRQoIPE0ngiBQFGd0cFAuA7CRoprMB6UtcNLODuAh3bWxOPQLgvn+4RzFHB5kV07K8Z5k3oFbbvzlo22V7tihDEibUTikIxzYCAsH5YvjD5zwtYYW1tb6fV6je0i79y5k+3t7STJy5cv0+v1Gt3CHH63a0qm02meP3+ejY2N8kycl7HzymGjOT8fsuWsDOdPFsbTOxD8dYUtmc/nv5bk11ovfyfJ33jHZ8+S/O0vcn4GGePAw3qNCh4c2I1XISvBGgSn5DhMzjmU4TtJLTxeC8M5EHLiYu6NCXNohJGyATOPAmqoqqqxpSRKiDLgIbnW5eVl1tfXc3R0VJa3cx6UHuODl7TxYEwI9/BUhEgYCHM2fAfojrHEgCVpPKOzJxgZxg4Fs8K7tseKzP8YbBTFdSUYaMr6HZIahToVi6zxOS8hgK/C+C8vL+fNmzcFXb548aK0DpxOp1lfXy+rqQmd7MA4L9fCgGPIVldXC4Il9MSJWq56vV4Ju5AdoyTuGTTJ8xI6E2K9j/G41RWmbdItqZdcM1D8vby8nJ/92Z9thDpGAAwSwgUJioI6LQjJZviLwPJdDBpxLf8jyDDs9phVVZXWf1VVlXoFUE+/389wOCyhE9cCLaDYjAVCdHh4WAQU4YKkYwXudDotMJ2xAQE4I8R1MVhJvdUBKWUvNDTXQTMiDKQL4xxaOJ3Jd5lvk7LOINiQ8lmcS5KSMfIzmJfxfICEeG7m1ytOmXeey8v3r66uMh6P8/DhwyTN/i37+/s5OTlJt7vYJBslZR55RorXMGiQvjs7O8V48Qzr6+slbAHlGi2Dys3LubTAi+icfbShvMlxq41H0uyeTmm2PQZw7Wtf+1r+4A/+oEBup1iBsxwouWNuBAg4nKSQVa6UxDgQ+19eXjb2+aDWwugE2G3iD3ISpSRz4tgVI4AXd7YIbidJgczuWEb6jxW4hCcs9GJM4CG8ZsbEnLNb3LszScB5wkunJxFuerPaSGHEjMRsINooDycBeejQCniPAUZuCOPY2Itn47N8rp3O98I45hVUx/Oenp7m4OCgEJ4g1PX19YYyQ1S7TQDGjLALFOgMDMaRqunZbFYyOBhj81DIDA4TA0x6mDG1HqA/Nz1uvfFwhV3ydhclBmU4HDbIUggqQ2AG2WjCE2UiFC8Gz8Kg41UcL+M9ied5n8kyyeoQZDgclkY7VgxCHUNQjBjXxGNhrBAqtihAQVAifuBt6C4G3OeZ3WEMeJ/UHFC7Y5qfledwaEPWhkrZ+XzeaMprSM/nOTBY9pjn5+c5ODgoNSlsykSHLM7FdZx1gRgFWTCHniuHiRhIUrWEl2yCfXJyUrreU7FJTQ1ys7q6Wmps6PuR1NtFEvZB0l5dXeXVq1cFTTNOFJyZwyCk474ZQ+QduXXlKXPH2L4PaXqrjQeCgGJhTW08sJx/+Id/2NhkKEnxknhviKv2oiBnE/Aa7RoCEI4RS5tQNNRGKS8vLxuEJsLZTr+1DY0zSBx47iQNo+CwgeIp+nLwN/dO6hblwHBwfSM9Nmrm2pCKnM9rbeAuGDMjCJTYYYCV1eQ0CIXxYe5IeXY6nUZWB5lgfl++fFkUz+EQNTL9fr+x+AyFRLFMUppvMVpxPQkGhkV0x8fHhQPrdhebSz969KggQch8+oJYjrg2qWE7NcLcpN7kGoTHeFl+OFjmYL6QwwjvJsetNh7Ehm4E6/Z3CNxkMimeLakJKYctKDjkklPAIBinHpk4ro2XYwJRABhwzsW5mVzQAoupIL7Yyb6dsUDgTe4ampqAxCsSfsBBEK5Qg4CSuvgKZQCK+765LuOB0HNNntVjxlhgwBw2WsBRWLgbozeQmWN0YLcrVa04GC4MNK8B8fHobQSJcXEWDEPFM1KT4xCZ+pBOp1MaKjOuXjtDyLG0tJT9/f0yV4ylq2AZI5PO7KniBstcx2PG+iHk3nPT5uoYU56Hz9/0uNXGI6kJUrwxpBcHAolQuojIhgFj4Sa+sOmuxrNyJSmlyHgWlM3MN8VEVVVXk6JUQHRgNe9zTQQK5XB2xp7EXpTvg6LwnCa/2pyQGwe5mIrP8DlCF4QO5SE2J4Pl+3FjZzy6C7+skIwTCAZDQE2Myc7ZbLH600oBkjEsNwLBaNGI5+DgoIyrP+dydeSGzEeb9+F5OEiXX14uurGxUfZ4PH4r7UvmrtPplNDaIbPL67kv7hVj4fu38+DAWPM55sYcobM9PhzWfdHjVhsPe7eqqkqxE0YB4nM2m2UymTTy9B50pwF5P0mZaBePIaiOxx2bJml4ZgQBj4HiEyaAJrzga319vVFy3Wb7bczs7ZOa53EGyYiF+zOfg6LxfV/XhwWQBW1tw5DUGyrZ6DhzZSMG+sK7OhQkCwWyIEPE+XmtHc449HPWrN/vN4rxMCYHBwflWrRk9MGckbpkHr2grZ1m5cBAzGazDIfDzGazkv1yep7w4cMPPyx9T0FqOEgcE3P06tWrcm1kHQdGloiwFeOGo+R7DsUcAiE3P7HZFoSbB/7kk0+KEuC1Pejuv8EA2ThgaFAulMKKwHVRdApvuA6KgNezB/G+pFY8BJ7X2/fPtUFNGBHCLaCmBRdldHquLSC8z706nMAIQaCagDURTaaize7z2/wJYQbK7nM6jPH88szOqngtEfMIcjQfAFJwj1eMF/cKsnCmwuiQOWec4ajIrphQTupSdxCnw1ZXgbZJfZQaFLi5uVnW57gzHXu58PwYXHgWG09zRg5FkCeSBsyPdYqxMj/yRY9bbTySFMi4ubnZiHmTZuZlPp+Xfhek2lwZaQhteGsjggADYS8vL4un4nwIJ8pkpfT34EnMu3g9S1K3oEPgfA5SjL43p1XhBfDI9thJvcYBeM5zmlzDKGB020oPPCeUQkg9Hl6mn9TdsxBYhN+hh4uneDYT4+arXEeCstrgc580ZeY+XNoOOkhSPDcKa8ifLIrLGCOHjoyZ0/Bs30C2ZWtrq8zR0dFRCRswHvT+4DxGBRCthLYYsp2dncKTgEpBrSZBMT5khNxnxrLa5q1YdHeT49YbD2LjTz/99K20EwLAgQHZ2dlJsjAu9Ao19EP4bTCAlkBAhID9OLwqMmkWruFpEAogL0ah3++X6lfYdMfrV1d1811Sm+ZLDEFRXl+fsAHhwRvyzFwLBcEL8jm+b4hsojlJ6cLt8eZ+MDi+Bys4SorXs/HnubhHzo+SQnB7hbVXPFNlyRqhg4OD8sydTqexFB5egRCXtSd0E4eANNqjzoLPocgYXcZ0e3s74/E4GxsbJeUKauFcbPB0cnLSQD2MM87RaJltF6gl4llAwey9g9H0uOM82rUeTjhgZG5y3Hrj0el0srm5mTdv3jQ8ADl7wzAMwOVl3a8TAgwvkCwGzUU/Rgj2guPxON1uXYAFhLcSMxEoGVkf4Hh7f5Okrnwks4OnB1bzfEndOBglwvAZeZh3QLhtDPA8CCb1GDaWvG/mn2Iooy6MonkB7qvNJbhfqu+/7RUxmkmdAQBRmtsAUdpQo8QQqcvLy6VVA7LAOBJCgYpALQ6RMBLUF8EpmLdAIS1bZEjG43FBxCg/92Uu4ujoqDgoIzUjNOYE7oPXkRcSCPz2eDEv5prM/XlMbnq819qWH8VRVVVJdRmSu0TcBzHz4eFh7t+/n7W1tZycnJTSarxS21smtVKjvHhbiq5IAQJtCVEcb5svwRgdHh6WrAUGhPtmTQnxO8+cpAgn94X3ms/nJY3o+J/rYoR4nVDIgsf5/FqShieiUxbnaWdqQGgoIbULZID4jrkNXuO5EHLuEYXxmhfmZ2Njoxh01/6gnMy9m+YwFtRegFYxRO6mRi9YE7s4p16vl8lkUvqoVtWiERUFXZeXlxmNRmUMkkVRGBs/nZ6eZjQa5eDgoBGeQoKCbs/OzgoZyticnZ1lY2OjMbZt8poxxGl43NpZO77bJsy/6HHrjcfa2lq+973v5e7du6XdIDGssxBJ7cnm83l2dnby4MGDomyOpbHyVh7DUTIfrAXB8KBg9u6j0aik2hxSUQ8wnU6zsbFR2HAUBsOC1zWRijC3vQnEGXEvSolXN6JJakKS72MAzf9wYHQQRtAJgsa429DAibTfoz0B6Ihz4xF5VowDhhZD5AIwxquqqrI1pw0lCgDKQHlMPHMvhCZLS4t+rg7xbKjv3r1bwiAgv8MlFq9VVb2I8fj4OOfn5/nOd75TwtSTk5OMx+PMZotNxWk4zToX7hejhwGghwzjyvjBsfj7JoUJq1wEiQHBeH8WEX2T41Ybj/l8njdv3hQlRLEIWZI67ua32fVXr17l4cOHWVpaKq3cjD6A4PbQ7UFnIRjXsgHBaEwmk9KYJUnjMwgx3/WaETIdKBAGwY2G8JhJzbc4dMLrYCisfPxvlp7/gfMcJtRccXp5eZl+v98oy+cavkenpTG4HCCDdqMmF6lx3iSlFNv1OWRyMIZ4bFAk9SDMi/kT/nbrR4wIaIfx6/V6pYycENI9P5knPyOELTLKMzx//jzd7mLflNevX5dVyiBYZG44HGZvby+np6cZDoclg4TsLC0t5eDgIA8fPizGEBSCAQZBIacgTZCi9YTDyYebHLfaeHQ6nezt7TXSj0nKgiQPFr9dRLO7u1ta5eP9fR6ugTHBmqOkZGt4jcFHqVzT4VCF9Q9J7f1pPkPXKK834b5MMKKI5kmAo+YdjJ6crbDi8Rr3g8eyAXD4Y+NJOOI0LZ9zirYtiJwPg4qxshGA0GOMTICbowCCUzNC2ApCIIwD4iMDd+7cyb179xq1FKBIruXruzrUJfAgJWA/yIw5cxkAYdRP/dRPpdvtlpXTFxcXOTg4KIbfqVzQDnyK+RFk2rJVVfXSBhs/yzP3wb35PD7eJ3S51cajqqq8efOmVOZVVVU2Gv4si2kre3Fxkd3d3aytrRXSj5LhNtRGIeAzYN7NlreLylgjYSXH42NsQDiEQF6I5KpYpzq5lqEywkZY4ec3v5CkIJw2J8RzYoDMExgicz4yRyArrmvuCeXCiHoHPH+GsAHl8HtcL6nT5I7Nqeh1Ja3RQ7/fLw2DQUAYZTy4jb8zKkY+zJ/vCb4EXopxdwbL8jEajfLq1aucn5/n8ePH2dnZKWNHyOFtDyB4cZBHR0cZjUYlU4jhR04h+DF+rlimj4j1AAPTvl8+8z7HrTYeKNbBwUGShbAZdXzWw1v5CSsGg0EODg6yvr6eyWTSaBAE9IXYItPhLIWRANacTAqeDA9DBSFtBhE4k1xJiiLYK9DgBuXlb/pZ2gAQsjgkQeG9BQJoA3TFveDdjexQAsajvZUBCoxiusu3S7Mx1Dw7sN+cjw0VBt3rXCCtec99Tsh2QERvbGwUPoHnhnx06MZ9MgauJMXYUzMDGuJ83qGPefd8wo8MBoOCPqfTRatMlH9tbS337t3LeDzOs2fPyrmSlBaTx8fHZVydETs6Oipy65XYTq0n9RYPLuBjjp0hRAZvetzqVC2KmCwment7+y04/FmHOYGqqrKzs5PpdFqKcdjGwGSk92v163guL/pyFSQTPJ1OCyQndmaBGobHMTuxOALvkIkuXSAQd4Pis0YLpHwpRmK9DfCXa6OY9kImb/mcy8K93sNhISFNUtcXcA5DfBtpEAoKQ2bEz8B9cq8uwsNos0qW95AReqdQAwFisXFifhwKkVUyYnTmwu0Enc7mWXnGdkEWCyL7/X62trYyHA6zsrKSR48e5ad/+qeLsUwWWyzQJhF0wntUvvr63KtXk2MocDCMi/XFBuN90MetNh5MFhD2xYsXSX7wA9vDQD6CQugJcnp6Wir5UHYEgFSc6w5QUuBfOyZmkoGQ1HpgrMgWODRJ6swL4QP37VShM0M8k4WD6sR2ypjsA97JS8tRdFCOx4sDXsWpbf7nPCgriAZF5nms7FVVd2gzwUoPEqdfeV4jKshVj3ebr0L5nW1aWloqRuTOnTulAA8kYxQHr8D3k3rR2buyTxh2E7PMG/PiEgHGiJ/BYJCvfe1rpeMYz2TUjDxQy+H7Rd65d2cSCZc9r8y7j59Y41FVVYkHsa5+zz8+nH4ipoeRnk7r5cyHh4c5PDzM3t5eY2WjwxV7IE+QW9aZm+h06iIsF53BlXB+FIDP2KBwjxygA56b39yXu1Tx28oOzDXhS4tCjAlKwL0a4uJFbUAtqKCldijkgiejJ/8YkdiQobQoo+eaMQTBIBsYEvNQVrR3hY2ur2EckRuTxIRRzF2v1ytz1l4iwGsgURtc7hujcufOnXzlK1/JV77ylRL2IottpzGbzXJ2dlaK4DifeR3G0vU4ll/G7335juSWGw+TZ+PxuLz+w+I0ZwC8yIjzTKfTEruiqCyG8hoCs/8WIhTfYQiVftPptBRvOUXLPZERsHHAwFVVvaSf0MQxrY2RK0p5zSQp7yFAxMAQhDybWXoOe0CMJyiDZ25XmHLeNqfhUMlL8LkGGY72a0mTc3JLAXtlL3QjrIQgd3jF+eFieCYMx+rqakEIhEOMC3ICme12j8ynZYR5Rk43NjbKOHMukBlGcmNjI1/72tdKyh/+zTIymy1Wj1OSfnFx0WiBaQOLgcdQOHX9PjyHj1ttPIDykGEcn8dqmhxz6pEaEYQaRcAAnJycNErZgfugFRNwzkpgTJzN4HUTrsmCGINH4PPt2o22ApOZcTEawmuBsYfhvHg7jx0xPGiC73DOpFYclMGGiO9wv26uw/dtOEkbGj0w/oQVjBWkIN6X2gbuh3CE9xhPxszZMNesQByz1IBQifccEpprM/Fso4TB5n8Ibc7V6XQatShGaigyKAQ0+OjRo0K0MjesiTEi8dwZfRul+T6c+v3/jfEgV/6DwpTkbWtqpaUXh6sTqQi0YeF7SXJ4eFjIK68TwCu2K14Ji0jFofSuEcCL0k2b87U9PMKB4rb/fxf05DPOSIBcEP6k7kVC7wsbl6qqyjPbU9mwQaS694ah87sMDygFdIIgO2wCFWA0/P7q6mpp5IySoRieU2TA52iTqV5E6Pedigc1YmyM5hh/7pnnS+pQy9kZOp8PBoMMBoMSJiGPyO6dO3cKqfqlL32p9EG9e/duDg4OSqrYSK+NMpziR/55dn/mXfpyk+NWG49ksQUfvMcPO9oChDBfXV1lMpkUL4Mg03OyzRkAF5msk5OT4lW4DueZTqeFNKyqqqTbgKcYqKTuAG6BxXD4vlCIdpES50jqPpr2MggjRsPXhaRMmux7G9EZsidpfAdBxQhQfwES8HecPvS9cX48MfdqhQd9cC3qLRhTruXwgXnxPVi5QUsOvzCsINGVlZXG+hp6tbbnm/sBkTkFzfxAmnIfODDvKkdXPK7FdQaDQT766KPyGVphmizme4wHRqONRI0SOYwM3+e41cajqhZFYijbDzvalpVBMgR2cdHKykppVwcCIJZM0rDyKLk9mnmVqqpKbt+QHEV1HEyRElDcLDjhBNtI8j5GBWX1WgWHS21S0XDVz+h0IlCfc5hw5J4YR0N1xhz47PvAa6PwhvoYFeaJa+KR4Xtc3ESKldoa4DnZHLgIxon7496NrgzlMVRwF4Sa8Bt8HmIXxMW8MgfmT7wIkWejmI1xTFKMGsgA4rzb7WZtbS0ff/xxORfz4+8b3XkpAg2HSBHbGfj773vceuPRXoPxwz7vvy2gV1dXefnyZWazWYHAKysr+fjjj3N0dNQgSc2XIIzEtFhz8xIo0NbWVmlcy3URcM7Z7XZLtyjgN0aGiUfIk2aPSZc0W0ld+WmIjWDxmgle7tl1HUmdJXBIldTrUxz7o6TO0AD5uZ4LzhhHDLlTqjY2PLeXB2AI4QDs8Rlv5sXZKbc2sDEC0WE4nHJ1WpjxxggZncHNrK6uNrJDoF0rNwiWjvRGbA5LGeOqWrTd/Pjjj8scOIxKaqfI4sz19fUMBoOsrKyU+SFVvb29Xdpfmpf5iU3VJnnLan7eA+OBAODJdnd3GynBTqdT9tWgjSAlwIQwCAGVpEkdPmBgIL54D6W2QLOq05WiCK4Vle/2er1SiowCuu4ED2kE0g4ZGENCIRtHn+ddAm1SEATB573a1EbE4Zc5CRQKT8/BeekSZgTFsziM47ngUZxZwqBwXmd3MFztJQXMM8YK2bHxhVw1l0RI64I9G2gIcYcSjAcl78iNw1x4JX5Go1FjZbjnAv7n6uoqg8GggVRPT08zHo8LskZuMSbOrt30uNXGA0G9yeGMQ1J3+X716lWBhYPBoJG26/f7ZT3BbFY3y6WaFC9q8hGhYILIwdt4IZTsF2vS0OQWAoGS4omB6giwMzHmcZxqRjhMpNlYMT4+QAJ8lmsRLjgjg9DDp3j9ibkdowOuT20G97G8vFxIZOadewa9YHRQIM7rTBWKy3VNRmMcIEsd3jkty/l5jXmwccJwMQ5e3IhB4HOEOtwv92nPj6wwTrPZogE1RuXRo0eFT0EOGEfP28nJSSaTSQOhggT39/dzdHT0znm/6XGrjYcPw8jP+/kkjVgRxXr16lUGg0Fp1IJnoxwexSaFmDRhL7/xHpeXl7l3714Raiso57ZwGP4TOqDsCLbRA57NIYpJz/ZKSxQYI5Q0qx+5B/MaFjSTpJyL1zmnUUE7fOBenPFK0jC8Jv3c4MakphXfFZNJHY44Y2EC1yErh8MRZzy4F/7nejZOzA2fZTwJwbgOSIDDK3pBSVTUzufzRm8OCsrgw0AfKysrZXGoyfB+v1924wNZmMshDAJNzmazgkQ43geB/NgYjy+aZnIsZ4RweXmZ58+fJ1kU77gbl8k98wlW/m63XouCErm8HKXkftspP5QCYUUwOXc7lGiTY5Bwzgq0ParTk5zXi8G4BmNkUtid1BlHX4u0pO+Ne7LndT8MFAheifOaYEXpkxTUZwNEQZqJYpPRoEmjBs4Fj8P4uB8MKM6owvdpmTCHhSFgLHlGuCyv/GWtkddAudmPeRjGxcZ9MBhka2urXN9Itc3jYDAcvniejUjf97jVxuMHPeDnRSEMclJb5fF4XAwAMehgMChIAxLNHAKkqzM2KNXV1VVJ73rRF/GrY1vuG8GDX2kbDIQUxXJGBA/ENfDW9jo2dvbE5lZsREzC4sXeZeCc8nRoY47F3tjcDk1+uDYeHCNhkhNjYX6GClIX3/GeEZJDBJf9M0aMvVFLUtezsDK5rYxGKtwvc8NnGAfvAQP34vsmmwZys6w6k8d1VlZWsra21jDCzH0bmXKYc2nrDb9/ognT5G1D8cNYYkNlPu+sxcXFRZ49e5alpaWyuVG7LLht1e3B3OkK9r+qqrJiFwX3ehKnPXkekAvvOwY3IQucdzgB9LUncS0Az28iFeU0sctzGaHQOBqBpLybeNvrYmy0vL6Hz6MASRNdoZgYxU6nU9Z7YKRd8MYz2zu3M17cG+OIQptQ5VmpeHVhlzko5AjUkaRhKJzu5Z5tjKnn4N58TxjQfr9fCG/IW5biM1a+9ubmZhkL5KPb7ZYFjnZONnrMNTLh877P8WNhPGwIPi/3YQUiNmWiEFxqMvAYTAiKRlztxU2gDGJSYnaExrG0wx4+Y+jNOUExCC1CndSLrvB4fJ4OWlzf1Y8YOKcfXQ9CpsT8jAucUC7GyvUhCC4ZI4yaeQKHGkZa9vj2htwjRttj3eYcqGFAEW2YnUJ28RiVsIzf8vJyRqNReRa+x98oXTv96lQz57ZTcvhjFGlD5jlnHAihNjc3G6lUj+/KykpZyu+sl8Nl7t+hIeEq5/urSNGW533vM/xzPr4o6uAzn/U/g/hnf/Zn2dvbK8iDScVrMiEgBYR/dXW1sbCwdgixAAAMmElEQVSMdQgs0WeCEDyqER2bWzERVq7t7BDGB74EYXO2g9dRGFAKCm/0YIPBNXh2BB4PiJK8izBs16JwPW/QxNhxXsJDF97Z0GB4zSnZcNrYYrhMpnobBocYpDmTZk9ZZy2c0eGn0+mU2hZkgbEw/4IzAc0xpiAHFwo6zOK6LHIjvQuCGY/HZRyTlOs/ePCg7DNjAh6H5zCVa+IUrEs/EuRRVdV/VVXVq6qq/ole26yq6veqqvqL69+j69erqqr+QVVVf1lV1T+uqurn9J1vXH/+L6qq+sZ733l93rcGwv970lhjkCwEcDKZFAOQ1IaDdO3h4WGJXfE6eEcq+NinA7IRz4kw0unKng0hp9eIERKhEgYDg2BuA+EwuYqhMcfAfaGs5gNANlzbwg/kbWdanEY25OfzPDNEsZWBObCHtOHkXjmvv2MOA4NhWG6+hbL2pN6My8SsiUVzS3j2dhWqkZA9PWtNOp1FHYg9PCErxpRxgkshpMNgklGxcWe5flKX9Pf7/fJ591vh3hx6mf/y8n7G5keFPP6bJP9u67W/l+T35/P5x0l+//r/JPlbST6+/vnVJL+eLIxNkl9L8q8m+RtJfg2D88MOK/9nvd/OwnzW5+fzeenQ1Ol08qd/+qeZzRYbFFMIlNTt6VwDMplMijciXsYwIDiukESp+Unqfg4ObyzsCADeAyOBAcArJykpZT9bUqcjESCHThhAeAiMHEqCciKQ7bU2VuL2ePN9k9CQnCAXyL/2eczH4IWZBxS9qqqi3M7uoDBGbEkaRWTUePjaLqZz+pP5BN0xXzwP/FaSxl61oASHMa51MadlRMK9Wma5FityCXld6Mb5HKo4a4hTdajbJk/f9/ihZ5rP5/93kr3Wy7+U5Leu//6tJL+s1//hfHH8YZJhVVWPkvw7SX5vPp/vzefz/SS/l7cN0g+6hxvBLKMSJg5Bq6oqu7u7+e53v1vYa3tqqkGx8m5v1+0uCnfYgMj3iAAyiZ+V6XA4YGEypAXKcq6k3tEdZeYwb2MDhPDCgRiJoBTwHRhNxovnMrFrLoBm0pyL+zLP4XCD873r76TumNUOtwgFnGHiPkEgRkI8q405c4KB85i3xx4lxSDYqfB8TlU7BF1aWioGhnCHeeN8OCru1y0FMNhwO/1+v8ilHVybBzR5jhFphyn8b6fzPsdNzdCD+Xz+Ikmuf9+/fv1Jkqf63LPr1z7r9beOqqp+taqqb1ZV9U2/bnTxg5BF8u7BaXMlCMK3vvWtBh9hqMrf/I9BIbzgHBCF9tD0YECgXDhksosGw+6DYY/IyktXjeINUTTCC8bFhU9GL8B21xZwfxhUIx0KmMx5cC0bGowUigkxDFLyQsP5fN4okTdvwnkcyhgVMC48uwlZh03mXZxKbmdL2iiPH6MOQ37e82rbJIW34FqESm1EwfgzxzbGGHD2Rmbe/OxwTQ5L2ql560G7Utiy/1dx/FUTpu/S6vkPeP3tF+fz35jP51+fz+df/6yLvA9hmtQohH1E2UTbqVWU7+rqquyCzrk++uijTKfTt/qdsky+16u7bnc6nfJ626hhcBAGFMjLqM13cC0bChs8DBlcCvdictXwFcHE2KDsKDFxve/X6VV7c7/fFnJCERTcVa3mNRw+4d05zP8QChGS8RzeHsHIzAaHJQJWYt73XDi0chhFpsqGF8SVpPAaDifo9uWw1IaFzbCqarG1CKiTnemMSoyeHZIyXsgOoQ5L+h0+/1UdN916Yaeqqkfz+fzFdVjy6vr1Z0k+1Oc+SPL8+vV/s/X6//VFLuiQ4IcdCI3DHX8Xr4sQf/vb3y7eAy6CHgreXxQSbG9vrxQy0Safay0vL6ff7+fw8LAQZxgl4mM8Y1Lvw9LO2CAwJhCTFOUzZ9Cu1Wh7UgwHAuvQCUEDNRhNMe4maZ3uJKa/uLgotQ4mJhlrx+sOwXh2MjPOTPgzNnDMja9jI0J444pVX6edTuZ6Rn2M83Q6zdraWtltEN6FcA0jgUHzRlJGYKALV7maWMaYYNggSieTSSO04nl4RiM/HMu9e/cyGo1y7969nJ+f55NPPsnLly/z7W9/+0ah/w86boo8fjcJGZNvJPkdvf53q8XxC0kOr8Oa/z3J36yqanRNlP7N69c+1/F5DMa7DgShPWiw5d3uovHKZDLJJ598Uq7DPqUorr3ehx9+mG63m6Ojo8JJtBluwglSqN460AjBfAhl8r1er9FL0zwGBgjh9zXtlbguAjeb1WXxKB2Gys1urGzmTEAbKBvXwRNTHu+1KyYHYfxBWjZg/197ZxNbRRXF8d8B+voqrS20QmqwlibGhLAQ4gLUmMaPxBBjN5hg3Ch2oxs/FgZ05cKFxhgWGsFoiDGKIBI0TYwLZUmqELWigJYiiFQEzFOjISH0uLj3TC+1pX0T3rz7kvtLXt7MnXlv/nNm7pk7Z+49Y87AKnh4CxTax1oz4Uu7wtZJ+DjXdNhvbDrcv7B1Ex43i6tYS2liYiJ7EZht0+IWFt+wfbZbzTBAarZT1Ww0a+i8Q3tZC82Orem3hFV2jMOLnLVsWlpa6O3tZeXKlcybN4/Tp08zPj5Oa2srg4OD9Pf3Z72gp8YC8yKzVUwR2YFrNXQBZ3BPTfYCu4Ae4CTwoKr+IU7Ja7hg6L/Ao6p6wP/PRuA5/7cvqur2WcWJqB0EmLsTCT27EbY8rDt6c3MzbW1tLFmyhPPnz9PT00N7e3vWzLOWiVXo7u7uy2IiXV1d2T25pSUsl8t0dHRkL1cOH1+GwbcwcGfLrOKYowhPvLD5bL8tlUpZB7awA1wYewgf38GkQ7XKb/sT3rqEQc+pwcBwnE/YCSrMi2H/H9rdHEZYac2pWKUM4xVhT12r2GGLJLy6h70nrXJOPVfCGJBdqcP9D8vNaZlu02J6bDv2KNhsEb6Dxvatqanpshc32TYtDmQXlImJiewcgsm35F24cIFKpcLSpUsz+9nvyuVy9kKpzs5OTpw4wdjYWPZEsFKpZOsNDAywf/9+RkZGuHTpUpYq09vm4JXCBDPWs7xX9SIQkb+Bo/XWMUe6gHP1FjEHGkUnNI7WRtEJ02u9UVWvq/aPon7dJHA0j0esByJyoBG0NopOaBytjaITrq7W6LunJxKJOEnOI5FI5CJ25/FmvQVUQaNobRSd0DhaG0UnXEWtUQdME4lEvMTe8kgkEpGSnEcikchFtM5DRO4TkaPicoNsmv0XNdVyg4jsE5HDIvK9iDzpy6vOa1KQ3vki8rWIDPn55SIy7HXuFJGSL2/286N+eW/BOjtEZLeIHPG2XRuxTZ/2x/6QiOwQkXIMdpV65tuZbhxEvT/AfOAY0AeUgG+BFXXU0w2s9tNtwI/ACuBlYJMv3wS85KfXAZ/iBgSuAYYL1vsM8D4w5Od3ARv89FbgcT/9BLDVT28Adhas8x1g0E+XgI4YbYobAX4caAns+UgMdgXuBFYDh4KyqmwILAbG/PciP71o1m0XebJUYZC1wGfB/GZgc711BXo+Bu7F9X7t9mXduE5tANuAh4L1s/UK0LYMl6DpLmDInyjngAVTbYsbX7TWTy/w60lBOq/1FVKmlMdoU0spsdjbaQiXoyYKuwK9U5xHVTYEHgK2BeWXrTfTJ9bbljnn/yga3wRdBQxTfV6TItgCPAvY+OtOoKKqNrY+1JLp9Mv/9OsXQR9wFtjub7HeEpGFRGhTVf0VeAU3jmscZ6eDxGlXqGG+nZBYncec838UiYi0Ah8BT6nqX1dadZqymusXkfuB31X14By11NPOC3DN7TdUdRXwD5PpLKejblp9zGAAWA5cDyzEpdycSU+U5y9XId9OSKzOY6a8IHVDRJpwjuM9Vd3ji8+Iy2eCzC2vSa25HXhARH4GPsDdumzBpYO0cUyhlkynX97O/1NO1opTwClVHfbzu3HOJDabAtwDHFfVs6p6EdgD3EacdoXqbZjLtrE6j6+Am3w0u4QLOn1SLzHixkG/DRxW1VeDRdXmNakpqrpZVZepai/OZl+o6sPAPmD9DDpN/3q/fiFXSFX9DfhFRG72RXcDPxCZTT0ngTUico0/F0xrdHadZvu1y7dTRMApZxBoHe6pxjHg+TpruQPXjBsBvvGfdbj72M+Bn/z3Yr++AK977d8Bt9ZBcz+TT1v6gC+BUeBDoNmXl/38qF/eV7DGW4AD3q57cZH+KG0KvAAcAQ4B7wLNMdgV2IGLw1zEtSAey2NDYKPXO4rLwzPrtlP39EQikYtYb1sSiUTkJOeRSCRykZxHIpHIRXIeiUQiF8l5JBKJXCTnkUgkcpGcRyKRyMV/Hk2yZyUE9woAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "ex_patient_id = test_dcm_fps[0]\n",
    "ex_img_path = os.path.join(img_dir, \"{}.jpg\".format(ex_patient_id))\n",
    "\n",
    "plt.imshow(cv2.imread(ex_img_path))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "8ac8f43aae2bc82599fd1bb4a66d1dd444b66a28"
   },
   "source": [
    "## 3. Prepare Configuration Files for Using YOLOv3\n",
    "We should prepare and modify config files, and bring pre-trained weights necessary for training. This proceeds with following four steps.\n",
    "```\n",
    " cfg/rsna.data\n",
    " cfg/rsna.names\n",
    " darknet53.conv.74\n",
    " cfg/rsna_yolov3.cfg_train\n",
    "```"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "72fb38e755495fca3bf456c4494c05b5300d3411"
   },
   "source": [
    "### - cfg/rsna.data\n",
    "This file point to RSNA data path\n",
    "  * train: Path to training image list textfile\n",
    "  * val: Path to validation image list textfile\n",
    "  * names: RSNA class name list (see <a href='#3.1.-cfg/rsna.names'>3.1</a>)\n",
    "  * backup: A directory where trained weights(checkpoints) will be stored as training progresses."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "_uuid": "f78dcf8c107c67e49f8bba8f483ab6a5b409ec0b"
   },
   "outputs": [],
   "source": [
    "data_extention_file_path = os.path.join(cfg_dir, 'rsna.data')\n",
    "with open(data_extention_file_path, 'w') as f:\n",
    "    contents = \"\"\"classes= 1\n",
    "train  = {}\n",
    "valid  = {}\n",
    "names  = {}\n",
    "backup = {}\n",
    "    \"\"\".format(os.path.join(metadata_dir, \"tr_list.txt\"),\n",
    "               os.path.join(metadata_dir, \"val_list.txt\"),\n",
    "               os.path.join(cfg_dir, 'rsna.names'),\n",
    "               backup_dir)\n",
    "    f.write(contents)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "_uuid": "684b3e9dde58f56a746952d2934661712e417161"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "classes= 1\r\n",
      "train  = /home/qinghe/cri/projects/AI/yolo/metadata/tr_list.txt\r\n",
      "valid  = /home/qinghe/cri/projects/AI/yolo/metadata/val_list.txt\r\n",
      "names  = /home/qinghe/cri/projects/AI/yolo/cfg/rsna.names\r\n",
      "backup = /home/qinghe/cri/projects/AI/yolo/backup\r\n",
      "    "
     ]
    }
   ],
   "source": [
    "!cat cfg/rsna.data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "46cd3f02af89917b9715f51f188a47d6030bd758"
   },
   "source": [
    "### - cfg/rsna.names"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {
    "_uuid": "917144c1cffd160a997db563cf9506db13e69414"
   },
   "outputs": [],
   "source": [
    "# Label list of bounding box.\n",
    "!echo \"pneumonia\" > cfg/rsna.names"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "6415f6efcae6fca1812cb402af668428d2918e46"
   },
   "source": [
    "### - darknet53.conv.74  (Download Pre-trained Model)\n",
    "For training, we would download the pre-trained model weights(darknet53.conv.74) using following wget command. I recommend you to use this pre-trained weight too. Author of darknet also uses this pre-trained weights in different fields of image recognition."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "_uuid": "ec2d5fbdfe6f3eebd553495d10f9adcdae29f16f"
   },
   "outputs": [],
   "source": [
    "!wget -q https://pjreddie.com/media/files/darknet53.conv.74"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "978655a61edee55f635c8bf8c5befae03c838652"
   },
   "source": [
    "### - cfg/rsna_yolov3.cfg_train\n",
    "* Basically, you can use darknet/cfg/yolov3.cfg files. However it won't work for RSNA. you need to edit for RSNA.\n",
    "* You can just download a cfg file I edited for RSNA with following wget command.\n",
    "\n",
    "\n",
    "* I refer to the following articles for editing cfg files.\n",
    "  * [YOLOv3 blog](https://pjreddie.com/darknet/yolo/)\n",
    "  * [YOLOv3 paper](https://pjreddie.com/media/files/papers/YOLOv3.pdf)\n",
    "  * [how to train yolov2 blog](https://medium.com/@manivannan_data/how-to-train-yolov2-to-detect-custom-objects-9010df784f36)\n",
    "  * [darknet github issues/236](https://github.com/pjreddie/darknet/issues/236)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "_uuid": "f64517332ee89f69af1b013ef6376b19d639f0a6"
   },
   "outputs": [],
   "source": [
    "!wget --no-check-certificate -q \"https://docs.google.com/uc?export=download&id=18ptTK4Vbeokqpux8Onr0OmwUP9ipmcYO\" -O cfg/rsna_yolov3.cfg_train"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "f03592ee228f9ae16edac192071955d01233a147"
   },
   "source": [
    "## 4. Training YOLOv3"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "69f52f31dea66f3da1952b537c63dc5376d64492"
   },
   "source": [
    "### 4.0. Command for training with Pre-trained CNN Weights (darknet53.conv.74)\n",
    "* I didn't run following command on kaggle kernel becuase of the long output.\n",
    "* If you crash with  'CUDA Error: out of memory', Solve it by Editing 'batch' and 'subdivisions' in 'cfg/rsna_yolov3.cfg_train'\n",
    "* If 'batch' and 'subdivisions' are 64 and 64 respectively, for every iteration only one image will be loaded on GPU memory. So it will use less GPU memory."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "_uuid": "0be53a581ab12776c1c276f72acbf08b2db16d77",
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "#!./darknet_gpu detector train cfg/rsna.data cfg/rsna_yolov3.cfg_train darknet53.conv.74 -i 0 | tee train_log.txt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/bin/sh: 1: ./darknet: Permission denied\r\n"
     ]
    }
   ],
   "source": [
    "!./darknet detector train cfg/rsna.data cfg/rsna_yolov3.cfg_train darknet53.conv.74 -i 0 | tee train_log.txt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "1c43f25ac894192e9c0e03cbcb4c36072210be12"
   },
   "source": [
    "### 4.1. Command for training with Multi-gpu after 1000 iteration\n",
    "\n",
    "If you are trying to train with multi-gpu, there are three things to watch out.\n",
    "* (The # of gpus)x('learning rate' in 'cfg/rsna_yolov3.cfg_train') is the real learning rate for training\n",
    "* I don't recommend you to use multi-gpu for first 1000 iterations. with multi-gpu, training would not be stable. Use single gpu before 1000 and after 1000, continue with more gpus.\n",
    "* By the way, If the # of gpus is over 5, training is not stable.\n",
    "\n",
    "```\n",
    "Above things will depend on your environment. The best way to find the most appropriate method is to just give it a try :)\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "_uuid": "a3694016ec0d7372b131837b8e6279a6f59a12a2"
   },
   "outputs": [],
   "source": [
    "# !./darknet_gpu detector train cfg/rsna.data cfg/rsna_yolov3.cfg_train backup/rsna_yolov3_1000.weights -gpus 0,1,2,3 | tee train_log.txt"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "3cfd26f1c657bd30952f04b4ebbe6ca6ff7dfea8"
   },
   "source": [
    "### 4.2. My Plot of Training Loss\n",
    "It's a loss graph up to about 2000 iteration. Since it tooks too long on kaggle kernel, I brought it. When learning, don't be surprised of big loss values at the beginning. Stay calm and It'll go down. Please See the following loss graph."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "_uuid": "417c9157d98d5a552f519a3935670fd4bd437a4b"
   },
   "outputs": [],
   "source": [
    "!wget --no-check-certificate -q \"https://docs.google.com/uc?export=download&id=1OhnlV3s7r6xsEme6DKkNYjcYjsl-C_Av\" -O train_log.txt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {
    "_uuid": "f761373c3c46f5c7875d78cc3023becc5924f6a4"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1440x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "iters = []\n",
    "losses = []\n",
    "total_losses = []\n",
    "with open(\"train_log.txt\", 'r') as f:\n",
    "    for i,line in enumerate(f):\n",
    "        if \"images\" in line:\n",
    "            iters.append(int(line.strip().split()[0].split(\":\")[0]))\n",
    "            losses.append(float(line.strip().split()[2]))        \n",
    "            total_losses.append(float(line.strip().split()[1].split(',')[0]))\n",
    "\n",
    "plt.figure(figsize=(20, 5))\n",
    "plt.subplot(1,2,1)\n",
    "sns.lineplot(iters, total_losses, label=\"total loss\")\n",
    "sns.lineplot(iters, losses, label=\"avg loss\")\n",
    "plt.xlabel(\"Iteration\")\n",
    "plt.ylabel(\"Loss\")\n",
    "\n",
    "plt.subplot(1,2,2)\n",
    "sns.lineplot(iters, total_losses, label=\"total loss\")\n",
    "sns.lineplot(iters, losses, label=\"avg loss\")\n",
    "plt.xlabel(\"Iteration\")\n",
    "plt.ylabel(\"Loss\")\n",
    "plt.ylim([0, 4.05])\n",
    "plt.savefig(\"loss.png\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "ef7409909b9403f43b2c92760220b23a38d42029"
   },
   "source": [
    "## 5. How to use trainined YOLOv3 for test images (command line)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "c97da5c5f52e0257a3b1a2568e6260a9560700a0"
   },
   "source": [
    "### 5.0. Copy sample test image"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "_uuid": "688344d91c7f7fb87418486b9b6b5f91a8b26e9a"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "00704310-78a8-4b38-8475-49f4573b2dbb\n"
     ]
    }
   ],
   "source": [
    "ex_patient_id = annots[annots.Target == 1].patientId.values[2]\n",
    "shutil.copy(ex_img_path, \"test.jpg\")\n",
    "print(ex_patient_id)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "245c607d7753b947a91fe4d7ce8daf983afdfd0e"
   },
   "source": [
    "### 5.1. Load trained model (at 15300 iteration)\n",
    "Since i uploaded the weights file (large big file) on my google drive, the command is very very long ...\n",
    "* It's a weight file at 15300 iteration, which I made submission file with. If you use this weight, you'll get a score of 0.141LB.\n",
    "  * Up to 15300 iteration, It takes about 8 hours.\n",
    "    * In .cfg file, I set 'batch' and 'subdivisions' as 64 and 8 respectively.\n",
    "    * Up to 1000 iteration from 0, it takes about 1h with **one** Tesla P100 GPU.      **(1000 iter/h)**\n",
    "    * Up to 15300 iteration from 1000, it takes about 7h with **four** Tesla P100 GPU. **(2043 iter/h)**"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "_uuid": "91d60e964d4cbd5397f6094fd811d115977abc3e"
   },
   "outputs": [],
   "source": [
    "!wget --load-cookies /tmp/cookies.txt -q \"https://docs.google.com/uc?export=download&confirm=$(wget --quiet --save-cookies /tmp/cookies.txt --keep-session-cookies --no-check-certificate 'https://docs.google.com/uc?export=download&id=1FDzMN-kGVYCvBeDKwemAazldSVkAEFyd' -O- | sed -rn 's/.*confirm=([0-9A-Za-z_]+).*/\\1\\n/p')&id=1FDzMN-kGVYCvBeDKwemAazldSVkAEFyd\" -O backup/rsna_yolov3_15300.weights && rm -rf /tmp/cookies.txt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {
    "_uuid": "d745c3d07102f9f1f27bf57a0c009311f9e0fd3b"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "total 235M\r\n",
      "235M -rw-r--r--  1 qinghe qinghe 235M janv. 21 16:56 rsna_yolov3_15300.weights\r\n",
      "4,0K drwxr-xr-x 10 qinghe qinghe 4,0K janv. 21 16:55 ..\r\n",
      "4,0K drwxrwxrwx  2 qinghe qinghe 4,0K janv. 21 16:54 .\r\n"
     ]
    }
   ],
   "source": [
    "!ls -alsth backup"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "41b4ca297869af6808b758824960e9dad12d00c0"
   },
   "source": [
    "### 5.2. cfg file for test (not for training)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {
    "_uuid": "17708ffdc9b834060ccb2b8cb4aefaa3c00b96b6"
   },
   "outputs": [],
   "source": [
    "!wget --no-check-certificate -q \"https://docs.google.com/uc?export=download&id=10Yk6ZMAKGz5LeBbikciALy82aK3lX-57\" -O cfg/rsna_yolov3.cfg_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7fd403fdaf98>"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "plt.imshow(cv2.imread(os.path.join(img_dir, \"{}.jpg\".format(val_series.tolist()[2]))))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7fd403fb5e10>"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "plt.imshow(cv2.imread(aa))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [],
   "source": [
    "aa = os.path.join(img_dir, \"{}.jpg\".format(val_series.tolist()[1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 68,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "/home/qinghe/cri/projects/AI/yolo/images/d373b335-d9d9-4594-b97e-03ec9aa8daa2.jpg\n"
     ]
    }
   ],
   "source": [
    "print(aa)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "layer     filters    size              input                output\n",
      "    0 conv     32  3 x 3 / 1   608 x 608 x   3   ->   608 x 608 x  32  0.639 BFLOPs\n",
      "    1 conv     64  3 x 3 / 2   608 x 608 x  32   ->   304 x 304 x  64  3.407 BFLOPs\n",
      "    2 conv     32  1 x 1 / 1   304 x 304 x  64   ->   304 x 304 x  32  0.379 BFLOPs\n",
      "    3 conv     64  3 x 3 / 1   304 x 304 x  32   ->   304 x 304 x  64  3.407 BFLOPs\n",
      "    4 res    1                 304 x 304 x  64   ->   304 x 304 x  64\n",
      "    5 conv    128  3 x 3 / 2   304 x 304 x  64   ->   152 x 152 x 128  3.407 BFLOPs\n",
      "    6 conv     64  1 x 1 / 1   152 x 152 x 128   ->   152 x 152 x  64  0.379 BFLOPs\n",
      "    7 conv    128  3 x 3 / 1   152 x 152 x  64   ->   152 x 152 x 128  3.407 BFLOPs\n",
      "    8 res    5                 152 x 152 x 128   ->   152 x 152 x 128\n",
      "    9 conv     64  1 x 1 / 1   152 x 152 x 128   ->   152 x 152 x  64  0.379 BFLOPs\n",
      "   10 conv    128  3 x 3 / 1   152 x 152 x  64   ->   152 x 152 x 128  3.407 BFLOPs\n",
      "   11 res    8                 152 x 152 x 128   ->   152 x 152 x 128\n",
      "   12 conv    256  3 x 3 / 2   152 x 152 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "   13 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "   14 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "   15 res   12                  76 x  76 x 256   ->    76 x  76 x 256\n",
      "   16 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "   17 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "   18 res   15                  76 x  76 x 256   ->    76 x  76 x 256\n",
      "   19 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "   20 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "   21 res   18                  76 x  76 x 256   ->    76 x  76 x 256\n",
      "   22 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "   23 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "   24 res   21                  76 x  76 x 256   ->    76 x  76 x 256\n",
      "   25 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "   26 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "   27 res   24                  76 x  76 x 256   ->    76 x  76 x 256\n",
      "   28 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "   29 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "   30 res   27                  76 x  76 x 256   ->    76 x  76 x 256\n",
      "   31 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "   32 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "   33 res   30                  76 x  76 x 256   ->    76 x  76 x 256\n",
      "   34 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "   35 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "   36 res   33                  76 x  76 x 256   ->    76 x  76 x 256\n",
      "   37 conv    512  3 x 3 / 2    76 x  76 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   38 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   39 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   40 res   37                  38 x  38 x 512   ->    38 x  38 x 512\n",
      "   41 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   42 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   43 res   40                  38 x  38 x 512   ->    38 x  38 x 512\n",
      "   44 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   45 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   46 res   43                  38 x  38 x 512   ->    38 x  38 x 512\n",
      "   47 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   48 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   49 res   46                  38 x  38 x 512   ->    38 x  38 x 512\n",
      "   50 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   51 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   52 res   49                  38 x  38 x 512   ->    38 x  38 x 512\n",
      "   53 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   54 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   55 res   52                  38 x  38 x 512   ->    38 x  38 x 512\n",
      "   56 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   57 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   58 res   55                  38 x  38 x 512   ->    38 x  38 x 512\n",
      "   59 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   60 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   61 res   58                  38 x  38 x 512   ->    38 x  38 x 512\n",
      "   62 conv   1024  3 x 3 / 2    38 x  38 x 512   ->    19 x  19 x1024  3.407 BFLOPs\n",
      "   63 conv    512  1 x 1 / 1    19 x  19 x1024   ->    19 x  19 x 512  0.379 BFLOPs\n",
      "   64 conv   1024  3 x 3 / 1    19 x  19 x 512   ->    19 x  19 x1024  3.407 BFLOPs\n",
      "   65 res   62                  19 x  19 x1024   ->    19 x  19 x1024\n",
      "   66 conv    512  1 x 1 / 1    19 x  19 x1024   ->    19 x  19 x 512  0.379 BFLOPs\n",
      "   67 conv   1024  3 x 3 / 1    19 x  19 x 512   ->    19 x  19 x1024  3.407 BFLOPs\n",
      "   68 res   65                  19 x  19 x1024   ->    19 x  19 x1024\n",
      "   69 conv    512  1 x 1 / 1    19 x  19 x1024   ->    19 x  19 x 512  0.379 BFLOPs\n",
      "   70 conv   1024  3 x 3 / 1    19 x  19 x 512   ->    19 x  19 x1024  3.407 BFLOPs\n",
      "   71 res   68                  19 x  19 x1024   ->    19 x  19 x1024\n",
      "   72 conv    512  1 x 1 / 1    19 x  19 x1024   ->    19 x  19 x 512  0.379 BFLOPs\n",
      "   73 conv   1024  3 x 3 / 1    19 x  19 x 512   ->    19 x  19 x1024  3.407 BFLOPs\n",
      "   74 res   71                  19 x  19 x1024   ->    19 x  19 x1024\n",
      "   75 conv    512  1 x 1 / 1    19 x  19 x1024   ->    19 x  19 x 512  0.379 BFLOPs\n",
      "   76 conv   1024  3 x 3 / 1    19 x  19 x 512   ->    19 x  19 x1024  3.407 BFLOPs\n",
      "   77 conv    512  1 x 1 / 1    19 x  19 x1024   ->    19 x  19 x 512  0.379 BFLOPs\n",
      "   78 conv   1024  3 x 3 / 1    19 x  19 x 512   ->    19 x  19 x1024  3.407 BFLOPs\n",
      "   79 conv    512  1 x 1 / 1    19 x  19 x1024   ->    19 x  19 x 512  0.379 BFLOPs\n",
      "   80 conv   1024  3 x 3 / 1    19 x  19 x 512   ->    19 x  19 x1024  3.407 BFLOPs\n",
      "   81 conv     18  1 x 1 / 1    19 x  19 x1024   ->    19 x  19 x  18  0.013 BFLOPs\n",
      "   82 yolo\n",
      "   83 route  79\n",
      "   84 conv    256  1 x 1 / 1    19 x  19 x 512   ->    19 x  19 x 256  0.095 BFLOPs\n",
      "   85 upsample            2x    19 x  19 x 256   ->    38 x  38 x 256\n",
      "   86 route  85 61\n",
      "   87 conv    256  1 x 1 / 1    38 x  38 x 768   ->    38 x  38 x 256  0.568 BFLOPs\n",
      "   88 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   89 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   90 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   91 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   92 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   93 conv     18  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x  18  0.027 BFLOPs\n",
      "   94 yolo\n",
      "   95 route  91\n",
      "   96 conv    128  1 x 1 / 1    38 x  38 x 256   ->    38 x  38 x 128  0.095 BFLOPs\n",
      "   97 upsample            2x    38 x  38 x 128   ->    76 x  76 x 128\n",
      "   98 route  97 36\n",
      "   99 conv    128  1 x 1 / 1    76 x  76 x 384   ->    76 x  76 x 128  0.568 BFLOPs\n",
      "  100 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "  101 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "  102 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "  103 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "  104 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "  105 conv     18  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x  18  0.053 BFLOPs\n",
      "  106 yolo\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loading weights from ../backup/rsna_yolov3_15300.weights...Done!\n",
      "/home/qinghe/cri/projects/AI/yolo/images/d373b335-d9d9-4594-b97e-03ec9aa8daa2.jpg: Predicted in 21.916307 seconds.\n",
      "pneumonia: 94%\n",
      "pneumonia: 76%\n",
      "^C\n"
     ]
    }
   ],
   "source": [
    "!cd darknet && ./darknet detector test ../cfg/rsna.data ../cfg/rsna_yolov3.cfg_test ../backup/rsna_yolov3_15300.weights /home/qinghe/cri/projects/AI/yolo/images/d373b335-d9d9-4594-b97e-03ec9aa8daa2.jpg -thresh 0.5"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 73,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "layer     filters    size              input                output\n",
      "    0 conv     32  3 x 3 / 1   608 x 608 x   3   ->   608 x 608 x  32  0.639 BFLOPs\n",
      "    1 conv     64  3 x 3 / 2   608 x 608 x  32   ->   304 x 304 x  64  3.407 BFLOPs\n",
      "    2 conv     32  1 x 1 / 1   304 x 304 x  64   ->   304 x 304 x  32  0.379 BFLOPs\n",
      "    3 conv     64  3 x 3 / 1   304 x 304 x  32   ->   304 x 304 x  64  3.407 BFLOPs\n",
      "    4 res    1                 304 x 304 x  64   ->   304 x 304 x  64\n",
      "    5 conv    128  3 x 3 / 2   304 x 304 x  64   ->   152 x 152 x 128  3.407 BFLOPs\n",
      "    6 conv     64  1 x 1 / 1   152 x 152 x 128   ->   152 x 152 x  64  0.379 BFLOPs\n",
      "    7 conv    128  3 x 3 / 1   152 x 152 x  64   ->   152 x 152 x 128  3.407 BFLOPs\n",
      "    8 res    5                 152 x 152 x 128   ->   152 x 152 x 128\n",
      "    9 conv     64  1 x 1 / 1   152 x 152 x 128   ->   152 x 152 x  64  0.379 BFLOPs\n",
      "   10 conv    128  3 x 3 / 1   152 x 152 x  64   ->   152 x 152 x 128  3.407 BFLOPs\n",
      "   11 res    8                 152 x 152 x 128   ->   152 x 152 x 128\n",
      "   12 conv    256  3 x 3 / 2   152 x 152 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "   13 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "   14 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "   15 res   12                  76 x  76 x 256   ->    76 x  76 x 256\n",
      "   16 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "   17 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "   18 res   15                  76 x  76 x 256   ->    76 x  76 x 256\n",
      "   19 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "   20 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "   21 res   18                  76 x  76 x 256   ->    76 x  76 x 256\n",
      "   22 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "   23 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "   24 res   21                  76 x  76 x 256   ->    76 x  76 x 256\n",
      "   25 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "   26 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "   27 res   24                  76 x  76 x 256   ->    76 x  76 x 256\n",
      "   28 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "   29 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "   30 res   27                  76 x  76 x 256   ->    76 x  76 x 256\n",
      "   31 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "   32 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "   33 res   30                  76 x  76 x 256   ->    76 x  76 x 256\n",
      "   34 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "   35 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "   36 res   33                  76 x  76 x 256   ->    76 x  76 x 256\n",
      "   37 conv    512  3 x 3 / 2    76 x  76 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   38 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   39 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   40 res   37                  38 x  38 x 512   ->    38 x  38 x 512\n",
      "   41 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   42 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   43 res   40                  38 x  38 x 512   ->    38 x  38 x 512\n",
      "   44 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   45 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   46 res   43                  38 x  38 x 512   ->    38 x  38 x 512\n",
      "   47 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   48 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   49 res   46                  38 x  38 x 512   ->    38 x  38 x 512\n",
      "   50 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   51 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   52 res   49                  38 x  38 x 512   ->    38 x  38 x 512\n",
      "   53 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   54 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   55 res   52                  38 x  38 x 512   ->    38 x  38 x 512\n",
      "   56 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   57 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   58 res   55                  38 x  38 x 512   ->    38 x  38 x 512\n",
      "   59 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   60 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   61 res   58                  38 x  38 x 512   ->    38 x  38 x 512\n",
      "   62 conv   1024  3 x 3 / 2    38 x  38 x 512   ->    19 x  19 x1024  3.407 BFLOPs\n",
      "   63 conv    512  1 x 1 / 1    19 x  19 x1024   ->    19 x  19 x 512  0.379 BFLOPs\n",
      "   64 conv   1024  3 x 3 / 1    19 x  19 x 512   ->    19 x  19 x1024  3.407 BFLOPs\n",
      "   65 res   62                  19 x  19 x1024   ->    19 x  19 x1024\n",
      "   66 conv    512  1 x 1 / 1    19 x  19 x1024   ->    19 x  19 x 512  0.379 BFLOPs\n",
      "   67 conv   1024  3 x 3 / 1    19 x  19 x 512   ->    19 x  19 x1024  3.407 BFLOPs\n",
      "   68 res   65                  19 x  19 x1024   ->    19 x  19 x1024\n",
      "   69 conv    512  1 x 1 / 1    19 x  19 x1024   ->    19 x  19 x 512  0.379 BFLOPs\n",
      "   70 conv   1024  3 x 3 / 1    19 x  19 x 512   ->    19 x  19 x1024  3.407 BFLOPs\n",
      "   71 res   68                  19 x  19 x1024   ->    19 x  19 x1024\n",
      "   72 conv    512  1 x 1 / 1    19 x  19 x1024   ->    19 x  19 x 512  0.379 BFLOPs\n",
      "   73 conv   1024  3 x 3 / 1    19 x  19 x 512   ->    19 x  19 x1024  3.407 BFLOPs\n",
      "   74 res   71                  19 x  19 x1024   ->    19 x  19 x1024\n",
      "   75 conv    512  1 x 1 / 1    19 x  19 x1024   ->    19 x  19 x 512  0.379 BFLOPs\n",
      "   76 conv   1024  3 x 3 / 1    19 x  19 x 512   ->    19 x  19 x1024  3.407 BFLOPs\n",
      "   77 conv    512  1 x 1 / 1    19 x  19 x1024   ->    19 x  19 x 512  0.379 BFLOPs\n",
      "   78 conv   1024  3 x 3 / 1    19 x  19 x 512   ->    19 x  19 x1024  3.407 BFLOPs\n",
      "   79 conv    512  1 x 1 / 1    19 x  19 x1024   ->    19 x  19 x 512  0.379 BFLOPs\n",
      "   80 conv   1024  3 x 3 / 1    19 x  19 x 512   ->    19 x  19 x1024  3.407 BFLOPs\n",
      "   81 conv     18  1 x 1 / 1    19 x  19 x1024   ->    19 x  19 x  18  0.013 BFLOPs\n",
      "   82 yolo\n",
      "   83 route  79\n",
      "   84 conv    256  1 x 1 / 1    19 x  19 x 512   ->    19 x  19 x 256  0.095 BFLOPs\n",
      "   85 upsample            2x    19 x  19 x 256   ->    38 x  38 x 256\n",
      "   86 route  85 61\n",
      "   87 conv    256  1 x 1 / 1    38 x  38 x 768   ->    38 x  38 x 256  0.568 BFLOPs\n",
      "   88 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   89 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   90 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   91 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   92 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   93 conv     18  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x  18  0.027 BFLOPs\n",
      "   94 yolo\n",
      "   95 route  91\n",
      "   96 conv    128  1 x 1 / 1    38 x  38 x 256   ->    38 x  38 x 128  0.095 BFLOPs\n",
      "   97 upsample            2x    38 x  38 x 128   ->    76 x  76 x 128\n",
      "   98 route  97 36\n",
      "   99 conv    128  1 x 1 / 1    76 x  76 x 384   ->    76 x  76 x 128  0.568 BFLOPs\n",
      "  100 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "  101 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "  102 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "  103 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "  104 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "  105 conv     18  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x  18  0.053 BFLOPs\n",
      "  106 yolo\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loading weights from ../backup/rsna_yolov3_15300.weights...Done!\r\n",
      "Learning Rate: 0.001, Momentum: 0.9, Decay: 0.0005\r\n",
      "Couldn't open file: data/coco_val_5k.list\r\n"
     ]
    }
   ],
   "source": [
    "!cd darknet && ./darknet detector recall ../cfg/rsna.data ../cfg/rsna_yolov3.cfg_test ../backup/rsna_yolov3_15300.weights -out \"\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "_uuid": "de9f7f2f30feeeaa3417ebd338c402e57a9b1f38"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "layer     filters    size              input                output\n",
      "    0 conv     32  3 x 3 / 1   608 x 608 x   3   ->   608 x 608 x  32  0.639 BFLOPs\n",
      "    1 conv     64  3 x 3 / 2   608 x 608 x  32   ->   304 x 304 x  64  3.407 BFLOPs\n",
      "    2 conv     32  1 x 1 / 1   304 x 304 x  64   ->   304 x 304 x  32  0.379 BFLOPs\n",
      "    3 conv     64  3 x 3 / 1   304 x 304 x  32   ->   304 x 304 x  64  3.407 BFLOPs\n",
      "    4 res    1                 304 x 304 x  64   ->   304 x 304 x  64\n",
      "    5 conv    128  3 x 3 / 2   304 x 304 x  64   ->   152 x 152 x 128  3.407 BFLOPs\n",
      "    6 conv     64  1 x 1 / 1   152 x 152 x 128   ->   152 x 152 x  64  0.379 BFLOPs\n",
      "    7 conv    128  3 x 3 / 1   152 x 152 x  64   ->   152 x 152 x 128  3.407 BFLOPs\n",
      "    8 res    5                 152 x 152 x 128   ->   152 x 152 x 128\n",
      "    9 conv     64  1 x 1 / 1   152 x 152 x 128   ->   152 x 152 x  64  0.379 BFLOPs\n",
      "   10 conv    128  3 x 3 / 1   152 x 152 x  64   ->   152 x 152 x 128  3.407 BFLOPs\n",
      "   11 res    8                 152 x 152 x 128   ->   152 x 152 x 128\n",
      "   12 conv    256  3 x 3 / 2   152 x 152 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "   13 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "   14 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "   15 res   12                  76 x  76 x 256   ->    76 x  76 x 256\n",
      "   16 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "   17 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "   18 res   15                  76 x  76 x 256   ->    76 x  76 x 256\n",
      "   19 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "   20 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "   21 res   18                  76 x  76 x 256   ->    76 x  76 x 256\n",
      "   22 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "   23 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "   24 res   21                  76 x  76 x 256   ->    76 x  76 x 256\n",
      "   25 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "   26 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "   27 res   24                  76 x  76 x 256   ->    76 x  76 x 256\n",
      "   28 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "   29 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "   30 res   27                  76 x  76 x 256   ->    76 x  76 x 256\n",
      "   31 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "   32 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "   33 res   30                  76 x  76 x 256   ->    76 x  76 x 256\n",
      "   34 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "   35 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "   36 res   33                  76 x  76 x 256   ->    76 x  76 x 256\n",
      "   37 conv    512  3 x 3 / 2    76 x  76 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   38 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   39 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   40 res   37                  38 x  38 x 512   ->    38 x  38 x 512\n",
      "   41 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   42 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   43 res   40                  38 x  38 x 512   ->    38 x  38 x 512\n",
      "   44 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   45 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   46 res   43                  38 x  38 x 512   ->    38 x  38 x 512\n",
      "   47 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   48 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   49 res   46                  38 x  38 x 512   ->    38 x  38 x 512\n",
      "   50 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   51 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   52 res   49                  38 x  38 x 512   ->    38 x  38 x 512\n",
      "   53 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   54 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   55 res   52                  38 x  38 x 512   ->    38 x  38 x 512\n",
      "   56 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   57 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   58 res   55                  38 x  38 x 512   ->    38 x  38 x 512\n",
      "   59 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   60 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   61 res   58                  38 x  38 x 512   ->    38 x  38 x 512\n",
      "   62 conv   1024  3 x 3 / 2    38 x  38 x 512   ->    19 x  19 x1024  3.407 BFLOPs\n",
      "   63 conv    512  1 x 1 / 1    19 x  19 x1024   ->    19 x  19 x 512  0.379 BFLOPs\n",
      "   64 conv   1024  3 x 3 / 1    19 x  19 x 512   ->    19 x  19 x1024  3.407 BFLOPs\n",
      "   65 res   62                  19 x  19 x1024   ->    19 x  19 x1024\n",
      "   66 conv    512  1 x 1 / 1    19 x  19 x1024   ->    19 x  19 x 512  0.379 BFLOPs\n",
      "   67 conv   1024  3 x 3 / 1    19 x  19 x 512   ->    19 x  19 x1024  3.407 BFLOPs\n",
      "   68 res   65                  19 x  19 x1024   ->    19 x  19 x1024\n",
      "   69 conv    512  1 x 1 / 1    19 x  19 x1024   ->    19 x  19 x 512  0.379 BFLOPs\n",
      "   70 conv   1024  3 x 3 / 1    19 x  19 x 512   ->    19 x  19 x1024  3.407 BFLOPs\n",
      "   71 res   68                  19 x  19 x1024   ->    19 x  19 x1024\n",
      "   72 conv    512  1 x 1 / 1    19 x  19 x1024   ->    19 x  19 x 512  0.379 BFLOPs\n",
      "   73 conv   1024  3 x 3 / 1    19 x  19 x 512   ->    19 x  19 x1024  3.407 BFLOPs\n",
      "   74 res   71                  19 x  19 x1024   ->    19 x  19 x1024\n",
      "   75 conv    512  1 x 1 / 1    19 x  19 x1024   ->    19 x  19 x 512  0.379 BFLOPs\n",
      "   76 conv   1024  3 x 3 / 1    19 x  19 x 512   ->    19 x  19 x1024  3.407 BFLOPs\n",
      "   77 conv    512  1 x 1 / 1    19 x  19 x1024   ->    19 x  19 x 512  0.379 BFLOPs\n",
      "   78 conv   1024  3 x 3 / 1    19 x  19 x 512   ->    19 x  19 x1024  3.407 BFLOPs\n",
      "   79 conv    512  1 x 1 / 1    19 x  19 x1024   ->    19 x  19 x 512  0.379 BFLOPs\n",
      "   80 conv   1024  3 x 3 / 1    19 x  19 x 512   ->    19 x  19 x1024  3.407 BFLOPs\n",
      "   81 conv     18  1 x 1 / 1    19 x  19 x1024   ->    19 x  19 x  18  0.013 BFLOPs\n",
      "   82 yolo\n",
      "   83 route  79\n",
      "   84 conv    256  1 x 1 / 1    19 x  19 x 512   ->    19 x  19 x 256  0.095 BFLOPs\n",
      "   85 upsample            2x    19 x  19 x 256   ->    38 x  38 x 256\n",
      "   86 route  85 61\n",
      "   87 conv    256  1 x 1 / 1    38 x  38 x 768   ->    38 x  38 x 256  0.568 BFLOPs\n",
      "   88 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   89 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   90 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   91 conv    256  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x 256  0.379 BFLOPs\n",
      "   92 conv    512  3 x 3 / 1    38 x  38 x 256   ->    38 x  38 x 512  3.407 BFLOPs\n",
      "   93 conv     18  1 x 1 / 1    38 x  38 x 512   ->    38 x  38 x  18  0.027 BFLOPs\n",
      "   94 yolo\n",
      "   95 route  91\n",
      "   96 conv    128  1 x 1 / 1    38 x  38 x 256   ->    38 x  38 x 128  0.095 BFLOPs\n",
      "   97 upsample            2x    38 x  38 x 128   ->    76 x  76 x 128\n",
      "   98 route  97 36\n",
      "   99 conv    128  1 x 1 / 1    76 x  76 x 384   ->    76 x  76 x 128  0.568 BFLOPs\n",
      "  100 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "  101 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "  102 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "  103 conv    128  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x 128  0.379 BFLOPs\n",
      "  104 conv    256  3 x 3 / 1    76 x  76 x 128   ->    76 x  76 x 256  3.407 BFLOPs\n",
      "  105 conv     18  1 x 1 / 1    76 x  76 x 256   ->    76 x  76 x  18  0.053 BFLOPs\n",
      "  106 yolo\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Loading weights from ../backup/rsna_yolov3_15300.weights...Done!\n",
      "../test.jpg: Predicted in 22.670948 seconds.\n",
      "pneumonia: 66%\n",
      "^C\n"
     ]
    }
   ],
   "source": [
    "!cd darknet && ./darknet detector test ../cfg/rsna.data ../cfg/rsna_yolov3.cfg_test ../backup/rsna_yolov3_15300.weights ../test.jpg -thresh 0.005"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "_uuid": "0dec1cc6f5bd1f8c854697128091ec7bd69d6e3e"
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7fe3a92f90f0>"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "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": [
    "# ![](predictions.jpg)\n",
    "plt.imshow(cv2.imread(\"./darknet/predictions.jpg\"))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "064029258018a3aa0ee22c59d74e6f321b767bf8"
   },
   "source": [
    "## 6. Generate Submission Files with YOLOv3 Python Wrapper"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "a4848803d5f41f1082de0b0a1e348a26f5d38d58"
   },
   "source": [
    "### 6.0. Download darknet python wrapper (darknet.py)\n",
    "* Basically, you can use darknet/python/darknet.py files. However it'll show error.\n",
    "* So, I edited the darknet.py. There are two main modifications.\n",
    "  * Change print statement to print function for python3\n",
    "  * Edit dynamic library('libdarknet.so') file path\n",
    "* I leaved '# ===' marks where i edited in darknet.py. For example,\n",
    "```\n",
    "# ==============================================================================\n",
    "#lib = CDLL(\"/home/pjreddie/documents/darknet/libdarknet.so\", RTLD_GLOBAL)\n",
    "darknet_lib_path = os.path.join(os.getcwd(), \"darknet\", \"libdarknet.so\")\n",
    "lib = CDLL(darknet_lib_path, RTLD_GLOBAL)\n",
    "# ==============================================================================\n",
    "```"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {
    "_uuid": "6cef00c59783cf8ff052b3c8b79903a9dfa24fc5"
   },
   "outputs": [],
   "source": [
    "!wget --no-check-certificate -q \"https://docs.google.com/uc?export=download&id=1-KTV7K9G1bl3SmnLnzmpkDyNt6tDmH7j\" -O darknet.py"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "91eafd163f9e78557c7acb1be7176845cc15d4c1"
   },
   "source": [
    "### 6.1. Load darknet python wrapper module"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {
    "_uuid": "238a70858c290db54f72700fe57215bd7934f6d8"
   },
   "outputs": [],
   "source": [
    "from darknet import *"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "1f6193268a5bd5b27d3ba855d6d36715a63d4ee0"
   },
   "source": [
    "### 6.2. Generate submission files\n",
    "* When making submission files, be aware of label format which is different in yolo."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "_uuid": "877c1dc79f4a4cb0dc3c322a05282d3d902ed615"
   },
   "outputs": [],
   "source": [
    "threshold = 0.2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {
    "_uuid": "b776bab5a37d180d2e5185555b47cafb1a964a71"
   },
   "outputs": [],
   "source": [
    "submit_file_path = \"submission.csv\"\n",
    "cfg_path = os.path.join(cfg_dir, \"rsna_yolov3.cfg_test\")\n",
    "weight_path = os.path.join(backup_dir, \"rsna_yolov3_15300.weights\")\n",
    "\n",
    "test_img_list_path = os.path.join(metadata_dir, \"te_list.txt\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "'/home/qinghe/cri/projects/AI/yolo/metadata/te_list.txt'"
      ]
     },
     "execution_count": 43,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_img_list_path"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {
    "_uuid": "d82a944502827338091976c4912f5eb6af6e886e"
   },
   "outputs": [],
   "source": [
    "gpu_index = 0\n",
    "net = load_net(cfg_path.encode(),\n",
    "               weight_path.encode(), \n",
    "               gpu_index)\n",
    "meta = load_meta(data_extention_file_path.encode())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "#\"/home/qinghe/cri/projects/AI/yolo/images/2fe3ba48-ce4d-42a4-8eda-99e3eb40d5fe.jpg\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "_uuid": "c11fa319a577beff0a0d53f107489ee79771bc22",
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "\n",
      "0it [00:00, ?it/s]\u001b[A\n",
      "1it [00:23, 23.93s/it]\u001b[A\n",
      "2it [00:47, 23.82s/it]\u001b[A\n",
      "3it [01:10, 23.69s/it]\u001b[A\n",
      "4it [01:33, 23.33s/it]\u001b[A\n",
      "5it [01:57, 23.57s/it]\u001b[A\n",
      "6it [02:20, 23.44s/it]\u001b[A\n",
      "7it [02:43, 23.29s/it]\u001b[A\n",
      "8it [03:05, 23.00s/it]\u001b[A\n",
      "9it [03:28, 22.77s/it]\u001b[A\n",
      "10it [03:50, 22.57s/it]\u001b[A\n",
      "11it [04:12, 22.42s/it]\u001b[A\n",
      "12it [04:34, 22.32s/it]\u001b[A\n",
      "13it [04:56, 22.36s/it]\u001b[A\n",
      "14it [05:18, 22.20s/it]\u001b[A\n",
      "15it [05:40, 22.10s/it]\u001b[A\n",
      "16it [06:02, 22.03s/it]\u001b[A\n",
      "17it [06:24, 22.15s/it]\u001b[A\n",
      "18it [06:50, 23.26s/it]\u001b[A\n",
      "19it [07:15, 23.63s/it]\u001b[A\n",
      "20it [07:38, 23.67s/it]\u001b[A\n",
      "21it [08:04, 24.25s/it]\u001b[A\n",
      "22it [08:30, 24.81s/it]\u001b[A\n",
      "23it [08:56, 25.23s/it]\u001b[A\n",
      "24it [09:23, 25.64s/it]\u001b[A\n",
      "25it [09:50, 26.05s/it]\u001b[A\n",
      "26it [10:15, 25.59s/it]\u001b[A\n",
      "27it [10:39, 25.39s/it]\u001b[A\n",
      "28it [11:05, 25.52s/it]\u001b[A\n",
      "29it [11:32, 25.75s/it]\u001b[A\n",
      "30it [11:57, 25.54s/it]\u001b[A\n",
      "31it [12:19, 24.70s/it]\u001b[A\n",
      "32it [12:41, 23.81s/it]\u001b[A\n",
      "33it [13:03, 23.28s/it]\u001b[A\n",
      "34it [13:26, 23.17s/it]\u001b[A\n",
      "35it [13:49, 23.19s/it]\u001b[A\n",
      "36it [14:12, 22.93s/it]\u001b[A\n",
      "37it [14:34, 22.84s/it]\u001b[A\n",
      "38it [14:57, 22.76s/it]\u001b[A\n",
      "39it [15:18, 22.42s/it]\u001b[A\n",
      "40it [15:41, 22.51s/it]\u001b[A\n",
      "41it [16:03, 22.24s/it]\u001b[A\n",
      "42it [16:24, 22.07s/it]\u001b[A\n",
      "43it [16:46, 21.95s/it]\u001b[A\n",
      "44it [17:08, 21.88s/it]\u001b[A\n",
      "45it [17:31, 22.25s/it]\u001b[A\n",
      "46it [17:54, 22.38s/it]\u001b[A\n",
      "47it [18:15, 22.19s/it]\u001b[A\n",
      "48it [18:37, 21.98s/it]\u001b[A\n",
      "49it [18:58, 21.88s/it]\u001b[A\n",
      "50it [19:20, 21.81s/it]\u001b[A\n",
      "51it [19:42, 21.77s/it]\u001b[A\n",
      "52it [20:05, 22.06s/it]\u001b[A\n",
      "53it [20:27, 22.13s/it]\u001b[A\n",
      "54it [20:49, 22.11s/it]\u001b[A\n",
      "55it [21:11, 22.10s/it]\u001b[A\n",
      "56it [21:33, 22.14s/it]\u001b[A\n",
      "57it [21:55, 22.09s/it]\u001b[A\n",
      "58it [22:17, 22.15s/it]\u001b[A\n",
      "59it [22:39, 22.09s/it]\u001b[A\n",
      "60it [23:01, 22.08s/it]\u001b[A\n",
      "61it [23:23, 22.03s/it]\u001b[A\n",
      "62it [23:45, 22.01s/it]\u001b[A\n",
      "63it [24:07, 21.99s/it]\u001b[A\n",
      "64it [24:29, 21.96s/it]\u001b[A\n",
      "65it [24:51, 22.00s/it]\u001b[A\n",
      "66it [25:15, 22.56s/it]\u001b[A\n",
      "67it [25:37, 22.34s/it]\u001b[A\n",
      "68it [26:02, 23.26s/it]\u001b[A\n",
      "69it [26:25, 22.93s/it]\u001b[A\n",
      "70it [26:47, 22.68s/it]\u001b[A\n",
      "71it [27:09, 22.48s/it]\u001b[A\n",
      "72it [27:37, 24.23s/it]\u001b[A"
     ]
    },
    {
     "ename": "KeyboardInterrupt",
     "evalue": "",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
      "\u001b[0;32m<ipython-input-44-f7a8a390101f>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m\u001b[0m\n\u001b[1;32m      6\u001b[0m         \u001b[0mpatient_id\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mline\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstrip\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msplit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'/'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m-\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstrip\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msplit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m'.'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m      7\u001b[0m         \u001b[0mpatient_id\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m----> 8\u001b[0;31m         \u001b[0minfer_result\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdetect\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mnet\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mmeta\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mline\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstrip\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mencode\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mthresh\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mthreshold\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m      9\u001b[0m         \u001b[0minfer_result\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     10\u001b[0m         \u001b[0msubmit_line\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m\"\"\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;32m~/cri/projects/AI/yolo/darknet.py\u001b[0m in \u001b[0;36mdetect\u001b[0;34m(net, meta, image, thresh, hier_thresh, nms)\u001b[0m\n\u001b[1;32m    136\u001b[0m     \u001b[0mnum\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mc_int\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    137\u001b[0m     \u001b[0mpnum\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mpointer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mnum\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 138\u001b[0;31m     \u001b[0mpredict_image\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mnet\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mim\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    139\u001b[0m     \u001b[0mdets\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mget_network_boxes\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mnet\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mim\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mw\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mim\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mh\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mthresh\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mhier_thresh\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;32mNone\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0;36m0\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mpnum\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    140\u001b[0m     \u001b[0mnum\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mpnum\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;36m0\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
      "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
     ]
    }
   ],
   "source": [
    "submit_dict = {\"patientId\": [], \"PredictionString\": []}\n",
    "\n",
    "with open(test_img_list_path, \"r\") as test_img_list_f:\n",
    "    # tqdm run up to 1000(The # of test set)\n",
    "    for line in tqdm(test_img_list_f):\n",
    "        patient_id = line.strip().split('/')[-1].strip().split('.')[0]\n",
    "        patient_id\n",
    "        infer_result = detect(net, meta, line.strip().encode(), thresh=threshold)\n",
    "        infer_result\n",
    "        submit_line = \"\"\n",
    "        for e in infer_result:\n",
    "            confi = e[1]\n",
    "            w = e[2][2]\n",
    "            h = e[2][3]\n",
    "            x = e[2][0]-w/2\n",
    "            y = e[2][1]-h/2\n",
    "            submit_line += \"{} {} {} {} {} \".format(confi, x, y, w, h)\n",
    "\n",
    "        submit_dict[\"patientId\"].append(patient_id)\n",
    "        submit_dict[\"PredictionString\"].append(submit_line)\n",
    "\n",
    "pd.DataFrame(submit_dict).to_csv(submit_file_path, index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {},
   "outputs": [],
   "source": [
    "pd.DataFrame(submit_dict).to_csv(submit_file_path, index=False)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "_uuid": "7ffbb3e3205f76f5b8f216266539469c65d56cc1"
   },
   "outputs": [],
   "source": [
    "# !ls -lsht\n",
    "!rm -rf darknet images labels metadata backup cfg\n",
    "!rm -rf train_log.txt darknet53.conv.74 darknet.py darknet_gpu\n",
    "!rm -rf test.jpg\n",
    "!rm -rf __pycache__ .ipynb_checkpoints"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "_uuid": "74cd76e2b853568a7f2962f0c220a64fe87e67c1"
   },
   "outputs": [],
   "source": [
    "!ls -alsht"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {
    "_uuid": "2fbf050ee377907eb544c04382d07e1449ef577a"
   },
   "source": [
    "## 7. Future works & Etc\n",
    "\n",
    "### Future works (Things to try)\n",
    "* Image augmentation\n",
    "* More training\n",
    "* Utilizing the not labeled images because we got rid of not labeled images above\n",
    "\n",
    "### ETC\n",
    "* For a private matter, i can not proceed RSNA task after 09/27. If you have any ideas, questions and problems with this kernel after 09/27, Please leave those things anyway~! Collaborator '@John Byun' will reply to your comments."
   ]
  }
 ],
 "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": "3.6.3"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {
    "height": "calc(100% - 180px)",
    "left": "10px",
    "top": "150px",
    "width": "296.475px"
   },
   "toc_section_display": true,
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 1
}