Gproject6 (1).ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# import packages"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from Bio.Seq import Seq"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Read the file whose contain the index of splicing "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "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>#NAME</th>\n",
       "      <th>CHROM</th>\n",
       "      <th>STRAND</th>\n",
       "      <th>TX_START</th>\n",
       "      <th>TX_END</th>\n",
       "      <th>EXON_START</th>\n",
       "      <th>EXON_END</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>OR4F5</td>\n",
       "      <td>1</td>\n",
       "      <td>+</td>\n",
       "      <td>69090</td>\n",
       "      <td>70008</td>\n",
       "      <td>69090,</td>\n",
       "      <td>70008,</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>OR4F16</td>\n",
       "      <td>1</td>\n",
       "      <td>-</td>\n",
       "      <td>685715</td>\n",
       "      <td>686654</td>\n",
       "      <td>685715,</td>\n",
       "      <td>686654,</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>SAMD11</td>\n",
       "      <td>1</td>\n",
       "      <td>+</td>\n",
       "      <td>925737</td>\n",
       "      <td>944575</td>\n",
       "      <td>925737,925921,930154,931038,935771,939039,9392...</td>\n",
       "      <td>925800,926013,930336,931089,935896,939129,9394...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>NOC2L</td>\n",
       "      <td>1</td>\n",
       "      <td>-</td>\n",
       "      <td>944203</td>\n",
       "      <td>959290</td>\n",
       "      <td>944203,945056,945517,946172,946401,948130,9484...</td>\n",
       "      <td>944800,945146,945653,946286,946545,948232,9486...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>KLHL17</td>\n",
       "      <td>1</td>\n",
       "      <td>+</td>\n",
       "      <td>960586</td>\n",
       "      <td>965715</td>\n",
       "      <td>960586,961292,961628,961825,962354,962703,9631...</td>\n",
       "      <td>960800,961552,961750,962047,962471,962917,9632...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19300</th>\n",
       "      <td>BPY2B</td>\n",
       "      <td>Y</td>\n",
       "      <td>+</td>\n",
       "      <td>24618003</td>\n",
       "      <td>24639207</td>\n",
       "      <td>24618003,24618309,24621497,24626043,24628221,2...</td>\n",
       "      <td>24618034,24618432,24621634,24626162,24628339,2...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19301</th>\n",
       "      <td>DAZ3</td>\n",
       "      <td>Y</td>\n",
       "      <td>-</td>\n",
       "      <td>24763068</td>\n",
       "      <td>24813479</td>\n",
       "      <td>24763068,24768898,24773569,24774349,24784136,2...</td>\n",
       "      <td>24764925,24768997,24773604,24774421,24784208,2...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19302</th>\n",
       "      <td>DAZ4</td>\n",
       "      <td>Y</td>\n",
       "      <td>+</td>\n",
       "      <td>24834126</td>\n",
       "      <td>24905240</td>\n",
       "      <td>24834126,24840730,24841151,24841803,24842291,2...</td>\n",
       "      <td>24834129,24840877,24841243,24841855,24842355,2...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19303</th>\n",
       "      <td>BPY2C</td>\n",
       "      <td>Y</td>\n",
       "      <td>-</td>\n",
       "      <td>25030900</td>\n",
       "      <td>25052104</td>\n",
       "      <td>25030900,25031316,25037991,25038808,25041768,2...</td>\n",
       "      <td>25031222,25031441,25038116,25038914,25041886,2...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19304</th>\n",
       "      <td>CDY1</td>\n",
       "      <td>Y</td>\n",
       "      <td>+</td>\n",
       "      <td>25622161</td>\n",
       "      <td>25624338</td>\n",
       "      <td>25622161,</td>\n",
       "      <td>25624338,</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>19305 rows × 7 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "        #NAME CHROM STRAND  TX_START    TX_END  \\\n",
       "0       OR4F5     1      +     69090     70008   \n",
       "1      OR4F16     1      -    685715    686654   \n",
       "2      SAMD11     1      +    925737    944575   \n",
       "3       NOC2L     1      -    944203    959290   \n",
       "4      KLHL17     1      +    960586    965715   \n",
       "...       ...   ...    ...       ...       ...   \n",
       "19300   BPY2B     Y      +  24618003  24639207   \n",
       "19301    DAZ3     Y      -  24763068  24813479   \n",
       "19302    DAZ4     Y      +  24834126  24905240   \n",
       "19303   BPY2C     Y      -  25030900  25052104   \n",
       "19304    CDY1     Y      +  25622161  25624338   \n",
       "\n",
       "                                              EXON_START  \\\n",
       "0                                                 69090,   \n",
       "1                                                685715,   \n",
       "2      925737,925921,930154,931038,935771,939039,9392...   \n",
       "3      944203,945056,945517,946172,946401,948130,9484...   \n",
       "4      960586,961292,961628,961825,962354,962703,9631...   \n",
       "...                                                  ...   \n",
       "19300  24618003,24618309,24621497,24626043,24628221,2...   \n",
       "19301  24763068,24768898,24773569,24774349,24784136,2...   \n",
       "19302  24834126,24840730,24841151,24841803,24842291,2...   \n",
       "19303  25030900,25031316,25037991,25038808,25041768,2...   \n",
       "19304                                          25622161,   \n",
       "\n",
       "                                                EXON_END  \n",
       "0                                                 70008,  \n",
       "1                                                686654,  \n",
       "2      925800,926013,930336,931089,935896,939129,9394...  \n",
       "3      944800,945146,945653,946286,946545,948232,9486...  \n",
       "4      960800,961552,961750,962047,962471,962917,9632...  \n",
       "...                                                  ...  \n",
       "19300  24618034,24618432,24621634,24626162,24628339,2...  \n",
       "19301  24764925,24768997,24773604,24774421,24784208,2...  \n",
       "19302  24834129,24840877,24841243,24841855,24842355,2...  \n",
       "19303  25031222,25031441,25038116,25038914,25041886,2...  \n",
       "19304                                          25624338,  \n",
       "\n",
       "[19305 rows x 7 columns]"
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "index_file=pd.read_table(\"grch38.txt\")\n",
    "index_file"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Read the chromosomes files whose contain the sequence of introns and exons"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "for n in range(1,23): \n",
    "    locals()[\"ch\"+str(n)] = Seq(\"\".join([l.strip() for l in open(\"Homo_sapiens.GRCh38.dna.chromosome.%i.fa\"%n).readlines()[1:]]))\n",
    "chx=Seq(\"\".join([l.strip() for l in open(\"Homo_sapiens.GRCh38.dna.chromosome.X.fa\").readlines()[1:]])) \n",
    "chy=Seq(\"\".join([l.strip() for l in open(\"Homo_sapiens.GRCh38.dna.chromosome.Y.fa\").readlines()[1:]]))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# make reverce complement for these chromosomes"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "for n in range(1,23): \n",
    "    locals()[\"ch\"+str(n)+\"_rev\"]=locals()[\"ch\"+str(n)].reverse_complement()\n",
    "chx_rev=chx.reverse_complement()\n",
    "chy_rev=chy.reverse_complement()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# make function to extract sequences and identify class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "seq_splice=[]\n",
    "class_type=[]\n",
    "def splice(seq,indexs):\n",
    "    for x,y in indexs:\n",
    "         seq_splice.append(seq[x:y])\n",
    "count=1             \n",
    "for i in range(len(index_file)):\n",
    "    #for n in range(1,23):\n",
    "     #if index_file['CHROM'][i]=='%s'%i or index_file['CHROM'][i]=='X'or index_file['CHROM'][i]=='Y':\n",
    "    if index_file['CHROM'][i]=='%s'%count:\n",
    "        E=index_file['EXON_END'][i].split(\",\")\n",
    "        E.remove(E[-1])\n",
    "           \n",
    "        s=index_file['EXON_START'][i].split(\",\")\n",
    "        s.remove(s[-1])\n",
    "            \n",
    "        lis_EXON_START=[]\n",
    "        lis_EXON_END=[]\n",
    "            \n",
    "        for j in E :\n",
    "            lis_EXON_END.append(int(j)) \n",
    "        for k in s :\n",
    "            lis_EXON_START.append(int(k))\n",
    "        for m in lis_EXON_START :\n",
    "            x=int((lis_EXON_START[0]+lis_EXON_END[0])/2)\n",
    "            if class_type.count('Neither')<=177723:\n",
    "                if index_file['STRAND'][i]=='+': \n",
    "                    splice(locals()[\"ch\"+str(count)],[(x-200,x+201)])\n",
    "                    class_type.append(\"Neither\")\n",
    "        #elif index_file['STRAND'][i]=='-':\n",
    "                else :  \n",
    "                    splice(locals()[\"ch\"+str(count)+\"_rev\"],[(x-200,x+201)])\n",
    "                    class_type.append(\"Neither\")\n",
    "            else:\n",
    "                break\n",
    "                   \n",
    "        for r in range(len(lis_EXON_START)):\n",
    "            if lis_EXON_START[r]==index_file['TX_START'][i] and lis_EXON_END[r]==index_file['TX_END'][i]:\n",
    "                continue   \n",
    "            else :\n",
    "                if index_file['STRAND'][i]=='+':\n",
    "                    splice(locals()[\"ch\"+str(count)],[(lis_EXON_START[r]-200,lis_EXON_START[r]+201)])\n",
    "                    class_type.append(\"Acceptor\")\n",
    "                    splice(locals()[\"ch\"+str(count)],[(lis_EXON_END[r]-200,lis_EXON_END[r]+201)])\n",
    "                    class_type.append(\"Doner\")\n",
    "                                  \n",
    "                else:\n",
    "                        #if index_file['STRAND'][i]=='-': \n",
    "                    splice(locals()[\"ch\"+str(count)+\"_rev\"],[(lis_EXON_START[r]-200,lis_EXON_START[r]+201)])\n",
    "                    class_type.append(\"Acceptor\")\n",
    "                    splice(locals()[\"ch\"+str(count)+\"_rev\"],[(lis_EXON_END[r]-200,lis_EXON_END[r]+201)])\n",
    "                    class_type.append(\"Doner\")\n",
    "                        \n",
    "    elif index_file['CHROM'][i]==\"X\" or index_file['CHROM'][i]==\"Y\":\n",
    "        E=index_file['EXON_END'][i].split(\",\")\n",
    "        E.remove(E[-1])\n",
    "        s=index_file['EXON_START'][i].split(\",\")   \n",
    "        s.remove(s[-1])  \n",
    "        lis_EXON_START=[]   \n",
    "        lis_EXON_END=[]\n",
    "        for j in E :\n",
    "            lis_EXON_END.append(int(j))\n",
    "        for k in s :    \n",
    "             lis_EXON_START.append(int(k))\n",
    "        for m in lis_EXON_START :\n",
    "            x=int((lis_EXON_START[0]+lis_EXON_END[0])/2)\n",
    "            if class_type.count('Neither')<=177723: \n",
    "                if index_file['STRAND'][i]=='+':\n",
    "                    if index_file['CHROM'][i]=='X':\n",
    "                        splice(\"chx\",[(x-200,x+201)])\n",
    "                    else:\n",
    "                        splice(\"chy\",[(x-200,x+201)])\n",
    "                        class_type.append(\"Neither\")\n",
    "                else :\n",
    "                    if index_file['CHROM'][i]=='X':\n",
    "                        splice(locals()[\"chx_rev\"],[(x-200,x+201)])\n",
    "                    else :\n",
    "                        splice(locals()[\"chy_rev\"],[(x-200,x+201)])\n",
    "                        class_type.append(\"Neither\") \n",
    "            else:\n",
    "                break\n",
    "        for r in range(len(lis_EXON_START)):\n",
    "            if lis_EXON_START[r]==index_file['TX_START'][i] and lis_EXON_END[r]==index_file['TX_END'][i]:\n",
    "                continue   \n",
    "            else :\n",
    "                if index_file['STRAND'][i]=='+':\n",
    "                    if index_file['CHROM'][i]=='X':\n",
    "                        splice(\"chx\",[(lis_EXON_START[r]-200,lis_EXON_START[r]+201)])\n",
    "                        class_type.append(\"Acceptor\")\n",
    "                        splice(\"chx\",[(lis_EXON_END[r]-200,lis_EXON_END[r]+201)])\n",
    "                        class_type.append(\"Doner\") \n",
    "                    else : \n",
    "                        splice(\"chy\",[(lis_EXON_START[r]-200,lis_EXON_START[r]+201)])\n",
    "                        class_type.append(\"Acceptor\")\n",
    "                        splice(\"chy\",[(lis_EXON_END[r]-200,lis_EXON_END[r]+201)])\n",
    "                        class_type.append(\"Doner\")        \n",
    "    else :\n",
    "        count=count+1\n",
    "        "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# show the length of list of sequences and class"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "the list sequences length is : 563604\n",
      "the list class length is : 563604\n"
     ]
    }
   ],
   "source": [
    "print(\"the list sequences length is :\",len(seq_splice))\n",
    "print(\"the list class length is :\",len( class_type))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Acceptor =  192940\n",
      "Doner =  192940\n",
      "Neither =  177724\n"
     ]
    }
   ],
   "source": [
    "Acceptor=0\n",
    "Doner=0\n",
    "Neither=0\n",
    "for i in class_type:\n",
    "    if i==\"Acceptor\":\n",
    "        Acceptor=Acceptor+1\n",
    "    elif i==\"Doner\" :\n",
    "        Doner=Doner+1\n",
    "    else:\n",
    "        Neither=Neither+1\n",
    "        #print(class_type.index(i))\n",
    "        \n",
    "print(\"Acceptor = \",Acceptor)\n",
    "print(\"Doner = \",Doner)\n",
    "print(\"Neither = \",Neither)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Convert sequences to binary matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "for j in range(563604):\n",
    "    locals()[\"x\"+str(j)]=np.empty((0,5),dtype=np.float32)\n",
    "    for i in seq_splice[j]:\n",
    "        if i=='A':\n",
    "            new_row=np.array([[1,0,0,0,0]],dtype=np.float32) \n",
    "        elif i=='T':\n",
    "            new_row=np.array([[0,1,0,0,0]],dtype=np.float32)\n",
    "        elif i=='C':\n",
    "            new_row=np.array([[0,0,1,0,0]],dtype=np.float32)\n",
    "        elif i=='G':\n",
    "            new_row=np.array([[0,0,0,1,0]],dtype=np.float32)\n",
    "        else :\n",
    "            new_row=np.array([[0,0,0,0,1]],dtype=np.float32)\n",
    "        \n",
    "        locals()[\"x\"+str(j)]=np.concatenate((locals()[\"x\"+str(j)],new_row))     "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# show one binary matrix from sequence"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[0. 1. 0. 0. 0.]\n",
      " [0. 0. 1. 0. 0.]\n",
      " [1. 0. 0. 0. 0.]\n",
      " ...\n",
      " [0. 0. 1. 0. 0.]\n",
      " [1. 0. 0. 0. 0.]\n",
      " [0. 1. 0. 0. 0.]]\n"
     ]
    }
   ],
   "source": [
    "print(x0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# make matrix of zeros"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[[0. 0. 0. 0. 0.]\n",
      "  [0. 0. 0. 0. 0.]\n",
      "  [0. 0. 0. 0. 0.]\n",
      "  ...\n",
      "  [0. 0. 0. 0. 0.]\n",
      "  [0. 0. 0. 0. 0.]\n",
      "  [0. 0. 0. 0. 0.]]\n",
      "\n",
      " [[0. 0. 0. 0. 0.]\n",
      "  [0. 0. 0. 0. 0.]\n",
      "  [0. 0. 0. 0. 0.]\n",
      "  ...\n",
      "  [0. 0. 0. 0. 0.]\n",
      "  [0. 0. 0. 0. 0.]\n",
      "  [0. 0. 0. 0. 0.]]\n",
      "\n",
      " [[0. 0. 0. 0. 0.]\n",
      "  [0. 0. 0. 0. 0.]\n",
      "  [0. 0. 0. 0. 0.]\n",
      "  ...\n",
      "  [0. 0. 0. 0. 0.]\n",
      "  [0. 0. 0. 0. 0.]\n",
      "  [0. 0. 0. 0. 0.]]\n",
      "\n",
      " ...\n",
      "\n",
      " [[0. 0. 0. 0. 0.]\n",
      "  [0. 0. 0. 0. 0.]\n",
      "  [0. 0. 0. 0. 0.]\n",
      "  ...\n",
      "  [0. 0. 0. 0. 0.]\n",
      "  [0. 0. 0. 0. 0.]\n",
      "  [0. 0. 0. 0. 0.]]\n",
      "\n",
      " [[0. 0. 0. 0. 0.]\n",
      "  [0. 0. 0. 0. 0.]\n",
      "  [0. 0. 0. 0. 0.]\n",
      "  ...\n",
      "  [0. 0. 0. 0. 0.]\n",
      "  [0. 0. 0. 0. 0.]\n",
      "  [0. 0. 0. 0. 0.]]\n",
      "\n",
      " [[0. 0. 0. 0. 0.]\n",
      "  [0. 0. 0. 0. 0.]\n",
      "  [0. 0. 0. 0. 0.]\n",
      "  ...\n",
      "  [0. 0. 0. 0. 0.]\n",
      "  [0. 0. 0. 0. 0.]\n",
      "  [0. 0. 0. 0. 0.]]]\n"
     ]
    }
   ],
   "source": [
    "z=np.zeros((563604,401,5),dtype=np.float32)\n",
    "print(z)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# collect matrixs of sequences in one matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "for i in range(563604):\n",
    "    z[i,:,:]=locals()[\"x\"+str(i)] "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Convert class to binary matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "class_type_matrix=np.empty((0,3),dtype=np.float32)\n",
    "for i in range(len(class_type)):  \n",
    "    if class_type[i]=='Acceptor':\n",
    "        new_row=np.array([[1,0,0]],dtype=np.float32)\n",
    "    elif class_type[i]=='Doner':\n",
    "        new_row=np.array([[0,1,0]],dtype=np.float32)\n",
    "    else :\n",
    "        new_row=np.array([[0,0,1]],dtype=np.float32)\n",
    "    class_type_matrix = np.concatenate((class_type_matrix,new_row))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# make shuffle to data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "from sklearn.utils import shuffle\n",
    "z,class_type_matrix = shuffle(z,class_type_matrix, random_state=0)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# store data in file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [],
   "source": [
    "z.tofile('GProject1')\n",
    "class_type_matrix.tofile('GProject2')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Read data from file"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "data=np.fromfile('GProject1',dtype=np.float32)\n",
    "label=np.fromfile('GProject2',dtype=np.float32)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Split data to Train and Test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "data_train=data.reshape(563604,401,5)[0:338162]\n",
    "data_test=data.reshape(563604,401,5)[338162:563604]\n",
    "lable_train=label.reshape(563604,3)[0:338162]\n",
    "lable_test=label.reshape(563604,3)[338162:563604]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "338162"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(data_train)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Build CNN"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras.models import Sequential\n",
    "from keras.layers import Conv2D\n",
    "from keras.layers import MaxPooling2D\n",
    "from keras.layers import Flatten\n",
    "from keras.layers import Dense"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "model=Sequential()\n",
    "model.add(Conv2D(filters=32,kernel_size=2,padding=\"same\",input_shape=(401,5,1),activation =\"relu\"))\n",
    "model.add(MaxPooling2D(pool_size=2))\n",
    "model.add(Conv2D(filters=32,kernel_size=2,padding=\"same\",activation=\"relu\"))\n",
    "model.add(MaxPooling2D(pool_size=2))\n",
    "model.add(Flatten())\n",
    "model.add(Dense(10,activation=\"relu\"))\n",
    "model.add(Dense(3,activation=\"relu\"))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(optimizer='adam',loss='binary_crossentropy', metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      " Layer (type)                Output Shape              Param #   \n",
      "=================================================================\n",
      " conv2d (Conv2D)             (None, 401, 5, 32)        160       \n",
      "                                                                 \n",
      " max_pooling2d (MaxPooling2D  (None, 200, 2, 32)       0         \n",
      " )                                                               \n",
      "                                                                 \n",
      " conv2d_1 (Conv2D)           (None, 200, 2, 32)        4128      \n",
      "                                                                 \n",
      " max_pooling2d_1 (MaxPooling  (None, 100, 1, 32)       0         \n",
      " 2D)                                                             \n",
      "                                                                 \n",
      " flatten (Flatten)           (None, 3200)              0         \n",
      "                                                                 \n",
      " dense (Dense)               (None, 10)                32010     \n",
      "                                                                 \n",
      " dense_1 (Dense)             (None, 3)                 33        \n",
      "                                                                 \n",
      "=================================================================\n",
      "Total params: 36,331\n",
      "Trainable params: 36,331\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Trainning the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "10568/10568 [==============================] - 196s 18ms/step - loss: 0.6834 - accuracy: 0.5209\n",
      "Epoch 2/10\n",
      "10568/10568 [==============================] - 197s 19ms/step - loss: 0.5527 - accuracy: 0.5640\n",
      "Epoch 3/10\n",
      "10568/10568 [==============================] - 195s 18ms/step - loss: 0.5730 - accuracy: 0.5378\n",
      "Epoch 4/10\n",
      "10568/10568 [==============================] - 200s 19ms/step - loss: 0.5408 - accuracy: 0.5681\n",
      "Epoch 5/10\n",
      "10568/10568 [==============================] - 203s 19ms/step - loss: 0.5209 - accuracy: 0.5826\n",
      "Epoch 6/10\n",
      "10568/10568 [==============================] - 204s 19ms/step - loss: 0.5340 - accuracy: 0.5661\n",
      "Epoch 7/10\n",
      "10568/10568 [==============================] - 205s 19ms/step - loss: 0.5216 - accuracy: 0.5754\n",
      "Epoch 8/10\n",
      "10568/10568 [==============================] - 190s 18ms/step - loss: 0.5126 - accuracy: 0.5875\n",
      "Epoch 9/10\n",
      "10568/10568 [==============================] - 204s 19ms/step - loss: 0.5084 - accuracy: 0.5850\n",
      "Epoch 10/10\n",
      "10568/10568 [==============================] - 208s 20ms/step - loss: 0.5120 - accuracy: 0.5798\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<keras.callbacks.History at 0x2b3658583d0>"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.fit(data_train,lable_train,shuffle=True,epochs=10)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# save model weights"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 79,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.save_weights('CNN')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# load model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.training.tracking.util.CheckpointLoadStatus at 0x183c8620c10>"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.load_weights('CNN')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Testing the model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 69,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "7046/7046 [==============================] - 28s 4ms/step - loss: 0.5253 - accuracy: 0.5205\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "[0.5252529978752136, 0.5205374360084534]"
      ]
     },
     "execution_count": 69,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "model.evaluate(data_test,lable_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 70,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(1000, 3)\n"
     ]
    }
   ],
   "source": [
    "predictions=model.predict(data_test[0:1000])\n",
    "print(predictions.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 71,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[0., 1., 0.],\n",
       "       [1., 0., 0.],\n",
       "       [0., 0., 0.],\n",
       "       ...,\n",
       "       [0., 0., 0.],\n",
       "       [1., 0., 0.],\n",
       "       [0., 0., 0.]], dtype=float32)"
      ]
     },
     "execution_count": 71,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "np.round(predictions)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Visualizing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1/10\n",
      "10568/10568 [==============================] - 210s 20ms/step - loss: 0.5291 - accuracy: 0.5512 - val_loss: 0.5845 - val_accuracy: 0.4469\n",
      "Epoch 2/10\n",
      "10568/10568 [==============================] - 213s 20ms/step - loss: 0.5079 - accuracy: 0.5902 - val_loss: 0.4940 - val_accuracy: 0.6056\n",
      "Epoch 3/10\n",
      "10568/10568 [==============================] - 216s 20ms/step - loss: 0.6083 - accuracy: 0.4216 - val_loss: 0.5677 - val_accuracy: 0.5337\n",
      "Epoch 4/10\n",
      "10568/10568 [==============================] - 213s 20ms/step - loss: 0.6209 - accuracy: 0.3853 - val_loss: 0.6289 - val_accuracy: 0.3977\n",
      "Epoch 5/10\n",
      "10568/10568 [==============================] - 211s 20ms/step - loss: 0.5851 - accuracy: 0.4783 - val_loss: 0.6052 - val_accuracy: 0.4321\n",
      "Epoch 6/10\n",
      "10568/10568 [==============================] - 216s 20ms/step - loss: 0.6081 - accuracy: 0.4334 - val_loss: 0.5790 - val_accuracy: 0.5289\n",
      "Epoch 7/10\n",
      "10568/10568 [==============================] - 220s 21ms/step - loss: 0.9135 - accuracy: 0.5101 - val_loss: 3.9058 - val_accuracy: 0.3155\n",
      "Epoch 8/10\n",
      "10568/10568 [==============================] - 222s 21ms/step - loss: 1.1912 - accuracy: 0.4707 - val_loss: 0.5282 - val_accuracy: 0.5863\n",
      "Epoch 9/10\n",
      "10568/10568 [==============================] - 226s 21ms/step - loss: 0.5551 - accuracy: 0.5274 - val_loss: 0.5989 - val_accuracy: 0.5004\n",
      "Epoch 10/10\n",
      "10568/10568 [==============================] - 234s 22ms/step - loss: 0.5535 - accuracy: 0.5282 - val_loss: 0.6166 - val_accuracy: 0.4372\n"
     ]
    }
   ],
   "source": [
    "history = model.fit(data_train,lable_train,epochs = 10 , validation_data = (data_test, lable_test))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAA3AAAAGiCAYAAACmgyTaAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8vihELAAAACXBIWXMAAAsTAAALEwEAmpwYAAC2QUlEQVR4nOzdd3yV5fn48c+dvU8gCSMnQBhBIAmEqYIDR62rQh2tW7St1Vq1W9uv1i5/2m/bb6vW1tpq3VIHIu5W6wQHJARIGLJCFiMBcjLIPvfvj+c8IYaMc3LGc8b1fr3yOsk5z7gSQp5zPfd9X5fSWiOEEEIIIYQQIvhFWR2AEEIIIYQQQgj3SAInhBBCCCGEECFCEjghhBBCCCGECBGSwAkhhBBCCCFEiJAETgghhBBCCCFChCRwQgghhBBCCBEiJIETPqOUekMpdY2vt7WSUqpCKXWmH477nlLqm67Pr1BK/dudbYdxnvFKqWalVPRwYxVCCGEdubZ6dFy5toqIIAlchHP9ATI/nEqp1l5fX+HJsbTW52itH/f1tsFIKfVTpdQH/TyfqZTqUEoVuHssrfXTWuuzfBTXFy6KWutKrXWK1rrbF8fv53xKKbVLKbXZH8cXQohQJNfW4ZFrKyiltFJqiq+PK8KLJHARzvUHKEVrnQJUAl/p9dzT5nZKqRjrogxKTwILlVIT+zx/KbBJa11mQUxWOAUYBUxSSs0P5Inld1IIEazk2jpscm0Vwg2SwIl+KaUWK6WqlVK3KaX2Af9USo1QSr2qlKpTSh12fZ7Ta5/eUxeWKaU+Ukr93rXtbqXUOcPcdqJS6gOlVJNS6m2l1INKqacGiNudGH+tlFrtOt6/lVKZvV6/Sim1Ryl1UCn1PwP9fLTW1cB/gav6vHQ18PhQcfSJeZlS6qNeX39JKbVVKeVQSv0ZUL1em6yU+q8rvnql1NNKqXTXa08C44FXXHd5f6KUynXdzYtxbZOtlFqllDqklNqhlPpWr2P/Qin1nFLqCdfPplwpNW+gn4HLNcDLwOuuz3t/X/lKqf+4zrVfKfUz1/PRSqmfKaV2us5TrJQa1zdW17Z9f09WK6X+qJQ6BPxisJ+Ha59xSqkVrn+Hg0qpPyul4l0xFfbabpQy7pBnDfH9CiHEsMm1Va6tbl5b+/t+bK5j1Ll+lncopaJcr01RSr3v+t7qlVL/cj2vXNfMA67XNioPRjFF8JIETgxmDDASmABcj/H78k/X1+OBVuDPg+x/PLANyAT+F3hEKaWGse0zwGdABvALjv3D3ps7MV4OXIsxchQH/AhAKTUD+Kvr+Nmu8/V7YXB5vHcsSqnjgCLgWTfjOIbrgvcicAfGz2InsKj3JsA9rvimA+MwfiZora/ii3d6/7efUzwLVLv2vxj4f0qpM3q9fgGwHEgHVg0Ws1IqyXWMp10flyql4lyvpQJvA2+6zjUFeMe16w+Ay4BzgTTgOuDIYD+XXo4HdmH8293NID8PZaxNeBXYA+QCdmC51rrd9T1e2eu4lwFva63r3IxDCCGGS66tcm0dMuZ+PADYgEnAqRhJ7bWu134N/BsYgfGzfcD1/FkYM2Wmus79deDgMM4tgo3WWj7kA601QAVwpuvzxUAHkDDI9kXA4V5fvwd80/X5MmBHr9eSAA2M8WRbjD/QXUBSr9efAp5y83vqL8Y7en39HeBN1+c/x3iDb76W7PoZnDnAsZOARmCh6+u7gZeH+bP6yPX51cAnvbZTGBeFbw5w3KXA+v7+DV1f57p+ljEYF6RuILXX6/cAj7k+/wVGEmO+NgNoHeRneyVQ5zp2PNAAfNX12mW94+qz3zZgST/P98Q6yM+pcoh/756fB3CiGV8/2x0PVAFRrq/XAV/z9/8x+ZAP+Yi8D+TaKtdWz66tGpjS57looB2Y0eu5bwPvuT5/AngYyOmz3+nA58AJuK538hEeHzICJwZTp7VuM79QSiUppf7mGrpvBD4A0tXAVZj2mZ9orc0RlhQPt80GDvV6Dow33v1yM8Z9vT4/0ium7N7H1lq3MMidKldMzwNXu+5oXoFx53A4PytT3xh076+VMdVvuVKqxnXcpzDuJrrD/Fk29XpuD8bIlKnvzyZBDbxG4xrgOa11lzZGtVZwdBrlOIw7nP0Z7LWhfOHffoifxzhgj9a6q+9BtNafAi3AqUqpaRgjhKuGGZMQQnhCrq1ybR3s2tqfTIxRzT0DnOMnGEnpZ64pmtcBaK3/izHa9yCwXyn1sFIqzYPziiAlCZwYjO7z9Q+B44DjtdZpGMPy0GseuR/sBUa6puuZxg2yvTcx7u19bNc5M4bY53Hga8CXgFSMKXvexNE3BsUXv997MP5dZrqOe2WfY/b9N+utFuNnmdrrufFAzRAxHUMZaw5OB65USu1TxlqOi4FzXVNVqoDJA+w+0Gstrsfe/9Zj+mzT9/sb7OdRBYwf5CL5uGv7q4AXer+hEkIIP5Jrq1xbPVUPdGJMHT3mHFrrfVrrb2mtszFG5v6iXJUstdb3a63nAvkYUyl/7MO4hEUkgROeSMWYb96glBoJ3OXvE2qt92BMb/uFUipOKXUi8BU/xfgCcL5S6iTXWq5fMfT/kQ8xpg4+jDFFpMPLOF4D8pVSF7oSj1v4YhKTCjS7jmvn2D/E+zHmxx9Da10FrAHuUUolKKVmAt/AWL/mqaswpmWYaxOKMC4M1RjTJ18FxiilvqeMoiGpSqnjXfv+A/i1UirPtcB6plIqQxvrz2owksJo1x3EgZJA02A/j88wLtr3KqWSXd9z7zUPTwJfxbhQPzGMn4EQQviCXFuPFanXVlOc61gJSqkE13PPAXe7rqcTMNaTPwWglLpEHS3mchgj4exWSs1XSh2vlIrFuEnahjHdU4Q4SeCEJ/4EJGLcCfoEo0BFIFyBsZ7pIPAb4F8Yc8H78yeGGaPWuhy4CWNh916MP4LVQ+yjMd78T+CLScCw4tBa1wOXAPdifL95wOpem/wSmAM4MC5IK/oc4h7gDqVUg1LqR/2c4jKMufu1wEvAXVrr/7gTWx/XAH9x3fXr+QAeAq5xTSX5EsYbgn3AduA0177/h3Eh+jfGOodHMH5WAN/CuHAexLhbuGaIOAb8eWijP89XMKZHVmL8W3691+vVQAnGhe5Dz38EQgjhE39Crq1994nUa6upHCNRNT+uBW7GSMJ2AR9h/DwfdW0/H/hUKdWMsRzgVq31boxCYX/H+Jnvwfjef+9FXCJIKOP/iBChQxnlcbdqrf1+l1KEN6XUo0Ct1voOq2MRQggrybVViNAhI3Ai6LmmAExWSkUppc4GlgArLQ5LhDilVC5wIcYIoBBCRBS5tgoRujypgCOEVcZgTGfIwJh2caPWer21IYlQppT6NfB94B7XNBMhhIg0cm0VIkTJFEohhBBCCCGECBEyhVIIIYQQQgghQoQkcEIIIYQQQggRIoJyDVxmZqbOzc21OgwhhBB+VlxcXK+1zrI6jlAh10chhIgcA10jgzKBy83NZd26dVaHIYQQws+UUnusjiGUyPVRCCEix0DXSJlCKYQQQgghhBAhQhI4IYQQQgghhAgRksAJIYQQfqKUilZKrVdKvdrPa0opdb9SaodSaqNSao4VMQohhAgtQbkGTgghhAgTtwJbgLR+XjsHyHN9HA/81fUohBDD0tnZSXV1NW1tbVaHIjyQkJBATk4OsbGxbm0vCZwQQgjhB0qpHOA84G7gB/1ssgR4QmutgU+UUulKqbFa672BjFMIET6qq6tJTU0lNzcXpZTV4Qg3aK05ePAg1dXVTJw40a19ZAqlEEII4R9/An4COAd43Q5U9fq62vWcEEIMS1tbGxkZGZK8hRClFBkZGR6NmkoCJ4QQQviYUup84IDWuniwzfp5TvdzrOuVUuuUUuvq6up8FqMQIjxJ8hZ6PP03kwROCCGE8L1FwAVKqQpgOXC6UuqpPttUA+N6fZ0D1PY9kNb6Ya31PK31vKws6XkuhAheBw8epKioiKKiIsaMGYPdbu/5uqOjY9B9161bxy233DLkORYuXOiTWN977z3OP/98nxwr0GQNnBBCCOFjWuufAj8FUEotBn6ktb6yz2argO8qpZZjFC9xyPo3IUQoy8jIoLS0FIBf/OIXpKSk8KMf/ajn9a6uLmJi+k8/5s2bx7x584Y8x5o1a3wSayiTETghhBAiQJRSNyilbnB9+TqwC9gB/B34jmWBCSGEnyxbtowf/OAHnHbaadx222189tlnLFy4kNmzZ7Nw4UK2bdsGfHFE7Be/+AXXXXcdixcvZtKkSdx///09x0tJSenZfvHixVx88cVMmzaNK664AqMmFLz++utMmzaNk046iVtuucWjkbZnn32WwsJCCgoKuO222wDo7u5m2bJlFBQUUFhYyB//+EcA7r//fmbMmMHMmTO59NJLvf9huUlG4IQQQgg/0lq/B7zn+vyhXs9r4CZrohJCiMD5/PPPefvtt4mOjqaxsZEPPviAmJgY3n77bX72s5/x4osvHrPP1q1beffdd2lqauK4447jxhtvPKbM/vr16ykvLyc7O5tFixaxevVq5s2bx7e//W0++OADJk6cyGWXXeZ2nLW1tdx2220UFxczYsQIzjrrLFauXMm4ceOoqamhrKwMgIaGBgDuvfdedu/eTXx8fM9zgSAJnBBCCCGEEGHml6+Us7m20afHnJGdxl1fyfd4v0suuYTo6GgAHA4H11xzDdu3b0cpRWdnZ7/7nHfeecTHxxMfH8+oUaPYv38/OTk5X9hmwYIFPc8VFRVRUVFBSkoKkyZN6inJf9lll/Hwww+7FefatWtZvHgx5nrjK664gg8++IA777yTXbt2cfPNN3Peeedx1llnATBz5kyuuOIKli5dytKlSz3+uQyXTKEUQgghhBBC+E1ycnLP53feeSennXYaZWVlvPLKKwOWz4+Pj+/5PDo6mq6uLre2MadRDsdA+44YMYINGzawePFiHnzwQb75zW8C8Nprr3HTTTdRXFzM3Llz+43RH2QETgghhBBCiDAznJGyQHA4HNjtRsvLxx57zOfHnzZtGrt27aKiooLc3Fz+9a9/ub3v8ccfz6233kp9fT0jRozg2Wef5eabb6a+vp64uDguuugiJk+ezLJly3A6nVRVVXHaaadx0kkn8cwzz9Dc3Ex6errPv6e+JIELpNpSKHkczvkdRMuPXgghhBBCAOufAhUFRZdbHYnf/eQnP+Gaa67h//7v/zj99NN9fvzExET+8pe/cPbZZ5OZmcmCBQsG3Padd975wrTM559/nnvuuYfTTjsNrTXnnnsuS5YsYcOGDVx77bU4nU4A7rnnHrq7u7nyyitxOBxorfn+978fkOQNQHkzzOgv8+bN0+vWrbM6DN97/cfw2cNw0SNQeLHV0QghhOWUUsVa66HrRgsgjK+PQkS6v5wIUTFww4deHWbLli1Mnz7dR0GFrubmZlJSUtBac9NNN5GXl8f3v/99q8MaVH//dgNdI2UNXCDVlBiPq/8EQZg4CyGEEEIICziqobHG6ijCxt///neKiorIz8/H4XDw7W9/2+qQfErm8QVKVwfs2wi2cbBvE+x6Fyb7fthYCCGEEEKEkDYHtLuqRXYcgbgka+MJA9///veDfsTNG26NwCmlzlZKbVNK7VBK3T7ANouVUqVKqXKl1Pue7BsRDpRDdwecfgekjIbV91kdkRBCCCGEsJqj+ujnjbXWxSFCxpAJnFIqGngQOAeYAVymlJrRZ5t04C/ABVrrfOASd/eNGOb0yfEnwgk3wq73jKImQgghhBAicvVO4BxV1sUhQoY7I3ALgB1a611a6w5gObCkzzaXAyu01pUAWusDHuwbGWpKICkD0sfD3GshLhXWPGB1VEIIIYQQwkq9kzZZByfc4E4CZwd63w6odj3X21RghFLqPaVUsVLqag/2jQy1JWCfC0pBYjrMWwblL8HhCosDE0IIIYQQlnHUgIoG1BdH44QYgDsJnOrnub4lFGOAucB5wJeBO5VSU93c1ziJUtcrpdYppdbV1dW5EVYIaW+Guq2QPefoc8ffaCRzH//FuriEEEIIIYS1HNVgs0PKqJBP4BYvXsxbb731hef+9Kc/8Z3vfGfQfcz2KOeeey4NDQ3HbPOLX/yC3//+94Oee+XKlWzevLnn65///Oe8/fbbHkTfv/fee4/zzz/f6+P4kjsJXDUwrtfXOUDfFZbVwJta6xatdT3wATDLzX0B0Fo/rLWep7Wel5WV5W78oWHvBtBOsPdK4Gx2KPwalDwBLQeti00IIYQQQljHUW1UKbflhHwCd9lll7F8+fIvPLd8+XIuu+wyt/Z//fXXh90Mu28C96tf/YozzzxzWMcKdu4kcGuBPKXURKVUHHApsKrPNi8DJyulYpRSScDxwBY39w1/ta4CJr1H4AAW3gxdrbD2H4GPSQghhBBCWM9RbSRvafaQXwN38cUX8+qrr9Le3g5ARUUFtbW1nHTSSdx4443MmzeP/Px87rrrrn73z83Npb6+HoC7776b4447jjPPPJNt27b1bPP3v/+d+fPnM2vWLC666CKOHDnCmjVrWLVqFT/+8Y8pKipi586dLFu2jBdeeAGAd955h9mzZ1NYWMh1113XE19ubi533XUXc+bMobCwkK1bt7r9vT777LMUFhZSUFDAbbfdBkB3dzfLli2joKCAwsJC/vjHPwJw//33M2PGDGbOnMmll17q4U/1WEMmcFrrLuC7wFsYSdlzWutypdQNSqkbXNtsAd4ENgKfAf/QWpcNtK/XUYeammKwjYeUPiOLo2dA3pfhs79BZ6s1sQkhhBBCCGs4u42kzZZjjMI5qkH3u9ooJGRkZLBgwQLefPNNwBh9+/rXv45Sirvvvpt169axceNG3n//fTZu3DjgcYqLi1m+fDnr169nxYoVrF27tue1Cy+8kLVr17JhwwamT5/OI488wsKFC7ngggv43e9+R2lpKZMnT+7Zvq2tjWXLlvGvf/2LTZs20dXVxV//+tee1zMzMykpKeHGG28ccpqmqba2lttuu43//ve/lJaWsnbtWlauXElpaSk1NTWUlZWxadMmrr32WgDuvfde1q9fz8aNG3nooYc8+pn2x61G3lrr14HX+zz3UJ+vfwf8zp19I05NCdhn9//aolvhsXOh9GmY/83AxiWEEEIIIazTtA90t5HAdbZC5xFoPQxJI70/9hu3w75N3h+ntzGFcM69g25iTqNcsmQJy5cv59FHHwXgueee4+GHH6arq4u9e/eyefNmZs6c2e8xPvzwQ7761a+SlGQ0Nb/gggt6XisrK+OOO+6goaGB5uZmvvzlLw8az7Zt25g4cSJTp04F4JprruHBBx/ke9/7HmAkhABz585lxYoVQ/8MgLVr17J48WLMZV9XXHEFH3zwAXfeeSe7du3i5ptv5rzzzuOss84CYObMmVxxxRUsXbqUpUuXunWOwbjVyFt4oeUgNOwxKlD2Z8JC47U1Dxh3YYQQQgghRGQw17yZa+B6Pxeili5dyjvvvENJSQmtra3MmTOH3bt38/vf/5533nmHjRs3ct5559HW1jbocZTqrxYiLFu2jD//+c9s2rSJu+66a8jj6CFGNOPj4wGIjo6mq6tr0G2HOuaIESPYsGEDixcv5sEHH+Sb3zQGZ1577TVuuukmiouLmTt3rtvnGYhbI3DCCwOtfzMpZYzCPXc1bFkF+V8NXGxCCCGEEMI6Zg84Ww50HDE+b6yBsf2PTHlkiJEyf0lJSWHx4sVcd911PcVLGhsbSU5OxmazsX//ft544w0WL1484DFOOeUUli1bxu23305XVxevvPIK3/72twFoampi7NixdHZ28vTTT2O3Gx3KUlNTaWpqOuZY06ZNo6Kigh07djBlyhSefPJJTj31VK++x+OPP55bb72V+vp6RowYwbPPPsvNN99MfX09cXFxXHTRRUyePJlly5bhdDqpqqritNNO46STTuKZZ56hubl52MVaQBI4/6spARRkFw28zbTzYeQkWH0fzFhqJHVCCCGEECK8maNtafaj9RBCfAQOjGmUF154YU9FylmzZjF79mzy8/OZNGkSixYtGnT/OXPm8PWvf52ioiImTJjAySef3PPar3/9a44//ngmTJhAYWFhT9J26aWX8q1vfYv777+/p3gJQEJCAv/85z+55JJL6OrqYv78+dxwww0efT/vvPMOOTk5PV8///zz3HPPPZx22mlorTn33HNZsmQJGzZs4Nprr8XpdAJwzz330N3dzZVXXonD4UBrzfe//32vkjcANdSwohXmzZunzX4QIe/prxlTKG/6dPDt1j0Kr34frnkVJp48+LZCCBEmlFLFWut5VscRKsLq+iiEgNd+BJueg9srwemE34yCE2+CL/1yWIfbsmUL06dP93GQIhD6+7cb6Bopa+D8SWtjCuVA0yd7m3UZJGcZo3BCCCGEECL8OaohzTWyExVl9AkOgxE44V+SwPmToxpa6noaeP93635ueqaEbmc/o56xibDg27DjP7CvLMCBCiGEEEKIgDN7wJnSckK+F5zwP0ng/Kmm2Hi0z2FTtYPvPF3Caxv3suNAc//bz/8GxCYZFSmFEEIIIUR4c1R9MYGz5cgInBiSJHD+VFsCUbHsTZjMNx5fS0JsNABlNY7+t08aCXOugbIX5D+vEEIIIUQ4a2+CtoY+CZwdGmu9ai0VjPUtxOA8/TeTBM6fakroHl3AN57axJGObp755gkkxkZTVjtAAgdw4neMtXOf/HXgbYQQQgghRGhzuKZK2sYdfc6WYzT2bto3rEMmJCRw8OBBSeJCiNaagwcPkpCQ4PY+0kbAX5xOdG0p78efxtb6Rh5dNp8Z2WnMyE6jvKZx4P3Sx0PBRVD8GJzyI0gcEbCQhYhYjhqIT4EEm9WRCCGEiBQ9Tbz7rIEDYx2cze7xIXNycqiurqaurs4HAYpASUhI+EKbgqFIAucvB7ejOpp4vWUsd30ln8XHjQKgIDuNF4qrcTo1UVED9HtbdItRUnbdo3DyDwMYtBAR6okLIGsaXPq01ZEIIYSIFL2beJvMzx1VMG6Bx4eMjY1l4sSJPghOBDOZQuknH3/wbwByC0/mmoW5Pc/n2220dHSz+2DLwDuPKYTJp8MnD0Fnm58jFSLCdXfCoV2w7Q1o2m91NEIIISJFYw2oKEgde/Q5c9TNIZUoxcAkgfODj7bXs730A1pVIjdcdPYXXivINqZoDVjIxLToVmg5ABuX+ytMIQQYF1DtNNYcbHjW6miEEEJECkc1pGZDdK8JcQk2iE+TYnZiUJLA+diOA03c+HQxC+IqiB03h5jY2C+8njc6hbiYKMprB1kHBzDxVBg7y2gp4HT6MWIhIlyDawpLXAqsf9IoIiSEEEL4W98ecKY0u/SCE4OSBM6HDrV0cN1j60iOdnIcFcTkzD1mm9joKKaPSR16BE4pYxTu4A7Y9rqfIhZC0FBpPB5/g/H/repTa+MRQggRGfr2gDNJLzgxBEngfKS9q5tvP7mO/Y1tPHZeEqq7A+xz+t02326jrMYxdInX6UsgfQKs/pOMCgjhL+Yi8hNvMkbhSp60Nh4hhBDhz+k01rn1m8DZJYETg5IEzge01tz+4ibWVhzmD1+bxbTu7cYL9mNH4MBYB9fY1kXVodbBDxwdAyd+F6rXQuUnPo5aCAEYUyhTx0LSSMj/KpS/ZDRXFUIIIfyl5QA4OwcegTtSD51DvE8UEUsSOB/483938NL6Gn74pamcPzMbakogKfOLjRl7KbS7CpkM1tDbNPsKSBwJq+/zZchCCFPDnqP/V+dcDZ0tRhInhBBC+EtPD7h+3iv29IKrDVw8IqRIAuelVzfW8of/fM6Fs+189/QpxpM1Jcb0SdV/n7epY1KIiVJDr4MDiEuGBdfD52/Aga0+jFwIARhTKNNdF9Cc+ZB5nEyjFEII4V/99YAz9fSCk2mUon+SwHmhpPIwP3huA/NzR3DPRYUopYypV3VbB5w+CRAfE83U0amUDVWJ0rTgWxCTCB8/4KPIhRBArzUIrgROKZh9JVR/BnXbrI1NCCFE+OoZgRtgDVzvbYToQxK4Yao6dITrn1jHmLQE/nbVPOJjoo0X9m4ANGT3X8DEVGBPc6+QCUBypvGmcsO/oHGv98ELIQzN+4w1COnjjz4361KIijFaCgghhBD+4KiGuFSj71tfaa4ETloJiAFIAjcMTW2dfPPxdXR0OXl02XxGJscdfbGmxHgcoAKlqcBu41BLB3sdbe6d9MSbjEbDn/51mFELIY5hthDoncCljIKpZ8OG5dDdaU1cQgghwpvZA66/5TYx8ZA86ug0SyH6kATOQ13dTr77zHp21jXz1yvnMmVUyhc3qCk23gwmZw56nAKzkIk76+AARk6EGUtg3T+hzc19hBCDM5t4911EPvsqaKmDz98KfExCCCHCn6Pq6FTJ/thyjCn+QvRDEjgP/frVzbz/eR2/XlrAoin9JGm1JUNOnwSYPiaNKIX76+AAFt4C7Y1Q/Jj7+wghBuYwR+D6JHBTzoSUMTKNUgghhH+YI3ADkV5wYhCSwHng8TUVPP7xHr518kQuWzD+2A1a6o0pWYMUMDElxkUzZVSK+yNwYEzLnHgKfPJX6OrwIHIhRL8aqiApw6j22lt0DBRdBtv/LetOxbAopRKUUp8ppTYopcqVUr/sZ5vFSimHUqrU9fFzK2IVQgRYxxE4cnCIBG6csQbOnVoJIuJIAuemd7cd4JevlHPm9NHcfs70/jdyc/2bqSDb5lkCB7DoVmjaC5ue92w/IcSxGioH7NfI7KtAO2HDs4GNSYSLduB0rfUsoAg4Wyl1Qj/bfai1LnJ9/CqgEQohrGEWJxno+gNGIZOOZmhrCEhIIrRIAueGrfsaufmZ9Uwbk8Z9lxYRHdV/fzdqSwAFY2e5ddwCu40DTe0caHSzkAnA5DNgdAGsud8ogS6EGL7ePeD6ypgMExbB+qfkDqjwmDY0u76MdX3IL5IQYvAecKaeXnCyDk4cSxK4IRxoauMbj60jOT6aR5bNIzk+ZuCNa4ohaxrEp7p1bLOQSbkn6+CUMtbC1W01pncJIYZHa2MKZfqEgbeZfSUc2gmVHwcuLhE2lFLRSqlS4ADwH631p/1sdqJrmuUbSqn8wEYohLDEYD3gTNLMWwxCErhBtHV2c/0TxRxq6eAfV89nrC1x4I21NqZQujl9EmBGdhrgQSVKU8GFkJYDq+/zbD8hxFEt9dDVOvgUlhlLjD49JVLMRHhOa92ttS4CcoAFSqmCPpuUABNc0ywfAFb2dxyl1PVKqXVKqXV1dXX+DFkIEQiOGkBBavbA2/T0gpMEThxLErgBOJ2aHz6/gQ3VDfzp0iIKc/pptNibowqO1HuUwKXExzApM5lNniZw0bFGX7jKNVC11rN9hRCGgSpQ9haXbNww2bwS2jwYKReiF611A/AecHaf5xvNaZZa69eBWKXUMeWNtdYPa63naa3nZWVlBSBiIYRfOaohdQzExA28TcpoiIqVETjRL0ngBvDHtz/ntY17uf3saXw5f8zQO9QUG49utBDoLd9u82wKpWnO1ZCQDmtkFE6IYTF7wKX3U1G2tzlXQ+cRKF/h/5hE2FBKZSml0l2fJwJnAlv7bDNGKaOLr1JqAcY1+WCAQxVCBJqjavDpkwBRUZA2VtbAiX5JAtePFSXVPPDfHXx93jiuP2WSezvVlEB0nFFgxAOF9jRqGlo51OJhW4D4FJj/TdjyKtTv8GxfIUSvReSDjMCB0RYka7pMoxSeGgu8q5TaCKzFWAP3qlLqBqXUDa5tLgbKlFIbgPuBS7WWijlChL2hesCZbONkBE70SxK4PtZWHOL2Fzdx4qQMfr20ANfN0aHVrocxhYMPh/ejINssZOLhNEqA479tJI0fP+D5vkJEuoZKiE+DxPTBt1PKKGZSsw4ObAlIaCL0aa03aq1na61naq0LzBYBWuuHtNYPuT7/s9Y6X2s9S2t9gtZ6jbVRCyH8Tmv3E7g0u6yBE/2SBK6XPQdbuP6JdeSMSOShK+cSF+Pmj8fZbSRwHk6fBMh3JXAer4MDSBllNBsufRaaD3i+vxCRrKFq6NE306xLISrGaCkghBBCDFdLPXS3u3f9seVAY63xPlOIXiSBc3Ec6eTax9aigUeXzceWFOv+zvXbjWaLHhQwMdmSYhk3MpHymmEWSDjxZujugE//Nrz9hYhUjqqh17+ZkjPhuHOMpt5dHk53FkIIIUzu9IAz2ezg7JKb9OIYksABnd1Obny6mKpDR/jblXPJzUz27ABmARP73GGdv9Buo2w4UygBMqfA9PNh7d+hvXno7YUQhobKwStQ9jX7ajhyED5/038xCSGECG/u9IAzmaN0sg5O9BHxCZzWmp+/XMaanQe598KZHD8pw/OD1JYYvaIy8oYVQ362jT0Hj+Bo7RzW/iz6HrQ5oOSJ4e0vRKRpbYD2RvenUAJMOcPo2bNeipkIIYQYpp4Ezo3rj/SCEwOI+ATuHx/u5tnPqrjptMlcNNeNuyH9qSmB7CKj5OswFNiNdXCbh9NOACBnHoxfCB8/CN3DTAKFiCQON1sI9BYVbaw53fG2sSZBCCGE8JSjGmKTIHHE0Nuao3QyAif6iOgE7t/l+/h/b2zh3MIx/PBLxw3vIF3tsG/TsNa/mfKz0wAoG04hE9OiW407NGXSq0qIIfX0gPNgBA6MapTaCaXP+D4mIYQQ4c/sAedOlfMEG8SlSC84cYyITeDKahzcuryUmXYbf7ikiKgoN9sF9LW/DJydw6pAacpMiWesLWH46+AA8s6CrGmw5n6jRK0QYmANlcajzYMROICRkyD3ZKMapdPp+7iEEEKEN3dbCICR5Nlyjs4aEcIlIhO4fY42vvn4OkYkxfL3a+aRGBc9/IPVlBiPwyxgYiqw27wbgYuKgoW3GAnlzne8ikWIsOeogphEo7qkp2ZfCYd3Q6W07BJCCOEhR/XRtW3uSLNDo4zAiS+KuATuSEcX33xiLU1tnTyybD6jUhO8O2BNCSRnuX83ZQAF2TZ21bfQ0t41/IMUXgKpY2H1fV7FIkTYMytQujOFpa/pFxgNwEukmIkQQggPdLZBywHPCmjZcmQNnDhGRCVwTqfme8tL2VzbyAOXz2b62DTvD1pbYkyfHM4bwV4K7GloDZv3DrOQCUBMHJzwHdj9wdGRQSHEsRoqPbuA9haXBAUXweaXjeqvQgghhDvMkTRPbvrbcqClzkj+hHCJqATut29u5d+b93Pn+TM4fdpo7w/Y3gR127yePglHK1F6NY0SYO4yY3Rgzf1exyRE2HJUeV7ApLc5V0FXK5S96LuYhBBChDdPesCZzG1lGqXoJWISuOWfVfK3D3Zx1QkTWLYw1zcHrS0FtFcVKE2j0xLISo2nrMaLETiAhDSYd60xOnBot9dxCRF2OlqMhtzDHYEDY9R9VL5MoxRCCOG+4SRwPb3gJIETR0VEArdmRz13rCzjlKlZ3PWVGSgvpzv2qHVNU/SiAmVvBdlplHtTidJ0/I2goo2+cEKILzIvoOkThn8MpYxiJrUlsL/cN3EJIYQIb+b1x5MiJtILTvQj7BO4nXXN3PBUMRMzk/nz5bOJifbht1xTbLwJTM7wyeEK7Da2H2imrbPbuwOljYVZXzdKnbfU+yQ2IcKG2ULAmymUADO/DlGxxv8zIYQQYiiN1ZA8CmI9KKBnJnvSC070EtYJ3OGWDq57bC2x0VE8umw+aQmxvj1BzXqfTJ805Wfb6HZqtnhTyMS08BZjjc5nf/f+WEKEk54ecF4mcMkZMO1c2LAcutq9j0sIIUR486QHnCk2wah2Lr3gRC9hm8C1d3Xz7SeL2eto4+Gr5zFuZJJvT9BcB45KnxQwMRXmuAqZ1Poggcs6DqaeA589DB1HvD+eEOHCUWWMnKWO8f5Ys6+G1kOw7Q3vjyWEECK8DSeBA+kFJ47hVgKnlDpbKbVNKbVDKXV7P68vVko5lFKlro+f93qtQim1yfX8Ol8GPxCtNT9dsYnPKg7xu4tnMnfCCN+fxMfr3wCybQmMSIql3NtKlKZFtxpvLkuf9s3xhAgHDZVgs0NUtPfHmnyacWFdL8VMhBBCDEJrVwI3jNkf0gtO9DFkAqeUigYeBM4BZgCXKaVm9LPph1rrItfHr/q8dprr+Xnehzy0v7y3kxUlNXz/zKksKfJgoagnakpARcHYWT47pFKKAruNMl8UMgEYfwLkLIA1D0C3Fw3ChQgnDVXeT580RUVD0eWw4x25uAohhBhY62HoPDK8ETgzgdPa93GJkOTOCNwCYIfWepfWugNYDizxb1jD19rRzYvF1SwtyuaWM6b470S1JZA1DeJTfHrY/Gwb2/Y10d7lZSETMCrlLboVGvbAlpe9P54Q4cBRBenjfXe8oisADaXP+u6YQgghwou5hm24CVxHM7T56Aa/CHnuJHB2oPfKyWrXc32dqJTaoJR6QymV3+t5DfxbKVWslLrei1jdkhgXzYrvLOTei2b6rl1AX1obFSh9OH3SVGBPo7Nbs31/s28OeNy5kDEFVt8nd26E6GqHpn2+TeBGToSJpxjTKJ1O3x1XCCFE+BhODziT9IITfbiTwPWXBfXNBEqACVrrWcADwMpery3SWs/BmIJ5k1LqlH5PotT1Sql1Sql1dXV1boQ1sPSkOBJifbC+ZSANlUYjYPtsnx+60O4qZOKrdXBRUbDwZti7AXa/75tjChGqHNWA9t0UStPsq4yR7ooPfXtcIYQQ4aEngRvOGrhxXzyGiHjuJHDVQO/fthygtvcGWutGrXWz6/PXgVilVKbr61rX4wHgJYwpmcfQWj+stZ6ntZ6XlZXl8TcSUDXFxqMPK1Caxo9MIjUhxnfr4ABmXmr0HVl9v++OKUQoMqeweNsDrq/pX4F4m/SEE0II0T9HFUTHQ3Km5/vazF5wksAJgzsJ3FogTyk1USkVB1wKrOq9gVJqjHLNV1RKLXAd96BSKlkplep6Phk4Cyjz5TdgidoSiI6DUflDb+shpRT52WmU1figlYApNgFOuAF2vgP7NvnuuEKEmgYzgfPhFEqA2EQovBi2rILWBt8eWwghROgzWwgMZ3lPymiIipEETvQYMoHTWncB3wXeArYAz2mty5VSNyilbnBtdjFQppTaANwPXKq11sBo4CPX858Br2mt3/THNxJQNethzEyIifPL4QuybWzZ20hXtw/X08y7DuJSZBRORLaGSqN6bJofqtPOuQq62qDsBd8fWwghRGgbbg84MCoep2bLGjjRw60+cFrr17XWU7XWk7XWd7uee0hr/ZDr8z9rrfO11rO01idorde4nt/lem6W6/W7/fetBIizG2rXg933BUxMhTk22ruc7KjzUSETgMQRMHcZlL1ovIkVIhI5qiB1LETH+v7YY4tgdCGUSE84IYQQfQy3B5xJesGJXtxK4EQv9Z9DZ4tfKlCa8rPNQiY+nEYJcMKNxtD9x3/x7XGFCBW+7AHXl1Iw+0rYWypTlYUQQhzV1WFUQLZ5MfvDZpcETvSQBM5TNSXGox8KmJgmZiaTFBftu0qUJlsOFFwMJU/AkUO+PbYQocBR6fv1b73N/JqxPlaKmQghhDA11WJUQB7mFEow9m2slXY1ApAEznM1xRCfZvRW85PoKMWMsWm+T+AAFt1ijCCue8T3xxYimHV3gaPG9xUoe0saCdPOg43/MnrOCSGEEN70gDOl2cHZCS0HfBOTCGmSwHmqtgTGzjL6q/lRgd3G5r2NdDt93Hx7dD5M+RJ8+jfobPXtsYUIZk17QXf7bwqlafZV0HoYtr7q3/MIIYQIDd70gDNJLzjRiyRwnuhqh31lfp0+aSqw2zjS0c3u+hbfH3zRrdBSBxue9f2xhQhWDj+1EOhr0mLjQivTKIUQQsDR6483FZClF5zoRRI4T+wrM4av/ViB0lRgTwOg3JcNvU25JxlFWNY8YFTVFCISmNVX/Z3ARUVD0eWw812p+CqEEMKYvp+UAXFJwz+GOf1SEjiBJHCeqfV/ARPTlKwU4mOi/LMOTiljLdyhXbD1Nd8fX4hgZDbx9mYNgruKrjAeS2WUWwghIp43PeBMCekQmyy94AQgCZxnaooheZR/mgD3ERMdxbSxaWzyRwIHMP0CGJELq/8E2sfr7IQIRo5KSM6C2ET/n2vEBJh0KpQ+JRXDhBAi0nnbAw6Mm++2nKPTMUVEkwTOEzUlxvRJpQJyukJ7GuU1jTh9XcgEjGleC282ktI9a3x/fCGCTUOV/6dP9jb7KmMK5e73A3dOIYQQwUVrI+nyxewPm92YjikiniRw7mprNJp4B2D6pKkg20ZTexdVh4/45wRFV0BSJqy+zz/HFyKYNFT6vwJlb9POhwSbFDMRQohI1uaAjmYfJXA5sgZOAJLAuW9vKaCN4h8BUmC3AVBW0+ifE8QmwvHfhu1vwf7N/jmHEMHA6TQuev7sAddXbAIUfg22vGK0FRBCCBF5fNEDzpSWY/SBkz6jEU8SOHfVuAqYZM8O2CnzRqcQG638tw4OYP43ITbJqEgpRLhqqYPudkifENjzzrnKOO/G5wN7XiGEEMHBFz3gTGYSKIVMIp4kcO6qKTaKfiRnBOyU8THRTB2d6p9WAqakkcZanU3Py7xqEb7Mcv6BnEIJMHYWjJkJ658M7HmFEEIEB4cPKyD39IKT92uRThI4d9WuD+j0SVOh3UZZjQPtz0qRJ94Ezi4ofdp/5xDCSg6zB1yAEzgwbpDs2wh7NwT+3EIIIazlqIaoWKOKubfMm5CyDi7iSQLnjuYDxh2UADTw7ivfbuPwkU5qHW3+O8mICTCmACo+9N85hLBSTw84CxK4woshOl6KmQghRCRyVBsjZ1E+eMudlm08NkoCF+kkgXNHTeAaePdVkJ0G4J+G3r1NWARVa6Grw7/nEcIKjiqjCWpCWuDPnTQSpp8PG5+DTj/eiBFBRSmVoJT6TCm1QSlVrpT6ZT/bKKXU/UqpHUqpjUqpwN8lFEL4ly96wJliE43q4TICF/EkgXNHbQmoKGM9S4BNH5tGdJQKTALX1eqqtilEmGmotGb6pGn2VdDWAFtftS4GEWjtwOla61lAEXC2UuqEPtucA+S5Pq4H/hrQCIUQ/ueo9s36N5P0ghNIAueemhLImg5xyQE/dUJsNHmjUgKQwC00Hves9u95hLBCQxXYAtjEu6+Jpxrnl2ImEUMbml1fxro++i5mXgI84dr2EyBdKTU2kHEKIfyouwuaan2cwI2TETghCdyQtDYqUNoD1z6gr/xsG2W1fuoFZ0rOhMzjoEISuJC37Q3YIiM9PbQ2plCmW5jARUXB7Ctg13tweI91cYiAUkpFK6VKgQPAf7TWn/bZxA5U9fq62vWcECIcNO0F7YQ0H/63TrNLGwFBjNUBBL2GPdB6yJIKlKYCexovllRzoLGNUWkJ/jvRhIWw6QVwdkNUtP/OI/zH2Q2vfA8SbMa6K2E00e5otnYKJUDRFfDevUa119N+NuzDOJ2aTqeTzm5NR5eTzm5nz6P5XEe3+fXR1zq6NZ09n5vP6z7b9N5HG1/3fq1Lf2GbaxbmcuUJAe6tF0K01t1AkVIqHXhJKVWgtS7rtYnqb7e+TyilrseYYsn48RbeiBBCeMaXPeBMthxob4Q2h3GtFxFJErih1BQbjxYUMDEV2I3/oJtqHJzhzwQu9yQo/ifs2wTZRf47j/CfPWugeR8cOWhM3YiW/+KW9YDrK30cTD4N1j8Np97m1k2Sd7bs586VZbR0dPckTV1O37cUiVIQFxNFbHQUcdHGY2yM6vncfC02WpEWF0tctCI2OoqRyXE+jyUcaa0blFLvAWcDvRO4aqD3L2YOUNvP/g8DDwPMmzfPjz1lhBA+1ZPA+XgNHBjr4CSBi1jy7m4oNSVGCfDR+ZaFMGNsGkpBWU0jZ0wf7b8TjT/ReNyzWhK4UFW+wnh0dhqjxxmTrY0nGJhNVK0egQOYfSW8cJ0xlXLKGYNu2tzexU9XbCIlIYaz8scQG616JVJmoqWIi4k+9rUY12t9ErA4V2JmPme+Hh3V30CQ8IZSKgvodCVvicCZwG/7bLYK+K5SajlwPODQWu8NcKhCCH/paeLtwymU5s3IxhoYPcN3xxUhRRK4odSuhzGFEB1rWQjJ8TFMykymrNbPhUxsdhiRa4zinHiTf88lfK+7Cza/bPwbHq6A+u2SwMHRHnDpQTDVb9r5kDjC6Ak3RAL34Ls7ONDUzt+umsvs8SMCFKDwobHA40qpaIz15s9prV9VSt0AoLV+CHgdOBfYARwBrrUqWCGEHzTWGC1s4lN9d0xzPZ2javDtRFiTIiaDcXZDbaml0ydNBXYb5f6uRAkw4SQjgXM6/X8u4Vu73zemTp78I+Pr+s+tjSdYNFRCbLKROFktJh4Kv2a0EzhyaMDNKupbeOTD3Vw4xy7JW4jSWm/UWs/WWs/UWhdorX/lev4hV/JmVqq8SWs9WWtdqLVeZ23UQgif8mUPOFPqGFDR0kogwkkCN5i6bdDZAnbre6sWZNuodbRxsLndvyeasNAo2lK/zb/nEb5XvgLi06DwEkjOkgTO5Kgypk+qIJkmOOcq6O4wGnsP4DevbSE2WnH72dMCGJgQQgif8nUPODDWT6dlSyuBCCcJ3GBqS4zHIBiBy7enAfi/nYDZD67iI/+eR/hWVwdseQWmnQexCZA51ZhCKVxNvIOoct+YQhhbZPSE08fWo3j/8zre3rKfm8/I82/VWSGEEP7lqPJ9AgfGMaWVQESTBG4wNcXGiMZI69cR5WcblYb83tB7RC6kZhvTKEXo2Plfo6Rw/oXG1xlT4KAkcIDrAhoEBUx6m30l7C+DvaVfeLqz28mvXiknNyOJaxflWhKaEEIIH2hzlfr3RwKXZpc1cBFOErjB1JQY1RijrP8x2RJjmZCRRLm/C5koBbmLjASun9EBEaTKXjTWeE1abHydOdVYD9dy0NKwLNfeZPSBC4YKlL0VXgIxCVDy5BeefnxNBTvrWvj5V2YQHyO9GIUQImSZI2R+G4GrlXoFEcz6zCRYdbbB/vKgmD5pKsi2UVbj5ymUYEyjbN4Hh3b5/1zCe52tsO11mP4ViHH15cqcajxG+iicWYEy2EbgEtONf69NLxj/fkBdUzv3vb2dxcdlcfo0P7YLEUII4X/+aOJtsuUYa6lb6nx/bBESJIEbyP4yo5dWtvUFTEz59jQqDx3BcaTTvyeasMh43LPav+cRvrH939DRDAUXHX0uc4rxGOnr4BxB1EKgr9lXQbvDWLsI/P6tbbR2dnPn+dLXRwghQl5PDzg/TaEEaJRCJpFKEriB1JgFTIIngStwrYPz+zTKzKmQlCnr4EJF2Qqj6uSEk44+lz4BouOkEmVDpfEYbFMoAXJPNv6d1j/JxuoGniuu4rqTJjI5K8XqyIQQQnjLUW2U+08d4/tjm0mhVKKMWJLADaSmGFJGH73LEQQK7K5CJoFYBzdhoYzAhYL2Zvj8LZixBKJjjj4fFW0UMon0EbiGSiORTR5ldSTHiooyipns/oC/vvQ2Gcnx3Hz6FKujEkII4QuOauM9ZJQf1jP3JHBSiTJSSQI3kNoSY/pksPSOAkYmx2FPTwzQOrhFxpvfBqlyFNQ+fxO6Wr84fdIklSiPVqAMgkJE/Sq6HI1i+v5X+MnZx5GaEGt1REIIIXzBHz3gTIkjIDZJRuAiWJC+q7FYm8MYuQiiAiam/Ow0/4/AwdF+cDKNMriVrTDaPow74djXMqfCod1Gj7hI1VAVnNMnXVoSxvCxKuLyuI+4uGis1eEIIYTwFX/1gANjcCHNLmvgIpgkcP2pLQU02GdbHckxCuw2dte30Nze5d8Tjc6HBJtMowxmrQ2w4z+Q/9X+R5gyp4LuhsO7Ax5a0GioDL4KlL08+O4Onmo/mUxnPVG737M6HCGEEL7g7DbK/PsrgQPj2DICF7EkgetPrauASRBVoDQV2m1oDZtr/TyNMioaxp8oI3DBbNvrRhnhggv7fz0zz3iM1HVwnW3QcgDSx1sdSb8q6lv4x4e7SZ75FUgcCeufsDokIYQQvtC8H5xdfk7g7LIGLoJJAtefmhIYMRGSRlodyTHy7WkAlNUEaBrlwe3QtN//5xKeK3vRSE4GmuqbYbYSiNBKlOadySBN4H7z2hZioxU/PncmzPw6bH1dGq8LIUQ46OkB588EbpyRKEbyMokIJglcf2pKgqp9QG+jUhMYlRofoHVwrrL0lTIKF3RaDsKu9yD/woEL7SSkQerYyB2Ba9hjPAbhFMr3P6/j7S37ufmMPEalJcCcq4y+kxv/ZXVoQgghvOXPHnCmNDugoanWf+cQQUsSuL6a9huLQoNw+qSpwG4LzAjc2JkQmyzTKIPRllXG9Iz+qk/2lpkXwSNwZhPv4ErgOrud/OqVcnIzkrh2Ua7x5Oh842/O+qdAa0vjE0II4aWAjMBJL7hIJglcX+b6tyCsQGkqsNvYcaCZ1o5u/54oOhbGLYAKKWQSdMpXGFMkxxQOvl1GnjENNhKTgoYqVxPVbKsj+YLH11Sws66Fn39lBvExvfoDzb4SDpQf/RskhBAiNDlqID7NKAbnL9ILLqJJAtdXTQmoKGP0KUgVZKfh1LBlX4D6wR0ohyOH/H8u4Z6m/VDx0eDTJ02ZU422GC11gYltEJUHj3Dzs+upOnQkMCd0VBlTTHo3OLdYfXM79729ncXHZXH6tNFffLHwYohJgJInrQlOCCGEb/izB5wpze46l/TrjUSSwPVVWwKjZkBcstWRDKjAbtzRKQ/ENMrcRcZj5Sf+P5dwz+aXQTuHnj4JvSpRWj+N8i/v7eCVDbV88/F1NLV1+v+EDZVBN33yd29uo7WzmzvPn3Hsiwk2mLHEKE7TEaAkVwghhO/5swecKS7JqGDcKCNwkUgSuN60hppiyA6+/m+9jbUlMDI5jrKaAIzAZc+B6HjpBxdMylcYNxlGTRt62yBpJXCopYOX1tcwd8IIdtQ1c+vyUrqdfp7W2VAVVAVMNlU7eK64imsX5TI5K6X/jWZfBe2NxhpHIYQQoSkQI3AgveAimCRwvR2ugNbDQVuB0qSUIj87jU2BGIGLTYCceZLABQtHNVR+PHDvt77SciAm0fIEbvnaStq7nNxzYSG/vCCf/249wP97fYv/TtjdaVTmCpIWAlpr7lpVRkZyPLeckTfwhrknGS1MZBqlEEKEpo4WaD0UwARORuAikSRwvYVAARNTod3G5/ubaO/ycyETMPrB7d0A7U3+P5cYXPlK4zHfzQQuKgoyp1g6hbKr28mTH+9h0ZQMpo5O5coTJrBsYS6PfLSbZz+r9M9JG2uMaaZBMoVyZWkNJZUN/OTs40hNiB14Q6Vg9hWw5yM4uDNwAQohhPANM6EKxAwQGYGLWJLA9VZTYhQRGNXP+pQgU2C30eXUfL6v2f8nm7DIeDNc9an/zyUGV/YijC2CjMnu72NWorTIW+X72etoY9nCiT3P3XHedE6dmsWdK8tYs7Pe9ydtMHvwWJ/AtbR3ce8bW5mVY+PiOW7ckS26wiikVPq0/4MTQgjhW4HoAWdKs0O7A9oCsKRGBBVJ4HqrKTHKskcPcoc8SBRkG4VMAtLQe9wCiIqRfnBWO7TbGCV2d/qkKXMqHN4DnW3+iWsIj63ZzfiRSZw+bVTPczHRUTxw+WwmZiZz41Ml7Krz8Y2Inh5w1k+hfPDdHexvbOeuC/KJihqiaihAWjZMORNKnwFnAEbYhRBC+E4gesCZzHNIIZOIIwmcqbsL9paGxPRJgHEjE0lLiAnMOri4ZGPUR/rBWat8hfGY/1XP9svMAzQcCvyUvLIaB2srDnP1iROI7pO8pCXE8sg184mOUnzz8XU4jviwMmVDAO+ADqKivoV/fLibC+fYmTN+hPs7zr4SmvbCjnf8F5wQQgjfc1QbsyhSx/r/XNILLmJJAmeq3wadR4yqiyFAKUWB3RaYVgJgtBOoKYbO1sCcTxyr7CXIWeD5qFLmVOPRgnVw/1xdQVJcNF+b3/9UxvEZSTx05VyqDh/hxqeL6ex2+ubEDZWQMgZi4n1zvGH6zWtbiI1W3H62GxVDe5t6DiRlwPon/BOYEEII/3BUG8lbIGZz9SRw0gsu0riVwCmlzlZKbVNK7VBK3d7P64uVUg6lVKnr4+fu7hs0aswCJqGRwIGxDm7LvibfvekdzIRF4OyE6nX+P5c4Vt3nsH+T59Mn4eh6ufodvo1pCHVN7byyoZaL5+aQNkjhjgUTR3LPhTNZs/Mgd60qR2sftBdwWN8D7v3P63h7y35uPiOPUWkJnu0cEwczL4Vtb0CLH9YICiGE8I9A9IAzpYwxRvtkCmXEGTKBU0pFAw8C5wAzgMuUUv1V+fhQa13k+viVh/tar6YY4m0w0oPiEBbLz06jo8vJjgMBKGQy7nhASTsBq5SvABTMWOr5vnHJRjGPAI/APftZJR3dTq5ZmDvkthfPzeGGUyfzzKeV/HN1hfcnb6iydP1bZ7eTX71STm5GEtcuyh3eQeZcBc4u2LDcp7EJIYTwo0D1gAOIjoHUbKlEGYHcGYFbAOzQWu/SWncAy4Elbh7fm30Dq7YEsouMsushosDuKmQSiGmUiekwpkASOCtobVSfnLAI0oY5pz4zL6AJXEeXkyc/2cOpU7MGblrdx0++fBxnzRjNb17bzLvbDgz/5E6n6wJq3Qjc42sq2FnXwp3nzyA+Jnp4Bxk1HezzYP2Txu+AEEKI4OZ0GqNhgVx/bbNLAheB3MlW7EDvybXVruf6OlEptUEp9YZSKt/Dfa3V2Qb7y0OmgIlpYkYyyXHRgUngACacBFVroasjMOcThv3lRvI1nOmTpow8OLgjYInAG2V7qWtq92j0KSpK8cevFzFtTBo3P7Oez/cPs+9g8z5juq9FUyjrm9u57+3tLD4u6wuVN4dl7jKo2wqfv+WT2IQQQvhRSx10dwT2BqL0gotI7iRw/dW97vsusASYoLWeBTwArPRgX2NDpa5XSq1TSq2rq6tzIywf2rfJmKoUQuvfwHjDm59to6w2QP0/JiyErlajWqcInPIVoKJhhheD15l50NFsVDYMgEdXVzApM5lT8rI82i85PoZHls0jMS6abzy+loPN7Z6f3KxAmT7B83194HdvbqO1s5s7z5+BUm60DRjMrEthRC7899fGnV0hhBDBK5AtBExpdmislWtEhHEngasGet9KyAFqe2+gtW7UWje7Pn8diFVKZbqzb69jPKy1nqe1npeV5dmbPq/VugqYhEgFyt7y7Wlsrm2k2xmAkZUJC43Hio/8fy5hMKdPTjoVkjOHf5wAVqJcX3mYDVUNXLMw172+Z32MtSXyj6vncaCxnW8/WUx7l4e90BzWNfHeVO3gueIqrl2U6/bU0UFFx8Jp/wP7y462kRBCCBGczOtPWgAnm9nGQXc7HJGCV5HEnQRuLZCnlJqolIoDLgVW9d5AKTVGuW41K6UWuI570J19g0JNsVHJJy3b6kg8VpBto7Wzm931AShkkpwJWdOkoXcg1a6HwxWQ78X0SXD1ggPqt3sd0lAeW1NBanwMF80d/h3IWePS+cPXZrFuz2F+umKTZ5UpG/YYjwGeQqm15q5VZWQkx3PLGXm+O3DBRTBqBrx7N3T7sFeeEEII37JiBM5m/+K5RUQYMoHTWncB3wXeArYAz2mty5VSNyilbnBtdjFQppTaANwPXKoN/e7rj2/EKzUlxvRJb6c7WcAsZBKQht5gjMJVfgJOD0dFxPCUvQhRsTD9fO+OkzoW4lL8nsDtb2zjtY17uWTeOFLiY7w61vkzs/n+mVNZUVLDX9/3oAl5QxUkjjSqbwbQytIaSiob+MnZx5E6SNsEj0VFw+l3wKFdUPqM744rhBDCtxprIDYZEkcE7pw9veAkgYskbpVc1Fq/rrWeqrWerLW+2/XcQ1rrh1yf/1lrna+1nqW1PkFrvWawfYNKmwMObg/J6ZMAk7OSSYiNoqwmUOvgFkFHk7FuUPiX0wnlK2HKGd5fDJQKSCXKpz/ZQ7fWXLPQN+vPbjljChfMyuZ/39zGm2X73NvJEfgWAi3tXdz7xlZm5di4eI4f7rwed65RZOn93xpFl4QQQgQfswdcIAcE0lzXHOkFF1FCp2a+v9SuNx5DrICJKSY6iulj0wJYidK1Dk7aCfhf9WfQWO399ElT5lS/jsC1dXbz9KeVnDFtFBMyfDP6pZTify+eSdG4dL7/r1L3fs8bAt/E+8F3d7C/sZ27Lsgf1rq/ISkFZ/zcuECve9T3xxdCCOG9QPaAMyWNhJhEGYGLMJLA1ZgFTGZbG4cXCrJtbK5txBmIQiZp2TBioqyDC4SyFRCTAMed45vjZeQZCWFHi2+O18erG/dysKWDaxdN9OlxE2KjefjquYxIiuWbj69jf+MgI1BaG1MobYEbgauob+EfH+7mwjl25oz347SZSYth4inw4R+gfZgtFoQQQviPFQmcUtILLgJJAldbAiMnGXcwQlSBPY2m9i4qDx0JzAknLDISOClZ6z/Obti8EvK+BAlpvjmmWcjk4A7fHK8XrTX/XL2bqaNTWDg5w+fHH5WawD+umU9jWyffemIdrR0DrME8ctBodRHAKZS/eW0LsdGK28+e5v+Tnf5zo9LYJw/5/1xCCCHc19lq9IGzoAKy9IKLPJLA1ZSE7Po3U362BYVMWg8ZDYaFf+xZDc37jQqEvtLTSsD30yjX7TlMeW0jyxZO9L732QBmZKdx36Wz2VTj4EfPb+h/xLmh0ngM0BTK9z+v4+0t+7n5jDxGpSX4/4Tj5hvr4dbcD0cO+f98Qggh3NPo6pIV6BE4MNbByRq4iBLZCVzTPuMXPkTXv5mmjk4lLjqKslpZBxc2yl40Klnlfdl3xxw5CVB+SeD+uXo3tsRYls72byuOL80Yze1nT+O1TXv509v9FGQxE7gA3AHt7Hbyq1fKyc1I4tpFuX4/X4/T/seYQrn6vsCdUwghxOB6epBakMDZcoz3tF0dgT+3sERkJ3Dm+jf7XGvj8FJcTBTHjUmlPFCVKEfkGk0qZR2cf3R3wuZVxtq3uCTfHTc2AUZM8HklypqGVt4q38+l88eRFOdd6wB3XH/KJC6Zm8P9/93By6V97jiaF9AAjMA9vqaCnXUt3Hn+DOJjov1+vh5jCqDwYvj0b8YFWwghhPWs6AFnstkBDU17A39uYYnITuBqS0BFw5iZVkfitQJ7GmW1Ds8aHg+XUsYo3J41RtEI4Vu73zemqBb4qPpkb36oRPnkx3vQWnPVib5pHTAUpRR3f7WQBRNH8uMXNlK85/DRFxuqID4NEtL9GkN9czv3vb2dxcdlcfq0UX49V78W/xScnfDB7wN/biGEEMdyVAPKKPYWaNILLuJEdgJXUwKjpvt2lMMi+dk2Go50Un24NTAnnLAQmvcZzYWFb5WtMJKQKWf6/tgZeUYREx8VoGnt6Gb52krOmjGGnBGB+38UFxPFQ1fOZUxaAt9+ch3Vh10FfBoqjemTfu7B87s3t9Ha2c2d58/w25q/QWVMhtlXQfFjcLgi8OcXQ1JKjVNKvauU2qKUKldK3drPNouVUg6lVKnr4+dWxCqE8AFHFaSMhpj4wJ9besFFnMhN4LQ2RuBCfP2bqdBuFDIpD9g6uJOMR1kH51td7bDlVZh2vn8uApl5RpXGRt/cpXu5tIaGI52BXQPmMjI5jkeXzaO9y8k3H19Hc3uXq4m3f6dPbqp28FxxFdcuymVyVopfzzWoU38CKgre+611MYjBdAE/1FpPB04AblJKzehnuw+11kWuj18FNkQhhM9Y0ULAZLO7Yqiy5vwi4CI3gTu8G1oPh3wFStNxY1KJjlKUBWodXGYeJGXKOjhf2/EOtDv8M30SelWi9H4dnNE6oILpY9NYMNGaNhxTRqXy4OVz2H6gmVufXY9uqPRrCwGtNXetKiMjOY6bz8jz23nckpYNC74FG5fDAakIG2y01nu11iWuz5uALYDd2qiEEH5jZQIXlwyJI8AhI3CRInITuDApYGJKiI0mb1RK4CpR9qyDkxE4nypfYfwRnrTYP8fvSeC87wX38a6DbNvfxLWLcq2ZRuhyytQs7vrKDNZu3Y1qb/RrBcqVpTWUVDbwk7OnkZYQ67fzuO2kH0BsErx7t9WRiEEopXKB2cCn/bx8olJqg1LqDaVUfmAjE0L4hNbWJnAgveAiTGQncDEJxhq4MFFgt1FWE6BCJmA09G6oNApHCO91HIGtr8P0CyDaT8lBciYk2HwyAvfP1RWMTI7jglkWLNju4+oTc/n2LONntvpgol/O0dLexb1vbGVWjo2L51h4ke4tOQNO/C5sWXX0ppQIKkqpFOBF4Hta675TJEqACVrrWcADwMoBjnG9UmqdUmpdXV2dX+MVQgzDkYPQ1WZNE2+T9IKLKJGbwNWWGNUn/fVG2QIF2WnUN3ewv7E9MCfMXWQ8yjRK39j+b+hs8W3z7r6UclWi9C6Bqzp0hLe37OfyBeNJiA1gCf1B3OBK4P7waSuf7Dro8+M/+O4O9je2c9cF+URFWTfieIwTbzJGbf/7G6sjEX0opWIxkrentdYr+r6utW7UWje7Pn8diFVKZfaz3cNa63la63lZWVl+j1sI4SEre8CZbDmyBi6CRGYC190FezeEzfRJU2GOUcikrCZA0yhHzTBGc2QapW+UvQjJoyD3JP+exwetBB5fU0GUUlx5QmBaB7gjusm486jSx3PDU8VU1Lf47NgV9S3848PdXDjHzpzxI3x2XJ9ISDOmUu58Byo+sjoa4aKMecWPAFu01v83wDZjXNuhlFqAcU32/d0HIYR/WdkDzmSzQ5sD2pusi0EETGQmcHVbofNI2FSgNE0fm4ZSBG4dXFQ0jD9REjhfaG8yRuDylxo/V3/KmGK0gGgbXsGblvYu/rWuinMKxjDGluDj4LzQUAkxifzftWeigG88vhZHa6dPDv2b17YQG624/expPjmezy34FqSOhXd+Lb0Zg8ci4Crg9F5tAs5VSt2glLrBtc3FQJlSagNwP3CpDtgceCGEzwRFAueavimFTCJCZCZwta61ImFSgdKUFBfD5KyUwFWiBGMd3MEd0LQ/cOcMR9veNObP5/up+mRvZiGTg8MbhVtRUk1TWxfXLprow6B8oKESbDlMyEzhoSvnUnnoCN99poSubu963r3/eR1vb9nPd0/PY1RaECWsvcUmwik/hqpPYPt/rI5GAFrrj7TWSms9s1ebgNe11g9prR9ybfNnrXW+1nqW1voErbXMRxciFDmqjboKSRnWxZDmKnLrozZBIrhFZgJXU2xM/Rs5yepIfK4gOy1wUyjBSOAAKuV9h1fKXoTUbBh3vP/P1VOJ0vMEzunUPLamgpk5NuaMT/dtXN5yVPW0EDh+UgZ3Ly3kw+31/PKVzcM+ZGe3k1+9Uk5uRhLXnZTro0D9ZPZVMCIX/vsrnzVqF0II4QazAqWFFZl7Rv+kEmVEiNAErgSyZ0NU+H37BXYb+xrbqGsKUCGTsTMhNlkKmXij9TDseNvo/RaI38kRuaCih5XAfbijnp11LZa3DuhXwxebeH9t/jiuP2UST36yh8fXVAzrkI+vqWBnXQt3nj+D+JjgKNYyoJg4WPwz2LcJNq+0OhohhIgcVrcQAGMavYqSKZQRIvwymKF0tsKBzWFXwMRUYDcKmZQHah1cdCyMWwAVsg5u2La+Bs7OwEyfBOON/siJw6pE+djq3WSmxHNu4Vg/BOaFjhY4Un9MCefbzp7GmdNH88tXynn/c8/Kr9c3t3Pf29s5dWoWp08b5cto/afwYsiabvSF6+6yOhohhIgMwZDARccYSZyMwEWEyEvg9m0CZ1fYrX8zzchOA6C8NoDr4HIXwYFyOHIocOe0WGNbJ99bvt4301XLVkD6hMAW1RlGJcpddc28u62OK08YH3yjUeYFyzWF0hQdpbjv0iKmjk7lu0+XsH2/+9W5fvfmNlo7u/n5V2YE32jjQKKi4fQ7jHWpG561OhohhAh/Xe1GYTAre8CZ0uyyBi5CRF4CZza7DbMKlKa0hFhyM5LYVG3FOrhPAndOi71cWsvK0lpufLqYxjYvKh221MOu94zpk4FMEjLz4NBOcHa7vcsTH+8hNlpx+fHjh9440Mxm8unHxpYcH8Mjy+YTHxvNNx5fx6GWjiEPt6nawXPFVVy7KJfJWSm+jta/pp1n3KB6717obLM6GiGECG+Ntcaj1SNwZgwyAhcRIi+Bqy0xhpjTsq2OxG/y7bbAtRIA481idHxEtRN4eX0No1LjqW1o46crNjHsyt9bVoHu9m/z7v5k5EF3BzTscWvzprZOnl9XxVdmZjMqNQgrMZrfxwB3QO3piTx89Vz2NbZxw5PFtHcNnLhqrblrVRkZyXHcfEaeP6L1L6XgjJ8bd2GL/2l1NEIIEd6CoYWAyWY31sBJN5KwF3kJXE1x2E6fNBXabVQfbqXhyNAjDT4RmwA58yMmgas6dIR1ew5zzcJcfnjWVF7buJdnP6sa3sHKVhjJ1OgC3wY5FA8rUT6/rpqWjm6uWZjrv5i84aiCqBhIHTPgJnPGj+B3F8/ks4pD/M9LZQMm3StLayipbOAnZ08jLSHWXxH716TFkHsyfPB7aG+2OhohhAhfPQlcEEyhtI2D7nZjdo8Ia5GVwLU2GGtD7LOtjsSvCrLNQiaB7Ae3EPZuMBpSh7lVG4zpEhfMyuaGUyZzcl4mv3ylnK37PPx5N+2Dio+M0bdAr7HKdI0suVHIxOnUPP5xBXPGpzNrXLp/4xquhirj7ucQTdCXFNm55Yw8Xiiu5m8f7Drm9Zb2Lu59YyuzcmxcPCcI7qYOlzkKd6QePv2r1dEIIUT4MhO4YJjZJb3gIkZkJXC1643HMK1Aacp3FTIJbD+4haCdUPVp4M5pAa01K9fXMD93BONGJhEVpfi/rxWRlhjLd59Zz5EODyr/la8EtLH+LdCSRhoNR90YgXt32wH2HDwSfI27e3NUuX3383tn5HHezLH89s2tvFW+7wuvPfjuDvY3tnPXBflERYVI4ZKBjFsAU8+B1Q9EVIEhIYQIKEcVJGdBbKLVkUgvuAgSYQmcq4BJdniPwI1IjsOensimQCZw4xYYU9jCvJ3A5r2NbD/QzJIie89zWanx/OnrReysa+YXq8rdP1j5ChiVD1nH+SFSN7hZifKxNRWMSUvg7IKBpydarqGy3wIm/YmKUvzhklnMtNv43vLSnpYbFfUt/OPD3Vw4x86c8SP8GW3gnH4HtDtgzf1WRyKEEOEpGFoImHoSOOkFF+4iK4GrKYGRkyAxTN6cDaLQbgvsFMq4ZCMxDvOG3i+X1hITpTivTx+0RVMyuWnxFJ5bV83LpW784WyoMkYrrRh9M2XmDTmFcvv+Jj7cXs9VJ04gNjpI/1x0dRjTUT1Yf5AQG83fr55HelIs33x8HQea2vjNa1uIjVbcfvY0PwYbYGMKoOBi+OQhaNpvdTRCCBF+gimBS8qAmASZQhkBgvQdmZ/UlIT99ElTgT2N3fUtNHlT4t5TExYaRWI6WwN3zgDqdmpWlday+LgsRiTHHfP6987MY37uCH62YhO761sGP1j5S8ajlQlcRp6xRmqQ6XWPrakgLiaKS+cHweLsgTRWA9rtETjTqLQE/n71PBqOdHLRX9fw9pb9fPf0PEalBWGVTW+c9jOj4uiHv7c6EiGECC9auxK4ILlGKmWsg5MplGEvchK4pn3QVBv2FShN+XajkMnmgBYyWQTOTqheF7hzBtCnuw+yr7HtC9Mne4uJjuK+S2cTEx3Fzc+WDFqqnvIVxojlyEl+itYNZiXKgzv6fdlxpJMVJTUsLcomIyU+gIF5qKHSeEz3/AJaYLfxp0uLqDrUSm5GEtedlOvb2IJBxmSYfSWs+yccdq9thBBCCDe0HobOluAZgQNXLziZQhnuIieB62ngHSEjcK5KlIFdB3c8oMK2ncDK9TUkx0Vz5vTRA26TnZ7I7y+ZRVlNI/e+sbX/jQ7uNArq5Fs4+gZDVqL817pKWju7WbYwiIuXwNEm3sO8A/rl/DE8cd0Co9l3zOBVLEPWqbeBioL3f2t1JEIIET6CqQecSZp5R4QISuCKQUXDmEKrIwmIrNR4xqQlBHYdXGK68fMNwwSurbObNzbt48sFY0iMG/xN/pdmjGbZwlz+ubqC/2zuZ92ROX0y/6t+iNQD6RMgOq7fQiZd3U4eX7OH4yeOZIarqmnQclQZyUla/yOj7jhlahaTs1J8GFSQsdlhwbdgw7NQt83qaIQQIjwEawLXvA+6A7iERgRc5CRwtSUwagbEJVkdScAU2NMC20oAjGmUVWuNwhJh5N2tB2hq72LpANMn+/rpudMosKfx4xc2UNvQZ01g2QpjtHIYU/58KjrGmMLZTwL39pYD1DS0cu2i3MDH5amGKkgdCzHHrksUvZz0fYhNgnfvtjoSIYQID8HUxNuUZjfaOjXttToS4UeRkcBp7SpgEhnr30z52TZ21jV71pvMWxMWQlfr0Z57YWJlaQ2ZKfEsnJzh1vbxMdE8cNkcOruc3PLserq6ncYLB7bCgXKjeXcwGKAS5WNrdmNPTxx0umjQaKgMrotnsErOhBNvgs0vh93/TyGEsERjtTGTJSnT6kiOsrluNMs6uLAWGQncoV3Q1hBxCVyB3YZTw5a9TYE76YSFxmMYTaN0HOnk3a11XDArmxgPSulPzEzm/11YyLo9h/nT265RrvIVgIIZS/wTrKcyp8Lh3V+YarFlbyOf7DrE1SdO8Oj7tYyj0vrRzFBx4k1GG5X//sbqSIQQIvQ5qo0Rr6ggulaaNzRlHVxYC6LfOD8y7zZHSAVKU4HdWLsU0GmUyZmQNS2s+sG9UbaXjm4nS2dne7zvkiI7X5uXw4Pv7eCjz+uM6ZO5J0FqkDTFzsgDZxccruh56rHVFSTERvH1YG4dYHJ2Q2Otxy0EIlaCzZhKueNtqAifmyxCCGGJYOoBZzLXg0svuLAWGQlcTTHEJMKo6VZHElBj0hLITImzYB3cQqj8xHhzHQZWltYwKTOZQldrBk/94oJ8Jmel8Jd/rYSD263t/daX2UrANY3yUEsHK0truHBODulJIbCmrGmvkYDKFEr3zf8WpIyB//7amF4uhBBieIKpB5wpPgUS0mUELsxFSAJXAmNnQnSs1ZEElFKK/GwbZYGsRAlGIZOOJti3MbDn9YO9jlY+3X2IJUV2lFLDOkZSXAwPXj6HUzs/pJsonNMu8HGUXsicYjy6ErhnP6ukvcvJsoW51sXkCS96wEWsuCQ49cdQ+bExEieEEMJz3Z3GTcRgG4ED6QUXAcI/gevugr0bIqb/W18F9jS272+irTOAo2E96+BCfxrlqtJatIYlRZ5Pn+ztuNEpXJG8jo+6C3hoXYNvgvOFBBukjIb6HXR2O3ny4z2cNCWTqaNTrY7MPWYPuPQJ1sYRamZfbfzM3vkVOJ1WRyOEEKGnaa9R7TFoEzgZgQtn4Z/A1W0xqiJG2Po3U0G2jS6nZtu+ABYyScuGERPDIoFbWVpL0bh0cjOTvTtQTQkprTVUZp/NH/79OcV7DvkmQF/InAr1n/NW+T72NbaFzugbGAVMIDgvoMEsJg5O+5kxSr7lZaujEUKI0BOMPeBMaXZZAxfmwj+BqykxHiOsAqWpwLVuq6zWgn5we9aE9N39bfua2LK3kaVejr4BRvXJqFiWXPZtstMTuOXZUhxHgqTJpquVwD8/2s2EjCROnzbK6ojc11AJyVkQm2h1JKGn8BKj4NB/7zZmKgghhHBfMPaAM9lyoPUwdLRYHYnwkwhI4IqNaWIjJ1kdiSVyRiRiS4ylrCbA6+ByF0HrIajbGtjz+tDK0hqioxTnz/IygXM6jeqTU84kLT2TBy6bw/7GNn7y4gZ0MBSRyJwKbQ3srqzk6hNziYoa3lo/SzRUBefFMxRERcPpdxiFdTYutzoaIYQILQ7XFH6z71owMUcFZR1c2Ar/BK62xJg+OcwCFKFOKUWBPY3ygI/AhXY/OKdTs6q0lpPzMslMiffuYFWfQlNtT/PuonHp3Hb2NN4q38+Tn+zxQbReysgDID9uH5fMC8KpIINxVEkLAW9MOx+yZ8N790JXu9XRCCFE6HBUQ+JIiPNyiYU/9CRwVdbGIfwmvBO4zlbYvzlip0+aCrJtbN3bRGd3AKczpk8w5mCH6Dq4dXsOU9PQytIiH9xZK3sRYhLguLN7nvrGSRM57bgsfvPqlsAn133UJxoJ0CUTWklLCKFKrVobF1CpQDl8SsEZPzcu8sWPWR2NEEKEjmDsAWfq6QUnI3DhKrwTuL0bQXdHbAVKU77dRke3k8/3B7CQiVLGKNye1SHZa2plaQ2JsdF8acZo7w7U3QWbV0LeWRB/tLJjVJTiD18rYkRyLDc/s56WduvWID29RdOmYzklo8GyGIal+QB0tYFNRuC8Muk0yD0ZPvidrJcQQgh3BWMPOFNaNqCkEmUYC+8ErtZVwCRCK1CazAbU5YFeBzdhETTvh0O7AnteL3V0OXl9017Oyh9NcnyMdwfb8xG01PVMn+xtZHIc9106m4qDLdy5ssy78wxTR5eTpz6r4kDcONJbKiyJYdjMqSEyAucdpeD0O43f008fsjoaIYQIDcE8AhcdC6ljZA1cGAvvBK6mBFKzIW2s1ZFYasLIJFLiY6ypRAkhtw7u/c/raDjS6aPpkysgNtkYgevHCZMyuOWMPFasr+GF4sDfKXt9017qmtpJGDutp5l3yOhp4i0jcF4bfzxMPRtW32dULhNCCDGwNge0NwZvAgeuXnCyBi5chXkCVxzx69/AmK43IzuNspoAJ3CZeZCUGXLr4FaW1jAyOY6T8jK9O1B3J2xZBdPOhbikATe7+fQ8Tpg0kjtXlrHjQLN35/SA1pp/rt7NpKxksnILoGFPaBWyMBO4YJ3CEmpOv8N4U7L6fqsjEUKI4BbMPeBMaXZZAxfGwjeBaz0Mh3YaFdYEBdk2Nu9tpCuQhUzMdXAVoTMC19TWydub93P+zLHERnv532PXe8bvYf6Fg24WHaW479LZJMZF891nSmjr7PbuvG5aX9XAhmoHyxbmojKngnaG1nRXR5XRIiQhzepIwsOYQmOq76cPQdN+q6MRQojgFcw94Ey2HCPOEKxDIIYWvglc7XrjMcILmJgKc9Jo63Syqz7ARQpyTwJH5dHRkiD3Ztk+2rucLPHV9Ml4G0w5Y8hNR6cl8IevzWLrviZ+89pm78/thn+uriA1PoaL5uQYo6UQWtMoG6SFgM8t/pkxCvvhH6yORAghgldPD7ggHoGz5RiFvo4csjoS4QduJXBKqbOVUtuUUjuUUrcPst18pVS3UuriXs9VKKU2KaVKlVLrfBG0W2rMAiYyAgfGCBwQ+GmUPf3gPg7seYfp5dJaxo9MYs74dO8O1NkGW1+F6edDjHt95E47bhTXnzKJpz6p5I1Ne707/xD2Odp4Y9NevjZ/nFGoJWOK8UIoJXCOKqlA6WuZU2D2FbDu0ZC56SKEEAHnqIGoWEjxslK1P0kvuLA2ZAKnlIoGHgTOAWYAlymlZgyw3W+Bt/o5zGla6yKt9Twv43Vf7XoYORkS0wN2ymA2KSuFhNgoygJdiXLUDGOaWwgUMjnQ2MaanfUsLcpGedv4fec7xgLngsGnT/b1o7OOY9a4dH7y4kaqDh3xLoZBPPXJHrq15poTc40n4lMgLQfqd/jtnD6ltZFgSAVK3zv1NmP683u/tToSIYQITo5qo0BeVBBPZJNecGHNnd+8BcAOrfUurXUHsBxY0s92NwMvAgd8GN/w1RTL9MleoqMUM8amBb4SZVQ0jF8YEgncqg21ODUsme2j5t2JI2HiqR7tFhcTxQOXzgYNtyxf75fm622d3TzzWSVnTBvN+IxexVUyp4TOCFzrYehoDu71B6HKlgPzvwkbnoG6EPl9EEKIQArmHnAmMz7pBReW3Eng7EDv8ddq13M9lFJ24KtAf02ENPBvpVSxUur64Qbqkca90LRXKlD2UWC3sbm2EaczwAtaJyyEgzuCvjDCy6W1FNptTM5K8e5AHS2w7Q2YcYHRi8VD4zOSuPeimayvbOD3/97mXSz9WLWhlkMtHVy7KPeLL2ROhfrtobHguacHnEyh9IuTfgAxifDu3VZHIoQQwSeYe8CZkjMhOl4SuDDlTgLX31yyvu/w/gTcprXur3zeIq31HIwpmDcppU7p9yRKXa+UWqeUWldXV+dGWIMwG3jLCNwXFNhtNLd3UXEwwIVMzH5wlcHbTmBnXTObahwsKcr2/mCfvwWdR/pt3u2u82aO5fLjx/O393fx3jbfDWprrXlsdQVTR6ewcHLGF1/MnAodTUbz9WDX0wMuyO+AhqqULDjxJti8EmpLrY5GCCGCh7PbmJYY7AmcUmCzSwIXptxJ4KqB3u+ScoDaPtvMA5YrpSqAi4G/KKWWAmita12PB4CXMKZkHkNr/bDWep7Wel5WVpYn38OxaoohKsYoiy169BQyqQ3wOrixM41m1kHcTuDl9TVEKbhglg8SuPIVxsJmM3Edpp+fP4NpY1L54XMb2N/Y5n1cwGe7D7F5byPLFk48dp1fKBUyaTArgMkInN8s/C4kpMN/f2N1JEIIETya9oHuDv4EDqQXXBhzJ4FbC+QppSYqpeKAS4FVvTfQWk/UWudqrXOBF4DvaK1XKqWSlVKpAEqpZOAsoMyn30F/akpg1HSITfT7qUJJ3ugU4qKjKA90JcroWBh/fNA29NZas7K0loWTMxmVluDdwdoa4fN/w4ylxvo/LyTERvPny2dzpKOb7y0vpdsHU18fW1OBLTGWr/a3zi9zqvEYCgmco8q4KZA00upIwleCDU76Puz4T9D+3xVCiIALhR5wJts4GYELU0MmcFrrLuC7GNUltwDPaa3LlVI3KKVuGGL30cBHSqkNwGfAa1rrN70NeoiAYd9GmT7Zj9joKKaNTWVToBM4MNbBHSgPyn4k66saqDx0xDfTJ7e9Ad3tHlefHMiUUan8ckk+H+86yIPvelchsqahlbfK93HpgnEkxvWTXKZlG0lR/XavzhMQZgVKb6uFisEtuN4YTX7nV6GxNjKIKKXGKaXeVUptUUqVK6Vu7WcbpZS639WiZ6NSShZuCxHsQqEHnMlmN2pCdHdZHYnwsRh3NtJavw683ue5/gqWoLVe1uvzXcAsL+LznFLwvU3Q4b8S7KGswG7j1Q21aK29L5XviZ51cJ/AtHMDd143rFxfQ3xMFGcXjPH+YOUrjHL8Of3OFB6WS+bmsGZHPX96+3OOnziS4ydlDL1TP574uAKAq83WAX0p5apEGSIJXCjc/Qx1cUlwyo/h9R/Bjncg70yrIwolXcAPtdYlrpkoxUqp/2itN/fa5hwgz/VxPPBX16MQIliZI1ppPqhY7W+2HNBOI4mTNeNhJYgbWHghLtlYhC+OUZBto7Gti+rDrYE9sX2uUQ0pyNoJdHY7eXXjXs6cPprUBM8rRn5B62HjTW7+Up/2hlFK8ZuvFjIhI5lbl5dyqKXD89A6uln+WRVfzh+DPX2QqcVmJcpg56iSi1GgzLnGqPb5zi/B6fu2FuFKa71Xa13i+rwJYwZL33d8S4AntOETIF0pNTbAoQohPOGoNqaYJ6RZHcnQ0lyjhLIOLuyEZwInBlRgN/7glAV6GmVMPOTMD7oE7qPt9Rxq6WCpL3q/bXkVnJ1eVZ8cSEp8DA9cNptDLR38+PkNaA+ns720vgZHayfLFuYOvmHmVHBUBvcIdnuTkSxLC4HAiImDxT8zpqZvWTX09uIYSqlcYDbwaZ+XhmzTI4QIMqHQA85kTvOUdXBhRxK4CDN1dCoxUSrwDb3BWAe3d4PxBjxIrCytIT0pllOn+mDEtuxFGJEL2bO9P1Y/Cuw2fnbuNN7ZeoBHV1e4vZ/WmsfW7GbG2DQWTByi6IdZifLQzuEH6m89FShD5AIaDmZ+DTKPM/rCyVoKjyilUoAXge9prfuWAHanTY9v2+wIIbwTCj3gTDbX/SBJ4MKOJHARJiE2mrzRqWyqCXArATASOO2Eqr43oa3R0t7Fv8v3c27hWOJivPyv0FwHuz+A/Av9WljjmoW5fGnGaO59Ywsbqxvc2mfNzoN8vr+ZZYtyh173GAqVKKWJd+BFRcPpdxi/Fxv/ZXU0IUMpFYuRvD2ttV7RzybutOnxbZsdIYR3HFWhk8DFpxrTPSWBCzuSwEWgQnsa5TUOj6fheW3cAqM/X5D0g/vP5v20dnaztMgX0ydfNvrC+GH6ZG9KKX538UyyUuK5+dn1NLV1DrnPP1dXMDI5zr0edxmTARXc6+B6mnhLAhdQ078CY4vgvXuhq93qaIKeMu6WPAJs0Vr/3wCbrQKudlWjPAFwaK33BixIIYRn2pugrSF0Ejgw1sHJGriwIwlcBCqw2zjY0sE+HzWHdltcsjG9MEh6Sq0srcGensi8CSO8P1jZS8bo1eh87481hPSkOO6/bDbVh1v52Utlgybiew628M7W/Vy+YDwJsW70pYtNNBKjYE7gHFUQHQfJo6yOJLIoBWf83FgjWfy41dGEgkXAVcDpSqlS18e5fVrwvA7sAnYAfwe+Y1GsQgh3OFyJUChN4bflHJ25IsKGJHARKD/bBkCZVdMoa4qhM8BVMPuob27nw+31XFCUTVSUl1MeG/caxVkKLgpYX7J5uSP5wZem8sqGWv61duA/zE98vIdopbjyhAnuHzwzL7inUDZUGhckH1b6FG6afDpMOAk++B10tFgdTVDTWn+ktVZa65la6yLXx+ta64fMNjyu6pM3aa0na60LtdbrrI5bCDGInibeITQCZ7MfTTxF2JB3QBFoxtg0ohQWNfQ+yajUWL028Ofu5bWNe+l2at9Mn9y8EtDG+rcAuvHUyZw0JZNfvFLO5/uPLQzT3N7Fc2urOKdwLGNsCe4fOHMqHNwRvCXjG6pC6+5nOFEKzrgTWg7Ap3+zOhohhAisUGribbLlQOuh4K4uLTwmCVwESoyLZsqoFMqtSODGHw8oy6dRriytYdqYVI4bk+r9wcpehNGFkDXV+2N5ICpK8X9fn0VKfAw3PV1Ca0f3F15fUVJNU3sX1y7K9ezAmXnQeSR458w7qmT9m5XGnwB5X4bVf4LWBqujEUKIwGmsARUNKWOsjsR90gsuLEkCF6EKsm3WtBJIsMGYQkv7wVXUt7C+soGv+qL3W0OlMZpY8FXvjzUMo1IT+OPXi9hR18wvXynved7p1Dy2poJZOTZmj0v37KAZecbjwSBcB9fZBs37JYGz2ul3QJsD1jxgdSRCCBE4jmpIy4boGKsjcV9PLzhZBxdOJIGLUPl2G/sb2znQFOBCJgATFkHVWujqCPy5gZdLa1EKLihyoyrjUMpfMh4DPH2yt5Pzsrjx1MksX1vFqg1GBfIPttexq67FvdYBffW0EgjCBK5n/YFMobTU2Jnwlfth3nVWRyKEEIHjqIY0H9z8DaSeXnAyAhdOJIGLUAXZaQCU11pQyCR3EXS1Qu36gJ9aa83LpTUcP3EkY22J3h+w7EXIngMjJ3p/LC/84EtTmTthBD9bsYk9B1t4bE0FWanxnFc4jCQ1ZRTE24KzkIlDWggEjbnXHH1jIIQQkSCUesCZUrMBJb3gwowkcBFqhiuBK6u2Yh3cicajBdMoN9U42FXf4pviJQd3wt4NUGDd6JspJjqK+y+bTXSU4rrH1vLetjquOH788BqUKwWZU4JzBK7BbOItI3BCCCECyOk0RrFCLYGLiYOU0dAoCVw4kQQuQqUmxDIpM9madXDJmZA1zZJCJivX1xIXHcU5hWO9P1jZCuMx35r1b33Z0xP534tnsrOuhdhoxRXHe9A6oK/MqUGawFUaC8hTfTD9VQghhHBXywGjinaoJXDg6gUnCVw4CaFVmMLX8u02SvYctubkExbCxuehuytgi4G7nZpXNtZy2rQsbImx3h+wfAWMOyGo/ph/OX8Md54/AwVkpcYP/0CZebDhWWhvgngfVOr0FUdV6C0gF0IIEfpCeQ22zQ77N1sdhfAhGYGLYAXZadQ0tHK4xYJiIhMWQUcT7N8UsFOu2VlPXVO7b6ZPbnweDmw2mncHmW+cNJHrTvJyTZ5ZyOTgDu8D8qUGaSEghBDCAqHYA85kG2ckoFpbHYnwEUngIliB3QZgzTTKCQuNxwBOo1y5vpbUhBhOmzbKuwNtXgUvfdtIQudc5Zvggo3ZSiDYplE2VIbm3U8hhBChrWcELgQTuDS7UTyu1aJZV8LnJIGLYAXZrgSuxoJKlGnZMGJiwBK41o5u3irfxzkFY0iIjR7+gT7/N7xwHdjnwOX/glgfVLIMRiMnGmvNgqkSZXcnNNVKARMhhBCB56iGuFSjn22okV5wYUcSuAhmS4pl3MhEa0bgwGgnsGe1UdnJz97esp/m9i6WetO8e+e78K8rYfQMuOKF4Fob5msx8TAiN7hG4BprQTtlCqUQQojAc1QbiZCnvVWDgfSCCzuSwEW4gmwb5TUWJXATFhnD+XVb/X6ql0trGJOWwAkTM4Z3gD1r4NnLIGMyXLUSEtN9GV5wyswLrgSuZ/2BjMAJIYQIsFDsAWcyr5tSiTJsSAIX4QrsNioOHqGxrTPwJ+9ZB+fffnCHWzp4b1sdFxRlExU1jDtn1evg6UuMP9xXvwxJI30fZDDKzDOKmDi7rY7E0CBNvIUQQljEHIELRUmZEB0nveDCiCRwES7f1dC73Ip1cOkTjIW1fl4H99qmvXQ5NUuKhtE7bO8GeOpCo3fdNasgxcsCKKEkcyp0tx9NnKxmNvFO80EVUSGEEMJdHUfgyMHQTeCiooxrp4zAhQ1J4CKcWYmy3Ip1cEoZ0yj3rPZraduXS2vIG5XCjLFpnu14YAs8sdRYtHz1KqPwSiQxK1EGSysBRyWkjIHYBKsjEUIIEUkaXWvHQnkKvy1H1sCFEUngIlxmSjxjbQmUWbYObiE074dDu/xy+OrDR1hbcZils+0oTxYe1++Axy8wphxcswpGTPBLfEHN7AUXLJUoGyqlAqUQQojAC+UecCZbjozAhRFJ4AT52TbKai2YQgnGCBz4bR3cy6W1AFwwy4PRs8MV8MQFoLuNNW8Zk/0SW9BLzoDEkUGUwFWF9t1PIYQQoSmUe8CZ0uzQtBe6u6yORPiAJHCCAnsaO+uaaWm34D91Zh4kZ0GF7xM4rTUvl9Ywb8IIxo1Mcm8nR40x8tbRYiRvo6b5PK6QkjnVGI20mtNpTGGRAiZCCCECzVEDqNBeSmHLMW5MN++zOhLhA5LACQrtNrSGLXstGIVTyphG6YdCJlv2NvH5/mb3e7817TdG3loPw1UrYEyhz2MKOZlTgmMErnk/dHfIFEohhBCB56iG1LEQHWt1JMPX08xb1sGFA0ngRE8hk3e2HrAmgAmLjAIVPq52uLK0hpgoxXmFY4feuOUgPLHEaBZ9xfNgn+vTWEJW5lRoOQCtDdbGYf5u2GQETgghRICFcg84U08CV2VtHMInJIETjE5L4IJZ2fz1vZ28WGzBAteefnAf++yQ3U7NqtJaFh+XxYjkuME3bj0MTy6Bw7vhsuUw/gSfxRHyzEImVleiNC84MgInhBAi0BzVYAvxFjZmC55GGYELB5LACQB+d8lMFk3J4CcvbuS/W/cH9uSj8iHBBns+8tkhP919kH2NbSwpGuIPblsjPHURHNgKX38aJp3qsxjCgtlKwOpplD0jcJLACSGECCCtQ7uJtykhDeJtUokyTEgCJwCIj4nmb1fNY8bYNL7zdAnFew4F7uRRUTDet+vgXl5fS3JcNGdOHz3wRh0t8MzXobYULnkM8s702fnDxogJEBUbHAlc4kiIT7E2DiGEEJGlpR6628PjBqLNLmvgwoQkcKJHSnwMj107n2xbItc9to7P9zcF7uQTFhrT9Jq8H/1r6+zm9bK9fLlgDIlx0f1v1NkGyy+Hqk/gor/D9PO9Pm9Yio6FkZOgfru1cTiqZPqkEEKIwAuHHnAmW46sgQsTksCJL8hIiefx6xaQEBvF1Y98RvXhI4E5sdkPrtL7Ubj3th2gqa2LpQNNn+zqgOeuhl3vwZIHoeAir88Z1jLzrE/gGqqkhYAQQojAC4cecKY0u6yBCxOSwIljjBuZxOPXLeBIRxdXP/IZB5vb/X/SsbMgNtkn/eBWrq8lMyWehZMzjn2xuwtevA62vwXn/R8UXe71+cJeZh4c2gXdndacX2tXBTBJ4IQQQgRYTwIXBrNAbHY4chA6W62ORHhJEjjRr2lj0nh02XxqGlq57rG1/m/yHR0D44/3eh2co7WT/249wFdmjSUmus+vt7MbVt4AW16BL98D87/h1bkiRuZUcHbC4T3WnP/IQeg8IlMohRBCBJ6jGmKTIHGE1ZF4z0xCZR1cyJMETgxoXu5I/nLFHMpqG7nhqWI6upz+PeGEhXCgHI4Mv4DKm2V76eh28tW+zbudTnjlFtj0PJzxczjxO14GG0F6WglYNI1SKlAKIYSwitkDTimrI/FeTysBqUQZ6iSBE4M6Y/pofnvRTD7cXs8PnivF6dT+O9mEk4zHyuH3g3tpfQ2TMpMpdDUnB4wpeG/8GNY/Baf8BE7+oZeBRpiMKcajVZUoe3rAyRRKIYQQARYOLQRMPc28JYELdZLAiSFdPDeHn507jVc37uWXr5SjtZ+SOPsciI4f9jTKvY5WPt19iCVFdpR5p0xr+PcdsPYfsPBmOO1nPgw4QiSmQ/Io6xK4BmniLYQQwiLhlMClZRuPMoUy5MVYHYAIDdefMpn65g4e/mAXmSnx3HxGnu9PEhMPOfNhz/AKmawqrUVrWFKUffTJd++Gj/8M878FX/p1eEyBsELmVKjfYc25GyohLhUS0q05vxBCiMjU2QYtB8JnCn9MPKSMllYCYUBG4ITbbj97GhfNyeEP//mcpz/1U0GLCQth7wZo97wH3crSWorGpZObmWw88cHv4YPfweyr4Jz/leTNG5lTrJ1CmT5e/v2EEEIElllyP1xG4EBaCYQJSeCE26KiFPdeVMgZ00Zxx8oy3ti01/cnyV0E2gmVn3q02+f7m9iyt5Gl5ujbx3+B//4aCi+Br9wHUfKr7pXMqdB6CFoOBv7cDdLEWwghhAXCqQecyZYja+DCgLyrFR6JjY7iz5fPYe74Edy6vJQ1O+t9e4Kc+RAV4/E0ypXra4iOUpw/KxvWPgJv/RSmXwBLH4KoaN/GGImsrETZUBk+01eEEEKEjrBN4GqMGgEiZEkCJzyWGBfNI9fMZ2JmMtc/UUxZjcN3B49LhuzZHhUycTo1L5fWctKUTDK3vwCv/QDyvgwXPWL0lxPes6oSZZsD2h0yAieEECLwzKmGafbBtwslthzobIHWw1ZHIrwgCZwYFltSLI9ftwBbYizXPPoZu+tbfHfwCYugphg6W93avLjyMDUNrdyUVQqrvguTFsPXnoCYON/FFOnSxxsVQgOdwDVICwEhhBAWcVQZRT9i4q2OxHd6esHJOrhQJgmcGLYxtgSe/MYCNHDVI59yoLHNNweesAicnVC91q3NV66v4fzYEuaX3A7jToBLn4HYBN/EIgxR0cYoXH2Ap1CalbJsksAJIYQIsHBqIWAylyTIOriQJgmc8MqkrBQeu3Y+h1s6uPrRz3C0dnp/0PHHA8qtaZQdXU4aNr7On6L/hMqeDVc8Z0zDFL6XmRf4BK6h0niUKZRCCCECLSwTONcInCRwIU0SOOG1mTnp/O2qeeysa+Zbj6+jrbPbuwMm2GBMoVuFTDZ+uIo/OH9H64ipcOWLEJ/q3bnFwDLz4HAFdLUH7pwNlRCTAMlZgTunEEIIobWR5KSFWQKXPAqiYiWBC3GSwAmfOCkvkz9+vYi1ew7x3WfW09Xt9O6AuSdB1Vro6hh4mz0fU/jBt6lWY0i4bhUkpnt3TjG4zKmgu+HQ7sCd01FlTPeQHnBCCCECqfUwdB4JvxG4qChIy5Y1cCFOEjjhM+fPzOZXF+Tz9pb9/OylTWhvStROWAhdrVC7vv/Xa4rRT19MrXMEKwoeJDZVRmj8LjPPeAxkK4GGSpk+KUKSUupRpdQBpVTZAK8vVko5lFKlro+fBzpGIcQgetZgh1kCB8aNURmBC2luJXBKqbOVUtuUUjuUUrcPst18pVS3UupiT/cV4eGqE3O59Yw8nltXzf++tW34Bxq/0Hjsbxrlvk3w5IW0xKRzWfvPOGP+zOGfR7jPilYCDVXSA06EqseAs4fY5kOtdZHr41cBiEkI4a5w7AFnstmNXnAiZA2ZwCmlooEHgXOAGcBlSqkZA2z3W+AtT/cV4eV7Z+ZxxfHj+et7O/nHh7uGd5DkDMiafmwhkwNb4YklEJfMnbb/R9zIccwZn+51zMIN8amQmh24QiYdR+BIvbQQECFJa/0BcMjqOIQQw9STwIXhTURbjjGF0ullzQJhGXdG4BYAO7TWu7TWHcByYEk/290MvAgcGMa+IowopfjVkgLOLRzDb17bwkvrhzlMP2EhVH4C3V3G1wd3whMXQFQMBy9+gZcrYlhSlI2S9VGBE8hKlObFUxI4Eb5OVEptUEq9oZTKtzoYIUQvjiqj/2lyptWR+F6a3VjT3rzf6kjEMLmTwNmBql5fV7ue66GUsgNfBR7ydF8RnqKjFH/8ehELJ2fw4+c38u7WA0Pv1NeEhdDRBPs3weE98PgF4OyCq19mZWUiTg1LiuTXKaDMBM6b9Y3uMlsIhOPdTyGgBJigtZ4FPACsHGhDpdT1Sql1Sql1dXV1gYpPiMhmthAIx5vE0gsu5LmTwPX3m9v33dufgNu01n3HYt3Z19hQLlBhJz4mmoevnse0sanc+HQxxXsOe3aACYuMx00vwONfMZK5q1bCqOm8XFpDod3GlFEpPo9bDCJzKrQ7oHkYCbmnHGYPOBmBE+FHa92otW52ff46EKuU6vdWv9b6Ya31PK31vKwsKdgkRECEYw84k/SCC3nuJHDVQO9b4DlAbZ9t5gHLlVIVwMXAX5RSS93cF5ALVLhKiY/hsWsXMCYtgeseW8vn+5vc3zltLIycBB//GY4cgitfgrEz2VnXzMZqB0uKsv0XuOifWYkyEIVMGqogKgZSx/j/XEIEmFJqjHLN/1ZKLcC4Hh+0NiohRA9HdfjOADETU0ngQpY7CdxaIE8pNVEpFQdcCqzqvYHWeqLWOldrnQu8AHxHa73SnX1F+MtMiefJbxxPfEwUVz/yGTUNre7vPOVMiE2CK56HnLkAvLy+higFF8ySBC7gMqcaj4FoJdBQaczTj4r2/7mE8DGl1LPAx8BxSqlqpdQ3lFI3KKVucG1yMVCmlNoA3A9cqr3qvSKE8JmuDmjaF74jcAk2iEuVXnAhLGaoDbTWXUqp72JUl4wGHtVal5sXIa1133VvQ+7rm9BFKBk3MonHr1vA1/72MVc98ikv3LCQkclxQ+/4pV/DqbcbVSkBrTUrS2tZODmTUWkJfo5aHCM120ioA1HIxFEl0ydFyNJaXzbE638G/hygcIQQnmiqBXT4JnBgfG8yAhey3OoDp7V+XWs9VWs9WWt9t+u5h/pL3rTWy7TWLwy2r4hM08em8cg186k53Mq1j62lpb1r6J1iE3qSN4D1VQ1UHjoi0yetEhVl9IML1BRKSeCEEEIEWjj3gDPZ7JLAhTC3EjghfGXBxJE8ePkcymoc3PBUMR1dTo/2f3l9DfExUZxdIOuiLJM51f8jcF0d0LQ3fNcfCCGECF7h3APOJCNwIU0SOBFwZ84YzT0XFvLh9np+9PwGnE73ln10djt5deNezpw+mtSEWD9HKQaUmWesT+v0YC2jpxqrAQ3pYXzxFEIIEZx6ErgwblWUlgNH6v17LRd+IwmcsMTX5o3j9nOmsWpDLb96dTPurN3/aEc9B1s6ZPqk1TLzAG00VveXBlf7SJlCKYQQItAc1ZCUCbGJVkfiP+b00MZ+i8OLICcJnLDMt0+ZxLdOnshjayp48N0dQ27/8voa0pNiWXzcqABEJwYUiEqUDlcCF87TV4QQQgSncO4BZ5JecCFtyCqUQviLUoqfnjOdg80d/P7fnzMyOZ7Lj+9/xOVIRxf/3ryfpbPtxMXIfQdLjZxsPPpzHVxDJaCMNgJCCCFEIDmqIWOy1VH4l/SCC2mSwAlLRUUpfnvxTA4f6eCOlZsYmRzL2QVjj9nuP5v3c6Sjm6VF8obecnFJYBvv30qUDVWQOhZi3Gg1IYQQQviK1sYskEmLrY7Ev8wbpNILLiTJUIawXGx0FH+5Yi6zx4/glmdLWbOz/phtXlpfgz09kXkTRlgQoThGZp5/EzjpASeEEMIKbQ7oaA7vAiYAMfGQPOrokgURUiSBE0EhMS6aR66ZR25mEtc/UUxZjaPntfrmdj7cXs8FRdlERSkLoxQ9MvOgfodxp9IfGiqlAqUQQojAi4QecCabHRwyAheKJIETQSM9KY7Hr1uALTGWZf/8jIr6FgBe27iXbqeW6ZPBJDMPOlv8U73K2W1M6ZACJkIIIQItEnrAmaQXXMiSBE4ElbG2RJ74xgKcGq5+9DMONLaxsrSGaWNSOW5MqtXhCZNZidIf0yib9oKzS6ZQCiGECLyeKsgRMAKXlmPcMPXXbBrhN5LAiaAzOSuFfy6bT31zO5c+/AnrKxtYOltG34JKTyuBods/eKynB1wE3P0UQggRXBzVEBVrrA8Ld7YcY71fW4PVkQgPSQIngtKscen87aq5VB0+glJwwSxp3h1UUkZDXKp/RuAaKo1Hm4zACSGECDBHtbE2LCoC3iL39IKTdXChRtoIiKB1cl4Wf796HhX1LWSnJ1odjuhNKf9VonSYCVwETF8RQggRXBzVkbH+DY5+n45qGFNgbSzCI5LAiaC2+LhRcJzVUYh+ZU6Fio98f9yGKkjOMvrNCSGEEIHkqIaJJ1sdRWD09IKTQiahJgLGh4UQfpE5xfij397s2+M6qiLn7qcQQojg0d0FTbWRMwMkZbSx3k8qUYYcSeCEEMPjr0Im0gNOCCGEFZr2gnZGTgIXFQVpY2UNXAiSBE4IMTz+SOC0Nu4ESgsBIYQQgRZJTbxNtnEyAheCJIETQgzPyEmgonxbyKSlDrrapAKlEEKIwIukJt6mNLusgQtBksAJIYYnJh7SJ/g2gTNbCMgUSiGEEIFmJjJpEdR71pYDjbXg7LY6EuEBSeCEEMOXORXqt/vueD094CSBE0IIEWCOakgcAfEpVkcSODY7OLug+YDVkQgPSAInhBi+zDxjDZzT6ZvjOaqMRxmBE0IIEWiO6sha/wZf7AUnQoYkcEKI4cvMM9asmYmXtxqqIMFmfAghhBCBFElNvE3SCy4kSQInhBg+sxKlr6ZRNlRKARMhhBDWcFRF4Aic6/uVEbiQIgmcEGL4eloJ+CiBc1RJCwEhhBCB19YIbY7IS+ASbBCXMmAvuLIaB/e8voWW9q4AByYGE2N1AEKIEJaUAQnpvqlEqbUxhXLiKd4fSwghhPBEoyuBiaQKlABKGUlrn6UQTqfmHx/t4ndvbaOzW+PUmv85b4ZFQYq+JIETQgyfUr6rRNl6GDqaIm/9gRBCCOtFYg84U5r9aAIL7HO08cPnS1m94yBfzh9NYmw0j66u4MI5OUwfm2ZhoMIkUyiFEN7xVQInFSiFEEJYxbwGRdoUSnCNwBkJ7Fvl+zj7vg8o2dPAvRcW8tCVc7nrK/nYEmO5Y2UZTqe2OFgBksAJIbyVOQWa9xlrB7zRYCZwsgZOCCFEgDmqQUVD6hirIwk8Ww601HHnC+v49pPFjBuRxGu3nMSlC8ajlGJEchy3nzON4j2HeaFYip0EA0nghBDe6alEucO74/Tc/ZQETgghRIA5qo2phFHRVkcScNXOkQB8WLyBG06dzIs3LmRS1hebmV88J4f5uSO4540tHG7psCJM0YskcEII7/QkcF4WMmmohNgkSBrpfUxCCCGEJyKwibfTqXno/Z389J1DAPz5/FHcfs404mKOTQ+iohS/XlpAY1sXv31za6BDFX1IAieE8M6IXIiK8b6VQEOlMX1SKZ+EJYQQQrgtwnrA7XO0ceUjn3LvG1uZOGkaAAXJTYPuM21MGt84aSLL11ZRvOdQIMIUA5AETgjhnehYGDHR+xE4R1VkVv8SQghhLWc3NNZGTAL3ZplRqGR9ZQO/vaiQX171JeOFAXrB9XbrGXmMtSXwPy+V0dXt9HOkYiCSwAkhvOeLSpQNlVKBUgghROA17wdnV9gncEc6urj9xY3c8FQx40cahUq+Pn88KjYRkrOO6QXXn+T4GO76ygy27mvisTUV/g9a9EsSOCGE9zLz4NAu6O4a3v7tzUYfOBmBE0IIEWgR0ANuU7WD8+//iH+tq+LGxZN54YY+hUr69IIbzJfzx3DacVn88T+fs9fR6qeIxWAkgRNCeC8zD7o7oGHP8PZ3SAsBIYQQFgnjHnBmoZIL/7qa1s5unvnmCdx2dj+FSnr1ghuKUopfXlBAl1Pzm1e3+CFqMRRJ4IQQ3uupRDnMaZTSA04IIYRVekbgwiuB2+to5Yp/GIVKvjRjNG/cejInTs7of2MzgdPuNeoen5HEd0+bwmub9vL+53U+jFq4QxI4IYT3MqYYj8OtRGmO3IXx9BUhhBBBylED8TZISLM6Ep95Y9Nezv7Th2yobuB/L57Jg5fPIT0pbuAd0uzQ0QxtDrfPcf2pk5iUmczPXy6jrbPbB1ELd0kCJ4TwXtJISMocfiVKRxVEx0HKaN/GJYQQQgwljHrAtbR3cdsLG7nx6RJyM5J47ZaT+dq8caihWvSY37+b6+AA4mOi+fXSAvYcPMJf39vpRdTCU5LACSF8w5tKlA2u/jtR8idJCCFEgIVJD7iN1Q2c/8BHPFdcxXcWT+aFGxcyMTPZvZ3N79+NVgK9LZqSyQWzsvnrezvZXd/iYcRiuOTdkhDCNzLzvBuBk+mTIowopR5VSh1QSpUN8LpSSt2vlNqhlNqolJoT6BiFEC4hPgLX7dT85b0dXPiXNbR1dvPst07gJ2dPIzbag7f5PQnc0K0E+rrjvOnEx0Tx85fL0G6uoRPekQROCOEbmVPhyEE4csjzfaUHnAg/jwFnD/L6OUCe6+N64K8BiEkI0VdHC7QeCtkEzihU8gn/++Y2vpw/hjdvPYUTJg1QqGQwKaMhKsajKZSmUWkJ/PCsqXy4vZ7XNu31/NzCY5LACSF8IzPPePR0GmVnm9FENX2C72MSwiJa6w+Awe5mLAGe0IZPgHSl1NjARCeE6GFOGQzBWSBmoZKN1Q5+d/FM/nz5bGxJscM7WFQ0pGa73Uqgr6tOzCU/O41fvbKZprbO4cUg3CYJnBDCN3oSOA+nUTaG7sVTCC/Ygd5zlapdzwkhAqmnB1zo/Pdrae/iJy9s6ClU8votJ3OJO4VKhmKze7wGzhQdpbj7q4XUNbfzx/8Mcz28cJskcEII30ifYFSS9LSVgNlCQKZQisjS3zutfhePKKWuV0qtU0qtq6uTfktC+FSI9YDbUGUUKnm+uJrvnjaFF25cSK67hUqGYssZ1ho4U9G4dC5fMJ7H1uymvNb9dgTCc5LACSF8IyoaRk72fAql2cRbRuBEZKkGev/S5wC1/W2otX5Yaz1Paz0vKysrIMEJETEc1aCiIDW4ZzB3OzUPvruDi/66hvbObpZ/6wR+9OXjPCtUMpQ0OzTWgtM57EP85MvTGJEUxx0ry3A6paCJv0gCJ4TwneFUonRUgYo2LhxCRI5VwNWuapQnAA6ttaz+FyLQHNVG8hY9zLVjAVDb0Mrlf/+E3721jS8XjOGNW0/h+OEUKhmKLQecndByYPiHSIrlZ+dOZ31lA/9aN/zRPDG4GKsDEEKEkcypsO116O50/2LYUAVp2RAtf45E+FBKPQssBjKVUtXAXUAsgNb6IeB14FxgB3AEuNaaSIWIcEHeA+61jXv56YqNdDs1v79kFhfNsXu/1m0gvXvBpY4Z9mEunGPnX+uquPeNrZw1YzQZKfE+ClCY5B2TEMJ3MvPA2QWHdkPWVPf2aaiU6ZMi7GitLxvidQ3cFKBwhBADcVSDPfjaMLa0d/GLVeU8X1zNrHHp3Pf1It+tdRtI715wOXOHfRilFL9ZWsC5933IvW9s5XeXzPJRgMIkUyiFEL4znEqUjipIH++feIQQQoiBOJ1GJeQgG4ErrWrgvPs/5MWSam4+fQov3HCi/5M3OLqUYRi94PqaOjqVb548ieeLq1lbMYz+sGJQbiVwSqmzlVLblFI7lFK39/P6EqXURqVUqatS1km9XqtQSm0yX/Nl8EKIIJPhSuDcrUTZ3WUsmJYKlEIIIQKtpQ66O4JmFohZqOTiv66hs1uz/PoT+eFZPi5UMpjEERCbPOxecH3dcsYU7OmJ3PFSGZ3dwy+MIo415G+EUioaeBA4B5gBXKaUmtFns3eAWVrrIuA64B99Xj9Na12ktZ7nfchCiKCVkGYsBne3EmVjDejuoLl4CiGEiCBB1EKgd6GSswvG8PqtJ7Ng4sjABqGUqxecbxK4pLgY7vrKDLbtb+Kfq3f75JjC4M4auAXADq31LgCl1HJgCbDZ3EBr3dxr+2QG6GUjhIgAGVPcn0Jp9puRETghxP9v787Do6zuho9/z8xk3xfWJJAAYQthCIQdWRptrSC7VSwKUlGpimitT2tttbVer89bH1ttKy2upaVQX2tQfBAVAbGsgrgQ2SFAQEJIIPsyy3n/uCeTEAJJIDOTSX6f65rrnrnX39yZzLl/c859jhDe5h7E27cJ3OaDBSxetQeb3cn/3GJlpic7KmlKVGKrJXAA303ryvUDOvOH9YeYMrg73aNDWm3fHVlz6mQTgPr9gOa55l1EKTVDKbUf+F+MWrhaGvhQKbVbKXXPtQQrhPAD8X2NBE4343ec2jHgont6NiYhhBCiIR/XwDldTSbnvb6TLhHBvLf4OmYNS/Rd8gauseCu/R64+p68OQ2n1vxmzTdNryyapTkJXGOfokuuzLTW2Vrr/sB04Ol6i8ZqrYdiNMG8Xyk1vtGDKHWP6/65XQUFBc0ISwjRJsX3hapiKD/X9Lq1v37KGHBCCCG8reQUBIZDcLT3D11l495/7OZ3Hxzg5sHdyb5/DCne6KikKVFJUJYP9upW22VSbCgPfieVdTln2Lj/6seYE3Wak8DlAfXbNyUCpy+3stZ6M9BbKRXven3aNT0LZGM0yWxsu2Va60ytdWanTp2aGb4Qos2J72NMm9OM8sJxCO8CAcGejUkIIYRoqHYMOC/XeB3ML2X6n7awcf9ZfjVlIC/cNoTQwDYysldUbU+Ul73UvyoLr+tF705h/OrdvVTWOFp13x1RcxK4z4BUpVSKUioQuA14t/4KSqk+ylXfq5QaCgQChUqpMKVUhGt+GPBdYG9rvgEhRBsT7xr/rVkJnAwhIIQQwkeK87zefHLNl6eZ/uctlFTZ+efCUSwYl+LbJpMNuceCa7374AACLSaenj6Ik0WVvLTpcKvuuyNqMt3XWtuVUg8AHwBm4DWtdY5S6j7X8r8As4A7lVI2oBK4VWutlVJdgGzXB9MC/FNrvc5D70UI0RZEJoIlBAqb8QVdfBK6DfF4SEIIIcQlivOgm3cGmbY5nDz7/n5e/c8xhvWM4aUfDqVLZBtsfRLpSuBa+T44gDG945mRkcBfPjnC9IwEencKb/VjdBTNqq/VWq8F1jaY95d6z/8b+O9GtjsKyPDrQnQkJlPzeqJ0Oo3Cc8DN3olLCCGEqGWrNMaB80INXEFpNQ/883N2HCti/phkHr9pAIEWL43t1lK1TSiLT155vav0+E0DWL8vn1+9s5d//Ghk26p99CNt9NMjhPBr8alNJ3Bl+W1qAFUhhBAdSO09Xh4ugz4/cZ4pf/yUL/Mu8PtbrTw1Na3tJm8AASEQGgfFrV8DB9ApIojHvtePLYcLeffL1r3PriNpw58gIYTfiu8LF06Arery6xTLEAJCCCF8xMNjwGmt+fv249z6120EWcy8vWgsMzJ8P2B4s7TyWHAN3T6yJ4MTo/jt/+6jpMrmseO0Z5LACSFaX3wqaCcUHb38OhdOGFMZxFsIIYS31SYoHhjGpsrm4NH/9xW/XL2XcX3iWfPAOAZ2j2z143hMZKJH7oGrZTYpfjt9EOfKqnn+w2Z0eCYuIQmcEKL1xaca0ys1o6xN4KQJpRBCCG8rzgMURHZv1d2eLKpg5ktbeXtPHkuuT+XVecOJCg1o1WN4nIdr4AAGJ0Zzx6ieLN+Wy95TxR49VnskCZwQovXF1Y4Fd+jy6xSfhJBYCJJeqIQQQnhZ8UljHFJLUKvtctOBs0z543/IO1/Ba/OGs+T6vphMfthJR1QCVJdAlWcTq598tx+xYUH8IvtrHE7t0WO1N5LACSFaX2CYUbNWeIUE7sJJaT4phBDCN1pxDDinU/Pix4e4643P6BYVzJoHxzGpf+dW2bdPuMeC81wzSoCokACemDyAL/OKWbnzhEeP1d5IAieE8IymhhK4cEKaTwohhPCNVkrgiitt3PP3XTz/0UGmD0kg+8dj6RkX1goB+pAHx4JraNqQ7ozuFcf/XbefgtJqjx+vvZAETgjhGfF9jSaUupFmEVobzVeie3g/LiGEEB2b1q2SwO0/U8K0P/2HTQcK+PXUNJ7/gZWQQHMrBelD7ho4z4wFV59SiqenD6LS5uD/vL/P48drLySBE0J4Rnwq1JRB6ZlLl1UUga1CEjghhBDeV1EI9qpragXyzhenmPHnrVTUOFh1zyjmjUluP4NSR3QFZfZ4E8pafTqHc8/4Xrz9+Sm2Hy30yjH9nSRwQgjPuFJPlMXSA6UQQggfuYYx4GwOJ79ek8NDq74gPSGK9xaPIzM5tpUD9DGT2eid08M9Udb3wKRUEmNCeGL1XmrsTq8d119JAieE8Iz4vsa0sQROxoATQgjhK7WJSQsTuLMlVdz+8nZe35LLXWOTWbFwJJ0jgj0QYBsQmeCVe+BqhQSa+fXUNA6fLePV/xzz2nH9lSRwQgjPiOgGgeFQePjSZRdcv35KE0ohhBDe5k7gmv8j4q7cIqb88T/sPVXCC7cN4cmb0wgwt+PL6KhEr9wDV1/WgC58d2AXXvz4EHnnK7x6bH/Tjj95QgifUuryPVEWn4TACAiO9npYQgghOrjiPLCEQGjTTR+11ryx5Ri3LdtOSKCZ7PvHMG1IgheC9LGoBCg5DU7vNmd8cmoaAE+9+41Xj+tvJIETQnhObU+UDV04YTSfbC83fAshhPAftT1QNlEGVdY4eOTNL3lqzTdM7NeJdx8YR/+ukV4K0seiksBRA+UFXj1sQnQID12fyvp9+Xz0Tb5Xj+1PLL4OQAjR/thsNvLy8qjq9SPoPhu+yQFV7/ei3vcYN0nvky6DO4rg4GASExMJCAjwdShCiI6uGUMIHC8s596/7+ZAfik/uaEv90/qg8nUgX50jHTVMpbkQUQXrx76R+NS+PfuPJ56N4exfeIIDZR0pSE5I0KIVpeXl0dERATJXaJQF45DfAoEhtat8K0NQmKkE5MOQmtNYWEheXl5pKSk+DocIURHV5wHqTdcdvGG/fksWfUFSilenz+cif06ezG4NsI9FlweJAzz6qEDzCZ+O30Qty7bzp82HOaxG/t79fj+QJpQCiFaXVVVFXFxcaiAEGOGvapuodMO2gGWQN8EJ7xOKUVcXBxVVVVNryyEEJ5kr4ayM412YOJ0an7/0UEWvLGLxJhQ3ntwXMdM3qBeAue9nijrG9krjllDE3n506McPlvqkxjaMknghBAeoZQCc5Dxwl5dt8BhM6ZmSeA6knYzwK0Qwr+VnDamDZpQFlfY+NHfPuOFjw8xa2gib/94DEmxoY3soIMIiYGAUK+OBdfQz2/qT2ighSdW70Vr7bM42iJJ4IQQnmMyGYnaRQlcjTH1UAJXWFjIkCFDGDJkCF27diUhIcH9uqam5orb7tq1i8WLFzd5jDFjxrRWuAA89NBDJCQk4PRyb19CCNHhNDIG3DenS7j5T//hP4fP8fT0QTx3y2CCA8w+CrCNUMo1FpzvErj48CAeu7Ef248WsfoL39QEtlVyD5wQwrMswRc3obR7NoGLi4vjiy++AOCpp54iPDycRx99tO7wdjsWS+NffZmZmWRmZjZ5jK1bt7ZKrABOp5Ps7GySkpLYvHkzEydObLV91+dwODCbO/gFiRBCNEjgsvfk8fO3vyYqJIBV94xmWM8YHwbXxkQl+rQGDmDO8B68uSuPZ/53H9/p14WoUOkIC6QGTgjhaZYgowautvmDowZQYPLe70fz58/nkUceYdKkSfzXf/0XO3fuZMyYMWRkZDBmzBgOHDgAwKZNm5gyZQpgJH8LFixg4sSJ9OrVixdffNG9v/DwcPf6EydOZPbs2fTv358f/vCH7mYea9eupX///owbN47Fixe799vQxo0bGTRoEIsWLWLlypXu+fn5+cyYMQOr1YrVanUnjcuXL2fw4MFYrVbuuOMO9/t76623Go1v0qRJ3H777aSnpwMwffp0hg0bRlpaGsuWLXNvs27dOoYOHYrVaiUrKwun00lqaioFBUYX0k6nkz59+nDu3Lmr/TMIIVqT1lBWABVF4LD7Ohr/4UpIasK68eQ7e3n4X19iTYzmvQevk+StoagEn90DV8tkUjwzfRBF5TU89+EBn8bSlkgNnBDCo379cT7fnCqCgG1Gkwx7FWgnBGy/6n0O7B7JkzentWibgwcPsn79esxmMyUlJWzevBmLxcL69et5/PHH+fe//33JNvv372fjxo2UlpbSr18/Fi1adEk3+Hv27CEnJ4fu3bszduxYtmzZQmZmJvfeey+bN28mJSWFOXPmXDaulStXMmfOHKZNm8bjjz+OzWYjICCAxYsXM2HCBLKzs3E4HJSVlZGTk8MzzzzDli1biI+Pp6ioqMn3vXPnTvbu3evu/fG1114jNjaWyspKhg8fzqxZs3A6nSxcuNAdb1FRESaTiblz57JixQqWLFnC+vXrsVqtxMfHt+i8CyFaTmtNtd1Jtc1JdXUV9nNHofAgpsJDBBQdIvDCEUJKjmCxldVtExAKwdGo4EgIjjIeQbXPIxu8jr50XkBIxxibs/gkjtB45rz+BbuPn+fucSn87Pv9sZilTuMSUUlQlm+0nPFhx2ODEqK4c3Qyf9uWy+xhiViTon0WS1shCZwQwrNMrmZ72gnK7Jp6/yLhlltucTchLC4uZt68eRw6dAilFDabrdFtJk+eTFBQEEFBQXTu3Jn8/HwSEy++8X3EiBHueUOGDCE3N5fw8HB69erlTprmzJlzUW1XrZqaGtauXcvvf/97IiIiGDlyJB9++CGTJ09mw4YNLF++HACz2UxUVBTLly9n9uzZ7iQqNja2yfc9YsSIi7ruf/HFF8nOzgbg5MmTHDp0iIKCAsaPH+9er3a/CxYsYNq0aSxZsoTXXnuNu+66q8njCdHeaK0pKKs2kim7gyrX1HjtpMrmMJItu6Pu9ZWW1SZm9fdld2KuKaW7I48kRx7JOo/e6jS91Wl6qnwClMMdz7c6ln3ObhzWY8jVXQGIoIJIewVRVZXEWyqJtVQSqc4TQTmhzgqCHaWYdRO1dKaARhK92uQv+tJksGGCGBRZ933fRjmcmtIzxzhVEcW+shL+OCeDm63dfR1W2xWZAGgoPQ0xyT4N5Sff7cvar7/lidV7WX3/WMytPSafww62cqipgJpyqCkDW73njc6v97DVf11hrPfoQTB7psmnJHBCCI968uY0yN9rFAThneHM10aBH93Dq3GEhYW5n//yl79k0qRJZGdnk5ube9n7zoKCgtzPzWYzdvulF0CNrdPc3rLWrVtHcXGxu3ljRUUFoaGhTJ48udH1tdaN9uZosVjcHaBorS/qrKX++960aRPr169n27ZthIaGMnHiRKqqqi6736SkJLp06cKGDRvYsWMHK1asaNb7EsKfaa05XljB1iOFbD1yju1HCzlXduUOkBpjUhAcYCY4wEyQxWQ8zCYSzOcZyCl66jwSHXkk2E/S1XaSKIerebIZHMpMcUgPSsIGcCDyJioielMV3Rt7TB8sIZEEWUxkBJgZYTZRUmWjoLSac2XV5JZWs6us2vW6xphWVGN3Ogmmxkj0VAWRVBBrriQhuIauQTV0DqwizlJFjKmSSFVJuLOckLIygorPYLGVoKpLUTVlV37DYCRx9ZO/kGgj+QuJNno1rH0e7Hpdf3kTF7pOp6a02k5plY3SKrvrYXNPS6rslFxmWe28smo7HwUeoTCgB6sXjaVvl4gW/107lPpjwfk4gYsIDuCJKQNZvHIPK3Yc585RPY0eRcsLGk+0bOWXJlmXS7RsFRffq98kBYHhxvi2gWHGIyDM+CxHdnctCwOnQxI4IYSfMlmMmjd7tfFl5rT7fAiB4uJiEhISAHjjjTdaff/9+/fn6NGj5ObmkpyczL/+9a9G11u5ciWvvPKKu4lleXk5KSkpVFRUkJWVxdKlS1myZAkOh4Py8nKysrKYMWMGDz/8MHFxcRQVFREbG0tycjK7d+/mBz/4Ae+8885laxSLi4uJiYkhNDSU/fv3s3270Yx19OjR3H///Rw7dszdhLK2Fu7uu+9m7ty53HHHHdIJimi3vi2uZOvhQrYeKWTbkXOcLjYu5rpEBjE+tRPWpGjCgix1iViAmWDXNMhiujhJCzATrOxYLuTCuQNw7iCcOwQFB6DwsHHBWCsoCjr1hfjvQXwqxPeD+L6YY3oSaw6g6Tr2pjmdmuJKG+dciV1BgwRve1k150qrKbhQTWFZNc5Gfn8KspjoGm6hR5idxFAb3YOMxK9TQCVxlkqiVBWRqpwwXUGArRSqio3HhZNQ+RW66kKTCWCNKYQKcwTlpghKCaOYMM47wyhyhnLOHspZewjF2phff1pCGA6M76ZAs4mIYIvrEUBEsIX4+DD384ggC8k7iuiRMYUgSd6a5uOx4NxslXB2HzfbvyYofgOx6w7i/OQ0pqoLTWyo6pKp2mTLnWgl1CVftfPd67m2CQhtsH24Ma8NNDeWBE4I4VlKuToyqWozY8A99thjzJs3j+eff57vfOc7rb7/kJAQXnrpJW688Ubi4+MZMWLEJetUVFTwwQcf8Ne//tU9LywsjHHjxrFmzRpeeOEF7rnnHl599VXMZjNLly5l9OjR/OIXv2DChAmYzWYyMjJ44403WLhwIdOmTWPEiBFkZWVdVOtW34033shf/vIXBg8eTL9+/Rg1ahQAnTp1YtmyZcycOROn00nnzp356KOPAJg6dSp33XWXNJ9sL07uhA1PQ4/RkDQSEocbzd86mMKyarYfLWLrkXNsPVLIsXPlAMSEBjC6dxyLesczpnccveLDrjyGYeV5OLcfzhw0ErUC1/R8Lui6Zo9EJRkJWo+5EN+37hHe2eMXgiaTIiYskJiwQFKbSFocTs35ihp3sneuYW1eWTV7Sqv56NtqCstr0Drqkn2EBpqJDw8iJMBcV/tVbceCnUgqiFZlRFFOlConknL361hTBXG6klhTOdGqnG58S19dRpizlEBVDVeoyHAGRqBColFN1fRZgmBrJcT1vPoT2pFEGj90UnzSO8fTGkpOwZm9Rsud/L2Qn2P88KGdKOAGSyhf6gS2BV/H2O9MMGq8Gku0AsOMXrDb6X2dqi0OjJeZmal37drl6zCEEFdp3759DBgwoG7G+eNQXWo0myw6AnGpEBTuuwC9oKysjPDwcLTW3H///aSmpvLwww/7OqwW27VrFw8//DCffvrpNe/rks8FoJTarbVueuwGAbRC+Xj4Y1j/lHFhpJ2gTNAlrS6h6zHa6HmunSmpsrHzaJG7WeT+M6UAhAdZGJkSy+jecYzpHU//rhGYGt5bo7VxAVs/QTt3yJiWn61bzxwIcX0uqkkjPtWY1w6/7+wOJ0UVNZckeAWlxqPa7qir+QoOILJBzVjdPOP1Fcdds1dD5QWoumAkze7nrte1zy9Zfr5u7NH65qyCft9v7VPSPv13CqRNhym/b939umrV3ElabdJWv1Ytuid0TTe+o7qkQZdBEJPC8+sP8eKGw/zz7pGM6dO+O9a6XBkpNXBCCM+zBEFlEdgrjdc+roHzhpdffpm//e1v1NTUkJGRwb333uvrkFrs2WefZenSpXLvW3vSJ8t4VJdC3mdwYgec2AZ7VsBOV0c7UUnQY5TxSBoFnQe0+c4pGqqscbDreG3CVsjXeRdwaqMp4PDkWH76ve6M6R1HekLUxb0P2msgf79xr+6Zr1zTr6G6pG6dkBgjQev7vbqatE59jYtNPztP18JiNtE5IpjOEcFeOFgQRHQxHi2htZEo1E/wHDWQfJ0HgmynrnUsuCZq1QCj+WKXgZA2A7oOMhK1zgMv2zrgx5P6sPqL0zzxzl7ef+g6giwd5/+ultTACSFa3SU1LZUX4PwxCIww7v/oZm23zRrE5UkN3LXzWPnosEP+13UJ3YntUHbGWBYUBUnD6xK6hGFGM6U2pMbu5IuTF9xNIvecOI/NobGYFBk9ohntahKZ0SO67mKvqsS4mPy2NlH7Es7uB6erqXdAqHEh2W2w8et/p/5GshYaJ99fouNYOcdoRfPjrU2va6uEs9/Uq1HLaaJWbZAxjUkBU8uGcdh44Cx3vf4ZCdEhhASaMSlQKPe/plIKhfGvqlzLTK4Z7vkN18OYWX9f6jL7NamLt6V2/Xr7enFOBoGWaxueQmrghBC+Y3H9QltTZvTIJBc/QrQtZgt0zzAeo+4zfjW/cNxI5GofG35rrGuyGD/C9Bhdl9SFd/JquA6nJud0MVuPFLLl8Dl25Z6n0uZAKUhPiGLBuBTG9I4ns2cMYYFmKD0DZ3bB1q/qErbzx+p2GNYJug6G0VlGwtZ1MMT26lA1akI0KioRcrdcPE9ro1YuP8f44ac2YSs60qBWLa3ZtWotNalfZ56YPIDPT5xHayMkjXZNjde4Xju1ds8zptr9Ntzb1FvP6QSN86L162+P1jgbbNvofvFcJZkkcEIIz3MPAKo7RPNJIQCUUjcCLwBm4BWt9bMNlk8E3gFqM4m3tda/8WRMZ0urOHimjPBgC+FBxj1J4UEWQgPNF3fWoZTRbXhMMlhvM+ZVFLmaXboSup0vw7Y/Gctie1/c7DI+tVV/qNFaczC/zF3Dtv1oIaVVxrAefbuEc+vwJEb3jmNUz2iiKk8YzR+PfwXbXclaxbm6ncX2MpK0jLlGotZtMIR3kR+WhGhMZAJUFxv/7+cO1TWDrCquWycm2UjQBs0ykraugyA6ucW1ai1193W9PLr/tkwSOCGE5ykTmIPAUS0JnOgQlFJm4M/ADUAe8JlS6l2t9TcNVv1Uaz3FW3HtOFrEgyv3XDJfKQgPtLgTu4YJXnhQAOHBFiKCUgmPHED4sIVEjHLQpewAcUWfE3F2F0EH3sf0het+ydA4V6coo4yaum5W4z6mZmo4Ftu2I4UUlhudUfSMC2XK4G6MTQ5nbHg+MSX74cxq2PY1rM4xxnQCY2DqzgOg341GotbV1RSyA/a6KcRVi3UlSWsfNXp37DywLlHrkm78j8n/lNdJAieE8A6LJHCiQxkBHNZaHwVQSq0CpgENEzivGtsnnjfvHU1Ztc09sHGZa1paZae82jXP9frb4ir38rLqSweyN6S5HnfSS33LmIBDjKg8RMbBL0g6sBYAmwrgRMgATkcM5mxMBhfihhEYEUtEUF3CGBpo5lB+WaNjsX2/VyA3xF3AajlBdPF+oxnk1wfrmmsFRRn31Qyd52oCmW50NGKR7xshrkm/m+DOdyE6ySu1aqJ5JIETQniHJdjoyc0LF1QTJ07k5z//Od/73vfc8/7whz9w8OBBXnrppctu89xzz5GZmclNN93EP//5T6Kjoy9a56mnniI8PJxHH330ssdevXo1ffv2ZeDAgQD86le/Yvz48Vx//fXX/saAhx56iLfeeouTJ09ikoK0LUsA6g+elAeMbGS90UqpL4HTwKNa6xxPBhUbFsiIlKsbHtrp1JTX1CV9pfWSv7rX/SirHsu2ajsfVtkxVRSQWPoVvav20r/yG0ZVrCQg/x8AHHAmstvZl0+c/dil+3JSdwZgYMgF5nctZFTiKXo7jhJatA91KA8OuQKJTDAStAFTjWm3wUbHCNIEUojWZ7ZArwm+jkI0IAmcEMI7ajsy8UIN3Jw5c1i1atVFCdyqVav43e9+16zt165de9XHXr16NVOmTHEncL/5Tevd0uR0OsnOziYpKYnNmzczceLEVtt3fQ6HA7NZOm+4Ro1lEw3vaP8c6Km1LlNK3QSsBlIv2ZFS9wD3APTo0aOVw2w+k0m5xuwKgEvHb76Cuv9DairQp3Zjz91G8ont9Dn1GbfXbACgOrgTFmcN5ppi+Baj6XVcKvQcbSRqXV01a2Hte9wnIYRoivx8K4TwjuBICI03eqbysNmzZ/Pee+9RXV0NQG5uLqdPn2bcuHEsWrSIzMxM0tLSePLJJxvdPjk5mXPnjE4PnnnmGfr168f111/PgQMH3Ou8/PLLDB8+HKvVyqxZs6ioqGDr1q28++67/PSnP2XIkCEcOXKE+fPn89ZbbwHw8ccfk5GRQXp6OgsWLHDHl5yczJNPPsnQoUNJT09n//79jca1ceNGBg0axKJFi1i5cqV7fn5+PjNmzMBqtWK1Wtm61ejuefny5QwePBir1codd9wBcFE8AOHhxgDDmzZtYtKkSdx+++2kp6cDMH36dIYNG0ZaWhrLli1zb7Nu3TqGDh2K1WolKysLp9NJamoqBQUFgJFo9unTx30OO6g8IKne60SMWjY3rXWJ1rrM9XwtEKCUuiQ70Vov01pnaq0zO3Xybm+PrS4wFJVyHQGTHiNo3tuYf5YL922Byf9DUN8szOkzjQGD7/4Yfn4KHtgJs16BsQ9B70mSvAkhBFIDJ4TwtPd/ZvQC15q6psP3n73s4ri4OEaMGMG6deuYNm0aq1at4tZbb0UpxTPPPENsbCwOh4OsrCy++uorBg8e3Oh+du/ezapVq9izZw92u52hQ4cybNgwAGbOnMnChQsBeOKJJ3j11Vd58MEHmTp1KlOmTGH27NkX7auqqor58+fz8ccf07dvX+68806WLl3KkiVLAIiPj+fzzz/npZde4rnnnuOVV165JJ6VK1cyZ84cpk2bxuOPP47NZiMgIIDFixczYcIEsrOzcTgclJWVkZOTwzPPPMOWLVuIj4+nqKioydO6c+dO9u7dS0pKCgCvvfYasbGxVFZWMnz4cGbNmoXT6WThwoVs3ryZlJQUioqKMJlMzJ07lxUrVrBkyRLWr1+P1WolPr5DX2x/BqQqpVKAU8BtwO31V1BKdQXytdZaKTUC40fVQq9H6ksms9FjXddBMPxuX0cjhBB+QWrghBDtUm0zSjCaT86ZMweAN998k6FDh5KRkUFOTg7ffHP5PiU+/fRTZsyYQWhoKJGRkUydOtW9bO/evVx33XWkp6ezYsUKcnKufOvSgQMHSElJoW/fvgDMmzePzZs3u5fPnDkTgGHDhpGbm3vJ9jU1Naxdu5bp06cTGRnJyJEj+fDDDwHYsGEDixYtAsBsNhMVFcWGDRuYPXu2O4mKjW36vqcRI0a4kzeAF198EavVyqhRozh58iSHDh1i+/btjB8/3r1e7X4XLFjA8uXLASPxu+uuu5o8XnumtbYDDwAfAPuAN7XWOUqp+5RS97lWmw3sdd0D9yJwm64dSEgIIYS4DKmBE0J41hVqyjxp+vTpPPLII3z++edUVlYydOhQjh07xnPPPcdnn31GTEwM8+fPp6qq6or7UZfpGGH+/PmsXr0aq9XKG2+8waZNm664n6auy4OCjC7WzWYzdvulvf2tW7eO4uJid/PGiooKQkNDmTx58mWP11jsFosFp9PpXqempsa9LCysrnnrpk2bWL9+Pdu2bSM0NJSJEydSVVV12f0mJSXRpUsXNmzYwI4dO1ixYsUV329H4GoWubbBvL/Ue/4n4E/ejksIIYR/kxo4IUS7FB4ezsSJE1mwYIG79q2kpISwsDCioqLIz8/n/fffv+I+xo8fT3Z2NpWVlZSWlrJmzRr3stLSUrp164bNZrsoWYmIiKC0tPSSffXv35/c3FwOHz4MwN///ncmTGh+z14rV67klVdeITc3l9zcXI4dO8aHH35IRUUFWVlZLF26FDA6ICkpKSErK4s333yTwkKjRV5tE8rk5GR2794NwDvvvIPNZmv0eMXFxcTExBAaGsr+/fvZvn07AKNHj+aTTz7h2LFjF+0X4O6772bu3Ln84Ac/kE5QhBBCCA+RBE4I0W7NmTOHL7/8kttuuw0Aq9VKRkYGaWlpLFiwgLFjx15x+6FDh3LrrbcyZMgQZs2axXXXXede9vTTTzNy5EhuuOEG+vfv755/22238bvf/Y6MjAyOHDninh8cHMzrr7/OLbfcQnp6OiaTifvuu4/mqKio4IMPPrioti0sLIxx48axZs0aXnjhBTZu3Eh6ejrDhg0jJyeHtLQ0fvGLXzBhwgSsViuPPPIIAAsXLuSTTz5hxIgR7Nix46Jat/puvPFG7HY7gwcP5pe//CWjRo0CoFOnTixbtoyZM2ditVq59dZb3dtMnTqVsrKyDt98UgghhPAk1Rab22dmZupdu3b5OgwhxFXat28fAwYM8HUYwst27drFww8/zKefftro8sY+F0qp3VrrTG/E1x5I+SiEEB3H5cpIuQdOCCHENXv22WdZunSp3PsmhBBCeJg0oRRCCHHNfvazn3H8+HHGjRvn61CEEEKIdk0SOCGEEEIIIYTwE5LACSE8oi3eXyt8Rz4PQgghROuQBE4I0eqCg4MpLCyUi3YBGMlbYWEhwcHBvg5FCCGE8HvSiYkQotUlJiaSl5dHQUGBr0MRbURwcDCJiYm+DkMIIYTwe5LACSFaXUBAACkpKb4OQwghhBCi3ZEmlEIIIYQQQgjhJySBE0IIIYQQQgg/IQmcEEIIIYQQQvgJ1RZ7iVNKFQDHr3E38cC5VginI5Fz1nJyzlpGzlfLtfdz1lNr3cnXQfgLKR99Rs5Zy8k5azk5Zy3X3s9Zo2Vkm0zgWoNSapfWOtPXcfgTOWctJ+esZeR8tZycM9Ha5DPVcnLOWk7OWcvJOWu5jnrOpAmlEEIIIYQQQvgJSeCEEEIIIYQQwk+05wRuma8D8ENyzlpOzlnLyPlqOTlnorXJZ6rl5Jy1nJyzlpNz1nId8py123vghBBCCCGEEKK9ac81cEIIIYQQQgjRrrS7BE4pdaNS6oBS6rBS6me+jqetU0olKaU2KqX2KaVylFIP+Tomf6GUMiul9iil3vN1LP5AKRWtlHpLKbXf9Xkb7euY2jql1MOu/8u9SqmVSqlgX8ck/JuUkS0jZeTVkfKx5aSMbJmOXj62qwROKWUG/gx8HxgIzFFKDfRtVG2eHfiJ1noAMAq4X85Zsz0E7PN1EH7kBWCd1ro/YEXO3RUppRKAxUCm1noQYAZu821Uwp9JGXlVpIy8OlI+tpyUkc0k5WM7S+CAEcBhrfVRrXUNsAqY5uOY2jSt9bda689dz0sxvjASfBtV26eUSgQmA6/4OhZ/oJSKBMYDrwJorWu01hd8GpR/sAAhSikLEAqc9nE8wr9JGdlCUka2nJSPLSdl5FXp0OVje0vgEoCT9V7nIV+0zaaUSgYygB0+DsUf/AF4DHD6OA5/0QsoAF53Nat5RSkV5uug2jKt9SngOeAE8C1QrLX+0LdRCT8nZeQ1kDKy2f6AlI8tJWVkC0j52P4SONXIPOlmsxmUUuHAv4ElWusSX8fTlimlpgBntda7fR2LH7EAQ4GlWusMoByQ+2+uQCkVg1E7kgJ0B8KUUnN9G5Xwc1JGXiUpI5tHyserJmVkC0j52P4SuDwgqd7rRDpYlerVUEoFYBRMK7TWb/s6Hj8wFpiqlMrFaIL0HaXUP3wbUpuXB+RprWt/uX4Lo7ASl3c9cExrXaC1tgFvA2N8HJPwb1JGXgUpI1tEyserI2Vky3T48rG9JXCfAalKqRSlVCDGDY3v+jimNk0ppTDaXO/TWj/v63j8gdb651rrRK11MsZnbIPWukP98tNSWuszwEmlVD/XrCzgGx+G5A9OAKOUUqGu/9Ms5KZ2cW2kjGwhKSNbRsrHqyNlZIt1+PLR4usAWpPW2q6UegD4AKNHmte01jk+DqutGwvcAXytlPrCNe9xrfVa34Uk2qkHgRWuC8ejwF0+jqdN01rvUEq9BXyO0RPeHmCZb6MS/kzKyKsiZaTwFikjm0nKR1BaS/N3IYQQQgghhPAH7a0JpRBCCCGEEEK0W5LACSGEEEIIIYSfkAROCCGEEEIIIfyEJHBCCCGEEEII4SckgRNCCCGEEEIIPyEJnBBCCCGEEEL4CUnghBBCCCGEEMJPSAInhBBCCCGEEH7i/wMjmvAVRkyX3gAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 1080x1080 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "acc = history.history['accuracy']\n",
    "val_acc = history.history['val_accuracy']\n",
    "loss = history.history['loss']\n",
    "val_loss = history.history['val_loss']\n",
    "\n",
    "epochs_range = range(10)\n",
    "\n",
    "plt.figure(figsize=(15, 15))\n",
    "plt.subplot(2, 2, 1)\n",
    "plt.plot(epochs_range, acc, label='Training Accuracy')\n",
    "plt.plot(epochs_range, val_acc, label='Validation Accuracy')\n",
    "plt.legend(loc='lower right')\n",
    "plt.title('Training and Validation Accuracy')\n",
    "\n",
    "plt.subplot(2, 2, 2)\n",
    "plt.plot(epochs_range, loss, label='Training Loss')\n",
    "plt.plot(epochs_range, val_loss, label='Validation Loss')\n",
    "plt.legend(loc='upper right')\n",
    "plt.title('Training and Validation Loss')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.8.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}