en_text_norm.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "from __future__ import unicode_literals, print_function, division\n",
    "\n",
    "from io import open\n",
    "import unicodedata\n",
    "import string\n",
    "import re\n",
    "import random\n",
    "\n",
    "import pandas as pd\n",
    "\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "from torch import optim\n",
    "import torch.nn.functional as F\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "plt.switch_backend('agg')\n",
    "import matplotlib.ticker as ticker\n",
    "import numpy as np\n",
    "%matplotlib inline\n",
    "\n",
    "device = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = pd.read_csv('./data/en_train.csv',encoding='utf-8')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "df = df[df['class']=='DATE']"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 154,
   "metadata": {
    "scrolled": true
   },
   "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>sentence_id</th>\n",
       "      <th>token_id</th>\n",
       "      <th>class</th>\n",
       "      <th>before</th>\n",
       "      <th>after</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>1</td>\n",
       "      <td>0</td>\n",
       "      <td>DATE</td>\n",
       "      <td>2006</td>\n",
       "      <td>two thousand six</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51</th>\n",
       "      <td>3</td>\n",
       "      <td>7</td>\n",
       "      <td>DATE</td>\n",
       "      <td>2007</td>\n",
       "      <td>two thousand seven</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>80</th>\n",
       "      <td>5</td>\n",
       "      <td>0</td>\n",
       "      <td>DATE</td>\n",
       "      <td>2008</td>\n",
       "      <td>two thousand eight</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>111</th>\n",
       "      <td>8</td>\n",
       "      <td>1</td>\n",
       "      <td>DATE</td>\n",
       "      <td>4 March 2014</td>\n",
       "      <td>the fourth of march twenty fourteen</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>147</th>\n",
       "      <td>11</td>\n",
       "      <td>1</td>\n",
       "      <td>DATE</td>\n",
       "      <td>April 10, 2013</td>\n",
       "      <td>april tenth twenty thirteen</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>151</th>\n",
       "      <td>12</td>\n",
       "      <td>2</td>\n",
       "      <td>DATE</td>\n",
       "      <td>7 August 2007</td>\n",
       "      <td>the seventh of august two thousand seven</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>158</th>\n",
       "      <td>14</td>\n",
       "      <td>1</td>\n",
       "      <td>DATE</td>\n",
       "      <td>1987</td>\n",
       "      <td>nineteen eighty seven</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>211</th>\n",
       "      <td>18</td>\n",
       "      <td>12</td>\n",
       "      <td>DATE</td>\n",
       "      <td>1996</td>\n",
       "      <td>nineteen ninety six</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>256</th>\n",
       "      <td>22</td>\n",
       "      <td>1</td>\n",
       "      <td>DATE</td>\n",
       "      <td>22 September 2015</td>\n",
       "      <td>the twenty second of september twenty fifteen</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>367</th>\n",
       "      <td>29</td>\n",
       "      <td>2</td>\n",
       "      <td>DATE</td>\n",
       "      <td>November 4, 2014</td>\n",
       "      <td>november fourth twenty fourteen</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>381</th>\n",
       "      <td>30</td>\n",
       "      <td>12</td>\n",
       "      <td>DATE</td>\n",
       "      <td>1916</td>\n",
       "      <td>nineteen sixteen</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>432</th>\n",
       "      <td>34</td>\n",
       "      <td>2</td>\n",
       "      <td>DATE</td>\n",
       "      <td>1982</td>\n",
       "      <td>nineteen eighty two</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>464</th>\n",
       "      <td>36</td>\n",
       "      <td>1</td>\n",
       "      <td>DATE</td>\n",
       "      <td>7 December 2015</td>\n",
       "      <td>the seventh of december twenty fifteen</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>470</th>\n",
       "      <td>37</td>\n",
       "      <td>4</td>\n",
       "      <td>DATE</td>\n",
       "      <td>11/17/09</td>\n",
       "      <td>november seventeenth o nine</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>513</th>\n",
       "      <td>41</td>\n",
       "      <td>0</td>\n",
       "      <td>DATE</td>\n",
       "      <td>August 16, 2005</td>\n",
       "      <td>august sixteenth two thousand five</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>555</th>\n",
       "      <td>44</td>\n",
       "      <td>5</td>\n",
       "      <td>DATE</td>\n",
       "      <td>1573</td>\n",
       "      <td>fifteen seventy three</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>557</th>\n",
       "      <td>44</td>\n",
       "      <td>7</td>\n",
       "      <td>DATE</td>\n",
       "      <td>1595</td>\n",
       "      <td>fifteen ninety five</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>582</th>\n",
       "      <td>45</td>\n",
       "      <td>9</td>\n",
       "      <td>DATE</td>\n",
       "      <td>1929</td>\n",
       "      <td>nineteen twenty nine</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>645</th>\n",
       "      <td>49</td>\n",
       "      <td>10</td>\n",
       "      <td>DATE</td>\n",
       "      <td>1933</td>\n",
       "      <td>nineteen thirty three</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>670</th>\n",
       "      <td>51</td>\n",
       "      <td>1</td>\n",
       "      <td>DATE</td>\n",
       "      <td>3 February 2016</td>\n",
       "      <td>the third of february twenty sixteen</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>677</th>\n",
       "      <td>52</td>\n",
       "      <td>5</td>\n",
       "      <td>DATE</td>\n",
       "      <td>September 5, 1895</td>\n",
       "      <td>september fifth eighteen ninety five</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>753</th>\n",
       "      <td>57</td>\n",
       "      <td>4</td>\n",
       "      <td>DATE</td>\n",
       "      <td>December 29, 2012</td>\n",
       "      <td>december twenty ninth twenty twelve</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>775</th>\n",
       "      <td>58</td>\n",
       "      <td>6</td>\n",
       "      <td>DATE</td>\n",
       "      <td>1946</td>\n",
       "      <td>nineteen forty six</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>777</th>\n",
       "      <td>58</td>\n",
       "      <td>8</td>\n",
       "      <td>DATE</td>\n",
       "      <td>1975</td>\n",
       "      <td>nineteen seventy five</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>822</th>\n",
       "      <td>61</td>\n",
       "      <td>9</td>\n",
       "      <td>DATE</td>\n",
       "      <td>1919</td>\n",
       "      <td>nineteen nineteen</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>829</th>\n",
       "      <td>61</td>\n",
       "      <td>16</td>\n",
       "      <td>DATE</td>\n",
       "      <td>1990</td>\n",
       "      <td>nineteen ninety</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>831</th>\n",
       "      <td>61</td>\n",
       "      <td>18</td>\n",
       "      <td>DATE</td>\n",
       "      <td>2000</td>\n",
       "      <td>two thousand</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>851</th>\n",
       "      <td>62</td>\n",
       "      <td>18</td>\n",
       "      <td>DATE</td>\n",
       "      <td>2004</td>\n",
       "      <td>two thousand four</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>863</th>\n",
       "      <td>63</td>\n",
       "      <td>10</td>\n",
       "      <td>DATE</td>\n",
       "      <td>2008</td>\n",
       "      <td>two thousand eight</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>866</th>\n",
       "      <td>64</td>\n",
       "      <td>1</td>\n",
       "      <td>DATE</td>\n",
       "      <td>25 July 2012</td>\n",
       "      <td>the twenty fifth of july twenty twelve</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "     sentence_id  token_id class             before  \\\n",
       "10             1         0  DATE               2006   \n",
       "51             3         7  DATE               2007   \n",
       "80             5         0  DATE               2008   \n",
       "111            8         1  DATE       4 March 2014   \n",
       "147           11         1  DATE     April 10, 2013   \n",
       "151           12         2  DATE      7 August 2007   \n",
       "158           14         1  DATE               1987   \n",
       "211           18        12  DATE               1996   \n",
       "256           22         1  DATE  22 September 2015   \n",
       "367           29         2  DATE   November 4, 2014   \n",
       "381           30        12  DATE               1916   \n",
       "432           34         2  DATE               1982   \n",
       "464           36         1  DATE    7 December 2015   \n",
       "470           37         4  DATE           11/17/09   \n",
       "513           41         0  DATE    August 16, 2005   \n",
       "555           44         5  DATE               1573   \n",
       "557           44         7  DATE               1595   \n",
       "582           45         9  DATE               1929   \n",
       "645           49        10  DATE               1933   \n",
       "670           51         1  DATE    3 February 2016   \n",
       "677           52         5  DATE  September 5, 1895   \n",
       "753           57         4  DATE  December 29, 2012   \n",
       "775           58         6  DATE               1946   \n",
       "777           58         8  DATE               1975   \n",
       "822           61         9  DATE               1919   \n",
       "829           61        16  DATE               1990   \n",
       "831           61        18  DATE               2000   \n",
       "851           62        18  DATE               2004   \n",
       "863           63        10  DATE               2008   \n",
       "866           64         1  DATE       25 July 2012   \n",
       "\n",
       "                                             after  \n",
       "10                                two thousand six  \n",
       "51                              two thousand seven  \n",
       "80                              two thousand eight  \n",
       "111            the fourth of march twenty fourteen  \n",
       "147                    april tenth twenty thirteen  \n",
       "151       the seventh of august two thousand seven  \n",
       "158                          nineteen eighty seven  \n",
       "211                            nineteen ninety six  \n",
       "256  the twenty second of september twenty fifteen  \n",
       "367                november fourth twenty fourteen  \n",
       "381                               nineteen sixteen  \n",
       "432                            nineteen eighty two  \n",
       "464         the seventh of december twenty fifteen  \n",
       "470                    november seventeenth o nine  \n",
       "513             august sixteenth two thousand five  \n",
       "555                          fifteen seventy three  \n",
       "557                            fifteen ninety five  \n",
       "582                           nineteen twenty nine  \n",
       "645                          nineteen thirty three  \n",
       "670           the third of february twenty sixteen  \n",
       "677           september fifth eighteen ninety five  \n",
       "753            december twenty ninth twenty twelve  \n",
       "775                             nineteen forty six  \n",
       "777                          nineteen seventy five  \n",
       "822                              nineteen nineteen  \n",
       "829                                nineteen ninety  \n",
       "831                                   two thousand  \n",
       "851                              two thousand four  \n",
       "863                             two thousand eight  \n",
       "866         the twenty fifth of july twenty twelve  "
      ]
     },
     "execution_count": 154,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df.head(30)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "MAX_LENGTH = df.after.str.len().max()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "SOS_token = 0\n",
    "EOS_token = 1\n",
    "\n",
    "\n",
    "class Lang:\n",
    "    def __init__(self, name):\n",
    "        self.name = name\n",
    "        self.char2index = {}\n",
    "        self.char2count = {}\n",
    "        self.index2char = {0: \"SOS\", 1: \"EOS\"}\n",
    "        self.n_chars = 2  # Count SOS and EOS\n",
    "\n",
    "    def add_sentence(self, sentence):\n",
    "        for char in sentence:\n",
    "            self.add_char(char)\n",
    "\n",
    "    def add_char(self, char):\n",
    "        if char not in self.char2index:\n",
    "            self.char2index[char] = self.n_chars\n",
    "            self.char2count[char] = 1\n",
    "            self.index2char[self.n_chars] = char\n",
    "            self.n_chars += 1\n",
    "        else:\n",
    "            self.char2count[char] += 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "def readLangs(df,lang1, lang2, reverse=False):\n",
    "    pairs = [[a,b] for b,a in zip(df[lang1],df[lang2])]\n",
    "    # Reverse pairs, make Lang instances\n",
    "    if reverse:\n",
    "        pairs = [list(reversed(p)) for p in pairs]\n",
    "        input_lang = Lang(lang2)\n",
    "        output_lang = Lang(lang1)\n",
    "    else:\n",
    "        input_lang = Lang(lang1)\n",
    "        output_lang = Lang(lang2)\n",
    "\n",
    "    return input_lang, output_lang, pairs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def prepareData(df,lang1, lang2, reverse=False):\n",
    "    MAX_LENGTH = df[lang2].str.len().max()\n",
    "    input_lang, output_lang, pairs = readLangs(df,lang1, lang2, reverse)\n",
    "    print(\"Read %s sentence pairs\" % len(pairs))\n",
    "    print(\"Counting chars...\")\n",
    "    for pair in pairs:\n",
    "        input_lang.add_sentence(pair[0])\n",
    "        output_lang.add_sentence(pair[1])\n",
    "    print(\"Counted chars:\")\n",
    "    print(input_lang.name, input_lang.n_chars)\n",
    "    print(output_lang.name, output_lang.n_chars)\n",
    "    return input_lang, output_lang, pairs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Read 258348 sentence pairs\n",
      "Counting chars...\n",
      "Counted chars:\n",
      "after 62\n",
      "before 26\n",
      "['1937', 'nineteen thirty seven']\n"
     ]
    }
   ],
   "source": [
    "input_lang, output_lang, pairs = prepareData(df,'before', 'after', True)\n",
    "print(random.choice(pairs))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "class EncoderRNN(nn.Module):\n",
    "    def __init__(self, input_size, hidden_size):\n",
    "        super(EncoderRNN, self).__init__()\n",
    "        self.hidden_size = hidden_size\n",
    "\n",
    "        self.embedding = nn.Embedding(input_size, hidden_size) # embedding_size (vocab_size), embedding_dim\n",
    "        self.gru = nn.GRU(hidden_size, hidden_size)\n",
    "\n",
    "    def forward(self, input, hidden):\n",
    "        embedded = self.embedding(input).view(1, 1, -1)\n",
    "        output = embedded\n",
    "        output, hidden = self.gru(output, hidden)\n",
    "        return output, hidden\n",
    "\n",
    "    def initHidden(self):\n",
    "        return torch.zeros(1, 1, self.hidden_size, device=device)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "class AttnDecoderRNN(nn.Module):\n",
    "    def __init__(self, hidden_size, output_size, dropout_p=0.1, max_length=MAX_LENGTH):\n",
    "        super(AttnDecoderRNN, self).__init__()\n",
    "        self.hidden_size = hidden_size\n",
    "        self.output_size = output_size\n",
    "        self.dropout_p = dropout_p\n",
    "        self.max_length = max_length\n",
    "\n",
    "        self.embedding = nn.Embedding(self.output_size, self.hidden_size)\n",
    "        self.attn = nn.Linear(self.hidden_size * 2, self.max_length)\n",
    "        self.attn_combine = nn.Linear(self.hidden_size * 2, self.hidden_size)\n",
    "        self.dropout = nn.Dropout(self.dropout_p)\n",
    "        self.gru = nn.GRU(self.hidden_size, self.hidden_size)\n",
    "        self.out = nn.Linear(self.hidden_size, self.output_size)\n",
    "\n",
    "    def forward(self, input, hidden, encoder_outputs):\n",
    "        embedded = self.embedding(input).view(1, 1, -1)\n",
    "        embedded = self.dropout(embedded)\n",
    "\n",
    "        attn_weights = F.softmax(\n",
    "            self.attn(torch.cat((embedded[0], hidden[0]), 1)), dim=1)\n",
    "        attn_applied = torch.bmm(attn_weights.unsqueeze(0),\n",
    "                                 encoder_outputs.unsqueeze(0))\n",
    "\n",
    "        output = torch.cat((embedded[0], attn_applied[0]), 1)\n",
    "        output = self.attn_combine(output).unsqueeze(0)\n",
    "\n",
    "        output = F.relu(output)\n",
    "        output, hidden = self.gru(output, hidden)\n",
    "\n",
    "        output = F.log_softmax(self.out(output[0]), dim=1)\n",
    "        return output, hidden, attn_weights\n",
    "\n",
    "    def initHidden(self):\n",
    "        return torch.zeros(1, 1, self.hidden_size, device=device)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [],
   "source": [
    "def indexesFromSentence(lang, sentence):\n",
    "    return [lang.char2index[char] for char in sentence]\n",
    "\n",
    "\n",
    "def tensorFromSentence(lang, sentence):\n",
    "    indexes = indexesFromSentence(lang, sentence)\n",
    "    indexes.append(EOS_token)\n",
    "    return torch.tensor(indexes, dtype=torch.long, device=device).view(-1, 1)\n",
    "\n",
    "\n",
    "def tensorsFromPair(pair):\n",
    "    input_tensor = tensorFromSentence(input_lang, pair[0])\n",
    "    target_tensor = tensorFromSentence(output_lang, pair[1])\n",
    "    return (input_tensor, target_tensor)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "teacher_forcing_ratio = 0.5\n",
    "\n",
    "\n",
    "def train(input_tensor, target_tensor, encoder, decoder, encoder_optimizer, decoder_optimizer, criterion, max_length=MAX_LENGTH):\n",
    "    encoder_hidden = encoder.initHidden()\n",
    "\n",
    "    encoder_optimizer.zero_grad()\n",
    "    decoder_optimizer.zero_grad()\n",
    "\n",
    "    input_length = input_tensor.size(0)\n",
    "    target_length = target_tensor.size(0)\n",
    "\n",
    "    encoder_outputs = torch.zeros(max_length, encoder.hidden_size, device=device)\n",
    "\n",
    "    loss = 0\n",
    "\n",
    "    for ei in range(input_length):\n",
    "        encoder_output, encoder_hidden = encoder(\n",
    "            input_tensor[ei], encoder_hidden)\n",
    "        encoder_outputs[ei] = encoder_output[0, 0]\n",
    "\n",
    "    decoder_input = torch.tensor([[SOS_token]], device=device)\n",
    "\n",
    "    decoder_hidden = encoder_hidden\n",
    "\n",
    "    use_teacher_forcing = True if random.random() < teacher_forcing_ratio else False\n",
    "\n",
    "    if use_teacher_forcing:\n",
    "        # Teacher forcing: Feed the target as the next input\n",
    "        for di in range(target_length):\n",
    "            decoder_output, decoder_hidden, decoder_attention = decoder(\n",
    "                decoder_input, decoder_hidden, encoder_outputs)\n",
    "            loss += criterion(decoder_output, target_tensor[di])\n",
    "            decoder_input = target_tensor[di]  # Teacher forcing\n",
    "\n",
    "    else:\n",
    "        # Without teacher forcing: use its own predictions as the next input\n",
    "        for di in range(target_length):\n",
    "            decoder_output, decoder_hidden, decoder_attention = decoder(\n",
    "                decoder_input, decoder_hidden, encoder_outputs)\n",
    "            topv, topi = decoder_output.topk(1)\n",
    "            decoder_input = topi.squeeze().detach()  # detach from history as input\n",
    "\n",
    "            loss += criterion(decoder_output, target_tensor[di])\n",
    "            if decoder_input.item() == EOS_token:\n",
    "                break\n",
    "\n",
    "    loss.backward()\n",
    "\n",
    "    encoder_optimizer.step()\n",
    "    decoder_optimizer.step()\n",
    "\n",
    "    return loss.item() / target_length\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "######################################################################\n",
    "# This is a helper function to print time elapsed and estimated time\n",
    "# remaining given the current time and progress %.\n",
    "#\n",
    "\n",
    "import time\n",
    "import math\n",
    "\n",
    "\n",
    "def asMinutes(s):\n",
    "    m = math.floor(s / 60)\n",
    "    s -= m * 60\n",
    "    return '%dm %ds' % (m, s)\n",
    "\n",
    "\n",
    "def timeSince(since, percent):\n",
    "    now = time.time()\n",
    "    s = now - since\n",
    "    es = s / (percent)\n",
    "    rs = es - s\n",
    "    return '%s (- %s)' % (asMinutes(s), asMinutes(rs))\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "def trainIters(encoder, decoder, n_iters, print_every=1000, plot_every=100, learning_rate=0.01):\n",
    "    start = time.time()\n",
    "    plot_losses = []\n",
    "    print_loss_total = 0  # Reset every print_every\n",
    "    plot_loss_total = 0  # Reset every plot_every\n",
    "\n",
    "    encoder_optimizer = optim.SGD(encoder.parameters(), lr=learning_rate)\n",
    "    decoder_optimizer = optim.SGD(decoder.parameters(), lr=learning_rate)\n",
    "    training_pairs = [tensorsFromPair(random.choice(pairs))\n",
    "                      for i in range(n_iters)]\n",
    "    criterion = nn.NLLLoss()\n",
    "\n",
    "    for iter in range(1, n_iters + 1):\n",
    "        training_pair = training_pairs[iter - 1]\n",
    "        input_tensor = training_pair[0]\n",
    "        target_tensor = training_pair[1]\n",
    "\n",
    "        loss = train(input_tensor, target_tensor, encoder,\n",
    "                     decoder, encoder_optimizer, decoder_optimizer, criterion)\n",
    "        print_loss_total += loss\n",
    "        plot_loss_total += loss\n",
    "\n",
    "        if iter % print_every == 0:\n",
    "            print_loss_avg = print_loss_total / print_every\n",
    "            print_loss_total = 0\n",
    "            print('%s (%d %d%%) %.4f' % (timeSince(start, iter / n_iters),\n",
    "                                         iter, iter / n_iters * 100, print_loss_avg))\n",
    "\n",
    "        if iter % plot_every == 0:\n",
    "            plot_loss_avg = plot_loss_total / plot_every\n",
    "            plot_losses.append(plot_loss_avg)\n",
    "            plot_loss_total = 0\n",
    "\n",
    "    showPlot(plot_losses)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "\n",
    "def showPlot(points):\n",
    "    plt.figure()\n",
    "    fig, ax = plt.subplots()\n",
    "    # this locator puts ticks at regular intervals\n",
    "    loc = ticker.MultipleLocator(base=0.2)\n",
    "    ax.yaxis.set_major_locator(loc)\n",
    "    plt.plot(points)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "def evaluate(encoder, decoder, sentence, max_length=MAX_LENGTH):\n",
    "    with torch.no_grad():\n",
    "        input_tensor = tensorFromSentence(input_lang, sentence)\n",
    "        input_length = input_tensor.size()[0]\n",
    "        encoder_hidden = encoder.initHidden()\n",
    "\n",
    "        encoder_outputs = torch.zeros(max_length, encoder.hidden_size, device=device)\n",
    "\n",
    "        for ei in range(input_length):\n",
    "            encoder_output, encoder_hidden = encoder(input_tensor[ei],\n",
    "                                                     encoder_hidden)\n",
    "            encoder_outputs[ei] += encoder_output[0, 0]\n",
    "\n",
    "        decoder_input = torch.tensor([[SOS_token]], device=device)  # SOS\n",
    "\n",
    "        decoder_hidden = encoder_hidden\n",
    "\n",
    "        decoded_chars = []\n",
    "        decoder_attentions = torch.zeros(max_length, max_length)\n",
    "\n",
    "        for di in range(max_length):\n",
    "            decoder_output, decoder_hidden, decoder_attention = decoder(\n",
    "                decoder_input, decoder_hidden, encoder_outputs)\n",
    "            decoder_attentions[di] = decoder_attention.data\n",
    "            topv, topi = decoder_output.data.topk(1)\n",
    "            if topi.item() == EOS_token:\n",
    "                decoded_chars.append('<EOS>')\n",
    "                break\n",
    "            else:\n",
    "                decoded_chars.append(output_lang.index2char[topi.item()])\n",
    "\n",
    "            decoder_input = topi.squeeze().detach()\n",
    "\n",
    "        return decoded_chars, decoder_attentions[:di + 1]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 241,
   "metadata": {},
   "outputs": [],
   "source": [
    "def evaluateRandomly(encoder, decoder, n=10):\n",
    "    for i in range(n):\n",
    "        pair = random.choice(pairs)\n",
    "        print('>', pair[0])\n",
    "        print('=', pair[1])\n",
    "        output_words, attentions = evaluate(encoder, decoder, pair[0])\n",
    "        output_sentence = ''.join(output_words)\n",
    "        print('<', output_sentence)\n",
    "        print('')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5m 4s (- 45m 37s) (5000 10%) 1.2282\n",
      "10m 27s (- 41m 51s) (10000 20%) 0.5710\n",
      "15m 45s (- 36m 45s) (15000 30%) 0.3557\n",
      "21m 4s (- 31m 36s) (20000 40%) 0.2314\n",
      "26m 40s (- 26m 40s) (25000 50%) 0.2284\n",
      "31m 58s (- 21m 19s) (30000 60%) 0.2077\n",
      "37m 22s (- 16m 1s) (35000 70%) 0.1454\n",
      "42m 46s (- 10m 41s) (40000 80%) 0.1503\n",
      "48m 11s (- 5m 21s) (45000 90%) 0.1153\n",
      "53m 37s (- 0m 0s) (50000 100%) 0.1268\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "hidden_size = 256\n",
    "encoder1 = EncoderRNN(input_lang.n_chars, hidden_size).to(device) # Size of the vocabulary, i.e. maximum integer index + 1.\n",
    "attn_decoder1 = AttnDecoderRNN(hidden_size, output_lang.n_chars, dropout_p=0.1).to(device)\n",
    "\n",
    "trainIters(encoder1, attn_decoder1, 50000, print_every=5000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 240,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "5m 25s (- 48m 52s) (5000 10%) 0.0639\n",
      "11m 8s (- 44m 35s) (10000 20%) 0.0441\n",
      "16m 44s (- 39m 4s) (15000 30%) 0.0466\n",
      "22m 17s (- 33m 25s) (20000 40%) 0.0400\n",
      "27m 45s (- 27m 45s) (25000 50%) 0.0363\n",
      "33m 26s (- 22m 17s) (30000 60%) 0.0423\n",
      "38m 59s (- 16m 42s) (35000 70%) 0.0350\n",
      "44m 37s (- 11m 9s) (40000 80%) 0.0289\n",
      "50m 16s (- 5m 35s) (45000 90%) 0.0304\n",
      "55m 52s (- 0m 0s) (50000 100%) 0.0293\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<Figure size 432x288 with 0 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAD8CAYAAACMwORRAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADl0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uIDIuMi4yLCBodHRwOi8vbWF0cGxvdGxpYi5vcmcvhp/UCwAAIABJREFUeJztfXeYHMWZ/ls9M5uUIwghEFkGAwIEGJsMZ2PAxulszuEccLzgdD7/wIE723c254R9Z5/D2cbncGfA+chJ5CiRERKSkIQklFBerXZ3Zrp+f3RX91fVVdXVs7NhhnqfZ5+d6elQ3V311Vvv99VXjHMODw8PD4/2QjDaBfDw8PDwaD68cffw8PBoQ3jj7uHh4dGG8Mbdw8PDow3hjbuHh4dHG8Ibdw8PD482hDfuHh4eHm0IJ+POGDuPMbaMMbaCMXap5vdPM8aWMMaeZIzdzhg7sPlF9fDw8PBwRa5xZ4yVAHwfwOsBHAngrxhjRyq7PQZgAef8GAC/BfD1ZhfUw8PDw8MdZYd9TgKwgnP+PAAwxn4D4CIAS8QOnPOFZP8HAbw776TTp0/nc+fOLVRYDw8Pj5c7Fi9e/BLnfEbefi7GfTaAteT7OgAnW/a/BMCNuh8YYx8G8GEAOOCAA7Bo0SKHy3t4eHh4CDDG1rjs11SHKmPs3QAWAPiG7nfO+Y855ws45wtmzMjteDw8PDw8GoQLc18PYA75vn+8TQJj7FwAnwdwBud8oDnF8/Dw8PBoBC7M/REAhzHGDmKMdQC4GMCf6Q6MseMA/AjAGznnm5tfTA8PDw+PIsg17pzzGoC/A3AzgGcBXMM5f4Yx9mXG2Bvj3b4BYDyAaxljjzPG/mw4nYeHh4fHCMBFlgHn/AYANyjbLiefz21yuTw8PDw8hgA/Q9XDw8OjDeGNu4eHh0cbomWN+8otvXhg5dbRLoaHh4fHmIST5j4Wcc637gIArL7iglEuiYeHh8fYQ8sydw8PDw8PM7xx9/Dw8GhDeOPu4eHh0Ybwxt3Dw8OjDeGNu4eHh0cbwht3Dw8PjzaEN+4eHh4ebQhv3D08PDzaEN64e3h4eLQhvHH38PDwaEN44+7h4eHRhvDG3YBFq7dhxebdo10MDw8Pj4bQsonDhhtv++EDAHxiMg8Pj9aEZ+4eHh4ebYiWN+6c89EugoeHh8eYQ8sb99Dbdg8PD48M2sC4e+vu4eHhoaLljXvdU3cPDw+PDFreuHvi7uHh4ZFFyxt3L8t4eHh4ZOGNu4eHh0cbovWNezjaJfDw8PAYe2h94+6Zu4eHh0cG3rh7eHh4tCHawLiPdgk8PDw8xh7awLh76+7h4eGhwht3Dw8PjzZEGxj30S6Bh4eHx9hD6xt3b909PDw8Mmh94+5lGQ8PD48M2sC4j3YJPDw8PMYe2sC4e+vu4eHhoaLljbtficnDw8Mji5Y37nWfW8bDw8Mjg5Y37kVkmf5qHYM13xt4eHi0P15Wxn3eF2/C2d+6c/gK4+Hh4TFG0PLGvajkvm773uEpiIeHh8cYQssbd7+GqoeHh0cWLW/cfSikh4eHRxZtYNxHuwQeHh4eYw8tb9x9nLuHh4dHFi1v3EdKc987WMfm3f0jci0PDw+PoaLljftIyTIX//gBnPSvt4/MxTw8PDyGiJY37iMlyzyxbueIXMfDw8OjGWh54+4dqh4eHh5ZtIFx99bdw8PDQ0XLG/f6GDXu9yzfgrmXXo+dfdXRLoqHh8fLEC1v3MdqKOT37lgBAHhmg9fqPTw8Rh4tb9zDYUjyOFY7DA8PDw9XtL5xHwZD7G27h4dHq8Mbdw2aeUYG1sSzeXh4eLihDYz7cJxz6Cf15N/Dw2M00QbGfWRlmaJ6PPPE3cPDYxTQssY9iI3mcDB3buHdrT5p6ql1O/Hf968e7WJ4eHgMM8qjXYBGwRgDOEc4DNbWRs5DzlFy0dHHaCfwhu/dCwB476vnjm5BPDw8hhUty9yFef3k1Y+jb7A2YtctKgN5VcbDw2M00LrGnVjNF3c0d11Uu+be1Et5eHh4DAta2Lin1r2zXMrdv4gj1K65e+vu4eEx9tG6xp18LgX54kcRad62r+t5bB2Eh4eHx3CjdY07secqmz7rm3fiylufk7YVYdw2lm87z58eX4+5l16PvsFaIt8wHwvp4eExCmhd4064u2pvV720B9+9fbm0rZBxt/1myWUjOpSNO/1yfB4eHqOL1jXuhBC72O0iCcbyQiGNx8X/PVv38PAYbbSscQ+IAXVh5YUcoY0adyHFkG3eznt4eIwGWta4U5vZbOPe6AxVcRxjY3YOUwKf1tjDo73RusZdcqjm718kWqbR3DJC+qG7NJO437N8C375wOqmnMvbdg+P9kZrpx+I4cJCizBVG8t36SRqZKdm2tD3/PTh6P8pc4d8rpBzBH7+rIdH26JlmXswnMzdep78MEm6z1hlyGO0WB4eHk1Cyxp3NowO1YajZeKfanWejBTG6ozWsVouDw+P5qB1jTv57GTcC1B3m0PVxfCHPD3DWLWhY7VcHh4ezUHLGneacsApzr2IMRtinHtd0tzHphX1xt3Do73RssYdAKaP7wTgatybM0PVGgopZJmwFTT3MVowDw+PpqBljTsHUIpLP3Y099ZxqLb6ilIeHh52tK5x50Apdqq6GO4iRtbFgGt/i/9HDtX8c40m/CSmsYWBWh391brTvpt29WN9k9cw8Gg/tKxxBziCQBj3/L1HQpbRhkI6X3Vk4Zn72MKZ37gT8754k9O+J3/1drzmijus+1yzaC3ef9XDzSiaR4uiZScxcZ46Vbkkg+itVr1ItEyDKX91mrtn7h4u2NDkTKKf/e2TTT2fR+uhZZk7B5Vl0u0mG96s9AO27JIJcw+Ju3KM2lBv2z082huta9w5lWXymXuzmKqVibcQcx+r5fLw8GgOWte4Q+9QHW7m7mDb5Tj3MWpDx2ixPDw8moTWNe6cZIYklsoUv928lL/5kTT1YUoc1kx45u7h0d5oYePOE4cqZeUmmzUSce6iHHXOk5OMJSMqpWAYO8Xy8PAYBrSucQeIcc+XQYoss9doyl/B+OvkYmPItqNKyuVDIV8e8FFRL1+0rHEHT5fakzX3Zsgyjf2aMHfSkYylxtUKOW88mosxVP08Rhgta9wpc+cOakPzZBnbcWkoZF55hoJGO4xqPd/x3O7Y0TeIx17YPtrFGDHUvXV/2aJ1jTvnhmgZE3MvdHbpOtJ5LCeik5jEXsOhuTdqmClzL5ICuZ3w3qsewZv/8/6Xzf2PJZ+Px8iidY07gCBJHEa2G+pyEbbLLeeza+4R6jzNLTMcbavRBlurF3A8tCme3bALANDnmMel1VHE1+TRXmhd485NDtWhM3du+Gw7P/2tTozocPCmRo17tQUmVw03ejpKAIDe/tool2TocCEsL9f37NHKxh08cajKuWX0+7tU8h19g/j0NY+jdyBt+BlZxkGPr0vMv/mNq9FT1uv5z6nd0VOJjftA6xt3l3xJ3ri/fNGyicMAEOOebjNq7g4N4bu3L8fvH12fnBfIMm+XxjLcoZCNM3caCvnybPTdHe1j3GshR7lk32ckZRnOOVZv7cNB08eN3EU9jGhd5s4BsdKeaW4OZc2N+s+ymruLcU9DDceqQ/XladqBcZ0Rnykqy9RDPqbCWgE5h5EJI9mJ/+7R9Tjrm3fivhUvjdg1PcxoXeMOveZuyjPTaCikepzLacYqc3fxTbQ7uhuQZXb0DeKQz92An967ariK1RBcHOQjGQr55LodAIAVm3tH7JoeZrSscTdNYqKUtN6gA7GR3DJU9qFtbjiYE29wqB1Kk6uKH7+1dwCX/u5J5xWDxiJ6GpBlNu8eAAD85pG1w1KmRkHnLZgwksxdXIqomh6jiJY17hxcO4kpNBh3lzqu28c1FJLq2fUwTEMh8y8rYWvvQK7xbAZzb0Ta+cbNy/CbR9biT4+vb+j6YwE9iSxTdT5G1LOxFkrq5FAdSc09ru3eto8NtK5x59DncyfmtNaoA1EaCKjRMibmnn6uD0H+OOFfbsMl//2IdZ8f3/N8Q7KKi+PZBnFMKys6Ilpmz6D76KMST6hwYcojiapDZzMazN1jbKB1jTvyV2KSDG6RZfbIZ3GYGGoaFwOROhW3SUxbewdw0ffuzSx2fN+Krdby/eDOlVi5ZY91n7wymsq1u7+Kzbuau+TbWEK5FL3I3QUcquLd18bYjKCxFgqZXMnrMmMCrWvcpZS/eqZcazALou58LPnNdAz57BiV8ofH1uOJdTvxswYcdY0x7/zjX3vl3Tjpq7drf2NNHnAvXLYZDz5v78iaDVElegfcZZkkrcQYY+4unc3IhkJG/xmAR1/YjrmXXo+HRvj9eqRoXeMO5E5ikjX3xhqma5w73V7UodqIoW7EzLpcx2Wh5maZuPdf9Qgu/vGDTTqbG8QzKGKohczmIoOMJMZaKKSoGYwB98fhkHcv3zKC1/egaF3jbopzp8a9QQeiLrcM00hA0jGSBOQWCqmbhOWKRka+ctx/Ax1KfM0f3rUSGx06gbEI8f6K3H/SIYyxZGMuHdRIhkKmzJ1ltnmMPFrWuAOmlL9ElqnrjdkLW/uw1+JQ45ovyYp+Dpp7naffbCOGIEfHt6O4dXdJsOaCNVv78PH/fazxEzQZtXqInXvdZBZRD4p19sXZ/kjApbMZjfkMXnIfG2hJ4y4qrC5axhQKSfc5/RsL8aFfLMo9Pz2OMfm7CmMIpvEq2dHAcDfE0PA8XEEb7UBt7MS6f+4PT+HYL91SyMFY5PbFaatjzKHqNIlplDR3j9FHixr36H9Jq7nLUSsCqjG7VzNFWmdcxRYx1DS1b1lzJ9e1GJwkAgfFDU4j7KiZS6hWSmOn6vzxsRcBuIYGRv+LdKTifY41iWGsae6caO7M0/dRx9hpoQUgqqtugWxTREgR0kWbQzYrpEGWIZtrJA9JEeberNWizMcM1cGcNtixZNxFXv8izL0RzX2swUUm8nHuL1+MnRZaAIkso0s/QMwpZXLFDKdGVkn08fxjpFBIq0NVPrbRcE33Y/SfG0FHeexUHTGCc2KyYSOae0PFGna4yEQjO0M1QrNDZj0aw9hpoQWQMvfov8loVRvMX66bycmU7yrodW1yEIVoBGKX4WaTcvx+4cMlKWgsGfegQHqA0WTu2/cMNjWFQX2sMndv28cExk4LLQBRiYJAp7mn+/UNprMQXSq5dg/hJNKEXcq7FXdWBspooEg7LDLjVsBlrVnr8eSaHWNJlolfjpssE/0vcvfNiICs1kMc95Vb8cU/PT30k8VwmcQ0oqGQmtwyY3TQ87LA2GmhBSAqUd4C2TQBV8Nx7vH/xKHqwtwNIZgqVFnJlo3SVsZGjmnk+EHCOosw95VbejH/y7fgRSXNQrMgfC/VArJMIw7VoUCEat709MYhn0tgzIVCJkTIU3cdVr20B2/4j3uxs899dvRQ0JrGXTB3zcQiWpf3DjaouVtysBjXaDVIMdbLKqOBIjZkqMy9kUZPZa4izP3XD76AHX1V3PDUhsLXdEHC3AvIFIUc7E0wkKJjm9hdGfK5BJwmMY1C9KY37cC/XLcEP7prpbTtP+5YjqfW78QtS5rXwdvQ2svsaVP+pl/2Sszd3hDufm4LfvHAGuP58mQZCvl4837JxKgGVm1qZLg91FDIai21FJXy2GnCop9xyrXSkH+jkVKlWLxmG976gwcAABO6mtfkxl4oZARP3IGfxPmiPnLGIcm2xMc2QmVoaeauk2Uo9lLNPach/PXPHk7PT68V/893qKbbpZS/llepph8osghHI2xyqJo7jT7qKOUs3kmga+zNlAuCItEyIxSZRPHUup3J5wmdduZe5LnYnLN5k+6GA1whQkNFGHL8+O6V2F0g9/5YxtBmpDdwvRG5SpORaO5x6U05UyhzLzJzXNamRYW155ZRI3ZcnKRBEu1TnLn/8K7nrSkUdBjqmrJUcx8qc29mnpbEuBeQZWwNbGdfFT8hOfNdFle3QazbCuQz9yKXsk3aSsjIKIdCDsWO3fncZnz1hqX48v8tGVrBxgiYEkAx3GhN4y40d80kJpPmXiwELauZ5+aWUfTsdNap+U0OJRTy1iWbCmfcCzWdVhFQY1IOiht3eslmOCkFkpWSCsR9267+peuewb9c/2wyi3moRaXvdUJX85j7oKUzYzmj2uFA0laaxNxF+y2yJOJYxkjLMi2puYuHkxctQ5l73tR0xtzYtku0jKtDNU0/kD2HCwZrxWhZkTh3znkm6qHReQO6tt7M9LmpcTcXqm+whj88tj6RzGxGT0RZ7YijGlRHdNFokN6BtB6OFHMPGFDHSIdCZuH19xQjzdxb0rgL6Jyc9LlRzb3IEml0z1Ch7sY4d2mGKj3efB2VXRVl00UNpPyc7NcKOVBSGqY847fQpTNoJnNPVkqyvOOvXLcE//twusC17fLdlahZCNmLGvfBeojOsru/AQD2EOYZ5Fi7vPdC60jV0rlHLJE3NEIz4bYlmzBlXAdOOHBKbtnSbY1fj+aqaQeouaSGGy0qy8QvHQwBM+dMKcTcNeePPsu/G3PLkM+mNV0zxyhGvai9K2rcTR2QDrr7pCOFhtZwlVJDNFGWYfmyzJbdg3JZLOXv7oiahag/Q5WTqKywdnufdRnD/BFV+nnQ9v7jClukimza1W8t2wd/sQhv/cH95rKRMjbDIKdtr02suyLDDjda07jH/xmLmJBJbthbJZp7TqOkQ21qd7jyu+nFqGVIdXTzNVUZqKg+atNcbdcD8nU/YcQO//yN+PfblwNoPFePrqEPi+ZueR6uCeAAoKcjYu59MXOnZW3EEdw7UENPR8T2b12yybiMYVRO+7nUUYQJeWREh5O/eru1bLkQ9bnZzLTFbLupbqsy7HCjNY07eTqMmXPL0GiSPH2a1h858Zc8NDRFTshSDE8blaVxqfHtRY27bViuLWOBUEjx82A9xLdvfS66njTzttClM2jmYtMu6QfUX2y3312JDLGQ9eizamTBjj0DNcyc0Ilp4zpy9817L1LupFp+WYYa6VMESX1u0qsduZI3F6YRdWJjfCikBWK4xhgYY9rQRQDYW02HwzZjcuNTGyRGJi/yIe9rDoWUDacLc08iN/L7AS2GpLk7MESV7dLrDZV5N3NVIxFSanseqtG0M/fIuPclmnv6m64erdveh7mXXo9bl2zSnm/PQA3jOstOjti8p0LLbXeo2kN3hxNc+tx4AdTF6VsFAwbS5Zm7A2iCoozmHv8vBUxi7jaW87FfPyp9pxEG5377LvzywTWFhrkhJ/qj5VWqE2qG37jrfROmfVXDUK2HeOfJB2BSd6UxzV0yko1X8UdWb8PiNduS7yUH5u7aSQNpI+xLNHc7c38ynqT0+0fXac+3uz8y7i4ZG1xHVIB9NDo6k5iG55qtlqvG9F7U0OfhRmsa94S5ZzV38bmrHEg9aBFDqA5l//nPz+SuiaqWwWU5t/S3xmSZ4pq7u6wS8ixLrYUc5YChFLBRlWX+8ocPJNP5gXS+gy1xmPrebJ2TqCr9QnMn++o6EDVFhYo9gzVM6CznRspE5bL/7srcG9Hch4rkUq2qpzQJJl9IGgrpZRkjxKOJmLtiaOLPqgFyyRgooMYGcweZRWyPfAA8w8p1SDV3cQ534wQAm3b247u3LXfWVYuEQnLOMyw1DDkCFkUoFXOoZidvNFOWSZl7gQ7ccnlxHp0sozOoeVEdfYN1dHeUHI17/ohKwOpQHY1JTA2SFOP5kufaWjAx95GWyloyzp2mBGBMZc3R/3IpcI4JVqEaS06uKWeM5PjeHSvwlhP2T34vBwxhSGdC5ksFIefor9bt+rzmt6sXRXHbC+ZOwWsOnZ5zV9nntLu/irXb9uLI/SZqr6dKJ5xHFZSpHWoOdI2zmdEyotHYwitVe2MzQMJm6mQZXbmTLcqNrt3Wh96BGuohR6UUJL4BG/KZe/qZdjRLN+7C+M4y9p/SIxVlNBbI5mg8fJFzjruXv4StvQNYtz3KpNliqkxu8MZIdbctadyvum81gOil7+6v4ar7VuPiEw/AEftOIHlnZLmmiAyQZe760MbnX9qDb936HG59dhP+6Q1Haa9ra6xivzuXbcG8L96E//ngybn76uA6U1VNP3DJfy/Cw6u24fmvnp9IG9L1QvV4joBl/RyNoKnRMg5rqKrPz1Z88f4HqtlJTLYORLVBp319IQBg9uRulALmxNxzWa9Bcz/vO/cAAFZfcYFUmNHICjmUa96yZBM+8svF0rYWs+1mzd3LMvn43sIVAOSX/rM4xWbC3AMmNfYi+rSO7SQVVxNVs3ewnrywUhy94zLrVP1JMBUd7EzT7d7UyVmPvbAdANBfyyYgCznPGOA65whiIzVUo2GSZQZrIX7/6LpCDaDksMxe1rhbOoJQlhfoY9B1SnllrQtfRdOjZfI7mhENheTy/0awyTKJqlkIQ46r7lslLebTTAzW9ecd6clYLWncE5DGUi4JT3TK3KXojCE4VOl5qVRBHbtiq2DuTpp7gfA8W4NxjTyR4vfBk2n0uuySnGc7jTCWZTJ+jgYgP8f08w/uXIlPX/MErnvSbWGPb9+yDPet2Jo5p4qsLGM+Z115d3QkZ7uGKaqjFnKUAuYkL+THuRfV3POv2TzkE5pG0OxomT8/8SK+9H9L8J3bljf1vAKDhsg8nxWyAOgrr8RxZuLBqfJIkWgZLTsTv2lmaTKwxHCWS4EURmiXZfTX0MFF3smDNNkrBLoq0TPr0xj3UONQ5bEso/o5XGEKhaSdyEu9AwCAbXvSdAG/fHAN1m7r057zsbU70nM2SXMX71KUKy8UMs/xVw9DlB1lmTzqnufcFRjNUEjpik24fDNN+1X3rcInr34cQLr0YbNhjJaJ//vcMg6gbaUimDvR3OuScXd/oDq2I05VNTB3sTmIZZlUo7cZHHcduBmyjJzzBglz1w1P6yHPsNR6Ei3Dig33NQmTaCdJ35OQWMQ99Q7U8MU/Po13/uRB7ampsbUyd6VB2YpfU4x63iSmvARXtTCSs0qBi+Zu/53WGZ22+/YfPYAbn9owOqGQ4poF6sbql/Y0JUNorR7imzcvyzXYv3xwTfJ5OBy1V923Cn/3P49qf/PMvQCohlWOmbtoeyJqRaBIBbLFMmuZO2NJAy/HIwaXoWkR+2hlmo61he4Wco7Ospm5R7KM/MxCjsRIFbPt2VZEDTG9jDCAakqG7XuyjZYrfgGb9KY+Inuce8zclTJE12iEuUeau9sM1TxZJv2sq9MPr9qGj/360VSWGYUpqq5X3LBzL8785p34+k1Lk23aJ+RghK9/agO+t3CFdK68Uw2HAv6l/1uC3f36/PO6kODhRMtFy9AGLDP3WJaJv5eCAJzTlL9DM+6JLKP5jUGWg+qcTmJyN8ouYZOu5c27HuccHWU5+6G6r04XH4osQyGneMgy95Q9m88xWA+lMtqYe5FoGdWRSk+rrRvJCM6muQdwIO4Ok8vcRqPiWk2cTpALOtJxYcUbdkbO00dWb7fuZ3NEhiHHwZ+7ATMndALIbwv0HQ13iKU6UktkGc/c9aCzTiXNXWF8ZUWWKTJpRmu4ePY8siwTX7cUx7knsoz+Gi9s7ctoc7aX/sfH1ht/ExV68+5+vLjDFnEjX6urYnaohly+V3ENIcsMtYJWDbKMmgTM1lgHaqEiy1iYu/LdReZKjTw1qNlr5M1QTaJlHKz73sEavnbjs8ZIDsmhag2BlQMMRgLiSq7XFPcoErU1gmr8zjfvjnw1lZwcDyMZr5JZrUwjTw7r9UfkKk0ErdC0MZW1DlVy3JCZeyzLEAMitlFjVwoYqrVQO+lJYHd/Fad/Y2H2GpZG8U9/fia3vCf9a5SuNYl1VqBOYuoswNxDcn9FZ6jaygzIxrOsaO42gz1QDSVja2fu6vd8lq/L1mkb1ZkYZj2JltH/Tt/7f929ClcvWotJ3RX8zZmHavaN/neUg5xoGXN5hwtFQyGFcReOfRNsDFu9Vp5xp07t4Q5NVMvic8vkQGbu6cu5dtFaPLdptxIKKTN312U/dQ0/cahq2KzE3ONOJXWo2u+BotGG6NpxyQnWeMLc9Zo7lzR3yk4bjXOXQ1OzHQeQdajqIlYEBmp1WZYpkM/dGlpaT+UFtXzaNBZkBGdCOe4U9WVLPw/Ecw5MrFw8985y4JhbxlymZiNh7o7MtD9eb6G7gzB3zUO0NVu1zQip0QR6+pGWZcRXP4nJgAE64YY8u+df2oPXXnm3FG9eV4bTeS9eQMcARaPSacU07bAwfLbEYaaQuEYzJTYyQzWPuf/X3auU/O3pKKVw+gHN7ZpCIU3GXYeBWij9ru67cksv+uKc7EVCIRNHqqZj0eWvoVlKTSiVzJOYdCUxFU/cYmc5sC+zNyqhkGmnqA2LVCDkwK6cZQttRlidTV5R14YcRWRUmeSdjND1R+YyzcOgQXMXUBm0QLUe5g7ZBGxDbzmnefSfyhTlUnTdNM7d3FGoMBn357f0WstrGglYr8s50dyz3v2rF63Fsxt2kWOj/yWWXdqwEehGBQAx7prOVIUqy9DPnHOc8627kqnsKpu0lZ7OUP3pvavw0Ko0vbDOiUl9LybY4tx1KatN4AlzL+WsxDSK0TLcjb0LUtHV0bjmrt5fruYuyTLDC/V9i3cyUlJZyxl3SZbRNBZZc6fMPQ39y4N+hmr0X+fAk6NlooRlonLrbKDJuNOXvrMvCv3b0TeIs791l7W8Q2bug/rjqVyjyjJFKqiuEZkYdxIKmWjuNuZe1zp9o/JG/+9Z/lL8m3ysnbmL/xxfuW6JtAiH7b5tGm4pCIzGX3V0A2YjLzH3ujnkttks0aUzT6VIbpW9BFwdqrbnqtaPIg5Vkw/kt4vXYfGaNILnM9c+gVf+083W82qvpZy/2Vkz89CCxj01ODbmnkk/EBZg7paHTx184iMjGrSIc0+Yu6aZmnyElHke++VbsGX3gJPhdnYWK6GQAn1VfVzuAJFrhLENGEOQE+e+u7+aO9qQnbXEuCeLXcvMXXe5gZocCqnzhyTXU56RLW9Zwtw1++hi6V2aajlgxv20LNdQB8V7E3W5HnIrO7fV5SIoehqXEUMiyxCTuKLGAAAgAElEQVSHqq5N20ZE6nXypFeXzJyfufYJaSHw3y5eJy1y7gpVlhFF9czdgAFDtIwKmjiMc45qnWtfvK4S2hq+HFedMncqW0SaY6o/qjA1OPWlL1y22clwNMrcxXddKCQgP2uxr8gKaWMff/nDB3JHG1LkTpjdnnWo6ssnRS9J0UDyAarz0WUS04AmoZpuJOEiy9jCICW1zHwKAOl7ELmU6pxr65MYdTXLeefCNtM1VN1cqr2xHOiUlsEA9d47cjT3kUzepd5WUrc9c9cjT1+mzD2ZVRq3iA4Nc9c1VitzNzgZRbMsqcxdJ8sYem5Vz63VuVMv727c5bKLhj9Q1R8vG/f0/vLi3Jdu3K3dLkUvUVlGs93JoVpVZBmrcZe/2x6rKIOOrWlnqMbvfkdfVdshAJqYZ3q8JMuYRyoA9e2kM7J1z2hXfyTrNcIST/23O/Cd255Trpt/HJWUXCbx9Q3Uncpodaiqx+Z0FHnRMjc/s9F6fBGoRUtkK8/c9aCGSPdyqOaehi9Gx+hkGV3FslU2Ka5aNHQmM6ow5NKoQYVZcw8z35tr3PXlMHVmOuPOGlqJKfpPD9GlTqbb60rHrOOCA7UQpkW71XtVmbtLnHu/ptPTRlLF2256ZiPef9Uj2nMGATNabDXnD2CLlomNO5m0Z5s124gdWbd9byZjohNzT67pprnviTvPfCabNvTFa7Zj7qXX47lNEYFQR9l5IxU5/YBsQJZt3J3JJT8UiLLctmQTbn92E0lI17RLWNF6k5ho+gHNEIs6NlUDUdHIMkUXjdBpxQwys5WSi2nOYWpwquHYumcQX/jj07llaiTOnY4uTEyiX9Lco/9pKGRxq0EvY0r5m0TJ1EVDMN9bTen8aJEymrvy3VZ6W4eq09zpue9fuVV7nJW5a7fpyyDusaxEFRnPPYKyDA3/5eSzCXtiWYbWPx1ho9uue/JFAMDdz23Bhf9xL6aP65D2fXTNdlxw9CxMG9+pv6gl/cCWeJZrsyBu64O/WAQA+MBrDoq3e+auBXXyMQZMV16i6tjcubeKT/0mSvFZ0TSwosNWur9o1IzJjY6yRN2LNF1T3f6d25bjrue25JYpj7kvXrMNnHOZIfJ85r6HyBJp5yU092j7E2t34EO/WOSUtMu00LRuu3iGtn6rVufGeHnVqKkx4UVy/gCpc0w3icmlDtk1d6rLiG32slUSWYYnHaEOjc6dyF43f5901EGIg+W4vdX8d2wCYwyDtRAv7pQX9/jj4y/iLcQZCgD3Ln8JH/j5I/EawOZzNuI0tcG0XoN3qBqgau6LvnCu9F2NlvnhXStx+9LNAPSyTNHKr5MB6IxNNUpHq7kbWm6jDdHmh7htySa89QcP4FcPrkHIadbF/MqmD4UUmnv0/ZNXP45bl2zCmjjfujQ6UCJdTGuR6j6L0YhtZKWOWGTNXd63qslwaYLueZTi3DC6kYQLEysHgZaNr93Whxueyi5MYjpjxqEa6h2qAs0yJE7Mnbxvl7C/wVp2GUMdirpA12yVc/9f8t+P4I6lm7G3WrdmhaRrCDQDWc3dTqaajdaTZXLi3HtjJ40YtlLGVtZ40otW/j0DNfRX6+iqlKT0A1Trp9BVXKNxb1CMs8kya7dHFX3F5t4kXW895NiyeyDJymeqa3vI5CYqy9A4d3VdyN2E/dQ5R4C04zPlaJHklPiLeM/WaJmqmY1nVpEqEueuqROMMZSZ3qGqPv7HXtiOG5+WHXMm5v6m79+HrcSo2OZH0HKX45i+OudWjWmoEsDcS6/H2fNm4sq3z8/dl77nNKDAfP1qIr3p60KzIMhPyLk17bJYKKZZUJ99ngzabLQ0c9e9pt1xlIBoTJRpN4O5b++r4pgv3QJAZu7pIiHyNcTZf3bvKlx563PScUMti8CAkj6ArloknhFH1HBEHPnP7luFx17YYS3PXg1zLwXZxUmi34Fv3LwUC+NREj2vzrlnZO51mbnnpR+gsMkyKluyxobrmDtjKAfMmppC4M3/eT9+fPfz0rZyoI8w2mpgi2bNXRj3NAmVrd7Ynt/6HXvxnduey9Xl71i62amTEP1epLnHny37Jx244gtSQW9hKMa/VufKJCb592Ybd7Ws4p16WcaAd550AA6aPg6A3vnSGyfKF8adLoytOrVufGoDPvar4t5xlVUCKTPMOM7iXb583RJ89/YoAsFUQW2Jr0QY59+fnc0UqOaGOe3rC5NOLSBGIORc69gTjUtt5BJz52lHRpOyJZOO6hzfX7gSn4j9G/QYThhdcq8azb2/Wk8WCdc9YxVq2CF9fKoxV8+jO+ulv3sS37x5mXbYHLAo/FAdXf36oTVYsdk+YQuIcstQ/Jdi/JNyJRqW/jzJCJHKMg0a94/+cjG+c9tyrMyZcAa4SQlybhkXWSZm1JpoKd15KRqJVq+FXAmFlM9CjbuunRRl3Cbm7mUZAyb1VHDozPFY9dIebbRMX2zoXJj7x36tXw7LFalDlUnGj2IoDlWKzjjF62H7TMj8ZorHrpTkdRs5zxoZgMzIVC4v4pCj39J7pQtki9vt18R3q041+ijkSUzR54/+ajHuXBY5kKuJ5m5+JmqoohwNRK5lCRWk+M0jawEAx+4/KfNboGHunHN8/g/50UxA1lj86w3P4kOnHyw5p2m5dHf9zIs78fP7VwNIgwPqIbcaUFudkmU3u8EpFi3jtkC8bnSm4zfNii6ph3ZZhq6gpJvwKGRGV2SYu3gmnrmbIdii7j31xYaurDXuQ5udZtLTaQiz2oh19dLUc9tSuHbGU7R1uvwejXGvprkRknKERJbRlUdtRDJzj/6LxGFqZ6ab5Wpb9EKXK14YdkCjucfbqQHPMHcDA9Q9b5eskBSMRT4bUy6bPJQM6QfKGRnPfM4L/v1eXPdk5HxNJjHx7Fq3FDaWKH5yCW11sa+ial69aC1e6h2Mj4sO1NXRhLlLnZtdlhFoZFJrtR5aTTN9BlrjPlTmHj+fkVodqyWNu2gPugq5Z1Aw92gnajDLjrllTFA7B9HQo8RhsUxRymfuJo3TztxL0jUp9gxkDatwJIvSiKFyoJNlFCMsQJlx6l8QxiDaLt6FLid8mjI3Pgfn6BusYaBWN05iEhgwyDL0q01zty2wIcJkTdD1sUHAUA4CKeqmyPBaNeLJdqW+uPrUK0SWKeo/EOCkg86zW26TmKJ9dvRV8Yd45TDOgaUbd+Gof7oZf37iRWl/6ui0lbe5zJ2c1+J0181mL1oOdXcdyRlOtKZxV5ZioxD5uwWDHqyl+wyVuauyTqhpHBnmrjmPqQFrF4KIITI46ljansFapsOgMfiiJGGoZ+623PNJ2eJCixmqqubep0kbrI4IOAeOvPxmvP679xgTh6nlzxp3wtwVWcZkJNTz5y3wrQt3LDEWSWO1bIfnAnXxGAG1TonQzzwnp+gsQu4WCvng81sx99LrsWJzmhoikdaQb7hcbtVkmJ/bFGn6NysRRCIU0jazGGheBE0t5JKUm/HLDDdzFyTHG3cz0lhtnXE3a+4m9tQodJOYVOlG10jNmruZtnUkxl2TlZBnnaqDCXOXHaoi8ZeuPLYGnjhomRzXL0YCNuauhkI+v2WP7ETTGXdVc+fZfVVZRtLZyWcdc3dJHEbBGENnpaQdzbjANENVJRy20E8KOl/B5ogXz+v/Ytb8AJlBS99L3vVc2KYp0mVch1jxSyYASShkbrRMum0oM27zmTsx7jrmXjBSWS3pSMe5t6ZxT5h79jfhBNQa9yEyd1UTTxu3nPKXQvceTRXU1kgFczeteK86VYVRTJf2ihoaY9m1PEWltbXf65+MWJdIHKaGQmqNu7DLGsOsm1lKH13VkH6APjpXWUZtlOpEM1O5KQIWpaalHUpR5q6Cc54hHKKO5Z1akmUsN5NIh5qqT99LnnxYJLcMRch5sozeHqWOCIdqXrRMs9h8tR5Kz6Ewc+ccO/oG8bUbnsWT63bkXi/L3EdWlmm5aBkgZYt6zV0OhaS5PnS9cRGohlWexCQzWQGtLDMEzd3E7lXdPZVRRDl4rLlHhorubXKoUvzsvlUARG6ZtIKK29Wt5pSeF9J/+hu9bilgCONnLEYoagdOn5Fq3EPOsX7HXsyc0ClLNMp9lUsBQm6eak6jJgQ4og52oEHmrnOo1kOeMfouoygg9R/VQ7tDVT0P1/wW8vz4cidZRstk0hEzdbrTEE5TKgq1nLT8rvq36tuh0WzqPdF9TUkGf3r/Kvzo7ufRX63jmP0nW6/NuT6CyzN3C0R70PWAgkHqhsEqc7fl+9BBnSSTGHekL069rloJI6alP79Vc6/oNXfB4MTkreRcdZMswzIhpIkU4DDsZExOp8xszD2RGOT/AKR8KMKA0xGFibnT59mvSFG79lbxmivuwOV/ekZi66oRrtZCq7HSTWaph9GyhP0NMnfdJKZayDOyTFXp3EyoEILj4lDVhQ3T9BPahUkMSyGaYJJURBlo9NWgJuNo9Dl7Xt1I13XCH72meoz63Oqc49RDp+MITbixKOeOvqr2NxNo0X1uGQeo62xSiJArneFWh8BFjTtFleRa5yAORjW0javHhcbGaNXcSyIUUj52UncFALBrb1bPXLZxdxLdEQpZBuZFBHSN87gDZHaiyjIli3HPhkKmv+kcqvR1bNsziLmXXp9o3GmuknQflbnvihn3HUs3KQYj+nzIjGjym0iRwEljO+rymzLlp6jVwyxzL8DAdHUt5DzDEF3y2AMyc7dmsVR+k41Nek2tYSaPV1rnlXzeuLMfn7n2iTgCKnt9OiqgzN3kmHZdPMfVQKqjBUog1HPUQ6CrUsKcqd3a51EPeUKabESMQvYXZLcNJ1pTlhFT3jUPWDR4XWNSWVI5YGg0VdCu/ip+dPdKAHIWvDzmPlAz52i3ae7jOvWvamJXBS/1DiaLMwjs7q/iTd+/L/nOwRPmrk60MkkBV73/RGzc2Z+kKQCILJPIUNF2e7SM/F1cq6MUTcwS19dF8qSGOL4Pcg41G6ZIw0AdvvT+jtl/MlZu2UPKAJRYVHZVD87ci4a5296XCp0zvxbyTHhuMnkr59zUoWozdOkIK/rODcZGH7FE2XW6nfP0fP/852dw0zMbcfa8mfnMnfiFBurE6GrYre4eKJyNe5W+r1AiELqUFKUA4Fy/RnBk3HlyLhdw6bNn7rlImLvlIellGXfmnqfPf/PmZcmLpuwkbzQwWAuNPbdtqPnh0w/G+149F5ecepC0fULM3HfulY17RlrgUUXTOdZ0mjgAnHXEzMz9iAWyU2Mb/dcxdzWnt+pQFY42YaR1y62pBlzS3BVZRhhopuwnPp9yyDR8/53H4yNnHCyVz8VIV0OOrnJJznFfhLmXspp7GGbTQTgzd7pYh6UcLnHuJqnQlP9H1sD129PfUyNKDa0ky+Rq7rTM0X9XWUYy7qGcW0a9VJ1zMjIVo035GQw6dr5p2emoRP4/3GhJ406TVVFQZq7KI9Hv8jbbAgrC8FAcTyQKmh6UVvy8rJDVusW4W9jAuM4y/vmNR2UY/PjOqJzqDNHte2RjL6QjldUCWW2cQmXT6gxVUcn1M1QhnVdaojDkyar3Sdim5nWojlqbLCNAJ1lFx6QjgwuOmYWJXRXpXC6LndRDjs5KIF2zSKI3bU6fkGf8QEl8fwGHqmrA5XC/eFv8/canN+KMbyyU6mEUCqlj7qQzlpaXjP6v3dYnrTege4yc+ARoQIJpMXOtT1YrGTUmy9CjsplDeZI7SfykTloTkwNdZRmt5u4dqmakxl01nOl3XWNSZRkby6Yrsgu87YQ5+MpFR2W202EtvW7AshVo0OLMsxkLet6ZEzoxd1oPACQGUo1z394nC06R4y16duqoRJVPpOsqzyxQmI0wjC6a+6BkGEP0xB2oCC/UvQ9xX6JotvQDAlQ2ouUQ5xfGT+xjWkNWvZeuisLcCxh33b3VNcxdSCF5HYdYCDrUGHdRJ+j5BFZs7sWarX3o7a8RzV3/7ul5qRwlntu5374r8YkwzbWi8ulHD5LmrnlX0jk0IwVbqg4KlbnbInMEc6fRYCohSWUzt+vrnMVelrFA2CbbQ3JxqNqSCOniXBkDJsYyiKSl8fTF0et2V0oZWaFab0xzpyGWD3/+XPzp704FAHTGDVmVKNSFB0S0DGPZ5QbVyUYUakegph8QDfqB57PLy4VKp0GNcRimo6MBiywjVuvhmg7IZJRp46TlEM9QXEfc7mDdrrcLdJYj5p6MRIoyd40xUU8h6oBtdSUgfe91njXudISaLPMb37PojHsHaoosY2futFPjyftUJDNNmU16Pm0XD6/almSmzAvJFLAtUEPRr2juEpPOOFQ5Sglzz77jOk81d9N8ExX0evUkvNcbdyNEI1V73ik9leSz1rirLNQij+s0d4Z0MpE6wUN8pUysu6OcGfIPNKi5q/KIKF9X2Y25c0SNJGAsOyvSEi2jSlmBkGXEULtmLnOGuddNzF1OT0whZJmQx+GnEnM3GHcwhQ1G/8UzDBTmrlsIW4euSgmcu+WaV6Grj7V61jCnoZA5xr0s0g9k3xt9Z6oBE8wzSlkR72MwwNRY02dkqr/qaldArLlrHq/aoZ7zrbu05aXXu3PZZuwdDOPjHZn7oCyj6aKoku9hlHupxNL6Qxm6pLk7CudnfvNO6friPCOBljTuopGqFeHOz5yV7uMgy6hM8aSDpuLUQ6fH++qZewdpVAK0cVCtv6cjy9wHLZq7LRRSvR9xL90d8QQR1birmjtPJzGpIxjblHeV5QcKs9E1aAGxT8L0pDDCyFgCEYt76Pmt2gWK6X2pSbJMDVwlyaksE30XMd+6Tkdgv0ldACIJTEAY1P5qY8Zd3fv//e5JPL5WnumYTt6yGw8xqS0Ms5OYOkg9r4Ucv390XRKpIjqPPQO13Dh32kG6OJJ1z0PtkAUpEh1zpzqK1Jyac2DF5t1431WP4HePrgOQv26wgFp/ZCdx9H/5pt04/esL8eLOfpTEAvDxb1LIbkjX9nV797ROFz12qGhN455Ey8jbx3elzka95p41VCrEJrXSAZFREOfoU5agS+Pc0/21skwt1DIZIIe5K/cjtEGhr6rsU8/cI4eRKjnZZBl1UXERSpnIMg55TYROOiDFNoeolIIkHPIdP35Qew7qEKuF+TlQgGwK2zSOXtbcxR6qvHPmETPwqb84HIBcjxIJTCS8KuAYKwdBpux09rRA3ZW5V8wOVdohP7thFz59zRO4dvG6ZH8gmtGcaMAm5k4svmzc9WXSr1Ilk7Arb30OYciTdqEGLphmqPYqM7AbMe5q/RH3/NvF6/BCvHpZtBRl+pu0ZjInce71VC50zXdTS97tyITLtHScu1oRqB3SDfFzE4fx9Nw65i5iswF5ijqVZShz7+4oZWaODlonMbnLMowxnHH4DMyfMwUBW5Vh7pnFfnnU0JhVlsleN8PcA9lhWa2HOGnuVDy8epvmfqL/oi7LeVmiDqpDmRikYq8y3d/FoDLFka06VBPNPT61ytzpXADGGH79wZPRO1DDrjjcdKAB5u6akbTqqMsK8qEzzLq6q4Iyd24w7mbNXdMRcH3He8fSzdKx/37HCpxyyHR85tonAQDjOsrSrE99Pves49nVodo/KGvuOkc7ff/Ro0vj3Cl5WbutL5EhxfaDLrsBr3/lvvjBu0/ILYtn7g5IJnBkQsBIKKSGlbs0MFGHdA7Vaj1MtqvGXRSFVsKejlLGcFjj3C0sWLeC0s/ffxIuOGYWKqUg41DVRsvwKCtkVpZJ91Gh19zTOPdqnaNLYV9dhFXS8w4qzL0UjyJsDs29yvRxl9hyBlmWSZKoJcYdUrnUZxewdHJWEACvOXQ6XnfUvlnmXqCR2pz3FIlRIee+Y+kmbIwXMxcQklYY8ky9ccmhFDlUxTX1HTs9rypvmMqtg26E0l+tY3JPBa+YNTH3PBEpkbe5GnfaDmoKOVDDeQGkskziUE2v8/f/+xhWbY0mwdFQSHUxdBNc/SnNQksa94vm74eOUoA3Hz/buI+OpKuTmNThEQe3MvdqnSfGnc4IpU4tOmLorpQyTr9qPTSyT9twTddZCXSUggxzV735YkgaMJaNliEMToXOTxEozL2HhN5d85FT8N/vP0k6bxotI7PwUinKkW5zaKozDF2GwCHXT7oRHa+aeC6ruafMnUp3XUPQ3F2R5tSJn13I8YGfL8Lbfni/tJ8w4FrmriEmKvoG65LmnpeaWnaoZs9XZLZupRQx47cdv38mqEGvuWelJ5domYXLNuM/71yZfFcna+kiYoRDNY0Gk68rCIouFDKvbtZ8tEw+Dpw2Ds/96+txyIzxxn10erqqH4uGNHtyd7JNMCwTc+/UMHfO08kR1Ah3axyqA5Y4d7F93r4TcPrhM+T7sbypSo6BBNIhacBkhxtAGbbm3JpQSBHn/ofH1mHn3qqkm5aCbL591YkmrimY+1bLqvNSPpK6PeGXwEC1Dt3wO1meMd6edDpVVZYh+jzZ3pX4N4ozd0CezZkHYVRE+gWxcLgAjZZR7Yz6fnXoHahJGRa1BptspExZP+HJXUcul4J4AlegnUikIuRZsuKiuT+6Zrv0Xc3rlM4GJrJMHDCgk2UoanWeMfB51aHmmXtxTOzKug60xl0x2NVaiIvm74dvvf1Yclz0nw5t3/fquQCil9JREql3KTMUsz9l3V/IMrQxVOv2LH4AcMPHT8Pk7oq0zcbcKyWmnSFKISQNxlhGlrE6VFXjHrAkzv1TVz8BIDV6QORzEB2kfRJTpKN2lgNs2mU27rTT6q+a5whQqOGmSXriZBJTrLlDz9xFRJD4LCCeRSOhkEURhhwrt/Ri2cbd2t+pNKlG1rgsSpOJlslh7nIWx+z5Bgsw93rIUQ1DVEosY+hMuWVUQ+oSCrnPxK7MdXXpACRZRpnEZJJ/qmGIfk3KaRtGot5QtKRDVcVtnz4DL8aa5Lmv2Ae3PbuJDKvTyqgaqsF6FLFBG3CgYe7pQhlhEqUAAAdNH4dp4zqSIa66EEZPRxmcyz11b381t4dnLBvtY5tNWykF0gxCHWp1Hse5ayYxCfmE1FVxGyZZhnZYO/emumaJsYQypNEyEeRJTFFMcUc5yOjJFLQR9yuM3IRov/S7kMEyDtWEucvPjpFOmvapHeXoiyqdDAdqIU9iv3Wg+ZVs0TIm9A3Wc+PcP/DzR5LPA5Jxz+6rPkMbBmvRZKJyEGQYv66PaJS5q6eqKbJMEsuuyjIh0xIS6Vx1niFUefUhYe6uC+UOEW1h3GdO7MLMuJf+wbuPx56BGpZs2AVAzhAo8rAIVOuhZEQ5Tx2X1Kgl65fWeTJpCACmjevA1HEd6B3oi/VsmbmribEAYMmGXTjckC8aiIwJI8wx3W7X3J8n2Q4FxneWkxWaorwaHEEQZGWZRBsnlVx0cgZZhja29UQyCAIAoRyqKuo8PUYw945SgJd63XJzDtRCaz4gup8uWoZ2+FG5zMxdp7kLcpBZAtARBSInc+PcacSYahBdHarptfRx7nQf+oy0xt0xNBFIUxmUSywje2iTj3GemU/h4lBVOxx1ngQduQiUGEMYpJOY+gydVq0eZtYTyHu/XnMfIiqlAJN7OrSNc6IidYRcyAzRd47IYAMKc49lB5W5d1VKSeSICDNU0w8AslPwyXU7MxX40Jmp70CUt8iSgKbQt/EkyVg1DJMZqkcqEQqi3dBiidswzVCl2HdSOvwtB0HiH0gnR+mNQSlgyWQcF7gy91rIpcavZuxUE8+pmrvIfCk+C4hnkS6FN3zGPa/joLKMaujECMOGpRt3JZ9NsgwFJSi6XQsZ95jxCscqhWmGajUTmOA2ggOA6z8epepQQyF10guVHQF9QjwgipZRjbvrM/Sa+xCha5yTFOMOZOWP6eM7M/vQ9Us7y4E00SkIRJw7T1itgJheLypIwIDnNu3GfSvk0LDbPn1GUjZRnCILiZg6AjqpSwzfGQM+fs5heOOx+6W/aZj7frGTWZc4bAth2mfPm4mvvzX1WVCHKtecV2CgVk/i3PMgHqktF76KPim+OTbuSp0QjVtl7rSTlpk7i/cffsdY34Bd5qAdlGqARCdkq0JPr0+Nu0mWodB1lhSmJG46CLITyTIOmntojlqxob8a5W8/ctZElAOWhEJecPQsnHXEjKTeq8y9FKR1Q5cQD4g6CjVCTZzv7Hkz8YULXpE9JsxebzjRxsZd/KfySpYlqkZUGHcaDSM6gFoYgjGWSDNdlVISExtyDgYmdSZClhE9/Pw5kxFySGlS1fKKqfEuTjEBU2WRmHs9YngdpQDlUoBTDpmWOV40rDcfNxu/+fCrAOijZZZvSp18Z8+biUlSTp9AkgwAaIf8gwlzz94nnfIPIHEuv/dnDydyWx4oqxLGWzxSujQg5zzDOqPkaPK+AGHu8f5FFzqePaU7f6cYW/eYncxAei8bd+7Fs8ozEeV0rUOmOHeKPIeqS2ZNAWEUI4eqorlrThPy7OjEZaSwt1pP2mg5HiVwHrX5UhDgha196K/WpY6jFEDKLaMacIFanWci1MSkuFMPnY6psQJAUSVynuus1qGgfY17bC3zGHA5kFeXnD4+eim7yOIX5VLK3IF0kk5nOUhiYlPN3SzLzJ8zxVgONR1tEeZu0h+pca+HoTQJS42+CYmz6aL5+2HWpMgQqca9xJgU+tijTGAqEb3atMJTtA1JKCTFW46fjZ+970Rp2+SetKH85J5Vcbnsz4c2fvF8UjYebX/dd+7GtYvXZVigHC2Tbk80d8e0vCq+9fZjccHRs5z2zZMdxPv79ztW4JYlm6TfknfsWIdM+dwpTMviJb8XcBIKo1guBRnNXVcOrnWo5o8U+qv1pA2KUYIIB2YsWpbxY79aLHUwQpbhPCqLbuH3gEUyJzX8H/nloqRDUO2AAL2HkSDv7WvcNbKMdj/JocoxJe5x6cpGHYrWKkL/OitBMhWo6NAAAB1tSURBVKFH6Nk2WWbWJDk0i0IwxERzL2Tc9TVlHHEg1+pRPo9kyK6cv04aOL0HnWP3ynfMx4/fcwK+e/F8XDRfnkhWKqWGUY2WUVEqZY17ibFMLn0qpwkjkzfFnjrTrntiQ3Ju9f6eXLcjIynkae5ClinK3Cd2VXC+o3HPg81wlzUdkw2mOHcKary1PpQCzL0/kWVkzV1NMkbL10goZH81TNpqKWCoxem2A8aS3FALl23JzFClSxjqZJmejjLqoRwtc/Mz6bq9IpzShpHIL9PGxl38d2DuZBexSs8uIsvMj1dgOvcVMwGkxn1cRzmJxgk5B5RoGSEDiXjYIGC45iOnaMuhpqNtDnNPjWLkZEwXZFZtYz1Mswvarh2wSLp67VH74qL5s7MJzRhL7kVdiUlFiWUdqqKTpJhMZJ/+ZEifY9wJ07x96eao7EHWYE/t6cgMr6m8ptPcq0NwjDWyCs+MCZ0ZDVeXO0mg6MLvLg7VgTxZpojmPpi+Q2rcq3V9J6OTZZwcqrV6EgAhYup5HERBF5RXV1KjYaY6h2p3RwnVOscX/vi0XM74PIxl1ylWMRK6exsbdzf2oi7Hd+C0HnRXSvjUuYcn2w7fZwKW/+vrcd4rZ8XHRCed2F2R0oMGSpy7YKWigpRYVk8WEAnGxDqgaqoEG0TFVydzTSDfa/UQg0SWUStfyHnCZlSphSLPcARBqger6QdUiElMFHS2rwCd0DXgyNzVSAZAz9z7BuvYvFvN2xIQ56tGc08moxRnX41ord2VUhKxJWCb1Gb7TQcX457nUHXN0ggQ5q5MYlJnkAroZBmnyWzVeuIfE7NOhSxD04csIjNZxQLwgGgTOuYenVNdp1gUyY25e+PeMEQDdtHcBTgiVv7sV87DBcfIw2dqTKgxFRN6RKWhl+tIcpGky8h1apbvA1KjLjRC3TJ/JoiG9XdnHyptp7LMYD3EYC1MYtzV51IP07Sq4zvN0x/yEmCViUPVNvMViBiU2iHpsgtSzT2VZdw1dwHVrwFE/hA6iWrBgVPw2fPmZZKMRdcUmnvjUQ+NHNNVCaSl86JyWYx7gTBaIJZlcmwzNd66jjPPwUnrG42WUWfBGmeoNtCRRg7V1Ln8zIu7sGFnPwLGMgvK03Im60XExl01Ieq7EMjT3KV9C8zobRTta9yTO2M4YGpPsn3hZ87EFW85OvkeVbpijUEMyyd2V5LFdMUMVTpcFlq9qMxBwKRJUDr86oMnA0DuflJ54opCZRgA0mLaewfrOQ5VJAs6jNekcxDIHQkRzdKWShiIOr8JXdm5B9OUSAPa2Qit1RT+2al0qFLZNeGNewfrUvqDL1x4JCZ1V5IaoQ2FHIIsQ/MY2UCfc1ellOnsbYEwRZn7V29YiruXZyO4KKhxf/N/3p8ZgeTJMnRilSlaZtCwBKVuhqqALvWIQH81TJz/pYDhqfU7AQhZRm/cxRrBQNQR91frGbKjLlKflJNMlstrJ565DwFUw77lU6fjyX9+LYAoZcBrj9o33S9gySSc0w+bkT2RBoMJc49kGcE21R5bsPRUljEzdwFRWdVFDGwQI4kJqixDKmFftZ7rUBXG3VR5gfyRUKnEMszdJEV0lUuY2C1fSzi1n7j8tZi3bzSTV/csTLKMui6rVDbCyAS29A5IMzHTfaLv9G6jvDys4UlMAHDywdPww3cfn7sfDaXrLAc49dDpkiGzyjJK52rCB089KPl87aK11n1VB6ZqbF2Y+6fjBVDoSJYa85O/ert2gpROcxe46ZOnJ/VERT+RZaj8FwVBGMpJCNrXblyKJ9fvyBCQGZq5MKKc0flZ7gjXa+5DACOyTFellDhKAVlTLgcMsyd346HPnYNPnHOY07kHCHNPZRnhSEn366lEjVEMAQOHGZliFm0RWUawANW400gUkeMmcagqla8e8iS2f1yHjbnnGHcmO6TEtXXoqpQyDUfsO6mnktxPl2YYbJpi312RjfuFRF4TbJfOYVDTNqgTsNT7rZQCKV65EdhCYgXoZLquSgmTezrw4OfOyZRTB/FbXqZQKt/syUk8p1vonSIvWiZgwIdOO1gqV6WUncSkdkilgMWauyFooKusnZzIecS6Rd2hbcHeMUa+MQD4n4dewNpteyV5EwD2mWgy7mmZ88ZOPlpmCLBFy9BeXDSEfSZ2WSMQKAYT4y6iZaLKxCCH/e07qQuMIVnCixo+E0QnVESWEVCNuy5TnykG+un1O7FnoIbuSslaxrwRv6xZIv6vN4Kd5UDqdNV9xXBY19GJTkotj5qWl3ZUolx0SL5+h5xKV8g9ic1Rzl8ppXl1ioZCCrjMLZo2PmXuaYRTWhjGzE47V1mmQgqS5xDNRquok4rsnQONQpFCIZU6um3PIE49dDped9Q+AKI6EoVC6p91OciuTQsISSVNF6I67qePz04yAmJJRan/qiwzk2SbPOuIdLR/VrwYNg2lNaFI/vtG0bbGXec8E5BWbCoYNgbIsoyIc+dxCJ+IZf/mXx6LjnKA/SZ1Y9VLe5yvJQyZjq3mQZVT1HwcQMp41efyN79+FHsGa1a9HXBg7gEDi2vVso278aO7VuKZF/WzSiPmrjhUicEU96Mb7QgjXFEsZZfC3OVc87Fx79frrUBKCjhxjlF0lIszd/W9m2aOfvfi+fjfD70Kt336DIm5p3Mfsnl+tPeQU8/OmReF9L42NqAuUGUXdeZmXtx5KUiXd0wcqqUAHzr9YGm/tdv6Ej8WEL3PSJYxS3s6614LOXbtrSaGmTL3gVqIGz5+Gn770VOkMFtRTvW5jldlGRLxpkufUQqYsQMXz2AkNPe2yAqpgyCALsao0XNP6CojCFg8uzOKlpk1qRsrv3p+ct7ZU7rx8KpofVGXkYHoeKg2/9uP6mPjBb578Xxc/chaqINB3dBPVEZVE9xbrWNr76A1UgbIZ4UBS/e5OkfH7SwH2mRuAoJ161ihYLOVEgNVFLorskOVSnDi+c/bN0qcNqWngu3x+p1C/+XJaEPcT9YwV+shNu/ux3duW269v+S6yiOzPUORFmLaOGrczeepayybOgGup6MkhfTNnzMZP1VmAdvQUQ4yzL63X565mS/LsCRnz16SOOzj5xyGznKAr924FEBk9EoBS67XVQ5Q1SRHAyKiYmpTT63fid0DNRwXz1GhBKG/Wk8yyV54zCz86sEXkt/oCENAlX10I3/dveoQyXp1r7kPBaGBeQkII2ebCbqfYUbpUftFxiGdxJRmhQTkF04bRZEoBjqzbsHcqdZ9L5o/G//zoVdlVvo584iZmX0TOUNznmde3JXRF1Xk3YIuXbEJOuZOaZhg7nsGarj2o6fgM69N5x6IEYg6H0Bl7tS4i+d/4TGzcM9nz8I5r0iZq+jUxNXFs8xo7uUoTe2typR/G9SGroYqCpZOh+rTJ6SyQToKlY+j7JDWVfX5z1DmVrjMobhofppYrouMVgR2DyjG3TEUshSwZF8xEnnz8bNx0PRxyb4BS9tNV6UEbgiFFEZWt8LVPctfAhCtf0v3Vcuqjgp1kS5qRA4dVevIozpTnULYHa+5DwEm5iXQmeiY5kdw86dOx4OXnZPZ/qtLTsYf/ubVcdhUlPf6d4+uw5bd2WRPn33dEcnnAvOSEsdgkR5esM4j9pmA1VdcoM0bnzL37PHrd+zVOlO/987jks8uCz3njZYEOitBxrjT2xWs64Cp43Di3Kl4ddxQgXR4q8a7i4a3YnMvAKCbau7ESM6Z2iPFKyfGXWHu6q1USgEG62HuCIdC7dTp96vedyLOe2XUydA3/YZj9kscd6bnSc+z3+RuXPW+E/H581+R2V9dkchlofhXHZwmlusoB5nokgxzz9Hck7WJFeYOADMndOGOfzgjkTRLAcMH4kieo2ZPQsiBwVq2HYiJXTqXzoYdezGuo4SZE7qSexCgYbKZ9BcaWUYdXfZopD75XrPlERDkyjP3ISDRTA1PWuSQsb2ICV0VKVc5Pfa4A6bExxOWrhk6UoPkavSAxjR3AdtlRIMy+fOzTBq48Jj9NHuakZe+QKCrXNKkH0gr/RuO3Q83ffI0nPfKKHR1KpnMJN6fql2qE0wkWUZ5MOI3xtL4c1F2wfTUmPtKLMsUkfMyKRrI93KJ4R9fNw8fOu0gKQ3znKk9+Ozr5iXl00HU7bnTevDD95yAs+bNxIdOPzhzPdW4U5ZsAnVi64zn9j55gRVX5l4uBdi4qz/5LMAYw5wp0XyUUw+djr84ch+svuICTBvXkUxiUt9Fytyz6B2oSR27LMukZc0uRpMdeU7squCVsyfigqNn4ar3nYgejZNevVdTEIFYx9nHuQ8B4tGZZJe3nbA/gMg7PxS4MFk6JDXhlk+djjv+4Yzke5FQSAGXsGuTQ1VAl6oUAL544ZGZxmUCvc1TCAME5OgV4Ve457Nn4QfvimK/VUYj9HEAmEKMuzBY6kxDtVOkz1F9/sLZ2l0p4XvvOg6XX3gkDpkRGb6TD5qKr1x0FL7ypldKx1TKDDc/sykZGbggo5VT4x4EmNRdwecvODLDIsU7Ust99OxJANJonbedsL/kgFX33zceAUzoKmP1FRdIcpQJ1Bj2DmQzI27aJadsyKt7okz0falt89Lz5+G8o/bFXy6Yk2xjLF2QZJLi/BTvVjePYnd/TZIYJeZORhnqvBOtQ7WzhOv+/jR8/13H46x5M7V+HIpoAR+5TOcfHRGUDfFzGwnm3rYO1UNnjMeHTjsI737VgdrfP3rGIaiUgsTINwr6bu/9f2dp9+ksB9E0ZotxVyWURkIhBVuwLslXNmvuAHD0/pO12y859SBcQia92ECvv7+Sw7yns5ToteIe50ztSSKKbEaCjir2jY27ur/aWOlzVB+/kFZCzjFzQlciBYh7eM8pczNlEAmnXJ2pQNYA0K82iUQnLd5/6dmJg0+wP3VClzCaJ86dAsYYLpo/G/91zyopvj8PtFMUI9L5cybj8bU7AMC6qDnF3551CL6/cKV2hKyW+/gDpuCH7zlB2iZWOqvWeYZliw5IV2V2D9Qkht0pyTJm5s45z4ZCKqNZnR9HKnPAMjLSCQdOxQ1PbUzqqw+FHAKCgOHzFxyJA6fph6Ad5QAfO/MQKW9JI6AsSR3+CiTOUaUiXHziHFz6+nnWY4pASBULDjRPkkkalMGmHLv/pMLXtUFdoKJHw9xpuWzGLggYvnzRUTho+jjjrERVlqHXUDs9IcXkTfah2N7ASE9lgrQcNucmnfEosN/k7sTRLH5XGb8wTgdOG4drPnIK5pD0G64QnWI5nkQEAG8iTlaVuZsgnKY6SdJlKclKKcBALVqLQO0MbKPbJ9bukIyw5FC1aO5inQEK1Q9F6/D7T52buXbAssz8FUp99Zp7C0A01HEdJeOUeFGx1OHyFW89Bh894xDrMUUwe3I3bvnU6bj8DUcm2xZ94VzJKWxa1u7EuVOw4MApkgxSBKby7j9FNiz0CVBWfdJBU/GRMw7Gv731GOt1/vqUuVj4mTONurHa4G0joCIrIwmoUSIusPlabNFaYmhvsv8m5p5MIot/t+VfMaEzSWCXPr+DZ4zH7z72agDAZkfmLjprnSlT5yjoMHVcBYP1EFt7B9FdKWH1FRfg9MOjiUMJczfYSWrc6TOi/oGsVMqNIzzdeeftOxG/vOQk6XedLNNRDnDTJ0/DT9+7AMDIRMu0rSwzUhAVQfWoU4gGIhr5R04/GM+Rpeq0520g/h7IyjvTx3dKMynTUEj5/JecelCS0rgR3PWPZ2nZnLpACZ1e3qno4Ze9PrvupAkm34Bq6NRUuRSuSbxsuO/Ss7G1dwBv/N59xn1sUVK21MXCP2+LlqlxnumwT4hHbhceG71PQUBOJc59Fd9++7Ho6Sjho796FACdTBdAZLYd11lOOvFNux2Ze3x/OgejC3MXI+vnt/Ti1MOi8ovDRP0xcWBqhOkjpNEyKksX+d4p1MmB6uhQfYeTeyoZhzNjDPP2nZiEeHrNvQUgGp4uykRAZe6Xne9uxJoBXaZKUadPnDsF//i6eTjpIHssfR72ndSljyxSZC/ayBsZnQiUSwH2mdiJd598IObNmoj7VryEn9+/OsOEbdcwdRCumD6+E7Mnd+d2Ejpd9u0L9sc1i9ZZc+cLo22qW5O6K9i6ZzCjG8+dPg6rr7hA2vb0l15nzMcDAG85fn+s2JwSjs5ylrmP6ywlz3fTrn5M6Crn6vhifx27dok4EhFSewbrSV0Sx+X5pUw5kr759nRBd7Xj5NDIMopxV42/2rnOmtSNlYacReK/n6HaAhDvWU2ARdGZMPdi5z798BlS7opmQOdQHaph1+EDrzkIlTLDK2ZNwJyp3Th69iTJoQS4RRrZ8NDnzk0+L1oTzQJW5y3MnT4Oh+8zPtPJiOt3VQKccXhjz5hqt8cdMBmPvbBDu586iQgAvvaWY/DOkw+06uFvmr8fXtyxFx88Te/IntQTG3eHTtIlLp++D9EJS8adGMv+aoh9J3ZJxn1iV1lawQxI65uOqeYtuAKkfiT6WRjkZOSn9BwHTuvBmq190shAfPrkuYfhLDK5T+1gQs4zqQPySIjaaU7pqWTyDokOQ/ggvObeApgUGw1bSKWoHC5Lg1H84gMn4f2vcYtQyYPQXZM4d2ZmVM3A5W84Epe9/hVgjOGez56Nr745yqF/Us5s20bxrpMOxH6TuvCm4+SY/PGdZdzyqTNwtWF5w6VfeT1+9J4Fztf5+ftPTJZbpPr7rz94Mh647OzM/v/21qPxX+/Nnr8UMMyfo49MEiiXAnz8nMMkBx6FWKHKxUi6gBopEVr59gVpNFlPR0mS0tSRz+wp2Y5KJIbThVM6MXdyjakG5q7ayQ/G2Sdd2ptaBpEjimKaIcmYQDaElWWMtxrW6pl7C+DtC/bHXcs2W9lfkqmwwDqTzcarDp6GW5ZsSoy6qGzDX8UiTO7pwM2fPB0HTuvB7x9b3/TzHzCtB/drZhM3G2ceMRMHThuH257dLG3v6ShrjfA7Tjxg2Moi9OhGlu7TgcpLU8d1YOlXzkNnOcBXb4jyvozrLEvOyAOm9uBRMlqZPbkLxx8wGSu39OLB57ehUmKYGcfYNzqfRDdxTXQ8prURBJFxWfpPHTxyyDH+qrylg65zVX0MySSu+H8jSzQWhTfuQ0RnuYSfvNeehEkw9wHN6kAjhSvfMR8Ll21OokyOmzMZf3XSHGO0znDgiDgc7D/fdTye3aDPFNkKmNNAlM1wQDB305JxRaFqyWo4ripPHKY47/eb3I0vXxRN+lq7rQ89HSXsUMo2d1oP+gbruI1M2LOB+humxJOYxNwJke8mm9jLvGCLCjXTJuccc6YWe7/Uz3NwPAlOXWBGXfbTx7m3CQ7fZzwAuy4/3BjXWZbSCJRLAb72lmOM8wCGE+cfPQv/8Noj8nccoxARILpFIm779Om48ROnAYAxFr9ZmB/n3tH5ExrFf/zVcXjL8bO1vzEmL2guZvMK7EeY/5ypPZg2vjMz9+POfzwLD33unEwefxNohyPS84oQ1i27o9HAle+Yjy9ckAYpiDLmpSEGspFMnEcO0SIQLH3auA7c8PHo3Z//ylm4/MI0JFlcR/gBfLRMm+CT5x6O4w+YkmSo8xhefO78ebhm0TrjggzNwAOXna3NM3/ozMigP375XwwpP5AL3vOqA3H4PhNwchMd4m84dj+84VhzLiHGGF45eyKeXr8LB88YL/0mSAyFzpFb1JH+h795Nb52w9JkpCA6jJfiGM0ZEzrxwdMOxuI123Hj0xuTrK0Xnzgncy5VwRL38L5Xz8Vja3fgrHkznXwBD33unNSoxzLR3599aBr2HCc/+487lmN7XzWTMXbMaO6MsfMAfBdACcBPOOdXKL93AvgFgBMAbAXwDs756uYWtXVRKQVO+Tw8moMPn34IPnz68MpNeexuqDOfXcAYk7I3NoKfvndB7ojy8H3G47lNaS6dX3zgZNz49AYcNlM25sNFXo47YAquIWsaCFlGncj2/XcejzqPlpJUtfKLTzoANzy9Ee9QDP7h+0zAQ587BzMndBbqdOiIZHxn2ajNz9t3Ih54fmuSCXPauE488vlzraHTzQLLc8YwxkoAngPwFwDWAXgEwF9xzpeQff4GwDGc848yxi4G8GbO+Tts512wYAFftGjRUMvv4eExzBioRYtL6BzGK7f04om1O1AKohw2Oty2ZBPGdZaThUiagcVrtmHevhOti7kPBc9v6cXOvdUk+2ujeKl3AL98YA0+cc5hDU9MVMEYW8w5zw3xcjHupwD4Z8756+LvlwEA5/xrZJ+b430eYIyVAWwEMINbTu6Nu4eHh0dxuBp3F4fqbAB0vbR18TbtPpzzGoCdADLdNGPsw4yxRYyxRVu2bHG4tIeHh4dHIxjRaBnO+Y855ws45wtmzGjuzEsPDw8PjxQuxn09AOqF2D/ept0nlmUmIXKsenh4eHiMAlyM+yMADmOMHcQY6wBwMYA/K/v8GcB7489vA3CHTW/38PDw8Bhe5LqaOec1xtjfAbgZUSjkzzjnzzDGvgxgEef8zwB+CuCXjLEVALYh6gA8PDw8PEYJTnFEnPMbANygbLucfO4H8JfNLZqHh4eHR6Pw6Qc8PDw82hDeuHt4eHi0IXInMQ3bhRnbAmBNg4dPB/BSE4vTCvD3/PKAv+eXB4ZyzwdyznNjyUfNuA8FjLFFLjO02gn+nl8e8Pf88sBI3LOXZTw8PDzaEN64e3h4eLQhWtW4/3i0CzAK8Pf88oC/55cHhv2eW1Jz9/Dw8PCwo1WZu4eHh4eHBS1n3Blj5zHGljHGVjDGLh3t8jQLjLGfMcY2M8aeJtumMsZuZYwtj/9Pibczxti/x8/gScbY8aNX8sbBGJvDGFvIGFvCGHuGMfaJeHvb3jdjrIsx9jBj7In4nr8Ubz+IMfZQfG9Xx3mcwBjrjL+viH+fO5rlbxSMsRJj7DHG2HXx97a+XwBgjK1mjD3FGHucMbYo3jZidbuljHu8KtT3AbwewJEA/ooxdqT9qJbBzwGcp2y7FMDtnPPDANwefwei+z8s/vswgB+MUBmbjRqAf+CcHwngVQD+Nn6f7XzfAwDO5pwfC2A+gPMYY68C8G8AruScHwpgO4BL4v0vAbA93n5lvF8r4hMAniXf2/1+Bc7inM8nYY8jV7c55y3zB+AUADeT75cBuGy0y9XE+5sL4GnyfRmAWfHnWQCWxZ9/hGipw8x+rfwH4E+IlnN8Wdw3gB4AjwI4GdGElnK8PanniBL2nRJ/Lsf7sdEue8H73D82ZGcDuA4Aa+f7Jfe9GsB0ZduI1e2WYu5wWxWqnbAP53xD/HkjALHKdts9h3j4fRyAh9Dm9x1LFI8D2AzgVgArAezg0SpmgHxfTqucjXF8B8BnAYTx92lo7/sV4ABuYYwtZox9ON42YnV7+Jfg9mgKOOecMdaWoU2MsfEAfgfgk5zzXXQV+na8b855HcB8xthkAH8AMG+UizRsYIxdCGAz53wxY+zM0S7PCONUzvl6xthMALcyxpbSH4e7brcac3dZFaqdsIkxNgsA4v+b4+1t8xwYYxVEhv3XnPPfx5vb/r4BgHO+A8BCRLLE5HgVM0C+r1Zf5ew1AN7IGFsN4DeIpJnvon3vNwHnfH38fzOiTvwkjGDdbjXj7rIqVDuBrnD1XkSatNj+17GH/VUAdpKhXsuARRT9pwCe5Zx/m/zUtvfNGJsRM3YwxroR+RieRWTk3xbvpt5zy65yxjm/jHO+P+d8LqL2egfn/F1o0/sVYIyNY4xNEJ8BvBbA0xjJuj3aTocGnBTnA3gOkU75+dEuTxPv638BbABQRaS3XYJIa7wdwHIAtwGYGu/LEEUNrQTwFIAFo13+Bu/5VES65JMAHo//zm/n+wZwDIDH4nt+GsDl8faDATwMYAWAawF0xtu74u8r4t8PHu17GMK9nwngupfD/cb390T894ywVSNZt/0MVQ8PD482RKvJMh4eHh4eDvDG3cPDw6MN4Y27h4eHRxvCG3cPDw+PNoQ37h4eHh5tCG/cPTw8PNoQ3rh7eHh4tCG8cffw8PBoQ/x/0dur4zPVvIoAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "trainIters(encoder1, attn_decoder1, 50000, print_every=5000,learning_rate=0.005)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 242,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "> 1919\n",
      "= nineteen nineteen\n",
      "< nineteen nineteen<EOS>\n",
      "\n",
      "> 1958\n",
      "= nineteen fifty eight\n",
      "< nineteen fifty eight<EOS>\n",
      "\n",
      "> December 19, 2000\n",
      "= december nineteenth two thousand\n",
      "< december nineteenth two thousand<EOS>\n",
      "\n",
      "> October 2009\n",
      "= october two thousand nine\n",
      "< october two thousand nine<EOS>\n",
      "\n",
      "> 1976\n",
      "= nineteen seventy six\n",
      "< nineteen seventy six<EOS>\n",
      "\n",
      "> 2009\n",
      "= two thousand nine\n",
      "< two thousand nine<EOS>\n",
      "\n",
      "> 1995\n",
      "= nineteen ninety five\n",
      "< nineteen ninety five<EOS>\n",
      "\n",
      "> 1985\n",
      "= nineteen eighty five\n",
      "< nineteen eighty five<EOS>\n",
      "\n",
      "> 8 April 2010\n",
      "= the eighth of april twenty ten\n",
      "< the eighth of april twenty ten<EOS>\n",
      "\n",
      "> 1976\n",
      "= nineteen seventy six\n",
      "< nineteen seventy six<EOS>\n",
      "\n"
     ]
    }
   ],
   "source": [
    "evaluateRandomly(encoder1, attn_decoder1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 243,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.image.AxesImage at 0x7f606255dfd0>"
      ]
     },
     "execution_count": 243,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1098x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "output_words, attentions = evaluate(\n",
    "    encoder1, attn_decoder1, \"2013\")\n",
    "plt.matshow(attentions.numpy())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 244,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plot_head_map(mma, target_labels, source_labels):\n",
    "    fig, ax = plt.subplots()\n",
    "    heatmap = ax.pcolor(mma, cmap=plt.cm.Blues)\n",
    "\n",
    "    # put the major ticks at the middle of each cell\n",
    "    ax.set_xticks(np.arange(mma.shape[1])+0.5, minor=False)\n",
    "    ax.set_yticks(np.arange(mma.shape[0])+0.5, minor=False)\n",
    "\n",
    "    # without this I get some extra columns rows\n",
    "    # http://stackoverflow.com/questions/31601351/why-does-this-matplotlib-heatmap-have-an-extra-blank-column\n",
    "    ax.set_xlim(0, len(source_labels)+2)\n",
    "    ax.set_ylim(0, len(target_labels)+1)\n",
    "\n",
    "    # want a more natural, table-like display\n",
    "    ax.invert_yaxis()\n",
    "    ax.xaxis.tick_top()\n",
    "\n",
    "    # source words -> column labels\n",
    "    ax.set_xticklabels(['']+list(source_labels), minor=False)\n",
    "    # target words -> row labels\n",
    "    ax.set_yticklabels(list(target_labels), minor=False)\n",
    "\n",
    "    plt.yticks(rotation=90)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 245,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input = 4 March 2014\n",
      "output = the fourth of march twenty fourteen<EOS>\n",
      "torch.Size([36, 61])\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "def evaluateAndShowAttention(input_sentence):\n",
    "    output_words, attentions = evaluate(\n",
    "        encoder1, attn_decoder1, input_sentence)\n",
    "    print('input =', input_sentence)\n",
    "    print('output =', ''.join(output_words))\n",
    "    print(attentions.shape)\n",
    "    plot_head_map(attentions,output_words,input_sentence)\n",
    "\n",
    "\n",
    "evaluateAndShowAttention(\"4 March 2014\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 246,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input = 1982\n",
      "output = nineteen eighty two<EOS>\n",
      "torch.Size([20, 61])\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "evaluateAndShowAttention(\"1982\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 247,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input = August 16, 2005\n",
      "output = august sixteenth two thousand five<EOS>\n",
      "torch.Size([35, 61])\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "evaluateAndShowAttention(\"August 16, 2005\")"
   ]
  },
  {
   "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}