2. Seq2Seq with attention - tensorflow.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "# Seq2Seq\n",
    "\n",
    "- https://medium.com/@dev.elect.iitd/neural-machine-translation-using-word-level-seq2seq-model-47538cba8cd7\n",
    "- https://github.com/keras-team/keras/blob/master/examples/lstm_seq2seq.py\n",
    "- https://machinelearningmastery.com/define-encoder-decoder-sequence-sequence-model-neural-machine-translation-keras/\n",
    "- https://machinelearningmastery.com/develop-encoder-decoder-model-sequence-sequence-prediction-keras/\n",
    "- https://blog.keras.io/a-ten-minute-introduction-to-sequence-to-sequence-learning-in-keras.html\n",
    "- https://www.tensorflow.org/tutorials/text/nmt_with_attention\n",
    "\n",
    "---\n",
    "\n",
    "\n",
    "\n",
    "- **`encoder`: it processes the input sequence and returns its own internal state.** \n",
    "\n",
    "    - encoder RNN의 아웃풋은 버린다! only recovering the state.\n",
    "    - This state will serve as the \"context\", or \"conditioning\"      \n",
    "    \n",
    "    \n",
    "- **`decoder`: it is trained to predict the next characters of the target sequence, given previous characters of the target sequence.** \n",
    "    - it is trained to **turn the target sequences into the same sequences** \n",
    "    - but offset by one timestep in the future, \n",
    "    - a training process called \"teacher forcing\" in this context. \n",
    "    \n",
    "    - `Importantly, the encoder uses as initial state the state vectors from the encoder, which is how the decoder obtains information about what it is supposed to generate.` \n",
    "    - Effectively, the decoder learns to generate targets[t+1...] given targets[...t], conditioned on the input sequence."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-09-29T09:05:12.012553Z",
     "start_time": "2020-09-29T09:05:10.329068Z"
    },
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "import tensorflow as tf\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import matplotlib.ticker as ticker\n",
    "from sklearn.model_selection import train_test_split\n",
    "\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "import pickle\n",
    "import re\n",
    "import os\n",
    "import io\n",
    "import time\n",
    "import math\n",
    "import random"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import warnings\n",
    "warnings.filterwarnings('ignore')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Load Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-09-29T09:05:12.119116Z",
     "start_time": "2020-09-29T09:05:12.014192Z"
    },
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "before:  (602325, 4)\n"
     ]
    },
    {
     "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>desc</th>\n",
       "      <th>story</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>Large tree with many outstretching branches an...</td>\n",
       "      <td>Our landmark tree in town was about to be dest...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>A green sign is describing a historic tree and...</td>\n",
       "      <td>So we decided to take the day to go out and en...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>A large tree with roots that look like crocodi...</td>\n",
       "      <td>To see the final glimpse of the roots, extendi...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>Big old tree being photographed on a sunny day</td>\n",
       "      <td>And its magnificent trunk, larger than life it...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>Huge brown tree roots rose above the ground.</td>\n",
       "      <td>One last picture of its beauty so we could cap...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>Large tree with many outstretching branches an...</td>\n",
       "      <td>We found this tree when we were walking in a n...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>A green sign is describing a historic tree and...</td>\n",
       "      <td>It turns out it is a popular attraction here.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>A large tree with roots that look like crocodi...</td>\n",
       "      <td>The tree is very unusual, with its roots exposed.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Big old tree being photographed on a sunny day</td>\n",
       "      <td>The trunk was really wide, as much as 12 feet!</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Huge brown tree roots rose above the ground.</td>\n",
       "      <td>You can see how big these roots are - pretty a...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>Large tree with many outstretching branches an...</td>\n",
       "      <td>Pictures of a tree are taken.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>A green sign is describing a historic tree and...</td>\n",
       "      <td>The top of the tree is taken.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>A large tree with roots that look like crocodi...</td>\n",
       "      <td>Another part of the tree mostly the roots.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>Big old tree being photographed on a sunny day</td>\n",
       "      <td>Some more different parts of the tree.</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>Huge brown tree roots rose above the ground.</td>\n",
       "      <td>And some more parts of the tree is taking.</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                 desc  \\\n",
       "0   Large tree with many outstretching branches an...   \n",
       "1   A green sign is describing a historic tree and...   \n",
       "2   A large tree with roots that look like crocodi...   \n",
       "3      Big old tree being photographed on a sunny day   \n",
       "4        Huge brown tree roots rose above the ground.   \n",
       "5   Large tree with many outstretching branches an...   \n",
       "6   A green sign is describing a historic tree and...   \n",
       "7   A large tree with roots that look like crocodi...   \n",
       "8      Big old tree being photographed on a sunny day   \n",
       "9        Huge brown tree roots rose above the ground.   \n",
       "10  Large tree with many outstretching branches an...   \n",
       "11  A green sign is describing a historic tree and...   \n",
       "12  A large tree with roots that look like crocodi...   \n",
       "13     Big old tree being photographed on a sunny day   \n",
       "14       Huge brown tree roots rose above the ground.   \n",
       "\n",
       "                                                story  \n",
       "0   Our landmark tree in town was about to be dest...  \n",
       "1   So we decided to take the day to go out and en...  \n",
       "2   To see the final glimpse of the roots, extendi...  \n",
       "3   And its magnificent trunk, larger than life it...  \n",
       "4   One last picture of its beauty so we could cap...  \n",
       "5   We found this tree when we were walking in a n...  \n",
       "6       It turns out it is a popular attraction here.  \n",
       "7   The tree is very unusual, with its roots exposed.  \n",
       "8      The trunk was really wide, as much as 12 feet!  \n",
       "9   You can see how big these roots are - pretty a...  \n",
       "10                      Pictures of a tree are taken.  \n",
       "11                      The top of the tree is taken.  \n",
       "12         Another part of the tree mostly the roots.  \n",
       "13             Some more different parts of the tree.  \n",
       "14         And some more parts of the tree is taking.  "
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# load\n",
    "with open('dataset/desc_story_data_df.pkl', 'rb') as f:\n",
    "    df_story_and_desc = pickle.load(f)\n",
    "\n",
    "print (\"before: \", df_story_and_desc.shape)\n",
    "# df_story_and_desc = df_story_and_desc[:150000]\n",
    "\n",
    "desc_story_text = df_story_and_desc[['desc', 'story']]\n",
    "desc_story_text.head(15)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "==> 15개 단위로 같은 사진 set에 대한 내용"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-09-29T09:05:12.153931Z",
     "start_time": "2020-09-29T09:05:12.121519Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "pair 0:Large tree with many outstretching branches and leaves. ==> Our landmark tree in town was about to be destroyed and cleared for a new mall. \n",
      "pair 5:Large tree with many outstretching branches and leaves. ==> We found this tree when we were walking in a nearby town. \n",
      "pair 10:Large tree with many outstretching branches and leaves. ==> Pictures of a tree are taken.\n",
      "pair 15:Large tree with many outstretching branches and leaves. ==> They went to the botanic gardens specifically to see the large tree.\n",
      "pair 20:Large tree with many outstretching branches and leaves. ==> We went to see the largest tree in the country. \n"
     ]
    }
   ],
   "source": [
    "check_sent = \"Large tree with many outstretching branches and leaves.\"\n",
    "for i, row in desc_story_text[desc_story_text['desc'] == check_sent].iterrows():\n",
    "    print(\"pair \" + str(i) + \":\" + row['desc'] + \" ==> \" + row['story'])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "==> 하나의 Description Text에 5개의 Story가 대응됨"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## cleaning"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-09-29T09:05:13.290141Z",
     "start_time": "2020-09-29T09:05:13.284824Z"
    }
   },
   "outputs": [],
   "source": [
    "def drop_duplicate(df):\n",
    "    drop_df = df.drop_duplicates(\"desc\")\n",
    "    drop_df = drop_df.drop_duplicates(\"story\")\n",
    "    drop_df = drop_df.reset_index(drop=True)\n",
    "    return drop_df"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-09-29T09:05:13.622737Z",
     "start_time": "2020-09-29T09:05:13.501184Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(39771, 2)"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "desc_story_text = drop_duplicate(desc_story_text)\n",
    "desc_story_text.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Prepare Data"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## preprocess text\n",
    "\n",
    "- Add a start and end token to each sentence.\n",
    "- Clean the sentences by removing special characters."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-09-29T09:05:14.547024Z",
     "start_time": "2020-09-29T09:05:14.541203Z"
    }
   },
   "outputs": [],
   "source": [
    "def preprocess_sentence(w):\n",
    "    w = w.lower()\n",
    "    w = re.sub('[^a-zA-Z]+', ' ', w)\n",
    "    w = re.sub('[^a-zA-Z.,!?]+', ' ', w)\n",
    "    w = w.strip()\n",
    "    \n",
    "    # adding a start and an end token to the sentence\n",
    "    w = '<start> ' + w + ' <end>'\n",
    "    return w"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-09-29T09:05:15.290271Z",
     "start_time": "2020-09-29T09:05:14.765919Z"
    }
   },
   "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>in_desc</th>\n",
       "      <th>out_story</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>&lt;start&gt; large tree with many outstretching bra...</td>\n",
       "      <td>&lt;start&gt; our landmark tree in town was about to...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>&lt;start&gt; a green sign is describing a historic ...</td>\n",
       "      <td>&lt;start&gt; so we decided to take the day to go ou...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>&lt;start&gt; a large tree with roots that look like...</td>\n",
       "      <td>&lt;start&gt; to see the final glimpse of the roots ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>&lt;start&gt; big old tree being photographed on a s...</td>\n",
       "      <td>&lt;start&gt; and its magnificent trunk larger than ...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>&lt;start&gt; huge brown tree roots rose above the g...</td>\n",
       "      <td>&lt;start&gt; one last picture of its beauty so we c...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39766</th>\n",
       "      <td>&lt;start&gt; an advertisement on glass for a brewin...</td>\n",
       "      <td>&lt;start&gt; a group of friends visited a brewery &lt;...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39767</th>\n",
       "      <td>&lt;start&gt; a man holds the woman s hand under the...</td>\n",
       "      <td>&lt;start&gt; they were very excited &lt;end&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39768</th>\n",
       "      <td>&lt;start&gt; a young man and older woman sitting do...</td>\n",
       "      <td>&lt;start&gt; they sampled many different beers &lt;end&gt;</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39769</th>\n",
       "      <td>&lt;start&gt; an elderly couple dances outside of a ...</td>\n",
       "      <td>&lt;start&gt; after becoming a little buzzed they ev...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39770</th>\n",
       "      <td>&lt;start&gt; man and two women standing at a counte...</td>\n",
       "      <td>&lt;start&gt; they had a great time and claimed that...</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>39771 rows × 2 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                 in_desc  \\\n",
       "0      <start> large tree with many outstretching bra...   \n",
       "1      <start> a green sign is describing a historic ...   \n",
       "2      <start> a large tree with roots that look like...   \n",
       "3      <start> big old tree being photographed on a s...   \n",
       "4      <start> huge brown tree roots rose above the g...   \n",
       "...                                                  ...   \n",
       "39766  <start> an advertisement on glass for a brewin...   \n",
       "39767  <start> a man holds the woman s hand under the...   \n",
       "39768  <start> a young man and older woman sitting do...   \n",
       "39769  <start> an elderly couple dances outside of a ...   \n",
       "39770  <start> man and two women standing at a counte...   \n",
       "\n",
       "                                               out_story  \n",
       "0      <start> our landmark tree in town was about to...  \n",
       "1      <start> so we decided to take the day to go ou...  \n",
       "2      <start> to see the final glimpse of the roots ...  \n",
       "3      <start> and its magnificent trunk larger than ...  \n",
       "4      <start> one last picture of its beauty so we c...  \n",
       "...                                                  ...  \n",
       "39766  <start> a group of friends visited a brewery <...  \n",
       "39767               <start> they were very excited <end>  \n",
       "39768    <start> they sampled many different beers <end>  \n",
       "39769  <start> after becoming a little buzzed they ev...  \n",
       "39770  <start> they had a great time and claimed that...  \n",
       "\n",
       "[39771 rows x 2 columns]"
      ]
     },
     "execution_count": 8,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "clean_data = desc_story_text.copy()\n",
    "clean_data['desc'] = desc_story_text['desc'].apply(lambda x: preprocess_sentence(x))\n",
    "clean_data['story'] = desc_story_text['story'].apply(lambda x: preprocess_sentence(x))\n",
    "clean_data.columns = ['in_desc','out_story']\n",
    "clean_data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-09-29T09:05:15.358745Z",
     "start_time": "2020-09-29T09:05:15.353507Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<start> large tree with many outstretching branches and leaves <end>\n",
      "<start> our landmark tree in town was about to be destroyed and cleared for a new mall <end>\n"
     ]
    }
   ],
   "source": [
    "print(clean_data[\"in_desc\"][0])\n",
    "print(clean_data[\"out_story\"][0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Indexing and Padding\n",
    "\n",
    "- Create a word index and reverse word index (dictionaries mapping from word → id and id → word).   \n",
    ": 단어 색인과 역방향 단어 색인을 만듭니다 (단어 → ID 및 ID → 단어에서 매핑 된 사전).\n",
    "\n",
    "- Pad each sentence to a maximum length."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-09-29T09:05:16.387938Z",
     "start_time": "2020-09-29T09:05:16.379201Z"
    }
   },
   "outputs": [],
   "source": [
    "def tokenize(input_desc, output_story, flag):\n",
    "    '''Args flags : pre / post 차이 \n",
    "    '''\n",
    "    tokenizer = tf.keras.preprocessing.text.Tokenizer(filters='')\n",
    "    \n",
    "    \n",
    "    if flag == \"input\":\n",
    "        tokenizer.fit_on_texts(input_desc)\n",
    "        # sentence 최대 길이로 패딩 자동으로 해줌\n",
    "        input_seq = tokenizer.texts_to_sequences(input_desc)\n",
    "        padding_flag = 'post'\n",
    "#        padding_flag = 'pre' # input은 앞에서 0을 채움\n",
    "        input_pad_seq = tf.keras.preprocessing.sequence.pad_sequences(input_seq, padding=padding_flag)\n",
    "        \n",
    "        return input_pad_seq, tokenizer\n",
    "    else:\n",
    "        tokenizer.fit_on_texts(output_story)\n",
    "        output_seq = tokenizer.texts_to_sequences(output_story)\n",
    "        padding_flag = 'post' # output은 뒤에서 0을 채움\n",
    "        output_pad_seq = tf.keras.preprocessing.sequence.pad_sequences(output_seq, padding=padding_flag)\n",
    "    \n",
    "        return output_pad_seq, tokenizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-09-29T09:05:17.117731Z",
     "start_time": "2020-09-29T09:05:17.108773Z"
    }
   },
   "outputs": [],
   "source": [
    "def get_token_data(data):\n",
    "    # creating cleaned input, output pairs\n",
    "    input_desc, output_story = data['in_desc'], data['out_story']\n",
    "    \n",
    "    tokenize_input_desc, input_tokenizer = tokenize(input_desc, output_story, \"input\")\n",
    "    tokenize_output_story, output_tokenizer= tokenize(input_desc, output_story, \"output\")\n",
    "    \n",
    "    input_desc_train, input_desc_test, output_story_train, output_story_test = train_test_split(tokenize_input_desc, tokenize_output_story, test_size=0.2)\n",
    "    \n",
    "    print(\"input_desc_train: \", input_desc_train.shape)\n",
    "    print(\"input_desc_test: \", input_desc_test.shape)\n",
    "    print(\"output_story_train: \", output_story_train.shape)\n",
    "    print(\"output_story_test: \", output_story_test.shape)\n",
    "    \n",
    "    return input_desc_train, input_desc_test, output_story_train, output_story_test, input_tokenizer, output_tokenizer"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-09-29T09:05:19.302707Z",
     "start_time": "2020-09-29T09:05:17.491358Z"
    }
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input_desc_train:  (31816, 72)\n",
      "input_desc_test:  (7955, 72)\n",
      "output_story_train:  (31816, 77)\n",
      "output_story_test:  (7955, 77)\n"
     ]
    }
   ],
   "source": [
    "# Creating training and validation sets using an 80-20 split\n",
    "\n",
    "input_desc_train, input_desc_test, output_story_train, output_story_test, input_tokenizer, output_tokenizer = get_token_data(clean_data)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-09-29T09:05:19.307572Z",
     "start_time": "2020-09-29T09:05:19.304267Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[   2,   18,   30, ...,    0,    0,    0],\n",
       "       [   2,    1,   69, ...,    0,    0,    0],\n",
       "       [   2,  164, 1313, ...,    0,    0,    0],\n",
       "       ...,\n",
       "       [   2,    1,   58, ...,    0,    0,    0],\n",
       "       [   2,    1,   16, ...,    0,    0,    0],\n",
       "       [   2,    1,   25, ...,    0,    0,    0]])"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "input_desc_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-09-29T09:05:19.315055Z",
     "start_time": "2020-09-29T09:05:19.309247Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[   1,   83, 2101, ...,    0,    0,    0],\n",
       "       [   1,  526,    9, ...,    0,    0,    0],\n",
       "       [   1,    3, 2050, ...,    0,    0,    0],\n",
       "       ...,\n",
       "       [   1,    3, 1038, ...,    0,    0,    0],\n",
       "       [   1,   25,    3, ...,    0,    0,    0],\n",
       "       [   1,  450, 3504, ...,    0,    0,    0]])"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "output_story_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Calculate max_length of the target tensors\n",
    "max_length_story, max_length_desc = output_story_train.shape[1], input_desc_train.shape[1]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-09-29T09:05:19.323326Z",
     "start_time": "2020-09-29T09:05:19.316620Z"
    }
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[1, 2, 4, 8, 41, 82, 97, 164, 194, 328, 388, 776, 3977, 7954, 15908, 31816]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# 배치 단위가 0으로 떨어지게끔 최소공약수 계산해서 배치 설정 숫자 구함\n",
    "\n",
    "def get_divisor(num):\n",
    "    divisors = []\n",
    "    length = int(math.sqrt(num)) + 1\n",
    "    for i in range(1, length):\n",
    "        if num % i == 0:\n",
    "            divisors.append(i)\n",
    "            divisors.append(num // i) # 나누기 연산 후 정수부분만 구하기,\n",
    "\n",
    "    divisors.sort()\n",
    "\n",
    "    return divisors\n",
    "\n",
    "get_divisor(len(input_desc_train))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "BUFFER_SIZE = len(input_desc_train)\n",
    "BATCH_SIZE = 164\n",
    "steps_per_epoch = len(input_desc_train)//BATCH_SIZE\n",
    "embedding_dim = 256\n",
    "units = 1024\n",
    "vocab_inp_size = len(input_tokenizer.word_index)+1\n",
    "vocab_tar_size = len(output_tokenizer.word_index)+1\n",
    "\n",
    "dataset = tf.data.Dataset.from_tensor_slices((input_desc_train, output_story_train)).shuffle(BUFFER_SIZE)\n",
    "dataset = dataset.batch(BATCH_SIZE, drop_remainder=True)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(13341, 14192)"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "vocab_inp_size, vocab_tar_size"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-10-03T05:16:00.985517Z",
     "start_time": "2020-10-03T05:16:00.977109Z"
    }
   },
   "outputs": [],
   "source": [
    "# Prepare training data\n",
    "\n",
    "# prefetch = gpu에 올라갈 데이터 slices\n",
    "train_ds = tf.data.Dataset.from_tensor_slices((input_desc_train, output_story_train)).shuffle(len(input_desc_train)).batch(BATCH_SIZE).prefetch(BATCH_SIZE)\n",
    "test_ds = tf.data.Dataset.from_tensor_slices((input_desc_test, output_story_test)).batch(1) # 하나씩 출력하기 위해서 batch 1"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(TensorShape([164, 72]), TensorShape([164, 77]))"
      ]
     },
     "execution_count": 20,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "example_input_batch, example_target_batch = next(iter(train_ds))\n",
    "example_input_batch.shape, example_target_batch.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Write the encoder and decoder model with attention\n",
    "\n",
    "Implement an encoder-decoder model with attention which you can read about in the TensorFlow [Neural Machine Translation (seq2seq) tutorial](https://github.com/tensorflow/nmt). This example uses a more recent set of APIs. This notebook implements the [attention equations](https://github.com/tensorflow/nmt#background-on-the-attention-mechanism) from the seq2seq tutorial. The following diagram shows that each input words is assigned a weight by the attention mechanism which is then used by the decoder to predict the next word in the sentence. The below picture and formulas are an example of attention mechanism from [Luong's paper](https://arxiv.org/abs/1508.04025v5). \n",
    "\n",
    "<img src=\"https://www.tensorflow.org/images/seq2seq/attention_mechanism.jpg\" width=\"500\">\n",
    "\n",
    "The input is put through an encoder model which gives us the encoder output of shape *(batch_size, max_length, hidden_size)* and the encoder hidden state of shape *(batch_size, hidden_size)*.\n",
    "\n",
    "Here are the equations that are implemented:\n",
    "\n",
    "<img src=\"https://www.tensorflow.org/images/seq2seq/attention_equation_0.jpg\" alt=\"attention equation 0\" width=\"800\">\n",
    "<img src=\"https://www.tensorflow.org/images/seq2seq/attention_equation_1.jpg\" alt=\"attention equation 1\" width=\"800\">\n",
    "\n",
    "This tutorial uses [Bahdanau attention](https://arxiv.org/pdf/1409.0473.pdf) for the encoder. Let's decide on notation before writing the simplified form:\n",
    "\n",
    "* FC = Fully connected (dense) layer\n",
    "* EO = Encoder output\n",
    "* H = hidden state\n",
    "* X = input to the decoder\n",
    "\n",
    "And the pseudo-code:\n",
    "\n",
    "* `score = FC(tanh(FC(EO) + FC(H)))`\n",
    "* `attention weights = softmax(score, axis = 1)`. Softmax by default is applied on the last axis but here we want to apply it on the *1st axis*, since the shape of score is *(batch_size, max_length, hidden_size)*. `Max_length` is the length of our input. Since we are trying to assign a weight to each input, softmax should be applied on that axis.\n",
    "* `context vector = sum(attention weights * EO, axis = 1)`. Same reason as above for choosing axis as 1.\n",
    "* `embedding output` = The input to the decoder X is passed through an embedding layer.\n",
    "* `merged vector = concat(embedding output, context vector)`\n",
    "* This merged vector is then given to the GRU\n",
    "\n",
    "The shapes of all the vectors at each step have been specified in the comments in the code:\n",
    "\n",
    "**Reference**\n",
    "- https://www.tensorflow.org/tutorials/text/nmt_with_attention\n",
    "- https://heung-bae-lee.github.io/2020/01/22/deep_learning_11/"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Encoder(tf.keras.Model):\n",
    "    def __init__(self, vocab_size, embedding_dim, enc_units, batch_sz):\n",
    "        super(Encoder, self).__init__()\n",
    "        self.batch_sz = batch_sz\n",
    "        self.enc_units = enc_units\n",
    "        self.embedding = tf.keras.layers.Embedding(vocab_size, embedding_dim)\n",
    "        self.gru = tf.keras.layers.GRU(self.enc_units,\n",
    "                                       return_sequences=True,  # output을 3차원으로 반환 (batch, sequence_length, #units)\n",
    "                                       return_state=True, # returns the hidden state output and cell state for the last input time step\n",
    "                                       recurrent_initializer='glorot_uniform')\n",
    "    def call(self, x, hidden):\n",
    "        x = self.embedding(x)\n",
    "        output, state = self.gru(x, initial_state = hidden)\n",
    "        return output, state\n",
    "\n",
    "    def initialize_hidden_state(self):\n",
    "        return tf.zeros((self.batch_sz, self.enc_units))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Encoder output shape: (batch size, sequence length, units) (164, 72, 1024)\n",
      "Encoder Hidden state shape: (batch size, units) (164, 1024)\n"
     ]
    }
   ],
   "source": [
    "encoder = Encoder(vocab_inp_size, embedding_dim, units, BATCH_SIZE)\n",
    "\n",
    "# sample input\n",
    "sample_hidden = encoder.initialize_hidden_state()\n",
    "sample_output, sample_hidden = encoder(example_input_batch, sample_hidden)\n",
    "print ('Encoder output shape: (batch size, sequence length, units) {}'.format(sample_output.shape))\n",
    "print ('Encoder Hidden state shape: (batch size, units) {}'.format(sample_hidden.shape))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "---\n",
    "바다나우 어텐션 설명 참고\n",
    "\n",
    "https://www.youtube.com/watch?v=WsQLdu2JMgI\n",
    "\n",
    "<img src=\"imgs/bahdanauAttention.png\" width=\"700\" align='center'>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "class BahdanauAttention(tf.keras.layers.Layer):\n",
    "    def __init__(self, units):\n",
    "        super(BahdanauAttention, self).__init__()\n",
    "        self.W1 = tf.keras.layers.Dense(units) # time step 별 hidden state\n",
    "        self.W2 = tf.keras.layers.Dense(units) # decoder hidden state \n",
    "        self.V = tf.keras.layers.Dense(1)\n",
    "\n",
    "    def call(self, query, values):\n",
    "        # query hidden state shape == (batch_size, hidden size)\n",
    "        # query_with_time_axis shape == (batch_size, 1, hidden size)\n",
    "        # values shape == (batch_size, max_len, hidden size)\n",
    "        # we are doing this to broadcast addition along the time axis to calculate the score\n",
    "        query_with_time_axis = tf.expand_dims(query, 1) # expand_dims를 통해 encoder의 hidden state들을 합쳐줌\n",
    "\n",
    "        # score shape == (batch_size, max_length, 1)\n",
    "        # we get 1 at the last axis because we are applying score to self.V\n",
    "        # the shape of the tensor before applying self.V is (batch_size, max_length, units)\n",
    "        score = self.V(tf.nn.tanh(\n",
    "            self.W1(query_with_time_axis) + self.W2(values))) #FC(tanh(FC(h1,h2,h3)))\n",
    "\n",
    "        # attention_weights shape == (batch_size, max_length, 1)\n",
    "        attention_weights = tf.nn.softmax(score, axis=1)\n",
    "\n",
    "        # context_vector shape after sum == (batch_size, hidden_size)\n",
    "        context_vector = attention_weights * values\n",
    "        context_vector = tf.reduce_sum(context_vector, axis=1)\n",
    "\n",
    "        return context_vector, attention_weights"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Attention result shape: (batch size, units) (164, 1024)\n",
      "Attention weights shape: (batch_size, sequence_length, 1) (164, 72, 1)\n"
     ]
    }
   ],
   "source": [
    "# test\n",
    "attention_layer = BahdanauAttention(10)\n",
    "attention_result, attention_weights = attention_layer(sample_hidden, sample_output)\n",
    "\n",
    "print(\"Attention result shape: (batch size, units) {}\".format(attention_result.shape))\n",
    "print(\"Attention weights shape: (batch_size, sequence_length, 1) {}\".format(attention_weights.shape))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Decoder(tf.keras.Model):\n",
    "    def __init__(self, vocab_size, embedding_dim, dec_units, batch_sz):\n",
    "        super(Decoder, self).__init__()\n",
    "        self.batch_sz = batch_sz\n",
    "        self.dec_units = dec_units\n",
    "        self.embedding = tf.keras.layers.Embedding(vocab_size, embedding_dim)\n",
    "        self.gru = tf.keras.layers.GRU(self.dec_units,\n",
    "                                       return_sequences=True,\n",
    "                                       return_state=True,\n",
    "                                       recurrent_initializer='glorot_uniform')\n",
    "        self.fc = tf.keras.layers.Dense(vocab_size)\n",
    "\n",
    "        # used for attention\n",
    "        self.attention = BahdanauAttention(self.dec_units)\n",
    "\n",
    "    def call(self, x, hidden, enc_output):\n",
    "        # enc_output shape == (batch_size, max_length, hidden_size)\n",
    "        context_vector, attention_weights = self.attention(hidden, enc_output) \n",
    "\n",
    "        # x shape after passing through embedding == (batch_size, 1, embedding_dim)\n",
    "        x = self.embedding(x)\n",
    "\n",
    "        # x shape after concatenation == (batch_size, 1, embedding_dim + hidden_size)\n",
    "        x = tf.concat([tf.expand_dims(context_vector, 1), x], axis=-1)\n",
    "\n",
    "        # passing the concatenated vector to the GRU\n",
    "        output, state = self.gru(x)\n",
    "\n",
    "        # output shape == (batch_size * 1, hidden_size)\n",
    "        output = tf.reshape(output, (-1, output.shape[2]))\n",
    "\n",
    "        # output shape == (batch_size, vocab)\n",
    "        x = self.fc(output)\n",
    "\n",
    "        return x, state, attention_weights"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Decoder output shape: (batch_size, vocab size) (164, 14192)\n"
     ]
    }
   ],
   "source": [
    "decoder = Decoder(vocab_tar_size, embedding_dim, units, BATCH_SIZE)\n",
    "\n",
    "sample_decoder_output, _, _ = decoder(tf.random.uniform((BATCH_SIZE, 1)),\n",
    "                                      sample_hidden, sample_output)\n",
    "\n",
    "print ('Decoder output shape: (batch_size, vocab size) {}'.format(sample_decoder_output.shape))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-09-29T10:27:41.489869Z",
     "start_time": "2020-09-29T09:05:30.051079Z"
    },
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "optimizer = tf.keras.optimizers.Adam()\n",
    "loss_object = tf.keras.losses.SparseCategoricalCrossentropy(\n",
    "    from_logits=True, reduction='none')\n",
    "\n",
    "def loss_function(real, pred):\n",
    "    mask = tf.math.logical_not(tf.math.equal(real, 0))\n",
    "    loss_ = loss_object(real, pred)\n",
    "\n",
    "    mask = tf.cast(mask, dtype=loss_.dtype)\n",
    "    loss_ *= mask\n",
    "\n",
    "    return tf.reduce_mean(loss_)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-10-03T05:07:47.110167Z",
     "start_time": "2020-10-03T05:07:47.104984Z"
    }
   },
   "outputs": [],
   "source": [
    "checkpoint_dir = './training_checkpoints'\n",
    "checkpoint_prefix = os.path.join(checkpoint_dir, \"ckpt\")\n",
    "checkpoint = tf.train.Checkpoint(optimizer=optimizer,\n",
    "                                 encoder=encoder,\n",
    "                                 decoder=decoder)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Training"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {
    "ExecuteTime": {
     "end_time": "2020-10-03T05:16:09.833751Z",
     "start_time": "2020-10-03T05:16:08.178507Z"
    }
   },
   "outputs": [],
   "source": [
    "@tf.function\n",
    "def train_step(inp, targ, enc_hidden):\n",
    "    loss = 0\n",
    "\n",
    "    with tf.GradientTape() as tape:\n",
    "        enc_output, enc_hidden = encoder(inp, enc_hidden)\n",
    "\n",
    "        dec_hidden = enc_hidden\n",
    "\n",
    "        dec_input = tf.expand_dims([output_tokenizer.word_index['<start>']] * BATCH_SIZE, 1)\n",
    "\n",
    "        # Teacher forcing - feeding the target as the next input\n",
    "        for t in range(1, targ.shape[1]):\n",
    "            # passing enc_output to the decoder\n",
    "            predictions, dec_hidden, _ = decoder(dec_input, dec_hidden, enc_output)\n",
    "\n",
    "            loss += loss_function(targ[:, t], predictions)\n",
    "\n",
    "            # using teacher forcing\n",
    "            dec_input = tf.expand_dims(targ[:, t], 1)\n",
    "\n",
    "    batch_loss = (loss / int(targ.shape[1]))\n",
    "\n",
    "    variables = encoder.trainable_variables + decoder.trainable_variables\n",
    "\n",
    "    gradients = tape.gradient(loss, variables)\n",
    "\n",
    "    optimizer.apply_gradients(zip(gradients, variables))\n",
    "\n",
    "    return batch_loss"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Epoch 1 Batch 0 Loss 1.4453\n",
      "Epoch 1 Batch 100 Loss 0.9004\n",
      "Epoch 1 Loss 0.8816\n",
      "Time taken for 1 epoch 5459.4469385147095 sec\n",
      "\n",
      "Epoch 2 Batch 0 Loss 0.7949\n",
      "Epoch 2 Batch 100 Loss 0.7523\n",
      "Epoch 2 Loss 0.7737\n",
      "Time taken for 1 epoch 5795.9613926410675 sec\n",
      "\n",
      "Epoch 3 Batch 0 Loss 0.7723\n",
      "Epoch 3 Batch 100 Loss 0.7081\n",
      "Epoch 3 Loss 0.7195\n",
      "Time taken for 1 epoch 6183.159811735153 sec\n",
      "\n",
      "Epoch 4 Batch 0 Loss 0.7033\n",
      "Epoch 4 Batch 100 Loss 0.6740\n",
      "Epoch 4 Loss 0.6842\n",
      "Time taken for 1 epoch 6078.722980260849 sec\n",
      "\n",
      "Epoch 5 Batch 0 Loss 0.6144\n",
      "Epoch 5 Batch 100 Loss 0.6647\n",
      "Epoch 5 Loss 0.6571\n",
      "Time taken for 1 epoch 6152.46605014801 sec\n",
      "\n",
      "Epoch 6 Batch 0 Loss 0.6154\n",
      "Epoch 6 Batch 100 Loss 0.6473\n",
      "Epoch 6 Loss 0.6324\n",
      "Time taken for 1 epoch 6100.823043346405 sec\n",
      "\n",
      "Epoch 7 Batch 0 Loss 0.5947\n",
      "Epoch 7 Batch 100 Loss 0.6144\n",
      "Epoch 7 Loss 0.6102\n",
      "Time taken for 1 epoch 6099.277186155319 sec\n",
      "\n",
      "Epoch 8 Batch 0 Loss 0.6021\n",
      "Epoch 8 Batch 100 Loss 0.6091\n",
      "Epoch 8 Loss 0.5892\n",
      "Time taken for 1 epoch 6139.639679670334 sec\n",
      "\n",
      "Epoch 9 Batch 0 Loss 0.5583\n",
      "Epoch 9 Batch 100 Loss 0.5723\n",
      "Epoch 9 Loss 0.5685\n",
      "Time taken for 1 epoch 6179.33754324913 sec\n",
      "\n",
      "Epoch 10 Batch 0 Loss 0.5556\n",
      "Epoch 10 Batch 100 Loss 0.5457\n",
      "Epoch 10 Loss 0.5482\n",
      "Time taken for 1 epoch 6238.055964946747 sec\n",
      "\n",
      "Epoch 11 Batch 0 Loss 0.5134\n",
      "Epoch 11 Batch 100 Loss 0.5596\n",
      "Epoch 11 Loss 0.5281\n",
      "Time taken for 1 epoch 6137.887465238571 sec\n",
      "\n",
      "Epoch 12 Batch 0 Loss 0.4883\n",
      "Epoch 12 Batch 100 Loss 0.4828\n",
      "Epoch 12 Loss 0.5079\n",
      "Time taken for 1 epoch 6072.6499009132385 sec\n",
      "\n",
      "Epoch 13 Batch 0 Loss 0.4726\n",
      "Epoch 13 Batch 100 Loss 0.4848\n",
      "Epoch 13 Loss 0.4882\n",
      "Time taken for 1 epoch 6077.006765127182 sec\n",
      "\n",
      "Epoch 14 Batch 0 Loss 0.4652\n",
      "Epoch 14 Batch 100 Loss 0.4338\n",
      "Epoch 14 Loss 0.4680\n",
      "Time taken for 1 epoch 6107.802958488464 sec\n",
      "\n",
      "Epoch 15 Batch 0 Loss 0.4672\n",
      "Epoch 15 Batch 100 Loss 0.4577\n",
      "Epoch 15 Loss 0.4475\n",
      "Time taken for 1 epoch 6111.44526553154 sec\n",
      "\n",
      "Epoch 16 Batch 0 Loss 0.4172\n",
      "Epoch 16 Batch 100 Loss 0.4157\n",
      "Epoch 16 Loss 0.4272\n",
      "Time taken for 1 epoch 6152.159466505051 sec\n",
      "\n",
      "Epoch 17 Batch 0 Loss 0.4010\n",
      "Epoch 17 Batch 100 Loss 0.4130\n",
      "Epoch 17 Loss 0.4071\n",
      "Time taken for 1 epoch 6248.126893043518 sec\n",
      "\n",
      "Epoch 18 Batch 0 Loss 0.3764\n",
      "Epoch 18 Batch 100 Loss 0.3966\n",
      "Epoch 18 Loss 0.3868\n",
      "Time taken for 1 epoch 6030.1285581588745 sec\n",
      "\n",
      "Epoch 19 Batch 0 Loss 0.3442\n",
      "Epoch 19 Batch 100 Loss 0.3639\n",
      "Epoch 19 Loss 0.3669\n",
      "Time taken for 1 epoch 6059.508865118027 sec\n",
      "\n",
      "Epoch 20 Batch 0 Loss 0.3144\n",
      "Epoch 20 Batch 100 Loss 0.3481\n",
      "Epoch 20 Loss 0.3467\n",
      "Time taken for 1 epoch 6031.765414237976 sec\n",
      "\n"
     ]
    }
   ],
   "source": [
    "EPOCHS = 20\n",
    "\n",
    "for epoch in range(EPOCHS):\n",
    "    start = time.time()\n",
    "\n",
    "    enc_hidden = encoder.initialize_hidden_state()\n",
    "    total_loss = 0\n",
    "\n",
    "    for (batch, (inp, targ)) in enumerate(dataset.take(steps_per_epoch)):\n",
    "        batch_loss = train_step(inp, targ, enc_hidden)\n",
    "        total_loss += batch_loss\n",
    "\n",
    "        if batch % 100 == 0:\n",
    "            print('Epoch {} Batch {} Loss {:.4f}'.format(epoch + 1,\n",
    "                                                         batch,\n",
    "                                                         batch_loss.numpy()))\n",
    "    # saving (checkpoint) the model every 2 epochs\n",
    "    if (epoch + 1) % 2 == 0:\n",
    "        checkpoint.save(file_prefix = checkpoint_prefix)\n",
    "\n",
    "    print('Epoch {} Loss {:.4f}'.format(epoch + 1,\n",
    "                                      total_loss / steps_per_epoch))\n",
    "    print('Time taken for 1 epoch {} sec\\n'.format(time.time() - start))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Generate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [],
   "source": [
    "def evaluate(sentence):\n",
    "    attention_plot = np.zeros((max_length_story, max_length_desc))\n",
    "\n",
    "    sentence = preprocess_sentence(sentence)\n",
    "\n",
    "    inputs = [input_tokenizer.word_index[i] for i in sentence.split(' ')]\n",
    "    inputs = tf.keras.preprocessing.sequence.pad_sequences([inputs],\n",
    "                                                         maxlen=max_length_desc,\n",
    "                                                         padding='post')\n",
    "    inputs = tf.convert_to_tensor(inputs)\n",
    "\n",
    "    result = ''\n",
    "\n",
    "    hidden = [tf.zeros((1, units))]\n",
    "    enc_out, enc_hidden = encoder(inputs, hidden)\n",
    "\n",
    "    dec_hidden = enc_hidden\n",
    "    dec_input = tf.expand_dims([output_tokenizer.word_index['<start>']], 0)\n",
    "\n",
    "    for t in range(max_length_story):\n",
    "        predictions, dec_hidden, attention_weights = decoder(dec_input,\n",
    "                                                             dec_hidden,\n",
    "                                                             enc_out)\n",
    "\n",
    "        # storing the attention weights to plot later on\n",
    "        attention_weights = tf.reshape(attention_weights, (-1, ))\n",
    "        attention_plot[t] = attention_weights.numpy()\n",
    "\n",
    "        predicted_id = tf.argmax(predictions[0]).numpy()\n",
    "\n",
    "        result += output_tokenizer.index_word[predicted_id] + ' '\n",
    "\n",
    "        if output_tokenizer.index_word[predicted_id] == '<end>':\n",
    "            return result, sentence, attention_plot\n",
    "\n",
    "        # the predicted ID is fed back into the model\n",
    "        dec_input = tf.expand_dims([predicted_id], 0)\n",
    "\n",
    "    return result, sentence, attention_plot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [],
   "source": [
    "# function for plotting the attention weights\n",
    "def plot_attention(attention, sentence, predicted_sentence):\n",
    "    fig = plt.figure(figsize=(10,10))\n",
    "    ax = fig.add_subplot(1, 1, 1)\n",
    "    ax.matshow(attention, cmap='viridis')\n",
    "\n",
    "    fontdict = {'fontsize': 14}\n",
    "\n",
    "    ax.set_xticklabels([''] + sentence, fontdict=fontdict, rotation=90)\n",
    "    ax.set_yticklabels([''] + predicted_sentence, fontdict=fontdict)\n",
    "\n",
    "    ax.xaxis.set_major_locator(ticker.MultipleLocator(1))\n",
    "    ax.yaxis.set_major_locator(ticker.MultipleLocator(1))\n",
    "\n",
    "    plt.show()\n",
    "\n",
    "def generate(sentence):\n",
    "    result , sentence, attention_plot = evaluate(sentence)\n",
    "\n",
    "    print('Input: %s' % (sentence))\n",
    "    print('Predicted translation: {}'.format(result))\n",
    "\n",
    "    attention_plot = attention_plot[:len(result.split(' ')), :len(sentence.split(' '))]\n",
    "    plot_attention(attention_plot, sentence.split(' '), result.split(' '))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 41,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<tensorflow.python.training.tracking.util.CheckpointLoadStatus at 0x1d114c0c780>"
      ]
     },
     "execution_count": 41,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# restoring the latest checkpoint in checkpoint_dir\n",
    "checkpoint.restore(tf.train.latest_checkpoint(checkpoint_dir))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Input: <start> an empty road that does not have any cars <end>\n",
      "Predicted translation: the hotel was beautifully constructed and clean <end> \n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Input: <start> the men are gathered around for the picture <end>\n",
      "Predicted translation: the friends talked about the people <end> \n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Input: <start> view of open hymn book inside a very large cathedral <end>\n",
      "Predicted translation: the park was auctioned off <end> \n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Input: <start> males smiling at camera man in middle wearing white tank top <end>\n",
      "Predicted translation: i had a party <end> \n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Input: <start> several women hold onto a sign for a parade as a frog walks alongside <end>\n",
      "Predicted translation: the family and i went to the end of the race <end> \n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Input: <start> a guy in red black and white is wheeling his bike up <end>\n",
      "Predicted translation: the parade was about to begin their team <end> \n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Input: <start> a cross is a very large monument at the grave yard <end>\n",
      "Predicted translation: the grounds were a sight <end> \n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Input: <start> an artistic structure that is in a building <end>\n",
      "Predicted translation: the grounds were amazing <end> \n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAn8AAAGnCAYAAADGyzEkAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy86wFpkAAAACXBIWXMAAAsTAAALEwEAmpwYAAArZElEQVR4nO3dd7xsVX338c/3FkCaLVJEFOyiKMi1YAMidmIi9oiK+IiPwURjsBfUJ2gwWLDEQBJRbEFFxUJULGCDIKACogIq0kRAEbiUey/we/7Y+8gwnNu4Z86emf15v17ndWfW3jPzW5zDzHfWXnvtVBWSJEnqhwVdFyBJkqT5Y/iTJEnqEcOfJElSjxj+JEmSesTwJ0mS1COGP0mSpB4x/EmSJPWI4U+SJKlHDH+S1liSzZP4viFJE8w3cUmrlGRxkncluQq4ENimbT8oyd91Wpwkaa0Z/iStzgHAXwF7AcsG2k8C9u6iIEnSrbeo6wIkjb3nAvtU1fFJbhxoPwO4d0c1SZJuJUf+JK3OnYHfztK+CL9AStLEMfxJWp2fAY+Zpf1ZwCnzXIskaR35rV3S6rwN+ESSrYGFwDOT3Bf4W+ApnVYmSVprqaqua5A05pI8AXgDsBPNEYNTgbdX1Tc6LUyStNYMf5JWKsli4EDgQ1U127w/SdKEMfxJWqUkS4EHVNW5XdciabokmW0+MUAB1wG/qqo/zmNJvWD4G0NJ7gUcCryiqk7vuh71W5KjgK9W1Ue6rkXSdGmXj5oJImn/Hbx/I/Al4PlVdfU8lze1PNt3PL0Q2BXYp+M6JIBvAe9I8r4kz0+y5+BP18VJmmhPAX5Os4j8PdufvWhWGXh6+7MD8C8d1TeVHPkbM0kCnAscS3NVhTtX1Q2dFqVeG1rYeVhV1cJ5K0bSVElyCvCaqvrWUPvuwEFVtVOSPYAPVNW2nRQ5hRz5Gz+7AZsA/wBcDzy523LUd1W1YBU/Bj9J62I7mmuGD7uw3QZwOrDFvFXUA4a/8fMC4HNVdQ3waZpDwJIkTaMzgTcmWX+mob39hnYbwNbAxR3UNrVc5HmMJNkI2JObFs79BHBCkttX1eXdVaY+S/KqVW2vqvfMVy2Sps7fAV8GLkxyBs3JHtvTnOixR7vP3YF/66a86eScvzGS5AXAW6vq7gNtpwEfrqoPd1eZVifJM4HlVXX0UPtfA4ur6nPdVLbukvxmqGkxsCVwLXDJ4N+rJK2tduBjL+A+NGf4/hz45DSd3dv28enA0VV1Rdf1OPI3Xp5PM9o36BM0h34Nf+PtrcBsI2RXA+8DJjb8zTbJOsnmwOHAf8x/RZKmSRvyDu26jhF7FvCfwCuAD3ZciyN/46K9bupvgPtV1dkD7XehOft3u6o6q6PytBpJrqX53Z071L4NcGZVbdhFXaOUZEfgM1V1r65rkTS52s+/RwObMXQuwrRMK0lyHE3/rqmqJR2X48jfuKiq85nl91FVF8zWrrFzOXAvmqA+6N7AVfNezfxYAGzedRGSJleS5wEfoVnd4lJuWuCZ9vbEh792EOARwEOBE5NsV1VnrvpRI67Jkb/xkeSuwPk1yy8lyV2r6rwOytIaSPJhmm+ue86M0Ca5D3AU8IOqemmX9a2LWRZyDs2cv/2AX1fVU275KElavSS/Ao4E3jyta9omeTOwa1U9NsnngbOr6rWd1mT4Gx9JbgC2rKpLhtrvSDOx3jXVxlSSTYCvAQ8Dftc2bwmcBDyxqq7sqrZ1Ncsiz0XzDf3bwD9V1e9u+ShJWr322uEPrKpfd13LqCQ5Gziwqj7afpl+P7D1bAM988XDieMl3HzIe8bGNBe41piqqquARyZ5HM2liAKcCnyry//B50JVuR6opFE5huZL81SGvySPoBkI+Gzb9BWaEz92p7mSVycMf2MgyfvbmwW8M8k1A5sX0swT+Ml816W1V1XH0uH/0KPQLkF0ZFUtG2pfD3hOVR3RTWWSpsCxwEFJ7k9zJY8Vgxur6vOdVDV3XkizvMvVAFW1PMlngL3p8LPCw75jIMl32pu7ACcAywc2L6c5ieDgwbOA1b128eN/q6rrpnkhZKcjSBqVab52eHulkouB51bV1wbaHwV8Hdi8qpZ2UpvhbzwkCfAZYJ/2EKLGXLv48ZKq+sMsCyEPqkleCLl9c968qi4dat+R5rD2HbqpTJLGV5K/AJ4MfKKqbhzathfwzarq5LJ1hr8xkWQhzby+B3V9CrgEkOR0mqkI9wd+SbMUw4yFwN2AY6rqWR2UJ0m6lZzzNyaq6oYkvwXW67oWrb0pnRc3c1WSBwBfBQYPT8xMRzhqnmuSNOH6MmVmnDnyN0aSvBB4LrBXVV3WdT1ac9M8L679u/zv4WArSbfGtE+Zafu0RuGqq/458jde9ge2BS5McgHNdWH/rKoe2ElVWhMrW6bnrkDnF/FeR+cCDweOH2xMsgvNm/N3uyhK0mQavF74bNcOnwKD1+7dmOa67yfRnNAJsDPNKh7vnue6/szwN14+t/pdNE4G5sUVcHySWefFdVHbHHov8PZZ2jcF3grsNK/VSNIYq6o/h7okHwUOqqp3DO6T5PU086k74WFfaR0kOaC9eQDNt7hZ58VV1XImVJKrgQdU1W+G2rcBzqiqjTspTNJESvKWNd23qmb74jkxklwJPLiqzhlqvydwalVt2kVdjvxJ66Cq3gaQ5Fymd17ctcCdgeG5OXfh5mtSStKaeObQ/bsBGwIXtffvDFxD8+V5osMfzfStXYFzhtp3peljJwx/Y6Q9M/SNNCd93BVYPLh9kk8a6IFjaA6DXgqQZHvg2cDPqurTXRY2B74O/EuSp1bV5QBJ7gC8o90mSWusqrafuZ3kRcALgBdW1Xlt212Bw4FPdlPhnHov8KEkS4AT27aH01z5461dFeVh3zGS5CCawPBOmj+YNwHbAM8B3lxVh3ZXnValvUrLx6vqI+3CnmfTfIu9C/D2wTkgkybJlsB3gc2A09rmBwKXALtU1UUre6wkrUp7ZuzfVNVPh9p3oLks2t06KWwOJXkW8Argfm3Tz4FDquozndVk+Bsf7f8EL6uqryW5Ctihqn6V5GXAY6vqGR2XqJVI8gfg0VV1ZpL/C7y4qh6S5K+Bf62qe3dc4jpJsiHwPGAHmjObTwU+VVWdHbaYK+0ow/k19GbYXnVn65nRCElzr72W/V9W1YlD7Q+nuYLQRt1UNt087DteNgdmru6xFLhde/trwEFdFKQ1dhtuOtljd+BL7e1Tga07qWgOtSHvP7quY0R+A2xJM5I56A7tNqdbSKNzLPAfSV4C/KhtewhwaLttaiS5HbBgsK2q/thFLYa/8XIezUTX82gmhz4BOIVmTaBrO6xLq3c2sGeSo4DHA//atm8O/KmrouZCkj1Xtb2qPj9ftYzIytZo3JjmkouSRuf/AB8Dfgjc0LYtoJlP/JKuiporSe4G/DuwGzefxz/zvtPJl0vD33j5AvBYmkmhhwCfbr8NbcVNYULj6W3Ap2mWe/lWVf1v2/4E4MedVTU3Vrb+5ExgmsiRsSTvb28W8M728NOMhTSLsP5kvuuS+qSqLgWenOTewH1pQtHPq+qsbiubM4fTHMXbh2Ye+FjMtXPO3xhL8jDgkcBZVfWVruvRqiXZnGbk9qdVdWPb9jDgiqr6RafFzaEki4Adab6QvLGqftBxSbdKe5IOwC40K+8PLlszs0bjwVV19jyXJmlKJFkKPLyqzui6lkGGvzGS5DHAD6vq+qH2RcAjvIyWxkmSRwAfrqoHdV3LukhyOPCKqrqy61qkPhgYdV+tqvqHUdYyau1VoPauqlO6rmWQ4W+MJLkB2LKqLhlqvyNwiev8jZf2Dez1VXX16t7MJv0NbDZJtgNO8gofktbGwKj76lRV/eVIixmxJH8JvA74u+GrfHTJOX/jZWUTz+9Is0q4xsv23DSBd/tV7TjJkjx4uInm7NjXMvnzGQFIshs3La6+3uC2Sf/wkcZNVe3WdQ3z6GhgfeCXSZYBNzuy5+XdeizJzLIgBXyi/QOZsRB4AM2ZUBojg29gU/5mdjLN32aG2k8EXjT/5cytJHvTnI33BZpLLh0N3BvYFvhEZ4VJmgYv77qA2Rj+xsMf2n8DXM7Nl3VZDnyf6V1jbSq0Fyo/eHjR4yS3AV494Rcn33bo/o3ApVU1Lcug7A+8vKr+s11c/fVV9eskH+SmtRslzZE+TZmpqo91XcNsnPM3RpIcQBMgPMQ7YaZ1vmaSxTSjX2+oql91Xc8otEu8bFdV5ya5jOZqA6cluS9wXFVt0XGJ0lRp5/w9rar+tJr5fxM/5w/+vBLE84F70Fyq9bIkjwQuqqrfdFGTI3/j5f8N3kmyBbAHcGZVedh3vK1svuaOQCcruM+FqlqR5PHA67uuZYT+AGzS3r6QZprFaTRzbW/TVVHStOrRlBmS7AR8i+ZqQfenWSLrMuBxNNNL/raLugx/4+WrNJdyOyTJxjRzrTYCNk7y4qo6otPq1lF7fdgdgM245SVuJvIqEe1hwmp/fp1kMAAuBDagmU82yT4P7Akc3HUhI/I9mquynA58Bnh/ksfRLLg+VZeXkjTvDgYOqaoD2s+LGV+nwznThr/xshPwmvb2nsCVNPOtnkczL2liw1+S3WmugHHHWTZ3dombOfBymlG/jwBvBK4Y2LYcOLeqTuiisDl0HvCmJI+m+UJys2kJVfWeTqqaOy+nCekA76Q5G++RNEHwn7sqSuqDaZ/zR/O5/uJZ2n9Hc/nPThj+xssm3HQd2McDX2gPu30b+FBnVc2NQ2hGNt9QVRd1XcxcqaqPtYtwbwQcXVUXdF3TCOxNcyLSA9ufQQVMdPgbvLB6e2WWgzosR+qb4WWyFtNc5m0RcOr8lzPnrgVuP0v7fYFLZmmfF4a/8XIe8MgkX6a5Juwz2/Y7ANes9FGTYRvgqdMU/GZU1fVJ3gVM5SX4qmr4bN+plOTOzD4lYRo+gKSxNNucvyQbAP9FMyVj0h0NHJBk5vO8kmxD8yXzqK6KWrD6XTSP3gN8HLiAZuL5zOXcHkMzH2mS/QC4T9dFjNCJNMP7UyfJW9r5msPtt2mXuJloSXZM8jPgfJqRhpMHfn7UZW1SH7XLSB1IM5Vm0u1PM4BzKbAhzdJt59BMEXpTV0W51MuYac8MuitwbFUtbdueAvypqn7QaXHrIMmeNPOn3kMTZFcMbp/00ZUkzwHeAbwfOIVbzoub2P5N6zI2M5L8iOaM37cDFzF01nZV/baLuqQ+S7IL8MWqmu2Q6cRpL/P2YJpBt1Or6pud1mP4Gw9Jbgs8sKpuMczdrgd0ZlVdPv+VzY0kN65ic01BgJja/rV927yqLh1q3x34dFXdqZvK5kaSq4Edq+qsrmvRrZPk2TRnZ8922P6pnRSlNZLkVcNNNJePfB7w7ap63vxXNTfG+XPdOX/j40bgf5I8YXCEL8kOwLeBrboqbI5M+7yxqetfT5axgWYkegtgKsNfku2AG6rql+39xwEvBH4GvKuqbuiyvnWV5F+BVwLfYZaRW429vx+6fyPNIdLDac6+n2Rj+7nuyN8YSfJJYGlVvXSg7WDg3tPw7bU9K/ahNIe11xvYVFX18W6qmjvT1r8kL+SmZWxeyRQtY5PkDgN3d6A5ZP8mZp+SMLGLdAMkOYFmnbH/TnIXmpB7HM2Z2x+vqolewDvJ74H9qupzXdcyKqt4b2HS138d1K5vy8yUp2kwrp/rhr8xkuQJNGvhbd4u8bKA5uSPl0/qIsgz2ktlfZlmhCzADTQjzyuAZVW1aYflrbNp7l+S/YDvVtXp7f2JHzlqD2UPvvml/Xe4baIP2QMk+RPw0Ko6K8k/0px1v1uS3YDDq2qbTgtcR0kuBXauqnO6rmUUpvm9ZUaSVwKv4qaRsIto5oe/ryY8pIzr57qHfcfLsTRLuvwVzVUVHkvzLe/LXRY1R95HcyLEDsDF7b+3BT5Mh2c8zaH3Mb3924vmhIjT25GjLwLHA/sBmzKZl34bXF5iG5ozfYdD7AKakZZJt5BmpBaa95Rj2tu/osNFZufQYTR/o2/tuI5ReR/T+95Cu0zWvjSXPZs5krAz8BaauX+vWclDJ8VYfq478jdmkhwE3Keq/ibJEcBVVbVf13WtqyR/AHapqjOSXEEzEvHL9oyuD1TV8OLBE2Wa+9eDkaNpP5v5BJplo74CfIPmd3l6kp2Bz1TV1p0WuI6SfIjm+qhn0lyTefiw/URfIWKa31sAkvwR2Hf4sH2SZwCHVtVsV4WaKOP4ue7I3/g5AjglydbA02i+JUyDcNNC1ZfSDO//kmb4+55dFTWHprl/0z5yFGY/SWBj4Lp5rmUUXkszWvtq4KMzh++BpwIndVXUHNoO+El7+74d1jEq0/zeMuO0lbRNy1rEY/e5bvgbM1X1sySnA58CLqiqaXhzBjgDeBDwa5oPnNe2Iy4voVnwctJNc//OAF6W5Cs0b1ozh3m3Ai7rrKp1NHBN0QLemWTwKjoLaSbY/2S+65prVfXdJHcCNh1aVuJQhtajnESzXSFiykzzews0wWg/4BVD7S+juejBxBvHz3XD33j6OM08j2lY3XzGgTTXv4VmnspXaJZmuAx4VldFzaFp7t/MyNH+wMemaORo5pqiAe7HTaObtLdPBQ6e76LmQpIvAXtV1ZXt7Zn22XafuJUEVta/WVRV/fV81TUiU/feMvDFC5ocsld7YsSJbdvDgDsDn5zv2kZorD7XDX/j6RM0F4I+vOtC5kpVfX3g9q+B7drlNi6f9LO5YLr7t5qRo4m95vTMiFGSw4FXVNWVHZc0l/7ATYey/9BlISMy7f37syl9b9l+6P4p7b93a/+9uP2ZpsP4Y/W57gkfkiRJPTItkyklSZK0Bgx/YyrJvl3XMEr2b7JNc/+muW9g/yad/Ztc49Q3w9/4Gps/khGxf5Ntmvs3zX0D+zfp7N/kGpu+Gf4kSZJ6xBM+1tB6Wb82+PPZ9qO3gmUsZv15e735Nu/9m3WFi9FZUctYnPnpX9af/7+T5Tdcw3oLN5yX16qF8/sddcWKq1m8eP7+X59vK66/msWL5rF/8/3/3nz//ub5I3Ref3/z/LuD+f391cL57eCK5VezeL35+9tcesWFl1XVnWbb5lIva2gDNuJhC3bvugzdSlk40VfoWqUFd9+26xJG6vrbzU/I7EwHH7Dzava1BafHNA+gTPnvbtnt1+u6hJH6/pdf89uVbfOwryRJUo8Y/iRJknrE8CdJktQjhj9JkqQeMfxJkiT1iOFPkiSpRwx/kiRJPWL4kyRJ6hHDnyRJUo8Y/iRJknrE8CdJktQjhj9JkqQeMfxJkiT1iOFPkiSpRwx/kiRJPWL4kyRJ6hHDnyRJUo8Y/iRJknrE8CdJktQjhj9JkqQeMfxJkiT1yNSEvyS7Jqkkf9F1LZIkSeNqYsNfkuOSfLDrOiRJkibJxIY/SZIkrb2JDH9JPgrsAuzXHuotYJt284OS/G+Sa5KcnOTBQ499RJLj2+0XJvlwkk3ntweSJEndmMjwB7wCOAE4HNiy/Tm/3fZO4HXAg4E/AJ9MEoAk2wPfAL4EPAjYE9gB+Mg81i5JktSZRV0XcGtU1RVJlgPXVNXFAEnu225+c1V9p217O/B9YCvgAuDVwJFV9e6Z50ryMuDHSTarqksGXyfJvsC+ABuw4Yh7JUmSNHoTGf5W47SB2xe1/25GE/52Au6Z5NkD+6T99x7AzcJfVR0GHAawae5QI6lWkiRpHk1j+FsxcHsmsC0Y+Pc/gffO8rgLR1mUJEnSOJjk8LccWLiWjzkVuH9VnTOCeiRJksbepJ7wAXAu8NAk27QLO69JXw5qH/PvSXZMcs8keyQ5dKSVSpIkjYlJDn8H04z+nQlcCtx1dQ+oqtOAx9AsC3M88FOas4N/P7IqJUmSxsjEHvatqrOAnYeaPzq0z7ncdELHTNvJwBNHWZskSdK4muSRP0mSJK0lw58kSVKPGP4kSZJ6xPAnSZLUI4Y/SZKkHjH8SZIk9YjhT5IkqUcMf5IkST1i+JMkSeoRw58kSVKPGP4kSZJ6xPAnSZLUI4Y/SZKkHjH8SZIk9YjhT5IkqUcMf5IkST1i+JMkSeoRw58kSVKPGP4kSZJ6xPAnSZLUI4Y/SZKkHlnUdQETJdOblbN4yv8UHnjvrisYmed94piuSxipd3z02V2XMFJbnnhd1yWM1ILrbui6hJFadMW1XZcwMrluedcljNSCZRt3XUJnpjfNSJIk6RYMf5IkST1i+JMkSeoRw58kSVKPGP4kSZJ6xPAnSZLUI4Y/SZKkHjH8SZIk9YjhT5IkqUcMf5IkST1i+JMkSeoRw58kSVKPGP4kSZJ6xPAnSZLUI4Y/SZKkHjH8SZIk9YjhT5IkqUcMf5IkST1i+JMkSeoRw58kSVKPGP4kSZJ6xPAnSZLUI70If0mWJtm76zokSZK61ovwJ0mSpMbIw1+S9Ub9GpIkSVozax3+kmyU5Ij2UOrvk7w+yVeSfLTdfm6Styb5SJI/AZ9s2/dMcnqSZUnOT/LGJBl43nOT7D/0Wscl+eDQPm9KcmiSK5NckOTVQ4+5Z/u465L8Mskes/ThLUl+29ZycZIj1va/gyRJ0iS6NSN/7wZ2AZ4G/CXwIODRQ/u8CvgFsAR4Q5KdgM8Cnwe2B14HvB54+a14/X8ETgceDBwEvCvJzgBJFgBfoOnXzsA+wFuB9WcenOTpwP7A3wH3AvYATroVdUiSJE2cRWuzc5KNaQLVC6rq2LbtxcAFQ7seX1XvGnjcJ9u2A9qms5LcC3gt8IG1rPkbVTUzGviBJP8APBY4Adgd2A7YtqrOa1/7lcD3Bh5/N+B37fOsAM4DTl5Jf/cF9gXYgA3XskxJkqTxs7Yjf/cAFjMwUlZVVwNnDO03HKbuB/xgqO37wFZJNl3LGk4bun8RsNnA61w4E/xa/wvcOHD/s8AGwG+S/FeSZyZZn1lU1WFVtaSqlixm1l0kSZImytqGv5k5erWa/a6e5XEre8xM+40Dzz9j8Sz7r5jl8TP9GH78LV+s6nzgPsBLgStpDmOfkmSj1T1WkiRp0q1t+DuHJnw9dKYhyYbAA1bzuDOBRw21PQq4oKquau9fCmw58LwbAPddy/rOpBlN3Hqg7aEM9bOqrquqr1bVPwIPAe4PPHItX0uSJGnirNWcv6pamuQjwEFJLqOZO/cmmnC1qtHAdwM/SvJW4FM0geufgDcM7PNtYJ8kX6IJgm9k9pG/VfkmzYkmRyT5R+A2wHuB62d2aBd7XkRzOHgp8GyaQHv2Wr6WJEnSxLk1Z/vuT3MCxZeA79DMwTsZuG5lD6iqU4FnAk+nmR/4L+3PBwd2eydNADwa+AbNnMBT16awqrqR5izkBTTh7gjgn4FlA7v9CXhx24cz2pr2rKrfrM1rSZIkTaK1GvmDZvQPeH77Q3uyxCuBY9rt26zkcZ+nWeplZc97JfDcoeZ/G9rnFs9dVbsO3T+LZimaQRsPbP8i8MWV1SFJkjTN1jr8JdmR5qzak4BNaJZr2QQ4cm5LkyRJ0lxb6/DXehXNGbPXAz8BHlNVw2v9SZIkaczcmsO+P6a5cockSZImzK054UOSJEkTyvAnSZLUI4Y/SZKkHjH8SZIk9YjhT5IkqUcMf5IkST1i+JMkSeoRw58kSVKPGP4kSZJ6xPAnSZLUI4Y/SZKkHjH8SZIk9YjhT5IkqUcMf5IkST1i+JMkSeoRw58kSVKPGP4kSZJ6xPAnSZLUI4u6LmBSZMECFmy0YddljMyNV1/TdQkjddEum3RdwsiccOU9uy5hpLY84bquSxip9S69uusSRuv6G7quYKRyzRT/fa5Y0XUFI7Xwhun+21wVR/4kSZJ6xPAnSZLUI4Y/SZKkHjH8SZIk9YjhT5IkqUcMf5IkST1i+JMkSeoRw58kSVKPGP4kSZJ6xPAnSZLUI4Y/SZKkHjH8SZIk9YjhT5IkqUcMf5IkST1i+JMkSeoRw58kSVKPGP4kSZJ6xPAnSZLUI4Y/SZKkHjH8SZIk9YjhT5IkqUcMf5IkST1i+JMkSeoRw58kSVKP9Cb8JVmv6xokSZK6NjbhL8mTklyVZFF7/15JKsmHB/Y5MMmx7e3tkny1fcwlST6dZIuBfT+a5CtJXpvkAuCCtn2rJP+d5PL256tJ7jXP3ZUkSerE2IQ/4HvABsCS9v6uwGXAbgP77Aocl2RL4LvAGcBDgd2BjYEvJRns0y7AA4EnAo9NsiHwHeC6dtvOwO+Ab7bbJEmSptrYhL+qWgqcyk1hb1fgg8DdkmzZhrOHAMcBLwN+WlWvraqfV9VpwAva7UsGnvY6YJ+qOqOqTgeeAwR4UVWdVlW/AF5KExz3GK4pyb5JTk5y8vK6bu47LUmSNM/GJvy1jqMJfdCMzP0PcFLb9khgRXt/J+AxSZbO/ADnt4+7x8DznVFVywbu7wRsC1w18LgrgNsPPQ6AqjqsqpZU1ZL1ssHc9FCSJKlDi7ouYMhxwH5JtgM2AU5p23YDLgV+WFUr2kO7XwX2n+U5fj9w++qhbQuAn9CMAA7747oULkmSNAnGLfx9D1gfeA3w/aq6IclxwGHAJcAx7X6nAs8CfltVK9bi+U8FngtcVlV/mquiJUmSJsVYHfYdmPe3F82JGQAnAFsDD6MZBQT4EHBb4MgkD0ty9yS7JzksySareIlP0owMHp1klyTbJnlMknd7xq8kSeqDsQp/re8AC2mDXlVdB5wILKOZ70dVXUQzB/BG4GvAz2gC4bL2Z1ZVdQ3wGODXwGeBXwAfo5nzd/koOiNJkjROxu2wL1X1OuB1Q227zrLf2cAzVvE8e6+k/ffAi9apSEmSpAk1jiN/kiRJGhHDnyRJUo8Y/iRJknrE8CdJktQjhj9JkqQeMfxJkiT1iOFPkiSpRwx/kiRJPWL4kyRJ6hHDnyRJUo8Y/iRJknrE8CdJktQjhj9JkqQeMfxJkiT1iOFPkiSpRwx/kiRJPWL4kyRJ6hHDnyRJUo8Y/iRJknrE8CdJktQji7ouYFIsv9NtuPD523ddxshsfdQFXZcwWum6gNE55sz7d13CSN3lttP9NrX+ecu6LmG0li3vuoKRquuv77qEkcmCKR8fmvb+rUJ/ey5JktRDhj9JkqQeMfxJkiT1iOFPkiSpRwx/kiRJPWL4kyRJ6hHDnyRJUo8Y/iRJknrE8CdJktQjhj9JkqQeMfxJkiT1iOFPkiSpRwx/kiRJPWL4kyRJ6hHDnyRJUo8Y/iRJknrE8CdJktQjhj9JkqQeMfxJkiT1iOFPkiSpRwx/kiRJPWL4kyRJ6hHDnyRJUo9MbPhLsneSpV3XIUmSNEkmNvwBRwJ377oISZKkSbKo6wJuraq6Fri26zokSZImyRqN/CV5YpLvJbk8yR+TfD3J/dpt2ySpJM9JcnySa5P8OMkDkzwgyQ+TXJ3k+0m2HXjOeyQ5OsnF7fZTk+wxsH3v9nmHfz46sH3pwP5vTXJGW8evklyV5ItJ/mJgn0VJ3tv24/L29oeTHLfu/yklSZLG35oe9t0IeB/wUGBX4Argy0nWG9jnbcBBwI7An4BPAR8A3tg+bgPg/QP7bwz8D/A44EHAUcDnk9y33X4ksOXAzxOA5cDxq6hzG+DZwNOAx7e1HDiwfX9gb+D/AA+n6f/frrb3kiRJU2KNDvtW1VGD95O8CLiSJtRd0Da/p6qOabe/G/gy8PSq+k7b9kHggwPP+VPgpwNPe2CSvwKeAfzz4GHdJHcCDgM+XFWHr6Y/e1fVFe3jDgNeNLD9FcBBM/1J8kqaUDmrJPsC+wIs3uT2q3hZSZKkybCmh33vkeRT7eHUK4Hft4+968Bupw3c/n377+lDbRsl2bB9zo2SvCvJme0h2KXAkqHnpB1d/ALwc+CfVlPqb2eCX+siYLP2eW4LbAGcNLOxqgr40cqerKoOq6olVbVk4YYbrealJUmSxt+anvDxZeBC4KXtv9cDZwKDh31XDNyuVbTNBM6DgSfSHIo9G7gGOGLoOQH+Hbg98OSqumE1da4Yul/cMuAWkiRJPbXakb8kdwTuB7yjqr5ZVT8HNmHdzxR+FHBEVR1VVafRHD6+x9Br7w/8FbBHVV25Li/WjgheTHOoeub5AzxkXZ5XkiRpkqxJgLscuAx4SZLzga2Af6UZ/VsXZwFPS3I0zYjdATQnhQCQZHfgHcDzgGuTbNFuunbo0O7aOAR4TZKzaEYuX0pzMsnvbuXzSZIkTZTVjvxV1Y00Z9A+EDgD+BDwZmDZOr72q4BLgO/RnPV7Ynt7xqOAxcBnaMLZzM8h6/CaBwMfBw5vXw+a+YTXrcNzSpIkTYw05zz0V5JTgR9U1d+var/bbLF13eP5r5qnqubf1kddsPqdJth5z7xL1yWMzNUPmO7vLnf5/MSuRb9GNjn9kq5LGK1ly7uuYKTq+nU9CDa+smCSLwK2BjZYv+sKRuprv373KVW1ZLZt0/2uOiTJ3WiWdjmepu/70qwxuG+XdUmSJM2XXoU/4EbgBTRzFhfQzPt7UlWd3GlVkiRJ86RX4a+qzqeZSyhJktRLU35AX5IkSYMMf5IkST1i+JMkSeoRw58kSVKPGP4kSZJ6xPAnSZLUI4Y/SZKkHjH8SZIk9YjhT5IkqUcMf5IkST1i+JMkSeoRw58kSVKPGP4kSZJ6xPAnSZLUI4Y/SZKkHjH8SZIk9YjhT5IkqUcWdV3ApNhu80v44as+0HUZI7P4nxZ2XcJIPenxz+m6hJHJV27ouoSRytJruy5hpOra6e4fy1d0XcFI1fLlXZcwMrVguseHsuI2XZfQmen+zUqSJOlmDH+SJEk9YviTJEnqEcOfJElSjxj+JEmSesTwJ0mS1COGP0mSpB4x/EmSJPWI4U+SJKlHDH+SJEk9YviTJEnqEcOfJElSjxj+JEmSesTwJ0mS1COGP0mSpB4x/EmSJPWI4U+SJKlHDH+SJEk9YviTJEnqEcOfJElSjxj+JEmSesTwJ0mS1COGP0mSpB6ZuvCXZP8k53ZdhyRJ0jiauvAnSZKklZvX8Jdk0yS3m+fXvFOSDebzNSVJksbVyMNfkoVJnpDkU8DFwIPa9tsmOSzJJUmuSnJ8kiUDj9s7ydIkj01yRpKrk3wnybZDz/+aJBe3+x4BbDxUwpOBi9vXeuSIuytJkjTWRhb+ktw/ybuA84AjgauBJwLfTRLgq8BWwB7AjsB3gW8n2XLgadYHXg/sA+wM3A7494HXeBbwz8ABwIOBXwKvGirlk8DfApsAxyY5J8lbhkOkJElSH8xp+EtyxyT/kORk4MfAfYFXAptX1Uuq6rtVVcBuwA7AM6rqpKo6p6reDPwaeP7AUy4C9mv3OQ04GNgtyUzdrwQ+VlWHVtVZVXUgcNJgTVV1fVUdU1XPBTYH3tG+/tntaOM+SYZHC2f6s2+Sk5OcfOkfbpiD/0KSJEndmuuRv78HDgGWAfeqqqdW1WeratnQfjsBGwKXtodrlyZZCjwAuMfAfsuq6pcD9y8CFtOMAALcDzhh6LmH7/9ZVV1VVR+pqt2AhwCbAf8FPGMl+x9WVUuqasmd7rhw5b2WJEmaEIvm+PkOA1YALwB+luQLwMeBb1XV4NDZAuD3wKNneY4rB25fP7StBh6/1pKsDzyFZnTxycDPaEYPj741zydJkjRp5nTkr6ouqqoDq+o+wO7AUuC/gQuSvDvJju2up9Icgr2xPeQ7+HPJWrzkz4GHD7Xd7H4aj0pyKM0JJx8EzgF2qqoHV9UhVXX52vdWkiRp8ozshI+qOrGqXgZsSXM4+N7ASUkeDXwT+AFwdJInJdk2yc5J3tZuX1OHAC9M8pIk90ryeuBhQ/vsBXwD2BR4LrB1Vb26qs5Yxy5KkiRNnLk+7HsL7Xy/zwGfS7IZcENVVZIn05yp+x80c+9+TxMIj1iL5z4yyd2BA2nmEH4JeA+w98Bu3wK2qKorb/kMkiRJ/ZLm5Futzk4PWr9++LWtui5jZBZnuk9oedLjn9N1CSOTG6b7TPQsvbbrEkaqrp3u/rF8RdcVjFQtX951CaOzYLovApYNb9N1CSP19UsPPaWqlsy2bbp/s5IkSboZw58kSVKPGP4kSZJ6xPAnSZLUI4Y/SZKkHjH8SZIk9YjhT5IkqUcMf5IkST1i+JMkSeoRw58kSVKPGP4kSZJ6xPAnSZLUI4Y/SZKkHjH8SZIk9YjhT5IkqUcMf5IkST1i+JMkSeoRw58kSVKPGP4kSZJ6xPAnSZLUI4u6LmBSnH3aRuyx1U5dl6Fb7RddFyBJGifXXNN1BZ1x5E+SJKlHDH+SJEk9YviTJEnqEcOfJElSjxj+JEmSesTwJ0mS1COGP0mSpB4x/EmSJPWI4U+SJKlHDH+SJEk9YviTJEnqEcOfJElSjxj+JEmSesTwJ0mS1COGP0mSpB4x/EmSJPWI4U+SJKlHDH+SJEk9YviTJEnqEcOfJElSjxj+JEmSesTwJ0mS1COGP0mSpB4x/EmSJPWI4U+SJKlHDH+SJEk9YviTJEnqkUVdFzDOkuwL7AuwARt2XI0kSdK6c+RvFarqsKpaUlVLFrN+1+VIkiStM8OfJElSjxj+JEmSesTwJ0mS1COGP0mSpB4x/EmSJPWI4U+SJKlHDH+SJEk9YviTJEnqEcOfJElSjxj+JEmSesTwJ0mS1COGP0mSpB4x/EmSJPWI4U+SJKlHDH+SJEk9YviTJEnqEcOfJElSjxj+JEmSesTwJ0mS1COGP0mSpB4x/EmSJPWI4U+SJKlHDH+SJEk9YviTJEnqEcOfJElSjxj+JEmSeiRV1XUNEyHJpcBv5/El/wK4bB5fb77Zv8k2zf2b5r6B/Zt09m9yzXff7lZVd5ptg+FvTCU5uaqWdF3HqNi/yTbN/ZvmvoH9m3T2b3KNU9887CtJktQjhj9JkqQeMfyNr8O6LmDE7N9km+b+TXPfwP5NOvs3ucamb875kyRJ6hFH/iRJknrE8CdJktQjhj9JkqQeMfxJkiT1iOFPkiSpR/4/0LycCJBJLH4AAAAASUVORK5CYII=\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Input: <start> the woman wearing a blue shirt posed with the university of florida mascot <end>\n",
      "Predicted translation: the rest of the day was the highlight of the night <end> \n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAoEAAAJCCAYAAAC/Lx2iAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy86wFpkAAAACXBIWXMAAAsTAAALEwEAmpwYAABD7UlEQVR4nO3deZhkdXn28e89PTPsuAKiUVE0EoyyiuKKGuNu8hqjSURFE4m8xvV1TeISDW7BqAkuEFncozExGk3URMEVMQiCCqKILIJsigoMMMPM8/5xTktNdc9MD3TXOTXn+7muvqr7nFNVT3VXV911fluqCkmSJA3Lsq4LkCRJ0uQZAiVJkgbIEChJkjRAhkBJkqQBMgRKkiQNkCFQkiRpgAyBkiRJA2QIlCRJGiBDoCRJ0gAZAiVJ2kxJduq6BunmMgRKkrT5Lkry8SSPTpKui5FuCkOgJEmb77HAauBfgQuTvD7J7h3XJG2WVFXXNUiSNJWS3BJ4KvBMYB/gS8AxwL9W1XUdliZtkiFQkqRFkOS5wFuBlcAvgKOBv62qq7usS9oQQ6AkSTdRkl2BZ9CcCbwD8HGaM4G3B14JXFFVv9NdhdKGGQIlSdpMSZ4IPAv4XeC7wHuBD1bVr0aO2RP4dlWt7KZKaeOWd12AJElT6DjgI8CBVfWtDRzzY+DwyZUkbR7PBEqStJmSbFtVq7quQ7o5nCJGkqTNd1WSncc3JrlNkrVdFCRtLkOgJEmbb0MTRG9FM3+g1Hv2CZQkaYGSvLj9toDnJBmd/mUGeBDw/YkXJt0E9gmUJGmBkvy4/fbOwE+A0abf1cB5wKur6uQJlyZtNkOgJEmbKckJwBOr6squa5FuKkOgtABJ1tE0/8znOuAc4Jiq+ofJVSVJ0k1nn8AOJLk7cBTwgqr6Ttf1aEH+Angt8AlgtpnnvsDvA28G7gi8KUlV1T92UaCkpZXkH4BXVtU17fcbVFXPn1BZ0k1mCOzGM4CDaGabf1G3pWiBHknz4n/MyLZjk3wTeEJV/V6Ss4HnAYZAact0L2DFyPcbYhObpoLNwROWJDQdh/8beDxw+6pyTqmea0cA7l1V54xtvxtwelVtl2R34DtVtW0nRUqStBmcJ3DyHgrsADwfuAF4TLflaIF+RtP0O+73gSva77cHfjmheiT1TJK7Jdm66zqkhbI5ePKeDny8qlYl+QhN0/B/dFxTryS5Nc16mw8Hdmbsw0pV7dhBWX8D/FOShwHfpGnuOYBm8fhnt8c8AvhSB7VJmrAkbwDOrqr3tS08n6d5zfplkkdX1Te6rVAbk+Rc4D5V9bOx7bcETq2qu3ZS2ITZHDxBSbYDfgo8tqq+kmRv4CSaJmGnGWgl+QSwD3A0cDFj/Wuq6n0d1XUgTZ+/PWhWCzgL+Adf7KXhSXI+8JSq+kaSxwDvAx4LPBW4d1U9tNMCtVHtjA+3q6rLxrbvAlxQVVt1U9lkeSZwsv4AuKKqvgJQVd9O8kPgj4B3d1pZvzwceETfJlutqpNoQrsk7UIzWTQ03Xo+VlXfTPJz4JTuytLGJHniyI+PTTLahWeG5v3nvIkWNaY9YfQHwCerakm7GBkCJ+tpwAfHtn2QpknYEHijy4CrN3lUB5LcnvmbqE/tpiJJHfkZN64a8rvAK9vty9nwusLq3sfbywKOGdu3hiYA/r9JFjSPJwPvBV4AHLmUd2Rz8IQkuSPwY+C3quqHI9t/g+ZJt2dV/aCj8nolyVNo/gmeUVW9CINJ9qEJ7LNNwaOqqmYmX5WkrrTzBP4e8AOa7it3bucP/CPgpVW1X6cFaqPa5f/uU1VXbPLgCUtyIs3JhlVVtf+S3pchUH2T5DvAbjSn5s+n+XT2a1V17w5q+l+aT/6vY/5+iudPuiZJ3UmynGaWhzsDx1fVae32FwFXVdV7u6xP0ynJbjQfLA4AvgHsW1VnLtX92Rw8QUnuBFxY8yTvJHeqqgs6KKuPPr7pQyZuT2Afz9ZKSrKCZgaDd45/AKyqt3VTVT+1sz08CrgTsHJ0X1W9rpOiWkkeC7yc5vW9gDOBN1fVf3ZY1tOAr7RjBv6TprvYy5fqzjwTOEFJ1gK7zjMa6TbAZTYp9leSbwAvq6ovd12LpO61E8j/dlWd13UtfZXkfsBngOuBnYCLgF3bn8/rolVnpLY/A94FfAj4arv5QcAfA4dV1bEd1fVD4PCqOr4dxPIPwB3nO3m0KPdnCJycdkj6LlV1+dj2OwNnVtV23VSmTWnnB3wD8NfAd5jbRP3zLuqS1I0k/wp8pquwMA2SfAU4jWaAw6+AvYBrgI8Ax1TVhzqs7YfAO6rqyLHtzwOeV1W/2UFN96eZb3KXtn/pSuASmqmI/ntJ7tMQuPRGFhp/LnAcsGpk9wxN2//qqnrApGvro/aJ/1c0n8juxI1rdQLQxRnTNsD/uoTRXTgwRBqcJP8XeDXwz8C3aMLNr1XVv3VRV5+006/cp6p+kOQXwIFVdVaS+wAfrqq7d1jb9cA9N7AU6Pe6mCcwyVHA9lX11JFt7wF2GN22mOwTOBmzC40H+C1g9ci+1cCpwBGTLqrHXg88BXgj8DbgpTQDRf4IeFVHNTnxq2629sPEgj55+8Gi92bPID1/nn1F8wF/6Ebf6y6lGURzFs0UYLfvpKIbXUCzytM5Y9t/l2ZA4kQl2YpmVow/Htv1QeBzSbZfitkyDIETUFUPbZcV+hjwrKq6quuaeu7JwHOq6rNJjqCZMPNHSc6i+ac9atIFVZXLwWkxPJkbQ+AuNKPNP8GNk5AfSLMe9WsmXpk2S1Ut2/RRg3cqcB+a0a4nAn/brshxMHBGh3VBc+LlH5PsC3yd5v/ygTQDM57XQT070DSbf350Y1V9Ncmf06xNv+gh0ObgCUkyA1wH7LWUw723BElWAXtU1QVJfgo8rqq+leQuwOmTWju4fXH4dlWta7/fICeL1uZK8ingP6rqn8a2Pxv4/ap6bDeVSYsjyf40TZknJNkJeD/wAJpQ+Myq+k7H9f0fmomhf6vddBbwd1X1ye6qmixD4AQlOQd4UlV9u+ta+izJ94FD2jU5vwL8V1W9IcmfAG+rql0mVMev15YcacabbyUA+wRqs7WjS/feQJ+k0x0o1m9t685hNH2970IzUvjcJK8Azq2qj3VaoLQANgdP1uuBNyU5uI+zlPfIJ2jWb/wG8A7gI+3ZkTsAfzfBOu4CXD7yvbSYrgCeBLxpbPuTuPF5p/56AfAy4M2s/ze8CPgLmu4/6qkkD4G5XX3a7TWp6cDalUsW2k/4rot+/54JnJx2JYy70Ix2/QlzR5N1NmdSn7VzTd0f+EFVfbqD+9/gxLDSTZXk6TSzBfwPN/YJvB/wO8CfVtX7uqpNm9a2WPy/qvpMkqtouvqcm+SewJer6jYdl9iJrkPNQiU5FXhdVf372PbHA6+d1LJ/SUbXKd4eeDHwTdbvJ3wA8NalmFzbM4GT1ceVMHqvqr5Bc1awq/tf004H8a6uatCWp6ren+RsmtGlT6DpanAm8ICqOrnT4rQQdwa+O8/2NcA2E66lT0bn3dtoqJlwXePuAZw+z/bvtPsmoqp+/XtIcjzNiiVvGD0mySuBey7F/RsCJ6iq/qbrGqZFktvRnP3bGVhvFF5VdRHGPgc8DHBiWC2aNuwtyfxfWnLnAvsydzqRx9CE+UHqOtRshmtppqn58dj232D9qW0m6Yk0z6lx/wK8cinu0BCo3klyMPBemjMjV7J+00LRzRm5LwBvSHJvnBh26mxo3e62c/8du1q3u50u42nAXYFXV9UVSR4AXFxV429O6pcjgCOTbEvzWnVgkqfR9BN8VqeV9cfEQ81m+BxNH/0nVNWV8Ot1jt/Q7uvCNcBBzJ278CDWX2Ri0RgCJ6iPK2H01OHAW2j6a9zQdTEtJ4adbj+mWbP0srHtt273dbEKzX40Hy5+THNW5AiawSKPAH4T+JNJ16SFq6rjkiynCQ3bAh+gGRTy/Kr6aKfF9cfEQ81meAnwZeC8JLNzFt6b5jXijzqq6W3AO9updWa7QN0PeAbw2qW4QweGTFCSN7P+Shh/zchKGFU18UmQ+yjJlcB+VXVu17Voy9DHdbuTnEAzgOA1YwMLDgT+uaruPOmadNMkuS2wrKrGP2QMWpKX0cyKcRzzhJqqenNXtQG0Z3GfCuxNczb3VJrl7DoLqEmeTDPyfHTuwncs1ZRDhsAJakdNHdauhHEVzRxhP0pyGPDwqnpSxyX2QpIjgbOr6h+7rkU3TZJ7AX8O7E6zSs5Pk/w+cH5VnTbBOnq7bneSX9G8Bpw7FgJ3A75fVVtPuqa2rqfQTNE0X3/cJ3RRUx8leRvwASeK37hJhxptHpuDJ2sXbuwwfDVwy/b7z9LMNaXGi4F/T/JwmpFaa0Z3LsUw+YVo+4s8iqYpf2UfauqjJL8LfAr4L5rBNLMjJXcHDqFZFm1S+rxu97XArebZvgdzm60nIsnfAS8ETgAuZoFTfQzUfYEXtCO8P0BzBum8bkvqnzbs9S7wteH0F1X1+fbnVwOHAt+jWazgpx3Xd0vmfgj7+aLfj2cCJ6cvK2H0XZLn0UwSfQXNm+F6A0O6mE+xnavwM8D1wE40fX92bX8+zzkeb5TkZOB9VfWusTNc+9EskzbRheP7um53kqOB2wF/SPNcvzfNc/2TwBer6kUd1HQp8NyqcjqrBWiXsnwqTf/Ne9CsQftB4GOzgw3UT0nOBF5YVZ8fWT/41TQf9C+pqon3yW27p7wHeCjrjxkIS7QylSFwgpK8Ebi6qg5P8iTgIzSTRt+BZr3Cv+q0wJ5Ichnwxqp6W9e1zGpD+2k0zRq/Avai6fT8EeCYqvpQh+X1Srsc2m9X1XljIfAuwFmTbubs67rdSXYE/pMm/G0HXELTWvA14DFVdc1Grr5UNV0OHDi+lJ02rQ0Sf0LTx/s2VTXIuQLbbg53bUe6X8VGzibXhNaBn0+Sa4A9q+r8JK8H7l5Vf5Rkb+BzXZyUSfJFmhbCI5jnTPz46iaLwebgCaqqV458//EkF9Iupt3FShg9NkPTnNgn96ZZxaGSrAW2aoPNy4EPA4bAG11J88HmvLHt+9J86Jmoqlqb5HzGmvC7VlW/Ah6Y5GE0v5tlwKlV9T8dlnU0cDBLNBJxC7cC2Irmeba241q69Dxg9oz7X3RZyCZcB+zQfv9wbpwD9pcj2yftAOB+VTXfJORLwhA4QUkeDHx9dtqTdqLYk5MsT/LgSa1VOAWOo2li6VM/u9G+ZJfSrBZwFk3fzok2b06BDwN/1/a5KWB5ux7nETR/2y70dt3uqvoi8EX49RKFEzUyeAaaIPrUJI8AzmBuf9z5pkgarCS/yY3NwbvR9KV8CfCvHZbVqdnlDtvpcy4HTq6qn3Vb1by+Arw1yVeB/WnW7IZmeqYLO6rpxzQfJCbG5uAJas8g7To+jUCS2wCXdTVPYDu9we7At6vq+i5qGKvnXTQvqt+jJ29EST4HvL+qPpTkKGA/4B9pzppsX1UHTrqmvmqDzPE0zWIB1rWXH6bpEzvxsyR9XLc7yfOBi6rqX9ufj6GZOuNHwBOq6uwJ1XHCQo+tqocuZS0b08PXqVOAfWiWHvsgzcCQS7qtql+SXAfs0ccBM0l+A3g3zUC/d1TVse32t9NM99PF+8zDgFcA/3dSXTIMgRO0kbnKfhM4ZdL9I5LsABxD8wmoaPpEnJvkPTQdY187yXpG6trYm1JV1cMmVkyrnbxzh6o6IclOwPtpm/KBZ1bVdyZdU98l2Z3mTXIZcFpV/bDDWl6zsf3VwZKOSc6hGazy5baV4DPAnwJ/AGxXVY+bdE191OPXqTfQTBFzVhf3Pw3aQWJ/1XEXh6nR9qHciqZL1PXAeoslLEVGMAROQJLZ/m2PBf6H5o87awb4bZoO84+acF3vohng8Fzgq8C92xfXxwGHV9Vek6xHGpIk1wL3qKoL2qlZblNVz0ryW8BXquq2HdR0LPCC8VHUSbYD/rGqJr4cmq9T0yvJo4E3Aa9h/uU2F33Kk2mW5Bkb2z/b1L6Y7BM4GbP9IWbXwr12ZN9qmhe2f5p0UcATgP9TVd9OMvpp4CyatUw7lWRr4G40n/5/VFXXdVzS7BnB3YFPV9U17Zvj9dWf5e06N9bHbA77lf3ar2imG7qAZqm4v2u3rwE6mSiapjn6FdzYsX/WNsDT6WZN3N68TrXP7Ve2//s+zzftM+3lv7H+SNfQ8XKb6eEyrksR8jbFEDgBVfVMgCTnAUd0MfXDBtyKGwPqqB3ocHRb26fsDTQjy1bSvGBcn+QfaZoW1mzs+ktU0y40I5bvQ9skBZwL/D3NKLMXTLqmHrvX2M8raCZAXk4zOfNETMFUFZ8H/inJaTQfdv6r3X5Pmg7iE9NOhJ7261ZJRj/UzNC0Ylw6yZpG9Ol16l7cGBbGn+ejbGJrdNaHdAFez/rLuL6UkWVcuyqqfa95Gs3Jhle1r18PAC6uqkV/XTAETtbrR39IcjvgcTRrl369g3r+l+ZT9tvbn2dfuP6cZuLMrryZ5tPZc2jOkgI8iOafdRnN6LtJexvNPG63oTlzM+tfaAaIqDXf4IH2rO4xNCPyJqXvU1U8Fzic5izEk0aaxvalmX9ykq6g+f8vblzVaFTRNOl1oTevU6PP7S4HyUyLpZjXbhE9GXhONcu4HgF8spplXM+iOTN/1KQLaifU/wLNh8B70rQOXNHW85s0AyYX9z7tEzg5Sf4L+GxVvSPJ9sD3aSaJ3Z5mDrr3T7ie+wOfA/6ZZpTre2meeAcAD66O1sRMcglNh/n/HNv+WOC9VbVrBzVdSrO+83fnmQD5u1W13aRrmjZJ9qSZhPWOXdei9bVT+IRmqpo/AEb7aq2mWfP54o5q6+XrlBamPbP1XGBPmgD/PeDdVdXVmeXZulbRjFy+IMlPgcdV1bfa1/TTu2gdaAdFfrmqXjP2PnMg8M9VdefFvk/PBE7WfsDL2u+fSNMnaHbZoZfQjDidmKr6evsC+xKaaSkeTtNcd2DHo11v0dYz7kfcuN7ypG3D+nMFztqJpjlYm7YTzQeeTmVCa3IusJataP7/R98gPzLpKVBmz9i0b4AXVI/ODvT4dYokT2nr2Zm5z6kndFJUj7TNmJ+l6UpwUrv5YODFSR5ZVSdt8MpL7wKaOV4vAM4BHkkzeOVA1u+3P0n70cwQMO6nNKsJLTpD4GTtAPyi/f53gU9U1Zp2qZh3dlFQ+yK60RFJHTgdeD7Np8dRLwC+PfFqGl8GDgH+sv252uXIXk5z+l6tJC8e30SzzvJTaZZJm7hsYk1OOuig3p4Z/SywIzAbZp4N/E2SR01q6pE0y519u6rW0XR3uE2SeY/t6qxbH1+n2hHdL6SZIHrOEl8CmgniP0LT7LoOIMkymv/FtwL377C2T9AE+G/QrFX/kSTPpl3GtaOarqXpAztuD+CyebbfbDYHT1CSs2n61fwHzZJaf1hVJ6ZZq/C/q2qnjuq6PfN/ku2qOfjBNGHhYppPj0Xz6ez2wKOr6qsbufpS1bQn8CWaEPoQ4NM0TVK3AB5QVfOduRykJOOdl9fRrBzwRZo1ocdHnk6ipomvybmAmv4bWAU8rZol5GbXE/4gzbKEj5xQHeuA21XVZe33RROOx1VXE9pDL1+nLgWeW1Uf7+L+R+r4IvDEqvpFkqcDH+3DZNrw62mQ9h6f+DzJHjRzh/ZmfeUk96XjZVyTHA3cDvhDmr6A96b5f/wk8MWqetGi36chcHKS/DlwJM1SY+cD+1bVunblgN+f9CTISfahecPZg7kv+n14wf+/wG/R1HYm8K6u+iW1Nd0OOIzmlP0ymiapd1bVT7uqSQuT5GomvCbnprR9ku5TVd8b234v4BuT6mfaniW9oKqq/X6Dqur8SdQ0qq+vU0kup2mSnsjKDhup43rgLlV1cTawKlVX2v7dh1TVZ8e2Pxo4tov+3WN13I7mbOT4h4uqqnd3UM+ONCdA7k0zXuASmmbgr9OcAFn0mUVsDp6gqjoqzVJDd6I587eu3fUjuhmSfjTNGonPpkfNGWmWaDuBZo6p11QHy4yNG6nps8Dr+lCTNsvE1+RcgOuYv4/rLZhgP9OxYHc0zfP8ROB/e/I87+XrFE1dBwOv7biO7wNvaAcVBHhyOz3SHJMefEgzmOeYJC+jCTIFPJBmAulJj4BfT5LZQUaz8/eOPq+KZkm5iWpbBB6YZvm4fWlPNtQSrrjimcAJSXILmpnu50yR0XaePbOqrpxwTdcA+1TVDyZ5v5uS5HCaJtf70AzG+DrNm9KJwDe7eGPqY019sqmJc0d1MYluOliTc1OSvI/m+fRsmn5J0HR7OIrmOfXMDmrq3fO8x69T76SZsuNMOlzjvB008w6auSZ3pOlXNt8be016xGs7IfPf0Uz3NXvSaQ1NwHp5Vc032G5StZ0PvI/mQ33nk/13lREMgROSZv3LnwKPrKqvjWzfGzgZuENVXTHhmr4BvKyqvjzJ+12oJNvQ9NE4qP06ALiuo4l9e1tTH2Tj6z2Pqkl1e5hnguitmeCanJvSjlJ+H/B4bpz0eIam/88hVfXLSdc0q0/P876+Tm3qOV8dzCM42r9z0ve9MUm2pZn8OMA5VbWq45JIciWwX1Wd23Ut0F1GsDl4QqrqqiSfpFl66Wsjuw6mmTttIgGwXRlg1l8Cb0ny1zSjE8c/yXa9ruOONKMVd6Lps7GWZgh/l3pVU5LlNG/Qd6JZXeXXJtn0s6E3vHY+TKrq6knVMqKPE0T/WlX9Avi9JHej6fsKzaf9Pgwy6vR5Pg2vU12EvAW4CzCT5HWsP+3Qu7oMhm3o63Q6n3l8iGYlnF5M9t9VRvBM4AQleSRNP4hd2qlhlgE/Af6iqv5tQjXMjv779ab2cs66jh12uH4nzVQedwa+STMq90TgpK5GvfW0pj1oRprfheZvtpbmg90amvWMuzxj+kLgxTTTLUDTl+vvgbd3MQddO7p77ewoxSSPoJly5EzgzV015/fw99SL5/mUvE59aiO7q6p+b2LFtDYwL9+BNEF+IvPybeL3sp7qcC7Ftqn632m6Pcz34eJ1HdQ08YzgmcDJmp0S4vE0C2o/nObszX9MsIbRT6+70XS4Hn8DXEZzZqkrh9FMKfImmvVUv9XFG+KYPtb0dpqzM3vTjCLbm2ZQwbuBv+6qqCRvAQ6l6Qs0+kb0apr5Al+2gasupWNo+k2dneQ3aF78v0QzAn0H4JWTLqinv6e+PM+n4XVqfD3jFcBewB1pXt+70Id5+e5F05e0s/5+C/TnwKNopmK5G3MHhkw8BNJBRvBM4IQleTNwj6r6/STvB66qqvFJkSdVy7zTCSS5DXBZh5+w78aNfZEeQrPKxFdpRy12MS9YT2v6GfCQapay+yVwQFWdnWYJsH+sqntPuqa2rp8Dh9bY/GlJngQcVVW36aCmX9D8fn6Q5EXAE6rqoUkeChxXVbt1UFMff099fJ738nVqQ5K8leZ1/bUd3Hfn8/KNzTt5Ls00SOOBuXNJLqOZt/RtXdcyatIZwTOBk/d+4FtJ7gj8H5qk35XZ1RLGbU+HS6G1ozfPoRm+T5Lfojkr8maaT/8Tf9HvY000f7/ZDtaX0zQpnk3TfHC3DuoZdcYGti2bZ/skzHDjmYmHc+PKJT9iiZZjWqBe/Z56/Dzv3evURhxFE5xf28F9/5Kme8jZY9vvwo2rVS21n7f3dxnNWdyu/uc3ZQZYcNP1BE00IxgCJ6yqvpfkO8CHgZ9U1TcnXcPIdB4FvLGdtHbWDM1Ag29Puq5ZbfPF/jRNQgfRjFLcmqbpc6GjULf4moDv0jQ/nUvTf+vl7VmTZ9O8kXfl/TRL/r1gbPthwAcmXw7Q/K4OS/JpmhfV2ebfO9A0B3Whd7+nPj3P+/46tRH36PC++zAv378CX04yO6fjKe3r0hxVddcJ1TSf42iWsuyi2XeDJp0RDIHd+ABNf66/6uj+79VehmZU4mjfjdU0K2EcMemiRvyCZmLf02g6pb8D+EotwWzpU17T4TSzykPTB/DTNG/UVwBPnmQhY/MELgcObjs5z85/d1+aZf8+NMm6Rrycph/gS4D3VbMWLcATaAJ0F7YC/mRDv6fR3+kE51b8Bf15nvf6dWqeuTFn18h+NHDs5CsCmrO2ae9/fF6+V0yohufQnGG7O80gp+OAiS8VuQDbAn/W/v91Ns/jBkwsI9gnsAPt9AfPo+n3c0mHdRwHvKDadUv7Ismj6D5graePNc2nfW5dOenO/H2cJ3Bckhlgx9EJV5PsBqzqYvqMPv7O+vg87/Hr1Pjfb3SN7GOrwwmI+zIvX/u3e351sF74pmzi/6+z1ymYbEYwBEqSJA1QXztsSpIkaQkZAiVJkgbIENihJId2XcM4a1oYa1q4PtZlTQtjTQvXx7qsaWGGXJMhsFu9e+JhTQtlTQvXx7qsaWGsaeH6WJc1LcxgazIESpIkDZCjgzfTymxVW/96arabZw3Xs4KtFuW2FsuWXlOWL87UmKvXXcvKZYu0AlMW52YWtaZFtGh1ZfE+s65et4qVy7a9+TfULM+6KBb3ObU4T6rVa69l5UzfalrFyplF+NsBLFvE59QN17By+c1/b6hF+j0BrLnhGlYsQk2L9RoFsGbNNaxYsTjvoYtlS6/pqmsuvqKqdppvn5NFb6at2Y77psuV3qbEIr6QLaaZW9+26xLmyIwn5Bdkm627rmCu667vuoL5rVzRdQVzzfRqiV8A1m3fv+dUbdW/v9265b5GTbMvfP1V529on39ZSZKkATIESpIkDZAhUJIkaYAMgZIkSQNkCJQkSRogQ6AkSdIAGQIlSZIGyBAoSZI0QIZASZKkATIESpIkDZAhUJIkaYC2qBCY5KAklaR/C8RKkiT1yFSHwCQnJjmy6zokSZKmzVSHQEmSJN00UxsCkxwPPAR4btsEXMBu7e69kpycZFWSU5LsO3bd+yf5Urv/oiTvTrLjZB+BJElSd6Y2BAIvAE4CjgN2bb8ubPe9EXgFsC/wM+BDSQKQ5F7A54FPAXsBTwT2Bo6dYO2SJEmdWt51ATdVVf0yyWpgVVVdApBkj3b3q6rqhHbb64CvAncAfgK8FPhoVb119raSHAaclmTnqrps/L6SHAocCrA12y7ho5IkSZqMqQ2Bm3DGyPcXt5c704TA/YC7JXnKyDFpL3cH5oTAqjoaOBpgx9y6Fr1aSZKkCdtSQ+Cake9nQ9uykcv3Am+b53oXLWVRkiRJfTHtIXA1MLOZ1zkVuGdVnbME9UiSJE2FaR4YAnAecECS3doJohfyeN7cXuc9SfZJcrckj0ty1JJWKkmS1CPTHgKPoDkbeCZwOXCnTV2hqs4AHkwzncyXgNNpRhNfumRVSpIk9cxUNwdX1Q+AA8c2Hz92zHncOPBjdtspwKOWsjZJkqQ+m/YzgZIkSboJDIGSJEkDZAiUJEkaIEOgJEnSABkCJUmSBsgQKEmSNECGQEmSpAEyBEqSJA2QIVCSJGmADIGSJEkDZAiUJEkaoKleO7gzyaaPmaDMzHRdwhzLtt+u6xLm9f233rHrEua49Ze36rqEObb5+bquS5hjx/+9qOsS5qjVq7suYX7XXtt1BXOlf+cc8qurui5hjtxwQ9clzNG/v1xreQ8jTA/fjzemt39bSZIkLR1DoCRJ0gAZAiVJkgbIEChJkjRAhkBJkqQBMgRKkiQNkCFQkiRpgAyBkiRJA2QIlCRJGiBDoCRJ0gAZAiVJkgbIEChJkjRAhkBJkqQBMgRKkiQN0KBDYJKDklSS23ZdiyRJ0iRNdQhMsrLrGiRJkqbRVIXAJCcmeXeSI5JcDnwtyZ5JPpPkqiSXJflIktuNXOdeSb6Q5FftMacneWiS3YAT2sMub88IHt/Bw5IkSZq4qQqBrYOBAA8Cng98GfgucADwO8D2wKeSzD62DwM/bffvA7wWuA64EPiD9ph7ArsCL5jII5AkSerY8q4LuAl+XFX/DyDJ64DTq+rlszuTPB34ObA/8E3gzsARVfX99pBzRo79efvtZVV1xYbuMMmhwKEAW7PtIj4USZKkbkzjmcBvjXy/H/DgJFfPftGc4QPYvb38e+C9Sb6Y5K+S7LG5d1hVR1fV/lW1/wq2unnVS5Ik9cA0hsBrRr5fBnwG2Hvs6+7ApwGq6rXAnsC/A/cHzkjyrMmUKkmS1E/T2Bw86lTgycD5VbVmQwdV1Q+BHwL/kOTdwJ8BxwKr20NmlrpQSZKkPpnGM4Gj3gncAvhokvsmuWuS30lydJIdkmyT5J3tfIC7Jbkv8EDgzPb65wMFPDbJTkm27+hxSJIkTdRUh8Cquhh4ALAO+CzwPZpgeH37tRa4FfA+4GzgE8BJwIvb618EvAY4HLgUOHKyj0CSJKkbU9UcXFUHzbPth8CTNnK1P9nEbb4eeP3Nq0ySJGm6TPWZQEmSJN00hkBJkqQBMgRKkiQNkCFQkiRpgAyBkiRJA2QIlCRJGiBDoCRJ0gAZAiVJkgbIEChJkjRAhkBJkqQBMgRKkiQN0FStHdwLgczMdF3Femrt2q5LmOPqg+7RdQnz2mqba7suYY6dvrmq6xLmWPbzq7ouYa7l/fq/A2DNDV1XML+VK7quYK6167quYI51v+zh83xZuq5gjqR/NQFUXdd1CXP09Xe1IZ4JlCRJGiBDoCRJ0gAZAiVJkgbIEChJkjRAhkBJkqQBMgRKkiQNkCFQkiRpgAyBkiRJA2QIlCRJGiBDoCRJ0gAZAiVJkgbIEChJkjRAhkBJkqQBMgQCSR6Q5Iwkq5Oc2HU9kiRJS2151wX0xDuA04HHAtd0XIskSdKS80xg427AF6vqwqr6edfFSJIkLbVBhMAkWyV5e5JLk1yX5BtJHphktyQF3AI4NkklOaTjciVJkpbcIEIg8BbgKcCzgH2A7wCfBdYAuwKrgBe233+0mxIlSZImZ4sPgUm2Aw4DXl5Vn6mqs4DnAJcCh1XVJUABv6yqS6rq2g7LlSRJmogtPgQCuwMrgK/NbqiqtcBJwJ4LuYEkhyY5Jckpa+r6palSkiRpgoYQAtNe1jz75ts296Cqo6tq/6raf0W2WrzKJEmSOjKEEHgOsBp44OyGJDPAgcCZXRUlSZLUpS1+nsCquibJu4E3JbkC+DHwImAX4F2dFidJktSRLT4Etl7eXh4H3BI4DXhUVf20s4okSZI6NIgQWFXX00wB88IN7N9+kvVIkiR1bQh9AiVJkjTGEChJkjRAhkBJkqQBMgRKkiQNkCFQkiRpgAyBkiRJA2QIlCRJGiBDoCRJ0gAZAiVJkgbIEChJkjRAhkBJkqQBMgRKkiQN0PKuC5g2N9x2Oy77w/t0XcZ6dv33H3ddwhwrf3FD1yXM6/qLt+u6hDlu2PG6rkuYY+WVV3ddwhzrfnZl1yXMUddd33UJ87u+p3X1zbJ0XYFuhizvX4TJiv7VxLUb3uWZQEmSpAEyBEqSJA2QIVCSJGmADIGSJEkDZAiUJEkaIEOgJEnSABkCJUmSBsgQKEmSNECGQEmSpAEyBEqSJA2QIVCSJGmADIGSJEkDZAiUJEkaoC0qBCY5KEkluW3XtUiSJPXZVIfAJCcmObLrOiRJkqbNVIdASZIk3TRTGwKTHA88BHhu2wRcwG7t7r2SnJxkVZJTkuw7dt37J/lSu/+iJO9OsuNkH4EkSVJ3pjYEAi8ATgKOA3Ztvy5s970ReAWwL/Az4ENJApDkXsDngU8BewFPBPYGjp1g7ZIkSZ1a3nUBN1VV/TLJamBVVV0CkGSPdverquqEdtvrgK8CdwB+ArwU+GhVvXX2tpIcBpyWZOequmz8vpIcChwKsGL7Wy3ho5IkSZqMaT4TuDFnjHx/cXu5c3u5H3Bwkqtnv4Cvtft2n+/Gquroqtq/qvZfvs12S1OxJEnSBE3tmcBNWDPyfbWXy0Yu3wu8bZ7rXbSURUmSJPXFtIfA1cDMZl7nVOCeVXXOEtQjSZI0Faa9Ofg84IAku7UTRC/k8by5vc57kuyT5G5JHpfkqCWtVJIkqUemPQQeQXM28EzgcuBOm7pCVZ0BPJhmOpkvAafTjCa+dMmqlCRJ6pmpbg6uqh8AB45tPn7smPOAjG07BXjUUtYmSZLUZ9N+JlCSJEk3gSFQkiRpgAyBkiRJA2QIlCRJGiBDoCRJ0gAZAiVJkgbIEChJkjRAhkBJkqQBMgRKkiQNkCFQkiRpgAyBkiRJAzTVawd3Yfm167jNd6/ruoz11HbbdF3CHOc9dquuS5jX1ldk0wdN2IrvX9h1CXOsu3511yXMkZn+fWatrgvYgCzv4Uv7sv79/Vi3rusK5siKHv7t+vh8ArJiRdclzDUz03UFc1254V09/K+UJEnSUjMESpIkDZAhUJIkaYAMgZIkSQNkCJQkSRogQ6AkSdIAGQIlSZIGyBAoSZI0QIZASZKkATIESpIkDZAhUJIkaYAMgZIkSQNkCJQkSRqgLSYEJvl0kuO7rkOSJGkabDEhUJIkSQtnCJQkSRqgqQyBSbZNcnySq5NcmuQvx/YfnOR/k1yV5LIk/5LkDu2+JDknyUvGrnP3JJVk30k+FkmSpC5MZQgEjgAeAfwB8HBgH+DBI/tXAq8B9gIeB9wW+AhAVRVwDPCssdt8FvDtqjp1SSuXJEnqgakLgUm2B/4UeFlVfa6qvgs8E1g3e0xVHVtV/1lV51bVN4HDgAcl+Y32kOOAuye5X3ubM8DTacLhfPd5aJJTkpyyes01S/fgJEmSJmTqQiCwO82ZvpNmN1TV1cB3Zn9Osm+STyY5P8lVwCntrju1x18CfJobzwY+CrgN8KH57rCqjq6q/atq/5UrtlvsxyNJkjRx0xgCs9GdyXbA54BVwNOA+9CEPGjC46z3Ak9Jsi1NGPy3qrpy8cuVJEnqn2kMgecAa4D7zW5og99vtz/uQdMH8C+r6stV9X1g53lu57PAr4DnAI8Hjl3KoiVJkvpk6kJg2/R7DPDmJI9Ick+aADfTHnIBcD3wF0numuSxwOvnuZ217fXeCFwEfGES9UuSJPXB1IXA1kuAE4BPtJffBb4MUFWXA88Afh84k2aU8Is3cDvH0jQRH9eOGpYkSRqE5V0XcFNU1TU0o3mfvoH9HwU+OrZ5vr6EtwPWAscvZn2SJEl9N5Uh8OZKshVwR+BvgU9U1QUdlyRJkjRR09ocfHP9MXA2zbQwG2oqliRJ2mINMgRW1fFVNVNV+1bVhV3XI0mSNGmDDIGSJElDZwiUJEkaIEOgJEnSABkCJUmSBsgQKEmSNECGQEmSpAEyBEqSJA2QIVCSJGmABrls3M2R1Tew8oIrui5jPWtvvWPXJcyR6rqC+W1zaQ8Lu9Utuq5gjmXXXt91CVMha27ouoR5ZeXKrkuYq9Z1XcEcta6HNfXwOZWuC9iAWr2m6xLmWjFdscozgZIkSQNkCJQkSRogQ6AkSdIAGQIlSZIGyBAoSZI0QIZASZKkATIESpIkDZAhUJIkaYAMgZIkSQNkCJQkSRogQ6AkSdIAGQIlSZIGyBAoSZI0QIZASZKkATIESpIkDZAhUJIkaYCmJgQmeXSSq5Isb3++e5JK8u6RYw5P8t9JZpIck+THSa5N8sMkL0uybOTYeyX5QpJftbd7epKHdvHYJEmSJm151wVshq8AWwP7A98ADgKuAEaD20HAf9KE24uAJwOXAwcARwM/A45pj/0wcHq77wbgXsB1S/sQJEmS+mFqQmBVXZ3kVJrQNxsCjwRekWRX4JfAfYCXVdUa4NUjVz8vyb7AH3NjCLwzcERVfb/9+ZwN3XeSQ4FDAbae2WHRHpMkSVJXpqY5uHUiTfgDeAjwX8A3220PANa0P5PkOUlOSXJ5kquBFwF3Grmtvwfem+SLSf4qyR4butOqOrqq9q+q/VfObLPID0mSJGnypjEEPiDJnsAOwLfabQ+lCYJfr6o1SZ4CvB04HngksDfwLmDl7A1V1WuBPYF/B+4PnJHkWRN4DJIkSZ2bmubg1leArYCXAV+tqrVJTqTp73cZTX9AgAcCJ1fVkbNXTLL7+I1V1Q+BHwL/0A4w+TPg2CV9BJIkST0wVWcCq+pq4FTgYOCEdvNJwB2B+9KcFQT4AbBvO6L47kleRdN8DECSbZK8M8lBSXZLcl+a4HjmhB6KJElSp6YqBLZOAGZoA19VXUczUOR62v6AwFHAx2hGAP8vsBvw1pHbWAvcCngfcDbwCZow+eKlLl6SJKkPpq05mKp6BfCKsW0Hjf28GvjT9mvU60b2/8nSVSlJktRv03gmUJIkSTeTIVCSJGmADIGSJEkDZAiUJEkaIEOgJEnSABkCJUmSBsgQKEmSNECGQEmSpAEyBEqSJA2QIVCSJGmADIGSJEkDNHVrB3dt7XYr+cV979B1Geu5xXd/3nUJc9zm9Oq6hHmt3jFdlzDH2ltt13UJcyxfdV3XJcxR11/fdQlzVT+f53XDDV2XMEdmPOewEJmZ6boE3Rzr+vmasCH+V0qSJA2QIVCSJGmADIGSJEkDZAiUJEkaIEOgJEnSABkCJUmSBsgQKEmSNECGQEmSpAEyBEqSJA2QIVCSJGmADIGSJEkDZAiUJEkaIEOgJEnSAG1RITDJQUkqyW27rkWSJKnPpjoEJjkxyZFd1yFJkjRtpjoESpIk6aaZ2hCY5HjgIcBz2ybgAnZrd++V5OQkq5KckmTfseveP8mX2v0XJXl3kh0n+wgkSZK6M7UhEHgBcBJwHLBr+3Vhu++NwCuAfYGfAR9KEoAk9wI+D3wK2At4IrA3cOwEa5ckSerU8q4LuKmq6pdJVgOrquoSgCR7tLtfVVUntNteB3wVuAPwE+ClwEer6q2zt5XkMOC0JDtX1WXj95XkUOBQgJXb3moJH5UkSdJkTPOZwI05Y+T7i9vLndvL/YCDk1w9+wV8rd23+3w3VlVHV9X+VbX/iq22W5qKJUmSJmhqzwRuwpqR76u9XDZy+V7gbfNc76KlLEqSJKkvpj0ErgZmNvM6pwL3rKpzlqAeSZKkqTDtzcHnAQck2a2dIHohj+fN7XXek2SfJHdL8rgkRy1ppZIkST0y7SHwCJqzgWcClwN32tQVquoM4ME008l8CTidZjTxpUtWpSRJUs9MdXNwVf0AOHBs8/Fjx5wHZGzbKcCjlrI2SZKkPpv2M4GSJEm6CQyBkiRJA2QIlCRJGiBDoCRJ0gAZAiVJkgbIEChJkjRAhkBJkqQBMgRKkiQNkCFQkiRpgAyBkiRJA2QIlCRJGiBDoCRJ0gAt77qAabNme7jkwK6rWN/X3v6xrkuY49H3eFDXJcwrM/373LPu2uu6LmGOtV0XMJ911XUFc9W6riuY38xM1xXMlXRdwRxZubLrEuaotf3778syo8KC9fDvtzH9e0eUJEnSkjMESpIkDZAhUJIkaYAMgZIkSQNkCJQkSRogQ6AkSdIAGQIlSZIGyBAoSZI0QIZASZKkATIESpIkDZAhUJIkaYAMgZIkSQO0yRCY5MQkR97U/Ru4TiV50uYcs5DrjF3/oPY6t92c2iRJkoZg+SLcxhOBNYtwO5uyK3DlYt5gkkOAI6tq+8W8XUmSpL672SGwqn6+GIUs4H4umcT9SJIkDcFC+wQuS/KGJFckuSzJEUmWwdzm4CS7JPlUkmuTnJ/kmUm+m+S1Y7d56yT/kuSaJOcmOXhjBczTPHzfJKcmuS7JaUke0x5z0NhV90pycpJVSU5Jsm97/YOA44Dt2uvVPDVKkiRtkRYaAp8K3ADcH/gL4IXAUzZw7PuAOwMPA34POLj9edyrgU8CewEfBY5NMt9xcyTZHvg08H1gP+BlwN9t4PA3Aq8A9gV+BnwoSYCvt49jFU1T867AEQu5f0mSpGm30BB4ZlW9uqp+UFUfA04AHj5+UJJ7AI8E/ryqTqqqbwOHANvOc5sfqKoPVtU5wKtoQuaDFljPU4EZ4E+r6ntV9d/A4Rs49lVVdUJVfR94HbAHcIeqWg38EqiquqT9unqB9y9JkjTVFhoCzxj7+WJg53mO2wNYB5wyu6GqLmyP3+BtVtUNwOUbuM357AF8t6quHdl28gaOHa19to6F3g8ASQ5tm5JPWXv1NZtzVUmSpF5aaAgcH/1bG7huNuO+F3qb80l7/Obez+x1Nmt+xKo6uqr2r6r9Z7bfbnOuKkmS1EuLPVn0We1t7je7IclvALdfgvu5V5JtRrYdcBNuZzVNs7IkSdKgLGoIrKqzgc8B70lyvyR704zAXcXCz9wtxIeAtcA/Jdkzye8AfzlbxmbcznnA1kkekeS2SebruyhJkrTFWYpl4w4BfgKcCHyKJrBdBly3WHfQDuB4PHBP4DSakcGvbXcv+H6q6uvAe4CP0PRJfNli1ShJktRnm5wsuqoOmmfbIRva307q/PjZn9tl244Gzhk5Zk7fwarabeznbOLnbwD7jNzP79GcBfxRu/9ExvooVtV582w7DDhsvB5JkqQt2WIsG7eeJA8DdgC+QzMK93DgCuCzi3w/zwDOBS4Efht4O/AfVXXFYt6PJEnSlmjRQyCwAvhb4K40fQFPBh5cVYs9t8ouwN/QTPJ8CfAZ4OWLfB+SJElbpEUPgVX1OZrBIUuqqt4CvGWp70eSJGlLtBQDQyRJktRzhkBJkqQBMgRKkiQNkCFQkiRpgAyBkiRJA2QIlCRJGiBDoCRJ0gAZAiVJkgbIEChJkjRAS7Fs3BZt5TZruNNv/7TrMtZz35cf1nUJc9x2p379jvosl17edQlz1Oo1XZcwR2b695l13eq1XZcwr8x0XcF0WLdqVdclzDXTvz9eelhTb03Z76p/r6qSJElacoZASZKkATIESpIkDZAhUJIkaYAMgZIkSQNkCJQkSRogQ6AkSdIAGQIlSZIGyBAoSZI0QIZASZKkATIESpIkDZAhUJIkaYAMgZIkSQNkCASSPCDJGUlWJzmx63okSZKW2vKuC+iJdwCnA48Frum4FkmSpCXnmcDG3YAvVtWFVfXzrouRJElaaoMIgUm2SvL2JJcmuS7JN5I8MMluSQq4BXBskkpySMflSpIkLblBhEDgLcBTgGcB+wDfAT4LrAF2BVYBL2y//2g3JUqSJE3OFh8Ck2wHHAa8vKo+U1VnAc8BLgUOq6pLgAJ+WVWXVNW189zGoUlOSXLKml+smmj9kiRJS2GLD4HA7sAK4GuzG6pqLXASsOdCbqCqjq6q/atq/xW33HZpqpQkSZqgIYTAtJc1z775tkmSJG3xhhACzwFWAw+c3ZBkBjgQOLOroiRJkrq0xc8TWFXXJHk38KYkVwA/Bl4E7AK8q9PiJEmSOrLFh8DWy9vL44BbAqcBj6qqn3ZWkSRJUocGEQKr6nqaKWBeuIH920+yHkmSpK4NoU+gJEmSxhgCJUmSBsgQKEmSNECGQEmSpAEyBEqSJA2QIVCSJGmADIGSJEkDZAiUJEkaIEOgJEnSABkCJUmSBsgQKEmSNECDWDt4Ma2+dgXnn7lr12WsZ+Xd0nUJcyy74XZdlzCvHS64rusS5lhx/equS5ijrrq66xLmqLVruy5hjiy7oesS5pX07zWBZf0755CVK7suYSpUVdclzG/duq4rmKuvv6sN6N9/pSRJkpacIVCSJGmADIGSJEkDZAiUJEkaIEOgJEnSABkCJUmSBsgQKEmSNECGQEmSpAEyBEqSJA2QIVCSJGmADIGSJEkDZAiUJEkaIEOgJEnSAG1RITDJQUkqyW27rkWSJKnPpjoEJjkxyZFd1yFJkjRtpjoESpIk6aaZ2hCY5HjgIcBz2ybgAnZrd++V5OQkq5KckmTfseveP8mX2v0XJXl3kh0n+wgkSZK6M7UhEHgBcBJwHLBr+3Vhu++NwCuAfYGfAR9KEoAk9wI+D3wK2At4IrA3cOwEa5ckSerU8q4LuKmq6pdJVgOrquoSgCR7tLtfVVUntNteB3wVuAPwE+ClwEer6q2zt5XkMOC0JDtX1WXj95XkUOBQgJlb3WoJH5UkSdJkTPOZwI05Y+T7i9vLndvL/YCDk1w9+wV8rd23+3w3VlVHV9X+VbX/zPbbLU3FkiRJEzS1ZwI3Yc3I99VeLhu5fC/wtnmud9FSFiVJktQX0x4CVwMzm3mdU4F7VtU5S1CPJEnSVJj25uDzgAOS7NZOEL2Qx/Pm9jrvSbJPkrsleVySo5a0UkmSpB6Z9hB4BM3ZwDOBy4E7beoKVXUG8GCa6WS+BJxOM5r40iWrUpIkqWemujm4qn4AHDi2+fixY84DMrbtFOBRS1mbJElSn037mUBJkiTdBIZASZKkATIESpIkDZAhUJIkaYAMgZIkSQNkCJQkSRogQ6AkSdIAGQIlSZIGyBAoSZI0QIZASZKkATIESpIkDdBUrx3ciZmCW6zpuor1bPXjrbouYY5tLu/X72jW8st+1XUJc9RVV3ddwhx1ww1dlzBHrV3bdQlzzcx0XcH8lvn5fiGyckXXJUyH5T2NClVdVzDXsh6+Jmzkbc9XCkmSpAEyBEqSJA2QIVCSJGmADIGSJEkDZAiUJEkaIEOgJEnSABkCJUmSBsgQKEmSNECGQEmSpAEyBEqSJA2QIVCSJGmADIGSJEkDZAiUJEkaoKkJgUlOTHLkZl6nkjxpqWqSJEmaVsu7LmAzPBFYs5g3mGQ34MfAfarqlMW8bUmSpD6bmhBYVT/vugZJkqQtRW+ag9vm3ncleUOSK5JcluSIJMtG9h85cvwuST6V5Nok5yd5ZpLvJnnt2E3fOsm/JLkmyblJDh7Z9+P28n/bpuMTl/RBSpIk9URvQmDrqcANwP2BvwBeCDxlA8e+D7gz8DDg94CD25/HvRr4JLAX8FHg2CSzxx3QXj4K2JWmyVmSJGmL17cQeGZVvbqqflBVHwNOAB4+flCSewCPBP68qk6qqm8DhwDbznObH6iqD1bVOcCraELmg9p9l7eXP6uqSzbU5Jzk0CSnJDll7VXX3JzHJ0mS1At9C4FnjP18MbDzPMftAawDfj2Yo6oubI/f4G1W1Q00wW++29ygqjq6qvavqv1ndthuc64qSZLUS30LgeOjf4v5a8wS3KYkSdJgTGsYOoum9v1mNyT5DeD2m3k7q9vLmUWqS5IkaSpMZQisqrOBzwHvSXK/JHsDxwGraM70LdRlwLXAI9vRxrdY9GIlSZJ6aCpDYOsQ4CfAicCngA/RhLrrFnoDbR/B5wN/RtOf8JOLXaQkSVIf9Way6Ko6aJ5th2xof1VdAjx+9ucktwWOBs4ZOWZO38Gq2m3s5/cC772pdUuSJE2j3oTAzZXkYcAOwHdoRvseDlwBfLbLuiRJkqbB1IZAYAXwt8BdafoCngw8uKqcyE+SJGkTpjYEVtXnaAaHSJIkaTNN88AQSZIk3USGQEmSpAEyBEqSJA2QIVCSJGmADIGSJEkDZAiUJEkaIEOgJEnSABkCJUmSBmhqJ4vuyo7bXMsj9jyz6zLWc/Kp+3Rdwhxrt5npuoT5XXFl1xXMsW7Vqq5LmKPWVdclzJFlc5YC71wff08ArF3bdQVTodbc0HUJUyEzPT1ftKx/dWX5dMWq/v0GJUmStOQMgZIkSQNkCJQkSRogQ6AkSdIAGQIlSZIGyBAoSZI0QIZASZKkATIESpIkDZAhUJIkaYAMgZIkSQNkCJQkSRogQ6AkSdIAGQIlSZIGaIsNgUlekuS8ruuQJEnqoy02BEqSJGnDOgmBSXZMcssJ3+dOSbae5H1KkiT11cRCYJKZJI9M8mHgEmCvdvstkhyd5LIkVyX5UpL9R653SJKrkzw8yXeTXJPkhCR3Gbv9lyW5pD32/cD2YyU8Brikva8HLPHDlSRJ6rUlD4FJ7pnkLcAFwEeBa4BHAV9OEuAzwB2AxwH7AF8Gvphk15Gb2Qp4JfAs4EDglsB7Ru7jycDfAq8B9gXOBl48VsqHgD8BdgD+O8k5SV49HiYlSZKGYElCYJLbJHl+klOA04A9gBcCu1TVs6vqy1VVwEOBvYEnVdU3q+qcqnoVcC7wtJGbXA48tz3mDOAI4KFJZut/IfC+qjqqqn5QVYcD3xytqapuqKr/rKo/BnYB3tDe/w/bs4/PSjJ+9lCSJGmLtFRnAp8HvAO4Hrh7VT2hqv6lqq4fO24/YFvg8rYZ9+okVwO/Dew+ctz1VXX2yM8XAytozggC/BZw0thtj//8a1V1VVUdW1UPBe4D7AwcAzxpvuOTHJrklCSnXHfl+EOQJEmaPsuX6HaPBtYATwe+l+QTwAeAL1TV2pHjlgGXAg+a5zZ+NfL9DWP7auT6my3JVsBjac42Pgb4Hs3ZxE/Od3xVHU3zmNhpz9vUfMdIkiRNkyU5E1hVF1fV4VV1D+B3gKuBfwZ+kuStSfZpDz2Vpml2XdsUPPp12Wbc5VnA/ca2rfdzGg9MchTNwJQjgXOA/apq36p6R1VdufmPVpIkafos+cCQqvpGVR0G7ErTTPybwDeTPAj4H+BrwCeTPDrJXZIcmORv2v0L9Q7gGUmeneTuSV4J3HfsmIOBzwM7An8M3LGqXlpV372ZD1GSJGnqLFVz8Bxtf8CPAx9PsjOwtqoqyWNoRvb+E03fvEtpguH7N+O2P5rkrsDhNH0MPwX8PXDIyGFfAG5XVb+aewuSJEnDMrEQOGq0qbeqrgJe0H7Nd+zxwPFj204EMrbtjcAbx67+2pH9F9/0iiVJkrYsLhsnSZI0QIZASZKkATIESpIkDZAhUJIkaYAMgZIkSQNkCJQkSRogQ6AkSdIAGQIlSZIGyBAoSZI0QIZASZKkATIESpIkDZAhUJIkaYCWd13AtLn+rOK8A67tuoz17MLXuy5haqztugDdZLWu6wqmh78rLaZa03UFWiqeCZQkSRogQ6AkSdIAGQIlSZIGyBAoSZI0QIZASZKkATIESpIkDZAhUJIkaYAMgZIkSQNkCJQkSRogQ6AkSdIAGQIlSZIGyBAoSZI0QIZASZKkATIESpIkDZAhUJIkaYAMgZIkSQNkCJQkSRogQ6AkSdIALe+6gGmQ5FDgUICt2bbjaiRJkm4+zwQuQFUdXVX7V9X+K9iq63IkSZJuNkOgJEnSABkCJUmSBsgQKEmSNECGQEmSpAEyBEqSJA2QIVCSJGmADIGSJEkDZAiUJEkaIEOgJEnSABkCJUmSBsgQKEmSNECGQEmSpAEyBEqSJA2QIVCSJGmADIGSJEkDZAiUJEkaoFRV1zVMlSSXA+cv0s3dFrhikW5rsVjTwljTwvWxLmtaGGtauD7WZU0Ls6XXdOeq2mm+HYbADiU5par277qOUda0MNa0cH2sy5oWxpoWro91WdPCDLkmm4MlSZIGyBAoSZI0QIbAbh3ddQHzsKaFsaaF62Nd1rQw1rRwfazLmhZmsDXZJ1CSJGmAPBMoSZI0QIZASZKkATIESpIkDZAhUJIkaYAMgZIkSQP0/wEzLEEEo8P0bAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Input: <start> a beautifully ornate church on a beautiful day <end>\n",
      "Predicted translation: i also saw a castle that we saw beautiful flowers <end> \n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Input: <start> four men stand outside a run down building without noticing the medical assistance poster on the light post <end>\n",
      "Predicted translation: we went to the beach i m stuck in the woods <end> \n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Input: <start> two little girls play on a sandy hill <end>\n",
      "Predicted translation: we got to go to the beach <end> \n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Input: <start> three men stand on a patio with drinks in their hands <end>\n",
      "Predicted translation: the guys were all very happy <end> \n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Input: <start> a group of musicians that are getting ready for an event <end>\n",
      "Predicted translation: the band was obstensibly christian and the crows enters the band played songs <end> \n"
     ]
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAjwAAAJ/CAYAAAB8wwlYAAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjMuMCwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy86wFpkAAAACXBIWXMAAAsTAAALEwEAmpwYAABLS0lEQVR4nO3deZwkdX3/8dd7D0AuMSKIeOCNB3KtCp4YYzRq8jN4JCoe0YgajVe8FQ8M4oE3iYoKeKCiiTfxIAp4BNAVBRUVERABOUVZWNjz8/ujatie3tndWXamq6f29Xw85jHdVdVVn+7tnX73t771/aaqkCRJ6rN5XRcgSZI02ww8kiSp9ww8kiSp9ww8kiSp9ww8kiSp9ww8kiSp9ww8kiSp9ww8kiSp9ww8kiSp9ww8kiSp9ww80mYoyUOT3H/g/jOTfD/Jh5Ns22VtkjQbDDzS5um9wK0Bktwd+DBwFrA/8M7uypKk2WHgkTZPdwZ+1t5+PHBiVf0L8BzgbzurSpJmiYFH2jwVML+9/XDgG+3tS4FbdlKRJM0iA4+0efoRcEiSpwEPBr7eLt+NJvRIUq8YeKTN00uAvYAjgcOq6rft8icC/9dRTZI0a1JVXdcgaUwk2QpYVVUruq5FkmaSgUeSJPXegq4LkDR6Sf4COIymw/JODJ3erqrtu6hLkmaLgUfaPH0M2Bs4CriE5qotSeotT2lJm6Ek1wCPqKrTu65FkkbBq7SkzdPlwLVdFyFJo2LgkTZPrwMOdd4sSZsLT2lJm6EkP6MZZHA+8Dtg0mXoVXWfDsqSpFljp2Vp8/RfXRcgSaNkC88cluSuNLNcv7iqfrah7SVJ2lzZh2duewZwAPCsjuuQJGms2cIzRyUJcAFwIvC3wG2qalWnRWnOSLIFTcflJwO3BxYOrq+q+VM9TpLmKlt45q6HAdsBLwJWAo/uthzNMW+haSF8F7AaeAXwH8BVwL90WJckzQpbeOaoJMcCy6vq4CRHALtV1RM6LktzRJLzgedX1TeSLAH2qqrfJnk+8HDfS5L6xhaeOSjJNsCBwCfbRZ8CHpPkFt1VpTlmZ+Ds9va1wA7t7W8Af91FQVq3JKuTrFrHz3VJzkzyoq7rlMaZgWduejxwZVV9D6Cqfgr8BvjHLovSnHIhcJv29rnAI9vb+wPXd1KR1ueFNKcbPwo8p/35KHAlcAjwHeBtSf61swqlmyjJNkmenuTms3kcA8/c9DSaVp1Bn6LpkyFNxxdpZkoHeB/w5vY017E0H6QaL48EXlNVz62qo9uf5wKvBR5aVS8FXgY8r9MqpZvmScAxNJ9ts8Y+PHNMktsB5wP3qKrfDCy/Lc1VW/esqnM6Kk9zVJL9gAcA51TV17quR5MluZamn9W5Q8vvApxZVdskuTPws6raupMipZsoycnATsDSqlo0W8dxpOU5pqp+zxT/blV10VTLpemoqtOA07quQ+t0FfA44Iih5Y+jOa0FsC3w59GVJG26JLvRfNm6H3BakntW1dnrf9RN4wfkHJTk9sDva4rmuSS3r6oLOyhLYy7JgcBXq2pFe3udquoLIypL0/Nm4CNJ/hL4IVA0HxB/TdOfB+ARwCndlCfdZE8DvldVP03yPzRdM141GwfylNYclGQVsEtVXT60/JbA5Q4ap6kkWQ3cuqoub2+vS/keGj9J9gf+FdgdCPBL4P1t65w0JyX5DXBYVR3bfhF7P3C7qb7Qb/KxDDxzT/thtXNVXTG0/A7A2VW1TTeVSZI0PUkeAHyL5vPsunYE+EuBf6iqE2f6eJ7SmkOSvL+9WcDhSZYOrJ5P08T901HXJWk0ktyGpnPnpCtsq+qMbiqSNskzgC9X1XUAVbU8yeeAZ9JMmzSjvCx9btmj/Qlwj4H7ewB3Ac6geaNs9pI8Lsl3k1zZ/nwvyd93Xde4SHJYkrUuYU7yvCRv6aImrVuSvZP8Avg9zf/zxQM/P+qyNs0NSd6QZK0r+JLcLMkbOqhnS5rL0T85tOpTwOOSbDvjx/SU1tzSThr6OeBZVbWk63rGUZJ/A94KfAI4tV28P3AQcEhVDV/pstlJciHwxKo6fWj5fYH/qqo7dFOZppLkRzRXah0KXELTynujqvpdBzU9jqYTvJMWzwHj1vczyY40c0B+qqpWD607CPjfqrp0Ro9p4JlbkswHbgD2nK1L9+a6JH8A3lBVHxla/hzg0KrapZvKxkeSG2jGbDpvaPmdaPqBbdVNZZpKkuuAvcdpjK22piXAx4Gjq+rXHZek9VhP38+/Aj5TVbfqprLR8ZTWHNN+m/odsEXXtYyxbYGTplh+UrtOzdQSD55i+UOAi0ZcizbsZ8Ctuy5iyK2BNwIPBc5O8v0k/9TO9acxkWRJkmtoWgXPS3LNwM91wDdpzhr0ni08c1CSZwBPBg6qqis3tP3mJskngV9U1duGlr8KuE9VPbWbysZHe9rvdTTjXXynXfxw4HDg7VX1jq5q09ra8XfeCryeJvysGFxfVX/soq4JSe4JPBt4KrA1cDzwMS+Z7177eRHgaOAlTB6ccjlwQVWdOsVDZ6ue8xk6JbsuVXWnGT22gWfuSfIz4I7AQppv49cNrq+q+3RR17hI8kbgpcDprOnDs1/7826a2cEBqKp3j7zAMZHkcJo/gBOthcuB91XVqzsrSlMaGjdp8I92GJNxk9rpbQ4GXknzXroZTQfr51TVWV3WJkjyUOD/qmrFBjee3Tr+beDutjRzwP2Qyf0t7we8q6oOndFjG3jmnvYDfZ2q6s2jqmUctd8gpqNm+hvEXNOefrgnzQfn2VV17QYeog60H1brVFWdjLCcZCHw98CzaFoIT6eZfPZ44BbA24FFVXWPLurT2sZpaIMkx9LM3/fWoeWvAe5VVQfN6PEMPJKkjZXkAzSn1ovm0uKPDl9I0U6Dc0FV2V+0Y0n2prnke2Kk7kGdtBK2fYv2WcekuGdU1fYzeTwHHpQ2E0m+QtPv65r29jpV1d+NqCytQ5J9gJ9W1er29jp1NPDgPYEXAl+oquXr2OYS4GGjK0nrcRTNOE7PYYqhDTpyHXAAcO7Q8gOApcMbbyoDzxzUDr/9OppvV7en6ctzo3E4n9+lgRGpp1RVLxpVLWPmKtb8kbuqy0I0LYtproS6vL1drP3NnHb5yP/PV9XDp7HNSpzQdFzckzEb2gB4D/AfSRYBEx3c96MZgflNM30wA8/c9BbgH2iuqHkP8ApgN+AfgUO6K2ts7DF0fyFNM+4Cmk6Um6Wq+qepbmts3RG4YuB255I8fbrbVtUnZrMWbbSJoQ3GJvBU1TuSXAC8mGbUZWgmxX1GVc34pfL24ZmD2k65z6+qbyRZAuxVVb9N8nzg4VX1hI5LHDtJtgI+Bnyvqj7UdT3jJsnNgAcCvxn1qL1tP4/fD8+O3I4qfruqunCU9Yyjdb1GE+tG9Rq1f28GbUHzhWLiKrJ5NJfML5vp/hfaNOM+tMEo2JFsbtoZmOgceC2wQ3v7G8Bfd1HQuKuqG4DDaE4FbvaSHJvkX9rbW9BcFvot4NdJ/mbE5ZwPTDXK61+067SO16idFmBkr1FVbTfxQ9OifBbNAJZbtT8PppnA+CmjqknT9r80l3t/i6YPzxXtz5WsaUnsTJIdkvzF4M9MH8NTWnPThcBt2t/nAo8EfkwzfsH1HdY17m6FIy1PeCQw0dfp74DtaJq7n0Vz7vzrI6wlTN2BcluaaVQ0nq/RETRz+g0OWveDJC8BjgW+1kVRWqex6zye5A7Ah2hqG+yLOvF+n9G+aQaeuemLNGNenAa8D/hMO0/UrsA7uyxsHCR52fAiYBeaUWD/Z/QVjaVb0HSGBXgU8N9VdXmSzzKiVrCBzuUFHJ5k8KqM+TTfRn86ilrG1Zi/RrsxNOhpaynNxRQaI12N1bQBx9CcoXgWI7hyzMAzB1XVawZu/1eS39P0vzinqvxWBf86dH81TZPtMTQdvQWXAvduJ1p9JM0IudC0GIxqJNaJzuUB7kEzOu+E5TQdzDf3me3H+TU6HXh/kqdW1cUASXaluZDCKSXGUJI9gOcCd6ZpnftDO+v976rqJx2UdD9gv6r6+SgOZuCZg5I8hGaI8JUAVXU6cHqSBUkeUlXf7bbCblXVWFzRMuaOphkN9xJgFfDtdvn9gV+NooCqehhAkmOAF1fVNaM47lwy5q/Rs4EvARckubhdtivwa+BxHdWkdUjy18BXaE5X/yXN1B/QhJ9n0s2/2fnAlqM6mFdpzUFJVgG7VNXlQ8tvCVy+uY/DMyjJtjSjiE7V9L5ZS/J4mlMPn6+qi9plzwD+VFVf7rQ4TZLkaJrAs2Ro+TbAB6rqWR3VFeARrBm992zgf6e6mkzdSnI68PGq+s/2ars9q+q8JPsCX62q23RQ018Crwb+ZXi05Vk5nu/LuaedSHDnqrpiaPndgMVeDgpJXkAzE/iu7aKLaGYB/8/uqtK6JHkYawbS3GJwXVX9ZUc1/QvwApoxcO7dfji8GjhvNsYI2UAt6/qSsyNwaVXZWq/1SnItzfv4gqHAc0fgl1W1VQc1LaFp4ZkPLANWDq53aonN2MB0AAV8KsmygdXzgXsD/zfywsZMktcCr6Hp2/D9dvGDgbcl2b6q3tZZcWMiyYHrW19VXxhhLc+kuVLjizRDyn8ZuBtN0PjUqOoYquklNLN+vx0YfL9cTDOdwkgCT3tpbtqfWyQZ/ECYDzwGuGwUtUylre9RTB1UZ3Sma22yq2m+AF4wtHwfmi+EXXjhKA9mC88c0p7Hh2bY7c8x+RL05TRv5I9U1ZUjLm2sJLkQeFVVfWZo+VOBt1bVHbqpbHy0rYRTKRjt9CRJfg68t6o+OvTN80jg2qp69ahqGajpV8C/VdUJQzXdC/huVd1yRHWsZv1XrhTwxqo6bBT1DEqyH3ACzTfzW9GEwV3a+xdU1X1GVMcCmvHHTq8qp0xZhyRvp/ni9ySaU4+LaP69jgWO2RwCqi08c8jEdADtUNxH2C9lnXYCfjTF8h/SDNq42Ruevbr90NibZliDUQ/OeCeaQdGg+bCcGCvpSOBkmnP8o3YHYKorR1awprPnKDyMpnXnO8DjgcHRcJfTXF1zyQjrGfRO4DiaaQGuoekIex3wGZpRzUeiqlYm+QJNPyIDz7q9nibc/I41/a0CfJpmUNZOJNkZeBpN5+lDqurKJA8ELqmqGR1U08AzN71l8E6SWwOPBc6uqs3+lBbNXDFPAYa/sTyF5goSDWmv+PtRezrwg8CeIzz8VTQDH0LTSnBvmhF8b8low8Wg82ia+oen2Xg0a0Y5n3UTY6e0/SwuHLPOwPcBnl1V1fYx2rJtBXsVzYfocSOs5UzgLqx9ukatqloBPDXJG2i+3MwDflJVv+mqprbD9Ldprta6F02IvpKmI/zdmOERuw08c9MJNNNIvK+9CmkxsA2wbZJnO2kfbwI+116+/wOaZv8HAQ8FnthhXXPBn2i+aY3S92hOSfyM5lTt+5M8gmZwzRNHXMuEI4Ajk2xN8y14/yRPo+nXM/Iroqrqd0n2SDJOY6gMjgl0GU2r2C9pprsZ9RU/bwLeleSNNKPOT2r97mqeqPb9sxdNq/OkVtVR9pNra/l/wAlV9Vvgt6M89nocAbyvqt44NE/bN4EZn+DYwDM37UvzhxfgQJrm5DvSjCT8cmCzDjxV9YUk9wNeRtPyNdF8e7+OPhjGTpJ9hhfRnM9/FTDq1+iFNPMwQTMw5EqagTQ/B/z7iGsBoKqOaU/zvRXYGvgkTevTi6rq+FHXM6ZjqJwB3JemRfVk4N/b0xMH0bTQjdIJ7e8vMLnP06xMUTAdSf6K5vTeVP29uqjpM8DSJJ8HPjkmZwP2pRnPadgfmIXuBwaeuWk7mm/i0Hwz/mJVrUjyHeA/OqtqDCRZSHNlz2ur6qCu6xlji2n+6GZo+WmMuAVj8Nt3Va2muTKqM23QORj4UlV9pL30e97wJeEj9hbgZQNjqEw4Gfi3bkridaw5Ffl6mi9aH6AJQDP+7XwDxm6eKJppf06g+VvUVT+rQTsDT6A5TfTd9uKO44BPVVVXp/qvp5nmZtjurJn6ZsZ4ldYclOTXwBuBr9Kcs35iVZ2cZC/gxKqaaubpzUaSq4F9q+q8Dmv4DnBgVf0pydOB46tq2YYeNyrtpH2DVgNXtLPKdyLJbZi66f+MDmq5DrhnVQ334enEOI6hovVr30P3aU8hjZUku9CMe/UUmv48P66q+3VQx1E0kxY/kabvzn1ovoh9GfhOVb10Jo9nC8/c9G6aJvZraTpVTkwl8RCafhCbuy/QnOrrch6mB9KcCvkTzRxe32AWvrHcVG2fkFsDD2AgZDQD58IoB2hMsjdNq9zEaL2TSqWD0xE0LV37snan5a6M4xgqACRZRHNq7WtVdV07+vOyialvRljHzjQDRd6T5n3zC+CDVdXVOEU/AO7O+PSXuVHb/+tImvf362ne6114Oc2EzlfQ/L38Pk1L1P+1dc0oA88cVFUfTrKYZrCvE9vTAND8xzqku8rGxoXA65M8mObUzXAHxnePoIZfAW9NchLNh/iTkkw5D1IXncyTHAR8lKa2q5nc76GAUY5IfRTwe+A5jGDG5Gn6CHBEktszdSfYUbc6fRp4Z5In0bw+C5I8lCbUH7PeR86SNmB8haYfTwF3pbm67d3ADTSXq4+qlgfSfKm4DDi1XXwQ8LIkj6yqU9f54NnzIZr30G1ovohOmpS3i5ZLuHFU86fSDHMAzYCfL+uilnZuuAe1U0zsQ/PF64yq+t/1P/Km8ZTWHJPk5jTNpN+bYt0DaS5Nv3r0lY2PJOsbu6Gq6k4jqOEBNOfw7wJsT3Oueqr/bNXFVCBJfgd8HDh01N/Ep6jlOmDvqjqnyzoGrWdgRmj+zUba6tT2TTsW+EeakLqa5sPhOOCZVbVqlPW0NX2a5urQZ9J8yZg4zfZXNPN73WOEtZxKEyqeN/EFMMk8mtBx76p6wKhqGahp3N5D76R5/+xEcxXUp4Avd3WqvYvPMgPPHJNkO5oe7I+sqh8MLN8LOB3YdXMfaXnctH/4bt1xp9dJxqGf00AtpwGvrKrvbnDjEZmij9MkXfXtSXIn1nwT7noMlcuAh1fVz6foV/TzqtpmhLVcD+w13Pk2ye40r9PIx3Mat/dQkv+jCTmf7eoy/aF6Rv5Z5imtOaaqliT5MvB0mnPEEw4CvtlV2GmvbLkfU8+pM9JTNmlmlp5K0TS1n0vTiXhUV07cEZif5FAm9y/4zw5D0HE08zB9oIuDt3MwTXgt8I4kr2fqpv+R/3Fu+zit6z1dNH3oRmYd7+lHJenqPQ3NpfHLp1h+q7amUfozzf+z4auN7siaK1pHatzeQ1X1gCR/A3yiDc6PrKrfJ/ln4Pyq+vaI6xn5Z5ktPHNQkkfSjKmwc3s5+jyajosvHPVgVm09u9NcMXZHmub2VTRhegVN58WRnrJJ8lWaOWNWs2Z6gHu3tf2YZkTPbYEHV9VPR1DPVP0L9qdpWu6kf0GSLYAv0XxgTRUyZnVenaw9R9REZ+W1xlAZddM/+J6eZk0nAGdW1WvbFp770Jza+hywqqqeNIo62lreS3OlzytpOrxODDb6NuBzVTXyPipj+B56Ks0pvo8CzwPu1bbIPZfmitJHjrKetqaRfpbZwjM3nQgsBf6W5oqkh9N8e/hqR/W8l+aP7l7Ape3vm9NMUTDjPe2n4Qc0V7A9u6qWwo0jnn6EZgj6R9OMGfIumtduth1B8596qv4F76K5UmrUnkszy/WVNP2Mhjstz/ZEgoPjpuxG02l5uB/KPJpvxl14L76nN+TlNOO53BfYsj32vWhepwfO9sHTjKT+f20ftFfShIqjWfO5toLm36uLudhg/N5DrwSeU1WfbVt1JpzG7P9/X5fRfpZVlT9z8IdmcLYvtbc/AfxHh7VcRdMxEJqm5bu3tx8KnNVBPX8A7jHF8nsCf2hv7w1cNaJ6rp94TYaW7w5c39G/2eXAS7t6zwzVsgrYaYrlt6RpKeiiJt/T669nIU0/i/2ANwNfo7m8+N+BXUb9vqG5OuyWNJc270HT2rR1F++dMX4PLQXu0N5eAtypvX3nrv4Otccf2WeZLTxz1yeAHye5HfD3jO5b3VRC858JmvEUdqU5l34RTevBqG1LM03CL4eW35o1M3Ffw+haOMeufwHN2DZf6ejYwyaG/x+2LaPvCzLB9/R6VHP64Y7AH6vqjaM45hSupvk/dDlNK+G8alq/xmUssnF7D11CMyHncGfph9DtWEEj+ywz8MxRVfWLJD+jGZ/joqr6YYfl/Jxmdu3zgB8Cr0oze/JzaDpTjtoXgY8leSXwI5oP0/sB76BpNqW9P6rLoD87UM9w/4LPjKiGYcfQjMXRVVM2Sd7f3izg8CRLB1bPp/k3+umo62r5nt6wj9O8Hq8Y4TEH/TdwSpI/0Lwei9t/o7XUCIaimMK4vYeOopmYd+J01u3ascreQTP5aidG+Vlm4JnbPklznvh1HddxGM14HNCcm/4acBJN/5CRdVwc8Dyawc8+xZr3+Eqa8/svb+//kuYPzyiMY/+CrYF/bjsNnsXanZZfNIIa9mh/B7gHk6/4WU4zOWVXo2X7nt6wbYCnppnZfqrBGWf7PfQ8mlbKu9K8NsfQnKoZF2P1Hqqqd7Rj35xIM1nvScAy4Iiq6noOxpF8lnmV1hzWXtr7r8CHq+rSrusZ1NZ2dXX4BmuHuL8zzQfquVV13QYeMtv1bD1Uz9INPGQ2azlpPaurqv5yhLUcA7y4mlFXx5bv6bVqGbf30IuqapwCz1rG5D20NU3fr3k0g/td21UtE0b1WWbgkSRJvTdvw5tIkiTNbQYeSZLUewaeHkhycNc1DLKe9Ru3emD8arKe9bOeDRu3mqxn/UZRj4GnH8bqjYv1bMi41QPjV5P1rJ/1bNi41WQ962fgkSRJ2lRepdWRLbJlbXXjEA2bZgXLWMiWM7KvmdDXerJw4QxUA8tXX88W82626TtaMHNzai5fuZQtFmw9Y/vbVNazfn2up+ZnwxtNw4oV17Fw4ab/ja2MVz0zUw0sX34dW2wxM59Bq2fgT9HKZdexYMuZqee6P118ZVXdani5Aw92ZCu24f7pcjaIOWDeyCfJXq8FO+/cdQmTrL7VDl2XMNk4fnmaoQ+r3hrDf7NV247PlyWAlTcbr79D81aO37/Z8puPV5T4wZdeOTx9BuApLUmStBkw8EiSpN4z8EiSpN4z8EiSpN4z8EiSpN4z8EiSpN4z8EiSpN4z8EiSpN4z8EiSpN4z8EiSpN4z8KxDkgOSVJIdu65FkiRtGgNPK8nJSY7sug5JkjTzDDySJKn3DDxAkmOBhwIvaE9jFbBbu3rPJKcnWZpkcZJ9hh77gCSntOsvTvLBJNuP9hlIkqT1MfA0XgycChwD7NL+/L5ddzjwamAf4CrguCQBSLIH8C3gK8CewIHAXsDRI6xdkiRtwIKuCxgHVfXnJMuBpVV1KUCS3dvVh1TVSe2yQ4HvA7sCFwGvAI6vqndN7CvJ84GfJNmpqi4f5fOQJElTM/Bs2FkDty9pf+9EE3j2Be6S5B8Gtkn7+87ApMCT5GDgYICt2HpWipUkSWsz8GzYioHb1f6eN/D7o8B7pnjcxcMLquoo4CiA7fMXtdYjJEnSrDDwrLEcmL+RjzkDuFdVnTsL9UiSpBlip+U1LgDul2S3drDB6bw2b28f86Ekeye5S5LHJvnwrFYqSZI2ioFnjSNoWnnOBq4Abr+hB1TVWcBDaC5hPwU4k+aqrstmrUpJkrTRPKXVqqpzgP2HFh87tM0FrOmUPLFsMfCo2axNkiRtGlt4JElS7xl4JElS7xl4JElS7xl4JElS7xl4JElS7xl4JElS7xl4JElS7xl4JElS7xl4JElS7xl4JElS7xl4JElS7zmXVpfmze+6ghvN22Jh1yWsZd6tduy6hEmu+vBWXZcwydJv79B1CZPs8NuVXZewlq0vWtp1CZNk+Zi9RsmGtxmxBVde23UJkyz40zVdlzDZ6uq6grVsuc3Nui5hWmzhkSRJvWfgkSRJvWfgkSRJvWfgkSRJvWfgkSRJvWfgkSRJvWfgkSRJvWfgkSRJvWfgkSRJvWfgkSRJvWfgkSRJvWfgkSRJvWfgkSRJvdfLwJPk5CRHdnDcJyQZv6lsJUnazPUy8EiSJA0y8EiSpN7rc+BZkOR9Sa5uf96ZZB5AkoOS/CjJkiSXJ/l8kl0nHpjkgCSV5OFJTk+yNMniJPsMHiDJ05P8rl3/NWDnET9HSZI0DX0OPE+leX77A88FDgZe0q7bAngjsCfwWGBH4DNT7ONw4NXAPsBVwHFJApDk/sCxwFHAXsBXgUNn44lIkqRNs6DrAmbRH4AXVVUBv0pyN+BlwLur6uiB7c5L8nzgl0luW1UXDaw7pKpOAkhyKPB9YFfgIuDFwLer6rB223OS3Bd49roKSnIwTfBiK7aekScpSZI2rM8tPKe1YWfCqcCuSbZPsk+SL7eno5YAi9ttbj+0j7MGbl/S/t6p/X2Pdp+Dhu9PUlVHVdWiqlq0kC2n/0wkSdIm6XPgWZcA3wSWAk8D7gs8ql23xdC2KwZuT4Snidcss1WgJEmaWX0+pXX/JBlo5dmPppXmLjR9dl5bVecDJDnwJuz/7Hafg4bvS5KkMdDnFp7bAO9NcvckTwBeAbwHuBBYBrwwyZ2SPAZ4y03Y//uBv0rymiR3TfIc4O9nqnhJkjRz+hx4jgPmA6cDHwE+Brynqq4AngE8jqaV5o00nZk3SlWdRtNB+fk0fX0OBN40A3VLkqQZ1stTWlV1wMDdF06x/njg+KHFGVh/MkN9dKrqgimWHQMcM7SfkU9pIUmS1q/PLTySJEmAgUeSJG0GDDySJKn3DDySJKn3DDySJKn3DDySJKn3DDySJKn3DDySJKn3DDySJKn3DDySJKn3ejm1hDbe6htu6LqEtfz5gbftuoRJdtrmgq5LmGTJL27RdQmTLFyyousS1pIbVnZdwiTzrl3adQmT1JJruy5hbfPmd13BZKtXdV3BZKvGrB6g/jxe/8/WxRYeSZLUewYeSZLUewYeSZLUewYeSZLUewYeSZLUewYeSZLUewYeSZLUewYeSZLUewYeSZLUewYeSZLUewYeSZLUewYeSZLUewYeSZLUewYeSZLUewYeSZLUewYeSZLUe5tl4EnyN0mWJFnQ3r9rkkrywYFtDktyYpL5ST6W5Pwk1yf5TZJXJpk3sO0eSb6d5Jp2v2cmeVgXz02SJK1tQdcFdOR7wFbAIuA04ADgSmAwpBwA/A9NKLwYeBJwBXA/4CjgKuBj7bafBs5s160E9gBumN2nIEmSpmuzDDxVdW2SM2gCzkTgORJ4dZJdgD8D9wVeWVUrgDcMPPyCJPsAT2ZN4LkDcERV/aq9f+5Ux01yMHAwwFZsPaPPSZIkrdtmeUqrdTJN0AF4KPB14IftsgcCK9r7JHleksVJrkhyLfBS4PYD+3o38NEk30nyuiS7T3XAqjqqqhZV1aKFbDkLT0mSJE1lcw88D0xyT2A74MftsofRhJ7/q6oVSf4BeC9wLPBIYC/gP4EtJnZUVW8C7gl8CXgAcFaSZ43gOUiSpGnYLE9ptb4HbAm8Evh+Va1KcjJN/5zLafrvADwIOL2qjpx4YJI7D++sqn4D/AZ4f9v5+Z+Bo2f1GUiSpGnZbFt4qupa4AzgIOCkdvGpwO2A+9O09gCcA+zTXtl11ySH0JwCAyDJzZL8R5IDkuyW5P40IensET0VSZK0AZtt4GmdBMynDTdVdQNNJ+ZltP13gA8Dn6O5EutHwG7Auwb2sQq4BfBx4NfAF2mC08tmu3hJkjQ9m/MpLarq1cCrh5YdMHR/OfDs9mfQoQPrnzJ7VUqSpE21ubfwSJKkzYCBR5Ik9Z6BR5Ik9Z6BR5Ik9Z6BR5Ik9Z6BR5Ik9Z6BR5Ik9Z6BR5Ik9Z6BR5Ik9Z6BR5Ik9Z6BR5Ik9d5mPZdWl1bddUuu+cBuXZdxo1s8/ZquS1jLbV5wbtclTHL1sq27LmGSrc/8fdclTLL66j91XcJaauXKrkuYZOXq6rqEyWp11xWsLWP2PXzcXqNxe32AeVss7LqEaRm/V06SJGmGGXgkSVLvGXgkSVLvGXgkSVLvGXgkSVLvGXgkSVLvGXgkSVLvGXgkSVLvGXgkSVLvGXgkSVLvGXgkSVLvGXgkSVLvbXLgSXJAkkqy40wUNEpJ3pTk5wP3j03ytQ08ZoPbSJKk8TIWLTxtYHpCB4c+AnhoB8eVJEkjtKDrArpUVdcC13ZdhyRJml3TauFJsmWS9ya5LMkNSU5L8qChzfZL8tN2/Y+T7Dvw+Jsn+WSSy9v15yV5Sbvugnazz7ctPRcMPO5v233dkOT8JIcl2WJg/QVJXp/kw0muSXJRklcM1f7cJOe0+7giyTeTLGjXTTqlNfCY17fP9dokxyS52Tpel6cnuSrJlkPLj0vylWm8tJIkaQSme0rrHcA/AM8C9gZ+BnwjyS4D2xwBvApYBJwHnJBk63bdvwN7AI8Fdm/3c3G77r7t7+cAu0zcT/JI4DjgSOBe7WOeALx1qLaXtvXsA7wdeEeS/dt9LAL+A3gzcHfgr4BvbOC5PhTYE3g48Hjgr9v9TuXzNK/h/5tYkOTmwN8DH9vAcSRJ0ohsMPAk2QZ4PvCqqjqhqn4JPA+4DHjBwKZvqapvVtXPgX8CtgKe0q67A/CTqvphVV1QVSdX1ecBquqKdps/VdWlA/dfB7yzqo6pqt9W1Uk0gep5STJw3G9V1ZFVdW5VfQA4lyasANweuA74SlX9rqrOrKr3VNXK9TzlVcA/VdXPq+qb7TGf274Ok1TV9TSh7FkDi58CXAOcsJ5jSJKkEZpOC8+dgYXADyYWVNUq4FTgngPbnTqw/lqaVpeJ9R8EnpTkzCRHJJlOR+F9gde1p5WuTXIt8GlgG+DWA9udNfS4S4Cd2tsnAr8Dzm9PMz0jyXYbOO5Zbf2Dz2sLmtdhKh8BHpHktu39ZwEfnypUJTk4yeIki1f+eekGypAkSTNlOoFnojWlplg31bK1N6r6Ok0rzxHAjjSnu46ZRm1vBvYa+LkPcFfgioHtVkxR07z2uEtoTnU9CbgQeA3wqyS3mU7d01FVZwJnAM9Mcm+aU3pHr2Pbo6pqUVUtWnDzrafaRJIkzYLpBJ5zgeXAjZ2Uk8wH9gfOHthuv4H12wD3Bn45sayqrqyqT1bVM4FnA88Y6Oy7Apg/dNwzgN3bU1XDP+s7JTVJVa2squ9U1WtoAtM2NH2J1mWPodNX+9E8/9+u5zEfAZ4J/DPwg6r69XTrkyRJs2+Dl6VX1XVJPgi8LcmVwPk0HYV3Bv6TpjMwwOuTXEFzSukNNCHh0wBJDqUJML9oj3kgcF5VLWsfewHw8CSnAMuq6mrgUOBrSX4HfA5YSROi7ldVr5zOk0vyWJpTUd8F/gg8DNiOgSA2hQXA0W3NtwHeBnykqq5bz2M+A7ybpq/T86ZTmyRJGp3pjsPzqvb3McAOwE+AR1XVH5JMBJ5XA++iCUC/AB47EBKWAYcBdwRuAE4D/nZg//9GExh+T3P11m5V9c0kjwEOAV5OE3jOAY7diOf3J+BxNAFsa5pWmn+uqu+t5zGntPWf1D7mv4H1BqyqWpLkc8ATacKZJEkaI6maVjccbUCSrwMXVdVzprP9Nnfbpe79gWfMclXTd4unX9N1CWvZ+gvj9d68etl49bva6mnLuy5hktVX/6nrEtZSK6d99nskavV4vaep1V1XsLaMxQQAa4zbazRurw8wb4uFXZcwybeu/9SPq2rR8PLNeqTlmZDkL2jG9/lrmvF7JEnSmDHwbLozgL8AXtuOQSRJksaMgWcTVdVuXdcgSZLWb/xOBkqSJM0wA48kSeo9A48kSeo9A48kSeo9A48kSeo9A48kSeo9A48kSeo9A48kSeo9Bx7syBbzVnK77f7UdRk3+shPvtZ1CWt58l8e1HUJk2wxZvPFsGC85viZd4sdui5hLXX99V2XMN4WbtF1BWsbs/nPWLWq6womGbf54QCyYMyixDr+29vCI0mSes/AI0mSes/AI0mSes/AI0mSes/AI0mSes/AI0mSes/AI0mSes/AI0mSes/AI0mSes/AI0mSes/AI0mSes/AI0mSem+sAk+S3ZJUkkU34bHPTHLtJh7/TUl+vin7kCRJ42esAs8mOh6403Q2XE+wOgJ46IxXJkmSOjVmc7rfNEkWVtX1rHNS+OmpqmuBTWolkiRJ46eTFp40/i3Jb5IsS3JRksMHNrlDkhOTLE1ydpJHDDz2gLZ15tFJfphkOfDI4VNaSW6X5MtJ/tju51dJ/rFdfX77+0ftvk5uHzPplFaS+yb5VpIrk1yT5PtJ9h96LpXk4CSfT3JdkvOSHDTDL5kkSdoEXZ3SeitwCHA4cC/gicDvB9YfBrwf2BP4EfDZJNsO7ePtwOuB3YHTpzjGfwJbAw9rj/ES4E/tuvu1vx8F7AIcuI46twM+CTy4fcxPgf9JsuPQdm8AvtzWezxwdJI7rGOfkiRpxEZ+SqsNLi8FXlJVR7eLzwVOTbJbe/89VfXVdvvXAk8H9gK+P7CrN1XVtwb2O3yoOwD/XVVntvfPH1h3Rfv7qqq6dF21VtV3hmr/V+DxNEHpUwOrPllVn2q3OQR4MU1I+t269i1JkkanixaeewJbAt9ezzZnDdy+pP2909A2izdwnPcBr09yapJ/T7LvxpUJSXZK8uEk5yT5M7CkreP266q3qlbSBKrhemlPfS1OsnjZn27Y2HIkSdJN1EXgWaspZgorJm5UVbU3h2u9bn07qKqPAXcEjgHuBvxfkjdNv0wAPg7cl6ZF6gE0rUwXAVusq96Jw09RL1V1VFUtqqpFW+6w1UaWIkmSbqouAs/ZwDLg4bN9oKq6qA0ZT6LpZ3Nwu2p5+3v+BnbxIOADVXVCVf2CpoVnl9mpVpIkzZaR9+GpqiVJ3gccnmQZ8F3glsC+wNdn6jjtMb4OnANsT9Pv5ux29eU0l7A/MskFwA1V9ecpdnMOcFCS04FtgHewJixJkqQ5oqurtF5Dc5XVIcAvgf8GbjvDx5gHfIAm5JwIXAY8A27sZ/Mi4J9p+gh9eR37eBawLfBj4LPA0cAFM1ynJEmaZVnTRUajdIvdb1UPP/rxXZdxo4/s9rWuS1jLk/9yvIYzqi0Wdl3CJPP+PGZjZK5c2XUFa6nrN2ks0v5bONwdcQyM2/to1aquK5ikxu31AbJgvMYw/uafj/5xVa01RVWfppaQJEmakoFHkiT1noFHkiT1noFHkiT1noFHkiT1noFHkiT1noFHkiT1noFHkiT1noFHkiT1noFHkiT1noFHkiT13nhNgLEZ2XL+Ku64zVVdl3Gjm8+7WdclrG316q4rmGx+uq5gkrGbJ2oM5+WrVWP2Hho3K5Z2XcFaMn9+1yVMMnbzTWa8/g4B1JjNN7YutvBIkqTeM/BIkqTeM/BIkqTeM/BIkqTeM/BIkqTeM/BIkqTeM/BIkqTeM/BIkqTeM/BIkqTeM/BIkqTeM/BIkqTeM/BIkqTeM/BIkqTeM/DMsCSLklSS3bquRZIkNQw8kiSp9ww8kiSp9zbrwJPkUUm+l+TqJH9M8s0k92jX7daemnp8khOTLE1ydpJHTLGPXyW5Icn3gLt18mQkSdI6bdaBB9gGeC9wP+AA4M/AV5NsMbDNYcD7gT2BHwGfTbItQJLbAV8CTgT2Aj4AvGMklUuSpGlb0HUBXaqq/x68n+SfgGtoAtBF7eL3VNVX2/WvBZ5OE26+DzwfuBB4UVUV8KskdwPeMtXxkhwMHAyw3S5bz/TTkSRJ67BZt/AkuXOSTyf5bZJrgMtoXpPbD2x21sDtS9rfO7W/7wGc1oadCaeu63hVdVRVLaqqRVvfYssZeAaSJGk6NusWHuCrwMXAc9vfK4GzgcFTWismblRVJYE1QTGjKVOSJG2KzbaFJ8ktaVpo3lpV/1tVvwS2Y+NC4NnA/dOmoNZ+M1imJEmaAZtt4AGuBq4EnpPkLkkeCnyIppVnuj4E7Aa8N8ndkzwBeN6MVypJkjbJZht4qmo18A/AfYCfA/8BHAIs24h9XAgcCDwKOBN4KfDqGS9WkiRtks26D09VfQe499DibQdur9VHp6oydP8E4IShzY6bkQIlSdKM2GxbeCRJ0ubDwCNJknrPwCNJknrPwCNJknrPwCNJknrPwCNJknrPwCNJknrPwCNJknrPwCNJknrPwCNJknrPwCNJknpvs55Lq0tLLt2G775jv67LuNHet3xA1yWsZYc7L++6hEkWLF3VdQmTZJvbdl3CZNV1AWtbcM0NXZcwSa6f9tzEI7F6u5t1XcJaVs8br+/h85at6LqEyS6/qusK1lK33bnrEib7ydSLx+udJUmSNAsMPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfcMPOuQ5IAklWTHrmuRJEmbxsDTSnJykiO7rkOSJM08A48kSeo9Aw+Q5FjgocAL2tNYBezWrt4zyelJliZZnGSfocc+IMkp7fqLk3wwyfajfQaSJGl9DDyNFwOnAscAu7Q/v2/XHQ68GtgHuAo4LkkAkuwBfAv4CrAncCCwF3D0CGuXJEkbsKDrAsZBVf05yXJgaVVdCpBk93b1IVV1UrvsUOD7wK7ARcArgOOr6l0T+0ryfOAnSXaqqssHj5PkYOBggC22vsUsPytJkjTBFp4NO2vg9iXt753a3/sCByW5duIH+EG77s7DO6qqo6pqUVUtWrDVNrNXsSRJmsQWng1bMXC72t/zBn5/FHjPFI+7eDaLkiRJ02fgWWM5MH8jH3MGcK+qOncW6pEkSTPEU1prXADcL8lu7WCD03lt3t4+5kNJ9k5ylySPTfLhWa1UkiRtFAPPGkfQtPKcDVwB3H5DD6iqs4CH0FzCfgpwJs1VXZfNWpWSJGmjeUqrVVXnAPsPLT52aJsLgAwtWww8ajZrkyRJm8YWHkmS1HsGHkmS1HsGHkmS1HsGHkmS1HsGHkmS1HsGHkmS1HsGHkmS1HsGHkmS1HsGHkmS1HsGHkmS1HsGHkmS1HvOpdWR1QvhutuMT97c+fSlXZewlnP/abzenjudslXXJUxyi19c03UJk8xburzrEtaS65d1XcJYm3fN+P2/p6rrCiap667vuoTJxuz1AZh36VVdlzAt4/OJK0mSNEsMPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfc2u8CTZGHXNUiSpNHqReBJ49+S/CbJsiQXJTk8yW5JKsmTk3wnyfXAc5PMS3JIkt+32/8syf8b2N/xST44cP+wdj/3H1h2UZKntrf3SPLtJNckWZLkzCQPG+mLIEmS1mlB1wXMkLcCzwdeBnwXuBWw98D6w4GXA88GVgAvBl4BPA9YDBwEfCHJvlX1U+DkdpsJBwBXAg8DTk9yV2DXdjuATwNnAvcDVgJ7ADfM6DOUJEk32Zxv4UmyLfBS4NVVdXRVnVtVp1bVfw5s9oGq+q+qOr+qLqIJP0dU1aer6pyqegPwvXY5NEHm7kl2SbI1sAh4F03ggSYAnVtVF7f37wCcWFW/ao//xao6dYpaD06yOMniVUuvm9kXQpIkrdOcDzzAPYEtgW+vZ5vFEzeSbA/cBvjB0Dbfb/dFVf0SuIwm2DwQ+C3wWeCBbR+gA1jTugPwbuCj7Wmz1yXZfaoiquqoqlpUVYvmb73NdJ+fJEnaRH0IPJnGNlM1p9QGlp1C06JzAHBSVV1Ac1rrvsBDGQg8VfUmmrD0JeABwFlJnjWNuiRJ0gj0IfCcDSwDHj6djavqGuAS4EFDqx7U7mvCyawJPCe3y04BDmZy/52J/f6mqt5fVY8BPgb88/SfgiRJmk1zvtNyVS1J8j7g8CTLaDot3xLYF/j6Oh72TuDQJL8BfkzTafnB7WMmnAz8J7Aba8LNycBHGOi/k+RmwBHA54ELgJ1pwtPpM/D0JEnSDJjzgaf1GuBq4BDgtjT9bz6xnu3fD2wHvIMmoPwaeHx7hRbQ9ONJcilwVVVd0S4+CZjP5NadVcAtgI8DtwauAr7Gmg7QkiSpY70IPFW1Gnhb+zNsrT4+7fZvaX/Wt99dhu5fMLy/qloOPGXjKpYkSaPUhz48kiRJ62XgkSRJvWfgkSRJvWfgkSRJvWfgkSRJvWfgkSRJvWfgkSRJvWfgkSRJvWfgkSRJvWfgkSRJvWfgkSRJvdeLubTmogU3FH/xyxVdl3Gj+Tes7LqEteyw4w1dlzDJwutu3nUJk63uuoDJaovx+3OSJUu7LmGyleP1/6xWjdmbCGDeWtMfdqvG7DXKmL0+QG27ddclTIstPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfcMPJIkqfc2u8CT5IAklWTHrmuRJEmjsdkFnpmUZGHXNUiSpA2bc4EnjVcm+W2S65P8LMlB7brd2tabxyc5McnSJGcnecTEeuCkdldXtNseu6H9Du37yUm+k+R64LlJbp7kk0kuT3JDkvOSvGSUr4kkSVq/BV0XcBP8O/AE4AXAr4H9gY8kuRr4RbvNYcArgH8BXg98NskdgN8Djwf+G7gX8Efg+g3tt6pOGDj+4cDLgWcDK9rH7QE8Frgc2A241Uw/aUmSdNPNqcCTZBvgZcBfV9X32sXnJ7kfTVD5l3bZe6rqq+1jXgs8Hdirqr6f5I/tNpdX1ZXT3O9g4PlAVf3XQE13AH5SVT9sF10wc89YkiTNhDkVeIB7AlsB30hSA8sXMjlonDVw+5L2904zsF+AxUP3Pwj8V5J9gBOBr1bVKVMdJMnBwMEAW95sh/WUI0mSZtJcCzwTfY7+FrhwaN0KIAO3AaiqSjL42Juy30HXDd6pqq+3rTx/AzwcOCHJ56vqn4YPUlVHAUcBbLfDbWt4vSRJmh1zLfCcDSwD7lBV3xle2XZK3pDl7e/5093vhrSnxj4JfDLJ14HPJHleVS3b2H1JkqSZN6cCT1UtSXIEcESaZpvvAtsC+wGrgW9NYze/Awp4TJKvAtdvaL9ty8yUkhwKnEHTYXoBcCBwnmFHkqTxMecuSwcOAd5Ec6XUL2j6zTweOH86D66qi4E30lzJdRlw5Cbud1m7rzOBHwDb0ZwakyRJY2JOtfBA0ycH+ED7M5UML6iqDN1/C/CWjdlvVV2wjn0fRhN4JEnSmJqLLTySJEkbxcAjSZJ6z8AjSZJ6z8AjSZJ6z8AjSZJ6z8AjSZJ6z8AjSZJ6z8AjSZJ6z8AjSZJ6z8AjSZJ6z8AjSZJ6b87NpdUXWVls+cfxmVA9K1d3XcJa/vFOP+66hEk+u8sjui5hku3PXt51CZNd+aeuK1jL6qVLuy5hsqquK5iklo/ZewjIgjH7WJo3Xu0C2WJh1yWs7U/XdF3BtIzXv6QkSdIsMPBIkqTeM/BIkqTeM/BIkqTeM/BIkqTeM/BIkqTeM/BIkqTeM/BIkqTeM/BIkqTeM/BIkqTeM/BIkqTeM/BIkqTeM/CsQ5IDklSSHbuuRZIkbRoDTyvJyUmO7LoOSZI08ww8kiSp9ww8QJJjgYcCL2hPYxWwW7t6zySnJ1maZHGSfYYe+4Akp7TrL07ywSTbj/YZSJKk9THwNF4MnAocA+zS/vy+XXc48GpgH+Aq4LgkAUiyB/At4CvAnsCBwF7A0SOsXZIkbcCCrgsYB1X15yTLgaVVdSlAkt3b1YdU1UntskOB7wO7AhcBrwCOr6p3TewryfOBnyTZqaouH+XzkCRJUzPwbNhZA7cvaX/vRBN49gXukuQfBrZJ+/vOwKTAk+Rg4GCArba4+awUK0mS1mbg2bAVA7er/T1v4PdHgfdM8biLhxdU1VHAUQDbb7trrfUISZI0Kww8aywH5m/kY84A7lVV585CPZIkaYbYaXmNC4D7JdmtHWxwOq/N29vHfCjJ3knukuSxST48q5VKkqSNYuBZ4wiaVp6zgSuA22/oAVV1FvAQmkvYTwHOpLmq67JZq1KSJG00T2m1quocYP+hxccObXMBazolTyxbDDxqNmuTJEmbxhYeSZLUewYeSZLUewYeSZLUewYeSZLUewYeSZLUewYeSZLUewYeSZLUewYeSZLUewYeSZLUewYeSZLUewYeSZLUewYeSZLUe04e2pGVW8/j8n227bqMG93qzKVdl7CWT33yEV2XMMnKHbuuYLLVW2/RdQmTzNtxh65LWMu8pVt1XcJky5Z3XcFk8+d3XcHaVq3quoJJatXqrkuYJPPHsJ1iHN9HUxjDV06SJGlmGXgkSVLvGXgkSVLvGXgkSVLvGXgkSVLvGXgkSVLvGXgkSVLvGXgkSVLvGXgkSVLvGXgkSVLvGXgkSVLvGXgkSVLv9TLwJDk5yZEdHPcJSWrUx5UkSevXy8AjSZI0yMAjSZJ6r8+BZ0GS9yW5uv15Z5J5AEkOSvKjJEuSXJ7k80l2nXhgkgOSVJKHJzk9ydIki5PsM3iAJE9P8rt2/deAnUf8HCVJ0jT0OfA8leb57Q88FzgYeEm7bgvgjcCewGOBHYHPTLGPw4FXA/sAVwHHJQlAkvsDxwJHAXsBXwUOnY0nIkmSNs2CrguYRX8AXlRVBfwqyd2AlwHvrqqjB7Y7L8nzgV8muW1VXTSw7pCqOgkgyaHA94FdgYuAFwPfrqrD2m3PSXJf4NnrKijJwTTBi4Xb3mJGnqQkSdqwPrfwnNaGnQmnArsm2T7JPkm+3J6OWgIsbre5/dA+zhq4fUn7e6f29z3afQ4avj9JVR1VVYuqatGCm20z/WciSZI2SZ8Dz7oE+CawFHgacF/gUe26LYa2XTFweyI8Tbxmma0CJUnSzOrzKa37J8lAK89+NK00d6Hps/PaqjofIMmBN2H/Z7f7HDR8X5IkjYE+t/DcBnhvkrsneQLwCuA9wIXAMuCFSe6U5DHAW27C/t8P/FWS1yS5a5LnAH8/U8VLkqSZ0+fAcxwwHzgd+AjwMeA9VXUF8AzgcTStNG+k6cy8UarqNJoOys+n6etzIPCmGahbkiTNsF6e0qqqAwbuvnCK9ccDxw8tzsD6kxnqo1NVF0yx7BjgmKH9jHxKC0mStH59buGRJEkCDDySJGkzYOCRJEm9Z+CRJEm9Z+CRJEm9Z+CRJEm9Z+CRJEm9Z+CRJEm9Z+CRJEm9Z+CRJEm9Z+CRJEm918u5tOaCBdevZsdfXN91GTda8Mfrui5hLbd+5FVdlzDJ1Z+9bdclTDLvhpVdlzBZsuFtRm3V6q4rGGt1ww1dl7C2efO7rmCy1au6rmCSWrG86xLWkm226bqEabGFR5Ik9Z6BR5Ik9Z6BR5Ik9Z6BR5Ik9Z6BR5Ik9Z6BR5Ik9Z6BR5Ik9Z6BR5Ik9Z6BR5Ik9Z6BR5Ik9Z6BR5Ik9Z6BR5Ik9V4vAk+Sk5McOQZ1PCFJdV2HJEmarBeBR5IkaX0MPJIkqffmROBpT1l9KMn7klzd/rwzyZT1JzkoyY+SLElyeZLPJ9m1XZck5yZ5+dBj7pqkkuzT3r95kqPaxy9JckqSRUOPeXqS3yVZmuRrwM6z9BJIkqRNMCcCT+upNPXuDzwXOBh4yTq23QJ4I7An8FhgR+AzAFVVwMeAZw095lnAT6vqjCQBTgB2bR+/N/Bd4DtJdgFIcn/gWOAoYC/gq8Chm/wsJUnSjFvQdQEb4Q/Ai9rA8qskdwNeBrx7eMOqOnrg7nlJng/8Msltq+oi4Bjg0CT7VdVpSeYDTwcObx/zMJoQc6uqur5ddkiSvwWeBrwDeDHw7ao6rF1/TpL7As9e1xNIcjBNUGPLLW++8a+AJEm6SeZSC89pbdiZcCqwa5LthzdMsk+SL7enm5YAi9tVtweoqkuBr7GmledRwC2B49r7+wJbA1ckuXbiB7g3cOd2m3u0NQwavj9JVR1VVYuqatEWC7eZxlOWJEkzYS618ExLkm2AbwL/S9MacznNKa3v0ZzqmvBR4NNJXkITfL5QVVe36+YBlwEPnuIQ10wcasaLlyRJs2IuBZ77J8lAK89+wCVVdU3T5eZGu9MEnNdW1fkASQ6cYn/foAkvzwP+Fnj0wLozaDogr66q89ZRz9ltDYOG70uSpDEwl05p3QZ4b5K7J3kC8ArgPVNsdyGwDHhhkjsleQzwluGNqmoVcDRNv52LgW8PrP5f4AfAl5P8TZI7Jtk/yZuTTLT6vB/4qySvaa/weg7w9zP0XCVJ0gyaS4HnOGA+cDrwEZorrdYKPFV1BfAM4HE0rTBvpOncPJWjaU5zHTPYP6i9/WjgO+2xfg18Drg7cEm7zWk0HZSfD5wFHAi8aZOeoSRJmhVz6ZTWyqp6IfDC4RVVdcDQ/eOB44c2m6rPza2BVTSXlw/vcwnNlVgvXldBVXUMzRVfgzqf4kKSJE02lwLPjEmyJXA74N+BL1bVhR2XJEmSZtFcOqU1k55Mc5rqlqz7dJckSeqJOdHCM3zKagb2dyxTnMaSJEn9tLm28EiSpM2IgUeSJPWegUeSJPWegUeSJPWegUeSJPWegUeSJPWegUeSJPWegUeSJPXenBh4sI9ql1WseN3VXZdxo7x5h65LWMvCedd1XcIkS27fdQWT7fSFK7ouYZK6/oauS1jL6uXLuy5BGytj9j28VnddwWTz53ddwVqyYmXXJUzLmL2zJEmSZp6BR5Ik9Z6BR5Ik9Z6BR5Ik9Z6BR5Ik9Z6BR5Ik9Z6BR5Ik9Z6BR5Ik9Z6BR5Ik9Z6BR5Ik9Z6BR5Ik9Z6BR5Ik9Z6BR5Ik9Z6BR5Ik9Z6BR5Ik9d6cDzxJHpLktCTXJvlzktOT3Ltdd2CSnyVZluT3SV6XJAOPvSDJ65N8OMk1SS5K8oqh/d8tySlJbkjy6ySPbo/1zIFt3pDkd+1xLk3yiZG9AJIkaYPmdOBJsgD4MvB9YE/g/sD7gFVJ9gU+D3wB2AN4NfAa4IVDu3kp8DNgH+DtwDuS7N/ufx7wRWAlsB/wTOCNwJYDNTweeDnwL8BdgccCP5zxJytJkm6yBV0XsIm2B3YAvlpVv22X/QogyXHAKVX1xnb5OUnuCrwK+MDAPr5VVUe2tz+Q5EXAw4FTgUcAdwf+uqoubvf7UuAHA4+/A/CHdj8rgAuBxTP6LCVJ0iaZ0y08VfVH4Fjgm0lOSPKyJLdrV9+DycEEmpagXZNsP7DsrKFtLgF2am/vDlwyEXZaPwJWD9z/PLAVcH6SjyV5YpItmUKSg5MsTrJ4xZ+XTvNZSpKkTTWnAw9AVf0Tzams7wJ/R9OS80ggQK3rYQO3V0yxbuJ1Wd8+Jo7/e5pWoOcC1wDvAn6cZJsptj2qqhZV1aKFN996vc9LkiTNnDkfeACq6syqentVHQCcDDwDOBt40NCmDwIuqqol09z1L2lahG4zsGwRQ69bVd1QVSdU1UuB+wL3Ah640U9EkiTNijndhyfJHWlaVr4CXAzcCbgP8EHgf4AfJXkT8GmaIPJvwGs34hAnAr8GPp7k5cDNgHfTdGKutoZn0ryOpwPXAv9A02r0m016cpIkacbM6cADLAXuRtOPZkfgMuA44O1VtSLJE4E304Scy4C3AUeuY19rqarVSf4e+CjNlVcX0ISmLwA3tJv9iaYj9BHAQpqWpQOr6vxNfG6SJGmGzOnAU1WXAQeuZ/0XaMLJutbvNsWyA4bunwM8ZOJ+kj1pgs257fovAV/amLolSdJozenAMwptC891NKeodqM5pXUmcEaHZUmSpI1g4Nmw7WgGJLwdcDVNp+iXVtV6r96SJEnjw8CzAVX1CcCpIiRJmsN6cVm6JEnS+hh4JElS7xl4JElS7xl4JElS7xl4JElS7xl4JElS7xl4JElS7xl4JElS7znwYEdWE5atmt91GTe67CWrui5hLTt+4HZdlzDJnX/9x65LmCRbbdV1CZOMWz0AtXJl1yVMNm71LBjDj4DVYzaI/YrlXVcw2fzx+dyYkIULuy5hsmunXmwLjyRJ6j0DjyRJ6j0DjyRJ6j0DjyRJ6j0DjyRJ6j0DjyRJ6j0DjyRJ6j0DjyRJ6j0DjyRJ6j0DjyRJ6j0DjyRJ6j0DjyRJ6j0DzwYkeXmSC7quQ5Ik3XQGHkmS1HtzOvAk2T7JDiM+5q2SbDXKY0qSpE0z5wJPkvlJHpnk08ClwJ7t8psnOSrJ5UmWJDklyaKBxz0zybVJHp7k50muS3JSkjsO7f+VSS5tt/0EsO1QCY8GLm2P9cBZfrqSJGkGzJnAk+ReSd4BXAgcD1wHPAr4bpIAJwC7Ao8F9ga+C3wnyS4Du9kSeA3wLGB/YAfgQwPHeBLw78AbgX2AXwMvGyrlOOApwHbAiUnOTfKG4eAkSZLGx1gHniS3TPKiJIuBnwC7Ay8Bdq6q51TVd6uqgIcBewFPqKofVtW5VXUIcB7wtIFdLgBe0G5zFnAE8LAkE6/DS4CPV9WHq+qcqjoM+OFgTVW1sqr+p6qeDOwMvLU9/m/aVqVnJRluFZIkSR0a68AD/CvwPmAZcNeq+ruq+nxVLRvabl9ga+CK9lTUtUmuBe4N3Hlgu2VV9euB+5cAC2laegDuAZw6tO/h+zeqqiVVdXRVPQy4L7AT8DHgCVNtn+TgJIuTLF7556XrftaSJGlGLei6gA04ClgBPB34RZIvAp8Evl1Vqwa2mwdcBjx4in1cM3B75dC6Gnj8RkuyJfAYmlakRwO/oGkl+vJU21fVUTTPiW3utktNtY0kSZp5Y93CU1WXVNVhVXV34K+Aa4HPAhcleVeSvdtNz6A5vbS6PZ01+HP5Rhzyl8B+Q8sm3U/jQUk+TNNp+kjgXGDfqtqnqt5XVVdv/LOVJEmzZawDz6CqOq2qng/sQnOq627AD5M8GPhf4AfAl5P8TZI7Jtk/yZvb9dP1PuAZSZ6T5K5JXgPcf2ibg4BvAdsDTwZuV1WvqKqfb+JTlCRJs2TcT2mtpe2/81/AfyXZCVhVVZXk0TRXWH2Epi/NZTQh6BMbse/jk9wJOIymT9BXgHcDzxzY7NvAravqmrX3IEmSxtGcCzyDBk9XVdUS4MXtz1TbHgscO7TsZCBDyw4HDh96+JsG1l9y0yuWJEldmDOntCRJkm4qA48kSeo9A48kSeo9A48kSeo9A48kSeo9A48kSeo9A48kSeo9A48kSeo9A48kSeo9A48kSeo9A48kSeo9A48kSeq9OT156Fw2/zfL2P5vftt1GTfavusC5oBVXRcgSbrJbOGRJEm9Z+CRJEm9Z+CRJEm9Z+CRJEm9Z+CRJEm9Z+CRJEm9Z+CRJEm9Z+CRJEm9Z+CRJEm9Z+CRJEm9Z+CRJEm9Z+CRJEm9Z+CRJEm9Z+CRJEm9Z+CRJEm9Z+CRJEm9Z+CRJEm9t6DrAjYnSQ4GDgbYiq07rkaSpM2HLTwjVFVHVdWiqlq0kC27LkeSpM2GgUeSJPWegUeSJPWegUeSJPWegUeSJPWegUeSJPWegUeSJPWegUeSJPWegUeSJPWegUeSJPWegUeSJPWegUeSJPWegUeSJPWegUeSJPWegUeSJPWegUeSJPWegUeSJPVeqqrrGjZLSa4AfjdDu9sRuHKG9jUTrGf9xq0eGL+arGf9rGfDxq0m61m/maznDlV1q+GFBp4eSLK4qhZ1XccE61m/casHxq8m61k/69mwcavJetZvFPV4SkuSJPWegUeSJPWegacfjuq6gCHWs37jVg+MX03Ws37Ws2HjVpP1rN+s12MfHkmS1Hu28EiSpN4z8EiSpN4z8EiSpN4z8EiSpN4z8EiSpN77/3RYqms4QvEMAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Input: <start> this room has windows but is dark inside and packed with antique furniture <end>\n",
      "Predicted translation: the rooms were very creative with beautiful clothing to eat <end> \n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Input: <start> a gravestone with a large tree in the background <end>\n",
      "Predicted translation: i love to live <end> \n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Input: <start> a strange colorful box with gears is being held out for view <end>\n",
      "Predicted translation: i was really happy to the finish line for the race <end> \n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Input: <start> an award is being presented in front of a crowd <end>\n",
      "Predicted translation: the audience listened intently <end> \n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Input: <start> the crop circle had very symmetrical lines and was designed perfectly <end>\n",
      "Predicted translation: the pipe was lit up in the middle of the building <end> \n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Input: <start> a horse is pulling a large cart down the street <end>\n",
      "Predicted translation: i love horses and ready to get this beautiful day <end> \n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "for i in range(20):\n",
    "    random_index = random.randrange(start=0,stop=len(clean_data))\n",
    "    generate(\" \".join(clean_data.loc[random_index,'in_desc'].split(\" \")[1:-1]))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Input: <start> large tree with many outstretching branches and leaves <end>\n",
      "Predicted translation: we were all kinds of poverty and crime <end> \n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "generate('large tree with many outstretching branches and leaves')"
   ]
  },
  {
   "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": "3.6.10"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {},
   "toc_section_display": true,
   "toc_window_display": false
  },
  "varInspector": {
   "cols": {
    "lenName": 16,
    "lenType": 16,
    "lenVar": 40
   },
   "kernels_config": {
    "python": {
     "delete_cmd_postfix": "",
     "delete_cmd_prefix": "del ",
     "library": "var_list.py",
     "varRefreshCmd": "print(var_dic_list())"
    },
    "r": {
     "delete_cmd_postfix": ") ",
     "delete_cmd_prefix": "rm(",
     "library": "var_list.r",
     "varRefreshCmd": "cat(var_dic_list()) "
    }
   },
   "types_to_exclude": [
    "module",
    "function",
    "builtin_function_or_method",
    "instance",
    "_Feature"
   ],
   "window_display": false
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}