02_rnn_spacy_wikitext103.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "from fastai.vision.all import *\n",
    "from fastai.learner import *\n",
    "from fastai.data.all import *\n",
    "from fastai.callback.tracker import SaveModelCallback\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "from pathlib2 import Path\n",
    "import numpy as np\n",
    "import random\n",
    "from torch.nn import MSELoss\n",
    "from IPython.lib.display import Audio\n",
    "from python_speech_features.base import mfcc\n",
    "import librosa"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "df_train = pd.read_csv('data/wikitext-103/train.csv', header=None)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(29539, 1)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "texts = df_train[0].tolist()\n",
    "texts = texts[:2000]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/opt/conda/lib/python3.7/site-packages/sklearn/utils/deprecation.py:143: FutureWarning: The sklearn.datasets.base module is  deprecated in version 0.22 and will be removed in version 0.24. The corresponding classes / functions should instead be imported from sklearn.datasets. Anything that cannot be imported from sklearn.datasets is now part of the private API.\n",
      "  warnings.warn(message, FutureWarning)\n"
     ]
    }
   ],
   "source": [
    "import logging\n",
    "from six import iteritems\n",
    "from web.datasets.similarity import fetch_MEN, fetch_WS353, fetch_SimLex999\n",
    "from web.embeddings import fetch_GloVe\n",
    "from web.evaluate import evaluate_similarity\n",
    "from web.embedding import Embedding, Vocabulary\n",
    "from gensim.models import Word2Vec\n",
    "from gensim.models import KeyedVectors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "37622"
      ]
     },
     "execution_count": 6,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "speech2vec = KeyedVectors.load_word2vec_format('../speech2vec-pretrained-vectors/speech2vec/50.vec', binary=False) \n",
    "\n",
    "vocab = set(speech2vec.vocab.keys())\n",
    "len(vocab)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "# mkdir -p data/audio"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "# from time import sleep"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "# %%time\n",
    "\n",
    "# from google.cloud import texttospeech\n",
    "\n",
    "# client = texttospeech.TextToSpeechClient()\n",
    "\n",
    "# for i, word in enumerate(vocab):\n",
    "#     if i % 300 == 0: sleep(60)\n",
    "\n",
    "#     synthesis_input = texttospeech.SynthesisInput(text=word)\n",
    "\n",
    "#     voice = texttospeech.VoiceSelectionParams(\n",
    "#         language_code=\"en-US\", ssml_gender=texttospeech.SsmlVoiceGender.NEUTRAL\n",
    "#     )\n",
    "\n",
    "#     audio_config = texttospeech.AudioConfig(\n",
    "#         audio_encoding=texttospeech.AudioEncoding.MP3\n",
    "#     )\n",
    "\n",
    "#     response = client.synthesize_speech(\n",
    "#         input=synthesis_input, voice=voice, audio_config=audio_config\n",
    "#     )\n",
    "\n",
    "#     # The response's audio_content is binary.\n",
    "#     with open(f\"data/audio/{word}.mp3\", \"wb\") as out:\n",
    "#         out.write(response.audio_content)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "# %%time\n",
    "\n",
    "# import warnings\n",
    "# warnings.filterwarnings('ignore')\n",
    "\n",
    "# word2ary = {}\n",
    "\n",
    "# for word in vocab:\n",
    "#     ary, _ = librosa.load(f'data/audio/{word}.mp3', sr=16000)\n",
    "#     word2ary[word] = ary\n",
    "    \n",
    "# pd.to_pickle(word2ary, 'word2ary.pkl')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "word2ary = pd.read_pickle('word2ary.pkl')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 7min 14s, sys: 6.64 s, total: 7min 20s\n",
      "Wall time: 1min 13s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "word2features = {}\n",
    "for k, v in word2ary.items():\n",
    "    word2features[k] = mfcc(v)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [],
   "source": [
    "from joblib import Parallel, delayed\n",
    "from functools import partial\n",
    "from spacy.util import minibatch\n",
    "import spacy\n",
    "from spacy.attrs import ORTH, LEMMA\n",
    "\n",
    "nlp = spacy.load(\"en_core_web_sm\")\n",
    "\n",
    "nlp.tokenizer.add_special_case('<unk>', [{ORTH: \"<unk>\"}])\n",
    "nlp.tokenizer.add_special_case('<eos>', [{ORTH: \"<eos>\"}])\n",
    "\n",
    "batch_size= 200\n",
    "partitions = minibatch(texts, size=batch_size)\n",
    "\n",
    "executor = Parallel(n_jobs=10, backend=\"multiprocessing\", prefer=\"threads\")\n",
    "\n",
    "def transform_texts(nlp, vocab, text):\n",
    "    docs = nlp.tokenizer.pipe(texts)\n",
    "    word_pairs = []\n",
    "    \n",
    "    for doc in docs:\n",
    "        words = [token.text.lower() for token in doc if token.is_alpha]\n",
    "        words = [w if w in vocab else '<UNK>' for w in words]\n",
    "        \n",
    "        for i, word in enumerate(words):\n",
    "            if word == '<UNK>': continue\n",
    "            for offset in [-2, -1, 1, 2]:\n",
    "                if i + offset < 0 or i + offset >= len(words): continue\n",
    "                target_word = words[i+offset]\n",
    "                if target_word == '<UNK>': continue\n",
    "                word_pairs.append([word, target_word])\n",
    "                 \n",
    "    return word_pairs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 1min 56s, sys: 17.7 s, total: 2min 13s\n",
      "Wall time: 4min 1s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "do = delayed(partial(transform_texts, nlp, vocab))\n",
    "tasks = (do(batch) for batch in partitions)\n",
    "word_pairs_hierarchical = executor(tasks)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [],
   "source": [
    "word_pairs = []\n",
    "\n",
    "for pairs in word_pairs_hierarchical: word_pairs += pairs"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[['is', 'a'],\n",
       " ['a', 'is'],\n",
       " ['a', 'genus'],\n",
       " ['genus', 'a'],\n",
       " ['genus', 'in'],\n",
       " ['genus', 'the'],\n",
       " ['in', 'genus'],\n",
       " ['in', 'the'],\n",
       " ['in', 'family'],\n",
       " ['the', 'genus']]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "word_pairs[:10]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "184410700"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(word_pairs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "184410700"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(word_pairs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [],
   "source": [
    "vocab = list(vocab)\n",
    "word2index = {w: i for i, w in enumerate(vocab)}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "def prepare_features(word, pad_to=70, pad_left=False):\n",
    "    ary = word2features[word][:pad_to]\n",
    "    example = np.zeros((pad_to, 13))\n",
    "    if pad_left:\n",
    "        example[-ary.shape[0]:, :] = ary\n",
    "    else: example[:ary.shape[0], :] = ary\n",
    "    return example.astype(np.float32)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "dataset_mean = -2\n",
    "dataset_std = 10\n",
    "\n",
    "def normalize_data(ary):\n",
    "    return (ary - dataset_mean) / dataset_std"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "fn2features = {}\n",
    "for word in vocab:\n",
    "    fn2features[word] = normalize_data(prepare_features(word))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Dataset():\n",
    "    def __init__(self, pairs, length=10_000_000):\n",
    "        self.pairs = pairs\n",
    "        self.len = length\n",
    "    def __len__(self):\n",
    "        return self.len\n",
    "    def __getitem__(self, idx):\n",
    "        source_word, target_word = self.pairs[np.random.randint(len(self.pairs))]\n",
    "        return fn2features[source_word], word2index[target_word]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [],
   "source": [
    "# 14 mln -> 180 mln"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "184410700"
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(word_pairs)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_ds = Dataset(word_pairs)\n",
    "valid_ds = Dataset(word_pairs, 1_000_000)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(10000000, 1000000)"
      ]
     },
     "execution_count": 24,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "len(train_ds), len(valid_ds)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [],
   "source": [
    "BS = 2048\n",
    "NUM_WORKERS = 8\n",
    "\n",
    "train_dl = DataLoader(train_ds, BS, NUM_WORKERS, shuffle=True)\n",
    "valid_dl = DataLoader(valid_ds, BS, NUM_WORKERS)\n",
    "\n",
    "dls = DataLoaders(train_dl, valid_dl)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Model(Module):\n",
    "    def __init__(self, hidden_size=25):\n",
    "        self.return_embeddings = False\n",
    "        self.hidden_size = hidden_size\n",
    "        \n",
    "        self.encoder= nn.LSTM(\n",
    "            input_size=13,\n",
    "            hidden_size=hidden_size,\n",
    "            num_layers=3,\n",
    "            batch_first=True,\n",
    "            dropout=0,\n",
    "            bidirectional=True\n",
    "        )\n",
    "        \n",
    "        self.linear = nn.Linear(2*hidden_size, len(vocab))\n",
    "            \n",
    "    def forward(self, source_features):\n",
    "        _, (embeddings, _) = self.encoder(source_features)        \n",
    "        embeddings = torch.cat((embeddings[-1], embeddings[-2]), 1)\n",
    "        \n",
    "        if self.return_embeddings: return embeddings\n",
    "\n",
    "        return self.linear(embeddings)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [],
   "source": [
    "learn = Learner(\n",
    "    dls.cuda(),\n",
    "    Model().cuda(),\n",
    "    loss_func=CrossEntropyLossFlat(),\n",
    "    opt_func=Adam,\n",
    "    metrics=[accuracy]\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "    <div>\n",
       "        <style>\n",
       "            /* Turns off some styling */\n",
       "            progress {\n",
       "                /* gets rid of default border in Firefox and Opera. */\n",
       "                border: none;\n",
       "                /* Needs to be in here for Safari polyfill so background images work as expected. */\n",
       "                background-size: auto;\n",
       "            }\n",
       "            .progress-bar-interrupted, .progress-bar-interrupted::-webkit-progress-bar {\n",
       "                background: #F44336;\n",
       "            }\n",
       "        </style>\n",
       "      <progress value='168' class='' max='200' style='width:300px; height:20px; vertical-align: middle;'></progress>\n",
       "      84.00% [168/200 23:03:53<4:23:35]\n",
       "    </div>\n",
       "    \n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: left;\">\n",
       "      <th>epoch</th>\n",
       "      <th>train_loss</th>\n",
       "      <th>valid_loss</th>\n",
       "      <th>accuracy</th>\n",
       "      <th>time</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <td>0</td>\n",
       "      <td>6.937828</td>\n",
       "      <td>6.943904</td>\n",
       "      <td>0.093364</td>\n",
       "      <td>07:57</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>1</td>\n",
       "      <td>6.894700</td>\n",
       "      <td>6.889863</td>\n",
       "      <td>0.094532</td>\n",
       "      <td>08:04</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>2</td>\n",
       "      <td>6.850228</td>\n",
       "      <td>6.861458</td>\n",
       "      <td>0.095167</td>\n",
       "      <td>06:44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>3</td>\n",
       "      <td>6.839160</td>\n",
       "      <td>6.839796</td>\n",
       "      <td>0.095905</td>\n",
       "      <td>06:21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>4</td>\n",
       "      <td>6.817207</td>\n",
       "      <td>6.825046</td>\n",
       "      <td>0.096110</td>\n",
       "      <td>07:06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>5</td>\n",
       "      <td>6.808103</td>\n",
       "      <td>6.811345</td>\n",
       "      <td>0.096223</td>\n",
       "      <td>06:26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>6</td>\n",
       "      <td>6.796218</td>\n",
       "      <td>6.804134</td>\n",
       "      <td>0.096441</td>\n",
       "      <td>06:45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>7</td>\n",
       "      <td>6.792039</td>\n",
       "      <td>6.788680</td>\n",
       "      <td>0.097343</td>\n",
       "      <td>06:44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>8</td>\n",
       "      <td>6.782446</td>\n",
       "      <td>6.785172</td>\n",
       "      <td>0.096652</td>\n",
       "      <td>06:39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>9</td>\n",
       "      <td>6.780403</td>\n",
       "      <td>6.775949</td>\n",
       "      <td>0.097271</td>\n",
       "      <td>07:29</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>10</td>\n",
       "      <td>6.785173</td>\n",
       "      <td>6.771817</td>\n",
       "      <td>0.097297</td>\n",
       "      <td>06:42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>11</td>\n",
       "      <td>6.775565</td>\n",
       "      <td>6.767179</td>\n",
       "      <td>0.097519</td>\n",
       "      <td>07:05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>12</td>\n",
       "      <td>6.769447</td>\n",
       "      <td>6.761656</td>\n",
       "      <td>0.097597</td>\n",
       "      <td>06:51</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>13</td>\n",
       "      <td>6.761017</td>\n",
       "      <td>6.763788</td>\n",
       "      <td>0.097168</td>\n",
       "      <td>06:49</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>14</td>\n",
       "      <td>6.757686</td>\n",
       "      <td>6.760779</td>\n",
       "      <td>0.097510</td>\n",
       "      <td>07:36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>15</td>\n",
       "      <td>6.757205</td>\n",
       "      <td>6.755786</td>\n",
       "      <td>0.097848</td>\n",
       "      <td>06:48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>16</td>\n",
       "      <td>6.752042</td>\n",
       "      <td>6.751215</td>\n",
       "      <td>0.097841</td>\n",
       "      <td>07:15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>17</td>\n",
       "      <td>6.745518</td>\n",
       "      <td>6.751219</td>\n",
       "      <td>0.097769</td>\n",
       "      <td>06:56</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>18</td>\n",
       "      <td>6.749759</td>\n",
       "      <td>6.749159</td>\n",
       "      <td>0.098032</td>\n",
       "      <td>06:58</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>19</td>\n",
       "      <td>6.761431</td>\n",
       "      <td>6.746167</td>\n",
       "      <td>0.098127</td>\n",
       "      <td>07:29</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>20</td>\n",
       "      <td>6.750870</td>\n",
       "      <td>6.744608</td>\n",
       "      <td>0.098509</td>\n",
       "      <td>06:53</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>21</td>\n",
       "      <td>6.743846</td>\n",
       "      <td>6.746496</td>\n",
       "      <td>0.098127</td>\n",
       "      <td>07:22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>22</td>\n",
       "      <td>6.741595</td>\n",
       "      <td>6.741887</td>\n",
       "      <td>0.097732</td>\n",
       "      <td>06:54</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>23</td>\n",
       "      <td>6.748423</td>\n",
       "      <td>6.747015</td>\n",
       "      <td>0.097708</td>\n",
       "      <td>07:06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>24</td>\n",
       "      <td>6.741353</td>\n",
       "      <td>6.742543</td>\n",
       "      <td>0.097764</td>\n",
       "      <td>07:10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>25</td>\n",
       "      <td>6.748846</td>\n",
       "      <td>6.736854</td>\n",
       "      <td>0.098585</td>\n",
       "      <td>06:52</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>26</td>\n",
       "      <td>6.747440</td>\n",
       "      <td>6.741305</td>\n",
       "      <td>0.097945</td>\n",
       "      <td>07:12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>27</td>\n",
       "      <td>6.738541</td>\n",
       "      <td>6.737718</td>\n",
       "      <td>0.098175</td>\n",
       "      <td>06:51</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>28</td>\n",
       "      <td>6.740782</td>\n",
       "      <td>6.733574</td>\n",
       "      <td>0.098808</td>\n",
       "      <td>07:09</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>29</td>\n",
       "      <td>6.741628</td>\n",
       "      <td>6.732259</td>\n",
       "      <td>0.098071</td>\n",
       "      <td>06:55</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>30</td>\n",
       "      <td>6.733377</td>\n",
       "      <td>6.735437</td>\n",
       "      <td>0.098119</td>\n",
       "      <td>07:07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>31</td>\n",
       "      <td>6.732092</td>\n",
       "      <td>6.730402</td>\n",
       "      <td>0.098323</td>\n",
       "      <td>07:12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>32</td>\n",
       "      <td>6.733955</td>\n",
       "      <td>6.734420</td>\n",
       "      <td>0.097242</td>\n",
       "      <td>07:15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>33</td>\n",
       "      <td>6.732849</td>\n",
       "      <td>6.731400</td>\n",
       "      <td>0.098292</td>\n",
       "      <td>07:17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>34</td>\n",
       "      <td>6.730484</td>\n",
       "      <td>6.728946</td>\n",
       "      <td>0.099009</td>\n",
       "      <td>07:15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>35</td>\n",
       "      <td>6.729080</td>\n",
       "      <td>6.730906</td>\n",
       "      <td>0.098047</td>\n",
       "      <td>07:20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>36</td>\n",
       "      <td>6.725754</td>\n",
       "      <td>6.733629</td>\n",
       "      <td>0.098763</td>\n",
       "      <td>07:17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>37</td>\n",
       "      <td>6.732082</td>\n",
       "      <td>6.725678</td>\n",
       "      <td>0.098345</td>\n",
       "      <td>07:22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>38</td>\n",
       "      <td>6.728032</td>\n",
       "      <td>6.727202</td>\n",
       "      <td>0.098274</td>\n",
       "      <td>07:18</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>39</td>\n",
       "      <td>6.740142</td>\n",
       "      <td>6.728988</td>\n",
       "      <td>0.098373</td>\n",
       "      <td>07:22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>40</td>\n",
       "      <td>6.715055</td>\n",
       "      <td>6.731730</td>\n",
       "      <td>0.098128</td>\n",
       "      <td>07:20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>41</td>\n",
       "      <td>6.727978</td>\n",
       "      <td>6.729724</td>\n",
       "      <td>0.097904</td>\n",
       "      <td>07:23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>42</td>\n",
       "      <td>6.728645</td>\n",
       "      <td>6.724854</td>\n",
       "      <td>0.098411</td>\n",
       "      <td>07:21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>43</td>\n",
       "      <td>6.738086</td>\n",
       "      <td>6.724919</td>\n",
       "      <td>0.098600</td>\n",
       "      <td>07:23</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>44</td>\n",
       "      <td>6.730297</td>\n",
       "      <td>6.736581</td>\n",
       "      <td>0.097721</td>\n",
       "      <td>07:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>45</td>\n",
       "      <td>6.721392</td>\n",
       "      <td>6.722812</td>\n",
       "      <td>0.098354</td>\n",
       "      <td>07:31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>46</td>\n",
       "      <td>6.727945</td>\n",
       "      <td>6.722617</td>\n",
       "      <td>0.097973</td>\n",
       "      <td>07:30</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>47</td>\n",
       "      <td>6.722115</td>\n",
       "      <td>6.725631</td>\n",
       "      <td>0.098671</td>\n",
       "      <td>07:34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>48</td>\n",
       "      <td>6.723235</td>\n",
       "      <td>6.722464</td>\n",
       "      <td>0.098609</td>\n",
       "      <td>07:34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>49</td>\n",
       "      <td>6.711812</td>\n",
       "      <td>6.724281</td>\n",
       "      <td>0.098611</td>\n",
       "      <td>07:38</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>50</td>\n",
       "      <td>6.716583</td>\n",
       "      <td>6.726024</td>\n",
       "      <td>0.098166</td>\n",
       "      <td>07:50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>51</td>\n",
       "      <td>6.715496</td>\n",
       "      <td>6.723745</td>\n",
       "      <td>0.098579</td>\n",
       "      <td>07:49</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>52</td>\n",
       "      <td>6.724834</td>\n",
       "      <td>6.720129</td>\n",
       "      <td>0.098707</td>\n",
       "      <td>07:40</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>53</td>\n",
       "      <td>6.729920</td>\n",
       "      <td>6.719212</td>\n",
       "      <td>0.098216</td>\n",
       "      <td>07:37</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>54</td>\n",
       "      <td>6.720644</td>\n",
       "      <td>6.716640</td>\n",
       "      <td>0.098545</td>\n",
       "      <td>07:34</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>55</td>\n",
       "      <td>6.720489</td>\n",
       "      <td>6.724748</td>\n",
       "      <td>0.098429</td>\n",
       "      <td>07:35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>56</td>\n",
       "      <td>6.720835</td>\n",
       "      <td>6.716567</td>\n",
       "      <td>0.099007</td>\n",
       "      <td>07:38</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>57</td>\n",
       "      <td>6.720689</td>\n",
       "      <td>6.721659</td>\n",
       "      <td>0.098457</td>\n",
       "      <td>07:46</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>58</td>\n",
       "      <td>6.729625</td>\n",
       "      <td>6.722971</td>\n",
       "      <td>0.098560</td>\n",
       "      <td>07:48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>59</td>\n",
       "      <td>6.717187</td>\n",
       "      <td>6.714963</td>\n",
       "      <td>0.098301</td>\n",
       "      <td>07:53</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>60</td>\n",
       "      <td>6.717500</td>\n",
       "      <td>6.721678</td>\n",
       "      <td>0.097995</td>\n",
       "      <td>07:50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>61</td>\n",
       "      <td>6.715535</td>\n",
       "      <td>6.722034</td>\n",
       "      <td>0.098483</td>\n",
       "      <td>07:52</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>62</td>\n",
       "      <td>6.718349</td>\n",
       "      <td>6.719652</td>\n",
       "      <td>0.098488</td>\n",
       "      <td>07:49</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>63</td>\n",
       "      <td>6.724133</td>\n",
       "      <td>6.714582</td>\n",
       "      <td>0.098496</td>\n",
       "      <td>07:53</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>64</td>\n",
       "      <td>6.708770</td>\n",
       "      <td>6.721063</td>\n",
       "      <td>0.098336</td>\n",
       "      <td>07:56</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>65</td>\n",
       "      <td>6.714809</td>\n",
       "      <td>6.716144</td>\n",
       "      <td>0.098190</td>\n",
       "      <td>08:17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>66</td>\n",
       "      <td>6.711950</td>\n",
       "      <td>6.717043</td>\n",
       "      <td>0.098455</td>\n",
       "      <td>09:13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>67</td>\n",
       "      <td>6.723290</td>\n",
       "      <td>6.713805</td>\n",
       "      <td>0.098626</td>\n",
       "      <td>08:42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>68</td>\n",
       "      <td>6.720173</td>\n",
       "      <td>6.717369</td>\n",
       "      <td>0.098197</td>\n",
       "      <td>07:47</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>69</td>\n",
       "      <td>6.715395</td>\n",
       "      <td>6.713726</td>\n",
       "      <td>0.098853</td>\n",
       "      <td>08:24</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>70</td>\n",
       "      <td>6.709816</td>\n",
       "      <td>6.714726</td>\n",
       "      <td>0.098340</td>\n",
       "      <td>07:44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>71</td>\n",
       "      <td>6.705814</td>\n",
       "      <td>6.721614</td>\n",
       "      <td>0.098048</td>\n",
       "      <td>08:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>72</td>\n",
       "      <td>6.718107</td>\n",
       "      <td>6.716190</td>\n",
       "      <td>0.097975</td>\n",
       "      <td>07:49</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>73</td>\n",
       "      <td>6.708908</td>\n",
       "      <td>6.717282</td>\n",
       "      <td>0.098501</td>\n",
       "      <td>08:27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>74</td>\n",
       "      <td>6.710528</td>\n",
       "      <td>6.717884</td>\n",
       "      <td>0.097873</td>\n",
       "      <td>07:45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>75</td>\n",
       "      <td>6.707027</td>\n",
       "      <td>6.715613</td>\n",
       "      <td>0.098031</td>\n",
       "      <td>08:21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>76</td>\n",
       "      <td>6.726900</td>\n",
       "      <td>6.718438</td>\n",
       "      <td>0.098109</td>\n",
       "      <td>07:50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>77</td>\n",
       "      <td>6.729160</td>\n",
       "      <td>6.715364</td>\n",
       "      <td>0.098335</td>\n",
       "      <td>08:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>78</td>\n",
       "      <td>6.721984</td>\n",
       "      <td>6.716440</td>\n",
       "      <td>0.098234</td>\n",
       "      <td>07:51</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>79</td>\n",
       "      <td>6.713536</td>\n",
       "      <td>6.718938</td>\n",
       "      <td>0.097854</td>\n",
       "      <td>08:27</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>80</td>\n",
       "      <td>6.724396</td>\n",
       "      <td>6.716909</td>\n",
       "      <td>0.098707</td>\n",
       "      <td>07:55</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>81</td>\n",
       "      <td>6.713537</td>\n",
       "      <td>6.708681</td>\n",
       "      <td>0.098574</td>\n",
       "      <td>08:38</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>82</td>\n",
       "      <td>6.712824</td>\n",
       "      <td>6.715304</td>\n",
       "      <td>0.098182</td>\n",
       "      <td>08:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>83</td>\n",
       "      <td>6.707284</td>\n",
       "      <td>6.710594</td>\n",
       "      <td>0.098735</td>\n",
       "      <td>08:40</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>84</td>\n",
       "      <td>6.700303</td>\n",
       "      <td>6.713943</td>\n",
       "      <td>0.098837</td>\n",
       "      <td>08:03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>85</td>\n",
       "      <td>6.715702</td>\n",
       "      <td>6.706846</td>\n",
       "      <td>0.098255</td>\n",
       "      <td>08:45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>86</td>\n",
       "      <td>6.722137</td>\n",
       "      <td>6.709610</td>\n",
       "      <td>0.098221</td>\n",
       "      <td>08:06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>87</td>\n",
       "      <td>6.715804</td>\n",
       "      <td>6.712278</td>\n",
       "      <td>0.098021</td>\n",
       "      <td>08:44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>88</td>\n",
       "      <td>6.713221</td>\n",
       "      <td>6.712051</td>\n",
       "      <td>0.098583</td>\n",
       "      <td>08:07</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>89</td>\n",
       "      <td>6.714155</td>\n",
       "      <td>6.712141</td>\n",
       "      <td>0.098479</td>\n",
       "      <td>08:49</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>90</td>\n",
       "      <td>6.718365</td>\n",
       "      <td>6.710437</td>\n",
       "      <td>0.098510</td>\n",
       "      <td>08:12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>91</td>\n",
       "      <td>6.710355</td>\n",
       "      <td>6.710189</td>\n",
       "      <td>0.097914</td>\n",
       "      <td>09:03</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>92</td>\n",
       "      <td>6.713025</td>\n",
       "      <td>6.713628</td>\n",
       "      <td>0.098113</td>\n",
       "      <td>08:29</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>93</td>\n",
       "      <td>6.702594</td>\n",
       "      <td>6.709874</td>\n",
       "      <td>0.098428</td>\n",
       "      <td>08:52</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>94</td>\n",
       "      <td>6.723342</td>\n",
       "      <td>6.711872</td>\n",
       "      <td>0.098423</td>\n",
       "      <td>08:17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>95</td>\n",
       "      <td>6.710260</td>\n",
       "      <td>6.713288</td>\n",
       "      <td>0.097931</td>\n",
       "      <td>08:58</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>96</td>\n",
       "      <td>6.710083</td>\n",
       "      <td>6.714245</td>\n",
       "      <td>0.098114</td>\n",
       "      <td>08:19</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>97</td>\n",
       "      <td>6.722625</td>\n",
       "      <td>6.716434</td>\n",
       "      <td>0.097782</td>\n",
       "      <td>08:54</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>98</td>\n",
       "      <td>6.713799</td>\n",
       "      <td>6.714789</td>\n",
       "      <td>0.098272</td>\n",
       "      <td>08:11</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>99</td>\n",
       "      <td>6.715990</td>\n",
       "      <td>6.711741</td>\n",
       "      <td>0.098384</td>\n",
       "      <td>08:54</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>100</td>\n",
       "      <td>6.702667</td>\n",
       "      <td>6.704270</td>\n",
       "      <td>0.098850</td>\n",
       "      <td>08:10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>101</td>\n",
       "      <td>6.703899</td>\n",
       "      <td>6.706736</td>\n",
       "      <td>0.099257</td>\n",
       "      <td>08:55</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>102</td>\n",
       "      <td>6.712865</td>\n",
       "      <td>6.711374</td>\n",
       "      <td>0.098858</td>\n",
       "      <td>08:12</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>103</td>\n",
       "      <td>6.723797</td>\n",
       "      <td>6.709304</td>\n",
       "      <td>0.098836</td>\n",
       "      <td>08:48</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>104</td>\n",
       "      <td>6.710118</td>\n",
       "      <td>6.711255</td>\n",
       "      <td>0.098983</td>\n",
       "      <td>08:10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>105</td>\n",
       "      <td>6.709251</td>\n",
       "      <td>6.704594</td>\n",
       "      <td>0.099045</td>\n",
       "      <td>08:53</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>106</td>\n",
       "      <td>6.710431</td>\n",
       "      <td>6.717253</td>\n",
       "      <td>0.097881</td>\n",
       "      <td>08:42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>107</td>\n",
       "      <td>6.710737</td>\n",
       "      <td>6.705627</td>\n",
       "      <td>0.098621</td>\n",
       "      <td>08:45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>108</td>\n",
       "      <td>6.720900</td>\n",
       "      <td>6.710648</td>\n",
       "      <td>0.099113</td>\n",
       "      <td>08:42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>109</td>\n",
       "      <td>6.716515</td>\n",
       "      <td>6.708785</td>\n",
       "      <td>0.098583</td>\n",
       "      <td>08:21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>110</td>\n",
       "      <td>6.705203</td>\n",
       "      <td>6.711572</td>\n",
       "      <td>0.098429</td>\n",
       "      <td>08:54</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>111</td>\n",
       "      <td>6.703488</td>\n",
       "      <td>6.705663</td>\n",
       "      <td>0.098846</td>\n",
       "      <td>08:13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>112</td>\n",
       "      <td>6.707486</td>\n",
       "      <td>6.709652</td>\n",
       "      <td>0.098426</td>\n",
       "      <td>08:55</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>113</td>\n",
       "      <td>6.693609</td>\n",
       "      <td>6.708553</td>\n",
       "      <td>0.098128</td>\n",
       "      <td>08:20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>114</td>\n",
       "      <td>6.710360</td>\n",
       "      <td>6.712616</td>\n",
       "      <td>0.098549</td>\n",
       "      <td>08:45</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>115</td>\n",
       "      <td>6.708791</td>\n",
       "      <td>6.709264</td>\n",
       "      <td>0.098025</td>\n",
       "      <td>08:32</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>116</td>\n",
       "      <td>6.716254</td>\n",
       "      <td>6.702915</td>\n",
       "      <td>0.098735</td>\n",
       "      <td>08:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>117</td>\n",
       "      <td>6.693410</td>\n",
       "      <td>6.713050</td>\n",
       "      <td>0.098210</td>\n",
       "      <td>08:49</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>118</td>\n",
       "      <td>6.712681</td>\n",
       "      <td>6.716629</td>\n",
       "      <td>0.098218</td>\n",
       "      <td>08:21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>119</td>\n",
       "      <td>6.715830</td>\n",
       "      <td>6.710740</td>\n",
       "      <td>0.098457</td>\n",
       "      <td>08:53</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>120</td>\n",
       "      <td>6.698483</td>\n",
       "      <td>6.715737</td>\n",
       "      <td>0.097007</td>\n",
       "      <td>09:06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>121</td>\n",
       "      <td>6.706649</td>\n",
       "      <td>6.703876</td>\n",
       "      <td>0.098667</td>\n",
       "      <td>09:16</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>122</td>\n",
       "      <td>6.704764</td>\n",
       "      <td>6.708820</td>\n",
       "      <td>0.098451</td>\n",
       "      <td>08:55</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>123</td>\n",
       "      <td>6.709310</td>\n",
       "      <td>6.711109</td>\n",
       "      <td>0.098500</td>\n",
       "      <td>08:31</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>124</td>\n",
       "      <td>6.711231</td>\n",
       "      <td>6.707022</td>\n",
       "      <td>0.098746</td>\n",
       "      <td>08:41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>125</td>\n",
       "      <td>6.705013</td>\n",
       "      <td>6.703653</td>\n",
       "      <td>0.098851</td>\n",
       "      <td>08:43</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>126</td>\n",
       "      <td>6.710001</td>\n",
       "      <td>6.709577</td>\n",
       "      <td>0.098740</td>\n",
       "      <td>08:29</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>127</td>\n",
       "      <td>6.704128</td>\n",
       "      <td>6.710823</td>\n",
       "      <td>0.098886</td>\n",
       "      <td>08:49</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>128</td>\n",
       "      <td>6.706517</td>\n",
       "      <td>6.709991</td>\n",
       "      <td>0.098331</td>\n",
       "      <td>08:36</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>129</td>\n",
       "      <td>6.702250</td>\n",
       "      <td>6.708593</td>\n",
       "      <td>0.098632</td>\n",
       "      <td>08:42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>130</td>\n",
       "      <td>6.702076</td>\n",
       "      <td>6.706769</td>\n",
       "      <td>0.098263</td>\n",
       "      <td>08:49</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>131</td>\n",
       "      <td>6.690513</td>\n",
       "      <td>6.711047</td>\n",
       "      <td>0.098531</td>\n",
       "      <td>08:40</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>132</td>\n",
       "      <td>6.703617</td>\n",
       "      <td>6.706165</td>\n",
       "      <td>0.098369</td>\n",
       "      <td>08:50</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>133</td>\n",
       "      <td>6.707214</td>\n",
       "      <td>6.707808</td>\n",
       "      <td>0.098506</td>\n",
       "      <td>08:43</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>134</td>\n",
       "      <td>6.714911</td>\n",
       "      <td>6.715313</td>\n",
       "      <td>0.098091</td>\n",
       "      <td>08:39</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>135</td>\n",
       "      <td>6.710434</td>\n",
       "      <td>6.710309</td>\n",
       "      <td>0.098510</td>\n",
       "      <td>09:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>136</td>\n",
       "      <td>6.706131</td>\n",
       "      <td>6.698198</td>\n",
       "      <td>0.099599</td>\n",
       "      <td>09:35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>137</td>\n",
       "      <td>6.713230</td>\n",
       "      <td>6.709128</td>\n",
       "      <td>0.098660</td>\n",
       "      <td>09:13</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>138</td>\n",
       "      <td>6.713501</td>\n",
       "      <td>6.710007</td>\n",
       "      <td>0.098031</td>\n",
       "      <td>08:43</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>139</td>\n",
       "      <td>6.705344</td>\n",
       "      <td>6.708174</td>\n",
       "      <td>0.099030</td>\n",
       "      <td>08:54</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>140</td>\n",
       "      <td>6.713291</td>\n",
       "      <td>6.708629</td>\n",
       "      <td>0.098884</td>\n",
       "      <td>08:35</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>141</td>\n",
       "      <td>6.706723</td>\n",
       "      <td>6.704213</td>\n",
       "      <td>0.098301</td>\n",
       "      <td>08:49</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>142</td>\n",
       "      <td>6.708839</td>\n",
       "      <td>6.703331</td>\n",
       "      <td>0.098997</td>\n",
       "      <td>08:55</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>143</td>\n",
       "      <td>6.701344</td>\n",
       "      <td>6.704261</td>\n",
       "      <td>0.098715</td>\n",
       "      <td>08:42</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>144</td>\n",
       "      <td>6.707780</td>\n",
       "      <td>6.711459</td>\n",
       "      <td>0.098278</td>\n",
       "      <td>08:54</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>145</td>\n",
       "      <td>6.701902</td>\n",
       "      <td>6.710982</td>\n",
       "      <td>0.098214</td>\n",
       "      <td>08:53</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>146</td>\n",
       "      <td>6.705715</td>\n",
       "      <td>6.707316</td>\n",
       "      <td>0.098801</td>\n",
       "      <td>08:44</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>147</td>\n",
       "      <td>6.699770</td>\n",
       "      <td>6.699169</td>\n",
       "      <td>0.098598</td>\n",
       "      <td>08:57</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>148</td>\n",
       "      <td>6.708000</td>\n",
       "      <td>6.707365</td>\n",
       "      <td>0.098636</td>\n",
       "      <td>09:10</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>149</td>\n",
       "      <td>6.703630</td>\n",
       "      <td>6.706052</td>\n",
       "      <td>0.098726</td>\n",
       "      <td>08:54</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>150</td>\n",
       "      <td>6.699873</td>\n",
       "      <td>6.706533</td>\n",
       "      <td>0.098568</td>\n",
       "      <td>09:06</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>151</td>\n",
       "      <td>6.714195</td>\n",
       "      <td>6.705225</td>\n",
       "      <td>0.098793</td>\n",
       "      <td>09:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>152</td>\n",
       "      <td>6.697634</td>\n",
       "      <td>6.715111</td>\n",
       "      <td>0.098403</td>\n",
       "      <td>08:56</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>153</td>\n",
       "      <td>6.711045</td>\n",
       "      <td>6.707883</td>\n",
       "      <td>0.098143</td>\n",
       "      <td>09:05</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>154</td>\n",
       "      <td>6.699325</td>\n",
       "      <td>6.699723</td>\n",
       "      <td>0.098391</td>\n",
       "      <td>09:26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>155</td>\n",
       "      <td>6.709980</td>\n",
       "      <td>6.702397</td>\n",
       "      <td>0.098173</td>\n",
       "      <td>12:56</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>156</td>\n",
       "      <td>6.707081</td>\n",
       "      <td>6.704415</td>\n",
       "      <td>0.099135</td>\n",
       "      <td>14:29</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>157</td>\n",
       "      <td>6.690387</td>\n",
       "      <td>6.710542</td>\n",
       "      <td>0.098785</td>\n",
       "      <td>09:51</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>158</td>\n",
       "      <td>6.705828</td>\n",
       "      <td>6.702172</td>\n",
       "      <td>0.098653</td>\n",
       "      <td>08:41</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>159</td>\n",
       "      <td>6.713968</td>\n",
       "      <td>6.700732</td>\n",
       "      <td>0.098408</td>\n",
       "      <td>09:28</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>160</td>\n",
       "      <td>6.707677</td>\n",
       "      <td>6.702272</td>\n",
       "      <td>0.098552</td>\n",
       "      <td>09:00</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>161</td>\n",
       "      <td>6.700186</td>\n",
       "      <td>6.706286</td>\n",
       "      <td>0.098348</td>\n",
       "      <td>08:52</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>162</td>\n",
       "      <td>6.694974</td>\n",
       "      <td>6.704957</td>\n",
       "      <td>0.098381</td>\n",
       "      <td>09:26</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>163</td>\n",
       "      <td>6.706378</td>\n",
       "      <td>6.705928</td>\n",
       "      <td>0.097973</td>\n",
       "      <td>09:15</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>164</td>\n",
       "      <td>6.703895</td>\n",
       "      <td>6.706660</td>\n",
       "      <td>0.098647</td>\n",
       "      <td>09:01</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>165</td>\n",
       "      <td>6.710333</td>\n",
       "      <td>6.701441</td>\n",
       "      <td>0.098878</td>\n",
       "      <td>09:21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>166</td>\n",
       "      <td>6.698319</td>\n",
       "      <td>6.700220</td>\n",
       "      <td>0.098666</td>\n",
       "      <td>09:17</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <td>167</td>\n",
       "      <td>6.708936</td>\n",
       "      <td>6.706348</td>\n",
       "      <td>0.098530</td>\n",
       "      <td>09:20</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table><p>\n",
       "\n",
       "    <div>\n",
       "        <style>\n",
       "            /* Turns off some styling */\n",
       "            progress {\n",
       "                /* gets rid of default border in Firefox and Opera. */\n",
       "                border: none;\n",
       "                /* Needs to be in here for Safari polyfill so background images work as expected. */\n",
       "                background-size: auto;\n",
       "            }\n",
       "            .progress-bar-interrupted, .progress-bar-interrupted::-webkit-progress-bar {\n",
       "                background: #F44336;\n",
       "            }\n",
       "        </style>\n",
       "      <progress value='4517' class='' max='4883' style='width:300px; height:20px; vertical-align: middle;'></progress>\n",
       "      92.50% [4517/4883 05:14<00:25 6.6999]\n",
       "    </div>\n",
       "    "
      ],
      "text/plain": [
       "<IPython.core.display.HTML object>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n",
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n",
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n",
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n",
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n",
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n",
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n",
      "IOPub message rate exceeded.\n",
      "The notebook server will temporarily stop sending output\n",
      "to the client in order to avoid crashing it.\n",
      "To change this limit, set the config variable\n",
      "`--NotebookApp.iopub_msg_rate_limit`.\n",
      "\n",
      "Current values:\n",
      "NotebookApp.iopub_msg_rate_limit=1000.0 (msgs/sec)\n",
      "NotebookApp.rate_limit_window=3.0 (secs)\n",
      "\n"
     ]
    }
   ],
   "source": [
    "NUM_EPOCHS = 200\n",
    "learn.fit(NUM_EPOCHS, lr=1e-3, cbs=SaveModelCallback(fname='rnn_encoder', every_epoch=True))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Evalute embeddings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [],
   "source": [
    "from utils import Embeddings"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "class DatasetPred():\n",
    "    def __len__(self):\n",
    "        return len(vocab)\n",
    "    def __getitem__(self, idx):\n",
    "        _, target_word = word_pairs[np.random.randint(len(word_pairs))]\n",
    "        source_word = vocab[idx]\n",
    "        return fn2features[source_word], word2index[target_word]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [],
   "source": [
    "pred_dl = DataLoader(DatasetPred(), BS, NUM_WORKERS)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<fastai.learner.Learner at 0x7ef770551e90>"
      ]
     },
     "execution_count": 31,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "learn.load('rnn_encoder_199')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 224 ms, sys: 3.68 s, total: 3.9 s\n",
      "Wall time: 4.87 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "learn.model.return_embeddings = True\n",
    "learn.model.eval()\n",
    "\n",
    "all_embeddings = []\n",
    "\n",
    "with torch.no_grad():    \n",
    "    for batch in pred_dl:\n",
    "        embeddings = learn.model(batch[0].cuda())\n",
    "        all_embeddings.append(embeddings.detach().cpu().squeeze(0))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [],
   "source": [
    "embeddings = torch.cat(all_embeddings)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [],
   "source": [
    "e = Embeddings(\n",
    "    embeddings,\n",
    "    vocab\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "for w in ['fast', 'lost', 'small', 'true', 'crazy', 'slow']:\n",
    "    print(f'{w}: {e.nn_words_to(e[w])}')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Evaluating embeddings using [word-embeddings-benchmarks](https://github.com/kudkudak/word-embeddings-benchmarks)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "import logging\n",
    "from six import iteritems\n",
    "from web.datasets.similarity import fetch_MEN, fetch_WS353, fetch_SimLex999\n",
    "from web.embeddings import fetch_GloVe\n",
    "from web.evaluate import evaluate_similarity\n",
    "from web.embedding import Embedding, Vocabulary\n",
    "from gensim.models import Word2Vec\n",
    "from gensim.models import KeyedVectors"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [],
   "source": [
    "tasks = {\n",
    "    \"MEN\": fetch_MEN(),\n",
    "    \"WS353\": fetch_WS353(),\n",
    "    \"SIMLEX999\": fetch_SimLex999()\n",
    "}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "our_embeddings = Embedding(\n",
    "    Vocabulary(vocab),\n",
    "    embeddings\n",
    ")\n",
    "\n",
    "speech2vec = KeyedVectors.load_word2vec_format('../speech2vec-pretrained-vectors/word2vec/50.vec', binary=False) \n",
    "speech2vec_embeddings = Embedding(Vocabulary(list(speech2vec.vocab.keys())), speech2vec.vectors)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Missing 392 words. Will replace them with mean vector\n",
      "/opt/conda/lib/python3.7/site-packages/web-0.0.1-py3.7.egg/web/evaluate.py:336: FutureWarning: arrays to stack must be passed as a \"sequence\" type such as list or tuple. Support for non-sequence iterables such as generators is deprecated as of NumPy 1.16 and will raise an error in the future.\n",
      "  A = np.vstack(w.get(word, mean_vector) for word in X[:, 0])\n",
      "/opt/conda/lib/python3.7/site-packages/web-0.0.1-py3.7.egg/web/evaluate.py:337: FutureWarning: arrays to stack must be passed as a \"sequence\" type such as list or tuple. Support for non-sequence iterables such as generators is deprecated as of NumPy 1.16 and will raise an error in the future.\n",
      "  B = np.vstack(w.get(word, mean_vector) for word in X[:, 1])\n",
      "Missing 61 words. Will replace them with mean vector\n",
      "Missing 24 words. Will replace them with mean vector\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Spearman correlation of scores on MEN 0.13716165687788728\n",
      "Spearman correlation of scores on WS353 0.11457108430299541\n",
      "Spearman correlation of scores on SIMLEX999 0.05293036082814397\n"
     ]
    }
   ],
   "source": [
    "for name, data in iteritems(tasks):\n",
    "    print(\"Spearman correlation of scores on {} {}\".format(name, evaluate_similarity(our_embeddings, data.X, data.y)))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Missing 392 words. Will replace them with mean vector\n",
      "Missing 61 words. Will replace them with mean vector\n",
      "Missing 24 words. Will replace them with mean vector\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Spearman correlation of scores on MEN 0.6056592803599269\n",
      "Spearman correlation of scores on WS353 0.43349390636024643\n",
      "Spearman correlation of scores on SIMLEX999 0.25938770901422736\n"
     ]
    }
   ],
   "source": [
    "for name, data in iteritems(tasks):\n",
    "    print(\"Spearman correlation of scores on {} {}\".format(name, evaluate_similarity(speech2vec_embeddings, data.X, data.y)))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Loss decrease and improvements on semantic tasks as training progresses"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 46,
   "metadata": {},
   "outputs": [],
   "source": [
    "%%capture\n",
    "\n",
    "val_losses, accuracies, task_perf = [], [], []\n",
    "for i in range(NUM_EPOCHS):\n",
    "    learn.load(f'rnn_encoder_{i}')\n",
    "#     loss, accuracy = learn.validate()\n",
    "#     val_losses.append(loss)\n",
    "#     accuracies.append(accuracy)\n",
    "    \n",
    "    learn.model.return_embeddings = True\n",
    "    learn.model.eval()\n",
    "\n",
    "    all_embeddings = []\n",
    "\n",
    "    with torch.no_grad():    \n",
    "        for batch in pred_dl:\n",
    "            embeddings = learn.model(batch[0].cuda())\n",
    "            all_embeddings.append(embeddings.detach().cpu().squeeze(0))\n",
    "\n",
    "    embeddings = torch.cat(all_embeddings)\n",
    "    our_embeddings = Embedding(\n",
    "        Vocabulary([w.lower() for w in vocab]),\n",
    "        embeddings\n",
    "    )\n",
    "\n",
    "    task_perf.append([evaluate_similarity(our_embeddings, data.X, data.y) for name, data in iteritems(tasks)])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 47,
   "metadata": {},
   "outputs": [],
   "source": [
    "men, ws353, simlex999 = list(zip(*task_perf))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {},
   "outputs": [],
   "source": [
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "No handles with labels found to put in legend.\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig, ax1 = plt.subplots()\n",
    "\n",
    "# ax1.plot(val_losses, label='val loss')\n",
    "\n",
    "ax2 = ax1.twinx()\n",
    "ax2.plot(men, label='MEN', c='g')\n",
    "ax2.plot(ws353, label='WS353', c='m')\n",
    "ax2.plot(simlex999, label='SIMLEX999', c='y')\n",
    "\n",
    "ax1.legend(loc=[0.07, 0.9])\n",
    "ax2.legend(loc=[0.7, 0.15])\n",
    "\n",
    "ax1.set_xlabel('epochs');"
   ]
  }
 ],
 "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": "3.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}