src/suomi/finnish-ending-pos-bilstm-case-tagger.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import pandas as pd\n",
    "\n",
    "import seaborn as sns\n",
    "\n",
    "import itertools\n",
    "import pickle\n",
    "\n",
    "from keras.utils import to_categorical\n",
    "from keras.preprocessing.text import Tokenizer\n",
    "from keras.preprocessing.sequence import pad_sequences\n",
    "from keras.models import Model\n",
    "from keras.layers import Dense, LSTM, Input, Embedding, Bidirectional\n",
    "\n",
    "from sklearn.metrics import classification_report, confusion_matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "DATASET_PATH = '../../data/ftb1u-v1/sentences'"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load train data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(f'{DATASET_PATH}/case_train_pos.pkl', 'rb') as f:\n",
    "    train = pickle.load(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[[('Tammikuussa|NOUN', 'Ine'),\n",
       "  ('hän|PRON', 'Nom'),\n",
       "  ('muutti|VERB', 'Unk'),\n",
       "  ('Ilkan|PROPN', 'Gen'),\n",
       "  ('luokse|ADP', 'Unk'),\n",
       "  ('Saviselkään|PROPN', 'Ill'),\n",
       "  (',|PUNCT', 'Unk'),\n",
       "  ('kevättalvi|NOUN', 'Nom'),\n",
       "  ('koeteltiin|VERB', 'Unk'),\n",
       "  ('toinen|PRON', 'Nom'),\n",
       "  ('toisensa|DET', 'Gen'),\n",
       "  ('luonnetta|NOUN', 'Par'),\n",
       "  ('ja|CONJ', 'Unk'),\n",
       "  ('kesällä|NOUN', 'Ade'),\n",
       "  ('mentiin|VERB', 'Unk'),\n",
       "  ('naimisiin|ADV', 'Unk'),\n",
       "  ('.|PUNCT', 'Unk')],\n",
       " [('rakkaat|ADJ', 'Nom'),\n",
       "  ('isä|NOUN', 'Nom'),\n",
       "  ('ja|CONJ', 'Unk'),\n",
       "  ('äiti|NOUN', 'Nom'),\n",
       "  ('.|PUNCT', 'Unk')],\n",
       " [('Pitipä|VERB', 'Unk'), ('sattua|VERB', 'Lat'), ('.|PUNCT', 'Unk')],\n",
       " [('Nyt|ADV', 'Unk'), ('säästäminen|NOUN', 'Nom'), ('kannattaa|VERB', 'Unk')],\n",
       " [('Monet|PRON', 'Nom'),\n",
       "  ('eivät|VERB', 'Unk'),\n",
       "  ('vielä|ADV', 'Unk'),\n",
       "  ('ole|VERB', 'Unk'),\n",
       "  ('epäonnekseen|NOUN', 'Tra'),\n",
       "  ('saaneet|VERB', 'Nom'),\n",
       "  ('työpaikkaa|NOUN', 'Par'),\n",
       "  ('.|PUNCT', 'Unk')]]"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "72"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "MAX_LENGTH = len(max(train, key=len))\n",
    "MAX_LENGTH"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 86.8 ms, sys: 567 µs, total: 87.4 ms\n",
      "Wall time: 83.4 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "train_sentences, train_sentence_tags = [], [] \n",
    "for tagged_sentence in train:\n",
    "    sentence, tags = zip(*tagged_sentence)\n",
    "    train_sentences.append(np.array(sentence))\n",
    "    train_sentence_tags.append(np.array(tags))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Encode data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "suff_len = 4\n",
    "words, tags = set([]), set([])\n",
    " \n",
    "for s in train_sentences:\n",
    "    for w in s:\n",
    "        w = w.lower()\n",
    "        word, pos = w.rsplit('|', 1)\n",
    "        ending = word[-suff_len:] if len(word) >= suff_len else word\n",
    "        words.add(ending + f'|{pos}')\n",
    "        words.add(f'-OOV-|{pos}')\n",
    "\n",
    "for ts in train_sentence_tags:\n",
    "    for t in ts:\n",
    "        tags.add(t)\n",
    "\n",
    "word2index = {w: i + 2 for i, w in enumerate(list(words))}\n",
    "word2index['-OOV-'] = 1\n",
    "word2index['-PAD-'] = 0\n",
    " \n",
    "tag2index = {t: i + 1 for i, t in enumerate(list(tags))}\n",
    "tag2index['-PAD-'] = 0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def encode_sentences(sentences, sentence_tags, suff_len=3):\n",
    "    sentences_X, tags_y = [], []\n",
    "\n",
    "    for s in sentences:\n",
    "        s_int = []\n",
    "        for t in s:\n",
    "            t = t.lower()\n",
    "            w, p = t.rsplit('|', 1)\n",
    "            ending = w[-suff_len:] if len(w) >= suff_len else w\n",
    "            end = f'{ending}|{p}'\n",
    "            if end not in word2index:\n",
    "                end = f'-OOV-|{p}'\n",
    "            s_int.append(word2index.get(end, word2index['-OOV-']))\n",
    "\n",
    "        sentences_X.append(s_int)\n",
    "\n",
    "    for s in sentence_tags:\n",
    "        tags_y.append([tag2index[t] for t in s])\n",
    "    return sentences_X, tags_y"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 233 ms, sys: 0 ns, total: 233 ms\n",
      "Wall time: 232 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "train_sentences_X, train_tags_y = encode_sentences(train_sentences, train_sentence_tags, suff_len=suff_len)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Load test data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "with open(f'{DATASET_PATH}/case_test_pos.pkl', 'rb') as f:\n",
    "    test = pickle.load(f)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[[('Bajul|PROPN', 'Nom'),\n",
       "  ('ei|VERB', 'Unk'),\n",
       "  ('kauan|ADV', 'Unk'),\n",
       "  ('viihdy|VERB', 'Unk'),\n",
       "  ('ammattilaisena|NOUN', 'Ess'),\n",
       "  ('.|PUNCT', 'Unk')],\n",
       " [('mennä|VERB', 'Lat'), ('päivälliselle|NOUN', 'All')],\n",
       " [('Pihassamme|NOUN', 'Ine'),\n",
       "  ('kasvaa|VERB', 'Unk'),\n",
       "  ('koivun|NOUN', 'Gen'),\n",
       "  ('juurella|ADP', 'Unk'),\n",
       "  ('sitkeä|ADJ', 'Nom'),\n",
       "  (',|PUNCT', 'Unk'),\n",
       "  ('ikivihreä|ADJ', 'Nom'),\n",
       "  ('kasvi|NOUN', 'Nom'),\n",
       "  ('.|PUNCT', 'Unk')],\n",
       " [('En|VERB', 'Unk'),\n",
       "  ('oo|VERB', 'Unk'),\n",
       "  ('itte|PRON', 'Nom'),\n",
       "  ('ainakaan|ADV', 'Unk'),\n",
       "  ('vielä|ADV', 'Unk'),\n",
       "  ('harrastanut|VERB', 'Nom'),\n",
       "  ('tota|DET', 'Par'),\n",
       "  ('yöllä|NOUN', 'Ade'),\n",
       "  ('syömistä|NOUN', 'Par'),\n",
       "  ('.|PUNCT', 'Unk')],\n",
       " [('Työttömyyteen|NOUN', 'Ill'),\n",
       "  ('on|VERB', 'Unk'),\n",
       "  ('käytävä|VERB', 'Nom'),\n",
       "  ('käsiksi|ADV', 'Unk'),\n",
       "  ('kaikella|DET', 'Ade'),\n",
       "  ('sillä|DET', 'Ade'),\n",
       "  ('tarmolla|NOUN', 'Ade'),\n",
       "  (',|PUNCT', 'Unk'),\n",
       "  ('joka|PRON', 'Nom'),\n",
       "  ('meistä|DET', 'Ela'),\n",
       "  ('suomalaisista|ADJ', 'Ela'),\n",
       "  ('löytyy|VERB', 'Unk'),\n",
       "  ('.|PUNCT', 'Unk')]]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test[:5]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 34.6 ms, sys: 3.72 ms, total: 38.3 ms\n",
      "Wall time: 35.3 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "test_sentences, test_sentence_tags = [], [] \n",
    "for tagged_sentence in test:\n",
    "    sentence, tags = zip(*tagged_sentence)\n",
    "    test_sentences.append(np.array(sentence))\n",
    "    test_sentence_tags.append(np.array(tags))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 62.1 ms, sys: 0 ns, total: 62.1 ms\n",
      "Wall time: 61.7 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "test_sentences_X, test_tags_y = encode_sentences(test_sentences, test_sentence_tags, suff_len=suff_len)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Padding"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [],
   "source": [
    "train_sentences_X = pad_sequences(train_sentences_X, maxlen=MAX_LENGTH, padding='post')\n",
    "train_tags_y = pad_sequences(train_tags_y, maxlen=MAX_LENGTH, padding='post')\n",
    "test_sentences_X = pad_sequences(test_sentences_X, maxlen=MAX_LENGTH, padding='post')\n",
    "test_tags_y = pad_sequences(test_tags_y, maxlen=MAX_LENGTH, padding='post')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((15277, 72), (15277, 72))"
      ]
     },
     "execution_count": 15,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "train_sentences_X.shape, train_tags_y.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "((3820, 72), (3820, 72))"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "test_sentences_X.shape, test_tags_y.shape"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Define model and evaluation functions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "from keras import backend as K\n",
    "\n",
    "def recall_m(y_true, y_pred):\n",
    "        true_positives = K.sum(K.round(K.clip(y_true * y_pred, 0, 1)))\n",
    "        possible_positives = K.sum(K.round(K.clip(y_true, 0, 1)))\n",
    "        recall = true_positives / (possible_positives + K.epsilon())\n",
    "        return recall\n",
    "\n",
    "def precision_m(y_true, y_pred):\n",
    "        true_positives = K.sum(K.round(K.clip(y_true * y_pred, 0, 1)))\n",
    "        predicted_positives = K.sum(K.round(K.clip(y_pred, 0, 1)))\n",
    "        precision = true_positives / (predicted_positives + K.epsilon())\n",
    "        return precision\n",
    "\n",
    "def f1_m(y_true, y_pred):\n",
    "    precision = precision_m(y_true, y_pred)\n",
    "    recall = recall_m(y_true, y_pred)\n",
    "    return 2*((precision*recall)/(precision+recall+K.epsilon()))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [],
   "source": [
    "batch_size = 64\n",
    "\n",
    "vocab_size = len(word2index)\n",
    "classes = len(tag2index)\n",
    "\n",
    "dropout = 0.2\n",
    "lstm_recurrent_dropout = 0.2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Word Input\n",
    "word_feats_in = Input(shape=(MAX_LENGTH,), name='word_ending_input')\n",
    "\n",
    "# Word Embedding Using Thai2Fit\n",
    "word_feats_embeddings = Embedding(input_dim=vocab_size,\n",
    "                            output_dim=50,\n",
    "                            input_length=MAX_LENGTH,\n",
    "                            mask_zero=False,\n",
    "                            name='word_ending_embedding')(word_feats_in)\n",
    "word_feats_enc = Bidirectional(LSTM(units=50, return_sequences=True, \n",
    "                dropout=dropout, recurrent_dropout=lstm_recurrent_dropout,\n",
    "                name='word_ending_encoder'))(word_feats_embeddings)\n",
    "\n",
    "# CRF\n",
    "out = Dense(classes, activation='softmax', name='case_decoder')(word_feats_enc)\n",
    "\n",
    "# Model\n",
    "model = Model(word_feats_in, out)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [],
   "source": [
    "model.compile(loss='categorical_crossentropy',\n",
    "              optimizer='adam',\n",
    "              metrics=['accuracy', precision_m, recall_m, f1_m])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"model_1\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "word_ending_input (InputLaye (None, 72)                0         \n",
      "_________________________________________________________________\n",
      "word_ending_embedding (Embed (None, 72, 50)            474600    \n",
      "_________________________________________________________________\n",
      "bidirectional_1 (Bidirection (None, 72, 100)           40400     \n",
      "_________________________________________________________________\n",
      "case_decoder (Dense)         (None, 72, 18)            1818      \n",
      "=================================================================\n",
      "Total params: 516,818\n",
      "Trainable params: 516,818\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n"
     ]
    }
   ],
   "source": [
    "model.summary()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Fit model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "history = model.fit(train_sentences_X, to_categorical(train_tags_y, len(tag2index)),\n",
    "                    epochs=10, verbose=1,\n",
    "                    validation_data=(test_sentences_X, to_categorical(test_tags_y, len(tag2index))))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Evaluate train model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "15277/15277 [==============================] - 3s 200us/step\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(0.004518307117757351,\n",
       " 0.9986390471458435,\n",
       " 0.9989011883735657,\n",
       " 0.9984021186828613,\n",
       " 0.9986503720283508)"
      ]
     },
     "execution_count": 23,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scores, acc, p, r, f = model.evaluate(train_sentences_X, to_categorical(train_tags_y, len(tag2index)))\n",
    "scores, acc, p, r, f"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "#### Plot scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "from matplotlib import pyplot as plt"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history['accuracy'])\n",
    "plt.plot(history.history['val_accuracy'])\n",
    "plt.title('model accuracy')\n",
    "plt.ylabel('accuracy')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'val'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history['precision_m'])\n",
    "plt.plot(history.history['val_precision_m'])\n",
    "plt.title('model precision')\n",
    "plt.ylabel('precision')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'val'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history['recall_m'])\n",
    "plt.plot(history.history['val_recall_m'])\n",
    "plt.title('model recall')\n",
    "plt.ylabel('recall')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'val'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history['f1_m'])\n",
    "plt.plot(history.history['val_f1_m'])\n",
    "plt.title('model F-score')\n",
    "plt.ylabel('F-score')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'val'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.plot(history.history['loss'])\n",
    "plt.plot(history.history['val_loss'])\n",
    "plt.title('model loss')\n",
    "plt.ylabel('loss')\n",
    "plt.xlabel('epoch')\n",
    "plt.legend(['train', 'val'], loc='upper left')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Evaluate test dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3820/3820 [==============================] - 1s 201us/step\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "(0.018404287302681288,\n",
       " 0.9945716857910156,\n",
       " 0.9956685304641724,\n",
       " 0.9940922856330872,\n",
       " 0.9948796629905701)"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "scores, acc, p, r, f = model.evaluate(test_sentences_X, to_categorical(test_tags_y, len(tag2index)))\n",
    "scores, acc, p, r, f"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Make predictions"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [],
   "source": [
    "def logits_to_tokens(sequences, index):\n",
    "    token_sequences = []\n",
    "    for categorical_sequence in sequences:\n",
    "        token_sequences.append([index[cat] for cat in np.argmax(categorical_sequence, axis=1)])\n",
    " \n",
    "    return token_sequences"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "inverted_tag_index = {i: t for t, i in tag2index.items()}"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 3.46 s, sys: 279 ms, total: 3.74 s\n",
      "Wall time: 865 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "predictions = model.predict(test_sentences_X)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 74.3 ms, sys: 0 ns, total: 74.3 ms\n",
      "Wall time: 73.9 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "pred_labels = logits_to_tokens(predictions, inverted_tag_index)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[['Nom',\n",
       "  'Unk',\n",
       "  'Unk',\n",
       "  'Unk',\n",
       "  'Ess',\n",
       "  'Unk',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-'],\n",
       " ['Lat',\n",
       "  'All',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-'],\n",
       " ['Nom',\n",
       "  'Unk',\n",
       "  'Gen',\n",
       "  'Unk',\n",
       "  'Nom',\n",
       "  'Unk',\n",
       "  'Nom',\n",
       "  'Nom',\n",
       "  'Unk',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-'],\n",
       " ['Unk',\n",
       "  'Unk',\n",
       "  'Nom',\n",
       "  'Unk',\n",
       "  'Unk',\n",
       "  'Nom',\n",
       "  'Par',\n",
       "  'Ade',\n",
       "  'Ela',\n",
       "  'Unk',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-'],\n",
       " ['Gen',\n",
       "  'Unk',\n",
       "  'Nom',\n",
       "  'Unk',\n",
       "  'Ade',\n",
       "  'Ade',\n",
       "  'Ade',\n",
       "  'Unk',\n",
       "  'Nom',\n",
       "  'Ela',\n",
       "  'Ela',\n",
       "  'Unk',\n",
       "  'Unk',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-',\n",
       "  '-PAD-']]\n"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "pred_labels[:5]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Classification report"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "y_true = [inverted_tag_index[t] for l in test_tags_y for t in l]\n",
    "y_pred = list(itertools.chain(*pred_labels))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "              precision    recall  f1-score   support\n",
      "\n",
      "       -PAD-       1.00      1.00      1.00    242650\n",
      "         Abe       1.00      0.79      0.88        42\n",
      "         Abl       0.94      0.83      0.88       189\n",
      "         Acc       1.00      1.00      1.00        25\n",
      "         Ade       0.98      0.96      0.97       649\n",
      "         All       0.97      0.96      0.96       359\n",
      "         Com       0.70      0.78      0.74        18\n",
      "         Ela       0.82      0.88      0.85       703\n",
      "         Ess       0.97      0.90      0.94       333\n",
      "         Gen       0.93      0.92      0.93      2795\n",
      "         Ill       0.89      0.86      0.88       817\n",
      "         Ine       0.97      0.97      0.97       842\n",
      "         Ins       0.78      0.69      0.73        67\n",
      "         Lat       0.92      0.92      0.92       649\n",
      "         Nom       0.91      0.96      0.93      6089\n",
      "         Par       0.93      0.90      0.92      2428\n",
      "         Tra       0.98      0.94      0.96       214\n",
      "         Unk       0.99      0.98      0.98     16171\n",
      "\n",
      "   micro avg       0.99      0.99      0.99    275040\n",
      "   macro avg       0.93      0.90      0.91    275040\n",
      "weighted avg       0.99      0.99      0.99    275040\n",
      "\n"
     ]
    }
   ],
   "source": [
    "print(classification_report(y_true, y_pred))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Build confusion matrices"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "labels = list(sorted(set(y_true)))\n",
    "array = confusion_matrix(y_true=y_true, y_pred=y_pred)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Precision"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_cm_1 = pd.DataFrame(array / array.sum(axis=0), index = labels, columns = labels)\n",
    "\n",
    "plt.figure(figsize = (10,10))\n",
    "sns.heatmap(df_cm_1, annot=True, cmap=plt.cm.Blues)\n",
    "plt.show() "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Recall"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 40,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "df_cm_2 = pd.DataFrame(array / array.sum(axis=1), index = labels, columns = labels)\n",
    "\n",
    "plt.figure(figsize = (10,10))\n",
    "sns.heatmap(df_cm_2, annot=True, cmap=plt.cm.Blues)\n",
    "plt.show()"
   ]
  }
 ],
 "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": "3.6.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}