hahnn-for-document-classification.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "[Paper] HAHNN for Document Classification",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": [
        "IFyTXuEK9TJ3",
        "5exIp4egmwwA",
        "JLle9TkY1T_X",
        "voCjs6ewiSoF",
        "QTR1_43CqGSW",
        "Iu_O63OD2h8P",
        "qsTKHWjKqXmo"
      ]
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "metadata": {
        "id": "EjNTyCbkZcKM",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "# Hierarchical Attentional Hybrid Neural Networks for Document Classification\n",
        "\n",
        "Document classification is a challenging task with important applications. Deep learning approaches to the problem have gained much attention. Despite the progress, the proposed models do not incorporate the knowledge of the document structure in the architecture efficiently and not take into account the contexting dependent importance of words and sentences. In this paper, we propose a new approach based on convolutional neural networks, gated recurrent units and attention mechanisms for document classification tasks. The datasets IMDB Movie Reviews and Yelp were used in experiments. The proposed method improves the results of current attention-based approaches\n",
        "\n",
        "Please, cite:\n",
        "\n",
        "Abreu, J., Fred, L., Macêdo, D., & Zanchettin, C. (2019). Hierarchical Attentional Hybrid Neural Networks for Document Classification. arXiv preprint arXiv:1901.06610."
      ]
    },
    {
      "metadata": {
        "id": "dzZHZGMH4X5-",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "# Prompts for kaggle username and API key\n",
        "\n",
        "Displays a field in prompt and waits for the user to input your kaggle username. By press ENTER, other field prompts for input Kaggle API key."
      ]
    },
    {
      "metadata": {
        "id": "J4z5K6u43ey_",
        "colab_type": "code",
        "outputId": "373cd022-248c-40f7-9a9d-f0be8682e925",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "cell_type": "code",
      "source": [
        "import os\n",
        "from getpass import getpass\n",
        "\n",
        "user = getpass('Kaggle Username: ')\n",
        "key = getpass('Kaggle API key: ')\n",
        "\n",
        "if '.kaggle' not in os.listdir('/root'):\n",
        "    !mkdir ~/.kaggle\n",
        "    \n",
        "!touch /root/.kaggle/kaggle.json\n",
        "!chmod 666 /root/.kaggle/kaggle.json\n",
        "\n",
        "with open('/root/.kaggle/kaggle.json', 'w') as f:\n",
        "    f.write('{\"username\":\"%s\",\"key\":\"%s\"}' % (user, key))\n",
        "!chmod 600 /root/.kaggle/kaggle.json"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Kaggle Username: ··········\n",
            "Kaggle API key: ··········\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "4_pgmoOw4gkH",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "# Download dataset"
      ]
    },
    {
      "metadata": {
        "id": "n-mnEILI3hRR",
        "colab_type": "code",
        "outputId": "9493a976-404e-4dfd-bc4f-7f69274aa1b9",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 119
        }
      },
      "cell_type": "code",
      "source": [
        "!kaggle datasets download -d luisfredgs/hahnn-for-document-classification\n",
        "!unzip -o hahnn-for-document-classification.zip\n",
        "# fixing in 400k reviews\n",
        "!head -n400000 yelp_reviews.json > yelp_reviews_sampling.json"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Downloading hahnn-for-document-classification.zip to /content\n",
            " 97% 344M/355M [00:02<00:00, 112MB/s] \n",
            "100% 355M/355M [00:02<00:00, 125MB/s]\n",
            "Archive:  hahnn-for-document-classification.zip\n",
            "  inflating: imdb_reviews.csv        \n",
            "  inflating: yelp_reviews.json       \n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "GOqzxs-ynuCb",
        "colab_type": "code",
        "trusted": true,
        "cellView": "form",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "dataset = \"yelp\" #@param [\"yelp\", \"imdb\"]\n",
        "\n",
        "word_embedding_type = \"from_scratch\" #@param [\"from_scratch\", \"pre_trained\"]\n",
        "word_vector_model = \"fasttext\" #@param [\"fasttext\"]\n",
        "rnn_type = \"GRU\" #@param [\"LSTM\", \"GRU\"]\n",
        "learning_rate = 0.001\n",
        "epochs = 8\n",
        "batch_size = 64"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "IFyTXuEK9TJ3",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "# Imports"
      ]
    },
    {
      "metadata": {
        "id": "n4g5Ufug8kyi",
        "colab_type": "code",
        "outputId": "1146b9d9-58e7-49e5-fa89-7be63c7cb0cd",
        "trusted": true,
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 0
        }
      },
      "cell_type": "code",
      "source": [
        "!pip -q install gensim\n",
        "!python -m spacy download en_core_web_md\n",
        "!pip -q install paramiko\n",
        "\n",
        "import datetime, pickle, os, codecs, re, string\n",
        "import json\n",
        "import random\n",
        "import numpy as np\n",
        "import keras\n",
        "from keras.models import *\n",
        "from keras.layers import *\n",
        "from keras.optimizers import *\n",
        "from keras.callbacks import *\n",
        "from keras import regularizers\n",
        "from keras.preprocessing.text import Tokenizer\n",
        "from keras.preprocessing.sequence import pad_sequences\n",
        "from keras import backend as K\n",
        "from keras.utils import CustomObjectScope\n",
        "from keras.engine.topology import Layer\n",
        "\n",
        "#\n",
        "from keras.engine import InputSpec\n",
        "\n",
        "from keras import initializers\n",
        "\n",
        "import pandas as pd\n",
        "from tqdm import tqdm\n",
        "\n",
        "import string\n",
        "from spacy.lang.en import English\n",
        "import gensim, nltk, logging\n",
        "\n",
        "from nltk.corpus import stopwords\n",
        "from nltk import tokenize\n",
        "from nltk.stem import SnowballStemmer\n",
        "\n",
        "nltk.download('punkt')\n",
        "nltk.download('stopwords')\n",
        "\n",
        "from sklearn.manifold import TSNE\n",
        "import matplotlib.pyplot as plt\n",
        "import en_core_web_sm\n",
        "\n",
        "from IPython.display import HTML, display\n",
        "\n",
        "import tensorflow as tf\n",
        "\n",
        "from numpy.random import seed\n",
        "from tensorflow import set_random_seed\n",
        "os.environ['PYTHONHASHSEED'] = str(1024)\n",
        "set_random_seed(1024)\n",
        "seed(1024)\n",
        "np.random.seed(1024)\n",
        "random.seed(1024)"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Collecting en_core_web_md==2.0.0 from https://github.com/explosion/spacy-models/releases/download/en_core_web_md-2.0.0/en_core_web_md-2.0.0.tar.gz#egg=en_core_web_md==2.0.0\n",
            "\u001b[?25l  Downloading https://github.com/explosion/spacy-models/releases/download/en_core_web_md-2.0.0/en_core_web_md-2.0.0.tar.gz (120.8MB)\n",
            "\u001b[K    100% |████████████████████████████████| 120.9MB 101.2MB/s \n",
            "\u001b[?25hInstalling collected packages: en-core-web-md\n",
            "  Running setup.py install for en-core-web-md ... \u001b[?25ldone\n",
            "\u001b[?25hSuccessfully installed en-core-web-md-2.0.0\n",
            "\n",
            "\u001b[93m    Linking successful\u001b[0m\n",
            "    /usr/local/lib/python3.6/dist-packages/en_core_web_md -->\n",
            "    /usr/local/lib/python3.6/dist-packages/spacy/data/en_core_web_md\n",
            "\n",
            "    You can now load the model via spacy.load('en_core_web_md')\n",
            "\n",
            "\u001b[K    100% |████████████████████████████████| 194kB 7.9MB/s \n",
            "\u001b[K    100% |████████████████████████████████| 768kB 25.1MB/s \n",
            "\u001b[K    100% |████████████████████████████████| 2.3MB 11.5MB/s \n",
            "\u001b[K    100% |████████████████████████████████| 61kB 27.2MB/s \n",
            "\u001b[K    100% |████████████████████████████████| 102kB 32.9MB/s \n",
            "\u001b[?25h"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "Using TensorFlow backend.\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "[nltk_data] Downloading package punkt to /root/nltk_data...\n",
            "[nltk_data]   Unzipping tokenizers/punkt.zip.\n",
            "[nltk_data] Downloading package stopwords to /root/nltk_data...\n",
            "[nltk_data]   Unzipping corpora/stopwords.zip.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "5exIp4egmwwA",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "# Text preprocessing"
      ]
    },
    {
      "metadata": {
        "id": "zY1TBBBbqEbp",
        "colab_type": "code",
        "trusted": true,
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "def clean_str(string):\n",
        "    string = re.sub(r\"\\'s\", \" \\'s\", string)\n",
        "    string = re.sub(r\"\\'ve\", \" \\'ve\", string)\n",
        "    string = re.sub(r\"n\\'t\", \" n\\'t\", string)\n",
        "    string = re.sub(r\"\\'re\", \" \\'re\", string)\n",
        "    string = re.sub(r\"\\'d\", \" \\'d\", string)\n",
        "    string = re.sub(r\"\\'ll\", \" \\'ll\", string)\n",
        "    string = re.sub(r\",\", \" , \", string)\n",
        "    string = re.sub(r\"!\", \" ! \", string)\n",
        "    string = re.sub(r\"\\(\", \" \\( \", string)\n",
        "    string = re.sub(r\"\\)\", \" \\) \", string)\n",
        "    string = re.sub(r\"\\?\", \" \\? \", string)\n",
        "    string = re.sub(r\"\\s{2,}\", \" \", string)\n",
        "\n",
        "    cleanr = re.compile('<.*?>')\n",
        "\n",
        "    # string = re.sub(r'\\d+', '', string)\n",
        "    string = re.sub(cleanr, '', string)\n",
        "    # string = re.sub(\"'\", '', string)\n",
        "    # string = re.sub(r'\\W+', ' ', string)\n",
        "    string = string.replace('_', '')\n",
        "\n",
        "\n",
        "    return string.strip().lower()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "JLle9TkY1T_X",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "# Use pre-trained word embeddings"
      ]
    },
    {
      "metadata": {
        "id": "oLodIgWw14kC",
        "colab_type": "code",
        "trusted": true,
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "def load_subword_embedding_300d(word_index):\n",
        "    print('load_subword_embedding...')\n",
        "    embeddings_index = {}\n",
        "    f = codecs.open(\"wiki-news-300d-1M-subword.vec\", encoding='utf-8')\n",
        "    for line in tqdm(f):\n",
        "        values = line.rstrip().rsplit(' ')\n",
        "        word = values[0]\n",
        "        coefs = np.asarray(values[1:], dtype='float32')\n",
        "        embeddings_index[word] = coefs\n",
        "    f.close()\n",
        "    print('found %s word vectors' % len(embeddings_index))\n",
        "    \n",
        "    #embedding matrix\n",
        "    print('preparing embedding matrix...')\n",
        "    words_not_found = []\n",
        "    \n",
        "    embedding_matrix = np.zeros((len(word_index) + 1, 300))\n",
        "    \n",
        "    for word, i in word_index.items():\n",
        "        embedding_vector = embeddings_index.get(word)\n",
        "        if (embedding_vector is not None) and len(embedding_vector) > 0:\n",
        "            \n",
        "            embedding_matrix[i] = embedding_vector\n",
        "        else:\n",
        "            words_not_found.append(word)\n",
        "    \n",
        "    return embedding_matrix"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "voCjs6ewiSoF",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "# Plot word embedding chart"
      ]
    },
    {
      "metadata": {
        "id": "0rXXQmEMibaF",
        "colab_type": "code",
        "trusted": true,
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "def plot_with_labels(low_dim_embs, labels, filename='tsne.png'):    \n",
        "    plt.figure(figsize=(18, 18))\n",
        "    for i, label in enumerate(labels):\n",
        "        x, y = low_dim_embs[i, :]\n",
        "        plt.scatter(x, y)\n",
        "        plt.annotate(label,\n",
        "                 xy=(x, y),\n",
        "                 xytext=(5, 2),\n",
        "                 textcoords='offset points',\n",
        "                 ha='right',\n",
        "                 va='bottom')\n",
        "    plt.savefig(filename)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "QTR1_43CqGSW",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "# Normalize texts"
      ]
    },
    {
      "metadata": {
        "id": "ePLZEjNxqPKe",
        "colab_type": "code",
        "trusted": true,
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "nlp = en_core_web_sm.load()\n",
        "\n",
        "\n",
        "puncts = [',', '.', '\"', ':', ')', '(', '-', '!', '?', '|', ';', \"'\", '$', '&', '/', '[', ']', '>', '%', '=', '#', '*', '+', '\\\\', '•',  '~', '@', '£', \n",
        " '·', '_', '{', '}', '©', '^', '®', '`',  '<', '→', '°', '€', '™', '›',  '♥', '←', '×', '§', '″', '′', 'Â', '█', '½', 'à', '…', \n",
        " '“', '★', '”', '–', '●', 'â', '►', '−', '¢', '²', '¬', '░', '¶', '↑', '±', '¿', '▾', '═', '¦', '║', '―', '¥', '▓', '—', '‹', '─', \n",
        " '▒', '：', '¼', '⊕', '▼', '▪', '†', '■', '’', '▀', '¨', '▄', '♫', '☆', 'é', '¯', '♦', '¤', '▲', 'è', '¸', '¾', 'Ã', '⋅', '‘', '∞', \n",
        " '∙', '）', '↓', '、', '│', '（', '»', '，', '♪', '╩', '╚', '³', '・', '╦', '╣', '╔', '╗', '▬', '❤', 'ï', 'Ø', '¹', '≤', '‡', '√', '#', '—–']\n",
        "\n",
        "\n",
        "def clean_puncts(x):\n",
        "    x = str(x)\n",
        "    for punct in puncts:\n",
        "        x = x.replace(punct, f' {punct} ')\n",
        "    return x\n",
        "\n",
        "def remove_stopwords(text):\n",
        "    text = str(text)    \n",
        "    ## Convert words to lower case and split them\n",
        "    text = text.lower().split()\n",
        "    \n",
        "    ## Remove stop words\n",
        "    stops = set(stopwords.words(\"english\"))\n",
        "    text = [w for w in text if not w in stops and len(w) >= 3]\n",
        "    text = \" \".join(text)\n",
        "    \n",
        "    return text\n",
        "\n",
        "\n",
        "def normalize(text):\n",
        "    text = text.lower().strip()\n",
        "    doc = nlp(text)\n",
        "    filtered_sentences = []\n",
        "    for sentence in doc.sents:                    \n",
        "        sentence = clean_puncts(sentence)\n",
        "        sentence = clean_str(sentence)            \n",
        "        #sentence = remove_stopwords(sentence)                \n",
        "        filtered_sentences.append(sentence)\n",
        "    return filtered_sentences"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "Iu_O63OD2h8P",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "# Training word embeddings"
      ]
    },
    {
      "metadata": {
        "id": "HXyS2m7C2pZ7",
        "colab_type": "code",
        "trusted": true,
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "def create_fasttext(embed_dim, data):\n",
        "    \n",
        "    filename = './fasttext_model.txt'\n",
        "    \n",
        "    if not os.path.isfile(filename):    \n",
        "        print('create_fasttext...')\n",
        "        sent_lst = []\n",
        "\n",
        "        for doc in data['text']:\n",
        "            doc = clean_str(doc)\n",
        "            sentences = nltk.tokenize.sent_tokenize(doc)\n",
        "            for sent in sentences:\n",
        "                word_lst = [w for w in nltk.tokenize.word_tokenize(sent) if w.isalnum()]\n",
        "                sent_lst.append(word_lst)\n",
        "\n",
        "\n",
        "        logging.basicConfig(format='%(asctime)s : %(levelname)s : %(message)s', level=logging.INFO)\n",
        "\n",
        "        fasttext_model = gensim.models.FastText(\n",
        "            word_ngrams=1,\n",
        "            sentences=sent_lst, \n",
        "            size = embed_dim, \n",
        "            workers=os.cpu_count(), \n",
        "            window = 1)\n",
        "        fasttext_model.save(\"./fasttext_model.txt\")"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "724247Q4phPE",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "# Load datasets\n",
        "\n",
        "**1)** Yelp Dataset with 500k reviews\n",
        "\n",
        "**2)** IMDb with 50k reviews"
      ]
    },
    {
      "metadata": {
        "id": "0Owr7XARpnO7",
        "colab_type": "code",
        "trusted": true,
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "def load_data_yelp(path, train_ratio=1, size=400000):\n",
        "    vector_dim = 200\n",
        "    dim = 5\n",
        "    \n",
        "    with open(path) as f:\n",
        "        reviews=f.read().strip().split(\"\\n\")\n",
        "        \n",
        "    df = pd.DataFrame([json.loads(review) for review in reviews])        \n",
        "\n",
        "    text_tokens = []\n",
        "    for row in tqdm(df['text']):    \n",
        "        text_tokens.append(normalize(row))  \n",
        "    \n",
        "    df['text_tokens'] = text_tokens    \n",
        "    del text_tokens\n",
        "    \n",
        "    vector_dim = 200     \n",
        "    if word_embedding_type is 'from_scratch':\n",
        "        create_fasttext(vector_dim, df)\n",
        "        \n",
        "    ###\n",
        "    \n",
        "    X = df['text_tokens'].values\n",
        "    Y = pd.get_dummies(df['stars']).values\n",
        "    \n",
        "    return (X, Y)\n",
        "\n",
        "def load_data_imdb(path, size=49000, train_ratio=1):    \n",
        "    df = pd.read_csv(path, nrows=size, usecols=['text', 'sentiment'])        \n",
        "    \n",
        "    vector_dim = 200 \n",
        "    \n",
        "    if word_embedding_type is 'from_scratch':\n",
        "        # Fasttext\n",
        "        create_fasttext(vector_dim, df)\n",
        "    \n",
        "    ### \n",
        "    \n",
        "    text_tokens = []\n",
        "    for row in tqdm(df['text']):    \n",
        "        text_tokens.append(normalize(row))  \n",
        "    \n",
        "    df['text_tokens'] = text_tokens\n",
        "    \n",
        "    del text_tokens\n",
        "    ###\n",
        "    \n",
        "    X = df['text_tokens'].values\n",
        "    Y = pd.get_dummies(df['sentiment']).values\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "qsTKHWjKqXmo",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "# Attention Layer\n",
        "Check [(Bahdanau et al., 2015)](https://arxiv.org/pdf/1409.0473.pdf)"
      ]
    },
    {
      "metadata": {
        "id": "C_HrvPCJ9HYs",
        "colab_type": "code",
        "trusted": true,
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "class Attention(Layer):\n",
        "    def __init__(self, **kwargs):\n",
        "        self.init = initializers.get('normal')\n",
        "        self.supports_masking = True\n",
        "        self.attention_dim = 50\n",
        "        super(Attention, self).__init__(**kwargs)\n",
        "\n",
        "    def build(self, input_shape):\n",
        "        assert len(input_shape) == 3\n",
        "        self.W = K.variable(self.init((input_shape[-1], 1)))\n",
        "        self.b = K.variable(self.init((self.attention_dim, )))\n",
        "        self.u = K.variable(self.init((self.attention_dim, 1)))\n",
        "        self.trainable_weights = [self.W, self.b, self.u]\n",
        "        super(Attention, self).build(input_shape)\n",
        "\n",
        "    def compute_mask(self, inputs, mask=None):\n",
        "        return mask\n",
        "\n",
        "    def call(self, x, mask=None):\n",
        "        uit = K.tanh(K.bias_add(K.dot(x, self.W), self.b))\n",
        "        ait = K.dot(uit, self.u)\n",
        "        ait = K.squeeze(ait, -1)\n",
        "        ait = K.exp(ait)\n",
        "\n",
        "        if mask is not None:\n",
        "            ait *= K.cast(mask, K.floatx())\n",
        "            \n",
        "        ait /= K.cast(K.sum(ait, axis=1, keepdims=True) + K.epsilon(), K.floatx())\n",
        "        ait = K.expand_dims(ait)\n",
        "        weighted_input = x * ait\n",
        "        output = K.sum(weighted_input, axis=1)\n",
        "        return output\n",
        "\n",
        "    def compute_output_shape(self, input_shape):\n",
        "        return (input_shape[0], input_shape[-1])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "lmmw7Qfvqx-f",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "# Model architecture"
      ]
    },
    {
      "metadata": {
        "id": "jHLATdJqEIiN",
        "colab_type": "code",
        "trusted": true,
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "!rm -rf saved_models\n",
        "!mkdir saved_models"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "V6bzUS1dqftf",
        "colab_type": "code",
        "trusted": true,
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "class HAHNetwork():\n",
        "    def __init__(self):\n",
        "        self.model = None\n",
        "        self.MAX_SENTENCE_LENGTH = 0\n",
        "        self.MAX_SENTENCE_COUNT = 0\n",
        "        self.VOCABULARY_SIZE = 0\n",
        "        self.word_embedding = None\n",
        "        self.model = None\n",
        "        self.word_attention_model = None\n",
        "        self.tokenizer = None\n",
        "        self.class_count = 2\n",
        "\n",
        "    def build_model(self, n_classes=2, embedding_dim=200, embeddings_path=False):\n",
        "        \n",
        "        l2_reg = regularizers.l2(0.001)\n",
        "        \n",
        "        embedding_weights = np.random.normal(0, 1, (len(self.tokenizer.word_index) + 1, embedding_dim))\n",
        "        \n",
        "        if embeddings_path is not None:\n",
        "\n",
        "            if word_embedding_type is 'from_scratch':\n",
        "                # FastText\n",
        "                filename = './fasttext_model.txt'                \n",
        "                model =  gensim.models.FastText.load(filename)\n",
        "\n",
        "                embeddings_index = model.wv                    \n",
        "                embedding_matrix = np.zeros( ( len(self.tokenizer.word_index) + 1, embedding_dim) )\n",
        "                for word, i in self.tokenizer.word_index.items():\n",
        "                    try:\n",
        "                        embedding_vector = embeddings_index[word]\n",
        "                        if embedding_vector is not None:\n",
        "                            embedding_matrix[i] = embedding_vector\n",
        "                    except Exception as e:\n",
        "                        #print(str(e))\n",
        "                        continue\n",
        "\n",
        "\n",
        "            else:                \n",
        "                embedding_dim = 300\n",
        "                embedding_matrix = load_subword_embedding_300d(self.tokenizer.word_index)\n",
        "\n",
        "            embedding_weights = embedding_matrix\n",
        "\n",
        "        sentence_in = Input(shape=(self.MAX_SENTENCE_LENGTH,), dtype='int32', name=\"input_1\")\n",
        "        \n",
        "        embedding_trainable = True\n",
        "        \n",
        "        \n",
        "        \n",
        "        if word_embedding_type is 'pre_trained':\n",
        "            embedding_trainable = False\n",
        "        \n",
        "        embedded_word_seq = Embedding(\n",
        "            self.VOCABULARY_SIZE,\n",
        "            embedding_dim,\n",
        "            weights=[embedding_weights],\n",
        "            input_length=self.MAX_SENTENCE_LENGTH,\n",
        "            trainable=embedding_trainable,\n",
        "            #mask_zero=True,\n",
        "            mask_zero=False,\n",
        "            name='word_embeddings',)(sentence_in) \n",
        "        \n",
        "        \n",
        "                     \n",
        "        dropout = Dropout(0.2)(embedded_word_seq)\n",
        "        filter_sizes = [3,4,5]\n",
        "        convs = []\n",
        "        for filter_size in filter_sizes:\n",
        "            conv = Conv1D(filters=64, kernel_size=filter_size, padding='same', activation='relu')(dropout)\n",
        "            pool = MaxPool1D(filter_size)(conv)\n",
        "            convs.append(pool)\n",
        "        \n",
        "        concatenate = Concatenate(axis=1)(convs)\n",
        "        \n",
        "        if rnn_type is 'GRU':\n",
        "            #word_encoder = Bidirectional(CuDNNGRU(50, return_sequences=True, dropout=0.2))(concatenate)                \n",
        "            dropout = Dropout(0.1)(concatenate)\n",
        "            word_encoder = Bidirectional(CuDNNGRU(50, return_sequences=True))(dropout)                \n",
        "        else:\n",
        "            word_encoder = Bidirectional(\n",
        "                LSTM(50, return_sequences=True, dropout=0.2))(embedded_word_seq)\n",
        "            \n",
        "        \n",
        "        dense_transform_word = Dense(\n",
        "            100, \n",
        "            activation='relu', \n",
        "            name='dense_transform_word', \n",
        "            kernel_regularizer=l2_reg)(word_encoder)\n",
        "        \n",
        "        # word attention\n",
        "        attention_weighted_sentence = Model(\n",
        "            sentence_in, Attention(name=\"word_attention\")(dense_transform_word))\n",
        "        \n",
        "        self.word_attention_model = attention_weighted_sentence\n",
        "        \n",
        "        attention_weighted_sentence.summary()\n",
        "\n",
        "        # sentence-attention-weighted document scores\n",
        "        \n",
        "        texts_in = Input(shape=(self.MAX_SENTENCE_COUNT, self.MAX_SENTENCE_LENGTH), dtype='int32', name=\"input_2\")\n",
        "        \n",
        "        attention_weighted_sentences = TimeDistributed(attention_weighted_sentence)(texts_in)\n",
        "        \n",
        "        \n",
        "        if rnn_type is 'GRU':\n",
        "            #sentence_encoder = Bidirectional(GRU(50, return_sequences=True, dropout=0.1, recurrent_dropout=0.2))(attention_weighted_sentences)\n",
        "            dropout = Dropout(0.1)(attention_weighted_sentences)\n",
        "            sentence_encoder = Bidirectional(CuDNNGRU(50, return_sequences=True))(dropout)\n",
        "        else:\n",
        "            sentence_encoder = Bidirectional(LSTM(50, return_sequences=True, dropout=0.1, recurrent_dropout=0.2))(attention_weighted_sentences)\n",
        "        \n",
        "        \n",
        "        dense_transform_sentence = Dense(\n",
        "            100, \n",
        "            activation='relu', \n",
        "            name='dense_transform_sentence',\n",
        "            kernel_regularizer=l2_reg)(sentence_encoder)\n",
        "        \n",
        "        # sentence attention\n",
        "        attention_weighted_text = Attention(name=\"sentence_attention\")(dense_transform_sentence)\n",
        "        \n",
        "        \n",
        "        prediction = Dense(n_classes, activation='softmax')(attention_weighted_text)\n",
        "        \n",
        "        model = Model(texts_in, prediction)\n",
        "        model.summary()\n",
        "        \n",
        "        \n",
        "        optimizer=Adam(lr=learning_rate, decay=0.0001)\n",
        "\n",
        "        model.compile(\n",
        "                      optimizer=optimizer,\n",
        "                      loss='categorical_crossentropy',\n",
        "                      metrics=['accuracy'])\n",
        "\n",
        "        return model\n",
        "\n",
        "\n",
        "    def get_tokenizer_filename(self, saved_model_filename):\n",
        "        return saved_model_filename + '.tokenizer'\n",
        "\n",
        "    def fit_on_texts(self, texts):\n",
        "        self.tokenizer = Tokenizer(filters='\"()*,-/;[\\]^_`{|}~', oov_token='UNK');\n",
        "        all_sentences = []\n",
        "        max_sentence_count = 0\n",
        "        max_sentence_length = 0\n",
        "        for text in texts:\n",
        "            sentence_count = len(text)\n",
        "            if sentence_count > max_sentence_count:\n",
        "                max_sentence_count = sentence_count\n",
        "            for sentence in text:\n",
        "                sentence_length = len(sentence)\n",
        "                if sentence_length > max_sentence_length:\n",
        "                    max_sentence_length = sentence_length\n",
        "                all_sentences.append(sentence)\n",
        "\n",
        "\n",
        "        self.MAX_SENTENCE_COUNT = min(max_sentence_count, 15)\n",
        "        self.MAX_SENTENCE_LENGTH = min(max_sentence_length, 50)\n",
        "        \n",
        "        self.tokenizer.fit_on_texts(all_sentences)\n",
        "        self.VOCABULARY_SIZE = len(self.tokenizer.word_index) + 1\n",
        "        self.create_reverse_word_index()\n",
        "\n",
        "    def create_reverse_word_index(self):\n",
        "        self.reverse_word_index = {value:key for key,value in self.tokenizer.word_index.items()}\n",
        "\n",
        "    def encode_texts(self, texts):\n",
        "        encoded_texts = np.zeros((len(texts), self.MAX_SENTENCE_COUNT, self.MAX_SENTENCE_LENGTH))\n",
        "        for i, text in enumerate(texts):\n",
        "            encoded_text = np.array(pad_sequences(\n",
        "                self.tokenizer.texts_to_sequences(text), \n",
        "                maxlen=self.MAX_SENTENCE_LENGTH))[:self.MAX_SENTENCE_COUNT]\n",
        "            encoded_texts[i][-len(encoded_text):] = encoded_text\n",
        "        return encoded_texts\n",
        "\n",
        "    def save_tokenizer_on_epoch_end(self, path, epoch):\n",
        "        if epoch == 0:\n",
        "            tokenizer_state = {\n",
        "                'tokenizer': self.tokenizer,\n",
        "                'maxSentenceCount': self.MAX_SENTENCE_COUNT,\n",
        "                'maxSentenceLength': self.MAX_SENTENCE_LENGTH,\n",
        "                'vocabularySize': self.VOCABULARY_SIZE\n",
        "            }\n",
        "            pickle.dump(tokenizer_state, open(path, \"wb\" ) )\n",
        "\n",
        "    def train(self, train_x, train_y,\n",
        "              batch_size=16, \n",
        "              epochs=1, \n",
        "              embedding_dim=200, \n",
        "              embeddings_path=False, \n",
        "              saved_model_dir='saved_models', \n",
        "              saved_model_filename=None,):\n",
        "        \n",
        "        self.fit_on_texts(train_x)\n",
        "        self.model = self.build_model(\n",
        "            n_classes=train_y.shape[-1], \n",
        "            embedding_dim=200,\n",
        "            embeddings_path=embeddings_path)\n",
        "        encoded_train_x = self.encode_texts(train_x)\n",
        "        callbacks = [\n",
        "            ReduceLROnPlateau(),\n",
        "            LambdaCallback(\n",
        "                on_epoch_end=lambda epoch, logs: self.save_tokenizer_on_epoch_end(\n",
        "                    os.path.join(saved_model_dir, \n",
        "                        self.get_tokenizer_filename(saved_model_filename)), epoch))\n",
        "        ]\n",
        "\n",
        "        if saved_model_filename:\n",
        "            callbacks.append(\n",
        "                ModelCheckpoint(\n",
        "                    filepath=os.path.join(saved_model_dir, saved_model_filename),\n",
        "                    monitor='val_acc',\n",
        "                    save_best_only=True,\n",
        "                    save_weights_only=False,\n",
        "                )\n",
        "            )\n",
        "        history = self.model.fit(\n",
        "                       x=encoded_train_x, \n",
        "                       y=train_y, \n",
        "                       batch_size=batch_size, \n",
        "                       epochs=epochs, \n",
        "                       verbose=1, \n",
        "                       callbacks=callbacks,\n",
        "                       validation_split=0.1,  \n",
        "                       shuffle=True)\n",
        "        \n",
        "        # Plot\n",
        "        print(history.history.keys())\n",
        "        \n",
        "        plt.plot(history.history['acc'])\n",
        "        plt.plot(history.history['val_acc'])\n",
        "        plt.title('model accuracy')\n",
        "        plt.ylabel('accuracy')\n",
        "        plt.xlabel('epoch')\n",
        "        plt.legend(['train', 'test'], loc='upper left')\n",
        "        plt.show()\n",
        "        \n",
        "        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', 'test'], loc='upper left')\n",
        "        plt.show()\n",
        "\n",
        "    def encode_input(self, x, log=False):\n",
        "        x = np.array(x)\n",
        "        if not x.shape:\n",
        "            x = np.expand_dims(x, 0)\n",
        "        texts = np.array([normalize(text) for text in x])\n",
        "        return self.encode_texts(texts)\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "obo_fTOvUsi8",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "# Load dataset\n",
        "\n",
        "This might take several minutes, depending your dataset"
      ]
    },
    {
      "metadata": {
        "id": "jFwuRFg3qxBW",
        "colab_type": "code",
        "outputId": "169edaf5-d3c2-406d-b0ed-62ac08f257c9",
        "trusted": true,
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "cell_type": "code",
      "source": [
        "YELP_DATA_PATH = 'yelp_reviews_sampling.json'\n",
        "IMDB_DATA_PATH = 'imdb_reviews.csv'\n",
        "SAVED_MODEL_DIR = 'saved_models'\n",
        "SAVED_MODEL_FILENAME = 'model.h5'\n",
        "\n",
        "if dataset is 'yelp':\n",
        "    (X, Y) = load_data_yelp(path=YELP_DATA_PATH, size=400000)\n",
        "else:\n",
        "    (X, Y) = load_data_imdb(path=IMDB_DATA_PATH, size=49000)\n"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            " 61%|██████    | 244394/400000 [3:07:41<1:32:16, 28.10it/s]"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Buffered data was truncated after reaching the output size limit."
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "D9Zcz1wv0wf3",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "# Plots Word Embedding chart\n"
      ]
    },
    {
      "metadata": {
        "id": "0TD83uVu04z7",
        "colab_type": "code",
        "trusted": true,
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1288
        },
        "outputId": "7cf44b2b-485c-4280-d72a-ea5ecf5adc93"
      },
      "cell_type": "code",
      "source": [
        "limit = 200\n",
        "vector_dim = 200\n",
        "\n",
        "# Fasttext\n",
        "filename = './fasttext_model.txt'                \n",
        "model =  gensim.models.FastText.load(filename)\n",
        "words = []\n",
        "embedding = np.array([])\n",
        "i = 0\n",
        "for word in model.wv.vocab:\n",
        "    if i == limit: break\n",
        "\n",
        "    words.append(word)\n",
        "    embedding = np.append(embedding, model[word])\n",
        "    i += 1\n",
        "\n",
        "embedding = embedding.reshape(limit, vector_dim)    \n",
        "tsne = TSNE(n_components=2)\n",
        "low_dim_embedding = tsne.fit_transform(embedding)\n",
        "       \n",
        "plot_with_labels(low_dim_embedding, words)"
      ],
      "execution_count": 15,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "2019-04-29 18:54:38,453 : INFO : loading FastText object from ./fasttext_model.txt\n",
            "2019-04-29 18:54:38,455 : WARNING : this function is deprecated, use smart_open.open instead\n",
            "2019-04-29 18:54:39,446 : INFO : loading wv recursively from ./fasttext_model.txt.wv.* with mmap=None\n",
            "2019-04-29 18:54:39,448 : INFO : loading vectors_ngrams from ./fasttext_model.txt.wv.vectors_ngrams.npy with mmap=None\n",
            "2019-04-29 18:54:39,542 : INFO : setting ignored attribute vectors_norm to None\n",
            "2019-04-29 18:54:39,544 : INFO : setting ignored attribute vectors_vocab_norm to None\n",
            "2019-04-29 18:54:39,545 : INFO : setting ignored attribute vectors_ngrams_norm to None\n",
            "2019-04-29 18:54:39,548 : INFO : setting ignored attribute buckets_word to None\n",
            "2019-04-29 18:54:39,550 : INFO : loading vocabulary recursively from ./fasttext_model.txt.vocabulary.* with mmap=None\n",
            "2019-04-29 18:54:39,553 : INFO : loading trainables recursively from ./fasttext_model.txt.trainables.* with mmap=None\n",
            "2019-04-29 18:54:39,559 : INFO : loading vectors_ngrams_lockf from ./fasttext_model.txt.trainables.vectors_ngrams_lockf.npy with mmap=None\n",
            "2019-04-29 18:54:39,652 : INFO : loaded ./fasttext_model.txt\n",
            "/usr/local/lib/python3.6/dist-packages/ipykernel_launcher.py:14: DeprecationWarning: Call to deprecated `__getitem__` (Method will be removed in 4.0.0, use self.wv.__getitem__() instead).\n",
            "  \n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAABBQAAAP1CAYAAADRoLw1AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<Figure size 1296x1296 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "id": "qOZg-uB4HV1H",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "\n",
        "# Taining model"
      ]
    },
    {
      "metadata": {
        "id": "7ZkHlV6rHNkJ",
        "colab_type": "code",
        "trusted": true,
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 2395
        },
        "outputId": "41baeda1-e852-4511-d1ac-51387f61842c"
      },
      "cell_type": "code",
      "source": [
        "K.clear_session()\n",
        "model = HAHNetwork()\n",
        "model.train(X, Y, batch_size=64, epochs=8, embeddings_path=True, saved_model_dir=SAVED_MODEL_DIR, saved_model_filename=SAVED_MODEL_FILENAME)"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "2019-04-29 18:55:44,843 : INFO : loading FastText object from ./fasttext_model.txt\n",
            "2019-04-29 18:55:44,845 : WARNING : this function is deprecated, use smart_open.open instead\n",
            "2019-04-29 18:55:45,983 : INFO : loading wv recursively from ./fasttext_model.txt.wv.* with mmap=None\n",
            "2019-04-29 18:55:45,984 : INFO : loading vectors_ngrams from ./fasttext_model.txt.wv.vectors_ngrams.npy with mmap=None\n",
            "2019-04-29 18:55:46,086 : INFO : setting ignored attribute vectors_norm to None\n",
            "2019-04-29 18:55:46,087 : INFO : setting ignored attribute vectors_vocab_norm to None\n",
            "2019-04-29 18:55:46,088 : INFO : setting ignored attribute vectors_ngrams_norm to None\n",
            "2019-04-29 18:55:46,090 : INFO : setting ignored attribute buckets_word to None\n",
            "2019-04-29 18:55:46,090 : INFO : loading vocabulary recursively from ./fasttext_model.txt.vocabulary.* with mmap=None\n",
            "2019-04-29 18:55:46,091 : INFO : loading trainables recursively from ./fasttext_model.txt.trainables.* with mmap=None\n",
            "2019-04-29 18:55:46,092 : INFO : loading vectors_ngrams_lockf from ./fasttext_model.txt.trainables.vectors_ngrams_lockf.npy with mmap=None\n",
            "2019-04-29 18:55:46,207 : INFO : loaded ./fasttext_model.txt\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/framework/op_def_library.py:263: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Colocations handled automatically by placer.\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "2019-04-29 18:55:54,856 : WARNING : From /usr/local/lib/python3.6/dist-packages/tensorflow/python/framework/op_def_library.py:263: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Colocations handled automatically by placer.\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:3445: calling dropout (from tensorflow.python.ops.nn_ops) with keep_prob is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Please use `rate` instead of `keep_prob`. Rate should be set to `rate = 1 - keep_prob`.\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "2019-04-29 18:55:55,680 : WARNING : From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:3445: calling dropout (from tensorflow.python.ops.nn_ops) with keep_prob is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Please use `rate` instead of `keep_prob`. Rate should be set to `rate = 1 - keep_prob`.\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "__________________________________________________________________________________________________\n",
            "Layer (type)                    Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            "input_1 (InputLayer)            (None, 50)           0                                            \n",
            "__________________________________________________________________________________________________\n",
            "word_embeddings (Embedding)     (None, 50, 200)      34060400    input_1[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "dropout_1 (Dropout)             (None, 50, 200)      0           word_embeddings[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv1d_1 (Conv1D)               (None, 50, 64)       38464       dropout_1[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "conv1d_2 (Conv1D)               (None, 50, 64)       51264       dropout_1[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "conv1d_3 (Conv1D)               (None, 50, 64)       64064       dropout_1[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "max_pooling1d_1 (MaxPooling1D)  (None, 16, 64)       0           conv1d_1[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "max_pooling1d_2 (MaxPooling1D)  (None, 12, 64)       0           conv1d_2[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "max_pooling1d_3 (MaxPooling1D)  (None, 10, 64)       0           conv1d_3[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_1 (Concatenate)     (None, 38, 64)       0           max_pooling1d_1[0][0]            \n",
            "                                                                 max_pooling1d_2[0][0]            \n",
            "                                                                 max_pooling1d_3[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "dropout_2 (Dropout)             (None, 38, 64)       0           concatenate_1[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "bidirectional_1 (Bidirectional) (None, 38, 100)      34800       dropout_2[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "dense_transform_word (Dense)    (None, 38, 100)      10100       bidirectional_1[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "word_attention (Attention)      (None, 100)          200         dense_transform_word[0][0]       \n",
            "==================================================================================================\n",
            "Total params: 34,259,292\n",
            "Trainable params: 34,259,292\n",
            "Non-trainable params: 0\n",
            "__________________________________________________________________________________________________\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "input_2 (InputLayer)         (None, 15, 50)            0         \n",
            "_________________________________________________________________\n",
            "time_distributed_1 (TimeDist (None, 15, 100)           34259292  \n",
            "_________________________________________________________________\n",
            "dropout_3 (Dropout)          (None, 15, 100)           0         \n",
            "_________________________________________________________________\n",
            "bidirectional_2 (Bidirection (None, 15, 100)           45600     \n",
            "_________________________________________________________________\n",
            "dense_transform_sentence (De (None, 15, 100)           10100     \n",
            "_________________________________________________________________\n",
            "sentence_attention (Attentio (None, 100)               200       \n",
            "_________________________________________________________________\n",
            "dense_1 (Dense)              (None, 5)                 505       \n",
            "=================================================================\n",
            "Total params: 34,315,697\n",
            "Trainable params: 34,315,697\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/ops/math_ops.py:3066: to_int32 (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use tf.cast instead.\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "2019-04-29 18:57:12,243 : WARNING : From /usr/local/lib/python3.6/dist-packages/tensorflow/python/ops/math_ops.py:3066: to_int32 (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use tf.cast instead.\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "stream",
          "text": [
            "Train on 360000 samples, validate on 40000 samples\n",
            "Epoch 1/8\n",
            "360000/360000 [==============================] - 483s 1ms/step - loss: 0.7650 - acc: 0.6915 - val_loss: 0.6750 - val_acc: 0.7219\n",
            "Epoch 2/8\n",
            "360000/360000 [==============================] - 463s 1ms/step - loss: 0.6592 - acc: 0.7261 - val_loss: 0.6380 - val_acc: 0.7350\n",
            "Epoch 3/8\n",
            "360000/360000 [==============================] - 471s 1ms/step - loss: 0.6282 - acc: 0.7372 - val_loss: 0.6300 - val_acc: 0.7371\n",
            "Epoch 4/8\n",
            "360000/360000 [==============================] - 466s 1ms/step - loss: 0.6064 - acc: 0.7467 - val_loss: 0.6246 - val_acc: 0.7412\n",
            "Epoch 5/8\n",
            "360000/360000 [==============================] - 466s 1ms/step - loss: 0.5894 - acc: 0.7545 - val_loss: 0.6269 - val_acc: 0.7414\n",
            "Epoch 6/8\n",
            "360000/360000 [==============================] - 466s 1ms/step - loss: 0.5767 - acc: 0.7605 - val_loss: 0.6254 - val_acc: 0.7409\n",
            "Epoch 7/8\n",
            "360000/360000 [==============================] - 465s 1ms/step - loss: 0.5646 - acc: 0.7659 - val_loss: 0.6266 - val_acc: 0.7418\n",
            "Epoch 8/8\n",
            "360000/360000 [==============================] - 465s 1ms/step - loss: 0.5551 - acc: 0.7703 - val_loss: 0.6310 - val_acc: 0.7405\n",
            "dict_keys(['val_loss', 'val_acc', 'loss', 'acc', 'lr'])\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAEWCAYAAACXGLsWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAWPEZkWi4CEvAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAEWCAYAAACXGLsWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\niIh0SGEhIiIdUliIiEiHFBYiItKh/wfkIGOOxYBGqgAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    }
  ]
}