SentimentAnalysis_copy.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "SentimentAnalysis.ipynb",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": [],
      "toc_visible": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "metadata": {
        "_uuid": "839fc1317e1b7253241839bbfa2d40303c53a3f1",
        "id": "CDx4CA09GZG8",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "## General information\n",
        "\n",
        "In this kernel I'll work with data from Movie Review Sentiment Analysis Playground Competition.\n"
      ]
    },
    {
      "metadata": {
        "id": "Vq4S5_HIGpBc",
        "colab_type": "code",
        "outputId": "92890594-d1dd-426e-a5ba-6f394c2ea339",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 280
        }
      },
      "cell_type": "code",
      "source": [
        "!pip install lightgbm wordcloud\n",
        "!pip install pydot && apt-get install graphviz\n"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Requirement already satisfied: lightgbm in /usr/local/lib/python3.6/dist-packages (2.2.2)\n",
            "Requirement already satisfied: wordcloud in /usr/local/lib/python3.6/dist-packages (1.5.0)\n",
            "Requirement already satisfied: scipy in /usr/local/lib/python3.6/dist-packages (from lightgbm) (1.1.0)\n",
            "Requirement already satisfied: numpy in /usr/local/lib/python3.6/dist-packages (from lightgbm) (1.14.6)\n",
            "Requirement already satisfied: scikit-learn in /usr/local/lib/python3.6/dist-packages (from lightgbm) (0.20.1)\n",
            "Requirement already satisfied: pillow in /usr/local/lib/python3.6/dist-packages (from wordcloud) (4.0.0)\n",
            "Requirement already satisfied: olefile in /usr/local/lib/python3.6/dist-packages (from pillow->wordcloud) (0.46)\n",
            "Requirement already satisfied: pydot in /usr/local/lib/python3.6/dist-packages (1.3.0)\n",
            "Requirement already satisfied: pyparsing>=2.1.4 in /usr/local/lib/python3.6/dist-packages (from pydot) (2.3.0)\n",
            "Reading package lists... Done\n",
            "Building dependency tree       \n",
            "Reading state information... Done\n",
            "graphviz is already the newest version (2.40.1-2).\n",
            "0 upgraded, 0 newly installed, 0 to remove and 8 not upgraded.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "_uuid": "8f2839f25d086af736a60e9eeb907d3b93b6e0e5",
        "_cell_guid": "b1076dfc-b9ad-4769-8c92-a6c4dae69d19",
        "trusted": true,
        "_kg_hide-input": true,
        "id": "9K7leB0DGZG9",
        "colab_type": "code",
        "outputId": "b520c044-e2b5-44fb-ac24-33a8de490c06",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        }
      },
      "cell_type": "code",
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "import seaborn as sns\n",
        "%matplotlib inline\n",
        "\n",
        "from nltk.tokenize import TweetTokenizer\n",
        "import datetime\n",
        "import lightgbm as lgb\n",
        "from scipy import stats\n",
        "from scipy.sparse import hstack, csr_matrix\n",
        "from sklearn.model_selection import train_test_split, cross_val_score\n",
        "from wordcloud import WordCloud\n",
        "from collections import Counter\n",
        "from nltk.corpus import stopwords\n",
        "from nltk.util import ngrams\n",
        "from sklearn.feature_extraction.text import TfidfVectorizer\n",
        "from sklearn.preprocessing import StandardScaler\n",
        "from sklearn.linear_model import LogisticRegression\n",
        "from sklearn.svm import LinearSVC\n",
        "from sklearn.multiclass import OneVsRestClassifier\n",
        "pd.set_option('max_colwidth',400)\n",
        "from google.colab import drive\n",
        "import nltk\n",
        "nltk.download('stopwords')\n",
        "from google.colab import files\n"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[nltk_data] Downloading package stopwords to /root/nltk_data...\n",
            "[nltk_data]   Unzipping corpora/stopwords.zip.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "yoOIXlOaxB9k",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "## Data Preparation:\n",
        "Import dataset are stored inside my google drive"
      ]
    },
    {
      "metadata": {
        "id": "NOvuWhQPLPmH",
        "colab_type": "code",
        "outputId": "155639e3-7355-4644-85dc-7f220c1718fa",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 124
        }
      },
      "cell_type": "code",
      "source": [
        "from google.colab import drive\n",
        "drive.mount('/content/drive')\n"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Go to this URL in a browser: https://accounts.google.com/o/oauth2/auth?client_id=947318989803-6bn6qk8qdgf4n4g3pfee6491hc0brc4i.apps.googleusercontent.com&redirect_uri=urn%3Aietf%3Awg%3Aoauth%3A2.0%3Aoob&scope=email%20https%3A%2F%2Fwww.googleapis.com%2Fauth%2Fdocs.test%20https%3A%2F%2Fwww.googleapis.com%2Fauth%2Fdrive%20https%3A%2F%2Fwww.googleapis.com%2Fauth%2Fdrive.photos.readonly%20https%3A%2F%2Fwww.googleapis.com%2Fauth%2Fpeopleapi.readonly&response_type=code\n",
            "\n",
            "Enter your authorization code:\n",
            "··········\n",
            "Mounted at /content/drive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "ZnmH5hSQxQYa",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "Read dataset: train test submission"
      ]
    },
    {
      "metadata": {
        "_cell_guid": "79c7e3d0-c299-4dcb-8224-4455121ee9b0",
        "_uuid": "d629ff2d2480ee46fbb7e2d37f6b5fab8052498a",
        "trusted": true,
        "scrolled": true,
        "id": "Qp7EQ0TrGZHB",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "train = pd.read_csv('/content/drive/My Drive/DeepLearning/train.tsv', sep=\"\\t\")\n",
        "test = pd.read_csv('/content/drive/My Drive/DeepLearning/test.tsv', sep=\"\\t\")\n",
        "sub = pd.read_csv('/content/drive/My Drive/DeepLearning/sampleSubmission.csv', sep=\",\")\n",
        "y = train['Sentiment']"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "Xyqvz9eLfST8",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "class2 = train[train['Sentiment']==2]\n",
        "class2Sample = class2.sample(frac=0.5) #, random_state=3\n",
        "train = pd.concat([train[train['Sentiment']!=2], class2Sample])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "rilTf7HZbM4y",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "This dataset is interesting for NLP researching. Sentences from original dataset were split in separate phrases and each of them has a sentiment label. Also a lot of phrases are really short which makes classifying them quite challenging.\n",
        "We can see than sentences were split in 18-20 phrases at average and a lot of phrases contain each other. Sometimes one word or even one punctuation mark influences the sentiment"
      ]
    },
    {
      "metadata": {
        "trusted": true,
        "_uuid": "f9b8d8423bb09068cb168b67f4756ee8b250fc8c",
        "id": "xBg-49HYGZHE",
        "colab_type": "code",
        "outputId": "4a3df630-c856-4270-ec54-198e3161f767",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 713
        }
      },
      "cell_type": "code",
      "source": [
        "#print(train.head(10))\n",
        "print(train.loc[train.SentenceId == 20])\n",
        "print('Average count of phrases per sentence in train is {0:.0f}.'.format(train.groupby('SentenceId')['Phrase'].count().mean()))\n",
        "print('Average count of phrases per sentence in test is {0:.0f}.'.format(test.groupby('SentenceId')['Phrase'].count().mean()))\n",
        "print('Number of phrases in train: {}. Number of sentences in train: {}.'.format(train.shape[0], len(train.SentenceId.unique())))\n",
        "print('Number of phrases in test: {}. Number of sentences in test: {}.'.format(test.shape[0], len(test.SentenceId.unique())))\n",
        "print('Average word length of phrases in train is {0:.0f}.'.format(np.mean(train['Phrase'].apply(lambda x: len(x.split())))))\n",
        "print('Average word length of phrases in test is {0:.0f}.'.format(np.mean(test['Phrase'].apply(lambda x: len(x.split())))))"
      ],
      "execution_count": 19,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "     PhraseId  SentenceId  \\\n",
            "536       537          20   \n",
            "537       538          20   \n",
            "538       539          20   \n",
            "539       540          20   \n",
            "540       541          20   \n",
            "541       542          20   \n",
            "542       543          20   \n",
            "543       544          20   \n",
            "544       545          20   \n",
            "545       546          20   \n",
            "546       547          20   \n",
            "547       548          20   \n",
            "548       549          20   \n",
            "549       550          20   \n",
            "550       551          20   \n",
            "\n",
            "                                                   Phrase  Sentiment  fold_id  \n",
            "536  It 's everything you 'd expect -- but nothing more .          2        0  \n",
            "537     's everything you 'd expect -- but nothing more .          1        0  \n",
            "538       's everything you 'd expect -- but nothing more          2        0  \n",
            "539          everything you 'd expect -- but nothing more          1        0  \n",
            "540                                            everything          2        0  \n",
            "541                     you 'd expect -- but nothing more          1        0  \n",
            "542                         'd expect -- but nothing more          2        0  \n",
            "543                                                    'd          2        0  \n",
            "544                            expect -- but nothing more          2        0  \n",
            "545                                 expect -- but nothing          2        0  \n",
            "546                                             expect --          2        0  \n",
            "547                                                expect          2        0  \n",
            "548                                           but nothing          2        0  \n",
            "549                                               nothing          1        0  \n",
            "550                                                  more          2        0  \n",
            "Average count of phrases per sentence in train is 18.\n",
            "Average count of phrases per sentence in test is 20.\n",
            "Number of phrases in train: 156060. Number of sentences in train: 8529.\n",
            "Number of phrases in test: 66292. Number of sentences in test: 3310.\n",
            "Average word length of phrases in train is 7.\n",
            "Average word length of phrases in test is 7.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "tI8pN6fob_mC",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "On Overlapping sentences in both Train and Test set"
      ]
    },
    {
      "metadata": {
        "id": "7K_b0M-0Eye9",
        "colab_type": "code",
        "outputId": "cd7b539e-acf5-43be-b654-07ab25094608",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        }
      },
      "cell_type": "code",
      "source": [
        "save_test = pd.merge(test, train[[\"Phrase\", \"Sentiment\"]], on=\"Phrase\", how=\"inner\")\n",
        "print (\"Number of overlapping phrases  \", save_test.shape[0])\n",
        "print (\"% of neutral sentiment phrases\",save_test[(save_test['Sentiment'] == 2)].count()[0] /save_test.shape[0])\n",
        "save_test = save_test[save_test[\"Sentiment\"].notnull()]\n",
        "save_test.drop(['SentenceId', 'Phrase'], axis=1,inplace=True)\n",
        "save_test = save_test[save_test[\"Sentiment\"].notnull()]\n",
        "\n",
        "import math\n",
        "def get_sentiment(row):\n",
        "    old_s = row['Sentiment_x']\n",
        "    new_s = row['Sentiment_y']\n",
        "    if math.isnan(new_s):\n",
        "        return int(old_s)\n",
        "    else:\n",
        "        return int(new_s)"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Number of overlapping phrases   6597\n",
            "% of neutral sentiment phrases 0.5787479157192663\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "4ggBRYFGKih5",
        "colab_type": "code",
        "outputId": "6d83aa7b-602d-49a4-967a-163cd7b2da89",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1303
        }
      },
      "cell_type": "code",
      "source": [
        "from wordcloud import WordCloud,STOPWORDS\n",
        "\n",
        "def wordcloud_draw(data, color = 'black'):\n",
        "    words = ' '.join(data)\n",
        "    cleaned_word = \" \".join([word for word in words.split()\n",
        "                            if 'http' not in word\n",
        "                                and not word.startswith('@')\n",
        "                                and not word.startswith('#')\n",
        "                                and word != 'RT'\n",
        "                            ])\n",
        "    wordcloud = WordCloud(stopwords=STOPWORDS,\n",
        "                      background_color=color,\n",
        "                      width=2500,\n",
        "                      height=2000\n",
        "                     ).generate(cleaned_word)\n",
        "    plt.figure(1,figsize=(13, 13))\n",
        "    plt.imshow(wordcloud)\n",
        "    plt.axis('off')\n",
        "    plt.show()\n",
        "train_pos = train[ train['Sentiment'] == 4]\n",
        "train_pos = train_pos['Phrase']\n",
        "train_neg = train[ train['Sentiment'] == 0]\n",
        "train_neg = train_neg['Phrase']    \n",
        "print(\"Positive words\")\n",
        "wordcloud_draw(train_pos,'white')\n",
        "print(\"Negative words\")\n",
        "wordcloud_draw(train_neg)"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Positive words\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAwwAAAJnCAYAAAA6O9RWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSUVORK5CYII=\n",
            "text/plain": [
              "<matplotlib.figure.Figure at 0x7f21a97efcf8>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "Negative words\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAwwAAAJnCAYAAAA6O9RWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
            "text/plain": [
              "<matplotlib.figure.Figure at 0x7f21a935f6a0>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "metadata": {
        "_uuid": "398461363c7a395e2a982e07e8ac6fccaee139c1",
        "id": "D-qyZxyVGZIG",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "## Models"
      ]
    },
    {
      "metadata": {
        "trusted": true,
        "_uuid": "eb29ec027df57f6597dbef976645dc8d151e1618",
        "id": "AwnzDD_0GZIH",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "from keras.preprocessing.text import Tokenizer\n",
        "from keras.preprocessing.sequence import pad_sequences\n",
        "from keras.layers import Dense, Input, LSTM, Embedding, Dropout, Activation, Conv1D, GRU, CuDNNGRU, CuDNNLSTM, BatchNormalization\n",
        "from keras.layers import Bidirectional, GlobalMaxPool1D, MaxPooling1D, Add, Flatten, LeakyReLU\n",
        "from keras.layers import GlobalAveragePooling1D, GlobalMaxPooling1D, concatenate, SpatialDropout1D\n",
        "from keras.models import Model, load_model\n",
        "from keras import initializers, regularizers, constraints, optimizers, layers, callbacks\n",
        "from keras import backend as K\n",
        "from keras.engine import InputSpec, Layer\n",
        "from keras.optimizers import Adam\n",
        "from keras.callbacks import ModelCheckpoint, TensorBoard, Callback, EarlyStopping\n",
        "import pydot\n",
        "from IPython.display import SVG\n",
        "from keras.utils.vis_utils import model_to_dot"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "6IRBrjIkWDq8",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "###Embedding & Word Vectorization"
      ]
    },
    {
      "metadata": {
        "trusted": true,
        "_uuid": "a2881c29f82578b4a373b52d2c7b96a2e73bfd80",
        "id": "TP2z0XJhGZIL",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "tk = Tokenizer(lower = True, filters='')\n",
        "full_text = list(train['Phrase'].values) + list(test['Phrase'].values)\n",
        "tk.fit_on_texts(full_text)\n",
        "train_tokenized = tk.texts_to_sequences(train['Phrase'])\n",
        "test_tokenized = tk.texts_to_sequences(test['Phrase'])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "DbNFsH5cHoiw",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "Pre-trained word2vec:  this model is trained on thecontext on each word so that similar words will havesimilar numerical representations. Sentences are firsttokenized to create a number of pairs of words, de-pending on the window size.  Then the data it’s fedinto  a  neural  network  through  an  embedding  layerinitialized with random weights.  Once the model istrained to minimize the loss of predicting the targetwords  using  the  context  words,  the  weights  in  theembedding layer would represent the vocabulary ofword vectors"
      ]
    },
    {
      "metadata": {
        "trusted": true,
        "_uuid": "9dfd0b8fa2c79bfa206d2fe8e35fbec444418f5c",
        "id": "x2VpwGsgGZIS",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "embedding_path = \"/content/drive/My Drive/DeepLearning/crawl-300d-2M.vec\"\n",
        "#embedding_path = \"/content/drive/My Drive/DeepLearning/glove.twitter.27B.25d.txt\"\n",
        "#embed_size = 25\n",
        "embed_size = 300\n",
        "max_features = 30000\n",
        "\n",
        "def get_coefs(word,*arr): return word, np.asarray(arr, dtype='float32')\n",
        "embedding_index = dict(get_coefs(*o.strip().split(\" \")) for o in open(embedding_path))\n",
        "\n",
        "word_index = tk.word_index\n",
        "nb_words = min(max_features, len(word_index))\n",
        "embedding_matrix = np.zeros((nb_words + 1, embed_size))\n",
        "for word, i in word_index.items():\n",
        "    if i >= max_features: continue\n",
        "    embedding_vector = embedding_index.get(word)\n",
        "    if embedding_vector is not None: embedding_matrix[i] = embedding_vector"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "trusted": true,
        "_uuid": "bcb80cf8a59ca779a0be1ab235a1e9da2f4b175b",
        "id": "b6bErvirGZIP",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "max_len = 50\n",
        "X_train = pad_sequences(train_tokenized, maxlen = max_len)\n",
        "X_test = pad_sequences(test_tokenized, maxlen = max_len)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "trusted": true,
        "_uuid": "365c0d607d55a78c5890268b9c168eb12a211855",
        "id": "4AlRADppGZIa",
        "colab_type": "code",
        "outputId": "93c9145a-d1e0-4c30-d9bd-f1468b43ab0f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 228
        }
      },
      "cell_type": "code",
      "source": [
        "from sklearn.preprocessing import OneHotEncoder\n",
        "ohe = OneHotEncoder(sparse=False)\n",
        "y_ohe = ohe.fit_transform(y.values.reshape(-1, 1))\n",
        "ohe.fit(train[\"Sentiment\"].values.reshape(-1, 1))"
      ],
      "execution_count": 21,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/sklearn/preprocessing/_encoders.py:368: FutureWarning: The handling of integer data will change in version 0.22. Currently, the categories are determined based on the range [0, max(values)], while in the future they will be determined based on the unique values.\n",
            "If you want the future behaviour and silence this warning, you can specify \"categories='auto'\".\n",
            "In case you used a LabelEncoder before this OneHotEncoder to convert the categories to integers, then you can now use the OneHotEncoder directly.\n",
            "  warnings.warn(msg, FutureWarning)\n",
            "/usr/local/lib/python3.6/dist-packages/sklearn/preprocessing/_encoders.py:368: FutureWarning: The handling of integer data will change in version 0.22. Currently, the categories are determined based on the range [0, max(values)], while in the future they will be determined based on the unique values.\n",
            "If you want the future behaviour and silence this warning, you can specify \"categories='auto'\".\n",
            "In case you used a LabelEncoder before this OneHotEncoder to convert the categories to integers, then you can now use the OneHotEncoder directly.\n",
            "  warnings.warn(msg, FutureWarning)\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "OneHotEncoder(categorical_features=None, categories=None,\n",
              "       dtype=<class 'numpy.float64'>, handle_unknown='error',\n",
              "       n_values=None, sparse=False)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 21
        }
      ]
    },
    {
      "metadata": {
        "id": "ynCM0G_xBJN1",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "file_path = \"/content/drive/My Drive/DeepLearning/best_model.hdf5\"\n",
        "check_point = ModelCheckpoint(file_path, monitor = \"val_loss\", verbose = 1,\n",
        "                              save_best_only = True, mode = \"min\")\n",
        "early_stop = EarlyStopping(monitor = \"val_loss\", mode = \"min\", patience = 2)\n",
        "NUM_FOLDS = 2\n",
        "train[\"fold_id\"] = train[\"SentenceId\"].apply(lambda x: x%NUM_FOLDS)\n",
        "\n",
        "def Train_And_Prediction(model):  \n",
        "  test_preds = np.zeros((test.shape[0], 5))\n",
        "  for i in range(NUM_FOLDS):\n",
        "      print(\"FOLD\", i+1)    \n",
        "      print(\"Splitting the data into train and validation...\")\n",
        "      train_seq, val_seq = X_train[train[\"fold_id\"] != i], X_train[train[\"fold_id\"] == i]\n",
        "      y_train = ohe.transform(train[train[\"fold_id\"] != i][\"Sentiment\"].values.reshape(-1, 1))\n",
        "      y_val = ohe.transform(train[train[\"fold_id\"] == i][\"Sentiment\"].values.reshape(-1, 1))        \n",
        "\n",
        "      print(\"Training the model...\")\n",
        "      model.fit(train_seq, y_train, validation_data = (val_seq, y_val), batch_size = 128, epochs = 1, verbose = 1, callbacks = [early_stop]) \n",
        "      model.save_weights(file_path)  \n",
        "      test_preds += model.predict([X_test], batch_size=1024, verbose=1)    \n",
        "      print()\n",
        "\n",
        "  print(\"Save model after cross-validation...\")\n",
        "  #model.save_weights(file_path) \n",
        "  model.save(file_path)\n",
        "  test_preds /= NUM_FOLDS\n",
        "\n",
        "\n",
        "  print(\"Make the submission ready...\")\n",
        "  sub = pd.read_csv('/content/drive/My Drive/DeepLearning/sampleSubmission.csv', sep=\",\")\n",
        "\n",
        "  pred = model.predict(X_test, batch_size = 1024, verbose = 1)\n",
        "  predictions = np.round(np.argmax(pred, axis=1)).astype(int)\n",
        "  submission = pd.DataFrame({'PhraseId':test['PhraseId'],'Sentiment': predictions})\n",
        "  submission =pd.merge(submission, save_test, on='PhraseId', how='left')\n",
        "  submission[\"Sentiment\"] = submission.apply(get_sentiment, axis=1)\n",
        "  submission.drop(['Sentiment_x', 'Sentiment_y'], axis=1,inplace=True)\n",
        "  submission[\"Sentiment\"] = submission[\"Sentiment\"].astype(int)\n",
        "  submission.to_csv(\"/content/drive/My Drive/DeepLearning/blend.csv\", index=False)\n",
        "\n",
        "  predictions = np.round(np.argmax(test_preds, axis=1)).astype(int)\n",
        "  submission = pd.DataFrame({'PhraseId':test['PhraseId'],'Sentiment': predictions})\n",
        "  submission =pd.merge(submission, save_test, on='PhraseId', how='left')\n",
        "  submission[\"Sentiment\"] = submission.apply(get_sentiment, axis=1)\n",
        "  submission.drop(['Sentiment_x', 'Sentiment_y'], axis=1,inplace=True)\n",
        "  submission[\"Sentiment\"] = submission[\"Sentiment\"].astype(int)\n",
        "  submission.to_csv(\"/content/drive/My Drive/DeepLearning/avg_blend.csv\", index=False)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "x2cpGj5WcMsH",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "###Model 1: Embedding + {LSTM/GRU} + CNN"
      ]
    },
    {
      "metadata": {
        "trusted": true,
        "_uuid": "8187e167ce93f0eb69f59cb9d7fedc4637a77cfe",
        "id": "jUBolZU1GZIl",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "def build_model1(lr=0.0, lr_d=0.0, units=0, spatial_dr=0.0, kernel_size1=3, kernel_size2=2, dense_units=128, dr=0.1, conv_size=32):\n",
        "    inp = Input(shape = (max_len,))\n",
        "    x = Embedding(19479, embed_size, weights = [embedding_matrix], trainable = False)(inp)\n",
        "    x1 = SpatialDropout1D(spatial_dr)(x)\n",
        "\n",
        "    x_gru = Bidirectional(CuDNNGRU(units, return_sequences = True))(x1)\n",
        "    x_lstm = Bidirectional(CuDNNLSTM(units, return_sequences = True))(x1)\n",
        "    \n",
        "    x_conv1 = Conv1D(conv_size, kernel_size=kernel_size1, padding='valid', kernel_initializer='he_uniform')(x_gru)\n",
        "    avg_pool1_gru = GlobalAveragePooling1D()(x_conv1)\n",
        "    max_pool1_gru = GlobalMaxPooling1D()(x_conv1)\n",
        "    \n",
        "    x_conv2 = Conv1D(conv_size, kernel_size=kernel_size2, padding='valid', kernel_initializer='he_uniform')(x_gru)\n",
        "    avg_pool2_gru = GlobalAveragePooling1D()(x_conv2)\n",
        "    max_pool2_gru = GlobalMaxPooling1D()(x_conv2)\n",
        "    \n",
        "    \n",
        "    x_conv3 = Conv1D(conv_size, kernel_size=kernel_size1, padding='valid', kernel_initializer='he_uniform')(x_lstm)\n",
        "    avg_pool1_lstm = GlobalAveragePooling1D()(x_conv3)\n",
        "    max_pool1_lstm = GlobalMaxPooling1D()(x_conv3)\n",
        "    \n",
        "    x_conv4 = Conv1D(conv_size, kernel_size=kernel_size2, padding='valid', kernel_initializer='he_uniform')(x_lstm)\n",
        "    avg_pool2_lstm = GlobalAveragePooling1D()(x_conv4)\n",
        "    max_pool2_lstm = GlobalMaxPooling1D()(x_conv4)\n",
        "    \n",
        "    \n",
        "    x = concatenate([avg_pool1_gru, max_pool1_gru, avg_pool2_gru, max_pool2_gru,\n",
        "                    avg_pool1_lstm, max_pool1_lstm, avg_pool2_lstm, max_pool2_lstm])\n",
        "    x = BatchNormalization()(x)\n",
        "    x = Dropout(dr)(Dense(dense_units, activation='relu') (x))\n",
        "    x = BatchNormalization()(x)\n",
        "    x = Dropout(dr)(Dense(int(dense_units / 2), activation='relu') (x))\n",
        "    x = Dense(5, activation = \"sigmoid\")(x)\n",
        "    model = Model(inputs = inp, outputs = x)\n",
        "    model.compile(loss = \"binary_crossentropy\", optimizer = Adam(lr = lr, decay = lr_d), metrics = [\"accuracy\"])\n",
        "    return model"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "wjh0IK94gE7e",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "Print the Summary and Arch of the model"
      ]
    },
    {
      "metadata": {
        "id": "yrvsOTvmOPvZ",
        "colab_type": "code",
        "outputId": "28df70e8-e2ef-4f52-f0c7-fb1eca8c98be",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 2663
        }
      },
      "cell_type": "code",
      "source": [
        "model1 = build_model1(lr = 1e-3, lr_d = 0, units = 64, spatial_dr = 0.5, kernel_size1=3, kernel_size2=3, dense_units=64, dr=0.3, conv_size=32)\n",
        "model1.summary()\n",
        "SVG(model_to_dot(model1,  show_shapes=True, show_layer_names=True, rankdir='HB').create(prog='dot', format='svg'))"
      ],
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "__________________________________________________________________________________________________\n",
            "Layer (type)                    Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            "input_3 (InputLayer)            (None, 50)           0                                            \n",
            "__________________________________________________________________________________________________\n",
            "embedding_3 (Embedding)         (None, 50, 300)      5843700     input_3[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "spatial_dropout1d_3 (SpatialDro (None, 50, 300)      0           embedding_3[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "bidirectional_5 (Bidirectional) (None, 50, 128)      140544      spatial_dropout1d_3[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "bidirectional_6 (Bidirectional) (None, 50, 128)      187392      spatial_dropout1d_3[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv1d_9 (Conv1D)               (None, 48, 32)       12320       bidirectional_5[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv1d_10 (Conv1D)              (None, 48, 32)       12320       bidirectional_5[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv1d_11 (Conv1D)              (None, 48, 32)       12320       bidirectional_6[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv1d_12 (Conv1D)              (None, 48, 32)       12320       bidirectional_6[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "global_average_pooling1d_9 (Glo (None, 32)           0           conv1d_9[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "global_max_pooling1d_9 (GlobalM (None, 32)           0           conv1d_9[0][0]                   \n",
            "__________________________________________________________________________________________________\n",
            "global_average_pooling1d_10 (Gl (None, 32)           0           conv1d_10[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "global_max_pooling1d_10 (Global (None, 32)           0           conv1d_10[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "global_average_pooling1d_11 (Gl (None, 32)           0           conv1d_11[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "global_max_pooling1d_11 (Global (None, 32)           0           conv1d_11[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "global_average_pooling1d_12 (Gl (None, 32)           0           conv1d_12[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "global_max_pooling1d_12 (Global (None, 32)           0           conv1d_12[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_3 (Concatenate)     (None, 256)          0           global_average_pooling1d_9[0][0] \n",
            "                                                                 global_max_pooling1d_9[0][0]     \n",
            "                                                                 global_average_pooling1d_10[0][0]\n",
            "                                                                 global_max_pooling1d_10[0][0]    \n",
            "                                                                 global_average_pooling1d_11[0][0]\n",
            "                                                                 global_max_pooling1d_11[0][0]    \n",
            "                                                                 global_average_pooling1d_12[0][0]\n",
            "                                                                 global_max_pooling1d_12[0][0]    \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_5 (BatchNor (None, 256)          1024        concatenate_3[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "dense_7 (Dense)                 (None, 64)           16448       batch_normalization_5[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "dropout_5 (Dropout)             (None, 64)           0           dense_7[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_6 (BatchNor (None, 64)           256         dropout_5[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "dense_8 (Dense)                 (None, 32)           2080        batch_normalization_6[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "dropout_6 (Dropout)             (None, 32)           0           dense_8[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "dense_9 (Dense)                 (None, 5)            165         dropout_6[0][0]                  \n",
            "==================================================================================================\n",
            "Total params: 6,240,889\n",
            "Trainable params: 396,549\n",
            "Non-trainable params: 5,844,340\n",
            "__________________________________________________________________________________________________\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<IPython.core.display.SVG object>"
            ],
            "image/svg+xml": "<svg height=\"1134pt\" viewBox=\"0.00 0.00 4047.50 1134.00\" width=\"4048pt\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n<g class=\"graph\" id=\"graph0\" transform=\"scale(1 1) rotate(0) translate(4 1130)\">\n<title>G</title>\n<polygon fill=\"#ffffff\" points=\"-4,4 -4,-1130 4043.5,-1130 4043.5,4 -4,4\" stroke=\"transparent\"/>\n<!-- 140068280049224 -->\n<g class=\"node\" id=\"node1\">\n<title>140068280049224</title>\n<polygon fill=\"none\" points=\"1865.5,-1079.5 1865.5,-1125.5 2136.5,-1125.5 2136.5,-1079.5 1865.5,-1079.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1932\" y=\"-1098.8\">input_3: InputLayer</text>\n<polyline fill=\"none\" points=\"1998.5,-1079.5 1998.5,-1125.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2027.5\" y=\"-1110.3\">input:</text>\n<polyline fill=\"none\" points=\"1998.5,-1102.5 2056.5,-1102.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2027.5\" y=\"-1087.3\">output:</text>\n<polyline fill=\"none\" points=\"2056.5,-1079.5 2056.5,-1125.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2096.5\" y=\"-1110.3\">(None, 50)</text>\n<polyline fill=\"none\" points=\"2056.5,-1102.5 2136.5,-1102.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2096.5\" y=\"-1087.3\">(None, 50)</text>\n</g>\n<!-- 140068280049448 -->\n<g class=\"node\" id=\"node2\">\n<title>140068280049448</title>\n<polygon fill=\"none\" points=\"1831.5,-996.5 1831.5,-1042.5 2170.5,-1042.5 2170.5,-996.5 1831.5,-996.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1917\" y=\"-1015.8\">embedding_3: Embedding</text>\n<polyline fill=\"none\" points=\"2002.5,-996.5 2002.5,-1042.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2031.5\" y=\"-1027.3\">input:</text>\n<polyline fill=\"none\" points=\"2002.5,-1019.5 2060.5,-1019.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2031.5\" y=\"-1004.3\">output:</text>\n<polyline fill=\"none\" points=\"2060.5,-996.5 2060.5,-1042.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2115.5\" y=\"-1027.3\">(None, 50)</text>\n<polyline fill=\"none\" points=\"2060.5,-1019.5 2170.5,-1019.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2115.5\" y=\"-1004.3\">(None, 50, 300)</text>\n</g>\n<!-- 140068280049224&#45;&gt;140068280049448 -->\n<g class=\"edge\" id=\"edge1\">\n<title>140068280049224-&gt;140068280049448</title>\n<path d=\"M2001,-1079.3799C2001,-1071.1745 2001,-1061.7679 2001,-1052.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2004.5001,-1052.784 2001,-1042.784 1997.5001,-1052.784 2004.5001,-1052.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140068280049392 -->\n<g class=\"node\" id=\"node3\">\n<title>140068280049392</title>\n<polygon fill=\"none\" points=\"1792.5,-913.5 1792.5,-959.5 2209.5,-959.5 2209.5,-913.5 1792.5,-913.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1917\" y=\"-932.8\">spatial_dropout1d_3: SpatialDropout1D</text>\n<polyline fill=\"none\" points=\"2041.5,-913.5 2041.5,-959.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2070.5\" y=\"-944.3\">input:</text>\n<polyline fill=\"none\" points=\"2041.5,-936.5 2099.5,-936.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2070.5\" y=\"-921.3\">output:</text>\n<polyline fill=\"none\" points=\"2099.5,-913.5 2099.5,-959.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2154.5\" y=\"-944.3\">(None, 50, 300)</text>\n<polyline fill=\"none\" points=\"2099.5,-936.5 2209.5,-936.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2154.5\" y=\"-921.3\">(None, 50, 300)</text>\n</g>\n<!-- 140068280049448&#45;&gt;140068280049392 -->\n<g class=\"edge\" id=\"edge2\">\n<title>140068280049448-&gt;140068280049392</title>\n<path d=\"M2001,-996.3799C2001,-988.1745 2001,-978.7679 2001,-969.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2004.5001,-969.784 2001,-959.784 1997.5001,-969.784 2004.5001,-969.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140068279531280 -->\n<g class=\"node\" id=\"node4\">\n<title>140068279531280</title>\n<polygon fill=\"none\" points=\"1228.5,-830.5 1228.5,-876.5 1755.5,-876.5 1755.5,-830.5 1228.5,-830.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1408\" y=\"-849.8\">bidirectional_5(cu_dnngru_3): Bidirectional(CuDNNGRU)</text>\n<polyline fill=\"none\" points=\"1587.5,-830.5 1587.5,-876.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1616.5\" y=\"-861.3\">input:</text>\n<polyline fill=\"none\" points=\"1587.5,-853.5 1645.5,-853.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1616.5\" y=\"-838.3\">output:</text>\n<polyline fill=\"none\" points=\"1645.5,-830.5 1645.5,-876.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1700.5\" y=\"-861.3\">(None, 50, 300)</text>\n<polyline fill=\"none\" points=\"1645.5,-853.5 1755.5,-853.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1700.5\" y=\"-838.3\">(None, 50, 128)</text>\n</g>\n<!-- 140068280049392&#45;&gt;140068279531280 -->\n<g class=\"edge\" id=\"edge3\">\n<title>140068280049392-&gt;140068279531280</title>\n<path d=\"M1859.8911,-913.4901C1792.5904,-902.5157 1711.9918,-889.3729 1643.4116,-878.1899\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"1643.8164,-874.7098 1633.3834,-876.5547 1642.6898,-881.6185 1643.8164,-874.7098\" stroke=\"#000000\"/>\n</g>\n<!-- 140067975612624 -->\n<g class=\"node\" id=\"node5\">\n<title>140067975612624</title>\n<polygon fill=\"none\" points=\"2261,-830.5 2261,-876.5 2801,-876.5 2801,-830.5 2261,-830.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2447\" y=\"-849.8\">bidirectional_6(cu_dnnlstm_3): Bidirectional(CuDNNLSTM)</text>\n<polyline fill=\"none\" points=\"2633,-830.5 2633,-876.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2662\" y=\"-861.3\">input:</text>\n<polyline fill=\"none\" points=\"2633,-853.5 2691,-853.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2662\" y=\"-838.3\">output:</text>\n<polyline fill=\"none\" points=\"2691,-830.5 2691,-876.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2746\" y=\"-861.3\">(None, 50, 300)</text>\n<polyline fill=\"none\" points=\"2691,-853.5 2801,-853.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2746\" y=\"-838.3\">(None, 50, 128)</text>\n</g>\n<!-- 140068280049392&#45;&gt;140067975612624 -->\n<g class=\"edge\" id=\"edge4\">\n<title>140068280049392-&gt;140067975612624</title>\n<path d=\"M2147.9307,-913.4901C2218.1517,-902.4932 2302.2759,-889.319 2373.7804,-878.1212\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2374.4454,-881.5598 2383.7835,-876.5547 2373.3623,-874.6441 2374.4454,-881.5598\" stroke=\"#000000\"/>\n</g>\n<!-- 140067975713400 -->\n<g class=\"node\" id=\"node6\">\n<title>140067975713400</title>\n<polygon fill=\"none\" points=\"603.5,-747.5 603.5,-793.5 904.5,-793.5 904.5,-747.5 603.5,-747.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"670\" y=\"-766.8\">conv1d_9: Conv1D</text>\n<polyline fill=\"none\" points=\"736.5,-747.5 736.5,-793.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"765.5\" y=\"-778.3\">input:</text>\n<polyline fill=\"none\" points=\"736.5,-770.5 794.5,-770.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"765.5\" y=\"-755.3\">output:</text>\n<polyline fill=\"none\" points=\"794.5,-747.5 794.5,-793.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"849.5\" y=\"-778.3\">(None, 50, 128)</text>\n<polyline fill=\"none\" points=\"794.5,-770.5 904.5,-770.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"849.5\" y=\"-755.3\">(None, 48, 32)</text>\n</g>\n<!-- 140068279531280&#45;&gt;140067975713400 -->\n<g class=\"edge\" id=\"edge5\">\n<title>140068279531280-&gt;140067975713400</title>\n<path d=\"M1287.4059,-830.4901C1169.8223,-817.2659 1024.2425,-800.8931 914.8916,-788.5949\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"915.0867,-785.0948 904.7582,-787.4552 914.3044,-792.051 915.0867,-785.0948\" stroke=\"#000000\"/>\n</g>\n<!-- 140067973474288 -->\n<g class=\"node\" id=\"node7\">\n<title>140067973474288</title>\n<polygon fill=\"none\" points=\"1338,-747.5 1338,-793.5 1646,-793.5 1646,-747.5 1338,-747.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1408\" y=\"-766.8\">conv1d_10: Conv1D</text>\n<polyline fill=\"none\" points=\"1478,-747.5 1478,-793.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1507\" y=\"-778.3\">input:</text>\n<polyline fill=\"none\" points=\"1478,-770.5 1536,-770.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1507\" y=\"-755.3\">output:</text>\n<polyline fill=\"none\" points=\"1536,-747.5 1536,-793.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1591\" y=\"-778.3\">(None, 50, 128)</text>\n<polyline fill=\"none\" points=\"1536,-770.5 1646,-770.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1591\" y=\"-755.3\">(None, 48, 32)</text>\n</g>\n<!-- 140068279531280&#45;&gt;140067973474288 -->\n<g class=\"edge\" id=\"edge6\">\n<title>140068279531280-&gt;140067973474288</title>\n<path d=\"M1492,-830.3799C1492,-822.1745 1492,-812.7679 1492,-803.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"1495.5001,-803.784 1492,-793.784 1488.5001,-803.784 1495.5001,-803.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140067973189984 -->\n<g class=\"node\" id=\"node8\">\n<title>140067973189984</title>\n<polygon fill=\"none\" points=\"2377,-747.5 2377,-793.5 2685,-793.5 2685,-747.5 2377,-747.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2447\" y=\"-766.8\">conv1d_11: Conv1D</text>\n<polyline fill=\"none\" points=\"2517,-747.5 2517,-793.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2546\" y=\"-778.3\">input:</text>\n<polyline fill=\"none\" points=\"2517,-770.5 2575,-770.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2546\" y=\"-755.3\">output:</text>\n<polyline fill=\"none\" points=\"2575,-747.5 2575,-793.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2630\" y=\"-778.3\">(None, 50, 128)</text>\n<polyline fill=\"none\" points=\"2575,-770.5 2685,-770.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2630\" y=\"-755.3\">(None, 48, 32)</text>\n</g>\n<!-- 140067975612624&#45;&gt;140067973189984 -->\n<g class=\"edge\" id=\"edge7\">\n<title>140067975612624-&gt;140067973189984</title>\n<path d=\"M2531,-830.3799C2531,-822.1745 2531,-812.7679 2531,-803.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2534.5001,-803.784 2531,-793.784 2527.5001,-803.784 2534.5001,-803.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140067972907360 -->\n<g class=\"node\" id=\"node9\">\n<title>140067972907360</title>\n<polygon fill=\"none\" points=\"3141,-747.5 3141,-793.5 3449,-793.5 3449,-747.5 3141,-747.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3211\" y=\"-766.8\">conv1d_12: Conv1D</text>\n<polyline fill=\"none\" points=\"3281,-747.5 3281,-793.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3310\" y=\"-778.3\">input:</text>\n<polyline fill=\"none\" points=\"3281,-770.5 3339,-770.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3310\" y=\"-755.3\">output:</text>\n<polyline fill=\"none\" points=\"3339,-747.5 3339,-793.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3394\" y=\"-778.3\">(None, 50, 128)</text>\n<polyline fill=\"none\" points=\"3339,-770.5 3449,-770.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3394\" y=\"-755.3\">(None, 48, 32)</text>\n</g>\n<!-- 140067975612624&#45;&gt;140067972907360 -->\n<g class=\"edge\" id=\"edge8\">\n<title>140067975612624-&gt;140067972907360</title>\n<path d=\"M2742.802,-830.4901C2865.2937,-817.1827 3017.1336,-800.6871 3130.5716,-788.3633\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"3131.1723,-791.8187 3140.7358,-787.2591 3130.4163,-784.8597 3131.1723,-791.8187\" stroke=\"#000000\"/>\n</g>\n<!-- 140067973474232 -->\n<g class=\"node\" id=\"node10\">\n<title>140067973474232</title>\n<polygon fill=\"none\" points=\"0,-664.5 0,-710.5 504,-710.5 504,-664.5 0,-664.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"172\" y=\"-683.8\">global_average_pooling1d_9: GlobalAveragePooling1D</text>\n<polyline fill=\"none\" points=\"344,-664.5 344,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"373\" y=\"-695.3\">input:</text>\n<polyline fill=\"none\" points=\"344,-687.5 402,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"373\" y=\"-672.3\">output:</text>\n<polyline fill=\"none\" points=\"402,-664.5 402,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"453\" y=\"-695.3\">(None, 48, 32)</text>\n<polyline fill=\"none\" points=\"402,-687.5 504,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"453\" y=\"-672.3\">(None, 32)</text>\n</g>\n<!-- 140067975713400&#45;&gt;140067973474232 -->\n<g class=\"edge\" id=\"edge9\">\n<title>140067975713400-&gt;140067973474232</title>\n<path d=\"M614.8317,-747.4901C548.4565,-736.5157 468.9664,-723.3729 401.3293,-712.1899\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"401.8761,-708.7329 391.4391,-710.5547 400.7342,-715.6391 401.8761,-708.7329\" stroke=\"#000000\"/>\n</g>\n<!-- 140067973474344 -->\n<g class=\"node\" id=\"node11\">\n<title>140067973474344</title>\n<polygon fill=\"none\" points=\"522,-664.5 522,-710.5 986,-710.5 986,-664.5 522,-664.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"674\" y=\"-683.8\">global_max_pooling1d_9: GlobalMaxPooling1D</text>\n<polyline fill=\"none\" points=\"826,-664.5 826,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"855\" y=\"-695.3\">input:</text>\n<polyline fill=\"none\" points=\"826,-687.5 884,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"855\" y=\"-672.3\">output:</text>\n<polyline fill=\"none\" points=\"884,-664.5 884,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"935\" y=\"-695.3\">(None, 48, 32)</text>\n<polyline fill=\"none\" points=\"884,-687.5 986,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"935\" y=\"-672.3\">(None, 32)</text>\n</g>\n<!-- 140067975713400&#45;&gt;140067973474344 -->\n<g class=\"edge\" id=\"edge10\">\n<title>140067975713400-&gt;140067973474344</title>\n<path d=\"M754,-747.3799C754,-739.1745 754,-729.7679 754,-720.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"757.5001,-720.784 754,-710.784 750.5001,-720.784 757.5001,-720.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140067973190376 -->\n<g class=\"node\" id=\"node12\">\n<title>140067973190376</title>\n<polygon fill=\"none\" points=\"1003.5,-664.5 1003.5,-710.5 1514.5,-710.5 1514.5,-664.5 1003.5,-664.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1179\" y=\"-683.8\">global_average_pooling1d_10: GlobalAveragePooling1D</text>\n<polyline fill=\"none\" points=\"1354.5,-664.5 1354.5,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1383.5\" y=\"-695.3\">input:</text>\n<polyline fill=\"none\" points=\"1354.5,-687.5 1412.5,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1383.5\" y=\"-672.3\">output:</text>\n<polyline fill=\"none\" points=\"1412.5,-664.5 1412.5,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1463.5\" y=\"-695.3\">(None, 48, 32)</text>\n<polyline fill=\"none\" points=\"1412.5,-687.5 1514.5,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1463.5\" y=\"-672.3\">(None, 32)</text>\n</g>\n<!-- 140067973474288&#45;&gt;140067973190376 -->\n<g class=\"edge\" id=\"edge11\">\n<title>140067973474288-&gt;140067973190376</title>\n<path d=\"M1427.4059,-747.4901C1398.2397,-737.1004 1363.6176,-724.7672 1333.3731,-713.9934\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"1334.3144,-710.6134 1323.7197,-710.5547 1331.9654,-717.2075 1334.3144,-710.6134\" stroke=\"#000000\"/>\n</g>\n<!-- 140067973190152 -->\n<g class=\"node\" id=\"node13\">\n<title>140067973190152</title>\n<polygon fill=\"none\" points=\"1532.5,-664.5 1532.5,-710.5 2003.5,-710.5 2003.5,-664.5 1532.5,-664.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1688\" y=\"-683.8\">global_max_pooling1d_10: GlobalMaxPooling1D</text>\n<polyline fill=\"none\" points=\"1843.5,-664.5 1843.5,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1872.5\" y=\"-695.3\">input:</text>\n<polyline fill=\"none\" points=\"1843.5,-687.5 1901.5,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1872.5\" y=\"-672.3\">output:</text>\n<polyline fill=\"none\" points=\"1901.5,-664.5 1901.5,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1952.5\" y=\"-695.3\">(None, 48, 32)</text>\n<polyline fill=\"none\" points=\"1901.5,-687.5 2003.5,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1952.5\" y=\"-672.3\">(None, 32)</text>\n</g>\n<!-- 140067973474288&#45;&gt;140067973190152 -->\n<g class=\"edge\" id=\"edge12\">\n<title>140067973474288-&gt;140067973190152</title>\n<path d=\"M1568.5149,-747.4901C1603.6619,-736.9205 1645.4975,-724.3395 1681.7578,-713.4352\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"1682.7679,-716.7863 1691.3363,-710.5547 1680.752,-710.0829 1682.7679,-716.7863\" stroke=\"#000000\"/>\n</g>\n<!-- 140067973285760 -->\n<g class=\"node\" id=\"node14\">\n<title>140067973285760</title>\n<polygon fill=\"none\" points=\"2022,-664.5 2022,-710.5 2532,-710.5 2532,-664.5 2022,-664.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2197\" y=\"-683.8\">global_average_pooling1d_11: GlobalAveragePooling1D</text>\n<polyline fill=\"none\" points=\"2372,-664.5 2372,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2401\" y=\"-695.3\">input:</text>\n<polyline fill=\"none\" points=\"2372,-687.5 2430,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2401\" y=\"-672.3\">output:</text>\n<polyline fill=\"none\" points=\"2430,-664.5 2430,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2481\" y=\"-695.3\">(None, 48, 32)</text>\n<polyline fill=\"none\" points=\"2430,-687.5 2532,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2481\" y=\"-672.3\">(None, 32)</text>\n</g>\n<!-- 140067973189984&#45;&gt;140067973285760 -->\n<g class=\"edge\" id=\"edge13\">\n<title>140067973189984-&gt;140067973285760</title>\n<path d=\"M2460.5841,-747.4901C2428.5139,-737.0105 2390.3927,-724.5535 2357.221,-713.7139\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2358.1454,-710.3339 2347.5528,-710.5547 2355.9711,-716.9877 2358.1454,-710.3339\" stroke=\"#000000\"/>\n</g>\n<!-- 140067973287328 -->\n<g class=\"node\" id=\"node15\">\n<title>140067973287328</title>\n<polygon fill=\"none\" points=\"2550.5,-664.5 2550.5,-710.5 3021.5,-710.5 3021.5,-664.5 2550.5,-664.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2706\" y=\"-683.8\">global_max_pooling1d_11: GlobalMaxPooling1D</text>\n<polyline fill=\"none\" points=\"2861.5,-664.5 2861.5,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2890.5\" y=\"-695.3\">input:</text>\n<polyline fill=\"none\" points=\"2861.5,-687.5 2919.5,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2890.5\" y=\"-672.3\">output:</text>\n<polyline fill=\"none\" points=\"2919.5,-664.5 2919.5,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2970.5\" y=\"-695.3\">(None, 48, 32)</text>\n<polyline fill=\"none\" points=\"2919.5,-687.5 3021.5,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2970.5\" y=\"-672.3\">(None, 32)</text>\n</g>\n<!-- 140067973189984&#45;&gt;140067973287328 -->\n<g class=\"edge\" id=\"edge14\">\n<title>140067973189984-&gt;140067973287328</title>\n<path d=\"M2601.6931,-747.4901C2633.8895,-737.0105 2672.1609,-724.5535 2705.4632,-713.7139\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2706.7437,-716.978 2715.1694,-710.5547 2704.5771,-710.3217 2706.7437,-716.978\" stroke=\"#000000\"/>\n</g>\n<!-- 140067972996792 -->\n<g class=\"node\" id=\"node16\">\n<title>140067972996792</title>\n<polygon fill=\"none\" points=\"3039.5,-664.5 3039.5,-710.5 3550.5,-710.5 3550.5,-664.5 3039.5,-664.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3215\" y=\"-683.8\">global_average_pooling1d_12: GlobalAveragePooling1D</text>\n<polyline fill=\"none\" points=\"3390.5,-664.5 3390.5,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3419.5\" y=\"-695.3\">input:</text>\n<polyline fill=\"none\" points=\"3390.5,-687.5 3448.5,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3419.5\" y=\"-672.3\">output:</text>\n<polyline fill=\"none\" points=\"3448.5,-664.5 3448.5,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3499.5\" y=\"-695.3\">(None, 48, 32)</text>\n<polyline fill=\"none\" points=\"3448.5,-687.5 3550.5,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3499.5\" y=\"-672.3\">(None, 32)</text>\n</g>\n<!-- 140067972907360&#45;&gt;140067972996792 -->\n<g class=\"edge\" id=\"edge15\">\n<title>140067972907360-&gt;140067972996792</title>\n<path d=\"M3295,-747.3799C3295,-739.1745 3295,-729.7679 3295,-720.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"3298.5001,-720.784 3295,-710.784 3291.5001,-720.784 3298.5001,-720.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140067972996400 -->\n<g class=\"node\" id=\"node17\">\n<title>140067972996400</title>\n<polygon fill=\"none\" points=\"3568.5,-664.5 3568.5,-710.5 4039.5,-710.5 4039.5,-664.5 3568.5,-664.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3724\" y=\"-683.8\">global_max_pooling1d_12: GlobalMaxPooling1D</text>\n<polyline fill=\"none\" points=\"3879.5,-664.5 3879.5,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3908.5\" y=\"-695.3\">input:</text>\n<polyline fill=\"none\" points=\"3879.5,-687.5 3937.5,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3908.5\" y=\"-672.3\">output:</text>\n<polyline fill=\"none\" points=\"3937.5,-664.5 3937.5,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3988.5\" y=\"-695.3\">(None, 48, 32)</text>\n<polyline fill=\"none\" points=\"3937.5,-687.5 4039.5,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3988.5\" y=\"-672.3\">(None, 32)</text>\n</g>\n<!-- 140067972907360&#45;&gt;140067972996400 -->\n<g class=\"edge\" id=\"edge16\">\n<title>140067972907360-&gt;140067972996400</title>\n<path d=\"M3436.1089,-747.4901C3503.4096,-736.5157 3584.0082,-723.3729 3652.5884,-712.1899\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"3653.3102,-715.6185 3662.6166,-710.5547 3652.1836,-708.7098 3653.3102,-715.6185\" stroke=\"#000000\"/>\n</g>\n<!-- 140067972996232 -->\n<g class=\"node\" id=\"node18\">\n<title>140067972996232</title>\n<polygon fill=\"none\" points=\"1614.5,-581.5 1614.5,-627.5 2429.5,-627.5 2429.5,-581.5 1614.5,-581.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1702\" y=\"-600.8\">concatenate_3: Concatenate</text>\n<polyline fill=\"none\" points=\"1789.5,-581.5 1789.5,-627.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1818.5\" y=\"-612.3\">input:</text>\n<polyline fill=\"none\" points=\"1789.5,-604.5 1847.5,-604.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1818.5\" y=\"-589.3\">output:</text>\n<polyline fill=\"none\" points=\"1847.5,-581.5 1847.5,-627.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2138.5\" y=\"-612.3\">[(None, 32), (None, 32), (None, 32), (None, 32), (None, 32), (None, 32), (None, 32), (None, 32)]</text>\n<polyline fill=\"none\" points=\"1847.5,-604.5 2429.5,-604.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2138.5\" y=\"-589.3\">(None, 256)</text>\n</g>\n<!-- 140067973474232&#45;&gt;140067972996232 -->\n<g class=\"edge\" id=\"edge17\">\n<title>140067973474232-&gt;140067972996232</title>\n<path d=\"M504.0945,-664.5887C507.0795,-664.3873 510.0489,-664.1909 513,-664 881.3206,-640.1723 1300.8747,-624.5579 1603.936,-615.3293\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"1604.328,-618.8191 1614.2172,-615.0173 1604.1156,-611.8223 1604.328,-618.8191\" stroke=\"#000000\"/>\n</g>\n<!-- 140067973474344&#45;&gt;140067972996232 -->\n<g class=\"edge\" id=\"edge18\">\n<title>140067973474344-&gt;140067972996232</title>\n<path d=\"M986.1291,-664.6213C988.7669,-664.4104 991.3913,-664.2031 994,-664 1194.5971,-648.3801 1417.5102,-634.994 1604.3444,-624.8967\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"1604.5412,-628.3913 1614.3383,-624.3579 1604.1643,-621.4015 1604.5412,-628.3913\" stroke=\"#000000\"/>\n</g>\n<!-- 140067973190376&#45;&gt;140067972996232 -->\n<g class=\"edge\" id=\"edge19\">\n<title>140067973190376-&gt;140067972996232</title>\n<path d=\"M1470.5248,-664.4901C1573.1704,-653.3242 1696.4516,-639.9135 1800.3971,-628.6062\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"1800.8318,-632.0797 1810.3946,-627.5187 1800.0747,-625.1207 1800.8318,-632.0797\" stroke=\"#000000\"/>\n</g>\n<!-- 140067973190152&#45;&gt;140067972996232 -->\n<g class=\"edge\" id=\"edge20\">\n<title>140067973190152-&gt;140067972996232</title>\n<path d=\"M1838.4159,-664.4901C1870.4861,-654.0105 1908.6073,-641.5535 1941.779,-630.7139\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"1943.0289,-633.9877 1951.4472,-627.5547 1940.8546,-627.3339 1943.0289,-633.9877\" stroke=\"#000000\"/>\n</g>\n<!-- 140067973285760&#45;&gt;140067972996232 -->\n<g class=\"edge\" id=\"edge21\">\n<title>140067973285760-&gt;140067972996232</title>\n<path d=\"M2206.3069,-664.4901C2174.1105,-654.0105 2135.8391,-641.5535 2102.5368,-630.7139\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2103.4229,-627.3217 2092.8306,-627.5547 2101.2563,-633.978 2103.4229,-627.3217\" stroke=\"#000000\"/>\n</g>\n<!-- 140067973287328&#45;&gt;140067972996232 -->\n<g class=\"edge\" id=\"edge22\">\n<title>140067973287328-&gt;140067972996232</title>\n<path d=\"M2574.198,-664.4901C2471.4178,-653.3242 2347.9751,-639.9135 2243.8934,-628.6062\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2244.2023,-625.1193 2233.8827,-627.5187 2243.4462,-632.0783 2244.2023,-625.1193\" stroke=\"#000000\"/>\n</g>\n<!-- 140067972996792&#45;&gt;140067972996232 -->\n<g class=\"edge\" id=\"edge23\">\n<title>140067972996792-&gt;140067972996232</title>\n<path d=\"M3039.3852,-664.701C3036.2387,-664.4634 3033.1092,-664.2296 3030,-664 2836.12,-649.6844 2621.0581,-636.6061 2439.5484,-626.3796\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2439.745,-622.8852 2429.5642,-625.818 2439.3518,-629.8741 2439.745,-622.8852\" stroke=\"#000000\"/>\n</g>\n<!-- 140067972996400&#45;&gt;140067972996232 -->\n<g class=\"edge\" id=\"edge24\">\n<title>140067972996400-&gt;140067972996232</title>\n<path d=\"M3568.3286,-664.5751C3565.537,-664.378 3562.7599,-664.1861 3560,-664 3181.1977,-638.4523 2749.1155,-622.9349 2439.5791,-614.1836\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2439.6336,-610.6839 2429.5391,-613.9009 2439.4366,-617.6811 2439.6336,-610.6839\" stroke=\"#000000\"/>\n</g>\n<!-- 140067972599536 -->\n<g class=\"node\" id=\"node19\">\n<title>140067972599536</title>\n<polygon fill=\"none\" points=\"1814,-498.5 1814,-544.5 2230,-544.5 2230,-498.5 1814,-498.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1949.5\" y=\"-517.8\">batch_normalization_5: BatchNormalization</text>\n<polyline fill=\"none\" points=\"2085,-498.5 2085,-544.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2114\" y=\"-529.3\">input:</text>\n<polyline fill=\"none\" points=\"2085,-521.5 2143,-521.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2114\" y=\"-506.3\">output:</text>\n<polyline fill=\"none\" points=\"2143,-498.5 2143,-544.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2186.5\" y=\"-529.3\">(None, 256)</text>\n<polyline fill=\"none\" points=\"2143,-521.5 2230,-521.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2186.5\" y=\"-506.3\">(None, 256)</text>\n</g>\n<!-- 140067972996232&#45;&gt;140067972599536 -->\n<g class=\"edge\" id=\"edge25\">\n<title>140067972996232-&gt;140067972599536</title>\n<path d=\"M2022,-581.3799C2022,-573.1745 2022,-563.7679 2022,-554.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2025.5001,-554.784 2022,-544.784 2018.5001,-554.784 2025.5001,-554.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140067972686456 -->\n<g class=\"node\" id=\"node20\">\n<title>140067972686456</title>\n<polygon fill=\"none\" points=\"1896,-415.5 1896,-461.5 2148,-461.5 2148,-415.5 1896,-415.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1949.5\" y=\"-434.8\">dense_7: Dense</text>\n<polyline fill=\"none\" points=\"2003,-415.5 2003,-461.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2032\" y=\"-446.3\">input:</text>\n<polyline fill=\"none\" points=\"2003,-438.5 2061,-438.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2032\" y=\"-423.3\">output:</text>\n<polyline fill=\"none\" points=\"2061,-415.5 2061,-461.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2104.5\" y=\"-446.3\">(None, 256)</text>\n<polyline fill=\"none\" points=\"2061,-438.5 2148,-438.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2104.5\" y=\"-423.3\">(None, 64)</text>\n</g>\n<!-- 140067972599536&#45;&gt;140067972686456 -->\n<g class=\"edge\" id=\"edge26\">\n<title>140067972599536-&gt;140067972686456</title>\n<path d=\"M2022,-498.3799C2022,-490.1745 2022,-480.7679 2022,-471.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2025.5001,-471.784 2022,-461.784 2018.5001,-471.784 2025.5001,-471.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140067972598808 -->\n<g class=\"node\" id=\"node21\">\n<title>140067972598808</title>\n<polygon fill=\"none\" points=\"1886,-332.5 1886,-378.5 2158,-378.5 2158,-332.5 1886,-332.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1953\" y=\"-351.8\">dropout_5: Dropout</text>\n<polyline fill=\"none\" points=\"2020,-332.5 2020,-378.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2049\" y=\"-363.3\">input:</text>\n<polyline fill=\"none\" points=\"2020,-355.5 2078,-355.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2049\" y=\"-340.3\">output:</text>\n<polyline fill=\"none\" points=\"2078,-332.5 2078,-378.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2118\" y=\"-363.3\">(None, 64)</text>\n<polyline fill=\"none\" points=\"2078,-355.5 2158,-355.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2118\" y=\"-340.3\">(None, 64)</text>\n</g>\n<!-- 140067972686456&#45;&gt;140067972598808 -->\n<g class=\"edge\" id=\"edge27\">\n<title>140067972686456-&gt;140067972598808</title>\n<path d=\"M2022,-415.3799C2022,-407.1745 2022,-397.7679 2022,-388.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2025.5001,-388.784 2022,-378.784 2018.5001,-388.784 2025.5001,-388.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140067972835144 -->\n<g class=\"node\" id=\"node22\">\n<title>140067972835144</title>\n<polygon fill=\"none\" points=\"1817.5,-249.5 1817.5,-295.5 2226.5,-295.5 2226.5,-249.5 1817.5,-249.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1953\" y=\"-268.8\">batch_normalization_6: BatchNormalization</text>\n<polyline fill=\"none\" points=\"2088.5,-249.5 2088.5,-295.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2117.5\" y=\"-280.3\">input:</text>\n<polyline fill=\"none\" points=\"2088.5,-272.5 2146.5,-272.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2117.5\" y=\"-257.3\">output:</text>\n<polyline fill=\"none\" points=\"2146.5,-249.5 2146.5,-295.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2186.5\" y=\"-280.3\">(None, 64)</text>\n<polyline fill=\"none\" points=\"2146.5,-272.5 2226.5,-272.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2186.5\" y=\"-257.3\">(None, 64)</text>\n</g>\n<!-- 140067972598808&#45;&gt;140067972835144 -->\n<g class=\"edge\" id=\"edge28\">\n<title>140067972598808-&gt;140067972835144</title>\n<path d=\"M2022,-332.3799C2022,-324.1745 2022,-314.7679 2022,-305.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2025.5001,-305.784 2022,-295.784 2018.5001,-305.784 2025.5001,-305.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140067971606904 -->\n<g class=\"node\" id=\"node23\">\n<title>140067971606904</title>\n<polygon fill=\"none\" points=\"1899.5,-166.5 1899.5,-212.5 2144.5,-212.5 2144.5,-166.5 1899.5,-166.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1953\" y=\"-185.8\">dense_8: Dense</text>\n<polyline fill=\"none\" points=\"2006.5,-166.5 2006.5,-212.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2035.5\" y=\"-197.3\">input:</text>\n<polyline fill=\"none\" points=\"2006.5,-189.5 2064.5,-189.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2035.5\" y=\"-174.3\">output:</text>\n<polyline fill=\"none\" points=\"2064.5,-166.5 2064.5,-212.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2104.5\" y=\"-197.3\">(None, 64)</text>\n<polyline fill=\"none\" points=\"2064.5,-189.5 2144.5,-189.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2104.5\" y=\"-174.3\">(None, 32)</text>\n</g>\n<!-- 140067972835144&#45;&gt;140067971606904 -->\n<g class=\"edge\" id=\"edge29\">\n<title>140067972835144-&gt;140067971606904</title>\n<path d=\"M2022,-249.3799C2022,-241.1745 2022,-231.7679 2022,-222.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2025.5001,-222.784 2022,-212.784 2018.5001,-222.784 2025.5001,-222.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140067971984296 -->\n<g class=\"node\" id=\"node24\">\n<title>140067971984296</title>\n<polygon fill=\"none\" points=\"1886,-83.5 1886,-129.5 2158,-129.5 2158,-83.5 1886,-83.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1953\" y=\"-102.8\">dropout_6: Dropout</text>\n<polyline fill=\"none\" points=\"2020,-83.5 2020,-129.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2049\" y=\"-114.3\">input:</text>\n<polyline fill=\"none\" points=\"2020,-106.5 2078,-106.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2049\" y=\"-91.3\">output:</text>\n<polyline fill=\"none\" points=\"2078,-83.5 2078,-129.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2118\" y=\"-114.3\">(None, 32)</text>\n<polyline fill=\"none\" points=\"2078,-106.5 2158,-106.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2118\" y=\"-91.3\">(None, 32)</text>\n</g>\n<!-- 140067971606904&#45;&gt;140067971984296 -->\n<g class=\"edge\" id=\"edge30\">\n<title>140067971606904-&gt;140067971984296</title>\n<path d=\"M2022,-166.3799C2022,-158.1745 2022,-148.7679 2022,-139.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2025.5001,-139.784 2022,-129.784 2018.5001,-139.784 2025.5001,-139.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140067971384600 -->\n<g class=\"node\" id=\"node25\">\n<title>140067971384600</title>\n<polygon fill=\"none\" points=\"1899.5,-.5 1899.5,-46.5 2144.5,-46.5 2144.5,-.5 1899.5,-.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1953\" y=\"-19.8\">dense_9: Dense</text>\n<polyline fill=\"none\" points=\"2006.5,-.5 2006.5,-46.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2035.5\" y=\"-31.3\">input:</text>\n<polyline fill=\"none\" points=\"2006.5,-23.5 2064.5,-23.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2035.5\" y=\"-8.3\">output:</text>\n<polyline fill=\"none\" points=\"2064.5,-.5 2064.5,-46.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2104.5\" y=\"-31.3\">(None, 32)</text>\n<polyline fill=\"none\" points=\"2064.5,-23.5 2144.5,-23.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2104.5\" y=\"-8.3\">(None, 5)</text>\n</g>\n<!-- 140067971984296&#45;&gt;140067971384600 -->\n<g class=\"edge\" id=\"edge31\">\n<title>140067971984296-&gt;140067971384600</title>\n<path d=\"M2022,-83.3799C2022,-75.1745 2022,-65.7679 2022,-56.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2025.5001,-56.784 2022,-46.784 2018.5001,-56.784 2025.5001,-56.784\" stroke=\"#000000\"/>\n</g>\n</g>\n</svg>"
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 30
        }
      ]
    },
    {
      "metadata": {
        "id": "q7cn0Rl5R6KJ",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "K folds Validation"
      ]
    },
    {
      "metadata": {
        "id": "O9OLbhyJTfMc",
        "colab_type": "code",
        "outputId": "58b88c23-3fbc-4483-ab99-0e98f8121c78",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 967
        }
      },
      "cell_type": "code",
      "source": [
        "NUM_FOLDS = 5\n",
        "train[\"fold_id\"] = train[\"SentenceId\"].apply(lambda x: x%NUM_FOLDS)\n",
        "test_preds = np.zeros((test.shape[0], 5))\n",
        "file_path = \"/content/drive/My Drive/DeepLearning/best_model.hdf5\"\n",
        "check_point = ModelCheckpoint(file_path, monitor = \"val_loss\", verbose = 1,\n",
        "                              save_best_only = True, mode = \"min\")\n",
        "early_stop = EarlyStopping(monitor = \"val_loss\", mode = \"min\", patience = 2)\n",
        "\n",
        "print(\"Building the model...\")\n",
        "model1 = build_model1(lr = 1e-3, lr_d = 0, units = 64, spatial_dr = 0.5, kernel_size1=3, kernel_size2=3, dense_units=64, dr=0.3, conv_size=32)\n",
        "\n",
        "for i in range(NUM_FOLDS):\n",
        "    print(\"FOLD\", i+1)    \n",
        "    print(\"Splitting the data into train and validation...\")\n",
        "    train_seq, val_seq = X_train[train[\"fold_id\"] != i], X_train[train[\"fold_id\"] == i]\n",
        "    y_train = ohe.transform(train[train[\"fold_id\"] != i][\"Sentiment\"].values.reshape(-1, 1))\n",
        "    y_val = ohe.transform(train[train[\"fold_id\"] == i][\"Sentiment\"].values.reshape(-1, 1))        \n",
        "    \n",
        "    print(\"Training the model...\")\n",
        "    model1.fit(train_seq, y_train, validation_data = (val_seq, y_val), batch_size = 128, epochs = 15, verbose = 1, callbacks = [early_stop]) \n",
        "    model1.save_weights(file_path)  \n",
        "    test_preds += model1.predict([X_test], batch_size=1024, verbose=1)    \n",
        "    print()\n",
        "    \n",
        "print(\"Save model after cross-validation...\")\n",
        "model1.save_weights(file_path)   \n",
        "test_preds /= NUM_FOLDS\n",
        "\n",
        "\n",
        "print(\"Make the submission ready...\")\n",
        "sub = pd.read_csv('/content/drive/My Drive/DeepLearning/sampleSubmission.csv', sep=\",\")\n",
        "\n",
        "pred = model1.predict(X_test, batch_size = 1024, verbose = 1)\n",
        "predictions = np.round(np.argmax(pred, axis=1)).astype(int)\n",
        "submission = pd.DataFrame({'PhraseId':test['PhraseId'],'Sentiment': predictions})\n",
        "submission =pd.merge(submission, save_test, on='PhraseId', how='left')\n",
        "submission[\"Sentiment\"] = submission.apply(get_sentiment, axis=1)\n",
        "submission.drop(['Sentiment_x', 'Sentiment_y'], axis=1,inplace=True)\n",
        "submission[\"Sentiment\"] = submission[\"Sentiment\"].astype(int)\n",
        "submission.to_csv(\"/content/drive/My Drive/DeepLearning/blend.csv\", index=False)\n",
        "\n",
        "predictions = np.round(np.argmax(test_preds, axis=1)).astype(int)\n",
        "submission = pd.DataFrame({'PhraseId':test['PhraseId'],'Sentiment': predictions})\n",
        "submission =pd.merge(submission, save_test, on='PhraseId', how='left')\n",
        "submission[\"Sentiment\"] = submission.apply(get_sentiment, axis=1)\n",
        "submission.drop(['Sentiment_x', 'Sentiment_y'], axis=1,inplace=True)\n",
        "submission[\"Sentiment\"] = submission[\"Sentiment\"].astype(int)\n",
        "submission.to_csv(\"/content/drive/My Drive/DeepLearning/avg_blend.csv\", index=False)"
      ],
      "execution_count": 31,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Building the model...\n",
            "FOLD 1\n",
            "Splitting the data into train and validation...\n",
            "Training the model...\n",
            "Train on 125230 samples, validate on 30830 samples\n",
            "Epoch 1/15\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "error",
          "ename": "KeyboardInterrupt",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-31-3d7852ee45af>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     18\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     19\u001b[0m     \u001b[0mprint\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\"Training the model...\"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 20\u001b[0;31m     \u001b[0mmodel1\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mfit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mtrain_seq\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_train\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mvalidation_data\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m(\u001b[0m\u001b[0mval_seq\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0my_val\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbatch_size\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m128\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mepochs\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m15\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mverbose\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m1\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcallbacks\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;34m[\u001b[0m\u001b[0mearly_stop\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     21\u001b[0m     \u001b[0mmodel1\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msave_weights\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mfile_path\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     22\u001b[0m     \u001b[0mtest_preds\u001b[0m \u001b[0;34m+=\u001b[0m \u001b[0mmodel1\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mpredict\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mX_test\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mbatch_size\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m1024\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mverbose\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;36m1\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/keras/engine/training.py\u001b[0m in \u001b[0;36mfit\u001b[0;34m(self, x, y, batch_size, epochs, verbose, callbacks, validation_split, validation_data, shuffle, class_weight, sample_weight, initial_epoch, steps_per_epoch, validation_steps, **kwargs)\u001b[0m\n\u001b[1;32m   1037\u001b[0m                                         \u001b[0minitial_epoch\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0minitial_epoch\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1038\u001b[0m                                         \u001b[0msteps_per_epoch\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0msteps_per_epoch\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1039\u001b[0;31m                                         validation_steps=validation_steps)\n\u001b[0m\u001b[1;32m   1040\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1041\u001b[0m     def evaluate(self, x=None, y=None,\n",
            "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/keras/engine/training_arrays.py\u001b[0m in \u001b[0;36mfit_loop\u001b[0;34m(model, f, ins, out_labels, batch_size, epochs, verbose, callbacks, val_f, val_ins, shuffle, callback_metrics, initial_epoch, steps_per_epoch, validation_steps)\u001b[0m\n\u001b[1;32m    197\u001b[0m                     \u001b[0mins_batch\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mins_batch\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0mi\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mtoarray\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    198\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m--> 199\u001b[0;31m                 \u001b[0mouts\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mf\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mins_batch\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    200\u001b[0m                 \u001b[0mouts\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mto_list\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mouts\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    201\u001b[0m                 \u001b[0;32mfor\u001b[0m \u001b[0ml\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mo\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mzip\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mout_labels\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mouts\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, inputs)\u001b[0m\n\u001b[1;32m   2713\u001b[0m                 \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_legacy_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minputs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2714\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2715\u001b[0;31m             \u001b[0;32mreturn\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_call\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0minputs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   2716\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2717\u001b[0m             \u001b[0;32mif\u001b[0m \u001b[0mpy_any\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mis_tensor\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mx\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mx\u001b[0m \u001b[0;32min\u001b[0m \u001b[0minputs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py\u001b[0m in \u001b[0;36m_call\u001b[0;34m(self, inputs)\u001b[0m\n\u001b[1;32m   2673\u001b[0m             \u001b[0mfetched\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_callable_fn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0marray_vals\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mrun_metadata\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mrun_metadata\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2674\u001b[0m         \u001b[0;32melse\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 2675\u001b[0;31m             \u001b[0mfetched\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_callable_fn\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0marray_vals\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m   2676\u001b[0m         \u001b[0;32mreturn\u001b[0m \u001b[0mfetched\u001b[0m\u001b[0;34m[\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0mlen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0moutputs\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m]\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   2677\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/tensorflow/python/client/session.py\u001b[0m in \u001b[0;36m__call__\u001b[0;34m(self, *args, **kwargs)\u001b[0m\n\u001b[1;32m   1437\u001b[0m           ret = tf_session.TF_SessionRunCallable(\n\u001b[1;32m   1438\u001b[0m               \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_session\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_session\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mself\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0m_handle\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0margs\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mstatus\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m-> 1439\u001b[0;31m               run_metadata_ptr)\n\u001b[0m\u001b[1;32m   1440\u001b[0m         \u001b[0;32mif\u001b[0m \u001b[0mrun_metadata\u001b[0m\u001b[0;34m:\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m   1441\u001b[0m           \u001b[0mproto_data\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtf_session\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mTF_GetBuffer\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mrun_metadata_ptr\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
          ]
        }
      ]
    },
    {
      "metadata": {
        "id": "w_Qgo7qZjPbh",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "###Model 2: Embedding + LSTM + multi CNN\n"
      ]
    },
    {
      "metadata": {
        "id": "Xby86sUKw4yg",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "def build_model2(lr=0.0, lr_d=0.0, units=0, spatial_dr=0.0, kernel_size1=3, kernel_size2=2, dense_units=128, dr=0.1, conv_size=32):\n",
        "    input_words = Input((max_len, ))\n",
        "    x_words = Embedding(19479, embed_size,weights=[embedding_matrix],trainable=False)(input_words)\n",
        "    x_words = SpatialDropout1D(0.3)(x_words)\n",
        "    x_words = Bidirectional(CuDNNLSTM(50, return_sequences=True))(x_words)\n",
        "    x_words = Dropout(0.2)(x_words)\n",
        "    x_words = Conv1D(256, 3, strides = 1,  padding='causal', activation='relu', )(x_words)\n",
        "    x_words = Conv1D(128, 3, strides = 1,  padding='causal', activation='relu', )(x_words)\n",
        "    x_words = Conv1D(64, 3, strides = 1,   padding='causal', activation='relu', )(x_words)\n",
        "    x_words = GlobalMaxPool1D()(x_words)\n",
        "    x_words = Dropout(0.2)(x_words)\n",
        "\n",
        "    x = Dense(50, activation=\"relu\")(x_words)\n",
        "    x = Dropout(0.2)(x)\n",
        "    predictions = Dense(5, activation=\"softmax\")(x)\n",
        "\n",
        "    model = Model(inputs=[input_words], outputs=predictions)\n",
        "    model.compile(optimizer='nadam' ,loss='categorical_crossentropy', metrics=['accuracy'])\n",
        "\n",
        "    return model"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "nZB9ZOYE47gN",
        "colab_type": "code",
        "outputId": "5120f18e-5dfc-4465-8007-06d6b18089e5",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1275
        }
      },
      "cell_type": "code",
      "source": [
        "test_preds = np.zeros((test.shape[0], 5))\n",
        "file_path = \"/content/drive/My Drive/DeepLearning/best_model.hdf5\"\n",
        "check_point = ModelCheckpoint(file_path, monitor = \"val_loss\", verbose = 1,\n",
        "                              save_best_only = True, mode = \"min\")\n",
        "early_stop = EarlyStopping(monitor = \"val_loss\", mode=\"min\", patience = 3, verbose=1)\n",
        "\n",
        "print(\"Building the model...\")\n",
        "model2 = build_model2(lr = 1e-3, lr_d = 0, units = 64, spatial_dr = 0.5, kernel_size1=3, kernel_size2=3, dense_units=64, dr=0.3, conv_size=32)\n",
        "\n",
        "for i in range(NUM_FOLDS):\n",
        "    print(\"FOLD\", i+1)    \n",
        "    print(\"Splitting the data into train and validation...\")\n",
        "    train_seq, val_seq = X_train[train[\"fold_id\"] != i], X_train[train[\"fold_id\"] == i]\n",
        "    y_train = ohe.transform(train[train[\"fold_id\"] != i][\"Sentiment\"].values.reshape(-1, 1))\n",
        "    y_val = ohe.transform(train[train[\"fold_id\"] == i][\"Sentiment\"].values.reshape(-1, 1))        \n",
        "    \n",
        "    print(\"Training the model...\")\n",
        "    model2.fit(train_seq, y_train, validation_data = (val_seq, y_val), batch_size = 128, epochs = 15, verbose = 1, callbacks = [early_stop]) \n",
        "    model2.save_weights(file_path)  \n",
        "    test_preds += model3.predict([X_test], batch_size=1024, verbose=1)    \n",
        "    print()\n",
        "    \n",
        "print(\"Save model after cross-validation...\")\n",
        "model2.save_weights(file_path)   \n",
        "test_preds /= NUM_FOLDS\n",
        "\n",
        "\n",
        "print(\"Make the submission ready...\")\n",
        "sub = pd.read_csv('/content/drive/My Drive/DeepLearning/sampleSubmission.csv', sep=\",\")\n",
        "\n",
        "pred = model2.predict(X_test, batch_size = 1024, verbose = 1)\n",
        "predictions = np.round(np.argmax(pred, axis=1)).astype(int)\n",
        "sub['Sentiment'] = predictions\n",
        "sub.to_csv(\"/content/drive/My Drive/DeepLearning/blend.csv\", index=False)\n",
        "\n",
        "predictions = np.round(np.argmax(test_preds, axis=1)).astype(int)\n",
        "sub['Sentiment'] = predictions\n",
        "sub.to_csv(\"/content/drive/My Drive/DeepLearning/avg_blend.csv\", index=False)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Building the model...\n",
            "FOLD 1\n",
            "Splitting the data into train and validation...\n",
            "Training the model...\n",
            "Train on 125230 samples, validate on 30830 samples\n",
            "Epoch 1/15\n",
            "125230/125230 [==============================] - 46s 366us/step - loss: 0.9709 - acc: 0.5990 - val_loss: 0.8935 - val_acc: 0.6251\n",
            "Epoch 2/15\n",
            "125230/125230 [==============================] - 40s 322us/step - loss: 0.8800 - acc: 0.6357 - val_loss: 0.8489 - val_acc: 0.6488\n",
            "Epoch 3/15\n",
            "125230/125230 [==============================] - 40s 321us/step - loss: 0.8414 - acc: 0.6511 - val_loss: 0.8371 - val_acc: 0.6549\n",
            "Epoch 4/15\n",
            "125230/125230 [==============================] - 40s 321us/step - loss: 0.8159 - acc: 0.6616 - val_loss: 0.8279 - val_acc: 0.6561\n",
            "Epoch 5/15\n",
            "125230/125230 [==============================] - 40s 321us/step - loss: 0.7955 - acc: 0.6676 - val_loss: 0.8275 - val_acc: 0.6587\n",
            "Epoch 6/15\n",
            "125230/125230 [==============================] - 40s 322us/step - loss: 0.7816 - acc: 0.6741 - val_loss: 0.8235 - val_acc: 0.6600\n",
            "Epoch 7/15\n",
            "125230/125230 [==============================] - 40s 323us/step - loss: 0.7671 - acc: 0.6804 - val_loss: 0.8327 - val_acc: 0.6608\n",
            "Epoch 8/15\n",
            "125230/125230 [==============================] - 40s 322us/step - loss: 0.7576 - acc: 0.6847 - val_loss: 0.8334 - val_acc: 0.6623\n",
            "66292/66292 [==============================] - 4s 54us/step\n",
            "\n",
            "FOLD 2\n",
            "Splitting the data into train and validation...\n",
            "Training the model...\n",
            "Train on 124608 samples, validate on 31452 samples\n",
            "Epoch 1/15\n",
            "124608/124608 [==============================] - 41s 326us/step - loss: 0.7761 - acc: 0.6783 - val_loss: 0.7067 - val_acc: 0.7130\n",
            "Epoch 2/15\n",
            "124608/124608 [==============================] - 40s 322us/step - loss: 0.7584 - acc: 0.6843 - val_loss: 0.7185 - val_acc: 0.7009\n",
            "Epoch 3/15\n",
            "124608/124608 [==============================] - 40s 322us/step - loss: 0.7486 - acc: 0.6889 - val_loss: 0.7210 - val_acc: 0.7015\n",
            "66292/66292 [==============================] - 3s 47us/step\n",
            "\n",
            "FOLD 3\n",
            "Splitting the data into train and validation...\n",
            "Training the model...\n",
            "Train on 125022 samples, validate on 31038 samples\n",
            "Epoch 1/15\n",
            "125022/125022 [==============================] - 41s 325us/step - loss: 0.7590 - acc: 0.6859 - val_loss: 0.6695 - val_acc: 0.7249\n",
            "Epoch 2/15\n",
            "125022/125022 [==============================] - 40s 321us/step - loss: 0.7492 - acc: 0.6886 - val_loss: 0.6859 - val_acc: 0.7142\n",
            "Epoch 3/15\n",
            "125022/125022 [==============================] - 40s 320us/step - loss: 0.7390 - acc: 0.6918 - val_loss: 0.6826 - val_acc: 0.7144\n",
            "66292/66292 [==============================] - 3s 48us/step\n",
            "\n",
            "FOLD 4\n",
            "Splitting the data into train and validation...\n",
            "Training the model...\n",
            "Train on 124609 samples, validate on 31451 samples\n",
            "Epoch 1/15\n",
            "124609/124609 [==============================] - 41s 327us/step - loss: 0.7406 - acc: 0.6922 - val_loss: 0.6621 - val_acc: 0.7266\n",
            "Epoch 2/15\n",
            "124609/124609 [==============================] - 40s 322us/step - loss: 0.7346 - acc: 0.6959 - val_loss: 0.6731 - val_acc: 0.7186\n",
            "Epoch 3/15\n",
            "124609/124609 [==============================] - 40s 321us/step - loss: 0.7292 - acc: 0.6982 - val_loss: 0.6777 - val_acc: 0.7146\n",
            "66292/66292 [==============================] - 3s 47us/step\n",
            "\n",
            "FOLD 5\n",
            "Splitting the data into train and validation...\n",
            "Training the model...\n",
            "Train on 124771 samples, validate on 31289 samples\n",
            "Epoch 1/15\n",
            "124771/124771 [==============================] - 41s 326us/step - loss: 0.7309 - acc: 0.6980 - val_loss: 0.6523 - val_acc: 0.7281\n",
            "Epoch 2/15\n",
            "124771/124771 [==============================] - 40s 321us/step - loss: 0.7263 - acc: 0.6981 - val_loss: 0.6621 - val_acc: 0.7269\n",
            "Epoch 3/15\n",
            "124771/124771 [==============================] - 40s 321us/step - loss: 0.7205 - acc: 0.6998 - val_loss: 0.6625 - val_acc: 0.7189\n",
            "66292/66292 [==============================] - 3s 48us/step\n",
            "\n",
            "Save model after cross-validation...\n",
            "Make the submission ready...\n",
            "66292/66292 [==============================] - 3s 47us/step\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "9I07QVa8OYB4",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "###Model 3: Embedding + {LSTM/GRU} + multi CNN"
      ]
    },
    {
      "metadata": {
        "id": "rpqwTRtOn3Be",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "def build_model3(lr=0.0, lr_d=0.0, units=0, spatial_dr=0.0, kernel_size1=3, kernel_size2=2, dense_units=128, dr=0.1, conv_size=32):\n",
        "    inp = Input(shape = (max_len,))\n",
        "    x = Embedding(19479, embed_size, weights = [embedding_matrix], trainable = False)(inp)\n",
        "    x1 = SpatialDropout1D(spatial_dr)(x)\n",
        "\n",
        "    x_gru = Bidirectional(CuDNNGRU(units, return_sequences = True))(x1)\n",
        "    x_lstm = Bidirectional(CuDNNLSTM(units, return_sequences = True))(x1)\n",
        "    \n",
        "    x_conv1 = Conv1D(128, kernel_size=kernel_size1, padding='causal', kernel_initializer='he_uniform')(x_gru)\n",
        "    x_conv1 = Conv1D(64, kernel_size=kernel_size1, padding='causal', kernel_initializer='he_uniform')(x_conv1)\n",
        "    x_conv1 = Conv1D(conv_size, kernel_size=kernel_size1, padding='causal', kernel_initializer='he_uniform')(x_conv1)\n",
        "\n",
        "      \n",
        "    avg_pool1_gru = GlobalAveragePooling1D()(x_conv1)\n",
        "    max_pool1_gru = GlobalMaxPooling1D()(x_conv1)\n",
        "    \n",
        "    x_conv2 = Conv1D(128, kernel_size=kernel_size2, padding='causal', kernel_initializer='he_uniform')(x_gru)\n",
        "    x_conv2 = Conv1D(64, kernel_size=kernel_size2, padding='causal', kernel_initializer='he_uniform')(x_conv2)\n",
        "    x_conv2 = Conv1D(conv_size, kernel_size=kernel_size2, padding='causal', kernel_initializer='he_uniform')(x_conv2)    \n",
        "    avg_pool2_gru = GlobalAveragePooling1D()(x_conv2)\n",
        "    max_pool2_gru = GlobalMaxPooling1D()(x_conv2)\n",
        "    \n",
        "    \n",
        "    x_conv3 = Conv1D(128, kernel_size=kernel_size1, padding='causal', kernel_initializer='he_uniform')(x_lstm)\n",
        "    x_conv3 = Conv1D(64, kernel_size=kernel_size1, padding='causal', kernel_initializer='he_uniform')(x_conv3)\n",
        "    x_conv3 = Conv1D(conv_size, kernel_size=kernel_size1, padding='causal', kernel_initializer='he_uniform')(x_conv3)\n",
        "    avg_pool1_lstm = GlobalAveragePooling1D()(x_conv3)\n",
        "    max_pool1_lstm = GlobalMaxPooling1D()(x_conv3)\n",
        "    \n",
        "    x_conv4 = Conv1D(128, kernel_size=kernel_size2, padding='causal', kernel_initializer='he_uniform')(x_lstm)\n",
        "    x_conv4 = Conv1D(64, kernel_size=kernel_size2, padding='causal', kernel_initializer='he_uniform')(x_conv4)\n",
        "    x_conv4 = Conv1D(conv_size, kernel_size=kernel_size2, padding='causal', kernel_initializer='he_uniform')(x_conv4)\n",
        "\n",
        "    avg_pool2_lstm = GlobalAveragePooling1D()(x_conv4)\n",
        "    max_pool2_lstm = GlobalMaxPooling1D()(x_conv4)\n",
        "    \n",
        "    \n",
        "    x = concatenate([avg_pool1_gru, max_pool1_gru, avg_pool2_gru, max_pool2_gru,\n",
        "                    avg_pool1_lstm, max_pool1_lstm, avg_pool2_lstm, max_pool2_lstm])\n",
        "    x = BatchNormalization()(x)\n",
        "    x = Dropout(dr)(Dense(dense_units, activation='relu') (x))\n",
        "    x = BatchNormalization()(x)\n",
        "    x = Dropout(dr)(Dense(int(dense_units / 2), activation='relu') (x))\n",
        "    x = Dense(5, activation = \"sigmoid\")(x)\n",
        "    model = Model(inputs = inp, outputs = x)\n",
        "    model.compile(loss = \"binary_crossentropy\", optimizer = Adam(lr = lr, decay = lr_d), metrics = [\"accuracy\"])\n",
        "    return model"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "FZamFj3U7z5o",
        "colab_type": "code",
        "outputId": "2b0ec71d-a9d5-4d66-c75a-768d9ddcdcaa",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 3134
        }
      },
      "cell_type": "code",
      "source": [
        "trained_model3 = build_model3(lr = 1e-3, lr_d = 0, units = 64, spatial_dr = 0.5, kernel_size1=3, kernel_size2=3, dense_units=64, dr=0.3, conv_size=32)\n",
        "trained_model3.summary()\n",
        "SVG(model_to_dot(trained_model3,  show_shapes=True, show_layer_names=True, rankdir='HB').create(prog='dot', format='svg'))"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "__________________________________________________________________________________________________\n",
            "Layer (type)                    Output Shape         Param #     Connected to                     \n",
            "==================================================================================================\n",
            "input_2 (InputLayer)            (None, 50)           0                                            \n",
            "__________________________________________________________________________________________________\n",
            "embedding_2 (Embedding)         (None, 50, 300)      5843700     input_2[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "spatial_dropout1d_2 (SpatialDro (None, 50, 300)      0           embedding_2[0][0]                \n",
            "__________________________________________________________________________________________________\n",
            "bidirectional_3 (Bidirectional) (None, 50, 128)      140544      spatial_dropout1d_2[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "bidirectional_4 (Bidirectional) (None, 50, 128)      187392      spatial_dropout1d_2[0][0]        \n",
            "__________________________________________________________________________________________________\n",
            "conv1d_13 (Conv1D)              (None, 50, 128)      49280       bidirectional_3[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv1d_16 (Conv1D)              (None, 50, 128)      49280       bidirectional_3[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv1d_19 (Conv1D)              (None, 50, 128)      49280       bidirectional_4[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv1d_22 (Conv1D)              (None, 50, 128)      49280       bidirectional_4[0][0]            \n",
            "__________________________________________________________________________________________________\n",
            "conv1d_14 (Conv1D)              (None, 50, 64)       24640       conv1d_13[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "conv1d_17 (Conv1D)              (None, 50, 64)       24640       conv1d_16[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "conv1d_20 (Conv1D)              (None, 50, 64)       24640       conv1d_19[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "conv1d_23 (Conv1D)              (None, 50, 64)       24640       conv1d_22[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "conv1d_15 (Conv1D)              (None, 50, 32)       6176        conv1d_14[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "conv1d_18 (Conv1D)              (None, 50, 32)       6176        conv1d_17[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "conv1d_21 (Conv1D)              (None, 50, 32)       6176        conv1d_20[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "conv1d_24 (Conv1D)              (None, 50, 32)       6176        conv1d_23[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "global_average_pooling1d_5 (Glo (None, 32)           0           conv1d_15[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "global_max_pooling1d_5 (GlobalM (None, 32)           0           conv1d_15[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "global_average_pooling1d_6 (Glo (None, 32)           0           conv1d_18[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "global_max_pooling1d_6 (GlobalM (None, 32)           0           conv1d_18[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "global_average_pooling1d_7 (Glo (None, 32)           0           conv1d_21[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "global_max_pooling1d_7 (GlobalM (None, 32)           0           conv1d_21[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "global_average_pooling1d_8 (Glo (None, 32)           0           conv1d_24[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "global_max_pooling1d_8 (GlobalM (None, 32)           0           conv1d_24[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "concatenate_2 (Concatenate)     (None, 256)          0           global_average_pooling1d_5[0][0] \n",
            "                                                                 global_max_pooling1d_5[0][0]     \n",
            "                                                                 global_average_pooling1d_6[0][0] \n",
            "                                                                 global_max_pooling1d_6[0][0]     \n",
            "                                                                 global_average_pooling1d_7[0][0] \n",
            "                                                                 global_max_pooling1d_7[0][0]     \n",
            "                                                                 global_average_pooling1d_8[0][0] \n",
            "                                                                 global_max_pooling1d_8[0][0]     \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_3 (BatchNor (None, 256)          1024        concatenate_2[0][0]              \n",
            "__________________________________________________________________________________________________\n",
            "dense_4 (Dense)                 (None, 64)           16448       batch_normalization_3[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "dropout_3 (Dropout)             (None, 64)           0           dense_4[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "batch_normalization_4 (BatchNor (None, 64)           256         dropout_3[0][0]                  \n",
            "__________________________________________________________________________________________________\n",
            "dense_5 (Dense)                 (None, 32)           2080        batch_normalization_4[0][0]      \n",
            "__________________________________________________________________________________________________\n",
            "dropout_4 (Dropout)             (None, 32)           0           dense_5[0][0]                    \n",
            "__________________________________________________________________________________________________\n",
            "dense_6 (Dense)                 (None, 5)            165         dropout_4[0][0]                  \n",
            "==================================================================================================\n",
            "Total params: 6,511,993\n",
            "Trainable params: 667,653\n",
            "Non-trainable params: 5,844,340\n",
            "__________________________________________________________________________________________________\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "<IPython.core.display.SVG object>"
            ],
            "image/svg+xml": "<svg height=\"1300pt\" viewBox=\"0.00 0.00 4006.00 1300.00\" width=\"4006pt\" xmlns=\"http://www.w3.org/2000/svg\" xmlns:xlink=\"http://www.w3.org/1999/xlink\">\n<g class=\"graph\" id=\"graph0\" transform=\"scale(1 1) rotate(0) translate(4 1296)\">\n<title>G</title>\n<polygon fill=\"#ffffff\" points=\"-4,4 -4,-1296 4002,-1296 4002,4 -4,4\" stroke=\"transparent\"/>\n<!-- 140252887122888 -->\n<g class=\"node\" id=\"node1\">\n<title>140252887122888</title>\n<polygon fill=\"none\" points=\"2039.5,-1245.5 2039.5,-1291.5 2310.5,-1291.5 2310.5,-1245.5 2039.5,-1245.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2106\" y=\"-1264.8\">input_2: InputLayer</text>\n<polyline fill=\"none\" points=\"2172.5,-1245.5 2172.5,-1291.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2201.5\" y=\"-1276.3\">input:</text>\n<polyline fill=\"none\" points=\"2172.5,-1268.5 2230.5,-1268.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2201.5\" y=\"-1253.3\">output:</text>\n<polyline fill=\"none\" points=\"2230.5,-1245.5 2230.5,-1291.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2270.5\" y=\"-1276.3\">(None, 50)</text>\n<polyline fill=\"none\" points=\"2230.5,-1268.5 2310.5,-1268.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2270.5\" y=\"-1253.3\">(None, 50)</text>\n</g>\n<!-- 140252887303392 -->\n<g class=\"node\" id=\"node2\">\n<title>140252887303392</title>\n<polygon fill=\"none\" points=\"2005.5,-1162.5 2005.5,-1208.5 2344.5,-1208.5 2344.5,-1162.5 2005.5,-1162.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2091\" y=\"-1181.8\">embedding_2: Embedding</text>\n<polyline fill=\"none\" points=\"2176.5,-1162.5 2176.5,-1208.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2205.5\" y=\"-1193.3\">input:</text>\n<polyline fill=\"none\" points=\"2176.5,-1185.5 2234.5,-1185.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2205.5\" y=\"-1170.3\">output:</text>\n<polyline fill=\"none\" points=\"2234.5,-1162.5 2234.5,-1208.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2289.5\" y=\"-1193.3\">(None, 50)</text>\n<polyline fill=\"none\" points=\"2234.5,-1185.5 2344.5,-1185.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2289.5\" y=\"-1170.3\">(None, 50, 300)</text>\n</g>\n<!-- 140252887122888&#45;&gt;140252887303392 -->\n<g class=\"edge\" id=\"edge1\">\n<title>140252887122888-&gt;140252887303392</title>\n<path d=\"M2175,-1245.3799C2175,-1237.1745 2175,-1227.7679 2175,-1218.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2178.5001,-1218.784 2175,-1208.784 2171.5001,-1218.784 2178.5001,-1218.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140252887303728 -->\n<g class=\"node\" id=\"node3\">\n<title>140252887303728</title>\n<polygon fill=\"none\" points=\"1966.5,-1079.5 1966.5,-1125.5 2383.5,-1125.5 2383.5,-1079.5 1966.5,-1079.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2091\" y=\"-1098.8\">spatial_dropout1d_2: SpatialDropout1D</text>\n<polyline fill=\"none\" points=\"2215.5,-1079.5 2215.5,-1125.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2244.5\" y=\"-1110.3\">input:</text>\n<polyline fill=\"none\" points=\"2215.5,-1102.5 2273.5,-1102.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2244.5\" y=\"-1087.3\">output:</text>\n<polyline fill=\"none\" points=\"2273.5,-1079.5 2273.5,-1125.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2328.5\" y=\"-1110.3\">(None, 50, 300)</text>\n<polyline fill=\"none\" points=\"2273.5,-1102.5 2383.5,-1102.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2328.5\" y=\"-1087.3\">(None, 50, 300)</text>\n</g>\n<!-- 140252887303392&#45;&gt;140252887303728 -->\n<g class=\"edge\" id=\"edge2\">\n<title>140252887303392-&gt;140252887303728</title>\n<path d=\"M2175,-1162.3799C2175,-1154.1745 2175,-1144.7679 2175,-1135.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2178.5001,-1135.784 2175,-1125.784 2171.5001,-1135.784 2178.5001,-1135.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140252886456976 -->\n<g class=\"node\" id=\"node4\">\n<title>140252886456976</title>\n<polygon fill=\"none\" points=\"1409.5,-996.5 1409.5,-1042.5 1936.5,-1042.5 1936.5,-996.5 1409.5,-996.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1589\" y=\"-1015.8\">bidirectional_3(cu_dnngru_2): Bidirectional(CuDNNGRU)</text>\n<polyline fill=\"none\" points=\"1768.5,-996.5 1768.5,-1042.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1797.5\" y=\"-1027.3\">input:</text>\n<polyline fill=\"none\" points=\"1768.5,-1019.5 1826.5,-1019.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1797.5\" y=\"-1004.3\">output:</text>\n<polyline fill=\"none\" points=\"1826.5,-996.5 1826.5,-1042.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1881.5\" y=\"-1027.3\">(None, 50, 300)</text>\n<polyline fill=\"none\" points=\"1826.5,-1019.5 1936.5,-1019.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1881.5\" y=\"-1004.3\">(None, 50, 128)</text>\n</g>\n<!-- 140252887303728&#45;&gt;140252886456976 -->\n<g class=\"edge\" id=\"edge3\">\n<title>140252887303728-&gt;140252886456976</title>\n<path d=\"M2035.8317,-1079.4901C1969.4565,-1068.5157 1889.9664,-1055.3729 1822.3293,-1044.1899\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"1822.8761,-1040.7329 1812.4391,-1042.5547 1821.7342,-1047.6391 1822.8761,-1040.7329\" stroke=\"#000000\"/>\n</g>\n<!-- 140252876423856 -->\n<g class=\"node\" id=\"node5\">\n<title>140252876423856</title>\n<polygon fill=\"none\" points=\"2266,-996.5 2266,-1042.5 2806,-1042.5 2806,-996.5 2266,-996.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2452\" y=\"-1015.8\">bidirectional_4(cu_dnnlstm_2): Bidirectional(CuDNNLSTM)</text>\n<polyline fill=\"none\" points=\"2638,-996.5 2638,-1042.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2667\" y=\"-1027.3\">input:</text>\n<polyline fill=\"none\" points=\"2638,-1019.5 2696,-1019.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2667\" y=\"-1004.3\">output:</text>\n<polyline fill=\"none\" points=\"2696,-996.5 2696,-1042.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2751\" y=\"-1027.3\">(None, 50, 300)</text>\n<polyline fill=\"none\" points=\"2696,-1019.5 2806,-1019.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2751\" y=\"-1004.3\">(None, 50, 128)</text>\n</g>\n<!-- 140252887303728&#45;&gt;140252876423856 -->\n<g class=\"edge\" id=\"edge4\">\n<title>140252887303728-&gt;140252876423856</title>\n<path d=\"M2275.0792,-1079.4901C2321.9309,-1068.7181 2377.8691,-1055.857 2425.9141,-1044.8106\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2426.7646,-1048.2065 2435.7261,-1042.5547 2425.1961,-1041.3844 2426.7646,-1048.2065\" stroke=\"#000000\"/>\n</g>\n<!-- 140252876426264 -->\n<g class=\"node\" id=\"node6\">\n<title>140252876426264</title>\n<polygon fill=\"none\" points=\"1024,-913.5 1024,-959.5 1332,-959.5 1332,-913.5 1024,-913.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1094\" y=\"-932.8\">conv1d_13: Conv1D</text>\n<polyline fill=\"none\" points=\"1164,-913.5 1164,-959.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1193\" y=\"-944.3\">input:</text>\n<polyline fill=\"none\" points=\"1164,-936.5 1222,-936.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1193\" y=\"-921.3\">output:</text>\n<polyline fill=\"none\" points=\"1222,-913.5 1222,-959.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1277\" y=\"-944.3\">(None, 50, 128)</text>\n<polyline fill=\"none\" points=\"1222,-936.5 1332,-936.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1277\" y=\"-921.3\">(None, 50, 128)</text>\n</g>\n<!-- 140252886456976&#45;&gt;140252876426264 -->\n<g class=\"edge\" id=\"edge5\">\n<title>140252886456976-&gt;140252876426264</title>\n<path d=\"M1535.7723,-996.4901C1470.4568,-985.5382 1392.2621,-972.4268 1325.6572,-961.2587\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"1325.9358,-957.7566 1315.4947,-959.5547 1324.7782,-964.6602 1325.9358,-957.7566\" stroke=\"#000000\"/>\n</g>\n<!-- 140252874037792 -->\n<g class=\"node\" id=\"node7\">\n<title>140252874037792</title>\n<polygon fill=\"none\" points=\"1519,-913.5 1519,-959.5 1827,-959.5 1827,-913.5 1519,-913.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1589\" y=\"-932.8\">conv1d_16: Conv1D</text>\n<polyline fill=\"none\" points=\"1659,-913.5 1659,-959.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1688\" y=\"-944.3\">input:</text>\n<polyline fill=\"none\" points=\"1659,-936.5 1717,-936.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1688\" y=\"-921.3\">output:</text>\n<polyline fill=\"none\" points=\"1717,-913.5 1717,-959.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1772\" y=\"-944.3\">(None, 50, 128)</text>\n<polyline fill=\"none\" points=\"1717,-936.5 1827,-936.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1772\" y=\"-921.3\">(None, 50, 128)</text>\n</g>\n<!-- 140252886456976&#45;&gt;140252874037792 -->\n<g class=\"edge\" id=\"edge6\">\n<title>140252886456976-&gt;140252874037792</title>\n<path d=\"M1673,-996.3799C1673,-988.1745 1673,-978.7679 1673,-969.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"1676.5001,-969.784 1673,-959.784 1669.5001,-969.784 1676.5001,-969.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140252873065528 -->\n<g class=\"node\" id=\"node8\">\n<title>140252873065528</title>\n<polygon fill=\"none\" points=\"2382,-913.5 2382,-959.5 2690,-959.5 2690,-913.5 2382,-913.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2452\" y=\"-932.8\">conv1d_19: Conv1D</text>\n<polyline fill=\"none\" points=\"2522,-913.5 2522,-959.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2551\" y=\"-944.3\">input:</text>\n<polyline fill=\"none\" points=\"2522,-936.5 2580,-936.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2551\" y=\"-921.3\">output:</text>\n<polyline fill=\"none\" points=\"2580,-913.5 2580,-959.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2635\" y=\"-944.3\">(None, 50, 128)</text>\n<polyline fill=\"none\" points=\"2580,-936.5 2690,-936.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2635\" y=\"-921.3\">(None, 50, 128)</text>\n</g>\n<!-- 140252876423856&#45;&gt;140252873065528 -->\n<g class=\"edge\" id=\"edge7\">\n<title>140252876423856-&gt;140252873065528</title>\n<path d=\"M2536,-996.3799C2536,-988.1745 2536,-978.7679 2536,-969.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2539.5001,-969.784 2536,-959.784 2532.5001,-969.784 2539.5001,-969.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140252872716360 -->\n<g class=\"node\" id=\"node9\">\n<title>140252872716360</title>\n<polygon fill=\"none\" points=\"2909,-913.5 2909,-959.5 3217,-959.5 3217,-913.5 2909,-913.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2979\" y=\"-932.8\">conv1d_22: Conv1D</text>\n<polyline fill=\"none\" points=\"3049,-913.5 3049,-959.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3078\" y=\"-944.3\">input:</text>\n<polyline fill=\"none\" points=\"3049,-936.5 3107,-936.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3078\" y=\"-921.3\">output:</text>\n<polyline fill=\"none\" points=\"3107,-913.5 3107,-959.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3162\" y=\"-944.3\">(None, 50, 128)</text>\n<polyline fill=\"none\" points=\"3107,-936.5 3217,-936.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3162\" y=\"-921.3\">(None, 50, 128)</text>\n</g>\n<!-- 140252876423856&#45;&gt;140252872716360 -->\n<g class=\"edge\" id=\"edge8\">\n<title>140252876423856-&gt;140252872716360</title>\n<path d=\"M2682.099,-996.4901C2751.9225,-985.4932 2835.5706,-972.319 2906.6703,-961.1212\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2907.2831,-964.5679 2916.6168,-959.5547 2906.194,-957.6531 2907.2831,-964.5679\" stroke=\"#000000\"/>\n</g>\n<!-- 140252876426488 -->\n<g class=\"node\" id=\"node10\">\n<title>140252876426488</title>\n<polygon fill=\"none\" points=\"685,-830.5 685,-876.5 993,-876.5 993,-830.5 685,-830.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"755\" y=\"-849.8\">conv1d_14: Conv1D</text>\n<polyline fill=\"none\" points=\"825,-830.5 825,-876.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"854\" y=\"-861.3\">input:</text>\n<polyline fill=\"none\" points=\"825,-853.5 883,-853.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"854\" y=\"-838.3\">output:</text>\n<polyline fill=\"none\" points=\"883,-830.5 883,-876.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"938\" y=\"-861.3\">(None, 50, 128)</text>\n<polyline fill=\"none\" points=\"883,-853.5 993,-853.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"938\" y=\"-838.3\">(None, 50, 64)</text>\n</g>\n<!-- 140252876426264&#45;&gt;140252876426488 -->\n<g class=\"edge\" id=\"edge9\">\n<title>140252876426264-&gt;140252876426488</title>\n<path d=\"M1084.0198,-913.4901C1040.2071,-902.7631 987.9326,-889.9643 942.9426,-878.9491\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"943.7085,-875.5333 933.163,-876.5547 942.0438,-882.3325 943.7085,-875.5333\" stroke=\"#000000\"/>\n</g>\n<!-- 140252873658440 -->\n<g class=\"node\" id=\"node11\">\n<title>140252873658440</title>\n<polygon fill=\"none\" points=\"1519,-830.5 1519,-876.5 1827,-876.5 1827,-830.5 1519,-830.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1589\" y=\"-849.8\">conv1d_17: Conv1D</text>\n<polyline fill=\"none\" points=\"1659,-830.5 1659,-876.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1688\" y=\"-861.3\">input:</text>\n<polyline fill=\"none\" points=\"1659,-853.5 1717,-853.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1688\" y=\"-838.3\">output:</text>\n<polyline fill=\"none\" points=\"1717,-830.5 1717,-876.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1772\" y=\"-861.3\">(None, 50, 128)</text>\n<polyline fill=\"none\" points=\"1717,-853.5 1827,-853.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1772\" y=\"-838.3\">(None, 50, 64)</text>\n</g>\n<!-- 140252874037792&#45;&gt;140252873658440 -->\n<g class=\"edge\" id=\"edge10\">\n<title>140252874037792-&gt;140252873658440</title>\n<path d=\"M1673,-913.3799C1673,-905.1745 1673,-895.7679 1673,-886.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"1676.5001,-886.784 1673,-876.784 1669.5001,-886.784 1676.5001,-886.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140252873301800 -->\n<g class=\"node\" id=\"node12\">\n<title>140252873301800</title>\n<polygon fill=\"none\" points=\"2382,-830.5 2382,-876.5 2690,-876.5 2690,-830.5 2382,-830.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2452\" y=\"-849.8\">conv1d_20: Conv1D</text>\n<polyline fill=\"none\" points=\"2522,-830.5 2522,-876.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2551\" y=\"-861.3\">input:</text>\n<polyline fill=\"none\" points=\"2522,-853.5 2580,-853.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2551\" y=\"-838.3\">output:</text>\n<polyline fill=\"none\" points=\"2580,-830.5 2580,-876.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2635\" y=\"-861.3\">(None, 50, 128)</text>\n<polyline fill=\"none\" points=\"2580,-853.5 2690,-853.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2635\" y=\"-838.3\">(None, 50, 64)</text>\n</g>\n<!-- 140252873065528&#45;&gt;140252873301800 -->\n<g class=\"edge\" id=\"edge11\">\n<title>140252873065528-&gt;140252873301800</title>\n<path d=\"M2536,-913.3799C2536,-905.1745 2536,-895.7679 2536,-886.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2539.5001,-886.784 2536,-876.784 2532.5001,-886.784 2539.5001,-886.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140252872318312 -->\n<g class=\"node\" id=\"node13\">\n<title>140252872318312</title>\n<polygon fill=\"none\" points=\"3010,-830.5 3010,-876.5 3318,-876.5 3318,-830.5 3010,-830.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3080\" y=\"-849.8\">conv1d_23: Conv1D</text>\n<polyline fill=\"none\" points=\"3150,-830.5 3150,-876.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3179\" y=\"-861.3\">input:</text>\n<polyline fill=\"none\" points=\"3150,-853.5 3208,-853.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3179\" y=\"-838.3\">output:</text>\n<polyline fill=\"none\" points=\"3208,-830.5 3208,-876.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3263\" y=\"-861.3\">(None, 50, 128)</text>\n<polyline fill=\"none\" points=\"3208,-853.5 3318,-853.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3263\" y=\"-838.3\">(None, 50, 64)</text>\n</g>\n<!-- 140252872716360&#45;&gt;140252872318312 -->\n<g class=\"edge\" id=\"edge12\">\n<title>140252872716360-&gt;140252872318312</title>\n<path d=\"M3091.1341,-913.3799C3102.5298,-904.0151 3115.8297,-893.0855 3127.9245,-883.1462\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"3130.1627,-885.8371 3135.6665,-876.784 3125.7183,-880.429 3130.1627,-885.8371\" stroke=\"#000000\"/>\n</g>\n<!-- 140252874209656 -->\n<g class=\"node\" id=\"node14\">\n<title>140252874209656</title>\n<polygon fill=\"none\" points=\"604,-747.5 604,-793.5 904,-793.5 904,-747.5 604,-747.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"674\" y=\"-766.8\">conv1d_15: Conv1D</text>\n<polyline fill=\"none\" points=\"744,-747.5 744,-793.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"773\" y=\"-778.3\">input:</text>\n<polyline fill=\"none\" points=\"744,-770.5 802,-770.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"773\" y=\"-755.3\">output:</text>\n<polyline fill=\"none\" points=\"802,-747.5 802,-793.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"853\" y=\"-778.3\">(None, 50, 64)</text>\n<polyline fill=\"none\" points=\"802,-770.5 904,-770.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"853\" y=\"-755.3\">(None, 50, 32)</text>\n</g>\n<!-- 140252876426488&#45;&gt;140252874209656 -->\n<g class=\"edge\" id=\"edge13\">\n<title>140252876426488-&gt;140252874209656</title>\n<path d=\"M815.3228,-830.3799C806.0063,-821.2827 795.1777,-810.7088 785.2354,-801.0005\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"787.4451,-798.2663 777.8451,-793.784 782.5546,-803.2746 787.4451,-798.2663\" stroke=\"#000000\"/>\n</g>\n<!-- 140252873759488 -->\n<g class=\"node\" id=\"node15\">\n<title>140252873759488</title>\n<polygon fill=\"none\" points=\"1523,-747.5 1523,-793.5 1823,-793.5 1823,-747.5 1523,-747.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1593\" y=\"-766.8\">conv1d_18: Conv1D</text>\n<polyline fill=\"none\" points=\"1663,-747.5 1663,-793.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1692\" y=\"-778.3\">input:</text>\n<polyline fill=\"none\" points=\"1663,-770.5 1721,-770.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1692\" y=\"-755.3\">output:</text>\n<polyline fill=\"none\" points=\"1721,-747.5 1721,-793.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1772\" y=\"-778.3\">(None, 50, 64)</text>\n<polyline fill=\"none\" points=\"1721,-770.5 1823,-770.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1772\" y=\"-755.3\">(None, 50, 32)</text>\n</g>\n<!-- 140252873658440&#45;&gt;140252873759488 -->\n<g class=\"edge\" id=\"edge14\">\n<title>140252873658440-&gt;140252873759488</title>\n<path d=\"M1673,-830.3799C1673,-822.1745 1673,-812.7679 1673,-803.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"1676.5001,-803.784 1673,-793.784 1669.5001,-803.784 1676.5001,-803.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140252872917400 -->\n<g class=\"node\" id=\"node16\">\n<title>140252872917400</title>\n<polygon fill=\"none\" points=\"2386,-747.5 2386,-793.5 2686,-793.5 2686,-747.5 2386,-747.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2456\" y=\"-766.8\">conv1d_21: Conv1D</text>\n<polyline fill=\"none\" points=\"2526,-747.5 2526,-793.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2555\" y=\"-778.3\">input:</text>\n<polyline fill=\"none\" points=\"2526,-770.5 2584,-770.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2555\" y=\"-755.3\">output:</text>\n<polyline fill=\"none\" points=\"2584,-747.5 2584,-793.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2635\" y=\"-778.3\">(None, 50, 64)</text>\n<polyline fill=\"none\" points=\"2584,-770.5 2686,-770.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2635\" y=\"-755.3\">(None, 50, 32)</text>\n</g>\n<!-- 140252873301800&#45;&gt;140252872917400 -->\n<g class=\"edge\" id=\"edge15\">\n<title>140252873301800-&gt;140252872917400</title>\n<path d=\"M2536,-830.3799C2536,-822.1745 2536,-812.7679 2536,-803.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2539.5001,-803.784 2536,-793.784 2532.5001,-803.784 2539.5001,-803.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140252872427000 -->\n<g class=\"node\" id=\"node17\">\n<title>140252872427000</title>\n<polygon fill=\"none\" points=\"3114,-747.5 3114,-793.5 3414,-793.5 3414,-747.5 3114,-747.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3184\" y=\"-766.8\">conv1d_24: Conv1D</text>\n<polyline fill=\"none\" points=\"3254,-747.5 3254,-793.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3283\" y=\"-778.3\">input:</text>\n<polyline fill=\"none\" points=\"3254,-770.5 3312,-770.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3283\" y=\"-755.3\">output:</text>\n<polyline fill=\"none\" points=\"3312,-747.5 3312,-793.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3363\" y=\"-778.3\">(None, 50, 64)</text>\n<polyline fill=\"none\" points=\"3312,-770.5 3414,-770.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3363\" y=\"-755.3\">(None, 50, 32)</text>\n</g>\n<!-- 140252872318312&#45;&gt;140252872427000 -->\n<g class=\"edge\" id=\"edge16\">\n<title>140252872318312-&gt;140252872427000</title>\n<path d=\"M3191.8556,-830.3799C3203.031,-821.1043 3216.056,-810.2936 3227.9392,-800.4304\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"3230.4875,-802.8639 3235.947,-793.784 3226.0168,-797.4775 3230.4875,-802.8639\" stroke=\"#000000\"/>\n</g>\n<!-- 140252874312728 -->\n<g class=\"node\" id=\"node18\">\n<title>140252874312728</title>\n<polygon fill=\"none\" points=\"0,-664.5 0,-710.5 504,-710.5 504,-664.5 0,-664.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"172\" y=\"-683.8\">global_average_pooling1d_5: GlobalAveragePooling1D</text>\n<polyline fill=\"none\" points=\"344,-664.5 344,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"373\" y=\"-695.3\">input:</text>\n<polyline fill=\"none\" points=\"344,-687.5 402,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"373\" y=\"-672.3\">output:</text>\n<polyline fill=\"none\" points=\"402,-664.5 402,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"453\" y=\"-695.3\">(None, 50, 32)</text>\n<polyline fill=\"none\" points=\"402,-687.5 504,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"453\" y=\"-672.3\">(None, 32)</text>\n</g>\n<!-- 140252874209656&#45;&gt;140252874312728 -->\n<g class=\"edge\" id=\"edge17\">\n<title>140252874209656-&gt;140252874312728</title>\n<path d=\"M614.8317,-747.4901C548.4565,-736.5157 468.9664,-723.3729 401.3293,-712.1899\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"401.8761,-708.7329 391.4391,-710.5547 400.7342,-715.6391 401.8761,-708.7329\" stroke=\"#000000\"/>\n</g>\n<!-- 140252874035328 -->\n<g class=\"node\" id=\"node19\">\n<title>140252874035328</title>\n<polygon fill=\"none\" points=\"522,-664.5 522,-710.5 986,-710.5 986,-664.5 522,-664.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"674\" y=\"-683.8\">global_max_pooling1d_5: GlobalMaxPooling1D</text>\n<polyline fill=\"none\" points=\"826,-664.5 826,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"855\" y=\"-695.3\">input:</text>\n<polyline fill=\"none\" points=\"826,-687.5 884,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"855\" y=\"-672.3\">output:</text>\n<polyline fill=\"none\" points=\"884,-664.5 884,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"935\" y=\"-695.3\">(None, 50, 32)</text>\n<polyline fill=\"none\" points=\"884,-687.5 986,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"935\" y=\"-672.3\">(None, 32)</text>\n</g>\n<!-- 140252874209656&#45;&gt;140252874035328 -->\n<g class=\"edge\" id=\"edge18\">\n<title>140252874209656-&gt;140252874035328</title>\n<path d=\"M754,-747.3799C754,-739.1745 754,-729.7679 754,-720.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"757.5001,-720.784 754,-710.784 750.5001,-720.784 757.5001,-720.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140252873362624 -->\n<g class=\"node\" id=\"node20\">\n<title>140252873362624</title>\n<polygon fill=\"none\" points=\"1004,-664.5 1004,-710.5 1508,-710.5 1508,-664.5 1004,-664.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1176\" y=\"-683.8\">global_average_pooling1d_6: GlobalAveragePooling1D</text>\n<polyline fill=\"none\" points=\"1348,-664.5 1348,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1377\" y=\"-695.3\">input:</text>\n<polyline fill=\"none\" points=\"1348,-687.5 1406,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1377\" y=\"-672.3\">output:</text>\n<polyline fill=\"none\" points=\"1406,-664.5 1406,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1457\" y=\"-695.3\">(None, 50, 32)</text>\n<polyline fill=\"none\" points=\"1406,-687.5 1508,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1457\" y=\"-672.3\">(None, 32)</text>\n</g>\n<!-- 140252873759488&#45;&gt;140252873362624 -->\n<g class=\"edge\" id=\"edge19\">\n<title>140252873759488-&gt;140252873362624</title>\n<path d=\"M1557.396,-747.4901C1502.8246,-736.6282 1437.581,-723.642 1381.7745,-712.5343\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"1382.3197,-709.0742 1371.8289,-710.5547 1380.9532,-715.9395 1382.3197,-709.0742\" stroke=\"#000000\"/>\n</g>\n<!-- 140252873065864 -->\n<g class=\"node\" id=\"node21\">\n<title>140252873065864</title>\n<polygon fill=\"none\" points=\"1526,-664.5 1526,-710.5 1990,-710.5 1990,-664.5 1526,-664.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1678\" y=\"-683.8\">global_max_pooling1d_6: GlobalMaxPooling1D</text>\n<polyline fill=\"none\" points=\"1830,-664.5 1830,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1859\" y=\"-695.3\">input:</text>\n<polyline fill=\"none\" points=\"1830,-687.5 1888,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1859\" y=\"-672.3\">output:</text>\n<polyline fill=\"none\" points=\"1888,-664.5 1888,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1939\" y=\"-695.3\">(None, 50, 32)</text>\n<polyline fill=\"none\" points=\"1888,-687.5 1990,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1939\" y=\"-672.3\">(None, 32)</text>\n</g>\n<!-- 140252873759488&#45;&gt;140252873065864 -->\n<g class=\"edge\" id=\"edge20\">\n<title>140252873759488-&gt;140252873065864</title>\n<path d=\"M1696.6772,-747.3799C1705.9937,-738.2827 1716.8223,-727.7088 1726.7646,-718.0005\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"1729.4454,-720.2746 1734.1549,-710.784 1724.5549,-715.2663 1729.4454,-720.2746\" stroke=\"#000000\"/>\n</g>\n<!-- 140252872996792 -->\n<g class=\"node\" id=\"node22\">\n<title>140252872996792</title>\n<polygon fill=\"none\" points=\"2008,-664.5 2008,-710.5 2512,-710.5 2512,-664.5 2008,-664.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2180\" y=\"-683.8\">global_average_pooling1d_7: GlobalAveragePooling1D</text>\n<polyline fill=\"none\" points=\"2352,-664.5 2352,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2381\" y=\"-695.3\">input:</text>\n<polyline fill=\"none\" points=\"2352,-687.5 2410,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2381\" y=\"-672.3\">output:</text>\n<polyline fill=\"none\" points=\"2410,-664.5 2410,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2461\" y=\"-695.3\">(None, 50, 32)</text>\n<polyline fill=\"none\" points=\"2410,-687.5 2512,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2461\" y=\"-672.3\">(None, 32)</text>\n</g>\n<!-- 140252872917400&#45;&gt;140252872996792 -->\n<g class=\"edge\" id=\"edge21\">\n<title>140252872917400-&gt;140252872996792</title>\n<path d=\"M2459.4851,-747.4901C2424.3381,-736.9205 2382.5025,-724.3395 2346.2422,-713.4352\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2347.248,-710.0829 2336.6637,-710.5547 2345.2321,-716.7863 2347.248,-710.0829\" stroke=\"#000000\"/>\n</g>\n<!-- 140252872718656 -->\n<g class=\"node\" id=\"node23\">\n<title>140252872718656</title>\n<polygon fill=\"none\" points=\"2530,-664.5 2530,-710.5 2994,-710.5 2994,-664.5 2530,-664.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2682\" y=\"-683.8\">global_max_pooling1d_7: GlobalMaxPooling1D</text>\n<polyline fill=\"none\" points=\"2834,-664.5 2834,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2863\" y=\"-695.3\">input:</text>\n<polyline fill=\"none\" points=\"2834,-687.5 2892,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2863\" y=\"-672.3\">output:</text>\n<polyline fill=\"none\" points=\"2892,-664.5 2892,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2943\" y=\"-695.3\">(None, 50, 32)</text>\n<polyline fill=\"none\" points=\"2892,-687.5 2994,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2943\" y=\"-672.3\">(None, 32)</text>\n</g>\n<!-- 140252872917400&#45;&gt;140252872718656 -->\n<g class=\"edge\" id=\"edge22\">\n<title>140252872917400-&gt;140252872718656</title>\n<path d=\"M2598.6535,-747.4901C2626.821,-737.1454 2660.2347,-724.874 2689.4801,-714.1334\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2691.0443,-717.2876 2699.2246,-710.5547 2688.631,-710.7167 2691.0443,-717.2876\" stroke=\"#000000\"/>\n</g>\n<!-- 140252872027384 -->\n<g class=\"node\" id=\"node24\">\n<title>140252872027384</title>\n<polygon fill=\"none\" points=\"3012,-664.5 3012,-710.5 3516,-710.5 3516,-664.5 3012,-664.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3184\" y=\"-683.8\">global_average_pooling1d_8: GlobalAveragePooling1D</text>\n<polyline fill=\"none\" points=\"3356,-664.5 3356,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3385\" y=\"-695.3\">input:</text>\n<polyline fill=\"none\" points=\"3356,-687.5 3414,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3385\" y=\"-672.3\">output:</text>\n<polyline fill=\"none\" points=\"3414,-664.5 3414,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3465\" y=\"-695.3\">(None, 50, 32)</text>\n<polyline fill=\"none\" points=\"3414,-687.5 3516,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3465\" y=\"-672.3\">(None, 32)</text>\n</g>\n<!-- 140252872427000&#45;&gt;140252872027384 -->\n<g class=\"edge\" id=\"edge23\">\n<title>140252872427000-&gt;140252872027384</title>\n<path d=\"M3264,-747.3799C3264,-739.1745 3264,-729.7679 3264,-720.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"3267.5001,-720.784 3264,-710.784 3260.5001,-720.784 3267.5001,-720.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140252872257888 -->\n<g class=\"node\" id=\"node25\">\n<title>140252872257888</title>\n<polygon fill=\"none\" points=\"3534,-664.5 3534,-710.5 3998,-710.5 3998,-664.5 3534,-664.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3686\" y=\"-683.8\">global_max_pooling1d_8: GlobalMaxPooling1D</text>\n<polyline fill=\"none\" points=\"3838,-664.5 3838,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3867\" y=\"-695.3\">input:</text>\n<polyline fill=\"none\" points=\"3838,-687.5 3896,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3867\" y=\"-672.3\">output:</text>\n<polyline fill=\"none\" points=\"3896,-664.5 3896,-710.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3947\" y=\"-695.3\">(None, 50, 32)</text>\n<polyline fill=\"none\" points=\"3896,-687.5 3998,-687.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"3947\" y=\"-672.3\">(None, 32)</text>\n</g>\n<!-- 140252872427000&#45;&gt;140252872257888 -->\n<g class=\"edge\" id=\"edge24\">\n<title>140252872427000-&gt;140252872257888</title>\n<path d=\"M3403.1683,-747.4901C3469.5435,-736.5157 3549.0336,-723.3729 3616.6707,-712.1899\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"3617.2658,-715.6391 3626.5609,-710.5547 3616.1239,-708.7329 3617.2658,-715.6391\" stroke=\"#000000\"/>\n</g>\n<!-- 140252872260352 -->\n<g class=\"node\" id=\"node26\">\n<title>140252872260352</title>\n<polygon fill=\"none\" points=\"1601.5,-581.5 1601.5,-627.5 2416.5,-627.5 2416.5,-581.5 1601.5,-581.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1689\" y=\"-600.8\">concatenate_2: Concatenate</text>\n<polyline fill=\"none\" points=\"1776.5,-581.5 1776.5,-627.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1805.5\" y=\"-612.3\">input:</text>\n<polyline fill=\"none\" points=\"1776.5,-604.5 1834.5,-604.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1805.5\" y=\"-589.3\">output:</text>\n<polyline fill=\"none\" points=\"1834.5,-581.5 1834.5,-627.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2125.5\" y=\"-612.3\">[(None, 32), (None, 32), (None, 32), (None, 32), (None, 32), (None, 32), (None, 32), (None, 32)]</text>\n<polyline fill=\"none\" points=\"1834.5,-604.5 2416.5,-604.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2125.5\" y=\"-589.3\">(None, 256)</text>\n</g>\n<!-- 140252874312728&#45;&gt;140252872260352 -->\n<g class=\"edge\" id=\"edge25\">\n<title>140252874312728-&gt;140252872260352</title>\n<path d=\"M504.0945,-664.5901C507.0795,-664.3883 510.049,-664.1914 513,-664 876.7058,-640.4083 1290.8022,-624.8098 1590.9716,-615.526\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"1591.2679,-619.0187 1601.1553,-615.2121 1591.0522,-612.022 1591.2679,-619.0187\" stroke=\"#000000\"/>\n</g>\n<!-- 140252874035328&#45;&gt;140252872260352 -->\n<g class=\"edge\" id=\"edge26\">\n<title>140252874035328-&gt;140252872260352</title>\n<path d=\"M986.1427,-664.7018C989.1127,-664.4631 992.0662,-664.229 995,-664 1190.8309,-648.7124 1408.217,-635.4638 1591.3256,-625.3741\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"1591.6046,-628.8641 1601.3974,-624.8204 1591.2203,-621.8747 1591.6046,-628.8641\" stroke=\"#000000\"/>\n</g>\n<!-- 140252873362624&#45;&gt;140252872260352 -->\n<g class=\"edge\" id=\"edge27\">\n<title>140252873362624-&gt;140252872260352</title>\n<path d=\"M1464.7525,-664.4901C1565.9507,-653.3354 1687.473,-639.9406 1789.9912,-628.6404\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"1790.6116,-632.0933 1800.168,-627.5187 1789.8447,-625.1355 1790.6116,-632.0933\" stroke=\"#000000\"/>\n</g>\n<!-- 140252873065864&#45;&gt;140252872260352 -->\n<g class=\"edge\" id=\"edge28\">\n<title>140252873065864-&gt;140252872260352</title>\n<path d=\"M1827.5842,-664.4901C1859.2756,-654.0105 1896.9466,-641.5535 1929.7265,-630.7139\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"1930.885,-634.0173 1939.2805,-627.5547 1928.6872,-627.3713 1930.885,-634.0173\" stroke=\"#000000\"/>\n</g>\n<!-- 140252872996792&#45;&gt;140252872260352 -->\n<g class=\"edge\" id=\"edge29\">\n<title>140252872996792-&gt;140252872260352</title>\n<path d=\"M2190.4158,-664.4901C2158.7244,-654.0105 2121.0534,-641.5535 2088.2735,-630.7139\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2089.3128,-627.3713 2078.7195,-627.5547 2087.115,-634.0173 2089.3128,-627.3713\" stroke=\"#000000\"/>\n</g>\n<!-- 140252872718656&#45;&gt;140252872260352 -->\n<g class=\"edge\" id=\"edge30\">\n<title>140252872718656-&gt;140252872260352</title>\n<path d=\"M2553.2475,-664.4901C2452.0493,-653.3354 2330.527,-639.9406 2228.0088,-628.6404\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2228.1553,-625.1355 2217.832,-627.5187 2227.3884,-632.0933 2228.1553,-625.1355\" stroke=\"#000000\"/>\n</g>\n<!-- 140252872027384&#45;&gt;140252872260352 -->\n<g class=\"edge\" id=\"edge31\">\n<title>140252872027384-&gt;140252872260352</title>\n<path d=\"M3011.8996,-664.6744C3008.9164,-664.4459 3005.949,-664.221 3003,-664 2813.9231,-649.8321 2604.4106,-636.8829 2426.6772,-626.7086\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2426.8146,-623.2109 2416.6312,-626.1346 2426.4152,-630.1995 2426.8146,-623.2109\" stroke=\"#000000\"/>\n</g>\n<!-- 140252872257888&#45;&gt;140252872260352 -->\n<g class=\"edge\" id=\"edge32\">\n<title>140252872257888-&gt;140252872260352</title>\n<path d=\"M3533.8632,-664.621C3530.8914,-664.4078 3527.9359,-664.2006 3525,-664 3154.2504,-638.6698 2731.7023,-623.1876 2427.0557,-614.392\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2426.8175,-610.8838 2416.721,-614.0949 2426.6163,-617.8809 2426.8175,-610.8838\" stroke=\"#000000\"/>\n</g>\n<!-- 140252871885096 -->\n<g class=\"node\" id=\"node27\">\n<title>140252871885096</title>\n<polygon fill=\"none\" points=\"1801,-498.5 1801,-544.5 2217,-544.5 2217,-498.5 1801,-498.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1936.5\" y=\"-517.8\">batch_normalization_3: BatchNormalization</text>\n<polyline fill=\"none\" points=\"2072,-498.5 2072,-544.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2101\" y=\"-529.3\">input:</text>\n<polyline fill=\"none\" points=\"2072,-521.5 2130,-521.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2101\" y=\"-506.3\">output:</text>\n<polyline fill=\"none\" points=\"2130,-498.5 2130,-544.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2173.5\" y=\"-529.3\">(None, 256)</text>\n<polyline fill=\"none\" points=\"2130,-521.5 2217,-521.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2173.5\" y=\"-506.3\">(None, 256)</text>\n</g>\n<!-- 140252872260352&#45;&gt;140252871885096 -->\n<g class=\"edge\" id=\"edge33\">\n<title>140252872260352-&gt;140252871885096</title>\n<path d=\"M2009,-581.3799C2009,-573.1745 2009,-563.7679 2009,-554.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2012.5001,-554.784 2009,-544.784 2005.5001,-554.784 2012.5001,-554.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140252871974360 -->\n<g class=\"node\" id=\"node28\">\n<title>140252871974360</title>\n<polygon fill=\"none\" points=\"1883,-415.5 1883,-461.5 2135,-461.5 2135,-415.5 1883,-415.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1936.5\" y=\"-434.8\">dense_4: Dense</text>\n<polyline fill=\"none\" points=\"1990,-415.5 1990,-461.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2019\" y=\"-446.3\">input:</text>\n<polyline fill=\"none\" points=\"1990,-438.5 2048,-438.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2019\" y=\"-423.3\">output:</text>\n<polyline fill=\"none\" points=\"2048,-415.5 2048,-461.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2091.5\" y=\"-446.3\">(None, 256)</text>\n<polyline fill=\"none\" points=\"2048,-438.5 2135,-438.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2091.5\" y=\"-423.3\">(None, 64)</text>\n</g>\n<!-- 140252871885096&#45;&gt;140252871974360 -->\n<g class=\"edge\" id=\"edge34\">\n<title>140252871885096-&gt;140252871974360</title>\n<path d=\"M2009,-498.3799C2009,-490.1745 2009,-480.7679 2009,-471.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2012.5001,-471.784 2009,-461.784 2005.5001,-471.784 2012.5001,-471.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140252871887504 -->\n<g class=\"node\" id=\"node29\">\n<title>140252871887504</title>\n<polygon fill=\"none\" points=\"1873,-332.5 1873,-378.5 2145,-378.5 2145,-332.5 1873,-332.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1940\" y=\"-351.8\">dropout_3: Dropout</text>\n<polyline fill=\"none\" points=\"2007,-332.5 2007,-378.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2036\" y=\"-363.3\">input:</text>\n<polyline fill=\"none\" points=\"2007,-355.5 2065,-355.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2036\" y=\"-340.3\">output:</text>\n<polyline fill=\"none\" points=\"2065,-332.5 2065,-378.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2105\" y=\"-363.3\">(None, 64)</text>\n<polyline fill=\"none\" points=\"2065,-355.5 2145,-355.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2105\" y=\"-340.3\">(None, 64)</text>\n</g>\n<!-- 140252871974360&#45;&gt;140252871887504 -->\n<g class=\"edge\" id=\"edge35\">\n<title>140252871974360-&gt;140252871887504</title>\n<path d=\"M2009,-415.3799C2009,-407.1745 2009,-397.7679 2009,-388.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2012.5001,-388.784 2009,-378.784 2005.5001,-388.784 2012.5001,-388.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140252871587880 -->\n<g class=\"node\" id=\"node30\">\n<title>140252871587880</title>\n<polygon fill=\"none\" points=\"1804.5,-249.5 1804.5,-295.5 2213.5,-295.5 2213.5,-249.5 1804.5,-249.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1940\" y=\"-268.8\">batch_normalization_4: BatchNormalization</text>\n<polyline fill=\"none\" points=\"2075.5,-249.5 2075.5,-295.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2104.5\" y=\"-280.3\">input:</text>\n<polyline fill=\"none\" points=\"2075.5,-272.5 2133.5,-272.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2104.5\" y=\"-257.3\">output:</text>\n<polyline fill=\"none\" points=\"2133.5,-249.5 2133.5,-295.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2173.5\" y=\"-280.3\">(None, 64)</text>\n<polyline fill=\"none\" points=\"2133.5,-272.5 2213.5,-272.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2173.5\" y=\"-257.3\">(None, 64)</text>\n</g>\n<!-- 140252871887504&#45;&gt;140252871587880 -->\n<g class=\"edge\" id=\"edge36\">\n<title>140252871887504-&gt;140252871587880</title>\n<path d=\"M2009,-332.3799C2009,-324.1745 2009,-314.7679 2009,-305.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2012.5001,-305.784 2009,-295.784 2005.5001,-305.784 2012.5001,-305.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140252870904408 -->\n<g class=\"node\" id=\"node31\">\n<title>140252870904408</title>\n<polygon fill=\"none\" points=\"1886.5,-166.5 1886.5,-212.5 2131.5,-212.5 2131.5,-166.5 1886.5,-166.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1940\" y=\"-185.8\">dense_5: Dense</text>\n<polyline fill=\"none\" points=\"1993.5,-166.5 1993.5,-212.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2022.5\" y=\"-197.3\">input:</text>\n<polyline fill=\"none\" points=\"1993.5,-189.5 2051.5,-189.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2022.5\" y=\"-174.3\">output:</text>\n<polyline fill=\"none\" points=\"2051.5,-166.5 2051.5,-212.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2091.5\" y=\"-197.3\">(None, 64)</text>\n<polyline fill=\"none\" points=\"2051.5,-189.5 2131.5,-189.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2091.5\" y=\"-174.3\">(None, 32)</text>\n</g>\n<!-- 140252871587880&#45;&gt;140252870904408 -->\n<g class=\"edge\" id=\"edge37\">\n<title>140252871587880-&gt;140252870904408</title>\n<path d=\"M2009,-249.3799C2009,-241.1745 2009,-231.7679 2009,-222.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2012.5001,-222.784 2009,-212.784 2005.5001,-222.784 2012.5001,-222.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140252870753864 -->\n<g class=\"node\" id=\"node32\">\n<title>140252870753864</title>\n<polygon fill=\"none\" points=\"1873,-83.5 1873,-129.5 2145,-129.5 2145,-83.5 1873,-83.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1940\" y=\"-102.8\">dropout_4: Dropout</text>\n<polyline fill=\"none\" points=\"2007,-83.5 2007,-129.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2036\" y=\"-114.3\">input:</text>\n<polyline fill=\"none\" points=\"2007,-106.5 2065,-106.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2036\" y=\"-91.3\">output:</text>\n<polyline fill=\"none\" points=\"2065,-83.5 2065,-129.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2105\" y=\"-114.3\">(None, 32)</text>\n<polyline fill=\"none\" points=\"2065,-106.5 2145,-106.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2105\" y=\"-91.3\">(None, 32)</text>\n</g>\n<!-- 140252870904408&#45;&gt;140252870753864 -->\n<g class=\"edge\" id=\"edge38\">\n<title>140252870904408-&gt;140252870753864</title>\n<path d=\"M2009,-166.3799C2009,-158.1745 2009,-148.7679 2009,-139.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2012.5001,-139.784 2009,-129.784 2005.5001,-139.784 2012.5001,-139.784\" stroke=\"#000000\"/>\n</g>\n<!-- 140252870690296 -->\n<g class=\"node\" id=\"node33\">\n<title>140252870690296</title>\n<polygon fill=\"none\" points=\"1886.5,-.5 1886.5,-46.5 2131.5,-46.5 2131.5,-.5 1886.5,-.5\" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"1940\" y=\"-19.8\">dense_6: Dense</text>\n<polyline fill=\"none\" points=\"1993.5,-.5 1993.5,-46.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2022.5\" y=\"-31.3\">input:</text>\n<polyline fill=\"none\" points=\"1993.5,-23.5 2051.5,-23.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2022.5\" y=\"-8.3\">output:</text>\n<polyline fill=\"none\" points=\"2051.5,-.5 2051.5,-46.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2091.5\" y=\"-31.3\">(None, 32)</text>\n<polyline fill=\"none\" points=\"2051.5,-23.5 2131.5,-23.5 \" stroke=\"#000000\"/>\n<text fill=\"#000000\" font-family=\"Times,serif\" font-size=\"14.00\" text-anchor=\"middle\" x=\"2091.5\" y=\"-8.3\">(None, 5)</text>\n</g>\n<!-- 140252870753864&#45;&gt;140252870690296 -->\n<g class=\"edge\" id=\"edge39\">\n<title>140252870753864-&gt;140252870690296</title>\n<path d=\"M2009,-83.3799C2009,-75.1745 2009,-65.7679 2009,-56.8786\" fill=\"none\" stroke=\"#000000\"/>\n<polygon fill=\"#000000\" points=\"2012.5001,-56.784 2009,-46.784 2005.5001,-56.784 2012.5001,-56.784\" stroke=\"#000000\"/>\n</g>\n</g>\n</svg>"
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 15
        }
      ]
    },
    {
      "metadata": {
        "id": "bHYz1kXFLhIB",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "test_preds = np.zeros((test.shape[0], 5))\n",
        "file_path = \"/content/drive/My Drive/DeepLearning/best_model.hdf5\"\n",
        "check_point = ModelCheckpoint(file_path, monitor = \"val_loss\", verbose = 1,\n",
        "                              save_best_only = True, mode = \"min\")\n",
        "early_stop = EarlyStopping(monitor = \"val_loss\", mode = \"min\", patience = 2)\n",
        "\n",
        "print(\"Building the model...\")\n",
        "model3 = build_model3(lr = 1e-3, lr_d = 0, units = 64, spatial_dr = 0.5, kernel_size1=3, kernel_size2=3, dense_units=64, dr=0.3, conv_size=32)\n",
        "\n",
        "for i in range(NUM_FOLDS):\n",
        "    print(\"FOLD\", i+1)    \n",
        "    print(\"Splitting the data into train and validation...\")\n",
        "    train_seq, val_seq = X_train[train[\"fold_id\"] != i], X_train[train[\"fold_id\"] == i]\n",
        "    y_train = ohe.transform(train[train[\"fold_id\"] != i][\"Sentiment\"].values.reshape(-1, 1))\n",
        "    y_val = ohe.transform(train[train[\"fold_id\"] == i][\"Sentiment\"].values.reshape(-1, 1))        \n",
        "    \n",
        "    print(\"Training the model...\")\n",
        "    model3.fit(train_seq, y_train, validation_data = (val_seq, y_val), batch_size = 128, epochs = 15, verbose = 1, callbacks = [early_stop]) \n",
        "    model3.save_weights(file_path)  \n",
        "    test_preds += model2.predict([X_test], batch_size=1024, verbose=1)    \n",
        "    print()\n",
        "    \n",
        "print(\"Save model after cross-validation...\")\n",
        "model3.save_weights(file_path)   \n",
        "test_preds /= NUM_FOLDS\n",
        "\n",
        "\n",
        "print(\"Make the submission ready...\")\n",
        "sub = pd.read_csv('/content/drive/My Drive/DeepLearning/sampleSubmission.csv', sep=\",\")\n",
        "\n",
        "pred = model3.predict(X_test, batch_size = 1024, verbose = 1)\n",
        "predictions = np.round(np.argmax(pred, axis=1)).astype(int)\n",
        "sub['Sentiment'] = predictions\n",
        "sub.to_csv(\"/content/drive/My Drive/DeepLearning/blend.csv\", index=False)\n",
        "\n",
        "predictions = np.round(np.argmax(test_preds, axis=1)).astype(int)\n",
        "sub['Sentiment'] = predictions\n",
        "sub.to_csv(\"/content/drive/My Drive/DeepLearning/avg_blend.csv\", index=False)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "5u0R1sLfthlo",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "def build_model4(lr=0.0, lr_d=0.0, units=0, spatial_dr=0.0, kernel_size1=3, kernel_size2=2, dense_units=128, dr=0.1, conv_size=32, ktop=5):\n",
        "    inp = Input(shape = (max_len,))\n",
        "    x = Embedding(19479, embed_size, weights = [embedding_matrix], trainable = False)(inp)\n",
        "    x1 = SpatialDropout1D(spatial_dr)(x)\n",
        "\n",
        "    x_gru = Bidirectional(CuDNNGRU(units, return_sequences = True))(x1)\n",
        "    x_lstm = Bidirectional(CuDNNLSTM(units, return_sequences = True))(x1)\n",
        "    \n",
        "    x_conv1 = Conv1D(conv_size, kernel_size=kernel_size1, padding='valid', kernel_initializer='he_uniform')(x_gru)\n",
        "    x_conv1 = LeakyReLU(0.2)(x_conv1)\n",
        "    avg_pool1_gru = GlobalAveragePooling1D()(x_conv1)\n",
        "    max_pool1_gru = GlobalMaxPooling1D()(x_conv1)\n",
        "    #dyn_pool1_gru = DynamicKMaxPoolLayer(x_conv1,ktop,nroflayers=2,layernr=1)\n",
        "    \n",
        "    x_conv2 = Conv1D(conv_size, kernel_size=kernel_size2, padding='valid', kernel_initializer='he_uniform')(x_gru)\n",
        "    x_conv2 = LeakyReLU(0.2)(x_conv2)\n",
        "    avg_pool2_gru = GlobalAveragePooling1D()(x_conv2)\n",
        "    max_pool2_gru = GlobalMaxPooling1D()(x_conv2)\n",
        "    #dyn_pool2_gru = DynamicKMaxPoolLayer(x_conv2,ktop,nroflayers=2,layernr=1)\n",
        "\n",
        "    \n",
        "    x_conv3 = Conv1D(conv_size, kernel_size=kernel_size1, padding='valid', kernel_initializer='he_uniform')(x_lstm)\n",
        "    x_conv3 = LeakyReLU(0.2)(x_conv3)\n",
        "    avg_pool1_lstm = GlobalAveragePooling1D()(x_conv3)\n",
        "    max_pool1_lstm = GlobalMaxPooling1D()(x_conv3)\n",
        "    #dyn_pool1_lstm = DynamicKMaxPoolLayer(x_conv3,ktop,nroflayers=2,layernr=1)\n",
        "    \n",
        "    x_conv4 = Conv1D(conv_size, kernel_size=kernel_size2, padding='valid', kernel_initializer='he_uniform')(x_lstm)\n",
        "    x_conv4 = LeakyReLU(0.2)(x_conv4)\n",
        "    avg_pool2_lstm = GlobalAveragePooling1D()(x_conv4)\n",
        "    max_pool2_lstm = GlobalMaxPooling1D()(x_conv4)\n",
        "    #dyn_pool2_lstm = DynamicKMaxPoolLayer(x_conv4,ktop,nroflayers=2,layernr=1)\n",
        "    \n",
        "    x = concatenate([avg_pool1_gru, max_pool1_gru, avg_pool2_gru, max_pool2_gru,\n",
        "                    avg_pool1_lstm, max_pool1_lstm, avg_pool2_lstm, max_pool2_lstm])\n",
        "    x = BatchNormalization()(x)\n",
        "    x = Dropout(dr)(Dense(dense_units, activation='relu') (x))\n",
        "    x = BatchNormalization()(x)\n",
        "    x = Dropout(dr)(Dense(int(dense_units / 2), activation='relu') (x))\n",
        "    x = Dense(5, activation = \"sigmoid\")(x)\n",
        "    model = Model(inputs = inp, outputs = x)\n",
        "    model.compile(loss = \"binary_crossentropy\", optimizer = Adam(lr = lr, decay = lr_d), metrics = [\"accuracy\"])\n",
        "    return model"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "JnzQurfeg0Mo",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "###Model 4: Embedding + LSTM/GRU + Multi-layer CNN"
      ]
    },
    {
      "metadata": {
        "id": "L3kxtJabtkYT",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "model4 = build_model4(lr = 1e-3, lr_d = 0, units = 64, spatial_dr = 0.5, kernel_size1=3, kernel_size2=3, dense_units=64, dr=0.3, conv_size=32, ktop=5)\n",
        "\n",
        "model4.summary()\n",
        "SVG(model_to_dot(model4,  show_shapes=True, show_layer_names=True, rankdir='HB').create(prog='dot', format='svg'))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "sjHlbhfUtkci",
        "colab_type": "code",
        "outputId": "e04e3c21-4ae2-41a4-d908-220341c3b8a2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 3313
        }
      },
      "cell_type": "code",
      "source": [
        "NUM_FOLDS = 20\n",
        "train[\"fold_id\"] = train[\"SentenceId\"].apply(lambda x: x%NUM_FOLDS)\n",
        "test_preds = np.zeros((test.shape[0], 5))\n",
        "file_path = \"/content/drive/My Drive/DeepLearning/best_model.hdf5\"\n",
        "check_point = ModelCheckpoint(file_path, monitor = \"val_loss\", verbose = 1,\n",
        "                              save_best_only = True, mode = \"min\")\n",
        "early_stop = EarlyStopping(monitor = \"val_loss\", mode = \"min\", patience = 2)\n",
        "\n",
        "print(\"Building the model...\")\n",
        "model4 = build_model4(lr = 1e-3, lr_d = 0, units = 64, spatial_dr = 0.5, kernel_size1=3, kernel_size2=4, dense_units=64, dr=0.3, conv_size=32)\n",
        "\n",
        "for i in range(NUM_FOLDS):\n",
        "    print(\"FOLD\", i+1)    \n",
        "    print(\"Splitting the data into train and validation...\")\n",
        "    train_seq, val_seq = X_train[train[\"fold_id\"] != i], X_train[train[\"fold_id\"] == i]\n",
        "    y_train = ohe.transform(train[train[\"fold_id\"] != i][\"Sentiment\"].values.reshape(-1, 1))\n",
        "    y_val = ohe.transform(train[train[\"fold_id\"] == i][\"Sentiment\"].values.reshape(-1, 1))        \n",
        "    \n",
        "    print(\"Training the model...\")\n",
        "    model4.fit(train_seq, y_train, validation_data = (val_seq, y_val), batch_size = 128, epochs = 15, verbose = 1, callbacks = [early_stop]) \n",
        "    model4.save_weights(file_path)  \n",
        "    test_preds += model4.predict([X_test], batch_size=1024, verbose=1)    \n",
        "    print()\n",
        "    \n",
        "print(\"Save model after cross-validation...\")\n",
        "model4.save_weights(file_path)   \n",
        "test_preds /= NUM_FOLDS\n",
        "\n",
        "\n",
        "print(\"Make the submission ready...\")\n",
        "sub = pd.read_csv('/content/drive/My Drive/DeepLearning/sampleSubmission.csv', sep=\",\")\n",
        "\n",
        "pred = model4.predict(X_test, batch_size = 1024, verbose = 1)\n",
        "predictions = np.round(np.argmax(pred, axis=1)).astype(int)\n",
        "submission = pd.DataFrame({'PhraseId':test['PhraseId'],'Sentiment': predictions})\n",
        "submission =pd.merge(submission, save_test, on='PhraseId', how='left')\n",
        "submission[\"Sentiment\"] = submission.apply(get_sentiment, axis=1)\n",
        "submission.drop(['Sentiment_x', 'Sentiment_y'], axis=1,inplace=True)\n",
        "submission[\"Sentiment\"] = submission[\"Sentiment\"].astype(int)\n",
        "submission.to_csv(\"/content/drive/My Drive/DeepLearning/blend.csv\", index=False)\n",
        "\n",
        "predictions = np.round(np.argmax(test_preds, axis=1)).astype(int)\n",
        "submission = pd.DataFrame({'PhraseId':test['PhraseId'],'Sentiment': predictions})\n",
        "submission =pd.merge(submission, save_test, on='PhraseId', how='left')\n",
        "submission[\"Sentiment\"] = submission.apply(get_sentiment, axis=1)\n",
        "submission.drop(['Sentiment_x', 'Sentiment_y'], axis=1,inplace=True)\n",
        "submission[\"Sentiment\"] = submission[\"Sentiment\"].astype(int)\n",
        "submission.to_csv(\"/content/drive/My Drive/DeepLearning/avg_blend.csv\", index=False)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Building the model...\n",
            "FOLD 1\n",
            "Splitting the data into train and validation...\n",
            "Training the model...\n",
            "Train on 148193 samples, validate on 7867 samples\n",
            "Epoch 1/15\n",
            "148193/148193 [==============================] - 82s 553us/step - loss: 0.3688 - acc: 0.8348 - val_loss: 0.3077 - val_acc: 0.8578\n",
            "Epoch 2/15\n",
            "148193/148193 [==============================] - 72s 485us/step - loss: 0.3255 - acc: 0.8525 - val_loss: 0.3013 - val_acc: 0.8608\n",
            "Epoch 3/15\n",
            "148193/148193 [==============================] - 72s 485us/step - loss: 0.3152 - acc: 0.8566 - val_loss: 0.3010 - val_acc: 0.8609\n",
            "Epoch 4/15\n",
            "148193/148193 [==============================] - 72s 486us/step - loss: 0.3069 - acc: 0.8599 - val_loss: 0.3034 - val_acc: 0.8624\n",
            "Epoch 5/15\n",
            "148193/148193 [==============================] - 72s 487us/step - loss: 0.3018 - acc: 0.8617 - val_loss: 0.2919 - val_acc: 0.8679\n",
            "Epoch 6/15\n",
            "148193/148193 [==============================] - 71s 476us/step - loss: 0.2963 - acc: 0.8638 - val_loss: 0.2904 - val_acc: 0.8672\n",
            "Epoch 7/15\n",
            "148193/148193 [==============================] - 71s 479us/step - loss: 0.2922 - acc: 0.8662 - val_loss: 0.2911 - val_acc: 0.8676\n",
            "Epoch 8/15\n",
            "148193/148193 [==============================] - 73s 490us/step - loss: 0.2885 - acc: 0.8681 - val_loss: 0.2927 - val_acc: 0.8675\n",
            "66292/66292 [==============================] - 4s 59us/step\n",
            "\n",
            "FOLD 2\n",
            "Splitting the data into train and validation...\n",
            "Training the model...\n",
            "Train on 148316 samples, validate on 7744 samples\n",
            "Epoch 1/15\n",
            "148316/148316 [==============================] - 73s 494us/step - loss: 0.2862 - acc: 0.8692 - val_loss: 0.2620 - val_acc: 0.8816\n",
            "Epoch 2/15\n",
            "148316/148316 [==============================] - 72s 486us/step - loss: 0.2835 - acc: 0.8705 - val_loss: 0.2649 - val_acc: 0.8789\n",
            "Epoch 3/15\n",
            "148316/148316 [==============================] - 72s 483us/step - loss: 0.2817 - acc: 0.8716 - val_loss: 0.2667 - val_acc: 0.8789\n",
            "66292/66292 [==============================] - 3s 50us/step\n",
            "\n",
            "FOLD 3\n",
            "Splitting the data into train and validation...\n",
            "Training the model...\n",
            "Train on 148487 samples, validate on 7573 samples\n",
            "Epoch 1/15\n",
            "148487/148487 [==============================] - 73s 494us/step - loss: 0.2793 - acc: 0.8726 - val_loss: 0.2469 - val_acc: 0.8883\n",
            "Epoch 2/15\n",
            "148487/148487 [==============================] - 72s 487us/step - loss: 0.2774 - acc: 0.8739 - val_loss: 0.2498 - val_acc: 0.8864\n",
            "Epoch 3/15\n",
            "148487/148487 [==============================] - 72s 488us/step - loss: 0.2761 - acc: 0.8748 - val_loss: 0.2508 - val_acc: 0.8862\n",
            "66292/66292 [==============================] - 3s 50us/step\n",
            "\n",
            "FOLD 4\n",
            "Splitting the data into train and validation...\n",
            "Training the model...\n",
            "Train on 148073 samples, validate on 7987 samples\n",
            "Epoch 1/15\n",
            "148073/148073 [==============================] - 73s 494us/step - loss: 0.2748 - acc: 0.8758 - val_loss: 0.2422 - val_acc: 0.8906\n",
            "Epoch 2/15\n",
            "148073/148073 [==============================] - 72s 487us/step - loss: 0.2727 - acc: 0.8764 - val_loss: 0.2441 - val_acc: 0.8889\n",
            "Epoch 3/15\n",
            "148073/148073 [==============================] - 72s 487us/step - loss: 0.2710 - acc: 0.8777 - val_loss: 0.2463 - val_acc: 0.8875\n",
            "66292/66292 [==============================] - 3s 50us/step\n",
            "\n",
            "FOLD 5\n",
            "Splitting the data into train and validation...\n",
            "Training the model...\n",
            "Train on 148054 samples, validate on 8006 samples\n",
            "Epoch 1/15\n",
            "148054/148054 [==============================] - 73s 496us/step - loss: 0.2698 - acc: 0.8782 - val_loss: 0.2416 - val_acc: 0.8935\n",
            "Epoch 2/15\n",
            "148054/148054 [==============================] - 72s 486us/step - loss: 0.2683 - acc: 0.8788 - val_loss: 0.2398 - val_acc: 0.8924\n",
            "Epoch 3/15\n",
            "148054/148054 [==============================] - 74s 502us/step - loss: 0.2668 - acc: 0.8794 - val_loss: 0.2445 - val_acc: 0.8901\n",
            "Epoch 4/15\n",
            "148054/148054 [==============================] - 73s 492us/step - loss: 0.2663 - acc: 0.8800 - val_loss: 0.2467 - val_acc: 0.8892\n",
            "66292/66292 [==============================] - 3s 50us/step\n",
            "\n",
            "FOLD 6\n",
            "Splitting the data into train and validation...\n",
            "Training the model...\n",
            "Train on 148575 samples, validate on 7485 samples\n",
            "Epoch 1/15\n",
            "148575/148575 [==============================] - 74s 496us/step - loss: 0.2655 - acc: 0.8804 - val_loss: 0.2380 - val_acc: 0.8942\n",
            "Epoch 2/15\n",
            "148575/148575 [==============================] - 74s 499us/step - loss: 0.2644 - acc: 0.8812 - val_loss: 0.2420 - val_acc: 0.8915\n",
            "Epoch 3/15\n",
            "148575/148575 [==============================] - 72s 487us/step - loss: 0.2630 - acc: 0.8818 - val_loss: 0.2433 - val_acc: 0.8915\n",
            "66292/66292 [==============================] - 3s 50us/step\n",
            "\n",
            "FOLD 7\n",
            "Splitting the data into train and validation...\n",
            "Training the model...\n",
            "Train on 148126 samples, validate on 7934 samples\n",
            "Epoch 1/15\n",
            "148126/148126 [==============================] - 74s 497us/step - loss: 0.2630 - acc: 0.8820 - val_loss: 0.2343 - val_acc: 0.8967\n",
            "Epoch 2/15\n",
            "148126/148126 [==============================] - 74s 500us/step - loss: 0.2615 - acc: 0.8823 - val_loss: 0.2370 - val_acc: 0.8955\n",
            "Epoch 3/15\n",
            "148126/148126 [==============================] - 72s 488us/step - loss: 0.2610 - acc: 0.8834 - val_loss: 0.2386 - val_acc: 0.8938\n",
            "66292/66292 [==============================] - 3s 50us/step\n",
            "\n",
            "FOLD 8\n",
            "Splitting the data into train and validation...\n",
            "Training the model...\n",
            "Train on 148199 samples, validate on 7861 samples\n",
            "Epoch 1/15\n",
            "148199/148199 [==============================] - 73s 495us/step - loss: 0.2612 - acc: 0.8826 - val_loss: 0.2253 - val_acc: 0.9020\n",
            "Epoch 2/15\n",
            "148199/148199 [==============================] - 74s 499us/step - loss: 0.2593 - acc: 0.8839 - val_loss: 0.2315 - val_acc: 0.8985\n",
            "Epoch 3/15\n",
            "148199/148199 [==============================] - 74s 502us/step - loss: 0.2590 - acc: 0.8841 - val_loss: 0.2306 - val_acc: 0.8986\n",
            "66292/66292 [==============================] - 3s 50us/step\n",
            "\n",
            "FOLD 9\n",
            "Splitting the data into train and validation...\n",
            "Training the model...\n",
            "Train on 148078 samples, validate on 7982 samples\n",
            "Epoch 1/15\n",
            "148078/148078 [==============================] - 74s 499us/step - loss: 0.2591 - acc: 0.8842 - val_loss: 0.2253 - val_acc: 0.9005\n",
            "Epoch 2/15\n",
            "148078/148078 [==============================] - 72s 487us/step - loss: 0.2579 - acc: 0.8844 - val_loss: 0.2246 - val_acc: 0.9013\n",
            "Epoch 3/15\n",
            "148078/148078 [==============================] - 72s 488us/step - loss: 0.2569 - acc: 0.8852 - val_loss: 0.2300 - val_acc: 0.8980\n",
            "Epoch 4/15\n",
            "148078/148078 [==============================] - 73s 494us/step - loss: 0.2569 - acc: 0.8850 - val_loss: 0.2293 - val_acc: 0.8986\n",
            "66292/66292 [==============================] - 3s 50us/step\n",
            "\n",
            "FOLD 10\n",
            "Splitting the data into train and validation...\n",
            "Training the model...\n",
            "Train on 148142 samples, validate on 7918 samples\n",
            "Epoch 1/15\n",
            "148142/148142 [==============================] - 73s 492us/step - loss: 0.2570 - acc: 0.8853 - val_loss: 0.2174 - val_acc: 0.9064\n",
            "Epoch 2/15\n",
            "148142/148142 [==============================] - 72s 487us/step - loss: 0.2560 - acc: 0.8857 - val_loss: 0.2184 - val_acc: 0.9055\n",
            "Epoch 3/15\n",
            "148142/148142 [==============================] - 73s 492us/step - loss: 0.2556 - acc: 0.8860 - val_loss: 0.2177 - val_acc: 0.9063\n",
            "66292/66292 [==============================] - 3s 50us/step\n",
            "\n",
            "FOLD 11\n",
            "Splitting the data into train and validation...\n",
            "Training the model...\n",
            "Train on 148164 samples, validate on 7896 samples\n",
            "Epoch 1/15\n",
            "148164/148164 [==============================] - 74s 496us/step - loss: 0.2551 - acc: 0.8861 - val_loss: 0.2121 - val_acc: 0.9096\n",
            "Epoch 2/15\n",
            "148164/148164 [==============================] - 73s 490us/step - loss: 0.2543 - acc: 0.8863 - val_loss: 0.2162 - val_acc: 0.9085\n",
            "Epoch 3/15\n",
            "148164/148164 [==============================] - 72s 485us/step - loss: 0.2536 - acc: 0.8868 - val_loss: 0.2170 - val_acc: 0.9074\n",
            "66292/66292 [==============================] - 3s 50us/step\n",
            "\n",
            "FOLD 12\n",
            "Splitting the data into train and validation...\n",
            "Training the model...\n",
            "Train on 148072 samples, validate on 7988 samples\n",
            "Epoch 1/15\n",
            "148072/148072 [==============================] - 73s 494us/step - loss: 0.2537 - acc: 0.8865 - val_loss: 0.2132 - val_acc: 0.9088\n",
            "Epoch 2/15\n",
            "148072/148072 [==============================] - 72s 486us/step - loss: 0.2534 - acc: 0.8875 - val_loss: 0.2157 - val_acc: 0.9084\n",
            "Epoch 3/15\n",
            "148072/148072 [==============================] - 73s 496us/step - loss: 0.2522 - acc: 0.8877 - val_loss: 0.2175 - val_acc: 0.9065\n",
            "66292/66292 [==============================] - 3s 50us/step\n",
            "\n",
            "FOLD 13\n",
            "Splitting the data into train and validation...\n",
            "Training the model...\n",
            "Train on 148135 samples, validate on 7925 samples\n",
            "Epoch 1/15\n",
            "148135/148135 [==============================] - 71s 480us/step - loss: 0.2536 - acc: 0.8876 - val_loss: 0.2040 - val_acc: 0.9124\n",
            "Epoch 2/15\n",
            "148135/148135 [==============================] - 70s 473us/step - loss: 0.2534 - acc: 0.8872 - val_loss: 0.2070 - val_acc: 0.9097\n",
            "Epoch 3/15\n",
            "148135/148135 [==============================] - 71s 479us/step - loss: 0.2521 - acc: 0.8878 - val_loss: 0.2117 - val_acc: 0.9092\n",
            "66292/66292 [==============================] - 3s 50us/step\n",
            "\n",
            "FOLD 14\n",
            "Splitting the data into train and validation...\n",
            "Training the model...\n",
            "Train on 148300 samples, validate on 7760 samples\n",
            "Epoch 1/15\n",
            "148300/148300 [==============================] - 71s 476us/step - loss: 0.2510 - acc: 0.8883 - val_loss: 0.2099 - val_acc: 0.9081\n",
            "Epoch 2/15\n",
            "148300/148300 [==============================] - 71s 480us/step - loss: 0.2511 - acc: 0.8882 - val_loss: 0.2124 - val_acc: 0.9064\n",
            "Epoch 3/15\n",
            "148300/148300 [==============================] - 71s 476us/step - loss: 0.2505 - acc: 0.8886 - val_loss: 0.2164 - val_acc: 0.9053\n",
            "66292/66292 [==============================] - 3s 50us/step\n",
            "\n",
            "FOLD 15\n",
            "Splitting the data into train and validation...\n",
            "Training the model...\n",
            "Train on 148443 samples, validate on 7617 samples\n",
            "Epoch 1/15\n",
            " 79744/148443 [===============>..............] - ETA: 31s - loss: 0.2501 - acc: 0.8893Buffered data was truncated after reaching the output size limit."
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "RnHdy1BuNbsE",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "The following code is trying to use dynamic maximum pooling"
      ]
    },
    {
      "metadata": {
        "id": "dLPox7LMfnTo",
        "colab_type": "text"
      },
      "cell_type": "markdown",
      "source": [
        "### Model5 Undersampling Embedding + {LSTM/GRU} + CNN"
      ]
    },
    {
      "metadata": {
        "id": "aYuVU-5bfmMO",
        "colab_type": "code",
        "outputId": "a58cec13-d203-4dce-a15d-dbf02e0f7efc",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 620
        }
      },
      "cell_type": "code",
      "source": [
        "train = pd.read_csv('/content/drive/My Drive/DeepLearning/train.tsv', sep=\"\\t\")\n",
        "test = pd.read_csv('/content/drive/My Drive/DeepLearning/test.tsv', sep=\"\\t\")\n",
        "sub = pd.read_csv('/content/drive/My Drive/DeepLearning/sampleSubmission.csv', sep=\",\")\n",
        "y = train['Sentiment']\n",
        "class2 = train[train['Sentiment']==2]\n",
        "class2Sample = class2.sample(frac=0.5) #, random_state=3\n",
        "train = pd.concat([train[train['Sentiment']!=2], class2Sample])\n",
        "\n",
        "tk = Tokenizer(lower = True, filters='')\n",
        "full_text = list(train['Phrase'].values) + list(test['Phrase'].values)\n",
        "tk.fit_on_texts(full_text)\n",
        "train_tokenized = tk.texts_to_sequences(train['Phrase'])\n",
        "test_tokenized = tk.texts_to_sequences(test['Phrase'])\n",
        "\n",
        "embedding_path = \"/content/drive/My Drive/DeepLearning/crawl-300d-2M.vec\"\n",
        "#embedding_path = \"/content/drive/My Drive/DeepLearning/glove.twitter.27B.25d.txt\"\n",
        "#embed_size = 25\n",
        "embed_size = 300\n",
        "max_features = 30000\n",
        "\n",
        "def get_coefs(word,*arr): return word, np.asarray(arr, dtype='float32')\n",
        "embedding_index = dict(get_coefs(*o.strip().split(\" \")) for o in open(embedding_path))\n",
        "\n",
        "word_index = tk.word_index\n",
        "nb_words = min(max_features, len(word_index))\n",
        "embedding_matrix = np.zeros((nb_words + 1, embed_size))\n",
        "for word, i in word_index.items():\n",
        "    if i >= max_features: continue\n",
        "    embedding_vector = embedding_index.get(word)\n",
        "    if embedding_vector is not None: embedding_matrix[i] = embedding_vector\n",
        "      \n",
        "      \n",
        "max_len = 50\n",
        "X_train = pad_sequences(train_tokenized, maxlen = max_len)\n",
        "X_test = pad_sequences(test_tokenized, maxlen = max_len)\n",
        "\n",
        "\n",
        "from sklearn.preprocessing import OneHotEncoder\n",
        "ohe = OneHotEncoder(sparse=False)\n",
        "y_ohe = ohe.fit_transform(y.values.reshape(-1, 1))\n",
        "ohe.fit(train[\"Sentiment\"].values.reshape(-1, 1))"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "error",
          "ename": "KeyboardInterrupt",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
            "\u001b[0;31mKeyboardInterrupt\u001b[0m                         Traceback (most recent call last)",
            "\u001b[0;32m<ipython-input-9-e32830b160c8>\u001b[0m in \u001b[0;36m<module>\u001b[0;34m()\u001b[0m\n\u001b[1;32m     20\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     21\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mget_coefs\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mword\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0marr\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;32mreturn\u001b[0m \u001b[0mword\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0masarray\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0marr\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdtype\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'float32'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 22\u001b[0;31m \u001b[0membedding_index\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdict\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mget_coefs\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0mo\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstrip\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msplit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\" \"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mo\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mopen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0membedding_path\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     23\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     24\u001b[0m \u001b[0mword_index\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtk\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mword_index\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m<ipython-input-9-e32830b160c8>\u001b[0m in \u001b[0;36m<genexpr>\u001b[0;34m(.0)\u001b[0m\n\u001b[1;32m     20\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     21\u001b[0m \u001b[0;32mdef\u001b[0m \u001b[0mget_coefs\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mword\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0marr\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;32mreturn\u001b[0m \u001b[0mword\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0masarray\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0marr\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdtype\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'float32'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 22\u001b[0;31m \u001b[0membedding_index\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdict\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mget_coefs\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0mo\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstrip\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msplit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\" \"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mo\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mopen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0membedding_path\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     23\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     24\u001b[0m \u001b[0mword_index\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mtk\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mword_index\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m<ipython-input-9-e32830b160c8>\u001b[0m in \u001b[0;36mget_coefs\u001b[0;34m(word, *arr)\u001b[0m\n\u001b[1;32m     19\u001b[0m \u001b[0mmax_features\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0;36m30000\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     20\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0;32m---> 21\u001b[0;31m \u001b[0;32mdef\u001b[0m \u001b[0mget_coefs\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mword\u001b[0m\u001b[0;34m,\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0marr\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m:\u001b[0m \u001b[0;32mreturn\u001b[0m \u001b[0mword\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mnp\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0masarray\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0marr\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdtype\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;34m'float32'\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m     22\u001b[0m \u001b[0membedding_index\u001b[0m \u001b[0;34m=\u001b[0m \u001b[0mdict\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0mget_coefs\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m*\u001b[0m\u001b[0mo\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0mstrip\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m.\u001b[0m\u001b[0msplit\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0;34m\" \"\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m \u001b[0;32mfor\u001b[0m \u001b[0mo\u001b[0m \u001b[0;32min\u001b[0m \u001b[0mopen\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0membedding_path\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m     23\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;32m/usr/local/lib/python3.6/dist-packages/numpy/core/numeric.py\u001b[0m in \u001b[0;36masarray\u001b[0;34m(a, dtype, order)\u001b[0m\n\u001b[1;32m    490\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    491\u001b[0m     \"\"\"\n\u001b[0;32m--> 492\u001b[0;31m     \u001b[0;32mreturn\u001b[0m \u001b[0marray\u001b[0m\u001b[0;34m(\u001b[0m\u001b[0ma\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mdtype\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0mcopy\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0;32mFalse\u001b[0m\u001b[0;34m,\u001b[0m \u001b[0morder\u001b[0m\u001b[0;34m=\u001b[0m\u001b[0morder\u001b[0m\u001b[0;34m)\u001b[0m\u001b[0;34m\u001b[0m\u001b[0;34m\u001b[0m\u001b[0m\n\u001b[0m\u001b[1;32m    493\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n\u001b[1;32m    494\u001b[0m \u001b[0;34m\u001b[0m\u001b[0m\n",
            "\u001b[0;31mKeyboardInterrupt\u001b[0m: "
          ]
        }
      ]
    },
    {
      "metadata": {
        "id": "--QxBTEFBA6I",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "def build_model5(lr=0.0, lr_d=0.0, units=0, spatial_dr=0.0, kernel_size1=3, kernel_size2=2, dense_units=128, dr=0.1, conv_size=32):\n",
        "    inp = Input(shape = (max_len,))\n",
        "    x = Embedding(19452, embed_size, weights = [embedding_matrix], trainable = False)(inp)\n",
        "    x1 = SpatialDropout1D(spatial_dr)(x)\n",
        "\n",
        "    x_gru = Bidirectional(CuDNNGRU(units, return_sequences = True))(x1)\n",
        "    x_lstm = Bidirectional(CuDNNLSTM(units, return_sequences = True))(x1)\n",
        "    \n",
        "    x_conv1 = Conv1D(conv_size, kernel_size=kernel_size1, padding='valid', kernel_initializer='he_uniform')(x_gru)\n",
        "    avg_pool1_gru = GlobalAveragePooling1D()(x_conv1)\n",
        "    max_pool1_gru = GlobalMaxPooling1D()(x_conv1)\n",
        "    \n",
        "    x_conv2 = Conv1D(conv_size, kernel_size=kernel_size2, padding='valid', kernel_initializer='he_uniform')(x_gru)\n",
        "    avg_pool2_gru = GlobalAveragePooling1D()(x_conv2)\n",
        "    max_pool2_gru = GlobalMaxPooling1D()(x_conv2)\n",
        "    \n",
        "    \n",
        "    x_conv3 = Conv1D(conv_size, kernel_size=kernel_size1, padding='valid', kernel_initializer='he_uniform')(x_lstm)\n",
        "    avg_pool1_lstm = GlobalAveragePooling1D()(x_conv3)\n",
        "    max_pool1_lstm = GlobalMaxPooling1D()(x_conv3)\n",
        "    \n",
        "    x_conv4 = Conv1D(conv_size, kernel_size=kernel_size2, padding='valid', kernel_initializer='he_uniform')(x_lstm)\n",
        "    avg_pool2_lstm = GlobalAveragePooling1D()(x_conv4)\n",
        "    max_pool2_lstm = GlobalMaxPooling1D()(x_conv4)\n",
        "    \n",
        "    \n",
        "    x = concatenate([avg_pool1_gru, max_pool1_gru, avg_pool2_gru, max_pool2_gru,\n",
        "                    avg_pool1_lstm, max_pool1_lstm, avg_pool2_lstm, max_pool2_lstm])\n",
        "    x = BatchNormalization()(x)\n",
        "    x = Dropout(dr)(Dense(dense_units, activation='relu') (x))\n",
        "    x = BatchNormalization()(x)\n",
        "    x = Dropout(dr)(Dense(int(dense_units / 2), activation='relu') (x))\n",
        "    x = Dense(5, activation = \"sigmoid\")(x)\n",
        "    model = Model(inputs = inp, outputs = x)\n",
        "    model.compile(loss = \"binary_crossentropy\", optimizer = Adam(lr = lr, decay = lr_d), metrics = [\"accuracy\"])\n",
        "    return model"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "sBnmScRvB-bV",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 384
        },
        "outputId": "61f97769-196d-4a31-cc39-940f706d5401"
      },
      "cell_type": "code",
      "source": [
        "print(\"Building the model...\")\n",
        "model5 = build_model5(lr = 1e-3, lr_d = 0, units = 64, spatial_dr = 0.5, kernel_size1=3, kernel_size2=4, dense_units=64, dr=0.3, conv_size=32)\n",
        "Train_And_Prediction(model5)\n",
        "#print(train)"
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Building the model...\n",
            "FOLD 1\n",
            "Splitting the data into train and validation...\n",
            "Training the model...\n",
            "Train on 58108 samples, validate on 58161 samples\n",
            "Epoch 1/1\n",
            "58108/58108 [==============================] - 43s 744us/step - loss: 0.4338 - acc: 0.7990 - val_loss: 0.3582 - val_acc: 0.8349\n",
            "66292/66292 [==============================] - 4s 58us/step\n",
            "\n",
            "FOLD 2\n",
            "Splitting the data into train and validation...\n",
            "Training the model...\n",
            "Train on 58161 samples, validate on 58108 samples\n",
            "Epoch 1/1\n",
            "58161/58161 [==============================] - 37s 629us/step - loss: 0.3720 - acc: 0.8274 - val_loss: 0.3388 - val_acc: 0.8406\n",
            "66292/66292 [==============================] - 3s 49us/step\n",
            "\n",
            "Save model after cross-validation...\n",
            "Make the submission ready...\n",
            "66292/66292 [==============================] - 3s 49us/step\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "metadata": {
        "id": "4noaMsufYjd1",
        "colab_type": "code",
        "colab": {}
      },
      "cell_type": "code",
      "source": [
        "from keras.layers import LSTM, Embedding, Dense, TimeDistributed, Bidirectional\n",
        "\n",
        "def build_model6(lr=0.0, lr_d=0.0, units=0, spatial_dr=0.0, kernel_size1=3, kernel_size2=2, dense_units=128, dr=0.1, conv_size=32):\n",
        "    input_words = Input((max_len, ))\n",
        "    x_words = Embedding(19453, embed_size,weights=[embedding_matrix],trainable=False)(input_words)\n",
        "    x_words = SpatialDropout1D(0.3)(x_words)\n",
        "    x_words = Bidirectional(CuDNNLSTM(50, return_sequences=True))(x_words)\n",
        "    x_words = LSTM(units = 64, return_sequences = True, recurrent_dropout = 0.2)(x_words)\n",
        "    x_words = Dropout(0.2)(x_words)\n",
        "    x_words = Conv1D(256, 3, strides = 1,  padding='causal', activation='relu', )(x_words)\n",
        "    x_words = GlobalMaxPool1D()(x_words)\n",
        "    x_words = Dropout(0.2)(x_words)\n",
        "\n",
        "    x = Dense(50, activation=\"relu\")(x_words)\n",
        "    x = Dropout(0.2)(x)\n",
        "    predictions = Dense(5, activation=\"softmax\")(x)\n",
        "\n",
        "    model = Model(inputs=[input_words], outputs=predictions)\n",
        "    model.compile(optimizer='nadam' ,loss='categorical_crossentropy', metrics=['accuracy'])\n",
        "\n",
        "    return model"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "metadata": {
        "id": "xze_K-cHb84-",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 121
        },
        "outputId": "6a147409-d739-4c9b-dc53-c951373c07ad"
      },
      "cell_type": "code",
      "source": [
        "print(\"Building the model...\")\n",
        "model6 = build_model6(lr = 1e-3, lr_d = 0, units = 64, spatial_dr = 0.5, kernel_size1=3, kernel_size2=4, dense_units=64, dr=0.3, conv_size=32)\n",
        "Train_And_Prediction(model6)\n",
        "#print(train)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Building the model...\n",
            "FOLD 1\n",
            "Splitting the data into train and validation...\n",
            "Training the model...\n",
            "Train on 58045 samples, validate on 58224 samples\n",
            "Epoch 1/1\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}