Emojify_[Emoji_Generation_From_Text].ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "Emojify [Emoji_Generation_From_Text].ipynb",
      "provenance": [],
      "collapsed_sections": []
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "metadata": {
        "id": "BJISfQ071PGn",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import pandas as pd\n",
        "import numpy as np\n",
        "import matplotlib.pyplot as plt"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qmxnw-uG3Btq",
        "colab_type": "code",
        "outputId": "515f7324-b0f4-4fef-a98f-429e3b932f49",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "import tensorflow as tf\n",
        "tf.__version__"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'2.2.0-rc3'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 2
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "h7_xmv853FmI",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from tensorflow.keras.preprocessing.text import Tokenizer\n",
        "from tensorflow.keras.preprocessing.sequence import pad_sequences\n",
        "import tensorflow.keras.utils as ku"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rMFjwJOwJB2Q",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import Dense\n",
        "from tensorflow.keras.layers import LSTM\n",
        "from tensorflow.keras.layers import Dropout\n",
        "from tensorflow.keras.layers import Bidirectional\n",
        "from tensorflow.keras.layers import Embedding"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JKVuf7xbZV_Y",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from sklearn.metrics import classification_report\n",
        "from sklearn.metrics import confusion_matrix"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OfezPacYl-OS",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import emoji"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "dee3uqPGI8Kf",
        "colab_type": "text"
      },
      "source": [
        "### Preparing Data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JHKZqVdO3bYl",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "df1=pd.read_csv('/content/drive/My Drive/Colab Notebooks/Emojify_Dataset/train_emoji.csv')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vzUlRA-S3oZg",
        "colab_type": "code",
        "outputId": "5a62a7cd-aab0-4585-e730-1305c0f05d4a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 197
        }
      },
      "source": [
        "df1.head()"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Text</th>\n",
              "      <th>Labels</th>\n",
              "      <th>Unnamed: 2</th>\n",
              "      <th>Unnamed: 3</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>I am proud of your achievements</td>\n",
              "      <td>2</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>It is the worst day in my life</td>\n",
              "      <td>3</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Miss you so much</td>\n",
              "      <td>0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[0]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>food is life</td>\n",
              "      <td>4</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>I love you mum</td>\n",
              "      <td>0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                              Text  Labels  Unnamed: 2 Unnamed: 3\n",
              "0  I am proud of your achievements       2         NaN        NaN\n",
              "1   It is the worst day in my life       3         NaN        NaN\n",
              "2                 Miss you so much       0         NaN        [0]\n",
              "3                     food is life       4         NaN        NaN\n",
              "4                   I love you mum       0         NaN        NaN"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 8
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_kCPd2ek553p",
        "colab_type": "code",
        "outputId": "3f7d1edc-bdec-4993-8004-20f02b62ee7c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "df1.shape"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(132, 4)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 9
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hq4tCLgqaHoR",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 122
        },
        "outputId": "69108f3a-7072-4a16-9e7c-be8a65aa5399"
      },
      "source": [
        "df1['Labels'].value_counts()"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "2    38\n",
              "3    36\n",
              "0    22\n",
              "1    19\n",
              "4    17\n",
              "Name: Labels, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 10
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cpFtpfPR7fTl",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "df2=pd.read_csv('/content/drive/My Drive/Colab Notebooks/Emojify_Dataset/emojify_data.csv')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6m2jROZA7m8S",
        "colab_type": "code",
        "outputId": "4b1e2d29-0877-40c0-bee1-7f72b2a15d8f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 197
        }
      },
      "source": [
        "df2.head()"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Text</th>\n",
              "      <th>Labels</th>\n",
              "      <th>Unnamed: 2</th>\n",
              "      <th>Unnamed: 3</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>work is horrible</td>\n",
              "      <td>3</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>I am upset</td>\n",
              "      <td>3</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[3]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>throw the ball</td>\n",
              "      <td>1</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[2]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>Good joke</td>\n",
              "      <td>2</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>what is your favorite baseball game</td>\n",
              "      <td>1</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                                  Text  Labels  Unnamed: 2 Unnamed: 3\n",
              "0                     work is horrible       3         NaN        NaN\n",
              "1                           I am upset       3         NaN        [3]\n",
              "2                       throw the ball       1         NaN        [2]\n",
              "3                            Good joke       2         NaN        NaN\n",
              "4  what is your favorite baseball game       1         NaN        NaN"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 12
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EWocKlecaBG6",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 122
        },
        "outputId": "7fedf2b4-9209-41b8-a3c5-fec690d5c146"
      },
      "source": [
        "df2['Labels'].value_counts()"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "2    56\n",
              "3    52\n",
              "0    33\n",
              "4    23\n",
              "1    19\n",
              "Name: Labels, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "B47aPwGJ77BM",
        "colab_type": "code",
        "outputId": "33be1b82-1094-4491-9e9e-e80cb0185d1f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "df2.shape"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(183, 4)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 14
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7vj99JJL7oY5",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "train_df=pd.concat([df1,df2],ignore_index=True)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "iTPIv8yC79SS",
        "colab_type": "code",
        "outputId": "bf975dee-a394-4604-bf28-ebfb5512517d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 407
        }
      },
      "source": [
        "train_df"
      ],
      "execution_count": 16,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Text</th>\n",
              "      <th>Labels</th>\n",
              "      <th>Unnamed: 2</th>\n",
              "      <th>Unnamed: 3</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>I am proud of your achievements</td>\n",
              "      <td>2</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>It is the worst day in my life</td>\n",
              "      <td>3</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>Miss you so much</td>\n",
              "      <td>0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>[0]</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>food is life</td>\n",
              "      <td>4</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>I love you mum</td>\n",
              "      <td>0</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>...</th>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "      <td>...</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>310</th>\n",
              "      <td>dance with me</td>\n",
              "      <td>2</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>311</th>\n",
              "      <td>she is a bully</td>\n",
              "      <td>3</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>312</th>\n",
              "      <td>she plays baseball</td>\n",
              "      <td>1</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>313</th>\n",
              "      <td>I like it when people smile</td>\n",
              "      <td>2</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>314</th>\n",
              "      <td>French macaroon is so tasty</td>\n",
              "      <td>4</td>\n",
              "      <td>NaN</td>\n",
              "      <td>NaN</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "<p>315 rows × 4 columns</p>\n",
              "</div>"
            ],
            "text/plain": [
              "                                Text  Labels  Unnamed: 2 Unnamed: 3\n",
              "0    I am proud of your achievements       2         NaN        NaN\n",
              "1     It is the worst day in my life       3         NaN        NaN\n",
              "2                   Miss you so much       0         NaN        [0]\n",
              "3                       food is life       4         NaN        NaN\n",
              "4                     I love you mum       0         NaN        NaN\n",
              "..                               ...     ...         ...        ...\n",
              "310                    dance with me       2         NaN        NaN\n",
              "311                   she is a bully       3         NaN        NaN\n",
              "312               she plays baseball       1         NaN        NaN\n",
              "313      I like it when people smile       2         NaN        NaN\n",
              "314      French macaroon is so tasty       4         NaN        NaN\n",
              "\n",
              "[315 rows x 4 columns]"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 16
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "AduHC6cR-8el",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "test_df=pd.read_csv('/content/drive/My Drive/Colab Notebooks/Emojify_Dataset/test_emoji.csv')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fsao0AcJ_ByU",
        "colab_type": "code",
        "outputId": "4032c0f9-7cf6-4e3b-82f3-7d396d54950c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 197
        }
      },
      "source": [
        "test_df.head()"
      ],
      "execution_count": 18,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/html": [
              "<div>\n",
              "<style scoped>\n",
              "    .dataframe tbody tr th:only-of-type {\n",
              "        vertical-align: middle;\n",
              "    }\n",
              "\n",
              "    .dataframe tbody tr th {\n",
              "        vertical-align: top;\n",
              "    }\n",
              "\n",
              "    .dataframe thead th {\n",
              "        text-align: right;\n",
              "    }\n",
              "</style>\n",
              "<table border=\"1\" class=\"dataframe\">\n",
              "  <thead>\n",
              "    <tr style=\"text-align: right;\">\n",
              "      <th></th>\n",
              "      <th>Text</th>\n",
              "      <th>Unnamed: 1</th>\n",
              "      <th>Labels</th>\n",
              "    </tr>\n",
              "  </thead>\n",
              "  <tbody>\n",
              "    <tr>\n",
              "      <th>0</th>\n",
              "      <td>he did not answer</td>\n",
              "      <td>NaN</td>\n",
              "      <td>3</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>1</th>\n",
              "      <td>he got a raise</td>\n",
              "      <td>NaN</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>2</th>\n",
              "      <td>she got me a present</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>3</th>\n",
              "      <td>ha ha ha it was so funny</td>\n",
              "      <td>NaN</td>\n",
              "      <td>2</td>\n",
              "    </tr>\n",
              "    <tr>\n",
              "      <th>4</th>\n",
              "      <td>he is a good friend</td>\n",
              "      <td>NaN</td>\n",
              "      <td>0</td>\n",
              "    </tr>\n",
              "  </tbody>\n",
              "</table>\n",
              "</div>"
            ],
            "text/plain": [
              "                       Text  Unnamed: 1  Labels\n",
              "0         he did not answer         NaN       3\n",
              "1            he got a raise         NaN       2\n",
              "2      she got me a present         NaN       0\n",
              "3  ha ha ha it was so funny         NaN       2\n",
              "4       he is a good friend         NaN       0"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 18
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "7KRBU3dhJYff",
        "colab_type": "text"
      },
      "source": [
        "### Preparing Training and Testing Data"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "koEurpIO-yu0",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "x_train=train_df['Text']"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "HZLyaYB2JqQ1",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "x_test=test_df['Text']"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Cgm00X6aJ0KX",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "y_train=train_df['Labels']"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aibTM06fNWMX",
        "colab_type": "code",
        "outputId": "a0b6e0e7-ac85-4ba4-b7c6-e2bfb350d2f1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 122
        }
      },
      "source": [
        "y_train.value_counts()"
      ],
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "2    94\n",
              "3    88\n",
              "0    55\n",
              "4    40\n",
              "1    38\n",
              "Name: Labels, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 22
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WNdYTHDiKCXr",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "y_test=test_df['Labels']"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sFHLtUvXYGDJ",
        "colab_type": "code",
        "outputId": "39e4fa5b-0b32-46e9-8fa0-5a3338929138",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 122
        }
      },
      "source": [
        "y_test.value_counts()"
      ],
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "2    18\n",
              "3    15\n",
              "0    12\n",
              "4     6\n",
              "1     5\n",
              "Name: Labels, dtype: int64"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "HpkRDpO9ezU4",
        "colab_type": "text"
      },
      "source": [
        "### Changing Target Variable to categorical"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GIVWamfudPh3",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "y_train_categorical=ku.to_categorical(y_train,num_classes=5)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EHlu4-VudPe0",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "y_test_categorical=ku.to_categorical(y_test,num_classes=5)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "EErACjb8KIvK",
        "colab_type": "text"
      },
      "source": [
        "### Tokenizing And Text Preprocessing"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "id3qJahFKHwn",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "tokenizer=Tokenizer()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "XIRVXcTnKWHp",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "tokenizer.fit_on_texts(x_train)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "H162roozKhst",
        "colab_type": "code",
        "outputId": "e203d92a-e5df-4009-d6f0-1c5ca80d557b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "word_index=tokenizer.word_index\n",
        "word_index"
      ],
      "execution_count": 29,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "{'a': 5,\n",
              " 'acceptance': 109,\n",
              " 'achievements': 104,\n",
              " 'adorable': 189,\n",
              " 'adore': 151,\n",
              " 'after': 133,\n",
              " 'afternoon': 236,\n",
              " 'again': 58,\n",
              " 'algorithm': 115,\n",
              " 'all': 279,\n",
              " 'alone': 165,\n",
              " 'always': 237,\n",
              " 'am': 6,\n",
              " 'amazing': 188,\n",
              " 'an': 187,\n",
              " 'and': 60,\n",
              " 'announcement': 184,\n",
              " 'answer': 86,\n",
              " 'any': 301,\n",
              " 'anything': 206,\n",
              " 'approved': 118,\n",
              " 'are': 10,\n",
              " 'around': 269,\n",
              " 'ask': 138,\n",
              " 'assignment': 110,\n",
              " 'at': 39,\n",
              " 'attractive': 140,\n",
              " 'away': 305,\n",
              " 'awesome': 257,\n",
              " 'awful': 168,\n",
              " 'baby': 95,\n",
              " 'back': 283,\n",
              " 'bad': 41,\n",
              " 'ball': 54,\n",
              " 'base': 229,\n",
              " 'baseball': 29,\n",
              " 'be': 289,\n",
              " 'best': 171,\n",
              " 'birthday': 308,\n",
              " 'boiled': 299,\n",
              " 'boring': 316,\n",
              " 'bravo': 81,\n",
              " 'break': 304,\n",
              " 'breakfast': 311,\n",
              " 'breaks': 296,\n",
              " 'brighten': 317,\n",
              " 'brunch': 135,\n",
              " 'bullshit': 108,\n",
              " 'bully': 318,\n",
              " 'but': 154,\n",
              " 'by': 79,\n",
              " 'cake': 233,\n",
              " 'can': 97,\n",
              " 'cancelled': 276,\n",
              " 'candy': 128,\n",
              " 'cannot': 197,\n",
              " 'catcher': 226,\n",
              " 'celebrate': 123,\n",
              " 'charming': 183,\n",
              " 'cheese': 232,\n",
              " 'chicago': 129,\n",
              " 'chinese': 181,\n",
              " 'code': 153,\n",
              " 'come': 198,\n",
              " 'coming': 126,\n",
              " 'congrats': 224,\n",
              " 'congratulation': 309,\n",
              " 'congratulations': 65,\n",
              " 'cooked': 169,\n",
              " 'cookies': 182,\n",
              " 'could': 222,\n",
              " 'cubs': 130,\n",
              " 'cute': 49,\n",
              " 'cutest': 238,\n",
              " 'dad': 219,\n",
              " 'dance': 258,\n",
              " 'dare': 137,\n",
              " 'date': 210,\n",
              " 'day': 44,\n",
              " 'dear': 208,\n",
              " 'dearest': 127,\n",
              " 'dedication': 200,\n",
              " 'deserve': 285,\n",
              " 'did': 23,\n",
              " 'dinner': 50,\n",
              " 'disappointed': 205,\n",
              " 'do': 25,\n",
              " 'dog': 177,\n",
              " 'dogs': 77,\n",
              " 'down': 52,\n",
              " 'during': 307,\n",
              " 'eat': 100,\n",
              " 'end': 163,\n",
              " 'enjoy': 303,\n",
              " 'ever': 240,\n",
              " 'exam': 134,\n",
              " 'excited': 72,\n",
              " 'exercise': 92,\n",
              " 'failed': 213,\n",
              " 'failing': 271,\n",
              " 'family': 278,\n",
              " 'favorite': 176,\n",
              " 'feeling': 274,\n",
              " 'few': 178,\n",
              " 'final': 263,\n",
              " 'finished': 122,\n",
              " 'first': 228,\n",
              " 'fon': 310,\n",
              " 'food': 26,\n",
              " 'for': 12,\n",
              " 'forever': 75,\n",
              " 'french': 147,\n",
              " 'friend': 293,\n",
              " 'friendly': 196,\n",
              " 'frustrated': 211,\n",
              " 'fun': 56,\n",
              " 'funny': 36,\n",
              " 'game': 47,\n",
              " 'gave': 155,\n",
              " 'get': 145,\n",
              " 'girl': 90,\n",
              " 'give': 215,\n",
              " 'go': 55,\n",
              " 'going': 172,\n",
              " 'good': 40,\n",
              " 'got': 24,\n",
              " 'grade': 160,\n",
              " 'grader': 91,\n",
              " 'grandmother': 297,\n",
              " 'great': 43,\n",
              " 'guy': 220,\n",
              " 'ha': 15,\n",
              " 'had': 61,\n",
              " 'haha': 158,\n",
              " 'handsome': 244,\n",
              " 'happy': 82,\n",
              " 'has': 113,\n",
              " 'hate': 180,\n",
              " 'have': 22,\n",
              " 'having': 294,\n",
              " 'he': 14,\n",
              " 'healthy': 243,\n",
              " 'her': 62,\n",
              " 'here': 246,\n",
              " 'him': 46,\n",
              " 'home': 250,\n",
              " 'homework': 231,\n",
              " 'horrible': 266,\n",
              " 'hours': 191,\n",
              " 'how': 87,\n",
              " 'hug': 216,\n",
              " 'humiliated': 166,\n",
              " 'hungry': 132,\n",
              " 'i': 1,\n",
              " 'impressed': 199,\n",
              " 'in': 83,\n",
              " 'incredibly': 173,\n",
              " 'indian': 248,\n",
              " 'intelligent': 174,\n",
              " 'is': 3,\n",
              " 'it': 20,\n",
              " 'jacket': 287,\n",
              " 'job': 35,\n",
              " 'join': 139,\n",
              " 'joke': 38,\n",
              " 'just': 71,\n",
              " 'killing': 157,\n",
              " 'kind': 195,\n",
              " 'laugh': 313,\n",
              " 'laughing': 143,\n",
              " 'lectures': 234,\n",
              " 'let': 270,\n",
              " 'lets': 48,\n",
              " 'life': 31,\n",
              " 'like': 27,\n",
              " 'likes': 70,\n",
              " 'limit': 114,\n",
              " 'lol': 57,\n",
              " 'long': 67,\n",
              " 'looking': 209,\n",
              " 'loser': 247,\n",
              " 'lost': 98,\n",
              " 'lot': 51,\n",
              " 'love': 18,\n",
              " 'lovely': 291,\n",
              " 'macaroon': 148,\n",
              " 'made': 150,\n",
              " 'make': 249,\n",
              " 'making': 314,\n",
              " 'man': 230,\n",
              " 'many': 259,\n",
              " 'me': 17,\n",
              " 'mean': 152,\n",
              " 'meat': 170,\n",
              " 'messing': 255,\n",
              " 'midterm': 214,\n",
              " 'miss': 32,\n",
              " 'missed': 161,\n",
              " 'moment': 202,\n",
              " 'much': 84,\n",
              " 'mum': 106,\n",
              " 'my': 8,\n",
              " 'near': 290,\n",
              " 'never': 103,\n",
              " 'new': 225,\n",
              " 'no': 68,\n",
              " 'not': 16,\n",
              " 'of': 30,\n",
              " 'on': 45,\n",
              " 'one': 121,\n",
              " 'ordering': 252,\n",
              " 'party': 275,\n",
              " 'people': 193,\n",
              " 'performs': 116,\n",
              " 'person': 239,\n",
              " 'pitch': 280,\n",
              " 'pizza': 159,\n",
              " 'plan': 120,\n",
              " 'play': 85,\n",
              " 'player': 93,\n",
              " 'playing': 262,\n",
              " 'plays': 186,\n",
              " 'points': 260,\n",
              " 'poorly': 117,\n",
              " 'position': 218,\n",
              " 'present': 288,\n",
              " 'prize': 286,\n",
              " 'project': 201,\n",
              " 'proud': 64,\n",
              " 'puppies': 179,\n",
              " 'qualified': 217,\n",
              " 'raise': 277,\n",
              " 'ready': 146,\n",
              " 'really': 281,\n",
              " 'restaurant': 102,\n",
              " 'rice': 300,\n",
              " 'rules': 264,\n",
              " 'run': 251,\n",
              " 'sad': 125,\n",
              " 'said': 73,\n",
              " 'saying': 107,\n",
              " 'score': 261,\n",
              " 'see': 88,\n",
              " 'seen': 241,\n",
              " 'serious': 312,\n",
              " 'she': 13,\n",
              " 'shouting': 69,\n",
              " 'sister': 167,\n",
              " 'smile': 63,\n",
              " 'smiles': 142,\n",
              " 'so': 7,\n",
              " 'solve': 223,\n",
              " 'some': 136,\n",
              " 'soon': 124,\n",
              " 'sounds': 119,\n",
              " 'specialization': 221,\n",
              " 'stadium': 94,\n",
              " 'stars': 282,\n",
              " 'starving': 298,\n",
              " 'stop': 33,\n",
              " 'stupid': 315,\n",
              " 'stupidity': 112,\n",
              " 'such': 99,\n",
              " 'suck': 141,\n",
              " 'sucks': 227,\n",
              " 'suggestions': 302,\n",
              " 'sushi': 203,\n",
              " 'takes': 144,\n",
              " 'taking': 295,\n",
              " 'talented': 175,\n",
              " 'talk': 254,\n",
              " 'tasty': 149,\n",
              " 'text': 111,\n",
              " 'that': 37,\n",
              " 'the': 4,\n",
              " 'they': 194,\n",
              " 'think': 162,\n",
              " 'this': 11,\n",
              " 'though': 235,\n",
              " 'throw': 268,\n",
              " 'to': 9,\n",
              " 'together': 207,\n",
              " 'tonight': 292,\n",
              " 'too': 66,\n",
              " 'totally': 284,\n",
              " 'traction': 185,\n",
              " 'two': 78,\n",
              " 'up': 164,\n",
              " 'upset': 267,\n",
              " 'us': 96,\n",
              " 'valentine': 256,\n",
              " 'vegetables': 242,\n",
              " 'very': 204,\n",
              " 'waiting': 190,\n",
              " 'wallet': 212,\n",
              " 'want': 21,\n",
              " 'was': 28,\n",
              " 'we': 76,\n",
              " 'well': 89,\n",
              " 'were': 245,\n",
              " 'what': 42,\n",
              " 'when': 192,\n",
              " 'where': 101,\n",
              " 'who': 80,\n",
              " 'why': 273,\n",
              " 'will': 34,\n",
              " 'with': 53,\n",
              " 'won': 131,\n",
              " 'work': 265,\n",
              " 'worked': 306,\n",
              " 'working': 59,\n",
              " 'worst': 105,\n",
              " 'wrong': 253,\n",
              " 'yes': 74,\n",
              " 'yesterday': 272,\n",
              " 'you': 2,\n",
              " 'your': 19,\n",
              " 'zero': 156}"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 29
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bEbdtJpRKp9Y",
        "colab_type": "code",
        "outputId": "b7f9f92e-2868-4871-ca3a-aab33cf4ea94",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "vocab_size=len(word_index)+1\n",
        "vocab_size"
      ],
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "319"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 30
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "axQjycFHKcUf",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "train_sequence=tokenizer.texts_to_sequences(x_train)\n",
        "train_padded=pad_sequences(train_sequence,maxlen=10)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "lDC8mo_ILGS_",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "test_sequence=tokenizer.texts_to_sequences(x_test)\n",
        "test_padded=pad_sequences(test_sequence,maxlen=10)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ga-zKeZHLNqw",
        "colab_type": "text"
      },
      "source": [
        "### Globe Embedding"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dIulF1WsLMiH",
        "colab_type": "code",
        "outputId": "521b2ae6-d7b1-421c-9635-d719a6b0b3a1",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 209
        }
      },
      "source": [
        "# This is the 100 dimension version of GloVe from Stanford\n",
        "# I am using a api for faster access\n",
        "!wget --no-check-certificate \\\n",
        "    https://storage.googleapis.com/laurencemoroney-blog.appspot.com/glove.6B.100d.txt \\\n",
        "    -O /tmp/glove.6B.100d.txt\n",
        "embeddings_index = {};\n",
        "with open('/tmp/glove.6B.100d.txt') as f:\n",
        "    for line in f:\n",
        "        values = line.split();\n",
        "        word = values[0];\n",
        "        coefs = np.asarray(values[1:], dtype='float32');\n",
        "        embeddings_index[word] = coefs;\n",
        "\n",
        "embeddings_matrix = np.zeros((vocab_size,100));\n",
        "for word, i in word_index.items():\n",
        "    embedding_vector = embeddings_index.get(word);\n",
        "    if embedding_vector is not None:\n",
        "        embeddings_matrix[i] = embedding_vector;"
      ],
      "execution_count": 33,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "--2020-04-30 19:28:36--  https://storage.googleapis.com/laurencemoroney-blog.appspot.com/glove.6B.100d.txt\n",
            "Resolving storage.googleapis.com (storage.googleapis.com)... 108.177.119.128, 2a00:1450:4013:c05::80\n",
            "Connecting to storage.googleapis.com (storage.googleapis.com)|108.177.119.128|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 347116733 (331M) [text/plain]\n",
            "Saving to: ‘/tmp/glove.6B.100d.txt’\n",
            "\n",
            "/tmp/glove.6B.100d. 100%[===================>] 331.04M  98.4MB/s    in 3.4s    \n",
            "\n",
            "2020-04-30 19:28:39 (98.4 MB/s) - ‘/tmp/glove.6B.100d.txt’ saved [347116733/347116733]\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "auDjlEiiMAJI",
        "colab_type": "text"
      },
      "source": [
        "### Building Model and Testing Accuracy\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "IL9BydyqMEsX",
        "colab_type": "text"
      },
      "source": [
        "#### Single Layer Bidirectional LSTM\n"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FPgrJnKLL8tX",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model1=Sequential()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "z-3FkEL_MP6q",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model1.add(Embedding(vocab_size,100,input_length=10,weights=[embeddings_matrix],trainable=False))\n",
        "model1.add(Dropout(0.3))\n",
        "model1.add(Bidirectional(LSTM(32)))\n",
        "model1.add(Dense(5,activation='softmax'))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0O6oE-iQNDzd",
        "colab_type": "code",
        "outputId": "69e8f3f2-28f9-41f6-a548-dcc10a72a93a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 297
        }
      },
      "source": [
        "model1.summary()"
      ],
      "execution_count": 36,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "embedding (Embedding)        (None, 10, 100)           31900     \n",
            "_________________________________________________________________\n",
            "dropout (Dropout)            (None, 10, 100)           0         \n",
            "_________________________________________________________________\n",
            "bidirectional (Bidirectional (None, 64)                34048     \n",
            "_________________________________________________________________\n",
            "dense (Dense)                (None, 5)                 325       \n",
            "=================================================================\n",
            "Total params: 66,273\n",
            "Trainable params: 34,373\n",
            "Non-trainable params: 31,900\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "DzMdwUaJNG0J",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model1.compile(optimizer='adam',loss='categorical_crossentropy',metrics=['accuracy'])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jpJBPKc0Nl1j",
        "colab_type": "code",
        "outputId": "82cf32a5-c281-429f-d267-9ed9ad45e0e6",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "history=model1.fit(train_padded,y_train_categorical,epochs=30,validation_data=(test_padded,y_test_categorical))"
      ],
      "execution_count": 38,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/30\n",
            "10/10 [==============================] - 1s 80ms/step - loss: 1.5601 - accuracy: 0.2508 - val_loss: 1.5069 - val_accuracy: 0.3036\n",
            "Epoch 2/30\n",
            "10/10 [==============================] - 0s 8ms/step - loss: 1.4476 - accuracy: 0.3651 - val_loss: 1.4229 - val_accuracy: 0.4286\n",
            "Epoch 3/30\n",
            "10/10 [==============================] - 0s 8ms/step - loss: 1.3752 - accuracy: 0.4444 - val_loss: 1.3524 - val_accuracy: 0.5000\n",
            "Epoch 4/30\n",
            "10/10 [==============================] - 0s 8ms/step - loss: 1.2860 - accuracy: 0.5206 - val_loss: 1.2763 - val_accuracy: 0.5357\n",
            "Epoch 5/30\n",
            "10/10 [==============================] - 0s 8ms/step - loss: 1.1871 - accuracy: 0.6222 - val_loss: 1.1829 - val_accuracy: 0.6250\n",
            "Epoch 6/30\n",
            "10/10 [==============================] - 0s 8ms/step - loss: 1.0817 - accuracy: 0.6508 - val_loss: 1.0812 - val_accuracy: 0.6071\n",
            "Epoch 7/30\n",
            "10/10 [==============================] - 0s 9ms/step - loss: 0.9449 - accuracy: 0.7206 - val_loss: 0.9885 - val_accuracy: 0.6250\n",
            "Epoch 8/30\n",
            "10/10 [==============================] - 0s 7ms/step - loss: 0.8382 - accuracy: 0.7365 - val_loss: 0.9204 - val_accuracy: 0.6786\n",
            "Epoch 9/30\n",
            "10/10 [==============================] - 0s 8ms/step - loss: 0.7318 - accuracy: 0.7841 - val_loss: 0.8403 - val_accuracy: 0.6429\n",
            "Epoch 10/30\n",
            "10/10 [==============================] - 0s 8ms/step - loss: 0.6483 - accuracy: 0.8063 - val_loss: 0.7633 - val_accuracy: 0.7143\n",
            "Epoch 11/30\n",
            "10/10 [==============================] - 0s 8ms/step - loss: 0.5762 - accuracy: 0.8381 - val_loss: 0.7267 - val_accuracy: 0.7321\n",
            "Epoch 12/30\n",
            "10/10 [==============================] - 0s 8ms/step - loss: 0.5133 - accuracy: 0.8190 - val_loss: 0.6442 - val_accuracy: 0.7500\n",
            "Epoch 13/30\n",
            "10/10 [==============================] - 0s 9ms/step - loss: 0.4917 - accuracy: 0.8254 - val_loss: 0.6420 - val_accuracy: 0.7500\n",
            "Epoch 14/30\n",
            "10/10 [==============================] - 0s 8ms/step - loss: 0.4379 - accuracy: 0.8571 - val_loss: 0.5567 - val_accuracy: 0.7679\n",
            "Epoch 15/30\n",
            "10/10 [==============================] - 0s 9ms/step - loss: 0.4508 - accuracy: 0.8540 - val_loss: 0.5550 - val_accuracy: 0.7679\n",
            "Epoch 16/30\n",
            "10/10 [==============================] - 0s 9ms/step - loss: 0.4115 - accuracy: 0.8508 - val_loss: 0.5582 - val_accuracy: 0.7679\n",
            "Epoch 17/30\n",
            "10/10 [==============================] - 0s 9ms/step - loss: 0.4061 - accuracy: 0.8825 - val_loss: 0.4725 - val_accuracy: 0.8750\n",
            "Epoch 18/30\n",
            "10/10 [==============================] - 0s 8ms/step - loss: 0.3552 - accuracy: 0.8825 - val_loss: 0.4500 - val_accuracy: 0.7857\n",
            "Epoch 19/30\n",
            "10/10 [==============================] - 0s 7ms/step - loss: 0.3234 - accuracy: 0.8889 - val_loss: 0.4626 - val_accuracy: 0.7857\n",
            "Epoch 20/30\n",
            "10/10 [==============================] - 0s 7ms/step - loss: 0.3191 - accuracy: 0.9048 - val_loss: 0.3673 - val_accuracy: 0.8750\n",
            "Epoch 21/30\n",
            "10/10 [==============================] - 0s 7ms/step - loss: 0.2806 - accuracy: 0.8921 - val_loss: 0.4502 - val_accuracy: 0.8036\n",
            "Epoch 22/30\n",
            "10/10 [==============================] - 0s 8ms/step - loss: 0.3039 - accuracy: 0.8984 - val_loss: 0.3654 - val_accuracy: 0.8393\n",
            "Epoch 23/30\n",
            "10/10 [==============================] - 0s 8ms/step - loss: 0.2516 - accuracy: 0.9175 - val_loss: 0.3229 - val_accuracy: 0.8750\n",
            "Epoch 24/30\n",
            "10/10 [==============================] - 0s 7ms/step - loss: 0.2381 - accuracy: 0.9333 - val_loss: 0.3643 - val_accuracy: 0.8393\n",
            "Epoch 25/30\n",
            "10/10 [==============================] - 0s 8ms/step - loss: 0.2638 - accuracy: 0.9048 - val_loss: 0.2808 - val_accuracy: 0.8929\n",
            "Epoch 26/30\n",
            "10/10 [==============================] - 0s 7ms/step - loss: 0.2465 - accuracy: 0.9143 - val_loss: 0.2674 - val_accuracy: 0.9107\n",
            "Epoch 27/30\n",
            "10/10 [==============================] - 0s 7ms/step - loss: 0.2262 - accuracy: 0.9143 - val_loss: 0.2652 - val_accuracy: 0.9107\n",
            "Epoch 28/30\n",
            "10/10 [==============================] - 0s 8ms/step - loss: 0.2123 - accuracy: 0.9238 - val_loss: 0.2501 - val_accuracy: 0.9286\n",
            "Epoch 29/30\n",
            "10/10 [==============================] - 0s 7ms/step - loss: 0.2030 - accuracy: 0.9302 - val_loss: 0.2444 - val_accuracy: 0.9107\n",
            "Epoch 30/30\n",
            "10/10 [==============================] - 0s 8ms/step - loss: 0.1788 - accuracy: 0.9460 - val_loss: 0.2266 - val_accuracy: 0.9464\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TUW1LuJkN6KY",
        "colab_type": "code",
        "outputId": "e2ca84a1-8c48-4e6a-a3c3-63b2c3885046",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 374
        }
      },
      "source": [
        "fig,(ax1,ax2)=plt.subplots(nrows=1,ncols=2,figsize=(20,6))\n",
        "ax1.plot(history.history['accuracy'],label='train_accuracy')\n",
        "ax1.plot(history.history['val_accuracy'],label='test_accuracy')\n",
        "ax1.legend()\n",
        "ax2.plot(history.history['loss'],label='train_loss')\n",
        "ax2.plot(history.history['val_loss'],label='test_loss')\n",
        "ax2.legend()\n",
        "plt.show()"
      ],
      "execution_count": 39,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1440x432 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ypwM9_8QOc9U",
        "colab_type": "text"
      },
      "source": [
        "### Multi layer Bidirectional LSTM"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SAGlDNpqOhk5",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model2=Sequential()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Gmf7MUpKPz_f",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model2.add(Embedding(vocab_size,100,input_length=10,weights=[embeddings_matrix],trainable=False))\n",
        "model2.add(Dropout(0.3))\n",
        "model2.add(Bidirectional(LSTM(128,return_sequences=True)))\n",
        "model2.add(Dropout(0.3))\n",
        "model2.add(Bidirectional(LSTM(128)))\n",
        "model2.add(Dense(5,activation='softmax'))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vb61wLB2P_pu",
        "colab_type": "code",
        "outputId": "0c27e8d9-40a1-4f1d-8b95-e2a2b0dfa4a2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 367
        }
      },
      "source": [
        "model2.summary()"
      ],
      "execution_count": 61,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_2\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "embedding_2 (Embedding)      (None, 10, 100)           31900     \n",
            "_________________________________________________________________\n",
            "dropout_3 (Dropout)          (None, 10, 100)           0         \n",
            "_________________________________________________________________\n",
            "bidirectional_3 (Bidirection (None, 10, 256)           234496    \n",
            "_________________________________________________________________\n",
            "dropout_4 (Dropout)          (None, 10, 256)           0         \n",
            "_________________________________________________________________\n",
            "bidirectional_4 (Bidirection (None, 256)               394240    \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 5)                 1285      \n",
            "=================================================================\n",
            "Total params: 661,921\n",
            "Trainable params: 630,021\n",
            "Non-trainable params: 31,900\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gb003YEGQFCR",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model2.compile(optimizer='adam',loss='categorical_crossentropy',metrics=['accuracy'])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QJuUItuyQIIy",
        "colab_type": "code",
        "outputId": "9522e771-ffd7-4eea-eb6a-02786e57838a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "history=model2.fit(train_padded,y_train_categorical,epochs=50,validation_data=(test_padded,y_test_categorical))"
      ],
      "execution_count": 63,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/50\n",
            "10/10 [==============================] - 1s 120ms/step - loss: 1.5362 - accuracy: 0.2444 - val_loss: 1.4019 - val_accuracy: 0.3571\n",
            "Epoch 2/50\n",
            "10/10 [==============================] - 0s 13ms/step - loss: 1.3006 - accuracy: 0.5143 - val_loss: 1.1530 - val_accuracy: 0.5893\n",
            "Epoch 3/50\n",
            "10/10 [==============================] - 0s 12ms/step - loss: 0.9309 - accuracy: 0.6730 - val_loss: 0.9855 - val_accuracy: 0.5357\n",
            "Epoch 4/50\n",
            "10/10 [==============================] - 0s 31ms/step - loss: 0.7069 - accuracy: 0.7397 - val_loss: 0.6746 - val_accuracy: 0.6964\n",
            "Epoch 5/50\n",
            "10/10 [==============================] - 0s 12ms/step - loss: 0.4872 - accuracy: 0.8190 - val_loss: 0.5136 - val_accuracy: 0.8036\n",
            "Epoch 6/50\n",
            "10/10 [==============================] - 0s 12ms/step - loss: 0.4177 - accuracy: 0.8444 - val_loss: 0.5344 - val_accuracy: 0.7679\n",
            "Epoch 7/50\n",
            "10/10 [==============================] - 0s 12ms/step - loss: 0.5374 - accuracy: 0.8000 - val_loss: 0.6254 - val_accuracy: 0.7500\n",
            "Epoch 8/50\n",
            "10/10 [==============================] - 0s 11ms/step - loss: 0.3868 - accuracy: 0.8444 - val_loss: 0.4428 - val_accuracy: 0.8393\n",
            "Epoch 9/50\n",
            "10/10 [==============================] - 0s 11ms/step - loss: 0.3501 - accuracy: 0.8730 - val_loss: 0.4496 - val_accuracy: 0.7857\n",
            "Epoch 10/50\n",
            "10/10 [==============================] - 0s 10ms/step - loss: 0.3113 - accuracy: 0.8857 - val_loss: 0.4873 - val_accuracy: 0.7679\n",
            "Epoch 11/50\n",
            "10/10 [==============================] - 0s 11ms/step - loss: 0.3679 - accuracy: 0.8635 - val_loss: 0.3953 - val_accuracy: 0.8571\n",
            "Epoch 12/50\n",
            "10/10 [==============================] - 0s 11ms/step - loss: 0.2874 - accuracy: 0.8825 - val_loss: 0.3197 - val_accuracy: 0.8929\n",
            "Epoch 13/50\n",
            "10/10 [==============================] - 0s 11ms/step - loss: 0.2895 - accuracy: 0.9048 - val_loss: 0.3875 - val_accuracy: 0.8393\n",
            "Epoch 14/50\n",
            "10/10 [==============================] - 0s 10ms/step - loss: 0.3102 - accuracy: 0.8889 - val_loss: 0.2740 - val_accuracy: 0.9286\n",
            "Epoch 15/50\n",
            "10/10 [==============================] - 0s 11ms/step - loss: 0.2840 - accuracy: 0.9143 - val_loss: 0.2462 - val_accuracy: 0.9107\n",
            "Epoch 16/50\n",
            "10/10 [==============================] - 0s 10ms/step - loss: 0.1711 - accuracy: 0.9460 - val_loss: 0.2699 - val_accuracy: 0.9286\n",
            "Epoch 17/50\n",
            "10/10 [==============================] - 0s 11ms/step - loss: 0.1805 - accuracy: 0.9429 - val_loss: 0.2766 - val_accuracy: 0.9107\n",
            "Epoch 18/50\n",
            "10/10 [==============================] - 0s 11ms/step - loss: 0.1408 - accuracy: 0.9556 - val_loss: 0.2385 - val_accuracy: 0.9286\n",
            "Epoch 19/50\n",
            "10/10 [==============================] - 0s 12ms/step - loss: 0.1483 - accuracy: 0.9524 - val_loss: 0.2048 - val_accuracy: 0.9286\n",
            "Epoch 20/50\n",
            "10/10 [==============================] - 0s 12ms/step - loss: 0.1887 - accuracy: 0.9333 - val_loss: 0.2011 - val_accuracy: 0.9464\n",
            "Epoch 21/50\n",
            "10/10 [==============================] - 0s 14ms/step - loss: 0.1468 - accuracy: 0.9524 - val_loss: 0.2312 - val_accuracy: 0.9286\n",
            "Epoch 22/50\n",
            "10/10 [==============================] - 0s 13ms/step - loss: 0.1460 - accuracy: 0.9524 - val_loss: 0.2227 - val_accuracy: 0.9286\n",
            "Epoch 23/50\n",
            "10/10 [==============================] - 0s 12ms/step - loss: 0.1644 - accuracy: 0.9460 - val_loss: 0.1814 - val_accuracy: 0.9464\n",
            "Epoch 24/50\n",
            "10/10 [==============================] - 0s 13ms/step - loss: 0.1630 - accuracy: 0.9397 - val_loss: 0.2270 - val_accuracy: 0.9286\n",
            "Epoch 25/50\n",
            "10/10 [==============================] - 0s 12ms/step - loss: 0.1853 - accuracy: 0.9238 - val_loss: 0.1786 - val_accuracy: 0.9464\n",
            "Epoch 26/50\n",
            "10/10 [==============================] - 0s 12ms/step - loss: 0.1178 - accuracy: 0.9587 - val_loss: 0.1615 - val_accuracy: 0.9643\n",
            "Epoch 27/50\n",
            "10/10 [==============================] - 0s 11ms/step - loss: 0.0887 - accuracy: 0.9619 - val_loss: 0.1672 - val_accuracy: 0.9643\n",
            "Epoch 28/50\n",
            "10/10 [==============================] - 0s 12ms/step - loss: 0.0831 - accuracy: 0.9778 - val_loss: 0.1620 - val_accuracy: 0.9643\n",
            "Epoch 29/50\n",
            "10/10 [==============================] - 0s 12ms/step - loss: 0.1246 - accuracy: 0.9492 - val_loss: 0.2921 - val_accuracy: 0.8571\n",
            "Epoch 30/50\n",
            "10/10 [==============================] - 0s 12ms/step - loss: 0.1528 - accuracy: 0.9492 - val_loss: 0.3166 - val_accuracy: 0.9286\n",
            "Epoch 31/50\n",
            "10/10 [==============================] - 0s 12ms/step - loss: 0.0602 - accuracy: 0.9841 - val_loss: 0.2015 - val_accuracy: 0.9464\n",
            "Epoch 32/50\n",
            "10/10 [==============================] - 0s 11ms/step - loss: 0.1173 - accuracy: 0.9651 - val_loss: 0.2393 - val_accuracy: 0.9464\n",
            "Epoch 33/50\n",
            "10/10 [==============================] - 0s 11ms/step - loss: 0.0727 - accuracy: 0.9746 - val_loss: 0.1889 - val_accuracy: 0.9643\n",
            "Epoch 34/50\n",
            "10/10 [==============================] - 0s 11ms/step - loss: 0.0895 - accuracy: 0.9714 - val_loss: 0.1756 - val_accuracy: 0.9643\n",
            "Epoch 35/50\n",
            "10/10 [==============================] - 0s 11ms/step - loss: 0.0772 - accuracy: 0.9778 - val_loss: 0.1801 - val_accuracy: 0.9643\n",
            "Epoch 36/50\n",
            "10/10 [==============================] - 0s 11ms/step - loss: 0.0684 - accuracy: 0.9810 - val_loss: 0.1886 - val_accuracy: 0.9643\n",
            "Epoch 37/50\n",
            "10/10 [==============================] - 0s 11ms/step - loss: 0.0492 - accuracy: 0.9873 - val_loss: 0.1691 - val_accuracy: 0.9643\n",
            "Epoch 38/50\n",
            "10/10 [==============================] - 0s 12ms/step - loss: 0.0624 - accuracy: 0.9810 - val_loss: 0.1936 - val_accuracy: 0.9643\n",
            "Epoch 39/50\n",
            "10/10 [==============================] - 0s 11ms/step - loss: 0.0612 - accuracy: 0.9746 - val_loss: 0.2004 - val_accuracy: 0.9643\n",
            "Epoch 40/50\n",
            "10/10 [==============================] - 0s 13ms/step - loss: 0.0479 - accuracy: 0.9810 - val_loss: 0.2083 - val_accuracy: 0.9643\n",
            "Epoch 41/50\n",
            "10/10 [==============================] - 0s 13ms/step - loss: 0.0330 - accuracy: 0.9873 - val_loss: 0.1889 - val_accuracy: 0.9643\n",
            "Epoch 42/50\n",
            "10/10 [==============================] - 0s 11ms/step - loss: 0.0373 - accuracy: 0.9873 - val_loss: 0.1725 - val_accuracy: 0.9643\n",
            "Epoch 43/50\n",
            "10/10 [==============================] - 0s 12ms/step - loss: 0.0620 - accuracy: 0.9714 - val_loss: 0.2012 - val_accuracy: 0.9643\n",
            "Epoch 44/50\n",
            "10/10 [==============================] - 0s 12ms/step - loss: 0.0528 - accuracy: 0.9778 - val_loss: 0.1647 - val_accuracy: 0.9643\n",
            "Epoch 45/50\n",
            "10/10 [==============================] - 0s 13ms/step - loss: 0.0789 - accuracy: 0.9778 - val_loss: 0.1585 - val_accuracy: 0.9643\n",
            "Epoch 46/50\n",
            "10/10 [==============================] - 0s 12ms/step - loss: 0.0568 - accuracy: 0.9873 - val_loss: 0.1634 - val_accuracy: 0.9643\n",
            "Epoch 47/50\n",
            "10/10 [==============================] - 0s 13ms/step - loss: 0.0807 - accuracy: 0.9651 - val_loss: 0.1820 - val_accuracy: 0.9643\n",
            "Epoch 48/50\n",
            "10/10 [==============================] - 0s 13ms/step - loss: 0.0578 - accuracy: 0.9810 - val_loss: 0.2095 - val_accuracy: 0.9643\n",
            "Epoch 49/50\n",
            "10/10 [==============================] - 0s 12ms/step - loss: 0.0437 - accuracy: 0.9841 - val_loss: 0.2093 - val_accuracy: 0.9643\n",
            "Epoch 50/50\n",
            "10/10 [==============================] - 0s 13ms/step - loss: 0.0713 - accuracy: 0.9778 - val_loss: 0.1897 - val_accuracy: 0.9643\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4mcidir0QNCw",
        "colab_type": "code",
        "outputId": "78b66310-1926-430b-a5b9-707a6b0d38a4",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 376
        }
      },
      "source": [
        "fig,(ax1,ax2)=plt.subplots(nrows=1,ncols=2,figsize=(20,6))\n",
        "ax1.plot(history.history['accuracy'],label='train_accuracy')\n",
        "ax1.plot(history.history['val_accuracy'],label='test_accuracy')\n",
        "ax1.legend()\n",
        "ax2.plot(history.history['loss'],label='train_loss')\n",
        "ax2.plot(history.history['val_loss'],label='test_loss')\n",
        "ax2.legend()\n",
        "plt.show()"
      ],
      "execution_count": 64,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 1440x432 with 2 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "YoavwPStQnKG",
        "colab_type": "text"
      },
      "source": [
        "### Multilayer Bidirectional LSTM is a better model and gives a good accuracy of 96%"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "coKppUu-ZAzD",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "prediction=model2.predict_classes(test_padded)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "02A9pWPQZUvp",
        "colab_type": "code",
        "outputId": "0520de39-7bd2-4438-ac1b-00bdabd0dd6c",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 227
        }
      },
      "source": [
        "print(classification_report(y_test,prediction))"
      ],
      "execution_count": 66,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       1.00      0.92      0.96        12\n",
            "           1       1.00      1.00      1.00         5\n",
            "           2       0.95      1.00      0.97        18\n",
            "           3       0.93      0.93      0.93        15\n",
            "           4       1.00      1.00      1.00         6\n",
            "\n",
            "    accuracy                           0.96        56\n",
            "   macro avg       0.98      0.97      0.97        56\n",
            "weighted avg       0.97      0.96      0.96        56\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kU_2wfcuZmqv",
        "colab_type": "code",
        "outputId": "cc73b6bf-f31d-492d-f8aa-92d349084a34",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 104
        }
      },
      "source": [
        "confusion_matrix(y_test,prediction)"
      ],
      "execution_count": 67,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([[11,  0,  0,  1,  0],\n",
              "       [ 0,  5,  0,  0,  0],\n",
              "       [ 0,  0, 18,  0,  0],\n",
              "       [ 0,  0,  1, 14,  0],\n",
              "       [ 0,  0,  0,  0,  6]])"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 67
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "eweQs7uxQ0yS",
        "colab_type": "text"
      },
      "source": [
        "### Predicting With some new values"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "r4tDVbxfWaMC",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "emoji_dict={0:['heart',':heart:'],1:['baseball',':baseball:'],2:['smile',':smile:'],3:['disappoined',':disappointed:'],4:['fork and knife',':fork_and_knife:']}"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "X7C4FwqdQlm0",
        "colab_type": "code",
        "outputId": "b503764d-3a09-4a3c-de3f-cb4a173a7d2f",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "text=x_test[8]\n",
        "text"
      ],
      "execution_count": 69,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'Stop making this joke ha ha ha'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 69
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "GaTjheOwYk4S",
        "colab_type": "code",
        "outputId": "02fbabd8-6c0d-4231-c567-1bfc633127c6",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "y_test[8]"
      ],
      "execution_count": 70,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "2"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 70
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QWWZeJy1bTDn",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def predict(text):\n",
        "  text_sequence=tokenizer.texts_to_sequences([text])\n",
        "  text_padded=pad_sequences(text_sequence,maxlen=30)\n",
        "  prediction=model2.predict_classes(text_padded)\n",
        "  x=prediction[0]\n",
        "  print(emoji_dict[x][0])\n",
        "  print(emoji.emojize(emoji_dict[x][1], use_aliases=True))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Vyvtt8rKbpda",
        "colab_type": "code",
        "outputId": "9c17af44-93f7-49e6-9cb4-ee383111f248",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 89
        }
      },
      "source": [
        "ans=predict(text)\n",
        "ans"
      ],
      "execution_count": 72,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:Model was constructed with shape (None, 10) for input Tensor(\"embedding_2_input:0\", shape=(None, 10), dtype=float32), but it was called on an input with incompatible shape (None, 30).\n",
            "smile\n",
            "😄\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JW6RveKlnS81",
        "colab_type": "text"
      },
      "source": [
        "### Predicting New Values"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0n37Sm6MgS0C",
        "colab_type": "code",
        "outputId": "57b6ac6f-126b-4d12-d06d-d876b42cc156",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        }
      },
      "source": [
        "msg='you brighten my day'\n",
        "ans=predict(msg)\n",
        "ans"
      ],
      "execution_count": 73,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "smile\n",
            "😄\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ySBOYvAnmb_i",
        "colab_type": "code",
        "outputId": "2761bc36-dc92-4113-8fd5-fb74b9b9eaf3",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        }
      },
      "source": [
        "msg='she got me a nice present'\n",
        "ans=predict(msg)\n",
        "ans"
      ],
      "execution_count": 74,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "heart\n",
            "❤\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "qeWqcpJ5mxd_",
        "colab_type": "code",
        "outputId": "167966e0-18f2-4402-e662-c84bdc4b3e17",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        }
      },
      "source": [
        "msg='This girl is messing with me'\n",
        "ans=predict(msg)\n",
        "ans"
      ],
      "execution_count": 75,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "disappoined\n",
            "😞\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Jrlbeeo5nJk1",
        "colab_type": "code",
        "outputId": "8cfd7442-abd3-4341-8e63-3cfe12cd7592",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        }
      },
      "source": [
        "msg='will you be my valentine'\n",
        "ans=predict(msg)\n",
        "ans"
      ],
      "execution_count": 76,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "heart\n",
            "❤\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5_AHBIgSZdPm",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        },
        "outputId": "aa858d58-bc4d-4b95-de5a-5ff8c80dc5f5"
      },
      "source": [
        "msg='Honey lets go out for a date'\n",
        "ans=predict(msg)\n",
        "ans"
      ],
      "execution_count": 77,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "fork and knife\n",
            "🍴\n"
          ],
          "name": "stdout"
        }
      ]
    }
  ]
}