TensorFlow5_Text_Generation.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "TensorFlow5_Text-Generation.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "authorship_tag": "ABX9TyPtLmcY3RGQZ0d0fbglXhBj",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/utkarsh0702/Tensorflow/blob/master/TensorFlow5_Text_Generation.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Wr6M7l3fw7fG",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import tensorflow as tf\n",
        "from tensorflow.keras.preprocessing.text import Tokenizer\n",
        "from tensorflow.keras.preprocessing.sequence import pad_sequences"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pr6RDxDFm4p4",
        "colab_type": "code",
        "outputId": "e6db66fb-5f82-4bcb-a090-5fa91b74ba86",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 204
        }
      },
      "source": [
        "!wget --no-check-certificate https://storage.googleapis.com/laurencemoroney-blog.appspot.com/irish-lyrics-eof.txt -O /content/text.txt"
      ],
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "--2020-04-09 12:35:55--  https://storage.googleapis.com/laurencemoroney-blog.appspot.com/irish-lyrics-eof.txt\n",
            "Resolving storage.googleapis.com (storage.googleapis.com)... 74.125.130.128, 2404:6800:4003:c02::80\n",
            "Connecting to storage.googleapis.com (storage.googleapis.com)|74.125.130.128|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 68970 (67K) [text/plain]\n",
            "Saving to: ‘/content/text.txt’\n",
            "\n",
            "\r/content/text.txt     0%[                    ]       0  --.-KB/s               \r/content/text.txt   100%[===================>]  67.35K  --.-KB/s    in 0s      \n",
            "\n",
            "2020-04-09 12:35:55 (163 MB/s) - ‘/content/text.txt’ saved [68970/68970]\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6ajHraJbz3-T",
        "colab_type": "code",
        "outputId": "f4e79a84-33db-4ee8-c208-b54913872289",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "data= open('/content/text.txt').readlines()\n",
        "print(data[12])"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "And stole my bunch of thyme away\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "eVhPmlcL1eoB",
        "colab_type": "code",
        "outputId": "ccaa451b-6fac-4a0b-881c-a7c2a34ba414",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "token= Tokenizer()\n",
        "token.fit_on_texts(data)\n",
        "total= len(token.word_index)+1\n",
        "print(total)"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "2690\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "psQEB6px2bRp",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import numpy as np\n",
        "\n",
        "input_sentence=[]\n",
        "for line in data:\n",
        "  token_list= token.texts_to_sequences([line])[0]\n",
        "  for i in range(1, len(token_list)):\n",
        "    n_gram_sequence= token_list[:i+1]\n",
        "    input_sentence.append(n_gram_sequence)\n",
        "\n",
        "max_sequence_length= max([len(x) for x in input_sentence])\n",
        "\n",
        "input_sequence= np.array(pad_sequences( input_sentence, maxlen= max_sequence_length, padding='pre'))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "4PsBPHoo4AOI",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "xs= input_sequence[:,:-1]\n",
        "labels= input_sequence[:,-1]\n",
        "\n",
        "#one hot encoding\n",
        "ys= tf.keras.utils.to_categorical(labels, num_classes=total)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uHoayacYIV6A",
        "colab_type": "code",
        "outputId": "a8621280-9632-4c6c-c88e-bb18dbd0a6c2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 289
        }
      },
      "source": [
        "from tensorflow.keras.optimizers import Adam\n",
        "#model building\n",
        "model= tf.keras.Sequential([\n",
        "                            tf.keras.layers.Embedding(total, 16, input_length=max_sequence_length-1),\n",
        "                            tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(32, return_sequences=True)),\n",
        "                            tf.keras.layers.Bidirectional(tf.keras.layers.LSTM(16)),\n",
        "                            tf.keras.layers.Dense(total, activation='softmax')\n",
        "])\n",
        "adam= Adam(lr=0.01)\n",
        "model.compile(loss='categorical_crossentropy', optimizer= adam, metrics=['accuracy'])\n",
        "model.summary()"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "embedding (Embedding)        (None, 15, 16)            43040     \n",
            "_________________________________________________________________\n",
            "bidirectional (Bidirectional (None, 15, 64)            12544     \n",
            "_________________________________________________________________\n",
            "bidirectional_1 (Bidirection (None, 32)                10368     \n",
            "_________________________________________________________________\n",
            "dense (Dense)                (None, 2690)              88770     \n",
            "=================================================================\n",
            "Total params: 154,722\n",
            "Trainable params: 154,722\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3eFCnEoqJrDp",
        "colab_type": "code",
        "outputId": "d9b95508-9f6c-4a75-d827-185c2eb8c72d",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "history= model.fit(xs,ys, epochs=100, verbose=1)"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 6.7485 - accuracy: 0.0640\n",
            "Epoch 2/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 6.2268 - accuracy: 0.0750\n",
            "Epoch 3/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 5.9759 - accuracy: 0.0953\n",
            "Epoch 4/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 5.7531 - accuracy: 0.1048\n",
            "Epoch 5/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 5.5613 - accuracy: 0.1121\n",
            "Epoch 6/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 5.3905 - accuracy: 0.1238\n",
            "Epoch 7/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 5.2364 - accuracy: 0.1337\n",
            "Epoch 8/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 5.1056 - accuracy: 0.1400\n",
            "Epoch 9/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 4.9731 - accuracy: 0.1482\n",
            "Epoch 10/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 4.8435 - accuracy: 0.1565\n",
            "Epoch 11/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 4.7163 - accuracy: 0.1681\n",
            "Epoch 12/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 4.6024 - accuracy: 0.1746\n",
            "Epoch 13/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 4.4923 - accuracy: 0.1811\n",
            "Epoch 14/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 4.3871 - accuracy: 0.1903\n",
            "Epoch 15/100\n",
            "377/377 [==============================] - 4s 11ms/step - loss: 4.2875 - accuracy: 0.1968\n",
            "Epoch 16/100\n",
            "377/377 [==============================] - 4s 11ms/step - loss: 4.1765 - accuracy: 0.2065\n",
            "Epoch 17/100\n",
            "377/377 [==============================] - 4s 11ms/step - loss: 4.0821 - accuracy: 0.2131\n",
            "Epoch 18/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 3.9692 - accuracy: 0.2215\n",
            "Epoch 19/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 3.8610 - accuracy: 0.2297\n",
            "Epoch 20/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 3.7639 - accuracy: 0.2402\n",
            "Epoch 21/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 3.6687 - accuracy: 0.2533\n",
            "Epoch 22/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 3.5735 - accuracy: 0.2608\n",
            "Epoch 23/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 3.4935 - accuracy: 0.2732\n",
            "Epoch 24/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 3.4136 - accuracy: 0.2837\n",
            "Epoch 25/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 3.3322 - accuracy: 0.2954\n",
            "Epoch 26/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 3.2515 - accuracy: 0.3068\n",
            "Epoch 27/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 3.1591 - accuracy: 0.3247\n",
            "Epoch 28/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 3.0735 - accuracy: 0.3318\n",
            "Epoch 29/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 3.0380 - accuracy: 0.3403\n",
            "Epoch 30/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 2.9611 - accuracy: 0.3498\n",
            "Epoch 31/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 2.8867 - accuracy: 0.3652\n",
            "Epoch 32/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 2.8147 - accuracy: 0.3775\n",
            "Epoch 33/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 2.7824 - accuracy: 0.3835\n",
            "Epoch 34/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 2.6966 - accuracy: 0.3942\n",
            "Epoch 35/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 2.6225 - accuracy: 0.4109\n",
            "Epoch 36/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 2.6112 - accuracy: 0.4155\n",
            "Epoch 37/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 2.5739 - accuracy: 0.4185\n",
            "Epoch 38/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 2.5208 - accuracy: 0.4282\n",
            "Epoch 39/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 2.4462 - accuracy: 0.4454\n",
            "Epoch 40/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 2.3736 - accuracy: 0.4574\n",
            "Epoch 41/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 2.3538 - accuracy: 0.4619\n",
            "Epoch 42/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 2.3088 - accuracy: 0.4728\n",
            "Epoch 43/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 2.2898 - accuracy: 0.4689\n",
            "Epoch 44/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 2.2608 - accuracy: 0.4760\n",
            "Epoch 45/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 2.2233 - accuracy: 0.4836\n",
            "Epoch 46/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 2.1962 - accuracy: 0.4912\n",
            "Epoch 47/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 2.1445 - accuracy: 0.5032\n",
            "Epoch 48/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 2.1001 - accuracy: 0.5082\n",
            "Epoch 49/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 2.1219 - accuracy: 0.5051\n",
            "Epoch 50/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 2.1061 - accuracy: 0.5098\n",
            "Epoch 51/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 2.0620 - accuracy: 0.5156\n",
            "Epoch 52/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 2.0573 - accuracy: 0.5140\n",
            "Epoch 53/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 2.0227 - accuracy: 0.5180\n",
            "Epoch 54/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.9772 - accuracy: 0.5352\n",
            "Epoch 55/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.9473 - accuracy: 0.5424\n",
            "Epoch 56/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.9213 - accuracy: 0.5425\n",
            "Epoch 57/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.9289 - accuracy: 0.5463\n",
            "Epoch 58/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.9494 - accuracy: 0.5358\n",
            "Epoch 59/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.9281 - accuracy: 0.5391\n",
            "Epoch 60/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.8844 - accuracy: 0.5489\n",
            "Epoch 61/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.8609 - accuracy: 0.5527\n",
            "Epoch 62/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.8644 - accuracy: 0.5564\n",
            "Epoch 63/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.8388 - accuracy: 0.5575\n",
            "Epoch 64/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.8183 - accuracy: 0.5618\n",
            "Epoch 65/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.8000 - accuracy: 0.5678\n",
            "Epoch 66/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.8195 - accuracy: 0.5610\n",
            "Epoch 67/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.7394 - accuracy: 0.5762\n",
            "Epoch 68/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.7330 - accuracy: 0.5829\n",
            "Epoch 69/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.7712 - accuracy: 0.5721\n",
            "Epoch 70/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.7751 - accuracy: 0.5744\n",
            "Epoch 71/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.7707 - accuracy: 0.5655\n",
            "Epoch 72/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.7223 - accuracy: 0.5856\n",
            "Epoch 73/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.7207 - accuracy: 0.5817\n",
            "Epoch 74/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.6967 - accuracy: 0.5826\n",
            "Epoch 75/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.6447 - accuracy: 0.6019\n",
            "Epoch 76/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.6830 - accuracy: 0.5924\n",
            "Epoch 77/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.6730 - accuracy: 0.5891\n",
            "Epoch 78/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.7343 - accuracy: 0.5784\n",
            "Epoch 79/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.7164 - accuracy: 0.5848\n",
            "Epoch 80/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.6728 - accuracy: 0.5935\n",
            "Epoch 81/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.6759 - accuracy: 0.5859\n",
            "Epoch 82/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.6716 - accuracy: 0.5900\n",
            "Epoch 83/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.6637 - accuracy: 0.5886\n",
            "Epoch 84/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.6173 - accuracy: 0.6034\n",
            "Epoch 85/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.5759 - accuracy: 0.6097\n",
            "Epoch 86/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.6455 - accuracy: 0.5918\n",
            "Epoch 87/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.6979 - accuracy: 0.5862\n",
            "Epoch 88/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.6585 - accuracy: 0.5914\n",
            "Epoch 89/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.6573 - accuracy: 0.5930\n",
            "Epoch 90/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.5544 - accuracy: 0.6145\n",
            "Epoch 91/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.5482 - accuracy: 0.6188\n",
            "Epoch 92/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.5502 - accuracy: 0.6175\n",
            "Epoch 93/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.6009 - accuracy: 0.6048\n",
            "Epoch 94/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.5948 - accuracy: 0.6041\n",
            "Epoch 95/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.5919 - accuracy: 0.6082\n",
            "Epoch 96/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.5907 - accuracy: 0.6061\n",
            "Epoch 97/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.5602 - accuracy: 0.6095\n",
            "Epoch 98/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.5491 - accuracy: 0.6132\n",
            "Epoch 99/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.5521 - accuracy: 0.6159\n",
            "Epoch 100/100\n",
            "377/377 [==============================] - 4s 10ms/step - loss: 1.5268 - accuracy: 0.6260\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "64P5E1lFXkMN",
        "colab_type": "code",
        "outputId": "53101b2b-7258-4e3a-a0f5-69107ca819a6",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 281
        }
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "plt.plot(history.history['accuracy'])\n",
        "plt.plot(history.history['loss'])\n",
        "plt.xlabel(\"Epochs\")\n",
        "plt.ylabel('Accuracy or loss')\n",
        "plt.legend(['Accuracy', 'Loss'])\n",
        "plt.show()"
      ],
      "execution_count": 9,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXgAAAEICAYAAABVv+9nAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n/V39SQAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vAWxctCYYM97",
        "colab_type": "code",
        "outputId": "6aa926ca-81c6-4b3d-fa69-385eebbf83ab",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 105
        }
      },
      "source": [
        "seed_text='Utkarsh Lives in India'\n",
        "for _ in range(50):\n",
        "  token_list= token.texts_to_sequences([seed_text])[0]\n",
        "  token_list= pad_sequences([token_list], maxlen= max_sequence_length-1, padding='pre')\n",
        "  predict= model.predict_classes(token_list, verbose=0)\n",
        "  for word, index in token.word_index.items():\n",
        "    if index== predict:\n",
        "      output_word= word\n",
        "      break\n",
        "  seed_text+=' '+ output_word\n",
        "\n",
        "print(seed_text)"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From <ipython-input-10-d7fe4208e5b3>:5: Sequential.predict_classes (from tensorflow.python.keras.engine.sequential) is deprecated and will be removed after 2021-01-01.\n",
            "Instructions for updating:\n",
            "Please use instead:* `np.argmax(model.predict(x), axis=-1)`,   if your model does multi-class classification   (e.g. if it uses a `softmax` last-layer activation).* `(model.predict(x) > 0.5).astype(\"int32\")`,   if your model does binary classification   (e.g. if it uses a `sigmoid` last-layer activation).\n",
            "Utkarsh Lives in India my heart my true love they they sat to do you you say now how rings rings rest do we died so glen im rest to save me in black away oer your bower oer the hurley or now into i meet to the bower oer the name of now\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uN-Gy46MqwLI",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}