NLP_Embedding.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "NLP_Embedding.ipynb",
      "provenance": [],
      "authorship_tag": "ABX9TyMjEwNoUz8uQAIQrd0xyCWU",
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/antreev-brar/ML-projects/blob/master/NLP_Embedding.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pWyUnteg4aDW",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 208
        },
        "outputId": "4a26ad86-4534-4945-a7f0-a2502b91b752"
      },
      "source": [
        "import csv\n",
        "import tensorflow as tf\n",
        "import numpy as np\n",
        "from tensorflow.keras.preprocessing.text import Tokenizer\n",
        "from tensorflow.keras.preprocessing.sequence import pad_sequences\n",
        "\n",
        "!wget --no-check-certificate \\\n",
        "    https://storage.googleapis.com/laurencemoroney-blog.appspot.com/bbc-text.csv \\\n",
        "    -O /tmp/bbc-text.csv"
      ],
      "execution_count": 1,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "--2020-05-06 12:10:59--  https://storage.googleapis.com/laurencemoroney-blog.appspot.com/bbc-text.csv\n",
            "Resolving storage.googleapis.com (storage.googleapis.com)... 142.250.98.128, 2607:f8b0:400c:c00::80\n",
            "Connecting to storage.googleapis.com (storage.googleapis.com)|142.250.98.128|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 5057493 (4.8M) [application/octet-stream]\n",
            "Saving to: ‘/tmp/bbc-text.csv’\n",
            "\n",
            "\r/tmp/bbc-text.csv     0%[                    ]       0  --.-KB/s               \r/tmp/bbc-text.csv   100%[===================>]   4.82M  --.-KB/s    in 0.05s   \n",
            "\n",
            "2020-05-06 12:10:59 (106 MB/s) - ‘/tmp/bbc-text.csv’ saved [5057493/5057493]\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NWHue0hB4blU",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "vocab_size = 1000\n",
        "embedding_dim = 16\n",
        "max_length = 120\n",
        "trunc_type = 'post'\n",
        "padding_type = 'post'\n",
        "oov_tok = '<OOV>'\n",
        "training_portion = .8"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "l1mNsmqG4bnv",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "c3477d1e-ed54-4115-a8c8-8f313a27acbe"
      },
      "source": [
        "sentences = []\n",
        "labels = []\n",
        "stopwords = [ \"a\", \"about\", \"above\", \"after\", \"again\", \"against\", \"all\", \"am\", \"an\", \"and\", \"any\", \"are\", \"as\", \"at\", \"be\", \"because\", \"been\", \"before\", \"being\", \"below\", \"between\", \"both\", \"but\", \"by\", \"could\", \"did\", \"do\", \"does\", \"doing\", \"down\", \"during\", \"each\", \"few\", \"for\", \"from\", \"further\", \"had\", \"has\", \"have\", \"having\", \"he\", \"he'd\", \"he'll\", \"he's\", \"her\", \"here\", \"here's\", \"hers\", \"herself\", \"him\", \"himself\", \"his\", \"how\", \"how's\", \"i\", \"i'd\", \"i'll\", \"i'm\", \"i've\", \"if\", \"in\", \"into\", \"is\", \"it\", \"it's\", \"its\", \"itself\", \"let's\", \"me\", \"more\", \"most\", \"my\", \"myself\", \"nor\", \"of\", \"on\", \"once\", \"only\", \"or\", \"other\", \"ought\", \"our\", \"ours\", \"ourselves\", \"out\", \"over\", \"own\", \"same\", \"she\", \"she'd\", \"she'll\", \"she's\", \"should\", \"so\", \"some\", \"such\", \"than\", \"that\", \"that's\", \"the\", \"their\", \"theirs\", \"them\", \"themselves\", \"then\", \"there\", \"there's\", \"these\", \"they\", \"they'd\", \"they'll\", \"they're\", \"they've\", \"this\", \"those\", \"through\", \"to\", \"too\", \"under\", \"until\", \"up\", \"very\", \"was\", \"we\", \"we'd\", \"we'll\", \"we're\", \"we've\", \"were\", \"what\", \"what's\", \"when\", \"when's\", \"where\", \"where's\", \"which\", \"while\", \"who\", \"who's\", \"whom\", \"why\", \"why's\", \"with\", \"would\", \"you\", \"you'd\", \"you'll\", \"you're\", \"you've\", \"your\", \"yours\", \"yourself\", \"yourselves\" ]\n",
        "print(len(stopwords))"
      ],
      "execution_count": 3,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "153\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "V15AvGF74bqM",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 89
        },
        "outputId": "5e3bcc4c-08fe-4cb5-d92a-24b78d161e97"
      },
      "source": [
        "with open(\"/tmp/bbc-text.csv\", 'r') as csvfile:\n",
        "    data = csv.reader(csvfile , delimiter = ',')\n",
        "    next(data)\n",
        "    \n",
        "    for row in data :\n",
        "      sentence = row[1]\n",
        "      labels.append(row[0])\n",
        "      for word in stopwords :\n",
        "        token =' '+ word + ' '\n",
        "        sentence = sentence.replace(token , ' ')\n",
        "      sentences.append(sentence)\n",
        "\n",
        "\n",
        "    \n",
        "print(len(labels))\n",
        "print(len(sentences))\n",
        "print(sentences[0])"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "2225\n",
            "2225\n",
            "tv future hands viewers home theatre systems  plasma high-definition tvs  digital video recorders moving living room  way people watch tv will radically different five years  time.  according expert panel gathered annual consumer electronics show las vegas discuss new technologies will impact one favourite pastimes. us leading trend  programmes content will delivered viewers via home networks  cable  satellite  telecoms companies  broadband service providers front rooms portable devices.  one talked-about technologies ces digital personal video recorders (dvr pvr). set-top boxes  like us s tivo uk s sky+ system  allow people record  store  play  pause forward wind tv programmes want.  essentially  technology allows much personalised tv. also built-in high-definition tv sets  big business japan us  slower take off europe lack high-definition programming. not can people forward wind adverts  can also forget abiding network channel schedules  putting together a-la-carte entertainment. us networks cable satellite companies worried means terms advertising revenues well  brand identity  viewer loyalty channels. although us leads technology moment  also concern raised europe  particularly growing uptake services like sky+.  happens today  will see nine months years  time uk   adam hume  bbc broadcast s futurologist told bbc news website. likes bbc  no issues lost advertising revenue yet. pressing issue moment commercial uk broadcasters  brand loyalty important everyone.  will talking content brands rather network brands   said tim hanlon  brand communications firm starcom mediavest.  reality broadband connections  anybody can producer content.  added:  challenge now hard promote programme much choice.   means  said stacey jolna  senior vice president tv guide tv group  way people find content want watch simplified tv viewers. means networks  us terms  channels take leaf google s book search engine future  instead scheduler help people find want watch. kind channel model might work younger ipod generation used taking control gadgets play them. might not suit everyone  panel recognised. older generations comfortable familiar schedules channel brands know getting. perhaps not want much choice put hands  mr hanlon suggested.  end  kids just diapers pushing buttons already - everything possible available   said mr hanlon.  ultimately  consumer will tell market want.   50 000 new gadgets technologies showcased ces  many enhancing tv-watching experience. high-definition tv sets everywhere many new models lcd (liquid crystal display) tvs launched dvr capability built  instead external boxes. one example launched show humax s 26-inch lcd tv 80-hour tivo dvr dvd recorder. one us s biggest satellite tv companies  directtv  even launched branded dvr show 100-hours recording capability  instant replay  search function. set can pause rewind tv 90 hours. microsoft chief bill gates announced pre-show keynote speech partnership tivo  called tivotogo  means people can play recorded programmes windows pcs mobile devices. reflect increasing trend freeing multimedia people can watch want  want.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "muAHzZxT4bsh",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 104
        },
        "outputId": "65a3f203-a818-4695-d60f-d21f1fda99bc"
      },
      "source": [
        "train_size = round(training_portion * len(sentences))\n",
        "\n",
        "train_sentences = sentences[:train_size]\n",
        "train_labels = labels[:train_size]\n",
        "\n",
        "validation_sentences = sentences[train_size :]\n",
        "validation_labels = labels[train_size :]\n",
        "\n",
        "print(train_size)\n",
        "print(len(train_sentences))\n",
        "print(len(train_labels))\n",
        "print(len(validation_sentences))\n",
        "print(len(validation_labels))\n"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "1780\n",
            "1780\n",
            "1780\n",
            "445\n",
            "445\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7kpxd5V_4buu",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 121
        },
        "outputId": "69c84de0-c276-4e19-ea43-6f09fa024e58"
      },
      "source": [
        "tokenizer = Tokenizer(num_words = vocab_size,oov_token= oov_tok)\n",
        "tokenizer.fit_on_texts(train_sentences)\n",
        "word_index = tokenizer.word_index\n",
        "\n",
        "train_sequences = tokenizer.texts_to_sequences(train_sentences)\n",
        "train_padded = pad_sequences(train_sequences ,\n",
        "                             maxlen =max_length , \n",
        "                             padding =padding_type,\n",
        "                             truncating = trunc_type)\n",
        "\n",
        "print(len(train_sequences[0]))\n",
        "print(len(train_padded[0]))\n",
        "\n",
        "print(len(train_sequences[1]))\n",
        "print(len(train_padded[1]))\n",
        "\n",
        "print(len(train_sequences[10]))\n",
        "print(len(train_padded[10]))\n"
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "449\n",
            "120\n",
            "200\n",
            "120\n",
            "192\n",
            "120\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Zt0FLK3n4bxB",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        },
        "outputId": "ab0acc15-4fb6-4580-cb7e-dc9260a52fb1"
      },
      "source": [
        "validation_sequences = tokenizer.texts_to_sequences(validation_sentences)\n",
        "validation_padded = pad_sequences(validation_sequences, padding=padding_type, maxlen=max_length)\n",
        "\n",
        "print(len(validation_sequences))\n",
        "print(validation_padded.shape)\n"
      ],
      "execution_count": 7,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "445\n",
            "(445, 120)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ypY0L_wY4bzX",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 156
        },
        "outputId": "737aef52-a1e0-4994-854b-a65c99ec6902"
      },
      "source": [
        "label_tokenizer = Tokenizer()\n",
        "label_tokenizer.fit_on_texts(labels)\n",
        "\n",
        "training_label_seq = np.array(label_tokenizer.texts_to_sequences(train_labels))\n",
        "validation_label_seq = np.array(label_tokenizer.texts_to_sequences(validation_labels))\n",
        "\n",
        "print(training_label_seq[0])\n",
        "print(training_label_seq[1])\n",
        "print(training_label_seq[2])\n",
        "print(training_label_seq.shape)\n",
        "\n",
        "print(validation_label_seq[0])\n",
        "print(validation_label_seq[1])\n",
        "print(validation_label_seq[2])\n",
        "print(validation_label_seq.shape)\n"
      ],
      "execution_count": 8,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[4]\n",
            "[2]\n",
            "[1]\n",
            "(1780, 1)\n",
            "[5]\n",
            "[4]\n",
            "[3]\n",
            "(445, 1)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "I2N-FdO543Dg",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 294
        },
        "outputId": "2901a4a5-75c5-428a-bccb-d521bd9f9565"
      },
      "source": [
        "model = tf.keras.Sequential([\n",
        "        tf.keras.layers.Embedding(vocab_size , embedding_dim , input_length = max_length),\n",
        "        tf.keras.layers.GlobalAveragePooling1D(),\n",
        "        tf.keras.layers.Dense(24 , activation ='relu'),\n",
        "        tf.keras.layers.Dense(6 , activation = 'softmax')\n",
        "])\n",
        "model.compile(loss='sparse_categorical_crossentropy',optimizer='adam',metrics=['accuracy'])\n",
        "model.summary()\n"
      ],
      "execution_count": 10,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Model: \"sequential_1\"\n",
            "_________________________________________________________________\n",
            "Layer (type)                 Output Shape              Param #   \n",
            "=================================================================\n",
            "embedding_1 (Embedding)      (None, 120, 16)           16000     \n",
            "_________________________________________________________________\n",
            "global_average_pooling1d_1 ( (None, 16)                0         \n",
            "_________________________________________________________________\n",
            "dense_2 (Dense)              (None, 24)                408       \n",
            "_________________________________________________________________\n",
            "dense_3 (Dense)              (None, 6)                 150       \n",
            "=================================================================\n",
            "Total params: 16,558\n",
            "Trainable params: 16,558\n",
            "Non-trainable params: 0\n",
            "_________________________________________________________________\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "pjmFvFJ743HY",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        },
        "outputId": "d7812685-d493-4279-c0b7-8bb93b5d70f5"
      },
      "source": [
        "num_epochs = 30\n",
        "history = model.fit(train_padded , training_label_seq , epochs = num_epochs ,  validation_data= (validation_padded , validation_label_seq) ,  verbose =2)"
      ],
      "execution_count": 11,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/30\n",
            "56/56 - 0s - loss: 1.7574 - accuracy: 0.2938 - val_loss: 1.7175 - val_accuracy: 0.2270\n",
            "Epoch 2/30\n",
            "56/56 - 0s - loss: 1.6650 - accuracy: 0.2303 - val_loss: 1.6206 - val_accuracy: 0.2270\n",
            "Epoch 3/30\n",
            "56/56 - 0s - loss: 1.5511 - accuracy: 0.2860 - val_loss: 1.5099 - val_accuracy: 0.4202\n",
            "Epoch 4/30\n",
            "56/56 - 0s - loss: 1.4065 - accuracy: 0.5612 - val_loss: 1.3666 - val_accuracy: 0.6360\n",
            "Epoch 5/30\n",
            "56/56 - 0s - loss: 1.2200 - accuracy: 0.7702 - val_loss: 1.1928 - val_accuracy: 0.8000\n",
            "Epoch 6/30\n",
            "56/56 - 0s - loss: 1.0078 - accuracy: 0.8590 - val_loss: 1.0084 - val_accuracy: 0.8674\n",
            "Epoch 7/30\n",
            "56/56 - 0s - loss: 0.8091 - accuracy: 0.9219 - val_loss: 0.8412 - val_accuracy: 0.8899\n",
            "Epoch 8/30\n",
            "56/56 - 0s - loss: 0.6450 - accuracy: 0.9309 - val_loss: 0.7083 - val_accuracy: 0.8921\n",
            "Epoch 9/30\n",
            "56/56 - 0s - loss: 0.5132 - accuracy: 0.9404 - val_loss: 0.6020 - val_accuracy: 0.8944\n",
            "Epoch 10/30\n",
            "56/56 - 0s - loss: 0.4116 - accuracy: 0.9528 - val_loss: 0.5144 - val_accuracy: 0.9034\n",
            "Epoch 11/30\n",
            "56/56 - 0s - loss: 0.3335 - accuracy: 0.9567 - val_loss: 0.4480 - val_accuracy: 0.9124\n",
            "Epoch 12/30\n",
            "56/56 - 0s - loss: 0.2755 - accuracy: 0.9691 - val_loss: 0.4004 - val_accuracy: 0.9236\n",
            "Epoch 13/30\n",
            "56/56 - 0s - loss: 0.2314 - accuracy: 0.9702 - val_loss: 0.3577 - val_accuracy: 0.9258\n",
            "Epoch 14/30\n",
            "56/56 - 0s - loss: 0.1971 - accuracy: 0.9758 - val_loss: 0.3278 - val_accuracy: 0.9303\n",
            "Epoch 15/30\n",
            "56/56 - 0s - loss: 0.1702 - accuracy: 0.9781 - val_loss: 0.3024 - val_accuracy: 0.9348\n",
            "Epoch 16/30\n",
            "56/56 - 0s - loss: 0.1479 - accuracy: 0.9815 - val_loss: 0.2812 - val_accuracy: 0.9348\n",
            "Epoch 17/30\n",
            "56/56 - 0s - loss: 0.1307 - accuracy: 0.9826 - val_loss: 0.2682 - val_accuracy: 0.9303\n",
            "Epoch 18/30\n",
            "56/56 - 0s - loss: 0.1147 - accuracy: 0.9865 - val_loss: 0.2507 - val_accuracy: 0.9348\n",
            "Epoch 19/30\n",
            "56/56 - 0s - loss: 0.1017 - accuracy: 0.9882 - val_loss: 0.2429 - val_accuracy: 0.9326\n",
            "Epoch 20/30\n",
            "56/56 - 0s - loss: 0.0911 - accuracy: 0.9899 - val_loss: 0.2309 - val_accuracy: 0.9348\n",
            "Epoch 21/30\n",
            "56/56 - 0s - loss: 0.0813 - accuracy: 0.9916 - val_loss: 0.2222 - val_accuracy: 0.9371\n",
            "Epoch 22/30\n",
            "56/56 - 0s - loss: 0.0732 - accuracy: 0.9927 - val_loss: 0.2153 - val_accuracy: 0.9371\n",
            "Epoch 23/30\n",
            "56/56 - 0s - loss: 0.0664 - accuracy: 0.9938 - val_loss: 0.2094 - val_accuracy: 0.9438\n",
            "Epoch 24/30\n",
            "56/56 - 0s - loss: 0.0596 - accuracy: 0.9949 - val_loss: 0.2056 - val_accuracy: 0.9416\n",
            "Epoch 25/30\n",
            "56/56 - 0s - loss: 0.0543 - accuracy: 0.9949 - val_loss: 0.1989 - val_accuracy: 0.9393\n",
            "Epoch 26/30\n",
            "56/56 - 0s - loss: 0.0491 - accuracy: 0.9972 - val_loss: 0.1946 - val_accuracy: 0.9371\n",
            "Epoch 27/30\n",
            "56/56 - 0s - loss: 0.0447 - accuracy: 0.9972 - val_loss: 0.1933 - val_accuracy: 0.9416\n",
            "Epoch 28/30\n",
            "56/56 - 0s - loss: 0.0408 - accuracy: 0.9972 - val_loss: 0.1890 - val_accuracy: 0.9393\n",
            "Epoch 29/30\n",
            "56/56 - 0s - loss: 0.0374 - accuracy: 0.9983 - val_loss: 0.1873 - val_accuracy: 0.9461\n",
            "Epoch 30/30\n",
            "56/56 - 0s - loss: 0.0343 - accuracy: 0.9994 - val_loss: 0.1847 - val_accuracy: 0.9416\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-Q-oBc8z43JR",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 540
        },
        "outputId": "e924682b-8a10-493d-bd7c-7519e79f9409"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "\n",
        "\n",
        "def plot_graphs(history, string):\n",
        "  plt.plot(history.history[string])\n",
        "  plt.plot(history.history['val_'+string])\n",
        "  plt.xlabel(\"Epochs\")\n",
        "  plt.ylabel(string)\n",
        "  plt.legend([string, 'val_'+string])\n",
        "  plt.show()\n",
        "  \n",
        "plot_graphs(history, \"accuracy\")\n",
        "plot_graphs(history, \"loss\")"
      ],
      "execution_count": 12,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NzL1CBeJ5Ckk",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "reverse_word_index = dict([(value, key) for (key, value) in word_index.items()])\n",
        "\n",
        "def decode_sentence(text):\n",
        "    return ' '.join([reverse_word_index.get(i, '?') for i in text])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WbvidLFQ5Cnr",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "0245c0a5-0c9a-43d8-aedb-36a801e13736"
      },
      "source": [
        "e = model.layers[0]\n",
        "weights = e.get_weights()[0]\n",
        "print(weights.shape) # shape: (vocab_size, embedding_dim)\n"
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(1000, 16)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vXsUHmEj5JcI",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import io\n",
        "\n",
        "out_v = io.open('vecs.tsv', 'w', encoding='utf-8')\n",
        "out_m = io.open('meta.tsv', 'w', encoding='utf-8')\n",
        "for word_num in range(1, vocab_size):\n",
        "  word = reverse_word_index[word_num]\n",
        "  embeddings = weights[word_num]\n",
        "  out_m.write(word + \"\\n\")\n",
        "  out_v.write('\\t'.join([str(x) for x in embeddings]) + \"\\n\")\n",
        "out_v.close()\n",
        "out_m.close()"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "94vhcMLc5Q2A",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "try:\n",
        "  from google.colab import files\n",
        "except ImportError:\n",
        "  pass\n",
        "else:\n",
        "  files.download('vecs.tsv')\n",
        "  files.download('meta.tsv')"
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}