The_chat_bot_using_DL_seq2seq.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "name": "The chat bot using DL",
      "version": "0.3.2",
      "provenance": [],
      "collapsed_sections": [],
      "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/moustafa-7/ChatBot-Project/blob/master/The_chat_bot_using_DL_seq2seq.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "bGik37-3W8SZ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import numpy as np\n",
        "import pandas as pd\n",
        "import string\n",
        "import pickle\n",
        "import operator\n",
        "import matplotlib.pyplot as plt\n",
        "%matplotlib inline"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EmJ1ASv2hmOa",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import codecs\n",
        "#!wget -c http://www.cs.cornell.edu/~cristian/data/cornell_movie_dialogs_corpus.zip\n",
        "#!unzip cornell_movie_dialogs_corpus.zip\n",
        "with codecs.open(\"cornell movie-dialogs corpus/movie_lines.txt\", \"rb\", encoding=\"utf-8\", errors=\"ignore\") as f:\n",
        "    lines = f.read().split(\"\\n\")\n",
        "    conversations = []\n",
        "    for line in lines:\n",
        "        data = line.split(\" +++$+++ \")\n",
        "        conversations.append(data)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6nAYtzh0h31a",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "chats = {}\n",
        "for tokens in conversations:\n",
        "  if len(tokens)>4:\n",
        "    idx = tokens[0][1:]\n",
        "    chat = tokens[4]\n",
        "    chats [int(idx)] = chat"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_MhIx5gzi7PV",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "sorted_chats = sorted(chats.items(), key = lambda x: x[0])\n",
        "#sorted_chats"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TcmObjb7sYMk",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "conves_dict = {}\n",
        "counter = 1\n",
        "conves_ids = []\n",
        "for i in range(1, len(sorted_chats)+1):\n",
        "    if i < len(sorted_chats):\n",
        "        if (sorted_chats[i][0] - sorted_chats[i-1][0]) == 1:\n",
        "            if sorted_chats[i-1][1] not in conves_ids:\n",
        "                conves_ids.append(sorted_chats[i-1][1])\n",
        "            conves_ids.append(sorted_chats[i][1])\n",
        "        elif (sorted_chats[i][0] - sorted_chats[i-1][0]) > 1:            \n",
        "            conves_dict[counter] = conves_ids\n",
        "            conves_ids = []\n",
        "        counter += 1\n",
        "    else:\n",
        "        pass"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zU2vo04Us4Ia",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#conves_dict"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "36Z9SVsGtQS9",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "context_and_target = []\n",
        "\n",
        "for conves in conves_dict.values():\n",
        "  if (len(conves) % 2) != 0:\n",
        "    conves = conves[:-1]\n",
        "    \n",
        "  for i in range(0, len(conves), 2):\n",
        "    context_and_target.append((conves[i], conves[i+1]))\n",
        "  \n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fmJQ4Pifwkf4",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#context_and_target"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aXB_paQTwnXd",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "context, target = zip(*context_and_target)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "3dJDAX0Fw3jB",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "context = list(context)\n",
        "target = list(target)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IwSDDK5Ew8Lp",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import re\n",
        "def clean_text(text):\n",
        "    '''Clean text by removing unnecessary characters and altering the format of words.'''\n",
        "\n",
        "    text = text.lower()\n",
        "    \n",
        "    text = re.sub(r\"i'm\", \"i am\", text)\n",
        "    text = re.sub(r\"he's\", \"he is\", text)\n",
        "    text = re.sub(r\"she's\", \"she is\", text)\n",
        "    text = re.sub(r\"it's\", \"it is\", text)\n",
        "    text = re.sub(r\"that's\", \"that is\", text)\n",
        "    text = re.sub(r\"what's\", \"that is\", text)\n",
        "    text = re.sub(r\"where's\", \"where is\", text)\n",
        "    text = re.sub(r\"how's\", \"how is\", text)\n",
        "    text = re.sub(r\"\\'ll\", \" will\", text)\n",
        "    text = re.sub(r\"\\'ve\", \" have\", text)\n",
        "    text = re.sub(r\"\\'re\", \" are\", text)\n",
        "    text = re.sub(r\"\\'d\", \" would\", text)\n",
        "    text = re.sub(r\"\\'re\", \" are\", text)\n",
        "    text = re.sub(r\"won't\", \"will not\", text)\n",
        "    text = re.sub(r\"can't\", \"cannot\", text)\n",
        "    text = re.sub(r\"n't\", \" not\", text)\n",
        "    text = re.sub(r\"n'\", \"ng\", text)\n",
        "    text = re.sub(r\"'bout\", \"about\", text)\n",
        "    text = re.sub(r\"'til\", \"until\", text)\n",
        "    text = re.sub(r\"[-()\\\"#/@;:<>{}`+=~|.!?,]\", \"\", text)\n",
        "    \n",
        "    return text"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "F81oSOANyAiC",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "clean_target = []\n",
        "for targ in target:\n",
        "  clean_target.append(clean_text(targ))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hQn8TUs1ypl5",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "clean_context =[]\n",
        "for cont in context:\n",
        "  clean_context.append(clean_text(cont))"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "5Y7C6KBNzFBH",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "#clean_context"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "oF0lMIx7zHOh",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# Beggining of a sentence <BOS>  and End of Senctence <EOS>\n",
        "bos = '<BOS> '\n",
        "eos = ' <EOS>'\n",
        "\n",
        "final_target = [bos + targ + eos for targ in clean_target]\n",
        "final_context = [bos + cont + eos for cont in clean_context]\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "OUUFkDH04AU9",
        "colab_type": "code",
        "outputId": "bb2dd975-af64-4806-e752-96de7084a9c7",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 289
        }
      },
      "source": [
        "import codecs\n",
        "!wget https://github.com/samurainote/Automatic-Encoder-Decoder_Seq2Seq_Chatbot/raw/master/encoder_inputs.txt\n",
        "with codecs.open(\"encoder_inputs.txt\", 'rb', encoding = \"utf-8\", errors = \"ignore\") as f:\n",
        "  lines = f.read().split('\\n')\n",
        "  encoder_text = []\n",
        "  for line in lines:\n",
        "    data = line.split('\\n')[0]\n",
        "    encoder_text.append(data)"
      ],
      "execution_count": 17,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "--2019-06-07 04:16:44--  https://github.com/samurainote/Automatic-Encoder-Decoder_Seq2Seq_Chatbot/raw/master/encoder_inputs.txt\n",
            "Resolving github.com (github.com)... 192.30.253.112\n",
            "Connecting to github.com (github.com)|192.30.253.112|:443... connected.\n",
            "HTTP request sent, awaiting response... 302 Found\n",
            "Location: https://raw.githubusercontent.com/samurainote/Automatic-Encoder-Decoder_Seq2Seq_Chatbot/master/encoder_inputs.txt [following]\n",
            "--2019-06-07 04:16:45--  https://raw.githubusercontent.com/samurainote/Automatic-Encoder-Decoder_Seq2Seq_Chatbot/master/encoder_inputs.txt\n",
            "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 151.101.0.133, 151.101.64.133, 151.101.128.133, ...\n",
            "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|151.101.0.133|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 7854930 (7.5M) [text/plain]\n",
            "Saving to: ‘encoder_inputs.txt.2’\n",
            "\n",
            "\rencoder_inputs.txt.   0%[                    ]       0  --.-KB/s               \rencoder_inputs.txt. 100%[===================>]   7.49M  --.-KB/s    in 0.1s    \n",
            "\n",
            "2019-06-07 04:16:46 (75.8 MB/s) - ‘encoder_inputs.txt.2’ saved [7854930/7854930]\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "sfH4oy1x715L",
        "colab_type": "code",
        "outputId": "c819c3a6-5f10-4ab4-bafe-cc7d66f33dbd",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "encoder_text = encoder_text[0:14499]\n",
        "len(encoder_text)"
      ],
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "14499"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 23
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uXQ1J0klFUKw",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "encoder_text"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "J7ONj0kz9Lsm",
        "colab_type": "code",
        "outputId": "724529c7-87a5-4f3c-8258-bd80e311f8cf",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 289
        }
      },
      "source": [
        "!wget https://github.com/samurainote/Automatic-Encoder-Decoder_Seq2Seq_Chatbot/raw/master/decoder_inputs.txt\n",
        "\n",
        "with codecs.open(\"decoder_inputs.txt\", 'rb', encoding = \"utf-8\", errors = \"ignore\") as f:\n",
        "  lines  = f.read().split('\\n')\n",
        "  decoder_text = []\n",
        "  for line in lines:\n",
        "    data = line.split('\\n')[0]\n",
        "    decoder_text.append(data) "
      ],
      "execution_count": 20,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "--2019-06-07 04:16:51--  https://github.com/samurainote/Automatic-Encoder-Decoder_Seq2Seq_Chatbot/raw/master/decoder_inputs.txt\n",
            "Resolving github.com (github.com)... 192.30.253.112\n",
            "Connecting to github.com (github.com)|192.30.253.112|:443... connected.\n",
            "HTTP request sent, awaiting response... 302 Found\n",
            "Location: https://raw.githubusercontent.com/samurainote/Automatic-Encoder-Decoder_Seq2Seq_Chatbot/master/decoder_inputs.txt [following]\n",
            "--2019-06-07 04:16:52--  https://raw.githubusercontent.com/samurainote/Automatic-Encoder-Decoder_Seq2Seq_Chatbot/master/decoder_inputs.txt\n",
            "Resolving raw.githubusercontent.com (raw.githubusercontent.com)... 151.101.0.133, 151.101.64.133, 151.101.128.133, ...\n",
            "Connecting to raw.githubusercontent.com (raw.githubusercontent.com)|151.101.0.133|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 9270194 (8.8M) [text/plain]\n",
            "Saving to: ‘decoder_inputs.txt.2’\n",
            "\n",
            "decoder_inputs.txt. 100%[===================>]   8.84M  --.-KB/s    in 0.1s    \n",
            "\n",
            "2019-06-07 04:16:53 (83.9 MB/s) - ‘decoder_inputs.txt.2’ saved [9270194/9270194]\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jmukqbaj-KjD",
        "colab_type": "code",
        "outputId": "51cf86ff-2acc-4e47-943d-5c8ad7b8bc20",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "decoder_text = decoder_text[0:14499]\n",
        "len(decoder_text)"
      ],
      "execution_count": 24,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "14499"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 24
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jhQaN-8wHn_p",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "decoder_text"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "P5tpJb2d-i8I",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "full_text = encoder_text + decoder_text"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "y4FS6r8M_WPH",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "# dictionary = []\n",
        "# print(\"Making dictionary of words.\\n\")\n",
        "# for text in full_text:\n",
        "#   words = text.split()\n",
        "#   for i in range(0,len(words)):\n",
        "#     if words[i] not in dictionary:\n",
        "#       dictionary.append(words[i])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6NVEww9p_2iK",
        "colab_type": "code",
        "outputId": "75192bd5-b3ab-42bd-efe1-4a027016cd64",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "from keras.preprocessing.text import Tokenizer"
      ],
      "execution_count": 28,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Using TensorFlow backend.\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZXA1tDiyCUqq",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "vocab_size = 14500\n",
        "tokenizer = Tokenizer(num_words = vocab_size)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "EMANi9XkComI",
        "colab_type": "code",
        "outputId": "0903be49-62c5-485d-e674-eb0faac717aa",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "tokenizer.fit_on_texts(full_text)\n",
        "word_index = tokenizer.word_index\n",
        "len(word_index)"
      ],
      "execution_count": 30,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "17294"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 30
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "MEyxHTyQDBul",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "index2word = {}\n",
        "for k,v in word_index.items():\n",
        "  if v<14500:\n",
        "    index2word[v] = k\n",
        "  if v>14500:\n",
        "    continue"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uSbHsQVgGvns",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "index2word"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FIRuk9r3GwvT",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "word2index = {}\n",
        "for k,v in index2word.items():\n",
        "  word2index[v] = k\n",
        "  "
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "BWUL46-OH-Bv",
        "colab_type": "code",
        "outputId": "d7a46730-808f-47e1-c585-3ae427f21a48",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "len(word2index)"
      ],
      "execution_count": 38,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "14499"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 38
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PiGsMdMdH_pS",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "encoder_sequences = tokenizer.texts_to_sequences(encoder_text)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "U7dr1-FaIP0W",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "decoder_sequences = tokenizer.texts_to_sequences(decoder_text)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "chu64s2eIXEq",
        "colab_type": "code",
        "outputId": "7b822a4d-c861-498f-a8f0-ae0119697d8e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "len(encoder_sequences)"
      ],
      "execution_count": 41,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "14499"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 41
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "M4kE8Bg3IeNM",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "for seqs in encoder_sequences:\n",
        "  for seq in seqs:\n",
        "    if seq>14499:\n",
        "      print(seq)\n",
        "      break"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "doYf5uABJLOE",
        "colab_type": "code",
        "outputId": "5454e496-37c9-45d3-fcb3-2b763a3f18bc",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "vocab_size = len(word2index)+1\n",
        "vocab_size "
      ],
      "execution_count": 43,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "14500"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 43
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "IiYdFe_W9LPg",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "decoder_output_data = None"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hgR5lSfxJhFK",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "import numpy as np\n",
        "max_len = 20\n",
        "num_samples = len(encoder_sequences)\n",
        "num_samples\n",
        "decoder_output_data = np.zeros((num_samples, max_len, vocab_size), dtype=\"float16\")\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dtyVx2mTJcoa",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "decoder_output_data"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nVuaYBikSjyJ",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "for i, seqs in enumerate(decoder_input_data):\n",
        "    for j, seq in enumerate(seqs):\n",
        "        if j > 0:\n",
        "            decoder_output_data[i][j][seq] = 1."
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JezJHojfKHMv",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from keras.preprocessing.sequence import pad_sequences\n",
        "encoder_input_data = pad_sequences(encoder_sequences, maxlen = max_len, dtype = 'int32', padding='post', truncating='post')\n",
        "decoder_input_data = pad_sequences(decoder_sequences, maxlen = max_len, dtype = 'int32', padding = 'post', truncating = 'post')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-ujlij4DSWEX",
        "colab_type": "code",
        "outputId": "62c855ff-07bf-4c79-d6a6-0a380dbb4ae2",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 51
        }
      },
      "source": [
        "decoder_input_data[0]"
      ],
      "execution_count": 48,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "array([ 1, 32,  2,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,  0,\n",
              "        0,  0,  0], dtype=int32)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 48
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "rzxNtuj7kcT3",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 476
        },
        "outputId": "6ee6ec2a-8db2-46b1-82f2-e569bec286a4"
      },
      "source": [
        "!wget http://nlp.stanford.edu/data/glove.6B.zip\n",
        "!unzip glove.6B.zip\n",
        "embeddings_index = {}\n",
        "with open('glove.6B.50d.txt', encoding='utf-8') 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",
        "    f.close()\n",
        "\n",
        "print(\"Glove Loded!\")"
      ],
      "execution_count": 53,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "--2019-06-07 04:26:02--  http://nlp.stanford.edu/data/glove.6B.zip\n",
            "Resolving nlp.stanford.edu (nlp.stanford.edu)... 171.64.67.140\n",
            "Connecting to nlp.stanford.edu (nlp.stanford.edu)|171.64.67.140|:80... connected.\n",
            "HTTP request sent, awaiting response... 302 Found\n",
            "Location: https://nlp.stanford.edu/data/glove.6B.zip [following]\n",
            "--2019-06-07 04:26:02--  https://nlp.stanford.edu/data/glove.6B.zip\n",
            "Connecting to nlp.stanford.edu (nlp.stanford.edu)|171.64.67.140|:443... connected.\n",
            "HTTP request sent, awaiting response... 301 Moved Permanently\n",
            "Location: http://downloads.cs.stanford.edu/nlp/data/glove.6B.zip [following]\n",
            "--2019-06-07 04:26:02--  http://downloads.cs.stanford.edu/nlp/data/glove.6B.zip\n",
            "Resolving downloads.cs.stanford.edu (downloads.cs.stanford.edu)... 171.64.64.22\n",
            "Connecting to downloads.cs.stanford.edu (downloads.cs.stanford.edu)|171.64.64.22|:80... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 862182613 (822M) [application/zip]\n",
            "Saving to: ‘glove.6B.zip’\n",
            "\n",
            "glove.6B.zip        100%[===================>] 822.24M  14.1MB/s    in 67s     \n",
            "\n",
            "2019-06-07 04:27:09 (12.3 MB/s) - ‘glove.6B.zip’ saved [862182613/862182613]\n",
            "\n",
            "Archive:  glove.6B.zip\n",
            "replace glove.6B.50d.txt? [y]es, [n]o, [A]ll, [N]one, [r]ename: y\n",
            "  inflating: glove.6B.50d.txt        \n",
            "  inflating: glove.6B.100d.txt       \n",
            "  inflating: glove.6B.200d.txt       \n",
            "  inflating: glove.6B.300d.txt       \n",
            "Glove Loded!\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yL0f_3GVkcW5",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "embedding_dimention = 50\n",
        "def embedding_matrix_creater(embedding_dimention, word_index):\n",
        "    embedding_matrix = np.zeros((len(word_index) + 1, embedding_dimention))\n",
        "    for word, i in word_index.items():\n",
        "        embedding_vector = embeddings_index.get(word)\n",
        "        if embedding_vector is not None:\n",
        "          # words not found in embedding index will be all-zeros.\n",
        "            embedding_matrix[i] = embedding_vector\n",
        "    return embedding_matrix"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "LAADQOAOkcZ8",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "embedding_matrix = embedding_matrix_creater(50, word_index=word2index)\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Q48Ckh6QNSRG",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from keras.layers import Embedding\n",
        "from keras.layers import Input, Dense, LSTM, TimeDistributed\n",
        "from keras.models import Model"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "yfJXyqHzkcdx",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 88
        },
        "outputId": "0e4c3a0b-eac8-4f8f-fe3f-8821eaf34d0d"
      },
      "source": [
        "embed_layer = Embedding(input_dim=vocab_size, output_dim=50, trainable=True,)\n",
        "embed_layer.build((None,))\n",
        "embed_layer.set_weights([embedding_matrix])"
      ],
      "execution_count": 59,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/framework/op_def_library.py:263: colocate_with (from tensorflow.python.framework.ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Colocations handled automatically by placer.\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "1jw1C5xukcgt",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "def seq2seq_model_builder(HIDDEN_DIM=300):\n",
        "    \n",
        "    encoder_inputs = Input(shape=(max_len, ), dtype='int32',)\n",
        "    encoder_embedding = embed_layer(encoder_inputs)\n",
        "    encoder_LSTM = LSTM(HIDDEN_DIM, return_state=True)\n",
        "    encoder_outputs, state_h, state_c = encoder_LSTM(encoder_embedding)\n",
        "    \n",
        "    decoder_inputs = Input(shape=(max_len, ), dtype='int32',)\n",
        "    decoder_embedding = embed_layer(decoder_inputs)\n",
        "    decoder_LSTM = LSTM(HIDDEN_DIM, return_state=True, return_sequences=True)\n",
        "    decoder_outputs, _, _ = decoder_LSTM(decoder_embedding, initial_state=[state_h, state_c])\n",
        "    \n",
        "    # dense_layer = Dense(VOCAB_SIZE, activation='softmax')\n",
        "    outputs = TimeDistributed(Dense(vocab_size, activation='softmax'))(decoder_outputs)\n",
        "    model = Model([encoder_inputs, decoder_inputs], outputs)\n",
        "    \n",
        "    return model"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "cx7jQ1Vrkcjd",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model = seq2seq_model_builder(HIDDEN_DIM=300)\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NfFT2GvsNsSB",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "from keras.utils import plot_model\n",
        "plot_model(model)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0IjUl05NNsVW",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "model.compile(optimizer='adam', loss ='categorical_crossentropy', metrics = ['accuracy'])\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "TVuz0gN6NsY4",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "BATCH_SIZE = 32\n",
        "EPOCHS = 5"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "WVWpcnjcNscA",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        },
        "outputId": "28e266e9-d38c-4a82-cb01-db6a92c16199"
      },
      "source": [
        "encoder_input_data.shape\n"
      ],
      "execution_count": 67,
      "outputs": [
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "(14499, 20)"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 67
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gXMW9mRyNsis",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 258
        },
        "outputId": "ec0160c2-9bb3-49bb-bbf6-5d3ae3b4111a"
      },
      "source": [
        "history = model.fit([encoder_input_data, decoder_input_data], \n",
        "                     decoder_output_data, \n",
        "                     epochs=EPOCHS, \n",
        "                     batch_size=BATCH_SIZE)"
      ],
      "execution_count": 68,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/tensorflow/python/ops/math_ops.py:3066: to_int32 (from tensorflow.python.ops.math_ops) is deprecated and will be removed in a future version.\n",
            "Instructions for updating:\n",
            "Use tf.cast instead.\n",
            "Epoch 1/5\n",
            "14499/14499 [==============================] - 57s 4ms/step - loss: 3.0342 - acc: 0.5312\n",
            "Epoch 2/5\n",
            "14499/14499 [==============================] - 54s 4ms/step - loss: 2.0992 - acc: 0.6408\n",
            "Epoch 3/5\n",
            "14499/14499 [==============================] - 54s 4ms/step - loss: 1.5192 - acc: 0.7556\n",
            "Epoch 4/5\n",
            "14499/14499 [==============================] - 54s 4ms/step - loss: 1.1340 - acc: 0.8103\n",
            "Epoch 5/5\n",
            "14499/14499 [==============================] - 55s 4ms/step - loss: 0.8065 - acc: 0.8488\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "xDuz3ONDOOdl",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "BATCH_SIZE = 64\n",
        "EPOCHS = 8"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "NSiBVSN8OOg0",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 289
        },
        "outputId": "5270187c-efe9-4076-bf93-2d27e91d209b"
      },
      "source": [
        "history = model.fit([encoder_input_data, decoder_input_data], \n",
        "                     decoder_output_data, \n",
        "                     epochs=EPOCHS, \n",
        "                     batch_size=BATCH_SIZE)"
      ],
      "execution_count": 70,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Epoch 1/8\n",
            "14499/14499 [==============================] - 43s 3ms/step - loss: 0.5812 - acc: 0.8768\n",
            "Epoch 2/8\n",
            "14499/14499 [==============================] - 43s 3ms/step - loss: 0.4651 - acc: 0.8915\n",
            "Epoch 3/8\n",
            "14499/14499 [==============================] - 43s 3ms/step - loss: 0.3672 - acc: 0.9044\n",
            "Epoch 4/8\n",
            "14499/14499 [==============================] - 43s 3ms/step - loss: 0.2855 - acc: 0.9153\n",
            "Epoch 5/8\n",
            "14499/14499 [==============================] - 43s 3ms/step - loss: 0.2188 - acc: 0.9245\n",
            "Epoch 6/8\n",
            "14499/14499 [==============================] - 43s 3ms/step - loss: 0.1651 - acc: 0.9321\n",
            "Epoch 7/8\n",
            "14499/14499 [==============================] - 43s 3ms/step - loss: 0.1224 - acc: 0.9382\n",
            "Epoch 8/8\n",
            "14499/14499 [==============================] - 43s 3ms/step - loss: 0.0889 - acc: 0.9441\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "hCCOE1tnOOkJ",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 404
        },
        "outputId": "802d9235-4348-4f09-8465-87275c241295"
      },
      "source": [
        "import matplotlib.pyplot as plt\n",
        "%matplotlib inline\n",
        "\n",
        "plt.figure(figsize=(10, 6))\n",
        "plt.plot(history.history['acc'])\n",
        "#plt.plot(history.history['val_acc'])\n",
        "plt.title('model accuracy')\n",
        "plt.ylabel('accuracy')\n",
        "plt.xlabel('epoch')\n",
        "# plt.legend(['train', 'test'], loc='upper left')\n",
        "plt.show()"
      ],
      "execution_count": 71,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmsAAAGDCAYAAAB0s1eWAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ntAAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 720x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FAa2wOphOOqc",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 404
        },
        "outputId": "b5372e21-2436-4a12-f06d-1b007972192d"
      },
      "source": [
        "\n",
        "# 損失関数の可視化\n",
        "plt.figure(figsize=(10, 6))\n",
        "plt.plot(history.history['loss'])\n",
        "# plt.plot(history.history['val_loss'])\n",
        "plt.title('model loss')\n",
        "plt.ylabel('loss')\n",
        "plt.xlabel('epoch')\n",
        "# plt.legend(['train', 'test'], loc='upper left')\n",
        "plt.show()"
      ],
      "execution_count": 72,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAmQAAAGDCAYAAACFuAwbAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nxs5aE3ABAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 720x432 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "58x36LasOOth",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 129
        },
        "outputId": "14d9694b-446c-4ff2-f703-6b8d73f82710"
      },
      "source": [
        "with open('seq2seq.json',\"w\").write(model.to_json())\n",
        "\n",
        "# 重みの読み込み\n",
        "model.load_weights('seq2seq.h5')\n",
        "print(\"Saved Model!\")"
      ],
      "execution_count": 73,
      "outputs": [
        {
          "output_type": "error",
          "ename": "SyntaxError",
          "evalue": "ignored",
          "traceback": [
            "\u001b[0;36m  File \u001b[0;32m\"<ipython-input-73-15e957d0646d>\"\u001b[0;36m, line \u001b[0;32m1\u001b[0m\n\u001b[0;31m    with open('seq2seq.json',\"w\").write(model.to_json())\u001b[0m\n\u001b[0m                                                        ^\u001b[0m\n\u001b[0;31mSyntaxError\u001b[0m\u001b[0;31m:\u001b[0m invalid syntax\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "PS3-gu8DOOwy",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 88
        },
        "outputId": "a88a180b-2e01-49cf-f870-37ebf467ac95"
      },
      "source": [
        "model_json = model.to_json()\n",
        "with open(\"model.json\", \"w\") as json_file:\n",
        "    json_file.write(model_json)\n",
        "\n",
        "model.save_weights(\"chatbot_model.h5\")\n",
        "print(\"Saved Model!\")"
      ],
      "execution_count": 74,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Saved Model!\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/keras/engine/network.py:877: UserWarning: Layer lstm_2 was passed non-serializable keyword arguments: {'initial_state': [<tf.Tensor 'lstm_1/while/Exit_3:0' shape=(?, 300) dtype=float32>, <tf.Tensor 'lstm_1/while/Exit_4:0' shape=(?, 300) dtype=float32>]}. They will not be included in the serialized model (and thus will be missing at deserialization time).\n",
            "  '. They will not be included '\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "0y_fQqPFOO0A",
        "colab_type": "code",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 71
        },
        "outputId": "d6c5669b-cbea-4df5-e807-3ba2feb0269c"
      },
      "source": [
        "json_string = model.to_json()\n",
        "open('seq2seq.json', 'w').write(json_string)\n",
        "model.save_weights('seq2seq_weights.h5')"
      ],
      "execution_count": 75,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.6/dist-packages/keras/engine/network.py:877: UserWarning: Layer lstm_2 was passed non-serializable keyword arguments: {'initial_state': [<tf.Tensor 'lstm_1/while/Exit_3:0' shape=(?, 300) dtype=float32>, <tf.Tensor 'lstm_1/while/Exit_4:0' shape=(?, 300) dtype=float32>]}. They will not be included in the serialized model (and thus will be missing at deserialization time).\n",
            "  '. They will not be included '\n"
          ],
          "name": "stderr"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "C3_GMImLOO3U",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-8HUNaRROO65",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "v9P8VUoEOO-u",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fn4h8k1AOPBu",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "vVlD3zATOPIO",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "jAeKd4wdOPL7",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "H0pUf7v8OPOx",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "SaK_ARAcOPSR",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wcrY7qz3OPE_",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "uNjmrF-9OOnq",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "g2CKnE6uSyN0",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "!wget https://www.dropbox.com/sh/o0rze9dulwmon8b/AAAEKe0FpShNMsLAJsuTY8Pwa/my_model_weights_bot.h5"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "nDPzXXtmVwzd",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "!wget https://www.dropbox.com/sh/o0rze9dulwmon8b/AADiuIuRIHbB2-i9l_BewYoFa/my_model_weights_discriminator.h5"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "kjchlUNkV-33",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "!wget https://www.dropbox.com/sh/o0rze9dulwmon8b/AAC3o1KWTAi0uETciZcy4-ava/my_model_weights.h5"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wzA9VAvtWeQo",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "!wget https://www.dropbox.com/sh/o0rze9dulwmon8b/AAAaJ1dHJ4Xke_QHl3gUeuyBa/my_model_weights20.h5"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "VPqLxZKQWk2w",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "gyVKKu0bka40",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "7kCkNgLIka-s",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "fyUcs4TYkbBi",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zZsojo_uka7u",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        ""
      ],
      "execution_count": 0,
      "outputs": []
    }
  ]
}