rnn/ex_1.ipynb

{
  "cells": [
    {
      "cell_type": "code",
      "execution_count": 1,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "BX1a7kd3jMGs",
        "outputId": "a3477675-ad8f-463f-9bd6-aa2709fed7c9"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Requirement already satisfied: tensorflow in /usr/local/lib/python3.10/dist-packages (2.15.0)\n",
            "Collecting tensorflow\n",
            "  Downloading tensorflow-2.15.0.post1-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (475.2 MB)\n",
            "\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m475.2/475.2 MB\u001b[0m \u001b[31m3.3 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: absl-py>=1.0.0 in /usr/local/lib/python3.10/dist-packages (from tensorflow) (1.4.0)\n",
            "Requirement already satisfied: astunparse>=1.6.0 in /usr/local/lib/python3.10/dist-packages (from tensorflow) (1.6.3)\n",
            "Requirement already satisfied: flatbuffers>=23.5.26 in /usr/local/lib/python3.10/dist-packages (from tensorflow) (23.5.26)\n",
            "Requirement already satisfied: gast!=0.5.0,!=0.5.1,!=0.5.2,>=0.2.1 in /usr/local/lib/python3.10/dist-packages (from tensorflow) (0.5.4)\n",
            "Requirement already satisfied: google-pasta>=0.1.1 in /usr/local/lib/python3.10/dist-packages (from tensorflow) (0.2.0)\n",
            "Requirement already satisfied: h5py>=2.9.0 in /usr/local/lib/python3.10/dist-packages (from tensorflow) (3.9.0)\n",
            "Requirement already satisfied: libclang>=13.0.0 in /usr/local/lib/python3.10/dist-packages (from tensorflow) (16.0.6)\n",
            "Requirement already satisfied: ml-dtypes~=0.2.0 in /usr/local/lib/python3.10/dist-packages (from tensorflow) (0.2.0)\n",
            "Requirement already satisfied: numpy<2.0.0,>=1.23.5 in /usr/local/lib/python3.10/dist-packages (from tensorflow) (1.25.2)\n",
            "Requirement already satisfied: opt-einsum>=2.3.2 in /usr/local/lib/python3.10/dist-packages (from tensorflow) (3.3.0)\n",
            "Requirement already satisfied: packaging in /usr/local/lib/python3.10/dist-packages (from tensorflow) (23.2)\n",
            "Requirement already satisfied: protobuf!=4.21.0,!=4.21.1,!=4.21.2,!=4.21.3,!=4.21.4,!=4.21.5,<5.0.0dev,>=3.20.3 in /usr/local/lib/python3.10/dist-packages (from tensorflow) (3.20.3)\n",
            "Requirement already satisfied: setuptools in /usr/local/lib/python3.10/dist-packages (from tensorflow) (67.7.2)\n",
            "Requirement already satisfied: six>=1.12.0 in /usr/local/lib/python3.10/dist-packages (from tensorflow) (1.16.0)\n",
            "Requirement already satisfied: termcolor>=1.1.0 in /usr/local/lib/python3.10/dist-packages (from tensorflow) (2.4.0)\n",
            "Requirement already satisfied: typing-extensions>=3.6.6 in /usr/local/lib/python3.10/dist-packages (from tensorflow) (4.9.0)\n",
            "Requirement already satisfied: wrapt<1.15,>=1.11.0 in /usr/local/lib/python3.10/dist-packages (from tensorflow) (1.14.1)\n",
            "Requirement already satisfied: tensorflow-io-gcs-filesystem>=0.23.1 in /usr/local/lib/python3.10/dist-packages (from tensorflow) (0.36.0)\n",
            "Requirement already satisfied: grpcio<2.0,>=1.24.3 in /usr/local/lib/python3.10/dist-packages (from tensorflow) (1.60.1)\n",
            "Requirement already satisfied: tensorboard<2.16,>=2.15 in /usr/local/lib/python3.10/dist-packages (from tensorflow) (2.15.2)\n",
            "Requirement already satisfied: tensorflow-estimator<2.16,>=2.15.0 in /usr/local/lib/python3.10/dist-packages (from tensorflow) (2.15.0)\n",
            "Requirement already satisfied: keras<2.16,>=2.15.0 in /usr/local/lib/python3.10/dist-packages (from tensorflow) (2.15.0)\n",
            "Requirement already satisfied: wheel<1.0,>=0.23.0 in /usr/local/lib/python3.10/dist-packages (from astunparse>=1.6.0->tensorflow) (0.42.0)\n",
            "Requirement already satisfied: google-auth<3,>=1.6.3 in /usr/local/lib/python3.10/dist-packages (from tensorboard<2.16,>=2.15->tensorflow) (2.27.0)\n",
            "Requirement already satisfied: google-auth-oauthlib<2,>=0.5 in /usr/local/lib/python3.10/dist-packages (from tensorboard<2.16,>=2.15->tensorflow) (1.2.0)\n",
            "Requirement already satisfied: markdown>=2.6.8 in /usr/local/lib/python3.10/dist-packages (from tensorboard<2.16,>=2.15->tensorflow) (3.5.2)\n",
            "Requirement already satisfied: requests<3,>=2.21.0 in /usr/local/lib/python3.10/dist-packages (from tensorboard<2.16,>=2.15->tensorflow) (2.31.0)\n",
            "Requirement already satisfied: tensorboard-data-server<0.8.0,>=0.7.0 in /usr/local/lib/python3.10/dist-packages (from tensorboard<2.16,>=2.15->tensorflow) (0.7.2)\n",
            "Requirement already satisfied: werkzeug>=1.0.1 in /usr/local/lib/python3.10/dist-packages (from tensorboard<2.16,>=2.15->tensorflow) (3.0.1)\n",
            "Requirement already satisfied: cachetools<6.0,>=2.0.0 in /usr/local/lib/python3.10/dist-packages (from google-auth<3,>=1.6.3->tensorboard<2.16,>=2.15->tensorflow) (5.3.2)\n",
            "Requirement already satisfied: pyasn1-modules>=0.2.1 in /usr/local/lib/python3.10/dist-packages (from google-auth<3,>=1.6.3->tensorboard<2.16,>=2.15->tensorflow) (0.3.0)\n",
            "Requirement already satisfied: rsa<5,>=3.1.4 in /usr/local/lib/python3.10/dist-packages (from google-auth<3,>=1.6.3->tensorboard<2.16,>=2.15->tensorflow) (4.9)\n",
            "Requirement already satisfied: requests-oauthlib>=0.7.0 in /usr/local/lib/python3.10/dist-packages (from google-auth-oauthlib<2,>=0.5->tensorboard<2.16,>=2.15->tensorflow) (1.3.1)\n",
            "Requirement already satisfied: charset-normalizer<4,>=2 in /usr/local/lib/python3.10/dist-packages (from requests<3,>=2.21.0->tensorboard<2.16,>=2.15->tensorflow) (3.3.2)\n",
            "Requirement already satisfied: idna<4,>=2.5 in /usr/local/lib/python3.10/dist-packages (from requests<3,>=2.21.0->tensorboard<2.16,>=2.15->tensorflow) (3.6)\n",
            "Requirement already satisfied: urllib3<3,>=1.21.1 in /usr/local/lib/python3.10/dist-packages (from requests<3,>=2.21.0->tensorboard<2.16,>=2.15->tensorflow) (2.0.7)\n",
            "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.10/dist-packages (from requests<3,>=2.21.0->tensorboard<2.16,>=2.15->tensorflow) (2024.2.2)\n",
            "Requirement already satisfied: MarkupSafe>=2.1.1 in /usr/local/lib/python3.10/dist-packages (from werkzeug>=1.0.1->tensorboard<2.16,>=2.15->tensorflow) (2.1.5)\n",
            "Requirement already satisfied: pyasn1<0.6.0,>=0.4.6 in /usr/local/lib/python3.10/dist-packages (from pyasn1-modules>=0.2.1->google-auth<3,>=1.6.3->tensorboard<2.16,>=2.15->tensorflow) (0.5.1)\n",
            "Requirement already satisfied: oauthlib>=3.0.0 in /usr/local/lib/python3.10/dist-packages (from requests-oauthlib>=0.7.0->google-auth-oauthlib<2,>=0.5->tensorboard<2.16,>=2.15->tensorflow) (3.2.2)\n",
            "Installing collected packages: tensorflow\n",
            "  Attempting uninstall: tensorflow\n",
            "    Found existing installation: tensorflow 2.15.0\n",
            "    Uninstalling tensorflow-2.15.0:\n",
            "      Successfully uninstalled tensorflow-2.15.0\n",
            "Successfully installed tensorflow-2.15.0.post1\n"
          ]
        }
      ],
      "source": [
        "!pip install --upgrade tensorflow"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 2,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "nRe15Be4lyAJ",
        "outputId": "63e4acf3-f01c-4f5b-a606-2d3416751886"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Collecting fasttext\n",
            "  Downloading fasttext-0.9.2.tar.gz (68 kB)\n",
            "\u001b[?25l     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m0.0/68.8 kB\u001b[0m \u001b[31m?\u001b[0m eta \u001b[36m-:--:--\u001b[0m\r\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m68.8/68.8 kB\u001b[0m \u001b[31m2.2 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25h  Preparing metadata (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "Collecting pybind11>=2.2 (from fasttext)\n",
            "  Using cached pybind11-2.11.1-py3-none-any.whl (227 kB)\n",
            "Requirement already satisfied: setuptools>=0.7.0 in /usr/local/lib/python3.10/dist-packages (from fasttext) (67.7.2)\n",
            "Requirement already satisfied: numpy in /usr/local/lib/python3.10/dist-packages (from fasttext) (1.25.2)\n",
            "Building wheels for collected packages: fasttext\n",
            "  Building wheel for fasttext (setup.py) ... \u001b[?25l\u001b[?25hdone\n",
            "  Created wheel for fasttext: filename=fasttext-0.9.2-cp310-cp310-linux_x86_64.whl size=4199770 sha256=5e34758177e40df4295501ee96abd227ca557dfc0aeb396f2f038c9d1a7a934e\n",
            "  Stored in directory: /root/.cache/pip/wheels/a5/13/75/f811c84a8ab36eedbaef977a6a58a98990e8e0f1967f98f394\n",
            "Successfully built fasttext\n",
            "Installing collected packages: pybind11, fasttext\n",
            "Successfully installed fasttext-0.9.2 pybind11-2.11.1\n"
          ]
        }
      ],
      "source": [
        "!pip install -U fasttext"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 3,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "9NRKjJ8ymI_p",
        "outputId": "6788a5e5-f228-454f-f2aa-8f3b48347f23"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Mon Feb 19 17:32:30 2024       \n",
            "+---------------------------------------------------------------------------------------+\n",
            "| NVIDIA-SMI 535.104.05             Driver Version: 535.104.05   CUDA Version: 12.2     |\n",
            "|-----------------------------------------+----------------------+----------------------+\n",
            "| GPU  Name                 Persistence-M | Bus-Id        Disp.A | Volatile Uncorr. ECC |\n",
            "| Fan  Temp   Perf          Pwr:Usage/Cap |         Memory-Usage | GPU-Util  Compute M. |\n",
            "|                                         |                      |               MIG M. |\n",
            "|=========================================+======================+======================|\n",
            "|   0  Tesla T4                       Off | 00000000:00:04.0 Off |                    0 |\n",
            "| N/A   57C    P8              10W /  70W |      0MiB / 15360MiB |      0%      Default |\n",
            "|                                         |                      |                  N/A |\n",
            "+-----------------------------------------+----------------------+----------------------+\n",
            "                                                                                         \n",
            "+---------------------------------------------------------------------------------------+\n",
            "| Processes:                                                                            |\n",
            "|  GPU   GI   CI        PID   Type   Process name                            GPU Memory |\n",
            "|        ID   ID                                                             Usage      |\n",
            "|=======================================================================================|\n",
            "|  No running processes found                                                           |\n",
            "+---------------------------------------------------------------------------------------+\n"
          ]
        }
      ],
      "source": [
        "!nvidia-smi"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "cSdFYuu_sEtR"
      },
      "source": [
        "##Load the dataset"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 4,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-sYe0nTQrS1B",
        "outputId": "e5b64dcf-486c-48c8-8c93-fd3c44e56f9f"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "--2024-02-19 17:32:30--  https://www.cs.cornell.edu/people/pabo/movie-review-data/review_polarity.tar.gz\n",
            "Resolving www.cs.cornell.edu (www.cs.cornell.edu)... 132.236.207.36\n",
            "Connecting to www.cs.cornell.edu (www.cs.cornell.edu)|132.236.207.36|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 3127238 (3.0M) [application/x-gzip]\n",
            "Saving to: ‘review_polarity.tar.gz’\n",
            "\n",
            "review_polarity.tar 100%[===================>]   2.98M  1.90MB/s    in 1.6s    \n",
            "\n",
            "2024-02-19 17:32:33 (1.90 MB/s) - ‘review_polarity.tar.gz’ saved [3127238/3127238]\n",
            "\n"
          ]
        }
      ],
      "source": [
        "!wget https://www.cs.cornell.edu/people/pabo/movie-review-data/review_polarity.tar.gz\n",
        "!tar xzf review_polarity.tar.gz"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 5,
      "metadata": {
        "id": "ajTyfDWSwgq5"
      },
      "outputs": [],
      "source": [
        "import os\n",
        "from sklearn.datasets import load_files\n",
        "\n",
        "\n",
        "dataset_path = 'txt_sentoken'\n",
        "movie_reviews = load_files(container_path = dataset_path, encoding = 'utf-8')\n",
        "\n",
        "\n",
        "x = movie_reviews.data            #the data\n",
        "y = movie_reviews.target          #the labels\n",
        "z = movie_reviews.target_names    #the names of labels"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 6,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "GDj-7nUvwj5R",
        "outputId": "dc01e16e-51f4-4e61-89b5-beb67f1e2b55"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "[\"arnold schwarzenegger has been an icon for action enthusiasts , since the late 80's , but lately his films have been very sloppy and the one-liners are getting worse . \\nit's hard seeing arnold as mr . freeze in batman and robin , especially when he says tons of ice jokes , but hey he got 15 million , what's it matter to him ? \\nonce again arnold has signed to do another expensive blockbuster , that can't compare with the likes of the terminator series , true lies and even eraser . \\nin this so called dark thriller , the devil ( gabriel byrne ) has come upon earth , to impregnate a woman ( robin tunney ) which happens every 1000 years , and basically destroy the world , but apparently god has chosen one man , and that one man is jericho cane ( arnold himself ) . \\nwith the help of a trusty sidekick ( kevin pollack ) , they will stop at nothing to let the devil take over the world ! \\nparts of this are actually so absurd , that they would fit right in with dogma . \\nyes , the film is that weak , but it's better than the other blockbuster right now ( sleepy hollow ) , but it makes the world is not enough look like a 4 star film . \\nanyway , this definitely doesn't seem like an arnold movie . \\nit just wasn't the type of film you can see him doing . \\nsure he gave us a few chuckles with his well known one-liners , but he seemed confused as to where his character and the film was going . \\nit's understandable , especially when the ending had to be changed according to some sources . \\naside form that , he still walked through it , much like he has in the past few films . \\ni'm sorry to say this arnold but maybe these are the end of your action days . \\nspeaking of action , where was it in this film ? \\nthere was hardly any explosions or fights . \\nthe devil made a few places explode , but arnold wasn't kicking some devil butt . \\nthe ending was changed to make it more spiritual , which undoubtedly ruined the film . \\ni was at least hoping for a cool ending if nothing else occurred , but once again i was let down . \\ni also don't know why the film took so long and cost so much . \\nthere was really no super affects at all , unless you consider an invisible devil , who was in it for 5 minutes tops , worth the overpriced budget . \\nthe budget should have gone into a better script , where at least audiences could be somewhat entertained instead of facing boredom . \\nit's pitiful to see how scripts like these get bought and made into a movie . \\ndo they even read these things anymore ? \\nit sure doesn't seem like it . \\nthankfully gabriel's performance gave some light to this poor film . \\nwhen he walks down the street searching for robin tunney , you can't help but feel that he looked like a devil . \\nthe guy is creepy looking anyway ! \\nwhen it's all over , you're just glad it's the end of the movie . \\ndon't bother to see this , if you're expecting a solid action flick , because it's neither solid nor does it have action . \\nit's just another movie that we are suckered in to seeing , due to a strategic marketing campaign . \\nsave your money and see the world is not enough for an entertaining experience . \\n\"]"
            ]
          },
          "execution_count": 6,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "x[:1]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 7,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1baTcgdfwlUR",
        "outputId": "821023d1-e37a-4541-b929-e851d2eb44e8"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "array([0, 1, 1, ..., 1, 0, 0])"
            ]
          },
          "execution_count": 7,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "y"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 8,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "AB6FzTDpwmzJ",
        "outputId": "c4cbd2ef-309d-44b6-be56-dd771fc57639"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "['neg', 'pos']"
            ]
          },
          "execution_count": 8,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "z"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 9,
      "metadata": {
        "id": "9DcborDYiXqs"
      },
      "outputs": [],
      "source": [
        "def average_doc_length(corpus):\n",
        "    ''' Takes as input a whole corpus\n",
        "      Returns the average number of words and chars per document '''\n",
        "\n",
        "    document_word_lengths = [len(doc.split()) for doc in corpus]                            #length of each doc (in words)\n",
        "    average_doc_length_words = sum(document_word_lengths) / len(document_word_lengths)      #average doc length (in words)\n",
        "\n",
        "    document_char_lengths = [len(doc) for doc in corpus]                                    #length of each doc (in characters)\n",
        "    average_doc_length_chars = sum(document_char_lengths) / len(document_char_lengths)      #average doc length (in characters)\n",
        "\n",
        "    return average_doc_length_words, average_doc_length_chars"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 10,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dFCB-r9biYyj",
        "outputId": "627fe857-6c2a-40e7-9f41-0cccf902628b"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "---Before preprocessing---\n",
            "Average Document Length (in words): 746.3405\n",
            "Average Document Length (in characters): 3893.002\n"
          ]
        }
      ],
      "source": [
        "print(\"---Before preprocessing---\")\n",
        "avg_length_words, avg_length_chars = average_doc_length(x)\n",
        "print(\"Average Document Length (in words):\", avg_length_words)\n",
        "print(\"Average Document Length (in characters):\", avg_length_chars)"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "LPyKL89hw9SZ"
      },
      "source": [
        "##Pre-processing"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "kN30flXUIy1J"
      },
      "source": [
        "The english stopwords is a package of 179 words that in general, would not help in a sentiment analysis problem. But, since they include terms that are negative, removing them could prove harmful for our case.\n",
        "\n",
        "e.g. imagine the phrase \"I didn't like the film\" to end up \"like film\".\n",
        "\n",
        "So, the plan is to remove all the stop words that include negative meaning before the preprocessing."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 11,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "wmU8jzwswobg",
        "outputId": "0b3cc531-2ca7-4937-9f8a-ad2c701c8ee3"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "[nltk_data] Downloading package stopwords to /root/nltk_data...\n",
            "[nltk_data]   Unzipping corpora/stopwords.zip.\n"
          ]
        },
        {
          "data": {
            "text/plain": [
              "['i',\n",
              " 'me',\n",
              " 'my',\n",
              " 'myself',\n",
              " 'we',\n",
              " 'our',\n",
              " 'ours',\n",
              " 'ourselves',\n",
              " 'you',\n",
              " \"you're\",\n",
              " \"you've\",\n",
              " \"you'll\",\n",
              " \"you'd\",\n",
              " 'your',\n",
              " 'yours',\n",
              " 'yourself',\n",
              " 'yourselves',\n",
              " 'he',\n",
              " 'him',\n",
              " 'his',\n",
              " 'himself',\n",
              " 'she',\n",
              " \"she's\",\n",
              " 'her',\n",
              " 'hers',\n",
              " 'herself',\n",
              " 'it',\n",
              " \"it's\",\n",
              " 'its',\n",
              " 'itself',\n",
              " 'they',\n",
              " 'them',\n",
              " 'their',\n",
              " 'theirs',\n",
              " 'themselves',\n",
              " 'what',\n",
              " 'which',\n",
              " 'who',\n",
              " 'whom',\n",
              " 'this',\n",
              " 'that',\n",
              " \"that'll\",\n",
              " 'these',\n",
              " 'those',\n",
              " 'am',\n",
              " 'is',\n",
              " 'are',\n",
              " 'was',\n",
              " 'were',\n",
              " 'be',\n",
              " 'been',\n",
              " 'being',\n",
              " 'have',\n",
              " 'has',\n",
              " 'had',\n",
              " 'having',\n",
              " 'do',\n",
              " 'does',\n",
              " 'did',\n",
              " 'doing',\n",
              " 'a',\n",
              " 'an',\n",
              " 'the',\n",
              " 'and',\n",
              " 'but',\n",
              " 'if',\n",
              " 'or',\n",
              " 'because',\n",
              " 'as',\n",
              " 'until',\n",
              " 'while',\n",
              " 'of',\n",
              " 'at',\n",
              " 'by',\n",
              " 'for',\n",
              " 'with',\n",
              " 'about',\n",
              " 'against',\n",
              " 'between',\n",
              " 'into',\n",
              " 'through',\n",
              " 'during',\n",
              " 'before',\n",
              " 'after',\n",
              " 'above',\n",
              " 'below',\n",
              " 'to',\n",
              " 'from',\n",
              " 'up',\n",
              " 'down',\n",
              " 'in',\n",
              " 'out',\n",
              " 'on',\n",
              " 'off',\n",
              " 'over',\n",
              " 'under',\n",
              " 'again',\n",
              " 'further',\n",
              " 'then',\n",
              " 'once',\n",
              " 'here',\n",
              " 'there',\n",
              " 'when',\n",
              " 'where',\n",
              " 'why',\n",
              " 'how',\n",
              " 'all',\n",
              " 'any',\n",
              " 'both',\n",
              " 'each',\n",
              " 'few',\n",
              " 'more',\n",
              " 'most',\n",
              " 'other',\n",
              " 'some',\n",
              " 'such',\n",
              " 'no',\n",
              " 'nor',\n",
              " 'not',\n",
              " 'only',\n",
              " 'own',\n",
              " 'same',\n",
              " 'so',\n",
              " 'than',\n",
              " 'too',\n",
              " 'very',\n",
              " 's',\n",
              " 't',\n",
              " 'can',\n",
              " 'will',\n",
              " 'just',\n",
              " 'don',\n",
              " \"don't\",\n",
              " 'should',\n",
              " \"should've\",\n",
              " 'now',\n",
              " 'd',\n",
              " 'll',\n",
              " 'm',\n",
              " 'o',\n",
              " 're',\n",
              " 've',\n",
              " 'y',\n",
              " 'ain',\n",
              " 'aren',\n",
              " \"aren't\",\n",
              " 'couldn',\n",
              " \"couldn't\",\n",
              " 'didn',\n",
              " \"didn't\",\n",
              " 'doesn',\n",
              " \"doesn't\",\n",
              " 'hadn',\n",
              " \"hadn't\",\n",
              " 'hasn',\n",
              " \"hasn't\",\n",
              " 'haven',\n",
              " \"haven't\",\n",
              " 'isn',\n",
              " \"isn't\",\n",
              " 'ma',\n",
              " 'mightn',\n",
              " \"mightn't\",\n",
              " 'mustn',\n",
              " \"mustn't\",\n",
              " 'needn',\n",
              " \"needn't\",\n",
              " 'shan',\n",
              " \"shan't\",\n",
              " 'shouldn',\n",
              " \"shouldn't\",\n",
              " 'wasn',\n",
              " \"wasn't\",\n",
              " 'weren',\n",
              " \"weren't\",\n",
              " 'won',\n",
              " \"won't\",\n",
              " 'wouldn',\n",
              " \"wouldn't\"]"
            ]
          },
          "execution_count": 11,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "import nltk\n",
        "nltk.download('stopwords')\n",
        "from nltk.corpus import stopwords\n",
        "\n",
        "stopwords.words('english')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "rqZSeeqHxiDg"
      },
      "source": [
        "From these words, we will decide which ones to keep because in fact they have a meaningful impact in our sentiment analysis problem, as we stated earlier."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 12,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "A1mKiosyxaNh",
        "outputId": "be562701-4178-445d-bbc5-f80256c1f99b"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "['not',\n",
              " \"don't\",\n",
              " \"aren't\",\n",
              " \"couldn't\",\n",
              " \"didn't\",\n",
              " \"doesn't\",\n",
              " \"hadn't\",\n",
              " \"hasn't\",\n",
              " \"shouldn't\",\n",
              " \"haven't\",\n",
              " \"wasn't\",\n",
              " \"weren't\",\n",
              " \"isn't\",\n",
              " 'doesn']"
            ]
          },
          "execution_count": 12,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "set_stop_words = set(stopwords.words('english'))\n",
        "\n",
        "to_keep_words = ['not', \"don't\", \"aren't\", \"couldn't\", \"didn't\", \"doesn't\", \"hadn't\", \"hasn't\" , \"shouldn't\", \"haven't\", \"wasn't\", \"weren't\",  \"isn't\", \"doesn\"]\n",
        "to_keep_words"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 13,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "2etFA9fWyE05",
        "outputId": "5da32c9e-8577-4305-ccbb-099a204eab4d"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "179\n",
            "14\n",
            "165\n"
          ]
        }
      ],
      "source": [
        "stopwords_updated = set(stopwords.words('english')) - set(to_keep_words)\n",
        "print(len(stopwords.words('english')))\n",
        "print(len(to_keep_words))\n",
        "print(len(stopwords_updated))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 14,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dKIgoYWCyOPw",
        "outputId": "10dd88b5-0f4f-4ae3-9d0b-8d2c588ea258"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "[nltk_data] Downloading package wordnet to /root/nltk_data...\n"
          ]
        }
      ],
      "source": [
        "from nltk.stem import WordNetLemmatizer\n",
        "import re\n",
        "from tqdm.auto import tqdm\n",
        "import string\n",
        "nltk.download('wordnet')\n",
        "\n",
        "def pre_process_text(text):\n",
        "    ''' Function to preprocess text.\n",
        "     input: initial text\n",
        "     output: processed text\n",
        "     Performs pre-processing methods:\n",
        "        1. Combination to a single document.\n",
        "        2. Convertion to lowercase.\n",
        "        3. Lemmatization and stop words extraction\n",
        "        4. Punctuation removal\n",
        "        5. Number removal\n",
        "        6. Single characters removal\n",
        "        7. Converting multiple spaces to single ones\n",
        "        '''\n",
        "    lemmatizer = WordNetLemmatizer()\n",
        "    all_docs = []\n",
        "\n",
        "    single_char = re.compile(r'\\s+[a-z]\\s+')                                          #6. Remove single characters\n",
        "    multiple_space= re.compile(r'\\s+')                                                 #7. Replace multiple space with a single one\n",
        "\n",
        "    stopwords_updated = set(stopwords.words('english')) - set(to_keep_words)\n",
        "    for document in tqdm(x):\n",
        "\n",
        "        combined_text = ' '.join(text)            #1.Combine in one single document\n",
        "\n",
        "        combined_text = combined_text.lower()    #2. Convert to lowercase\n",
        "        combined_text = [lemmatizer.lemmatize(word) for word in document.split() if word not in stopwords_updated]  # 3.Lemmatize and remove stop words\n",
        "\n",
        "\n",
        "        combined_text = ' '.join(combined_text)\n",
        "\n",
        "        combined_text = ''.join([char for char in combined_text if char not in string.punctuation])   #4.remove punctuation\n",
        "        combined_text = ''.join([char for char in combined_text if not char.isdigit()])     #5.remove numbers\n",
        "\n",
        "        res = single_char.sub(combined_text, '')\n",
        "        res2 = multiple_space.sub(combined_text, ' ')\n",
        "        all_docs.append(combined_text)\n",
        "\n",
        "    return all_docs"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 15,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 292,
          "referenced_widgets": [
            "1ee6b2099b5b4fa59d83ca4f8479a830",
            "0cf1af116cd1439ca7c38b4855d7e9a9",
            "07870395d12a4795a897c9c4e81bccaa",
            "8f9432d67fc74d4daafa43c7e144127b",
            "6c25a81bdd5b4979b98bee24bd613816",
            "0b289fb9c8e948a8af718fb65f78d975",
            "316615ccbf44426d818ff140331e6a69",
            "c0e082767739480da73bd1c54c6f16db",
            "f94d417290f44f0bbb844c652f8dfe93",
            "e6bb239862224171b616751c072677e7",
            "143c4bafc3d0415c8e1d435af380aea6"
          ]
        },
        "id": "r0lgYGgDyfEY",
        "outputId": "295fb02e-f346-43da-9db9-19e720713e5e"
      },
      "outputs": [
        {
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "1ee6b2099b5b4fa59d83ca4f8479a830",
              "version_major": 2,
              "version_minor": 0
            },
            "text/plain": [
              "  0%|          | 0/2000 [00:00<?, ?it/s]"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "text/plain": [
              "['arnold schwarzenegger icon action enthusiast  since late s  lately film sloppy oneliner getting worse  hard seeing arnold mr  freeze batman robin  especially say ton ice joke  hey got  million  whats matter  arnold signed another expensive blockbuster  cant compare like terminator series  true lie even eraser  called dark thriller  devil  gabriel byrne  come upon earth  impregnate woman  robin tunney  happens every  year  basically destroy world  apparently god chosen one man  one man jericho cane  arnold   help trusty sidekick  kevin pollack   stop nothing let devil take world  part actually absurd  would fit right dogma  yes  film weak  better blockbuster right  sleepy hollow   make world not enough look like  star film  anyway  definitely doesnt seem like arnold movie  wasnt type film see  sure gave u chuckle well known oneliner  seemed confused character film going  understandable  especially ending changed according source  aside form  still walked  much like past film  im sorry say arnold maybe end action day  speaking action  film  hardly explosion fight  devil made place explode  arnold wasnt kicking devil butt  ending changed make spiritual  undoubtedly ruined film  least hoping cool ending nothing else occurred  let  also dont know film took long cost much  really super affect  unless consider invisible devil   minute top  worth overpriced budget  budget gone better script  least audience could somewhat entertained instead facing boredom  pitiful see script like get bought made movie  even read thing anymore  sure doesnt seem like  thankfully gabriels performance gave light poor film  walk street searching robin tunney  cant help feel looked like devil  guy creepy looking anyway   glad end movie  dont bother see  expecting solid action flick  neither solid action  another movie suckered seeing  due strategic marketing campaign  save money see world not enough entertaining experience ']"
            ]
          },
          "execution_count": 15,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "processed_text = pre_process_text(x)\n",
        "\n",
        "processed_text[:1]"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "vsP4h-94y6RY"
      },
      "source": [
        "##Splitting into training set (70%), development set (15%) and test set (15%)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 16,
      "metadata": {
        "id": "w2GVixrAy0M5"
      },
      "outputs": [],
      "source": [
        "from sklearn.model_selection import train_test_split\n",
        "\n",
        "X_train, X_temp, y_train, y_temp = train_test_split(processed_text, y, test_size=0.3, random_state=17)\n",
        "X_dev, X_test, y_dev, y_test = train_test_split(X_temp, y_temp, test_size=0.5, random_state=25)\n",
        "\n",
        "training_text = ' '.join(X_train)                    #Flatten into a single string\n",
        "development_text = ' '.join(X_dev)\n",
        "test_text = ' '.join(X_test)\n",
        "\n",
        "training_words = training_text.split()\n",
        "development_words = development_text.split()\n",
        "test_words = test_text.split()\n",
        "\n",
        "training_vocab = set(training_words)\n",
        "development_vocab = set(development_words)\n",
        "test_vocab = set(test_words)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 17,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "nMqHNi20zPOf",
        "outputId": "1b5b0e3f-7074-42bb-caab-33385c5672cd"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Training set size (in documents):  1400\n",
            "Development set size (in documents):  300\n",
            "Test set size (in documents):  300\n",
            "Full size (sanity check):  2000\n",
            "---------------------------------\n",
            "Training vocabulary size (in words):  36624\n",
            "Development vocabulary size (in words):  16948\n",
            "Test vocabulary size (in words):  16780\n",
            "Full vocabulary size (in words):  70352\n"
          ]
        }
      ],
      "source": [
        "print(\"Training set size (in documents): \", len(y_train))\n",
        "print(\"Development set size (in documents): \", len(y_dev))\n",
        "print(\"Test set size (in documents): \", len(y_test))\n",
        "print(\"Full size (sanity check): \", len(y_train) + len(y_dev) + len(y_test))\n",
        "print(\"---------------------------------\")\n",
        "print(\"Training vocabulary size (in words): \" , len(training_vocab))\n",
        "print(\"Development vocabulary size (in words): \", len(development_vocab))\n",
        "print(\"Test vocabulary size (in words): \", len(test_vocab))\n",
        "print(\"Full vocabulary size (in words): \", len(training_vocab) + len(development_vocab) + len(test_vocab))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 18,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "8H2cQKI4zbJw",
        "outputId": "367579c9-9545-4999-99cc-436b217283fb"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "{'same', 'becomes', 'even', 'besides', 'though', 'being', 'why', 'ourselves', 'indeed', 'this', 'could', 'such', 'can', 'during', '’ll', 'get', 'each', 'behind', 'if', 'anyway', 'meanwhile', 'n‘t', 'what', 'must', 'itself', \"'ll\", 'another', 'our', 'although', 'keep', '’ve', 'first', '’m', 'would', 'anyhow', 'without', 'does', 'move', 'now', 'once', 'thereafter', 'either', 'then', 'through', 'will', 'below', 'however', 'these', 'within', 'nine', 'much', 'thence', 'become', 'front', 'three', 'everything', 'beyond', 'latter', 'latterly', 'mostly', 'yourselves', 'might', 'up', 'anywhere', 'us', 'somewhere', 'amongst', 'enough', 'too', 'several', 'regarding', 'whereupon', 'i', 'whence', 'part', 'sixty', 'as', 'whereby', 'whether', 'still', 'others', 'yet', 'not', 'seeming', \"'s\", 'just', 'him', 'thru', 'you', 'whither', 're', '’d', 'hereby', 'in', 'done', 'various', 'whenever', 'be', 'herself', 'every', 'side', 'who', 'yours', 'about', 'only', 'while', 'thus', 'say', 'wherever', 'and', 'else', 'one', 'nothing', 'wherein', \"'d\", 'noone', 'except', 'whole', 'also', 'twenty', 'which', 'so', 'should', 'few', 'again', 'moreover', 'seems', 'across', 'twelve', \"'m\", 'have', 'hence', 'go', 'own', 'well', 'make', 'he', 'are', 'former', 'is', 'them', 'onto', '‘ll', 'am', 'never', 'since', 'been', 'third', 'no', 'name', 'it', 'became', 'show', 'the', 'towards', 'around', 'hers', 'amount', 'seem', 'perhaps', 'a', 'may', 'all', 'where', 'by', 'serious', 'nobody', 'fifteen', 'everywhere', 'formerly', 'himself', 'see', '‘m', 'of', 'seemed', 'full', '‘re', 'their', '‘d', 'otherwise', 'his', 'next', 'she', 'more', 'further', 'empty', 'least', 'whoever', 'forty', 'at', 'due', 'top', 'or', 'her', 'eleven', 'doing', 'anyone', 'ten', 'herein', '‘s', 'rather', 'using', 'six', 'thereupon', 'my', 'two', 'when', 'give', 'hereupon', 'elsewhere', 'down', 'mine', 'over', 'any', 'to', 'themselves', 'against', 'out', 'than', 'me', 'toward', 'between', 'upon', 'therein', 'but', 'whom', 'did', 'beside', 'under', 'because', 'most', 'on', 'ours', 'had', 'made', 'take', 'among', 'that', 'very', 'call', 'above', 'last', 'often', 'into', 'beforehand', 'hundred', 'were', 'until', 'somehow', \"n't\", 'whose', 'has', 'before', 'becoming', 'many', 'none', 'put', 'those', 'whereafter', 'back', '’s', 'something', 'unless', 'already', 'via', 'almost', 'everyone', \"'re\", 'therefore', 'namely', 'eight', \"'ve\", 'n’t', 'ever', 'was', 'someone', 'fifty', 'less', 'along', 'cannot', 'there', 'its', 'anything', 'myself', 'for', 'yourself', 'quite', 'neither', 'used', 'we', 'with', 'four', 'whatever', 'your', 'do', 'together', 'how', 'off', 'sometimes', 'nevertheless', 'ca', 'here', 'thereby', '’re', 'whereas', '‘ve', 'an', 'alone', 'per', 'really', 'bottom', 'throughout', 'both', 'other', 'hereafter', 'some', 'nor', 'five', 'nowhere', 'always', 'please', 'after', 'sometime', 'afterwards', 'they', 'from'}\n",
            "326\n"
          ]
        },
        {
          "data": {
            "text/plain": [
              "['not',\n",
              " \"don't\",\n",
              " \"aren't\",\n",
              " \"couldn't\",\n",
              " \"didn't\",\n",
              " \"doesn't\",\n",
              " \"hadn't\",\n",
              " \"hasn't\",\n",
              " \"shouldn't\",\n",
              " \"haven't\",\n",
              " \"wasn't\",\n",
              " \"weren't\",\n",
              " \"isn't\",\n",
              " 'doesn',\n",
              " \"n't\",\n",
              " 'not',\n",
              " 'no']"
            ]
          },
          "execution_count": 18,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "import spacy\n",
        "from spacy.lang.en.stop_words import STOP_WORDS\n",
        "\n",
        "print(STOP_WORDS)\n",
        "print(len(STOP_WORDS))\n",
        "extra_to_keep = [\"n't\", \"not\", \"no\"]\n",
        "for i in range(len(extra_to_keep)):\n",
        "  to_keep_words.append(extra_to_keep[i])\n",
        "\n",
        "to_keep_words\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 19,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "qzDTTrtX4A7J",
        "outputId": "c30e423d-65cd-4181-b6b5-3d6e7dfa2ca6"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "323\n"
          ]
        }
      ],
      "source": [
        "stop_words_updated = STOP_WORDS - set(to_keep_words)\n",
        "print(len(stop_words_updated))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": null,
      "metadata": {},
      "outputs": [],
      "source": [
        "!python -m spacy download en_core_web_sm"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 20,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "dqiO44uW5Cfh",
        "outputId": "cb2eac09-e2f2-4ee6-e5ff-ccfc1c8be505"
      },
      "outputs": [
        {
          "data": {
            "text/plain": [
              "<spacy.pipeline.sentencizer.Sentencizer at 0x7f4c50122fc0>"
            ]
          },
          "execution_count": 20,
          "metadata": {},
          "output_type": "execute_result"
        }
      ],
      "source": [
        "nlp = spacy.load('en_core_web_sm',disable=[\"tagger\", \"parser\",\"ner\"])\n",
        "nlp.add_pipe('sentencizer')"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "ggXTl3Ff4VWQ"
      },
      "source": [
        "##Use spaCy for sentence splitting & tokenization"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 21,
      "metadata": {
        "id": "imPBqgCy4K4x"
      },
      "outputs": [],
      "source": [
        "def tokenize_sent(x):\n",
        "\n",
        "  X_tokenized = []\n",
        "  for idx in tqdm(range(len(x))):\n",
        "    doc = nlp(x[idx])\n",
        "    tokens = []\n",
        "    for sent in doc.sents:\n",
        "      for tok in sent:\n",
        "        if '\\n' in tok.text or \"\\t\" in tok.text or \"--\" in tok.text or \"*\" in tok.text or tok.text.lower() in stop_words_updated or tok.text in string.punctuation or all(x in string.punctuation for x in tok.text):\n",
        "          continue\n",
        "        if tok.text.strip():\n",
        "          tokens.append(tok.text.replace('\"',\"'\").strip().lower())\n",
        "    X_tokenized.append(tokens)\n",
        "  return X_tokenized"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 22,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 168,
          "referenced_widgets": [
            "3d712ec28c9c405ab625ad74ca72c7a4",
            "c69fc550f2344cac92607cc70c1fbd30",
            "649916d14e044e16ac04051feb5cefe3",
            "f814c9373cad46a3a77f3961d783cddb",
            "37d787fac5874a8aa0342b73a6d7c3df",
            "f0659f1993e34924817f779df5761b41",
            "da534940a0234c4497f484fa36c2c86a",
            "d75197b7aaf44523b467e4a19bb4bbf9",
            "37bedfb3159948aa9296986647813cc4",
            "7d12bfacf7f34e5684a2ad21a609ff9b",
            "300e1ca78baa4457b906f58170b19d06",
            "adfe3d33b08f40348508cfa717d0008c",
            "94655d0a32e6440fb5bfd0853de28d97",
            "41500907f8724730b3ac15a7f060d32c",
            "6ab52aa715f7473ebf7bada52805c3c9",
            "151b618982024df2bf250c08f2fdc3e6",
            "174ceef81b444b4b9071ad658a5ace66",
            "b596dcc695414e2f951b34e911f93970",
            "cf8516cba4064213a7977b9a62e4c523",
            "3f13e910892e4b3b8dffdc546f9d6692",
            "5f44aca6e17e40a888a2aab692aae14b",
            "af6c19f76e58484a8ccbcda95056debc",
            "bff56443f4ac4861977a3a7f2e66abd3",
            "e017afc5b5c847c4a3360bf5bb0d1f6e",
            "79ca2685fda64168aff4ed04bd7f47eb",
            "f2a2ab3d7a804129a2ca662875e7fc04",
            "0f2a5d2bfa024fceb2a809017ab4713d",
            "694294f380574008a713a7e803f60a98",
            "dbe786c66e134f44b67a5e4b63aca3d5",
            "e6e1a14c25f8468c9c5dedcd6d184860",
            "3278459baebc4a2b87430120b9563eb2",
            "5c1e1bf2b3c84d149d0bb8937d087246",
            "bd44fcb7420d4298b782a6601df0472d"
          ]
        },
        "id": "JUurnAiE5y55",
        "outputId": "01896941-ea48-41c3-80a4-747f80079983"
      },
      "outputs": [
        {
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "3d712ec28c9c405ab625ad74ca72c7a4",
              "version_major": 2,
              "version_minor": 0
            },
            "text/plain": [
              "  0%|          | 0/1400 [00:00<?, ?it/s]"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/spacy/pipeline/lemmatizer.py:211: UserWarning: [W108] The rule-based lemmatizer did not find POS annotation for one or more tokens. Check that your pipeline includes components that assign token.pos, typically 'tagger'+'attribute_ruler' or 'morphologizer'.\n",
            "  warnings.warn(Warnings.W108)\n"
          ]
        },
        {
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "adfe3d33b08f40348508cfa717d0008c",
              "version_major": 2,
              "version_minor": 0
            },
            "text/plain": [
              "  0%|          | 0/300 [00:00<?, ?it/s]"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "application/vnd.jupyter.widget-view+json": {
              "model_id": "bff56443f4ac4861977a3a7f2e66abd3",
              "version_major": 2,
              "version_minor": 0
            },
            "text/plain": [
              "  0%|          | 0/300 [00:00<?, ?it/s]"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "X_train_tokenized = tokenize_sent(X_train)\n",
        "X_dev_tokenized = tokenize_sent(X_dev)\n",
        "X_test_tokenized = tokenize_sent(X_test)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 23,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "EQzno_K4hU0r",
        "outputId": "bdde50fb-96a8-45cc-e192-547d1f84e10e"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Total word length before tokenization: [1910, 4209, 3093, 1382, 1574, 1239, 4053, 1042, 6206, 2550, 2693, 4096, 3678, 2388, 943, 2316, 2228, 4459, 4586, 2112, 2227, 2743, 1606, 791, 2321, 2903, 2693, 4476, 3583, 2241, 2490, 1719, 2804, 3176, 1336, 800, 5545, 2102, 2411, 4811, 2947, 2195, 1294, 1774, 2754, 1520, 3547, 1888, 2094, 1584, 4735, 1274, 2724, 3922, 1968, 1124, 2736, 3521, 2311, 2329, 2434, 4925, 1544, 2104, 935, 746, 1455, 593, 2325, 3718, 1768, 2206, 2989, 1180, 3287, 1359, 2650, 2204, 2339, 5212, 5447, 2112, 3601, 2824, 1912, 2388, 1482, 2185, 5191, 2883, 2981, 2002, 3331, 1782, 2157, 2141, 2590, 3800, 3393, 1933, 2925, 2294, 3405, 2047, 1251, 2903, 2111, 2902, 2805, 3191, 2118, 2706, 1877, 1000, 1259, 3983, 4724, 2286, 3712, 1555, 2490, 1256, 2542, 2394, 2488, 3550, 3433, 553, 2682, 2831, 941, 3457, 1049, 5903, 3145, 3218, 2378, 2035, 2661, 2495, 1958, 2728, 2125, 2924, 2159, 3182, 5699, 1172, 2943, 866, 667, 3179, 3595, 2169, 1539, 1868, 3993, 2336, 2226, 2436, 1634, 3292, 1746, 2510, 1828, 3609, 2696, 2959, 3035, 2162, 3749, 1306, 1820, 3709, 1937, 1251, 3789, 5213, 2301, 4402, 6045, 4612, 3294, 1965, 5337, 1834, 3045, 2461, 2239, 3473, 3102, 2896, 2171, 1480, 1807, 4032, 2611, 4450, 4420, 1900, 2349, 2308, 984, 2880, 2913, 3608, 4107, 1841, 1837, 4601, 1603, 2753, 4141, 1778, 3537, 2173, 2295, 919, 1847, 4775, 3923, 2313, 2009, 1717, 1940, 2361, 7195, 1708, 1662, 56, 3263, 2443, 1629, 2104, 1406, 3158, 2261, 1511, 2450, 1299, 1749, 1492, 3528, 1297, 2395, 5254, 2184, 1362, 1628, 2211, 1623, 1422, 1508, 4136, 3335, 1533, 1917, 1592, 2708, 1691, 3111, 762, 8513, 3143, 2176, 2555, 1953, 1317, 1940, 1878, 1877, 1451, 2577, 3062, 1766, 1131, 1192, 4542, 3541, 3118, 3037, 1998, 1969, 1973, 2000, 3548, 2761, 2646, 1990, 1256, 3962, 4467, 3851, 3321, 5291, 2552, 1882, 2625, 3686, 2857, 1912, 2245, 2187, 2715, 2090, 1622, 2356, 1513, 3356, 3022, 4334, 2716, 830, 2766, 3692, 1398, 2845, 1787, 2183, 1513, 3760, 5357, 4889, 2284, 1658, 2967, 3855, 2080, 1346, 2052, 2023, 2125, 2457, 2849, 3614, 2035, 2850, 2168, 930, 1817, 2718, 1130, 2592, 2976, 3836, 2517, 1640, 2886, 3256, 2632, 1996, 1739, 2915, 1404, 2528, 3502, 2605, 639, 3393, 1755, 2467, 4687, 2070, 2772, 3038, 2578, 4127, 4263, 4204, 3010, 1790, 2149, 3396, 4222, 834, 2353, 1256, 2998, 2334, 1953, 3232, 2827, 3372, 2412, 2019, 1778, 1720, 3643, 2539, 1167, 4183, 2123, 1690, 2328, 3065, 1129, 2000, 2395, 3278, 1639, 1453, 1044, 4470, 4000, 2291, 2839, 2820, 1647, 2695, 2902, 1920, 1467, 3482, 3061, 907, 2147, 5013, 2068, 2300, 1793, 1640, 1833, 2930, 2252, 2443, 4770, 2727, 3471, 4125, 2184, 3138, 1942, 3050, 3183, 3257, 3193, 2182, 2696, 1620, 4430, 2971, 1364, 2833, 3842, 5301, 1428, 3293, 2590, 2904, 3273, 1618, 4001, 2250, 4396, 1859, 2365, 2376, 1868, 2458, 2719, 2763, 2277, 1873, 2067, 3105, 2734, 2911, 3518, 1461, 1498, 2873, 2765, 2280, 3677, 1564, 1279, 3381, 1714, 2747, 1612, 2165, 1949, 2766, 1322, 3739, 2516, 1727, 4239, 1079, 2140, 2690, 2216, 1696, 3916, 841, 1281, 3491, 3055, 2300, 3554, 2776, 3226, 1945, 3076, 2591, 3526, 1855, 2398, 2667, 2446, 3273, 1880, 1432, 3878, 2937, 3144, 2927, 1571, 2532, 2297, 2884, 3797, 2858, 2199, 1846, 1184, 2536, 2675, 1526, 2621, 2765, 2575, 2610, 2512, 3096, 4456, 1426, 1829, 1865, 1965, 2272, 2080, 3517, 3359, 2189, 3538, 1050, 1597, 3646, 3890, 1739, 1001, 1244, 2706, 1706, 1984, 2768, 648, 2733, 1469, 1566, 3741, 3715, 4252, 1669, 3537, 3335, 509, 2713, 1908, 1143, 3500, 3204, 2497, 3064, 1278, 3257, 2285, 2772, 2273, 2501, 1895, 1464, 4776, 3229, 1505, 1569, 2981, 3695, 2004, 3797, 1151, 3322, 2577, 2105, 2183, 2625, 2248, 3060, 2487, 3109, 1326, 3478, 842, 1428, 2144, 2664, 2496, 798, 3822, 2544, 1831, 2192, 3041, 1673, 826, 2946, 2711, 2312, 1974, 4182, 2424, 5579, 3564, 2095, 2444, 2817, 6027, 4996, 1326, 1334, 5085, 4483, 2394, 2967, 2776, 2632, 3056, 1721, 2094, 1449, 2833, 2323, 3070, 2165, 3300, 2718, 2351, 2964, 2386, 2094, 2638, 2476, 3303, 2816, 2562, 2678, 3086, 2475, 3444, 2206, 974, 2336, 2732, 2545, 1263, 2870, 3026, 2497, 1910, 4035, 1388, 3024, 1703, 3238, 1425, 2997, 2901, 970, 2593, 2528, 1782, 1632, 4127, 2456, 3255, 2298, 2111, 3182, 2361, 3703, 2290, 3846, 3609, 3453, 1208, 2598, 3049, 1422, 1528, 2685, 1668, 2920, 3142, 2342, 3019, 1554, 2220, 2821, 3335, 3130, 2467, 2544, 1690, 2017, 4513, 2596, 2322, 2130, 6853, 2163, 2460, 3380, 1307, 3548, 2299, 2506, 3837, 2456, 4062, 1641, 2813, 4096, 2773, 2340, 2015, 5545, 1896, 3494, 3596, 2747, 2195, 2259, 2239, 840, 2031, 2645, 1861, 1555, 1900, 3175, 2697, 1680, 1415, 1854, 1583, 1095, 2177, 2821, 1559, 5709, 3412, 3581, 5972, 3850, 3064, 1965, 2486, 3766, 1989, 2552, 2557, 3201, 1153, 3214, 3105, 1625, 1798, 2145, 1807, 2006, 2775, 2133, 2910, 2135, 787, 3024, 2439, 3716, 1823, 2063, 3502, 1810, 4137, 3260, 1742, 1050, 1735, 2065, 2051, 1438, 2111, 2220, 2179, 2953, 1727, 3592, 1682, 1894, 3065, 2945, 1810, 2877, 3126, 2063, 1429, 4186, 6136, 2320, 1676, 2413, 1776, 3113, 3342, 953, 1314, 2181, 2914, 1775, 2184, 2303, 10678, 4371, 5296, 3828, 5079, 2802, 2244, 928, 4931, 4658, 1741, 3130, 2252, 4618, 3241, 3418, 1418, 4631, 2528, 1313, 1659, 1811, 2955, 1239, 1299, 1221, 3902, 2167, 4018, 1779, 2687, 2696, 882, 1462, 2582, 2563, 3138, 1503, 768, 2956, 2918, 2206, 1456, 4844, 2189, 1559, 4361, 1544, 1818, 1627, 2310, 5002, 3190, 6845, 2689, 1407, 878, 2202, 1351, 5024, 2985, 2683, 1934, 2101, 894, 3240, 2489, 2559, 2705, 1753, 1487, 2586, 1837, 3140, 1442, 2949, 3792, 2842, 1784, 1774, 1440, 3066, 2307, 1967, 2478, 2208, 5353, 3474, 2296, 3597, 916, 3921, 2932, 2790, 2059, 587, 4628, 1654, 2877, 2052, 2327, 3407, 2156, 2816, 2448, 3317, 2126, 2270, 4166, 2022, 1625, 2459, 2797, 2677, 1520, 2890, 2467, 1036, 2744, 898, 1256, 1936, 2351, 972, 967, 2461, 1624, 3048, 2124, 5797, 2791, 2352, 1418, 1659, 4102, 2228, 2548, 1847, 4141, 2351, 868, 1852, 2430, 3662, 1946, 1680, 4688, 3376, 3254, 3334, 2636, 1860, 1125, 1359, 3115, 1911, 2127, 3812, 2982, 3473, 3809, 2944, 1745, 1372, 3579, 4347, 2857, 1908, 4399, 2418, 1945, 1592, 1124, 2856, 2985, 1687, 3297, 1260, 4915, 5111, 2369, 1840, 2098, 2503, 2012, 1027, 3231, 1600, 1275, 2547, 1123, 2034, 1904, 1243, 5098, 2670, 2304, 3420, 5034, 1920, 2130, 4172, 2696, 2212, 5167, 3214, 1598, 2677, 3684, 1579, 2338, 1034, 3335, 2763, 2870, 3653, 1799, 2144, 2650, 2949, 3311, 3458, 2559, 2293, 2337, 1484, 4405, 1234, 2961, 5940, 2055, 701, 1290, 2046, 3122, 1858, 2005, 2416, 3250, 1278, 3237, 1251, 3025, 3050, 2083, 3567, 1272, 938, 2290, 1196, 3167, 2796, 2114, 2785, 1970, 1925, 1704, 1700, 2520, 1414, 5759, 632, 1426, 1966, 2263, 1507, 1250, 2100, 1737, 2349, 2651, 2839, 2546, 4722, 5547, 1354, 3587, 3386, 2592, 1735, 935, 3312, 1719, 3536, 3712, 2125, 3544, 5471, 2456, 2024, 7426, 2959, 1844, 1561, 2798, 2017, 3094, 1495, 2750, 3283, 2124, 3031, 2850, 1576, 4090, 2024, 2436, 1857, 1411, 2535, 1759, 2475, 2607, 1067, 2815, 3805, 1296, 4550, 2686, 1882, 1938, 2894, 2197, 2835, 1221, 2057, 3440, 3045, 2915, 3405, 2015, 3125, 3446, 2991, 2442, 1992, 1594, 2575, 3756, 3065, 1764, 3266, 1382, 308, 3651, 674, 1234, 2264, 3103, 5162, 4938, 1903, 2000, 1030, 3003, 3185, 1344, 3437, 1856, 2235, 1824, 3583, 3668, 1944, 2946, 1087, 901, 2432, 3891, 1532, 2966, 4989, 1990, 2219, 2614, 2707, 1056, 1502, 1934, 3928, 2534, 2561, 1806, 3979, 2461, 1440, 2133, 1560, 1808, 1320, 1876, 1063, 3600, 976, 2281, 1832, 1150, 1295, 3579, 2172, 2141, 4015, 2613, 2361, 2257, 2152, 3628, 2028, 944, 5781, 3951, 2520, 5461, 3954, 3015, 2551, 1890, 2499, 2697, 1608, 3880, 8411, 3574, 2560, 1924, 1954, 3352, 1964, 2779, 2930, 2327, 1408, 1863, 2642, 2497, 5572, 3532, 1847, 2457, 2133, 1766, 2141, 1664, 2640, 6099, 3385, 2842, 2014, 3890, 1815, 2408, 5103, 1605, 2518, 858, 1841, 2107, 4293, 2495, 1763, 2535, 1075, 1341, 1571, 2637, 2409, 1367, 1761, 1756, 2699, 3079, 2154, 4417, 1170, 785, 2773, 2111, 1310, 1747, 1799, 5750, 2138, 1092, 2333, 2900, 2755, 1773, 3926, 2926, 2308, 2309, 2155, 2370, 1557, 4451, 428, 1024, 7080, 1355, 1466, 3911, 4970, 1537, 1962, 2307, 2683, 2135, 2304, 3449, 1310, 3894, 3401, 3212, 4311, 1275, 3030, 3184, 2539, 3442, 5213, 4217, 1545, 4188, 3277, 2126, 2344, 1520, 2483, 3458, 1620, 3660, 3322, 3005, 3500, 1043, 1692, 2829, 4535, 3053, 2618, 1990, 2111, 3487, 2501, 2048, 780, 2824, 4349, 2172, 3268, 1330, 1763, 1939, 1858, 1457, 3631, 2184, 2917, 1429, 2587, 3066, 3499, 4415, 2460, 5539, 2272, 2455, 2403, 3091]\n",
            "Total word length after tokenization: [230, 506, 373, 177, 180, 148, 480, 130, 704, 317, 332, 475, 425, 276, 111, 297, 273, 530, 506, 277, 273, 335, 194, 99, 264, 349, 306, 483, 423, 269, 297, 192, 333, 390, 150, 102, 667, 252, 297, 556, 353, 258, 156, 219, 340, 176, 422, 235, 246, 208, 555, 170, 337, 488, 246, 144, 317, 413, 274, 272, 284, 636, 177, 248, 114, 91, 186, 73, 273, 455, 207, 289, 351, 145, 406, 163, 314, 256, 305, 675, 661, 245, 447, 349, 239, 289, 175, 266, 661, 323, 358, 247, 419, 205, 252, 259, 320, 440, 423, 238, 346, 288, 413, 250, 142, 373, 247, 354, 356, 371, 269, 316, 234, 130, 165, 451, 575, 278, 469, 188, 325, 156, 299, 296, 298, 464, 416, 68, 297, 356, 122, 437, 128, 731, 383, 358, 276, 251, 357, 322, 243, 323, 277, 357, 251, 385, 668, 141, 337, 100, 76, 381, 430, 264, 191, 227, 474, 275, 274, 295, 189, 396, 210, 303, 236, 447, 329, 352, 352, 281, 444, 163, 206, 472, 221, 161, 445, 603, 292, 548, 733, 555, 407, 240, 621, 223, 358, 293, 260, 422, 388, 368, 255, 178, 225, 506, 327, 538, 530, 241, 286, 282, 123, 352, 349, 420, 493, 213, 217, 565, 180, 360, 493, 207, 434, 257, 278, 110, 228, 567, 457, 275, 243, 208, 231, 296, 829, 205, 205, 8, 365, 296, 203, 251, 175, 366, 262, 189, 280, 154, 213, 185, 414, 157, 302, 608, 259, 166, 192, 251, 198, 166, 191, 512, 404, 197, 244, 198, 320, 200, 358, 94, 1044, 379, 262, 315, 240, 168, 240, 227, 229, 164, 302, 369, 203, 133, 153, 537, 443, 388, 364, 242, 241, 258, 228, 423, 333, 323, 260, 158, 482, 548, 456, 389, 634, 309, 234, 288, 442, 349, 246, 274, 266, 338, 248, 196, 273, 182, 405, 371, 540, 365, 98, 331, 439, 156, 340, 237, 282, 199, 440, 657, 593, 279, 194, 363, 450, 239, 167, 235, 255, 257, 301, 331, 429, 249, 337, 246, 118, 239, 326, 134, 332, 351, 468, 290, 199, 343, 394, 330, 247, 201, 379, 170, 320, 415, 306, 79, 420, 216, 313, 563, 235, 353, 360, 317, 443, 524, 495, 366, 220, 243, 418, 530, 95, 288, 151, 362, 290, 220, 375, 346, 422, 283, 257, 212, 209, 426, 309, 144, 497, 261, 203, 281, 353, 138, 243, 283, 363, 199, 177, 136, 542, 492, 272, 354, 342, 212, 315, 353, 240, 180, 422, 388, 105, 277, 572, 248, 275, 230, 207, 208, 362, 266, 301, 573, 330, 421, 480, 246, 362, 226, 349, 382, 397, 377, 276, 337, 195, 537, 382, 178, 333, 456, 609, 173, 390, 308, 361, 410, 193, 475, 276, 488, 216, 303, 283, 232, 291, 302, 336, 265, 239, 242, 383, 339, 363, 448, 186, 187, 334, 336, 275, 455, 187, 156, 390, 208, 341, 193, 268, 241, 324, 157, 448, 330, 206, 509, 138, 271, 316, 255, 196, 469, 106, 148, 397, 351, 275, 433, 314, 406, 222, 367, 317, 429, 221, 301, 294, 300, 406, 229, 182, 454, 350, 373, 363, 185, 303, 266, 336, 454, 344, 274, 221, 153, 306, 340, 196, 327, 331, 326, 325, 317, 367, 518, 184, 237, 237, 250, 266, 250, 429, 390, 259, 445, 135, 199, 461, 469, 207, 120, 158, 331, 204, 240, 347, 73, 319, 179, 185, 444, 432, 509, 203, 404, 394, 70, 322, 226, 145, 439, 375, 320, 371, 163, 367, 287, 335, 254, 282, 232, 181, 544, 389, 185, 204, 367, 430, 235, 476, 147, 403, 316, 253, 257, 331, 265, 361, 300, 372, 156, 420, 100, 178, 253, 315, 305, 95, 452, 327, 227, 270, 357, 214, 91, 346, 342, 284, 250, 489, 304, 674, 455, 252, 304, 323, 715, 610, 162, 168, 634, 538, 297, 368, 325, 300, 377, 214, 253, 183, 320, 280, 359, 266, 385, 336, 295, 355, 277, 249, 301, 314, 373, 342, 309, 326, 375, 273, 424, 276, 124, 286, 313, 315, 154, 363, 370, 311, 227, 490, 167, 386, 203, 387, 163, 367, 364, 123, 313, 319, 206, 208, 497, 303, 367, 281, 260, 384, 294, 439, 273, 447, 445, 444, 144, 311, 381, 166, 185, 337, 200, 354, 372, 278, 371, 195, 280, 355, 388, 377, 300, 316, 217, 243, 552, 287, 287, 269, 826, 257, 301, 425, 160, 425, 264, 295, 442, 280, 467, 209, 329, 482, 324, 278, 241, 683, 229, 411, 444, 334, 257, 257, 281, 97, 243, 324, 211, 189, 226, 386, 335, 210, 174, 232, 185, 131, 260, 342, 181, 692, 417, 410, 700, 483, 373, 242, 273, 450, 277, 300, 300, 405, 143, 385, 367, 192, 215, 251, 232, 241, 339, 257, 351, 257, 98, 362, 284, 460, 223, 243, 444, 220, 487, 407, 204, 133, 214, 238, 260, 172, 259, 264, 248, 380, 213, 410, 194, 235, 372, 351, 219, 337, 369, 222, 169, 483, 735, 281, 208, 286, 202, 397, 406, 120, 161, 262, 356, 222, 306, 282, 1289, 531, 618, 460, 639, 327, 274, 119, 578, 540, 218, 365, 292, 556, 387, 408, 177, 557, 323, 165, 213, 218, 354, 151, 160, 146, 471, 249, 496, 224, 313, 366, 105, 180, 304, 325, 387, 185, 101, 351, 361, 250, 173, 577, 258, 199, 508, 195, 217, 196, 289, 587, 408, 823, 330, 175, 101, 281, 168, 610, 387, 323, 223, 247, 122, 370, 303, 312, 333, 208, 182, 306, 224, 364, 175, 386, 455, 336, 226, 221, 170, 365, 267, 243, 314, 248, 658, 421, 284, 454, 114, 488, 363, 333, 250, 72, 576, 197, 343, 235, 262, 406, 262, 351, 305, 414, 257, 256, 502, 228, 190, 286, 334, 313, 194, 349, 297, 133, 339, 112, 155, 230, 307, 120, 118, 290, 189, 371, 282, 717, 341, 282, 164, 205, 566, 272, 330, 224, 497, 289, 105, 214, 288, 457, 240, 208, 566, 408, 401, 408, 321, 227, 132, 175, 389, 235, 254, 453, 365, 416, 449, 375, 220, 153, 418, 517, 328, 237, 552, 297, 225, 195, 126, 348, 371, 206, 389, 153, 631, 595, 291, 224, 260, 314, 244, 127, 394, 186, 155, 309, 141, 235, 230, 159, 601, 317, 284, 424, 582, 239, 263, 467, 342, 256, 621, 384, 202, 306, 444, 183, 276, 129, 419, 364, 346, 433, 239, 251, 334, 362, 415, 400, 314, 270, 286, 169, 503, 146, 373, 728, 249, 94, 159, 268, 390, 234, 242, 304, 366, 163, 398, 150, 358, 376, 255, 436, 175, 120, 279, 150, 402, 328, 250, 330, 238, 214, 211, 199, 299, 172, 690, 84, 170, 243, 285, 186, 167, 266, 212, 279, 328, 344, 306, 555, 671, 161, 424, 423, 314, 204, 116, 397, 200, 404, 437, 262, 435, 686, 303, 277, 926, 350, 208, 191, 342, 244, 364, 174, 354, 399, 254, 346, 341, 193, 488, 242, 282, 225, 177, 325, 211, 301, 320, 124, 346, 477, 166, 533, 333, 228, 228, 368, 257, 327, 146, 255, 423, 357, 332, 417, 244, 365, 416, 356, 291, 251, 202, 318, 465, 366, 213, 407, 173, 37, 433, 79, 165, 281, 377, 631, 584, 228, 265, 117, 359, 383, 157, 405, 226, 278, 231, 427, 439, 240, 365, 133, 120, 282, 439, 202, 378, 597, 235, 265, 302, 329, 120, 193, 227, 475, 303, 312, 218, 485, 311, 169, 266, 184, 221, 177, 226, 127, 438, 122, 283, 243, 138, 149, 422, 269, 272, 484, 313, 274, 248, 261, 458, 241, 121, 712, 457, 307, 655, 449, 372, 332, 228, 294, 328, 203, 496, 1045, 419, 296, 237, 251, 400, 253, 345, 352, 287, 179, 229, 343, 308, 653, 407, 232, 299, 265, 225, 281, 198, 307, 738, 415, 346, 244, 463, 206, 308, 609, 196, 322, 107, 227, 256, 525, 285, 212, 301, 127, 162, 191, 306, 298, 181, 230, 195, 324, 384, 257, 578, 153, 98, 318, 257, 158, 216, 228, 704, 271, 125, 269, 352, 326, 223, 475, 357, 292, 273, 268, 297, 187, 551, 48, 141, 830, 172, 185, 477, 596, 176, 248, 271, 331, 241, 270, 439, 154, 444, 406, 393, 504, 153, 355, 371, 314, 394, 630, 518, 200, 515, 393, 255, 295, 183, 297, 410, 205, 442, 413, 353, 420, 118, 200, 330, 521, 390, 322, 244, 259, 438, 297, 236, 92, 342, 513, 260, 389, 160, 209, 242, 234, 185, 467, 259, 373, 184, 329, 363, 429, 558, 312, 662, 268, 298, 290, 353]\n",
            "-----------------------------------------------------------\n",
            "Average word length before tokenization: 2586.9\n",
            "Average word length after tokenization: 312.97285714285715\n"
          ]
        }
      ],
      "source": [
        "import numpy as np\n",
        "\n",
        "print(\"Total word length before tokenization:\", [len(x) for x in X_train])\n",
        "print(\"Total word length after tokenization:\", [len(x) for x in X_train_tokenized])\n",
        "\n",
        "print(\"-----------------------------------------------------------\")\n",
        "\n",
        "print(\"Average word length before tokenization:\", np.mean([len(x) for x in X_train]))\n",
        "print(\"Average word length after tokenization:\", np.mean([len(x) for x in X_train_tokenized]))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 24,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "fi0bHmHx6ayR",
        "outputId": "6bf79870-02e0-4902-8a5e-92e32fdcc606"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Mean of sequence length on training set: 312.97285714285715\n",
            "Standard deviation of sequence length on training set: 134.5485652324931\n",
            "-----------------------------------------------------------\n",
            "Mean of sequence length on development set: 315.10333333333335\n",
            "Standard deviation of sequence length on development set: 139.33664027176133\n",
            "-----------------------------------------------------------\n",
            "Mean of sequence length on test set: 305.0566666666667\n",
            "Standard deviation of sequence length on test set: 131.16630711513616\n"
          ]
        }
      ],
      "source": [
        "import numpy as np\n",
        "\n",
        "# Get mean and std of sequence length\n",
        "train_mean = np.mean([len(x) for x in X_train_tokenized])\n",
        "train_std = np.std([len(x) for x in X_train_tokenized])\n",
        "print(\"Mean of sequence length on training set:\", train_mean)\n",
        "print(\"Standard deviation of sequence length on training set:\", train_std)\n",
        "\n",
        "print(\"-----------------------------------------------------------\")\n",
        "\n",
        "dev_mean = np.mean([len(x) for x in X_dev_tokenized])\n",
        "dev_std = np.std([len(x) for x in X_dev_tokenized])\n",
        "print(\"Mean of sequence length on development set:\", dev_mean)\n",
        "print(\"Standard deviation of sequence length on development set:\", dev_std)\n",
        "\n",
        "print(\"-----------------------------------------------------------\")\n",
        "\n",
        "test_mean = np.mean([len(x) for x in X_test_tokenized])\n",
        "test_std = np.std([len(x) for x in X_test_tokenized])\n",
        "print(\"Mean of sequence length on test set:\", test_mean)\n",
        "print(\"Standard deviation of sequence length on test set:\", test_std)"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 25,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "FowDIKShmCz9",
        "outputId": "313f8bf8-b476-48b6-ffee-c2588ed91cdb"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "not many people know james whale  safe bet good chunk seen movie  believe semi biographical god monster  whale would wanted way  insightful  haunting exploration last day frankenstein bride frankenstein director  notable introducing one first complicated gay character hollywood movie  god monster interest biopic whales life track final day life  probably better movie  focus whales  ian mckellen  untraditional deceptive lust heterosexual gardener clayton boone  brendan fraser   begin whale  fairly talented artist  aside director  asking boone  sit    pose portrait   first  boone doesnt realize whale gay grows fascinated old man  discover whales sexual orientation dedicated protective maid  lynn redgrave   frightened  refusing sit guy  come back  storm  disgusted whales  locker room talk   boone return yet another time  whale promise tone aforementioned  locker room talk   find platonic relationship strengthening  meanwhile  jimmy suffering hallucination mental attack result stroke not long ago  ailment lead desperation depression  build point forced ask boone incredulously big favor  perhaps sensational god monster brendan frasers bravura performance  seeing film intensely disliked young actor tendency ruin movie  blast past  example   surprised expressive actor proved  impressive characters dialogue minimal  mckellan talking  requires fraser act body language  difficult skill master  something prof capable well  not ignore mckellans oscar nominated turn  earned nomination touching performance  although  risk sounding like philistine  would chosen performance nazi apt pupil nomination   understated  compelling gem  whale dreaded remembered solely work hollywood horror genre  thought hollywood equivalent battlefield  rejoiced free  reluctantly going reception fellow movie people  end  perhaps  killed  memory hollywood war blending together form living nightmare    eugene novikov \n",
            "['not', 'people', 'know', 'james', 'whale', 'safe', 'bet', 'good', 'chunk', 'seen', 'movie', 'believe', 'semi', 'biographical', 'god', 'monster', 'whale', 'wanted', 'way', 'insightful', 'haunting', 'exploration', 'day', 'frankenstein', 'bride', 'frankenstein', 'director', 'notable', 'introducing', 'complicated', 'gay', 'character', 'hollywood', 'movie', 'god', 'monster', 'interest', 'biopic', 'whales', 'life', 'track', 'final', 'day', 'life', 'probably', 'better', 'movie', 'focus', 'whales', 'ian', 'mckellen', 'untraditional', 'deceptive', 'lust', 'heterosexual', 'gardener', 'clayton', 'boone', 'brendan', 'fraser', 'begin', 'whale', 'fairly', 'talented', 'artist', 'aside', 'director', 'asking', 'boone', 'sit', 'pose', 'portrait', 'boone', 'nt', 'realize', 'whale', 'gay', 'grows', 'fascinated', 'old', 'man', 'discover', 'whales', 'sexual', 'orientation', 'dedicated', 'protective', 'maid', 'lynn', 'redgrave', 'frightened', 'refusing', 'sit', 'guy', 'come', 'storm', 'disgusted', 'whales', 'locker', 'room', 'talk', 'boone', 'return', 'time', 'whale', 'promise', 'tone', 'aforementioned', 'locker', 'room', 'talk', 'find', 'platonic', 'relationship', 'strengthening', 'jimmy', 'suffering', 'hallucination', 'mental', 'attack', 'result', 'stroke', 'not', 'long', 'ago', 'ailment', 'lead', 'desperation', 'depression', 'build', 'point', 'forced', 'ask', 'boone', 'incredulously', 'big', 'favor', 'sensational', 'god', 'monster', 'brendan', 'frasers', 'bravura', 'performance', 'seeing', 'film', 'intensely', 'disliked', 'young', 'actor', 'tendency', 'ruin', 'movie', 'blast', 'past', 'example', 'surprised', 'expressive', 'actor', 'proved', 'impressive', 'characters', 'dialogue', 'minimal', 'mckellan', 'talking', 'requires', 'fraser', 'act', 'body', 'language', 'difficult', 'skill', 'master', 'prof', 'capable', 'not', 'ignore', 'mckellans', 'oscar', 'nominated', 'turn', 'earned', 'nomination', 'touching', 'performance', 'risk', 'sounding', 'like', 'philistine', 'chosen', 'performance', 'nazi', 'apt', 'pupil', 'nomination', 'understated', 'compelling', 'gem', 'whale', 'dreaded', 'remembered', 'solely', 'work', 'hollywood', 'horror', 'genre', 'thought', 'hollywood', 'equivalent', 'battlefield', 'rejoiced', 'free', 'reluctantly', 'going', 'reception', 'fellow', 'movie', 'people', 'end', 'killed', 'memory', 'hollywood', 'war', 'blending', 'form', 'living', 'nightmare', 'eugene', 'novikov']\n"
          ]
        }
      ],
      "source": [
        "print(X_train[0])\n",
        "print(X_train_tokenized[0])\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "4xeBK1x2xDpM"
      },
      "source": [
        "##Tokenize, convert text (sequence of words) to sequence of indexes and PAD the sequences"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 26,
      "metadata": {
        "id": "glx4fvhHxBFV"
      },
      "outputs": [],
      "source": [
        "from tensorflow.keras.preprocessing.text import Tokenizer\n",
        "from tensorflow.keras.preprocessing.sequence import pad_sequences\n",
        "import math\n",
        "\n",
        "MAX_WORDS = 100000\n",
        "MAX_SEQUENCE_LENGTH =  math.ceil(train_mean + train_std)\n",
        "EMBEDDING_DIM = 300              #to be changed\n",
        "\n",
        "tokenizer = Tokenizer(num_words = MAX_WORDS, oov_token ='UNK')\n",
        "tokenizer.fit_on_texts([\" \".join(x) for x in X_train_tokenized])\n",
        "\n",
        "word_idx = tokenizer.word_index\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 27,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "1CsFs6b3goPV",
        "outputId": "f4419f3c-9ec7-4955-91ed-bd826adaa49f"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Number of unique words in the word index:  36337\n"
          ]
        }
      ],
      "source": [
        "print(\"Number of unique words in the word index: \", len(word_idx))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 28,
      "metadata": {
        "id": "GlqPE_iZ08Gs"
      },
      "outputs": [],
      "source": [
        "train_sequences = tokenizer.texts_to_sequences([\" \".join(x) for x in X_train_tokenized])\n",
        "dev_sequences = tokenizer.texts_to_sequences([\" \".join(x) for x in X_dev_tokenized])\n",
        "test_sequences = tokenizer.texts_to_sequences([\" \".join(x) for x in X_test_tokenized])\n",
        "\n",
        "train_pad = pad_sequences(train_sequences, maxlen = MAX_SEQUENCE_LENGTH, padding ='post')\n",
        "dev_pad = pad_sequences(dev_sequences, maxlen = MAX_SEQUENCE_LENGTH, padding ='post')\n",
        "test_pad = pad_sequences(test_sequences, maxlen = MAX_SEQUENCE_LENGTH, padding ='post')"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 29,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "DXkC9kDZ14l8",
        "outputId": "9b7379bc-1887-4260-9e4f-806c7dbc5eed"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "[5, 21, 19, 172, 2072, 1378, 1887, 11, 4996, 47, 3, 226, 10828, 12688, 363, 567, 2072, 615, 13, 3289, 1486, 4069, 45, 2294, 2744, 2294, 34, 1888, 4491, 2295, 1015, 7, 110, 3, 363, 567, 343, 9480, 7054, 14, 715, 206, 45, 14, 114, 46, 3, 524, 7054, 1983, 15531, 20585, 12689, 4492, 6522, 7684, 6523, 4493, 4754, 3290, 106, 2072, 736, 670, 975, 859, 34, 1648, 4493, 765, 3906, 2406, 4493, 4, 593, 2072, 1015, 1349, 4755, 71, 26, 950, 7054, 568, 15532, 3050, 7685, 4494, 8470, 7686, 3742, 4495, 765, 48, 17, 1649, 7687, 7054, 6524, 284, 288, 4493, 302, 9, 2072, 976, 705, 1841, 6524, 284, 288, 37, 20586, 151, 20587, 1431, 2034, 6525, 1790, 688, 244, 3424, 5, 72, 442, 15533, 136, 3743, 4070, 1016, 54, 443, 766, 4493, 20588, 42, 1683, 6526, 363, 567, 4754, 15534, 10829, 24, 212, 2, 5663, 6058, 65, 30, 2838, 2073, 3, 4071, 219, 245, 1159, 4496, 30, 3291, 430, 767, 134, 4072, 10830, 569, 1729, 3290, 224, 276, 847, 463, 1002, 737, 832, 1541, 5, 3292, 12690, 431, 4073, 67, 3907, 1487, 1309, 24, 1350, 4281, 8, 20589, 1213, 24, 2839, 4997, 5329, 1487, 4074, 1330, 2236, 2072, 6527, 2407, 2840, 22, 110, 141, 303, 165, 110, 3744, 2237, 20590, 833, 3581, 55, 7688, 977, 3, 21, 25, 420, 811, 110, 119, 8471, 354, 369, 1140, 3168, 6528]\n",
            "------------------------------\n",
            "[    5    21    19   172  2072  1378  1887    11  4996    47     3   226\n",
            " 10828 12688   363   567  2072   615    13  3289  1486  4069    45  2294\n",
            "  2744  2294    34  1888  4491  2295  1015     7   110     3   363   567\n",
            "   343  9480  7054    14   715   206    45    14   114    46     3   524\n",
            "  7054  1983 15531 20585 12689  4492  6522  7684  6523  4493  4754  3290\n",
            "   106  2072   736   670   975   859    34  1648  4493   765  3906  2406\n",
            "  4493     4   593  2072  1015  1349  4755    71    26   950  7054   568\n",
            " 15532  3050  7685  4494  8470  7686  3742  4495   765    48    17  1649\n",
            "  7687  7054  6524   284   288  4493   302     9  2072   976   705  1841\n",
            "  6524   284   288    37 20586   151 20587  1431  2034  6525  1790   688\n",
            "   244  3424     5    72   442 15533   136  3743  4070  1016    54   443\n",
            "   766  4493 20588    42  1683  6526   363   567  4754 15534 10829    24\n",
            "   212     2  5663  6058    65    30  2838  2073     3  4071   219   245\n",
            "  1159  4496    30  3291   430   767   134  4072 10830   569  1729  3290\n",
            "   224   276   847   463  1002   737   832  1541     5  3292 12690   431\n",
            "  4073    67  3907  1487  1309    24  1350  4281     8 20589  1213    24\n",
            "  2839  4997  5329  1487  4074  1330  2236  2072  6527  2407  2840    22\n",
            "   110   141   303   165   110  3744  2237 20590   833  3581    55  7688\n",
            "   977     3    21    25   420   811   110   119  8471   354   369  1140\n",
            "  3168  6528     0     0     0     0     0     0     0     0     0     0\n",
            "     0     0     0     0     0     0     0     0     0     0     0     0\n",
            "     0     0     0     0     0     0     0     0     0     0     0     0\n",
            "     0     0     0     0     0     0     0     0     0     0     0     0\n",
            "     0     0     0     0     0     0     0     0     0     0     0     0\n",
            "     0     0     0     0     0     0     0     0     0     0     0     0\n",
            "     0     0     0     0     0     0     0     0     0     0     0     0\n",
            "     0     0     0     0     0     0     0     0     0     0     0     0\n",
            "     0     0     0     0     0     0     0     0     0     0     0     0\n",
            "     0     0     0     0     0     0     0     0     0     0     0     0\n",
            "     0     0     0     0     0     0     0     0     0     0     0     0\n",
            "     0     0     0     0     0     0     0     0     0     0     0     0\n",
            "     0     0     0     0     0     0     0     0     0     0     0     0\n",
            "     0     0     0     0     0     0     0     0     0     0     0     0\n",
            "     0     0     0     0     0     0     0     0     0     0     0     0\n",
            "     0     0     0     0     0     0     0     0     0     0     0     0\n",
            "     0     0     0     0     0     0     0     0     0     0     0     0\n",
            "     0     0     0     0     0     0     0     0     0     0     0     0\n",
            "     0     0     0     0]\n"
          ]
        }
      ],
      "source": [
        "print(train_sequences[0])\n",
        "print(\"------------------------------\")\n",
        "print(train_pad[0])"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "-6fWrvm5rC7Y"
      },
      "source": [
        "##Download and unzip fasttext binary model for word embeddings"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 30,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "MYdz-1ysqxQg",
        "outputId": "e4997c8d-8f0a-4fd2-a9fd-98779d68235d"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "--2024-02-19 17:34:35--  https://dl.fbaipublicfiles.com/fasttext/vectors-crawl/cc.en.300.bin.gz\n",
            "Resolving dl.fbaipublicfiles.com (dl.fbaipublicfiles.com)... 108.157.254.15, 108.157.254.124, 108.157.254.102, ...\n",
            "Connecting to dl.fbaipublicfiles.com (dl.fbaipublicfiles.com)|108.157.254.15|:443... connected.\n",
            "HTTP request sent, awaiting response... 200 OK\n",
            "Length: 4503593528 (4.2G) [application/octet-stream]\n",
            "Saving to: ‘cc.en.300.bin.gz’\n",
            "\n",
            "cc.en.300.bin.gz    100%[===================>]   4.19G   176MB/s    in 24s     \n",
            "\n",
            "2024-02-19 17:34:59 (176 MB/s) - ‘cc.en.300.bin.gz’ saved [4503593528/4503593528]\n",
            "\n"
          ]
        }
      ],
      "source": [
        "!wget https://dl.fbaipublicfiles.com/fasttext/vectors-crawl/cc.en.300.bin.gz\n",
        "!gzip -d cc.en.300.bin.gz"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "JhufSqPelZ0r"
      },
      "source": [
        "##Create embedding matrix"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "2uAkNWh2oQhS"
      },
      "source": [
        "**Process**: The loop iterates over the word_index dictionary, which contains word-to-index mappings generated by the tokenizer. For each word in the word_index, it checks if the index is within the limit of *MAX_WORDS*. If so, it retrieves the corresponding word vector from the FastText model using *fasttext_model.get_word_vector(word=key)* and assigns it to the corresponding row in the embedding_matrix."
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 31,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_7CLC2WIkh_6",
        "outputId": "064ec3b2-12f6-4abd-ab00-375dc3313ec4"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Loading embeddings model..\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "Warning : `load_model` does not return WordVectorModel or SupervisedModel any more, but a `FastText` object which is very similar.\n"
          ]
        }
      ],
      "source": [
        "import fasttext\n",
        "\n",
        "print(\"Loading embeddings model..\")\n",
        "fasttext_model = fasttext.load_model('cc.en.300.bin')\n",
        "embedding_matrix = np.zeros(shape=((MAX_WORDS + 2), 300))          # +2 because we have reserved indices for padding and out-of-vocabulary tokens\n",
        "\n",
        "for key, value in word_idx.items():\n",
        "    if value <= MAX_WORDS:\n",
        "        embedding_matrix[value] = fasttext_model.get_word_vector(word=key)            #create embedding matrix\n",
        "\n",
        "del fasttext_model      #save memory"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 32,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "aCVGePpfnQsi",
        "outputId": "c2a5587d-c1a0-4f51-eeed-90e1488f723c"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "y_dev_1_hot: [0] [1] [1] [0] [0] [1] [0] [1] [1] [1] [0] [1] [0] [1] [0] [1] [0] [1] [0] [0] [0] [1] [0] [1] [1] [1] [0] [0] [0] [0] [0] [1] [0] [0] [1] [1] [1] [0] [1] [1] [1] [0] [1] [1] [1] [1] [0] [0] [1] [0] [0] [0] [0] [1] [1] [1] [0] [0] [0] [0] [1] [1] [1] [1] [0] [0] [0] [0] [0] [0] [1] [0] [1] [1] [0] [1] [1] [0] [1] [0] [0] [0] [1] [0] [1] [1] [1] [0] [1] [1] [0] [0] [0] [0] [0] [1] [0] [1] [1] [0] [1] [0] [0] [1] [1] [1] [0] [0] [0] [0] [1] [0] [1] [0] [1] [0] [0] [1] [0] [1] [0] [0] [0] [0] [0] [1] [0] [0] [0] [0] [0] [0] [1] [0] [1] [0] [0] [0] [0] [0] [0] [0] [0] [1] [1] [0] [1] [0] [1] [1] [1] [1] [1] [0] [0] [0] [0] [1] [0] [1] [0] [1] [0] [1] [0] [1] [0] [1] [1] [1] [0] [0] [0] [1] [0] [0] [1] [1] [1] [1] [1] [1] [1] [0] [0] [1] [0] [0] [0] [1] [1] [1] [1] [1] [1] [1] [1] [1] [0] [1] [1] [0] [0] [0] [0] [1] [0] [1] [0] [1] [0] [1] [0] [0] [1] [1] [1] [1] [0] [0] [0] [0] [0] [1] [1] [1] [0] [1] [1] [1] [0] [0] [0] [0] [1] [0] [1] [0] [1] [1] [0] [0] [1] [0] [0] [0] [0] [0] [1] [1] [0] [1] [1] [1] [1] [0] [0] [0] [1] [0] [1] [0] [0] [1] [0] [1] [0] [0] [0] [0] [0] [0] [0] [0] [1] [1] [1] [1] [1] [1] [1] [0] [0] [1] [1] [1] [1] [1] [0] [0] [1] [1] [0] [0] [0] [1] [1] [1] [0] [1]\n"
          ]
        }
      ],
      "source": [
        "from sklearn.preprocessing import LabelBinarizer\n",
        "\n",
        "lb = LabelBinarizer()\n",
        "target_list = z\n",
        "\n",
        "y_train_1_hot = lb.fit_transform([target_list[x] for x in y_train])\n",
        "y_dev_1_hot = lb.transform([target_list[x] for x in y_dev])\n",
        "\n",
        "#y_train_1_hot = np.argmax(y_train_1_hot, axis=1)\n",
        "#y_dev_1_hot = np.argmax(y_dev_1_hot, axis=1)\n",
        "\n",
        "#print('y_dev_1_hot:', y_dev_1_hot)                             #prints vertically\n",
        "print('y_dev_1_hot:', ' '.join(map(str, y_dev_1_hot)))          #prints horizontally\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 33,
      "metadata": {
        "id": "ixmTyQKcpT_q"
      },
      "outputs": [],
      "source": [
        "import tensorflow as tf\n",
        "\n",
        "class Metrics(tf.keras.callbacks.Callback):\n",
        "\n",
        "    def __init__(self, valid_data):\n",
        "        super(Metrics, self).__init__()\n",
        "        self.validation_data = valid_data\n",
        "\n",
        "    def on_epoch_end(self, epoch, logs=None):\n",
        "        logs = logs or {}\n",
        "        val_predict = np.argmax(self.model.predict(self.validation_data[0]), -1)\n",
        "        val_targ = self.validation_data[1]\n",
        "        if len(val_targ.shape) == 2 and val_targ.shape[1] != 1:\n",
        "            val_targ = np.argmax(val_targ, -1)\n",
        "        val_targ = tf.cast(val_targ,dtype=tf.float32)\n",
        "        _val_f1 = f1_score(val_targ, val_predict,average=\"weighted\")\n",
        "        _val_recall = recall_score(val_targ, val_predict,average=\"weighted\")\n",
        "        _val_precision = precision_score(val_targ, val_predict,average=\"weighted\")\n",
        "        logs['val_f1'] = _val_f1\n",
        "        logs['val_recall'] = _val_recall\n",
        "        logs['val_precision'] = _val_precision\n",
        "        print(\" — val_f1: %f — val_precision: %f — val_recall: %f\" % (_val_f1, _val_precision, _val_recall))\n",
        "        return"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "KENECgWVtMBM"
      },
      "source": [
        "##Self attention class"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 34,
      "metadata": {
        "id": "zD27L6NItPC7"
      },
      "outputs": [],
      "source": [
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import Dense, Dropout, Layer\n",
        "\n",
        "class SelfAttention(tf.keras.layers.Layer):\n",
        "  def __init__(self, mlp_layers=0, units=0, dropout_rate=0, return_attention=False, **kwargs):\n",
        "    super(SelfAttention, self).__init__(**kwargs)\n",
        "    self.mlp_layers = mlp_layers\n",
        "    self.mlp_units = units\n",
        "    self.return_attention = return_attention\n",
        "    self.dropout_rate = dropout_rate\n",
        "    self.attention_mlp = self.build_mlp()\n",
        "\n",
        "  def build_mlp(self):\n",
        "    mlp = Sequential()\n",
        "    for i in range(self.mlp_layers):\n",
        "      mlp.add(Dense(self.mlp_units, activation='relu'))\n",
        "      mlp.add(Dropout(self.dropout_rate))\n",
        "    mlp.add(Dense(1))\n",
        "    return mlp\n",
        "\n",
        "  def call(self, x, mask=None):\n",
        "    a = self.attention_mlp(x)\n",
        "    a = tf.squeeze(a, axis=2)\n",
        "\n",
        "    if mask is not None:\n",
        "      mask = tf.keras.backend.cast(mask, tf.keras.backend.floatx())\n",
        "      a -= 100000.0 * (1.0 - mask)\n",
        "\n",
        "    a = tf.keras.backend.expand_dims(tf.keras.backend.softmax(a, axis=-1))\n",
        "    weighted_input = x * a\n",
        "    result = tf.keras.backend.sum(weighted_input, axis=1)\n",
        "\n",
        "    if self.return_attention:\n",
        "      return [result, a]\n",
        "    return result"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Hw7_uglhosv7"
      },
      "source": [
        "##Create one-hot vectors"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 35,
      "metadata": {
        "id": "zjeoicRn2rd7"
      },
      "outputs": [],
      "source": [
        "y_train = tf.keras.utils.to_categorical(y_train, num_classes=2)\n",
        "y_dev = tf.keras.utils.to_categorical(y_dev, num_classes=2)\n",
        "y_test = tf.keras.utils.to_categorical(y_test, num_classes=2)\n"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "AnC0EFNQ4-5S"
      },
      "source": [
        "##RNN model"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 39,
      "metadata": {
        "id": "Ih5jqOZz4uUK"
      },
      "outputs": [],
      "source": [
        "def build_model(hp):\n",
        "    GRU_SIZE = hp.Int('gru_size', min_value=100, max_value=500, step=50)\n",
        "    learning_rate = hp.Choice('learning_rate', values=[1e-3, 1e-2, 1e-1])\n",
        "    num_hidden_layers = hp.Int('num_hidden_layers', min_value=1, max_value=3)\n",
        "    mlp_units = hp.Choice('mlp_units', values=[64, 128, 256])\n",
        "    dropout_rate = hp.Float('dropout_rate', min_value=0.2, max_value=0.5, step=0.05)\n",
        "    hidden_layer_size = hp.Choice('hidden_layer_size', values=[64, 128, 256])\n",
        "\n",
        "    model = tf.keras.Sequential()\n",
        "\n",
        "    model.add(\n",
        "        tf.keras.layers.Embedding(\n",
        "            MAX_WORDS+2,\n",
        "            EMBEDDING_DIM,\n",
        "            weights=[embedding_matrix],\n",
        "            input_length=MAX_SEQUENCE_LENGTH,\n",
        "            mask_zero=True,\n",
        "            trainable=False\n",
        "        )\n",
        "    )\n",
        "    model.add(tf.keras.layers.Dropout(dropout_rate))\n",
        "\n",
        "    model.add(\n",
        "        tf.keras.layers.Bidirectional(\n",
        "            tf.keras.layers.GRU(\n",
        "                GRU_SIZE,\n",
        "                return_sequences=True,\n",
        "                recurrent_dropout=dropout_rate\n",
        "            )\n",
        "        )\n",
        "    )\n",
        "    model.add(tf.keras.layers.Dropout(dropout_rate))\n",
        "\n",
        "    model.add(SelfAttention(mlp_layers=num_hidden_layers, units=mlp_units))\n",
        "\n",
        "    model.add(tf.keras.layers.Dense(hidden_layer_size, activation='relu'))\n",
        "    model.add(tf.keras.layers.Dropout(dropout_rate))\n",
        "\n",
        "    model.add(tf.keras.layers.Dense(2, activation='sigmoid'))\n",
        "\n",
        "    model.compile(\n",
        "        loss='binary_crossentropy',\n",
        "        optimizer=tf.keras.optimizers.Adam(learning_rate=learning_rate),\n",
        "        metrics=[\"binary_accuracy\"]\n",
        "    )\n",
        "\n",
        "    return model"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "i2xh27gH5vcq"
      },
      "source": [
        "##Hyperparameter search with Keras Tuner"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 37,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SbM7UBKW7IrC",
        "outputId": "64488557-355c-4640-b41a-4e931164f946"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Collecting keras-tuner\n",
            "  Downloading keras_tuner-1.4.6-py3-none-any.whl (128 kB)\n",
            "\u001b[?25l     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m0.0/128.9 kB\u001b[0m \u001b[31m?\u001b[0m eta \u001b[36m-:--:--\u001b[0m\r\u001b[2K     \u001b[90m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m128.9/128.9 kB\u001b[0m \u001b[31m3.6 MB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m\n",
            "\u001b[?25hRequirement already satisfied: keras in /usr/local/lib/python3.10/dist-packages (from keras-tuner) (2.15.0)\n",
            "Requirement already satisfied: packaging in /usr/local/lib/python3.10/dist-packages (from keras-tuner) (23.2)\n",
            "Requirement already satisfied: requests in /usr/local/lib/python3.10/dist-packages (from keras-tuner) (2.31.0)\n",
            "Collecting kt-legacy (from keras-tuner)\n",
            "  Downloading kt_legacy-1.0.5-py3-none-any.whl (9.6 kB)\n",
            "Requirement already satisfied: charset-normalizer<4,>=2 in /usr/local/lib/python3.10/dist-packages (from requests->keras-tuner) (3.3.2)\n",
            "Requirement already satisfied: idna<4,>=2.5 in /usr/local/lib/python3.10/dist-packages (from requests->keras-tuner) (3.6)\n",
            "Requirement already satisfied: urllib3<3,>=1.21.1 in /usr/local/lib/python3.10/dist-packages (from requests->keras-tuner) (2.0.7)\n",
            "Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.10/dist-packages (from requests->keras-tuner) (2024.2.2)\n",
            "Installing collected packages: kt-legacy, keras-tuner\n",
            "Successfully installed keras-tuner-1.4.6 kt-legacy-1.0.5\n"
          ]
        }
      ],
      "source": [
        "!pip install keras-tuner"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 40,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "Vmt3o1f95uzp",
        "outputId": "9113bf4c-ae19-48ad-abff-572366f0cfb7"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Trial 5 Complete [00h 04m 19s]\n",
            "val_binary_accuracy: 0.5\n",
            "\n",
            "Best val_binary_accuracy So Far: 0.8566666841506958\n",
            "Total elapsed time: 00h 34m 30s\n"
          ]
        }
      ],
      "source": [
        "from kerastuner.tuners import RandomSearch\n",
        "\n",
        "tuner = RandomSearch(\n",
        "    build_model,\n",
        "    objective='val_binary_accuracy',\n",
        "    max_trials=5,\n",
        "    executions_per_trial=1,\n",
        "    directory='keras_tuner_logs',\n",
        "    project_name='hyperparameter_tuning_gru_self_attention'\n",
        ")\n",
        "\n",
        "\n",
        "tuner.search(\n",
        "    train_pad,\n",
        "    y_train,\n",
        "    validation_data=(dev_pad, y_dev),\n",
        "    batch_size=256,\n",
        "    epochs=50,\n",
        "    shuffle=True,\n",
        "    callbacks=[\n",
        "        tf.keras.callbacks.EarlyStopping(monitor='val_binary_accuracy', patience=10, restore_best_weights=True)\n",
        "    ]\n",
        ")\n",
        "\n",
        "\n",
        "best_hp = tuner.get_best_hyperparameters(num_trials=1)[0]"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 41,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "mV8wW-T0mBKS",
        "outputId": "de90c4bb-780c-4d56-8afe-c6d1be4d495d"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Best hyperparameters:\n",
            "GRU Size: 250\n",
            "Dropout Rate: 0.30000000000000004\n",
            "MLP Layers: 1\n",
            "MLP Units: 64\n",
            "Hidden layer size:  256\n",
            "Learning rate:  0.01\n"
          ]
        }
      ],
      "source": [
        "best_gru_size = best_hp.get('gru_size')\n",
        "best_dropout_rate = best_hp.get('dropout_rate')\n",
        "best_mlp_layers = best_hp.get('num_hidden_layers')\n",
        "best_mlp_units = best_hp.get('mlp_units')\n",
        "best_hid_size = best_hp.get('hidden_layer_size')\n",
        "best_learning_rate = best_hp.get('learning_rate')\n",
        "\n",
        "\n",
        "print(\"Best hyperparameters:\")\n",
        "print(\"GRU Size:\", best_gru_size)\n",
        "print(\"Dropout Rate:\", best_dropout_rate)\n",
        "print(\"MLP Layers:\", best_mlp_layers)\n",
        "print(\"MLP Units:\", best_mlp_units)\n",
        "print(\"Hidden layer size: \", best_hid_size)\n",
        "print(\"Learning rate: \", best_learning_rate)\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 42,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "-5Wb1XXbmsbR",
        "outputId": "87c65b9b-e0d5-46b2-f53d-f695844e5a92"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:Layer gru will not use cuDNN kernels since it doesn't meet the criteria. It will use a generic GPU kernel as fallback when running on GPU.\n",
            "WARNING:tensorflow:Layer gru will not use cuDNN kernels since it doesn't meet the criteria. It will use a generic GPU kernel as fallback when running on GPU.\n",
            "WARNING:tensorflow:Layer gru will not use cuDNN kernels since it doesn't meet the criteria. It will use a generic GPU kernel as fallback when running on GPU.\n"
          ]
        }
      ],
      "source": [
        "best_model = tuner.get_best_models(num_models=1)[0]\n",
        "#best_trial = tuner.get_best_trials(num_trials=1)[0]\n",
        "\n",
        "\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 43,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "MUeEOsweSFYG",
        "outputId": "7b93d864-b761-453f-eb78-4191f6d69a2c"
      },
      "outputs": [
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:Layer gru_1 will not use cuDNN kernels since it doesn't meet the criteria. It will use a generic GPU kernel as fallback when running on GPU.\n",
            "WARNING:tensorflow:Layer gru_1 will not use cuDNN kernels since it doesn't meet the criteria. It will use a generic GPU kernel as fallback when running on GPU.\n",
            "WARNING:tensorflow:Layer gru_1 will not use cuDNN kernels since it doesn't meet the criteria. It will use a generic GPU kernel as fallback when running on GPU.\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Epoch 1/50\n"
          ]
        },
        {
          "name": "stderr",
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:Detecting that an object or model or tf.train.Checkpoint is being deleted with unrestored values. See the following logs for the specific values in question. To silence these warnings, use `status.expect_partial()`. See https://www.tensorflow.org/api_docs/python/tf/train/Checkpoint#restorefor details about the status object returned by the restore function.\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.1\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.2\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.3\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.4\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.5\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.6\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.7\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.8\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.9\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.10\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.11\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.12\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.13\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.14\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.15\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.16\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.17\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.18\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.19\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.20\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.21\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.22\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.23\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.24\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.25\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.26\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.27\n",
            "WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).optimizer._variables.28\n"
          ]
        },
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "22/22 [==============================] - 69s 3s/step - loss: 0.7021 - binary_accuracy: 0.5314 - val_loss: 0.6726 - val_binary_accuracy: 0.5617\n",
            "Epoch 2/50\n",
            "22/22 [==============================] - 59s 3s/step - loss: 0.6548 - binary_accuracy: 0.6193 - val_loss: 0.6068 - val_binary_accuracy: 0.7083\n",
            "Epoch 3/50\n",
            "22/22 [==============================] - 60s 3s/step - loss: 0.5914 - binary_accuracy: 0.6896 - val_loss: 0.5338 - val_binary_accuracy: 0.7317\n",
            "Epoch 4/50\n",
            "22/22 [==============================] - 62s 3s/step - loss: 0.4481 - binary_accuracy: 0.7996 - val_loss: 0.4021 - val_binary_accuracy: 0.8317\n",
            "Epoch 5/50\n",
            "22/22 [==============================] - 61s 3s/step - loss: 0.4004 - binary_accuracy: 0.8289 - val_loss: 0.3884 - val_binary_accuracy: 0.8450\n",
            "Epoch 6/50\n",
            "22/22 [==============================] - 62s 3s/step - loss: 0.3641 - binary_accuracy: 0.8586 - val_loss: 0.3726 - val_binary_accuracy: 0.8467\n",
            "Epoch 7/50\n",
            "22/22 [==============================] - 61s 3s/step - loss: 0.3001 - binary_accuracy: 0.8771 - val_loss: 0.4168 - val_binary_accuracy: 0.8233\n",
            "Epoch 8/50\n",
            "22/22 [==============================] - 60s 3s/step - loss: 0.2559 - binary_accuracy: 0.9082 - val_loss: 0.4256 - val_binary_accuracy: 0.8117\n",
            "Epoch 9/50\n",
            "22/22 [==============================] - 61s 3s/step - loss: 0.2590 - binary_accuracy: 0.8886 - val_loss: 0.3569 - val_binary_accuracy: 0.8433\n",
            "Epoch 10/50\n",
            "22/22 [==============================] - 62s 3s/step - loss: 0.2159 - binary_accuracy: 0.9243 - val_loss: 0.4611 - val_binary_accuracy: 0.8367\n",
            "Epoch 11/50\n",
            "22/22 [==============================] - 62s 3s/step - loss: 0.1992 - binary_accuracy: 0.9186 - val_loss: 0.4705 - val_binary_accuracy: 0.8200\n",
            "Epoch 12/50\n",
            "22/22 [==============================] - 61s 3s/step - loss: 0.1557 - binary_accuracy: 0.9364 - val_loss: 0.4919 - val_binary_accuracy: 0.8300\n",
            "Epoch 13/50\n",
            "22/22 [==============================] - 60s 3s/step - loss: 0.1824 - binary_accuracy: 0.9271 - val_loss: 0.4828 - val_binary_accuracy: 0.8183\n",
            "Epoch 14/50\n",
            "22/22 [==============================] - 62s 3s/step - loss: 0.1556 - binary_accuracy: 0.9364 - val_loss: 0.5007 - val_binary_accuracy: 0.8233\n",
            "Epoch 15/50\n",
            "22/22 [==============================] - 62s 3s/step - loss: 0.1401 - binary_accuracy: 0.9489 - val_loss: 0.5480 - val_binary_accuracy: 0.8433\n",
            "Epoch 16/50\n",
            "22/22 [==============================] - 61s 3s/step - loss: 0.1333 - binary_accuracy: 0.9475 - val_loss: 0.5379 - val_binary_accuracy: 0.8150\n"
          ]
        }
      ],
      "source": [
        "\n",
        "best_model = tuner.hypermodel.build(best_hp)       #build model with the best hyperparameters\n",
        "\n",
        "history = best_model.fit(\n",
        "    train_pad,\n",
        "    y_train,\n",
        "    validation_data=(dev_pad, y_dev),\n",
        "    batch_size=64,                            #the optimal from the MLP model\n",
        "    epochs=50,\n",
        "    shuffle=True,\n",
        "    callbacks=[\n",
        "        tf.keras.callbacks.EarlyStopping(monitor='val_binary_accuracy', patience=10, restore_best_weights=True)\n",
        "    ]\n",
        ")"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "w2--OyLgivIO"
      },
      "source": [
        "##Visualize Model's Training History"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 44,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 927
        },
        "id": "98j0z4o1ixKy",
        "outputId": "7b73a64e-a6d0-4b74-fcad-69be74e3d318"
      },
      "outputs": [
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        },
        {
          "data": {
            "image/png": "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",
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ]
          },
          "metadata": {},
          "output_type": "display_data"
        }
      ],
      "source": [
        "%matplotlib inline\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "# summarize history for accuracy\n",
        "plt.plot(history.history['binary_accuracy'])\n",
        "plt.plot(history.history['val_binary_accuracy'])\n",
        "plt.title('model accuracy')\n",
        "plt.ylabel('accuracy')\n",
        "plt.xlabel('epoch')\n",
        "plt.legend(['train', 'dev'], loc='upper left')\n",
        "plt.show()\n",
        "\n",
        "# summarize history for loss\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', 'dev'], loc='upper right')\n",
        "plt.show()"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "TJRhWykgMMxu"
      },
      "source": [
        "##Metrics - Classification Report"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 45,
      "metadata": {
        "id": "wV14A90wNE8G"
      },
      "outputs": [],
      "source": [
        "from sklearn.metrics import roc_auc_score\n",
        "\n",
        "def calculate_auc_scores(model, data, labels):\n",
        "    predictions = np.argmax(model.predict(data, verbose=0), -1)\n",
        "    true_labels = np.argmax(labels, -1)\n",
        "    auc_scores = []\n",
        "    for class_index in range(labels.shape[1]):\n",
        "        class_true = (true_labels == class_index).astype(int)\n",
        "        class_probs = model.predict(data, verbose=0)[:, class_index]\n",
        "        roc_auc = roc_auc_score(class_true, class_probs)\n",
        "        auc_scores.append(roc_auc)\n",
        "    macro_auc = np.mean(auc_scores)\n",
        "    return auc_scores, macro_auc"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 46,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "_evHO2OZafBt",
        "outputId": "d201fdef-c2da-4e5b-bf44-c1d85b92a00c"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Classification report - Training set\n",
            "------------------------\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.87      0.95      0.91       690\n",
            "           1       0.94      0.86      0.90       710\n",
            "\n",
            "    accuracy                           0.90      1400\n",
            "   macro avg       0.91      0.90      0.90      1400\n",
            "weighted avg       0.91      0.90      0.90      1400\n",
            "\n"
          ]
        }
      ],
      "source": [
        "from sklearn.metrics import classification_report\n",
        "\n",
        "\n",
        "predictions_train = np.argmax(best_model.predict(train_pad, verbose=0), -1)\n",
        "y_train_ = np.argmax(y_train, -1)\n",
        "print('Classification report - Training set')\n",
        "print('------------------------')\n",
        "print(classification_report(y_train_, predictions_train))\n"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 47,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "KzGWtrWfa4R2",
        "outputId": "87da0ca6-626f-487e-b024-197c89088f10"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Classification report - Development set\n",
            "------------------------\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.82      0.90      0.86       157\n",
            "           1       0.88      0.79      0.83       143\n",
            "\n",
            "    accuracy                           0.85       300\n",
            "   macro avg       0.85      0.84      0.85       300\n",
            "weighted avg       0.85      0.85      0.85       300\n",
            "\n"
          ]
        }
      ],
      "source": [
        "predictions_dev = np.argmax(best_model.predict(dev_pad, verbose=0), -1)\n",
        "y_dev_ = np.argmax(y_dev, -1)\n",
        "print('Classification report - Development set')\n",
        "print('------------------------')\n",
        "print(classification_report(y_dev_, predictions_dev))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 48,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "ldyBnG7IbEwA",
        "outputId": "93627b3f-8de1-4eef-e2d8-1111a6004487"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Classification report - Test set\n",
            "---------------------------\n",
            "              precision    recall  f1-score   support\n",
            "\n",
            "           0       0.82      0.85      0.84       153\n",
            "           1       0.84      0.81      0.82       147\n",
            "\n",
            "    accuracy                           0.83       300\n",
            "   macro avg       0.83      0.83      0.83       300\n",
            "weighted avg       0.83      0.83      0.83       300\n",
            "\n"
          ]
        }
      ],
      "source": [
        "predictions_test = np.argmax(best_model.predict(test_pad, verbose=0), -1)\n",
        "y_test_ = np.argmax(y_test, -1)\n",
        "print('Classification report - Test set')\n",
        "print('---------------------------')\n",
        "print(classification_report(y_test_, predictions_test))"
      ]
    },
    {
      "cell_type": "code",
      "execution_count": 49,
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "gUvSc5YzNhO9",
        "outputId": "fe3d55b2-f508-43d7-b572-9136801ff762"
      },
      "outputs": [
        {
          "name": "stdout",
          "output_type": "stream",
          "text": [
            "Train AUC Scores:\n",
            "Class 0: 0.9641926923862013\n",
            "Class 1: 0.9640845070422535\n",
            "Train Macro AUC: 0.9641385997142273\n",
            "\n",
            "Dev AUC Scores:\n",
            "Class 0: 0.9170638278918534\n",
            "Class 1: 0.9173756180125607\n",
            "Dev Macro AUC: 0.917219722952207\n",
            "\n",
            "Test AUC Scores:\n",
            "Class 0: 0.9057400738073006\n",
            "Class 1: 0.9056511493486283\n",
            "Test Macro AUC: 0.9056956115779644\n"
          ]
        }
      ],
      "source": [
        "train_auc_scores, train_macro_auc = calculate_auc_scores(best_model, train_pad, y_train)\n",
        "dev_auc_scores, dev_macro_auc = calculate_auc_scores(best_model, dev_pad, y_dev)\n",
        "test_auc_scores, test_macro_auc = calculate_auc_scores(best_model, test_pad, y_test)\n",
        "\n",
        "print(\"Train AUC Scores:\")\n",
        "for class_index, auc_score in enumerate(train_auc_scores):\n",
        "    print(f\"Class {class_index}: {auc_score}\")\n",
        "print(f\"Train Macro AUC: {train_macro_auc}\")\n",
        "\n",
        "print(\"\\nDev AUC Scores:\")\n",
        "for class_index, auc_score in enumerate(dev_auc_scores):\n",
        "    print(f\"Class {class_index}: {auc_score}\")\n",
        "print(f\"Dev Macro AUC: {dev_macro_auc}\")\n",
        "\n",
        "print(\"\\nTest AUC Scores:\")\n",
        "for class_index, auc_score in enumerate(test_auc_scores):\n",
        "    print(f\"Class {class_index}: {auc_score}\")\n",
        "print(f\"Test Macro AUC: {test_macro_auc}\")"
      ]
    }
  ],
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "gpuType": "T4",
      "provenance": []
    },
    "kernelspec": {
      "display_name": "Python 3",
      "name": "python3"
    },
    "language_info": {
      "name": "python"
    },
    "widgets": {
      "application/vnd.jupyter.widget-state+json": {
        "07870395d12a4795a897c9c4e81bccaa": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "FloatProgressModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_c0e082767739480da73bd1c54c6f16db",
            "max": 2000,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_f94d417290f44f0bbb844c652f8dfe93",
            "value": 2000
          }
        },
        "0b289fb9c8e948a8af718fb65f78d975": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "0cf1af116cd1439ca7c38b4855d7e9a9": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_0b289fb9c8e948a8af718fb65f78d975",
            "placeholder": "​",
            "style": "IPY_MODEL_316615ccbf44426d818ff140331e6a69",
            "value": "100%"
          }
        },
        "0f2a5d2bfa024fceb2a809017ab4713d": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "143c4bafc3d0415c8e1d435af380aea6": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "151b618982024df2bf250c08f2fdc3e6": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "174ceef81b444b4b9071ad658a5ace66": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "1ee6b2099b5b4fa59d83ca4f8479a830": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HBoxModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_0cf1af116cd1439ca7c38b4855d7e9a9",
              "IPY_MODEL_07870395d12a4795a897c9c4e81bccaa",
              "IPY_MODEL_8f9432d67fc74d4daafa43c7e144127b"
            ],
            "layout": "IPY_MODEL_6c25a81bdd5b4979b98bee24bd613816"
          }
        },
        "300e1ca78baa4457b906f58170b19d06": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "316615ccbf44426d818ff140331e6a69": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "3278459baebc4a2b87430120b9563eb2": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "ProgressStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "37bedfb3159948aa9296986647813cc4": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "ProgressStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "37d787fac5874a8aa0342b73a6d7c3df": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "3d712ec28c9c405ab625ad74ca72c7a4": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HBoxModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_c69fc550f2344cac92607cc70c1fbd30",
              "IPY_MODEL_649916d14e044e16ac04051feb5cefe3",
              "IPY_MODEL_f814c9373cad46a3a77f3961d783cddb"
            ],
            "layout": "IPY_MODEL_37d787fac5874a8aa0342b73a6d7c3df"
          }
        },
        "3f13e910892e4b3b8dffdc546f9d6692": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "ProgressStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        },
        "41500907f8724730b3ac15a7f060d32c": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "FloatProgressModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_cf8516cba4064213a7977b9a62e4c523",
            "max": 300,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_3f13e910892e4b3b8dffdc546f9d6692",
            "value": 300
          }
        },
        "5c1e1bf2b3c84d149d0bb8937d087246": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "5f44aca6e17e40a888a2aab692aae14b": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "649916d14e044e16ac04051feb5cefe3": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "FloatProgressModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_d75197b7aaf44523b467e4a19bb4bbf9",
            "max": 1400,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_37bedfb3159948aa9296986647813cc4",
            "value": 1400
          }
        },
        "694294f380574008a713a7e803f60a98": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "6ab52aa715f7473ebf7bada52805c3c9": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_5f44aca6e17e40a888a2aab692aae14b",
            "placeholder": "​",
            "style": "IPY_MODEL_af6c19f76e58484a8ccbcda95056debc",
            "value": " 300/300 [00:08&lt;00:00, 20.83it/s]"
          }
        },
        "6c25a81bdd5b4979b98bee24bd613816": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "79ca2685fda64168aff4ed04bd7f47eb": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "FloatProgressModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "FloatProgressModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "ProgressView",
            "bar_style": "success",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_e6e1a14c25f8468c9c5dedcd6d184860",
            "max": 300,
            "min": 0,
            "orientation": "horizontal",
            "style": "IPY_MODEL_3278459baebc4a2b87430120b9563eb2",
            "value": 300
          }
        },
        "7d12bfacf7f34e5684a2ad21a609ff9b": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "8f9432d67fc74d4daafa43c7e144127b": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_e6bb239862224171b616751c072677e7",
            "placeholder": "​",
            "style": "IPY_MODEL_143c4bafc3d0415c8e1d435af380aea6",
            "value": " 2000/2000 [00:48&lt;00:00, 70.19it/s]"
          }
        },
        "94655d0a32e6440fb5bfd0853de28d97": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_174ceef81b444b4b9071ad658a5ace66",
            "placeholder": "​",
            "style": "IPY_MODEL_b596dcc695414e2f951b34e911f93970",
            "value": "100%"
          }
        },
        "adfe3d33b08f40348508cfa717d0008c": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HBoxModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_94655d0a32e6440fb5bfd0853de28d97",
              "IPY_MODEL_41500907f8724730b3ac15a7f060d32c",
              "IPY_MODEL_6ab52aa715f7473ebf7bada52805c3c9"
            ],
            "layout": "IPY_MODEL_151b618982024df2bf250c08f2fdc3e6"
          }
        },
        "af6c19f76e58484a8ccbcda95056debc": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "b596dcc695414e2f951b34e911f93970": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "bd44fcb7420d4298b782a6601df0472d": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "bff56443f4ac4861977a3a7f2e66abd3": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HBoxModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HBoxModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HBoxView",
            "box_style": "",
            "children": [
              "IPY_MODEL_e017afc5b5c847c4a3360bf5bb0d1f6e",
              "IPY_MODEL_79ca2685fda64168aff4ed04bd7f47eb",
              "IPY_MODEL_f2a2ab3d7a804129a2ca662875e7fc04"
            ],
            "layout": "IPY_MODEL_0f2a5d2bfa024fceb2a809017ab4713d"
          }
        },
        "c0e082767739480da73bd1c54c6f16db": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "c69fc550f2344cac92607cc70c1fbd30": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_f0659f1993e34924817f779df5761b41",
            "placeholder": "​",
            "style": "IPY_MODEL_da534940a0234c4497f484fa36c2c86a",
            "value": "100%"
          }
        },
        "cf8516cba4064213a7977b9a62e4c523": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "d75197b7aaf44523b467e4a19bb4bbf9": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "da534940a0234c4497f484fa36c2c86a": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "dbe786c66e134f44b67a5e4b63aca3d5": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "DescriptionStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "DescriptionStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "description_width": ""
          }
        },
        "e017afc5b5c847c4a3360bf5bb0d1f6e": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_694294f380574008a713a7e803f60a98",
            "placeholder": "​",
            "style": "IPY_MODEL_dbe786c66e134f44b67a5e4b63aca3d5",
            "value": "100%"
          }
        },
        "e6bb239862224171b616751c072677e7": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "e6e1a14c25f8468c9c5dedcd6d184860": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "f0659f1993e34924817f779df5761b41": {
          "model_module": "@jupyter-widgets/base",
          "model_module_version": "1.2.0",
          "model_name": "LayoutModel",
          "state": {
            "_model_module": "@jupyter-widgets/base",
            "_model_module_version": "1.2.0",
            "_model_name": "LayoutModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "LayoutView",
            "align_content": null,
            "align_items": null,
            "align_self": null,
            "border": null,
            "bottom": null,
            "display": null,
            "flex": null,
            "flex_flow": null,
            "grid_area": null,
            "grid_auto_columns": null,
            "grid_auto_flow": null,
            "grid_auto_rows": null,
            "grid_column": null,
            "grid_gap": null,
            "grid_row": null,
            "grid_template_areas": null,
            "grid_template_columns": null,
            "grid_template_rows": null,
            "height": null,
            "justify_content": null,
            "justify_items": null,
            "left": null,
            "margin": null,
            "max_height": null,
            "max_width": null,
            "min_height": null,
            "min_width": null,
            "object_fit": null,
            "object_position": null,
            "order": null,
            "overflow": null,
            "overflow_x": null,
            "overflow_y": null,
            "padding": null,
            "right": null,
            "top": null,
            "visibility": null,
            "width": null
          }
        },
        "f2a2ab3d7a804129a2ca662875e7fc04": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_5c1e1bf2b3c84d149d0bb8937d087246",
            "placeholder": "​",
            "style": "IPY_MODEL_bd44fcb7420d4298b782a6601df0472d",
            "value": " 300/300 [00:08&lt;00:00, 37.54it/s]"
          }
        },
        "f814c9373cad46a3a77f3961d783cddb": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "HTMLModel",
          "state": {
            "_dom_classes": [],
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "HTMLModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/controls",
            "_view_module_version": "1.5.0",
            "_view_name": "HTMLView",
            "description": "",
            "description_tooltip": null,
            "layout": "IPY_MODEL_7d12bfacf7f34e5684a2ad21a609ff9b",
            "placeholder": "​",
            "style": "IPY_MODEL_300e1ca78baa4457b906f58170b19d06",
            "value": " 1400/1400 [00:42&lt;00:00, 35.21it/s]"
          }
        },
        "f94d417290f44f0bbb844c652f8dfe93": {
          "model_module": "@jupyter-widgets/controls",
          "model_module_version": "1.5.0",
          "model_name": "ProgressStyleModel",
          "state": {
            "_model_module": "@jupyter-widgets/controls",
            "_model_module_version": "1.5.0",
            "_model_name": "ProgressStyleModel",
            "_view_count": null,
            "_view_module": "@jupyter-widgets/base",
            "_view_module_version": "1.2.0",
            "_view_name": "StyleView",
            "bar_color": null,
            "description_width": ""
          }
        }
      }
    }
  },
  "nbformat": 4,
  "nbformat_minor": 0
}