automatic_essay_grading.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "accelerator": "GPU",
    "colab": {
      "name": "automatic_essay_grading.ipynb",
      "provenance": [],
      "collapsed_sections": [],
      "machine_shape": "hm"
    },
    "kernelspec": {
      "display_name": "Python 3",
      "language": "python",
      "name": "python3"
    },
    "language_info": {
      "codemirror_mode": {
        "name": "ipython",
        "version": 3
      },
      "file_extension": ".py",
      "mimetype": "text/x-python",
      "name": "python",
      "nbconvert_exporter": "python",
      "pygments_lexer": "ipython3",
      "version": "3.7.4"
    }
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "RcFR8EacppUp"
      },
      "source": [
        "# Automatic Essay Grading and Feedback\n",
        " - Patrick Myers, Gaurav Kumar Jindal, Sanchit Sinha, Aniruddha Dave"
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Ic8F0gMXUFm4",
        "colab_type": "text"
      },
      "source": [
        "# Importing Libraries:\n",
        " - We first import all necessary libraries for our code."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "uR5mEik19lj5",
        "outputId": "08ca27d6-d058-4eca-bde7-d310e4e8ea45",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 86
        }
      },
      "source": [
        "# import libraries\n",
        "import warnings\n",
        "from sklearn import model_selection\n",
        "from sklearn.metrics import mean_squared_error\n",
        "import torch\n",
        "import torch.nn as nn\n",
        "import torch.nn.functional as F\n",
        "from torch import optim\n",
        "from torch.utils.data import TensorDataset, DataLoader\n",
        "import math\n",
        "#nltk.download()\n",
        "import nltk\n",
        "from nltk.corpus import stopwords\n",
        "from nltk.tokenize import word_tokenize\n",
        "import numpy as np\n",
        "import pandas as pd\n",
        "nltk.download('stopwords')\n",
        "nltk.download('punkt')\n",
        "\n",
        "import matplotlib.pyplot as plt\n",
        "\n",
        "# Supress unnecessary warnings so that presentation looks clean\n",
        "warnings.filterwarnings('ignore')\n",
        "\n",
        "google_colab=False"
      ],
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "[nltk_data] Downloading package stopwords to /root/nltk_data...\n",
            "[nltk_data]   Package stopwords is already up-to-date!\n",
            "[nltk_data] Downloading package punkt to /root/nltk_data...\n",
            "[nltk_data]   Package punkt is already up-to-date!\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "dsyHONuffyvp",
        "outputId": "89f3a722-4eb1-491c-a5a8-14bb475b350e",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 124
        }
      },
      "source": [
        "if google_colab:\n",
        "    # Google Colab stuff\n",
        "    from google.colab import drive\n",
        "    drive.mount('/content/drive')"
      ],
      "execution_count": 5,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Go to this URL in a browser: https://accounts.google.com/o/oauth2/auth?client_id=947318989803-6bn6qk8qdgf4n4g3pfee6491hc0brc4i.apps.googleusercontent.com&redirect_uri=urn%3aietf%3awg%3aoauth%3a2.0%3aoob&response_type=code&scope=email%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdocs.test%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive%20https%3a%2f%2fwww.googleapis.com%2fauth%2fdrive.photos.readonly%20https%3a%2f%2fwww.googleapis.com%2fauth%2fpeopleapi.readonly\n",
            "\n",
            "Enter your authorization code:\n",
            "··········\n",
            "Mounted at /content/drive\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "o8GdLUX9UFnE",
        "colab_type": "text"
      },
      "source": [
        "## Data Preprocessing:\n",
        " - We establish a class to handle loading, preprocessing, and chunking of our dataset."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "OQDlhxcSgud1",
        "outputId": "4c020d67-fafd-44a6-d487-3b7e0775ebae",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 52
        }
      },
      "source": [
        "class Dataset:\n",
        "    def __init__(self, path):\n",
        "\n",
        "        if google_colab:\n",
        "            self.df = pd.read_csv(\"/content/drive/My Drive/Colab Notebooks/NLP/data/training_data.csv\")\n",
        "        else:\n",
        "            self.df = pd.read_csv(\"training_data.csv\")\n",
        "        self.data = self.df.to_numpy()\n",
        "\n",
        "        self.stopwords = set(stopwords.words('english'))\n",
        "        self.essay_id = self.data[:,0]\n",
        "        self.text = self.data[:,1]\n",
        "        self.scores = self.data[:,2:8]\n",
        "        self.new_data = []\n",
        "        self.new_scores = []\n",
        "        self.vocab = set()\n",
        "        self.word_to_id = None\n",
        "\n",
        "    def preprocess(self):\n",
        "        for i in range(len(self.essay_id)):\n",
        "            text = self.text[i].lower()\n",
        "            text = \" \".join([word for word in text.split() if '@' not in word])\n",
        "            text = word_tokenize(text)\n",
        "            text = [word for word in text if word not in self.stopwords]\n",
        "            self.text[i] = text \n",
        "\n",
        "    def create_vocab(self):\n",
        "        for line in self.text:\n",
        "            for word in line:\n",
        "                self.vocab.add(word)\n",
        "\n",
        "        self.vocab = sorted(list(self.vocab))\n",
        "        self.word_to_id = {word:i for i, word in enumerate(self.vocab)}\n",
        "\n",
        "    def text_num(self):\n",
        "        for i, line in enumerate(self.text):\n",
        "            x = []\n",
        "            for word in line:\n",
        "                x.append(self.word_to_id[word])\n",
        "            self.text[i] = x\n",
        "\n",
        "    def create_chunks(self):\n",
        "        for idx in range(len(self.essay_id)):\n",
        "            ess = self.text[idx]\n",
        "            n = len(ess)\n",
        "            self.new_data.append([ess[:n//3]])\n",
        "            self.new_data.append([ess[n//3:2*n//3]])\n",
        "            self.new_data.append([ess[2*n//3:]])\n",
        "            self.new_scores.append(self.scores[idx])\n",
        "            self.new_scores.append(self.scores[idx])\n",
        "            self.new_scores.append( self.scores[idx])\n",
        "\n",
        "import pandas as pd\n",
        "import nltk\n",
        "#nltk.download()\n",
        "from nltk.corpus import stopwords\n",
        "from nltk.tokenize import word_tokenize\n",
        "import numpy as np\n",
        "\n",
        "\n",
        "dataset = Dataset(\"/content/drive/My Drive/Colab Notebooks/NLP/data/training_data.csv\")\n",
        "print(dataset.text.shape)\n",
        "dataset.preprocess()\n",
        "dataset.create_vocab()\n",
        "# print(len(dataset.vocab))\n",
        "# print(dataset.vocab[:10])\n",
        "# print(dataset.word_to_id['!'])\n",
        "# dataset.create_chunks()\n",
        "# print(dataset.new_data.shape)\n",
        "# print(dataset.text[0])\n",
        "\"\"\"\n",
        "print(len(dataset.new_data), len(dataset.essay_id))\n",
        "print(\"0: \", dataset.new_data[0, 0])\n",
        "print(\"1: \", dataset.new_data[1, 0])\n",
        "print(\"2: \", dataset.new_data[2, 0])\n",
        "print(dataset.text[0])\n",
        "\"\"\""
      ],
      "execution_count": 6,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(723,)\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "'\\nprint(len(dataset.new_data), len(dataset.essay_id))\\nprint(\"0: \", dataset.new_data[0, 0])\\nprint(\"1: \", dataset.new_data[1, 0])\\nprint(\"2: \", dataset.new_data[2, 0])\\nprint(dataset.text[0])\\n'"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 6
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "XbdvYqDTRDJq",
        "colab": {}
      },
      "source": [
        "dataset.text_num()\n",
        "dataset.create_chunks()\n",
        "train_text = dataset.new_data\n",
        "\n",
        "max_len = 0\n",
        "for row in train_text:\n",
        "  leng = len(row[0])\n",
        "  if leng > max_len:\n",
        "    max_len = leng\n",
        "\n",
        "train_text = [train_text[i][0] for i in range(len(train_text))]\n",
        "\n",
        "# Padding with space (0)\n",
        "for i in range(len(train_text)):\n",
        "    while len(train_text[i])<=max_len:\n",
        "        train_text[i].append(0)\n",
        "\n",
        "\n",
        "train_label = dataset.new_scores\n",
        "len(train_label)\n",
        "\n",
        "from sklearn import model_selection\n",
        "train_x, valid_x, train_y, valid_y = model_selection.train_test_split(train_text, train_label)\n",
        "\n",
        "train_x = np.array(train_x)\n",
        "valid_x = np.array(valid_x)\n",
        "train_y = np.array(train_y)\n",
        "valid_y = np.array(valid_y)\n",
        "\n",
        "vocab_len = len(dataset.vocab)\n",
        "\n",
        "device = torch.device('cuda:0')# if torch.cuda.is_available() else 'cpu')"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "k6ADZ04WWqnf",
        "colab": {}
      },
      "source": [
        "batch_size=1\n",
        "def load_category_data(cat, train_x, valid_x, train_y, valid_y):\n",
        "    if google_colab:\n",
        "        train_x = torch.from_numpy(np.array(train_x)).to(device)\n",
        "        train_y = torch.from_numpy(np.array(train_y)[:,cat].astype(float)).to(device)\n",
        "        valid_x = torch.from_numpy(np.array(valid_x)).to(device)\n",
        "        valid_y = torch.from_numpy(np.array(valid_y)[:,cat].astype(float)).to(device)\n",
        "    else:\n",
        "        train_x = torch.from_numpy(np.array(train_x)).to(torch.int64).to(device)\n",
        "        train_y = torch.from_numpy(np.array(train_y)[:,cat].astype(float)).to(torch.int64).to(device)\n",
        "        valid_x = torch.from_numpy(np.array(valid_x)).to(torch.int64).to(device)\n",
        "        valid_y = torch.from_numpy(np.array(valid_y)[:,cat].astype(float)).to(torch.int64).to(device)\n",
        "\n",
        "    train_loader = DataLoader(TensorDataset(train_x, train_y), batch_size = 1, shuffle = False)\n",
        "    valid_loader = DataLoader(TensorDataset(valid_x, valid_y), batch_size = 1, shuffle = False)\n",
        "    \n",
        "    return train_loader, valid_loader"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "Z47VHgg4UFnN",
        "colab_type": "text"
      },
      "source": [
        "## PyTorch LSTM Model:\n",
        " - Here is our LSTM model and helper functions for training, validation, and calculating accuracy. We round predictions to the nearest integer in order to determine accuracy."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "2FFL2oRnVW2G",
        "colab": {}
      },
      "source": [
        "class LSTM_Model(nn.Module):\n",
        "    def __init__(self, vocab_size, embedding_size, hidden_num, output_num, layer_num):\n",
        "        super().__init__()\n",
        "        self.vocab_size = vocab_size\n",
        "        self.layer_num = layer_num\n",
        "        self.hidden_num = hidden_num\n",
        "        self.output_num = output_num\n",
        "            \n",
        "        self.embedding = nn.Embedding(vocab_size, embedding_size)\n",
        "        self.lstm = nn.LSTM(embedding_size, hidden_num, num_layers=layer_num)\n",
        "        self.fc = nn.Linear(hidden_num,output_num)\n",
        "        self.relu = nn.ReLU()\n",
        "\n",
        "    def make_hidden(self):\n",
        "        return (torch.randn(self.layer_num, max_len+1, self.hidden_num).cuda(), torch.randn(self.layer_num, max_len+1, self.hidden_num).cuda())\n",
        "        \n",
        "    def forward(self, word_seq, hidden):\n",
        "        word_emb = self.embedding(word_seq)\n",
        "        lstm_out, hidden = self.lstm(word_emb, hidden)\n",
        "        lstm_out = lstm_out.contiguous().view(-1, self.hidden_num)[-1]\n",
        "        fc_out = self.fc(lstm_out)\n",
        "        relu_out = self.relu(fc_out)\n",
        "\n",
        "        return relu_out, hidden"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "Ta36AU1WLLMl",
        "colab": {}
      },
      "source": [
        "def get_accuracy(predictions):\n",
        "    correct = 0\n",
        "    total = 0\n",
        "    for batch in predictions:\n",
        "        for item in batch:\n",
        "            if int(round(item[0])) == int(item[1]):\n",
        "                correct += 1\n",
        "            total += 1\n",
        "    return correct/total"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "52hmyt9hVW_h",
        "colab": {}
      },
      "source": [
        "def train(data_loader, classifier, loss_function, optimizer):\n",
        "    classifier.train()\n",
        "    loss = 0\n",
        "    losses = []\n",
        "    prediction_list = []\n",
        "    accuracy = 0\n",
        "    accuracies = []\n",
        "    for i, (texts, labels) in enumerate(data_loader):\n",
        "#         skipping broken data entry\n",
        "        if texts[0][0] != 3785 or labels != 5:\n",
        "            if(texts.shape[0] != batch_size):\n",
        "                break\n",
        "                \n",
        "            labels = labels.float().cuda()\n",
        "            texts = texts.cuda()\n",
        "            optimizer.zero_grad()\n",
        "\n",
        "            hidden = rnn_model.make_hidden()\n",
        "            predictions, hidden = classifier(texts, hidden)\n",
        "            loss = loss_function(predictions, labels)\n",
        "\n",
        "            loss.backward()\n",
        "            optimizer.step()\n",
        "            losses.append(loss.item()) \n",
        "            prediction_list.append([(predictions[i].item(), labels[i].item()) for i in range(len(predictions))])\n",
        "      \n",
        "    return prediction_list, sum(losses)/len(losses)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "zpn2meSknDHB",
        "colab": {}
      },
      "source": [
        "def validation(data_loader, classifier, loss_function):\n",
        "    classifier.eval()\n",
        "    loss = 0\n",
        "    losses = []\n",
        "    prediction_list = []\n",
        "    accuracy = 0\n",
        "    accuracies = []\n",
        "    for i, (texts, labels) in enumerate(data_loader):\n",
        "#         skipping broken data entry\n",
        "        if texts[0][0] != 3785 or labels != 5:\n",
        "            if(texts.shape[0] != batch_size):\n",
        "                break\n",
        "                \n",
        "            labels = labels.float().cuda()\n",
        "            texts = texts.cuda()\n",
        "            \n",
        "            hidden = rnn_model.make_hidden()\n",
        "            predictions, hidden = classifier(texts, hidden)\n",
        "            loss = loss_function(predictions, labels)\n",
        "\n",
        "            losses.append(loss.item()) \n",
        "            prediction_list.append([(predictions[i].item(), labels[i].item()) for i in range(len(predictions))])\n",
        "            \n",
        "    return prediction_list, sum(losses)/len(losses)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "OLwWZ9TuUFna",
        "colab_type": "text"
      },
      "source": [
        "## Training Our Model:\n",
        " - We train six separate models, one for each of the rubric categories, and store their loss, accuracy, and RMSE values in lists."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "nmt60RhcVW92",
        "outputId": "556d6a93-d8e1-4ccb-d649-f3f732ea6f0b",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "trn_loss_list = []\n",
        "val_loss_list = []\n",
        "trn_acc_list = []\n",
        "val_acc_list = []\n",
        "trn_rmse_list = []\n",
        "val_rmse_list = []\n",
        "\n",
        "model_list = []\n",
        "\n",
        "for cat in range(6):\n",
        "    trn_loss = []\n",
        "    val_loss = []\n",
        "    trn_acc = []\n",
        "    val_acc = []\n",
        "    trn_rmse = []\n",
        "    val_rmse = []\n",
        "    \n",
        "    \n",
        "    train_loader, valid_loader = load_category_data(cat, train_x, valid_x, train_y, valid_y)\n",
        "    n_vocab = vocab_len\n",
        "    n_embed = 100\n",
        "    n_hidden = 32\n",
        "    n_output = 1\n",
        "    n_layers = 2\n",
        "\n",
        "    rnn_model = LSTM_Model(n_vocab, n_embed, n_hidden, n_output, n_layers)\n",
        "    rnn_model.cuda()\n",
        "    loss_function = nn.MSELoss()\n",
        "    optimizer = torch.optim.Adam(rnn_model.parameters(), lr=.0001)\n",
        "\n",
        "    epochs = 10\n",
        "\n",
        "    training_losses = []\n",
        "    validation_losses = []\n",
        "    for epoch in range(0, epochs):\n",
        "        print(\"category:\", cat, \"epoch:\", epoch + 1)\n",
        "        train_predictions, training_loss = train(train_loader, rnn_model, loss_function, optimizer)\n",
        "        validation_predictions, validation_loss = validation(valid_loader, rnn_model, loss_function)\n",
        "        print(\"training_loss:\", training_loss)\n",
        "        print(\"validation_loss:\", validation_loss)\n",
        "        trn_loss.append(training_loss)\n",
        "        val_loss.append(validation_loss)\n",
        "\n",
        "        training_accuracy = get_accuracy(train_predictions)\n",
        "        validation_accuracy = get_accuracy(validation_predictions)\n",
        "        print(\"train_accuracy:\", training_accuracy)\n",
        "        print(\"validation_accuracy:\", validation_accuracy)\n",
        "        trn_acc.append(training_accuracy)\n",
        "        val_acc.append(validation_accuracy)\n",
        "\n",
        "        training_rmse = math.sqrt(mean_squared_error([pred[0][0] for pred in train_predictions], [pred[0][1] for pred in train_predictions]))\n",
        "        validation_rmse = math.sqrt(mean_squared_error([pred[0][0] for pred in validation_predictions], [pred[0][1] for pred in validation_predictions]))\n",
        "        print(\"training_rmse\", training_rmse)\n",
        "        print(\"validation_rmse\", validation_rmse)\n",
        "        trn_rmse.append(training_rmse)\n",
        "        val_rmse.append(validation_rmse)\n",
        "        \n",
        "        print(\"\\n\")\n",
        "        \n",
        "    trn_loss_list.append(trn_loss)\n",
        "    val_loss_list.append(val_loss)\n",
        "    trn_acc_list.append(trn_acc)\n",
        "    val_acc_list.append(val_acc)\n",
        "    trn_rmse_list.append(trn_rmse)\n",
        "    val_rmse_list.append(val_rmse)\n",
        "    \n",
        "    model_list.append(rnn_model)"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "category: 0 epoch: 1\n",
            "training_loss: 6.787516865047649\n",
            "validation_loss: 0.7979272252045888\n",
            "train_accuracy: 0.09471094710947109\n",
            "validation_accuracy: 0.3020257826887661\n",
            "training_rmse 2.605286331418289\n",
            "validation_rmse 0.8932677233916992\n",
            "\n",
            "\n",
            "category: 0 epoch: 2\n",
            "training_loss: 0.5496929801467934\n",
            "validation_loss: 0.5476000995971388\n",
            "train_accuracy: 0.498769987699877\n",
            "validation_accuracy: 0.5322283609576427\n",
            "training_rmse 0.7414128271649104\n",
            "validation_rmse 0.7400000674408113\n",
            "\n",
            "\n",
            "category: 0 epoch: 3\n",
            "training_loss: 0.5099897414853719\n",
            "validation_loss: 0.5393421434036574\n",
            "train_accuracy: 0.5615006150061501\n",
            "validation_accuracy: 0.5506445672191529\n",
            "training_rmse 0.7141356609654339\n",
            "validation_rmse 0.7343991717449826\n",
            "\n",
            "\n",
            "category: 0 epoch: 4\n",
            "training_loss: 0.5068431735543172\n",
            "validation_loss: 0.5519455597207678\n",
            "train_accuracy: 0.5541205412054121\n",
            "validation_accuracy: 0.5488029465930019\n",
            "training_rmse 0.7119291917809052\n",
            "validation_rmse 0.7429303863473251\n",
            "\n",
            "\n",
            "category: 0 epoch: 5\n",
            "training_loss: 0.5152475570970503\n",
            "validation_loss: 0.5318904268087109\n",
            "train_accuracy: 0.5535055350553506\n",
            "validation_accuracy: 0.5543278084714549\n",
            "training_rmse 0.7178074651554438\n",
            "validation_rmse 0.7293081846399924\n",
            "\n",
            "\n",
            "category: 0 epoch: 6\n",
            "training_loss: 0.5104407638612184\n",
            "validation_loss: 0.5516258236323097\n",
            "train_accuracy: 0.5547355473554736\n",
            "validation_accuracy: 0.5524861878453039\n",
            "training_rmse 0.714451373156271\n",
            "validation_rmse 0.7427151700237383\n",
            "\n",
            "\n",
            "category: 0 epoch: 7\n",
            "training_loss: 0.5155579924597298\n",
            "validation_loss: 0.5313389411570871\n",
            "train_accuracy: 0.5547355473554736\n",
            "validation_accuracy: 0.5580110497237569\n",
            "training_rmse 0.7180236707816351\n",
            "validation_rmse 0.7289299975058311\n",
            "\n",
            "\n",
            "category: 0 epoch: 8\n",
            "training_loss: 0.5115384791219966\n",
            "validation_loss: 0.5279641537335367\n",
            "train_accuracy: 0.5584255842558425\n",
            "validation_accuracy: 0.567219152854512\n",
            "training_rmse 0.7152191823247201\n",
            "validation_rmse 0.7266114192980812\n",
            "\n",
            "\n",
            "category: 0 epoch: 9\n",
            "training_loss: 0.5040736388526913\n",
            "validation_loss: 0.5424180255523621\n",
            "train_accuracy: 0.5608856088560885\n",
            "validation_accuracy: 0.5616942909760589\n",
            "training_rmse 0.7099814355490093\n",
            "validation_rmse 0.7364903431170993\n",
            "\n",
            "\n",
            "category: 0 epoch: 10\n",
            "training_loss: 0.5092593566918443\n",
            "validation_loss: 0.5331622945902865\n",
            "train_accuracy: 0.5565805658056581\n",
            "validation_accuracy: 0.569060773480663\n",
            "training_rmse 0.7136241001342125\n",
            "validation_rmse 0.7301796314758774\n",
            "\n",
            "\n",
            "category: 1 epoch: 1\n",
            "training_loss: 7.14750364574547\n",
            "validation_loss: 0.7138039828965648\n",
            "train_accuracy: 0.1002460024600246\n",
            "validation_accuracy: 0.31307550644567217\n",
            "training_rmse 2.6734815559179665\n",
            "validation_rmse 0.8448692103876855\n",
            "\n",
            "\n",
            "category: 1 epoch: 2\n",
            "training_loss: 0.5332094648405327\n",
            "validation_loss: 0.5465245463500626\n",
            "train_accuracy: 0.503690036900369\n",
            "validation_accuracy: 0.49171270718232046\n",
            "training_rmse 0.7302119319914884\n",
            "validation_rmse 0.7392729843497531\n",
            "\n",
            "\n",
            "category: 1 epoch: 3\n",
            "training_loss: 0.4982233611471826\n",
            "validation_loss: 0.5328112014613333\n",
            "train_accuracy: 0.5571955719557196\n",
            "validation_accuracy: 0.5414364640883977\n",
            "training_rmse 0.7058493894918938\n",
            "validation_rmse 0.7299391761422306\n",
            "\n",
            "\n",
            "category: 1 epoch: 4\n",
            "training_loss: 0.4980997539034365\n",
            "validation_loss: 0.5155858903563526\n",
            "train_accuracy: 0.5633456334563346\n",
            "validation_accuracy: 0.5340699815837937\n",
            "training_rmse 0.7057618261309947\n",
            "validation_rmse 0.7180430979449554\n",
            "\n",
            "\n",
            "category: 1 epoch: 5\n",
            "training_loss: 0.4831238197158538\n",
            "validation_loss: 0.5282343950630277\n",
            "train_accuracy: 0.5731857318573186\n",
            "validation_accuracy: 0.5469613259668509\n",
            "training_rmse 0.695071090152994\n",
            "validation_rmse 0.7267973557961708\n",
            "\n",
            "\n",
            "category: 1 epoch: 6\n",
            "training_loss: 0.4913859695172504\n",
            "validation_loss: 0.5143829664604783\n",
            "train_accuracy: 0.5731857318573186\n",
            "validation_accuracy: 0.5469613259668509\n",
            "training_rmse 0.7009892794638591\n",
            "validation_rmse 0.7172049668472108\n",
            "\n",
            "\n",
            "category: 1 epoch: 7\n",
            "training_loss: 0.48545934469306873\n",
            "validation_loss: 0.505520028173649\n",
            "train_accuracy: 0.5768757687576875\n",
            "validation_accuracy: 0.5469613259668509\n",
            "training_rmse 0.6967491266869367\n",
            "validation_rmse 0.7109993160592964\n",
            "\n",
            "\n",
            "category: 1 epoch: 8\n",
            "training_loss: 0.4859575778568064\n",
            "validation_loss: 0.5050555016831744\n",
            "train_accuracy: 0.5750307503075031\n",
            "validation_accuracy: 0.5616942909760589\n",
            "training_rmse 0.6971065762656328\n",
            "validation_rmse 0.7106725708296467\n",
            "\n",
            "\n",
            "category: 1 epoch: 9\n",
            "training_loss: 0.48782890733588613\n",
            "validation_loss: 0.5161151107262157\n",
            "train_accuracy: 0.5768757687576875\n",
            "validation_accuracy: 0.5488029465930019\n",
            "training_rmse 0.6984474979919886\n",
            "validation_rmse 0.7184115196781911\n",
            "\n",
            "\n",
            "category: 1 epoch: 10\n",
            "training_loss: 0.4850953178950869\n",
            "validation_loss: 0.5213294211298255\n",
            "train_accuracy: 0.5817958179581796\n",
            "validation_accuracy: 0.5432780847145487\n",
            "training_rmse 0.6964878447842062\n",
            "validation_rmse 0.7220314549293887\n",
            "\n",
            "\n",
            "category: 2 epoch: 1\n",
            "training_loss: 6.067703695323068\n",
            "validation_loss: 0.5732798015539979\n",
            "train_accuracy: 0.11377613776137761\n",
            "validation_accuracy: 0.5469613259668509\n",
            "training_rmse 2.4632709357776603\n",
            "validation_rmse 0.7571524296329574\n",
            "\n",
            "\n",
            "category: 2 epoch: 2\n",
            "training_loss: 0.4442674411222199\n",
            "validation_loss: 0.4544751342464183\n",
            "train_accuracy: 0.6365313653136532\n",
            "validation_accuracy: 0.6482504604051565\n",
            "training_rmse 0.6665339007308291\n",
            "validation_rmse 0.6741477087526133\n",
            "\n",
            "\n",
            "category: 2 epoch: 3\n",
            "training_loss: 0.4330818932584567\n",
            "validation_loss: 0.4338667617486949\n",
            "train_accuracy: 0.6562115621156211\n",
            "validation_accuracy: 0.6427255985267035\n",
            "training_rmse 0.6580895785606115\n",
            "validation_rmse 0.6586856316509161\n",
            "\n",
            "\n",
            "category: 2 epoch: 4\n",
            "training_loss: 0.42224055513704\n",
            "validation_loss: 0.4494653201338511\n",
            "train_accuracy: 0.6592865928659286\n",
            "validation_accuracy: 0.6482504604051565\n",
            "training_rmse 0.6498003964270359\n",
            "validation_rmse 0.6704217480104776\n",
            "\n",
            "\n",
            "category: 2 epoch: 5\n",
            "training_loss: 0.41665316948631004\n",
            "validation_loss: 0.4498289336914912\n",
            "train_accuracy: 0.6605166051660517\n",
            "validation_accuracy: 0.6445672191528545\n",
            "training_rmse 0.6454867699626484\n",
            "validation_rmse 0.6706928772165089\n",
            "\n",
            "\n",
            "category: 2 epoch: 6\n",
            "training_loss: 0.4222738673420139\n",
            "validation_loss: 0.44490841637982265\n",
            "train_accuracy: 0.6611316113161132\n",
            "validation_accuracy: 0.6500920810313076\n",
            "training_rmse 0.6498260285962515\n",
            "validation_rmse 0.6670145546092484\n",
            "\n",
            "\n",
            "category: 2 epoch: 7\n",
            "training_loss: 0.421884828613632\n",
            "validation_loss: 0.4356673966015643\n",
            "train_accuracy: 0.6599015990159902\n",
            "validation_accuracy: 0.6500920810313076\n",
            "training_rmse 0.6495266184751229\n",
            "validation_rmse 0.6600510544523827\n",
            "\n",
            "\n",
            "category: 2 epoch: 8\n",
            "training_loss: 0.41585790978204357\n",
            "validation_loss: 0.433512751026248\n",
            "train_accuracy: 0.6617466174661747\n",
            "validation_accuracy: 0.6500920810313076\n",
            "training_rmse 0.6448704598908723\n",
            "validation_rmse 0.6584168525703423\n",
            "\n",
            "\n",
            "category: 2 epoch: 9\n",
            "training_loss: 0.4169170524605666\n",
            "validation_loss: 0.4326563263911813\n",
            "train_accuracy: 0.6611316113161132\n",
            "validation_accuracy: 0.6500920810313076\n",
            "training_rmse 0.6456911435546003\n",
            "validation_rmse 0.6577661638525129\n",
            "\n",
            "\n",
            "category: 2 epoch: 10\n",
            "training_loss: 0.4144843774769364\n",
            "validation_loss: 0.4393495857207425\n",
            "train_accuracy: 0.6611316113161132\n",
            "validation_accuracy: 0.6500920810313076\n",
            "training_rmse 0.643804611069932\n",
            "validation_rmse 0.6628345095259488\n",
            "\n",
            "\n",
            "category: 3 epoch: 1\n",
            "training_loss: 7.300654385405058\n",
            "validation_loss: 0.7168018076904152\n",
            "train_accuracy: 0.07564575645756458\n",
            "validation_accuracy: 0.27440147329650094\n",
            "training_rmse 2.7019723152783386\n",
            "validation_rmse 0.8466414885698522\n",
            "\n",
            "\n",
            "category: 3 epoch: 2\n",
            "training_loss: 0.4636316096327369\n",
            "validation_loss: 0.4341306632120308\n",
            "train_accuracy: 0.6076260762607626\n",
            "validation_accuracy: 0.6482504604051565\n",
            "training_rmse 0.6809049925896652\n",
            "validation_rmse 0.6588859255284927\n",
            "\n",
            "\n",
            "category: 3 epoch: 3\n",
            "training_loss: 0.4335250213488414\n",
            "validation_loss: 0.4344500995733856\n",
            "train_accuracy: 0.6617466174661747\n",
            "validation_accuracy: 0.6519337016574586\n",
            "training_rmse 0.6584261702595515\n",
            "validation_rmse 0.6591282884628521\n",
            "\n",
            "\n",
            "category: 3 epoch: 4\n",
            "training_loss: 0.4176669338135825\n",
            "validation_loss: 0.43702088424515334\n",
            "train_accuracy: 0.6623616236162362\n",
            "validation_accuracy: 0.6519337016574586\n",
            "training_rmse 0.6462715634929569\n",
            "validation_rmse 0.6610755512418793\n",
            "\n",
            "\n",
            "category: 3 epoch: 5\n",
            "training_loss: 0.4166912299684383\n",
            "validation_loss: 0.42877293233047115\n",
            "train_accuracy: 0.6678966789667896\n",
            "validation_accuracy: 0.6519337016574586\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "stream",
          "text": [
            "training_rmse 0.645516250740042\n",
            "validation_rmse 0.6548075530998541\n",
            "\n",
            "\n",
            "category: 3 epoch: 6\n",
            "training_loss: 0.41364426743202876\n",
            "validation_loss: 0.4376144431223797\n",
            "train_accuracy: 0.6654366543665436\n",
            "validation_accuracy: 0.6519337016574586\n",
            "training_rmse 0.6431518226364472\n",
            "validation_rmse 0.6615243311261612\n",
            "\n",
            "\n",
            "category: 3 epoch: 7\n",
            "training_loss: 0.4168746388707678\n",
            "validation_loss: 0.4384407489458239\n",
            "train_accuracy: 0.6648216482164822\n",
            "validation_accuracy: 0.6519337016574586\n",
            "training_rmse 0.6456582983520859\n",
            "validation_rmse 0.6621485859857574\n",
            "\n",
            "\n",
            "category: 3 epoch: 8\n",
            "training_loss: 0.414603421586761\n",
            "validation_loss: 0.4293413471550765\n",
            "train_accuracy: 0.6654366543665436\n",
            "validation_accuracy: 0.6556169429097606\n",
            "training_rmse 0.6438970579072937\n",
            "validation_rmse 0.655241440682818\n",
            "\n",
            "\n",
            "category: 3 epoch: 9\n",
            "training_loss: 0.4068600275569723\n",
            "validation_loss: 0.42622044165413975\n",
            "train_accuracy: 0.6672816728167281\n",
            "validation_accuracy: 0.6519337016574586\n",
            "training_rmse 0.6378558051896089\n",
            "validation_rmse 0.6528556054055236\n",
            "\n",
            "\n",
            "category: 3 epoch: 10\n",
            "training_loss: 0.4162262019876376\n",
            "validation_loss: 0.42727064136223836\n",
            "train_accuracy: 0.6666666666666666\n",
            "validation_accuracy: 0.6556169429097606\n",
            "training_rmse 0.6451559517330583\n",
            "validation_rmse 0.653659422949896\n",
            "\n",
            "\n",
            "category: 4 epoch: 1\n",
            "training_loss: 6.326429024930817\n",
            "validation_loss: 0.7689565822126176\n",
            "train_accuracy: 0.11131611316113162\n",
            "validation_accuracy: 0.3149171270718232\n",
            "training_rmse 2.5152393568261933\n",
            "validation_rmse 0.8769016952986993\n",
            "\n",
            "\n",
            "category: 4 epoch: 2\n",
            "training_loss: 0.5517552732269573\n",
            "validation_loss: 0.5497819559963549\n",
            "train_accuracy: 0.5289052890528906\n",
            "validation_accuracy: 0.5377532228360957\n",
            "training_rmse 0.7428023111331923\n",
            "validation_rmse 0.7414728288967808\n",
            "\n",
            "\n",
            "category: 4 epoch: 3\n",
            "training_loss: 0.5284783020183212\n",
            "validation_loss: 0.5598821037504256\n",
            "train_accuracy: 0.552890528905289\n",
            "validation_accuracy: 0.4972375690607735\n",
            "training_rmse 0.7269651312777181\n",
            "validation_rmse 0.7482527007771531\n",
            "\n",
            "\n",
            "category: 4 epoch: 4\n",
            "training_loss: 0.5284088958032478\n",
            "validation_loss: 0.5544550466688022\n",
            "train_accuracy: 0.5608856088560885\n",
            "validation_accuracy: 0.5303867403314917\n",
            "training_rmse 0.7269173924299277\n",
            "validation_rmse 0.744617380669114\n",
            "\n",
            "\n",
            "category: 4 epoch: 5\n",
            "training_loss: 0.5246856908969981\n",
            "validation_loss: 0.549858772020931\n",
            "train_accuracy: 0.5639606396063961\n",
            "validation_accuracy: 0.5377532228360957\n",
            "training_rmse 0.7243519107411294\n",
            "validation_rmse 0.7415246268474515\n",
            "\n",
            "\n",
            "category: 4 epoch: 6\n",
            "training_loss: 0.5268283718506394\n",
            "validation_loss: 0.5421762120177623\n",
            "train_accuracy: 0.5713407134071341\n",
            "validation_accuracy: 0.5414364640883977\n",
            "training_rmse 0.7258294370073407\n",
            "validation_rmse 0.7363261589717092\n",
            "\n",
            "\n",
            "category: 4 epoch: 7\n",
            "training_loss: 0.5240986892002023\n",
            "validation_loss: 0.536698704777529\n",
            "train_accuracy: 0.5694956949569495\n",
            "validation_accuracy: 0.5359116022099447\n",
            "training_rmse 0.7239466066499991\n",
            "validation_rmse 0.7325972325192797\n",
            "\n",
            "\n",
            "category: 4 epoch: 8\n",
            "training_loss: 0.5185596711459933\n",
            "validation_loss: 0.5519485639229856\n",
            "train_accuracy: 0.5756457564575646\n",
            "validation_accuracy: 0.5322283609576427\n",
            "training_rmse 0.720110874327365\n",
            "validation_rmse 0.7429324083017914\n",
            "\n",
            "\n",
            "category: 4 epoch: 9\n",
            "training_loss: 0.5204001462748601\n",
            "validation_loss: 0.5382614692443258\n",
            "train_accuracy: 0.5756457564575646\n",
            "validation_accuracy: 0.5414364640883977\n",
            "training_rmse 0.7213876538596984\n",
            "validation_rmse 0.7336630491135123\n",
            "\n",
            "\n",
            "category: 4 epoch: 10\n",
            "training_loss: 0.515163847456941\n",
            "validation_loss: 0.5418220082260188\n",
            "train_accuracy: 0.5787207872078721\n",
            "validation_accuracy: 0.5322283609576427\n",
            "training_rmse 0.7177491536917203\n",
            "validation_rmse 0.7360855987046169\n",
            "\n",
            "\n",
            "category: 5 epoch: 1\n",
            "training_loss: 5.823443317413396\n",
            "validation_loss: 0.6716140323171025\n",
            "train_accuracy: 0.13776137761377613\n",
            "validation_accuracy: 0.42357274401473294\n",
            "training_rmse 2.4131811616112704\n",
            "validation_rmse 0.8195206119566605\n",
            "\n",
            "\n",
            "category: 5 epoch: 2\n",
            "training_loss: 0.518847596759773\n",
            "validation_loss: 0.5265733823894126\n",
            "train_accuracy: 0.44772447724477243\n",
            "validation_accuracy: 0.4530386740331492\n",
            "training_rmse 0.7203107650375447\n",
            "validation_rmse 0.7256537615513878\n",
            "\n",
            "\n",
            "category: 5 epoch: 3\n",
            "training_loss: 0.5049806307694723\n",
            "validation_loss: 0.5311184875104529\n",
            "train_accuracy: 0.45387453874538747\n",
            "validation_accuracy: 0.43646408839779005\n",
            "training_rmse 0.7106198922948694\n",
            "validation_rmse 0.7287787651880588\n",
            "\n",
            "\n",
            "category: 5 epoch: 4\n",
            "training_loss: 0.49682816905616745\n",
            "validation_loss: 0.5255542217503417\n",
            "train_accuracy: 0.45571955719557194\n",
            "validation_accuracy: 0.430939226519337\n",
            "training_rmse 0.7048603893072322\n",
            "validation_rmse 0.7249511851354246\n",
            "\n",
            "\n",
            "category: 5 epoch: 5\n",
            "training_loss: 0.4959006852004677\n",
            "validation_loss: 0.5220681237773781\n",
            "train_accuracy: 0.46494464944649444\n",
            "validation_accuracy: 0.4604051565377532\n",
            "training_rmse 0.7042021623889497\n",
            "validation_rmse 0.7225428186266913\n",
            "\n",
            "\n",
            "category: 5 epoch: 6\n",
            "training_loss: 0.4886424916316406\n",
            "validation_loss: 0.526371075303515\n",
            "train_accuracy: 0.48093480934809346\n",
            "validation_accuracy: 0.43278084714548803\n",
            "training_rmse 0.6990296786268388\n",
            "validation_rmse 0.7255143526733221\n",
            "\n",
            "\n",
            "category: 5 epoch: 7\n",
            "training_loss: 0.49165188069403776\n",
            "validation_loss: 0.5083034288045534\n",
            "train_accuracy: 0.46371463714637146\n",
            "validation_accuracy: 0.4622467771639042\n",
            "training_rmse 0.7011789223348689\n",
            "validation_rmse 0.7129540154188695\n",
            "\n",
            "\n",
            "category: 5 epoch: 8\n",
            "training_loss: 0.4926300338672206\n",
            "validation_loss: 0.5135972443727937\n",
            "train_accuracy: 0.44341943419434193\n",
            "validation_accuracy: 0.4677716390423573\n",
            "training_rmse 0.7018760817710751\n",
            "validation_rmse 0.716656991753556\n",
            "\n",
            "\n",
            "category: 5 epoch: 9\n",
            "training_loss: 0.4953376020260183\n",
            "validation_loss: 0.5230134133503362\n",
            "train_accuracy: 0.4575645756457565\n",
            "validation_accuracy: 0.43830570902394106\n",
            "training_rmse 0.7038022463843644\n",
            "validation_rmse 0.7231966622048173\n",
            "\n",
            "\n",
            "category: 5 epoch: 10\n",
            "training_loss: 0.4815071526259587\n",
            "validation_loss: 0.5213862205984184\n",
            "train_accuracy: 0.46924969249692494\n",
            "validation_accuracy: 0.43278084714548803\n",
            "training_rmse 0.6939071641948324\n",
            "validation_rmse 0.7220707852872221\n",
            "\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "0PkKLGcEUFne",
        "colab_type": "text"
      },
      "source": [
        "## Plotting Results:\n",
        " - Here we plot the loss, accuracy, and RMSE results from each of the six models."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "XoXDNZjsLLMz",
        "colab": {}
      },
      "source": [
        "index_to_cat = {\n",
        "    0 : \"ideas/content\",\n",
        "    1 : \"organization\",\n",
        "    2 : \"voice\",\n",
        "    3 : \"word choice\",\n",
        "    4 : \"sentence fluency\",\n",
        "    5 : \"conventions\",\n",
        "             }"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "O9MPQlXH8nVG",
        "outputId": "6bc9b710-08ab-4a7e-890e-416431d27a11",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "for index in range(6):\n",
        "    plt.plot(trn_loss_list[index], label=str(index_to_cat[index]))\n",
        "plt.title(\"Training Loss vs Epoch\")\n",
        "plt.xlabel(\"epoch\")\n",
        "plt.ylabel(\"loss\")\n",
        "plt.legend()\n",
        "plt.show()\n",
        "\n",
        "for index in range(6):\n",
        "    plt.plot(val_loss_list[index], label=str(index_to_cat[index]))\n",
        "plt.title(\"Validation Loss vs Epoch\")\n",
        "plt.xlabel(\"epoch\")\n",
        "plt.ylabel(\"loss\")\n",
        "plt.legend()\n",
        "plt.show()\n",
        "\n",
        "for index in range(6):\n",
        "    plt.plot(trn_acc_list[index], label=str(index_to_cat[index]))\n",
        "plt.title(\"Training Accuracy vs Epoch\")\n",
        "plt.xlabel(\"epoch\")\n",
        "plt.ylabel(\"accuracy\")\n",
        "plt.legend()\n",
        "plt.show()\n",
        "\n",
        "for index in range(6):\n",
        "    plt.plot(val_acc_list[index], label=str(index_to_cat[index]))\n",
        "plt.title(\"Validation Accuracy vs Epoch\")\n",
        "plt.xlabel(\"epoch\")\n",
        "plt.ylabel(\"accuracy\")\n",
        "plt.legend()\n",
        "plt.show()\n",
        "\n",
        "for index in range(6):\n",
        "    plt.plot(trn_rmse_list[index], label=str(index_to_cat[index]))\n",
        "plt.title(\"Training RMSE vs Epoch\")\n",
        "plt.xlabel(\"epoch\")\n",
        "plt.ylabel(\"RMSE\")\n",
        "plt.legend()\n",
        "plt.show()\n",
        "\n",
        "for index in range(6):\n",
        "    plt.plot(val_rmse_list[index], label=str(index_to_cat[index]))\n",
        "plt.title(\"Validation RMSE vs Epoch\")\n",
        "plt.xlabel(\"epoch\")\n",
        "plt.ylabel(\"rmse\")\n",
        "plt.legend()\n",
        "plt.show()\n"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEWCAYAAABrDZDcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjEsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy8QZhcZAAAgAElEQVR4nOy9eXxdZZ34//6cc5fs+560TZe0TQptgVpWQUAURkWKKDjMKDiAzgiozBe/rgi4DAM/URkdGVTAL8sogrLIjmwCXWihhbY3bdp0S5O0aZKb/W7nPL8/zrk3N2nSpiU3Sdvn/Xqd13m285zPOcn9fJ7zfJ5FlFJoNBqN5tjFmGwBNBqNRjO5aEOg0Wg0xzjaEGg0Gs0xjjYEGo1Gc4yjDYFGo9Ec42hDoNFoNMc42hBoxhURMUWkV0Smj2dZzdGHiFwlIq9OthwabQiOeVxFHD9sERlIil9+qPUppSylVJZSaud4lj1cXGWjROTiVN3jaEBEPur+/XuHHR+abNk0qccz2QJoJhelVFY8LCLbgauUUi+NVl5EPEqp2ETINk58Eehwz3+eyBuLiKmUsibynh+QnUqp6skWQjPx6C8CzQERkR+JyB9F5H9FpAf4JxE5VURWiEhQRFpE5C4R8brlPW4LvNqNP+jmPysiPSKyXERmHmpZN/8CEdksIl0i8l8i8qaIXHEA2WcBpwNfBi4QkeJh+ReLyFoR6RaRLSLyMTe9UETud5+tU0Qec9OHdGWMIv+vROQ5EekDPiwiF7r36BGRnSLy/WEynOm+yy4R2SUi/+y+32YRMZLKXSoiq0d4xjNEZPewsp8VkXfc8Cki8o77jHtE5I7R3teBEJE3ROTHIrLalfUvIpKflH+RiGxw/ydeFpF5SXkzRORxEWkTkX0i8ouhVcvP3Osa438DzcSiDYFmLCwDHgZygT8CMeBrQBGOoj0fR9mOxj8C3wcKgJ3ADw+1rIiUAI8AN7r33QYsPYjcXwRWKKUeBbYCn49niMhpwL3AvwN5wNnADjf7YcAH1AGlQLLiOhj/CNwCZAPLgV7gn3De3aeAr4nIJ10ZZgJPA3cChcAJwPtKqeVAD3BuUr3/BDwwwv3eBKLAWcNkeNgN/xdwh1IqB5gDPHoIzzKcL7hHBSDAz9znqAUeBK4DioGXgKdExCsiHvcZtwDVwDScv2Oc04D3cZ7/Z8DvPoB8msNFKaUPfaCUAtgOfHRY2o+Alw9y3f8B/uSGPYACqt34g8DdSWUvBNYfRtkvAX9PyhOgBbhiFJkEaASudePfB9Yk5f8OR0EOv24ajqHLHSHvKuDVpPhI8t97kHf1y/h9XZn+NEq57wK/d8NFQD9QMkrZ24B73HCeW7bKjb8F3AQUHkSujwI2EBx2+N38N4AfJZVfCITc93wL8HBSngG0AmcAH3bD5ijvsz4pnuO+z6LJ/i0ca4f+ItCMhV3JERGZLyJPi0iriHQDt+Ioq9FoTQr3A1mjFTxA2YpkOZSjOZoOUM+ZDG19PgycKCLHufFpOF8Jw5kG7FNKdR2g7gMx/F2dKiKvut0iXTjKL/6uRpMBnNb/RSKSAVwGvKKU2jtK2YeBz7jdc58BViql4u/mSpwvm00iskpE/uEAsu9USuUNO8KjPNsOwI/z5VbB4NcUSikb529T6T7jdjW6r2T43xsO/P+hSQHaEGjGwvAlav8HWA/MUU6Xw004LcNU0gJUxSMiIjiKZjS+iPP//Z6ItOJ0oSicrg1wlNrsEa7bBRSJSM4IeX1ARlK8bIQyw9/VH4DHgGlKqVzgtwy+q9FkQDkjqVYDnwb+mZG7heJl38N5Px9naLcQSqlNSqnLgBLgp8BjIpI2Wl0HYVpSeDoQxnHENwMz4hmuv6IK2I3zjDNExDzMe2omAG0INIdDNtAF9Ln9wwfyD4wXf8Vp0X/K7Xf+Gk5/9H64rehLgH8BFicd38Bxdps4XUNXicjZImKISJWIzFNK7cLp4/6ViOS5/dxnulWvAxaKyPEikg78YAxyZwMdSqmQiJyC07qP8yBwvoh8xnU8F4nIoqT8/wd8G5gPPHGQ+/yv+3ynkuQHcJ3PRW4rvQvHUNljkHskvuB+DWbidAc94n6ZPQJcKCIfcb9KbsTxcazE8ZO0Az8RkQwRSReR0w/z/poUoQ2B5nD4d5wWdw/O18EfU31DpdQe4FIcx2o7Tkv6XZxW6XAudmV7UCnVGj+A3wDpwHlKqbeAq4G7cBTkKwy2eP/JPW8G9uA4QVFKbQR+ArwKbAJeH4Po/wr8hzgjrr5DkqNUKbUNx4H8f3Fa1u8Axydd+xgwC3hUKTVwkPs8DJwDvKiU6kxK/wcg4N7//wMuVUpFRqljuuw/j+CipPwHcIxXC2ACX3efYwPO/8OvgTacwQMXKqWiyhlq/EmgFufrYCeOkdZMIcQx6BrNkYXbqm8GLlFK/X2y5UkFbvfXNhyH+KuTLMsbwG+VUvdPphya1KC/CDRHDCJyvojkiogfZ8RNDFg1yWKlks/hfPG8NtmCaI5u9MxizZHEGcBDOGP8NwAXDRvVctTgtsBrgMuV/mzXpBjdNaTRaDTHOLprSKPRaI5xjriuoaKiIlVdXT3ZYmg0Gs0RxZo1a/YppUYccn3EGYLq6mpWr95v7S2NRqPRHAAR2TFanu4a0mg0mmMcbQg0Go3mGEcbAo1GoznG0YZAo9FojnG0IdBoNJpjHG0INBqN5hhHGwKNRqM5xjni5hFoNIeKikax+/qwevuw+3qxe53D6u3F7u1z4v39YBoYPh/i9YLXi7iH4fPtF08uc8C4pHq/Hs3BULYNsRgqflgWKhp10iwLFY2hYlGwLCc/GgMrqXws5uziYIjz9zQMEAMExDBAZMS4GOKEDQOQQ7veMJyyifJO2MzOxkhPH/d3pA2BZkqilEJFIvsr7b4RlHhvL3Zfr6Poe5MUfZ8TV6HQ5D1I3Dh4veA7sHHB60U8XleBxBWCqzwOEEcEEcNVGK4yYVCBiCH7x/erc3h94ihQBdg2oEaO2wqUAmU7+w6PJW7boBRKJdU3WtyyHUVsxVyF7SrvWBSi8XBshHgsofwdeY8Oym7+AfmXXXbwgoeINgTHIEopel95lUjj1gP/WA/nx38g5TFieYUaGMDqG6rYrb4+iEYP/jAeD2ZWFkbiyMRTVIRRXZ2Im1lZGJmZGJlZQ9LIzICMdFRmOirNj1I2VjiMikawIxFUNIoVCWFHIhCNYUfD2GEnXUWj2NEIKjIYVxEnDTesYu45GsWKRh2lFI2h3PrirVIi/RDqhqijvEQBSiG2syBk/IytADUkLqhEuvN3SwqrpMO2nT00XSU7JB5POxDDWqeDLVsSrdnE2RCU2wKGoXElAIIy3D09xUlTibP7PypOvjIE5TFRpgGmifK4Z58BHhNl+tyzCR4TTCcdjwcMJ015TCSe7/GA6ZwlKS4eD3idsOHxDKZ7PYjpGfyys4e+08TfQQ1732qEv4Ha/+8jI6QdqI6BBTPJP/iv4pDRhuAYo2/FCvb+9E5C778/tgsMY+iPP6nVqOJx94c8eB78USuGnu2kdFsUCoj6DSJpHiJpJuEck0haFuG0HMJ+k1CaQchvEEoTQn5hwC/0+yDkF/p9ioipsLCxlY1lh7BVP5ZqxVKWk6YsbNs9D9hY/Rb2Hic+7hg427n7x7/qD4opJoYYmOJxzyaGYSSlm5gYmBh4xEQQIipKVMWwsInZMSzbIqZiRO0olm2hiE32Y40flntMFdyPuOFe3O/7ZvI5Th7322lDcIwwsGEDe++8k/433yJWnMfGq89k46JcwipG2A4TtqNEVISQHSFiRQjZYSJ2lIgVIWyFE2d12NvdDuI3/fhMn3M2fJiG6SiiEZTT8LNpOOFMMckeXma4YouHjVHqMwbLxNMBDDEQHMMXDxtiDMkzxOnDHUtevK4D5olguL96GxvLtoYas2FGbURDp4bl2ftfP1K9yeGYHcNWNgqFx/DgEY9zjh/iwTTMIfF42BQn3Wt4nTJunmmYeA3vmK4dniciKKWwcb4mlVIoVELG5DiQSLfdr8/9yibXxWDcaZDb++e5104VZuTMSEm92hAcpUTtKI3BRras/zu+3z1G1crt9KYLj51r8MKJPeB7m6JgEX7Tn1DMPtNHpjeb/CQlHVfYftOP1/QOLX+Q/JHqiP+4NRrN1EEbgqOAUCxEQ2cDgY6Ac7QHaNu5mQv/HuKcdQrLhDfPK6fz4rM4ZdoiriyoZVbeLLyGd7JF12g0UwBtCI4w+qJ91HfUE2gPJBR/Y7Ax0eddamdx+TtZLHk9ihkTPMv+gRnX38ji0tJJllyj0UxVtCGYwgRDwSGt/EBHgB3dg0uKF6YVUldYx0eqPkJddg2z/raJ6H1/wO5qIucTn6D4a9fjmz59Ep9Ao9EcCWhDMAVQStE20DbYynfPLX0tiTIVmRXUFtbyqVmforawltqCWoozilGxGF1PPEHbf91OuLWVzDPOoOSGb5BWVzeJT6TRaI4ktCGYYJRS7O7dPUThB9oDtIfaARCEGTkzWFyymM8XfD6h9HP9ufvV0/3ii7T9/BdEtm4lbeFCKm67jcxTxn9omUajObrRhmCC+dk7P+O+9fcBztju2XmzOaPyjITCn1cwj0xv5gHr6Fu1iraf3snAunX4Zs6k8q5fkH3eeXo0jkajOSy0IZhgljcvZ0HhAr5/yveZkz8Hvzn22UehQIC9d/6Mvr//HU9pKeU/+iG5F13kzIrUaDSaw0RrkAkkYkXYEtzCF+u+yIKiBWO/btcu2n5xF91//StGbi4lN/4f8i+/HCMtLYXSajSaYwVtCCaQLcEtxOwYtYW1Yyof27ePfb++m85HHkFMk8JrrqHwqn/BzMlJsaQajeZYQhuCCSTQHgCgruDAI3qs3l467r2X9vt/jwqHybvkEor+7d/wlpZMhJgajeYYQxuCCSTQESDLm0VlduWI+XYkQufDD9N+9/9gBYNkX3A+JV/7Gr7q6okVVKPRHFNoQzCBBDoCzC+Yn1ikLI6yLLqefIq2/7qLWHMLmaedRvENN5B+3Nj9CBqNRnO4aEMwQVi2xeaOzVwy95JEmrMvwCu0/exnhBu2kHbccVT8+MdknnrqJEqq0WiONbQhmCC2d28nZIWoK3T8A/1r1rD3p3cy8M47+Kqrqfz5z8n++Mf0XACNRjPhaEMwQWxs3wjA/HABu77yr/S++iqe4mLKbrmFvIuXOXvcajQazSRgHLzI4SMi54vIJhHZIiLfGqXM50Rko4hsEJGHUynPZFLfUY/f9JN29x/pW7WK4htuYPYLz5N/6ee0EdBoNJNKyr4IRMQEfgWcBzQBb4vIk0qpjUllaoBvA6crpTpF5KgdHxnoCDAvfx7hDRvJ/shZFF1z9WSLpNFoNEBqvwiWAluUUo1KqQjwB+DTw8pcDfxKKdUJoJTam0J5Jg2lFPXt9Rzvn0V0927888c2oUyj0WgmglQagkpgV1K8yU1LZi4wV0TeFJEVInL+SBWJyDUislpEVre1taVI3NTR1NtET7SHhcFsANJqtSHQaDRTh1QagpGGvwzfBdoD1AAfAT4P/FZE8va7SKl7lFJLlFJLiouLx13QVBOfUTxjj7PBdlqdNgQajWbqkEpD0ARMS4pXAc0jlHlCKRVVSm0DNuEYhqOK+o56POIhZ0c7npISPIWFky2SRqPRJEilIXgbqBGRmSLiAy4DnhxW5nHgbAARKcLpKmpMoUyTwsaOjczOm000sEl3C2k0milHygyBUioGXAs8DwSAR5RSG0TkVhG50C32PNAuIhuBV4AblVLtqZJpMlBKEWgPcFz2XMKNjfhr50+2SBqNRjOElE4oU0o9AzwzLO2mpLACbnCPo5K2gTY6Qh0s6skHyyKtVu8lrNEckygFsRCEeyESP/qcczgpPFI80ueknXYd1H5y3EXTM4tTTNxRPLvN+fjSjmLNlCSupKIDztmKguEB0+ucE2EvGCYcbUuhKAW2BXYM7KhzttywFR2TArfDvVgDA1ihMFY4RCwcxgpHnSMaJRaJYSkPlvISU14s5cXC58Z9WHiwlBvHj2VkYkkRlqQ5cfwsKvcyMwUqRBuCFLOxYyOCkL8zSH9WFt6qqskWSXOkYFuOYo4OQGwAoiGI9g8q7ET6wKACHxLud64Za5lDwfC4RsED5mjh4UbEg4WPiMokrDKJWGmEVQYRK52w5Xfilp9IzEc45iMS82JZBqAcRa3spLAC7KTwSGWcUXooO6mMPXr5A+Ao6UHl7SjyLCxVgIWTbo+DOjVMwfQaeDwGptc5PF4D042riukf+B4joQ1Biqlvr2dGzgys+gbS5s/Xi8pNBrEI9O5xjp5W6G2Fnj3OuXevk9a3D5Q14aLZSojYGYSsTEJWBhHbj8RCmFY/hgojYmEQw8DCEGvY2U1PShNRICZ4M8CbBp508KYPhn1ZkFkMnrSkMmn7lzc8buvYPawoyooRjdhEQopwGCIRCIeESMQgHBEiEZNwyCQc9RBxj3DMO6jYLT8x+2DLqdj4ZACf0Y9fejElCojzBSIyGAbAcNOTwsj+5Y1R0hP1CIjh1iMj1mOajkL2ez14fCamz4fp82L6fZh+Hx6v6ShuT5Lyjsc9w+LDlHsi32MgxuToB20IUkygI8AJRYsIbf4beZ+95OAXaMZOuGdQofe4Sj1ZycfPA50jXCyOQswuhawyKKlzWrOHiVIQjXkIRX3OEXHO4Yh3v7RQxEfYTQtHvYw85ebwEAHDNBBTME3BMAUxnLNhGhiJsLhhYzCelKaUIjIQIzwQc879zvkgDWcMj+BP9+DP8OLL9eBPN8lK9+JPN/FluOcR4r5007nGb06aMjyW0YYghQRDQVr6Wrgy+2OogQHtKB4Ltg0DHUmt9xHOvXscJR/t2/960wdZpc5ROBtmnAbZZU48uwyyShzFn1k8quKPRSxCfVH3iBHqjRLud+O9UUL9blqiTJRwXwzbHl1L+tJM/Jle0jK9pBV5yHXDibQsL/4MD750DyiFZSmUpbAthW27Z8tOCielDSuj7GHXD7/OHnatpYhF4mWc60HwpZtk5afhqzDxx5V1stIeIe7xmuP0j6CZSLQhSCGBDsdRPLfNec3HlKPYth1nWqgLwt3OOdSdFA8OxvvakpT8XsdBh+u/w+M40rz5xDLKsdPLiWXXYRUXY/mKsPwFxLz5WN48LE8OMUnHiimsqI0Vs4hFbax9CqvFxopaxGI2VnQfVqzNiUdtrKhNLGoT7o8R6otiRe1RH8vjNZKUt4eC8kz8WV7SMgbT/BmOYk9LKHsPppnShX41mg+ENgQpJG4Iipt66fd68c+aNckSOSzf2s7NT25gZ0c/eRle8jJ85Gd4h4QL/EKxd4BCT5h8s59cGSCHfjJUH2ake5hST1byXRDqQg10E1YZhOwcQnYOA3a2e84hpOLnPEJqOlHJwCJt0Alnm1i2gWUdShdB0D2GIQzpjx2pb9ab5cX0uAo+w5NQ4kMVuoe0TC8en27xTjS2rYhYNlHLJhKziVrKCQ9Js4nEnPR4mpM/QlpSubB7bdQtG0mqL+qWjVg2HkNI95lk+EwyfB4n7DVJ95mJcCI9nubzkO5NjpukeUyMKdj1pQ1BCqlvr6ciswI2N+KvqZn0fQfae8P8+JkAT7+zjWuyl/OR0lbMSDeeUC++3m78Vh+Zdi9Zqo80iQ65Nqp8hOwcOlyl3m/n0a0K6VUFDKgqQiqXsMohamcRsdKJxnwoNXIr2DCF9GwvaVk+0rO8pPuHOdqSRkwkO9KGxL3Dyo6k6L1On7h20I8vtq0IxSwGIhahmE0o6oTDMYtQ1HbTnXAoaiUdNgNJ4UR6vK6o7VyXVG84ZmMdoMvtcDENwWsKXtPA7zHwmvFj/7RsrwfLVvSEYuztDtMfjTEQseiPWAxErYP6TYYTNw5p3rhhSTIcCaNikuYzyfB6hhiSE6bnM7Moc9zfhzYEKSTQEaC2YD6hwNtknXvOpMlh24pHVu/i9mc28LHYK6zKeYL0UJCe4FxCnjIGjLmEvIV0e/LYqXIYiGXSH02nP+IjFPYSDpvYo7TOFRA2YEAUvdgMuOF+n8WAxBLxkKHwZjjdJNmZXvIz/eRmeMnP8OE1DQx3gIYh4g4EsTFEOaNmBEQJEgWJCkZosCw4Z+d6cQeIuHEkUa+IIAzWn7hPUllLOf3nlq0Gw+455qbbSmHZuOfktKQj6VrLZkg9lnLqsodfq5xy9qFqlRQRsxXh/ZS3E49Yo3edHQhDHCWYljiMRDjD56Eg08DvdVrN6T6DNI+Jz2MMHuagcvZ5HKWdSHPzfR5JUurJCl0SZbymgTlOrXKlFOGYTX/Eoj8yaCAcIxFLhEPRpPRIzD27aVHH+O3tCQ1JH4hY+73rHy87ThuCI4m+aB/bu7ezLPdMrM4XJm2NoU2tPXz3z+9RtvNVfm0ECIdKeK7vJtpD5SO2ZHxpJmnZTks9p8RLqevITM/2JbpJ0rPctCwf/gxPYpSHZSu6BqJ09kcI9kcJuudEfCBCp5ve2h2ivrWHYH+EqKVQKGzlKNgpogvHjAh4DMEQwTQEU5zRN2YijUSax3DzJDk/njZooCYbQ4S8DB9pXmOY8h5U4OnDlHma1yTNY5Dui4fdfLdLxGsefV9nIpJ49oJM37jXH7Ns+t0vroGIRX7G+N8DtCFIGZs6NgFQ154GTOweBEopWpt6ePjpBnoCW7kgBiHrVFZzKh6Pomx2AUvm5JJfnkm6q/TjSt70HL5T0zSEgkzfuPwglGsQbKVQDBoIpUgYDaUGz8PLonDyGFpm/zrj14JpMKiY40rd2F+5O2ESyvxoU26aqYPHNMgxDXLSUtutrA1Biog7ist3h+gXIW3evJTdy7Zs9jX10twQpGVLFzs2dWANWKQBeeKlLGM7lQu9lJ9xFkXVeUfECBaJd+GM4xh7jUYzMtoQpIhAe4DCtELMrTvxzZiBkTl+/XrRsMWebV00b+miZUuQ1m3dxMLOrFjLb5HHBk7KeYOSrGaKzvtn5EP/Bh7/uN1fo9EcXWhDkCICHQFqC2sJbQyQvmjhB6proDdCi6v0m7d0sW9njzN5SaCwMot5J5eyM7SHzMa7uch8FsuTgXH69XhO/yr4s8fpiTQazdGKNgQpIGyF2Rrcyrl5pxDd/Sp5l1465muVUvS0hxJKv2VLkM7WfsCZvl9ancPij02nYk4eZbNyaGjaTuNjt3BZ6BnEY9B3wpfJ+eg3IVPvgqbRaMaGNgQpYEvnFixlcVxnBnBgR7GyFe3Nfa7id/r4+4JhAHzpHspn5zLvlDLK5+RRMiM7MYW/O9jO8t//Ox9q/QPzJUrzzM8wbdnN5OTq1U01Gs2hoQ1BCtjYsRGAqpYoYYYuLWFbNq3bumlxlX7L1i4iAzEAMvP8VMzJpXxOHuVz8iioyNxvFqKK9BN44qdUbLibj9DL+vxzqf7cj5lecQwtX6HRaMYVbQhSQH17Pdm+bPwbWrCGbVb/+h8b2PD6bgDyyzKYc1JJQvlnF6aNPhTRitL+xu+Q12+nzmpnjfckcj5xK8ctPmMiHkmj0RzFaEOQApwZxbWEA4H99ijeuaGdabX5nPelBaRnj2G8vW0Tfe9R+p67lcLQLt5V81j1oTs47x8uHrfZkRqN5thm6g8oP8KI2TE2d26mLmsO4cbGIf6Bno4QPe0hqhcWHdwIKAWbX6Dvv07D+/jVtPTDr8t/QsUNr3H+Jz+jjYBGoxk39BfBOLOtaxthK8zCnrz9NqtvbnBWx6yoyTtwJTuWE33hB3h3r2SfXcJP077Oh5d9hX+tLU+l6BqN5hhFG4JxJj6juLpVEWOoo7i5IYg/w0NBRdbIF7e+j/rbD5GG5wmSxy+tK8k57V+48dw60vXyxxqNJkVoQzDOBNoDpHvSyWzYS09WFt7KykRec0OQ8tm5+69H3r4VXvkJrH+UPsniV9HLWF95GTd9Zgk1pXpCmEajSS3aEIwzgY4Ac/PnEgnUO5vVG44bpr87QnBPP7WnJ3XvdLfAa/+JevcBYpj81vo0/+u5iGuXfYhvnlSlFzPTaDQTQkqdxSJyvohsEpEtIvKtEfKvEJE2EVnrHlelUp5UYyub+o56avPmEdq8Gf+wbiFw/QP9HfDiTXDXCdjvPMBf5KOc1n8n2xb9O4//n0/yuSXTtBHQaDQTRsq+CETEBH4FnAc0AW+LyJNKqY3Div5RKXVtquSYSJp6muiL9rEoVLLfZvXNW4J4fAbF/W/AL/4VFe5mZda53Nj+CfzFs/nl5cdx8iy9LIRGo5l4Utk1tBTYopRqBBCRPwCfBoYbgqOG+Izi2Xud1nxa0hyC5oYgZbNyMd76EV1GDl+0v0cgOI3rP1bD1R+ehe8D7AOg0Wg0H4RUap9KYFdSvMlNG85nROQ9EXlURKaNVJGIXCMiq0VkdVtbWypkHRcC7QE8hoe8HZ2I14t/9mwAQn1R2nf3UjkrA7X7HR7oOZG8mYt58Rtn8dWz52gjoNFoJpVUaqCROrmHb0L4FFCtlFoIvAT8fqSKlFL3KKWWKKWWFBcXj7OY40d9Rz01eTVEN20asll9y9YuUFCRvQsDi1Dladx3xYeYXpgxyRJrNBpNag1BE5Dcwq8CmpMLKKXalVJhN/ob4KQUypNSlFIE2uOb1dfv5yg2PEJ612tElEnpgjO1M1ij0UwZUmkI3gZqRGSmiPiAy4AnkwuISPJU2QuBQArlSSl7+vfQGe7keFWJ1dk5ZGmJ5oYgpdU52NtfZZ2azZIavVS0RqOZOqTMECilYsC1wPM4Cv4RpdQGEblVRC50i10vIhtEZB1wPXBFquRJNYF2x4bNbXP873FDEAnFaNvZQ8WsDPKCG1hrHMc8PUlMo9FMIVI6oUwp9QzwzLC0m5LC3wa+nUoZJopARwBDDIp29dCVtFn9nsZulK2oyG7CxKK3/JT9ZxZrNIBtW4T7+0nLzNJdh5oJRc8sHicCHQGqc6qx3946ZLP65i1BxBAye18nokyKaj88yZJqJptoOERH8246mpvo2L2Ljt1NdDQ30dmyGysaJS0zi8JpMyiaNoPCadOdc9V0MnJyJ1t0zVGKNgTjRKA9wJKyJR2YvZkAACAASURBVIQ2rh6yWX1zQ5DiaVnIjpdZq+bwobnaP3AsoJRioLuLjt1NtO/e5Sh9V/F3t+1NlBMxyC0ppaCyiupFJ5KRm0ewpZl9TTupf/M1wv19ibIZuXkUTZvuGImqGa6xmI4/I3MyHlFzFKENwTjQEepgT/8ejvPNJLr7icRm9bGoxZ5t3Rz/4WLyAhtZZyzjX0q0f+BowrYtuvfuHVT2u91WfnMTod6eRDmPz09BRRUVc2s57uzzKKiYRkFlFfllFXh8I+9NoZSit7Od9p072Ne0k327dtDetJP1L79INBxKlMsqLKKoKm4gBr8gvGlpKX9+zdGBNgTjQH17PQC17c4PL+4o3ru9BytmU5HTjIFNb/mp2j9whHKw7pw4Gbl5FFRUMfeU0ymomEZhZRUFldPILixKLEA4VkSE7IIisguKqF48OLJa2Tbd+9oShqF91w727dpJ08aniUUjiXK5JaUUxg2D29VUUFE1quHRTE2UUgz0dNOzr42sgkIy8/LH/R7aEIwD8aUlyltCdDG4B0FzQycIZPW/Rlh5KKo9fRKl1ByMQ+rOKS2loMLpzimorEq08NOzUv/FJ4bTnZRbUsrsk5Ym0m3bomtPK/uadia+Itp37WD7unexrVhC9ryycsdATJ+RMBT55RWYHm/KZdfsTzQUoru9jZ59bfS076N7Xxs9SfGe9n3EIs50q49e9W8sOu8fxl0GbQjGgfqOeiqzKmFVI56kzeqbG4IUVmRh7nT8A0vnjriChmYCiIZD9HcF6Qt20tcVpD/ohPu7gvR3BentbKezpXlod47fT0H50O6cwsoq8g7QnTOZGIZJfnkl+eWV1Hzo1ES6FYsRbG1m365495LzBbF19UqUsp1rTefauN+haHo1pTPnOF8yegTTYWPFYvR1dhxQ0Sf/zwEgQlZ+AdmFRRRXz2LWSUvJKSomu7CI0lk1KZFTG4JxINAeoK6wjtDGwc3qLcumpbGb2qUF5DUEWGdezFUlo+xMpjksBpV7kL6uTvqDwYSy7+8KOgq/q5O+YJBoaGDEOtKyc8jMzSMzL4+5p5xOYeU0CioOvztnKmJ6PBRWTaewajrzTj0jkR6LROhobqI9yf+wp7GBzSvecPbMBtJzcimdNYeyWXMomTVHG4ckkrtsRlP0fZ2dCWMbJy0zi+zCIrKLiqmYW0t2UXFC0ecUlZCZX4DpmVjVrA3BB6Qn0sPOnp0sm/4Jwo0vkHXuOQDs29lLLGxRkduCgU2f9g+MiWTlPppS7+9y0iIDB1buGbl5lM2eS4YbdhR+vhPPyyMjJ2/Cf3BTCY/PR0n1LEqqZw1Jj4ZC7Nu1gz2NW2htbGBv4xZWvvcuynYUWnpOLmWz5lB6lBsH27LoaW+js7WF7ra9rnLfR0/7XqfLZt++IT4ZANPrTSj1GcefQHZREdmFxeQUFpFdVEJ2URG+tPRJeqLROXZ/BePEpo5NACzoyh6yWX18I5rs/r8TVh6Ka88YtY6JRilFZGCAWCSMbVvYMQvbimFbtnu2sC0Ly4qhLAvLGpYfi2HbtnN2y8av269scn7Mcu8Xw7KsRN2h3u6Ewh9VuWdlJ5R46ayaRDjTVeqZuflauY8T3rQ0ymvmUV4zL5EWjYRp276NPdu2sKfRObYnGYeM3DxKZ85OGIeyWTVkFRROeeMQi0bpbttDsLWFYGszwT2t7rmFrr17sC0rUVbEIDM/n+yiYoqrZzN7ySmJln1OYTHZRcWkZ+dM+WceCf2L+YDEN6uvao3SS5KjeEuQvNIMvLtfcuYP1Iy0AvfhYcWihPv7Cff3Ee7rc87Dw4l4P+H+Xufc10ek30kb/rmaCgzTg2GazuHxYBgGhseDGU9z89Oysh3lnptHRl7+CMo9VzsyJxmvz0/F3PlUzB3cY2OIcdi6hT3bRjEOs2sonel8QUyGcYiGQ46C39MyqPBbWwjuaaVnX9uQ34IvPZ280gqKZ8yiZulp5JVVkFdaRm5J2aR02UwUR+dTTSCB9gDF6cV4Arsw3M3qla1o2RJk9sJc8rbXs9b8DNeUDvoHIqEBR0n39Q4q9FEVeVJ4wFHm8REEoyKCPz0DX0YGaRmZ+DIyHcfTtBn4MjJJy3TSvD7/iIraUdAGhukobTFNzGSlbnowPCaGYWJ4nDynzKByF8M4IltGmrEzonEIh2jbsc39atjqGId17yaUrfNF5xiF8TQO4f4+V7k7yr6ztZkut3Xf29kxpGxadg75peVUzJ1P3pnnkF9WTl5ZOXml5aTn5B6T/7faEHxAAh0BagtrCW8MJDar39fUS7g/RrnrH+ivODXxz/Xag/ey+qk/H7BO0+PBn5mFPyMTf2Ym/oxMsguL8WdkOOnpGYl0f2Ym/vTMIXFfWvpR4eTUHHl4/WlUzK2lYu7g6rtDjYPbrbT2nZGNw6waSmfNJit/qHGIO2YHlX1SN05rCwM93UPkyMwvIK+0jBkLT3SUvKvo80rLScvSgzaGow3BByAUC7GtaxvnVp1NaPP95H32EsCdPwDkht8irLwJ/4BtWax/9SUq59dRd+Y5+DOyXOXuKnH3mIpDEzWaw+VAxqF16xb2bhvdOHjT0hPKPjLQP1ipCNmFReSXlVOz9DRyS8vIL6sgr6yc3NKyKemQncpoQ/ABaOhswFIWx/cXDtmsvrkhSHZBGv7mF3lXzWFpTQUATYH1hHq6OekTF1Gz9LTJFF2jmVRGNA6hEHvdL4e927bQurUBKxYlr6zC6cYprUi07nOLS3WDaRzRhuADEHcUz9ijGMDZrF4pRXNDkOnzsslrCrDW/CxfducPbF7xJh6/n+pFJ06i1BrN1MSblkblvFoq59UevLBmXNEdyR+Aje0byfHlkN7YktisPrinn4GeKOV5rRgoBlz/gG1bNKx6i1knfAivXy8GptFopg7aEHwA6jvqHUdxfX1is/r4/IHc8HLCykuJu75Q86YA/V1B5p6i1xvSaDRTizEZAhF5TEQ+ISLacLhE7SibOzdTl19LKBBIbFbfvCVIeo6PjNbnWWPXJPwDm1e+icfrY+YJSyZTbI1Go9mPsSr2XwP/CDSIyG0iMv9gFxztNAYbidpRFqhyZ7P6+fEVR4NUzMwgr6ee9zzHM6ckC2XbNKx8i+rFJ+rRDBqNZsoxJkOglHpJKXU5cCKwHXhRRN4SkStF5Jic8hl3FM/e67zCtLpautsH6O0IU5G/1/EPVDr+gZYtm+ntaGfuybpbSKPRTD3G3NUjIoXAFcBVwLvAL3AMw4spkWyKE2gPkO5JJ2f7PnA3q2+J+weiKwkpLyV1zvyBzSvfxDA9zEpaO16j0WimCmMaPioifwbmAw8An1JKtbhZfxSR1akSbipT31HP/IL5RFZsSmxWv7thJ/4MD9l7n2WNPZeT55ShlKJh5ZtULzpB7y2r0WimJGP9IvilUqpOKfUfSUYAAKXUMef9tJXtLC1RUDtkD4LmhiDlcf+A9zhmF2exd9tWutv2UqO7hTQazRRlrIagVkTy4hERyReRf0uRTFOeHd07GIgNUOebQXT3btJq6+jrCtO1d4CKgjZ3/sBpiAibV7yBYZrMXnLyZIut0Wg0IzJWQ3C1UioYjyilOoGrD3aRiJwvIptEZIuIfOsA5S4RESUiR8TXRX2Hs1n9vH3OFPe02trB+QPRtxlQPkrrTkcpxeaVbzJtwcIJ2ctWo9FoDoexGgJDkpYCFBETOOBCH26ZXwEXAHXA50WkboRy2cD1wMqxCj3ZBNoDeA0vhU3OiodptfNpaQji8Zvk7nuaNXYNJ88po23HNoKtLXoSmUajmdKM1RA8DzwiIueKyDnA/wLPHeSapcAWpVSjUioC/AH49AjlfgjcDoTGKMukE+gIUJNfQ7R+E57iYjxFRTRvCVJenUFhbz3vexcyuziLhlVvIWIwJ2kjcY1Go5lqjNUQ/F/gZeBfga8CfwO+eZBrKoFdSfEmNy2BiJwATFNK/XWMckw6SqmhjuK6WkJ9Udp391FesA8gMX9g84o3qao7joyc3EmWWqPRaEZnTMNHlbNI+K/dY6yMtM2PSmQ6y1X8DGduwoErErkGuAZg+vTphyDC+NPS10JXuIu6zDmEGx8l69xzaNni+AfyYqsZUD7Kak+nvWknHbt3ccLHPzmp8mo0Gs3BGOtaQzUi8qiIbBSRxvhxkMuagGlJ8SqgOSmeDRwHvCoi24FTgCdHchgrpe5RSi1RSi0pLi4ei8gpIz6juLY7M7FZ/e6GIKbHoKDjr6y257J0ThmbV74JIsxZqruFNBrN1GasXUP34XwNxICzgf+HM7nsQLwN1IjITBHxAZcBT8YzlVJdSqkipVS1UqoaWAFcqJSa0hPUAu0BTDEp3T0AOEtLtDQEKZ2RTmFfPe97j2d2cSYNK96kcl4tWfkFkyyxRqPRHJixGoJ0pdTfAFFK7VBK3Qycc6ALlFIx4FocR3MAeEQptUFEbhWRCz+I0JNJoCPAzNyZWJsaMLKyUIWltO3soaLQ2SA7VHU6wdZm2nZu12sLaTSaI4Kx7lAWcvv0G0TkWmA3UHKwi5RSzwDPDEu7aZSyHxmjLJNKfXs9J5efnNisfs+2HpSC3Nga+pWf8tpT2bzyLQDm6O0oNRrNEcBYvwi+DmTgjPc/Cfgn4IupEmqqsm9gH3sH9jI/dy6hzZvx1zkTyQxDKAoO+gcaVr5J+Zx55BRNrj9Do9FoxsJBDYE7MexzSqlepVSTUupKpdRnlFIrJkC+KUV8RnHdQH5is/rmLUGKq9Ip6g/wvnchhXYPexq3UKMnkWk0miOEgxoCpZQFnJQ8s/hYJdDujBia1hIFwFMzjz3buykv6gQgVHUaW1YtB2DuybpbSKPRHBmM1UfwLvCEiPwJ6IsnKqX+nBKppiiBjgDTsqdBwzbE6yVoFGHHWsiz36Vf+amoPZXNT/83JTNnk1tSNtniajQazZgYq4+gAGjHGSn0Kfc45mZKBdqdGcXhgLNZfcu2XhAo7nqa1fZcFhZ6aGnYpEcLaTSaI4qxziy+MtWCTHW6wl009TbxmZqLCQXuI+vcc2huCFJUnkbJwHoe8V7OqVvWAui9BzQazRHFWHcou4+k5SHiKKW+NO4STVE2dWwCoM4uw+rsxDu3lta3u6itHYC9EKk6jYaVb1E0vZqCisqD1KbRaDRTh7F2Df0VeNo9/gbkAL2pEmoqEl9aorrVBqCvuIZY1CZfraNP+SmZvoDdmzbqbiGNRnPEMdauoceS4yLyv8BLKZFoihLoCFCSUYK3cTeI0B7LB3op7nmGt+35lHdvZ49Seu8BjUZzxDHWL4Lh1ACTuwzoBFPfXk9dQR3hQADfjBm07Ownv8RPWWgd633H077+bQoqqiisOqZei0ajOQoY6+qjPSLSHT+Ap3D2KDgm6I/2s617G7WFzh4EvvnzadnSRXmxs0NZuOREmjau118DGo3miGSsXUPH9Ia7mzs3YyubWu90ort3oz71z0TqY+TzHr0qjTzDT1DZerSQRqM5IhnrF8EyEclNiueJyEWpE2tqEV9aYs4+x24GM6sBKOl9lrftefh2B8grLad4xszJElGj0WgOm7H6CH6glOqKR5RSQeAHqRFp6hHoCJDvzyejsRWAtoFMcgq8VITfZb2xgH2b11NzyunoVTg0Gs2RyFgNwUjlxro8xRFPoD3A/IL5hOsDmMXFtOwcoLzEGT0b8VRiW5YeNqrRaI5YxqrMV4vIncCvcCaWXQesSZlUU4ioFaUh2MAX6r5AaOMrROuWEuqNki/r6VHppPX14CsuoXTWnMkWVaM5KNFolKamJkKh0GSLokkRaWlpVFVV4fV6x3zNWA3BdcD3gT+68ReA7x2aeEcmW4JbiNkxarPmEG78Hd0nXgJ7oLTvOVZE5xHZtYkFH/+k7hbSHBE0NTWRnZ1NdXW1/p89ClFK0d7eTlNTEzNnjt1nOdZRQ33Atw5XuCOZuKN4bmcaYcuiw1tBRo6HyvBqHg1fjB1r08NGNUcMoVBIG4GjGBGhsLCQtra2Q7purKOGXhSRvKR4vog8f4gyHpFsbN9IpjeTvB0dKKCty0tFSR8iYEXTyMovoHzOvMkWU6MZM9oIHN0czt93rM7iInekEABKqU7GsGfx0UCgI8C8/HmE6+uJFEyjr8ci39hAu5WJ7Gul5uTTEeNwJ2hrNBrN5DNWDWaLSGLtBBGpZoTVSI82LNtic+dm6grrCAfq6Z3rdAGV9r/AKz11KCuqRwtpNIfIaaeNvHvfFVdcwaOPPjru9zvppJOIRCLjUlcwGOS///u/P1Ad999/P83NzeMiz3gxVkPwXeANEXlARB4AXgO+nTqxpgY7uncwEBtwNqvftImuwnn4M0ymR1exq6+QjNw8KubXTraYGs0RxVtvvTVh99q+fTuVlZX4fL5xqe9oNQRjdRY/JyJLgGuAtcATwEAqBZsKxJeent+fiz0wwD67kIqyAWJ9gt0XpuacMzEMc5Kl1GgOj1ue2sDG5u5xrbOuIocffGrBActkZWXR29uLUorrrruOl19+mZkzZ6LUYCfDmjVruOGGG+jt7aWoqIj777+f8vJyfvOb33DPPfcQiUSYM2cODzzwABkZGfzpT3/illtuwTRNcnNzef311wF49tlnOf/88wF47rnn+M53voNlWRQVFfG3v/2Njo4OvvSlL9HY2EhGRgb33HMPCxcu5Oabb2bnzp00Njayc+dOvv71r3P99dfzrW99i61bt7J48WLOO+887rjjDu644w4eeeQRwuEwy5Yt45ZbbmH79u1ccMEFnHHGGbz11ltUVlbyxBNP8PTTT7N69Wouv/xy0tPTWb58Oenp6eP6NzgcxuosvgpnH4J/d48HgJtTJ9bUINAewG/6KWnqJezLpXfApMDcyMbeUrBiem0hjeYD8Je//IVNmzbx/vvv85vf/CbxpRCNRrnuuut49NFHWbNmDV/60pf47ne/C8DFF1/M22+/zbp166itreV3v/sdALfeeivPP/8869at48knn0zc47nnnuP888+nra2Nq6++mscee4x169bxpz/9CYAf/OAHnHDCCbz33nv85Cc/4Qtf+ELi2vr6ep5//nlWrVrFLbfcQjQa5bbbbmP27NmsXbuWO+64gxdeeIGGhgZWrVrF2rVrWbNmTcIINTQ08NWvfpUNGzaQl5fHY489xiWXXMKSJUt46KGHWLt27ZQwAjD2eQRfAz4ErFBKnS0i84FbUifW1CDQEaAmr4ZIYBPBQmdkUOnAi7zcPQNfZjbT6o6fZAk1msPnYC33VPP666/z+c9/HtM0qaio4JxzzgFg06ZNrF+/nvPOOw8Ay7IoLy8HYP369Xzve98jGAzS29vLxz/+cQBOP/10rrjiCj73uc9x8cUXAxCJRGhqamLWrFk89dRTnHnmmYmx9QUFBQC88cYbPPaYs93KOeecQ3t7O11dzmo6n/jEJ/D7/fj9fkpKStizZ89+z/DCCy/wwgsvcMIJJwDQ29tLQ0MD06dPZ+bMmSxevBhw/BTbt28f93c4XozVEISUUiERQUT8Sql6ETnomEkROR/4BWACv1VK3TYs/yvAVwELZ8eza5RSGw/tEVKDUopAR4Dzq88nHKinZ9qJeP0GlZGVBPtO57izTsEwdbeQRvNBGGmoo1KKBQsWsHz58v3yrrjiCh5//HEWLVrE/fffz6uvvgrA3XffzcqVK3n66adZvHgxa9euZe3atZxxxhmJOke712gy+f3+RJppmsRisRGv//a3v82Xv/zlIenbt2/f7/qBganbmz5WZ3GTO4/gceBFEXkCOKC3Q0RMnCUpLgDqgM+LSN2wYg8rpY5XSi0GbgfuPCTpU8ju3t30RHqoLZhPKBCgM3sm5SUhdvXngq2Ye8oZky2iRnNEc+aZZ/KHP/wBy7JoaWnhlVdeAWDevHm0tbUlDEE0GmXDhg0A9PT0UF5eTjQa5aGHHkrUtXXrVk4++WRuvfVWioqK2LVrF8899xwXXHABAKeeeiqvvfYa27ZtA6CjoyMhQ7yeV199laKiInJyckaVOTs7m56enkT84x//OPfeey+9vc7aY7t372bv3r0HfO7hdUwFxuosXuYGbxaRV4Bc4LmDXLYU2KKUagQQkT8AnwYSLX6lVLKnKpMpNCQ17iiutUoZ6I3QY2Ux2/seG3tKEX86049bOMkSajRHNsuWLePll1/m+OOPZ+7cuZx11lkA+Hw+Hn30Ua6//nq6urqIxWJ8/etfZ8GCBfzwhz/k5JNPZsaMGRx//PEJhXrjjTfS0NCAUopzzz2XRYsWcfXVV3PrrbcCUFxczD333MPFF1+MbduUlJTw4osvcvPNN3PllVeycOFCMjIy+P3vf39AmQsLCzn99NM57rjjuOCCC7jjjjsIBAKceuqpgOMIf/DBBzEP0FtwxRVX8JWvfGVKOYtlpE+jcalY5BLgfKXUVW78n4GTlVLXDiv3VeAGwAeco5RqGKGua3BGLDF9+vSTduzYkRKZk7nrnbu4d/29vFxxO+tv/h3vH/dlziv7Gc9ssqn60FlccsM3Uy6DRjPeBAIBamuP/iHPTU1NXH311Tz77LOTLcqkMNLfWUTWKKWWjFQ+lVNiR5rnvJ/VUUr9Sik1G2fryxEXslNK3aOUWqKUWlJcXDzOYo5MoCPArLxZWJu3EMyrwfQIZs8GLFtYfOZZEyKDRqM5PKqqqo5ZI3A4pNIQNAHTkuJVHNiv8Adgyux6Vt9RT21BLeFAgK7iOkpLwmztzQfTQ/XCEyZbPI1Goxk3UmkI3gZqRGSmiPiAy4AnkwuISE1S9BPAft1Ck0Fbfxv7BvZRV1hHT30j3f5SCjz1NPQUkVGzCM84zVLUaDSaqUDKdhlTSsVE5FrgeZzho/cqpTaIyK3AaqXUk8C1IvJRIAp0Al9MlTyHQmJGsXca7X1pgODpfIWQ5eW0Mz8yqbJpNBrNeJPS7SaVUs8AzwxLuykp/LVU3v9wCbQ7hqB6r2J57hxEFH3dzSBlnHT6KZMsnUaj0Ywvev3kEQh0BJiRMwM2byOYN4ei/AEae/KhdDq+tMkf6qXRaDTjiTYEIxB3FPdt3ER3djXpvEOf5aNwybmTLZpGozkMRlv6+mD85Cc/GZd6pjraEAyjK9zF7t7d1BbWsmdrF8owiXSsQURx5rl62KhGk2qUUti2Pa51Hu7S18MNwUQuoT2RpNRHcCSScBRnzaatfyNK2ezrHUBl5zOronCSpdNoxpFnvwWt749vnWXHwwW3HbTYnXfeyb333gvAVVddxUUXXcQFF1zA2WefzfLly3n88cd56aWX+M///E8qKiqoqanB7/fzy1/+kqeeeoof/ehHRCIRCgsLeeihhygtLR116WgYXPr6pptuSqxO2tbWxsc+9jHuu+8+LrroInbt2kUoFOJrX/sa11xzDd/61rcYGBhg8eLFLFiwgIceemjIEtrf/OY3efbZZxERvve973HppZfy6quvcvPNN1NUVMT69es56aSTePDBB6f89qD6i2AY9e3OZvVz2r0Ec2aT5WmkN+ZDzTxpkiXTaI4O1qxZw3333cfKlStZsWIFv/nNb+js7GTTpk184Qtf4N1338Xr9fLDH/6QFStW8OKLL1JfX5+4/owzzmDFihW8++67XHbZZdx+++2JvJGWjk7m1ltvZe3atbz22msUFhZy7bXOQgf33nsva9asYfXq1dx11120t7dz2223kZ6eztq1a4esawTw5z//mbVr17Ju3TpeeuklbrzxRlpaWgB49913+fnPf87GjRtpbGzkzTffTNWrHDf0F8EwNnZspDyzHKOhia6cmWTE/oqBTeVZn55s0TSa8WUMLfdU8MYbb7Bs2TIyMzMBZ4+Bv//978yYMYNTTnFG5a1atYqzzjorsVz0Zz/7WTZv3gw4y0dceumltLS0EIlEEktLw8hLR1dVVQ25v1KKyy+/nG984xucdJLTwLvrrrv4y1/+AsCuXbtoaGigsHD0HoA33ngjsYR2aWkpZ511Fm+//TY5OTksXbo0cc/Fixezffv2xCqoUxX9RTCMQHuA+QXzaX2/Ccvw0te7g4xMxam10w5+sUajOSijrW8WNwwHKgNw3XXXce211/L+++/zP//zP4RCoUTeWJaOvvnmm6mqquLKK68EnFVHX3rpJZYvX866des44YQThtR5KM8wVhmmGtoQJNEf7WdH9w5qC2tpbYqgrL0MxBR9+RVMK8iYbPE0mqOCM888k8cff5z+/n76+vr4y1/+woc//OEhZZYuXcprr71GZ2cnsVgssXkMQFdXF5WVlQAHXS10OH/961958cUXueuuu4bUl5+fT0ZGBvX19axYsSKR5/V69+teij/DH//4RyzLoq2tjddff52lS5cekixTCd01lMSmzk0oFLW582gNBzB5HwMbu+6jky2aRnPUcOKJJ3LFFVckFOdVV11Ffn7+kDKVlZV85zvf4eSTT6aiooK6ujpyc3MBp0X/2c9+lsrKSk455ZTEHgNj4ac//SnNzc2Je1944YV897vf5e6772bhwoXMmzcv0T0FcM0117Bw4UJOPPHEIX6CZcuWsXz5chYtWoSIcPvtt1NWVjbEl3EkkbJlqFPFkiVL1OrVq1NS98OBh/mPVf/Bcyf8lid/0UZ04PcU+/fg/eqDXLJkekruqdFMJEfSMtS9vb1kZWURi/3/7d19fE73/fjx11vuLoSgSdugCKubIIkgWeue0lFTzEZ/VM1Xb7aWtb62VDubtdV2anu0bKO6YqbTWPRGV1/6KKFoq0kkImnci5umiEQiCZIrl8/vjyuuBYmbNpcTOe/nXznnOudz3tcheZ/z+Zzz/pQzatQoJk+ezKhRo669o6pVZahvOVn5WTRzNKN01ynKpQSn6xwmsD4/bBdsdWhK2c7s2bOJioqiS5cuhIWFMXJkrSlOXOdo11AlWXlZdGrWcaOUpAAAGitJREFUiZzNp3A5jyAYTgR3pGVTHR9Q6mabN2+e1SHYht4RVChzlXGg4ID7jeKTBlO2h5YNCnG17W91aEop5VWaCCrsK9hHuSmnQ5MOnHTW48KFApo3Kiask75IppSq2zQRVLhYerrlidsou3AMgLyGIcTq+IBSqo7TRFBhd/5uGvk14vyuIlzOfYT4nudgYKSODyil6jxNBBWy8rLo0KwD2buPYly5dArK4ULr2v1auFJ1XU5ODmPGjLE6jDpPEwFQfqGcvaf30qlZJ07kHwHg9ibndHxAKYs1b96chIQEq8Oo8/TxUSC7MJvzrvP8oF442WXv09D4keHzAx0fUHXaH7/6I7vza/ZN2I7NOhIXE3fVbeLi4mjdujW//OUvAff7Ao0aNeL48eNXlHXOzs5m+PDhZGRk4HK5iIuLY/369YgIjz76KFOnTiUlJYXp06dTXFxMcHAwy5YtIzQ0tEa/V12ndwT8dw6Chvv8MK4TtJF89jiiaNFEp6VUqqaNGzeO+Ph4z/KqVasIDg6utqzzRYsXL+bQoUOkpqaSnp7O+PHjcTqdTJ06lYSEBFJSUpg8eTLPP//8zf5Ktzy9I8CdCBw+DnJT3JN0dG10kIzW0y2OSinvutaVu7d069aNkydPkpOTQ25uLk2bNiUtLa3Kss4RERGe/T799FOeeOIJfH3df7aaNWtGRkYGGRkZDB48GACXy6V3A9+BJgLcA8Xtm7Yn7/hu/EwgfrcJ7cKjrQ5LqTprzJgxJCQkcPz4ccaNG8eBAweuuY8x5oqZvowxdO7cmS+++MJbodqC7buGLpgL7M7fTXh5R8qdx7mt1Iev/DsS21bHB5TylnHjxvHuu++SkJDAmDFjrqus85AhQ1i0aJGnvn9+fj4dOnQgNzfXkwicTieZmZk3/fvc6myfCL4p+oZiZzEhe3wAaOf8hr31o2iu4wNKeU3nzp0pKiqiRYsWhIaGMmrUKCIiIoiMjGTgwIGess6VTZkyhVatWnm2+9e//oW/vz8JCQnExcURGRlJVFRUnZ1g3pu8WoZaRH4EvAH4AH83xrx62efTgSlAOZALTDbGHL5amzVdhnp99npmbJ7B45/3ouzMWcaXJrJi9F/53wkjauwYStUWt1IZavXd1Zoy1CLiA/wVGAqEAw+JSPhlm6UCPYwxEUACMJebbHf+bgJL/SktOEZgeVOcTXxpG97tZoehlFKW8WbXUAyw3xhz0BhTBrwLXDIDvDEm0RhztmLxS6AlN1lWXhY9T7cD4PaiUnYFhen8A0opW/FmImgBHK20fKxiXXX+B/i/qj4QkcdEJFlEknNzc2ssQGMMWflZtDziQOo15c4z2WQF9yA0SMcHlFL24c1EIFWsq3JAQkQmAD2A16r63Biz2BjTwxjTIyQkpMYCPHn2JCWFBfjkncbHrx13+B7AtOlz7R2VUqoO8eZ7BMeAuyottwRyLt9IRO4Dngf6GWNKvRjPFbLys2h1sgEYQ5NSBwQZ2nXS8QGllL14844gCbhbRMJExB8YB6ypvIGIdAPeBEYYY056MZYqZeVn0TanEVIviOCCk2Q3uVPHB5RStuO1RGCMKQeeAtYDWcAqY0ymiLwgIhefzXwNCAT+LSJpIrKmmua8Yk9OBnfm+1PP726aFu5n/51duTPIcTNDUErVgE2bNjF8+PDr3r5NmzacOnXqurdfs2YNr7766rU3vEV5tcSEMWYtsPaydb+r9PN93jz+tZzO3M9dxg9f/3YEFa+mKOJ/rAxHKXWdXC4XPj4+N+14I0aMYMSIuvtukW1rDZ0+f5qmh52Ib1OaOMvwa1RGuy46/4Cyj+Mvv0xpVs2WoQ7o1JE7n3uu2s/nzp2Lw+Fg2rRpPPPMM+zcuZONGzeyYcMGli5dyooVK1i5ciUvv/wyxhgeeOAB/vjHPwIQGBjI9OnTWb9+PX/6058oLi7m6aefJjg4mOjoqmuDVVe6GmDBggV89NFHOJ1O/v3vf9OxY0fy8/OZPHkyBw8epEGDBixevJiIiAiWLVtGcnIyf/nLXzhx4gRPPPEEBw8eBGDhwoXce++9rFixgvnz51NWVkZsbCx/+9vfbmqy+j5sW2JiV04azXPrU8+3PU3z9pAf1FjrCynlZX379mXLli0AJCcnU1xcjNPpZOvWrfTp04ecnBzi4uLYuHEjaWlpJCUl8cEHHwBQUlJCly5d2L59Oz169ODRRx/lo48+YsuWLRw/frzK41VVuvqi4OBgduzYwS9+8QvmzZsHwO9//3u6detGeno6L7/8MhMnTryizWnTptGvXz927tzJjh076Ny5M1lZWcTHx7Nt2zbS0tLw8fHhnXfeqenT5zW2vSPI3L4ZHyP4+LUnKP99jv2gLT/U8QFlI1e7cveW7t27k5KSQlFREQEBAURHR5OcnMyWLVuYP38+SUlJ9O/fn4uPiY8fP57PPvuMkSNH4uPjw09+8hMAdu/eTVhYGHfffTcAEyZMYPHixVccr6rS1ReNHj3aE9N7770HwNatW1m9ejUAAwcOJC8vj8LCwkva3LhxI8uXLwfAx8eHoKAg/vnPf5KSkkLPnj0BOHfuHLfffnvNnLSbwLaJIH/nHnz9/HD43ElQ4UH2dNbxAaW8zc/PjzZt2rB06VLuvfdeIiIiSExM5MCBA3Tq1Im9e/dWu6/D4bikq+XyktRVqap09UUBAQGA+4/5xYqmVdVeu97jPPLII7zyyivX3LY2smXXUNn5c/gdKcKnYVuCOIGv6xwhfYZYHZZSttC3b1/mzZtH37596dOnD4sWLSIqKgoRITY2ls2bN3Pq1ClcLhcrV66kX79+V7TRsWNHDh065JnHYOXKlVUeq6rS1deK7WKXzqZNmwgODqZx48aXbDNo0CAWLlwIuMcgzpw5w6BBg0hISODkyZOe4xw+fNX6mbWKLRNBVtI2fFwQYLrSrDCTskA/Yru0sjospWyhT58+fPvtt9xzzz3ccccdOBwO+vRxv9EfGhrKK6+8woABA4iMjCQ6OpoHH3zwijYcDgeLFy/mgQceoHfv3rRu3brKY1VVuvpqZs+eTXJyMhERETz77LP84x//uGKbN954g8TERLp27Ur37t3JzMwkPDycl156iSFDhhAREcHgwYOvmGqzNvNqGWpvqIky1MtfjeNYRiaNGz5Nt/2LOO9XwKB1n9ZQhErVXlqG2h5qTRnq2spZVsqpjD0UNQpCpB6NTmRzpl2U1WEppZRlbJcIsnfuwDjLcQSE09j3FH7lJQT27GV1WEopZRnbJYJ9X27D6Q9NnN0IPrsLgK79Yq6xl1JK1V22SgTlTicHUr7ieDPwdTm4rWA35+s7uDPsatMkKKVU3WarRHBkVxpl585S2uAOAAJPHORMyzCLo1JKKWvZKhHs/XIb4vCjoelM/YACJL8E3y46/4BSyt5s82axq7ycA8lfUtamMc3z2hNqUsAIrWL0iSGlbjVpaWnk5OQwbNgwS47/61//mrVr1zJs2DAaNmxIYGAgM2bMsCSWmmCbO4KjmemcLynm28ZCQHkDbj+bCcAd3SMtjkwpdaPS0tJYu3bttTf0kjfffJMdO3bw2mtVzq57y7HNHUH+N0dxBAaS72wIQJP8/ZQEOPBr2dLiyJSyxpZVezl1tLhG2wy+K5A+P2tf7eclJSX87Gc/49ixY7hcLmbNmsXYsWNJSUlh+vTpFBcXExwczLJlywgNDaV///7ExsaSmJhIQUEBb7/9NrGxsfzud7/j3LlzbN26lZkzZzJ8+HCmTp3Krl27KC8vZ/bs2Tz44IMsW7aMNWvWcPbsWQ4cOMCoUaOYO3cuAOvWreO5557D5XIRHBzMhg0bKCkpqbKdykaMGEFJSQmxsbHMnDnzks/69+/PvHnz6NGjB6dOnaJHjx5kZ2fjcrl49tln2bRpE6WlpTz55JM8/vjjbNq0idmzZxMcHExGRgbdu3dnxYoViAhJSUn86le/oqSkhICAADZs2MCwYcNYsGABUVHunoxevXqxcOFCIiIivte/m20SQfSwB6nf4wfsf+P/qOd/hvJTJZS3jULq2eamSCnLrVu3jubNm/Pxxx8DUFhYiNPpZOrUqXz44YeEhIQQHx/P888/z5IlSwAoLy/nq6++Yu3atfzhD3/g008/5YUXXvDMDwDw3HPPMXDgQJYsWUJBQQExMTHcd5973qu0tDRSU1MJCAigQ4cOTJ06FYfDwaOPPspnn31GWFiYpwbRnDlzqmynYcOGnu+wZs0aAgMDSUtLA9xlKa7l7bffJigoiKSkJEpLS+nVqxdDhrjrm6WmppKZmUnz5s3p1asX27ZtIyYmhrFjxxIfH0/Pnj05c+YM9evXZ8qUKSxbtozXX3+dvXv3Ulpa+r2TANgoEQDsKdxP6Jl2hPhnUl7gQ9P7OlsdklKWudqVu7d07dqVGTNmEBcXx/Dhw+nTpw8ZGRlkZGQwePBgwF3ILTQ01LNP5XLR2dnZVbb7ySefsGbNGs+8AufPn+fIkSOAu0hcUFAQAOHh4Rw+fJjTp0/Tt29fwsLcTw1eLE9dXTvftyzHJ598Qnp6OgkJCYA7Ae7btw9/f39iYmJoWdEzERUVRXZ2NkFBQYSGhnrKWl8sfPfTn/6UF198kddee40lS5YwadKk7xXXRbZKBHuzD9HQGUE7Saeey9A0sqvVISllK+3btyclJYW1a9cyc+ZMhgwZwqhRo+jcuTNffPFFlftUVS76csYYVq9eTYcOHS5Zv337ds/+lduorjx1de1cL19fXy5cuAC4k0jldhcsWMD9999/yfabNm26ofgaNGjA4MGD+fDDD1m1ahXft+7aRbbqFzlxoAiA24rdNc8d4Vp8S6mbKScnhwYNGjBhwgRmzJjBjh076NChA7m5uZ5E4HQ6yczMvGo7jRo1oqioyLN8//33s2DBAs98AqmpqVfd/5577mHz5s0cOnQI+G956htt53Jt2rQhJSUFwHP1f7HdhQsX4nQ6Adi7dy8lJSXVttOxY0dycnJISkoCoKioyJMEp0yZwrRp0+jZs+clE+18H7ZJBBfMBUxOfS74llCWX4Lx9SOgbVurw1LKVnbt2kVMTAxRUVHMmTOH3/72t/j7+5OQkEBcXByRkZFERUXx+eefX7WdAQMG8PXXXxMVFUV8fDyzZs3C6XQSERFBly5dmDVr1lX3DwkJYfHixYwePZrIyEjGjh0LcMPtXG7GjBmeOYxPnTrlWT9lyhTCw8OJjo6mS5cuPP7449Xe3QD4+/sTHx/P1KlTiYyMZPDgwZ47jO7du9O4cWN+/vOf31BsV2ObMtSHzxzmX7O+Itgvm6iklQQ1aU37D9/3QoRK1V5ahvrWl5OTQ//+/dm9ezf1qnnYRctQVyP9UBaNS2+jtW86fgUuAruEWx2SUkrdkOXLlxMbG8ucOXOqTQLfhW0Gi08eKAaaEe7MoKi0DEcnTQRKqVvLxIkTmThxYo23a5s7gt4tetOs4QnKCt0DNDpQrJRSbl5NBCLyIxHZIyL7ReTZKj7vKyI7RKRcRMZ4M5YOPW9nVNM4juSHgAgB7b/b42FKKVXXeC0RiIgP8FdgKBAOPCQil/fHHAEmAVefUbomnPwaR3khF4rr49+qFT6BDa+9j1JK2YA37whigP3GmIPGmDLgXeCSoh3GmGxjTDpwwYtxAOA6tAWAoKLzBGi3kFJKeXgzEbQAjlZaPlax7oaJyGMikiwiybm5ud8pmL0BXfhzyWganM7TgWKllMfrr7/O2bNnPcvDhg2joKDAwohuPm8mgivfj4bv9NKCMWaxMaaHMaZHSEjIdwom8Uwoifnuin0OfY5aKVXh8kSwdu1amjRpYmFEN583Hx89BtxVabklkOPF413VqG4tiPhCYBs4OnW0Kgylao3EZYs5efhgjbZ5e+u2DJj02DW3W758OfPmzUNEiIiI4KWXXmLy5Mnk5uYSEhLC0qVLadWqFZMmTaJx48YkJydz/Phx5s6dy5gxYxg7diyPPPKIZ2KaSZMm8eMf/5iRI0feULnnBQsWkJOTw4ABAwgODiYxMZE2bdqQnJxMcHAwf/7znz1VUKdMmcLTTz9NdnY2Q4cOpXfv3nz++ee0aNGCDz/8kPr16zN//nwWLVqEr68v4eHhvPvuuzV6fr3Fm3cEScDdIhImIv7AOGCNF493VaFB9Wlb8A2+ISH4BgdbFYZStpeZmcmcOXPYuHEjO3fu5I033uCpp55i4sSJpKenM378eKZNm+bZ/ttvv2Xr1q385z//4dln3Q8fjhs3jvj4eADKyso8tforl3tOSkrirbfe8tQTSk1N5fXXX+frr7/m4MGDbNu2jWnTptG8eXMSExNJTEy8JM6UlBSWLl3K9u3b+fLLL3nrrbc8tYf27dvHk08+SWZmJk2aNGH16tUAvPrqq6SmppKens6iRYu8fi5ritfuCIwx5SLyFLAe8AGWGGMyReQFINkYs0ZEegLvA02BH4vIH4wxXqsNfT5rtw4UK1Xheq7cvWHjxo2MGTOG4IoLsmbNmvHFF1/w3nvvAfDwww/zm9/8xrP9yJEjqVevHuHh4Zw4cQKAoUOHMm3aNEpLS1m3bh19+/alfv36N1zuuXfv3tXGuXXrVkaNGuWZi2D06NFs2bKFESNGEBYW5pkcpnJ57IiICMaPH8/IkSMZOXJkDZ417/Lqm8XGmLXA2svW/a7Sz0m4u4y87kJpKaUHDhA4cMDNOJxSqhrVlViurPLnlcs0X6yN5nA46N+/P+vXryc+Pp6HHnrI8/mNlHu+VpzVubytc+fOAfDxxx/z2WefsWbNGl588UUyMzPx9a39BRxs82Zx6d594HLh6Kh3BEpZadCgQaxatYq8vDzAXQL63nvv9fSnv/POO1e9Ur9o3LhxLF26lC1btnj+8N9ouWe4sqT1RX379uWDDz7g7NmzlJSU8P7779OnT59q27lw4QJHjx5lwIABzJ07l4KCAoqLa3YqUG+p/amqhpzP+hrQ0hJKWa1z5848//zz9OvXDx8fH7p168b8+fOZPHkyr732mmew+FqGDBnCxIkTGTFiBP7+/oB7QDc7O5vo6GiMMYSEhPDBBx9ctZ3HHnuMoUOHEhoaesk4QXR0NJMmTSImJsbTdrdu3aqdJc3lcjFhwgQKCwsxxvDMM8/cMk8f2aYMddGGDRS89z4tF8zXeYqVbWkZanu40TLUtrkjaDRoEI0GDbI6DKWUqnX00lgppWxOE4FSNnOrdQerG/Nd/n01EShlIw6Hg7y8PE0GdZQxhry8PBwOxw3tZ5sxAqUUtGzZkmPHjvFdizeq2s/hcHhenLtemgiUshE/Pz/CwsKsDkPVMto1pJRSNqeJQCmlbE4TgVJK2dwt92axiOQCh7/j7sHAqRoM51an5+NSej7+S8/FperC+WhtjKlyZq9bLhF8HyKSXN0r1nak5+NSej7+S8/Fper6+dCuIaWUsjlNBEopZXN2SwSLrQ6gltHzcSk9H/+l5+JSdfp82GqMQCml1JXsdkeglFLqMpoIlFLK5myTCETkRyKyR0T2i8izVsdjFRG5S0QSRSRLRDJF5FdWx1QbiIiPiKSKyH+sjsVqItJERBJEZHfF/5N7rI7JKiLyTMXvSYaIrBSRGyvreYuwRSIQER/gr8BQIBx4SETCrY3KMuXA/xpjOgE/BJ608bmo7FdAltVB1BJvAOuMMR2BSGx6XkSkBTAN6GGM6QL4AOOsjco7bJEIgBhgvzHmoDGmDHgXeNDimCxhjPnWGLOj4uci3L/kLayNyloi0hJ4APi71bFYTUQaA32BtwGMMWXGmAJro7KUL1BfRHyBBkCOxfF4hV0SQQvgaKXlY9j8jx+AiLQBugHbrY3Ecq8DvwEuWB1ILdAWyAWWVnSV/V1EGlodlBWMMd8A84AjwLdAoTHmE2uj8g67JAKpYp2tn5sVkUBgNfC0MeaM1fFYRUSGAyeNMSlWx1JL+ALRwEJjTDegBLDlmJqINMXdcxAGNAcaisgEa6PyDrskgmPAXZWWW1JHb/Guh4j44U4C7xhj3rM6Hov1AkaISDbuLsOBIrLC2pAsdQw4Zoy5eJeYgDsx2NF9wCFjTK4xxgm8B9xrcUxeYZdEkATcLSJhIuKPe8BnjcUxWUJEBHf/b5Yx5s9Wx2M1Y8xMY0xLY0wb3P8vNhpj6uRV3/UwxhwHjopIh4pVg4CvLQzJSkeAH4pIg4rfm0HU0YFzW0xVaYwpF5GngPW4R/6XGGMyLQ7LKr2Ah4FdIpJWse45Y8xaC2NStctU4J2Ki6aDwM8tjscSxpjtIpIA7MD9tF0qdbTUhJaYUEopm7NL15BSSqlqaCJQSimb00SglFI2p4lAKaVsThOBUkrZnCYCpW4iEemvFU5VbaOJQCmlbE4TgVJVEJEJIvKViKSJyJsV8xUUi8ifRGSHiGwQkZCKbaNE5EsRSReR9ytq1CAiPxCRT0VkZ8U+7SqaD6xU7/+dirdWlbKMJgKlLiMinYCxQC9jTBTgAsYDDYEdxphoYDPw+4pdlgNxxpgIYFel9e8AfzXGROKuUfNtxfpuwNO458Zoi/ttb6UsY4sSE0rdoEFAdyCp4mK9PnASd5nq+IptVgDviUgQ0MQYs7li/T+Af4tII6CFMeZ9AGPMeYCK9r4yxhyrWE4D2gBbvf+1lKqaJgKlriTAP4wxMy9ZKTLrsu2uVp/lat09pZV+dqG/h8pi2jWk1JU2AGNE5HYAEWkmIq1x/76Mqdjm/wFbjTGFwGkR6VOx/mFgc8UcD8dEZGRFGwEi0uCmfgulrpNeiSh1GWPM1yLyW+ATEakHOIEncU/S0llEUoBC3OMIAI8Aiyr+0Feu1vkw8KaIvFDRxk9v4tdQ6rpp9VGlrpOIFBtjAq2OQ6mapl1DSillc3pHoJRSNqd3BEopZXOaCJRSyuY0ESillM1pIlBKKZvTRKCUUjb3/wGFdlxdbozG1gAAAABJRU5ErkJggg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "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\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": [],
            "needs_background": "light"
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "i1u-6cfqUFnk",
        "colab_type": "text"
      },
      "source": [
        "## Grading an Example Essay\n",
        " - Here we grade a single essay using the trained models and present our output feedback."
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FzLHS0WPUFnl",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "selected_essay_index = 0\n",
        "\n",
        "data_example = Dataset(\"/content/drive/My Drive/Colab Notebooks/NLP/data/training_data.csv\")\n",
        "data_example.preprocess()\n",
        "data_example.text = [data_example.text[selected_essay_index]]\n",
        "data_example.essay_id = [data_example.essay_id[selected_essay_index]]\n",
        "data_example.scores = [data_example.scores[selected_essay_index]]\n",
        "data_example.create_vocab()\n",
        "data_example.text_num()\n",
        "data_example.create_chunks()\n",
        "train_text_example = data_example.new_data\n",
        "\n",
        "for i in range(len(train_text_example)):\n",
        "    while len(train_text_example[i][0])<=max_len:\n",
        "        train_text_example[i][0].append(0)"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "aYQwQ15LUFno",
        "colab_type": "code",
        "colab": {},
        "outputId": "1062d3cb-d57a-40de-904a-49f089658c0a"
      },
      "source": [
        "for cat in range(6):\n",
        "    print(\"Category: \", index_to_cat[cat])\n",
        "    for chunk in range(3):\n",
        "        hidden = model_list[cat].make_hidden()\n",
        "        prediction, hidden = model_list[cat](torch.from_numpy(np.array(train_text_example[chunk])).to(torch.int64).to(device), hidden)\n",
        "        print(\"Section:\", chunk, \" Predicted Score:\", prediction.item(), \" True Score:\", data_example.scores[0][cat])\n",
        "    print(\"\\n\")"
      ],
      "execution_count": 0,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Category:  ideas/content\n",
            "Section: 0  Predicted Score: 3.588099718093872  True Score: 4\n",
            "Section: 1  Predicted Score: 3.5635805130004883  True Score: 4\n",
            "Section: 2  Predicted Score: 3.639218330383301  True Score: 4\n",
            "\n",
            "\n",
            "Category:  organization\n",
            "Section: 0  Predicted Score: 3.6750357151031494  True Score: 4\n",
            "Section: 1  Predicted Score: 3.901684522628784  True Score: 4\n",
            "Section: 2  Predicted Score: 3.6916704177856445  True Score: 4\n",
            "\n",
            "\n",
            "Category:  voice\n",
            "Section: 0  Predicted Score: 4.0161051750183105  True Score: 4\n",
            "Section: 1  Predicted Score: 4.017720699310303  True Score: 4\n",
            "Section: 2  Predicted Score: 3.9705214500427246  True Score: 4\n",
            "\n",
            "\n",
            "Category:  word choice\n",
            "Section: 0  Predicted Score: 3.7575244903564453  True Score: 4\n",
            "Section: 1  Predicted Score: 3.8006322383880615  True Score: 4\n",
            "Section: 2  Predicted Score: 3.8926503658294678  True Score: 4\n",
            "\n",
            "\n",
            "Category:  sentence fluency\n",
            "Section: 0  Predicted Score: 3.7069079875946045  True Score: 4\n",
            "Section: 1  Predicted Score: 3.415090560913086  True Score: 4\n",
            "Section: 2  Predicted Score: 3.4983415603637695  True Score: 4\n",
            "\n",
            "\n",
            "Category:  conventions\n",
            "Section: 0  Predicted Score: 3.450746536254883  True Score: 3\n",
            "Section: 1  Predicted Score: 3.572643280029297  True Score: 3\n",
            "Section: 2  Predicted Score: 3.5628910064697266  True Score: 3\n",
            "\n",
            "\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "0y2O_w0nLXGq",
        "colab": {}
      },
      "source": [
        "## Keras Implementation Using Glove Embedding:\n",
        " - Here we provide an alternative implementation using Keras and a Glove embedding."
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "rwT8dOgcLXLB",
        "outputId": "131385f6-6235-46f8-9954-303764191a9a",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 168
        }
      },
      "source": [
        "# Supress unnecessary warnings so that presentation looks clean\n",
        "import warnings\n",
        "warnings.filterwarnings('ignore')\n",
        "\n",
        "import keras\n",
        "from keras import layers, models, optimizers\n",
        "import tensorflow as tf\n",
        "config = tf.ConfigProto( device_count = {'GPU': 1} ) \n",
        "sess = tf.Session(config=config) \n",
        "keras.backend.set_session(sess)\n",
        "\n",
        "from keras import backend as K\n",
        "K.tensorflow_backend._get_available_gpus()"
      ],
      "execution_count": 13,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Using TensorFlow backend.\n"
          ],
          "name": "stderr"
        },
        {
          "output_type": "display_data",
          "data": {
            "text/html": [
              "<p style=\"color: red;\">\n",
              "The default version of TensorFlow in Colab will soon switch to TensorFlow 2.x.<br>\n",
              "We recommend you <a href=\"https://www.tensorflow.org/guide/migrate\" target=\"_blank\">upgrade</a> now \n",
              "or ensure your notebook will continue to use TensorFlow 1.x via the <code>%tensorflow_version 1.x</code> magic:\n",
              "<a href=\"https://colab.research.google.com/notebooks/tensorflow_version.ipynb\" target=\"_blank\">more info</a>.</p>\n"
            ],
            "text/plain": [
              "<IPython.core.display.HTML object>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "stream",
          "text": [
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:190: The name tf.get_default_session is deprecated. Please use tf.compat.v1.get_default_session instead.\n",
            "\n",
            "WARNING:tensorflow:From /usr/local/lib/python3.6/dist-packages/keras/backend/tensorflow_backend.py:207: The name tf.global_variables is deprecated. Please use tf.compat.v1.global_variables instead.\n",
            "\n"
          ],
          "name": "stdout"
        },
        {
          "output_type": "execute_result",
          "data": {
            "text/plain": [
              "['/job:localhost/replica:0/task:0/device:GPU:0']"
            ]
          },
          "metadata": {
            "tags": []
          },
          "execution_count": 13
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "5kb1HfIyLXFC",
        "outputId": "7b43e953-9f4b-4a7c-bc04-c4b8c386e251",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 34
        }
      },
      "source": [
        "word_embeddings = {}\n",
        "for i, row in enumerate(open('/content/drive/My Drive/Colab Notebooks/NLP/glove.6B/glove.6B.300d.txt',encoding=\"utf8\")):\n",
        "    row_val = row.split()\n",
        "    word_embeddings[row_val[0]] = np.array(row_val[1:])\n",
        "    \n",
        "trn_embedding = []\n",
        "trn_tokens = dataset.vocab\n",
        "for row in trn_tokens:\n",
        "    embedd = word_embeddings.get(row,0)\n",
        "    if embedd != 0:\n",
        "        trn_embedding.append(embedd.tolist())\n",
        "    \n",
        "trn_embedding = np.array(trn_embedding)\n",
        "print(trn_embedding.shape)    "
      ],
      "execution_count": 14,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "(10884, 300)\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "P9EwlZ3cUqXf",
        "colab": {}
      },
      "source": [
        "def lstm_model():\n",
        "    input_layer = layers.Input((max_len+1, ))\n",
        "\n",
        "    embedding_layer = layers.Embedding(len(trn_embedding), 300, weights=[trn_embedding], trainable=False)(input_layer)\n",
        "    embedding_layer = layers.SpatialDropout1D(0.3)(embedding_layer)\n",
        "    lstm_layer = layers.LSTM(256)(embedding_layer)\n",
        "\n",
        "    output_layer1 = layers.Dense(64, activation=\"relu\")(lstm_layer)\n",
        "    output_layer1 = layers.Dropout(0.2)(output_layer1)\n",
        "    output_layer2 = layers.Dense(32, activation=\"relu\")(output_layer1)\n",
        "    output_layer2 = layers.Dropout(0.2)(output_layer2)\n",
        "    output_layer3 = layers.Dense(1, activation=\"relu\")(output_layer2)\n",
        "\n",
        "    model = models.Model(inputs=input_layer, outputs=output_layer3)\n",
        "    model.compile(optimizer=optimizers.Adam(), loss='mean_squared_error')\n",
        "    \n",
        "    return model\n"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "tDom58OmUqcF",
        "outputId": "9eaf6577-5cb1-46bf-b94d-9780f3a08e37",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 1000
        }
      },
      "source": [
        "trn_loss_list = []\n",
        "val_loss_list = []\n",
        "trn_acc_list = []\n",
        "val_acc_list = []\n",
        "trn_rmse_list = []\n",
        "val_rmse_list = []\n",
        "\n",
        "history_list = []\n",
        "for cat in range(6):\n",
        "    trn_loss = []\n",
        "    val_loss = []\n",
        "    trn_acc = []\n",
        "    val_acc = []\n",
        "    trn_rmse = []\n",
        "    val_rmse = []\n",
        "\n",
        "    epochs = 10\n",
        "    model = lstm_model()\n",
        "    history = model.fit(train_x, train_y[:,cat], epochs= epochs, verbose=1, \n",
        "                        validation_split= 0.2, batch_size= 100)\n",
        "    \n",
        "    history_list.append(history)\n",
        "    train_predictions = model.predict(train_x, batch_size= 100)\n",
        "    valid_predictions = model.predict(valid_x, batch_size= 100)\n",
        "\n",
        "    trn_rmse = math.sqrt(mean_squared_error(train_predictions, train_y[:,cat]))\n",
        "    val_rmse = math.sqrt(mean_squared_error(valid_predictions, valid_y[:,cat]))\n",
        "\n",
        "    trn_rmse_list.append(trn_rmse)\n",
        "    val_rmse_list.append(val_rmse)"
      ],
      "execution_count": 25,
      "outputs": [
        {
          "output_type": "stream",
          "text": [
            "Train on 1300 samples, validate on 326 samples\n",
            "Epoch 1/10\n",
            "1300/1300 [==============================] - 5s 4ms/step - loss: 3.7908 - val_loss: 0.8713\n",
            "Epoch 2/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 1.3894 - val_loss: 0.4682\n",
            "Epoch 3/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 1.2932 - val_loss: 0.4719\n",
            "Epoch 4/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 1.1496 - val_loss: 0.4905\n",
            "Epoch 5/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 1.0633 - val_loss: 0.4694\n",
            "Epoch 6/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 1.1482 - val_loss: 0.4693\n",
            "Epoch 7/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 1.0600 - val_loss: 0.4667\n",
            "Epoch 8/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 0.9559 - val_loss: 0.4676\n",
            "Epoch 9/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 0.9556 - val_loss: 0.4686\n",
            "Epoch 10/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 1.0323 - val_loss: 0.4672\n",
            "Train on 1300 samples, validate on 326 samples\n",
            "Epoch 1/10\n",
            "1300/1300 [==============================] - 5s 4ms/step - loss: 3.7039 - val_loss: 1.0052\n",
            "Epoch 2/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 1.4878 - val_loss: 0.4445\n",
            "Epoch 3/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 1.2280 - val_loss: 0.4540\n",
            "Epoch 4/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 1.2150 - val_loss: 0.4860\n",
            "Epoch 5/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 1.0813 - val_loss: 0.4523\n",
            "Epoch 6/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 0.9997 - val_loss: 0.4481\n",
            "Epoch 7/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 1.0218 - val_loss: 0.4477\n",
            "Epoch 8/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 0.9779 - val_loss: 0.4697\n",
            "Epoch 9/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 0.9225 - val_loss: 0.4916\n",
            "Epoch 10/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 0.8742 - val_loss: 0.5039\n",
            "Train on 1300 samples, validate on 326 samples\n",
            "Epoch 1/10\n",
            "1300/1300 [==============================] - 5s 4ms/step - loss: 4.6855 - val_loss: 0.7177\n",
            "Epoch 2/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 1.3260 - val_loss: 0.3444\n",
            "Epoch 3/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 1.1659 - val_loss: 0.3431\n",
            "Epoch 4/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 1.0041 - val_loss: 0.3415\n",
            "Epoch 5/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 1.0851 - val_loss: 0.3441\n",
            "Epoch 6/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 1.0373 - val_loss: 0.3420\n",
            "Epoch 7/10\n",
            "1300/1300 [==============================] - 4s 3ms/step - loss: 0.9389 - val_loss: 0.3454\n",
            "Epoch 8/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 0.9411 - val_loss: 0.3468\n",
            "Epoch 9/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 0.9791 - val_loss: 0.3483\n",
            "Epoch 10/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 0.9376 - val_loss: 0.3531\n",
            "Train on 1300 samples, validate on 326 samples\n",
            "Epoch 1/10\n",
            "1300/1300 [==============================] - 5s 4ms/step - loss: 2.6471 - val_loss: 0.3375\n",
            "Epoch 2/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 1.1582 - val_loss: 0.3340\n",
            "Epoch 3/10\n",
            "1300/1300 [==============================] - 4s 3ms/step - loss: 0.9970 - val_loss: 0.3556\n",
            "Epoch 4/10\n",
            "1300/1300 [==============================] - 4s 3ms/step - loss: 0.9488 - val_loss: 0.3501\n",
            "Epoch 5/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 0.8742 - val_loss: 0.3351\n",
            "Epoch 6/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 0.8771 - val_loss: 0.3754\n",
            "Epoch 7/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 0.8292 - val_loss: 0.4678\n",
            "Epoch 8/10\n",
            "1300/1300 [==============================] - 4s 3ms/step - loss: 0.8077 - val_loss: 0.3548\n",
            "Epoch 9/10\n",
            "1300/1300 [==============================] - 4s 3ms/step - loss: 0.7862 - val_loss: 0.3393\n",
            "Epoch 10/10\n",
            "1300/1300 [==============================] - 4s 3ms/step - loss: 0.8275 - val_loss: 0.4099\n",
            "Train on 1300 samples, validate on 326 samples\n",
            "Epoch 1/10\n",
            "1300/1300 [==============================] - 6s 4ms/step - loss: 3.5198 - val_loss: 0.7719\n",
            "Epoch 2/10\n",
            "1300/1300 [==============================] - 4s 3ms/step - loss: 1.2752 - val_loss: 0.5898\n",
            "Epoch 3/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 1.1485 - val_loss: 0.5679\n",
            "Epoch 4/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 1.0894 - val_loss: 0.5001\n",
            "Epoch 5/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 1.0626 - val_loss: 0.5242\n",
            "Epoch 6/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 1.0314 - val_loss: 0.5228\n",
            "Epoch 7/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 1.0790 - val_loss: 0.4719\n",
            "Epoch 8/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 0.9433 - val_loss: 0.4502\n",
            "Epoch 9/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 0.9420 - val_loss: 0.4788\n",
            "Epoch 10/10\n",
            "1300/1300 [==============================] - 4s 3ms/step - loss: 0.9352 - val_loss: 0.5842\n",
            "Train on 1300 samples, validate on 326 samples\n",
            "Epoch 1/10\n",
            "1300/1300 [==============================] - 6s 5ms/step - loss: 3.3900 - val_loss: 0.6391\n",
            "Epoch 2/10\n",
            "1300/1300 [==============================] - 4s 3ms/step - loss: 1.2325 - val_loss: 0.6058\n",
            "Epoch 3/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 1.1309 - val_loss: 0.5452\n",
            "Epoch 4/10\n",
            "1300/1300 [==============================] - 4s 3ms/step - loss: 1.0966 - val_loss: 0.5234\n",
            "Epoch 5/10\n",
            "1300/1300 [==============================] - 4s 3ms/step - loss: 1.0025 - val_loss: 0.4755\n",
            "Epoch 6/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 0.9495 - val_loss: 0.4855\n",
            "Epoch 7/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 0.9256 - val_loss: 0.4779\n",
            "Epoch 8/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 0.9162 - val_loss: 0.4899\n",
            "Epoch 9/10\n",
            "1300/1300 [==============================] - 3s 3ms/step - loss: 0.9083 - val_loss: 0.6643\n",
            "Epoch 10/10\n",
            "1300/1300 [==============================] - 4s 3ms/step - loss: 0.8674 - val_loss: 0.6807\n"
          ],
          "name": "stdout"
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Kufol9LJWQWh",
        "colab_type": "code",
        "colab": {}
      },
      "source": [
        "trn_loss_list = []\n",
        "val_loss_list = []\n",
        "for index in range(6):\n",
        "    history = history_list[index]\n",
        "    trn_loss_list.append(history.history['loss'])\n",
        "    val_loss_list.append(history.history['val_loss'])"
      ],
      "execution_count": 0,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "colab_type": "code",
        "id": "Ony9fmkNUqUb",
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 573
        },
        "outputId": "1175d9fd-05f9-4f37-ef04-57bd8ece40ba"
      },
      "source": [
        "index_to_cat = {\n",
        "    0 : \"ideas/content\",\n",
        "    1 : \"organization\",\n",
        "    2 : \"voice\",\n",
        "    3 : \"word choice\",\n",
        "    4 : \"sentence fluency\",\n",
        "    5 : \"conventions\",\n",
        "             }\n",
        "\n",
        "for index in range(6):\n",
        "    plt.plot(trn_loss_list[index], label=str(index_to_cat[index]))\n",
        "plt.title(\"Training Loss vs Epoch\")\n",
        "plt.xlabel(\"epoch\")\n",
        "plt.ylabel(\"loss\")\n",
        "plt.legend()\n",
        "plt.show()\n",
        "\n",
        "for index in range(6):\n",
        "    plt.plot(val_loss_list[index], label=str(index_to_cat[index]))\n",
        "plt.title(\"Validation Loss vs Epoch\")\n",
        "plt.xlabel(\"epoch\")\n",
        "plt.ylabel(\"loss\")\n",
        "plt.legend()\n",
        "plt.show()\n"
      ],
      "execution_count": 27,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEWCAYAAABrDZDcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        },
        {
          "output_type": "display_data",
          "data": {
            "image/png": "iVBORw0KGgoAAAANSUhEUgAAAYIAAAEWCAYAAABrDZDcAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nhUKhGOcoIVAoFIpxjhIChUKhGOf8fxC7c8YpGnqxAAAAAElFTkSuQmCC\n",
            "text/plain": [
              "<Figure size 432x288 with 1 Axes>"
            ]
          },
          "metadata": {
            "tags": []
          }
        }
      ]
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "colab_type": "text",
        "id": "aKAndDymnJrS"
      },
      "source": [
        ""
      ]
    }
  ]
}