Shakespearean_Text_Prediction.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "include_colab_link": true
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "view-in-github",
        "colab_type": "text"
      },
      "source": [
        "<a href=\"https://colab.research.google.com/github/Ishwin9/text-prediction/blob/main/Shakespearean_Text_Prediction.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open In Colab\"/></a>"
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Shakespeare Bot: An NLP Text Prediction Neural Network\n"
      ],
      "metadata": {
        "id": "uGMaXzlEy_FR"
      }
    },
    {
      "cell_type": "markdown",
      "metadata": {
        "id": "qF_ESTSJT9IX"
      },
      "source": [
        "## Step 1: Data Collection + Data Cleaning\n"
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "dataurl = \"https://gist.githubusercontent.com/Ishwin9/a79c0a9e6af14e4cf819768d2f560f1a/raw/0df4aff0453895874b2d1300793be3537489ed8f/gistfile1.txt\""
      ],
      "metadata": {
        "id": "XXXl0HszeDVY"
      },
      "execution_count": 21,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "-bxV3YjC7psj",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "5101c3a3-ad29-4320-fe31-c6f67fcb8dad"
      },
      "source": [
        "\n",
        "from urllib.request import urlopen\n",
        "data = urlopen(dataurl).read().decode('utf-8').split(\"\\n\")\n",
        "print(\"Number of monologues: {}\".format(len(data)))\n",
        "\n",
        "unique = set([w for s in data for w in s.split()])\n",
        "print(\"Number of unique words: {}\".format(len(unique)))"
      ],
      "execution_count": 22,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Number of monologues: 639\n",
            "Number of unique words: 4772\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "import pandas as pd\n",
        "import matplotlib.pyplot as plt\n",
        "from wordcloud import WordCloud\n",
        "\n",
        "# get rid of blank lines\n",
        "data = [line for line in data if line != '']\n",
        "\n",
        "# plot word cloud to see most ised words\n",
        "wordcloud = WordCloud().generate(' '.join(data))\n",
        "plt.imshow(wordcloud, interpolation='bilinear')\n",
        "plt.axis(\"off\")\n",
        "plt.show()"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/",
          "height": 285
        },
        "id": "rp-vTM5a4ysR",
        "outputId": "95f83021-8b93-4e89-dcde-a326c2b40a8f"
      },
      "execution_count": 23,
      "outputs": [
        {
          "output_type": "display_data",
          "data": {
            "text/plain": [
              "<Figure size 640x480 with 1 Axes>"
            ],
            "image/png": "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\n"
          },
          "metadata": {}
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "Considering the data is the script from the Romeo and Juliet play, it seems fitting that \"love\" is one of the most used words."
      ],
      "metadata": {
        "id": "7u6R9031L5wv"
      }
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "_6RgRCeEATzo",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "5e0a54d5-3513-487a-9c08-a9b89ec3133e"
      },
      "source": [
        "import spacy\n",
        "\n",
        "# tokenization\n",
        "nlp = spacy.load(\"en_core_web_sm\", disable=[\"parser\", \"tagger\", \"ner\", \"textcat\"])\n",
        "data = [nlp(s) for s in data]\n",
        "\n",
        "# adding <n> tag to specify a new monologue\n",
        "data = [ [\"<n>\"] + [str(w) for w in s] + [\"</n>\"] for s in data]\n",
        "\n",
        "#split into train and validate\n",
        "train = data[:-5]\n",
        "valid = data[-5:]\n",
        "\n",
        "# flatten the lists into one long string and remove extra whitespace\n",
        "train = [w for s in train for w in s if not w.isspace()]\n",
        "valid = [w for s in valid for w in s if not w.isspace()]"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "/usr/local/lib/python3.10/dist-packages/spacy/pipeline/lemmatizer.py:211: UserWarning: [W108] The rule-based lemmatizer did not find POS annotation for one or more tokens. Check that your pipeline includes components that assign token.pos, typically 'tagger'+'attribute_ruler' or 'morphologizer'.\n",
            "  warnings.warn(Warnings.W108)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "JzYMScnsDQ2V",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "0dbb762e-6aa2-489f-83de-557e9c9f47fa"
      },
      "source": [
        "print(\"Training dataset contains {} lexical types (unique words)\".format(len(set(train))))\n",
        "print(\"Training dataset contains {} lexical tokens (number of words)\".format(len(train)))"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Training dataset contains 3247 lexical types (unique words)\n",
            "Training dataset contains 22039 lexical tokens (number of words)\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "6H0t1qQ0AQhI"
      },
      "source": [
        "import re\n",
        "def simplify_word(w):\n",
        "    # remove extra punctuation\n",
        "    w = w.replace(\"-\", \"\").replace(\"~\",\"\")\n",
        "\n",
        "    # replace numbers with # sign\n",
        "    w = re.sub('\\d', '#', w)\n",
        "\n",
        "    # split words into root and stem\n",
        "    if len(w) > 3 and w[-2:] in set([\"ed\", \"er\",\"ly\"]):\n",
        "        return [w[:-2], w[-2:]]\n",
        "    elif len(w) > 4 and w[-3:] in set([\"ing\",\"'re\"]):\n",
        "        return [w[:-3], w[-3:]]\n",
        "    return [w]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "Gp6CiTzOQnua"
      },
      "source": [
        "# loop through all the words and perform data cleaning\n",
        "train_clean = []\n",
        "for w in train:\n",
        "    for piece in simplify_word(w):\n",
        "        train_clean.append(piece)\n",
        "\n",
        "valid_clean = []\n",
        "for w in valid:\n",
        "    for piece in simplify_word(w):\n",
        "        valid_clean.append(piece)\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QKCex8fuRoLR"
      },
      "source": [
        "from collections import Counter\n",
        "\n",
        "counts_clean = Counter(train_clean)\n",
        "\n",
        "# only keeping the tokens in validation set that are already in train\n",
        "train_new = [w for w in train_clean]\n",
        "valid_new = [w for w in valid_clean if w in counts_clean]\n"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "ZIcK-thVBcNW"
      },
      "source": [
        "# create a mapping from words <-> numbers\n",
        "vocab = set(train_new)\n",
        "word_to_num = {}\n",
        "num_to_word = {}\n",
        "for num, word in enumerate(vocab):\n",
        "  word_to_num[word] = num\n",
        "  num_to_word[num] = word\n",
        "\n",
        "# convert our dataset tokens into integers\n",
        "import torch\n",
        "train = torch.LongTensor(len(train_new))\n",
        "for i in range(len(train_new)):\n",
        "  train[i] = word_to_num[train_new[i]]\n",
        "\n",
        "valid = torch.LongTensor(len(valid_new))\n",
        "for i in range(len(valid_new)):\n",
        "  valid[i] = word_to_num[valid_new[i]]"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "FetTV331Vqsq"
      },
      "source": [
        "num_batches = 30\n",
        "num_tokens_in_batch = 200\n",
        "\n",
        "# use a GPU for computation\n",
        "device = torch.device(\"cuda\")\n",
        "\n",
        "# divide data into `num_batches` batches\n",
        "def batchify(data, bsz):\n",
        "    nbatch = data.size(0) // bsz\n",
        "    data = data.narrow(0, 0, nbatch * bsz)\n",
        "    data = data.view(bsz, -1).t().contiguous() # reshape\n",
        "    return data.to(device)\n",
        "\n",
        "def get_batch(source, i, seq_len):\n",
        "    seq_len = min(seq_len, len(source) - 1 - i)\n",
        "    data = source[i:i+seq_len]\n",
        "\n",
        "    # targets are the same sequce of words but index shifed to the right, i.e. the next word\n",
        "    target = source[i+1:i+1+seq_len].view(-1)\n",
        "    return data, target\n",
        "\n",
        "def repackage_hidden(h):\n",
        "    if isinstance(h, torch.Tensor):\n",
        "        return h.detach()\n",
        "    else:\n",
        "        return tuple(repackage_hidden(v) for v in h)\n",
        "\n",
        "train = batchify(train, num_batches)\n",
        "valid = batchify(valid, num_batches)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "dAazfehEBgT4"
      },
      "source": [
        "import torch.nn as nn\n",
        "\n",
        "class EncoderDecoder(nn.Module):\n",
        "  def __init__(self):\n",
        "    super(EncoderDecoder, self).__init__()\n",
        "    self.embed_size = 300\n",
        "    self.hidden_size = 600\n",
        "\n",
        "    # create embedding matrix, i.e. map each word to vector of ints\n",
        "    self.embedding = nn.Embedding(len(vocab), self.embed_size)\n",
        "\n",
        "    # initialize embedding matrix with a random uniform distribution\n",
        "    initrange = 0.1\n",
        "    self.embedding.weight.data.uniform_(-initrange, initrange)\n",
        "\n",
        "\n",
        "    # RNN (LSTM) with dropout\n",
        "    self.rnn = nn.LSTM(input_size=self.embed_size,\n",
        "                       hidden_size=self.hidden_size)\n",
        "    self.shrink = nn.Linear(self.hidden_size,\n",
        "                            self.embed_size)\n",
        "    self.drop = nn.Dropout(p=0.5)\n",
        "\n",
        "    # decode layer predicts next words from our model\n",
        "    self.decode = nn.Linear(self.embed_size, self.embedding.weight.size(0))\n",
        "    self.decode.weight = self.embedding.weight\n",
        "    self.decode.bias.data.zero_()\n",
        "\n",
        "\n",
        "  def forward(self, input, hidden=None):\n",
        "    x = self.embedding(input)\n",
        "    if hidden is not None:\n",
        "      x, hidden = self.rnn(x, hidden)\n",
        "    else:\n",
        "      x, hidden = self.rnn(x)\n",
        "    x = F.relu(self.shrink(self.drop(x)))\n",
        "    x = self.decode(x)\n",
        "    return hidden, x"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "UPA-jER1YPN1"
      },
      "source": [
        "import torch.nn.functional as F\n",
        "def training(model, data, targets, lr, hidden):\n",
        "  # reset gradient\n",
        "  model.zero_grad()\n",
        "\n",
        "  # get prediction\n",
        "  hidden, prediction_vector = model(data, hidden)\n",
        "  prediction_vector = prediction_vector.view(targets.size(0), -1)\n",
        "\n",
        "  # calculate loss\n",
        "  loss = F.cross_entropy(input = prediction_vector, target = targets)\n",
        "\n",
        "  # get gradients and perform back-propagation\n",
        "  loss.backward()\n",
        "  torch.nn.utils.clip_grad_norm_(model.parameters(), 0.25)\n",
        "  for p in model.parameters():\n",
        "      if p.grad is not None:\n",
        "        p.data.add_(-lr, p.grad.data)\n",
        "\n",
        "  return loss.item(), repackage_hidden(hidden)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "wEE6G8FhZ_px"
      },
      "source": [
        "def validation(model):\n",
        " # checks the predictions without updating parameters\n",
        "  model.eval()\n",
        "  hidden = None\n",
        "  valid_loss = 0\n",
        "  for i in range(0, valid.size(0) - num_tokens_in_batch, num_tokens_in_batch):\n",
        "    data, targets = get_batch(valid, i, num_tokens_in_batch)\n",
        "    hidden, prd = model(data, hidden)\n",
        "    hidden = repackage_hidden(hidden)\n",
        "\n",
        "    # softmax calculation\n",
        "    softmax = torch.exp(prd) / torch.exp(prd).sum(dim=1, keepdim=True)\n",
        "\n",
        "    # cross-entropy loss calculation\n",
        "    log_softmax = torch.log(softmax)\n",
        "    loss = -log_softmax[torch.arange(log_softmax.size(0)), targets].sum()\n",
        "    validation_loss += loss.item()\n",
        "  return valid_loss / (valid.size(0)/num_tokens_in_batch)"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "zUM8SRBgBjBQ"
      },
      "source": [
        "import numpy as np\n",
        "import time\n",
        "\n",
        "model = EncoderDecoder().float().to(device)\n",
        "prev_valid_loss = 1e100\n",
        "\n",
        "learning_rate = 20\n",
        "num_batches = 20"
      ],
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "import time\n",
        "\n",
        "num_epochs = 150\n",
        "timing = time.time()\n",
        "\n",
        "# train specified number of epochs\n",
        "for epoch in range(num_epochs):\n",
        "\n",
        "  model.train()\n",
        "  hidden = None\n",
        "  train_loss = 0\n",
        "  start_time = time.time()\n",
        "  for i in range(0, train.size(0) - 1, num_tokens_in_batch):\n",
        "    data, targets = get_batch(train, i, num_tokens_in_batch)\n",
        "\n",
        "    # run the model and perform backpropagation\n",
        "    loss, hidden = training(model, data, targets, learning_rate, hidden)\n",
        "    train_loss += loss\n",
        "\n",
        "  # evaluate how well the model predicts unseen validation data\n",
        "  valid_loss = validation(model)\n",
        "\n",
        "# updte learning rate if model is getting worse\n",
        "  if valid_loss > prev_valid_loss:\n",
        "    learning_rate /= 4.0\n",
        "\n",
        "  train_loss /= (train.size(0)/num_tokens_in_batch)\n",
        "  finish_time = time.time()\n",
        "  print(\"Epoch {:2} took {:3.2f}s with train accuracy: {:7.2f}\"\\\n",
        "        \" and validation: {:7.2f}\".format(epoch, finish_time - start_time,\n",
        "                                          np.exp(train_loss),\n",
        "                                          np.exp(valid_loss)))\n",
        "\n",
        "  prev_valid_loss = valid_loss\n",
        "\n",
        "total_time = (time.time() - timing)/60\n",
        "print(\"Completed {} epochs in {:5.3f} minutes\".format(num_epochs, total_time))"
      ],
      "metadata": {
        "id": "UDxlTO2hrQ3S",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "42542f87-fb64-49b1-8a41-4452e0fd0d3a"
      },
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stderr",
          "text": [
            "<ipython-input-12-bbf4e0e731c6>:18: UserWarning: This overload of add_ is deprecated:\n",
            "\tadd_(Number alpha, Tensor other)\n",
            "Consider using one of the following signatures instead:\n",
            "\tadd_(Tensor other, *, Number alpha) (Triggered internally at ../torch/csrc/utils/python_arg_parser.cpp:1485.)\n",
            "  p.data.add_(-lr, p.grad.data)\n"
          ]
        },
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Epoch  0 took 1.25s with train accuracy: 186304.96 and validation:    1.00\n",
            "Epoch  1 took 0.19s with train accuracy: 28264.41 and validation:    1.00\n",
            "Epoch  2 took 0.19s with train accuracy: 948389.08 and validation:    1.00\n",
            "Epoch  3 took 0.19s with train accuracy: 7341.78 and validation:    1.00\n",
            "Epoch  4 took 0.19s with train accuracy: 8497.49 and validation:    1.00\n",
            "Epoch  5 took 0.19s with train accuracy: 1693.77 and validation:    1.00\n",
            "Epoch  6 took 0.19s with train accuracy: 5062.26 and validation:    1.00\n",
            "Epoch  7 took 0.19s with train accuracy: 1234.88 and validation:    1.00\n",
            "Epoch  8 took 0.19s with train accuracy: 1530.73 and validation:    1.00\n",
            "Epoch  9 took 0.19s with train accuracy: 3222.08 and validation:    1.00\n",
            "Epoch 10 took 0.19s with train accuracy: 1521.44 and validation:    1.00\n",
            "Epoch 11 took 0.20s with train accuracy: 6034.60 and validation:    1.00\n",
            "Epoch 12 took 0.19s with train accuracy: 8156.07 and validation:    1.00\n",
            "Epoch 13 took 0.19s with train accuracy: 3869.06 and validation:    1.00\n",
            "Epoch 14 took 0.19s with train accuracy: 1079.86 and validation:    1.00\n",
            "Epoch 15 took 0.19s with train accuracy: 1033.55 and validation:    1.00\n",
            "Epoch 16 took 0.19s with train accuracy:  827.78 and validation:    1.00\n",
            "Epoch 17 took 0.19s with train accuracy:  811.66 and validation:    1.00\n",
            "Epoch 18 took 0.19s with train accuracy:  656.22 and validation:    1.00\n",
            "Epoch 19 took 0.19s with train accuracy:  814.76 and validation:    1.00\n",
            "Epoch 20 took 0.19s with train accuracy:  644.83 and validation:    1.00\n",
            "Epoch 21 took 0.19s with train accuracy:  591.89 and validation:    1.00\n",
            "Epoch 22 took 0.19s with train accuracy:  672.27 and validation:    1.00\n",
            "Epoch 23 took 0.19s with train accuracy:  726.43 and validation:    1.00\n",
            "Epoch 24 took 0.19s with train accuracy:  573.59 and validation:    1.00\n",
            "Epoch 25 took 0.19s with train accuracy:  436.42 and validation:    1.00\n",
            "Epoch 26 took 0.19s with train accuracy:  514.17 and validation:    1.00\n",
            "Epoch 27 took 0.19s with train accuracy:  509.86 and validation:    1.00\n",
            "Epoch 28 took 0.19s with train accuracy:  418.67 and validation:    1.00\n",
            "Epoch 29 took 0.19s with train accuracy:  434.41 and validation:    1.00\n",
            "Epoch 30 took 0.19s with train accuracy:  423.08 and validation:    1.00\n",
            "Epoch 31 took 0.19s with train accuracy:  390.42 and validation:    1.00\n",
            "Epoch 32 took 0.20s with train accuracy:  429.24 and validation:    1.00\n",
            "Epoch 33 took 0.19s with train accuracy:  431.03 and validation:    1.00\n",
            "Epoch 34 took 0.20s with train accuracy:  417.35 and validation:    1.00\n",
            "Epoch 35 took 0.19s with train accuracy:  380.31 and validation:    1.00\n",
            "Epoch 36 took 0.19s with train accuracy:  394.65 and validation:    1.00\n",
            "Epoch 37 took 0.20s with train accuracy:  366.07 and validation:    1.00\n",
            "Epoch 38 took 0.19s with train accuracy:  357.14 and validation:    1.00\n",
            "Epoch 39 took 0.19s with train accuracy:  328.08 and validation:    1.00\n",
            "Epoch 40 took 0.20s with train accuracy:  337.97 and validation:    1.00\n",
            "Epoch 41 took 0.20s with train accuracy:  329.86 and validation:    1.00\n",
            "Epoch 42 took 0.19s with train accuracy:  294.22 and validation:    1.00\n",
            "Epoch 43 took 0.19s with train accuracy:  283.85 and validation:    1.00\n",
            "Epoch 44 took 0.19s with train accuracy:  310.17 and validation:    1.00\n",
            "Epoch 45 took 0.20s with train accuracy:  274.69 and validation:    1.00\n",
            "Epoch 46 took 0.19s with train accuracy:  278.56 and validation:    1.00\n",
            "Epoch 47 took 0.20s with train accuracy:  288.96 and validation:    1.00\n",
            "Epoch 48 took 0.20s with train accuracy:  257.48 and validation:    1.00\n",
            "Epoch 49 took 0.19s with train accuracy:  270.06 and validation:    1.00\n",
            "Epoch 50 took 0.19s with train accuracy:  258.63 and validation:    1.00\n",
            "Epoch 51 took 0.20s with train accuracy:  239.58 and validation:    1.00\n",
            "Epoch 52 took 0.20s with train accuracy:  244.71 and validation:    1.00\n",
            "Epoch 53 took 0.19s with train accuracy:  248.28 and validation:    1.00\n",
            "Epoch 54 took 0.20s with train accuracy:  259.50 and validation:    1.00\n",
            "Epoch 55 took 0.20s with train accuracy:  245.74 and validation:    1.00\n",
            "Epoch 56 took 0.19s with train accuracy:  216.56 and validation:    1.00\n",
            "Epoch 57 took 0.19s with train accuracy:  220.75 and validation:    1.00\n",
            "Epoch 58 took 0.20s with train accuracy:  209.75 and validation:    1.00\n",
            "Epoch 59 took 0.20s with train accuracy:  202.37 and validation:    1.00\n",
            "Epoch 60 took 0.20s with train accuracy:  205.60 and validation:    1.00\n",
            "Epoch 61 took 0.20s with train accuracy:  204.04 and validation:    1.00\n",
            "Epoch 62 took 0.20s with train accuracy:  178.18 and validation:    1.00\n",
            "Epoch 63 took 0.20s with train accuracy:  171.88 and validation:    1.00\n",
            "Epoch 64 took 0.19s with train accuracy:  192.57 and validation:    1.00\n",
            "Epoch 65 took 0.20s with train accuracy:  187.44 and validation:    1.00\n",
            "Epoch 66 took 0.20s with train accuracy:  184.26 and validation:    1.00\n",
            "Epoch 67 took 0.20s with train accuracy:  177.50 and validation:    1.00\n",
            "Epoch 68 took 0.20s with train accuracy:  174.10 and validation:    1.00\n",
            "Epoch 69 took 0.19s with train accuracy:  196.99 and validation:    1.00\n",
            "Epoch 70 took 0.20s with train accuracy:  167.88 and validation:    1.00\n",
            "Epoch 71 took 0.19s with train accuracy:  165.08 and validation:    1.00\n",
            "Epoch 72 took 0.19s with train accuracy:  155.76 and validation:    1.00\n",
            "Epoch 73 took 0.20s with train accuracy:  165.76 and validation:    1.00\n",
            "Epoch 74 took 0.19s with train accuracy:  150.73 and validation:    1.00\n",
            "Epoch 75 took 0.20s with train accuracy:  150.73 and validation:    1.00\n",
            "Epoch 76 took 0.19s with train accuracy:  141.64 and validation:    1.00\n",
            "Epoch 77 took 0.20s with train accuracy:  139.98 and validation:    1.00\n",
            "Epoch 78 took 0.20s with train accuracy:  139.73 and validation:    1.00\n",
            "Epoch 79 took 0.20s with train accuracy:  137.99 and validation:    1.00\n",
            "Epoch 80 took 0.20s with train accuracy:  125.75 and validation:    1.00\n",
            "Epoch 81 took 0.20s with train accuracy:  126.15 and validation:    1.00\n",
            "Epoch 82 took 0.20s with train accuracy:  125.65 and validation:    1.00\n",
            "Epoch 83 took 0.20s with train accuracy:  119.62 and validation:    1.00\n",
            "Epoch 84 took 0.19s with train accuracy:  115.86 and validation:    1.00\n",
            "Epoch 85 took 0.20s with train accuracy:  116.17 and validation:    1.00\n",
            "Epoch 86 took 0.20s with train accuracy:  103.22 and validation:    1.00\n",
            "Epoch 87 took 0.20s with train accuracy:  114.42 and validation:    1.00\n",
            "Epoch 88 took 0.19s with train accuracy:  105.82 and validation:    1.00\n",
            "Epoch 89 took 0.20s with train accuracy:  105.94 and validation:    1.00\n",
            "Epoch 90 took 0.20s with train accuracy:  101.64 and validation:    1.00\n",
            "Epoch 91 took 0.20s with train accuracy:  100.21 and validation:    1.00\n",
            "Epoch 92 took 0.20s with train accuracy:  104.74 and validation:    1.00\n",
            "Epoch 93 took 0.20s with train accuracy:   95.97 and validation:    1.00\n",
            "Epoch 94 took 0.20s with train accuracy:   91.70 and validation:    1.00\n",
            "Epoch 95 took 0.20s with train accuracy:   88.19 and validation:    1.00\n",
            "Epoch 96 took 0.20s with train accuracy:   86.14 and validation:    1.00\n",
            "Epoch 97 took 0.20s with train accuracy:   85.60 and validation:    1.00\n",
            "Epoch 98 took 0.19s with train accuracy:   78.71 and validation:    1.00\n",
            "Epoch 99 took 0.20s with train accuracy:   79.72 and validation:    1.00\n",
            "Epoch 100 took 0.20s with train accuracy:   75.07 and validation:    1.00\n",
            "Epoch 101 took 0.20s with train accuracy:   75.25 and validation:    1.00\n",
            "Epoch 102 took 0.20s with train accuracy:   72.18 and validation:    1.00\n",
            "Epoch 103 took 0.20s with train accuracy:   70.97 and validation:    1.00\n",
            "Epoch 104 took 0.20s with train accuracy:   67.87 and validation:    1.00\n",
            "Epoch 105 took 0.20s with train accuracy:   68.61 and validation:    1.00\n",
            "Epoch 106 took 0.20s with train accuracy:   67.46 and validation:    1.00\n",
            "Epoch 107 took 0.20s with train accuracy:   63.90 and validation:    1.00\n",
            "Epoch 108 took 0.20s with train accuracy:   61.18 and validation:    1.00\n",
            "Epoch 109 took 0.20s with train accuracy:   61.03 and validation:    1.00\n",
            "Epoch 110 took 0.20s with train accuracy:   58.62 and validation:    1.00\n",
            "Epoch 111 took 0.20s with train accuracy:   54.59 and validation:    1.00\n",
            "Epoch 112 took 0.20s with train accuracy:   53.28 and validation:    1.00\n",
            "Epoch 113 took 0.20s with train accuracy:   50.86 and validation:    1.00\n",
            "Epoch 114 took 0.19s with train accuracy:   52.66 and validation:    1.00\n",
            "Epoch 115 took 0.20s with train accuracy:   50.61 and validation:    1.00\n",
            "Epoch 116 took 0.20s with train accuracy:   45.96 and validation:    1.00\n",
            "Epoch 117 took 0.20s with train accuracy:   45.51 and validation:    1.00\n",
            "Epoch 118 took 0.20s with train accuracy:   43.36 and validation:    1.00\n",
            "Epoch 119 took 0.20s with train accuracy:   41.80 and validation:    1.00\n",
            "Epoch 120 took 0.20s with train accuracy:   40.69 and validation:    1.00\n",
            "Epoch 121 took 0.20s with train accuracy:   40.46 and validation:    1.00\n",
            "Epoch 122 took 0.20s with train accuracy:   36.69 and validation:    1.00\n",
            "Epoch 123 took 0.20s with train accuracy:   36.33 and validation:    1.00\n",
            "Epoch 124 took 0.20s with train accuracy:   34.40 and validation:    1.00\n",
            "Epoch 125 took 0.20s with train accuracy:   33.25 and validation:    1.00\n",
            "Epoch 126 took 0.20s with train accuracy:   31.98 and validation:    1.00\n",
            "Epoch 127 took 0.21s with train accuracy:   31.76 and validation:    1.00\n",
            "Epoch 128 took 0.20s with train accuracy:   30.68 and validation:    1.00\n",
            "Epoch 129 took 0.20s with train accuracy:   27.61 and validation:    1.00\n",
            "Epoch 130 took 0.20s with train accuracy:   29.06 and validation:    1.00\n",
            "Epoch 131 took 0.20s with train accuracy:   26.24 and validation:    1.00\n",
            "Epoch 132 took 0.20s with train accuracy:   24.70 and validation:    1.00\n",
            "Epoch 133 took 0.20s with train accuracy:   24.48 and validation:    1.00\n",
            "Epoch 134 took 0.20s with train accuracy:   22.60 and validation:    1.00\n",
            "Epoch 135 took 0.20s with train accuracy:   21.67 and validation:    1.00\n",
            "Epoch 136 took 0.20s with train accuracy:   22.18 and validation:    1.00\n",
            "Epoch 137 took 0.20s with train accuracy:   20.20 and validation:    1.00\n",
            "Epoch 138 took 0.20s with train accuracy:   18.83 and validation:    1.00\n",
            "Epoch 139 took 0.20s with train accuracy:   18.49 and validation:    1.00\n",
            "Epoch 140 took 0.20s with train accuracy:   17.06 and validation:    1.00\n",
            "Epoch 141 took 0.20s with train accuracy:   16.40 and validation:    1.00\n",
            "Epoch 142 took 0.20s with train accuracy:   16.29 and validation:    1.00\n",
            "Epoch 143 took 0.20s with train accuracy:   15.75 and validation:    1.00\n",
            "Epoch 144 took 0.20s with train accuracy:   15.60 and validation:    1.00\n",
            "Epoch 145 took 0.20s with train accuracy:   14.25 and validation:    1.00\n",
            "Epoch 146 took 0.20s with train accuracy:   13.55 and validation:    1.00\n",
            "Epoch 147 took 0.20s with train accuracy:   12.85 and validation:    1.00\n",
            "Epoch 148 took 0.20s with train accuracy:   12.36 and validation:    1.00\n",
            "Epoch 149 took 0.20s with train accuracy:   11.57 and validation:    1.00\n",
            "Completed 150 epochs in 0.511 minutes\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "metadata": {
        "id": "QkSJrjFhig8l",
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "outputId": "820a0039-3192-4fe2-9b7e-86b47eb9f4d9"
      },
      "source": [
        "from textwrap import wrap\n",
        "\n",
        "# words to start the sentence with\n",
        "prefixes = ['<n> love', '<n> shall', '<n> thou', '<n> the']\n",
        "\n",
        "for prefix in prefixes:\n",
        "  words_to_generate = 15\n",
        "\n",
        "  # set the model to be in evaluation mode (i.e. no backprop!)\n",
        "  model.eval()\n",
        "\n",
        "  argmax_sent = None\n",
        "  argmax_prob = 0\n",
        "  collection = []\n",
        "  for item in range(10):\n",
        "    # make our starting words test data\n",
        "    test = [word_to_num[word] for word in prefix.split()]\n",
        "    probabilities = []\n",
        "\n",
        "    # run the model on the same initial input and it's own generations until\n",
        "    # we reach `word_to_generate`\n",
        "    for w in range(words_to_generate):\n",
        "      input = torch.from_numpy(np.array(test)).to(device)\n",
        "      _, output = model(input.view(-1,1))\n",
        "\n",
        "      # get the predicted next word\n",
        "      last_pred = output[-1,:,:].squeeze()\n",
        "\n",
        "      # sample words from the probability prediction, rather than using the highest probability each time\n",
        "      if item > 0:\n",
        "        last_pred /= 0.70\n",
        "\n",
        "        # create distribution\n",
        "        dist = torch.distributions.categorical.Categorical(logits=last_pred)\n",
        "\n",
        "        predicted_idx = dist.sample().item()\n",
        "\n",
        "      else:\n",
        "        # take the most probable word\n",
        "        _, predicted_idx = last_pred.max(0)\n",
        "        predicted_idx = predicted_idx.item()\n",
        "\n",
        "      # save this probability\n",
        "      value = F.log_softmax(last_pred,-1)[predicted_idx].item()\n",
        "      test.append(predicted_idx)\n",
        "      probabilities.append(value)\n",
        "\n",
        "    if item > 0:\n",
        "      # save generated sentence and its score to a list\n",
        "      generation = (np.exp(np.sum(probabilities)), \\\n",
        "                        \" \".join([num_to_word[w] for w in test]))\n",
        "      if generation not in collection:\n",
        "        collection.append(generation)\n",
        "    else:\n",
        "      argmax_sent = \" \".join([num_to_word[w] for w in test])\n",
        "      argmax_prob = np.exp(np.sum(probabilities))\n",
        "\n",
        "  # get the best model predictions\n",
        "  collection.sort()\n",
        "  collection.reverse()\n",
        "  print(\"Choosing Best Probability Generation:\")\n",
        "  print(\"{:.2E}:  {}\\n\".format(argmax_prob,\"\\n\\t\\t\".join(wrap(argmax_sent))))\n",
        "  print(\"\\nSampled Generations:\")\n",
        "  for probability, sent in collection[:10]:\n",
        "    print(\"{:.2E}:  {}\\n\".format(probability, \"\\n\\t\\t\".join(wrap(sent))))\n",
        "\n",
        "  print(\"\\n\\n\\n\")"
      ],
      "execution_count": null,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Choosing Best Probability Generation:\n",
            "1.68E-07:  <n> love , to my ghost ly confessor , And , in a lenten thread , in\n",
            "\n",
            "\n",
            "Sampled Generations:\n",
            "1.74E-06:  <n> love , to my ghost ly confessor 's voice , And , with a martial\n",
            "\t\tflow\n",
            "\n",
            "2.50E-08:  <n> love , to my ghost ly confessor 's the very fair soon of the\n",
            "\t\tgolden fair\n",
            "\n",
            "8.47E-09:  <n> love , I 'll tell my Romeo , that is my Romeo is the variable of\n",
            "\n",
            "9.07E-11:  <n> love , so , be my lie , sir , what , be gone , Tybalt\n",
            "\n",
            "4.24E-11:  <n> love , I would have thee at his wedd ing eyes ! Why , death !\n",
            "\n",
            "3.58E-13:  <n> love , I 'll have no long ; these thy tongue is a Montague , Who\n",
            "\n",
            "1.09E-13:  <n> love , you are this night . </n> <n> Here comes such have this ! a\n",
            "\n",
            "3.80E-14:  <n> love , to my ghost ly confessor </n> <n> Of this , Romeo ; Who\n",
            "\t\twhere\n",
            "\n",
            "3.21E-16:  <n> love , by this news ? O , how my heart abhors To hear you wear\n",
            "\n",
            "\n",
            "\n",
            "\n",
            "\n",
            "Choosing Best Probability Generation:\n",
            "2.91E-09:  <n> shall you like the cause , and you shall be a man ; For , in\n",
            "\n",
            "\n",
            "Sampled Generations:\n",
            "4.18E-06:  <n> shall go with him . </n> <n> How now , Romeo ! what a beast was\n",
            "\n",
            "1.60E-08:  <n> shall you go to him at our distemperature ; Or if thou wilt , thou\n",
            "\t\thadst\n",
            "\n",
            "1.96E-09:  <n> shall you like the cause ? </n> <n> That now , so , I pray you\n",
            "\n",
            "1.38E-09:  <n> shall go with him . </n> <n> O , banishment ! come , she could up\n",
            "\n",
            "1.80E-10:  <n> shall you like the cause , or ' I hate , Is it did the prettiest\n",
            "\n",
            "1.28E-10:  <n> shall you like the cause is two ; For Romeo would my lady ! O ,\n",
            "\n",
            "1.26E-15:  <n> shall go along ; An Romeo is my lord , I 'll talk some aqua hand\n",
            "\n",
            "3.00E-16:  <n> shall we d , Shuts up my food , as shall say , now it were\n",
            "\n",
            "5.82E-20:  <n> shall be my convoy That thou hast more , though thou first , All\n",
            "\t\tsuit ;\n",
            "\n",
            "\n",
            "\n",
            "\n",
            "\n",
            "Choosing Best Probability Generation:\n",
            "3.26E-07:  <n> thou wilt , for the men 's death , And she shall scant . </n> <n>\n",
            "\n",
            "\n",
            "Sampled Generations:\n",
            "5.43E-08:  <n> thou wilt , swear , if you do ; Wilt thou not , Jule ? </n>\n",
            "\n",
            "6.47E-10:  <n> thou art up him . </n> <n> By my troth , I 'll tell thee And\n",
            "\n",
            "8.44E-11:  <n> thou art like new tithe  door ; For ' tis a ' ? O ,\n",
            "\n",
            "4.15E-11:  <n> thou wilt , for a prince , and what learn ing  gentle eye , And\n",
            "\n",
            "1.61E-11:  <n> thou art like that hand , Drawn with thee , coward ! What , shall\n",
            "\t\tyou\n",
            "\n",
            "1.39E-11:  <n> thou wilt , swear , I 'll warrant a candle back . </n> <n> No ,\n",
            "\n",
            "3.42E-12:  <n> thou wilt , thy wit : go , you be even in this night . </n>\n",
            "\n",
            "1.76E-13:  <n> thou wilt , when the tailor doom , more in the night , Thou appear\n",
            "\t\ton\n",
            "\n",
            "4.57E-17:  <n> thou art not well a form . I am none of his hand , Hath we\n",
            "\n",
            "\n",
            "\n",
            "\n",
            "\n",
            "Choosing Best Probability Generation:\n",
            "6.42E-09:  <n> the very pin of my ach ing dew of thy joy is my will : go\n",
            "\n",
            "\n",
            "Sampled Generations:\n",
            "2.01E-07:  <n> the fee  simple of my joy hath the fairies ' midwife , and his\n",
            "\t\tsorrows\n",
            "\n",
            "5.13E-10:  <n> the very pink of his last , in that spar ing fair , For all the\n",
            "\n",
            "3.71E-10:  <n> the very butch er of his joy in the year , And , be my three\n",
            "\n",
            "1.35E-11:  <n> the sweetest honey Is my heart 's love . This , two , come , go\n",
            "\n",
            "1.10E-11:  <n> the rest my husbandaGod With my consent profess'd , thy three\n",
            "\t\thours I see thee\n",
            "\n",
            "2.81E-12:  <n> the rest my very Valentio ! the very butch ing of my weal or light\n",
            "\t\t,\n",
            "\n",
            "4.95E-13:  <n> the very butch er of my behavior , wert thou a poperin locks in\n",
            "\t\tthe golden\n",
            "\n",
            "1.61E-14:  <n> the world , as they say , Her view , measure , of love , as\n",
            "\n",
            "6.45E-17:  <n> the house , of his my man , but a handsome , or a Of death\n",
            "\n",
            "\n",
            "\n",
            "\n",
            "\n"
          ]
        }
      ]
    },
    {
      "cell_type": "markdown",
      "source": [
        "# Flask Deployment"
      ],
      "metadata": {
        "id": "Hnbzv9KJJSYW"
      }
    },
    {
      "cell_type": "code",
      "source": [
        "def prediction(prefix):\n",
        "  # words to start the sentence with\n",
        "  prefixes = [prefix]\n",
        "\n",
        "  # What do we want the model to start the sentence with?\n",
        "  for prefix in prefixes:\n",
        "    words_to_generate = 15\n",
        "\n",
        "    # set the model to be in evaluation mode (i.e. no backprop!)\n",
        "    model.eval()\n",
        "\n",
        "    argmax_sent = None\n",
        "    argmax_prob = 0\n",
        "    collection = []\n",
        "    for item in range(10):\n",
        "      # make our starting words test data\n",
        "      test = [word_to_num[word] for word in prefix.split()]\n",
        "      probabilities = []\n",
        "\n",
        "      # run the model on the same initial input and it's own generations until\n",
        "      # we reach `word_to_generate`\n",
        "      for w in range(words_to_generate):\n",
        "        input = torch.from_numpy(np.array(test)).to(device)\n",
        "        _, output = model(input.view(-1,1))\n",
        "\n",
        "        # get the predicted next word\n",
        "        last_pred = output[-1,:,:].squeeze()\n",
        "\n",
        "        # sample words from the probability prediction, rather than using the highest probability each time\n",
        "        if item > 0:\n",
        "          last_pred /= 0.70\n",
        "\n",
        "          # create distribution\n",
        "          dist = torch.distributions.categorical.Categorical(logits=last_pred)\n",
        "\n",
        "          predicted_idx = dist.sample().item()\n",
        "\n",
        "        else:\n",
        "          # take the most probable word\n",
        "          _, predicted_idx = last_pred.max(0)\n",
        "          predicted_idx = predicted_idx.item()\n",
        "\n",
        "        # save this probability\n",
        "        value = F.log_softmax(last_pred,-1)[predicted_idx].item()\n",
        "        test.append(predicted_idx)\n",
        "        probabilities.append(value)\n",
        "\n",
        "      if item > 0:\n",
        "        # save generated sentence and its score to a list\n",
        "        generation = (np.exp(np.sum(probabilities)), \\\n",
        "                          \" \".join([num_to_word[w] for w in test]))\n",
        "        if generation not in collection:\n",
        "          collection.append(generation)\n",
        "      else:\n",
        "        argmax_sent = \" \".join([num_to_word[w] for w in test])\n",
        "        argmax_prob = np.exp(np.sum(probabilities))\n",
        "\n",
        "    # get the best model predictions\n",
        "    collection.sort()\n",
        "    collection.reverse()\n",
        "\n",
        "    return collection\n"
      ],
      "metadata": {
        "id": "vyaGt6kehC_w"
      },
      "execution_count": null,
      "outputs": []
    },
    {
      "cell_type": "code",
      "source": [
        "from flask import Flask\n",
        "from flask_ngrok import run_with_ngrok\n",
        "app = Flask(__name__)\n",
        "run_with_ngrok(app)\n",
        "@app.route(\"/\")\n",
        "def hom():\n",
        "    return \"Hello World\"\n",
        "\n",
        "@app.route(\"/<prefix>\")\n",
        "def home(prefix):\n",
        "    predictions = prediction(prefix)\n",
        "    html = \"\"\n",
        "    for pred in predictions:\n",
        "      html += \"<h1>\"+pred[-1]+\"</h1>\"\n",
        "    return html\n",
        "\n",
        "if __name__ == '__main__':\n",
        "    app.run()"
      ],
      "metadata": {
        "id": "4w_jPcDqx5ci"
      },
      "execution_count": null,
      "outputs": []
    }
  ]
}