lstm.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Librairies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:From c:\\Users\\yanis\\anaconda3\\Lib\\site-packages\\keras\\src\\losses.py:2976: The name tf.losses.sparse_softmax_cross_entropy is deprecated. Please use tf.compat.v1.losses.sparse_softmax_cross_entropy instead.\n",
      "\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "import torch\n",
    "import torch.nn as nn\n",
    "import torch.optim as optim\n",
    "import torch.nn.functional as F\n",
    "from torch.utils.data import DataLoader\n",
    "from tensorflow.keras.datasets import imdb\n",
    "import matplotlib.pyplot as plt\n",
    "from tensorflow.keras.preprocessing import sequence\n",
    "import random\n",
    "\n",
    "from src.dataset import Dataset\n",
    "from src.lstm import LSTM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Device: GPU if available, CPU else\n",
    "device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Import and preprocessing"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<START> although i had seen <UNK> in a theater way back in <UNK> i couldn't remember anything of the plot except for vague images of kurt thomas running and fighting against a backdrop of stone walls and disappointment regarding the ending br br after reading some of the other reviews i picked up a copy of the newly released dvd to once again enter the world of <UNK> br br it turns out this is one of those films produced during the <UNK> that would go directly to video today the film stars <UNK> <UNK> kurt thomas as jonathan <UNK> <UNK> out of the blue to <UNK> the nation of <UNK> to enter and hopefully win the game a <UNK> <UNK> <UNK> by the khan who <UNK> his people by yelling what sounds like <UNK> power the goal of the mission involves the star wars defense system jonathan is trained in the martial arts by princess <UNK> who never speaks or leaves the house once trained tries to blend in with the <UNK> by wearing a bright red <UNK> with <UNK> of blue and white needless to say <UNK> finds himself running and fighting for his life along the stone streets of <UNK> on his way to a date with destiny and the game br br star kurt thomas was ill served by director robert <UNK> who it looks like was never on the set the so called script is just this side of incompetent see other reviews for the many <UNK> throughout the town of <UNK> has a few good moments but is ultimately ruined by bad editing the ending <UNK> still there's the <UNK> of a good action adventure here a hong kong version with more <UNK> action and faster pace might even be pretty good\n",
      "X shape: (200, 170)\n",
      "len(X[0]): 170\n",
      "len(X[1]): 170\n",
      "X[0]: [   1   14   22   16   43  530  973 1622 1385   65  458 4468   66 3941\n",
      "    4  173   36  256    5   25  100   43  838  112   50  670    2    9\n",
      "   35  480  284    5  150    4  172  112  167    2  336  385   39    4\n",
      "  172 4536 1111   17  546   38   13  447    4  192   50   16    6  147\n",
      " 2025   19   14   22    4 1920 4613  469    4   22   71   87   12   16\n",
      "   43  530   38   76   15   13 1247    4   22   17  515   17   12   16\n",
      "  626   18    2    5   62  386   12    8  316    8  106    5    4 2223\n",
      "    2   16  480   66 3785   33    4  130   12   16   38  619    5   25\n",
      "  124   51   36  135   48   25 1415   33    6   22   12  215   28   77\n",
      "   52    5   14  407   16   82    2    8    4  107  117    2   15  256\n",
      "    4    2    7 3766    5  723   36   71   43  530  476   26  400  317\n",
      "   46    7    4    2 1029   13  104   88    4  381   15  297   98   32\n",
      " 2071   56]\n"
     ]
    }
   ],
   "source": [
    "NUM_WORDS = 5000\n",
    "INDEX_FROM = 3\n",
    "TOTAL_LENGTH = 170\n",
    "\n",
    "# --- Import the IMDB data and only consider the ``top_words``` most used words\n",
    "np.load.__defaults__=(None, True, True, 'ASCII')\n",
    "(X, y), (X_test, y_test) = imdb.load_data(num_words=NUM_WORDS, index_from=INDEX_FROM)\n",
    "np.load.__defaults__=(None, False, True, 'ASCII')\n",
    "\n",
    "word_to_id = imdb.get_word_index()\n",
    "word_to_id = {k:(v+INDEX_FROM) for k,v in word_to_id.items()}\n",
    "word_to_id[\"<PAD>\"] = 0\n",
    "word_to_id[\"<START>\"] = 1\n",
    "word_to_id[\"<UNK>\"] = 2\n",
    "\n",
    "id_to_word = {value:key for key,value in word_to_id.items()}\n",
    "print(' '.join(id_to_word[id] for id in X[1000] ))\n",
    "\n",
    "# --- filter sequences with at least 170 words\n",
    "X = [lst for lst in X if len(lst) >= TOTAL_LENGTH]\n",
    "# --- truncate input sequences\n",
    "X = sequence.pad_sequences(X, maxlen=TOTAL_LENGTH, truncating='post')\n",
    "# --- Keep only 200 sequences\n",
    "N = 200\n",
    "X = X[:N]\n",
    "\n",
    "print(\"X shape:\", X.shape)\n",
    "print(\"len(X[0]):\", len(X[0]))\n",
    "print(\"len(X[1]):\", len(X[1]))\n",
    "print(\"X[0]:\", X[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "LENGTH_TRAIN = 100\n",
    "LENGTH_TEST = 70\n",
    "X_train = X[:, :LENGTH_TRAIN]\n",
    "X_test = X[:, LENGTH_TRAIN:]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Train and test LSTM"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [],
   "source": [
    "HIDDEN_SIZE = 64\n",
    "MODEL_LENGTH = 10\n",
    "BATCH_SIZE = 1\n",
    "N_EPOCHS = 5\n",
    "\n",
    "lstm_model = LSTM(input_size=NUM_WORDS, hidden_size=HIDDEN_SIZE, output_size=NUM_WORDS, model_length=MODEL_LENGTH, batch_size=BATCH_SIZE, device=device)\n",
    "criterion = nn.CrossEntropyLoss(reduction='sum')\n",
    "optimizer = optim.Adam(lstm_model.parameters())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [],
   "source": [
    "X_train_torch = torch.tensor(X_train, dtype = torch.long)\n",
    "X_test_torch = torch.tensor(X_test, dtype = torch.long)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "C:\\Users\\yanis\\AppData\\Local\\Temp\\ipykernel_143216\\2488810157.py:10: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).\n",
      "  input_sequence=torch.tensor(self.topics[index:index+self.model_length, :])\n",
      "C:\\Users\\yanis\\AppData\\Local\\Temp\\ipykernel_143216\\2488810157.py:11: UserWarning: To copy construct from a tensor, it is recommended to use sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than torch.tensor(sourceTensor).\n",
      "  target_sequence=torch.tensor(self.topics[index+1:index+self.model_length+1, :])\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "4929.605200047662\n",
      "4908.216237531931\n",
      "4905.4120459960595\n",
      "4905.9568967408595\n",
      "4904.568143124338\n"
     ]
    }
   ],
   "source": [
    "list_perplexity = []\n",
    "for n in range(N_EPOCHS):\n",
    "    test_loss = 0\n",
    "    shuffle_sample = random.sample(range(N), N)\n",
    "    for i in shuffle_sample:\n",
    "        # Train LSTM model\n",
    "        loss_i = 0\n",
    "        X_train_torch_i = X_train_torch[i]\n",
    "        x_one_hot_train_i = F.one_hot(X_train_torch_i-1, num_classes=NUM_WORDS).float()\n",
    "        dataset_i=Dataset(topics=x_one_hot_train_i, model_length=MODEL_LENGTH)\n",
    "        dataloader_i = DataLoader(dataset_i, batch_size=BATCH_SIZE)\n",
    "        lstm_model.train_model(dataloader_i, optimizer, criterion)\n",
    "\n",
    "        for t in range(LENGTH_TEST):\n",
    "            # Compute p(x_t | x_1:t-1) and perplexity on test sequences\n",
    "            input_sequence = X_train_torch_i\n",
    "            input_seq_one_hot = x_one_hot_train_i\n",
    "            p_next = lstm_model.predict_next_probability(input_seq_one_hot)\n",
    "            true_x_it_one_hot = F.one_hot(X_test_torch[i][t]-1, num_classes=NUM_WORDS).float()\n",
    "            cross_entropy = F.cross_entropy(p_next, true_x_it_one_hot, reduction=\"none\").item()\n",
    "            x_next = torch.multinomial(p_next, 1)+1\n",
    "            input_sequence = torch.cat([input_sequence, x_next])\n",
    "            input_seq_one_hot = F.one_hot(input_sequence-1, num_classes=NUM_WORDS).float()\n",
    "            loss_i+=cross_entropy\n",
    "        test_loss+=loss_i/LENGTH_TEST     \n",
    "    mean_test_loss = test_loss/N\n",
    "    perplexity = np.exp(mean_test_loss)\n",
    "    print(perplexity)\n",
    "    list_perplexity.append(perplexity)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 1000x600 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(10,6))\n",
    "plt.plot(range(1, N_EPOCHS + 1), list_perplexity, 'o-', color='red', label='Perplexity')\n",
    "plt.title('Perplexity curve - LSTM')\n",
    "plt.xlabel('Epochs')\n",
    "plt.ylabel('Perplexity')\n",
    "plt.legend()\n",
    "plt.grid()\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.11.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}