generator/1.char-generator-lstm.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy as np\n",
    "import tensorflow as tf\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "from tqdm import tqdm\n",
    "import random\n",
    "import time\n",
    "sns.set()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "def get_vocab(file, lower = False):\n",
    "    with open(file, 'r') as fopen:\n",
    "        data = fopen.read()\n",
    "    if lower:\n",
    "        data = data.lower()\n",
    "    vocab = list(set(data))\n",
    "    return data, vocab\n",
    "\n",
    "def embed_to_onehot(data, vocab):\n",
    "    onehot = np.zeros((len(data), len(vocab)), dtype = np.float32)\n",
    "    for i in range(len(data)):\n",
    "        onehot[i, vocab.index(data[i])] = 1.0\n",
    "    return onehot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "text, text_vocab = get_vocab('shakespeare.txt', lower = False)\n",
    "onehot = embed_to_onehot(text, text_vocab)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [],
   "source": [
    "learning_rate = 0.01\n",
    "batch_size = 128\n",
    "sequence_length = 64\n",
    "epoch = 3000\n",
    "num_layers = 2\n",
    "size_layer = 512\n",
    "possible_batch_id = range(len(text) - sequence_length - 1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [],
   "source": [
    "class Model:\n",
    "    def __init__(\n",
    "        self, num_layers, size_layer, dimension, sequence_length, learning_rate\n",
    "    ):\n",
    "        def lstm_cell():\n",
    "            return tf.nn.rnn_cell.LSTMCell(\n",
    "                size_layer, sequence_length, state_is_tuple = False\n",
    "            )\n",
    "\n",
    "        self.rnn_cells = tf.nn.rnn_cell.MultiRNNCell(\n",
    "            [lstm_cell() for _ in range(num_layers)], state_is_tuple = False\n",
    "        )\n",
    "        self.X = tf.placeholder(tf.float32, (None, None, dimension))\n",
    "        self.Y = tf.placeholder(tf.float32, (None, None, dimension))\n",
    "        self.hidden_layer = tf.placeholder(\n",
    "            tf.float32, (None, num_layers * 2 * size_layer)\n",
    "        )\n",
    "        self.outputs, self.last_state = tf.nn.dynamic_rnn(\n",
    "            self.rnn_cells,\n",
    "            self.X,\n",
    "            initial_state = self.hidden_layer,\n",
    "            dtype = tf.float32,\n",
    "        )\n",
    "        rnn_W = tf.Variable(tf.random_normal((size_layer, dimension)))\n",
    "        rnn_B = tf.Variable(tf.random_normal([dimension]))\n",
    "        self.logits = (\n",
    "            tf.matmul(tf.reshape(self.outputs, [-1, size_layer]), rnn_W) + rnn_B\n",
    "        )\n",
    "        y_batch_long = tf.reshape(self.Y, [-1, dimension])\n",
    "        self.cost = tf.reduce_mean(\n",
    "            tf.nn.softmax_cross_entropy_with_logits(\n",
    "                logits = self.logits, labels = y_batch_long\n",
    "            )\n",
    "        )\n",
    "        self.optimizer = tf.train.RMSPropOptimizer(learning_rate, 0.9).minimize(\n",
    "            self.cost\n",
    "        )\n",
    "        self.correct_pred = tf.equal(\n",
    "            tf.argmax(self.logits, 1), tf.argmax(y_batch_long, 1)\n",
    "        )\n",
    "        self.accuracy = tf.reduce_mean(tf.cast(self.correct_pred, tf.float32))\n",
    "        seq_shape = tf.shape(self.outputs)\n",
    "        self.final_outputs = tf.reshape(\n",
    "            tf.nn.softmax(self.logits), (seq_shape[0], seq_shape[1], dimension)\n",
    "        )\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "WARNING:tensorflow:<tensorflow.python.ops.rnn_cell_impl.LSTMCell object at 0x7f77a4760a58>: Using a concatenated state is slower and will soon be deprecated.  Use state_is_tuple=True.\n",
      "WARNING:tensorflow:<tensorflow.python.ops.rnn_cell_impl.LSTMCell object at 0x7f77a4760a90>: Using a concatenated state is slower and will soon be deprecated.  Use state_is_tuple=True.\n",
      "WARNING:tensorflow:From <ipython-input-5-ab798ebdc858>:32: softmax_cross_entropy_with_logits (from tensorflow.python.ops.nn_ops) is deprecated and will be removed in a future version.\n",
      "Instructions for updating:\n",
      "\n",
      "Future major versions of TensorFlow will allow gradients to flow\n",
      "into the labels input on backprop by default.\n",
      "\n",
      "See `tf.nn.softmax_cross_entropy_with_logits_v2`.\n",
      "\n"
     ]
    }
   ],
   "source": [
    "tf.reset_default_graph()\n",
    "sess = tf.InteractiveSession()\n",
    "model = Model(num_layers, size_layer, len(text_vocab), sequence_length, learning_rate)\n",
    "sess.run(tf.global_variables_initializer())"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "that\n"
     ]
    }
   ],
   "source": [
    "split_text = text.split()\n",
    "tag = split_text[np.random.randint(0, len(split_text))]\n",
    "print(tag)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [],
   "source": [
    "def train_random_sequence():\n",
    "    LOST, ACCURACY = [], []\n",
    "    pbar = tqdm(range(epoch), desc = 'epoch')\n",
    "    for i in pbar:\n",
    "        last_time = time.time()\n",
    "        init_value = np.zeros((batch_size, num_layers * 2 * size_layer))\n",
    "        batch_x = np.zeros((batch_size, sequence_length, len(text_vocab)))\n",
    "        batch_y = np.zeros((batch_size, sequence_length, len(text_vocab)))\n",
    "        batch_id = random.sample(possible_batch_id, batch_size)\n",
    "        for n in range(sequence_length):\n",
    "            id1 = [k + n for k in batch_id]\n",
    "            id2 = [k + n + 1 for k in batch_id]\n",
    "            batch_x[:,n,:] = onehot[id1, :]\n",
    "            batch_y[:,n,:] = onehot[id2, :]\n",
    "        last_state, _, loss = sess.run([model.last_state, model.optimizer, model.cost], \n",
    "                                       feed_dict = {model.X: batch_x, \n",
    "                                                    model.Y: batch_y,\n",
    "                                                    model.hidden_layer: init_value})\n",
    "        accuracy = sess.run(model.accuracy, feed_dict = {model.X: batch_x, \n",
    "                                                         model.Y: batch_y, \n",
    "                                                         model.hidden_layer: init_value})\n",
    "        ACCURACY.append(accuracy); LOST.append(loss)\n",
    "        init_value = last_state\n",
    "        pbar.set_postfix(cost = loss, accuracy = accuracy)\n",
    "    return LOST, ACCURACY"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "epoch: 100%|██████████| 3000/3000 [13:49<00:00,  3.61it/s, accuracy=0.887, cost=0.83] \n"
     ]
    }
   ],
   "source": [
    "LOST, ACCURACY = train_random_sequence()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 1080x360 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize = (15, 5))\n",
    "plt.subplot(1, 2, 1)\n",
    "EPOCH = np.arange(len(LOST))\n",
    "plt.plot(EPOCH, LOST)\n",
    "plt.xlabel('epoch'); plt.ylabel('loss')\n",
    "plt.subplot(1, 2, 2)\n",
    "plt.plot(EPOCH, ACCURACY)\n",
    "plt.xlabel('epoch'); plt.ylabel('accuracy')\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "def generate_based_sequence(length_sentence, argmax = False):\n",
    "    sentence_generated = tag\n",
    "    onehot = embed_to_onehot(tag, text_vocab)\n",
    "    init_value = np.zeros((1, num_layers * 2 * size_layer))\n",
    "    for i in range(len(tag)):\n",
    "        batch_x = np.zeros((1, 1, len(text_vocab)))\n",
    "        batch_x[:, 0, :] = onehot[i, :]\n",
    "        last_state, prob = sess.run(\n",
    "            [model.last_state, model.final_outputs],\n",
    "            feed_dict = {model.X: batch_x, model.hidden_layer: init_value},\n",
    "        )\n",
    "        init_value = last_state\n",
    "\n",
    "    for i in range(length_sentence):\n",
    "        if argmax:\n",
    "            char = np.argmax(prob[0][0])\n",
    "        else:\n",
    "            char = np.random.choice(range(len(text_vocab)), p = prob[0][0])\n",
    "        element = text_vocab[char]\n",
    "        sentence_generated += element\n",
    "        onehot = embed_to_onehot(element, text_vocab)\n",
    "        batch_x = np.zeros((1, 1, len(text_vocab)))\n",
    "        batch_x[:, 0, :] = onehot[0, :]\n",
    "        last_state, prob = sess.run(\n",
    "            [model.last_state, model.final_outputs],\n",
    "            feed_dict = {model.X: batch_x, model.hidden_layer: init_value},\n",
    "        )\n",
    "        init_value = last_state\n",
    "\n",
    "    return sentence_generated"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "that the world goes well, who rather\n",
      "had been so brief with him, and all of slanderous stoutness\n",
      "He shall not be so bold to be of morrow.\n",
      "\n",
      "KING EDWARD IV:\n",
      "What says his majesty?\n",
      "\n",
      "NORTHUMBERLAND:\n",
      "Here come the children of the blood of my sweet boy\n",
      "Thy brother wrought by the boldness of my heart,\n",
      "That thou hast made me by my brother's son,\n",
      "The son of England's face, my dear lord's death!\n",
      "Lo, ere I can repeat this cursed pale.\n",
      "\n",
      "QUEEN MARGARET:\n",
      "Why, then, the whate that I should be proclaim'd\n",
      "In exchange than the rest of the complexion of the king,\n",
      "And be in barded steps your promise proud to hear\n",
      "My brother wrought by the fire\n",
      "Of both of your highness complaining.\n",
      "\n",
      "Servant:\n",
      "My gracious lord, I think he cannot choose.\n",
      "\n",
      "Second Murderer:\n",
      "Who made thee fair. But if he thus\n",
      "death to use me for the best of it.\n",
      "\n",
      "KATHARINA:\n",
      "Why, sir, you must part your form.\n",
      "\n",
      "MIRANDA:\n",
      "But the removed friend,\n",
      "My tears shall share with shame to be my son,\n",
      "The sons of Edward will be sure of you.\n",
      "\n",
      "GLOUCESTER:\n",
      "I go. What \n"
     ]
    }
   ],
   "source": [
    "print(generate_based_sequence(1000,True))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "that a big and mercy enemies.\n",
      "My sons shall never brook pronounce in his\n",
      "counstles far their loss, a boon that slew thyself?\n",
      "\n",
      "First Citizen:\n",
      "And but to strike, madam, in his sple ncess,\n",
      "His grave like mine office in the prison, love me,\n",
      "My bed with your pleasure.\n",
      "\n",
      "AUFIDIUS:\n",
      "Call me too, yet your grace\n",
      "For fair without not: but, I'll oft beautiful\n",
      "Wife is gone, and well see--was't that is found dagger.\n",
      "Gardenerer, thou hadst been said: 'I thank myself\n",
      "Fool to his councell'd spirit in peace be made,\n",
      "Some of those weakness are his children\n",
      "dress'd his beauty had been silent. Please you to make thy hell,\n",
      "In breast that had their soldiers usurp.\n",
      "And brief, so heartily he be safet,\n",
      "With something dissension by the swording friend,\n",
      "Which butts my stay to his foel wanders to all noble father?\n",
      "Pay to thy sweet son: since the contents remember\n",
      ": yet I had fought to breathe,--play edity\n",
      "but little a army piece of earth.\n",
      "\n",
      "DUKE OF YORK:\n",
      "My counsel, let us shame his back in pray\n",
      "To stale more precious t\n"
     ]
    }
   ],
   "source": [
    "print(generate_based_sequence(1000,False))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "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": "3.6.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}