Tratamiento de Lenguaje Natural (PLN)/Cuadernos/nlp_Shakespeare-Original.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<figure>\n",
    "<img src=\"../Imagenes/logo-final-ap.png\"  width=\"80\" height=\"80\" align=\"left\"/> \n",
    "</figure>\n",
    "\n",
    "# <span style=\"color:blue\"><left>Aprendizaje Profundo</left></span>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# <span style=\"color:red\"><center>Construcción automática de texto </center></span>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<center>Introducción</center>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "<figure>\n",
    "<center>\n",
    "<img src=\"../Imagenes/text_generation_sampling.png\" width=\"800\" height=\"600\" align=\"center\"/>\n",
    "</center>\n",
    "<figcaption>\n",
    "<p style=\"text-align:center\">Muestreo  de palabras</p>\n",
    "</figcaption>\n",
    "</figure>\n",
    "\n",
    "Fuente: [Text generation with an RNN](https://www.tensorflow.org/tutorials/text/text_generation)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##   <span style=\"color:blue\">Autores</span>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. Alvaro Mauricio Montenegro Díaz, ammontenegrod@unal.edu.co\n",
    "2. Daniel Mauricio Montenegro Reyes, dextronomo@gmail.com "
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##   <span style=\"color:blue\">Referencias</span>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "1. Basado en el blog de laurencemoroney: (https://storage.googleapis.com/laurencemoroney-blog.appspot.com/sonnets.txt)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##   <span style=\"color:blue\">Contenido</span>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "* [Introducción](#Introducción)\n",
    "* [Datos](#Datos)\n",
    "* [Preprocesamiento](#Preprocesamiento)\n",
    "* [Librerías](#Librerías)\n",
    "* [Lee los datos](#Lee-los-datos)\n",
    "* [Tokeniza el texto](#Lee-los-datos)\n",
    "* [Crea n-Grams para datos de entrenamiento y etiquetas](#Crea-n-Grams-para-datos-de-entrenamiento-y-etiquetas)\n",
    "* [Crea-el-modelo](#Crea-el-modelo)\n",
    "* [Genera-texto-automáticamente](#Genera-texto-automáticamente)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "id": "P-AhVYeBWgQ3"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "2.4.1\n"
     ]
    }
   ],
   "source": [
    "import tensorflow as tf\n",
    "print(tf.__version__)\n",
    "\n",
    "# !pip install -q tensorflow-datasets"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##   <span style=\"color:blue\">Introducción</span>"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "Gran conjunto de datos basado en sonetos de Shekespeare.\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "##   <span style=\"color:blue\">Librerías</span>\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "id": "BOwsuGQQY9OL"
   },
   "outputs": [],
   "source": [
    "from tensorflow.keras.preprocessing.sequence import pad_sequences\n",
    "from tensorflow.keras.layers import Embedding, LSTM, Dense, Dropout, Bidirectional\n",
    "from tensorflow.keras.preprocessing.text import Tokenizer\n",
    "from tensorflow.keras.models import Sequential\n",
    "from tensorflow.keras.optimizers import Adam\n",
    "from tensorflow.keras import regularizers\n",
    "from tensorflow.keras.utils import to_categorical\n",
    "import numpy as np \n",
    "\n",
    "from tensorflow.keras.utils import plot_model"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## <span style=\"color:blue\">Lee los datos</span>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--2022-05-16 16:22:14--  https://storage.googleapis.com/laurencemoroney-blog.appspot.com/sonnets.txt\n",
      "Resolving storage.googleapis.com (storage.googleapis.com)... 142.250.78.144, 142.250.78.112, 142.250.78.80, ...\n",
      "Connecting to storage.googleapis.com (storage.googleapis.com)|142.250.78.144|:443... connected.\n",
      "HTTP request sent, awaiting response... 200 OK\n",
      "Length: 93578 (91K) [text/plain]\n",
      "Saving to: ‘/tmp/sonnets.txt’\n",
      "\n",
      "/tmp/sonnets.txt    100%[===================>]  91,38K  --.-KB/s    in 0,08s   \n",
      "\n",
      "2022-05-16 16:22:14 (1,10 MB/s) - ‘/tmp/sonnets.txt’ saved [93578/93578]\n",
      "\n"
     ]
    }
   ],
   "source": [
    "!wget --no-check-certificate \\\n",
    "    https://storage.googleapis.com/laurencemoroney-blog.appspot.com/sonnets.txt \\\n",
    "    -O /tmp/sonnets.txt\n",
    "data = open('/tmp/sonnets.txt').read()\n",
    "\n",
    "corpus = data.lower().split(\"\\n\")"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## <span style=\"color:blue\">Tokeniza el texto</span>\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "id": "PRnDnCW-Z7qv"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "index of rage =  443\n",
      "\n",
      "Total Words in the Tokenizer (+OOV) = 3211 \n",
      "\n",
      "First lines of the Corpus: \n",
      "\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "['from fairest creatures we desire increase,',\n",
       " \"that thereby beauty's rose might never die,\",\n",
       " 'but as the riper should by time decease,',\n",
       " 'his tender heir might bear his memory:',\n",
       " 'but thou, contracted to thine own bright eyes,',\n",
       " \"feed'st thy light'st flame with self-substantial fuel,\",\n",
       " 'making a famine where abundance lies,',\n",
       " 'thyself thy foe, to thy sweet self too cruel.',\n",
       " \"thou that art now the world's fresh ornament\",\n",
       " 'and only herald to the gaudy spring,']"
      ]
     },
     "execution_count": 10,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "tokenizer = Tokenizer()\n",
    "\n",
    "corpus = data.lower().split(\"\\n\")\n",
    "tokenizer.fit_on_texts(corpus)\n",
    "total_words = len(tokenizer.word_index) + 1\n",
    "\n",
    "#print('Dictionary: ',tokenizer.word_index,'\\n')\n",
    "print('índice de rage = ',tokenizer.word_index['rage'])\n",
    "print('\\nTNúmero total d palabras en el Tokenizador (+OOV) =',total_words,'\\n')\n",
    "\n",
    "# Muestra parte del corpus\n",
    "print('Primeras líneas del corpus: \\n')\n",
    "corpus[0:10]"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## <span style=\"color:blue\">Crea n-Grams para datos de entrenamiento y etiquetas</span>\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "id": "PRnDnCW-Z7qv"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "input_sequences[0:10] =  [[34, 417], [34, 417, 877], [34, 417, 877, 166], [34, 417, 877, 166, 213], [34, 417, 877, 166, 213, 517], [8, 878], [8, 878, 134], [8, 878, 134, 351], [8, 878, 134, 351, 102], [8, 878, 134, 351, 102, 156]] \n",
      "\n",
      "\n",
      "long. n-grama más largo: 11\n"
     ]
    }
   ],
   "source": [
    "# crear secuencias de entrada usando una lista de tokens\n",
    "input_sequences = []\n",
    "for line in corpus:\n",
    "    token_list = tokenizer.texts_to_sequences([line])[0]\n",
    "    for i in range(1, len(token_list)):\n",
    "        n_gram_sequence = token_list[:i+1]\n",
    "        input_sequences.append(n_gram_sequence)\n",
    "        \n",
    "print('input_sequences[0:10] = ',input_sequences[:10],'\\n')\n",
    "\n",
    "# Determine el n-gram más largo\n",
    "max_seq_test_len = max([len(x) for x in input_sequences])\n",
    "print('\\nlongitud del  n-grama más largo:',max_seq_test_len)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### ¿Qué hizo tokenizer?"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "from fairest creatures we desire increase,\n",
      "[[34, 417, 877, 166, 213, 517]]\n",
      "[34, 417, 877, 166, 213, 517]\n"
     ]
    }
   ],
   "source": [
    "line = corpus[0]\n",
    "token_list = tokenizer.texts_to_sequences([line])\n",
    "print(line)\n",
    "print(token_list)\n",
    "print(token_list[0])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Agrega PAD para tener secuencias de la misma longitud"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {
    "id": "PRnDnCW-Z7qv"
   },
   "outputs": [],
   "source": [
    "# pad sequences \n",
    "max_sequence_len = max([len(x) for x in input_sequences])\n",
    "input_sequences = np.array(pad_sequences(input_sequences, maxlen=max_sequence_len, padding='pre'))\n",
    "\n",
    "# crea predictores y etiqueta\n",
    "predictors, label = input_sequences[:,:-1],input_sequences[:,-1]\n",
    "\n",
    "#"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(15462, 11)\n",
      "(15462, 10)\n",
      "(15462,)\n"
     ]
    }
   ],
   "source": [
    "print(input_sequences.shape)\n",
    "print(predictors.shape)\n",
    "print(label.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "15462\n",
      "15462\n"
     ]
    }
   ],
   "source": [
    "print(len(predictors))\n",
    "print(len(label))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "[[  0   0   0   0   0   0   0   0   0  34 417]\n",
      " [  0   0   0   0   0   0   0   0  34 417 877]\n",
      " [  0   0   0   0   0   0   0  34 417 877 166]]\n",
      "\n",
      "\n",
      "[[  0   0   0   0   0   0   0   0   0  34]\n",
      " [  0   0   0   0   0   0   0   0  34 417]\n",
      " [  0   0   0   0   0   0   0  34 417 877]]\n",
      "\n",
      "\n",
      "[417 877 166]\n"
     ]
    }
   ],
   "source": [
    "print(input_sequences[0:3]) # secuencia completa\n",
    "print('\\n')\n",
    "print(predictors[0:3]) # secuencia hasta el anterior al último\n",
    "print('\\n')\n",
    "print(label[0:3]) # el último de la secuencia"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "### Covierte etiquetas a 1-hot"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(15462, 3211)\n"
     ]
    }
   ],
   "source": [
    "label = to_categorical(label, num_classes=total_words)\n",
    "print(label.shape)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## <span style=\"color:blue\">Crea el modelo</span>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "id": "w9vH8Y59ajYL"
   },
   "outputs": [],
   "source": [
    "model = Sequential()\n",
    "model.add(Embedding(total_words, 100, input_length=max_sequence_len-1))\n",
    "model.add(Bidirectional(LSTM(150, return_sequences = True)))\n",
    "model.add(Dropout(0.2))\n",
    "model.add(LSTM(100))\n",
    "model.add(Dense(total_words/2, activation='relu', kernel_regularizer=regularizers.l2(0.01)))\n",
    "model.add(Dense(total_words, activation='softmax'))\n",
    "model.compile(loss='categorical_crossentropy', optimizer='adam', metrics=['accuracy'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {
    "id": "w9vH8Y59ajYL"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Model: \"sequential\"\n",
      "_________________________________________________________________\n",
      "Layer (type)                 Output Shape              Param #   \n",
      "=================================================================\n",
      "embedding (Embedding)        (None, 10, 100)           321100    \n",
      "_________________________________________________________________\n",
      "bidirectional (Bidirectional (None, 10, 300)           301200    \n",
      "_________________________________________________________________\n",
      "dropout (Dropout)            (None, 10, 300)           0         \n",
      "_________________________________________________________________\n",
      "lstm_1 (LSTM)                (None, 100)               160400    \n",
      "_________________________________________________________________\n",
      "dense (Dense)                (None, 1605)              162105    \n",
      "_________________________________________________________________\n",
      "dense_1 (Dense)              (None, 3211)              5156866   \n",
      "=================================================================\n",
      "Total params: 6,101,671\n",
      "Trainable params: 6,101,671\n",
      "Non-trainable params: 0\n",
      "_________________________________________________________________\n",
      "None\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<IPython.core.display.Image object>"
      ]
     },
     "execution_count": 9,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "print(model.summary())\n",
    "plot_model(model, to_file='../Imagenes/sonetos.png', \n",
    "           show_shapes=True)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## <span style=\"color:blue\">Entrena</span>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {
    "id": "AIg2f1HBxqof"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Train on 15462 samples\n",
      "Epoch 1/100\n",
      "15462/15462 [==============================] - 124s 8ms/sample - loss: 6.9038 - accuracy: 0.0216\n",
      "Epoch 2/100\n",
      "15462/15462 [==============================] - 113s 7ms/sample - loss: 6.5005 - accuracy: 0.0227\n",
      "Epoch 3/100\n",
      "15462/15462 [==============================] - 104s 7ms/sample - loss: 6.4022 - accuracy: 0.0240\n",
      "Epoch 4/100\n",
      "15462/15462 [==============================] - 130s 8ms/sample - loss: 6.2774 - accuracy: 0.0270\n",
      "Epoch 5/100\n",
      "15462/15462 [==============================] - 126s 8ms/sample - loss: 6.1789 - accuracy: 0.0349\n",
      "Epoch 6/100\n",
      "15462/15462 [==============================] - 89s 6ms/sample - loss: 6.0936 - accuracy: 0.0380\n",
      "Epoch 7/100\n",
      "15462/15462 [==============================] - 91s 6ms/sample - loss: 6.0174 - accuracy: 0.0416\n",
      "Epoch 8/100\n",
      "15462/15462 [==============================] - 90s 6ms/sample - loss: 5.9288 - accuracy: 0.0453\n",
      "Epoch 9/100\n",
      "15462/15462 [==============================] - 90s 6ms/sample - loss: 5.8338 - accuracy: 0.0519\n",
      "Epoch 10/100\n",
      "15462/15462 [==============================] - 91s 6ms/sample - loss: 5.7278 - accuracy: 0.0586\n",
      "Epoch 11/100\n",
      "15462/15462 [==============================] - 90s 6ms/sample - loss: 5.6250 - accuracy: 0.0624\n",
      "Epoch 12/100\n",
      "15462/15462 [==============================] - 90s 6ms/sample - loss: 5.5199 - accuracy: 0.0696\n",
      "Epoch 13/100\n",
      "15462/15462 [==============================] - 90s 6ms/sample - loss: 5.4180 - accuracy: 0.0770\n",
      "Epoch 14/100\n",
      "15462/15462 [==============================] - 90s 6ms/sample - loss: 5.3231 - accuracy: 0.0812\n",
      "Epoch 15/100\n",
      "15462/15462 [==============================] - 90s 6ms/sample - loss: 5.2243 - accuracy: 0.0882\n",
      "Epoch 16/100\n",
      "15462/15462 [==============================] - 91s 6ms/sample - loss: 5.1296 - accuracy: 0.0944\n",
      "Epoch 17/100\n",
      "15462/15462 [==============================] - 89s 6ms/sample - loss: 5.0357 - accuracy: 0.1003\n",
      "Epoch 18/100\n",
      "15462/15462 [==============================] - 83s 5ms/sample - loss: 4.9465 - accuracy: 0.1104\n",
      "Epoch 19/100\n",
      "15462/15462 [==============================] - 83s 5ms/sample - loss: 4.8444 - accuracy: 0.1178\n",
      "Epoch 20/100\n",
      "15462/15462 [==============================] - 83s 5ms/sample - loss: 4.7512 - accuracy: 0.1236\n",
      "Epoch 21/100\n",
      "15462/15462 [==============================] - 83s 5ms/sample - loss: 4.6581 - accuracy: 0.1330\n",
      "Epoch 22/100\n",
      "15462/15462 [==============================] - 83s 5ms/sample - loss: 4.5687 - accuracy: 0.1409\n",
      "Epoch 23/100\n",
      "15462/15462 [==============================] - 83s 5ms/sample - loss: 4.4696 - accuracy: 0.1528\n",
      "Epoch 24/100\n",
      "15462/15462 [==============================] - 83s 5ms/sample - loss: 4.3714 - accuracy: 0.1665\n",
      "Epoch 25/100\n",
      "15462/15462 [==============================] - 83s 5ms/sample - loss: 4.2818 - accuracy: 0.1766\n",
      "Epoch 26/100\n",
      "15462/15462 [==============================] - 83s 5ms/sample - loss: 4.1756 - accuracy: 0.1907\n",
      "Epoch 27/100\n",
      "15462/15462 [==============================] - 83s 5ms/sample - loss: 4.0885 - accuracy: 0.2006\n",
      "Epoch 28/100\n",
      "15462/15462 [==============================] - 83s 5ms/sample - loss: 3.9861 - accuracy: 0.2143\n",
      "Epoch 29/100\n",
      "15462/15462 [==============================] - 83s 5ms/sample - loss: 3.8985 - accuracy: 0.2286\n",
      "Epoch 30/100\n",
      "15462/15462 [==============================] - 83s 5ms/sample - loss: 3.8006 - accuracy: 0.2441\n",
      "Epoch 31/100\n",
      "15462/15462 [==============================] - 83s 5ms/sample - loss: 3.7157 - accuracy: 0.2591\n",
      "Epoch 32/100\n",
      "15462/15462 [==============================] - 87s 6ms/sample - loss: 3.6145 - accuracy: 0.2742\n",
      "Epoch 33/100\n",
      "15462/15462 [==============================] - 78s 5ms/sample - loss: 3.5389 - accuracy: 0.2943\n",
      "Epoch 34/100\n",
      "15462/15462 [==============================] - 79s 5ms/sample - loss: 3.4463 - accuracy: 0.3128\n",
      "Epoch 35/100\n",
      "15462/15462 [==============================] - 73s 5ms/sample - loss: 3.3642 - accuracy: 0.3313\n",
      "Epoch 36/100\n",
      "15462/15462 [==============================] - 73s 5ms/sample - loss: 3.2903 - accuracy: 0.3464\n",
      "Epoch 37/100\n",
      "15462/15462 [==============================] - 72s 5ms/sample - loss: 3.2163 - accuracy: 0.3610\n",
      "Epoch 38/100\n",
      "15462/15462 [==============================] - 71s 5ms/sample - loss: 3.1351 - accuracy: 0.3817\n",
      "Epoch 39/100\n",
      "15462/15462 [==============================] - 71s 5ms/sample - loss: 3.0525 - accuracy: 0.4012\n",
      "Epoch 40/100\n",
      "15462/15462 [==============================] - 72s 5ms/sample - loss: 2.9784 - accuracy: 0.4148\n",
      "Epoch 41/100\n",
      "15462/15462 [==============================] - 78s 5ms/sample - loss: 2.9153 - accuracy: 0.4321\n",
      "Epoch 42/100\n",
      "15462/15462 [==============================] - 73s 5ms/sample - loss: 2.8443 - accuracy: 0.4483\n",
      "Epoch 43/100\n",
      "15462/15462 [==============================] - 75s 5ms/sample - loss: 2.7891 - accuracy: 0.4567\n",
      "Epoch 44/100\n",
      "15462/15462 [==============================] - 80s 5ms/sample - loss: 2.7309 - accuracy: 0.4733\n",
      "Epoch 45/100\n",
      "15462/15462 [==============================] - 72s 5ms/sample - loss: 2.6593 - accuracy: 0.4876\n",
      "Epoch 46/100\n",
      "15462/15462 [==============================] - 72s 5ms/sample - loss: 2.6006 - accuracy: 0.5024\n",
      "Epoch 47/100\n",
      "15462/15462 [==============================] - 75s 5ms/sample - loss: 2.5596 - accuracy: 0.5072\n",
      "Epoch 48/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 2.5005 - accuracy: 0.5247\n",
      "Epoch 49/100\n",
      "15462/15462 [==============================] - 72s 5ms/sample - loss: 2.4416 - accuracy: 0.5376\n",
      "Epoch 50/100\n",
      "15462/15462 [==============================] - 73s 5ms/sample - loss: 2.3894 - accuracy: 0.5477\n",
      "Epoch 51/100\n",
      "15462/15462 [==============================] - 79s 5ms/sample - loss: 2.3365 - accuracy: 0.5613\n",
      "Epoch 52/100\n",
      "15462/15462 [==============================] - 82s 5ms/sample - loss: 2.2947 - accuracy: 0.5695\n",
      "Epoch 53/100\n",
      "15462/15462 [==============================] - 80s 5ms/sample - loss: 2.2460 - accuracy: 0.5846\n",
      "Epoch 54/100\n",
      "15462/15462 [==============================] - 79s 5ms/sample - loss: 2.2029 - accuracy: 0.5894\n",
      "Epoch 55/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 2.1632 - accuracy: 0.6045\n",
      "Epoch 56/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 2.1208 - accuracy: 0.6071\n",
      "Epoch 57/100\n",
      "15462/15462 [==============================] - 75s 5ms/sample - loss: 2.0766 - accuracy: 0.6209\n",
      "Epoch 58/100\n",
      "15462/15462 [==============================] - 75s 5ms/sample - loss: 2.0474 - accuracy: 0.6249\n",
      "Epoch 59/100\n",
      "15462/15462 [==============================] - 75s 5ms/sample - loss: 2.0070 - accuracy: 0.6370\n",
      "Epoch 60/100\n",
      "15462/15462 [==============================] - 88s 6ms/sample - loss: 1.9635 - accuracy: 0.6484\n",
      "Epoch 61/100\n",
      "15462/15462 [==============================] - 78s 5ms/sample - loss: 1.9387 - accuracy: 0.6510\n",
      "Epoch 62/100\n",
      "15462/15462 [==============================] - 78s 5ms/sample - loss: 1.9077 - accuracy: 0.6552\n",
      "Epoch 63/100\n",
      "15462/15462 [==============================] - 74s 5ms/sample - loss: 1.8570 - accuracy: 0.6684\n",
      "Epoch 64/100\n",
      "15462/15462 [==============================] - 74s 5ms/sample - loss: 1.8286 - accuracy: 0.6716\n",
      "Epoch 65/100\n",
      "15462/15462 [==============================] - 74s 5ms/sample - loss: 1.8109 - accuracy: 0.6757\n",
      "Epoch 66/100\n",
      "15462/15462 [==============================] - 81s 5ms/sample - loss: 1.7666 - accuracy: 0.6857\n",
      "Epoch 67/100\n",
      "15462/15462 [==============================] - 79s 5ms/sample - loss: 1.7576 - accuracy: 0.6891\n",
      "Epoch 68/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 1.7132 - accuracy: 0.6981\n",
      "Epoch 69/100\n",
      "15462/15462 [==============================] - 75s 5ms/sample - loss: 1.6938 - accuracy: 0.7019\n",
      "Epoch 70/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 1.6917 - accuracy: 0.7002\n",
      "Epoch 71/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 1.6407 - accuracy: 0.7139\n",
      "Epoch 72/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 1.6013 - accuracy: 0.7255\n",
      "Epoch 73/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 1.5896 - accuracy: 0.7219\n",
      "Epoch 74/100\n",
      "15462/15462 [==============================] - 75s 5ms/sample - loss: 1.5665 - accuracy: 0.7262\n",
      "Epoch 75/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 1.5539 - accuracy: 0.7311\n",
      "Epoch 76/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 1.5267 - accuracy: 0.7395\n",
      "Epoch 77/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 1.4955 - accuracy: 0.7423\n",
      "Epoch 78/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 1.4860 - accuracy: 0.7445\n",
      "Epoch 79/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 1.4752 - accuracy: 0.7475\n",
      "Epoch 80/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 1.4457 - accuracy: 0.7528\n",
      "Epoch 81/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 1.4415 - accuracy: 0.7493\n",
      "Epoch 82/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 1.4143 - accuracy: 0.7614\n",
      "Epoch 83/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 1.3932 - accuracy: 0.7652\n",
      "Epoch 84/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 1.3788 - accuracy: 0.7636\n",
      "Epoch 85/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 1.3747 - accuracy: 0.7641\n",
      "Epoch 86/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 1.3530 - accuracy: 0.7689\n",
      "Epoch 87/100\n",
      "15462/15462 [==============================] - 75s 5ms/sample - loss: 1.3217 - accuracy: 0.7740\n",
      "Epoch 88/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 1.3201 - accuracy: 0.7736\n",
      "Epoch 89/100\n",
      "15462/15462 [==============================] - 75s 5ms/sample - loss: 1.3068 - accuracy: 0.7748\n",
      "Epoch 90/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 1.2978 - accuracy: 0.7760\n",
      "Epoch 91/100\n",
      "15462/15462 [==============================] - 75s 5ms/sample - loss: 1.2877 - accuracy: 0.7757\n",
      "Epoch 92/100\n",
      "15462/15462 [==============================] - 75s 5ms/sample - loss: 1.2707 - accuracy: 0.7809\n",
      "Epoch 93/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 1.2564 - accuracy: 0.7842\n",
      "Epoch 94/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 1.2484 - accuracy: 0.7831\n",
      "Epoch 95/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 1.2275 - accuracy: 0.7870\n",
      "Epoch 96/100\n",
      "15462/15462 [==============================] - 75s 5ms/sample - loss: 1.2120 - accuracy: 0.7913\n",
      "Epoch 97/100\n",
      "15462/15462 [==============================] - 75s 5ms/sample - loss: 1.2155 - accuracy: 0.7881\n",
      "Epoch 98/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 1.1953 - accuracy: 0.7950\n",
      "Epoch 99/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 1.1939 - accuracy: 0.7934\n",
      "Epoch 100/100\n",
      "15462/15462 [==============================] - 76s 5ms/sample - loss: 1.1778 - accuracy: 0.7976\n"
     ]
    }
   ],
   "source": [
    " history = model.fit(predictors, label, epochs=100, verbose=1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {
    "id": "1fXTEO3GJ282"
   },
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "import matplotlib.pyplot as plt\n",
    "acc = history.history['accuracy']\n",
    "loss = history.history['loss']\n",
    "\n",
    "epochs = range(len(acc))\n",
    "\n",
    "plt.plot(epochs, acc, 'b', label='Exactitud de entrenamiento')\n",
    "plt.title('Exactitud de entrenamiento')\n",
    "\n",
    "plt.figure()\n",
    "\n",
    "plt.plot(epochs, loss, 'b', label='Pérdida en entrenamiento')\n",
    "plt.title('Pérdida en entrenamiento')\n",
    "plt.legend()\n",
    "\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## <span style=\"color:blue\">Genera texto automáticamente</span>"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "id": "6Vc6PHgxa6Hm"
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Who are you, so too cruel? in you have free tend show thee me heaven dead deem'd lie on her long alone must not must strive to write self hue most scope still true shown pride words behold old hate old part my view you are pride prove me 'no new prove be worth and thee doth tender'd made write you bright 'no new part me young me old write for me i love not thought thy sight must new must find torn clears dyed new breast lack where how to thee now swearing write old love me prove me write old for thee in thee\n"
     ]
    }
   ],
   "source": [
    "seed_text = \"Who are you, so too cruel?\"\n",
    "next_words = 100\n",
    "  \n",
    "for _ in range(next_words):\n",
    "    token_list = tokenizer.texts_to_sequences([seed_text])[0]\n",
    "    token_list = pad_sequences([token_list], maxlen=max_sequence_len-1, padding='pre')\n",
    "    predicted = model.predict_classes(token_list, verbose=0)\n",
    "    output_word = \"\"\n",
    "    for word, index in tokenizer.word_index.items():\n",
    "        if index == predicted:\n",
    "            output_word = word\n",
    "            break\n",
    "    seed_text += \" \" + output_word\n",
    "print(seed_text)"
   ]
  },
  {
   "cell_type": "raw",
   "metadata": {},
   "source": [
    "¿Quién eres, tan cruel? en ti tienes libre, ten cuidado, muéstrame el cielo muerto, consideró que yacía en ella mucho tiempo solo, no debes esforzarte por escribir el tono más amplio, todavía verdadero, palabras de orgullo, he aquí, viejo odio, parte vieja, mi punto de vista, eres orgullo, demuéstrame 'no hay nada nuevo. vale la pena y tu ternura te hizo escribirte brillante 'ninguna parte nueva yo joven yo viejo escribe para mí amo no pensé que tu vista debe ser nueva debe encontrar rasgados claros teñidos nuevos pechos falta dónde cómo estás ahora jurando escribe viejo amor demuéstrame escribe antiguo para ti en ti"
   ]
  }
 ],
 "metadata": {
  "accelerator": "GPU",
  "colab": {
   "name": "NLP_Week4_Exercise_Shakespeare_Answer.ipynb",
   "provenance": [],
   "toc_visible": true
  },
  "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": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}