Generating_Poetic_Texts_with_Recurrent_Neural_Networks_in_Python.ipynb

{
  "nbformat": 4,
  "nbformat_minor": 0,
  "metadata": {
    "colab": {
      "provenance": [],
      "gpuType": "T4"
    },
    "kernelspec": {
      "name": "python3",
      "display_name": "Python 3"
    },
    "language_info": {
      "name": "python"
    },
    "accelerator": "GPU"
  },
  "cells": [
    {
      "cell_type": "code",
      "source": [
        "import random\n",
        "import numpy as np\n",
        "import tensorflow as tf\n",
        "from tensorflow.keras.models import Sequential\n",
        "from tensorflow.keras.layers import LSTM, Dense, Activation, Input\n",
        "from tensorflow.keras.optimizers import RMSprop\n",
        "\n",
        "# Download and load Shakespeare text\n",
        "filepath = tf.keras.utils.get_file(\n",
        "    'shakespeare.txt',\n",
        "    'https://storage.googleapis.com/download.tensorflow.org/data/shakespeare.txt'\n",
        ")\n",
        "\n",
        "# Read and preprocess the text\n",
        "with open(filepath, 'rb') as file:\n",
        "    text = file.read().decode(encoding='utf-8').lower()\n",
        "\n",
        "# Use a subset of the text for training\n",
        "text = text[300000:800000]\n",
        "\n",
        "# Create mappings from characters to indices and vice versa\n",
        "characters = sorted(set(text))\n",
        "char_to_index = {c: i for i, c in enumerate(characters)}\n",
        "index_to_char = {i: c for i, c in enumerate(characters)}\n",
        "\n",
        "# Define sequence parameters\n",
        "SEQ_LENGTH = 40\n",
        "STEP_SIZE = 3\n",
        "\n",
        "# Prepare input-output sequences\n",
        "sentences = []\n",
        "next_characters = []\n",
        "\n",
        "for i in range(0, len(text) - SEQ_LENGTH, STEP_SIZE):\n",
        "    sentences.append(text[i: i + SEQ_LENGTH])\n",
        "    next_characters.append(text[i + SEQ_LENGTH])\n",
        "\n",
        "# Initialize input (x) and output (y) arrays\n",
        "x = np.zeros((len(sentences), SEQ_LENGTH, len(characters)), dtype=bool)\n",
        "y = np.zeros((len(sentences), len(characters)), dtype=bool)\n",
        "\n",
        "# One-hot encode the input and output\n",
        "for i, sentence in enumerate(sentences):\n",
        "    for t, character in enumerate(sentence):\n",
        "        x[i, t, char_to_index[character]] = 1\n",
        "    y[i, char_to_index[next_characters[i]]] = 1\n",
        "\n",
        "# Build the model\n",
        "model = Sequential()\n",
        "model.add(Input(shape=(SEQ_LENGTH, len(characters))))  # Explicit input shape\n",
        "model.add(LSTM(128))  # LSTM layer for sequence processing\n",
        "model.add(Dense(len(characters)))  # Dense output layer\n",
        "model.add(Activation('softmax'))  # Softmax for categorical output\n",
        "\n",
        "# Compile the model\n",
        "model.compile(\n",
        "    loss='categorical_crossentropy',\n",
        "    optimizer=RMSprop(learning_rate=0.01)\n",
        ")\n",
        "\n",
        "# Train the model\n",
        "model.fit(x, y, batch_size=256, epochs=4)\n",
        "\n",
        "# Save the model in the recommended Keras format\n",
        "model.save('textgenerator.keras')\n",
        "\n",
        "print(\"Model trained and saved successfully.\")"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "lclm35N-sq_w",
        "outputId": "036b7186-0cd3-4fff-a574-7ef4a5ac6103"
      },
      "execution_count": 2,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "Downloading data from https://storage.googleapis.com/download.tensorflow.org/data/shakespeare.txt\n",
            "\u001b[1m1115394/1115394\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m0s\u001b[0m 0us/step\n",
            "Epoch 1/4\n",
            "\u001b[1m651/651\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m5s\u001b[0m 6ms/step - loss: 2.4796\n",
            "Epoch 2/4\n",
            "\u001b[1m651/651\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 5ms/step - loss: 1.7651\n",
            "Epoch 3/4\n",
            "\u001b[1m651/651\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m4s\u001b[0m 6ms/step - loss: 1.6073\n",
            "Epoch 4/4\n",
            "\u001b[1m651/651\u001b[0m \u001b[32m━━━━━━━━━━━━━━━━━━━━\u001b[0m\u001b[37m\u001b[0m \u001b[1m3s\u001b[0m 5ms/step - loss: 1.5188\n",
            "Model trained and saved successfully.\n"
          ]
        }
      ]
    },
    {
      "cell_type": "code",
      "source": [
        "def sample(preds, temperature=1.0):\n",
        "    preds = np.asarray(preds).astype('float64')\n",
        "    preds = np.log(preds) / temperature\n",
        "    exp_preds = np.exp(preds)\n",
        "    preds = exp_preds / np.sum(exp_preds)\n",
        "    probas = np.random.multinomial(1, preds, 1)\n",
        "    return np.argmax(probas)\n",
        "\n",
        "def generate_text(length, temperature):\n",
        "    start_index = random.randint(0, len(text) - SEQ_LENGTH - 1)  # Fixed typo\n",
        "    generated = ''\n",
        "    sentence = text[start_index: start_index + SEQ_LENGTH]\n",
        "    generated += sentence\n",
        "    for i in range(length):\n",
        "        x = np.zeros((1, SEQ_LENGTH, len(characters)))\n",
        "        for t, character in enumerate(sentence):\n",
        "            x[0, t, char_to_index[character]] = 1\n",
        "\n",
        "        predictions = model.predict(x, verbose=0)[0]\n",
        "        next_index = sample(predictions, temperature)\n",
        "        next_character = index_to_char[next_index]\n",
        "\n",
        "        generated += next_character\n",
        "        sentence = sentence[1:] + next_character\n",
        "    return generated\n",
        "\n",
        "# Generate text with different temperature values\n",
        "print('________0.2_________')\n",
        "print(generate_text(300, 0.2))\n",
        "print('________0.4_________')\n",
        "print(generate_text(300, 0.4))\n",
        "print('________0.6_________')\n",
        "print(generate_text(300, 0.6))\n",
        "print('________0.8_________')\n",
        "print(generate_text(300, 0.8))\n",
        "print('________1.0_________')\n",
        "print(generate_text(300, 1.0))"
      ],
      "metadata": {
        "colab": {
          "base_uri": "https://localhost:8080/"
        },
        "id": "SVgCIIkLsu10",
        "outputId": "0e62f351-feb1-4cd9-b2e4-695a675c2f9e"
      },
      "execution_count": 4,
      "outputs": [
        {
          "output_type": "stream",
          "name": "stdout",
          "text": [
            "________0.2_________\n",
            "lord warwick, but my son,\n",
            "whom i unnatured the soul and make me the soul,\n",
            "and the soul and the soner's days the sears,\n",
            "and the hands the soul soul some the seat,\n",
            "and so the soul and the soul and to the signing the brother,\n",
            "the soul some seen and the hands the first,\n",
            "and that the heart the love the soul soul, and the hands the soul soul,\n",
            "a\n",
            "________0.4_________\n",
            "elay this marriage for a month, a week;\n",
            "i do the soul and that speak the servented\n",
            "whose the wass in the brides of the servant.\n",
            "\n",
            "king richard ii:\n",
            "thou wears the soul happy set the forther warwick,\n",
            "how now that here the bray the seat the broof and the traitors,\n",
            "and no have the king richard and him that stander'd with the soul,\n",
            "and for a se\n",
            "________0.6_________\n",
            "ashy too;\n",
            "but time will not permit: all that the sinely;\n",
            "how it impood hath no brother, but i have see:\n",
            "now with her not, but their house,\n",
            "but love the world him but was a seat youth and.\n",
            "\n",
            "britoous:\n",
            "past the sent she have baneness of hands.\n",
            "thy hould to me as that here the soners of marria thee.\n",
            "\n",
            "juliet:\n",
            "to brothers of brather heaven, why\n",
            "________0.8_________\n",
            " hither;\n",
            "you that have been so tenderly swayner,\n",
            "shall stay, that before on the house, and that brow' from mands\n",
            "and bid a kingering it nam or right, of marcariay:\n",
            "i she but hersele hath forth, too only farewell,\n",
            "and prayer the streathen die.\n",
            "\n",
            "clordes:\n",
            "esery that speak will tell a deason,\n",
            "the fairets loves\n",
            "a seat or stawhore you'll brothe\n",
            "________1.0_________\n",
            "d;\n",
            "and if thou fail us, all our hope is henderiedy\n",
            "way thy fears!\n",
            "why, what thee away, henry, and wellxour.\n",
            "\n",
            "affock:\n",
            "'aintenis pepones the awnsow, and need!\n",
            "but mannarstive some brainy, have slanderbut,\n",
            "for plaise murseer by brother live.\n",
            "\n",
            "prince:\n",
            "death, dryou more comps a durg? shall shall.\n",
            "\n",
            "warwick:\n",
            "or soney, with his rance of thal i ry\n"
          ]
        }
      ]
    }
  ]
}