{ "nbformat": 4, "nbformat_minor": 0, "metadata": { "accelerator": "GPU", "colab": { "name": "Course 3 - Week 4 - Lesson 2 - Notebook.ipynb", "provenance": [], "toc_visible": true }, "kernelspec": { "display_name": "Python 3", "name": "python3" } }, "cells": [ { "cell_type": "code", "metadata": { "id": "zX4Kg8DUTKWO" }, "source": [ "#@title Licensed under the Apache License, Version 2.0 (the \"License\");\n", "# you may not use this file except in compliance with the License.\n", "# You may obtain a copy of the License at\n", "#\n", "# https://www.apache.org/licenses/LICENSE-2.0\n", "#\n", "# Unless required by applicable law or agreed to in writing, software\n", "# distributed under the License is distributed on an \"AS IS\" BASIS,\n", "# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.\n", "# See the License for the specific language governing permissions and\n", "# limitations under the License." ], "execution_count": null, "outputs": [] }, { "cell_type": "markdown", "metadata": { "id": "view-in-github" }, "source": [ "\"Open" ] }, { "cell_type": "code", "metadata": { "id": "BOwsuGQQY9OL" }, "source": [ "import tensorflow as tf\n", "\n", "from tensorflow.keras.preprocessing.sequence import pad_sequences\n", "from tensorflow.keras.layers import Embedding, LSTM, Dense, Bidirectional\n", "from tensorflow.keras.preprocessing.text import Tokenizer\n", "from tensorflow.keras.models import Sequential\n", "from tensorflow.keras.optimizers import Adam\n", "import numpy as np " ], "execution_count": null, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "pylt5qZYsWPh" }, "source": [ "!wget --no-check-certificate \\\n", " https://storage.googleapis.com/laurencemoroney-blog.appspot.com/irish-lyrics-eof.txt \\\n", " -O /tmp/irish-lyrics-eof.txt" ], "execution_count": null, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "PRnDnCW-Z7qv" }, "source": [ "tokenizer = Tokenizer()\n", "\n", "data = open('/tmp/irish-lyrics-eof.txt').read()\n", "\n", "corpus = data.lower().split(\"\\n\")\n", "\n", "tokenizer.fit_on_texts(corpus)\n", "total_words = len(tokenizer.word_index) + 1\n", "\n", "print(tokenizer.word_index)\n", "print(total_words)\n" ], "execution_count": null, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "soPGVheskaQP" }, "source": [ "input_sequences = []\n", "for line in corpus:\n", "\ttoken_list = tokenizer.texts_to_sequences([line])[0]\n", "\tfor i in range(1, len(token_list)):\n", "\t\tn_gram_sequence = token_list[:i+1]\n", "\t\tinput_sequences.append(n_gram_sequence)\n", "\n", "# 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", "# create predictors and label\n", "xs, labels = input_sequences[:,:-1],input_sequences[:,-1]\n", "\n", "ys = tf.keras.utils.to_categorical(labels, num_classes=total_words)" ], "execution_count": null, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "pJtwVB2NbOAP" }, "source": [ "print(tokenizer.word_index['in'])\n", "print(tokenizer.word_index['the'])\n", "print(tokenizer.word_index['town'])\n", "print(tokenizer.word_index['of'])\n", "print(tokenizer.word_index['athy'])\n", "print(tokenizer.word_index['one'])\n", "print(tokenizer.word_index['jeremy'])\n", "print(tokenizer.word_index['lanigan'])" ], "execution_count": null, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "49Cv68JOakwv" }, "source": [ "print(xs[6])" ], "execution_count": null, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "iY-jwvfgbEF8" }, "source": [ "print(ys[6])" ], "execution_count": null, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "wtzlUMYadhKt" }, "source": [ "print(xs[5])\n", "print(ys[5])" ], "execution_count": null, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "H4myRpB1c4Gg" }, "source": [ "print(tokenizer.word_index)" ], "execution_count": null, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "w9vH8Y59ajYL" }, "source": [ "model = Sequential()\n", "model.add(Embedding(total_words, 100, input_length=max_sequence_len-1))\n", "model.add(Bidirectional(LSTM(150)))\n", "model.add(Dense(total_words, activation='softmax'))\n", "adam = Adam(lr=0.01)\n", "model.compile(loss='categorical_crossentropy', optimizer=adam, metrics=['accuracy'])\n", "#earlystop = EarlyStopping(monitor='val_loss', min_delta=0, patience=5, verbose=0, mode='auto')\n", "history = model.fit(xs, ys, epochs=100, verbose=1)\n", "#print model.summary()\n", "print(model)\n" ], "execution_count": null, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "3YXGelKThoTT" }, "source": [ "import matplotlib.pyplot as plt\n", "\n", "\n", "def plot_graphs(history, string):\n", " plt.plot(history.history[string])\n", " plt.xlabel(\"Epochs\")\n", " plt.ylabel(string)\n", " plt.show()" ], "execution_count": null, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "poeprYK8h-c7" }, "source": [ "plot_graphs(history, 'accuracy')\n" ], "execution_count": null, "outputs": [] }, { "cell_type": "code", "metadata": { "id": "6Vc6PHgxa6Hm" }, "source": [ "seed_text = \"I've got a bad feeling about this\"\n", "next_words = 100\n", " \n", "for _ in range(next_words):\n", "\ttoken_list = tokenizer.texts_to_sequences([seed_text])[0]\n", "\ttoken_list = pad_sequences([token_list], maxlen=max_sequence_len-1, padding='pre')\n", "\tpredicted = model.predict_classes(token_list, verbose=0)\n", "\toutput_word = \"\"\n", "\tfor word, index in tokenizer.word_index.items():\n", "\t\tif index == predicted:\n", "\t\t\toutput_word = word\n", "\t\t\tbreak\n", "\tseed_text += \" \" + output_word\n", "print(seed_text)" ], "execution_count": null, "outputs": [] } ] }