TV Scripts Generation

Table Of Contents

• Introduction
• Project Overview
• Requirements
• Neccessary Files
• How to Run

Introduction

This repository contains all my work for the Udacity's Deep Learning Nanodegree Program.

The goal in this project was to perform sentiment analysis using a Recurrent Neural Network (RNN) to generate a TV script that resembles a training set.

To accomplish this, I applied data pre-processing techniques (lookup table and tokenize punctuation). Then, implemented an RNN with PyTorch using a Long Short-Term Memory (LSTM) network to avoid the long-term dependency problem and trained the neural network on the pre-processed data tracking the loss to avoid overfitting. Finally, I modified model hyperparameters (learning rate, batch size, layers in RNN) to achieve a better loss and faster converge.

Project Overview

Automated text generation is becoming increasingly common. From robot journalism to product descriptions, their use has become more and more ubiquitous. As demand continues to grow for individual online content, automated text generation is to become increasingly important.

In this project, you'll generate your own Seinfeld TV scripts using RNNs. You'll be using part of the Seinfeld dataset of scripts from 9 seasons. The Neural Network you'll build will generate a new ,"fake" TV script, based on patterns it recognizes in this training data.

Requirements

This project uses the following software and Python libraries:

• Python
• NumPy
• pandas
• scikit-learn (v0.17)
• Matplotlib

You will also need to have software installed to run and execute a Jupyter Notebook.

If you do not have Python installed yet, it is highly recommended that you install the Anaconda distribution of Python, which already has the above packages and more included.

Neccessary Files

This repository contains three files needed to solve the project.

1. dlnd_tv_script_generation.ipynb: This is the main file where I performed the work on the project.
2. dlnd_tv_script_generation.html: This is an HTML report of the Jupyter notebook.
3. data/Seinfeld_Scripts.txt: Seinfeld dataset of scripts from 9 seasons.
4. helper.py: Utilities functions to use in the project.
5. problem_unittests.py: Unit tests to verify correct functionality of the algorithms created.

How to Run

In the Terminal or Command Prompt, navigate to the folder on your machine where you've put the project files, and then use the command:

jupyter notebook dlnd_tv_script_generation.ipynb

to open up a browser window or tab to work with your notebook. Alternatively, you can use the command:

jupyter notebook

or

ipython notebook

and navigate to the notebook file (dlnd_tv_script_generation.ipynb) in the browser window that opens.