In this project, you will use PyTorch to build and train a text generation model using Long-Short Term Memory (LSTM) networks. Specifically, you’ll build a language model to generate the next token end-to-end by:
- using a real-world text dataset
- performing text tokenization and preprocessing
- building and training an LSTM network
- generating the next character-based token
We’ll be using a small portion of text from the ebook Frankenstein by Mary Shelley to train a relatively small language model. But in general, text generation models can assist us in tasks like answering questions, analyzing texts, translating text, and brainstorming ideas!