Deep N-Grams

This is an exploration of Recurrent Neural Networks (RNN) using trax. We're going to predict the next set of characters in a sentence given the previous characters.

Since this is so long I'm going to break it up into separate posts.

• Loading the Data: Load the data and convert it to tensors
• Generating Data: Create a batch generator for the tensors
• Creating the Model: Create a Gated Recurrent Unit (GRU) model
• Training the Model: Train the model
• Evaluating the Model: Evaluate the model's perplexity
• Generating Sentences: Generate new sentences using the model

First up: - Loading the Data.