Code to run (i) Article / sentence classification (ii) Train a language model (iii) Summarization (this one is coming) 1. The runs directory is diagnostic information from the language model scripts 2. The output stores the results of both the language model and article classification