This Python library implements a set of probabilistic topic models that allow integration of arbitrary document-level metainformation into the generative process for the data. The model currently implemented is the Structural Topic Model (STM) of Roberts, Stewart, and Tingley. A nonparametric variant is forthcoming. While the software can be run in serial on a local machine, the software's primary goal is use for parallel computation with Apache Spark.