Code and data for the paper "Contextual morphologically-guided tokenization for Latin encoder models" by Marisa Hudspeth, Patrick J. Burns, and Brendan O'Connor.