Utilities and scripts to transform plWordNet into machine‑learning‑ready artifacts: build graphs, create balanced embedding datasets from lexical relations, split into train/test, and prepare indices for downstream retrieval tasks.