Datasets:
The toxic-data-bg dataset consists of 4,384 manually annotated sentences across four categories: toxic language, medical terminology, non-toxic language, and terms related to minority communities.