---
language:
- bg
task_categories:
- text-classification
---

#### Warning: This dataset contains content that includes toxic, offensive, or otherwise inappropriate language. 

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.