Dataset · Paul
German hate speech detection test suite
Paul/hatecheck-german
A suite of functional tests for evaluating hate speech detection models in German.
What's inside
Includes over 25 functional tests for the German language, corresponding to various types of hate speech and complex non-hate content. Designed for in-depth diagnostics of text classifier performance.
What you can do with it
- Test and benchmark text classification models
- Evaluate the reliability of safety filters and hate speech detection systems
- Tasks
- Text classification
- Languages
- German
- Size
- 1K–10K rows
- Downloads this month
- 283
- Likes
- 9
Models trained on it
How to load it
from datasets import load_dataset
ds = load_dataset("Paul/hatecheck-german")