This model is trained (base: BERTweet) to detect hate speech, specifically for the purpose of a research project conducted by the authors.

The model was tested against HateCheck:
[HateCheck: Functional Tests for Hate Speech Detection Models](https://aclanthology.org/2021.acl-long.4/) (Röttger et al., ACL-IJCNLP 2021)

Some details are still being developed. Use at your own risk.



---
license: cc-by-4.0
datasets:
- manueltonneau/english-hate-speech-superset
base_model:
- vinai/bertweet-base
---