Fine-tuned DeHateBERT-style classifier developed during the PrivHSD Challenge
for harmful or hate speech detection in the Glimo privacy-preserving pipeline.
Do not publish private challenge samples, raw admin uploads, or generated
outputs containing private source text in this repository.
The classifier can produce false positives and false negatives, especially for
dialectal language, reclaimed terms, counterspeech, quoted speech, contextual
ambiguity, and emerging coded language. Model outputs and restatements require
human/admin review before consequential action.
1from transformers import pipeline
2
3clf = pipeline("text-classification", model="batinium/glimo-dehatebert-hsd")
4print(clf("The comment uses abusive language toward a protected group."))
