MYBully
Collection
Collection of PLM model fine-tuned on MYBully Dataset • 9 items • Updated
MYBully-HateBERT (Manual + HITL)
Model Overview
This model is MYBully-HateBERT, fine-tuned on manual + HITL annotated data from the MYBully dataset.
It captures more nuanced hate/offensive patterns than the manual baseline.
Intended Use
- Detecting hate speech and abusive language in Bahasa Malaysia tweets.
Training Data
- Dataset: MYBully (Bahasa Malaysia tweets).
- Annotation: Manual + HITL
Model Details
- Base model: roberta-base-bahasa-cased
- Fine-tuning: Binary classification head
- Labels: Hate Speech (1), Non-Hate Speech (0)
Performance
| Metric | Value |
|---|---|
| Accuracy | 0.80 |
| Precision | 0.78 |
| Recall | 0.80 |
| Macro F1 | 0.79 |
- Downloads last month
- 18
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support
Model tree for mohanrj/mybully-hatebert-manual-hitl
Base model
mesolitica/roberta-base-bahasa-cased