News classification model in Reliable or Unreliable. This model has been obtained by fine-tuning the RoBERTa language model in Catalan (https://huggingface.co/PlanTL-GOB-ES/roberta-base-ca).
We train this model with an automatic translation of RUN_AS dataset (https://github.com/rsepulveda911112/NewsReliabilityAnnotation/tree/main).
Results:
| F_1-macro | F_1 Unreliable | F_1 Reliable |
|---|---|---|
| 0.791 | 0.833 | 0.75 |
- Downloads last month
- 7