The ASSIN (Avaliação de Similaridade Semântica e INferência textual) corpus is a corpus annotated with pairs of sentences written in
Portuguese that is suitable for the exploration of textual entailment and paraphrasing classifiers. The corpus contains pairs of sentences
extracted from news articles written in European Portuguese (EP) and Brazilian Portuguese (BP), obtained from Google News Portugal
and Brazil, respectively. To… See the full description on the dataset page: https://huggingface.co/datasets/nilc-nlp/assin.