The dhivehi-english-pairs-with-metadata dataset contains 56,029 aligned sentence pairs between Dhivehi (written in Thaana script) and English, automatically generated using a gemma-3 language model. Each sentence pair includes additional metadata such as the opinion expressed in the sentence, a general sentiment classification (Positive, Negative, or Neutral), the intent behind the statement (e.g., Inform, Request, Accuse), and its… See the full description on the dataset page:
https://huggingface.co/datasets/alakxender/dhivehi-english-pairs-with-metadata.
