This dataset is a French translation of the original work SummEval.
The translation was made using DeepL from English to French.
We used a LLM to rate the quality of translations, we verified random samples rated above 9/10 manually and corrected all those rated under 9/10. We also checked the correlation of ROUGE and BLEU scores between SummEval and SummEvalFr. For more details about the quality checks of this dataset, please refer to our paper.
We use this dataset… See the full description on the dataset page:
https://huggingface.co/datasets/lyon-nlp/summarization-summeval-fr-p2p.