Compar:IA Datasets
Ministere de la Culture (France)Compar:IA Datasets is human-preference data released from the French government's Compar:IA chatbot arena and published by the Ministere de la Culture. Its comparia-votes dataset holds conversation-level pairwise preference votes on French-language conversations.
Openness
3 high confidence- license
- Etalab-2.0(permissive French-gov open license)
- gated
- terms-acceptance+contact-info(HF contact-info agreement required)
- dataset_card
- present
Openly licensed and documented, but access is gated behind an agreement, which caps it at gated. The gate is a click-through agreement that also collects contact details, so it asks for terms acceptance and contact information together.
- https://huggingface.co/datasets/ministere-culture/comparia-votes recorded 2026-09-29
Dataset page tags the license as "license:etalab-2.0", shows the gate notice "You need to agree to share your contact information to access this dataset", and renders a dataset card beginning "conversation-level preference datasets in French".
Adoption
1 high confidenceHugging Face downloads are the only usage signal for ministere-culture/comparia-votes, the collection's sole declared artifact.
- https://huggingface.co/api/datasets/ministere-culture/comparia-votes recorded 2026-08-13
31 downloads in the trailing 30 days for ministere-culture/comparia-votes
Capability
not assessedA dataset is not 'capable', so capability is left unscored. The repository publishes a static corpus of pairwise preference annotations, with no performance or feature claim to score.
- https://huggingface.co/datasets/ministere-culture/comparia-votes recorded 2026-09-29
Dataset page describes a static corpus of pairwise preference votes from the Compar:IA arena as "conversation-level preference datasets in French"; searched the body for leaderboard, accuracy, SOTA and throughput: none.
Verified 2026-09-27