OpenLLM-France/Luciole-Audio-Evaluation-Dataset

Tasks

• automatic-speech-recognition
• translation
• question-answering
• audio-classification

Citation text from the card

This dataset builds on the work of many others. Several sources are under **attribution
(CC-BY*)** licenses and **require credit**. Please attribute every source dataset you use;
the attribution-required ones are highlighted here:

The following datasets carry a **CC-BY / attribution** license on their text and/or audio — you **must credit** them (and cite their papers) when you use this dataset:
- **AMI Meeting Corpus** (CC-BY-4.0) — [source](https://groups.inf.ed.ac.uk/ami/corpus/)
- **FLEURS** (CC-BY-4.0) — [source](https://huggingface.co/datasets/google/fleurs)
- **MTJ-Jamendo (via AIR-Bench)** (CC-BY-NC-SA-4.0 (metadata) (text), Mixed per-track CC (Jamendo) — authorization required (audio)) — [source](https://github.com/MTG/mtg-jamendo-dataset)
- **MusicCaps** (CC-BY-SA-4.0 (text), Copyrighted (YouTube / music industry rights) (audio)) — [source](https://www.kaggle.com/datasets/googleai/musiccaps)
- **SLUE-SQA-5 (SLUE phase-2)** (Apache-2.0 / CC-BY-SA-4.0) — [source](https://huggingface.co/datasets/asapp/slue-phase-2)
- **SpokenWOZ (via AIR-Bench)** (CC-BY-NC-4.0) — [source](https://spokenwoz.github.io/)
- **SUMM-RE** (CC-BY-SA-4.0) — [source](https://huggingface.co/datasets/linagora/SUMM-RE)
- **Vigogne (French Alpaca)** (CC-BY-NC-4.0 (text), Synthesized by OpenLLM-France (Chatterbox TTS) (audio)) — [source](https://github.com/bofenghuang/vigogne)

These also require keeping their text license/attribution notice (Apache-2.0 / MIT / CDLA):
- **aya_collection** (Apache-2.

All datasets marked Use restrictions | JSON