Skip to content

Dataset · linagora

French meeting transcripts with audio

linagora/SUMM-RE

A collection of French conversational meeting transcripts aligned with audio signals.

What's inside

The corpus contains audio recordings and text transcripts of meetings in French with a total size of under 1,000 items, created as part of the SUMM-RE project. The text and audio are precisely aligned with each other.

What you can do with it

  • Train French ASR systems
  • Voice activity detection tasks
Tasks
Automatic speech recognition, Voice activity detection
Languages
French
Size
up to 1K rows
Downloads this month
2.8K
Likes
7

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("linagora/SUMM-RE")