README.md

metadata

language:
  - fr
license: unknown
size_categories:
  - 100K<n<1M
task_categories:
  - automatic-speech-recognition
pretty_name: Transcribed French Podcast
tags:
  - Podcast
  - Audio
  - Transcribed
configs:
  - config_name: default
    data_files:
      - split: train
        path: data/train-*
dataset_info:
  features:
    - name: programme_id
      dtype: string
    - name: programme_entry_date
      dtype: string
    - name: programme_rss_link
      dtype: string
    - name: podcast_title
      dtype: string
    - name: podcast_date
      dtype: string
    - name: podcast_duration
      dtype: string
    - name: audio_podcast_link
      dtype: string
  splits:
    - name: train
      num_bytes: 96627005
      num_examples: 281759
  download_size: 28777088
  dataset_size: 96627005

French Transcribed Podcast

Dataset Summary

Dataset of 280,000 mp3 links to French podcasts. Transcription using whisper is underway. However, due to the large number of podcasts, it will not be possible to transcribe all of them. We are therefore counting on the help of the community to help us finish this colossal task.

The total duration of the podcasts is estimated at approximately 2958 days (4259523 minutes). However, this value is only an indication, as some links no longer seem to work and not all podcasts have the indicated duration.

N.B. The podcast links are available on the French government's data gouv website.