TristanBehrens's picture
Upload folder using huggingface_hub
ab68132 verified
metadata
task_categories:
  - text-generation
language:
  - de

Heilbronn Podcast Transcripts Dataset

Logo

A German dataset of podcast transcripts from Heilbronn. Provided by Robert Mucha.

Contains:

Visit https://www.ki-salon.net/.

Statistics

The dataset was created from 78 transcripts. The statistics are based on the tokenize estimation strategy.

Statistics

Statistics (strategy: tokenize, window_length: 30, hop_length: 10)

  • Window length: 30
  • Hop length: 10
  • Min length: 665.00
  • Mean length: 3080.09
  • STD length: 1708.43
  • Max length: 12596.00
  • Lines: 1485