metadata

dataset_info:
  features:
    - name: audio
      dtype:
        audio:
          sampling_rate: 16000
    - name: text
      dtype: string
  splits:
    - name: development
      num_bytes: 6030729
      num_examples: 59
    - name: test
      num_bytes: 8229224
      num_examples: 71
    - name: train
      num_bytes: 29128904
      num_examples: 300
  download_size: 43004020
  dataset_size: 43388857
license: cc-by-4.0
task_categories:
  - automatic-speech-recognition
language:
  - gos

Gronings transcribed speech

Demonstration dataset with Gronings transcribed speech based on the dataset released by San et al. (2021).

For more information see the corresponding ASRU 2021 paper.