metadata
dataset_info:
features:
- name: audio
sequence: float32
- name: speaker_id
dtype: string
- name: transcription
dtype: string
- name: filename
dtype: string
splits:
- name: train
num_bytes: 3879044515
num_examples: 13998
- name: dev
num_bytes: 279984004
num_examples: 1120
download_size: 2072416720
dataset_size: 4159028519
task_categories:
- automatic-speech-recognition
language:
- wo