metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 44100
- name: transcription
dtype: string
- name: speaker_source
dtype: string
- name: gender
dtype: int64
splits:
- name: train
num_bytes: 25034429253.44
num_examples: 27704
download_size: 22800231862
dataset_size: 25034429253.44
configs:
- config_name: default
data_files:
- split: train
path: data/train-*