metadata
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
dataset_info:
features:
- name: transcription
dtype: string
- name: speaker_id
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 16000
splits:
- name: train
num_bytes: 34531110279
num_examples: 17585
download_size: 28972670658
dataset_size: 34531110279