metadata
dataset_info:
features:
- name: audio
dtype: audio
- name: speech_id
dtype: string
- name: speaker_id
dtype: string
- name: transcript
dtype: string
splits:
- name: train
num_bytes: 187936686
num_examples: 1000
download_size: 180001519
dataset_size: 187936686
configs:
- config_name: default
data_files:
- split: train
path: data/train-*