metadata
dataset_info:
features:
- name: audio
dtype: audio
- name: text
dtype: string
- name: language
dtype: string
- name: source
dtype: string
- name: split
dtype: string
- name: speaker_id
dtype: string
- name: duration
dtype: float64
- name: meta
dtype: string
splits:
- name: train
num_bytes: 2719859106682.288
num_examples: 1667296
download_size: 1775584962641
dataset_size: 2719859106682.288
configs:
- config_name: default
data_files:
- split: train
path: data/train-*