metadata
dataset_info:
features:
- name: text
dtype: string
- name: language
dtype: string
- name: speaker
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 24000
splits:
- name: train
num_bytes: 704409548
num_examples: 2000
download_size: 638440845
dataset_size: 704409548
configs:
- config_name: default
data_files:
- split: train
path: data/train-*