metadata
dataset_info:
features:
- name: arabic_text
dtype: string
- name: audio
sequence: float32
splits:
- name: train
num_bytes: 3773822740
num_examples: 12619
download_size: 2595977608
dataset_size: 3773822740
configs:
- config_name: default
data_files:
- split: train
path: data/train-*