metadata
dataset_info:
features:
- name: audio
dtype: audio
- name: transcript
dtype: string
- name: transliteration
dtype: string
- name: translation
dtype: string
- name: language
dtype: string
- name: duration_ms
dtype: int64
splits:
- name: train
num_bytes: 5702051263.756
num_examples: 3727
download_size: 4842006173
dataset_size: 5702051263.756
configs:
- config_name: default
data_files:
- split: train
path: data/train-*