metadata
dataset_info:
features:
- name: text
dtype: string
- name: audio_file
dtype: string
- name: audio_array16k
sequence: float32
splits:
- name: train
num_bytes: 18767276319
num_examples: 15023
download_size: 18769901061
dataset_size: 18767276319
configs:
- config_name: default
data_files:
- split: train
path: data/train-*