metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: text
dtype: string
- name: duration
dtype: float32
splits:
- name: train
num_bytes: 2641304240.175
num_examples: 7933
download_size: 2515289684
dataset_size: 2641304240.175
configs:
- config_name: default
data_files:
- split: train
path: data/train-*