metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: text
dtype: string
- name: start_time
dtype: float64
- name: end_time
dtype: float64
- name: original_audio
dtype: string
- name: source_json
dtype: string
splits:
- name: train
num_bytes: 18183253200.906
num_examples: 83691
download_size: 20075498567
dataset_size: 18183253200.906
configs:
- config_name: default
data_files:
- split: train
path: data/train-*