metadata
dataset_info:
features:
- name: audio
struct:
- name: array
sequence: float64
- name: path
dtype: 'null'
- name: sampling_rate
dtype: int64
- name: text
dtype: string
- name: start_time
dtype: float64
- name: end_time
dtype: float64
splits:
- name: train
num_bytes: 4630387550
num_examples: 9726
download_size: 1156488335
dataset_size: 4630387550
configs:
- config_name: default
data_files:
- split: train
path: data/train-*