metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: input_features
sequence:
sequence: float32
- name: labels
sequence: int64
splits:
- name: train
num_bytes: 860995896
num_examples: 500
download_size: 231749503
dataset_size: 860995896
configs:
- config_name: default
data_files:
- split: train
path: data/train-*