metadata
dataset_info:
features:
- name: compressed_audio
sequence:
sequence: float32
- name: segment_size
dtype: int32
- name: sample_rate
dtype: int32
splits:
- name: train
num_bytes: 8168952
num_examples: 319
download_size: 8698648
dataset_size: 8168952
configs:
- config_name: default
data_files:
- split: train
path: data/train-*