metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: logits
sequence: float32
splits:
- name: train
num_bytes: 573986574.0302557
num_examples: 17904
download_size: 520117745
dataset_size: 573986574.0302557
configs:
- config_name: default
data_files:
- split: train
path: data/train-*