metadata
dataset_info:
features:
- name: audio
sequence: binary
- name: transcript
sequence: string
splits:
- name: train
num_bytes: 8641276
num_examples: 3
download_size: 8634102
dataset_size: 8641276
configs:
- config_name: default
data_files:
- split: train
path: data/train-*