metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcripts
dtype: string
splits:
- name: train
num_bytes: 5309145.5
num_examples: 33
- name: test
num_bytes: 632889.1666666666
num_examples: 5
- name: valid
num_bytes: 877374.3333333334
num_examples: 4
download_size: 6716063
dataset_size: 6819409
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: valid
path: data/valid-*