metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcripts
dtype: string
splits:
- name: train
num_bytes: 176730120
num_examples: 1600
- name: test
num_bytes: 22174948.2
num_examples: 200
- name: valid
num_bytes: 21947524.2
num_examples: 200
download_size: 190687583
dataset_size: 220852592.39999998
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: valid
path: data/valid-*