metadata
dataset_info:
features:
- name: id
dtype: string
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcription
dtype: string
splits:
- name: train
num_bytes: 17538341989.70426
num_examples: 343353
- name: test
num_bytes: 4450337596.076739
num_examples: 85839
download_size: 21300390861
dataset_size: 21988679585.781002
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*