metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 44100
- name: transcription
dtype: string
splits:
- name: train
num_bytes: 5344181847.948263
num_examples: 17952
- name: test
num_bytes: 593897214.0517371
num_examples: 1995
download_size: 4623275090
dataset_size: 5938079062
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*