metadata
dataset_info:
features:
- name: audio_filename
dtype: string
- name: audio
dtype: audio
- name: transcription
dtype: string
splits:
- name: train
num_bytes: 18553093667.842
num_examples: 37533
- name: test
num_bytes: 430887929
num_examples: 871
- name: validation
num_bytes: 420056211
num_examples: 826
download_size: 18905484278
dataset_size: 19404037807.842
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: validation
path: data/validation-*