metadata
dataset_info:
features:
- name: audio
dtype:
audio:
sampling_rate: 16000
- name: transcription
dtype: string
splits:
- name: train
num_bytes: 2942676359.46
num_examples: 6435
- name: test
num_bytes: 611245558.458
num_examples: 2797
- name: validation
num_bytes: 588074226.268
num_examples: 3162
download_size: 3825117564
dataset_size: 4141996144.1860003
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: validation
path: data/validation-*