metadata
dataset_info:
features:
- name: audio_filename
dtype: string
- name: audio
dtype: audio
- name: transcription
dtype: string
splits:
- name: train
num_bytes: 11662279531.75
num_examples: 70698
- name: test
num_bytes: 3339197359.6
num_examples: 20200
- name: validation
num_bytes: 1676259572.3
num_examples: 10100
download_size: 16194203297
dataset_size: 16677736463.65
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
- split: test
path: data/test-*
- split: validation
path: data/validation-*